diff --git "a/sft/hyperrouter/trainer_state.json" "b/sft/hyperrouter/trainer_state.json" new file mode 100644--- /dev/null +++ "b/sft/hyperrouter/trainer_state.json" @@ -0,0 +1,74877 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999398785546805, + "eval_steps": 500, + "global_step": 8316, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00012024289063909097, + "flos": 18263040018000.0, + "grad_norm": 13.817966121833583, + "learning_rate": 0.0, + "loss": 1.4267, + "num_input_tokens_seen": 20095, + "step": 1 + }, + { + "epoch": 0.00024048578127818193, + "flos": 22506996816480.0, + "grad_norm": 18.108283812304055, + "learning_rate": 5.021476677069823e-07, + "loss": 1.3813, + "num_input_tokens_seen": 38475, + "step": 2 + }, + { + "epoch": 0.0003607286719172729, + "flos": 13893881084160.0, + "grad_norm": 17.568114336986252, + "learning_rate": 7.958852231401551e-07, + "loss": 1.2402, + "num_input_tokens_seen": 56760, + "step": 3 + }, + { + "epoch": 0.00048097156255636386, + "flos": 14108309722320.0, + "grad_norm": 11.07942172748333, + "learning_rate": 1.0042953354139647e-06, + "loss": 1.3203, + "num_input_tokens_seen": 75345, + "step": 4 + }, + { + "epoch": 0.0006012144531954548, + "flos": 10167680360400.0, + "grad_norm": 14.930126461201223, + "learning_rate": 1.1659507774310057e-06, + "loss": 1.3867, + "num_input_tokens_seen": 92950, + "step": 5 + }, + { + "epoch": 0.0007214573438345458, + "flos": 16944982125000.0, + "grad_norm": 16.141306594091233, + "learning_rate": 1.2980328908471373e-06, + "loss": 1.2601, + "num_input_tokens_seen": 112915, + "step": 6 + }, + { + "epoch": 0.0008417002344736367, + "flos": 49893916281720.0, + "grad_norm": 3.0801871116577098, + "learning_rate": 1.4097067265369432e-06, + "loss": 0.7714, + "num_input_tokens_seen": 177630, + "step": 7 + }, + { + "epoch": 0.0009619431251127277, + "flos": 15561860892120.0, + "grad_norm": 23.371684935799586, + "learning_rate": 1.506443003120947e-06, + "loss": 1.2059, + "num_input_tokens_seen": 192850, + "step": 8 + }, + { + "epoch": 0.0010821860157518186, + "flos": 17079926077680.0, + "grad_norm": 10.773063935351715, + "learning_rate": 1.5917704462803102e-06, + "loss": 1.3108, + "num_input_tokens_seen": 209115, + "step": 9 + }, + { + "epoch": 0.0012024289063909096, + "flos": 12400537836000.0, + "grad_norm": 11.279973092554343, + "learning_rate": 1.6680984451379884e-06, + "loss": 1.3537, + "num_input_tokens_seen": 221905, + "step": 10 + }, + { + "epoch": 0.0013226717970300007, + "flos": 15567812964600.0, + "grad_norm": 8.092702201271033, + "learning_rate": 1.7371455188905097e-06, + "loss": 1.138, + "num_input_tokens_seen": 241555, + "step": 11 + }, + { + "epoch": 0.0014429146876690916, + "flos": 19971392888400.0, + "grad_norm": 33.78233043484974, + "learning_rate": 1.8001805585541196e-06, + "loss": 1.1182, + "num_input_tokens_seen": 262405, + "step": 12 + }, + { + "epoch": 0.0015631575783081825, + "flos": 13920407639280.0, + "grad_norm": 6.876950906169764, + "learning_rate": 1.8581671739548328e-06, + "loss": 1.1964, + "num_input_tokens_seen": 279860, + "step": 13 + }, + { + "epoch": 0.0016834004689472734, + "flos": 35439300493320.0, + "grad_norm": 5.773836898926085, + "learning_rate": 1.9118543942439254e-06, + "loss": 1.0317, + "num_input_tokens_seen": 301765, + "step": 14 + }, + { + "epoch": 0.0018036433595863645, + "flos": 25693295089680.0, + "grad_norm": 5.786681227306793, + "learning_rate": 1.961836000571161e-06, + "loss": 1.0432, + "num_input_tokens_seen": 323140, + "step": 15 + }, + { + "epoch": 0.0019238862502254555, + "flos": 44049128081160.0, + "grad_norm": 2.5706490061589915, + "learning_rate": 2.0085906708279293e-06, + "loss": 0.6258, + "num_input_tokens_seen": 378230, + "step": 16 + }, + { + "epoch": 0.0020441291408645466, + "flos": 15217427255040.0, + "grad_norm": 5.213731740052087, + "learning_rate": 2.0525099325728135e-06, + "loss": 1.0917, + "num_input_tokens_seen": 396130, + "step": 17 + }, + { + "epoch": 0.0021643720315036373, + "flos": 47075380544760.0, + "grad_norm": 2.249954541065026, + "learning_rate": 2.0939181139872922e-06, + "loss": 0.7098, + "num_input_tokens_seen": 457565, + "step": 18 + }, + { + "epoch": 0.0022846149221427284, + "flos": 22966116523200.0, + "grad_norm": 5.084587046791643, + "learning_rate": 2.1330868934640175e-06, + "loss": 0.9383, + "num_input_tokens_seen": 477960, + "step": 19 + }, + { + "epoch": 0.002404857812781819, + "flos": 37621768786680.0, + "grad_norm": 2.0895544119215606, + "learning_rate": 2.170246112844971e-06, + "loss": 0.7579, + "num_input_tokens_seen": 532020, + "step": 20 + }, + { + "epoch": 0.0025251007034209102, + "flos": 11517594786600.0, + "grad_norm": 7.5789698808179535, + "learning_rate": 2.2055919496770983e-06, + "loss": 0.966, + "num_input_tokens_seen": 549880, + "step": 21 + }, + { + "epoch": 0.0026453435940600014, + "flos": 27825966757320.0, + "grad_norm": 5.763089415884376, + "learning_rate": 2.2392931865974923e-06, + "loss": 0.8608, + "num_input_tokens_seen": 572290, + "step": 22 + }, + { + "epoch": 0.002765586484699092, + "flos": 15459617186760.0, + "grad_norm": 7.636480974559204, + "learning_rate": 2.271496085962064e-06, + "loss": 0.9816, + "num_input_tokens_seen": 589705, + "step": 23 + }, + { + "epoch": 0.002885829375338183, + "flos": 15108344998320.0, + "grad_norm": 5.005354874841114, + "learning_rate": 2.3023282262611022e-06, + "loss": 0.9934, + "num_input_tokens_seen": 608200, + "step": 24 + }, + { + "epoch": 0.003006072265977274, + "flos": 25585099311840.0, + "grad_norm": 5.104600641394964, + "learning_rate": 2.3319015548620114e-06, + "loss": 0.8753, + "num_input_tokens_seen": 629060, + "step": 25 + }, + { + "epoch": 0.003126315156616365, + "flos": 17886550437600.0, + "grad_norm": 3.186201841006131, + "learning_rate": 2.3603148416618152e-06, + "loss": 0.8974, + "num_input_tokens_seen": 648295, + "step": 26 + }, + { + "epoch": 0.003246558047255456, + "flos": 17295146214840.0, + "grad_norm": 11.13022653628503, + "learning_rate": 2.3876556694204647e-06, + "loss": 0.9761, + "num_input_tokens_seen": 668170, + "step": 27 + }, + { + "epoch": 0.003366800937894547, + "flos": 13001123447160.0, + "grad_norm": 5.1525362454564245, + "learning_rate": 2.414002061950908e-06, + "loss": 0.8651, + "num_input_tokens_seen": 686765, + "step": 28 + }, + { + "epoch": 0.003487043828533638, + "flos": 17890159673040.0, + "grad_norm": 4.789056326782828, + "learning_rate": 2.4394238264681557e-06, + "loss": 0.9775, + "num_input_tokens_seen": 706220, + "step": 29 + }, + { + "epoch": 0.003607286719172729, + "flos": 19158689816160.0, + "grad_norm": 3.0290220272438297, + "learning_rate": 2.4639836682781433e-06, + "loss": 0.967, + "num_input_tokens_seen": 726070, + "step": 30 + }, + { + "epoch": 0.00372752960981182, + "flos": 15053170670760.0, + "grad_norm": 3.383246343336847, + "learning_rate": 2.487738122623307e-06, + "loss": 0.9538, + "num_input_tokens_seen": 744700, + "step": 31 + }, + { + "epoch": 0.003847772500450911, + "flos": 16593899896320.0, + "grad_norm": 5.524775276964839, + "learning_rate": 2.510738338534912e-06, + "loss": 0.9442, + "num_input_tokens_seen": 763145, + "step": 32 + }, + { + "epoch": 0.003968015391090002, + "flos": 13110522303480.0, + "grad_norm": 5.514101331297901, + "learning_rate": 2.5330307420306648e-06, + "loss": 1.0015, + "num_input_tokens_seen": 779715, + "step": 33 + }, + { + "epoch": 0.004088258281729093, + "flos": 20019443724960.0, + "grad_norm": 5.9010215909324035, + "learning_rate": 2.554657600279796e-06, + "loss": 0.8433, + "num_input_tokens_seen": 800710, + "step": 34 + }, + { + "epoch": 0.004208501172368184, + "flos": 17026113128400.0, + "grad_norm": 4.295900546466009, + "learning_rate": 2.5756575039679493e-06, + "loss": 1.0193, + "num_input_tokens_seen": 819980, + "step": 35 + }, + { + "epoch": 0.0043287440630072746, + "flos": 12625414260960.0, + "grad_norm": 7.186673058747025, + "learning_rate": 2.5960657816942747e-06, + "loss": 0.9287, + "num_input_tokens_seen": 838615, + "step": 36 + }, + { + "epoch": 0.004448986953646365, + "flos": 39108748382880.0, + "grad_norm": 1.061511477461804, + "learning_rate": 2.6159148575788668e-06, + "loss": 0.6319, + "num_input_tokens_seen": 896370, + "step": 37 + }, + { + "epoch": 0.004569229844285457, + "flos": 9762816842400.0, + "grad_norm": 5.4406360921905526, + "learning_rate": 2.635234561171e-06, + "loss": 0.9627, + "num_input_tokens_seen": 914485, + "step": 38 + }, + { + "epoch": 0.0046894727349245475, + "flos": 11808373774200.0, + "grad_norm": 5.571640391392913, + "learning_rate": 2.6540523970949877e-06, + "loss": 0.9251, + "num_input_tokens_seen": 929050, + "step": 39 + }, + { + "epoch": 0.004809715625563638, + "flos": 17511189510960.0, + "grad_norm": 4.942158686227952, + "learning_rate": 2.6723937805519533e-06, + "loss": 0.9125, + "num_input_tokens_seen": 946270, + "step": 40 + }, + { + "epoch": 0.00492995851620273, + "flos": 15185993406120.0, + "grad_norm": 4.11023288373305, + "learning_rate": 2.690282243737839e-06, + "loss": 0.9144, + "num_input_tokens_seen": 964925, + "step": 41 + }, + { + "epoch": 0.0050502014068418205, + "flos": 14865015308040.0, + "grad_norm": 4.578626585684053, + "learning_rate": 2.7077396173840807e-06, + "loss": 0.9691, + "num_input_tokens_seen": 982930, + "step": 42 + }, + { + "epoch": 0.005170444297480911, + "flos": 19051127237520.0, + "grad_norm": 8.79427970153042, + "learning_rate": 2.7247861909342594e-06, + "loss": 0.898, + "num_input_tokens_seen": 1003575, + "step": 43 + }, + { + "epoch": 0.005290687188120003, + "flos": 15319765940280.0, + "grad_norm": 4.446974619582612, + "learning_rate": 2.7414408543044743e-06, + "loss": 0.8088, + "num_input_tokens_seen": 1018935, + "step": 44 + }, + { + "epoch": 0.005410930078759093, + "flos": 11544216321600.0, + "grad_norm": 9.348319374854356, + "learning_rate": 2.7577212237113157e-06, + "loss": 0.7757, + "num_input_tokens_seen": 1035695, + "step": 45 + }, + { + "epoch": 0.005531172969398184, + "flos": 15432172492800.0, + "grad_norm": 4.741288277400874, + "learning_rate": 2.7736437536690466e-06, + "loss": 1.0219, + "num_input_tokens_seen": 1055045, + "step": 46 + }, + { + "epoch": 0.005651415860037276, + "flos": 15239014856400.0, + "grad_norm": 4.133535166909097, + "learning_rate": 2.789223836941131e-06, + "loss": 1.0517, + "num_input_tokens_seen": 1074900, + "step": 47 + }, + { + "epoch": 0.005771658750676366, + "flos": 9626448191520.0, + "grad_norm": 5.947806959148138, + "learning_rate": 2.8044758939680847e-06, + "loss": 1.0403, + "num_input_tokens_seen": 1090690, + "step": 48 + }, + { + "epoch": 0.005891901641315457, + "flos": 17889526473840.0, + "grad_norm": 11.095980739629047, + "learning_rate": 2.8194134530738863e-06, + "loss": 0.9874, + "num_input_tokens_seen": 1109180, + "step": 49 + }, + { + "epoch": 0.006012144531954548, + "flos": 16891137515760.0, + "grad_norm": 8.98428498596212, + "learning_rate": 2.834049222568994e-06, + "loss": 0.8812, + "num_input_tokens_seen": 1126250, + "step": 50 + }, + { + "epoch": 0.006132387422593639, + "flos": 16485989058120.0, + "grad_norm": 12.226682307719713, + "learning_rate": 2.848395155712969e-06, + "loss": 0.909, + "num_input_tokens_seen": 1146190, + "step": 51 + }, + { + "epoch": 0.00625263031323273, + "flos": 20258815920240.0, + "grad_norm": 8.391225100746984, + "learning_rate": 2.8624625093687977e-06, + "loss": 0.9724, + "num_input_tokens_seen": 1163045, + "step": 52 + }, + { + "epoch": 0.006372873203871821, + "flos": 16916366012520.0, + "grad_norm": 8.988484539068976, + "learning_rate": 2.876261897070029e-06, + "loss": 0.8662, + "num_input_tokens_seen": 1182895, + "step": 53 + }, + { + "epoch": 0.006493116094510912, + "flos": 16728020690040.0, + "grad_norm": 4.492974413074048, + "learning_rate": 2.889803337127447e-06, + "loss": 0.9103, + "num_input_tokens_seen": 1201215, + "step": 54 + }, + { + "epoch": 0.006613358985150003, + "flos": 16887180020760.0, + "grad_norm": 7.953218290909745, + "learning_rate": 2.903096296321516e-06, + "loss": 0.8334, + "num_input_tokens_seen": 1219080, + "step": 55 + }, + { + "epoch": 0.006733601875789094, + "flos": 19452191560320.0, + "grad_norm": 7.936220801896522, + "learning_rate": 2.9161497296578907e-06, + "loss": 0.9143, + "num_input_tokens_seen": 1238190, + "step": 56 + }, + { + "epoch": 0.006853844766428185, + "flos": 11301329870760.0, + "grad_norm": 4.395616402478639, + "learning_rate": 2.928972116604173e-06, + "loss": 0.8485, + "num_input_tokens_seen": 1254185, + "step": 57 + }, + { + "epoch": 0.006974087657067276, + "flos": 17755215720360.0, + "grad_norm": 8.873661853757119, + "learning_rate": 2.9415714941751377e-06, + "loss": 0.9943, + "num_input_tokens_seen": 1275125, + "step": 58 + }, + { + "epoch": 0.007094330547706367, + "flos": 18886079154240.0, + "grad_norm": 4.186188221905894, + "learning_rate": 2.9539554871897396e-06, + "loss": 0.9224, + "num_input_tokens_seen": 1295015, + "step": 59 + }, + { + "epoch": 0.007214573438345458, + "flos": 15589875465360.0, + "grad_norm": 3.373425575648723, + "learning_rate": 2.9661313359851253e-06, + "loss": 0.958, + "num_input_tokens_seen": 1312620, + "step": 60 + }, + { + "epoch": 0.007334816328984549, + "flos": 18268338336120.0, + "grad_norm": 3.5857938849917925, + "learning_rate": 2.978105921839922e-06, + "loss": 0.9307, + "num_input_tokens_seen": 1332885, + "step": 61 + }, + { + "epoch": 0.00745505921962364, + "flos": 13511586626280.0, + "grad_norm": 5.812450449742594, + "learning_rate": 2.9898857903302893e-06, + "loss": 0.7234, + "num_input_tokens_seen": 1351555, + "step": 62 + }, + { + "epoch": 0.007575302110262731, + "flos": 13487719507800.0, + "grad_norm": 5.276221595801576, + "learning_rate": 3.001477172817253e-06, + "loss": 0.872, + "num_input_tokens_seen": 1369165, + "step": 63 + }, + { + "epoch": 0.007695545000901822, + "flos": 18294991531080.0, + "grad_norm": 6.243728848441557, + "learning_rate": 3.012886006241894e-06, + "loss": 0.9586, + "num_input_tokens_seen": 1388270, + "step": 64 + }, + { + "epoch": 0.007815787891540913, + "flos": 15594466159560.0, + "grad_norm": 3.660141919412314, + "learning_rate": 3.0241179513858383e-06, + "loss": 0.8854, + "num_input_tokens_seen": 1407300, + "step": 65 + }, + { + "epoch": 0.007936030782180003, + "flos": 15780088725480.0, + "grad_norm": 3.7789390374843888, + "learning_rate": 3.035178409737647e-06, + "loss": 0.8737, + "num_input_tokens_seen": 1424470, + "step": 66 + }, + { + "epoch": 0.008056273672819095, + "flos": 14703924719760.0, + "grad_norm": 5.94667901714964, + "learning_rate": 3.046072539090907e-06, + "loss": 0.8728, + "num_input_tokens_seen": 1442155, + "step": 67 + }, + { + "epoch": 0.008176516563458186, + "flos": 13380948428160.0, + "grad_norm": 5.134053676810564, + "learning_rate": 3.056805267986779e-06, + "loss": 1.0399, + "num_input_tokens_seen": 1459345, + "step": 68 + }, + { + "epoch": 0.008296759454097276, + "flos": 15999994536720.0, + "grad_norm": 5.50090938332737, + "learning_rate": 3.0673813091022194e-06, + "loss": 0.9399, + "num_input_tokens_seen": 1478285, + "step": 69 + }, + { + "epoch": 0.008417002344736368, + "flos": 46746871867560.0, + "grad_norm": 1.2141129236538757, + "learning_rate": 3.0778051716749317e-06, + "loss": 0.6547, + "num_input_tokens_seen": 1541550, + "step": 70 + }, + { + "epoch": 0.008537245235375458, + "flos": 16809879872520.0, + "grad_norm": 4.141205829692235, + "learning_rate": 3.0880811730470094e-06, + "loss": 0.8977, + "num_input_tokens_seen": 1560725, + "step": 71 + }, + { + "epoch": 0.008657488126014549, + "flos": 45692707042680.0, + "grad_norm": 0.9566476792822703, + "learning_rate": 3.098213449401257e-06, + "loss": 0.6079, + "num_input_tokens_seen": 1627375, + "step": 72 + }, + { + "epoch": 0.00877773101665364, + "flos": 22234354494600.0, + "grad_norm": 4.712959222654002, + "learning_rate": 3.1082059657570015e-06, + "loss": 0.9707, + "num_input_tokens_seen": 1646330, + "step": 73 + }, + { + "epoch": 0.00889797390729273, + "flos": 17215186629960.0, + "grad_norm": 10.965363504113474, + "learning_rate": 3.1180625252858496e-06, + "loss": 0.9408, + "num_input_tokens_seen": 1664480, + "step": 74 + }, + { + "epoch": 0.009018216797931822, + "flos": 16890820916160.0, + "grad_norm": 3.9283860100920722, + "learning_rate": 3.1277867780021663e-06, + "loss": 0.7866, + "num_input_tokens_seen": 1680835, + "step": 75 + }, + { + "epoch": 0.009138459688570914, + "flos": 11593375256760.0, + "grad_norm": 2.8915168005138425, + "learning_rate": 3.1373822288779824e-06, + "loss": 0.9457, + "num_input_tokens_seen": 1697135, + "step": 76 + }, + { + "epoch": 0.009258702579210003, + "flos": 19807231284000.0, + "grad_norm": 3.3892857591915284, + "learning_rate": 3.1468522454274533e-06, + "loss": 0.7749, + "num_input_tokens_seen": 1717210, + "step": 77 + }, + { + "epoch": 0.009378945469849095, + "flos": 19722807644760.0, + "grad_norm": 3.1498169255416593, + "learning_rate": 3.15620006480197e-06, + "loss": 0.9012, + "num_input_tokens_seen": 1736200, + "step": 78 + }, + { + "epoch": 0.009499188360488187, + "flos": 26228480206200.0, + "grad_norm": 5.08425711487492, + "learning_rate": 3.1654288004333087e-06, + "loss": 0.7407, + "num_input_tokens_seen": 1754585, + "step": 79 + }, + { + "epoch": 0.009619431251127276, + "flos": 15726434076000.0, + "grad_norm": 3.9785422322840036, + "learning_rate": 3.1745414482589353e-06, + "loss": 0.7381, + "num_input_tokens_seen": 1773515, + "step": 80 + }, + { + "epoch": 0.009739674141766368, + "flos": 12705975385080.0, + "grad_norm": 3.1308173392764487, + "learning_rate": 3.1835408925606204e-06, + "loss": 0.8432, + "num_input_tokens_seen": 1791375, + "step": 81 + }, + { + "epoch": 0.00985991703240546, + "flos": 20185061687520.0, + "grad_norm": 5.157053566942035, + "learning_rate": 3.1924299114448214e-06, + "loss": 0.8754, + "num_input_tokens_seen": 1811575, + "step": 82 + }, + { + "epoch": 0.00998015992304455, + "flos": 10032008228640.0, + "grad_norm": 4.003129450179879, + "learning_rate": 3.2012111819909055e-06, + "loss": 0.8268, + "num_input_tokens_seen": 1828625, + "step": 83 + }, + { + "epoch": 0.010100402813683641, + "flos": 14755838071440.0, + "grad_norm": 3.313172038301558, + "learning_rate": 3.2098872850910627e-06, + "loss": 0.9393, + "num_input_tokens_seen": 1845020, + "step": 84 + }, + { + "epoch": 0.010220645704322733, + "flos": 12543523418520.0, + "grad_norm": 3.353935344293502, + "learning_rate": 3.2184607100038194e-06, + "loss": 0.8808, + "num_input_tokens_seen": 1863380, + "step": 85 + }, + { + "epoch": 0.010340888594961822, + "flos": 15702693597360.0, + "grad_norm": 3.1380323649889688, + "learning_rate": 3.2269338586412414e-06, + "loss": 0.9127, + "num_input_tokens_seen": 1880685, + "step": 86 + }, + { + "epoch": 0.010461131485600914, + "flos": 16836786347160.0, + "grad_norm": 15.044603788382533, + "learning_rate": 3.2353090496083106e-06, + "loss": 0.9546, + "num_input_tokens_seen": 1898240, + "step": 87 + }, + { + "epoch": 0.010581374376240005, + "flos": 24640016743680.0, + "grad_norm": 3.476606938796039, + "learning_rate": 3.2435885220114572e-06, + "loss": 0.8107, + "num_input_tokens_seen": 1919310, + "step": 88 + }, + { + "epoch": 0.010701617266879095, + "flos": 15919401752640.0, + "grad_norm": 3.221154361203631, + "learning_rate": 3.2517744390519113e-06, + "loss": 0.9245, + "num_input_tokens_seen": 1937895, + "step": 89 + }, + { + "epoch": 0.010821860157518187, + "flos": 13918666341480.0, + "grad_norm": 3.57502098877621, + "learning_rate": 3.259868891418298e-06, + "loss": 0.7398, + "num_input_tokens_seen": 1955380, + "step": 90 + }, + { + "epoch": 0.010942103048157278, + "flos": 18645915459960.0, + "grad_norm": 2.850145463410185, + "learning_rate": 3.2678739004917757e-06, + "loss": 0.8238, + "num_input_tokens_seen": 1974835, + "step": 91 + }, + { + "epoch": 0.011062345938796368, + "flos": 20160972949320.0, + "grad_norm": 2.6337257861385353, + "learning_rate": 3.275791421376029e-06, + "loss": 0.9096, + "num_input_tokens_seen": 1998000, + "step": 92 + }, + { + "epoch": 0.01118258882943546, + "flos": 11707016547720.0, + "grad_norm": 3.419591690585627, + "learning_rate": 3.2836233457634622e-06, + "loss": 0.9398, + "num_input_tokens_seen": 2015855, + "step": 93 + }, + { + "epoch": 0.011302831720074551, + "flos": 15109421436960.0, + "grad_norm": 2.9625103077125696, + "learning_rate": 3.2913715046481135e-06, + "loss": 0.8301, + "num_input_tokens_seen": 2035320, + "step": 94 + }, + { + "epoch": 0.011423074610713641, + "flos": 9486565285080.0, + "grad_norm": 3.6233140379091586, + "learning_rate": 3.299037670895023e-06, + "loss": 0.8683, + "num_input_tokens_seen": 2051655, + "step": 95 + }, + { + "epoch": 0.011543317501352733, + "flos": 22263445506480.0, + "grad_norm": 2.706996548693077, + "learning_rate": 3.3066235616750667e-06, + "loss": 0.7987, + "num_input_tokens_seen": 2072610, + "step": 96 + }, + { + "epoch": 0.011663560391991824, + "flos": 11301139911000.0, + "grad_norm": 3.3290104650743753, + "learning_rate": 3.3141308407736276e-06, + "loss": 0.9006, + "num_input_tokens_seen": 2088965, + "step": 97 + }, + { + "epoch": 0.011783803282630914, + "flos": 14542137612360.0, + "grad_norm": 3.1030764788227265, + "learning_rate": 3.321561120780869e-06, + "loss": 0.8482, + "num_input_tokens_seen": 2107395, + "step": 98 + }, + { + "epoch": 0.011904046173270006, + "flos": 16345979511840.0, + "grad_norm": 3.6309801956767105, + "learning_rate": 3.3289159651708192e-06, + "loss": 1.0016, + "num_input_tokens_seen": 2124690, + "step": 99 + }, + { + "epoch": 0.012024289063909096, + "flos": 13948643832240.0, + "grad_norm": 3.5272053198196534, + "learning_rate": 3.3361968902759768e-06, + "loss": 0.9633, + "num_input_tokens_seen": 2144090, + "step": 100 + }, + { + "epoch": 0.012144531954548187, + "flos": 10923151207680.0, + "grad_norm": 4.604635407535338, + "learning_rate": 3.343405367163663e-06, + "loss": 0.9351, + "num_input_tokens_seen": 2160740, + "step": 101 + }, + { + "epoch": 0.012264774845187279, + "flos": 11004472170840.0, + "grad_norm": 3.5824568068249074, + "learning_rate": 3.350542823419951e-06, + "loss": 0.8038, + "num_input_tokens_seen": 2177060, + "step": 102 + }, + { + "epoch": 0.012385017735826368, + "flos": 10135803272040.0, + "grad_norm": 4.713867086343742, + "learning_rate": 3.3576106448465615e-06, + "loss": 0.8571, + "num_input_tokens_seen": 2190160, + "step": 103 + }, + { + "epoch": 0.01250526062646546, + "flos": 17295969373800.0, + "grad_norm": 2.677911286523508, + "learning_rate": 3.3646101770757797e-06, + "loss": 0.8653, + "num_input_tokens_seen": 2208670, + "step": 104 + }, + { + "epoch": 0.012625503517104552, + "flos": 25448888960760.0, + "grad_norm": 2.798810406433796, + "learning_rate": 3.371542727108104e-06, + "loss": 0.8352, + "num_input_tokens_seen": 2230565, + "step": 105 + }, + { + "epoch": 0.012745746407743641, + "flos": 13002548145360.0, + "grad_norm": 4.574011298735708, + "learning_rate": 3.3784095647770114e-06, + "loss": 0.884, + "num_input_tokens_seen": 2248930, + "step": 106 + }, + { + "epoch": 0.012865989298382733, + "flos": 15054911968560.0, + "grad_norm": 3.487873732801168, + "learning_rate": 3.3852119241449547e-06, + "loss": 0.8663, + "num_input_tokens_seen": 2267770, + "step": 107 + }, + { + "epoch": 0.012986232189021825, + "flos": 17536639627440.0, + "grad_norm": 3.242975125528243, + "learning_rate": 3.3919510048344295e-06, + "loss": 0.945, + "num_input_tokens_seen": 2285500, + "step": 108 + }, + { + "epoch": 0.013106475079660914, + "flos": 17370546765480.0, + "grad_norm": 4.2261033011437945, + "learning_rate": 3.3986279732976907e-06, + "loss": 0.8398, + "num_input_tokens_seen": 2303695, + "step": 109 + }, + { + "epoch": 0.013226717970300006, + "flos": 15429924635640.0, + "grad_norm": 4.043200382102523, + "learning_rate": 3.4052439640284983e-06, + "loss": 0.9371, + "num_input_tokens_seen": 2322330, + "step": 110 + }, + { + "epoch": 0.013346960860939098, + "flos": 18214145467320.0, + "grad_norm": 4.027525134585357, + "learning_rate": 3.4118000807190217e-06, + "loss": 0.8031, + "num_input_tokens_seen": 2342930, + "step": 111 + }, + { + "epoch": 0.013467203751578187, + "flos": 20669409891000.0, + "grad_norm": 3.6971547089477514, + "learning_rate": 3.4182973973648723e-06, + "loss": 0.7544, + "num_input_tokens_seen": 2363860, + "step": 112 + }, + { + "epoch": 0.013587446642217279, + "flos": 13812401821200.0, + "grad_norm": 3.9076008406481315, + "learning_rate": 3.424736959321014e-06, + "loss": 0.9226, + "num_input_tokens_seen": 2381385, + "step": 113 + }, + { + "epoch": 0.01370768953285637, + "flos": 17566427158440.0, + "grad_norm": 3.03921014186877, + "learning_rate": 3.431119784311155e-06, + "loss": 0.8678, + "num_input_tokens_seen": 2400780, + "step": 114 + }, + { + "epoch": 0.01382793242349546, + "flos": 28825685433720.0, + "grad_norm": 14.510905407052375, + "learning_rate": 3.43744686339307e-06, + "loss": 0.7627, + "num_input_tokens_seen": 2422820, + "step": 115 + }, + { + "epoch": 0.013948175314134552, + "flos": 30417283232280.0, + "grad_norm": 4.7220946325879805, + "learning_rate": 3.44371916188212e-06, + "loss": 0.8903, + "num_input_tokens_seen": 2443295, + "step": 116 + }, + { + "epoch": 0.014068418204773643, + "flos": 16429991571600.0, + "grad_norm": 3.158379770975929, + "learning_rate": 3.449937620235143e-06, + "loss": 0.8523, + "num_input_tokens_seen": 2463610, + "step": 117 + }, + { + "epoch": 0.014188661095412733, + "flos": 17430501747000.0, + "grad_norm": 4.019049345884337, + "learning_rate": 3.456103154896722e-06, + "loss": 0.8646, + "num_input_tokens_seen": 2484605, + "step": 118 + }, + { + "epoch": 0.014308903986051825, + "flos": 17323414067760.0, + "grad_norm": 2.7940144321994596, + "learning_rate": 3.462216659109757e-06, + "loss": 0.9095, + "num_input_tokens_seen": 2504505, + "step": 119 + }, + { + "epoch": 0.014429146876690916, + "flos": 15108883217640.0, + "grad_norm": 6.324216940433651, + "learning_rate": 3.4682790036921077e-06, + "loss": 0.8315, + "num_input_tokens_seen": 2522885, + "step": 120 + }, + { + "epoch": 0.014549389767330006, + "flos": 14784010944480.0, + "grad_norm": 3.398480013877398, + "learning_rate": 3.4742910377810193e-06, + "loss": 0.8223, + "num_input_tokens_seen": 2540065, + "step": 121 + }, + { + "epoch": 0.014669632657969098, + "flos": 13137428778120.0, + "grad_norm": 5.4478643450099655, + "learning_rate": 3.4802535895469042e-06, + "loss": 0.867, + "num_input_tokens_seen": 2558535, + "step": 122 + }, + { + "epoch": 0.01478987554860819, + "flos": 16642805551800.0, + "grad_norm": 3.017696321512022, + "learning_rate": 3.4861674668779934e-06, + "loss": 0.8838, + "num_input_tokens_seen": 2576485, + "step": 123 + }, + { + "epoch": 0.01491011843924728, + "flos": 12540832321920.0, + "grad_norm": 3.2389351215701625, + "learning_rate": 3.492033458037272e-06, + "loss": 0.8353, + "num_input_tokens_seen": 2594775, + "step": 124 + }, + { + "epoch": 0.01503036132988637, + "flos": 12892706049600.0, + "grad_norm": 4.679628196684444, + "learning_rate": 3.497852332293018e-06, + "loss": 0.8498, + "num_input_tokens_seen": 2610070, + "step": 125 + }, + { + "epoch": 0.015150604220525462, + "flos": 13784070648360.0, + "grad_norm": 7.34467222226358, + "learning_rate": 3.5036248405242356e-06, + "loss": 0.9649, + "num_input_tokens_seen": 2628545, + "step": 126 + }, + { + "epoch": 0.015270847111164552, + "flos": 28986554402280.0, + "grad_norm": 3.6089088832549816, + "learning_rate": 3.509351715802146e-06, + "loss": 0.8139, + "num_input_tokens_seen": 2649150, + "step": 127 + }, + { + "epoch": 0.015391090001803644, + "flos": 32200202389560.0, + "grad_norm": 6.493564060401882, + "learning_rate": 3.5150336739488763e-06, + "loss": 0.7712, + "num_input_tokens_seen": 2671155, + "step": 128 + }, + { + "epoch": 0.015511332892442733, + "flos": 13812496801080.0, + "grad_norm": 8.66026504991985, + "learning_rate": 3.5206714140744143e-06, + "loss": 0.8159, + "num_input_tokens_seen": 2690930, + "step": 129 + }, + { + "epoch": 0.015631575783081827, + "flos": 17970910756920.0, + "grad_norm": 4.0911663725896075, + "learning_rate": 3.5262656190928208e-06, + "loss": 0.8493, + "num_input_tokens_seen": 2708950, + "step": 130 + }, + { + "epoch": 0.015751818673720917, + "flos": 45944770390560.0, + "grad_norm": 0.9399841774461044, + "learning_rate": 3.5318169562186737e-06, + "loss": 0.7087, + "num_input_tokens_seen": 2777515, + "step": 131 + }, + { + "epoch": 0.015872061564360006, + "flos": 17484884575560.0, + "grad_norm": 3.691314276301299, + "learning_rate": 3.5373260774446292e-06, + "loss": 0.8141, + "num_input_tokens_seen": 2797685, + "step": 132 + }, + { + "epoch": 0.0159923044549991, + "flos": 17106990852120.0, + "grad_norm": 3.5402008314331384, + "learning_rate": 3.542793620000961e-06, + "loss": 0.8935, + "num_input_tokens_seen": 2816880, + "step": 133 + }, + { + "epoch": 0.01611254734563819, + "flos": 13029486279960.0, + "grad_norm": 4.677219153841471, + "learning_rate": 3.5482202067978894e-06, + "loss": 0.8464, + "num_input_tokens_seen": 2833810, + "step": 134 + }, + { + "epoch": 0.01623279023627728, + "flos": 15320399139480.0, + "grad_norm": 4.266674094392351, + "learning_rate": 3.553606446851471e-06, + "loss": 0.7391, + "num_input_tokens_seen": 2850270, + "step": 135 + }, + { + "epoch": 0.016353033126916373, + "flos": 11463623537520.0, + "grad_norm": 3.655207558725398, + "learning_rate": 3.5589529356937613e-06, + "loss": 0.8232, + "num_input_tokens_seen": 2868385, + "step": 136 + }, + { + "epoch": 0.016473276017555463, + "flos": 13731840697080.0, + "grad_norm": 2.5369245243055034, + "learning_rate": 3.5642602557679627e-06, + "loss": 0.7556, + "num_input_tokens_seen": 2886555, + "step": 137 + }, + { + "epoch": 0.016593518908194552, + "flos": 17835713524560.0, + "grad_norm": 4.654622657634122, + "learning_rate": 3.569528976809202e-06, + "loss": 0.8306, + "num_input_tokens_seen": 2903490, + "step": 138 + }, + { + "epoch": 0.016713761798833646, + "flos": 16350791825760.0, + "grad_norm": 5.422779722337261, + "learning_rate": 3.5747596562115522e-06, + "loss": 0.8679, + "num_input_tokens_seen": 2923825, + "step": 139 + }, + { + "epoch": 0.016834004689472735, + "flos": 13002484825440.0, + "grad_norm": 4.424777422980722, + "learning_rate": 3.5799528393819138e-06, + "loss": 0.8939, + "num_input_tokens_seen": 2942625, + "step": 140 + }, + { + "epoch": 0.016954247580111825, + "flos": 14998629542400.0, + "grad_norm": 10.642236707968749, + "learning_rate": 3.585109060081286e-06, + "loss": 0.8597, + "num_input_tokens_seen": 2962145, + "step": 141 + }, + { + "epoch": 0.017074490470750915, + "flos": 16160135326200.0, + "grad_norm": 3.8955489410118886, + "learning_rate": 3.590228840753992e-06, + "loss": 0.7718, + "num_input_tokens_seen": 2982295, + "step": 142 + }, + { + "epoch": 0.01719473336139001, + "flos": 11652158819760.0, + "grad_norm": 21.423409238001927, + "learning_rate": 3.5953126928453423e-06, + "loss": 0.8573, + "num_input_tokens_seen": 2999565, + "step": 143 + }, + { + "epoch": 0.017314976252029098, + "flos": 16458765983880.0, + "grad_norm": 3.285644349840703, + "learning_rate": 3.600361117108239e-06, + "loss": 0.7898, + "num_input_tokens_seen": 3019085, + "step": 144 + }, + { + "epoch": 0.017435219142668188, + "flos": 16108032014760.0, + "grad_norm": 3.520134640199411, + "learning_rate": 3.6053746038991616e-06, + "loss": 0.9494, + "num_input_tokens_seen": 3037890, + "step": 145 + }, + { + "epoch": 0.01755546203330728, + "flos": 53278591593360.0, + "grad_norm": 0.9713907857065575, + "learning_rate": 3.6103536334639843e-06, + "loss": 0.6136, + "num_input_tokens_seen": 3090875, + "step": 146 + }, + { + "epoch": 0.01767570492394637, + "flos": 18564151257360.0, + "grad_norm": 2.988292579380076, + "learning_rate": 3.615298676214041e-06, + "loss": 0.8369, + "num_input_tokens_seen": 3110875, + "step": 147 + }, + { + "epoch": 0.01779594781458546, + "flos": 14946969470400.0, + "grad_norm": 3.225647434037604, + "learning_rate": 3.6202101929928317e-06, + "loss": 0.8797, + "num_input_tokens_seen": 3129185, + "step": 148 + }, + { + "epoch": 0.017916190705224554, + "flos": 11841992160360.0, + "grad_norm": 2.9853881394451287, + "learning_rate": 3.6250886353337413e-06, + "loss": 0.8718, + "num_input_tokens_seen": 3146435, + "step": 149 + }, + { + "epoch": 0.018036433595863644, + "flos": 17079261218520.0, + "grad_norm": 4.929706845494926, + "learning_rate": 3.6299344457091488e-06, + "loss": 0.8562, + "num_input_tokens_seen": 3167015, + "step": 150 + }, + { + "epoch": 0.018156676486502734, + "flos": 13569515370360.0, + "grad_norm": 3.3389518751598963, + "learning_rate": 3.634748057771256e-06, + "loss": 0.9073, + "num_input_tokens_seen": 3182675, + "step": 151 + }, + { + "epoch": 0.018276919377141827, + "flos": 18647086878480.0, + "grad_norm": 3.94498785212862, + "learning_rate": 3.639529896584965e-06, + "loss": 0.8454, + "num_input_tokens_seen": 3203770, + "step": 152 + }, + { + "epoch": 0.018397162267780917, + "flos": 14649478571280.0, + "grad_norm": 6.89100448705239, + "learning_rate": 3.6442803788531233e-06, + "loss": 0.8609, + "num_input_tokens_seen": 3221450, + "step": 153 + }, + { + "epoch": 0.018517405158420007, + "flos": 20213551160160.0, + "grad_norm": 5.30421789503668, + "learning_rate": 3.6489999131344357e-06, + "loss": 0.9378, + "num_input_tokens_seen": 3243945, + "step": 154 + }, + { + "epoch": 0.0186376480490591, + "flos": 14136419275440.0, + "grad_norm": 4.6808846141380185, + "learning_rate": 3.653688900054313e-06, + "loss": 0.8932, + "num_input_tokens_seen": 3262195, + "step": 155 + }, + { + "epoch": 0.01875789093969819, + "flos": 19563933253680.0, + "grad_norm": 4.5950417291948495, + "learning_rate": 3.6583477325089526e-06, + "loss": 0.7488, + "num_input_tokens_seen": 3282455, + "step": 156 + }, + { + "epoch": 0.01887813383033728, + "flos": 17836030124160.0, + "grad_norm": 3.970115847110452, + "learning_rate": 3.6629767958628916e-06, + "loss": 1.0233, + "num_input_tokens_seen": 3299550, + "step": 157 + }, + { + "epoch": 0.018998376720976373, + "flos": 10653294962280.0, + "grad_norm": 4.3476834497689145, + "learning_rate": 3.667576468140291e-06, + "loss": 0.8365, + "num_input_tokens_seen": 3317085, + "step": 158 + }, + { + "epoch": 0.019118619611615463, + "flos": 21499648089720.0, + "grad_norm": 3.523624660486108, + "learning_rate": 3.672147120210184e-06, + "loss": 0.8687, + "num_input_tokens_seen": 3333405, + "step": 159 + }, + { + "epoch": 0.019238862502254553, + "flos": 15270733644960.0, + "grad_norm": 3.9997799335978366, + "learning_rate": 3.6766891159659177e-06, + "loss": 0.8586, + "num_input_tokens_seen": 3351535, + "step": 160 + }, + { + "epoch": 0.019359105392893646, + "flos": 15621435954120.0, + "grad_norm": 10.175053206920929, + "learning_rate": 3.6812028124990075e-06, + "loss": 0.8592, + "num_input_tokens_seen": 3368525, + "step": 161 + }, + { + "epoch": 0.019479348283532736, + "flos": 11863389801960.0, + "grad_norm": 4.382431040872512, + "learning_rate": 3.6856885602676016e-06, + "loss": 0.8058, + "num_input_tokens_seen": 3384280, + "step": 162 + }, + { + "epoch": 0.019599591174171826, + "flos": 16162224883560.0, + "grad_norm": 3.77697726881213, + "learning_rate": 3.6901467032597733e-06, + "loss": 0.9314, + "num_input_tokens_seen": 3402485, + "step": 163 + }, + { + "epoch": 0.01971983406481092, + "flos": 14325619416840.0, + "grad_norm": 3.0422523840365825, + "learning_rate": 3.694577579151804e-06, + "loss": 0.8507, + "num_input_tokens_seen": 3420615, + "step": 164 + }, + { + "epoch": 0.01984007695545001, + "flos": 13948010633040.0, + "grad_norm": 4.325415689414194, + "learning_rate": 3.6989815194616703e-06, + "loss": 0.7339, + "num_input_tokens_seen": 3437530, + "step": 165 + }, + { + "epoch": 0.0199603198460891, + "flos": 15242117532480.0, + "grad_norm": 12.692833291577315, + "learning_rate": 3.703358849697888e-06, + "loss": 0.7898, + "num_input_tokens_seen": 3457160, + "step": 166 + }, + { + "epoch": 0.020080562736728192, + "flos": 15998823118200.0, + "grad_norm": 2.4375412766311797, + "learning_rate": 3.7077098895038803e-06, + "loss": 0.8195, + "num_input_tokens_seen": 3476250, + "step": 167 + }, + { + "epoch": 0.020200805627367282, + "flos": 15864227425080.0, + "grad_norm": 3.9530750618294626, + "learning_rate": 3.712034952798045e-06, + "loss": 0.9567, + "num_input_tokens_seen": 3494085, + "step": 168 + }, + { + "epoch": 0.02032104851800637, + "flos": 24636660787920.0, + "grad_norm": 3.668835749753182, + "learning_rate": 3.7163343479096656e-06, + "loss": 0.8156, + "num_input_tokens_seen": 3515380, + "step": 169 + }, + { + "epoch": 0.020441291408645465, + "flos": 23263702402200.0, + "grad_norm": 3.0124203746608598, + "learning_rate": 3.720608377710802e-06, + "loss": 0.8085, + "num_input_tokens_seen": 3535190, + "step": 170 + }, + { + "epoch": 0.020561534299284555, + "flos": 15270701985000.0, + "grad_norm": 3.61718473166572, + "learning_rate": 3.7248573397443277e-06, + "loss": 0.8468, + "num_input_tokens_seen": 3553835, + "step": 171 + }, + { + "epoch": 0.020681777189923645, + "flos": 15352117928040.0, + "grad_norm": 6.106735088403738, + "learning_rate": 3.729081526348224e-06, + "loss": 0.9664, + "num_input_tokens_seen": 3572085, + "step": 172 + }, + { + "epoch": 0.020802020080562738, + "flos": 20725723977120.0, + "grad_norm": 3.008506318364507, + "learning_rate": 3.7332812247762777e-06, + "loss": 0.8251, + "num_input_tokens_seen": 3593105, + "step": 173 + }, + { + "epoch": 0.020922262971201828, + "flos": 14378419247400.0, + "grad_norm": 2.92885209443298, + "learning_rate": 3.737456717315293e-06, + "loss": 0.9382, + "num_input_tokens_seen": 3611790, + "step": 174 + }, + { + "epoch": 0.021042505861840918, + "flos": 11406897871920.0, + "grad_norm": 2.4415218722621295, + "learning_rate": 3.7416082813989552e-06, + "loss": 0.8989, + "num_input_tokens_seen": 3628505, + "step": 175 + }, + { + "epoch": 0.02116274875248001, + "flos": 15459648846720.0, + "grad_norm": 7.25044038684704, + "learning_rate": 3.745736189718439e-06, + "loss": 0.8936, + "num_input_tokens_seen": 3647480, + "step": 176 + }, + { + "epoch": 0.0212829916431191, + "flos": 18104208391680.0, + "grad_norm": 3.3418346873317377, + "learning_rate": 3.749840710329894e-06, + "loss": 0.6951, + "num_input_tokens_seen": 3667905, + "step": 177 + }, + { + "epoch": 0.02140323453375819, + "flos": 12130681590600.0, + "grad_norm": 5.581285623092852, + "learning_rate": 3.7539221067588938e-06, + "loss": 0.9607, + "num_input_tokens_seen": 3681600, + "step": 178 + }, + { + "epoch": 0.021523477424397284, + "flos": 14835512716680.0, + "grad_norm": 11.586723727958947, + "learning_rate": 3.757980638101964e-06, + "loss": 0.9063, + "num_input_tokens_seen": 3694815, + "step": 179 + }, + { + "epoch": 0.021643720315036374, + "flos": 19131846661440.0, + "grad_norm": 4.234239414198027, + "learning_rate": 3.7620165591252806e-06, + "loss": 0.8824, + "num_input_tokens_seen": 3712635, + "step": 180 + }, + { + "epoch": 0.021763963205675464, + "flos": 18157261501920.0, + "grad_norm": 25.464788323014538, + "learning_rate": 3.766030120360636e-06, + "loss": 0.9203, + "num_input_tokens_seen": 3731985, + "step": 181 + }, + { + "epoch": 0.021884206096314557, + "flos": 18917101423680.0, + "grad_norm": 8.002763747612299, + "learning_rate": 3.7700215681987578e-06, + "loss": 0.8687, + "num_input_tokens_seen": 3751475, + "step": 182 + }, + { + "epoch": 0.022004448986953647, + "flos": 14675530227000.0, + "grad_norm": 3.5275899097817818, + "learning_rate": 3.7739911449800767e-06, + "loss": 0.8102, + "num_input_tokens_seen": 3771250, + "step": 183 + }, + { + "epoch": 0.022124691877592736, + "flos": 14969506870560.0, + "grad_norm": 3.7002353902636216, + "learning_rate": 3.7779390890830114e-06, + "loss": 0.7917, + "num_input_tokens_seen": 3789620, + "step": 184 + }, + { + "epoch": 0.02224493476823183, + "flos": 17269031239200.0, + "grad_norm": 4.094090571323645, + "learning_rate": 3.7818656350098723e-06, + "loss": 0.8415, + "num_input_tokens_seen": 3810290, + "step": 185 + }, + { + "epoch": 0.02236517765887092, + "flos": 12327670082160.0, + "grad_norm": 6.46014982539891, + "learning_rate": 3.7857710134704447e-06, + "loss": 0.7493, + "num_input_tokens_seen": 3828945, + "step": 186 + }, + { + "epoch": 0.02248542054951001, + "flos": 32012173666680.0, + "grad_norm": 3.9089427974377084, + "learning_rate": 3.7896554514633234e-06, + "loss": 0.7712, + "num_input_tokens_seen": 3852950, + "step": 187 + }, + { + "epoch": 0.022605663440149103, + "flos": 17106769232400.0, + "grad_norm": 4.513436919187675, + "learning_rate": 3.7935191723550955e-06, + "loss": 0.8429, + "num_input_tokens_seen": 3871625, + "step": 188 + }, + { + "epoch": 0.022725906330788193, + "flos": 21289398566280.0, + "grad_norm": 4.465374078234581, + "learning_rate": 3.797362395957408e-06, + "loss": 0.8766, + "num_input_tokens_seen": 3891910, + "step": 189 + }, + { + "epoch": 0.022846149221427282, + "flos": 17941598125320.0, + "grad_norm": 4.258998544671696, + "learning_rate": 3.8011853386020055e-06, + "loss": 0.7666, + "num_input_tokens_seen": 3912535, + "step": 190 + }, + { + "epoch": 0.022966392112066376, + "flos": 11301646470360.0, + "grad_norm": 4.710135775996305, + "learning_rate": 3.804988213213804e-06, + "loss": 0.8712, + "num_input_tokens_seen": 3930495, + "step": 191 + }, + { + "epoch": 0.023086635002705466, + "flos": 41002332794880.0, + "grad_norm": 1.033305979694802, + "learning_rate": 3.808771229382049e-06, + "loss": 0.6454, + "num_input_tokens_seen": 3989680, + "step": 192 + }, + { + "epoch": 0.023206877893344555, + "flos": 14105776925520.0, + "grad_norm": 20.39269475845992, + "learning_rate": 3.8125345934296324e-06, + "loss": 0.8414, + "num_input_tokens_seen": 4007710, + "step": 193 + }, + { + "epoch": 0.02332712078398365, + "flos": 16888288119360.0, + "grad_norm": 4.3272338390424325, + "learning_rate": 3.81627850848061e-06, + "loss": 0.8518, + "num_input_tokens_seen": 4028025, + "step": 194 + }, + { + "epoch": 0.02344736367462274, + "flos": 17890033033200.0, + "grad_norm": 6.654878143639405, + "learning_rate": 3.820003174525994e-06, + "loss": 0.8272, + "num_input_tokens_seen": 4047170, + "step": 195 + }, + { + "epoch": 0.02356760656526183, + "flos": 15783634641000.0, + "grad_norm": 14.866011267842662, + "learning_rate": 3.823708788487851e-06, + "loss": 0.8233, + "num_input_tokens_seen": 4063890, + "step": 196 + }, + { + "epoch": 0.02368784945590092, + "flos": 18458329976520.0, + "grad_norm": 3.465962157433028, + "learning_rate": 3.827395544281781e-06, + "loss": 0.8254, + "num_input_tokens_seen": 4085950, + "step": 197 + }, + { + "epoch": 0.02380809234654001, + "flos": 20211081683280.0, + "grad_norm": 6.831950846439402, + "learning_rate": 3.831063632877802e-06, + "loss": 0.7803, + "num_input_tokens_seen": 4105990, + "step": 198 + }, + { + "epoch": 0.0239283352371791, + "flos": 13326628919520.0, + "grad_norm": 13.73681119071865, + "learning_rate": 3.834713242359712e-06, + "loss": 0.7421, + "num_input_tokens_seen": 4123540, + "step": 199 + }, + { + "epoch": 0.02404857812781819, + "flos": 15646632790920.0, + "grad_norm": 4.032956966988739, + "learning_rate": 3.838344557982959e-06, + "loss": 0.8588, + "num_input_tokens_seen": 4144640, + "step": 200 + }, + { + "epoch": 0.024168821018457284, + "flos": 12138406620840.0, + "grad_norm": 6.402213560203922, + "learning_rate": 3.841957762231063e-06, + "loss": 0.8328, + "num_input_tokens_seen": 4161890, + "step": 201 + }, + { + "epoch": 0.024289063909096374, + "flos": 16702190654040.0, + "grad_norm": 3.010191454707514, + "learning_rate": 3.8455530348706454e-06, + "loss": 0.8668, + "num_input_tokens_seen": 4181210, + "step": 202 + }, + { + "epoch": 0.024409306799735464, + "flos": 12948260296680.0, + "grad_norm": 2.880885526894011, + "learning_rate": 3.849130553005099e-06, + "loss": 0.7622, + "num_input_tokens_seen": 4199145, + "step": 203 + }, + { + "epoch": 0.024529549690374557, + "flos": 15810572775600.0, + "grad_norm": 3.4276130676540393, + "learning_rate": 3.852690491126933e-06, + "loss": 0.8214, + "num_input_tokens_seen": 4218915, + "step": 204 + }, + { + "epoch": 0.024649792581013647, + "flos": 18722360789280.0, + "grad_norm": 5.187774834763557, + "learning_rate": 3.856233021168845e-06, + "loss": 0.8875, + "num_input_tokens_seen": 4238330, + "step": 205 + }, + { + "epoch": 0.024770035471652737, + "flos": 25342149541080.0, + "grad_norm": 5.733480191382933, + "learning_rate": 3.859758312553544e-06, + "loss": 0.909, + "num_input_tokens_seen": 4260270, + "step": 206 + }, + { + "epoch": 0.02489027836229183, + "flos": 15728396993520.0, + "grad_norm": 8.340873765433598, + "learning_rate": 3.8632665322423735e-06, + "loss": 0.9094, + "num_input_tokens_seen": 4279645, + "step": 207 + }, + { + "epoch": 0.02501052125293092, + "flos": 16996388917320.0, + "grad_norm": 32.16222270515422, + "learning_rate": 3.866757844782762e-06, + "loss": 0.8417, + "num_input_tokens_seen": 4299870, + "step": 208 + }, + { + "epoch": 0.02513076414357001, + "flos": 19341716265360.0, + "grad_norm": 5.94519123048955, + "learning_rate": 3.870232412354527e-06, + "loss": 0.9011, + "num_input_tokens_seen": 4316435, + "step": 209 + }, + { + "epoch": 0.025251007034209103, + "flos": 9871107600120.0, + "grad_norm": 3.392298553490418, + "learning_rate": 3.873690394815086e-06, + "loss": 0.9128, + "num_input_tokens_seen": 4332735, + "step": 210 + }, + { + "epoch": 0.025371249924848193, + "flos": 10950627561600.0, + "grad_norm": 6.470321095913905, + "learning_rate": 3.877131949743587e-06, + "loss": 0.8981, + "num_input_tokens_seen": 4349320, + "step": 211 + }, + { + "epoch": 0.025491492815487283, + "flos": 18724292046840.0, + "grad_norm": 4.261177399235895, + "learning_rate": 3.880557232483993e-06, + "loss": 0.7714, + "num_input_tokens_seen": 4368480, + "step": 212 + }, + { + "epoch": 0.025611735706126376, + "flos": 15081216903960.0, + "grad_norm": 33.99973072618554, + "learning_rate": 3.883966396187164e-06, + "loss": 0.8727, + "num_input_tokens_seen": 4387470, + "step": 213 + }, + { + "epoch": 0.025731978596765466, + "flos": 13920502619160.0, + "grad_norm": 2.9494239179980695, + "learning_rate": 3.887359591851937e-06, + "loss": 0.893, + "num_input_tokens_seen": 4404795, + "step": 214 + }, + { + "epoch": 0.025852221487404556, + "flos": 16215879533040.0, + "grad_norm": 4.66285063624262, + "learning_rate": 3.890736968365265e-06, + "loss": 0.9213, + "num_input_tokens_seen": 4424830, + "step": 215 + }, + { + "epoch": 0.02597246437804365, + "flos": 19455990755520.0, + "grad_norm": 22.001822656209928, + "learning_rate": 3.894098672541412e-06, + "loss": 0.845, + "num_input_tokens_seen": 4445455, + "step": 216 + }, + { + "epoch": 0.02609270726868274, + "flos": 23992393414680.0, + "grad_norm": 3.2469754450427324, + "learning_rate": 3.89744484916025e-06, + "loss": 0.7406, + "num_input_tokens_seen": 4466285, + "step": 217 + }, + { + "epoch": 0.02621295015932183, + "flos": 19234217006640.0, + "grad_norm": 3.636687543921564, + "learning_rate": 3.900775641004673e-06, + "loss": 0.8627, + "num_input_tokens_seen": 4485320, + "step": 218 + }, + { + "epoch": 0.026333193049960922, + "flos": 31577490957720.0, + "grad_norm": 6.491333098486613, + "learning_rate": 3.904091188897156e-06, + "loss": 0.7191, + "num_input_tokens_seen": 4504175, + "step": 219 + }, + { + "epoch": 0.026453435940600012, + "flos": 13107039707880.0, + "grad_norm": 3.9206249174911316, + "learning_rate": 3.90739163173548e-06, + "loss": 0.8025, + "num_input_tokens_seen": 4521730, + "step": 220 + }, + { + "epoch": 0.026573678831239102, + "flos": 13862098975680.0, + "grad_norm": 6.290399520630657, + "learning_rate": 3.910677106527646e-06, + "loss": 0.868, + "num_input_tokens_seen": 4538640, + "step": 221 + }, + { + "epoch": 0.026693921721878195, + "flos": 21453718470480.0, + "grad_norm": 4.1256556134416575, + "learning_rate": 3.913947748426004e-06, + "loss": 0.829, + "num_input_tokens_seen": 4555180, + "step": 222 + }, + { + "epoch": 0.026814164612517285, + "flos": 10267644548640.0, + "grad_norm": 22.238999980151704, + "learning_rate": 3.9172036907606136e-06, + "loss": 0.748, + "num_input_tokens_seen": 4568380, + "step": 223 + }, + { + "epoch": 0.026934407503156375, + "flos": 17213350352280.0, + "grad_norm": 8.100102233241556, + "learning_rate": 3.920445065071855e-06, + "loss": 0.9329, + "num_input_tokens_seen": 4589265, + "step": 224 + }, + { + "epoch": 0.027054650393795468, + "flos": 21238403353440.0, + "grad_norm": 10.000373833344463, + "learning_rate": 3.923672001142322e-06, + "loss": 0.7597, + "num_input_tokens_seen": 4609295, + "step": 225 + }, + { + "epoch": 0.027174893284434558, + "flos": 23074692220560.0, + "grad_norm": 7.858694968561047, + "learning_rate": 3.926884627027996e-06, + "loss": 0.8391, + "num_input_tokens_seen": 4632785, + "step": 226 + }, + { + "epoch": 0.027295136175073648, + "flos": 16134970149360.0, + "grad_norm": 2.831703072310705, + "learning_rate": 3.930083069088744e-06, + "loss": 0.7694, + "num_input_tokens_seen": 4652505, + "step": 227 + }, + { + "epoch": 0.02741537906571274, + "flos": 44814761775600.0, + "grad_norm": 1.0205057682582694, + "learning_rate": 3.933267452018137e-06, + "loss": 0.6265, + "num_input_tokens_seen": 4712020, + "step": 228 + }, + { + "epoch": 0.02753562195635183, + "flos": 18023172368160.0, + "grad_norm": 2.8999973428648254, + "learning_rate": 3.936437898872622e-06, + "loss": 0.8288, + "num_input_tokens_seen": 4731715, + "step": 229 + }, + { + "epoch": 0.02765586484699092, + "flos": 25045133541360.0, + "grad_norm": 12.282690531570676, + "learning_rate": 3.9395945311000525e-06, + "loss": 0.784, + "num_input_tokens_seen": 4753060, + "step": 230 + }, + { + "epoch": 0.027776107737630014, + "flos": 10788270574920.0, + "grad_norm": 3.799082173381805, + "learning_rate": 3.942737468567608e-06, + "loss": 0.8986, + "num_input_tokens_seen": 4770795, + "step": 231 + }, + { + "epoch": 0.027896350628269104, + "flos": 35277101806440.0, + "grad_norm": 6.2837394981076455, + "learning_rate": 3.9458668295891026e-06, + "loss": 0.8501, + "num_input_tokens_seen": 4792800, + "step": 232 + }, + { + "epoch": 0.028016593518908194, + "flos": 15860491549800.0, + "grad_norm": 4.587677139895199, + "learning_rate": 3.948982730951712e-06, + "loss": 0.8539, + "num_input_tokens_seen": 4810615, + "step": 233 + }, + { + "epoch": 0.028136836409547287, + "flos": 13459230035160.0, + "grad_norm": 2.945438129366458, + "learning_rate": 3.9520852879421254e-06, + "loss": 0.807, + "num_input_tokens_seen": 4827680, + "step": 234 + }, + { + "epoch": 0.028257079300186377, + "flos": 23182191479280.0, + "grad_norm": 2.938526866884137, + "learning_rate": 3.955174614372137e-06, + "loss": 0.7992, + "num_input_tokens_seen": 4847165, + "step": 235 + }, + { + "epoch": 0.028377322190825467, + "flos": 17214553430760.0, + "grad_norm": 3.173806490310878, + "learning_rate": 3.9582508226037045e-06, + "loss": 0.8347, + "num_input_tokens_seen": 4867025, + "step": 236 + }, + { + "epoch": 0.02849756508146456, + "flos": 14968620391680.0, + "grad_norm": 3.351688836124144, + "learning_rate": 3.9613140235734636e-06, + "loss": 0.9091, + "num_input_tokens_seen": 4883725, + "step": 237 + }, + { + "epoch": 0.02861780797210365, + "flos": 10383881956320.0, + "grad_norm": 7.9579739581015385, + "learning_rate": 3.96436432681674e-06, + "loss": 0.8029, + "num_input_tokens_seen": 4901435, + "step": 238 + }, + { + "epoch": 0.02873805086274274, + "flos": 18912669029280.0, + "grad_norm": 3.198206703047425, + "learning_rate": 3.967401840491044e-06, + "loss": 0.8818, + "num_input_tokens_seen": 4920435, + "step": 239 + }, + { + "epoch": 0.028858293753381833, + "flos": 12618480729720.0, + "grad_norm": 2.7225222670920455, + "learning_rate": 3.97042667139909e-06, + "loss": 0.8653, + "num_input_tokens_seen": 4937480, + "step": 240 + }, + { + "epoch": 0.028978536644020923, + "flos": 16862932982760.0, + "grad_norm": 2.62186062505371, + "learning_rate": 3.973438925011327e-06, + "loss": 0.8649, + "num_input_tokens_seen": 4955485, + "step": 241 + }, + { + "epoch": 0.029098779534660012, + "flos": 20779030367040.0, + "grad_norm": 3.3239515042491368, + "learning_rate": 3.976438705488002e-06, + "loss": 0.9123, + "num_input_tokens_seen": 4974865, + "step": 242 + }, + { + "epoch": 0.029219022425299106, + "flos": 10086549357000.0, + "grad_norm": 2.8760294522407537, + "learning_rate": 3.9794261157007744e-06, + "loss": 0.9177, + "num_input_tokens_seen": 4992340, + "step": 243 + }, + { + "epoch": 0.029339265315938196, + "flos": 14186433029520.0, + "grad_norm": 2.677344451717199, + "learning_rate": 3.982401257253887e-06, + "loss": 0.8182, + "num_input_tokens_seen": 5010400, + "step": 244 + }, + { + "epoch": 0.029459508206577285, + "flos": 11409715608360.0, + "grad_norm": 41.92681321082757, + "learning_rate": 3.985364230504893e-06, + "loss": 0.8819, + "num_input_tokens_seen": 5028005, + "step": 245 + }, + { + "epoch": 0.02957975109721638, + "flos": 20698659202680.0, + "grad_norm": 2.7020175848387953, + "learning_rate": 3.988315134584976e-06, + "loss": 0.8248, + "num_input_tokens_seen": 5047405, + "step": 246 + }, + { + "epoch": 0.02969999398785547, + "flos": 17808458790360.0, + "grad_norm": 2.3090089789241177, + "learning_rate": 3.991254067418851e-06, + "loss": 0.7936, + "num_input_tokens_seen": 5067665, + "step": 247 + }, + { + "epoch": 0.02982023687849456, + "flos": 25773507954240.0, + "grad_norm": 4.437955376268776, + "learning_rate": 3.994181125744254e-06, + "loss": 0.8211, + "num_input_tokens_seen": 5089190, + "step": 248 + }, + { + "epoch": 0.02994047976913365, + "flos": 19186007870280.0, + "grad_norm": 5.082189441655439, + "learning_rate": 3.99709640513106e-06, + "loss": 0.7378, + "num_input_tokens_seen": 5109790, + "step": 249 + }, + { + "epoch": 0.03006072265977274, + "flos": 18776711957880.0, + "grad_norm": 3.074869050750008, + "learning_rate": 4e-06, + "loss": 0.8389, + "num_input_tokens_seen": 5129345, + "step": 250 + }, + { + "epoch": 0.03018096555041183, + "flos": 16188656458800.0, + "grad_norm": 8.217389012194468, + "learning_rate": 3.999999848300794e-06, + "loss": 0.8805, + "num_input_tokens_seen": 5148050, + "step": 251 + }, + { + "epoch": 0.030301208441050925, + "flos": 22154109970080.0, + "grad_norm": 1.9718402647897704, + "learning_rate": 3.999999393203203e-06, + "loss": 0.8776, + "num_input_tokens_seen": 5170180, + "step": 252 + }, + { + "epoch": 0.030421451331690014, + "flos": 17294323055880.0, + "grad_norm": 2.5577961617648817, + "learning_rate": 3.999998634707293e-06, + "loss": 0.8426, + "num_input_tokens_seen": 5189450, + "step": 253 + }, + { + "epoch": 0.030541694222329104, + "flos": 20482077687240.0, + "grad_norm": 3.618689338557774, + "learning_rate": 3.999997572813182e-06, + "loss": 0.961, + "num_input_tokens_seen": 5206980, + "step": 254 + }, + { + "epoch": 0.030661937112968194, + "flos": 13569895289880.0, + "grad_norm": 3.0540265004745106, + "learning_rate": 3.999996207521028e-06, + "loss": 0.864, + "num_input_tokens_seen": 5225410, + "step": 255 + }, + { + "epoch": 0.030782180003607287, + "flos": 9411164734440.0, + "grad_norm": 20.414939257148106, + "learning_rate": 3.999994538831039e-06, + "loss": 0.8008, + "num_input_tokens_seen": 5241715, + "step": 256 + }, + { + "epoch": 0.030902422894246377, + "flos": 17083471993200.0, + "grad_norm": 3.7829989337319656, + "learning_rate": 3.99999256674347e-06, + "loss": 0.8289, + "num_input_tokens_seen": 5261585, + "step": 257 + }, + { + "epoch": 0.031022665784885467, + "flos": 40589395987080.0, + "grad_norm": 0.8942019926904305, + "learning_rate": 3.999990291258618e-06, + "loss": 0.5527, + "num_input_tokens_seen": 5319995, + "step": 258 + }, + { + "epoch": 0.03114290867552456, + "flos": 14109607780680.0, + "grad_norm": 2.9059479259070518, + "learning_rate": 3.999987712376829e-06, + "loss": 0.8535, + "num_input_tokens_seen": 5338035, + "step": 259 + }, + { + "epoch": 0.031263151566163654, + "flos": 15324514934280.0, + "grad_norm": 2.8035816785991203, + "learning_rate": 3.999984830098494e-06, + "loss": 0.7955, + "num_input_tokens_seen": 5357335, + "step": 260 + }, + { + "epoch": 0.03138339445680274, + "flos": 10760889200880.0, + "grad_norm": 7.129473971401167, + "learning_rate": 3.999981644424051e-06, + "loss": 0.9483, + "num_input_tokens_seen": 5371855, + "step": 261 + }, + { + "epoch": 0.03150363734744183, + "flos": 8439049051800.0, + "grad_norm": 3.003857199239886, + "learning_rate": 3.999978155353982e-06, + "loss": 0.8583, + "num_input_tokens_seen": 5388720, + "step": 262 + }, + { + "epoch": 0.03162388023808092, + "flos": 24773662638000.0, + "grad_norm": 2.622504183144956, + "learning_rate": 3.9999743628888186e-06, + "loss": 0.7909, + "num_input_tokens_seen": 5410230, + "step": 263 + }, + { + "epoch": 0.03174412312872001, + "flos": 15214261259040.0, + "grad_norm": 2.7668018315707226, + "learning_rate": 3.999970267029133e-06, + "loss": 0.8588, + "num_input_tokens_seen": 5428910, + "step": 264 + }, + { + "epoch": 0.0318643660193591, + "flos": 17372762962680.0, + "grad_norm": 2.936468064877417, + "learning_rate": 3.999965867775548e-06, + "loss": 0.7891, + "num_input_tokens_seen": 5449025, + "step": 265 + }, + { + "epoch": 0.0319846089099982, + "flos": 10113297531840.0, + "grad_norm": 3.3726323617986536, + "learning_rate": 3.9999611651287315e-06, + "loss": 0.8473, + "num_input_tokens_seen": 5466900, + "step": 266 + }, + { + "epoch": 0.03210485180063729, + "flos": 10731006690000.0, + "grad_norm": 10.731201413488401, + "learning_rate": 3.999956159089396e-06, + "loss": 0.7871, + "num_input_tokens_seen": 5484070, + "step": 267 + }, + { + "epoch": 0.03222509469127638, + "flos": 21210768699720.0, + "grad_norm": 4.431924675504504, + "learning_rate": 3.999950849658302e-06, + "loss": 0.7761, + "num_input_tokens_seen": 5502710, + "step": 268 + }, + { + "epoch": 0.03234533758191547, + "flos": 12354639876720.0, + "grad_norm": 4.8889826250241555, + "learning_rate": 3.999945236836254e-06, + "loss": 0.8316, + "num_input_tokens_seen": 5521395, + "step": 269 + }, + { + "epoch": 0.03246558047255456, + "flos": 13866309750360.0, + "grad_norm": 3.053766370269581, + "learning_rate": 3.999939320624103e-06, + "loss": 0.9184, + "num_input_tokens_seen": 5536265, + "step": 270 + }, + { + "epoch": 0.03258582336319365, + "flos": 17373111222240.0, + "grad_norm": 6.029000767145758, + "learning_rate": 3.999933101022749e-06, + "loss": 0.8857, + "num_input_tokens_seen": 5556390, + "step": 271 + }, + { + "epoch": 0.032706066253832745, + "flos": 20291104588080.0, + "grad_norm": 2.562809809122951, + "learning_rate": 3.999926578033132e-06, + "loss": 0.8539, + "num_input_tokens_seen": 5575925, + "step": 272 + }, + { + "epoch": 0.032826309144471835, + "flos": 33578574628440.0, + "grad_norm": 2.719724984570907, + "learning_rate": 3.999919751656244e-06, + "loss": 0.6314, + "num_input_tokens_seen": 5602545, + "step": 273 + }, + { + "epoch": 0.032946552035110925, + "flos": 18916056645000.0, + "grad_norm": 3.1808602984348715, + "learning_rate": 3.9999126218931195e-06, + "loss": 0.764, + "num_input_tokens_seen": 5620300, + "step": 274 + }, + { + "epoch": 0.033066794925750015, + "flos": 11004440510880.0, + "grad_norm": 3.617452722964166, + "learning_rate": 3.99990518874484e-06, + "loss": 0.8904, + "num_input_tokens_seen": 5636460, + "step": 275 + }, + { + "epoch": 0.033187037816389105, + "flos": 16668318988200.0, + "grad_norm": 3.169077273591039, + "learning_rate": 3.999897452212534e-06, + "loss": 0.9113, + "num_input_tokens_seen": 5653510, + "step": 276 + }, + { + "epoch": 0.033307280707028195, + "flos": 17079989397600.0, + "grad_norm": 2.313267451602348, + "learning_rate": 3.999889412297374e-06, + "loss": 0.9829, + "num_input_tokens_seen": 5672655, + "step": 277 + }, + { + "epoch": 0.03342752359766729, + "flos": 21156860770560.0, + "grad_norm": 3.1873402274984826, + "learning_rate": 3.999881069000581e-06, + "loss": 0.7748, + "num_input_tokens_seen": 5692105, + "step": 278 + }, + { + "epoch": 0.03354776648830638, + "flos": 14158133516640.0, + "grad_norm": 2.924069972410415, + "learning_rate": 3.99987242232342e-06, + "loss": 0.8616, + "num_input_tokens_seen": 5706830, + "step": 279 + }, + { + "epoch": 0.03366800937894547, + "flos": 13029771219600.0, + "grad_norm": 2.628388080288801, + "learning_rate": 3.9998634722672026e-06, + "loss": 0.7867, + "num_input_tokens_seen": 5726605, + "step": 280 + }, + { + "epoch": 0.03378825226958456, + "flos": 26418725126280.0, + "grad_norm": 2.0372591777374303, + "learning_rate": 3.999854218833286e-06, + "loss": 0.7728, + "num_input_tokens_seen": 5747145, + "step": 281 + }, + { + "epoch": 0.03390849516022365, + "flos": 18834672361920.0, + "grad_norm": 2.5193617652842595, + "learning_rate": 3.999844662023075e-06, + "loss": 0.8081, + "num_input_tokens_seen": 5766740, + "step": 282 + }, + { + "epoch": 0.03402873805086274, + "flos": 15564615308640.0, + "grad_norm": 1.9462465548622643, + "learning_rate": 3.999834801838018e-06, + "loss": 0.9084, + "num_input_tokens_seen": 5785440, + "step": 283 + }, + { + "epoch": 0.03414898094150183, + "flos": 16620743051040.0, + "grad_norm": 4.65276436621914, + "learning_rate": 3.9998246382796115e-06, + "loss": 0.7357, + "num_input_tokens_seen": 5804740, + "step": 284 + }, + { + "epoch": 0.03426922383214093, + "flos": 13785558666480.0, + "grad_norm": 3.9334931346522577, + "learning_rate": 3.999814171349399e-06, + "loss": 0.8912, + "num_input_tokens_seen": 5822320, + "step": 285 + }, + { + "epoch": 0.03438946672278002, + "flos": 25532077861560.0, + "grad_norm": 2.325005918925965, + "learning_rate": 3.9998034010489655e-06, + "loss": 0.7275, + "num_input_tokens_seen": 5845730, + "step": 286 + }, + { + "epoch": 0.03450970961341911, + "flos": 16216132812720.0, + "grad_norm": 3.159568175495352, + "learning_rate": 3.999792327379946e-06, + "loss": 0.7549, + "num_input_tokens_seen": 5864825, + "step": 287 + }, + { + "epoch": 0.034629952504058197, + "flos": 15891640459080.0, + "grad_norm": 2.5296105621117384, + "learning_rate": 3.999780950344021e-06, + "loss": 0.9601, + "num_input_tokens_seen": 5882735, + "step": 288 + }, + { + "epoch": 0.034750195394697286, + "flos": 14650080110520.0, + "grad_norm": 2.3970485977510902, + "learning_rate": 3.999769269942916e-06, + "loss": 0.8034, + "num_input_tokens_seen": 5902495, + "step": 289 + }, + { + "epoch": 0.034870438285336376, + "flos": 20509902300720.0, + "grad_norm": 2.24707731521888, + "learning_rate": 3.999757286178402e-06, + "loss": 0.806, + "num_input_tokens_seen": 5924650, + "step": 290 + }, + { + "epoch": 0.03499068117597547, + "flos": 16107557115360.0, + "grad_norm": 2.104343432061367, + "learning_rate": 3.999744999052299e-06, + "loss": 0.8942, + "num_input_tokens_seen": 5945760, + "step": 291 + }, + { + "epoch": 0.03511092406661456, + "flos": 42178397668920.0, + "grad_norm": 0.9622534448488158, + "learning_rate": 3.9997324085664675e-06, + "loss": 0.6571, + "num_input_tokens_seen": 6005710, + "step": 292 + }, + { + "epoch": 0.03523116695725365, + "flos": 16781200440120.0, + "grad_norm": 2.6475382040435447, + "learning_rate": 3.999719514722821e-06, + "loss": 0.9122, + "num_input_tokens_seen": 6025560, + "step": 293 + }, + { + "epoch": 0.03535140984789274, + "flos": 27123928939800.0, + "grad_norm": 3.0616995874862822, + "learning_rate": 3.999706317523314e-06, + "loss": 0.747, + "num_input_tokens_seen": 6043840, + "step": 294 + }, + { + "epoch": 0.03547165273853183, + "flos": 14946937810440.0, + "grad_norm": 2.672730872063935, + "learning_rate": 3.999692816969948e-06, + "loss": 0.8407, + "num_input_tokens_seen": 6063095, + "step": 295 + }, + { + "epoch": 0.03559189562917092, + "flos": 51512420554920.0, + "grad_norm": 0.9458578759144756, + "learning_rate": 3.999679013064772e-06, + "loss": 0.7015, + "num_input_tokens_seen": 6129560, + "step": 296 + }, + { + "epoch": 0.03571213851981002, + "flos": 15836909370960.0, + "grad_norm": 4.026125941149152, + "learning_rate": 3.99966490580988e-06, + "loss": 0.8487, + "num_input_tokens_seen": 6146640, + "step": 297 + }, + { + "epoch": 0.03583238141044911, + "flos": 32335842861360.0, + "grad_norm": 7.367363289508643, + "learning_rate": 3.999650495207411e-06, + "loss": 0.654, + "num_input_tokens_seen": 6172385, + "step": 298 + }, + { + "epoch": 0.0359526243010882, + "flos": 13807621167240.0, + "grad_norm": 3.6305584877765713, + "learning_rate": 3.999635781259553e-06, + "loss": 0.894, + "num_input_tokens_seen": 6187370, + "step": 299 + }, + { + "epoch": 0.03607286719172729, + "flos": 45456179752440.0, + "grad_norm": 0.9302988046106397, + "learning_rate": 3.999620763968535e-06, + "loss": 0.5524, + "num_input_tokens_seen": 6245965, + "step": 300 + }, + { + "epoch": 0.03619311008236638, + "flos": 20401136643600.0, + "grad_norm": 2.057269080959554, + "learning_rate": 3.999605443336638e-06, + "loss": 0.8677, + "num_input_tokens_seen": 6267815, + "step": 301 + }, + { + "epoch": 0.03631335297300547, + "flos": 9893011801080.0, + "grad_norm": 2.9757424974709896, + "learning_rate": 3.999589819366185e-06, + "loss": 0.8891, + "num_input_tokens_seen": 6281325, + "step": 302 + }, + { + "epoch": 0.036433595863644565, + "flos": 20261855276400.0, + "grad_norm": 2.4982691495190665, + "learning_rate": 3.999573892059547e-06, + "loss": 0.8226, + "num_input_tokens_seen": 6300175, + "step": 303 + }, + { + "epoch": 0.036553838754283655, + "flos": 17998133831160.0, + "grad_norm": 2.7999853814969606, + "learning_rate": 3.999557661419138e-06, + "loss": 0.8016, + "num_input_tokens_seen": 6320045, + "step": 304 + }, + { + "epoch": 0.036674081644922744, + "flos": 17133928986720.0, + "grad_norm": 2.224983592455085, + "learning_rate": 3.9995411274474225e-06, + "loss": 0.7912, + "num_input_tokens_seen": 6339045, + "step": 305 + }, + { + "epoch": 0.036794324535561834, + "flos": 20157237074040.0, + "grad_norm": 3.262334744883011, + "learning_rate": 3.999524290146908e-06, + "loss": 0.8194, + "num_input_tokens_seen": 6358970, + "step": 306 + }, + { + "epoch": 0.036914567426200924, + "flos": 14217391979040.0, + "grad_norm": 2.781784006351968, + "learning_rate": 3.9995071495201485e-06, + "loss": 0.8896, + "num_input_tokens_seen": 6375795, + "step": 307 + }, + { + "epoch": 0.037034810316840014, + "flos": 16323505431600.0, + "grad_norm": 3.3225322887726048, + "learning_rate": 3.999489705569744e-06, + "loss": 0.9472, + "num_input_tokens_seen": 6393215, + "step": 308 + }, + { + "epoch": 0.03715505320747911, + "flos": 13569800310000.0, + "grad_norm": 4.057253927594796, + "learning_rate": 3.999471958298341e-06, + "loss": 0.8499, + "num_input_tokens_seen": 6411845, + "step": 309 + }, + { + "epoch": 0.0372752960981182, + "flos": 26422524321480.0, + "grad_norm": 4.434397727902687, + "learning_rate": 3.999453907708631e-06, + "loss": 0.7586, + "num_input_tokens_seen": 6433970, + "step": 310 + }, + { + "epoch": 0.03739553898875729, + "flos": 15216984015600.0, + "grad_norm": 2.4282594267996886, + "learning_rate": 3.999435553803353e-06, + "loss": 0.8129, + "num_input_tokens_seen": 6453090, + "step": 311 + }, + { + "epoch": 0.03751578187939638, + "flos": 14809651020720.0, + "grad_norm": 3.642676434455871, + "learning_rate": 3.999416896585292e-06, + "loss": 0.8361, + "num_input_tokens_seen": 6469840, + "step": 312 + }, + { + "epoch": 0.03763602477003547, + "flos": 15108946537560.0, + "grad_norm": 3.616960753219757, + "learning_rate": 3.9993979360572775e-06, + "loss": 0.8467, + "num_input_tokens_seen": 6489700, + "step": 313 + }, + { + "epoch": 0.03775626766067456, + "flos": 12165946294680.0, + "grad_norm": 3.5178251036802566, + "learning_rate": 3.999378672222185e-06, + "loss": 0.8049, + "num_input_tokens_seen": 6507205, + "step": 314 + }, + { + "epoch": 0.03787651055131366, + "flos": 15459458886960.0, + "grad_norm": 2.357926789503742, + "learning_rate": 3.9993591050829385e-06, + "loss": 0.824, + "num_input_tokens_seen": 6524790, + "step": 315 + }, + { + "epoch": 0.037996753441952746, + "flos": 16108063674720.0, + "grad_norm": 2.5220471026796396, + "learning_rate": 3.999339234642506e-06, + "loss": 0.7916, + "num_input_tokens_seen": 6544260, + "step": 316 + }, + { + "epoch": 0.038116996332591836, + "flos": 20319942320280.0, + "grad_norm": 3.79945304849427, + "learning_rate": 3.9993190609038994e-06, + "loss": 0.8287, + "num_input_tokens_seen": 6562745, + "step": 317 + }, + { + "epoch": 0.038237239223230926, + "flos": 15486270381720.0, + "grad_norm": 2.830790407044167, + "learning_rate": 3.999298583870182e-06, + "loss": 0.8335, + "num_input_tokens_seen": 6582050, + "step": 318 + }, + { + "epoch": 0.038357482113870016, + "flos": 18726729863760.0, + "grad_norm": 2.7882460139028997, + "learning_rate": 3.999277803544458e-06, + "loss": 0.7592, + "num_input_tokens_seen": 6601925, + "step": 319 + }, + { + "epoch": 0.038477725004509106, + "flos": 43649900036040.0, + "grad_norm": 0.9438276372642573, + "learning_rate": 3.999256719929882e-06, + "loss": 0.6623, + "num_input_tokens_seen": 6662920, + "step": 320 + }, + { + "epoch": 0.0385979678951482, + "flos": 49635738070200.0, + "grad_norm": 1.255009341262157, + "learning_rate": 3.999235333029651e-06, + "loss": 0.7441, + "num_input_tokens_seen": 6716580, + "step": 321 + }, + { + "epoch": 0.03871821078578729, + "flos": 16647174626280.0, + "grad_norm": 2.9061768225204196, + "learning_rate": 3.999213642847009e-06, + "loss": 0.8141, + "num_input_tokens_seen": 6736885, + "step": 322 + }, + { + "epoch": 0.03883845367642638, + "flos": 19261725020520.0, + "grad_norm": 2.1544046660334857, + "learning_rate": 3.999191649385247e-06, + "loss": 0.9105, + "num_input_tokens_seen": 6757780, + "step": 323 + }, + { + "epoch": 0.03895869656706547, + "flos": 41973467018760.0, + "grad_norm": 0.8670651117512483, + "learning_rate": 3.999169352647702e-06, + "loss": 0.6279, + "num_input_tokens_seen": 6818680, + "step": 324 + }, + { + "epoch": 0.03907893945770456, + "flos": 18213670567920.0, + "grad_norm": 4.741341982129522, + "learning_rate": 3.999146752637755e-06, + "loss": 0.8063, + "num_input_tokens_seen": 6839445, + "step": 325 + }, + { + "epoch": 0.03919918234834365, + "flos": 13407063403800.0, + "grad_norm": 2.885942119766115, + "learning_rate": 3.999123849358836e-06, + "loss": 0.8915, + "num_input_tokens_seen": 6856830, + "step": 326 + }, + { + "epoch": 0.03931942523898275, + "flos": 18481215636240.0, + "grad_norm": 2.3250126869101626, + "learning_rate": 3.999100642814418e-06, + "loss": 0.7334, + "num_input_tokens_seen": 6876990, + "step": 327 + }, + { + "epoch": 0.03943966812962184, + "flos": 17025511589160.0, + "grad_norm": 3.4197284811555813, + "learning_rate": 3.999077133008022e-06, + "loss": 0.8795, + "num_input_tokens_seen": 6895295, + "step": 328 + }, + { + "epoch": 0.03955991102026093, + "flos": 21400032161040.0, + "grad_norm": 2.8641875577388016, + "learning_rate": 3.9990533199432145e-06, + "loss": 0.8937, + "num_input_tokens_seen": 6916510, + "step": 329 + }, + { + "epoch": 0.03968015391090002, + "flos": 12840064518840.0, + "grad_norm": 2.9814993700061665, + "learning_rate": 3.999029203623608e-06, + "loss": 0.7498, + "num_input_tokens_seen": 6933950, + "step": 330 + }, + { + "epoch": 0.03980039680153911, + "flos": 15946149927480.0, + "grad_norm": 2.2320796674231036, + "learning_rate": 3.99900478405286e-06, + "loss": 0.8549, + "num_input_tokens_seen": 6952980, + "step": 331 + }, + { + "epoch": 0.0399206396921782, + "flos": 11058601719720.0, + "grad_norm": 2.688839955405701, + "learning_rate": 3.998980061234676e-06, + "loss": 0.7976, + "num_input_tokens_seen": 6970615, + "step": 332 + }, + { + "epoch": 0.040040882582817294, + "flos": 10486252301520.0, + "grad_norm": 7.009411615555835, + "learning_rate": 3.9989550351728055e-06, + "loss": 0.7495, + "num_input_tokens_seen": 6987265, + "step": 333 + }, + { + "epoch": 0.040161125473456384, + "flos": 14082258066600.0, + "grad_norm": 2.8751223994198067, + "learning_rate": 3.998929705871046e-06, + "loss": 0.8232, + "num_input_tokens_seen": 7004340, + "step": 334 + }, + { + "epoch": 0.040281368364095474, + "flos": 13001851626240.0, + "grad_norm": 3.6134627013138827, + "learning_rate": 3.99890407333324e-06, + "loss": 0.8777, + "num_input_tokens_seen": 7022590, + "step": 335 + }, + { + "epoch": 0.040401611254734564, + "flos": 14299124521680.0, + "grad_norm": 4.0064054699991445, + "learning_rate": 3.998878137563275e-06, + "loss": 0.8525, + "num_input_tokens_seen": 7041860, + "step": 336 + }, + { + "epoch": 0.040521854145373654, + "flos": 16134970149360.0, + "grad_norm": 2.955615250104931, + "learning_rate": 3.998851898565085e-06, + "loss": 0.8285, + "num_input_tokens_seen": 7061385, + "step": 337 + }, + { + "epoch": 0.04064209703601274, + "flos": 16593931556280.0, + "grad_norm": 2.798799758203029, + "learning_rate": 3.998825356342653e-06, + "loss": 0.8296, + "num_input_tokens_seen": 7081280, + "step": 338 + }, + { + "epoch": 0.04076233992665183, + "flos": 28366534067040.0, + "grad_norm": 3.132019551446333, + "learning_rate": 3.998798510900003e-06, + "loss": 0.7103, + "num_input_tokens_seen": 7103800, + "step": 339 + }, + { + "epoch": 0.04088258281729093, + "flos": 18970027894080.0, + "grad_norm": 3.5833860142110368, + "learning_rate": 3.998771362241207e-06, + "loss": 0.8302, + "num_input_tokens_seen": 7123925, + "step": 340 + }, + { + "epoch": 0.04100282570793002, + "flos": 14458632111960.0, + "grad_norm": 3.191806950154671, + "learning_rate": 3.998743910370385e-06, + "loss": 0.8631, + "num_input_tokens_seen": 7142505, + "step": 341 + }, + { + "epoch": 0.04112306859856911, + "flos": 16512895532760.0, + "grad_norm": 3.532527726180004, + "learning_rate": 3.998716155291702e-06, + "loss": 0.7229, + "num_input_tokens_seen": 7160065, + "step": 342 + }, + { + "epoch": 0.0412433114892082, + "flos": 18344150466240.0, + "grad_norm": 2.227326165126184, + "learning_rate": 3.998688097009366e-06, + "loss": 0.8911, + "num_input_tokens_seen": 7180550, + "step": 343 + }, + { + "epoch": 0.04136355437984729, + "flos": 18456367059000.0, + "grad_norm": 3.0494231694805727, + "learning_rate": 3.998659735527636e-06, + "loss": 0.795, + "num_input_tokens_seen": 7199360, + "step": 344 + }, + { + "epoch": 0.04148379727048638, + "flos": 16810164812160.0, + "grad_norm": 2.182072254400686, + "learning_rate": 3.998631070850813e-06, + "loss": 0.762, + "num_input_tokens_seen": 7219700, + "step": 345 + }, + { + "epoch": 0.041604040161125476, + "flos": 10221018410280.0, + "grad_norm": 3.371242267966111, + "learning_rate": 3.9986021029832455e-06, + "loss": 0.8241, + "num_input_tokens_seen": 7236735, + "step": 346 + }, + { + "epoch": 0.041724283051764566, + "flos": 8761705127760.0, + "grad_norm": 4.441958699219444, + "learning_rate": 3.9985728319293285e-06, + "loss": 0.9077, + "num_input_tokens_seen": 7250430, + "step": 347 + }, + { + "epoch": 0.041844525942403656, + "flos": 8979078142200.0, + "grad_norm": 3.1724349406179497, + "learning_rate": 3.998543257693501e-06, + "loss": 0.8394, + "num_input_tokens_seen": 7266905, + "step": 348 + }, + { + "epoch": 0.041964768833042745, + "flos": 17404101831720.0, + "grad_norm": 2.2346626576741992, + "learning_rate": 3.998513380280251e-06, + "loss": 0.8709, + "num_input_tokens_seen": 7286905, + "step": 349 + }, + { + "epoch": 0.042085011723681835, + "flos": 8601785958000.0, + "grad_norm": 2.8870116047852235, + "learning_rate": 3.99848319969411e-06, + "loss": 0.9399, + "num_input_tokens_seen": 7304225, + "step": 350 + }, + { + "epoch": 0.042205254614320925, + "flos": 12301111867080.0, + "grad_norm": 2.9539222246700088, + "learning_rate": 3.9984527159396564e-06, + "loss": 0.7701, + "num_input_tokens_seen": 7322585, + "step": 351 + }, + { + "epoch": 0.04232549750496002, + "flos": 18402459129840.0, + "grad_norm": 6.043488571998217, + "learning_rate": 3.9984219290215154e-06, + "loss": 0.8239, + "num_input_tokens_seen": 7342480, + "step": 352 + }, + { + "epoch": 0.04244574039559911, + "flos": 19591124667960.0, + "grad_norm": 2.9407238415637242, + "learning_rate": 3.998390838944356e-06, + "loss": 0.8849, + "num_input_tokens_seen": 7363705, + "step": 353 + }, + { + "epoch": 0.0425659832862382, + "flos": 15297766759440.0, + "grad_norm": 2.63691215653517, + "learning_rate": 3.998359445712895e-06, + "loss": 0.8723, + "num_input_tokens_seen": 7382530, + "step": 354 + }, + { + "epoch": 0.04268622617687729, + "flos": 17079736117920.0, + "grad_norm": 3.3820637850078783, + "learning_rate": 3.9983277493318955e-06, + "loss": 0.7897, + "num_input_tokens_seen": 7401545, + "step": 355 + }, + { + "epoch": 0.04280646906751638, + "flos": 19050367398480.0, + "grad_norm": 1.9513070152305578, + "learning_rate": 3.998295749806165e-06, + "loss": 0.799, + "num_input_tokens_seen": 7422490, + "step": 356 + }, + { + "epoch": 0.04292671195815547, + "flos": 19726005300720.0, + "grad_norm": 2.919925957616293, + "learning_rate": 3.998263447140558e-06, + "loss": 0.8247, + "num_input_tokens_seen": 7442410, + "step": 357 + }, + { + "epoch": 0.04304695484879457, + "flos": 29014125736080.0, + "grad_norm": 2.7419632605352766, + "learning_rate": 3.998230841339976e-06, + "loss": 0.8102, + "num_input_tokens_seen": 7464140, + "step": 358 + }, + { + "epoch": 0.04316719773943366, + "flos": 14352747511200.0, + "grad_norm": 2.749749519038043, + "learning_rate": 3.998197932409363e-06, + "loss": 0.8357, + "num_input_tokens_seen": 7481870, + "step": 359 + }, + { + "epoch": 0.04328744063007275, + "flos": 16428915132960.0, + "grad_norm": 2.369325886878722, + "learning_rate": 3.9981647203537125e-06, + "loss": 0.8509, + "num_input_tokens_seen": 7499090, + "step": 360 + }, + { + "epoch": 0.04340768352071184, + "flos": 15563317250280.0, + "grad_norm": 2.9287251023436216, + "learning_rate": 3.998131205178063e-06, + "loss": 0.9439, + "num_input_tokens_seen": 7517280, + "step": 361 + }, + { + "epoch": 0.04352792641135093, + "flos": 8385995941560.0, + "grad_norm": 3.0932128131616365, + "learning_rate": 3.998097386887498e-06, + "loss": 0.7438, + "num_input_tokens_seen": 7534075, + "step": 362 + }, + { + "epoch": 0.04364816930199002, + "flos": 17296222653480.0, + "grad_norm": 2.1088264034519866, + "learning_rate": 3.998063265487148e-06, + "loss": 0.8298, + "num_input_tokens_seen": 7554845, + "step": 363 + }, + { + "epoch": 0.043768412192629114, + "flos": 10492267693920.0, + "grad_norm": 2.2745619677045466, + "learning_rate": 3.99802884098219e-06, + "loss": 0.8035, + "num_input_tokens_seen": 7572675, + "step": 364 + }, + { + "epoch": 0.043888655083268203, + "flos": 19401892866600.0, + "grad_norm": 3.290296744324203, + "learning_rate": 3.997994113377845e-06, + "loss": 0.8084, + "num_input_tokens_seen": 7591295, + "step": 365 + }, + { + "epoch": 0.04400889797390729, + "flos": 19969081711320.0, + "grad_norm": 2.4285069384947993, + "learning_rate": 3.9979590826793815e-06, + "loss": 0.8267, + "num_input_tokens_seen": 7612205, + "step": 366 + }, + { + "epoch": 0.04412914086454638, + "flos": 14702816621160.0, + "grad_norm": 2.3919933118184464, + "learning_rate": 3.997923748892113e-06, + "loss": 0.7869, + "num_input_tokens_seen": 7631245, + "step": 367 + }, + { + "epoch": 0.04424938375518547, + "flos": 16542524763960.0, + "grad_norm": 2.5264469306667303, + "learning_rate": 3.9978881120214015e-06, + "loss": 0.8808, + "num_input_tokens_seen": 7652485, + "step": 368 + }, + { + "epoch": 0.04436962664582456, + "flos": 17672501718960.0, + "grad_norm": 2.3596956148893398, + "learning_rate": 3.997852172072652e-06, + "loss": 0.796, + "num_input_tokens_seen": 7673420, + "step": 369 + }, + { + "epoch": 0.04448986953646366, + "flos": 13299279205440.0, + "grad_norm": 4.951264522370361, + "learning_rate": 3.9978159290513155e-06, + "loss": 0.8749, + "num_input_tokens_seen": 7691220, + "step": 370 + }, + { + "epoch": 0.04461011242710275, + "flos": 22102481558040.0, + "grad_norm": 2.1368526437710686, + "learning_rate": 3.997779382962892e-06, + "loss": 0.8006, + "num_input_tokens_seen": 7713825, + "step": 371 + }, + { + "epoch": 0.04473035531774184, + "flos": 21831992113440.0, + "grad_norm": 2.699586790749613, + "learning_rate": 3.997742533812924e-06, + "loss": 0.7248, + "num_input_tokens_seen": 7736810, + "step": 372 + }, + { + "epoch": 0.04485059820838093, + "flos": 9546172007040.0, + "grad_norm": 5.873557705066504, + "learning_rate": 3.997705381607001e-06, + "loss": 0.9, + "num_input_tokens_seen": 7753345, + "step": 373 + }, + { + "epoch": 0.04497084109902002, + "flos": 50211886683600.0, + "grad_norm": 0.9888860789240038, + "learning_rate": 3.997667926350761e-06, + "loss": 0.6274, + "num_input_tokens_seen": 7811395, + "step": 374 + }, + { + "epoch": 0.04509108398965911, + "flos": 47436055741320.0, + "grad_norm": 0.854726286292233, + "learning_rate": 3.997630168049886e-06, + "loss": 0.5973, + "num_input_tokens_seen": 7869480, + "step": 375 + }, + { + "epoch": 0.045211326880298205, + "flos": 16294984299000.0, + "grad_norm": 2.7209214056539466, + "learning_rate": 3.997592106710101e-06, + "loss": 0.7695, + "num_input_tokens_seen": 7888660, + "step": 376 + }, + { + "epoch": 0.045331569770937295, + "flos": 23613106653000.0, + "grad_norm": 4.09698544840036, + "learning_rate": 3.997553742337182e-06, + "loss": 0.6556, + "num_input_tokens_seen": 7907805, + "step": 377 + }, + { + "epoch": 0.045451812661576385, + "flos": 16215562933440.0, + "grad_norm": 1.9163818631523601, + "learning_rate": 3.997515074936949e-06, + "loss": 0.9042, + "num_input_tokens_seen": 7928400, + "step": 378 + }, + { + "epoch": 0.045572055552215475, + "flos": 12085005251040.0, + "grad_norm": 2.486863431517755, + "learning_rate": 3.997476104515268e-06, + "loss": 0.8378, + "num_input_tokens_seen": 7946310, + "step": 379 + }, + { + "epoch": 0.045692298442854565, + "flos": 12840887677800.0, + "grad_norm": 2.108969077153549, + "learning_rate": 3.9974368310780485e-06, + "loss": 0.7632, + "num_input_tokens_seen": 7963205, + "step": 380 + }, + { + "epoch": 0.045812541333493655, + "flos": 19618189442400.0, + "grad_norm": 3.146625309634657, + "learning_rate": 3.997397254631251e-06, + "loss": 0.7247, + "num_input_tokens_seen": 7983545, + "step": 381 + }, + { + "epoch": 0.04593278422413275, + "flos": 44406985541280.0, + "grad_norm": 0.8747688737519179, + "learning_rate": 3.997357375180878e-06, + "loss": 0.6462, + "num_input_tokens_seen": 8047545, + "step": 382 + }, + { + "epoch": 0.04605302711477184, + "flos": 15946118267520.0, + "grad_norm": 2.401609429500908, + "learning_rate": 3.997317192732979e-06, + "loss": 0.7322, + "num_input_tokens_seen": 8066045, + "step": 383 + }, + { + "epoch": 0.04617327000541093, + "flos": 14214067683240.0, + "grad_norm": 1.9762514495361725, + "learning_rate": 3.99727670729365e-06, + "loss": 0.8107, + "num_input_tokens_seen": 8084325, + "step": 384 + }, + { + "epoch": 0.04629351289605002, + "flos": 18619420564800.0, + "grad_norm": 2.0796402304627546, + "learning_rate": 3.997235918869033e-06, + "loss": 0.7745, + "num_input_tokens_seen": 8105080, + "step": 385 + }, + { + "epoch": 0.04641375578668911, + "flos": 15027657234360.0, + "grad_norm": 2.890337460624884, + "learning_rate": 3.997194827465315e-06, + "loss": 0.8243, + "num_input_tokens_seen": 8123395, + "step": 386 + }, + { + "epoch": 0.0465339986773282, + "flos": 9573711680880.0, + "grad_norm": 4.662901971992874, + "learning_rate": 3.997153433088728e-06, + "loss": 0.9081, + "num_input_tokens_seen": 8140240, + "step": 387 + }, + { + "epoch": 0.0466542415679673, + "flos": 18726666543840.0, + "grad_norm": 2.4022117516454813, + "learning_rate": 3.997111735745554e-06, + "loss": 0.7949, + "num_input_tokens_seen": 8162930, + "step": 388 + }, + { + "epoch": 0.04677448445860639, + "flos": 16269850782120.0, + "grad_norm": 2.714747672816571, + "learning_rate": 3.997069735442118e-06, + "loss": 0.8196, + "num_input_tokens_seen": 8182345, + "step": 389 + }, + { + "epoch": 0.04689472734924548, + "flos": 20644497993840.0, + "grad_norm": 1.9415044932385876, + "learning_rate": 3.997027432184792e-06, + "loss": 0.7897, + "num_input_tokens_seen": 8206725, + "step": 390 + }, + { + "epoch": 0.04701497023988457, + "flos": 16944950465040.0, + "grad_norm": 2.732910897788335, + "learning_rate": 3.99698482597999e-06, + "loss": 0.8736, + "num_input_tokens_seen": 8224125, + "step": 391 + }, + { + "epoch": 0.04713521313052366, + "flos": 47796857577720.0, + "grad_norm": 1.1912933296096921, + "learning_rate": 3.99694191683418e-06, + "loss": 0.6526, + "num_input_tokens_seen": 8284645, + "step": 392 + }, + { + "epoch": 0.047255456021162746, + "flos": 13705535761680.0, + "grad_norm": 3.306510837123572, + "learning_rate": 3.996898704753867e-06, + "loss": 0.8118, + "num_input_tokens_seen": 8302315, + "step": 393 + }, + { + "epoch": 0.04737569891180184, + "flos": 16134400270080.0, + "grad_norm": 4.373631726606673, + "learning_rate": 3.996855189745609e-06, + "loss": 0.874, + "num_input_tokens_seen": 8321300, + "step": 394 + }, + { + "epoch": 0.04749594180244093, + "flos": 21317254839720.0, + "grad_norm": 2.5086656111668457, + "learning_rate": 3.996811371816007e-06, + "loss": 0.9184, + "num_input_tokens_seen": 8343445, + "step": 395 + }, + { + "epoch": 0.04761618469308002, + "flos": 25798039931880.0, + "grad_norm": 3.090072449757607, + "learning_rate": 3.996767250971707e-06, + "loss": 0.7621, + "num_input_tokens_seen": 8365905, + "step": 396 + }, + { + "epoch": 0.04773642758371911, + "flos": 18781714231560.0, + "grad_norm": 2.2291346633847993, + "learning_rate": 3.996722827219403e-06, + "loss": 0.845, + "num_input_tokens_seen": 8387240, + "step": 397 + }, + { + "epoch": 0.0478566704743582, + "flos": 15081881763120.0, + "grad_norm": 3.462369532141191, + "learning_rate": 3.996678100565833e-06, + "loss": 0.7992, + "num_input_tokens_seen": 8406015, + "step": 398 + }, + { + "epoch": 0.04797691336499729, + "flos": 13751750320560.0, + "grad_norm": 3.1393066979644213, + "learning_rate": 3.996633071017783e-06, + "loss": 0.8601, + "num_input_tokens_seen": 8422365, + "step": 399 + }, + { + "epoch": 0.04809715625563638, + "flos": 15427486818720.0, + "grad_norm": 2.6679267489030143, + "learning_rate": 3.996587738582084e-06, + "loss": 0.8043, + "num_input_tokens_seen": 8438885, + "step": 400 + }, + { + "epoch": 0.04821739914627548, + "flos": 17430944986440.0, + "grad_norm": 3.31959852732667, + "learning_rate": 3.9965421032656115e-06, + "loss": 0.8361, + "num_input_tokens_seen": 8458535, + "step": 401 + }, + { + "epoch": 0.04833764203691457, + "flos": 16243134267240.0, + "grad_norm": 4.1258719707527245, + "learning_rate": 3.99649616507529e-06, + "loss": 0.9225, + "num_input_tokens_seen": 8477350, + "step": 402 + }, + { + "epoch": 0.04845788492755366, + "flos": 48591166213200.0, + "grad_norm": 0.9387970217472649, + "learning_rate": 3.996449924018088e-06, + "loss": 0.6702, + "num_input_tokens_seen": 8537530, + "step": 403 + }, + { + "epoch": 0.04857812781819275, + "flos": 14082606326160.0, + "grad_norm": 2.511596202085503, + "learning_rate": 3.99640338010102e-06, + "loss": 0.7866, + "num_input_tokens_seen": 8556355, + "step": 404 + }, + { + "epoch": 0.04869837070883184, + "flos": 17621126586600.0, + "grad_norm": 2.2682459962368675, + "learning_rate": 3.996356533331146e-06, + "loss": 0.7672, + "num_input_tokens_seen": 8577945, + "step": 405 + }, + { + "epoch": 0.04881861359947093, + "flos": 18452662843680.0, + "grad_norm": 4.67158390018574, + "learning_rate": 3.996309383715573e-06, + "loss": 0.5883, + "num_input_tokens_seen": 8596445, + "step": 406 + }, + { + "epoch": 0.048938856490110025, + "flos": 11922869884080.0, + "grad_norm": 2.6839331977194507, + "learning_rate": 3.996261931261454e-06, + "loss": 0.7097, + "num_input_tokens_seen": 8614745, + "step": 407 + }, + { + "epoch": 0.049059099380749115, + "flos": 21937021895280.0, + "grad_norm": 2.431698963124244, + "learning_rate": 3.996214175975987e-06, + "loss": 0.854, + "num_input_tokens_seen": 8634985, + "step": 408 + }, + { + "epoch": 0.049179342271388204, + "flos": 26395522866960.0, + "grad_norm": 2.8186461368211333, + "learning_rate": 3.996166117866417e-06, + "loss": 0.7863, + "num_input_tokens_seen": 8656640, + "step": 409 + }, + { + "epoch": 0.049299585162027294, + "flos": 10626451807560.0, + "grad_norm": 3.177461111819697, + "learning_rate": 3.996117756940035e-06, + "loss": 0.8612, + "num_input_tokens_seen": 8673045, + "step": 410 + }, + { + "epoch": 0.049419828052666384, + "flos": 14293995608160.0, + "grad_norm": 2.4866226872841035, + "learning_rate": 3.996069093204175e-06, + "loss": 0.9627, + "num_input_tokens_seen": 8688725, + "step": 411 + }, + { + "epoch": 0.049540070943305474, + "flos": 9922039493040.0, + "grad_norm": 2.9493553685761444, + "learning_rate": 3.996020126666221e-06, + "loss": 0.8616, + "num_input_tokens_seen": 8705425, + "step": 412 + }, + { + "epoch": 0.04966031383394457, + "flos": 15969858746160.0, + "grad_norm": 4.699266441401282, + "learning_rate": 3.995970857333601e-06, + "loss": 0.8152, + "num_input_tokens_seen": 8725555, + "step": 413 + }, + { + "epoch": 0.04978055672458366, + "flos": 20992572526320.0, + "grad_norm": 2.678394773198341, + "learning_rate": 3.995921285213789e-06, + "loss": 0.7858, + "num_input_tokens_seen": 8745535, + "step": 414 + }, + { + "epoch": 0.04990079961522275, + "flos": 14487311544360.0, + "grad_norm": 3.803903988602288, + "learning_rate": 3.995871410314305e-06, + "loss": 0.7855, + "num_input_tokens_seen": 8763815, + "step": 415 + }, + { + "epoch": 0.05002104250586184, + "flos": 46246345424520.0, + "grad_norm": 1.012090666173632, + "learning_rate": 3.995821232642714e-06, + "loss": 0.651, + "num_input_tokens_seen": 8821940, + "step": 416 + }, + { + "epoch": 0.05014128539650093, + "flos": 20482141007160.0, + "grad_norm": 3.3177487517110125, + "learning_rate": 3.995770752206629e-06, + "loss": 0.808, + "num_input_tokens_seen": 8842735, + "step": 417 + }, + { + "epoch": 0.05026152828714002, + "flos": 12915496729440.0, + "grad_norm": 3.5335525162808805, + "learning_rate": 3.995719969013709e-06, + "loss": 0.9634, + "num_input_tokens_seen": 8859635, + "step": 418 + }, + { + "epoch": 0.05038177117777912, + "flos": 13972764230400.0, + "grad_norm": 4.049181189004957, + "learning_rate": 3.995668883071655e-06, + "loss": 0.8338, + "num_input_tokens_seen": 8875580, + "step": 419 + }, + { + "epoch": 0.050502014068418206, + "flos": 15108629937960.0, + "grad_norm": 2.7952460121704292, + "learning_rate": 3.995617494388219e-06, + "loss": 0.8907, + "num_input_tokens_seen": 8893420, + "step": 420 + }, + { + "epoch": 0.050622256959057296, + "flos": 15621815873640.0, + "grad_norm": 3.9618021305361775, + "learning_rate": 3.995565802971196e-06, + "loss": 0.7845, + "num_input_tokens_seen": 8913115, + "step": 421 + }, + { + "epoch": 0.050742499849696386, + "flos": 20293542405000.0, + "grad_norm": 3.5149074245975824, + "learning_rate": 3.995513808828427e-06, + "loss": 0.6658, + "num_input_tokens_seen": 8935630, + "step": 422 + }, + { + "epoch": 0.050862742740335476, + "flos": 14515452757440.0, + "grad_norm": 5.050166423812499, + "learning_rate": 3.9954615119678e-06, + "loss": 0.7478, + "num_input_tokens_seen": 8953905, + "step": 423 + }, + { + "epoch": 0.050982985630974566, + "flos": 16156811030400.0, + "grad_norm": 3.1843624754894306, + "learning_rate": 3.995408912397248e-06, + "loss": 0.7901, + "num_input_tokens_seen": 8971520, + "step": 424 + }, + { + "epoch": 0.05110322852161366, + "flos": 15162886126680.0, + "grad_norm": 4.751702656006952, + "learning_rate": 3.99535601012475e-06, + "loss": 0.9187, + "num_input_tokens_seen": 8986570, + "step": 425 + }, + { + "epoch": 0.05122347141225275, + "flos": 20939867675640.0, + "grad_norm": 3.0599223597032834, + "learning_rate": 3.995302805158333e-06, + "loss": 0.7455, + "num_input_tokens_seen": 9008945, + "step": 426 + }, + { + "epoch": 0.05134371430289184, + "flos": 14409758116440.0, + "grad_norm": 2.743635615786454, + "learning_rate": 3.9952492975060665e-06, + "loss": 0.8242, + "num_input_tokens_seen": 9028735, + "step": 427 + }, + { + "epoch": 0.05146395719353093, + "flos": 25315021446720.0, + "grad_norm": 3.722487935186937, + "learning_rate": 3.995195487176067e-06, + "loss": 0.8517, + "num_input_tokens_seen": 9048685, + "step": 428 + }, + { + "epoch": 0.05158420008417002, + "flos": 15916932275760.0, + "grad_norm": 2.498977452011935, + "learning_rate": 3.995141374176499e-06, + "loss": 0.8476, + "num_input_tokens_seen": 9066800, + "step": 429 + }, + { + "epoch": 0.05170444297480911, + "flos": 53513599205520.0, + "grad_norm": 0.9332536532813118, + "learning_rate": 3.995086958515572e-06, + "loss": 0.6692, + "num_input_tokens_seen": 9124540, + "step": 430 + }, + { + "epoch": 0.05182468586544821, + "flos": 46011690563280.0, + "grad_norm": 0.9932890356723935, + "learning_rate": 3.995032240201538e-06, + "loss": 0.6383, + "num_input_tokens_seen": 9186655, + "step": 431 + }, + { + "epoch": 0.0519449287560873, + "flos": 30325518973680.0, + "grad_norm": 0.9945950172280326, + "learning_rate": 3.9949772192427e-06, + "loss": 0.6466, + "num_input_tokens_seen": 9233000, + "step": 432 + }, + { + "epoch": 0.05206517164672639, + "flos": 12760421533560.0, + "grad_norm": 3.6345111116225537, + "learning_rate": 3.994921895647405e-06, + "loss": 0.7817, + "num_input_tokens_seen": 9250890, + "step": 433 + }, + { + "epoch": 0.05218541453736548, + "flos": 47183228063040.0, + "grad_norm": 0.8095526054847891, + "learning_rate": 3.994866269424043e-06, + "loss": 0.5651, + "num_input_tokens_seen": 9306980, + "step": 434 + }, + { + "epoch": 0.05230565742800457, + "flos": 14109196201200.0, + "grad_norm": 3.4892123048827113, + "learning_rate": 3.9948103405810545e-06, + "loss": 0.7675, + "num_input_tokens_seen": 9325650, + "step": 435 + }, + { + "epoch": 0.05242590031864366, + "flos": 18535440165000.0, + "grad_norm": 3.1579314936623546, + "learning_rate": 3.994754109126923e-06, + "loss": 0.8467, + "num_input_tokens_seen": 9346865, + "step": 436 + }, + { + "epoch": 0.052546143209282754, + "flos": 19211236367040.0, + "grad_norm": 2.683514856236607, + "learning_rate": 3.994697575070181e-06, + "loss": 0.9164, + "num_input_tokens_seen": 9366045, + "step": 437 + }, + { + "epoch": 0.052666386099921844, + "flos": 16210877259360.0, + "grad_norm": 2.2778514472812637, + "learning_rate": 3.994640738419402e-06, + "loss": 0.8967, + "num_input_tokens_seen": 9385140, + "step": 438 + }, + { + "epoch": 0.052786628990560934, + "flos": 17487765631920.0, + "grad_norm": 4.1237765723161806, + "learning_rate": 3.9945835991832075e-06, + "loss": 0.7935, + "num_input_tokens_seen": 9406745, + "step": 439 + }, + { + "epoch": 0.052906871881200024, + "flos": 18022064269560.0, + "grad_norm": 4.380486018767794, + "learning_rate": 3.994526157370268e-06, + "loss": 0.9126, + "num_input_tokens_seen": 9425080, + "step": 440 + }, + { + "epoch": 0.053027114771839114, + "flos": 41603899864800.0, + "grad_norm": 0.8614963012655764, + "learning_rate": 3.994468412989296e-06, + "loss": 0.6055, + "num_input_tokens_seen": 9486210, + "step": 441 + }, + { + "epoch": 0.053147357662478203, + "flos": 12624527782080.0, + "grad_norm": 6.885821062717479, + "learning_rate": 3.994410366049052e-06, + "loss": 0.9335, + "num_input_tokens_seen": 9503790, + "step": 442 + }, + { + "epoch": 0.0532676005531173, + "flos": 12516363664200.0, + "grad_norm": 4.05693283093903, + "learning_rate": 3.994352016558341e-06, + "loss": 0.8227, + "num_input_tokens_seen": 9520815, + "step": 443 + }, + { + "epoch": 0.05338784344375639, + "flos": 20401928142600.0, + "grad_norm": 3.5534447142826604, + "learning_rate": 3.994293364526014e-06, + "loss": 0.7272, + "num_input_tokens_seen": 9541420, + "step": 444 + }, + { + "epoch": 0.05350808633439548, + "flos": 15730391571000.0, + "grad_norm": 2.415993816058348, + "learning_rate": 3.99423440996097e-06, + "loss": 0.8183, + "num_input_tokens_seen": 9560680, + "step": 445 + }, + { + "epoch": 0.05362832922503457, + "flos": 14676923265240.0, + "grad_norm": 19.01873337628669, + "learning_rate": 3.994175152872152e-06, + "loss": 0.7943, + "num_input_tokens_seen": 9579485, + "step": 446 + }, + { + "epoch": 0.05374857211567366, + "flos": 19155967059600.0, + "grad_norm": 3.0606849118320887, + "learning_rate": 3.994115593268548e-06, + "loss": 0.7787, + "num_input_tokens_seen": 9598985, + "step": 447 + }, + { + "epoch": 0.05386881500631275, + "flos": 20154704277240.0, + "grad_norm": 7.811103879160433, + "learning_rate": 3.994055731159195e-06, + "loss": 0.8079, + "num_input_tokens_seen": 9616175, + "step": 448 + }, + { + "epoch": 0.053989057896951846, + "flos": 17267479901160.0, + "grad_norm": 3.549589353639062, + "learning_rate": 3.993995566553172e-06, + "loss": 0.8613, + "num_input_tokens_seen": 9634860, + "step": 449 + }, + { + "epoch": 0.054109300787590936, + "flos": 18484254992400.0, + "grad_norm": 3.2616175835733117, + "learning_rate": 3.993935099459607e-06, + "loss": 0.7547, + "num_input_tokens_seen": 9656195, + "step": 450 + }, + { + "epoch": 0.054229543678230026, + "flos": 17457408221640.0, + "grad_norm": 2.8374350721014063, + "learning_rate": 3.993874329887673e-06, + "loss": 0.7311, + "num_input_tokens_seen": 9674570, + "step": 451 + }, + { + "epoch": 0.054349786568869116, + "flos": 11891024455680.0, + "grad_norm": 4.125865267544811, + "learning_rate": 3.993813257846589e-06, + "loss": 0.8439, + "num_input_tokens_seen": 9691045, + "step": 452 + }, + { + "epoch": 0.054470029459508205, + "flos": 13623296659680.0, + "grad_norm": 3.2668250402349157, + "learning_rate": 3.993751883345619e-06, + "loss": 0.9178, + "num_input_tokens_seen": 9709125, + "step": 453 + }, + { + "epoch": 0.054590272350147295, + "flos": 12975451710960.0, + "grad_norm": 4.7397551229455175, + "learning_rate": 3.993690206394073e-06, + "loss": 0.8571, + "num_input_tokens_seen": 9725145, + "step": 454 + }, + { + "epoch": 0.054710515240786385, + "flos": 12976021590240.0, + "grad_norm": 2.8393419916516995, + "learning_rate": 3.993628227001307e-06, + "loss": 0.8706, + "num_input_tokens_seen": 9743065, + "step": 455 + }, + { + "epoch": 0.05483075813142548, + "flos": 35491942024080.0, + "grad_norm": 2.4942984415309217, + "learning_rate": 3.993565945176726e-06, + "loss": 0.7083, + "num_input_tokens_seen": 9763810, + "step": 456 + }, + { + "epoch": 0.05495100102206457, + "flos": 14379274066320.0, + "grad_norm": 2.38320549397522, + "learning_rate": 3.993503360929776e-06, + "loss": 0.8094, + "num_input_tokens_seen": 9782415, + "step": 457 + }, + { + "epoch": 0.05507124391270366, + "flos": 19321426722360.0, + "grad_norm": 2.0732117128221277, + "learning_rate": 3.99344047426995e-06, + "loss": 0.8033, + "num_input_tokens_seen": 9803395, + "step": 458 + }, + { + "epoch": 0.05519148680334275, + "flos": 16539612047640.0, + "grad_norm": 4.005297252658449, + "learning_rate": 3.993377285206789e-06, + "loss": 0.899, + "num_input_tokens_seen": 9822900, + "step": 459 + }, + { + "epoch": 0.05531172969398184, + "flos": 29824739250960.0, + "grad_norm": 2.3240624126779514, + "learning_rate": 3.99331379374988e-06, + "loss": 0.8525, + "num_input_tokens_seen": 9846225, + "step": 460 + }, + { + "epoch": 0.05543197258462093, + "flos": 17188470115080.0, + "grad_norm": 2.70683008192373, + "learning_rate": 3.993249999908852e-06, + "loss": 0.7857, + "num_input_tokens_seen": 9866095, + "step": 461 + }, + { + "epoch": 0.05555221547526003, + "flos": 13596611804760.0, + "grad_norm": 2.478341218961045, + "learning_rate": 3.993185903693384e-06, + "loss": 0.8555, + "num_input_tokens_seen": 9882615, + "step": 462 + }, + { + "epoch": 0.05567245836589912, + "flos": 17269252858920.0, + "grad_norm": 3.8526690210970114, + "learning_rate": 3.9931215051131995e-06, + "loss": 0.818, + "num_input_tokens_seen": 9902980, + "step": 463 + }, + { + "epoch": 0.05579270125653821, + "flos": 20347640293920.0, + "grad_norm": 6.21868099001832, + "learning_rate": 3.993056804178068e-06, + "loss": 0.7909, + "num_input_tokens_seen": 9924245, + "step": 464 + }, + { + "epoch": 0.0559129441471773, + "flos": 19805141726640.0, + "grad_norm": 12.1851378670439, + "learning_rate": 3.992991800897803e-06, + "loss": 0.8302, + "num_input_tokens_seen": 9943770, + "step": 465 + }, + { + "epoch": 0.05603318703781639, + "flos": 11083988516280.0, + "grad_norm": 4.354779846723337, + "learning_rate": 3.9929264952822665e-06, + "loss": 0.8782, + "num_input_tokens_seen": 9961025, + "step": 466 + }, + { + "epoch": 0.05615342992845548, + "flos": 16290900164160.0, + "grad_norm": 2.4798864927773057, + "learning_rate": 3.992860887341366e-06, + "loss": 0.8695, + "num_input_tokens_seen": 9978915, + "step": 467 + }, + { + "epoch": 0.056273672819094574, + "flos": 17267384921280.0, + "grad_norm": 2.604030595403379, + "learning_rate": 3.992794977085052e-06, + "loss": 0.7924, + "num_input_tokens_seen": 9996635, + "step": 468 + }, + { + "epoch": 0.056393915709733664, + "flos": 14509025785560.0, + "grad_norm": 3.1935955928496873, + "learning_rate": 3.992728764523326e-06, + "loss": 0.8391, + "num_input_tokens_seen": 10015300, + "step": 469 + }, + { + "epoch": 0.05651415860037275, + "flos": 16215847873080.0, + "grad_norm": 2.54331037391842, + "learning_rate": 3.99266224966623e-06, + "loss": 0.7859, + "num_input_tokens_seen": 10035935, + "step": 470 + }, + { + "epoch": 0.05663440149101184, + "flos": 14217455298960.0, + "grad_norm": 2.30578700848216, + "learning_rate": 3.992595432523855e-06, + "loss": 0.8631, + "num_input_tokens_seen": 10052945, + "step": 471 + }, + { + "epoch": 0.05675464438165093, + "flos": 16590037381200.0, + "grad_norm": 2.308205818849968, + "learning_rate": 3.992528313106338e-06, + "loss": 0.847, + "num_input_tokens_seen": 10070865, + "step": 472 + }, + { + "epoch": 0.05687488727229002, + "flos": 12319945051920.0, + "grad_norm": 3.3112229098647945, + "learning_rate": 3.9924608914238595e-06, + "loss": 0.7984, + "num_input_tokens_seen": 10085580, + "step": 473 + }, + { + "epoch": 0.05699513016292912, + "flos": 21400158800880.0, + "grad_norm": 3.368219699770615, + "learning_rate": 3.992393167486648e-06, + "loss": 0.823, + "num_input_tokens_seen": 10104450, + "step": 474 + }, + { + "epoch": 0.05711537305356821, + "flos": 13812370161240.0, + "grad_norm": 4.641211886815804, + "learning_rate": 3.992325141304977e-06, + "loss": 0.7888, + "num_input_tokens_seen": 10122125, + "step": 475 + }, + { + "epoch": 0.0572356159442073, + "flos": 19616321504760.0, + "grad_norm": 2.930640470008753, + "learning_rate": 3.992256812889166e-06, + "loss": 0.8491, + "num_input_tokens_seen": 10137950, + "step": 476 + }, + { + "epoch": 0.05735585883484639, + "flos": 25801459207560.0, + "grad_norm": 3.3245520956909203, + "learning_rate": 3.992188182249582e-06, + "loss": 0.7484, + "num_input_tokens_seen": 10159565, + "step": 477 + }, + { + "epoch": 0.05747610172548548, + "flos": 13677901107960.0, + "grad_norm": 2.469572871725434, + "learning_rate": 3.992119249396633e-06, + "loss": 0.8915, + "num_input_tokens_seen": 10177970, + "step": 478 + }, + { + "epoch": 0.05759634461612457, + "flos": 20508920841960.0, + "grad_norm": 2.954159187196231, + "learning_rate": 3.992050014340778e-06, + "loss": 0.8083, + "num_input_tokens_seen": 10198045, + "step": 479 + }, + { + "epoch": 0.057716587506763666, + "flos": 51097457509680.0, + "grad_norm": 1.126726997206704, + "learning_rate": 3.99198047709252e-06, + "loss": 0.5662, + "num_input_tokens_seen": 10259285, + "step": 480 + }, + { + "epoch": 0.057836830397402755, + "flos": 18321581406120.0, + "grad_norm": 2.416380426840529, + "learning_rate": 3.991910637662408e-06, + "loss": 0.7651, + "num_input_tokens_seen": 10279295, + "step": 481 + }, + { + "epoch": 0.057957073288041845, + "flos": 18754427837400.0, + "grad_norm": 2.519235330399457, + "learning_rate": 3.9918404960610355e-06, + "loss": 0.7905, + "num_input_tokens_seen": 10298045, + "step": 482 + }, + { + "epoch": 0.058077316178680935, + "flos": 15189855921240.0, + "grad_norm": 2.7367358669711637, + "learning_rate": 3.991770052299043e-06, + "loss": 0.7546, + "num_input_tokens_seen": 10315995, + "step": 483 + }, + { + "epoch": 0.058197559069320025, + "flos": 13812623440920.0, + "grad_norm": 3.9909860852371, + "learning_rate": 3.991699306387118e-06, + "loss": 0.8597, + "num_input_tokens_seen": 10334185, + "step": 484 + }, + { + "epoch": 0.058317801959959115, + "flos": 18213638907960.0, + "grad_norm": 2.476318828737293, + "learning_rate": 3.991628258335991e-06, + "loss": 0.7796, + "num_input_tokens_seen": 10356110, + "step": 485 + }, + { + "epoch": 0.05843804485059821, + "flos": 17025574909080.0, + "grad_norm": 15.096901258612798, + "learning_rate": 3.991556908156442e-06, + "loss": 0.8534, + "num_input_tokens_seen": 10372355, + "step": 486 + }, + { + "epoch": 0.0585582877412373, + "flos": 16946343503280.0, + "grad_norm": 2.175654354538506, + "learning_rate": 3.9914852558592914e-06, + "loss": 0.8571, + "num_input_tokens_seen": 10393125, + "step": 487 + }, + { + "epoch": 0.05867853063187639, + "flos": 17208443058480.0, + "grad_norm": 11.17275592025153, + "learning_rate": 3.991413301455413e-06, + "loss": 0.7702, + "num_input_tokens_seen": 10409295, + "step": 488 + }, + { + "epoch": 0.05879877352251548, + "flos": 21641810513280.0, + "grad_norm": 3.8747275472042517, + "learning_rate": 3.991341044955719e-06, + "loss": 0.7576, + "num_input_tokens_seen": 10428770, + "step": 489 + }, + { + "epoch": 0.05891901641315457, + "flos": 14731116134040.0, + "grad_norm": 2.29438085793766, + "learning_rate": 3.991268486371172e-06, + "loss": 0.8091, + "num_input_tokens_seen": 10447045, + "step": 490 + }, + { + "epoch": 0.05903925930379366, + "flos": 18051440221080.0, + "grad_norm": 3.6338143884749665, + "learning_rate": 3.991195625712779e-06, + "loss": 0.8511, + "num_input_tokens_seen": 10463730, + "step": 491 + }, + { + "epoch": 0.05915950219443276, + "flos": 15539703411480.0, + "grad_norm": 3.8102271387272766, + "learning_rate": 3.991122462991592e-06, + "loss": 0.7962, + "num_input_tokens_seen": 10482970, + "step": 492 + }, + { + "epoch": 0.05927974508507185, + "flos": 7141586196600.0, + "grad_norm": 4.18430878613185, + "learning_rate": 3.991048998218712e-06, + "loss": 0.7828, + "num_input_tokens_seen": 10495995, + "step": 493 + }, + { + "epoch": 0.05939998797571094, + "flos": 13326375639840.0, + "grad_norm": 3.174890447213359, + "learning_rate": 3.990975231405281e-06, + "loss": 0.7416, + "num_input_tokens_seen": 10514165, + "step": 494 + }, + { + "epoch": 0.05952023086635003, + "flos": 20724900818160.0, + "grad_norm": 2.8307887225771755, + "learning_rate": 3.990901162562491e-06, + "loss": 0.7853, + "num_input_tokens_seen": 10534575, + "step": 495 + }, + { + "epoch": 0.05964047375698912, + "flos": 10841925224400.0, + "grad_norm": 7.164125864462145, + "learning_rate": 3.9908267917015765e-06, + "loss": 0.8819, + "num_input_tokens_seen": 10552355, + "step": 496 + }, + { + "epoch": 0.059760716647628206, + "flos": 16972078559400.0, + "grad_norm": 3.425730027410672, + "learning_rate": 3.990752118833821e-06, + "loss": 0.9153, + "num_input_tokens_seen": 10569515, + "step": 497 + }, + { + "epoch": 0.0598809595382673, + "flos": 16647364586040.0, + "grad_norm": 2.2441174644556603, + "learning_rate": 3.990677143970553e-06, + "loss": 0.7699, + "num_input_tokens_seen": 10590045, + "step": 498 + }, + { + "epoch": 0.06000120242890639, + "flos": 16188783098640.0, + "grad_norm": 3.32278142352906, + "learning_rate": 3.990601867123144e-06, + "loss": 0.7906, + "num_input_tokens_seen": 10609490, + "step": 499 + }, + { + "epoch": 0.06012144531954548, + "flos": 14001538642680.0, + "grad_norm": 3.7904985046665556, + "learning_rate": 3.990526288303014e-06, + "loss": 0.8416, + "num_input_tokens_seen": 10628000, + "step": 500 + }, + { + "epoch": 0.06024168821018457, + "flos": 16674271060680.0, + "grad_norm": 2.3949841417970243, + "learning_rate": 3.9904504075216295e-06, + "loss": 0.8868, + "num_input_tokens_seen": 10648480, + "step": 501 + }, + { + "epoch": 0.06036193110082366, + "flos": 13704744262680.0, + "grad_norm": 3.363500855971004, + "learning_rate": 3.990374224790501e-06, + "loss": 0.931, + "num_input_tokens_seen": 10666405, + "step": 502 + }, + { + "epoch": 0.06048217399146275, + "flos": 12543460098600.0, + "grad_norm": 2.6893211227080527, + "learning_rate": 3.990297740121185e-06, + "loss": 0.6969, + "num_input_tokens_seen": 10684060, + "step": 503 + }, + { + "epoch": 0.06060241688210185, + "flos": 17592573794040.0, + "grad_norm": 2.3532242590650276, + "learning_rate": 3.990220953525284e-06, + "loss": 0.7646, + "num_input_tokens_seen": 10700890, + "step": 504 + }, + { + "epoch": 0.06072265977274094, + "flos": 10626420147600.0, + "grad_norm": 3.258282305097567, + "learning_rate": 3.9901438650144465e-06, + "loss": 0.7233, + "num_input_tokens_seen": 10716860, + "step": 505 + }, + { + "epoch": 0.06084290266338003, + "flos": 15027657234360.0, + "grad_norm": 2.805780479387285, + "learning_rate": 3.990066474600367e-06, + "loss": 0.8959, + "num_input_tokens_seen": 10734550, + "step": 506 + }, + { + "epoch": 0.06096314555401912, + "flos": 16323410451720.0, + "grad_norm": 2.3920053722933656, + "learning_rate": 3.989988782294786e-06, + "loss": 0.6689, + "num_input_tokens_seen": 10754360, + "step": 507 + }, + { + "epoch": 0.06108338844465821, + "flos": 13971086252520.0, + "grad_norm": 1.9787690281871813, + "learning_rate": 3.989910788109489e-06, + "loss": 0.9315, + "num_input_tokens_seen": 10770730, + "step": 508 + }, + { + "epoch": 0.0612036313352973, + "flos": 24667619737440.0, + "grad_norm": 3.185097929977564, + "learning_rate": 3.989832492056307e-06, + "loss": 0.7355, + "num_input_tokens_seen": 10791475, + "step": 509 + }, + { + "epoch": 0.06132387422593639, + "flos": 20157996913080.0, + "grad_norm": 3.1525565717251207, + "learning_rate": 3.989753894147119e-06, + "loss": 0.7802, + "num_input_tokens_seen": 10811320, + "step": 510 + }, + { + "epoch": 0.061444117116575485, + "flos": 18969806274360.0, + "grad_norm": 2.23792992409365, + "learning_rate": 3.989674994393846e-06, + "loss": 0.7816, + "num_input_tokens_seen": 10830515, + "step": 511 + }, + { + "epoch": 0.061564360007214575, + "flos": 20913942659760.0, + "grad_norm": 3.0350470603046933, + "learning_rate": 3.98959579280846e-06, + "loss": 0.9283, + "num_input_tokens_seen": 10848635, + "step": 512 + }, + { + "epoch": 0.061684602897853665, + "flos": 8763889665000.0, + "grad_norm": 3.063234824947875, + "learning_rate": 3.989516289402973e-06, + "loss": 0.8126, + "num_input_tokens_seen": 10863985, + "step": 513 + }, + { + "epoch": 0.061804845788492754, + "flos": 14268038932320.0, + "grad_norm": 3.3109439577955104, + "learning_rate": 3.989436484189447e-06, + "loss": 0.8001, + "num_input_tokens_seen": 10881650, + "step": 514 + }, + { + "epoch": 0.061925088679131844, + "flos": 11166639197760.0, + "grad_norm": 3.663554532561146, + "learning_rate": 3.9893563771799885e-06, + "loss": 0.7849, + "num_input_tokens_seen": 10897845, + "step": 515 + }, + { + "epoch": 0.062045331569770934, + "flos": 18995383030680.0, + "grad_norm": 4.576972061121961, + "learning_rate": 3.989275968386749e-06, + "loss": 0.8542, + "num_input_tokens_seen": 10915475, + "step": 516 + }, + { + "epoch": 0.06216557446041003, + "flos": 20616863340120.0, + "grad_norm": 3.0428275588533573, + "learning_rate": 3.989195257821926e-06, + "loss": 0.7499, + "num_input_tokens_seen": 10933680, + "step": 517 + }, + { + "epoch": 0.06228581735104912, + "flos": 17188723394760.0, + "grad_norm": 2.6716028264494107, + "learning_rate": 3.989114245497765e-06, + "loss": 0.8259, + "num_input_tokens_seen": 10953200, + "step": 518 + }, + { + "epoch": 0.06240606024168821, + "flos": 11058665039640.0, + "grad_norm": 2.9968480400328654, + "learning_rate": 3.989032931426554e-06, + "loss": 0.9327, + "num_input_tokens_seen": 10970075, + "step": 519 + }, + { + "epoch": 0.06252630313232731, + "flos": 15082166702760.0, + "grad_norm": 2.6971584025291766, + "learning_rate": 3.9889513156206295e-06, + "loss": 0.8528, + "num_input_tokens_seen": 10989235, + "step": 520 + }, + { + "epoch": 0.06264654602296639, + "flos": 15190425800520.0, + "grad_norm": 6.410737043310743, + "learning_rate": 3.988869398092371e-06, + "loss": 0.6872, + "num_input_tokens_seen": 11008865, + "step": 521 + }, + { + "epoch": 0.06276678891360549, + "flos": 21723289776240.0, + "grad_norm": 2.429087672140574, + "learning_rate": 3.988787178854206e-06, + "loss": 0.76, + "num_input_tokens_seen": 11028120, + "step": 522 + }, + { + "epoch": 0.06288703180424457, + "flos": 16187959939680.0, + "grad_norm": 2.578679897818141, + "learning_rate": 3.988704657918608e-06, + "loss": 0.8567, + "num_input_tokens_seen": 11046900, + "step": 523 + }, + { + "epoch": 0.06300727469488367, + "flos": 10897004572080.0, + "grad_norm": 3.0715635434780664, + "learning_rate": 3.988621835298094e-06, + "loss": 0.7982, + "num_input_tokens_seen": 11063835, + "step": 524 + }, + { + "epoch": 0.06312751758552275, + "flos": 17971005736800.0, + "grad_norm": 2.3523330186531313, + "learning_rate": 3.988538711005229e-06, + "loss": 0.9084, + "num_input_tokens_seen": 11083010, + "step": 525 + }, + { + "epoch": 0.06324776047616185, + "flos": 15729790031760.0, + "grad_norm": 2.541829828330407, + "learning_rate": 3.988455285052622e-06, + "loss": 0.8772, + "num_input_tokens_seen": 11098910, + "step": 526 + }, + { + "epoch": 0.06336800336680094, + "flos": 15863562565920.0, + "grad_norm": 2.7082475404111093, + "learning_rate": 3.98837155745293e-06, + "loss": 0.8236, + "num_input_tokens_seen": 11116670, + "step": 527 + }, + { + "epoch": 0.06348824625744003, + "flos": 14245216592520.0, + "grad_norm": 3.803462338434245, + "learning_rate": 3.988287528218854e-06, + "loss": 0.7542, + "num_input_tokens_seen": 11135175, + "step": 528 + }, + { + "epoch": 0.06360848914807912, + "flos": 11269516102320.0, + "grad_norm": 2.76984458625072, + "learning_rate": 3.98820319736314e-06, + "loss": 0.8844, + "num_input_tokens_seen": 11151510, + "step": 529 + }, + { + "epoch": 0.0637287320387182, + "flos": 15053423950440.0, + "grad_norm": 2.1402055092647543, + "learning_rate": 3.988118564898582e-06, + "loss": 0.8417, + "num_input_tokens_seen": 11170770, + "step": 530 + }, + { + "epoch": 0.0638489749293573, + "flos": 12698187034920.0, + "grad_norm": 3.684265269683031, + "learning_rate": 3.988033630838019e-06, + "loss": 0.8838, + "num_input_tokens_seen": 11184530, + "step": 531 + }, + { + "epoch": 0.0639692178199964, + "flos": 17432717944200.0, + "grad_norm": 3.2832873604229773, + "learning_rate": 3.987948395194334e-06, + "loss": 0.8625, + "num_input_tokens_seen": 11206630, + "step": 532 + }, + { + "epoch": 0.06408946071063548, + "flos": 13487339588280.0, + "grad_norm": 2.80142855645824, + "learning_rate": 3.987862857980458e-06, + "loss": 0.767, + "num_input_tokens_seen": 11222295, + "step": 533 + }, + { + "epoch": 0.06420970360127458, + "flos": 19915015482360.0, + "grad_norm": 2.468614502097693, + "learning_rate": 3.987777019209368e-06, + "loss": 0.748, + "num_input_tokens_seen": 11242530, + "step": 534 + }, + { + "epoch": 0.06432994649191366, + "flos": 17322907508400.0, + "grad_norm": 4.352596796965361, + "learning_rate": 3.987690878894084e-06, + "loss": 0.8003, + "num_input_tokens_seen": 11261965, + "step": 535 + }, + { + "epoch": 0.06445018938255276, + "flos": 17133612387120.0, + "grad_norm": 3.233021770947938, + "learning_rate": 3.987604437047673e-06, + "loss": 0.8309, + "num_input_tokens_seen": 11281485, + "step": 536 + }, + { + "epoch": 0.06457043227319184, + "flos": 14353095770760.0, + "grad_norm": 2.8004256572067305, + "learning_rate": 3.987517693683251e-06, + "loss": 0.7685, + "num_input_tokens_seen": 11299780, + "step": 537 + }, + { + "epoch": 0.06469067516383094, + "flos": 12378000435840.0, + "grad_norm": 2.5698344705019482, + "learning_rate": 3.9874306488139745e-06, + "loss": 0.9484, + "num_input_tokens_seen": 11314760, + "step": 538 + }, + { + "epoch": 0.06481091805447003, + "flos": 17054286001440.0, + "grad_norm": 3.871949625720067, + "learning_rate": 3.987343302453049e-06, + "loss": 0.8474, + "num_input_tokens_seen": 11335755, + "step": 539 + }, + { + "epoch": 0.06493116094510912, + "flos": 21403134837120.0, + "grad_norm": 2.1024256638892553, + "learning_rate": 3.987255654613724e-06, + "loss": 0.8099, + "num_input_tokens_seen": 11359240, + "step": 540 + }, + { + "epoch": 0.06505140383574821, + "flos": 14514629598480.0, + "grad_norm": 2.695544335166443, + "learning_rate": 3.987167705309296e-06, + "loss": 0.6942, + "num_input_tokens_seen": 11378235, + "step": 541 + }, + { + "epoch": 0.0651716467263873, + "flos": 13077695416320.0, + "grad_norm": 2.5610885454545866, + "learning_rate": 3.987079454553108e-06, + "loss": 0.936, + "num_input_tokens_seen": 11395905, + "step": 542 + }, + { + "epoch": 0.0652918896170264, + "flos": 15237210238680.0, + "grad_norm": 2.02450380250626, + "learning_rate": 3.986990902358546e-06, + "loss": 0.9043, + "num_input_tokens_seen": 11412565, + "step": 543 + }, + { + "epoch": 0.06541213250766549, + "flos": 16000374456240.0, + "grad_norm": 2.5904286272193033, + "learning_rate": 3.986902048739045e-06, + "loss": 0.9119, + "num_input_tokens_seen": 11432230, + "step": 544 + }, + { + "epoch": 0.06553237539830457, + "flos": 16916429332440.0, + "grad_norm": 4.041455266608437, + "learning_rate": 3.986812893708082e-06, + "loss": 0.7827, + "num_input_tokens_seen": 11448140, + "step": 545 + }, + { + "epoch": 0.06565261828894367, + "flos": 13076872257360.0, + "grad_norm": 6.975953352115998, + "learning_rate": 3.9867234372791826e-06, + "loss": 0.8018, + "num_input_tokens_seen": 11465815, + "step": 546 + }, + { + "epoch": 0.06577286117958275, + "flos": 16674017781000.0, + "grad_norm": 2.1593285297522136, + "learning_rate": 3.986633679465918e-06, + "loss": 0.8606, + "num_input_tokens_seen": 11485690, + "step": 547 + }, + { + "epoch": 0.06589310407022185, + "flos": 17349940622880.0, + "grad_norm": 2.9998472492748585, + "learning_rate": 3.986543620281904e-06, + "loss": 0.8018, + "num_input_tokens_seen": 11505060, + "step": 548 + }, + { + "epoch": 0.06601334696086093, + "flos": 19694096552400.0, + "grad_norm": 1.7215947277897723, + "learning_rate": 3.986453259740802e-06, + "loss": 0.8998, + "num_input_tokens_seen": 11522950, + "step": 549 + }, + { + "epoch": 0.06613358985150003, + "flos": 9114053754840.0, + "grad_norm": 3.6241711951571274, + "learning_rate": 3.986362597856319e-06, + "loss": 0.7678, + "num_input_tokens_seen": 11539170, + "step": 550 + }, + { + "epoch": 0.06625383274213913, + "flos": 13380410208840.0, + "grad_norm": 2.9788802927378457, + "learning_rate": 3.986271634642211e-06, + "loss": 0.7934, + "num_input_tokens_seen": 11555870, + "step": 551 + }, + { + "epoch": 0.06637407563277821, + "flos": 11191867694520.0, + "grad_norm": 2.9685448826652343, + "learning_rate": 3.986180370112274e-06, + "loss": 0.8053, + "num_input_tokens_seen": 11572110, + "step": 552 + }, + { + "epoch": 0.0664943185234173, + "flos": 17593555252800.0, + "grad_norm": 2.531170219765356, + "learning_rate": 3.986088804280354e-06, + "loss": 0.734, + "num_input_tokens_seen": 11591560, + "step": 553 + }, + { + "epoch": 0.06661456141405639, + "flos": 15322805296440.0, + "grad_norm": 3.316016436021885, + "learning_rate": 3.985996937160342e-06, + "loss": 0.9204, + "num_input_tokens_seen": 11610470, + "step": 554 + }, + { + "epoch": 0.06673480430469549, + "flos": 38461817081640.0, + "grad_norm": 2.577361392596482, + "learning_rate": 3.985904768766173e-06, + "loss": 0.6794, + "num_input_tokens_seen": 11632965, + "step": 555 + }, + { + "epoch": 0.06685504719533458, + "flos": 11814959045880.0, + "grad_norm": 3.669809482836926, + "learning_rate": 3.98581229911183e-06, + "loss": 0.7396, + "num_input_tokens_seen": 11651605, + "step": 556 + }, + { + "epoch": 0.06697529008597367, + "flos": 16458322744440.0, + "grad_norm": 2.207297412048131, + "learning_rate": 3.985719528211341e-06, + "loss": 0.8997, + "num_input_tokens_seen": 11670695, + "step": 557 + }, + { + "epoch": 0.06709553297661276, + "flos": 46210225901520.0, + "grad_norm": 0.9252762523741491, + "learning_rate": 3.985626456078777e-06, + "loss": 0.6653, + "num_input_tokens_seen": 11735070, + "step": 558 + }, + { + "epoch": 0.06721577586725185, + "flos": 8113986818880.0, + "grad_norm": 2.8924249766914567, + "learning_rate": 3.985533082728259e-06, + "loss": 0.8498, + "num_input_tokens_seen": 11750445, + "step": 559 + }, + { + "epoch": 0.06733601875789094, + "flos": 18997377608160.0, + "grad_norm": 2.096672864742043, + "learning_rate": 3.985439408173951e-06, + "loss": 0.7295, + "num_input_tokens_seen": 11770390, + "step": 560 + }, + { + "epoch": 0.06745626164853002, + "flos": 15216635756040.0, + "grad_norm": 2.8217683453011264, + "learning_rate": 3.9853454324300634e-06, + "loss": 0.6876, + "num_input_tokens_seen": 11789320, + "step": 561 + }, + { + "epoch": 0.06757650453916912, + "flos": 14488229683200.0, + "grad_norm": 3.509910856457166, + "learning_rate": 3.985251155510852e-06, + "loss": 0.7584, + "num_input_tokens_seen": 11808070, + "step": 562 + }, + { + "epoch": 0.06769674742980822, + "flos": 18861895436160.0, + "grad_norm": 2.8439357822641806, + "learning_rate": 3.98515657743062e-06, + "loss": 0.7985, + "num_input_tokens_seen": 11827255, + "step": 563 + }, + { + "epoch": 0.0678169903204473, + "flos": 9488813142240.0, + "grad_norm": 2.454822706047718, + "learning_rate": 3.985061698203711e-06, + "loss": 0.7584, + "num_input_tokens_seen": 11844090, + "step": 564 + }, + { + "epoch": 0.0679372332110864, + "flos": 52262889128520.0, + "grad_norm": 0.8626638422712917, + "learning_rate": 3.984966517844523e-06, + "loss": 0.661, + "num_input_tokens_seen": 11899055, + "step": 565 + }, + { + "epoch": 0.06805747610172548, + "flos": 20724742518360.0, + "grad_norm": 3.51540472400194, + "learning_rate": 3.984871036367492e-06, + "loss": 0.8111, + "num_input_tokens_seen": 11918800, + "step": 566 + }, + { + "epoch": 0.06817771899236458, + "flos": 14703703100040.0, + "grad_norm": 2.277164825163232, + "learning_rate": 3.984775253787102e-06, + "loss": 0.8246, + "num_input_tokens_seen": 11936810, + "step": 567 + }, + { + "epoch": 0.06829796188300366, + "flos": 13082919309720.0, + "grad_norm": 3.723563020793957, + "learning_rate": 3.984679170117885e-06, + "loss": 0.8624, + "num_input_tokens_seen": 11952735, + "step": 568 + }, + { + "epoch": 0.06841820477364276, + "flos": 10545320804160.0, + "grad_norm": 2.843554311873215, + "learning_rate": 3.984582785374415e-06, + "loss": 0.7674, + "num_input_tokens_seen": 11969895, + "step": 569 + }, + { + "epoch": 0.06853844766428185, + "flos": 16048203673080.0, + "grad_norm": 2.1233940788135466, + "learning_rate": 3.9844860995713155e-06, + "loss": 0.7965, + "num_input_tokens_seen": 11989155, + "step": 570 + }, + { + "epoch": 0.06865869055492094, + "flos": 12246792358440.0, + "grad_norm": 3.694821992543022, + "learning_rate": 3.9843891127232524e-06, + "loss": 0.7984, + "num_input_tokens_seen": 12006410, + "step": 571 + }, + { + "epoch": 0.06877893344556003, + "flos": 14567967648360.0, + "grad_norm": 3.441458070605161, + "learning_rate": 3.984291824844938e-06, + "loss": 0.6512, + "num_input_tokens_seen": 12021225, + "step": 572 + }, + { + "epoch": 0.06889917633619912, + "flos": 16863851121600.0, + "grad_norm": 2.7876282324611634, + "learning_rate": 3.984194235951132e-06, + "loss": 0.8253, + "num_input_tokens_seen": 12037090, + "step": 573 + }, + { + "epoch": 0.06901941922683821, + "flos": 15325496393040.0, + "grad_norm": 3.7824624379795977, + "learning_rate": 3.9840963460566375e-06, + "loss": 0.8206, + "num_input_tokens_seen": 12055590, + "step": 574 + }, + { + "epoch": 0.06913966211747731, + "flos": 18182236719000.0, + "grad_norm": 2.1788932419665943, + "learning_rate": 3.983998155176305e-06, + "loss": 0.8823, + "num_input_tokens_seen": 12075670, + "step": 575 + }, + { + "epoch": 0.06925990500811639, + "flos": 43015542230280.0, + "grad_norm": 0.8029901324481784, + "learning_rate": 3.9838996633250305e-06, + "loss": 0.5855, + "num_input_tokens_seen": 12135905, + "step": 576 + }, + { + "epoch": 0.06938014789875549, + "flos": 9248902727640.0, + "grad_norm": 2.64066373356508, + "learning_rate": 3.983800870517753e-06, + "loss": 0.8728, + "num_input_tokens_seen": 12152415, + "step": 577 + }, + { + "epoch": 0.06950039078939457, + "flos": 16701620774760.0, + "grad_norm": 4.1838549791516755, + "learning_rate": 3.983701776769463e-06, + "loss": 0.7771, + "num_input_tokens_seen": 12169545, + "step": 578 + }, + { + "epoch": 0.06962063368003367, + "flos": 16050736469880.0, + "grad_norm": 2.864551298944934, + "learning_rate": 3.9836023820951885e-06, + "loss": 0.8439, + "num_input_tokens_seen": 12188480, + "step": 579 + }, + { + "epoch": 0.06974087657067275, + "flos": 15137246050440.0, + "grad_norm": 2.377992150584902, + "learning_rate": 3.983502686510011e-06, + "loss": 0.6752, + "num_input_tokens_seen": 12209030, + "step": 580 + }, + { + "epoch": 0.06986111946131185, + "flos": 16566993421680.0, + "grad_norm": 2.2263423342354574, + "learning_rate": 3.9834026900290525e-06, + "loss": 0.7108, + "num_input_tokens_seen": 12228145, + "step": 581 + }, + { + "epoch": 0.06998136235195095, + "flos": 19753133395080.0, + "grad_norm": 3.703173544521278, + "learning_rate": 3.983302392667482e-06, + "loss": 0.9787, + "num_input_tokens_seen": 12248710, + "step": 582 + }, + { + "epoch": 0.07010160524259003, + "flos": 16459177563360.0, + "grad_norm": 2.317987786636893, + "learning_rate": 3.983201794440517e-06, + "loss": 0.9263, + "num_input_tokens_seen": 12268005, + "step": 583 + }, + { + "epoch": 0.07022184813322913, + "flos": 13380441868800.0, + "grad_norm": 2.5772252217743796, + "learning_rate": 3.9831008953634165e-06, + "loss": 0.6581, + "num_input_tokens_seen": 12287015, + "step": 584 + }, + { + "epoch": 0.07034209102386821, + "flos": 18073914301320.0, + "grad_norm": 2.5555026652442203, + "learning_rate": 3.9829996954514864e-06, + "loss": 0.7992, + "num_input_tokens_seen": 12305875, + "step": 585 + }, + { + "epoch": 0.0704623339145073, + "flos": 19052013716400.0, + "grad_norm": 2.3400219577182946, + "learning_rate": 3.982898194720079e-06, + "loss": 0.8328, + "num_input_tokens_seen": 12326325, + "step": 586 + }, + { + "epoch": 0.0705825768051464, + "flos": 18565259355960.0, + "grad_norm": 2.3527210714669327, + "learning_rate": 3.982796393184592e-06, + "loss": 0.7916, + "num_input_tokens_seen": 12345125, + "step": 587 + }, + { + "epoch": 0.07070281969578548, + "flos": 48696734214360.0, + "grad_norm": 0.8072840085230369, + "learning_rate": 3.98269429086047e-06, + "loss": 0.6439, + "num_input_tokens_seen": 12402685, + "step": 588 + }, + { + "epoch": 0.07082306258642458, + "flos": 17317525315200.0, + "grad_norm": 2.9017014543645305, + "learning_rate": 3.982591887763199e-06, + "loss": 0.8378, + "num_input_tokens_seen": 12419865, + "step": 589 + }, + { + "epoch": 0.07094330547706366, + "flos": 10085124658800.0, + "grad_norm": 3.1610521100584346, + "learning_rate": 3.982489183908316e-06, + "loss": 0.7979, + "num_input_tokens_seen": 12436005, + "step": 590 + }, + { + "epoch": 0.07106354836770276, + "flos": 18052231720080.0, + "grad_norm": 2.1077923524203306, + "learning_rate": 3.982386179311399e-06, + "loss": 0.8297, + "num_input_tokens_seen": 12456245, + "step": 591 + }, + { + "epoch": 0.07118379125834184, + "flos": 11815085685720.0, + "grad_norm": 3.4868906288596713, + "learning_rate": 3.982282873988075e-06, + "loss": 0.8517, + "num_input_tokens_seen": 12473840, + "step": 592 + }, + { + "epoch": 0.07130403414898094, + "flos": 14407383619440.0, + "grad_norm": 2.005840238277446, + "learning_rate": 3.982179267954016e-06, + "loss": 0.8646, + "num_input_tokens_seen": 12493990, + "step": 593 + }, + { + "epoch": 0.07142427703962004, + "flos": 15999867896880.0, + "grad_norm": 3.600418273150799, + "learning_rate": 3.982075361224937e-06, + "loss": 0.9478, + "num_input_tokens_seen": 12512075, + "step": 594 + }, + { + "epoch": 0.07154451993025912, + "flos": 13353915313680.0, + "grad_norm": 3.0339298004196658, + "learning_rate": 3.981971153816602e-06, + "loss": 0.8796, + "num_input_tokens_seen": 12529400, + "step": 595 + }, + { + "epoch": 0.07166476282089822, + "flos": 16213125116520.0, + "grad_norm": 1.9512733532365227, + "learning_rate": 3.981866645744819e-06, + "loss": 0.9486, + "num_input_tokens_seen": 12549835, + "step": 596 + }, + { + "epoch": 0.0717850057115373, + "flos": 10274926339440.0, + "grad_norm": 3.1417114504211745, + "learning_rate": 3.9817618370254416e-06, + "loss": 0.7907, + "num_input_tokens_seen": 12566210, + "step": 597 + }, + { + "epoch": 0.0719052486021764, + "flos": 22694044080600.0, + "grad_norm": 3.0761188611503303, + "learning_rate": 3.9816567276743684e-06, + "loss": 0.8571, + "num_input_tokens_seen": 12585795, + "step": 598 + }, + { + "epoch": 0.0720254914928155, + "flos": 15566831505840.0, + "grad_norm": 2.1648320231125227, + "learning_rate": 3.9815513177075466e-06, + "loss": 0.7485, + "num_input_tokens_seen": 12604300, + "step": 599 + }, + { + "epoch": 0.07214573438345458, + "flos": 20401231623480.0, + "grad_norm": 2.408934841695226, + "learning_rate": 3.9814456071409646e-06, + "loss": 0.697, + "num_input_tokens_seen": 12624555, + "step": 600 + }, + { + "epoch": 0.07226597727409367, + "flos": 18672663634800.0, + "grad_norm": 3.2659466598742917, + "learning_rate": 3.981339595990659e-06, + "loss": 0.8366, + "num_input_tokens_seen": 12642805, + "step": 601 + }, + { + "epoch": 0.07238622016473276, + "flos": 17214996670200.0, + "grad_norm": 2.5457794079685825, + "learning_rate": 3.981233284272713e-06, + "loss": 0.804, + "num_input_tokens_seen": 12662270, + "step": 602 + }, + { + "epoch": 0.07250646305537185, + "flos": 18651329313120.0, + "grad_norm": 1.7814771273180778, + "learning_rate": 3.981126672003253e-06, + "loss": 0.8876, + "num_input_tokens_seen": 12684665, + "step": 603 + }, + { + "epoch": 0.07262670594601094, + "flos": 19908683490360.0, + "grad_norm": 3.010751056689012, + "learning_rate": 3.981019759198451e-06, + "loss": 0.7676, + "num_input_tokens_seen": 12703335, + "step": 604 + }, + { + "epoch": 0.07274694883665003, + "flos": 19537153418880.0, + "grad_norm": 3.4307674714880356, + "learning_rate": 3.980912545874528e-06, + "loss": 0.8076, + "num_input_tokens_seen": 12723220, + "step": 605 + }, + { + "epoch": 0.07286719172728913, + "flos": 21913313076600.0, + "grad_norm": 3.0075225293971686, + "learning_rate": 3.980805032047746e-06, + "loss": 0.8393, + "num_input_tokens_seen": 12744410, + "step": 606 + }, + { + "epoch": 0.07298743461792821, + "flos": 12676757733360.0, + "grad_norm": 2.3069147242678794, + "learning_rate": 3.980697217734415e-06, + "loss": 0.7849, + "num_input_tokens_seen": 12761870, + "step": 607 + }, + { + "epoch": 0.07310767750856731, + "flos": 14242842095520.0, + "grad_norm": 1.8959400719857364, + "learning_rate": 3.980589102950891e-06, + "loss": 0.8949, + "num_input_tokens_seen": 12779755, + "step": 608 + }, + { + "epoch": 0.07322792039920639, + "flos": 21399905521200.0, + "grad_norm": 3.5012221246254565, + "learning_rate": 3.9804806877135755e-06, + "loss": 0.7641, + "num_input_tokens_seen": 12797520, + "step": 609 + }, + { + "epoch": 0.07334816328984549, + "flos": 17188406795160.0, + "grad_norm": 3.2080758887373517, + "learning_rate": 3.980371972038915e-06, + "loss": 0.8514, + "num_input_tokens_seen": 12817730, + "step": 610 + }, + { + "epoch": 0.07346840618048459, + "flos": 16806365616960.0, + "grad_norm": 2.301727980953271, + "learning_rate": 3.980262955943399e-06, + "loss": 0.8289, + "num_input_tokens_seen": 12837115, + "step": 611 + }, + { + "epoch": 0.07358864907112367, + "flos": 12892737709560.0, + "grad_norm": 3.3551131782828003, + "learning_rate": 3.980153639443569e-06, + "loss": 0.8672, + "num_input_tokens_seen": 12852820, + "step": 612 + }, + { + "epoch": 0.07370889196176277, + "flos": 17646513383160.0, + "grad_norm": 2.8665557913814514, + "learning_rate": 3.980044022556005e-06, + "loss": 0.7932, + "num_input_tokens_seen": 12872225, + "step": 613 + }, + { + "epoch": 0.07382913485240185, + "flos": 18970566113400.0, + "grad_norm": 3.6438423315163737, + "learning_rate": 3.9799341052973375e-06, + "loss": 0.713, + "num_input_tokens_seen": 12891780, + "step": 614 + }, + { + "epoch": 0.07394937774304094, + "flos": 12301016887200.0, + "grad_norm": 2.8864243787027797, + "learning_rate": 3.979823887684241e-06, + "loss": 0.7172, + "num_input_tokens_seen": 12910440, + "step": 615 + }, + { + "epoch": 0.07406962063368003, + "flos": 15134998193280.0, + "grad_norm": 3.784779438360139, + "learning_rate": 3.979713369733434e-06, + "loss": 0.8485, + "num_input_tokens_seen": 12928025, + "step": 616 + }, + { + "epoch": 0.07418986352431912, + "flos": 15673096026120.0, + "grad_norm": 3.2851643783341773, + "learning_rate": 3.979602551461683e-06, + "loss": 0.8267, + "num_input_tokens_seen": 12948525, + "step": 617 + }, + { + "epoch": 0.07431010641495822, + "flos": 8709665136240.0, + "grad_norm": 3.299997017598972, + "learning_rate": 3.979491432885799e-06, + "loss": 0.9018, + "num_input_tokens_seen": 12964510, + "step": 618 + }, + { + "epoch": 0.0744303493055973, + "flos": 15322868616360.0, + "grad_norm": 2.434061412666559, + "learning_rate": 3.97938001402264e-06, + "loss": 0.8154, + "num_input_tokens_seen": 12983355, + "step": 619 + }, + { + "epoch": 0.0745505921962364, + "flos": 11841580580880.0, + "grad_norm": 4.417459716613202, + "learning_rate": 3.979268294889105e-06, + "loss": 0.7913, + "num_input_tokens_seen": 12998625, + "step": 620 + }, + { + "epoch": 0.07467083508687548, + "flos": 37514708276040.0, + "grad_norm": 3.333784422551915, + "learning_rate": 3.979156275502143e-06, + "loss": 0.7428, + "num_input_tokens_seen": 13022005, + "step": 621 + }, + { + "epoch": 0.07479107797751458, + "flos": 12786251569560.0, + "grad_norm": 3.1159379409760652, + "learning_rate": 3.979043955878749e-06, + "loss": 0.8984, + "num_input_tokens_seen": 13039570, + "step": 622 + }, + { + "epoch": 0.07491132086815366, + "flos": 17184987519480.0, + "grad_norm": 2.899212592202434, + "learning_rate": 3.978931336035959e-06, + "loss": 0.8119, + "num_input_tokens_seen": 13058100, + "step": 623 + }, + { + "epoch": 0.07503156375879276, + "flos": 14731179453960.0, + "grad_norm": 3.147964934293116, + "learning_rate": 3.9788184159908595e-06, + "loss": 0.8084, + "num_input_tokens_seen": 13074950, + "step": 624 + }, + { + "epoch": 0.07515180664943186, + "flos": 10998741718080.0, + "grad_norm": 3.0533214744048833, + "learning_rate": 3.97870519576058e-06, + "loss": 0.8018, + "num_input_tokens_seen": 13091095, + "step": 625 + }, + { + "epoch": 0.07527204954007094, + "flos": 15567433045080.0, + "grad_norm": 2.939918496435919, + "learning_rate": 3.978591675362295e-06, + "loss": 0.7875, + "num_input_tokens_seen": 13109530, + "step": 626 + }, + { + "epoch": 0.07539229243071004, + "flos": 15594434499600.0, + "grad_norm": 2.1367208700500795, + "learning_rate": 3.978477854813226e-06, + "loss": 0.86, + "num_input_tokens_seen": 13128590, + "step": 627 + }, + { + "epoch": 0.07551253532134912, + "flos": 9465547563000.0, + "grad_norm": 2.8571082964577994, + "learning_rate": 3.97836373413064e-06, + "loss": 0.8059, + "num_input_tokens_seen": 13146365, + "step": 628 + }, + { + "epoch": 0.07563277821198822, + "flos": 14028983336640.0, + "grad_norm": 2.639471309927228, + "learning_rate": 3.978249313331848e-06, + "loss": 0.7394, + "num_input_tokens_seen": 13164315, + "step": 629 + }, + { + "epoch": 0.07575302110262731, + "flos": 14272059747240.0, + "grad_norm": 4.070546793280856, + "learning_rate": 3.978134592434208e-06, + "loss": 0.592, + "num_input_tokens_seen": 13181785, + "step": 630 + }, + { + "epoch": 0.0758732639932664, + "flos": 50115816670440.0, + "grad_norm": 0.9858666879703285, + "learning_rate": 3.978019571455123e-06, + "loss": 0.6381, + "num_input_tokens_seen": 13233450, + "step": 631 + }, + { + "epoch": 0.07599350688390549, + "flos": 13866404730240.0, + "grad_norm": 2.785941378451144, + "learning_rate": 3.977904250412042e-06, + "loss": 0.8284, + "num_input_tokens_seen": 13252125, + "step": 632 + }, + { + "epoch": 0.07611374977454458, + "flos": 15405614277720.0, + "grad_norm": 2.6386832597714243, + "learning_rate": 3.97778862932246e-06, + "loss": 0.8417, + "num_input_tokens_seen": 13269010, + "step": 633 + }, + { + "epoch": 0.07623399266518367, + "flos": 13515069221880.0, + "grad_norm": 2.829624179108441, + "learning_rate": 3.9776727082039144e-06, + "loss": 0.9273, + "num_input_tokens_seen": 13285700, + "step": 634 + }, + { + "epoch": 0.07635423555582276, + "flos": 32870209310280.0, + "grad_norm": 0.8115480054310843, + "learning_rate": 3.977556487073991e-06, + "loss": 0.5722, + "num_input_tokens_seen": 13339975, + "step": 635 + }, + { + "epoch": 0.07647447844646185, + "flos": 15917692114800.0, + "grad_norm": 2.197852786131691, + "learning_rate": 3.97743996595032e-06, + "loss": 0.7913, + "num_input_tokens_seen": 13359735, + "step": 636 + }, + { + "epoch": 0.07659472133710095, + "flos": 17431958105160.0, + "grad_norm": 2.338172073441031, + "learning_rate": 3.9773231448505804e-06, + "loss": 0.8007, + "num_input_tokens_seen": 13381245, + "step": 637 + }, + { + "epoch": 0.07671496422774003, + "flos": 15702313677840.0, + "grad_norm": 3.1649637328084306, + "learning_rate": 3.977206023792491e-06, + "loss": 0.7499, + "num_input_tokens_seen": 13400855, + "step": 638 + }, + { + "epoch": 0.07683520711837913, + "flos": 12379963353360.0, + "grad_norm": 3.072372405566475, + "learning_rate": 3.97708860279382e-06, + "loss": 0.8005, + "num_input_tokens_seen": 13418685, + "step": 639 + }, + { + "epoch": 0.07695545000901821, + "flos": 17187456996360.0, + "grad_norm": 2.5980002268731632, + "learning_rate": 3.97697088187238e-06, + "loss": 0.774, + "num_input_tokens_seen": 13438920, + "step": 640 + }, + { + "epoch": 0.07707569289965731, + "flos": 12862950178560.0, + "grad_norm": 3.3501485763169425, + "learning_rate": 3.976852861046029e-06, + "loss": 0.8967, + "num_input_tokens_seen": 13455255, + "step": 641 + }, + { + "epoch": 0.0771959357902964, + "flos": 18888865230720.0, + "grad_norm": 2.6312873565849064, + "learning_rate": 3.97673454033267e-06, + "loss": 0.786, + "num_input_tokens_seen": 13477075, + "step": 642 + }, + { + "epoch": 0.07731617868093549, + "flos": 14487691463880.0, + "grad_norm": 2.54361192858089, + "learning_rate": 3.976615919750254e-06, + "loss": 0.8023, + "num_input_tokens_seen": 13494495, + "step": 643 + }, + { + "epoch": 0.07743642157157458, + "flos": 15594846079080.0, + "grad_norm": 2.476589744249544, + "learning_rate": 3.976496999316775e-06, + "loss": 0.845, + "num_input_tokens_seen": 13512970, + "step": 644 + }, + { + "epoch": 0.07755666446221367, + "flos": 14590884968040.0, + "grad_norm": 3.8459480017818324, + "learning_rate": 3.976377779050271e-06, + "loss": 0.827, + "num_input_tokens_seen": 13530820, + "step": 645 + }, + { + "epoch": 0.07767690735285276, + "flos": 17295937713840.0, + "grad_norm": 3.1907870230041797, + "learning_rate": 3.976258258968831e-06, + "loss": 0.8153, + "num_input_tokens_seen": 13549085, + "step": 646 + }, + { + "epoch": 0.07779715024349185, + "flos": 16269407542680.0, + "grad_norm": 2.7085686703674225, + "learning_rate": 3.976138439090583e-06, + "loss": 0.7129, + "num_input_tokens_seen": 13566885, + "step": 647 + }, + { + "epoch": 0.07791739313413094, + "flos": 15320684079120.0, + "grad_norm": 2.5380229054132077, + "learning_rate": 3.976018319433706e-06, + "loss": 0.8278, + "num_input_tokens_seen": 13584150, + "step": 648 + }, + { + "epoch": 0.07803763602477004, + "flos": 14104795466760.0, + "grad_norm": 3.2555386225665424, + "learning_rate": 3.9758979000164205e-06, + "loss": 0.9082, + "num_input_tokens_seen": 13600690, + "step": 649 + }, + { + "epoch": 0.07815787891540912, + "flos": 16620648071160.0, + "grad_norm": 2.379764896769785, + "learning_rate": 3.975777180856995e-06, + "loss": 0.6927, + "num_input_tokens_seen": 13619530, + "step": 650 + }, + { + "epoch": 0.07827812180604822, + "flos": 16620933010800.0, + "grad_norm": 4.879330203570325, + "learning_rate": 3.975656161973742e-06, + "loss": 0.8338, + "num_input_tokens_seen": 13638335, + "step": 651 + }, + { + "epoch": 0.0783983646966873, + "flos": 15891513819240.0, + "grad_norm": 3.552263242280343, + "learning_rate": 3.9755348433850194e-06, + "loss": 0.865, + "num_input_tokens_seen": 13653395, + "step": 652 + }, + { + "epoch": 0.0785186075873264, + "flos": 44694471893040.0, + "grad_norm": 1.0591542354327437, + "learning_rate": 3.975413225109232e-06, + "loss": 0.7082, + "num_input_tokens_seen": 13713665, + "step": 653 + }, + { + "epoch": 0.0786388504779655, + "flos": 17484314696280.0, + "grad_norm": 3.66960404965244, + "learning_rate": 3.975291307164829e-06, + "loss": 0.9068, + "num_input_tokens_seen": 13732030, + "step": 654 + }, + { + "epoch": 0.07875909336860458, + "flos": 11031631925160.0, + "grad_norm": 2.5872769761835044, + "learning_rate": 3.975169089570306e-06, + "loss": 0.8448, + "num_input_tokens_seen": 13750125, + "step": 655 + }, + { + "epoch": 0.07887933625924368, + "flos": 16269945762000.0, + "grad_norm": 3.0658127909002624, + "learning_rate": 3.975046572344202e-06, + "loss": 0.9051, + "num_input_tokens_seen": 13766305, + "step": 656 + }, + { + "epoch": 0.07899957914988276, + "flos": 15190140860880.0, + "grad_norm": 3.1956407895995356, + "learning_rate": 3.974923755505103e-06, + "loss": 0.6992, + "num_input_tokens_seen": 13785255, + "step": 657 + }, + { + "epoch": 0.07911982204052186, + "flos": 16836976306920.0, + "grad_norm": 2.0761722195069683, + "learning_rate": 3.974800639071641e-06, + "loss": 0.8981, + "num_input_tokens_seen": 13805695, + "step": 658 + }, + { + "epoch": 0.07924006493116094, + "flos": 16916999211720.0, + "grad_norm": 2.399112767891306, + "learning_rate": 3.974677223062492e-06, + "loss": 0.9916, + "num_input_tokens_seen": 13822630, + "step": 659 + }, + { + "epoch": 0.07936030782180004, + "flos": 12003684287880.0, + "grad_norm": 6.338665206904732, + "learning_rate": 3.974553507496378e-06, + "loss": 0.7161, + "num_input_tokens_seen": 13840925, + "step": 660 + }, + { + "epoch": 0.07948055071243913, + "flos": 17377480296720.0, + "grad_norm": 3.3709443767601592, + "learning_rate": 3.974429492392068e-06, + "loss": 0.8723, + "num_input_tokens_seen": 13860670, + "step": 661 + }, + { + "epoch": 0.07960079360307822, + "flos": 14298744602160.0, + "grad_norm": 4.319963934359295, + "learning_rate": 3.974305177768373e-06, + "loss": 0.8855, + "num_input_tokens_seen": 13878600, + "step": 662 + }, + { + "epoch": 0.07972103649371731, + "flos": 17214616750680.0, + "grad_norm": 3.1483410235974594, + "learning_rate": 3.974180563644152e-06, + "loss": 0.838, + "num_input_tokens_seen": 13896885, + "step": 663 + }, + { + "epoch": 0.0798412793843564, + "flos": 12192092930280.0, + "grad_norm": 3.509313996923602, + "learning_rate": 3.97405565003831e-06, + "loss": 0.875, + "num_input_tokens_seen": 13912690, + "step": 664 + }, + { + "epoch": 0.07996152227499549, + "flos": 13299659124960.0, + "grad_norm": 3.5196912519558072, + "learning_rate": 3.973930436969794e-06, + "loss": 0.7667, + "num_input_tokens_seen": 13930865, + "step": 665 + }, + { + "epoch": 0.08008176516563459, + "flos": 15054753668760.0, + "grad_norm": 2.8989553477337826, + "learning_rate": 3.973804924457602e-06, + "loss": 0.8415, + "num_input_tokens_seen": 13948665, + "step": 666 + }, + { + "epoch": 0.08020200805627367, + "flos": 23372942958720.0, + "grad_norm": 1.9551547680445778, + "learning_rate": 3.973679112520771e-06, + "loss": 0.8458, + "num_input_tokens_seen": 13970100, + "step": 667 + }, + { + "epoch": 0.08032225094691277, + "flos": 12973868712960.0, + "grad_norm": 2.6373859832851734, + "learning_rate": 3.973553001178389e-06, + "loss": 0.9793, + "num_input_tokens_seen": 13987325, + "step": 668 + }, + { + "epoch": 0.08044249383755185, + "flos": 17620525047360.0, + "grad_norm": 3.550133372205785, + "learning_rate": 3.973426590449585e-06, + "loss": 0.7423, + "num_input_tokens_seen": 14005000, + "step": 669 + }, + { + "epoch": 0.08056273672819095, + "flos": 13299912404640.0, + "grad_norm": 2.9961774632820886, + "learning_rate": 3.9732998803535364e-06, + "loss": 0.7418, + "num_input_tokens_seen": 14022780, + "step": 670 + }, + { + "epoch": 0.08068297961883003, + "flos": 14374525072320.0, + "grad_norm": 3.1795607782929527, + "learning_rate": 3.973172870909465e-06, + "loss": 0.8495, + "num_input_tokens_seen": 14037265, + "step": 671 + }, + { + "epoch": 0.08080322250946913, + "flos": 16944760505280.0, + "grad_norm": 5.7760091585208215, + "learning_rate": 3.973045562136638e-06, + "loss": 0.7929, + "num_input_tokens_seen": 14053800, + "step": 672 + }, + { + "epoch": 0.08092346540010822, + "flos": 15919876652040.0, + "grad_norm": 2.5399284881377104, + "learning_rate": 3.972917954054368e-06, + "loss": 0.8969, + "num_input_tokens_seen": 14072075, + "step": 673 + }, + { + "epoch": 0.08104370829074731, + "flos": 15378581163240.0, + "grad_norm": 2.7815507019019043, + "learning_rate": 3.972790046682013e-06, + "loss": 0.7988, + "num_input_tokens_seen": 14090470, + "step": 674 + }, + { + "epoch": 0.0811639511813864, + "flos": 14675941806480.0, + "grad_norm": 2.5794889598949577, + "learning_rate": 3.972661840038977e-06, + "loss": 0.7729, + "num_input_tokens_seen": 14110480, + "step": 675 + }, + { + "epoch": 0.08128419407202549, + "flos": 12273445553400.0, + "grad_norm": 4.05667776899822, + "learning_rate": 3.972533334144707e-06, + "loss": 0.8084, + "num_input_tokens_seen": 14127125, + "step": 676 + }, + { + "epoch": 0.08140443696266458, + "flos": 17404038511800.0, + "grad_norm": 2.9850447137956535, + "learning_rate": 3.972404529018699e-06, + "loss": 0.7683, + "num_input_tokens_seen": 14146705, + "step": 677 + }, + { + "epoch": 0.08152467985330367, + "flos": 17863601457960.0, + "grad_norm": 2.1335661398499024, + "learning_rate": 3.972275424680493e-06, + "loss": 0.8378, + "num_input_tokens_seen": 14166535, + "step": 678 + }, + { + "epoch": 0.08164492274394276, + "flos": 14109576120720.0, + "grad_norm": 3.5290189463668113, + "learning_rate": 3.972146021149673e-06, + "loss": 0.8916, + "num_input_tokens_seen": 14184530, + "step": 679 + }, + { + "epoch": 0.08176516563458186, + "flos": 10788745474320.0, + "grad_norm": 2.6778317801736007, + "learning_rate": 3.972016318445868e-06, + "loss": 0.7746, + "num_input_tokens_seen": 14202250, + "step": 680 + }, + { + "epoch": 0.08188540852522094, + "flos": 16540498526520.0, + "grad_norm": 2.9897074568409256, + "learning_rate": 3.971886316588757e-06, + "loss": 0.8934, + "num_input_tokens_seen": 14222475, + "step": 681 + }, + { + "epoch": 0.08200565141586004, + "flos": 14217613598760.0, + "grad_norm": 3.3781065595407593, + "learning_rate": 3.9717560155980595e-06, + "loss": 0.7167, + "num_input_tokens_seen": 14237845, + "step": 682 + }, + { + "epoch": 0.08212589430649912, + "flos": 15054563709000.0, + "grad_norm": 2.994038380409352, + "learning_rate": 3.971625415493542e-06, + "loss": 0.8978, + "num_input_tokens_seen": 14255885, + "step": 683 + }, + { + "epoch": 0.08224613719713822, + "flos": 19019915008320.0, + "grad_norm": 2.4225824122784774, + "learning_rate": 3.971494516295017e-06, + "loss": 0.8618, + "num_input_tokens_seen": 14275055, + "step": 684 + }, + { + "epoch": 0.08236638008777732, + "flos": 17403563612400.0, + "grad_norm": 2.7901311798430593, + "learning_rate": 3.971363318022341e-06, + "loss": 0.8255, + "num_input_tokens_seen": 14296115, + "step": 685 + }, + { + "epoch": 0.0824866229784164, + "flos": 28526833156320.0, + "grad_norm": 2.49271207926865, + "learning_rate": 3.971231820695417e-06, + "loss": 0.6748, + "num_input_tokens_seen": 14319450, + "step": 686 + }, + { + "epoch": 0.0826068658690555, + "flos": 16913801555760.0, + "grad_norm": 2.2979302090902607, + "learning_rate": 3.971100024334193e-06, + "loss": 0.8037, + "num_input_tokens_seen": 14336690, + "step": 687 + }, + { + "epoch": 0.08272710875969458, + "flos": 15454678233000.0, + "grad_norm": 2.537824029387811, + "learning_rate": 3.970967928958663e-06, + "loss": 0.8537, + "num_input_tokens_seen": 14353525, + "step": 688 + }, + { + "epoch": 0.08284735165033368, + "flos": 13921135818360.0, + "grad_norm": 1.87943493548654, + "learning_rate": 3.970835534588865e-06, + "loss": 0.8221, + "num_input_tokens_seen": 14370740, + "step": 689 + }, + { + "epoch": 0.08296759454097276, + "flos": 12192599489640.0, + "grad_norm": 2.318874234472735, + "learning_rate": 3.970702841244883e-06, + "loss": 0.8424, + "num_input_tokens_seen": 14388780, + "step": 690 + }, + { + "epoch": 0.08308783743161186, + "flos": 13138093637280.0, + "grad_norm": 2.805519596872397, + "learning_rate": 3.970569848946847e-06, + "loss": 0.801, + "num_input_tokens_seen": 14408315, + "step": 691 + }, + { + "epoch": 0.08320808032225095, + "flos": 10972405122720.0, + "grad_norm": 2.7870885023526415, + "learning_rate": 3.970436557714932e-06, + "loss": 0.7886, + "num_input_tokens_seen": 14424555, + "step": 692 + }, + { + "epoch": 0.08332832321289003, + "flos": 16378204859760.0, + "grad_norm": 2.456821936104918, + "learning_rate": 3.970302967569358e-06, + "loss": 0.8382, + "num_input_tokens_seen": 14442865, + "step": 693 + }, + { + "epoch": 0.08344856610352913, + "flos": 18105633089880.0, + "grad_norm": 3.235405504379783, + "learning_rate": 3.9701690785303896e-06, + "loss": 0.6671, + "num_input_tokens_seen": 14461780, + "step": 694 + }, + { + "epoch": 0.08356880899416821, + "flos": 18588208335600.0, + "grad_norm": 2.6448502003123515, + "learning_rate": 3.970034890618339e-06, + "loss": 0.8614, + "num_input_tokens_seen": 14481190, + "step": 695 + }, + { + "epoch": 0.08368905188480731, + "flos": 17836093444080.0, + "grad_norm": 2.7546910243193032, + "learning_rate": 3.969900403853562e-06, + "loss": 0.8578, + "num_input_tokens_seen": 14499950, + "step": 696 + }, + { + "epoch": 0.08380929477544641, + "flos": 13161485856360.0, + "grad_norm": 2.8790102646152143, + "learning_rate": 3.96976561825646e-06, + "loss": 0.7673, + "num_input_tokens_seen": 14516760, + "step": 697 + }, + { + "epoch": 0.08392953766608549, + "flos": 19266727294200.0, + "grad_norm": 2.5206983125535185, + "learning_rate": 3.969630533847479e-06, + "loss": 0.86, + "num_input_tokens_seen": 14535440, + "step": 698 + }, + { + "epoch": 0.08404978055672459, + "flos": 16458860963760.0, + "grad_norm": 2.827095729619854, + "learning_rate": 3.969495150647113e-06, + "loss": 0.8257, + "num_input_tokens_seen": 14553330, + "step": 699 + }, + { + "epoch": 0.08417002344736367, + "flos": 18186669113400.0, + "grad_norm": 2.6485166284673975, + "learning_rate": 3.969359468675899e-06, + "loss": 0.7519, + "num_input_tokens_seen": 14573180, + "step": 700 + }, + { + "epoch": 0.08429026633800277, + "flos": 12354228297240.0, + "grad_norm": 3.0400032999623106, + "learning_rate": 3.969223487954418e-06, + "loss": 0.8763, + "num_input_tokens_seen": 14590360, + "step": 701 + }, + { + "epoch": 0.08441050922864185, + "flos": 17457756481200.0, + "grad_norm": 2.2605627923895812, + "learning_rate": 3.969087208503301e-06, + "loss": 0.8147, + "num_input_tokens_seen": 14610160, + "step": 702 + }, + { + "epoch": 0.08453075211928095, + "flos": 18699791729160.0, + "grad_norm": 4.638112219528275, + "learning_rate": 3.968950630343219e-06, + "loss": 0.8221, + "num_input_tokens_seen": 14626865, + "step": 703 + }, + { + "epoch": 0.08465099500992004, + "flos": 14267912292480.0, + "grad_norm": 2.527755680150662, + "learning_rate": 3.968813753494892e-06, + "loss": 0.917, + "num_input_tokens_seen": 14644745, + "step": 704 + }, + { + "epoch": 0.08477123790055913, + "flos": 21535071093600.0, + "grad_norm": 2.894662377891837, + "learning_rate": 3.968676577979084e-06, + "loss": 0.7428, + "num_input_tokens_seen": 14664015, + "step": 705 + }, + { + "epoch": 0.08489148079119822, + "flos": 13596516824880.0, + "grad_norm": 2.6135870465235818, + "learning_rate": 3.968539103816605e-06, + "loss": 0.7672, + "num_input_tokens_seen": 14681535, + "step": 706 + }, + { + "epoch": 0.0850117236818373, + "flos": 17182739662320.0, + "grad_norm": 2.4562634843839573, + "learning_rate": 3.9684013310283085e-06, + "loss": 0.8787, + "num_input_tokens_seen": 14699940, + "step": 707 + }, + { + "epoch": 0.0851319665724764, + "flos": 29878362240480.0, + "grad_norm": 2.24512409011689, + "learning_rate": 3.9682632596350956e-06, + "loss": 0.6159, + "num_input_tokens_seen": 14720825, + "step": 708 + }, + { + "epoch": 0.0852522094631155, + "flos": 11563872665400.0, + "grad_norm": 2.053612605564674, + "learning_rate": 3.968124889657911e-06, + "loss": 0.7683, + "num_input_tokens_seen": 14735645, + "step": 709 + }, + { + "epoch": 0.08537245235375458, + "flos": 10592706781560.0, + "grad_norm": 2.8112278074318917, + "learning_rate": 3.967986221117746e-06, + "loss": 0.8959, + "num_input_tokens_seen": 14751305, + "step": 710 + }, + { + "epoch": 0.08549269524439368, + "flos": 19401924526560.0, + "grad_norm": 2.7308188720109134, + "learning_rate": 3.967847254035635e-06, + "loss": 0.8557, + "num_input_tokens_seen": 14770410, + "step": 711 + }, + { + "epoch": 0.08561293813503276, + "flos": 9870917640360.0, + "grad_norm": 3.54112187103604, + "learning_rate": 3.967707988432661e-06, + "loss": 0.85, + "num_input_tokens_seen": 14787835, + "step": 712 + }, + { + "epoch": 0.08573318102567186, + "flos": 19753196715000.0, + "grad_norm": 2.740633880357263, + "learning_rate": 3.967568424329949e-06, + "loss": 0.8609, + "num_input_tokens_seen": 14807980, + "step": 713 + }, + { + "epoch": 0.08585342391631094, + "flos": 49627732591680.0, + "grad_norm": 0.8034630505576782, + "learning_rate": 3.967428561748671e-06, + "loss": 0.5872, + "num_input_tokens_seen": 14875670, + "step": 714 + }, + { + "epoch": 0.08597366680695004, + "flos": 16432682668200.0, + "grad_norm": 2.151886214626759, + "learning_rate": 3.967288400710045e-06, + "loss": 0.849, + "num_input_tokens_seen": 14894855, + "step": 715 + }, + { + "epoch": 0.08609390969758914, + "flos": 17242156424520.0, + "grad_norm": 2.6874229952403623, + "learning_rate": 3.9671479412353335e-06, + "loss": 0.8674, + "num_input_tokens_seen": 14913040, + "step": 716 + }, + { + "epoch": 0.08621415258822822, + "flos": 18970502793480.0, + "grad_norm": 3.3518211425005178, + "learning_rate": 3.967007183345843e-06, + "loss": 0.7261, + "num_input_tokens_seen": 14932615, + "step": 717 + }, + { + "epoch": 0.08633439547886732, + "flos": 9870347761080.0, + "grad_norm": 3.164760484355782, + "learning_rate": 3.966866127062927e-06, + "loss": 0.8764, + "num_input_tokens_seen": 14949460, + "step": 718 + }, + { + "epoch": 0.0864546383695064, + "flos": 42569113676160.0, + "grad_norm": 0.9082173007966258, + "learning_rate": 3.966724772407982e-06, + "loss": 0.6747, + "num_input_tokens_seen": 15006695, + "step": 719 + }, + { + "epoch": 0.0865748812601455, + "flos": 14649161971680.0, + "grad_norm": 14.23467032401518, + "learning_rate": 3.966583119402454e-06, + "loss": 0.8786, + "num_input_tokens_seen": 15023180, + "step": 720 + }, + { + "epoch": 0.08669512415078459, + "flos": 25909781625240.0, + "grad_norm": 1.975575196731492, + "learning_rate": 3.9664411680678305e-06, + "loss": 0.8044, + "num_input_tokens_seen": 15044655, + "step": 721 + }, + { + "epoch": 0.08681536704142367, + "flos": 48407348265000.0, + "grad_norm": 0.8881396701682985, + "learning_rate": 3.966298918425644e-06, + "loss": 0.6339, + "num_input_tokens_seen": 15101865, + "step": 722 + }, + { + "epoch": 0.08693560993206277, + "flos": 25367283057960.0, + "grad_norm": 2.6090880824434124, + "learning_rate": 3.966156370497476e-06, + "loss": 0.8165, + "num_input_tokens_seen": 15125195, + "step": 723 + }, + { + "epoch": 0.08705585282270185, + "flos": 16945362044520.0, + "grad_norm": 2.8917409031671224, + "learning_rate": 3.96601352430495e-06, + "loss": 0.868, + "num_input_tokens_seen": 15144685, + "step": 724 + }, + { + "epoch": 0.08717609571334095, + "flos": 21643393511280.0, + "grad_norm": 2.031345775987213, + "learning_rate": 3.965870379869735e-06, + "loss": 0.8157, + "num_input_tokens_seen": 15166450, + "step": 725 + }, + { + "epoch": 0.08729633860398003, + "flos": 15108439978200.0, + "grad_norm": 2.5675687203872717, + "learning_rate": 3.965726937213547e-06, + "loss": 0.8451, + "num_input_tokens_seen": 15184805, + "step": 726 + }, + { + "epoch": 0.08741658149461913, + "flos": 13407854902800.0, + "grad_norm": 2.5124245893254424, + "learning_rate": 3.965583196358144e-06, + "loss": 0.7938, + "num_input_tokens_seen": 15203560, + "step": 727 + }, + { + "epoch": 0.08753682438525823, + "flos": 13673753653200.0, + "grad_norm": 3.327918104447901, + "learning_rate": 3.965439157325335e-06, + "loss": 0.7243, + "num_input_tokens_seen": 15220645, + "step": 728 + }, + { + "epoch": 0.08765706727589731, + "flos": 20368436396280.0, + "grad_norm": 2.72068875343925, + "learning_rate": 3.965294820136968e-06, + "loss": 0.7358, + "num_input_tokens_seen": 15242165, + "step": 729 + }, + { + "epoch": 0.08777731016653641, + "flos": 17862714979080.0, + "grad_norm": 3.470636902769326, + "learning_rate": 3.965150184814938e-06, + "loss": 0.8424, + "num_input_tokens_seen": 15261370, + "step": 730 + }, + { + "epoch": 0.08789755305717549, + "flos": 16081283839920.0, + "grad_norm": 2.5184244253101054, + "learning_rate": 3.965005251381189e-06, + "loss": 0.7357, + "num_input_tokens_seen": 15279025, + "step": 731 + }, + { + "epoch": 0.08801779594781459, + "flos": 43009083598440.0, + "grad_norm": 1.0369091235725991, + "learning_rate": 3.964860019857705e-06, + "loss": 0.659, + "num_input_tokens_seen": 15343660, + "step": 732 + }, + { + "epoch": 0.08813803883845367, + "flos": 17052861303240.0, + "grad_norm": 10.296939091345491, + "learning_rate": 3.964714490266518e-06, + "loss": 0.8209, + "num_input_tokens_seen": 15364025, + "step": 733 + }, + { + "epoch": 0.08825828172909277, + "flos": 46756428684120.0, + "grad_norm": 0.8681848306821497, + "learning_rate": 3.964568662629706e-06, + "loss": 0.6519, + "num_input_tokens_seen": 15425050, + "step": 734 + }, + { + "epoch": 0.08837852461973186, + "flos": 19587990331920.0, + "grad_norm": 2.575672161926645, + "learning_rate": 3.9644225369693895e-06, + "loss": 0.8232, + "num_input_tokens_seen": 15445070, + "step": 735 + }, + { + "epoch": 0.08849876751037095, + "flos": 19990511012880.0, + "grad_norm": 2.3540338349353664, + "learning_rate": 3.964276113307735e-06, + "loss": 0.8563, + "num_input_tokens_seen": 15464755, + "step": 736 + }, + { + "epoch": 0.08861901040101004, + "flos": 14487786443760.0, + "grad_norm": 2.7967624748226605, + "learning_rate": 3.9641293916669574e-06, + "loss": 0.7803, + "num_input_tokens_seen": 15483435, + "step": 737 + }, + { + "epoch": 0.08873925329164913, + "flos": 17322052689480.0, + "grad_norm": 1.8590407511865208, + "learning_rate": 3.9639823720693115e-06, + "loss": 0.8098, + "num_input_tokens_seen": 15505010, + "step": 738 + }, + { + "epoch": 0.08885949618228822, + "flos": 52979997087000.0, + "grad_norm": 0.9014870345928282, + "learning_rate": 3.963835054537102e-06, + "loss": 0.6451, + "num_input_tokens_seen": 15573695, + "step": 739 + }, + { + "epoch": 0.08897973907292732, + "flos": 16350823485720.0, + "grad_norm": 3.0591767633601075, + "learning_rate": 3.963687439092676e-06, + "loss": 0.5972, + "num_input_tokens_seen": 15594100, + "step": 740 + }, + { + "epoch": 0.0890999819635664, + "flos": 15540969809880.0, + "grad_norm": 2.32418848867511, + "learning_rate": 3.963539525758427e-06, + "loss": 0.7902, + "num_input_tokens_seen": 15613380, + "step": 741 + }, + { + "epoch": 0.0892202248542055, + "flos": 18588809874840.0, + "grad_norm": 5.644221039269876, + "learning_rate": 3.9633913145567925e-06, + "loss": 0.6583, + "num_input_tokens_seen": 15633590, + "step": 742 + }, + { + "epoch": 0.08934046774484458, + "flos": 17913013672800.0, + "grad_norm": 2.1685769267652426, + "learning_rate": 3.9632428055102575e-06, + "loss": 0.8051, + "num_input_tokens_seen": 15653320, + "step": 743 + }, + { + "epoch": 0.08946071063548368, + "flos": 26287390409040.0, + "grad_norm": 2.1671863690671054, + "learning_rate": 3.9630939986413495e-06, + "loss": 0.6501, + "num_input_tokens_seen": 15674840, + "step": 744 + }, + { + "epoch": 0.08958095352612276, + "flos": 10438106485080.0, + "grad_norm": 2.8774474962947814, + "learning_rate": 3.962944893972643e-06, + "loss": 0.769, + "num_input_tokens_seen": 15693010, + "step": 745 + }, + { + "epoch": 0.08970119641676186, + "flos": 13025053885560.0, + "grad_norm": 2.7713442357705094, + "learning_rate": 3.962795491526756e-06, + "loss": 0.8961, + "num_input_tokens_seen": 15709890, + "step": 746 + }, + { + "epoch": 0.08982143930740095, + "flos": 15214894458240.0, + "grad_norm": 2.323535260912477, + "learning_rate": 3.962645791326354e-06, + "loss": 0.8694, + "num_input_tokens_seen": 15728865, + "step": 747 + }, + { + "epoch": 0.08994168219804004, + "flos": 17647114922400.0, + "grad_norm": 2.200781258666689, + "learning_rate": 3.962495793394146e-06, + "loss": 0.8163, + "num_input_tokens_seen": 15747775, + "step": 748 + }, + { + "epoch": 0.09006192508867913, + "flos": 42141776077920.0, + "grad_norm": 0.6837422236919171, + "learning_rate": 3.9623454977528864e-06, + "loss": 0.6044, + "num_input_tokens_seen": 15806150, + "step": 749 + }, + { + "epoch": 0.09018216797931822, + "flos": 14975585582880.0, + "grad_norm": 2.9525583520819247, + "learning_rate": 3.962194904425375e-06, + "loss": 0.8398, + "num_input_tokens_seen": 15826500, + "step": 750 + }, + { + "epoch": 0.09030241086995731, + "flos": 16566898441800.0, + "grad_norm": 2.323132891206678, + "learning_rate": 3.9620440134344566e-06, + "loss": 0.6629, + "num_input_tokens_seen": 15844375, + "step": 751 + }, + { + "epoch": 0.09042265376059641, + "flos": 15999487977360.0, + "grad_norm": 4.460772456174123, + "learning_rate": 3.9618928248030215e-06, + "loss": 0.8138, + "num_input_tokens_seen": 15863605, + "step": 752 + }, + { + "epoch": 0.0905428966512355, + "flos": 17808743730000.0, + "grad_norm": 3.029742780538167, + "learning_rate": 3.961741338554005e-06, + "loss": 0.8201, + "num_input_tokens_seen": 15881665, + "step": 753 + }, + { + "epoch": 0.09066313954187459, + "flos": 26340791778840.0, + "grad_norm": 3.7940123013310547, + "learning_rate": 3.9615895547103865e-06, + "loss": 0.7425, + "num_input_tokens_seen": 15905030, + "step": 754 + }, + { + "epoch": 0.09078338243251367, + "flos": 21399937181160.0, + "grad_norm": 2.6704633229330934, + "learning_rate": 3.961437473295193e-06, + "loss": 0.7653, + "num_input_tokens_seen": 15924895, + "step": 755 + }, + { + "epoch": 0.09090362532315277, + "flos": 16026236152200.0, + "grad_norm": 2.870319038761605, + "learning_rate": 3.961285094331495e-06, + "loss": 0.7001, + "num_input_tokens_seen": 15942530, + "step": 756 + }, + { + "epoch": 0.09102386821379185, + "flos": 20049896115120.0, + "grad_norm": 2.590503453255159, + "learning_rate": 3.961132417842406e-06, + "loss": 0.8405, + "num_input_tokens_seen": 15962035, + "step": 757 + }, + { + "epoch": 0.09114411110443095, + "flos": 15216509116200.0, + "grad_norm": 3.0425603590081836, + "learning_rate": 3.960979443851089e-06, + "loss": 0.7412, + "num_input_tokens_seen": 15978780, + "step": 758 + }, + { + "epoch": 0.09126435399507005, + "flos": 19163660429880.0, + "grad_norm": 1.9644146369141766, + "learning_rate": 3.96082617238075e-06, + "loss": 0.7788, + "num_input_tokens_seen": 16001125, + "step": 759 + }, + { + "epoch": 0.09138459688570913, + "flos": 17862525019320.0, + "grad_norm": 2.7237965017255785, + "learning_rate": 3.960672603454639e-06, + "loss": 0.769, + "num_input_tokens_seen": 16020825, + "step": 760 + }, + { + "epoch": 0.09150483977634823, + "flos": 15378739463040.0, + "grad_norm": 3.732482047322069, + "learning_rate": 3.960518737096054e-06, + "loss": 0.744, + "num_input_tokens_seen": 16040175, + "step": 761 + }, + { + "epoch": 0.09162508266698731, + "flos": 16728875508960.0, + "grad_norm": 2.9734411819924325, + "learning_rate": 3.960364573328334e-06, + "loss": 0.7057, + "num_input_tokens_seen": 16059220, + "step": 762 + }, + { + "epoch": 0.0917453255576264, + "flos": 15890469040560.0, + "grad_norm": 2.4165090860314007, + "learning_rate": 3.9602101121748675e-06, + "loss": 0.8718, + "num_input_tokens_seen": 16079435, + "step": 763 + }, + { + "epoch": 0.0918655684482655, + "flos": 10624678849800.0, + "grad_norm": 2.4319057335402974, + "learning_rate": 3.960055353659085e-06, + "loss": 0.7049, + "num_input_tokens_seen": 16096265, + "step": 764 + }, + { + "epoch": 0.09198581133890459, + "flos": 17155833187680.0, + "grad_norm": 2.1011265866302375, + "learning_rate": 3.959900297804465e-06, + "loss": 0.8256, + "num_input_tokens_seen": 16116155, + "step": 765 + }, + { + "epoch": 0.09210605422954368, + "flos": 12242613243720.0, + "grad_norm": 3.1504012200496048, + "learning_rate": 3.9597449446345276e-06, + "loss": 0.7551, + "num_input_tokens_seen": 16133120, + "step": 766 + }, + { + "epoch": 0.09222629712018277, + "flos": 16593583296720.0, + "grad_norm": 2.754970912883648, + "learning_rate": 3.95958929417284e-06, + "loss": 0.8137, + "num_input_tokens_seen": 16150995, + "step": 767 + }, + { + "epoch": 0.09234654001082186, + "flos": 51608178459840.0, + "grad_norm": 0.7760766211367277, + "learning_rate": 3.9594333464430145e-06, + "loss": 0.6033, + "num_input_tokens_seen": 16220205, + "step": 768 + }, + { + "epoch": 0.09246678290146094, + "flos": 14623015336080.0, + "grad_norm": 3.3923998990678137, + "learning_rate": 3.959277101468709e-06, + "loss": 0.8647, + "num_input_tokens_seen": 16239475, + "step": 769 + }, + { + "epoch": 0.09258702579210004, + "flos": 12947247177960.0, + "grad_norm": 6.171673194386636, + "learning_rate": 3.959120559273624e-06, + "loss": 0.7691, + "num_input_tokens_seen": 16256980, + "step": 770 + }, + { + "epoch": 0.09270726868273914, + "flos": 15270575345160.0, + "grad_norm": 2.389954017771423, + "learning_rate": 3.958963719881509e-06, + "loss": 0.8296, + "num_input_tokens_seen": 16274790, + "step": 771 + }, + { + "epoch": 0.09282751157337822, + "flos": 12405413469840.0, + "grad_norm": 2.972134397720398, + "learning_rate": 3.958806583316154e-06, + "loss": 0.912, + "num_input_tokens_seen": 16292480, + "step": 772 + }, + { + "epoch": 0.09294775446401732, + "flos": 23882551318920.0, + "grad_norm": 2.003564043179571, + "learning_rate": 3.9586491496013985e-06, + "loss": 0.7768, + "num_input_tokens_seen": 16314595, + "step": 773 + }, + { + "epoch": 0.0930679973546564, + "flos": 13326597259560.0, + "grad_norm": 2.214360475376782, + "learning_rate": 3.958491418761124e-06, + "loss": 0.7991, + "num_input_tokens_seen": 16331885, + "step": 774 + }, + { + "epoch": 0.0931882402452955, + "flos": 15427550138640.0, + "grad_norm": 3.8911987862064326, + "learning_rate": 3.958333390819258e-06, + "loss": 0.7209, + "num_input_tokens_seen": 16348535, + "step": 775 + }, + { + "epoch": 0.0933084831359346, + "flos": 17728214265840.0, + "grad_norm": 2.522090304234974, + "learning_rate": 3.9581750657997754e-06, + "loss": 0.7828, + "num_input_tokens_seen": 16367620, + "step": 776 + }, + { + "epoch": 0.09342872602657368, + "flos": 18669750918480.0, + "grad_norm": 1.824036027468864, + "learning_rate": 3.95801644372669e-06, + "loss": 0.885, + "num_input_tokens_seen": 16387245, + "step": 777 + }, + { + "epoch": 0.09354896891721277, + "flos": 16945045444920.0, + "grad_norm": 2.005049173223, + "learning_rate": 3.957857524624068e-06, + "loss": 0.8222, + "num_input_tokens_seen": 16405845, + "step": 778 + }, + { + "epoch": 0.09366921180785186, + "flos": 17781615635640.0, + "grad_norm": 1.7039697634382818, + "learning_rate": 3.957698308516016e-06, + "loss": 0.8888, + "num_input_tokens_seen": 16426865, + "step": 779 + }, + { + "epoch": 0.09378945469849095, + "flos": 13674070252800.0, + "grad_norm": 2.1436947162050384, + "learning_rate": 3.957538795426688e-06, + "loss": 0.8161, + "num_input_tokens_seen": 16444010, + "step": 780 + }, + { + "epoch": 0.09390969758913004, + "flos": 16996230617520.0, + "grad_norm": 2.944962781790767, + "learning_rate": 3.9573789853802804e-06, + "loss": 0.7576, + "num_input_tokens_seen": 16462205, + "step": 781 + }, + { + "epoch": 0.09402994047976913, + "flos": 14353127430720.0, + "grad_norm": 2.379052268872747, + "learning_rate": 3.957218878401037e-06, + "loss": 0.7407, + "num_input_tokens_seen": 16480415, + "step": 782 + }, + { + "epoch": 0.09415018337040823, + "flos": 21585718046880.0, + "grad_norm": 2.2727805955006795, + "learning_rate": 3.957058474513246e-06, + "loss": 0.8707, + "num_input_tokens_seen": 16499990, + "step": 783 + }, + { + "epoch": 0.09427042626104731, + "flos": 17998197151080.0, + "grad_norm": 3.294430518018147, + "learning_rate": 3.956897773741241e-06, + "loss": 0.7679, + "num_input_tokens_seen": 16518700, + "step": 784 + }, + { + "epoch": 0.09439066915168641, + "flos": 19321015142880.0, + "grad_norm": 1.8228843104806083, + "learning_rate": 3.956736776109398e-06, + "loss": 0.7097, + "num_input_tokens_seen": 16539595, + "step": 785 + }, + { + "epoch": 0.09451091204232549, + "flos": 14190643804200.0, + "grad_norm": 2.0436094603206985, + "learning_rate": 3.956575481642143e-06, + "loss": 0.8175, + "num_input_tokens_seen": 16558205, + "step": 786 + }, + { + "epoch": 0.09463115493296459, + "flos": 18586308738000.0, + "grad_norm": 4.279562958837309, + "learning_rate": 3.956413890363943e-06, + "loss": 0.7213, + "num_input_tokens_seen": 16574905, + "step": 787 + }, + { + "epoch": 0.09475139782360369, + "flos": 7305019621920.0, + "grad_norm": 2.8032550181979525, + "learning_rate": 3.956252002299312e-06, + "loss": 0.8081, + "num_input_tokens_seen": 16590525, + "step": 788 + }, + { + "epoch": 0.09487164071424277, + "flos": 12564319520880.0, + "grad_norm": 2.3282480911383048, + "learning_rate": 3.956089817472807e-06, + "loss": 0.8958, + "num_input_tokens_seen": 16607550, + "step": 789 + }, + { + "epoch": 0.09499188360488187, + "flos": 22644220286280.0, + "grad_norm": 2.3473227164019272, + "learning_rate": 3.955927335909032e-06, + "loss": 0.8441, + "num_input_tokens_seen": 16630480, + "step": 790 + }, + { + "epoch": 0.09511212649552095, + "flos": 21535419353160.0, + "grad_norm": 2.3365943010836188, + "learning_rate": 3.955764557632634e-06, + "loss": 0.7516, + "num_input_tokens_seen": 16650010, + "step": 791 + }, + { + "epoch": 0.09523236938616005, + "flos": 7494061463520.0, + "grad_norm": 3.9066542759548017, + "learning_rate": 3.955601482668309e-06, + "loss": 0.9265, + "num_input_tokens_seen": 16667590, + "step": 792 + }, + { + "epoch": 0.09535261227679913, + "flos": 13920091039680.0, + "grad_norm": 2.9232101245986666, + "learning_rate": 3.955438111040794e-06, + "loss": 0.8683, + "num_input_tokens_seen": 16685585, + "step": 793 + }, + { + "epoch": 0.09547285516743823, + "flos": 15297545139720.0, + "grad_norm": 2.3349451106806254, + "learning_rate": 3.955274442774873e-06, + "loss": 0.7948, + "num_input_tokens_seen": 16703885, + "step": 794 + }, + { + "epoch": 0.09559309805807732, + "flos": 22129483012560.0, + "grad_norm": 2.948913477578172, + "learning_rate": 3.9551104778953725e-06, + "loss": 0.7034, + "num_input_tokens_seen": 16723900, + "step": 795 + }, + { + "epoch": 0.0957133409487164, + "flos": 15403619700240.0, + "grad_norm": 2.2996175572421635, + "learning_rate": 3.954946216427167e-06, + "loss": 0.8374, + "num_input_tokens_seen": 16744080, + "step": 796 + }, + { + "epoch": 0.0958335838393555, + "flos": 52581904309080.0, + "grad_norm": 0.8181931137860982, + "learning_rate": 3.954781658395176e-06, + "loss": 0.6405, + "num_input_tokens_seen": 16800055, + "step": 797 + }, + { + "epoch": 0.09595382672999458, + "flos": 16000786035720.0, + "grad_norm": 2.581666028946399, + "learning_rate": 3.95461680382436e-06, + "loss": 0.9047, + "num_input_tokens_seen": 16818700, + "step": 798 + }, + { + "epoch": 0.09607406962063368, + "flos": 13648841756040.0, + "grad_norm": 2.8109798137709445, + "learning_rate": 3.9544516527397295e-06, + "loss": 0.8417, + "num_input_tokens_seen": 16834770, + "step": 799 + }, + { + "epoch": 0.09619431251127276, + "flos": 16515808249080.0, + "grad_norm": 3.3982891883496036, + "learning_rate": 3.954286205166338e-06, + "loss": 0.7954, + "num_input_tokens_seen": 16855655, + "step": 800 + }, + { + "epoch": 0.09631455540191186, + "flos": 10356310622520.0, + "grad_norm": 2.9178834466070436, + "learning_rate": 3.954120461129282e-06, + "loss": 0.814, + "num_input_tokens_seen": 16872785, + "step": 801 + }, + { + "epoch": 0.09643479829255096, + "flos": 15162379567320.0, + "grad_norm": 2.520744185539636, + "learning_rate": 3.953954420653706e-06, + "loss": 0.8354, + "num_input_tokens_seen": 16889530, + "step": 802 + }, + { + "epoch": 0.09655504118319004, + "flos": 17885189059320.0, + "grad_norm": 2.0430461219738656, + "learning_rate": 3.953788083764798e-06, + "loss": 0.8704, + "num_input_tokens_seen": 16908485, + "step": 803 + }, + { + "epoch": 0.09667528407382914, + "flos": 13461414572400.0, + "grad_norm": 5.179113203751247, + "learning_rate": 3.953621450487792e-06, + "loss": 0.9062, + "num_input_tokens_seen": 16926825, + "step": 804 + }, + { + "epoch": 0.09679552696446822, + "flos": 52229686813200.0, + "grad_norm": 0.8384665264006176, + "learning_rate": 3.953454520847964e-06, + "loss": 0.6425, + "num_input_tokens_seen": 16991390, + "step": 805 + }, + { + "epoch": 0.09691576985510732, + "flos": 16054409025240.0, + "grad_norm": 2.7987068187249866, + "learning_rate": 3.9532872948706395e-06, + "loss": 0.7161, + "num_input_tokens_seen": 17010605, + "step": 806 + }, + { + "epoch": 0.09703601274574641, + "flos": 13108242786360.0, + "grad_norm": 2.6336922010715966, + "learning_rate": 3.9531197725811845e-06, + "loss": 0.8195, + "num_input_tokens_seen": 17025710, + "step": 807 + }, + { + "epoch": 0.0971562556363855, + "flos": 16214423174880.0, + "grad_norm": 2.1470460026809333, + "learning_rate": 3.952951954005013e-06, + "loss": 0.8612, + "num_input_tokens_seen": 17045115, + "step": 808 + }, + { + "epoch": 0.0972764985270246, + "flos": 18942931459680.0, + "grad_norm": 2.1623291345026985, + "learning_rate": 3.952783839167584e-06, + "loss": 0.8385, + "num_input_tokens_seen": 17064880, + "step": 809 + }, + { + "epoch": 0.09739674141766368, + "flos": 14865648507240.0, + "grad_norm": 3.150335832889146, + "learning_rate": 3.952615428094398e-06, + "loss": 0.719, + "num_input_tokens_seen": 17084120, + "step": 810 + }, + { + "epoch": 0.09751698430830277, + "flos": 11464193416800.0, + "grad_norm": 3.0950303944218414, + "learning_rate": 3.952446720811004e-06, + "loss": 0.7293, + "num_input_tokens_seen": 17102165, + "step": 811 + }, + { + "epoch": 0.09763722719894186, + "flos": 46972440320280.0, + "grad_norm": 0.8090019485723655, + "learning_rate": 3.952277717342995e-06, + "loss": 0.6655, + "num_input_tokens_seen": 17168320, + "step": 812 + }, + { + "epoch": 0.09775747008958095, + "flos": 16161939943920.0, + "grad_norm": 2.4728611383628567, + "learning_rate": 3.952108417716009e-06, + "loss": 0.8425, + "num_input_tokens_seen": 17187495, + "step": 813 + }, + { + "epoch": 0.09787771298022005, + "flos": 15565976686920.0, + "grad_norm": 2.0117538300946243, + "learning_rate": 3.951938821955727e-06, + "loss": 0.8378, + "num_input_tokens_seen": 17206615, + "step": 814 + }, + { + "epoch": 0.09799795587085913, + "flos": 16135065129240.0, + "grad_norm": 1.7669580204353568, + "learning_rate": 3.9517689300878786e-06, + "loss": 0.753, + "num_input_tokens_seen": 17226070, + "step": 815 + }, + { + "epoch": 0.09811819876149823, + "flos": 16215816213120.0, + "grad_norm": 2.1178167892982263, + "learning_rate": 3.951598742138236e-06, + "loss": 0.7692, + "num_input_tokens_seen": 17244515, + "step": 816 + }, + { + "epoch": 0.09823844165213731, + "flos": 16264880168400.0, + "grad_norm": 2.6387373819415436, + "learning_rate": 3.951428258132615e-06, + "loss": 0.7763, + "num_input_tokens_seen": 17262355, + "step": 817 + }, + { + "epoch": 0.09835868454277641, + "flos": 16454966788680.0, + "grad_norm": 2.5826075570086267, + "learning_rate": 3.951257478096879e-06, + "loss": 0.8305, + "num_input_tokens_seen": 17280440, + "step": 818 + }, + { + "epoch": 0.0984789274334155, + "flos": 11922553284480.0, + "grad_norm": 4.853289654573628, + "learning_rate": 3.951086402056936e-06, + "loss": 0.6814, + "num_input_tokens_seen": 17294760, + "step": 819 + }, + { + "epoch": 0.09859917032405459, + "flos": 17755025760600.0, + "grad_norm": 2.417349740756294, + "learning_rate": 3.950915030038735e-06, + "loss": 0.8234, + "num_input_tokens_seen": 17314275, + "step": 820 + }, + { + "epoch": 0.09871941321469369, + "flos": 12705215546040.0, + "grad_norm": 4.690734154475074, + "learning_rate": 3.9507433620682765e-06, + "loss": 0.8167, + "num_input_tokens_seen": 17330930, + "step": 821 + }, + { + "epoch": 0.09883965610533277, + "flos": 20888524203240.0, + "grad_norm": 1.9020321569281549, + "learning_rate": 3.9505713981716e-06, + "loss": 0.87, + "num_input_tokens_seen": 17353480, + "step": 822 + }, + { + "epoch": 0.09895989899597187, + "flos": 17347946045400.0, + "grad_norm": 3.703638998808617, + "learning_rate": 3.950399138374795e-06, + "loss": 0.7963, + "num_input_tokens_seen": 17372280, + "step": 823 + }, + { + "epoch": 0.09908014188661095, + "flos": 18077491876800.0, + "grad_norm": 2.199267425811512, + "learning_rate": 3.95022658270399e-06, + "loss": 0.7174, + "num_input_tokens_seen": 17392365, + "step": 824 + }, + { + "epoch": 0.09920038477725004, + "flos": 10221746589360.0, + "grad_norm": 2.164827287405546, + "learning_rate": 3.9500537311853635e-06, + "loss": 0.7677, + "num_input_tokens_seen": 17410040, + "step": 825 + }, + { + "epoch": 0.09932062766788914, + "flos": 9735277168560.0, + "grad_norm": 4.01766120553986, + "learning_rate": 3.949880583845136e-06, + "loss": 0.8103, + "num_input_tokens_seen": 17427835, + "step": 826 + }, + { + "epoch": 0.09944087055852822, + "flos": 14245089952680.0, + "grad_norm": 2.437858750768036, + "learning_rate": 3.949707140709575e-06, + "loss": 0.8004, + "num_input_tokens_seen": 17447285, + "step": 827 + }, + { + "epoch": 0.09956111344916732, + "flos": 12948703536120.0, + "grad_norm": 2.349071108076718, + "learning_rate": 3.949533401804991e-06, + "loss": 0.8127, + "num_input_tokens_seen": 17463910, + "step": 828 + }, + { + "epoch": 0.0996813563398064, + "flos": 13110332343720.0, + "grad_norm": 2.4785775732260276, + "learning_rate": 3.949359367157739e-06, + "loss": 0.9021, + "num_input_tokens_seen": 17482325, + "step": 829 + }, + { + "epoch": 0.0998015992304455, + "flos": 12732533600160.0, + "grad_norm": 4.247888826082877, + "learning_rate": 3.949185036794222e-06, + "loss": 0.7471, + "num_input_tokens_seen": 17500055, + "step": 830 + }, + { + "epoch": 0.0999218421210846, + "flos": 18973067250240.0, + "grad_norm": 1.9314146767932308, + "learning_rate": 3.949010410740884e-06, + "loss": 0.7621, + "num_input_tokens_seen": 17522600, + "step": 831 + }, + { + "epoch": 0.10004208501172368, + "flos": 15514474914720.0, + "grad_norm": 2.2030287917753952, + "learning_rate": 3.948835489024216e-06, + "loss": 0.853, + "num_input_tokens_seen": 17542055, + "step": 832 + }, + { + "epoch": 0.10016232790236278, + "flos": 12652257415680.0, + "grad_norm": 2.2261212898929053, + "learning_rate": 3.948660271670755e-06, + "loss": 0.8804, + "num_input_tokens_seen": 17558925, + "step": 833 + }, + { + "epoch": 0.10028257079300186, + "flos": 18808145806800.0, + "grad_norm": 2.5855833427645667, + "learning_rate": 3.948484758707079e-06, + "loss": 0.8311, + "num_input_tokens_seen": 17578245, + "step": 834 + }, + { + "epoch": 0.10040281368364096, + "flos": 18430378723200.0, + "grad_norm": 2.531789086474929, + "learning_rate": 3.948308950159815e-06, + "loss": 0.8251, + "num_input_tokens_seen": 17596645, + "step": 835 + }, + { + "epoch": 0.10052305657428004, + "flos": 12840824357880.0, + "grad_norm": 2.602893294226475, + "learning_rate": 3.9481328460556326e-06, + "loss": 0.7387, + "num_input_tokens_seen": 17613585, + "step": 836 + }, + { + "epoch": 0.10064329946491914, + "flos": 13622790100320.0, + "grad_norm": 3.013408894168222, + "learning_rate": 3.9479564464212455e-06, + "loss": 0.8641, + "num_input_tokens_seen": 17632465, + "step": 837 + }, + { + "epoch": 0.10076354235555823, + "flos": 12541243901400.0, + "grad_norm": 3.0606243072627386, + "learning_rate": 3.947779751283414e-06, + "loss": 0.7515, + "num_input_tokens_seen": 17649355, + "step": 838 + }, + { + "epoch": 0.10088378524619732, + "flos": 16806175657200.0, + "grad_norm": 2.3468217175729955, + "learning_rate": 3.947602760668944e-06, + "loss": 0.7405, + "num_input_tokens_seen": 17668865, + "step": 839 + }, + { + "epoch": 0.10100402813683641, + "flos": 27850752014640.0, + "grad_norm": 1.9106112759252918, + "learning_rate": 3.947425474604684e-06, + "loss": 0.6969, + "num_input_tokens_seen": 17692520, + "step": 840 + }, + { + "epoch": 0.1011242710274755, + "flos": 15756759826320.0, + "grad_norm": 2.22784943448209, + "learning_rate": 3.947247893117528e-06, + "loss": 0.9116, + "num_input_tokens_seen": 17710745, + "step": 841 + }, + { + "epoch": 0.10124451391811459, + "flos": 9893296740720.0, + "grad_norm": 3.627580447734434, + "learning_rate": 3.947070016234413e-06, + "loss": 0.6691, + "num_input_tokens_seen": 17726255, + "step": 842 + }, + { + "epoch": 0.10136475680875369, + "flos": 12133974226440.0, + "grad_norm": 3.428744453650984, + "learning_rate": 3.946891843982326e-06, + "loss": 0.7264, + "num_input_tokens_seen": 17743640, + "step": 843 + }, + { + "epoch": 0.10148499969939277, + "flos": 14215239101760.0, + "grad_norm": 3.0995507622352854, + "learning_rate": 3.9467133763882935e-06, + "loss": 0.7343, + "num_input_tokens_seen": 17761825, + "step": 844 + }, + { + "epoch": 0.10160524259003187, + "flos": 15432172492800.0, + "grad_norm": 2.5821076120693336, + "learning_rate": 3.9465346134793905e-06, + "loss": 0.8452, + "num_input_tokens_seen": 17781355, + "step": 845 + }, + { + "epoch": 0.10172548548067095, + "flos": 12921765401520.0, + "grad_norm": 2.08491073655167, + "learning_rate": 3.9463555552827335e-06, + "loss": 0.7846, + "num_input_tokens_seen": 17798245, + "step": 846 + }, + { + "epoch": 0.10184572837131005, + "flos": 15432045852960.0, + "grad_norm": 3.7087561691330584, + "learning_rate": 3.946176201825487e-06, + "loss": 0.8352, + "num_input_tokens_seen": 17816000, + "step": 847 + }, + { + "epoch": 0.10196597126194913, + "flos": 19104908526840.0, + "grad_norm": 2.463662905999559, + "learning_rate": 3.9459965531348575e-06, + "loss": 0.8168, + "num_input_tokens_seen": 17835375, + "step": 848 + }, + { + "epoch": 0.10208621415258823, + "flos": 21507911339280.0, + "grad_norm": 3.0669161038611015, + "learning_rate": 3.945816609238098e-06, + "loss": 0.8388, + "num_input_tokens_seen": 17854505, + "step": 849 + }, + { + "epoch": 0.10220645704322733, + "flos": 17431071626280.0, + "grad_norm": 2.1855078208781658, + "learning_rate": 3.945636370162507e-06, + "loss": 0.8406, + "num_input_tokens_seen": 17874335, + "step": 850 + }, + { + "epoch": 0.10232669993386641, + "flos": 16995977337840.0, + "grad_norm": 1.8845021118006484, + "learning_rate": 3.945455835935425e-06, + "loss": 0.7817, + "num_input_tokens_seen": 17893240, + "step": 851 + }, + { + "epoch": 0.1024469428245055, + "flos": 16776673065840.0, + "grad_norm": 4.074114819204117, + "learning_rate": 3.94527500658424e-06, + "loss": 0.7292, + "num_input_tokens_seen": 17910625, + "step": 852 + }, + { + "epoch": 0.10256718571514459, + "flos": 23021480810520.0, + "grad_norm": 2.8309359412078994, + "learning_rate": 3.945093882136382e-06, + "loss": 0.7898, + "num_input_tokens_seen": 17934120, + "step": 853 + }, + { + "epoch": 0.10268742860578368, + "flos": 17185810678440.0, + "grad_norm": 1.8422457490857223, + "learning_rate": 3.944912462619329e-06, + "loss": 0.8313, + "num_input_tokens_seen": 17952805, + "step": 854 + }, + { + "epoch": 0.10280767149642277, + "flos": 18699886709040.0, + "grad_norm": 2.4844776828106703, + "learning_rate": 3.9447307480606025e-06, + "loss": 0.7957, + "num_input_tokens_seen": 17972610, + "step": 855 + }, + { + "epoch": 0.10292791438706186, + "flos": 12651624216480.0, + "grad_norm": 4.0756353620173105, + "learning_rate": 3.944548738487767e-06, + "loss": 0.8846, + "num_input_tokens_seen": 17989845, + "step": 856 + }, + { + "epoch": 0.10304815727770096, + "flos": 19833947798880.0, + "grad_norm": 2.378402080259623, + "learning_rate": 3.944366433928434e-06, + "loss": 0.8935, + "num_input_tokens_seen": 18009545, + "step": 857 + }, + { + "epoch": 0.10316840016834004, + "flos": 16673606201520.0, + "grad_norm": 1.7982722944778609, + "learning_rate": 3.9441838344102594e-06, + "loss": 0.8191, + "num_input_tokens_seen": 18028990, + "step": 858 + }, + { + "epoch": 0.10328864305897914, + "flos": 15135979652040.0, + "grad_norm": 2.9491083187623777, + "learning_rate": 3.944000939960943e-06, + "loss": 0.666, + "num_input_tokens_seen": 18047435, + "step": 859 + }, + { + "epoch": 0.10340888594961822, + "flos": 20889822261600.0, + "grad_norm": 5.138597268201277, + "learning_rate": 3.943817750608229e-06, + "loss": 0.7941, + "num_input_tokens_seen": 18069705, + "step": 860 + }, + { + "epoch": 0.10352912884025732, + "flos": 9708054094320.0, + "grad_norm": 2.6072440211755805, + "learning_rate": 3.943634266379908e-06, + "loss": 0.8008, + "num_input_tokens_seen": 18086320, + "step": 861 + }, + { + "epoch": 0.10364937173089642, + "flos": 18728439501600.0, + "grad_norm": 1.8785388611141545, + "learning_rate": 3.943450487303815e-06, + "loss": 0.8407, + "num_input_tokens_seen": 18106535, + "step": 862 + }, + { + "epoch": 0.1037696146215355, + "flos": 15514094995200.0, + "grad_norm": 2.209459977087239, + "learning_rate": 3.943266413407827e-06, + "loss": 0.8343, + "num_input_tokens_seen": 18125530, + "step": 863 + }, + { + "epoch": 0.1038898575121746, + "flos": 18911655910560.0, + "grad_norm": 2.17007734152089, + "learning_rate": 3.94308204471987e-06, + "loss": 0.8296, + "num_input_tokens_seen": 18144265, + "step": 864 + }, + { + "epoch": 0.10401010040281368, + "flos": 13920629259000.0, + "grad_norm": 2.51682020902141, + "learning_rate": 3.942897381267912e-06, + "loss": 0.7303, + "num_input_tokens_seen": 18160350, + "step": 865 + }, + { + "epoch": 0.10413034329345278, + "flos": 11915999672760.0, + "grad_norm": 3.4551813553332025, + "learning_rate": 3.942712423079965e-06, + "loss": 0.6635, + "num_input_tokens_seen": 18176460, + "step": 866 + }, + { + "epoch": 0.10425058618409186, + "flos": 12569100174840.0, + "grad_norm": 3.0166467704608597, + "learning_rate": 3.942527170184088e-06, + "loss": 0.8954, + "num_input_tokens_seen": 18192800, + "step": 867 + }, + { + "epoch": 0.10437082907473096, + "flos": 13110142383960.0, + "grad_norm": 2.4020937409606953, + "learning_rate": 3.942341622608385e-06, + "loss": 0.7579, + "num_input_tokens_seen": 18209550, + "step": 868 + }, + { + "epoch": 0.10449107196537005, + "flos": 26665189152600.0, + "grad_norm": 2.551487402853274, + "learning_rate": 3.942155780381001e-06, + "loss": 0.7618, + "num_input_tokens_seen": 18233005, + "step": 869 + }, + { + "epoch": 0.10461131485600914, + "flos": 17427810650400.0, + "grad_norm": 2.33059653361197, + "learning_rate": 3.94196964353013e-06, + "loss": 0.7451, + "num_input_tokens_seen": 18252175, + "step": 870 + }, + { + "epoch": 0.10473155774664823, + "flos": 13434413117880.0, + "grad_norm": 3.1403618697587494, + "learning_rate": 3.941783212084008e-06, + "loss": 0.7915, + "num_input_tokens_seen": 18269650, + "step": 871 + }, + { + "epoch": 0.10485180063728732, + "flos": 18753319738800.0, + "grad_norm": 3.6959416376105083, + "learning_rate": 3.941596486070916e-06, + "loss": 0.7522, + "num_input_tokens_seen": 18287415, + "step": 872 + }, + { + "epoch": 0.10497204352792641, + "flos": 19860569333880.0, + "grad_norm": 6.709181390393258, + "learning_rate": 3.941409465519182e-06, + "loss": 0.5652, + "num_input_tokens_seen": 18307660, + "step": 873 + }, + { + "epoch": 0.10509228641856551, + "flos": 24131263202400.0, + "grad_norm": 1.7853232541420603, + "learning_rate": 3.941222150457176e-06, + "loss": 0.8397, + "num_input_tokens_seen": 18330635, + "step": 874 + }, + { + "epoch": 0.10521252930920459, + "flos": 10410281871600.0, + "grad_norm": 4.852565692376366, + "learning_rate": 3.941034540913311e-06, + "loss": 0.6996, + "num_input_tokens_seen": 18347885, + "step": 875 + }, + { + "epoch": 0.10533277219984369, + "flos": 15862676087040.0, + "grad_norm": 1.9854287180846701, + "learning_rate": 3.940846636916051e-06, + "loss": 0.8108, + "num_input_tokens_seen": 18367640, + "step": 876 + }, + { + "epoch": 0.10545301509048277, + "flos": 16293812880480.0, + "grad_norm": 2.475605102705647, + "learning_rate": 3.940658438493899e-06, + "loss": 0.8467, + "num_input_tokens_seen": 18385205, + "step": 877 + }, + { + "epoch": 0.10557325798112187, + "flos": 16242374428200.0, + "grad_norm": 3.3536882394968166, + "learning_rate": 3.940469945675405e-06, + "loss": 0.745, + "num_input_tokens_seen": 18403310, + "step": 878 + }, + { + "epoch": 0.10569350087176095, + "flos": 18889055190480.0, + "grad_norm": 2.12396271015918, + "learning_rate": 3.940281158489163e-06, + "loss": 0.8918, + "num_input_tokens_seen": 18422260, + "step": 879 + }, + { + "epoch": 0.10581374376240005, + "flos": 12624781061760.0, + "grad_norm": 2.116100750260702, + "learning_rate": 3.940092076963812e-06, + "loss": 0.8143, + "num_input_tokens_seen": 18439475, + "step": 880 + }, + { + "epoch": 0.10593398665303914, + "flos": 25233985423200.0, + "grad_norm": 2.356461399622616, + "learning_rate": 3.9399027011280355e-06, + "loss": 0.7752, + "num_input_tokens_seen": 18461290, + "step": 881 + }, + { + "epoch": 0.10605422954367823, + "flos": 17025574909080.0, + "grad_norm": 2.5870488017595545, + "learning_rate": 3.939713031010561e-06, + "loss": 0.755, + "num_input_tokens_seen": 18479375, + "step": 882 + }, + { + "epoch": 0.10617447243431732, + "flos": 16701525794880.0, + "grad_norm": 2.4212792432887027, + "learning_rate": 3.939523066640163e-06, + "loss": 0.7625, + "num_input_tokens_seen": 18497990, + "step": 883 + }, + { + "epoch": 0.10629471532495641, + "flos": 17860467121920.0, + "grad_norm": 2.2213265945375094, + "learning_rate": 3.939332808045657e-06, + "loss": 0.793, + "num_input_tokens_seen": 18517360, + "step": 884 + }, + { + "epoch": 0.1064149582155955, + "flos": 15432615732240.0, + "grad_norm": 4.581895544932029, + "learning_rate": 3.939142255255906e-06, + "loss": 0.8282, + "num_input_tokens_seen": 18537965, + "step": 885 + }, + { + "epoch": 0.1065352011062346, + "flos": 15133731794880.0, + "grad_norm": 2.4745700655421627, + "learning_rate": 3.938951408299817e-06, + "loss": 0.8628, + "num_input_tokens_seen": 18556525, + "step": 886 + }, + { + "epoch": 0.10665544399687368, + "flos": 46190189638200.0, + "grad_norm": 0.8179001350371006, + "learning_rate": 3.938760267206342e-06, + "loss": 0.5719, + "num_input_tokens_seen": 18618065, + "step": 887 + }, + { + "epoch": 0.10677568688751278, + "flos": 19158753136080.0, + "grad_norm": 2.5071969390923488, + "learning_rate": 3.938568832004475e-06, + "loss": 0.7592, + "num_input_tokens_seen": 18636490, + "step": 888 + }, + { + "epoch": 0.10689592977815186, + "flos": 9275999162040.0, + "grad_norm": 2.3117610678710543, + "learning_rate": 3.938377102723257e-06, + "loss": 0.7402, + "num_input_tokens_seen": 18653345, + "step": 889 + }, + { + "epoch": 0.10701617266879096, + "flos": 16188308199240.0, + "grad_norm": 2.5485585271092117, + "learning_rate": 3.938185079391774e-06, + "loss": 0.8194, + "num_input_tokens_seen": 18670110, + "step": 890 + }, + { + "epoch": 0.10713641555943004, + "flos": 14424823766040.0, + "grad_norm": 2.8951680542127063, + "learning_rate": 3.937992762039157e-06, + "loss": 1.0528, + "num_input_tokens_seen": 18683155, + "step": 891 + }, + { + "epoch": 0.10725665845006914, + "flos": 17540185542960.0, + "grad_norm": 2.421096873642961, + "learning_rate": 3.937800150694577e-06, + "loss": 0.7836, + "num_input_tokens_seen": 18704050, + "step": 892 + }, + { + "epoch": 0.10737690134070824, + "flos": 13542672215640.0, + "grad_norm": 2.447843564071499, + "learning_rate": 3.937607245387255e-06, + "loss": 0.7339, + "num_input_tokens_seen": 18723135, + "step": 893 + }, + { + "epoch": 0.10749714423134732, + "flos": 16617830334720.0, + "grad_norm": 2.7726346425418096, + "learning_rate": 3.937414046146455e-06, + "loss": 0.7026, + "num_input_tokens_seen": 18740810, + "step": 894 + }, + { + "epoch": 0.10761738712198642, + "flos": 15433122291600.0, + "grad_norm": 2.6609864136909187, + "learning_rate": 3.9372205530014845e-06, + "loss": 0.7362, + "num_input_tokens_seen": 18759010, + "step": 895 + }, + { + "epoch": 0.1077376300126255, + "flos": 17401790654640.0, + "grad_norm": 3.077493825576253, + "learning_rate": 3.937026765981696e-06, + "loss": 0.7122, + "num_input_tokens_seen": 18778800, + "step": 896 + }, + { + "epoch": 0.1078578729032646, + "flos": 15294537443520.0, + "grad_norm": 2.1283700810274384, + "learning_rate": 3.936832685116488e-06, + "loss": 0.7727, + "num_input_tokens_seen": 18796615, + "step": 897 + }, + { + "epoch": 0.10797811579390369, + "flos": 10788238914960.0, + "grad_norm": 2.7935916659485382, + "learning_rate": 3.936638310435301e-06, + "loss": 0.874, + "num_input_tokens_seen": 18814200, + "step": 898 + }, + { + "epoch": 0.10809835868454278, + "flos": 14271869787480.0, + "grad_norm": 3.0587024109150596, + "learning_rate": 3.936443641967623e-06, + "loss": 0.8122, + "num_input_tokens_seen": 18832750, + "step": 899 + }, + { + "epoch": 0.10821860157518187, + "flos": 13461731172000.0, + "grad_norm": 2.6645997699849975, + "learning_rate": 3.936248679742983e-06, + "loss": 0.8151, + "num_input_tokens_seen": 18850965, + "step": 900 + }, + { + "epoch": 0.10833884446582095, + "flos": 36344437174680.0, + "grad_norm": 1.006574654184535, + "learning_rate": 3.936053423790959e-06, + "loss": 0.7468, + "num_input_tokens_seen": 18899005, + "step": 901 + }, + { + "epoch": 0.10845908735646005, + "flos": 14919271496760.0, + "grad_norm": 2.2654316786796684, + "learning_rate": 3.935857874141168e-06, + "loss": 0.7546, + "num_input_tokens_seen": 18917560, + "step": 902 + }, + { + "epoch": 0.10857933024709913, + "flos": 10194681814920.0, + "grad_norm": 3.7261242238917047, + "learning_rate": 3.935662030823279e-06, + "loss": 0.8217, + "num_input_tokens_seen": 18933465, + "step": 903 + }, + { + "epoch": 0.10869957313773823, + "flos": 9706882675800.0, + "grad_norm": 3.732618868673095, + "learning_rate": 3.935465893866998e-06, + "loss": 0.6937, + "num_input_tokens_seen": 18951410, + "step": 904 + }, + { + "epoch": 0.10881981602837733, + "flos": 18911402630880.0, + "grad_norm": 2.1631555963670848, + "learning_rate": 3.935269463302079e-06, + "loss": 0.7958, + "num_input_tokens_seen": 18969335, + "step": 905 + }, + { + "epoch": 0.10894005891901641, + "flos": 15189982561080.0, + "grad_norm": 2.2345785835192395, + "learning_rate": 3.935072739158322e-06, + "loss": 0.7404, + "num_input_tokens_seen": 18988765, + "step": 906 + }, + { + "epoch": 0.10906030180965551, + "flos": 19535633740800.0, + "grad_norm": 1.638689261561331, + "learning_rate": 3.934875721465569e-06, + "loss": 0.7785, + "num_input_tokens_seen": 19008905, + "step": 907 + }, + { + "epoch": 0.10918054470029459, + "flos": 26850400139040.0, + "grad_norm": 2.6632483348563714, + "learning_rate": 3.9346784102537076e-06, + "loss": 0.6844, + "num_input_tokens_seen": 19030760, + "step": 908 + }, + { + "epoch": 0.10930078759093369, + "flos": 15918926853240.0, + "grad_norm": 2.316954678075324, + "learning_rate": 3.934480805552669e-06, + "loss": 0.7574, + "num_input_tokens_seen": 19051490, + "step": 909 + }, + { + "epoch": 0.10942103048157277, + "flos": 16540023627120.0, + "grad_norm": 2.399646764827537, + "learning_rate": 3.93428290739243e-06, + "loss": 0.8564, + "num_input_tokens_seen": 19070580, + "step": 910 + }, + { + "epoch": 0.10954127337221187, + "flos": 10947588205440.0, + "grad_norm": 2.920166821874758, + "learning_rate": 3.9340847158030125e-06, + "loss": 0.7836, + "num_input_tokens_seen": 19083880, + "step": 911 + }, + { + "epoch": 0.10966151626285096, + "flos": 15836054552040.0, + "grad_norm": 2.1253148014037633, + "learning_rate": 3.9338862308144814e-06, + "loss": 0.7452, + "num_input_tokens_seen": 19102420, + "step": 912 + }, + { + "epoch": 0.10978175915349005, + "flos": 14704431279120.0, + "grad_norm": 2.1398848764702034, + "learning_rate": 3.933687452456946e-06, + "loss": 0.8401, + "num_input_tokens_seen": 19122040, + "step": 913 + }, + { + "epoch": 0.10990200204412914, + "flos": 14919936355920.0, + "grad_norm": 4.235609036624227, + "learning_rate": 3.933488380760562e-06, + "loss": 0.8372, + "num_input_tokens_seen": 19141120, + "step": 914 + }, + { + "epoch": 0.11002224493476823, + "flos": 12786631489080.0, + "grad_norm": 2.7837767535235747, + "learning_rate": 3.9332890157555286e-06, + "loss": 0.8647, + "num_input_tokens_seen": 19157775, + "step": 915 + }, + { + "epoch": 0.11014248782540732, + "flos": 8844260829360.0, + "grad_norm": 3.712889854435086, + "learning_rate": 3.933089357472088e-06, + "loss": 0.7438, + "num_input_tokens_seen": 19175525, + "step": 916 + }, + { + "epoch": 0.11026273071604642, + "flos": 16378109879880.0, + "grad_norm": 2.261909121076547, + "learning_rate": 3.932889405940529e-06, + "loss": 0.8486, + "num_input_tokens_seen": 19193340, + "step": 917 + }, + { + "epoch": 0.1103829736066855, + "flos": 14488039723440.0, + "grad_norm": 2.81139368974887, + "learning_rate": 3.932689161191184e-06, + "loss": 0.7926, + "num_input_tokens_seen": 19210765, + "step": 918 + }, + { + "epoch": 0.1105032164973246, + "flos": 16589499161880.0, + "grad_norm": 2.395383698661293, + "learning_rate": 3.93248862325443e-06, + "loss": 0.862, + "num_input_tokens_seen": 19229390, + "step": 919 + }, + { + "epoch": 0.11062345938796368, + "flos": 49019105350680.0, + "grad_norm": 0.9762551396194347, + "learning_rate": 3.932287792160688e-06, + "loss": 0.648, + "num_input_tokens_seen": 19287570, + "step": 920 + }, + { + "epoch": 0.11074370227860278, + "flos": 16026299472120.0, + "grad_norm": 2.947833613830934, + "learning_rate": 3.932086667940424e-06, + "loss": 0.7891, + "num_input_tokens_seen": 19303995, + "step": 921 + }, + { + "epoch": 0.11086394516924186, + "flos": 21022075117680.0, + "grad_norm": 2.0800649436568857, + "learning_rate": 3.93188525062415e-06, + "loss": 0.8042, + "num_input_tokens_seen": 19324180, + "step": 922 + }, + { + "epoch": 0.11098418805988096, + "flos": 17970879096960.0, + "grad_norm": 10.130806470666279, + "learning_rate": 3.931683540242418e-06, + "loss": 0.8552, + "num_input_tokens_seen": 19344965, + "step": 923 + }, + { + "epoch": 0.11110443095052006, + "flos": 16804529339280.0, + "grad_norm": 5.507138388212948, + "learning_rate": 3.9314815368258295e-06, + "loss": 0.8779, + "num_input_tokens_seen": 19361165, + "step": 924 + }, + { + "epoch": 0.11122467384115914, + "flos": 13837028778720.0, + "grad_norm": 1.9564900211398535, + "learning_rate": 3.9312792404050275e-06, + "loss": 0.7572, + "num_input_tokens_seen": 19378940, + "step": 925 + }, + { + "epoch": 0.11134491673179824, + "flos": 18376280834280.0, + "grad_norm": 2.6812354418663906, + "learning_rate": 3.9310766510107e-06, + "loss": 0.7654, + "num_input_tokens_seen": 19397835, + "step": 926 + }, + { + "epoch": 0.11146515962243732, + "flos": 17944194242040.0, + "grad_norm": 2.0732794464975153, + "learning_rate": 3.9308737686735806e-06, + "loss": 0.9167, + "num_input_tokens_seen": 19417515, + "step": 927 + }, + { + "epoch": 0.11158540251307641, + "flos": 16348702268400.0, + "grad_norm": 3.584780437816731, + "learning_rate": 3.9306705934244455e-06, + "loss": 0.8081, + "num_input_tokens_seen": 19437315, + "step": 928 + }, + { + "epoch": 0.11170564540371551, + "flos": 14542295912160.0, + "grad_norm": 2.1295155064631124, + "learning_rate": 3.930467125294116e-06, + "loss": 0.8703, + "num_input_tokens_seen": 19456585, + "step": 929 + }, + { + "epoch": 0.1118258882943546, + "flos": 47614998055680.0, + "grad_norm": 0.9733544719917623, + "learning_rate": 3.930263364313458e-06, + "loss": 0.6283, + "num_input_tokens_seen": 19506875, + "step": 930 + }, + { + "epoch": 0.11194613118499369, + "flos": 12543808358160.0, + "grad_norm": 2.739458862107211, + "learning_rate": 3.930059310513384e-06, + "loss": 0.8224, + "num_input_tokens_seen": 19525635, + "step": 931 + }, + { + "epoch": 0.11206637407563277, + "flos": 23393485781400.0, + "grad_norm": 1.7795505782482899, + "learning_rate": 3.929854963924846e-06, + "loss": 0.8185, + "num_input_tokens_seen": 19545620, + "step": 932 + }, + { + "epoch": 0.11218661696627187, + "flos": 16054504005120.0, + "grad_norm": 2.4152875341308904, + "learning_rate": 3.929650324578845e-06, + "loss": 0.7553, + "num_input_tokens_seen": 19564805, + "step": 933 + }, + { + "epoch": 0.11230685985691095, + "flos": 18963759222000.0, + "grad_norm": 3.077508390215322, + "learning_rate": 3.929445392506423e-06, + "loss": 0.8135, + "num_input_tokens_seen": 19582465, + "step": 934 + }, + { + "epoch": 0.11242710274755005, + "flos": 16265228427960.0, + "grad_norm": 2.1396346202771572, + "learning_rate": 3.92924016773867e-06, + "loss": 0.7482, + "num_input_tokens_seen": 19598680, + "step": 935 + }, + { + "epoch": 0.11254734563818915, + "flos": 12921322162080.0, + "grad_norm": 3.0332251337337395, + "learning_rate": 3.9290346503067175e-06, + "loss": 0.7266, + "num_input_tokens_seen": 19615065, + "step": 936 + }, + { + "epoch": 0.11266758852882823, + "flos": 40465686828840.0, + "grad_norm": 2.5171426495582443, + "learning_rate": 3.9288288402417415e-06, + "loss": 0.7703, + "num_input_tokens_seen": 19641045, + "step": 937 + }, + { + "epoch": 0.11278783141946733, + "flos": 13783975668480.0, + "grad_norm": 3.4685560292489943, + "learning_rate": 3.928622737574964e-06, + "loss": 0.6796, + "num_input_tokens_seen": 19656100, + "step": 938 + }, + { + "epoch": 0.11290807431010641, + "flos": 19402209466200.0, + "grad_norm": 2.3366818470083137, + "learning_rate": 3.928416342337652e-06, + "loss": 0.8997, + "num_input_tokens_seen": 19675555, + "step": 939 + }, + { + "epoch": 0.1130283172007455, + "flos": 16620458111400.0, + "grad_norm": 2.2767738427459068, + "learning_rate": 3.928209654561113e-06, + "loss": 0.8091, + "num_input_tokens_seen": 19696110, + "step": 940 + }, + { + "epoch": 0.1131485600913846, + "flos": 16997180416320.0, + "grad_norm": 3.1617026488636695, + "learning_rate": 3.928002674276703e-06, + "loss": 0.7855, + "num_input_tokens_seen": 19715220, + "step": 941 + }, + { + "epoch": 0.11326880298202369, + "flos": 10221714929400.0, + "grad_norm": 2.654527352426928, + "learning_rate": 3.92779540151582e-06, + "loss": 0.7384, + "num_input_tokens_seen": 19732025, + "step": 942 + }, + { + "epoch": 0.11338904587266278, + "flos": 11894950290720.0, + "grad_norm": 2.2654664082761387, + "learning_rate": 3.927587836309907e-06, + "loss": 0.8465, + "num_input_tokens_seen": 19749575, + "step": 943 + }, + { + "epoch": 0.11350928876330187, + "flos": 17890286312880.0, + "grad_norm": 3.363546646758727, + "learning_rate": 3.927379978690452e-06, + "loss": 0.7617, + "num_input_tokens_seen": 19768560, + "step": 944 + }, + { + "epoch": 0.11362953165394096, + "flos": 17942389624320.0, + "grad_norm": 2.541166777516975, + "learning_rate": 3.927171828688987e-06, + "loss": 0.8489, + "num_input_tokens_seen": 19787805, + "step": 945 + }, + { + "epoch": 0.11374977454458005, + "flos": 17619765208320.0, + "grad_norm": 2.3006057024630637, + "learning_rate": 3.926963386337088e-06, + "loss": 0.7951, + "num_input_tokens_seen": 19805755, + "step": 946 + }, + { + "epoch": 0.11387001743521914, + "flos": 29014410675720.0, + "grad_norm": 4.4756214157919025, + "learning_rate": 3.926754651666375e-06, + "loss": 0.6841, + "num_input_tokens_seen": 19826035, + "step": 947 + }, + { + "epoch": 0.11399026032585824, + "flos": 18375837594840.0, + "grad_norm": 3.936469291327875, + "learning_rate": 3.926545624708513e-06, + "loss": 0.7626, + "num_input_tokens_seen": 19844995, + "step": 948 + }, + { + "epoch": 0.11411050321649732, + "flos": 13105266750120.0, + "grad_norm": 3.9316779796722727, + "learning_rate": 3.926336305495213e-06, + "loss": 0.8478, + "num_input_tokens_seen": 19863275, + "step": 949 + }, + { + "epoch": 0.11423074610713642, + "flos": 16431637889520.0, + "grad_norm": 2.6257180639607736, + "learning_rate": 3.926126694058226e-06, + "loss": 0.8664, + "num_input_tokens_seen": 19882145, + "step": 950 + }, + { + "epoch": 0.1143509889977755, + "flos": 14405357382000.0, + "grad_norm": 1.9846216484032393, + "learning_rate": 3.92591679042935e-06, + "loss": 0.8061, + "num_input_tokens_seen": 19901755, + "step": 951 + }, + { + "epoch": 0.1144712318884146, + "flos": 14482594210320.0, + "grad_norm": 1.6088696028822127, + "learning_rate": 3.92570659464043e-06, + "loss": 0.8114, + "num_input_tokens_seen": 19919535, + "step": 952 + }, + { + "epoch": 0.1145914747790537, + "flos": 10869116638680.0, + "grad_norm": 2.312931742187511, + "learning_rate": 3.925496106723349e-06, + "loss": 0.7861, + "num_input_tokens_seen": 19936695, + "step": 953 + }, + { + "epoch": 0.11471171766969278, + "flos": 14514597938520.0, + "grad_norm": 2.5095418986189637, + "learning_rate": 3.9252853267100405e-06, + "loss": 0.8238, + "num_input_tokens_seen": 19955660, + "step": 954 + }, + { + "epoch": 0.11483196056033187, + "flos": 16483962820680.0, + "grad_norm": 2.2694584478628217, + "learning_rate": 3.9250742546324786e-06, + "loss": 0.8235, + "num_input_tokens_seen": 19975615, + "step": 955 + }, + { + "epoch": 0.11495220345097096, + "flos": 20698057663440.0, + "grad_norm": 2.3267015035183625, + "learning_rate": 3.924862890522683e-06, + "loss": 0.8514, + "num_input_tokens_seen": 19995345, + "step": 956 + }, + { + "epoch": 0.11507244634161005, + "flos": 12758426956080.0, + "grad_norm": 3.611907202845579, + "learning_rate": 3.9246512344127174e-06, + "loss": 0.8488, + "num_input_tokens_seen": 20012725, + "step": 957 + }, + { + "epoch": 0.11519268923224914, + "flos": 16512705573000.0, + "grad_norm": 2.1679832664209946, + "learning_rate": 3.9244392863346895e-06, + "loss": 0.809, + "num_input_tokens_seen": 20031850, + "step": 958 + }, + { + "epoch": 0.11531293212288823, + "flos": 12381609671280.0, + "grad_norm": 2.3629227799863597, + "learning_rate": 3.9242270463207524e-06, + "loss": 0.8959, + "num_input_tokens_seen": 20049960, + "step": 959 + }, + { + "epoch": 0.11543317501352733, + "flos": 9006269556480.0, + "grad_norm": 3.5122539719184367, + "learning_rate": 3.924014514403102e-06, + "loss": 0.8285, + "num_input_tokens_seen": 20065835, + "step": 960 + }, + { + "epoch": 0.11555341790416641, + "flos": 14481517771680.0, + "grad_norm": 3.305476966871139, + "learning_rate": 3.92380169061398e-06, + "loss": 0.9079, + "num_input_tokens_seen": 20083335, + "step": 961 + }, + { + "epoch": 0.11567366079480551, + "flos": 18861578836560.0, + "grad_norm": 2.341922403491591, + "learning_rate": 3.9235885749856705e-06, + "loss": 0.8242, + "num_input_tokens_seen": 20101735, + "step": 962 + }, + { + "epoch": 0.1157939036854446, + "flos": 13299627465000.0, + "grad_norm": 2.437136598542904, + "learning_rate": 3.9233751675505035e-06, + "loss": 0.817, + "num_input_tokens_seen": 20120165, + "step": 963 + }, + { + "epoch": 0.11591414657608369, + "flos": 16888984638480.0, + "grad_norm": 4.272777778430287, + "learning_rate": 3.923161468340853e-06, + "loss": 0.8264, + "num_input_tokens_seen": 20139720, + "step": 964 + }, + { + "epoch": 0.11603438946672277, + "flos": 14215840641000.0, + "grad_norm": 2.101127607876032, + "learning_rate": 3.9229474773891374e-06, + "loss": 0.7996, + "num_input_tokens_seen": 20157980, + "step": 965 + }, + { + "epoch": 0.11615463235736187, + "flos": 19670356073760.0, + "grad_norm": 3.064488135405888, + "learning_rate": 3.922733194727818e-06, + "loss": 0.8334, + "num_input_tokens_seen": 20177495, + "step": 966 + }, + { + "epoch": 0.11627487524800097, + "flos": 13562740138920.0, + "grad_norm": 4.975407776342916, + "learning_rate": 3.922518620389402e-06, + "loss": 0.8293, + "num_input_tokens_seen": 20194080, + "step": 967 + }, + { + "epoch": 0.11639511813864005, + "flos": 13245751195800.0, + "grad_norm": 3.1510335286353244, + "learning_rate": 3.922303754406439e-06, + "loss": 0.8916, + "num_input_tokens_seen": 20211640, + "step": 968 + }, + { + "epoch": 0.11651536102927915, + "flos": 15296690320800.0, + "grad_norm": 3.271246201684632, + "learning_rate": 3.922088596811526e-06, + "loss": 0.7781, + "num_input_tokens_seen": 20230490, + "step": 969 + }, + { + "epoch": 0.11663560391991823, + "flos": 12030654082440.0, + "grad_norm": 4.344003573913228, + "learning_rate": 3.9218731476373e-06, + "loss": 0.8431, + "num_input_tokens_seen": 20246395, + "step": 970 + }, + { + "epoch": 0.11675584681055733, + "flos": 14514597938520.0, + "grad_norm": 3.2215029600663585, + "learning_rate": 3.9216574069164455e-06, + "loss": 0.8391, + "num_input_tokens_seen": 20265090, + "step": 971 + }, + { + "epoch": 0.11687608970119642, + "flos": 16053395906520.0, + "grad_norm": 2.692871138018472, + "learning_rate": 3.921441374681691e-06, + "loss": 0.79, + "num_input_tokens_seen": 20284870, + "step": 972 + }, + { + "epoch": 0.1169963325918355, + "flos": 17620050147960.0, + "grad_norm": 3.074793372926683, + "learning_rate": 3.921225050965808e-06, + "loss": 0.6362, + "num_input_tokens_seen": 20304475, + "step": 973 + }, + { + "epoch": 0.1171165754824746, + "flos": 17107212471840.0, + "grad_norm": 2.5344350109215585, + "learning_rate": 3.921008435801612e-06, + "loss": 0.7207, + "num_input_tokens_seen": 20323280, + "step": 974 + }, + { + "epoch": 0.11723681837311369, + "flos": 13543147115040.0, + "grad_norm": 2.787684508918584, + "learning_rate": 3.920791529221963e-06, + "loss": 0.7314, + "num_input_tokens_seen": 20341675, + "step": 975 + }, + { + "epoch": 0.11735706126375278, + "flos": 17242188084480.0, + "grad_norm": 2.247002512320526, + "learning_rate": 3.920574331259768e-06, + "loss": 0.7512, + "num_input_tokens_seen": 20362595, + "step": 976 + }, + { + "epoch": 0.11747730415439187, + "flos": 16376843481480.0, + "grad_norm": 2.9067124849902077, + "learning_rate": 3.9203568419479716e-06, + "loss": 0.7758, + "num_input_tokens_seen": 20382870, + "step": 977 + }, + { + "epoch": 0.11759754704503096, + "flos": 16243260907080.0, + "grad_norm": 2.291442779239439, + "learning_rate": 3.92013906131957e-06, + "loss": 0.7298, + "num_input_tokens_seen": 20401520, + "step": 978 + }, + { + "epoch": 0.11771778993567006, + "flos": 16323885351120.0, + "grad_norm": 2.2143145001205986, + "learning_rate": 3.9199209894076e-06, + "loss": 0.7984, + "num_input_tokens_seen": 20421555, + "step": 979 + }, + { + "epoch": 0.11783803282630914, + "flos": 15567844624560.0, + "grad_norm": 2.125948862395622, + "learning_rate": 3.919702626245142e-06, + "loss": 0.8831, + "num_input_tokens_seen": 20440930, + "step": 980 + }, + { + "epoch": 0.11795827571694824, + "flos": 18589379754120.0, + "grad_norm": 3.4396146338871856, + "learning_rate": 3.919483971865322e-06, + "loss": 0.6358, + "num_input_tokens_seen": 20460645, + "step": 981 + }, + { + "epoch": 0.11807851860758732, + "flos": 17295051234960.0, + "grad_norm": 2.674134665305941, + "learning_rate": 3.91926502630131e-06, + "loss": 0.8528, + "num_input_tokens_seen": 20480980, + "step": 982 + }, + { + "epoch": 0.11819876149822642, + "flos": 18294389991840.0, + "grad_norm": 4.127171493941738, + "learning_rate": 3.91904578958632e-06, + "loss": 0.7022, + "num_input_tokens_seen": 20500115, + "step": 983 + }, + { + "epoch": 0.11831900438886551, + "flos": 16837134606720.0, + "grad_norm": 3.149067477295858, + "learning_rate": 3.918826261753608e-06, + "loss": 0.8356, + "num_input_tokens_seen": 20519415, + "step": 984 + }, + { + "epoch": 0.1184392472795046, + "flos": 20509269101520.0, + "grad_norm": 3.921349235643372, + "learning_rate": 3.918606442836478e-06, + "loss": 0.6817, + "num_input_tokens_seen": 20541355, + "step": 985 + }, + { + "epoch": 0.1185594901701437, + "flos": 14538939956400.0, + "grad_norm": 2.2478580876148886, + "learning_rate": 3.918386332868277e-06, + "loss": 0.7627, + "num_input_tokens_seen": 20559045, + "step": 986 + }, + { + "epoch": 0.11867973306078278, + "flos": 13808982545520.0, + "grad_norm": 2.9322495307853687, + "learning_rate": 3.918165931882394e-06, + "loss": 0.9354, + "num_input_tokens_seen": 20577165, + "step": 987 + }, + { + "epoch": 0.11879997595142187, + "flos": 12381578011320.0, + "grad_norm": 3.775057679803031, + "learning_rate": 3.917945239912264e-06, + "loss": 0.7324, + "num_input_tokens_seen": 20594360, + "step": 988 + }, + { + "epoch": 0.11892021884206096, + "flos": 12786853108800.0, + "grad_norm": 4.18706248260954, + "learning_rate": 3.917724256991367e-06, + "loss": 0.7492, + "num_input_tokens_seen": 20612825, + "step": 989 + }, + { + "epoch": 0.11904046173270005, + "flos": 22722470233320.0, + "grad_norm": 2.6887038440014717, + "learning_rate": 3.9175029831532245e-06, + "loss": 0.7895, + "num_input_tokens_seen": 20632060, + "step": 990 + }, + { + "epoch": 0.11916070462333915, + "flos": 14730894514320.0, + "grad_norm": 2.7339931548394976, + "learning_rate": 3.917281418431404e-06, + "loss": 0.8616, + "num_input_tokens_seen": 20650825, + "step": 991 + }, + { + "epoch": 0.11928094751397823, + "flos": 17242694643840.0, + "grad_norm": 2.6269863424363606, + "learning_rate": 3.917059562859516e-06, + "loss": 0.7488, + "num_input_tokens_seen": 20669870, + "step": 992 + }, + { + "epoch": 0.11940119040461733, + "flos": 17506060597440.0, + "grad_norm": 2.251873190289103, + "learning_rate": 3.916837416471218e-06, + "loss": 0.8747, + "num_input_tokens_seen": 20686210, + "step": 993 + }, + { + "epoch": 0.11952143329525641, + "flos": 10058978023200.0, + "grad_norm": 3.6503023205277123, + "learning_rate": 3.916614979300207e-06, + "loss": 0.7007, + "num_input_tokens_seen": 20700775, + "step": 994 + }, + { + "epoch": 0.11964167618589551, + "flos": 19806344805120.0, + "grad_norm": 1.7296327246286487, + "learning_rate": 3.9163922513802274e-06, + "loss": 0.7663, + "num_input_tokens_seen": 20722830, + "step": 995 + }, + { + "epoch": 0.1197619190765346, + "flos": 9113990434920.0, + "grad_norm": 3.1342309038198626, + "learning_rate": 3.916169232745067e-06, + "loss": 0.8064, + "num_input_tokens_seen": 20740225, + "step": 996 + }, + { + "epoch": 0.11988216196717369, + "flos": 12327226842720.0, + "grad_norm": 3.4332186306308032, + "learning_rate": 3.915945923428559e-06, + "loss": 0.8833, + "num_input_tokens_seen": 20756470, + "step": 997 + }, + { + "epoch": 0.12000240485781279, + "flos": 11814515806440.0, + "grad_norm": 3.3949006716955226, + "learning_rate": 3.915722323464577e-06, + "loss": 0.8126, + "num_input_tokens_seen": 20774795, + "step": 998 + }, + { + "epoch": 0.12012264774845187, + "flos": 36331583230920.0, + "grad_norm": 2.9608482810395675, + "learning_rate": 3.91549843288704e-06, + "loss": 0.6914, + "num_input_tokens_seen": 20798195, + "step": 999 + }, + { + "epoch": 0.12024289063909097, + "flos": 19779691610160.0, + "grad_norm": 2.3531114919814295, + "learning_rate": 3.915274251729916e-06, + "loss": 0.7859, + "num_input_tokens_seen": 20819205, + "step": 1000 + }, + { + "epoch": 0.12036313352973005, + "flos": 14272154727120.0, + "grad_norm": 3.0865661753313676, + "learning_rate": 3.91504978002721e-06, + "loss": 0.888, + "num_input_tokens_seen": 20837980, + "step": 1001 + }, + { + "epoch": 0.12048337642036915, + "flos": 12591922514640.0, + "grad_norm": 2.74549809340378, + "learning_rate": 3.914825017812974e-06, + "loss": 0.753, + "num_input_tokens_seen": 20854350, + "step": 1002 + }, + { + "epoch": 0.12060361931100824, + "flos": 16728843849000.0, + "grad_norm": 2.855831932005934, + "learning_rate": 3.9145999651213065e-06, + "loss": 0.7163, + "num_input_tokens_seen": 20873310, + "step": 1003 + }, + { + "epoch": 0.12072386220164733, + "flos": 12192029610360.0, + "grad_norm": 4.253597083648663, + "learning_rate": 3.9143746219863465e-06, + "loss": 0.86, + "num_input_tokens_seen": 20890135, + "step": 1004 + }, + { + "epoch": 0.12084410509228642, + "flos": 40628333246520.0, + "grad_norm": 1.0565756391753636, + "learning_rate": 3.914148988442278e-06, + "loss": 0.7116, + "num_input_tokens_seen": 20945645, + "step": 1005 + }, + { + "epoch": 0.1209643479829255, + "flos": 19617429603360.0, + "grad_norm": 4.854818581394361, + "learning_rate": 3.91392306452333e-06, + "loss": 0.935, + "num_input_tokens_seen": 20962440, + "step": 1006 + }, + { + "epoch": 0.1210845908735646, + "flos": 7979739385320.0, + "grad_norm": 3.571213756640405, + "learning_rate": 3.913696850263774e-06, + "loss": 0.6306, + "num_input_tokens_seen": 20976525, + "step": 1007 + }, + { + "epoch": 0.1212048337642037, + "flos": 14973052786080.0, + "grad_norm": 2.2732709581127857, + "learning_rate": 3.913470345697929e-06, + "loss": 0.7731, + "num_input_tokens_seen": 20994875, + "step": 1008 + }, + { + "epoch": 0.12132507665484278, + "flos": 16350728505840.0, + "grad_norm": 2.516052967890669, + "learning_rate": 3.913243550860153e-06, + "loss": 0.8435, + "num_input_tokens_seen": 21012360, + "step": 1009 + }, + { + "epoch": 0.12144531954548188, + "flos": 21508196278920.0, + "grad_norm": 2.2813771878673488, + "learning_rate": 3.913016465784852e-06, + "loss": 0.752, + "num_input_tokens_seen": 21032755, + "step": 1010 + }, + { + "epoch": 0.12156556243612096, + "flos": 14973591005400.0, + "grad_norm": 5.0231658245087525, + "learning_rate": 3.912789090506474e-06, + "loss": 0.6978, + "num_input_tokens_seen": 21051735, + "step": 1011 + }, + { + "epoch": 0.12168580532676006, + "flos": 12003557648040.0, + "grad_norm": 3.2472601859429067, + "learning_rate": 3.9125614250595114e-06, + "loss": 0.71, + "num_input_tokens_seen": 21067665, + "step": 1012 + }, + { + "epoch": 0.12180604821739914, + "flos": 11166797497560.0, + "grad_norm": 6.53748983524619, + "learning_rate": 3.912333469478502e-06, + "loss": 0.8775, + "num_input_tokens_seen": 21085350, + "step": 1013 + }, + { + "epoch": 0.12192629110803824, + "flos": 14109829400400.0, + "grad_norm": 2.4256819798402627, + "learning_rate": 3.912105223798025e-06, + "loss": 0.7648, + "num_input_tokens_seen": 21104490, + "step": 1014 + }, + { + "epoch": 0.12204653399867733, + "flos": 35136178612680.0, + "grad_norm": 1.0209368374311236, + "learning_rate": 3.9118766880527065e-06, + "loss": 0.7115, + "num_input_tokens_seen": 21158645, + "step": 1015 + }, + { + "epoch": 0.12216677688931642, + "flos": 13298424386520.0, + "grad_norm": 1.8083466923420934, + "learning_rate": 3.9116478622772145e-06, + "loss": 0.7186, + "num_input_tokens_seen": 21176940, + "step": 1016 + }, + { + "epoch": 0.12228701977995551, + "flos": 20183763629160.0, + "grad_norm": 2.861993310052822, + "learning_rate": 3.911418746506261e-06, + "loss": 0.8653, + "num_input_tokens_seen": 21196790, + "step": 1017 + }, + { + "epoch": 0.1224072626705946, + "flos": 15945136808760.0, + "grad_norm": 2.6108697504628293, + "learning_rate": 3.911189340774604e-06, + "loss": 0.7719, + "num_input_tokens_seen": 21216640, + "step": 1018 + }, + { + "epoch": 0.1225275055612337, + "flos": 15135251472960.0, + "grad_norm": 2.1561251618003046, + "learning_rate": 3.910959645117043e-06, + "loss": 0.7814, + "num_input_tokens_seen": 21235695, + "step": 1019 + }, + { + "epoch": 0.12264774845187278, + "flos": 42554839525560.0, + "grad_norm": 0.784250524606429, + "learning_rate": 3.910729659568423e-06, + "loss": 0.5877, + "num_input_tokens_seen": 21292600, + "step": 1020 + }, + { + "epoch": 0.12276799134251187, + "flos": 19348206557160.0, + "grad_norm": 2.3491346429827766, + "learning_rate": 3.9104993841636344e-06, + "loss": 0.8081, + "num_input_tokens_seen": 21312890, + "step": 1021 + }, + { + "epoch": 0.12288823423315097, + "flos": 15401371843080.0, + "grad_norm": 1.994888031095243, + "learning_rate": 3.910268818937608e-06, + "loss": 0.7939, + "num_input_tokens_seen": 21330765, + "step": 1022 + }, + { + "epoch": 0.12300847712379005, + "flos": 8925170213040.0, + "grad_norm": 3.8500669101994442, + "learning_rate": 3.9100379639253196e-06, + "loss": 0.8676, + "num_input_tokens_seen": 21347205, + "step": 1023 + }, + { + "epoch": 0.12312872001442915, + "flos": 12218112926040.0, + "grad_norm": 3.2310126329964293, + "learning_rate": 3.909806819161791e-06, + "loss": 0.8425, + "num_input_tokens_seen": 21362400, + "step": 1024 + }, + { + "epoch": 0.12324896290506823, + "flos": 13433938218480.0, + "grad_norm": 2.5618069339522256, + "learning_rate": 3.909575384682086e-06, + "loss": 0.8421, + "num_input_tokens_seen": 21381000, + "step": 1025 + }, + { + "epoch": 0.12336920579570733, + "flos": 13811673642120.0, + "grad_norm": 2.0323285100645134, + "learning_rate": 3.9093436605213144e-06, + "loss": 0.6704, + "num_input_tokens_seen": 21401220, + "step": 1026 + }, + { + "epoch": 0.12348944868634643, + "flos": 17484726275760.0, + "grad_norm": 2.351633875577919, + "learning_rate": 3.909111646714627e-06, + "loss": 0.7719, + "num_input_tokens_seen": 21421785, + "step": 1027 + }, + { + "epoch": 0.12360969157698551, + "flos": 13893754444320.0, + "grad_norm": 2.0676657979002893, + "learning_rate": 3.9088793432972206e-06, + "loss": 0.6967, + "num_input_tokens_seen": 21440325, + "step": 1028 + }, + { + "epoch": 0.1237299344676246, + "flos": 9600238236000.0, + "grad_norm": 2.7106597494264464, + "learning_rate": 3.908646750304336e-06, + "loss": 0.797, + "num_input_tokens_seen": 21457730, + "step": 1029 + }, + { + "epoch": 0.12385017735826369, + "flos": 14973970924920.0, + "grad_norm": 1.6224834322561068, + "learning_rate": 3.908413867771257e-06, + "loss": 0.8479, + "num_input_tokens_seen": 21476360, + "step": 1030 + }, + { + "epoch": 0.12397042024890279, + "flos": 12651750856320.0, + "grad_norm": 2.2937152953471838, + "learning_rate": 3.908180695733311e-06, + "loss": 0.7902, + "num_input_tokens_seen": 21495570, + "step": 1031 + }, + { + "epoch": 0.12409066313954187, + "flos": 14919651416280.0, + "grad_norm": 3.2874421321286746, + "learning_rate": 3.907947234225871e-06, + "loss": 0.8144, + "num_input_tokens_seen": 21514300, + "step": 1032 + }, + { + "epoch": 0.12421090603018096, + "flos": 15159118591440.0, + "grad_norm": 2.317974723908505, + "learning_rate": 3.907713483284352e-06, + "loss": 0.8591, + "num_input_tokens_seen": 21533495, + "step": 1033 + }, + { + "epoch": 0.12433114892082006, + "flos": 17943940962360.0, + "grad_norm": 3.1834622408612385, + "learning_rate": 3.907479442944216e-06, + "loss": 0.9646, + "num_input_tokens_seen": 21551620, + "step": 1034 + }, + { + "epoch": 0.12445139181145914, + "flos": 14379337386240.0, + "grad_norm": 2.673098813691935, + "learning_rate": 3.907245113240963e-06, + "loss": 0.907, + "num_input_tokens_seen": 21569460, + "step": 1035 + }, + { + "epoch": 0.12457163470209824, + "flos": 34168839092640.0, + "grad_norm": 1.994447331469237, + "learning_rate": 3.907010494210144e-06, + "loss": 0.7315, + "num_input_tokens_seen": 21591840, + "step": 1036 + }, + { + "epoch": 0.12469187759273732, + "flos": 14893029881280.0, + "grad_norm": 3.5966722631779198, + "learning_rate": 3.9067755858873495e-06, + "loss": 0.9189, + "num_input_tokens_seen": 21608360, + "step": 1037 + }, + { + "epoch": 0.12481212048337642, + "flos": 51048393554400.0, + "grad_norm": 0.8402535791476926, + "learning_rate": 3.906540388308214e-06, + "loss": 0.6559, + "num_input_tokens_seen": 21667665, + "step": 1038 + }, + { + "epoch": 0.12493236337401552, + "flos": 13299975724560.0, + "grad_norm": 12.128664701701052, + "learning_rate": 3.906304901508417e-06, + "loss": 0.8046, + "num_input_tokens_seen": 21686285, + "step": 1039 + }, + { + "epoch": 0.12505260626465461, + "flos": 22047972089640.0, + "grad_norm": 4.102083532754103, + "learning_rate": 3.9060691255236835e-06, + "loss": 0.7394, + "num_input_tokens_seen": 21706570, + "step": 1040 + }, + { + "epoch": 0.1251728491552937, + "flos": 17897093204280.0, + "grad_norm": 1.9410773445768188, + "learning_rate": 3.905833060389778e-06, + "loss": 0.8092, + "num_input_tokens_seen": 21730410, + "step": 1041 + }, + { + "epoch": 0.12529309204593278, + "flos": 19882790134440.0, + "grad_norm": 4.138803678882822, + "learning_rate": 3.905596706142513e-06, + "loss": 0.7529, + "num_input_tokens_seen": 21751540, + "step": 1042 + }, + { + "epoch": 0.12541333493657186, + "flos": 22588064499960.0, + "grad_norm": 2.337639638228389, + "learning_rate": 3.9053600628177435e-06, + "loss": 0.8301, + "num_input_tokens_seen": 21770870, + "step": 1043 + }, + { + "epoch": 0.12553357782721097, + "flos": 17322812528520.0, + "grad_norm": 2.4863519060626253, + "learning_rate": 3.905123130451367e-06, + "loss": 0.8284, + "num_input_tokens_seen": 21791690, + "step": 1044 + }, + { + "epoch": 0.12565382071785006, + "flos": 18214018827480.0, + "grad_norm": 2.3927883756471915, + "learning_rate": 3.904885909079326e-06, + "loss": 0.7734, + "num_input_tokens_seen": 21810195, + "step": 1045 + }, + { + "epoch": 0.12577406360848914, + "flos": 15621530934000.0, + "grad_norm": 28.897255779771935, + "learning_rate": 3.904648398737607e-06, + "loss": 0.7567, + "num_input_tokens_seen": 21828480, + "step": 1046 + }, + { + "epoch": 0.12589430649912825, + "flos": 26557784873760.0, + "grad_norm": 2.538789892967905, + "learning_rate": 3.9044105994622406e-06, + "loss": 0.7659, + "num_input_tokens_seen": 21849345, + "step": 1047 + }, + { + "epoch": 0.12601454938976733, + "flos": 18564689476680.0, + "grad_norm": 2.92065306219215, + "learning_rate": 3.9041725112893005e-06, + "loss": 0.7928, + "num_input_tokens_seen": 21870290, + "step": 1048 + }, + { + "epoch": 0.12613479228040642, + "flos": 11328711244800.0, + "grad_norm": 3.2570349042015962, + "learning_rate": 3.903934134254904e-06, + "loss": 0.7367, + "num_input_tokens_seen": 21887800, + "step": 1049 + }, + { + "epoch": 0.1262550351710455, + "flos": 15702851897160.0, + "grad_norm": 2.8590000928229644, + "learning_rate": 3.903695468395213e-06, + "loss": 0.8289, + "num_input_tokens_seen": 21905390, + "step": 1050 + }, + { + "epoch": 0.1263752780616846, + "flos": 23182001519520.0, + "grad_norm": 5.944760029973022, + "learning_rate": 3.903456513746434e-06, + "loss": 0.5519, + "num_input_tokens_seen": 21926085, + "step": 1051 + }, + { + "epoch": 0.1264955209523237, + "flos": 21103649360520.0, + "grad_norm": 2.483957795434192, + "learning_rate": 3.903217270344815e-06, + "loss": 0.8614, + "num_input_tokens_seen": 21946055, + "step": 1052 + }, + { + "epoch": 0.12661576384296278, + "flos": 21453433530840.0, + "grad_norm": 2.433494210695051, + "learning_rate": 3.902977738226648e-06, + "loss": 0.8023, + "num_input_tokens_seen": 21966510, + "step": 1053 + }, + { + "epoch": 0.12673600673360189, + "flos": 15243827170320.0, + "grad_norm": 2.259975571964543, + "learning_rate": 3.902737917428273e-06, + "loss": 0.9026, + "num_input_tokens_seen": 21984395, + "step": 1054 + }, + { + "epoch": 0.12685624962424097, + "flos": 18509800088760.0, + "grad_norm": 2.0436231101305755, + "learning_rate": 3.902497807986068e-06, + "loss": 0.8155, + "num_input_tokens_seen": 22004135, + "step": 1055 + }, + { + "epoch": 0.12697649251488005, + "flos": 20185061687520.0, + "grad_norm": 2.5048035250304816, + "learning_rate": 3.902257409936458e-06, + "loss": 0.8185, + "num_input_tokens_seen": 22024620, + "step": 1056 + }, + { + "epoch": 0.12709673540551916, + "flos": 15541001469840.0, + "grad_norm": 2.4281904017846987, + "learning_rate": 3.902016723315912e-06, + "loss": 0.8172, + "num_input_tokens_seen": 22042280, + "step": 1057 + }, + { + "epoch": 0.12721697829615825, + "flos": 18564721136640.0, + "grad_norm": 5.088216572752712, + "learning_rate": 3.901775748160941e-06, + "loss": 0.6715, + "num_input_tokens_seen": 22061180, + "step": 1058 + }, + { + "epoch": 0.12733722118679733, + "flos": 45661938052920.0, + "grad_norm": 0.8100947210477285, + "learning_rate": 3.901534484508101e-06, + "loss": 0.6304, + "num_input_tokens_seen": 22123575, + "step": 1059 + }, + { + "epoch": 0.1274574640774364, + "flos": 19777317113160.0, + "grad_norm": 2.5952990398942144, + "learning_rate": 3.901292932393991e-06, + "loss": 0.7217, + "num_input_tokens_seen": 22142175, + "step": 1060 + }, + { + "epoch": 0.12757770696807552, + "flos": 16269470862600.0, + "grad_norm": 2.7519616153280135, + "learning_rate": 3.9010510918552555e-06, + "loss": 0.8412, + "num_input_tokens_seen": 22160970, + "step": 1061 + }, + { + "epoch": 0.1276979498587146, + "flos": 20940405894960.0, + "grad_norm": 3.2941712777796837, + "learning_rate": 3.900808962928581e-06, + "loss": 0.7332, + "num_input_tokens_seen": 22178305, + "step": 1062 + }, + { + "epoch": 0.1278181927493537, + "flos": 12705342185880.0, + "grad_norm": 3.1860090770604104, + "learning_rate": 3.900566545650698e-06, + "loss": 0.8832, + "num_input_tokens_seen": 22195695, + "step": 1063 + }, + { + "epoch": 0.1279384356399928, + "flos": 15456672810480.0, + "grad_norm": 2.6217309487091605, + "learning_rate": 3.900323840058381e-06, + "loss": 0.7998, + "num_input_tokens_seen": 22213125, + "step": 1064 + }, + { + "epoch": 0.12805867853063188, + "flos": 19481472531960.0, + "grad_norm": 3.004973262978934, + "learning_rate": 3.900080846188449e-06, + "loss": 0.7997, + "num_input_tokens_seen": 22231435, + "step": 1065 + }, + { + "epoch": 0.12817892142127096, + "flos": 11977157732760.0, + "grad_norm": 2.284100671831003, + "learning_rate": 3.8998375640777625e-06, + "loss": 0.7947, + "num_input_tokens_seen": 22249025, + "step": 1066 + }, + { + "epoch": 0.12829916431191005, + "flos": 44781396669120.0, + "grad_norm": 0.7899577661282836, + "learning_rate": 3.899593993763229e-06, + "loss": 0.5456, + "num_input_tokens_seen": 22309705, + "step": 1067 + }, + { + "epoch": 0.12841940720254916, + "flos": 21857125630320.0, + "grad_norm": 2.9179295303589194, + "learning_rate": 3.899350135281796e-06, + "loss": 0.7835, + "num_input_tokens_seen": 22330425, + "step": 1068 + }, + { + "epoch": 0.12853965009318824, + "flos": 19018363670280.0, + "grad_norm": 2.369106710746284, + "learning_rate": 3.8991059886704585e-06, + "loss": 0.7881, + "num_input_tokens_seen": 22349650, + "step": 1069 + }, + { + "epoch": 0.12865989298382732, + "flos": 22641845789280.0, + "grad_norm": 2.697108104617003, + "learning_rate": 3.898861553966252e-06, + "loss": 0.8109, + "num_input_tokens_seen": 22369020, + "step": 1070 + }, + { + "epoch": 0.12878013587446643, + "flos": 18970439473560.0, + "grad_norm": 2.9903809195700166, + "learning_rate": 3.898616831206257e-06, + "loss": 0.8643, + "num_input_tokens_seen": 22389165, + "step": 1071 + }, + { + "epoch": 0.12890037876510552, + "flos": 17081509075680.0, + "grad_norm": 2.206371323488922, + "learning_rate": 3.8983718204276e-06, + "loss": 0.7535, + "num_input_tokens_seen": 22411105, + "step": 1072 + }, + { + "epoch": 0.1290206216557446, + "flos": 17269379498760.0, + "grad_norm": 2.1182949646547087, + "learning_rate": 3.898126521667446e-06, + "loss": 0.8226, + "num_input_tokens_seen": 22430980, + "step": 1073 + }, + { + "epoch": 0.12914086454638368, + "flos": 17701054511520.0, + "grad_norm": 2.258405033488919, + "learning_rate": 3.897880934963007e-06, + "loss": 0.8286, + "num_input_tokens_seen": 22450250, + "step": 1074 + }, + { + "epoch": 0.1292611074370228, + "flos": 14812310457360.0, + "grad_norm": 2.519186212142277, + "learning_rate": 3.89763506035154e-06, + "loss": 0.7697, + "num_input_tokens_seen": 22467820, + "step": 1075 + }, + { + "epoch": 0.12938135032766188, + "flos": 20073984853320.0, + "grad_norm": 4.020480652727367, + "learning_rate": 3.897388897870343e-06, + "loss": 0.8019, + "num_input_tokens_seen": 22488180, + "step": 1076 + }, + { + "epoch": 0.12950159321830096, + "flos": 21479991745920.0, + "grad_norm": 1.935384703154257, + "learning_rate": 3.89714244755676e-06, + "loss": 0.7229, + "num_input_tokens_seen": 22509260, + "step": 1077 + }, + { + "epoch": 0.12962183610894007, + "flos": 17970562497360.0, + "grad_norm": 2.8770146152945855, + "learning_rate": 3.896895709448175e-06, + "loss": 0.854, + "num_input_tokens_seen": 22528730, + "step": 1078 + }, + { + "epoch": 0.12974207899957915, + "flos": 8113416939600.0, + "grad_norm": 3.858561318062921, + "learning_rate": 3.896648683582019e-06, + "loss": 0.7424, + "num_input_tokens_seen": 22543785, + "step": 1079 + }, + { + "epoch": 0.12986232189021824, + "flos": 18106076329320.0, + "grad_norm": 3.801249178303332, + "learning_rate": 3.896401369995766e-06, + "loss": 0.7962, + "num_input_tokens_seen": 22563310, + "step": 1080 + }, + { + "epoch": 0.12998256478085732, + "flos": 17512170969720.0, + "grad_norm": 2.742013779052571, + "learning_rate": 3.896153768726932e-06, + "loss": 0.7671, + "num_input_tokens_seen": 22583340, + "step": 1081 + }, + { + "epoch": 0.13010280767149643, + "flos": 13596263545200.0, + "grad_norm": 2.6376092607908896, + "learning_rate": 3.8959058798130806e-06, + "loss": 0.8526, + "num_input_tokens_seen": 22601035, + "step": 1082 + }, + { + "epoch": 0.1302230505621355, + "flos": 16674840939960.0, + "grad_norm": 2.056753302201838, + "learning_rate": 3.895657703291814e-06, + "loss": 0.7242, + "num_input_tokens_seen": 22620860, + "step": 1083 + }, + { + "epoch": 0.1303432934527746, + "flos": 15593927940240.0, + "grad_norm": 3.4120399192898723, + "learning_rate": 3.895409239200781e-06, + "loss": 0.7782, + "num_input_tokens_seen": 22636465, + "step": 1084 + }, + { + "epoch": 0.1304635363434137, + "flos": 15297101900280.0, + "grad_norm": 2.9302977879974543, + "learning_rate": 3.895160487577673e-06, + "loss": 0.8969, + "num_input_tokens_seen": 22653755, + "step": 1085 + }, + { + "epoch": 0.1305837792340528, + "flos": 46624207488000.0, + "grad_norm": 0.7747944959312069, + "learning_rate": 3.894911448460226e-06, + "loss": 0.6247, + "num_input_tokens_seen": 22712790, + "step": 1086 + }, + { + "epoch": 0.13070402212469187, + "flos": 19371757076040.0, + "grad_norm": 2.4268582175531948, + "learning_rate": 3.8946621218862195e-06, + "loss": 0.7172, + "num_input_tokens_seen": 22733510, + "step": 1087 + }, + { + "epoch": 0.13082426501533098, + "flos": 20293384105200.0, + "grad_norm": 2.65202417260225, + "learning_rate": 3.894412507893475e-06, + "loss": 0.8682, + "num_input_tokens_seen": 22753510, + "step": 1088 + }, + { + "epoch": 0.13094450790597006, + "flos": 18186574133520.0, + "grad_norm": 2.8115717733666328, + "learning_rate": 3.894162606519859e-06, + "loss": 0.6993, + "num_input_tokens_seen": 22772180, + "step": 1089 + }, + { + "epoch": 0.13106475079660915, + "flos": 13920787558800.0, + "grad_norm": 2.0887625205705143, + "learning_rate": 3.893912417803282e-06, + "loss": 0.7495, + "num_input_tokens_seen": 22791615, + "step": 1090 + }, + { + "epoch": 0.13118499368724823, + "flos": 21210800359680.0, + "grad_norm": 2.7875474934316737, + "learning_rate": 3.8936619417816975e-06, + "loss": 0.7459, + "num_input_tokens_seen": 22811665, + "step": 1091 + }, + { + "epoch": 0.13130523657788734, + "flos": 10383913616280.0, + "grad_norm": 2.2919150015630265, + "learning_rate": 3.8934111784931015e-06, + "loss": 0.7082, + "num_input_tokens_seen": 22828835, + "step": 1092 + }, + { + "epoch": 0.13142547946852642, + "flos": 51751222870920.0, + "grad_norm": 0.9649932374865512, + "learning_rate": 3.893160127975535e-06, + "loss": 0.6357, + "num_input_tokens_seen": 22889245, + "step": 1093 + }, + { + "epoch": 0.1315457223591655, + "flos": 33712600442280.0, + "grad_norm": 3.0457688108638443, + "learning_rate": 3.8929087902670826e-06, + "loss": 0.7966, + "num_input_tokens_seen": 22910595, + "step": 1094 + }, + { + "epoch": 0.13166596524980462, + "flos": 46354509542400.0, + "grad_norm": 0.8700992740795114, + "learning_rate": 3.8926571654058715e-06, + "loss": 0.6184, + "num_input_tokens_seen": 22966960, + "step": 1095 + }, + { + "epoch": 0.1317862081404437, + "flos": 17268746299560.0, + "grad_norm": 2.5883327120919186, + "learning_rate": 3.892405253430074e-06, + "loss": 0.744, + "num_input_tokens_seen": 22984200, + "step": 1096 + }, + { + "epoch": 0.13190645103108278, + "flos": 14806674984480.0, + "grad_norm": 2.1741259868652265, + "learning_rate": 3.892153054377904e-06, + "loss": 0.8119, + "num_input_tokens_seen": 23001325, + "step": 1097 + }, + { + "epoch": 0.13202669392172187, + "flos": 39377084950200.0, + "grad_norm": 0.9781939802130032, + "learning_rate": 3.891900568287619e-06, + "loss": 0.6302, + "num_input_tokens_seen": 23053430, + "step": 1098 + }, + { + "epoch": 0.13214693681236098, + "flos": 11544121341720.0, + "grad_norm": 5.549442352597663, + "learning_rate": 3.891647795197523e-06, + "loss": 0.7065, + "num_input_tokens_seen": 23069190, + "step": 1099 + }, + { + "epoch": 0.13226717970300006, + "flos": 14136229315680.0, + "grad_norm": 2.58315309285436, + "learning_rate": 3.8913947351459605e-06, + "loss": 0.6675, + "num_input_tokens_seen": 23086450, + "step": 1100 + }, + { + "epoch": 0.13238742259363914, + "flos": 15130249199280.0, + "grad_norm": 1.9941127674275418, + "learning_rate": 3.89114138817132e-06, + "loss": 0.6658, + "num_input_tokens_seen": 23102835, + "step": 1101 + }, + { + "epoch": 0.13250766548427825, + "flos": 15378549503280.0, + "grad_norm": 1.9999989271161063, + "learning_rate": 3.890887754312035e-06, + "loss": 0.8364, + "num_input_tokens_seen": 23120800, + "step": 1102 + }, + { + "epoch": 0.13262790837491734, + "flos": 16566676822080.0, + "grad_norm": 2.562414772663295, + "learning_rate": 3.890633833606581e-06, + "loss": 0.8573, + "num_input_tokens_seen": 23140210, + "step": 1103 + }, + { + "epoch": 0.13274815126555642, + "flos": 14380065565320.0, + "grad_norm": 3.001782364663889, + "learning_rate": 3.890379626093477e-06, + "loss": 0.6756, + "num_input_tokens_seen": 23159680, + "step": 1104 + }, + { + "epoch": 0.1328683941561955, + "flos": 15588704046840.0, + "grad_norm": 2.7429413805242335, + "learning_rate": 3.890125131811287e-06, + "loss": 0.8915, + "num_input_tokens_seen": 23177450, + "step": 1105 + }, + { + "epoch": 0.1329886370468346, + "flos": 9951447104520.0, + "grad_norm": 2.4216118454246796, + "learning_rate": 3.889870350798618e-06, + "loss": 0.7338, + "num_input_tokens_seen": 23194515, + "step": 1106 + }, + { + "epoch": 0.1331088799374737, + "flos": 15378517843320.0, + "grad_norm": 1.5920541754279465, + "learning_rate": 3.889615283094119e-06, + "loss": 0.7715, + "num_input_tokens_seen": 23213425, + "step": 1107 + }, + { + "epoch": 0.13322912282811278, + "flos": 13327198798800.0, + "grad_norm": 3.2886334862073037, + "learning_rate": 3.889359928736485e-06, + "loss": 0.8301, + "num_input_tokens_seen": 23231090, + "step": 1108 + }, + { + "epoch": 0.1333493657187519, + "flos": 17915641449480.0, + "grad_norm": 2.3768845158802003, + "learning_rate": 3.889104287764451e-06, + "loss": 0.8884, + "num_input_tokens_seen": 23251185, + "step": 1109 + }, + { + "epoch": 0.13346960860939097, + "flos": 16211573778480.0, + "grad_norm": 2.288870949916643, + "learning_rate": 3.888848360216798e-06, + "loss": 0.8892, + "num_input_tokens_seen": 23268550, + "step": 1110 + }, + { + "epoch": 0.13358985150003005, + "flos": 50094007449360.0, + "grad_norm": 0.7938020583975955, + "learning_rate": 3.888592146132351e-06, + "loss": 0.575, + "num_input_tokens_seen": 23329540, + "step": 1111 + }, + { + "epoch": 0.13371009439066917, + "flos": 19672382311200.0, + "grad_norm": 2.0000216959730195, + "learning_rate": 3.888335645549978e-06, + "loss": 0.7635, + "num_input_tokens_seen": 23349680, + "step": 1112 + }, + { + "epoch": 0.13383033728130825, + "flos": 19293792068640.0, + "grad_norm": 2.858972573233173, + "learning_rate": 3.888078858508588e-06, + "loss": 0.802, + "num_input_tokens_seen": 23369260, + "step": 1113 + }, + { + "epoch": 0.13395058017194733, + "flos": 16511819094120.0, + "grad_norm": 2.50910312799763, + "learning_rate": 3.8878217850471365e-06, + "loss": 0.8288, + "num_input_tokens_seen": 23388895, + "step": 1114 + }, + { + "epoch": 0.13407082306258641, + "flos": 18914790246600.0, + "grad_norm": 1.730571115191487, + "learning_rate": 3.887564425204621e-06, + "loss": 0.7171, + "num_input_tokens_seen": 23410300, + "step": 1115 + }, + { + "epoch": 0.13419106595322552, + "flos": 50390987297760.0, + "grad_norm": 0.8065610019650712, + "learning_rate": 3.887306779020083e-06, + "loss": 0.5628, + "num_input_tokens_seen": 23464675, + "step": 1116 + }, + { + "epoch": 0.1343113088438646, + "flos": 14946874490520.0, + "grad_norm": 2.018680828787251, + "learning_rate": 3.887048846532608e-06, + "loss": 0.6878, + "num_input_tokens_seen": 23481370, + "step": 1117 + }, + { + "epoch": 0.1344315517345037, + "flos": 49688981140200.0, + "grad_norm": 0.7468143598128075, + "learning_rate": 3.8867906277813224e-06, + "loss": 0.5903, + "num_input_tokens_seen": 23539245, + "step": 1118 + }, + { + "epoch": 0.1345517946251428, + "flos": 29959303284120.0, + "grad_norm": 2.442860601975646, + "learning_rate": 3.886532122805399e-06, + "loss": 0.7248, + "num_input_tokens_seen": 23561445, + "step": 1119 + }, + { + "epoch": 0.13467203751578188, + "flos": 16698644738520.0, + "grad_norm": 2.298952577344922, + "learning_rate": 3.886273331644053e-06, + "loss": 0.8828, + "num_input_tokens_seen": 23580035, + "step": 1120 + }, + { + "epoch": 0.13479228040642097, + "flos": 12461980835640.0, + "grad_norm": 2.113004373685803, + "learning_rate": 3.886014254336542e-06, + "loss": 0.8107, + "num_input_tokens_seen": 23596230, + "step": 1121 + }, + { + "epoch": 0.13491252329706005, + "flos": 17375042479800.0, + "grad_norm": 1.9216495784894145, + "learning_rate": 3.885754890922168e-06, + "loss": 0.9072, + "num_input_tokens_seen": 23616280, + "step": 1122 + }, + { + "epoch": 0.13503276618769916, + "flos": 25068905679960.0, + "grad_norm": 2.3277949156796676, + "learning_rate": 3.885495241440277e-06, + "loss": 0.7736, + "num_input_tokens_seen": 23640095, + "step": 1123 + }, + { + "epoch": 0.13515300907833824, + "flos": 12921448801920.0, + "grad_norm": 2.0806742928676965, + "learning_rate": 3.885235305930257e-06, + "loss": 0.73, + "num_input_tokens_seen": 23658015, + "step": 1124 + }, + { + "epoch": 0.13527325196897733, + "flos": 14806769964360.0, + "grad_norm": 2.127873105281176, + "learning_rate": 3.884975084431539e-06, + "loss": 0.8586, + "num_input_tokens_seen": 23672685, + "step": 1125 + }, + { + "epoch": 0.13539349485961644, + "flos": 13272341070840.0, + "grad_norm": 2.940288437688003, + "learning_rate": 3.8847145769836e-06, + "loss": 0.9105, + "num_input_tokens_seen": 23688825, + "step": 1126 + }, + { + "epoch": 0.13551373775025552, + "flos": 14109322841040.0, + "grad_norm": 2.9825986998375194, + "learning_rate": 3.884453783625959e-06, + "loss": 0.6561, + "num_input_tokens_seen": 23706155, + "step": 1127 + }, + { + "epoch": 0.1356339806408946, + "flos": 15244238749800.0, + "grad_norm": 2.469539178185678, + "learning_rate": 3.884192704398176e-06, + "loss": 0.8394, + "num_input_tokens_seen": 23723075, + "step": 1128 + }, + { + "epoch": 0.13575422353153369, + "flos": 37168723300920.0, + "grad_norm": 2.257454497257905, + "learning_rate": 3.883931339339858e-06, + "loss": 0.7267, + "num_input_tokens_seen": 23747180, + "step": 1129 + }, + { + "epoch": 0.1358744664221728, + "flos": 13246036135440.0, + "grad_norm": 2.435661502884693, + "learning_rate": 3.883669688490654e-06, + "loss": 0.7694, + "num_input_tokens_seen": 23764670, + "step": 1130 + }, + { + "epoch": 0.13599470931281188, + "flos": 13462301051280.0, + "grad_norm": 2.254403679234066, + "learning_rate": 3.883407751890256e-06, + "loss": 0.8346, + "num_input_tokens_seen": 23782995, + "step": 1131 + }, + { + "epoch": 0.13611495220345096, + "flos": 19557696241560.0, + "grad_norm": 1.8178239437672306, + "learning_rate": 3.8831455295783994e-06, + "loss": 0.8366, + "num_input_tokens_seen": 23801965, + "step": 1132 + }, + { + "epoch": 0.13623519509409007, + "flos": 15861979567920.0, + "grad_norm": 2.203193825803263, + "learning_rate": 3.882883021594864e-06, + "loss": 0.7306, + "num_input_tokens_seen": 23819825, + "step": 1133 + }, + { + "epoch": 0.13635543798472916, + "flos": 10788682154400.0, + "grad_norm": 2.409890925364831, + "learning_rate": 3.8826202279794705e-06, + "loss": 0.8478, + "num_input_tokens_seen": 23836605, + "step": 1134 + }, + { + "epoch": 0.13647568087536824, + "flos": 16753090887000.0, + "grad_norm": 2.417285533019327, + "learning_rate": 3.882357148772085e-06, + "loss": 0.6785, + "num_input_tokens_seen": 23853750, + "step": 1135 + }, + { + "epoch": 0.13659592376600732, + "flos": 14568537527640.0, + "grad_norm": 3.0008594553399406, + "learning_rate": 3.882093784012617e-06, + "loss": 0.8368, + "num_input_tokens_seen": 23872110, + "step": 1136 + }, + { + "epoch": 0.13671616665664643, + "flos": 15670974808800.0, + "grad_norm": 1.8839691534817526, + "learning_rate": 3.881830133741019e-06, + "loss": 0.8173, + "num_input_tokens_seen": 23890695, + "step": 1137 + }, + { + "epoch": 0.13683640954728551, + "flos": 16669680366480.0, + "grad_norm": 2.1555278508655213, + "learning_rate": 3.881566197997285e-06, + "loss": 0.7535, + "num_input_tokens_seen": 23906850, + "step": 1138 + }, + { + "epoch": 0.1369566524379246, + "flos": 15891767098920.0, + "grad_norm": 1.8961797404742027, + "learning_rate": 3.881301976821456e-06, + "loss": 0.7348, + "num_input_tokens_seen": 23926600, + "step": 1139 + }, + { + "epoch": 0.1370768953285637, + "flos": 13596485164920.0, + "grad_norm": 2.3093309080795406, + "learning_rate": 3.881037470253612e-06, + "loss": 0.887, + "num_input_tokens_seen": 23945835, + "step": 1140 + }, + { + "epoch": 0.1371971382192028, + "flos": 10869401578320.0, + "grad_norm": 2.5153905155056155, + "learning_rate": 3.88077267833388e-06, + "loss": 0.7826, + "num_input_tokens_seen": 23962070, + "step": 1141 + }, + { + "epoch": 0.13731738110984187, + "flos": 13891728206880.0, + "grad_norm": 6.403252919354252, + "learning_rate": 3.880507601102427e-06, + "loss": 0.8188, + "num_input_tokens_seen": 23979725, + "step": 1142 + }, + { + "epoch": 0.13743762400048098, + "flos": 13273195889760.0, + "grad_norm": 1.6834286212704705, + "learning_rate": 3.880242238599467e-06, + "loss": 0.8066, + "num_input_tokens_seen": 23995970, + "step": 1143 + }, + { + "epoch": 0.13755786689112007, + "flos": 15378137923800.0, + "grad_norm": 1.9129986979215412, + "learning_rate": 3.879976590865254e-06, + "loss": 0.821, + "num_input_tokens_seen": 24015145, + "step": 1144 + }, + { + "epoch": 0.13767810978175915, + "flos": 15621404294160.0, + "grad_norm": 2.468090307243532, + "learning_rate": 3.879710657940087e-06, + "loss": 0.8638, + "num_input_tokens_seen": 24033815, + "step": 1145 + }, + { + "epoch": 0.13779835267239823, + "flos": 22453373826960.0, + "grad_norm": 2.7453891072578847, + "learning_rate": 3.879444439864308e-06, + "loss": 0.6827, + "num_input_tokens_seen": 24053110, + "step": 1146 + }, + { + "epoch": 0.13791859556303734, + "flos": 16589720781600.0, + "grad_norm": 1.9579746918956853, + "learning_rate": 3.879177936678301e-06, + "loss": 0.8465, + "num_input_tokens_seen": 24071835, + "step": 1147 + }, + { + "epoch": 0.13803883845367643, + "flos": 25881197172720.0, + "grad_norm": 2.0627448196762184, + "learning_rate": 3.878911148422496e-06, + "loss": 0.7653, + "num_input_tokens_seen": 24093030, + "step": 1148 + }, + { + "epoch": 0.1381590813443155, + "flos": 23506018973760.0, + "grad_norm": 2.828745905544067, + "learning_rate": 3.878644075137364e-06, + "loss": 0.6886, + "num_input_tokens_seen": 24113400, + "step": 1149 + }, + { + "epoch": 0.13827932423495462, + "flos": 13002516485400.0, + "grad_norm": 2.500237262911184, + "learning_rate": 3.878376716863418e-06, + "loss": 0.7783, + "num_input_tokens_seen": 24129420, + "step": 1150 + }, + { + "epoch": 0.1383995671255937, + "flos": 14190485504400.0, + "grad_norm": 2.1119179550187237, + "learning_rate": 3.878109073641219e-06, + "loss": 0.6861, + "num_input_tokens_seen": 24148170, + "step": 1151 + }, + { + "epoch": 0.13851981001623279, + "flos": 20755004948760.0, + "grad_norm": 1.6617088328335947, + "learning_rate": 3.877841145511366e-06, + "loss": 0.798, + "num_input_tokens_seen": 24170630, + "step": 1152 + }, + { + "epoch": 0.13864005290687187, + "flos": 15511973777880.0, + "grad_norm": 2.64347180284845, + "learning_rate": 3.8775729325145035e-06, + "loss": 0.8186, + "num_input_tokens_seen": 24189585, + "step": 1153 + }, + { + "epoch": 0.13876029579751098, + "flos": 52848757349640.0, + "grad_norm": 0.7758657695601624, + "learning_rate": 3.877304434691321e-06, + "loss": 0.6388, + "num_input_tokens_seen": 24256155, + "step": 1154 + }, + { + "epoch": 0.13888053868815006, + "flos": 16050736469880.0, + "grad_norm": 2.37887658412332, + "learning_rate": 3.877035652082548e-06, + "loss": 0.7841, + "num_input_tokens_seen": 24275320, + "step": 1155 + }, + { + "epoch": 0.13900078157878915, + "flos": 14324701278000.0, + "grad_norm": 2.040791330379074, + "learning_rate": 3.87676658472896e-06, + "loss": 0.8335, + "num_input_tokens_seen": 24293850, + "step": 1156 + }, + { + "epoch": 0.13912102446942826, + "flos": 16566835121880.0, + "grad_norm": 3.123409171498364, + "learning_rate": 3.876497232671372e-06, + "loss": 0.8341, + "num_input_tokens_seen": 24313525, + "step": 1157 + }, + { + "epoch": 0.13924126736006734, + "flos": 21751209369600.0, + "grad_norm": 2.5669065594550076, + "learning_rate": 3.876227595950647e-06, + "loss": 0.8113, + "num_input_tokens_seen": 24332675, + "step": 1158 + }, + { + "epoch": 0.13936151025070642, + "flos": 20104057323960.0, + "grad_norm": 1.8096906001321849, + "learning_rate": 3.875957674607686e-06, + "loss": 0.7703, + "num_input_tokens_seen": 24354670, + "step": 1159 + }, + { + "epoch": 0.1394817531413455, + "flos": 11949808018680.0, + "grad_norm": 2.21878503046455, + "learning_rate": 3.8756874686834386e-06, + "loss": 0.868, + "num_input_tokens_seen": 24372605, + "step": 1160 + }, + { + "epoch": 0.13960199603198462, + "flos": 22698033235560.0, + "grad_norm": 1.953205747842544, + "learning_rate": 3.875416978218893e-06, + "loss": 0.7921, + "num_input_tokens_seen": 24395520, + "step": 1161 + }, + { + "epoch": 0.1397222389226237, + "flos": 13218559781520.0, + "grad_norm": 2.836664448194462, + "learning_rate": 3.8751462032550835e-06, + "loss": 0.8111, + "num_input_tokens_seen": 24412245, + "step": 1162 + }, + { + "epoch": 0.13984248181326278, + "flos": 12300067088400.0, + "grad_norm": 2.683919251092834, + "learning_rate": 3.874875143833085e-06, + "loss": 0.8057, + "num_input_tokens_seen": 24430205, + "step": 1163 + }, + { + "epoch": 0.1399627247039019, + "flos": 39866082676440.0, + "grad_norm": 2.8375602682157846, + "learning_rate": 3.874603799994019e-06, + "loss": 0.6687, + "num_input_tokens_seen": 24453460, + "step": 1164 + }, + { + "epoch": 0.14008296759454097, + "flos": 8520464994840.0, + "grad_norm": 2.1885552177269916, + "learning_rate": 3.874332171779046e-06, + "loss": 0.8676, + "num_input_tokens_seen": 24468060, + "step": 1165 + }, + { + "epoch": 0.14020321048518006, + "flos": 16107557115360.0, + "grad_norm": 2.3458480599609097, + "learning_rate": 3.874060259229373e-06, + "loss": 0.7449, + "num_input_tokens_seen": 24489355, + "step": 1166 + }, + { + "epoch": 0.14032345337581917, + "flos": 17133928986720.0, + "grad_norm": 3.9708365414670865, + "learning_rate": 3.873788062386249e-06, + "loss": 0.9007, + "num_input_tokens_seen": 24507335, + "step": 1167 + }, + { + "epoch": 0.14044369626645825, + "flos": 21753362246880.0, + "grad_norm": 1.8919124790861042, + "learning_rate": 3.873515581290965e-06, + "loss": 0.8174, + "num_input_tokens_seen": 24531860, + "step": 1168 + }, + { + "epoch": 0.14056393915709733, + "flos": 13380821788320.0, + "grad_norm": 2.6862204410991124, + "learning_rate": 3.8732428159848575e-06, + "loss": 0.7416, + "num_input_tokens_seen": 24550555, + "step": 1169 + }, + { + "epoch": 0.14068418204773642, + "flos": 19563648314040.0, + "grad_norm": 2.261513603538944, + "learning_rate": 3.872969766509304e-06, + "loss": 0.7704, + "num_input_tokens_seen": 24570830, + "step": 1170 + }, + { + "epoch": 0.14080442493837553, + "flos": 48115429518840.0, + "grad_norm": 0.8236466722993513, + "learning_rate": 3.872696432905726e-06, + "loss": 0.5794, + "num_input_tokens_seen": 24631370, + "step": 1171 + }, + { + "epoch": 0.1409246678290146, + "flos": 18889783369560.0, + "grad_norm": 2.57137177041307, + "learning_rate": 3.872422815215589e-06, + "loss": 0.6853, + "num_input_tokens_seen": 24650170, + "step": 1172 + }, + { + "epoch": 0.1410449107196537, + "flos": 15996860200680.0, + "grad_norm": 2.4540186912427937, + "learning_rate": 3.8721489134803994e-06, + "loss": 0.7342, + "num_input_tokens_seen": 24668680, + "step": 1173 + }, + { + "epoch": 0.1411651536102928, + "flos": 12159582642720.0, + "grad_norm": 3.315150514065418, + "learning_rate": 3.871874727741707e-06, + "loss": 0.7075, + "num_input_tokens_seen": 24685630, + "step": 1174 + }, + { + "epoch": 0.1412853965009319, + "flos": 15348287072880.0, + "grad_norm": 2.2298415282524866, + "learning_rate": 3.871600258041108e-06, + "loss": 0.9551, + "num_input_tokens_seen": 24704875, + "step": 1175 + }, + { + "epoch": 0.14140563939157097, + "flos": 14862039271800.0, + "grad_norm": 2.787490273360316, + "learning_rate": 3.871325504420238e-06, + "loss": 0.8358, + "num_input_tokens_seen": 24723585, + "step": 1176 + }, + { + "epoch": 0.14152588228221005, + "flos": 15405455977920.0, + "grad_norm": 2.201590686885686, + "learning_rate": 3.871050466920776e-06, + "loss": 0.8056, + "num_input_tokens_seen": 24743210, + "step": 1177 + }, + { + "epoch": 0.14164612517284916, + "flos": 13299500825160.0, + "grad_norm": 2.230631433222743, + "learning_rate": 3.870775145584447e-06, + "loss": 0.7869, + "num_input_tokens_seen": 24760710, + "step": 1178 + }, + { + "epoch": 0.14176636806348825, + "flos": 16645148388840.0, + "grad_norm": 3.115828980528437, + "learning_rate": 3.8704995404530145e-06, + "loss": 0.6192, + "num_input_tokens_seen": 24776055, + "step": 1179 + }, + { + "epoch": 0.14188661095412733, + "flos": 16162098243720.0, + "grad_norm": 2.5061082604870144, + "learning_rate": 3.87022365156829e-06, + "loss": 0.8289, + "num_input_tokens_seen": 24796490, + "step": 1180 + }, + { + "epoch": 0.14200685384476644, + "flos": 17835491904840.0, + "grad_norm": 2.619553592448942, + "learning_rate": 3.869947478972123e-06, + "loss": 0.7955, + "num_input_tokens_seen": 24817235, + "step": 1181 + }, + { + "epoch": 0.14212709673540552, + "flos": 17590737516360.0, + "grad_norm": 2.2810766075812845, + "learning_rate": 3.869671022706412e-06, + "loss": 0.7968, + "num_input_tokens_seen": 24835685, + "step": 1182 + }, + { + "epoch": 0.1422473396260446, + "flos": 19374353192760.0, + "grad_norm": 2.0686912090326235, + "learning_rate": 3.869394282813092e-06, + "loss": 0.6433, + "num_input_tokens_seen": 24854605, + "step": 1183 + }, + { + "epoch": 0.1423675825166837, + "flos": 12435169340880.0, + "grad_norm": 3.140235535115705, + "learning_rate": 3.869117259334147e-06, + "loss": 0.8612, + "num_input_tokens_seen": 24872250, + "step": 1184 + }, + { + "epoch": 0.1424878254073228, + "flos": 13081906191000.0, + "grad_norm": 1.8723520654931252, + "learning_rate": 3.868839952311599e-06, + "loss": 0.8095, + "num_input_tokens_seen": 24889925, + "step": 1185 + }, + { + "epoch": 0.14260806829796188, + "flos": 14839280251920.0, + "grad_norm": 3.6386207662976666, + "learning_rate": 3.868562361787516e-06, + "loss": 0.7857, + "num_input_tokens_seen": 24908775, + "step": 1186 + }, + { + "epoch": 0.14272831118860096, + "flos": 16972205199240.0, + "grad_norm": 2.2503394824459004, + "learning_rate": 3.868284487804009e-06, + "loss": 0.6855, + "num_input_tokens_seen": 24927725, + "step": 1187 + }, + { + "epoch": 0.14284855407924008, + "flos": 19966390614720.0, + "grad_norm": 1.8846147071621575, + "learning_rate": 3.86800633040323e-06, + "loss": 0.778, + "num_input_tokens_seen": 24948035, + "step": 1188 + }, + { + "epoch": 0.14296879696987916, + "flos": 20671404468480.0, + "grad_norm": 2.2806967495530026, + "learning_rate": 3.867727889627376e-06, + "loss": 0.7698, + "num_input_tokens_seen": 24967370, + "step": 1189 + }, + { + "epoch": 0.14308903986051824, + "flos": 14163199110240.0, + "grad_norm": 2.2582477295288848, + "learning_rate": 3.867449165518687e-06, + "loss": 0.7667, + "num_input_tokens_seen": 24983560, + "step": 1190 + }, + { + "epoch": 0.14320928275115732, + "flos": 12732945179640.0, + "grad_norm": 2.1795003393618835, + "learning_rate": 3.867170158119444e-06, + "loss": 0.6799, + "num_input_tokens_seen": 25002280, + "step": 1191 + }, + { + "epoch": 0.14332952564179643, + "flos": 15698862742200.0, + "grad_norm": 2.4778868205999895, + "learning_rate": 3.866890867471972e-06, + "loss": 0.7426, + "num_input_tokens_seen": 25020470, + "step": 1192 + }, + { + "epoch": 0.14344976853243552, + "flos": 11947750121280.0, + "grad_norm": 3.510473384832894, + "learning_rate": 3.86661129361864e-06, + "loss": 0.8803, + "num_input_tokens_seen": 25034680, + "step": 1193 + }, + { + "epoch": 0.1435700114230746, + "flos": 13812845060640.0, + "grad_norm": 2.7072868281564975, + "learning_rate": 3.866331436601859e-06, + "loss": 0.8366, + "num_input_tokens_seen": 25052395, + "step": 1194 + }, + { + "epoch": 0.1436902543137137, + "flos": 14433435275160.0, + "grad_norm": 2.4954624004401467, + "learning_rate": 3.866051296464083e-06, + "loss": 0.732, + "num_input_tokens_seen": 25070950, + "step": 1195 + }, + { + "epoch": 0.1438104972043528, + "flos": 10680992935920.0, + "grad_norm": 3.0916013095524724, + "learning_rate": 3.86577087324781e-06, + "loss": 0.8324, + "num_input_tokens_seen": 25087160, + "step": 1196 + }, + { + "epoch": 0.14393074009499188, + "flos": 12462645694800.0, + "grad_norm": 2.299974802127968, + "learning_rate": 3.865490166995578e-06, + "loss": 0.7532, + "num_input_tokens_seen": 25105110, + "step": 1197 + }, + { + "epoch": 0.144050982985631, + "flos": 22366702330560.0, + "grad_norm": 1.985337752234616, + "learning_rate": 3.86520917774997e-06, + "loss": 0.8168, + "num_input_tokens_seen": 25124265, + "step": 1198 + }, + { + "epoch": 0.14417122587627007, + "flos": 13029644579760.0, + "grad_norm": 2.496535666542753, + "learning_rate": 3.864927905553614e-06, + "loss": 0.7367, + "num_input_tokens_seen": 25141895, + "step": 1199 + }, + { + "epoch": 0.14429146876690915, + "flos": 15808546538160.0, + "grad_norm": 2.4419196725766583, + "learning_rate": 3.8646463504491765e-06, + "loss": 0.869, + "num_input_tokens_seen": 25161750, + "step": 1200 + }, + { + "epoch": 0.14441171165754824, + "flos": 17025701548920.0, + "grad_norm": 1.9076767256877762, + "learning_rate": 3.8643645124793705e-06, + "loss": 0.8204, + "num_input_tokens_seen": 25180370, + "step": 1201 + }, + { + "epoch": 0.14453195454818735, + "flos": 31528585302240.0, + "grad_norm": 1.989339550588129, + "learning_rate": 3.8640823916869515e-06, + "loss": 0.7297, + "num_input_tokens_seen": 25204400, + "step": 1202 + }, + { + "epoch": 0.14465219743882643, + "flos": 19969619930640.0, + "grad_norm": 3.3004755544463658, + "learning_rate": 3.863799988114714e-06, + "loss": 0.7697, + "num_input_tokens_seen": 25226150, + "step": 1203 + }, + { + "epoch": 0.1447724403294655, + "flos": 12165376415400.0, + "grad_norm": 3.464571110988092, + "learning_rate": 3.863517301805502e-06, + "loss": 0.6845, + "num_input_tokens_seen": 25244260, + "step": 1204 + }, + { + "epoch": 0.14489268322010462, + "flos": 14667615237000.0, + "grad_norm": 2.532894683389409, + "learning_rate": 3.863234332802196e-06, + "loss": 0.9481, + "num_input_tokens_seen": 25256185, + "step": 1205 + }, + { + "epoch": 0.1450129261107437, + "flos": 19887919047960.0, + "grad_norm": 4.342289082044849, + "learning_rate": 3.862951081147723e-06, + "loss": 0.7058, + "num_input_tokens_seen": 25276070, + "step": 1206 + }, + { + "epoch": 0.1451331690013828, + "flos": 18834545722080.0, + "grad_norm": 2.7671072053810595, + "learning_rate": 3.862667546885053e-06, + "loss": 0.7626, + "num_input_tokens_seen": 25294340, + "step": 1207 + }, + { + "epoch": 0.14525341189202187, + "flos": 18860280778200.0, + "grad_norm": 2.0013316013109277, + "learning_rate": 3.8623837300571965e-06, + "loss": 0.7104, + "num_input_tokens_seen": 25313045, + "step": 1208 + }, + { + "epoch": 0.14537365478266098, + "flos": 16889111278320.0, + "grad_norm": 2.2448505178527283, + "learning_rate": 3.8620996307072085e-06, + "loss": 0.8275, + "num_input_tokens_seen": 25333470, + "step": 1209 + }, + { + "epoch": 0.14549389767330007, + "flos": 15054880308600.0, + "grad_norm": 2.0335601581110145, + "learning_rate": 3.861815248878188e-06, + "loss": 0.6304, + "num_input_tokens_seen": 25350675, + "step": 1210 + }, + { + "epoch": 0.14561414056393915, + "flos": 11004282211080.0, + "grad_norm": 3.6823488884795825, + "learning_rate": 3.861530584613274e-06, + "loss": 0.7837, + "num_input_tokens_seen": 25368395, + "step": 1211 + }, + { + "epoch": 0.14573438345457826, + "flos": 14190833763960.0, + "grad_norm": 2.7222900144305817, + "learning_rate": 3.86124563795565e-06, + "loss": 0.8019, + "num_input_tokens_seen": 25386930, + "step": 1212 + }, + { + "epoch": 0.14585462634521734, + "flos": 18188030491680.0, + "grad_norm": 2.023295865201649, + "learning_rate": 3.860960408948543e-06, + "loss": 0.6951, + "num_input_tokens_seen": 25408400, + "step": 1213 + }, + { + "epoch": 0.14597486923585642, + "flos": 11245933923480.0, + "grad_norm": 2.637205991584108, + "learning_rate": 3.860674897635222e-06, + "loss": 0.8815, + "num_input_tokens_seen": 25424605, + "step": 1214 + }, + { + "epoch": 0.1460951121264955, + "flos": 12139198119840.0, + "grad_norm": 2.460607791188365, + "learning_rate": 3.860389104058998e-06, + "loss": 0.8144, + "num_input_tokens_seen": 25442555, + "step": 1215 + }, + { + "epoch": 0.14621535501713462, + "flos": 18213607248000.0, + "grad_norm": 5.463255121688736, + "learning_rate": 3.860103028263227e-06, + "loss": 0.7043, + "num_input_tokens_seen": 25465380, + "step": 1216 + }, + { + "epoch": 0.1463355979077737, + "flos": 18483495153360.0, + "grad_norm": 2.147507344811772, + "learning_rate": 3.859816670291304e-06, + "loss": 0.675, + "num_input_tokens_seen": 25484195, + "step": 1217 + }, + { + "epoch": 0.14645584079841278, + "flos": 16134685209720.0, + "grad_norm": 2.36808110357271, + "learning_rate": 3.859530030186672e-06, + "loss": 0.884, + "num_input_tokens_seen": 25500925, + "step": 1218 + }, + { + "epoch": 0.1465760836890519, + "flos": 17296507593120.0, + "grad_norm": 2.6547398481770346, + "learning_rate": 3.859243107992813e-06, + "loss": 0.8047, + "num_input_tokens_seen": 25519450, + "step": 1219 + }, + { + "epoch": 0.14669632657969098, + "flos": 27496535449920.0, + "grad_norm": 2.881908905749805, + "learning_rate": 3.858955903753252e-06, + "loss": 0.7697, + "num_input_tokens_seen": 25537810, + "step": 1220 + }, + { + "epoch": 0.14681656947033006, + "flos": 20805778541880.0, + "grad_norm": 2.239027820858992, + "learning_rate": 3.858668417511559e-06, + "loss": 0.8198, + "num_input_tokens_seen": 25560280, + "step": 1221 + }, + { + "epoch": 0.14693681236096917, + "flos": 13489334165760.0, + "grad_norm": 2.957518523570181, + "learning_rate": 3.8583806493113445e-06, + "loss": 0.7451, + "num_input_tokens_seen": 25578345, + "step": 1222 + }, + { + "epoch": 0.14705705525160825, + "flos": 15190204180800.0, + "grad_norm": 2.0607140640481565, + "learning_rate": 3.858092599196263e-06, + "loss": 0.8164, + "num_input_tokens_seen": 25596020, + "step": 1223 + }, + { + "epoch": 0.14717729814224734, + "flos": 21965226428280.0, + "grad_norm": 2.4389174814751957, + "learning_rate": 3.857804267210012e-06, + "loss": 0.8071, + "num_input_tokens_seen": 25615040, + "step": 1224 + }, + { + "epoch": 0.14729754103288642, + "flos": 14649636871080.0, + "grad_norm": 2.119516140764083, + "learning_rate": 3.857515653396331e-06, + "loss": 0.8639, + "num_input_tokens_seen": 25631970, + "step": 1225 + }, + { + "epoch": 0.14741778392352553, + "flos": 14082796285920.0, + "grad_norm": 2.957746814737998, + "learning_rate": 3.857226757799002e-06, + "loss": 0.8552, + "num_input_tokens_seen": 25649245, + "step": 1226 + }, + { + "epoch": 0.1475380268141646, + "flos": 18619103965200.0, + "grad_norm": 2.892591312849748, + "learning_rate": 3.85693758046185e-06, + "loss": 0.7105, + "num_input_tokens_seen": 25667255, + "step": 1227 + }, + { + "epoch": 0.1476582697048037, + "flos": 15241452673320.0, + "grad_norm": 2.876095438704039, + "learning_rate": 3.8566481214287435e-06, + "loss": 0.816, + "num_input_tokens_seen": 25685095, + "step": 1228 + }, + { + "epoch": 0.1477785125954428, + "flos": 10194998414520.0, + "grad_norm": 3.0653335541632023, + "learning_rate": 3.8563583807435935e-06, + "loss": 0.8911, + "num_input_tokens_seen": 25700960, + "step": 1229 + }, + { + "epoch": 0.1478987554860819, + "flos": 14996160065520.0, + "grad_norm": 3.7274301986248, + "learning_rate": 3.856068358450353e-06, + "loss": 0.7555, + "num_input_tokens_seen": 25720630, + "step": 1230 + }, + { + "epoch": 0.14801899837672097, + "flos": 13028346521400.0, + "grad_norm": 1.9248766525185161, + "learning_rate": 3.8557780545930186e-06, + "loss": 0.8437, + "num_input_tokens_seen": 25738765, + "step": 1231 + }, + { + "epoch": 0.14813924126736006, + "flos": 15266047970880.0, + "grad_norm": 2.0051032761304888, + "learning_rate": 3.855487469215628e-06, + "loss": 0.7698, + "num_input_tokens_seen": 25757415, + "step": 1232 + }, + { + "epoch": 0.14825948415799917, + "flos": 27501980963040.0, + "grad_norm": 3.0171988535923746, + "learning_rate": 3.855196602362264e-06, + "loss": 0.7035, + "num_input_tokens_seen": 25780055, + "step": 1233 + }, + { + "epoch": 0.14837972704863825, + "flos": 16105024318560.0, + "grad_norm": 2.470485801336347, + "learning_rate": 3.854905454077051e-06, + "loss": 0.933, + "num_input_tokens_seen": 25797385, + "step": 1234 + }, + { + "epoch": 0.14849996993927733, + "flos": 15351864648360.0, + "grad_norm": 3.546633136585991, + "learning_rate": 3.854614024404155e-06, + "loss": 0.8557, + "num_input_tokens_seen": 25815415, + "step": 1235 + }, + { + "epoch": 0.14862021282991644, + "flos": 14650080110520.0, + "grad_norm": 2.1821887250730487, + "learning_rate": 3.8543223133877865e-06, + "loss": 0.8777, + "num_input_tokens_seen": 25833730, + "step": 1236 + }, + { + "epoch": 0.14874045572055553, + "flos": 16621724509800.0, + "grad_norm": 2.4456641581144964, + "learning_rate": 3.854030321072198e-06, + "loss": 0.8586, + "num_input_tokens_seen": 25853355, + "step": 1237 + }, + { + "epoch": 0.1488606986111946, + "flos": 18619262265000.0, + "grad_norm": 2.420550470246388, + "learning_rate": 3.853738047501682e-06, + "loss": 0.7202, + "num_input_tokens_seen": 25873635, + "step": 1238 + }, + { + "epoch": 0.1489809415018337, + "flos": 12408104566440.0, + "grad_norm": 3.1940515543790537, + "learning_rate": 3.85344549272058e-06, + "loss": 0.7563, + "num_input_tokens_seen": 25891335, + "step": 1239 + }, + { + "epoch": 0.1491011843924728, + "flos": 24723743863800.0, + "grad_norm": 1.7639280958982964, + "learning_rate": 3.853152656773269e-06, + "loss": 0.8166, + "num_input_tokens_seen": 25912490, + "step": 1240 + }, + { + "epoch": 0.14922142728311188, + "flos": 15487093540680.0, + "grad_norm": 2.100238482239932, + "learning_rate": 3.852859539704174e-06, + "loss": 0.818, + "num_input_tokens_seen": 25931510, + "step": 1241 + }, + { + "epoch": 0.14934167017375097, + "flos": 21911730078600.0, + "grad_norm": 2.4637590975039436, + "learning_rate": 3.85256614155776e-06, + "loss": 0.7425, + "num_input_tokens_seen": 25951360, + "step": 1242 + }, + { + "epoch": 0.14946191306439008, + "flos": 12408516145920.0, + "grad_norm": 2.68126700869414, + "learning_rate": 3.852272462378535e-06, + "loss": 0.7269, + "num_input_tokens_seen": 25968955, + "step": 1243 + }, + { + "epoch": 0.14958215595502916, + "flos": 11409272368920.0, + "grad_norm": 2.0796318614232256, + "learning_rate": 3.85197850221105e-06, + "loss": 0.7739, + "num_input_tokens_seen": 25984975, + "step": 1244 + }, + { + "epoch": 0.14970239884566824, + "flos": 24315587709960.0, + "grad_norm": 2.762811986644651, + "learning_rate": 3.851684261099899e-06, + "loss": 0.7447, + "num_input_tokens_seen": 26006435, + "step": 1245 + }, + { + "epoch": 0.14982264173630733, + "flos": 13001978266080.0, + "grad_norm": 1.9782871366708088, + "learning_rate": 3.851389739089718e-06, + "loss": 0.8516, + "num_input_tokens_seen": 26022775, + "step": 1246 + }, + { + "epoch": 0.14994288462694644, + "flos": 23797462820520.0, + "grad_norm": 3.2485483302026013, + "learning_rate": 3.851094936225186e-06, + "loss": 0.7961, + "num_input_tokens_seen": 26043380, + "step": 1247 + }, + { + "epoch": 0.15006312751758552, + "flos": 23343155427720.0, + "grad_norm": 1.711878408962726, + "learning_rate": 3.850799852551024e-06, + "loss": 0.7586, + "num_input_tokens_seen": 26065520, + "step": 1248 + }, + { + "epoch": 0.1501833704082246, + "flos": 12111626786040.0, + "grad_norm": 2.620286251475753, + "learning_rate": 3.850504488111995e-06, + "loss": 0.8402, + "num_input_tokens_seen": 26081915, + "step": 1249 + }, + { + "epoch": 0.15030361329886371, + "flos": 17182549702560.0, + "grad_norm": 1.9366981323671242, + "learning_rate": 3.850208842952907e-06, + "loss": 0.8107, + "num_input_tokens_seen": 26100440, + "step": 1250 + }, + { + "epoch": 0.1504238561895028, + "flos": 18780574473000.0, + "grad_norm": 2.5898417246999057, + "learning_rate": 3.849912917118608e-06, + "loss": 0.7781, + "num_input_tokens_seen": 26121200, + "step": 1251 + }, + { + "epoch": 0.15054409908014188, + "flos": 38374078129320.0, + "grad_norm": 0.9770141777720039, + "learning_rate": 3.849616710653992e-06, + "loss": 0.6196, + "num_input_tokens_seen": 26182390, + "step": 1252 + }, + { + "epoch": 0.150664341970781, + "flos": 13785843606120.0, + "grad_norm": 2.8613224589321034, + "learning_rate": 3.84932022360399e-06, + "loss": 0.7429, + "num_input_tokens_seen": 26200775, + "step": 1253 + }, + { + "epoch": 0.15078458486142007, + "flos": 16215372973680.0, + "grad_norm": 3.1790455913028413, + "learning_rate": 3.849023456013581e-06, + "loss": 0.8112, + "num_input_tokens_seen": 26218055, + "step": 1254 + }, + { + "epoch": 0.15090482775205916, + "flos": 19537406698560.0, + "grad_norm": 2.4745850957027167, + "learning_rate": 3.848726407927784e-06, + "loss": 0.6083, + "num_input_tokens_seen": 26238160, + "step": 1255 + }, + { + "epoch": 0.15102507064269824, + "flos": 15945896647800.0, + "grad_norm": 2.7513271943857216, + "learning_rate": 3.84842907939166e-06, + "loss": 0.852, + "num_input_tokens_seen": 26257105, + "step": 1256 + }, + { + "epoch": 0.15114531353333735, + "flos": 16702222314000.0, + "grad_norm": 3.1659502590992834, + "learning_rate": 3.8481314704503146e-06, + "loss": 0.6905, + "num_input_tokens_seen": 26276655, + "step": 1257 + }, + { + "epoch": 0.15126555642397643, + "flos": 14379653985840.0, + "grad_norm": 3.919537459352018, + "learning_rate": 3.847833581148895e-06, + "loss": 0.8579, + "num_input_tokens_seen": 26295285, + "step": 1258 + }, + { + "epoch": 0.15138579931461552, + "flos": 21071613972360.0, + "grad_norm": 3.1384778630434793, + "learning_rate": 3.84753541153259e-06, + "loss": 0.7866, + "num_input_tokens_seen": 26314575, + "step": 1259 + }, + { + "epoch": 0.15150604220525463, + "flos": 16188593138880.0, + "grad_norm": 1.6978302291992027, + "learning_rate": 3.847236961646633e-06, + "loss": 0.8227, + "num_input_tokens_seen": 26333275, + "step": 1260 + }, + { + "epoch": 0.1516262850958937, + "flos": 9410784814920.0, + "grad_norm": 2.9773257420920416, + "learning_rate": 3.846938231536296e-06, + "loss": 0.7722, + "num_input_tokens_seen": 26348615, + "step": 1261 + }, + { + "epoch": 0.1517465279865328, + "flos": 15944250329880.0, + "grad_norm": 1.7255342623664667, + "learning_rate": 3.8466392212468995e-06, + "loss": 0.8, + "num_input_tokens_seen": 26368525, + "step": 1262 + }, + { + "epoch": 0.15186677087717187, + "flos": 42871321909320.0, + "grad_norm": 0.8082641814498343, + "learning_rate": 3.8463399308238e-06, + "loss": 0.6404, + "num_input_tokens_seen": 26427350, + "step": 1263 + }, + { + "epoch": 0.15198701376781099, + "flos": 23990652116880.0, + "grad_norm": 1.8808693256689337, + "learning_rate": 3.846040360312402e-06, + "loss": 0.6254, + "num_input_tokens_seen": 26450330, + "step": 1264 + }, + { + "epoch": 0.15210725665845007, + "flos": 20832938296200.0, + "grad_norm": 2.336196527823327, + "learning_rate": 3.8457405097581485e-06, + "loss": 0.7996, + "num_input_tokens_seen": 26469040, + "step": 1265 + }, + { + "epoch": 0.15222749954908915, + "flos": 14568759147360.0, + "grad_norm": 2.3731037399951163, + "learning_rate": 3.8454403792065275e-06, + "loss": 0.77, + "num_input_tokens_seen": 26487580, + "step": 1266 + }, + { + "epoch": 0.15234774243972826, + "flos": 15594466159560.0, + "grad_norm": 2.852582357222369, + "learning_rate": 3.845139968703068e-06, + "loss": 0.8393, + "num_input_tokens_seen": 26504820, + "step": 1267 + }, + { + "epoch": 0.15246798533036734, + "flos": 19023619223640.0, + "grad_norm": 2.7955522019723826, + "learning_rate": 3.844839278293342e-06, + "loss": 0.8189, + "num_input_tokens_seen": 26525390, + "step": 1268 + }, + { + "epoch": 0.15258822822100643, + "flos": 18915645065520.0, + "grad_norm": 2.36173549574235, + "learning_rate": 3.8445383080229654e-06, + "loss": 0.7481, + "num_input_tokens_seen": 26541125, + "step": 1269 + }, + { + "epoch": 0.1527084711116455, + "flos": 18511351426800.0, + "grad_norm": 2.9255871370958553, + "learning_rate": 3.844237057937593e-06, + "loss": 0.7185, + "num_input_tokens_seen": 26559850, + "step": 1270 + }, + { + "epoch": 0.15282871400228462, + "flos": 21452642031840.0, + "grad_norm": 3.4266637450575588, + "learning_rate": 3.843935528082926e-06, + "loss": 0.7604, + "num_input_tokens_seen": 26580595, + "step": 1271 + }, + { + "epoch": 0.1529489568929237, + "flos": 15267409349160.0, + "grad_norm": 2.091816023196673, + "learning_rate": 3.843633718504704e-06, + "loss": 0.8397, + "num_input_tokens_seen": 26598760, + "step": 1272 + }, + { + "epoch": 0.1530691997835628, + "flos": 14676511685760.0, + "grad_norm": 4.152658158149113, + "learning_rate": 3.843331629248715e-06, + "loss": 0.8767, + "num_input_tokens_seen": 26616080, + "step": 1273 + }, + { + "epoch": 0.1531894426742019, + "flos": 21096240929880.0, + "grad_norm": 4.183126414578192, + "learning_rate": 3.843029260360782e-06, + "loss": 0.7612, + "num_input_tokens_seen": 26634170, + "step": 1274 + }, + { + "epoch": 0.15330968556484098, + "flos": 16269439202640.0, + "grad_norm": 2.2767825343005628, + "learning_rate": 3.8427266118867755e-06, + "loss": 0.7731, + "num_input_tokens_seen": 26653640, + "step": 1275 + }, + { + "epoch": 0.15342992845548006, + "flos": 20185409947080.0, + "grad_norm": 2.2934223616677767, + "learning_rate": 3.842423683872608e-06, + "loss": 0.8074, + "num_input_tokens_seen": 26673935, + "step": 1276 + }, + { + "epoch": 0.15355017134611917, + "flos": 14325556096920.0, + "grad_norm": 2.918836023014662, + "learning_rate": 3.842120476364232e-06, + "loss": 0.7648, + "num_input_tokens_seen": 26692105, + "step": 1277 + }, + { + "epoch": 0.15367041423675826, + "flos": 13488542666760.0, + "grad_norm": 2.7448854569125145, + "learning_rate": 3.841816989407644e-06, + "loss": 0.8109, + "num_input_tokens_seen": 26707315, + "step": 1278 + }, + { + "epoch": 0.15379065712739734, + "flos": 30472330920000.0, + "grad_norm": 2.962594487095847, + "learning_rate": 3.841513223048884e-06, + "loss": 0.7524, + "num_input_tokens_seen": 26727720, + "step": 1279 + }, + { + "epoch": 0.15391090001803642, + "flos": 16134748529640.0, + "grad_norm": 2.9044956942181774, + "learning_rate": 3.841209177334031e-06, + "loss": 0.7554, + "num_input_tokens_seen": 26745800, + "step": 1280 + }, + { + "epoch": 0.15403114290867553, + "flos": 11247453601560.0, + "grad_norm": 2.8961310656695307, + "learning_rate": 3.84090485230921e-06, + "loss": 0.7372, + "num_input_tokens_seen": 26763760, + "step": 1281 + }, + { + "epoch": 0.15415138579931462, + "flos": 13081621251360.0, + "grad_norm": 4.553888633642908, + "learning_rate": 3.840600248020588e-06, + "loss": 0.7371, + "num_input_tokens_seen": 26780420, + "step": 1282 + }, + { + "epoch": 0.1542716286899537, + "flos": 8270550032880.0, + "grad_norm": 2.6530609734204234, + "learning_rate": 3.840295364514371e-06, + "loss": 0.7754, + "num_input_tokens_seen": 26797520, + "step": 1283 + }, + { + "epoch": 0.1543918715805928, + "flos": 12705342185880.0, + "grad_norm": 2.802889021484842, + "learning_rate": 3.83999020183681e-06, + "loss": 0.7611, + "num_input_tokens_seen": 26815935, + "step": 1284 + }, + { + "epoch": 0.1545121144712319, + "flos": 12976401509760.0, + "grad_norm": 2.2789734232525842, + "learning_rate": 3.839684760034199e-06, + "loss": 0.7658, + "num_input_tokens_seen": 26833860, + "step": 1285 + }, + { + "epoch": 0.15463235736187098, + "flos": 20697962683560.0, + "grad_norm": 2.9011356103412065, + "learning_rate": 3.8393790391528716e-06, + "loss": 0.6285, + "num_input_tokens_seen": 26854275, + "step": 1286 + }, + { + "epoch": 0.15475260025251006, + "flos": 16728780529080.0, + "grad_norm": 2.9755369006461563, + "learning_rate": 3.8390730392392075e-06, + "loss": 0.8706, + "num_input_tokens_seen": 26873975, + "step": 1287 + }, + { + "epoch": 0.15487284314314917, + "flos": 12840444438360.0, + "grad_norm": 2.7996519144860055, + "learning_rate": 3.838766760339626e-06, + "loss": 0.7849, + "num_input_tokens_seen": 26892220, + "step": 1288 + }, + { + "epoch": 0.15499308603378825, + "flos": 14674928687760.0, + "grad_norm": 3.278380383488587, + "learning_rate": 3.838460202500587e-06, + "loss": 0.7579, + "num_input_tokens_seen": 26907730, + "step": 1289 + }, + { + "epoch": 0.15511332892442733, + "flos": 11592647077680.0, + "grad_norm": 3.5122402597417617, + "learning_rate": 3.838153365768599e-06, + "loss": 0.7232, + "num_input_tokens_seen": 26923960, + "step": 1290 + }, + { + "epoch": 0.15523357181506645, + "flos": 30364989961080.0, + "grad_norm": 2.8564276174665593, + "learning_rate": 3.837846250190206e-06, + "loss": 0.7263, + "num_input_tokens_seen": 26946545, + "step": 1291 + }, + { + "epoch": 0.15535381470570553, + "flos": 13488700966560.0, + "grad_norm": 2.7067316062854028, + "learning_rate": 3.837538855811998e-06, + "loss": 0.7522, + "num_input_tokens_seen": 26964440, + "step": 1292 + }, + { + "epoch": 0.1554740575963446, + "flos": 10113550811520.0, + "grad_norm": 2.6725785963100486, + "learning_rate": 3.837231182680606e-06, + "loss": 0.6945, + "num_input_tokens_seen": 26982125, + "step": 1293 + }, + { + "epoch": 0.1555943004869837, + "flos": 15240946113960.0, + "grad_norm": 2.6056469995014178, + "learning_rate": 3.836923230842706e-06, + "loss": 0.7549, + "num_input_tokens_seen": 27000960, + "step": 1294 + }, + { + "epoch": 0.1557145433776228, + "flos": 16157507549520.0, + "grad_norm": 2.451161661924689, + "learning_rate": 3.836615000345011e-06, + "loss": 0.7971, + "num_input_tokens_seen": 27018860, + "step": 1295 + }, + { + "epoch": 0.1558347862682619, + "flos": 14460531709560.0, + "grad_norm": 2.547476013040716, + "learning_rate": 3.836306491234282e-06, + "loss": 0.7648, + "num_input_tokens_seen": 27036430, + "step": 1296 + }, + { + "epoch": 0.15595502915890097, + "flos": 12569606734200.0, + "grad_norm": 2.6325530425231696, + "learning_rate": 3.835997703557317e-06, + "loss": 0.7233, + "num_input_tokens_seen": 27052890, + "step": 1297 + }, + { + "epoch": 0.15607527204954008, + "flos": 14406972039960.0, + "grad_norm": 1.8736646347500887, + "learning_rate": 3.83568863736096e-06, + "loss": 0.7877, + "num_input_tokens_seen": 27071480, + "step": 1298 + }, + { + "epoch": 0.15619551494017916, + "flos": 13516050680640.0, + "grad_norm": 2.5921120125032733, + "learning_rate": 3.8353792926920975e-06, + "loss": 0.8617, + "num_input_tokens_seen": 27089850, + "step": 1299 + }, + { + "epoch": 0.15631575783081825, + "flos": 14542137612360.0, + "grad_norm": 3.7370549884330395, + "learning_rate": 3.835069669597655e-06, + "loss": 0.7907, + "num_input_tokens_seen": 27107960, + "step": 1300 + }, + { + "epoch": 0.15643600072145733, + "flos": 15190077540960.0, + "grad_norm": 2.516523496921712, + "learning_rate": 3.834759768124603e-06, + "loss": 0.7864, + "num_input_tokens_seen": 27126555, + "step": 1301 + }, + { + "epoch": 0.15655624361209644, + "flos": 13537986541560.0, + "grad_norm": 5.844255558115546, + "learning_rate": 3.834449588319953e-06, + "loss": 0.7408, + "num_input_tokens_seen": 27144310, + "step": 1302 + }, + { + "epoch": 0.15667648650273552, + "flos": 18484856531640.0, + "grad_norm": 1.996178253833683, + "learning_rate": 3.834139130230758e-06, + "loss": 0.8295, + "num_input_tokens_seen": 27163335, + "step": 1303 + }, + { + "epoch": 0.1567967293933746, + "flos": 18186859073160.0, + "grad_norm": 1.7551858902429285, + "learning_rate": 3.833828393904117e-06, + "loss": 0.7945, + "num_input_tokens_seen": 27183335, + "step": 1304 + }, + { + "epoch": 0.15691697228401372, + "flos": 13995174990720.0, + "grad_norm": 2.846422553513342, + "learning_rate": 3.833517379387165e-06, + "loss": 0.7504, + "num_input_tokens_seen": 27199510, + "step": 1305 + }, + { + "epoch": 0.1570372151746528, + "flos": 18160142558280.0, + "grad_norm": 1.9804728430831717, + "learning_rate": 3.833206086727085e-06, + "loss": 0.8743, + "num_input_tokens_seen": 27218580, + "step": 1306 + }, + { + "epoch": 0.15715745806529188, + "flos": 18213923847600.0, + "grad_norm": 2.9119963902251955, + "learning_rate": 3.8328945159710994e-06, + "loss": 0.6852, + "num_input_tokens_seen": 27238480, + "step": 1307 + }, + { + "epoch": 0.157277700955931, + "flos": 16000057856640.0, + "grad_norm": 2.1272015106918865, + "learning_rate": 3.832582667166473e-06, + "loss": 0.8627, + "num_input_tokens_seen": 27258010, + "step": 1308 + }, + { + "epoch": 0.15739794384657008, + "flos": 17969201119080.0, + "grad_norm": 2.0182454426427046, + "learning_rate": 3.8322705403605125e-06, + "loss": 0.8034, + "num_input_tokens_seen": 27278075, + "step": 1309 + }, + { + "epoch": 0.15751818673720916, + "flos": 12945379240320.0, + "grad_norm": 2.1645915902952915, + "learning_rate": 3.831958135600568e-06, + "loss": 0.7975, + "num_input_tokens_seen": 27295345, + "step": 1310 + }, + { + "epoch": 0.15763842962784824, + "flos": 13029676239720.0, + "grad_norm": 2.251103236674304, + "learning_rate": 3.831645452934032e-06, + "loss": 0.7863, + "num_input_tokens_seen": 27313495, + "step": 1311 + }, + { + "epoch": 0.15775867251848735, + "flos": 19780989668520.0, + "grad_norm": 1.891822372727845, + "learning_rate": 3.831332492408336e-06, + "loss": 0.7925, + "num_input_tokens_seen": 27334625, + "step": 1312 + }, + { + "epoch": 0.15787891540912644, + "flos": 14051710696560.0, + "grad_norm": 2.090447967743054, + "learning_rate": 3.831019254070957e-06, + "loss": 0.6729, + "num_input_tokens_seen": 27352130, + "step": 1313 + }, + { + "epoch": 0.15799915829976552, + "flos": 19995513286560.0, + "grad_norm": 3.572082481465753, + "learning_rate": 3.8307057379694135e-06, + "loss": 0.9276, + "num_input_tokens_seen": 27371185, + "step": 1314 + }, + { + "epoch": 0.15811940119040463, + "flos": 14913636023880.0, + "grad_norm": 2.1898533832121774, + "learning_rate": 3.830391944151264e-06, + "loss": 0.8041, + "num_input_tokens_seen": 27386785, + "step": 1315 + }, + { + "epoch": 0.1582396440810437, + "flos": 23938960384920.0, + "grad_norm": 2.3248139460783985, + "learning_rate": 3.830077872664114e-06, + "loss": 0.6525, + "num_input_tokens_seen": 27407630, + "step": 1316 + }, + { + "epoch": 0.1583598869716828, + "flos": 24827538907200.0, + "grad_norm": 2.402167585163086, + "learning_rate": 3.829763523555604e-06, + "loss": 0.7244, + "num_input_tokens_seen": 27427750, + "step": 1317 + }, + { + "epoch": 0.15848012986232188, + "flos": 18079011554880.0, + "grad_norm": 3.7320378224213373, + "learning_rate": 3.829448896873423e-06, + "loss": 0.767, + "num_input_tokens_seen": 27446570, + "step": 1318 + }, + { + "epoch": 0.158600372752961, + "flos": 16540276906800.0, + "grad_norm": 2.9242499995721225, + "learning_rate": 3.829133992665299e-06, + "loss": 0.7745, + "num_input_tokens_seen": 27465415, + "step": 1319 + }, + { + "epoch": 0.15872061564360007, + "flos": 20481634447800.0, + "grad_norm": 3.071674229375754, + "learning_rate": 3.828818810979002e-06, + "loss": 0.8663, + "num_input_tokens_seen": 27483465, + "step": 1320 + }, + { + "epoch": 0.15884085853423915, + "flos": 17350478842200.0, + "grad_norm": 2.4155662815710195, + "learning_rate": 3.8285033518623454e-06, + "loss": 0.7886, + "num_input_tokens_seen": 27503435, + "step": 1321 + }, + { + "epoch": 0.15896110142487826, + "flos": 16917854030640.0, + "grad_norm": 3.04825361314095, + "learning_rate": 3.8281876153631845e-06, + "loss": 0.7942, + "num_input_tokens_seen": 27519910, + "step": 1322 + }, + { + "epoch": 0.15908134431551735, + "flos": 10681372855440.0, + "grad_norm": 2.5717166671127987, + "learning_rate": 3.827871601529416e-06, + "loss": 0.6214, + "num_input_tokens_seen": 27538150, + "step": 1323 + }, + { + "epoch": 0.15920158720615643, + "flos": 14757800988960.0, + "grad_norm": 1.9695912940236546, + "learning_rate": 3.827555310408979e-06, + "loss": 0.7929, + "num_input_tokens_seen": 27557265, + "step": 1324 + }, + { + "epoch": 0.1593218300967955, + "flos": 18186447493680.0, + "grad_norm": 1.6903404772053423, + "learning_rate": 3.827238742049854e-06, + "loss": 0.8051, + "num_input_tokens_seen": 27577280, + "step": 1325 + }, + { + "epoch": 0.15944207298743462, + "flos": 20778143888160.0, + "grad_norm": 2.7491715657249416, + "learning_rate": 3.826921896500066e-06, + "loss": 0.5115, + "num_input_tokens_seen": 27598285, + "step": 1326 + }, + { + "epoch": 0.1595623158780737, + "flos": 16808455174320.0, + "grad_norm": 2.166142864978906, + "learning_rate": 3.826604773807678e-06, + "loss": 0.7667, + "num_input_tokens_seen": 27615980, + "step": 1327 + }, + { + "epoch": 0.1596825587687128, + "flos": 14399057049960.0, + "grad_norm": 3.2298215796733163, + "learning_rate": 3.826287374020798e-06, + "loss": 0.7184, + "num_input_tokens_seen": 27630505, + "step": 1328 + }, + { + "epoch": 0.1598028016593519, + "flos": 16566550182240.0, + "grad_norm": 1.9107110401630212, + "learning_rate": 3.825969697187575e-06, + "loss": 0.8051, + "num_input_tokens_seen": 27649555, + "step": 1329 + }, + { + "epoch": 0.15992304454999098, + "flos": 14971533108000.0, + "grad_norm": 2.2763185892763897, + "learning_rate": 3.8256517433562015e-06, + "loss": 0.6922, + "num_input_tokens_seen": 27667215, + "step": 1330 + }, + { + "epoch": 0.16004328744063007, + "flos": 12894668967120.0, + "grad_norm": 5.233860093027771, + "learning_rate": 3.82533351257491e-06, + "loss": 0.8984, + "num_input_tokens_seen": 27684885, + "step": 1331 + }, + { + "epoch": 0.16016353033126918, + "flos": 17648001401280.0, + "grad_norm": 2.370729622698297, + "learning_rate": 3.825015004891975e-06, + "loss": 0.876, + "num_input_tokens_seen": 27703345, + "step": 1332 + }, + { + "epoch": 0.16028377322190826, + "flos": 20264546373000.0, + "grad_norm": 2.300543625482421, + "learning_rate": 3.824696220355716e-06, + "loss": 0.7433, + "num_input_tokens_seen": 27724655, + "step": 1333 + }, + { + "epoch": 0.16040401611254734, + "flos": 15326066272320.0, + "grad_norm": 1.8076383567819372, + "learning_rate": 3.824377159014491e-06, + "loss": 0.7774, + "num_input_tokens_seen": 27745270, + "step": 1334 + }, + { + "epoch": 0.16052425900318643, + "flos": 15536759035200.0, + "grad_norm": 2.0841977297138077, + "learning_rate": 3.824057820916702e-06, + "loss": 0.8352, + "num_input_tokens_seen": 27762195, + "step": 1335 + }, + { + "epoch": 0.16064450189382554, + "flos": 11302058049840.0, + "grad_norm": 2.895796649801141, + "learning_rate": 3.8237382061107904e-06, + "loss": 0.7008, + "num_input_tokens_seen": 27778635, + "step": 1336 + }, + { + "epoch": 0.16076474478446462, + "flos": 15486650301240.0, + "grad_norm": 3.7945298224531494, + "learning_rate": 3.823418314645243e-06, + "loss": 0.7656, + "num_input_tokens_seen": 27797230, + "step": 1337 + }, + { + "epoch": 0.1608849876751037, + "flos": 13405195466160.0, + "grad_norm": 2.4300205211008388, + "learning_rate": 3.823098146568588e-06, + "loss": 0.7395, + "num_input_tokens_seen": 27816655, + "step": 1338 + }, + { + "epoch": 0.1610052305657428, + "flos": 21642886951920.0, + "grad_norm": 1.924507571067992, + "learning_rate": 3.822777701929394e-06, + "loss": 0.7065, + "num_input_tokens_seen": 27838200, + "step": 1339 + }, + { + "epoch": 0.1611254734563819, + "flos": 19645444176600.0, + "grad_norm": 2.3019945866896467, + "learning_rate": 3.8224569807762714e-06, + "loss": 0.7239, + "num_input_tokens_seen": 27857240, + "step": 1340 + }, + { + "epoch": 0.16124571634702098, + "flos": 16405206314280.0, + "grad_norm": 1.995311068517161, + "learning_rate": 3.822135983157873e-06, + "loss": 0.7557, + "num_input_tokens_seen": 27876235, + "step": 1341 + }, + { + "epoch": 0.16136595923766006, + "flos": 7953244490160.0, + "grad_norm": 3.0720115406197097, + "learning_rate": 3.821814709122896e-06, + "loss": 0.8276, + "num_input_tokens_seen": 27894005, + "step": 1342 + }, + { + "epoch": 0.16148620212829917, + "flos": 15513651755760.0, + "grad_norm": 2.9270701198800033, + "learning_rate": 3.821493158720076e-06, + "loss": 0.8396, + "num_input_tokens_seen": 27912830, + "step": 1343 + }, + { + "epoch": 0.16160644501893826, + "flos": 12214788630240.0, + "grad_norm": 4.487007883444574, + "learning_rate": 3.821171331998191e-06, + "loss": 0.7028, + "num_input_tokens_seen": 27929080, + "step": 1344 + }, + { + "epoch": 0.16172668790957734, + "flos": 47511108032400.0, + "grad_norm": 0.8024089738434347, + "learning_rate": 3.820849229006064e-06, + "loss": 0.5845, + "num_input_tokens_seen": 27996550, + "step": 1345 + }, + { + "epoch": 0.16184693080021645, + "flos": 17025321629400.0, + "grad_norm": 2.3744986155464343, + "learning_rate": 3.8205268497925564e-06, + "loss": 0.6965, + "num_input_tokens_seen": 28016740, + "step": 1346 + }, + { + "epoch": 0.16196717369085553, + "flos": 12727309706760.0, + "grad_norm": 2.826858995870105, + "learning_rate": 3.8202041944065725e-06, + "loss": 0.7683, + "num_input_tokens_seen": 28032280, + "step": 1347 + }, + { + "epoch": 0.16208741658149461, + "flos": 17480325541320.0, + "grad_norm": 2.40382481022676, + "learning_rate": 3.819881262897061e-06, + "loss": 0.7243, + "num_input_tokens_seen": 28050135, + "step": 1348 + }, + { + "epoch": 0.1622076594721337, + "flos": 18969584654640.0, + "grad_norm": 2.6187933251340865, + "learning_rate": 3.819558055313008e-06, + "loss": 0.7188, + "num_input_tokens_seen": 28070540, + "step": 1349 + }, + { + "epoch": 0.1623279023627728, + "flos": 15753498850440.0, + "grad_norm": 2.1579516786498476, + "learning_rate": 3.819234571703444e-06, + "loss": 0.7543, + "num_input_tokens_seen": 28089085, + "step": 1350 + }, + { + "epoch": 0.1624481452534119, + "flos": 16157095970040.0, + "grad_norm": 1.924260651429978, + "learning_rate": 3.8189108121174435e-06, + "loss": 0.8395, + "num_input_tokens_seen": 28108570, + "step": 1351 + }, + { + "epoch": 0.16256838814405097, + "flos": 19859461235280.0, + "grad_norm": 1.9171524952473027, + "learning_rate": 3.818586776604118e-06, + "loss": 0.8198, + "num_input_tokens_seen": 28128930, + "step": 1352 + }, + { + "epoch": 0.16268863103469008, + "flos": 14703544800240.0, + "grad_norm": 3.404812430706493, + "learning_rate": 3.818262465212625e-06, + "loss": 0.6054, + "num_input_tokens_seen": 28148775, + "step": 1353 + }, + { + "epoch": 0.16280887392532917, + "flos": 13380663488520.0, + "grad_norm": 3.171963093854406, + "learning_rate": 3.817937877992161e-06, + "loss": 0.7609, + "num_input_tokens_seen": 28165790, + "step": 1354 + }, + { + "epoch": 0.16292911681596825, + "flos": 8600899479120.0, + "grad_norm": 2.839801100629707, + "learning_rate": 3.817613014991967e-06, + "loss": 0.8434, + "num_input_tokens_seen": 28181650, + "step": 1355 + }, + { + "epoch": 0.16304935970660733, + "flos": 19131371762040.0, + "grad_norm": 2.3214355217127456, + "learning_rate": 3.817287876261323e-06, + "loss": 0.7439, + "num_input_tokens_seen": 28201705, + "step": 1356 + }, + { + "epoch": 0.16316960259724644, + "flos": 21535640972880.0, + "grad_norm": 1.9025906944964022, + "learning_rate": 3.816962461849553e-06, + "loss": 0.7811, + "num_input_tokens_seen": 28223295, + "step": 1357 + }, + { + "epoch": 0.16328984548788553, + "flos": 15270765304920.0, + "grad_norm": 2.544362238657768, + "learning_rate": 3.8166367718060235e-06, + "loss": 0.824, + "num_input_tokens_seen": 28242905, + "step": 1358 + }, + { + "epoch": 0.1634100883785246, + "flos": 13164810152160.0, + "grad_norm": 2.9903378658711777, + "learning_rate": 3.816310806180139e-06, + "loss": 0.7267, + "num_input_tokens_seen": 28261035, + "step": 1359 + }, + { + "epoch": 0.16353033126916372, + "flos": 17998228811040.0, + "grad_norm": 1.9414112850868435, + "learning_rate": 3.81598456502135e-06, + "loss": 0.7925, + "num_input_tokens_seen": 28280775, + "step": 1360 + }, + { + "epoch": 0.1636505741598028, + "flos": 14534001002640.0, + "grad_norm": 2.124780980211181, + "learning_rate": 3.8156580483791455e-06, + "loss": 0.856, + "num_input_tokens_seen": 28295685, + "step": 1361 + }, + { + "epoch": 0.16377081705044189, + "flos": 20832969956160.0, + "grad_norm": 2.9197531763035496, + "learning_rate": 3.815331256303059e-06, + "loss": 0.7518, + "num_input_tokens_seen": 28315435, + "step": 1362 + }, + { + "epoch": 0.163891059941081, + "flos": 16026552751800.0, + "grad_norm": 2.3413162820031483, + "learning_rate": 3.815004188842665e-06, + "loss": 0.7372, + "num_input_tokens_seen": 28333195, + "step": 1363 + }, + { + "epoch": 0.16401130283172008, + "flos": 19645222556880.0, + "grad_norm": 3.2785387736266505, + "learning_rate": 3.814676846047578e-06, + "loss": 0.7761, + "num_input_tokens_seen": 28353790, + "step": 1364 + }, + { + "epoch": 0.16413154572235916, + "flos": 24233095328280.0, + "grad_norm": 1.9549772101242262, + "learning_rate": 3.8143492279674565e-06, + "loss": 0.6862, + "num_input_tokens_seen": 28376205, + "step": 1365 + }, + { + "epoch": 0.16425178861299825, + "flos": 29504937062760.0, + "grad_norm": 0.9377334621926826, + "learning_rate": 3.8140213346519997e-06, + "loss": 0.6499, + "num_input_tokens_seen": 28426520, + "step": 1366 + }, + { + "epoch": 0.16437203150363736, + "flos": 18646010439840.0, + "grad_norm": 1.9014703206372383, + "learning_rate": 3.813693166150948e-06, + "loss": 0.7462, + "num_input_tokens_seen": 28446450, + "step": 1367 + }, + { + "epoch": 0.16449227439427644, + "flos": 17187583636200.0, + "grad_norm": 5.080949297238499, + "learning_rate": 3.813364722514086e-06, + "loss": 0.8284, + "num_input_tokens_seen": 28464505, + "step": 1368 + }, + { + "epoch": 0.16461251728491552, + "flos": 9836412775320.0, + "grad_norm": 2.823071511914471, + "learning_rate": 3.8130360037912368e-06, + "loss": 0.7883, + "num_input_tokens_seen": 28480670, + "step": 1369 + }, + { + "epoch": 0.16473276017555463, + "flos": 16837197926640.0, + "grad_norm": 2.4379764971456894, + "learning_rate": 3.812707010032268e-06, + "loss": 0.7932, + "num_input_tokens_seen": 28499445, + "step": 1370 + }, + { + "epoch": 0.16485300306619372, + "flos": 18159920938560.0, + "grad_norm": 2.4906188428699774, + "learning_rate": 3.8123777412870863e-06, + "loss": 0.7757, + "num_input_tokens_seen": 28518665, + "step": 1371 + }, + { + "epoch": 0.1649732459568328, + "flos": 15433502211120.0, + "grad_norm": 2.1902206940255127, + "learning_rate": 3.812048197605643e-06, + "loss": 0.7649, + "num_input_tokens_seen": 28537280, + "step": 1372 + }, + { + "epoch": 0.16509348884747188, + "flos": 14811930537840.0, + "grad_norm": 3.2473886709534505, + "learning_rate": 3.8117183790379277e-06, + "loss": 0.7891, + "num_input_tokens_seen": 28555450, + "step": 1373 + }, + { + "epoch": 0.165213731738111, + "flos": 7980150964800.0, + "grad_norm": 3.0416587585625114, + "learning_rate": 3.811388285633976e-06, + "loss": 0.9137, + "num_input_tokens_seen": 28571155, + "step": 1374 + }, + { + "epoch": 0.16533397462875007, + "flos": 21994539059880.0, + "grad_norm": 2.9385007411663113, + "learning_rate": 3.811057917443861e-06, + "loss": 0.6077, + "num_input_tokens_seen": 28590140, + "step": 1375 + }, + { + "epoch": 0.16545421751938916, + "flos": 48336348448800.0, + "grad_norm": 0.8640569938498027, + "learning_rate": 3.8107272745177e-06, + "loss": 0.6978, + "num_input_tokens_seen": 28662190, + "step": 1376 + }, + { + "epoch": 0.16557446041002827, + "flos": 16459050923520.0, + "grad_norm": 2.3517507686846906, + "learning_rate": 3.8103963569056513e-06, + "loss": 0.773, + "num_input_tokens_seen": 28681045, + "step": 1377 + }, + { + "epoch": 0.16569470330066735, + "flos": 18020038032120.0, + "grad_norm": 1.7007456266276035, + "learning_rate": 3.8100651646579146e-06, + "loss": 0.8695, + "num_input_tokens_seen": 28699975, + "step": 1378 + }, + { + "epoch": 0.16581494619130643, + "flos": 10917863994360.0, + "grad_norm": 2.0837134459549653, + "learning_rate": 3.8097336978247317e-06, + "loss": 0.9039, + "num_input_tokens_seen": 28716400, + "step": 1379 + }, + { + "epoch": 0.16593518908194552, + "flos": 12705342185880.0, + "grad_norm": 2.87046057537722, + "learning_rate": 3.8094019564563854e-06, + "loss": 0.8675, + "num_input_tokens_seen": 28733050, + "step": 1380 + }, + { + "epoch": 0.16605543197258463, + "flos": 14919904695960.0, + "grad_norm": 3.5267031944512897, + "learning_rate": 3.809069940603201e-06, + "loss": 0.7568, + "num_input_tokens_seen": 28750725, + "step": 1381 + }, + { + "epoch": 0.1661756748632237, + "flos": 10329309168000.0, + "grad_norm": 2.472329357649938, + "learning_rate": 3.8087376503155452e-06, + "loss": 0.76, + "num_input_tokens_seen": 28767930, + "step": 1382 + }, + { + "epoch": 0.1662959177538628, + "flos": 48720949592400.0, + "grad_norm": 1.5776951270476536, + "learning_rate": 3.808405085643826e-06, + "loss": 0.6011, + "num_input_tokens_seen": 28832530, + "step": 1383 + }, + { + "epoch": 0.1664161606445019, + "flos": 15162537867120.0, + "grad_norm": 2.234252499510157, + "learning_rate": 3.8080722466384925e-06, + "loss": 0.8834, + "num_input_tokens_seen": 28850100, + "step": 1384 + }, + { + "epoch": 0.166536403535141, + "flos": 18507362271840.0, + "grad_norm": 2.4370131373186363, + "learning_rate": 3.8077391333500376e-06, + "loss": 0.6817, + "num_input_tokens_seen": 28868960, + "step": 1385 + }, + { + "epoch": 0.16665664642578007, + "flos": 18645978779880.0, + "grad_norm": 1.8400290491050115, + "learning_rate": 3.8074057458289934e-06, + "loss": 0.7421, + "num_input_tokens_seen": 28889370, + "step": 1386 + }, + { + "epoch": 0.16677688931641918, + "flos": 16243292567040.0, + "grad_norm": 2.7790971631953547, + "learning_rate": 3.807072084125934e-06, + "loss": 0.8014, + "num_input_tokens_seen": 28910940, + "step": 1387 + }, + { + "epoch": 0.16689713220705826, + "flos": 12354259957200.0, + "grad_norm": 2.612702857359733, + "learning_rate": 3.806738148291477e-06, + "loss": 0.7726, + "num_input_tokens_seen": 28927485, + "step": 1388 + }, + { + "epoch": 0.16701737509769735, + "flos": 26636066480760.0, + "grad_norm": 2.3892336517761383, + "learning_rate": 3.8064039383762793e-06, + "loss": 0.6926, + "num_input_tokens_seen": 28949570, + "step": 1389 + }, + { + "epoch": 0.16713761798833643, + "flos": 17026176448320.0, + "grad_norm": 2.4024309840297438, + "learning_rate": 3.8060694544310396e-06, + "loss": 0.7605, + "num_input_tokens_seen": 28967800, + "step": 1390 + }, + { + "epoch": 0.16725786087897554, + "flos": 18538859440680.0, + "grad_norm": 2.368347991407627, + "learning_rate": 3.8057346965065006e-06, + "loss": 0.7701, + "num_input_tokens_seen": 28988750, + "step": 1391 + }, + { + "epoch": 0.16737810376961462, + "flos": 23370505141800.0, + "grad_norm": 2.471367522760658, + "learning_rate": 3.805399664653443e-06, + "loss": 0.8401, + "num_input_tokens_seen": 29010610, + "step": 1392 + }, + { + "epoch": 0.1674983466602537, + "flos": 20506356385200.0, + "grad_norm": 3.143519647550518, + "learning_rate": 3.805064358922692e-06, + "loss": 0.733, + "num_input_tokens_seen": 29028620, + "step": 1393 + }, + { + "epoch": 0.16761858955089282, + "flos": 15919021833120.0, + "grad_norm": 2.8324428916891113, + "learning_rate": 3.8047287793651136e-06, + "loss": 0.7816, + "num_input_tokens_seen": 29049785, + "step": 1394 + }, + { + "epoch": 0.1677388324415319, + "flos": 17430818346600.0, + "grad_norm": 2.1641137292630472, + "learning_rate": 3.8043929260316137e-06, + "loss": 0.867, + "num_input_tokens_seen": 29067660, + "step": 1395 + }, + { + "epoch": 0.16785907533217098, + "flos": 15028163793720.0, + "grad_norm": 2.198452045145836, + "learning_rate": 3.8040567989731417e-06, + "loss": 0.82, + "num_input_tokens_seen": 29085325, + "step": 1396 + }, + { + "epoch": 0.16797931822281006, + "flos": 11409810588240.0, + "grad_norm": 2.3180047577882377, + "learning_rate": 3.8037203982406876e-06, + "loss": 0.7956, + "num_input_tokens_seen": 29103210, + "step": 1397 + }, + { + "epoch": 0.16809956111344918, + "flos": 12054141281400.0, + "grad_norm": 2.3706784332392457, + "learning_rate": 3.8033837238852835e-06, + "loss": 0.7161, + "num_input_tokens_seen": 29119630, + "step": 1398 + }, + { + "epoch": 0.16821980400408826, + "flos": 17025669888960.0, + "grad_norm": 3.8821260701324287, + "learning_rate": 3.8030467759580017e-06, + "loss": 0.6774, + "num_input_tokens_seen": 29140270, + "step": 1399 + }, + { + "epoch": 0.16834004689472734, + "flos": 15187513084200.0, + "grad_norm": 2.347747343965055, + "learning_rate": 3.802709554509958e-06, + "loss": 0.8615, + "num_input_tokens_seen": 29157790, + "step": 1400 + }, + { + "epoch": 0.16846028978536645, + "flos": 19563458354280.0, + "grad_norm": 4.040452020870807, + "learning_rate": 3.8023720595923083e-06, + "loss": 0.7667, + "num_input_tokens_seen": 29176765, + "step": 1401 + }, + { + "epoch": 0.16858053267600553, + "flos": 13758430572120.0, + "grad_norm": 1.9267946869095653, + "learning_rate": 3.80203429125625e-06, + "loss": 0.857, + "num_input_tokens_seen": 29194660, + "step": 1402 + }, + { + "epoch": 0.16870077556664462, + "flos": 20345772356280.0, + "grad_norm": 4.228069269527736, + "learning_rate": 3.8016962495530225e-06, + "loss": 0.688, + "num_input_tokens_seen": 29213570, + "step": 1403 + }, + { + "epoch": 0.1688210184572837, + "flos": 9974174464440.0, + "grad_norm": 3.1709791901000948, + "learning_rate": 3.8013579345339063e-06, + "loss": 0.7446, + "num_input_tokens_seen": 29228155, + "step": 1404 + }, + { + "epoch": 0.1689412613479228, + "flos": 19401797886720.0, + "grad_norm": 2.8241998400127644, + "learning_rate": 3.801019346250224e-06, + "loss": 0.6751, + "num_input_tokens_seen": 29248020, + "step": 1405 + }, + { + "epoch": 0.1690615042385619, + "flos": 15456862770240.0, + "grad_norm": 3.0989548798015596, + "learning_rate": 3.8006804847533395e-06, + "loss": 0.8088, + "num_input_tokens_seen": 29267255, + "step": 1406 + }, + { + "epoch": 0.16918174712920098, + "flos": 15242940691440.0, + "grad_norm": 2.438385924708425, + "learning_rate": 3.8003413500946556e-06, + "loss": 0.8322, + "num_input_tokens_seen": 29287085, + "step": 1407 + }, + { + "epoch": 0.1693019900198401, + "flos": 12381989590800.0, + "grad_norm": 3.224913686144337, + "learning_rate": 3.8000019423256216e-06, + "loss": 0.8041, + "num_input_tokens_seen": 29304570, + "step": 1408 + }, + { + "epoch": 0.16942223291047917, + "flos": 19644589357680.0, + "grad_norm": 1.9876228246583976, + "learning_rate": 3.7996622614977234e-06, + "loss": 0.8646, + "num_input_tokens_seen": 29325480, + "step": 1409 + }, + { + "epoch": 0.16954247580111825, + "flos": 13565716175160.0, + "grad_norm": 2.069700676105575, + "learning_rate": 3.799322307662492e-06, + "loss": 0.7819, + "num_input_tokens_seen": 29343020, + "step": 1410 + }, + { + "epoch": 0.16966271869175734, + "flos": 10160968448880.0, + "grad_norm": 2.7239649075391705, + "learning_rate": 3.798982080871496e-06, + "loss": 0.8296, + "num_input_tokens_seen": 29357880, + "step": 1411 + }, + { + "epoch": 0.16978296158239645, + "flos": 27556237151760.0, + "grad_norm": 2.3078536704697155, + "learning_rate": 3.798641581176349e-06, + "loss": 0.6612, + "num_input_tokens_seen": 29379880, + "step": 1412 + }, + { + "epoch": 0.16990320447303553, + "flos": 20778365507880.0, + "grad_norm": 2.8487185157016937, + "learning_rate": 3.7983008086287044e-06, + "loss": 0.7257, + "num_input_tokens_seen": 29400920, + "step": 1413 + }, + { + "epoch": 0.1700234473636746, + "flos": 14753020335000.0, + "grad_norm": 2.4389052619567937, + "learning_rate": 3.797959763280257e-06, + "loss": 0.7725, + "num_input_tokens_seen": 29419325, + "step": 1414 + }, + { + "epoch": 0.17014369025431372, + "flos": 18209523113160.0, + "grad_norm": 6.061135364596223, + "learning_rate": 3.797618445182743e-06, + "loss": 0.7748, + "num_input_tokens_seen": 29440440, + "step": 1415 + }, + { + "epoch": 0.1702639331449528, + "flos": 11999346873360.0, + "grad_norm": 2.857017504122063, + "learning_rate": 3.79727685438794e-06, + "loss": 0.8254, + "num_input_tokens_seen": 29454350, + "step": 1416 + }, + { + "epoch": 0.1703841760355919, + "flos": 38676851750400.0, + "grad_norm": 0.850795435371224, + "learning_rate": 3.796934990947667e-06, + "loss": 0.6284, + "num_input_tokens_seen": 29515755, + "step": 1417 + }, + { + "epoch": 0.170504418926231, + "flos": 36354283422240.0, + "grad_norm": 0.9251590360068978, + "learning_rate": 3.7965928549137854e-06, + "loss": 0.6495, + "num_input_tokens_seen": 29572290, + "step": 1418 + }, + { + "epoch": 0.17062466181687008, + "flos": 18943374699120.0, + "grad_norm": 4.450365998722241, + "learning_rate": 3.7962504463381953e-06, + "loss": 0.7335, + "num_input_tokens_seen": 29593500, + "step": 1419 + }, + { + "epoch": 0.17074490470750917, + "flos": 15325369753200.0, + "grad_norm": 2.174165464756454, + "learning_rate": 3.7959077652728412e-06, + "loss": 0.7722, + "num_input_tokens_seen": 29611675, + "step": 1420 + }, + { + "epoch": 0.17086514759814825, + "flos": 15324799873920.0, + "grad_norm": 2.562735383922861, + "learning_rate": 3.795564811769707e-06, + "loss": 0.7571, + "num_input_tokens_seen": 29629750, + "step": 1421 + }, + { + "epoch": 0.17098539048878736, + "flos": 20885643146880.0, + "grad_norm": 2.3868283310628167, + "learning_rate": 3.795221585880818e-06, + "loss": 0.7683, + "num_input_tokens_seen": 29650150, + "step": 1422 + }, + { + "epoch": 0.17110563337942644, + "flos": 11868930294960.0, + "grad_norm": 2.384032390572858, + "learning_rate": 3.794878087658242e-06, + "loss": 0.8982, + "num_input_tokens_seen": 29667640, + "step": 1423 + }, + { + "epoch": 0.17122587627006552, + "flos": 21773271870360.0, + "grad_norm": 1.7693793966386258, + "learning_rate": 3.7945343171540873e-06, + "loss": 0.7651, + "num_input_tokens_seen": 29688235, + "step": 1424 + }, + { + "epoch": 0.17134611916070464, + "flos": 18565322675880.0, + "grad_norm": 2.2332007250156556, + "learning_rate": 3.7941902744205033e-06, + "loss": 0.7684, + "num_input_tokens_seen": 29708990, + "step": 1425 + }, + { + "epoch": 0.17146636205134372, + "flos": 10139855746920.0, + "grad_norm": 3.1807288261943505, + "learning_rate": 3.7938459595096817e-06, + "loss": 0.8208, + "num_input_tokens_seen": 29727255, + "step": 1426 + }, + { + "epoch": 0.1715866049419828, + "flos": 17512044329880.0, + "grad_norm": 2.371144025542979, + "learning_rate": 3.7935013724738545e-06, + "loss": 0.8346, + "num_input_tokens_seen": 29747475, + "step": 1427 + }, + { + "epoch": 0.17170684783262188, + "flos": 16619381672760.0, + "grad_norm": 2.234711094900936, + "learning_rate": 3.7931565133652945e-06, + "loss": 0.7598, + "num_input_tokens_seen": 29767270, + "step": 1428 + }, + { + "epoch": 0.171827090723261, + "flos": 19508537306400.0, + "grad_norm": 2.8088293069627914, + "learning_rate": 3.792811382236317e-06, + "loss": 0.6611, + "num_input_tokens_seen": 29785500, + "step": 1429 + }, + { + "epoch": 0.17194733361390008, + "flos": 20645099533080.0, + "grad_norm": 4.662923446939858, + "learning_rate": 3.792465979139279e-06, + "loss": 0.7577, + "num_input_tokens_seen": 29807825, + "step": 1430 + }, + { + "epoch": 0.17206757650453916, + "flos": 48313363317840.0, + "grad_norm": 1.0605831807422346, + "learning_rate": 3.792120304126576e-06, + "loss": 0.7323, + "num_input_tokens_seen": 29870920, + "step": 1431 + }, + { + "epoch": 0.17218781939517827, + "flos": 16297073856360.0, + "grad_norm": 2.0912727390291175, + "learning_rate": 3.791774357250649e-06, + "loss": 0.8246, + "num_input_tokens_seen": 29889470, + "step": 1432 + }, + { + "epoch": 0.17230806228581735, + "flos": 10275781158360.0, + "grad_norm": 3.808507360799076, + "learning_rate": 3.7914281385639757e-06, + "loss": 0.7716, + "num_input_tokens_seen": 29907065, + "step": 1433 + }, + { + "epoch": 0.17242830517645644, + "flos": 15136201271760.0, + "grad_norm": 2.309301793818396, + "learning_rate": 3.7910816481190784e-06, + "loss": 0.7715, + "num_input_tokens_seen": 29926600, + "step": 1434 + }, + { + "epoch": 0.17254854806709552, + "flos": 22588444419480.0, + "grad_norm": 2.195612007969207, + "learning_rate": 3.7907348859685193e-06, + "loss": 0.7378, + "num_input_tokens_seen": 29948025, + "step": 1435 + }, + { + "epoch": 0.17266879095773463, + "flos": 19509803704800.0, + "grad_norm": 2.412517855879567, + "learning_rate": 3.790387852164902e-06, + "loss": 0.7819, + "num_input_tokens_seen": 29968475, + "step": 1436 + }, + { + "epoch": 0.1727890338483737, + "flos": 14811550618320.0, + "grad_norm": 4.075647529688366, + "learning_rate": 3.7900405467608707e-06, + "loss": 0.7473, + "num_input_tokens_seen": 29987740, + "step": 1437 + }, + { + "epoch": 0.1729092767390128, + "flos": 13266452318280.0, + "grad_norm": 2.9090126606864706, + "learning_rate": 3.7896929698091114e-06, + "loss": 0.7491, + "num_input_tokens_seen": 30000275, + "step": 1438 + }, + { + "epoch": 0.1730295196296519, + "flos": 19616796404160.0, + "grad_norm": 3.512421229012068, + "learning_rate": 3.7893451213623518e-06, + "loss": 0.6762, + "num_input_tokens_seen": 30017225, + "step": 1439 + }, + { + "epoch": 0.173149762520291, + "flos": 17458294700520.0, + "grad_norm": 2.7163180293320974, + "learning_rate": 3.7889970014733606e-06, + "loss": 0.8079, + "num_input_tokens_seen": 30036050, + "step": 1440 + }, + { + "epoch": 0.17327000541093007, + "flos": 17107244131800.0, + "grad_norm": 2.0365812324714865, + "learning_rate": 3.7886486101949463e-06, + "loss": 0.7594, + "num_input_tokens_seen": 30056950, + "step": 1441 + }, + { + "epoch": 0.17339024830156918, + "flos": 13297854507240.0, + "grad_norm": 2.1871655889755197, + "learning_rate": 3.7882999475799594e-06, + "loss": 0.8598, + "num_input_tokens_seen": 30074705, + "step": 1442 + }, + { + "epoch": 0.17351049119220827, + "flos": 17080527616920.0, + "grad_norm": 2.0309153354490097, + "learning_rate": 3.787951013681293e-06, + "loss": 0.8021, + "num_input_tokens_seen": 30092470, + "step": 1443 + }, + { + "epoch": 0.17363073408284735, + "flos": 17429045388840.0, + "grad_norm": 2.1346875566480583, + "learning_rate": 3.787601808551879e-06, + "loss": 0.766, + "num_input_tokens_seen": 30112005, + "step": 1444 + }, + { + "epoch": 0.17375097697348643, + "flos": 13515734081040.0, + "grad_norm": 2.6696902260403053, + "learning_rate": 3.7872523322446926e-06, + "loss": 0.815, + "num_input_tokens_seen": 30130610, + "step": 1445 + }, + { + "epoch": 0.17387121986412554, + "flos": 28585933318920.0, + "grad_norm": 2.395968380867473, + "learning_rate": 3.7869025848127478e-06, + "loss": 0.5807, + "num_input_tokens_seen": 30154525, + "step": 1446 + }, + { + "epoch": 0.17399146275476463, + "flos": 14892523321920.0, + "grad_norm": 3.0068303870422137, + "learning_rate": 3.786552566309102e-06, + "loss": 0.7712, + "num_input_tokens_seen": 30172455, + "step": 1447 + }, + { + "epoch": 0.1741117056454037, + "flos": 13994985030960.0, + "grad_norm": 3.2346032103593165, + "learning_rate": 3.7862022767868517e-06, + "loss": 0.8546, + "num_input_tokens_seen": 30189765, + "step": 1448 + }, + { + "epoch": 0.17423194853604282, + "flos": 18591817571040.0, + "grad_norm": 2.276329377333726, + "learning_rate": 3.7858517162991367e-06, + "loss": 0.8264, + "num_input_tokens_seen": 30209560, + "step": 1449 + }, + { + "epoch": 0.1743521914266819, + "flos": 18456810298440.0, + "grad_norm": 3.506752176753099, + "learning_rate": 3.7855008848991363e-06, + "loss": 0.5952, + "num_input_tokens_seen": 30227485, + "step": 1450 + }, + { + "epoch": 0.17447243431732098, + "flos": 18808177466760.0, + "grad_norm": 2.2552359375941817, + "learning_rate": 3.7851497826400714e-06, + "loss": 0.7657, + "num_input_tokens_seen": 30247345, + "step": 1451 + }, + { + "epoch": 0.17459267720796007, + "flos": 26663986074120.0, + "grad_norm": 2.3958869817504516, + "learning_rate": 3.7847984095752034e-06, + "loss": 0.7377, + "num_input_tokens_seen": 30270520, + "step": 1452 + }, + { + "epoch": 0.17471292009859918, + "flos": 14622730396440.0, + "grad_norm": 2.276389705721688, + "learning_rate": 3.784446765757836e-06, + "loss": 0.7791, + "num_input_tokens_seen": 30288885, + "step": 1453 + }, + { + "epoch": 0.17483316298923826, + "flos": 20401294943400.0, + "grad_norm": 2.6946614681642354, + "learning_rate": 3.7840948512413133e-06, + "loss": 0.769, + "num_input_tokens_seen": 30306190, + "step": 1454 + }, + { + "epoch": 0.17495340587987734, + "flos": 32389275891120.0, + "grad_norm": 3.008872264094264, + "learning_rate": 3.7837426660790196e-06, + "loss": 0.7627, + "num_input_tokens_seen": 30327325, + "step": 1455 + }, + { + "epoch": 0.17507364877051645, + "flos": 15269942145960.0, + "grad_norm": 2.556783580353237, + "learning_rate": 3.783390210324382e-06, + "loss": 0.7978, + "num_input_tokens_seen": 30346770, + "step": 1456 + }, + { + "epoch": 0.17519389166115554, + "flos": 18105823049640.0, + "grad_norm": 2.7728429022611643, + "learning_rate": 3.7830374840308676e-06, + "loss": 0.7006, + "num_input_tokens_seen": 30366645, + "step": 1457 + }, + { + "epoch": 0.17531413455179462, + "flos": 17424328054800.0, + "grad_norm": 2.639567597986382, + "learning_rate": 3.7826844872519842e-06, + "loss": 0.8016, + "num_input_tokens_seen": 30384220, + "step": 1458 + }, + { + "epoch": 0.1754343774424337, + "flos": 17998450430760.0, + "grad_norm": 2.5412663212227207, + "learning_rate": 3.782331220041282e-06, + "loss": 0.7138, + "num_input_tokens_seen": 30404005, + "step": 1459 + }, + { + "epoch": 0.17555462033307281, + "flos": 13219287960600.0, + "grad_norm": 3.5627963294211082, + "learning_rate": 3.7819776824523504e-06, + "loss": 0.82, + "num_input_tokens_seen": 30421590, + "step": 1460 + }, + { + "epoch": 0.1756748632237119, + "flos": 20805873521760.0, + "grad_norm": 2.1521326089580373, + "learning_rate": 3.7816238745388213e-06, + "loss": 0.8297, + "num_input_tokens_seen": 30440855, + "step": 1461 + }, + { + "epoch": 0.17579510611435098, + "flos": 18856354943160.0, + "grad_norm": 2.1410425154889157, + "learning_rate": 3.781269796354367e-06, + "loss": 0.8633, + "num_input_tokens_seen": 30460195, + "step": 1462 + }, + { + "epoch": 0.1759153490049901, + "flos": 13569578690280.0, + "grad_norm": 3.6539198794629675, + "learning_rate": 3.7809154479527006e-06, + "loss": 0.859, + "num_input_tokens_seen": 30479120, + "step": 1463 + }, + { + "epoch": 0.17603559189562917, + "flos": 13591957790640.0, + "grad_norm": 2.3929155135203146, + "learning_rate": 3.780560829387577e-06, + "loss": 0.8147, + "num_input_tokens_seen": 30497340, + "step": 1464 + }, + { + "epoch": 0.17615583478626826, + "flos": 44612897650200.0, + "grad_norm": 0.8549581597781926, + "learning_rate": 3.7802059407127915e-06, + "loss": 0.607, + "num_input_tokens_seen": 30555610, + "step": 1465 + }, + { + "epoch": 0.17627607767690734, + "flos": 17289764021640.0, + "grad_norm": 2.2674599418760844, + "learning_rate": 3.7798507819821797e-06, + "loss": 0.8386, + "num_input_tokens_seen": 30572455, + "step": 1466 + }, + { + "epoch": 0.17639632056754645, + "flos": 12867192613200.0, + "grad_norm": 3.065105653220664, + "learning_rate": 3.7794953532496197e-06, + "loss": 0.7778, + "num_input_tokens_seen": 30588080, + "step": 1467 + }, + { + "epoch": 0.17651656345818553, + "flos": 42694152552720.0, + "grad_norm": 1.5880166692315298, + "learning_rate": 3.7791396545690295e-06, + "loss": 0.6312, + "num_input_tokens_seen": 30649035, + "step": 1468 + }, + { + "epoch": 0.17663680634882462, + "flos": 16782466838520.0, + "grad_norm": 2.689594605823851, + "learning_rate": 3.7787836859943685e-06, + "loss": 0.7851, + "num_input_tokens_seen": 30667480, + "step": 1469 + }, + { + "epoch": 0.17675704923946373, + "flos": 16566771801960.0, + "grad_norm": 4.085489106237174, + "learning_rate": 3.7784274475796363e-06, + "loss": 0.7672, + "num_input_tokens_seen": 30685830, + "step": 1470 + }, + { + "epoch": 0.1768772921301028, + "flos": 19888710546960.0, + "grad_norm": 2.61514995458638, + "learning_rate": 3.7780709393788745e-06, + "loss": 0.7535, + "num_input_tokens_seen": 30706025, + "step": 1471 + }, + { + "epoch": 0.1769975350207419, + "flos": 14001855242280.0, + "grad_norm": 2.2982387516439857, + "learning_rate": 3.777714161446165e-06, + "loss": 0.7429, + "num_input_tokens_seen": 30725450, + "step": 1472 + }, + { + "epoch": 0.177117777911381, + "flos": 26555030457240.0, + "grad_norm": 3.574281852456969, + "learning_rate": 3.7773571138356304e-06, + "loss": 0.6556, + "num_input_tokens_seen": 30745340, + "step": 1473 + }, + { + "epoch": 0.17723802080202009, + "flos": 16161655004280.0, + "grad_norm": 2.3722204708159405, + "learning_rate": 3.776999796601435e-06, + "loss": 0.8781, + "num_input_tokens_seen": 30763820, + "step": 1474 + }, + { + "epoch": 0.17735826369265917, + "flos": 22179338466840.0, + "grad_norm": 2.2289624581134055, + "learning_rate": 3.776642209797783e-06, + "loss": 0.695, + "num_input_tokens_seen": 30785370, + "step": 1475 + }, + { + "epoch": 0.17747850658329825, + "flos": 15648817328160.0, + "grad_norm": 2.7304409827920204, + "learning_rate": 3.7762843534789205e-06, + "loss": 0.7543, + "num_input_tokens_seen": 30803840, + "step": 1476 + }, + { + "epoch": 0.17759874947393736, + "flos": 12382052910720.0, + "grad_norm": 2.397088581700323, + "learning_rate": 3.7759262276991343e-06, + "loss": 0.8521, + "num_input_tokens_seen": 30821170, + "step": 1477 + }, + { + "epoch": 0.17771899236457644, + "flos": 8358013028280.0, + "grad_norm": 2.510092296195242, + "learning_rate": 3.7755678325127506e-06, + "loss": 0.7818, + "num_input_tokens_seen": 30838570, + "step": 1478 + }, + { + "epoch": 0.17783923525521553, + "flos": 13732252276560.0, + "grad_norm": 1.9112115297007253, + "learning_rate": 3.7752091679741393e-06, + "loss": 0.7299, + "num_input_tokens_seen": 30856080, + "step": 1479 + }, + { + "epoch": 0.17795947814585464, + "flos": 22317226795800.0, + "grad_norm": 2.900237067168194, + "learning_rate": 3.774850234137708e-06, + "loss": 0.7563, + "num_input_tokens_seen": 30873095, + "step": 1480 + }, + { + "epoch": 0.17807972103649372, + "flos": 17862651659160.0, + "grad_norm": 2.823287282640976, + "learning_rate": 3.7744910310579076e-06, + "loss": 0.8103, + "num_input_tokens_seen": 30891740, + "step": 1481 + }, + { + "epoch": 0.1781999639271328, + "flos": 14837475634200.0, + "grad_norm": 2.401147460813383, + "learning_rate": 3.774131558789229e-06, + "loss": 0.8442, + "num_input_tokens_seen": 30910790, + "step": 1482 + }, + { + "epoch": 0.1783202068177719, + "flos": 11598314210520.0, + "grad_norm": 3.968234291512282, + "learning_rate": 3.773771817386203e-06, + "loss": 0.6798, + "num_input_tokens_seen": 30927840, + "step": 1483 + }, + { + "epoch": 0.178440449708411, + "flos": 15082103382840.0, + "grad_norm": 1.8548559843438654, + "learning_rate": 3.773411806903403e-06, + "loss": 0.7811, + "num_input_tokens_seen": 30946640, + "step": 1484 + }, + { + "epoch": 0.17856069259905008, + "flos": 15861821268120.0, + "grad_norm": 3.8692093929243407, + "learning_rate": 3.7730515273954415e-06, + "loss": 0.92, + "num_input_tokens_seen": 30964970, + "step": 1485 + }, + { + "epoch": 0.17868093548968916, + "flos": 19478148236160.0, + "grad_norm": 2.0039398012085283, + "learning_rate": 3.772690978916973e-06, + "loss": 0.8226, + "num_input_tokens_seen": 30984445, + "step": 1486 + }, + { + "epoch": 0.17880117838032827, + "flos": 13563341678160.0, + "grad_norm": 2.3213080636909003, + "learning_rate": 3.772330161522693e-06, + "loss": 0.8493, + "num_input_tokens_seen": 31002075, + "step": 1487 + }, + { + "epoch": 0.17892142127096736, + "flos": 19455325896360.0, + "grad_norm": 2.9436751046341825, + "learning_rate": 3.7719690752673365e-06, + "loss": 0.7871, + "num_input_tokens_seen": 31022590, + "step": 1488 + }, + { + "epoch": 0.17904166416160644, + "flos": 17479660682160.0, + "grad_norm": 2.299566858390087, + "learning_rate": 3.7716077202056796e-06, + "loss": 0.7629, + "num_input_tokens_seen": 31040785, + "step": 1489 + }, + { + "epoch": 0.17916190705224552, + "flos": 13973777349120.0, + "grad_norm": 2.284997240060688, + "learning_rate": 3.7712460963925404e-06, + "loss": 0.9187, + "num_input_tokens_seen": 31056445, + "step": 1490 + }, + { + "epoch": 0.17928214994288463, + "flos": 18426927787560.0, + "grad_norm": 1.9300184699642322, + "learning_rate": 3.7708842038827775e-06, + "loss": 0.7387, + "num_input_tokens_seen": 31075125, + "step": 1491 + }, + { + "epoch": 0.17940239283352372, + "flos": 16379566238040.0, + "grad_norm": 2.4573047770729493, + "learning_rate": 3.770522042731288e-06, + "loss": 0.8349, + "num_input_tokens_seen": 31096740, + "step": 1492 + }, + { + "epoch": 0.1795226357241628, + "flos": 16965999847080.0, + "grad_norm": 2.1942510787928184, + "learning_rate": 3.7701596129930122e-06, + "loss": 0.8666, + "num_input_tokens_seen": 31115185, + "step": 1493 + }, + { + "epoch": 0.1796428786148019, + "flos": 16161496704480.0, + "grad_norm": 2.0795176722588966, + "learning_rate": 3.7697969147229315e-06, + "loss": 0.7112, + "num_input_tokens_seen": 31133065, + "step": 1494 + }, + { + "epoch": 0.179763121505441, + "flos": 15594814419120.0, + "grad_norm": 2.2139840316058765, + "learning_rate": 3.7694339479760647e-06, + "loss": 0.8439, + "num_input_tokens_seen": 31151815, + "step": 1495 + }, + { + "epoch": 0.17988336439608008, + "flos": 50261203918560.0, + "grad_norm": 0.7796993889217568, + "learning_rate": 3.769070712807476e-06, + "loss": 0.5996, + "num_input_tokens_seen": 31213565, + "step": 1496 + }, + { + "epoch": 0.18000360728671919, + "flos": 16053997445760.0, + "grad_norm": 2.262782661948253, + "learning_rate": 3.768707209272266e-06, + "loss": 0.7658, + "num_input_tokens_seen": 31233415, + "step": 1497 + }, + { + "epoch": 0.18012385017735827, + "flos": 13863871933440.0, + "grad_norm": 10.165935510911513, + "learning_rate": 3.768343437425579e-06, + "loss": 0.7454, + "num_input_tokens_seen": 31251705, + "step": 1498 + }, + { + "epoch": 0.18024409306799735, + "flos": 14427008303280.0, + "grad_norm": 2.4748612104862433, + "learning_rate": 3.7679793973225987e-06, + "loss": 0.8414, + "num_input_tokens_seen": 31267235, + "step": 1499 + }, + { + "epoch": 0.18036433595863643, + "flos": 49571450165520.0, + "grad_norm": 0.8779912636350491, + "learning_rate": 3.767615089018549e-06, + "loss": 0.6321, + "num_input_tokens_seen": 31329300, + "step": 1500 + }, + { + "epoch": 0.18048457884927555, + "flos": 13267908676440.0, + "grad_norm": 2.252649654451007, + "learning_rate": 3.7672505125686966e-06, + "loss": 0.8419, + "num_input_tokens_seen": 31345385, + "step": 1501 + }, + { + "epoch": 0.18060482173991463, + "flos": 11516201748360.0, + "grad_norm": 3.5475195748539807, + "learning_rate": 3.7668856680283455e-06, + "loss": 0.8426, + "num_input_tokens_seen": 31362130, + "step": 1502 + }, + { + "epoch": 0.1807250646305537, + "flos": 13568977151040.0, + "grad_norm": 2.442249659005324, + "learning_rate": 3.7665205554528437e-06, + "loss": 0.8107, + "num_input_tokens_seen": 31381205, + "step": 1503 + }, + { + "epoch": 0.18084530752119282, + "flos": 16945108764840.0, + "grad_norm": 1.956694735835308, + "learning_rate": 3.7661551748975782e-06, + "loss": 0.732, + "num_input_tokens_seen": 31399100, + "step": 1504 + }, + { + "epoch": 0.1809655504118319, + "flos": 44074704837480.0, + "grad_norm": 0.8381530777497229, + "learning_rate": 3.7657895264179772e-06, + "loss": 0.6093, + "num_input_tokens_seen": 31454795, + "step": 1505 + }, + { + "epoch": 0.181085793302471, + "flos": 32525676201960.0, + "grad_norm": 1.775956301907623, + "learning_rate": 3.765423610069509e-06, + "loss": 0.7337, + "num_input_tokens_seen": 31479905, + "step": 1506 + }, + { + "epoch": 0.18120603619311007, + "flos": 25640463599160.0, + "grad_norm": 2.466094314912539, + "learning_rate": 3.765057425907683e-06, + "loss": 0.7152, + "num_input_tokens_seen": 31501085, + "step": 1507 + }, + { + "epoch": 0.18132627908374918, + "flos": 15729568412040.0, + "grad_norm": 2.239355977664594, + "learning_rate": 3.764690973988048e-06, + "loss": 0.7718, + "num_input_tokens_seen": 31521145, + "step": 1508 + }, + { + "epoch": 0.18144652197438826, + "flos": 21317539779360.0, + "grad_norm": 2.1462859876900353, + "learning_rate": 3.7643242543661967e-06, + "loss": 0.7293, + "num_input_tokens_seen": 31543525, + "step": 1509 + }, + { + "epoch": 0.18156676486502735, + "flos": 50641345499160.0, + "grad_norm": 0.8430602115088457, + "learning_rate": 3.7639572670977573e-06, + "loss": 0.635, + "num_input_tokens_seen": 31598740, + "step": 1510 + }, + { + "epoch": 0.18168700775566646, + "flos": 19404204043680.0, + "grad_norm": 1.6115505735419458, + "learning_rate": 3.7635900122384042e-06, + "loss": 0.7536, + "num_input_tokens_seen": 31621455, + "step": 1511 + }, + { + "epoch": 0.18180725064630554, + "flos": 10917389094960.0, + "grad_norm": 2.663644797057017, + "learning_rate": 3.7632224898438477e-06, + "loss": 0.8496, + "num_input_tokens_seen": 31637650, + "step": 1512 + }, + { + "epoch": 0.18192749353694462, + "flos": 14379400706160.0, + "grad_norm": 1.8367107305185346, + "learning_rate": 3.762854699969842e-06, + "loss": 0.7801, + "num_input_tokens_seen": 31657880, + "step": 1513 + }, + { + "epoch": 0.1820477364275837, + "flos": 15134333334120.0, + "grad_norm": 2.9221471975902866, + "learning_rate": 3.762486642672179e-06, + "loss": 0.7109, + "num_input_tokens_seen": 31674540, + "step": 1514 + }, + { + "epoch": 0.18216797931822282, + "flos": 12489552169440.0, + "grad_norm": 2.529102309118837, + "learning_rate": 3.7621183180066946e-06, + "loss": 0.8379, + "num_input_tokens_seen": 31692220, + "step": 1515 + }, + { + "epoch": 0.1822882222088619, + "flos": 21454699929240.0, + "grad_norm": 1.8959384372068158, + "learning_rate": 3.7617497260292625e-06, + "loss": 0.7221, + "num_input_tokens_seen": 31713995, + "step": 1516 + }, + { + "epoch": 0.18240846509950098, + "flos": 12912995592600.0, + "grad_norm": 3.499106429337783, + "learning_rate": 3.7613808667957967e-06, + "loss": 0.7751, + "num_input_tokens_seen": 31726405, + "step": 1517 + }, + { + "epoch": 0.1825287079901401, + "flos": 10758261424200.0, + "grad_norm": 2.930113250314392, + "learning_rate": 3.7610117403622547e-06, + "loss": 0.9031, + "num_input_tokens_seen": 31742685, + "step": 1518 + }, + { + "epoch": 0.18264895088077918, + "flos": 16055042224440.0, + "grad_norm": 2.209545007016378, + "learning_rate": 3.7606423467846313e-06, + "loss": 0.884, + "num_input_tokens_seen": 31762010, + "step": 1519 + }, + { + "epoch": 0.18276919377141826, + "flos": 15270765304920.0, + "grad_norm": 2.0057459783398457, + "learning_rate": 3.760272686118964e-06, + "loss": 0.7769, + "num_input_tokens_seen": 31779950, + "step": 1520 + }, + { + "epoch": 0.18288943666205737, + "flos": 15702218697960.0, + "grad_norm": 2.4549525554291183, + "learning_rate": 3.7599027584213297e-06, + "loss": 0.902, + "num_input_tokens_seen": 31798550, + "step": 1521 + }, + { + "epoch": 0.18300967955269645, + "flos": 15753150590880.0, + "grad_norm": 3.100025786015723, + "learning_rate": 3.7595325637478465e-06, + "loss": 0.7719, + "num_input_tokens_seen": 31816295, + "step": 1522 + }, + { + "epoch": 0.18312992244333554, + "flos": 21183387325680.0, + "grad_norm": 1.825750324736588, + "learning_rate": 3.7591621021546723e-06, + "loss": 0.8096, + "num_input_tokens_seen": 31838010, + "step": 1523 + }, + { + "epoch": 0.18325016533397462, + "flos": 14703323180520.0, + "grad_norm": 5.070859059644944, + "learning_rate": 3.7587913736980062e-06, + "loss": 0.7982, + "num_input_tokens_seen": 31857370, + "step": 1524 + }, + { + "epoch": 0.18337040822461373, + "flos": 17078026480080.0, + "grad_norm": 2.0581502418580384, + "learning_rate": 3.7584203784340865e-06, + "loss": 0.8336, + "num_input_tokens_seen": 31876260, + "step": 1525 + }, + { + "epoch": 0.1834906511152528, + "flos": 18321898005720.0, + "grad_norm": 3.377956367116018, + "learning_rate": 3.7580491164191938e-06, + "loss": 0.8393, + "num_input_tokens_seen": 31894290, + "step": 1526 + }, + { + "epoch": 0.1836108940058919, + "flos": 49587718893600.0, + "grad_norm": 0.7564304766984115, + "learning_rate": 3.757677587709648e-06, + "loss": 0.6248, + "num_input_tokens_seen": 31957275, + "step": 1527 + }, + { + "epoch": 0.183731136896531, + "flos": 18861452196720.0, + "grad_norm": 2.248955802215193, + "learning_rate": 3.7573057923618095e-06, + "loss": 0.7419, + "num_input_tokens_seen": 31977090, + "step": 1528 + }, + { + "epoch": 0.1838513797871701, + "flos": 14947159430160.0, + "grad_norm": 2.443752778580981, + "learning_rate": 3.7569337304320793e-06, + "loss": 0.7203, + "num_input_tokens_seen": 31996395, + "step": 1529 + }, + { + "epoch": 0.18397162267780917, + "flos": 50559929556120.0, + "grad_norm": 0.8462372254008951, + "learning_rate": 3.756561401976899e-06, + "loss": 0.6525, + "num_input_tokens_seen": 32055820, + "step": 1530 + }, + { + "epoch": 0.18409186556844825, + "flos": 23451604485240.0, + "grad_norm": 2.5414945721340363, + "learning_rate": 3.7561888070527514e-06, + "loss": 0.8141, + "num_input_tokens_seen": 32077580, + "step": 1531 + }, + { + "epoch": 0.18421210845908736, + "flos": 14703703100040.0, + "grad_norm": 2.4840070344023686, + "learning_rate": 3.7558159457161577e-06, + "loss": 0.7775, + "num_input_tokens_seen": 32095265, + "step": 1532 + }, + { + "epoch": 0.18433235134972645, + "flos": 16916207712720.0, + "grad_norm": 2.949790242383356, + "learning_rate": 3.755442818023681e-06, + "loss": 0.7596, + "num_input_tokens_seen": 32114610, + "step": 1533 + }, + { + "epoch": 0.18445259424036553, + "flos": 13349292959520.0, + "grad_norm": 2.814933741113263, + "learning_rate": 3.7550694240319246e-06, + "loss": 0.7429, + "num_input_tokens_seen": 32132205, + "step": 1534 + }, + { + "epoch": 0.18457283713100464, + "flos": 15594339519720.0, + "grad_norm": 3.1843973655027566, + "learning_rate": 3.7546957637975326e-06, + "loss": 0.7365, + "num_input_tokens_seen": 32149335, + "step": 1535 + }, + { + "epoch": 0.18469308002164372, + "flos": 14892491661960.0, + "grad_norm": 2.1745458413750343, + "learning_rate": 3.7543218373771873e-06, + "loss": 0.733, + "num_input_tokens_seen": 32168380, + "step": 1536 + }, + { + "epoch": 0.1848133229122828, + "flos": 19377962428200.0, + "grad_norm": 1.4641170886436723, + "learning_rate": 3.753947644827615e-06, + "loss": 0.7778, + "num_input_tokens_seen": 32191560, + "step": 1537 + }, + { + "epoch": 0.1849335658029219, + "flos": 52028677506720.0, + "grad_norm": 1.0310310940609266, + "learning_rate": 3.753573186205579e-06, + "loss": 0.5929, + "num_input_tokens_seen": 32259400, + "step": 1538 + }, + { + "epoch": 0.185053808693561, + "flos": 12678594011040.0, + "grad_norm": 2.6063903368300654, + "learning_rate": 3.753198461567885e-06, + "loss": 0.7518, + "num_input_tokens_seen": 32276365, + "step": 1539 + }, + { + "epoch": 0.18517405158420008, + "flos": 21264803268720.0, + "grad_norm": 2.0655453480470642, + "learning_rate": 3.7528234709713783e-06, + "loss": 0.9024, + "num_input_tokens_seen": 32298830, + "step": 1540 + }, + { + "epoch": 0.18529429447483917, + "flos": 19642879719840.0, + "grad_norm": 2.9730766490367118, + "learning_rate": 3.7524482144729447e-06, + "loss": 0.8238, + "num_input_tokens_seen": 32318005, + "step": 1541 + }, + { + "epoch": 0.18541453736547828, + "flos": 9863319249960.0, + "grad_norm": 2.9083219956083197, + "learning_rate": 3.7520726921295106e-06, + "loss": 0.8256, + "num_input_tokens_seen": 32334445, + "step": 1542 + }, + { + "epoch": 0.18553478025611736, + "flos": 17593713552600.0, + "grad_norm": 2.5709308471816175, + "learning_rate": 3.751696903998042e-06, + "loss": 0.7205, + "num_input_tokens_seen": 32352800, + "step": 1543 + }, + { + "epoch": 0.18565502314675644, + "flos": 18969901254240.0, + "grad_norm": 2.2827543759110323, + "learning_rate": 3.7513208501355456e-06, + "loss": 0.6785, + "num_input_tokens_seen": 32373625, + "step": 1544 + }, + { + "epoch": 0.18577526603739553, + "flos": 14325967676400.0, + "grad_norm": 2.462382446426526, + "learning_rate": 3.750944530599069e-06, + "loss": 0.8138, + "num_input_tokens_seen": 32392915, + "step": 1545 + }, + { + "epoch": 0.18589550892803464, + "flos": 13485281690880.0, + "grad_norm": 3.8384450885660737, + "learning_rate": 3.7505679454456992e-06, + "loss": 0.7847, + "num_input_tokens_seen": 32409245, + "step": 1546 + }, + { + "epoch": 0.18601575181867372, + "flos": 17241681525120.0, + "grad_norm": 2.1127549959935497, + "learning_rate": 3.750191094732564e-06, + "loss": 0.6747, + "num_input_tokens_seen": 32429830, + "step": 1547 + }, + { + "epoch": 0.1861359947093128, + "flos": 19321553362200.0, + "grad_norm": 2.0345952584237943, + "learning_rate": 3.7498139785168313e-06, + "loss": 0.7379, + "num_input_tokens_seen": 32450155, + "step": 1548 + }, + { + "epoch": 0.1862562375999519, + "flos": 17079799437840.0, + "grad_norm": 2.344536407780273, + "learning_rate": 3.749436596855709e-06, + "loss": 0.7623, + "num_input_tokens_seen": 32469175, + "step": 1549 + }, + { + "epoch": 0.186376480490591, + "flos": 12133087747560.0, + "grad_norm": 2.3080622388754923, + "learning_rate": 3.749058949806446e-06, + "loss": 0.8996, + "num_input_tokens_seen": 32485620, + "step": 1550 + }, + { + "epoch": 0.18649672338123008, + "flos": 15701142259320.0, + "grad_norm": 2.1953829825276325, + "learning_rate": 3.748681037426331e-06, + "loss": 0.8247, + "num_input_tokens_seen": 32504550, + "step": 1551 + }, + { + "epoch": 0.1866169662718692, + "flos": 8925423492720.0, + "grad_norm": 2.3203093146613765, + "learning_rate": 3.7483028597726936e-06, + "loss": 0.8995, + "num_input_tokens_seen": 32521040, + "step": 1552 + }, + { + "epoch": 0.18673720916250827, + "flos": 17264123945400.0, + "grad_norm": 2.2759683162180053, + "learning_rate": 3.7479244169029017e-06, + "loss": 0.6148, + "num_input_tokens_seen": 32540550, + "step": 1553 + }, + { + "epoch": 0.18685745205314735, + "flos": 14405958921240.0, + "grad_norm": 3.3836854861041785, + "learning_rate": 3.7475457088743658e-06, + "loss": 0.6926, + "num_input_tokens_seen": 32557520, + "step": 1554 + }, + { + "epoch": 0.18697769494378644, + "flos": 25126517824440.0, + "grad_norm": 2.7711461017353605, + "learning_rate": 3.7471667357445348e-06, + "loss": 0.722, + "num_input_tokens_seen": 32577070, + "step": 1555 + }, + { + "epoch": 0.18709793783442555, + "flos": 25154152478160.0, + "grad_norm": 2.6602295628778867, + "learning_rate": 3.7467874975709e-06, + "loss": 0.7088, + "num_input_tokens_seen": 32597595, + "step": 1556 + }, + { + "epoch": 0.18721818072506463, + "flos": 29959398264000.0, + "grad_norm": 3.851985389753068, + "learning_rate": 3.7464079944109904e-06, + "loss": 0.7764, + "num_input_tokens_seen": 32619175, + "step": 1557 + }, + { + "epoch": 0.18733842361570371, + "flos": 16210212400200.0, + "grad_norm": 2.561603389281938, + "learning_rate": 3.746028226322376e-06, + "loss": 0.7646, + "num_input_tokens_seen": 32634775, + "step": 1558 + }, + { + "epoch": 0.18745866650634282, + "flos": 13811262062640.0, + "grad_norm": 2.4431703587398386, + "learning_rate": 3.745648193362669e-06, + "loss": 0.741, + "num_input_tokens_seen": 32653850, + "step": 1559 + }, + { + "epoch": 0.1875789093969819, + "flos": 14106821704200.0, + "grad_norm": 3.655578682068471, + "learning_rate": 3.745267895589518e-06, + "loss": 0.7163, + "num_input_tokens_seen": 32672110, + "step": 1560 + }, + { + "epoch": 0.187699152287621, + "flos": 12408294526200.0, + "grad_norm": 2.8135243563712122, + "learning_rate": 3.7448873330606154e-06, + "loss": 0.7914, + "num_input_tokens_seen": 32689600, + "step": 1561 + }, + { + "epoch": 0.18781939517826007, + "flos": 16351013445480.0, + "grad_norm": 5.086520628893025, + "learning_rate": 3.7445065058336914e-06, + "loss": 0.8512, + "num_input_tokens_seen": 32708190, + "step": 1562 + }, + { + "epoch": 0.18793963806889918, + "flos": 10626673427280.0, + "grad_norm": 2.161912862639698, + "learning_rate": 3.7441254139665176e-06, + "loss": 0.845, + "num_input_tokens_seen": 32724095, + "step": 1563 + }, + { + "epoch": 0.18805988095953827, + "flos": 12732565260120.0, + "grad_norm": 2.886990298247941, + "learning_rate": 3.743744057516905e-06, + "loss": 0.81, + "num_input_tokens_seen": 32741875, + "step": 1564 + }, + { + "epoch": 0.18818012385017735, + "flos": 10945910227560.0, + "grad_norm": 3.234271503948746, + "learning_rate": 3.743362436542706e-06, + "loss": 0.8601, + "num_input_tokens_seen": 32756285, + "step": 1565 + }, + { + "epoch": 0.18830036674081646, + "flos": 35003984585160.0, + "grad_norm": 2.2950509022904053, + "learning_rate": 3.7429805511018115e-06, + "loss": 0.7534, + "num_input_tokens_seen": 32777665, + "step": 1566 + }, + { + "epoch": 0.18842060963145554, + "flos": 22048193709360.0, + "grad_norm": 3.1838357845112513, + "learning_rate": 3.7425984012521524e-06, + "loss": 0.7543, + "num_input_tokens_seen": 32797585, + "step": 1567 + }, + { + "epoch": 0.18854085252209463, + "flos": 51857709010920.0, + "grad_norm": 0.7328166660605235, + "learning_rate": 3.7422159870517025e-06, + "loss": 0.6186, + "num_input_tokens_seen": 32862560, + "step": 1568 + }, + { + "epoch": 0.1886610954127337, + "flos": 15568857743280.0, + "grad_norm": 1.849049468324148, + "learning_rate": 3.7418333085584717e-06, + "loss": 0.7734, + "num_input_tokens_seen": 32883465, + "step": 1569 + }, + { + "epoch": 0.18878133830337282, + "flos": 12590656116240.0, + "grad_norm": 2.3669765869452473, + "learning_rate": 3.7414503658305128e-06, + "loss": 0.8951, + "num_input_tokens_seen": 32900420, + "step": 1570 + }, + { + "epoch": 0.1889015811940119, + "flos": 18888453651240.0, + "grad_norm": 3.0421539508554734, + "learning_rate": 3.7410671589259185e-06, + "loss": 0.7639, + "num_input_tokens_seen": 32918740, + "step": 1571 + }, + { + "epoch": 0.18902182408465099, + "flos": 15378327883560.0, + "grad_norm": 3.5593049294504886, + "learning_rate": 3.7406836879028205e-06, + "loss": 0.7713, + "num_input_tokens_seen": 32938685, + "step": 1572 + }, + { + "epoch": 0.1891420669752901, + "flos": 16295870777880.0, + "grad_norm": 2.1718011198582454, + "learning_rate": 3.7402999528193907e-06, + "loss": 0.7505, + "num_input_tokens_seen": 32957905, + "step": 1573 + }, + { + "epoch": 0.18926230986592918, + "flos": 16107810395040.0, + "grad_norm": 3.3145601595858345, + "learning_rate": 3.739915953733842e-06, + "loss": 0.8162, + "num_input_tokens_seen": 32975670, + "step": 1574 + }, + { + "epoch": 0.18938255275656826, + "flos": 17917224447480.0, + "grad_norm": 2.0472021642237466, + "learning_rate": 3.7395316907044264e-06, + "loss": 0.8006, + "num_input_tokens_seen": 32996175, + "step": 1575 + }, + { + "epoch": 0.18950279564720737, + "flos": 17890982832000.0, + "grad_norm": 3.7299386331863995, + "learning_rate": 3.7391471637894364e-06, + "loss": 0.7801, + "num_input_tokens_seen": 33018160, + "step": 1576 + }, + { + "epoch": 0.18962303853784646, + "flos": 14434226774160.0, + "grad_norm": 2.2589437957026632, + "learning_rate": 3.738762373047205e-06, + "loss": 0.8372, + "num_input_tokens_seen": 33037800, + "step": 1577 + }, + { + "epoch": 0.18974328142848554, + "flos": 15378201243720.0, + "grad_norm": 2.1518779049736576, + "learning_rate": 3.738377318536103e-06, + "loss": 0.8269, + "num_input_tokens_seen": 33057405, + "step": 1578 + }, + { + "epoch": 0.18986352431912462, + "flos": 9408505297800.0, + "grad_norm": 2.9294000510025047, + "learning_rate": 3.7379920003145447e-06, + "loss": 0.6986, + "num_input_tokens_seen": 33071400, + "step": 1579 + }, + { + "epoch": 0.18998376720976373, + "flos": 17402835433320.0, + "grad_norm": 2.2072627393463136, + "learning_rate": 3.7376064184409817e-06, + "loss": 0.8288, + "num_input_tokens_seen": 33090700, + "step": 1580 + }, + { + "epoch": 0.19010401010040281, + "flos": 16809499953000.0, + "grad_norm": 1.5649471955333634, + "learning_rate": 3.7372205729739063e-06, + "loss": 0.8557, + "num_input_tokens_seen": 33112235, + "step": 1581 + }, + { + "epoch": 0.1902242529910419, + "flos": 13974853787760.0, + "grad_norm": 3.299751509356202, + "learning_rate": 3.7368344639718514e-06, + "loss": 0.7041, + "num_input_tokens_seen": 33129890, + "step": 1582 + }, + { + "epoch": 0.190344495881681, + "flos": 18672980234400.0, + "grad_norm": 1.9675596362410257, + "learning_rate": 3.7364480914933895e-06, + "loss": 0.7967, + "num_input_tokens_seen": 33149850, + "step": 1583 + }, + { + "epoch": 0.1904647387723201, + "flos": 19641011782200.0, + "grad_norm": 2.4352645404468913, + "learning_rate": 3.7360614555971325e-06, + "loss": 0.8024, + "num_input_tokens_seen": 33169225, + "step": 1584 + }, + { + "epoch": 0.19058498166295917, + "flos": 17565667319400.0, + "grad_norm": 2.81173356527425, + "learning_rate": 3.735674556341733e-06, + "loss": 0.828, + "num_input_tokens_seen": 33188560, + "step": 1585 + }, + { + "epoch": 0.19070522455359826, + "flos": 20752092232440.0, + "grad_norm": 2.1771307275410603, + "learning_rate": 3.7352873937858835e-06, + "loss": 0.8186, + "num_input_tokens_seen": 33209815, + "step": 1586 + }, + { + "epoch": 0.19082546744423737, + "flos": 18802573653840.0, + "grad_norm": 2.2489356596526555, + "learning_rate": 3.734899967988316e-06, + "loss": 0.701, + "num_input_tokens_seen": 33227715, + "step": 1587 + }, + { + "epoch": 0.19094571033487645, + "flos": 14406940380000.0, + "grad_norm": 2.083328043613076, + "learning_rate": 3.7345122790078026e-06, + "loss": 0.8251, + "num_input_tokens_seen": 33245000, + "step": 1588 + }, + { + "epoch": 0.19106595322551553, + "flos": 15810667755480.0, + "grad_norm": 3.737180669404936, + "learning_rate": 3.7341243269031556e-06, + "loss": 0.9121, + "num_input_tokens_seen": 33263710, + "step": 1589 + }, + { + "epoch": 0.19118619611615464, + "flos": 21938604893280.0, + "grad_norm": 2.385678927956598, + "learning_rate": 3.7337361117332275e-06, + "loss": 0.7634, + "num_input_tokens_seen": 33285170, + "step": 1590 + }, + { + "epoch": 0.19130643900679373, + "flos": 12596481548880.0, + "grad_norm": 3.2887567300992253, + "learning_rate": 3.7333476335569087e-06, + "loss": 0.7484, + "num_input_tokens_seen": 33302890, + "step": 1591 + }, + { + "epoch": 0.1914266818974328, + "flos": 18186194214000.0, + "grad_norm": 4.093005163559032, + "learning_rate": 3.7329588924331325e-06, + "loss": 0.6466, + "num_input_tokens_seen": 33323815, + "step": 1592 + }, + { + "epoch": 0.1915469247880719, + "flos": 13839181656000.0, + "grad_norm": 2.3827475617111804, + "learning_rate": 3.732569888420871e-06, + "loss": 0.8008, + "num_input_tokens_seen": 33343070, + "step": 1593 + }, + { + "epoch": 0.191667167678711, + "flos": 15378992742720.0, + "grad_norm": 2.349277438010123, + "learning_rate": 3.732180621579134e-06, + "loss": 0.808, + "num_input_tokens_seen": 33362005, + "step": 1594 + }, + { + "epoch": 0.1917874105693501, + "flos": 25150923162240.0, + "grad_norm": 2.041027890401244, + "learning_rate": 3.7317910919669745e-06, + "loss": 0.7907, + "num_input_tokens_seen": 33382920, + "step": 1595 + }, + { + "epoch": 0.19190765345998917, + "flos": 17242251404400.0, + "grad_norm": 2.8288382795164724, + "learning_rate": 3.7314012996434826e-06, + "loss": 0.7465, + "num_input_tokens_seen": 33401370, + "step": 1596 + }, + { + "epoch": 0.19202789635062828, + "flos": 14511336962640.0, + "grad_norm": 1.975214683357007, + "learning_rate": 3.7310112446677907e-06, + "loss": 0.7835, + "num_input_tokens_seen": 33419000, + "step": 1597 + }, + { + "epoch": 0.19214813924126736, + "flos": 15297513479760.0, + "grad_norm": 2.3436873364018815, + "learning_rate": 3.7306209270990695e-06, + "loss": 0.6722, + "num_input_tokens_seen": 33436725, + "step": 1598 + }, + { + "epoch": 0.19226838213190645, + "flos": 19321141782720.0, + "grad_norm": 2.4941743685621174, + "learning_rate": 3.7302303469965292e-06, + "loss": 0.8471, + "num_input_tokens_seen": 33455985, + "step": 1599 + }, + { + "epoch": 0.19238862502254553, + "flos": 15243985470120.0, + "grad_norm": 2.0760202863874397, + "learning_rate": 3.7298395044194206e-06, + "loss": 0.6906, + "num_input_tokens_seen": 33474515, + "step": 1600 + }, + { + "epoch": 0.19250886791318464, + "flos": 15891925398720.0, + "grad_norm": 1.9697278939556109, + "learning_rate": 3.7294483994270356e-06, + "loss": 0.9089, + "num_input_tokens_seen": 33492560, + "step": 1601 + }, + { + "epoch": 0.19262911080382372, + "flos": 17106927532200.0, + "grad_norm": 2.4420337084319037, + "learning_rate": 3.7290570320787033e-06, + "loss": 0.7485, + "num_input_tokens_seen": 33511860, + "step": 1602 + }, + { + "epoch": 0.1927493536944628, + "flos": 16052636067480.0, + "grad_norm": 2.823940743551394, + "learning_rate": 3.728665402433793e-06, + "loss": 0.6967, + "num_input_tokens_seen": 33530150, + "step": 1603 + }, + { + "epoch": 0.19286959658510192, + "flos": 12058098776400.0, + "grad_norm": 3.2531587948908043, + "learning_rate": 3.7282735105517164e-06, + "loss": 0.8396, + "num_input_tokens_seen": 33547995, + "step": 1604 + }, + { + "epoch": 0.192989839475741, + "flos": 15811237634760.0, + "grad_norm": 2.5360519637190215, + "learning_rate": 3.727881356491922e-06, + "loss": 0.6678, + "num_input_tokens_seen": 33566125, + "step": 1605 + }, + { + "epoch": 0.19311008236638008, + "flos": 14082796285920.0, + "grad_norm": 2.080980332691794, + "learning_rate": 3.7274889403139002e-06, + "loss": 0.7486, + "num_input_tokens_seen": 33583470, + "step": 1606 + }, + { + "epoch": 0.1932303252570192, + "flos": 21017009524080.0, + "grad_norm": 3.065557549499672, + "learning_rate": 3.727096262077179e-06, + "loss": 0.7737, + "num_input_tokens_seen": 33602185, + "step": 1607 + }, + { + "epoch": 0.19335056814765827, + "flos": 13407411663360.0, + "grad_norm": 1.8540197878607707, + "learning_rate": 3.7267033218413285e-06, + "loss": 0.8356, + "num_input_tokens_seen": 33619700, + "step": 1608 + }, + { + "epoch": 0.19347081103829736, + "flos": 9627049730760.0, + "grad_norm": 2.4692630476159345, + "learning_rate": 3.726310119665957e-06, + "loss": 0.7961, + "num_input_tokens_seen": 33635755, + "step": 1609 + }, + { + "epoch": 0.19359105392893644, + "flos": 14836209235800.0, + "grad_norm": 1.9849328886357078, + "learning_rate": 3.725916655610713e-06, + "loss": 0.8369, + "num_input_tokens_seen": 33654805, + "step": 1610 + }, + { + "epoch": 0.19371129681957555, + "flos": 14973021126120.0, + "grad_norm": 3.031917695118577, + "learning_rate": 3.725522929735284e-06, + "loss": 0.7288, + "num_input_tokens_seen": 33671460, + "step": 1611 + }, + { + "epoch": 0.19383153971021463, + "flos": 22344798129600.0, + "grad_norm": 3.1635643837751313, + "learning_rate": 3.725128942099399e-06, + "loss": 0.7188, + "num_input_tokens_seen": 33691580, + "step": 1612 + }, + { + "epoch": 0.19395178260085372, + "flos": 17996297553480.0, + "grad_norm": 2.049339616056469, + "learning_rate": 3.7247346927628245e-06, + "loss": 0.7881, + "num_input_tokens_seen": 33711235, + "step": 1613 + }, + { + "epoch": 0.19407202549149283, + "flos": 21238624973160.0, + "grad_norm": 2.4972160096940486, + "learning_rate": 3.7243401817853694e-06, + "loss": 0.768, + "num_input_tokens_seen": 33731645, + "step": 1614 + }, + { + "epoch": 0.1941922683821319, + "flos": 13137713717760.0, + "grad_norm": 2.593866736301408, + "learning_rate": 3.723945409226879e-06, + "loss": 0.709, + "num_input_tokens_seen": 33749855, + "step": 1615 + }, + { + "epoch": 0.194312511272771, + "flos": 7006610583960.0, + "grad_norm": 2.58408282287777, + "learning_rate": 3.723550375147241e-06, + "loss": 0.7969, + "num_input_tokens_seen": 33764350, + "step": 1616 + }, + { + "epoch": 0.19443275416341008, + "flos": 19806946344360.0, + "grad_norm": 1.9817032136212605, + "learning_rate": 3.7231550796063816e-06, + "loss": 0.7798, + "num_input_tokens_seen": 33784080, + "step": 1617 + }, + { + "epoch": 0.1945529970540492, + "flos": 11540005546920.0, + "grad_norm": 2.1278820689012847, + "learning_rate": 3.722759522664266e-06, + "loss": 0.6271, + "num_input_tokens_seen": 33801100, + "step": 1618 + }, + { + "epoch": 0.19467323994468827, + "flos": 14107138303800.0, + "grad_norm": 2.05953025869362, + "learning_rate": 3.7223637043809016e-06, + "loss": 0.7909, + "num_input_tokens_seen": 33819800, + "step": 1619 + }, + { + "epoch": 0.19479348283532735, + "flos": 17673261558000.0, + "grad_norm": 1.8534693174536738, + "learning_rate": 3.7219676248163322e-06, + "loss": 0.8457, + "num_input_tokens_seen": 33836685, + "step": 1620 + }, + { + "epoch": 0.19491372572596646, + "flos": 18888833570760.0, + "grad_norm": 2.0823467080069573, + "learning_rate": 3.721571284030643e-06, + "loss": 0.905, + "num_input_tokens_seen": 33856215, + "step": 1621 + }, + { + "epoch": 0.19503396861660555, + "flos": 14351797712400.0, + "grad_norm": 2.434228707105915, + "learning_rate": 3.7211746820839587e-06, + "loss": 0.7598, + "num_input_tokens_seen": 33873030, + "step": 1622 + }, + { + "epoch": 0.19515421150724463, + "flos": 15379056062640.0, + "grad_norm": 1.9912378777986885, + "learning_rate": 3.7207778190364437e-06, + "loss": 0.808, + "num_input_tokens_seen": 33891175, + "step": 1623 + }, + { + "epoch": 0.1952744543978837, + "flos": 24206378813400.0, + "grad_norm": 1.9163281791450928, + "learning_rate": 3.720380694948302e-06, + "loss": 0.7324, + "num_input_tokens_seen": 33913780, + "step": 1624 + }, + { + "epoch": 0.19539469728852282, + "flos": 47214693571920.0, + "grad_norm": 0.9824684389883707, + "learning_rate": 3.719983309879777e-06, + "loss": 0.7362, + "num_input_tokens_seen": 33973280, + "step": 1625 + }, + { + "epoch": 0.1955149401791619, + "flos": 9708560653680.0, + "grad_norm": 1.8677773282238834, + "learning_rate": 3.719585663891151e-06, + "loss": 0.7662, + "num_input_tokens_seen": 33990535, + "step": 1626 + }, + { + "epoch": 0.195635183069801, + "flos": 13673057134080.0, + "grad_norm": 3.0377153485101624, + "learning_rate": 3.719187757042747e-06, + "loss": 0.7648, + "num_input_tokens_seen": 34008075, + "step": 1627 + }, + { + "epoch": 0.1957554259604401, + "flos": 51116801745240.0, + "grad_norm": 0.7753556865824806, + "learning_rate": 3.7187895893949275e-06, + "loss": 0.5895, + "num_input_tokens_seen": 34074265, + "step": 1628 + }, + { + "epoch": 0.19587566885107918, + "flos": 15405835897440.0, + "grad_norm": 2.242716550715999, + "learning_rate": 3.7183911610080937e-06, + "loss": 0.719, + "num_input_tokens_seen": 34090850, + "step": 1629 + }, + { + "epoch": 0.19599591174171827, + "flos": 16237910373840.0, + "grad_norm": 3.020750127939949, + "learning_rate": 3.7179924719426872e-06, + "loss": 0.723, + "num_input_tokens_seen": 34108465, + "step": 1630 + }, + { + "epoch": 0.19611615463235738, + "flos": 17403247012800.0, + "grad_norm": 2.572248106877278, + "learning_rate": 3.7175935222591885e-06, + "loss": 0.7331, + "num_input_tokens_seen": 34127485, + "step": 1631 + }, + { + "epoch": 0.19623639752299646, + "flos": 20992065966960.0, + "grad_norm": 2.084483236222125, + "learning_rate": 3.717194312018118e-06, + "loss": 0.7422, + "num_input_tokens_seen": 34146190, + "step": 1632 + }, + { + "epoch": 0.19635664041363554, + "flos": 15378771123000.0, + "grad_norm": 2.5862721575252916, + "learning_rate": 3.716794841280036e-06, + "loss": 0.7383, + "num_input_tokens_seen": 34164615, + "step": 1633 + }, + { + "epoch": 0.19647688330427462, + "flos": 13785400366680.0, + "grad_norm": 2.1789905158401535, + "learning_rate": 3.7163951101055407e-06, + "loss": 0.7532, + "num_input_tokens_seen": 34182395, + "step": 1634 + }, + { + "epoch": 0.19659712619491373, + "flos": 17754107621760.0, + "grad_norm": 2.1525038090228006, + "learning_rate": 3.715995118555273e-06, + "loss": 0.7698, + "num_input_tokens_seen": 34202090, + "step": 1635 + }, + { + "epoch": 0.19671736908555282, + "flos": 18105664749840.0, + "grad_norm": 4.7009414216988645, + "learning_rate": 3.71559486668991e-06, + "loss": 0.8244, + "num_input_tokens_seen": 34220670, + "step": 1636 + }, + { + "epoch": 0.1968376119761919, + "flos": 17458073080800.0, + "grad_norm": 2.543286281748753, + "learning_rate": 3.715194354570169e-06, + "loss": 0.7542, + "num_input_tokens_seen": 34240395, + "step": 1637 + }, + { + "epoch": 0.196957854866831, + "flos": 13218686421360.0, + "grad_norm": 3.163678331953761, + "learning_rate": 3.714793582256809e-06, + "loss": 0.8148, + "num_input_tokens_seen": 34257180, + "step": 1638 + }, + { + "epoch": 0.1970780977574701, + "flos": 15837954149640.0, + "grad_norm": 2.5675276769466784, + "learning_rate": 3.7143925498106253e-06, + "loss": 0.8258, + "num_input_tokens_seen": 34275440, + "step": 1639 + }, + { + "epoch": 0.19719834064810918, + "flos": 15215021098080.0, + "grad_norm": 2.098559421340616, + "learning_rate": 3.7139912572924558e-06, + "loss": 0.7798, + "num_input_tokens_seen": 34294190, + "step": 1640 + }, + { + "epoch": 0.19731858353874826, + "flos": 17155643227920.0, + "grad_norm": 2.993302817185574, + "learning_rate": 3.7135897047631744e-06, + "loss": 0.7979, + "num_input_tokens_seen": 34311795, + "step": 1641 + }, + { + "epoch": 0.19743882642938737, + "flos": 17566047238920.0, + "grad_norm": 2.5059254043692487, + "learning_rate": 3.713187892283698e-06, + "loss": 0.7281, + "num_input_tokens_seen": 34331125, + "step": 1642 + }, + { + "epoch": 0.19755906932002645, + "flos": 10916217676440.0, + "grad_norm": 3.8149637485046677, + "learning_rate": 3.71278581991498e-06, + "loss": 0.8446, + "num_input_tokens_seen": 34346705, + "step": 1643 + }, + { + "epoch": 0.19767931221066554, + "flos": 14239549459680.0, + "grad_norm": 3.421881549428307, + "learning_rate": 3.712383487718015e-06, + "loss": 0.7721, + "num_input_tokens_seen": 34364665, + "step": 1644 + }, + { + "epoch": 0.19779955510130465, + "flos": 18860724017640.0, + "grad_norm": 2.32043775226833, + "learning_rate": 3.7119808957538365e-06, + "loss": 0.8484, + "num_input_tokens_seen": 34383380, + "step": 1645 + }, + { + "epoch": 0.19791979799194373, + "flos": 15190109200920.0, + "grad_norm": 2.4413276354530353, + "learning_rate": 3.711578044083517e-06, + "loss": 0.7785, + "num_input_tokens_seen": 34399900, + "step": 1646 + }, + { + "epoch": 0.1980400408825828, + "flos": 18750122082840.0, + "grad_norm": 2.379388769664556, + "learning_rate": 3.7111749327681698e-06, + "loss": 0.7395, + "num_input_tokens_seen": 34419655, + "step": 1647 + }, + { + "epoch": 0.1981602837732219, + "flos": 17214996670200.0, + "grad_norm": 5.719382996260187, + "learning_rate": 3.7107715618689455e-06, + "loss": 0.8403, + "num_input_tokens_seen": 34438350, + "step": 1648 + }, + { + "epoch": 0.198280526663861, + "flos": 16971888599640.0, + "grad_norm": 1.6092988523968081, + "learning_rate": 3.710367931447035e-06, + "loss": 0.8141, + "num_input_tokens_seen": 34459850, + "step": 1649 + }, + { + "epoch": 0.1984007695545001, + "flos": 15864828964320.0, + "grad_norm": 4.2258437809520055, + "learning_rate": 3.70996404156367e-06, + "loss": 0.8436, + "num_input_tokens_seen": 34479205, + "step": 1650 + }, + { + "epoch": 0.19852101244513917, + "flos": 26502832165920.0, + "grad_norm": 2.0625450822931604, + "learning_rate": 3.7095598922801187e-06, + "loss": 0.7089, + "num_input_tokens_seen": 34501000, + "step": 1651 + }, + { + "epoch": 0.19864125533577828, + "flos": 16912155237840.0, + "grad_norm": 4.844274873905802, + "learning_rate": 3.7091554836576914e-06, + "loss": 0.7466, + "num_input_tokens_seen": 34517395, + "step": 1652 + }, + { + "epoch": 0.19876149822641737, + "flos": 18025135285680.0, + "grad_norm": 2.024383088319784, + "learning_rate": 3.708750815757736e-06, + "loss": 0.8086, + "num_input_tokens_seen": 34537885, + "step": 1653 + }, + { + "epoch": 0.19888174111705645, + "flos": 23640741306720.0, + "grad_norm": 3.2484033036920614, + "learning_rate": 3.7083458886416407e-06, + "loss": 0.6966, + "num_input_tokens_seen": 34556800, + "step": 1654 + }, + { + "epoch": 0.19900198400769553, + "flos": 18025103625720.0, + "grad_norm": 3.3298528779668044, + "learning_rate": 3.707940702370832e-06, + "loss": 0.8584, + "num_input_tokens_seen": 34577365, + "step": 1655 + }, + { + "epoch": 0.19912222689833464, + "flos": 50080425326520.0, + "grad_norm": 0.7738293127352185, + "learning_rate": 3.707535257006777e-06, + "loss": 0.5893, + "num_input_tokens_seen": 34642710, + "step": 1656 + }, + { + "epoch": 0.19924246978897373, + "flos": 11463560217600.0, + "grad_norm": 3.122107884118626, + "learning_rate": 3.707129552610981e-06, + "loss": 0.8712, + "num_input_tokens_seen": 34661080, + "step": 1657 + }, + { + "epoch": 0.1993627126796128, + "flos": 12813601283640.0, + "grad_norm": 3.8134511792929233, + "learning_rate": 3.70672358924499e-06, + "loss": 0.7223, + "num_input_tokens_seen": 34680040, + "step": 1658 + }, + { + "epoch": 0.19948295557025192, + "flos": 30040244327760.0, + "grad_norm": 2.263277297564615, + "learning_rate": 3.706317366970386e-06, + "loss": 0.769, + "num_input_tokens_seen": 34700760, + "step": 1659 + }, + { + "epoch": 0.199603198460891, + "flos": 18375805934880.0, + "grad_norm": 2.24326226929124, + "learning_rate": 3.705910885848795e-06, + "loss": 0.8272, + "num_input_tokens_seen": 34718855, + "step": 1660 + }, + { + "epoch": 0.19972344135153008, + "flos": 14676764965440.0, + "grad_norm": 2.8036129734060795, + "learning_rate": 3.705504145941879e-06, + "loss": 0.8245, + "num_input_tokens_seen": 34736745, + "step": 1661 + }, + { + "epoch": 0.1998436842421692, + "flos": 17372351383200.0, + "grad_norm": 2.330910162406776, + "learning_rate": 3.7050971473113403e-06, + "loss": 0.7682, + "num_input_tokens_seen": 34756240, + "step": 1662 + }, + { + "epoch": 0.19996392713280828, + "flos": 26531479938360.0, + "grad_norm": 1.8963235144550679, + "learning_rate": 3.7046898900189196e-06, + "loss": 0.7924, + "num_input_tokens_seen": 34780295, + "step": 1663 + }, + { + "epoch": 0.20008417002344736, + "flos": 17321134550640.0, + "grad_norm": 1.9098601004409264, + "learning_rate": 3.704282374126398e-06, + "loss": 0.8229, + "num_input_tokens_seen": 34799695, + "step": 1664 + }, + { + "epoch": 0.20020441291408644, + "flos": 15999994536720.0, + "grad_norm": 2.317849852852564, + "learning_rate": 3.7038745996955954e-06, + "loss": 0.8627, + "num_input_tokens_seen": 34818760, + "step": 1665 + }, + { + "epoch": 0.20032465580472555, + "flos": 16966886325960.0, + "grad_norm": 2.897674706946139, + "learning_rate": 3.703466566788371e-06, + "loss": 0.6891, + "num_input_tokens_seen": 34837610, + "step": 1666 + }, + { + "epoch": 0.20044489869536464, + "flos": 17480452181160.0, + "grad_norm": 4.847769171101208, + "learning_rate": 3.703058275466622e-06, + "loss": 0.7304, + "num_input_tokens_seen": 34856565, + "step": 1667 + }, + { + "epoch": 0.20056514158600372, + "flos": 16054155745560.0, + "grad_norm": 1.9493142756292838, + "learning_rate": 3.7026497257922877e-06, + "loss": 0.7573, + "num_input_tokens_seen": 34876595, + "step": 1668 + }, + { + "epoch": 0.20068538447664283, + "flos": 17485422794880.0, + "grad_norm": 2.2252664642222046, + "learning_rate": 3.7022409178273436e-06, + "loss": 0.8229, + "num_input_tokens_seen": 34897295, + "step": 1669 + }, + { + "epoch": 0.2008056273672819, + "flos": 13462174411440.0, + "grad_norm": 1.9307582216741002, + "learning_rate": 3.7018318516338054e-06, + "loss": 0.7643, + "num_input_tokens_seen": 34916175, + "step": 1670 + }, + { + "epoch": 0.200925870257921, + "flos": 17322780868560.0, + "grad_norm": 2.7129802647852523, + "learning_rate": 3.7014225272737284e-06, + "loss": 0.7876, + "num_input_tokens_seen": 34935120, + "step": 1671 + }, + { + "epoch": 0.20104611314856008, + "flos": 11814990705840.0, + "grad_norm": 2.5839232929878957, + "learning_rate": 3.701012944809207e-06, + "loss": 0.7183, + "num_input_tokens_seen": 34951955, + "step": 1672 + }, + { + "epoch": 0.2011663560391992, + "flos": 15648627368400.0, + "grad_norm": 2.176827033864604, + "learning_rate": 3.700603104302374e-06, + "loss": 0.7728, + "num_input_tokens_seen": 34971485, + "step": 1673 + }, + { + "epoch": 0.20128659892983827, + "flos": 45872915755440.0, + "grad_norm": 0.8650061174959163, + "learning_rate": 3.7001930058154027e-06, + "loss": 0.5769, + "num_input_tokens_seen": 35036165, + "step": 1674 + }, + { + "epoch": 0.20140684182047736, + "flos": 20859654811080.0, + "grad_norm": 3.0220944282527586, + "learning_rate": 3.6997826494105037e-06, + "loss": 0.7831, + "num_input_tokens_seen": 35056330, + "step": 1675 + }, + { + "epoch": 0.20152708471111647, + "flos": 20589798565680.0, + "grad_norm": 2.5719693765671128, + "learning_rate": 3.6993720351499286e-06, + "loss": 0.679, + "num_input_tokens_seen": 35077175, + "step": 1676 + }, + { + "epoch": 0.20164732760175555, + "flos": 17241491565360.0, + "grad_norm": 1.828951656881069, + "learning_rate": 3.6989611630959666e-06, + "loss": 0.757, + "num_input_tokens_seen": 35095450, + "step": 1677 + }, + { + "epoch": 0.20176757049239463, + "flos": 52438891557960.0, + "grad_norm": 0.7139165315912531, + "learning_rate": 3.6985500333109474e-06, + "loss": 0.607, + "num_input_tokens_seen": 35163500, + "step": 1678 + }, + { + "epoch": 0.20188781338303372, + "flos": 15672367847040.0, + "grad_norm": 2.6899025402353067, + "learning_rate": 3.6981386458572385e-06, + "loss": 0.7478, + "num_input_tokens_seen": 35181195, + "step": 1679 + }, + { + "epoch": 0.20200805627367283, + "flos": 8358108008160.0, + "grad_norm": 2.3660636223101306, + "learning_rate": 3.6977270007972468e-06, + "loss": 0.7364, + "num_input_tokens_seen": 35198450, + "step": 1680 + }, + { + "epoch": 0.2021282991643119, + "flos": 20938189697760.0, + "grad_norm": 2.5824971435176565, + "learning_rate": 3.6973150981934196e-06, + "loss": 0.6978, + "num_input_tokens_seen": 35219400, + "step": 1681 + }, + { + "epoch": 0.202248542054951, + "flos": 13076587317720.0, + "grad_norm": 3.0844971977538997, + "learning_rate": 3.6969029381082415e-06, + "loss": 0.8109, + "num_input_tokens_seen": 35235115, + "step": 1682 + }, + { + "epoch": 0.2023687849455901, + "flos": 14514249678960.0, + "grad_norm": 2.107051671897696, + "learning_rate": 3.696490520604237e-06, + "loss": 0.7916, + "num_input_tokens_seen": 35253525, + "step": 1683 + }, + { + "epoch": 0.20248902783622919, + "flos": 16512515613240.0, + "grad_norm": 1.8656908365514804, + "learning_rate": 3.696077845743968e-06, + "loss": 0.7908, + "num_input_tokens_seen": 35272835, + "step": 1684 + }, + { + "epoch": 0.20260927072686827, + "flos": 16619698272360.0, + "grad_norm": 2.4300092379861438, + "learning_rate": 3.69566491359004e-06, + "loss": 0.7055, + "num_input_tokens_seen": 35289200, + "step": 1685 + }, + { + "epoch": 0.20272951361750738, + "flos": 37574346657960.0, + "grad_norm": 1.8690843085372162, + "learning_rate": 3.695251724205092e-06, + "loss": 0.6891, + "num_input_tokens_seen": 35313280, + "step": 1686 + }, + { + "epoch": 0.20284975650814646, + "flos": 19482707270400.0, + "grad_norm": 1.904393652388617, + "learning_rate": 3.6948382776518054e-06, + "loss": 0.8482, + "num_input_tokens_seen": 35333705, + "step": 1687 + }, + { + "epoch": 0.20296999939878554, + "flos": 11679571853760.0, + "grad_norm": 3.621909303807162, + "learning_rate": 3.6944245739929e-06, + "loss": 0.7801, + "num_input_tokens_seen": 35349585, + "step": 1688 + }, + { + "epoch": 0.20309024228942463, + "flos": 14024234342640.0, + "grad_norm": 2.3146528584373125, + "learning_rate": 3.6940106132911332e-06, + "loss": 0.7104, + "num_input_tokens_seen": 35366490, + "step": 1689 + }, + { + "epoch": 0.20321048518006374, + "flos": 16702380613800.0, + "grad_norm": 2.261915129205182, + "learning_rate": 3.6935963956093037e-06, + "loss": 0.8639, + "num_input_tokens_seen": 35386295, + "step": 1690 + }, + { + "epoch": 0.20333072807070282, + "flos": 13945414516320.0, + "grad_norm": 1.9904576589052407, + "learning_rate": 3.6931819210102474e-06, + "loss": 0.6781, + "num_input_tokens_seen": 35405410, + "step": 1691 + }, + { + "epoch": 0.2034509709613419, + "flos": 13267465437000.0, + "grad_norm": 2.0753192747901883, + "learning_rate": 3.6927671895568402e-06, + "loss": 0.8399, + "num_input_tokens_seen": 35424190, + "step": 1692 + }, + { + "epoch": 0.20357121385198101, + "flos": 16777527884760.0, + "grad_norm": 2.01713777299099, + "learning_rate": 3.692352201311996e-06, + "loss": 0.8559, + "num_input_tokens_seen": 35442760, + "step": 1693 + }, + { + "epoch": 0.2036914567426201, + "flos": 15296785300680.0, + "grad_norm": 2.4052523241498736, + "learning_rate": 3.6919369563386687e-06, + "loss": 0.7479, + "num_input_tokens_seen": 35462280, + "step": 1694 + }, + { + "epoch": 0.20381169963325918, + "flos": 11297783955240.0, + "grad_norm": 2.7310250824733617, + "learning_rate": 3.69152145469985e-06, + "loss": 0.7801, + "num_input_tokens_seen": 35479045, + "step": 1695 + }, + { + "epoch": 0.20393194252389826, + "flos": 21150560438520.0, + "grad_norm": 2.300529634612744, + "learning_rate": 3.691105696458572e-06, + "loss": 0.8088, + "num_input_tokens_seen": 35496060, + "step": 1696 + }, + { + "epoch": 0.20405218541453737, + "flos": 16455694967760.0, + "grad_norm": 3.466615724800639, + "learning_rate": 3.690689681677904e-06, + "loss": 0.6561, + "num_input_tokens_seen": 35514250, + "step": 1697 + }, + { + "epoch": 0.20417242830517646, + "flos": 18591912550920.0, + "grad_norm": 1.9851857128756778, + "learning_rate": 3.690273410420956e-06, + "loss": 0.8686, + "num_input_tokens_seen": 35533735, + "step": 1698 + }, + { + "epoch": 0.20429267119581554, + "flos": 10760952520800.0, + "grad_norm": 2.6350886526522075, + "learning_rate": 3.689856882750875e-06, + "loss": 0.7528, + "num_input_tokens_seen": 35548655, + "step": 1699 + }, + { + "epoch": 0.20441291408645465, + "flos": 12972412354800.0, + "grad_norm": 3.1099963310195524, + "learning_rate": 3.6894400987308486e-06, + "loss": 0.7741, + "num_input_tokens_seen": 35565895, + "step": 1700 + }, + { + "epoch": 0.20453315697709373, + "flos": 11976682833360.0, + "grad_norm": 2.2304302686588544, + "learning_rate": 3.6890230584241024e-06, + "loss": 0.825, + "num_input_tokens_seen": 35582545, + "step": 1701 + }, + { + "epoch": 0.20465339986773282, + "flos": 49190517085920.0, + "grad_norm": 0.9286733272730795, + "learning_rate": 3.6886057618939016e-06, + "loss": 0.693, + "num_input_tokens_seen": 35645085, + "step": 1702 + }, + { + "epoch": 0.2047736427583719, + "flos": 30876624558720.0, + "grad_norm": 2.296423835300324, + "learning_rate": 3.6881882092035492e-06, + "loss": 0.6713, + "num_input_tokens_seen": 35666190, + "step": 1703 + }, + { + "epoch": 0.204893885649011, + "flos": 51576550159800.0, + "grad_norm": 1.0984760630364905, + "learning_rate": 3.6877704004163873e-06, + "loss": 0.6732, + "num_input_tokens_seen": 35726315, + "step": 1704 + }, + { + "epoch": 0.2050141285396501, + "flos": 16242627707880.0, + "grad_norm": 2.3571106611257617, + "learning_rate": 3.6873523355957984e-06, + "loss": 0.7698, + "num_input_tokens_seen": 35745035, + "step": 1705 + }, + { + "epoch": 0.20513437143028918, + "flos": 34071792112080.0, + "grad_norm": 1.0588957296014623, + "learning_rate": 3.686934014805201e-06, + "loss": 0.7144, + "num_input_tokens_seen": 35795385, + "step": 1706 + }, + { + "epoch": 0.20525461432092829, + "flos": 16023260115960.0, + "grad_norm": 2.076294745526341, + "learning_rate": 3.6865154381080552e-06, + "loss": 0.7947, + "num_input_tokens_seen": 35815790, + "step": 1707 + }, + { + "epoch": 0.20537485721156737, + "flos": 15513018556560.0, + "grad_norm": 2.1907884131414392, + "learning_rate": 3.6860966055678585e-06, + "loss": 0.8092, + "num_input_tokens_seen": 35831865, + "step": 1708 + }, + { + "epoch": 0.20549510010220645, + "flos": 14755679771640.0, + "grad_norm": 1.8256845078895811, + "learning_rate": 3.685677517248147e-06, + "loss": 0.8495, + "num_input_tokens_seen": 35850475, + "step": 1709 + }, + { + "epoch": 0.20561534299284553, + "flos": 12406299948720.0, + "grad_norm": 1.9394755597464066, + "learning_rate": 3.6852581732124967e-06, + "loss": 0.7929, + "num_input_tokens_seen": 35867540, + "step": 1710 + }, + { + "epoch": 0.20573558588348465, + "flos": 16753217526840.0, + "grad_norm": 2.3690481395862353, + "learning_rate": 3.6848385735245213e-06, + "loss": 0.7473, + "num_input_tokens_seen": 35886350, + "step": 1711 + }, + { + "epoch": 0.20585582877412373, + "flos": 18050490422280.0, + "grad_norm": 2.185576001502466, + "learning_rate": 3.6844187182478734e-06, + "loss": 0.8495, + "num_input_tokens_seen": 35906925, + "step": 1712 + }, + { + "epoch": 0.2059760716647628, + "flos": 17727929326200.0, + "grad_norm": 2.8692742427617093, + "learning_rate": 3.683998607446246e-06, + "loss": 0.7352, + "num_input_tokens_seen": 35925295, + "step": 1713 + }, + { + "epoch": 0.20609631455540192, + "flos": 14784295884120.0, + "grad_norm": 2.2907808383529913, + "learning_rate": 3.6835782411833686e-06, + "loss": 0.7278, + "num_input_tokens_seen": 35944535, + "step": 1714 + }, + { + "epoch": 0.206216557446041, + "flos": 14514534618600.0, + "grad_norm": 2.1687859172210446, + "learning_rate": 3.68315761952301e-06, + "loss": 0.7275, + "num_input_tokens_seen": 35961485, + "step": 1715 + }, + { + "epoch": 0.2063368003366801, + "flos": 17645880183960.0, + "grad_norm": 2.784545317812789, + "learning_rate": 3.6827367425289797e-06, + "loss": 0.8176, + "num_input_tokens_seen": 35980980, + "step": 1716 + }, + { + "epoch": 0.2064570432273192, + "flos": 14866345026360.0, + "grad_norm": 3.001438272934793, + "learning_rate": 3.6823156102651225e-06, + "loss": 0.7061, + "num_input_tokens_seen": 35998855, + "step": 1717 + }, + { + "epoch": 0.20657728611795828, + "flos": 14999389381440.0, + "grad_norm": 2.3868054563065844, + "learning_rate": 3.6818942227953257e-06, + "loss": 0.6917, + "num_input_tokens_seen": 36019120, + "step": 1718 + }, + { + "epoch": 0.20669752900859736, + "flos": 15946149927480.0, + "grad_norm": 3.002284928241195, + "learning_rate": 3.681472580183512e-06, + "loss": 0.6786, + "num_input_tokens_seen": 36037490, + "step": 1719 + }, + { + "epoch": 0.20681777189923645, + "flos": 11004472170840.0, + "grad_norm": 2.3315287037358114, + "learning_rate": 3.6810506824936455e-06, + "loss": 0.855, + "num_input_tokens_seen": 36055290, + "step": 1720 + }, + { + "epoch": 0.20693801478987556, + "flos": 41615831178360.0, + "grad_norm": 1.06073513432537, + "learning_rate": 3.680628529789726e-06, + "loss": 0.6553, + "num_input_tokens_seen": 36107420, + "step": 1721 + }, + { + "epoch": 0.20705825768051464, + "flos": 15808799817840.0, + "grad_norm": 2.490177408202578, + "learning_rate": 3.680206122135796e-06, + "loss": 0.8364, + "num_input_tokens_seen": 36127745, + "step": 1722 + }, + { + "epoch": 0.20717850057115372, + "flos": 18943089759480.0, + "grad_norm": 2.3158293805037737, + "learning_rate": 3.6797834595959323e-06, + "loss": 0.767, + "num_input_tokens_seen": 36147365, + "step": 1723 + }, + { + "epoch": 0.20729874346179283, + "flos": 21371669328240.0, + "grad_norm": 3.233930916637519, + "learning_rate": 3.679360542234254e-06, + "loss": 0.7594, + "num_input_tokens_seen": 36166430, + "step": 1724 + }, + { + "epoch": 0.20741898635243192, + "flos": 21292437922440.0, + "grad_norm": 3.1361963732601934, + "learning_rate": 3.678937370114916e-06, + "loss": 0.7112, + "num_input_tokens_seen": 36185955, + "step": 1725 + }, + { + "epoch": 0.207539229243071, + "flos": 11328141365520.0, + "grad_norm": 2.482027204040169, + "learning_rate": 3.678513943302114e-06, + "loss": 0.7769, + "num_input_tokens_seen": 36202450, + "step": 1726 + }, + { + "epoch": 0.20765947213371008, + "flos": 15000022580640.0, + "grad_norm": 2.7766162117918727, + "learning_rate": 3.678090261860082e-06, + "loss": 0.8341, + "num_input_tokens_seen": 36221900, + "step": 1727 + }, + { + "epoch": 0.2077797150243492, + "flos": 14136735875040.0, + "grad_norm": 2.288851679444967, + "learning_rate": 3.6776663258530906e-06, + "loss": 0.7621, + "num_input_tokens_seen": 36240270, + "step": 1728 + }, + { + "epoch": 0.20789995791498828, + "flos": 15968022468480.0, + "grad_norm": 2.1891177054066784, + "learning_rate": 3.6772421353454516e-06, + "loss": 0.7003, + "num_input_tokens_seen": 36258585, + "step": 1729 + }, + { + "epoch": 0.20802020080562736, + "flos": 16944855485160.0, + "grad_norm": 4.4548918849443435, + "learning_rate": 3.6768176904015153e-06, + "loss": 0.8644, + "num_input_tokens_seen": 36278110, + "step": 1730 + }, + { + "epoch": 0.20814044369626647, + "flos": 16888034839680.0, + "grad_norm": 2.47346991820788, + "learning_rate": 3.6763929910856674e-06, + "loss": 0.5912, + "num_input_tokens_seen": 36296280, + "step": 1731 + }, + { + "epoch": 0.20826068658690555, + "flos": 14324163058680.0, + "grad_norm": 2.608311419247677, + "learning_rate": 3.6759680374623365e-06, + "loss": 0.7562, + "num_input_tokens_seen": 36313915, + "step": 1732 + }, + { + "epoch": 0.20838092947754464, + "flos": 18591595951320.0, + "grad_norm": 2.9109131513591846, + "learning_rate": 3.675542829595986e-06, + "loss": 0.7252, + "num_input_tokens_seen": 36333300, + "step": 1733 + }, + { + "epoch": 0.20850117236818372, + "flos": 17620556707320.0, + "grad_norm": 1.569944727708298, + "learning_rate": 3.6751173675511213e-06, + "loss": 0.7757, + "num_input_tokens_seen": 36355065, + "step": 1734 + }, + { + "epoch": 0.20862141525882283, + "flos": 14671192812480.0, + "grad_norm": 2.4180485183795524, + "learning_rate": 3.674691651392283e-06, + "loss": 0.8729, + "num_input_tokens_seen": 36372455, + "step": 1735 + }, + { + "epoch": 0.2087416581494619, + "flos": 28686309086640.0, + "grad_norm": 2.5418220430836724, + "learning_rate": 3.674265681184053e-06, + "loss": 0.7448, + "num_input_tokens_seen": 36395435, + "step": 1736 + }, + { + "epoch": 0.208861901040101, + "flos": 19130073703680.0, + "grad_norm": 1.9241762727299667, + "learning_rate": 3.6738394569910504e-06, + "loss": 0.8484, + "num_input_tokens_seen": 36415695, + "step": 1737 + }, + { + "epoch": 0.2089821439307401, + "flos": 20829550680480.0, + "grad_norm": 2.554517734498134, + "learning_rate": 3.6734129788779333e-06, + "loss": 0.8108, + "num_input_tokens_seen": 36434590, + "step": 1738 + }, + { + "epoch": 0.2091023868213792, + "flos": 15406152497040.0, + "grad_norm": 2.179767894287722, + "learning_rate": 3.6729862469093976e-06, + "loss": 0.8927, + "num_input_tokens_seen": 36453405, + "step": 1739 + }, + { + "epoch": 0.20922262971201827, + "flos": 16431701209440.0, + "grad_norm": 3.856004733751486, + "learning_rate": 3.6725592611501782e-06, + "loss": 0.7891, + "num_input_tokens_seen": 36471800, + "step": 1740 + }, + { + "epoch": 0.20934287260265738, + "flos": 20425225381800.0, + "grad_norm": 2.111666642540036, + "learning_rate": 3.6721320216650496e-06, + "loss": 0.7481, + "num_input_tokens_seen": 36492135, + "step": 1741 + }, + { + "epoch": 0.20946311549329646, + "flos": 11976366233760.0, + "grad_norm": 2.1289091617669307, + "learning_rate": 3.6717045285188215e-06, + "loss": 0.831, + "num_input_tokens_seen": 36509550, + "step": 1742 + }, + { + "epoch": 0.20958335838393555, + "flos": 16459367523120.0, + "grad_norm": 3.5596802912965537, + "learning_rate": 3.671276781776346e-06, + "loss": 0.849, + "num_input_tokens_seen": 36527925, + "step": 1743 + }, + { + "epoch": 0.20970360127457463, + "flos": 18480645756960.0, + "grad_norm": 2.8420143280205026, + "learning_rate": 3.6708487815025128e-06, + "loss": 0.6542, + "num_input_tokens_seen": 36548225, + "step": 1744 + }, + { + "epoch": 0.20982384416521374, + "flos": 13489049226120.0, + "grad_norm": 2.552815255807918, + "learning_rate": 3.6704205277622463e-06, + "loss": 0.7214, + "num_input_tokens_seen": 36566385, + "step": 1745 + }, + { + "epoch": 0.20994408705585282, + "flos": 18591152711880.0, + "grad_norm": 1.8561687875572122, + "learning_rate": 3.6699920206205146e-06, + "loss": 0.7833, + "num_input_tokens_seen": 36586845, + "step": 1746 + }, + { + "epoch": 0.2100643299464919, + "flos": 15591585103200.0, + "grad_norm": 1.817532360518173, + "learning_rate": 3.669563260142321e-06, + "loss": 0.8095, + "num_input_tokens_seen": 36605455, + "step": 1747 + }, + { + "epoch": 0.21018457283713102, + "flos": 14136704215080.0, + "grad_norm": 2.537724820404908, + "learning_rate": 3.6691342463927083e-06, + "loss": 0.817, + "num_input_tokens_seen": 36624170, + "step": 1748 + }, + { + "epoch": 0.2103048157277701, + "flos": 20779315306680.0, + "grad_norm": 2.1472537576986963, + "learning_rate": 3.668704979436758e-06, + "loss": 0.8027, + "num_input_tokens_seen": 36643985, + "step": 1749 + }, + { + "epoch": 0.21042505861840918, + "flos": 12732755219880.0, + "grad_norm": 2.4676015572871814, + "learning_rate": 3.668275459339588e-06, + "loss": 0.7715, + "num_input_tokens_seen": 36662185, + "step": 1750 + }, + { + "epoch": 0.21054530150904827, + "flos": 10329372487920.0, + "grad_norm": 2.137813866309142, + "learning_rate": 3.667845686166358e-06, + "loss": 0.7856, + "num_input_tokens_seen": 36678830, + "step": 1751 + }, + { + "epoch": 0.21066554439968738, + "flos": 13590976331880.0, + "grad_norm": 1.8629921742887654, + "learning_rate": 3.6674156599822634e-06, + "loss": 0.8462, + "num_input_tokens_seen": 36694345, + "step": 1752 + }, + { + "epoch": 0.21078578729032646, + "flos": 17322210989280.0, + "grad_norm": 2.0895536132199988, + "learning_rate": 3.666985380852539e-06, + "loss": 0.795, + "num_input_tokens_seen": 36713070, + "step": 1753 + }, + { + "epoch": 0.21090603018096554, + "flos": 21530733679080.0, + "grad_norm": 2.5673496596792056, + "learning_rate": 3.6665548488424576e-06, + "loss": 0.7329, + "num_input_tokens_seen": 36731550, + "step": 1754 + }, + { + "epoch": 0.21102627307160465, + "flos": 17028455965440.0, + "grad_norm": 1.8744999536660292, + "learning_rate": 3.6661240640173307e-06, + "loss": 0.8629, + "num_input_tokens_seen": 36752740, + "step": 1755 + }, + { + "epoch": 0.21114651596224374, + "flos": 40250376203160.0, + "grad_norm": 0.939190104061514, + "learning_rate": 3.6656930264425085e-06, + "loss": 0.6301, + "num_input_tokens_seen": 36816505, + "step": 1756 + }, + { + "epoch": 0.21126675885288282, + "flos": 15756633186480.0, + "grad_norm": 2.073028395307861, + "learning_rate": 3.665261736183378e-06, + "loss": 0.742, + "num_input_tokens_seen": 36836260, + "step": 1757 + }, + { + "epoch": 0.2113870017435219, + "flos": 7925578176480.0, + "grad_norm": 4.094908106405117, + "learning_rate": 3.664830193305366e-06, + "loss": 0.8612, + "num_input_tokens_seen": 36853755, + "step": 1758 + }, + { + "epoch": 0.211507244634161, + "flos": 12138026701320.0, + "grad_norm": 3.3774392532092494, + "learning_rate": 3.6643983978739373e-06, + "loss": 0.7454, + "num_input_tokens_seen": 36870090, + "step": 1759 + }, + { + "epoch": 0.2116274875248001, + "flos": 15319797600240.0, + "grad_norm": 1.753975983658619, + "learning_rate": 3.663966349954596e-06, + "loss": 0.7993, + "num_input_tokens_seen": 36889990, + "step": 1760 + }, + { + "epoch": 0.21174773041543918, + "flos": 50286970634640.0, + "grad_norm": 1.0371112715709176, + "learning_rate": 3.6635340496128816e-06, + "loss": 0.6107, + "num_input_tokens_seen": 36946640, + "step": 1761 + }, + { + "epoch": 0.2118679733060783, + "flos": 15109168157280.0, + "grad_norm": 1.9101916974919015, + "learning_rate": 3.6631014969143747e-06, + "loss": 0.9042, + "num_input_tokens_seen": 36966050, + "step": 1762 + }, + { + "epoch": 0.21198821619671737, + "flos": 17000283092400.0, + "grad_norm": 1.9003640277879266, + "learning_rate": 3.662668691924693e-06, + "loss": 0.8691, + "num_input_tokens_seen": 36986820, + "step": 1763 + }, + { + "epoch": 0.21210845908735645, + "flos": 17943687682680.0, + "grad_norm": 3.023749283675779, + "learning_rate": 3.6622356347094927e-06, + "loss": 0.6963, + "num_input_tokens_seen": 37008105, + "step": 1764 + }, + { + "epoch": 0.21222870197799554, + "flos": 19861487472720.0, + "grad_norm": 2.1536461408461904, + "learning_rate": 3.6618023253344684e-06, + "loss": 0.7715, + "num_input_tokens_seen": 37026685, + "step": 1765 + }, + { + "epoch": 0.21234894486863465, + "flos": 12301016887200.0, + "grad_norm": 1.900415277272175, + "learning_rate": 3.6613687638653527e-06, + "loss": 0.8275, + "num_input_tokens_seen": 37044575, + "step": 1766 + }, + { + "epoch": 0.21246918775927373, + "flos": 17185873998360.0, + "grad_norm": 2.3512729200334066, + "learning_rate": 3.660934950367916e-06, + "loss": 0.765, + "num_input_tokens_seen": 37063540, + "step": 1767 + }, + { + "epoch": 0.21258943064991281, + "flos": 16377508340640.0, + "grad_norm": 1.8462460965558876, + "learning_rate": 3.660500884907968e-06, + "loss": 0.8204, + "num_input_tokens_seen": 37084000, + "step": 1768 + }, + { + "epoch": 0.21270967354055192, + "flos": 43806336610200.0, + "grad_norm": 0.8566338256621763, + "learning_rate": 3.660066567551356e-06, + "loss": 0.6176, + "num_input_tokens_seen": 37143865, + "step": 1769 + }, + { + "epoch": 0.212829916431191, + "flos": 15757266385680.0, + "grad_norm": 2.0610750072958197, + "learning_rate": 3.6596319983639657e-06, + "loss": 0.8136, + "num_input_tokens_seen": 37162165, + "step": 1770 + }, + { + "epoch": 0.2129501593218301, + "flos": 21266196306960.0, + "grad_norm": 1.99925253728868, + "learning_rate": 3.6591971774117214e-06, + "loss": 0.849, + "num_input_tokens_seen": 37184860, + "step": 1771 + }, + { + "epoch": 0.2130704022124692, + "flos": 13731587417400.0, + "grad_norm": 2.874668102180297, + "learning_rate": 3.6587621047605833e-06, + "loss": 0.7932, + "num_input_tokens_seen": 37201750, + "step": 1772 + }, + { + "epoch": 0.21319064510310828, + "flos": 10113550811520.0, + "grad_norm": 2.0446044908756527, + "learning_rate": 3.6583267804765542e-06, + "loss": 0.8579, + "num_input_tokens_seen": 37215805, + "step": 1773 + }, + { + "epoch": 0.21331088799374737, + "flos": 15324736554000.0, + "grad_norm": 2.1431128530972976, + "learning_rate": 3.6578912046256702e-06, + "loss": 0.8422, + "num_input_tokens_seen": 37234045, + "step": 1774 + }, + { + "epoch": 0.21343113088438645, + "flos": 13596643464720.0, + "grad_norm": 2.193596980743779, + "learning_rate": 3.6574553772740083e-06, + "loss": 0.7471, + "num_input_tokens_seen": 37251695, + "step": 1775 + }, + { + "epoch": 0.21355137377502556, + "flos": 49709560114200.0, + "grad_norm": 0.8834905805317715, + "learning_rate": 3.657019298487684e-06, + "loss": 0.6546, + "num_input_tokens_seen": 37316425, + "step": 1776 + }, + { + "epoch": 0.21367161666566464, + "flos": 25369182655560.0, + "grad_norm": 1.9506384317949492, + "learning_rate": 3.6565829683328495e-06, + "loss": 0.7993, + "num_input_tokens_seen": 37338770, + "step": 1777 + }, + { + "epoch": 0.21379185955630373, + "flos": 13866531370080.0, + "grad_norm": 1.873616598316198, + "learning_rate": 3.6561463868756965e-06, + "loss": 0.8405, + "num_input_tokens_seen": 37357190, + "step": 1778 + }, + { + "epoch": 0.21391210244694284, + "flos": 20696474665440.0, + "grad_norm": 1.8033216187430963, + "learning_rate": 3.655709554182452e-06, + "loss": 0.775, + "num_input_tokens_seen": 37377250, + "step": 1779 + }, + { + "epoch": 0.21403234533758192, + "flos": 12730792302360.0, + "grad_norm": 2.0164433673574287, + "learning_rate": 3.6552724703193855e-06, + "loss": 0.8456, + "num_input_tokens_seen": 37394160, + "step": 1780 + }, + { + "epoch": 0.214152588228221, + "flos": 38032679357040.0, + "grad_norm": 0.7781331640221231, + "learning_rate": 3.654835135352801e-06, + "loss": 0.5756, + "num_input_tokens_seen": 37448690, + "step": 1781 + }, + { + "epoch": 0.21427283111886009, + "flos": 14241037477800.0, + "grad_norm": 2.1374523849091718, + "learning_rate": 3.654397549349043e-06, + "loss": 0.8637, + "num_input_tokens_seen": 37465785, + "step": 1782 + }, + { + "epoch": 0.2143930740094992, + "flos": 14677366504680.0, + "grad_norm": 2.3975157317758207, + "learning_rate": 3.653959712374491e-06, + "loss": 0.73, + "num_input_tokens_seen": 37483610, + "step": 1783 + }, + { + "epoch": 0.21451331690013828, + "flos": 15945548388240.0, + "grad_norm": 2.0788818357941423, + "learning_rate": 3.6535216244955663e-06, + "loss": 0.8134, + "num_input_tokens_seen": 37503225, + "step": 1784 + }, + { + "epoch": 0.21463355979077736, + "flos": 24127210727520.0, + "grad_norm": 1.8045914529450249, + "learning_rate": 3.653083285778726e-06, + "loss": 0.686, + "num_input_tokens_seen": 37524315, + "step": 1785 + }, + { + "epoch": 0.21475380268141647, + "flos": 15757076425920.0, + "grad_norm": 2.7767616659935808, + "learning_rate": 3.6526446962904653e-06, + "loss": 0.7873, + "num_input_tokens_seen": 37542750, + "step": 1786 + }, + { + "epoch": 0.21487404557205556, + "flos": 23612790053400.0, + "grad_norm": 1.7398361966432077, + "learning_rate": 3.652205856097318e-06, + "loss": 0.7226, + "num_input_tokens_seen": 37565655, + "step": 1787 + }, + { + "epoch": 0.21499428846269464, + "flos": 9191480542920.0, + "grad_norm": 3.1967253977207077, + "learning_rate": 3.651766765265856e-06, + "loss": 0.7768, + "num_input_tokens_seen": 37582385, + "step": 1788 + }, + { + "epoch": 0.21511453135333372, + "flos": 17182897962120.0, + "grad_norm": 4.242593707172034, + "learning_rate": 3.65132742386269e-06, + "loss": 0.7829, + "num_input_tokens_seen": 37597325, + "step": 1789 + }, + { + "epoch": 0.21523477424397283, + "flos": 19753228374960.0, + "grad_norm": 2.510618832205275, + "learning_rate": 3.6508878319544656e-06, + "loss": 0.8257, + "num_input_tokens_seen": 37617260, + "step": 1790 + }, + { + "epoch": 0.21535501713461191, + "flos": 13812338501280.0, + "grad_norm": 3.4193593711265757, + "learning_rate": 3.65044798960787e-06, + "loss": 0.7914, + "num_input_tokens_seen": 37635320, + "step": 1791 + }, + { + "epoch": 0.215475260025251, + "flos": 13056772674120.0, + "grad_norm": 5.828701941848141, + "learning_rate": 3.650007896889627e-06, + "loss": 0.7627, + "num_input_tokens_seen": 37653620, + "step": 1792 + }, + { + "epoch": 0.2155955029158901, + "flos": 12138438280800.0, + "grad_norm": 4.127404639137746, + "learning_rate": 3.6495675538664974e-06, + "loss": 0.7858, + "num_input_tokens_seen": 37672355, + "step": 1793 + }, + { + "epoch": 0.2157157458065292, + "flos": 17294164756080.0, + "grad_norm": 2.026851292888224, + "learning_rate": 3.649126960605282e-06, + "loss": 0.8105, + "num_input_tokens_seen": 37693060, + "step": 1794 + }, + { + "epoch": 0.21583598869716827, + "flos": 16188656458800.0, + "grad_norm": 2.5652345162721715, + "learning_rate": 3.6486861171728174e-06, + "loss": 0.806, + "num_input_tokens_seen": 37711175, + "step": 1795 + }, + { + "epoch": 0.21595623158780738, + "flos": 17161310360760.0, + "grad_norm": 2.0951924835128515, + "learning_rate": 3.6482450236359803e-06, + "loss": 0.7723, + "num_input_tokens_seen": 37732750, + "step": 1796 + }, + { + "epoch": 0.21607647447844647, + "flos": 19725720361080.0, + "grad_norm": 2.5037169481629817, + "learning_rate": 3.647803680061683e-06, + "loss": 0.7505, + "num_input_tokens_seen": 37752885, + "step": 1797 + }, + { + "epoch": 0.21619671736908555, + "flos": 10539811971120.0, + "grad_norm": 3.306322058773986, + "learning_rate": 3.6473620865168776e-06, + "loss": 0.7216, + "num_input_tokens_seen": 37769475, + "step": 1798 + }, + { + "epoch": 0.21631696025972463, + "flos": 13083330889200.0, + "grad_norm": 2.2487077180832524, + "learning_rate": 3.646920243068554e-06, + "loss": 0.8071, + "num_input_tokens_seen": 37787090, + "step": 1799 + }, + { + "epoch": 0.21643720315036374, + "flos": 17916971167800.0, + "grad_norm": 1.7744885661513792, + "learning_rate": 3.6464781497837384e-06, + "loss": 0.7241, + "num_input_tokens_seen": 37808785, + "step": 1800 + }, + { + "epoch": 0.21655744604100283, + "flos": 20885959746480.0, + "grad_norm": 5.713628758404521, + "learning_rate": 3.6460358067294965e-06, + "loss": 0.7182, + "num_input_tokens_seen": 37829735, + "step": 1801 + }, + { + "epoch": 0.2166776889316419, + "flos": 14727126979080.0, + "grad_norm": 2.375013150630481, + "learning_rate": 3.645593213972932e-06, + "loss": 0.7575, + "num_input_tokens_seen": 37848360, + "step": 1802 + }, + { + "epoch": 0.21679793182228102, + "flos": 11057145361560.0, + "grad_norm": 2.57420819154464, + "learning_rate": 3.6451503715811852e-06, + "loss": 0.7638, + "num_input_tokens_seen": 37866390, + "step": 1803 + }, + { + "epoch": 0.2169181747129201, + "flos": 12678878950680.0, + "grad_norm": 2.337427792989329, + "learning_rate": 3.6447072796214345e-06, + "loss": 0.7934, + "num_input_tokens_seen": 37884675, + "step": 1804 + }, + { + "epoch": 0.21703841760355919, + "flos": 43305656358720.0, + "grad_norm": 0.96274916903951, + "learning_rate": 3.644263938160898e-06, + "loss": 0.6625, + "num_input_tokens_seen": 37940360, + "step": 1805 + }, + { + "epoch": 0.21715866049419827, + "flos": 16404224855520.0, + "grad_norm": 3.7351981177633498, + "learning_rate": 3.6438203472668293e-06, + "loss": 0.706, + "num_input_tokens_seen": 37959725, + "step": 1806 + }, + { + "epoch": 0.21727890338483738, + "flos": 12570176613480.0, + "grad_norm": 5.500517560709078, + "learning_rate": 3.6433765070065206e-06, + "loss": 0.81, + "num_input_tokens_seen": 37977235, + "step": 1807 + }, + { + "epoch": 0.21739914627547646, + "flos": 9754553592840.0, + "grad_norm": 3.1111261238765286, + "learning_rate": 3.6429324174473025e-06, + "loss": 0.8669, + "num_input_tokens_seen": 37990495, + "step": 1808 + }, + { + "epoch": 0.21751938916611555, + "flos": 15324578254200.0, + "grad_norm": 3.3267286353241885, + "learning_rate": 3.6424880786565425e-06, + "loss": 0.8316, + "num_input_tokens_seen": 38006360, + "step": 1809 + }, + { + "epoch": 0.21763963205675466, + "flos": 20238937956720.0, + "grad_norm": 2.479784677448586, + "learning_rate": 3.6420434907016482e-06, + "loss": 0.7719, + "num_input_tokens_seen": 38025770, + "step": 1810 + }, + { + "epoch": 0.21775987494739374, + "flos": 15672589466760.0, + "grad_norm": 2.0140210060834907, + "learning_rate": 3.6415986536500606e-06, + "loss": 0.7994, + "num_input_tokens_seen": 38043820, + "step": 1811 + }, + { + "epoch": 0.21788011783803282, + "flos": 13380821788320.0, + "grad_norm": 2.1580264748477482, + "learning_rate": 3.641153567569263e-06, + "loss": 0.8065, + "num_input_tokens_seen": 38061855, + "step": 1812 + }, + { + "epoch": 0.2180003607286719, + "flos": 22209505917360.0, + "grad_norm": 4.494153145014612, + "learning_rate": 3.640708232526774e-06, + "loss": 0.9375, + "num_input_tokens_seen": 38080230, + "step": 1813 + }, + { + "epoch": 0.21812060361931102, + "flos": 18669750918480.0, + "grad_norm": 2.1551997626312285, + "learning_rate": 3.6402626485901504e-06, + "loss": 0.7641, + "num_input_tokens_seen": 38099045, + "step": 1814 + }, + { + "epoch": 0.2182408465099501, + "flos": 16026647731680.0, + "grad_norm": 2.0971126597952563, + "learning_rate": 3.639816815826988e-06, + "loss": 0.7548, + "num_input_tokens_seen": 38118090, + "step": 1815 + }, + { + "epoch": 0.21836108940058918, + "flos": 17320849611000.0, + "grad_norm": 2.191736206672667, + "learning_rate": 3.6393707343049176e-06, + "loss": 0.7621, + "num_input_tokens_seen": 38138140, + "step": 1816 + }, + { + "epoch": 0.2184813322912283, + "flos": 18078726615240.0, + "grad_norm": 2.9624026254352307, + "learning_rate": 3.6389244040916104e-06, + "loss": 0.7275, + "num_input_tokens_seen": 38156935, + "step": 1817 + }, + { + "epoch": 0.21860157518186737, + "flos": 19479636254280.0, + "grad_norm": 2.6760770536770417, + "learning_rate": 3.6384778252547747e-06, + "loss": 0.7812, + "num_input_tokens_seen": 38172535, + "step": 1818 + }, + { + "epoch": 0.21872181807250646, + "flos": 15270353725440.0, + "grad_norm": 2.822273903804899, + "learning_rate": 3.638030997862155e-06, + "loss": 0.759, + "num_input_tokens_seen": 38191190, + "step": 1819 + }, + { + "epoch": 0.21884206096314554, + "flos": 45118268067120.0, + "grad_norm": 0.830966680329268, + "learning_rate": 3.6375839219815356e-06, + "loss": 0.6178, + "num_input_tokens_seen": 38248710, + "step": 1820 + }, + { + "epoch": 0.21896230385378465, + "flos": 17184892539600.0, + "grad_norm": 2.4067280282134766, + "learning_rate": 3.6371365976807375e-06, + "loss": 0.8131, + "num_input_tokens_seen": 38268825, + "step": 1821 + }, + { + "epoch": 0.21908254674442373, + "flos": 18375362695440.0, + "grad_norm": 2.1974715615299574, + "learning_rate": 3.6366890250276185e-06, + "loss": 0.8118, + "num_input_tokens_seen": 38289500, + "step": 1822 + }, + { + "epoch": 0.21920278963506282, + "flos": 17214933350280.0, + "grad_norm": 1.9413052998936102, + "learning_rate": 3.6362412040900764e-06, + "loss": 0.8793, + "num_input_tokens_seen": 38309010, + "step": 1823 + }, + { + "epoch": 0.21932303252570193, + "flos": 21454414989600.0, + "grad_norm": 2.0067097645099286, + "learning_rate": 3.635793134936044e-06, + "loss": 0.7852, + "num_input_tokens_seen": 38329740, + "step": 1824 + }, + { + "epoch": 0.219443275416341, + "flos": 15210652023600.0, + "grad_norm": 2.4839908134855717, + "learning_rate": 3.635344817633494e-06, + "loss": 0.7158, + "num_input_tokens_seen": 38348775, + "step": 1825 + }, + { + "epoch": 0.2195635183069801, + "flos": 10545257484240.0, + "grad_norm": 4.162950435243488, + "learning_rate": 3.634896252250436e-06, + "loss": 0.7425, + "num_input_tokens_seen": 38365260, + "step": 1826 + }, + { + "epoch": 0.2196837611976192, + "flos": 17754962440680.0, + "grad_norm": 2.551108183070831, + "learning_rate": 3.6344474388549157e-06, + "loss": 0.8022, + "num_input_tokens_seen": 38384635, + "step": 1827 + }, + { + "epoch": 0.2198040040882583, + "flos": 13189120510080.0, + "grad_norm": 3.471589447007555, + "learning_rate": 3.6339983775150183e-06, + "loss": 0.7972, + "num_input_tokens_seen": 38400915, + "step": 1828 + }, + { + "epoch": 0.21992424697889737, + "flos": 12974470252200.0, + "grad_norm": 4.044674779693211, + "learning_rate": 3.6335490682988664e-06, + "loss": 0.8336, + "num_input_tokens_seen": 38416245, + "step": 1829 + }, + { + "epoch": 0.22004448986953645, + "flos": 12866686053840.0, + "grad_norm": 2.277299465056395, + "learning_rate": 3.63309951127462e-06, + "loss": 0.812, + "num_input_tokens_seen": 38432875, + "step": 1830 + }, + { + "epoch": 0.22016473276017556, + "flos": 16298878474080.0, + "grad_norm": 2.118653300281387, + "learning_rate": 3.6326497065104757e-06, + "loss": 0.7355, + "num_input_tokens_seen": 38453060, + "step": 1831 + }, + { + "epoch": 0.22028497565081465, + "flos": 18726413264160.0, + "grad_norm": 2.812179292831543, + "learning_rate": 3.6321996540746697e-06, + "loss": 0.7603, + "num_input_tokens_seen": 38471855, + "step": 1832 + }, + { + "epoch": 0.22040521854145373, + "flos": 26638630937520.0, + "grad_norm": 1.9126270127350367, + "learning_rate": 3.6317493540354733e-06, + "loss": 0.7817, + "num_input_tokens_seen": 38494990, + "step": 1833 + }, + { + "epoch": 0.22052546143209284, + "flos": 8574404583960.0, + "grad_norm": 2.199919079246033, + "learning_rate": 3.6312988064611976e-06, + "loss": 0.7545, + "num_input_tokens_seen": 38513020, + "step": 1834 + }, + { + "epoch": 0.22064570432273192, + "flos": 17729955563640.0, + "grad_norm": 2.000187388224587, + "learning_rate": 3.6308480114201896e-06, + "loss": 0.7894, + "num_input_tokens_seen": 38534660, + "step": 1835 + }, + { + "epoch": 0.220765947213371, + "flos": 13083552508920.0, + "grad_norm": 2.05336820651244, + "learning_rate": 3.630396968980835e-06, + "loss": 0.7501, + "num_input_tokens_seen": 38552255, + "step": 1836 + }, + { + "epoch": 0.2208861901040101, + "flos": 19614611866920.0, + "grad_norm": 3.155584586223375, + "learning_rate": 3.6299456792115575e-06, + "loss": 0.8339, + "num_input_tokens_seen": 38573230, + "step": 1837 + }, + { + "epoch": 0.2210064329946492, + "flos": 12999097209720.0, + "grad_norm": 2.35587004980301, + "learning_rate": 3.629494142180815e-06, + "loss": 0.794, + "num_input_tokens_seen": 38591695, + "step": 1838 + }, + { + "epoch": 0.22112667588528828, + "flos": 13110174043920.0, + "grad_norm": 2.7739523826093135, + "learning_rate": 3.6290423579571075e-06, + "loss": 0.8323, + "num_input_tokens_seen": 38607955, + "step": 1839 + }, + { + "epoch": 0.22124691877592736, + "flos": 13407601623120.0, + "grad_norm": 1.9073386913803418, + "learning_rate": 3.6285903266089694e-06, + "loss": 0.7824, + "num_input_tokens_seen": 38626950, + "step": 1840 + }, + { + "epoch": 0.22136716166656648, + "flos": 15216319156440.0, + "grad_norm": 2.1136707786003095, + "learning_rate": 3.628138048204974e-06, + "loss": 0.757, + "num_input_tokens_seen": 38647355, + "step": 1841 + }, + { + "epoch": 0.22148740455720556, + "flos": 12894320707560.0, + "grad_norm": 2.023287380938926, + "learning_rate": 3.6276855228137304e-06, + "loss": 0.7508, + "num_input_tokens_seen": 38665280, + "step": 1842 + }, + { + "epoch": 0.22160764744784464, + "flos": 15892495278000.0, + "grad_norm": 2.1968214611838888, + "learning_rate": 3.6272327505038874e-06, + "loss": 0.8019, + "num_input_tokens_seen": 38681465, + "step": 1843 + }, + { + "epoch": 0.22172789033848372, + "flos": 17400017696880.0, + "grad_norm": 2.178693975624527, + "learning_rate": 3.626779731344131e-06, + "loss": 0.7708, + "num_input_tokens_seen": 38700975, + "step": 1844 + }, + { + "epoch": 0.22184813322912283, + "flos": 12381134771880.0, + "grad_norm": 3.543911002133101, + "learning_rate": 3.6263264654031814e-06, + "loss": 0.8356, + "num_input_tokens_seen": 38717595, + "step": 1845 + }, + { + "epoch": 0.22196837611976192, + "flos": 45572037240600.0, + "grad_norm": 0.6954383950432114, + "learning_rate": 3.6258729527498008e-06, + "loss": 0.5936, + "num_input_tokens_seen": 38778160, + "step": 1846 + }, + { + "epoch": 0.222088619010401, + "flos": 18727964602200.0, + "grad_norm": 2.93954591319497, + "learning_rate": 3.6254191934527854e-06, + "loss": 0.6061, + "num_input_tokens_seen": 38797235, + "step": 1847 + }, + { + "epoch": 0.2222088619010401, + "flos": 14110335959760.0, + "grad_norm": 2.9596833564891214, + "learning_rate": 3.6249651875809715e-06, + "loss": 0.6254, + "num_input_tokens_seen": 38816835, + "step": 1848 + }, + { + "epoch": 0.2223291047916792, + "flos": 13948010633040.0, + "grad_norm": 2.115015471201194, + "learning_rate": 3.62451093520323e-06, + "loss": 0.8713, + "num_input_tokens_seen": 38834460, + "step": 1849 + }, + { + "epoch": 0.22244934768231828, + "flos": 15243890490240.0, + "grad_norm": 4.159599159442689, + "learning_rate": 3.6240564363884714e-06, + "loss": 0.8814, + "num_input_tokens_seen": 38854125, + "step": 1850 + }, + { + "epoch": 0.2225695905729574, + "flos": 11382587514000.0, + "grad_norm": 2.24480304287346, + "learning_rate": 3.623601691205643e-06, + "loss": 0.683, + "num_input_tokens_seen": 38872920, + "step": 1851 + }, + { + "epoch": 0.22268983346359647, + "flos": 18591184371840.0, + "grad_norm": 2.4331488286793435, + "learning_rate": 3.623146699723729e-06, + "loss": 0.7896, + "num_input_tokens_seen": 38892100, + "step": 1852 + }, + { + "epoch": 0.22281007635423555, + "flos": 9627556290120.0, + "grad_norm": 2.3351936304157785, + "learning_rate": 3.6226914620117507e-06, + "loss": 0.7708, + "num_input_tokens_seen": 38910440, + "step": 1853 + }, + { + "epoch": 0.22293031924487464, + "flos": 11166132638400.0, + "grad_norm": 2.3826185804160205, + "learning_rate": 3.622235978138768e-06, + "loss": 0.7889, + "num_input_tokens_seen": 38927785, + "step": 1854 + }, + { + "epoch": 0.22305056213551375, + "flos": 16512452293320.0, + "grad_norm": 2.5755374754910267, + "learning_rate": 3.621780248173877e-06, + "loss": 0.8049, + "num_input_tokens_seen": 38945705, + "step": 1855 + }, + { + "epoch": 0.22317080502615283, + "flos": 47833035929280.0, + "grad_norm": 0.8816568355396782, + "learning_rate": 3.6213242721862125e-06, + "loss": 0.6479, + "num_input_tokens_seen": 39003880, + "step": 1856 + }, + { + "epoch": 0.2232910479167919, + "flos": 18889150170360.0, + "grad_norm": 1.7526178534220365, + "learning_rate": 3.620868050244945e-06, + "loss": 0.7311, + "num_input_tokens_seen": 39024080, + "step": 1857 + }, + { + "epoch": 0.22341129080743102, + "flos": 17020129395960.0, + "grad_norm": 2.818086797337275, + "learning_rate": 3.6204115824192817e-06, + "loss": 0.7524, + "num_input_tokens_seen": 39041275, + "step": 1858 + }, + { + "epoch": 0.2235315336980701, + "flos": 15513271836240.0, + "grad_norm": 3.610600251147167, + "learning_rate": 3.619954868778471e-06, + "loss": 0.7496, + "num_input_tokens_seen": 39057690, + "step": 1859 + }, + { + "epoch": 0.2236517765887092, + "flos": 14541979312560.0, + "grad_norm": 2.3482874372485694, + "learning_rate": 3.6194979093917944e-06, + "loss": 0.8083, + "num_input_tokens_seen": 39076825, + "step": 1860 + }, + { + "epoch": 0.22377201947934827, + "flos": 16992716361960.0, + "grad_norm": 2.4075514011303003, + "learning_rate": 3.6190407043285724e-06, + "loss": 0.8722, + "num_input_tokens_seen": 39094280, + "step": 1861 + }, + { + "epoch": 0.22389226236998738, + "flos": 19642246520640.0, + "grad_norm": 2.4537844409361975, + "learning_rate": 3.618583253658163e-06, + "loss": 0.7305, + "num_input_tokens_seen": 39114100, + "step": 1862 + }, + { + "epoch": 0.22401250526062647, + "flos": 17700832891800.0, + "grad_norm": 2.000856216259562, + "learning_rate": 3.618125557449961e-06, + "loss": 0.8476, + "num_input_tokens_seen": 39131875, + "step": 1863 + }, + { + "epoch": 0.22413274815126555, + "flos": 12215928388800.0, + "grad_norm": 2.0716723312746503, + "learning_rate": 3.6176676157733983e-06, + "loss": 0.8149, + "num_input_tokens_seen": 39146605, + "step": 1864 + }, + { + "epoch": 0.22425299104190466, + "flos": 15621024374640.0, + "grad_norm": 2.6646029911305353, + "learning_rate": 3.6172094286979443e-06, + "loss": 0.7374, + "num_input_tokens_seen": 39163695, + "step": 1865 + }, + { + "epoch": 0.22437323393254374, + "flos": 23618488846200.0, + "grad_norm": 1.6946393623528484, + "learning_rate": 3.6167509962931064e-06, + "loss": 0.7971, + "num_input_tokens_seen": 39189115, + "step": 1866 + }, + { + "epoch": 0.22449347682318282, + "flos": 13136225699640.0, + "grad_norm": 7.198122616888634, + "learning_rate": 3.6162923186284276e-06, + "loss": 0.7581, + "num_input_tokens_seen": 39204795, + "step": 1867 + }, + { + "epoch": 0.2246137197138219, + "flos": 13650361434120.0, + "grad_norm": 2.482781721035964, + "learning_rate": 3.6158333957734888e-06, + "loss": 0.8433, + "num_input_tokens_seen": 39223105, + "step": 1868 + }, + { + "epoch": 0.22473396260446102, + "flos": 11350488805920.0, + "grad_norm": 2.51403018795513, + "learning_rate": 3.6153742277979088e-06, + "loss": 0.8178, + "num_input_tokens_seen": 39240255, + "step": 1869 + }, + { + "epoch": 0.2248542054951001, + "flos": 10518635949240.0, + "grad_norm": 2.5556760036260964, + "learning_rate": 3.6149148147713434e-06, + "loss": 0.7688, + "num_input_tokens_seen": 39258210, + "step": 1870 + }, + { + "epoch": 0.22497444838573918, + "flos": 14054781712680.0, + "grad_norm": 2.2129283339544146, + "learning_rate": 3.614455156763484e-06, + "loss": 0.8585, + "num_input_tokens_seen": 39276235, + "step": 1871 + }, + { + "epoch": 0.2250946912763783, + "flos": 12327986681760.0, + "grad_norm": 2.252885134305618, + "learning_rate": 3.613995253844061e-06, + "loss": 0.6911, + "num_input_tokens_seen": 39293635, + "step": 1872 + }, + { + "epoch": 0.22521493416701738, + "flos": 18079043214840.0, + "grad_norm": 2.3119459777315243, + "learning_rate": 3.6135351060828414e-06, + "loss": 0.7909, + "num_input_tokens_seen": 39313830, + "step": 1873 + }, + { + "epoch": 0.22533517705765646, + "flos": 13002611465280.0, + "grad_norm": 2.551246121677888, + "learning_rate": 3.6130747135496285e-06, + "loss": 0.6645, + "num_input_tokens_seen": 39332550, + "step": 1874 + }, + { + "epoch": 0.22545541994829554, + "flos": 24746787823320.0, + "grad_norm": 2.2512948231150705, + "learning_rate": 3.6126140763142646e-06, + "loss": 0.653, + "num_input_tokens_seen": 39357300, + "step": 1875 + }, + { + "epoch": 0.22557566283893465, + "flos": 14001380342880.0, + "grad_norm": 3.412511966670367, + "learning_rate": 3.6121531944466275e-06, + "loss": 0.831, + "num_input_tokens_seen": 39374345, + "step": 1876 + }, + { + "epoch": 0.22569590572957374, + "flos": 15186215025840.0, + "grad_norm": 2.8247842183031193, + "learning_rate": 3.611692068016633e-06, + "loss": 0.7683, + "num_input_tokens_seen": 39390395, + "step": 1877 + }, + { + "epoch": 0.22581614862021282, + "flos": 13462079431560.0, + "grad_norm": 2.396113558980567, + "learning_rate": 3.611230697094233e-06, + "loss": 0.7306, + "num_input_tokens_seen": 39406815, + "step": 1878 + }, + { + "epoch": 0.22593639151085193, + "flos": 14650428370080.0, + "grad_norm": 19.907340745970362, + "learning_rate": 3.6107690817494173e-06, + "loss": 0.864, + "num_input_tokens_seen": 39426755, + "step": 1879 + }, + { + "epoch": 0.226056634401491, + "flos": 9519202212480.0, + "grad_norm": 2.2307690954334816, + "learning_rate": 3.6103072220522117e-06, + "loss": 0.6941, + "num_input_tokens_seen": 39442005, + "step": 1880 + }, + { + "epoch": 0.2261768772921301, + "flos": 13865676551160.0, + "grad_norm": 1.9661957656549347, + "learning_rate": 3.609845118072682e-06, + "loss": 0.8971, + "num_input_tokens_seen": 39460395, + "step": 1881 + }, + { + "epoch": 0.2262971201827692, + "flos": 14595222382560.0, + "grad_norm": 9.558932966753696, + "learning_rate": 3.6093827698809276e-06, + "loss": 0.7725, + "num_input_tokens_seen": 39479215, + "step": 1882 + }, + { + "epoch": 0.2264173630734083, + "flos": 12138659900520.0, + "grad_norm": 3.107207496484343, + "learning_rate": 3.6089201775470864e-06, + "loss": 0.8331, + "num_input_tokens_seen": 39494390, + "step": 1883 + }, + { + "epoch": 0.22653760596404737, + "flos": 17863316518320.0, + "grad_norm": 1.9217160787237384, + "learning_rate": 3.6084573411413334e-06, + "loss": 0.7703, + "num_input_tokens_seen": 39513505, + "step": 1884 + }, + { + "epoch": 0.22665784885468646, + "flos": 13380441868800.0, + "grad_norm": 2.661089269999275, + "learning_rate": 3.607994260733881e-06, + "loss": 0.7877, + "num_input_tokens_seen": 39532465, + "step": 1885 + }, + { + "epoch": 0.22677809174532557, + "flos": 17617359051360.0, + "grad_norm": 2.007607772423565, + "learning_rate": 3.6075309363949776e-06, + "loss": 0.7343, + "num_input_tokens_seen": 39551355, + "step": 1886 + }, + { + "epoch": 0.22689833463596465, + "flos": 14891225263560.0, + "grad_norm": 2.895520328282394, + "learning_rate": 3.6070673681949094e-06, + "loss": 0.7903, + "num_input_tokens_seen": 39569440, + "step": 1887 + }, + { + "epoch": 0.22701857752660373, + "flos": 22104666095280.0, + "grad_norm": 2.8211359317217357, + "learning_rate": 3.606603556203999e-06, + "loss": 0.7831, + "num_input_tokens_seen": 39594105, + "step": 1888 + }, + { + "epoch": 0.22713882041724284, + "flos": 16458892623720.0, + "grad_norm": 2.150393840291931, + "learning_rate": 3.6061395004926066e-06, + "loss": 0.8323, + "num_input_tokens_seen": 39612760, + "step": 1889 + }, + { + "epoch": 0.22725906330788193, + "flos": 15000434160120.0, + "grad_norm": 3.0954662148109664, + "learning_rate": 3.605675201131129e-06, + "loss": 0.8222, + "num_input_tokens_seen": 39630940, + "step": 1890 + }, + { + "epoch": 0.227379306198521, + "flos": 13867037929440.0, + "grad_norm": 3.4831868014627463, + "learning_rate": 3.60521065819e-06, + "loss": 0.7739, + "num_input_tokens_seen": 39647970, + "step": 1891 + }, + { + "epoch": 0.2274995490891601, + "flos": 15945326768520.0, + "grad_norm": 2.0563052230886614, + "learning_rate": 3.60474587173969e-06, + "loss": 0.8566, + "num_input_tokens_seen": 39666175, + "step": 1892 + }, + { + "epoch": 0.2276197919797992, + "flos": 13917304963200.0, + "grad_norm": 2.7680546492325067, + "learning_rate": 3.6042808418507084e-06, + "loss": 0.8224, + "num_input_tokens_seen": 39683580, + "step": 1893 + }, + { + "epoch": 0.22774003487043828, + "flos": 13731397457640.0, + "grad_norm": 2.1224787846056494, + "learning_rate": 3.6038155685935976e-06, + "loss": 0.7574, + "num_input_tokens_seen": 39699870, + "step": 1894 + }, + { + "epoch": 0.22786027776107737, + "flos": 16836533067480.0, + "grad_norm": 2.67380289047739, + "learning_rate": 3.6033500520389404e-06, + "loss": 0.686, + "num_input_tokens_seen": 39716260, + "step": 1895 + }, + { + "epoch": 0.22798052065171648, + "flos": 49185039912840.0, + "grad_norm": 0.8193197873948387, + "learning_rate": 3.6028842922573553e-06, + "loss": 0.6802, + "num_input_tokens_seen": 39780125, + "step": 1896 + }, + { + "epoch": 0.22810076354235556, + "flos": 45762598760280.0, + "grad_norm": 0.8144991196193819, + "learning_rate": 3.602418289319497e-06, + "loss": 0.6557, + "num_input_tokens_seen": 39838400, + "step": 1897 + }, + { + "epoch": 0.22822100643299464, + "flos": 17482953318000.0, + "grad_norm": 1.876474309495869, + "learning_rate": 3.601952043296059e-06, + "loss": 0.7245, + "num_input_tokens_seen": 39858115, + "step": 1898 + }, + { + "epoch": 0.22834124932363373, + "flos": 15347463913920.0, + "grad_norm": 2.235593023628711, + "learning_rate": 3.6014855542577696e-06, + "loss": 0.7963, + "num_input_tokens_seen": 39875045, + "step": 1899 + }, + { + "epoch": 0.22846149221427284, + "flos": 18242191700520.0, + "grad_norm": 1.9507761144360403, + "learning_rate": 3.6010188222753943e-06, + "loss": 0.8249, + "num_input_tokens_seen": 39895535, + "step": 1900 + }, + { + "epoch": 0.22858173510491192, + "flos": 41360031955200.0, + "grad_norm": 1.0109103592733057, + "learning_rate": 3.6005518474197372e-06, + "loss": 0.6671, + "num_input_tokens_seen": 39947300, + "step": 1901 + }, + { + "epoch": 0.228701977995551, + "flos": 17701244471280.0, + "grad_norm": 2.424004977872666, + "learning_rate": 3.6000846297616373e-06, + "loss": 0.7695, + "num_input_tokens_seen": 39965320, + "step": 1902 + }, + { + "epoch": 0.22882222088619011, + "flos": 15640269138960.0, + "grad_norm": 3.070395121542597, + "learning_rate": 3.5996171693719717e-06, + "loss": 0.7199, + "num_input_tokens_seen": 39981135, + "step": 1903 + }, + { + "epoch": 0.2289424637768292, + "flos": 47617974091920.0, + "grad_norm": 0.9215684154653087, + "learning_rate": 3.5991494663216528e-06, + "loss": 0.6619, + "num_input_tokens_seen": 40043840, + "step": 1904 + }, + { + "epoch": 0.22906270666746828, + "flos": 16215531273480.0, + "grad_norm": 2.2175391748316065, + "learning_rate": 3.5986815206816314e-06, + "loss": 0.8529, + "num_input_tokens_seen": 40062380, + "step": 1905 + }, + { + "epoch": 0.2291829495581074, + "flos": 18885920854440.0, + "grad_norm": 2.3942699694319014, + "learning_rate": 3.598213332522895e-06, + "loss": 0.7278, + "num_input_tokens_seen": 40082130, + "step": 1906 + }, + { + "epoch": 0.22930319244874647, + "flos": 22882674342720.0, + "grad_norm": 2.392225141372392, + "learning_rate": 3.597744901916466e-06, + "loss": 0.7662, + "num_input_tokens_seen": 40103135, + "step": 1907 + }, + { + "epoch": 0.22942343533938556, + "flos": 17022155633400.0, + "grad_norm": 2.281761948087056, + "learning_rate": 3.5972762289334058e-06, + "loss": 0.7426, + "num_input_tokens_seen": 40122485, + "step": 1908 + }, + { + "epoch": 0.22954367823002464, + "flos": 10626166867920.0, + "grad_norm": 2.8432418872338823, + "learning_rate": 3.5968073136448116e-06, + "loss": 0.8307, + "num_input_tokens_seen": 40140225, + "step": 1909 + }, + { + "epoch": 0.22966392112066375, + "flos": 12219316004520.0, + "grad_norm": 2.061008376377044, + "learning_rate": 3.596338156121818e-06, + "loss": 0.8923, + "num_input_tokens_seen": 40158830, + "step": 1910 + }, + { + "epoch": 0.22978416401130283, + "flos": 49753115236440.0, + "grad_norm": 0.75906009816606, + "learning_rate": 3.595868756435595e-06, + "loss": 0.6126, + "num_input_tokens_seen": 40226230, + "step": 1911 + }, + { + "epoch": 0.22990440690194192, + "flos": 14514692918400.0, + "grad_norm": 2.2472671230290615, + "learning_rate": 3.5953991146573504e-06, + "loss": 0.7872, + "num_input_tokens_seen": 40244595, + "step": 1912 + }, + { + "epoch": 0.23002464979258103, + "flos": 9646896034320.0, + "grad_norm": 2.607812013620339, + "learning_rate": 3.5949292308583294e-06, + "loss": 0.8126, + "num_input_tokens_seen": 40257560, + "step": 1913 + }, + { + "epoch": 0.2301448926832201, + "flos": 16215847873080.0, + "grad_norm": 2.313037397472361, + "learning_rate": 3.594459105109811e-06, + "loss": 0.7844, + "num_input_tokens_seen": 40276460, + "step": 1914 + }, + { + "epoch": 0.2302651355738592, + "flos": 15135884672160.0, + "grad_norm": 2.4921663576383275, + "learning_rate": 3.593988737483115e-06, + "loss": 0.7973, + "num_input_tokens_seen": 40296120, + "step": 1915 + }, + { + "epoch": 0.23038537846449827, + "flos": 13570338529320.0, + "grad_norm": 2.181018928596112, + "learning_rate": 3.5935181280495947e-06, + "loss": 0.771, + "num_input_tokens_seen": 40314420, + "step": 1916 + }, + { + "epoch": 0.23050562135513739, + "flos": 47347516307280.0, + "grad_norm": 0.901601213387618, + "learning_rate": 3.5930472768806412e-06, + "loss": 0.5798, + "num_input_tokens_seen": 40372810, + "step": 1917 + }, + { + "epoch": 0.23062586424577647, + "flos": 12626554019520.0, + "grad_norm": 2.1134205770484833, + "learning_rate": 3.5925761840476826e-06, + "loss": 0.7472, + "num_input_tokens_seen": 40391140, + "step": 1918 + }, + { + "epoch": 0.23074610713641555, + "flos": 20428518017640.0, + "grad_norm": 2.3203480521923834, + "learning_rate": 3.592104849622183e-06, + "loss": 0.8058, + "num_input_tokens_seen": 40413115, + "step": 1919 + }, + { + "epoch": 0.23086635002705466, + "flos": 21158190488880.0, + "grad_norm": 1.9848127708248686, + "learning_rate": 3.591633273675644e-06, + "loss": 0.714, + "num_input_tokens_seen": 40435070, + "step": 1920 + }, + { + "epoch": 0.23098659291769374, + "flos": 46381664805360.0, + "grad_norm": 0.9770970216834738, + "learning_rate": 3.591161456279602e-06, + "loss": 0.6204, + "num_input_tokens_seen": 40480335, + "step": 1921 + }, + { + "epoch": 0.23110683580833283, + "flos": 17187520316280.0, + "grad_norm": 1.6853514521571482, + "learning_rate": 3.590689397505633e-06, + "loss": 0.782, + "num_input_tokens_seen": 40500965, + "step": 1922 + }, + { + "epoch": 0.2312270786989719, + "flos": 19995481626600.0, + "grad_norm": 1.704906546752754, + "learning_rate": 3.590217097425347e-06, + "loss": 0.8495, + "num_input_tokens_seen": 40520585, + "step": 1923 + }, + { + "epoch": 0.23134732158961102, + "flos": 9627461310240.0, + "grad_norm": 2.8543530331395535, + "learning_rate": 3.589744556110391e-06, + "loss": 0.6787, + "num_input_tokens_seen": 40538295, + "step": 1924 + }, + { + "epoch": 0.2314675644802501, + "flos": 27179198247240.0, + "grad_norm": 1.9895521497695987, + "learning_rate": 3.58927177363245e-06, + "loss": 0.8302, + "num_input_tokens_seen": 40560840, + "step": 1925 + }, + { + "epoch": 0.2315878073708892, + "flos": 17458009760880.0, + "grad_norm": 2.9432419194692434, + "learning_rate": 3.5887987500632447e-06, + "loss": 0.719, + "num_input_tokens_seen": 40578565, + "step": 1926 + }, + { + "epoch": 0.2317080502615283, + "flos": 16863882781560.0, + "grad_norm": 2.1420738718738663, + "learning_rate": 3.5883254854745325e-06, + "loss": 0.8214, + "num_input_tokens_seen": 40596675, + "step": 1927 + }, + { + "epoch": 0.23182829315216738, + "flos": 8142381311640.0, + "grad_norm": 3.1706051845070444, + "learning_rate": 3.587851979938107e-06, + "loss": 0.7433, + "num_input_tokens_seen": 40613285, + "step": 1928 + }, + { + "epoch": 0.23194853604280646, + "flos": 14487944743560.0, + "grad_norm": 2.456209127071517, + "learning_rate": 3.5873782335257985e-06, + "loss": 0.7598, + "num_input_tokens_seen": 40631170, + "step": 1929 + }, + { + "epoch": 0.23206877893344555, + "flos": 11139986002800.0, + "grad_norm": 5.814010587899796, + "learning_rate": 3.5869042463094744e-06, + "loss": 0.772, + "num_input_tokens_seen": 40648605, + "step": 1930 + }, + { + "epoch": 0.23218902182408466, + "flos": 16621249610400.0, + "grad_norm": 2.2551842409980494, + "learning_rate": 3.586430018361038e-06, + "loss": 0.7597, + "num_input_tokens_seen": 40668095, + "step": 1931 + }, + { + "epoch": 0.23230926471472374, + "flos": 16620584751240.0, + "grad_norm": 2.6889854584141477, + "learning_rate": 3.5859555497524283e-06, + "loss": 0.7408, + "num_input_tokens_seen": 40685050, + "step": 1932 + }, + { + "epoch": 0.23242950760536282, + "flos": 14892333362160.0, + "grad_norm": 2.4349060928754165, + "learning_rate": 3.5854808405556237e-06, + "loss": 0.9077, + "num_input_tokens_seen": 40704005, + "step": 1933 + }, + { + "epoch": 0.23254975049600193, + "flos": 12326245383960.0, + "grad_norm": 3.130409746254212, + "learning_rate": 3.5850058908426355e-06, + "loss": 0.7368, + "num_input_tokens_seen": 40722275, + "step": 1934 + }, + { + "epoch": 0.23266999338664102, + "flos": 17052892963200.0, + "grad_norm": 2.6606601680324826, + "learning_rate": 3.584530700685514e-06, + "loss": 0.8275, + "num_input_tokens_seen": 40742255, + "step": 1935 + }, + { + "epoch": 0.2327902362772801, + "flos": 14295800225880.0, + "grad_norm": 2.6283136071057736, + "learning_rate": 3.5840552701563448e-06, + "loss": 0.8709, + "num_input_tokens_seen": 40758175, + "step": 1936 + }, + { + "epoch": 0.2329104791679192, + "flos": 12192662809560.0, + "grad_norm": 4.037574479210371, + "learning_rate": 3.5835795993272513e-06, + "loss": 0.7984, + "num_input_tokens_seen": 40776180, + "step": 1937 + }, + { + "epoch": 0.2330307220585583, + "flos": 16215816213120.0, + "grad_norm": 4.930153705297769, + "learning_rate": 3.583103688270391e-06, + "loss": 0.6916, + "num_input_tokens_seen": 40795680, + "step": 1938 + }, + { + "epoch": 0.23315096494919738, + "flos": 14109196201200.0, + "grad_norm": 2.87777128582995, + "learning_rate": 3.58262753705796e-06, + "loss": 0.8686, + "num_input_tokens_seen": 40810290, + "step": 1939 + }, + { + "epoch": 0.23327120783983646, + "flos": 39064908321000.0, + "grad_norm": 0.7697217289217011, + "learning_rate": 3.5821511457621902e-06, + "loss": 0.5664, + "num_input_tokens_seen": 40867310, + "step": 1940 + }, + { + "epoch": 0.23339145073047557, + "flos": 12488032491360.0, + "grad_norm": 3.7720803294199183, + "learning_rate": 3.5816745144553497e-06, + "loss": 0.7844, + "num_input_tokens_seen": 40882350, + "step": 1941 + }, + { + "epoch": 0.23351169362111465, + "flos": 9492454037640.0, + "grad_norm": 3.4235802151049146, + "learning_rate": 3.5811976432097424e-06, + "loss": 0.728, + "num_input_tokens_seen": 40899740, + "step": 1942 + }, + { + "epoch": 0.23363193651175373, + "flos": 11544089681760.0, + "grad_norm": 2.4132259363792334, + "learning_rate": 3.58072053209771e-06, + "loss": 0.8286, + "num_input_tokens_seen": 40916015, + "step": 1943 + }, + { + "epoch": 0.23375217940239285, + "flos": 15372660750720.0, + "grad_norm": 11.740355733911517, + "learning_rate": 3.5802431811916296e-06, + "loss": 0.7785, + "num_input_tokens_seen": 40932345, + "step": 1944 + }, + { + "epoch": 0.23387242229303193, + "flos": 15054595368960.0, + "grad_norm": 2.3180372597676033, + "learning_rate": 3.579765590563916e-06, + "loss": 0.7833, + "num_input_tokens_seen": 40951465, + "step": 1945 + }, + { + "epoch": 0.233992665183671, + "flos": 17781552315720.0, + "grad_norm": 2.8507496968742303, + "learning_rate": 3.579287760287017e-06, + "loss": 0.7937, + "num_input_tokens_seen": 40971935, + "step": 1946 + }, + { + "epoch": 0.2341129080743101, + "flos": 22129672972320.0, + "grad_norm": 1.7673655345222696, + "learning_rate": 3.578809690433421e-06, + "loss": 0.7071, + "num_input_tokens_seen": 40993365, + "step": 1947 + }, + { + "epoch": 0.2342331509649492, + "flos": 16674967579800.0, + "grad_norm": 3.4723704921006946, + "learning_rate": 3.578331381075651e-06, + "loss": 0.7841, + "num_input_tokens_seen": 41013585, + "step": 1948 + }, + { + "epoch": 0.2343533938555883, + "flos": 17295874393920.0, + "grad_norm": 4.6755526047766995, + "learning_rate": 3.5778528322862646e-06, + "loss": 0.6691, + "num_input_tokens_seen": 41032125, + "step": 1949 + }, + { + "epoch": 0.23447363674622737, + "flos": 17996645813040.0, + "grad_norm": 1.8563231739559174, + "learning_rate": 3.5773740441378585e-06, + "loss": 0.8515, + "num_input_tokens_seen": 41052600, + "step": 1950 + }, + { + "epoch": 0.23459387963686648, + "flos": 39140589319920.0, + "grad_norm": 2.989725001010133, + "learning_rate": 3.5768950167030633e-06, + "loss": 0.7242, + "num_input_tokens_seen": 41077020, + "step": 1951 + }, + { + "epoch": 0.23471412252750556, + "flos": 17538792504720.0, + "grad_norm": 2.9165411422603094, + "learning_rate": 3.576415750054548e-06, + "loss": 0.7632, + "num_input_tokens_seen": 41096860, + "step": 1952 + }, + { + "epoch": 0.23483436541814465, + "flos": 11437191962280.0, + "grad_norm": 2.190968950943907, + "learning_rate": 3.5759362442650172e-06, + "loss": 0.8375, + "num_input_tokens_seen": 41113330, + "step": 1953 + }, + { + "epoch": 0.23495460830878373, + "flos": 18267610157040.0, + "grad_norm": 3.3025797441717843, + "learning_rate": 3.5754564994072113e-06, + "loss": 0.8193, + "num_input_tokens_seen": 41131890, + "step": 1954 + }, + { + "epoch": 0.23507485119942284, + "flos": 22372401123360.0, + "grad_norm": 2.6638595190496526, + "learning_rate": 3.5749765155539067e-06, + "loss": 0.5922, + "num_input_tokens_seen": 41152095, + "step": 1955 + }, + { + "epoch": 0.23519509409006192, + "flos": 13377434172600.0, + "grad_norm": 2.9225471253915267, + "learning_rate": 3.574496292777917e-06, + "loss": 0.9052, + "num_input_tokens_seen": 41170025, + "step": 1956 + }, + { + "epoch": 0.235315336980701, + "flos": 21751114389720.0, + "grad_norm": 2.5062335501342656, + "learning_rate": 3.574015831152092e-06, + "loss": 0.6882, + "num_input_tokens_seen": 41190160, + "step": 1957 + }, + { + "epoch": 0.23543557987134012, + "flos": 13542672215640.0, + "grad_norm": 2.6016207851719795, + "learning_rate": 3.573535130749316e-06, + "loss": 0.8123, + "num_input_tokens_seen": 41207830, + "step": 1958 + }, + { + "epoch": 0.2355558227619792, + "flos": 18076985317440.0, + "grad_norm": 1.9036626297557209, + "learning_rate": 3.5730541916425127e-06, + "loss": 0.7166, + "num_input_tokens_seen": 41229030, + "step": 1959 + }, + { + "epoch": 0.23567606565261828, + "flos": 16053775826040.0, + "grad_norm": 2.046080456923934, + "learning_rate": 3.572573013904639e-06, + "loss": 0.8394, + "num_input_tokens_seen": 41248660, + "step": 1960 + }, + { + "epoch": 0.2357963085432574, + "flos": 9870157801320.0, + "grad_norm": 2.720202704618063, + "learning_rate": 3.572091597608689e-06, + "loss": 0.9057, + "num_input_tokens_seen": 41266505, + "step": 1961 + }, + { + "epoch": 0.23591655143389648, + "flos": 16159977026400.0, + "grad_norm": 2.8202704006444956, + "learning_rate": 3.571609942827694e-06, + "loss": 0.7036, + "num_input_tokens_seen": 41285340, + "step": 1962 + }, + { + "epoch": 0.23603679432453556, + "flos": 12406869828000.0, + "grad_norm": 2.280489232963532, + "learning_rate": 3.57112804963472e-06, + "loss": 0.8612, + "num_input_tokens_seen": 41303275, + "step": 1963 + }, + { + "epoch": 0.23615703721517464, + "flos": 14001127063200.0, + "grad_norm": 2.2183939688420704, + "learning_rate": 3.57064591810287e-06, + "loss": 0.7463, + "num_input_tokens_seen": 41320495, + "step": 1964 + }, + { + "epoch": 0.23627728010581375, + "flos": 13947535733640.0, + "grad_norm": 3.8455445357171567, + "learning_rate": 3.570163548305284e-06, + "loss": 0.7839, + "num_input_tokens_seen": 41339145, + "step": 1965 + }, + { + "epoch": 0.23639752299645284, + "flos": 10382868837600.0, + "grad_norm": 3.0034944527353975, + "learning_rate": 3.569680940315135e-06, + "loss": 0.6809, + "num_input_tokens_seen": 41355265, + "step": 1966 + }, + { + "epoch": 0.23651776588709192, + "flos": 16755813643560.0, + "grad_norm": 3.073776349004783, + "learning_rate": 3.5691980942056356e-06, + "loss": 0.7977, + "num_input_tokens_seen": 41374355, + "step": 1967 + }, + { + "epoch": 0.23663800877773103, + "flos": 13596548484840.0, + "grad_norm": 3.087005816613276, + "learning_rate": 3.5687150100500332e-06, + "loss": 0.7724, + "num_input_tokens_seen": 41393775, + "step": 1968 + }, + { + "epoch": 0.2367582516683701, + "flos": 18726033344640.0, + "grad_norm": 2.059287951283766, + "learning_rate": 3.568231687921611e-06, + "loss": 0.7333, + "num_input_tokens_seen": 41413670, + "step": 1969 + }, + { + "epoch": 0.2368784945590092, + "flos": 17053114582920.0, + "grad_norm": 1.6648154547661866, + "learning_rate": 3.5677481278936883e-06, + "loss": 0.7993, + "num_input_tokens_seen": 41432970, + "step": 1970 + }, + { + "epoch": 0.23699873744964828, + "flos": 51520113925200.0, + "grad_norm": 0.833403552593718, + "learning_rate": 3.5672643300396214e-06, + "loss": 0.5864, + "num_input_tokens_seen": 41501835, + "step": 1971 + }, + { + "epoch": 0.2371189803402874, + "flos": 15972518182800.0, + "grad_norm": 5.420340109144786, + "learning_rate": 3.566780294432802e-06, + "loss": 0.6641, + "num_input_tokens_seen": 41518730, + "step": 1972 + }, + { + "epoch": 0.23723922323092647, + "flos": 16026932671320.0, + "grad_norm": 2.8864397422570858, + "learning_rate": 3.566296021146657e-06, + "loss": 0.7152, + "num_input_tokens_seen": 41537830, + "step": 1973 + }, + { + "epoch": 0.23735946612156555, + "flos": 24019964748480.0, + "grad_norm": 2.0694435455297557, + "learning_rate": 3.565811510254652e-06, + "loss": 0.7087, + "num_input_tokens_seen": 41558430, + "step": 1974 + }, + { + "epoch": 0.23747970901220466, + "flos": 52028234267280.0, + "grad_norm": 0.7843668371132964, + "learning_rate": 3.5653267618302845e-06, + "loss": 0.5855, + "num_input_tokens_seen": 41625730, + "step": 1975 + }, + { + "epoch": 0.23759995190284375, + "flos": 15243352270920.0, + "grad_norm": 2.1158455399888187, + "learning_rate": 3.564841775947093e-06, + "loss": 0.8425, + "num_input_tokens_seen": 41646340, + "step": 1976 + }, + { + "epoch": 0.23772019479348283, + "flos": 24176622942360.0, + "grad_norm": 2.6505834836795987, + "learning_rate": 3.5643565526786475e-06, + "loss": 0.7397, + "num_input_tokens_seen": 41666000, + "step": 1977 + }, + { + "epoch": 0.2378404376841219, + "flos": 23748905424600.0, + "grad_norm": 2.5861377465660773, + "learning_rate": 3.5638710920985574e-06, + "loss": 0.7534, + "num_input_tokens_seen": 41687180, + "step": 1978 + }, + { + "epoch": 0.23796068057476102, + "flos": 16832195652960.0, + "grad_norm": 4.879419524267085, + "learning_rate": 3.5633853942804655e-06, + "loss": 0.7876, + "num_input_tokens_seen": 41705225, + "step": 1979 + }, + { + "epoch": 0.2380809234654001, + "flos": 9790071576600.0, + "grad_norm": 4.170250737178462, + "learning_rate": 3.5628994592980527e-06, + "loss": 0.7473, + "num_input_tokens_seen": 41722850, + "step": 1980 + }, + { + "epoch": 0.2382011663560392, + "flos": 12299212269480.0, + "grad_norm": 2.608345056899499, + "learning_rate": 3.562413287225034e-06, + "loss": 0.6934, + "num_input_tokens_seen": 41740680, + "step": 1981 + }, + { + "epoch": 0.2383214092466783, + "flos": 13460686393320.0, + "grad_norm": 3.101499465570891, + "learning_rate": 3.5619268781351623e-06, + "loss": 0.8689, + "num_input_tokens_seen": 41758470, + "step": 1982 + }, + { + "epoch": 0.23844165213731738, + "flos": 14433720214800.0, + "grad_norm": 3.133210801373295, + "learning_rate": 3.5614402321022256e-06, + "loss": 0.7568, + "num_input_tokens_seen": 41776020, + "step": 1983 + }, + { + "epoch": 0.23856189502795647, + "flos": 17106515952720.0, + "grad_norm": 2.4440761768000825, + "learning_rate": 3.5609533492000463e-06, + "loss": 0.8443, + "num_input_tokens_seen": 41794630, + "step": 1984 + }, + { + "epoch": 0.23868213791859555, + "flos": 17185968978240.0, + "grad_norm": 6.218210168152288, + "learning_rate": 3.560466229502485e-06, + "loss": 0.7602, + "num_input_tokens_seen": 41813695, + "step": 1985 + }, + { + "epoch": 0.23880238080923466, + "flos": 12111025246800.0, + "grad_norm": 3.1709835509827276, + "learning_rate": 3.5599788730834384e-06, + "loss": 0.8836, + "num_input_tokens_seen": 41831375, + "step": 1986 + }, + { + "epoch": 0.23892262369987374, + "flos": 12652257415680.0, + "grad_norm": 4.9163754694854225, + "learning_rate": 3.559491280016836e-06, + "loss": 0.7777, + "num_input_tokens_seen": 41849040, + "step": 1987 + }, + { + "epoch": 0.23904286659051283, + "flos": 16324011990960.0, + "grad_norm": 2.233526749087998, + "learning_rate": 3.5590034503766465e-06, + "loss": 0.696, + "num_input_tokens_seen": 41868425, + "step": 1988 + }, + { + "epoch": 0.23916310948115194, + "flos": 15486428681520.0, + "grad_norm": 2.8677733443863636, + "learning_rate": 3.558515384236874e-06, + "loss": 0.7948, + "num_input_tokens_seen": 41885575, + "step": 1989 + }, + { + "epoch": 0.23928335237179102, + "flos": 10275812818320.0, + "grad_norm": 1.8265314574332647, + "learning_rate": 3.558027081671556e-06, + "loss": 0.814, + "num_input_tokens_seen": 41902280, + "step": 1990 + }, + { + "epoch": 0.2394035952624301, + "flos": 17404323451440.0, + "grad_norm": 2.19938049263687, + "learning_rate": 3.557538542754769e-06, + "loss": 0.668, + "num_input_tokens_seen": 41921695, + "step": 1991 + }, + { + "epoch": 0.2395238381530692, + "flos": 17727264467040.0, + "grad_norm": 2.2028142831713704, + "learning_rate": 3.557049767560623e-06, + "loss": 0.6503, + "num_input_tokens_seen": 41940330, + "step": 1992 + }, + { + "epoch": 0.2396440810437083, + "flos": 18532907368200.0, + "grad_norm": 2.174154718068184, + "learning_rate": 3.5565607561632655e-06, + "loss": 0.838, + "num_input_tokens_seen": 41958890, + "step": 1993 + }, + { + "epoch": 0.23976432393434738, + "flos": 20936163460320.0, + "grad_norm": 2.4738839764044993, + "learning_rate": 3.5560715086368787e-06, + "loss": 0.7672, + "num_input_tokens_seen": 41976480, + "step": 1994 + }, + { + "epoch": 0.23988456682498646, + "flos": 14239454479800.0, + "grad_norm": 1.898631772748468, + "learning_rate": 3.5555820250556816e-06, + "loss": 0.8056, + "num_input_tokens_seen": 41993400, + "step": 1995 + }, + { + "epoch": 0.24000480971562557, + "flos": 14811582278280.0, + "grad_norm": 2.619296462552537, + "learning_rate": 3.5550923054939278e-06, + "loss": 0.6728, + "num_input_tokens_seen": 42012575, + "step": 1996 + }, + { + "epoch": 0.24012505260626466, + "flos": 18642211244640.0, + "grad_norm": 4.353741687434867, + "learning_rate": 3.5546023500259083e-06, + "loss": 0.7114, + "num_input_tokens_seen": 42033390, + "step": 1997 + }, + { + "epoch": 0.24024529549690374, + "flos": 11322759172320.0, + "grad_norm": 2.867168873406799, + "learning_rate": 3.5541121587259477e-06, + "loss": 0.7876, + "num_input_tokens_seen": 42050945, + "step": 1998 + }, + { + "epoch": 0.24036553838754285, + "flos": 42093440301720.0, + "grad_norm": 0.8023684519203281, + "learning_rate": 3.553621731668408e-06, + "loss": 0.5834, + "num_input_tokens_seen": 42113875, + "step": 1999 + }, + { + "epoch": 0.24048578127818193, + "flos": 18292142134680.0, + "grad_norm": 1.9314238115136737, + "learning_rate": 3.553131068927688e-06, + "loss": 0.8218, + "num_input_tokens_seen": 42132000, + "step": 2000 + }, + { + "epoch": 0.24060602416882101, + "flos": 17079577818120.0, + "grad_norm": 2.329697455545462, + "learning_rate": 3.552640170578219e-06, + "loss": 0.7874, + "num_input_tokens_seen": 42151970, + "step": 2001 + }, + { + "epoch": 0.2407262670594601, + "flos": 10302307713480.0, + "grad_norm": 3.2549417045752302, + "learning_rate": 3.5521490366944703e-06, + "loss": 0.7629, + "num_input_tokens_seen": 42169340, + "step": 2002 + }, + { + "epoch": 0.2408465099500992, + "flos": 9924857229480.0, + "grad_norm": 4.654507509396596, + "learning_rate": 3.5516576673509474e-06, + "loss": 0.7806, + "num_input_tokens_seen": 42187060, + "step": 2003 + }, + { + "epoch": 0.2409667528407383, + "flos": 22938703489200.0, + "grad_norm": 1.7573132971710146, + "learning_rate": 3.5511660626221896e-06, + "loss": 0.8457, + "num_input_tokens_seen": 42207420, + "step": 2004 + }, + { + "epoch": 0.24108699573137737, + "flos": 16242564387960.0, + "grad_norm": 2.7069258235862677, + "learning_rate": 3.5506742225827744e-06, + "loss": 0.8539, + "num_input_tokens_seen": 42223995, + "step": 2005 + }, + { + "epoch": 0.24120723862201648, + "flos": 19130991842520.0, + "grad_norm": 2.748865240230549, + "learning_rate": 3.5501821473073116e-06, + "loss": 0.8816, + "num_input_tokens_seen": 42240300, + "step": 2006 + }, + { + "epoch": 0.24132748151265557, + "flos": 13596896744400.0, + "grad_norm": 2.230793249572642, + "learning_rate": 3.54968983687045e-06, + "loss": 0.8522, + "num_input_tokens_seen": 42256890, + "step": 2007 + }, + { + "epoch": 0.24144772440329465, + "flos": 11112224709240.0, + "grad_norm": 4.447839859858414, + "learning_rate": 3.549197291346872e-06, + "loss": 0.8743, + "num_input_tokens_seen": 42273135, + "step": 2008 + }, + { + "epoch": 0.24156796729393373, + "flos": 17592700433880.0, + "grad_norm": 2.494950917894506, + "learning_rate": 3.548704510811297e-06, + "loss": 0.7816, + "num_input_tokens_seen": 42292050, + "step": 2009 + }, + { + "epoch": 0.24168821018457284, + "flos": 19267170533640.0, + "grad_norm": 3.563611576599233, + "learning_rate": 3.5482114953384787e-06, + "loss": 0.7244, + "num_input_tokens_seen": 42311000, + "step": 2010 + }, + { + "epoch": 0.24180845307521193, + "flos": 13299532485120.0, + "grad_norm": 2.5871446448411186, + "learning_rate": 3.5477182450032077e-06, + "loss": 0.8226, + "num_input_tokens_seen": 42329320, + "step": 2011 + }, + { + "epoch": 0.241928695965851, + "flos": 14947032790320.0, + "grad_norm": 2.2433031875187073, + "learning_rate": 3.5472247598803097e-06, + "loss": 0.8117, + "num_input_tokens_seen": 42348385, + "step": 2012 + }, + { + "epoch": 0.24204893885649012, + "flos": 18726603223920.0, + "grad_norm": 2.166986686352301, + "learning_rate": 3.546731040044645e-06, + "loss": 0.8329, + "num_input_tokens_seen": 42363275, + "step": 2013 + }, + { + "epoch": 0.2421691817471292, + "flos": 22503102641400.0, + "grad_norm": 1.9328288713654622, + "learning_rate": 3.546237085571112e-06, + "loss": 0.7383, + "num_input_tokens_seen": 42381430, + "step": 2014 + }, + { + "epoch": 0.24228942463776829, + "flos": 16053965785800.0, + "grad_norm": 2.214792079423738, + "learning_rate": 3.5457428965346425e-06, + "loss": 0.7011, + "num_input_tokens_seen": 42400090, + "step": 2015 + }, + { + "epoch": 0.2424096675284074, + "flos": 24963875898120.0, + "grad_norm": 2.145000135515115, + "learning_rate": 3.545248473010205e-06, + "loss": 0.7394, + "num_input_tokens_seen": 42422615, + "step": 2016 + }, + { + "epoch": 0.24252991041904648, + "flos": 15838207429320.0, + "grad_norm": 4.33013494761431, + "learning_rate": 3.544753815072802e-06, + "loss": 0.8529, + "num_input_tokens_seen": 42440990, + "step": 2017 + }, + { + "epoch": 0.24265015330968556, + "flos": 15998221578960.0, + "grad_norm": 3.33179397006477, + "learning_rate": 3.544258922797474e-06, + "loss": 0.8774, + "num_input_tokens_seen": 42458830, + "step": 2018 + }, + { + "epoch": 0.24277039620032465, + "flos": 18779561354280.0, + "grad_norm": 2.4841491788521375, + "learning_rate": 3.543763796259295e-06, + "loss": 0.7692, + "num_input_tokens_seen": 42478505, + "step": 2019 + }, + { + "epoch": 0.24289063909096376, + "flos": 19266980573880.0, + "grad_norm": 3.1136079279380144, + "learning_rate": 3.5432684355333754e-06, + "loss": 0.8899, + "num_input_tokens_seen": 42496880, + "step": 2020 + }, + { + "epoch": 0.24301088198160284, + "flos": 18369315643080.0, + "grad_norm": 3.040848939966568, + "learning_rate": 3.5427728406948613e-06, + "loss": 0.7466, + "num_input_tokens_seen": 42515715, + "step": 2021 + }, + { + "epoch": 0.24313112487224192, + "flos": 51551674413960.0, + "grad_norm": 0.767779957101449, + "learning_rate": 3.542277011818934e-06, + "loss": 0.5956, + "num_input_tokens_seen": 42579270, + "step": 2022 + }, + { + "epoch": 0.24325136776288103, + "flos": 29906060214120.0, + "grad_norm": 3.2751060002768333, + "learning_rate": 3.5417809489808104e-06, + "loss": 0.7162, + "num_input_tokens_seen": 42600600, + "step": 2023 + }, + { + "epoch": 0.24337161065352012, + "flos": 18349216059840.0, + "grad_norm": 2.2741072150045376, + "learning_rate": 3.5412846522557422e-06, + "loss": 0.706, + "num_input_tokens_seen": 42621210, + "step": 2024 + }, + { + "epoch": 0.2434918535441592, + "flos": 13623676579200.0, + "grad_norm": 2.9721518851857622, + "learning_rate": 3.540788121719018e-06, + "loss": 0.7306, + "num_input_tokens_seen": 42639350, + "step": 2025 + }, + { + "epoch": 0.24361209643479828, + "flos": 17511981009960.0, + "grad_norm": 2.372797697823167, + "learning_rate": 3.5402913574459604e-06, + "loss": 0.8079, + "num_input_tokens_seen": 42658975, + "step": 2026 + }, + { + "epoch": 0.2437323393254374, + "flos": 21021536898360.0, + "grad_norm": 2.0826363605674576, + "learning_rate": 3.5397943595119297e-06, + "loss": 0.8367, + "num_input_tokens_seen": 42680115, + "step": 2027 + }, + { + "epoch": 0.24385258221607647, + "flos": 17241998124720.0, + "grad_norm": 3.613059405140702, + "learning_rate": 3.5392971279923177e-06, + "loss": 0.7432, + "num_input_tokens_seen": 42698055, + "step": 2028 + }, + { + "epoch": 0.24397282510671556, + "flos": 18563296438440.0, + "grad_norm": 3.1593915959647725, + "learning_rate": 3.5387996629625557e-06, + "loss": 0.811, + "num_input_tokens_seen": 42715365, + "step": 2029 + }, + { + "epoch": 0.24409306799735467, + "flos": 48062376408600.0, + "grad_norm": 0.8431154620152115, + "learning_rate": 3.5383019644981083e-06, + "loss": 0.5922, + "num_input_tokens_seen": 42778780, + "step": 2030 + }, + { + "epoch": 0.24421331088799375, + "flos": 14271964767360.0, + "grad_norm": 3.2351254684431985, + "learning_rate": 3.5378040326744763e-06, + "loss": 0.6975, + "num_input_tokens_seen": 42797985, + "step": 2031 + }, + { + "epoch": 0.24433355377863283, + "flos": 15405234358200.0, + "grad_norm": 2.7484913068734165, + "learning_rate": 3.5373058675671946e-06, + "loss": 0.8423, + "num_input_tokens_seen": 42815710, + "step": 2032 + }, + { + "epoch": 0.24445379666927192, + "flos": 16566075282840.0, + "grad_norm": 2.370655100938342, + "learning_rate": 3.536807469251836e-06, + "loss": 0.7033, + "num_input_tokens_seen": 42834585, + "step": 2033 + }, + { + "epoch": 0.24457403955991103, + "flos": 15540779850120.0, + "grad_norm": 2.6623758027387017, + "learning_rate": 3.5363088378040055e-06, + "loss": 0.8032, + "num_input_tokens_seen": 42853195, + "step": 2034 + }, + { + "epoch": 0.2446942824505501, + "flos": 49400228390040.0, + "grad_norm": 0.8082028978512473, + "learning_rate": 3.5358099732993463e-06, + "loss": 0.6743, + "num_input_tokens_seen": 42912025, + "step": 2035 + }, + { + "epoch": 0.2448145253411892, + "flos": 14918574977640.0, + "grad_norm": 3.56242303597819, + "learning_rate": 3.535310875813535e-06, + "loss": 0.8704, + "num_input_tokens_seen": 42930140, + "step": 2036 + }, + { + "epoch": 0.2449347682318283, + "flos": 21130682475000.0, + "grad_norm": 2.4161044376612786, + "learning_rate": 3.5348115454222843e-06, + "loss": 0.8054, + "num_input_tokens_seen": 42952445, + "step": 2037 + }, + { + "epoch": 0.2450550111224674, + "flos": 16486179017880.0, + "grad_norm": 2.4266035618105004, + "learning_rate": 3.5343119822013425e-06, + "loss": 0.831, + "num_input_tokens_seen": 42971275, + "step": 2038 + }, + { + "epoch": 0.24517525401310647, + "flos": 15915919157040.0, + "grad_norm": 2.875191889454351, + "learning_rate": 3.533812186226493e-06, + "loss": 0.7654, + "num_input_tokens_seen": 42991705, + "step": 2039 + }, + { + "epoch": 0.24529549690374555, + "flos": 18346493303280.0, + "grad_norm": 2.7768818703273874, + "learning_rate": 3.5333121575735545e-06, + "loss": 0.7482, + "num_input_tokens_seen": 43011065, + "step": 2040 + }, + { + "epoch": 0.24541573979438466, + "flos": 23586706737720.0, + "grad_norm": 2.35587703272534, + "learning_rate": 3.532811896318381e-06, + "loss": 0.7371, + "num_input_tokens_seen": 43032855, + "step": 2041 + }, + { + "epoch": 0.24553598268502375, + "flos": 23019771172680.0, + "grad_norm": 2.5719444404581266, + "learning_rate": 3.5323114025368615e-06, + "loss": 0.7864, + "num_input_tokens_seen": 43047640, + "step": 2042 + }, + { + "epoch": 0.24565622557566283, + "flos": 10194428535240.0, + "grad_norm": 2.7080953248909627, + "learning_rate": 3.53181067630492e-06, + "loss": 0.7953, + "num_input_tokens_seen": 43064830, + "step": 2043 + }, + { + "epoch": 0.24577646846630194, + "flos": 12084562011600.0, + "grad_norm": 6.064210001668473, + "learning_rate": 3.5313097176985175e-06, + "loss": 0.7445, + "num_input_tokens_seen": 43082860, + "step": 2044 + }, + { + "epoch": 0.24589671135694102, + "flos": 13731967336920.0, + "grad_norm": 2.3457061742148166, + "learning_rate": 3.5308085267936482e-06, + "loss": 0.7875, + "num_input_tokens_seen": 43100295, + "step": 2045 + }, + { + "epoch": 0.2460169542475801, + "flos": 14569075746960.0, + "grad_norm": 2.2541144157554216, + "learning_rate": 3.530307103666342e-06, + "loss": 0.8923, + "num_input_tokens_seen": 43119095, + "step": 2046 + }, + { + "epoch": 0.24613719713821922, + "flos": 17701592730840.0, + "grad_norm": 2.741034719186736, + "learning_rate": 3.5298054483926658e-06, + "loss": 0.7799, + "num_input_tokens_seen": 43139510, + "step": 2047 + }, + { + "epoch": 0.2462574400288583, + "flos": 22177913768640.0, + "grad_norm": 2.795903724508187, + "learning_rate": 3.5293035610487187e-06, + "loss": 0.8147, + "num_input_tokens_seen": 43158595, + "step": 2048 + }, + { + "epoch": 0.24637768291949738, + "flos": 50841436666800.0, + "grad_norm": 0.6981451647381157, + "learning_rate": 3.5288014417106374e-06, + "loss": 0.6368, + "num_input_tokens_seen": 43224335, + "step": 2049 + }, + { + "epoch": 0.24649792581013646, + "flos": 25259277239880.0, + "grad_norm": 2.055895780066707, + "learning_rate": 3.528299090454593e-06, + "loss": 0.7433, + "num_input_tokens_seen": 43244590, + "step": 2050 + }, + { + "epoch": 0.24661816870077558, + "flos": 14377691068320.0, + "grad_norm": 3.2521293706932366, + "learning_rate": 3.527796507356792e-06, + "loss": 0.8085, + "num_input_tokens_seen": 43258200, + "step": 2051 + }, + { + "epoch": 0.24673841159141466, + "flos": 14614942046280.0, + "grad_norm": 3.324799644915917, + "learning_rate": 3.527293692493475e-06, + "loss": 0.8615, + "num_input_tokens_seen": 43273785, + "step": 2052 + }, + { + "epoch": 0.24685865448205374, + "flos": 15832192036920.0, + "grad_norm": 3.374259867575505, + "learning_rate": 3.52679064594092e-06, + "loss": 0.7116, + "num_input_tokens_seen": 43290845, + "step": 2053 + }, + { + "epoch": 0.24697889737269285, + "flos": 13104063671640.0, + "grad_norm": 2.648818177177945, + "learning_rate": 3.5262873677754375e-06, + "loss": 0.7245, + "num_input_tokens_seen": 43308570, + "step": 2054 + }, + { + "epoch": 0.24709914026333193, + "flos": 20049991095000.0, + "grad_norm": 1.9998937817024285, + "learning_rate": 3.5257838580733745e-06, + "loss": 0.7893, + "num_input_tokens_seen": 43327895, + "step": 2055 + }, + { + "epoch": 0.24721938315397102, + "flos": 14077509072600.0, + "grad_norm": 3.41560171944798, + "learning_rate": 3.5252801169111138e-06, + "loss": 0.8486, + "num_input_tokens_seen": 43345280, + "step": 2056 + }, + { + "epoch": 0.2473396260446101, + "flos": 16972268519160.0, + "grad_norm": 2.2581093140961883, + "learning_rate": 3.524776144365072e-06, + "loss": 0.7759, + "num_input_tokens_seen": 43363455, + "step": 2057 + }, + { + "epoch": 0.2474598689352492, + "flos": 15460028766240.0, + "grad_norm": 2.4525169056223204, + "learning_rate": 3.5242719405117016e-06, + "loss": 0.7862, + "num_input_tokens_seen": 43382980, + "step": 2058 + }, + { + "epoch": 0.2475801118258883, + "flos": 15833743374960.0, + "grad_norm": 4.661728471426515, + "learning_rate": 3.5237675054274893e-06, + "loss": 0.7187, + "num_input_tokens_seen": 43401900, + "step": 2059 + }, + { + "epoch": 0.24770035471652738, + "flos": 16593804916440.0, + "grad_norm": 5.022827301221931, + "learning_rate": 3.5232628391889584e-06, + "loss": 0.7861, + "num_input_tokens_seen": 43419910, + "step": 2060 + }, + { + "epoch": 0.2478205976071665, + "flos": 16215879533040.0, + "grad_norm": 3.2836717660309485, + "learning_rate": 3.522757941872666e-06, + "loss": 0.6269, + "num_input_tokens_seen": 43437785, + "step": 2061 + }, + { + "epoch": 0.24794084049780557, + "flos": 18294928211160.0, + "grad_norm": 1.8931919673184276, + "learning_rate": 3.5222528135552042e-06, + "loss": 0.8121, + "num_input_tokens_seen": 43458965, + "step": 2062 + }, + { + "epoch": 0.24806108338844465, + "flos": 13353725353920.0, + "grad_norm": 3.146570300415445, + "learning_rate": 3.521747454313201e-06, + "loss": 0.7928, + "num_input_tokens_seen": 43477365, + "step": 2063 + }, + { + "epoch": 0.24818132627908374, + "flos": 14083366165200.0, + "grad_norm": 3.4398193857550443, + "learning_rate": 3.521241864223319e-06, + "loss": 0.6567, + "num_input_tokens_seen": 43496045, + "step": 2064 + }, + { + "epoch": 0.24830156916972285, + "flos": 51834190152000.0, + "grad_norm": 0.80807338316684, + "learning_rate": 3.5207360433622552e-06, + "loss": 0.6229, + "num_input_tokens_seen": 43557765, + "step": 2065 + }, + { + "epoch": 0.24842181206036193, + "flos": 29718981290040.0, + "grad_norm": 2.1675945887006534, + "learning_rate": 3.5202299918067437e-06, + "loss": 0.7324, + "num_input_tokens_seen": 43581080, + "step": 2066 + }, + { + "epoch": 0.248542054951001, + "flos": 14675276947320.0, + "grad_norm": 2.748661322314704, + "learning_rate": 3.519723709633551e-06, + "loss": 0.6835, + "num_input_tokens_seen": 43599560, + "step": 2067 + }, + { + "epoch": 0.24866229784164012, + "flos": 17215091650080.0, + "grad_norm": 2.5109022839491417, + "learning_rate": 3.519217196919479e-06, + "loss": 0.8074, + "num_input_tokens_seen": 43618265, + "step": 2068 + }, + { + "epoch": 0.2487825407322792, + "flos": 14514661258440.0, + "grad_norm": 2.1109546539264734, + "learning_rate": 3.518710453741367e-06, + "loss": 0.7135, + "num_input_tokens_seen": 43637185, + "step": 2069 + }, + { + "epoch": 0.2489027836229183, + "flos": 16107652095240.0, + "grad_norm": 2.5198942174356107, + "learning_rate": 3.518203480176086e-06, + "loss": 0.656, + "num_input_tokens_seen": 43655835, + "step": 2070 + }, + { + "epoch": 0.2490230265135574, + "flos": 17052481383720.0, + "grad_norm": 2.367363750077944, + "learning_rate": 3.517696276300545e-06, + "loss": 0.7787, + "num_input_tokens_seen": 43677095, + "step": 2071 + }, + { + "epoch": 0.24914326940419648, + "flos": 14486045145960.0, + "grad_norm": 7.096956712999005, + "learning_rate": 3.517188842191685e-06, + "loss": 0.6808, + "num_input_tokens_seen": 43694965, + "step": 2072 + }, + { + "epoch": 0.24926351229483557, + "flos": 14784074264400.0, + "grad_norm": 2.229161402140548, + "learning_rate": 3.5166811779264837e-06, + "loss": 0.7221, + "num_input_tokens_seen": 43715005, + "step": 2073 + }, + { + "epoch": 0.24938375518547465, + "flos": 17052671343480.0, + "grad_norm": 3.3675039899387524, + "learning_rate": 3.5161732835819545e-06, + "loss": 0.7609, + "num_input_tokens_seen": 43734035, + "step": 2074 + }, + { + "epoch": 0.24950399807611376, + "flos": 12625034341440.0, + "grad_norm": 4.269128722876629, + "learning_rate": 3.515665159235143e-06, + "loss": 0.8049, + "num_input_tokens_seen": 43752640, + "step": 2075 + }, + { + "epoch": 0.24962424096675284, + "flos": 13892678005680.0, + "grad_norm": 2.254012450785465, + "learning_rate": 3.5151568049631318e-06, + "loss": 0.7298, + "num_input_tokens_seen": 43771075, + "step": 2076 + }, + { + "epoch": 0.24974448385739192, + "flos": 24530142987960.0, + "grad_norm": 2.549712770778399, + "learning_rate": 3.5146482208430385e-06, + "loss": 0.781, + "num_input_tokens_seen": 43792625, + "step": 2077 + }, + { + "epoch": 0.24986472674803104, + "flos": 22020432415800.0, + "grad_norm": 3.1569583211328003, + "learning_rate": 3.514139406952014e-06, + "loss": 0.6662, + "num_input_tokens_seen": 43814370, + "step": 2078 + }, + { + "epoch": 0.24998496963867012, + "flos": 19508347346640.0, + "grad_norm": 3.202477515303261, + "learning_rate": 3.5136303633672454e-06, + "loss": 0.8283, + "num_input_tokens_seen": 43834220, + "step": 2079 + }, + { + "epoch": 0.25010521252930923, + "flos": 17245227440640.0, + "grad_norm": 2.3363103154429563, + "learning_rate": 3.5131210901659544e-06, + "loss": 0.7382, + "num_input_tokens_seen": 43855695, + "step": 2080 + }, + { + "epoch": 0.2502254554199483, + "flos": 17161278700800.0, + "grad_norm": 3.3041169235378427, + "learning_rate": 3.5126115874253967e-06, + "loss": 0.7995, + "num_input_tokens_seen": 43874970, + "step": 2081 + }, + { + "epoch": 0.2503456983105874, + "flos": 21097887247800.0, + "grad_norm": 2.1912527319569595, + "learning_rate": 3.5121018552228644e-06, + "loss": 0.7885, + "num_input_tokens_seen": 43893195, + "step": 2082 + }, + { + "epoch": 0.2504659412012265, + "flos": 13704364343160.0, + "grad_norm": 3.4891521737874522, + "learning_rate": 3.5115918936356827e-06, + "loss": 0.7583, + "num_input_tokens_seen": 43909670, + "step": 2083 + }, + { + "epoch": 0.25058618409186556, + "flos": 12300605307720.0, + "grad_norm": 2.3809431904660823, + "learning_rate": 3.5110817027412123e-06, + "loss": 0.7673, + "num_input_tokens_seen": 43928480, + "step": 2084 + }, + { + "epoch": 0.25070642698250467, + "flos": 17889653113680.0, + "grad_norm": 2.6786225541094875, + "learning_rate": 3.5105712826168493e-06, + "loss": 0.6655, + "num_input_tokens_seen": 43947850, + "step": 2085 + }, + { + "epoch": 0.2508266698731437, + "flos": 14806833284280.0, + "grad_norm": 2.718311230561635, + "learning_rate": 3.5100606333400235e-06, + "loss": 0.6926, + "num_input_tokens_seen": 43964705, + "step": 2086 + }, + { + "epoch": 0.25094691276378284, + "flos": 14056048111080.0, + "grad_norm": 3.180239617778707, + "learning_rate": 3.5095497549882006e-06, + "loss": 0.7576, + "num_input_tokens_seen": 43982870, + "step": 2087 + }, + { + "epoch": 0.25106715565442195, + "flos": 19752943435320.0, + "grad_norm": 3.4175841022403866, + "learning_rate": 3.50903864763888e-06, + "loss": 0.6998, + "num_input_tokens_seen": 44003380, + "step": 2088 + }, + { + "epoch": 0.251187398545061, + "flos": 35602353999120.0, + "grad_norm": 2.8693325739378555, + "learning_rate": 3.5085273113695965e-06, + "loss": 0.7377, + "num_input_tokens_seen": 44027670, + "step": 2089 + }, + { + "epoch": 0.2513076414357001, + "flos": 19806851364480.0, + "grad_norm": 3.4820150414787805, + "learning_rate": 3.508015746257919e-06, + "loss": 0.7692, + "num_input_tokens_seen": 44046430, + "step": 2090 + }, + { + "epoch": 0.2514278843263392, + "flos": 14217518618880.0, + "grad_norm": 2.801011080070679, + "learning_rate": 3.5075039523814518e-06, + "loss": 0.8136, + "num_input_tokens_seen": 44065340, + "step": 2091 + }, + { + "epoch": 0.2515481272169783, + "flos": 12294178335840.0, + "grad_norm": 4.226561371808422, + "learning_rate": 3.506991929817834e-06, + "loss": 0.8016, + "num_input_tokens_seen": 44081780, + "step": 2092 + }, + { + "epoch": 0.2516683701076174, + "flos": 17376752117640.0, + "grad_norm": 2.0336793896891385, + "learning_rate": 3.506479678644738e-06, + "loss": 0.8074, + "num_input_tokens_seen": 44101895, + "step": 2093 + }, + { + "epoch": 0.2517886129982565, + "flos": 20265527831760.0, + "grad_norm": 3.269073417812357, + "learning_rate": 3.505967198939873e-06, + "loss": 0.7196, + "num_input_tokens_seen": 44118655, + "step": 2094 + }, + { + "epoch": 0.25190885588889556, + "flos": 28012634101920.0, + "grad_norm": 3.001480055336535, + "learning_rate": 3.5054544907809813e-06, + "loss": 0.7585, + "num_input_tokens_seen": 44138875, + "step": 2095 + }, + { + "epoch": 0.25202909877953467, + "flos": 16293907860360.0, + "grad_norm": 2.700282106849854, + "learning_rate": 3.50494155424584e-06, + "loss": 0.7835, + "num_input_tokens_seen": 44157500, + "step": 2096 + }, + { + "epoch": 0.2521493416701738, + "flos": 15918230334120.0, + "grad_norm": 1.9984010026475698, + "learning_rate": 3.504428389412262e-06, + "loss": 0.8206, + "num_input_tokens_seen": 44178030, + "step": 2097 + }, + { + "epoch": 0.25226958456081283, + "flos": 20347703613840.0, + "grad_norm": 9.074245654824514, + "learning_rate": 3.5039149963580927e-06, + "loss": 0.7168, + "num_input_tokens_seen": 44197770, + "step": 2098 + }, + { + "epoch": 0.25238982745145194, + "flos": 22556219071560.0, + "grad_norm": 3.82994764884847, + "learning_rate": 3.503401375161215e-06, + "loss": 0.6732, + "num_input_tokens_seen": 44217235, + "step": 2099 + }, + { + "epoch": 0.252510070342091, + "flos": 14811582278280.0, + "grad_norm": 3.2884757272857823, + "learning_rate": 3.502887525899544e-06, + "loss": 0.8335, + "num_input_tokens_seen": 44235935, + "step": 2100 + }, + { + "epoch": 0.2526303132327301, + "flos": 16647744505560.0, + "grad_norm": 2.0126255399038566, + "learning_rate": 3.50237344865103e-06, + "loss": 0.8143, + "num_input_tokens_seen": 44256655, + "step": 2101 + }, + { + "epoch": 0.2527505561233692, + "flos": 22210139116560.0, + "grad_norm": 3.047287820090462, + "learning_rate": 3.501859143493658e-06, + "loss": 0.7483, + "num_input_tokens_seen": 44277005, + "step": 2102 + }, + { + "epoch": 0.2528707990140083, + "flos": 43105501871160.0, + "grad_norm": 0.850590141977984, + "learning_rate": 3.5013446105054488e-06, + "loss": 0.6398, + "num_input_tokens_seen": 44329645, + "step": 2103 + }, + { + "epoch": 0.2529910419046474, + "flos": 18052263380040.0, + "grad_norm": 3.2387341595596926, + "learning_rate": 3.5008298497644555e-06, + "loss": 0.7421, + "num_input_tokens_seen": 44348410, + "step": 2104 + }, + { + "epoch": 0.2531112847952865, + "flos": 17457946440960.0, + "grad_norm": 2.318343743365583, + "learning_rate": 3.500314861348767e-06, + "loss": 0.8692, + "num_input_tokens_seen": 44368765, + "step": 2105 + }, + { + "epoch": 0.25323152768592555, + "flos": 11760702857160.0, + "grad_norm": 4.003989852404621, + "learning_rate": 3.499799645336507e-06, + "loss": 0.7589, + "num_input_tokens_seen": 44385380, + "step": 2106 + }, + { + "epoch": 0.25335177057656466, + "flos": 20835217813320.0, + "grad_norm": 1.70272409462675, + "learning_rate": 3.4992842018058336e-06, + "loss": 0.8577, + "num_input_tokens_seen": 44408000, + "step": 2107 + }, + { + "epoch": 0.25347201346720377, + "flos": 13725128785560.0, + "grad_norm": 2.541894172950698, + "learning_rate": 3.4987685308349384e-06, + "loss": 0.8572, + "num_input_tokens_seen": 44425450, + "step": 2108 + }, + { + "epoch": 0.2535922563578428, + "flos": 11517721426440.0, + "grad_norm": 3.1948801508290763, + "learning_rate": 3.4982526325020497e-06, + "loss": 0.6033, + "num_input_tokens_seen": 44442140, + "step": 2109 + }, + { + "epoch": 0.25371249924848194, + "flos": 11889251497920.0, + "grad_norm": 2.708862653994171, + "learning_rate": 3.4977365068854273e-06, + "loss": 0.8069, + "num_input_tokens_seen": 44457480, + "step": 2110 + }, + { + "epoch": 0.25383274213912105, + "flos": 15918103694280.0, + "grad_norm": 2.973248576879261, + "learning_rate": 3.4972201540633676e-06, + "loss": 0.707, + "num_input_tokens_seen": 44476555, + "step": 2111 + }, + { + "epoch": 0.2539529850297601, + "flos": 15648184128960.0, + "grad_norm": 2.113941928444856, + "learning_rate": 3.4967035741142008e-06, + "loss": 0.832, + "num_input_tokens_seen": 44495095, + "step": 2112 + }, + { + "epoch": 0.2540732279203992, + "flos": 18483906732840.0, + "grad_norm": 3.2870044929513145, + "learning_rate": 3.4961867671162917e-06, + "loss": 0.8075, + "num_input_tokens_seen": 44514745, + "step": 2113 + }, + { + "epoch": 0.2541934708110383, + "flos": 14190738784080.0, + "grad_norm": 4.210666400059826, + "learning_rate": 3.4956697331480402e-06, + "loss": 0.7447, + "num_input_tokens_seen": 44533035, + "step": 2114 + }, + { + "epoch": 0.2543137137016774, + "flos": 17537051206920.0, + "grad_norm": 1.917771774090804, + "learning_rate": 3.495152472287879e-06, + "loss": 0.7878, + "num_input_tokens_seen": 44553465, + "step": 2115 + }, + { + "epoch": 0.2544339565923165, + "flos": 18753889618080.0, + "grad_norm": 2.3488153271728516, + "learning_rate": 3.4946349846142766e-06, + "loss": 0.7108, + "num_input_tokens_seen": 44572325, + "step": 2116 + }, + { + "epoch": 0.25455419948295555, + "flos": 15864828964320.0, + "grad_norm": 2.5214703808432346, + "learning_rate": 3.4941172702057353e-06, + "loss": 0.7414, + "num_input_tokens_seen": 44592105, + "step": 2117 + }, + { + "epoch": 0.25467444237359466, + "flos": 19240200739080.0, + "grad_norm": 2.3711409084903114, + "learning_rate": 3.4935993291407924e-06, + "loss": 0.7871, + "num_input_tokens_seen": 44610650, + "step": 2118 + }, + { + "epoch": 0.25479468526423377, + "flos": 19779913229880.0, + "grad_norm": 2.630703769264939, + "learning_rate": 3.4930811614980183e-06, + "loss": 0.6902, + "num_input_tokens_seen": 44632065, + "step": 2119 + }, + { + "epoch": 0.2549149281548728, + "flos": 17186633837400.0, + "grad_norm": 2.761931458392727, + "learning_rate": 3.4925627673560198e-06, + "loss": 0.7788, + "num_input_tokens_seen": 44652445, + "step": 2120 + }, + { + "epoch": 0.25503517104551193, + "flos": 18915993325080.0, + "grad_norm": 2.7660079949484775, + "learning_rate": 3.4920441467934357e-06, + "loss": 0.8653, + "num_input_tokens_seen": 44672680, + "step": 2121 + }, + { + "epoch": 0.25515541393615104, + "flos": 19531422966120.0, + "grad_norm": 2.442749825608479, + "learning_rate": 3.491525299888941e-06, + "loss": 0.802, + "num_input_tokens_seen": 44691245, + "step": 2122 + }, + { + "epoch": 0.2552756568267901, + "flos": 48631464850920.0, + "grad_norm": 0.9939772316111495, + "learning_rate": 3.491006226721244e-06, + "loss": 0.6791, + "num_input_tokens_seen": 44755175, + "step": 2123 + }, + { + "epoch": 0.2553958997174292, + "flos": 13083330889200.0, + "grad_norm": 2.599876430215978, + "learning_rate": 3.4904869273690882e-06, + "loss": 0.7527, + "num_input_tokens_seen": 44772785, + "step": 2124 + }, + { + "epoch": 0.2555161426080683, + "flos": 17106895872240.0, + "grad_norm": 2.5377760224949606, + "learning_rate": 3.489967401911251e-06, + "loss": 0.874, + "num_input_tokens_seen": 44791805, + "step": 2125 + }, + { + "epoch": 0.2556363854987074, + "flos": 29878425560400.0, + "grad_norm": 2.177065458330819, + "learning_rate": 3.4894476504265428e-06, + "loss": 0.6813, + "num_input_tokens_seen": 44815765, + "step": 2126 + }, + { + "epoch": 0.2557566283893465, + "flos": 50156300776560.0, + "grad_norm": 0.7665260617926096, + "learning_rate": 3.4889276729938104e-06, + "loss": 0.5649, + "num_input_tokens_seen": 44874015, + "step": 2127 + }, + { + "epoch": 0.2558768712799856, + "flos": 16564713904560.0, + "grad_norm": 4.076891014515629, + "learning_rate": 3.488407469691934e-06, + "loss": 0.7814, + "num_input_tokens_seen": 44894430, + "step": 2128 + }, + { + "epoch": 0.25599711417062465, + "flos": 19348269877080.0, + "grad_norm": 3.1407057718546842, + "learning_rate": 3.487887040599828e-06, + "loss": 0.7805, + "num_input_tokens_seen": 44913950, + "step": 2129 + }, + { + "epoch": 0.25611735706126376, + "flos": 16724506434480.0, + "grad_norm": 3.039064853539818, + "learning_rate": 3.4873663857964407e-06, + "loss": 0.7441, + "num_input_tokens_seen": 44930885, + "step": 2130 + }, + { + "epoch": 0.2562375999519028, + "flos": 17106959192160.0, + "grad_norm": 2.4908648481914, + "learning_rate": 3.4868455053607556e-06, + "loss": 0.6525, + "num_input_tokens_seen": 44950220, + "step": 2131 + }, + { + "epoch": 0.2563578428425419, + "flos": 16728843849000.0, + "grad_norm": 3.0330171448235266, + "learning_rate": 3.486324399371789e-06, + "loss": 0.6911, + "num_input_tokens_seen": 44969240, + "step": 2132 + }, + { + "epoch": 0.25647808573318104, + "flos": 15838619008800.0, + "grad_norm": 6.004985645286561, + "learning_rate": 3.485803067908593e-06, + "loss": 0.7645, + "num_input_tokens_seen": 44988470, + "step": 2133 + }, + { + "epoch": 0.2565983286238201, + "flos": 24776163774840.0, + "grad_norm": 2.395507215335739, + "learning_rate": 3.485281511050253e-06, + "loss": 0.778, + "num_input_tokens_seen": 45010325, + "step": 2134 + }, + { + "epoch": 0.2567185715144592, + "flos": 11814009247080.0, + "grad_norm": 2.961715233754672, + "learning_rate": 3.484759728875889e-06, + "loss": 0.8856, + "num_input_tokens_seen": 45025410, + "step": 2135 + }, + { + "epoch": 0.2568388144050983, + "flos": 12733799998560.0, + "grad_norm": 2.143034866319944, + "learning_rate": 3.4842377214646543e-06, + "loss": 0.7925, + "num_input_tokens_seen": 45043425, + "step": 2136 + }, + { + "epoch": 0.25695905729573737, + "flos": 15271366844160.0, + "grad_norm": 3.402176126934295, + "learning_rate": 3.483715488895737e-06, + "loss": 0.6617, + "num_input_tokens_seen": 45063475, + "step": 2137 + }, + { + "epoch": 0.2570793001863765, + "flos": 18105538110000.0, + "grad_norm": 2.303144966773927, + "learning_rate": 3.48319303124836e-06, + "loss": 0.7656, + "num_input_tokens_seen": 45083575, + "step": 2138 + }, + { + "epoch": 0.2571995430770156, + "flos": 19726068620640.0, + "grad_norm": 2.837322811022347, + "learning_rate": 3.4826703486017798e-06, + "loss": 0.6601, + "num_input_tokens_seen": 45102920, + "step": 2139 + }, + { + "epoch": 0.25731978596765465, + "flos": 14461038268920.0, + "grad_norm": 2.534891123493293, + "learning_rate": 3.4821474410352867e-06, + "loss": 0.7483, + "num_input_tokens_seen": 45121300, + "step": 2140 + }, + { + "epoch": 0.25744002885829376, + "flos": 52038962502360.0, + "grad_norm": 0.9020178400335043, + "learning_rate": 3.481624308628205e-06, + "loss": 0.6632, + "num_input_tokens_seen": 45182390, + "step": 2141 + }, + { + "epoch": 0.25756027174893287, + "flos": 13162245695400.0, + "grad_norm": 5.231483138892787, + "learning_rate": 3.481100951459893e-06, + "loss": 0.9856, + "num_input_tokens_seen": 45195130, + "step": 2142 + }, + { + "epoch": 0.2576805146395719, + "flos": 16593424996920.0, + "grad_norm": 1.953321035083858, + "learning_rate": 3.4805773696097453e-06, + "loss": 0.7665, + "num_input_tokens_seen": 45215740, + "step": 2143 + }, + { + "epoch": 0.25780075753021103, + "flos": 12003747607800.0, + "grad_norm": 2.403057705615355, + "learning_rate": 3.4800535631571874e-06, + "loss": 0.856, + "num_input_tokens_seen": 45230990, + "step": 2144 + }, + { + "epoch": 0.25792100042085014, + "flos": 16701684094680.0, + "grad_norm": 2.760312425096671, + "learning_rate": 3.4795295321816804e-06, + "loss": 0.7525, + "num_input_tokens_seen": 45249535, + "step": 2145 + }, + { + "epoch": 0.2580412433114892, + "flos": 13650361434120.0, + "grad_norm": 3.421039357333357, + "learning_rate": 3.47900527676272e-06, + "loss": 0.9028, + "num_input_tokens_seen": 45267590, + "step": 2146 + }, + { + "epoch": 0.2581614862021283, + "flos": 10383691996560.0, + "grad_norm": 2.505299863246944, + "learning_rate": 3.478480796979835e-06, + "loss": 0.8713, + "num_input_tokens_seen": 45285195, + "step": 2147 + }, + { + "epoch": 0.25828172909276736, + "flos": 21643995050520.0, + "grad_norm": 1.877043373566037, + "learning_rate": 3.4779560929125894e-06, + "loss": 0.7604, + "num_input_tokens_seen": 45306460, + "step": 2148 + }, + { + "epoch": 0.2584019719834065, + "flos": 49487501425680.0, + "grad_norm": 0.6924915524253239, + "learning_rate": 3.4774311646405783e-06, + "loss": 0.5771, + "num_input_tokens_seen": 45376085, + "step": 2149 + }, + { + "epoch": 0.2585222148740456, + "flos": 16755940283400.0, + "grad_norm": 2.1053978619646996, + "learning_rate": 3.476906012243435e-06, + "loss": 0.8173, + "num_input_tokens_seen": 45394715, + "step": 2150 + }, + { + "epoch": 0.25864245776468464, + "flos": 21207286104120.0, + "grad_norm": 2.850701188949266, + "learning_rate": 3.476380635800824e-06, + "loss": 0.8018, + "num_input_tokens_seen": 45415635, + "step": 2151 + }, + { + "epoch": 0.25876270065532375, + "flos": 10761047500680.0, + "grad_norm": 3.515706108005362, + "learning_rate": 3.475855035392444e-06, + "loss": 0.8371, + "num_input_tokens_seen": 45430675, + "step": 2152 + }, + { + "epoch": 0.25888294354596286, + "flos": 44562562805160.0, + "grad_norm": 2.5886351007578274, + "learning_rate": 3.475329211098029e-06, + "loss": 0.6894, + "num_input_tokens_seen": 45453550, + "step": 2153 + }, + { + "epoch": 0.2590031864366019, + "flos": 20424275583000.0, + "grad_norm": 2.2607387073069867, + "learning_rate": 3.4748031629973453e-06, + "loss": 0.8132, + "num_input_tokens_seen": 45474000, + "step": 2154 + }, + { + "epoch": 0.25912342932724103, + "flos": 46017231056280.0, + "grad_norm": 0.9514728673052204, + "learning_rate": 3.4742768911701944e-06, + "loss": 0.566, + "num_input_tokens_seen": 45536415, + "step": 2155 + }, + { + "epoch": 0.25924367221788014, + "flos": 8973379349400.0, + "grad_norm": 4.864155175306192, + "learning_rate": 3.4737503956964113e-06, + "loss": 0.677, + "num_input_tokens_seen": 45548440, + "step": 2156 + }, + { + "epoch": 0.2593639151085192, + "flos": 10599861932520.0, + "grad_norm": 3.0145584354984987, + "learning_rate": 3.473223676655865e-06, + "loss": 0.6486, + "num_input_tokens_seen": 45566160, + "step": 2157 + }, + { + "epoch": 0.2594841579991583, + "flos": 11085001635000.0, + "grad_norm": 3.067535153710628, + "learning_rate": 3.472696734128459e-06, + "loss": 0.7785, + "num_input_tokens_seen": 45583745, + "step": 2158 + }, + { + "epoch": 0.2596044008897974, + "flos": 17296254313440.0, + "grad_norm": 2.3143078722221033, + "learning_rate": 3.4721695681941286e-06, + "loss": 0.7369, + "num_input_tokens_seen": 45602505, + "step": 2159 + }, + { + "epoch": 0.25972464378043647, + "flos": 10005608313360.0, + "grad_norm": 4.5155225695968, + "learning_rate": 3.471642178932845e-06, + "loss": 0.8166, + "num_input_tokens_seen": 45620870, + "step": 2160 + }, + { + "epoch": 0.2598448866710756, + "flos": 14298966221880.0, + "grad_norm": 3.353007114183278, + "learning_rate": 3.471114566424613e-06, + "loss": 0.8904, + "num_input_tokens_seen": 45639050, + "step": 2161 + }, + { + "epoch": 0.25996512956171464, + "flos": 15838049129520.0, + "grad_norm": 3.16403281115508, + "learning_rate": 3.4705867307494715e-06, + "loss": 0.7457, + "num_input_tokens_seen": 45657840, + "step": 2162 + }, + { + "epoch": 0.26008537245235375, + "flos": 13299785764800.0, + "grad_norm": 3.689807622386588, + "learning_rate": 3.470058671987492e-06, + "loss": 0.8096, + "num_input_tokens_seen": 45675825, + "step": 2163 + }, + { + "epoch": 0.26020561534299286, + "flos": 18052484999760.0, + "grad_norm": 2.3090848706930363, + "learning_rate": 3.4695303902187805e-06, + "loss": 0.8199, + "num_input_tokens_seen": 45695100, + "step": 2164 + }, + { + "epoch": 0.2603258582336319, + "flos": 18887092272960.0, + "grad_norm": 2.7148866060016696, + "learning_rate": 3.469001885523478e-06, + "loss": 0.7701, + "num_input_tokens_seen": 45715540, + "step": 2165 + }, + { + "epoch": 0.260446101124271, + "flos": 21101971382640.0, + "grad_norm": 1.8355282591179989, + "learning_rate": 3.4684731579817568e-06, + "loss": 0.7984, + "num_input_tokens_seen": 45736250, + "step": 2166 + }, + { + "epoch": 0.26056634401491013, + "flos": 18808335766560.0, + "grad_norm": 3.2933039556832653, + "learning_rate": 3.4679442076738247e-06, + "loss": 0.7571, + "num_input_tokens_seen": 45755685, + "step": 2167 + }, + { + "epoch": 0.2606865869055492, + "flos": 19834011118800.0, + "grad_norm": 3.527581934528179, + "learning_rate": 3.4674150346799245e-06, + "loss": 0.8231, + "num_input_tokens_seen": 45775105, + "step": 2168 + }, + { + "epoch": 0.2608068297961883, + "flos": 12921702081600.0, + "grad_norm": 4.08469933737574, + "learning_rate": 3.4668856390803295e-06, + "loss": 0.7713, + "num_input_tokens_seen": 45792705, + "step": 2169 + }, + { + "epoch": 0.2609270726868274, + "flos": 13542608895720.0, + "grad_norm": 2.945263444810732, + "learning_rate": 3.4663560209553495e-06, + "loss": 0.8866, + "num_input_tokens_seen": 45810490, + "step": 2170 + }, + { + "epoch": 0.26104731557746647, + "flos": 15973088062080.0, + "grad_norm": 2.5836052187517127, + "learning_rate": 3.4658261803853267e-06, + "loss": 0.7658, + "num_input_tokens_seen": 45828135, + "step": 2171 + }, + { + "epoch": 0.2611675584681056, + "flos": 15864987264120.0, + "grad_norm": 2.4905191894416903, + "learning_rate": 3.4652961174506383e-06, + "loss": 0.7978, + "num_input_tokens_seen": 45847725, + "step": 2172 + }, + { + "epoch": 0.2612878013587447, + "flos": 53001774648120.0, + "grad_norm": 1.0650856308696028, + "learning_rate": 3.464765832231694e-06, + "loss": 0.6006, + "num_input_tokens_seen": 45901610, + "step": 2173 + }, + { + "epoch": 0.26140804424938374, + "flos": 14782522926360.0, + "grad_norm": 2.603947551410048, + "learning_rate": 3.4642353248089373e-06, + "loss": 0.6916, + "num_input_tokens_seen": 45920090, + "step": 2174 + }, + { + "epoch": 0.26152828714002285, + "flos": 18727141443240.0, + "grad_norm": 2.4829266242847434, + "learning_rate": 3.463704595262846e-06, + "loss": 0.7887, + "num_input_tokens_seen": 45940690, + "step": 2175 + }, + { + "epoch": 0.26164853003066196, + "flos": 18645155620920.0, + "grad_norm": 2.5879201059641534, + "learning_rate": 3.463173643673931e-06, + "loss": 0.6852, + "num_input_tokens_seen": 45962935, + "step": 2176 + }, + { + "epoch": 0.261768772921301, + "flos": 39741179422440.0, + "grad_norm": 1.0570640632177115, + "learning_rate": 3.4626424701227387e-06, + "loss": 0.6559, + "num_input_tokens_seen": 46017715, + "step": 2177 + }, + { + "epoch": 0.26188901581194013, + "flos": 52131586071240.0, + "grad_norm": 0.9511168227037861, + "learning_rate": 3.4621110746898452e-06, + "loss": 0.6124, + "num_input_tokens_seen": 46085295, + "step": 2178 + }, + { + "epoch": 0.2620092587025792, + "flos": 15406120837080.0, + "grad_norm": 3.113009221961224, + "learning_rate": 3.4615794574558654e-06, + "loss": 0.7292, + "num_input_tokens_seen": 46104025, + "step": 2179 + }, + { + "epoch": 0.2621295015932183, + "flos": 13407949882680.0, + "grad_norm": 3.594154534169437, + "learning_rate": 3.4610476185014436e-06, + "loss": 0.8043, + "num_input_tokens_seen": 46121005, + "step": 2180 + }, + { + "epoch": 0.2622497444838574, + "flos": 17323192448040.0, + "grad_norm": 3.5408759786566977, + "learning_rate": 3.4605155579072597e-06, + "loss": 0.7835, + "num_input_tokens_seen": 46140580, + "step": 2181 + }, + { + "epoch": 0.26236998737449646, + "flos": 16185775402440.0, + "grad_norm": 2.2962944698440415, + "learning_rate": 3.459983275754027e-06, + "loss": 0.7029, + "num_input_tokens_seen": 46159195, + "step": 2182 + }, + { + "epoch": 0.26249023026513557, + "flos": 13056835994040.0, + "grad_norm": 4.346572727249206, + "learning_rate": 3.4594507721224918e-06, + "loss": 0.7728, + "num_input_tokens_seen": 46177565, + "step": 2183 + }, + { + "epoch": 0.2626104731557747, + "flos": 13380790128360.0, + "grad_norm": 3.104153797370467, + "learning_rate": 3.4589180470934353e-06, + "loss": 0.8002, + "num_input_tokens_seen": 46197150, + "step": 2184 + }, + { + "epoch": 0.26273071604641374, + "flos": 14109544460760.0, + "grad_norm": 3.26619514708331, + "learning_rate": 3.4583851007476713e-06, + "loss": 0.7457, + "num_input_tokens_seen": 46215340, + "step": 2185 + }, + { + "epoch": 0.26285095893705285, + "flos": 13375977814440.0, + "grad_norm": 3.012872101528855, + "learning_rate": 3.4578519331660464e-06, + "loss": 0.6618, + "num_input_tokens_seen": 46232055, + "step": 2186 + }, + { + "epoch": 0.26297120182769196, + "flos": 14757990948720.0, + "grad_norm": 2.833239909109068, + "learning_rate": 3.4573185444294426e-06, + "loss": 0.8072, + "num_input_tokens_seen": 46250140, + "step": 2187 + }, + { + "epoch": 0.263091444718331, + "flos": 16404193195560.0, + "grad_norm": 2.6264649799132007, + "learning_rate": 3.456784934618774e-06, + "loss": 0.7703, + "num_input_tokens_seen": 46271025, + "step": 2188 + }, + { + "epoch": 0.2632116876089701, + "flos": 13892646345720.0, + "grad_norm": 5.699848226567809, + "learning_rate": 3.4562511038149897e-06, + "loss": 0.7834, + "num_input_tokens_seen": 46286240, + "step": 2189 + }, + { + "epoch": 0.26333193049960923, + "flos": 49632418265760.0, + "grad_norm": 0.8803202804313686, + "learning_rate": 3.4557170520990705e-06, + "loss": 0.595, + "num_input_tokens_seen": 46346635, + "step": 2190 + }, + { + "epoch": 0.2634521733902483, + "flos": 18351337277160.0, + "grad_norm": 2.0530820891426385, + "learning_rate": 3.4551827795520324e-06, + "loss": 0.8495, + "num_input_tokens_seen": 46369240, + "step": 2191 + }, + { + "epoch": 0.2635724162808874, + "flos": 15054848648640.0, + "grad_norm": 2.636576602071115, + "learning_rate": 3.4546482862549226e-06, + "loss": 0.8343, + "num_input_tokens_seen": 46389275, + "step": 2192 + }, + { + "epoch": 0.2636926591715265, + "flos": 14055446571840.0, + "grad_norm": 8.313508166218709, + "learning_rate": 3.4541135722888253e-06, + "loss": 0.7765, + "num_input_tokens_seen": 46405585, + "step": 2193 + }, + { + "epoch": 0.26381290206216557, + "flos": 21130809114840.0, + "grad_norm": 2.919222846928493, + "learning_rate": 3.453578637734854e-06, + "loss": 0.7869, + "num_input_tokens_seen": 46426495, + "step": 2194 + }, + { + "epoch": 0.2639331449528047, + "flos": 18321328126440.0, + "grad_norm": 3.091727934964578, + "learning_rate": 3.4530434826741605e-06, + "loss": 0.7695, + "num_input_tokens_seen": 46447155, + "step": 2195 + }, + { + "epoch": 0.26405338784344373, + "flos": 34252692852600.0, + "grad_norm": 8.068884871912095, + "learning_rate": 3.452508107187926e-06, + "loss": 0.6743, + "num_input_tokens_seen": 46470250, + "step": 2196 + }, + { + "epoch": 0.26417363073408284, + "flos": 15487505120160.0, + "grad_norm": 2.5033781116542637, + "learning_rate": 3.451972511357366e-06, + "loss": 0.7429, + "num_input_tokens_seen": 46489515, + "step": 2197 + }, + { + "epoch": 0.26429387362472195, + "flos": 16594026536160.0, + "grad_norm": 2.5929545825963145, + "learning_rate": 3.45143669526373e-06, + "loss": 0.841, + "num_input_tokens_seen": 46508995, + "step": 2198 + }, + { + "epoch": 0.264414116515361, + "flos": 49535488942320.0, + "grad_norm": 0.740706240548159, + "learning_rate": 3.450900658988302e-06, + "loss": 0.6251, + "num_input_tokens_seen": 46570265, + "step": 2199 + }, + { + "epoch": 0.2645343594060001, + "flos": 18806531148840.0, + "grad_norm": 9.837354538265023, + "learning_rate": 3.450364402612397e-06, + "loss": 0.7557, + "num_input_tokens_seen": 46587140, + "step": 2200 + }, + { + "epoch": 0.26465460229663923, + "flos": 16458576024120.0, + "grad_norm": 2.558870117170915, + "learning_rate": 3.449827926217366e-06, + "loss": 0.8096, + "num_input_tokens_seen": 46606295, + "step": 2201 + }, + { + "epoch": 0.2647748451872783, + "flos": 21562294167840.0, + "grad_norm": 3.243192879934451, + "learning_rate": 3.449291229884591e-06, + "loss": 0.7846, + "num_input_tokens_seen": 46627255, + "step": 2202 + }, + { + "epoch": 0.2648950880779174, + "flos": 19645380856680.0, + "grad_norm": 2.8948228202718385, + "learning_rate": 3.4487543136954887e-06, + "loss": 0.8421, + "num_input_tokens_seen": 46646595, + "step": 2203 + }, + { + "epoch": 0.2650153309685565, + "flos": 21157493969760.0, + "grad_norm": 3.2536285391789077, + "learning_rate": 3.448217177731509e-06, + "loss": 0.8922, + "num_input_tokens_seen": 46666800, + "step": 2204 + }, + { + "epoch": 0.26513557385919556, + "flos": 14838963652320.0, + "grad_norm": 3.06907943542239, + "learning_rate": 3.4476798220741348e-06, + "loss": 0.7574, + "num_input_tokens_seen": 46685400, + "step": 2205 + }, + { + "epoch": 0.26525581674983467, + "flos": 12894827266920.0, + "grad_norm": 2.426157358593215, + "learning_rate": 3.4471422468048826e-06, + "loss": 0.7651, + "num_input_tokens_seen": 46703845, + "step": 2206 + }, + { + "epoch": 0.2653760596404738, + "flos": 19671907411800.0, + "grad_norm": 3.163505750110381, + "learning_rate": 3.4466044520053022e-06, + "loss": 0.7171, + "num_input_tokens_seen": 46722570, + "step": 2207 + }, + { + "epoch": 0.26549630253111284, + "flos": 16673226282000.0, + "grad_norm": 2.8841407424509686, + "learning_rate": 3.446066437756977e-06, + "loss": 0.5855, + "num_input_tokens_seen": 46741495, + "step": 2208 + }, + { + "epoch": 0.26561654542175195, + "flos": 17241713185080.0, + "grad_norm": 2.987545546180498, + "learning_rate": 3.4455282041415224e-06, + "loss": 0.7331, + "num_input_tokens_seen": 46760425, + "step": 2209 + }, + { + "epoch": 0.265736788312391, + "flos": 19725878660880.0, + "grad_norm": 2.945213741194897, + "learning_rate": 3.4449897512405894e-06, + "loss": 0.8517, + "num_input_tokens_seen": 46779295, + "step": 2210 + }, + { + "epoch": 0.2658570312030301, + "flos": 17187805255920.0, + "grad_norm": 3.0287902920034875, + "learning_rate": 3.444451079135859e-06, + "loss": 0.727, + "num_input_tokens_seen": 46798525, + "step": 2211 + }, + { + "epoch": 0.2659772740936692, + "flos": 15994992263040.0, + "grad_norm": 2.7978916984195807, + "learning_rate": 3.4439121879090493e-06, + "loss": 0.7235, + "num_input_tokens_seen": 46816025, + "step": 2212 + }, + { + "epoch": 0.2660975169843083, + "flos": 14461513168320.0, + "grad_norm": 2.5802142376627466, + "learning_rate": 3.4433730776419082e-06, + "loss": 0.8156, + "num_input_tokens_seen": 46834670, + "step": 2213 + }, + { + "epoch": 0.2662177598749474, + "flos": 21288575407320.0, + "grad_norm": 2.7994643857107366, + "learning_rate": 3.4428337484162183e-06, + "loss": 0.7801, + "num_input_tokens_seen": 46855200, + "step": 2214 + }, + { + "epoch": 0.2663380027655865, + "flos": 15918926853240.0, + "grad_norm": 2.5654716938414768, + "learning_rate": 3.442294200313797e-06, + "loss": 0.8253, + "num_input_tokens_seen": 46872950, + "step": 2215 + }, + { + "epoch": 0.26645824565622556, + "flos": 49389975054360.0, + "grad_norm": 0.9045000939630109, + "learning_rate": 3.4417544334164916e-06, + "loss": 0.5753, + "num_input_tokens_seen": 46936815, + "step": 2216 + }, + { + "epoch": 0.26657848854686467, + "flos": 18510243328200.0, + "grad_norm": 2.1626548182464354, + "learning_rate": 3.4412144478061854e-06, + "loss": 0.7537, + "num_input_tokens_seen": 46958945, + "step": 2217 + }, + { + "epoch": 0.2666987314375038, + "flos": 17350447182240.0, + "grad_norm": 2.2770874507363086, + "learning_rate": 3.4406742435647925e-06, + "loss": 0.7432, + "num_input_tokens_seen": 46978730, + "step": 2218 + }, + { + "epoch": 0.26681897432814283, + "flos": 19830655163040.0, + "grad_norm": 3.8927182614012135, + "learning_rate": 3.440133820774263e-06, + "loss": 0.7667, + "num_input_tokens_seen": 46998260, + "step": 2219 + }, + { + "epoch": 0.26693921721878194, + "flos": 21264613308960.0, + "grad_norm": 3.8011749157801784, + "learning_rate": 3.439593179516578e-06, + "loss": 0.7905, + "num_input_tokens_seen": 47017890, + "step": 2220 + }, + { + "epoch": 0.26705946010942105, + "flos": 15730518210840.0, + "grad_norm": 2.381688587718728, + "learning_rate": 3.4390523198737524e-06, + "loss": 0.7895, + "num_input_tokens_seen": 47036770, + "step": 2221 + }, + { + "epoch": 0.2671797030000601, + "flos": 15703453436400.0, + "grad_norm": 5.005794791148442, + "learning_rate": 3.4385112419278333e-06, + "loss": 0.7209, + "num_input_tokens_seen": 47057715, + "step": 2222 + }, + { + "epoch": 0.2672999458906992, + "flos": 47322414450360.0, + "grad_norm": 0.8243227346191154, + "learning_rate": 3.4379699457609033e-06, + "loss": 0.687, + "num_input_tokens_seen": 47115260, + "step": 2223 + }, + { + "epoch": 0.26742018878133833, + "flos": 12326878583160.0, + "grad_norm": 2.8520032568186693, + "learning_rate": 3.4374284314550755e-06, + "loss": 0.8752, + "num_input_tokens_seen": 47134020, + "step": 2224 + }, + { + "epoch": 0.2675404316719774, + "flos": 15108503298120.0, + "grad_norm": 3.578057434243696, + "learning_rate": 3.436886699092498e-06, + "loss": 0.7855, + "num_input_tokens_seen": 47152255, + "step": 2225 + }, + { + "epoch": 0.2676606745626165, + "flos": 12752031644160.0, + "grad_norm": 3.9754267721377894, + "learning_rate": 3.4363447487553502e-06, + "loss": 0.6915, + "num_input_tokens_seen": 47165290, + "step": 2226 + }, + { + "epoch": 0.26778091745325555, + "flos": 20425605301320.0, + "grad_norm": 2.4147394212680036, + "learning_rate": 3.4358025805258455e-06, + "loss": 0.7682, + "num_input_tokens_seen": 47184715, + "step": 2227 + }, + { + "epoch": 0.26790116034389466, + "flos": 15321475578120.0, + "grad_norm": 2.3039051861791267, + "learning_rate": 3.435260194486232e-06, + "loss": 0.8149, + "num_input_tokens_seen": 47202405, + "step": 2228 + }, + { + "epoch": 0.2680214032345338, + "flos": 13164430232640.0, + "grad_norm": 3.6421701975001928, + "learning_rate": 3.4347175907187875e-06, + "loss": 0.7962, + "num_input_tokens_seen": 47219115, + "step": 2229 + }, + { + "epoch": 0.26814164612517283, + "flos": 16405079674440.0, + "grad_norm": 2.428069333257037, + "learning_rate": 3.4341747693058254e-06, + "loss": 0.8602, + "num_input_tokens_seen": 47237310, + "step": 2230 + }, + { + "epoch": 0.26826188901581194, + "flos": 26179859490360.0, + "grad_norm": 4.193194261275402, + "learning_rate": 3.4336317303296916e-06, + "loss": 0.7547, + "num_input_tokens_seen": 47258005, + "step": 2231 + }, + { + "epoch": 0.26838213190645105, + "flos": 12867825812400.0, + "grad_norm": 3.0487073816978154, + "learning_rate": 3.4330884738727635e-06, + "loss": 0.7339, + "num_input_tokens_seen": 47275900, + "step": 2232 + }, + { + "epoch": 0.2685023747970901, + "flos": 16594089856080.0, + "grad_norm": 2.7837263933815817, + "learning_rate": 3.4325450000174535e-06, + "loss": 0.7013, + "num_input_tokens_seen": 47292260, + "step": 2233 + }, + { + "epoch": 0.2686226176877292, + "flos": 14703703100040.0, + "grad_norm": 4.2306871859724815, + "learning_rate": 3.4320013088462067e-06, + "loss": 0.7243, + "num_input_tokens_seen": 47309340, + "step": 2234 + }, + { + "epoch": 0.2687428605783683, + "flos": 15999962876760.0, + "grad_norm": 2.7018833726607165, + "learning_rate": 3.431457400441499e-06, + "loss": 0.7952, + "num_input_tokens_seen": 47329455, + "step": 2235 + }, + { + "epoch": 0.2688631034690074, + "flos": 51578703037080.0, + "grad_norm": 1.0100459700216653, + "learning_rate": 3.4309132748858424e-06, + "loss": 0.6792, + "num_input_tokens_seen": 47390165, + "step": 2236 + }, + { + "epoch": 0.2689833463596465, + "flos": 16728558909360.0, + "grad_norm": 2.066327098924208, + "learning_rate": 3.430368932261779e-06, + "loss": 0.8246, + "num_input_tokens_seen": 47410240, + "step": 2237 + }, + { + "epoch": 0.2691035892502856, + "flos": 12542288680080.0, + "grad_norm": 3.308124171130251, + "learning_rate": 3.429824372651886e-06, + "loss": 0.7342, + "num_input_tokens_seen": 47428110, + "step": 2238 + }, + { + "epoch": 0.26922383214092466, + "flos": 12947278837920.0, + "grad_norm": 2.890247217184099, + "learning_rate": 3.4292795961387732e-06, + "loss": 0.8168, + "num_input_tokens_seen": 47445730, + "step": 2239 + }, + { + "epoch": 0.26934407503156377, + "flos": 11781593939400.0, + "grad_norm": 4.742206403766894, + "learning_rate": 3.4287346028050818e-06, + "loss": 0.861, + "num_input_tokens_seen": 47461520, + "step": 2240 + }, + { + "epoch": 0.2694643179222028, + "flos": 17377132037160.0, + "grad_norm": 1.8986178967828173, + "learning_rate": 3.4281893927334866e-06, + "loss": 0.7796, + "num_input_tokens_seen": 47481150, + "step": 2241 + }, + { + "epoch": 0.26958456081284193, + "flos": 18106297949040.0, + "grad_norm": 3.7364133754322975, + "learning_rate": 3.4276439660066963e-06, + "loss": 0.7346, + "num_input_tokens_seen": 47500570, + "step": 2242 + }, + { + "epoch": 0.26970480370348104, + "flos": 13217515002840.0, + "grad_norm": 5.521132013650661, + "learning_rate": 3.427098322707452e-06, + "loss": 0.8269, + "num_input_tokens_seen": 47516255, + "step": 2243 + }, + { + "epoch": 0.2698250465941201, + "flos": 7818110577720.0, + "grad_norm": 2.711162722683463, + "learning_rate": 3.426552462918526e-06, + "loss": 0.8675, + "num_input_tokens_seen": 47533910, + "step": 2244 + }, + { + "epoch": 0.2699452894847592, + "flos": 12621900005400.0, + "grad_norm": 3.5974740014319777, + "learning_rate": 3.426006386722726e-06, + "loss": 0.7237, + "num_input_tokens_seen": 47551690, + "step": 2245 + }, + { + "epoch": 0.2700655323753983, + "flos": 13192318166040.0, + "grad_norm": 3.0390293933467456, + "learning_rate": 3.4254600942028914e-06, + "loss": 0.9071, + "num_input_tokens_seen": 47569285, + "step": 2246 + }, + { + "epoch": 0.2701857752660374, + "flos": 13272467710680.0, + "grad_norm": 2.869712279249233, + "learning_rate": 3.424913585441893e-06, + "loss": 0.7918, + "num_input_tokens_seen": 47586840, + "step": 2247 + }, + { + "epoch": 0.2703060181566765, + "flos": 11889758057280.0, + "grad_norm": 2.6510578347383222, + "learning_rate": 3.4243668605226374e-06, + "loss": 0.8578, + "num_input_tokens_seen": 47603585, + "step": 2248 + }, + { + "epoch": 0.2704262610473156, + "flos": 14297889783240.0, + "grad_norm": 2.857035028602745, + "learning_rate": 3.423819919528061e-06, + "loss": 0.8154, + "num_input_tokens_seen": 47621390, + "step": 2249 + }, + { + "epoch": 0.27054650393795465, + "flos": 15162696166920.0, + "grad_norm": 2.8078512979840253, + "learning_rate": 3.4232727625411355e-06, + "loss": 0.7694, + "num_input_tokens_seen": 47640215, + "step": 2250 + }, + { + "epoch": 0.27066674682859376, + "flos": 13621175442360.0, + "grad_norm": 2.703646482760173, + "learning_rate": 3.4227253896448626e-06, + "loss": 0.8377, + "num_input_tokens_seen": 47657795, + "step": 2251 + }, + { + "epoch": 0.2707869897192329, + "flos": 16836723027240.0, + "grad_norm": 3.0580014255737415, + "learning_rate": 3.42217780092228e-06, + "loss": 0.7881, + "num_input_tokens_seen": 47675855, + "step": 2252 + }, + { + "epoch": 0.27090723260987193, + "flos": 42981733884360.0, + "grad_norm": 0.8473053306889133, + "learning_rate": 3.421629996456456e-06, + "loss": 0.6295, + "num_input_tokens_seen": 47734195, + "step": 2253 + }, + { + "epoch": 0.27102747550051104, + "flos": 8682568701840.0, + "grad_norm": 2.5222669786057037, + "learning_rate": 3.421081976330491e-06, + "loss": 0.8013, + "num_input_tokens_seen": 47752430, + "step": 2254 + }, + { + "epoch": 0.27114771839115015, + "flos": 14540617934280.0, + "grad_norm": 3.9353408948888244, + "learning_rate": 3.4205337406275207e-06, + "loss": 0.8529, + "num_input_tokens_seen": 47772270, + "step": 2255 + }, + { + "epoch": 0.2712679612817892, + "flos": 13379935309440.0, + "grad_norm": 3.8278357571690957, + "learning_rate": 3.4199852894307114e-06, + "loss": 0.7249, + "num_input_tokens_seen": 47788740, + "step": 2256 + }, + { + "epoch": 0.2713882041724283, + "flos": 17915134890120.0, + "grad_norm": 2.5626508156659673, + "learning_rate": 3.419436622823262e-06, + "loss": 0.77, + "num_input_tokens_seen": 47809180, + "step": 2257 + }, + { + "epoch": 0.27150844706306737, + "flos": 16863787801680.0, + "grad_norm": 2.4572077507340184, + "learning_rate": 3.4188877408884063e-06, + "loss": 0.7254, + "num_input_tokens_seen": 47829605, + "step": 2258 + }, + { + "epoch": 0.2716286899537065, + "flos": 16511565814440.0, + "grad_norm": 3.219028201626418, + "learning_rate": 3.4183386437094088e-06, + "loss": 0.6333, + "num_input_tokens_seen": 47845990, + "step": 2259 + }, + { + "epoch": 0.2717489328443456, + "flos": 9519138892560.0, + "grad_norm": 2.790556520858511, + "learning_rate": 3.417789331369565e-06, + "loss": 0.793, + "num_input_tokens_seen": 47861500, + "step": 2260 + }, + { + "epoch": 0.27186917573498465, + "flos": 21481384784160.0, + "grad_norm": 2.238863298455523, + "learning_rate": 3.4172398039522088e-06, + "loss": 0.8832, + "num_input_tokens_seen": 47882505, + "step": 2261 + }, + { + "epoch": 0.27198941862562376, + "flos": 19078888531080.0, + "grad_norm": 2.0097295611438737, + "learning_rate": 3.4166900615407e-06, + "loss": 0.7807, + "num_input_tokens_seen": 47900140, + "step": 2262 + }, + { + "epoch": 0.27210966151626287, + "flos": 24073271138400.0, + "grad_norm": 2.6381975562251982, + "learning_rate": 3.416140104218436e-06, + "loss": 0.7269, + "num_input_tokens_seen": 47919225, + "step": 2263 + }, + { + "epoch": 0.2722299044069019, + "flos": 48271106253960.0, + "grad_norm": 0.8516157291119941, + "learning_rate": 3.4155899320688437e-06, + "loss": 0.7365, + "num_input_tokens_seen": 47985020, + "step": 2264 + }, + { + "epoch": 0.27235014729754103, + "flos": 11160497165520.0, + "grad_norm": 2.357864907139212, + "learning_rate": 3.415039545175384e-06, + "loss": 0.7273, + "num_input_tokens_seen": 48000465, + "step": 2265 + }, + { + "epoch": 0.27247039018818014, + "flos": 16000247816400.0, + "grad_norm": 7.761529283750063, + "learning_rate": 3.414488943621551e-06, + "loss": 0.6385, + "num_input_tokens_seen": 48018850, + "step": 2266 + }, + { + "epoch": 0.2725906330788192, + "flos": 13648778436120.0, + "grad_norm": 2.5240471165547076, + "learning_rate": 3.41393812749087e-06, + "loss": 0.7241, + "num_input_tokens_seen": 48036615, + "step": 2267 + }, + { + "epoch": 0.2727108759694583, + "flos": 13048826024160.0, + "grad_norm": 5.6758440910793375, + "learning_rate": 3.4133870968668984e-06, + "loss": 0.713, + "num_input_tokens_seen": 48051135, + "step": 2268 + }, + { + "epoch": 0.2728311188600974, + "flos": 17916337968600.0, + "grad_norm": 2.59771673375896, + "learning_rate": 3.412835851833229e-06, + "loss": 0.7551, + "num_input_tokens_seen": 48073050, + "step": 2269 + }, + { + "epoch": 0.2729513617507365, + "flos": 22750294846800.0, + "grad_norm": 2.355311855069589, + "learning_rate": 3.4122843924734834e-06, + "loss": 0.7633, + "num_input_tokens_seen": 48095070, + "step": 2270 + }, + { + "epoch": 0.2730716046413756, + "flos": 13943863178280.0, + "grad_norm": 3.1425336665107997, + "learning_rate": 3.411732718871319e-06, + "loss": 0.8702, + "num_input_tokens_seen": 48110630, + "step": 2271 + }, + { + "epoch": 0.27319184753201464, + "flos": 19754114853840.0, + "grad_norm": 2.0500567219027865, + "learning_rate": 3.4111808311104227e-06, + "loss": 0.7652, + "num_input_tokens_seen": 48132665, + "step": 2272 + }, + { + "epoch": 0.27331209042265375, + "flos": 23317610331360.0, + "grad_norm": 3.17574461368013, + "learning_rate": 3.410628729274517e-06, + "loss": 0.6721, + "num_input_tokens_seen": 48153905, + "step": 2273 + }, + { + "epoch": 0.27343233331329286, + "flos": 18861768796320.0, + "grad_norm": 3.8226945000034265, + "learning_rate": 3.4100764134473546e-06, + "loss": 0.8107, + "num_input_tokens_seen": 48172910, + "step": 2274 + }, + { + "epoch": 0.2735525762039319, + "flos": 17862968258760.0, + "grad_norm": 4.283965426065961, + "learning_rate": 3.4095238837127215e-06, + "loss": 0.825, + "num_input_tokens_seen": 48191770, + "step": 2275 + }, + { + "epoch": 0.27367281909457103, + "flos": 10518509309400.0, + "grad_norm": 3.047781159071515, + "learning_rate": 3.4089711401544355e-06, + "loss": 0.771, + "num_input_tokens_seen": 48209085, + "step": 2276 + }, + { + "epoch": 0.27379306198521014, + "flos": 17187678616080.0, + "grad_norm": 3.622402839279073, + "learning_rate": 3.4084181828563486e-06, + "loss": 0.6526, + "num_input_tokens_seen": 48225525, + "step": 2277 + }, + { + "epoch": 0.2739133048758492, + "flos": 12733515058920.0, + "grad_norm": 2.4909002632992157, + "learning_rate": 3.4078650119023428e-06, + "loss": 0.6872, + "num_input_tokens_seen": 48243560, + "step": 2278 + }, + { + "epoch": 0.2740335477664883, + "flos": 14075387855280.0, + "grad_norm": 3.3501006097061934, + "learning_rate": 3.4073116273763337e-06, + "loss": 0.7307, + "num_input_tokens_seen": 48257725, + "step": 2279 + }, + { + "epoch": 0.2741537906571274, + "flos": 19132479860640.0, + "grad_norm": 2.6614048026139248, + "learning_rate": 3.40675802936227e-06, + "loss": 0.7892, + "num_input_tokens_seen": 48278230, + "step": 2280 + }, + { + "epoch": 0.27427403354776647, + "flos": 25096477013760.0, + "grad_norm": 2.6433927037553633, + "learning_rate": 3.4062042179441318e-06, + "loss": 0.6969, + "num_input_tokens_seen": 48298420, + "step": 2281 + }, + { + "epoch": 0.2743942764384056, + "flos": 13701483286800.0, + "grad_norm": 2.1612957506428465, + "learning_rate": 3.4056501932059314e-06, + "loss": 0.7918, + "num_input_tokens_seen": 48316215, + "step": 2282 + }, + { + "epoch": 0.2745145193290447, + "flos": 47848703118120.0, + "grad_norm": 0.8366910443674358, + "learning_rate": 3.405095955231715e-06, + "loss": 0.6075, + "num_input_tokens_seen": 48367590, + "step": 2283 + }, + { + "epoch": 0.27463476221968375, + "flos": 11754624144840.0, + "grad_norm": 4.990952507647588, + "learning_rate": 3.4045415041055585e-06, + "loss": 0.921, + "num_input_tokens_seen": 48382950, + "step": 2284 + }, + { + "epoch": 0.27475500511032286, + "flos": 7492320165720.0, + "grad_norm": 3.7107713521605095, + "learning_rate": 3.4039868399115728e-06, + "loss": 0.7562, + "num_input_tokens_seen": 48397310, + "step": 2285 + }, + { + "epoch": 0.27487524800096197, + "flos": 12624496122120.0, + "grad_norm": 3.4370974218275787, + "learning_rate": 3.4034319627339003e-06, + "loss": 0.7838, + "num_input_tokens_seen": 48413895, + "step": 2286 + }, + { + "epoch": 0.274995490891601, + "flos": 19883138394000.0, + "grad_norm": 3.132074663979872, + "learning_rate": 3.402876872656715e-06, + "loss": 0.6811, + "num_input_tokens_seen": 48431935, + "step": 2287 + }, + { + "epoch": 0.27511573378224013, + "flos": 17156782986480.0, + "grad_norm": 2.4869581504586216, + "learning_rate": 3.402321569764223e-06, + "loss": 0.8743, + "num_input_tokens_seen": 48450960, + "step": 2288 + }, + { + "epoch": 0.2752359766728792, + "flos": 12187818835680.0, + "grad_norm": 2.362540136733038, + "learning_rate": 3.4017660541406635e-06, + "loss": 0.8136, + "num_input_tokens_seen": 48466745, + "step": 2289 + }, + { + "epoch": 0.2753562195635183, + "flos": 18534648666000.0, + "grad_norm": 2.578077281163497, + "learning_rate": 3.4012103258703092e-06, + "loss": 0.7266, + "num_input_tokens_seen": 48485220, + "step": 2290 + }, + { + "epoch": 0.2754764624541574, + "flos": 20045020481280.0, + "grad_norm": 2.597818699396492, + "learning_rate": 3.4006543850374616e-06, + "loss": 0.8187, + "num_input_tokens_seen": 48499990, + "step": 2291 + }, + { + "epoch": 0.27559670534479647, + "flos": 12570746492760.0, + "grad_norm": 4.022629147406999, + "learning_rate": 3.400098231726458e-06, + "loss": 0.7405, + "num_input_tokens_seen": 48516810, + "step": 2292 + }, + { + "epoch": 0.2757169482354356, + "flos": 16048741892400.0, + "grad_norm": 2.663694026221091, + "learning_rate": 3.3995418660216657e-06, + "loss": 0.8622, + "num_input_tokens_seen": 48533985, + "step": 2293 + }, + { + "epoch": 0.2758371911260747, + "flos": 15244112109960.0, + "grad_norm": 3.6085094388373467, + "learning_rate": 3.3989852880074848e-06, + "loss": 0.7964, + "num_input_tokens_seen": 48555135, + "step": 2294 + }, + { + "epoch": 0.27595743401671374, + "flos": 51082391860080.0, + "grad_norm": 0.768752968014826, + "learning_rate": 3.398428497768348e-06, + "loss": 0.6352, + "num_input_tokens_seen": 48620025, + "step": 2295 + }, + { + "epoch": 0.27607767690735285, + "flos": 15513683415720.0, + "grad_norm": 2.4104781221838243, + "learning_rate": 3.3978714953887205e-06, + "loss": 0.698, + "num_input_tokens_seen": 48639500, + "step": 2296 + }, + { + "epoch": 0.27619791979799196, + "flos": 18185307735120.0, + "grad_norm": 2.445665132976604, + "learning_rate": 3.397314280953098e-06, + "loss": 0.8478, + "num_input_tokens_seen": 48660045, + "step": 2297 + }, + { + "epoch": 0.276318162688631, + "flos": 18132349604760.0, + "grad_norm": 2.739880409136879, + "learning_rate": 3.3967568545460108e-06, + "loss": 0.7803, + "num_input_tokens_seen": 48679305, + "step": 2298 + }, + { + "epoch": 0.27643840557927013, + "flos": 13245339616320.0, + "grad_norm": 2.1321520778459364, + "learning_rate": 3.3961992162520185e-06, + "loss": 0.7738, + "num_input_tokens_seen": 48697650, + "step": 2299 + }, + { + "epoch": 0.27655864846990924, + "flos": 18184072996680.0, + "grad_norm": 2.5598353976737633, + "learning_rate": 3.3956413661557156e-06, + "loss": 0.7058, + "num_input_tokens_seen": 48717545, + "step": 2300 + }, + { + "epoch": 0.2766788913605483, + "flos": 14811803898000.0, + "grad_norm": 3.192964843269723, + "learning_rate": 3.3950833043417273e-06, + "loss": 0.6467, + "num_input_tokens_seen": 48735410, + "step": 2301 + }, + { + "epoch": 0.2767991342511874, + "flos": 15702566957520.0, + "grad_norm": 5.216120440046257, + "learning_rate": 3.3945250308947105e-06, + "loss": 0.7077, + "num_input_tokens_seen": 48751435, + "step": 2302 + }, + { + "epoch": 0.2769193771418265, + "flos": 48662292669240.0, + "grad_norm": 1.2337172482055592, + "learning_rate": 3.3939665458993556e-06, + "loss": 0.6958, + "num_input_tokens_seen": 48805575, + "step": 2303 + }, + { + "epoch": 0.27703962003246557, + "flos": 15135853012200.0, + "grad_norm": 3.8992620601400287, + "learning_rate": 3.3934078494403843e-06, + "loss": 0.7391, + "num_input_tokens_seen": 48824870, + "step": 2304 + }, + { + "epoch": 0.2771598629231047, + "flos": 16782815098080.0, + "grad_norm": 2.671668775307372, + "learning_rate": 3.3928489416025495e-06, + "loss": 0.7912, + "num_input_tokens_seen": 48845435, + "step": 2305 + }, + { + "epoch": 0.27728010581374374, + "flos": 13407823242840.0, + "grad_norm": 3.189281937910159, + "learning_rate": 3.392289822470638e-06, + "loss": 0.7632, + "num_input_tokens_seen": 48863135, + "step": 2306 + }, + { + "epoch": 0.27740034870438285, + "flos": 14190960403800.0, + "grad_norm": 3.030505216673638, + "learning_rate": 3.3917304921294674e-06, + "loss": 0.741, + "num_input_tokens_seen": 48881020, + "step": 2307 + }, + { + "epoch": 0.27752059159502196, + "flos": 15809306377200.0, + "grad_norm": 2.1824940759812193, + "learning_rate": 3.3911709506638876e-06, + "loss": 0.7882, + "num_input_tokens_seen": 48900050, + "step": 2308 + }, + { + "epoch": 0.277640834485661, + "flos": 19503883292280.0, + "grad_norm": 4.165152974153014, + "learning_rate": 3.390611198158781e-06, + "loss": 0.7935, + "num_input_tokens_seen": 48917645, + "step": 2309 + }, + { + "epoch": 0.2777610773763001, + "flos": 14238093101520.0, + "grad_norm": 4.263809879881191, + "learning_rate": 3.3900512346990612e-06, + "loss": 0.8782, + "num_input_tokens_seen": 48933355, + "step": 2310 + }, + { + "epoch": 0.27788132026693924, + "flos": 28150839030480.0, + "grad_norm": 2.5817937467593466, + "learning_rate": 3.389491060369674e-06, + "loss": 0.6338, + "num_input_tokens_seen": 48958750, + "step": 2311 + }, + { + "epoch": 0.2780015631575783, + "flos": 16377729960360.0, + "grad_norm": 3.419766407847285, + "learning_rate": 3.388930675255598e-06, + "loss": 0.8696, + "num_input_tokens_seen": 48978320, + "step": 2312 + }, + { + "epoch": 0.2781218060482174, + "flos": 8844355809240.0, + "grad_norm": 14.44864689916661, + "learning_rate": 3.388370079441843e-06, + "loss": 0.7587, + "num_input_tokens_seen": 48993555, + "step": 2313 + }, + { + "epoch": 0.2782420489388565, + "flos": 13213145928360.0, + "grad_norm": 2.454747726313502, + "learning_rate": 3.3878092730134505e-06, + "loss": 0.9175, + "num_input_tokens_seen": 49011260, + "step": 2314 + }, + { + "epoch": 0.27836229182949557, + "flos": 13515259181640.0, + "grad_norm": 2.9323659631092367, + "learning_rate": 3.3872482560554947e-06, + "loss": 0.7876, + "num_input_tokens_seen": 49029755, + "step": 2315 + }, + { + "epoch": 0.2784825347201347, + "flos": 49461671389680.0, + "grad_norm": 0.8163970143649071, + "learning_rate": 3.386687028653082e-06, + "loss": 0.5871, + "num_input_tokens_seen": 49092320, + "step": 2316 + }, + { + "epoch": 0.2786027776107738, + "flos": 16560724749600.0, + "grad_norm": 2.7030314170205543, + "learning_rate": 3.386125590891349e-06, + "loss": 0.829, + "num_input_tokens_seen": 49108915, + "step": 2317 + }, + { + "epoch": 0.27872302050141284, + "flos": 11490561672120.0, + "grad_norm": 2.5321591475310736, + "learning_rate": 3.3855639428554657e-06, + "loss": 0.8092, + "num_input_tokens_seen": 49126165, + "step": 2318 + }, + { + "epoch": 0.27884326339205195, + "flos": 16188561478920.0, + "grad_norm": 2.4979636481933265, + "learning_rate": 3.385002084630635e-06, + "loss": 0.7906, + "num_input_tokens_seen": 49144855, + "step": 2319 + }, + { + "epoch": 0.278963506282691, + "flos": 15028037153880.0, + "grad_norm": 2.3575526393219883, + "learning_rate": 3.384440016302088e-06, + "loss": 0.8236, + "num_input_tokens_seen": 49163250, + "step": 2320 + }, + { + "epoch": 0.2790837491733301, + "flos": 16052129508120.0, + "grad_norm": 3.3301527266662054, + "learning_rate": 3.3838777379550923e-06, + "loss": 0.5983, + "num_input_tokens_seen": 49182415, + "step": 2321 + }, + { + "epoch": 0.27920399206396923, + "flos": 19267170533640.0, + "grad_norm": 2.792050425404765, + "learning_rate": 3.383315249674944e-06, + "loss": 0.7857, + "num_input_tokens_seen": 49200700, + "step": 2322 + }, + { + "epoch": 0.2793242349546083, + "flos": 18610175856480.0, + "grad_norm": 2.7058171443691843, + "learning_rate": 3.3827525515469715e-06, + "loss": 0.8488, + "num_input_tokens_seen": 49215325, + "step": 2323 + }, + { + "epoch": 0.2794444778452474, + "flos": 15243827170320.0, + "grad_norm": 2.5038013168461917, + "learning_rate": 3.3821896436565367e-06, + "loss": 0.6875, + "num_input_tokens_seen": 49234705, + "step": 2324 + }, + { + "epoch": 0.2795647207358865, + "flos": 15780500304960.0, + "grad_norm": 2.7612041765575164, + "learning_rate": 3.381626526089032e-06, + "loss": 0.6869, + "num_input_tokens_seen": 49253990, + "step": 2325 + }, + { + "epoch": 0.27968496362652556, + "flos": 15703865015880.0, + "grad_norm": 2.3923609794637577, + "learning_rate": 3.3810631989298815e-06, + "loss": 0.7715, + "num_input_tokens_seen": 49273320, + "step": 2326 + }, + { + "epoch": 0.2798052065171647, + "flos": 17026208108280.0, + "grad_norm": 3.407836824423877, + "learning_rate": 3.3804996622645423e-06, + "loss": 0.8309, + "num_input_tokens_seen": 49291040, + "step": 2327 + }, + { + "epoch": 0.2799254494078038, + "flos": 15753118930920.0, + "grad_norm": 2.692680245602848, + "learning_rate": 3.3799359161785015e-06, + "loss": 0.8694, + "num_input_tokens_seen": 49310410, + "step": 2328 + }, + { + "epoch": 0.28004569229844284, + "flos": 19346116999800.0, + "grad_norm": 3.552200822574496, + "learning_rate": 3.3793719607572798e-06, + "loss": 0.8381, + "num_input_tokens_seen": 49331095, + "step": 2329 + }, + { + "epoch": 0.28016593518908195, + "flos": 24640744922760.0, + "grad_norm": 3.579876466889392, + "learning_rate": 3.378807796086428e-06, + "loss": 0.749, + "num_input_tokens_seen": 49353675, + "step": 2330 + }, + { + "epoch": 0.28028617807972106, + "flos": 11166385918080.0, + "grad_norm": 2.5058039051793655, + "learning_rate": 3.37824342225153e-06, + "loss": 0.7477, + "num_input_tokens_seen": 49369815, + "step": 2331 + }, + { + "epoch": 0.2804064209703601, + "flos": 18699855049080.0, + "grad_norm": 2.4967705849215, + "learning_rate": 3.3776788393382006e-06, + "loss": 0.7523, + "num_input_tokens_seen": 49389015, + "step": 2332 + }, + { + "epoch": 0.2805266638609992, + "flos": 21535482673080.0, + "grad_norm": 3.958625314216658, + "learning_rate": 3.3771140474320872e-06, + "loss": 0.7502, + "num_input_tokens_seen": 49408685, + "step": 2333 + }, + { + "epoch": 0.28064690675163834, + "flos": 15696804844800.0, + "grad_norm": 2.568072795471931, + "learning_rate": 3.3765490466188664e-06, + "loss": 0.7651, + "num_input_tokens_seen": 49425805, + "step": 2334 + }, + { + "epoch": 0.2807671496422774, + "flos": 15351263109120.0, + "grad_norm": 3.4743784012342953, + "learning_rate": 3.3759838369842508e-06, + "loss": 0.7072, + "num_input_tokens_seen": 49443600, + "step": 2335 + }, + { + "epoch": 0.2808873925329165, + "flos": 15729410112240.0, + "grad_norm": 3.731597250366179, + "learning_rate": 3.375418418613981e-06, + "loss": 0.7104, + "num_input_tokens_seen": 49462345, + "step": 2336 + }, + { + "epoch": 0.28100763542355556, + "flos": 11706668288160.0, + "grad_norm": 4.235995387318867, + "learning_rate": 3.374852791593831e-06, + "loss": 0.8151, + "num_input_tokens_seen": 49478265, + "step": 2337 + }, + { + "epoch": 0.28112787831419467, + "flos": 13920787558800.0, + "grad_norm": 12.251714469668686, + "learning_rate": 3.374286956009605e-06, + "loss": 0.5083, + "num_input_tokens_seen": 49496550, + "step": 2338 + }, + { + "epoch": 0.2812481212048338, + "flos": 9303538835880.0, + "grad_norm": 3.0740046929526508, + "learning_rate": 3.3737209119471405e-06, + "loss": 0.7481, + "num_input_tokens_seen": 49512780, + "step": 2339 + }, + { + "epoch": 0.28136836409547283, + "flos": 11382714153840.0, + "grad_norm": 3.629752628105027, + "learning_rate": 3.373154659492306e-06, + "loss": 0.6326, + "num_input_tokens_seen": 49530640, + "step": 2340 + }, + { + "epoch": 0.28148860698611194, + "flos": 14564833312320.0, + "grad_norm": 2.1361599876765878, + "learning_rate": 3.3725881987310016e-06, + "loss": 0.8321, + "num_input_tokens_seen": 49547895, + "step": 2341 + }, + { + "epoch": 0.28160884987675106, + "flos": 12732691899960.0, + "grad_norm": 2.4610035720917076, + "learning_rate": 3.372021529749159e-06, + "loss": 0.8522, + "num_input_tokens_seen": 49566675, + "step": 2342 + }, + { + "epoch": 0.2817290927673901, + "flos": 12271704255600.0, + "grad_norm": 2.0795792388824754, + "learning_rate": 3.3714546526327405e-06, + "loss": 0.9018, + "num_input_tokens_seen": 49584395, + "step": 2343 + }, + { + "epoch": 0.2818493356580292, + "flos": 11220642106800.0, + "grad_norm": 2.8825112782478537, + "learning_rate": 3.3708875674677423e-06, + "loss": 0.8485, + "num_input_tokens_seen": 49602090, + "step": 2344 + }, + { + "epoch": 0.28196957854866833, + "flos": 14919746396160.0, + "grad_norm": 3.143955267998076, + "learning_rate": 3.37032027434019e-06, + "loss": 0.8112, + "num_input_tokens_seen": 49621330, + "step": 2345 + }, + { + "epoch": 0.2820898214393074, + "flos": 14594399223600.0, + "grad_norm": 3.6858529760768923, + "learning_rate": 3.369752773336141e-06, + "loss": 0.8184, + "num_input_tokens_seen": 49640530, + "step": 2346 + }, + { + "epoch": 0.2822100643299465, + "flos": 16485577478640.0, + "grad_norm": 2.8262820482693565, + "learning_rate": 3.3691850645416864e-06, + "loss": 0.7605, + "num_input_tokens_seen": 49659960, + "step": 2347 + }, + { + "epoch": 0.2823303072205856, + "flos": 8358519587640.0, + "grad_norm": 2.5286152631887195, + "learning_rate": 3.368617148042945e-06, + "loss": 0.8245, + "num_input_tokens_seen": 49677350, + "step": 2348 + }, + { + "epoch": 0.28245055011122466, + "flos": 13326723899400.0, + "grad_norm": 2.141819864080469, + "learning_rate": 3.368049023926071e-06, + "loss": 0.8314, + "num_input_tokens_seen": 49696065, + "step": 2349 + }, + { + "epoch": 0.2825707930018638, + "flos": 18024787026120.0, + "grad_norm": 2.036646204198923, + "learning_rate": 3.3674806922772476e-06, + "loss": 0.8213, + "num_input_tokens_seen": 49716670, + "step": 2350 + }, + { + "epoch": 0.28269103589250283, + "flos": 18482608674480.0, + "grad_norm": 2.795013660813797, + "learning_rate": 3.3669121531826904e-06, + "loss": 0.729, + "num_input_tokens_seen": 49737370, + "step": 2351 + }, + { + "epoch": 0.28281127878314194, + "flos": 14082827945880.0, + "grad_norm": 2.991210140436149, + "learning_rate": 3.366343406728647e-06, + "loss": 0.8083, + "num_input_tokens_seen": 49756540, + "step": 2352 + }, + { + "epoch": 0.28293152167378105, + "flos": 17484757935720.0, + "grad_norm": 2.2274045188639437, + "learning_rate": 3.3657744530013946e-06, + "loss": 0.6665, + "num_input_tokens_seen": 49775495, + "step": 2353 + }, + { + "epoch": 0.2830517645644201, + "flos": 32276837678640.0, + "grad_norm": 2.8854442461993273, + "learning_rate": 3.3652052920872437e-06, + "loss": 0.699, + "num_input_tokens_seen": 49798080, + "step": 2354 + }, + { + "epoch": 0.2831720074550592, + "flos": 19537280058720.0, + "grad_norm": 2.675503503348378, + "learning_rate": 3.3646359240725355e-06, + "loss": 0.8386, + "num_input_tokens_seen": 49816990, + "step": 2355 + }, + { + "epoch": 0.2832922503456983, + "flos": 22777454601120.0, + "grad_norm": 2.4186023874114433, + "learning_rate": 3.364066349043643e-06, + "loss": 0.6613, + "num_input_tokens_seen": 49837915, + "step": 2356 + }, + { + "epoch": 0.2834124932363374, + "flos": 14913794323680.0, + "grad_norm": 2.045754744916623, + "learning_rate": 3.363496567086969e-06, + "loss": 0.8083, + "num_input_tokens_seen": 49854730, + "step": 2357 + }, + { + "epoch": 0.2835327361269765, + "flos": 28960566066480.0, + "grad_norm": 3.564599606365171, + "learning_rate": 3.3629265782889506e-06, + "loss": 0.7357, + "num_input_tokens_seen": 49876275, + "step": 2358 + }, + { + "epoch": 0.2836529790176156, + "flos": 22208492798640.0, + "grad_norm": 3.100240045913406, + "learning_rate": 3.362356382736054e-06, + "loss": 0.6874, + "num_input_tokens_seen": 49896600, + "step": 2359 + }, + { + "epoch": 0.28377322190825466, + "flos": 9195627997680.0, + "grad_norm": 2.723438210851125, + "learning_rate": 3.361785980514777e-06, + "loss": 0.8947, + "num_input_tokens_seen": 49912520, + "step": 2360 + }, + { + "epoch": 0.28389346479889377, + "flos": 13353978633600.0, + "grad_norm": 2.715094992449377, + "learning_rate": 3.361215371711649e-06, + "loss": 0.7494, + "num_input_tokens_seen": 49931335, + "step": 2361 + }, + { + "epoch": 0.2840137076895329, + "flos": 14914934082240.0, + "grad_norm": 2.035078929523287, + "learning_rate": 3.3606445564132326e-06, + "loss": 0.8146, + "num_input_tokens_seen": 49948350, + "step": 2362 + }, + { + "epoch": 0.28413395058017193, + "flos": 14650111770480.0, + "grad_norm": 2.4546030050652967, + "learning_rate": 3.360073534706118e-06, + "loss": 0.8024, + "num_input_tokens_seen": 49965225, + "step": 2363 + }, + { + "epoch": 0.28425419347081105, + "flos": 27685830571200.0, + "grad_norm": 4.084505081732096, + "learning_rate": 3.35950230667693e-06, + "loss": 0.7431, + "num_input_tokens_seen": 49986640, + "step": 2364 + }, + { + "epoch": 0.28437443636145016, + "flos": 10059452922600.0, + "grad_norm": 3.480911517852355, + "learning_rate": 3.358930872412323e-06, + "loss": 0.8403, + "num_input_tokens_seen": 50003525, + "step": 2365 + }, + { + "epoch": 0.2844946792520892, + "flos": 16648092765120.0, + "grad_norm": 2.044854251359191, + "learning_rate": 3.3583592319989825e-06, + "loss": 0.7865, + "num_input_tokens_seen": 50022615, + "step": 2366 + }, + { + "epoch": 0.2846149221427283, + "flos": 23803129953360.0, + "grad_norm": 3.8254314054662184, + "learning_rate": 3.357787385523627e-06, + "loss": 0.6674, + "num_input_tokens_seen": 50043740, + "step": 2367 + }, + { + "epoch": 0.2847351650333674, + "flos": 20887321124760.0, + "grad_norm": 15.888992990954511, + "learning_rate": 3.3572153330730048e-06, + "loss": 0.8215, + "num_input_tokens_seen": 50064555, + "step": 2368 + }, + { + "epoch": 0.2848554079240065, + "flos": 41077638365640.0, + "grad_norm": 0.801810222218399, + "learning_rate": 3.3566430747338956e-06, + "loss": 0.6614, + "num_input_tokens_seen": 50119480, + "step": 2369 + }, + { + "epoch": 0.2849756508146456, + "flos": 8573043205680.0, + "grad_norm": 3.452186836773707, + "learning_rate": 3.35607061059311e-06, + "loss": 0.8451, + "num_input_tokens_seen": 50134130, + "step": 2370 + }, + { + "epoch": 0.28509589370528465, + "flos": 18429650544120.0, + "grad_norm": 2.413604199012793, + "learning_rate": 3.3554979407374917e-06, + "loss": 0.7304, + "num_input_tokens_seen": 50155960, + "step": 2371 + }, + { + "epoch": 0.28521613659592376, + "flos": 14595887241720.0, + "grad_norm": 2.8049572998169485, + "learning_rate": 3.3549250652539134e-06, + "loss": 0.7197, + "num_input_tokens_seen": 50174775, + "step": 2372 + }, + { + "epoch": 0.2853363794865629, + "flos": 17107054172040.0, + "grad_norm": 2.368304200106212, + "learning_rate": 3.3543519842292794e-06, + "loss": 0.8012, + "num_input_tokens_seen": 50194150, + "step": 2373 + }, + { + "epoch": 0.28545662237720193, + "flos": 14511875181960.0, + "grad_norm": 2.1765227501081355, + "learning_rate": 3.353778697750527e-06, + "loss": 0.8216, + "num_input_tokens_seen": 50212275, + "step": 2374 + }, + { + "epoch": 0.28557686526784104, + "flos": 17485296155040.0, + "grad_norm": 2.2021093529998867, + "learning_rate": 3.353205205904622e-06, + "loss": 0.8709, + "num_input_tokens_seen": 50231105, + "step": 2375 + }, + { + "epoch": 0.28569710815848015, + "flos": 33034493063160.0, + "grad_norm": 9.36538488050336, + "learning_rate": 3.3526315087785637e-06, + "loss": 0.7111, + "num_input_tokens_seen": 50251940, + "step": 2376 + }, + { + "epoch": 0.2858173510491192, + "flos": 19668393156240.0, + "grad_norm": 1.9816194761393322, + "learning_rate": 3.3520576064593805e-06, + "loss": 0.7952, + "num_input_tokens_seen": 50271615, + "step": 2377 + }, + { + "epoch": 0.2859375939397583, + "flos": 16944918805080.0, + "grad_norm": 2.02172814709969, + "learning_rate": 3.3514834990341337e-06, + "loss": 0.8102, + "num_input_tokens_seen": 50291660, + "step": 2378 + }, + { + "epoch": 0.2860578368303974, + "flos": 8789972980680.0, + "grad_norm": 4.297382338656826, + "learning_rate": 3.3509091865899144e-06, + "loss": 0.9232, + "num_input_tokens_seen": 50306570, + "step": 2379 + }, + { + "epoch": 0.2861780797210365, + "flos": 14568632507520.0, + "grad_norm": 2.4895312464757207, + "learning_rate": 3.350334669213846e-06, + "loss": 0.6861, + "num_input_tokens_seen": 50323695, + "step": 2380 + }, + { + "epoch": 0.2862983226116756, + "flos": 20211651562560.0, + "grad_norm": 2.7510562082176193, + "learning_rate": 3.3497599469930816e-06, + "loss": 0.7547, + "num_input_tokens_seen": 50341625, + "step": 2381 + }, + { + "epoch": 0.28641856550231465, + "flos": 16134526909920.0, + "grad_norm": 2.889693321402853, + "learning_rate": 3.349185020014807e-06, + "loss": 0.8145, + "num_input_tokens_seen": 50358610, + "step": 2382 + }, + { + "epoch": 0.28653880839295376, + "flos": 16374405664560.0, + "grad_norm": 3.151832737681878, + "learning_rate": 3.348609888366237e-06, + "loss": 0.7306, + "num_input_tokens_seen": 50377345, + "step": 2383 + }, + { + "epoch": 0.28665905128359287, + "flos": 17107054172040.0, + "grad_norm": 4.561077575519007, + "learning_rate": 3.348034552134619e-06, + "loss": 0.6107, + "num_input_tokens_seen": 50396470, + "step": 2384 + }, + { + "epoch": 0.2867792941742319, + "flos": 15266301250560.0, + "grad_norm": 3.9620196126484206, + "learning_rate": 3.3474590114072316e-06, + "loss": 0.8329, + "num_input_tokens_seen": 50414190, + "step": 2385 + }, + { + "epoch": 0.28689953706487104, + "flos": 15105368962080.0, + "grad_norm": 2.496210468649706, + "learning_rate": 3.3468832662713836e-06, + "loss": 0.8057, + "num_input_tokens_seen": 50432155, + "step": 2386 + }, + { + "epoch": 0.28701977995551015, + "flos": 9193633420200.0, + "grad_norm": 3.0902182717374482, + "learning_rate": 3.346307316814415e-06, + "loss": 0.8255, + "num_input_tokens_seen": 50447045, + "step": 2387 + }, + { + "epoch": 0.2871400228461492, + "flos": 15541444709280.0, + "grad_norm": 4.417827573589841, + "learning_rate": 3.3457311631236965e-06, + "loss": 0.7222, + "num_input_tokens_seen": 50467750, + "step": 2388 + }, + { + "epoch": 0.2872602657367883, + "flos": 18403123989000.0, + "grad_norm": 2.4548238715177386, + "learning_rate": 3.345154805286631e-06, + "loss": 0.83, + "num_input_tokens_seen": 50487730, + "step": 2389 + }, + { + "epoch": 0.2873805086274274, + "flos": 12131663049360.0, + "grad_norm": 3.97839265388932, + "learning_rate": 3.344578243390651e-06, + "loss": 0.7383, + "num_input_tokens_seen": 50503010, + "step": 2390 + }, + { + "epoch": 0.2875007515180665, + "flos": 12705152226120.0, + "grad_norm": 2.938924890244398, + "learning_rate": 3.3440014775232206e-06, + "loss": 0.7571, + "num_input_tokens_seen": 50520785, + "step": 2391 + }, + { + "epoch": 0.2876209944087056, + "flos": 17155105008600.0, + "grad_norm": 2.6332382627620543, + "learning_rate": 3.343424507771834e-06, + "loss": 0.705, + "num_input_tokens_seen": 50538715, + "step": 2392 + }, + { + "epoch": 0.2877412372993447, + "flos": 9978353579160.0, + "grad_norm": 2.9889574270729407, + "learning_rate": 3.342847334224018e-06, + "loss": 0.8562, + "num_input_tokens_seen": 50555835, + "step": 2393 + }, + { + "epoch": 0.28786148018998375, + "flos": 42803171489520.0, + "grad_norm": 0.876513738516915, + "learning_rate": 3.342269956967329e-06, + "loss": 0.6583, + "num_input_tokens_seen": 50617460, + "step": 2394 + }, + { + "epoch": 0.28798172308062286, + "flos": 17155864847640.0, + "grad_norm": 4.8630659308509, + "learning_rate": 3.341692376089355e-06, + "loss": 0.6928, + "num_input_tokens_seen": 50632735, + "step": 2395 + }, + { + "epoch": 0.288101965971262, + "flos": 18394987379280.0, + "grad_norm": 3.347280189641846, + "learning_rate": 3.3411145916777146e-06, + "loss": 0.8146, + "num_input_tokens_seen": 50646615, + "step": 2396 + }, + { + "epoch": 0.28822220886190103, + "flos": 11841074021520.0, + "grad_norm": 3.285200084706808, + "learning_rate": 3.3405366038200566e-06, + "loss": 0.8839, + "num_input_tokens_seen": 50665270, + "step": 2397 + }, + { + "epoch": 0.28834245175254014, + "flos": 17971227356520.0, + "grad_norm": 3.3691498244892517, + "learning_rate": 3.3399584126040617e-06, + "loss": 0.8241, + "num_input_tokens_seen": 50684490, + "step": 2398 + }, + { + "epoch": 0.2884626946431792, + "flos": 17997627271800.0, + "grad_norm": 3.009196490641166, + "learning_rate": 3.339380018117441e-06, + "loss": 0.8825, + "num_input_tokens_seen": 50705045, + "step": 2399 + }, + { + "epoch": 0.2885829375338183, + "flos": 12057275617440.0, + "grad_norm": 6.468076809117419, + "learning_rate": 3.3388014204479366e-06, + "loss": 0.7558, + "num_input_tokens_seen": 50722570, + "step": 2400 + }, + { + "epoch": 0.2887031804244574, + "flos": 17620113467880.0, + "grad_norm": 3.3691702048067445, + "learning_rate": 3.338222619683321e-06, + "loss": 0.896, + "num_input_tokens_seen": 50742255, + "step": 2401 + }, + { + "epoch": 0.2888234233150965, + "flos": 17079356198400.0, + "grad_norm": 47.84081575750721, + "learning_rate": 3.337643615911398e-06, + "loss": 0.7154, + "num_input_tokens_seen": 50761600, + "step": 2402 + }, + { + "epoch": 0.2889436662057356, + "flos": 16296440657160.0, + "grad_norm": 3.804707803221623, + "learning_rate": 3.3370644092200026e-06, + "loss": 0.7689, + "num_input_tokens_seen": 50778595, + "step": 2403 + }, + { + "epoch": 0.2890639090963747, + "flos": 15811205974800.0, + "grad_norm": 9.973349726794364, + "learning_rate": 3.3364849996969985e-06, + "loss": 0.7748, + "num_input_tokens_seen": 50798335, + "step": 2404 + }, + { + "epoch": 0.28918415198701375, + "flos": 20968420468200.0, + "grad_norm": 4.0558165519957114, + "learning_rate": 3.335905387430283e-06, + "loss": 0.8402, + "num_input_tokens_seen": 50819490, + "step": 2405 + }, + { + "epoch": 0.28930439487765286, + "flos": 16053680846160.0, + "grad_norm": 2.885534639850279, + "learning_rate": 3.335325572507782e-06, + "loss": 0.8052, + "num_input_tokens_seen": 50839710, + "step": 2406 + }, + { + "epoch": 0.28942463776829197, + "flos": 14083144545480.0, + "grad_norm": 2.9087336602631146, + "learning_rate": 3.3347455550174537e-06, + "loss": 0.7249, + "num_input_tokens_seen": 50858770, + "step": 2407 + }, + { + "epoch": 0.289544880658931, + "flos": 10651680304320.0, + "grad_norm": 3.313758035403265, + "learning_rate": 3.3341653350472864e-06, + "loss": 0.6696, + "num_input_tokens_seen": 50875320, + "step": 2408 + }, + { + "epoch": 0.28966512354957014, + "flos": 20995295282880.0, + "grad_norm": 4.383878581628035, + "learning_rate": 3.333584912685298e-06, + "loss": 0.6851, + "num_input_tokens_seen": 50893660, + "step": 2409 + }, + { + "epoch": 0.28978536644020925, + "flos": 47710186081320.0, + "grad_norm": 0.8523109716446599, + "learning_rate": 3.3330042880195385e-06, + "loss": 0.5808, + "num_input_tokens_seen": 50947730, + "step": 2410 + }, + { + "epoch": 0.2899056093308483, + "flos": 13596643464720.0, + "grad_norm": 10.355777812131654, + "learning_rate": 3.3324234611380888e-06, + "loss": 0.764, + "num_input_tokens_seen": 50966180, + "step": 2411 + }, + { + "epoch": 0.2900258522214874, + "flos": 16756098583200.0, + "grad_norm": 2.9253516091836747, + "learning_rate": 3.3318424321290596e-06, + "loss": 0.8054, + "num_input_tokens_seen": 50985615, + "step": 2412 + }, + { + "epoch": 0.2901460951121265, + "flos": 52438507147080.0, + "grad_norm": 0.8399639664596104, + "learning_rate": 3.3312612010805917e-06, + "loss": 0.6383, + "num_input_tokens_seen": 51044910, + "step": 2413 + }, + { + "epoch": 0.2902663380027656, + "flos": 23614468031280.0, + "grad_norm": 2.5738551066772306, + "learning_rate": 3.330679768080858e-06, + "loss": 0.6916, + "num_input_tokens_seen": 51068515, + "step": 2414 + }, + { + "epoch": 0.2903865808934047, + "flos": 21535387693200.0, + "grad_norm": 3.23760898362467, + "learning_rate": 3.3300981332180627e-06, + "loss": 0.8241, + "num_input_tokens_seen": 51087440, + "step": 2415 + }, + { + "epoch": 0.29050682378404374, + "flos": 12459732978480.0, + "grad_norm": 3.2836313961175434, + "learning_rate": 3.3295162965804373e-06, + "loss": 0.7932, + "num_input_tokens_seen": 51105655, + "step": 2416 + }, + { + "epoch": 0.29062706667468285, + "flos": 13029422960040.0, + "grad_norm": 4.359954467325699, + "learning_rate": 3.328934258256247e-06, + "loss": 0.7562, + "num_input_tokens_seen": 51123440, + "step": 2417 + }, + { + "epoch": 0.29074730956532197, + "flos": 17781837255360.0, + "grad_norm": 5.859647454603078, + "learning_rate": 3.3283520183337856e-06, + "loss": 0.6499, + "num_input_tokens_seen": 51142865, + "step": 2418 + }, + { + "epoch": 0.290867552455961, + "flos": 16346232791520.0, + "grad_norm": 3.4901667472550675, + "learning_rate": 3.3277695769013797e-06, + "loss": 0.6841, + "num_input_tokens_seen": 51162030, + "step": 2419 + }, + { + "epoch": 0.29098779534660013, + "flos": 16972585118760.0, + "grad_norm": 4.024494987697942, + "learning_rate": 3.327186934047385e-06, + "loss": 0.762, + "num_input_tokens_seen": 51180445, + "step": 2420 + }, + { + "epoch": 0.29110803823723924, + "flos": 11139257823720.0, + "grad_norm": 11.918470624060856, + "learning_rate": 3.3266040898601877e-06, + "loss": 0.629, + "num_input_tokens_seen": 51198000, + "step": 2421 + }, + { + "epoch": 0.2912282811278783, + "flos": 16534356494280.0, + "grad_norm": 3.1890160938047036, + "learning_rate": 3.3260210444282045e-06, + "loss": 0.758, + "num_input_tokens_seen": 51215675, + "step": 2422 + }, + { + "epoch": 0.2913485240185174, + "flos": 17943117803400.0, + "grad_norm": 4.825422895013727, + "learning_rate": 3.325437797839883e-06, + "loss": 0.7063, + "num_input_tokens_seen": 51233765, + "step": 2423 + }, + { + "epoch": 0.2914687669091565, + "flos": 13083267569280.0, + "grad_norm": 8.791758267134872, + "learning_rate": 3.3248543501837015e-06, + "loss": 0.7262, + "num_input_tokens_seen": 51250690, + "step": 2424 + }, + { + "epoch": 0.2915890097997956, + "flos": 16486463957520.0, + "grad_norm": 3.799139658257424, + "learning_rate": 3.3242707015481684e-06, + "loss": 0.7573, + "num_input_tokens_seen": 51270345, + "step": 2425 + }, + { + "epoch": 0.2917092526904347, + "flos": 10059579562440.0, + "grad_norm": 2.411393155781883, + "learning_rate": 3.323686852021823e-06, + "loss": 0.7938, + "num_input_tokens_seen": 51287575, + "step": 2426 + }, + { + "epoch": 0.2918294955810738, + "flos": 16594121516040.0, + "grad_norm": 3.9762827357000248, + "learning_rate": 3.323102801693235e-06, + "loss": 0.7867, + "num_input_tokens_seen": 51306060, + "step": 2427 + }, + { + "epoch": 0.29194973847171285, + "flos": 17158714244040.0, + "grad_norm": 4.259053404681686, + "learning_rate": 3.322518550651003e-06, + "loss": 0.7774, + "num_input_tokens_seen": 51325090, + "step": 2428 + }, + { + "epoch": 0.29206998136235196, + "flos": 16027439230680.0, + "grad_norm": 4.757990490820908, + "learning_rate": 3.3219340989837586e-06, + "loss": 0.7899, + "num_input_tokens_seen": 51344800, + "step": 2429 + }, + { + "epoch": 0.292190224252991, + "flos": 16993666160760.0, + "grad_norm": 2.886337317017015, + "learning_rate": 3.3213494467801625e-06, + "loss": 0.7897, + "num_input_tokens_seen": 51363695, + "step": 2430 + }, + { + "epoch": 0.2923104671436301, + "flos": 15162759486840.0, + "grad_norm": 4.484570861154478, + "learning_rate": 3.3207645941289063e-06, + "loss": 0.6951, + "num_input_tokens_seen": 51381760, + "step": 2431 + }, + { + "epoch": 0.29243071003426924, + "flos": 26314265223720.0, + "grad_norm": 3.4872973037936394, + "learning_rate": 3.320179541118711e-06, + "loss": 0.7897, + "num_input_tokens_seen": 51403980, + "step": 2432 + }, + { + "epoch": 0.2925509529249083, + "flos": 42804849467400.0, + "grad_norm": 1.0322599955966598, + "learning_rate": 3.3195942878383293e-06, + "loss": 0.6367, + "num_input_tokens_seen": 51459800, + "step": 2433 + }, + { + "epoch": 0.2926711958155474, + "flos": 15648405748680.0, + "grad_norm": 4.015804064131384, + "learning_rate": 3.319008834376543e-06, + "loss": 0.7618, + "num_input_tokens_seen": 51479210, + "step": 2434 + }, + { + "epoch": 0.2927914387061865, + "flos": 16972268519160.0, + "grad_norm": 3.326108222930296, + "learning_rate": 3.3184231808221654e-06, + "loss": 0.8664, + "num_input_tokens_seen": 51493255, + "step": 2435 + }, + { + "epoch": 0.29291168159682557, + "flos": 16431764529360.0, + "grad_norm": 5.762906061458347, + "learning_rate": 3.3178373272640394e-06, + "loss": 0.5995, + "num_input_tokens_seen": 51512070, + "step": 2436 + }, + { + "epoch": 0.2930319244874647, + "flos": 15479716770000.0, + "grad_norm": 6.5441722890283325, + "learning_rate": 3.3172512737910387e-06, + "loss": 0.8409, + "num_input_tokens_seen": 51529300, + "step": 2437 + }, + { + "epoch": 0.2931521673781038, + "flos": 23020024452360.0, + "grad_norm": 3.154660268347672, + "learning_rate": 3.3166650204920674e-06, + "loss": 0.8598, + "num_input_tokens_seen": 51550190, + "step": 2438 + }, + { + "epoch": 0.29327241026874284, + "flos": 17722673772840.0, + "grad_norm": 3.4000016044164125, + "learning_rate": 3.316078567456059e-06, + "loss": 0.8092, + "num_input_tokens_seen": 51567750, + "step": 2439 + }, + { + "epoch": 0.29339265315938196, + "flos": 17754202601640.0, + "grad_norm": 2.457676837569831, + "learning_rate": 3.3154919147719786e-06, + "loss": 0.7497, + "num_input_tokens_seen": 51588485, + "step": 2440 + }, + { + "epoch": 0.29351289605002107, + "flos": 12354861496440.0, + "grad_norm": 3.7559985186134157, + "learning_rate": 3.31490506252882e-06, + "loss": 0.8556, + "num_input_tokens_seen": 51607585, + "step": 2441 + }, + { + "epoch": 0.2936331389406601, + "flos": 14488071383400.0, + "grad_norm": 10.333736493849473, + "learning_rate": 3.31431801081561e-06, + "loss": 0.8365, + "num_input_tokens_seen": 51626240, + "step": 2442 + }, + { + "epoch": 0.29375338183129923, + "flos": 52671203582160.0, + "grad_norm": 0.9391543732193662, + "learning_rate": 3.313730759721402e-06, + "loss": 0.6795, + "num_input_tokens_seen": 51688890, + "step": 2443 + }, + { + "epoch": 0.29387362472193834, + "flos": 16135001809320.0, + "grad_norm": 3.5071872757204026, + "learning_rate": 3.313143309335282e-06, + "loss": 0.8395, + "num_input_tokens_seen": 51707100, + "step": 2444 + }, + { + "epoch": 0.2939938676125774, + "flos": 16809848212560.0, + "grad_norm": 3.5532361767233662, + "learning_rate": 3.3125556597463665e-06, + "loss": 0.8207, + "num_input_tokens_seen": 51726125, + "step": 2445 + }, + { + "epoch": 0.2941141105032165, + "flos": 23020784291400.0, + "grad_norm": 3.0045354256156305, + "learning_rate": 3.311967811043801e-06, + "loss": 0.6443, + "num_input_tokens_seen": 51747765, + "step": 2446 + }, + { + "epoch": 0.29423435339385556, + "flos": 16999143333840.0, + "grad_norm": 4.402788683011695, + "learning_rate": 3.3113797633167617e-06, + "loss": 0.7849, + "num_input_tokens_seen": 51765780, + "step": 2447 + }, + { + "epoch": 0.2943545962844947, + "flos": 19693463353200.0, + "grad_norm": 3.6613507026636367, + "learning_rate": 3.310791516654455e-06, + "loss": 0.6618, + "num_input_tokens_seen": 51782560, + "step": 2448 + }, + { + "epoch": 0.2944748391751338, + "flos": 14786195481720.0, + "grad_norm": 2.7505463144202817, + "learning_rate": 3.3102030711461177e-06, + "loss": 0.7759, + "num_input_tokens_seen": 51801855, + "step": 2449 + }, + { + "epoch": 0.29459508206577284, + "flos": 11625315665040.0, + "grad_norm": 3.0286089216209593, + "learning_rate": 3.3096144268810156e-06, + "loss": 0.6607, + "num_input_tokens_seen": 51820335, + "step": 2450 + }, + { + "epoch": 0.29471532495641195, + "flos": 14920126315680.0, + "grad_norm": 3.2870681769180012, + "learning_rate": 3.3090255839484462e-06, + "loss": 0.7081, + "num_input_tokens_seen": 51838050, + "step": 2451 + }, + { + "epoch": 0.29483556784705106, + "flos": 14893283160960.0, + "grad_norm": 2.246768538243519, + "learning_rate": 3.3084365424377366e-06, + "loss": 0.8356, + "num_input_tokens_seen": 51856535, + "step": 2452 + }, + { + "epoch": 0.2949558107376901, + "flos": 50551571326680.0, + "grad_norm": 0.7857783116535828, + "learning_rate": 3.307847302438245e-06, + "loss": 0.5759, + "num_input_tokens_seen": 51910235, + "step": 2453 + }, + { + "epoch": 0.2950760536283292, + "flos": 11733479782920.0, + "grad_norm": 4.237175835107293, + "learning_rate": 3.3072578640393562e-06, + "loss": 0.7507, + "num_input_tokens_seen": 51927290, + "step": 2454 + }, + { + "epoch": 0.29519629651896834, + "flos": 14972166307200.0, + "grad_norm": 2.4310422597748143, + "learning_rate": 3.3066682273304886e-06, + "loss": 0.7772, + "num_input_tokens_seen": 51944655, + "step": 2455 + }, + { + "epoch": 0.2953165394096074, + "flos": 13812623440920.0, + "grad_norm": 2.740882671774655, + "learning_rate": 3.3060783924010904e-06, + "loss": 0.7487, + "num_input_tokens_seen": 51962300, + "step": 2456 + }, + { + "epoch": 0.2954367823002465, + "flos": 15075676410960.0, + "grad_norm": 3.2016255660615265, + "learning_rate": 3.3054883593406387e-06, + "loss": 0.834, + "num_input_tokens_seen": 51976770, + "step": 2457 + }, + { + "epoch": 0.2955570251908856, + "flos": 22885112159640.0, + "grad_norm": 5.362806995202964, + "learning_rate": 3.3048981282386404e-06, + "loss": 0.6222, + "num_input_tokens_seen": 51997800, + "step": 2458 + }, + { + "epoch": 0.29567726808152467, + "flos": 15835484672760.0, + "grad_norm": 3.4715084071232694, + "learning_rate": 3.304307699184634e-06, + "loss": 0.8063, + "num_input_tokens_seen": 52016110, + "step": 2459 + }, + { + "epoch": 0.2957975109721638, + "flos": 17755310700240.0, + "grad_norm": 2.380910045156566, + "learning_rate": 3.3037170722681866e-06, + "loss": 0.7819, + "num_input_tokens_seen": 52036665, + "step": 2460 + }, + { + "epoch": 0.29591775386280283, + "flos": 9788995137960.0, + "grad_norm": 3.8144842518651263, + "learning_rate": 3.3031262475788956e-06, + "loss": 0.6671, + "num_input_tokens_seen": 52053325, + "step": 2461 + }, + { + "epoch": 0.29603799675344195, + "flos": 12949305075360.0, + "grad_norm": 2.4538610046743132, + "learning_rate": 3.3025352252063897e-06, + "loss": 0.719, + "num_input_tokens_seen": 52071740, + "step": 2462 + }, + { + "epoch": 0.29615823964408106, + "flos": 16674777620040.0, + "grad_norm": 3.0321539091537533, + "learning_rate": 3.3019440052403252e-06, + "loss": 0.7287, + "num_input_tokens_seen": 52091325, + "step": 2463 + }, + { + "epoch": 0.2962784825347201, + "flos": 17215313269800.0, + "grad_norm": 2.9061300601615727, + "learning_rate": 3.30135258777039e-06, + "loss": 0.6916, + "num_input_tokens_seen": 52110415, + "step": 2464 + }, + { + "epoch": 0.2963987254253592, + "flos": 11922743244240.0, + "grad_norm": 3.127396999404189, + "learning_rate": 3.3007609728863024e-06, + "loss": 0.6823, + "num_input_tokens_seen": 52128225, + "step": 2465 + }, + { + "epoch": 0.29651896831599833, + "flos": 24584589136440.0, + "grad_norm": 3.160721133073066, + "learning_rate": 3.300169160677809e-06, + "loss": 0.7167, + "num_input_tokens_seen": 52151860, + "step": 2466 + }, + { + "epoch": 0.2966392112066374, + "flos": 17430691706760.0, + "grad_norm": 3.406573638939797, + "learning_rate": 3.2995771512346878e-06, + "loss": 0.732, + "num_input_tokens_seen": 52169930, + "step": 2467 + }, + { + "epoch": 0.2967594540972765, + "flos": 14569012427040.0, + "grad_norm": 3.52486348530363, + "learning_rate": 3.298984944646746e-06, + "loss": 0.7013, + "num_input_tokens_seen": 52188330, + "step": 2468 + }, + { + "epoch": 0.2968796969879156, + "flos": 17377068717240.0, + "grad_norm": 2.465402385684707, + "learning_rate": 3.298392541003822e-06, + "loss": 0.801, + "num_input_tokens_seen": 52207455, + "step": 2469 + }, + { + "epoch": 0.29699993987855466, + "flos": 16756066923240.0, + "grad_norm": 2.2648994073151125, + "learning_rate": 3.2977999403957806e-06, + "loss": 0.8738, + "num_input_tokens_seen": 52225935, + "step": 2470 + }, + { + "epoch": 0.2971201827691938, + "flos": 24850582866720.0, + "grad_norm": 3.294433557564936, + "learning_rate": 3.2972071429125207e-06, + "loss": 0.6495, + "num_input_tokens_seen": 52246875, + "step": 2471 + }, + { + "epoch": 0.2972404256598329, + "flos": 16134875169480.0, + "grad_norm": 2.564719936555707, + "learning_rate": 3.2966141486439682e-06, + "loss": 0.8639, + "num_input_tokens_seen": 52265785, + "step": 2472 + }, + { + "epoch": 0.29736066855047194, + "flos": 23479492418640.0, + "grad_norm": 3.2953387835030075, + "learning_rate": 3.29602095768008e-06, + "loss": 0.6371, + "num_input_tokens_seen": 52286020, + "step": 2473 + }, + { + "epoch": 0.29748091144111105, + "flos": 24613046949120.0, + "grad_norm": 3.201510287557716, + "learning_rate": 3.2954275701108437e-06, + "loss": 0.6291, + "num_input_tokens_seen": 52306920, + "step": 2474 + }, + { + "epoch": 0.29760115433175016, + "flos": 30365781460080.0, + "grad_norm": 2.4715960551400658, + "learning_rate": 3.294833986026275e-06, + "loss": 0.6731, + "num_input_tokens_seen": 52329880, + "step": 2475 + }, + { + "epoch": 0.2977213972223892, + "flos": 17939096988480.0, + "grad_norm": 2.721791374945333, + "learning_rate": 3.29424020551642e-06, + "loss": 0.8343, + "num_input_tokens_seen": 52348235, + "step": 2476 + }, + { + "epoch": 0.2978416401130283, + "flos": 15565153527960.0, + "grad_norm": 3.1760963949857848, + "learning_rate": 3.2936462286713546e-06, + "loss": 0.6932, + "num_input_tokens_seen": 52366305, + "step": 2477 + }, + { + "epoch": 0.2979618830036674, + "flos": 18885762554640.0, + "grad_norm": 2.8137106938711614, + "learning_rate": 3.2930520555811846e-06, + "loss": 0.7559, + "num_input_tokens_seen": 52385650, + "step": 2478 + }, + { + "epoch": 0.2980821258943065, + "flos": 17187678616080.0, + "grad_norm": 3.0490226805716985, + "learning_rate": 3.292457686336046e-06, + "loss": 0.7793, + "num_input_tokens_seen": 52404690, + "step": 2479 + }, + { + "epoch": 0.2982023687849456, + "flos": 51441294098880.0, + "grad_norm": 0.9338938657652086, + "learning_rate": 3.291863121026105e-06, + "loss": 0.6431, + "num_input_tokens_seen": 52468190, + "step": 2480 + }, + { + "epoch": 0.29832261167558466, + "flos": 21886374942000.0, + "grad_norm": 2.7721268256708167, + "learning_rate": 3.2912683597415547e-06, + "loss": 0.7443, + "num_input_tokens_seen": 52491995, + "step": 2481 + }, + { + "epoch": 0.29844285456622377, + "flos": 24908669910600.0, + "grad_norm": 2.8108007596033495, + "learning_rate": 3.2906734025726213e-06, + "loss": 0.7559, + "num_input_tokens_seen": 52510980, + "step": 2482 + }, + { + "epoch": 0.2985630974568629, + "flos": 17483459877360.0, + "grad_norm": 2.7889613712900534, + "learning_rate": 3.290078249609559e-06, + "loss": 0.852, + "num_input_tokens_seen": 52530120, + "step": 2483 + }, + { + "epoch": 0.29868334034750194, + "flos": 15945991627680.0, + "grad_norm": 2.771630786687217, + "learning_rate": 3.2894829009426514e-06, + "loss": 0.8486, + "num_input_tokens_seen": 52547675, + "step": 2484 + }, + { + "epoch": 0.29880358323814105, + "flos": 18808240786680.0, + "grad_norm": 2.6818610182046303, + "learning_rate": 3.288887356662213e-06, + "loss": 0.7555, + "num_input_tokens_seen": 52568730, + "step": 2485 + }, + { + "epoch": 0.29892382612878016, + "flos": 52365671053200.0, + "grad_norm": 0.7979393003820959, + "learning_rate": 3.288291616858588e-06, + "loss": 0.6048, + "num_input_tokens_seen": 52623840, + "step": 2486 + }, + { + "epoch": 0.2990440690194192, + "flos": 18670510757520.0, + "grad_norm": 2.724962874693347, + "learning_rate": 3.287695681622149e-06, + "loss": 0.7599, + "num_input_tokens_seen": 52642910, + "step": 2487 + }, + { + "epoch": 0.2991643119100583, + "flos": 17376657137760.0, + "grad_norm": 2.921461433068598, + "learning_rate": 3.2870995510432982e-06, + "loss": 0.7963, + "num_input_tokens_seen": 52661110, + "step": 2488 + }, + { + "epoch": 0.29928455480069743, + "flos": 20103962344080.0, + "grad_norm": 4.717511753538659, + "learning_rate": 3.2865032252124697e-06, + "loss": 0.7429, + "num_input_tokens_seen": 52681345, + "step": 2489 + }, + { + "epoch": 0.2994047976913365, + "flos": 24748149201600.0, + "grad_norm": 2.811559647078178, + "learning_rate": 3.2859067042201243e-06, + "loss": 0.7626, + "num_input_tokens_seen": 52703105, + "step": 2490 + }, + { + "epoch": 0.2995250405819756, + "flos": 12219632604120.0, + "grad_norm": 3.227330003393775, + "learning_rate": 3.2853099881567544e-06, + "loss": 0.7529, + "num_input_tokens_seen": 52721225, + "step": 2491 + }, + { + "epoch": 0.29964528347261465, + "flos": 16807695335280.0, + "grad_norm": 2.408011052128468, + "learning_rate": 3.284713077112881e-06, + "loss": 0.7716, + "num_input_tokens_seen": 52740375, + "step": 2492 + }, + { + "epoch": 0.29976552636325376, + "flos": 12347263106040.0, + "grad_norm": 3.0672956015240698, + "learning_rate": 3.284115971179056e-06, + "loss": 0.8267, + "num_input_tokens_seen": 52754125, + "step": 2493 + }, + { + "epoch": 0.2998857692538929, + "flos": 12436182459600.0, + "grad_norm": 3.032285061331845, + "learning_rate": 3.283518670445859e-06, + "loss": 0.778, + "num_input_tokens_seen": 52771755, + "step": 2494 + }, + { + "epoch": 0.30000601214453193, + "flos": 50758627685520.0, + "grad_norm": 1.9583798328587652, + "learning_rate": 3.2829211750038995e-06, + "loss": 0.575, + "num_input_tokens_seen": 52840105, + "step": 2495 + }, + { + "epoch": 0.30012625503517104, + "flos": 12840792697920.0, + "grad_norm": 2.3099782442331716, + "learning_rate": 3.2823234849438183e-06, + "loss": 0.8619, + "num_input_tokens_seen": 52857860, + "step": 2496 + }, + { + "epoch": 0.30024649792581015, + "flos": 15541634669040.0, + "grad_norm": 3.865488746603988, + "learning_rate": 3.2817256003562836e-06, + "loss": 0.7219, + "num_input_tokens_seen": 52877955, + "step": 2497 + }, + { + "epoch": 0.3003667408164492, + "flos": 16837166266680.0, + "grad_norm": 2.4940385312160736, + "learning_rate": 3.281127521331995e-06, + "loss": 0.6452, + "num_input_tokens_seen": 52898855, + "step": 2498 + }, + { + "epoch": 0.3004869837070883, + "flos": 47354639798280.0, + "grad_norm": 0.9045327128287971, + "learning_rate": 3.2805292479616798e-06, + "loss": 0.6365, + "num_input_tokens_seen": 52957440, + "step": 2499 + }, + { + "epoch": 0.30060722659772743, + "flos": 19238712720960.0, + "grad_norm": 5.16678038851371, + "learning_rate": 3.2799307803360955e-06, + "loss": 0.8852, + "num_input_tokens_seen": 52973090, + "step": 2500 + }, + { + "epoch": 0.3007274694883665, + "flos": 18292965293640.0, + "grad_norm": 2.5329135090985586, + "learning_rate": 3.27933211854603e-06, + "loss": 0.8029, + "num_input_tokens_seen": 52991865, + "step": 2501 + }, + { + "epoch": 0.3008477123790056, + "flos": 12435644240280.0, + "grad_norm": 2.1393443265382635, + "learning_rate": 3.278733262682299e-06, + "loss": 0.8566, + "num_input_tokens_seen": 53009440, + "step": 2502 + }, + { + "epoch": 0.3009679552696447, + "flos": 15729251812440.0, + "grad_norm": 4.18817884721682, + "learning_rate": 3.2781342128357484e-06, + "loss": 0.7968, + "num_input_tokens_seen": 53028515, + "step": 2503 + }, + { + "epoch": 0.30108819816028376, + "flos": 15452588675640.0, + "grad_norm": 3.871728192641223, + "learning_rate": 3.2775349690972547e-06, + "loss": 0.7823, + "num_input_tokens_seen": 53042385, + "step": 2504 + }, + { + "epoch": 0.30120844105092287, + "flos": 52456490004360.0, + "grad_norm": 0.767728873245092, + "learning_rate": 3.276935531557722e-06, + "loss": 0.5556, + "num_input_tokens_seen": 53107325, + "step": 2505 + }, + { + "epoch": 0.301328683941562, + "flos": 14810030940240.0, + "grad_norm": 3.702983740836557, + "learning_rate": 3.2763359003080837e-06, + "loss": 0.7833, + "num_input_tokens_seen": 53124000, + "step": 2506 + }, + { + "epoch": 0.30144892683220104, + "flos": 52102020159960.0, + "grad_norm": 0.9344237400981206, + "learning_rate": 3.2757360754393047e-06, + "loss": 0.6693, + "num_input_tokens_seen": 53187790, + "step": 2507 + }, + { + "epoch": 0.30156916972284015, + "flos": 16701905714400.0, + "grad_norm": 3.4118522280750527, + "learning_rate": 3.2751360570423767e-06, + "loss": 0.6052, + "num_input_tokens_seen": 53205895, + "step": 2508 + }, + { + "epoch": 0.3016894126134792, + "flos": 21940631130720.0, + "grad_norm": 2.8647656931774788, + "learning_rate": 3.2745358452083236e-06, + "loss": 0.7473, + "num_input_tokens_seen": 53228515, + "step": 2509 + }, + { + "epoch": 0.3018096555041183, + "flos": 15759292623120.0, + "grad_norm": 1.5675103536747574, + "learning_rate": 3.2739354400281955e-06, + "loss": 0.8071, + "num_input_tokens_seen": 53249455, + "step": 2510 + }, + { + "epoch": 0.3019298983947574, + "flos": 43584915612240.0, + "grad_norm": 0.8955025875558962, + "learning_rate": 3.2733348415930744e-06, + "loss": 0.683, + "num_input_tokens_seen": 53311045, + "step": 2511 + }, + { + "epoch": 0.3020501412853965, + "flos": 25288558211520.0, + "grad_norm": 3.1483181535615046, + "learning_rate": 3.27273404999407e-06, + "loss": 0.7879, + "num_input_tokens_seen": 53332985, + "step": 2512 + }, + { + "epoch": 0.3021703841760356, + "flos": 52367823930480.0, + "grad_norm": 0.7971024656085117, + "learning_rate": 3.272133065322322e-06, + "loss": 0.6183, + "num_input_tokens_seen": 53390975, + "step": 2513 + }, + { + "epoch": 0.3022906270666747, + "flos": 15732892707840.0, + "grad_norm": 1.9108509758413341, + "learning_rate": 3.271531887669e-06, + "loss": 0.7745, + "num_input_tokens_seen": 53410755, + "step": 2514 + }, + { + "epoch": 0.30241086995731375, + "flos": 22851905352960.0, + "grad_norm": 3.5760562696787495, + "learning_rate": 3.2709305171253015e-06, + "loss": 0.6286, + "num_input_tokens_seen": 53430595, + "step": 2515 + }, + { + "epoch": 0.30253111284795287, + "flos": 17212970432760.0, + "grad_norm": 2.168049248927793, + "learning_rate": 3.2703289537824536e-06, + "loss": 0.7665, + "num_input_tokens_seen": 53450115, + "step": 2516 + }, + { + "epoch": 0.302651355738592, + "flos": 13669954458000.0, + "grad_norm": 3.4191111372150407, + "learning_rate": 3.269727197731714e-06, + "loss": 0.7719, + "num_input_tokens_seen": 53462600, + "step": 2517 + }, + { + "epoch": 0.30277159862923103, + "flos": 16404604775040.0, + "grad_norm": 2.351315608957255, + "learning_rate": 3.269125249064367e-06, + "loss": 0.7643, + "num_input_tokens_seen": 53482015, + "step": 2518 + }, + { + "epoch": 0.30289184151987014, + "flos": 16296757256760.0, + "grad_norm": 2.0861977404156775, + "learning_rate": 3.2685231078717297e-06, + "loss": 0.8108, + "num_input_tokens_seen": 53501925, + "step": 2519 + }, + { + "epoch": 0.30301208441050925, + "flos": 18481373936040.0, + "grad_norm": 2.8491389596831658, + "learning_rate": 3.267920774245145e-06, + "loss": 0.7131, + "num_input_tokens_seen": 53521050, + "step": 2520 + }, + { + "epoch": 0.3031323273011483, + "flos": 16863724481760.0, + "grad_norm": 2.4504147275897443, + "learning_rate": 3.2673182482759876e-06, + "loss": 0.8307, + "num_input_tokens_seen": 53539885, + "step": 2521 + }, + { + "epoch": 0.3032525701917874, + "flos": 13782709270080.0, + "grad_norm": 3.777608962743872, + "learning_rate": 3.266715530055659e-06, + "loss": 0.635, + "num_input_tokens_seen": 53557755, + "step": 2522 + }, + { + "epoch": 0.30337281308242653, + "flos": 12973108873920.0, + "grad_norm": 2.498610873525977, + "learning_rate": 3.2661126196755927e-06, + "loss": 0.7907, + "num_input_tokens_seen": 53576585, + "step": 2523 + }, + { + "epoch": 0.3034930559730656, + "flos": 42614161307880.0, + "grad_norm": 0.8424311326176344, + "learning_rate": 3.265509517227248e-06, + "loss": 0.5998, + "num_input_tokens_seen": 53633120, + "step": 2524 + }, + { + "epoch": 0.3036132988637047, + "flos": 10733064587400.0, + "grad_norm": 2.4512131159379216, + "learning_rate": 3.264906222802115e-06, + "loss": 0.7889, + "num_input_tokens_seen": 53650690, + "step": 2525 + }, + { + "epoch": 0.30373354175434375, + "flos": 15379214362440.0, + "grad_norm": 3.0448570415844687, + "learning_rate": 3.264302736491715e-06, + "loss": 0.7531, + "num_input_tokens_seen": 53670530, + "step": 2526 + }, + { + "epoch": 0.30385378464498286, + "flos": 15460725285360.0, + "grad_norm": 2.3198050512422466, + "learning_rate": 3.263699058387594e-06, + "loss": 0.8494, + "num_input_tokens_seen": 53687685, + "step": 2527 + }, + { + "epoch": 0.30397402753562197, + "flos": 15079887185640.0, + "grad_norm": 2.9469040028832754, + "learning_rate": 3.2630951885813315e-06, + "loss": 0.8925, + "num_input_tokens_seen": 53704800, + "step": 2528 + }, + { + "epoch": 0.304094270426261, + "flos": 10977597356160.0, + "grad_norm": 2.7371343150689604, + "learning_rate": 3.262491127164533e-06, + "loss": 0.7566, + "num_input_tokens_seen": 53723335, + "step": 2529 + }, + { + "epoch": 0.30421451331690014, + "flos": 10059579562440.0, + "grad_norm": 5.470218657424416, + "learning_rate": 3.2618868742288337e-06, + "loss": 0.7875, + "num_input_tokens_seen": 53739980, + "step": 2530 + }, + { + "epoch": 0.30433475620753925, + "flos": 12678657330960.0, + "grad_norm": 2.7775336306619236, + "learning_rate": 3.261282429865899e-06, + "loss": 0.7095, + "num_input_tokens_seen": 53757705, + "step": 2531 + }, + { + "epoch": 0.3044549990981783, + "flos": 13813035020400.0, + "grad_norm": 2.4151262328523475, + "learning_rate": 3.2606777941674225e-06, + "loss": 0.7128, + "num_input_tokens_seen": 53776080, + "step": 2532 + }, + { + "epoch": 0.3045752419888174, + "flos": 15405107718360.0, + "grad_norm": 4.644964071243666, + "learning_rate": 3.2600729672251276e-06, + "loss": 0.8247, + "num_input_tokens_seen": 53793515, + "step": 2533 + }, + { + "epoch": 0.3046954848794565, + "flos": 21346504151400.0, + "grad_norm": 6.294595234699462, + "learning_rate": 3.259467949130765e-06, + "loss": 0.6306, + "num_input_tokens_seen": 53814645, + "step": 2534 + }, + { + "epoch": 0.3048157277700956, + "flos": 14831998461120.0, + "grad_norm": 4.040351234124674, + "learning_rate": 3.2588627399761164e-06, + "loss": 0.817, + "num_input_tokens_seen": 53830360, + "step": 2535 + }, + { + "epoch": 0.3049359706607347, + "flos": 16641159233880.0, + "grad_norm": 2.090994772832933, + "learning_rate": 3.2582573398529903e-06, + "loss": 0.6999, + "num_input_tokens_seen": 53847435, + "step": 2536 + }, + { + "epoch": 0.3050562135513738, + "flos": 13454955940560.0, + "grad_norm": 3.408013263649718, + "learning_rate": 3.2576517488532265e-06, + "loss": 0.7287, + "num_input_tokens_seen": 53863505, + "step": 2537 + }, + { + "epoch": 0.30517645644201286, + "flos": 14887774327920.0, + "grad_norm": 2.4738909153368764, + "learning_rate": 3.257045967068692e-06, + "loss": 0.8476, + "num_input_tokens_seen": 53882480, + "step": 2538 + }, + { + "epoch": 0.30529669933265197, + "flos": 16054060765680.0, + "grad_norm": 2.6297761064675647, + "learning_rate": 3.2564399945912848e-06, + "loss": 0.8009, + "num_input_tokens_seen": 53901990, + "step": 2539 + }, + { + "epoch": 0.305416942223291, + "flos": 15972834782400.0, + "grad_norm": 2.6534578795251877, + "learning_rate": 3.2558338315129287e-06, + "loss": 0.8055, + "num_input_tokens_seen": 53919855, + "step": 2540 + }, + { + "epoch": 0.30553718511393013, + "flos": 24910126268760.0, + "grad_norm": 2.844682812968167, + "learning_rate": 3.2552274779255785e-06, + "loss": 0.7389, + "num_input_tokens_seen": 53940505, + "step": 2541 + }, + { + "epoch": 0.30565742800456924, + "flos": 16292704781880.0, + "grad_norm": 3.6365669128648195, + "learning_rate": 3.2546209339212184e-06, + "loss": 0.75, + "num_input_tokens_seen": 53959245, + "step": 2542 + }, + { + "epoch": 0.3057776708952083, + "flos": 16512800552880.0, + "grad_norm": 2.170953019760577, + "learning_rate": 3.25401419959186e-06, + "loss": 0.7517, + "num_input_tokens_seen": 53979575, + "step": 2543 + }, + { + "epoch": 0.3058979137858474, + "flos": 15946086607560.0, + "grad_norm": 2.8515730348157287, + "learning_rate": 3.253407275029545e-06, + "loss": 0.751, + "num_input_tokens_seen": 53998200, + "step": 2544 + }, + { + "epoch": 0.3060181566764865, + "flos": 19779691610160.0, + "grad_norm": 3.5947155526098036, + "learning_rate": 3.2528001603263425e-06, + "loss": 0.7871, + "num_input_tokens_seen": 54019990, + "step": 2545 + }, + { + "epoch": 0.3061383995671256, + "flos": 14514661258440.0, + "grad_norm": 2.41190832126412, + "learning_rate": 3.2521928555743514e-06, + "loss": 0.7957, + "num_input_tokens_seen": 54037055, + "step": 2546 + }, + { + "epoch": 0.3062586424577647, + "flos": 16189036378320.0, + "grad_norm": 2.465741606107749, + "learning_rate": 3.2515853608657e-06, + "loss": 0.6508, + "num_input_tokens_seen": 54054775, + "step": 2547 + }, + { + "epoch": 0.3063788853484038, + "flos": 15239553075720.0, + "grad_norm": 3.273371220102035, + "learning_rate": 3.250977676292545e-06, + "loss": 0.7125, + "num_input_tokens_seen": 54072735, + "step": 2548 + }, + { + "epoch": 0.30649912823904285, + "flos": 11808468754080.0, + "grad_norm": 2.292864353591317, + "learning_rate": 3.2503698019470712e-06, + "loss": 0.7798, + "num_input_tokens_seen": 54088225, + "step": 2549 + }, + { + "epoch": 0.30661937112968196, + "flos": 13590628072320.0, + "grad_norm": 3.051707343332677, + "learning_rate": 3.249761737921492e-06, + "loss": 0.7623, + "num_input_tokens_seen": 54104475, + "step": 2550 + }, + { + "epoch": 0.30673961402032107, + "flos": 23043923230800.0, + "grad_norm": 4.971653311994219, + "learning_rate": 3.249153484308051e-06, + "loss": 0.7295, + "num_input_tokens_seen": 54122810, + "step": 2551 + }, + { + "epoch": 0.3068598569109601, + "flos": 14784549163800.0, + "grad_norm": 2.522091060491934, + "learning_rate": 3.2485450411990194e-06, + "loss": 0.7558, + "num_input_tokens_seen": 54141885, + "step": 2552 + }, + { + "epoch": 0.30698009980159924, + "flos": 21719933820480.0, + "grad_norm": 2.648454514170321, + "learning_rate": 3.2479364086866983e-06, + "loss": 0.7989, + "num_input_tokens_seen": 54161860, + "step": 2553 + }, + { + "epoch": 0.30710034269223835, + "flos": 17458199720640.0, + "grad_norm": 2.3456957052273886, + "learning_rate": 3.247327586863416e-06, + "loss": 0.8025, + "num_input_tokens_seen": 54182460, + "step": 2554 + }, + { + "epoch": 0.3072205855828774, + "flos": 18969742954440.0, + "grad_norm": 3.3718077079273487, + "learning_rate": 3.2467185758215304e-06, + "loss": 0.7586, + "num_input_tokens_seen": 54201920, + "step": 2555 + }, + { + "epoch": 0.3073408284735165, + "flos": 16269534182520.0, + "grad_norm": 3.744364000004055, + "learning_rate": 3.246109375653428e-06, + "loss": 0.8305, + "num_input_tokens_seen": 54218405, + "step": 2556 + }, + { + "epoch": 0.30746107136415557, + "flos": 14244678373200.0, + "grad_norm": 2.5520541650052424, + "learning_rate": 3.2454999864515243e-06, + "loss": 0.7756, + "num_input_tokens_seen": 54237500, + "step": 2557 + }, + { + "epoch": 0.3075813142547947, + "flos": 15890817300120.0, + "grad_norm": 2.390409073396746, + "learning_rate": 3.244890408308263e-06, + "loss": 0.6766, + "num_input_tokens_seen": 54257925, + "step": 2558 + }, + { + "epoch": 0.3077015571454338, + "flos": 17647336542120.0, + "grad_norm": 2.941845540632024, + "learning_rate": 3.2442806413161165e-06, + "loss": 0.591, + "num_input_tokens_seen": 54277290, + "step": 2559 + }, + { + "epoch": 0.30782180003607285, + "flos": 13434413117880.0, + "grad_norm": 2.992722266793581, + "learning_rate": 3.243670685567586e-06, + "loss": 0.7615, + "num_input_tokens_seen": 54294410, + "step": 2560 + }, + { + "epoch": 0.30794204292671196, + "flos": 17484884575560.0, + "grad_norm": 2.6049573588038872, + "learning_rate": 3.2430605411552012e-06, + "loss": 0.7764, + "num_input_tokens_seen": 54314245, + "step": 2561 + }, + { + "epoch": 0.30806228581735107, + "flos": 50090999753160.0, + "grad_norm": 0.9015254484129663, + "learning_rate": 3.2424502081715205e-06, + "loss": 0.7109, + "num_input_tokens_seen": 54377080, + "step": 2562 + }, + { + "epoch": 0.3081825287079901, + "flos": 17458896239760.0, + "grad_norm": 3.1013298812228163, + "learning_rate": 3.241839686709132e-06, + "loss": 0.764, + "num_input_tokens_seen": 54397735, + "step": 2563 + }, + { + "epoch": 0.30830277159862923, + "flos": 11808690373800.0, + "grad_norm": 2.888206336092307, + "learning_rate": 3.2412289768606495e-06, + "loss": 0.7966, + "num_input_tokens_seen": 54414025, + "step": 2564 + }, + { + "epoch": 0.30842301448926834, + "flos": 21533488095600.0, + "grad_norm": 1.9429605506076677, + "learning_rate": 3.240618078718718e-06, + "loss": 0.8007, + "num_input_tokens_seen": 54435205, + "step": 2565 + }, + { + "epoch": 0.3085432573799074, + "flos": 16054124085600.0, + "grad_norm": 2.7887916754015643, + "learning_rate": 3.240006992376011e-06, + "loss": 0.7198, + "num_input_tokens_seen": 54454550, + "step": 2566 + }, + { + "epoch": 0.3086635002705465, + "flos": 16135191769080.0, + "grad_norm": 2.79581955756154, + "learning_rate": 3.2393957179252284e-06, + "loss": 0.7453, + "num_input_tokens_seen": 54470805, + "step": 2567 + }, + { + "epoch": 0.3087837431611856, + "flos": 23987327821080.0, + "grad_norm": 2.7194157804650314, + "learning_rate": 3.2387842554591016e-06, + "loss": 0.7939, + "num_input_tokens_seen": 54491340, + "step": 2568 + }, + { + "epoch": 0.3089039860518247, + "flos": 12837246782400.0, + "grad_norm": 6.328676552931197, + "learning_rate": 3.238172605070388e-06, + "loss": 0.8476, + "num_input_tokens_seen": 54506475, + "step": 2569 + }, + { + "epoch": 0.3090242289424638, + "flos": 10456971329880.0, + "grad_norm": 3.2146927530418767, + "learning_rate": 3.2375607668518745e-06, + "loss": 0.7584, + "num_input_tokens_seen": 54519230, + "step": 2570 + }, + { + "epoch": 0.30914447183310284, + "flos": 11704737030600.0, + "grad_norm": 2.5426540853945574, + "learning_rate": 3.236948740896377e-06, + "loss": 0.8837, + "num_input_tokens_seen": 54533750, + "step": 2571 + }, + { + "epoch": 0.30926471472374195, + "flos": 23665178304480.0, + "grad_norm": 2.012327113337138, + "learning_rate": 3.2363365272967384e-06, + "loss": 0.8245, + "num_input_tokens_seen": 54556040, + "step": 2572 + }, + { + "epoch": 0.30938495761438106, + "flos": 14888122587480.0, + "grad_norm": 2.485180418346112, + "learning_rate": 3.235724126145832e-06, + "loss": 0.803, + "num_input_tokens_seen": 54571795, + "step": 2573 + }, + { + "epoch": 0.3095052005050201, + "flos": 17620050147960.0, + "grad_norm": 2.15711205982899, + "learning_rate": 3.235111537536558e-06, + "loss": 0.7518, + "num_input_tokens_seen": 54592330, + "step": 2574 + }, + { + "epoch": 0.30962544339565923, + "flos": 17131491169800.0, + "grad_norm": 2.2523258691475574, + "learning_rate": 3.2344987615618456e-06, + "loss": 0.8219, + "num_input_tokens_seen": 54611885, + "step": 2575 + }, + { + "epoch": 0.30974568628629834, + "flos": 24827032347840.0, + "grad_norm": 1.9914761937677357, + "learning_rate": 3.2338857983146533e-06, + "loss": 0.7561, + "num_input_tokens_seen": 54633105, + "step": 2576 + }, + { + "epoch": 0.3098659291769374, + "flos": 14784359204040.0, + "grad_norm": 2.378470545363628, + "learning_rate": 3.233272647887966e-06, + "loss": 0.7469, + "num_input_tokens_seen": 54651715, + "step": 2577 + }, + { + "epoch": 0.3099861720675765, + "flos": 18127948870320.0, + "grad_norm": 2.0838078403511764, + "learning_rate": 3.2326593103747985e-06, + "loss": 0.8759, + "num_input_tokens_seen": 54670450, + "step": 2578 + }, + { + "epoch": 0.3101064149582156, + "flos": 8520433334880.0, + "grad_norm": 2.9016238894770434, + "learning_rate": 3.2320457858681936e-06, + "loss": 0.8315, + "num_input_tokens_seen": 54688560, + "step": 2579 + }, + { + "epoch": 0.31022665784885467, + "flos": 16858437268440.0, + "grad_norm": 2.7122502295470863, + "learning_rate": 3.2314320744612228e-06, + "loss": 0.8452, + "num_input_tokens_seen": 54703580, + "step": 2580 + }, + { + "epoch": 0.3103469007394938, + "flos": 11868202115880.0, + "grad_norm": 2.0027216036474735, + "learning_rate": 3.2308181762469854e-06, + "loss": 0.7538, + "num_input_tokens_seen": 54721245, + "step": 2581 + }, + { + "epoch": 0.3104671436301329, + "flos": 22395983302200.0, + "grad_norm": 2.761471317759095, + "learning_rate": 3.230204091318609e-06, + "loss": 0.768, + "num_input_tokens_seen": 54741505, + "step": 2582 + }, + { + "epoch": 0.31058738652077195, + "flos": 14649415251360.0, + "grad_norm": 3.0608350938195965, + "learning_rate": 3.2295898197692503e-06, + "loss": 0.8343, + "num_input_tokens_seen": 54760185, + "step": 2583 + }, + { + "epoch": 0.31070762941141106, + "flos": 20590305125040.0, + "grad_norm": 2.3800618288680275, + "learning_rate": 3.228975361692094e-06, + "loss": 0.7737, + "num_input_tokens_seen": 54780925, + "step": 2584 + }, + { + "epoch": 0.31082787230205017, + "flos": 15000750759720.0, + "grad_norm": 2.9623828269223798, + "learning_rate": 3.228360717180352e-06, + "loss": 0.784, + "num_input_tokens_seen": 54798170, + "step": 2585 + }, + { + "epoch": 0.3109481151926892, + "flos": 46032771605280.0, + "grad_norm": 0.8584558600096454, + "learning_rate": 3.227745886327266e-06, + "loss": 0.6356, + "num_input_tokens_seen": 54856585, + "step": 2586 + }, + { + "epoch": 0.31106835808332833, + "flos": 32933678547360.0, + "grad_norm": 0.8333912511407996, + "learning_rate": 3.227130869226105e-06, + "loss": 0.5809, + "num_input_tokens_seen": 54913465, + "step": 2587 + }, + { + "epoch": 0.3111886009739674, + "flos": 17132820888120.0, + "grad_norm": 7.247233948365663, + "learning_rate": 3.226515665970167e-06, + "loss": 0.8007, + "num_input_tokens_seen": 54930725, + "step": 2588 + }, + { + "epoch": 0.3113088438646065, + "flos": 12787169708400.0, + "grad_norm": 4.7879848543467745, + "learning_rate": 3.225900276652777e-06, + "loss": 0.8347, + "num_input_tokens_seen": 54947220, + "step": 2589 + }, + { + "epoch": 0.3114290867552456, + "flos": 20805683562000.0, + "grad_norm": 1.8511162715110783, + "learning_rate": 3.2252847013672906e-06, + "loss": 0.7414, + "num_input_tokens_seen": 54969600, + "step": 2590 + }, + { + "epoch": 0.31154932964588467, + "flos": 20075282911680.0, + "grad_norm": 2.7813148062367197, + "learning_rate": 3.224668940207089e-06, + "loss": 0.7427, + "num_input_tokens_seen": 54988305, + "step": 2591 + }, + { + "epoch": 0.3116695725365238, + "flos": 19455895775640.0, + "grad_norm": 2.135783491184399, + "learning_rate": 3.2240529932655828e-06, + "loss": 0.8474, + "num_input_tokens_seen": 55007290, + "step": 2592 + }, + { + "epoch": 0.3117898154271629, + "flos": 15485858802240.0, + "grad_norm": 4.451946972746254, + "learning_rate": 3.223436860636211e-06, + "loss": 0.8511, + "num_input_tokens_seen": 55022645, + "step": 2593 + }, + { + "epoch": 0.31191005831780194, + "flos": 19996684705080.0, + "grad_norm": 1.678894139490698, + "learning_rate": 3.2228205424124403e-06, + "loss": 0.725, + "num_input_tokens_seen": 55045520, + "step": 2594 + }, + { + "epoch": 0.31203030120844105, + "flos": 9406194120720.0, + "grad_norm": 3.128844223645895, + "learning_rate": 3.222204038687765e-06, + "loss": 0.7381, + "num_input_tokens_seen": 55058375, + "step": 2595 + }, + { + "epoch": 0.31215054409908016, + "flos": 20210670103800.0, + "grad_norm": 2.0125934365895706, + "learning_rate": 3.221587349555709e-06, + "loss": 0.8709, + "num_input_tokens_seen": 55078355, + "step": 2596 + }, + { + "epoch": 0.3122707869897192, + "flos": 15729188492520.0, + "grad_norm": 2.326881332532292, + "learning_rate": 3.2209704751098236e-06, + "loss": 0.6757, + "num_input_tokens_seen": 55097105, + "step": 2597 + }, + { + "epoch": 0.31239102988035833, + "flos": 11051446568760.0, + "grad_norm": 2.3561537017166945, + "learning_rate": 3.2203534154436875e-06, + "loss": 0.8004, + "num_input_tokens_seen": 55111180, + "step": 2598 + }, + { + "epoch": 0.31251127277099744, + "flos": 16134305290200.0, + "grad_norm": 3.0132272309497194, + "learning_rate": 3.2197361706509084e-06, + "loss": 0.7539, + "num_input_tokens_seen": 55131655, + "step": 2599 + }, + { + "epoch": 0.3126315156616365, + "flos": 11057588601000.0, + "grad_norm": 4.7451092487778315, + "learning_rate": 3.2191187408251228e-06, + "loss": 0.8111, + "num_input_tokens_seen": 55148535, + "step": 2600 + }, + { + "epoch": 0.3127517585522756, + "flos": 13241223821520.0, + "grad_norm": 6.688382669826623, + "learning_rate": 3.218501126059993e-06, + "loss": 0.755, + "num_input_tokens_seen": 55163650, + "step": 2601 + }, + { + "epoch": 0.31287200144291466, + "flos": 16027059311160.0, + "grad_norm": 2.308433772448542, + "learning_rate": 3.2178833264492116e-06, + "loss": 0.7946, + "num_input_tokens_seen": 55182075, + "step": 2602 + }, + { + "epoch": 0.31299224433355377, + "flos": 21938699873160.0, + "grad_norm": 2.4222242534212857, + "learning_rate": 3.217265342086498e-06, + "loss": 0.7449, + "num_input_tokens_seen": 55202285, + "step": 2603 + }, + { + "epoch": 0.3131124872241929, + "flos": 8439302331480.0, + "grad_norm": 2.5544637879578636, + "learning_rate": 3.216647173065599e-06, + "loss": 0.716, + "num_input_tokens_seen": 55217470, + "step": 2604 + }, + { + "epoch": 0.31323273011483194, + "flos": 36703524881880.0, + "grad_norm": 3.717342462953533, + "learning_rate": 3.216028819480292e-06, + "loss": 0.7191, + "num_input_tokens_seen": 55238530, + "step": 2605 + }, + { + "epoch": 0.31335297300547105, + "flos": 16539580387680.0, + "grad_norm": 3.2551220366165383, + "learning_rate": 3.2154102814243793e-06, + "loss": 0.7447, + "num_input_tokens_seen": 55257390, + "step": 2606 + }, + { + "epoch": 0.31347321589611016, + "flos": 25500960612240.0, + "grad_norm": 3.164777058106092, + "learning_rate": 3.2147915589916937e-06, + "loss": 0.6617, + "num_input_tokens_seen": 55278670, + "step": 2607 + }, + { + "epoch": 0.3135934587867492, + "flos": 14433783534720.0, + "grad_norm": 4.4081920916039365, + "learning_rate": 3.2141726522760938e-06, + "loss": 0.8125, + "num_input_tokens_seen": 55296450, + "step": 2608 + }, + { + "epoch": 0.3137137016773883, + "flos": 48528334666680.0, + "grad_norm": 0.781268348478381, + "learning_rate": 3.213553561371469e-06, + "loss": 0.5513, + "num_input_tokens_seen": 55359905, + "step": 2609 + }, + { + "epoch": 0.31383394456802743, + "flos": 11841232321320.0, + "grad_norm": 4.0121499547650865, + "learning_rate": 3.212934286371733e-06, + "loss": 0.9554, + "num_input_tokens_seen": 55376335, + "step": 2610 + }, + { + "epoch": 0.3139541874586665, + "flos": 28523635500360.0, + "grad_norm": 3.7825156371505044, + "learning_rate": 3.2123148273708304e-06, + "loss": 0.7991, + "num_input_tokens_seen": 55396245, + "step": 2611 + }, + { + "epoch": 0.3140744303493056, + "flos": 18349152739920.0, + "grad_norm": 2.4487285701481922, + "learning_rate": 3.211695184462733e-06, + "loss": 0.7596, + "num_input_tokens_seen": 55417140, + "step": 2612 + }, + { + "epoch": 0.3141946732399447, + "flos": 53474946885720.0, + "grad_norm": 0.8954170205985521, + "learning_rate": 3.2110753577414383e-06, + "loss": 0.6375, + "num_input_tokens_seen": 55478440, + "step": 2613 + }, + { + "epoch": 0.31431491613058377, + "flos": 14051172477240.0, + "grad_norm": 2.555570677237991, + "learning_rate": 3.2104553473009757e-06, + "loss": 0.7741, + "num_input_tokens_seen": 55496280, + "step": 2614 + }, + { + "epoch": 0.3144351590212229, + "flos": 26611407863280.0, + "grad_norm": 2.4782083619998216, + "learning_rate": 3.209835153235399e-06, + "loss": 0.6645, + "num_input_tokens_seen": 55517555, + "step": 2615 + }, + { + "epoch": 0.314555401911862, + "flos": 13542703875600.0, + "grad_norm": 2.877578242718823, + "learning_rate": 3.2092147756387916e-06, + "loss": 0.6698, + "num_input_tokens_seen": 55537600, + "step": 2616 + }, + { + "epoch": 0.31467564480250104, + "flos": 11922838224120.0, + "grad_norm": 3.247024347653989, + "learning_rate": 3.208594214605264e-06, + "loss": 0.8243, + "num_input_tokens_seen": 55555865, + "step": 2617 + }, + { + "epoch": 0.31479588769314015, + "flos": 15837289290480.0, + "grad_norm": 4.223312962058138, + "learning_rate": 3.2079734702289553e-06, + "loss": 0.7614, + "num_input_tokens_seen": 55574480, + "step": 2618 + }, + { + "epoch": 0.3149161305837792, + "flos": 48699836890440.0, + "grad_norm": 0.8159759701762236, + "learning_rate": 3.207352542604031e-06, + "loss": 0.6353, + "num_input_tokens_seen": 55636535, + "step": 2619 + }, + { + "epoch": 0.3150363734744183, + "flos": 21265721407560.0, + "grad_norm": 2.196940731188482, + "learning_rate": 3.2067314318246864e-06, + "loss": 0.7649, + "num_input_tokens_seen": 55656970, + "step": 2620 + }, + { + "epoch": 0.31515661636505743, + "flos": 20266097711040.0, + "grad_norm": 2.2760466714508643, + "learning_rate": 3.206110137985143e-06, + "loss": 0.7591, + "num_input_tokens_seen": 55676895, + "step": 2621 + }, + { + "epoch": 0.3152768592556965, + "flos": 18022602488880.0, + "grad_norm": 2.150861336921007, + "learning_rate": 3.2054886611796505e-06, + "loss": 0.9039, + "num_input_tokens_seen": 55695610, + "step": 2622 + }, + { + "epoch": 0.3153971021463356, + "flos": 50492471164080.0, + "grad_norm": 0.946079951468408, + "learning_rate": 3.204867001502487e-06, + "loss": 0.6842, + "num_input_tokens_seen": 55753985, + "step": 2623 + }, + { + "epoch": 0.3155173450369747, + "flos": 18754269537600.0, + "grad_norm": 2.8315926047689706, + "learning_rate": 3.2042451590479567e-06, + "loss": 0.7839, + "num_input_tokens_seen": 55774220, + "step": 2624 + }, + { + "epoch": 0.31563758792761376, + "flos": 17802696677640.0, + "grad_norm": 1.982821960667134, + "learning_rate": 3.203623133910394e-06, + "loss": 0.8418, + "num_input_tokens_seen": 55792245, + "step": 2625 + }, + { + "epoch": 0.31575783081825287, + "flos": 23423938171560.0, + "grad_norm": 3.9444147580481825, + "learning_rate": 3.203000926184158e-06, + "loss": 0.7533, + "num_input_tokens_seen": 55810890, + "step": 2626 + }, + { + "epoch": 0.315878073708892, + "flos": 22615350894120.0, + "grad_norm": 2.513714935953526, + "learning_rate": 3.202378535963639e-06, + "loss": 0.7584, + "num_input_tokens_seen": 55831525, + "step": 2627 + }, + { + "epoch": 0.31599831659953104, + "flos": 16242849327600.0, + "grad_norm": 2.5642077640202574, + "learning_rate": 3.2017559633432516e-06, + "loss": 0.8252, + "num_input_tokens_seen": 55850875, + "step": 2628 + }, + { + "epoch": 0.31611855949017015, + "flos": 18754047917880.0, + "grad_norm": 2.420261362322401, + "learning_rate": 3.2011332084174398e-06, + "loss": 0.6436, + "num_input_tokens_seen": 55871465, + "step": 2629 + }, + { + "epoch": 0.31623880238080926, + "flos": 15054057149640.0, + "grad_norm": 2.003354359081118, + "learning_rate": 3.2005102712806756e-06, + "loss": 0.8748, + "num_input_tokens_seen": 55890015, + "step": 2630 + }, + { + "epoch": 0.3163590452714483, + "flos": 9274574463840.0, + "grad_norm": 5.309649072750179, + "learning_rate": 3.1998871520274575e-06, + "loss": 0.7054, + "num_input_tokens_seen": 55905070, + "step": 2631 + }, + { + "epoch": 0.3164792881620874, + "flos": 16865529099480.0, + "grad_norm": 5.463095046396495, + "learning_rate": 3.199263850752312e-06, + "loss": 0.8216, + "num_input_tokens_seen": 55925625, + "step": 2632 + }, + { + "epoch": 0.31659953105272653, + "flos": 13353630374040.0, + "grad_norm": 3.4620070344092064, + "learning_rate": 3.198640367549795e-06, + "loss": 0.833, + "num_input_tokens_seen": 55944240, + "step": 2633 + }, + { + "epoch": 0.3167197739433656, + "flos": 18835495520880.0, + "grad_norm": 2.606646819059449, + "learning_rate": 3.198016702514487e-06, + "loss": 0.8446, + "num_input_tokens_seen": 55964240, + "step": 2634 + }, + { + "epoch": 0.3168400168340047, + "flos": 17238198929520.0, + "grad_norm": 2.0789290858766694, + "learning_rate": 3.1973928557409972e-06, + "loss": 0.8325, + "num_input_tokens_seen": 55982000, + "step": 2635 + }, + { + "epoch": 0.31696025972464376, + "flos": 20806253441280.0, + "grad_norm": 2.7991685041410412, + "learning_rate": 3.1967688273239636e-06, + "loss": 0.6894, + "num_input_tokens_seen": 56001525, + "step": 2636 + }, + { + "epoch": 0.31708050261528287, + "flos": 11949618058920.0, + "grad_norm": 2.7688686509036557, + "learning_rate": 3.1961446173580503e-06, + "loss": 0.8071, + "num_input_tokens_seen": 56018185, + "step": 2637 + }, + { + "epoch": 0.317200745505922, + "flos": 19211869566240.0, + "grad_norm": 2.174108660231061, + "learning_rate": 3.1955202259379502e-06, + "loss": 0.7537, + "num_input_tokens_seen": 56039635, + "step": 2638 + }, + { + "epoch": 0.31732098839656103, + "flos": 23013882420120.0, + "grad_norm": 2.209855316915114, + "learning_rate": 3.194895653158381e-06, + "loss": 0.8049, + "num_input_tokens_seen": 56058295, + "step": 2639 + }, + { + "epoch": 0.31744123128720014, + "flos": 43475738375640.0, + "grad_norm": 0.7803833923401847, + "learning_rate": 3.194270899114093e-06, + "loss": 0.5874, + "num_input_tokens_seen": 56123810, + "step": 2640 + }, + { + "epoch": 0.31756147417783925, + "flos": 12702904368960.0, + "grad_norm": 2.842036475079726, + "learning_rate": 3.193645963899858e-06, + "loss": 0.8111, + "num_input_tokens_seen": 56141145, + "step": 2641 + }, + { + "epoch": 0.3176817170684783, + "flos": 18670985656920.0, + "grad_norm": 2.2612759645910336, + "learning_rate": 3.193020847610479e-06, + "loss": 0.8159, + "num_input_tokens_seen": 56161185, + "step": 2642 + }, + { + "epoch": 0.3178019599591174, + "flos": 18294484971720.0, + "grad_norm": 3.370390959115152, + "learning_rate": 3.192395550340787e-06, + "loss": 0.6944, + "num_input_tokens_seen": 56178855, + "step": 2643 + }, + { + "epoch": 0.31792220284975653, + "flos": 9006459516240.0, + "grad_norm": 3.044453915892979, + "learning_rate": 3.191770072185638e-06, + "loss": 0.7543, + "num_input_tokens_seen": 56195570, + "step": 2644 + }, + { + "epoch": 0.3180424457403956, + "flos": 11273505257280.0, + "grad_norm": 3.957875168368587, + "learning_rate": 3.191144413239916e-06, + "loss": 0.7077, + "num_input_tokens_seen": 56211860, + "step": 2645 + }, + { + "epoch": 0.3181626886310347, + "flos": 19184076612720.0, + "grad_norm": 2.801180035705978, + "learning_rate": 3.190518573598534e-06, + "loss": 0.8613, + "num_input_tokens_seen": 56228185, + "step": 2646 + }, + { + "epoch": 0.3182829315216738, + "flos": 18672252055320.0, + "grad_norm": 1.8709933387397184, + "learning_rate": 3.1898925533564308e-06, + "loss": 0.7702, + "num_input_tokens_seen": 56249375, + "step": 2647 + }, + { + "epoch": 0.31840317441231286, + "flos": 13218496461600.0, + "grad_norm": 3.197526309432296, + "learning_rate": 3.1892663526085733e-06, + "loss": 0.6293, + "num_input_tokens_seen": 56267470, + "step": 2648 + }, + { + "epoch": 0.31852341730295197, + "flos": 47731615382880.0, + "grad_norm": 0.7606634802006281, + "learning_rate": 3.188639971449956e-06, + "loss": 0.5937, + "num_input_tokens_seen": 56333240, + "step": 2649 + }, + { + "epoch": 0.318643660193591, + "flos": 15109009857480.0, + "grad_norm": 3.4205977612384437, + "learning_rate": 3.1880134099756e-06, + "loss": 0.7135, + "num_input_tokens_seen": 56352595, + "step": 2650 + }, + { + "epoch": 0.31876390308423014, + "flos": 19753038415200.0, + "grad_norm": 2.0911578173707692, + "learning_rate": 3.1873866682805535e-06, + "loss": 0.6922, + "num_input_tokens_seen": 56373010, + "step": 2651 + }, + { + "epoch": 0.31888414597486925, + "flos": 13165538331240.0, + "grad_norm": 2.3684975804128725, + "learning_rate": 3.186759746459894e-06, + "loss": 0.8645, + "num_input_tokens_seen": 56391840, + "step": 2652 + }, + { + "epoch": 0.3190043888655083, + "flos": 18424458310680.0, + "grad_norm": 2.8537241646077174, + "learning_rate": 3.1861326446087246e-06, + "loss": 0.7874, + "num_input_tokens_seen": 56410300, + "step": 2653 + }, + { + "epoch": 0.3191246317561474, + "flos": 16134463590000.0, + "grad_norm": 3.4991815154949233, + "learning_rate": 3.1855053628221763e-06, + "loss": 0.6981, + "num_input_tokens_seen": 56429275, + "step": 2654 + }, + { + "epoch": 0.3192448746467865, + "flos": 10841323685160.0, + "grad_norm": 5.106551670609004, + "learning_rate": 3.184877901195407e-06, + "loss": 0.868, + "num_input_tokens_seen": 56445690, + "step": 2655 + }, + { + "epoch": 0.3193651175374256, + "flos": 49578193737000.0, + "grad_norm": 0.8402851098286303, + "learning_rate": 3.184250259823602e-06, + "loss": 0.6616, + "num_input_tokens_seen": 56507940, + "step": 2656 + }, + { + "epoch": 0.3194853604280647, + "flos": 8865531831120.0, + "grad_norm": 2.8382834407462987, + "learning_rate": 3.183622438801974e-06, + "loss": 0.7889, + "num_input_tokens_seen": 56522950, + "step": 2657 + }, + { + "epoch": 0.3196056033187038, + "flos": 10869749837880.0, + "grad_norm": 3.2353985774674077, + "learning_rate": 3.1829944382257637e-06, + "loss": 0.7471, + "num_input_tokens_seen": 56540800, + "step": 2658 + }, + { + "epoch": 0.31972584620934286, + "flos": 17403690252240.0, + "grad_norm": 2.5923808517224467, + "learning_rate": 3.1823662581902373e-06, + "loss": 0.7998, + "num_input_tokens_seen": 56558205, + "step": 2659 + }, + { + "epoch": 0.31984608909998197, + "flos": 15540716530200.0, + "grad_norm": 2.4182419630264347, + "learning_rate": 3.1817378987906896e-06, + "loss": 0.7198, + "num_input_tokens_seen": 56577430, + "step": 2660 + }, + { + "epoch": 0.3199663319906211, + "flos": 13353598714080.0, + "grad_norm": 2.9122416762368033, + "learning_rate": 3.181109360122442e-06, + "loss": 0.7922, + "num_input_tokens_seen": 56594740, + "step": 2661 + }, + { + "epoch": 0.32008657488126013, + "flos": 13677236248800.0, + "grad_norm": 2.861265797636018, + "learning_rate": 3.1804806422808445e-06, + "loss": 0.7679, + "num_input_tokens_seen": 56611595, + "step": 2662 + }, + { + "epoch": 0.32020681777189924, + "flos": 15055165248240.0, + "grad_norm": 3.466124219514145, + "learning_rate": 3.1798517453612714e-06, + "loss": 0.7087, + "num_input_tokens_seen": 56631120, + "step": 2663 + }, + { + "epoch": 0.32032706066253835, + "flos": 25909116766080.0, + "grad_norm": 2.4845495582531196, + "learning_rate": 3.1792226694591265e-06, + "loss": 0.7422, + "num_input_tokens_seen": 56652225, + "step": 2664 + }, + { + "epoch": 0.3204473035531774, + "flos": 11139289483680.0, + "grad_norm": 2.4189320064543702, + "learning_rate": 3.178593414669841e-06, + "loss": 0.7898, + "num_input_tokens_seen": 56670530, + "step": 2665 + }, + { + "epoch": 0.3205675464438165, + "flos": 17917129467600.0, + "grad_norm": 5.88288966479106, + "learning_rate": 3.1779639810888707e-06, + "loss": 0.6893, + "num_input_tokens_seen": 56689845, + "step": 2666 + }, + { + "epoch": 0.3206877893344556, + "flos": 16432366068600.0, + "grad_norm": 2.7276186933137305, + "learning_rate": 3.1773343688117013e-06, + "loss": 0.7463, + "num_input_tokens_seen": 56710475, + "step": 2667 + }, + { + "epoch": 0.3208080322250947, + "flos": 14919556436400.0, + "grad_norm": 3.0875284850007323, + "learning_rate": 3.1767045779338445e-06, + "loss": 0.8217, + "num_input_tokens_seen": 56727855, + "step": 2668 + }, + { + "epoch": 0.3209282751157338, + "flos": 15918546933720.0, + "grad_norm": 5.609140205250987, + "learning_rate": 3.176074608550839e-06, + "loss": 0.8865, + "num_input_tokens_seen": 56743395, + "step": 2669 + }, + { + "epoch": 0.32104851800637285, + "flos": 16135476708720.0, + "grad_norm": 2.3043308822675455, + "learning_rate": 3.17544446075825e-06, + "loss": 0.8118, + "num_input_tokens_seen": 56762280, + "step": 2670 + }, + { + "epoch": 0.32116876089701196, + "flos": 27204585043800.0, + "grad_norm": 1.7264293084187268, + "learning_rate": 3.174814134651671e-06, + "loss": 0.6994, + "num_input_tokens_seen": 56784550, + "step": 2671 + }, + { + "epoch": 0.3212890037876511, + "flos": 16078972662840.0, + "grad_norm": 2.573976542955056, + "learning_rate": 3.1741836303267215e-06, + "loss": 0.7967, + "num_input_tokens_seen": 56803805, + "step": 2672 + }, + { + "epoch": 0.32140924667829013, + "flos": 7467186648840.0, + "grad_norm": 4.607024029691646, + "learning_rate": 3.1735529478790496e-06, + "loss": 0.7292, + "num_input_tokens_seen": 56821515, + "step": 2673 + }, + { + "epoch": 0.32152948956892924, + "flos": 37406892417720.0, + "grad_norm": 2.579157304721125, + "learning_rate": 3.172922087404328e-06, + "loss": 0.7817, + "num_input_tokens_seen": 56843495, + "step": 2674 + }, + { + "epoch": 0.32164973245956835, + "flos": 47080889377800.0, + "grad_norm": 0.9253902655907155, + "learning_rate": 3.1722910489982586e-06, + "loss": 0.5774, + "num_input_tokens_seen": 56903575, + "step": 2675 + }, + { + "epoch": 0.3217699753502074, + "flos": 17215091650080.0, + "grad_norm": 2.0587602870107435, + "learning_rate": 3.1716598327565694e-06, + "loss": 0.7948, + "num_input_tokens_seen": 56922935, + "step": 2676 + }, + { + "epoch": 0.3218902182408465, + "flos": 13920724238880.0, + "grad_norm": 2.0122638445366197, + "learning_rate": 3.171028438775015e-06, + "loss": 0.8266, + "num_input_tokens_seen": 56941850, + "step": 2677 + }, + { + "epoch": 0.3220104611314856, + "flos": 14892396682080.0, + "grad_norm": 2.782474719649888, + "learning_rate": 3.170396867149377e-06, + "loss": 0.8265, + "num_input_tokens_seen": 56959575, + "step": 2678 + }, + { + "epoch": 0.3221307040221247, + "flos": 14700600423960.0, + "grad_norm": 2.191612954684209, + "learning_rate": 3.1697651179754653e-06, + "loss": 0.8362, + "num_input_tokens_seen": 56977955, + "step": 2679 + }, + { + "epoch": 0.3222509469127638, + "flos": 17565698979360.0, + "grad_norm": 3.5867619782113067, + "learning_rate": 3.1691331913491153e-06, + "loss": 0.7292, + "num_input_tokens_seen": 57000245, + "step": 2680 + }, + { + "epoch": 0.32237118980340285, + "flos": 12894352367520.0, + "grad_norm": 2.9367467036113064, + "learning_rate": 3.1685010873661898e-06, + "loss": 0.8278, + "num_input_tokens_seen": 57019140, + "step": 2681 + }, + { + "epoch": 0.32249143269404196, + "flos": 16944095646120.0, + "grad_norm": 2.2275519063362728, + "learning_rate": 3.167868806122578e-06, + "loss": 0.777, + "num_input_tokens_seen": 57037910, + "step": 2682 + }, + { + "epoch": 0.32261167558468107, + "flos": 17886803717280.0, + "grad_norm": 2.6217515918664653, + "learning_rate": 3.1672363477141968e-06, + "loss": 0.656, + "num_input_tokens_seen": 57056925, + "step": 2683 + }, + { + "epoch": 0.3227319184753201, + "flos": 22286077886520.0, + "grad_norm": 2.8632491085231297, + "learning_rate": 3.1666037122369903e-06, + "loss": 0.8254, + "num_input_tokens_seen": 57077305, + "step": 2684 + }, + { + "epoch": 0.32285216136595923, + "flos": 12354449916960.0, + "grad_norm": 3.6319988693201934, + "learning_rate": 3.165970899786928e-06, + "loss": 0.8396, + "num_input_tokens_seen": 57094940, + "step": 2685 + }, + { + "epoch": 0.32297240425659834, + "flos": 16080998900280.0, + "grad_norm": 2.7490532719295815, + "learning_rate": 3.1653379104600067e-06, + "loss": 0.7184, + "num_input_tokens_seen": 57114805, + "step": 2686 + }, + { + "epoch": 0.3230926471472374, + "flos": 16648314384840.0, + "grad_norm": 2.240629392483927, + "learning_rate": 3.164704744352251e-06, + "loss": 0.6796, + "num_input_tokens_seen": 57135330, + "step": 2687 + }, + { + "epoch": 0.3232128900378765, + "flos": 12351632180520.0, + "grad_norm": 2.554272445886249, + "learning_rate": 3.164071401559713e-06, + "loss": 0.794, + "num_input_tokens_seen": 57152450, + "step": 2688 + }, + { + "epoch": 0.3233331329285156, + "flos": 17592225534480.0, + "grad_norm": 4.156171296027023, + "learning_rate": 3.1634378821784674e-06, + "loss": 0.6953, + "num_input_tokens_seen": 57172385, + "step": 2689 + }, + { + "epoch": 0.3234533758191547, + "flos": 13218623101440.0, + "grad_norm": 2.816570219010082, + "learning_rate": 3.1628041863046208e-06, + "loss": 0.717, + "num_input_tokens_seen": 57189520, + "step": 2690 + }, + { + "epoch": 0.3235736187097938, + "flos": 12354418257000.0, + "grad_norm": 3.9053307023730346, + "learning_rate": 3.162170314034304e-06, + "loss": 0.8999, + "num_input_tokens_seen": 57206655, + "step": 2691 + }, + { + "epoch": 0.3236938616004329, + "flos": 16188719778720.0, + "grad_norm": 2.0505942995144535, + "learning_rate": 3.1615362654636738e-06, + "loss": 0.7896, + "num_input_tokens_seen": 57227115, + "step": 2692 + }, + { + "epoch": 0.32381410449107195, + "flos": 12516047064600.0, + "grad_norm": 2.231374259501982, + "learning_rate": 3.1609020406889163e-06, + "loss": 0.864, + "num_input_tokens_seen": 57244270, + "step": 2693 + }, + { + "epoch": 0.32393434738171106, + "flos": 12081617635320.0, + "grad_norm": 3.005341225544615, + "learning_rate": 3.1602676398062416e-06, + "loss": 0.8352, + "num_input_tokens_seen": 57262900, + "step": 2694 + }, + { + "epoch": 0.3240545902723502, + "flos": 18672695294760.0, + "grad_norm": 3.4750029419811064, + "learning_rate": 3.1596330629118886e-06, + "loss": 0.5985, + "num_input_tokens_seen": 57282590, + "step": 2695 + }, + { + "epoch": 0.32417483316298923, + "flos": 26341488297960.0, + "grad_norm": 2.3905394295730122, + "learning_rate": 3.1589983101021223e-06, + "loss": 0.7188, + "num_input_tokens_seen": 57300940, + "step": 2696 + }, + { + "epoch": 0.32429507605362834, + "flos": 22074656944560.0, + "grad_norm": 10.440659461696717, + "learning_rate": 3.1583633814732337e-06, + "loss": 0.8155, + "num_input_tokens_seen": 57320180, + "step": 2697 + }, + { + "epoch": 0.3244153189442674, + "flos": 13299595805040.0, + "grad_norm": 3.456216248387134, + "learning_rate": 3.157728277121541e-06, + "loss": 0.6913, + "num_input_tokens_seen": 57338075, + "step": 2698 + }, + { + "epoch": 0.3245355618349065, + "flos": 12919675844160.0, + "grad_norm": 3.4931637894883876, + "learning_rate": 3.1570929971433897e-06, + "loss": 0.7648, + "num_input_tokens_seen": 57353580, + "step": 2699 + }, + { + "epoch": 0.3246558047255456, + "flos": 17160930441240.0, + "grad_norm": 2.7930328575192904, + "learning_rate": 3.1564575416351504e-06, + "loss": 0.8104, + "num_input_tokens_seen": 57372000, + "step": 2700 + }, + { + "epoch": 0.32477604761618467, + "flos": 15917438835120.0, + "grad_norm": 2.610406089410159, + "learning_rate": 3.155821910693221e-06, + "loss": 0.7215, + "num_input_tokens_seen": 57391135, + "step": 2701 + }, + { + "epoch": 0.3248962905068238, + "flos": 14487406524240.0, + "grad_norm": 3.348076846253743, + "learning_rate": 3.1551861044140275e-06, + "loss": 0.836, + "num_input_tokens_seen": 57410490, + "step": 2702 + }, + { + "epoch": 0.3250165333974629, + "flos": 17535974768280.0, + "grad_norm": 2.3780602264605974, + "learning_rate": 3.15455012289402e-06, + "loss": 0.7507, + "num_input_tokens_seen": 57429855, + "step": 2703 + }, + { + "epoch": 0.32513677628810195, + "flos": 17567060357640.0, + "grad_norm": 2.0993048016735423, + "learning_rate": 3.153913966229677e-06, + "loss": 0.8233, + "num_input_tokens_seen": 57448695, + "step": 2704 + }, + { + "epoch": 0.32525701917874106, + "flos": 52108700411520.0, + "grad_norm": 0.6430979426246516, + "learning_rate": 3.1532776345175027e-06, + "loss": 0.5082, + "num_input_tokens_seen": 57513560, + "step": 2705 + }, + { + "epoch": 0.32537726206938017, + "flos": 14379622325880.0, + "grad_norm": 2.4993024806657678, + "learning_rate": 3.1526411278540285e-06, + "loss": 0.7624, + "num_input_tokens_seen": 57531710, + "step": 2706 + }, + { + "epoch": 0.3254975049600192, + "flos": 21097285708560.0, + "grad_norm": 2.4229348109564426, + "learning_rate": 3.1520044463358116e-06, + "loss": 0.7943, + "num_input_tokens_seen": 57548160, + "step": 2707 + }, + { + "epoch": 0.32561774785065833, + "flos": 13783310809320.0, + "grad_norm": 1.9745605945693276, + "learning_rate": 3.151367590059436e-06, + "loss": 0.783, + "num_input_tokens_seen": 57566305, + "step": 2708 + }, + { + "epoch": 0.32573799074129745, + "flos": 16917632410920.0, + "grad_norm": 2.1477488020808204, + "learning_rate": 3.1507305591215117e-06, + "loss": 0.8481, + "num_input_tokens_seen": 57583935, + "step": 2709 + }, + { + "epoch": 0.3258582336319365, + "flos": 52539045705960.0, + "grad_norm": 0.7482013194355245, + "learning_rate": 3.150093353618677e-06, + "loss": 0.5762, + "num_input_tokens_seen": 57648385, + "step": 2710 + }, + { + "epoch": 0.3259784765225756, + "flos": 16432176108840.0, + "grad_norm": 2.9158221702510994, + "learning_rate": 3.149455973647596e-06, + "loss": 0.8566, + "num_input_tokens_seen": 57666165, + "step": 2711 + }, + { + "epoch": 0.32609871941321467, + "flos": 14973211085880.0, + "grad_norm": 2.5733493847749913, + "learning_rate": 3.1488184193049563e-06, + "loss": 0.7438, + "num_input_tokens_seen": 57685420, + "step": 2712 + }, + { + "epoch": 0.3262189623038538, + "flos": 16402768497360.0, + "grad_norm": 2.0099172284748006, + "learning_rate": 3.1481806906874767e-06, + "loss": 0.7104, + "num_input_tokens_seen": 57706450, + "step": 2713 + }, + { + "epoch": 0.3263392051944929, + "flos": 15297988379160.0, + "grad_norm": 2.0563458034842, + "learning_rate": 3.147542787891899e-06, + "loss": 0.8641, + "num_input_tokens_seen": 57725515, + "step": 2714 + }, + { + "epoch": 0.32645944808513194, + "flos": 17593491932880.0, + "grad_norm": 2.7183920964528596, + "learning_rate": 3.1469047110149926e-06, + "loss": 0.7252, + "num_input_tokens_seen": 57743975, + "step": 2715 + }, + { + "epoch": 0.32657969097577105, + "flos": 15378517843320.0, + "grad_norm": 3.2310533651881777, + "learning_rate": 3.146266460153554e-06, + "loss": 0.8354, + "num_input_tokens_seen": 57763405, + "step": 2716 + }, + { + "epoch": 0.32669993386641016, + "flos": 16620173171760.0, + "grad_norm": 2.72357416209485, + "learning_rate": 3.145628035404404e-06, + "loss": 0.786, + "num_input_tokens_seen": 57782325, + "step": 2717 + }, + { + "epoch": 0.3268201767570492, + "flos": 53180400362880.0, + "grad_norm": 0.8584693992195426, + "learning_rate": 3.1449894368643922e-06, + "loss": 0.5935, + "num_input_tokens_seen": 57844360, + "step": 2718 + }, + { + "epoch": 0.32694041964768833, + "flos": 17970815777040.0, + "grad_norm": 2.02791756390542, + "learning_rate": 3.1443506646303934e-06, + "loss": 0.7076, + "num_input_tokens_seen": 57865380, + "step": 2719 + }, + { + "epoch": 0.32706066253832744, + "flos": 24370888677360.0, + "grad_norm": 2.594753025481352, + "learning_rate": 3.1437117187993086e-06, + "loss": 0.6563, + "num_input_tokens_seen": 57887420, + "step": 2720 + }, + { + "epoch": 0.3271809054289665, + "flos": 17620461727440.0, + "grad_norm": 1.9789795101082333, + "learning_rate": 3.143072599468065e-06, + "loss": 0.7866, + "num_input_tokens_seen": 57906965, + "step": 2721 + }, + { + "epoch": 0.3273011483196056, + "flos": 28123489316400.0, + "grad_norm": 1.6349369082395657, + "learning_rate": 3.1424333067336174e-06, + "loss": 0.7436, + "num_input_tokens_seen": 57929450, + "step": 2722 + }, + { + "epoch": 0.3274213912102447, + "flos": 21314627063040.0, + "grad_norm": 2.0943443251924503, + "learning_rate": 3.141793840692945e-06, + "loss": 0.7571, + "num_input_tokens_seen": 57949920, + "step": 2723 + }, + { + "epoch": 0.32754163410088377, + "flos": 21373822205520.0, + "grad_norm": 3.477128229295179, + "learning_rate": 3.1411542014430553e-06, + "loss": 0.5903, + "num_input_tokens_seen": 57970720, + "step": 2724 + }, + { + "epoch": 0.3276618769915229, + "flos": 15081501843600.0, + "grad_norm": 1.9958887760157342, + "learning_rate": 3.1405143890809804e-06, + "loss": 0.8053, + "num_input_tokens_seen": 57989735, + "step": 2725 + }, + { + "epoch": 0.327782119882162, + "flos": 13620383943360.0, + "grad_norm": 2.1691304161783673, + "learning_rate": 3.1398744037037796e-06, + "loss": 0.6909, + "num_input_tokens_seen": 58008790, + "step": 2726 + }, + { + "epoch": 0.32790236277280105, + "flos": 15943078911360.0, + "grad_norm": 2.277115196942769, + "learning_rate": 3.139234245408538e-06, + "loss": 0.8227, + "num_input_tokens_seen": 58027390, + "step": 2727 + }, + { + "epoch": 0.32802260566344016, + "flos": 17079989397600.0, + "grad_norm": 1.8133591555879218, + "learning_rate": 3.1385939142923666e-06, + "loss": 0.7492, + "num_input_tokens_seen": 58049500, + "step": 2728 + }, + { + "epoch": 0.3281428485540792, + "flos": 17727960986160.0, + "grad_norm": 2.810994911137673, + "learning_rate": 3.137953410452405e-06, + "loss": 0.76, + "num_input_tokens_seen": 58069490, + "step": 2729 + }, + { + "epoch": 0.3282630914447183, + "flos": 25070583657840.0, + "grad_norm": 2.2409963600643206, + "learning_rate": 3.1373127339858146e-06, + "loss": 0.7312, + "num_input_tokens_seen": 58091810, + "step": 2730 + }, + { + "epoch": 0.32838333433535744, + "flos": 20239634475840.0, + "grad_norm": 2.7116445009732013, + "learning_rate": 3.136671884989787e-06, + "loss": 0.7183, + "num_input_tokens_seen": 58111440, + "step": 2731 + }, + { + "epoch": 0.3285035772259965, + "flos": 8952678226920.0, + "grad_norm": 3.991552134379217, + "learning_rate": 3.1360308635615383e-06, + "loss": 0.8441, + "num_input_tokens_seen": 58129700, + "step": 2732 + }, + { + "epoch": 0.3286238201166356, + "flos": 17808648750120.0, + "grad_norm": 2.2659267882247933, + "learning_rate": 3.135389669798311e-06, + "loss": 0.773, + "num_input_tokens_seen": 58147480, + "step": 2733 + }, + { + "epoch": 0.3287440630072747, + "flos": 15644606553480.0, + "grad_norm": 3.1146272552012984, + "learning_rate": 3.134748303797373e-06, + "loss": 0.7891, + "num_input_tokens_seen": 58164570, + "step": 2734 + }, + { + "epoch": 0.32886430589791377, + "flos": 17376815437560.0, + "grad_norm": 3.294425162619362, + "learning_rate": 3.1341067656560203e-06, + "loss": 0.7939, + "num_input_tokens_seen": 58182135, + "step": 2735 + }, + { + "epoch": 0.3289845487885529, + "flos": 16404446475240.0, + "grad_norm": 3.1127402339283012, + "learning_rate": 3.133465055471572e-06, + "loss": 0.833, + "num_input_tokens_seen": 58201640, + "step": 2736 + }, + { + "epoch": 0.329104791679192, + "flos": 14379590665920.0, + "grad_norm": 2.816291796174745, + "learning_rate": 3.1328231733413767e-06, + "loss": 0.6405, + "num_input_tokens_seen": 58218000, + "step": 2737 + }, + { + "epoch": 0.32922503456983104, + "flos": 11652127159800.0, + "grad_norm": 3.0969215859182118, + "learning_rate": 3.1321811193628067e-06, + "loss": 0.8839, + "num_input_tokens_seen": 58235865, + "step": 2738 + }, + { + "epoch": 0.32934527746047015, + "flos": 19672192351440.0, + "grad_norm": 3.2214671397566264, + "learning_rate": 3.131538893633261e-06, + "loss": 0.6841, + "num_input_tokens_seen": 58255145, + "step": 2739 + }, + { + "epoch": 0.32946552035110926, + "flos": 17133454087320.0, + "grad_norm": 4.7980347265884635, + "learning_rate": 3.130896496250165e-06, + "loss": 0.758, + "num_input_tokens_seen": 58274690, + "step": 2740 + }, + { + "epoch": 0.3295857632417483, + "flos": 10301611194360.0, + "grad_norm": 3.2332146816178478, + "learning_rate": 3.1302539273109693e-06, + "loss": 0.8422, + "num_input_tokens_seen": 58291235, + "step": 2741 + }, + { + "epoch": 0.32970600613238743, + "flos": 16239714991560.0, + "grad_norm": 4.460638620816259, + "learning_rate": 3.1296111869131513e-06, + "loss": 0.7957, + "num_input_tokens_seen": 58308380, + "step": 2742 + }, + { + "epoch": 0.32982624902302654, + "flos": 16134368610120.0, + "grad_norm": 2.4769820089660923, + "learning_rate": 3.1289682751542153e-06, + "loss": 0.8351, + "num_input_tokens_seen": 58327660, + "step": 2743 + }, + { + "epoch": 0.3299464919136656, + "flos": 13624088158680.0, + "grad_norm": 2.7869458969279552, + "learning_rate": 3.1283251921316883e-06, + "loss": 0.7014, + "num_input_tokens_seen": 58345125, + "step": 2744 + }, + { + "epoch": 0.3300667348043047, + "flos": 9735530448240.0, + "grad_norm": 2.8607812977373657, + "learning_rate": 3.1276819379431277e-06, + "loss": 0.8051, + "num_input_tokens_seen": 58362935, + "step": 2745 + }, + { + "epoch": 0.33018697769494376, + "flos": 11463750177360.0, + "grad_norm": 2.9903523766315203, + "learning_rate": 3.1270385126861134e-06, + "loss": 0.7328, + "num_input_tokens_seen": 58380640, + "step": 2746 + }, + { + "epoch": 0.3303072205855829, + "flos": 13325584140840.0, + "grad_norm": 3.15493278589004, + "learning_rate": 3.1263949164582533e-06, + "loss": 0.8039, + "num_input_tokens_seen": 58400010, + "step": 2747 + }, + { + "epoch": 0.330427463476222, + "flos": 12948988475760.0, + "grad_norm": 2.932908710899425, + "learning_rate": 3.1257511493571797e-06, + "loss": 0.7519, + "num_input_tokens_seen": 58418235, + "step": 2748 + }, + { + "epoch": 0.33054770636686104, + "flos": 19915332081960.0, + "grad_norm": 4.58167511969867, + "learning_rate": 3.125107211480552e-06, + "loss": 0.7721, + "num_input_tokens_seen": 58437890, + "step": 2749 + }, + { + "epoch": 0.33066794925750015, + "flos": 14700758723760.0, + "grad_norm": 2.051981256983116, + "learning_rate": 3.124463102926054e-06, + "loss": 0.775, + "num_input_tokens_seen": 58456945, + "step": 2750 + }, + { + "epoch": 0.33078819214813926, + "flos": 52097081206200.0, + "grad_norm": 0.7583136720169608, + "learning_rate": 3.1238188237913984e-06, + "loss": 0.62, + "num_input_tokens_seen": 58521205, + "step": 2751 + }, + { + "epoch": 0.3309084350387783, + "flos": 15459585526800.0, + "grad_norm": 4.302635742386873, + "learning_rate": 3.1231743741743202e-06, + "loss": 0.7406, + "num_input_tokens_seen": 58540430, + "step": 2752 + }, + { + "epoch": 0.3310286779294174, + "flos": 10328960908440.0, + "grad_norm": 2.489794595817899, + "learning_rate": 3.122529754172582e-06, + "loss": 0.81, + "num_input_tokens_seen": 58557035, + "step": 2753 + }, + { + "epoch": 0.33114892082005654, + "flos": 15190489120440.0, + "grad_norm": 2.702310163692295, + "learning_rate": 3.1218849638839736e-06, + "loss": 0.7105, + "num_input_tokens_seen": 58576015, + "step": 2754 + }, + { + "epoch": 0.3312691637106956, + "flos": 12461252656560.0, + "grad_norm": 2.074805340196287, + "learning_rate": 3.121240003406307e-06, + "loss": 0.7665, + "num_input_tokens_seen": 58594585, + "step": 2755 + }, + { + "epoch": 0.3313894066013347, + "flos": 21612086302200.0, + "grad_norm": 4.440321978673618, + "learning_rate": 3.120594872837425e-06, + "loss": 0.7041, + "num_input_tokens_seen": 58612975, + "step": 2756 + }, + { + "epoch": 0.3315096494919738, + "flos": 45272235164400.0, + "grad_norm": 0.8374637681326133, + "learning_rate": 3.1199495722751906e-06, + "loss": 0.6407, + "num_input_tokens_seen": 58672225, + "step": 2757 + }, + { + "epoch": 0.33162989238261287, + "flos": 15837890829720.0, + "grad_norm": 4.923326260185628, + "learning_rate": 3.1193041018174972e-06, + "loss": 0.8308, + "num_input_tokens_seen": 58692660, + "step": 2758 + }, + { + "epoch": 0.331750135273252, + "flos": 16594058196120.0, + "grad_norm": 4.380733330010741, + "learning_rate": 3.118658461562261e-06, + "loss": 0.9244, + "num_input_tokens_seen": 58708480, + "step": 2759 + }, + { + "epoch": 0.33187037816389103, + "flos": 16647301266120.0, + "grad_norm": 1.5392393213929856, + "learning_rate": 3.118012651607426e-06, + "loss": 0.8356, + "num_input_tokens_seen": 58729805, + "step": 2760 + }, + { + "epoch": 0.33199062105453014, + "flos": 14024582602200.0, + "grad_norm": 3.396169322111771, + "learning_rate": 3.1173666720509603e-06, + "loss": 0.8181, + "num_input_tokens_seen": 58746460, + "step": 2761 + }, + { + "epoch": 0.33211086394516925, + "flos": 23182254799200.0, + "grad_norm": 2.1887707969945422, + "learning_rate": 3.116720522990859e-06, + "loss": 0.6662, + "num_input_tokens_seen": 58767055, + "step": 2762 + }, + { + "epoch": 0.3322311068358083, + "flos": 13084565627640.0, + "grad_norm": 2.8976674663607396, + "learning_rate": 3.116074204525142e-06, + "loss": 0.6026, + "num_input_tokens_seen": 58784950, + "step": 2763 + }, + { + "epoch": 0.3323513497264474, + "flos": 23694934175520.0, + "grad_norm": 1.6909992530049591, + "learning_rate": 3.1154277167518553e-06, + "loss": 0.813, + "num_input_tokens_seen": 58806285, + "step": 2764 + }, + { + "epoch": 0.33247159261708653, + "flos": 38797078313040.0, + "grad_norm": 0.8488128084924136, + "learning_rate": 3.114781059769072e-06, + "loss": 0.6186, + "num_input_tokens_seen": 58857330, + "step": 2765 + }, + { + "epoch": 0.3325918355077256, + "flos": 19887570788400.0, + "grad_norm": 3.6137437679061537, + "learning_rate": 3.1141342336748874e-06, + "loss": 0.6625, + "num_input_tokens_seen": 58876610, + "step": 2766 + }, + { + "epoch": 0.3327120783983647, + "flos": 17327086623120.0, + "grad_norm": 1.8524739400289354, + "learning_rate": 3.1134872385674253e-06, + "loss": 0.8004, + "num_input_tokens_seen": 58900485, + "step": 2767 + }, + { + "epoch": 0.3328323212890038, + "flos": 14001190383120.0, + "grad_norm": 2.3715248784323215, + "learning_rate": 3.1128400745448353e-06, + "loss": 0.8505, + "num_input_tokens_seen": 58919585, + "step": 2768 + }, + { + "epoch": 0.33295256417964286, + "flos": 27718340858760.0, + "grad_norm": 2.5380424445881524, + "learning_rate": 3.11219274170529e-06, + "loss": 0.6164, + "num_input_tokens_seen": 58941115, + "step": 2769 + }, + { + "epoch": 0.333072807070282, + "flos": 19429432540440.0, + "grad_norm": 3.373364935514564, + "learning_rate": 3.1115452401469903e-06, + "loss": 0.8039, + "num_input_tokens_seen": 58961235, + "step": 2770 + }, + { + "epoch": 0.3331930499609211, + "flos": 15673032706200.0, + "grad_norm": 1.871321630438367, + "learning_rate": 3.1108975699681613e-06, + "loss": 0.8519, + "num_input_tokens_seen": 58978350, + "step": 2771 + }, + { + "epoch": 0.33331329285156014, + "flos": 15101791386600.0, + "grad_norm": 1.8450890543878433, + "learning_rate": 3.1102497312670542e-06, + "loss": 0.7021, + "num_input_tokens_seen": 58996075, + "step": 2772 + }, + { + "epoch": 0.33343353574219925, + "flos": 20536238896080.0, + "grad_norm": 2.7741890762018326, + "learning_rate": 3.109601724141946e-06, + "loss": 0.7783, + "num_input_tokens_seen": 59014790, + "step": 2773 + }, + { + "epoch": 0.33355377863283836, + "flos": 17400239316600.0, + "grad_norm": 2.061355504469839, + "learning_rate": 3.108953548691138e-06, + "loss": 0.678, + "num_input_tokens_seen": 59034595, + "step": 2774 + }, + { + "epoch": 0.3336740215234774, + "flos": 27772660367400.0, + "grad_norm": 3.759538091735208, + "learning_rate": 3.108305205012959e-06, + "loss": 0.7001, + "num_input_tokens_seen": 59055010, + "step": 2775 + }, + { + "epoch": 0.3337942644141165, + "flos": 18698778610440.0, + "grad_norm": 2.5200933731279576, + "learning_rate": 3.107656693205761e-06, + "loss": 0.8689, + "num_input_tokens_seen": 59074170, + "step": 2776 + }, + { + "epoch": 0.3339145073047556, + "flos": 19046694843120.0, + "grad_norm": 2.602406774311413, + "learning_rate": 3.107008013367924e-06, + "loss": 0.6723, + "num_input_tokens_seen": 59092685, + "step": 2777 + }, + { + "epoch": 0.3340347501953947, + "flos": 13920755898840.0, + "grad_norm": 4.906168748705194, + "learning_rate": 3.1063591655978507e-06, + "loss": 0.8421, + "num_input_tokens_seen": 59108355, + "step": 2778 + }, + { + "epoch": 0.3341549930860338, + "flos": 13214982206040.0, + "grad_norm": 3.1333342234292356, + "learning_rate": 3.105710149993972e-06, + "loss": 0.7747, + "num_input_tokens_seen": 59127405, + "step": 2779 + }, + { + "epoch": 0.33427523597667286, + "flos": 16594216495920.0, + "grad_norm": 2.5959018834973553, + "learning_rate": 3.1050609666547427e-06, + "loss": 0.8359, + "num_input_tokens_seen": 59146685, + "step": 2780 + }, + { + "epoch": 0.33439547886731197, + "flos": 16567088401560.0, + "grad_norm": 2.3562494252657, + "learning_rate": 3.104411615678644e-06, + "loss": 0.7623, + "num_input_tokens_seen": 59165255, + "step": 2781 + }, + { + "epoch": 0.3345157217579511, + "flos": 17645943503880.0, + "grad_norm": 3.889336430062452, + "learning_rate": 3.1037620971641803e-06, + "loss": 0.7119, + "num_input_tokens_seen": 59184765, + "step": 2782 + }, + { + "epoch": 0.33463596464859013, + "flos": 13866658009920.0, + "grad_norm": 4.400256686959616, + "learning_rate": 3.1031124112098844e-06, + "loss": 0.6313, + "num_input_tokens_seen": 59202695, + "step": 2783 + }, + { + "epoch": 0.33475620753922924, + "flos": 14892428342040.0, + "grad_norm": 2.402334619520251, + "learning_rate": 3.1024625579143127e-06, + "loss": 0.7068, + "num_input_tokens_seen": 59219935, + "step": 2784 + }, + { + "epoch": 0.33487645042986836, + "flos": 13268890135200.0, + "grad_norm": 2.5258036288288586, + "learning_rate": 3.101812537376048e-06, + "loss": 0.711, + "num_input_tokens_seen": 59238675, + "step": 2785 + }, + { + "epoch": 0.3349966933205074, + "flos": 18943026439560.0, + "grad_norm": 2.2000343883601094, + "learning_rate": 3.1011623496936973e-06, + "loss": 0.8198, + "num_input_tokens_seen": 59256690, + "step": 2786 + }, + { + "epoch": 0.3351169362111465, + "flos": 20617591519200.0, + "grad_norm": 2.474861788503023, + "learning_rate": 3.100511994965893e-06, + "loss": 0.6781, + "num_input_tokens_seen": 59276365, + "step": 2787 + }, + { + "epoch": 0.33523717910178563, + "flos": 16592760137760.0, + "grad_norm": 1.9021338752418693, + "learning_rate": 3.0998614732912947e-06, + "loss": 0.8447, + "num_input_tokens_seen": 59295460, + "step": 2788 + }, + { + "epoch": 0.3353574219924247, + "flos": 11409937228080.0, + "grad_norm": 3.1033547553590877, + "learning_rate": 3.0992107847685855e-06, + "loss": 0.6563, + "num_input_tokens_seen": 59312895, + "step": 2789 + }, + { + "epoch": 0.3354776648830638, + "flos": 18159889278600.0, + "grad_norm": 2.9691279120587755, + "learning_rate": 3.0985599294964736e-06, + "loss": 0.7805, + "num_input_tokens_seen": 59332170, + "step": 2790 + }, + { + "epoch": 0.33559790777370285, + "flos": 21049298191920.0, + "grad_norm": 3.2184092517223886, + "learning_rate": 3.097908907573695e-06, + "loss": 0.6953, + "num_input_tokens_seen": 59349870, + "step": 2791 + }, + { + "epoch": 0.33571815066434196, + "flos": 16268995963200.0, + "grad_norm": 4.428296498145566, + "learning_rate": 3.0972577190990067e-06, + "loss": 0.8783, + "num_input_tokens_seen": 59368070, + "step": 2792 + }, + { + "epoch": 0.3358383935549811, + "flos": 17458073080800.0, + "grad_norm": 2.7721360282604617, + "learning_rate": 3.096606364171196e-06, + "loss": 0.7849, + "num_input_tokens_seen": 59387580, + "step": 2793 + }, + { + "epoch": 0.33595863644562013, + "flos": 16291628343240.0, + "grad_norm": 3.1560795804084365, + "learning_rate": 3.0959548428890703e-06, + "loss": 0.8308, + "num_input_tokens_seen": 59406170, + "step": 2794 + }, + { + "epoch": 0.33607887933625924, + "flos": 14703038240880.0, + "grad_norm": 2.7110090878403614, + "learning_rate": 3.095303155351468e-06, + "loss": 0.8179, + "num_input_tokens_seen": 59426095, + "step": 2795 + }, + { + "epoch": 0.33619912222689835, + "flos": 14109892720320.0, + "grad_norm": 3.2022571411482037, + "learning_rate": 3.0946513016572464e-06, + "loss": 0.7677, + "num_input_tokens_seen": 59444720, + "step": 2796 + }, + { + "epoch": 0.3363193651175374, + "flos": 12246824018400.0, + "grad_norm": 2.694948033555462, + "learning_rate": 3.0939992819052938e-06, + "loss": 0.7461, + "num_input_tokens_seen": 59461950, + "step": 2797 + }, + { + "epoch": 0.3364396080081765, + "flos": 17241966464760.0, + "grad_norm": 2.360719209446996, + "learning_rate": 3.0933470961945193e-06, + "loss": 0.7918, + "num_input_tokens_seen": 59479965, + "step": 2798 + }, + { + "epoch": 0.3365598508988156, + "flos": 20563082050800.0, + "grad_norm": 2.4859422262879334, + "learning_rate": 3.0926947446238597e-06, + "loss": 0.6727, + "num_input_tokens_seen": 59499255, + "step": 2799 + }, + { + "epoch": 0.3366800937894547, + "flos": 12381546351360.0, + "grad_norm": 2.743635876482772, + "learning_rate": 3.092042227292276e-06, + "loss": 0.8088, + "num_input_tokens_seen": 59515810, + "step": 2800 + }, + { + "epoch": 0.3368003366800938, + "flos": 17512139309760.0, + "grad_norm": 1.9323821616424721, + "learning_rate": 3.0913895442987557e-06, + "loss": 0.8676, + "num_input_tokens_seen": 59536495, + "step": 2801 + }, + { + "epoch": 0.3369205795707329, + "flos": 18160395837960.0, + "grad_norm": 1.9152441620114702, + "learning_rate": 3.090736695742308e-06, + "loss": 0.8392, + "num_input_tokens_seen": 59557345, + "step": 2802 + }, + { + "epoch": 0.33704082246137196, + "flos": 13083330889200.0, + "grad_norm": 3.65179876657619, + "learning_rate": 3.0900836817219713e-06, + "loss": 0.489, + "num_input_tokens_seen": 59573495, + "step": 2803 + }, + { + "epoch": 0.33716106535201107, + "flos": 15566388266400.0, + "grad_norm": 2.2180627107620947, + "learning_rate": 3.089430502336807e-06, + "loss": 0.8269, + "num_input_tokens_seen": 59593185, + "step": 2804 + }, + { + "epoch": 0.3372813082426502, + "flos": 13432196920680.0, + "grad_norm": 3.916263776037178, + "learning_rate": 3.088777157685902e-06, + "loss": 0.88, + "num_input_tokens_seen": 59608495, + "step": 2805 + }, + { + "epoch": 0.33740155113328923, + "flos": 12543301798800.0, + "grad_norm": 2.423588497295618, + "learning_rate": 3.088123647868367e-06, + "loss": 0.8416, + "num_input_tokens_seen": 59624765, + "step": 2806 + }, + { + "epoch": 0.33752179402392835, + "flos": 21318647877960.0, + "grad_norm": 3.1775915187618784, + "learning_rate": 3.0874699729833405e-06, + "loss": 0.7968, + "num_input_tokens_seen": 59645855, + "step": 2807 + }, + { + "epoch": 0.3376420369145674, + "flos": 18374317916760.0, + "grad_norm": 2.608517625857019, + "learning_rate": 3.086816133129983e-06, + "loss": 0.7856, + "num_input_tokens_seen": 59665835, + "step": 2808 + }, + { + "epoch": 0.3377622798052065, + "flos": 20158060233000.0, + "grad_norm": 2.3427184823171, + "learning_rate": 3.0861621284074826e-06, + "loss": 0.7535, + "num_input_tokens_seen": 59686080, + "step": 2809 + }, + { + "epoch": 0.3378825226958456, + "flos": 16075426747320.0, + "grad_norm": 2.3712888131511893, + "learning_rate": 3.085507958915051e-06, + "loss": 0.7139, + "num_input_tokens_seen": 59704230, + "step": 2810 + }, + { + "epoch": 0.3380027655864847, + "flos": 31283229374520.0, + "grad_norm": 2.3669193732869553, + "learning_rate": 3.084853624751925e-06, + "loss": 0.6931, + "num_input_tokens_seen": 59725535, + "step": 2811 + }, + { + "epoch": 0.3381230084771238, + "flos": 19591599567360.0, + "grad_norm": 3.1773923796549823, + "learning_rate": 3.0841991260173668e-06, + "loss": 0.831, + "num_input_tokens_seen": 59745160, + "step": 2812 + }, + { + "epoch": 0.3382432513677629, + "flos": 16620268151640.0, + "grad_norm": 2.840641712437516, + "learning_rate": 3.0835444628106634e-06, + "loss": 0.7849, + "num_input_tokens_seen": 59763860, + "step": 2813 + }, + { + "epoch": 0.33836349425840195, + "flos": 16184540664000.0, + "grad_norm": 2.375302446080849, + "learning_rate": 3.082889635231126e-06, + "loss": 0.8298, + "num_input_tokens_seen": 59782240, + "step": 2814 + }, + { + "epoch": 0.33848373714904106, + "flos": 20022957980520.0, + "grad_norm": 3.6272887535288736, + "learning_rate": 3.0822346433780925e-06, + "loss": 0.7289, + "num_input_tokens_seen": 59802685, + "step": 2815 + }, + { + "epoch": 0.3386039800396802, + "flos": 18942931459680.0, + "grad_norm": 2.0595118522034426, + "learning_rate": 3.0815794873509237e-06, + "loss": 0.8292, + "num_input_tokens_seen": 59820690, + "step": 2816 + }, + { + "epoch": 0.33872422293031923, + "flos": 13786191865680.0, + "grad_norm": 3.236261969139388, + "learning_rate": 3.0809241672490066e-06, + "loss": 0.7088, + "num_input_tokens_seen": 59838580, + "step": 2817 + }, + { + "epoch": 0.33884446582095834, + "flos": 16943209167240.0, + "grad_norm": 2.4581498115270852, + "learning_rate": 3.080268683171753e-06, + "loss": 0.8332, + "num_input_tokens_seen": 59858590, + "step": 2818 + }, + { + "epoch": 0.33896470871159745, + "flos": 11652222139680.0, + "grad_norm": 3.4339418896583402, + "learning_rate": 3.0796130352185985e-06, + "loss": 0.8663, + "num_input_tokens_seen": 59875165, + "step": 2819 + }, + { + "epoch": 0.3390849516022365, + "flos": 25342307840880.0, + "grad_norm": 2.185841503857627, + "learning_rate": 3.0789572234890057e-06, + "loss": 0.665, + "num_input_tokens_seen": 59896525, + "step": 2820 + }, + { + "epoch": 0.3392051944928756, + "flos": 11787704311680.0, + "grad_norm": 2.22037382208955, + "learning_rate": 3.0783012480824596e-06, + "loss": 0.7625, + "num_input_tokens_seen": 59915390, + "step": 2821 + }, + { + "epoch": 0.33932543738351467, + "flos": 12457643421120.0, + "grad_norm": 2.8966728658430565, + "learning_rate": 3.077645109098471e-06, + "loss": 0.7288, + "num_input_tokens_seen": 59931380, + "step": 2822 + }, + { + "epoch": 0.3394456802741538, + "flos": 16188561478920.0, + "grad_norm": 2.548333060757523, + "learning_rate": 3.076988806636577e-06, + "loss": 0.7016, + "num_input_tokens_seen": 59948860, + "step": 2823 + }, + { + "epoch": 0.3395659231647929, + "flos": 18482513694600.0, + "grad_norm": 2.1279618434568444, + "learning_rate": 3.0763323407963377e-06, + "loss": 0.8644, + "num_input_tokens_seen": 59968190, + "step": 2824 + }, + { + "epoch": 0.33968616605543195, + "flos": 21346567471320.0, + "grad_norm": 2.3100614319926813, + "learning_rate": 3.075675711677337e-06, + "loss": 0.7751, + "num_input_tokens_seen": 59988005, + "step": 2825 + }, + { + "epoch": 0.33980640894607106, + "flos": 15675723802800.0, + "grad_norm": 9.888616131891572, + "learning_rate": 3.0750189193791865e-06, + "loss": 0.7624, + "num_input_tokens_seen": 60007310, + "step": 2826 + }, + { + "epoch": 0.33992665183671017, + "flos": 23858082661200.0, + "grad_norm": 3.874808645138233, + "learning_rate": 3.0743619640015203e-06, + "loss": 0.6921, + "num_input_tokens_seen": 60027280, + "step": 2827 + }, + { + "epoch": 0.3400468947273492, + "flos": 12435264320760.0, + "grad_norm": 2.3160706867242458, + "learning_rate": 3.073704845643999e-06, + "loss": 0.9082, + "num_input_tokens_seen": 60044125, + "step": 2828 + }, + { + "epoch": 0.34016713761798834, + "flos": 12300257048160.0, + "grad_norm": 3.343480036880592, + "learning_rate": 3.0730475644063063e-06, + "loss": 0.7625, + "num_input_tokens_seen": 60058945, + "step": 2829 + }, + { + "epoch": 0.34028738050862745, + "flos": 16025761252800.0, + "grad_norm": 2.3541835232921695, + "learning_rate": 3.072390120388151e-06, + "loss": 0.6457, + "num_input_tokens_seen": 60076990, + "step": 2830 + }, + { + "epoch": 0.3404076233992665, + "flos": 16646541427080.0, + "grad_norm": 3.0272868883268793, + "learning_rate": 3.071732513689267e-06, + "loss": 0.6826, + "num_input_tokens_seen": 60095245, + "step": 2831 + }, + { + "epoch": 0.3405278662899056, + "flos": 12432573224160.0, + "grad_norm": 3.4913311277388264, + "learning_rate": 3.0710747444094134e-06, + "loss": 0.6585, + "num_input_tokens_seen": 60112995, + "step": 2832 + }, + { + "epoch": 0.3406481091805447, + "flos": 31497531372840.0, + "grad_norm": 2.3780031792095677, + "learning_rate": 3.070416812648372e-06, + "loss": 0.6431, + "num_input_tokens_seen": 60136165, + "step": 2833 + }, + { + "epoch": 0.3407683520711838, + "flos": 19618506042000.0, + "grad_norm": 2.4331452031329346, + "learning_rate": 3.069758718505951e-06, + "loss": 0.6342, + "num_input_tokens_seen": 60157625, + "step": 2834 + }, + { + "epoch": 0.3408885949618229, + "flos": 21103142801160.0, + "grad_norm": 2.7585531218245074, + "learning_rate": 3.0691004620819836e-06, + "loss": 0.7825, + "num_input_tokens_seen": 60177475, + "step": 2835 + }, + { + "epoch": 0.341008837852462, + "flos": 46869183496200.0, + "grad_norm": 0.8052253037046572, + "learning_rate": 3.0684420434763254e-06, + "loss": 0.6274, + "num_input_tokens_seen": 60243380, + "step": 2836 + }, + { + "epoch": 0.34112908074310105, + "flos": 15215337697680.0, + "grad_norm": 2.8251391081192105, + "learning_rate": 3.06778346278886e-06, + "loss": 0.7488, + "num_input_tokens_seen": 60261935, + "step": 2837 + }, + { + "epoch": 0.34124932363374016, + "flos": 18297872587440.0, + "grad_norm": 2.5625392631686177, + "learning_rate": 3.0671247201194906e-06, + "loss": 0.7783, + "num_input_tokens_seen": 60283790, + "step": 2838 + }, + { + "epoch": 0.3413695665243792, + "flos": 20833286555760.0, + "grad_norm": 4.152672167015955, + "learning_rate": 3.066465815568151e-06, + "loss": 0.7295, + "num_input_tokens_seen": 60304340, + "step": 2839 + }, + { + "epoch": 0.34148980941501833, + "flos": 18538416201240.0, + "grad_norm": 2.3339893803882, + "learning_rate": 3.0658067492347947e-06, + "loss": 0.6753, + "num_input_tokens_seen": 60326700, + "step": 2840 + }, + { + "epoch": 0.34161005230565744, + "flos": 12786599829120.0, + "grad_norm": 3.137937392184635, + "learning_rate": 3.065147521219402e-06, + "loss": 0.6586, + "num_input_tokens_seen": 60345675, + "step": 2841 + }, + { + "epoch": 0.3417302951962965, + "flos": 32116633569240.0, + "grad_norm": 1.88243903417261, + "learning_rate": 3.064488131621977e-06, + "loss": 0.7386, + "num_input_tokens_seen": 60368720, + "step": 2842 + }, + { + "epoch": 0.3418505380869356, + "flos": 22667834125080.0, + "grad_norm": 2.400822848087795, + "learning_rate": 3.063828580542549e-06, + "loss": 0.7208, + "num_input_tokens_seen": 60389635, + "step": 2843 + }, + { + "epoch": 0.3419707809775747, + "flos": 14217740238600.0, + "grad_norm": 2.3979598075991775, + "learning_rate": 3.0631688680811706e-06, + "loss": 0.7022, + "num_input_tokens_seen": 60408980, + "step": 2844 + }, + { + "epoch": 0.3420910238682138, + "flos": 21073196970360.0, + "grad_norm": 2.6074355907293754, + "learning_rate": 3.062508994337921e-06, + "loss": 0.7452, + "num_input_tokens_seen": 60428305, + "step": 2845 + }, + { + "epoch": 0.3422112667588529, + "flos": 15648532388520.0, + "grad_norm": 3.101784186524025, + "learning_rate": 3.0618489594129013e-06, + "loss": 0.7644, + "num_input_tokens_seen": 60446165, + "step": 2846 + }, + { + "epoch": 0.342331509649492, + "flos": 10086359397240.0, + "grad_norm": 2.736019141581986, + "learning_rate": 3.061188763406239e-06, + "loss": 0.696, + "num_input_tokens_seen": 60462030, + "step": 2847 + }, + { + "epoch": 0.34245175254013105, + "flos": 20994915363360.0, + "grad_norm": 13.011367522841445, + "learning_rate": 3.060528406418085e-06, + "loss": 0.8039, + "num_input_tokens_seen": 60481600, + "step": 2848 + }, + { + "epoch": 0.34257199543077016, + "flos": 25068905679960.0, + "grad_norm": 2.201495711191381, + "learning_rate": 3.0598678885486145e-06, + "loss": 0.6002, + "num_input_tokens_seen": 60503860, + "step": 2849 + }, + { + "epoch": 0.34269223832140927, + "flos": 14595190722600.0, + "grad_norm": 2.6223823574797023, + "learning_rate": 3.0592072098980282e-06, + "loss": 0.7272, + "num_input_tokens_seen": 60523240, + "step": 2850 + }, + { + "epoch": 0.3428124812120483, + "flos": 19969429970880.0, + "grad_norm": 4.363430092372342, + "learning_rate": 3.0585463705665514e-06, + "loss": 0.7076, + "num_input_tokens_seen": 60543335, + "step": 2851 + }, + { + "epoch": 0.34293272410268744, + "flos": 17997120712440.0, + "grad_norm": 3.5637841838052196, + "learning_rate": 3.0578853706544304e-06, + "loss": 0.6858, + "num_input_tokens_seen": 60560445, + "step": 2852 + }, + { + "epoch": 0.34305296699332655, + "flos": 15729663391920.0, + "grad_norm": 2.602334306116298, + "learning_rate": 3.0572242102619404e-06, + "loss": 0.6356, + "num_input_tokens_seen": 60577320, + "step": 2853 + }, + { + "epoch": 0.3431732098839656, + "flos": 17616092652960.0, + "grad_norm": 6.2159472719997595, + "learning_rate": 3.0565628894893784e-06, + "loss": 0.7992, + "num_input_tokens_seen": 60597675, + "step": 2854 + }, + { + "epoch": 0.3432934527746047, + "flos": 12247013978160.0, + "grad_norm": 2.3685591849819696, + "learning_rate": 3.0559014084370655e-06, + "loss": 0.7331, + "num_input_tokens_seen": 60615920, + "step": 2855 + }, + { + "epoch": 0.34341369566524377, + "flos": 17159980642440.0, + "grad_norm": 2.3303283915131963, + "learning_rate": 3.055239767205349e-06, + "loss": 0.7707, + "num_input_tokens_seen": 60637390, + "step": 2856 + }, + { + "epoch": 0.3435339385558829, + "flos": 12591225995520.0, + "grad_norm": 2.405170743595383, + "learning_rate": 3.054577965894599e-06, + "loss": 0.7574, + "num_input_tokens_seen": 60653255, + "step": 2857 + }, + { + "epoch": 0.343654181446522, + "flos": 16242057828600.0, + "grad_norm": 2.8108850717028653, + "learning_rate": 3.0539160046052094e-06, + "loss": 0.7056, + "num_input_tokens_seen": 60672675, + "step": 2858 + }, + { + "epoch": 0.34377442433716104, + "flos": 14541852672720.0, + "grad_norm": 2.761311070676487, + "learning_rate": 3.0532538834376003e-06, + "loss": 0.6859, + "num_input_tokens_seen": 60691955, + "step": 2859 + }, + { + "epoch": 0.34389466722780015, + "flos": 16240569810480.0, + "grad_norm": 3.1980201795560417, + "learning_rate": 3.0525916024922143e-06, + "loss": 0.7622, + "num_input_tokens_seen": 60710860, + "step": 2860 + }, + { + "epoch": 0.34401491011843927, + "flos": 13596580144800.0, + "grad_norm": 5.596202802733847, + "learning_rate": 3.0519291618695193e-06, + "loss": 0.8089, + "num_input_tokens_seen": 60727980, + "step": 2861 + }, + { + "epoch": 0.3441351530090783, + "flos": 13029391300080.0, + "grad_norm": 3.0091440399370017, + "learning_rate": 3.0512665616700065e-06, + "loss": 0.7323, + "num_input_tokens_seen": 60746765, + "step": 2862 + }, + { + "epoch": 0.34425539589971743, + "flos": 16918202290200.0, + "grad_norm": 3.327390777835951, + "learning_rate": 3.0506038019941933e-06, + "loss": 0.8786, + "num_input_tokens_seen": 60766495, + "step": 2863 + }, + { + "epoch": 0.34437563879035654, + "flos": 16026236152200.0, + "grad_norm": 13.243455530181041, + "learning_rate": 3.049940882942617e-06, + "loss": 0.6514, + "num_input_tokens_seen": 60785000, + "step": 2864 + }, + { + "epoch": 0.3444958816809956, + "flos": 16891770714960.0, + "grad_norm": 2.507906998134401, + "learning_rate": 3.0492778046158448e-06, + "loss": 0.776, + "num_input_tokens_seen": 60806140, + "step": 2865 + }, + { + "epoch": 0.3446161245716347, + "flos": 16027059311160.0, + "grad_norm": 2.6507867616865406, + "learning_rate": 3.0486145671144633e-06, + "loss": 0.7555, + "num_input_tokens_seen": 60825650, + "step": 2866 + }, + { + "epoch": 0.3447363674622738, + "flos": 18397456856160.0, + "grad_norm": 2.7693378935170117, + "learning_rate": 3.047951170539086e-06, + "loss": 0.7282, + "num_input_tokens_seen": 60844995, + "step": 2867 + }, + { + "epoch": 0.3448566103529129, + "flos": 8682695341680.0, + "grad_norm": 6.694948700643497, + "learning_rate": 3.047287614990349e-06, + "loss": 0.8205, + "num_input_tokens_seen": 60862635, + "step": 2868 + }, + { + "epoch": 0.344976853243552, + "flos": 29554914665520.0, + "grad_norm": 3.9294102231560477, + "learning_rate": 3.046623900568914e-06, + "loss": 0.5782, + "num_input_tokens_seen": 60884920, + "step": 2869 + }, + { + "epoch": 0.34509709613419104, + "flos": 21069524415000.0, + "grad_norm": 5.862224567507015, + "learning_rate": 3.045960027375465e-06, + "loss": 0.6779, + "num_input_tokens_seen": 60902475, + "step": 2870 + }, + { + "epoch": 0.34521733902483015, + "flos": 21990613224840.0, + "grad_norm": 4.082895579845016, + "learning_rate": 3.045295995510711e-06, + "loss": 0.7992, + "num_input_tokens_seen": 60919165, + "step": 2871 + }, + { + "epoch": 0.34533758191546926, + "flos": 20477328693240.0, + "grad_norm": 2.4643455521799935, + "learning_rate": 3.0446318050753865e-06, + "loss": 0.7213, + "num_input_tokens_seen": 60939365, + "step": 2872 + }, + { + "epoch": 0.3454578248061083, + "flos": 19888140667680.0, + "grad_norm": 2.6998286793361745, + "learning_rate": 3.0439674561702474e-06, + "loss": 0.7624, + "num_input_tokens_seen": 60958585, + "step": 2873 + }, + { + "epoch": 0.3455780676967474, + "flos": 13892709665640.0, + "grad_norm": 2.743336493890857, + "learning_rate": 3.043302948896076e-06, + "loss": 0.8629, + "num_input_tokens_seen": 60976910, + "step": 2874 + }, + { + "epoch": 0.34569831058738654, + "flos": 25342529460600.0, + "grad_norm": 4.483073720305016, + "learning_rate": 3.0426382833536756e-06, + "loss": 0.582, + "num_input_tokens_seen": 60999985, + "step": 2875 + }, + { + "epoch": 0.3458185534780256, + "flos": 22857794105520.0, + "grad_norm": 3.389405483901529, + "learning_rate": 3.041973459643877e-06, + "loss": 0.7704, + "num_input_tokens_seen": 61019160, + "step": 2876 + }, + { + "epoch": 0.3459387963686647, + "flos": 23830163067840.0, + "grad_norm": 2.4800294502109392, + "learning_rate": 3.0413084778675334e-06, + "loss": 0.6464, + "num_input_tokens_seen": 61040130, + "step": 2877 + }, + { + "epoch": 0.3460590392593038, + "flos": 18074009281200.0, + "grad_norm": 2.873093677920108, + "learning_rate": 3.0406433381255214e-06, + "loss": 0.8174, + "num_input_tokens_seen": 61057885, + "step": 2878 + }, + { + "epoch": 0.34617928214994287, + "flos": 13731872357040.0, + "grad_norm": 3.53644443035436, + "learning_rate": 3.0399780405187425e-06, + "loss": 0.807, + "num_input_tokens_seen": 61076600, + "step": 2879 + }, + { + "epoch": 0.346299525040582, + "flos": 18208636634280.0, + "grad_norm": 2.953452117402623, + "learning_rate": 3.0393125851481216e-06, + "loss": 0.7661, + "num_input_tokens_seen": 61096195, + "step": 2880 + }, + { + "epoch": 0.3464197679312211, + "flos": 11976017974200.0, + "grad_norm": 2.6190650695780895, + "learning_rate": 3.038646972114608e-06, + "loss": 0.8415, + "num_input_tokens_seen": 61112240, + "step": 2881 + }, + { + "epoch": 0.34654001082186014, + "flos": 16377666640440.0, + "grad_norm": 3.1414030259577617, + "learning_rate": 3.037981201519174e-06, + "loss": 0.6635, + "num_input_tokens_seen": 61132560, + "step": 2882 + }, + { + "epoch": 0.34666025371249926, + "flos": 14299251161520.0, + "grad_norm": 3.326346195153139, + "learning_rate": 3.0373152734628175e-06, + "loss": 0.6817, + "num_input_tokens_seen": 61150560, + "step": 2883 + }, + { + "epoch": 0.34678049660313837, + "flos": 11112193049280.0, + "grad_norm": 2.6030867410071594, + "learning_rate": 3.0366491880465584e-06, + "loss": 0.7358, + "num_input_tokens_seen": 61168300, + "step": 2884 + }, + { + "epoch": 0.3469007394937774, + "flos": 15489341397840.0, + "grad_norm": 2.3902135756763268, + "learning_rate": 3.035982945371443e-06, + "loss": 0.8081, + "num_input_tokens_seen": 61189715, + "step": 2885 + }, + { + "epoch": 0.34702098238441653, + "flos": 16371492948240.0, + "grad_norm": 2.9513285991020943, + "learning_rate": 3.035316545538537e-06, + "loss": 0.8427, + "num_input_tokens_seen": 61208230, + "step": 2886 + }, + { + "epoch": 0.3471412252750556, + "flos": 16782403518600.0, + "grad_norm": 2.929793292360728, + "learning_rate": 3.034649988648935e-06, + "loss": 0.7647, + "num_input_tokens_seen": 61227715, + "step": 2887 + }, + { + "epoch": 0.3472614681656947, + "flos": 15594497819520.0, + "grad_norm": 2.7766981273149542, + "learning_rate": 3.033983274803752e-06, + "loss": 0.799, + "num_input_tokens_seen": 61247225, + "step": 2888 + }, + { + "epoch": 0.3473817110563338, + "flos": 17186570517480.0, + "grad_norm": 3.7500474291027333, + "learning_rate": 3.0333164041041283e-06, + "loss": 0.7021, + "num_input_tokens_seen": 61263260, + "step": 2889 + }, + { + "epoch": 0.34750195394697286, + "flos": 16350696845880.0, + "grad_norm": 4.854235319817235, + "learning_rate": 3.032649376651228e-06, + "loss": 0.7071, + "num_input_tokens_seen": 61282400, + "step": 2890 + }, + { + "epoch": 0.347622196837612, + "flos": 21345269412960.0, + "grad_norm": 2.7284202040177723, + "learning_rate": 3.031982192546238e-06, + "loss": 0.7406, + "num_input_tokens_seen": 61305215, + "step": 2891 + }, + { + "epoch": 0.3477424397282511, + "flos": 16431922829160.0, + "grad_norm": 3.6167714351357168, + "learning_rate": 3.0313148518903696e-06, + "loss": 0.9287, + "num_input_tokens_seen": 61324760, + "step": 2892 + }, + { + "epoch": 0.34786268261889014, + "flos": 11490530012160.0, + "grad_norm": 3.041269163172469, + "learning_rate": 3.030647354784859e-06, + "loss": 0.7851, + "num_input_tokens_seen": 61341520, + "step": 2893 + }, + { + "epoch": 0.34798292550952925, + "flos": 15190045881000.0, + "grad_norm": 2.157245572007688, + "learning_rate": 3.029979701330964e-06, + "loss": 0.7585, + "num_input_tokens_seen": 61360665, + "step": 2894 + }, + { + "epoch": 0.34810316840016836, + "flos": 14568442547760.0, + "grad_norm": 2.7815093873184615, + "learning_rate": 3.029311891629966e-06, + "loss": 0.7794, + "num_input_tokens_seen": 61378840, + "step": 2895 + }, + { + "epoch": 0.3482234112908074, + "flos": 17296001033760.0, + "grad_norm": 2.4605925787547696, + "learning_rate": 3.0286439257831744e-06, + "loss": 0.723, + "num_input_tokens_seen": 61398030, + "step": 2896 + }, + { + "epoch": 0.3483436541814465, + "flos": 17564749180560.0, + "grad_norm": 3.169494336829818, + "learning_rate": 3.0279758038919156e-06, + "loss": 0.6939, + "num_input_tokens_seen": 61415975, + "step": 2897 + }, + { + "epoch": 0.34846389707208564, + "flos": 16567088401560.0, + "grad_norm": 2.4143262799290013, + "learning_rate": 3.0273075260575455e-06, + "loss": 0.7578, + "num_input_tokens_seen": 61434595, + "step": 2898 + }, + { + "epoch": 0.3485841399627247, + "flos": 15940482794640.0, + "grad_norm": 2.6382304513438917, + "learning_rate": 3.0266390923814396e-06, + "loss": 0.7788, + "num_input_tokens_seen": 61452375, + "step": 2899 + }, + { + "epoch": 0.3487043828533638, + "flos": 12408357846120.0, + "grad_norm": 3.282064282333474, + "learning_rate": 3.0259705029650008e-06, + "loss": 0.8067, + "num_input_tokens_seen": 61470025, + "step": 2900 + }, + { + "epoch": 0.34882462574400286, + "flos": 16539453747840.0, + "grad_norm": 2.740661196322848, + "learning_rate": 3.025301757909652e-06, + "loss": 0.7136, + "num_input_tokens_seen": 61489940, + "step": 2901 + }, + { + "epoch": 0.34894486863464197, + "flos": 21912584897520.0, + "grad_norm": 2.314799840525478, + "learning_rate": 3.024632857316842e-06, + "loss": 0.7957, + "num_input_tokens_seen": 61510975, + "step": 2902 + }, + { + "epoch": 0.3490651115252811, + "flos": 16184414024160.0, + "grad_norm": 3.4342390371940335, + "learning_rate": 3.0239638012880412e-06, + "loss": 0.7523, + "num_input_tokens_seen": 61530590, + "step": 2903 + }, + { + "epoch": 0.34918535441592014, + "flos": 9194646538920.0, + "grad_norm": 3.9099702743219855, + "learning_rate": 3.0232945899247466e-06, + "loss": 0.7994, + "num_input_tokens_seen": 61547245, + "step": 2904 + }, + { + "epoch": 0.34930559730655925, + "flos": 16972110219360.0, + "grad_norm": 3.9007858511654843, + "learning_rate": 3.022625223328476e-06, + "loss": 0.7589, + "num_input_tokens_seen": 61568705, + "step": 2905 + }, + { + "epoch": 0.34942584019719836, + "flos": 16727134211160.0, + "grad_norm": 2.008569953032651, + "learning_rate": 3.0219557016007723e-06, + "loss": 0.6826, + "num_input_tokens_seen": 61588555, + "step": 2906 + }, + { + "epoch": 0.3495460830878374, + "flos": 17890096353120.0, + "grad_norm": 2.0955136969884873, + "learning_rate": 3.021286024843202e-06, + "loss": 0.684, + "num_input_tokens_seen": 61606470, + "step": 2907 + }, + { + "epoch": 0.3496663259784765, + "flos": 51626057354520.0, + "grad_norm": 1.1071624182805704, + "learning_rate": 3.0206161931573526e-06, + "loss": 0.6966, + "num_input_tokens_seen": 61658740, + "step": 2908 + }, + { + "epoch": 0.34978656886911563, + "flos": 21047493574200.0, + "grad_norm": 2.087082337952743, + "learning_rate": 3.0199462066448388e-06, + "loss": 0.9218, + "num_input_tokens_seen": 61680655, + "step": 2909 + }, + { + "epoch": 0.3499068117597547, + "flos": 15459775486560.0, + "grad_norm": 3.116511603865188, + "learning_rate": 3.019276065407296e-06, + "loss": 0.6744, + "num_input_tokens_seen": 61699495, + "step": 2910 + }, + { + "epoch": 0.3500270546503938, + "flos": 16673701181400.0, + "grad_norm": 2.57483180709709, + "learning_rate": 3.018605769546385e-06, + "loss": 0.7953, + "num_input_tokens_seen": 61719770, + "step": 2911 + }, + { + "epoch": 0.3501472975410329, + "flos": 16756383522840.0, + "grad_norm": 2.108795877713992, + "learning_rate": 3.017935319163788e-06, + "loss": 0.7849, + "num_input_tokens_seen": 61738450, + "step": 2912 + }, + { + "epoch": 0.35026754043167196, + "flos": 18645218940840.0, + "grad_norm": 4.929820208569311, + "learning_rate": 3.017264714361213e-06, + "loss": 0.6979, + "num_input_tokens_seen": 61757820, + "step": 2913 + }, + { + "epoch": 0.3503877833223111, + "flos": 14299124521680.0, + "grad_norm": 2.621086109224146, + "learning_rate": 3.016593955240389e-06, + "loss": 0.804, + "num_input_tokens_seen": 61776230, + "step": 2914 + }, + { + "epoch": 0.3505080262129502, + "flos": 47976179811600.0, + "grad_norm": 0.8076530919056626, + "learning_rate": 3.015923041903071e-06, + "loss": 0.649, + "num_input_tokens_seen": 61842075, + "step": 2915 + }, + { + "epoch": 0.35062826910358924, + "flos": 21507848019360.0, + "grad_norm": 3.922151472217832, + "learning_rate": 3.0152519744510347e-06, + "loss": 0.8158, + "num_input_tokens_seen": 61861595, + "step": 2916 + }, + { + "epoch": 0.35074851199422835, + "flos": 17565635659440.0, + "grad_norm": 3.5708087601292813, + "learning_rate": 3.014580752986081e-06, + "loss": 0.8144, + "num_input_tokens_seen": 61880190, + "step": 2917 + }, + { + "epoch": 0.3508687548848674, + "flos": 11079271182240.0, + "grad_norm": 2.3265476900119944, + "learning_rate": 3.0139093776100345e-06, + "loss": 0.7661, + "num_input_tokens_seen": 61896500, + "step": 2918 + }, + { + "epoch": 0.3509889977755065, + "flos": 15622385752920.0, + "grad_norm": 2.6141135311674986, + "learning_rate": 3.013237848424741e-06, + "loss": 0.7372, + "num_input_tokens_seen": 61915605, + "step": 2919 + }, + { + "epoch": 0.35110924066614563, + "flos": 13974600508080.0, + "grad_norm": 5.658037609224805, + "learning_rate": 3.012566165532072e-06, + "loss": 0.7407, + "num_input_tokens_seen": 61934115, + "step": 2920 + }, + { + "epoch": 0.3512294835567847, + "flos": 16080429021000.0, + "grad_norm": 3.375010596364659, + "learning_rate": 3.0118943290339207e-06, + "loss": 0.7408, + "num_input_tokens_seen": 61954045, + "step": 2921 + }, + { + "epoch": 0.3513497264474238, + "flos": 12998495670480.0, + "grad_norm": 2.8182233432607857, + "learning_rate": 3.011222339032204e-06, + "loss": 0.6533, + "num_input_tokens_seen": 61971915, + "step": 2922 + }, + { + "epoch": 0.3514699693380629, + "flos": 19753101735120.0, + "grad_norm": 2.3868725816065663, + "learning_rate": 3.0105501956288626e-06, + "loss": 0.6713, + "num_input_tokens_seen": 61992105, + "step": 2923 + }, + { + "epoch": 0.35159021222870196, + "flos": 11112129729360.0, + "grad_norm": 4.297534572957118, + "learning_rate": 3.0098778989258602e-06, + "loss": 0.7087, + "num_input_tokens_seen": 62010435, + "step": 2924 + }, + { + "epoch": 0.35171045511934107, + "flos": 10161411688320.0, + "grad_norm": 2.9127601802020373, + "learning_rate": 3.009205449025183e-06, + "loss": 0.857, + "num_input_tokens_seen": 62026350, + "step": 2925 + }, + { + "epoch": 0.3518306980099802, + "flos": 10383945276240.0, + "grad_norm": 4.65503671699245, + "learning_rate": 3.008532846028842e-06, + "loss": 0.6099, + "num_input_tokens_seen": 62042830, + "step": 2926 + }, + { + "epoch": 0.35195094090061924, + "flos": 19834106098680.0, + "grad_norm": 4.642099062893509, + "learning_rate": 3.0078600900388694e-06, + "loss": 0.6883, + "num_input_tokens_seen": 62062855, + "step": 2927 + }, + { + "epoch": 0.35207118379125835, + "flos": 18780447833160.0, + "grad_norm": 2.5236897534401197, + "learning_rate": 3.007187181157323e-06, + "loss": 0.726, + "num_input_tokens_seen": 62082585, + "step": 2928 + }, + { + "epoch": 0.35219142668189746, + "flos": 13137903677520.0, + "grad_norm": 3.1619272067662583, + "learning_rate": 3.006514119486282e-06, + "loss": 0.6625, + "num_input_tokens_seen": 62099135, + "step": 2929 + }, + { + "epoch": 0.3523116695725365, + "flos": 10194998414520.0, + "grad_norm": 2.6477997948890883, + "learning_rate": 3.005840905127849e-06, + "loss": 0.6821, + "num_input_tokens_seen": 62115760, + "step": 2930 + }, + { + "epoch": 0.3524319124631756, + "flos": 15675723802800.0, + "grad_norm": 2.7960509039811505, + "learning_rate": 3.0051675381841516e-06, + "loss": 0.8458, + "num_input_tokens_seen": 62132790, + "step": 2931 + }, + { + "epoch": 0.3525521553538147, + "flos": 19293728748720.0, + "grad_norm": 3.1143891817306266, + "learning_rate": 3.0044940187573363e-06, + "loss": 0.7505, + "num_input_tokens_seen": 62153520, + "step": 2932 + }, + { + "epoch": 0.3526723982444538, + "flos": 15756633186480.0, + "grad_norm": 2.781672584918998, + "learning_rate": 3.003820346949578e-06, + "loss": 0.6568, + "num_input_tokens_seen": 62171320, + "step": 2933 + }, + { + "epoch": 0.3527926411350929, + "flos": 17377796896320.0, + "grad_norm": 3.370931504324605, + "learning_rate": 3.003146522863071e-06, + "loss": 0.7755, + "num_input_tokens_seen": 62191925, + "step": 2934 + }, + { + "epoch": 0.35291288402573195, + "flos": 22344766469640.0, + "grad_norm": 3.2539770294954278, + "learning_rate": 3.0024725466000345e-06, + "loss": 0.8316, + "num_input_tokens_seen": 62211600, + "step": 2935 + }, + { + "epoch": 0.35303312691637107, + "flos": 16917695730840.0, + "grad_norm": 2.868058490726015, + "learning_rate": 3.0017984182627087e-06, + "loss": 0.7736, + "num_input_tokens_seen": 62230645, + "step": 2936 + }, + { + "epoch": 0.3531533698070102, + "flos": 15973088062080.0, + "grad_norm": 4.172576273866119, + "learning_rate": 3.00112413795336e-06, + "loss": 0.7968, + "num_input_tokens_seen": 62250200, + "step": 2937 + }, + { + "epoch": 0.35327361269764923, + "flos": 11490688311960.0, + "grad_norm": 2.7815329589920785, + "learning_rate": 3.000449705774275e-06, + "loss": 0.7935, + "num_input_tokens_seen": 62268160, + "step": 2938 + }, + { + "epoch": 0.35339385558828834, + "flos": 16161496704480.0, + "grad_norm": 3.866009157007131, + "learning_rate": 2.9997751218277654e-06, + "loss": 0.6951, + "num_input_tokens_seen": 62286035, + "step": 2939 + }, + { + "epoch": 0.35351409847892745, + "flos": 17696115557760.0, + "grad_norm": 3.6743493366983633, + "learning_rate": 2.999100386216166e-06, + "loss": 0.757, + "num_input_tokens_seen": 62304695, + "step": 2940 + }, + { + "epoch": 0.3536343413695665, + "flos": 19833599539320.0, + "grad_norm": 2.759844498745732, + "learning_rate": 2.998425499041831e-06, + "loss": 0.7213, + "num_input_tokens_seen": 62324930, + "step": 2941 + }, + { + "epoch": 0.3537545842602056, + "flos": 48657263226960.0, + "grad_norm": 1.030368659513995, + "learning_rate": 2.997750460407142e-06, + "loss": 0.6219, + "num_input_tokens_seen": 62386005, + "step": 2942 + }, + { + "epoch": 0.35387482715084473, + "flos": 13456950518040.0, + "grad_norm": 11.574579970316755, + "learning_rate": 2.997075270414501e-06, + "loss": 0.6856, + "num_input_tokens_seen": 62402940, + "step": 2943 + }, + { + "epoch": 0.3539950700414838, + "flos": 48359234108520.0, + "grad_norm": 0.7025044139709815, + "learning_rate": 2.9963999291663347e-06, + "loss": 0.5933, + "num_input_tokens_seen": 62468440, + "step": 2944 + }, + { + "epoch": 0.3541153129321229, + "flos": 15000529140000.0, + "grad_norm": 4.277070615922639, + "learning_rate": 2.9957244367650915e-06, + "loss": 0.7281, + "num_input_tokens_seen": 62484405, + "step": 2945 + }, + { + "epoch": 0.354235555822762, + "flos": 14298966221880.0, + "grad_norm": 3.1745610632071264, + "learning_rate": 2.9950487933132425e-06, + "loss": 0.8216, + "num_input_tokens_seen": 62501540, + "step": 2946 + }, + { + "epoch": 0.35435579871340106, + "flos": 15189729281400.0, + "grad_norm": 4.229229154579278, + "learning_rate": 2.994372998913283e-06, + "loss": 0.6933, + "num_input_tokens_seen": 62519765, + "step": 2947 + }, + { + "epoch": 0.35447604160404017, + "flos": 17539140764280.0, + "grad_norm": 7.751121685932735, + "learning_rate": 2.99369705366773e-06, + "loss": 0.6111, + "num_input_tokens_seen": 62539730, + "step": 2948 + }, + { + "epoch": 0.3545962844946792, + "flos": 17156403066960.0, + "grad_norm": 2.866487755478319, + "learning_rate": 2.9930209576791244e-06, + "loss": 0.7932, + "num_input_tokens_seen": 62557925, + "step": 2949 + }, + { + "epoch": 0.35471652738531834, + "flos": 16156842690360.0, + "grad_norm": 3.0135187096528426, + "learning_rate": 2.9923447110500285e-06, + "loss": 0.6336, + "num_input_tokens_seen": 62576390, + "step": 2950 + }, + { + "epoch": 0.35483677027595745, + "flos": 20047584938040.0, + "grad_norm": 2.07107668034654, + "learning_rate": 2.9916683138830295e-06, + "loss": 0.7353, + "num_input_tokens_seen": 62596775, + "step": 2951 + }, + { + "epoch": 0.3549570131665965, + "flos": 9815964932520.0, + "grad_norm": 9.709581766913491, + "learning_rate": 2.9909917662807353e-06, + "loss": 0.7863, + "num_input_tokens_seen": 62614295, + "step": 2952 + }, + { + "epoch": 0.3550772560572356, + "flos": 15271461824040.0, + "grad_norm": 4.739593349348391, + "learning_rate": 2.9903150683457783e-06, + "loss": 0.6619, + "num_input_tokens_seen": 62632560, + "step": 2953 + }, + { + "epoch": 0.3551974989478747, + "flos": 14758180908480.0, + "grad_norm": 3.624919101207006, + "learning_rate": 2.9896382201808126e-06, + "loss": 0.6261, + "num_input_tokens_seen": 62649680, + "step": 2954 + }, + { + "epoch": 0.3553177418385138, + "flos": 14487849763680.0, + "grad_norm": 2.979085178639059, + "learning_rate": 2.988961221888516e-06, + "loss": 0.7839, + "num_input_tokens_seen": 62666075, + "step": 2955 + }, + { + "epoch": 0.3554379847291529, + "flos": 10787574055800.0, + "grad_norm": 3.977901091046499, + "learning_rate": 2.988284073571589e-06, + "loss": 0.7768, + "num_input_tokens_seen": 62681880, + "step": 2956 + }, + { + "epoch": 0.355558227619792, + "flos": 14973685985280.0, + "grad_norm": 5.917958114161235, + "learning_rate": 2.9876067753327528e-06, + "loss": 0.7025, + "num_input_tokens_seen": 62699330, + "step": 2957 + }, + { + "epoch": 0.35567847051043106, + "flos": 27685830571200.0, + "grad_norm": 2.9671885349297993, + "learning_rate": 2.986929327274754e-06, + "loss": 0.7776, + "num_input_tokens_seen": 62719630, + "step": 2958 + }, + { + "epoch": 0.35579871340107017, + "flos": 19752785135520.0, + "grad_norm": 2.7013017095005756, + "learning_rate": 2.9862517295003617e-06, + "loss": 0.764, + "num_input_tokens_seen": 62739765, + "step": 2959 + }, + { + "epoch": 0.3559189562917093, + "flos": 20752123892400.0, + "grad_norm": 6.648889044412032, + "learning_rate": 2.9855739821123654e-06, + "loss": 0.7151, + "num_input_tokens_seen": 62761065, + "step": 2960 + }, + { + "epoch": 0.35603919918234833, + "flos": 18805961269560.0, + "grad_norm": 2.5419629242216883, + "learning_rate": 2.98489608521358e-06, + "loss": 0.8036, + "num_input_tokens_seen": 62780725, + "step": 2961 + }, + { + "epoch": 0.35615944207298744, + "flos": 16835045049360.0, + "grad_norm": 3.4316686934146636, + "learning_rate": 2.9842180389068425e-06, + "loss": 0.7768, + "num_input_tokens_seen": 62797755, + "step": 2962 + }, + { + "epoch": 0.35627968496362655, + "flos": 50329707089280.0, + "grad_norm": 0.7798080870923961, + "learning_rate": 2.98353984329501e-06, + "loss": 0.6146, + "num_input_tokens_seen": 62861820, + "step": 2963 + }, + { + "epoch": 0.3563999278542656, + "flos": 16570064437800.0, + "grad_norm": 4.224434195108125, + "learning_rate": 2.982861498480965e-06, + "loss": 0.6943, + "num_input_tokens_seen": 62883920, + "step": 2964 + }, + { + "epoch": 0.3565201707449047, + "flos": 19018965209520.0, + "grad_norm": 2.399184318849427, + "learning_rate": 2.9821830045676122e-06, + "loss": 0.8047, + "num_input_tokens_seen": 62903340, + "step": 2965 + }, + { + "epoch": 0.3566404136355438, + "flos": 20887099505040.0, + "grad_norm": 2.6117790249445645, + "learning_rate": 2.9815043616578793e-06, + "loss": 0.6963, + "num_input_tokens_seen": 62923855, + "step": 2966 + }, + { + "epoch": 0.3567606565261829, + "flos": 28204398700080.0, + "grad_norm": 4.170367593640742, + "learning_rate": 2.9808255698547145e-06, + "loss": 0.7619, + "num_input_tokens_seen": 62946375, + "step": 2967 + }, + { + "epoch": 0.356880899416822, + "flos": 16080555660840.0, + "grad_norm": 4.509293813628896, + "learning_rate": 2.9801466292610913e-06, + "loss": 0.767, + "num_input_tokens_seen": 62965980, + "step": 2968 + }, + { + "epoch": 0.35700114230746105, + "flos": 13866341410320.0, + "grad_norm": 3.443221168968473, + "learning_rate": 2.979467539980003e-06, + "loss": 0.7868, + "num_input_tokens_seen": 62982490, + "step": 2969 + }, + { + "epoch": 0.35712138519810016, + "flos": 14434068474360.0, + "grad_norm": 3.119392704914395, + "learning_rate": 2.978788302114468e-06, + "loss": 0.7485, + "num_input_tokens_seen": 62999325, + "step": 2970 + }, + { + "epoch": 0.35724162808873927, + "flos": 25850079923400.0, + "grad_norm": 3.679679724812645, + "learning_rate": 2.9781089157675255e-06, + "loss": 0.7996, + "num_input_tokens_seen": 63017505, + "step": 2971 + }, + { + "epoch": 0.3573618709793783, + "flos": 18726824843640.0, + "grad_norm": 1.8392779539264659, + "learning_rate": 2.977429381042238e-06, + "loss": 0.8681, + "num_input_tokens_seen": 63037900, + "step": 2972 + }, + { + "epoch": 0.35748211387001744, + "flos": 21373030706520.0, + "grad_norm": 3.7634263327240536, + "learning_rate": 2.9767496980416913e-06, + "loss": 0.8713, + "num_input_tokens_seen": 63056915, + "step": 2973 + }, + { + "epoch": 0.35760235676065655, + "flos": 10140140686560.0, + "grad_norm": 4.347785666733978, + "learning_rate": 2.9760698668689914e-06, + "loss": 0.7912, + "num_input_tokens_seen": 63072860, + "step": 2974 + }, + { + "epoch": 0.3577225996512956, + "flos": 32389940750280.0, + "grad_norm": 2.6026352512126034, + "learning_rate": 2.975389887627269e-06, + "loss": 0.699, + "num_input_tokens_seen": 63095180, + "step": 2975 + }, + { + "epoch": 0.3578428425419347, + "flos": 12435074361000.0, + "grad_norm": 4.508867216434684, + "learning_rate": 2.9747097604196764e-06, + "loss": 0.8834, + "num_input_tokens_seen": 63111545, + "step": 2976 + }, + { + "epoch": 0.3579630854325738, + "flos": 52864012959000.0, + "grad_norm": 0.6908763859505277, + "learning_rate": 2.9740294853493875e-06, + "loss": 0.59, + "num_input_tokens_seen": 63182825, + "step": 2977 + }, + { + "epoch": 0.3580833283232129, + "flos": 18349437679560.0, + "grad_norm": 3.4887413413508583, + "learning_rate": 2.9733490625196008e-06, + "loss": 0.6608, + "num_input_tokens_seen": 63202405, + "step": 2978 + }, + { + "epoch": 0.358203571213852, + "flos": 10139222547720.0, + "grad_norm": 5.676215746552981, + "learning_rate": 2.9726684920335353e-06, + "loss": 0.7124, + "num_input_tokens_seen": 63219990, + "step": 2979 + }, + { + "epoch": 0.35832381410449105, + "flos": 14838267133200.0, + "grad_norm": 3.1038307332459496, + "learning_rate": 2.971987773994432e-06, + "loss": 0.8048, + "num_input_tokens_seen": 63235895, + "step": 2980 + }, + { + "epoch": 0.35844405699513016, + "flos": 12381957930840.0, + "grad_norm": 2.5482444591838096, + "learning_rate": 2.9713069085055566e-06, + "loss": 0.8046, + "num_input_tokens_seen": 63253925, + "step": 2981 + }, + { + "epoch": 0.35856429988576927, + "flos": 16994489319720.0, + "grad_norm": 3.011369618169428, + "learning_rate": 2.9706258956701958e-06, + "loss": 0.7756, + "num_input_tokens_seen": 63273635, + "step": 2982 + }, + { + "epoch": 0.3586845427764083, + "flos": 16859608686960.0, + "grad_norm": 3.378029911421522, + "learning_rate": 2.9699447355916575e-06, + "loss": 0.7479, + "num_input_tokens_seen": 63292165, + "step": 2983 + }, + { + "epoch": 0.35880478566704743, + "flos": 14622350476920.0, + "grad_norm": 3.723815801422622, + "learning_rate": 2.969263428373275e-06, + "loss": 0.7198, + "num_input_tokens_seen": 63310235, + "step": 2984 + }, + { + "epoch": 0.35892502855768654, + "flos": 9951415444560.0, + "grad_norm": 4.064055863799969, + "learning_rate": 2.9685819741184007e-06, + "loss": 0.7605, + "num_input_tokens_seen": 63328395, + "step": 2985 + }, + { + "epoch": 0.3590452714483256, + "flos": 13218844721160.0, + "grad_norm": 4.145769486183278, + "learning_rate": 2.967900372930411e-06, + "loss": 0.6613, + "num_input_tokens_seen": 63346625, + "step": 2986 + }, + { + "epoch": 0.3591655143389647, + "flos": 12948925155840.0, + "grad_norm": 4.132365500031972, + "learning_rate": 2.9672186249127046e-06, + "loss": 0.7687, + "num_input_tokens_seen": 63365810, + "step": 2987 + }, + { + "epoch": 0.3592857572296038, + "flos": 18480329157360.0, + "grad_norm": 4.14353708033762, + "learning_rate": 2.9665367301687014e-06, + "loss": 0.7724, + "num_input_tokens_seen": 63383775, + "step": 2988 + }, + { + "epoch": 0.3594060001202429, + "flos": 21559096511880.0, + "grad_norm": 2.801110156550588, + "learning_rate": 2.965854688801845e-06, + "loss": 0.7441, + "num_input_tokens_seen": 63405555, + "step": 2989 + }, + { + "epoch": 0.359526243010882, + "flos": 12433048123560.0, + "grad_norm": 2.7066621955433767, + "learning_rate": 2.9651725009156005e-06, + "loss": 0.7463, + "num_input_tokens_seen": 63423020, + "step": 2990 + }, + { + "epoch": 0.3596464859015211, + "flos": 16808676794040.0, + "grad_norm": 2.5577559868611344, + "learning_rate": 2.964490166613454e-06, + "loss": 0.7278, + "num_input_tokens_seen": 63442665, + "step": 2991 + }, + { + "epoch": 0.35976672879216015, + "flos": 40328341210560.0, + "grad_norm": 0.8320776339930948, + "learning_rate": 2.963807685998917e-06, + "loss": 0.5934, + "num_input_tokens_seen": 63498250, + "step": 2992 + }, + { + "epoch": 0.35988697168279926, + "flos": 31738359926280.0, + "grad_norm": 2.102688512526972, + "learning_rate": 2.9631250591755196e-06, + "loss": 0.7694, + "num_input_tokens_seen": 63520685, + "step": 2993 + }, + { + "epoch": 0.36000721457343837, + "flos": 26341266678240.0, + "grad_norm": 4.69156460331988, + "learning_rate": 2.962442286246817e-06, + "loss": 0.5698, + "num_input_tokens_seen": 63543235, + "step": 2994 + }, + { + "epoch": 0.3601274574640774, + "flos": 13349071339800.0, + "grad_norm": 2.118637545591757, + "learning_rate": 2.9617593673163853e-06, + "loss": 0.6859, + "num_input_tokens_seen": 63561775, + "step": 2995 + }, + { + "epoch": 0.36024770035471654, + "flos": 9681147619680.0, + "grad_norm": 2.566262994539163, + "learning_rate": 2.9610763024878216e-06, + "loss": 0.7519, + "num_input_tokens_seen": 63577000, + "step": 2996 + }, + { + "epoch": 0.3603679432453556, + "flos": 14812183817520.0, + "grad_norm": 2.482273580592246, + "learning_rate": 2.960393091864747e-06, + "loss": 0.8975, + "num_input_tokens_seen": 63595100, + "step": 2997 + }, + { + "epoch": 0.3604881861359947, + "flos": 16427870354280.0, + "grad_norm": 3.1740235817218907, + "learning_rate": 2.959709735550804e-06, + "loss": 0.726, + "num_input_tokens_seen": 63614415, + "step": 2998 + }, + { + "epoch": 0.3606084290266338, + "flos": 16134780189600.0, + "grad_norm": 3.6141151909258857, + "learning_rate": 2.9590262336496575e-06, + "loss": 0.7469, + "num_input_tokens_seen": 63633865, + "step": 2999 + }, + { + "epoch": 0.36072867191727287, + "flos": 11382587514000.0, + "grad_norm": 3.4454100718273457, + "learning_rate": 2.9583425862649936e-06, + "loss": 0.8362, + "num_input_tokens_seen": 63651720, + "step": 3000 + }, + { + "epoch": 0.360848914807912, + "flos": 14374651712160.0, + "grad_norm": 3.3630731143858705, + "learning_rate": 2.9576587935005215e-06, + "loss": 0.7285, + "num_input_tokens_seen": 63669520, + "step": 3001 + }, + { + "epoch": 0.3609691576985511, + "flos": 13783690728840.0, + "grad_norm": 4.439591465618407, + "learning_rate": 2.9569748554599713e-06, + "loss": 0.6974, + "num_input_tokens_seen": 63684850, + "step": 3002 + }, + { + "epoch": 0.36108940058919015, + "flos": 31060854086400.0, + "grad_norm": 4.459332339344056, + "learning_rate": 2.956290772247097e-06, + "loss": 0.7159, + "num_input_tokens_seen": 63703245, + "step": 3003 + }, + { + "epoch": 0.36120964347982926, + "flos": 16971698639880.0, + "grad_norm": 3.0487727588462556, + "learning_rate": 2.9556065439656724e-06, + "loss": 0.725, + "num_input_tokens_seen": 63722015, + "step": 3004 + }, + { + "epoch": 0.36132988637046837, + "flos": 13217926582320.0, + "grad_norm": 2.4751722603393227, + "learning_rate": 2.9549221707194952e-06, + "loss": 0.8059, + "num_input_tokens_seen": 63740585, + "step": 3005 + }, + { + "epoch": 0.3614501292611074, + "flos": 20395881090240.0, + "grad_norm": 4.565509796255729, + "learning_rate": 2.954237652612384e-06, + "loss": 0.7257, + "num_input_tokens_seen": 63759355, + "step": 3006 + }, + { + "epoch": 0.36157037215174653, + "flos": 16564080705360.0, + "grad_norm": 3.3583366741103853, + "learning_rate": 2.9535529897481796e-06, + "loss": 0.8348, + "num_input_tokens_seen": 63776620, + "step": 3007 + }, + { + "epoch": 0.36169061504238564, + "flos": 9138110833080.0, + "grad_norm": 4.193461714954026, + "learning_rate": 2.9528681822307446e-06, + "loss": 0.7485, + "num_input_tokens_seen": 63793190, + "step": 3008 + }, + { + "epoch": 0.3618108579330247, + "flos": 19561115517240.0, + "grad_norm": 7.521110325271286, + "learning_rate": 2.952183230163964e-06, + "loss": 0.8067, + "num_input_tokens_seen": 63812485, + "step": 3009 + }, + { + "epoch": 0.3619311008236638, + "flos": 16698803038320.0, + "grad_norm": 2.732146221269234, + "learning_rate": 2.9514981336517448e-06, + "loss": 0.7219, + "num_input_tokens_seen": 63831975, + "step": 3010 + }, + { + "epoch": 0.36205134371430286, + "flos": 18995098091040.0, + "grad_norm": 2.589199624720982, + "learning_rate": 2.950812892798015e-06, + "loss": 0.793, + "num_input_tokens_seen": 63852590, + "step": 3011 + }, + { + "epoch": 0.362171586604942, + "flos": 19158056616960.0, + "grad_norm": 5.010150333973877, + "learning_rate": 2.9501275077067256e-06, + "loss": 0.8574, + "num_input_tokens_seen": 63872930, + "step": 3012 + }, + { + "epoch": 0.3622918294955811, + "flos": 20589925205520.0, + "grad_norm": 2.3018465797809733, + "learning_rate": 2.949441978481848e-06, + "loss": 0.8721, + "num_input_tokens_seen": 63893550, + "step": 3013 + }, + { + "epoch": 0.36241207238622014, + "flos": 14487121584600.0, + "grad_norm": 3.552404588656989, + "learning_rate": 2.9487563052273778e-06, + "loss": 0.7862, + "num_input_tokens_seen": 63910030, + "step": 3014 + }, + { + "epoch": 0.36253231527685925, + "flos": 15648437408640.0, + "grad_norm": 3.385160340606662, + "learning_rate": 2.94807048804733e-06, + "loss": 0.8509, + "num_input_tokens_seen": 63929370, + "step": 3015 + }, + { + "epoch": 0.36265255816749836, + "flos": 13402029470160.0, + "grad_norm": 2.643526368118076, + "learning_rate": 2.9473845270457434e-06, + "loss": 0.8823, + "num_input_tokens_seen": 63945905, + "step": 3016 + }, + { + "epoch": 0.3627728010581374, + "flos": 13622695120440.0, + "grad_norm": 2.743389333599146, + "learning_rate": 2.946698422326677e-06, + "loss": 0.6652, + "num_input_tokens_seen": 63963085, + "step": 3017 + }, + { + "epoch": 0.36289304394877653, + "flos": 20182560550680.0, + "grad_norm": 3.3398628814327784, + "learning_rate": 2.946012173994213e-06, + "loss": 0.7836, + "num_input_tokens_seen": 63982590, + "step": 3018 + }, + { + "epoch": 0.36301328683941564, + "flos": 25370195774280.0, + "grad_norm": 2.0118354837566597, + "learning_rate": 2.945325782152454e-06, + "loss": 0.6712, + "num_input_tokens_seen": 64005345, + "step": 3019 + }, + { + "epoch": 0.3631335297300547, + "flos": 13893659464440.0, + "grad_norm": 2.7655105998198715, + "learning_rate": 2.9446392469055257e-06, + "loss": 0.7798, + "num_input_tokens_seen": 64023100, + "step": 3020 + }, + { + "epoch": 0.3632537726206938, + "flos": 14271679827720.0, + "grad_norm": 2.3913454453702454, + "learning_rate": 2.9439525683575745e-06, + "loss": 0.7827, + "num_input_tokens_seen": 64041740, + "step": 3021 + }, + { + "epoch": 0.3633740155113329, + "flos": 15405266018160.0, + "grad_norm": 3.167321906980559, + "learning_rate": 2.9432657466127694e-06, + "loss": 0.7281, + "num_input_tokens_seen": 64061030, + "step": 3022 + }, + { + "epoch": 0.36349425840197197, + "flos": 14834974497360.0, + "grad_norm": 2.3329700005438347, + "learning_rate": 2.9425787817753007e-06, + "loss": 0.7575, + "num_input_tokens_seen": 64079410, + "step": 3023 + }, + { + "epoch": 0.3636145012926111, + "flos": 21805623858120.0, + "grad_norm": 2.4330920931582227, + "learning_rate": 2.94189167394938e-06, + "loss": 0.689, + "num_input_tokens_seen": 64101565, + "step": 3024 + }, + { + "epoch": 0.3637347441832502, + "flos": 15673855865160.0, + "grad_norm": 3.1136592314854323, + "learning_rate": 2.941204423239241e-06, + "loss": 0.7978, + "num_input_tokens_seen": 64120160, + "step": 3025 + }, + { + "epoch": 0.36385498707388925, + "flos": 21670268325960.0, + "grad_norm": 2.3587124127828343, + "learning_rate": 2.9405170297491395e-06, + "loss": 0.7491, + "num_input_tokens_seen": 64139875, + "step": 3026 + }, + { + "epoch": 0.36397522996452836, + "flos": 16269692482320.0, + "grad_norm": 2.877137757674412, + "learning_rate": 2.939829493583353e-06, + "loss": 0.7876, + "num_input_tokens_seen": 64156240, + "step": 3027 + }, + { + "epoch": 0.3640954728551674, + "flos": 15729346792320.0, + "grad_norm": 2.7765887343455864, + "learning_rate": 2.939141814846179e-06, + "loss": 0.8019, + "num_input_tokens_seen": 64173375, + "step": 3028 + }, + { + "epoch": 0.3642157157458065, + "flos": 12921923701320.0, + "grad_norm": 2.1845264387776933, + "learning_rate": 2.938453993641938e-06, + "loss": 0.812, + "num_input_tokens_seen": 64191470, + "step": 3029 + }, + { + "epoch": 0.36433595863644563, + "flos": 12867604192680.0, + "grad_norm": 5.852130286830854, + "learning_rate": 2.937766030074973e-06, + "loss": 0.6759, + "num_input_tokens_seen": 64208445, + "step": 3030 + }, + { + "epoch": 0.3644562015270847, + "flos": 19671495832320.0, + "grad_norm": 2.3098317810834357, + "learning_rate": 2.937077924249646e-06, + "loss": 0.7994, + "num_input_tokens_seen": 64230755, + "step": 3031 + }, + { + "epoch": 0.3645764444177238, + "flos": 10383312077040.0, + "grad_norm": 4.2585773306043535, + "learning_rate": 2.9363896762703443e-06, + "loss": 0.7328, + "num_input_tokens_seen": 64247540, + "step": 3032 + }, + { + "epoch": 0.3646966873083629, + "flos": 15108439978200.0, + "grad_norm": 1.910462586971771, + "learning_rate": 2.9357012862414725e-06, + "loss": 0.8288, + "num_input_tokens_seen": 64266620, + "step": 3033 + }, + { + "epoch": 0.36481693019900197, + "flos": 20374166849040.0, + "grad_norm": 4.945584209866402, + "learning_rate": 2.9350127542674593e-06, + "loss": 0.7038, + "num_input_tokens_seen": 64288550, + "step": 3034 + }, + { + "epoch": 0.3649371730896411, + "flos": 14400861667680.0, + "grad_norm": 2.773029909943131, + "learning_rate": 2.934324080452755e-06, + "loss": 0.7517, + "num_input_tokens_seen": 64306060, + "step": 3035 + }, + { + "epoch": 0.3650574159802802, + "flos": 18129658508160.0, + "grad_norm": 2.276138850624403, + "learning_rate": 2.9336352649018307e-06, + "loss": 0.7675, + "num_input_tokens_seen": 64325850, + "step": 3036 + }, + { + "epoch": 0.36517765887091924, + "flos": 24126862467960.0, + "grad_norm": 3.9907769923289926, + "learning_rate": 2.9329463077191783e-06, + "loss": 0.6868, + "num_input_tokens_seen": 64348945, + "step": 3037 + }, + { + "epoch": 0.36529790176155835, + "flos": 14703513140280.0, + "grad_norm": 5.938310748512445, + "learning_rate": 2.9322572090093135e-06, + "loss": 0.6267, + "num_input_tokens_seen": 64367370, + "step": 3038 + }, + { + "epoch": 0.36541814465219746, + "flos": 12867635852640.0, + "grad_norm": 4.22497754514246, + "learning_rate": 2.9315679688767713e-06, + "loss": 0.7264, + "num_input_tokens_seen": 64385100, + "step": 3039 + }, + { + "epoch": 0.3655383875428365, + "flos": 16593678276600.0, + "grad_norm": 2.5887575014780237, + "learning_rate": 2.9308785874261085e-06, + "loss": 0.6508, + "num_input_tokens_seen": 64405010, + "step": 3040 + }, + { + "epoch": 0.36565863043347563, + "flos": 16080682300680.0, + "grad_norm": 4.614077894144444, + "learning_rate": 2.9301890647619045e-06, + "loss": 0.791, + "num_input_tokens_seen": 64424025, + "step": 3041 + }, + { + "epoch": 0.36577887332411474, + "flos": 18187333972560.0, + "grad_norm": 2.750897261153997, + "learning_rate": 2.929499400988759e-06, + "loss": 0.77, + "num_input_tokens_seen": 64444905, + "step": 3042 + }, + { + "epoch": 0.3658991162147538, + "flos": 20751997252560.0, + "grad_norm": 2.4178808932758873, + "learning_rate": 2.9288095962112927e-06, + "loss": 0.6265, + "num_input_tokens_seen": 64465330, + "step": 3043 + }, + { + "epoch": 0.3660193591053929, + "flos": 12975673330680.0, + "grad_norm": 2.7335507267665005, + "learning_rate": 2.9281196505341503e-06, + "loss": 0.8281, + "num_input_tokens_seen": 64482220, + "step": 3044 + }, + { + "epoch": 0.36613960199603196, + "flos": 7467566568360.0, + "grad_norm": 3.1503208481403964, + "learning_rate": 2.9274295640619946e-06, + "loss": 0.7767, + "num_input_tokens_seen": 64499070, + "step": 3045 + }, + { + "epoch": 0.36625984488667107, + "flos": 14433625234920.0, + "grad_norm": 2.686070705304092, + "learning_rate": 2.9267393368995103e-06, + "loss": 0.7733, + "num_input_tokens_seen": 64518020, + "step": 3046 + }, + { + "epoch": 0.3663800877773102, + "flos": 12893054309160.0, + "grad_norm": 4.195565043827506, + "learning_rate": 2.926048969151407e-06, + "loss": 0.7203, + "num_input_tokens_seen": 64535025, + "step": 3047 + }, + { + "epoch": 0.36650033066794924, + "flos": 14839280251920.0, + "grad_norm": 3.0458212570876437, + "learning_rate": 2.92535846092241e-06, + "loss": 0.6614, + "num_input_tokens_seen": 64553760, + "step": 3048 + }, + { + "epoch": 0.36662057355858835, + "flos": 18104619971160.0, + "grad_norm": 3.770062250083788, + "learning_rate": 2.9246678123172704e-06, + "loss": 0.8061, + "num_input_tokens_seen": 64573570, + "step": 3049 + }, + { + "epoch": 0.36674081644922746, + "flos": 8979046482240.0, + "grad_norm": 4.52947058104, + "learning_rate": 2.9239770234407596e-06, + "loss": 0.7059, + "num_input_tokens_seen": 64591595, + "step": 3050 + }, + { + "epoch": 0.3668610593398665, + "flos": 15432995651760.0, + "grad_norm": 2.3020135400010986, + "learning_rate": 2.9232860943976686e-06, + "loss": 0.6646, + "num_input_tokens_seen": 64612050, + "step": 3051 + }, + { + "epoch": 0.3669813022305056, + "flos": 19615530005760.0, + "grad_norm": 2.3912510270367267, + "learning_rate": 2.9225950252928115e-06, + "loss": 0.8237, + "num_input_tokens_seen": 64632620, + "step": 3052 + }, + { + "epoch": 0.36710154512114473, + "flos": 13948168932840.0, + "grad_norm": 3.4706990033212723, + "learning_rate": 2.9219038162310217e-06, + "loss": 0.7863, + "num_input_tokens_seen": 64650540, + "step": 3053 + }, + { + "epoch": 0.3672217880117838, + "flos": 15215211057840.0, + "grad_norm": 3.3376291885804013, + "learning_rate": 2.921212467317157e-06, + "loss": 0.8072, + "num_input_tokens_seen": 64669705, + "step": 3054 + }, + { + "epoch": 0.3673420309024229, + "flos": 9871075940160.0, + "grad_norm": 4.190551556904439, + "learning_rate": 2.920520978656093e-06, + "loss": 0.7882, + "num_input_tokens_seen": 64686390, + "step": 3055 + }, + { + "epoch": 0.367462273793062, + "flos": 21265436467920.0, + "grad_norm": 2.796825216692064, + "learning_rate": 2.919829350352729e-06, + "loss": 0.7486, + "num_input_tokens_seen": 64707715, + "step": 3056 + }, + { + "epoch": 0.36758251668370107, + "flos": 46175657716560.0, + "grad_norm": 0.7543935041204577, + "learning_rate": 2.919137582511983e-06, + "loss": 0.6105, + "num_input_tokens_seen": 64763875, + "step": 3057 + }, + { + "epoch": 0.3677027595743402, + "flos": 9222281192640.0, + "grad_norm": 2.6499534027033427, + "learning_rate": 2.918445675238797e-06, + "loss": 0.6161, + "num_input_tokens_seen": 64780520, + "step": 3058 + }, + { + "epoch": 0.36782300246497923, + "flos": 18349374359640.0, + "grad_norm": 3.588136617817465, + "learning_rate": 2.917753628638132e-06, + "loss": 0.6676, + "num_input_tokens_seen": 64800545, + "step": 3059 + }, + { + "epoch": 0.36794324535561834, + "flos": 12704614006800.0, + "grad_norm": 6.256401445840386, + "learning_rate": 2.9170614428149716e-06, + "loss": 0.6781, + "num_input_tokens_seen": 64818600, + "step": 3060 + }, + { + "epoch": 0.36806348824625745, + "flos": 17639548191960.0, + "grad_norm": 4.161948062629493, + "learning_rate": 2.9163691178743195e-06, + "loss": 0.847, + "num_input_tokens_seen": 64836970, + "step": 3061 + }, + { + "epoch": 0.3681837311368965, + "flos": 15000719099760.0, + "grad_norm": 2.9516862866528437, + "learning_rate": 2.9156766539212006e-06, + "loss": 0.7684, + "num_input_tokens_seen": 64854335, + "step": 3062 + }, + { + "epoch": 0.3683039740275356, + "flos": 15699369301560.0, + "grad_norm": 2.9353499965167327, + "learning_rate": 2.9149840510606614e-06, + "loss": 0.6946, + "num_input_tokens_seen": 64872710, + "step": 3063 + }, + { + "epoch": 0.36842421691817473, + "flos": 51902593851480.0, + "grad_norm": 1.0765407919944339, + "learning_rate": 2.914291309397769e-06, + "loss": 0.6768, + "num_input_tokens_seen": 64929900, + "step": 3064 + }, + { + "epoch": 0.3685444598088138, + "flos": 17079957737640.0, + "grad_norm": 3.200045859485082, + "learning_rate": 2.9135984290376117e-06, + "loss": 0.7696, + "num_input_tokens_seen": 64948485, + "step": 3065 + }, + { + "epoch": 0.3686647026994529, + "flos": 16885913622360.0, + "grad_norm": 2.2700036192541635, + "learning_rate": 2.9129054100853e-06, + "loss": 0.8176, + "num_input_tokens_seen": 64967045, + "step": 3066 + }, + { + "epoch": 0.368784945590092, + "flos": 18403693868280.0, + "grad_norm": 3.0025317317610374, + "learning_rate": 2.912212252645963e-06, + "loss": 0.7392, + "num_input_tokens_seen": 64989350, + "step": 3067 + }, + { + "epoch": 0.36890518848073106, + "flos": 13461699512040.0, + "grad_norm": 3.415752490178074, + "learning_rate": 2.9115189568247523e-06, + "loss": 0.7415, + "num_input_tokens_seen": 65006630, + "step": 3068 + }, + { + "epoch": 0.36902543137137017, + "flos": 11922458304600.0, + "grad_norm": 2.5968254883380832, + "learning_rate": 2.910825522726841e-06, + "loss": 0.8929, + "num_input_tokens_seen": 65023875, + "step": 3069 + }, + { + "epoch": 0.3691456742620093, + "flos": 8897852158920.0, + "grad_norm": 3.6537709757372108, + "learning_rate": 2.9101319504574215e-06, + "loss": 0.7597, + "num_input_tokens_seen": 65040035, + "step": 3070 + }, + { + "epoch": 0.36926591715264834, + "flos": 12976433169720.0, + "grad_norm": 2.365399078506235, + "learning_rate": 2.909438240121709e-06, + "loss": 0.746, + "num_input_tokens_seen": 65060030, + "step": 3071 + }, + { + "epoch": 0.36938616004328745, + "flos": 21236598735720.0, + "grad_norm": 2.9142519810003815, + "learning_rate": 2.908744391824939e-06, + "loss": 0.6887, + "num_input_tokens_seen": 65080770, + "step": 3072 + }, + { + "epoch": 0.36950640293392656, + "flos": 21427096935480.0, + "grad_norm": 2.444795499023069, + "learning_rate": 2.908050405672367e-06, + "loss": 0.7645, + "num_input_tokens_seen": 65100035, + "step": 3073 + }, + { + "epoch": 0.3696266458245656, + "flos": 18187460612400.0, + "grad_norm": 2.698858698060113, + "learning_rate": 2.9073562817692703e-06, + "loss": 0.7781, + "num_input_tokens_seen": 65118440, + "step": 3074 + }, + { + "epoch": 0.3697468887152047, + "flos": 44138522334120.0, + "grad_norm": 0.7896232123524545, + "learning_rate": 2.9066620202209468e-06, + "loss": 0.6029, + "num_input_tokens_seen": 65180650, + "step": 3075 + }, + { + "epoch": 0.3698671316058438, + "flos": 19156283659200.0, + "grad_norm": 2.902085304305226, + "learning_rate": 2.905967621132716e-06, + "loss": 0.7731, + "num_input_tokens_seen": 65197980, + "step": 3076 + }, + { + "epoch": 0.3699873744964829, + "flos": 18024217146840.0, + "grad_norm": 2.55775533436329, + "learning_rate": 2.9052730846099172e-06, + "loss": 0.7287, + "num_input_tokens_seen": 65219045, + "step": 3077 + }, + { + "epoch": 0.370107617387122, + "flos": 47837373343800.0, + "grad_norm": 1.6553963674623648, + "learning_rate": 2.9045784107579123e-06, + "loss": 0.6412, + "num_input_tokens_seen": 65278870, + "step": 3078 + }, + { + "epoch": 0.37022786027776106, + "flos": 11625948864240.0, + "grad_norm": 3.0205894284055086, + "learning_rate": 2.9038835996820807e-06, + "loss": 0.6611, + "num_input_tokens_seen": 65296200, + "step": 3079 + }, + { + "epoch": 0.37034810316840017, + "flos": 13538461440960.0, + "grad_norm": 2.3876162276013235, + "learning_rate": 2.903188651487826e-06, + "loss": 0.7752, + "num_input_tokens_seen": 65314475, + "step": 3080 + }, + { + "epoch": 0.3704683460590393, + "flos": 13002579805320.0, + "grad_norm": 3.7068416438437426, + "learning_rate": 2.902493566280571e-06, + "loss": 0.851, + "num_input_tokens_seen": 65332300, + "step": 3081 + }, + { + "epoch": 0.37058858894967833, + "flos": 10272931761960.0, + "grad_norm": 2.3878463853344574, + "learning_rate": 2.9017983441657595e-06, + "loss": 0.7962, + "num_input_tokens_seen": 65349350, + "step": 3082 + }, + { + "epoch": 0.37070883184031744, + "flos": 10140425626200.0, + "grad_norm": 2.906058520500362, + "learning_rate": 2.9011029852488564e-06, + "loss": 0.7465, + "num_input_tokens_seen": 65366305, + "step": 3083 + }, + { + "epoch": 0.37082907473095655, + "flos": 38532124870080.0, + "grad_norm": 1.0004766639492981, + "learning_rate": 2.9004074896353465e-06, + "loss": 0.67, + "num_input_tokens_seen": 65420025, + "step": 3084 + }, + { + "epoch": 0.3709493176215956, + "flos": 11652886998840.0, + "grad_norm": 2.303883661293289, + "learning_rate": 2.8997118574307362e-06, + "loss": 0.7962, + "num_input_tokens_seen": 65436700, + "step": 3085 + }, + { + "epoch": 0.3710695605122347, + "flos": 15324863193840.0, + "grad_norm": 2.734391915405268, + "learning_rate": 2.899016088740553e-06, + "loss": 0.7307, + "num_input_tokens_seen": 65454530, + "step": 3086 + }, + { + "epoch": 0.37118980340287383, + "flos": 10437220006200.0, + "grad_norm": 2.723192816120359, + "learning_rate": 2.898320183670344e-06, + "loss": 0.7721, + "num_input_tokens_seen": 65471665, + "step": 3087 + }, + { + "epoch": 0.3713100462935129, + "flos": 18970091214000.0, + "grad_norm": 2.446977042729262, + "learning_rate": 2.8976241423256767e-06, + "loss": 0.8701, + "num_input_tokens_seen": 65491480, + "step": 3088 + }, + { + "epoch": 0.371430289184152, + "flos": 22399212618120.0, + "grad_norm": 2.4077710385128746, + "learning_rate": 2.896927964812142e-06, + "loss": 0.6663, + "num_input_tokens_seen": 65511765, + "step": 3089 + }, + { + "epoch": 0.37155053207479105, + "flos": 11463465237720.0, + "grad_norm": 4.7190972004665435, + "learning_rate": 2.8962316512353465e-06, + "loss": 0.7163, + "num_input_tokens_seen": 65529030, + "step": 3090 + }, + { + "epoch": 0.37167077496543016, + "flos": 17133707367000.0, + "grad_norm": 3.31704544000377, + "learning_rate": 2.8955352017009233e-06, + "loss": 0.7332, + "num_input_tokens_seen": 65547995, + "step": 3091 + }, + { + "epoch": 0.3717910178560693, + "flos": 16159945366440.0, + "grad_norm": 2.6379912294093075, + "learning_rate": 2.8948386163145212e-06, + "loss": 0.7533, + "num_input_tokens_seen": 65566925, + "step": 3092 + }, + { + "epoch": 0.3719112607467083, + "flos": 19750062378960.0, + "grad_norm": 2.624933332777402, + "learning_rate": 2.8941418951818135e-06, + "loss": 0.7732, + "num_input_tokens_seen": 65586205, + "step": 3093 + }, + { + "epoch": 0.37203150363734744, + "flos": 8817164394960.0, + "grad_norm": 3.2835741803541127, + "learning_rate": 2.8934450384084903e-06, + "loss": 0.6999, + "num_input_tokens_seen": 65603440, + "step": 3094 + }, + { + "epoch": 0.37215174652798655, + "flos": 17351175361320.0, + "grad_norm": 5.735336846700217, + "learning_rate": 2.8927480461002653e-06, + "loss": 0.6911, + "num_input_tokens_seen": 65623130, + "step": 3095 + }, + { + "epoch": 0.3722719894186256, + "flos": 13050282382320.0, + "grad_norm": 38.34063315062266, + "learning_rate": 2.892050918362872e-06, + "loss": 0.8205, + "num_input_tokens_seen": 65637905, + "step": 3096 + }, + { + "epoch": 0.3723922323092647, + "flos": 46009818134280.0, + "grad_norm": 0.9000697691683281, + "learning_rate": 2.8913536553020626e-06, + "loss": 0.6031, + "num_input_tokens_seen": 65691680, + "step": 3097 + }, + { + "epoch": 0.3725124751999038, + "flos": 16864167721200.0, + "grad_norm": 2.036775907322445, + "learning_rate": 2.8906562570236137e-06, + "loss": 0.8394, + "num_input_tokens_seen": 65709310, + "step": 3098 + }, + { + "epoch": 0.3726327180905429, + "flos": 15295392262440.0, + "grad_norm": 1.9453278812888084, + "learning_rate": 2.889958723633318e-06, + "loss": 0.7514, + "num_input_tokens_seen": 65727970, + "step": 3099 + }, + { + "epoch": 0.372752960981182, + "flos": 22453437146880.0, + "grad_norm": 2.7818827391538012, + "learning_rate": 2.889261055236992e-06, + "loss": 0.7227, + "num_input_tokens_seen": 65749905, + "step": 3100 + }, + { + "epoch": 0.3728732038718211, + "flos": 18400907791800.0, + "grad_norm": 2.2544813034084386, + "learning_rate": 2.8885632519404704e-06, + "loss": 0.8124, + "num_input_tokens_seen": 65769895, + "step": 3101 + }, + { + "epoch": 0.37299344676246016, + "flos": 18538447861200.0, + "grad_norm": 2.4894315495103023, + "learning_rate": 2.8878653138496107e-06, + "loss": 0.741, + "num_input_tokens_seen": 65790110, + "step": 3102 + }, + { + "epoch": 0.37311368965309927, + "flos": 17458294700520.0, + "grad_norm": 2.9345868863924647, + "learning_rate": 2.8871672410702878e-06, + "loss": 0.7363, + "num_input_tokens_seen": 65807190, + "step": 3103 + }, + { + "epoch": 0.3732339325437384, + "flos": 18915676725480.0, + "grad_norm": 2.1011630107567134, + "learning_rate": 2.8864690337084008e-06, + "loss": 0.8045, + "num_input_tokens_seen": 65826185, + "step": 3104 + }, + { + "epoch": 0.37335417543437743, + "flos": 19208418630600.0, + "grad_norm": 2.633313139906448, + "learning_rate": 2.885770691869866e-06, + "loss": 0.7668, + "num_input_tokens_seen": 65846785, + "step": 3105 + }, + { + "epoch": 0.37347441832501654, + "flos": 17592352174320.0, + "grad_norm": 2.608619169309599, + "learning_rate": 2.8850722156606207e-06, + "loss": 0.7184, + "num_input_tokens_seen": 65864895, + "step": 3106 + }, + { + "epoch": 0.3735946612156556, + "flos": 14402919565080.0, + "grad_norm": 2.4870288997024788, + "learning_rate": 2.8843736051866252e-06, + "loss": 0.646, + "num_input_tokens_seen": 65883540, + "step": 3107 + }, + { + "epoch": 0.3737149041062947, + "flos": 16864041081360.0, + "grad_norm": 2.0730318455989605, + "learning_rate": 2.8836748605538557e-06, + "loss": 0.6832, + "num_input_tokens_seen": 65904900, + "step": 3108 + }, + { + "epoch": 0.3738351469969338, + "flos": 25477251793560.0, + "grad_norm": 4.948438093584686, + "learning_rate": 2.882975981868313e-06, + "loss": 0.616, + "num_input_tokens_seen": 65925005, + "step": 3109 + }, + { + "epoch": 0.3739553898875729, + "flos": 32011983706920.0, + "grad_norm": 3.015873557802182, + "learning_rate": 2.882276969236016e-06, + "loss": 0.6548, + "num_input_tokens_seen": 65946085, + "step": 3110 + }, + { + "epoch": 0.374075632778212, + "flos": 9327342634440.0, + "grad_norm": 3.4288710054579643, + "learning_rate": 2.881577822763005e-06, + "loss": 0.731, + "num_input_tokens_seen": 65963755, + "step": 3111 + }, + { + "epoch": 0.3741958756688511, + "flos": 19071670060200.0, + "grad_norm": 2.532600325228723, + "learning_rate": 2.880878542555338e-06, + "loss": 0.8661, + "num_input_tokens_seen": 65981240, + "step": 3112 + }, + { + "epoch": 0.37431611855949015, + "flos": 15675660482880.0, + "grad_norm": 2.8794698383535455, + "learning_rate": 2.8801791287190976e-06, + "loss": 0.7907, + "num_input_tokens_seen": 65998955, + "step": 3113 + }, + { + "epoch": 0.37443636145012926, + "flos": 17728055966040.0, + "grad_norm": 3.7203019694339887, + "learning_rate": 2.8794795813603817e-06, + "loss": 0.8313, + "num_input_tokens_seen": 66014140, + "step": 3114 + }, + { + "epoch": 0.3745566043407684, + "flos": 11194527131160.0, + "grad_norm": 2.9165484994838176, + "learning_rate": 2.878779900585314e-06, + "loss": 0.7946, + "num_input_tokens_seen": 66031700, + "step": 3115 + }, + { + "epoch": 0.37467684723140743, + "flos": 18024882006000.0, + "grad_norm": 2.245123452953143, + "learning_rate": 2.8780800865000336e-06, + "loss": 0.7506, + "num_input_tokens_seen": 66052730, + "step": 3116 + }, + { + "epoch": 0.37479709012204654, + "flos": 47461949097240.0, + "grad_norm": 1.0338737172056662, + "learning_rate": 2.877380139210702e-06, + "loss": 0.6506, + "num_input_tokens_seen": 66111120, + "step": 3117 + }, + { + "epoch": 0.37491733301268565, + "flos": 17429456968320.0, + "grad_norm": 4.721917712298031, + "learning_rate": 2.876680058823501e-06, + "loss": 0.7315, + "num_input_tokens_seen": 66131240, + "step": 3118 + }, + { + "epoch": 0.3750375759033247, + "flos": 23614436371320.0, + "grad_norm": 3.299367820990633, + "learning_rate": 2.8759798454446314e-06, + "loss": 0.6515, + "num_input_tokens_seen": 66154125, + "step": 3119 + }, + { + "epoch": 0.3751578187939638, + "flos": 17106895872240.0, + "grad_norm": 2.3169990282842834, + "learning_rate": 2.8752794991803173e-06, + "loss": 0.7974, + "num_input_tokens_seen": 66171530, + "step": 3120 + }, + { + "epoch": 0.37527806168460287, + "flos": 10621607833680.0, + "grad_norm": 3.477307186307292, + "learning_rate": 2.8745790201367976e-06, + "loss": 0.7414, + "num_input_tokens_seen": 66187005, + "step": 3121 + }, + { + "epoch": 0.375398304575242, + "flos": 19343394243240.0, + "grad_norm": 3.0834258469457767, + "learning_rate": 2.8738784084203373e-06, + "loss": 0.8449, + "num_input_tokens_seen": 66206800, + "step": 3122 + }, + { + "epoch": 0.3755185474658811, + "flos": 16269724142280.0, + "grad_norm": 3.4557358205808364, + "learning_rate": 2.873177664137216e-06, + "loss": 0.7708, + "num_input_tokens_seen": 66227450, + "step": 3123 + }, + { + "epoch": 0.37563879035652015, + "flos": 22617092191920.0, + "grad_norm": 2.3780223287908857, + "learning_rate": 2.8724767873937384e-06, + "loss": 0.6718, + "num_input_tokens_seen": 66251290, + "step": 3124 + }, + { + "epoch": 0.37575903324715926, + "flos": 15186436645560.0, + "grad_norm": 3.7555201273594516, + "learning_rate": 2.871775778296225e-06, + "loss": 0.8545, + "num_input_tokens_seen": 66268100, + "step": 3125 + }, + { + "epoch": 0.37587927613779837, + "flos": 13650709693680.0, + "grad_norm": 5.6469077839240125, + "learning_rate": 2.8710746369510196e-06, + "loss": 0.7693, + "num_input_tokens_seen": 66285805, + "step": 3126 + }, + { + "epoch": 0.3759995190284374, + "flos": 9896241117000.0, + "grad_norm": 4.256865173785527, + "learning_rate": 2.8703733634644846e-06, + "loss": 0.8191, + "num_input_tokens_seen": 66300280, + "step": 3127 + }, + { + "epoch": 0.37611976191907653, + "flos": 14973242745840.0, + "grad_norm": 2.959312939421124, + "learning_rate": 2.869671957943002e-06, + "loss": 0.7862, + "num_input_tokens_seen": 66319155, + "step": 3128 + }, + { + "epoch": 0.37624000480971564, + "flos": 15459458886960.0, + "grad_norm": 3.356430537591135, + "learning_rate": 2.8689704204929747e-06, + "loss": 0.7341, + "num_input_tokens_seen": 66338055, + "step": 3129 + }, + { + "epoch": 0.3763602477003547, + "flos": 16512293993520.0, + "grad_norm": 2.810780317355243, + "learning_rate": 2.8682687512208253e-06, + "loss": 0.7918, + "num_input_tokens_seen": 66356785, + "step": 3130 + }, + { + "epoch": 0.3764804905909938, + "flos": 20184840067800.0, + "grad_norm": 3.2924058462481036, + "learning_rate": 2.8675669502329972e-06, + "loss": 0.7886, + "num_input_tokens_seen": 66378035, + "step": 3131 + }, + { + "epoch": 0.3766007334816329, + "flos": 16485387518880.0, + "grad_norm": 6.083261393639896, + "learning_rate": 2.866865017635952e-06, + "loss": 0.8225, + "num_input_tokens_seen": 66395575, + "step": 3132 + }, + { + "epoch": 0.376720976372272, + "flos": 19023049344360.0, + "grad_norm": 2.0064260245136634, + "learning_rate": 2.866162953536174e-06, + "loss": 0.7788, + "num_input_tokens_seen": 66416265, + "step": 3133 + }, + { + "epoch": 0.3768412192629111, + "flos": 13164873472080.0, + "grad_norm": 2.029048842104053, + "learning_rate": 2.8654607580401634e-06, + "loss": 0.736, + "num_input_tokens_seen": 66435720, + "step": 3134 + }, + { + "epoch": 0.3769614621535502, + "flos": 46434500787240.0, + "grad_norm": 0.9092229879948845, + "learning_rate": 2.8647584312544446e-06, + "loss": 0.6764, + "num_input_tokens_seen": 66500645, + "step": 3135 + }, + { + "epoch": 0.37708170504418925, + "flos": 17324680466160.0, + "grad_norm": 1.9345245278884615, + "learning_rate": 2.864055973285559e-06, + "loss": 0.8397, + "num_input_tokens_seen": 66522365, + "step": 3136 + }, + { + "epoch": 0.37720194793482836, + "flos": 17887563556320.0, + "grad_norm": 2.792094487293824, + "learning_rate": 2.8633533842400698e-06, + "loss": 0.8593, + "num_input_tokens_seen": 66542285, + "step": 3137 + }, + { + "epoch": 0.3773221908254674, + "flos": 15352022948160.0, + "grad_norm": 3.3948044573689327, + "learning_rate": 2.862650664224558e-06, + "loss": 0.7614, + "num_input_tokens_seen": 66560855, + "step": 3138 + }, + { + "epoch": 0.37744243371610653, + "flos": 27662881591560.0, + "grad_norm": 2.3301807518459148, + "learning_rate": 2.861947813345627e-06, + "loss": 0.6853, + "num_input_tokens_seen": 66583275, + "step": 3139 + }, + { + "epoch": 0.37756267660674564, + "flos": 19158879775920.0, + "grad_norm": 2.803106600935993, + "learning_rate": 2.8612448317098974e-06, + "loss": 0.6991, + "num_input_tokens_seen": 66603330, + "step": 3140 + }, + { + "epoch": 0.3776829194973847, + "flos": 14191182023520.0, + "grad_norm": 3.0897895817885406, + "learning_rate": 2.8605417194240114e-06, + "loss": 0.8245, + "num_input_tokens_seen": 66621410, + "step": 3141 + }, + { + "epoch": 0.3778031623880238, + "flos": 12676979353080.0, + "grad_norm": 3.6128718623654605, + "learning_rate": 2.8598384765946315e-06, + "loss": 0.7702, + "num_input_tokens_seen": 66639785, + "step": 3142 + }, + { + "epoch": 0.3779234052786629, + "flos": 19888109007720.0, + "grad_norm": 2.595534572482028, + "learning_rate": 2.8591351033284377e-06, + "loss": 0.6993, + "num_input_tokens_seen": 66659235, + "step": 3143 + }, + { + "epoch": 0.37804364816930197, + "flos": 14379622325880.0, + "grad_norm": 3.023689200399941, + "learning_rate": 2.8584315997321325e-06, + "loss": 0.8106, + "num_input_tokens_seen": 66677960, + "step": 3144 + }, + { + "epoch": 0.3781638910599411, + "flos": 16613682879960.0, + "grad_norm": 3.614404386255474, + "learning_rate": 2.8577279659124356e-06, + "loss": 0.7617, + "num_input_tokens_seen": 66695355, + "step": 3145 + }, + { + "epoch": 0.3782841339505802, + "flos": 10653358282200.0, + "grad_norm": 7.905733476133274, + "learning_rate": 2.857024201976089e-06, + "loss": 0.8051, + "num_input_tokens_seen": 66712635, + "step": 3146 + }, + { + "epoch": 0.37840437684121925, + "flos": 24100937452080.0, + "grad_norm": 4.914051243901059, + "learning_rate": 2.8563203080298516e-06, + "loss": 0.7131, + "num_input_tokens_seen": 66733130, + "step": 3147 + }, + { + "epoch": 0.37852461973185836, + "flos": 13407601623120.0, + "grad_norm": 7.100366405649952, + "learning_rate": 2.855616284180505e-06, + "loss": 0.8692, + "num_input_tokens_seen": 66749900, + "step": 3148 + }, + { + "epoch": 0.37864486262249747, + "flos": 43854518577960.0, + "grad_norm": 0.9354139006530602, + "learning_rate": 2.8549121305348477e-06, + "loss": 0.7195, + "num_input_tokens_seen": 66809405, + "step": 3149 + }, + { + "epoch": 0.3787651055131365, + "flos": 17102685097560.0, + "grad_norm": 3.300927442213743, + "learning_rate": 2.8542078471997006e-06, + "loss": 0.8007, + "num_input_tokens_seen": 66826740, + "step": 3150 + }, + { + "epoch": 0.37888534840377563, + "flos": 18018866613600.0, + "grad_norm": 2.848290476140839, + "learning_rate": 2.8535034342819013e-06, + "loss": 0.7428, + "num_input_tokens_seen": 66843870, + "step": 3151 + }, + { + "epoch": 0.37900559129441475, + "flos": 17564907480360.0, + "grad_norm": 4.401442915402724, + "learning_rate": 2.85279889188831e-06, + "loss": 0.7063, + "num_input_tokens_seen": 66863965, + "step": 3152 + }, + { + "epoch": 0.3791258341850538, + "flos": 18051851800560.0, + "grad_norm": 4.672717372518117, + "learning_rate": 2.852094220125805e-06, + "loss": 0.7848, + "num_input_tokens_seen": 66883195, + "step": 3153 + }, + { + "epoch": 0.3792460770756929, + "flos": 12705785425320.0, + "grad_norm": 2.9702873666373213, + "learning_rate": 2.8513894191012846e-06, + "loss": 0.6899, + "num_input_tokens_seen": 66901895, + "step": 3154 + }, + { + "epoch": 0.37936631996633197, + "flos": 17727581066640.0, + "grad_norm": 3.541341624584412, + "learning_rate": 2.8506844889216664e-06, + "loss": 0.7677, + "num_input_tokens_seen": 66921000, + "step": 3155 + }, + { + "epoch": 0.3794865628569711, + "flos": 51844131379440.0, + "grad_norm": 0.929752091159505, + "learning_rate": 2.849979429693887e-06, + "loss": 0.6623, + "num_input_tokens_seen": 66981705, + "step": 3156 + }, + { + "epoch": 0.3796068057476102, + "flos": 11491036571520.0, + "grad_norm": 3.8013846534931814, + "learning_rate": 2.8492742415249042e-06, + "loss": 0.7264, + "num_input_tokens_seen": 66999070, + "step": 3157 + }, + { + "epoch": 0.37972704863824924, + "flos": 18456651998640.0, + "grad_norm": 2.3088819707918393, + "learning_rate": 2.848568924521694e-06, + "loss": 0.743, + "num_input_tokens_seen": 67019570, + "step": 3158 + }, + { + "epoch": 0.37984729152888835, + "flos": 19210128268440.0, + "grad_norm": 2.4409219246160934, + "learning_rate": 2.8478634787912526e-06, + "loss": 0.7157, + "num_input_tokens_seen": 67037345, + "step": 3159 + }, + { + "epoch": 0.37996753441952746, + "flos": 18780479493120.0, + "grad_norm": 2.606651392336661, + "learning_rate": 2.847157904440596e-06, + "loss": 0.7583, + "num_input_tokens_seen": 67056795, + "step": 3160 + }, + { + "epoch": 0.3800877773101665, + "flos": 14701835162400.0, + "grad_norm": 2.891340507413399, + "learning_rate": 2.846452201576759e-06, + "loss": 0.7378, + "num_input_tokens_seen": 67075890, + "step": 3161 + }, + { + "epoch": 0.38020802020080563, + "flos": 46480715346120.0, + "grad_norm": 0.9173193074905756, + "learning_rate": 2.845746370306795e-06, + "loss": 0.6491, + "num_input_tokens_seen": 67140800, + "step": 3162 + }, + { + "epoch": 0.38032826309144474, + "flos": 15568066244280.0, + "grad_norm": 2.7198724293244556, + "learning_rate": 2.84504041073778e-06, + "loss": 0.7645, + "num_input_tokens_seen": 67158935, + "step": 3163 + }, + { + "epoch": 0.3804485059820838, + "flos": 13840638014160.0, + "grad_norm": 2.821349568838892, + "learning_rate": 2.844334322976806e-06, + "loss": 0.7863, + "num_input_tokens_seen": 67178870, + "step": 3164 + }, + { + "epoch": 0.3805687488727229, + "flos": 15971378424240.0, + "grad_norm": 2.6176265519089594, + "learning_rate": 2.8436281071309866e-06, + "loss": 0.8149, + "num_input_tokens_seen": 67197130, + "step": 3165 + }, + { + "epoch": 0.380688991763362, + "flos": 43147320186960.0, + "grad_norm": 4.332464253255566, + "learning_rate": 2.842921763307455e-06, + "loss": 0.5473, + "num_input_tokens_seen": 67259660, + "step": 3166 + }, + { + "epoch": 0.38080923465400107, + "flos": 17425531133280.0, + "grad_norm": 2.119788117526846, + "learning_rate": 2.842215291613361e-06, + "loss": 0.8137, + "num_input_tokens_seen": 67277760, + "step": 3167 + }, + { + "epoch": 0.3809294775446402, + "flos": 40500481124880.0, + "grad_norm": 0.8393126471815691, + "learning_rate": 2.8415086921558774e-06, + "loss": 0.6333, + "num_input_tokens_seen": 67340905, + "step": 3168 + }, + { + "epoch": 0.38104972043527924, + "flos": 18050996981640.0, + "grad_norm": 1.8775139168377186, + "learning_rate": 2.840801965042194e-06, + "loss": 0.7649, + "num_input_tokens_seen": 67360085, + "step": 3169 + }, + { + "epoch": 0.38116996332591835, + "flos": 16728685549200.0, + "grad_norm": 3.0501001748065555, + "learning_rate": 2.840095110379521e-06, + "loss": 0.8187, + "num_input_tokens_seen": 67379325, + "step": 3170 + }, + { + "epoch": 0.38129020621655746, + "flos": 44840813431320.0, + "grad_norm": 0.769483806871017, + "learning_rate": 2.8393881282750884e-06, + "loss": 0.5468, + "num_input_tokens_seen": 67441875, + "step": 3171 + }, + { + "epoch": 0.3814104491071965, + "flos": 15833521755240.0, + "grad_norm": 3.2202225066839265, + "learning_rate": 2.838681018836144e-06, + "loss": 0.7697, + "num_input_tokens_seen": 67458915, + "step": 3172 + }, + { + "epoch": 0.3815306919978356, + "flos": 13947789013320.0, + "grad_norm": 2.7880327476865925, + "learning_rate": 2.837973782169955e-06, + "loss": 0.7596, + "num_input_tokens_seen": 67477010, + "step": 3173 + }, + { + "epoch": 0.38165093488847474, + "flos": 49451698502280.0, + "grad_norm": 0.8571382022912923, + "learning_rate": 2.8372664183838096e-06, + "loss": 0.6187, + "num_input_tokens_seen": 67539750, + "step": 3174 + }, + { + "epoch": 0.3817711777791138, + "flos": 16346359431360.0, + "grad_norm": 3.280860223780251, + "learning_rate": 2.836558927585015e-06, + "loss": 0.6683, + "num_input_tokens_seen": 67556440, + "step": 3175 + }, + { + "epoch": 0.3818914206697529, + "flos": 16701969034320.0, + "grad_norm": 5.859550453102283, + "learning_rate": 2.8358513098808957e-06, + "loss": 0.8071, + "num_input_tokens_seen": 67576475, + "step": 3176 + }, + { + "epoch": 0.382011663560392, + "flos": 17860023882480.0, + "grad_norm": 3.8387532825014965, + "learning_rate": 2.835143565378798e-06, + "loss": 0.7497, + "num_input_tokens_seen": 67596660, + "step": 3177 + }, + { + "epoch": 0.38213190645103107, + "flos": 16081030560240.0, + "grad_norm": 2.38426159430276, + "learning_rate": 2.8344356941860847e-06, + "loss": 0.7603, + "num_input_tokens_seen": 67616010, + "step": 3178 + }, + { + "epoch": 0.3822521493416702, + "flos": 26098000307880.0, + "grad_norm": 9.754116265385212, + "learning_rate": 2.8337276964101403e-06, + "loss": 0.643, + "num_input_tokens_seen": 67636170, + "step": 3179 + }, + { + "epoch": 0.3823723922323093, + "flos": 15406405776720.0, + "grad_norm": 2.7807645909911636, + "learning_rate": 2.833019572158367e-06, + "loss": 0.7389, + "num_input_tokens_seen": 67654325, + "step": 3180 + }, + { + "epoch": 0.38249263512294834, + "flos": 14458030572720.0, + "grad_norm": 2.39794091666184, + "learning_rate": 2.8323113215381872e-06, + "loss": 0.8005, + "num_input_tokens_seen": 67672390, + "step": 3181 + }, + { + "epoch": 0.38261287801358745, + "flos": 15675692142840.0, + "grad_norm": 2.471119672920907, + "learning_rate": 2.831602944657042e-06, + "loss": 0.7389, + "num_input_tokens_seen": 67690190, + "step": 3182 + }, + { + "epoch": 0.38273312090422656, + "flos": 16081030560240.0, + "grad_norm": 3.2596298747198302, + "learning_rate": 2.830894441622391e-06, + "loss": 0.7307, + "num_input_tokens_seen": 67706560, + "step": 3183 + }, + { + "epoch": 0.3828533637948656, + "flos": 18160110898320.0, + "grad_norm": 2.8966083358609938, + "learning_rate": 2.8301858125417134e-06, + "loss": 0.7799, + "num_input_tokens_seen": 67726120, + "step": 3184 + }, + { + "epoch": 0.38297360668550473, + "flos": 16431669549480.0, + "grad_norm": 2.690790247079485, + "learning_rate": 2.8294770575225082e-06, + "loss": 0.7305, + "num_input_tokens_seen": 67745970, + "step": 3185 + }, + { + "epoch": 0.3830938495761438, + "flos": 18238075905720.0, + "grad_norm": 2.3623952166213384, + "learning_rate": 2.828768176672293e-06, + "loss": 0.8268, + "num_input_tokens_seen": 67764805, + "step": 3186 + }, + { + "epoch": 0.3832140924667829, + "flos": 24262154680200.0, + "grad_norm": 2.6054520956762515, + "learning_rate": 2.8280591700986044e-06, + "loss": 0.7082, + "num_input_tokens_seen": 67786390, + "step": 3187 + }, + { + "epoch": 0.383334335357422, + "flos": 23423811531720.0, + "grad_norm": 2.663509725459471, + "learning_rate": 2.827350037908999e-06, + "loss": 0.7328, + "num_input_tokens_seen": 67805550, + "step": 3188 + }, + { + "epoch": 0.38345457824806106, + "flos": 14460468389640.0, + "grad_norm": 3.4486653109664016, + "learning_rate": 2.8266407802110496e-06, + "loss": 0.7733, + "num_input_tokens_seen": 67823525, + "step": 3189 + }, + { + "epoch": 0.3835748211387002, + "flos": 16405048014480.0, + "grad_norm": 2.598137878283274, + "learning_rate": 2.8259313971123515e-06, + "loss": 0.7329, + "num_input_tokens_seen": 67844365, + "step": 3190 + }, + { + "epoch": 0.3836950640293393, + "flos": 18402395809920.0, + "grad_norm": 2.47140350667716, + "learning_rate": 2.8252218887205166e-06, + "loss": 0.76, + "num_input_tokens_seen": 67864775, + "step": 3191 + }, + { + "epoch": 0.38381530691997834, + "flos": 15946086607560.0, + "grad_norm": 2.5581172586785716, + "learning_rate": 2.824512255143178e-06, + "loss": 0.7903, + "num_input_tokens_seen": 67883730, + "step": 3192 + }, + { + "epoch": 0.38393554981061745, + "flos": 15541729648920.0, + "grad_norm": 6.003091333663243, + "learning_rate": 2.8238024964879855e-06, + "loss": 0.7774, + "num_input_tokens_seen": 67904345, + "step": 3193 + }, + { + "epoch": 0.38405579270125656, + "flos": 12408642785760.0, + "grad_norm": 3.4360206888839415, + "learning_rate": 2.8230926128626095e-06, + "loss": 0.7451, + "num_input_tokens_seen": 67922560, + "step": 3194 + }, + { + "epoch": 0.3841760355918956, + "flos": 15972803122440.0, + "grad_norm": 2.5533538043477253, + "learning_rate": 2.822382604374738e-06, + "loss": 0.7685, + "num_input_tokens_seen": 67941205, + "step": 3195 + }, + { + "epoch": 0.3842962784825347, + "flos": 18991140596040.0, + "grad_norm": 3.713471328618504, + "learning_rate": 2.8216724711320793e-06, + "loss": 0.6385, + "num_input_tokens_seen": 67960050, + "step": 3196 + }, + { + "epoch": 0.38441652137317384, + "flos": 18564531176880.0, + "grad_norm": 2.154425028415694, + "learning_rate": 2.820962213242361e-06, + "loss": 0.7844, + "num_input_tokens_seen": 67979100, + "step": 3197 + }, + { + "epoch": 0.3845367642638129, + "flos": 13218306501840.0, + "grad_norm": 3.596958486216167, + "learning_rate": 2.8202518308133264e-06, + "loss": 0.8403, + "num_input_tokens_seen": 67996095, + "step": 3198 + }, + { + "epoch": 0.384657007154452, + "flos": 18483843412920.0, + "grad_norm": 2.8813573788309315, + "learning_rate": 2.8195413239527426e-06, + "loss": 0.7257, + "num_input_tokens_seen": 68015555, + "step": 3199 + }, + { + "epoch": 0.38477725004509106, + "flos": 14515326117600.0, + "grad_norm": 2.3290260158738434, + "learning_rate": 2.8188306927683906e-06, + "loss": 0.7964, + "num_input_tokens_seen": 68034745, + "step": 3200 + }, + { + "epoch": 0.38489749293573017, + "flos": 13326280659960.0, + "grad_norm": 4.182829344483617, + "learning_rate": 2.818119937368074e-06, + "loss": 0.7234, + "num_input_tokens_seen": 68053100, + "step": 3201 + }, + { + "epoch": 0.3850177358263693, + "flos": 17862746639040.0, + "grad_norm": 4.343477528760415, + "learning_rate": 2.817409057859613e-06, + "loss": 0.6372, + "num_input_tokens_seen": 68071810, + "step": 3202 + }, + { + "epoch": 0.38513797871700833, + "flos": 12890648152200.0, + "grad_norm": 2.406656404334227, + "learning_rate": 2.8166980543508482e-06, + "loss": 0.7668, + "num_input_tokens_seen": 68087420, + "step": 3203 + }, + { + "epoch": 0.38525822160764744, + "flos": 18862212035760.0, + "grad_norm": 23.25733225066815, + "learning_rate": 2.815986926949638e-06, + "loss": 0.7693, + "num_input_tokens_seen": 68105640, + "step": 3204 + }, + { + "epoch": 0.38537846449828655, + "flos": 14757832648920.0, + "grad_norm": 3.3097528546313293, + "learning_rate": 2.8152756757638597e-06, + "loss": 0.7818, + "num_input_tokens_seen": 68123860, + "step": 3205 + }, + { + "epoch": 0.3854987073889256, + "flos": 16863819461640.0, + "grad_norm": 3.2113506116359347, + "learning_rate": 2.8145643009014093e-06, + "loss": 0.8211, + "num_input_tokens_seen": 68142075, + "step": 3206 + }, + { + "epoch": 0.3856189502795647, + "flos": 14755173212280.0, + "grad_norm": 3.0073771372216354, + "learning_rate": 2.813852802470202e-06, + "loss": 0.776, + "num_input_tokens_seen": 68159690, + "step": 3207 + }, + { + "epoch": 0.38573919317020383, + "flos": 18834799001760.0, + "grad_norm": 4.4634119977184445, + "learning_rate": 2.8131411805781717e-06, + "loss": 0.7017, + "num_input_tokens_seen": 68179535, + "step": 3208 + }, + { + "epoch": 0.3858594360608429, + "flos": 21882385787040.0, + "grad_norm": 6.025834414369872, + "learning_rate": 2.8124294353332707e-06, + "loss": 0.6168, + "num_input_tokens_seen": 68197930, + "step": 3209 + }, + { + "epoch": 0.385979678951482, + "flos": 18159699318840.0, + "grad_norm": 2.3593720695022675, + "learning_rate": 2.8117175668434713e-06, + "loss": 0.7531, + "num_input_tokens_seen": 68217310, + "step": 3210 + }, + { + "epoch": 0.3860999218421211, + "flos": 15567591344880.0, + "grad_norm": 4.578512858863596, + "learning_rate": 2.811005575216762e-06, + "loss": 0.6569, + "num_input_tokens_seen": 68235745, + "step": 3211 + }, + { + "epoch": 0.38622016473276016, + "flos": 17972177155320.0, + "grad_norm": 2.369700391639516, + "learning_rate": 2.8102934605611513e-06, + "loss": 0.7719, + "num_input_tokens_seen": 68257100, + "step": 3212 + }, + { + "epoch": 0.3863404076233993, + "flos": 15027593914440.0, + "grad_norm": 4.398171669226023, + "learning_rate": 2.8095812229846665e-06, + "loss": 0.6539, + "num_input_tokens_seen": 68276780, + "step": 3213 + }, + { + "epoch": 0.3864606505140384, + "flos": 16350918465600.0, + "grad_norm": 5.472048740036046, + "learning_rate": 2.808868862595355e-06, + "loss": 0.6748, + "num_input_tokens_seen": 68296745, + "step": 3214 + }, + { + "epoch": 0.38658089340467744, + "flos": 18835653820680.0, + "grad_norm": 4.025197771793429, + "learning_rate": 2.8081563795012795e-06, + "loss": 0.7664, + "num_input_tokens_seen": 68316090, + "step": 3215 + }, + { + "epoch": 0.38670113629531655, + "flos": 24829280205000.0, + "grad_norm": 2.83793246296801, + "learning_rate": 2.807443773810524e-06, + "loss": 0.7211, + "num_input_tokens_seen": 68337070, + "step": 3216 + }, + { + "epoch": 0.3868213791859556, + "flos": 17080084377480.0, + "grad_norm": 11.296969549223975, + "learning_rate": 2.80673104563119e-06, + "loss": 0.8737, + "num_input_tokens_seen": 68357415, + "step": 3217 + }, + { + "epoch": 0.3869416220765947, + "flos": 13461034652880.0, + "grad_norm": 2.0630087947389875, + "learning_rate": 2.8060181950713976e-06, + "loss": 0.7656, + "num_input_tokens_seen": 68373925, + "step": 3218 + }, + { + "epoch": 0.3870618649672338, + "flos": 11382650833920.0, + "grad_norm": 3.4380197305518148, + "learning_rate": 2.805305222239286e-06, + "loss": 0.7886, + "num_input_tokens_seen": 68390900, + "step": 3219 + }, + { + "epoch": 0.3871821078578729, + "flos": 17214965010240.0, + "grad_norm": 4.348395847637828, + "learning_rate": 2.8045921272430118e-06, + "loss": 0.7178, + "num_input_tokens_seen": 68410300, + "step": 3220 + }, + { + "epoch": 0.387302350748512, + "flos": 12969246358800.0, + "grad_norm": 9.098166530477918, + "learning_rate": 2.803878910190753e-06, + "loss": 0.7546, + "num_input_tokens_seen": 68426940, + "step": 3221 + }, + { + "epoch": 0.3874225936391511, + "flos": 8325534400680.0, + "grad_norm": 7.149409727260159, + "learning_rate": 2.8031655711907017e-06, + "loss": 0.8065, + "num_input_tokens_seen": 68440365, + "step": 3222 + }, + { + "epoch": 0.38754283652979016, + "flos": 16054092425640.0, + "grad_norm": 3.6792278286151237, + "learning_rate": 2.8024521103510723e-06, + "loss": 0.7839, + "num_input_tokens_seen": 68456855, + "step": 3223 + }, + { + "epoch": 0.38766307942042927, + "flos": 15484434104040.0, + "grad_norm": 2.979664385285361, + "learning_rate": 2.8017385277800952e-06, + "loss": 0.7388, + "num_input_tokens_seen": 68474930, + "step": 3224 + }, + { + "epoch": 0.3877833223110684, + "flos": 20103740724360.0, + "grad_norm": 3.97771684421159, + "learning_rate": 2.8010248235860213e-06, + "loss": 0.7195, + "num_input_tokens_seen": 68494765, + "step": 3225 + }, + { + "epoch": 0.38790356520170743, + "flos": 47554477686240.0, + "grad_norm": 0.9318304608227468, + "learning_rate": 2.8003109978771192e-06, + "loss": 0.6606, + "num_input_tokens_seen": 68555650, + "step": 3226 + }, + { + "epoch": 0.38802380809234654, + "flos": 16350158626560.0, + "grad_norm": 7.377533041228217, + "learning_rate": 2.799597050761674e-06, + "loss": 0.778, + "num_input_tokens_seen": 68571575, + "step": 3227 + }, + { + "epoch": 0.38814405098298566, + "flos": 18507742191360.0, + "grad_norm": 5.8404531715879475, + "learning_rate": 2.7988829823479924e-06, + "loss": 0.7787, + "num_input_tokens_seen": 68589685, + "step": 3228 + }, + { + "epoch": 0.3882642938736247, + "flos": 13756562634480.0, + "grad_norm": 3.6453774012720657, + "learning_rate": 2.7981687927443976e-06, + "loss": 0.6314, + "num_input_tokens_seen": 68606205, + "step": 3229 + }, + { + "epoch": 0.3883845367642638, + "flos": 15837669210000.0, + "grad_norm": 4.864048712662157, + "learning_rate": 2.797454482059231e-06, + "loss": 0.8365, + "num_input_tokens_seen": 68626080, + "step": 3230 + }, + { + "epoch": 0.3885047796549029, + "flos": 15026992375200.0, + "grad_norm": 3.117261536753917, + "learning_rate": 2.7967400504008537e-06, + "loss": 0.8263, + "num_input_tokens_seen": 68645100, + "step": 3231 + }, + { + "epoch": 0.388625022545542, + "flos": 47421270540000.0, + "grad_norm": 0.8414785752518478, + "learning_rate": 2.7960254978776456e-06, + "loss": 0.6169, + "num_input_tokens_seen": 68706910, + "step": 3232 + }, + { + "epoch": 0.3887452654361811, + "flos": 13218654761400.0, + "grad_norm": 2.825869677876234, + "learning_rate": 2.7953108245980006e-06, + "loss": 0.7835, + "num_input_tokens_seen": 68725145, + "step": 3233 + }, + { + "epoch": 0.38886550832682015, + "flos": 18296827808760.0, + "grad_norm": 2.389930474614781, + "learning_rate": 2.7945960306703365e-06, + "loss": 0.7236, + "num_input_tokens_seen": 68747850, + "step": 3234 + }, + { + "epoch": 0.38898575121745926, + "flos": 19942713456000.0, + "grad_norm": 5.387213027993638, + "learning_rate": 2.7938811162030865e-06, + "loss": 0.639, + "num_input_tokens_seen": 68767835, + "step": 3235 + }, + { + "epoch": 0.3891059941080984, + "flos": 21099691865520.0, + "grad_norm": 2.198489676414198, + "learning_rate": 2.793166081304702e-06, + "loss": 0.8077, + "num_input_tokens_seen": 68788050, + "step": 3236 + }, + { + "epoch": 0.38922623699873743, + "flos": 16755845303520.0, + "grad_norm": 5.167804326147104, + "learning_rate": 2.7924509260836543e-06, + "loss": 0.7958, + "num_input_tokens_seen": 68806895, + "step": 3237 + }, + { + "epoch": 0.38934647988937654, + "flos": 14460911629080.0, + "grad_norm": 2.3670900028911257, + "learning_rate": 2.791735650648431e-06, + "loss": 0.6636, + "num_input_tokens_seen": 68825735, + "step": 3238 + }, + { + "epoch": 0.38946672278001565, + "flos": 14023094584080.0, + "grad_norm": 2.5846286008089967, + "learning_rate": 2.791020255107538e-06, + "loss": 0.718, + "num_input_tokens_seen": 68842825, + "step": 3239 + }, + { + "epoch": 0.3895869656706547, + "flos": 18267800116800.0, + "grad_norm": 1.9824456875878944, + "learning_rate": 2.7903047395695023e-06, + "loss": 0.7887, + "num_input_tokens_seen": 68862445, + "step": 3240 + }, + { + "epoch": 0.3897072085612938, + "flos": 17673609817560.0, + "grad_norm": 8.992138289939197, + "learning_rate": 2.789589104142865e-06, + "loss": 0.8841, + "num_input_tokens_seen": 68879790, + "step": 3241 + }, + { + "epoch": 0.3898274514519329, + "flos": 12517566742680.0, + "grad_norm": 13.532758738780178, + "learning_rate": 2.7888733489361895e-06, + "loss": 0.7526, + "num_input_tokens_seen": 68897925, + "step": 3242 + }, + { + "epoch": 0.389947694342572, + "flos": 48715599059160.0, + "grad_norm": 0.7921957749154654, + "learning_rate": 2.788157474058054e-06, + "loss": 0.6423, + "num_input_tokens_seen": 68959920, + "step": 3243 + }, + { + "epoch": 0.3900679372332111, + "flos": 18863383454280.0, + "grad_norm": 5.9856503873081355, + "learning_rate": 2.7874414796170555e-06, + "loss": 0.6873, + "num_input_tokens_seen": 68981130, + "step": 3244 + }, + { + "epoch": 0.3901881801238502, + "flos": 8547118189800.0, + "grad_norm": 5.220171158129397, + "learning_rate": 2.7867253657218113e-06, + "loss": 0.8105, + "num_input_tokens_seen": 68994740, + "step": 3245 + }, + { + "epoch": 0.39030842301448926, + "flos": 20023717819560.0, + "grad_norm": 4.516830758542803, + "learning_rate": 2.7860091324809544e-06, + "loss": 0.7105, + "num_input_tokens_seen": 69015520, + "step": 3246 + }, + { + "epoch": 0.39042866590512837, + "flos": 19915996941120.0, + "grad_norm": 2.982073790416947, + "learning_rate": 2.7852927800031377e-06, + "loss": 0.7991, + "num_input_tokens_seen": 69035405, + "step": 3247 + }, + { + "epoch": 0.3905489087957674, + "flos": 21805053978840.0, + "grad_norm": 3.1446331505228375, + "learning_rate": 2.7845763083970298e-06, + "loss": 0.8171, + "num_input_tokens_seen": 69055525, + "step": 3248 + }, + { + "epoch": 0.39066915168640653, + "flos": 17943751002600.0, + "grad_norm": 2.8385660059640756, + "learning_rate": 2.7838597177713205e-06, + "loss": 0.7967, + "num_input_tokens_seen": 69076335, + "step": 3249 + }, + { + "epoch": 0.39078939457704565, + "flos": 15027910514040.0, + "grad_norm": 2.719154941612666, + "learning_rate": 2.7831430082347143e-06, + "loss": 0.7181, + "num_input_tokens_seen": 69095260, + "step": 3250 + }, + { + "epoch": 0.3909096374676847, + "flos": 16674334380600.0, + "grad_norm": 2.712490549378698, + "learning_rate": 2.7824261798959373e-06, + "loss": 0.8049, + "num_input_tokens_seen": 69113160, + "step": 3251 + }, + { + "epoch": 0.3910298803583238, + "flos": 16837039626840.0, + "grad_norm": 3.144039139239495, + "learning_rate": 2.78170923286373e-06, + "loss": 0.7798, + "num_input_tokens_seen": 69132480, + "step": 3252 + }, + { + "epoch": 0.3911501232489629, + "flos": 17808807049920.0, + "grad_norm": 14.700174260566289, + "learning_rate": 2.780992167246854e-06, + "loss": 0.8276, + "num_input_tokens_seen": 69149725, + "step": 3253 + }, + { + "epoch": 0.391270366139602, + "flos": 44864648889840.0, + "grad_norm": 1.0143764744511088, + "learning_rate": 2.7802749831540883e-06, + "loss": 0.7726, + "num_input_tokens_seen": 69208345, + "step": 3254 + }, + { + "epoch": 0.3913906090302411, + "flos": 15756791486280.0, + "grad_norm": 6.353046089128876, + "learning_rate": 2.7795576806942268e-06, + "loss": 0.8035, + "num_input_tokens_seen": 69226870, + "step": 3255 + }, + { + "epoch": 0.3915108519208802, + "flos": 36700901596560.0, + "grad_norm": 0.802972762822867, + "learning_rate": 2.778840259976085e-06, + "loss": 0.5694, + "num_input_tokens_seen": 69281820, + "step": 3256 + }, + { + "epoch": 0.39163109481151925, + "flos": 12028754484840.0, + "grad_norm": 3.2849434546680114, + "learning_rate": 2.778122721108495e-06, + "loss": 0.7535, + "num_input_tokens_seen": 69299770, + "step": 3257 + }, + { + "epoch": 0.39175133770215836, + "flos": 19104813546960.0, + "grad_norm": 2.9529370444898926, + "learning_rate": 2.7774050642003076e-06, + "loss": 0.8666, + "num_input_tokens_seen": 69320300, + "step": 3258 + }, + { + "epoch": 0.3918715805927975, + "flos": 15999994536720.0, + "grad_norm": 4.07580874790058, + "learning_rate": 2.7766872893603896e-06, + "loss": 0.9246, + "num_input_tokens_seen": 69339995, + "step": 3259 + }, + { + "epoch": 0.39199182348343653, + "flos": 14892934901400.0, + "grad_norm": 4.11468279110742, + "learning_rate": 2.7759693966976275e-06, + "loss": 0.713, + "num_input_tokens_seen": 69358220, + "step": 3260 + }, + { + "epoch": 0.39211206637407564, + "flos": 15859605070920.0, + "grad_norm": 2.266809403843219, + "learning_rate": 2.7752513863209242e-06, + "loss": 0.825, + "num_input_tokens_seen": 69376520, + "step": 3261 + }, + { + "epoch": 0.39223230926471475, + "flos": 15403208120760.0, + "grad_norm": 5.903033960305788, + "learning_rate": 2.774533258339203e-06, + "loss": 0.8229, + "num_input_tokens_seen": 69393700, + "step": 3262 + }, + { + "epoch": 0.3923525521553538, + "flos": 12840571078200.0, + "grad_norm": 5.629710450391656, + "learning_rate": 2.7738150128614014e-06, + "loss": 0.7724, + "num_input_tokens_seen": 69410825, + "step": 3263 + }, + { + "epoch": 0.3924727950459929, + "flos": 15027657234360.0, + "grad_norm": 3.0460878578084043, + "learning_rate": 2.7730966499964777e-06, + "loss": 0.8842, + "num_input_tokens_seen": 69427495, + "step": 3264 + }, + { + "epoch": 0.39259303793663197, + "flos": 11814104226960.0, + "grad_norm": 4.104056626343785, + "learning_rate": 2.772378169853408e-06, + "loss": 0.7856, + "num_input_tokens_seen": 69444785, + "step": 3265 + }, + { + "epoch": 0.3927132808272711, + "flos": 12243594702480.0, + "grad_norm": 2.7468908246553383, + "learning_rate": 2.771659572541183e-06, + "loss": 0.7292, + "num_input_tokens_seen": 69462435, + "step": 3266 + }, + { + "epoch": 0.3928335237179102, + "flos": 14812373777280.0, + "grad_norm": 2.978806978578001, + "learning_rate": 2.7709408581688143e-06, + "loss": 0.8539, + "num_input_tokens_seen": 69482140, + "step": 3267 + }, + { + "epoch": 0.39295376660854925, + "flos": 18295054851000.0, + "grad_norm": 4.401330893980327, + "learning_rate": 2.7702220268453307e-06, + "loss": 0.865, + "num_input_tokens_seen": 69502220, + "step": 3268 + }, + { + "epoch": 0.39307400949918836, + "flos": 13650804673560.0, + "grad_norm": 5.141278164538575, + "learning_rate": 2.7695030786797785e-06, + "loss": 0.82, + "num_input_tokens_seen": 69517835, + "step": 3269 + }, + { + "epoch": 0.39319425238982747, + "flos": 16401597078840.0, + "grad_norm": 2.5618505818007615, + "learning_rate": 2.7687840137812206e-06, + "loss": 0.7264, + "num_input_tokens_seen": 69535640, + "step": 3270 + }, + { + "epoch": 0.3933144952804665, + "flos": 48804265133040.0, + "grad_norm": 0.7996069508838667, + "learning_rate": 2.7680648322587395e-06, + "loss": 0.6264, + "num_input_tokens_seen": 69600235, + "step": 3271 + }, + { + "epoch": 0.39343473817110564, + "flos": 11274739995720.0, + "grad_norm": 10.801046843479527, + "learning_rate": 2.7673455342214334e-06, + "loss": 0.7975, + "num_input_tokens_seen": 69616945, + "step": 3272 + }, + { + "epoch": 0.39355498106174475, + "flos": 15595067698800.0, + "grad_norm": 3.082323802605574, + "learning_rate": 2.7666261197784198e-06, + "loss": 0.7368, + "num_input_tokens_seen": 69635480, + "step": 3273 + }, + { + "epoch": 0.3936752239523838, + "flos": 9653164706400.0, + "grad_norm": 4.112329385575025, + "learning_rate": 2.7659065890388336e-06, + "loss": 0.7533, + "num_input_tokens_seen": 69651200, + "step": 3274 + }, + { + "epoch": 0.3937954668430229, + "flos": 12246380778960.0, + "grad_norm": 3.0893940170292384, + "learning_rate": 2.7651869421118266e-06, + "loss": 0.8174, + "num_input_tokens_seen": 69667530, + "step": 3275 + }, + { + "epoch": 0.393915709733662, + "flos": 15401403503040.0, + "grad_norm": 2.3649064458041478, + "learning_rate": 2.76446717910657e-06, + "loss": 0.8083, + "num_input_tokens_seen": 69687955, + "step": 3276 + }, + { + "epoch": 0.3940359526243011, + "flos": 12516616943880.0, + "grad_norm": 3.766102234289921, + "learning_rate": 2.763747300132249e-06, + "loss": 0.7407, + "num_input_tokens_seen": 69705115, + "step": 3277 + }, + { + "epoch": 0.3941561955149402, + "flos": 15351453068880.0, + "grad_norm": 1.947229505565528, + "learning_rate": 2.7630273052980704e-06, + "loss": 0.8508, + "num_input_tokens_seen": 69725425, + "step": 3278 + }, + { + "epoch": 0.39427643840557924, + "flos": 13758525552000.0, + "grad_norm": 3.9115507247856063, + "learning_rate": 2.7623071947132554e-06, + "loss": 0.6575, + "num_input_tokens_seen": 69742175, + "step": 3279 + }, + { + "epoch": 0.39439668129621835, + "flos": 17026176448320.0, + "grad_norm": 4.1498129193525095, + "learning_rate": 2.7615869684870458e-06, + "loss": 0.7705, + "num_input_tokens_seen": 69761205, + "step": 3280 + }, + { + "epoch": 0.39451692418685746, + "flos": 19537754958120.0, + "grad_norm": 3.2197680298550484, + "learning_rate": 2.7608666267286986e-06, + "loss": 0.8303, + "num_input_tokens_seen": 69781155, + "step": 3281 + }, + { + "epoch": 0.3946371670774965, + "flos": 13325679120720.0, + "grad_norm": 2.9989803488743587, + "learning_rate": 2.760146169547489e-06, + "loss": 0.8314, + "num_input_tokens_seen": 69797640, + "step": 3282 + }, + { + "epoch": 0.39475740996813563, + "flos": 17727201147120.0, + "grad_norm": 1.6923778113430505, + "learning_rate": 2.75942559705271e-06, + "loss": 0.7477, + "num_input_tokens_seen": 69817095, + "step": 3283 + }, + { + "epoch": 0.39487765285877474, + "flos": 14109671100600.0, + "grad_norm": 2.5716985889990904, + "learning_rate": 2.7587049093536713e-06, + "loss": 0.8665, + "num_input_tokens_seen": 69833145, + "step": 3284 + }, + { + "epoch": 0.3949978957494138, + "flos": 12625097661360.0, + "grad_norm": 2.9991404573596068, + "learning_rate": 2.757984106559701e-06, + "loss": 0.7863, + "num_input_tokens_seen": 69851850, + "step": 3285 + }, + { + "epoch": 0.3951181386400529, + "flos": 26690480969280.0, + "grad_norm": 3.0309079950628877, + "learning_rate": 2.7572631887801446e-06, + "loss": 0.6979, + "num_input_tokens_seen": 69873195, + "step": 3286 + }, + { + "epoch": 0.395238381530692, + "flos": 16916081072880.0, + "grad_norm": 2.3496718806627754, + "learning_rate": 2.7565421561243654e-06, + "loss": 0.7367, + "num_input_tokens_seen": 69891080, + "step": 3287 + }, + { + "epoch": 0.3953586244213311, + "flos": 17831281130160.0, + "grad_norm": 4.153023062230683, + "learning_rate": 2.7558210087017413e-06, + "loss": 0.8023, + "num_input_tokens_seen": 69910735, + "step": 3288 + }, + { + "epoch": 0.3954788673119702, + "flos": 17160297242040.0, + "grad_norm": 2.6586854428451256, + "learning_rate": 2.7550997466216724e-06, + "loss": 0.7311, + "num_input_tokens_seen": 69928250, + "step": 3289 + }, + { + "epoch": 0.3955991102026093, + "flos": 12760009954080.0, + "grad_norm": 2.991798475190116, + "learning_rate": 2.7543783699935714e-06, + "loss": 0.787, + "num_input_tokens_seen": 69946000, + "step": 3290 + }, + { + "epoch": 0.39571935309324835, + "flos": 13297886167200.0, + "grad_norm": 3.614825076650021, + "learning_rate": 2.753656878926872e-06, + "loss": 0.8374, + "num_input_tokens_seen": 69961600, + "step": 3291 + }, + { + "epoch": 0.39583959598388746, + "flos": 12948450256440.0, + "grad_norm": 2.6414489363544242, + "learning_rate": 2.752935273531023e-06, + "loss": 0.7255, + "num_input_tokens_seen": 69979470, + "step": 3292 + }, + { + "epoch": 0.39595983887452657, + "flos": 14134931257320.0, + "grad_norm": 2.6025979670573602, + "learning_rate": 2.752213553915492e-06, + "loss": 0.7688, + "num_input_tokens_seen": 69997545, + "step": 3293 + }, + { + "epoch": 0.3960800817651656, + "flos": 44725810762080.0, + "grad_norm": 0.8013785390958703, + "learning_rate": 2.751491720189762e-06, + "loss": 0.6752, + "num_input_tokens_seen": 70055375, + "step": 3294 + }, + { + "epoch": 0.39620032465580474, + "flos": 12273445553400.0, + "grad_norm": 3.4177332508155853, + "learning_rate": 2.7507697724633364e-06, + "loss": 0.897, + "num_input_tokens_seen": 70071855, + "step": 3295 + }, + { + "epoch": 0.3963205675464438, + "flos": 50933992424400.0, + "grad_norm": 0.767163657383138, + "learning_rate": 2.7500477108457327e-06, + "loss": 0.5578, + "num_input_tokens_seen": 70123585, + "step": 3296 + }, + { + "epoch": 0.3964408104370829, + "flos": 18808937305800.0, + "grad_norm": 3.1070256391191164, + "learning_rate": 2.7493255354464877e-06, + "loss": 0.7939, + "num_input_tokens_seen": 70141115, + "step": 3297 + }, + { + "epoch": 0.396561053327722, + "flos": 17779779357960.0, + "grad_norm": 3.3767866421241695, + "learning_rate": 2.748603246375156e-06, + "loss": 0.7408, + "num_input_tokens_seen": 70158850, + "step": 3298 + }, + { + "epoch": 0.39668129621836107, + "flos": 15001099019280.0, + "grad_norm": 2.912170122972605, + "learning_rate": 2.7478808437413055e-06, + "loss": 0.6688, + "num_input_tokens_seen": 70177980, + "step": 3299 + }, + { + "epoch": 0.3968015391090002, + "flos": 19833852819000.0, + "grad_norm": 3.288413230620109, + "learning_rate": 2.7471583276545263e-06, + "loss": 0.6415, + "num_input_tokens_seen": 70198360, + "step": 3300 + }, + { + "epoch": 0.3969217819996393, + "flos": 9087590519640.0, + "grad_norm": 2.8791326591665745, + "learning_rate": 2.7464356982244224e-06, + "loss": 0.6821, + "num_input_tokens_seen": 70216080, + "step": 3301 + }, + { + "epoch": 0.39704202489027834, + "flos": 48841809354240.0, + "grad_norm": 0.8247330493786621, + "learning_rate": 2.745712955560617e-06, + "loss": 0.6381, + "num_input_tokens_seen": 70272005, + "step": 3302 + }, + { + "epoch": 0.39716226778091746, + "flos": 12381736311120.0, + "grad_norm": 5.129353604704945, + "learning_rate": 2.7449900997727496e-06, + "loss": 0.7551, + "num_input_tokens_seen": 70289835, + "step": 3303 + }, + { + "epoch": 0.39728251067155657, + "flos": 17187805255920.0, + "grad_norm": 2.699578665601586, + "learning_rate": 2.744267130970476e-06, + "loss": 0.8279, + "num_input_tokens_seen": 70309280, + "step": 3304 + }, + { + "epoch": 0.3974027535621956, + "flos": 15135853012200.0, + "grad_norm": 3.105554274814852, + "learning_rate": 2.7435440492634697e-06, + "loss": 0.7476, + "num_input_tokens_seen": 70328325, + "step": 3305 + }, + { + "epoch": 0.39752299645283473, + "flos": 15756601526520.0, + "grad_norm": 3.3066968648310944, + "learning_rate": 2.7428208547614228e-06, + "loss": 0.6482, + "num_input_tokens_seen": 70347540, + "step": 3306 + }, + { + "epoch": 0.39764323934347384, + "flos": 14028571757160.0, + "grad_norm": 2.715969373774717, + "learning_rate": 2.742097547574043e-06, + "loss": 0.7385, + "num_input_tokens_seen": 70365485, + "step": 3307 + }, + { + "epoch": 0.3977634822341129, + "flos": 15243700530480.0, + "grad_norm": 2.92982454106567, + "learning_rate": 2.7413741278110544e-06, + "loss": 0.7546, + "num_input_tokens_seen": 70383895, + "step": 3308 + }, + { + "epoch": 0.397883725124752, + "flos": 29257455426360.0, + "grad_norm": 3.214230225099224, + "learning_rate": 2.7406505955822016e-06, + "loss": 0.6649, + "num_input_tokens_seen": 70404640, + "step": 3309 + }, + { + "epoch": 0.39800396801539106, + "flos": 12674509876200.0, + "grad_norm": 4.693504048696129, + "learning_rate": 2.7399269509972415e-06, + "loss": 0.6453, + "num_input_tokens_seen": 70418515, + "step": 3310 + }, + { + "epoch": 0.3981242109060302, + "flos": 14023696123320.0, + "grad_norm": 9.284328374036196, + "learning_rate": 2.7392031941659514e-06, + "loss": 0.8258, + "num_input_tokens_seen": 70436080, + "step": 3311 + }, + { + "epoch": 0.3982444537966693, + "flos": 17992055118840.0, + "grad_norm": 2.457021934600783, + "learning_rate": 2.7384793251981244e-06, + "loss": 0.8447, + "num_input_tokens_seen": 70454785, + "step": 3312 + }, + { + "epoch": 0.39836469668730834, + "flos": 19213199284560.0, + "grad_norm": 2.3834348772553935, + "learning_rate": 2.737755344203571e-06, + "loss": 0.79, + "num_input_tokens_seen": 70474455, + "step": 3313 + }, + { + "epoch": 0.39848493957794745, + "flos": 20265686131560.0, + "grad_norm": 2.942107949014581, + "learning_rate": 2.7370312512921186e-06, + "loss": 0.7813, + "num_input_tokens_seen": 70495955, + "step": 3314 + }, + { + "epoch": 0.39860518246858656, + "flos": 8871040664160.0, + "grad_norm": 4.19685889359577, + "learning_rate": 2.736307046573611e-06, + "loss": 0.745, + "num_input_tokens_seen": 70511545, + "step": 3315 + }, + { + "epoch": 0.3987254253592256, + "flos": 16374722264160.0, + "grad_norm": 2.11581725651434, + "learning_rate": 2.73558273015791e-06, + "loss": 0.8109, + "num_input_tokens_seen": 70531095, + "step": 3316 + }, + { + "epoch": 0.3988456682498647, + "flos": 17215154970000.0, + "grad_norm": 37.17042346892721, + "learning_rate": 2.734858302154894e-06, + "loss": 0.6856, + "num_input_tokens_seen": 70552315, + "step": 3317 + }, + { + "epoch": 0.39896591114050384, + "flos": 14028983336640.0, + "grad_norm": 4.670097497743247, + "learning_rate": 2.734133762674457e-06, + "loss": 0.7345, + "num_input_tokens_seen": 70571625, + "step": 3318 + }, + { + "epoch": 0.3990861540311429, + "flos": 20831576917920.0, + "grad_norm": 2.581222892415688, + "learning_rate": 2.7334091118265124e-06, + "loss": 0.689, + "num_input_tokens_seen": 70593240, + "step": 3319 + }, + { + "epoch": 0.399206396921782, + "flos": 45524904542880.0, + "grad_norm": 0.6590063474708352, + "learning_rate": 2.732684349720989e-06, + "loss": 0.5798, + "num_input_tokens_seen": 70660920, + "step": 3320 + }, + { + "epoch": 0.3993266398124211, + "flos": 20590241805120.0, + "grad_norm": 3.1728733539860317, + "learning_rate": 2.7319594764678318e-06, + "loss": 0.7262, + "num_input_tokens_seen": 70682740, + "step": 3321 + }, + { + "epoch": 0.39944688270306017, + "flos": 17001739450560.0, + "grad_norm": 1.9906276205815143, + "learning_rate": 2.7312344921770044e-06, + "loss": 0.812, + "num_input_tokens_seen": 70704160, + "step": 3322 + }, + { + "epoch": 0.3995671255936993, + "flos": 14163325750080.0, + "grad_norm": 2.5098320265717433, + "learning_rate": 2.7305093969584857e-06, + "loss": 0.7676, + "num_input_tokens_seen": 70722705, + "step": 3323 + }, + { + "epoch": 0.3996873684843384, + "flos": 17458389680400.0, + "grad_norm": 2.7345519090372, + "learning_rate": 2.729784190922272e-06, + "loss": 0.7772, + "num_input_tokens_seen": 70743860, + "step": 3324 + }, + { + "epoch": 0.39980761137497745, + "flos": 49091091117000.0, + "grad_norm": 0.7742958315327572, + "learning_rate": 2.729058874178378e-06, + "loss": 0.5847, + "num_input_tokens_seen": 70814260, + "step": 3325 + }, + { + "epoch": 0.39992785426561656, + "flos": 20941513993560.0, + "grad_norm": 2.520789111881839, + "learning_rate": 2.7283334468368315e-06, + "loss": 0.6819, + "num_input_tokens_seen": 70835260, + "step": 3326 + }, + { + "epoch": 0.4000480971562556, + "flos": 10950785861400.0, + "grad_norm": 3.2285862226665687, + "learning_rate": 2.72760790900768e-06, + "loss": 0.7137, + "num_input_tokens_seen": 70851565, + "step": 3327 + }, + { + "epoch": 0.4001683400468947, + "flos": 17512392589440.0, + "grad_norm": 2.8219226245937956, + "learning_rate": 2.7268822608009875e-06, + "loss": 0.7841, + "num_input_tokens_seen": 70870660, + "step": 3328 + }, + { + "epoch": 0.40028858293753383, + "flos": 17835745184520.0, + "grad_norm": 2.8391234941916488, + "learning_rate": 2.726156502326834e-06, + "loss": 0.7572, + "num_input_tokens_seen": 70891680, + "step": 3329 + }, + { + "epoch": 0.4004088258281729, + "flos": 49245374813880.0, + "grad_norm": 0.7077154989911975, + "learning_rate": 2.725430633695316e-06, + "loss": 0.6221, + "num_input_tokens_seen": 70954480, + "step": 3330 + }, + { + "epoch": 0.400529068718812, + "flos": 43182991979160.0, + "grad_norm": 0.932822862883854, + "learning_rate": 2.7247046550165485e-06, + "loss": 0.6163, + "num_input_tokens_seen": 71006325, + "step": 3331 + }, + { + "epoch": 0.4006493116094511, + "flos": 18594603647520.0, + "grad_norm": 2.7851636974775515, + "learning_rate": 2.7239785664006606e-06, + "loss": 0.7465, + "num_input_tokens_seen": 71029585, + "step": 3332 + }, + { + "epoch": 0.40076955450009016, + "flos": 44428604802600.0, + "grad_norm": 0.8132452114889485, + "learning_rate": 2.7232523679578002e-06, + "loss": 0.6598, + "num_input_tokens_seen": 71092385, + "step": 3333 + }, + { + "epoch": 0.4008897973907293, + "flos": 12110961926880.0, + "grad_norm": 4.326452930698293, + "learning_rate": 2.7225260597981295e-06, + "loss": 0.7832, + "num_input_tokens_seen": 71109810, + "step": 3334 + }, + { + "epoch": 0.4010100402813684, + "flos": 11193419032560.0, + "grad_norm": 3.813944558551307, + "learning_rate": 2.721799642031831e-06, + "loss": 0.7704, + "num_input_tokens_seen": 71125700, + "step": 3335 + }, + { + "epoch": 0.40113028317200744, + "flos": 9655349243640.0, + "grad_norm": 2.4194472419361515, + "learning_rate": 2.721073114769101e-06, + "loss": 0.7546, + "num_input_tokens_seen": 71143095, + "step": 3336 + }, + { + "epoch": 0.40125052606264655, + "flos": 15108914877600.0, + "grad_norm": 2.4768610151629096, + "learning_rate": 2.7203464781201523e-06, + "loss": 0.7362, + "num_input_tokens_seen": 71162130, + "step": 3337 + }, + { + "epoch": 0.40137076895328566, + "flos": 18024407106600.0, + "grad_norm": 4.67735229715688, + "learning_rate": 2.719619732195215e-06, + "loss": 0.7561, + "num_input_tokens_seen": 71183490, + "step": 3338 + }, + { + "epoch": 0.4014910118439247, + "flos": 17727612726600.0, + "grad_norm": 1.955491364333442, + "learning_rate": 2.7188928771045377e-06, + "loss": 0.7182, + "num_input_tokens_seen": 71204530, + "step": 3339 + }, + { + "epoch": 0.4016112547345638, + "flos": 19591599567360.0, + "grad_norm": 2.4129395079636367, + "learning_rate": 2.7181659129583815e-06, + "loss": 0.7857, + "num_input_tokens_seen": 71223840, + "step": 3340 + }, + { + "epoch": 0.4017314976252029, + "flos": 15567717984720.0, + "grad_norm": 3.017497375790546, + "learning_rate": 2.7174388398670276e-06, + "loss": 0.7514, + "num_input_tokens_seen": 71242740, + "step": 3341 + }, + { + "epoch": 0.401851740515842, + "flos": 18673645093560.0, + "grad_norm": 2.4248650582119224, + "learning_rate": 2.716711657940773e-06, + "loss": 0.8986, + "num_input_tokens_seen": 71263470, + "step": 3342 + }, + { + "epoch": 0.4019719834064811, + "flos": 40813607552880.0, + "grad_norm": 0.8563613763476401, + "learning_rate": 2.7159843672899284e-06, + "loss": 0.5948, + "num_input_tokens_seen": 71327390, + "step": 3343 + }, + { + "epoch": 0.40209222629712016, + "flos": 13266705597960.0, + "grad_norm": 4.45150588198079, + "learning_rate": 2.715256968024825e-06, + "loss": 0.7938, + "num_input_tokens_seen": 71344185, + "step": 3344 + }, + { + "epoch": 0.40221246918775927, + "flos": 19026436960080.0, + "grad_norm": 2.766168659460552, + "learning_rate": 2.7145294602558083e-06, + "loss": 0.82, + "num_input_tokens_seen": 71364615, + "step": 3345 + }, + { + "epoch": 0.4023327120783984, + "flos": 24856186679640.0, + "grad_norm": 3.8717430948198457, + "learning_rate": 2.713801844093241e-06, + "loss": 0.6808, + "num_input_tokens_seen": 71385485, + "step": 3346 + }, + { + "epoch": 0.40245295496903744, + "flos": 19720211528040.0, + "grad_norm": 3.1154526397871685, + "learning_rate": 2.7130741196475014e-06, + "loss": 0.8504, + "num_input_tokens_seen": 71403335, + "step": 3347 + }, + { + "epoch": 0.40257319785967655, + "flos": 27123485700360.0, + "grad_norm": 3.0438002501927963, + "learning_rate": 2.7123462870289848e-06, + "loss": 0.7782, + "num_input_tokens_seen": 71423105, + "step": 3348 + }, + { + "epoch": 0.40269344075031566, + "flos": 17836188423960.0, + "grad_norm": 2.721593838348231, + "learning_rate": 2.711618346348102e-06, + "loss": 0.7899, + "num_input_tokens_seen": 71443350, + "step": 3349 + }, + { + "epoch": 0.4028136836409547, + "flos": 10462100243400.0, + "grad_norm": 2.2747181434836494, + "learning_rate": 2.7108902977152825e-06, + "loss": 0.6172, + "num_input_tokens_seen": 71460970, + "step": 3350 + }, + { + "epoch": 0.4029339265315938, + "flos": 19154922280920.0, + "grad_norm": 5.602996685135348, + "learning_rate": 2.7101621412409704e-06, + "loss": 0.7348, + "num_input_tokens_seen": 71480175, + "step": 3351 + }, + { + "epoch": 0.40305416942223293, + "flos": 17024118550920.0, + "grad_norm": 4.360470432485252, + "learning_rate": 2.7094338770356256e-06, + "loss": 0.8401, + "num_input_tokens_seen": 71498980, + "step": 3352 + }, + { + "epoch": 0.403174412312872, + "flos": 19861360832880.0, + "grad_norm": 2.7618818221447143, + "learning_rate": 2.708705505209726e-06, + "loss": 0.6223, + "num_input_tokens_seen": 71519475, + "step": 3353 + }, + { + "epoch": 0.4032946552035111, + "flos": 15913449680160.0, + "grad_norm": 3.180687899290861, + "learning_rate": 2.7079770258737646e-06, + "loss": 0.897, + "num_input_tokens_seen": 71537105, + "step": 3354 + }, + { + "epoch": 0.4034148980941502, + "flos": 12648331580640.0, + "grad_norm": 3.47772238595527, + "learning_rate": 2.707248439138251e-06, + "loss": 0.7266, + "num_input_tokens_seen": 71553060, + "step": 3355 + }, + { + "epoch": 0.40353514098478926, + "flos": 16107842055000.0, + "grad_norm": 6.145223434272257, + "learning_rate": 2.7065197451137114e-06, + "loss": 0.6341, + "num_input_tokens_seen": 71574160, + "step": 3356 + }, + { + "epoch": 0.4036553838754284, + "flos": 10356500582280.0, + "grad_norm": 5.99001212091764, + "learning_rate": 2.7057909439106894e-06, + "loss": 0.6648, + "num_input_tokens_seen": 71591735, + "step": 3357 + }, + { + "epoch": 0.40377562676606743, + "flos": 18159889278600.0, + "grad_norm": 2.8800266937502275, + "learning_rate": 2.7050620356397417e-06, + "loss": 0.769, + "num_input_tokens_seen": 71610405, + "step": 3358 + }, + { + "epoch": 0.40389586965670654, + "flos": 17620113467880.0, + "grad_norm": 2.4528798752307486, + "learning_rate": 2.7043330204114437e-06, + "loss": 0.7024, + "num_input_tokens_seen": 71628835, + "step": 3359 + }, + { + "epoch": 0.40401611254734565, + "flos": 11950061298360.0, + "grad_norm": 3.6642200948501196, + "learning_rate": 2.7036038983363862e-06, + "loss": 0.8351, + "num_input_tokens_seen": 71645160, + "step": 3360 + }, + { + "epoch": 0.4041363554379847, + "flos": 17566078898880.0, + "grad_norm": 2.603894262689695, + "learning_rate": 2.702874669525177e-06, + "loss": 0.8266, + "num_input_tokens_seen": 71663360, + "step": 3361 + }, + { + "epoch": 0.4042565983286238, + "flos": 20832400076880.0, + "grad_norm": 3.3706730550250747, + "learning_rate": 2.7021453340884394e-06, + "loss": 0.6739, + "num_input_tokens_seen": 71680805, + "step": 3362 + }, + { + "epoch": 0.40437684121926293, + "flos": 12920245723440.0, + "grad_norm": 3.389029835200179, + "learning_rate": 2.7014158921368125e-06, + "loss": 0.7068, + "num_input_tokens_seen": 71698850, + "step": 3363 + }, + { + "epoch": 0.404497084109902, + "flos": 17588014759800.0, + "grad_norm": 2.711154896047315, + "learning_rate": 2.700686343780953e-06, + "loss": 0.8476, + "num_input_tokens_seen": 71718440, + "step": 3364 + }, + { + "epoch": 0.4046173270005411, + "flos": 16782783438120.0, + "grad_norm": 3.478986876328933, + "learning_rate": 2.699956689131532e-06, + "loss": 0.856, + "num_input_tokens_seen": 71738145, + "step": 3365 + }, + { + "epoch": 0.4047375698911802, + "flos": 15109009857480.0, + "grad_norm": 3.180942447087745, + "learning_rate": 2.699226928299238e-06, + "loss": 0.8229, + "num_input_tokens_seen": 71755885, + "step": 3366 + }, + { + "epoch": 0.40485781278181926, + "flos": 21210515420040.0, + "grad_norm": 29.58921539143052, + "learning_rate": 2.698497061394774e-06, + "loss": 0.7646, + "num_input_tokens_seen": 71774090, + "step": 3367 + }, + { + "epoch": 0.40497805567245837, + "flos": 16944443905680.0, + "grad_norm": 1.9989702434758148, + "learning_rate": 2.6977670885288627e-06, + "loss": 0.7937, + "num_input_tokens_seen": 71795210, + "step": 3368 + }, + { + "epoch": 0.4050982985630975, + "flos": 11868392075640.0, + "grad_norm": 2.5805528710767316, + "learning_rate": 2.6970370098122378e-06, + "loss": 0.7395, + "num_input_tokens_seen": 71811915, + "step": 3369 + }, + { + "epoch": 0.40521854145373654, + "flos": 25315496346120.0, + "grad_norm": 2.4548883599965166, + "learning_rate": 2.6963068253556535e-06, + "loss": 0.8434, + "num_input_tokens_seen": 71833020, + "step": 3370 + }, + { + "epoch": 0.40533878434437565, + "flos": 18558895704000.0, + "grad_norm": 2.8397490473807028, + "learning_rate": 2.6955765352698763e-06, + "loss": 0.8349, + "num_input_tokens_seen": 71852885, + "step": 3371 + }, + { + "epoch": 0.40545902723501476, + "flos": 10923024567840.0, + "grad_norm": 5.219311164334942, + "learning_rate": 2.6948461396656923e-06, + "loss": 0.7152, + "num_input_tokens_seen": 71870015, + "step": 3372 + }, + { + "epoch": 0.4055792701256538, + "flos": 18700583228160.0, + "grad_norm": 2.8541939912773535, + "learning_rate": 2.6941156386539013e-06, + "loss": 0.7295, + "num_input_tokens_seen": 71889685, + "step": 3373 + }, + { + "epoch": 0.4056995130162929, + "flos": 14299852700760.0, + "grad_norm": 3.0469100558880755, + "learning_rate": 2.6933850323453203e-06, + "loss": 0.7721, + "num_input_tokens_seen": 71907850, + "step": 3374 + }, + { + "epoch": 0.405819755906932, + "flos": 11220388827120.0, + "grad_norm": 3.2836036596663756, + "learning_rate": 2.6926543208507806e-06, + "loss": 0.7354, + "num_input_tokens_seen": 71926250, + "step": 3375 + }, + { + "epoch": 0.4059399987975711, + "flos": 15675692142840.0, + "grad_norm": 3.5860718014236443, + "learning_rate": 2.6919235042811316e-06, + "loss": 0.7675, + "num_input_tokens_seen": 71944755, + "step": 3376 + }, + { + "epoch": 0.4060602416882102, + "flos": 18727014803400.0, + "grad_norm": 3.736061258152817, + "learning_rate": 2.691192582747237e-06, + "loss": 0.737, + "num_input_tokens_seen": 71964105, + "step": 3377 + }, + { + "epoch": 0.40618048457884925, + "flos": 17401822314600.0, + "grad_norm": 3.678369500679984, + "learning_rate": 2.6904615563599765e-06, + "loss": 0.7209, + "num_input_tokens_seen": 71983625, + "step": 3378 + }, + { + "epoch": 0.40630072746948837, + "flos": 12867730832520.0, + "grad_norm": 3.1203047002046356, + "learning_rate": 2.6897304252302477e-06, + "loss": 0.8115, + "num_input_tokens_seen": 72000665, + "step": 3379 + }, + { + "epoch": 0.4064209703601275, + "flos": 44840718451440.0, + "grad_norm": 0.8222185062490582, + "learning_rate": 2.688999189468962e-06, + "loss": 0.5621, + "num_input_tokens_seen": 72056815, + "step": 3380 + }, + { + "epoch": 0.40654121325076653, + "flos": 17592763753800.0, + "grad_norm": 3.7314649919331884, + "learning_rate": 2.6882678491870464e-06, + "loss": 0.7335, + "num_input_tokens_seen": 72076970, + "step": 3381 + }, + { + "epoch": 0.40666145614140564, + "flos": 20049674495400.0, + "grad_norm": 7.315578497372624, + "learning_rate": 2.6875364044954453e-06, + "loss": 0.6936, + "num_input_tokens_seen": 72096920, + "step": 3382 + }, + { + "epoch": 0.40678169903204475, + "flos": 19185627950760.0, + "grad_norm": 3.0468113036711384, + "learning_rate": 2.6868048555051185e-06, + "loss": 0.8043, + "num_input_tokens_seen": 72118170, + "step": 3383 + }, + { + "epoch": 0.4069019419226838, + "flos": 20995770182280.0, + "grad_norm": 4.86241775874629, + "learning_rate": 2.686073202327041e-06, + "loss": 0.8307, + "num_input_tokens_seen": 72136890, + "step": 3384 + }, + { + "epoch": 0.4070221848133229, + "flos": 18484381632240.0, + "grad_norm": 3.4818789175980784, + "learning_rate": 2.6853414450722043e-06, + "loss": 0.7203, + "num_input_tokens_seen": 72156275, + "step": 3385 + }, + { + "epoch": 0.40714242770396203, + "flos": 13434191498160.0, + "grad_norm": 2.6556497232171066, + "learning_rate": 2.684609583851616e-06, + "loss": 0.831, + "num_input_tokens_seen": 72174170, + "step": 3386 + }, + { + "epoch": 0.4072626705946011, + "flos": 22184847299880.0, + "grad_norm": 3.5849247584833375, + "learning_rate": 2.683877618776297e-06, + "loss": 0.7903, + "num_input_tokens_seen": 72196145, + "step": 3387 + }, + { + "epoch": 0.4073829134852402, + "flos": 15972074943360.0, + "grad_norm": 5.166791278607954, + "learning_rate": 2.6831455499572876e-06, + "loss": 0.7257, + "num_input_tokens_seen": 72213800, + "step": 3388 + }, + { + "epoch": 0.40750315637587925, + "flos": 18506729072640.0, + "grad_norm": 2.5575733741832423, + "learning_rate": 2.682413377505641e-06, + "loss": 0.7563, + "num_input_tokens_seen": 72232325, + "step": 3389 + }, + { + "epoch": 0.40762339926651836, + "flos": 14401019967480.0, + "grad_norm": 3.5937576957288524, + "learning_rate": 2.6816811015324284e-06, + "loss": 0.751, + "num_input_tokens_seen": 72250095, + "step": 3390 + }, + { + "epoch": 0.40774364215715747, + "flos": 52694659121160.0, + "grad_norm": 0.8039508892451798, + "learning_rate": 2.6809487221487343e-06, + "loss": 0.6122, + "num_input_tokens_seen": 72309300, + "step": 3391 + }, + { + "epoch": 0.4078638850477965, + "flos": 10976235977880.0, + "grad_norm": 14.914688632744525, + "learning_rate": 2.6802162394656605e-06, + "loss": 0.7981, + "num_input_tokens_seen": 72325730, + "step": 3392 + }, + { + "epoch": 0.40798412793843564, + "flos": 17458611300120.0, + "grad_norm": 4.211877721447963, + "learning_rate": 2.679483653594324e-06, + "loss": 0.7045, + "num_input_tokens_seen": 72347220, + "step": 3393 + }, + { + "epoch": 0.40810437082907475, + "flos": 15402764881320.0, + "grad_norm": 3.164747526635993, + "learning_rate": 2.678750964645857e-06, + "loss": 0.7355, + "num_input_tokens_seen": 72366020, + "step": 3394 + }, + { + "epoch": 0.4082246137197138, + "flos": 8190875387640.0, + "grad_norm": 3.051731718638443, + "learning_rate": 2.6780181727314094e-06, + "loss": 0.8295, + "num_input_tokens_seen": 72380645, + "step": 3395 + }, + { + "epoch": 0.4083448566103529, + "flos": 13920755898840.0, + "grad_norm": 3.5162149570100754, + "learning_rate": 2.6772852779621435e-06, + "loss": 0.7637, + "num_input_tokens_seen": 72398225, + "step": 3396 + }, + { + "epoch": 0.408465099500992, + "flos": 17241934804800.0, + "grad_norm": 3.5615641385287455, + "learning_rate": 2.676552280449239e-06, + "loss": 0.8282, + "num_input_tokens_seen": 72417830, + "step": 3397 + }, + { + "epoch": 0.4085853423916311, + "flos": 9248966047560.0, + "grad_norm": 3.0562502870774084, + "learning_rate": 2.6758191803038917e-06, + "loss": 0.7211, + "num_input_tokens_seen": 72436045, + "step": 3398 + }, + { + "epoch": 0.4087055852822702, + "flos": 17836093444080.0, + "grad_norm": 1.9796707983599573, + "learning_rate": 2.6750859776373125e-06, + "loss": 0.8115, + "num_input_tokens_seen": 72455220, + "step": 3399 + }, + { + "epoch": 0.4088258281729093, + "flos": 48950353391640.0, + "grad_norm": 0.7820124529581185, + "learning_rate": 2.674352672560727e-06, + "loss": 0.6072, + "num_input_tokens_seen": 72516385, + "step": 3400 + }, + { + "epoch": 0.40894607106354836, + "flos": 14947032790320.0, + "grad_norm": 3.44072752201294, + "learning_rate": 2.673619265185377e-06, + "loss": 0.7493, + "num_input_tokens_seen": 72535945, + "step": 3401 + }, + { + "epoch": 0.40906631395418747, + "flos": 19834770957840.0, + "grad_norm": 2.119908460044274, + "learning_rate": 2.672885755622521e-06, + "loss": 0.7561, + "num_input_tokens_seen": 72558080, + "step": 3402 + }, + { + "epoch": 0.4091865568448266, + "flos": 18673170194160.0, + "grad_norm": 3.710029442765451, + "learning_rate": 2.67215214398343e-06, + "loss": 0.6784, + "num_input_tokens_seen": 72577815, + "step": 3403 + }, + { + "epoch": 0.40930679973546563, + "flos": 21021853497960.0, + "grad_norm": 7.958400572913824, + "learning_rate": 2.671418430379393e-06, + "loss": 0.754, + "num_input_tokens_seen": 72596220, + "step": 3404 + }, + { + "epoch": 0.40942704262610474, + "flos": 15270353725440.0, + "grad_norm": 2.183613840191569, + "learning_rate": 2.670684614921715e-06, + "loss": 0.8012, + "num_input_tokens_seen": 72614915, + "step": 3405 + }, + { + "epoch": 0.4095472855167438, + "flos": 15812219093520.0, + "grad_norm": 2.773578736913228, + "learning_rate": 2.6699506977217128e-06, + "loss": 0.6676, + "num_input_tokens_seen": 72634810, + "step": 3406 + }, + { + "epoch": 0.4096675284073829, + "flos": 20480019789840.0, + "grad_norm": 3.5505372352812543, + "learning_rate": 2.6692166788907233e-06, + "loss": 0.6725, + "num_input_tokens_seen": 72654725, + "step": 3407 + }, + { + "epoch": 0.409787771298022, + "flos": 14027083739040.0, + "grad_norm": 2.941922531311254, + "learning_rate": 2.6684825585400957e-06, + "loss": 0.7513, + "num_input_tokens_seen": 72673390, + "step": 3408 + }, + { + "epoch": 0.4099080141886611, + "flos": 43681460524800.0, + "grad_norm": 0.8850298354405367, + "learning_rate": 2.6677483367811947e-06, + "loss": 0.7125, + "num_input_tokens_seen": 72733150, + "step": 3409 + }, + { + "epoch": 0.4100282570793002, + "flos": 16025128053600.0, + "grad_norm": 4.3592651790882515, + "learning_rate": 2.6670140137254028e-06, + "loss": 0.7444, + "num_input_tokens_seen": 72752345, + "step": 3410 + }, + { + "epoch": 0.4101484999699393, + "flos": 13542577235760.0, + "grad_norm": 3.7236082254794085, + "learning_rate": 2.666279589484115e-06, + "loss": 0.853, + "num_input_tokens_seen": 72769965, + "step": 3411 + }, + { + "epoch": 0.41026874286057835, + "flos": 13943768198400.0, + "grad_norm": 4.1158458977740695, + "learning_rate": 2.6655450641687435e-06, + "loss": 0.7807, + "num_input_tokens_seen": 72787250, + "step": 3412 + }, + { + "epoch": 0.41038898575121746, + "flos": 22909960736880.0, + "grad_norm": 4.242232686007622, + "learning_rate": 2.664810437890715e-06, + "loss": 0.6774, + "num_input_tokens_seen": 72808640, + "step": 3413 + }, + { + "epoch": 0.41050922864185657, + "flos": 10437473285880.0, + "grad_norm": 2.8265498711940187, + "learning_rate": 2.6640757107614714e-06, + "loss": 0.7894, + "num_input_tokens_seen": 72826455, + "step": 3414 + }, + { + "epoch": 0.4106294715324956, + "flos": 22722818492880.0, + "grad_norm": 5.533884401725001, + "learning_rate": 2.6633408828924697e-06, + "loss": 0.6909, + "num_input_tokens_seen": 72845040, + "step": 3415 + }, + { + "epoch": 0.41074971442313474, + "flos": 17912507113440.0, + "grad_norm": 4.286826852165739, + "learning_rate": 2.662605954395185e-06, + "loss": 0.6856, + "num_input_tokens_seen": 72864720, + "step": 3416 + }, + { + "epoch": 0.41086995731377385, + "flos": 15756348246840.0, + "grad_norm": 2.9485868269601707, + "learning_rate": 2.6618709253811027e-06, + "loss": 0.8239, + "num_input_tokens_seen": 72883895, + "step": 3417 + }, + { + "epoch": 0.4109902002044129, + "flos": 15134364994080.0, + "grad_norm": 2.161424923067901, + "learning_rate": 2.6611357959617277e-06, + "loss": 0.8678, + "num_input_tokens_seen": 72903235, + "step": 3418 + }, + { + "epoch": 0.411110443095052, + "flos": 13266610618080.0, + "grad_norm": 4.38040241744353, + "learning_rate": 2.660400566248578e-06, + "loss": 0.888, + "num_input_tokens_seen": 72921080, + "step": 3419 + }, + { + "epoch": 0.41123068598569107, + "flos": 10599893592480.0, + "grad_norm": 3.0442383924615233, + "learning_rate": 2.6596652363531876e-06, + "loss": 0.6555, + "num_input_tokens_seen": 72936675, + "step": 3420 + }, + { + "epoch": 0.4113509288763302, + "flos": 15486207061800.0, + "grad_norm": 2.061570391533611, + "learning_rate": 2.6589298063871055e-06, + "loss": 0.7701, + "num_input_tokens_seen": 72956570, + "step": 3421 + }, + { + "epoch": 0.4114711717669693, + "flos": 13461636192120.0, + "grad_norm": 3.0665975134419323, + "learning_rate": 2.658194276461895e-06, + "loss": 0.6823, + "num_input_tokens_seen": 72974215, + "step": 3422 + }, + { + "epoch": 0.41159141465760835, + "flos": 19967435393400.0, + "grad_norm": 3.3981714604920015, + "learning_rate": 2.6574586466891368e-06, + "loss": 0.6597, + "num_input_tokens_seen": 72994410, + "step": 3423 + }, + { + "epoch": 0.41171165754824746, + "flos": 14618994521160.0, + "grad_norm": 2.5809638836952384, + "learning_rate": 2.6567229171804247e-06, + "loss": 0.6158, + "num_input_tokens_seen": 73012015, + "step": 3424 + }, + { + "epoch": 0.41183190043888657, + "flos": 13137555417960.0, + "grad_norm": 3.1362777152876093, + "learning_rate": 2.655987088047368e-06, + "loss": 0.8424, + "num_input_tokens_seen": 73030080, + "step": 3425 + }, + { + "epoch": 0.4119521433295256, + "flos": 19916028601080.0, + "grad_norm": 5.535668973253848, + "learning_rate": 2.6552511594015912e-06, + "loss": 0.7667, + "num_input_tokens_seen": 73050190, + "step": 3426 + }, + { + "epoch": 0.41207238622016473, + "flos": 11004472170840.0, + "grad_norm": 6.9668195984525685, + "learning_rate": 2.654515131354735e-06, + "loss": 0.8431, + "num_input_tokens_seen": 73068175, + "step": 3427 + }, + { + "epoch": 0.41219262911080384, + "flos": 19834106098680.0, + "grad_norm": 3.746804465297312, + "learning_rate": 2.653779004018453e-06, + "loss": 0.8423, + "num_input_tokens_seen": 73088460, + "step": 3428 + }, + { + "epoch": 0.4123128720014429, + "flos": 18083443949280.0, + "grad_norm": 2.265377794130679, + "learning_rate": 2.653042777504417e-06, + "loss": 0.803, + "num_input_tokens_seen": 73110770, + "step": 3429 + }, + { + "epoch": 0.412433114892082, + "flos": 19235293445280.0, + "grad_norm": 2.948674152946355, + "learning_rate": 2.6523064519243105e-06, + "loss": 0.7829, + "num_input_tokens_seen": 73130060, + "step": 3430 + }, + { + "epoch": 0.4125533577827211, + "flos": 15621752553720.0, + "grad_norm": 4.452325260561056, + "learning_rate": 2.6515700273898333e-06, + "loss": 0.7663, + "num_input_tokens_seen": 73147655, + "step": 3431 + }, + { + "epoch": 0.4126736006733602, + "flos": 19104876866880.0, + "grad_norm": 4.007560261007371, + "learning_rate": 2.6508335040127018e-06, + "loss": 0.6704, + "num_input_tokens_seen": 73167070, + "step": 3432 + }, + { + "epoch": 0.4127938435639993, + "flos": 18807797547240.0, + "grad_norm": 2.8705491152749096, + "learning_rate": 2.6500968819046446e-06, + "loss": 0.7532, + "num_input_tokens_seen": 73187090, + "step": 3433 + }, + { + "epoch": 0.4129140864546384, + "flos": 13130146987320.0, + "grad_norm": 4.061265258948255, + "learning_rate": 2.649360161177408e-06, + "loss": 0.5657, + "num_input_tokens_seen": 73201870, + "step": 3434 + }, + { + "epoch": 0.41303432934527745, + "flos": 17377132037160.0, + "grad_norm": 4.690733544544295, + "learning_rate": 2.6486233419427504e-06, + "loss": 0.7119, + "num_input_tokens_seen": 73221405, + "step": 3435 + }, + { + "epoch": 0.41315457223591656, + "flos": 14433340295280.0, + "grad_norm": 2.916814328043488, + "learning_rate": 2.6478864243124484e-06, + "loss": 0.7373, + "num_input_tokens_seen": 73240790, + "step": 3436 + }, + { + "epoch": 0.4132748151265556, + "flos": 15297830079360.0, + "grad_norm": 2.2913667193546123, + "learning_rate": 2.6471494083982903e-06, + "loss": 0.8384, + "num_input_tokens_seen": 73259895, + "step": 3437 + }, + { + "epoch": 0.4133950580171947, + "flos": 23667204541920.0, + "grad_norm": 2.4260607572213626, + "learning_rate": 2.6464122943120818e-06, + "loss": 0.7361, + "num_input_tokens_seen": 73279840, + "step": 3438 + }, + { + "epoch": 0.41351530090783384, + "flos": 17050835065800.0, + "grad_norm": 6.991317541463567, + "learning_rate": 2.645675082165642e-06, + "loss": 0.8048, + "num_input_tokens_seen": 73295770, + "step": 3439 + }, + { + "epoch": 0.4136355437984729, + "flos": 18754142897760.0, + "grad_norm": 3.843972300482276, + "learning_rate": 2.644937772070806e-06, + "loss": 0.7451, + "num_input_tokens_seen": 73313935, + "step": 3440 + }, + { + "epoch": 0.413755786689112, + "flos": 14487659803920.0, + "grad_norm": 4.374859616888348, + "learning_rate": 2.6442003641394225e-06, + "loss": 0.8139, + "num_input_tokens_seen": 73331250, + "step": 3441 + }, + { + "epoch": 0.4138760295797511, + "flos": 19699098826080.0, + "grad_norm": 2.0541369494699553, + "learning_rate": 2.643462858483356e-06, + "loss": 0.8212, + "num_input_tokens_seen": 73351255, + "step": 3442 + }, + { + "epoch": 0.41399627247039017, + "flos": 11949744698760.0, + "grad_norm": 3.01653564618214, + "learning_rate": 2.6427252552144856e-06, + "loss": 0.7249, + "num_input_tokens_seen": 73369625, + "step": 3443 + }, + { + "epoch": 0.4141165153610293, + "flos": 16783100037720.0, + "grad_norm": 2.539872260448496, + "learning_rate": 2.6419875544447044e-06, + "loss": 0.7351, + "num_input_tokens_seen": 73390745, + "step": 3444 + }, + { + "epoch": 0.4142367582516684, + "flos": 18457411837680.0, + "grad_norm": 2.296047554706592, + "learning_rate": 2.6412497562859218e-06, + "loss": 0.6953, + "num_input_tokens_seen": 73411745, + "step": 3445 + }, + { + "epoch": 0.41435700114230745, + "flos": 15865715443200.0, + "grad_norm": 3.918497158332311, + "learning_rate": 2.6405118608500617e-06, + "loss": 0.7463, + "num_input_tokens_seen": 73430290, + "step": 3446 + }, + { + "epoch": 0.41447724403294656, + "flos": 19051380517200.0, + "grad_norm": 3.1596099796317696, + "learning_rate": 2.6397738682490613e-06, + "loss": 0.789, + "num_input_tokens_seen": 73450910, + "step": 3447 + }, + { + "epoch": 0.41459748692358567, + "flos": 13326533939640.0, + "grad_norm": 2.6300647557920365, + "learning_rate": 2.6390357785948734e-06, + "loss": 0.7389, + "num_input_tokens_seen": 73467745, + "step": 3448 + }, + { + "epoch": 0.4147177298142247, + "flos": 17699154913920.0, + "grad_norm": 2.6366504801636337, + "learning_rate": 2.6382975919994667e-06, + "loss": 0.778, + "num_input_tokens_seen": 73488040, + "step": 3449 + }, + { + "epoch": 0.41483797270486383, + "flos": 14677239864840.0, + "grad_norm": 2.235886282770754, + "learning_rate": 2.637559308574822e-06, + "loss": 0.7047, + "num_input_tokens_seen": 73507505, + "step": 3450 + }, + { + "epoch": 0.4149582155955029, + "flos": 22075131843960.0, + "grad_norm": 3.09701904824114, + "learning_rate": 2.6368209284329376e-06, + "loss": 0.7026, + "num_input_tokens_seen": 73527855, + "step": 3451 + }, + { + "epoch": 0.415078458486142, + "flos": 12220012523640.0, + "grad_norm": 4.715526768373505, + "learning_rate": 2.636082451685825e-06, + "loss": 0.7343, + "num_input_tokens_seen": 73545775, + "step": 3452 + }, + { + "epoch": 0.4151987013767811, + "flos": 19080186589440.0, + "grad_norm": 2.244246862966192, + "learning_rate": 2.6353438784455094e-06, + "loss": 0.8409, + "num_input_tokens_seen": 73568780, + "step": 3453 + }, + { + "epoch": 0.41531894426742016, + "flos": 18024692046240.0, + "grad_norm": 3.818900442203669, + "learning_rate": 2.6346052088240326e-06, + "loss": 0.6882, + "num_input_tokens_seen": 73588020, + "step": 3454 + }, + { + "epoch": 0.4154391871580593, + "flos": 10896213073080.0, + "grad_norm": 2.4952428379259897, + "learning_rate": 2.63386644293345e-06, + "loss": 0.7545, + "num_input_tokens_seen": 73604085, + "step": 3455 + }, + { + "epoch": 0.4155594300486984, + "flos": 10652978362680.0, + "grad_norm": 4.708822354377779, + "learning_rate": 2.633127580885833e-06, + "loss": 0.817, + "num_input_tokens_seen": 73618305, + "step": 3456 + }, + { + "epoch": 0.41567967293933744, + "flos": 21643108571640.0, + "grad_norm": 4.535479025644968, + "learning_rate": 2.632388622793265e-06, + "loss": 0.6184, + "num_input_tokens_seen": 73637180, + "step": 3457 + }, + { + "epoch": 0.41579991582997655, + "flos": 14049937738800.0, + "grad_norm": 3.979150076360553, + "learning_rate": 2.6316495687678457e-06, + "loss": 0.6702, + "num_input_tokens_seen": 73655550, + "step": 3458 + }, + { + "epoch": 0.41592015872061566, + "flos": 17916686228160.0, + "grad_norm": 2.734355991842414, + "learning_rate": 2.6309104189216887e-06, + "loss": 0.7393, + "num_input_tokens_seen": 73672835, + "step": 3459 + }, + { + "epoch": 0.4160404016112547, + "flos": 15188304583200.0, + "grad_norm": 3.064576982240473, + "learning_rate": 2.630171173366923e-06, + "loss": 0.7284, + "num_input_tokens_seen": 73688355, + "step": 3460 + }, + { + "epoch": 0.41616064450189383, + "flos": 10112062793400.0, + "grad_norm": 3.831764204109323, + "learning_rate": 2.629431832215691e-06, + "loss": 0.7204, + "num_input_tokens_seen": 73702880, + "step": 3461 + }, + { + "epoch": 0.41628088739253294, + "flos": 14622477116760.0, + "grad_norm": 3.4643232262087515, + "learning_rate": 2.628692395580151e-06, + "loss": 0.8474, + "num_input_tokens_seen": 73722690, + "step": 3462 + }, + { + "epoch": 0.416401130283172, + "flos": 21400000501080.0, + "grad_norm": 2.6832078770123244, + "learning_rate": 2.6279528635724747e-06, + "loss": 0.7784, + "num_input_tokens_seen": 73742565, + "step": 3463 + }, + { + "epoch": 0.4165213731738111, + "flos": 11836008427920.0, + "grad_norm": 4.816827982980807, + "learning_rate": 2.627213236304848e-06, + "loss": 0.7586, + "num_input_tokens_seen": 73759085, + "step": 3464 + }, + { + "epoch": 0.4166416160644502, + "flos": 24801740531160.0, + "grad_norm": 2.746161556356366, + "learning_rate": 2.626473513889472e-06, + "loss": 0.6929, + "num_input_tokens_seen": 73781185, + "step": 3465 + }, + { + "epoch": 0.41676185895508927, + "flos": 15292701165840.0, + "grad_norm": 4.542198266697981, + "learning_rate": 2.625733696438562e-06, + "loss": 0.8078, + "num_input_tokens_seen": 73798410, + "step": 3466 + }, + { + "epoch": 0.4168821018457284, + "flos": 13434982997160.0, + "grad_norm": 3.0566421850035956, + "learning_rate": 2.6249937840643476e-06, + "loss": 0.738, + "num_input_tokens_seen": 73816435, + "step": 3467 + }, + { + "epoch": 0.41700234473636744, + "flos": 13651342892880.0, + "grad_norm": 2.373425614388134, + "learning_rate": 2.6242537768790733e-06, + "loss": 0.6538, + "num_input_tokens_seen": 73835310, + "step": 3468 + }, + { + "epoch": 0.41712258762700655, + "flos": 22780367317440.0, + "grad_norm": 2.6366294111022235, + "learning_rate": 2.6235136749949975e-06, + "loss": 0.6746, + "num_input_tokens_seen": 73858480, + "step": 3469 + }, + { + "epoch": 0.41724283051764566, + "flos": 26391913631520.0, + "grad_norm": 2.923138513572569, + "learning_rate": 2.6227734785243924e-06, + "loss": 0.5931, + "num_input_tokens_seen": 73878160, + "step": 3470 + }, + { + "epoch": 0.4173630734082847, + "flos": 18561143561160.0, + "grad_norm": 2.515428997483697, + "learning_rate": 2.6220331875795466e-06, + "loss": 0.7793, + "num_input_tokens_seen": 73897230, + "step": 3471 + }, + { + "epoch": 0.4174833162989238, + "flos": 19561780376400.0, + "grad_norm": 2.365331948596387, + "learning_rate": 2.62129280227276e-06, + "loss": 0.744, + "num_input_tokens_seen": 73916950, + "step": 3472 + }, + { + "epoch": 0.41760355918956293, + "flos": 55314460577400.0, + "grad_norm": 3.798456812267725, + "learning_rate": 2.62055232271635e-06, + "loss": 0.6621, + "num_input_tokens_seen": 73943855, + "step": 3473 + }, + { + "epoch": 0.417723802080202, + "flos": 10408825513440.0, + "grad_norm": 3.2545871640405335, + "learning_rate": 2.619811749022646e-06, + "loss": 0.8712, + "num_input_tokens_seen": 73958885, + "step": 3474 + }, + { + "epoch": 0.4178440449708411, + "flos": 10650002326440.0, + "grad_norm": 5.016310220181529, + "learning_rate": 2.6190710813039917e-06, + "loss": 0.6997, + "num_input_tokens_seen": 73971730, + "step": 3475 + }, + { + "epoch": 0.4179642878614802, + "flos": 15730043311440.0, + "grad_norm": 4.940575330027345, + "learning_rate": 2.618330319672747e-06, + "loss": 0.8078, + "num_input_tokens_seen": 73990870, + "step": 3476 + }, + { + "epoch": 0.41808453075211927, + "flos": 13461446232360.0, + "grad_norm": 3.3449382631452154, + "learning_rate": 2.617589464241284e-06, + "loss": 0.9034, + "num_input_tokens_seen": 74004990, + "step": 3477 + }, + { + "epoch": 0.4182047736427584, + "flos": 14837000734800.0, + "grad_norm": 2.9292929421813176, + "learning_rate": 2.6168485151219914e-06, + "loss": 0.7179, + "num_input_tokens_seen": 74024330, + "step": 3478 + }, + { + "epoch": 0.4183250165333975, + "flos": 13782645950160.0, + "grad_norm": 3.44284391304554, + "learning_rate": 2.616107472427269e-06, + "loss": 0.6996, + "num_input_tokens_seen": 74038745, + "step": 3479 + }, + { + "epoch": 0.41844525942403654, + "flos": 12940725226200.0, + "grad_norm": 4.84154061412155, + "learning_rate": 2.615366336269533e-06, + "loss": 0.7479, + "num_input_tokens_seen": 74052130, + "step": 3480 + }, + { + "epoch": 0.41856550231467565, + "flos": 13400699751840.0, + "grad_norm": 4.9385925243145765, + "learning_rate": 2.6146251067612126e-06, + "loss": 0.7724, + "num_input_tokens_seen": 74067325, + "step": 3481 + }, + { + "epoch": 0.41868574520531476, + "flos": 16672909682400.0, + "grad_norm": 2.5937647072248695, + "learning_rate": 2.6138837840147525e-06, + "loss": 0.806, + "num_input_tokens_seen": 74086080, + "step": 3482 + }, + { + "epoch": 0.4188059880959538, + "flos": 9951478764480.0, + "grad_norm": 4.481686094976561, + "learning_rate": 2.6131423681426103e-06, + "loss": 0.7544, + "num_input_tokens_seen": 74101715, + "step": 3483 + }, + { + "epoch": 0.41892623098659293, + "flos": 27801529759560.0, + "grad_norm": 2.399728831230717, + "learning_rate": 2.6124008592572587e-06, + "loss": 0.7103, + "num_input_tokens_seen": 74125420, + "step": 3484 + }, + { + "epoch": 0.419046473877232, + "flos": 17026366408080.0, + "grad_norm": 5.131590722465647, + "learning_rate": 2.6116592574711835e-06, + "loss": 0.7898, + "num_input_tokens_seen": 74143440, + "step": 3485 + }, + { + "epoch": 0.4191667167678711, + "flos": 15162949446600.0, + "grad_norm": 2.7262329951774373, + "learning_rate": 2.6109175628968853e-06, + "loss": 0.8164, + "num_input_tokens_seen": 74162925, + "step": 3486 + }, + { + "epoch": 0.4192869596585102, + "flos": 17268651319680.0, + "grad_norm": 5.558605759715593, + "learning_rate": 2.610175775646878e-06, + "loss": 0.8174, + "num_input_tokens_seen": 74181225, + "step": 3487 + }, + { + "epoch": 0.41940720254914926, + "flos": 18374792816160.0, + "grad_norm": 4.773328015059877, + "learning_rate": 2.6094338958336907e-06, + "loss": 0.7278, + "num_input_tokens_seen": 74199615, + "step": 3488 + }, + { + "epoch": 0.41952744543978837, + "flos": 11323455691440.0, + "grad_norm": 2.8288462019612495, + "learning_rate": 2.608691923569867e-06, + "loss": 0.7973, + "num_input_tokens_seen": 74216210, + "step": 3489 + }, + { + "epoch": 0.4196476883304275, + "flos": 18051946780440.0, + "grad_norm": 3.8352634435932424, + "learning_rate": 2.6079498589679616e-06, + "loss": 0.7479, + "num_input_tokens_seen": 74237020, + "step": 3490 + }, + { + "epoch": 0.41976793122106654, + "flos": 17967428161320.0, + "grad_norm": 4.0012996469579845, + "learning_rate": 2.6072077021405465e-06, + "loss": 0.7379, + "num_input_tokens_seen": 74255575, + "step": 3491 + }, + { + "epoch": 0.41988817411170565, + "flos": 15483832564800.0, + "grad_norm": 3.213556785280017, + "learning_rate": 2.6064654532002054e-06, + "loss": 0.6787, + "num_input_tokens_seen": 74274305, + "step": 3492 + }, + { + "epoch": 0.42000841700234476, + "flos": 23235751148880.0, + "grad_norm": 2.9880836970034577, + "learning_rate": 2.6057231122595375e-06, + "loss": 0.7412, + "num_input_tokens_seen": 74295335, + "step": 3493 + }, + { + "epoch": 0.4201286598929838, + "flos": 15561987531960.0, + "grad_norm": 2.8460846424611517, + "learning_rate": 2.604980679431154e-06, + "loss": 0.7203, + "num_input_tokens_seen": 74313295, + "step": 3494 + }, + { + "epoch": 0.4202489027836229, + "flos": 13538113181400.0, + "grad_norm": 3.146575452834923, + "learning_rate": 2.604238154827684e-06, + "loss": 0.7445, + "num_input_tokens_seen": 74329640, + "step": 3495 + }, + { + "epoch": 0.42036914567426203, + "flos": 14109481140840.0, + "grad_norm": 4.164659525314404, + "learning_rate": 2.6034955385617656e-06, + "loss": 0.705, + "num_input_tokens_seen": 74347690, + "step": 3496 + }, + { + "epoch": 0.4204893885649011, + "flos": 50026517397360.0, + "grad_norm": 0.7635878885043024, + "learning_rate": 2.6027528307460544e-06, + "loss": 0.6551, + "num_input_tokens_seen": 74411415, + "step": 3497 + }, + { + "epoch": 0.4206096314555402, + "flos": 16027185951000.0, + "grad_norm": 2.221375000570133, + "learning_rate": 2.602010031493217e-06, + "loss": 0.8421, + "num_input_tokens_seen": 74429365, + "step": 3498 + }, + { + "epoch": 0.42072987434617926, + "flos": 21480941544720.0, + "grad_norm": 15.299326184962107, + "learning_rate": 2.6012671409159367e-06, + "loss": 0.8452, + "num_input_tokens_seen": 74450420, + "step": 3499 + }, + { + "epoch": 0.42085011723681837, + "flos": 20239634475840.0, + "grad_norm": 2.5871118374167277, + "learning_rate": 2.6005241591269097e-06, + "loss": 0.8141, + "num_input_tokens_seen": 74469510, + "step": 3500 + }, + { + "epoch": 0.4209703601274575, + "flos": 20400756724080.0, + "grad_norm": 1.848738371610664, + "learning_rate": 2.5997810862388454e-06, + "loss": 0.787, + "num_input_tokens_seen": 74489070, + "step": 3501 + }, + { + "epoch": 0.42109060301809653, + "flos": 20183668649280.0, + "grad_norm": 2.733643962685665, + "learning_rate": 2.599037922364467e-06, + "loss": 0.7453, + "num_input_tokens_seen": 74507690, + "step": 3502 + }, + { + "epoch": 0.42121084590873564, + "flos": 21507594739680.0, + "grad_norm": 3.183111722998247, + "learning_rate": 2.5982946676165112e-06, + "loss": 0.7444, + "num_input_tokens_seen": 74527180, + "step": 3503 + }, + { + "epoch": 0.42133108879937475, + "flos": 49696959450120.0, + "grad_norm": 0.7604456116554164, + "learning_rate": 2.5975513221077313e-06, + "loss": 0.594, + "num_input_tokens_seen": 74590870, + "step": 3504 + }, + { + "epoch": 0.4214513316900138, + "flos": 16912281877680.0, + "grad_norm": 2.793089451882675, + "learning_rate": 2.5968078859508897e-06, + "loss": 0.8656, + "num_input_tokens_seen": 74607790, + "step": 3505 + }, + { + "epoch": 0.4215715745806529, + "flos": 11162301783240.0, + "grad_norm": 3.5130622169464516, + "learning_rate": 2.5960643592587673e-06, + "loss": 0.7787, + "num_input_tokens_seen": 74624920, + "step": 3506 + }, + { + "epoch": 0.42169181747129203, + "flos": 16188814758600.0, + "grad_norm": 3.782445040931976, + "learning_rate": 2.5953207421441553e-06, + "loss": 0.8082, + "num_input_tokens_seen": 74643240, + "step": 3507 + }, + { + "epoch": 0.4218120603619311, + "flos": 16560376490040.0, + "grad_norm": 3.0324658973581444, + "learning_rate": 2.5945770347198603e-06, + "loss": 0.7273, + "num_input_tokens_seen": 74661115, + "step": 3508 + }, + { + "epoch": 0.4219323032525702, + "flos": 14379780625680.0, + "grad_norm": 2.6962180565227087, + "learning_rate": 2.593833237098701e-06, + "loss": 0.8168, + "num_input_tokens_seen": 74678435, + "step": 3509 + }, + { + "epoch": 0.4220525461432093, + "flos": 22156357827240.0, + "grad_norm": 2.5374316802951618, + "learning_rate": 2.593089349393512e-06, + "loss": 0.6162, + "num_input_tokens_seen": 74698645, + "step": 3510 + }, + { + "epoch": 0.42217278903384836, + "flos": 17808522110280.0, + "grad_norm": 3.076542326404307, + "learning_rate": 2.592345371717141e-06, + "loss": 0.8198, + "num_input_tokens_seen": 74717895, + "step": 3511 + }, + { + "epoch": 0.42229303192448747, + "flos": 12462772334640.0, + "grad_norm": 3.583757124096403, + "learning_rate": 2.591601304182448e-06, + "loss": 0.6902, + "num_input_tokens_seen": 74735585, + "step": 3512 + }, + { + "epoch": 0.4224132748151266, + "flos": 16675315839360.0, + "grad_norm": 2.7214552475744997, + "learning_rate": 2.5908571469023067e-06, + "loss": 0.7789, + "num_input_tokens_seen": 74754790, + "step": 3513 + }, + { + "epoch": 0.42253351770576564, + "flos": 13000078668480.0, + "grad_norm": 4.897507964162305, + "learning_rate": 2.5901128999896067e-06, + "loss": 0.7422, + "num_input_tokens_seen": 74769940, + "step": 3514 + }, + { + "epoch": 0.42265376059640475, + "flos": 20914259259360.0, + "grad_norm": 2.0511904353687975, + "learning_rate": 2.5893685635572487e-06, + "loss": 0.6718, + "num_input_tokens_seen": 74790510, + "step": 3515 + }, + { + "epoch": 0.4227740034870438, + "flos": 11841770540640.0, + "grad_norm": 2.9562174434361177, + "learning_rate": 2.5886241377181483e-06, + "loss": 0.6759, + "num_input_tokens_seen": 74809100, + "step": 3516 + }, + { + "epoch": 0.4228942463776829, + "flos": 18532844048280.0, + "grad_norm": 2.2888515752798146, + "learning_rate": 2.587879622585234e-06, + "loss": 0.7905, + "num_input_tokens_seen": 74827420, + "step": 3517 + }, + { + "epoch": 0.423014489268322, + "flos": 19347763317720.0, + "grad_norm": 3.074581061877323, + "learning_rate": 2.5871350182714486e-06, + "loss": 0.7365, + "num_input_tokens_seen": 74848020, + "step": 3518 + }, + { + "epoch": 0.4231347321589611, + "flos": 12596988108240.0, + "grad_norm": 3.5878875682570674, + "learning_rate": 2.586390324889748e-06, + "loss": 0.7823, + "num_input_tokens_seen": 74863640, + "step": 3519 + }, + { + "epoch": 0.4232549750496002, + "flos": 16833968610720.0, + "grad_norm": 2.2632727678460642, + "learning_rate": 2.5856455425531003e-06, + "loss": 0.6526, + "num_input_tokens_seen": 74884835, + "step": 3520 + }, + { + "epoch": 0.4233752179402393, + "flos": 15537898793760.0, + "grad_norm": 2.4773600644837885, + "learning_rate": 2.5849006713744902e-06, + "loss": 0.7948, + "num_input_tokens_seen": 74903350, + "step": 3521 + }, + { + "epoch": 0.42349546083087836, + "flos": 15135726372360.0, + "grad_norm": 3.6137705538629312, + "learning_rate": 2.5841557114669135e-06, + "loss": 0.7131, + "num_input_tokens_seen": 74919930, + "step": 3522 + }, + { + "epoch": 0.42361570372151747, + "flos": 13566919253640.0, + "grad_norm": 4.231829451717407, + "learning_rate": 2.58341066294338e-06, + "loss": 0.6413, + "num_input_tokens_seen": 74936315, + "step": 3523 + }, + { + "epoch": 0.4237359466121566, + "flos": 15324546594240.0, + "grad_norm": 4.046823862113556, + "learning_rate": 2.5826655259169124e-06, + "loss": 0.8358, + "num_input_tokens_seen": 74954690, + "step": 3524 + }, + { + "epoch": 0.42385618950279563, + "flos": 13162245695400.0, + "grad_norm": 2.7319327646753657, + "learning_rate": 2.5819203005005475e-06, + "loss": 0.8803, + "num_input_tokens_seen": 74971745, + "step": 3525 + }, + { + "epoch": 0.42397643239343474, + "flos": 17403911871960.0, + "grad_norm": 2.6895376066127117, + "learning_rate": 2.581174986807336e-06, + "loss": 0.7735, + "num_input_tokens_seen": 74991700, + "step": 3526 + }, + { + "epoch": 0.42409667528407385, + "flos": 12057655536960.0, + "grad_norm": 2.8343339910768512, + "learning_rate": 2.580429584950341e-06, + "loss": 0.9014, + "num_input_tokens_seen": 75007170, + "step": 3527 + }, + { + "epoch": 0.4242169181747129, + "flos": 11679540193800.0, + "grad_norm": 2.6263984406547087, + "learning_rate": 2.5796840950426397e-06, + "loss": 0.6375, + "num_input_tokens_seen": 75023975, + "step": 3528 + }, + { + "epoch": 0.424337161065352, + "flos": 14676669985560.0, + "grad_norm": 3.4206709463503744, + "learning_rate": 2.578938517197322e-06, + "loss": 0.6411, + "num_input_tokens_seen": 75041790, + "step": 3529 + }, + { + "epoch": 0.4244574039559911, + "flos": 17484726275760.0, + "grad_norm": 3.0788412277452797, + "learning_rate": 2.5781928515274916e-06, + "loss": 0.6062, + "num_input_tokens_seen": 75060230, + "step": 3530 + }, + { + "epoch": 0.4245776468466302, + "flos": 12813031404360.0, + "grad_norm": 3.0057818010518775, + "learning_rate": 2.577447098146265e-06, + "loss": 0.6641, + "num_input_tokens_seen": 75077125, + "step": 3531 + }, + { + "epoch": 0.4246978897372693, + "flos": 20368563036120.0, + "grad_norm": 2.331748242354362, + "learning_rate": 2.5767012571667724e-06, + "loss": 0.7698, + "num_input_tokens_seen": 75096325, + "step": 3532 + }, + { + "epoch": 0.42481813262790835, + "flos": 11355744359280.0, + "grad_norm": 12.145211589378805, + "learning_rate": 2.5759553287021587e-06, + "loss": 0.6617, + "num_input_tokens_seen": 75114375, + "step": 3533 + }, + { + "epoch": 0.42493837551854746, + "flos": 17538507565080.0, + "grad_norm": 3.1042882054919367, + "learning_rate": 2.5752093128655786e-06, + "loss": 0.7513, + "num_input_tokens_seen": 75132340, + "step": 3534 + }, + { + "epoch": 0.4250586184091866, + "flos": 15216477456240.0, + "grad_norm": 2.486454986668031, + "learning_rate": 2.574463209770204e-06, + "loss": 0.7161, + "num_input_tokens_seen": 75151375, + "step": 3535 + }, + { + "epoch": 0.42517886129982563, + "flos": 22289465502240.0, + "grad_norm": 2.633344194735891, + "learning_rate": 2.5737170195292165e-06, + "loss": 0.7764, + "num_input_tokens_seen": 75174430, + "step": 3536 + }, + { + "epoch": 0.42529910419046474, + "flos": 14673693949320.0, + "grad_norm": 2.7529153109659195, + "learning_rate": 2.572970742255814e-06, + "loss": 0.7617, + "num_input_tokens_seen": 75192640, + "step": 3537 + }, + { + "epoch": 0.42541934708110385, + "flos": 16561896168120.0, + "grad_norm": 3.7669005706764804, + "learning_rate": 2.5722243780632046e-06, + "loss": 0.8013, + "num_input_tokens_seen": 75210625, + "step": 3538 + }, + { + "epoch": 0.4255395899717429, + "flos": 48811072024440.0, + "grad_norm": 0.8352680556107817, + "learning_rate": 2.5714779270646125e-06, + "loss": 0.6511, + "num_input_tokens_seen": 75271115, + "step": 3539 + }, + { + "epoch": 0.425659832862382, + "flos": 13083964088400.0, + "grad_norm": 6.955736814147409, + "learning_rate": 2.5707313893732735e-06, + "loss": 0.7516, + "num_input_tokens_seen": 75289375, + "step": 3540 + }, + { + "epoch": 0.4257800757530211, + "flos": 17591180755800.0, + "grad_norm": 2.620820487455373, + "learning_rate": 2.5699847651024364e-06, + "loss": 0.7505, + "num_input_tokens_seen": 75309735, + "step": 3541 + }, + { + "epoch": 0.4259003186436602, + "flos": 17349782323080.0, + "grad_norm": 3.0926844756012533, + "learning_rate": 2.5692380543653627e-06, + "loss": 0.7628, + "num_input_tokens_seen": 75327610, + "step": 3542 + }, + { + "epoch": 0.4260205615342993, + "flos": 11105259518040.0, + "grad_norm": 3.7158970507844815, + "learning_rate": 2.5684912572753293e-06, + "loss": 0.6773, + "num_input_tokens_seen": 75343005, + "step": 3543 + }, + { + "epoch": 0.4261408044249384, + "flos": 22507598355720.0, + "grad_norm": 3.5714191082420315, + "learning_rate": 2.5677443739456245e-06, + "loss": 0.8337, + "num_input_tokens_seen": 75364385, + "step": 3544 + }, + { + "epoch": 0.42626104731557746, + "flos": 17025511589160.0, + "grad_norm": 3.298280822993867, + "learning_rate": 2.5669974044895495e-06, + "loss": 0.775, + "num_input_tokens_seen": 75380500, + "step": 3545 + }, + { + "epoch": 0.42638129020621657, + "flos": 18969552994680.0, + "grad_norm": 4.137422294693751, + "learning_rate": 2.5662503490204187e-06, + "loss": 0.7724, + "num_input_tokens_seen": 75400385, + "step": 3546 + }, + { + "epoch": 0.4265015330968556, + "flos": 19426234884480.0, + "grad_norm": 2.488261511707918, + "learning_rate": 2.5655032076515603e-06, + "loss": 0.7471, + "num_input_tokens_seen": 75419430, + "step": 3547 + }, + { + "epoch": 0.42662177598749473, + "flos": 17863063238640.0, + "grad_norm": 2.6989883788269577, + "learning_rate": 2.5647559804963155e-06, + "loss": 0.8008, + "num_input_tokens_seen": 75439080, + "step": 3548 + }, + { + "epoch": 0.42674201887813384, + "flos": 16944665525400.0, + "grad_norm": 5.775164396249767, + "learning_rate": 2.5640086676680364e-06, + "loss": 0.7766, + "num_input_tokens_seen": 75460295, + "step": 3549 + }, + { + "epoch": 0.4268622617687729, + "flos": 15864955604160.0, + "grad_norm": 3.827262687875871, + "learning_rate": 2.5632612692800923e-06, + "loss": 0.7833, + "num_input_tokens_seen": 75479080, + "step": 3550 + }, + { + "epoch": 0.426982504659412, + "flos": 17160360561960.0, + "grad_norm": 3.8553134071415203, + "learning_rate": 2.5625137854458603e-06, + "loss": 0.7287, + "num_input_tokens_seen": 75497815, + "step": 3551 + }, + { + "epoch": 0.4271027475500511, + "flos": 13812686760840.0, + "grad_norm": 2.490387939505289, + "learning_rate": 2.561766216278735e-06, + "loss": 0.7857, + "num_input_tokens_seen": 75515130, + "step": 3552 + }, + { + "epoch": 0.4272229904406902, + "flos": 19699193805960.0, + "grad_norm": 2.4920979546510953, + "learning_rate": 2.561018561892121e-06, + "loss": 0.7946, + "num_input_tokens_seen": 75533990, + "step": 3553 + }, + { + "epoch": 0.4273432333313293, + "flos": 17538760844760.0, + "grad_norm": 2.163547495903154, + "learning_rate": 2.5602708223994363e-06, + "loss": 0.7485, + "num_input_tokens_seen": 75555575, + "step": 3554 + }, + { + "epoch": 0.4274634762219684, + "flos": 21697206460560.0, + "grad_norm": 3.2674265003766756, + "learning_rate": 2.559522997914115e-06, + "loss": 0.6602, + "num_input_tokens_seen": 75574875, + "step": 3555 + }, + { + "epoch": 0.42758371911260745, + "flos": 15675882102600.0, + "grad_norm": 3.7514893435296734, + "learning_rate": 2.558775088549599e-06, + "loss": 0.8455, + "num_input_tokens_seen": 75594175, + "step": 3556 + }, + { + "epoch": 0.42770396200324656, + "flos": 10730626770480.0, + "grad_norm": 3.6118334039688076, + "learning_rate": 2.5580270944193467e-06, + "loss": 0.6364, + "num_input_tokens_seen": 75610715, + "step": 3557 + }, + { + "epoch": 0.4278242048938857, + "flos": 52107750612720.0, + "grad_norm": 0.7331166138620573, + "learning_rate": 2.557279015636827e-06, + "loss": 0.5673, + "num_input_tokens_seen": 75670845, + "step": 3558 + }, + { + "epoch": 0.42794444778452473, + "flos": 51152505197400.0, + "grad_norm": 0.8123009144355335, + "learning_rate": 2.5565308523155245e-06, + "loss": 0.6504, + "num_input_tokens_seen": 75730165, + "step": 3559 + }, + { + "epoch": 0.42806469067516384, + "flos": 13292598953880.0, + "grad_norm": 4.553916276968825, + "learning_rate": 2.5557826045689336e-06, + "loss": 0.8035, + "num_input_tokens_seen": 75746125, + "step": 3560 + }, + { + "epoch": 0.4281849335658029, + "flos": 40179787966440.0, + "grad_norm": 0.8368162011137279, + "learning_rate": 2.5550342725105643e-06, + "loss": 0.6011, + "num_input_tokens_seen": 75804010, + "step": 3561 + }, + { + "epoch": 0.428305176456442, + "flos": 12597716287320.0, + "grad_norm": 2.6721869392559494, + "learning_rate": 2.554285856253937e-06, + "loss": 0.8, + "num_input_tokens_seen": 75822565, + "step": 3562 + }, + { + "epoch": 0.4284254193470811, + "flos": 19320951822960.0, + "grad_norm": 2.4575036690918397, + "learning_rate": 2.5535373559125855e-06, + "loss": 0.749, + "num_input_tokens_seen": 75842650, + "step": 3563 + }, + { + "epoch": 0.42854566223772017, + "flos": 21803407660920.0, + "grad_norm": 2.1520849719661395, + "learning_rate": 2.552788771600057e-06, + "loss": 0.8101, + "num_input_tokens_seen": 75862680, + "step": 3564 + }, + { + "epoch": 0.4286659051283593, + "flos": 16108126994640.0, + "grad_norm": 3.0247731967108753, + "learning_rate": 2.5520401034299118e-06, + "loss": 0.799, + "num_input_tokens_seen": 75880160, + "step": 3565 + }, + { + "epoch": 0.4287861480189984, + "flos": 9681559199160.0, + "grad_norm": 3.806102325130368, + "learning_rate": 2.551291351515722e-06, + "loss": 0.8516, + "num_input_tokens_seen": 75896895, + "step": 3566 + }, + { + "epoch": 0.42890639090963745, + "flos": 19537090098960.0, + "grad_norm": 4.908412966813531, + "learning_rate": 2.5505425159710726e-06, + "loss": 0.8469, + "num_input_tokens_seen": 75916425, + "step": 3567 + }, + { + "epoch": 0.42902663380027656, + "flos": 17615301153960.0, + "grad_norm": 5.637275525548858, + "learning_rate": 2.549793596909561e-06, + "loss": 0.8203, + "num_input_tokens_seen": 75934765, + "step": 3568 + }, + { + "epoch": 0.42914687669091567, + "flos": 11382270914400.0, + "grad_norm": 7.05911795058238, + "learning_rate": 2.5490445944447976e-06, + "loss": 0.6442, + "num_input_tokens_seen": 75952980, + "step": 3569 + }, + { + "epoch": 0.4292671195815547, + "flos": 23101472055360.0, + "grad_norm": 9.818080602952607, + "learning_rate": 2.548295508690406e-06, + "loss": 0.6404, + "num_input_tokens_seen": 75973995, + "step": 3570 + }, + { + "epoch": 0.42938736247219383, + "flos": 22205105182920.0, + "grad_norm": 2.7538513610997164, + "learning_rate": 2.5475463397600217e-06, + "loss": 0.7512, + "num_input_tokens_seen": 75993795, + "step": 3571 + }, + { + "epoch": 0.42950760536283294, + "flos": 21533646395400.0, + "grad_norm": 4.687933939875197, + "learning_rate": 2.546797087767293e-06, + "loss": 0.7657, + "num_input_tokens_seen": 76013640, + "step": 3572 + }, + { + "epoch": 0.429627848253472, + "flos": 19698655586640.0, + "grad_norm": 3.724871685711094, + "learning_rate": 2.546047752825881e-06, + "loss": 0.8516, + "num_input_tokens_seen": 76033965, + "step": 3573 + }, + { + "epoch": 0.4297480911441111, + "flos": 10086201097440.0, + "grad_norm": 3.4377668797323597, + "learning_rate": 2.5452983350494595e-06, + "loss": 0.9173, + "num_input_tokens_seen": 76049240, + "step": 3574 + }, + { + "epoch": 0.4298683340347502, + "flos": 15163044426480.0, + "grad_norm": 3.609304980635095, + "learning_rate": 2.544548834551713e-06, + "loss": 0.6442, + "num_input_tokens_seen": 76067965, + "step": 3575 + }, + { + "epoch": 0.4299885769253893, + "flos": 15266681170080.0, + "grad_norm": 3.813119056345633, + "learning_rate": 2.5437992514463424e-06, + "loss": 0.9086, + "num_input_tokens_seen": 76081010, + "step": 3576 + }, + { + "epoch": 0.4301088198160284, + "flos": 18673581773640.0, + "grad_norm": 2.599362471252905, + "learning_rate": 2.5430495858470565e-06, + "loss": 0.8643, + "num_input_tokens_seen": 76100200, + "step": 3577 + }, + { + "epoch": 0.43022906270666744, + "flos": 13326312319920.0, + "grad_norm": 3.4116722136953856, + "learning_rate": 2.54229983786758e-06, + "loss": 0.756, + "num_input_tokens_seen": 76117865, + "step": 3578 + }, + { + "epoch": 0.43034930559730655, + "flos": 17129718212040.0, + "grad_norm": 2.5733063952200523, + "learning_rate": 2.541550007621651e-06, + "loss": 0.8287, + "num_input_tokens_seen": 76136075, + "step": 3579 + }, + { + "epoch": 0.43046954848794566, + "flos": 20671562768280.0, + "grad_norm": 3.1825496154897115, + "learning_rate": 2.5408000952230156e-06, + "loss": 0.7889, + "num_input_tokens_seen": 76154585, + "step": 3580 + }, + { + "epoch": 0.4305897913785847, + "flos": 20963956413840.0, + "grad_norm": 3.2554428300953977, + "learning_rate": 2.5400501007854357e-06, + "loss": 0.8811, + "num_input_tokens_seen": 76173750, + "step": 3581 + }, + { + "epoch": 0.43071003426922383, + "flos": 14946557890920.0, + "grad_norm": 2.6890519562536155, + "learning_rate": 2.539300024422685e-06, + "loss": 0.7412, + "num_input_tokens_seen": 76191415, + "step": 3582 + }, + { + "epoch": 0.43083027715986294, + "flos": 38298736407240.0, + "grad_norm": 0.7995857060092261, + "learning_rate": 2.538549866248549e-06, + "loss": 0.6317, + "num_input_tokens_seen": 76246115, + "step": 3583 + }, + { + "epoch": 0.430950520050502, + "flos": 12165534715200.0, + "grad_norm": 4.0460390383712825, + "learning_rate": 2.5377996263768274e-06, + "loss": 0.7914, + "num_input_tokens_seen": 76263915, + "step": 3584 + }, + { + "epoch": 0.4310707629411411, + "flos": 18025071965760.0, + "grad_norm": 7.5867772369283015, + "learning_rate": 2.5370493049213293e-06, + "loss": 0.6563, + "num_input_tokens_seen": 76283280, + "step": 3585 + }, + { + "epoch": 0.4311910058317802, + "flos": 19375397971440.0, + "grad_norm": 16.3268343912785, + "learning_rate": 2.536298901995878e-06, + "loss": 0.7826, + "num_input_tokens_seen": 76302210, + "step": 3586 + }, + { + "epoch": 0.43131124872241927, + "flos": 18430252083360.0, + "grad_norm": 2.897175393275218, + "learning_rate": 2.535548417714311e-06, + "loss": 0.7879, + "num_input_tokens_seen": 76321230, + "step": 3587 + }, + { + "epoch": 0.4314314916130584, + "flos": 15809338037160.0, + "grad_norm": 2.754932228680997, + "learning_rate": 2.534797852190474e-06, + "loss": 0.8542, + "num_input_tokens_seen": 76341130, + "step": 3588 + }, + { + "epoch": 0.4315517345036975, + "flos": 14077065833160.0, + "grad_norm": 2.445014521081783, + "learning_rate": 2.5340472055382283e-06, + "loss": 0.787, + "num_input_tokens_seen": 76356880, + "step": 3589 + }, + { + "epoch": 0.43167197739433655, + "flos": 17776581702000.0, + "grad_norm": 4.322594590042902, + "learning_rate": 2.5332964778714468e-06, + "loss": 0.796, + "num_input_tokens_seen": 76373785, + "step": 3590 + }, + { + "epoch": 0.43179222028497566, + "flos": 12295698013920.0, + "grad_norm": 2.3463742632469224, + "learning_rate": 2.5325456693040123e-06, + "loss": 0.6493, + "num_input_tokens_seen": 76390700, + "step": 3591 + }, + { + "epoch": 0.43191246317561477, + "flos": 12867319253040.0, + "grad_norm": 3.631074091578512, + "learning_rate": 2.531794779949824e-06, + "loss": 0.7377, + "num_input_tokens_seen": 76408320, + "step": 3592 + }, + { + "epoch": 0.4320327060662538, + "flos": 17485106195280.0, + "grad_norm": 2.9009541487490855, + "learning_rate": 2.5310438099227903e-06, + "loss": 0.8638, + "num_input_tokens_seen": 76425305, + "step": 3593 + }, + { + "epoch": 0.43215294895689293, + "flos": 48956495423880.0, + "grad_norm": 1.3162433738577939, + "learning_rate": 2.530292759336833e-06, + "loss": 0.5535, + "num_input_tokens_seen": 76485760, + "step": 3594 + }, + { + "epoch": 0.432273191847532, + "flos": 15054215449440.0, + "grad_norm": 3.9628236030788266, + "learning_rate": 2.5295416283058855e-06, + "loss": 0.6897, + "num_input_tokens_seen": 76504345, + "step": 3595 + }, + { + "epoch": 0.4323934347381711, + "flos": 14083334505240.0, + "grad_norm": 2.965711142572175, + "learning_rate": 2.5287904169438943e-06, + "loss": 0.6459, + "num_input_tokens_seen": 76523270, + "step": 3596 + }, + { + "epoch": 0.4325136776288102, + "flos": 15892147018440.0, + "grad_norm": 4.649458880369739, + "learning_rate": 2.528039125364817e-06, + "loss": 0.6197, + "num_input_tokens_seen": 76541795, + "step": 3597 + }, + { + "epoch": 0.43263392051944927, + "flos": 16346042831760.0, + "grad_norm": 2.8071408393461574, + "learning_rate": 2.5272877536826246e-06, + "loss": 0.73, + "num_input_tokens_seen": 76560310, + "step": 3598 + }, + { + "epoch": 0.4327541634100884, + "flos": 21399652241520.0, + "grad_norm": 3.544160047486528, + "learning_rate": 2.5265363020112986e-06, + "loss": 0.6726, + "num_input_tokens_seen": 76580350, + "step": 3599 + }, + { + "epoch": 0.4328744063007275, + "flos": 19104781887000.0, + "grad_norm": 2.4665809470395126, + "learning_rate": 2.5257847704648344e-06, + "loss": 0.8301, + "num_input_tokens_seen": 76601300, + "step": 3600 + }, + { + "epoch": 0.43299464919136654, + "flos": 12084467031720.0, + "grad_norm": 5.683763555238331, + "learning_rate": 2.525033159157239e-06, + "loss": 0.7541, + "num_input_tokens_seen": 76617335, + "step": 3601 + }, + { + "epoch": 0.43311489208200565, + "flos": 11733638082720.0, + "grad_norm": 2.9217653304953113, + "learning_rate": 2.52428146820253e-06, + "loss": 0.7523, + "num_input_tokens_seen": 76635310, + "step": 3602 + }, + { + "epoch": 0.43323513497264476, + "flos": 16783100037720.0, + "grad_norm": 3.147679239976386, + "learning_rate": 2.52352969771474e-06, + "loss": 0.8043, + "num_input_tokens_seen": 76654255, + "step": 3603 + }, + { + "epoch": 0.4333553778632838, + "flos": 18535788424560.0, + "grad_norm": 2.4356575874182713, + "learning_rate": 2.5227778478079106e-06, + "loss": 0.8731, + "num_input_tokens_seen": 76673385, + "step": 3604 + }, + { + "epoch": 0.43347562075392293, + "flos": 14160444693720.0, + "grad_norm": 2.3453043551878583, + "learning_rate": 2.522025918596098e-06, + "loss": 0.7566, + "num_input_tokens_seen": 76691405, + "step": 3605 + }, + { + "epoch": 0.43359586364456204, + "flos": 19296546485160.0, + "grad_norm": 2.218954264949538, + "learning_rate": 2.521273910193368e-06, + "loss": 0.6394, + "num_input_tokens_seen": 76714305, + "step": 3606 + }, + { + "epoch": 0.4337161065352011, + "flos": 11645098648680.0, + "grad_norm": 3.201055388457936, + "learning_rate": 2.5205218227138006e-06, + "loss": 0.8522, + "num_input_tokens_seen": 76726980, + "step": 3607 + }, + { + "epoch": 0.4338363494258402, + "flos": 14780053449480.0, + "grad_norm": 2.678127357366141, + "learning_rate": 2.519769656271486e-06, + "loss": 0.7743, + "num_input_tokens_seen": 76744120, + "step": 3608 + }, + { + "epoch": 0.43395659231647926, + "flos": 14676575005680.0, + "grad_norm": 4.046731482358029, + "learning_rate": 2.5190174109805285e-06, + "loss": 0.6571, + "num_input_tokens_seen": 76763665, + "step": 3609 + }, + { + "epoch": 0.43407683520711837, + "flos": 14541852672720.0, + "grad_norm": 2.9480835201280153, + "learning_rate": 2.518265086955042e-06, + "loss": 0.6194, + "num_input_tokens_seen": 76781105, + "step": 3610 + }, + { + "epoch": 0.4341970780977575, + "flos": 16914656374680.0, + "grad_norm": 7.831938707370968, + "learning_rate": 2.5175126843091534e-06, + "loss": 0.8116, + "num_input_tokens_seen": 76800195, + "step": 3611 + }, + { + "epoch": 0.43431732098839654, + "flos": 27496883709480.0, + "grad_norm": 3.507088703340094, + "learning_rate": 2.5167602031570034e-06, + "loss": 0.7302, + "num_input_tokens_seen": 76820100, + "step": 3612 + }, + { + "epoch": 0.43443756387903565, + "flos": 23397664896120.0, + "grad_norm": 2.6079964108354643, + "learning_rate": 2.51600764361274e-06, + "loss": 0.7224, + "num_input_tokens_seen": 76841345, + "step": 3613 + }, + { + "epoch": 0.43455780676967476, + "flos": 17187805255920.0, + "grad_norm": 2.763883052885794, + "learning_rate": 2.5152550057905283e-06, + "loss": 0.7755, + "num_input_tokens_seen": 76860955, + "step": 3614 + }, + { + "epoch": 0.4346780496603138, + "flos": 17728024306080.0, + "grad_norm": 3.212594822771819, + "learning_rate": 2.5145022898045415e-06, + "loss": 0.7332, + "num_input_tokens_seen": 76879860, + "step": 3615 + }, + { + "epoch": 0.4347982925509529, + "flos": 12462677354760.0, + "grad_norm": 2.828502650890352, + "learning_rate": 2.5137494957689664e-06, + "loss": 0.8805, + "num_input_tokens_seen": 76895190, + "step": 3616 + }, + { + "epoch": 0.43491853544159204, + "flos": 44921754474960.0, + "grad_norm": 0.7528284699997451, + "learning_rate": 2.5129966237980016e-06, + "loss": 0.5917, + "num_input_tokens_seen": 76957905, + "step": 3617 + }, + { + "epoch": 0.4350387783322311, + "flos": 16053649186200.0, + "grad_norm": 3.3235498062334137, + "learning_rate": 2.512243674005857e-06, + "loss": 0.7612, + "num_input_tokens_seen": 76976990, + "step": 3618 + }, + { + "epoch": 0.4351590212228702, + "flos": 18376059214560.0, + "grad_norm": 3.0068833538503132, + "learning_rate": 2.5114906465067537e-06, + "loss": 0.8487, + "num_input_tokens_seen": 76997695, + "step": 3619 + }, + { + "epoch": 0.4352792641135093, + "flos": 15729378452280.0, + "grad_norm": 8.061936469677324, + "learning_rate": 2.5107375414149264e-06, + "loss": 0.7255, + "num_input_tokens_seen": 77016660, + "step": 3620 + }, + { + "epoch": 0.43539950700414837, + "flos": 11841865520520.0, + "grad_norm": 3.40885587633448, + "learning_rate": 2.5099843588446197e-06, + "loss": 0.6854, + "num_input_tokens_seen": 77034700, + "step": 3621 + }, + { + "epoch": 0.4355197498947875, + "flos": 12166357874160.0, + "grad_norm": 2.0508297287796795, + "learning_rate": 2.509231098910091e-06, + "loss": 0.6014, + "num_input_tokens_seen": 77054290, + "step": 3622 + }, + { + "epoch": 0.4356399927854266, + "flos": 11922711584280.0, + "grad_norm": 3.0101938308203486, + "learning_rate": 2.508477761725611e-06, + "loss": 0.7179, + "num_input_tokens_seen": 77072285, + "step": 3623 + }, + { + "epoch": 0.43576023567606564, + "flos": 12543871678080.0, + "grad_norm": 2.185697848617054, + "learning_rate": 2.507724347405458e-06, + "loss": 0.7875, + "num_input_tokens_seen": 77089955, + "step": 3624 + }, + { + "epoch": 0.43588047856670475, + "flos": 11592298818120.0, + "grad_norm": 3.056729231942553, + "learning_rate": 2.5069708560639243e-06, + "loss": 0.7994, + "num_input_tokens_seen": 77107585, + "step": 3625 + }, + { + "epoch": 0.4360007214573438, + "flos": 17322812528520.0, + "grad_norm": 3.7589803613797677, + "learning_rate": 2.5062172878153158e-06, + "loss": 0.5968, + "num_input_tokens_seen": 77126580, + "step": 3626 + }, + { + "epoch": 0.4361209643479829, + "flos": 16078561083360.0, + "grad_norm": 2.858647154845838, + "learning_rate": 2.505463642773947e-06, + "loss": 0.8572, + "num_input_tokens_seen": 77146265, + "step": 3627 + }, + { + "epoch": 0.43624120723862203, + "flos": 12705183886080.0, + "grad_norm": 2.728623624942927, + "learning_rate": 2.504709921054146e-06, + "loss": 0.7321, + "num_input_tokens_seen": 77162800, + "step": 3628 + }, + { + "epoch": 0.4363614501292611, + "flos": 13056930973920.0, + "grad_norm": 3.4989528451921754, + "learning_rate": 2.50395612277025e-06, + "loss": 0.8305, + "num_input_tokens_seen": 77178375, + "step": 3629 + }, + { + "epoch": 0.4364816930199002, + "flos": 14838647052720.0, + "grad_norm": 2.7822299045423504, + "learning_rate": 2.503202248036612e-06, + "loss": 0.7221, + "num_input_tokens_seen": 77196950, + "step": 3630 + }, + { + "epoch": 0.4366019359105393, + "flos": 17620208447760.0, + "grad_norm": 2.3712134040417627, + "learning_rate": 2.5024482969675927e-06, + "loss": 0.7078, + "num_input_tokens_seen": 77216625, + "step": 3631 + }, + { + "epoch": 0.43672217880117836, + "flos": 15911486762640.0, + "grad_norm": 2.5918719374963066, + "learning_rate": 2.501694269677566e-06, + "loss": 0.8293, + "num_input_tokens_seen": 77234115, + "step": 3632 + }, + { + "epoch": 0.4368424216918175, + "flos": 13159617918720.0, + "grad_norm": 2.796458463712633, + "learning_rate": 2.500940166280918e-06, + "loss": 0.78, + "num_input_tokens_seen": 77252265, + "step": 3633 + }, + { + "epoch": 0.4369626645824566, + "flos": 18645693840240.0, + "grad_norm": 3.3132736454224068, + "learning_rate": 2.500185986892045e-06, + "loss": 0.7655, + "num_input_tokens_seen": 77271470, + "step": 3634 + }, + { + "epoch": 0.43708290747309564, + "flos": 18539017740480.0, + "grad_norm": 2.9509432367691013, + "learning_rate": 2.499431731625355e-06, + "loss": 0.7633, + "num_input_tokens_seen": 77290215, + "step": 3635 + }, + { + "epoch": 0.43720315036373475, + "flos": 23180576821320.0, + "grad_norm": 2.256113436045207, + "learning_rate": 2.4986774005952686e-06, + "loss": 0.7849, + "num_input_tokens_seen": 77312310, + "step": 3636 + }, + { + "epoch": 0.43732339325437386, + "flos": 16917537431040.0, + "grad_norm": 2.7140882140562783, + "learning_rate": 2.4979229939162166e-06, + "loss": 0.8257, + "num_input_tokens_seen": 77330810, + "step": 3637 + }, + { + "epoch": 0.4374436361450129, + "flos": 20347292034360.0, + "grad_norm": 1.799800382247801, + "learning_rate": 2.4971685117026433e-06, + "loss": 0.7851, + "num_input_tokens_seen": 77350295, + "step": 3638 + }, + { + "epoch": 0.437563879035652, + "flos": 17702605849560.0, + "grad_norm": 1.9112352939237032, + "learning_rate": 2.4964139540690018e-06, + "loss": 0.7533, + "num_input_tokens_seen": 77373350, + "step": 3639 + }, + { + "epoch": 0.4376841219262911, + "flos": 17187868575840.0, + "grad_norm": 2.733607505988315, + "learning_rate": 2.495659321129758e-06, + "loss": 0.7109, + "num_input_tokens_seen": 77390815, + "step": 3640 + }, + { + "epoch": 0.4378043648169302, + "flos": 18646295379480.0, + "grad_norm": 2.128295250803108, + "learning_rate": 2.494904612999389e-06, + "loss": 0.7399, + "num_input_tokens_seen": 77409245, + "step": 3641 + }, + { + "epoch": 0.4379246077075693, + "flos": 39719433521280.0, + "grad_norm": 0.7992909403515303, + "learning_rate": 2.4941498297923843e-06, + "loss": 0.6, + "num_input_tokens_seen": 77469535, + "step": 3642 + }, + { + "epoch": 0.43804485059820836, + "flos": 15048959896080.0, + "grad_norm": 3.482821952358612, + "learning_rate": 2.4933949716232424e-06, + "loss": 0.6933, + "num_input_tokens_seen": 77486780, + "step": 3643 + }, + { + "epoch": 0.43816509348884747, + "flos": 17482826678160.0, + "grad_norm": 2.5997572125271455, + "learning_rate": 2.492640038606476e-06, + "loss": 0.7109, + "num_input_tokens_seen": 77504865, + "step": 3644 + }, + { + "epoch": 0.4382853363794866, + "flos": 10680676336320.0, + "grad_norm": 6.188453061464174, + "learning_rate": 2.491885030856608e-06, + "loss": 0.7756, + "num_input_tokens_seen": 77522680, + "step": 3645 + }, + { + "epoch": 0.43840557927012563, + "flos": 12516648603840.0, + "grad_norm": 2.887112354498446, + "learning_rate": 2.4911299484881713e-06, + "loss": 0.8082, + "num_input_tokens_seen": 77539930, + "step": 3646 + }, + { + "epoch": 0.43852582216076474, + "flos": 14163547369800.0, + "grad_norm": 3.6740670661349557, + "learning_rate": 2.490374791615712e-06, + "loss": 0.802, + "num_input_tokens_seen": 77559675, + "step": 3647 + }, + { + "epoch": 0.43864606505140386, + "flos": 13189658729400.0, + "grad_norm": 3.3843457841536746, + "learning_rate": 2.4896195603537867e-06, + "loss": 0.7586, + "num_input_tokens_seen": 77574005, + "step": 3648 + }, + { + "epoch": 0.4387663079420429, + "flos": 14351512772760.0, + "grad_norm": 3.1543659592376665, + "learning_rate": 2.488864254816964e-06, + "loss": 0.7224, + "num_input_tokens_seen": 77592415, + "step": 3649 + }, + { + "epoch": 0.438886550832682, + "flos": 14406813740160.0, + "grad_norm": 3.1304533106720904, + "learning_rate": 2.4881088751198218e-06, + "loss": 0.654, + "num_input_tokens_seen": 77610295, + "step": 3650 + }, + { + "epoch": 0.43900679372332113, + "flos": 10571055860280.0, + "grad_norm": 5.642447982771419, + "learning_rate": 2.4873534213769517e-06, + "loss": 0.613, + "num_input_tokens_seen": 77625245, + "step": 3651 + }, + { + "epoch": 0.4391270366139602, + "flos": 17615712733440.0, + "grad_norm": 2.0619661044749056, + "learning_rate": 2.4865978937029547e-06, + "loss": 0.701, + "num_input_tokens_seen": 77643945, + "step": 3652 + }, + { + "epoch": 0.4392472795045993, + "flos": 23153480386920.0, + "grad_norm": 2.3307615366700025, + "learning_rate": 2.485842292212445e-06, + "loss": 0.6523, + "num_input_tokens_seen": 77664880, + "step": 3653 + }, + { + "epoch": 0.4393675223952384, + "flos": 10815177049560.0, + "grad_norm": 4.847443384604235, + "learning_rate": 2.485086617020045e-06, + "loss": 0.7778, + "num_input_tokens_seen": 77683095, + "step": 3654 + }, + { + "epoch": 0.43948776528587746, + "flos": 10784724659400.0, + "grad_norm": 3.7142181992944647, + "learning_rate": 2.4843308682403903e-06, + "loss": 0.7973, + "num_input_tokens_seen": 77699730, + "step": 3655 + }, + { + "epoch": 0.4396080081765166, + "flos": 10110353155560.0, + "grad_norm": 1.7504611770042477, + "learning_rate": 2.4835750459881294e-06, + "loss": 0.8126, + "num_input_tokens_seen": 77716075, + "step": 3656 + }, + { + "epoch": 0.43972825106715563, + "flos": 13299342525360.0, + "grad_norm": 2.1542313566646247, + "learning_rate": 2.4828191503779177e-06, + "loss": 0.795, + "num_input_tokens_seen": 77733895, + "step": 3657 + }, + { + "epoch": 0.43984849395779474, + "flos": 12300478667880.0, + "grad_norm": 22.751809519811413, + "learning_rate": 2.482063181524425e-06, + "loss": 0.8783, + "num_input_tokens_seen": 77749515, + "step": 3658 + }, + { + "epoch": 0.43996873684843385, + "flos": 13645264180560.0, + "grad_norm": 2.1708837209212155, + "learning_rate": 2.4813071395423307e-06, + "loss": 0.7874, + "num_input_tokens_seen": 77766800, + "step": 3659 + }, + { + "epoch": 0.4400889797390729, + "flos": 17317430335320.0, + "grad_norm": 2.8232127148182746, + "learning_rate": 2.4805510245463263e-06, + "loss": 0.6382, + "num_input_tokens_seen": 77786675, + "step": 3660 + }, + { + "epoch": 0.440209222629712, + "flos": 16945172084760.0, + "grad_norm": 2.888079703132008, + "learning_rate": 2.4797948366511137e-06, + "loss": 0.563, + "num_input_tokens_seen": 77806105, + "step": 3661 + }, + { + "epoch": 0.4403294655203511, + "flos": 18183598097280.0, + "grad_norm": 2.498079229627014, + "learning_rate": 2.4790385759714055e-06, + "loss": 0.7347, + "num_input_tokens_seen": 77824890, + "step": 3662 + }, + { + "epoch": 0.4404497084109902, + "flos": 16512832212840.0, + "grad_norm": 2.641228601212842, + "learning_rate": 2.478282242621926e-06, + "loss": 0.6981, + "num_input_tokens_seen": 77845070, + "step": 3663 + }, + { + "epoch": 0.4405699513016293, + "flos": 47899101283080.0, + "grad_norm": 0.9147960295312884, + "learning_rate": 2.477525836717411e-06, + "loss": 0.6271, + "num_input_tokens_seen": 77912555, + "step": 3664 + }, + { + "epoch": 0.4406901941922684, + "flos": 26205182967000.0, + "grad_norm": 3.224335647878214, + "learning_rate": 2.476769358372606e-06, + "loss": 0.7754, + "num_input_tokens_seen": 77933925, + "step": 3665 + }, + { + "epoch": 0.44081043708290746, + "flos": 13164525212520.0, + "grad_norm": 2.820228216121795, + "learning_rate": 2.4760128077022683e-06, + "loss": 0.7278, + "num_input_tokens_seen": 77951780, + "step": 3666 + }, + { + "epoch": 0.44093067997354657, + "flos": 22128691513560.0, + "grad_norm": 2.266232534047745, + "learning_rate": 2.4752561848211672e-06, + "loss": 0.6719, + "num_input_tokens_seen": 77973900, + "step": 3667 + }, + { + "epoch": 0.4410509228641857, + "flos": 17023010452320.0, + "grad_norm": 4.6826739072099235, + "learning_rate": 2.4744994898440797e-06, + "loss": 0.6947, + "num_input_tokens_seen": 77992410, + "step": 3668 + }, + { + "epoch": 0.44117116575482473, + "flos": 14244931652880.0, + "grad_norm": 2.465992316865979, + "learning_rate": 2.473742722885797e-06, + "loss": 0.8223, + "num_input_tokens_seen": 78011150, + "step": 3669 + }, + { + "epoch": 0.44129140864546385, + "flos": 19834327718400.0, + "grad_norm": 4.762536220462864, + "learning_rate": 2.4729858840611197e-06, + "loss": 0.6384, + "num_input_tokens_seen": 78029780, + "step": 3670 + }, + { + "epoch": 0.4414116515361029, + "flos": 19130168683560.0, + "grad_norm": 4.165101685673707, + "learning_rate": 2.4722289734848605e-06, + "loss": 0.7122, + "num_input_tokens_seen": 78049965, + "step": 3671 + }, + { + "epoch": 0.441531894426742, + "flos": 16024811454000.0, + "grad_norm": 4.707752270697847, + "learning_rate": 2.471471991271841e-06, + "loss": 0.7665, + "num_input_tokens_seen": 78066810, + "step": 3672 + }, + { + "epoch": 0.4416521373173811, + "flos": 17157669465360.0, + "grad_norm": 2.0162534695032, + "learning_rate": 2.470714937536896e-06, + "loss": 0.7805, + "num_input_tokens_seen": 78085255, + "step": 3673 + }, + { + "epoch": 0.4417723802080202, + "flos": 14861216112840.0, + "grad_norm": 2.9206640008517897, + "learning_rate": 2.469957812394868e-06, + "loss": 0.7001, + "num_input_tokens_seen": 78103785, + "step": 3674 + }, + { + "epoch": 0.4418926230986593, + "flos": 13786223525640.0, + "grad_norm": 2.5482916139423133, + "learning_rate": 2.4692006159606148e-06, + "loss": 0.7505, + "num_input_tokens_seen": 78121035, + "step": 3675 + }, + { + "epoch": 0.4420128659892984, + "flos": 14218215138000.0, + "grad_norm": 3.0565421871629286, + "learning_rate": 2.468443348349e-06, + "loss": 0.7662, + "num_input_tokens_seen": 78138630, + "step": 3676 + }, + { + "epoch": 0.44213310887993745, + "flos": 13056107814960.0, + "grad_norm": 3.293873075614304, + "learning_rate": 2.467686009674902e-06, + "loss": 0.8071, + "num_input_tokens_seen": 78152800, + "step": 3677 + }, + { + "epoch": 0.44225335177057656, + "flos": 14025374101200.0, + "grad_norm": 2.436242292616089, + "learning_rate": 2.466928600053209e-06, + "loss": 0.8348, + "num_input_tokens_seen": 78167825, + "step": 3678 + }, + { + "epoch": 0.4423735946612157, + "flos": 17182676342400.0, + "grad_norm": 2.9854271117668993, + "learning_rate": 2.466171119598818e-06, + "loss": 0.694, + "num_input_tokens_seen": 78187515, + "step": 3679 + }, + { + "epoch": 0.44249383755185473, + "flos": 19561780376400.0, + "grad_norm": 2.7469869927152866, + "learning_rate": 2.465413568426639e-06, + "loss": 0.7474, + "num_input_tokens_seen": 78208185, + "step": 3680 + }, + { + "epoch": 0.44261408044249384, + "flos": 16943810706480.0, + "grad_norm": 6.649864034768545, + "learning_rate": 2.464655946651591e-06, + "loss": 0.8084, + "num_input_tokens_seen": 78226910, + "step": 3681 + }, + { + "epoch": 0.44273432333313295, + "flos": 17917129467600.0, + "grad_norm": 2.5222225512235017, + "learning_rate": 2.4638982543886065e-06, + "loss": 0.7857, + "num_input_tokens_seen": 78246670, + "step": 3682 + }, + { + "epoch": 0.442854566223772, + "flos": 12785586710400.0, + "grad_norm": 4.172050329517415, + "learning_rate": 2.4631404917526254e-06, + "loss": 0.8463, + "num_input_tokens_seen": 78263345, + "step": 3683 + }, + { + "epoch": 0.4429748091144111, + "flos": 18237537686400.0, + "grad_norm": 2.3325288839074845, + "learning_rate": 2.4623826588586e-06, + "loss": 0.7718, + "num_input_tokens_seen": 78283335, + "step": 3684 + }, + { + "epoch": 0.4430950520050502, + "flos": 15809243057280.0, + "grad_norm": 2.365335678085805, + "learning_rate": 2.461624755821492e-06, + "loss": 0.8166, + "num_input_tokens_seen": 78302535, + "step": 3685 + }, + { + "epoch": 0.4432152948956893, + "flos": 17998355450880.0, + "grad_norm": 2.275324297417128, + "learning_rate": 2.4608667827562763e-06, + "loss": 0.761, + "num_input_tokens_seen": 78321585, + "step": 3686 + }, + { + "epoch": 0.4433355377863284, + "flos": 15918673573560.0, + "grad_norm": 2.4236226329025583, + "learning_rate": 2.460108739777936e-06, + "loss": 0.8771, + "num_input_tokens_seen": 78340440, + "step": 3687 + }, + { + "epoch": 0.44345578067696745, + "flos": 14677081565040.0, + "grad_norm": 2.8060079457448532, + "learning_rate": 2.4593506270014656e-06, + "loss": 0.7501, + "num_input_tokens_seen": 78359130, + "step": 3688 + }, + { + "epoch": 0.44357602356760656, + "flos": 17700073052760.0, + "grad_norm": 3.788112371814186, + "learning_rate": 2.45859244454187e-06, + "loss": 0.8027, + "num_input_tokens_seen": 78378640, + "step": 3689 + }, + { + "epoch": 0.44369626645824567, + "flos": 16617925314600.0, + "grad_norm": 2.7442054387919668, + "learning_rate": 2.4578341925141655e-06, + "loss": 0.6504, + "num_input_tokens_seen": 78397575, + "step": 3690 + }, + { + "epoch": 0.4438165093488847, + "flos": 27957396454440.0, + "grad_norm": 2.3089730456246436, + "learning_rate": 2.457075871033378e-06, + "loss": 0.701, + "num_input_tokens_seen": 78419170, + "step": 3691 + }, + { + "epoch": 0.44393675223952384, + "flos": 11301361530720.0, + "grad_norm": 2.719251915213733, + "learning_rate": 2.4563174802145445e-06, + "loss": 0.8654, + "num_input_tokens_seen": 78436140, + "step": 3692 + }, + { + "epoch": 0.44405699513016295, + "flos": 47610158573160.0, + "grad_norm": 0.6309555500050327, + "learning_rate": 2.455559020172712e-06, + "loss": 0.4911, + "num_input_tokens_seen": 78503215, + "step": 3693 + }, + { + "epoch": 0.444177238020802, + "flos": 17565350719800.0, + "grad_norm": 4.398667541234011, + "learning_rate": 2.4548004910229385e-06, + "loss": 0.8741, + "num_input_tokens_seen": 78520510, + "step": 3694 + }, + { + "epoch": 0.4442974809114411, + "flos": 16511059255080.0, + "grad_norm": 2.150658364739726, + "learning_rate": 2.4540418928802913e-06, + "loss": 0.8526, + "num_input_tokens_seen": 78538965, + "step": 3695 + }, + { + "epoch": 0.4444177238020802, + "flos": 12894320707560.0, + "grad_norm": 2.6909748944231664, + "learning_rate": 2.4532832258598506e-06, + "loss": 0.635, + "num_input_tokens_seen": 78556515, + "step": 3696 + }, + { + "epoch": 0.4445379666927193, + "flos": 20995042003200.0, + "grad_norm": 8.860373896508968, + "learning_rate": 2.4525244900767047e-06, + "loss": 0.7986, + "num_input_tokens_seen": 78577050, + "step": 3697 + }, + { + "epoch": 0.4446582095833584, + "flos": 51986040523320.0, + "grad_norm": 0.8245402789204387, + "learning_rate": 2.4517656856459536e-06, + "loss": 0.6253, + "num_input_tokens_seen": 78642615, + "step": 3698 + }, + { + "epoch": 0.4447784524739975, + "flos": 19428736021320.0, + "grad_norm": 3.8715384003741056, + "learning_rate": 2.4510068126827073e-06, + "loss": 0.6649, + "num_input_tokens_seen": 78663335, + "step": 3699 + }, + { + "epoch": 0.44489869536463655, + "flos": 8439587271120.0, + "grad_norm": 3.7192290181899383, + "learning_rate": 2.450247871302086e-06, + "loss": 0.8044, + "num_input_tokens_seen": 78680830, + "step": 3700 + }, + { + "epoch": 0.44501893825527566, + "flos": 14946336271200.0, + "grad_norm": 3.778068149539645, + "learning_rate": 2.44948886161922e-06, + "loss": 0.8167, + "num_input_tokens_seen": 78699565, + "step": 3701 + }, + { + "epoch": 0.4451391811459148, + "flos": 13327768678080.0, + "grad_norm": 2.3686642714313937, + "learning_rate": 2.4487297837492524e-06, + "loss": 0.8344, + "num_input_tokens_seen": 78718450, + "step": 3702 + }, + { + "epoch": 0.44525942403655383, + "flos": 12327860041920.0, + "grad_norm": 2.9293603332946723, + "learning_rate": 2.4479706378073323e-06, + "loss": 0.5923, + "num_input_tokens_seen": 78736710, + "step": 3703 + }, + { + "epoch": 0.44537966692719294, + "flos": 17026334748120.0, + "grad_norm": 2.114658548555189, + "learning_rate": 2.447211423908623e-06, + "loss": 0.825, + "num_input_tokens_seen": 78756475, + "step": 3704 + }, + { + "epoch": 0.445499909817832, + "flos": 15890247420840.0, + "grad_norm": 2.5016611302581593, + "learning_rate": 2.4464521421682966e-06, + "loss": 0.7316, + "num_input_tokens_seen": 78773785, + "step": 3705 + }, + { + "epoch": 0.4456201527084711, + "flos": 17565857279160.0, + "grad_norm": 1.8413391515380697, + "learning_rate": 2.4456927927015345e-06, + "loss": 0.8634, + "num_input_tokens_seen": 78794545, + "step": 3706 + }, + { + "epoch": 0.4457403955991102, + "flos": 13731650737320.0, + "grad_norm": 3.655984461344413, + "learning_rate": 2.4449333756235307e-06, + "loss": 0.7347, + "num_input_tokens_seen": 78810980, + "step": 3707 + }, + { + "epoch": 0.4458606384897493, + "flos": 14028096857760.0, + "grad_norm": 4.150445771692453, + "learning_rate": 2.4441738910494876e-06, + "loss": 0.7851, + "num_input_tokens_seen": 78825435, + "step": 3708 + }, + { + "epoch": 0.4459808813803884, + "flos": 15621974173440.0, + "grad_norm": 2.811333986811313, + "learning_rate": 2.4434143390946176e-06, + "loss": 0.8105, + "num_input_tokens_seen": 78843965, + "step": 3709 + }, + { + "epoch": 0.4461011242710275, + "flos": 17047732389720.0, + "grad_norm": 2.472354141268482, + "learning_rate": 2.4426547198741457e-06, + "loss": 0.837, + "num_input_tokens_seen": 78861890, + "step": 3710 + }, + { + "epoch": 0.44622136716166655, + "flos": 14757262769640.0, + "grad_norm": 2.9083490482404906, + "learning_rate": 2.441895033503305e-06, + "loss": 0.7257, + "num_input_tokens_seen": 78879530, + "step": 3711 + }, + { + "epoch": 0.44634161005230566, + "flos": 15563317250280.0, + "grad_norm": 2.9636086365295764, + "learning_rate": 2.4411352800973375e-06, + "loss": 0.8157, + "num_input_tokens_seen": 78897685, + "step": 3712 + }, + { + "epoch": 0.44646185294294477, + "flos": 16782340198680.0, + "grad_norm": 3.02607946760737, + "learning_rate": 2.4403754597715005e-06, + "loss": 0.7205, + "num_input_tokens_seen": 78916850, + "step": 3713 + }, + { + "epoch": 0.4465820958335838, + "flos": 16566518522280.0, + "grad_norm": 3.039538348681903, + "learning_rate": 2.4396155726410553e-06, + "loss": 0.9113, + "num_input_tokens_seen": 78935180, + "step": 3714 + }, + { + "epoch": 0.44670233872422294, + "flos": 16592063618640.0, + "grad_norm": 3.1084890972582544, + "learning_rate": 2.438855618821278e-06, + "loss": 0.9005, + "num_input_tokens_seen": 78950700, + "step": 3715 + }, + { + "epoch": 0.44682258161486205, + "flos": 17264915444400.0, + "grad_norm": 5.0682116654174605, + "learning_rate": 2.4380955984274517e-06, + "loss": 0.6552, + "num_input_tokens_seen": 78969075, + "step": 3716 + }, + { + "epoch": 0.4469428245055011, + "flos": 19425348405600.0, + "grad_norm": 2.1490180323622154, + "learning_rate": 2.4373355115748716e-06, + "loss": 0.7484, + "num_input_tokens_seen": 78989625, + "step": 3717 + }, + { + "epoch": 0.4470630673961402, + "flos": 15727953754080.0, + "grad_norm": 2.138315333079778, + "learning_rate": 2.436575358378842e-06, + "loss": 0.7125, + "num_input_tokens_seen": 79008835, + "step": 3718 + }, + { + "epoch": 0.44718331028677927, + "flos": 11782132158720.0, + "grad_norm": 6.840997011569636, + "learning_rate": 2.4358151389546782e-06, + "loss": 0.8093, + "num_input_tokens_seen": 79025240, + "step": 3719 + }, + { + "epoch": 0.4473035531774184, + "flos": 14378387587440.0, + "grad_norm": 3.597334028263275, + "learning_rate": 2.4350548534177035e-06, + "loss": 0.7333, + "num_input_tokens_seen": 79041790, + "step": 3720 + }, + { + "epoch": 0.4474237960680575, + "flos": 30472267600080.0, + "grad_norm": 2.3528416598536275, + "learning_rate": 2.434294501883254e-06, + "loss": 0.6597, + "num_input_tokens_seen": 79064605, + "step": 3721 + }, + { + "epoch": 0.44754403895869654, + "flos": 16753977365880.0, + "grad_norm": 1.9305209317597793, + "learning_rate": 2.433534084466674e-06, + "loss": 0.6484, + "num_input_tokens_seen": 79083545, + "step": 3722 + }, + { + "epoch": 0.44766428184933565, + "flos": 18781334312040.0, + "grad_norm": 1.9147122895566255, + "learning_rate": 2.4327736012833178e-06, + "loss": 0.7046, + "num_input_tokens_seen": 79104985, + "step": 3723 + }, + { + "epoch": 0.44778452473997477, + "flos": 14946779510640.0, + "grad_norm": 2.9844596810834445, + "learning_rate": 2.4320130524485506e-06, + "loss": 0.7453, + "num_input_tokens_seen": 79123500, + "step": 3724 + }, + { + "epoch": 0.4479047676306138, + "flos": 16075648367040.0, + "grad_norm": 2.4055249992154697, + "learning_rate": 2.431252438077746e-06, + "loss": 0.7881, + "num_input_tokens_seen": 79142720, + "step": 3725 + }, + { + "epoch": 0.44802501052125293, + "flos": 15699875860920.0, + "grad_norm": 4.866620418032713, + "learning_rate": 2.4304917582862906e-06, + "loss": 0.7526, + "num_input_tokens_seen": 79161620, + "step": 3726 + }, + { + "epoch": 0.44814525341189204, + "flos": 16188561478920.0, + "grad_norm": 2.9607242718820217, + "learning_rate": 2.4297310131895774e-06, + "loss": 0.8701, + "num_input_tokens_seen": 79179885, + "step": 3727 + }, + { + "epoch": 0.4482654963025311, + "flos": 12138121681200.0, + "grad_norm": 4.8801695972426575, + "learning_rate": 2.4289702029030113e-06, + "loss": 0.7327, + "num_input_tokens_seen": 79197075, + "step": 3728 + }, + { + "epoch": 0.4483857391931702, + "flos": 13756625954400.0, + "grad_norm": 2.8228509104672583, + "learning_rate": 2.4282093275420057e-06, + "loss": 0.8137, + "num_input_tokens_seen": 79215825, + "step": 3729 + }, + { + "epoch": 0.4485059820838093, + "flos": 14889768905400.0, + "grad_norm": 2.9915573213477256, + "learning_rate": 2.4274483872219863e-06, + "loss": 0.6854, + "num_input_tokens_seen": 79232905, + "step": 3730 + }, + { + "epoch": 0.4486262249744484, + "flos": 14649858490800.0, + "grad_norm": 2.341925457628952, + "learning_rate": 2.426687382058386e-06, + "loss": 0.9177, + "num_input_tokens_seen": 79250905, + "step": 3731 + }, + { + "epoch": 0.4487464678650875, + "flos": 47623324625160.0, + "grad_norm": 0.9565346811431111, + "learning_rate": 2.425926312166649e-06, + "loss": 0.6106, + "num_input_tokens_seen": 79303500, + "step": 3732 + }, + { + "epoch": 0.4488667107557266, + "flos": 15183144009720.0, + "grad_norm": 2.9240988362068494, + "learning_rate": 2.42516517766223e-06, + "loss": 0.7049, + "num_input_tokens_seen": 79321300, + "step": 3733 + }, + { + "epoch": 0.44898695364636565, + "flos": 17539077444360.0, + "grad_norm": 9.004314236282664, + "learning_rate": 2.4244039786605907e-06, + "loss": 0.6641, + "num_input_tokens_seen": 79342025, + "step": 3734 + }, + { + "epoch": 0.44910719653700476, + "flos": 13596390185040.0, + "grad_norm": 3.081786326067025, + "learning_rate": 2.4236427152772055e-06, + "loss": 0.8075, + "num_input_tokens_seen": 79360150, + "step": 3735 + }, + { + "epoch": 0.4492274394276438, + "flos": 42025380370440.0, + "grad_norm": 0.9304578458593894, + "learning_rate": 2.422881387627557e-06, + "loss": 0.6016, + "num_input_tokens_seen": 79412320, + "step": 3736 + }, + { + "epoch": 0.4493476823182829, + "flos": 17022852152520.0, + "grad_norm": 1.8859562251086643, + "learning_rate": 2.422119995827139e-06, + "loss": 0.757, + "num_input_tokens_seen": 79432165, + "step": 3737 + }, + { + "epoch": 0.44946792520892204, + "flos": 11517404826840.0, + "grad_norm": 4.643913773412217, + "learning_rate": 2.4213585399914528e-06, + "loss": 0.7223, + "num_input_tokens_seen": 79449090, + "step": 3738 + }, + { + "epoch": 0.4495881680995611, + "flos": 14325967676400.0, + "grad_norm": 3.3670456008927485, + "learning_rate": 2.4205970202360113e-06, + "loss": 0.8394, + "num_input_tokens_seen": 79468375, + "step": 3739 + }, + { + "epoch": 0.4497084109902002, + "flos": 19078223671920.0, + "grad_norm": 2.553637180758486, + "learning_rate": 2.4198354366763354e-06, + "loss": 0.7677, + "num_input_tokens_seen": 79486735, + "step": 3740 + }, + { + "epoch": 0.4498286538808393, + "flos": 10761490740120.0, + "grad_norm": 3.582293219019243, + "learning_rate": 2.4190737894279587e-06, + "loss": 0.7617, + "num_input_tokens_seen": 79503825, + "step": 3741 + }, + { + "epoch": 0.44994889677147837, + "flos": 11247295301760.0, + "grad_norm": 3.3406094928092487, + "learning_rate": 2.4183120786064203e-06, + "loss": 0.7894, + "num_input_tokens_seen": 79520420, + "step": 3742 + }, + { + "epoch": 0.4500691396621175, + "flos": 15945295108560.0, + "grad_norm": 3.9686178350594497, + "learning_rate": 2.417550304327273e-06, + "loss": 0.8433, + "num_input_tokens_seen": 79538180, + "step": 3743 + }, + { + "epoch": 0.4501893825527566, + "flos": 23778344696040.0, + "grad_norm": 2.1457032716710422, + "learning_rate": 2.4167884667060763e-06, + "loss": 0.7413, + "num_input_tokens_seen": 79560610, + "step": 3744 + }, + { + "epoch": 0.45030962544339564, + "flos": 12057528897120.0, + "grad_norm": 3.1037122833381545, + "learning_rate": 2.4160265658584e-06, + "loss": 0.8587, + "num_input_tokens_seen": 79575220, + "step": 3745 + }, + { + "epoch": 0.45042986833403476, + "flos": 14298491322480.0, + "grad_norm": 4.029294744834022, + "learning_rate": 2.4152646018998253e-06, + "loss": 0.6609, + "num_input_tokens_seen": 79593890, + "step": 3746 + }, + { + "epoch": 0.45055011122467387, + "flos": 16917949010520.0, + "grad_norm": 3.6004218490122413, + "learning_rate": 2.4145025749459403e-06, + "loss": 0.6977, + "num_input_tokens_seen": 79614635, + "step": 3747 + }, + { + "epoch": 0.4506703541153129, + "flos": 14565624811320.0, + "grad_norm": 2.6362460694661327, + "learning_rate": 2.413740485112344e-06, + "loss": 0.6864, + "num_input_tokens_seen": 79632695, + "step": 3748 + }, + { + "epoch": 0.45079059700595203, + "flos": 14244013514040.0, + "grad_norm": 3.950895261109539, + "learning_rate": 2.412978332514646e-06, + "loss": 0.8067, + "num_input_tokens_seen": 79651195, + "step": 3749 + }, + { + "epoch": 0.4509108398965911, + "flos": 20265844431360.0, + "grad_norm": 3.972564666202508, + "learning_rate": 2.4122161172684623e-06, + "loss": 0.6938, + "num_input_tokens_seen": 79671710, + "step": 3750 + }, + { + "epoch": 0.4510310827872302, + "flos": 15351326429040.0, + "grad_norm": 2.4777542279449594, + "learning_rate": 2.4114538394894216e-06, + "loss": 0.8113, + "num_input_tokens_seen": 79689070, + "step": 3751 + }, + { + "epoch": 0.4511513256778693, + "flos": 11814610786320.0, + "grad_norm": 3.1854236796476325, + "learning_rate": 2.410691499293161e-06, + "loss": 0.8204, + "num_input_tokens_seen": 79706945, + "step": 3752 + }, + { + "epoch": 0.45127156856850836, + "flos": 18452251264200.0, + "grad_norm": 2.3189936013801735, + "learning_rate": 2.409929096795326e-06, + "loss": 0.7207, + "num_input_tokens_seen": 79727035, + "step": 3753 + }, + { + "epoch": 0.4513918114591475, + "flos": 14919809716080.0, + "grad_norm": 2.379082634060343, + "learning_rate": 2.409166632111573e-06, + "loss": 0.7725, + "num_input_tokens_seen": 79744890, + "step": 3754 + }, + { + "epoch": 0.4515120543497866, + "flos": 19534399002360.0, + "grad_norm": 2.7613437943007906, + "learning_rate": 2.4084041053575674e-06, + "loss": 0.7705, + "num_input_tokens_seen": 79764030, + "step": 3755 + }, + { + "epoch": 0.45163229724042564, + "flos": 15054943628520.0, + "grad_norm": 2.371156895910605, + "learning_rate": 2.4076415166489834e-06, + "loss": 0.7119, + "num_input_tokens_seen": 79783160, + "step": 3756 + }, + { + "epoch": 0.45175254013106475, + "flos": 15864417384840.0, + "grad_norm": 3.4680201131992012, + "learning_rate": 2.406878866101506e-06, + "loss": 0.7864, + "num_input_tokens_seen": 79801845, + "step": 3757 + }, + { + "epoch": 0.45187278302170386, + "flos": 13784608867680.0, + "grad_norm": 3.773487920009616, + "learning_rate": 2.4061161538308273e-06, + "loss": 0.7687, + "num_input_tokens_seen": 79818410, + "step": 3758 + }, + { + "epoch": 0.4519930259123429, + "flos": 13564639736520.0, + "grad_norm": 2.8514844073746026, + "learning_rate": 2.4053533799526523e-06, + "loss": 0.8796, + "num_input_tokens_seen": 79833850, + "step": 3759 + }, + { + "epoch": 0.452113268802982, + "flos": 18457253537880.0, + "grad_norm": 3.8210040443636615, + "learning_rate": 2.404590544582691e-06, + "loss": 0.8373, + "num_input_tokens_seen": 79851805, + "step": 3760 + }, + { + "epoch": 0.45223351169362114, + "flos": 28955342173080.0, + "grad_norm": 2.9917431693182004, + "learning_rate": 2.403827647836666e-06, + "loss": 0.7944, + "num_input_tokens_seen": 79872080, + "step": 3761 + }, + { + "epoch": 0.4523537545842602, + "flos": 15785724198360.0, + "grad_norm": 2.5364527039723264, + "learning_rate": 2.4030646898303075e-06, + "loss": 0.6749, + "num_input_tokens_seen": 79893290, + "step": 3762 + }, + { + "epoch": 0.4524739974748993, + "flos": 20860351330200.0, + "grad_norm": 2.5295624935938195, + "learning_rate": 2.4023016706793566e-06, + "loss": 0.8008, + "num_input_tokens_seen": 79912805, + "step": 3763 + }, + { + "epoch": 0.4525942403655384, + "flos": 45372579272160.0, + "grad_norm": 0.8234562895221771, + "learning_rate": 2.401538590499561e-06, + "loss": 0.5955, + "num_input_tokens_seen": 79972980, + "step": 3764 + }, + { + "epoch": 0.45271448325617747, + "flos": 20187467844480.0, + "grad_norm": 6.504911181278216, + "learning_rate": 2.400775449406682e-06, + "loss": 0.688, + "num_input_tokens_seen": 79995895, + "step": 3765 + }, + { + "epoch": 0.4528347261468166, + "flos": 16429168412640.0, + "grad_norm": 2.323261172716932, + "learning_rate": 2.400012247516485e-06, + "loss": 0.7193, + "num_input_tokens_seen": 80016180, + "step": 3766 + }, + { + "epoch": 0.45295496903745563, + "flos": 15431191034040.0, + "grad_norm": 2.210889188959812, + "learning_rate": 2.3992489849447484e-06, + "loss": 0.8744, + "num_input_tokens_seen": 80034355, + "step": 3767 + }, + { + "epoch": 0.45307521192809475, + "flos": 16998288514920.0, + "grad_norm": 2.525703758894073, + "learning_rate": 2.3984856618072584e-06, + "loss": 0.7766, + "num_input_tokens_seen": 80054110, + "step": 3768 + }, + { + "epoch": 0.45319545481873386, + "flos": 11328711244800.0, + "grad_norm": 2.6073422309055867, + "learning_rate": 2.3977222782198098e-06, + "loss": 0.7201, + "num_input_tokens_seen": 80072465, + "step": 3769 + }, + { + "epoch": 0.4533156977093729, + "flos": 16053744166080.0, + "grad_norm": 3.7128601875642855, + "learning_rate": 2.3969588342982077e-06, + "loss": 0.7351, + "num_input_tokens_seen": 80091560, + "step": 3770 + }, + { + "epoch": 0.453435940600012, + "flos": 17753822682120.0, + "grad_norm": 2.4037268154158897, + "learning_rate": 2.396195330158267e-06, + "loss": 0.7125, + "num_input_tokens_seen": 80111170, + "step": 3771 + }, + { + "epoch": 0.45355618349065113, + "flos": 17160550521720.0, + "grad_norm": 3.172520914692699, + "learning_rate": 2.3954317659158094e-06, + "loss": 0.7758, + "num_input_tokens_seen": 80131225, + "step": 3772 + }, + { + "epoch": 0.4536764263812902, + "flos": 49332267930000.0, + "grad_norm": 0.9052791327798363, + "learning_rate": 2.394668141686667e-06, + "loss": 0.6164, + "num_input_tokens_seen": 80192910, + "step": 3773 + }, + { + "epoch": 0.4537966692719293, + "flos": 31443180204240.0, + "grad_norm": 2.27561138866684, + "learning_rate": 2.3939044575866813e-06, + "loss": 0.6629, + "num_input_tokens_seen": 80215380, + "step": 3774 + }, + { + "epoch": 0.4539169121625684, + "flos": 26121519166800.0, + "grad_norm": 2.9575765653681136, + "learning_rate": 2.3931407137317024e-06, + "loss": 0.7384, + "num_input_tokens_seen": 80235255, + "step": 3775 + }, + { + "epoch": 0.45403715505320746, + "flos": 13515227521680.0, + "grad_norm": 2.797838354162758, + "learning_rate": 2.3923769102375907e-06, + "loss": 0.8283, + "num_input_tokens_seen": 80253840, + "step": 3776 + }, + { + "epoch": 0.4541573979438466, + "flos": 18348709500480.0, + "grad_norm": 2.6017004211340984, + "learning_rate": 2.391613047220213e-06, + "loss": 0.7733, + "num_input_tokens_seen": 80273460, + "step": 3777 + }, + { + "epoch": 0.4542776408344857, + "flos": 13380315228960.0, + "grad_norm": 2.4382825720043075, + "learning_rate": 2.390849124795447e-06, + "loss": 0.7846, + "num_input_tokens_seen": 80289180, + "step": 3778 + }, + { + "epoch": 0.45439788372512474, + "flos": 15132971955840.0, + "grad_norm": 3.4649179272200095, + "learning_rate": 2.3900851430791804e-06, + "loss": 0.825, + "num_input_tokens_seen": 80306920, + "step": 3779 + }, + { + "epoch": 0.45451812661576385, + "flos": 16323600411480.0, + "grad_norm": 2.8805331184431386, + "learning_rate": 2.389321102187307e-06, + "loss": 0.8348, + "num_input_tokens_seen": 80325420, + "step": 3780 + }, + { + "epoch": 0.4546383695064029, + "flos": 15919591712400.0, + "grad_norm": 3.034479053414334, + "learning_rate": 2.3885570022357326e-06, + "loss": 0.8106, + "num_input_tokens_seen": 80344270, + "step": 3781 + }, + { + "epoch": 0.454758612397042, + "flos": 47363821186680.0, + "grad_norm": 1.404943495052004, + "learning_rate": 2.38779284334037e-06, + "loss": 0.6214, + "num_input_tokens_seen": 80408965, + "step": 3782 + }, + { + "epoch": 0.4548788552876811, + "flos": 20019507044880.0, + "grad_norm": 2.9157322249571904, + "learning_rate": 2.387028625617141e-06, + "loss": 0.77, + "num_input_tokens_seen": 80427900, + "step": 3783 + }, + { + "epoch": 0.4549990981783202, + "flos": 16729002148800.0, + "grad_norm": 2.694201962111691, + "learning_rate": 2.3862643491819766e-06, + "loss": 0.8377, + "num_input_tokens_seen": 80446185, + "step": 3784 + }, + { + "epoch": 0.4551193410689593, + "flos": 17026113128400.0, + "grad_norm": 3.463166976203417, + "learning_rate": 2.3855000141508186e-06, + "loss": 0.8204, + "num_input_tokens_seen": 80466060, + "step": 3785 + }, + { + "epoch": 0.4552395839595984, + "flos": 15189950901120.0, + "grad_norm": 3.8475604514961788, + "learning_rate": 2.3847356206396143e-06, + "loss": 0.8138, + "num_input_tokens_seen": 80483090, + "step": 3786 + }, + { + "epoch": 0.45535982685023746, + "flos": 17025163329600.0, + "grad_norm": 1.884736846545997, + "learning_rate": 2.3839711687643227e-06, + "loss": 0.7735, + "num_input_tokens_seen": 80504035, + "step": 3787 + }, + { + "epoch": 0.45548006974087657, + "flos": 14352589211400.0, + "grad_norm": 4.711519879075419, + "learning_rate": 2.38320665864091e-06, + "loss": 0.7099, + "num_input_tokens_seen": 80523105, + "step": 3788 + }, + { + "epoch": 0.4556003126315157, + "flos": 14649890150760.0, + "grad_norm": 4.070806843031426, + "learning_rate": 2.3824420903853516e-06, + "loss": 0.8074, + "num_input_tokens_seen": 80541290, + "step": 3789 + }, + { + "epoch": 0.45572055552215474, + "flos": 16803959460000.0, + "grad_norm": 3.336951930160832, + "learning_rate": 2.3816774641136324e-06, + "loss": 0.7874, + "num_input_tokens_seen": 80558265, + "step": 3790 + }, + { + "epoch": 0.45584079841279385, + "flos": 24316790788440.0, + "grad_norm": 1.9724033306883566, + "learning_rate": 2.380912779941745e-06, + "loss": 0.7072, + "num_input_tokens_seen": 80581105, + "step": 3791 + }, + { + "epoch": 0.45596104130343296, + "flos": 19996463085360.0, + "grad_norm": 3.14746623996024, + "learning_rate": 2.3801480379856918e-06, + "loss": 0.7929, + "num_input_tokens_seen": 80602535, + "step": 3792 + }, + { + "epoch": 0.456081284194072, + "flos": 15783634641000.0, + "grad_norm": 3.140610728065364, + "learning_rate": 2.379383238361484e-06, + "loss": 0.8217, + "num_input_tokens_seen": 80621615, + "step": 3793 + }, + { + "epoch": 0.4562015270847111, + "flos": 26395491207000.0, + "grad_norm": 2.710911115798301, + "learning_rate": 2.3786183811851407e-06, + "loss": 0.7743, + "num_input_tokens_seen": 80642040, + "step": 3794 + }, + { + "epoch": 0.45632176997535023, + "flos": 9870537720840.0, + "grad_norm": 2.533330462688359, + "learning_rate": 2.3778534665726892e-06, + "loss": 0.787, + "num_input_tokens_seen": 80658590, + "step": 3795 + }, + { + "epoch": 0.4564420128659893, + "flos": 23965581919920.0, + "grad_norm": 2.4533040169678135, + "learning_rate": 2.377088494640168e-06, + "loss": 0.7119, + "num_input_tokens_seen": 80680060, + "step": 3796 + }, + { + "epoch": 0.4565622557566284, + "flos": 14894264619720.0, + "grad_norm": 2.345126752213567, + "learning_rate": 2.3763234655036216e-06, + "loss": 0.7677, + "num_input_tokens_seen": 80698980, + "step": 3797 + }, + { + "epoch": 0.45668249864726745, + "flos": 18591880890960.0, + "grad_norm": 2.5527334410265126, + "learning_rate": 2.3755583792791046e-06, + "loss": 0.8564, + "num_input_tokens_seen": 80718570, + "step": 3798 + }, + { + "epoch": 0.45680274153790656, + "flos": 11328457965120.0, + "grad_norm": 2.5099050757707717, + "learning_rate": 2.3747932360826803e-06, + "loss": 0.7277, + "num_input_tokens_seen": 80735220, + "step": 3799 + }, + { + "epoch": 0.4569229844285457, + "flos": 14461323208560.0, + "grad_norm": 6.211704883312877, + "learning_rate": 2.3740280360304205e-06, + "loss": 0.8015, + "num_input_tokens_seen": 80752665, + "step": 3800 + }, + { + "epoch": 0.45704322731918473, + "flos": 17643790626600.0, + "grad_norm": 2.378659891736143, + "learning_rate": 2.3732627792384038e-06, + "loss": 0.6699, + "num_input_tokens_seen": 80773455, + "step": 3801 + }, + { + "epoch": 0.45716347020982384, + "flos": 22989888661800.0, + "grad_norm": 2.9094748704201665, + "learning_rate": 2.3724974658227207e-06, + "loss": 0.7378, + "num_input_tokens_seen": 80793965, + "step": 3802 + }, + { + "epoch": 0.45728371310046295, + "flos": 19425380065560.0, + "grad_norm": 3.6103321474237906, + "learning_rate": 2.3717320958994687e-06, + "loss": 0.6904, + "num_input_tokens_seen": 80811245, + "step": 3803 + }, + { + "epoch": 0.457403955991102, + "flos": 13081842871080.0, + "grad_norm": 2.251264428786701, + "learning_rate": 2.3709666695847534e-06, + "loss": 0.6912, + "num_input_tokens_seen": 80829145, + "step": 3804 + }, + { + "epoch": 0.4575241988817411, + "flos": 31066647859080.0, + "grad_norm": 2.184467284254677, + "learning_rate": 2.370201186994689e-06, + "loss": 0.6914, + "num_input_tokens_seen": 80852550, + "step": 3805 + }, + { + "epoch": 0.45764444177238023, + "flos": 22101943338720.0, + "grad_norm": 5.110727673581877, + "learning_rate": 2.369435648245399e-06, + "loss": 0.6759, + "num_input_tokens_seen": 80872485, + "step": 3806 + }, + { + "epoch": 0.4577646846630193, + "flos": 17619290308920.0, + "grad_norm": 3.19283894722357, + "learning_rate": 2.368670053453015e-06, + "loss": 0.8414, + "num_input_tokens_seen": 80893895, + "step": 3807 + }, + { + "epoch": 0.4578849275536584, + "flos": 12703790847840.0, + "grad_norm": 3.869247103847823, + "learning_rate": 2.3679044027336757e-06, + "loss": 0.7061, + "num_input_tokens_seen": 80909505, + "step": 3808 + }, + { + "epoch": 0.4580051704442975, + "flos": 9810677719200.0, + "grad_norm": 4.217040669336251, + "learning_rate": 2.3671386962035326e-06, + "loss": 0.6671, + "num_input_tokens_seen": 80926695, + "step": 3809 + }, + { + "epoch": 0.45812541333493656, + "flos": 13161739136040.0, + "grad_norm": 9.044307575028547, + "learning_rate": 2.3663729339787405e-06, + "loss": 0.6683, + "num_input_tokens_seen": 80943350, + "step": 3810 + }, + { + "epoch": 0.45824565622557567, + "flos": 14777837252280.0, + "grad_norm": 3.4657852287161295, + "learning_rate": 2.365607116175466e-06, + "loss": 0.7016, + "num_input_tokens_seen": 80958495, + "step": 3811 + }, + { + "epoch": 0.4583658991162148, + "flos": 14514471298680.0, + "grad_norm": 3.3096046029501247, + "learning_rate": 2.3648412429098825e-06, + "loss": 0.6334, + "num_input_tokens_seen": 80976565, + "step": 3812 + }, + { + "epoch": 0.45848614200685384, + "flos": 15376333306080.0, + "grad_norm": 2.4226538605541363, + "learning_rate": 2.364075314298172e-06, + "loss": 0.7981, + "num_input_tokens_seen": 80993740, + "step": 3813 + }, + { + "epoch": 0.45860638489749295, + "flos": 15405867557400.0, + "grad_norm": 2.1634288094156053, + "learning_rate": 2.3633093304565267e-06, + "loss": 0.6789, + "num_input_tokens_seen": 81012515, + "step": 3814 + }, + { + "epoch": 0.458726627788132, + "flos": 19672034051640.0, + "grad_norm": 2.7214390402278927, + "learning_rate": 2.3625432915011443e-06, + "loss": 0.6096, + "num_input_tokens_seen": 81034145, + "step": 3815 + }, + { + "epoch": 0.4588468706787711, + "flos": 17647304882160.0, + "grad_norm": 4.914474106184348, + "learning_rate": 2.3617771975482334e-06, + "loss": 0.6383, + "num_input_tokens_seen": 81052695, + "step": 3816 + }, + { + "epoch": 0.4589671135694102, + "flos": 13051675420560.0, + "grad_norm": 2.6258480881841697, + "learning_rate": 2.3610110487140083e-06, + "loss": 0.7298, + "num_input_tokens_seen": 81070850, + "step": 3817 + }, + { + "epoch": 0.4590873564600493, + "flos": 18778421595720.0, + "grad_norm": 2.013980283161465, + "learning_rate": 2.360244845114695e-06, + "loss": 0.8034, + "num_input_tokens_seen": 81090190, + "step": 3818 + }, + { + "epoch": 0.4592075993506884, + "flos": 13515290841600.0, + "grad_norm": 3.6586517531403335, + "learning_rate": 2.3594785868665245e-06, + "loss": 0.6726, + "num_input_tokens_seen": 81106215, + "step": 3819 + }, + { + "epoch": 0.4593278422413275, + "flos": 15081533503560.0, + "grad_norm": 2.905026085846432, + "learning_rate": 2.3587122740857386e-06, + "loss": 0.7927, + "num_input_tokens_seen": 81123035, + "step": 3820 + }, + { + "epoch": 0.45944808513196655, + "flos": 15619536356520.0, + "grad_norm": 2.235437207395732, + "learning_rate": 2.357945906888586e-06, + "loss": 0.767, + "num_input_tokens_seen": 81142195, + "step": 3821 + }, + { + "epoch": 0.45956832802260567, + "flos": 15670278289680.0, + "grad_norm": 2.8294795886225512, + "learning_rate": 2.357179485391324e-06, + "loss": 0.7698, + "num_input_tokens_seen": 81159770, + "step": 3822 + }, + { + "epoch": 0.4596885709132448, + "flos": 16377824940240.0, + "grad_norm": 2.3136461871252525, + "learning_rate": 2.3564130097102173e-06, + "loss": 0.8404, + "num_input_tokens_seen": 81179145, + "step": 3823 + }, + { + "epoch": 0.45980881380388383, + "flos": 21261099053400.0, + "grad_norm": 3.1168368723151114, + "learning_rate": 2.355646479961541e-06, + "loss": 0.727, + "num_input_tokens_seen": 81198175, + "step": 3824 + }, + { + "epoch": 0.45992905669452294, + "flos": 24528211730400.0, + "grad_norm": 2.7740531130699897, + "learning_rate": 2.354879896261576e-06, + "loss": 0.7046, + "num_input_tokens_seen": 81218105, + "step": 3825 + }, + { + "epoch": 0.46004929958516205, + "flos": 26690765908920.0, + "grad_norm": 3.7495840477726374, + "learning_rate": 2.3541132587266133e-06, + "loss": 0.556, + "num_input_tokens_seen": 81240545, + "step": 3826 + }, + { + "epoch": 0.4601695424758011, + "flos": 12570809812680.0, + "grad_norm": 2.404673369467143, + "learning_rate": 2.3533465674729515e-06, + "loss": 0.6795, + "num_input_tokens_seen": 81257495, + "step": 3827 + }, + { + "epoch": 0.4602897853664402, + "flos": 11571692675520.0, + "grad_norm": 3.0193488496513803, + "learning_rate": 2.352579822616895e-06, + "loss": 0.7155, + "num_input_tokens_seen": 81274650, + "step": 3828 + }, + { + "epoch": 0.4604100282570793, + "flos": 18620433683520.0, + "grad_norm": 2.7509259485714854, + "learning_rate": 2.351813024274761e-06, + "loss": 0.7665, + "num_input_tokens_seen": 81295725, + "step": 3829 + }, + { + "epoch": 0.4605302711477184, + "flos": 20260905477600.0, + "grad_norm": 2.5171731485517714, + "learning_rate": 2.3510461725628693e-06, + "loss": 0.7135, + "num_input_tokens_seen": 81315910, + "step": 3830 + }, + { + "epoch": 0.4606505140383575, + "flos": 17455761903720.0, + "grad_norm": 2.4782834015643243, + "learning_rate": 2.350279267597554e-06, + "loss": 0.692, + "num_input_tokens_seen": 81336270, + "step": 3831 + }, + { + "epoch": 0.46077075692899655, + "flos": 11733416463000.0, + "grad_norm": 3.7630777568097615, + "learning_rate": 2.3495123094951515e-06, + "loss": 0.8102, + "num_input_tokens_seen": 81354335, + "step": 3832 + }, + { + "epoch": 0.46089099981963566, + "flos": 35927036312520.0, + "grad_norm": 3.336626682962893, + "learning_rate": 2.34874529837201e-06, + "loss": 0.732, + "num_input_tokens_seen": 81377600, + "step": 3833 + }, + { + "epoch": 0.46101124271027477, + "flos": 13947915653160.0, + "grad_norm": 2.9277570137029167, + "learning_rate": 2.347978234344483e-06, + "loss": 0.7847, + "num_input_tokens_seen": 81393525, + "step": 3834 + }, + { + "epoch": 0.4611314856009138, + "flos": 28933184692440.0, + "grad_norm": 2.895050949412605, + "learning_rate": 2.347211117528935e-06, + "loss": 0.6844, + "num_input_tokens_seen": 81415545, + "step": 3835 + }, + { + "epoch": 0.46125172849155294, + "flos": 15214102959240.0, + "grad_norm": 2.056086191070273, + "learning_rate": 2.3464439480417374e-06, + "loss": 0.7019, + "num_input_tokens_seen": 81434785, + "step": 3836 + }, + { + "epoch": 0.46137197138219205, + "flos": 13083077609520.0, + "grad_norm": 3.998014434101408, + "learning_rate": 2.3456767259992676e-06, + "loss": 0.7428, + "num_input_tokens_seen": 81452150, + "step": 3837 + }, + { + "epoch": 0.4614922142728311, + "flos": 12273350573520.0, + "grad_norm": 3.7891755077381126, + "learning_rate": 2.3449094515179135e-06, + "loss": 0.8539, + "num_input_tokens_seen": 81469330, + "step": 3838 + }, + { + "epoch": 0.4616124571634702, + "flos": 19509962004600.0, + "grad_norm": 2.238040078902831, + "learning_rate": 2.34414212471407e-06, + "loss": 0.8046, + "num_input_tokens_seen": 81488845, + "step": 3839 + }, + { + "epoch": 0.4617327000541093, + "flos": 14866313366400.0, + "grad_norm": 2.5426823092864455, + "learning_rate": 2.3433747457041394e-06, + "loss": 0.7219, + "num_input_tokens_seen": 81507270, + "step": 3840 + }, + { + "epoch": 0.4618529429447484, + "flos": 21697554720120.0, + "grad_norm": 4.641093940032358, + "learning_rate": 2.342607314604533e-06, + "loss": 0.8258, + "num_input_tokens_seen": 81526740, + "step": 3841 + }, + { + "epoch": 0.4619731858353875, + "flos": 14456162635080.0, + "grad_norm": 13.77987255729575, + "learning_rate": 2.3418398315316694e-06, + "loss": 0.8276, + "num_input_tokens_seen": 81544280, + "step": 3842 + }, + { + "epoch": 0.4620934287260266, + "flos": 13838010237480.0, + "grad_norm": 3.7374621284122624, + "learning_rate": 2.3410722966019755e-06, + "loss": 0.7584, + "num_input_tokens_seen": 81559115, + "step": 3843 + }, + { + "epoch": 0.46221367161666566, + "flos": 27448484613360.0, + "grad_norm": 2.5377414494538493, + "learning_rate": 2.3403047099318848e-06, + "loss": 0.642, + "num_input_tokens_seen": 81582905, + "step": 3844 + }, + { + "epoch": 0.46233391450730477, + "flos": 10491222915240.0, + "grad_norm": 2.900586917284445, + "learning_rate": 2.3395370716378405e-06, + "loss": 0.7258, + "num_input_tokens_seen": 81600070, + "step": 3845 + }, + { + "epoch": 0.4624541573979438, + "flos": 16459652462760.0, + "grad_norm": 3.469753085979049, + "learning_rate": 2.338769381836292e-06, + "loss": 0.7126, + "num_input_tokens_seen": 81619400, + "step": 3846 + }, + { + "epoch": 0.46257440028858293, + "flos": 10518256029720.0, + "grad_norm": 2.924785449644578, + "learning_rate": 2.3380016406436984e-06, + "loss": 0.6997, + "num_input_tokens_seen": 81636600, + "step": 3847 + }, + { + "epoch": 0.46269464317922204, + "flos": 17080685916720.0, + "grad_norm": 3.068393253015553, + "learning_rate": 2.337233848176524e-06, + "loss": 0.8029, + "num_input_tokens_seen": 81654090, + "step": 3848 + }, + { + "epoch": 0.4628148860698611, + "flos": 13542957155280.0, + "grad_norm": 3.27700109134082, + "learning_rate": 2.3364660045512435e-06, + "loss": 0.8049, + "num_input_tokens_seen": 81672570, + "step": 3849 + }, + { + "epoch": 0.4629351289605002, + "flos": 52119084878400.0, + "grad_norm": 0.7749918337360953, + "learning_rate": 2.335698109884337e-06, + "loss": 0.6199, + "num_input_tokens_seen": 81737495, + "step": 3850 + }, + { + "epoch": 0.4630553718511393, + "flos": 43993135086000.0, + "grad_norm": 0.8623640962313085, + "learning_rate": 2.334930164292294e-06, + "loss": 0.6477, + "num_input_tokens_seen": 81799765, + "step": 3851 + }, + { + "epoch": 0.4631756147417784, + "flos": 11624999065440.0, + "grad_norm": 2.8167156414612204, + "learning_rate": 2.334162167891612e-06, + "loss": 0.7727, + "num_input_tokens_seen": 81816750, + "step": 3852 + }, + { + "epoch": 0.4632958576324175, + "flos": 12004349147040.0, + "grad_norm": 5.101242078734316, + "learning_rate": 2.333394120798795e-06, + "loss": 0.7167, + "num_input_tokens_seen": 81835205, + "step": 3853 + }, + { + "epoch": 0.4634161005230566, + "flos": 16351108425360.0, + "grad_norm": 2.8688863352041833, + "learning_rate": 2.3326260231303545e-06, + "loss": 0.7194, + "num_input_tokens_seen": 81853525, + "step": 3854 + }, + { + "epoch": 0.46353634341369565, + "flos": 11463465237720.0, + "grad_norm": 1.801135245591679, + "learning_rate": 2.331857875002811e-06, + "loss": 0.8627, + "num_input_tokens_seen": 81871845, + "step": 3855 + }, + { + "epoch": 0.46365658630433476, + "flos": 20779125346920.0, + "grad_norm": 17.042846539356695, + "learning_rate": 2.3310896765326916e-06, + "loss": 0.7434, + "num_input_tokens_seen": 81892565, + "step": 3856 + }, + { + "epoch": 0.46377682919497387, + "flos": 18024755366160.0, + "grad_norm": 2.1715598117813375, + "learning_rate": 2.330321427836531e-06, + "loss": 0.8248, + "num_input_tokens_seen": 81914155, + "step": 3857 + }, + { + "epoch": 0.4638970720856129, + "flos": 13919331200640.0, + "grad_norm": 2.9695947348761744, + "learning_rate": 2.3295531290308733e-06, + "loss": 0.8159, + "num_input_tokens_seen": 81932025, + "step": 3858 + }, + { + "epoch": 0.46401731497625204, + "flos": 13481862415200.0, + "grad_norm": 4.681599272671458, + "learning_rate": 2.3287847802322678e-06, + "loss": 0.7237, + "num_input_tokens_seen": 81947315, + "step": 3859 + }, + { + "epoch": 0.4641375578668911, + "flos": 19104433627440.0, + "grad_norm": 3.5536631690764513, + "learning_rate": 2.3280163815572723e-06, + "loss": 0.8174, + "num_input_tokens_seen": 81967630, + "step": 3860 + }, + { + "epoch": 0.4642578007575302, + "flos": 14296148485440.0, + "grad_norm": 3.2996916829200185, + "learning_rate": 2.3272479331224522e-06, + "loss": 0.7482, + "num_input_tokens_seen": 81984780, + "step": 3861 + }, + { + "epoch": 0.4643780436481693, + "flos": 20673209086200.0, + "grad_norm": 2.289238099533332, + "learning_rate": 2.3264794350443817e-06, + "loss": 0.7624, + "num_input_tokens_seen": 82006595, + "step": 3862 + }, + { + "epoch": 0.46449828653880837, + "flos": 18592672389960.0, + "grad_norm": 2.392106339007859, + "learning_rate": 2.3257108874396396e-06, + "loss": 0.7751, + "num_input_tokens_seen": 82027410, + "step": 3863 + }, + { + "epoch": 0.4646185294294475, + "flos": 11679666833640.0, + "grad_norm": 3.573074574590961, + "learning_rate": 2.3249422904248152e-06, + "loss": 0.7195, + "num_input_tokens_seen": 82045565, + "step": 3864 + }, + { + "epoch": 0.4647387723200866, + "flos": 19324117818960.0, + "grad_norm": 1.9558561227145543, + "learning_rate": 2.324173644116504e-06, + "loss": 0.8616, + "num_input_tokens_seen": 82068135, + "step": 3865 + }, + { + "epoch": 0.46485901521072565, + "flos": 19913242524600.0, + "grad_norm": 10.930567697089847, + "learning_rate": 2.3234049486313087e-06, + "loss": 0.8018, + "num_input_tokens_seen": 82089305, + "step": 3866 + }, + { + "epoch": 0.46497925810136476, + "flos": 17781868915320.0, + "grad_norm": 2.480514694148814, + "learning_rate": 2.322636204085839e-06, + "loss": 0.7488, + "num_input_tokens_seen": 82109095, + "step": 3867 + }, + { + "epoch": 0.46509950099200387, + "flos": 11841580580880.0, + "grad_norm": 2.8800204022002767, + "learning_rate": 2.3218674105967143e-06, + "loss": 0.7498, + "num_input_tokens_seen": 82127080, + "step": 3868 + }, + { + "epoch": 0.4652197438826429, + "flos": 17162228499600.0, + "grad_norm": 2.1575813053587107, + "learning_rate": 2.3210985682805593e-06, + "loss": 0.8157, + "num_input_tokens_seen": 82148580, + "step": 3869 + }, + { + "epoch": 0.46533998677328203, + "flos": 11814357506640.0, + "grad_norm": 3.0379641394140457, + "learning_rate": 2.320329677254007e-06, + "loss": 0.6544, + "num_input_tokens_seen": 82165630, + "step": 3870 + }, + { + "epoch": 0.46546022966392114, + "flos": 15459522206880.0, + "grad_norm": 3.230194190834051, + "learning_rate": 2.319560737633697e-06, + "loss": 0.705, + "num_input_tokens_seen": 82184070, + "step": 3871 + }, + { + "epoch": 0.4655804725545602, + "flos": 30283035798720.0, + "grad_norm": 2.243939396507346, + "learning_rate": 2.3187917495362775e-06, + "loss": 0.6721, + "num_input_tokens_seen": 82208200, + "step": 3872 + }, + { + "epoch": 0.4657007154451993, + "flos": 14296655044800.0, + "grad_norm": 3.101309278322162, + "learning_rate": 2.318022713078403e-06, + "loss": 0.7412, + "num_input_tokens_seen": 82222500, + "step": 3873 + }, + { + "epoch": 0.4658209583358384, + "flos": 11296612536720.0, + "grad_norm": 2.826062287985265, + "learning_rate": 2.3172536283767354e-06, + "loss": 0.8358, + "num_input_tokens_seen": 82235980, + "step": 3874 + }, + { + "epoch": 0.4659412012264775, + "flos": 10842526763640.0, + "grad_norm": 3.120984205631967, + "learning_rate": 2.3164844955479447e-06, + "loss": 0.7887, + "num_input_tokens_seen": 82251510, + "step": 3875 + }, + { + "epoch": 0.4660614441171166, + "flos": 17889716433600.0, + "grad_norm": 2.7632596520638666, + "learning_rate": 2.3157153147087082e-06, + "loss": 0.6814, + "num_input_tokens_seen": 82273120, + "step": 3876 + }, + { + "epoch": 0.46618168700775564, + "flos": 16162098243720.0, + "grad_norm": 2.3767745766701696, + "learning_rate": 2.314946085975709e-06, + "loss": 0.8251, + "num_input_tokens_seen": 82291820, + "step": 3877 + }, + { + "epoch": 0.46630192989839475, + "flos": 19185849570480.0, + "grad_norm": 3.389592448579661, + "learning_rate": 2.3141768094656393e-06, + "loss": 0.8097, + "num_input_tokens_seen": 82310115, + "step": 3878 + }, + { + "epoch": 0.46642217278903386, + "flos": 8331011573760.0, + "grad_norm": 2.940731949031929, + "learning_rate": 2.3134074852951966e-06, + "loss": 0.8174, + "num_input_tokens_seen": 82326425, + "step": 3879 + }, + { + "epoch": 0.4665424156796729, + "flos": 23722378869480.0, + "grad_norm": 3.397383883908127, + "learning_rate": 2.312638113581088e-06, + "loss": 0.7749, + "num_input_tokens_seen": 82346630, + "step": 3880 + }, + { + "epoch": 0.46666265857031203, + "flos": 13456317318840.0, + "grad_norm": 4.452421433935271, + "learning_rate": 2.311868694440027e-06, + "loss": 0.7685, + "num_input_tokens_seen": 82360770, + "step": 3881 + }, + { + "epoch": 0.46678290146095114, + "flos": 50469368376000.0, + "grad_norm": 0.7411387211093564, + "learning_rate": 2.3110992279887323e-06, + "loss": 0.6337, + "num_input_tokens_seen": 82432415, + "step": 3882 + }, + { + "epoch": 0.4669031443515902, + "flos": 12921607101720.0, + "grad_norm": 2.765457062013587, + "learning_rate": 2.310329714343932e-06, + "loss": 0.8332, + "num_input_tokens_seen": 82448285, + "step": 3883 + }, + { + "epoch": 0.4670233872422293, + "flos": 17535563188800.0, + "grad_norm": 2.255842147695313, + "learning_rate": 2.309560153622361e-06, + "loss": 0.7984, + "num_input_tokens_seen": 82464915, + "step": 3884 + }, + { + "epoch": 0.4671436301328684, + "flos": 20617433219400.0, + "grad_norm": 2.56680020773472, + "learning_rate": 2.3087905459407602e-06, + "loss": 0.7234, + "num_input_tokens_seen": 82483575, + "step": 3885 + }, + { + "epoch": 0.46726387302350747, + "flos": 51155607873480.0, + "grad_norm": 0.8794000213468617, + "learning_rate": 2.3080208914158795e-06, + "loss": 0.6745, + "num_input_tokens_seen": 82546295, + "step": 3886 + }, + { + "epoch": 0.4673841159141466, + "flos": 18699316829760.0, + "grad_norm": 2.831210313396743, + "learning_rate": 2.3072511901644753e-06, + "loss": 0.6884, + "num_input_tokens_seen": 82565085, + "step": 3887 + }, + { + "epoch": 0.4675043588047857, + "flos": 17944320881880.0, + "grad_norm": 5.0188850432169065, + "learning_rate": 2.306481442303309e-06, + "loss": 0.7929, + "num_input_tokens_seen": 82584380, + "step": 3888 + }, + { + "epoch": 0.46762460169542475, + "flos": 15325686352800.0, + "grad_norm": 2.376519670671099, + "learning_rate": 2.3057116479491515e-06, + "loss": 0.7154, + "num_input_tokens_seen": 82603510, + "step": 3889 + }, + { + "epoch": 0.46774484458606386, + "flos": 14001443662800.0, + "grad_norm": 2.5425308717786996, + "learning_rate": 2.30494180721878e-06, + "loss": 0.7506, + "num_input_tokens_seen": 82620570, + "step": 3890 + }, + { + "epoch": 0.4678650874767029, + "flos": 13110427323600.0, + "grad_norm": 2.4056148929438135, + "learning_rate": 2.3041719202289794e-06, + "loss": 0.8862, + "num_input_tokens_seen": 82636465, + "step": 3891 + }, + { + "epoch": 0.467985330367342, + "flos": 15621562593960.0, + "grad_norm": 3.308007991241434, + "learning_rate": 2.30340198709654e-06, + "loss": 0.7912, + "num_input_tokens_seen": 82656020, + "step": 3892 + }, + { + "epoch": 0.46810557325798113, + "flos": 15000687439800.0, + "grad_norm": 2.6416244872503407, + "learning_rate": 2.3026320079382605e-06, + "loss": 0.7425, + "num_input_tokens_seen": 82672675, + "step": 3893 + }, + { + "epoch": 0.4682258161486202, + "flos": 22102418238120.0, + "grad_norm": 5.392324163252039, + "learning_rate": 2.3018619828709454e-06, + "loss": 0.7393, + "num_input_tokens_seen": 82693935, + "step": 3894 + }, + { + "epoch": 0.4683460590392593, + "flos": 18530786150880.0, + "grad_norm": 4.734562130493683, + "learning_rate": 2.3010919120114084e-06, + "loss": 0.8062, + "num_input_tokens_seen": 82710185, + "step": 3895 + }, + { + "epoch": 0.4684663019298984, + "flos": 11185472382600.0, + "grad_norm": 3.2128843916162935, + "learning_rate": 2.3003217954764672e-06, + "loss": 0.6288, + "num_input_tokens_seen": 82724610, + "step": 3896 + }, + { + "epoch": 0.46858654482053747, + "flos": 20370842553240.0, + "grad_norm": 3.367459385774942, + "learning_rate": 2.299551633382949e-06, + "loss": 0.786, + "num_input_tokens_seen": 82744640, + "step": 3897 + }, + { + "epoch": 0.4687067877111766, + "flos": 13164398572680.0, + "grad_norm": 2.605094642603955, + "learning_rate": 2.2987814258476854e-06, + "loss": 0.8257, + "num_input_tokens_seen": 82762160, + "step": 3898 + }, + { + "epoch": 0.4688270306018157, + "flos": 12376797357360.0, + "grad_norm": 3.496857867569668, + "learning_rate": 2.2980111729875177e-06, + "loss": 0.6492, + "num_input_tokens_seen": 82778130, + "step": 3899 + }, + { + "epoch": 0.46894727349245474, + "flos": 13002326525640.0, + "grad_norm": 2.5155985576116593, + "learning_rate": 2.2972408749192917e-06, + "loss": 0.8038, + "num_input_tokens_seen": 82795580, + "step": 3900 + }, + { + "epoch": 0.46906751638309385, + "flos": 15703516756320.0, + "grad_norm": 2.727128459264671, + "learning_rate": 2.296470531759861e-06, + "loss": 0.6587, + "num_input_tokens_seen": 82813400, + "step": 3901 + }, + { + "epoch": 0.46918775927373296, + "flos": 14864287128960.0, + "grad_norm": 3.116860585291246, + "learning_rate": 2.2957001436260866e-06, + "loss": 0.7751, + "num_input_tokens_seen": 82830180, + "step": 3902 + }, + { + "epoch": 0.469308002164372, + "flos": 13432450200360.0, + "grad_norm": 2.355679175244046, + "learning_rate": 2.294929710634836e-06, + "loss": 0.7081, + "num_input_tokens_seen": 82847990, + "step": 3903 + }, + { + "epoch": 0.46942824505501113, + "flos": 27907635980040.0, + "grad_norm": 2.450309157815003, + "learning_rate": 2.2941592329029823e-06, + "loss": 0.5944, + "num_input_tokens_seen": 82868815, + "step": 3904 + }, + { + "epoch": 0.46954848794565024, + "flos": 16000374456240.0, + "grad_norm": 5.138594113163529, + "learning_rate": 2.2933887105474067e-06, + "loss": 0.7797, + "num_input_tokens_seen": 82886710, + "step": 3905 + }, + { + "epoch": 0.4696687308362893, + "flos": 16106702296440.0, + "grad_norm": 2.525561876383723, + "learning_rate": 2.2926181436849974e-06, + "loss": 0.7961, + "num_input_tokens_seen": 82905785, + "step": 3906 + }, + { + "epoch": 0.4697889737269284, + "flos": 15808324918440.0, + "grad_norm": 2.8498911518927734, + "learning_rate": 2.2918475324326478e-06, + "loss": 0.7134, + "num_input_tokens_seen": 82925225, + "step": 3907 + }, + { + "epoch": 0.46990921661756746, + "flos": 18483780093000.0, + "grad_norm": 2.89201527746869, + "learning_rate": 2.2910768769072603e-06, + "loss": 0.8899, + "num_input_tokens_seen": 82943725, + "step": 3908 + }, + { + "epoch": 0.47002945950820657, + "flos": 10058534783760.0, + "grad_norm": 2.9445880838862077, + "learning_rate": 2.2903061772257417e-06, + "loss": 0.7385, + "num_input_tokens_seen": 82961430, + "step": 3909 + }, + { + "epoch": 0.4701497023988457, + "flos": 19237256362800.0, + "grad_norm": 2.4565810623506112, + "learning_rate": 2.289535433505007e-06, + "loss": 0.7713, + "num_input_tokens_seen": 82982505, + "step": 3910 + }, + { + "epoch": 0.47026994528948474, + "flos": 18780669452880.0, + "grad_norm": 4.342943514596025, + "learning_rate": 2.2887646458619767e-06, + "loss": 0.6238, + "num_input_tokens_seen": 83003590, + "step": 3911 + }, + { + "epoch": 0.47039018818012385, + "flos": 15024237958680.0, + "grad_norm": 2.1616327299671276, + "learning_rate": 2.2879938144135797e-06, + "loss": 0.7498, + "num_input_tokens_seen": 83019415, + "step": 3912 + }, + { + "epoch": 0.47051043107076296, + "flos": 15781386783840.0, + "grad_norm": 2.160083624669185, + "learning_rate": 2.2872229392767496e-06, + "loss": 0.7488, + "num_input_tokens_seen": 83039240, + "step": 3913 + }, + { + "epoch": 0.470630673961402, + "flos": 13839814855200.0, + "grad_norm": 2.546873783041072, + "learning_rate": 2.286452020568428e-06, + "loss": 0.7371, + "num_input_tokens_seen": 83057035, + "step": 3914 + }, + { + "epoch": 0.4707509168520411, + "flos": 14569455666480.0, + "grad_norm": 3.82542567159639, + "learning_rate": 2.2856810584055637e-06, + "loss": 0.7286, + "num_input_tokens_seen": 83074290, + "step": 3915 + }, + { + "epoch": 0.47087115974268023, + "flos": 29504299372200.0, + "grad_norm": 1.8835019317487427, + "learning_rate": 2.2849100529051085e-06, + "loss": 0.6643, + "num_input_tokens_seen": 83100945, + "step": 3916 + }, + { + "epoch": 0.4709914026333193, + "flos": 9843029706960.0, + "grad_norm": 4.4643964568435175, + "learning_rate": 2.284139004184026e-06, + "loss": 0.7754, + "num_input_tokens_seen": 83117895, + "step": 3917 + }, + { + "epoch": 0.4711116455239584, + "flos": 14595728941920.0, + "grad_norm": 3.7017958947894045, + "learning_rate": 2.2833679123592814e-06, + "loss": 0.7239, + "num_input_tokens_seen": 83134875, + "step": 3918 + }, + { + "epoch": 0.4712318884145975, + "flos": 23587751516400.0, + "grad_norm": 3.5479463144087675, + "learning_rate": 2.2825967775478508e-06, + "loss": 0.6132, + "num_input_tokens_seen": 83155695, + "step": 3919 + }, + { + "epoch": 0.47135213130523657, + "flos": 14649953470680.0, + "grad_norm": 2.575176795890009, + "learning_rate": 2.2818255998667135e-06, + "loss": 0.8195, + "num_input_tokens_seen": 83173925, + "step": 3920 + }, + { + "epoch": 0.4714723741958757, + "flos": 14190390524520.0, + "grad_norm": 2.2456234646926503, + "learning_rate": 2.2810543794328566e-06, + "loss": 0.7758, + "num_input_tokens_seen": 83192680, + "step": 3921 + }, + { + "epoch": 0.4715926170865148, + "flos": 14890908663960.0, + "grad_norm": 2.86782638541455, + "learning_rate": 2.2802831163632735e-06, + "loss": 0.8026, + "num_input_tokens_seen": 83211120, + "step": 3922 + }, + { + "epoch": 0.47171285997715384, + "flos": 16592380218240.0, + "grad_norm": 2.071550797863736, + "learning_rate": 2.279511810774965e-06, + "loss": 0.7248, + "num_input_tokens_seen": 83232370, + "step": 3923 + }, + { + "epoch": 0.47183310286779295, + "flos": 15432552412320.0, + "grad_norm": 2.802400469412017, + "learning_rate": 2.2787404627849364e-06, + "loss": 0.698, + "num_input_tokens_seen": 83251300, + "step": 3924 + }, + { + "epoch": 0.471953345758432, + "flos": 15892368638160.0, + "grad_norm": 2.61343451764672, + "learning_rate": 2.277969072510202e-06, + "loss": 0.7706, + "num_input_tokens_seen": 83270000, + "step": 3925 + }, + { + "epoch": 0.4720735886490711, + "flos": 14511115342920.0, + "grad_norm": 2.0825826691227576, + "learning_rate": 2.2771976400677803e-06, + "loss": 0.8024, + "num_input_tokens_seen": 83288550, + "step": 3926 + }, + { + "epoch": 0.47219383153971023, + "flos": 14001570302640.0, + "grad_norm": 2.222918396090759, + "learning_rate": 2.2764261655746965e-06, + "loss": 0.7717, + "num_input_tokens_seen": 83305765, + "step": 3927 + }, + { + "epoch": 0.4723140744303493, + "flos": 17000789651760.0, + "grad_norm": 1.972375166101181, + "learning_rate": 2.2756546491479832e-06, + "loss": 0.7444, + "num_input_tokens_seen": 83326400, + "step": 3928 + }, + { + "epoch": 0.4724343173209884, + "flos": 13299595805040.0, + "grad_norm": 3.674373734239236, + "learning_rate": 2.274883090904679e-06, + "loss": 0.786, + "num_input_tokens_seen": 83343885, + "step": 3929 + }, + { + "epoch": 0.4725545602116275, + "flos": 15540273290760.0, + "grad_norm": 3.0095911885186792, + "learning_rate": 2.2741114909618283e-06, + "loss": 0.6573, + "num_input_tokens_seen": 83359500, + "step": 3930 + }, + { + "epoch": 0.47267480310226656, + "flos": 15675755462760.0, + "grad_norm": 2.4247718467746884, + "learning_rate": 2.2733398494364828e-06, + "loss": 0.704, + "num_input_tokens_seen": 83378465, + "step": 3931 + }, + { + "epoch": 0.47279504599290567, + "flos": 13704585962880.0, + "grad_norm": 2.4119055534457967, + "learning_rate": 2.272568166445699e-06, + "loss": 0.8327, + "num_input_tokens_seen": 83396750, + "step": 3932 + }, + { + "epoch": 0.4729152888835448, + "flos": 15432837351960.0, + "grad_norm": 2.580383236536756, + "learning_rate": 2.271796442106541e-06, + "loss": 0.6271, + "num_input_tokens_seen": 83415825, + "step": 3933 + }, + { + "epoch": 0.47303553177418384, + "flos": 51770499295200.0, + "grad_norm": 0.7975155369073267, + "learning_rate": 2.271024676536079e-06, + "loss": 0.6036, + "num_input_tokens_seen": 83475805, + "step": 3934 + }, + { + "epoch": 0.47315577466482295, + "flos": 16431384609840.0, + "grad_norm": 3.3308710222654634, + "learning_rate": 2.2702528698513894e-06, + "loss": 0.7175, + "num_input_tokens_seen": 83496650, + "step": 3935 + }, + { + "epoch": 0.47327601755546206, + "flos": 17835903484320.0, + "grad_norm": 4.246750430763669, + "learning_rate": 2.269481022169554e-06, + "loss": 0.7675, + "num_input_tokens_seen": 83514965, + "step": 3936 + }, + { + "epoch": 0.4733962604461011, + "flos": 16780060681560.0, + "grad_norm": 3.2353117687335424, + "learning_rate": 2.2687091336076614e-06, + "loss": 0.7895, + "num_input_tokens_seen": 83534025, + "step": 3937 + }, + { + "epoch": 0.4735165033367402, + "flos": 13375882834560.0, + "grad_norm": 3.8857803515419764, + "learning_rate": 2.267937204282807e-06, + "loss": 0.7812, + "num_input_tokens_seen": 83550885, + "step": 3938 + }, + { + "epoch": 0.4736367462273793, + "flos": 16861761564240.0, + "grad_norm": 3.0447997239369813, + "learning_rate": 2.2671652343120926e-06, + "loss": 0.7575, + "num_input_tokens_seen": 83571080, + "step": 3939 + }, + { + "epoch": 0.4737569891180184, + "flos": 18591849231000.0, + "grad_norm": 2.3175908319641536, + "learning_rate": 2.2663932238126236e-06, + "loss": 0.7871, + "num_input_tokens_seen": 83589360, + "step": 3940 + }, + { + "epoch": 0.4738772320086575, + "flos": 18943469679000.0, + "grad_norm": 2.3191620000567714, + "learning_rate": 2.265621172901515e-06, + "loss": 0.7811, + "num_input_tokens_seen": 83612195, + "step": 3941 + }, + { + "epoch": 0.47399747489929656, + "flos": 20212791321120.0, + "grad_norm": 3.527565983966882, + "learning_rate": 2.2648490816958854e-06, + "loss": 0.7017, + "num_input_tokens_seen": 83632910, + "step": 3942 + }, + { + "epoch": 0.47411771778993567, + "flos": 18213923847600.0, + "grad_norm": 3.531253510870074, + "learning_rate": 2.264076950312861e-06, + "loss": 0.7043, + "num_input_tokens_seen": 83651440, + "step": 3943 + }, + { + "epoch": 0.4742379606805748, + "flos": 16648282724880.0, + "grad_norm": 2.8857187234663666, + "learning_rate": 2.2633047788695727e-06, + "loss": 0.7997, + "num_input_tokens_seen": 83671465, + "step": 3944 + }, + { + "epoch": 0.47435820357121383, + "flos": 14378292607560.0, + "grad_norm": 3.7150083038114055, + "learning_rate": 2.262532567483159e-06, + "loss": 0.6294, + "num_input_tokens_seen": 83689745, + "step": 3945 + }, + { + "epoch": 0.47447844646185294, + "flos": 18484128352560.0, + "grad_norm": 7.944531067536752, + "learning_rate": 2.2617603162707635e-06, + "loss": 0.7816, + "num_input_tokens_seen": 83709875, + "step": 3946 + }, + { + "epoch": 0.47459868935249205, + "flos": 17996899092720.0, + "grad_norm": 2.3099572920050817, + "learning_rate": 2.2609880253495363e-06, + "loss": 0.8079, + "num_input_tokens_seen": 83729230, + "step": 3947 + }, + { + "epoch": 0.4747189322431311, + "flos": 15027847194120.0, + "grad_norm": 2.7698338266649944, + "learning_rate": 2.260215694836633e-06, + "loss": 0.8434, + "num_input_tokens_seen": 83748125, + "step": 3948 + }, + { + "epoch": 0.4748391751337702, + "flos": 19051443837120.0, + "grad_norm": 2.195391453310836, + "learning_rate": 2.2594433248492157e-06, + "loss": 0.6294, + "num_input_tokens_seen": 83766820, + "step": 3949 + }, + { + "epoch": 0.47495941802440933, + "flos": 16702348953840.0, + "grad_norm": 2.2612201885664343, + "learning_rate": 2.2586709155044527e-06, + "loss": 0.7838, + "num_input_tokens_seen": 83787140, + "step": 3950 + }, + { + "epoch": 0.4750796609150484, + "flos": 20454917932920.0, + "grad_norm": 1.9151548421623792, + "learning_rate": 2.2578984669195167e-06, + "loss": 0.7487, + "num_input_tokens_seen": 83807825, + "step": 3951 + }, + { + "epoch": 0.4751999038056875, + "flos": 26203536649080.0, + "grad_norm": 2.3218815183483854, + "learning_rate": 2.2571259792115887e-06, + "loss": 0.6585, + "num_input_tokens_seen": 83828765, + "step": 3952 + }, + { + "epoch": 0.4753201466963266, + "flos": 16161908283960.0, + "grad_norm": 3.325607708916666, + "learning_rate": 2.2563534524978544e-06, + "loss": 0.7787, + "num_input_tokens_seen": 83845955, + "step": 3953 + }, + { + "epoch": 0.47544038958696566, + "flos": 22156452807120.0, + "grad_norm": 3.230612515222267, + "learning_rate": 2.2555808868955052e-06, + "loss": 0.7115, + "num_input_tokens_seen": 83867805, + "step": 3954 + }, + { + "epoch": 0.47556063247760477, + "flos": 17183467841400.0, + "grad_norm": 4.180078563043151, + "learning_rate": 2.254808282521738e-06, + "loss": 0.7114, + "num_input_tokens_seen": 83886275, + "step": 3955 + }, + { + "epoch": 0.4756808753682438, + "flos": 18429650544120.0, + "grad_norm": 4.837730950256567, + "learning_rate": 2.2540356394937573e-06, + "loss": 0.7973, + "num_input_tokens_seen": 83904695, + "step": 3956 + }, + { + "epoch": 0.47580111825888294, + "flos": 11409873908160.0, + "grad_norm": 3.2658625612137744, + "learning_rate": 2.253262957928772e-06, + "loss": 0.8136, + "num_input_tokens_seen": 83921300, + "step": 3957 + }, + { + "epoch": 0.47592136114952205, + "flos": 12864976416000.0, + "grad_norm": 1.996406426174099, + "learning_rate": 2.2524902379439976e-06, + "loss": 0.6973, + "num_input_tokens_seen": 83939690, + "step": 3958 + }, + { + "epoch": 0.4760416040401611, + "flos": 45270398886720.0, + "grad_norm": 0.7584162794240081, + "learning_rate": 2.251717479656655e-06, + "loss": 0.6495, + "num_input_tokens_seen": 84004205, + "step": 3959 + }, + { + "epoch": 0.4761618469308002, + "flos": 13434761377440.0, + "grad_norm": 2.44497463829445, + "learning_rate": 2.2509446831839704e-06, + "loss": 0.7366, + "num_input_tokens_seen": 84023365, + "step": 3960 + }, + { + "epoch": 0.4762820898214393, + "flos": 13164651852360.0, + "grad_norm": 2.8420472129857104, + "learning_rate": 2.250171848643177e-06, + "loss": 0.7961, + "num_input_tokens_seen": 84040375, + "step": 3961 + }, + { + "epoch": 0.4764023327120784, + "flos": 14110019360160.0, + "grad_norm": 4.9093132418239165, + "learning_rate": 2.249398976151513e-06, + "loss": 0.8533, + "num_input_tokens_seen": 84057645, + "step": 3962 + }, + { + "epoch": 0.4765225756027175, + "flos": 16647617865720.0, + "grad_norm": 2.9769195582779053, + "learning_rate": 2.248626065826223e-06, + "loss": 0.7727, + "num_input_tokens_seen": 84075570, + "step": 3963 + }, + { + "epoch": 0.4766428184933566, + "flos": 48611582396040.0, + "grad_norm": 0.7564081767466843, + "learning_rate": 2.2478531177845564e-06, + "loss": 0.634, + "num_input_tokens_seen": 84136285, + "step": 3964 + }, + { + "epoch": 0.47676306138399566, + "flos": 17676142614360.0, + "grad_norm": 3.1873433691698048, + "learning_rate": 2.247080132143769e-06, + "loss": 0.8356, + "num_input_tokens_seen": 84158360, + "step": 3965 + }, + { + "epoch": 0.47688330427463477, + "flos": 9141181849200.0, + "grad_norm": 2.80642196026456, + "learning_rate": 2.246307109021121e-06, + "loss": 0.6682, + "num_input_tokens_seen": 84175485, + "step": 3966 + }, + { + "epoch": 0.4770035471652739, + "flos": 15642643635960.0, + "grad_norm": 1.8835255392564365, + "learning_rate": 2.2455340485338817e-06, + "loss": 0.8061, + "num_input_tokens_seen": 84192840, + "step": 3967 + }, + { + "epoch": 0.47712379005591293, + "flos": 18430568682960.0, + "grad_norm": 6.640540842757175, + "learning_rate": 2.244760950799322e-06, + "loss": 0.6651, + "num_input_tokens_seen": 84210830, + "step": 3968 + }, + { + "epoch": 0.47724403294655204, + "flos": 16134843509520.0, + "grad_norm": 3.444931043756933, + "learning_rate": 2.2439878159347203e-06, + "loss": 0.713, + "num_input_tokens_seen": 84229975, + "step": 3969 + }, + { + "epoch": 0.4773642758371911, + "flos": 51794750824560.0, + "grad_norm": 0.817468417662935, + "learning_rate": 2.2432146440573616e-06, + "loss": 0.6165, + "num_input_tokens_seen": 84295655, + "step": 3970 + }, + { + "epoch": 0.4774845187278302, + "flos": 17240478446640.0, + "grad_norm": 2.500338531461614, + "learning_rate": 2.242441435284534e-06, + "loss": 0.643, + "num_input_tokens_seen": 84314250, + "step": 3971 + }, + { + "epoch": 0.4776047616184693, + "flos": 16890409336680.0, + "grad_norm": 2.2402780091244265, + "learning_rate": 2.2416681897335337e-06, + "loss": 0.8314, + "num_input_tokens_seen": 84332120, + "step": 3972 + }, + { + "epoch": 0.4777250045091084, + "flos": 23418872577960.0, + "grad_norm": 2.4053360825936676, + "learning_rate": 2.240894907521661e-06, + "loss": 0.6507, + "num_input_tokens_seen": 84350920, + "step": 3973 + }, + { + "epoch": 0.4778452473997475, + "flos": 17780380897200.0, + "grad_norm": 2.677515957283461, + "learning_rate": 2.240121588766223e-06, + "loss": 0.6232, + "num_input_tokens_seen": 84370690, + "step": 3974 + }, + { + "epoch": 0.4779654902903866, + "flos": 23180513501400.0, + "grad_norm": 3.7333624197190143, + "learning_rate": 2.239348233584531e-06, + "loss": 0.6896, + "num_input_tokens_seen": 84391265, + "step": 3975 + }, + { + "epoch": 0.47808573318102565, + "flos": 14244805013040.0, + "grad_norm": 2.8466749973160335, + "learning_rate": 2.2385748420939013e-06, + "loss": 0.7889, + "num_input_tokens_seen": 84410180, + "step": 3976 + }, + { + "epoch": 0.47820597607166476, + "flos": 16539453747840.0, + "grad_norm": 4.669647604411884, + "learning_rate": 2.2378014144116583e-06, + "loss": 0.7126, + "num_input_tokens_seen": 84428710, + "step": 3977 + }, + { + "epoch": 0.4783262189623039, + "flos": 16836944646960.0, + "grad_norm": 2.6220174830280847, + "learning_rate": 2.23702795065513e-06, + "loss": 0.7816, + "num_input_tokens_seen": 84448010, + "step": 3978 + }, + { + "epoch": 0.47844646185294293, + "flos": 36652027601040.0, + "grad_norm": 1.0002425614861985, + "learning_rate": 2.2362544509416493e-06, + "loss": 0.6979, + "num_input_tokens_seen": 84499845, + "step": 3979 + }, + { + "epoch": 0.47856670474358204, + "flos": 14784327544080.0, + "grad_norm": 3.785272397679431, + "learning_rate": 2.2354809153885572e-06, + "loss": 0.8055, + "num_input_tokens_seen": 84516635, + "step": 3980 + }, + { + "epoch": 0.47868694763422115, + "flos": 15347115654360.0, + "grad_norm": 2.9483657512819743, + "learning_rate": 2.234707344113197e-06, + "loss": 0.8281, + "num_input_tokens_seen": 84534450, + "step": 3981 + }, + { + "epoch": 0.4788071905248602, + "flos": 13894166023800.0, + "grad_norm": 2.855412306534315, + "learning_rate": 2.233933737232919e-06, + "loss": 0.7619, + "num_input_tokens_seen": 84551950, + "step": 3982 + }, + { + "epoch": 0.4789274334154993, + "flos": 16836248127840.0, + "grad_norm": 2.4038521845815275, + "learning_rate": 2.2331600948650793e-06, + "loss": 0.7589, + "num_input_tokens_seen": 84571815, + "step": 3983 + }, + { + "epoch": 0.4790476763061384, + "flos": 16890599296440.0, + "grad_norm": 2.5329302638620224, + "learning_rate": 2.2323864171270386e-06, + "loss": 0.7866, + "num_input_tokens_seen": 84592805, + "step": 3984 + }, + { + "epoch": 0.4791679191967775, + "flos": 15487346820360.0, + "grad_norm": 4.1678083508703905, + "learning_rate": 2.231612704136164e-06, + "loss": 0.7138, + "num_input_tokens_seen": 84612895, + "step": 3985 + }, + { + "epoch": 0.4792881620874166, + "flos": 16316603560320.0, + "grad_norm": 2.3306764280578904, + "learning_rate": 2.2308389560098253e-06, + "loss": 0.7133, + "num_input_tokens_seen": 84628990, + "step": 3986 + }, + { + "epoch": 0.47940840497805565, + "flos": 12705817085280.0, + "grad_norm": 3.560827699301634, + "learning_rate": 2.2300651728654008e-06, + "loss": 0.7492, + "num_input_tokens_seen": 84643970, + "step": 3987 + }, + { + "epoch": 0.47952864786869476, + "flos": 48187094194200.0, + "grad_norm": 0.742124052849323, + "learning_rate": 2.229291354820272e-06, + "loss": 0.6265, + "num_input_tokens_seen": 84704700, + "step": 3988 + }, + { + "epoch": 0.47964889075933387, + "flos": 12246064179360.0, + "grad_norm": 8.026408951805353, + "learning_rate": 2.228517501991828e-06, + "loss": 0.7472, + "num_input_tokens_seen": 84723220, + "step": 3989 + }, + { + "epoch": 0.4797691336499729, + "flos": 51683040791160.0, + "grad_norm": 0.8375118083975427, + "learning_rate": 2.22774361449746e-06, + "loss": 0.6328, + "num_input_tokens_seen": 84779420, + "step": 3990 + }, + { + "epoch": 0.47988937654061203, + "flos": 13839941495040.0, + "grad_norm": 5.6872722873585575, + "learning_rate": 2.2269696924545668e-06, + "loss": 0.6849, + "num_input_tokens_seen": 84796970, + "step": 3991 + }, + { + "epoch": 0.48000961943125114, + "flos": 10515533273160.0, + "grad_norm": 3.272197479416442, + "learning_rate": 2.2261957359805523e-06, + "loss": 0.7557, + "num_input_tokens_seen": 84813925, + "step": 3992 + }, + { + "epoch": 0.4801298623218902, + "flos": 19832586420600.0, + "grad_norm": 2.4268006472577763, + "learning_rate": 2.225421745192823e-06, + "loss": 0.7391, + "num_input_tokens_seen": 84833800, + "step": 3993 + }, + { + "epoch": 0.4802501052125293, + "flos": 19317184287720.0, + "grad_norm": 2.800803212355052, + "learning_rate": 2.2246477202087955e-06, + "loss": 0.7529, + "num_input_tokens_seen": 84854200, + "step": 3994 + }, + { + "epoch": 0.4803703481031684, + "flos": 15349743431040.0, + "grad_norm": 1.9335239590466728, + "learning_rate": 2.223873661145887e-06, + "loss": 0.8173, + "num_input_tokens_seen": 84873975, + "step": 3995 + }, + { + "epoch": 0.4804905909938075, + "flos": 15135219813000.0, + "grad_norm": 2.188723848492755, + "learning_rate": 2.2230995681215226e-06, + "loss": 0.6988, + "num_input_tokens_seen": 84893220, + "step": 3996 + }, + { + "epoch": 0.4806108338844466, + "flos": 12138913180200.0, + "grad_norm": 3.4571002155903385, + "learning_rate": 2.2223254412531305e-06, + "loss": 0.768, + "num_input_tokens_seen": 84910310, + "step": 3997 + }, + { + "epoch": 0.4807310767750857, + "flos": 14622920356200.0, + "grad_norm": 2.2176212072093264, + "learning_rate": 2.221551280658146e-06, + "loss": 0.8006, + "num_input_tokens_seen": 84929090, + "step": 3998 + }, + { + "epoch": 0.48085131966572475, + "flos": 16971793619760.0, + "grad_norm": 1.9851165947674627, + "learning_rate": 2.2207770864540085e-06, + "loss": 0.7299, + "num_input_tokens_seen": 84947785, + "step": 3999 + }, + { + "epoch": 0.48097156255636386, + "flos": 15027973833960.0, + "grad_norm": 3.010886783697315, + "learning_rate": 2.220002858758162e-06, + "loss": 0.7063, + "num_input_tokens_seen": 84965495, + "step": 4000 + }, + { + "epoch": 0.481091805447003, + "flos": 52002245931480.0, + "grad_norm": 0.9136300979724634, + "learning_rate": 2.2192285976880573e-06, + "loss": 0.6205, + "num_input_tokens_seen": 85029470, + "step": 4001 + }, + { + "epoch": 0.48121204833764203, + "flos": 26772276831840.0, + "grad_norm": 1.8975653509212391, + "learning_rate": 2.2184543033611485e-06, + "loss": 0.7944, + "num_input_tokens_seen": 85050270, + "step": 4002 + }, + { + "epoch": 0.48133229122828114, + "flos": 20157743633400.0, + "grad_norm": 2.6374383241212866, + "learning_rate": 2.2176799758948957e-06, + "loss": 0.7971, + "num_input_tokens_seen": 85070150, + "step": 4003 + }, + { + "epoch": 0.4814525341189202, + "flos": 31690309089720.0, + "grad_norm": 2.448513288365661, + "learning_rate": 2.2169056154067635e-06, + "loss": 0.7182, + "num_input_tokens_seen": 85093790, + "step": 4004 + }, + { + "epoch": 0.4815727770095593, + "flos": 17749168668000.0, + "grad_norm": 2.268616063391647, + "learning_rate": 2.216131222014222e-06, + "loss": 0.8141, + "num_input_tokens_seen": 85111585, + "step": 4005 + }, + { + "epoch": 0.4816930199001984, + "flos": 13218433141680.0, + "grad_norm": 4.660221531067802, + "learning_rate": 2.2153567958347455e-06, + "loss": 0.8022, + "num_input_tokens_seen": 85127515, + "step": 4006 + }, + { + "epoch": 0.48181326279083747, + "flos": 12597621307440.0, + "grad_norm": 2.4389304829755507, + "learning_rate": 2.214582336985815e-06, + "loss": 0.7828, + "num_input_tokens_seen": 85145135, + "step": 4007 + }, + { + "epoch": 0.4819335056814766, + "flos": 10842590083560.0, + "grad_norm": 3.3243824148624332, + "learning_rate": 2.2138078455849142e-06, + "loss": 0.6319, + "num_input_tokens_seen": 85162850, + "step": 4008 + }, + { + "epoch": 0.4820537485721157, + "flos": 14055668191560.0, + "grad_norm": 2.7197886269848386, + "learning_rate": 2.2130333217495334e-06, + "loss": 0.7654, + "num_input_tokens_seen": 85181740, + "step": 4009 + }, + { + "epoch": 0.48217399146275475, + "flos": 11679191934240.0, + "grad_norm": 3.464772598929471, + "learning_rate": 2.2122587655971665e-06, + "loss": 0.6517, + "num_input_tokens_seen": 85196405, + "step": 4010 + }, + { + "epoch": 0.48229423435339386, + "flos": 17674116376920.0, + "grad_norm": 2.252826716430697, + "learning_rate": 2.211484177245314e-06, + "loss": 0.6259, + "num_input_tokens_seen": 85215715, + "step": 4011 + }, + { + "epoch": 0.48241447724403297, + "flos": 17430755026680.0, + "grad_norm": 2.470447584363984, + "learning_rate": 2.21070955681148e-06, + "loss": 0.7089, + "num_input_tokens_seen": 85234540, + "step": 4012 + }, + { + "epoch": 0.482534720134672, + "flos": 16916049412920.0, + "grad_norm": 2.3040777919338917, + "learning_rate": 2.209934904413174e-06, + "loss": 0.7613, + "num_input_tokens_seen": 85255865, + "step": 4013 + }, + { + "epoch": 0.48265496302531113, + "flos": 15298273318800.0, + "grad_norm": 5.361591381287143, + "learning_rate": 2.2091602201679095e-06, + "loss": 0.6971, + "num_input_tokens_seen": 85275195, + "step": 4014 + }, + { + "epoch": 0.48277520591595025, + "flos": 11085191594760.0, + "grad_norm": 3.2056200683936615, + "learning_rate": 2.208385504193206e-06, + "loss": 0.7996, + "num_input_tokens_seen": 85292415, + "step": 4015 + }, + { + "epoch": 0.4828954488065893, + "flos": 13029802879560.0, + "grad_norm": 2.765672241497118, + "learning_rate": 2.2076107566065873e-06, + "loss": 0.8019, + "num_input_tokens_seen": 85309920, + "step": 4016 + }, + { + "epoch": 0.4830156916972284, + "flos": 23559926902920.0, + "grad_norm": 3.3847471700034246, + "learning_rate": 2.2068359775255816e-06, + "loss": 0.7342, + "num_input_tokens_seen": 85327950, + "step": 4017 + }, + { + "epoch": 0.48313593458786747, + "flos": 15999677937120.0, + "grad_norm": 3.164799357196831, + "learning_rate": 2.206061167067723e-06, + "loss": 0.7544, + "num_input_tokens_seen": 85345780, + "step": 4018 + }, + { + "epoch": 0.4832561774785066, + "flos": 16539517067760.0, + "grad_norm": 2.313884938992165, + "learning_rate": 2.205286325350549e-06, + "loss": 0.7715, + "num_input_tokens_seen": 85364565, + "step": 4019 + }, + { + "epoch": 0.4833764203691457, + "flos": 9756801450000.0, + "grad_norm": 2.4805546783221684, + "learning_rate": 2.204511452491603e-06, + "loss": 0.7243, + "num_input_tokens_seen": 85380910, + "step": 4020 + }, + { + "epoch": 0.48349666325978474, + "flos": 32470850133960.0, + "grad_norm": 2.292679360631417, + "learning_rate": 2.2037365486084316e-06, + "loss": 0.7323, + "num_input_tokens_seen": 85403870, + "step": 4021 + }, + { + "epoch": 0.48361690615042385, + "flos": 19075500915360.0, + "grad_norm": 3.1306791961663976, + "learning_rate": 2.2029616138185886e-06, + "loss": 0.7601, + "num_input_tokens_seen": 85422590, + "step": 4022 + }, + { + "epoch": 0.48373714904106296, + "flos": 16296915556560.0, + "grad_norm": 2.372242983693205, + "learning_rate": 2.202186648239629e-06, + "loss": 0.8151, + "num_input_tokens_seen": 85442245, + "step": 4023 + }, + { + "epoch": 0.483857391931702, + "flos": 20751205753560.0, + "grad_norm": 2.2368929886714146, + "learning_rate": 2.201411651989117e-06, + "loss": 0.7036, + "num_input_tokens_seen": 85463945, + "step": 4024 + }, + { + "epoch": 0.48397763482234113, + "flos": 20104405583520.0, + "grad_norm": 3.235029661110414, + "learning_rate": 2.2006366251846167e-06, + "loss": 0.7603, + "num_input_tokens_seen": 85484305, + "step": 4025 + }, + { + "epoch": 0.48409787771298024, + "flos": 12243816322200.0, + "grad_norm": 2.304159537149704, + "learning_rate": 2.1998615679436997e-06, + "loss": 0.7372, + "num_input_tokens_seen": 85501565, + "step": 4026 + }, + { + "epoch": 0.4842181206036193, + "flos": 18376755733680.0, + "grad_norm": 3.347928867345551, + "learning_rate": 2.199086480383942e-06, + "loss": 0.7471, + "num_input_tokens_seen": 85520660, + "step": 4027 + }, + { + "epoch": 0.4843383634942584, + "flos": 22290605260800.0, + "grad_norm": 4.91586391986325, + "learning_rate": 2.1983113626229234e-06, + "loss": 0.6401, + "num_input_tokens_seen": 85539630, + "step": 4028 + }, + { + "epoch": 0.4844586063848975, + "flos": 14920727854920.0, + "grad_norm": 2.5931233947557697, + "learning_rate": 2.1975362147782293e-06, + "loss": 0.7709, + "num_input_tokens_seen": 85558545, + "step": 4029 + }, + { + "epoch": 0.48457884927553657, + "flos": 51849956812080.0, + "grad_norm": 0.8499618675990871, + "learning_rate": 2.196761036967448e-06, + "loss": 0.5623, + "num_input_tokens_seen": 85626230, + "step": 4030 + }, + { + "epoch": 0.4846990921661757, + "flos": 14565529831440.0, + "grad_norm": 2.501813612181702, + "learning_rate": 2.1959858293081743e-06, + "loss": 0.7668, + "num_input_tokens_seen": 85645085, + "step": 4031 + }, + { + "epoch": 0.4848193350568148, + "flos": 16890915896040.0, + "grad_norm": 5.442905589153775, + "learning_rate": 2.1952105919180056e-06, + "loss": 0.7523, + "num_input_tokens_seen": 85664060, + "step": 4032 + }, + { + "epoch": 0.48493957794745385, + "flos": 16431701209440.0, + "grad_norm": 3.8808328352137313, + "learning_rate": 2.1944353249145456e-06, + "loss": 0.6642, + "num_input_tokens_seen": 85682890, + "step": 4033 + }, + { + "epoch": 0.48505982083809296, + "flos": 18348962780160.0, + "grad_norm": 4.416492218794997, + "learning_rate": 2.193660028415401e-06, + "loss": 0.7314, + "num_input_tokens_seen": 85703390, + "step": 4034 + }, + { + "epoch": 0.485180063728732, + "flos": 19618442722080.0, + "grad_norm": 2.447096396377315, + "learning_rate": 2.1928847025381852e-06, + "loss": 0.8134, + "num_input_tokens_seen": 85723715, + "step": 4035 + }, + { + "epoch": 0.4853003066193711, + "flos": 17618277190200.0, + "grad_norm": 2.4555243160261027, + "learning_rate": 2.192109347400512e-06, + "loss": 0.826, + "num_input_tokens_seen": 85743650, + "step": 4036 + }, + { + "epoch": 0.48542054951001024, + "flos": 16891295815560.0, + "grad_norm": 2.6694821356279057, + "learning_rate": 2.191333963120004e-06, + "loss": 0.7772, + "num_input_tokens_seen": 85762350, + "step": 4037 + }, + { + "epoch": 0.4855407924006493, + "flos": 18806689448640.0, + "grad_norm": 3.376263134848542, + "learning_rate": 2.190558549814286e-06, + "loss": 0.6778, + "num_input_tokens_seen": 85782230, + "step": 4038 + }, + { + "epoch": 0.4856610352912884, + "flos": 17565825619200.0, + "grad_norm": 3.92935743519416, + "learning_rate": 2.1897831076009872e-06, + "loss": 0.7747, + "num_input_tokens_seen": 85801590, + "step": 4039 + }, + { + "epoch": 0.4857812781819275, + "flos": 17646323423400.0, + "grad_norm": 5.679349319905386, + "learning_rate": 2.1890076365977426e-06, + "loss": 0.7856, + "num_input_tokens_seen": 85821135, + "step": 4040 + }, + { + "epoch": 0.48590152107256657, + "flos": 41457811606200.0, + "grad_norm": 0.8842143516073863, + "learning_rate": 2.188232136922189e-06, + "loss": 0.5583, + "num_input_tokens_seen": 85878975, + "step": 4041 + }, + { + "epoch": 0.4860217639632057, + "flos": 14649193631640.0, + "grad_norm": 2.4195852961553337, + "learning_rate": 2.187456608691971e-06, + "loss": 0.7371, + "num_input_tokens_seen": 85897570, + "step": 4042 + }, + { + "epoch": 0.4861420068538448, + "flos": 13002769765080.0, + "grad_norm": 2.4237835654835913, + "learning_rate": 2.1866810520247334e-06, + "loss": 0.8604, + "num_input_tokens_seen": 85916160, + "step": 4043 + }, + { + "epoch": 0.48626224974448384, + "flos": 19240010779320.0, + "grad_norm": 5.339593174042591, + "learning_rate": 2.185905467038129e-06, + "loss": 0.6266, + "num_input_tokens_seen": 85934785, + "step": 4044 + }, + { + "epoch": 0.48638249263512295, + "flos": 16135001809320.0, + "grad_norm": 1.962724821005722, + "learning_rate": 2.1851298538498127e-06, + "loss": 0.7558, + "num_input_tokens_seen": 85954220, + "step": 4045 + }, + { + "epoch": 0.48650273552576206, + "flos": 18403725528240.0, + "grad_norm": 4.1229689250536445, + "learning_rate": 2.184354212577446e-06, + "loss": 0.7858, + "num_input_tokens_seen": 85974245, + "step": 4046 + }, + { + "epoch": 0.4866229784164011, + "flos": 12732438620280.0, + "grad_norm": 8.008165483836102, + "learning_rate": 2.1835785433386907e-06, + "loss": 0.5975, + "num_input_tokens_seen": 85992780, + "step": 4047 + }, + { + "epoch": 0.48674322130704023, + "flos": 17079736117920.0, + "grad_norm": 4.9762710174523, + "learning_rate": 2.182802846251216e-06, + "loss": 0.6301, + "num_input_tokens_seen": 86012770, + "step": 4048 + }, + { + "epoch": 0.4868634641976793, + "flos": 21130334215440.0, + "grad_norm": 2.5869367345764704, + "learning_rate": 2.182027121432696e-06, + "loss": 0.7011, + "num_input_tokens_seen": 86033115, + "step": 4049 + }, + { + "epoch": 0.4869837070883184, + "flos": 13893659464440.0, + "grad_norm": 3.6345120300018308, + "learning_rate": 2.1812513690008054e-06, + "loss": 0.8069, + "num_input_tokens_seen": 86051955, + "step": 4050 + }, + { + "epoch": 0.4871039499789575, + "flos": 11003933951520.0, + "grad_norm": 3.202056855535064, + "learning_rate": 2.180475589073227e-06, + "loss": 0.7783, + "num_input_tokens_seen": 86069375, + "step": 4051 + }, + { + "epoch": 0.48722419286959656, + "flos": 19183285113720.0, + "grad_norm": 2.7526325283315702, + "learning_rate": 2.1796997817676456e-06, + "loss": 0.7303, + "num_input_tokens_seen": 86090105, + "step": 4052 + }, + { + "epoch": 0.4873444357602357, + "flos": 17594346751800.0, + "grad_norm": 1.8077390558594428, + "learning_rate": 2.1789239472017494e-06, + "loss": 0.6637, + "num_input_tokens_seen": 86111475, + "step": 4053 + }, + { + "epoch": 0.4874646786508748, + "flos": 16701810734520.0, + "grad_norm": 4.3009834761720125, + "learning_rate": 2.1781480854932326e-06, + "loss": 0.71, + "num_input_tokens_seen": 86130960, + "step": 4054 + }, + { + "epoch": 0.48758492154151384, + "flos": 15567338065200.0, + "grad_norm": 6.564340387867101, + "learning_rate": 2.1773721967597933e-06, + "loss": 0.7846, + "num_input_tokens_seen": 86149130, + "step": 4055 + }, + { + "epoch": 0.48770516443215295, + "flos": 45880667954280.0, + "grad_norm": 0.9011970599176913, + "learning_rate": 2.1765962811191322e-06, + "loss": 0.6233, + "num_input_tokens_seen": 86203315, + "step": 4056 + }, + { + "epoch": 0.48782540732279206, + "flos": 48762134709000.0, + "grad_norm": 0.9647445936642162, + "learning_rate": 2.1758203386889566e-06, + "loss": 0.6756, + "num_input_tokens_seen": 86265805, + "step": 4057 + }, + { + "epoch": 0.4879456502134311, + "flos": 10623539091240.0, + "grad_norm": 4.217671800516266, + "learning_rate": 2.1750443695869746e-06, + "loss": 0.8243, + "num_input_tokens_seen": 86281095, + "step": 4058 + }, + { + "epoch": 0.4880658931040702, + "flos": 14244836673000.0, + "grad_norm": 2.7529729032558223, + "learning_rate": 2.174268373930901e-06, + "loss": 0.8404, + "num_input_tokens_seen": 86298330, + "step": 4059 + }, + { + "epoch": 0.48818613599470934, + "flos": 12188736974520.0, + "grad_norm": 2.660874234856955, + "learning_rate": 2.1734923518384537e-06, + "loss": 0.7878, + "num_input_tokens_seen": 86314655, + "step": 4060 + }, + { + "epoch": 0.4883063788853484, + "flos": 19614516887040.0, + "grad_norm": 3.0108527337392537, + "learning_rate": 2.1727163034273547e-06, + "loss": 0.8057, + "num_input_tokens_seen": 86332540, + "step": 4061 + }, + { + "epoch": 0.4884266217759875, + "flos": 12219126044760.0, + "grad_norm": 2.6945775011207043, + "learning_rate": 2.17194022881533e-06, + "loss": 0.7649, + "num_input_tokens_seen": 86348350, + "step": 4062 + }, + { + "epoch": 0.4885468646666266, + "flos": 17728087626000.0, + "grad_norm": 2.283895343567356, + "learning_rate": 2.1711641281201092e-06, + "loss": 0.6559, + "num_input_tokens_seen": 86368000, + "step": 4063 + }, + { + "epoch": 0.48866710755726567, + "flos": 10626071888040.0, + "grad_norm": 2.901076439330053, + "learning_rate": 2.1703880014594264e-06, + "loss": 0.7616, + "num_input_tokens_seen": 86385310, + "step": 4064 + }, + { + "epoch": 0.4887873504479048, + "flos": 21130429195320.0, + "grad_norm": 2.717047585849529, + "learning_rate": 2.1696118489510182e-06, + "loss": 0.7104, + "num_input_tokens_seen": 86405960, + "step": 4065 + }, + { + "epoch": 0.48890759333854383, + "flos": 16675189199520.0, + "grad_norm": 2.711177056801196, + "learning_rate": 2.1688356707126286e-06, + "loss": 0.698, + "num_input_tokens_seen": 86425300, + "step": 4066 + }, + { + "epoch": 0.48902783622918294, + "flos": 12976148230080.0, + "grad_norm": 3.197412970986271, + "learning_rate": 2.168059466862001e-06, + "loss": 0.6871, + "num_input_tokens_seen": 86443170, + "step": 4067 + }, + { + "epoch": 0.48914807911982205, + "flos": 16324360250520.0, + "grad_norm": 6.215888663827846, + "learning_rate": 2.167283237516887e-06, + "loss": 0.796, + "num_input_tokens_seen": 86461165, + "step": 4068 + }, + { + "epoch": 0.4892683220104611, + "flos": 11922964863960.0, + "grad_norm": 2.570709939856916, + "learning_rate": 2.1665069827950383e-06, + "loss": 0.7436, + "num_input_tokens_seen": 86478170, + "step": 4069 + }, + { + "epoch": 0.4893885649011002, + "flos": 11460742481160.0, + "grad_norm": 4.5813100105716345, + "learning_rate": 2.1657307028142126e-06, + "loss": 0.8569, + "num_input_tokens_seen": 86495430, + "step": 4070 + }, + { + "epoch": 0.48950880779173933, + "flos": 20965286132160.0, + "grad_norm": 3.630838493627757, + "learning_rate": 2.164954397692171e-06, + "loss": 0.654, + "num_input_tokens_seen": 86514575, + "step": 4071 + }, + { + "epoch": 0.4896290506823784, + "flos": 48798408040440.0, + "grad_norm": 1.1256128337399598, + "learning_rate": 2.164178067546678e-06, + "loss": 0.8204, + "num_input_tokens_seen": 86573460, + "step": 4072 + }, + { + "epoch": 0.4897492935730175, + "flos": 9086925660480.0, + "grad_norm": 2.3052372680815525, + "learning_rate": 2.163401712495504e-06, + "loss": 0.8977, + "num_input_tokens_seen": 86590875, + "step": 4073 + }, + { + "epoch": 0.4898695364636566, + "flos": 17187520316280.0, + "grad_norm": 1.8430790731115747, + "learning_rate": 2.1626253326564194e-06, + "loss": 0.7763, + "num_input_tokens_seen": 86609545, + "step": 4074 + }, + { + "epoch": 0.48998977935429566, + "flos": 19913242524600.0, + "grad_norm": 2.7185609247517175, + "learning_rate": 2.161848928147201e-06, + "loss": 0.754, + "num_input_tokens_seen": 86629535, + "step": 4075 + }, + { + "epoch": 0.4901100222449348, + "flos": 14865838467000.0, + "grad_norm": 2.2901336975715383, + "learning_rate": 2.161072499085629e-06, + "loss": 0.7751, + "num_input_tokens_seen": 86648250, + "step": 4076 + }, + { + "epoch": 0.4902302651355739, + "flos": 22345557968640.0, + "grad_norm": 1.9417117087688618, + "learning_rate": 2.160296045589487e-06, + "loss": 0.8132, + "num_input_tokens_seen": 86671430, + "step": 4077 + }, + { + "epoch": 0.49035050802621294, + "flos": 14003659860000.0, + "grad_norm": 2.876399155868642, + "learning_rate": 2.159519567776562e-06, + "loss": 0.6722, + "num_input_tokens_seen": 86690800, + "step": 4078 + }, + { + "epoch": 0.49047075091685205, + "flos": 16262568991320.0, + "grad_norm": 3.692341487222459, + "learning_rate": 2.1587430657646463e-06, + "loss": 0.6904, + "num_input_tokens_seen": 86703955, + "step": 4079 + }, + { + "epoch": 0.4905909938074911, + "flos": 14730261315120.0, + "grad_norm": 2.790317626975757, + "learning_rate": 2.157966539671533e-06, + "loss": 0.7557, + "num_input_tokens_seen": 86723315, + "step": 4080 + }, + { + "epoch": 0.4907112366981302, + "flos": 12544029977880.0, + "grad_norm": 3.280295969007539, + "learning_rate": 2.157189989615021e-06, + "loss": 0.6402, + "num_input_tokens_seen": 86741625, + "step": 4081 + }, + { + "epoch": 0.4908314795887693, + "flos": 15864385724880.0, + "grad_norm": 2.8118926346143858, + "learning_rate": 2.156413415712913e-06, + "loss": 0.7323, + "num_input_tokens_seen": 86763395, + "step": 4082 + }, + { + "epoch": 0.4909517224794084, + "flos": 19213040984760.0, + "grad_norm": 4.586092424781438, + "learning_rate": 2.155636818083014e-06, + "loss": 0.7637, + "num_input_tokens_seen": 86784485, + "step": 4083 + }, + { + "epoch": 0.4910719653700475, + "flos": 16944443905680.0, + "grad_norm": 2.8262030883836977, + "learning_rate": 2.154860196843134e-06, + "loss": 0.8234, + "num_input_tokens_seen": 86803400, + "step": 4084 + }, + { + "epoch": 0.4911922082606866, + "flos": 17080369317120.0, + "grad_norm": 2.342100045859889, + "learning_rate": 2.154083552111085e-06, + "loss": 0.752, + "num_input_tokens_seen": 86822290, + "step": 4085 + }, + { + "epoch": 0.49131245115132566, + "flos": 21425197337880.0, + "grad_norm": 8.369379235586264, + "learning_rate": 2.1533068840046834e-06, + "loss": 0.8006, + "num_input_tokens_seen": 86842275, + "step": 4086 + }, + { + "epoch": 0.49143269404196477, + "flos": 14722283005200.0, + "grad_norm": 8.433371204798807, + "learning_rate": 2.152530192641749e-06, + "loss": 0.5994, + "num_input_tokens_seen": 86856905, + "step": 4087 + }, + { + "epoch": 0.4915529369326039, + "flos": 17863316518320.0, + "grad_norm": 2.0758778748799305, + "learning_rate": 2.1517534781401068e-06, + "loss": 0.7067, + "num_input_tokens_seen": 86874505, + "step": 4088 + }, + { + "epoch": 0.49167317982324293, + "flos": 7601782341960.0, + "grad_norm": 3.066375771146297, + "learning_rate": 2.150976740617581e-06, + "loss": 0.6891, + "num_input_tokens_seen": 86890785, + "step": 4089 + }, + { + "epoch": 0.49179342271388204, + "flos": 18754016257920.0, + "grad_norm": 2.8924157446720646, + "learning_rate": 2.150199980192006e-06, + "loss": 0.711, + "num_input_tokens_seen": 86909625, + "step": 4090 + }, + { + "epoch": 0.49191366560452116, + "flos": 15429576376080.0, + "grad_norm": 2.2367472826343326, + "learning_rate": 2.1494231969812114e-06, + "loss": 0.7992, + "num_input_tokens_seen": 86928335, + "step": 4091 + }, + { + "epoch": 0.4920339084951602, + "flos": 19104845206920.0, + "grad_norm": 2.9770440140181615, + "learning_rate": 2.1486463911030372e-06, + "loss": 0.7799, + "num_input_tokens_seen": 86948705, + "step": 4092 + }, + { + "epoch": 0.4921541513857993, + "flos": 18374792816160.0, + "grad_norm": 2.984907332472361, + "learning_rate": 2.147869562675324e-06, + "loss": 0.7337, + "num_input_tokens_seen": 86967395, + "step": 4093 + }, + { + "epoch": 0.49227439427643843, + "flos": 17863316518320.0, + "grad_norm": 2.5555494994861143, + "learning_rate": 2.147092711815915e-06, + "loss": 0.7046, + "num_input_tokens_seen": 86986465, + "step": 4094 + }, + { + "epoch": 0.4923946371670775, + "flos": 8223069075600.0, + "grad_norm": 4.6400791450853776, + "learning_rate": 2.1463158386426593e-06, + "loss": 0.8508, + "num_input_tokens_seen": 87003995, + "step": 4095 + }, + { + "epoch": 0.4925148800577166, + "flos": 22345209709080.0, + "grad_norm": 9.5721976778475, + "learning_rate": 2.145538943273407e-06, + "loss": 0.7586, + "num_input_tokens_seen": 87023990, + "step": 4096 + }, + { + "epoch": 0.49263512294835565, + "flos": 15243922150200.0, + "grad_norm": 2.8328351162461614, + "learning_rate": 2.144762025826013e-06, + "loss": 0.705, + "num_input_tokens_seen": 87042800, + "step": 4097 + }, + { + "epoch": 0.49275536583899476, + "flos": 17402455513800.0, + "grad_norm": 2.464528204160789, + "learning_rate": 2.143985086418334e-06, + "loss": 0.8498, + "num_input_tokens_seen": 87057700, + "step": 4098 + }, + { + "epoch": 0.4928756087296339, + "flos": 16297105516320.0, + "grad_norm": 2.0262912965833957, + "learning_rate": 2.1432081251682324e-06, + "loss": 0.7601, + "num_input_tokens_seen": 87077790, + "step": 4099 + }, + { + "epoch": 0.49299585162027293, + "flos": 14352019332120.0, + "grad_norm": 2.523006059830593, + "learning_rate": 2.142431142193572e-06, + "loss": 0.8513, + "num_input_tokens_seen": 87095290, + "step": 4100 + }, + { + "epoch": 0.49311609451091204, + "flos": 28555164329160.0, + "grad_norm": 2.91946670914119, + "learning_rate": 2.1416541376122207e-06, + "loss": 0.7108, + "num_input_tokens_seen": 87115190, + "step": 4101 + }, + { + "epoch": 0.49323633740155115, + "flos": 20778492147720.0, + "grad_norm": 2.4251934328307727, + "learning_rate": 2.1408771115420496e-06, + "loss": 0.7192, + "num_input_tokens_seen": 87134770, + "step": 4102 + }, + { + "epoch": 0.4933565802921902, + "flos": 15453981713880.0, + "grad_norm": 2.220169580357693, + "learning_rate": 2.140100064100932e-06, + "loss": 0.6435, + "num_input_tokens_seen": 87150465, + "step": 4103 + }, + { + "epoch": 0.4934768231828293, + "flos": 13162973874480.0, + "grad_norm": 3.141491822521854, + "learning_rate": 2.139322995406746e-06, + "loss": 0.75, + "num_input_tokens_seen": 87167820, + "step": 4104 + }, + { + "epoch": 0.4935970660734684, + "flos": 17181473263920.0, + "grad_norm": 2.6790616831126917, + "learning_rate": 2.1385459055773727e-06, + "loss": 0.7908, + "num_input_tokens_seen": 87185730, + "step": 4105 + }, + { + "epoch": 0.4937173089641075, + "flos": 47531013164520.0, + "grad_norm": 2.3688800665341425, + "learning_rate": 2.137768794730696e-06, + "loss": 0.7197, + "num_input_tokens_seen": 87208900, + "step": 4106 + }, + { + "epoch": 0.4938375518547466, + "flos": 16350696845880.0, + "grad_norm": 3.2003259850400334, + "learning_rate": 2.1369916629846026e-06, + "loss": 0.7886, + "num_input_tokens_seen": 87228370, + "step": 4107 + }, + { + "epoch": 0.4939577947453857, + "flos": 13029834539520.0, + "grad_norm": 2.149994331174962, + "learning_rate": 2.136214510456983e-06, + "loss": 0.7349, + "num_input_tokens_seen": 87246545, + "step": 4108 + }, + { + "epoch": 0.49407803763602476, + "flos": 51672023125080.0, + "grad_norm": 0.947117650380491, + "learning_rate": 2.1354373372657296e-06, + "loss": 0.6977, + "num_input_tokens_seen": 87304705, + "step": 4109 + }, + { + "epoch": 0.49419828052666387, + "flos": 17809883488560.0, + "grad_norm": 3.178305285656267, + "learning_rate": 2.1346601435287404e-06, + "loss": 0.6944, + "num_input_tokens_seen": 87326695, + "step": 4110 + }, + { + "epoch": 0.494318523417303, + "flos": 21560552870040.0, + "grad_norm": 5.6334053817445415, + "learning_rate": 2.1338829293639144e-06, + "loss": 0.7862, + "num_input_tokens_seen": 87346775, + "step": 4111 + }, + { + "epoch": 0.49443876630794203, + "flos": 11112762928560.0, + "grad_norm": 3.1815796267268897, + "learning_rate": 2.1331056948891547e-06, + "loss": 0.8191, + "num_input_tokens_seen": 87363595, + "step": 4112 + }, + { + "epoch": 0.49455900919858115, + "flos": 8950873609200.0, + "grad_norm": 3.5176955165898773, + "learning_rate": 2.1323284402223666e-06, + "loss": 0.7492, + "num_input_tokens_seen": 87379305, + "step": 4113 + }, + { + "epoch": 0.4946792520892202, + "flos": 16670978424840.0, + "grad_norm": 2.0769679928674867, + "learning_rate": 2.1315511654814597e-06, + "loss": 0.8679, + "num_input_tokens_seen": 87397435, + "step": 4114 + }, + { + "epoch": 0.4947994949798593, + "flos": 16944222285960.0, + "grad_norm": 2.783136135432841, + "learning_rate": 2.1307738707843456e-06, + "loss": 0.763, + "num_input_tokens_seen": 87416820, + "step": 4115 + }, + { + "epoch": 0.4949197378704984, + "flos": 17323192448040.0, + "grad_norm": 7.490483413395788, + "learning_rate": 2.1299965562489385e-06, + "loss": 0.6671, + "num_input_tokens_seen": 87436345, + "step": 4116 + }, + { + "epoch": 0.4950399807611375, + "flos": 19730121095520.0, + "grad_norm": 1.7150019613983238, + "learning_rate": 2.129219221993158e-06, + "loss": 0.779, + "num_input_tokens_seen": 87460850, + "step": 4117 + }, + { + "epoch": 0.4951602236517766, + "flos": 49633616852880.0, + "grad_norm": 0.8220788014037755, + "learning_rate": 2.128441868134924e-06, + "loss": 0.6367, + "num_input_tokens_seen": 87522505, + "step": 4118 + }, + { + "epoch": 0.4952804665424157, + "flos": 14541536073120.0, + "grad_norm": 5.308410147326253, + "learning_rate": 2.1276644947921606e-06, + "loss": 0.8182, + "num_input_tokens_seen": 87541140, + "step": 4119 + }, + { + "epoch": 0.49540070943305475, + "flos": 13731429117600.0, + "grad_norm": 2.5947821528258936, + "learning_rate": 2.126887102082795e-06, + "loss": 0.8108, + "num_input_tokens_seen": 87560885, + "step": 4120 + }, + { + "epoch": 0.49552095232369386, + "flos": 18265773879360.0, + "grad_norm": 2.1454930328846316, + "learning_rate": 2.126109690124757e-06, + "loss": 0.6832, + "num_input_tokens_seen": 87581420, + "step": 4121 + }, + { + "epoch": 0.495641195214333, + "flos": 16729097128680.0, + "grad_norm": 3.3706727720921585, + "learning_rate": 2.1253322590359786e-06, + "loss": 0.696, + "num_input_tokens_seen": 87600475, + "step": 4122 + }, + { + "epoch": 0.49576143810497203, + "flos": 18883641337320.0, + "grad_norm": 2.904022511681349, + "learning_rate": 2.124554808934397e-06, + "loss": 0.7122, + "num_input_tokens_seen": 87620775, + "step": 4123 + }, + { + "epoch": 0.49588168099561114, + "flos": 16190207796840.0, + "grad_norm": 3.949555968424461, + "learning_rate": 2.1237773399379496e-06, + "loss": 0.7178, + "num_input_tokens_seen": 87641460, + "step": 4124 + }, + { + "epoch": 0.49600192388625025, + "flos": 17861131981080.0, + "grad_norm": 2.2133063576483027, + "learning_rate": 2.122999852164578e-06, + "loss": 0.8526, + "num_input_tokens_seen": 87661800, + "step": 4125 + }, + { + "epoch": 0.4961221667768893, + "flos": 16729255428480.0, + "grad_norm": 4.052979331829043, + "learning_rate": 2.122222345732227e-06, + "loss": 0.5634, + "num_input_tokens_seen": 87681435, + "step": 4126 + }, + { + "epoch": 0.4962424096675284, + "flos": 13029296320200.0, + "grad_norm": 2.4829611923989034, + "learning_rate": 2.121444820758843e-06, + "loss": 0.818, + "num_input_tokens_seen": 87699795, + "step": 4127 + }, + { + "epoch": 0.49636265255816747, + "flos": 15941147653800.0, + "grad_norm": 2.3564795101512543, + "learning_rate": 2.120667277362376e-06, + "loss": 0.7585, + "num_input_tokens_seen": 87718230, + "step": 4128 + }, + { + "epoch": 0.4964828954488066, + "flos": 11919070688880.0, + "grad_norm": 2.4032886662699573, + "learning_rate": 2.1198897156607796e-06, + "loss": 0.8341, + "num_input_tokens_seen": 87735305, + "step": 4129 + }, + { + "epoch": 0.4966031383394457, + "flos": 18100092596880.0, + "grad_norm": 2.736696663652224, + "learning_rate": 2.1191121357720085e-06, + "loss": 0.7261, + "num_input_tokens_seen": 87753085, + "step": 4130 + }, + { + "epoch": 0.49672338123008475, + "flos": 16783448297280.0, + "grad_norm": 2.459037992273122, + "learning_rate": 2.1183345378140206e-06, + "loss": 0.7295, + "num_input_tokens_seen": 87772550, + "step": 4131 + }, + { + "epoch": 0.49684362412072386, + "flos": 48644630902920.0, + "grad_norm": 1.0050202953130294, + "learning_rate": 2.1175569219047783e-06, + "loss": 0.6546, + "num_input_tokens_seen": 87833710, + "step": 4132 + }, + { + "epoch": 0.49696386701136297, + "flos": 14592784565640.0, + "grad_norm": 3.252279802473062, + "learning_rate": 2.1167792881622437e-06, + "loss": 0.7095, + "num_input_tokens_seen": 87852450, + "step": 4133 + }, + { + "epoch": 0.497084109902002, + "flos": 18129943447800.0, + "grad_norm": 2.016568931334056, + "learning_rate": 2.116001636704384e-06, + "loss": 0.796, + "num_input_tokens_seen": 87872555, + "step": 4134 + }, + { + "epoch": 0.49720435279264114, + "flos": 15999772917000.0, + "grad_norm": 2.140619765226549, + "learning_rate": 2.1152239676491685e-06, + "loss": 0.7787, + "num_input_tokens_seen": 87890380, + "step": 4135 + }, + { + "epoch": 0.49732459568328025, + "flos": 17431071626280.0, + "grad_norm": 4.301614706579647, + "learning_rate": 2.114446281114569e-06, + "loss": 0.7285, + "num_input_tokens_seen": 87909120, + "step": 4136 + }, + { + "epoch": 0.4974448385739193, + "flos": 14649858490800.0, + "grad_norm": 2.3832921280222186, + "learning_rate": 2.1136685772185587e-06, + "loss": 0.738, + "num_input_tokens_seen": 87927865, + "step": 4137 + }, + { + "epoch": 0.4975650814645584, + "flos": 18181793479560.0, + "grad_norm": 5.8407820238096715, + "learning_rate": 2.1128908560791163e-06, + "loss": 0.771, + "num_input_tokens_seen": 87947415, + "step": 4138 + }, + { + "epoch": 0.4976853243551975, + "flos": 14488419642960.0, + "grad_norm": 3.3927325892711093, + "learning_rate": 2.1121131178142203e-06, + "loss": 0.7617, + "num_input_tokens_seen": 87966500, + "step": 4139 + }, + { + "epoch": 0.4978055672458366, + "flos": 16939694911680.0, + "grad_norm": 1.8661419482931085, + "learning_rate": 2.1113353625418544e-06, + "loss": 0.8086, + "num_input_tokens_seen": 87984770, + "step": 4140 + }, + { + "epoch": 0.4979258101364757, + "flos": 11328552945000.0, + "grad_norm": 2.0301217099990994, + "learning_rate": 2.1105575903800017e-06, + "loss": 0.7766, + "num_input_tokens_seen": 88003210, + "step": 4141 + }, + { + "epoch": 0.4980460530271148, + "flos": 19318007446680.0, + "grad_norm": 3.6875164225584385, + "learning_rate": 2.1097798014466502e-06, + "loss": 0.8344, + "num_input_tokens_seen": 88022530, + "step": 4142 + }, + { + "epoch": 0.49816629591775385, + "flos": 12597463007640.0, + "grad_norm": 3.4905390251588977, + "learning_rate": 2.109001995859791e-06, + "loss": 0.5542, + "num_input_tokens_seen": 88041150, + "step": 4143 + }, + { + "epoch": 0.49828653880839296, + "flos": 47872416428160.0, + "grad_norm": 0.8370066784030548, + "learning_rate": 2.108224173737415e-06, + "loss": 0.6253, + "num_input_tokens_seen": 88104170, + "step": 4144 + }, + { + "epoch": 0.498406781699032, + "flos": 20152931319480.0, + "grad_norm": 2.0229835737526027, + "learning_rate": 2.1074463351975183e-06, + "loss": 0.7393, + "num_input_tokens_seen": 88122775, + "step": 4145 + }, + { + "epoch": 0.49852702458967113, + "flos": 23124579334800.0, + "grad_norm": 2.0561326894950875, + "learning_rate": 2.106668480358098e-06, + "loss": 0.6949, + "num_input_tokens_seen": 88142720, + "step": 4146 + }, + { + "epoch": 0.49864726748031024, + "flos": 16725234613560.0, + "grad_norm": 1.9776948843894895, + "learning_rate": 2.105890609337154e-06, + "loss": 0.6929, + "num_input_tokens_seen": 88160955, + "step": 4147 + }, + { + "epoch": 0.4987675103709493, + "flos": 51924375903960.0, + "grad_norm": 0.6897339073507778, + "learning_rate": 2.1051127222526883e-06, + "loss": 0.6392, + "num_input_tokens_seen": 88232790, + "step": 4148 + }, + { + "epoch": 0.4988877532615884, + "flos": 21103079481240.0, + "grad_norm": 2.1560368086976123, + "learning_rate": 2.1043348192227067e-06, + "loss": 0.7954, + "num_input_tokens_seen": 88252880, + "step": 4149 + }, + { + "epoch": 0.4990079961522275, + "flos": 12300003768480.0, + "grad_norm": 2.207343209423821, + "learning_rate": 2.1035569003652156e-06, + "loss": 0.6131, + "num_input_tokens_seen": 88271230, + "step": 4150 + }, + { + "epoch": 0.4991282390428666, + "flos": 9648479032320.0, + "grad_norm": 3.3209974592199716, + "learning_rate": 2.1027789657982255e-06, + "loss": 0.8001, + "num_input_tokens_seen": 88285165, + "step": 4151 + }, + { + "epoch": 0.4992484819335057, + "flos": 15751504272960.0, + "grad_norm": 2.7604251405597875, + "learning_rate": 2.1020010156397482e-06, + "loss": 0.7616, + "num_input_tokens_seen": 88302105, + "step": 4152 + }, + { + "epoch": 0.4993687248241448, + "flos": 18210789511560.0, + "grad_norm": 2.5792221971325766, + "learning_rate": 2.101223050007797e-06, + "loss": 0.7646, + "num_input_tokens_seen": 88320375, + "step": 4153 + }, + { + "epoch": 0.49948896771478385, + "flos": 46397463125400.0, + "grad_norm": 0.8577831784432738, + "learning_rate": 2.1004450690203904e-06, + "loss": 0.5744, + "num_input_tokens_seen": 88376175, + "step": 4154 + }, + { + "epoch": 0.49960921060542296, + "flos": 50353574207760.0, + "grad_norm": 1.0187569085571446, + "learning_rate": 2.099667072795546e-06, + "loss": 0.7064, + "num_input_tokens_seen": 88438015, + "step": 4155 + }, + { + "epoch": 0.49972945349606207, + "flos": 17322622568760.0, + "grad_norm": 2.3969525141908097, + "learning_rate": 2.0988890614512864e-06, + "loss": 0.776, + "num_input_tokens_seen": 88457625, + "step": 4156 + }, + { + "epoch": 0.4998496963867011, + "flos": 14433815194680.0, + "grad_norm": 3.1091414320743134, + "learning_rate": 2.098111035105635e-06, + "loss": 0.8238, + "num_input_tokens_seen": 88475770, + "step": 4157 + }, + { + "epoch": 0.49996993927734024, + "flos": 16290235305000.0, + "grad_norm": 2.0746281394045254, + "learning_rate": 2.0973329938766176e-06, + "loss": 0.728, + "num_input_tokens_seen": 88492920, + "step": 4158 + }, + { + "epoch": 0.5000901821679793, + "flos": 17076285182280.0, + "grad_norm": 2.233594851578624, + "learning_rate": 2.0965549378822618e-06, + "loss": 0.7807, + "num_input_tokens_seen": 88513930, + "step": 4159 + }, + { + "epoch": 0.5002104250586185, + "flos": 14865331907640.0, + "grad_norm": 4.409139132392393, + "learning_rate": 2.095776867240599e-06, + "loss": 0.8229, + "num_input_tokens_seen": 88530640, + "step": 4160 + }, + { + "epoch": 0.5003306679492575, + "flos": 10167775340280.0, + "grad_norm": 3.6230373989094407, + "learning_rate": 2.094998782069661e-06, + "loss": 0.7962, + "num_input_tokens_seen": 88548065, + "step": 4161 + }, + { + "epoch": 0.5004509108398966, + "flos": 20292814225920.0, + "grad_norm": 4.418006525797274, + "learning_rate": 2.0942206824874845e-06, + "loss": 0.7377, + "num_input_tokens_seen": 88570560, + "step": 4162 + }, + { + "epoch": 0.5005711537305357, + "flos": 10896561332640.0, + "grad_norm": 6.23195270337238, + "learning_rate": 2.093442568612105e-06, + "loss": 0.7735, + "num_input_tokens_seen": 88588085, + "step": 4163 + }, + { + "epoch": 0.5006913966211748, + "flos": 19427026383480.0, + "grad_norm": 3.1269645619227915, + "learning_rate": 2.0926644405615613e-06, + "loss": 0.8266, + "num_input_tokens_seen": 88608705, + "step": 4164 + }, + { + "epoch": 0.5008116395118138, + "flos": 14947064450280.0, + "grad_norm": 3.9164573667619647, + "learning_rate": 2.091886298453897e-06, + "loss": 0.8051, + "num_input_tokens_seen": 88626610, + "step": 4165 + }, + { + "epoch": 0.500931882402453, + "flos": 15783697960920.0, + "grad_norm": 3.6625273199006387, + "learning_rate": 2.091108142407153e-06, + "loss": 0.703, + "num_input_tokens_seen": 88645070, + "step": 4166 + }, + { + "epoch": 0.5010521252930921, + "flos": 42584690376480.0, + "grad_norm": 0.9059358907463568, + "learning_rate": 2.090329972539377e-06, + "loss": 0.6812, + "num_input_tokens_seen": 88703355, + "step": 4167 + }, + { + "epoch": 0.5011723681837311, + "flos": 13597403303760.0, + "grad_norm": 5.326765049076364, + "learning_rate": 2.089551788968616e-06, + "loss": 0.6665, + "num_input_tokens_seen": 88721040, + "step": 4168 + }, + { + "epoch": 0.5012926110743702, + "flos": 41176530606600.0, + "grad_norm": 0.878390046164132, + "learning_rate": 2.08877359181292e-06, + "loss": 0.634, + "num_input_tokens_seen": 88777325, + "step": 4169 + }, + { + "epoch": 0.5014128539650093, + "flos": 17750371746480.0, + "grad_norm": 26.40379981236129, + "learning_rate": 2.0879953811903396e-06, + "loss": 0.8258, + "num_input_tokens_seen": 88791930, + "step": 4170 + }, + { + "epoch": 0.5015330968556484, + "flos": 20185378287120.0, + "grad_norm": 3.6398232031424653, + "learning_rate": 2.08721715721893e-06, + "loss": 0.7657, + "num_input_tokens_seen": 88810975, + "step": 4171 + }, + { + "epoch": 0.5016533397462875, + "flos": 17426670891840.0, + "grad_norm": 2.8951015142322665, + "learning_rate": 2.0864389200167477e-06, + "loss": 0.7628, + "num_input_tokens_seen": 88828950, + "step": 4172 + }, + { + "epoch": 0.5017735826369266, + "flos": 18532242509040.0, + "grad_norm": 11.874251492651744, + "learning_rate": 2.0856606697018504e-06, + "loss": 0.7811, + "num_input_tokens_seen": 88846680, + "step": 4173 + }, + { + "epoch": 0.5018938255275657, + "flos": 12300447007920.0, + "grad_norm": 6.397979310040709, + "learning_rate": 2.084882406392297e-06, + "loss": 0.7096, + "num_input_tokens_seen": 88864360, + "step": 4174 + }, + { + "epoch": 0.5020140684182047, + "flos": 18695580954480.0, + "grad_norm": 8.840728782888753, + "learning_rate": 2.0841041302061496e-06, + "loss": 0.6817, + "num_input_tokens_seen": 88883540, + "step": 4175 + }, + { + "epoch": 0.5021343113088439, + "flos": 17322559248840.0, + "grad_norm": 8.280917409300038, + "learning_rate": 2.083325841261473e-06, + "loss": 0.7361, + "num_input_tokens_seen": 88902320, + "step": 4176 + }, + { + "epoch": 0.502254554199483, + "flos": 17970372537600.0, + "grad_norm": 2.516947049401995, + "learning_rate": 2.0825475396763322e-06, + "loss": 0.6451, + "num_input_tokens_seen": 88922690, + "step": 4177 + }, + { + "epoch": 0.502374797090122, + "flos": 25153234339320.0, + "grad_norm": 2.3431263411775762, + "learning_rate": 2.081769225568796e-06, + "loss": 0.6366, + "num_input_tokens_seen": 88944860, + "step": 4178 + }, + { + "epoch": 0.5024950399807612, + "flos": 19077907072320.0, + "grad_norm": 3.210676270785283, + "learning_rate": 2.0809908990569327e-06, + "loss": 0.7394, + "num_input_tokens_seen": 88966360, + "step": 4179 + }, + { + "epoch": 0.5026152828714002, + "flos": 15541508029200.0, + "grad_norm": 3.548509263287639, + "learning_rate": 2.0802125602588146e-06, + "loss": 0.7792, + "num_input_tokens_seen": 88985345, + "step": 4180 + }, + { + "epoch": 0.5027355257620393, + "flos": 22722660193080.0, + "grad_norm": 4.166646703036484, + "learning_rate": 2.0794342092925146e-06, + "loss": 0.6487, + "num_input_tokens_seen": 89006200, + "step": 4181 + }, + { + "epoch": 0.5028557686526784, + "flos": 18160332518040.0, + "grad_norm": 4.367160858455187, + "learning_rate": 2.078655846276108e-06, + "loss": 0.6706, + "num_input_tokens_seen": 89026250, + "step": 4182 + }, + { + "epoch": 0.5029760115433175, + "flos": 16810354771920.0, + "grad_norm": 4.271773940302307, + "learning_rate": 2.0778774713276727e-06, + "loss": 0.6707, + "num_input_tokens_seen": 89045445, + "step": 4183 + }, + { + "epoch": 0.5030962544339566, + "flos": 11140112642640.0, + "grad_norm": 3.969949976416149, + "learning_rate": 2.077099084565287e-06, + "loss": 0.6496, + "num_input_tokens_seen": 89062570, + "step": 4184 + }, + { + "epoch": 0.5032164973245957, + "flos": 17940426706800.0, + "grad_norm": 9.126727633034617, + "learning_rate": 2.0763206861070313e-06, + "loss": 0.6229, + "num_input_tokens_seen": 89081350, + "step": 4185 + }, + { + "epoch": 0.5033367402152348, + "flos": 11811761389920.0, + "grad_norm": 6.448778069248917, + "learning_rate": 2.0755422760709876e-06, + "loss": 0.7402, + "num_input_tokens_seen": 89098470, + "step": 4186 + }, + { + "epoch": 0.5034569831058738, + "flos": 15643340155080.0, + "grad_norm": 5.370549688353895, + "learning_rate": 2.0747638545752417e-06, + "loss": 0.7529, + "num_input_tokens_seen": 89116750, + "step": 4187 + }, + { + "epoch": 0.503577225996513, + "flos": 15028068813840.0, + "grad_norm": 10.109977598080592, + "learning_rate": 2.073985421737878e-06, + "loss": 0.8147, + "num_input_tokens_seen": 89133780, + "step": 4188 + }, + { + "epoch": 0.5036974688871521, + "flos": 19963699518120.0, + "grad_norm": 7.03671283012269, + "learning_rate": 2.0732069776769844e-06, + "loss": 0.7361, + "num_input_tokens_seen": 89150910, + "step": 4189 + }, + { + "epoch": 0.5038177117777911, + "flos": 14919493116480.0, + "grad_norm": 5.756508833608788, + "learning_rate": 2.072428522510651e-06, + "loss": 0.7143, + "num_input_tokens_seen": 89167195, + "step": 4190 + }, + { + "epoch": 0.5039379546684303, + "flos": 16026489431880.0, + "grad_norm": 8.598690873686001, + "learning_rate": 2.071650056356968e-06, + "loss": 0.7347, + "num_input_tokens_seen": 89184455, + "step": 4191 + }, + { + "epoch": 0.5040581975590693, + "flos": 14622382136880.0, + "grad_norm": 4.8372442054640885, + "learning_rate": 2.070871579334028e-06, + "loss": 0.7706, + "num_input_tokens_seen": 89203285, + "step": 4192 + }, + { + "epoch": 0.5041784404497084, + "flos": 15324863193840.0, + "grad_norm": 6.155975490588202, + "learning_rate": 2.0700930915599264e-06, + "loss": 0.7038, + "num_input_tokens_seen": 89222735, + "step": 4193 + }, + { + "epoch": 0.5042986833403476, + "flos": 9060367445400.0, + "grad_norm": 5.244816900935706, + "learning_rate": 2.0693145931527583e-06, + "loss": 0.77, + "num_input_tokens_seen": 89238935, + "step": 4194 + }, + { + "epoch": 0.5044189262309866, + "flos": 21424374178920.0, + "grad_norm": 9.314892762284039, + "learning_rate": 2.068536084230622e-06, + "loss": 0.7702, + "num_input_tokens_seen": 89260520, + "step": 4195 + }, + { + "epoch": 0.5045391691216257, + "flos": 17480167241520.0, + "grad_norm": 3.669019522181695, + "learning_rate": 2.067757564911616e-06, + "loss": 0.8636, + "num_input_tokens_seen": 89278815, + "step": 4196 + }, + { + "epoch": 0.5046594120122648, + "flos": 18052136740200.0, + "grad_norm": 3.521664056832652, + "learning_rate": 2.0669790353138407e-06, + "loss": 0.8982, + "num_input_tokens_seen": 89297500, + "step": 4197 + }, + { + "epoch": 0.5047796549029039, + "flos": 17102526797760.0, + "grad_norm": 4.53202865750443, + "learning_rate": 2.0662004955553995e-06, + "loss": 0.7167, + "num_input_tokens_seen": 89316920, + "step": 4198 + }, + { + "epoch": 0.5048998977935429, + "flos": 12618354089880.0, + "grad_norm": 4.483120945623254, + "learning_rate": 2.065421945754395e-06, + "loss": 0.7473, + "num_input_tokens_seen": 89334370, + "step": 4199 + }, + { + "epoch": 0.505020140684182, + "flos": 25609124730120.0, + "grad_norm": 5.681775576345054, + "learning_rate": 2.0646433860289344e-06, + "loss": 0.7737, + "num_input_tokens_seen": 89353015, + "step": 4200 + }, + { + "epoch": 0.5051403835748212, + "flos": 18214208787240.0, + "grad_norm": 4.458090493311864, + "learning_rate": 2.0638648164971233e-06, + "loss": 0.8045, + "num_input_tokens_seen": 89371200, + "step": 4201 + }, + { + "epoch": 0.5052606264654602, + "flos": 15324736554000.0, + "grad_norm": 3.3135021780987466, + "learning_rate": 2.06308623727707e-06, + "loss": 0.872, + "num_input_tokens_seen": 89391020, + "step": 4202 + }, + { + "epoch": 0.5053808693560993, + "flos": 14349296575560.0, + "grad_norm": 6.110339934886851, + "learning_rate": 2.0623076484868846e-06, + "loss": 0.7361, + "num_input_tokens_seen": 89408195, + "step": 4203 + }, + { + "epoch": 0.5055011122467384, + "flos": 49774987777440.0, + "grad_norm": 0.9145420479344761, + "learning_rate": 2.061529050244679e-06, + "loss": 0.6789, + "num_input_tokens_seen": 89467660, + "step": 4204 + }, + { + "epoch": 0.5056213551373775, + "flos": 12300225388200.0, + "grad_norm": 19.822685054589286, + "learning_rate": 2.060750442668565e-06, + "loss": 0.7271, + "num_input_tokens_seen": 89485135, + "step": 4205 + }, + { + "epoch": 0.5057415980280165, + "flos": 11160528825480.0, + "grad_norm": 15.644632685376527, + "learning_rate": 2.059971825876657e-06, + "loss": 0.6235, + "num_input_tokens_seen": 89499365, + "step": 4206 + }, + { + "epoch": 0.5058618409186557, + "flos": 13893659464440.0, + "grad_norm": 3.187630445017726, + "learning_rate": 2.0591931999870713e-06, + "loss": 0.7552, + "num_input_tokens_seen": 89518010, + "step": 4207 + }, + { + "epoch": 0.5059820838092948, + "flos": 46776116687880.0, + "grad_norm": 0.9405331500253137, + "learning_rate": 2.0584145651179234e-06, + "loss": 0.6281, + "num_input_tokens_seen": 89573440, + "step": 4208 + }, + { + "epoch": 0.5061023266999338, + "flos": 11239918531080.0, + "grad_norm": 3.7450191479996837, + "learning_rate": 2.0576359213873327e-06, + "loss": 0.7697, + "num_input_tokens_seen": 89588310, + "step": 4209 + }, + { + "epoch": 0.506222569590573, + "flos": 16427807034360.0, + "grad_norm": 3.982042295467931, + "learning_rate": 2.056857268913419e-06, + "loss": 0.6659, + "num_input_tokens_seen": 89608080, + "step": 4210 + }, + { + "epoch": 0.506342812481212, + "flos": 12806636092440.0, + "grad_norm": 4.726591756430632, + "learning_rate": 2.056078607814303e-06, + "loss": 0.826, + "num_input_tokens_seen": 89623585, + "step": 4211 + }, + { + "epoch": 0.5064630553718511, + "flos": 17131934409240.0, + "grad_norm": 3.7530934608017286, + "learning_rate": 2.055299938208106e-06, + "loss": 0.7685, + "num_input_tokens_seen": 89644295, + "step": 4212 + }, + { + "epoch": 0.5065832982624903, + "flos": 17565034120200.0, + "grad_norm": 3.7500385282444593, + "learning_rate": 2.0545212602129526e-06, + "loss": 0.8479, + "num_input_tokens_seen": 89663870, + "step": 4213 + }, + { + "epoch": 0.5067035411531293, + "flos": 15726307436160.0, + "grad_norm": 10.836417673661664, + "learning_rate": 2.0537425739469673e-06, + "loss": 0.6208, + "num_input_tokens_seen": 89682525, + "step": 4214 + }, + { + "epoch": 0.5068237840437684, + "flos": 48612468874920.0, + "grad_norm": 1.0349148450825632, + "learning_rate": 2.052963879528276e-06, + "loss": 0.6325, + "num_input_tokens_seen": 89742115, + "step": 4215 + }, + { + "epoch": 0.5069440269344075, + "flos": 19989434574240.0, + "grad_norm": 4.233950885774352, + "learning_rate": 2.052185177075007e-06, + "loss": 0.7515, + "num_input_tokens_seen": 89761405, + "step": 4216 + }, + { + "epoch": 0.5070642698250466, + "flos": 17105597813880.0, + "grad_norm": 3.6026149047909186, + "learning_rate": 2.051406466705288e-06, + "loss": 0.8107, + "num_input_tokens_seen": 89780665, + "step": 4217 + }, + { + "epoch": 0.5071845127156857, + "flos": 14866313366400.0, + "grad_norm": 4.920976054129646, + "learning_rate": 2.0506277485372486e-06, + "loss": 0.7905, + "num_input_tokens_seen": 89799210, + "step": 4218 + }, + { + "epoch": 0.5073047556063248, + "flos": 8762654926560.0, + "grad_norm": 4.181298289568987, + "learning_rate": 2.04984902268902e-06, + "loss": 0.6484, + "num_input_tokens_seen": 89816130, + "step": 4219 + }, + { + "epoch": 0.5074249984969639, + "flos": 14379812285640.0, + "grad_norm": 7.238295215937012, + "learning_rate": 2.0490702892787345e-06, + "loss": 0.7218, + "num_input_tokens_seen": 89834910, + "step": 4220 + }, + { + "epoch": 0.5075452413876029, + "flos": 21101433163320.0, + "grad_norm": 9.268200675546847, + "learning_rate": 2.0482915484245246e-06, + "loss": 0.613, + "num_input_tokens_seen": 89856250, + "step": 4221 + }, + { + "epoch": 0.5076654842782421, + "flos": 14865711827160.0, + "grad_norm": 11.64072986529947, + "learning_rate": 2.047512800244526e-06, + "loss": 0.8142, + "num_input_tokens_seen": 89871235, + "step": 4222 + }, + { + "epoch": 0.5077857271688812, + "flos": 19321078462800.0, + "grad_norm": 3.5161814609523803, + "learning_rate": 2.046734044856873e-06, + "loss": 0.7636, + "num_input_tokens_seen": 89890365, + "step": 4223 + }, + { + "epoch": 0.5079059700595202, + "flos": 15945864987840.0, + "grad_norm": 4.505285973146333, + "learning_rate": 2.045955282379702e-06, + "loss": 0.7879, + "num_input_tokens_seen": 89908745, + "step": 4224 + }, + { + "epoch": 0.5080262129501594, + "flos": 9573078481680.0, + "grad_norm": 6.116704188482105, + "learning_rate": 2.045176512931152e-06, + "loss": 0.7338, + "num_input_tokens_seen": 89923095, + "step": 4225 + }, + { + "epoch": 0.5081464558407984, + "flos": 18537941301840.0, + "grad_norm": 3.4658944688195774, + "learning_rate": 2.0443977366293604e-06, + "loss": 0.7379, + "num_input_tokens_seen": 89940855, + "step": 4226 + }, + { + "epoch": 0.5082666987314375, + "flos": 22718354438520.0, + "grad_norm": 5.231759855702872, + "learning_rate": 2.043618953592468e-06, + "loss": 0.7565, + "num_input_tokens_seen": 89963030, + "step": 4227 + }, + { + "epoch": 0.5083869416220766, + "flos": 14379495686040.0, + "grad_norm": 5.346696364842711, + "learning_rate": 2.0428401639386144e-06, + "loss": 0.7957, + "num_input_tokens_seen": 89983315, + "step": 4228 + }, + { + "epoch": 0.5085071845127157, + "flos": 52967649702600.0, + "grad_norm": 0.9012341964330768, + "learning_rate": 2.042061367785943e-06, + "loss": 0.6424, + "num_input_tokens_seen": 90036175, + "step": 4229 + }, + { + "epoch": 0.5086274274033548, + "flos": 26419231685640.0, + "grad_norm": 5.174064242028663, + "learning_rate": 2.041282565252594e-06, + "loss": 0.7333, + "num_input_tokens_seen": 90056060, + "step": 4230 + }, + { + "epoch": 0.5087476702939938, + "flos": 17214901690320.0, + "grad_norm": 3.428400376003887, + "learning_rate": 2.040503756456714e-06, + "loss": 0.7471, + "num_input_tokens_seen": 90074990, + "step": 4231 + }, + { + "epoch": 0.508867913184633, + "flos": 11004535490760.0, + "grad_norm": 7.01984671009645, + "learning_rate": 2.0397249415164456e-06, + "loss": 0.7848, + "num_input_tokens_seen": 90092065, + "step": 4232 + }, + { + "epoch": 0.508988156075272, + "flos": 18970027894080.0, + "grad_norm": 4.04853933862167, + "learning_rate": 2.0389461205499354e-06, + "loss": 0.7873, + "num_input_tokens_seen": 90110920, + "step": 4233 + }, + { + "epoch": 0.5091083989659111, + "flos": 10059262962840.0, + "grad_norm": 4.048556534472179, + "learning_rate": 2.03816729367533e-06, + "loss": 0.7151, + "num_input_tokens_seen": 90128795, + "step": 4234 + }, + { + "epoch": 0.5092286418565503, + "flos": 15432077512920.0, + "grad_norm": 5.35797065152129, + "learning_rate": 2.0373884610107765e-06, + "loss": 0.6841, + "num_input_tokens_seen": 90148830, + "step": 4235 + }, + { + "epoch": 0.5093488847471893, + "flos": 13593699088440.0, + "grad_norm": 6.211552888145328, + "learning_rate": 2.0366096226744225e-06, + "loss": 0.6712, + "num_input_tokens_seen": 90163745, + "step": 4236 + }, + { + "epoch": 0.5094691276378284, + "flos": 17429298668520.0, + "grad_norm": 3.627841230625853, + "learning_rate": 2.035830778784418e-06, + "loss": 0.746, + "num_input_tokens_seen": 90184140, + "step": 4237 + }, + { + "epoch": 0.5095893705284675, + "flos": 12705405505800.0, + "grad_norm": 12.868196041417443, + "learning_rate": 2.0350519294589134e-06, + "loss": 0.7927, + "num_input_tokens_seen": 90201900, + "step": 4238 + }, + { + "epoch": 0.5097096134191066, + "flos": 18943501338960.0, + "grad_norm": 3.7686059320093905, + "learning_rate": 2.0342730748160588e-06, + "loss": 0.8148, + "num_input_tokens_seen": 90222085, + "step": 4239 + }, + { + "epoch": 0.5098298563097456, + "flos": 20346278915640.0, + "grad_norm": 4.799788295527707, + "learning_rate": 2.033494214974006e-06, + "loss": 0.6684, + "num_input_tokens_seen": 90242950, + "step": 4240 + }, + { + "epoch": 0.5099500992003848, + "flos": 15619884616080.0, + "grad_norm": 3.784515232099375, + "learning_rate": 2.0327153500509067e-06, + "loss": 0.8176, + "num_input_tokens_seen": 90261695, + "step": 4241 + }, + { + "epoch": 0.5100703420910239, + "flos": 14515452757440.0, + "grad_norm": 3.19410365585835, + "learning_rate": 2.031936480164916e-06, + "loss": 0.8385, + "num_input_tokens_seen": 90279155, + "step": 4242 + }, + { + "epoch": 0.5101905849816629, + "flos": 18055492695960.0, + "grad_norm": 4.928376665867671, + "learning_rate": 2.0311576054341857e-06, + "loss": 0.7903, + "num_input_tokens_seen": 90299490, + "step": 4243 + }, + { + "epoch": 0.5103108278723021, + "flos": 16782973397880.0, + "grad_norm": 3.4565863888822674, + "learning_rate": 2.0303787259768715e-06, + "loss": 0.6172, + "num_input_tokens_seen": 90320110, + "step": 4244 + }, + { + "epoch": 0.5104310707629411, + "flos": 15729441772200.0, + "grad_norm": 5.484468203891074, + "learning_rate": 2.0295998419111294e-06, + "loss": 0.664, + "num_input_tokens_seen": 90337120, + "step": 4245 + }, + { + "epoch": 0.5105513136535802, + "flos": 10842526763640.0, + "grad_norm": 3.810033907401523, + "learning_rate": 2.028820953355115e-06, + "loss": 0.7117, + "num_input_tokens_seen": 90354940, + "step": 4246 + }, + { + "epoch": 0.5106715565442194, + "flos": 16540403546640.0, + "grad_norm": 12.59423948335359, + "learning_rate": 2.0280420604269834e-06, + "loss": 0.765, + "num_input_tokens_seen": 90374200, + "step": 4247 + }, + { + "epoch": 0.5107917994348584, + "flos": 52384888648920.0, + "grad_norm": 0.8298043192145382, + "learning_rate": 2.027263163244895e-06, + "loss": 0.6371, + "num_input_tokens_seen": 90443045, + "step": 4248 + }, + { + "epoch": 0.5109120423254975, + "flos": 18187144012800.0, + "grad_norm": 2.7912016713818275, + "learning_rate": 2.026484261927005e-06, + "loss": 0.7276, + "num_input_tokens_seen": 90462260, + "step": 4249 + }, + { + "epoch": 0.5110322852161366, + "flos": 15537297254520.0, + "grad_norm": 8.199370885766985, + "learning_rate": 2.025705356591475e-06, + "loss": 0.7187, + "num_input_tokens_seen": 90479670, + "step": 4250 + }, + { + "epoch": 0.5111525281067757, + "flos": 49001000344920.0, + "grad_norm": 0.8472812719733763, + "learning_rate": 2.024926447356462e-06, + "loss": 0.6141, + "num_input_tokens_seen": 90541675, + "step": 4251 + }, + { + "epoch": 0.5112727709974147, + "flos": 10815556969080.0, + "grad_norm": 6.69594774826974, + "learning_rate": 2.024147534340127e-06, + "loss": 0.7665, + "num_input_tokens_seen": 90559255, + "step": 4252 + }, + { + "epoch": 0.5113930138880539, + "flos": 15485858802240.0, + "grad_norm": 8.983784372569948, + "learning_rate": 2.02336861766063e-06, + "loss": 0.7746, + "num_input_tokens_seen": 90578035, + "step": 4253 + }, + { + "epoch": 0.511513256778693, + "flos": 14916802019880.0, + "grad_norm": 5.010612288348639, + "learning_rate": 2.0225896974361327e-06, + "loss": 0.7707, + "num_input_tokens_seen": 90597860, + "step": 4254 + }, + { + "epoch": 0.511633499669332, + "flos": 51533184997320.0, + "grad_norm": 0.9468216947148994, + "learning_rate": 2.0218107737847962e-06, + "loss": 0.6469, + "num_input_tokens_seen": 90659625, + "step": 4255 + }, + { + "epoch": 0.5117537425599712, + "flos": 17647241562240.0, + "grad_norm": 4.036250598443141, + "learning_rate": 2.0210318468247826e-06, + "loss": 0.748, + "num_input_tokens_seen": 90678315, + "step": 4256 + }, + { + "epoch": 0.5118739854506102, + "flos": 15324609914160.0, + "grad_norm": 3.95096610604059, + "learning_rate": 2.020252916674255e-06, + "loss": 0.8031, + "num_input_tokens_seen": 90697150, + "step": 4257 + }, + { + "epoch": 0.5119942283412493, + "flos": 12732786879840.0, + "grad_norm": 5.110127338233742, + "learning_rate": 2.019473983451375e-06, + "loss": 0.787, + "num_input_tokens_seen": 90715290, + "step": 4258 + }, + { + "epoch": 0.5121144712318885, + "flos": 15403619700240.0, + "grad_norm": 5.155619819741173, + "learning_rate": 2.0186950472743076e-06, + "loss": 0.6951, + "num_input_tokens_seen": 90734915, + "step": 4259 + }, + { + "epoch": 0.5122347141225275, + "flos": 14510862063240.0, + "grad_norm": 3.623731720145663, + "learning_rate": 2.0179161082612162e-06, + "loss": 0.7316, + "num_input_tokens_seen": 90754025, + "step": 4260 + }, + { + "epoch": 0.5123549570131666, + "flos": 16485735778440.0, + "grad_norm": 4.1229293712531385, + "learning_rate": 2.017137166530266e-06, + "loss": 0.6998, + "num_input_tokens_seen": 90773280, + "step": 4261 + }, + { + "epoch": 0.5124751999038056, + "flos": 14860361293920.0, + "grad_norm": 5.094845706598763, + "learning_rate": 2.0163582221996213e-06, + "loss": 0.7917, + "num_input_tokens_seen": 90791375, + "step": 4262 + }, + { + "epoch": 0.5125954427944448, + "flos": 29257392106440.0, + "grad_norm": 5.53403536391462, + "learning_rate": 2.015579275387446e-06, + "loss": 0.6665, + "num_input_tokens_seen": 90816415, + "step": 4263 + }, + { + "epoch": 0.5127156856850839, + "flos": 21427096935480.0, + "grad_norm": 14.537138108975029, + "learning_rate": 2.0148003262119085e-06, + "loss": 0.6708, + "num_input_tokens_seen": 90837105, + "step": 4264 + }, + { + "epoch": 0.5128359285757229, + "flos": 9843884525880.0, + "grad_norm": 5.313310449152877, + "learning_rate": 2.0140213747911728e-06, + "loss": 0.7372, + "num_input_tokens_seen": 90855360, + "step": 4265 + }, + { + "epoch": 0.5129561714663621, + "flos": 18457000258200.0, + "grad_norm": 6.547195171821283, + "learning_rate": 2.013242421243406e-06, + "loss": 0.7864, + "num_input_tokens_seen": 90873985, + "step": 4266 + }, + { + "epoch": 0.5130764143570011, + "flos": 13245529576080.0, + "grad_norm": 3.6265465955866247, + "learning_rate": 2.012463465686774e-06, + "loss": 0.7698, + "num_input_tokens_seen": 90893455, + "step": 4267 + }, + { + "epoch": 0.5131966572476402, + "flos": 44072588111520.0, + "grad_norm": 0.8424683479453302, + "learning_rate": 2.0116845082394446e-06, + "loss": 0.5921, + "num_input_tokens_seen": 90958010, + "step": 4268 + }, + { + "epoch": 0.5133169001382794, + "flos": 13515670761120.0, + "grad_norm": 4.474722758820992, + "learning_rate": 2.0109055490195836e-06, + "loss": 0.7678, + "num_input_tokens_seen": 90976185, + "step": 4269 + }, + { + "epoch": 0.5134371430289184, + "flos": 11301646470360.0, + "grad_norm": 4.283239640177228, + "learning_rate": 2.0101265881453605e-06, + "loss": 0.6276, + "num_input_tokens_seen": 90994380, + "step": 4270 + }, + { + "epoch": 0.5135573859195575, + "flos": 15675565503000.0, + "grad_norm": 5.929177323861462, + "learning_rate": 2.009347625734941e-06, + "loss": 0.7744, + "num_input_tokens_seen": 91014010, + "step": 4271 + }, + { + "epoch": 0.5136776288101966, + "flos": 12921733741560.0, + "grad_norm": 6.645237655810156, + "learning_rate": 2.0085686619064954e-06, + "loss": 0.7393, + "num_input_tokens_seen": 91030170, + "step": 4272 + }, + { + "epoch": 0.5137978717008357, + "flos": 12084751971360.0, + "grad_norm": 5.998422415282869, + "learning_rate": 2.00778969677819e-06, + "loss": 0.8174, + "num_input_tokens_seen": 91046925, + "step": 4273 + }, + { + "epoch": 0.5139181145914747, + "flos": 15109104837360.0, + "grad_norm": 4.456739384341585, + "learning_rate": 2.0070107304681934e-06, + "loss": 0.6307, + "num_input_tokens_seen": 91065600, + "step": 4274 + }, + { + "epoch": 0.5140383574821139, + "flos": 24181340276400.0, + "grad_norm": 4.476347327262163, + "learning_rate": 2.006231763094675e-06, + "loss": 0.7787, + "num_input_tokens_seen": 91086340, + "step": 4275 + }, + { + "epoch": 0.514158600372753, + "flos": 14271964767360.0, + "grad_norm": 6.696384410894195, + "learning_rate": 2.0054527947758027e-06, + "loss": 0.8694, + "num_input_tokens_seen": 91104860, + "step": 4276 + }, + { + "epoch": 0.514278843263392, + "flos": 46091143588800.0, + "grad_norm": 0.7764950437049852, + "learning_rate": 2.004673825629746e-06, + "loss": 0.584, + "num_input_tokens_seen": 91165360, + "step": 4277 + }, + { + "epoch": 0.5143990861540312, + "flos": 18967400117400.0, + "grad_norm": 7.120528842702336, + "learning_rate": 2.0038948557746744e-06, + "loss": 0.7154, + "num_input_tokens_seen": 91186935, + "step": 4278 + }, + { + "epoch": 0.5145193290446702, + "flos": 17079324538440.0, + "grad_norm": 6.176060129199069, + "learning_rate": 2.0031158853287558e-06, + "loss": 0.7258, + "num_input_tokens_seen": 91207090, + "step": 4279 + }, + { + "epoch": 0.5146395719353093, + "flos": 16728527249400.0, + "grad_norm": 10.759515853256799, + "learning_rate": 2.0023369144101593e-06, + "loss": 0.6982, + "num_input_tokens_seen": 91224980, + "step": 4280 + }, + { + "epoch": 0.5147598148259485, + "flos": 19344059102400.0, + "grad_norm": 4.53228053508991, + "learning_rate": 2.0015579431370555e-06, + "loss": 0.7466, + "num_input_tokens_seen": 91246380, + "step": 4281 + }, + { + "epoch": 0.5148800577165875, + "flos": 21988935246960.0, + "grad_norm": 5.764068972469428, + "learning_rate": 2.000778971627612e-06, + "loss": 0.6902, + "num_input_tokens_seen": 91265565, + "step": 4282 + }, + { + "epoch": 0.5150003006072266, + "flos": 13083584168880.0, + "grad_norm": 7.071305448662999, + "learning_rate": 2e-06, + "loss": 0.8856, + "num_input_tokens_seen": 91282880, + "step": 4283 + }, + { + "epoch": 0.5151205434978657, + "flos": 13488669306600.0, + "grad_norm": 9.085194341652024, + "learning_rate": 1.9992210283723878e-06, + "loss": 0.8412, + "num_input_tokens_seen": 91299840, + "step": 4284 + }, + { + "epoch": 0.5152407863885048, + "flos": 18567855472680.0, + "grad_norm": 2.476467768545066, + "learning_rate": 1.9984420568629448e-06, + "loss": 0.7689, + "num_input_tokens_seen": 91322325, + "step": 4285 + }, + { + "epoch": 0.5153610292791438, + "flos": 13377972391920.0, + "grad_norm": 7.040565389557764, + "learning_rate": 1.9976630855898405e-06, + "loss": 0.7663, + "num_input_tokens_seen": 91339800, + "step": 4286 + }, + { + "epoch": 0.515481272169783, + "flos": 22318619834040.0, + "grad_norm": 6.082184286770846, + "learning_rate": 1.9968841146712445e-06, + "loss": 0.7351, + "num_input_tokens_seen": 91359135, + "step": 4287 + }, + { + "epoch": 0.5156015150604221, + "flos": 16862014843920.0, + "grad_norm": 12.155093885817866, + "learning_rate": 1.996105144225326e-06, + "loss": 0.6969, + "num_input_tokens_seen": 91379090, + "step": 4288 + }, + { + "epoch": 0.5157217579510611, + "flos": 13029707899680.0, + "grad_norm": 10.380052921696814, + "learning_rate": 1.995326174370254e-06, + "loss": 0.7735, + "num_input_tokens_seen": 91397910, + "step": 4289 + }, + { + "epoch": 0.5158420008417003, + "flos": 14001285363000.0, + "grad_norm": 5.850436541579678, + "learning_rate": 1.994547205224197e-06, + "loss": 0.7101, + "num_input_tokens_seen": 91416535, + "step": 4290 + }, + { + "epoch": 0.5159622437323393, + "flos": 16405174654320.0, + "grad_norm": 5.103904291792254, + "learning_rate": 1.993768236905325e-06, + "loss": 0.663, + "num_input_tokens_seen": 91436925, + "step": 4291 + }, + { + "epoch": 0.5160824866229784, + "flos": 18020291311800.0, + "grad_norm": 7.136178332884518, + "learning_rate": 1.992989269531807e-06, + "loss": 0.6384, + "num_input_tokens_seen": 91455455, + "step": 4292 + }, + { + "epoch": 0.5162027295136175, + "flos": 13136162379720.0, + "grad_norm": 6.784713774612601, + "learning_rate": 1.99221030322181e-06, + "loss": 0.641, + "num_input_tokens_seen": 91471980, + "step": 4293 + }, + { + "epoch": 0.5163229724042566, + "flos": 20049991095000.0, + "grad_norm": 5.247365381296048, + "learning_rate": 1.991431338093505e-06, + "loss": 0.7943, + "num_input_tokens_seen": 91494055, + "step": 4294 + }, + { + "epoch": 0.5164432152948957, + "flos": 15919116813000.0, + "grad_norm": 6.736626834943528, + "learning_rate": 1.9906523742650587e-06, + "loss": 0.768, + "num_input_tokens_seen": 91512635, + "step": 4295 + }, + { + "epoch": 0.5165634581855347, + "flos": 18721189370760.0, + "grad_norm": 3.947829246391869, + "learning_rate": 1.9898734118546397e-06, + "loss": 0.7392, + "num_input_tokens_seen": 91532305, + "step": 4296 + }, + { + "epoch": 0.5166837010761739, + "flos": 14541314453400.0, + "grad_norm": 3.8052129676359696, + "learning_rate": 1.989094450980416e-06, + "loss": 0.7928, + "num_input_tokens_seen": 91552125, + "step": 4297 + }, + { + "epoch": 0.516803943966813, + "flos": 19532277785040.0, + "grad_norm": 6.1674461215804435, + "learning_rate": 1.9883154917605556e-06, + "loss": 0.7358, + "num_input_tokens_seen": 91571055, + "step": 4298 + }, + { + "epoch": 0.516924186857452, + "flos": 14379812285640.0, + "grad_norm": 6.214005850767998, + "learning_rate": 1.9875365343132262e-06, + "loss": 0.8038, + "num_input_tokens_seen": 91587895, + "step": 4299 + }, + { + "epoch": 0.5170444297480912, + "flos": 11621453149920.0, + "grad_norm": 20.43298850004494, + "learning_rate": 1.9867575787565946e-06, + "loss": 0.8286, + "num_input_tokens_seen": 91602275, + "step": 4300 + }, + { + "epoch": 0.5171646726387302, + "flos": 10303099212480.0, + "grad_norm": 11.411842459280171, + "learning_rate": 1.9859786252088275e-06, + "loss": 0.8442, + "num_input_tokens_seen": 91619175, + "step": 4301 + }, + { + "epoch": 0.5172849155293693, + "flos": 17261717788440.0, + "grad_norm": 7.264492151687125, + "learning_rate": 1.9851996737880914e-06, + "loss": 0.6518, + "num_input_tokens_seen": 91634080, + "step": 4302 + }, + { + "epoch": 0.5174051584200084, + "flos": 10384040256120.0, + "grad_norm": 7.159857094899643, + "learning_rate": 1.9844207246125537e-06, + "loss": 0.7338, + "num_input_tokens_seen": 91650380, + "step": 4303 + }, + { + "epoch": 0.5175254013106475, + "flos": 27853886350680.0, + "grad_norm": 6.490385721320686, + "learning_rate": 1.983641777800379e-06, + "loss": 0.6676, + "num_input_tokens_seen": 91672745, + "step": 4304 + }, + { + "epoch": 0.5176456442012866, + "flos": 50551512498120.0, + "grad_norm": 0.8719485754327245, + "learning_rate": 1.9828628334697343e-06, + "loss": 0.6199, + "num_input_tokens_seen": 91739675, + "step": 4305 + }, + { + "epoch": 0.5177658870919257, + "flos": 47244639402720.0, + "grad_norm": 0.8201226877680823, + "learning_rate": 1.982083891738784e-06, + "loss": 0.5857, + "num_input_tokens_seen": 91800265, + "step": 4306 + }, + { + "epoch": 0.5178861299825648, + "flos": 19536836819280.0, + "grad_norm": 6.502102731832429, + "learning_rate": 1.9813049527256923e-06, + "loss": 0.814, + "num_input_tokens_seen": 91820380, + "step": 4307 + }, + { + "epoch": 0.5180063728732038, + "flos": 13084027408320.0, + "grad_norm": 6.775617237760462, + "learning_rate": 1.9805260165486252e-06, + "loss": 0.7974, + "num_input_tokens_seen": 91839470, + "step": 4308 + }, + { + "epoch": 0.518126615763843, + "flos": 14245216592520.0, + "grad_norm": 4.893767593371652, + "learning_rate": 1.9797470833257457e-06, + "loss": 0.8475, + "num_input_tokens_seen": 91858890, + "step": 4309 + }, + { + "epoch": 0.5182468586544821, + "flos": 15135536412600.0, + "grad_norm": 7.596603297055338, + "learning_rate": 1.9789681531752177e-06, + "loss": 0.7579, + "num_input_tokens_seen": 91878830, + "step": 4310 + }, + { + "epoch": 0.5183671015451211, + "flos": 16917632410920.0, + "grad_norm": 4.94552443111989, + "learning_rate": 1.978189226215204e-06, + "loss": 0.7126, + "num_input_tokens_seen": 91899095, + "step": 4311 + }, + { + "epoch": 0.5184873444357603, + "flos": 12835347184800.0, + "grad_norm": 5.332798136242742, + "learning_rate": 1.9774103025638675e-06, + "loss": 0.7486, + "num_input_tokens_seen": 91916940, + "step": 4312 + }, + { + "epoch": 0.5186075873263993, + "flos": 18268085056440.0, + "grad_norm": 4.090017233068447, + "learning_rate": 1.9766313823393696e-06, + "loss": 0.7499, + "num_input_tokens_seen": 91937525, + "step": 4313 + }, + { + "epoch": 0.5187278302170384, + "flos": 11054169325320.0, + "grad_norm": 7.203871682082772, + "learning_rate": 1.975852465659873e-06, + "loss": 0.6655, + "num_input_tokens_seen": 91953225, + "step": 4314 + }, + { + "epoch": 0.5188480731076776, + "flos": 18321771365880.0, + "grad_norm": 7.665478185778784, + "learning_rate": 1.9750735526435377e-06, + "loss": 0.6834, + "num_input_tokens_seen": 91969890, + "step": 4315 + }, + { + "epoch": 0.5189683159983166, + "flos": 18159446039160.0, + "grad_norm": 7.080496485484407, + "learning_rate": 1.974294643408525e-06, + "loss": 0.7675, + "num_input_tokens_seen": 91987405, + "step": 4316 + }, + { + "epoch": 0.5190885588889557, + "flos": 18132982803960.0, + "grad_norm": 6.015524806389635, + "learning_rate": 1.9735157380729947e-06, + "loss": 0.6537, + "num_input_tokens_seen": 92007535, + "step": 4317 + }, + { + "epoch": 0.5192088017795948, + "flos": 18101453975160.0, + "grad_norm": 7.433615559692782, + "learning_rate": 1.9727368367551053e-06, + "loss": 0.8178, + "num_input_tokens_seen": 92025805, + "step": 4318 + }, + { + "epoch": 0.5193290446702339, + "flos": 19964079437640.0, + "grad_norm": 7.049127121585305, + "learning_rate": 1.9719579395730164e-06, + "loss": 0.6807, + "num_input_tokens_seen": 92044900, + "step": 4319 + }, + { + "epoch": 0.5194492875608729, + "flos": 8304200079000.0, + "grad_norm": 8.629042410650483, + "learning_rate": 1.9711790466448854e-06, + "loss": 0.9144, + "num_input_tokens_seen": 92058640, + "step": 4320 + }, + { + "epoch": 0.5195695304515121, + "flos": 14865236927760.0, + "grad_norm": 18.068688622723062, + "learning_rate": 1.9704001580888704e-06, + "loss": 0.6818, + "num_input_tokens_seen": 92077100, + "step": 4321 + }, + { + "epoch": 0.5196897733421512, + "flos": 14650365050160.0, + "grad_norm": 10.848008256914715, + "learning_rate": 1.9696212740231283e-06, + "loss": 0.8574, + "num_input_tokens_seen": 92095470, + "step": 4322 + }, + { + "epoch": 0.5198100162327902, + "flos": 17430818346600.0, + "grad_norm": 14.740156089339283, + "learning_rate": 1.9688423945658146e-06, + "loss": 0.798, + "num_input_tokens_seen": 92116055, + "step": 4323 + }, + { + "epoch": 0.5199302591234293, + "flos": 17592415494240.0, + "grad_norm": 9.577533548087738, + "learning_rate": 1.9680635198350845e-06, + "loss": 0.7037, + "num_input_tokens_seen": 92135485, + "step": 4324 + }, + { + "epoch": 0.5200505020140684, + "flos": 19320825183120.0, + "grad_norm": 4.802853316377432, + "learning_rate": 1.967284649949093e-06, + "loss": 0.707, + "num_input_tokens_seen": 92154415, + "step": 4325 + }, + { + "epoch": 0.5201707449047075, + "flos": 29123018033040.0, + "grad_norm": 6.531646862475342, + "learning_rate": 1.966505785025994e-06, + "loss": 0.717, + "num_input_tokens_seen": 92176040, + "step": 4326 + }, + { + "epoch": 0.5202909877953465, + "flos": 39541368703080.0, + "grad_norm": 6.310141547741941, + "learning_rate": 1.965726925183941e-06, + "loss": 0.7534, + "num_input_tokens_seen": 92198865, + "step": 4327 + }, + { + "epoch": 0.5204112306859857, + "flos": 14271901447440.0, + "grad_norm": 5.776098977421756, + "learning_rate": 1.964948070541087e-06, + "loss": 0.8239, + "num_input_tokens_seen": 92217245, + "step": 4328 + }, + { + "epoch": 0.5205314735766248, + "flos": 11139701063160.0, + "grad_norm": 6.311021990587089, + "learning_rate": 1.9641692212155816e-06, + "loss": 0.67, + "num_input_tokens_seen": 92234730, + "step": 4329 + }, + { + "epoch": 0.5206517164672638, + "flos": 43671419826120.0, + "grad_norm": 4.332839766539363, + "learning_rate": 1.9633903773255777e-06, + "loss": 0.7127, + "num_input_tokens_seen": 92256765, + "step": 4330 + }, + { + "epoch": 0.520771959357903, + "flos": 19699700365320.0, + "grad_norm": 4.208840493842966, + "learning_rate": 1.9626115389892237e-06, + "loss": 0.7312, + "num_input_tokens_seen": 92277275, + "step": 4331 + }, + { + "epoch": 0.520892202248542, + "flos": 19726226920440.0, + "grad_norm": 6.128200395539512, + "learning_rate": 1.96183270632467e-06, + "loss": 0.8276, + "num_input_tokens_seen": 92296845, + "step": 4332 + }, + { + "epoch": 0.5210124451391811, + "flos": 18943564658880.0, + "grad_norm": 4.487390441407212, + "learning_rate": 1.9610538794500644e-06, + "loss": 0.77, + "num_input_tokens_seen": 92316115, + "step": 4333 + }, + { + "epoch": 0.5211326880298203, + "flos": 52034281319640.0, + "grad_norm": 0.8013551469865703, + "learning_rate": 1.9602750584835542e-06, + "loss": 0.6256, + "num_input_tokens_seen": 92381770, + "step": 4334 + }, + { + "epoch": 0.5212529309204593, + "flos": 11379041598480.0, + "grad_norm": 11.247385017806119, + "learning_rate": 1.959496243543286e-06, + "loss": 0.8057, + "num_input_tokens_seen": 92399370, + "step": 4335 + }, + { + "epoch": 0.5213731738110984, + "flos": 19233425507640.0, + "grad_norm": 5.330367078039327, + "learning_rate": 1.9587174347474057e-06, + "loss": 0.777, + "num_input_tokens_seen": 92415600, + "step": 4336 + }, + { + "epoch": 0.5214934167017375, + "flos": 14182823794080.0, + "grad_norm": 4.235951710743638, + "learning_rate": 1.9579386322140574e-06, + "loss": 0.7895, + "num_input_tokens_seen": 92431000, + "step": 4337 + }, + { + "epoch": 0.5216136595923766, + "flos": 22722375253440.0, + "grad_norm": 7.654929035438386, + "learning_rate": 1.9571598360613854e-06, + "loss": 0.7905, + "num_input_tokens_seen": 92453595, + "step": 4338 + }, + { + "epoch": 0.5217339024830157, + "flos": 16053997445760.0, + "grad_norm": 4.253163226402195, + "learning_rate": 1.956381046407532e-06, + "loss": 0.6784, + "num_input_tokens_seen": 92473610, + "step": 4339 + }, + { + "epoch": 0.5218541453736548, + "flos": 15297735099480.0, + "grad_norm": 5.631188612252974, + "learning_rate": 1.9556022633706394e-06, + "loss": 0.8397, + "num_input_tokens_seen": 92492120, + "step": 4340 + }, + { + "epoch": 0.5219743882642939, + "flos": 17539014124440.0, + "grad_norm": 5.60159195342893, + "learning_rate": 1.954823487068848e-06, + "loss": 0.7805, + "num_input_tokens_seen": 92512050, + "step": 4341 + }, + { + "epoch": 0.5220946311549329, + "flos": 21125870161080.0, + "grad_norm": 7.885404948767989, + "learning_rate": 1.9540447176202976e-06, + "loss": 0.7896, + "num_input_tokens_seen": 92533015, + "step": 4342 + }, + { + "epoch": 0.5222148740455721, + "flos": 49543277292480.0, + "grad_norm": 0.8438393934013312, + "learning_rate": 1.9532659551431272e-06, + "loss": 0.6443, + "num_input_tokens_seen": 92599765, + "step": 4343 + }, + { + "epoch": 0.5223351169362112, + "flos": 45590296054800.0, + "grad_norm": 4.033562994689547, + "learning_rate": 1.9524871997554744e-06, + "loss": 0.6543, + "num_input_tokens_seen": 92627245, + "step": 4344 + }, + { + "epoch": 0.5224553598268502, + "flos": 10653168322440.0, + "grad_norm": 8.696332434707255, + "learning_rate": 1.951708451575475e-06, + "loss": 0.766, + "num_input_tokens_seen": 92644030, + "step": 4345 + }, + { + "epoch": 0.5225756027174894, + "flos": 10784851299240.0, + "grad_norm": 3.4145148564645447, + "learning_rate": 1.9509297107212657e-06, + "loss": 0.7916, + "num_input_tokens_seen": 92660520, + "step": 4346 + }, + { + "epoch": 0.5226958456081284, + "flos": 17213382012240.0, + "grad_norm": 6.41509815548308, + "learning_rate": 1.95015097731098e-06, + "loss": 0.781, + "num_input_tokens_seen": 92679730, + "step": 4347 + }, + { + "epoch": 0.5228160884987675, + "flos": 13921072498440.0, + "grad_norm": 6.709466242247544, + "learning_rate": 1.949372251462751e-06, + "loss": 0.8077, + "num_input_tokens_seen": 92696865, + "step": 4348 + }, + { + "epoch": 0.5229363313894067, + "flos": 15400517024160.0, + "grad_norm": 6.381462860293452, + "learning_rate": 1.9485935332947124e-06, + "loss": 0.8123, + "num_input_tokens_seen": 92714495, + "step": 4349 + }, + { + "epoch": 0.5230565742800457, + "flos": 10788682154400.0, + "grad_norm": 12.993072424575853, + "learning_rate": 1.947814822924993e-06, + "loss": 0.8118, + "num_input_tokens_seen": 92731725, + "step": 4350 + }, + { + "epoch": 0.5231768171706848, + "flos": 18916626524280.0, + "grad_norm": 5.84462335761534, + "learning_rate": 1.9470361204717236e-06, + "loss": 0.8086, + "num_input_tokens_seen": 92750585, + "step": 4351 + }, + { + "epoch": 0.5232970600613239, + "flos": 16644578509560.0, + "grad_norm": 5.247648529960266, + "learning_rate": 1.9462574260530326e-06, + "loss": 0.7948, + "num_input_tokens_seen": 92770585, + "step": 4352 + }, + { + "epoch": 0.523417302951963, + "flos": 12624179522520.0, + "grad_norm": 7.937849622821125, + "learning_rate": 1.9454787397870472e-06, + "loss": 0.7929, + "num_input_tokens_seen": 92787625, + "step": 4353 + }, + { + "epoch": 0.523537545842602, + "flos": 13542703875600.0, + "grad_norm": 13.458768823780023, + "learning_rate": 1.944700061791894e-06, + "loss": 0.7002, + "num_input_tokens_seen": 92805740, + "step": 4354 + }, + { + "epoch": 0.5236577887332411, + "flos": 14406750420240.0, + "grad_norm": 22.78917766511867, + "learning_rate": 1.943921392185698e-06, + "loss": 0.6465, + "num_input_tokens_seen": 92824085, + "step": 4355 + }, + { + "epoch": 0.5237780316238803, + "flos": 17241903144840.0, + "grad_norm": 12.635319184474712, + "learning_rate": 1.9431427310865814e-06, + "loss": 0.7581, + "num_input_tokens_seen": 92843410, + "step": 4356 + }, + { + "epoch": 0.5238982745145193, + "flos": 16458639344040.0, + "grad_norm": 5.602711920077187, + "learning_rate": 1.942364078612667e-06, + "loss": 0.7794, + "num_input_tokens_seen": 92861860, + "step": 4357 + }, + { + "epoch": 0.5240185174051584, + "flos": 19860727633680.0, + "grad_norm": 5.234642819057976, + "learning_rate": 1.9415854348820765e-06, + "loss": 0.7309, + "num_input_tokens_seen": 92881430, + "step": 4358 + }, + { + "epoch": 0.5241387602957975, + "flos": 16593963216240.0, + "grad_norm": 7.229825329708575, + "learning_rate": 1.940806800012929e-06, + "loss": 0.6528, + "num_input_tokens_seen": 92901220, + "step": 4359 + }, + { + "epoch": 0.5242590031864366, + "flos": 29825562409920.0, + "grad_norm": 5.694452752001318, + "learning_rate": 1.9400281741233432e-06, + "loss": 0.6255, + "num_input_tokens_seen": 92925830, + "step": 4360 + }, + { + "epoch": 0.5243792460770756, + "flos": 49163262351720.0, + "grad_norm": 0.687071276416869, + "learning_rate": 1.939249557331435e-06, + "loss": 0.5505, + "num_input_tokens_seen": 92991365, + "step": 4361 + }, + { + "epoch": 0.5244994889677148, + "flos": 20670929569080.0, + "grad_norm": 7.1663949012978385, + "learning_rate": 1.938470949755321e-06, + "loss": 0.709, + "num_input_tokens_seen": 93010965, + "step": 4362 + }, + { + "epoch": 0.5246197318583539, + "flos": 48626557557120.0, + "grad_norm": 0.8861096008099235, + "learning_rate": 1.937692351513115e-06, + "loss": 0.6149, + "num_input_tokens_seen": 93069680, + "step": 4363 + }, + { + "epoch": 0.5247399747489929, + "flos": 15379467642120.0, + "grad_norm": 4.8230518736813, + "learning_rate": 1.9369137627229297e-06, + "loss": 0.7866, + "num_input_tokens_seen": 93087800, + "step": 4364 + }, + { + "epoch": 0.5248602176396321, + "flos": 13893627804480.0, + "grad_norm": 5.309433714130547, + "learning_rate": 1.936135183502877e-06, + "loss": 0.8651, + "num_input_tokens_seen": 93104820, + "step": 4365 + }, + { + "epoch": 0.5249804605302711, + "flos": 16242691027800.0, + "grad_norm": 6.149578371559242, + "learning_rate": 1.935356613971066e-06, + "loss": 0.7864, + "num_input_tokens_seen": 93125200, + "step": 4366 + }, + { + "epoch": 0.5251007034209102, + "flos": 17430976646400.0, + "grad_norm": 19.170309411677557, + "learning_rate": 1.9345780542456047e-06, + "loss": 0.7589, + "num_input_tokens_seen": 93144295, + "step": 4367 + }, + { + "epoch": 0.5252209463115494, + "flos": 17052703003440.0, + "grad_norm": 4.553373431861849, + "learning_rate": 1.9337995044446007e-06, + "loss": 0.7043, + "num_input_tokens_seen": 93162855, + "step": 4368 + }, + { + "epoch": 0.5253411892021884, + "flos": 14487723123840.0, + "grad_norm": 11.082762012718959, + "learning_rate": 1.9330209646861596e-06, + "loss": 0.7859, + "num_input_tokens_seen": 93181725, + "step": 4369 + }, + { + "epoch": 0.5254614320928275, + "flos": 17673704797440.0, + "grad_norm": 4.7704255486446, + "learning_rate": 1.9322424350883843e-06, + "loss": 0.7672, + "num_input_tokens_seen": 93203280, + "step": 4370 + }, + { + "epoch": 0.5255816749834666, + "flos": 18052105080240.0, + "grad_norm": 8.611279174914742, + "learning_rate": 1.931463915769379e-06, + "loss": 0.77, + "num_input_tokens_seen": 93223115, + "step": 4371 + }, + { + "epoch": 0.5257019178741057, + "flos": 10275147959160.0, + "grad_norm": 7.381912095804552, + "learning_rate": 1.930685406847242e-06, + "loss": 0.7276, + "num_input_tokens_seen": 93237410, + "step": 4372 + }, + { + "epoch": 0.5258221607647448, + "flos": 17240795046240.0, + "grad_norm": 5.166344950514346, + "learning_rate": 1.9299069084400734e-06, + "loss": 0.8106, + "num_input_tokens_seen": 93257990, + "step": 4373 + }, + { + "epoch": 0.5259424036553839, + "flos": 18288754518960.0, + "grad_norm": 7.112416448923665, + "learning_rate": 1.9291284206659717e-06, + "loss": 0.673, + "num_input_tokens_seen": 93275895, + "step": 4374 + }, + { + "epoch": 0.526062646546023, + "flos": 21099850165320.0, + "grad_norm": 6.128791725348098, + "learning_rate": 1.928349943643032e-06, + "loss": 0.6968, + "num_input_tokens_seen": 93294715, + "step": 4375 + }, + { + "epoch": 0.526182889436662, + "flos": 16702633893480.0, + "grad_norm": 4.13107292038167, + "learning_rate": 1.9275714774893493e-06, + "loss": 0.8142, + "num_input_tokens_seen": 93313890, + "step": 4376 + }, + { + "epoch": 0.5263031323273012, + "flos": 16782435178560.0, + "grad_norm": 12.366710622320348, + "learning_rate": 1.9267930223230154e-06, + "loss": 0.7065, + "num_input_tokens_seen": 93332085, + "step": 4377 + }, + { + "epoch": 0.5264233752179402, + "flos": 12948291956640.0, + "grad_norm": 8.973127300434598, + "learning_rate": 1.9260145782621224e-06, + "loss": 0.7617, + "num_input_tokens_seen": 93349585, + "step": 4378 + }, + { + "epoch": 0.5265436181085793, + "flos": 17886202178040.0, + "grad_norm": 10.047591069689835, + "learning_rate": 1.925236145424758e-06, + "loss": 0.8641, + "num_input_tokens_seen": 93368125, + "step": 4379 + }, + { + "epoch": 0.5266638609992185, + "flos": 51038266858560.0, + "grad_norm": 0.702688738797331, + "learning_rate": 1.924457723929012e-06, + "loss": 0.5945, + "num_input_tokens_seen": 93438655, + "step": 4380 + }, + { + "epoch": 0.5267841038898575, + "flos": 15160891549200.0, + "grad_norm": 7.686883669081521, + "learning_rate": 1.9236793138929685e-06, + "loss": 0.8078, + "num_input_tokens_seen": 93457645, + "step": 4381 + }, + { + "epoch": 0.5269043467804966, + "flos": 12567232237200.0, + "grad_norm": 6.989774865275442, + "learning_rate": 1.9229009154347133e-06, + "loss": 0.7916, + "num_input_tokens_seen": 93474955, + "step": 4382 + }, + { + "epoch": 0.5270245896711357, + "flos": 13299500825160.0, + "grad_norm": 8.522780913213063, + "learning_rate": 1.922122528672327e-06, + "loss": 0.7814, + "num_input_tokens_seen": 93493340, + "step": 4383 + }, + { + "epoch": 0.5271448325617748, + "flos": 15567306405240.0, + "grad_norm": 7.944563478705894, + "learning_rate": 1.9213441537238914e-06, + "loss": 0.7612, + "num_input_tokens_seen": 93509935, + "step": 4384 + }, + { + "epoch": 0.5272650754524139, + "flos": 48287501621880.0, + "grad_norm": 0.9653200834747627, + "learning_rate": 1.920565790707485e-06, + "loss": 0.6458, + "num_input_tokens_seen": 93575045, + "step": 4385 + }, + { + "epoch": 0.527385318343053, + "flos": 14374145152800.0, + "grad_norm": 5.486650739224114, + "learning_rate": 1.9197874397411853e-06, + "loss": 0.6425, + "num_input_tokens_seen": 93591395, + "step": 4386 + }, + { + "epoch": 0.5275055612336921, + "flos": 9220634874720.0, + "grad_norm": 8.136098002005765, + "learning_rate": 1.919009100943067e-06, + "loss": 0.6101, + "num_input_tokens_seen": 93606805, + "step": 4387 + }, + { + "epoch": 0.5276258041243311, + "flos": 12948640216200.0, + "grad_norm": 5.911776747229384, + "learning_rate": 1.9182307744312043e-06, + "loss": 0.6387, + "num_input_tokens_seen": 93623630, + "step": 4388 + }, + { + "epoch": 0.5277460470149702, + "flos": 16620331471560.0, + "grad_norm": 5.812112118484298, + "learning_rate": 1.9174524603236676e-06, + "loss": 0.742, + "num_input_tokens_seen": 93642300, + "step": 4389 + }, + { + "epoch": 0.5278662899056094, + "flos": 14542644171720.0, + "grad_norm": 9.788629545119278, + "learning_rate": 1.916674158738527e-06, + "loss": 0.7517, + "num_input_tokens_seen": 93660925, + "step": 4390 + }, + { + "epoch": 0.5279865327962484, + "flos": 13138346916960.0, + "grad_norm": 6.3523576339628125, + "learning_rate": 1.9158958697938506e-06, + "loss": 0.5785, + "num_input_tokens_seen": 93679025, + "step": 4391 + }, + { + "epoch": 0.5281067756868875, + "flos": 11598282550560.0, + "grad_norm": 10.27879717012883, + "learning_rate": 1.9151175936077032e-06, + "loss": 0.8238, + "num_input_tokens_seen": 93693715, + "step": 4392 + }, + { + "epoch": 0.5282270185775266, + "flos": 14189947285080.0, + "grad_norm": 7.700948473490614, + "learning_rate": 1.9143393302981507e-06, + "loss": 0.7805, + "num_input_tokens_seen": 93711120, + "step": 4393 + }, + { + "epoch": 0.5283472614681657, + "flos": 11950061298360.0, + "grad_norm": 14.742816274846104, + "learning_rate": 1.913561079983252e-06, + "loss": 0.81, + "num_input_tokens_seen": 93729665, + "step": 4394 + }, + { + "epoch": 0.5284675043588047, + "flos": 19617682883040.0, + "grad_norm": 6.017699520541986, + "learning_rate": 1.9127828427810693e-06, + "loss": 0.7225, + "num_input_tokens_seen": 93749950, + "step": 4395 + }, + { + "epoch": 0.5285877472494439, + "flos": 14540174694840.0, + "grad_norm": 8.911368629157172, + "learning_rate": 1.9120046188096607e-06, + "loss": 0.7868, + "num_input_tokens_seen": 93767715, + "step": 4396 + }, + { + "epoch": 0.528707990140083, + "flos": 14622857036280.0, + "grad_norm": 13.293962780568208, + "learning_rate": 1.9112264081870804e-06, + "loss": 0.7358, + "num_input_tokens_seen": 93785825, + "step": 4397 + }, + { + "epoch": 0.528828233030722, + "flos": 15108851557680.0, + "grad_norm": 5.152579185769492, + "learning_rate": 1.9104482110313843e-06, + "loss": 0.7383, + "num_input_tokens_seen": 93805135, + "step": 4398 + }, + { + "epoch": 0.5289484759213612, + "flos": 18457316857800.0, + "grad_norm": 19.271227737377277, + "learning_rate": 1.909670027460623e-06, + "loss": 0.731, + "num_input_tokens_seen": 93822155, + "step": 4399 + }, + { + "epoch": 0.5290687188120002, + "flos": 22858300664880.0, + "grad_norm": 7.826834993041618, + "learning_rate": 1.908891857592847e-06, + "loss": 0.6975, + "num_input_tokens_seen": 93842945, + "step": 4400 + }, + { + "epoch": 0.5291889617026393, + "flos": 14703101560800.0, + "grad_norm": 7.68374566925309, + "learning_rate": 1.9081137015461034e-06, + "loss": 0.8764, + "num_input_tokens_seen": 93858740, + "step": 4401 + }, + { + "epoch": 0.5293092045932785, + "flos": 14350499654040.0, + "grad_norm": 4.83770690002372, + "learning_rate": 1.9073355594384383e-06, + "loss": 0.8842, + "num_input_tokens_seen": 93876700, + "step": 4402 + }, + { + "epoch": 0.5294294474839175, + "flos": 17810105108280.0, + "grad_norm": 6.532859786381872, + "learning_rate": 1.906557431387895e-06, + "loss": 0.7895, + "num_input_tokens_seen": 93895410, + "step": 4403 + }, + { + "epoch": 0.5295496903745566, + "flos": 13782044410920.0, + "grad_norm": 5.463571418072693, + "learning_rate": 1.905779317512516e-06, + "loss": 0.7737, + "num_input_tokens_seen": 93912675, + "step": 4404 + }, + { + "epoch": 0.5296699332651957, + "flos": 15297640119600.0, + "grad_norm": 4.973413929597284, + "learning_rate": 1.9050012179303385e-06, + "loss": 0.7811, + "num_input_tokens_seen": 93930905, + "step": 4405 + }, + { + "epoch": 0.5297901761558348, + "flos": 16128163257960.0, + "grad_norm": 9.053106662045632, + "learning_rate": 1.904223132759401e-06, + "loss": 0.6567, + "num_input_tokens_seen": 93949225, + "step": 4406 + }, + { + "epoch": 0.5299104190464738, + "flos": 15945453408360.0, + "grad_norm": 9.568779373373024, + "learning_rate": 1.9034450621177383e-06, + "loss": 0.6564, + "num_input_tokens_seen": 93967265, + "step": 4407 + }, + { + "epoch": 0.530030661937113, + "flos": 10707551151000.0, + "grad_norm": 6.454686010688826, + "learning_rate": 1.9026670061233824e-06, + "loss": 0.6879, + "num_input_tokens_seen": 93984420, + "step": 4408 + }, + { + "epoch": 0.5301509048277521, + "flos": 15540684870240.0, + "grad_norm": 4.195441842527235, + "learning_rate": 1.901888964894365e-06, + "loss": 0.7904, + "num_input_tokens_seen": 94003180, + "step": 4409 + }, + { + "epoch": 0.5302711477183911, + "flos": 19022954364480.0, + "grad_norm": 6.180372969350896, + "learning_rate": 1.9011109385487134e-06, + "loss": 0.6575, + "num_input_tokens_seen": 94024150, + "step": 4410 + }, + { + "epoch": 0.5303913906090303, + "flos": 16296757256760.0, + "grad_norm": 10.522209656584918, + "learning_rate": 1.900332927204454e-06, + "loss": 0.6346, + "num_input_tokens_seen": 94042320, + "step": 4411 + }, + { + "epoch": 0.5305116334996693, + "flos": 18267388537320.0, + "grad_norm": 9.959937143395566, + "learning_rate": 1.8995549309796097e-06, + "loss": 0.7485, + "num_input_tokens_seen": 94061345, + "step": 4412 + }, + { + "epoch": 0.5306318763903084, + "flos": 14754033453720.0, + "grad_norm": 6.4881392617854825, + "learning_rate": 1.8987769499922028e-06, + "loss": 0.7435, + "num_input_tokens_seen": 94080035, + "step": 4413 + }, + { + "epoch": 0.5307521192809476, + "flos": 14811613938240.0, + "grad_norm": 4.718486652225631, + "learning_rate": 1.897998984360252e-06, + "loss": 0.6827, + "num_input_tokens_seen": 94098725, + "step": 4414 + }, + { + "epoch": 0.5308723621715866, + "flos": 21160375026120.0, + "grad_norm": 2.7208960821404276, + "learning_rate": 1.897221034201775e-06, + "loss": 0.764, + "num_input_tokens_seen": 94122185, + "step": 4415 + }, + { + "epoch": 0.5309926050622257, + "flos": 20134256434440.0, + "grad_norm": 26.528695171087264, + "learning_rate": 1.8964430996347842e-06, + "loss": 0.656, + "num_input_tokens_seen": 94143455, + "step": 4416 + }, + { + "epoch": 0.5311128479528648, + "flos": 14994260467920.0, + "grad_norm": 5.184938096915686, + "learning_rate": 1.8956651807772931e-06, + "loss": 0.8169, + "num_input_tokens_seen": 94161210, + "step": 4417 + }, + { + "epoch": 0.5312330908435039, + "flos": 15648880648080.0, + "grad_norm": 4.384755185121202, + "learning_rate": 1.8948872777473115e-06, + "loss": 0.8143, + "num_input_tokens_seen": 94178885, + "step": 4418 + }, + { + "epoch": 0.531353333734143, + "flos": 18105601429920.0, + "grad_norm": 7.734486958627805, + "learning_rate": 1.8941093906628458e-06, + "loss": 0.6252, + "num_input_tokens_seen": 94196390, + "step": 4419 + }, + { + "epoch": 0.531473576624782, + "flos": 22370659825560.0, + "grad_norm": 6.2484214315090965, + "learning_rate": 1.893331519641902e-06, + "loss": 0.697, + "num_input_tokens_seen": 94218255, + "step": 4420 + }, + { + "epoch": 0.5315938195154212, + "flos": 16836912987000.0, + "grad_norm": 10.532278672868884, + "learning_rate": 1.8925536648024815e-06, + "loss": 0.7339, + "num_input_tokens_seen": 94235395, + "step": 4421 + }, + { + "epoch": 0.5317140624060602, + "flos": 16648789284240.0, + "grad_norm": 4.167247935122556, + "learning_rate": 1.8917758262625849e-06, + "loss": 0.7513, + "num_input_tokens_seen": 94255355, + "step": 4422 + }, + { + "epoch": 0.5318343052966993, + "flos": 16702380613800.0, + "grad_norm": 6.052624396487935, + "learning_rate": 1.8909980041402089e-06, + "loss": 0.7939, + "num_input_tokens_seen": 94273670, + "step": 4423 + }, + { + "epoch": 0.5319545481873384, + "flos": 9897380875560.0, + "grad_norm": 9.757444620642351, + "learning_rate": 1.8902201985533494e-06, + "loss": 0.6461, + "num_input_tokens_seen": 94290655, + "step": 4424 + }, + { + "epoch": 0.5320747910779775, + "flos": 16214739774480.0, + "grad_norm": 15.375667898666325, + "learning_rate": 1.8894424096199983e-06, + "loss": 0.7342, + "num_input_tokens_seen": 94309580, + "step": 4425 + }, + { + "epoch": 0.5321950339686166, + "flos": 13569578690280.0, + "grad_norm": 7.867996977269766, + "learning_rate": 1.8886646374581463e-06, + "loss": 0.8421, + "num_input_tokens_seen": 94328525, + "step": 4426 + }, + { + "epoch": 0.5323152768592557, + "flos": 16728717209160.0, + "grad_norm": 3.760751063025102, + "learning_rate": 1.8878868821857795e-06, + "loss": 0.6905, + "num_input_tokens_seen": 94347895, + "step": 4427 + }, + { + "epoch": 0.5324355197498948, + "flos": 24937475982840.0, + "grad_norm": 8.362069695639146, + "learning_rate": 1.8871091439208838e-06, + "loss": 0.7227, + "num_input_tokens_seen": 94369225, + "step": 4428 + }, + { + "epoch": 0.5325557626405338, + "flos": 17024466810480.0, + "grad_norm": 5.2288847868835235, + "learning_rate": 1.8863314227814414e-06, + "loss": 0.7429, + "num_input_tokens_seen": 94387255, + "step": 4429 + }, + { + "epoch": 0.532676005531173, + "flos": 19585964094480.0, + "grad_norm": 6.870140994206264, + "learning_rate": 1.8855537188854313e-06, + "loss": 0.4591, + "num_input_tokens_seen": 94405950, + "step": 4430 + }, + { + "epoch": 0.5327962484218121, + "flos": 13056329434680.0, + "grad_norm": 5.562793509636986, + "learning_rate": 1.8847760323508315e-06, + "loss": 0.7624, + "num_input_tokens_seen": 94424575, + "step": 4431 + }, + { + "epoch": 0.5329164913124511, + "flos": 13078296955560.0, + "grad_norm": 11.483392416511837, + "learning_rate": 1.883998363295616e-06, + "loss": 0.7364, + "num_input_tokens_seen": 94441775, + "step": 4432 + }, + { + "epoch": 0.5330367342030903, + "flos": 42189003755520.0, + "grad_norm": 0.963943648470018, + "learning_rate": 1.8832207118377565e-06, + "loss": 0.679, + "num_input_tokens_seen": 94496865, + "step": 4433 + }, + { + "epoch": 0.5331569770937293, + "flos": 12705500485680.0, + "grad_norm": 3.720634944421239, + "learning_rate": 1.882443078095222e-06, + "loss": 0.6693, + "num_input_tokens_seen": 94515465, + "step": 4434 + }, + { + "epoch": 0.5332772199843684, + "flos": 41817505344000.0, + "grad_norm": 0.8629615847766527, + "learning_rate": 1.8816654621859794e-06, + "loss": 0.6813, + "num_input_tokens_seen": 94574850, + "step": 4435 + }, + { + "epoch": 0.5333974628750076, + "flos": 13650773013600.0, + "grad_norm": 4.882489833088733, + "learning_rate": 1.8808878642279915e-06, + "loss": 0.6943, + "num_input_tokens_seen": 94589975, + "step": 4436 + }, + { + "epoch": 0.5335177057656466, + "flos": 17430818346600.0, + "grad_norm": 7.6964468647578546, + "learning_rate": 1.8801102843392209e-06, + "loss": 0.6413, + "num_input_tokens_seen": 94609100, + "step": 4437 + }, + { + "epoch": 0.5336379486562857, + "flos": 18372513299040.0, + "grad_norm": 3.848569748000497, + "learning_rate": 1.8793327226376238e-06, + "loss": 0.8448, + "num_input_tokens_seen": 94628140, + "step": 4438 + }, + { + "epoch": 0.5337581915469248, + "flos": 15648025829160.0, + "grad_norm": 15.644679989149367, + "learning_rate": 1.8785551792411569e-06, + "loss": 0.7824, + "num_input_tokens_seen": 94646870, + "step": 4439 + }, + { + "epoch": 0.5338784344375639, + "flos": 10814828790000.0, + "grad_norm": 11.260048003018168, + "learning_rate": 1.8777776542677733e-06, + "loss": 0.809, + "num_input_tokens_seen": 94664640, + "step": 4440 + }, + { + "epoch": 0.5339986773282029, + "flos": 15216509116200.0, + "grad_norm": 5.5760699421506414, + "learning_rate": 1.8770001478354216e-06, + "loss": 0.6961, + "num_input_tokens_seen": 94684035, + "step": 4441 + }, + { + "epoch": 0.5341189202188421, + "flos": 13111662062040.0, + "grad_norm": 5.458746780384076, + "learning_rate": 1.8762226600620504e-06, + "loss": 0.8099, + "num_input_tokens_seen": 94702370, + "step": 4442 + }, + { + "epoch": 0.5342391631094812, + "flos": 7976953308840.0, + "grad_norm": 9.173791324508368, + "learning_rate": 1.8754451910656031e-06, + "loss": 0.582, + "num_input_tokens_seen": 94715990, + "step": 4443 + }, + { + "epoch": 0.5343594060001202, + "flos": 11164897899960.0, + "grad_norm": 7.7446590601213385, + "learning_rate": 1.8746677409640212e-06, + "loss": 0.7981, + "num_input_tokens_seen": 94732810, + "step": 4444 + }, + { + "epoch": 0.5344796488907594, + "flos": 19720781407320.0, + "grad_norm": 5.655415678928017, + "learning_rate": 1.8738903098752432e-06, + "loss": 0.8263, + "num_input_tokens_seen": 94751660, + "step": 4445 + }, + { + "epoch": 0.5345998917813984, + "flos": 18619198945080.0, + "grad_norm": 3.667891788995587, + "learning_rate": 1.8731128979172052e-06, + "loss": 0.712, + "num_input_tokens_seen": 94770580, + "step": 4446 + }, + { + "epoch": 0.5347201346720375, + "flos": 24126767488080.0, + "grad_norm": 10.026349257998117, + "learning_rate": 1.8723355052078394e-06, + "loss": 0.6432, + "num_input_tokens_seen": 94790335, + "step": 4447 + }, + { + "epoch": 0.5348403775626767, + "flos": 13110648943320.0, + "grad_norm": 3.9324003831937913, + "learning_rate": 1.8715581318650765e-06, + "loss": 0.7525, + "num_input_tokens_seen": 94809110, + "step": 4448 + }, + { + "epoch": 0.5349606204533157, + "flos": 12840856017840.0, + "grad_norm": 10.248517185046211, + "learning_rate": 1.8707807780068422e-06, + "loss": 0.792, + "num_input_tokens_seen": 94826645, + "step": 4449 + }, + { + "epoch": 0.5350808633439548, + "flos": 21399145682160.0, + "grad_norm": 42.75393462339168, + "learning_rate": 1.8700034437510611e-06, + "loss": 0.6511, + "num_input_tokens_seen": 94846460, + "step": 4450 + }, + { + "epoch": 0.5352011062345938, + "flos": 14244425093520.0, + "grad_norm": 9.032836352351312, + "learning_rate": 1.8692261292156549e-06, + "loss": 0.7916, + "num_input_tokens_seen": 94865415, + "step": 4451 + }, + { + "epoch": 0.535321349125233, + "flos": 17188248495360.0, + "grad_norm": 5.119119759829664, + "learning_rate": 1.8684488345185401e-06, + "loss": 0.7981, + "num_input_tokens_seen": 94885310, + "step": 4452 + }, + { + "epoch": 0.535441592015872, + "flos": 14967259013400.0, + "grad_norm": 13.164134945089977, + "learning_rate": 1.8676715597776332e-06, + "loss": 0.7647, + "num_input_tokens_seen": 94903375, + "step": 4453 + }, + { + "epoch": 0.5355618349065111, + "flos": 14298586302360.0, + "grad_norm": 6.335891441243656, + "learning_rate": 1.8668943051108455e-06, + "loss": 0.7446, + "num_input_tokens_seen": 94920400, + "step": 4454 + }, + { + "epoch": 0.5356820777971503, + "flos": 17593017033480.0, + "grad_norm": 12.898839712081422, + "learning_rate": 1.8661170706360856e-06, + "loss": 0.7413, + "num_input_tokens_seen": 94939285, + "step": 4455 + }, + { + "epoch": 0.5358023206877893, + "flos": 15268929027240.0, + "grad_norm": 4.51164666981112, + "learning_rate": 1.8653398564712594e-06, + "loss": 0.8099, + "num_input_tokens_seen": 94957950, + "step": 4456 + }, + { + "epoch": 0.5359225635784284, + "flos": 16404699754920.0, + "grad_norm": 3.4562362532661326, + "learning_rate": 1.8645626627342704e-06, + "loss": 0.8059, + "num_input_tokens_seen": 94978435, + "step": 4457 + }, + { + "epoch": 0.5360428064690675, + "flos": 17647178242320.0, + "grad_norm": 7.884368410232457, + "learning_rate": 1.8637854895430172e-06, + "loss": 0.7804, + "num_input_tokens_seen": 94997420, + "step": 4458 + }, + { + "epoch": 0.5361630493597066, + "flos": 15676198702200.0, + "grad_norm": 4.986900144370925, + "learning_rate": 1.8630083370153978e-06, + "loss": 0.6684, + "num_input_tokens_seen": 95016780, + "step": 4459 + }, + { + "epoch": 0.5362832922503457, + "flos": 50798700212160.0, + "grad_norm": 0.7938632523849707, + "learning_rate": 1.8622312052693041e-06, + "loss": 0.5823, + "num_input_tokens_seen": 95077680, + "step": 4460 + }, + { + "epoch": 0.5364035351409848, + "flos": 7060961752560.0, + "grad_norm": 7.436267398025801, + "learning_rate": 1.8614540944226267e-06, + "loss": 0.715, + "num_input_tokens_seen": 95094070, + "step": 4461 + }, + { + "epoch": 0.5365237780316239, + "flos": 17048713848480.0, + "grad_norm": 5.692812272705435, + "learning_rate": 1.8606770045932537e-06, + "loss": 0.6721, + "num_input_tokens_seen": 95112905, + "step": 4462 + }, + { + "epoch": 0.5366440209222629, + "flos": 19482707270400.0, + "grad_norm": 5.276860506534157, + "learning_rate": 1.859899935899068e-06, + "loss": 0.7972, + "num_input_tokens_seen": 95132480, + "step": 4463 + }, + { + "epoch": 0.5367642638129021, + "flos": 14324194718640.0, + "grad_norm": 4.089240465966874, + "learning_rate": 1.8591228884579506e-06, + "loss": 0.7828, + "num_input_tokens_seen": 95150695, + "step": 4464 + }, + { + "epoch": 0.5368845067035412, + "flos": 17512044329880.0, + "grad_norm": 17.324675427323577, + "learning_rate": 1.8583458623877795e-06, + "loss": 0.8103, + "num_input_tokens_seen": 95169515, + "step": 4465 + }, + { + "epoch": 0.5370047495941802, + "flos": 12300383688000.0, + "grad_norm": 7.8184245114230215, + "learning_rate": 1.8575688578064281e-06, + "loss": 0.7155, + "num_input_tokens_seen": 95187360, + "step": 4466 + }, + { + "epoch": 0.5371249924848194, + "flos": 15162854466720.0, + "grad_norm": 3.6266462599890787, + "learning_rate": 1.8567918748317674e-06, + "loss": 0.7384, + "num_input_tokens_seen": 95206430, + "step": 4467 + }, + { + "epoch": 0.5372452353754584, + "flos": 13110807243120.0, + "grad_norm": 4.420236236680289, + "learning_rate": 1.8560149135816659e-06, + "loss": 0.8062, + "num_input_tokens_seen": 95222985, + "step": 4468 + }, + { + "epoch": 0.5373654782660975, + "flos": 11193134092920.0, + "grad_norm": 6.932845195068133, + "learning_rate": 1.8552379741739873e-06, + "loss": 0.8143, + "num_input_tokens_seen": 95240050, + "step": 4469 + }, + { + "epoch": 0.5374857211567367, + "flos": 50881414213560.0, + "grad_norm": 0.9155028968090185, + "learning_rate": 1.8544610567265935e-06, + "loss": 0.5741, + "num_input_tokens_seen": 95293710, + "step": 4470 + }, + { + "epoch": 0.5376059640473757, + "flos": 10977312416520.0, + "grad_norm": 4.055957867134556, + "learning_rate": 1.853684161357341e-06, + "loss": 0.8171, + "num_input_tokens_seen": 95311090, + "step": 4471 + }, + { + "epoch": 0.5377262069380148, + "flos": 14461069928880.0, + "grad_norm": 5.756757994197756, + "learning_rate": 1.852907288184085e-06, + "loss": 0.766, + "num_input_tokens_seen": 95329695, + "step": 4472 + }, + { + "epoch": 0.5378464498286539, + "flos": 22016728200480.0, + "grad_norm": 5.276098503969006, + "learning_rate": 1.8521304373246762e-06, + "loss": 0.6837, + "num_input_tokens_seen": 95350460, + "step": 4473 + }, + { + "epoch": 0.537966692719293, + "flos": 15540431590560.0, + "grad_norm": 6.869145623043391, + "learning_rate": 1.8513536088969626e-06, + "loss": 0.8679, + "num_input_tokens_seen": 95367845, + "step": 4474 + }, + { + "epoch": 0.538086935609932, + "flos": 15756506546640.0, + "grad_norm": 4.485698012998463, + "learning_rate": 1.8505768030187884e-06, + "loss": 0.7818, + "num_input_tokens_seen": 95387695, + "step": 4475 + }, + { + "epoch": 0.5382071785005712, + "flos": 16647586205760.0, + "grad_norm": 8.007347071077863, + "learning_rate": 1.849800019807995e-06, + "loss": 0.79, + "num_input_tokens_seen": 95408640, + "step": 4476 + }, + { + "epoch": 0.5383274213912103, + "flos": 18266122138920.0, + "grad_norm": 4.975594083500447, + "learning_rate": 1.8490232593824186e-06, + "loss": 0.7018, + "num_input_tokens_seen": 95424815, + "step": 4477 + }, + { + "epoch": 0.5384476642818493, + "flos": 16324265270640.0, + "grad_norm": 3.2926097597264987, + "learning_rate": 1.8482465218598935e-06, + "loss": 0.8279, + "num_input_tokens_seen": 95444480, + "step": 4478 + }, + { + "epoch": 0.5385679071724885, + "flos": 16620838030920.0, + "grad_norm": 4.546535833234546, + "learning_rate": 1.8474698073582508e-06, + "loss": 0.8211, + "num_input_tokens_seen": 95465570, + "step": 4479 + }, + { + "epoch": 0.5386881500631275, + "flos": 11618952013080.0, + "grad_norm": 4.622310371862552, + "learning_rate": 1.8466931159953166e-06, + "loss": 0.8453, + "num_input_tokens_seen": 95481925, + "step": 4480 + }, + { + "epoch": 0.5388083929537666, + "flos": 17619068689200.0, + "grad_norm": 6.764527123862588, + "learning_rate": 1.8459164478889158e-06, + "loss": 0.8275, + "num_input_tokens_seen": 95503040, + "step": 4481 + }, + { + "epoch": 0.5389286358444056, + "flos": 16269977421960.0, + "grad_norm": 3.2426061888448707, + "learning_rate": 1.8451398031568663e-06, + "loss": 0.7369, + "num_input_tokens_seen": 95522385, + "step": 4482 + }, + { + "epoch": 0.5390488787350448, + "flos": 17783515233240.0, + "grad_norm": 5.212320353021715, + "learning_rate": 1.844363181916986e-06, + "loss": 0.7321, + "num_input_tokens_seen": 95542830, + "step": 4483 + }, + { + "epoch": 0.5391691216256839, + "flos": 12084498691680.0, + "grad_norm": 7.086976703789058, + "learning_rate": 1.8435865842870868e-06, + "loss": 0.8151, + "num_input_tokens_seen": 95560490, + "step": 4484 + }, + { + "epoch": 0.5392893645163229, + "flos": 17078786319120.0, + "grad_norm": 9.668661591335868, + "learning_rate": 1.8428100103849787e-06, + "loss": 0.7001, + "num_input_tokens_seen": 95580005, + "step": 4485 + }, + { + "epoch": 0.5394096074069621, + "flos": 11409937228080.0, + "grad_norm": 5.927106574559918, + "learning_rate": 1.842033460328467e-06, + "loss": 0.6932, + "num_input_tokens_seen": 95598445, + "step": 4486 + }, + { + "epoch": 0.5395298502976011, + "flos": 16755845303520.0, + "grad_norm": 3.4371172605194946, + "learning_rate": 1.8412569342353541e-06, + "loss": 0.7366, + "num_input_tokens_seen": 95618320, + "step": 4487 + }, + { + "epoch": 0.5396500931882402, + "flos": 17457946440960.0, + "grad_norm": 10.75211153805872, + "learning_rate": 1.840480432223438e-06, + "loss": 0.8287, + "num_input_tokens_seen": 95637045, + "step": 4488 + }, + { + "epoch": 0.5397703360788794, + "flos": 19293918708480.0, + "grad_norm": 8.908046608126755, + "learning_rate": 1.8397039544105131e-06, + "loss": 0.7473, + "num_input_tokens_seen": 95655850, + "step": 4489 + }, + { + "epoch": 0.5398905789695184, + "flos": 15513303496200.0, + "grad_norm": 4.398425574820091, + "learning_rate": 1.8389275009143711e-06, + "loss": 0.685, + "num_input_tokens_seen": 95675310, + "step": 4490 + }, + { + "epoch": 0.5400108218601575, + "flos": 18373969657200.0, + "grad_norm": 4.821287183353644, + "learning_rate": 1.8381510718527988e-06, + "loss": 0.7105, + "num_input_tokens_seen": 95694640, + "step": 4491 + }, + { + "epoch": 0.5401310647507966, + "flos": 19319178865200.0, + "grad_norm": 4.829295587450791, + "learning_rate": 1.8373746673435812e-06, + "loss": 0.624, + "num_input_tokens_seen": 95715385, + "step": 4492 + }, + { + "epoch": 0.5402513076414357, + "flos": 20427948138360.0, + "grad_norm": 10.226715436451924, + "learning_rate": 1.8365982875044964e-06, + "loss": 0.7738, + "num_input_tokens_seen": 95735415, + "step": 4493 + }, + { + "epoch": 0.5403715505320748, + "flos": 16756130243160.0, + "grad_norm": 12.36377985763917, + "learning_rate": 1.8358219324533217e-06, + "loss": 0.734, + "num_input_tokens_seen": 95755400, + "step": 4494 + }, + { + "epoch": 0.5404917934227139, + "flos": 22180636525200.0, + "grad_norm": 6.721416281455366, + "learning_rate": 1.8350456023078292e-06, + "loss": 0.6825, + "num_input_tokens_seen": 95777495, + "step": 4495 + }, + { + "epoch": 0.540612036313353, + "flos": 14569044087000.0, + "grad_norm": 6.095939673197855, + "learning_rate": 1.8342692971857874e-06, + "loss": 0.7443, + "num_input_tokens_seen": 95796415, + "step": 4496 + }, + { + "epoch": 0.540732279203992, + "flos": 17781647295600.0, + "grad_norm": 11.835007311839274, + "learning_rate": 1.833493017204962e-06, + "loss": 0.6869, + "num_input_tokens_seen": 95816240, + "step": 4497 + }, + { + "epoch": 0.5408525220946312, + "flos": 14757452729400.0, + "grad_norm": 8.690638572136336, + "learning_rate": 1.8327167624831134e-06, + "loss": 0.7509, + "num_input_tokens_seen": 95833690, + "step": 4498 + }, + { + "epoch": 0.5409727649852702, + "flos": 17674337996640.0, + "grad_norm": 9.682614774410315, + "learning_rate": 1.831940533137999e-06, + "loss": 0.6908, + "num_input_tokens_seen": 95852315, + "step": 4499 + }, + { + "epoch": 0.5410930078759093, + "flos": 17509828132680.0, + "grad_norm": 3.9803911225870166, + "learning_rate": 1.8311643292873718e-06, + "loss": 0.7, + "num_input_tokens_seen": 95870855, + "step": 4500 + }, + { + "epoch": 0.5412132507665485, + "flos": 15432299132640.0, + "grad_norm": 4.792060946400168, + "learning_rate": 1.8303881510489818e-06, + "loss": 0.8698, + "num_input_tokens_seen": 95888965, + "step": 4501 + }, + { + "epoch": 0.5413334936571875, + "flos": 22183454261640.0, + "grad_norm": 4.408756922734544, + "learning_rate": 1.829611998540574e-06, + "loss": 0.667, + "num_input_tokens_seen": 95909890, + "step": 4502 + }, + { + "epoch": 0.5414537365478266, + "flos": 17781678955560.0, + "grad_norm": 16.5231844691028, + "learning_rate": 1.8288358718798914e-06, + "loss": 0.7872, + "num_input_tokens_seen": 95928800, + "step": 4503 + }, + { + "epoch": 0.5415739794384657, + "flos": 12138659900520.0, + "grad_norm": 4.307121793202451, + "learning_rate": 1.8280597711846703e-06, + "loss": 0.7086, + "num_input_tokens_seen": 95946760, + "step": 4504 + }, + { + "epoch": 0.5416942223291048, + "flos": 16972205199240.0, + "grad_norm": 9.140617487977682, + "learning_rate": 1.8272836965726455e-06, + "loss": 0.8147, + "num_input_tokens_seen": 95965415, + "step": 4505 + }, + { + "epoch": 0.5418144652197439, + "flos": 14838963652320.0, + "grad_norm": 7.899329251171153, + "learning_rate": 1.8265076481615461e-06, + "loss": 0.7605, + "num_input_tokens_seen": 95985050, + "step": 4506 + }, + { + "epoch": 0.541934708110383, + "flos": 9033397650840.0, + "grad_norm": 38.636786480390036, + "learning_rate": 1.8257316260690987e-06, + "loss": 0.8517, + "num_input_tokens_seen": 96002555, + "step": 4507 + }, + { + "epoch": 0.5420549510010221, + "flos": 15542711107680.0, + "grad_norm": 7.062192927073642, + "learning_rate": 1.8249556304130254e-06, + "loss": 0.7521, + "num_input_tokens_seen": 96023555, + "step": 4508 + }, + { + "epoch": 0.5421751938916611, + "flos": 21637156499160.0, + "grad_norm": 9.480999869238888, + "learning_rate": 1.824179661311044e-06, + "loss": 0.6673, + "num_input_tokens_seen": 96042025, + "step": 4509 + }, + { + "epoch": 0.5422954367823003, + "flos": 13677774468120.0, + "grad_norm": 13.610418794286213, + "learning_rate": 1.823403718880868e-06, + "loss": 0.772, + "num_input_tokens_seen": 96060505, + "step": 4510 + }, + { + "epoch": 0.5424156796729394, + "flos": 29392937598360.0, + "grad_norm": 3.890947075806675, + "learning_rate": 1.822627803240207e-06, + "loss": 0.6594, + "num_input_tokens_seen": 96082555, + "step": 4511 + }, + { + "epoch": 0.5425359225635784, + "flos": 8358392947800.0, + "grad_norm": 7.382732669459268, + "learning_rate": 1.8218519145067675e-06, + "loss": 0.8375, + "num_input_tokens_seen": 96097895, + "step": 4512 + }, + { + "epoch": 0.5426561654542175, + "flos": 14784454183920.0, + "grad_norm": 8.112337545170156, + "learning_rate": 1.8210760527982508e-06, + "loss": 0.8719, + "num_input_tokens_seen": 96117900, + "step": 4513 + }, + { + "epoch": 0.5427764083448566, + "flos": 15999646277160.0, + "grad_norm": 7.3140976945518865, + "learning_rate": 1.8203002182323552e-06, + "loss": 0.7294, + "num_input_tokens_seen": 96135175, + "step": 4514 + }, + { + "epoch": 0.5428966512354957, + "flos": 14348030177160.0, + "grad_norm": 5.078974162775646, + "learning_rate": 1.819524410926773e-06, + "loss": 0.7444, + "num_input_tokens_seen": 96152575, + "step": 4515 + }, + { + "epoch": 0.5430168941261347, + "flos": 16377413360760.0, + "grad_norm": 2.9220737654861457, + "learning_rate": 1.8187486309991944e-06, + "loss": 0.7551, + "num_input_tokens_seen": 96173175, + "step": 4516 + }, + { + "epoch": 0.5431371370167739, + "flos": 13699013809920.0, + "grad_norm": 4.782852602382106, + "learning_rate": 1.817972878567304e-06, + "loss": 0.7488, + "num_input_tokens_seen": 96191550, + "step": 4517 + }, + { + "epoch": 0.543257379907413, + "flos": 13730732598480.0, + "grad_norm": 22.193370289671037, + "learning_rate": 1.8171971537487834e-06, + "loss": 0.7476, + "num_input_tokens_seen": 96209920, + "step": 4518 + }, + { + "epoch": 0.543377622798052, + "flos": 12759408414840.0, + "grad_norm": 8.893976260408115, + "learning_rate": 1.8164214566613093e-06, + "loss": 0.7891, + "num_input_tokens_seen": 96228265, + "step": 4519 + }, + { + "epoch": 0.5434978656886912, + "flos": 13866499710120.0, + "grad_norm": 8.002079693363802, + "learning_rate": 1.8156457874225547e-06, + "loss": 0.6255, + "num_input_tokens_seen": 96246445, + "step": 4520 + }, + { + "epoch": 0.5436181085793302, + "flos": 12597811267200.0, + "grad_norm": 10.131485733781997, + "learning_rate": 1.814870146150187e-06, + "loss": 0.7859, + "num_input_tokens_seen": 96264275, + "step": 4521 + }, + { + "epoch": 0.5437383514699693, + "flos": 13947345773880.0, + "grad_norm": 5.098283119681823, + "learning_rate": 1.814094532961871e-06, + "loss": 0.7708, + "num_input_tokens_seen": 96282570, + "step": 4522 + }, + { + "epoch": 0.5438585943606085, + "flos": 16540846786080.0, + "grad_norm": 5.677210628470599, + "learning_rate": 1.8133189479752666e-06, + "loss": 0.8094, + "num_input_tokens_seen": 96301220, + "step": 4523 + }, + { + "epoch": 0.5439788372512475, + "flos": 15945453408360.0, + "grad_norm": 7.631043337461326, + "learning_rate": 1.8125433913080292e-06, + "loss": 0.8045, + "num_input_tokens_seen": 96318640, + "step": 4524 + }, + { + "epoch": 0.5440990801418866, + "flos": 11949839678640.0, + "grad_norm": 4.869547215274446, + "learning_rate": 1.811767863077811e-06, + "loss": 0.7963, + "num_input_tokens_seen": 96337310, + "step": 4525 + }, + { + "epoch": 0.5442193230325257, + "flos": 15810066216240.0, + "grad_norm": 4.095731997990742, + "learning_rate": 1.8109923634022577e-06, + "loss": 0.7682, + "num_input_tokens_seen": 96357055, + "step": 4526 + }, + { + "epoch": 0.5443395659231648, + "flos": 11269484442360.0, + "grad_norm": 7.651825291905949, + "learning_rate": 1.8102168923990128e-06, + "loss": 0.8441, + "num_input_tokens_seen": 96370320, + "step": 4527 + }, + { + "epoch": 0.5444598088138038, + "flos": 13704522642960.0, + "grad_norm": 4.936016004613714, + "learning_rate": 1.809441450185714e-06, + "loss": 0.7717, + "num_input_tokens_seen": 96388525, + "step": 4528 + }, + { + "epoch": 0.544580051704443, + "flos": 15459933786360.0, + "grad_norm": 6.350931546824542, + "learning_rate": 1.8086660368799958e-06, + "loss": 0.7149, + "num_input_tokens_seen": 96406295, + "step": 4529 + }, + { + "epoch": 0.5447002945950821, + "flos": 23859380719560.0, + "grad_norm": 8.131833064222052, + "learning_rate": 1.807890652599488e-06, + "loss": 0.7492, + "num_input_tokens_seen": 96430400, + "step": 4530 + }, + { + "epoch": 0.5448205374857211, + "flos": 8544205473480.0, + "grad_norm": 10.988148633886903, + "learning_rate": 1.8071152974618156e-06, + "loss": 0.8045, + "num_input_tokens_seen": 96447920, + "step": 4531 + }, + { + "epoch": 0.5449407803763603, + "flos": 17673736457400.0, + "grad_norm": 29.034367684451773, + "learning_rate": 1.806339971584599e-06, + "loss": 0.7684, + "num_input_tokens_seen": 96464300, + "step": 4532 + }, + { + "epoch": 0.5450610232669993, + "flos": 17026081468440.0, + "grad_norm": 5.460515040045534, + "learning_rate": 1.8055646750854546e-06, + "loss": 0.838, + "num_input_tokens_seen": 96483530, + "step": 4533 + }, + { + "epoch": 0.5451812661576384, + "flos": 12975261751200.0, + "grad_norm": 7.6162935172508925, + "learning_rate": 1.8047894080819945e-06, + "loss": 0.799, + "num_input_tokens_seen": 96500500, + "step": 4534 + }, + { + "epoch": 0.5453015090482776, + "flos": 52410528725160.0, + "grad_norm": 0.7511026700694204, + "learning_rate": 1.8040141706918258e-06, + "loss": 0.6572, + "num_input_tokens_seen": 96561460, + "step": 4535 + }, + { + "epoch": 0.5454217519389166, + "flos": 18722835688680.0, + "grad_norm": 6.37396890118101, + "learning_rate": 1.8032389630325525e-06, + "loss": 0.7533, + "num_input_tokens_seen": 96579930, + "step": 4536 + }, + { + "epoch": 0.5455419948295557, + "flos": 17321736089880.0, + "grad_norm": 3.9542187772049853, + "learning_rate": 1.8024637852217707e-06, + "loss": 0.7472, + "num_input_tokens_seen": 96599375, + "step": 4537 + }, + { + "epoch": 0.5456622377201948, + "flos": 17133644047080.0, + "grad_norm": 5.039582549472646, + "learning_rate": 1.8016886373770766e-06, + "loss": 0.8293, + "num_input_tokens_seen": 96617610, + "step": 4538 + }, + { + "epoch": 0.5457824806108339, + "flos": 17565920599080.0, + "grad_norm": 5.008469465609175, + "learning_rate": 1.8009135196160579e-06, + "loss": 0.7542, + "num_input_tokens_seen": 96636205, + "step": 4539 + }, + { + "epoch": 0.545902723501473, + "flos": 16513528731960.0, + "grad_norm": 4.083106028140358, + "learning_rate": 1.8001384320563e-06, + "loss": 0.8249, + "num_input_tokens_seen": 96656180, + "step": 4540 + }, + { + "epoch": 0.5460229663921121, + "flos": 40665809656440.0, + "grad_norm": 0.8112205187902694, + "learning_rate": 1.7993633748153833e-06, + "loss": 0.6013, + "num_input_tokens_seen": 96710505, + "step": 4541 + }, + { + "epoch": 0.5461432092827512, + "flos": 11220452147040.0, + "grad_norm": 10.571389334469595, + "learning_rate": 1.7985883480108834e-06, + "loss": 0.7106, + "num_input_tokens_seen": 96727860, + "step": 4542 + }, + { + "epoch": 0.5462634521733902, + "flos": 17593270313160.0, + "grad_norm": 5.396251274428616, + "learning_rate": 1.797813351760371e-06, + "loss": 0.7112, + "num_input_tokens_seen": 96749285, + "step": 4543 + }, + { + "epoch": 0.5463836950640293, + "flos": 16702253973960.0, + "grad_norm": 6.801597340286658, + "learning_rate": 1.7970383861814116e-06, + "loss": 0.7585, + "num_input_tokens_seen": 96768775, + "step": 4544 + }, + { + "epoch": 0.5465039379546685, + "flos": 14946177971400.0, + "grad_norm": 7.171425799096255, + "learning_rate": 1.7962634513915684e-06, + "loss": 0.728, + "num_input_tokens_seen": 96785845, + "step": 4545 + }, + { + "epoch": 0.5466241808453075, + "flos": 12647761701360.0, + "grad_norm": 4.324160746915075, + "learning_rate": 1.7954885475083969e-06, + "loss": 0.7843, + "num_input_tokens_seen": 96803235, + "step": 4546 + }, + { + "epoch": 0.5467444237359466, + "flos": 15811300954680.0, + "grad_norm": 9.00832617586212, + "learning_rate": 1.7947136746494513e-06, + "loss": 0.7201, + "num_input_tokens_seen": 96823870, + "step": 4547 + }, + { + "epoch": 0.5468646666265857, + "flos": 17700927871680.0, + "grad_norm": 11.576746396501317, + "learning_rate": 1.793938832932277e-06, + "loss": 0.8559, + "num_input_tokens_seen": 96841700, + "step": 4548 + }, + { + "epoch": 0.5469849095172248, + "flos": 20184966707640.0, + "grad_norm": 7.699278948822878, + "learning_rate": 1.7931640224744185e-06, + "loss": 0.6813, + "num_input_tokens_seen": 96861970, + "step": 4549 + }, + { + "epoch": 0.5471051524078638, + "flos": 20508857522040.0, + "grad_norm": 4.862627222902361, + "learning_rate": 1.7923892433934127e-06, + "loss": 0.7214, + "num_input_tokens_seen": 96882765, + "step": 4550 + }, + { + "epoch": 0.547225395298503, + "flos": 13245782855760.0, + "grad_norm": 4.207404811641427, + "learning_rate": 1.7916144958067939e-06, + "loss": 0.7645, + "num_input_tokens_seen": 96900345, + "step": 4551 + }, + { + "epoch": 0.5473456381891421, + "flos": 15622164133200.0, + "grad_norm": 4.066571586831492, + "learning_rate": 1.7908397798320905e-06, + "loss": 0.7713, + "num_input_tokens_seen": 96919800, + "step": 4552 + }, + { + "epoch": 0.5474658810797811, + "flos": 14563440274080.0, + "grad_norm": 4.624812766744337, + "learning_rate": 1.7900650955868265e-06, + "loss": 0.7366, + "num_input_tokens_seen": 96939165, + "step": 4553 + }, + { + "epoch": 0.5475861239704203, + "flos": 37168881600720.0, + "grad_norm": 3.304765560589953, + "learning_rate": 1.7892904431885202e-06, + "loss": 0.7577, + "num_input_tokens_seen": 96966060, + "step": 4554 + }, + { + "epoch": 0.5477063668610593, + "flos": 15136454551440.0, + "grad_norm": 3.8290321987867513, + "learning_rate": 1.788515822754686e-06, + "loss": 0.7381, + "num_input_tokens_seen": 96986200, + "step": 4555 + }, + { + "epoch": 0.5478266097516984, + "flos": 14325682736760.0, + "grad_norm": 23.726071158890147, + "learning_rate": 1.7877412344028335e-06, + "loss": 0.7572, + "num_input_tokens_seen": 97005725, + "step": 4556 + }, + { + "epoch": 0.5479468526423376, + "flos": 9357541744920.0, + "grad_norm": 16.03216657193389, + "learning_rate": 1.7869666782504668e-06, + "loss": 0.7517, + "num_input_tokens_seen": 97022025, + "step": 4557 + }, + { + "epoch": 0.5480670955329766, + "flos": 13570211889480.0, + "grad_norm": 5.949810236621555, + "learning_rate": 1.7861921544150867e-06, + "loss": 0.6764, + "num_input_tokens_seen": 97040595, + "step": 4558 + }, + { + "epoch": 0.5481873384236157, + "flos": 11619711852120.0, + "grad_norm": 4.457370380735661, + "learning_rate": 1.7854176630141856e-06, + "loss": 0.743, + "num_input_tokens_seen": 97057450, + "step": 4559 + }, + { + "epoch": 0.5483075813142548, + "flos": 16674682640160.0, + "grad_norm": 8.636237301427196, + "learning_rate": 1.784643204165255e-06, + "loss": 0.8139, + "num_input_tokens_seen": 97076490, + "step": 4560 + }, + { + "epoch": 0.5484278242048939, + "flos": 14109101221320.0, + "grad_norm": 7.230285411387779, + "learning_rate": 1.7838687779857783e-06, + "loss": 0.7555, + "num_input_tokens_seen": 97094085, + "step": 4561 + }, + { + "epoch": 0.5485480670955329, + "flos": 16697916559440.0, + "grad_norm": 5.142886884542781, + "learning_rate": 1.7830943845932366e-06, + "loss": 0.6231, + "num_input_tokens_seen": 97113130, + "step": 4562 + }, + { + "epoch": 0.5486683099861721, + "flos": 16590987180000.0, + "grad_norm": 10.91531930621549, + "learning_rate": 1.7823200241051044e-06, + "loss": 0.7404, + "num_input_tokens_seen": 97131765, + "step": 4563 + }, + { + "epoch": 0.5487885528768112, + "flos": 16945330384560.0, + "grad_norm": 4.12141331884476, + "learning_rate": 1.7815456966388513e-06, + "loss": 0.7847, + "num_input_tokens_seen": 97150580, + "step": 4564 + }, + { + "epoch": 0.5489087957674502, + "flos": 16134495249960.0, + "grad_norm": 7.483849493065937, + "learning_rate": 1.780771402311943e-06, + "loss": 0.8022, + "num_input_tokens_seen": 97169135, + "step": 4565 + }, + { + "epoch": 0.5490290386580894, + "flos": 17808553770240.0, + "grad_norm": 9.897285135417581, + "learning_rate": 1.7799971412418374e-06, + "loss": 0.7682, + "num_input_tokens_seen": 97190250, + "step": 4566 + }, + { + "epoch": 0.5491492815487284, + "flos": 13352458955520.0, + "grad_norm": 8.645127404742992, + "learning_rate": 1.7792229135459918e-06, + "loss": 0.7195, + "num_input_tokens_seen": 97206620, + "step": 4567 + }, + { + "epoch": 0.5492695244393675, + "flos": 47588439840600.0, + "grad_norm": 0.7675120783920946, + "learning_rate": 1.7784487193418538e-06, + "loss": 0.6265, + "num_input_tokens_seen": 97264190, + "step": 4568 + }, + { + "epoch": 0.5493897673300067, + "flos": 12674573196120.0, + "grad_norm": 30.828865456135425, + "learning_rate": 1.7776745587468698e-06, + "loss": 0.5881, + "num_input_tokens_seen": 97281335, + "step": 4569 + }, + { + "epoch": 0.5495100102206457, + "flos": 14541821012760.0, + "grad_norm": 9.845816840309883, + "learning_rate": 1.7769004318784776e-06, + "loss": 0.798, + "num_input_tokens_seen": 97298700, + "step": 4570 + }, + { + "epoch": 0.5496302531112848, + "flos": 12192377869920.0, + "grad_norm": 13.09889400559917, + "learning_rate": 1.776126338854113e-06, + "loss": 0.7918, + "num_input_tokens_seen": 97316210, + "step": 4571 + }, + { + "epoch": 0.5497504960019239, + "flos": 17998418770800.0, + "grad_norm": 10.536707338542053, + "learning_rate": 1.7753522797912044e-06, + "loss": 0.8285, + "num_input_tokens_seen": 97336015, + "step": 4572 + }, + { + "epoch": 0.549870738892563, + "flos": 11247580241400.0, + "grad_norm": 9.60710839312465, + "learning_rate": 1.7745782548071765e-06, + "loss": 0.6819, + "num_input_tokens_seen": 97352630, + "step": 4573 + }, + { + "epoch": 0.549990981783202, + "flos": 15405709257600.0, + "grad_norm": 5.136424287543808, + "learning_rate": 1.7738042640194482e-06, + "loss": 0.7228, + "num_input_tokens_seen": 97372015, + "step": 4574 + }, + { + "epoch": 0.5501112246738411, + "flos": 15647329310040.0, + "grad_norm": 5.838467255208059, + "learning_rate": 1.7730303075454335e-06, + "loss": 0.6844, + "num_input_tokens_seen": 97390625, + "step": 4575 + }, + { + "epoch": 0.5502314675644803, + "flos": 12732438620280.0, + "grad_norm": 4.363256988867756, + "learning_rate": 1.7722563855025402e-06, + "loss": 0.835, + "num_input_tokens_seen": 97408375, + "step": 4576 + }, + { + "epoch": 0.5503517104551193, + "flos": 17803646476440.0, + "grad_norm": 7.728570659789518, + "learning_rate": 1.7714824980081721e-06, + "loss": 0.6933, + "num_input_tokens_seen": 97427390, + "step": 4577 + }, + { + "epoch": 0.5504719533457584, + "flos": 16405079674440.0, + "grad_norm": 5.117500773151677, + "learning_rate": 1.7707086451797276e-06, + "loss": 0.7219, + "num_input_tokens_seen": 97447985, + "step": 4578 + }, + { + "epoch": 0.5505921962363975, + "flos": 49620165861240.0, + "grad_norm": 0.7093944362044707, + "learning_rate": 1.7699348271345993e-06, + "loss": 0.533, + "num_input_tokens_seen": 97510330, + "step": 4579 + }, + { + "epoch": 0.5507124391270366, + "flos": 33626693276280.0, + "grad_norm": 0.7378575961312448, + "learning_rate": 1.7691610439901753e-06, + "loss": 0.5565, + "num_input_tokens_seen": 97572985, + "step": 4580 + }, + { + "epoch": 0.5508326820176757, + "flos": 16297390455960.0, + "grad_norm": 10.81896285580973, + "learning_rate": 1.7683872958638367e-06, + "loss": 0.7397, + "num_input_tokens_seen": 97591585, + "step": 4581 + }, + { + "epoch": 0.5509529249083148, + "flos": 14623901814960.0, + "grad_norm": 6.387191185699375, + "learning_rate": 1.7676135828729614e-06, + "loss": 0.8193, + "num_input_tokens_seen": 97611015, + "step": 4582 + }, + { + "epoch": 0.5510731677989539, + "flos": 15972296563080.0, + "grad_norm": 5.320386709628374, + "learning_rate": 1.7668399051349205e-06, + "loss": 0.8143, + "num_input_tokens_seen": 97630415, + "step": 4583 + }, + { + "epoch": 0.5511934106895929, + "flos": 15700477400160.0, + "grad_norm": 8.334035411505903, + "learning_rate": 1.766066262767081e-06, + "loss": 0.8092, + "num_input_tokens_seen": 97647975, + "step": 4584 + }, + { + "epoch": 0.5513136535802321, + "flos": 15405550957800.0, + "grad_norm": 8.89206012211636, + "learning_rate": 1.765292655886803e-06, + "loss": 0.7542, + "num_input_tokens_seen": 97666340, + "step": 4585 + }, + { + "epoch": 0.5514338964708712, + "flos": 20397559068120.0, + "grad_norm": 5.5371586334890095, + "learning_rate": 1.764519084611443e-06, + "loss": 0.6986, + "num_input_tokens_seen": 97686515, + "step": 4586 + }, + { + "epoch": 0.5515541393615102, + "flos": 16026964331280.0, + "grad_norm": 4.667065853121012, + "learning_rate": 1.7637455490583505e-06, + "loss": 0.7528, + "num_input_tokens_seen": 97705560, + "step": 4587 + }, + { + "epoch": 0.5516743822521494, + "flos": 14973749305200.0, + "grad_norm": 9.385741730260238, + "learning_rate": 1.7629720493448701e-06, + "loss": 0.7609, + "num_input_tokens_seen": 97722575, + "step": 4588 + }, + { + "epoch": 0.5517946251427884, + "flos": 10870161417360.0, + "grad_norm": 10.87083618429158, + "learning_rate": 1.7621985855883418e-06, + "loss": 0.8353, + "num_input_tokens_seen": 97738995, + "step": 4589 + }, + { + "epoch": 0.5519148680334275, + "flos": 13433874898560.0, + "grad_norm": 6.467820547658906, + "learning_rate": 1.7614251579060983e-06, + "loss": 0.7012, + "num_input_tokens_seen": 97757310, + "step": 4590 + }, + { + "epoch": 0.5520351109240667, + "flos": 18397900095600.0, + "grad_norm": 5.7756541620630175, + "learning_rate": 1.76065176641547e-06, + "loss": 0.8378, + "num_input_tokens_seen": 97779740, + "step": 4591 + }, + { + "epoch": 0.5521553538147057, + "flos": 15406089177120.0, + "grad_norm": 5.6764425586867615, + "learning_rate": 1.759878411233777e-06, + "loss": 0.7602, + "num_input_tokens_seen": 97797920, + "step": 4592 + }, + { + "epoch": 0.5522755967053448, + "flos": 13785621986400.0, + "grad_norm": 4.529742812377879, + "learning_rate": 1.7591050924783388e-06, + "loss": 0.7481, + "num_input_tokens_seen": 97814830, + "step": 4593 + }, + { + "epoch": 0.5523958395959839, + "flos": 47683469566440.0, + "grad_norm": 0.8643676687993018, + "learning_rate": 1.7583318102664661e-06, + "loss": 0.6254, + "num_input_tokens_seen": 97882115, + "step": 4594 + }, + { + "epoch": 0.552516082486623, + "flos": 7872303446520.0, + "grad_norm": 4.659031389283412, + "learning_rate": 1.757558564715466e-06, + "loss": 0.7724, + "num_input_tokens_seen": 97899910, + "step": 4595 + }, + { + "epoch": 0.552636325377262, + "flos": 16242437748120.0, + "grad_norm": 8.9359424007407, + "learning_rate": 1.7567853559426386e-06, + "loss": 0.7089, + "num_input_tokens_seen": 97916680, + "step": 4596 + }, + { + "epoch": 0.5527565682679012, + "flos": 17566047238920.0, + "grad_norm": 3.988114102230179, + "learning_rate": 1.7560121840652797e-06, + "loss": 0.7373, + "num_input_tokens_seen": 97935785, + "step": 4597 + }, + { + "epoch": 0.5528768111585403, + "flos": 14406750420240.0, + "grad_norm": 4.5458664447383885, + "learning_rate": 1.7552390492006782e-06, + "loss": 0.6691, + "num_input_tokens_seen": 97953825, + "step": 4598 + }, + { + "epoch": 0.5529970540491793, + "flos": 19215067222200.0, + "grad_norm": 6.3226475498635, + "learning_rate": 1.7544659514661184e-06, + "loss": 0.6344, + "num_input_tokens_seen": 97976635, + "step": 4599 + }, + { + "epoch": 0.5531172969398185, + "flos": 17890033033200.0, + "grad_norm": 4.121020622870152, + "learning_rate": 1.7536928909788786e-06, + "loss": 0.7853, + "num_input_tokens_seen": 97995660, + "step": 4600 + }, + { + "epoch": 0.5532375398304575, + "flos": 47851584174480.0, + "grad_norm": 0.8869969943919217, + "learning_rate": 1.752919867856231e-06, + "loss": 0.6328, + "num_input_tokens_seen": 98047025, + "step": 4601 + }, + { + "epoch": 0.5533577827210966, + "flos": 14380033905360.0, + "grad_norm": 3.624915417966956, + "learning_rate": 1.7521468822154436e-06, + "loss": 0.7693, + "num_input_tokens_seen": 98065660, + "step": 4602 + }, + { + "epoch": 0.5534780256117358, + "flos": 23722093929840.0, + "grad_norm": 4.44190502469527, + "learning_rate": 1.751373934173777e-06, + "loss": 0.7192, + "num_input_tokens_seen": 98088125, + "step": 4603 + }, + { + "epoch": 0.5535982685023748, + "flos": 16999396613520.0, + "grad_norm": 7.458792333135952, + "learning_rate": 1.750601023848487e-06, + "loss": 0.7155, + "num_input_tokens_seen": 98108570, + "step": 4604 + }, + { + "epoch": 0.5537185113930139, + "flos": 17835618544680.0, + "grad_norm": 3.458758837402877, + "learning_rate": 1.749828151356823e-06, + "loss": 0.7198, + "num_input_tokens_seen": 98128485, + "step": 4605 + }, + { + "epoch": 0.553838754283653, + "flos": 17240890026120.0, + "grad_norm": 5.167922246814584, + "learning_rate": 1.7490553168160297e-06, + "loss": 0.7464, + "num_input_tokens_seen": 98149275, + "step": 4606 + }, + { + "epoch": 0.5539589971742921, + "flos": 12597906247080.0, + "grad_norm": 5.668997715223992, + "learning_rate": 1.748282520343345e-06, + "loss": 0.7412, + "num_input_tokens_seen": 98168025, + "step": 4607 + }, + { + "epoch": 0.5540792400649311, + "flos": 20212569701400.0, + "grad_norm": 5.52932326385658, + "learning_rate": 1.7475097620560023e-06, + "loss": 0.7662, + "num_input_tokens_seen": 98187810, + "step": 4608 + }, + { + "epoch": 0.5541994829555702, + "flos": 17484979555440.0, + "grad_norm": 5.799651681372265, + "learning_rate": 1.746737042071228e-06, + "loss": 0.6926, + "num_input_tokens_seen": 98206035, + "step": 4609 + }, + { + "epoch": 0.5543197258462094, + "flos": 14699460665400.0, + "grad_norm": 4.245062484371204, + "learning_rate": 1.7459643605062424e-06, + "loss": 0.7807, + "num_input_tokens_seen": 98223015, + "step": 4610 + }, + { + "epoch": 0.5544399687368484, + "flos": 15291878006880.0, + "grad_norm": 3.5234790469678385, + "learning_rate": 1.745191717478262e-06, + "loss": 0.7889, + "num_input_tokens_seen": 98241315, + "step": 4611 + }, + { + "epoch": 0.5545602116274875, + "flos": 18698430350880.0, + "grad_norm": 3.8586514061039674, + "learning_rate": 1.7444191131044948e-06, + "loss": 0.7876, + "num_input_tokens_seen": 98261310, + "step": 4612 + }, + { + "epoch": 0.5546804545181266, + "flos": 15351199789200.0, + "grad_norm": 3.8496715554180367, + "learning_rate": 1.7436465475021456e-06, + "loss": 0.7137, + "num_input_tokens_seen": 98281080, + "step": 4613 + }, + { + "epoch": 0.5548006974087657, + "flos": 19671812431920.0, + "grad_norm": 4.852159251493656, + "learning_rate": 1.7428740207884111e-06, + "loss": 0.6852, + "num_input_tokens_seen": 98301680, + "step": 4614 + }, + { + "epoch": 0.5549209402994048, + "flos": 24721116087120.0, + "grad_norm": 3.6158050465904172, + "learning_rate": 1.7421015330804833e-06, + "loss": 0.6079, + "num_input_tokens_seen": 98321320, + "step": 4615 + }, + { + "epoch": 0.5550411831900439, + "flos": 17404260131520.0, + "grad_norm": 4.9130652648454465, + "learning_rate": 1.7413290844955475e-06, + "loss": 0.7094, + "num_input_tokens_seen": 98341070, + "step": 4616 + }, + { + "epoch": 0.555161426080683, + "flos": 15836244511800.0, + "grad_norm": 3.5151431961519184, + "learning_rate": 1.7405566751507843e-06, + "loss": 0.7659, + "num_input_tokens_seen": 98358835, + "step": 4617 + }, + { + "epoch": 0.555281668971322, + "flos": 36492990418800.0, + "grad_norm": 3.4681761670802307, + "learning_rate": 1.7397843051633668e-06, + "loss": 0.6521, + "num_input_tokens_seen": 98381250, + "step": 4618 + }, + { + "epoch": 0.5554019118619612, + "flos": 15163456005960.0, + "grad_norm": 5.763460078495904, + "learning_rate": 1.739011974650464e-06, + "loss": 0.698, + "num_input_tokens_seen": 98400300, + "step": 4619 + }, + { + "epoch": 0.5555221547526003, + "flos": 18672347035200.0, + "grad_norm": 16.333519966979523, + "learning_rate": 1.7382396837292365e-06, + "loss": 0.7376, + "num_input_tokens_seen": 98420480, + "step": 4620 + }, + { + "epoch": 0.5556423976432393, + "flos": 15918895193280.0, + "grad_norm": 3.8590307661353895, + "learning_rate": 1.737467432516841e-06, + "loss": 0.7209, + "num_input_tokens_seen": 98440300, + "step": 4621 + }, + { + "epoch": 0.5557626405338785, + "flos": 18240956962080.0, + "grad_norm": 5.785894394481444, + "learning_rate": 1.7366952211304274e-06, + "loss": 0.7156, + "num_input_tokens_seen": 98457865, + "step": 4622 + }, + { + "epoch": 0.5558828834245175, + "flos": 13650551393880.0, + "grad_norm": 38.09423547424449, + "learning_rate": 1.735923049687139e-06, + "loss": 0.8181, + "num_input_tokens_seen": 98474160, + "step": 4623 + }, + { + "epoch": 0.5560031263151566, + "flos": 19996463085360.0, + "grad_norm": 3.2605013282842203, + "learning_rate": 1.7351509183041144e-06, + "loss": 0.7316, + "num_input_tokens_seen": 98494210, + "step": 4624 + }, + { + "epoch": 0.5561233692057957, + "flos": 17133422427360.0, + "grad_norm": 6.088011245095574, + "learning_rate": 1.7343788270984852e-06, + "loss": 0.7084, + "num_input_tokens_seen": 98513070, + "step": 4625 + }, + { + "epoch": 0.5562436120964348, + "flos": 27692225883120.0, + "grad_norm": 3.6261443600920042, + "learning_rate": 1.7336067761873764e-06, + "loss": 0.7329, + "num_input_tokens_seen": 98535215, + "step": 4626 + }, + { + "epoch": 0.5563638549870739, + "flos": 18429682204080.0, + "grad_norm": 4.351452920242829, + "learning_rate": 1.7328347656879076e-06, + "loss": 0.7403, + "num_input_tokens_seen": 98554795, + "step": 4627 + }, + { + "epoch": 0.556484097877713, + "flos": 9863509209720.0, + "grad_norm": 6.947340850379908, + "learning_rate": 1.7320627957171927e-06, + "loss": 0.6754, + "num_input_tokens_seen": 98569175, + "step": 4628 + }, + { + "epoch": 0.5566043407683521, + "flos": 18079043214840.0, + "grad_norm": 5.244167539502628, + "learning_rate": 1.7312908663923382e-06, + "loss": 0.7881, + "num_input_tokens_seen": 98585070, + "step": 4629 + }, + { + "epoch": 0.5567245836589911, + "flos": 15048833256240.0, + "grad_norm": 5.5883051622823885, + "learning_rate": 1.7305189778304463e-06, + "loss": 0.6583, + "num_input_tokens_seen": 98602965, + "step": 4630 + }, + { + "epoch": 0.5568448265496303, + "flos": 15135631392480.0, + "grad_norm": 5.048326785432925, + "learning_rate": 1.729747130148611e-06, + "loss": 0.786, + "num_input_tokens_seen": 98621880, + "step": 4631 + }, + { + "epoch": 0.5569650694402694, + "flos": 18539207700240.0, + "grad_norm": 8.02127108349555, + "learning_rate": 1.7289753234639208e-06, + "loss": 0.7498, + "num_input_tokens_seen": 98640575, + "step": 4632 + }, + { + "epoch": 0.5570853123309084, + "flos": 14400893327640.0, + "grad_norm": 3.6001596097531903, + "learning_rate": 1.7282035578934592e-06, + "loss": 0.7462, + "num_input_tokens_seen": 98658460, + "step": 4633 + }, + { + "epoch": 0.5572055552215476, + "flos": 11734366261800.0, + "grad_norm": 6.153752008965083, + "learning_rate": 1.727431833554301e-06, + "loss": 0.7651, + "num_input_tokens_seen": 98676655, + "step": 4634 + }, + { + "epoch": 0.5573257981121866, + "flos": 12489647149320.0, + "grad_norm": 5.284148194122793, + "learning_rate": 1.7266601505635175e-06, + "loss": 0.7481, + "num_input_tokens_seen": 98693715, + "step": 4635 + }, + { + "epoch": 0.5574460410028257, + "flos": 13731619077360.0, + "grad_norm": 4.610066197182772, + "learning_rate": 1.7258885090381717e-06, + "loss": 0.7372, + "num_input_tokens_seen": 98711475, + "step": 4636 + }, + { + "epoch": 0.5575662838934649, + "flos": 21750639490320.0, + "grad_norm": 8.20343330893835, + "learning_rate": 1.7251169090953213e-06, + "loss": 0.7722, + "num_input_tokens_seen": 98731670, + "step": 4637 + }, + { + "epoch": 0.5576865267841039, + "flos": 16134970149360.0, + "grad_norm": 13.172611745409963, + "learning_rate": 1.7243453508520168e-06, + "loss": 0.7407, + "num_input_tokens_seen": 98748375, + "step": 4638 + }, + { + "epoch": 0.557806769674743, + "flos": 12538584464760.0, + "grad_norm": 5.840261795237784, + "learning_rate": 1.7235738344253038e-06, + "loss": 0.8315, + "num_input_tokens_seen": 98761725, + "step": 4639 + }, + { + "epoch": 0.557927012565382, + "flos": 18101517295080.0, + "grad_norm": 4.98908606528748, + "learning_rate": 1.72280235993222e-06, + "loss": 0.8, + "num_input_tokens_seen": 98779750, + "step": 4640 + }, + { + "epoch": 0.5580472554560212, + "flos": 12381831291000.0, + "grad_norm": 11.642884638065969, + "learning_rate": 1.722030927489798e-06, + "loss": 0.6696, + "num_input_tokens_seen": 98796750, + "step": 4641 + }, + { + "epoch": 0.5581674983466602, + "flos": 17210817555480.0, + "grad_norm": 3.0576611652458543, + "learning_rate": 1.7212595372150634e-06, + "loss": 0.728, + "num_input_tokens_seen": 98816450, + "step": 4642 + }, + { + "epoch": 0.5582877412372993, + "flos": 9789501697320.0, + "grad_norm": 8.312166823973778, + "learning_rate": 1.720488189225035e-06, + "loss": 0.7134, + "num_input_tokens_seen": 98833870, + "step": 4643 + }, + { + "epoch": 0.5584079841279385, + "flos": 16022626916760.0, + "grad_norm": 4.67989678782579, + "learning_rate": 1.7197168836367265e-06, + "loss": 0.7752, + "num_input_tokens_seen": 98850400, + "step": 4644 + }, + { + "epoch": 0.5585282270185775, + "flos": 13758652191840.0, + "grad_norm": 4.039887865984223, + "learning_rate": 1.7189456205671433e-06, + "loss": 0.8039, + "num_input_tokens_seen": 98868965, + "step": 4645 + }, + { + "epoch": 0.5586484699092166, + "flos": 15996796880760.0, + "grad_norm": 2.918968182726564, + "learning_rate": 1.7181744001332866e-06, + "loss": 0.8078, + "num_input_tokens_seen": 98887295, + "step": 4646 + }, + { + "epoch": 0.5587687127998557, + "flos": 16755655343760.0, + "grad_norm": 4.736304767983645, + "learning_rate": 1.7174032224521493e-06, + "loss": 0.6231, + "num_input_tokens_seen": 98905725, + "step": 4647 + }, + { + "epoch": 0.5588889556904948, + "flos": 14838805352520.0, + "grad_norm": 3.6412266573827248, + "learning_rate": 1.7166320876407184e-06, + "loss": 0.6848, + "num_input_tokens_seen": 98924865, + "step": 4648 + }, + { + "epoch": 0.5590091985811338, + "flos": 12003684287880.0, + "grad_norm": 9.791153351876424, + "learning_rate": 1.7158609958159742e-06, + "loss": 0.6675, + "num_input_tokens_seen": 98941990, + "step": 4649 + }, + { + "epoch": 0.559129441471773, + "flos": 10567098365280.0, + "grad_norm": 4.135350506282926, + "learning_rate": 1.7150899470948911e-06, + "loss": 0.7669, + "num_input_tokens_seen": 98956975, + "step": 4650 + }, + { + "epoch": 0.5592496843624121, + "flos": 44609451205920.0, + "grad_norm": 0.8070906522146902, + "learning_rate": 1.7143189415944365e-06, + "loss": 0.593, + "num_input_tokens_seen": 99021155, + "step": 4651 + }, + { + "epoch": 0.5593699272530511, + "flos": 15270322065480.0, + "grad_norm": 4.136615236768514, + "learning_rate": 1.7135479794315714e-06, + "loss": 0.7441, + "num_input_tokens_seen": 99037830, + "step": 4652 + }, + { + "epoch": 0.5594901701436903, + "flos": 9357668384760.0, + "grad_norm": 4.500349878908613, + "learning_rate": 1.7127770607232502e-06, + "loss": 0.7683, + "num_input_tokens_seen": 99056095, + "step": 4653 + }, + { + "epoch": 0.5596104130343293, + "flos": 16917632410920.0, + "grad_norm": 5.192868683334243, + "learning_rate": 1.7120061855864204e-06, + "loss": 0.7848, + "num_input_tokens_seen": 99075825, + "step": 4654 + }, + { + "epoch": 0.5597306559249684, + "flos": 19023619223640.0, + "grad_norm": 4.229400302590861, + "learning_rate": 1.7112353541380233e-06, + "loss": 0.6979, + "num_input_tokens_seen": 99095405, + "step": 4655 + }, + { + "epoch": 0.5598508988156076, + "flos": 16458987603600.0, + "grad_norm": 3.934046972051567, + "learning_rate": 1.7104645664949931e-06, + "loss": 0.7154, + "num_input_tokens_seen": 99117595, + "step": 4656 + }, + { + "epoch": 0.5599711417062466, + "flos": 16918012330440.0, + "grad_norm": 3.550654342951811, + "learning_rate": 1.7096938227742584e-06, + "loss": 0.7091, + "num_input_tokens_seen": 99138445, + "step": 4657 + }, + { + "epoch": 0.5600913845968857, + "flos": 16344808093320.0, + "grad_norm": 4.481539264956808, + "learning_rate": 1.70892312309274e-06, + "loss": 0.8218, + "num_input_tokens_seen": 99156055, + "step": 4658 + }, + { + "epoch": 0.5602116274875248, + "flos": 12862538599080.0, + "grad_norm": 5.1408348156626, + "learning_rate": 1.7081524675673523e-06, + "loss": 0.655, + "num_input_tokens_seen": 99171265, + "step": 4659 + }, + { + "epoch": 0.5603318703781639, + "flos": 51710232205440.0, + "grad_norm": 0.9818990541033409, + "learning_rate": 1.7073818563150026e-06, + "loss": 0.6356, + "num_input_tokens_seen": 99233065, + "step": 4660 + }, + { + "epoch": 0.560452113268803, + "flos": 13537511642160.0, + "grad_norm": 4.458145684306972, + "learning_rate": 1.7066112894525935e-06, + "loss": 0.8399, + "num_input_tokens_seen": 99250865, + "step": 4661 + }, + { + "epoch": 0.5605723561594421, + "flos": 18510654907680.0, + "grad_norm": 2.860896805161828, + "learning_rate": 1.7058407670970177e-06, + "loss": 0.7145, + "num_input_tokens_seen": 99272060, + "step": 4662 + }, + { + "epoch": 0.5606925990500812, + "flos": 15054816988680.0, + "grad_norm": 3.7971831773562705, + "learning_rate": 1.7050702893651643e-06, + "loss": 0.5929, + "num_input_tokens_seen": 99291360, + "step": 4663 + }, + { + "epoch": 0.5608128419407202, + "flos": 25719884964720.0, + "grad_norm": 7.278215684731169, + "learning_rate": 1.7042998563739134e-06, + "loss": 0.7405, + "num_input_tokens_seen": 99309430, + "step": 4664 + }, + { + "epoch": 0.5609330848313594, + "flos": 18047577705960.0, + "grad_norm": 4.3221158063308485, + "learning_rate": 1.703529468240139e-06, + "loss": 0.6945, + "num_input_tokens_seen": 99328020, + "step": 4665 + }, + { + "epoch": 0.5610533277219985, + "flos": 13697874051360.0, + "grad_norm": 7.978602403521344, + "learning_rate": 1.7027591250807088e-06, + "loss": 0.711, + "num_input_tokens_seen": 99344915, + "step": 4666 + }, + { + "epoch": 0.5611735706126375, + "flos": 10923309507480.0, + "grad_norm": 8.698432903997798, + "learning_rate": 1.7019888270124825e-06, + "loss": 0.8057, + "num_input_tokens_seen": 99361800, + "step": 4667 + }, + { + "epoch": 0.5612938135032767, + "flos": 12000170032320.0, + "grad_norm": 4.345399173007967, + "learning_rate": 1.7012185741523147e-06, + "loss": 0.8085, + "num_input_tokens_seen": 99377845, + "step": 4668 + }, + { + "epoch": 0.5614140563939157, + "flos": 18808050826920.0, + "grad_norm": 5.130947848494916, + "learning_rate": 1.7004483666170514e-06, + "loss": 0.592, + "num_input_tokens_seen": 99398060, + "step": 4669 + }, + { + "epoch": 0.5615342992845548, + "flos": 18105791389680.0, + "grad_norm": 3.675049144390328, + "learning_rate": 1.699678204523533e-06, + "loss": 0.7747, + "num_input_tokens_seen": 99417645, + "step": 4670 + }, + { + "epoch": 0.5616545421751938, + "flos": 16106164077120.0, + "grad_norm": 6.838819983794862, + "learning_rate": 1.6989080879885918e-06, + "loss": 0.6553, + "num_input_tokens_seen": 99435225, + "step": 4671 + }, + { + "epoch": 0.561774785065833, + "flos": 39604589151840.0, + "grad_norm": 0.8924506699529157, + "learning_rate": 1.6981380171290544e-06, + "loss": 0.6305, + "num_input_tokens_seen": 99495970, + "step": 4672 + }, + { + "epoch": 0.5618950279564721, + "flos": 14428686281160.0, + "grad_norm": 4.287397217700239, + "learning_rate": 1.6973679920617396e-06, + "loss": 0.7246, + "num_input_tokens_seen": 99513225, + "step": 4673 + }, + { + "epoch": 0.5620152708471111, + "flos": 12246760698480.0, + "grad_norm": 7.306325016990844, + "learning_rate": 1.6965980129034603e-06, + "loss": 0.8393, + "num_input_tokens_seen": 99530330, + "step": 4674 + }, + { + "epoch": 0.5621355137377503, + "flos": 19645380856680.0, + "grad_norm": 3.674405269118538, + "learning_rate": 1.6958280797710209e-06, + "loss": 0.7529, + "num_input_tokens_seen": 99551975, + "step": 4675 + }, + { + "epoch": 0.5622557566283893, + "flos": 45116146849800.0, + "grad_norm": 0.7189867210329183, + "learning_rate": 1.6950581927812198e-06, + "loss": 0.5506, + "num_input_tokens_seen": 99611265, + "step": 4676 + }, + { + "epoch": 0.5623759995190284, + "flos": 19401576267000.0, + "grad_norm": 3.933166662553516, + "learning_rate": 1.6942883520508486e-06, + "loss": 0.766, + "num_input_tokens_seen": 99629720, + "step": 4677 + }, + { + "epoch": 0.5624962424096676, + "flos": 14163737329560.0, + "grad_norm": 4.555661445772293, + "learning_rate": 1.693518557696691e-06, + "loss": 0.7632, + "num_input_tokens_seen": 99648580, + "step": 4678 + }, + { + "epoch": 0.5626164853003066, + "flos": 15108313338360.0, + "grad_norm": 5.022953561873852, + "learning_rate": 1.6927488098355252e-06, + "loss": 0.8744, + "num_input_tokens_seen": 99665930, + "step": 4679 + }, + { + "epoch": 0.5627367281909457, + "flos": 46268534565120.0, + "grad_norm": 0.8901163623049478, + "learning_rate": 1.6919791085841201e-06, + "loss": 0.6664, + "num_input_tokens_seen": 99723060, + "step": 4680 + }, + { + "epoch": 0.5628569710815848, + "flos": 9276537381360.0, + "grad_norm": 7.30109762013495, + "learning_rate": 1.6912094540592396e-06, + "loss": 0.7782, + "num_input_tokens_seen": 99738300, + "step": 4681 + }, + { + "epoch": 0.5629772139722239, + "flos": 9996616884720.0, + "grad_norm": 7.015276452053081, + "learning_rate": 1.6904398463776393e-06, + "loss": 0.7851, + "num_input_tokens_seen": 99751820, + "step": 4682 + }, + { + "epoch": 0.5630974568628629, + "flos": 15700414080240.0, + "grad_norm": 3.7771600720029976, + "learning_rate": 1.6896702856560683e-06, + "loss": 0.7173, + "num_input_tokens_seen": 99770635, + "step": 4683 + }, + { + "epoch": 0.5632176997535021, + "flos": 10356025682880.0, + "grad_norm": 4.597828410513957, + "learning_rate": 1.6889007720112677e-06, + "loss": 0.6585, + "num_input_tokens_seen": 99788100, + "step": 4684 + }, + { + "epoch": 0.5633379426441412, + "flos": 15215401017600.0, + "grad_norm": 6.819438795340745, + "learning_rate": 1.6881313055599734e-06, + "loss": 0.7635, + "num_input_tokens_seen": 99807750, + "step": 4685 + }, + { + "epoch": 0.5634581855347802, + "flos": 16539232128120.0, + "grad_norm": 4.884413409432912, + "learning_rate": 1.6873618864189117e-06, + "loss": 0.8012, + "num_input_tokens_seen": 99823240, + "step": 4686 + }, + { + "epoch": 0.5635784284254194, + "flos": 15729758371800.0, + "grad_norm": 7.410074397445926, + "learning_rate": 1.686592514704803e-06, + "loss": 0.7647, + "num_input_tokens_seen": 99840355, + "step": 4687 + }, + { + "epoch": 0.5636986713160584, + "flos": 14486646685200.0, + "grad_norm": 3.919750960671987, + "learning_rate": 1.685823190534361e-06, + "loss": 0.6873, + "num_input_tokens_seen": 99858315, + "step": 4688 + }, + { + "epoch": 0.5638189142066975, + "flos": 14460943289040.0, + "grad_norm": 3.9050019979044435, + "learning_rate": 1.6850539140242907e-06, + "loss": 0.8071, + "num_input_tokens_seen": 99877295, + "step": 4689 + }, + { + "epoch": 0.5639391570973367, + "flos": 16756130243160.0, + "grad_norm": 18.396671723917294, + "learning_rate": 1.684284685291292e-06, + "loss": 0.7986, + "num_input_tokens_seen": 99898660, + "step": 4690 + }, + { + "epoch": 0.5640593999879757, + "flos": 17372003123640.0, + "grad_norm": 3.73767926861073, + "learning_rate": 1.683515504452055e-06, + "loss": 0.7887, + "num_input_tokens_seen": 99915755, + "step": 4691 + }, + { + "epoch": 0.5641796428786148, + "flos": 16620046531920.0, + "grad_norm": 6.798398884036932, + "learning_rate": 1.6827463716232648e-06, + "loss": 0.6463, + "num_input_tokens_seen": 99936135, + "step": 4692 + }, + { + "epoch": 0.5642998857692539, + "flos": 14460531709560.0, + "grad_norm": 3.7574444945519385, + "learning_rate": 1.6819772869215972e-06, + "loss": 0.7398, + "num_input_tokens_seen": 99954935, + "step": 4693 + }, + { + "epoch": 0.564420128659893, + "flos": 16972363499040.0, + "grad_norm": 4.228165807195793, + "learning_rate": 1.6812082504637228e-06, + "loss": 0.8038, + "num_input_tokens_seen": 99975975, + "step": 4694 + }, + { + "epoch": 0.564540371550532, + "flos": 17025828188760.0, + "grad_norm": 3.501596223055132, + "learning_rate": 1.6804392623663025e-06, + "loss": 0.7351, + "num_input_tokens_seen": 99996900, + "step": 4695 + }, + { + "epoch": 0.5646606144411712, + "flos": 18322341245160.0, + "grad_norm": 3.972088708538128, + "learning_rate": 1.6796703227459935e-06, + "loss": 0.7595, + "num_input_tokens_seen": 100014575, + "step": 4696 + }, + { + "epoch": 0.5647808573318103, + "flos": 26855022493200.0, + "grad_norm": 3.4739877832051866, + "learning_rate": 1.6789014317194407e-06, + "loss": 0.7441, + "num_input_tokens_seen": 100035775, + "step": 4697 + }, + { + "epoch": 0.5649011002224493, + "flos": 16485704118480.0, + "grad_norm": 13.417905635768594, + "learning_rate": 1.6781325894032853e-06, + "loss": 0.7105, + "num_input_tokens_seen": 100054455, + "step": 4698 + }, + { + "epoch": 0.5650213431130885, + "flos": 13515607441200.0, + "grad_norm": 4.543620724524669, + "learning_rate": 1.6773637959141608e-06, + "loss": 0.9004, + "num_input_tokens_seen": 100071150, + "step": 4699 + }, + { + "epoch": 0.5651415860037275, + "flos": 12783275533320.0, + "grad_norm": 4.905116636734615, + "learning_rate": 1.6765950513686915e-06, + "loss": 0.6321, + "num_input_tokens_seen": 100088980, + "step": 4700 + }, + { + "epoch": 0.5652618288943666, + "flos": 18699633429360.0, + "grad_norm": 4.601545505459916, + "learning_rate": 1.675826355883496e-06, + "loss": 0.7418, + "num_input_tokens_seen": 100107915, + "step": 4701 + }, + { + "epoch": 0.5653820717850057, + "flos": 14380033905360.0, + "grad_norm": 3.782528747731733, + "learning_rate": 1.6750577095751848e-06, + "loss": 0.7806, + "num_input_tokens_seen": 100126745, + "step": 4702 + }, + { + "epoch": 0.5655023146756448, + "flos": 19779944889840.0, + "grad_norm": 3.4005412288051606, + "learning_rate": 1.6742891125603605e-06, + "loss": 0.7219, + "num_input_tokens_seen": 100147370, + "step": 4703 + }, + { + "epoch": 0.5656225575662839, + "flos": 20289806529720.0, + "grad_norm": 5.161287378687173, + "learning_rate": 1.6735205649556185e-06, + "loss": 0.6994, + "num_input_tokens_seen": 100166960, + "step": 4704 + }, + { + "epoch": 0.5657428004569229, + "flos": 17833149067800.0, + "grad_norm": 3.3385028489703297, + "learning_rate": 1.6727520668775476e-06, + "loss": 0.8256, + "num_input_tokens_seen": 100186965, + "step": 4705 + }, + { + "epoch": 0.5658630433475621, + "flos": 16053110966880.0, + "grad_norm": 4.6394503869734915, + "learning_rate": 1.6719836184427275e-06, + "loss": 0.7407, + "num_input_tokens_seen": 100206990, + "step": 4706 + }, + { + "epoch": 0.5659832862382012, + "flos": 22317796675080.0, + "grad_norm": 2.928519461167372, + "learning_rate": 1.671215219767733e-06, + "loss": 0.6339, + "num_input_tokens_seen": 100226170, + "step": 4707 + }, + { + "epoch": 0.5661035291288402, + "flos": 9573300101400.0, + "grad_norm": 15.463820533212303, + "learning_rate": 1.670446870969127e-06, + "loss": 0.7459, + "num_input_tokens_seen": 100243555, + "step": 4708 + }, + { + "epoch": 0.5662237720194794, + "flos": 11760006338040.0, + "grad_norm": 4.477204974075099, + "learning_rate": 1.6696785721634685e-06, + "loss": 0.7963, + "num_input_tokens_seen": 100257760, + "step": 4709 + }, + { + "epoch": 0.5663440149101184, + "flos": 12895333826280.0, + "grad_norm": 4.576386730378175, + "learning_rate": 1.6689103234673086e-06, + "loss": 0.7157, + "num_input_tokens_seen": 100275800, + "step": 4710 + }, + { + "epoch": 0.5664642578007575, + "flos": 17107370771640.0, + "grad_norm": 13.390817576284434, + "learning_rate": 1.668142124997189e-06, + "loss": 0.7404, + "num_input_tokens_seen": 100295180, + "step": 4711 + }, + { + "epoch": 0.5665845006913967, + "flos": 48305991038520.0, + "grad_norm": 0.7600865801633933, + "learning_rate": 1.6673739768696453e-06, + "loss": 0.6208, + "num_input_tokens_seen": 100361470, + "step": 4712 + }, + { + "epoch": 0.5667047435820357, + "flos": 19159196375520.0, + "grad_norm": 6.137596723178947, + "learning_rate": 1.6666058792012052e-06, + "loss": 0.7504, + "num_input_tokens_seen": 100382075, + "step": 4713 + }, + { + "epoch": 0.5668249864726748, + "flos": 53006048742720.0, + "grad_norm": 0.8603239542139376, + "learning_rate": 1.6658378321083878e-06, + "loss": 0.6915, + "num_input_tokens_seen": 100446125, + "step": 4714 + }, + { + "epoch": 0.5669452293633139, + "flos": 16238796852720.0, + "grad_norm": 4.546476261444023, + "learning_rate": 1.6650698357077055e-06, + "loss": 0.8088, + "num_input_tokens_seen": 100462890, + "step": 4715 + }, + { + "epoch": 0.567065472253953, + "flos": 13299532485120.0, + "grad_norm": 5.457375168117539, + "learning_rate": 1.6643018901156632e-06, + "loss": 0.7955, + "num_input_tokens_seen": 100481705, + "step": 4716 + }, + { + "epoch": 0.567185715144592, + "flos": 14888850766560.0, + "grad_norm": 7.438805986391487, + "learning_rate": 1.6635339954487566e-06, + "loss": 0.7678, + "num_input_tokens_seen": 100497300, + "step": 4717 + }, + { + "epoch": 0.5673059580352312, + "flos": 16998446814720.0, + "grad_norm": 3.4700284784121314, + "learning_rate": 1.6627661518234765e-06, + "loss": 0.8088, + "num_input_tokens_seen": 100516275, + "step": 4718 + }, + { + "epoch": 0.5674262009258703, + "flos": 15886226605920.0, + "grad_norm": 5.12351279661476, + "learning_rate": 1.661998359356302e-06, + "loss": 0.8307, + "num_input_tokens_seen": 100535025, + "step": 4719 + }, + { + "epoch": 0.5675464438165093, + "flos": 49749695960760.0, + "grad_norm": 0.8007205102726476, + "learning_rate": 1.6612306181637077e-06, + "loss": 0.5778, + "num_input_tokens_seen": 100594070, + "step": 4720 + }, + { + "epoch": 0.5676666867071485, + "flos": 13784893807320.0, + "grad_norm": 4.884365378069537, + "learning_rate": 1.6604629283621598e-06, + "loss": 0.6325, + "num_input_tokens_seen": 100611720, + "step": 4721 + }, + { + "epoch": 0.5677869295977875, + "flos": 24639953423760.0, + "grad_norm": 5.164158714887058, + "learning_rate": 1.6596952900681152e-06, + "loss": 0.728, + "num_input_tokens_seen": 100632200, + "step": 4722 + }, + { + "epoch": 0.5679071724884266, + "flos": 20562512171520.0, + "grad_norm": 7.330626884185057, + "learning_rate": 1.658927703398025e-06, + "loss": 0.8042, + "num_input_tokens_seen": 100651985, + "step": 4723 + }, + { + "epoch": 0.5680274153790658, + "flos": 17241808164960.0, + "grad_norm": 4.0159311140771194, + "learning_rate": 1.6581601684683309e-06, + "loss": 0.7398, + "num_input_tokens_seen": 100672130, + "step": 4724 + }, + { + "epoch": 0.5681476582697048, + "flos": 16431416269800.0, + "grad_norm": 7.644269265279557, + "learning_rate": 1.6573926853954674e-06, + "loss": 0.6736, + "num_input_tokens_seen": 100689435, + "step": 4725 + }, + { + "epoch": 0.5682679011603439, + "flos": 14271806467560.0, + "grad_norm": 7.902650518869518, + "learning_rate": 1.6566252542958608e-06, + "loss": 0.8112, + "num_input_tokens_seen": 100708655, + "step": 4726 + }, + { + "epoch": 0.568388144050983, + "flos": 21101243203560.0, + "grad_norm": 3.369053157832012, + "learning_rate": 1.6558578752859305e-06, + "loss": 0.7567, + "num_input_tokens_seen": 100727335, + "step": 4727 + }, + { + "epoch": 0.5685083869416221, + "flos": 15508776121920.0, + "grad_norm": 4.831822235841713, + "learning_rate": 1.6550905484820865e-06, + "loss": 0.7683, + "num_input_tokens_seen": 100745515, + "step": 4728 + }, + { + "epoch": 0.5686286298322611, + "flos": 18186795753240.0, + "grad_norm": 6.723675860542112, + "learning_rate": 1.6543232740007328e-06, + "loss": 0.7819, + "num_input_tokens_seen": 100762350, + "step": 4729 + }, + { + "epoch": 0.5687488727229003, + "flos": 19510151964360.0, + "grad_norm": 9.341547167306063, + "learning_rate": 1.653556051958263e-06, + "loss": 0.6297, + "num_input_tokens_seen": 100781750, + "step": 4730 + }, + { + "epoch": 0.5688691156135394, + "flos": 15212298341520.0, + "grad_norm": 10.23323809882799, + "learning_rate": 1.6527888824710642e-06, + "loss": 0.7254, + "num_input_tokens_seen": 100801070, + "step": 4731 + }, + { + "epoch": 0.5689893585041784, + "flos": 18966197038920.0, + "grad_norm": 18.181052202215568, + "learning_rate": 1.6520217656555166e-06, + "loss": 0.7304, + "num_input_tokens_seen": 100820080, + "step": 4732 + }, + { + "epoch": 0.5691096013948175, + "flos": 17188090195560.0, + "grad_norm": 3.492985917887244, + "learning_rate": 1.65125470162799e-06, + "loss": 0.6879, + "num_input_tokens_seen": 100840155, + "step": 4733 + }, + { + "epoch": 0.5692298442854566, + "flos": 13190070308880.0, + "grad_norm": 4.191535485542559, + "learning_rate": 1.6504876905048485e-06, + "loss": 0.6876, + "num_input_tokens_seen": 100856835, + "step": 4734 + }, + { + "epoch": 0.5693500871760957, + "flos": 16863787801680.0, + "grad_norm": 5.585205883851969, + "learning_rate": 1.6497207324024464e-06, + "loss": 0.7151, + "num_input_tokens_seen": 100875455, + "step": 4735 + }, + { + "epoch": 0.5694703300667348, + "flos": 13866816309720.0, + "grad_norm": 7.662220647054, + "learning_rate": 1.6489538274371305e-06, + "loss": 0.8077, + "num_input_tokens_seen": 100893780, + "step": 4736 + }, + { + "epoch": 0.5695905729573739, + "flos": 16027122631080.0, + "grad_norm": 4.262311270107012, + "learning_rate": 1.6481869757252396e-06, + "loss": 0.8133, + "num_input_tokens_seen": 100911835, + "step": 4737 + }, + { + "epoch": 0.569710815848013, + "flos": 20887922664000.0, + "grad_norm": 3.6635869415033815, + "learning_rate": 1.647420177383105e-06, + "loss": 0.7037, + "num_input_tokens_seen": 100934425, + "step": 4738 + }, + { + "epoch": 0.569831058738652, + "flos": 20806000161600.0, + "grad_norm": 7.800313493346432, + "learning_rate": 1.646653432527049e-06, + "loss": 0.7157, + "num_input_tokens_seen": 100954785, + "step": 4739 + }, + { + "epoch": 0.5699513016292912, + "flos": 18944039558280.0, + "grad_norm": 5.585957304379228, + "learning_rate": 1.645886741273387e-06, + "loss": 0.7343, + "num_input_tokens_seen": 100976320, + "step": 4740 + }, + { + "epoch": 0.5700715445199303, + "flos": 13161960755760.0, + "grad_norm": 3.7953749468186246, + "learning_rate": 1.645120103738424e-06, + "loss": 0.7099, + "num_input_tokens_seen": 100993550, + "step": 4741 + }, + { + "epoch": 0.5701917874105693, + "flos": 8304200079000.0, + "grad_norm": 6.137861489024273, + "learning_rate": 1.6443535200384591e-06, + "loss": 0.8156, + "num_input_tokens_seen": 101011445, + "step": 4742 + }, + { + "epoch": 0.5703120303012085, + "flos": 15918230334120.0, + "grad_norm": 4.119425619512068, + "learning_rate": 1.6435869902897827e-06, + "loss": 0.6901, + "num_input_tokens_seen": 101029745, + "step": 4743 + }, + { + "epoch": 0.5704322731918475, + "flos": 41813516189040.0, + "grad_norm": 0.7983255460899553, + "learning_rate": 1.6428205146086764e-06, + "loss": 0.6309, + "num_input_tokens_seen": 101091445, + "step": 4744 + }, + { + "epoch": 0.5705525160824866, + "flos": 15163044426480.0, + "grad_norm": 4.699556029432919, + "learning_rate": 1.6420540931114142e-06, + "loss": 0.6931, + "num_input_tokens_seen": 101111755, + "step": 4745 + }, + { + "epoch": 0.5706727589731257, + "flos": 13705250822040.0, + "grad_norm": 8.716152009182032, + "learning_rate": 1.6412877259142616e-06, + "loss": 0.7692, + "num_input_tokens_seen": 101131395, + "step": 4746 + }, + { + "epoch": 0.5707930018637648, + "flos": 20264451393120.0, + "grad_norm": 14.90523040931228, + "learning_rate": 1.6405214131334757e-06, + "loss": 0.7208, + "num_input_tokens_seen": 101149640, + "step": 4747 + }, + { + "epoch": 0.5709132447544039, + "flos": 20236943379240.0, + "grad_norm": 5.118427435058446, + "learning_rate": 1.6397551548853052e-06, + "loss": 0.7902, + "num_input_tokens_seen": 101167525, + "step": 4748 + }, + { + "epoch": 0.571033487645043, + "flos": 15861852928080.0, + "grad_norm": 6.622076001071329, + "learning_rate": 1.6389889512859917e-06, + "loss": 0.6906, + "num_input_tokens_seen": 101186905, + "step": 4749 + }, + { + "epoch": 0.5711537305356821, + "flos": 51757649842800.0, + "grad_norm": 0.8867106500331275, + "learning_rate": 1.638222802451767e-06, + "loss": 0.6499, + "num_input_tokens_seen": 101248105, + "step": 4750 + }, + { + "epoch": 0.5712739734263211, + "flos": 18105728069760.0, + "grad_norm": 5.823946930486326, + "learning_rate": 1.6374567084988561e-06, + "loss": 0.7317, + "num_input_tokens_seen": 101269010, + "step": 4751 + }, + { + "epoch": 0.5713942163169603, + "flos": 19482738930360.0, + "grad_norm": 4.002245749428483, + "learning_rate": 1.6366906695434738e-06, + "loss": 0.7474, + "num_input_tokens_seen": 101291250, + "step": 4752 + }, + { + "epoch": 0.5715144592075994, + "flos": 15460028766240.0, + "grad_norm": 13.017560982031098, + "learning_rate": 1.6359246857018275e-06, + "loss": 0.8379, + "num_input_tokens_seen": 101308500, + "step": 4753 + }, + { + "epoch": 0.5716347020982384, + "flos": 17078976278880.0, + "grad_norm": 9.721167555022005, + "learning_rate": 1.6351587570901178e-06, + "loss": 0.7689, + "num_input_tokens_seen": 101328345, + "step": 4754 + }, + { + "epoch": 0.5717549449888776, + "flos": 12399999616680.0, + "grad_norm": 4.462350170708313, + "learning_rate": 1.634392883824534e-06, + "loss": 0.7447, + "num_input_tokens_seen": 101340065, + "step": 4755 + }, + { + "epoch": 0.5718751878795166, + "flos": 26099045086560.0, + "grad_norm": 3.9957361145179235, + "learning_rate": 1.6336270660212595e-06, + "loss": 0.6603, + "num_input_tokens_seen": 101361380, + "step": 4756 + }, + { + "epoch": 0.5719954307701557, + "flos": 28389419726760.0, + "grad_norm": 5.922382906669653, + "learning_rate": 1.6328613037964676e-06, + "loss": 0.6335, + "num_input_tokens_seen": 101384165, + "step": 4757 + }, + { + "epoch": 0.5721156736607949, + "flos": 15081533503560.0, + "grad_norm": 3.5325472989057554, + "learning_rate": 1.6320955972663241e-06, + "loss": 0.6717, + "num_input_tokens_seen": 101403480, + "step": 4758 + }, + { + "epoch": 0.5722359165514339, + "flos": 27502867441920.0, + "grad_norm": 7.681507286933909, + "learning_rate": 1.6313299465469857e-06, + "loss": 0.6349, + "num_input_tokens_seen": 101425930, + "step": 4759 + }, + { + "epoch": 0.572356159442073, + "flos": 16073368849920.0, + "grad_norm": 7.6820252315203215, + "learning_rate": 1.6305643517546014e-06, + "loss": 0.764, + "num_input_tokens_seen": 101441030, + "step": 4760 + }, + { + "epoch": 0.5724764023327121, + "flos": 13974758807880.0, + "grad_norm": 3.478247622018962, + "learning_rate": 1.629798813005311e-06, + "loss": 0.8328, + "num_input_tokens_seen": 101460470, + "step": 4761 + }, + { + "epoch": 0.5725966452233512, + "flos": 16701209195280.0, + "grad_norm": 4.80778887138226, + "learning_rate": 1.6290333304152473e-06, + "loss": 0.6971, + "num_input_tokens_seen": 101480065, + "step": 4762 + }, + { + "epoch": 0.5727168881139902, + "flos": 30522787913520.0, + "grad_norm": 5.780113974638775, + "learning_rate": 1.6282679041005314e-06, + "loss": 0.5583, + "num_input_tokens_seen": 101505375, + "step": 4763 + }, + { + "epoch": 0.5728371310046293, + "flos": 10653421602120.0, + "grad_norm": 4.654201594636883, + "learning_rate": 1.6275025341772789e-06, + "loss": 0.8541, + "num_input_tokens_seen": 101521400, + "step": 4764 + }, + { + "epoch": 0.5729573738952685, + "flos": 15729441772200.0, + "grad_norm": 9.0325187663061, + "learning_rate": 1.626737220761596e-06, + "loss": 0.8154, + "num_input_tokens_seen": 101538585, + "step": 4765 + }, + { + "epoch": 0.5730776167859075, + "flos": 17294576335560.0, + "grad_norm": 10.883710858358, + "learning_rate": 1.62597196396958e-06, + "loss": 0.7771, + "num_input_tokens_seen": 101556475, + "step": 4766 + }, + { + "epoch": 0.5731978596765466, + "flos": 18861895436160.0, + "grad_norm": 12.964350016020495, + "learning_rate": 1.6252067639173197e-06, + "loss": 0.8352, + "num_input_tokens_seen": 101578105, + "step": 4767 + }, + { + "epoch": 0.5733181025671857, + "flos": 19320951822960.0, + "grad_norm": 13.144309526361766, + "learning_rate": 1.6244416207208956e-06, + "loss": 0.6758, + "num_input_tokens_seen": 101598760, + "step": 4768 + }, + { + "epoch": 0.5734383454578248, + "flos": 21588725743080.0, + "grad_norm": 6.639932178932275, + "learning_rate": 1.6236765344963787e-06, + "loss": 0.7251, + "num_input_tokens_seen": 101619740, + "step": 4769 + }, + { + "epoch": 0.5735585883484638, + "flos": 25693231769760.0, + "grad_norm": 10.02399388922454, + "learning_rate": 1.6229115053598322e-06, + "loss": 0.6742, + "num_input_tokens_seen": 101641215, + "step": 4770 + }, + { + "epoch": 0.573678831239103, + "flos": 13704807582600.0, + "grad_norm": 4.675966442395642, + "learning_rate": 1.6221465334273108e-06, + "loss": 0.6904, + "num_input_tokens_seen": 101660145, + "step": 4771 + }, + { + "epoch": 0.5737990741297421, + "flos": 18835020621480.0, + "grad_norm": 9.28777579595028, + "learning_rate": 1.6213816188148593e-06, + "loss": 0.5976, + "num_input_tokens_seen": 101678570, + "step": 4772 + }, + { + "epoch": 0.5739193170203811, + "flos": 19993866968640.0, + "grad_norm": 2.6954557076708534, + "learning_rate": 1.6206167616385162e-06, + "loss": 0.7601, + "num_input_tokens_seen": 101699355, + "step": 4773 + }, + { + "epoch": 0.5740395599110203, + "flos": 8871388923720.0, + "grad_norm": 3.895816632947299, + "learning_rate": 1.6198519620143078e-06, + "loss": 0.73, + "num_input_tokens_seen": 101716230, + "step": 4774 + }, + { + "epoch": 0.5741598028016593, + "flos": 18996681089040.0, + "grad_norm": 3.443819652482999, + "learning_rate": 1.6190872200582546e-06, + "loss": 0.761, + "num_input_tokens_seen": 101737690, + "step": 4775 + }, + { + "epoch": 0.5742800456922984, + "flos": 14055003332400.0, + "grad_norm": 21.938094693461803, + "learning_rate": 1.6183225358863676e-06, + "loss": 0.761, + "num_input_tokens_seen": 101754305, + "step": 4776 + }, + { + "epoch": 0.5744002885829376, + "flos": 22696418577600.0, + "grad_norm": 8.04188445064825, + "learning_rate": 1.617557909614648e-06, + "loss": 0.7044, + "num_input_tokens_seen": 101773460, + "step": 4777 + }, + { + "epoch": 0.5745205314735766, + "flos": 17456078503320.0, + "grad_norm": 3.37435581982569, + "learning_rate": 1.6167933413590899e-06, + "loss": 0.8392, + "num_input_tokens_seen": 101792085, + "step": 4778 + }, + { + "epoch": 0.5746407743642157, + "flos": 8924695313640.0, + "grad_norm": 4.759498735849474, + "learning_rate": 1.6160288312356773e-06, + "loss": 0.8903, + "num_input_tokens_seen": 101808935, + "step": 4779 + }, + { + "epoch": 0.5747610172548548, + "flos": 17673641477520.0, + "grad_norm": 4.524358520562115, + "learning_rate": 1.6152643793603857e-06, + "loss": 0.8134, + "num_input_tokens_seen": 101829005, + "step": 4780 + }, + { + "epoch": 0.5748812601454939, + "flos": 18616761128160.0, + "grad_norm": 3.670242790901583, + "learning_rate": 1.6144999858491815e-06, + "loss": 0.869, + "num_input_tokens_seen": 101847355, + "step": 4781 + }, + { + "epoch": 0.575001503036133, + "flos": 22477177625520.0, + "grad_norm": 3.6752350712545576, + "learning_rate": 1.6137356508180232e-06, + "loss": 0.8419, + "num_input_tokens_seen": 101868785, + "step": 4782 + }, + { + "epoch": 0.5751217459267721, + "flos": 15892368638160.0, + "grad_norm": 15.293076508241043, + "learning_rate": 1.6129713743828593e-06, + "loss": 0.8035, + "num_input_tokens_seen": 101887515, + "step": 4783 + }, + { + "epoch": 0.5752419888174112, + "flos": 15836624431320.0, + "grad_norm": 4.43445450413775, + "learning_rate": 1.6122071566596306e-06, + "loss": 0.7445, + "num_input_tokens_seen": 101907510, + "step": 4784 + }, + { + "epoch": 0.5753622317080502, + "flos": 12967251781320.0, + "grad_norm": 6.9643327201402725, + "learning_rate": 1.6114429977642674e-06, + "loss": 0.8167, + "num_input_tokens_seen": 101921735, + "step": 4785 + }, + { + "epoch": 0.5754824745986894, + "flos": 14458442152200.0, + "grad_norm": 4.048493168579534, + "learning_rate": 1.6106788978126926e-06, + "loss": 0.7108, + "num_input_tokens_seen": 101940430, + "step": 4786 + }, + { + "epoch": 0.5756027174893285, + "flos": 22743551275320.0, + "grad_norm": 5.209462198627204, + "learning_rate": 1.6099148569208196e-06, + "loss": 0.7607, + "num_input_tokens_seen": 101957370, + "step": 4787 + }, + { + "epoch": 0.5757229603799675, + "flos": 20939487756120.0, + "grad_norm": 3.06357073069127, + "learning_rate": 1.6091508752045523e-06, + "loss": 0.6138, + "num_input_tokens_seen": 101977970, + "step": 4788 + }, + { + "epoch": 0.5758432032706067, + "flos": 16833430391400.0, + "grad_norm": 3.617995104017261, + "learning_rate": 1.608386952779787e-06, + "loss": 0.8535, + "num_input_tokens_seen": 101997060, + "step": 4789 + }, + { + "epoch": 0.5759634461612457, + "flos": 18862338675600.0, + "grad_norm": 3.105356658256115, + "learning_rate": 1.6076230897624098e-06, + "loss": 0.7295, + "num_input_tokens_seen": 102018985, + "step": 4790 + }, + { + "epoch": 0.5760836890518848, + "flos": 22453088887320.0, + "grad_norm": 5.214537056465975, + "learning_rate": 1.6068592862682974e-06, + "loss": 0.766, + "num_input_tokens_seen": 102036860, + "step": 4791 + }, + { + "epoch": 0.576203931942524, + "flos": 26854420953960.0, + "grad_norm": 5.087037513950036, + "learning_rate": 1.6060955424133187e-06, + "loss": 0.7166, + "num_input_tokens_seen": 102057505, + "step": 4792 + }, + { + "epoch": 0.576324174833163, + "flos": 18696910672800.0, + "grad_norm": 9.140955522870339, + "learning_rate": 1.6053318583133332e-06, + "loss": 0.8748, + "num_input_tokens_seen": 102078095, + "step": 4793 + }, + { + "epoch": 0.5764444177238021, + "flos": 18403123989000.0, + "grad_norm": 4.849415474501637, + "learning_rate": 1.6045682340841907e-06, + "loss": 0.7416, + "num_input_tokens_seen": 102096740, + "step": 4794 + }, + { + "epoch": 0.5765646606144411, + "flos": 45858130554120.0, + "grad_norm": 0.7901426042340631, + "learning_rate": 1.6038046698417336e-06, + "loss": 0.6017, + "num_input_tokens_seen": 102157355, + "step": 4795 + }, + { + "epoch": 0.5766849035050803, + "flos": 18402680749560.0, + "grad_norm": 4.6041244343795675, + "learning_rate": 1.6030411657017919e-06, + "loss": 0.6632, + "num_input_tokens_seen": 102176730, + "step": 4796 + }, + { + "epoch": 0.5768051463957193, + "flos": 11646713306640.0, + "grad_norm": 4.2595077079759625, + "learning_rate": 1.6022777217801903e-06, + "loss": 0.8338, + "num_input_tokens_seen": 102193405, + "step": 4797 + }, + { + "epoch": 0.5769253892863584, + "flos": 16215721233240.0, + "grad_norm": 5.260253702289575, + "learning_rate": 1.601514338192742e-06, + "loss": 0.7104, + "num_input_tokens_seen": 102213055, + "step": 4798 + }, + { + "epoch": 0.5770456321769976, + "flos": 16728337289640.0, + "grad_norm": 3.9488536088615325, + "learning_rate": 1.6007510150552514e-06, + "loss": 0.699, + "num_input_tokens_seen": 102230835, + "step": 4799 + }, + { + "epoch": 0.5771658750676366, + "flos": 34117970519640.0, + "grad_norm": 3.686835794921024, + "learning_rate": 1.599987752483515e-06, + "loss": 0.6059, + "num_input_tokens_seen": 102255000, + "step": 4800 + }, + { + "epoch": 0.5772861179582757, + "flos": 16212491917320.0, + "grad_norm": 4.074615956563446, + "learning_rate": 1.5992245505933184e-06, + "loss": 0.6725, + "num_input_tokens_seen": 102274420, + "step": 4801 + }, + { + "epoch": 0.5774063608489148, + "flos": 22939494988200.0, + "grad_norm": 3.992586537703017, + "learning_rate": 1.5984614095004388e-06, + "loss": 0.6991, + "num_input_tokens_seen": 102295275, + "step": 4802 + }, + { + "epoch": 0.5775266037395539, + "flos": 16484532699960.0, + "grad_norm": 7.600145699961582, + "learning_rate": 1.5976983293206438e-06, + "loss": 0.7863, + "num_input_tokens_seen": 102310800, + "step": 4803 + }, + { + "epoch": 0.577646846630193, + "flos": 15405424317960.0, + "grad_norm": 4.2896543556523055, + "learning_rate": 1.5969353101696928e-06, + "loss": 0.7031, + "num_input_tokens_seen": 102328960, + "step": 4804 + }, + { + "epoch": 0.5777670895208321, + "flos": 21803439320880.0, + "grad_norm": 4.943529913593301, + "learning_rate": 1.5961723521633341e-06, + "loss": 0.7855, + "num_input_tokens_seen": 102349920, + "step": 4805 + }, + { + "epoch": 0.5778873324114712, + "flos": 14244805013040.0, + "grad_norm": 3.6474686506360885, + "learning_rate": 1.5954094554173097e-06, + "loss": 0.8934, + "num_input_tokens_seen": 102367630, + "step": 4806 + }, + { + "epoch": 0.5780075753021102, + "flos": 10275369578880.0, + "grad_norm": 4.7804283483262715, + "learning_rate": 1.5946466200473482e-06, + "loss": 0.779, + "num_input_tokens_seen": 102385260, + "step": 4807 + }, + { + "epoch": 0.5781278181927494, + "flos": 11107380735360.0, + "grad_norm": 4.079772382632614, + "learning_rate": 1.5938838461691723e-06, + "loss": 0.8283, + "num_input_tokens_seen": 102401890, + "step": 4808 + }, + { + "epoch": 0.5782480610833884, + "flos": 12192061270320.0, + "grad_norm": 5.611712210575857, + "learning_rate": 1.593121133898494e-06, + "loss": 0.8037, + "num_input_tokens_seen": 102418815, + "step": 4809 + }, + { + "epoch": 0.5783683039740275, + "flos": 18671397236400.0, + "grad_norm": 3.1739272821049074, + "learning_rate": 1.592358483351016e-06, + "loss": 0.7645, + "num_input_tokens_seen": 102438710, + "step": 4810 + }, + { + "epoch": 0.5784885468646667, + "flos": 13434729717480.0, + "grad_norm": 3.3735397853424782, + "learning_rate": 1.5915958946424326e-06, + "loss": 0.7109, + "num_input_tokens_seen": 102457115, + "step": 4811 + }, + { + "epoch": 0.5786087897553057, + "flos": 33931113215280.0, + "grad_norm": 2.6536914845221995, + "learning_rate": 1.5908333678884271e-06, + "loss": 0.7346, + "num_input_tokens_seen": 102483255, + "step": 4812 + }, + { + "epoch": 0.5787290326459448, + "flos": 8979648021480.0, + "grad_norm": 5.519448141348389, + "learning_rate": 1.5900709032046743e-06, + "loss": 0.718, + "num_input_tokens_seen": 102501050, + "step": 4813 + }, + { + "epoch": 0.5788492755365839, + "flos": 17048935468200.0, + "grad_norm": 3.4593397460905506, + "learning_rate": 1.5893085007068391e-06, + "loss": 0.7748, + "num_input_tokens_seen": 102518330, + "step": 4814 + }, + { + "epoch": 0.578969518427223, + "flos": 17619955168080.0, + "grad_norm": 9.970121761638346, + "learning_rate": 1.5885461605105786e-06, + "loss": 0.7, + "num_input_tokens_seen": 102539650, + "step": 4815 + }, + { + "epoch": 0.579089761317862, + "flos": 15999772917000.0, + "grad_norm": 4.472052787482858, + "learning_rate": 1.5877838827315375e-06, + "loss": 0.7529, + "num_input_tokens_seen": 102557915, + "step": 4816 + }, + { + "epoch": 0.5792100042085012, + "flos": 16782625138320.0, + "grad_norm": 4.889089910150572, + "learning_rate": 1.587021667485355e-06, + "loss": 0.6791, + "num_input_tokens_seen": 102577005, + "step": 4817 + }, + { + "epoch": 0.5793302470991403, + "flos": 15702376997760.0, + "grad_norm": 8.581658752441708, + "learning_rate": 1.5862595148876559e-06, + "loss": 0.7648, + "num_input_tokens_seen": 102596830, + "step": 4818 + }, + { + "epoch": 0.5794504899897793, + "flos": 9219748395840.0, + "grad_norm": 3.982808601793527, + "learning_rate": 1.58549742505406e-06, + "loss": 0.7608, + "num_input_tokens_seen": 102611295, + "step": 4819 + }, + { + "epoch": 0.5795707328804185, + "flos": 10815841908720.0, + "grad_norm": 5.343613293778589, + "learning_rate": 1.5847353981001747e-06, + "loss": 0.737, + "num_input_tokens_seen": 102628195, + "step": 4820 + }, + { + "epoch": 0.5796909757710575, + "flos": 26773891489800.0, + "grad_norm": 3.608869508958509, + "learning_rate": 1.5839734341415993e-06, + "loss": 0.6821, + "num_input_tokens_seen": 102650115, + "step": 4821 + }, + { + "epoch": 0.5798112186616966, + "flos": 16864294361040.0, + "grad_norm": 3.460135545666494, + "learning_rate": 1.5832115332939238e-06, + "loss": 0.7599, + "num_input_tokens_seen": 102668275, + "step": 4822 + }, + { + "epoch": 0.5799314615523358, + "flos": 12136950262680.0, + "grad_norm": 3.8303709331598297, + "learning_rate": 1.5824496956727272e-06, + "loss": 0.729, + "num_input_tokens_seen": 102685200, + "step": 4823 + }, + { + "epoch": 0.5800517044429748, + "flos": 14973970924920.0, + "grad_norm": 3.400934954727035, + "learning_rate": 1.5816879213935797e-06, + "loss": 0.7057, + "num_input_tokens_seen": 102703730, + "step": 4824 + }, + { + "epoch": 0.5801719473336139, + "flos": 23153258767200.0, + "grad_norm": 4.121668770976338, + "learning_rate": 1.5809262105720416e-06, + "loss": 0.7801, + "num_input_tokens_seen": 102724490, + "step": 4825 + }, + { + "epoch": 0.580292190224253, + "flos": 14892744941640.0, + "grad_norm": 3.3044828888156745, + "learning_rate": 1.5801645633236644e-06, + "loss": 0.7759, + "num_input_tokens_seen": 102745195, + "step": 4826 + }, + { + "epoch": 0.5804124331148921, + "flos": 19510373584080.0, + "grad_norm": 4.836455728745305, + "learning_rate": 1.579402979763989e-06, + "loss": 0.7646, + "num_input_tokens_seen": 102765250, + "step": 4827 + }, + { + "epoch": 0.5805326760055312, + "flos": 9787918699320.0, + "grad_norm": 7.085079321999397, + "learning_rate": 1.578641460008548e-06, + "loss": 0.7754, + "num_input_tokens_seen": 102782705, + "step": 4828 + }, + { + "epoch": 0.5806529188961702, + "flos": 8761261888320.0, + "grad_norm": 3.869588210797446, + "learning_rate": 1.5778800041728613e-06, + "loss": 0.6438, + "num_input_tokens_seen": 102798715, + "step": 4829 + }, + { + "epoch": 0.5807731617868094, + "flos": 19213420904280.0, + "grad_norm": 10.023163099688103, + "learning_rate": 1.577118612372443e-06, + "loss": 0.6488, + "num_input_tokens_seen": 102820275, + "step": 4830 + }, + { + "epoch": 0.5808934046774484, + "flos": 27907699299960.0, + "grad_norm": 3.8678744852240574, + "learning_rate": 1.5763572847227943e-06, + "loss": 0.6841, + "num_input_tokens_seen": 102840880, + "step": 4831 + }, + { + "epoch": 0.5810136475680875, + "flos": 14974129224720.0, + "grad_norm": 4.168709800789477, + "learning_rate": 1.5755960213394091e-06, + "loss": 0.7854, + "num_input_tokens_seen": 102857700, + "step": 4832 + }, + { + "epoch": 0.5811338904587267, + "flos": 12786663149040.0, + "grad_norm": 15.004358294251228, + "learning_rate": 1.5748348223377703e-06, + "loss": 0.768, + "num_input_tokens_seen": 102874975, + "step": 4833 + }, + { + "epoch": 0.5812541333493657, + "flos": 14212579665120.0, + "grad_norm": 4.612034537961062, + "learning_rate": 1.5740736878333507e-06, + "loss": 0.7622, + "num_input_tokens_seen": 102892535, + "step": 4834 + }, + { + "epoch": 0.5813743762400048, + "flos": 15054816988680.0, + "grad_norm": 20.143509781766557, + "learning_rate": 1.5733126179416143e-06, + "loss": 0.762, + "num_input_tokens_seen": 102906740, + "step": 4835 + }, + { + "epoch": 0.5814946191306439, + "flos": 24366677902680.0, + "grad_norm": 5.10228066406127, + "learning_rate": 1.5725516127780137e-06, + "loss": 0.7011, + "num_input_tokens_seen": 102928595, + "step": 4836 + }, + { + "epoch": 0.581614862021283, + "flos": 11760227957760.0, + "grad_norm": 8.274598873131584, + "learning_rate": 1.5717906724579943e-06, + "loss": 0.8619, + "num_input_tokens_seen": 102945375, + "step": 4837 + }, + { + "epoch": 0.581735104911922, + "flos": 24829375184880.0, + "grad_norm": 4.84507073885176, + "learning_rate": 1.571029797096989e-06, + "loss": 0.6602, + "num_input_tokens_seen": 102966200, + "step": 4838 + }, + { + "epoch": 0.5818553478025612, + "flos": 17079767777880.0, + "grad_norm": 3.4849625848981, + "learning_rate": 1.570268986810423e-06, + "loss": 0.7678, + "num_input_tokens_seen": 102985815, + "step": 4839 + }, + { + "epoch": 0.5819755906932003, + "flos": 15351706348560.0, + "grad_norm": 5.720801147162944, + "learning_rate": 1.5695082417137096e-06, + "loss": 0.7403, + "num_input_tokens_seen": 103003410, + "step": 4840 + }, + { + "epoch": 0.5820958335838393, + "flos": 15673697565360.0, + "grad_norm": 3.916401847381826, + "learning_rate": 1.5687475619222539e-06, + "loss": 0.7415, + "num_input_tokens_seen": 103023085, + "step": 4841 + }, + { + "epoch": 0.5822160764744785, + "flos": 13110237363840.0, + "grad_norm": 4.284279746701661, + "learning_rate": 1.5679869475514496e-06, + "loss": 0.7218, + "num_input_tokens_seen": 103039740, + "step": 4842 + }, + { + "epoch": 0.5823363193651175, + "flos": 16858975487760.0, + "grad_norm": 10.653832303975433, + "learning_rate": 1.567226398716682e-06, + "loss": 0.7853, + "num_input_tokens_seen": 103059375, + "step": 4843 + }, + { + "epoch": 0.5824565622557566, + "flos": 24155510240400.0, + "grad_norm": 3.50829463100757, + "learning_rate": 1.566465915533326e-06, + "loss": 0.6177, + "num_input_tokens_seen": 103081125, + "step": 4844 + }, + { + "epoch": 0.5825768051463958, + "flos": 16263613770000.0, + "grad_norm": 5.294964020834664, + "learning_rate": 1.5657054981167458e-06, + "loss": 0.8601, + "num_input_tokens_seen": 103099740, + "step": 4845 + }, + { + "epoch": 0.5826970480370348, + "flos": 20536207236120.0, + "grad_norm": 4.258700496922195, + "learning_rate": 1.5649451465822965e-06, + "loss": 0.661, + "num_input_tokens_seen": 103120850, + "step": 4846 + }, + { + "epoch": 0.5828172909276739, + "flos": 13029264660240.0, + "grad_norm": 3.1281217622692568, + "learning_rate": 1.5641848610453218e-06, + "loss": 0.8161, + "num_input_tokens_seen": 103139230, + "step": 4847 + }, + { + "epoch": 0.582937533818313, + "flos": 14514787898280.0, + "grad_norm": 4.340980580034911, + "learning_rate": 1.563424641621158e-06, + "loss": 0.8375, + "num_input_tokens_seen": 103158130, + "step": 4848 + }, + { + "epoch": 0.5830577767089521, + "flos": 19699352105760.0, + "grad_norm": 4.395988690612518, + "learning_rate": 1.5626644884251282e-06, + "loss": 0.6823, + "num_input_tokens_seen": 103177370, + "step": 4849 + }, + { + "epoch": 0.5831780195995911, + "flos": 18535060245480.0, + "grad_norm": 4.8187090959804335, + "learning_rate": 1.5619044015725488e-06, + "loss": 0.8678, + "num_input_tokens_seen": 103196780, + "step": 4850 + }, + { + "epoch": 0.5832982624902303, + "flos": 10734299325840.0, + "grad_norm": 6.571112802868139, + "learning_rate": 1.5611443811787224e-06, + "loss": 0.8474, + "num_input_tokens_seen": 103210625, + "step": 4851 + }, + { + "epoch": 0.5834185053808694, + "flos": 14942885335560.0, + "grad_norm": 4.0611894107430695, + "learning_rate": 1.560384427358945e-06, + "loss": 0.69, + "num_input_tokens_seen": 103229890, + "step": 4852 + }, + { + "epoch": 0.5835387482715084, + "flos": 19943504955000.0, + "grad_norm": 2.6948308099637415, + "learning_rate": 1.5596245402284998e-06, + "loss": 0.7218, + "num_input_tokens_seen": 103253135, + "step": 4853 + }, + { + "epoch": 0.5836589911621476, + "flos": 12138723220440.0, + "grad_norm": 3.1569604357372074, + "learning_rate": 1.5588647199026619e-06, + "loss": 0.7976, + "num_input_tokens_seen": 103270590, + "step": 4854 + }, + { + "epoch": 0.5837792340527866, + "flos": 14944879913040.0, + "grad_norm": 8.052676344761197, + "learning_rate": 1.5581049664966956e-06, + "loss": 0.8611, + "num_input_tokens_seen": 103288070, + "step": 4855 + }, + { + "epoch": 0.5838994769434257, + "flos": 48655992337200.0, + "grad_norm": 1.0787696017079045, + "learning_rate": 1.5573452801258545e-06, + "loss": 0.6834, + "num_input_tokens_seen": 103334960, + "step": 4856 + }, + { + "epoch": 0.5840197198340649, + "flos": 15702788577240.0, + "grad_norm": 4.952248673279349, + "learning_rate": 1.5565856609053824e-06, + "loss": 0.6133, + "num_input_tokens_seen": 103353475, + "step": 4857 + }, + { + "epoch": 0.5841399627247039, + "flos": 13974410548320.0, + "grad_norm": 3.9766409216082335, + "learning_rate": 1.5558261089505127e-06, + "loss": 0.7826, + "num_input_tokens_seen": 103371925, + "step": 4858 + }, + { + "epoch": 0.584260205615343, + "flos": 19368970999560.0, + "grad_norm": 5.312246698062798, + "learning_rate": 1.5550666243764697e-06, + "loss": 0.7836, + "num_input_tokens_seen": 103389805, + "step": 4859 + }, + { + "epoch": 0.584380448505982, + "flos": 10086675996840.0, + "grad_norm": 13.17715062462232, + "learning_rate": 1.554307207298465e-06, + "loss": 0.7494, + "num_input_tokens_seen": 103407785, + "step": 4860 + }, + { + "epoch": 0.5845006913966212, + "flos": 15756981446040.0, + "grad_norm": 3.044037265263757, + "learning_rate": 1.553547857831704e-06, + "loss": 0.777, + "num_input_tokens_seen": 103424015, + "step": 4861 + }, + { + "epoch": 0.5846209342872603, + "flos": 43017437336520.0, + "grad_norm": 1.01126418340707, + "learning_rate": 1.5527885760913771e-06, + "loss": 0.7125, + "num_input_tokens_seen": 103473625, + "step": 4862 + }, + { + "epoch": 0.5847411771778993, + "flos": 13569547030320.0, + "grad_norm": 4.007131418280307, + "learning_rate": 1.552029362192668e-06, + "loss": 0.748, + "num_input_tokens_seen": 103492605, + "step": 4863 + }, + { + "epoch": 0.5848614200685385, + "flos": 17752587943680.0, + "grad_norm": 6.661722861479176, + "learning_rate": 1.5512702162507478e-06, + "loss": 0.6959, + "num_input_tokens_seen": 103512640, + "step": 4864 + }, + { + "epoch": 0.5849816629591775, + "flos": 52849639337160.0, + "grad_norm": 0.9861840473358543, + "learning_rate": 1.5505111383807792e-06, + "loss": 0.5626, + "num_input_tokens_seen": 103575030, + "step": 4865 + }, + { + "epoch": 0.5851019058498166, + "flos": 17427810650400.0, + "grad_norm": 3.1427691527843518, + "learning_rate": 1.5497521286979138e-06, + "loss": 0.7875, + "num_input_tokens_seen": 103594990, + "step": 4866 + }, + { + "epoch": 0.5852221487404557, + "flos": 17862176759760.0, + "grad_norm": 4.04983518288645, + "learning_rate": 1.5489931873172927e-06, + "loss": 0.7354, + "num_input_tokens_seen": 103616030, + "step": 4867 + }, + { + "epoch": 0.5853423916310948, + "flos": 19995798226200.0, + "grad_norm": 3.8851867416845267, + "learning_rate": 1.5482343143540467e-06, + "loss": 0.7779, + "num_input_tokens_seen": 103637015, + "step": 4868 + }, + { + "epoch": 0.5854626345217339, + "flos": 8681903842680.0, + "grad_norm": 4.03715756502443, + "learning_rate": 1.547475509923295e-06, + "loss": 0.8151, + "num_input_tokens_seen": 103653775, + "step": 4869 + }, + { + "epoch": 0.585582877412373, + "flos": 47436119061240.0, + "grad_norm": 0.7615819563667641, + "learning_rate": 1.5467167741401495e-06, + "loss": 0.5858, + "num_input_tokens_seen": 103714975, + "step": 4870 + }, + { + "epoch": 0.5857031203030121, + "flos": 12402089174040.0, + "grad_norm": 5.717880433056337, + "learning_rate": 1.5459581071197083e-06, + "loss": 0.7008, + "num_input_tokens_seen": 103730355, + "step": 4871 + }, + { + "epoch": 0.5858233631936511, + "flos": 15269815506120.0, + "grad_norm": 7.694626149198396, + "learning_rate": 1.5451995089770624e-06, + "loss": 0.805, + "num_input_tokens_seen": 103749860, + "step": 4872 + }, + { + "epoch": 0.5859436060842903, + "flos": 16977650712360.0, + "grad_norm": 2.838888687246589, + "learning_rate": 1.5444409798272885e-06, + "loss": 0.7045, + "num_input_tokens_seen": 103773670, + "step": 4873 + }, + { + "epoch": 0.5860638489749294, + "flos": 16459304203200.0, + "grad_norm": 5.400495520568069, + "learning_rate": 1.543682519785456e-06, + "loss": 0.7844, + "num_input_tokens_seen": 103791870, + "step": 4874 + }, + { + "epoch": 0.5861840918655684, + "flos": 12813316344000.0, + "grad_norm": 17.289668814210355, + "learning_rate": 1.5429241289666219e-06, + "loss": 0.7711, + "num_input_tokens_seen": 103809090, + "step": 4875 + }, + { + "epoch": 0.5863043347562076, + "flos": 18726318284280.0, + "grad_norm": 7.058772360859333, + "learning_rate": 1.5421658074858342e-06, + "loss": 0.682, + "num_input_tokens_seen": 103826915, + "step": 4876 + }, + { + "epoch": 0.5864245776468466, + "flos": 15108471638160.0, + "grad_norm": 17.472188594093705, + "learning_rate": 1.5414075554581298e-06, + "loss": 0.6371, + "num_input_tokens_seen": 103844680, + "step": 4877 + }, + { + "epoch": 0.5865448205374857, + "flos": 21211433558880.0, + "grad_norm": 8.941851216948042, + "learning_rate": 1.5406493729985348e-06, + "loss": 0.7564, + "num_input_tokens_seen": 103863595, + "step": 4878 + }, + { + "epoch": 0.5866650634281249, + "flos": 18937517606520.0, + "grad_norm": 4.430475181772685, + "learning_rate": 1.5398912602220644e-06, + "loss": 0.7017, + "num_input_tokens_seen": 103882590, + "step": 4879 + }, + { + "epoch": 0.5867853063187639, + "flos": 12432541564200.0, + "grad_norm": 4.698681731706154, + "learning_rate": 1.539133217243724e-06, + "loss": 0.759, + "num_input_tokens_seen": 103899330, + "step": 4880 + }, + { + "epoch": 0.586905549209403, + "flos": 18052168400160.0, + "grad_norm": 4.354964455460594, + "learning_rate": 1.5383752441785081e-06, + "loss": 0.7302, + "num_input_tokens_seen": 103918275, + "step": 4881 + }, + { + "epoch": 0.5870257921000421, + "flos": 10707329531280.0, + "grad_norm": 4.055096026881663, + "learning_rate": 1.5376173411414003e-06, + "loss": 0.8369, + "num_input_tokens_seen": 103936035, + "step": 4882 + }, + { + "epoch": 0.5871460349906812, + "flos": 17511854370120.0, + "grad_norm": 3.299286019508674, + "learning_rate": 1.5368595082473753e-06, + "loss": 0.769, + "num_input_tokens_seen": 103954055, + "step": 4883 + }, + { + "epoch": 0.5872662778813202, + "flos": 16216196132640.0, + "grad_norm": 4.025861583924604, + "learning_rate": 1.5361017456113935e-06, + "loss": 0.7711, + "num_input_tokens_seen": 103974125, + "step": 4884 + }, + { + "epoch": 0.5873865207719594, + "flos": 13461382912440.0, + "grad_norm": 5.264120820951788, + "learning_rate": 1.5353440533484085e-06, + "loss": 0.839, + "num_input_tokens_seen": 103992700, + "step": 4885 + }, + { + "epoch": 0.5875067636625985, + "flos": 39788054349120.0, + "grad_norm": 3.7567625740172375, + "learning_rate": 1.534586431573361e-06, + "loss": 0.6485, + "num_input_tokens_seen": 104017360, + "step": 4886 + }, + { + "epoch": 0.5876270065532375, + "flos": 20531046662640.0, + "grad_norm": 9.265028833055176, + "learning_rate": 1.5338288804011817e-06, + "loss": 0.7611, + "num_input_tokens_seen": 104036580, + "step": 4887 + }, + { + "epoch": 0.5877472494438767, + "flos": 15622132473240.0, + "grad_norm": 5.401200775579004, + "learning_rate": 1.533071399946791e-06, + "loss": 0.6813, + "num_input_tokens_seen": 104055045, + "step": 4888 + }, + { + "epoch": 0.5878674923345157, + "flos": 16378141539840.0, + "grad_norm": 5.896291582431183, + "learning_rate": 1.5323139903250977e-06, + "loss": 0.5557, + "num_input_tokens_seen": 104075370, + "step": 4889 + }, + { + "epoch": 0.5879877352251548, + "flos": 15996986840520.0, + "grad_norm": 4.078520189335153, + "learning_rate": 1.5315566516510002e-06, + "loss": 0.7591, + "num_input_tokens_seen": 104093260, + "step": 4890 + }, + { + "epoch": 0.5881079781157939, + "flos": 12759788334360.0, + "grad_norm": 4.217028005042473, + "learning_rate": 1.5307993840393857e-06, + "loss": 0.6664, + "num_input_tokens_seen": 104111060, + "step": 4891 + }, + { + "epoch": 0.588228221006433, + "flos": 16540086947040.0, + "grad_norm": 4.752802373506743, + "learning_rate": 1.530042187605132e-06, + "loss": 0.7936, + "num_input_tokens_seen": 104130035, + "step": 4892 + }, + { + "epoch": 0.5883484638970721, + "flos": 19185817910520.0, + "grad_norm": 3.8979008674131483, + "learning_rate": 1.5292850624631044e-06, + "loss": 0.8262, + "num_input_tokens_seen": 104151950, + "step": 4893 + }, + { + "epoch": 0.5884687067877111, + "flos": 22343246791560.0, + "grad_norm": 3.8756079658290212, + "learning_rate": 1.5285280087281593e-06, + "loss": 0.7673, + "num_input_tokens_seen": 104172400, + "step": 4894 + }, + { + "epoch": 0.5885889496783503, + "flos": 51998668356000.0, + "grad_norm": 0.670668405690634, + "learning_rate": 1.5277710265151398e-06, + "loss": 0.5623, + "num_input_tokens_seen": 104241600, + "step": 4895 + }, + { + "epoch": 0.5887091925689893, + "flos": 13947504073680.0, + "grad_norm": 10.789530617560043, + "learning_rate": 1.5270141159388803e-06, + "loss": 0.7376, + "num_input_tokens_seen": 104258340, + "step": 4896 + }, + { + "epoch": 0.5888294354596284, + "flos": 17052418063800.0, + "grad_norm": 3.210419030573373, + "learning_rate": 1.526257277114203e-06, + "loss": 0.7789, + "num_input_tokens_seen": 104279135, + "step": 4897 + }, + { + "epoch": 0.5889496783502676, + "flos": 16080998900280.0, + "grad_norm": 3.4760527847900247, + "learning_rate": 1.5255005101559201e-06, + "loss": 0.7843, + "num_input_tokens_seen": 104296465, + "step": 4898 + }, + { + "epoch": 0.5890699212409066, + "flos": 15861093089040.0, + "grad_norm": 3.3690177740283223, + "learning_rate": 1.524743815178833e-06, + "loss": 0.7511, + "num_input_tokens_seen": 104314145, + "step": 4899 + }, + { + "epoch": 0.5891901641315457, + "flos": 14218341777840.0, + "grad_norm": 4.426364089284039, + "learning_rate": 1.5239871922977315e-06, + "loss": 0.7944, + "num_input_tokens_seen": 104333780, + "step": 4900 + }, + { + "epoch": 0.5893104070221848, + "flos": 14325809376600.0, + "grad_norm": 4.430083618752656, + "learning_rate": 1.523230641627394e-06, + "loss": 0.8754, + "num_input_tokens_seen": 104352485, + "step": 4901 + }, + { + "epoch": 0.5894306499128239, + "flos": 21319629336720.0, + "grad_norm": 15.934830026812286, + "learning_rate": 1.5224741632825888e-06, + "loss": 0.7005, + "num_input_tokens_seen": 104372395, + "step": 4902 + }, + { + "epoch": 0.589550892803463, + "flos": 31116471653400.0, + "grad_norm": 6.226772022431521, + "learning_rate": 1.521717757378074e-06, + "loss": 0.6763, + "num_input_tokens_seen": 104392660, + "step": 4903 + }, + { + "epoch": 0.5896711356941021, + "flos": 10275559538640.0, + "grad_norm": 4.538550448137388, + "learning_rate": 1.5209614240285943e-06, + "loss": 0.6688, + "num_input_tokens_seen": 104410035, + "step": 4904 + }, + { + "epoch": 0.5897913785847412, + "flos": 12543270138840.0, + "grad_norm": 4.841110002269846, + "learning_rate": 1.520205163348887e-06, + "loss": 0.8287, + "num_input_tokens_seen": 104427690, + "step": 4905 + }, + { + "epoch": 0.5899116214753802, + "flos": 35928687121800.0, + "grad_norm": 0.7475496636672696, + "learning_rate": 1.519448975453674e-06, + "loss": 0.5718, + "num_input_tokens_seen": 104482510, + "step": 4906 + }, + { + "epoch": 0.5900318643660194, + "flos": 15431001074280.0, + "grad_norm": 13.439825380190026, + "learning_rate": 1.5186928604576696e-06, + "loss": 0.7448, + "num_input_tokens_seen": 104499425, + "step": 4907 + }, + { + "epoch": 0.5901521072566585, + "flos": 15486555321360.0, + "grad_norm": 4.378572585693149, + "learning_rate": 1.5179368184755752e-06, + "loss": 0.7604, + "num_input_tokens_seen": 104517230, + "step": 4908 + }, + { + "epoch": 0.5902723501472975, + "flos": 14781161548080.0, + "grad_norm": 4.5474871980341165, + "learning_rate": 1.5171808496220821e-06, + "loss": 0.8116, + "num_input_tokens_seen": 104535705, + "step": 4909 + }, + { + "epoch": 0.5903925930379367, + "flos": 16808075254800.0, + "grad_norm": 3.1871572572029003, + "learning_rate": 1.5164249540118708e-06, + "loss": 0.798, + "num_input_tokens_seen": 104554550, + "step": 4910 + }, + { + "epoch": 0.5905128359285757, + "flos": 17106832552320.0, + "grad_norm": 4.722371511655113, + "learning_rate": 1.5156691317596093e-06, + "loss": 0.82, + "num_input_tokens_seen": 104575695, + "step": 4911 + }, + { + "epoch": 0.5906330788192148, + "flos": 20558174757000.0, + "grad_norm": 8.949711919655527, + "learning_rate": 1.5149133829799556e-06, + "loss": 0.663, + "num_input_tokens_seen": 104593410, + "step": 4912 + }, + { + "epoch": 0.590753321709854, + "flos": 13487972787480.0, + "grad_norm": 4.345699174857756, + "learning_rate": 1.5141577077875556e-06, + "loss": 0.7858, + "num_input_tokens_seen": 104610455, + "step": 4913 + }, + { + "epoch": 0.590873564600493, + "flos": 12300890247360.0, + "grad_norm": 6.05430788880539, + "learning_rate": 1.5134021062970451e-06, + "loss": 0.6998, + "num_input_tokens_seen": 104628555, + "step": 4914 + }, + { + "epoch": 0.5909938074911321, + "flos": 9815964932520.0, + "grad_norm": 4.012603691876575, + "learning_rate": 1.5126465786230483e-06, + "loss": 0.7888, + "num_input_tokens_seen": 104645050, + "step": 4915 + }, + { + "epoch": 0.5911140503817712, + "flos": 19071701720160.0, + "grad_norm": 3.645627493724235, + "learning_rate": 1.5118911248801787e-06, + "loss": 0.7941, + "num_input_tokens_seen": 104662780, + "step": 4916 + }, + { + "epoch": 0.5912342932724103, + "flos": 17026239768240.0, + "grad_norm": 7.44857834942281, + "learning_rate": 1.5111357451830364e-06, + "loss": 0.7645, + "num_input_tokens_seen": 104681195, + "step": 4917 + }, + { + "epoch": 0.5913545361630493, + "flos": 14298491322480.0, + "grad_norm": 4.719508634303742, + "learning_rate": 1.5103804396462131e-06, + "loss": 0.6999, + "num_input_tokens_seen": 104700850, + "step": 4918 + }, + { + "epoch": 0.5914747790536885, + "flos": 19212914344920.0, + "grad_norm": 4.587876282105613, + "learning_rate": 1.5096252083842877e-06, + "loss": 0.7842, + "num_input_tokens_seen": 104719780, + "step": 4919 + }, + { + "epoch": 0.5915950219443276, + "flos": 20103962344080.0, + "grad_norm": 4.028640730627963, + "learning_rate": 1.5088700515118285e-06, + "loss": 0.8414, + "num_input_tokens_seen": 104738820, + "step": 4920 + }, + { + "epoch": 0.5917152648349666, + "flos": 16026394452000.0, + "grad_norm": 4.270756914649139, + "learning_rate": 1.508114969143392e-06, + "loss": 0.6491, + "num_input_tokens_seen": 104758525, + "step": 4921 + }, + { + "epoch": 0.5918355077256057, + "flos": 20615596941720.0, + "grad_norm": 4.081346430668902, + "learning_rate": 1.5073599613935238e-06, + "loss": 0.7641, + "num_input_tokens_seen": 104780365, + "step": 4922 + }, + { + "epoch": 0.5919557506162448, + "flos": 20671056208920.0, + "grad_norm": 4.805642332397274, + "learning_rate": 1.5066050283767574e-06, + "loss": 0.5617, + "num_input_tokens_seen": 104800765, + "step": 4923 + }, + { + "epoch": 0.5920759935068839, + "flos": 8763889665000.0, + "grad_norm": 5.82564641288241, + "learning_rate": 1.505850170207616e-06, + "loss": 0.8112, + "num_input_tokens_seen": 104817350, + "step": 4924 + }, + { + "epoch": 0.592196236397523, + "flos": 21589517242080.0, + "grad_norm": 4.012849672476715, + "learning_rate": 1.505095387000611e-06, + "loss": 0.7507, + "num_input_tokens_seen": 104839370, + "step": 4925 + }, + { + "epoch": 0.5923164792881621, + "flos": 17858820804000.0, + "grad_norm": 3.9171337633741863, + "learning_rate": 1.504340678870242e-06, + "loss": 0.737, + "num_input_tokens_seen": 104857305, + "step": 4926 + }, + { + "epoch": 0.5924367221788012, + "flos": 17592668773920.0, + "grad_norm": 4.4404403186383075, + "learning_rate": 1.5035860459309989e-06, + "loss": 0.8786, + "num_input_tokens_seen": 104874740, + "step": 4927 + }, + { + "epoch": 0.5925569650694402, + "flos": 19699225465920.0, + "grad_norm": 6.703718821580926, + "learning_rate": 1.5028314882973568e-06, + "loss": 0.625, + "num_input_tokens_seen": 104894865, + "step": 4928 + }, + { + "epoch": 0.5926772079600794, + "flos": 16317743318880.0, + "grad_norm": 3.67153110922884, + "learning_rate": 1.502077006083783e-06, + "loss": 0.8258, + "num_input_tokens_seen": 104913245, + "step": 4929 + }, + { + "epoch": 0.5927974508507184, + "flos": 14514787898280.0, + "grad_norm": 5.7195140557512, + "learning_rate": 1.5013225994047315e-06, + "loss": 0.7628, + "num_input_tokens_seen": 104930595, + "step": 4930 + }, + { + "epoch": 0.5929176937413575, + "flos": 11488725394440.0, + "grad_norm": 3.519289675264088, + "learning_rate": 1.5005682683746452e-06, + "loss": 0.7938, + "num_input_tokens_seen": 104948830, + "step": 4931 + }, + { + "epoch": 0.5930379366319967, + "flos": 12839209699920.0, + "grad_norm": 6.025480208511523, + "learning_rate": 1.4998140131079553e-06, + "loss": 0.7094, + "num_input_tokens_seen": 104964640, + "step": 4932 + }, + { + "epoch": 0.5931581795226357, + "flos": 12914356970880.0, + "grad_norm": 4.144609024511776, + "learning_rate": 1.4990598337190821e-06, + "loss": 0.7354, + "num_input_tokens_seen": 104980715, + "step": 4933 + }, + { + "epoch": 0.5932784224132748, + "flos": 17782090535040.0, + "grad_norm": 5.437860761929979, + "learning_rate": 1.4983057303224338e-06, + "loss": 0.6599, + "num_input_tokens_seen": 105000250, + "step": 4934 + }, + { + "epoch": 0.5933986653039139, + "flos": 16779997361640.0, + "grad_norm": 4.824944593674755, + "learning_rate": 1.4975517030324072e-06, + "loss": 0.8501, + "num_input_tokens_seen": 105017980, + "step": 4935 + }, + { + "epoch": 0.593518908194553, + "flos": 52452342549600.0, + "grad_norm": 0.8571915371445485, + "learning_rate": 1.4967977519633882e-06, + "loss": 0.6422, + "num_input_tokens_seen": 105075160, + "step": 4936 + }, + { + "epoch": 0.593639151085192, + "flos": 14946494571000.0, + "grad_norm": 6.3663274078786305, + "learning_rate": 1.4960438772297494e-06, + "loss": 0.7669, + "num_input_tokens_seen": 105091925, + "step": 4937 + }, + { + "epoch": 0.5937593939758312, + "flos": 22668910563720.0, + "grad_norm": 5.118390170574379, + "learning_rate": 1.495290078945855e-06, + "loss": 0.7237, + "num_input_tokens_seen": 105111410, + "step": 4938 + }, + { + "epoch": 0.5938796368664703, + "flos": 27119369905560.0, + "grad_norm": 2.9933244820897924, + "learning_rate": 1.4945363572260529e-06, + "loss": 0.731, + "num_input_tokens_seen": 105132125, + "step": 4939 + }, + { + "epoch": 0.5939998797571093, + "flos": 17458642960080.0, + "grad_norm": 3.052163879234391, + "learning_rate": 1.4937827121846845e-06, + "loss": 0.6484, + "num_input_tokens_seen": 105152100, + "step": 4940 + }, + { + "epoch": 0.5941201226477485, + "flos": 18456398718960.0, + "grad_norm": 4.396908211328462, + "learning_rate": 1.4930291439360755e-06, + "loss": 0.7229, + "num_input_tokens_seen": 105174385, + "step": 4941 + }, + { + "epoch": 0.5942403655383875, + "flos": 16404699754920.0, + "grad_norm": 3.209104288453007, + "learning_rate": 1.4922756525945427e-06, + "loss": 0.7773, + "num_input_tokens_seen": 105193415, + "step": 4942 + }, + { + "epoch": 0.5943606084290266, + "flos": 49869162684360.0, + "grad_norm": 0.7815330374135685, + "learning_rate": 1.4915222382743894e-06, + "loss": 0.6092, + "num_input_tokens_seen": 105251970, + "step": 4943 + }, + { + "epoch": 0.5944808513196658, + "flos": 13299500825160.0, + "grad_norm": 5.112168414641501, + "learning_rate": 1.4907689010899085e-06, + "loss": 0.7044, + "num_input_tokens_seen": 105269270, + "step": 4944 + }, + { + "epoch": 0.5946010942103048, + "flos": 18160047578400.0, + "grad_norm": 6.47242551310597, + "learning_rate": 1.4900156411553804e-06, + "loss": 0.6054, + "num_input_tokens_seen": 105288820, + "step": 4945 + }, + { + "epoch": 0.5947213371009439, + "flos": 11085603174240.0, + "grad_norm": 8.795447767352305, + "learning_rate": 1.4892624585850739e-06, + "loss": 0.8353, + "num_input_tokens_seen": 105306895, + "step": 4946 + }, + { + "epoch": 0.594841579991583, + "flos": 18942361580400.0, + "grad_norm": 5.9495278050880325, + "learning_rate": 1.4885093534932465e-06, + "loss": 0.7729, + "num_input_tokens_seen": 105324580, + "step": 4947 + }, + { + "epoch": 0.5949618228822221, + "flos": 17559810226800.0, + "grad_norm": 4.9522448217979305, + "learning_rate": 1.4877563259941433e-06, + "loss": 0.703, + "num_input_tokens_seen": 105342155, + "step": 4948 + }, + { + "epoch": 0.5950820657728612, + "flos": 29820306856560.0, + "grad_norm": 3.822039192521757, + "learning_rate": 1.4870033762019988e-06, + "loss": 0.6601, + "num_input_tokens_seen": 105362040, + "step": 4949 + }, + { + "epoch": 0.5952023086635003, + "flos": 16971097100640.0, + "grad_norm": 3.0397989121235405, + "learning_rate": 1.4862505042310334e-06, + "loss": 0.7171, + "num_input_tokens_seen": 105381045, + "step": 4950 + }, + { + "epoch": 0.5953225515541394, + "flos": 24718361670600.0, + "grad_norm": 2.699072296983403, + "learning_rate": 1.4854977101954587e-06, + "loss": 0.686, + "num_input_tokens_seen": 105402985, + "step": 4951 + }, + { + "epoch": 0.5954427944447784, + "flos": 17914533350880.0, + "grad_norm": 10.131706371536012, + "learning_rate": 1.4847449942094716e-06, + "loss": 0.8328, + "num_input_tokens_seen": 105421585, + "step": 4952 + }, + { + "epoch": 0.5955630373354175, + "flos": 13542798855480.0, + "grad_norm": 4.424870962211956, + "learning_rate": 1.4839923563872598e-06, + "loss": 0.8428, + "num_input_tokens_seen": 105439845, + "step": 4953 + }, + { + "epoch": 0.5956832802260567, + "flos": 14460278429880.0, + "grad_norm": 4.63960064715047, + "learning_rate": 1.483239796842997e-06, + "loss": 0.7478, + "num_input_tokens_seen": 105457595, + "step": 4954 + }, + { + "epoch": 0.5958035231166957, + "flos": 14429446120200.0, + "grad_norm": 3.305042004339984, + "learning_rate": 1.4824873156908462e-06, + "loss": 0.8191, + "num_input_tokens_seen": 105475240, + "step": 4955 + }, + { + "epoch": 0.5959237660073348, + "flos": 15837669210000.0, + "grad_norm": 2.9548154160736346, + "learning_rate": 1.4817349130449584e-06, + "loss": 0.7458, + "num_input_tokens_seen": 105494680, + "step": 4956 + }, + { + "epoch": 0.5960440088979739, + "flos": 15480223329360.0, + "grad_norm": 5.489710980643788, + "learning_rate": 1.4809825890194717e-06, + "loss": 0.8111, + "num_input_tokens_seen": 105513070, + "step": 4957 + }, + { + "epoch": 0.596164251788613, + "flos": 10734774225240.0, + "grad_norm": 3.0360717178337664, + "learning_rate": 1.4802303437285139e-06, + "loss": 0.7534, + "num_input_tokens_seen": 105530060, + "step": 4958 + }, + { + "epoch": 0.596284494679252, + "flos": 14974192544640.0, + "grad_norm": 4.724632184279893, + "learning_rate": 1.4794781772861994e-06, + "loss": 0.7715, + "num_input_tokens_seen": 105546275, + "step": 4959 + }, + { + "epoch": 0.5964047375698912, + "flos": 22912335233880.0, + "grad_norm": 4.492831560487017, + "learning_rate": 1.4787260898066324e-06, + "loss": 0.6485, + "num_input_tokens_seen": 105565995, + "step": 4960 + }, + { + "epoch": 0.5965249804605303, + "flos": 20151886540800.0, + "grad_norm": 5.914129259344966, + "learning_rate": 1.4779740814039023e-06, + "loss": 0.8379, + "num_input_tokens_seen": 105585800, + "step": 4961 + }, + { + "epoch": 0.5966452233511693, + "flos": 22587874540200.0, + "grad_norm": 5.137453904142353, + "learning_rate": 1.4772221521920894e-06, + "loss": 0.6642, + "num_input_tokens_seen": 105605545, + "step": 4962 + }, + { + "epoch": 0.5967654662418085, + "flos": 18670542417480.0, + "grad_norm": 3.3593839423481997, + "learning_rate": 1.4764703022852598e-06, + "loss": 0.7236, + "num_input_tokens_seen": 105625785, + "step": 4963 + }, + { + "epoch": 0.5968857091324475, + "flos": 13947820673280.0, + "grad_norm": 4.394563150925879, + "learning_rate": 1.4757185317974696e-06, + "loss": 0.7453, + "num_input_tokens_seen": 105643890, + "step": 4964 + }, + { + "epoch": 0.5970059520230866, + "flos": 17346869606760.0, + "grad_norm": 4.643979898874182, + "learning_rate": 1.474966840842761e-06, + "loss": 0.7031, + "num_input_tokens_seen": 105663190, + "step": 4965 + }, + { + "epoch": 0.5971261949137258, + "flos": 16972046899440.0, + "grad_norm": 5.930078305118237, + "learning_rate": 1.4742152295351655e-06, + "loss": 0.8544, + "num_input_tokens_seen": 105682065, + "step": 4966 + }, + { + "epoch": 0.5972464378043648, + "flos": 15027150675000.0, + "grad_norm": 6.47629391029314, + "learning_rate": 1.4734636979887016e-06, + "loss": 0.6213, + "num_input_tokens_seen": 105699245, + "step": 4967 + }, + { + "epoch": 0.5973666806950039, + "flos": 21561407688960.0, + "grad_norm": 4.999188357280348, + "learning_rate": 1.4727122463173755e-06, + "loss": 0.89, + "num_input_tokens_seen": 105717495, + "step": 4968 + }, + { + "epoch": 0.597486923585643, + "flos": 16297105516320.0, + "grad_norm": 3.6636190898495786, + "learning_rate": 1.471960874635183e-06, + "loss": 0.6263, + "num_input_tokens_seen": 105736775, + "step": 4969 + }, + { + "epoch": 0.5976071664762821, + "flos": 10005766613160.0, + "grad_norm": 4.8438088690349055, + "learning_rate": 1.4712095830561055e-06, + "loss": 0.6797, + "num_input_tokens_seen": 105754985, + "step": 4970 + }, + { + "epoch": 0.5977274093669211, + "flos": 13947377433840.0, + "grad_norm": 4.834087159576442, + "learning_rate": 1.4704583716941147e-06, + "loss": 0.7814, + "num_input_tokens_seen": 105773570, + "step": 4971 + }, + { + "epoch": 0.5978476522575603, + "flos": 14892776601600.0, + "grad_norm": 3.049231923431863, + "learning_rate": 1.4697072406631672e-06, + "loss": 0.6975, + "num_input_tokens_seen": 105793195, + "step": 4972 + }, + { + "epoch": 0.5979678951481994, + "flos": 21292247962680.0, + "grad_norm": 3.56952790502073, + "learning_rate": 1.4689561900772097e-06, + "loss": 0.7208, + "num_input_tokens_seen": 105812975, + "step": 4973 + }, + { + "epoch": 0.5980881380388384, + "flos": 13110427323600.0, + "grad_norm": 7.467407872104559, + "learning_rate": 1.4682052200501758e-06, + "loss": 0.7117, + "num_input_tokens_seen": 105829900, + "step": 4974 + }, + { + "epoch": 0.5982083809294776, + "flos": 16806682216560.0, + "grad_norm": 4.912347394208467, + "learning_rate": 1.4674543306959876e-06, + "loss": 0.7679, + "num_input_tokens_seen": 105849090, + "step": 4975 + }, + { + "epoch": 0.5983286238201166, + "flos": 15348002133240.0, + "grad_norm": 6.33142184808501, + "learning_rate": 1.4667035221285535e-06, + "loss": 0.8282, + "num_input_tokens_seen": 105866450, + "step": 4976 + }, + { + "epoch": 0.5984488667107557, + "flos": 20670739609320.0, + "grad_norm": 2.724727329996003, + "learning_rate": 1.4659527944617715e-06, + "loss": 0.7304, + "num_input_tokens_seen": 105886115, + "step": 4977 + }, + { + "epoch": 0.5985691096013949, + "flos": 12003367688280.0, + "grad_norm": 8.521907623642633, + "learning_rate": 1.465202147809526e-06, + "loss": 0.741, + "num_input_tokens_seen": 105904330, + "step": 4978 + }, + { + "epoch": 0.5986893524920339, + "flos": 19586280694080.0, + "grad_norm": 5.523514678286382, + "learning_rate": 1.4644515822856888e-06, + "loss": 0.7537, + "num_input_tokens_seen": 105922485, + "step": 4979 + }, + { + "epoch": 0.598809595382673, + "flos": 45412877909880.0, + "grad_norm": 0.7676365953845266, + "learning_rate": 1.4637010980041215e-06, + "loss": 0.5805, + "num_input_tokens_seen": 105984315, + "step": 4980 + }, + { + "epoch": 0.5989298382733121, + "flos": 8547213169680.0, + "grad_norm": 9.279249393098366, + "learning_rate": 1.4629506950786707e-06, + "loss": 0.8792, + "num_input_tokens_seen": 106000215, + "step": 4981 + }, + { + "epoch": 0.5990500811639512, + "flos": 44240006200440.0, + "grad_norm": 0.826194961261367, + "learning_rate": 1.4622003736231733e-06, + "loss": 0.5787, + "num_input_tokens_seen": 106058925, + "step": 4982 + }, + { + "epoch": 0.5991703240545903, + "flos": 13299120905640.0, + "grad_norm": 3.597261786192164, + "learning_rate": 1.461450133751451e-06, + "loss": 0.7763, + "num_input_tokens_seen": 106076715, + "step": 4983 + }, + { + "epoch": 0.5992905669452293, + "flos": 20320385559720.0, + "grad_norm": 4.018138529294452, + "learning_rate": 1.4606999755773153e-06, + "loss": 0.7466, + "num_input_tokens_seen": 106097640, + "step": 4984 + }, + { + "epoch": 0.5994108098358685, + "flos": 14946811170600.0, + "grad_norm": 3.0143877564851795, + "learning_rate": 1.4599498992145643e-06, + "loss": 0.7922, + "num_input_tokens_seen": 106117385, + "step": 4985 + }, + { + "epoch": 0.5995310527265075, + "flos": 16293971180280.0, + "grad_norm": 4.168782548906033, + "learning_rate": 1.4591999047769846e-06, + "loss": 0.7035, + "num_input_tokens_seen": 106135960, + "step": 4986 + }, + { + "epoch": 0.5996512956171466, + "flos": 13813035020400.0, + "grad_norm": 3.7521330806661557, + "learning_rate": 1.4584499923783486e-06, + "loss": 0.7418, + "num_input_tokens_seen": 106154260, + "step": 4987 + }, + { + "epoch": 0.5997715385077858, + "flos": 11187751899720.0, + "grad_norm": 4.813548778844303, + "learning_rate": 1.457700162132419e-06, + "loss": 0.7522, + "num_input_tokens_seen": 106170970, + "step": 4988 + }, + { + "epoch": 0.5998917813984248, + "flos": 18510813207480.0, + "grad_norm": 5.0514696305596845, + "learning_rate": 1.4569504141529433e-06, + "loss": 0.6998, + "num_input_tokens_seen": 106188525, + "step": 4989 + }, + { + "epoch": 0.6000120242890639, + "flos": 16135001809320.0, + "grad_norm": 11.959635558428596, + "learning_rate": 1.456200748553658e-06, + "loss": 0.7031, + "num_input_tokens_seen": 106206240, + "step": 4990 + }, + { + "epoch": 0.600132267179703, + "flos": 21914262875400.0, + "grad_norm": 3.663130977295752, + "learning_rate": 1.455451165448287e-06, + "loss": 0.7694, + "num_input_tokens_seen": 106228615, + "step": 4991 + }, + { + "epoch": 0.6002525100703421, + "flos": 18348297921000.0, + "grad_norm": 4.862101780292534, + "learning_rate": 1.4547016649505407e-06, + "loss": 0.6983, + "num_input_tokens_seen": 106246345, + "step": 4992 + }, + { + "epoch": 0.6003727529609811, + "flos": 15243162311160.0, + "grad_norm": 5.2081014759590385, + "learning_rate": 1.4539522471741193e-06, + "loss": 0.8283, + "num_input_tokens_seen": 106263490, + "step": 4993 + }, + { + "epoch": 0.6004929958516203, + "flos": 11353971401520.0, + "grad_norm": 3.9136028210218843, + "learning_rate": 1.4532029122327067e-06, + "loss": 0.6901, + "num_input_tokens_seen": 106279995, + "step": 4994 + }, + { + "epoch": 0.6006132387422594, + "flos": 15919560052440.0, + "grad_norm": 5.219937674782456, + "learning_rate": 1.4524536602399783e-06, + "loss": 0.7388, + "num_input_tokens_seen": 106298805, + "step": 4995 + }, + { + "epoch": 0.6007334816328984, + "flos": 16730236887240.0, + "grad_norm": 6.661803315400864, + "learning_rate": 1.4517044913095938e-06, + "loss": 0.7606, + "num_input_tokens_seen": 106318945, + "step": 4996 + }, + { + "epoch": 0.6008537245235376, + "flos": 20774502992760.0, + "grad_norm": 3.2327801739204394, + "learning_rate": 1.4509554055552022e-06, + "loss": 0.8009, + "num_input_tokens_seen": 106338895, + "step": 4997 + }, + { + "epoch": 0.6009739674141766, + "flos": 15270195425640.0, + "grad_norm": 6.350853161305562, + "learning_rate": 1.450206403090439e-06, + "loss": 0.8176, + "num_input_tokens_seen": 106356810, + "step": 4998 + }, + { + "epoch": 0.6010942103048157, + "flos": 14969855130120.0, + "grad_norm": 4.323791317374275, + "learning_rate": 1.4494574840289274e-06, + "loss": 0.8443, + "num_input_tokens_seen": 106373645, + "step": 4999 + }, + { + "epoch": 0.6012144531954549, + "flos": 17431229926080.0, + "grad_norm": 5.194924797563845, + "learning_rate": 1.4487086484842782e-06, + "loss": 0.7303, + "num_input_tokens_seen": 106392010, + "step": 5000 + }, + { + "epoch": 0.6013346960860939, + "flos": 13865676551160.0, + "grad_norm": 6.8711004121627735, + "learning_rate": 1.4479598965700878e-06, + "loss": 0.5765, + "num_input_tokens_seen": 106408995, + "step": 5001 + }, + { + "epoch": 0.601454938976733, + "flos": 17593618572720.0, + "grad_norm": 4.570099237145023, + "learning_rate": 1.4472112283999427e-06, + "loss": 0.665, + "num_input_tokens_seen": 106427370, + "step": 5002 + }, + { + "epoch": 0.6015751818673721, + "flos": 19370237397960.0, + "grad_norm": 6.219391909493801, + "learning_rate": 1.4464626440874143e-06, + "loss": 0.6869, + "num_input_tokens_seen": 106446205, + "step": 5003 + }, + { + "epoch": 0.6016954247580112, + "flos": 9519328852320.0, + "grad_norm": 5.918052224555674, + "learning_rate": 1.4457141437460636e-06, + "loss": 0.7203, + "num_input_tokens_seen": 106463150, + "step": 5004 + }, + { + "epoch": 0.6018156676486502, + "flos": 17403310332720.0, + "grad_norm": 8.397972370840847, + "learning_rate": 1.444965727489436e-06, + "loss": 0.7175, + "num_input_tokens_seen": 106482315, + "step": 5005 + }, + { + "epoch": 0.6019359105392894, + "flos": 19402304446080.0, + "grad_norm": 12.222761115321962, + "learning_rate": 1.444217395431066e-06, + "loss": 0.6223, + "num_input_tokens_seen": 106504575, + "step": 5006 + }, + { + "epoch": 0.6020561534299285, + "flos": 51023766596880.0, + "grad_norm": 0.8091368394796098, + "learning_rate": 1.4434691476844755e-06, + "loss": 0.5774, + "num_input_tokens_seen": 106565270, + "step": 5007 + }, + { + "epoch": 0.6021763963205675, + "flos": 15972739802520.0, + "grad_norm": 4.781201131733004, + "learning_rate": 1.4427209843631729e-06, + "loss": 0.6589, + "num_input_tokens_seen": 106582040, + "step": 5008 + }, + { + "epoch": 0.6022966392112067, + "flos": 19483150509840.0, + "grad_norm": 3.3326687309206346, + "learning_rate": 1.4419729055806534e-06, + "loss": 0.7991, + "num_input_tokens_seen": 106601195, + "step": 5009 + }, + { + "epoch": 0.6024168821018457, + "flos": 15135219813000.0, + "grad_norm": 4.370201749356444, + "learning_rate": 1.441224911450401e-06, + "loss": 0.8202, + "num_input_tokens_seen": 106616870, + "step": 5010 + }, + { + "epoch": 0.6025371249924848, + "flos": 18078694955280.0, + "grad_norm": 3.9970555196861617, + "learning_rate": 1.4404770020858851e-06, + "loss": 0.8206, + "num_input_tokens_seen": 106636075, + "step": 5011 + }, + { + "epoch": 0.602657367883124, + "flos": 19021846265880.0, + "grad_norm": 2.6467843536744655, + "learning_rate": 1.439729177600563e-06, + "loss": 0.8518, + "num_input_tokens_seen": 106656290, + "step": 5012 + }, + { + "epoch": 0.602777610773763, + "flos": 12165313095480.0, + "grad_norm": 4.34825570223021, + "learning_rate": 1.4389814381078793e-06, + "loss": 0.7136, + "num_input_tokens_seen": 106675250, + "step": 5013 + }, + { + "epoch": 0.6028978536644021, + "flos": 9681400899360.0, + "grad_norm": 5.3225618108526165, + "learning_rate": 1.438233783721265e-06, + "loss": 0.7776, + "num_input_tokens_seen": 106691135, + "step": 5014 + }, + { + "epoch": 0.6030180965550412, + "flos": 14351449452840.0, + "grad_norm": 4.78517857137651, + "learning_rate": 1.43748621455414e-06, + "loss": 0.7633, + "num_input_tokens_seen": 106707290, + "step": 5015 + }, + { + "epoch": 0.6031383394456803, + "flos": 10512367277160.0, + "grad_norm": 15.367247573122286, + "learning_rate": 1.4367387307199082e-06, + "loss": 0.7919, + "num_input_tokens_seen": 106723860, + "step": 5016 + }, + { + "epoch": 0.6032585823363193, + "flos": 10113234211920.0, + "grad_norm": 4.221202448339014, + "learning_rate": 1.4359913323319632e-06, + "loss": 0.8142, + "num_input_tokens_seen": 106740750, + "step": 5017 + }, + { + "epoch": 0.6033788252269584, + "flos": 17836251743880.0, + "grad_norm": 3.126897007225986, + "learning_rate": 1.4352440195036847e-06, + "loss": 0.766, + "num_input_tokens_seen": 106760645, + "step": 5018 + }, + { + "epoch": 0.6034990681175976, + "flos": 18700709868000.0, + "grad_norm": 3.0922618330800726, + "learning_rate": 1.4344967923484395e-06, + "loss": 0.7766, + "num_input_tokens_seen": 106782335, + "step": 5019 + }, + { + "epoch": 0.6036193110082366, + "flos": 19024125783000.0, + "grad_norm": 14.83002449306284, + "learning_rate": 1.433749650979581e-06, + "loss": 0.6994, + "num_input_tokens_seen": 106802040, + "step": 5020 + }, + { + "epoch": 0.6037395538988757, + "flos": 18754016257920.0, + "grad_norm": 3.6434765564274363, + "learning_rate": 1.433002595510451e-06, + "loss": 0.6683, + "num_input_tokens_seen": 106820540, + "step": 5021 + }, + { + "epoch": 0.6038597967895148, + "flos": 12997577531640.0, + "grad_norm": 10.482899864789099, + "learning_rate": 1.4322556260543757e-06, + "loss": 0.7071, + "num_input_tokens_seen": 106836835, + "step": 5022 + }, + { + "epoch": 0.6039800396801539, + "flos": 48046324808880.0, + "grad_norm": 1.0332608702591064, + "learning_rate": 1.4315087427246703e-06, + "loss": 0.6487, + "num_input_tokens_seen": 106890380, + "step": 5023 + }, + { + "epoch": 0.604100282570793, + "flos": 49686005103960.0, + "grad_norm": 0.896210073754889, + "learning_rate": 1.4307619456346372e-06, + "loss": 0.6106, + "num_input_tokens_seen": 106934405, + "step": 5024 + }, + { + "epoch": 0.6042205254614321, + "flos": 25935516681360.0, + "grad_norm": 2.734810581935682, + "learning_rate": 1.430015234897564e-06, + "loss": 0.7151, + "num_input_tokens_seen": 106957405, + "step": 5025 + }, + { + "epoch": 0.6043407683520712, + "flos": 33603011626200.0, + "grad_norm": 6.632011589877651, + "learning_rate": 1.4292686106267274e-06, + "loss": 0.6471, + "num_input_tokens_seen": 106979975, + "step": 5026 + }, + { + "epoch": 0.6044610112427102, + "flos": 11787451032000.0, + "grad_norm": 4.373843012821295, + "learning_rate": 1.4285220729353876e-06, + "loss": 0.749, + "num_input_tokens_seen": 106998235, + "step": 5027 + }, + { + "epoch": 0.6045812541333494, + "flos": 10028462313120.0, + "grad_norm": 27.384417303188254, + "learning_rate": 1.4277756219367957e-06, + "loss": 0.7658, + "num_input_tokens_seen": 107014980, + "step": 5028 + }, + { + "epoch": 0.6047014970239885, + "flos": 14596108861440.0, + "grad_norm": 4.579764739744951, + "learning_rate": 1.4270292577441864e-06, + "loss": 0.7823, + "num_input_tokens_seen": 107034205, + "step": 5029 + }, + { + "epoch": 0.6048217399146275, + "flos": 18564182917320.0, + "grad_norm": 2.73497351499353, + "learning_rate": 1.4262829804707836e-06, + "loss": 0.7042, + "num_input_tokens_seen": 107055915, + "step": 5030 + }, + { + "epoch": 0.6049419828052667, + "flos": 19077653792640.0, + "grad_norm": 4.740210634855056, + "learning_rate": 1.4255367902297958e-06, + "loss": 0.6778, + "num_input_tokens_seen": 107076965, + "step": 5031 + }, + { + "epoch": 0.6050622256959057, + "flos": 10653801521640.0, + "grad_norm": 4.829098501484116, + "learning_rate": 1.4247906871344215e-06, + "loss": 0.7822, + "num_input_tokens_seen": 107092080, + "step": 5032 + }, + { + "epoch": 0.6051824685865448, + "flos": 17080021057560.0, + "grad_norm": 10.100265620298604, + "learning_rate": 1.4240446712978415e-06, + "loss": 0.7372, + "num_input_tokens_seen": 107110785, + "step": 5033 + }, + { + "epoch": 0.605302711477184, + "flos": 20211809862360.0, + "grad_norm": 6.75941601730649, + "learning_rate": 1.423298742833227e-06, + "loss": 0.7207, + "num_input_tokens_seen": 107129165, + "step": 5034 + }, + { + "epoch": 0.605422954367823, + "flos": 11028117669600.0, + "grad_norm": 3.717958895098504, + "learning_rate": 1.4225529018537352e-06, + "loss": 0.717, + "num_input_tokens_seen": 107144390, + "step": 5035 + }, + { + "epoch": 0.6055431972584621, + "flos": 20293827344640.0, + "grad_norm": 3.3019669248517527, + "learning_rate": 1.4218071484725082e-06, + "loss": 0.7665, + "num_input_tokens_seen": 107166230, + "step": 5036 + }, + { + "epoch": 0.6056634401491012, + "flos": 14160729633360.0, + "grad_norm": 7.617652229413026, + "learning_rate": 1.4210614828026786e-06, + "loss": 0.7437, + "num_input_tokens_seen": 107183800, + "step": 5037 + }, + { + "epoch": 0.6057836830397403, + "flos": 18158844499920.0, + "grad_norm": 3.2035690372006718, + "learning_rate": 1.4203159049573605e-06, + "loss": 0.7326, + "num_input_tokens_seen": 107204755, + "step": 5038 + }, + { + "epoch": 0.6059039259303793, + "flos": 15027625574400.0, + "grad_norm": 5.558627034005235, + "learning_rate": 1.4195704150496593e-06, + "loss": 0.8555, + "num_input_tokens_seen": 107222190, + "step": 5039 + }, + { + "epoch": 0.6060241688210185, + "flos": 15406247476920.0, + "grad_norm": 3.978173669605744, + "learning_rate": 1.4188250131926639e-06, + "loss": 0.7151, + "num_input_tokens_seen": 107240710, + "step": 5040 + }, + { + "epoch": 0.6061444117116576, + "flos": 11921255226120.0, + "grad_norm": 4.668788541050027, + "learning_rate": 1.4180796994994525e-06, + "loss": 0.7919, + "num_input_tokens_seen": 107257845, + "step": 5041 + }, + { + "epoch": 0.6062646546022966, + "flos": 15730043311440.0, + "grad_norm": 3.843074956404628, + "learning_rate": 1.4173344740830877e-06, + "loss": 0.7074, + "num_input_tokens_seen": 107276695, + "step": 5042 + }, + { + "epoch": 0.6063848974929358, + "flos": 28663233467160.0, + "grad_norm": 2.9279656398403886, + "learning_rate": 1.4165893370566206e-06, + "loss": 0.6909, + "num_input_tokens_seen": 107300170, + "step": 5043 + }, + { + "epoch": 0.6065051403835748, + "flos": 14353000790880.0, + "grad_norm": 3.082574561981556, + "learning_rate": 1.4158442885330865e-06, + "loss": 0.7506, + "num_input_tokens_seen": 107318460, + "step": 5044 + }, + { + "epoch": 0.6066253832742139, + "flos": 17214331811040.0, + "grad_norm": 3.9725279360653745, + "learning_rate": 1.4150993286255094e-06, + "loss": 0.7772, + "num_input_tokens_seen": 107337430, + "step": 5045 + }, + { + "epoch": 0.6067456261648531, + "flos": 13973144149920.0, + "grad_norm": 5.393540984134095, + "learning_rate": 1.4143544574468993e-06, + "loss": 0.7681, + "num_input_tokens_seen": 107355510, + "step": 5046 + }, + { + "epoch": 0.6068658690554921, + "flos": 14999611001160.0, + "grad_norm": 2.7683132920830653, + "learning_rate": 1.4136096751102523e-06, + "loss": 0.8045, + "num_input_tokens_seen": 107373560, + "step": 5047 + }, + { + "epoch": 0.6069861119461312, + "flos": 20071230436800.0, + "grad_norm": 4.868329693301959, + "learning_rate": 1.4128649817285516e-06, + "loss": 0.8094, + "num_input_tokens_seen": 107391415, + "step": 5048 + }, + { + "epoch": 0.6071063548367702, + "flos": 18778010016240.0, + "grad_norm": 5.823210334082971, + "learning_rate": 1.412120377414766e-06, + "loss": 0.62, + "num_input_tokens_seen": 107411325, + "step": 5049 + }, + { + "epoch": 0.6072265977274094, + "flos": 17915324849880.0, + "grad_norm": 3.262312966316652, + "learning_rate": 1.4113758622818522e-06, + "loss": 0.6857, + "num_input_tokens_seen": 107431110, + "step": 5050 + }, + { + "epoch": 0.6073468406180484, + "flos": 13244674757160.0, + "grad_norm": 3.8598034578400875, + "learning_rate": 1.410631436442751e-06, + "loss": 0.8088, + "num_input_tokens_seen": 107449625, + "step": 5051 + }, + { + "epoch": 0.6074670835086875, + "flos": 15129837619800.0, + "grad_norm": 6.069314803487943, + "learning_rate": 1.4098871000103936e-06, + "loss": 0.8488, + "num_input_tokens_seen": 107467945, + "step": 5052 + }, + { + "epoch": 0.6075873263993267, + "flos": 17404640051040.0, + "grad_norm": 3.7906702150571325, + "learning_rate": 1.409142853097693e-06, + "loss": 0.8129, + "num_input_tokens_seen": 107487905, + "step": 5053 + }, + { + "epoch": 0.6077075692899657, + "flos": 17910005976600.0, + "grad_norm": 5.551647271739224, + "learning_rate": 1.408398695817553e-06, + "loss": 0.7834, + "num_input_tokens_seen": 107504850, + "step": 5054 + }, + { + "epoch": 0.6078278121806048, + "flos": 20077879028400.0, + "grad_norm": 17.003081098603957, + "learning_rate": 1.4076546282828593e-06, + "loss": 0.678, + "num_input_tokens_seen": 107527425, + "step": 5055 + }, + { + "epoch": 0.6079480550712439, + "flos": 28258718208720.0, + "grad_norm": 9.032182163587628, + "learning_rate": 1.4069106506064874e-06, + "loss": 0.6403, + "num_input_tokens_seen": 107548570, + "step": 5056 + }, + { + "epoch": 0.608068297961883, + "flos": 18563454738240.0, + "grad_norm": 5.52897105710614, + "learning_rate": 1.4061667629012989e-06, + "loss": 0.7615, + "num_input_tokens_seen": 107568960, + "step": 5057 + }, + { + "epoch": 0.608188540852522, + "flos": 17723971831200.0, + "grad_norm": 2.818140604327484, + "learning_rate": 1.40542296528014e-06, + "loss": 0.8146, + "num_input_tokens_seen": 107588340, + "step": 5058 + }, + { + "epoch": 0.6083087837431612, + "flos": 15564298709040.0, + "grad_norm": 3.011153310297846, + "learning_rate": 1.4046792578558452e-06, + "loss": 0.7412, + "num_input_tokens_seen": 107605955, + "step": 5059 + }, + { + "epoch": 0.6084290266338003, + "flos": 12003051088680.0, + "grad_norm": 5.165837477729383, + "learning_rate": 1.4039356407412325e-06, + "loss": 0.7464, + "num_input_tokens_seen": 107618915, + "step": 5060 + }, + { + "epoch": 0.6085492695244393, + "flos": 48989064540000.0, + "grad_norm": 0.8326573769563024, + "learning_rate": 1.40319211404911e-06, + "loss": 0.6109, + "num_input_tokens_seen": 107673635, + "step": 5061 + }, + { + "epoch": 0.6086695124150785, + "flos": 17292201838560.0, + "grad_norm": 4.1086244078945064, + "learning_rate": 1.4024486778922691e-06, + "loss": 0.8796, + "num_input_tokens_seen": 107691670, + "step": 5062 + }, + { + "epoch": 0.6087897553057176, + "flos": 14730736214520.0, + "grad_norm": 5.8230459054470245, + "learning_rate": 1.4017053323834884e-06, + "loss": 0.7581, + "num_input_tokens_seen": 107711220, + "step": 5063 + }, + { + "epoch": 0.6089099981963566, + "flos": 18671745495960.0, + "grad_norm": 4.032584982140314, + "learning_rate": 1.4009620776355333e-06, + "loss": 0.7423, + "num_input_tokens_seen": 107732540, + "step": 5064 + }, + { + "epoch": 0.6090302410869958, + "flos": 18560478702000.0, + "grad_norm": 4.6808065818134015, + "learning_rate": 1.4002189137611553e-06, + "loss": 0.7775, + "num_input_tokens_seen": 107751600, + "step": 5065 + }, + { + "epoch": 0.6091504839776348, + "flos": 17565319059840.0, + "grad_norm": 5.307644408997581, + "learning_rate": 1.3994758408730901e-06, + "loss": 0.6755, + "num_input_tokens_seen": 107770505, + "step": 5066 + }, + { + "epoch": 0.6092707268682739, + "flos": 21750956089920.0, + "grad_norm": 5.762908378247412, + "learning_rate": 1.3987328590840629e-06, + "loss": 0.7679, + "num_input_tokens_seen": 107791170, + "step": 5067 + }, + { + "epoch": 0.609390969758913, + "flos": 17593301973120.0, + "grad_norm": 3.8341981630966915, + "learning_rate": 1.397989968506783e-06, + "loss": 0.8473, + "num_input_tokens_seen": 107809900, + "step": 5068 + }, + { + "epoch": 0.6095112126495521, + "flos": 8028581720880.0, + "grad_norm": 6.375806271298623, + "learning_rate": 1.3972471692539458e-06, + "loss": 0.7105, + "num_input_tokens_seen": 107824335, + "step": 5069 + }, + { + "epoch": 0.6096314555401912, + "flos": 12589484697720.0, + "grad_norm": 12.383689218766442, + "learning_rate": 1.3965044614382348e-06, + "loss": 0.7382, + "num_input_tokens_seen": 107839505, + "step": 5070 + }, + { + "epoch": 0.6097516984308303, + "flos": 15831590497680.0, + "grad_norm": 15.254436910583896, + "learning_rate": 1.3957618451723162e-06, + "loss": 0.7411, + "num_input_tokens_seen": 107855255, + "step": 5071 + }, + { + "epoch": 0.6098719413214694, + "flos": 19942460176320.0, + "grad_norm": 4.978144180251323, + "learning_rate": 1.3950193205688457e-06, + "loss": 0.687, + "num_input_tokens_seen": 107874700, + "step": 5072 + }, + { + "epoch": 0.6099921842121084, + "flos": 14920189635600.0, + "grad_norm": 3.658482318413676, + "learning_rate": 1.3942768877404627e-06, + "loss": 0.8222, + "num_input_tokens_seen": 107893385, + "step": 5073 + }, + { + "epoch": 0.6101124271027476, + "flos": 17105439514080.0, + "grad_norm": 3.682724139092658, + "learning_rate": 1.393534546799795e-06, + "loss": 0.7397, + "num_input_tokens_seen": 107912805, + "step": 5074 + }, + { + "epoch": 0.6102326699933867, + "flos": 19563933253680.0, + "grad_norm": 3.5168703035330373, + "learning_rate": 1.3927922978594536e-06, + "loss": 0.6568, + "num_input_tokens_seen": 107933610, + "step": 5075 + }, + { + "epoch": 0.6103529128840257, + "flos": 44697764528880.0, + "grad_norm": 0.8397629100128584, + "learning_rate": 1.3920501410320387e-06, + "loss": 0.6159, + "num_input_tokens_seen": 107989445, + "step": 5076 + }, + { + "epoch": 0.6104731557746649, + "flos": 13889543669640.0, + "grad_norm": 4.89059672484588, + "learning_rate": 1.3913080764301333e-06, + "loss": 0.747, + "num_input_tokens_seen": 108006125, + "step": 5077 + }, + { + "epoch": 0.6105933986653039, + "flos": 17106104373240.0, + "grad_norm": 5.970173131838062, + "learning_rate": 1.3905661041663085e-06, + "loss": 0.7055, + "num_input_tokens_seen": 108027030, + "step": 5078 + }, + { + "epoch": 0.610713641555943, + "flos": 25446672763560.0, + "grad_norm": 5.477987106859592, + "learning_rate": 1.389824224353122e-06, + "loss": 0.6279, + "num_input_tokens_seen": 108048340, + "step": 5079 + }, + { + "epoch": 0.610833884446582, + "flos": 19532626044600.0, + "grad_norm": 17.28499337846995, + "learning_rate": 1.389082437103115e-06, + "loss": 0.7549, + "num_input_tokens_seen": 108067330, + "step": 5080 + }, + { + "epoch": 0.6109541273372212, + "flos": 15513778395600.0, + "grad_norm": 4.767867364357104, + "learning_rate": 1.3883407425288172e-06, + "loss": 0.7571, + "num_input_tokens_seen": 108087385, + "step": 5081 + }, + { + "epoch": 0.6110743702278603, + "flos": 14676733305480.0, + "grad_norm": 7.153439698769102, + "learning_rate": 1.3875991407427417e-06, + "loss": 0.761, + "num_input_tokens_seen": 108105330, + "step": 5082 + }, + { + "epoch": 0.6111946131184993, + "flos": 50364935642040.0, + "grad_norm": 0.7701390765661621, + "learning_rate": 1.38685763185739e-06, + "loss": 0.6027, + "num_input_tokens_seen": 108158710, + "step": 5083 + }, + { + "epoch": 0.6113148560091385, + "flos": 14568315907920.0, + "grad_norm": 4.527344803238122, + "learning_rate": 1.3861162159852476e-06, + "loss": 0.6517, + "num_input_tokens_seen": 108176565, + "step": 5084 + }, + { + "epoch": 0.6114350988997775, + "flos": 17376213898320.0, + "grad_norm": 3.6613236136181264, + "learning_rate": 1.3853748932387875e-06, + "loss": 0.7845, + "num_input_tokens_seen": 108196925, + "step": 5085 + }, + { + "epoch": 0.6115553417904166, + "flos": 17592193874520.0, + "grad_norm": 4.498990687455415, + "learning_rate": 1.3846336637304671e-06, + "loss": 0.7321, + "num_input_tokens_seen": 108214915, + "step": 5086 + }, + { + "epoch": 0.6116755846810558, + "flos": 17377701916440.0, + "grad_norm": 6.403331557033173, + "learning_rate": 1.3838925275727316e-06, + "loss": 0.8031, + "num_input_tokens_seen": 108235375, + "step": 5087 + }, + { + "epoch": 0.6117958275716948, + "flos": 13624088158680.0, + "grad_norm": 3.987677906815065, + "learning_rate": 1.3831514848780089e-06, + "loss": 0.7783, + "num_input_tokens_seen": 108254670, + "step": 5088 + }, + { + "epoch": 0.6119160704623339, + "flos": 12002829468960.0, + "grad_norm": 5.1421215348720155, + "learning_rate": 1.3824105357587152e-06, + "loss": 0.8993, + "num_input_tokens_seen": 108271495, + "step": 5089 + }, + { + "epoch": 0.612036313352973, + "flos": 17512265949600.0, + "grad_norm": 2.9881309321874614, + "learning_rate": 1.381669680327253e-06, + "loss": 0.8142, + "num_input_tokens_seen": 108292895, + "step": 5090 + }, + { + "epoch": 0.6121565562436121, + "flos": 19775575815360.0, + "grad_norm": 4.979213421760356, + "learning_rate": 1.380928918696008e-06, + "loss": 0.6938, + "num_input_tokens_seen": 108311385, + "step": 5091 + }, + { + "epoch": 0.6122767991342511, + "flos": 11409114069120.0, + "grad_norm": 6.870810046907395, + "learning_rate": 1.3801882509773548e-06, + "loss": 0.6979, + "num_input_tokens_seen": 108328965, + "step": 5092 + }, + { + "epoch": 0.6123970420248903, + "flos": 20508540922440.0, + "grad_norm": 4.919040214769694, + "learning_rate": 1.3794476772836503e-06, + "loss": 0.7938, + "num_input_tokens_seen": 108349785, + "step": 5093 + }, + { + "epoch": 0.6125172849155294, + "flos": 15701205579240.0, + "grad_norm": 3.3982583185073203, + "learning_rate": 1.3787071977272402e-06, + "loss": 0.8233, + "num_input_tokens_seen": 108368765, + "step": 5094 + }, + { + "epoch": 0.6126375278061684, + "flos": 11837338146240.0, + "grad_norm": 9.029104855802169, + "learning_rate": 1.3779668124204535e-06, + "loss": 0.6907, + "num_input_tokens_seen": 108384900, + "step": 5095 + }, + { + "epoch": 0.6127577706968076, + "flos": 14946684530760.0, + "grad_norm": 4.610818100368686, + "learning_rate": 1.3772265214756074e-06, + "loss": 0.7956, + "num_input_tokens_seen": 108404380, + "step": 5096 + }, + { + "epoch": 0.6128780135874466, + "flos": 13327167138840.0, + "grad_norm": 3.855659580629075, + "learning_rate": 1.3764863250050025e-06, + "loss": 0.725, + "num_input_tokens_seen": 108422340, + "step": 5097 + }, + { + "epoch": 0.6129982564780857, + "flos": 18268338336120.0, + "grad_norm": 4.44623147217947, + "learning_rate": 1.3757462231209272e-06, + "loss": 0.79, + "num_input_tokens_seen": 108442365, + "step": 5098 + }, + { + "epoch": 0.6131184993687249, + "flos": 16458765983880.0, + "grad_norm": 7.637760428875501, + "learning_rate": 1.3750062159356525e-06, + "loss": 0.8679, + "num_input_tokens_seen": 108461435, + "step": 5099 + }, + { + "epoch": 0.6132387422593639, + "flos": 11569413158400.0, + "grad_norm": 6.692258193813656, + "learning_rate": 1.3742663035614382e-06, + "loss": 0.8118, + "num_input_tokens_seen": 108478525, + "step": 5100 + }, + { + "epoch": 0.613358985150003, + "flos": 18619483884720.0, + "grad_norm": 4.075765460617897, + "learning_rate": 1.3735264861105283e-06, + "loss": 0.7879, + "num_input_tokens_seen": 108498885, + "step": 5101 + }, + { + "epoch": 0.6134792280406421, + "flos": 15621974173440.0, + "grad_norm": 4.619699275810686, + "learning_rate": 1.372786763695152e-06, + "loss": 0.7723, + "num_input_tokens_seen": 108517365, + "step": 5102 + }, + { + "epoch": 0.6135994709312812, + "flos": 15510359119920.0, + "grad_norm": 4.4009577575750445, + "learning_rate": 1.3720471364275257e-06, + "loss": 0.7597, + "num_input_tokens_seen": 108536730, + "step": 5103 + }, + { + "epoch": 0.6137197138219203, + "flos": 10761807339720.0, + "grad_norm": 20.89628128479331, + "learning_rate": 1.3713076044198486e-06, + "loss": 0.7611, + "num_input_tokens_seen": 108553260, + "step": 5104 + }, + { + "epoch": 0.6138399567125594, + "flos": 14677113225000.0, + "grad_norm": 4.720110968562285, + "learning_rate": 1.3705681677843086e-06, + "loss": 0.7859, + "num_input_tokens_seen": 108571575, + "step": 5105 + }, + { + "epoch": 0.6139601996031985, + "flos": 44316107761560.0, + "grad_norm": 0.7994500982278879, + "learning_rate": 1.3698288266330768e-06, + "loss": 0.6221, + "num_input_tokens_seen": 108631920, + "step": 5106 + }, + { + "epoch": 0.6140804424938375, + "flos": 17268904599360.0, + "grad_norm": 5.402311621492497, + "learning_rate": 1.3690895810783113e-06, + "loss": 0.7111, + "num_input_tokens_seen": 108650435, + "step": 5107 + }, + { + "epoch": 0.6142006853844767, + "flos": 15649355547480.0, + "grad_norm": 5.551807542240994, + "learning_rate": 1.3683504312321543e-06, + "loss": 0.6962, + "num_input_tokens_seen": 108670490, + "step": 5108 + }, + { + "epoch": 0.6143209282751158, + "flos": 8736444971040.0, + "grad_norm": 4.463662405914521, + "learning_rate": 1.3676113772067355e-06, + "loss": 0.7763, + "num_input_tokens_seen": 108687265, + "step": 5109 + }, + { + "epoch": 0.6144411711657548, + "flos": 18375679295040.0, + "grad_norm": 5.002905382986786, + "learning_rate": 1.3668724191141671e-06, + "loss": 0.7041, + "num_input_tokens_seen": 108706255, + "step": 5110 + }, + { + "epoch": 0.6145614140563939, + "flos": 14698004307240.0, + "grad_norm": 5.911370212609539, + "learning_rate": 1.3661335570665493e-06, + "loss": 0.6535, + "num_input_tokens_seen": 108723885, + "step": 5111 + }, + { + "epoch": 0.614681656947033, + "flos": 12246570738720.0, + "grad_norm": 5.454999104332785, + "learning_rate": 1.3653947911759676e-06, + "loss": 0.6731, + "num_input_tokens_seen": 108741155, + "step": 5112 + }, + { + "epoch": 0.6148018998376721, + "flos": 28528986033600.0, + "grad_norm": 3.227209437356596, + "learning_rate": 1.3646561215544904e-06, + "loss": 0.7279, + "num_input_tokens_seen": 108765515, + "step": 5113 + }, + { + "epoch": 0.6149221427283111, + "flos": 17076126882480.0, + "grad_norm": 4.164054255092681, + "learning_rate": 1.363917548314176e-06, + "loss": 0.7765, + "num_input_tokens_seen": 108784500, + "step": 5114 + }, + { + "epoch": 0.6150423856189503, + "flos": 16375038863760.0, + "grad_norm": 4.175580953963243, + "learning_rate": 1.3631790715670626e-06, + "loss": 0.7127, + "num_input_tokens_seen": 108802625, + "step": 5115 + }, + { + "epoch": 0.6151626285095894, + "flos": 13645770739920.0, + "grad_norm": 4.374961852861035, + "learning_rate": 1.3624406914251783e-06, + "loss": 0.8427, + "num_input_tokens_seen": 108819465, + "step": 5116 + }, + { + "epoch": 0.6152828714002284, + "flos": 11544469601280.0, + "grad_norm": 3.890710178653108, + "learning_rate": 1.3617024080005335e-06, + "loss": 0.8561, + "num_input_tokens_seen": 108836085, + "step": 5117 + }, + { + "epoch": 0.6154031142908676, + "flos": 18219337700760.0, + "grad_norm": 5.484418993825062, + "learning_rate": 1.3609642214051266e-06, + "loss": 0.7271, + "num_input_tokens_seen": 108860030, + "step": 5118 + }, + { + "epoch": 0.6155233571815066, + "flos": 14055541551720.0, + "grad_norm": 6.928184785012825, + "learning_rate": 1.3602261317509385e-06, + "loss": 0.6462, + "num_input_tokens_seen": 108876410, + "step": 5119 + }, + { + "epoch": 0.6156436000721457, + "flos": 13704206043360.0, + "grad_norm": 6.462020838551288, + "learning_rate": 1.3594881391499387e-06, + "loss": 0.8022, + "num_input_tokens_seen": 108895045, + "step": 5120 + }, + { + "epoch": 0.6157638429627849, + "flos": 13164905132040.0, + "grad_norm": 2.5801298726831776, + "learning_rate": 1.3587502437140778e-06, + "loss": 0.7792, + "num_input_tokens_seen": 108912930, + "step": 5121 + }, + { + "epoch": 0.6158840858534239, + "flos": 18726824843640.0, + "grad_norm": 22.597581636112775, + "learning_rate": 1.3580124455552952e-06, + "loss": 0.8286, + "num_input_tokens_seen": 108932015, + "step": 5122 + }, + { + "epoch": 0.616004328744063, + "flos": 18048432524880.0, + "grad_norm": 2.902298733142639, + "learning_rate": 1.3572747447855148e-06, + "loss": 0.8515, + "num_input_tokens_seen": 108952145, + "step": 5123 + }, + { + "epoch": 0.6161245716347021, + "flos": 15865018924080.0, + "grad_norm": 3.285107334421424, + "learning_rate": 1.356537141516644e-06, + "loss": 0.6708, + "num_input_tokens_seen": 108969285, + "step": 5124 + }, + { + "epoch": 0.6162448145253412, + "flos": 25855557096480.0, + "grad_norm": 17.341936666185365, + "learning_rate": 1.3557996358605775e-06, + "loss": 0.6061, + "num_input_tokens_seen": 108988925, + "step": 5125 + }, + { + "epoch": 0.6163650574159802, + "flos": 15809749616640.0, + "grad_norm": 4.533161049729339, + "learning_rate": 1.3550622279291941e-06, + "loss": 0.6862, + "num_input_tokens_seen": 109006790, + "step": 5126 + }, + { + "epoch": 0.6164853003066194, + "flos": 17998260471000.0, + "grad_norm": 2.229745510445378, + "learning_rate": 1.354324917834358e-06, + "loss": 0.8162, + "num_input_tokens_seen": 109027755, + "step": 5127 + }, + { + "epoch": 0.6166055431972585, + "flos": 15973341341760.0, + "grad_norm": 2.7927850409424835, + "learning_rate": 1.353587705687918e-06, + "loss": 0.7554, + "num_input_tokens_seen": 109045650, + "step": 5128 + }, + { + "epoch": 0.6167257860878975, + "flos": 12975420051000.0, + "grad_norm": 6.317213659163931, + "learning_rate": 1.3528505916017096e-06, + "loss": 0.6906, + "num_input_tokens_seen": 109063070, + "step": 5129 + }, + { + "epoch": 0.6168460289785367, + "flos": 16992684702000.0, + "grad_norm": 4.0864483895019985, + "learning_rate": 1.3521135756875514e-06, + "loss": 0.8602, + "num_input_tokens_seen": 109079105, + "step": 5130 + }, + { + "epoch": 0.6169662718691757, + "flos": 19212692725200.0, + "grad_norm": 3.9155940487697065, + "learning_rate": 1.3513766580572496e-06, + "loss": 0.8429, + "num_input_tokens_seen": 109101645, + "step": 5131 + }, + { + "epoch": 0.6170865147598148, + "flos": 13893786104280.0, + "grad_norm": 6.8065949491831885, + "learning_rate": 1.3506398388225924e-06, + "loss": 0.7555, + "num_input_tokens_seen": 109118685, + "step": 5132 + }, + { + "epoch": 0.617206757650454, + "flos": 13327135478880.0, + "grad_norm": 5.6525484059896085, + "learning_rate": 1.349903118095355e-06, + "loss": 0.6965, + "num_input_tokens_seen": 109137540, + "step": 5133 + }, + { + "epoch": 0.617327000541093, + "flos": 13272752650320.0, + "grad_norm": 3.8646220627700862, + "learning_rate": 1.349166495987298e-06, + "loss": 0.7139, + "num_input_tokens_seen": 109155825, + "step": 5134 + }, + { + "epoch": 0.6174472434317321, + "flos": 45580902029400.0, + "grad_norm": 0.9389280251491711, + "learning_rate": 1.348429972610166e-06, + "loss": 0.6685, + "num_input_tokens_seen": 109219850, + "step": 5135 + }, + { + "epoch": 0.6175674863223712, + "flos": 52533156953400.0, + "grad_norm": 0.860141724230831, + "learning_rate": 1.3476935480756897e-06, + "loss": 0.5994, + "num_input_tokens_seen": 109276320, + "step": 5136 + }, + { + "epoch": 0.6176877292130103, + "flos": 15972866442360.0, + "grad_norm": 5.415998896983273, + "learning_rate": 1.346957222495583e-06, + "loss": 0.7396, + "num_input_tokens_seen": 109293835, + "step": 5137 + }, + { + "epoch": 0.6178079721036493, + "flos": 12942213244320.0, + "grad_norm": 9.22240062781073, + "learning_rate": 1.3462209959815466e-06, + "loss": 0.7077, + "num_input_tokens_seen": 109308295, + "step": 5138 + }, + { + "epoch": 0.6179282149942885, + "flos": 16562497707360.0, + "grad_norm": 4.265470956109887, + "learning_rate": 1.345484868645265e-06, + "loss": 0.7157, + "num_input_tokens_seen": 109326825, + "step": 5139 + }, + { + "epoch": 0.6180484578849276, + "flos": 16324391910480.0, + "grad_norm": 3.5813409604010054, + "learning_rate": 1.3447488405984088e-06, + "loss": 0.7676, + "num_input_tokens_seen": 109344805, + "step": 5140 + }, + { + "epoch": 0.6181687007755666, + "flos": 25882336931280.0, + "grad_norm": 4.329977252010648, + "learning_rate": 1.3440129119526322e-06, + "loss": 0.6708, + "num_input_tokens_seen": 109366950, + "step": 5141 + }, + { + "epoch": 0.6182889436662057, + "flos": 45364537642320.0, + "grad_norm": 0.9350067044594029, + "learning_rate": 1.3432770828195762e-06, + "loss": 0.5589, + "num_input_tokens_seen": 109427655, + "step": 5142 + }, + { + "epoch": 0.6184091865568448, + "flos": 14325904356480.0, + "grad_norm": 5.027206786571562, + "learning_rate": 1.3425413533108635e-06, + "loss": 0.7036, + "num_input_tokens_seen": 109445975, + "step": 5143 + }, + { + "epoch": 0.6185294294474839, + "flos": 17269062899160.0, + "grad_norm": 4.098379764012051, + "learning_rate": 1.341805723538105e-06, + "loss": 0.6873, + "num_input_tokens_seen": 109465800, + "step": 5144 + }, + { + "epoch": 0.618649672338123, + "flos": 19619012601360.0, + "grad_norm": 19.116882596158625, + "learning_rate": 1.3410701936128948e-06, + "loss": 0.7573, + "num_input_tokens_seen": 109488300, + "step": 5145 + }, + { + "epoch": 0.6187699152287621, + "flos": 10510815939120.0, + "grad_norm": 7.420978307262812, + "learning_rate": 1.340334763646812e-06, + "loss": 0.8353, + "num_input_tokens_seen": 109502155, + "step": 5146 + }, + { + "epoch": 0.6188901581194012, + "flos": 14677334844720.0, + "grad_norm": 8.75916268305385, + "learning_rate": 1.3395994337514218e-06, + "loss": 0.7321, + "num_input_tokens_seen": 109522045, + "step": 5147 + }, + { + "epoch": 0.6190104010100402, + "flos": 18833121023880.0, + "grad_norm": 2.88740863738079, + "learning_rate": 1.3388642040382725e-06, + "loss": 0.7755, + "num_input_tokens_seen": 109542190, + "step": 5148 + }, + { + "epoch": 0.6191306439006794, + "flos": 22342043713080.0, + "grad_norm": 2.8513093182766425, + "learning_rate": 1.3381290746188975e-06, + "loss": 0.8184, + "num_input_tokens_seen": 109561280, + "step": 5149 + }, + { + "epoch": 0.6192508867913185, + "flos": 19562128635960.0, + "grad_norm": 16.87990199345345, + "learning_rate": 1.3373940456048152e-06, + "loss": 0.6599, + "num_input_tokens_seen": 109581025, + "step": 5150 + }, + { + "epoch": 0.6193711296819575, + "flos": 26989998105840.0, + "grad_norm": 2.8369698655225966, + "learning_rate": 1.3366591171075299e-06, + "loss": 0.5801, + "num_input_tokens_seen": 109604250, + "step": 5151 + }, + { + "epoch": 0.6194913725725967, + "flos": 18457411837680.0, + "grad_norm": 3.2135676172198924, + "learning_rate": 1.335924289238529e-06, + "loss": 0.8961, + "num_input_tokens_seen": 109623180, + "step": 5152 + }, + { + "epoch": 0.6196116154632357, + "flos": 15486935240880.0, + "grad_norm": 2.713088443758516, + "learning_rate": 1.3351895621092859e-06, + "loss": 0.744, + "num_input_tokens_seen": 109643245, + "step": 5153 + }, + { + "epoch": 0.6197318583538748, + "flos": 11842118800200.0, + "grad_norm": 6.048891505056228, + "learning_rate": 1.3344549358312567e-06, + "loss": 0.7494, + "num_input_tokens_seen": 109661365, + "step": 5154 + }, + { + "epoch": 0.619852101244514, + "flos": 17889716433600.0, + "grad_norm": 7.012031297438694, + "learning_rate": 1.3337204105158852e-06, + "loss": 0.7587, + "num_input_tokens_seen": 109679955, + "step": 5155 + }, + { + "epoch": 0.619972344135153, + "flos": 12192472849800.0, + "grad_norm": 3.400064854844813, + "learning_rate": 1.332985986274597e-06, + "loss": 0.6987, + "num_input_tokens_seen": 109697305, + "step": 5156 + }, + { + "epoch": 0.6200925870257921, + "flos": 9060462425280.0, + "grad_norm": 3.1712724085081123, + "learning_rate": 1.3322516632188047e-06, + "loss": 0.7394, + "num_input_tokens_seen": 109713920, + "step": 5157 + }, + { + "epoch": 0.6202128299164312, + "flos": 19453679578440.0, + "grad_norm": 3.2190009454475144, + "learning_rate": 1.3315174414599045e-06, + "loss": 0.6607, + "num_input_tokens_seen": 109734960, + "step": 5158 + }, + { + "epoch": 0.6203330728070703, + "flos": 13704585962880.0, + "grad_norm": 2.8751268773485545, + "learning_rate": 1.3307833211092768e-06, + "loss": 0.7434, + "num_input_tokens_seen": 109753345, + "step": 5159 + }, + { + "epoch": 0.6204533156977093, + "flos": 15080362085040.0, + "grad_norm": 2.490421255242665, + "learning_rate": 1.3300493022782873e-06, + "loss": 0.7293, + "num_input_tokens_seen": 109773635, + "step": 5160 + }, + { + "epoch": 0.6205735585883485, + "flos": 12651940816080.0, + "grad_norm": 4.588159182220739, + "learning_rate": 1.3293153850782855e-06, + "loss": 0.7029, + "num_input_tokens_seen": 109791675, + "step": 5161 + }, + { + "epoch": 0.6206938014789876, + "flos": 16808613474120.0, + "grad_norm": 2.9096833324842954, + "learning_rate": 1.3285815696206069e-06, + "loss": 0.6955, + "num_input_tokens_seen": 109812940, + "step": 5162 + }, + { + "epoch": 0.6208140443696266, + "flos": 17561646504480.0, + "grad_norm": 4.345061179763041, + "learning_rate": 1.32784785601657e-06, + "loss": 0.7568, + "num_input_tokens_seen": 109832070, + "step": 5163 + }, + { + "epoch": 0.6209342872602658, + "flos": 25772368195680.0, + "grad_norm": 3.2732785716658843, + "learning_rate": 1.3271142443774798e-06, + "loss": 0.7326, + "num_input_tokens_seen": 109854025, + "step": 5164 + }, + { + "epoch": 0.6210545301509048, + "flos": 19779596630280.0, + "grad_norm": 7.0790643131909015, + "learning_rate": 1.3263807348146228e-06, + "loss": 0.8001, + "num_input_tokens_seen": 109873600, + "step": 5165 + }, + { + "epoch": 0.6211747730415439, + "flos": 24692626614480.0, + "grad_norm": 4.7482666065209065, + "learning_rate": 1.3256473274392733e-06, + "loss": 0.7235, + "num_input_tokens_seen": 109894665, + "step": 5166 + }, + { + "epoch": 0.6212950159321831, + "flos": 25099041470520.0, + "grad_norm": 3.4198581552369856, + "learning_rate": 1.3249140223626873e-06, + "loss": 0.6892, + "num_input_tokens_seen": 109916005, + "step": 5167 + }, + { + "epoch": 0.6214152588228221, + "flos": 20509807320840.0, + "grad_norm": 2.9581424132156373, + "learning_rate": 1.3241808196961077e-06, + "loss": 0.748, + "num_input_tokens_seen": 109936850, + "step": 5168 + }, + { + "epoch": 0.6215355017134612, + "flos": 14784897423360.0, + "grad_norm": 3.7452978535709134, + "learning_rate": 1.3234477195507608e-06, + "loss": 0.6903, + "num_input_tokens_seen": 109955400, + "step": 5169 + }, + { + "epoch": 0.6216557446041003, + "flos": 30472742499480.0, + "grad_norm": 4.830912257698038, + "learning_rate": 1.322714722037857e-06, + "loss": 0.618, + "num_input_tokens_seen": 109976565, + "step": 5170 + }, + { + "epoch": 0.6217759874947394, + "flos": 20482362626880.0, + "grad_norm": 4.503976971880787, + "learning_rate": 1.321981827268591e-06, + "loss": 0.7501, + "num_input_tokens_seen": 109996940, + "step": 5171 + }, + { + "epoch": 0.6218962303853784, + "flos": 15921681269760.0, + "grad_norm": 2.9475191385231834, + "learning_rate": 1.3212490353541426e-06, + "loss": 0.8048, + "num_input_tokens_seen": 110018920, + "step": 5172 + }, + { + "epoch": 0.6220164732760175, + "flos": 15536062516080.0, + "grad_norm": 5.276483856780872, + "learning_rate": 1.3205163464056762e-06, + "loss": 0.787, + "num_input_tokens_seen": 110035245, + "step": 5173 + }, + { + "epoch": 0.6221367161666567, + "flos": 19155207220560.0, + "grad_norm": 4.3613394956188545, + "learning_rate": 1.319783760534339e-06, + "loss": 0.7061, + "num_input_tokens_seen": 110054210, + "step": 5174 + }, + { + "epoch": 0.6222569590572957, + "flos": 11863801381440.0, + "grad_norm": 3.749776960733668, + "learning_rate": 1.319051277851266e-06, + "loss": 0.7381, + "num_input_tokens_seen": 110070215, + "step": 5175 + }, + { + "epoch": 0.6223772019479348, + "flos": 13300102364400.0, + "grad_norm": 5.263796525657635, + "learning_rate": 1.3183188984675716e-06, + "loss": 0.8299, + "num_input_tokens_seen": 110088300, + "step": 5176 + }, + { + "epoch": 0.6224974448385739, + "flos": 20157332053920.0, + "grad_norm": 5.353657422419823, + "learning_rate": 1.3175866224943586e-06, + "loss": 0.7128, + "num_input_tokens_seen": 110106740, + "step": 5177 + }, + { + "epoch": 0.622617687729213, + "flos": 14460500049600.0, + "grad_norm": 4.957681287104571, + "learning_rate": 1.316854450042712e-06, + "loss": 0.7152, + "num_input_tokens_seen": 110124400, + "step": 5178 + }, + { + "epoch": 0.622737930619852, + "flos": 16863407882160.0, + "grad_norm": 12.460973623576166, + "learning_rate": 1.3161223812237024e-06, + "loss": 0.7327, + "num_input_tokens_seen": 110143475, + "step": 5179 + }, + { + "epoch": 0.6228581735104912, + "flos": 9162421191000.0, + "grad_norm": 6.850034850094543, + "learning_rate": 1.3153904161483842e-06, + "loss": 0.8293, + "num_input_tokens_seen": 110158495, + "step": 5180 + }, + { + "epoch": 0.6229784164011303, + "flos": 17428443849600.0, + "grad_norm": 3.155235958555651, + "learning_rate": 1.3146585549277953e-06, + "loss": 0.8375, + "num_input_tokens_seen": 110176855, + "step": 5181 + }, + { + "epoch": 0.6230986592917693, + "flos": 16400488980240.0, + "grad_norm": 4.382594900458339, + "learning_rate": 1.3139267976729591e-06, + "loss": 0.763, + "num_input_tokens_seen": 110196765, + "step": 5182 + }, + { + "epoch": 0.6232189021824085, + "flos": 25365763379880.0, + "grad_norm": 3.802590546241565, + "learning_rate": 1.3131951444948815e-06, + "loss": 0.704, + "num_input_tokens_seen": 110215885, + "step": 5183 + }, + { + "epoch": 0.6233391450730476, + "flos": 16270072401840.0, + "grad_norm": 3.91065681599433, + "learning_rate": 1.3124635955045546e-06, + "loss": 0.7294, + "num_input_tokens_seen": 110235420, + "step": 5184 + }, + { + "epoch": 0.6234593879636866, + "flos": 14859759754680.0, + "grad_norm": 4.869802002377345, + "learning_rate": 1.3117321508129537e-06, + "loss": 0.8265, + "num_input_tokens_seen": 110253220, + "step": 5185 + }, + { + "epoch": 0.6235796308543258, + "flos": 15106160461080.0, + "grad_norm": 3.422145049008033, + "learning_rate": 1.3110008105310388e-06, + "loss": 0.7551, + "num_input_tokens_seen": 110272760, + "step": 5186 + }, + { + "epoch": 0.6236998737449648, + "flos": 19513254640440.0, + "grad_norm": 3.681673392800133, + "learning_rate": 1.3102695747697526e-06, + "loss": 0.7691, + "num_input_tokens_seen": 110295350, + "step": 5187 + }, + { + "epoch": 0.6238201166356039, + "flos": 9192810261240.0, + "grad_norm": 3.6838266212098363, + "learning_rate": 1.3095384436400237e-06, + "loss": 0.8759, + "num_input_tokens_seen": 110306600, + "step": 5188 + }, + { + "epoch": 0.623940359526243, + "flos": 7548507612000.0, + "grad_norm": 4.626736289241645, + "learning_rate": 1.3088074172527633e-06, + "loss": 0.805, + "num_input_tokens_seen": 110323450, + "step": 5189 + }, + { + "epoch": 0.6240606024168821, + "flos": 21319154437320.0, + "grad_norm": 5.9582564697998786, + "learning_rate": 1.3080764957188684e-06, + "loss": 0.7021, + "num_input_tokens_seen": 110343415, + "step": 5190 + }, + { + "epoch": 0.6241808453075212, + "flos": 16108063674720.0, + "grad_norm": 3.975401703312227, + "learning_rate": 1.3073456791492192e-06, + "loss": 0.6885, + "num_input_tokens_seen": 110362845, + "step": 5191 + }, + { + "epoch": 0.6243010881981603, + "flos": 15456736130400.0, + "grad_norm": 4.589542252398147, + "learning_rate": 1.3066149676546801e-06, + "loss": 0.7732, + "num_input_tokens_seen": 110380745, + "step": 5192 + }, + { + "epoch": 0.6244213310887994, + "flos": 16349747047080.0, + "grad_norm": 5.917113632011038, + "learning_rate": 1.3058843613460985e-06, + "loss": 0.6451, + "num_input_tokens_seen": 110398405, + "step": 5193 + }, + { + "epoch": 0.6245415739794384, + "flos": 11085919773840.0, + "grad_norm": 3.9791763199028374, + "learning_rate": 1.3051538603343075e-06, + "loss": 0.7216, + "num_input_tokens_seen": 110416055, + "step": 5194 + }, + { + "epoch": 0.6246618168700776, + "flos": 13785527006520.0, + "grad_norm": 3.1208109435096185, + "learning_rate": 1.3044234647301235e-06, + "loss": 0.6613, + "num_input_tokens_seen": 110433800, + "step": 5195 + }, + { + "epoch": 0.6247820597607167, + "flos": 10410313531560.0, + "grad_norm": 6.933642165323922, + "learning_rate": 1.303693174644347e-06, + "loss": 0.6998, + "num_input_tokens_seen": 110450995, + "step": 5196 + }, + { + "epoch": 0.6249023026513557, + "flos": 16566771801960.0, + "grad_norm": 5.009141195246548, + "learning_rate": 1.3029629901877625e-06, + "loss": 0.79, + "num_input_tokens_seen": 110470090, + "step": 5197 + }, + { + "epoch": 0.6250225455419949, + "flos": 14811898877880.0, + "grad_norm": 4.0072461298764965, + "learning_rate": 1.3022329114711376e-06, + "loss": 0.7672, + "num_input_tokens_seen": 110488520, + "step": 5198 + }, + { + "epoch": 0.6251427884326339, + "flos": 17158334324520.0, + "grad_norm": 4.21252071930179, + "learning_rate": 1.3015029386052256e-06, + "loss": 0.6805, + "num_input_tokens_seen": 110508410, + "step": 5199 + }, + { + "epoch": 0.625263031323273, + "flos": 23290355597160.0, + "grad_norm": 3.428299677387978, + "learning_rate": 1.3007730717007622e-06, + "loss": 0.7067, + "num_input_tokens_seen": 110528945, + "step": 5200 + }, + { + "epoch": 0.6253832742139122, + "flos": 17674243016760.0, + "grad_norm": 3.8596655855938073, + "learning_rate": 1.3000433108684676e-06, + "loss": 0.7414, + "num_input_tokens_seen": 110549165, + "step": 5201 + }, + { + "epoch": 0.6255035171045512, + "flos": 20289331630320.0, + "grad_norm": 7.203238860779866, + "learning_rate": 1.2993136562190467e-06, + "loss": 0.7874, + "num_input_tokens_seen": 110568005, + "step": 5202 + }, + { + "epoch": 0.6256237599951903, + "flos": 14782776206040.0, + "grad_norm": 2.9608017379516514, + "learning_rate": 1.2985841078631871e-06, + "loss": 0.6854, + "num_input_tokens_seen": 110587045, + "step": 5203 + }, + { + "epoch": 0.6257440028858293, + "flos": 17701022851560.0, + "grad_norm": 6.918776531220303, + "learning_rate": 1.2978546659115608e-06, + "loss": 0.7699, + "num_input_tokens_seen": 110604845, + "step": 5204 + }, + { + "epoch": 0.6258642457764685, + "flos": 11544691221000.0, + "grad_norm": 2.8920841967301993, + "learning_rate": 1.2971253304748228e-06, + "loss": 0.8435, + "num_input_tokens_seen": 110622280, + "step": 5205 + }, + { + "epoch": 0.6259844886671075, + "flos": 8628375833040.0, + "grad_norm": 6.798278242809012, + "learning_rate": 1.296396101663614e-06, + "loss": 0.7359, + "num_input_tokens_seen": 110638560, + "step": 5206 + }, + { + "epoch": 0.6261047315577466, + "flos": 11571281096040.0, + "grad_norm": 3.822592462552819, + "learning_rate": 1.2956669795885565e-06, + "loss": 0.8268, + "num_input_tokens_seen": 110654910, + "step": 5207 + }, + { + "epoch": 0.6262249744483858, + "flos": 22939494988200.0, + "grad_norm": 4.998797844373787, + "learning_rate": 1.294937964360259e-06, + "loss": 0.6774, + "num_input_tokens_seen": 110674900, + "step": 5208 + }, + { + "epoch": 0.6263452173390248, + "flos": 19941921957000.0, + "grad_norm": 9.525983912795358, + "learning_rate": 1.2942090560893108e-06, + "loss": 0.6767, + "num_input_tokens_seen": 110694025, + "step": 5209 + }, + { + "epoch": 0.6264654602296639, + "flos": 27448832872920.0, + "grad_norm": 4.297463116712225, + "learning_rate": 1.2934802548862882e-06, + "loss": 0.5872, + "num_input_tokens_seen": 110716530, + "step": 5210 + }, + { + "epoch": 0.626585703120303, + "flos": 10734267665880.0, + "grad_norm": 4.1510849603033, + "learning_rate": 1.292751560861749e-06, + "loss": 0.818, + "num_input_tokens_seen": 110731155, + "step": 5211 + }, + { + "epoch": 0.6267059460109421, + "flos": 16351836604440.0, + "grad_norm": 3.495645266275844, + "learning_rate": 1.2920229741262354e-06, + "loss": 0.7806, + "num_input_tokens_seen": 110748880, + "step": 5212 + }, + { + "epoch": 0.6268261889015811, + "flos": 12948450256440.0, + "grad_norm": 8.712522627638299, + "learning_rate": 1.2912944947902739e-06, + "loss": 0.7392, + "num_input_tokens_seen": 110765085, + "step": 5213 + }, + { + "epoch": 0.6269464317922203, + "flos": 24120688775760.0, + "grad_norm": 12.94557711803322, + "learning_rate": 1.2905661229643742e-06, + "loss": 0.6901, + "num_input_tokens_seen": 110784565, + "step": 5214 + }, + { + "epoch": 0.6270666746828594, + "flos": 13081652911320.0, + "grad_norm": 4.264027949801625, + "learning_rate": 1.2898378587590299e-06, + "loss": 0.8085, + "num_input_tokens_seen": 110800885, + "step": 5215 + }, + { + "epoch": 0.6271869175734984, + "flos": 12732786879840.0, + "grad_norm": 4.461879543906475, + "learning_rate": 1.2891097022847173e-06, + "loss": 0.8555, + "num_input_tokens_seen": 110817950, + "step": 5216 + }, + { + "epoch": 0.6273071604641376, + "flos": 19697484168120.0, + "grad_norm": 4.101445310825845, + "learning_rate": 1.2883816536518978e-06, + "loss": 0.662, + "num_input_tokens_seen": 110838810, + "step": 5217 + }, + { + "epoch": 0.6274274033547766, + "flos": 19100317832640.0, + "grad_norm": 4.766773523231133, + "learning_rate": 1.2876537129710155e-06, + "loss": 0.8069, + "num_input_tokens_seen": 110856260, + "step": 5218 + }, + { + "epoch": 0.6275476462454157, + "flos": 14811518958360.0, + "grad_norm": 5.669363427329613, + "learning_rate": 1.286925880352499e-06, + "loss": 0.7298, + "num_input_tokens_seen": 110874840, + "step": 5219 + }, + { + "epoch": 0.6276678891360549, + "flos": 19291639191360.0, + "grad_norm": 2.5907441152100694, + "learning_rate": 1.2861981559067592e-06, + "loss": 0.6857, + "num_input_tokens_seen": 110895165, + "step": 5220 + }, + { + "epoch": 0.6277881320266939, + "flos": 10108580197800.0, + "grad_norm": 3.699476674160643, + "learning_rate": 1.2854705397441917e-06, + "loss": 0.7897, + "num_input_tokens_seen": 110910425, + "step": 5221 + }, + { + "epoch": 0.627908374917333, + "flos": 19830623503080.0, + "grad_norm": 6.584887293008218, + "learning_rate": 1.2847430319751747e-06, + "loss": 0.7555, + "num_input_tokens_seen": 110928240, + "step": 5222 + }, + { + "epoch": 0.6280286178079721, + "flos": 17403975191880.0, + "grad_norm": 5.426099730798746, + "learning_rate": 1.2840156327100712e-06, + "loss": 0.6725, + "num_input_tokens_seen": 110945085, + "step": 5223 + }, + { + "epoch": 0.6281488606986112, + "flos": 19320635223360.0, + "grad_norm": 3.4276264228015147, + "learning_rate": 1.2832883420592272e-06, + "loss": 0.7058, + "num_input_tokens_seen": 110963700, + "step": 5224 + }, + { + "epoch": 0.6282691035892503, + "flos": 26557025034720.0, + "grad_norm": 7.235731883330889, + "learning_rate": 1.282561160132972e-06, + "loss": 0.6213, + "num_input_tokens_seen": 110983940, + "step": 5225 + }, + { + "epoch": 0.6283893464798894, + "flos": 19451811640800.0, + "grad_norm": 2.5519233705382143, + "learning_rate": 1.2818340870416186e-06, + "loss": 0.8004, + "num_input_tokens_seen": 111004795, + "step": 5226 + }, + { + "epoch": 0.6285095893705285, + "flos": 16270262361600.0, + "grad_norm": 8.415717597052543, + "learning_rate": 1.2811071228954626e-06, + "loss": 0.7427, + "num_input_tokens_seen": 111023150, + "step": 5227 + }, + { + "epoch": 0.6286298322611675, + "flos": 19456244035200.0, + "grad_norm": 3.840296737013992, + "learning_rate": 1.2803802678047846e-06, + "loss": 0.7978, + "num_input_tokens_seen": 111043020, + "step": 5228 + }, + { + "epoch": 0.6287500751518067, + "flos": 15942572352000.0, + "grad_norm": 8.162588645371926, + "learning_rate": 1.279653521879848e-06, + "loss": 0.7158, + "num_input_tokens_seen": 111062805, + "step": 5229 + }, + { + "epoch": 0.6288703180424458, + "flos": 14621400678120.0, + "grad_norm": 5.282607767402266, + "learning_rate": 1.2789268852308997e-06, + "loss": 0.8273, + "num_input_tokens_seen": 111077735, + "step": 5230 + }, + { + "epoch": 0.6289905609330848, + "flos": 16186756861200.0, + "grad_norm": 5.424884939072243, + "learning_rate": 1.2782003579681688e-06, + "loss": 0.6864, + "num_input_tokens_seen": 111096985, + "step": 5231 + }, + { + "epoch": 0.629110803823724, + "flos": 18698272051080.0, + "grad_norm": 2.607225274950053, + "learning_rate": 1.2774739402018701e-06, + "loss": 0.732, + "num_input_tokens_seen": 111117540, + "step": 5232 + }, + { + "epoch": 0.629231046714363, + "flos": 15270638665080.0, + "grad_norm": 6.010138846267698, + "learning_rate": 1.2767476320422002e-06, + "loss": 0.7297, + "num_input_tokens_seen": 111137185, + "step": 5233 + }, + { + "epoch": 0.6293512896050021, + "flos": 47961905661000.0, + "grad_norm": 0.7148862732293392, + "learning_rate": 1.2760214335993392e-06, + "loss": 0.5898, + "num_input_tokens_seen": 111203550, + "step": 5234 + }, + { + "epoch": 0.6294715324956413, + "flos": 25477030173840.0, + "grad_norm": 3.601900097297687, + "learning_rate": 1.2752953449834514e-06, + "loss": 0.5641, + "num_input_tokens_seen": 111225720, + "step": 5235 + }, + { + "epoch": 0.6295917753862803, + "flos": 16675030899720.0, + "grad_norm": 8.82770843873408, + "learning_rate": 1.2745693663046836e-06, + "loss": 0.7804, + "num_input_tokens_seen": 111244510, + "step": 5236 + }, + { + "epoch": 0.6297120182769194, + "flos": 15243922150200.0, + "grad_norm": 5.383337429899344, + "learning_rate": 1.2738434976731662e-06, + "loss": 0.7957, + "num_input_tokens_seen": 111262415, + "step": 5237 + }, + { + "epoch": 0.6298322611675584, + "flos": 14242367196120.0, + "grad_norm": 4.241470079544231, + "learning_rate": 1.2731177391990125e-06, + "loss": 0.7366, + "num_input_tokens_seen": 111282060, + "step": 5238 + }, + { + "epoch": 0.6299525040581976, + "flos": 9141308489040.0, + "grad_norm": 9.977921718171817, + "learning_rate": 1.2723920909923203e-06, + "loss": 0.8036, + "num_input_tokens_seen": 111297525, + "step": 5239 + }, + { + "epoch": 0.6300727469488366, + "flos": 42537553187400.0, + "grad_norm": 0.9092350173651357, + "learning_rate": 1.2716665531631688e-06, + "loss": 0.6462, + "num_input_tokens_seen": 111351530, + "step": 5240 + }, + { + "epoch": 0.6301929898394757, + "flos": 16484690999760.0, + "grad_norm": 3.374247007788237, + "learning_rate": 1.270941125821623e-06, + "loss": 0.7548, + "num_input_tokens_seen": 111371675, + "step": 5241 + }, + { + "epoch": 0.6303132327301149, + "flos": 20752313852160.0, + "grad_norm": 3.144261773925051, + "learning_rate": 1.2702158090777278e-06, + "loss": 0.7441, + "num_input_tokens_seen": 111392485, + "step": 5242 + }, + { + "epoch": 0.6304334756207539, + "flos": 18510528267840.0, + "grad_norm": 3.1783487936949784, + "learning_rate": 1.2694906030415148e-06, + "loss": 0.7319, + "num_input_tokens_seen": 111409950, + "step": 5243 + }, + { + "epoch": 0.630553718511393, + "flos": 13158161560560.0, + "grad_norm": 15.941659900579431, + "learning_rate": 1.2687655078229958e-06, + "loss": 0.7882, + "num_input_tokens_seen": 111424000, + "step": 5244 + }, + { + "epoch": 0.6306739614020321, + "flos": 20019443724960.0, + "grad_norm": 4.5174603279875525, + "learning_rate": 1.2680405235321678e-06, + "loss": 0.6932, + "num_input_tokens_seen": 111445055, + "step": 5245 + }, + { + "epoch": 0.6307942042926712, + "flos": 11166924137400.0, + "grad_norm": 3.693351499487571, + "learning_rate": 1.267315650279011e-06, + "loss": 0.7715, + "num_input_tokens_seen": 111463245, + "step": 5246 + }, + { + "epoch": 0.6309144471833102, + "flos": 14322738360480.0, + "grad_norm": 10.08637962879215, + "learning_rate": 1.2665908881734874e-06, + "loss": 0.7217, + "num_input_tokens_seen": 111481800, + "step": 5247 + }, + { + "epoch": 0.6310346900739494, + "flos": 12759376754880.0, + "grad_norm": 3.473349056928291, + "learning_rate": 1.2658662373255432e-06, + "loss": 0.8339, + "num_input_tokens_seen": 111499910, + "step": 5248 + }, + { + "epoch": 0.6311549329645885, + "flos": 40573412198640.0, + "grad_norm": 0.8342714195393883, + "learning_rate": 1.2651416978451063e-06, + "loss": 0.5641, + "num_input_tokens_seen": 111565015, + "step": 5249 + }, + { + "epoch": 0.6312751758552275, + "flos": 30823856388120.0, + "grad_norm": 4.2291265526483635, + "learning_rate": 1.2644172698420903e-06, + "loss": 0.6373, + "num_input_tokens_seen": 111586840, + "step": 5250 + }, + { + "epoch": 0.6313954187458667, + "flos": 14353190750640.0, + "grad_norm": 4.670386648243569, + "learning_rate": 1.2636929534263892e-06, + "loss": 0.828, + "num_input_tokens_seen": 111605545, + "step": 5251 + }, + { + "epoch": 0.6315156616365057, + "flos": 16756573482600.0, + "grad_norm": 2.705504059568034, + "learning_rate": 1.2629687487078821e-06, + "loss": 0.757, + "num_input_tokens_seen": 111624075, + "step": 5252 + }, + { + "epoch": 0.6316359045271448, + "flos": 17371274944560.0, + "grad_norm": 4.14559833620981, + "learning_rate": 1.2622446557964293e-06, + "loss": 0.7562, + "num_input_tokens_seen": 111641800, + "step": 5253 + }, + { + "epoch": 0.631756147417784, + "flos": 24315461070120.0, + "grad_norm": 7.045236468925381, + "learning_rate": 1.261520674801876e-06, + "loss": 0.6964, + "num_input_tokens_seen": 111662115, + "step": 5254 + }, + { + "epoch": 0.631876390308423, + "flos": 22939051748760.0, + "grad_norm": 4.062315716598507, + "learning_rate": 1.2607968058340488e-06, + "loss": 0.7053, + "num_input_tokens_seen": 111681530, + "step": 5255 + }, + { + "epoch": 0.6319966331990621, + "flos": 18078853255080.0, + "grad_norm": 3.1428509315825677, + "learning_rate": 1.2600730490027583e-06, + "loss": 0.7158, + "num_input_tokens_seen": 111701490, + "step": 5256 + }, + { + "epoch": 0.6321168760897012, + "flos": 12757983716640.0, + "grad_norm": 3.3689671037743123, + "learning_rate": 1.2593494044177984e-06, + "loss": 0.7958, + "num_input_tokens_seen": 111719515, + "step": 5257 + }, + { + "epoch": 0.6322371189803403, + "flos": 13352838875040.0, + "grad_norm": 4.011734915072106, + "learning_rate": 1.2586258721889448e-06, + "loss": 0.7832, + "num_input_tokens_seen": 111736585, + "step": 5258 + }, + { + "epoch": 0.6323573618709794, + "flos": 14730926174280.0, + "grad_norm": 4.848377207984473, + "learning_rate": 1.2579024524259573e-06, + "loss": 0.7989, + "num_input_tokens_seen": 111752565, + "step": 5259 + }, + { + "epoch": 0.6324776047616185, + "flos": 14645141156760.0, + "grad_norm": 3.1417671518083004, + "learning_rate": 1.2571791452385768e-06, + "loss": 0.9013, + "num_input_tokens_seen": 111769550, + "step": 5260 + }, + { + "epoch": 0.6325978476522576, + "flos": 22643048867760.0, + "grad_norm": 3.3108748281973335, + "learning_rate": 1.2564559507365301e-06, + "loss": 0.7518, + "num_input_tokens_seen": 111791675, + "step": 5261 + }, + { + "epoch": 0.6327180905428966, + "flos": 17970752457120.0, + "grad_norm": 4.517685364053587, + "learning_rate": 1.2557328690295244e-06, + "loss": 0.77, + "num_input_tokens_seen": 111809585, + "step": 5262 + }, + { + "epoch": 0.6328383334335358, + "flos": 15779645486040.0, + "grad_norm": 5.710706045989513, + "learning_rate": 1.255009900227251e-06, + "loss": 0.7482, + "num_input_tokens_seen": 111828330, + "step": 5263 + }, + { + "epoch": 0.6329585763241748, + "flos": 16782783438120.0, + "grad_norm": 4.074241455248933, + "learning_rate": 1.254287044439383e-06, + "loss": 0.7847, + "num_input_tokens_seen": 111847655, + "step": 5264 + }, + { + "epoch": 0.6330788192148139, + "flos": 52315214059680.0, + "grad_norm": 0.9693041416343804, + "learning_rate": 1.2535643017755776e-06, + "loss": 0.5609, + "num_input_tokens_seen": 111909565, + "step": 5265 + }, + { + "epoch": 0.6331990621054531, + "flos": 15534511178040.0, + "grad_norm": 15.201682670213168, + "learning_rate": 1.2528416723454737e-06, + "loss": 0.7038, + "num_input_tokens_seen": 111925955, + "step": 5266 + }, + { + "epoch": 0.6333193049960921, + "flos": 25315433026200.0, + "grad_norm": 5.516175672150091, + "learning_rate": 1.2521191562586945e-06, + "loss": 0.6958, + "num_input_tokens_seen": 111949325, + "step": 5267 + }, + { + "epoch": 0.6334395478867312, + "flos": 13378099031760.0, + "grad_norm": 3.656245435402132, + "learning_rate": 1.2513967536248445e-06, + "loss": 0.7555, + "num_input_tokens_seen": 111965365, + "step": 5268 + }, + { + "epoch": 0.6335597907773702, + "flos": 17295874393920.0, + "grad_norm": 3.7031774074531962, + "learning_rate": 1.2506744645535117e-06, + "loss": 0.8032, + "num_input_tokens_seen": 111985515, + "step": 5269 + }, + { + "epoch": 0.6336800336680094, + "flos": 16620743051040.0, + "grad_norm": 6.869043162427003, + "learning_rate": 1.249952289154267e-06, + "loss": 0.584, + "num_input_tokens_seen": 112005275, + "step": 5270 + }, + { + "epoch": 0.6338002765586485, + "flos": 17295526134360.0, + "grad_norm": 3.7147233464156213, + "learning_rate": 1.2492302275366635e-06, + "loss": 0.7532, + "num_input_tokens_seen": 112024670, + "step": 5271 + }, + { + "epoch": 0.6339205194492875, + "flos": 19428957641040.0, + "grad_norm": 3.843702858736581, + "learning_rate": 1.2485082798102377e-06, + "loss": 0.616, + "num_input_tokens_seen": 112044805, + "step": 5272 + }, + { + "epoch": 0.6340407623399267, + "flos": 13536150263880.0, + "grad_norm": 3.5519560999004645, + "learning_rate": 1.2477864460845084e-06, + "loss": 0.6638, + "num_input_tokens_seen": 112060925, + "step": 5273 + }, + { + "epoch": 0.6341610052305657, + "flos": 12921670421640.0, + "grad_norm": 13.151937688638139, + "learning_rate": 1.2470647264689776e-06, + "loss": 0.7084, + "num_input_tokens_seen": 112079125, + "step": 5274 + }, + { + "epoch": 0.6342812481212048, + "flos": 17269189539000.0, + "grad_norm": 3.8935831000499563, + "learning_rate": 1.2463431210731282e-06, + "loss": 0.6955, + "num_input_tokens_seen": 112098430, + "step": 5275 + }, + { + "epoch": 0.634401491011844, + "flos": 13002579805320.0, + "grad_norm": 4.17426108942651, + "learning_rate": 1.2456216300064289e-06, + "loss": 0.7434, + "num_input_tokens_seen": 112115700, + "step": 5276 + }, + { + "epoch": 0.634521733902483, + "flos": 15619219756920.0, + "grad_norm": 3.107696386808172, + "learning_rate": 1.244900253378328e-06, + "loss": 0.7563, + "num_input_tokens_seen": 112135475, + "step": 5277 + }, + { + "epoch": 0.6346419767931221, + "flos": 12057623877000.0, + "grad_norm": 3.677368210661955, + "learning_rate": 1.2441789912982583e-06, + "loss": 0.68, + "num_input_tokens_seen": 112152280, + "step": 5278 + }, + { + "epoch": 0.6347622196837612, + "flos": 17834257166400.0, + "grad_norm": 3.5547984346958295, + "learning_rate": 1.2434578438756346e-06, + "loss": 0.6262, + "num_input_tokens_seen": 112172430, + "step": 5279 + }, + { + "epoch": 0.6348824625744003, + "flos": 47564568230760.0, + "grad_norm": 2.8269212384420075, + "learning_rate": 1.242736811219855e-06, + "loss": 0.7648, + "num_input_tokens_seen": 112198110, + "step": 5280 + }, + { + "epoch": 0.6350027054650393, + "flos": 20995516902600.0, + "grad_norm": 3.9955559123412376, + "learning_rate": 1.2420158934402988e-06, + "loss": 0.8107, + "num_input_tokens_seen": 112218445, + "step": 5281 + }, + { + "epoch": 0.6351229483556785, + "flos": 16836533067480.0, + "grad_norm": 4.049140678888146, + "learning_rate": 1.2412950906463286e-06, + "loss": 0.819, + "num_input_tokens_seen": 112235470, + "step": 5282 + }, + { + "epoch": 0.6352431912463176, + "flos": 16048931852160.0, + "grad_norm": 2.986842388734156, + "learning_rate": 1.2405744029472902e-06, + "loss": 0.8771, + "num_input_tokens_seen": 112254675, + "step": 5283 + }, + { + "epoch": 0.6353634341369566, + "flos": 9978131959440.0, + "grad_norm": 4.136372235619812, + "learning_rate": 1.2398538304525108e-06, + "loss": 0.7302, + "num_input_tokens_seen": 112273020, + "step": 5284 + }, + { + "epoch": 0.6354836770275958, + "flos": 14108183082480.0, + "grad_norm": 4.376969684261402, + "learning_rate": 1.2391333732713016e-06, + "loss": 0.7357, + "num_input_tokens_seen": 112290545, + "step": 5285 + }, + { + "epoch": 0.6356039199182348, + "flos": 15808071638760.0, + "grad_norm": 5.2090665784313295, + "learning_rate": 1.2384130315129543e-06, + "loss": 0.7608, + "num_input_tokens_seen": 112308590, + "step": 5286 + }, + { + "epoch": 0.6357241628088739, + "flos": 13216976783520.0, + "grad_norm": 3.9885488153238993, + "learning_rate": 1.2376928052867447e-06, + "loss": 0.7114, + "num_input_tokens_seen": 112327430, + "step": 5287 + }, + { + "epoch": 0.6358444056995131, + "flos": 18266723678160.0, + "grad_norm": 5.8378893181246445, + "learning_rate": 1.2369726947019299e-06, + "loss": 0.771, + "num_input_tokens_seen": 112347625, + "step": 5288 + }, + { + "epoch": 0.6359646485901521, + "flos": 17051594904840.0, + "grad_norm": 4.463216276412034, + "learning_rate": 1.2362526998677511e-06, + "loss": 0.6566, + "num_input_tokens_seen": 112363710, + "step": 5289 + }, + { + "epoch": 0.6360848914807912, + "flos": 15271366844160.0, + "grad_norm": 3.228787816291786, + "learning_rate": 1.2355328208934301e-06, + "loss": 0.842, + "num_input_tokens_seen": 112382305, + "step": 5290 + }, + { + "epoch": 0.6362051343714303, + "flos": 13489587445440.0, + "grad_norm": 5.67055594919486, + "learning_rate": 1.2348130578881728e-06, + "loss": 0.7165, + "num_input_tokens_seen": 112400245, + "step": 5291 + }, + { + "epoch": 0.6363253772620694, + "flos": 17863126558560.0, + "grad_norm": 3.9974135142710905, + "learning_rate": 1.2340934109611664e-06, + "loss": 0.7275, + "num_input_tokens_seen": 112420725, + "step": 5292 + }, + { + "epoch": 0.6364456201527084, + "flos": 19023999143160.0, + "grad_norm": 5.311057758905717, + "learning_rate": 1.2333738802215798e-06, + "loss": 0.6544, + "num_input_tokens_seen": 112440665, + "step": 5293 + }, + { + "epoch": 0.6365658630433476, + "flos": 15162664506960.0, + "grad_norm": 4.003826695093795, + "learning_rate": 1.2326544657785668e-06, + "loss": 0.7781, + "num_input_tokens_seen": 112460075, + "step": 5294 + }, + { + "epoch": 0.6366861059339867, + "flos": 15675882102600.0, + "grad_norm": 4.8942342974065065, + "learning_rate": 1.2319351677412608e-06, + "loss": 0.7296, + "num_input_tokens_seen": 112476840, + "step": 5295 + }, + { + "epoch": 0.6368063488246257, + "flos": 16291438383480.0, + "grad_norm": 3.5561222744013126, + "learning_rate": 1.2312159862187796e-06, + "loss": 0.7295, + "num_input_tokens_seen": 112494970, + "step": 5296 + }, + { + "epoch": 0.6369265917152649, + "flos": 16406219433000.0, + "grad_norm": 5.657606099305814, + "learning_rate": 1.2304969213202217e-06, + "loss": 0.743, + "num_input_tokens_seen": 112515950, + "step": 5297 + }, + { + "epoch": 0.6370468346059039, + "flos": 18106329609000.0, + "grad_norm": 5.057564672715632, + "learning_rate": 1.2297779731546692e-06, + "loss": 0.7636, + "num_input_tokens_seen": 112534765, + "step": 5298 + }, + { + "epoch": 0.637167077496543, + "flos": 18533635547280.0, + "grad_norm": 2.6708810307076747, + "learning_rate": 1.2290591418311853e-06, + "loss": 0.7613, + "num_input_tokens_seen": 112551880, + "step": 5299 + }, + { + "epoch": 0.637287320387182, + "flos": 20291357867760.0, + "grad_norm": 4.2145213091301725, + "learning_rate": 1.2283404274588172e-06, + "loss": 0.6993, + "num_input_tokens_seen": 112570545, + "step": 5300 + }, + { + "epoch": 0.6374075632778212, + "flos": 46570837778160.0, + "grad_norm": 0.7588978496712563, + "learning_rate": 1.227621830146592e-06, + "loss": 0.5438, + "num_input_tokens_seen": 112625630, + "step": 5301 + }, + { + "epoch": 0.6375278061684603, + "flos": 18728059582080.0, + "grad_norm": 9.581863644452415, + "learning_rate": 1.2269033500035217e-06, + "loss": 0.7829, + "num_input_tokens_seen": 112645485, + "step": 5302 + }, + { + "epoch": 0.6376480490590993, + "flos": 18807987507000.0, + "grad_norm": 4.35743872734583, + "learning_rate": 1.2261849871385988e-06, + "loss": 0.7282, + "num_input_tokens_seen": 112666310, + "step": 5303 + }, + { + "epoch": 0.6377682919497385, + "flos": 23152752207840.0, + "grad_norm": 5.486027047630551, + "learning_rate": 1.2254667416607972e-06, + "loss": 0.5989, + "num_input_tokens_seen": 112687630, + "step": 5304 + }, + { + "epoch": 0.6378885348403776, + "flos": 16863692821800.0, + "grad_norm": 4.6529480151763565, + "learning_rate": 1.2247486136790756e-06, + "loss": 0.8202, + "num_input_tokens_seen": 112706830, + "step": 5305 + }, + { + "epoch": 0.6380087777310166, + "flos": 13650171474360.0, + "grad_norm": 5.0185420505107, + "learning_rate": 1.2240306033023726e-06, + "loss": 0.7862, + "num_input_tokens_seen": 112724375, + "step": 5306 + }, + { + "epoch": 0.6381290206216558, + "flos": 17080116037440.0, + "grad_norm": 3.3465909208664835, + "learning_rate": 1.223312710639611e-06, + "loss": 0.7122, + "num_input_tokens_seen": 112742815, + "step": 5307 + }, + { + "epoch": 0.6382492635122948, + "flos": 13786033565880.0, + "grad_norm": 4.30741981152746, + "learning_rate": 1.2225949357996928e-06, + "loss": 0.8399, + "num_input_tokens_seen": 112760660, + "step": 5308 + }, + { + "epoch": 0.6383695064029339, + "flos": 20401199963520.0, + "grad_norm": 2.6742443265145965, + "learning_rate": 1.221877278891505e-06, + "loss": 0.7915, + "num_input_tokens_seen": 112779635, + "step": 5309 + }, + { + "epoch": 0.638489749293573, + "flos": 19347636677880.0, + "grad_norm": 3.5200165171669195, + "learning_rate": 1.221159740023915e-06, + "loss": 0.6819, + "num_input_tokens_seen": 112799185, + "step": 5310 + }, + { + "epoch": 0.6386099921842121, + "flos": 17566173878760.0, + "grad_norm": 34.54244727957908, + "learning_rate": 1.2204423193057735e-06, + "loss": 0.7084, + "num_input_tokens_seen": 112817735, + "step": 5311 + }, + { + "epoch": 0.6387302350748512, + "flos": 52901837628480.0, + "grad_norm": 0.9626277851143207, + "learning_rate": 1.2197250168459122e-06, + "loss": 0.6816, + "num_input_tokens_seen": 112873855, + "step": 5312 + }, + { + "epoch": 0.6388504779654903, + "flos": 10570264361280.0, + "grad_norm": 4.021663889646197, + "learning_rate": 1.2190078327531454e-06, + "loss": 0.7268, + "num_input_tokens_seen": 112889820, + "step": 5313 + }, + { + "epoch": 0.6389707208561294, + "flos": 16350791825760.0, + "grad_norm": 2.5165888675344332, + "learning_rate": 1.2182907671362697e-06, + "loss": 0.7158, + "num_input_tokens_seen": 112910235, + "step": 5314 + }, + { + "epoch": 0.6390909637467684, + "flos": 14190295544640.0, + "grad_norm": 4.323025891253176, + "learning_rate": 1.2175738201040626e-06, + "loss": 0.7649, + "num_input_tokens_seen": 112926995, + "step": 5315 + }, + { + "epoch": 0.6392112066374076, + "flos": 17641067870040.0, + "grad_norm": 3.1872519602289677, + "learning_rate": 1.2168569917652855e-06, + "loss": 0.7742, + "num_input_tokens_seen": 112946570, + "step": 5316 + }, + { + "epoch": 0.6393314495280467, + "flos": 19643038019640.0, + "grad_norm": 3.7069790256628243, + "learning_rate": 1.2161402822286797e-06, + "loss": 0.6288, + "num_input_tokens_seen": 112966975, + "step": 5317 + }, + { + "epoch": 0.6394516924186857, + "flos": 14806611664560.0, + "grad_norm": 15.126724018802472, + "learning_rate": 1.2154236916029703e-06, + "loss": 0.7687, + "num_input_tokens_seen": 112984670, + "step": 5318 + }, + { + "epoch": 0.6395719353093249, + "flos": 13407285023520.0, + "grad_norm": 8.1424818502194, + "learning_rate": 1.2147072199968627e-06, + "loss": 0.71, + "num_input_tokens_seen": 113003025, + "step": 5319 + }, + { + "epoch": 0.6396921781999639, + "flos": 12813379663920.0, + "grad_norm": 6.4264280045911475, + "learning_rate": 1.2139908675190454e-06, + "loss": 0.7008, + "num_input_tokens_seen": 113021955, + "step": 5320 + }, + { + "epoch": 0.639812421090603, + "flos": 15540463250520.0, + "grad_norm": 5.621608813919578, + "learning_rate": 1.2132746342781883e-06, + "loss": 0.7436, + "num_input_tokens_seen": 113042835, + "step": 5321 + }, + { + "epoch": 0.6399326639812422, + "flos": 8087745203400.0, + "grad_norm": 6.173186730167619, + "learning_rate": 1.2125585203829442e-06, + "loss": 0.7713, + "num_input_tokens_seen": 113058195, + "step": 5322 + }, + { + "epoch": 0.6400529068718812, + "flos": 17508371774520.0, + "grad_norm": 9.449770194998461, + "learning_rate": 1.211842525941946e-06, + "loss": 0.7222, + "num_input_tokens_seen": 113077710, + "step": 5323 + }, + { + "epoch": 0.6401731497625203, + "flos": 32389972410240.0, + "grad_norm": 7.184625996390335, + "learning_rate": 1.2111266510638105e-06, + "loss": 0.789, + "num_input_tokens_seen": 113100355, + "step": 5324 + }, + { + "epoch": 0.6402933926531594, + "flos": 15103785964080.0, + "grad_norm": 3.117553536476224, + "learning_rate": 1.2104108958571346e-06, + "loss": 0.7962, + "num_input_tokens_seen": 113118345, + "step": 5325 + }, + { + "epoch": 0.6404136355437985, + "flos": 18213638907960.0, + "grad_norm": 2.2111153143561117, + "learning_rate": 1.2096952604304975e-06, + "loss": 0.7396, + "num_input_tokens_seen": 113138495, + "step": 5326 + }, + { + "epoch": 0.6405338784344375, + "flos": 29770736341920.0, + "grad_norm": 5.00750322026622, + "learning_rate": 1.2089797448924616e-06, + "loss": 0.6807, + "num_input_tokens_seen": 113162090, + "step": 5327 + }, + { + "epoch": 0.6406541213250767, + "flos": 15270132105720.0, + "grad_norm": 3.9084781243047715, + "learning_rate": 1.2082643493515692e-06, + "loss": 0.6402, + "num_input_tokens_seen": 113180130, + "step": 5328 + }, + { + "epoch": 0.6407743642157158, + "flos": 17053906081920.0, + "grad_norm": 2.9651830828334496, + "learning_rate": 1.207549073916346e-06, + "loss": 0.8071, + "num_input_tokens_seen": 113200785, + "step": 5329 + }, + { + "epoch": 0.6408946071063548, + "flos": 10923531127200.0, + "grad_norm": 4.317496238932161, + "learning_rate": 1.2068339186952976e-06, + "loss": 0.771, + "num_input_tokens_seen": 113218045, + "step": 5330 + }, + { + "epoch": 0.6410148499969939, + "flos": 16485830758320.0, + "grad_norm": 3.6765071140045644, + "learning_rate": 1.2061188837969136e-06, + "loss": 0.7125, + "num_input_tokens_seen": 113237375, + "step": 5331 + }, + { + "epoch": 0.641135092887633, + "flos": 9006269556480.0, + "grad_norm": 5.706828380741472, + "learning_rate": 1.2054039693296631e-06, + "loss": 0.8099, + "num_input_tokens_seen": 113255090, + "step": 5332 + }, + { + "epoch": 0.6412553357782721, + "flos": 16188878078520.0, + "grad_norm": 4.404679471672025, + "learning_rate": 1.2046891754019992e-06, + "loss": 0.8059, + "num_input_tokens_seen": 113275420, + "step": 5333 + }, + { + "epoch": 0.6413755786689112, + "flos": 11571850975320.0, + "grad_norm": 3.549373603682152, + "learning_rate": 1.2039745021223548e-06, + "loss": 0.8043, + "num_input_tokens_seen": 113292560, + "step": 5334 + }, + { + "epoch": 0.6414958215595503, + "flos": 50173682094600.0, + "grad_norm": 0.8717240247277966, + "learning_rate": 1.2032599495991456e-06, + "loss": 0.6133, + "num_input_tokens_seen": 113357020, + "step": 5335 + }, + { + "epoch": 0.6416160644501894, + "flos": 32443627059720.0, + "grad_norm": 2.641587663145041, + "learning_rate": 1.2025455179407685e-06, + "loss": 0.6804, + "num_input_tokens_seen": 113377900, + "step": 5336 + }, + { + "epoch": 0.6417363073408284, + "flos": 15324863193840.0, + "grad_norm": 6.465465061722588, + "learning_rate": 1.2018312072556022e-06, + "loss": 0.7185, + "num_input_tokens_seen": 113396120, + "step": 5337 + }, + { + "epoch": 0.6418565502314676, + "flos": 16431384609840.0, + "grad_norm": 3.9191644428129795, + "learning_rate": 1.2011170176520077e-06, + "loss": 0.7305, + "num_input_tokens_seen": 113416755, + "step": 5338 + }, + { + "epoch": 0.6419767931221066, + "flos": 18348424560840.0, + "grad_norm": 2.5843944728721957, + "learning_rate": 1.2004029492383256e-06, + "loss": 0.7933, + "num_input_tokens_seen": 113437815, + "step": 5339 + }, + { + "epoch": 0.6420970360127457, + "flos": 14217581938800.0, + "grad_norm": 4.1963155389189755, + "learning_rate": 1.1996890021228814e-06, + "loss": 0.7251, + "num_input_tokens_seen": 113454310, + "step": 5340 + }, + { + "epoch": 0.6422172789033849, + "flos": 29716353513360.0, + "grad_norm": 3.4005809119377317, + "learning_rate": 1.1989751764139785e-06, + "loss": 0.6837, + "num_input_tokens_seen": 113477680, + "step": 5341 + }, + { + "epoch": 0.6423375217940239, + "flos": 20292909205800.0, + "grad_norm": 3.6467526975127345, + "learning_rate": 1.1982614722199044e-06, + "loss": 0.802, + "num_input_tokens_seen": 113498575, + "step": 5342 + }, + { + "epoch": 0.642457764684663, + "flos": 13407633283080.0, + "grad_norm": 3.350815198410155, + "learning_rate": 1.1975478896489276e-06, + "loss": 0.7675, + "num_input_tokens_seen": 113516130, + "step": 5343 + }, + { + "epoch": 0.6425780075753021, + "flos": 14429129520600.0, + "grad_norm": 4.218478949987691, + "learning_rate": 1.1968344288092981e-06, + "loss": 0.7491, + "num_input_tokens_seen": 113532430, + "step": 5344 + }, + { + "epoch": 0.6426982504659412, + "flos": 15027878854080.0, + "grad_norm": 3.530349751725293, + "learning_rate": 1.1961210898092468e-06, + "loss": 0.633, + "num_input_tokens_seen": 113551100, + "step": 5345 + }, + { + "epoch": 0.6428184933565803, + "flos": 13023914127000.0, + "grad_norm": 5.378414400283939, + "learning_rate": 1.1954078727569874e-06, + "loss": 0.7716, + "num_input_tokens_seen": 113568120, + "step": 5346 + }, + { + "epoch": 0.6429387362472194, + "flos": 16432999267800.0, + "grad_norm": 3.69185510560593, + "learning_rate": 1.1946947777607141e-06, + "loss": 0.7671, + "num_input_tokens_seen": 113588975, + "step": 5347 + }, + { + "epoch": 0.6430589791378585, + "flos": 18131494785840.0, + "grad_norm": 4.114833672871299, + "learning_rate": 1.1939818049286024e-06, + "loss": 0.7851, + "num_input_tokens_seen": 113606855, + "step": 5348 + }, + { + "epoch": 0.6431792220284975, + "flos": 18241906760880.0, + "grad_norm": 3.3148632616291036, + "learning_rate": 1.1932689543688101e-06, + "loss": 0.7281, + "num_input_tokens_seen": 113627680, + "step": 5349 + }, + { + "epoch": 0.6432994649191367, + "flos": 15378422863440.0, + "grad_norm": 12.141118926969957, + "learning_rate": 1.1925562261894756e-06, + "loss": 0.7111, + "num_input_tokens_seen": 113646480, + "step": 5350 + }, + { + "epoch": 0.6434197078097758, + "flos": 22669797042600.0, + "grad_norm": 4.006377143449148, + "learning_rate": 1.1918436204987207e-06, + "loss": 0.7665, + "num_input_tokens_seen": 113668060, + "step": 5351 + }, + { + "epoch": 0.6435399507004148, + "flos": 10923404487360.0, + "grad_norm": 5.018566564340635, + "learning_rate": 1.191131137404645e-06, + "loss": 0.7924, + "num_input_tokens_seen": 113684520, + "step": 5352 + }, + { + "epoch": 0.643660193591054, + "flos": 14542010972520.0, + "grad_norm": 6.80866350410697, + "learning_rate": 1.190418777015333e-06, + "loss": 0.758, + "num_input_tokens_seen": 113703150, + "step": 5353 + }, + { + "epoch": 0.643780436481693, + "flos": 17673609817560.0, + "grad_norm": 2.898187261508648, + "learning_rate": 1.1897065394388487e-06, + "loss": 0.7232, + "num_input_tokens_seen": 113723310, + "step": 5354 + }, + { + "epoch": 0.6439006793723321, + "flos": 16944823825200.0, + "grad_norm": 3.9347851794690567, + "learning_rate": 1.1889944247832385e-06, + "loss": 0.7488, + "num_input_tokens_seen": 113743270, + "step": 5355 + }, + { + "epoch": 0.6440209222629713, + "flos": 17291252039760.0, + "grad_norm": 3.977566995014553, + "learning_rate": 1.1882824331565283e-06, + "loss": 0.6879, + "num_input_tokens_seen": 113762450, + "step": 5356 + }, + { + "epoch": 0.6441411651536103, + "flos": 12057465577200.0, + "grad_norm": 4.149965741119536, + "learning_rate": 1.1875705646667287e-06, + "loss": 0.8714, + "num_input_tokens_seen": 113780060, + "step": 5357 + }, + { + "epoch": 0.6442614080442494, + "flos": 18618977325360.0, + "grad_norm": 8.065336452970167, + "learning_rate": 1.1868588194218282e-06, + "loss": 0.7336, + "num_input_tokens_seen": 113800160, + "step": 5358 + }, + { + "epoch": 0.6443816509348885, + "flos": 20752725431640.0, + "grad_norm": 3.9510689315170295, + "learning_rate": 1.1861471975297979e-06, + "loss": 0.7198, + "num_input_tokens_seen": 113821575, + "step": 5359 + }, + { + "epoch": 0.6445018938255276, + "flos": 26966985806280.0, + "grad_norm": 2.6925452138067345, + "learning_rate": 1.185435699098591e-06, + "loss": 0.7013, + "num_input_tokens_seen": 113847490, + "step": 5360 + }, + { + "epoch": 0.6446221367161666, + "flos": 10545130844400.0, + "grad_norm": 11.306159671656689, + "learning_rate": 1.1847243242361403e-06, + "loss": 0.7511, + "num_input_tokens_seen": 113865800, + "step": 5361 + }, + { + "epoch": 0.6447423796068057, + "flos": 18025895124720.0, + "grad_norm": 8.075832491735468, + "learning_rate": 1.1840130730503624e-06, + "loss": 0.7648, + "num_input_tokens_seen": 113886800, + "step": 5362 + }, + { + "epoch": 0.6448626224974449, + "flos": 18350007558840.0, + "grad_norm": 3.989836894743352, + "learning_rate": 1.1833019456491518e-06, + "loss": 0.7329, + "num_input_tokens_seen": 113908050, + "step": 5363 + }, + { + "epoch": 0.6449828653880839, + "flos": 16487382096360.0, + "grad_norm": 3.815394444103928, + "learning_rate": 1.1825909421403871e-06, + "loss": 0.7467, + "num_input_tokens_seen": 113926865, + "step": 5364 + }, + { + "epoch": 0.645103108278723, + "flos": 18828720289440.0, + "grad_norm": 4.154744728959576, + "learning_rate": 1.181880062631926e-06, + "loss": 0.741, + "num_input_tokens_seen": 113945920, + "step": 5365 + }, + { + "epoch": 0.6452233511693621, + "flos": 20127702822720.0, + "grad_norm": 5.612284211426237, + "learning_rate": 1.1811693072316093e-06, + "loss": 0.8406, + "num_input_tokens_seen": 113963320, + "step": 5366 + }, + { + "epoch": 0.6453435940600012, + "flos": 14028856696800.0, + "grad_norm": 4.246438722186988, + "learning_rate": 1.1804586760472574e-06, + "loss": 0.8123, + "num_input_tokens_seen": 113979505, + "step": 5367 + }, + { + "epoch": 0.6454638369506402, + "flos": 18860534057880.0, + "grad_norm": 3.6545327873476285, + "learning_rate": 1.1797481691866736e-06, + "loss": 0.7729, + "num_input_tokens_seen": 113996450, + "step": 5368 + }, + { + "epoch": 0.6455840798412794, + "flos": 15345564316320.0, + "grad_norm": 4.148935176788894, + "learning_rate": 1.1790377867576393e-06, + "loss": 0.8151, + "num_input_tokens_seen": 114013920, + "step": 5369 + }, + { + "epoch": 0.6457043227319185, + "flos": 19104876866880.0, + "grad_norm": 3.7549761498883676, + "learning_rate": 1.1783275288679203e-06, + "loss": 0.7439, + "num_input_tokens_seen": 114030805, + "step": 5370 + }, + { + "epoch": 0.6458245656225575, + "flos": 44497709512560.0, + "grad_norm": 1.1473577185124268, + "learning_rate": 1.177617395625262e-06, + "loss": 0.6452, + "num_input_tokens_seen": 114088500, + "step": 5371 + }, + { + "epoch": 0.6459448085131967, + "flos": 16891644075120.0, + "grad_norm": 3.1632286951132524, + "learning_rate": 1.1769073871373908e-06, + "loss": 0.7433, + "num_input_tokens_seen": 114108425, + "step": 5372 + }, + { + "epoch": 0.6460650514038357, + "flos": 16536857631120.0, + "grad_norm": 3.261708955121895, + "learning_rate": 1.176197503512015e-06, + "loss": 0.8234, + "num_input_tokens_seen": 114127860, + "step": 5373 + }, + { + "epoch": 0.6461852942944748, + "flos": 14811930537840.0, + "grad_norm": 5.481689317207401, + "learning_rate": 1.1754877448568223e-06, + "loss": 0.7936, + "num_input_tokens_seen": 114147035, + "step": 5374 + }, + { + "epoch": 0.646305537185114, + "flos": 17106294333000.0, + "grad_norm": 4.858751655658806, + "learning_rate": 1.1747781112794837e-06, + "loss": 0.8852, + "num_input_tokens_seen": 114163250, + "step": 5375 + }, + { + "epoch": 0.646425780075753, + "flos": 17779779357960.0, + "grad_norm": 2.5611235480605337, + "learning_rate": 1.1740686028876487e-06, + "loss": 0.8157, + "num_input_tokens_seen": 114181835, + "step": 5376 + }, + { + "epoch": 0.6465460229663921, + "flos": 15216920695680.0, + "grad_norm": 3.4947175625914544, + "learning_rate": 1.1733592197889507e-06, + "loss": 0.7306, + "num_input_tokens_seen": 114198465, + "step": 5377 + }, + { + "epoch": 0.6466662658570312, + "flos": 16725614533080.0, + "grad_norm": 4.187494534161544, + "learning_rate": 1.1726499620910014e-06, + "loss": 0.7081, + "num_input_tokens_seen": 114218465, + "step": 5378 + }, + { + "epoch": 0.6467865087476703, + "flos": 11139574423320.0, + "grad_norm": 3.4741135104559717, + "learning_rate": 1.1719408299013955e-06, + "loss": 0.7405, + "num_input_tokens_seen": 114236910, + "step": 5379 + }, + { + "epoch": 0.6469067516383094, + "flos": 14299029541800.0, + "grad_norm": 6.202152973570416, + "learning_rate": 1.1712318233277067e-06, + "loss": 0.7306, + "num_input_tokens_seen": 114255650, + "step": 5380 + }, + { + "epoch": 0.6470269945289485, + "flos": 47993873237880.0, + "grad_norm": 0.7515613278322338, + "learning_rate": 1.1705229424774916e-06, + "loss": 0.5842, + "num_input_tokens_seen": 114309640, + "step": 5381 + }, + { + "epoch": 0.6471472374195876, + "flos": 22530325715640.0, + "grad_norm": 2.9829406486947163, + "learning_rate": 1.1698141874582867e-06, + "loss": 0.6263, + "num_input_tokens_seen": 114330405, + "step": 5382 + }, + { + "epoch": 0.6472674803102266, + "flos": 15000687439800.0, + "grad_norm": 2.839224263748963, + "learning_rate": 1.169105558377609e-06, + "loss": 0.7035, + "num_input_tokens_seen": 114350215, + "step": 5383 + }, + { + "epoch": 0.6473877232008658, + "flos": 18103923452040.0, + "grad_norm": 3.287586451797055, + "learning_rate": 1.1683970553429587e-06, + "loss": 0.7694, + "num_input_tokens_seen": 114371390, + "step": 5384 + }, + { + "epoch": 0.6475079660915048, + "flos": 11569159878720.0, + "grad_norm": 3.546737466573938, + "learning_rate": 1.1676886784618128e-06, + "loss": 0.8026, + "num_input_tokens_seen": 114387775, + "step": 5385 + }, + { + "epoch": 0.6476282089821439, + "flos": 12676346153880.0, + "grad_norm": 4.4585155319109315, + "learning_rate": 1.1669804278416332e-06, + "loss": 0.8056, + "num_input_tokens_seen": 114402220, + "step": 5386 + }, + { + "epoch": 0.6477484518727831, + "flos": 14758212568440.0, + "grad_norm": 2.949342596220034, + "learning_rate": 1.1662723035898602e-06, + "loss": 0.6965, + "num_input_tokens_seen": 114421700, + "step": 5387 + }, + { + "epoch": 0.6478686947634221, + "flos": 18618502425960.0, + "grad_norm": 2.909520431800816, + "learning_rate": 1.165564305813915e-06, + "loss": 0.8053, + "num_input_tokens_seen": 114440420, + "step": 5388 + }, + { + "epoch": 0.6479889376540612, + "flos": 14676290066040.0, + "grad_norm": 2.968411878103166, + "learning_rate": 1.1648564346212019e-06, + "loss": 0.7946, + "num_input_tokens_seen": 114459260, + "step": 5389 + }, + { + "epoch": 0.6481091805447003, + "flos": 19615530005760.0, + "grad_norm": 7.319225120596223, + "learning_rate": 1.164148690119104e-06, + "loss": 0.7506, + "num_input_tokens_seen": 114480260, + "step": 5390 + }, + { + "epoch": 0.6482294234353394, + "flos": 17539394043960.0, + "grad_norm": 3.57447568876681, + "learning_rate": 1.163441072414985e-06, + "loss": 0.7259, + "num_input_tokens_seen": 114500185, + "step": 5391 + }, + { + "epoch": 0.6483496663259785, + "flos": 19209653369040.0, + "grad_norm": 4.34688032686084, + "learning_rate": 1.16273358161619e-06, + "loss": 0.6812, + "num_input_tokens_seen": 114520240, + "step": 5392 + }, + { + "epoch": 0.6484699092166175, + "flos": 15296880280560.0, + "grad_norm": 4.407065674515989, + "learning_rate": 1.1620262178300446e-06, + "loss": 0.8302, + "num_input_tokens_seen": 114538575, + "step": 5393 + }, + { + "epoch": 0.6485901521072567, + "flos": 24289282774560.0, + "grad_norm": 6.916936417187458, + "learning_rate": 1.1613189811638563e-06, + "loss": 0.735, + "num_input_tokens_seen": 114560020, + "step": 5394 + }, + { + "epoch": 0.6487103949978957, + "flos": 16297580415720.0, + "grad_norm": 3.3651748625396496, + "learning_rate": 1.1606118717249117e-06, + "loss": 0.7569, + "num_input_tokens_seen": 114579840, + "step": 5395 + }, + { + "epoch": 0.6488306378885348, + "flos": 16783258337520.0, + "grad_norm": 2.606509617011884, + "learning_rate": 1.1599048896204787e-06, + "loss": 0.6504, + "num_input_tokens_seen": 114599440, + "step": 5396 + }, + { + "epoch": 0.648950880779174, + "flos": 14892871581480.0, + "grad_norm": 3.6978358873311006, + "learning_rate": 1.1591980349578061e-06, + "loss": 0.7913, + "num_input_tokens_seen": 114617830, + "step": 5397 + }, + { + "epoch": 0.649071123669813, + "flos": 44171824120680.0, + "grad_norm": 0.7894419522558944, + "learning_rate": 1.158491307844123e-06, + "loss": 0.5673, + "num_input_tokens_seen": 114677470, + "step": 5398 + }, + { + "epoch": 0.6491913665604521, + "flos": 14944658293320.0, + "grad_norm": 4.04695512257798, + "learning_rate": 1.1577847083866387e-06, + "loss": 0.8343, + "num_input_tokens_seen": 114696225, + "step": 5399 + }, + { + "epoch": 0.6493116094510912, + "flos": 12355019796240.0, + "grad_norm": 5.1333659670564495, + "learning_rate": 1.1570782366925453e-06, + "loss": 0.7047, + "num_input_tokens_seen": 114714460, + "step": 5400 + }, + { + "epoch": 0.6494318523417303, + "flos": 13728009841920.0, + "grad_norm": 4.667893770645206, + "learning_rate": 1.1563718928690132e-06, + "loss": 0.7476, + "num_input_tokens_seen": 114731615, + "step": 5401 + }, + { + "epoch": 0.6495520952323693, + "flos": 13860800917320.0, + "grad_norm": 3.471620076844063, + "learning_rate": 1.1556656770231942e-06, + "loss": 0.7057, + "num_input_tokens_seen": 114747530, + "step": 5402 + }, + { + "epoch": 0.6496723381230085, + "flos": 16645971547800.0, + "grad_norm": 2.709282889554866, + "learning_rate": 1.1549595892622207e-06, + "loss": 0.7475, + "num_input_tokens_seen": 114766020, + "step": 5403 + }, + { + "epoch": 0.6497925810136476, + "flos": 45809034938880.0, + "grad_norm": 0.8610231028125583, + "learning_rate": 1.1542536296932047e-06, + "loss": 0.6257, + "num_input_tokens_seen": 114829275, + "step": 5404 + }, + { + "epoch": 0.6499128239042866, + "flos": 14730451274880.0, + "grad_norm": 2.4424608073999496, + "learning_rate": 1.1535477984232414e-06, + "loss": 0.6812, + "num_input_tokens_seen": 114848870, + "step": 5405 + }, + { + "epoch": 0.6500330667949258, + "flos": 17917002827760.0, + "grad_norm": 3.224935569415284, + "learning_rate": 1.152842095559404e-06, + "loss": 0.7511, + "num_input_tokens_seen": 114869250, + "step": 5406 + }, + { + "epoch": 0.6501533096855648, + "flos": 18667218121680.0, + "grad_norm": 2.964521428255155, + "learning_rate": 1.1521365212087474e-06, + "loss": 0.7606, + "num_input_tokens_seen": 114888955, + "step": 5407 + }, + { + "epoch": 0.6502735525762039, + "flos": 32982104812080.0, + "grad_norm": 3.5919271075237815, + "learning_rate": 1.1514310754783062e-06, + "loss": 0.683, + "num_input_tokens_seen": 114911625, + "step": 5408 + }, + { + "epoch": 0.6503937954668431, + "flos": 21022455037200.0, + "grad_norm": 3.494214726147951, + "learning_rate": 1.1507257584750964e-06, + "loss": 0.7041, + "num_input_tokens_seen": 114931525, + "step": 5409 + }, + { + "epoch": 0.6505140383574821, + "flos": 15297165220200.0, + "grad_norm": 6.531454420769231, + "learning_rate": 1.150020570306113e-06, + "loss": 0.745, + "num_input_tokens_seen": 114950385, + "step": 5410 + }, + { + "epoch": 0.6506342812481212, + "flos": 15055070268360.0, + "grad_norm": 3.9716805279151433, + "learning_rate": 1.1493155110783338e-06, + "loss": 0.7262, + "num_input_tokens_seen": 114968630, + "step": 5411 + }, + { + "epoch": 0.6507545241387603, + "flos": 22182694422600.0, + "grad_norm": 3.9799175386177006, + "learning_rate": 1.1486105808987155e-06, + "loss": 0.6915, + "num_input_tokens_seen": 114989840, + "step": 5412 + }, + { + "epoch": 0.6508747670293994, + "flos": 12489140589960.0, + "grad_norm": 2.9936953099617067, + "learning_rate": 1.1479057798741947e-06, + "loss": 0.796, + "num_input_tokens_seen": 115007615, + "step": 5413 + }, + { + "epoch": 0.6509950099200384, + "flos": 50555243882040.0, + "grad_norm": 0.819626803141477, + "learning_rate": 1.14720110811169e-06, + "loss": 0.5719, + "num_input_tokens_seen": 115064565, + "step": 5414 + }, + { + "epoch": 0.6511152528106776, + "flos": 16351963244280.0, + "grad_norm": 2.9968114438442828, + "learning_rate": 1.146496565718098e-06, + "loss": 0.7462, + "num_input_tokens_seen": 115084855, + "step": 5415 + }, + { + "epoch": 0.6512354957013167, + "flos": 15000972379440.0, + "grad_norm": 4.950271891823571, + "learning_rate": 1.1457921528002996e-06, + "loss": 0.7412, + "num_input_tokens_seen": 115103010, + "step": 5416 + }, + { + "epoch": 0.6513557385919557, + "flos": 23744124770640.0, + "grad_norm": 3.640161574833397, + "learning_rate": 1.1450878694651522e-06, + "loss": 0.7086, + "num_input_tokens_seen": 115123295, + "step": 5417 + }, + { + "epoch": 0.6514759814825949, + "flos": 8762876546280.0, + "grad_norm": 4.151219126865963, + "learning_rate": 1.1443837158194954e-06, + "loss": 0.6068, + "num_input_tokens_seen": 115138160, + "step": 5418 + }, + { + "epoch": 0.651596224373234, + "flos": 16484342740200.0, + "grad_norm": 2.761413989052186, + "learning_rate": 1.1436796919701484e-06, + "loss": 0.7291, + "num_input_tokens_seen": 115156595, + "step": 5419 + }, + { + "epoch": 0.651716467263873, + "flos": 20401041663720.0, + "grad_norm": 22.89571855659167, + "learning_rate": 1.1429757980239115e-06, + "loss": 0.6007, + "num_input_tokens_seen": 115176740, + "step": 5420 + }, + { + "epoch": 0.6518367101545122, + "flos": 17808838709880.0, + "grad_norm": 5.257373536421475, + "learning_rate": 1.1422720340875636e-06, + "loss": 0.784, + "num_input_tokens_seen": 115195210, + "step": 5421 + }, + { + "epoch": 0.6519569530451512, + "flos": 14623458575520.0, + "grad_norm": 4.431800513243412, + "learning_rate": 1.1415684002678671e-06, + "loss": 0.7798, + "num_input_tokens_seen": 115213690, + "step": 5422 + }, + { + "epoch": 0.6520771959357903, + "flos": 15780342005160.0, + "grad_norm": 13.424578748519274, + "learning_rate": 1.1408648966715617e-06, + "loss": 0.7621, + "num_input_tokens_seen": 115230930, + "step": 5423 + }, + { + "epoch": 0.6521974388264293, + "flos": 16620838030920.0, + "grad_norm": 3.888493210478451, + "learning_rate": 1.1401615234053683e-06, + "loss": 0.7106, + "num_input_tokens_seen": 115249470, + "step": 5424 + }, + { + "epoch": 0.6523176817170685, + "flos": 16836881327040.0, + "grad_norm": 3.2678069349804986, + "learning_rate": 1.1394582805759885e-06, + "loss": 0.7395, + "num_input_tokens_seen": 115268470, + "step": 5425 + }, + { + "epoch": 0.6524379246077076, + "flos": 15863752525680.0, + "grad_norm": 5.0065251687919305, + "learning_rate": 1.1387551682901022e-06, + "loss": 0.747, + "num_input_tokens_seen": 115288795, + "step": 5426 + }, + { + "epoch": 0.6525581674983466, + "flos": 14163768989520.0, + "grad_norm": 3.9881062109605785, + "learning_rate": 1.138052186654373e-06, + "loss": 0.6951, + "num_input_tokens_seen": 115305985, + "step": 5427 + }, + { + "epoch": 0.6526784103889858, + "flos": 12516775243680.0, + "grad_norm": 5.215977463375153, + "learning_rate": 1.1373493357754417e-06, + "loss": 0.8575, + "num_input_tokens_seen": 115324610, + "step": 5428 + }, + { + "epoch": 0.6527986532796248, + "flos": 13866373070280.0, + "grad_norm": 4.348748494033136, + "learning_rate": 1.1366466157599303e-06, + "loss": 0.757, + "num_input_tokens_seen": 115343605, + "step": 5429 + }, + { + "epoch": 0.6529188961702639, + "flos": 10349567051040.0, + "grad_norm": 3.8916883126309663, + "learning_rate": 1.1359440267144412e-06, + "loss": 0.7353, + "num_input_tokens_seen": 115360780, + "step": 5430 + }, + { + "epoch": 0.653039139060903, + "flos": 26960685474240.0, + "grad_norm": 3.5295868054161486, + "learning_rate": 1.1352415687455556e-06, + "loss": 0.7311, + "num_input_tokens_seen": 115381760, + "step": 5431 + }, + { + "epoch": 0.6531593819515421, + "flos": 18593558868840.0, + "grad_norm": 5.97607579462277, + "learning_rate": 1.1345392419598362e-06, + "loss": 0.6216, + "num_input_tokens_seen": 115400360, + "step": 5432 + }, + { + "epoch": 0.6532796248421812, + "flos": 15486681961200.0, + "grad_norm": 3.0668119320119005, + "learning_rate": 1.1338370464638263e-06, + "loss": 0.698, + "num_input_tokens_seen": 115419480, + "step": 5433 + }, + { + "epoch": 0.6533998677328203, + "flos": 12894542327280.0, + "grad_norm": 5.519840175139316, + "learning_rate": 1.1331349823640474e-06, + "loss": 0.6185, + "num_input_tokens_seen": 115436630, + "step": 5434 + }, + { + "epoch": 0.6535201106234594, + "flos": 20831007038640.0, + "grad_norm": 4.159411382480225, + "learning_rate": 1.132433049767003e-06, + "loss": 0.7618, + "num_input_tokens_seen": 115454265, + "step": 5435 + }, + { + "epoch": 0.6536403535140984, + "flos": 17268872939400.0, + "grad_norm": 3.206044276369023, + "learning_rate": 1.1317312487791748e-06, + "loss": 0.7927, + "num_input_tokens_seen": 115475635, + "step": 5436 + }, + { + "epoch": 0.6537605964047376, + "flos": 15783349701360.0, + "grad_norm": 4.851654297579103, + "learning_rate": 1.1310295795070253e-06, + "loss": 0.698, + "num_input_tokens_seen": 115495295, + "step": 5437 + }, + { + "epoch": 0.6538808392953767, + "flos": 19672129031520.0, + "grad_norm": 4.827863075040332, + "learning_rate": 1.1303280420569982e-06, + "loss": 0.7825, + "num_input_tokens_seen": 115516900, + "step": 5438 + }, + { + "epoch": 0.6540010821860157, + "flos": 22561569604800.0, + "grad_norm": 3.35126200575992, + "learning_rate": 1.1296266365355158e-06, + "loss": 0.7578, + "num_input_tokens_seen": 115540005, + "step": 5439 + }, + { + "epoch": 0.6541213250766549, + "flos": 19726290240360.0, + "grad_norm": 3.4564794757398656, + "learning_rate": 1.1289253630489806e-06, + "loss": 0.701, + "num_input_tokens_seen": 115560775, + "step": 5440 + }, + { + "epoch": 0.6542415679672939, + "flos": 14002076862000.0, + "grad_norm": 4.466755028225494, + "learning_rate": 1.1282242217037753e-06, + "loss": 0.7263, + "num_input_tokens_seen": 115577995, + "step": 5441 + }, + { + "epoch": 0.654361810857933, + "flos": 35464149070560.0, + "grad_norm": 4.341557669903583, + "learning_rate": 1.127523212606262e-06, + "loss": 0.6118, + "num_input_tokens_seen": 115600540, + "step": 5442 + }, + { + "epoch": 0.6544820537485722, + "flos": 19752880115400.0, + "grad_norm": 2.5038859683448775, + "learning_rate": 1.1268223358627835e-06, + "loss": 0.7107, + "num_input_tokens_seen": 115622750, + "step": 5443 + }, + { + "epoch": 0.6546022966392112, + "flos": 15270385385400.0, + "grad_norm": 9.605157785862167, + "learning_rate": 1.126121591579663e-06, + "loss": 0.6966, + "num_input_tokens_seen": 115641675, + "step": 5444 + }, + { + "epoch": 0.6547225395298503, + "flos": 18267990076560.0, + "grad_norm": 3.8788907606847083, + "learning_rate": 1.1254209798632018e-06, + "loss": 0.6768, + "num_input_tokens_seen": 115662415, + "step": 5445 + }, + { + "epoch": 0.6548427824204894, + "flos": 16513592051880.0, + "grad_norm": 3.7163680967369657, + "learning_rate": 1.124720500819683e-06, + "loss": 0.8334, + "num_input_tokens_seen": 115680290, + "step": 5446 + }, + { + "epoch": 0.6549630253111285, + "flos": 13461889471800.0, + "grad_norm": 4.773337405290056, + "learning_rate": 1.1240201545553682e-06, + "loss": 0.8053, + "num_input_tokens_seen": 115697810, + "step": 5447 + }, + { + "epoch": 0.6550832682017675, + "flos": 18452789483520.0, + "grad_norm": 4.393675915024303, + "learning_rate": 1.1233199411764987e-06, + "loss": 0.7252, + "num_input_tokens_seen": 115716965, + "step": 5448 + }, + { + "epoch": 0.6552035110924067, + "flos": 16648409364720.0, + "grad_norm": 6.572751421496123, + "learning_rate": 1.1226198607892978e-06, + "loss": 0.6759, + "num_input_tokens_seen": 115737245, + "step": 5449 + }, + { + "epoch": 0.6553237539830458, + "flos": 15946054947600.0, + "grad_norm": 6.9389984986880435, + "learning_rate": 1.1219199134999664e-06, + "loss": 0.7923, + "num_input_tokens_seen": 115755465, + "step": 5450 + }, + { + "epoch": 0.6554439968736848, + "flos": 15271176884400.0, + "grad_norm": 3.8643320966767027, + "learning_rate": 1.1212200994146863e-06, + "loss": 0.7599, + "num_input_tokens_seen": 115772940, + "step": 5451 + }, + { + "epoch": 0.655564239764324, + "flos": 11756492082480.0, + "grad_norm": 3.3342722682982964, + "learning_rate": 1.120520418639618e-06, + "loss": 0.7419, + "num_input_tokens_seen": 115791195, + "step": 5452 + }, + { + "epoch": 0.655684482654963, + "flos": 21697206460560.0, + "grad_norm": 4.260254549829822, + "learning_rate": 1.119820871280903e-06, + "loss": 0.8189, + "num_input_tokens_seen": 115811990, + "step": 5453 + }, + { + "epoch": 0.6558047255456021, + "flos": 21859341827520.0, + "grad_norm": 3.7857424948327663, + "learning_rate": 1.1191214574446614e-06, + "loss": 0.7022, + "num_input_tokens_seen": 115831955, + "step": 5454 + }, + { + "epoch": 0.6559249684362413, + "flos": 21318837837720.0, + "grad_norm": 2.7668214785300433, + "learning_rate": 1.118422177236995e-06, + "loss": 0.7886, + "num_input_tokens_seen": 115853500, + "step": 5455 + }, + { + "epoch": 0.6560452113268803, + "flos": 14780718308640.0, + "grad_norm": 4.924940529091482, + "learning_rate": 1.1177230307639835e-06, + "loss": 0.8342, + "num_input_tokens_seen": 115870760, + "step": 5456 + }, + { + "epoch": 0.6561654542175194, + "flos": 18348456220800.0, + "grad_norm": 4.383834420536954, + "learning_rate": 1.1170240181316865e-06, + "loss": 0.7736, + "num_input_tokens_seen": 115891925, + "step": 5457 + }, + { + "epoch": 0.6562856971081584, + "flos": 16728622229280.0, + "grad_norm": 3.2086956430608002, + "learning_rate": 1.1163251394461442e-06, + "loss": 0.7845, + "num_input_tokens_seen": 115910125, + "step": 5458 + }, + { + "epoch": 0.6564059399987976, + "flos": 13779068374680.0, + "grad_norm": 3.347178475149904, + "learning_rate": 1.1156263948133746e-06, + "loss": 0.8139, + "num_input_tokens_seen": 115926500, + "step": 5459 + }, + { + "epoch": 0.6565261828894366, + "flos": 18673201854120.0, + "grad_norm": 6.098093216345921, + "learning_rate": 1.1149277843393787e-06, + "loss": 0.7619, + "num_input_tokens_seen": 115947380, + "step": 5460 + }, + { + "epoch": 0.6566464257800757, + "flos": 14380382164920.0, + "grad_norm": 3.8687539983429837, + "learning_rate": 1.1142293081301342e-06, + "loss": 0.6184, + "num_input_tokens_seen": 115964980, + "step": 5461 + }, + { + "epoch": 0.6567666686707149, + "flos": 17215249949880.0, + "grad_norm": 2.5091289736947604, + "learning_rate": 1.1135309662915995e-06, + "loss": 0.6526, + "num_input_tokens_seen": 115984865, + "step": 5462 + }, + { + "epoch": 0.6568869115613539, + "flos": 24073587738000.0, + "grad_norm": 4.1991939952031965, + "learning_rate": 1.112832758929712e-06, + "loss": 0.5836, + "num_input_tokens_seen": 116007195, + "step": 5463 + }, + { + "epoch": 0.657007154451993, + "flos": 13461826151880.0, + "grad_norm": 4.135986493854126, + "learning_rate": 1.11213468615039e-06, + "loss": 0.7301, + "num_input_tokens_seen": 116026345, + "step": 5464 + }, + { + "epoch": 0.6571273973426321, + "flos": 18430980262440.0, + "grad_norm": 4.1435554930546585, + "learning_rate": 1.1114367480595292e-06, + "loss": 0.7495, + "num_input_tokens_seen": 116047145, + "step": 5465 + }, + { + "epoch": 0.6572476402332712, + "flos": 12786378209400.0, + "grad_norm": 5.813680549424927, + "learning_rate": 1.1107389447630086e-06, + "loss": 0.8056, + "num_input_tokens_seen": 116065565, + "step": 5466 + }, + { + "epoch": 0.6573678831239103, + "flos": 12404147071440.0, + "grad_norm": 2.8153691386189896, + "learning_rate": 1.1100412763666818e-06, + "loss": 0.7663, + "num_input_tokens_seen": 116080545, + "step": 5467 + }, + { + "epoch": 0.6574881260145494, + "flos": 17507516955600.0, + "grad_norm": 5.491206856186239, + "learning_rate": 1.1093437429763865e-06, + "loss": 0.7825, + "num_input_tokens_seen": 116100530, + "step": 5468 + }, + { + "epoch": 0.6576083689051885, + "flos": 8115633136800.0, + "grad_norm": 3.964987104859134, + "learning_rate": 1.1086463446979361e-06, + "loss": 0.7189, + "num_input_tokens_seen": 116118600, + "step": 5469 + }, + { + "epoch": 0.6577286117958275, + "flos": 16431764529360.0, + "grad_norm": 3.8253884012330155, + "learning_rate": 1.1079490816371277e-06, + "loss": 0.7596, + "num_input_tokens_seen": 116138085, + "step": 5470 + }, + { + "epoch": 0.6578488546864667, + "flos": 16000089516600.0, + "grad_norm": 4.407174087911902, + "learning_rate": 1.1072519538997352e-06, + "loss": 0.7154, + "num_input_tokens_seen": 116156945, + "step": 5471 + }, + { + "epoch": 0.6579690975771058, + "flos": 17236204352040.0, + "grad_norm": 4.169567598753051, + "learning_rate": 1.1065549615915095e-06, + "loss": 0.8053, + "num_input_tokens_seen": 116176495, + "step": 5472 + }, + { + "epoch": 0.6580893404677448, + "flos": 24045826444440.0, + "grad_norm": 6.933718088673863, + "learning_rate": 1.105858104818187e-06, + "loss": 0.7429, + "num_input_tokens_seen": 116197370, + "step": 5473 + }, + { + "epoch": 0.658209583358384, + "flos": 11571439395840.0, + "grad_norm": 5.450012018032997, + "learning_rate": 1.105161383685478e-06, + "loss": 0.7296, + "num_input_tokens_seen": 116213475, + "step": 5474 + }, + { + "epoch": 0.658329826249023, + "flos": 46217001132960.0, + "grad_norm": 0.7553019279273554, + "learning_rate": 1.1044647982990771e-06, + "loss": 0.5912, + "num_input_tokens_seen": 116275080, + "step": 5475 + }, + { + "epoch": 0.6584500691396621, + "flos": 23289975677640.0, + "grad_norm": 3.820971597868949, + "learning_rate": 1.1037683487646536e-06, + "loss": 0.6206, + "num_input_tokens_seen": 116295085, + "step": 5476 + }, + { + "epoch": 0.6585703120303013, + "flos": 13434951337200.0, + "grad_norm": 2.751932158975759, + "learning_rate": 1.1030720351878583e-06, + "loss": 0.7506, + "num_input_tokens_seen": 116312925, + "step": 5477 + }, + { + "epoch": 0.6586905549209403, + "flos": 42969956379240.0, + "grad_norm": 0.8219426986603029, + "learning_rate": 1.102375857674323e-06, + "loss": 0.6068, + "num_input_tokens_seen": 116374560, + "step": 5478 + }, + { + "epoch": 0.6588107978115794, + "flos": 16673637861480.0, + "grad_norm": 3.1530890924925217, + "learning_rate": 1.1016798163296561e-06, + "loss": 0.8895, + "num_input_tokens_seen": 116393480, + "step": 5479 + }, + { + "epoch": 0.6589310407022185, + "flos": 15108471638160.0, + "grad_norm": 3.182279762624566, + "learning_rate": 1.1009839112594471e-06, + "loss": 0.6388, + "num_input_tokens_seen": 116411225, + "step": 5480 + }, + { + "epoch": 0.6590512835928576, + "flos": 18781619251680.0, + "grad_norm": 5.31892007091902, + "learning_rate": 1.1002881425692638e-06, + "loss": 0.7049, + "num_input_tokens_seen": 116431375, + "step": 5481 + }, + { + "epoch": 0.6591715264834966, + "flos": 17371148304720.0, + "grad_norm": 3.859715496904016, + "learning_rate": 1.0995925103646532e-06, + "loss": 0.7504, + "num_input_tokens_seen": 116449695, + "step": 5482 + }, + { + "epoch": 0.6592917693741358, + "flos": 26288340207840.0, + "grad_norm": 4.717020488358902, + "learning_rate": 1.0988970147511437e-06, + "loss": 0.6602, + "num_input_tokens_seen": 116471295, + "step": 5483 + }, + { + "epoch": 0.6594120122647749, + "flos": 15648152469000.0, + "grad_norm": 4.880448255744748, + "learning_rate": 1.0982016558342405e-06, + "loss": 0.7858, + "num_input_tokens_seen": 116489985, + "step": 5484 + }, + { + "epoch": 0.6595322551554139, + "flos": 14134456357920.0, + "grad_norm": 4.214045182907851, + "learning_rate": 1.0975064337194291e-06, + "loss": 0.7002, + "num_input_tokens_seen": 116507750, + "step": 5485 + }, + { + "epoch": 0.6596524980460531, + "flos": 12273793812960.0, + "grad_norm": 6.142895771298033, + "learning_rate": 1.0968113485121743e-06, + "loss": 0.6925, + "num_input_tokens_seen": 116527060, + "step": 5486 + }, + { + "epoch": 0.6597727409366921, + "flos": 15945516728280.0, + "grad_norm": 7.333097482992324, + "learning_rate": 1.0961164003179185e-06, + "loss": 0.7776, + "num_input_tokens_seen": 116545290, + "step": 5487 + }, + { + "epoch": 0.6598929838273312, + "flos": 17375200779600.0, + "grad_norm": 3.198594923296384, + "learning_rate": 1.0954215892420884e-06, + "loss": 0.816, + "num_input_tokens_seen": 116565710, + "step": 5488 + }, + { + "epoch": 0.6600132267179702, + "flos": 14595633962040.0, + "grad_norm": 5.191512980056462, + "learning_rate": 1.094726915390082e-06, + "loss": 0.6919, + "num_input_tokens_seen": 116583765, + "step": 5489 + }, + { + "epoch": 0.6601334696086094, + "flos": 16347720809640.0, + "grad_norm": 3.651025283203515, + "learning_rate": 1.0940323788672836e-06, + "loss": 0.6812, + "num_input_tokens_seen": 116602660, + "step": 5490 + }, + { + "epoch": 0.6602537124992485, + "flos": 18835210581240.0, + "grad_norm": 3.631566644148714, + "learning_rate": 1.0933379797790522e-06, + "loss": 0.7283, + "num_input_tokens_seen": 116621795, + "step": 5491 + }, + { + "epoch": 0.6603739553898875, + "flos": 18942804819840.0, + "grad_norm": 5.906422768947133, + "learning_rate": 1.0926437182307293e-06, + "loss": 0.6939, + "num_input_tokens_seen": 116640325, + "step": 5492 + }, + { + "epoch": 0.6604941982805267, + "flos": 17996614153080.0, + "grad_norm": 2.456288235975851, + "learning_rate": 1.0919495943276338e-06, + "loss": 0.7593, + "num_input_tokens_seen": 116661065, + "step": 5493 + }, + { + "epoch": 0.6606144411711657, + "flos": 9627841229760.0, + "grad_norm": 4.974905555633754, + "learning_rate": 1.0912556081750611e-06, + "loss": 0.7268, + "num_input_tokens_seen": 116678715, + "step": 5494 + }, + { + "epoch": 0.6607346840618048, + "flos": 18429713864040.0, + "grad_norm": 3.493960483058979, + "learning_rate": 1.0905617598782909e-06, + "loss": 0.7432, + "num_input_tokens_seen": 116698640, + "step": 5495 + }, + { + "epoch": 0.660854926952444, + "flos": 12866274474360.0, + "grad_norm": 3.6372136845593768, + "learning_rate": 1.0898680495425775e-06, + "loss": 0.7928, + "num_input_tokens_seen": 116716650, + "step": 5496 + }, + { + "epoch": 0.660975169843083, + "flos": 12272812354200.0, + "grad_norm": 3.438551308519092, + "learning_rate": 1.0891744772731594e-06, + "loss": 0.7901, + "num_input_tokens_seen": 116734185, + "step": 5497 + }, + { + "epoch": 0.6610954127337221, + "flos": 19699067166120.0, + "grad_norm": 4.837343569254394, + "learning_rate": 1.088481043175248e-06, + "loss": 0.6389, + "num_input_tokens_seen": 116754475, + "step": 5498 + }, + { + "epoch": 0.6612156556243612, + "flos": 19399075130160.0, + "grad_norm": 4.236723864209527, + "learning_rate": 1.0877877473540368e-06, + "loss": 0.7428, + "num_input_tokens_seen": 116774780, + "step": 5499 + }, + { + "epoch": 0.6613358985150003, + "flos": 14460468389640.0, + "grad_norm": 3.465386762075281, + "learning_rate": 1.0870945899147002e-06, + "loss": 0.7097, + "num_input_tokens_seen": 116791145, + "step": 5500 + }, + { + "epoch": 0.6614561414056394, + "flos": 19670039474160.0, + "grad_norm": 4.040421336647543, + "learning_rate": 1.0864015709623879e-06, + "loss": 0.7386, + "num_input_tokens_seen": 116811735, + "step": 5501 + }, + { + "epoch": 0.6615763842962785, + "flos": 16756858422240.0, + "grad_norm": 3.9777691102836226, + "learning_rate": 1.0857086906022313e-06, + "loss": 0.7839, + "num_input_tokens_seen": 116829790, + "step": 5502 + }, + { + "epoch": 0.6616966271869176, + "flos": 18159604338960.0, + "grad_norm": 6.055725087505973, + "learning_rate": 1.0850159489393388e-06, + "loss": 0.7148, + "num_input_tokens_seen": 116848770, + "step": 5503 + }, + { + "epoch": 0.6618168700775566, + "flos": 12543934998000.0, + "grad_norm": 6.889700137673667, + "learning_rate": 1.0843233460787992e-06, + "loss": 0.8017, + "num_input_tokens_seen": 116865705, + "step": 5504 + }, + { + "epoch": 0.6619371129681958, + "flos": 18645662180280.0, + "grad_norm": 3.1763607683499826, + "learning_rate": 1.0836308821256805e-06, + "loss": 0.7659, + "num_input_tokens_seen": 116886225, + "step": 5505 + }, + { + "epoch": 0.6620573558588349, + "flos": 13164240272880.0, + "grad_norm": 9.837536791644194, + "learning_rate": 1.0829385571850282e-06, + "loss": 0.7727, + "num_input_tokens_seen": 116902925, + "step": 5506 + }, + { + "epoch": 0.6621775987494739, + "flos": 12975704990640.0, + "grad_norm": 10.194020437165106, + "learning_rate": 1.0822463713618679e-06, + "loss": 0.8085, + "num_input_tokens_seen": 116919500, + "step": 5507 + }, + { + "epoch": 0.6622978416401131, + "flos": 12758711895720.0, + "grad_norm": 3.2537295143517193, + "learning_rate": 1.0815543247612034e-06, + "loss": 0.8362, + "num_input_tokens_seen": 116936290, + "step": 5508 + }, + { + "epoch": 0.6624180845307521, + "flos": 15833743374960.0, + "grad_norm": 3.622068864375216, + "learning_rate": 1.0808624174880168e-06, + "loss": 0.8171, + "num_input_tokens_seen": 116956660, + "step": 5509 + }, + { + "epoch": 0.6625383274213912, + "flos": 17431166606160.0, + "grad_norm": 4.486521987811844, + "learning_rate": 1.080170649647272e-06, + "loss": 0.7828, + "num_input_tokens_seen": 116976185, + "step": 5510 + }, + { + "epoch": 0.6626585703120303, + "flos": 24431033618640.0, + "grad_norm": 2.9803708361619905, + "learning_rate": 1.0794790213439068e-06, + "loss": 0.6528, + "num_input_tokens_seen": 117002805, + "step": 5511 + }, + { + "epoch": 0.6627788132026694, + "flos": 16157412569640.0, + "grad_norm": 5.025297256711343, + "learning_rate": 1.078787532682843e-06, + "loss": 0.7662, + "num_input_tokens_seen": 117020000, + "step": 5512 + }, + { + "epoch": 0.6628990560933085, + "flos": 26583963169320.0, + "grad_norm": 4.257117426422324, + "learning_rate": 1.0780961837689773e-06, + "loss": 0.7497, + "num_input_tokens_seen": 117039230, + "step": 5513 + }, + { + "epoch": 0.6630192989839476, + "flos": 13514214402960.0, + "grad_norm": 6.231779471260032, + "learning_rate": 1.0774049747071883e-06, + "loss": 0.6879, + "num_input_tokens_seen": 117056830, + "step": 5514 + }, + { + "epoch": 0.6631395418745867, + "flos": 26314613483280.0, + "grad_norm": 4.050155193981697, + "learning_rate": 1.076713905602332e-06, + "loss": 0.6736, + "num_input_tokens_seen": 117077125, + "step": 5515 + }, + { + "epoch": 0.6632597847652257, + "flos": 14649858490800.0, + "grad_norm": 3.3471873076232908, + "learning_rate": 1.07602297655924e-06, + "loss": 0.7918, + "num_input_tokens_seen": 117095165, + "step": 5516 + }, + { + "epoch": 0.6633800276558649, + "flos": 15513208516320.0, + "grad_norm": 10.420508894254366, + "learning_rate": 1.0753321876827292e-06, + "loss": 0.7935, + "num_input_tokens_seen": 117114170, + "step": 5517 + }, + { + "epoch": 0.663500270546504, + "flos": 17565667319400.0, + "grad_norm": 4.312672708342499, + "learning_rate": 1.0746415390775893e-06, + "loss": 0.7252, + "num_input_tokens_seen": 117132020, + "step": 5518 + }, + { + "epoch": 0.663620513437143, + "flos": 13084470647760.0, + "grad_norm": 3.14675866132626, + "learning_rate": 1.0739510308485939e-06, + "loss": 0.7627, + "num_input_tokens_seen": 117148955, + "step": 5519 + }, + { + "epoch": 0.6637407563277821, + "flos": 42623053265280.0, + "grad_norm": 0.8217577237775026, + "learning_rate": 1.07326066310049e-06, + "loss": 0.6352, + "num_input_tokens_seen": 117212800, + "step": 5520 + }, + { + "epoch": 0.6638609992184212, + "flos": 20021121702840.0, + "grad_norm": 5.104432494838222, + "learning_rate": 1.0725704359380059e-06, + "loss": 0.7767, + "num_input_tokens_seen": 117232375, + "step": 5521 + }, + { + "epoch": 0.6639812421090603, + "flos": 13596833424480.0, + "grad_norm": 3.307611565124761, + "learning_rate": 1.0718803494658497e-06, + "loss": 0.7075, + "num_input_tokens_seen": 117250985, + "step": 5522 + }, + { + "epoch": 0.6641014849996993, + "flos": 11598314210520.0, + "grad_norm": 3.3780220065441813, + "learning_rate": 1.071190403788707e-06, + "loss": 0.8157, + "num_input_tokens_seen": 117266010, + "step": 5523 + }, + { + "epoch": 0.6642217278903385, + "flos": 19428989301000.0, + "grad_norm": 5.576537347097151, + "learning_rate": 1.0705005990112415e-06, + "loss": 0.7322, + "num_input_tokens_seen": 117285510, + "step": 5524 + }, + { + "epoch": 0.6643419707809776, + "flos": 11193387372600.0, + "grad_norm": 19.04911158935634, + "learning_rate": 1.0698109352380957e-06, + "loss": 0.7158, + "num_input_tokens_seen": 117302830, + "step": 5525 + }, + { + "epoch": 0.6644622136716166, + "flos": 18401667630840.0, + "grad_norm": 3.6650786139785687, + "learning_rate": 1.0691214125738909e-06, + "loss": 0.7712, + "num_input_tokens_seen": 117322755, + "step": 5526 + }, + { + "epoch": 0.6645824565622558, + "flos": 48811990163280.0, + "grad_norm": 0.7796408293369229, + "learning_rate": 1.0684320311232287e-06, + "loss": 0.6071, + "num_input_tokens_seen": 117385380, + "step": 5527 + }, + { + "epoch": 0.6647026994528948, + "flos": 18375552655200.0, + "grad_norm": 3.8320955821323333, + "learning_rate": 1.0677427909906865e-06, + "loss": 0.789, + "num_input_tokens_seen": 117405550, + "step": 5528 + }, + { + "epoch": 0.6648229423435339, + "flos": 13298646006240.0, + "grad_norm": 3.3803459496054056, + "learning_rate": 1.0670536922808216e-06, + "loss": 0.6891, + "num_input_tokens_seen": 117425395, + "step": 5529 + }, + { + "epoch": 0.6649431852341731, + "flos": 13354168593360.0, + "grad_norm": 5.798144589640124, + "learning_rate": 1.06636473509817e-06, + "loss": 0.6964, + "num_input_tokens_seen": 117441495, + "step": 5530 + }, + { + "epoch": 0.6650634281248121, + "flos": 12408832745520.0, + "grad_norm": 3.690812126667776, + "learning_rate": 1.0656759195472447e-06, + "loss": 0.7853, + "num_input_tokens_seen": 117458505, + "step": 5531 + }, + { + "epoch": 0.6651836710154512, + "flos": 51099926986560.0, + "grad_norm": 0.8122804418432216, + "learning_rate": 1.0649872457325414e-06, + "loss": 0.6358, + "num_input_tokens_seen": 117519510, + "step": 5532 + }, + { + "epoch": 0.6653039139060903, + "flos": 49314886611960.0, + "grad_norm": 0.8967791968028129, + "learning_rate": 1.0642987137585278e-06, + "loss": 0.6059, + "num_input_tokens_seen": 117578755, + "step": 5533 + }, + { + "epoch": 0.6654241567967294, + "flos": 15702376997760.0, + "grad_norm": 3.075953768309568, + "learning_rate": 1.0636103237296561e-06, + "loss": 0.8069, + "num_input_tokens_seen": 117597400, + "step": 5534 + }, + { + "epoch": 0.6655443996873684, + "flos": 18403345608720.0, + "grad_norm": 4.489781752211917, + "learning_rate": 1.062922075750353e-06, + "loss": 0.819, + "num_input_tokens_seen": 117617135, + "step": 5535 + }, + { + "epoch": 0.6656646425780076, + "flos": 12732786879840.0, + "grad_norm": 6.161062151010431, + "learning_rate": 1.0622339699250267e-06, + "loss": 0.6934, + "num_input_tokens_seen": 117634775, + "step": 5536 + }, + { + "epoch": 0.6657848854686467, + "flos": 17155105008600.0, + "grad_norm": 5.59627488761925, + "learning_rate": 1.0615460063580624e-06, + "loss": 0.7944, + "num_input_tokens_seen": 117652970, + "step": 5537 + }, + { + "epoch": 0.6659051283592857, + "flos": 8331264853440.0, + "grad_norm": 6.6179726235084875, + "learning_rate": 1.060858185153821e-06, + "loss": 0.7138, + "num_input_tokens_seen": 117670790, + "step": 5538 + }, + { + "epoch": 0.6660253712499249, + "flos": 15054468729120.0, + "grad_norm": 4.889953372163542, + "learning_rate": 1.0601705064166474e-06, + "loss": 0.7383, + "num_input_tokens_seen": 117688905, + "step": 5539 + }, + { + "epoch": 0.666145614140564, + "flos": 15540684870240.0, + "grad_norm": 3.7651585966752683, + "learning_rate": 1.0594829702508596e-06, + "loss": 0.7216, + "num_input_tokens_seen": 117706340, + "step": 5540 + }, + { + "epoch": 0.666265857031203, + "flos": 24772301259720.0, + "grad_norm": 3.461836239648907, + "learning_rate": 1.0587955767607592e-06, + "loss": 0.5449, + "num_input_tokens_seen": 117727920, + "step": 5541 + }, + { + "epoch": 0.6663860999218422, + "flos": 12732185340600.0, + "grad_norm": 8.71350858917312, + "learning_rate": 1.0581083260506206e-06, + "loss": 0.7568, + "num_input_tokens_seen": 117744425, + "step": 5542 + }, + { + "epoch": 0.6665063428124812, + "flos": 12895238846400.0, + "grad_norm": 4.445767076040892, + "learning_rate": 1.0574212182246993e-06, + "loss": 0.7444, + "num_input_tokens_seen": 117762840, + "step": 5543 + }, + { + "epoch": 0.6666265857031203, + "flos": 20293257465360.0, + "grad_norm": 9.7124513288577, + "learning_rate": 1.0567342533872303e-06, + "loss": 0.7176, + "num_input_tokens_seen": 117782590, + "step": 5544 + }, + { + "epoch": 0.6667468285937594, + "flos": 18349595979360.0, + "grad_norm": 5.150008614542154, + "learning_rate": 1.0560474316424255e-06, + "loss": 0.7984, + "num_input_tokens_seen": 117802070, + "step": 5545 + }, + { + "epoch": 0.6668670714843985, + "flos": 16671738263880.0, + "grad_norm": 4.85301159826308, + "learning_rate": 1.0553607530944746e-06, + "loss": 0.7173, + "num_input_tokens_seen": 117819845, + "step": 5546 + }, + { + "epoch": 0.6669873143750376, + "flos": 16215942852960.0, + "grad_norm": 4.906494717660329, + "learning_rate": 1.0546742178475463e-06, + "loss": 0.8811, + "num_input_tokens_seen": 117838560, + "step": 5547 + }, + { + "epoch": 0.6671075572656767, + "flos": 14994513747600.0, + "grad_norm": 4.8567562590543165, + "learning_rate": 1.0539878260057868e-06, + "loss": 0.8426, + "num_input_tokens_seen": 117857320, + "step": 5548 + }, + { + "epoch": 0.6672278001563158, + "flos": 13083425869080.0, + "grad_norm": 7.370416801188532, + "learning_rate": 1.0533015776733226e-06, + "loss": 0.6601, + "num_input_tokens_seen": 117873190, + "step": 5549 + }, + { + "epoch": 0.6673480430469548, + "flos": 16347277570200.0, + "grad_norm": 3.492843259136881, + "learning_rate": 1.0526154729542566e-06, + "loss": 0.7691, + "num_input_tokens_seen": 117892970, + "step": 5550 + }, + { + "epoch": 0.6674682859375939, + "flos": 15135314792880.0, + "grad_norm": 5.769675385386458, + "learning_rate": 1.0519295119526699e-06, + "loss": 0.7694, + "num_input_tokens_seen": 117908995, + "step": 5551 + }, + { + "epoch": 0.667588528828233, + "flos": 19206867292560.0, + "grad_norm": 2.4222880349662836, + "learning_rate": 1.0512436947726227e-06, + "loss": 0.8144, + "num_input_tokens_seen": 117930130, + "step": 5552 + }, + { + "epoch": 0.6677087717188721, + "flos": 16886578481520.0, + "grad_norm": 10.298679037478164, + "learning_rate": 1.0505580215181517e-06, + "loss": 0.6436, + "num_input_tokens_seen": 117948090, + "step": 5553 + }, + { + "epoch": 0.6678290146095112, + "flos": 52319108234760.0, + "grad_norm": 0.8154581944399556, + "learning_rate": 1.0498724922932753e-06, + "loss": 0.5848, + "num_input_tokens_seen": 118005925, + "step": 5554 + }, + { + "epoch": 0.6679492575001503, + "flos": 13623613259280.0, + "grad_norm": 13.284236796505285, + "learning_rate": 1.0491871072019851e-06, + "loss": 0.8372, + "num_input_tokens_seen": 118023535, + "step": 5555 + }, + { + "epoch": 0.6680695003907894, + "flos": 21801413083440.0, + "grad_norm": 5.558798426320027, + "learning_rate": 1.0485018663482555e-06, + "loss": 0.6201, + "num_input_tokens_seen": 118043275, + "step": 5556 + }, + { + "epoch": 0.6681897432814284, + "flos": 20696886244920.0, + "grad_norm": 5.046554508012744, + "learning_rate": 1.0478167698360354e-06, + "loss": 0.6823, + "num_input_tokens_seen": 118062295, + "step": 5557 + }, + { + "epoch": 0.6683099861720676, + "flos": 18349342699680.0, + "grad_norm": 5.092006455158898, + "learning_rate": 1.0471318177692556e-06, + "loss": 0.6787, + "num_input_tokens_seen": 118082315, + "step": 5558 + }, + { + "epoch": 0.6684302290627067, + "flos": 16831625773680.0, + "grad_norm": 6.605885696110562, + "learning_rate": 1.046447010251821e-06, + "loss": 0.7153, + "num_input_tokens_seen": 118099365, + "step": 5559 + }, + { + "epoch": 0.6685504719533457, + "flos": 19478908075200.0, + "grad_norm": 3.487861155196297, + "learning_rate": 1.0457623473876157e-06, + "loss": 0.7413, + "num_input_tokens_seen": 118118590, + "step": 5560 + }, + { + "epoch": 0.6686707148439849, + "flos": 21264739948800.0, + "grad_norm": 3.3829616311130413, + "learning_rate": 1.0450778292805046e-06, + "loss": 0.6909, + "num_input_tokens_seen": 118138295, + "step": 5561 + }, + { + "epoch": 0.6687909577346239, + "flos": 17295874393920.0, + "grad_norm": 4.4256982888158936, + "learning_rate": 1.0443934560343267e-06, + "loss": 0.7799, + "num_input_tokens_seen": 118159425, + "step": 5562 + }, + { + "epoch": 0.668911200625263, + "flos": 16944633865440.0, + "grad_norm": 3.4624686890436167, + "learning_rate": 1.0437092277529034e-06, + "loss": 0.7536, + "num_input_tokens_seen": 118178400, + "step": 5563 + }, + { + "epoch": 0.6690314435159022, + "flos": 13542862175400.0, + "grad_norm": 4.82712759437514, + "learning_rate": 1.0430251445400292e-06, + "loss": 0.7203, + "num_input_tokens_seen": 118196165, + "step": 5564 + }, + { + "epoch": 0.6691516864065412, + "flos": 23317357051680.0, + "grad_norm": 3.6940936604818386, + "learning_rate": 1.0423412064994787e-06, + "loss": 0.61, + "num_input_tokens_seen": 118216655, + "step": 5565 + }, + { + "epoch": 0.6692719292971803, + "flos": 25666578574800.0, + "grad_norm": 2.8301278288315626, + "learning_rate": 1.0416574137350064e-06, + "loss": 0.7224, + "num_input_tokens_seen": 118237080, + "step": 5566 + }, + { + "epoch": 0.6693921721878194, + "flos": 14946969470400.0, + "grad_norm": 6.260498080197629, + "learning_rate": 1.0409737663503428e-06, + "loss": 0.7899, + "num_input_tokens_seen": 118255180, + "step": 5567 + }, + { + "epoch": 0.6695124150784585, + "flos": 12108555769920.0, + "grad_norm": 33.544254962893795, + "learning_rate": 1.040290264449196e-06, + "loss": 0.8214, + "num_input_tokens_seen": 118273005, + "step": 5568 + }, + { + "epoch": 0.6696326579690975, + "flos": 19537628318280.0, + "grad_norm": 6.279507067827265, + "learning_rate": 1.0396069081352532e-06, + "loss": 0.62, + "num_input_tokens_seen": 118291880, + "step": 5569 + }, + { + "epoch": 0.6697529008597367, + "flos": 49376107991880.0, + "grad_norm": 0.8634132180818979, + "learning_rate": 1.0389236975121782e-06, + "loss": 0.5864, + "num_input_tokens_seen": 118346450, + "step": 5570 + }, + { + "epoch": 0.6698731437503758, + "flos": 15270923604720.0, + "grad_norm": 3.798356508923863, + "learning_rate": 1.0382406326836147e-06, + "loss": 0.6963, + "num_input_tokens_seen": 118365315, + "step": 5571 + }, + { + "epoch": 0.6699933866410148, + "flos": 14917118619480.0, + "grad_norm": 3.0746640339741127, + "learning_rate": 1.0375577137531828e-06, + "loss": 0.7241, + "num_input_tokens_seen": 118383595, + "step": 5572 + }, + { + "epoch": 0.670113629531654, + "flos": 21292184642760.0, + "grad_norm": 3.2903670308574857, + "learning_rate": 1.0368749408244802e-06, + "loss": 0.7004, + "num_input_tokens_seen": 118406235, + "step": 5573 + }, + { + "epoch": 0.670233872422293, + "flos": 14460278429880.0, + "grad_norm": 2.859401515149859, + "learning_rate": 1.0361923140010836e-06, + "loss": 0.7631, + "num_input_tokens_seen": 118424440, + "step": 5574 + }, + { + "epoch": 0.6703541153129321, + "flos": 17755152400440.0, + "grad_norm": 4.236759429404327, + "learning_rate": 1.0355098333865455e-06, + "loss": 0.6143, + "num_input_tokens_seen": 118443390, + "step": 5575 + }, + { + "epoch": 0.6704743582035713, + "flos": 19564534792920.0, + "grad_norm": 13.573313818757045, + "learning_rate": 1.0348274990844006e-06, + "loss": 0.6842, + "num_input_tokens_seen": 118465870, + "step": 5576 + }, + { + "epoch": 0.6705946010942103, + "flos": 17215471569600.0, + "grad_norm": 4.716611466011747, + "learning_rate": 1.034145311198155e-06, + "loss": 0.7083, + "num_input_tokens_seen": 118485605, + "step": 5577 + }, + { + "epoch": 0.6707148439848494, + "flos": 17619923508120.0, + "grad_norm": 5.2076099758100405, + "learning_rate": 1.0334632698312989e-06, + "loss": 0.6301, + "num_input_tokens_seen": 118506120, + "step": 5578 + }, + { + "epoch": 0.6708350868754885, + "flos": 16485894078240.0, + "grad_norm": 4.922147034892951, + "learning_rate": 1.032781375087295e-06, + "loss": 0.7373, + "num_input_tokens_seen": 118525740, + "step": 5579 + }, + { + "epoch": 0.6709553297661276, + "flos": 18483273533640.0, + "grad_norm": 5.396270537859776, + "learning_rate": 1.0320996270695891e-06, + "loss": 0.6551, + "num_input_tokens_seen": 118546530, + "step": 5580 + }, + { + "epoch": 0.6710755726567667, + "flos": 14946589550880.0, + "grad_norm": 3.244109170245613, + "learning_rate": 1.0314180258815998e-06, + "loss": 0.7128, + "num_input_tokens_seen": 118564890, + "step": 5581 + }, + { + "epoch": 0.6711958155474057, + "flos": 19051602136920.0, + "grad_norm": 3.4403986413763845, + "learning_rate": 1.0307365716267247e-06, + "loss": 0.7231, + "num_input_tokens_seen": 118585055, + "step": 5582 + }, + { + "epoch": 0.6713160584380449, + "flos": 14568347567880.0, + "grad_norm": 3.9117122454291087, + "learning_rate": 1.0300552644083423e-06, + "loss": 0.7613, + "num_input_tokens_seen": 118603700, + "step": 5583 + }, + { + "epoch": 0.6714363013286839, + "flos": 13299690784920.0, + "grad_norm": 10.022227570315463, + "learning_rate": 1.0293741043298036e-06, + "loss": 0.7003, + "num_input_tokens_seen": 118621770, + "step": 5584 + }, + { + "epoch": 0.671556544219323, + "flos": 18916056645000.0, + "grad_norm": 4.247297662021061, + "learning_rate": 1.0286930914944436e-06, + "loss": 0.7029, + "num_input_tokens_seen": 118641305, + "step": 5585 + }, + { + "epoch": 0.6716767871099621, + "flos": 11543361502680.0, + "grad_norm": 5.14711474732864, + "learning_rate": 1.0280122260055684e-06, + "loss": 0.7583, + "num_input_tokens_seen": 118656735, + "step": 5586 + }, + { + "epoch": 0.6717970300006012, + "flos": 14434068474360.0, + "grad_norm": 3.937438358853755, + "learning_rate": 1.0273315079664652e-06, + "loss": 0.7912, + "num_input_tokens_seen": 118674410, + "step": 5587 + }, + { + "epoch": 0.6719172728912403, + "flos": 18673961693160.0, + "grad_norm": 3.4378032203800823, + "learning_rate": 1.0266509374803992e-06, + "loss": 0.7232, + "num_input_tokens_seen": 118695290, + "step": 5588 + }, + { + "epoch": 0.6720375157818794, + "flos": 11564569184520.0, + "grad_norm": 5.217312552023592, + "learning_rate": 1.0259705146506123e-06, + "loss": 0.8274, + "num_input_tokens_seen": 118709905, + "step": 5589 + }, + { + "epoch": 0.6721577586725185, + "flos": 23502757997880.0, + "grad_norm": 6.801631832667673, + "learning_rate": 1.025290239580324e-06, + "loss": 0.7592, + "num_input_tokens_seen": 118730295, + "step": 5590 + }, + { + "epoch": 0.6722780015631575, + "flos": 15160163370120.0, + "grad_norm": 2.7883609348837957, + "learning_rate": 1.0246101123727313e-06, + "loss": 0.7415, + "num_input_tokens_seen": 118748995, + "step": 5591 + }, + { + "epoch": 0.6723982444537967, + "flos": 12327385142520.0, + "grad_norm": 4.576981437473774, + "learning_rate": 1.0239301331310085e-06, + "loss": 0.7741, + "num_input_tokens_seen": 118766335, + "step": 5592 + }, + { + "epoch": 0.6725184873444358, + "flos": 15108724917840.0, + "grad_norm": 8.61275597179483, + "learning_rate": 1.0232503019583088e-06, + "loss": 0.8716, + "num_input_tokens_seen": 118785665, + "step": 5593 + }, + { + "epoch": 0.6726387302350748, + "flos": 17372478023040.0, + "grad_norm": 3.1457890050598545, + "learning_rate": 1.0225706189577619e-06, + "loss": 0.6794, + "num_input_tokens_seen": 118803910, + "step": 5594 + }, + { + "epoch": 0.672758973125714, + "flos": 11052554667360.0, + "grad_norm": 3.6831541153012517, + "learning_rate": 1.021891084232475e-06, + "loss": 0.7204, + "num_input_tokens_seen": 118821565, + "step": 5595 + }, + { + "epoch": 0.672879216016353, + "flos": 13191526667040.0, + "grad_norm": 4.384010627162889, + "learning_rate": 1.0212116978855325e-06, + "loss": 0.7807, + "num_input_tokens_seen": 118839300, + "step": 5596 + }, + { + "epoch": 0.6729994589069921, + "flos": 17187108736800.0, + "grad_norm": 4.0228529422137385, + "learning_rate": 1.020532460019997e-06, + "loss": 0.7835, + "num_input_tokens_seen": 118858270, + "step": 5597 + }, + { + "epoch": 0.6731197017976313, + "flos": 19293063889560.0, + "grad_norm": 10.018511994331613, + "learning_rate": 1.0198533707389096e-06, + "loss": 0.6851, + "num_input_tokens_seen": 118878865, + "step": 5598 + }, + { + "epoch": 0.6732399446882703, + "flos": 15810762735360.0, + "grad_norm": 2.8743801277979206, + "learning_rate": 1.0191744301452853e-06, + "loss": 0.7173, + "num_input_tokens_seen": 118897885, + "step": 5599 + }, + { + "epoch": 0.6733601875789094, + "flos": 18965500519800.0, + "grad_norm": 4.246865631385109, + "learning_rate": 1.0184956383421208e-06, + "loss": 0.691, + "num_input_tokens_seen": 118916255, + "step": 5600 + }, + { + "epoch": 0.6734804304695485, + "flos": 16782530158440.0, + "grad_norm": 5.0404654508953755, + "learning_rate": 1.017816995432387e-06, + "loss": 0.6459, + "num_input_tokens_seen": 118935075, + "step": 5601 + }, + { + "epoch": 0.6736006733601876, + "flos": 13650962973360.0, + "grad_norm": 3.78127887809382, + "learning_rate": 1.0171385015190353e-06, + "loss": 0.7193, + "num_input_tokens_seen": 118954655, + "step": 5602 + }, + { + "epoch": 0.6737209162508266, + "flos": 14191055383680.0, + "grad_norm": 3.493308755652282, + "learning_rate": 1.0164601567049908e-06, + "loss": 0.7127, + "num_input_tokens_seen": 118972905, + "step": 5603 + }, + { + "epoch": 0.6738411591414658, + "flos": 14731622693400.0, + "grad_norm": 5.769822161758568, + "learning_rate": 1.015781961093158e-06, + "loss": 0.7887, + "num_input_tokens_seen": 118991945, + "step": 5604 + }, + { + "epoch": 0.6739614020321049, + "flos": 15838587348840.0, + "grad_norm": 3.0126035110608163, + "learning_rate": 1.0151039147864197e-06, + "loss": 0.7609, + "num_input_tokens_seen": 119011640, + "step": 5605 + }, + { + "epoch": 0.6740816449227439, + "flos": 14001475322760.0, + "grad_norm": 3.99584351592346, + "learning_rate": 1.0144260178876336e-06, + "loss": 0.6444, + "num_input_tokens_seen": 119030705, + "step": 5606 + }, + { + "epoch": 0.6742018878133831, + "flos": 15425143981680.0, + "grad_norm": 5.300033302922253, + "learning_rate": 1.0137482704996388e-06, + "loss": 0.6595, + "num_input_tokens_seen": 119044775, + "step": 5607 + }, + { + "epoch": 0.6743221307040221, + "flos": 17241681525120.0, + "grad_norm": 3.7240547363323246, + "learning_rate": 1.0130706727252461e-06, + "loss": 0.7801, + "num_input_tokens_seen": 119061550, + "step": 5608 + }, + { + "epoch": 0.6744423735946612, + "flos": 11838414584880.0, + "grad_norm": 6.911091898039306, + "learning_rate": 1.0123932246672468e-06, + "loss": 0.669, + "num_input_tokens_seen": 119075415, + "step": 5609 + }, + { + "epoch": 0.6745626164853004, + "flos": 42625712701920.0, + "grad_norm": 0.7942016863298537, + "learning_rate": 1.0117159264284114e-06, + "loss": 0.5743, + "num_input_tokens_seen": 119138305, + "step": 5610 + }, + { + "epoch": 0.6746828593759394, + "flos": 14973401045640.0, + "grad_norm": 5.618261242614016, + "learning_rate": 1.0110387781114837e-06, + "loss": 0.7594, + "num_input_tokens_seen": 119156640, + "step": 5611 + }, + { + "epoch": 0.6748031022665785, + "flos": 14028508437240.0, + "grad_norm": 4.577447730473798, + "learning_rate": 1.0103617798191872e-06, + "loss": 0.7462, + "num_input_tokens_seen": 119175835, + "step": 5612 + }, + { + "epoch": 0.6749233451572175, + "flos": 11058506739840.0, + "grad_norm": 12.5721847628769, + "learning_rate": 1.0096849316542217e-06, + "loss": 0.7929, + "num_input_tokens_seen": 119192105, + "step": 5613 + }, + { + "epoch": 0.6750435880478567, + "flos": 19423702087680.0, + "grad_norm": 6.388909517706402, + "learning_rate": 1.0090082337192643e-06, + "loss": 0.7176, + "num_input_tokens_seen": 119211470, + "step": 5614 + }, + { + "epoch": 0.6751638309384957, + "flos": 17133928986720.0, + "grad_norm": 8.960374703065488, + "learning_rate": 1.0083316861169705e-06, + "loss": 0.7664, + "num_input_tokens_seen": 119229925, + "step": 5615 + }, + { + "epoch": 0.6752840738291348, + "flos": 17161468660560.0, + "grad_norm": 5.142633016940236, + "learning_rate": 1.0076552889499713e-06, + "loss": 0.6966, + "num_input_tokens_seen": 119250410, + "step": 5616 + }, + { + "epoch": 0.675404316719774, + "flos": 22263603806280.0, + "grad_norm": 7.489137539884589, + "learning_rate": 1.006979042320876e-06, + "loss": 0.7178, + "num_input_tokens_seen": 119270345, + "step": 5617 + }, + { + "epoch": 0.675524559610413, + "flos": 17295399494520.0, + "grad_norm": 3.8933508948578885, + "learning_rate": 1.0063029463322702e-06, + "loss": 0.6207, + "num_input_tokens_seen": 119290340, + "step": 5618 + }, + { + "epoch": 0.6756448025010521, + "flos": 15538152073440.0, + "grad_norm": 4.6454481754961705, + "learning_rate": 1.0056270010867164e-06, + "loss": 0.7337, + "num_input_tokens_seen": 119307630, + "step": 5619 + }, + { + "epoch": 0.6757650453916912, + "flos": 15832128717000.0, + "grad_norm": 6.610436897555449, + "learning_rate": 1.004951206686758e-06, + "loss": 0.7592, + "num_input_tokens_seen": 119325625, + "step": 5620 + }, + { + "epoch": 0.6758852882823303, + "flos": 15942952271520.0, + "grad_norm": 3.3005485136606962, + "learning_rate": 1.0042755632349087e-06, + "loss": 0.6955, + "num_input_tokens_seen": 119342235, + "step": 5621 + }, + { + "epoch": 0.6760055311729694, + "flos": 19860600993840.0, + "grad_norm": 9.157547686278507, + "learning_rate": 1.0036000708336653e-06, + "loss": 0.6109, + "num_input_tokens_seen": 119361085, + "step": 5622 + }, + { + "epoch": 0.6761257740636085, + "flos": 13133566263000.0, + "grad_norm": 3.3957775712066818, + "learning_rate": 1.0029247295854984e-06, + "loss": 0.7881, + "num_input_tokens_seen": 119377425, + "step": 5623 + }, + { + "epoch": 0.6762460169542476, + "flos": 11004377190960.0, + "grad_norm": 4.278448817367916, + "learning_rate": 1.0022495395928588e-06, + "loss": 0.7017, + "num_input_tokens_seen": 119395625, + "step": 5624 + }, + { + "epoch": 0.6763662598448866, + "flos": 50060167443480.0, + "grad_norm": 0.7877930353310527, + "learning_rate": 1.0015745009581697e-06, + "loss": 0.6298, + "num_input_tokens_seen": 119456950, + "step": 5625 + }, + { + "epoch": 0.6764865027355258, + "flos": 15082166702760.0, + "grad_norm": 3.5323182237294826, + "learning_rate": 1.0008996137838343e-06, + "loss": 0.6545, + "num_input_tokens_seen": 119475645, + "step": 5626 + }, + { + "epoch": 0.6766067456261649, + "flos": 15514379934840.0, + "grad_norm": 3.2183530988700335, + "learning_rate": 1.000224878172234e-06, + "loss": 0.7809, + "num_input_tokens_seen": 119494490, + "step": 5627 + }, + { + "epoch": 0.6767269885168039, + "flos": 14568822467280.0, + "grad_norm": 5.47038375428904, + "learning_rate": 9.99550294225724e-07, + "loss": 0.712, + "num_input_tokens_seen": 119513365, + "step": 5628 + }, + { + "epoch": 0.6768472314074431, + "flos": 15217015675560.0, + "grad_norm": 9.191510836634038, + "learning_rate": 9.988758620466402e-07, + "loss": 0.7075, + "num_input_tokens_seen": 119531955, + "step": 5629 + }, + { + "epoch": 0.6769674742980821, + "flos": 16972458478920.0, + "grad_norm": 3.2427110364203986, + "learning_rate": 9.982015817372917e-07, + "loss": 0.7436, + "num_input_tokens_seen": 119552115, + "step": 5630 + }, + { + "epoch": 0.6770877171887212, + "flos": 17754487541280.0, + "grad_norm": 4.008316930365888, + "learning_rate": 9.975274533999657e-07, + "loss": 0.7975, + "num_input_tokens_seen": 119571365, + "step": 5631 + }, + { + "epoch": 0.6772079600793603, + "flos": 13237899525720.0, + "grad_norm": 6.995471034193078, + "learning_rate": 9.96853477136929e-07, + "loss": 0.8088, + "num_input_tokens_seen": 119585830, + "step": 5632 + }, + { + "epoch": 0.6773282029699994, + "flos": 16428693513240.0, + "grad_norm": 5.108465364543182, + "learning_rate": 9.96179653050422e-07, + "loss": 0.7322, + "num_input_tokens_seen": 119605710, + "step": 5633 + }, + { + "epoch": 0.6774484458606385, + "flos": 13647227098080.0, + "grad_norm": 6.232004733278488, + "learning_rate": 9.955059812426635e-07, + "loss": 0.7148, + "num_input_tokens_seen": 119622960, + "step": 5634 + }, + { + "epoch": 0.6775686887512776, + "flos": 19050399058440.0, + "grad_norm": 3.83438992796637, + "learning_rate": 9.948324618158493e-07, + "loss": 0.8113, + "num_input_tokens_seen": 119643020, + "step": 5635 + }, + { + "epoch": 0.6776889316419167, + "flos": 9868954722840.0, + "grad_norm": 5.266173382284822, + "learning_rate": 9.941590948721502e-07, + "loss": 0.7592, + "num_input_tokens_seen": 119659940, + "step": 5636 + }, + { + "epoch": 0.6778091745325557, + "flos": 20239982735400.0, + "grad_norm": 2.9794366531923013, + "learning_rate": 9.934858805137188e-07, + "loss": 0.7438, + "num_input_tokens_seen": 119680310, + "step": 5637 + }, + { + "epoch": 0.6779294174231949, + "flos": 13677616168320.0, + "grad_norm": 5.990334354806336, + "learning_rate": 9.92812818842677e-07, + "loss": 0.7965, + "num_input_tokens_seen": 119699205, + "step": 5638 + }, + { + "epoch": 0.678049660313834, + "flos": 33762044317080.0, + "grad_norm": 7.374464112548493, + "learning_rate": 9.921399099611306e-07, + "loss": 0.6135, + "num_input_tokens_seen": 119720090, + "step": 5639 + }, + { + "epoch": 0.678169903204473, + "flos": 14595665622000.0, + "grad_norm": 4.710980821603254, + "learning_rate": 9.914671539711588e-07, + "loss": 0.6854, + "num_input_tokens_seen": 119739330, + "step": 5640 + }, + { + "epoch": 0.6782901460951122, + "flos": 15647076030360.0, + "grad_norm": 4.560418268060846, + "learning_rate": 9.90794550974817e-07, + "loss": 0.7605, + "num_input_tokens_seen": 119759445, + "step": 5641 + }, + { + "epoch": 0.6784103889857512, + "flos": 15675755462760.0, + "grad_norm": 3.795701712968954, + "learning_rate": 9.901221010741407e-07, + "loss": 0.7939, + "num_input_tokens_seen": 119778485, + "step": 5642 + }, + { + "epoch": 0.6785306318763903, + "flos": 23992615034400.0, + "grad_norm": 3.5558339711961935, + "learning_rate": 9.894498043711375e-07, + "loss": 0.7328, + "num_input_tokens_seen": 119799950, + "step": 5643 + }, + { + "epoch": 0.6786508747670293, + "flos": 18782980629960.0, + "grad_norm": 4.061645183128129, + "learning_rate": 9.887776609677962e-07, + "loss": 0.6821, + "num_input_tokens_seen": 119821040, + "step": 5644 + }, + { + "epoch": 0.6787711176576685, + "flos": 14001633622560.0, + "grad_norm": 3.1411560520585073, + "learning_rate": 9.88105670966079e-07, + "loss": 0.7114, + "num_input_tokens_seen": 119839220, + "step": 5645 + }, + { + "epoch": 0.6788913605483076, + "flos": 10162298167200.0, + "grad_norm": 3.35698284765003, + "learning_rate": 9.874338344679283e-07, + "loss": 0.7746, + "num_input_tokens_seen": 119854785, + "step": 5646 + }, + { + "epoch": 0.6790116034389466, + "flos": 16107620435280.0, + "grad_norm": 3.6727406881363787, + "learning_rate": 9.86762151575259e-07, + "loss": 0.7309, + "num_input_tokens_seen": 119874500, + "step": 5647 + }, + { + "epoch": 0.6791318463295858, + "flos": 15297070240320.0, + "grad_norm": 2.2550036636288944, + "learning_rate": 9.860906223899651e-07, + "loss": 0.7935, + "num_input_tokens_seen": 119893615, + "step": 5648 + }, + { + "epoch": 0.6792520892202248, + "flos": 20914575858960.0, + "grad_norm": 7.187805169302759, + "learning_rate": 9.854192470139184e-07, + "loss": 0.7375, + "num_input_tokens_seen": 119914815, + "step": 5649 + }, + { + "epoch": 0.6793723321108639, + "flos": 14623426915560.0, + "grad_norm": 4.857605050639968, + "learning_rate": 9.847480255489645e-07, + "loss": 0.7053, + "num_input_tokens_seen": 119933560, + "step": 5650 + }, + { + "epoch": 0.6794925750015031, + "flos": 19535285481240.0, + "grad_norm": 3.350000119565136, + "learning_rate": 9.840769580969295e-07, + "loss": 0.6756, + "num_input_tokens_seen": 119953720, + "step": 5651 + }, + { + "epoch": 0.6796128178921421, + "flos": 15784141200360.0, + "grad_norm": 4.805480196930969, + "learning_rate": 9.834060447596114e-07, + "loss": 0.7743, + "num_input_tokens_seen": 119972710, + "step": 5652 + }, + { + "epoch": 0.6797330607827812, + "flos": 16458734323920.0, + "grad_norm": 4.3427219409261255, + "learning_rate": 9.827352856387868e-07, + "loss": 0.7642, + "num_input_tokens_seen": 119992140, + "step": 5653 + }, + { + "epoch": 0.6798533036734203, + "flos": 48889005371880.0, + "grad_norm": 0.8457438963473344, + "learning_rate": 9.820646808362118e-07, + "loss": 0.6478, + "num_input_tokens_seen": 120058115, + "step": 5654 + }, + { + "epoch": 0.6799735465640594, + "flos": 11787577671840.0, + "grad_norm": 5.142814378711983, + "learning_rate": 9.813942304536154e-07, + "loss": 0.6939, + "num_input_tokens_seen": 120075805, + "step": 5655 + }, + { + "epoch": 0.6800937894546984, + "flos": 16187485040280.0, + "grad_norm": 3.292853338678537, + "learning_rate": 9.807239345927043e-07, + "loss": 0.6139, + "num_input_tokens_seen": 120095535, + "step": 5656 + }, + { + "epoch": 0.6802140323453376, + "flos": 23208781354320.0, + "grad_norm": 10.517549290524542, + "learning_rate": 9.80053793355162e-07, + "loss": 0.7083, + "num_input_tokens_seen": 120113950, + "step": 5657 + }, + { + "epoch": 0.6803342752359767, + "flos": 12921987021240.0, + "grad_norm": 3.702965455300248, + "learning_rate": 9.793838068426472e-07, + "loss": 0.7212, + "num_input_tokens_seen": 120131365, + "step": 5658 + }, + { + "epoch": 0.6804545181266157, + "flos": 8195656041600.0, + "grad_norm": 5.844389365403657, + "learning_rate": 9.78713975156799e-07, + "loss": 0.5965, + "num_input_tokens_seen": 120146950, + "step": 5659 + }, + { + "epoch": 0.6805747610172549, + "flos": 21534786153960.0, + "grad_norm": 3.7539653158155564, + "learning_rate": 9.780442983992273e-07, + "loss": 0.7082, + "num_input_tokens_seen": 120165185, + "step": 5660 + }, + { + "epoch": 0.680695003907894, + "flos": 27663008231400.0, + "grad_norm": 4.203398855143871, + "learning_rate": 9.773747766715238e-07, + "loss": 0.7002, + "num_input_tokens_seen": 120185725, + "step": 5661 + }, + { + "epoch": 0.680815246798533, + "flos": 16188909738480.0, + "grad_norm": 9.781890750653211, + "learning_rate": 9.767054100752536e-07, + "loss": 0.7857, + "num_input_tokens_seen": 120205395, + "step": 5662 + }, + { + "epoch": 0.6809354896891722, + "flos": 12543681718320.0, + "grad_norm": 3.065503263243738, + "learning_rate": 9.760361987119584e-07, + "loss": 0.7965, + "num_input_tokens_seen": 120222850, + "step": 5663 + }, + { + "epoch": 0.6810557325798112, + "flos": 9032986071360.0, + "grad_norm": 4.4265241681185366, + "learning_rate": 9.753671426831592e-07, + "loss": 0.657, + "num_input_tokens_seen": 120238585, + "step": 5664 + }, + { + "epoch": 0.6811759754704503, + "flos": 16209135961560.0, + "grad_norm": 5.31978565357895, + "learning_rate": 9.746982420903483e-07, + "loss": 0.7777, + "num_input_tokens_seen": 120256500, + "step": 5665 + }, + { + "epoch": 0.6812962183610894, + "flos": 12782452374360.0, + "grad_norm": 2.544975086982066, + "learning_rate": 9.740294970349993e-07, + "loss": 0.7371, + "num_input_tokens_seen": 120272635, + "step": 5666 + }, + { + "epoch": 0.6814164612517285, + "flos": 44423095969560.0, + "grad_norm": 0.9713943752663188, + "learning_rate": 9.733609076185594e-07, + "loss": 0.6711, + "num_input_tokens_seen": 120328760, + "step": 5667 + }, + { + "epoch": 0.6815367041423676, + "flos": 14109576120720.0, + "grad_norm": 4.030086378965571, + "learning_rate": 9.72692473942455e-07, + "loss": 0.8253, + "num_input_tokens_seen": 120345705, + "step": 5668 + }, + { + "epoch": 0.6816569470330067, + "flos": 16214011595400.0, + "grad_norm": 4.576388605889926, + "learning_rate": 9.720241961080849e-07, + "loss": 0.7609, + "num_input_tokens_seen": 120364740, + "step": 5669 + }, + { + "epoch": 0.6817771899236458, + "flos": 30498762495240.0, + "grad_norm": 4.722659279039041, + "learning_rate": 9.713560742168259e-07, + "loss": 0.6973, + "num_input_tokens_seen": 120387085, + "step": 5670 + }, + { + "epoch": 0.6818974328142848, + "flos": 15433185611520.0, + "grad_norm": 27.6971162582576, + "learning_rate": 9.706881083700333e-07, + "loss": 0.6995, + "num_input_tokens_seen": 120406490, + "step": 5671 + }, + { + "epoch": 0.682017675704924, + "flos": 14940542498520.0, + "grad_norm": 6.104882971839306, + "learning_rate": 9.700202986690357e-07, + "loss": 0.8005, + "num_input_tokens_seen": 120424510, + "step": 5672 + }, + { + "epoch": 0.682137918595563, + "flos": 14646945774480.0, + "grad_norm": 4.309501600722624, + "learning_rate": 9.693526452151413e-07, + "loss": 0.6432, + "num_input_tokens_seen": 120443280, + "step": 5673 + }, + { + "epoch": 0.6822581614862021, + "flos": 23261581184880.0, + "grad_norm": 104.49389747432622, + "learning_rate": 9.686851481096305e-07, + "loss": 0.7456, + "num_input_tokens_seen": 120464310, + "step": 5674 + }, + { + "epoch": 0.6823784043768413, + "flos": 17188185175440.0, + "grad_norm": 20.755951177369436, + "learning_rate": 9.68017807453762e-07, + "loss": 0.7204, + "num_input_tokens_seen": 120482775, + "step": 5675 + }, + { + "epoch": 0.6824986472674803, + "flos": 10275622858560.0, + "grad_norm": 3.6110224411340255, + "learning_rate": 9.673506233487721e-07, + "loss": 0.71, + "num_input_tokens_seen": 120500460, + "step": 5676 + }, + { + "epoch": 0.6826188901581194, + "flos": 15728017074000.0, + "grad_norm": 3.07727286844777, + "learning_rate": 9.666835958958717e-07, + "loss": 0.8483, + "num_input_tokens_seen": 120519500, + "step": 5677 + }, + { + "epoch": 0.6827391330487584, + "flos": 15213533079960.0, + "grad_norm": 4.266715221327435, + "learning_rate": 9.660167251962484e-07, + "loss": 0.7826, + "num_input_tokens_seen": 120537580, + "step": 5678 + }, + { + "epoch": 0.6828593759393976, + "flos": 15864227425080.0, + "grad_norm": 3.517167901450977, + "learning_rate": 9.653500113510654e-07, + "loss": 0.761, + "num_input_tokens_seen": 120556415, + "step": 5679 + }, + { + "epoch": 0.6829796188300367, + "flos": 18564721136640.0, + "grad_norm": 4.294132869702699, + "learning_rate": 9.646834544614627e-07, + "loss": 0.6546, + "num_input_tokens_seen": 120576635, + "step": 5680 + }, + { + "epoch": 0.6830998617206757, + "flos": 15136169611800.0, + "grad_norm": 4.952390598280739, + "learning_rate": 9.64017054628558e-07, + "loss": 0.7463, + "num_input_tokens_seen": 120595180, + "step": 5681 + }, + { + "epoch": 0.6832201046113149, + "flos": 15892431958080.0, + "grad_norm": 3.3005116730708814, + "learning_rate": 9.63350811953441e-07, + "loss": 0.7845, + "num_input_tokens_seen": 120615275, + "step": 5682 + }, + { + "epoch": 0.6833403475019539, + "flos": 14271489867960.0, + "grad_norm": 8.646567979110719, + "learning_rate": 9.626847265371826e-07, + "loss": 0.6896, + "num_input_tokens_seen": 120634315, + "step": 5683 + }, + { + "epoch": 0.683460590392593, + "flos": 14134741297560.0, + "grad_norm": 4.91321550356512, + "learning_rate": 9.620187984808262e-07, + "loss": 0.7606, + "num_input_tokens_seen": 120652835, + "step": 5684 + }, + { + "epoch": 0.6835808332832322, + "flos": 17047320810240.0, + "grad_norm": 4.343350398941845, + "learning_rate": 9.613530278853919e-07, + "loss": 0.8482, + "num_input_tokens_seen": 120672530, + "step": 5685 + }, + { + "epoch": 0.6837010761738712, + "flos": 15838207429320.0, + "grad_norm": 8.944013023107667, + "learning_rate": 9.60687414851879e-07, + "loss": 0.7239, + "num_input_tokens_seen": 120693255, + "step": 5686 + }, + { + "epoch": 0.6838213190645103, + "flos": 12813442983840.0, + "grad_norm": 3.838104816701125, + "learning_rate": 9.600219594812575e-07, + "loss": 0.746, + "num_input_tokens_seen": 120710915, + "step": 5687 + }, + { + "epoch": 0.6839415619551494, + "flos": 16918012330440.0, + "grad_norm": 3.648220276339188, + "learning_rate": 9.593566618744786e-07, + "loss": 0.7099, + "num_input_tokens_seen": 120730785, + "step": 5688 + }, + { + "epoch": 0.6840618048457885, + "flos": 16189352977920.0, + "grad_norm": 3.5263648887506354, + "learning_rate": 9.58691522132466e-07, + "loss": 0.7214, + "num_input_tokens_seen": 120749315, + "step": 5689 + }, + { + "epoch": 0.6841820477364275, + "flos": 16106132417160.0, + "grad_norm": 6.284999658164271, + "learning_rate": 9.58026540356123e-07, + "loss": 0.8387, + "num_input_tokens_seen": 120767300, + "step": 5690 + }, + { + "epoch": 0.6843022906270667, + "flos": 18241273561680.0, + "grad_norm": 5.730279358982633, + "learning_rate": 9.573617166463246e-07, + "loss": 0.845, + "num_input_tokens_seen": 120788235, + "step": 5691 + }, + { + "epoch": 0.6844225335177058, + "flos": 14591581487160.0, + "grad_norm": 5.056991971607561, + "learning_rate": 9.56697051103924e-07, + "loss": 0.5835, + "num_input_tokens_seen": 120805395, + "step": 5692 + }, + { + "epoch": 0.6845427764083448, + "flos": 18968286596280.0, + "grad_norm": 3.4462468616801996, + "learning_rate": 9.560325438297522e-07, + "loss": 0.7929, + "num_input_tokens_seen": 120823425, + "step": 5693 + }, + { + "epoch": 0.684663019298984, + "flos": 13785748626240.0, + "grad_norm": 4.562276338294612, + "learning_rate": 9.553681949246127e-07, + "loss": 0.8645, + "num_input_tokens_seen": 120840770, + "step": 5694 + }, + { + "epoch": 0.684783262189623, + "flos": 39919515706200.0, + "grad_norm": 5.28072395187749, + "learning_rate": 9.547040044892886e-07, + "loss": 0.751, + "num_input_tokens_seen": 120868005, + "step": 5695 + }, + { + "epoch": 0.6849035050802621, + "flos": 47159740864080.0, + "grad_norm": 0.8645151563367115, + "learning_rate": 9.540399726245354e-07, + "loss": 0.6305, + "num_input_tokens_seen": 120924430, + "step": 5696 + }, + { + "epoch": 0.6850237479709013, + "flos": 18480075877680.0, + "grad_norm": 4.2670446695332345, + "learning_rate": 9.533760994310859e-07, + "loss": 0.6694, + "num_input_tokens_seen": 120944550, + "step": 5697 + }, + { + "epoch": 0.6851439908615403, + "flos": 14136482595360.0, + "grad_norm": 2.866789662879264, + "learning_rate": 9.527123850096508e-07, + "loss": 0.7337, + "num_input_tokens_seen": 120962630, + "step": 5698 + }, + { + "epoch": 0.6852642337521794, + "flos": 16969039203240.0, + "grad_norm": 3.6543482864001113, + "learning_rate": 9.520488294609142e-07, + "loss": 0.7004, + "num_input_tokens_seen": 120981130, + "step": 5699 + }, + { + "epoch": 0.6853844766428185, + "flos": 46179235292040.0, + "grad_norm": 0.9157622745450478, + "learning_rate": 9.513854328855368e-07, + "loss": 0.5901, + "num_input_tokens_seen": 121038725, + "step": 5700 + }, + { + "epoch": 0.6855047195334576, + "flos": 17157796105200.0, + "grad_norm": 3.268821648419694, + "learning_rate": 9.507221953841558e-07, + "loss": 0.7891, + "num_input_tokens_seen": 121056075, + "step": 5701 + }, + { + "epoch": 0.6856249624240967, + "flos": 15106097141160.0, + "grad_norm": 3.155410503880287, + "learning_rate": 9.500591170573824e-07, + "loss": 0.7609, + "num_input_tokens_seen": 121075815, + "step": 5702 + }, + { + "epoch": 0.6857452053147358, + "flos": 12458213300400.0, + "grad_norm": 3.3326946282411507, + "learning_rate": 9.493961980058078e-07, + "loss": 0.7331, + "num_input_tokens_seen": 121093130, + "step": 5703 + }, + { + "epoch": 0.6858654482053749, + "flos": 22642194048840.0, + "grad_norm": 4.257740601098794, + "learning_rate": 9.48733438329993e-07, + "loss": 0.6577, + "num_input_tokens_seen": 121113115, + "step": 5704 + }, + { + "epoch": 0.6859856910960139, + "flos": 21184337124480.0, + "grad_norm": 4.205319877369151, + "learning_rate": 9.480708381304807e-07, + "loss": 0.7203, + "num_input_tokens_seen": 121134130, + "step": 5705 + }, + { + "epoch": 0.6861059339866531, + "flos": 14136830854920.0, + "grad_norm": 3.944288305914432, + "learning_rate": 9.474083975077858e-07, + "loss": 0.8321, + "num_input_tokens_seen": 121150975, + "step": 5706 + }, + { + "epoch": 0.6862261768772921, + "flos": 16242596047920.0, + "grad_norm": 5.783278650617104, + "learning_rate": 9.467461165623994e-07, + "loss": 0.7811, + "num_input_tokens_seen": 121169745, + "step": 5707 + }, + { + "epoch": 0.6863464197679312, + "flos": 19266189074880.0, + "grad_norm": 5.515310478428383, + "learning_rate": 9.46083995394791e-07, + "loss": 0.78, + "num_input_tokens_seen": 121187275, + "step": 5708 + }, + { + "epoch": 0.6864666626585703, + "flos": 27798838662960.0, + "grad_norm": 4.357772477910966, + "learning_rate": 9.454220341054012e-07, + "loss": 0.6149, + "num_input_tokens_seen": 121211780, + "step": 5709 + }, + { + "epoch": 0.6865869055492094, + "flos": 14163958949280.0, + "grad_norm": 5.621064229096387, + "learning_rate": 9.447602327946512e-07, + "loss": 0.7967, + "num_input_tokens_seen": 121230140, + "step": 5710 + }, + { + "epoch": 0.6867071484398485, + "flos": 14892744941640.0, + "grad_norm": 5.018935967559256, + "learning_rate": 9.440985915629338e-07, + "loss": 0.7571, + "num_input_tokens_seen": 121247190, + "step": 5711 + }, + { + "epoch": 0.6868273913304875, + "flos": 11572325874720.0, + "grad_norm": 3.1186776865677084, + "learning_rate": 9.434371105106223e-07, + "loss": 0.7107, + "num_input_tokens_seen": 121264510, + "step": 5712 + }, + { + "epoch": 0.6869476342211267, + "flos": 17917066147680.0, + "grad_norm": 3.664597982551191, + "learning_rate": 9.427757897380602e-07, + "loss": 0.6746, + "num_input_tokens_seen": 121283630, + "step": 5713 + }, + { + "epoch": 0.6870678771117658, + "flos": 13461952791720.0, + "grad_norm": 4.449475100988476, + "learning_rate": 9.421146293455695e-07, + "loss": 0.8315, + "num_input_tokens_seen": 121299090, + "step": 5714 + }, + { + "epoch": 0.6871881200024048, + "flos": 16242944307480.0, + "grad_norm": 3.6747501885242597, + "learning_rate": 9.414536294334489e-07, + "loss": 0.6747, + "num_input_tokens_seen": 121318830, + "step": 5715 + }, + { + "epoch": 0.687308362893044, + "flos": 16189226338080.0, + "grad_norm": 4.3163737714417225, + "learning_rate": 9.407927901019708e-07, + "loss": 0.6866, + "num_input_tokens_seen": 121337680, + "step": 5716 + }, + { + "epoch": 0.687428605783683, + "flos": 18344055486360.0, + "grad_norm": 3.879647206263523, + "learning_rate": 9.401321114513854e-07, + "loss": 0.761, + "num_input_tokens_seen": 121356295, + "step": 5717 + }, + { + "epoch": 0.6875488486743221, + "flos": 17404735030920.0, + "grad_norm": 3.822992738110677, + "learning_rate": 9.394715935819155e-07, + "loss": 0.7345, + "num_input_tokens_seen": 121376405, + "step": 5718 + }, + { + "epoch": 0.6876690915649613, + "flos": 18696910672800.0, + "grad_norm": 3.414056354578279, + "learning_rate": 9.388112365937608e-07, + "loss": 0.5952, + "num_input_tokens_seen": 121395590, + "step": 5719 + }, + { + "epoch": 0.6877893344556003, + "flos": 14191277003400.0, + "grad_norm": 4.37737710951281, + "learning_rate": 9.381510405870985e-07, + "loss": 0.7992, + "num_input_tokens_seen": 121414325, + "step": 5720 + }, + { + "epoch": 0.6879095773462394, + "flos": 13623803219040.0, + "grad_norm": 18.34808329620648, + "learning_rate": 9.374910056620791e-07, + "loss": 0.7472, + "num_input_tokens_seen": 121433110, + "step": 5721 + }, + { + "epoch": 0.6880298202368785, + "flos": 15267567648960.0, + "grad_norm": 3.8051364015048943, + "learning_rate": 9.368311319188293e-07, + "loss": 0.7933, + "num_input_tokens_seen": 121450645, + "step": 5722 + }, + { + "epoch": 0.6881500631275176, + "flos": 22128501553800.0, + "grad_norm": 3.7539865284730354, + "learning_rate": 9.361714194574515e-07, + "loss": 0.78, + "num_input_tokens_seen": 121472700, + "step": 5723 + }, + { + "epoch": 0.6882703060181566, + "flos": 48799362330600.0, + "grad_norm": 0.7347666325174387, + "learning_rate": 9.355118683780228e-07, + "loss": 0.59, + "num_input_tokens_seen": 121542490, + "step": 5724 + }, + { + "epoch": 0.6883905489087958, + "flos": 13291997414640.0, + "grad_norm": 4.6110116930027205, + "learning_rate": 9.348524787805987e-07, + "loss": 0.7738, + "num_input_tokens_seen": 121557400, + "step": 5725 + }, + { + "epoch": 0.6885107917994349, + "flos": 10214654758320.0, + "grad_norm": 15.968642023546394, + "learning_rate": 9.341932507652053e-07, + "loss": 0.8318, + "num_input_tokens_seen": 121571610, + "step": 5726 + }, + { + "epoch": 0.6886310346900739, + "flos": 21045657296520.0, + "grad_norm": 2.662109912360089, + "learning_rate": 9.335341844318489e-07, + "loss": 0.7582, + "num_input_tokens_seen": 121591470, + "step": 5727 + }, + { + "epoch": 0.6887512775807131, + "flos": 17970942416880.0, + "grad_norm": 4.2754727107812025, + "learning_rate": 9.328752798805091e-07, + "loss": 0.7182, + "num_input_tokens_seen": 121609660, + "step": 5728 + }, + { + "epoch": 0.6888715204713521, + "flos": 16400615620080.0, + "grad_norm": 3.708551736179426, + "learning_rate": 9.322165372111399e-07, + "loss": 0.7346, + "num_input_tokens_seen": 121627525, + "step": 5729 + }, + { + "epoch": 0.6889917633619912, + "flos": 16134843509520.0, + "grad_norm": 3.9386407319956036, + "learning_rate": 9.315579565236747e-07, + "loss": 0.7462, + "num_input_tokens_seen": 121646350, + "step": 5730 + }, + { + "epoch": 0.6891120062526304, + "flos": 17537304486600.0, + "grad_norm": 4.104143997878518, + "learning_rate": 9.308995379180162e-07, + "loss": 0.7298, + "num_input_tokens_seen": 121665625, + "step": 5731 + }, + { + "epoch": 0.6892322491432694, + "flos": 47268759800880.0, + "grad_norm": 0.8016438745905494, + "learning_rate": 9.302412814940488e-07, + "loss": 0.6293, + "num_input_tokens_seen": 121728120, + "step": 5732 + }, + { + "epoch": 0.6893524920339085, + "flos": 16836438087600.0, + "grad_norm": 7.173293960255245, + "learning_rate": 9.295831873516276e-07, + "loss": 0.6852, + "num_input_tokens_seen": 121747115, + "step": 5733 + }, + { + "epoch": 0.6894727349245476, + "flos": 15647645909640.0, + "grad_norm": 5.874509101000217, + "learning_rate": 9.289252555905873e-07, + "loss": 0.7509, + "num_input_tokens_seen": 121766915, + "step": 5734 + }, + { + "epoch": 0.6895929778151867, + "flos": 14514851218200.0, + "grad_norm": 5.517578816371638, + "learning_rate": 9.282674863107334e-07, + "loss": 0.735, + "num_input_tokens_seen": 121784450, + "step": 5735 + }, + { + "epoch": 0.6897132207058257, + "flos": 13265882439000.0, + "grad_norm": 4.553083343420028, + "learning_rate": 9.276098796118488e-07, + "loss": 0.7521, + "num_input_tokens_seen": 121800655, + "step": 5736 + }, + { + "epoch": 0.6898334635964649, + "flos": 23910882491760.0, + "grad_norm": 3.108437401577971, + "learning_rate": 9.269524355936938e-07, + "loss": 0.6542, + "num_input_tokens_seen": 121823555, + "step": 5737 + }, + { + "epoch": 0.689953706487104, + "flos": 16700544336120.0, + "grad_norm": 4.8816146968334, + "learning_rate": 9.262951543560002e-07, + "loss": 0.8362, + "num_input_tokens_seen": 121842500, + "step": 5738 + }, + { + "epoch": 0.690073949377743, + "flos": 13515797400960.0, + "grad_norm": 7.626899607686758, + "learning_rate": 9.256380359984795e-07, + "loss": 0.8419, + "num_input_tokens_seen": 121859330, + "step": 5739 + }, + { + "epoch": 0.6901941922683821, + "flos": 25609314689880.0, + "grad_norm": 7.1034262113736135, + "learning_rate": 9.249810806208139e-07, + "loss": 0.7305, + "num_input_tokens_seen": 121878315, + "step": 5740 + }, + { + "epoch": 0.6903144351590212, + "flos": 11842055480280.0, + "grad_norm": 5.066249259687022, + "learning_rate": 9.243242883226627e-07, + "loss": 0.79, + "num_input_tokens_seen": 121897130, + "step": 5741 + }, + { + "epoch": 0.6904346780496603, + "flos": 20560644233880.0, + "grad_norm": 3.8392822848586676, + "learning_rate": 9.236676592036628e-07, + "loss": 0.6892, + "num_input_tokens_seen": 121916525, + "step": 5742 + }, + { + "epoch": 0.6905549209402994, + "flos": 17296855852680.0, + "grad_norm": 6.889048147240662, + "learning_rate": 9.230111933634228e-07, + "loss": 0.7165, + "num_input_tokens_seen": 121937840, + "step": 5743 + }, + { + "epoch": 0.6906751638309385, + "flos": 16920133547760.0, + "grad_norm": 4.52585822971225, + "learning_rate": 9.223548909015288e-07, + "loss": 0.7972, + "num_input_tokens_seen": 121959250, + "step": 5744 + }, + { + "epoch": 0.6907954067215776, + "flos": 20020646803440.0, + "grad_norm": 3.796657979908942, + "learning_rate": 9.216987519175407e-07, + "loss": 0.6955, + "num_input_tokens_seen": 121979145, + "step": 5745 + }, + { + "epoch": 0.6909156496122166, + "flos": 15864765644400.0, + "grad_norm": 7.262713893032068, + "learning_rate": 9.210427765109942e-07, + "loss": 0.6753, + "num_input_tokens_seen": 121998540, + "step": 5746 + }, + { + "epoch": 0.6910358925028558, + "flos": 16509634556880.0, + "grad_norm": 7.357816689118048, + "learning_rate": 9.20386964781402e-07, + "loss": 0.7897, + "num_input_tokens_seen": 122016280, + "step": 5747 + }, + { + "epoch": 0.6911561353934949, + "flos": 16134653549760.0, + "grad_norm": 3.039095763193606, + "learning_rate": 9.197313168282472e-07, + "loss": 0.8181, + "num_input_tokens_seen": 122033445, + "step": 5748 + }, + { + "epoch": 0.6912763782841339, + "flos": 17727581066640.0, + "grad_norm": 4.333975573165916, + "learning_rate": 9.190758327509935e-07, + "loss": 0.6997, + "num_input_tokens_seen": 122051910, + "step": 5749 + }, + { + "epoch": 0.6913966211747731, + "flos": 38544662214240.0, + "grad_norm": 0.8862010772287016, + "learning_rate": 9.184205126490767e-07, + "loss": 0.6629, + "num_input_tokens_seen": 122100525, + "step": 5750 + }, + { + "epoch": 0.6915168640654121, + "flos": 49210205089680.0, + "grad_norm": 0.9473717069708295, + "learning_rate": 9.177653566219075e-07, + "loss": 0.6296, + "num_input_tokens_seen": 122154970, + "step": 5751 + }, + { + "epoch": 0.6916371069560512, + "flos": 13353851993760.0, + "grad_norm": 4.205702661326438, + "learning_rate": 9.171103647688744e-07, + "loss": 0.7296, + "num_input_tokens_seen": 122173430, + "step": 5752 + }, + { + "epoch": 0.6917573498466904, + "flos": 14352304271760.0, + "grad_norm": 3.813871168364159, + "learning_rate": 9.164555371893367e-07, + "loss": 0.6674, + "num_input_tokens_seen": 122193080, + "step": 5753 + }, + { + "epoch": 0.6918775927373294, + "flos": 10329815727360.0, + "grad_norm": 2.935614447409246, + "learning_rate": 9.158008739826333e-07, + "loss": 0.7413, + "num_input_tokens_seen": 122210400, + "step": 5754 + }, + { + "epoch": 0.6919978356279685, + "flos": 17563641081960.0, + "grad_norm": 4.266802167727652, + "learning_rate": 9.151463752480744e-07, + "loss": 0.8487, + "num_input_tokens_seen": 122228850, + "step": 5755 + }, + { + "epoch": 0.6921180785186076, + "flos": 17295684434160.0, + "grad_norm": 6.801950107767234, + "learning_rate": 9.144920410849493e-07, + "loss": 0.7886, + "num_input_tokens_seen": 122249805, + "step": 5756 + }, + { + "epoch": 0.6922383214092467, + "flos": 15460187066040.0, + "grad_norm": 3.7249166184892335, + "learning_rate": 9.138378715925176e-07, + "loss": 0.782, + "num_input_tokens_seen": 122268620, + "step": 5757 + }, + { + "epoch": 0.6923585642998857, + "flos": 15702915217080.0, + "grad_norm": 4.223555702872706, + "learning_rate": 9.131838668700167e-07, + "loss": 0.787, + "num_input_tokens_seen": 122288410, + "step": 5758 + }, + { + "epoch": 0.6924788071905249, + "flos": 15432362452560.0, + "grad_norm": 3.607414737151168, + "learning_rate": 9.125300270166598e-07, + "loss": 0.8511, + "num_input_tokens_seen": 122308735, + "step": 5759 + }, + { + "epoch": 0.692599050081164, + "flos": 19239915799440.0, + "grad_norm": 4.436195289186312, + "learning_rate": 9.118763521316324e-07, + "loss": 0.8528, + "num_input_tokens_seen": 122329030, + "step": 5760 + }, + { + "epoch": 0.692719292971803, + "flos": 15269403926640.0, + "grad_norm": 4.0965980930864205, + "learning_rate": 9.112228423140987e-07, + "loss": 0.7522, + "num_input_tokens_seen": 122347670, + "step": 5761 + }, + { + "epoch": 0.6928395358624422, + "flos": 18997155988440.0, + "grad_norm": 6.916235699689688, + "learning_rate": 9.105694976631932e-07, + "loss": 0.8614, + "num_input_tokens_seen": 122365300, + "step": 5762 + }, + { + "epoch": 0.6929597787530812, + "flos": 17268777959520.0, + "grad_norm": 7.036685724393744, + "learning_rate": 9.099163182780283e-07, + "loss": 0.7122, + "num_input_tokens_seen": 122383175, + "step": 5763 + }, + { + "epoch": 0.6930800216437203, + "flos": 13322988024120.0, + "grad_norm": 12.270392435353408, + "learning_rate": 9.092633042576916e-07, + "loss": 0.481, + "num_input_tokens_seen": 122400160, + "step": 5764 + }, + { + "epoch": 0.6932002645343595, + "flos": 21399842201280.0, + "grad_norm": 5.539138255388169, + "learning_rate": 9.086104557012446e-07, + "loss": 0.5513, + "num_input_tokens_seen": 122420450, + "step": 5765 + }, + { + "epoch": 0.6933205074249985, + "flos": 17458484660280.0, + "grad_norm": 6.26342917609398, + "learning_rate": 9.079577727077239e-07, + "loss": 0.653, + "num_input_tokens_seen": 122439000, + "step": 5766 + }, + { + "epoch": 0.6934407503156376, + "flos": 17697603575880.0, + "grad_norm": 3.5000780641841835, + "learning_rate": 9.073052553761404e-07, + "loss": 0.6982, + "num_input_tokens_seen": 122458085, + "step": 5767 + }, + { + "epoch": 0.6935609932062767, + "flos": 15081723463320.0, + "grad_norm": 3.9229405388061007, + "learning_rate": 9.066529038054805e-07, + "loss": 0.7595, + "num_input_tokens_seen": 122477870, + "step": 5768 + }, + { + "epoch": 0.6936812360969158, + "flos": 13321689965760.0, + "grad_norm": 3.357285386396344, + "learning_rate": 9.060007180947071e-07, + "loss": 0.7276, + "num_input_tokens_seen": 122495645, + "step": 5769 + }, + { + "epoch": 0.6938014789875548, + "flos": 22989477082320.0, + "grad_norm": 4.293560511421072, + "learning_rate": 9.053486983427534e-07, + "loss": 0.7155, + "num_input_tokens_seen": 122516615, + "step": 5770 + }, + { + "epoch": 0.6939217218781939, + "flos": 12785143470960.0, + "grad_norm": 3.6758081027761853, + "learning_rate": 9.046968446485326e-07, + "loss": 0.6886, + "num_input_tokens_seen": 122534740, + "step": 5771 + }, + { + "epoch": 0.6940419647688331, + "flos": 13542735535560.0, + "grad_norm": 5.301083673495762, + "learning_rate": 9.040451571109295e-07, + "loss": 0.6899, + "num_input_tokens_seen": 122550080, + "step": 5772 + }, + { + "epoch": 0.6941622076594721, + "flos": 49347491879400.0, + "grad_norm": 0.9241548246846659, + "learning_rate": 9.033936358288042e-07, + "loss": 0.6406, + "num_input_tokens_seen": 122603535, + "step": 5773 + }, + { + "epoch": 0.6942824505501112, + "flos": 19482643950480.0, + "grad_norm": 4.48329771706049, + "learning_rate": 9.027422809009937e-07, + "loss": 0.8102, + "num_input_tokens_seen": 122623200, + "step": 5774 + }, + { + "epoch": 0.6944026934407503, + "flos": 15537867133800.0, + "grad_norm": 3.4698393889533037, + "learning_rate": 9.020910924263054e-07, + "loss": 0.8122, + "num_input_tokens_seen": 122641445, + "step": 5775 + }, + { + "epoch": 0.6945229363313894, + "flos": 52122404682840.0, + "grad_norm": 0.8627509304960191, + "learning_rate": 9.014400705035261e-07, + "loss": 0.6157, + "num_input_tokens_seen": 122698070, + "step": 5776 + }, + { + "epoch": 0.6946431792220285, + "flos": 13839593235480.0, + "grad_norm": 14.791315180224583, + "learning_rate": 9.00789215231414e-07, + "loss": 0.7645, + "num_input_tokens_seen": 122716185, + "step": 5777 + }, + { + "epoch": 0.6947634221126676, + "flos": 14865173607840.0, + "grad_norm": 3.030302705909249, + "learning_rate": 9.001385267087056e-07, + "loss": 0.8077, + "num_input_tokens_seen": 122735050, + "step": 5778 + }, + { + "epoch": 0.6948836650033067, + "flos": 15971346764280.0, + "grad_norm": 3.199900720963611, + "learning_rate": 8.994880050341072e-07, + "loss": 0.6864, + "num_input_tokens_seen": 122754875, + "step": 5779 + }, + { + "epoch": 0.6950039078939457, + "flos": 17321007910800.0, + "grad_norm": 24.043313677968506, + "learning_rate": 8.988376503063026e-07, + "loss": 0.7647, + "num_input_tokens_seen": 122775855, + "step": 5780 + }, + { + "epoch": 0.6951241507845849, + "flos": 15940134535080.0, + "grad_norm": 4.517753759285804, + "learning_rate": 8.981874626239521e-07, + "loss": 0.8024, + "num_input_tokens_seen": 122794150, + "step": 5781 + }, + { + "epoch": 0.695244393675224, + "flos": 10653326622240.0, + "grad_norm": 4.918866306660291, + "learning_rate": 8.975374420856872e-07, + "loss": 0.8643, + "num_input_tokens_seen": 122810765, + "step": 5782 + }, + { + "epoch": 0.695364636565863, + "flos": 12300447007920.0, + "grad_norm": 3.99595485233314, + "learning_rate": 8.968875887901157e-07, + "loss": 0.7084, + "num_input_tokens_seen": 122827865, + "step": 5783 + }, + { + "epoch": 0.6954848794565022, + "flos": 14136830854920.0, + "grad_norm": 3.641334955630179, + "learning_rate": 8.9623790283582e-07, + "loss": 0.5933, + "num_input_tokens_seen": 122845465, + "step": 5784 + }, + { + "epoch": 0.6956051223471412, + "flos": 13867291209120.0, + "grad_norm": 9.655212149045012, + "learning_rate": 8.955883843213561e-07, + "loss": 0.745, + "num_input_tokens_seen": 122864200, + "step": 5785 + }, + { + "epoch": 0.6957253652377803, + "flos": 11733606422760.0, + "grad_norm": 3.803441025604828, + "learning_rate": 8.949390333452569e-07, + "loss": 0.8683, + "num_input_tokens_seen": 122881865, + "step": 5786 + }, + { + "epoch": 0.6958456081284194, + "flos": 21562832387160.0, + "grad_norm": 4.467971627264002, + "learning_rate": 8.942898500060279e-07, + "loss": 0.6588, + "num_input_tokens_seen": 122901300, + "step": 5787 + }, + { + "epoch": 0.6959658510190585, + "flos": 18429333944520.0, + "grad_norm": 5.000434665859977, + "learning_rate": 8.936408344021493e-07, + "loss": 0.6831, + "num_input_tokens_seen": 122917935, + "step": 5788 + }, + { + "epoch": 0.6960860939096976, + "flos": 31498639471440.0, + "grad_norm": 5.842519375561248, + "learning_rate": 8.929919866320765e-07, + "loss": 0.6923, + "num_input_tokens_seen": 122938470, + "step": 5789 + }, + { + "epoch": 0.6962063368003367, + "flos": 12813601283640.0, + "grad_norm": 4.517371239318101, + "learning_rate": 8.923433067942385e-07, + "loss": 0.7979, + "num_input_tokens_seen": 122956755, + "step": 5790 + }, + { + "epoch": 0.6963265796909758, + "flos": 15540811510080.0, + "grad_norm": 3.19475343472184, + "learning_rate": 8.916947949870417e-07, + "loss": 0.6681, + "num_input_tokens_seen": 122976140, + "step": 5791 + }, + { + "epoch": 0.6964468225816148, + "flos": 50755778289120.0, + "grad_norm": 0.8304820076717667, + "learning_rate": 8.910464513088615e-07, + "loss": 0.6245, + "num_input_tokens_seen": 123039900, + "step": 5792 + }, + { + "epoch": 0.696567065472254, + "flos": 13837788617760.0, + "grad_norm": 4.456944163054212, + "learning_rate": 8.903982758580542e-07, + "loss": 0.7724, + "num_input_tokens_seen": 123058560, + "step": 5793 + }, + { + "epoch": 0.696687308362893, + "flos": 16728590569320.0, + "grad_norm": 3.5887802923153753, + "learning_rate": 8.897502687329457e-07, + "loss": 0.7964, + "num_input_tokens_seen": 123078080, + "step": 5794 + }, + { + "epoch": 0.6968075512535321, + "flos": 18294769911360.0, + "grad_norm": 5.397649196808425, + "learning_rate": 8.891024300318382e-07, + "loss": 0.7872, + "num_input_tokens_seen": 123096370, + "step": 5795 + }, + { + "epoch": 0.6969277941441713, + "flos": 15376776545520.0, + "grad_norm": 3.5005199863542735, + "learning_rate": 8.884547598530103e-07, + "loss": 0.7444, + "num_input_tokens_seen": 123116660, + "step": 5796 + }, + { + "epoch": 0.6970480370348103, + "flos": 15783286381440.0, + "grad_norm": 6.4760167682282175, + "learning_rate": 8.8780725829471e-07, + "loss": 0.738, + "num_input_tokens_seen": 123134285, + "step": 5797 + }, + { + "epoch": 0.6971682799254494, + "flos": 16404858054720.0, + "grad_norm": 7.9277110375518784, + "learning_rate": 8.87159925455165e-07, + "loss": 0.7457, + "num_input_tokens_seen": 123153835, + "step": 5798 + }, + { + "epoch": 0.6972885228160886, + "flos": 14618076382320.0, + "grad_norm": 9.466261533566195, + "learning_rate": 8.865127614325738e-07, + "loss": 0.7209, + "num_input_tokens_seen": 123171985, + "step": 5799 + }, + { + "epoch": 0.6974087657067276, + "flos": 27827359795560.0, + "grad_norm": 3.128127554817185, + "learning_rate": 8.85865766325113e-07, + "loss": 0.6494, + "num_input_tokens_seen": 123195635, + "step": 5800 + }, + { + "epoch": 0.6975290085973667, + "flos": 21636048400560.0, + "grad_norm": 6.918387815511668, + "learning_rate": 8.852189402309287e-07, + "loss": 0.6957, + "num_input_tokens_seen": 123214540, + "step": 5801 + }, + { + "epoch": 0.6976492514880057, + "flos": 9356686926000.0, + "grad_norm": 4.380099349266891, + "learning_rate": 8.845722832481441e-07, + "loss": 0.7011, + "num_input_tokens_seen": 123229690, + "step": 5802 + }, + { + "epoch": 0.6977694943786449, + "flos": 17835966804240.0, + "grad_norm": 4.338775420865888, + "learning_rate": 8.83925795474858e-07, + "loss": 0.7557, + "num_input_tokens_seen": 123249535, + "step": 5803 + }, + { + "epoch": 0.6978897372692839, + "flos": 21939997931520.0, + "grad_norm": 6.509119825265026, + "learning_rate": 8.832794770091414e-07, + "loss": 0.5922, + "num_input_tokens_seen": 123270090, + "step": 5804 + }, + { + "epoch": 0.698009980159923, + "flos": 15918072034320.0, + "grad_norm": 4.583464441447184, + "learning_rate": 8.826333279490401e-07, + "loss": 0.8076, + "num_input_tokens_seen": 123290445, + "step": 5805 + }, + { + "epoch": 0.6981302230505622, + "flos": 14507664407280.0, + "grad_norm": 4.054785108034563, + "learning_rate": 8.819873483925748e-07, + "loss": 0.6627, + "num_input_tokens_seen": 123307285, + "step": 5806 + }, + { + "epoch": 0.6982504659412012, + "flos": 16241424629400.0, + "grad_norm": 6.918744552596376, + "learning_rate": 8.81341538437739e-07, + "loss": 0.7317, + "num_input_tokens_seen": 123325295, + "step": 5807 + }, + { + "epoch": 0.6983707088318403, + "flos": 26150736818520.0, + "grad_norm": 3.2866252649213057, + "learning_rate": 8.80695898182503e-07, + "loss": 0.6774, + "num_input_tokens_seen": 123345995, + "step": 5808 + }, + { + "epoch": 0.6984909517224794, + "flos": 48248948773320.0, + "grad_norm": 0.897748002558127, + "learning_rate": 8.800504277248093e-07, + "loss": 0.6779, + "num_input_tokens_seen": 123410465, + "step": 5809 + }, + { + "epoch": 0.6986111946131185, + "flos": 13538461440960.0, + "grad_norm": 3.5083738697697266, + "learning_rate": 8.794051271625753e-07, + "loss": 0.7311, + "num_input_tokens_seen": 123427820, + "step": 5810 + }, + { + "epoch": 0.6987314375037575, + "flos": 16863661161840.0, + "grad_norm": 4.844427442861076, + "learning_rate": 8.787599965936925e-07, + "loss": 0.8211, + "num_input_tokens_seen": 123448470, + "step": 5811 + }, + { + "epoch": 0.6988516803943967, + "flos": 28232096673720.0, + "grad_norm": 4.580696919819105, + "learning_rate": 8.781150361160261e-07, + "loss": 0.7158, + "num_input_tokens_seen": 123470100, + "step": 5812 + }, + { + "epoch": 0.6989719232850358, + "flos": 17646481723200.0, + "grad_norm": 89.98716813906024, + "learning_rate": 8.774702458274181e-07, + "loss": 0.7209, + "num_input_tokens_seen": 123490225, + "step": 5813 + }, + { + "epoch": 0.6990921661756748, + "flos": 10815651948960.0, + "grad_norm": 14.936701158686487, + "learning_rate": 8.768256258256799e-07, + "loss": 0.6758, + "num_input_tokens_seen": 123506570, + "step": 5814 + }, + { + "epoch": 0.699212409066314, + "flos": 14757895968840.0, + "grad_norm": 4.256833809180186, + "learning_rate": 8.76181176208602e-07, + "loss": 0.7342, + "num_input_tokens_seen": 123524390, + "step": 5815 + }, + { + "epoch": 0.699332651956953, + "flos": 14191182023520.0, + "grad_norm": 4.119148033717852, + "learning_rate": 8.755368970739461e-07, + "loss": 0.7162, + "num_input_tokens_seen": 123543470, + "step": 5816 + }, + { + "epoch": 0.6994528948475921, + "flos": 11760607877280.0, + "grad_norm": 4.559393883627469, + "learning_rate": 8.748927885194479e-07, + "loss": 0.5915, + "num_input_tokens_seen": 123561495, + "step": 5817 + }, + { + "epoch": 0.6995731377382313, + "flos": 47886184019400.0, + "grad_norm": 0.7387309647381376, + "learning_rate": 8.742488506428209e-07, + "loss": 0.5757, + "num_input_tokens_seen": 123620305, + "step": 5818 + }, + { + "epoch": 0.6996933806288703, + "flos": 18240925302120.0, + "grad_norm": 4.503640503665327, + "learning_rate": 8.736050835417466e-07, + "loss": 0.7728, + "num_input_tokens_seen": 123640065, + "step": 5819 + }, + { + "epoch": 0.6998136235195094, + "flos": 15190204180800.0, + "grad_norm": 9.746409219593708, + "learning_rate": 8.729614873138862e-07, + "loss": 0.5983, + "num_input_tokens_seen": 123657420, + "step": 5820 + }, + { + "epoch": 0.6999338664101485, + "flos": 17377638596520.0, + "grad_norm": 4.411726990332861, + "learning_rate": 8.723180620568716e-07, + "loss": 0.7678, + "num_input_tokens_seen": 123676395, + "step": 5821 + }, + { + "epoch": 0.7000541093007876, + "flos": 14514218019000.0, + "grad_norm": 5.303839052465488, + "learning_rate": 8.716748078683116e-07, + "loss": 0.8476, + "num_input_tokens_seen": 123692890, + "step": 5822 + }, + { + "epoch": 0.7001743521914267, + "flos": 21777545964960.0, + "grad_norm": 5.787483442707147, + "learning_rate": 8.710317248457855e-07, + "loss": 0.6698, + "num_input_tokens_seen": 123712225, + "step": 5823 + }, + { + "epoch": 0.7002945950820658, + "flos": 20157300393960.0, + "grad_norm": 5.318347898674728, + "learning_rate": 8.703888130868482e-07, + "loss": 0.7004, + "num_input_tokens_seen": 123733795, + "step": 5824 + }, + { + "epoch": 0.7004148379727049, + "flos": 16211162199000.0, + "grad_norm": 3.954271474453513, + "learning_rate": 8.697460726890307e-07, + "loss": 0.8023, + "num_input_tokens_seen": 123750660, + "step": 5825 + }, + { + "epoch": 0.7005350808633439, + "flos": 14187541128120.0, + "grad_norm": 4.942776335505384, + "learning_rate": 8.691035037498354e-07, + "loss": 0.8827, + "num_input_tokens_seen": 123766370, + "step": 5826 + }, + { + "epoch": 0.7006553237539831, + "flos": 17187425336400.0, + "grad_norm": 3.8513406376262345, + "learning_rate": 8.684611063667391e-07, + "loss": 0.7171, + "num_input_tokens_seen": 123786555, + "step": 5827 + }, + { + "epoch": 0.7007755666446221, + "flos": 22912873453200.0, + "grad_norm": 4.547670904285678, + "learning_rate": 8.678188806371935e-07, + "loss": 0.7453, + "num_input_tokens_seen": 123808310, + "step": 5828 + }, + { + "epoch": 0.7008958095352612, + "flos": 13245244636440.0, + "grad_norm": 3.170716473451559, + "learning_rate": 8.671768266586228e-07, + "loss": 0.842, + "num_input_tokens_seen": 123826155, + "step": 5829 + }, + { + "epoch": 0.7010160524259004, + "flos": 20129444120520.0, + "grad_norm": 4.294519062937502, + "learning_rate": 8.665349445284275e-07, + "loss": 0.7702, + "num_input_tokens_seen": 123845615, + "step": 5830 + }, + { + "epoch": 0.7011362953165394, + "flos": 17457883121040.0, + "grad_norm": 10.393218441031069, + "learning_rate": 8.658932343439799e-07, + "loss": 0.7972, + "num_input_tokens_seen": 123865120, + "step": 5831 + }, + { + "epoch": 0.7012565382071785, + "flos": 18183629757240.0, + "grad_norm": 3.8962787774705725, + "learning_rate": 8.65251696202627e-07, + "loss": 0.7488, + "num_input_tokens_seen": 123881220, + "step": 5832 + }, + { + "epoch": 0.7013767810978175, + "flos": 15648690688320.0, + "grad_norm": 6.660492167573403, + "learning_rate": 8.646103302016896e-07, + "loss": 0.8596, + "num_input_tokens_seen": 123899910, + "step": 5833 + }, + { + "epoch": 0.7014970239884567, + "flos": 12162242079360.0, + "grad_norm": 3.989405190629853, + "learning_rate": 8.639691364384614e-07, + "loss": 0.8664, + "num_input_tokens_seen": 123917255, + "step": 5834 + }, + { + "epoch": 0.7016172668790958, + "flos": 9114560314200.0, + "grad_norm": 4.405733335247774, + "learning_rate": 8.633281150102136e-07, + "loss": 0.7064, + "num_input_tokens_seen": 123933825, + "step": 5835 + }, + { + "epoch": 0.7017375097697348, + "flos": 12728734404960.0, + "grad_norm": 4.3356614583442825, + "learning_rate": 8.626872660141855e-07, + "loss": 0.6624, + "num_input_tokens_seen": 123951455, + "step": 5836 + }, + { + "epoch": 0.701857752660374, + "flos": 13513739503560.0, + "grad_norm": 4.992666970690576, + "learning_rate": 8.620465895475957e-07, + "loss": 0.7353, + "num_input_tokens_seen": 123969395, + "step": 5837 + }, + { + "epoch": 0.701977995551013, + "flos": 17890033033200.0, + "grad_norm": 4.465001594910545, + "learning_rate": 8.614060857076333e-07, + "loss": 0.7415, + "num_input_tokens_seen": 123989785, + "step": 5838 + }, + { + "epoch": 0.7020982384416521, + "flos": 16836723027240.0, + "grad_norm": 5.699099272115528, + "learning_rate": 8.60765754591462e-07, + "loss": 0.7338, + "num_input_tokens_seen": 124009200, + "step": 5839 + }, + { + "epoch": 0.7022184813322913, + "flos": 14946874490520.0, + "grad_norm": 2.8999471922703255, + "learning_rate": 8.601255962962211e-07, + "loss": 0.7216, + "num_input_tokens_seen": 124027930, + "step": 5840 + }, + { + "epoch": 0.7023387242229303, + "flos": 14459233651200.0, + "grad_norm": 6.311719487347555, + "learning_rate": 8.594856109190194e-07, + "loss": 0.7141, + "num_input_tokens_seen": 124044680, + "step": 5841 + }, + { + "epoch": 0.7024589671135694, + "flos": 24427804302720.0, + "grad_norm": 8.512044563377579, + "learning_rate": 8.588457985569446e-07, + "loss": 0.6889, + "num_input_tokens_seen": 124067310, + "step": 5842 + }, + { + "epoch": 0.7025792100042085, + "flos": 13947694033440.0, + "grad_norm": 4.556120710801732, + "learning_rate": 8.582061593070542e-07, + "loss": 0.6989, + "num_input_tokens_seen": 124087760, + "step": 5843 + }, + { + "epoch": 0.7026994528948476, + "flos": 13839308295840.0, + "grad_norm": 7.311039966338872, + "learning_rate": 8.57566693266383e-07, + "loss": 0.7434, + "num_input_tokens_seen": 124105170, + "step": 5844 + }, + { + "epoch": 0.7028196957854866, + "flos": 14271584847840.0, + "grad_norm": 4.845635816113541, + "learning_rate": 8.569274005319354e-07, + "loss": 0.6798, + "num_input_tokens_seen": 124123290, + "step": 5845 + }, + { + "epoch": 0.7029399386761258, + "flos": 15239394775920.0, + "grad_norm": 3.1101177612774844, + "learning_rate": 8.562882812006913e-07, + "loss": 0.7862, + "num_input_tokens_seen": 124140500, + "step": 5846 + }, + { + "epoch": 0.7030601815667649, + "flos": 16135445048760.0, + "grad_norm": 3.7110387527305044, + "learning_rate": 8.556493353696066e-07, + "loss": 0.7557, + "num_input_tokens_seen": 124159220, + "step": 5847 + }, + { + "epoch": 0.7031804244574039, + "flos": 19942048596840.0, + "grad_norm": 6.709902309383113, + "learning_rate": 8.550105631356077e-07, + "loss": 0.6579, + "num_input_tokens_seen": 124178665, + "step": 5848 + }, + { + "epoch": 0.7033006673480431, + "flos": 16375355463360.0, + "grad_norm": 19.687100845408033, + "learning_rate": 8.543719645955961e-07, + "loss": 0.75, + "num_input_tokens_seen": 124196715, + "step": 5849 + }, + { + "epoch": 0.7034209102386821, + "flos": 18104398351440.0, + "grad_norm": 23.640001489216456, + "learning_rate": 8.537335398464467e-07, + "loss": 0.7275, + "num_input_tokens_seen": 124216755, + "step": 5850 + }, + { + "epoch": 0.7035411531293212, + "flos": 16504980542760.0, + "grad_norm": 7.783654266645955, + "learning_rate": 8.53095288985007e-07, + "loss": 0.8377, + "num_input_tokens_seen": 124230210, + "step": 5851 + }, + { + "epoch": 0.7036613960199604, + "flos": 16594216495920.0, + "grad_norm": 5.596989210572645, + "learning_rate": 8.524572121081009e-07, + "loss": 0.8054, + "num_input_tokens_seen": 124250030, + "step": 5852 + }, + { + "epoch": 0.7037816389105994, + "flos": 16458860963760.0, + "grad_norm": 3.6128816290838377, + "learning_rate": 8.518193093125232e-07, + "loss": 0.6054, + "num_input_tokens_seen": 124268805, + "step": 5853 + }, + { + "epoch": 0.7039018818012385, + "flos": 19859461235280.0, + "grad_norm": 3.7429798537702523, + "learning_rate": 8.511815806950436e-07, + "loss": 0.7909, + "num_input_tokens_seen": 124289555, + "step": 5854 + }, + { + "epoch": 0.7040221246918776, + "flos": 12949336735320.0, + "grad_norm": 5.710162609878111, + "learning_rate": 8.505440263524044e-07, + "loss": 0.7588, + "num_input_tokens_seen": 124308120, + "step": 5855 + }, + { + "epoch": 0.7041423675825167, + "flos": 11859970526280.0, + "grad_norm": 5.932755994389526, + "learning_rate": 8.49906646381322e-07, + "loss": 0.849, + "num_input_tokens_seen": 124320675, + "step": 5856 + }, + { + "epoch": 0.7042626104731557, + "flos": 18672315375240.0, + "grad_norm": 7.891410078518525, + "learning_rate": 8.492694408784884e-07, + "loss": 0.7048, + "num_input_tokens_seen": 124340650, + "step": 5857 + }, + { + "epoch": 0.7043828533637949, + "flos": 13028884740720.0, + "grad_norm": 4.494660918926259, + "learning_rate": 8.486324099405642e-07, + "loss": 0.5885, + "num_input_tokens_seen": 124357215, + "step": 5858 + }, + { + "epoch": 0.704503096254434, + "flos": 21640449135000.0, + "grad_norm": 4.490606464302603, + "learning_rate": 8.479955536641887e-07, + "loss": 0.7438, + "num_input_tokens_seen": 124378430, + "step": 5859 + }, + { + "epoch": 0.704623339145073, + "flos": 22696640197320.0, + "grad_norm": 3.425383419656152, + "learning_rate": 8.473588721459716e-07, + "loss": 0.6418, + "num_input_tokens_seen": 124398060, + "step": 5860 + }, + { + "epoch": 0.7047435820357122, + "flos": 17511632750400.0, + "grad_norm": 5.137279036102224, + "learning_rate": 8.467223654824967e-07, + "loss": 0.6769, + "num_input_tokens_seen": 124417235, + "step": 5861 + }, + { + "epoch": 0.7048638249263512, + "flos": 34221037383960.0, + "grad_norm": 5.360437048843523, + "learning_rate": 8.460860337703233e-07, + "loss": 0.6106, + "num_input_tokens_seen": 124437560, + "step": 5862 + }, + { + "epoch": 0.7049840678169903, + "flos": 15864607344600.0, + "grad_norm": 5.34278173765006, + "learning_rate": 8.454498771059797e-07, + "loss": 0.6867, + "num_input_tokens_seen": 124456655, + "step": 5863 + }, + { + "epoch": 0.7051043107076294, + "flos": 13434064858320.0, + "grad_norm": 8.785266107166617, + "learning_rate": 8.448138955859725e-07, + "loss": 0.8092, + "num_input_tokens_seen": 124472960, + "step": 5864 + }, + { + "epoch": 0.7052245535982685, + "flos": 14110715879280.0, + "grad_norm": 9.441574732800593, + "learning_rate": 8.44178089306778e-07, + "loss": 0.8909, + "num_input_tokens_seen": 124490615, + "step": 5865 + }, + { + "epoch": 0.7053447964889076, + "flos": 13920375979320.0, + "grad_norm": 3.803299731062399, + "learning_rate": 8.4354245836485e-07, + "loss": 0.7636, + "num_input_tokens_seen": 124508780, + "step": 5866 + }, + { + "epoch": 0.7054650393795466, + "flos": 20075346231600.0, + "grad_norm": 3.422102411107735, + "learning_rate": 8.429070028566108e-07, + "loss": 0.7103, + "num_input_tokens_seen": 124529810, + "step": 5867 + }, + { + "epoch": 0.7055852822701858, + "flos": 11729174028360.0, + "grad_norm": 3.8609053940404907, + "learning_rate": 8.422717228784586e-07, + "loss": 0.7336, + "num_input_tokens_seen": 124546405, + "step": 5868 + }, + { + "epoch": 0.7057055251608249, + "flos": 8466240466080.0, + "grad_norm": 6.121124286977115, + "learning_rate": 8.416366185267663e-07, + "loss": 0.682, + "num_input_tokens_seen": 124563625, + "step": 5869 + }, + { + "epoch": 0.7058257680514639, + "flos": 16621312930320.0, + "grad_norm": 3.09917952231681, + "learning_rate": 8.410016898978778e-07, + "loss": 0.761, + "num_input_tokens_seen": 124580820, + "step": 5870 + }, + { + "epoch": 0.7059460109421031, + "flos": 12786124929720.0, + "grad_norm": 4.161754433701088, + "learning_rate": 8.403669370881115e-07, + "loss": 0.7873, + "num_input_tokens_seen": 124599275, + "step": 5871 + }, + { + "epoch": 0.7060662538327421, + "flos": 17237027511000.0, + "grad_norm": 4.356347786505033, + "learning_rate": 8.397323601937587e-07, + "loss": 0.7638, + "num_input_tokens_seen": 124618895, + "step": 5872 + }, + { + "epoch": 0.7061864967233812, + "flos": 22207732959600.0, + "grad_norm": 4.3101950857526266, + "learning_rate": 8.390979593110838e-07, + "loss": 0.7633, + "num_input_tokens_seen": 124640745, + "step": 5873 + }, + { + "epoch": 0.7063067396140204, + "flos": 15133225235520.0, + "grad_norm": 3.371111431053357, + "learning_rate": 8.384637345363262e-07, + "loss": 0.7882, + "num_input_tokens_seen": 124659655, + "step": 5874 + }, + { + "epoch": 0.7064269825046594, + "flos": 23691704859600.0, + "grad_norm": 3.524990238690219, + "learning_rate": 8.378296859656964e-07, + "loss": 0.7559, + "num_input_tokens_seen": 124680530, + "step": 5875 + }, + { + "epoch": 0.7065472253952985, + "flos": 22183264301880.0, + "grad_norm": 6.007378174320048, + "learning_rate": 8.371958136953792e-07, + "loss": 0.6644, + "num_input_tokens_seen": 124700280, + "step": 5876 + }, + { + "epoch": 0.7066674682859376, + "flos": 11814642446280.0, + "grad_norm": 54.99432676706298, + "learning_rate": 8.365621178215326e-07, + "loss": 0.6388, + "num_input_tokens_seen": 124716470, + "step": 5877 + }, + { + "epoch": 0.7067877111765767, + "flos": 10788745474320.0, + "grad_norm": 5.768803245158189, + "learning_rate": 8.359285984402871e-07, + "loss": 0.7338, + "num_input_tokens_seen": 124733455, + "step": 5878 + }, + { + "epoch": 0.7069079540672157, + "flos": 18639583467960.0, + "grad_norm": 4.250098956582902, + "learning_rate": 8.352952556477489e-07, + "loss": 0.7284, + "num_input_tokens_seen": 124751085, + "step": 5879 + }, + { + "epoch": 0.7070281969578549, + "flos": 18024470426520.0, + "grad_norm": 3.669154941234846, + "learning_rate": 8.34662089539993e-07, + "loss": 0.7548, + "num_input_tokens_seen": 124770315, + "step": 5880 + }, + { + "epoch": 0.707148439848494, + "flos": 19591061348040.0, + "grad_norm": 5.487490385911727, + "learning_rate": 8.340291002130722e-07, + "loss": 0.7612, + "num_input_tokens_seen": 124789225, + "step": 5881 + }, + { + "epoch": 0.707268682739133, + "flos": 10977660676080.0, + "grad_norm": 4.866926109679505, + "learning_rate": 8.3339628776301e-07, + "loss": 0.7784, + "num_input_tokens_seen": 124807085, + "step": 5882 + }, + { + "epoch": 0.7073889256297722, + "flos": 25207142268480.0, + "grad_norm": 5.350533981057058, + "learning_rate": 8.327636522858033e-07, + "loss": 0.5489, + "num_input_tokens_seen": 124826410, + "step": 5883 + }, + { + "epoch": 0.7075091685204112, + "flos": 14676606665640.0, + "grad_norm": 5.36401113279412, + "learning_rate": 8.321311938774225e-07, + "loss": 0.749, + "num_input_tokens_seen": 124845220, + "step": 5884 + }, + { + "epoch": 0.7076294114110503, + "flos": 15189602641560.0, + "grad_norm": 3.5051089192879266, + "learning_rate": 8.314989126338104e-07, + "loss": 0.7744, + "num_input_tokens_seen": 124864950, + "step": 5885 + }, + { + "epoch": 0.7077496543016895, + "flos": 13110427323600.0, + "grad_norm": 4.79384455183501, + "learning_rate": 8.308668086508847e-07, + "loss": 0.836, + "num_input_tokens_seen": 124882750, + "step": 5886 + }, + { + "epoch": 0.7078698971923285, + "flos": 33469998931080.0, + "grad_norm": 5.372036849019798, + "learning_rate": 8.302348820245342e-07, + "loss": 0.7204, + "num_input_tokens_seen": 124905035, + "step": 5887 + }, + { + "epoch": 0.7079901400829676, + "flos": 19753355014800.0, + "grad_norm": 9.909929524776407, + "learning_rate": 8.296031328506232e-07, + "loss": 0.6804, + "num_input_tokens_seen": 124924505, + "step": 5888 + }, + { + "epoch": 0.7081103829736067, + "flos": 17887595216280.0, + "grad_norm": 3.9103234376329863, + "learning_rate": 8.289715612249857e-07, + "loss": 0.751, + "num_input_tokens_seen": 124944840, + "step": 5889 + }, + { + "epoch": 0.7082306258642458, + "flos": 13535263785000.0, + "grad_norm": 5.30527822270083, + "learning_rate": 8.283401672434305e-07, + "loss": 0.7531, + "num_input_tokens_seen": 124959785, + "step": 5890 + }, + { + "epoch": 0.7083508687548848, + "flos": 17188153515480.0, + "grad_norm": 5.570866839418198, + "learning_rate": 8.277089510017412e-07, + "loss": 0.6918, + "num_input_tokens_seen": 124980310, + "step": 5891 + }, + { + "epoch": 0.708471111645524, + "flos": 16405111334400.0, + "grad_norm": 2.9006620440658506, + "learning_rate": 8.270779125956719e-07, + "loss": 0.8001, + "num_input_tokens_seen": 125000410, + "step": 5892 + }, + { + "epoch": 0.7085913545361631, + "flos": 15297450159840.0, + "grad_norm": 7.69619804747684, + "learning_rate": 8.264470521209505e-07, + "loss": 0.7889, + "num_input_tokens_seen": 125018495, + "step": 5893 + }, + { + "epoch": 0.7087115974268021, + "flos": 10923309507480.0, + "grad_norm": 4.240323663629802, + "learning_rate": 8.258163696732785e-07, + "loss": 0.757, + "num_input_tokens_seen": 125035805, + "step": 5894 + }, + { + "epoch": 0.7088318403174413, + "flos": 15752865651240.0, + "grad_norm": 4.5982573318053825, + "learning_rate": 8.251858653483288e-07, + "loss": 0.7491, + "num_input_tokens_seen": 125053690, + "step": 5895 + }, + { + "epoch": 0.7089520832080803, + "flos": 11295694397880.0, + "grad_norm": 4.420392869776276, + "learning_rate": 8.245555392417501e-07, + "loss": 0.841, + "num_input_tokens_seen": 125068065, + "step": 5896 + }, + { + "epoch": 0.7090723260987194, + "flos": 14919936355920.0, + "grad_norm": 5.95627332524699, + "learning_rate": 8.239253914491613e-07, + "loss": 0.7767, + "num_input_tokens_seen": 125086110, + "step": 5897 + }, + { + "epoch": 0.7091925689893585, + "flos": 18809728804800.0, + "grad_norm": 3.150278618010437, + "learning_rate": 8.232954220661556e-07, + "loss": 0.7332, + "num_input_tokens_seen": 125108565, + "step": 5898 + }, + { + "epoch": 0.7093128118799976, + "flos": 17727644386560.0, + "grad_norm": 6.888928539749941, + "learning_rate": 8.226656311882989e-07, + "loss": 0.6501, + "num_input_tokens_seen": 125127595, + "step": 5899 + }, + { + "epoch": 0.7094330547706367, + "flos": 12131821349160.0, + "grad_norm": 17.448808182292048, + "learning_rate": 8.22036018911129e-07, + "loss": 0.7459, + "num_input_tokens_seen": 125145345, + "step": 5900 + }, + { + "epoch": 0.7095532976612757, + "flos": 11863073202360.0, + "grad_norm": 8.459954568366673, + "learning_rate": 8.214065853301599e-07, + "loss": 0.795, + "num_input_tokens_seen": 125160595, + "step": 5901 + }, + { + "epoch": 0.7096735405519149, + "flos": 52156814568000.0, + "grad_norm": 0.76812840894745, + "learning_rate": 8.207773305408734e-07, + "loss": 0.5968, + "num_input_tokens_seen": 125227535, + "step": 5902 + }, + { + "epoch": 0.709793783442554, + "flos": 17294418035760.0, + "grad_norm": 9.068318781206257, + "learning_rate": 8.201482546387288e-07, + "loss": 0.7852, + "num_input_tokens_seen": 125246730, + "step": 5903 + }, + { + "epoch": 0.709914026333193, + "flos": 19049892499080.0, + "grad_norm": 2.55536775543943, + "learning_rate": 8.195193577191553e-07, + "loss": 0.8951, + "num_input_tokens_seen": 125268280, + "step": 5904 + }, + { + "epoch": 0.7100342692238322, + "flos": 18212055909960.0, + "grad_norm": 3.413805500195356, + "learning_rate": 8.188906398775579e-07, + "loss": 0.832, + "num_input_tokens_seen": 125288545, + "step": 5905 + }, + { + "epoch": 0.7101545121144712, + "flos": 18264032581560.0, + "grad_norm": 5.52213635692365, + "learning_rate": 8.18262101209311e-07, + "loss": 0.6711, + "num_input_tokens_seen": 125307475, + "step": 5906 + }, + { + "epoch": 0.7102747550051103, + "flos": 17403658592280.0, + "grad_norm": 3.536680518892313, + "learning_rate": 8.176337418097626e-07, + "loss": 0.7002, + "num_input_tokens_seen": 125327665, + "step": 5907 + }, + { + "epoch": 0.7103949978957494, + "flos": 11139004544040.0, + "grad_norm": 3.627666807854745, + "learning_rate": 8.170055617742364e-07, + "loss": 0.778, + "num_input_tokens_seen": 125344665, + "step": 5908 + }, + { + "epoch": 0.7105152407863885, + "flos": 16567499981040.0, + "grad_norm": 3.5342497111137763, + "learning_rate": 8.163775611980252e-07, + "loss": 0.6871, + "num_input_tokens_seen": 125363495, + "step": 5909 + }, + { + "epoch": 0.7106354836770276, + "flos": 12570936452520.0, + "grad_norm": 5.068015402586067, + "learning_rate": 8.157497401763982e-07, + "loss": 0.7799, + "num_input_tokens_seen": 125380880, + "step": 5910 + }, + { + "epoch": 0.7107557265676667, + "flos": 14757832648920.0, + "grad_norm": 5.142270088749658, + "learning_rate": 8.151220988045935e-07, + "loss": 0.7651, + "num_input_tokens_seen": 125399855, + "step": 5911 + }, + { + "epoch": 0.7108759694583058, + "flos": 15729853351680.0, + "grad_norm": 6.121831268886525, + "learning_rate": 8.144946371778234e-07, + "loss": 0.8178, + "num_input_tokens_seen": 125419685, + "step": 5912 + }, + { + "epoch": 0.7109962123489448, + "flos": 17620715007120.0, + "grad_norm": 4.9379081074541205, + "learning_rate": 8.138673553912751e-07, + "loss": 0.7701, + "num_input_tokens_seen": 125439965, + "step": 5913 + }, + { + "epoch": 0.711116455239584, + "flos": 22370469865800.0, + "grad_norm": 4.429614300266006, + "learning_rate": 8.132402535401059e-07, + "loss": 0.5597, + "num_input_tokens_seen": 125460940, + "step": 5914 + }, + { + "epoch": 0.711236698130223, + "flos": 18348519540720.0, + "grad_norm": 3.316506895373587, + "learning_rate": 8.126133317194465e-07, + "loss": 0.731, + "num_input_tokens_seen": 125480850, + "step": 5915 + }, + { + "epoch": 0.7113569410208621, + "flos": 17727707706480.0, + "grad_norm": 3.911818906348669, + "learning_rate": 8.11986590024401e-07, + "loss": 0.7215, + "num_input_tokens_seen": 125500310, + "step": 5916 + }, + { + "epoch": 0.7114771839115013, + "flos": 26039881604040.0, + "grad_norm": 5.80987653696323, + "learning_rate": 8.113600285500442e-07, + "loss": 0.676, + "num_input_tokens_seen": 125520240, + "step": 5917 + }, + { + "epoch": 0.7115974268021403, + "flos": 15427296858960.0, + "grad_norm": 4.0070607333587045, + "learning_rate": 8.107336473914268e-07, + "loss": 0.7304, + "num_input_tokens_seen": 125538590, + "step": 5918 + }, + { + "epoch": 0.7117176696927794, + "flos": 41819341621680.0, + "grad_norm": 0.8206867272664832, + "learning_rate": 8.101074466435694e-07, + "loss": 0.5988, + "num_input_tokens_seen": 125597785, + "step": 5919 + }, + { + "epoch": 0.7118379125834186, + "flos": 11599010729640.0, + "grad_norm": 4.40876535895895, + "learning_rate": 8.094814264014662e-07, + "loss": 0.6699, + "num_input_tokens_seen": 125616260, + "step": 5920 + }, + { + "epoch": 0.7119581554740576, + "flos": 14757579369240.0, + "grad_norm": 3.3603647592646273, + "learning_rate": 8.088555867600844e-07, + "loss": 0.8032, + "num_input_tokens_seen": 125632145, + "step": 5921 + }, + { + "epoch": 0.7120783983646967, + "flos": 25504949767200.0, + "grad_norm": 4.474761547363358, + "learning_rate": 8.08229927814362e-07, + "loss": 0.603, + "num_input_tokens_seen": 125654755, + "step": 5922 + }, + { + "epoch": 0.7121986412553358, + "flos": 19320603563400.0, + "grad_norm": 4.799642851576099, + "learning_rate": 8.076044496592134e-07, + "loss": 0.6285, + "num_input_tokens_seen": 125676325, + "step": 5923 + }, + { + "epoch": 0.7123188841459749, + "flos": 8061503587920.0, + "grad_norm": 6.0955613794197365, + "learning_rate": 8.069791523895204e-07, + "loss": 0.7628, + "num_input_tokens_seen": 125692495, + "step": 5924 + }, + { + "epoch": 0.7124391270366139, + "flos": 15214894458240.0, + "grad_norm": 3.568058963818719, + "learning_rate": 8.063540361001422e-07, + "loss": 0.7591, + "num_input_tokens_seen": 125710785, + "step": 5925 + }, + { + "epoch": 0.7125593699272531, + "flos": 12841204277400.0, + "grad_norm": 5.626090728125837, + "learning_rate": 8.057291008859069e-07, + "loss": 0.7767, + "num_input_tokens_seen": 125728665, + "step": 5926 + }, + { + "epoch": 0.7126796128178922, + "flos": 21018497542200.0, + "grad_norm": 3.257209703768133, + "learning_rate": 8.051043468416187e-07, + "loss": 0.6666, + "num_input_tokens_seen": 125749635, + "step": 5927 + }, + { + "epoch": 0.7127998557085312, + "flos": 11679761813520.0, + "grad_norm": 8.379776688470452, + "learning_rate": 8.044797740620506e-07, + "loss": 0.8043, + "num_input_tokens_seen": 125767960, + "step": 5928 + }, + { + "epoch": 0.7129200985991703, + "flos": 17133010847880.0, + "grad_norm": 5.02841371948468, + "learning_rate": 8.038553826419494e-07, + "loss": 0.7719, + "num_input_tokens_seen": 125786390, + "step": 5929 + }, + { + "epoch": 0.7130403414898094, + "flos": 15648690688320.0, + "grad_norm": 3.275687307206646, + "learning_rate": 8.032311726760364e-07, + "loss": 0.7933, + "num_input_tokens_seen": 125807330, + "step": 5930 + }, + { + "epoch": 0.7131605843804485, + "flos": 55313985678000.0, + "grad_norm": 12.52973940298983, + "learning_rate": 8.026071442590022e-07, + "loss": 0.6786, + "num_input_tokens_seen": 125833980, + "step": 5931 + }, + { + "epoch": 0.7132808272710875, + "flos": 13407190043640.0, + "grad_norm": 4.074806470897551, + "learning_rate": 8.019832974855134e-07, + "loss": 0.7952, + "num_input_tokens_seen": 125851660, + "step": 5932 + }, + { + "epoch": 0.7134010701617267, + "flos": 17021522434200.0, + "grad_norm": 7.943642469829352, + "learning_rate": 8.013596324502052e-07, + "loss": 0.8173, + "num_input_tokens_seen": 125869845, + "step": 5933 + }, + { + "epoch": 0.7135213130523658, + "flos": 17317145395680.0, + "grad_norm": 5.891915177531709, + "learning_rate": 8.007361492476872e-07, + "loss": 0.767, + "num_input_tokens_seen": 125888890, + "step": 5934 + }, + { + "epoch": 0.7136415559430048, + "flos": 18159889278600.0, + "grad_norm": 3.5072681709834166, + "learning_rate": 8.001128479725426e-07, + "loss": 0.7713, + "num_input_tokens_seen": 125910515, + "step": 5935 + }, + { + "epoch": 0.713761798833644, + "flos": 13354231913280.0, + "grad_norm": 5.950075024043651, + "learning_rate": 7.994897287193248e-07, + "loss": 0.7895, + "num_input_tokens_seen": 125929615, + "step": 5936 + }, + { + "epoch": 0.713882041724283, + "flos": 11327096586840.0, + "grad_norm": 7.318756296365247, + "learning_rate": 7.988667915825605e-07, + "loss": 0.8114, + "num_input_tokens_seen": 125946400, + "step": 5937 + }, + { + "epoch": 0.7140022846149221, + "flos": 17619923508120.0, + "grad_norm": 4.237221074559644, + "learning_rate": 7.982440366567491e-07, + "loss": 0.7306, + "num_input_tokens_seen": 125964610, + "step": 5938 + }, + { + "epoch": 0.7141225275055613, + "flos": 20454854613000.0, + "grad_norm": 3.188846528114924, + "learning_rate": 7.97621464036361e-07, + "loss": 0.7375, + "num_input_tokens_seen": 125986090, + "step": 5939 + }, + { + "epoch": 0.7142427703962003, + "flos": 14379590665920.0, + "grad_norm": 6.398071725751543, + "learning_rate": 7.969990738158417e-07, + "loss": 0.6635, + "num_input_tokens_seen": 126004220, + "step": 5940 + }, + { + "epoch": 0.7143630132868394, + "flos": 15378486183360.0, + "grad_norm": 4.315064041175213, + "learning_rate": 7.963768660896062e-07, + "loss": 0.8306, + "num_input_tokens_seen": 126022350, + "step": 5941 + }, + { + "epoch": 0.7144832561774785, + "flos": 17670348841680.0, + "grad_norm": 5.553185368490449, + "learning_rate": 7.957548409520432e-07, + "loss": 0.8147, + "num_input_tokens_seen": 126041295, + "step": 5942 + }, + { + "epoch": 0.7146034990681176, + "flos": 11895330210240.0, + "grad_norm": 4.41548001644609, + "learning_rate": 7.951329984975135e-07, + "loss": 0.8206, + "num_input_tokens_seen": 126057955, + "step": 5943 + }, + { + "epoch": 0.7147237419587567, + "flos": 51346454332800.0, + "grad_norm": 0.7229939264646509, + "learning_rate": 7.94511338820349e-07, + "loss": 0.5547, + "num_input_tokens_seen": 126119980, + "step": 5944 + }, + { + "epoch": 0.7148439848493958, + "flos": 16291185103800.0, + "grad_norm": 3.4330287637769543, + "learning_rate": 7.938898620148575e-07, + "loss": 0.7743, + "num_input_tokens_seen": 126137460, + "step": 5945 + }, + { + "epoch": 0.7149642277400349, + "flos": 13083584168880.0, + "grad_norm": 5.126715768315209, + "learning_rate": 7.932685681753135e-07, + "loss": 0.7036, + "num_input_tokens_seen": 126154460, + "step": 5946 + }, + { + "epoch": 0.7150844706306739, + "flos": 23259270007800.0, + "grad_norm": 4.888413584367241, + "learning_rate": 7.92647457395969e-07, + "loss": 0.6072, + "num_input_tokens_seen": 126176005, + "step": 5947 + }, + { + "epoch": 0.7152047135213131, + "flos": 7899304901040.0, + "grad_norm": 8.592765834981437, + "learning_rate": 7.920265297710444e-07, + "loss": 0.7306, + "num_input_tokens_seen": 126193115, + "step": 5948 + }, + { + "epoch": 0.7153249564119522, + "flos": 15351263109120.0, + "grad_norm": 5.901542604153814, + "learning_rate": 7.914057853947363e-07, + "loss": 0.7087, + "num_input_tokens_seen": 126212015, + "step": 5949 + }, + { + "epoch": 0.7154451993025912, + "flos": 17754645841080.0, + "grad_norm": 3.9132523907073082, + "learning_rate": 7.907852243612089e-07, + "loss": 0.622, + "num_input_tokens_seen": 126232140, + "step": 5950 + }, + { + "epoch": 0.7155654421932304, + "flos": 17079577818120.0, + "grad_norm": 5.675280848876384, + "learning_rate": 7.901648467646009e-07, + "loss": 0.7182, + "num_input_tokens_seen": 126250800, + "step": 5951 + }, + { + "epoch": 0.7156856850838694, + "flos": 16621597869960.0, + "grad_norm": 3.245888970958802, + "learning_rate": 7.895446526990244e-07, + "loss": 0.7096, + "num_input_tokens_seen": 126270535, + "step": 5952 + }, + { + "epoch": 0.7158059279745085, + "flos": 14514946198080.0, + "grad_norm": 4.818198855060797, + "learning_rate": 7.889246422585609e-07, + "loss": 0.7419, + "num_input_tokens_seen": 126289640, + "step": 5953 + }, + { + "epoch": 0.7159261708651476, + "flos": 17674622936280.0, + "grad_norm": 4.731548208214228, + "learning_rate": 7.883048155372675e-07, + "loss": 0.7304, + "num_input_tokens_seen": 126307875, + "step": 5954 + }, + { + "epoch": 0.7160464137557867, + "flos": 12381926270880.0, + "grad_norm": 5.425544310656262, + "learning_rate": 7.876851726291698e-07, + "loss": 0.6852, + "num_input_tokens_seen": 126325895, + "step": 5955 + }, + { + "epoch": 0.7161666566464258, + "flos": 18483716773080.0, + "grad_norm": 3.6391782055121467, + "learning_rate": 7.870657136282666e-07, + "loss": 0.7619, + "num_input_tokens_seen": 126344475, + "step": 5956 + }, + { + "epoch": 0.7162868995370649, + "flos": 19401449627160.0, + "grad_norm": 4.710634844975802, + "learning_rate": 7.86446438628531e-07, + "loss": 0.8079, + "num_input_tokens_seen": 126365265, + "step": 5957 + }, + { + "epoch": 0.716407142427704, + "flos": 51622547590320.0, + "grad_norm": 0.7856996847629096, + "learning_rate": 7.858273477239059e-07, + "loss": 0.5911, + "num_input_tokens_seen": 126433405, + "step": 5958 + }, + { + "epoch": 0.716527385318343, + "flos": 15162442887240.0, + "grad_norm": 4.83192526378055, + "learning_rate": 7.852084410083067e-07, + "loss": 0.6979, + "num_input_tokens_seen": 126451945, + "step": 5959 + }, + { + "epoch": 0.7166476282089821, + "flos": 18589759673640.0, + "grad_norm": 3.5052029220894494, + "learning_rate": 7.84589718575621e-07, + "loss": 0.6234, + "num_input_tokens_seen": 126472110, + "step": 5960 + }, + { + "epoch": 0.7167678710996213, + "flos": 17673704797440.0, + "grad_norm": 4.514564680064765, + "learning_rate": 7.83971180519708e-07, + "loss": 0.6724, + "num_input_tokens_seen": 126490685, + "step": 5961 + }, + { + "epoch": 0.7168881139902603, + "flos": 22182789402480.0, + "grad_norm": 5.235919052815534, + "learning_rate": 7.833528269344008e-07, + "loss": 0.7323, + "num_input_tokens_seen": 126510310, + "step": 5962 + }, + { + "epoch": 0.7170083568808994, + "flos": 10622209372920.0, + "grad_norm": 8.640328317691406, + "learning_rate": 7.827346579135023e-07, + "loss": 0.7608, + "num_input_tokens_seen": 126527370, + "step": 5963 + }, + { + "epoch": 0.7171285997715385, + "flos": 17079704457960.0, + "grad_norm": 5.972075010550405, + "learning_rate": 7.821166735507885e-07, + "loss": 0.8231, + "num_input_tokens_seen": 126546120, + "step": 5964 + }, + { + "epoch": 0.7172488426621776, + "flos": 12056294158680.0, + "grad_norm": 5.519770201999875, + "learning_rate": 7.81498873940007e-07, + "loss": 0.672, + "num_input_tokens_seen": 126563055, + "step": 5965 + }, + { + "epoch": 0.7173690855528166, + "flos": 19455674155920.0, + "grad_norm": 5.12837164156126, + "learning_rate": 7.808812591748768e-07, + "loss": 0.748, + "num_input_tokens_seen": 126583155, + "step": 5966 + }, + { + "epoch": 0.7174893284434558, + "flos": 16674935919840.0, + "grad_norm": 4.834271416667137, + "learning_rate": 7.802638293490915e-07, + "loss": 0.6258, + "num_input_tokens_seen": 126602520, + "step": 5967 + }, + { + "epoch": 0.7176095713340949, + "flos": 17051784864600.0, + "grad_norm": 3.3518609756554745, + "learning_rate": 7.796465845563123e-07, + "loss": 0.7626, + "num_input_tokens_seen": 126621115, + "step": 5968 + }, + { + "epoch": 0.7177298142247339, + "flos": 18752433259920.0, + "grad_norm": 9.3316281895481, + "learning_rate": 7.790295248901766e-07, + "loss": 0.7884, + "num_input_tokens_seen": 126641965, + "step": 5969 + }, + { + "epoch": 0.7178500571153731, + "flos": 23238727185120.0, + "grad_norm": 5.443597641224139, + "learning_rate": 7.784126504442902e-07, + "loss": 0.6054, + "num_input_tokens_seen": 126664915, + "step": 5970 + }, + { + "epoch": 0.7179703000060121, + "flos": 14190612144240.0, + "grad_norm": 3.9218574918682902, + "learning_rate": 7.777959613122351e-07, + "loss": 0.6631, + "num_input_tokens_seen": 126684460, + "step": 5971 + }, + { + "epoch": 0.7180905428966512, + "flos": 21155942631720.0, + "grad_norm": 5.259511052602815, + "learning_rate": 7.771794575875604e-07, + "loss": 0.7574, + "num_input_tokens_seen": 126706050, + "step": 5972 + }, + { + "epoch": 0.7182107857872904, + "flos": 14649700191000.0, + "grad_norm": 4.645373243235652, + "learning_rate": 7.765631393637888e-07, + "loss": 0.7467, + "num_input_tokens_seen": 126723965, + "step": 5973 + }, + { + "epoch": 0.7183310286779294, + "flos": 16647997785240.0, + "grad_norm": 12.462333997961931, + "learning_rate": 7.75947006734417e-07, + "loss": 0.4523, + "num_input_tokens_seen": 126741465, + "step": 5974 + }, + { + "epoch": 0.7184512715685685, + "flos": 12510443251680.0, + "grad_norm": 5.467553928913383, + "learning_rate": 7.753310597929101e-07, + "loss": 0.8196, + "num_input_tokens_seen": 126757825, + "step": 5975 + }, + { + "epoch": 0.7185715144592076, + "flos": 48299943986160.0, + "grad_norm": 0.8163878151304385, + "learning_rate": 7.747152986327095e-07, + "loss": 0.5695, + "num_input_tokens_seen": 126818090, + "step": 5976 + }, + { + "epoch": 0.7186917573498467, + "flos": 11787894271440.0, + "grad_norm": 4.540073196896578, + "learning_rate": 7.740997233472228e-07, + "loss": 0.6653, + "num_input_tokens_seen": 126835430, + "step": 5977 + }, + { + "epoch": 0.7188120002404857, + "flos": 21454446649560.0, + "grad_norm": 4.6294035696716564, + "learning_rate": 7.734843340298329e-07, + "loss": 0.6935, + "num_input_tokens_seen": 126854975, + "step": 5978 + }, + { + "epoch": 0.7189322431311249, + "flos": 24532232545320.0, + "grad_norm": 3.851991084087507, + "learning_rate": 7.72869130773895e-07, + "loss": 0.733, + "num_input_tokens_seen": 126875295, + "step": 5979 + }, + { + "epoch": 0.719052486021764, + "flos": 45222632989800.0, + "grad_norm": 0.7930277158865351, + "learning_rate": 7.722541136727343e-07, + "loss": 0.6032, + "num_input_tokens_seen": 126931030, + "step": 5980 + }, + { + "epoch": 0.719172728912403, + "flos": 11355902659080.0, + "grad_norm": 5.7436651080378125, + "learning_rate": 7.716392828196483e-07, + "loss": 0.7864, + "num_input_tokens_seen": 126948550, + "step": 5981 + }, + { + "epoch": 0.7192929718030422, + "flos": 11322600872520.0, + "grad_norm": 5.071040167413052, + "learning_rate": 7.710246383079064e-07, + "loss": 0.752, + "num_input_tokens_seen": 126963655, + "step": 5982 + }, + { + "epoch": 0.7194132146936812, + "flos": 15991541327400.0, + "grad_norm": 8.711510718948102, + "learning_rate": 7.704101802307492e-07, + "loss": 0.9006, + "num_input_tokens_seen": 126975675, + "step": 5983 + }, + { + "epoch": 0.7195334575843203, + "flos": 20044957161360.0, + "grad_norm": 6.239130257368266, + "learning_rate": 7.697959086813912e-07, + "loss": 0.8621, + "num_input_tokens_seen": 126991560, + "step": 5984 + }, + { + "epoch": 0.7196537004749595, + "flos": 13704680942760.0, + "grad_norm": 4.3275791746727785, + "learning_rate": 7.691818237530145e-07, + "loss": 0.7901, + "num_input_tokens_seen": 127010140, + "step": 5985 + }, + { + "epoch": 0.7197739433655985, + "flos": 17967143221680.0, + "grad_norm": 8.022922577348538, + "learning_rate": 7.685679255387774e-07, + "loss": 0.7596, + "num_input_tokens_seen": 127028175, + "step": 5986 + }, + { + "epoch": 0.7198941862562376, + "flos": 13164366912720.0, + "grad_norm": 4.638214204395844, + "learning_rate": 7.679542141318065e-07, + "loss": 0.7466, + "num_input_tokens_seen": 127045000, + "step": 5987 + }, + { + "epoch": 0.7200144291468767, + "flos": 21291583103520.0, + "grad_norm": 4.113108748163456, + "learning_rate": 7.673406896252013e-07, + "loss": 0.7516, + "num_input_tokens_seen": 127066095, + "step": 5988 + }, + { + "epoch": 0.7201346720375158, + "flos": 18591785911080.0, + "grad_norm": 3.312766730168048, + "learning_rate": 7.667273521120347e-07, + "loss": 0.7711, + "num_input_tokens_seen": 127085375, + "step": 5989 + }, + { + "epoch": 0.7202549149281549, + "flos": 10437251666160.0, + "grad_norm": 14.384161907095292, + "learning_rate": 7.661142016853468e-07, + "loss": 0.7878, + "num_input_tokens_seen": 127102455, + "step": 5990 + }, + { + "epoch": 0.7203751578187939, + "flos": 16835773228440.0, + "grad_norm": 4.038453283056967, + "learning_rate": 7.655012384381543e-07, + "loss": 0.7297, + "num_input_tokens_seen": 127121660, + "step": 5991 + }, + { + "epoch": 0.7204954007094331, + "flos": 17346616327080.0, + "grad_norm": 4.39313628033134, + "learning_rate": 7.648884624634415e-07, + "loss": 0.7994, + "num_input_tokens_seen": 127139930, + "step": 5992 + }, + { + "epoch": 0.7206156436000721, + "flos": 11841485601000.0, + "grad_norm": 4.270503457828502, + "learning_rate": 7.642758738541683e-07, + "loss": 0.8747, + "num_input_tokens_seen": 127156230, + "step": 5993 + }, + { + "epoch": 0.7207358864907112, + "flos": 40061144401800.0, + "grad_norm": 0.8176325299222883, + "learning_rate": 7.636634727032621e-07, + "loss": 0.6111, + "num_input_tokens_seen": 127213055, + "step": 5994 + }, + { + "epoch": 0.7208561293813504, + "flos": 13974758807880.0, + "grad_norm": 4.32518924894444, + "learning_rate": 7.630512591036231e-07, + "loss": 0.7773, + "num_input_tokens_seen": 127232085, + "step": 5995 + }, + { + "epoch": 0.7209763722719894, + "flos": 12948196976760.0, + "grad_norm": 5.902624238454505, + "learning_rate": 7.624392331481255e-07, + "loss": 0.629, + "num_input_tokens_seen": 127249460, + "step": 5996 + }, + { + "epoch": 0.7210966151626285, + "flos": 49268830352880.0, + "grad_norm": 0.7576516757480072, + "learning_rate": 7.618273949296115e-07, + "loss": 0.533, + "num_input_tokens_seen": 127308690, + "step": 5997 + }, + { + "epoch": 0.7212168580532676, + "flos": 15459522206880.0, + "grad_norm": 7.027297608763194, + "learning_rate": 7.612157445408987e-07, + "loss": 0.6755, + "num_input_tokens_seen": 127326220, + "step": 5998 + }, + { + "epoch": 0.7213371009439067, + "flos": 16350443566200.0, + "grad_norm": 11.544200045369102, + "learning_rate": 7.606042820747716e-07, + "loss": 0.7217, + "num_input_tokens_seen": 127342345, + "step": 5999 + }, + { + "epoch": 0.7214573438345457, + "flos": 13516272300360.0, + "grad_norm": 4.269182559168763, + "learning_rate": 7.599930076239889e-07, + "loss": 0.834, + "num_input_tokens_seen": 127359350, + "step": 6000 + }, + { + "epoch": 0.7215775867251849, + "flos": 26260230654720.0, + "grad_norm": 7.024921468384034, + "learning_rate": 7.593819212812818e-07, + "loss": 0.6956, + "num_input_tokens_seen": 127380650, + "step": 6001 + }, + { + "epoch": 0.721697829615824, + "flos": 14889610605600.0, + "grad_norm": 4.46659767230443, + "learning_rate": 7.587710231393508e-07, + "loss": 0.705, + "num_input_tokens_seen": 127398725, + "step": 6002 + }, + { + "epoch": 0.721818072506463, + "flos": 14784549163800.0, + "grad_norm": 5.96288870578089, + "learning_rate": 7.581603132908685e-07, + "loss": 0.8166, + "num_input_tokens_seen": 127416415, + "step": 6003 + }, + { + "epoch": 0.7219383153971022, + "flos": 13272689330400.0, + "grad_norm": 5.595487186333018, + "learning_rate": 7.575497918284795e-07, + "loss": 0.7617, + "num_input_tokens_seen": 127433680, + "step": 6004 + }, + { + "epoch": 0.7220585582877412, + "flos": 12646716922680.0, + "grad_norm": 7.177146287558016, + "learning_rate": 7.569394588447984e-07, + "loss": 0.7366, + "num_input_tokens_seen": 127450415, + "step": 6005 + }, + { + "epoch": 0.7221788011783803, + "flos": 12375879218520.0, + "grad_norm": 5.637492062717441, + "learning_rate": 7.563293144324146e-07, + "loss": 0.7572, + "num_input_tokens_seen": 127465685, + "step": 6006 + }, + { + "epoch": 0.7222990440690195, + "flos": 19266790614120.0, + "grad_norm": 5.493935102137096, + "learning_rate": 7.557193586838834e-07, + "loss": 0.7895, + "num_input_tokens_seen": 127480770, + "step": 6007 + }, + { + "epoch": 0.7224192869596585, + "flos": 12839399659680.0, + "grad_norm": 10.257629532355264, + "learning_rate": 7.551095916917371e-07, + "loss": 0.6848, + "num_input_tokens_seen": 127497820, + "step": 6008 + }, + { + "epoch": 0.7225395298502976, + "flos": 9383878340280.0, + "grad_norm": 5.181887612493696, + "learning_rate": 7.545000135484758e-07, + "loss": 0.6494, + "num_input_tokens_seen": 127514975, + "step": 6009 + }, + { + "epoch": 0.7226597727409367, + "flos": 21751715928960.0, + "grad_norm": 10.595349396457287, + "learning_rate": 7.538906243465714e-07, + "loss": 0.62, + "num_input_tokens_seen": 127534830, + "step": 6010 + }, + { + "epoch": 0.7227800156315758, + "flos": 10005038434080.0, + "grad_norm": 4.6028126700646865, + "learning_rate": 7.5328142417847e-07, + "loss": 0.7691, + "num_input_tokens_seen": 127551315, + "step": 6011 + }, + { + "epoch": 0.7229002585222148, + "flos": 14837760573840.0, + "grad_norm": 3.7864854366516583, + "learning_rate": 7.526724131365838e-07, + "loss": 0.6824, + "num_input_tokens_seen": 127571990, + "step": 6012 + }, + { + "epoch": 0.723020501412854, + "flos": 12084372051840.0, + "grad_norm": 6.949025246304642, + "learning_rate": 7.520635913133017e-07, + "loss": 0.6834, + "num_input_tokens_seen": 127590340, + "step": 6013 + }, + { + "epoch": 0.7231407443034931, + "flos": 20940754154520.0, + "grad_norm": 3.4152486386132717, + "learning_rate": 7.514549588009798e-07, + "loss": 0.8051, + "num_input_tokens_seen": 127610935, + "step": 6014 + }, + { + "epoch": 0.7232609871941321, + "flos": 22021540514400.0, + "grad_norm": 3.231511566967739, + "learning_rate": 7.508465156919492e-07, + "loss": 0.7007, + "num_input_tokens_seen": 127634165, + "step": 6015 + }, + { + "epoch": 0.7233812300847713, + "flos": 12138185001120.0, + "grad_norm": 7.428625887367252, + "learning_rate": 7.502382620785083e-07, + "loss": 0.5929, + "num_input_tokens_seen": 127650435, + "step": 6016 + }, + { + "epoch": 0.7235014729754103, + "flos": 49594145865480.0, + "grad_norm": 0.8969869098558737, + "learning_rate": 7.496301980529289e-07, + "loss": 0.6849, + "num_input_tokens_seen": 127713365, + "step": 6017 + }, + { + "epoch": 0.7236217158660494, + "flos": 19752943435320.0, + "grad_norm": 7.937079951668604, + "learning_rate": 7.490223237074547e-07, + "loss": 0.7318, + "num_input_tokens_seen": 127732795, + "step": 6018 + }, + { + "epoch": 0.7237419587566886, + "flos": 21588282503640.0, + "grad_norm": 3.854037652133157, + "learning_rate": 7.484146391342989e-07, + "loss": 0.6436, + "num_input_tokens_seen": 127752310, + "step": 6019 + }, + { + "epoch": 0.7238622016473276, + "flos": 13002199885800.0, + "grad_norm": 4.882838867116309, + "learning_rate": 7.478071444256484e-07, + "loss": 0.5442, + "num_input_tokens_seen": 127769790, + "step": 6020 + }, + { + "epoch": 0.7239824445379667, + "flos": 18861800456280.0, + "grad_norm": 7.036669731964691, + "learning_rate": 7.471998396736579e-07, + "loss": 0.7637, + "num_input_tokens_seen": 127789890, + "step": 6021 + }, + { + "epoch": 0.7241026874286057, + "flos": 16944918805080.0, + "grad_norm": 4.476810042375885, + "learning_rate": 7.465927249704549e-07, + "loss": 0.7429, + "num_input_tokens_seen": 127807495, + "step": 6022 + }, + { + "epoch": 0.7242229303192449, + "flos": 14812088837640.0, + "grad_norm": 9.433694842884831, + "learning_rate": 7.459858004081398e-07, + "loss": 0.7518, + "num_input_tokens_seen": 127825185, + "step": 6023 + }, + { + "epoch": 0.724343173209884, + "flos": 45935055274200.0, + "grad_norm": 0.6733216967222403, + "learning_rate": 7.453790660787815e-07, + "loss": 0.5925, + "num_input_tokens_seen": 127893000, + "step": 6024 + }, + { + "epoch": 0.724463416100523, + "flos": 25720201564320.0, + "grad_norm": 9.364085036030573, + "learning_rate": 7.447725220744214e-07, + "loss": 0.6226, + "num_input_tokens_seen": 127914965, + "step": 6025 + }, + { + "epoch": 0.7245836589911622, + "flos": 15999899556840.0, + "grad_norm": 3.654181653423018, + "learning_rate": 7.441661684870717e-07, + "loss": 0.7407, + "num_input_tokens_seen": 127934940, + "step": 6026 + }, + { + "epoch": 0.7247039018818012, + "flos": 16840015663080.0, + "grad_norm": 4.90405924292371, + "learning_rate": 7.435600054087152e-07, + "loss": 0.8024, + "num_input_tokens_seen": 127956825, + "step": 6027 + }, + { + "epoch": 0.7248241447724403, + "flos": 23289944017680.0, + "grad_norm": 4.609717404853828, + "learning_rate": 7.42954032931308e-07, + "loss": 0.732, + "num_input_tokens_seen": 127977585, + "step": 6028 + }, + { + "epoch": 0.7249443876630794, + "flos": 25639228860720.0, + "grad_norm": 3.772633216078213, + "learning_rate": 7.423482511467733e-07, + "loss": 0.7328, + "num_input_tokens_seen": 127998075, + "step": 6029 + }, + { + "epoch": 0.7250646305537185, + "flos": 19320761863200.0, + "grad_norm": 4.590188650605172, + "learning_rate": 7.417426601470099e-07, + "loss": 0.6406, + "num_input_tokens_seen": 128018155, + "step": 6030 + }, + { + "epoch": 0.7251848734443576, + "flos": 22075638403320.0, + "grad_norm": 8.464997871520332, + "learning_rate": 7.411372600238841e-07, + "loss": 0.7625, + "num_input_tokens_seen": 128038490, + "step": 6031 + }, + { + "epoch": 0.7253051163349967, + "flos": 12975420051000.0, + "grad_norm": 5.50103420591026, + "learning_rate": 7.405320508692346e-07, + "loss": 0.714, + "num_input_tokens_seen": 128056950, + "step": 6032 + }, + { + "epoch": 0.7254253592256358, + "flos": 9168658203120.0, + "grad_norm": 3.459643808894758, + "learning_rate": 7.399270327748727e-07, + "loss": 0.7291, + "num_input_tokens_seen": 128074330, + "step": 6033 + }, + { + "epoch": 0.7255456021162748, + "flos": 19942460176320.0, + "grad_norm": 4.333682902003466, + "learning_rate": 7.39322205832577e-07, + "loss": 0.7292, + "num_input_tokens_seen": 128094940, + "step": 6034 + }, + { + "epoch": 0.725665845006914, + "flos": 15567876284520.0, + "grad_norm": 3.809543025728988, + "learning_rate": 7.387175701341009e-07, + "loss": 0.7936, + "num_input_tokens_seen": 128113330, + "step": 6035 + }, + { + "epoch": 0.7257860878975531, + "flos": 11679096954360.0, + "grad_norm": 6.532880807632152, + "learning_rate": 7.381131257711659e-07, + "loss": 0.7062, + "num_input_tokens_seen": 128130155, + "step": 6036 + }, + { + "epoch": 0.7259063307881921, + "flos": 8790099620520.0, + "grad_norm": 5.0412388559639645, + "learning_rate": 7.375088728354677e-07, + "loss": 0.8237, + "num_input_tokens_seen": 128144905, + "step": 6037 + }, + { + "epoch": 0.7260265736788313, + "flos": 22343341771440.0, + "grad_norm": 4.302267124538953, + "learning_rate": 7.369048114186691e-07, + "loss": 0.6496, + "num_input_tokens_seen": 128165670, + "step": 6038 + }, + { + "epoch": 0.7261468165694703, + "flos": 15459870466440.0, + "grad_norm": 3.3415534389526043, + "learning_rate": 7.363009416124055e-07, + "loss": 0.8238, + "num_input_tokens_seen": 128184715, + "step": 6039 + }, + { + "epoch": 0.7262670594601094, + "flos": 16320845994960.0, + "grad_norm": 4.028857800316474, + "learning_rate": 7.356972635082852e-07, + "loss": 0.6127, + "num_input_tokens_seen": 128203290, + "step": 6040 + }, + { + "epoch": 0.7263873023507486, + "flos": 18562378299600.0, + "grad_norm": 4.7765308103299455, + "learning_rate": 7.35093777197884e-07, + "loss": 0.7431, + "num_input_tokens_seen": 128223080, + "step": 6041 + }, + { + "epoch": 0.7265075452413876, + "flos": 17484947895480.0, + "grad_norm": 4.460389753340955, + "learning_rate": 7.344904827727525e-07, + "loss": 0.8403, + "num_input_tokens_seen": 128239980, + "step": 6042 + }, + { + "epoch": 0.7266277881320267, + "flos": 21070315914000.0, + "grad_norm": 4.257384224447015, + "learning_rate": 7.338873803244076e-07, + "loss": 0.698, + "num_input_tokens_seen": 128254935, + "step": 6043 + }, + { + "epoch": 0.7267480310226658, + "flos": 18213797207760.0, + "grad_norm": 5.054319766315829, + "learning_rate": 7.332844699443401e-07, + "loss": 0.7844, + "num_input_tokens_seen": 128273255, + "step": 6044 + }, + { + "epoch": 0.7268682739133049, + "flos": 19941826977120.0, + "grad_norm": 4.647340789415369, + "learning_rate": 7.326817517240121e-07, + "loss": 0.736, + "num_input_tokens_seen": 128294680, + "step": 6045 + }, + { + "epoch": 0.7269885168039439, + "flos": 24611052371640.0, + "grad_norm": 4.053588718307814, + "learning_rate": 7.320792257548545e-07, + "loss": 0.822, + "num_input_tokens_seen": 128315575, + "step": 6046 + }, + { + "epoch": 0.7271087596945831, + "flos": 17806590852720.0, + "grad_norm": 3.96174327471888, + "learning_rate": 7.314768921282704e-07, + "loss": 0.7506, + "num_input_tokens_seen": 128335950, + "step": 6047 + }, + { + "epoch": 0.7272290025852222, + "flos": 17430501747000.0, + "grad_norm": 6.243713879791042, + "learning_rate": 7.30874750935633e-07, + "loss": 0.6984, + "num_input_tokens_seen": 128355355, + "step": 6048 + }, + { + "epoch": 0.7273492454758612, + "flos": 12186362477520.0, + "grad_norm": 3.4983134974813446, + "learning_rate": 7.30272802268286e-07, + "loss": 0.7736, + "num_input_tokens_seen": 128372070, + "step": 6049 + }, + { + "epoch": 0.7274694883665004, + "flos": 20557288278120.0, + "grad_norm": 3.1242613873215848, + "learning_rate": 7.29671046217547e-07, + "loss": 0.7498, + "num_input_tokens_seen": 128390900, + "step": 6050 + }, + { + "epoch": 0.7275897312571394, + "flos": 22290953520360.0, + "grad_norm": 9.7071986592707, + "learning_rate": 7.290694828746988e-07, + "loss": 0.8027, + "num_input_tokens_seen": 128410285, + "step": 6051 + }, + { + "epoch": 0.7277099741477785, + "flos": 14025469081080.0, + "grad_norm": 4.802248253407305, + "learning_rate": 7.284681123310004e-07, + "loss": 0.8439, + "num_input_tokens_seen": 128428720, + "step": 6052 + }, + { + "epoch": 0.7278302170384175, + "flos": 15108788237760.0, + "grad_norm": 3.4878539093815952, + "learning_rate": 7.27866934677678e-07, + "loss": 0.7767, + "num_input_tokens_seen": 128448110, + "step": 6053 + }, + { + "epoch": 0.7279504599290567, + "flos": 13942596779880.0, + "grad_norm": 3.709511316200166, + "learning_rate": 7.272659500059297e-07, + "loss": 0.7596, + "num_input_tokens_seen": 128465170, + "step": 6054 + }, + { + "epoch": 0.7280707028196958, + "flos": 13920407639280.0, + "grad_norm": 4.778042184674626, + "learning_rate": 7.266651584069264e-07, + "loss": 0.7895, + "num_input_tokens_seen": 128482555, + "step": 6055 + }, + { + "epoch": 0.7281909457103348, + "flos": 27340985354640.0, + "grad_norm": 7.376270249453226, + "learning_rate": 7.260645599718045e-07, + "loss": 0.5647, + "num_input_tokens_seen": 128508630, + "step": 6056 + }, + { + "epoch": 0.728311188600974, + "flos": 15108091718640.0, + "grad_norm": 8.44106641206261, + "learning_rate": 7.254641547916767e-07, + "loss": 0.6605, + "num_input_tokens_seen": 128525845, + "step": 6057 + }, + { + "epoch": 0.728431431491613, + "flos": 21157209030120.0, + "grad_norm": 11.046037224022568, + "learning_rate": 7.248639429576226e-07, + "loss": 0.6734, + "num_input_tokens_seen": 128545020, + "step": 6058 + }, + { + "epoch": 0.7285516743822521, + "flos": 19048467800880.0, + "grad_norm": 3.196872022819438, + "learning_rate": 7.242639245606959e-07, + "loss": 0.7116, + "num_input_tokens_seen": 128564530, + "step": 6059 + }, + { + "epoch": 0.7286719172728913, + "flos": 11950156278240.0, + "grad_norm": 7.007820665233537, + "learning_rate": 7.236640996919168e-07, + "loss": 0.8067, + "num_input_tokens_seen": 128583295, + "step": 6060 + }, + { + "epoch": 0.7287921601635303, + "flos": 16108126994640.0, + "grad_norm": 4.985836185584362, + "learning_rate": 7.230644684422782e-07, + "loss": 0.6894, + "num_input_tokens_seen": 128603245, + "step": 6061 + }, + { + "epoch": 0.7289124030541694, + "flos": 18018455034120.0, + "grad_norm": 4.3243856975027795, + "learning_rate": 7.224650309027451e-07, + "loss": 0.808, + "num_input_tokens_seen": 128622715, + "step": 6062 + }, + { + "epoch": 0.7290326459448085, + "flos": 15645334732560.0, + "grad_norm": 4.508480134972473, + "learning_rate": 7.218657871642506e-07, + "loss": 0.6674, + "num_input_tokens_seen": 128641240, + "step": 6063 + }, + { + "epoch": 0.7291528888354476, + "flos": 13569230430720.0, + "grad_norm": 3.9914790232175394, + "learning_rate": 7.212667373177012e-07, + "loss": 0.5991, + "num_input_tokens_seen": 128655955, + "step": 6064 + }, + { + "epoch": 0.7292731317260867, + "flos": 13837946917560.0, + "grad_norm": 4.823551320227368, + "learning_rate": 7.206678814539704e-07, + "loss": 0.7395, + "num_input_tokens_seen": 128673975, + "step": 6065 + }, + { + "epoch": 0.7293933746167258, + "flos": 15404347879320.0, + "grad_norm": 3.768431950903626, + "learning_rate": 7.20069219663904e-07, + "loss": 0.7221, + "num_input_tokens_seen": 128693580, + "step": 6066 + }, + { + "epoch": 0.7295136175073649, + "flos": 16429801611840.0, + "grad_norm": 3.094332553462219, + "learning_rate": 7.1947075203832e-07, + "loss": 0.7702, + "num_input_tokens_seen": 128713280, + "step": 6067 + }, + { + "epoch": 0.7296338603980039, + "flos": 41354048222760.0, + "grad_norm": 0.8702714400133894, + "learning_rate": 7.188724786680049e-07, + "loss": 0.6053, + "num_input_tokens_seen": 128773470, + "step": 6068 + }, + { + "epoch": 0.7297541032886431, + "flos": 18483748433040.0, + "grad_norm": 3.2797487730251773, + "learning_rate": 7.182743996437162e-07, + "loss": 0.7324, + "num_input_tokens_seen": 128792725, + "step": 6069 + }, + { + "epoch": 0.7298743461792822, + "flos": 19400753108040.0, + "grad_norm": 9.644401119894681, + "learning_rate": 7.176765150561819e-07, + "loss": 0.6672, + "num_input_tokens_seen": 128811050, + "step": 6070 + }, + { + "epoch": 0.7299945890699212, + "flos": 14294977066920.0, + "grad_norm": 8.320186875518763, + "learning_rate": 7.170788249961002e-07, + "loss": 0.7687, + "num_input_tokens_seen": 128829280, + "step": 6071 + }, + { + "epoch": 0.7301148319605604, + "flos": 16781991939120.0, + "grad_norm": 6.713367666076533, + "learning_rate": 7.164813295541418e-07, + "loss": 0.8595, + "num_input_tokens_seen": 128848565, + "step": 6072 + }, + { + "epoch": 0.7302350748511994, + "flos": 18587765096160.0, + "grad_norm": 4.232144940394739, + "learning_rate": 7.15884028820944e-07, + "loss": 0.6811, + "num_input_tokens_seen": 128867340, + "step": 6073 + }, + { + "epoch": 0.7303553177418385, + "flos": 20401231623480.0, + "grad_norm": 5.670501794963957, + "learning_rate": 7.152869228871185e-07, + "loss": 0.5729, + "num_input_tokens_seen": 128889545, + "step": 6074 + }, + { + "epoch": 0.7304755606324776, + "flos": 17890507932600.0, + "grad_norm": 6.883870453772094, + "learning_rate": 7.146900118432457e-07, + "loss": 0.7041, + "num_input_tokens_seen": 128909010, + "step": 6075 + }, + { + "epoch": 0.7305958035231167, + "flos": 17456490082800.0, + "grad_norm": 3.9084030931397487, + "learning_rate": 7.140932957798753e-07, + "loss": 0.8462, + "num_input_tokens_seen": 128927170, + "step": 6076 + }, + { + "epoch": 0.7307160464137558, + "flos": 12192156250200.0, + "grad_norm": 5.535227062388873, + "learning_rate": 7.134967747875309e-07, + "loss": 0.6935, + "num_input_tokens_seen": 128945100, + "step": 6077 + }, + { + "epoch": 0.7308362893043949, + "flos": 15945263448600.0, + "grad_norm": 3.778568294193164, + "learning_rate": 7.129004489567014e-07, + "loss": 0.8071, + "num_input_tokens_seen": 128962300, + "step": 6078 + }, + { + "epoch": 0.730956532195034, + "flos": 7737359493840.0, + "grad_norm": 6.62643992772308, + "learning_rate": 7.123043183778512e-07, + "loss": 0.7645, + "num_input_tokens_seen": 128979350, + "step": 6079 + }, + { + "epoch": 0.731076775085673, + "flos": 14460531709560.0, + "grad_norm": 3.048203555251126, + "learning_rate": 7.117083831414114e-07, + "loss": 0.6348, + "num_input_tokens_seen": 128998345, + "step": 6080 + }, + { + "epoch": 0.7311970179763122, + "flos": 14945069872800.0, + "grad_norm": 3.606718596850222, + "learning_rate": 7.11112643337787e-07, + "loss": 0.688, + "num_input_tokens_seen": 129017110, + "step": 6081 + }, + { + "epoch": 0.7313172608669513, + "flos": 13514436022680.0, + "grad_norm": 6.447724454004344, + "learning_rate": 7.10517099057349e-07, + "loss": 0.7548, + "num_input_tokens_seen": 129033780, + "step": 6082 + }, + { + "epoch": 0.7314375037575903, + "flos": 11788084231200.0, + "grad_norm": 5.483671050797939, + "learning_rate": 7.099217503904411e-07, + "loss": 0.569, + "num_input_tokens_seen": 129051355, + "step": 6083 + }, + { + "epoch": 0.7315577466482295, + "flos": 13110648943320.0, + "grad_norm": 12.607855609551097, + "learning_rate": 7.093265974273788e-07, + "loss": 0.8938, + "num_input_tokens_seen": 129068970, + "step": 6084 + }, + { + "epoch": 0.7316779895388685, + "flos": 13434381457920.0, + "grad_norm": 3.729812452494915, + "learning_rate": 7.087316402584447e-07, + "loss": 0.7119, + "num_input_tokens_seen": 129087515, + "step": 6085 + }, + { + "epoch": 0.7317982324295076, + "flos": 13080861412320.0, + "grad_norm": 4.889068453313629, + "learning_rate": 7.081368789738953e-07, + "loss": 0.8429, + "num_input_tokens_seen": 129104435, + "step": 6086 + }, + { + "epoch": 0.7319184753201466, + "flos": 19964301057360.0, + "grad_norm": 4.806527729715945, + "learning_rate": 7.075423136639537e-07, + "loss": 0.7665, + "num_input_tokens_seen": 129123410, + "step": 6087 + }, + { + "epoch": 0.7320387182107858, + "flos": 27527209459800.0, + "grad_norm": 4.713613177296231, + "learning_rate": 7.069479444188149e-07, + "loss": 0.7228, + "num_input_tokens_seen": 129143720, + "step": 6088 + }, + { + "epoch": 0.7321589611014249, + "flos": 13030119479160.0, + "grad_norm": 5.931246707059149, + "learning_rate": 7.063537713286453e-07, + "loss": 0.8083, + "num_input_tokens_seen": 129161120, + "step": 6089 + }, + { + "epoch": 0.7322792039920639, + "flos": 19128838965240.0, + "grad_norm": 8.560923633122096, + "learning_rate": 7.057597944835803e-07, + "loss": 0.7955, + "num_input_tokens_seen": 129180115, + "step": 6090 + }, + { + "epoch": 0.7323994468827031, + "flos": 18587226876840.0, + "grad_norm": 18.12077503289991, + "learning_rate": 7.051660139737253e-07, + "loss": 0.7325, + "num_input_tokens_seen": 129198055, + "step": 6091 + }, + { + "epoch": 0.7325196897733421, + "flos": 19726226920440.0, + "grad_norm": 6.433190245675978, + "learning_rate": 7.045724298891565e-07, + "loss": 0.7475, + "num_input_tokens_seen": 129217245, + "step": 6092 + }, + { + "epoch": 0.7326399326639812, + "flos": 18888960210600.0, + "grad_norm": 6.201490518388156, + "learning_rate": 7.039790423199192e-07, + "loss": 0.6827, + "num_input_tokens_seen": 129236605, + "step": 6093 + }, + { + "epoch": 0.7327601755546204, + "flos": 15378391203480.0, + "grad_norm": 12.213551383702603, + "learning_rate": 7.033858513560322e-07, + "loss": 0.7584, + "num_input_tokens_seen": 129252620, + "step": 6094 + }, + { + "epoch": 0.7328804184452594, + "flos": 11868233775840.0, + "grad_norm": 4.988995267210134, + "learning_rate": 7.027928570874794e-07, + "loss": 0.7452, + "num_input_tokens_seen": 129270530, + "step": 6095 + }, + { + "epoch": 0.7330006613358985, + "flos": 13029422960040.0, + "grad_norm": 3.6678695584048713, + "learning_rate": 7.022000596042194e-07, + "loss": 0.8344, + "num_input_tokens_seen": 129287350, + "step": 6096 + }, + { + "epoch": 0.7331209042265376, + "flos": 16458797643840.0, + "grad_norm": 8.545820711654452, + "learning_rate": 7.016074589961784e-07, + "loss": 0.7989, + "num_input_tokens_seen": 129305635, + "step": 6097 + }, + { + "epoch": 0.7332411471171767, + "flos": 24289187794680.0, + "grad_norm": 6.659439683776366, + "learning_rate": 7.01015055353253e-07, + "loss": 0.6564, + "num_input_tokens_seen": 129327780, + "step": 6098 + }, + { + "epoch": 0.7333613900078157, + "flos": 16643881990440.0, + "grad_norm": 9.28858775857924, + "learning_rate": 7.004228487653123e-07, + "loss": 0.7566, + "num_input_tokens_seen": 129348305, + "step": 6099 + }, + { + "epoch": 0.7334816328984549, + "flos": 16350791825760.0, + "grad_norm": 6.985405830930851, + "learning_rate": 6.998308393221906e-07, + "loss": 0.769, + "num_input_tokens_seen": 129366430, + "step": 6100 + }, + { + "epoch": 0.733601875789094, + "flos": 15158643692040.0, + "grad_norm": 13.107381566815164, + "learning_rate": 6.992390271136977e-07, + "loss": 0.6971, + "num_input_tokens_seen": 129381860, + "step": 6101 + }, + { + "epoch": 0.733722118679733, + "flos": 16512199013640.0, + "grad_norm": 6.939007019769482, + "learning_rate": 6.986474122296094e-07, + "loss": 0.8467, + "num_input_tokens_seen": 129400695, + "step": 6102 + }, + { + "epoch": 0.7338423615703722, + "flos": 14676828285360.0, + "grad_norm": 4.27987203210634, + "learning_rate": 6.980559947596751e-07, + "loss": 0.7068, + "num_input_tokens_seen": 129418955, + "step": 6103 + }, + { + "epoch": 0.7339626044610112, + "flos": 15863245966320.0, + "grad_norm": 3.9961296669100093, + "learning_rate": 6.974647747936109e-07, + "loss": 0.7424, + "num_input_tokens_seen": 129437060, + "step": 6104 + }, + { + "epoch": 0.7340828473516503, + "flos": 11112731268600.0, + "grad_norm": 5.039783514913837, + "learning_rate": 6.968737524211039e-07, + "loss": 0.8101, + "num_input_tokens_seen": 129453590, + "step": 6105 + }, + { + "epoch": 0.7342030902422895, + "flos": 16783005057840.0, + "grad_norm": 5.4359723279977095, + "learning_rate": 6.962829277318132e-07, + "loss": 0.7698, + "num_input_tokens_seen": 129472905, + "step": 6106 + }, + { + "epoch": 0.7343233331329285, + "flos": 18942203280600.0, + "grad_norm": 4.26507743702541, + "learning_rate": 6.956923008153652e-07, + "loss": 0.8088, + "num_input_tokens_seen": 129492390, + "step": 6107 + }, + { + "epoch": 0.7344435760235676, + "flos": 13488890926320.0, + "grad_norm": 4.961310809650124, + "learning_rate": 6.951018717613593e-07, + "loss": 0.8256, + "num_input_tokens_seen": 129511125, + "step": 6108 + }, + { + "epoch": 0.7345638189142067, + "flos": 12868047432120.0, + "grad_norm": 3.6185558523959775, + "learning_rate": 6.945116406593614e-07, + "loss": 0.7536, + "num_input_tokens_seen": 129529700, + "step": 6109 + }, + { + "epoch": 0.7346840618048458, + "flos": 14806200085080.0, + "grad_norm": 12.978418482753458, + "learning_rate": 6.939216075989089e-07, + "loss": 0.7137, + "num_input_tokens_seen": 129547350, + "step": 6110 + }, + { + "epoch": 0.7348043046954849, + "flos": 21292279622640.0, + "grad_norm": 3.4852012031082884, + "learning_rate": 6.933317726695109e-07, + "loss": 0.6477, + "num_input_tokens_seen": 129568300, + "step": 6111 + }, + { + "epoch": 0.734924547586124, + "flos": 13083425869080.0, + "grad_norm": 6.54963629491769, + "learning_rate": 6.92742135960644e-07, + "loss": 0.7631, + "num_input_tokens_seen": 129585720, + "step": 6112 + }, + { + "epoch": 0.7350447904767631, + "flos": 46879821242760.0, + "grad_norm": 0.8816349987683759, + "learning_rate": 6.921526975617556e-07, + "loss": 0.5898, + "num_input_tokens_seen": 129644900, + "step": 6113 + }, + { + "epoch": 0.7351650333674021, + "flos": 15783824600760.0, + "grad_norm": 4.561093884486013, + "learning_rate": 6.915634575622631e-07, + "loss": 0.7286, + "num_input_tokens_seen": 129663135, + "step": 6114 + }, + { + "epoch": 0.7352852762580413, + "flos": 13272214431000.0, + "grad_norm": 4.180769792702094, + "learning_rate": 6.909744160515532e-07, + "loss": 0.7011, + "num_input_tokens_seen": 129680995, + "step": 6115 + }, + { + "epoch": 0.7354055191486804, + "flos": 28609135578240.0, + "grad_norm": 4.759606134484231, + "learning_rate": 6.903855731189849e-07, + "loss": 0.6522, + "num_input_tokens_seen": 129703350, + "step": 6116 + }, + { + "epoch": 0.7355257620393194, + "flos": 11868898635000.0, + "grad_norm": 4.052928270964103, + "learning_rate": 6.897969288538825e-07, + "loss": 0.7903, + "num_input_tokens_seen": 129721015, + "step": 6117 + }, + { + "epoch": 0.7356460049299585, + "flos": 13218971361000.0, + "grad_norm": 3.9480761224892467, + "learning_rate": 6.892084833455452e-07, + "loss": 0.7943, + "num_input_tokens_seen": 129740305, + "step": 6118 + }, + { + "epoch": 0.7357662478205976, + "flos": 15595320978480.0, + "grad_norm": 3.210076655248027, + "learning_rate": 6.886202366832384e-07, + "loss": 0.8285, + "num_input_tokens_seen": 129761710, + "step": 6119 + }, + { + "epoch": 0.7358864907112367, + "flos": 10356848841840.0, + "grad_norm": 24.647481657377924, + "learning_rate": 6.880321889561987e-07, + "loss": 0.7157, + "num_input_tokens_seen": 129779405, + "step": 6120 + }, + { + "epoch": 0.7360067336018757, + "flos": 16323822031200.0, + "grad_norm": 4.8905563593425905, + "learning_rate": 6.874443402536338e-07, + "loss": 0.6319, + "num_input_tokens_seen": 129798215, + "step": 6121 + }, + { + "epoch": 0.7361269764925149, + "flos": 18724956906000.0, + "grad_norm": 6.344574672671101, + "learning_rate": 6.868566906647177e-07, + "loss": 0.8029, + "num_input_tokens_seen": 129818885, + "step": 6122 + }, + { + "epoch": 0.736247219383154, + "flos": 14893219841040.0, + "grad_norm": 6.761698158096173, + "learning_rate": 6.862692402785984e-07, + "loss": 0.8131, + "num_input_tokens_seen": 129838855, + "step": 6123 + }, + { + "epoch": 0.736367462273793, + "flos": 51876356727360.0, + "grad_norm": 0.6970102392782709, + "learning_rate": 6.856819891843899e-07, + "loss": 0.5066, + "num_input_tokens_seen": 129903280, + "step": 6124 + }, + { + "epoch": 0.7364877051644322, + "flos": 16399159261920.0, + "grad_norm": 4.739890332462296, + "learning_rate": 6.8509493747118e-07, + "loss": 0.7057, + "num_input_tokens_seen": 129921810, + "step": 6125 + }, + { + "epoch": 0.7366079480550712, + "flos": 8790384560160.0, + "grad_norm": 5.742130699655297, + "learning_rate": 6.845080852280221e-07, + "loss": 0.8543, + "num_input_tokens_seen": 129938600, + "step": 6126 + }, + { + "epoch": 0.7367281909457103, + "flos": 10950722541480.0, + "grad_norm": 5.662553004675693, + "learning_rate": 6.839214325439409e-07, + "loss": 0.7305, + "num_input_tokens_seen": 129956015, + "step": 6127 + }, + { + "epoch": 0.7368484338363495, + "flos": 17212273913640.0, + "grad_norm": 5.464868879499049, + "learning_rate": 6.833349795079327e-07, + "loss": 0.7024, + "num_input_tokens_seen": 129974845, + "step": 6128 + }, + { + "epoch": 0.7369686767269885, + "flos": 20104215623760.0, + "grad_norm": 3.5894088398651594, + "learning_rate": 6.827487262089613e-07, + "loss": 0.6636, + "num_input_tokens_seen": 129995070, + "step": 6129 + }, + { + "epoch": 0.7370889196176276, + "flos": 51840807083640.0, + "grad_norm": 0.8725933287910893, + "learning_rate": 6.821626727359606e-07, + "loss": 0.5913, + "num_input_tokens_seen": 130060350, + "step": 6130 + }, + { + "epoch": 0.7372091625082667, + "flos": 13164588532440.0, + "grad_norm": 14.951625896315068, + "learning_rate": 6.815768191778348e-07, + "loss": 0.7492, + "num_input_tokens_seen": 130078150, + "step": 6131 + }, + { + "epoch": 0.7373294053989058, + "flos": 24772427899560.0, + "grad_norm": 4.58233636358924, + "learning_rate": 6.809911656234569e-07, + "loss": 0.7171, + "num_input_tokens_seen": 130099845, + "step": 6132 + }, + { + "epoch": 0.7374496482895448, + "flos": 15729188492520.0, + "grad_norm": 4.197605295047724, + "learning_rate": 6.804057121616707e-07, + "loss": 0.7652, + "num_input_tokens_seen": 130117770, + "step": 6133 + }, + { + "epoch": 0.737569891180184, + "flos": 18268180036320.0, + "grad_norm": 3.810801768808604, + "learning_rate": 6.798204588812888e-07, + "loss": 0.7046, + "num_input_tokens_seen": 130136905, + "step": 6134 + }, + { + "epoch": 0.7376901340708231, + "flos": 15105843861480.0, + "grad_norm": 4.559919071483239, + "learning_rate": 6.792354058710937e-07, + "loss": 0.7418, + "num_input_tokens_seen": 130154095, + "step": 6135 + }, + { + "epoch": 0.7378103769614621, + "flos": 17431071626280.0, + "grad_norm": 4.105220652886604, + "learning_rate": 6.786505532198374e-07, + "loss": 0.6426, + "num_input_tokens_seen": 130172760, + "step": 6136 + }, + { + "epoch": 0.7379306198521013, + "flos": 16270262361600.0, + "grad_norm": 5.994068392669214, + "learning_rate": 6.780659010162411e-07, + "loss": 0.8399, + "num_input_tokens_seen": 130191430, + "step": 6137 + }, + { + "epoch": 0.7380508627427403, + "flos": 10842463443720.0, + "grad_norm": 13.58302715557119, + "learning_rate": 6.774814493489975e-07, + "loss": 0.8236, + "num_input_tokens_seen": 130208825, + "step": 6138 + }, + { + "epoch": 0.7381711056333794, + "flos": 15861631308360.0, + "grad_norm": 3.8991793918476705, + "learning_rate": 6.768971983067655e-07, + "loss": 0.6587, + "num_input_tokens_seen": 130228875, + "step": 6139 + }, + { + "epoch": 0.7382913485240186, + "flos": 38600121481440.0, + "grad_norm": 1.042031828136115, + "learning_rate": 6.763131479781772e-07, + "loss": 0.6956, + "num_input_tokens_seen": 130278355, + "step": 6140 + }, + { + "epoch": 0.7384115914146576, + "flos": 15945738348000.0, + "grad_norm": 3.5800815879859136, + "learning_rate": 6.757292984518316e-07, + "loss": 0.7572, + "num_input_tokens_seen": 130297475, + "step": 6141 + }, + { + "epoch": 0.7385318343052967, + "flos": 45328106011080.0, + "grad_norm": 0.8365315260797803, + "learning_rate": 6.751456498162981e-07, + "loss": 0.6059, + "num_input_tokens_seen": 130356230, + "step": 6142 + }, + { + "epoch": 0.7386520771959358, + "flos": 12403703832000.0, + "grad_norm": 6.428572306557247, + "learning_rate": 6.745622021601174e-07, + "loss": 0.8412, + "num_input_tokens_seen": 130372975, + "step": 6143 + }, + { + "epoch": 0.7387723200865749, + "flos": 13704807582600.0, + "grad_norm": 5.360581153849537, + "learning_rate": 6.739789555717954e-07, + "loss": 0.6796, + "num_input_tokens_seen": 130389670, + "step": 6144 + }, + { + "epoch": 0.738892562977214, + "flos": 16483139661720.0, + "grad_norm": 8.85137675293303, + "learning_rate": 6.733959101398124e-07, + "loss": 0.7702, + "num_input_tokens_seen": 130407520, + "step": 6145 + }, + { + "epoch": 0.7390128058678531, + "flos": 15724914397920.0, + "grad_norm": 6.730288511443296, + "learning_rate": 6.728130659526143e-07, + "loss": 0.8043, + "num_input_tokens_seen": 130425050, + "step": 6146 + }, + { + "epoch": 0.7391330487584922, + "flos": 18889308470160.0, + "grad_norm": 4.534010053127941, + "learning_rate": 6.7223042309862e-07, + "loss": 0.6861, + "num_input_tokens_seen": 130444970, + "step": 6147 + }, + { + "epoch": 0.7392532916491312, + "flos": 20806506720960.0, + "grad_norm": 3.5893005690711215, + "learning_rate": 6.716479816662144e-07, + "loss": 0.7202, + "num_input_tokens_seen": 130466420, + "step": 6148 + }, + { + "epoch": 0.7393735345397703, + "flos": 17268176420280.0, + "grad_norm": 3.811141725818191, + "learning_rate": 6.710657417437531e-07, + "loss": 0.7135, + "num_input_tokens_seen": 130485845, + "step": 6149 + }, + { + "epoch": 0.7394937774304094, + "flos": 14595855581760.0, + "grad_norm": 20.26500005059909, + "learning_rate": 6.704837034195628e-07, + "loss": 0.7643, + "num_input_tokens_seen": 130504030, + "step": 6150 + }, + { + "epoch": 0.7396140203210485, + "flos": 17188596754920.0, + "grad_norm": 5.240774131841265, + "learning_rate": 6.699018667819376e-07, + "loss": 0.8358, + "num_input_tokens_seen": 130523150, + "step": 6151 + }, + { + "epoch": 0.7397342632116876, + "flos": 18725685085080.0, + "grad_norm": 2.8816319144252924, + "learning_rate": 6.693202319191415e-07, + "loss": 0.7128, + "num_input_tokens_seen": 130544605, + "step": 6152 + }, + { + "epoch": 0.7398545061023267, + "flos": 18133647663120.0, + "grad_norm": 4.068318118189754, + "learning_rate": 6.687387989194084e-07, + "loss": 0.7214, + "num_input_tokens_seen": 130563840, + "step": 6153 + }, + { + "epoch": 0.7399747489929658, + "flos": 12030654082440.0, + "grad_norm": 4.885294095950689, + "learning_rate": 6.681575678709404e-07, + "loss": 0.7793, + "num_input_tokens_seen": 130582250, + "step": 6154 + }, + { + "epoch": 0.7400949918836048, + "flos": 17646545043120.0, + "grad_norm": 4.243748948289519, + "learning_rate": 6.67576538861911e-07, + "loss": 0.6897, + "num_input_tokens_seen": 130600545, + "step": 6155 + }, + { + "epoch": 0.740215234774244, + "flos": 15948999323880.0, + "grad_norm": 11.787732513820579, + "learning_rate": 6.669957119804612e-07, + "loss": 0.8022, + "num_input_tokens_seen": 130621900, + "step": 6156 + }, + { + "epoch": 0.7403354776648831, + "flos": 13676856329280.0, + "grad_norm": 5.30102520509926, + "learning_rate": 6.66415087314702e-07, + "loss": 0.685, + "num_input_tokens_seen": 130636575, + "step": 6157 + }, + { + "epoch": 0.7404557205555221, + "flos": 12327353482560.0, + "grad_norm": 4.51313454348447, + "learning_rate": 6.65834664952714e-07, + "loss": 0.7064, + "num_input_tokens_seen": 130653745, + "step": 6158 + }, + { + "epoch": 0.7405759634461613, + "flos": 15513145196400.0, + "grad_norm": 3.647998595538622, + "learning_rate": 6.652544449825457e-07, + "loss": 0.7409, + "num_input_tokens_seen": 130673720, + "step": 6159 + }, + { + "epoch": 0.7406962063368003, + "flos": 14969063631120.0, + "grad_norm": 6.981203455757025, + "learning_rate": 6.646744274922182e-07, + "loss": 0.7464, + "num_input_tokens_seen": 130691885, + "step": 6160 + }, + { + "epoch": 0.7408164492274394, + "flos": 14460468389640.0, + "grad_norm": 16.4509621382288, + "learning_rate": 6.640946125697171e-07, + "loss": 0.7316, + "num_input_tokens_seen": 130709135, + "step": 6161 + }, + { + "epoch": 0.7409366921180786, + "flos": 21426622036080.0, + "grad_norm": 4.349249343120352, + "learning_rate": 6.635150003030017e-07, + "loss": 0.7482, + "num_input_tokens_seen": 130727380, + "step": 6162 + }, + { + "epoch": 0.7410569350087176, + "flos": 16782910077960.0, + "grad_norm": 6.43943239115558, + "learning_rate": 6.629355907799981e-07, + "loss": 0.8425, + "num_input_tokens_seen": 130746905, + "step": 6163 + }, + { + "epoch": 0.7411771778993567, + "flos": 22340302415280.0, + "grad_norm": 3.20025193295578, + "learning_rate": 6.623563840886015e-07, + "loss": 0.6843, + "num_input_tokens_seen": 130767550, + "step": 6164 + }, + { + "epoch": 0.7412974207899958, + "flos": 15000877399560.0, + "grad_norm": 3.7079179223944356, + "learning_rate": 6.617773803166795e-07, + "loss": 0.6862, + "num_input_tokens_seen": 130785595, + "step": 6165 + }, + { + "epoch": 0.7414176636806349, + "flos": 16161749984160.0, + "grad_norm": 6.264599038313613, + "learning_rate": 6.611985795520634e-07, + "loss": 0.808, + "num_input_tokens_seen": 130803860, + "step": 6166 + }, + { + "epoch": 0.7415379065712739, + "flos": 18429682204080.0, + "grad_norm": 3.8240544066608857, + "learning_rate": 6.606199818825588e-07, + "loss": 0.7587, + "num_input_tokens_seen": 130824035, + "step": 6167 + }, + { + "epoch": 0.7416581494619131, + "flos": 12299148949560.0, + "grad_norm": 4.852942032567457, + "learning_rate": 6.600415873959377e-07, + "loss": 0.7988, + "num_input_tokens_seen": 130841630, + "step": 6168 + }, + { + "epoch": 0.7417783923525522, + "flos": 20859781450920.0, + "grad_norm": 8.621470738245865, + "learning_rate": 6.594633961799437e-07, + "loss": 0.6292, + "num_input_tokens_seen": 130860390, + "step": 6169 + }, + { + "epoch": 0.7418986352431912, + "flos": 14676923265240.0, + "grad_norm": 3.448763340989538, + "learning_rate": 6.588854083222857e-07, + "loss": 0.8042, + "num_input_tokens_seen": 130879545, + "step": 6170 + }, + { + "epoch": 0.7420188781338304, + "flos": 13326438959760.0, + "grad_norm": 16.102437165411192, + "learning_rate": 6.583076239106444e-07, + "loss": 0.7781, + "num_input_tokens_seen": 130897770, + "step": 6171 + }, + { + "epoch": 0.7421391210244694, + "flos": 10005355033680.0, + "grad_norm": 4.950918963893081, + "learning_rate": 6.577300430326707e-07, + "loss": 0.7414, + "num_input_tokens_seen": 130912435, + "step": 6172 + }, + { + "epoch": 0.7422593639151085, + "flos": 11625948864240.0, + "grad_norm": 4.1892580355946025, + "learning_rate": 6.571526657759821e-07, + "loss": 0.7122, + "num_input_tokens_seen": 130927895, + "step": 6173 + }, + { + "epoch": 0.7423796068057477, + "flos": 22099378881960.0, + "grad_norm": 6.608301675278917, + "learning_rate": 6.565754922281663e-07, + "loss": 0.6975, + "num_input_tokens_seen": 130949860, + "step": 6174 + }, + { + "epoch": 0.7424998496963867, + "flos": 15000750759720.0, + "grad_norm": 6.144832654618959, + "learning_rate": 6.559985224767801e-07, + "loss": 0.7789, + "num_input_tokens_seen": 130967455, + "step": 6175 + }, + { + "epoch": 0.7426200925870258, + "flos": 15999551297280.0, + "grad_norm": 6.473506339865235, + "learning_rate": 6.55421756609349e-07, + "loss": 0.7423, + "num_input_tokens_seen": 130985430, + "step": 6176 + }, + { + "epoch": 0.7427403354776649, + "flos": 19374574812480.0, + "grad_norm": 3.3033614868853074, + "learning_rate": 6.54845194713369e-07, + "loss": 0.7786, + "num_input_tokens_seen": 131006100, + "step": 6177 + }, + { + "epoch": 0.742860578368304, + "flos": 14538844976520.0, + "grad_norm": 5.667420542815829, + "learning_rate": 6.542688368763034e-07, + "loss": 0.7905, + "num_input_tokens_seen": 131024225, + "step": 6178 + }, + { + "epoch": 0.742980821258943, + "flos": 18186954053040.0, + "grad_norm": 4.195114273133503, + "learning_rate": 6.536926831855854e-07, + "loss": 0.7608, + "num_input_tokens_seen": 131043110, + "step": 6179 + }, + { + "epoch": 0.7431010641495821, + "flos": 18484065032640.0, + "grad_norm": 6.424906439786478, + "learning_rate": 6.531167337286165e-07, + "loss": 0.7147, + "num_input_tokens_seen": 131062850, + "step": 6180 + }, + { + "epoch": 0.7432213070402213, + "flos": 15918768553440.0, + "grad_norm": 3.681355934367516, + "learning_rate": 6.52540988592768e-07, + "loss": 0.7852, + "num_input_tokens_seen": 131083590, + "step": 6181 + }, + { + "epoch": 0.7433415499308603, + "flos": 10761617379960.0, + "grad_norm": 5.598581113897663, + "learning_rate": 6.519654478653814e-07, + "loss": 0.8188, + "num_input_tokens_seen": 131101675, + "step": 6182 + }, + { + "epoch": 0.7434617928214994, + "flos": 49518650334960.0, + "grad_norm": 1.4870532318571623, + "learning_rate": 6.51390111633763e-07, + "loss": 0.5694, + "num_input_tokens_seen": 131166670, + "step": 6183 + }, + { + "epoch": 0.7435820357121385, + "flos": 20073889873440.0, + "grad_norm": 3.557334501232926, + "learning_rate": 6.508149799851932e-07, + "loss": 0.7453, + "num_input_tokens_seen": 131188055, + "step": 6184 + }, + { + "epoch": 0.7437022786027776, + "flos": 17565572339520.0, + "grad_norm": 11.368106304146826, + "learning_rate": 6.502400530069183e-07, + "loss": 0.6111, + "num_input_tokens_seen": 131207660, + "step": 6185 + }, + { + "epoch": 0.7438225214934167, + "flos": 15995308862640.0, + "grad_norm": 3.0572311854621503, + "learning_rate": 6.496653307861535e-07, + "loss": 0.6702, + "num_input_tokens_seen": 131228050, + "step": 6186 + }, + { + "epoch": 0.7439427643840558, + "flos": 14784802443480.0, + "grad_norm": 7.67950619302129, + "learning_rate": 6.490908134100857e-07, + "loss": 0.6383, + "num_input_tokens_seen": 131246235, + "step": 6187 + }, + { + "epoch": 0.7440630072746949, + "flos": 15243447250800.0, + "grad_norm": 4.892897982620827, + "learning_rate": 6.48516500965866e-07, + "loss": 0.6795, + "num_input_tokens_seen": 131265890, + "step": 6188 + }, + { + "epoch": 0.7441832501653339, + "flos": 19427469622920.0, + "grad_norm": 4.016859762739499, + "learning_rate": 6.479423935406192e-07, + "loss": 0.8045, + "num_input_tokens_seen": 131285595, + "step": 6189 + }, + { + "epoch": 0.7443034930559731, + "flos": 50588007449280.0, + "grad_norm": 0.9253679471600611, + "learning_rate": 6.473684912214357e-07, + "loss": 0.702, + "num_input_tokens_seen": 131348875, + "step": 6190 + }, + { + "epoch": 0.7444237359466122, + "flos": 13618231066080.0, + "grad_norm": 6.193812069548192, + "learning_rate": 6.467947940953778e-07, + "loss": 0.6914, + "num_input_tokens_seen": 131367120, + "step": 6191 + }, + { + "epoch": 0.7445439788372512, + "flos": 16699309597680.0, + "grad_norm": 4.83646124990938, + "learning_rate": 6.462213022494732e-07, + "loss": 0.7111, + "num_input_tokens_seen": 131386085, + "step": 6192 + }, + { + "epoch": 0.7446642217278904, + "flos": 49437266051880.0, + "grad_norm": 0.8017125366660102, + "learning_rate": 6.456480157707201e-07, + "loss": 0.6436, + "num_input_tokens_seen": 131450580, + "step": 6193 + }, + { + "epoch": 0.7447844646185294, + "flos": 12702809389080.0, + "grad_norm": 11.66542097434932, + "learning_rate": 6.450749347460866e-07, + "loss": 0.8342, + "num_input_tokens_seen": 131467275, + "step": 6194 + }, + { + "epoch": 0.7449047075091685, + "flos": 19510816823520.0, + "grad_norm": 4.7570190770342275, + "learning_rate": 6.445020592625083e-07, + "loss": 0.7791, + "num_input_tokens_seen": 131487645, + "step": 6195 + }, + { + "epoch": 0.7450249503998077, + "flos": 10299838236600.0, + "grad_norm": 6.4303463196225445, + "learning_rate": 6.4392938940689e-07, + "loss": 0.7914, + "num_input_tokens_seen": 131502780, + "step": 6196 + }, + { + "epoch": 0.7451451932904467, + "flos": 14322991640160.0, + "grad_norm": 11.213455183324964, + "learning_rate": 6.433569252661049e-07, + "loss": 0.682, + "num_input_tokens_seen": 131520500, + "step": 6197 + }, + { + "epoch": 0.7452654361810858, + "flos": 9060652385040.0, + "grad_norm": 5.145061955055771, + "learning_rate": 6.427846669269952e-07, + "loss": 0.6878, + "num_input_tokens_seen": 131537840, + "step": 6198 + }, + { + "epoch": 0.7453856790717249, + "flos": 16188878078520.0, + "grad_norm": 5.491706664306824, + "learning_rate": 6.422126144763729e-07, + "loss": 0.8, + "num_input_tokens_seen": 131556950, + "step": 6199 + }, + { + "epoch": 0.745505921962364, + "flos": 14622350476920.0, + "grad_norm": 4.457482491467932, + "learning_rate": 6.416407680010174e-07, + "loss": 0.7521, + "num_input_tokens_seen": 131571030, + "step": 6200 + }, + { + "epoch": 0.745626164853003, + "flos": 18075972198720.0, + "grad_norm": 5.447135368422914, + "learning_rate": 6.410691275876774e-07, + "loss": 0.7931, + "num_input_tokens_seen": 131590170, + "step": 6201 + }, + { + "epoch": 0.7457464077436422, + "flos": 10572955457880.0, + "grad_norm": 7.042336095406414, + "learning_rate": 6.404976933230704e-07, + "loss": 0.7403, + "num_input_tokens_seen": 131606410, + "step": 6202 + }, + { + "epoch": 0.7458666506342813, + "flos": 24991067312400.0, + "grad_norm": 4.140367953401096, + "learning_rate": 6.399264652938813e-07, + "loss": 0.7138, + "num_input_tokens_seen": 131627035, + "step": 6203 + }, + { + "epoch": 0.7459868935249203, + "flos": 17781140736240.0, + "grad_norm": 4.757793457794919, + "learning_rate": 6.393554435867679e-07, + "loss": 0.728, + "num_input_tokens_seen": 131647605, + "step": 6204 + }, + { + "epoch": 0.7461071364155595, + "flos": 16027185951000.0, + "grad_norm": 4.129268864901216, + "learning_rate": 6.387846282883502e-07, + "loss": 0.8114, + "num_input_tokens_seen": 131663855, + "step": 6205 + }, + { + "epoch": 0.7462273793061985, + "flos": 16752552667680.0, + "grad_norm": 3.5071614432265563, + "learning_rate": 6.38214019485223e-07, + "loss": 0.7562, + "num_input_tokens_seen": 131682400, + "step": 6206 + }, + { + "epoch": 0.7463476221968376, + "flos": 14590283428800.0, + "grad_norm": 3.532575375454027, + "learning_rate": 6.376436172639461e-07, + "loss": 0.7, + "num_input_tokens_seen": 131699965, + "step": 6207 + }, + { + "epoch": 0.7464678650874768, + "flos": 12273350573520.0, + "grad_norm": 6.20524687713167, + "learning_rate": 6.370734217110487e-07, + "loss": 0.6288, + "num_input_tokens_seen": 131718430, + "step": 6208 + }, + { + "epoch": 0.7465881079781158, + "flos": 35410431101160.0, + "grad_norm": 5.08272534262568, + "learning_rate": 6.36503432913031e-07, + "loss": 0.6302, + "num_input_tokens_seen": 131741295, + "step": 6209 + }, + { + "epoch": 0.7467083508687549, + "flos": 14374968311760.0, + "grad_norm": 6.159656801943158, + "learning_rate": 6.359336509563569e-07, + "loss": 0.6703, + "num_input_tokens_seen": 131757035, + "step": 6210 + }, + { + "epoch": 0.7468285937593939, + "flos": 13057057613760.0, + "grad_norm": 5.036163395079965, + "learning_rate": 6.353640759274641e-07, + "loss": 0.7925, + "num_input_tokens_seen": 131775645, + "step": 6211 + }, + { + "epoch": 0.7469488366500331, + "flos": 16938333533400.0, + "grad_norm": 4.383520929265045, + "learning_rate": 6.347947079127556e-07, + "loss": 0.7266, + "num_input_tokens_seen": 131793265, + "step": 6212 + }, + { + "epoch": 0.7470690795406721, + "flos": 12165503055240.0, + "grad_norm": 5.056132326772605, + "learning_rate": 6.342255469986053e-07, + "loss": 0.7476, + "num_input_tokens_seen": 131811730, + "step": 6213 + }, + { + "epoch": 0.7471893224313112, + "flos": 18457538477520.0, + "grad_norm": 12.654816421509613, + "learning_rate": 6.336565932713533e-07, + "loss": 0.7539, + "num_input_tokens_seen": 131830875, + "step": 6214 + }, + { + "epoch": 0.7473095653219504, + "flos": 16483741200960.0, + "grad_norm": 4.693616817936693, + "learning_rate": 6.330878468173088e-07, + "loss": 0.7646, + "num_input_tokens_seen": 131850660, + "step": 6215 + }, + { + "epoch": 0.7474298082125894, + "flos": 13218749741280.0, + "grad_norm": 4.067793932741437, + "learning_rate": 6.32519307722752e-07, + "loss": 0.7169, + "num_input_tokens_seen": 131868275, + "step": 6216 + }, + { + "epoch": 0.7475500511032285, + "flos": 45767569374000.0, + "grad_norm": 0.7928183652595443, + "learning_rate": 6.31950976073929e-07, + "loss": 0.5702, + "num_input_tokens_seen": 131922085, + "step": 6217 + }, + { + "epoch": 0.7476702939938676, + "flos": 12975704990640.0, + "grad_norm": 4.0954060007866495, + "learning_rate": 6.31382851957055e-07, + "loss": 0.7842, + "num_input_tokens_seen": 131938625, + "step": 6218 + }, + { + "epoch": 0.7477905368845067, + "flos": 20481634447800.0, + "grad_norm": 5.7092279257620095, + "learning_rate": 6.308149354583143e-07, + "loss": 0.689, + "num_input_tokens_seen": 131957750, + "step": 6219 + }, + { + "epoch": 0.7479107797751458, + "flos": 19699035506160.0, + "grad_norm": 8.788294671321648, + "learning_rate": 6.302472266638586e-07, + "loss": 0.7946, + "num_input_tokens_seen": 131978010, + "step": 6220 + }, + { + "epoch": 0.7480310226657849, + "flos": 24932790308760.0, + "grad_norm": 11.197655868450422, + "learning_rate": 6.296797256598101e-07, + "loss": 0.6917, + "num_input_tokens_seen": 131999210, + "step": 6221 + }, + { + "epoch": 0.748151265556424, + "flos": 18186384173760.0, + "grad_norm": 3.8319513621737724, + "learning_rate": 6.291124325322576e-07, + "loss": 0.7983, + "num_input_tokens_seen": 132019055, + "step": 6222 + }, + { + "epoch": 0.748271508447063, + "flos": 28231178534880.0, + "grad_norm": 3.1369756496020638, + "learning_rate": 6.285453473672595e-07, + "loss": 0.6142, + "num_input_tokens_seen": 132041345, + "step": 6223 + }, + { + "epoch": 0.7483917513377022, + "flos": 15755335128120.0, + "grad_norm": 5.772875850133509, + "learning_rate": 6.279784702508415e-07, + "loss": 0.7262, + "num_input_tokens_seen": 132061815, + "step": 6224 + }, + { + "epoch": 0.7485119942283412, + "flos": 45937588071000.0, + "grad_norm": 0.8358344255163246, + "learning_rate": 6.274118012689979e-07, + "loss": 0.6365, + "num_input_tokens_seen": 132123435, + "step": 6225 + }, + { + "epoch": 0.7486322371189803, + "flos": 21969342223080.0, + "grad_norm": 3.351218466007428, + "learning_rate": 6.268453405076943e-07, + "loss": 0.669, + "num_input_tokens_seen": 132145550, + "step": 6226 + }, + { + "epoch": 0.7487524800096195, + "flos": 13836490559400.0, + "grad_norm": 5.320007763620319, + "learning_rate": 6.262790880528592e-07, + "loss": 0.8116, + "num_input_tokens_seen": 132162890, + "step": 6227 + }, + { + "epoch": 0.7488727229002585, + "flos": 13650456414000.0, + "grad_norm": 6.82196423432462, + "learning_rate": 6.257130439903951e-07, + "loss": 0.7736, + "num_input_tokens_seen": 132179105, + "step": 6228 + }, + { + "epoch": 0.7489929657908976, + "flos": 17296064353680.0, + "grad_norm": 5.91040280591137, + "learning_rate": 6.251472084061695e-07, + "loss": 0.7874, + "num_input_tokens_seen": 132197745, + "step": 6229 + }, + { + "epoch": 0.7491132086815367, + "flos": 15021705161880.0, + "grad_norm": 3.709422748004477, + "learning_rate": 6.245815813860191e-07, + "loss": 0.8849, + "num_input_tokens_seen": 132212975, + "step": 6230 + }, + { + "epoch": 0.7492334515721758, + "flos": 16836944646960.0, + "grad_norm": 6.304279980730508, + "learning_rate": 6.240161630157495e-07, + "loss": 0.6717, + "num_input_tokens_seen": 132232050, + "step": 6231 + }, + { + "epoch": 0.7493536944628149, + "flos": 11949364779240.0, + "grad_norm": 4.559041423765886, + "learning_rate": 6.23450953381133e-07, + "loss": 0.6753, + "num_input_tokens_seen": 132249860, + "step": 6232 + }, + { + "epoch": 0.749473937353454, + "flos": 11164391340600.0, + "grad_norm": 7.2255190864127705, + "learning_rate": 6.228859525679131e-07, + "loss": 0.6693, + "num_input_tokens_seen": 132263995, + "step": 6233 + }, + { + "epoch": 0.7495941802440931, + "flos": 13838010237480.0, + "grad_norm": 7.423089949899359, + "learning_rate": 6.223211606617986e-07, + "loss": 0.7898, + "num_input_tokens_seen": 132282135, + "step": 6234 + }, + { + "epoch": 0.7497144231347321, + "flos": 16459145903400.0, + "grad_norm": 5.048004685328708, + "learning_rate": 6.217565777484701e-07, + "loss": 0.8189, + "num_input_tokens_seen": 132300950, + "step": 6235 + }, + { + "epoch": 0.7498346660253713, + "flos": 17754994100640.0, + "grad_norm": 3.738892699553494, + "learning_rate": 6.211922039135722e-07, + "loss": 0.7852, + "num_input_tokens_seen": 132320815, + "step": 6236 + }, + { + "epoch": 0.7499549089160104, + "flos": 17860910361360.0, + "grad_norm": 4.413502669490371, + "learning_rate": 6.206280392427201e-07, + "loss": 0.7923, + "num_input_tokens_seen": 132340120, + "step": 6237 + }, + { + "epoch": 0.7500751518066494, + "flos": 25017752167320.0, + "grad_norm": 3.8450550872071037, + "learning_rate": 6.200640838214983e-07, + "loss": 0.7189, + "num_input_tokens_seen": 132362615, + "step": 6238 + }, + { + "epoch": 0.7501953946972886, + "flos": 13758683851800.0, + "grad_norm": 4.875493733761661, + "learning_rate": 6.195003377354578e-07, + "loss": 0.6573, + "num_input_tokens_seen": 132381605, + "step": 6239 + }, + { + "epoch": 0.7503156375879276, + "flos": 14803952227920.0, + "grad_norm": 6.229846963977533, + "learning_rate": 6.189368010701183e-07, + "loss": 0.7048, + "num_input_tokens_seen": 132398385, + "step": 6240 + }, + { + "epoch": 0.7504358804785667, + "flos": 9789280077600.0, + "grad_norm": 8.79077544071745, + "learning_rate": 6.183734739109683e-07, + "loss": 0.7457, + "num_input_tokens_seen": 132415925, + "step": 6241 + }, + { + "epoch": 0.7505561233692057, + "flos": 21616328736840.0, + "grad_norm": 5.721047195397263, + "learning_rate": 6.178103563434629e-07, + "loss": 0.6764, + "num_input_tokens_seen": 132434645, + "step": 6242 + }, + { + "epoch": 0.7506763662598449, + "flos": 14838267133200.0, + "grad_norm": 3.556164109947667, + "learning_rate": 6.172474484530283e-07, + "loss": 0.8345, + "num_input_tokens_seen": 132453100, + "step": 6243 + }, + { + "epoch": 0.750796609150484, + "flos": 27366372151200.0, + "grad_norm": 6.1061825423482485, + "learning_rate": 6.166847503250563e-07, + "loss": 0.7376, + "num_input_tokens_seen": 132475060, + "step": 6244 + }, + { + "epoch": 0.750916852041123, + "flos": 14325397797120.0, + "grad_norm": 5.204793393361366, + "learning_rate": 6.161222620449078e-07, + "loss": 0.7708, + "num_input_tokens_seen": 132493555, + "step": 6245 + }, + { + "epoch": 0.7510370949317622, + "flos": 18397330216320.0, + "grad_norm": 6.080944829107434, + "learning_rate": 6.155599836979117e-07, + "loss": 0.7803, + "num_input_tokens_seen": 132511960, + "step": 6246 + }, + { + "epoch": 0.7511573378224012, + "flos": 13920502619160.0, + "grad_norm": 3.863283595131682, + "learning_rate": 6.149979153693649e-07, + "loss": 0.7984, + "num_input_tokens_seen": 132528935, + "step": 6247 + }, + { + "epoch": 0.7512775807130403, + "flos": 14568315907920.0, + "grad_norm": 17.372640984754053, + "learning_rate": 6.144360571445343e-07, + "loss": 0.7419, + "num_input_tokens_seen": 132547800, + "step": 6248 + }, + { + "epoch": 0.7513978236036795, + "flos": 15161967987840.0, + "grad_norm": 15.702794131309034, + "learning_rate": 6.138744091086509e-07, + "loss": 0.7873, + "num_input_tokens_seen": 132567105, + "step": 6249 + }, + { + "epoch": 0.7515180664943185, + "flos": 20211683222520.0, + "grad_norm": 5.355983183545926, + "learning_rate": 6.133129713469183e-07, + "loss": 0.7101, + "num_input_tokens_seen": 132586030, + "step": 6250 + }, + { + "epoch": 0.7516383093849576, + "flos": 24829850084280.0, + "grad_norm": 8.32742277981661, + "learning_rate": 6.127517439445053e-07, + "loss": 0.6204, + "num_input_tokens_seen": 132606595, + "step": 6251 + }, + { + "epoch": 0.7517585522755967, + "flos": 21826388300520.0, + "grad_norm": 6.542185120984163, + "learning_rate": 6.121907269865498e-07, + "loss": 0.8052, + "num_input_tokens_seen": 132625805, + "step": 6252 + }, + { + "epoch": 0.7518787951662358, + "flos": 51480036907200.0, + "grad_norm": 0.9606965546027709, + "learning_rate": 6.116299205581577e-07, + "loss": 0.721, + "num_input_tokens_seen": 132680355, + "step": 6253 + }, + { + "epoch": 0.7519990380568748, + "flos": 25125726325440.0, + "grad_norm": 3.654736458269835, + "learning_rate": 6.110693247444018e-07, + "loss": 0.6603, + "num_input_tokens_seen": 132701910, + "step": 6254 + }, + { + "epoch": 0.752119280947514, + "flos": 15891608799120.0, + "grad_norm": 6.320172773969258, + "learning_rate": 6.105089396303258e-07, + "loss": 0.8091, + "num_input_tokens_seen": 132720020, + "step": 6255 + }, + { + "epoch": 0.7522395238381531, + "flos": 24044750005800.0, + "grad_norm": 4.418781829676179, + "learning_rate": 6.099487653009383e-07, + "loss": 0.743, + "num_input_tokens_seen": 132739085, + "step": 6256 + }, + { + "epoch": 0.7523597667287921, + "flos": 17186823797160.0, + "grad_norm": 5.030606056847856, + "learning_rate": 6.093888018412192e-07, + "loss": 0.815, + "num_input_tokens_seen": 132754995, + "step": 6257 + }, + { + "epoch": 0.7524800096194313, + "flos": 49661350977840.0, + "grad_norm": 0.7409721864507449, + "learning_rate": 6.088290493361125e-07, + "loss": 0.5685, + "num_input_tokens_seen": 132819600, + "step": 6258 + }, + { + "epoch": 0.7526002525100703, + "flos": 9438546108480.0, + "grad_norm": 6.9358207801095055, + "learning_rate": 6.082695078705322e-07, + "loss": 0.6917, + "num_input_tokens_seen": 132836800, + "step": 6259 + }, + { + "epoch": 0.7527204954007094, + "flos": 15648500728560.0, + "grad_norm": 15.036027494080155, + "learning_rate": 6.077101775293618e-07, + "loss": 0.6684, + "num_input_tokens_seen": 132855345, + "step": 6260 + }, + { + "epoch": 0.7528407382913486, + "flos": 13834527641880.0, + "grad_norm": 8.512387393425078, + "learning_rate": 6.071510583974504e-07, + "loss": 0.8164, + "num_input_tokens_seen": 132870250, + "step": 6261 + }, + { + "epoch": 0.7529609811819876, + "flos": 11085761474040.0, + "grad_norm": 5.427382780027808, + "learning_rate": 6.065921505596161e-07, + "loss": 0.6988, + "num_input_tokens_seen": 132888250, + "step": 6262 + }, + { + "epoch": 0.7530812240726267, + "flos": 14136799194960.0, + "grad_norm": 3.0785660016318976, + "learning_rate": 6.060334541006445e-07, + "loss": 0.7626, + "num_input_tokens_seen": 132906465, + "step": 6263 + }, + { + "epoch": 0.7532014669632658, + "flos": 20349033332160.0, + "grad_norm": 3.9324707125223695, + "learning_rate": 6.05474969105289e-07, + "loss": 0.6741, + "num_input_tokens_seen": 132929175, + "step": 6264 + }, + { + "epoch": 0.7533217098539049, + "flos": 10275876138240.0, + "grad_norm": 6.619720082429963, + "learning_rate": 6.049166956582725e-07, + "loss": 0.7206, + "num_input_tokens_seen": 132947160, + "step": 6265 + }, + { + "epoch": 0.753441952744544, + "flos": 19371947035800.0, + "grad_norm": 4.967556408205263, + "learning_rate": 6.043586338442841e-07, + "loss": 0.8583, + "num_input_tokens_seen": 132965935, + "step": 6266 + }, + { + "epoch": 0.7535621956351831, + "flos": 17484947895480.0, + "grad_norm": 4.021217341031021, + "learning_rate": 6.038007837479815e-07, + "loss": 0.7173, + "num_input_tokens_seen": 132986760, + "step": 6267 + }, + { + "epoch": 0.7536824385258222, + "flos": 15943142231280.0, + "grad_norm": 5.3952360473129986, + "learning_rate": 6.032431454539897e-07, + "loss": 0.616, + "num_input_tokens_seen": 133005325, + "step": 6268 + }, + { + "epoch": 0.7538026814164612, + "flos": 21209438981400.0, + "grad_norm": 6.020164936632368, + "learning_rate": 6.026857190469014e-07, + "loss": 0.7985, + "num_input_tokens_seen": 133026800, + "step": 6269 + }, + { + "epoch": 0.7539229243071004, + "flos": 15431919213120.0, + "grad_norm": 5.420938108557101, + "learning_rate": 6.0212850461128e-07, + "loss": 0.7212, + "num_input_tokens_seen": 133045640, + "step": 6270 + }, + { + "epoch": 0.7540431671977395, + "flos": 11031441965400.0, + "grad_norm": 5.538023513606263, + "learning_rate": 6.015715022316516e-07, + "loss": 0.7258, + "num_input_tokens_seen": 133063340, + "step": 6271 + }, + { + "epoch": 0.7541634100883785, + "flos": 13704649282800.0, + "grad_norm": 5.9764115457856155, + "learning_rate": 6.010147119925154e-07, + "loss": 0.7585, + "num_input_tokens_seen": 133080815, + "step": 6272 + }, + { + "epoch": 0.7542836529790176, + "flos": 15054658688880.0, + "grad_norm": 22.312610102363152, + "learning_rate": 6.004581339783348e-07, + "loss": 0.6345, + "num_input_tokens_seen": 133098855, + "step": 6273 + }, + { + "epoch": 0.7544038958696567, + "flos": 13944274757760.0, + "grad_norm": 10.727882165157842, + "learning_rate": 5.999017682735425e-07, + "loss": 0.6578, + "num_input_tokens_seen": 133114965, + "step": 6274 + }, + { + "epoch": 0.7545241387602958, + "flos": 23290577216880.0, + "grad_norm": 3.6568951159888714, + "learning_rate": 5.993456149625387e-07, + "loss": 0.6502, + "num_input_tokens_seen": 133135835, + "step": 6275 + }, + { + "epoch": 0.7546443816509348, + "flos": 14833011579840.0, + "grad_norm": 7.74170536268382, + "learning_rate": 5.987896741296909e-07, + "loss": 0.8125, + "num_input_tokens_seen": 133153295, + "step": 6276 + }, + { + "epoch": 0.754764624541574, + "flos": 17349908962920.0, + "grad_norm": 4.388069902668271, + "learning_rate": 5.982339458593361e-07, + "loss": 0.7647, + "num_input_tokens_seen": 133172955, + "step": 6277 + }, + { + "epoch": 0.7548848674322131, + "flos": 18564277897200.0, + "grad_norm": 4.531014903316601, + "learning_rate": 5.976784302357767e-07, + "loss": 0.8383, + "num_input_tokens_seen": 133193240, + "step": 6278 + }, + { + "epoch": 0.7550051103228521, + "flos": 14298586302360.0, + "grad_norm": 5.4752899302179685, + "learning_rate": 5.971231273432855e-07, + "loss": 0.7119, + "num_input_tokens_seen": 133212445, + "step": 6279 + }, + { + "epoch": 0.7551253532134913, + "flos": 47292436959600.0, + "grad_norm": 0.8260252982622828, + "learning_rate": 5.965680372661e-07, + "loss": 0.5761, + "num_input_tokens_seen": 133269730, + "step": 6280 + }, + { + "epoch": 0.7552455961041303, + "flos": 19105256786400.0, + "grad_norm": 4.496703742015592, + "learning_rate": 5.960131600884266e-07, + "loss": 0.5371, + "num_input_tokens_seen": 133288720, + "step": 6281 + }, + { + "epoch": 0.7553658389947694, + "flos": 17943402743040.0, + "grad_norm": 4.99627126895824, + "learning_rate": 5.954584958944413e-07, + "loss": 0.7436, + "num_input_tokens_seen": 133307105, + "step": 6282 + }, + { + "epoch": 0.7554860818854086, + "flos": 15946308227280.0, + "grad_norm": 4.408102309021069, + "learning_rate": 5.949040447682854e-07, + "loss": 0.794, + "num_input_tokens_seen": 133326650, + "step": 6283 + }, + { + "epoch": 0.7556063247760476, + "flos": 11922173364960.0, + "grad_norm": 4.863690879146773, + "learning_rate": 5.943498067940686e-07, + "loss": 0.6669, + "num_input_tokens_seen": 133343395, + "step": 6284 + }, + { + "epoch": 0.7557265676666867, + "flos": 19968986731440.0, + "grad_norm": 5.002924254736705, + "learning_rate": 5.937957820558686e-07, + "loss": 0.793, + "num_input_tokens_seen": 133362460, + "step": 6285 + }, + { + "epoch": 0.7558468105573258, + "flos": 45842526685200.0, + "grad_norm": 0.8381733806280773, + "learning_rate": 5.932419706377296e-07, + "loss": 0.6624, + "num_input_tokens_seen": 133420485, + "step": 6286 + }, + { + "epoch": 0.7559670534479649, + "flos": 24418812874080.0, + "grad_norm": 7.993349648979446, + "learning_rate": 5.92688372623666e-07, + "loss": 0.7215, + "num_input_tokens_seen": 133438910, + "step": 6287 + }, + { + "epoch": 0.7560872963386039, + "flos": 10221809909280.0, + "grad_norm": 27.129538160837026, + "learning_rate": 5.921349880976574e-07, + "loss": 0.7254, + "num_input_tokens_seen": 133456465, + "step": 6288 + }, + { + "epoch": 0.7562075392292431, + "flos": 14919619756320.0, + "grad_norm": 3.763763474130219, + "learning_rate": 5.915818171436515e-07, + "loss": 0.7932, + "num_input_tokens_seen": 133475520, + "step": 6289 + }, + { + "epoch": 0.7563277821198822, + "flos": 14893473120720.0, + "grad_norm": 17.698441534883003, + "learning_rate": 5.910288598455642e-07, + "loss": 0.7296, + "num_input_tokens_seen": 133494590, + "step": 6290 + }, + { + "epoch": 0.7564480250105212, + "flos": 13569736990080.0, + "grad_norm": 5.088341398277517, + "learning_rate": 5.90476116287278e-07, + "loss": 0.7164, + "num_input_tokens_seen": 133511910, + "step": 6291 + }, + { + "epoch": 0.7565682679011604, + "flos": 15514126655160.0, + "grad_norm": 5.284137545889759, + "learning_rate": 5.899235865526456e-07, + "loss": 0.6612, + "num_input_tokens_seen": 133530925, + "step": 6292 + }, + { + "epoch": 0.7566885107917994, + "flos": 14946811170600.0, + "grad_norm": 3.3018295880526662, + "learning_rate": 5.893712707254825e-07, + "loss": 0.8035, + "num_input_tokens_seen": 133548105, + "step": 6293 + }, + { + "epoch": 0.7568087536824385, + "flos": 13893532824600.0, + "grad_norm": 9.400105950590357, + "learning_rate": 5.888191688895769e-07, + "loss": 0.6312, + "num_input_tokens_seen": 133565085, + "step": 6294 + }, + { + "epoch": 0.7569289965730777, + "flos": 11080157661120.0, + "grad_norm": 4.945869788035779, + "learning_rate": 5.882672811286813e-07, + "loss": 0.5882, + "num_input_tokens_seen": 133581085, + "step": 6295 + }, + { + "epoch": 0.7570492394637167, + "flos": 15182764090200.0, + "grad_norm": 4.317753784284048, + "learning_rate": 5.877156075265166e-07, + "loss": 0.6779, + "num_input_tokens_seen": 133597070, + "step": 6296 + }, + { + "epoch": 0.7571694823543558, + "flos": 11404903294440.0, + "grad_norm": 6.433156742709564, + "learning_rate": 5.871641481667715e-07, + "loss": 0.6799, + "num_input_tokens_seen": 133611235, + "step": 6297 + }, + { + "epoch": 0.7572897252449949, + "flos": 18617457647280.0, + "grad_norm": 11.187851884305417, + "learning_rate": 5.866129031331011e-07, + "loss": 0.8252, + "num_input_tokens_seen": 133630610, + "step": 6298 + }, + { + "epoch": 0.757409968135634, + "flos": 17781425675880.0, + "grad_norm": 7.154906034245783, + "learning_rate": 5.8606187250913e-07, + "loss": 0.8034, + "num_input_tokens_seen": 133648380, + "step": 6299 + }, + { + "epoch": 0.757530211026273, + "flos": 17673736457400.0, + "grad_norm": 5.927619503108547, + "learning_rate": 5.855110563784482e-07, + "loss": 0.8182, + "num_input_tokens_seen": 133666635, + "step": 6300 + }, + { + "epoch": 0.7576504539169122, + "flos": 17539140764280.0, + "grad_norm": 3.3839708419788126, + "learning_rate": 5.849604548246156e-07, + "loss": 0.6266, + "num_input_tokens_seen": 133687465, + "step": 6301 + }, + { + "epoch": 0.7577706968075513, + "flos": 15540558230400.0, + "grad_norm": 11.549788462365473, + "learning_rate": 5.844100679311565e-07, + "loss": 0.7858, + "num_input_tokens_seen": 133706145, + "step": 6302 + }, + { + "epoch": 0.7578909396981903, + "flos": 13353915313680.0, + "grad_norm": 5.128762329592518, + "learning_rate": 5.838598957815637e-07, + "loss": 0.7541, + "num_input_tokens_seen": 133723095, + "step": 6303 + }, + { + "epoch": 0.7580111825888295, + "flos": 18591437651520.0, + "grad_norm": 3.0814544176614067, + "learning_rate": 5.833099384592996e-07, + "loss": 0.8441, + "num_input_tokens_seen": 133743390, + "step": 6304 + }, + { + "epoch": 0.7581314254794685, + "flos": 17403500292480.0, + "grad_norm": 5.360725432882013, + "learning_rate": 5.827601960477913e-07, + "loss": 0.6915, + "num_input_tokens_seen": 133761035, + "step": 6305 + }, + { + "epoch": 0.7582516683701076, + "flos": 16135096789200.0, + "grad_norm": 3.8822224895979667, + "learning_rate": 5.822106686304344e-07, + "loss": 0.7012, + "num_input_tokens_seen": 133780045, + "step": 6306 + }, + { + "epoch": 0.7583719112607467, + "flos": 23236036088520.0, + "grad_norm": 4.350972490017228, + "learning_rate": 5.816613562905919e-07, + "loss": 0.5583, + "num_input_tokens_seen": 133800950, + "step": 6307 + }, + { + "epoch": 0.7584921541513858, + "flos": 24285325279560.0, + "grad_norm": 2.8784979397420316, + "learning_rate": 5.811122591115933e-07, + "loss": 0.6956, + "num_input_tokens_seen": 133821655, + "step": 6308 + }, + { + "epoch": 0.7586123970420249, + "flos": 17075525343240.0, + "grad_norm": 4.298958789827572, + "learning_rate": 5.805633771767376e-07, + "loss": 0.6787, + "num_input_tokens_seen": 133838770, + "step": 6309 + }, + { + "epoch": 0.7587326399326639, + "flos": 13381929886920.0, + "grad_norm": 2.7425070899725683, + "learning_rate": 5.800147105692888e-07, + "loss": 0.759, + "num_input_tokens_seen": 133858065, + "step": 6310 + }, + { + "epoch": 0.7588528828233031, + "flos": 12597937907040.0, + "grad_norm": 5.250532395752146, + "learning_rate": 5.794662593724795e-07, + "loss": 0.7774, + "num_input_tokens_seen": 133876790, + "step": 6311 + }, + { + "epoch": 0.7589731257139422, + "flos": 12922588560480.0, + "grad_norm": 30.309364770206543, + "learning_rate": 5.789180236695091e-07, + "loss": 0.7388, + "num_input_tokens_seen": 133893365, + "step": 6312 + }, + { + "epoch": 0.7590933686045812, + "flos": 11625410644920.0, + "grad_norm": 5.14639653921724, + "learning_rate": 5.78370003543544e-07, + "loss": 0.8444, + "num_input_tokens_seen": 133911840, + "step": 6313 + }, + { + "epoch": 0.7592136114952204, + "flos": 15404791118760.0, + "grad_norm": 5.0948191263814255, + "learning_rate": 5.778221990777203e-07, + "loss": 0.8174, + "num_input_tokens_seen": 133929300, + "step": 6314 + }, + { + "epoch": 0.7593338543858594, + "flos": 18534585346080.0, + "grad_norm": 8.309179575021888, + "learning_rate": 5.772746103551372e-07, + "loss": 0.8149, + "num_input_tokens_seen": 133944415, + "step": 6315 + }, + { + "epoch": 0.7594540972764985, + "flos": 23371233320880.0, + "grad_norm": 3.617237593301895, + "learning_rate": 5.767272374588648e-07, + "loss": 0.7127, + "num_input_tokens_seen": 133965540, + "step": 6316 + }, + { + "epoch": 0.7595743401671377, + "flos": 27637653094800.0, + "grad_norm": 8.428475534110635, + "learning_rate": 5.76180080471939e-07, + "loss": 0.7722, + "num_input_tokens_seen": 133988430, + "step": 6317 + }, + { + "epoch": 0.7596945830577767, + "flos": 13346095303560.0, + "grad_norm": 4.7755803419625416, + "learning_rate": 5.756331394773631e-07, + "loss": 0.703, + "num_input_tokens_seen": 134004365, + "step": 6318 + }, + { + "epoch": 0.7598148259484158, + "flos": 16266653126160.0, + "grad_norm": 6.451193458089085, + "learning_rate": 5.750864145581071e-07, + "loss": 0.7541, + "num_input_tokens_seen": 134023305, + "step": 6319 + }, + { + "epoch": 0.7599350688390549, + "flos": 19914572242920.0, + "grad_norm": 6.2531419104672015, + "learning_rate": 5.745399057971085e-07, + "loss": 0.8404, + "num_input_tokens_seen": 134044160, + "step": 6320 + }, + { + "epoch": 0.760055311729694, + "flos": 11329249464120.0, + "grad_norm": 6.802798020894486, + "learning_rate": 5.739936132772738e-07, + "loss": 0.731, + "num_input_tokens_seen": 134062445, + "step": 6321 + }, + { + "epoch": 0.760175554620333, + "flos": 18429903823800.0, + "grad_norm": 4.249712317210747, + "learning_rate": 5.734475370814733e-07, + "loss": 0.736, + "num_input_tokens_seen": 134081845, + "step": 6322 + }, + { + "epoch": 0.7602957975109722, + "flos": 17836726643280.0, + "grad_norm": 2.8750408003855936, + "learning_rate": 5.729016772925483e-07, + "loss": 0.7691, + "num_input_tokens_seen": 134103140, + "step": 6323 + }, + { + "epoch": 0.7604160404016113, + "flos": 18457506817560.0, + "grad_norm": 3.704386950802307, + "learning_rate": 5.723560339933038e-07, + "loss": 0.6981, + "num_input_tokens_seen": 134123195, + "step": 6324 + }, + { + "epoch": 0.7605362832922503, + "flos": 21912869837160.0, + "grad_norm": 4.4900674345054385, + "learning_rate": 5.71810607266513e-07, + "loss": 0.6275, + "num_input_tokens_seen": 134141500, + "step": 6325 + }, + { + "epoch": 0.7606565261828895, + "flos": 10113360851760.0, + "grad_norm": 4.6275874967459085, + "learning_rate": 5.712653971949184e-07, + "loss": 0.5835, + "num_input_tokens_seen": 134159340, + "step": 6326 + }, + { + "epoch": 0.7607767690735285, + "flos": 13542292296120.0, + "grad_norm": 7.013272645105142, + "learning_rate": 5.707204038612268e-07, + "loss": 0.753, + "num_input_tokens_seen": 134176490, + "step": 6327 + }, + { + "epoch": 0.7608970119641676, + "flos": 15297545139720.0, + "grad_norm": 4.003944359574732, + "learning_rate": 5.701756273481138e-07, + "loss": 0.7155, + "num_input_tokens_seen": 134193630, + "step": 6328 + }, + { + "epoch": 0.7610172548548068, + "flos": 17432401344600.0, + "grad_norm": 5.1912052755727895, + "learning_rate": 5.696310677382212e-07, + "loss": 0.7203, + "num_input_tokens_seen": 134214745, + "step": 6329 + }, + { + "epoch": 0.7611374977454458, + "flos": 49030629576120.0, + "grad_norm": 0.8275007378151361, + "learning_rate": 5.690867251141576e-07, + "loss": 0.6395, + "num_input_tokens_seen": 134281120, + "step": 6330 + }, + { + "epoch": 0.7612577406360849, + "flos": 11382777473760.0, + "grad_norm": 12.4422389831206, + "learning_rate": 5.685425995585013e-07, + "loss": 0.8943, + "num_input_tokens_seen": 134298765, + "step": 6331 + }, + { + "epoch": 0.761377983526724, + "flos": 44614073560080.0, + "grad_norm": 0.8037036696113518, + "learning_rate": 5.679986911537935e-07, + "loss": 0.613, + "num_input_tokens_seen": 134366015, + "step": 6332 + }, + { + "epoch": 0.7614982264173631, + "flos": 26287168789320.0, + "grad_norm": 3.8174145400318755, + "learning_rate": 5.674549999825462e-07, + "loss": 0.6587, + "num_input_tokens_seen": 134388550, + "step": 6333 + }, + { + "epoch": 0.7616184693080021, + "flos": 50088973515720.0, + "grad_norm": 0.9957330387211926, + "learning_rate": 5.669115261272363e-07, + "loss": 0.7648, + "num_input_tokens_seen": 134448590, + "step": 6334 + }, + { + "epoch": 0.7617387121986413, + "flos": 15000814079640.0, + "grad_norm": 8.879789107971584, + "learning_rate": 5.663682696703081e-07, + "loss": 0.7082, + "num_input_tokens_seen": 134466575, + "step": 6335 + }, + { + "epoch": 0.7618589550892804, + "flos": 13596516824880.0, + "grad_norm": 3.7120262074860784, + "learning_rate": 5.658252306941746e-07, + "loss": 0.8202, + "num_input_tokens_seen": 134485615, + "step": 6336 + }, + { + "epoch": 0.7619791979799194, + "flos": 12729620883840.0, + "grad_norm": 3.927331054861245, + "learning_rate": 5.65282409281212e-07, + "loss": 0.7406, + "num_input_tokens_seen": 134502800, + "step": 6337 + }, + { + "epoch": 0.7620994408705585, + "flos": 10275876138240.0, + "grad_norm": 3.734818946913377, + "learning_rate": 5.64739805513768e-07, + "loss": 0.6847, + "num_input_tokens_seen": 134520065, + "step": 6338 + }, + { + "epoch": 0.7622196837611976, + "flos": 52145100382800.0, + "grad_norm": 0.8111191165832388, + "learning_rate": 5.641974194741541e-07, + "loss": 0.5792, + "num_input_tokens_seen": 134575470, + "step": 6339 + }, + { + "epoch": 0.7623399266518367, + "flos": 44727457080000.0, + "grad_norm": 0.7436725415838848, + "learning_rate": 5.636552512446502e-07, + "loss": 0.6423, + "num_input_tokens_seen": 134636245, + "step": 6340 + }, + { + "epoch": 0.7624601695424758, + "flos": 19401892866600.0, + "grad_norm": 3.839369347760826, + "learning_rate": 5.631133009075027e-07, + "loss": 0.7732, + "num_input_tokens_seen": 134655150, + "step": 6341 + }, + { + "epoch": 0.7625804124331149, + "flos": 13974948767640.0, + "grad_norm": 4.319362542840003, + "learning_rate": 5.625715685449242e-07, + "loss": 0.6816, + "num_input_tokens_seen": 134672975, + "step": 6342 + }, + { + "epoch": 0.762700655323754, + "flos": 19212534425400.0, + "grad_norm": 2.911780724506889, + "learning_rate": 5.620300542390966e-07, + "loss": 0.7073, + "num_input_tokens_seen": 134693740, + "step": 6343 + }, + { + "epoch": 0.762820898214393, + "flos": 16377761620320.0, + "grad_norm": 7.630254717133619, + "learning_rate": 5.614887580721659e-07, + "loss": 0.8364, + "num_input_tokens_seen": 134713605, + "step": 6344 + }, + { + "epoch": 0.7629411411050322, + "flos": 11431841429040.0, + "grad_norm": 6.098545789135923, + "learning_rate": 5.609476801262481e-07, + "loss": 0.7153, + "num_input_tokens_seen": 134728185, + "step": 6345 + }, + { + "epoch": 0.7630613839956712, + "flos": 10003898675520.0, + "grad_norm": 7.115778795137475, + "learning_rate": 5.604068204834223e-07, + "loss": 0.617, + "num_input_tokens_seen": 134744800, + "step": 6346 + }, + { + "epoch": 0.7631816268863103, + "flos": 10594891318800.0, + "grad_norm": 5.687031150189232, + "learning_rate": 5.598661792257367e-07, + "loss": 0.7321, + "num_input_tokens_seen": 134761565, + "step": 6347 + }, + { + "epoch": 0.7633018697769495, + "flos": 13920439299240.0, + "grad_norm": 11.95839280879065, + "learning_rate": 5.593257564352071e-07, + "loss": 0.7512, + "num_input_tokens_seen": 134779725, + "step": 6348 + }, + { + "epoch": 0.7634221126675885, + "flos": 16133640431040.0, + "grad_norm": 4.338148938096345, + "learning_rate": 5.58785552193815e-07, + "loss": 0.7407, + "num_input_tokens_seen": 134799690, + "step": 6349 + }, + { + "epoch": 0.7635423555582276, + "flos": 21557608493760.0, + "grad_norm": 3.6557940337346473, + "learning_rate": 5.582455665835086e-07, + "loss": 0.7401, + "num_input_tokens_seen": 134819705, + "step": 6350 + }, + { + "epoch": 0.7636625984488667, + "flos": 12975103451400.0, + "grad_norm": 6.500200415235858, + "learning_rate": 5.577057996862036e-07, + "loss": 0.7105, + "num_input_tokens_seen": 134837050, + "step": 6351 + }, + { + "epoch": 0.7637828413395058, + "flos": 17378620055280.0, + "grad_norm": 4.039896364299132, + "learning_rate": 5.571662515837814e-07, + "loss": 0.738, + "num_input_tokens_seen": 134858730, + "step": 6352 + }, + { + "epoch": 0.7639030842301449, + "flos": 26665157492640.0, + "grad_norm": 4.9688156411495505, + "learning_rate": 5.566269223580926e-07, + "loss": 0.8188, + "num_input_tokens_seen": 134880160, + "step": 6353 + }, + { + "epoch": 0.764023327120784, + "flos": 21185033643600.0, + "grad_norm": 3.1943963936402993, + "learning_rate": 5.560878120909511e-07, + "loss": 0.7377, + "num_input_tokens_seen": 134902480, + "step": 6354 + }, + { + "epoch": 0.7641435700114231, + "flos": 47768558064840.0, + "grad_norm": 0.9200792441240235, + "learning_rate": 5.55548920864141e-07, + "loss": 0.6423, + "num_input_tokens_seen": 134962855, + "step": 6355 + }, + { + "epoch": 0.7642638129020621, + "flos": 12272622394440.0, + "grad_norm": 5.095234660396598, + "learning_rate": 5.550102487594113e-07, + "loss": 0.7635, + "num_input_tokens_seen": 134981245, + "step": 6356 + }, + { + "epoch": 0.7643840557927013, + "flos": 22317733355160.0, + "grad_norm": 4.165577224720759, + "learning_rate": 5.54471795858477e-07, + "loss": 0.701, + "num_input_tokens_seen": 135001035, + "step": 6357 + }, + { + "epoch": 0.7645042986833404, + "flos": 12381957930840.0, + "grad_norm": 3.789030660908433, + "learning_rate": 5.539335622430235e-07, + "loss": 0.8073, + "num_input_tokens_seen": 135019375, + "step": 6358 + }, + { + "epoch": 0.7646245415739794, + "flos": 12624876041640.0, + "grad_norm": 9.092266158337148, + "learning_rate": 5.533955479946975e-07, + "loss": 0.7295, + "num_input_tokens_seen": 135037875, + "step": 6359 + }, + { + "epoch": 0.7647447844646186, + "flos": 51921399867720.0, + "grad_norm": 0.9075986429728874, + "learning_rate": 5.528577531951173e-07, + "loss": 0.6659, + "num_input_tokens_seen": 135098000, + "step": 6360 + }, + { + "epoch": 0.7648650273552576, + "flos": 12894510667320.0, + "grad_norm": 4.90756460667125, + "learning_rate": 5.523201779258653e-07, + "loss": 0.73, + "num_input_tokens_seen": 135116695, + "step": 6361 + }, + { + "epoch": 0.7649852702458967, + "flos": 16214296535040.0, + "grad_norm": 5.287058147609232, + "learning_rate": 5.517828222684912e-07, + "loss": 0.8286, + "num_input_tokens_seen": 135137070, + "step": 6362 + }, + { + "epoch": 0.7651055131365359, + "flos": 51510457637400.0, + "grad_norm": 0.7806029119489553, + "learning_rate": 5.512456863045117e-07, + "loss": 0.5968, + "num_input_tokens_seen": 135197480, + "step": 6363 + }, + { + "epoch": 0.7652257560271749, + "flos": 14217866878440.0, + "grad_norm": 7.139388163021055, + "learning_rate": 5.507087701154089e-07, + "loss": 0.7241, + "num_input_tokens_seen": 135217120, + "step": 6364 + }, + { + "epoch": 0.765345998917814, + "flos": 11625442304880.0, + "grad_norm": 4.205794950586849, + "learning_rate": 5.50172073782634e-07, + "loss": 0.7275, + "num_input_tokens_seen": 135234820, + "step": 6365 + }, + { + "epoch": 0.7654662418084531, + "flos": 17323065808200.0, + "grad_norm": 4.227762274115651, + "learning_rate": 5.496355973876023e-07, + "loss": 0.8632, + "num_input_tokens_seen": 135253795, + "step": 6366 + }, + { + "epoch": 0.7655864846990922, + "flos": 30498857475120.0, + "grad_norm": 5.569490990109959, + "learning_rate": 5.490993410116984e-07, + "loss": 0.6877, + "num_input_tokens_seen": 135276505, + "step": 6367 + }, + { + "epoch": 0.7657067275897312, + "flos": 31741589242200.0, + "grad_norm": 4.286874682143773, + "learning_rate": 5.485633047362704e-07, + "loss": 0.6898, + "num_input_tokens_seen": 135298230, + "step": 6368 + }, + { + "epoch": 0.7658269704803703, + "flos": 12625129321320.0, + "grad_norm": 5.93596267875599, + "learning_rate": 5.480274886426341e-07, + "loss": 0.7683, + "num_input_tokens_seen": 135314590, + "step": 6369 + }, + { + "epoch": 0.7659472133710095, + "flos": 9114465334320.0, + "grad_norm": 5.719840857686153, + "learning_rate": 5.474918928120744e-07, + "loss": 0.7567, + "num_input_tokens_seen": 135330805, + "step": 6370 + }, + { + "epoch": 0.7660674562616485, + "flos": 16617672034920.0, + "grad_norm": 4.83999192670669, + "learning_rate": 5.469565173258392e-07, + "loss": 0.8647, + "num_input_tokens_seen": 135349040, + "step": 6371 + }, + { + "epoch": 0.7661876991522876, + "flos": 12436182459600.0, + "grad_norm": 3.864164770171261, + "learning_rate": 5.464213622651454e-07, + "loss": 0.6228, + "num_input_tokens_seen": 135366575, + "step": 6372 + }, + { + "epoch": 0.7663079420429267, + "flos": 14676859945320.0, + "grad_norm": 9.446579765327519, + "learning_rate": 5.458864277111753e-07, + "loss": 0.8245, + "num_input_tokens_seen": 135384130, + "step": 6373 + }, + { + "epoch": 0.7664281849335658, + "flos": 9195343058040.0, + "grad_norm": 6.178360170742239, + "learning_rate": 5.453517137450769e-07, + "loss": 0.6864, + "num_input_tokens_seen": 135400425, + "step": 6374 + }, + { + "epoch": 0.7665484278242048, + "flos": 16350031986720.0, + "grad_norm": 11.081928670877026, + "learning_rate": 5.448172204479684e-07, + "loss": 0.7445, + "num_input_tokens_seen": 135419425, + "step": 6375 + }, + { + "epoch": 0.766668670714844, + "flos": 17291252039760.0, + "grad_norm": 6.483358648229501, + "learning_rate": 5.442829479009294e-07, + "loss": 0.733, + "num_input_tokens_seen": 135437925, + "step": 6376 + }, + { + "epoch": 0.7667889136054831, + "flos": 14190517164360.0, + "grad_norm": 3.5114649595011196, + "learning_rate": 5.437488961850103e-07, + "loss": 0.7038, + "num_input_tokens_seen": 135457445, + "step": 6377 + }, + { + "epoch": 0.7669091564961221, + "flos": 19695489590640.0, + "grad_norm": 2.8512532941840183, + "learning_rate": 5.432150653812258e-07, + "loss": 0.7418, + "num_input_tokens_seen": 135477200, + "step": 6378 + }, + { + "epoch": 0.7670293993867613, + "flos": 8979078142200.0, + "grad_norm": 4.524630848700985, + "learning_rate": 5.42681455570557e-07, + "loss": 0.8149, + "num_input_tokens_seen": 135493450, + "step": 6379 + }, + { + "epoch": 0.7671496422774003, + "flos": 15919275112800.0, + "grad_norm": 6.170530844037391, + "learning_rate": 5.42148066833954e-07, + "loss": 0.639, + "num_input_tokens_seen": 135512415, + "step": 6380 + }, + { + "epoch": 0.7672698851680394, + "flos": 15405804237480.0, + "grad_norm": 6.080106044647589, + "learning_rate": 5.416148992523289e-07, + "loss": 0.7384, + "num_input_tokens_seen": 135530710, + "step": 6381 + }, + { + "epoch": 0.7673901280586786, + "flos": 12376195818120.0, + "grad_norm": 3.4431749144083486, + "learning_rate": 5.410819529065644e-07, + "loss": 0.7692, + "num_input_tokens_seen": 135548385, + "step": 6382 + }, + { + "epoch": 0.7675103709493176, + "flos": 21454794909120.0, + "grad_norm": 9.069487305001642, + "learning_rate": 5.405492278775079e-07, + "loss": 0.6378, + "num_input_tokens_seen": 135567885, + "step": 6383 + }, + { + "epoch": 0.7676306138399567, + "flos": 21292121322840.0, + "grad_norm": 3.7744717967366443, + "learning_rate": 5.400167242459732e-07, + "loss": 0.7842, + "num_input_tokens_seen": 135586565, + "step": 6384 + }, + { + "epoch": 0.7677508567305958, + "flos": 16512863872800.0, + "grad_norm": 7.244894137138901, + "learning_rate": 5.394844420927405e-07, + "loss": 0.7943, + "num_input_tokens_seen": 135605895, + "step": 6385 + }, + { + "epoch": 0.7678710996212349, + "flos": 18619388904840.0, + "grad_norm": 4.839199167944644, + "learning_rate": 5.389523814985562e-07, + "loss": 0.7054, + "num_input_tokens_seen": 135625035, + "step": 6386 + }, + { + "epoch": 0.767991342511874, + "flos": 19614421907160.0, + "grad_norm": 3.952728130939949, + "learning_rate": 5.384205425441344e-07, + "loss": 0.7425, + "num_input_tokens_seen": 135645665, + "step": 6387 + }, + { + "epoch": 0.7681115854025131, + "flos": 19320951822960.0, + "grad_norm": 3.8977740072910927, + "learning_rate": 5.378889253101537e-07, + "loss": 0.8286, + "num_input_tokens_seen": 135665940, + "step": 6388 + }, + { + "epoch": 0.7682318282931522, + "flos": 17024751750120.0, + "grad_norm": 4.970273917803039, + "learning_rate": 5.373575298772617e-07, + "loss": 0.7914, + "num_input_tokens_seen": 135684780, + "step": 6389 + }, + { + "epoch": 0.7683520711837912, + "flos": 51677057058720.0, + "grad_norm": 0.7215623979663405, + "learning_rate": 5.368263563260689e-07, + "loss": 0.6154, + "num_input_tokens_seen": 135749635, + "step": 6390 + }, + { + "epoch": 0.7684723140744304, + "flos": 13596865084440.0, + "grad_norm": 3.314765731105566, + "learning_rate": 5.362954047371537e-07, + "loss": 0.6156, + "num_input_tokens_seen": 135768465, + "step": 6391 + }, + { + "epoch": 0.7685925569650695, + "flos": 20130045659760.0, + "grad_norm": 3.6677792046173177, + "learning_rate": 5.357646751910627e-07, + "loss": 0.7146, + "num_input_tokens_seen": 135789365, + "step": 6392 + }, + { + "epoch": 0.7687127998557085, + "flos": 17971290676440.0, + "grad_norm": 4.948581091456983, + "learning_rate": 5.352341677683061e-07, + "loss": 0.7834, + "num_input_tokens_seen": 135810385, + "step": 6393 + }, + { + "epoch": 0.7688330427463477, + "flos": 18430315403280.0, + "grad_norm": 3.9143992144027413, + "learning_rate": 5.347038825493617e-07, + "loss": 0.7682, + "num_input_tokens_seen": 135831635, + "step": 6394 + }, + { + "epoch": 0.7689532856369867, + "flos": 15510390779880.0, + "grad_norm": 3.7257955789640453, + "learning_rate": 5.341738196146732e-07, + "loss": 0.6532, + "num_input_tokens_seen": 135849700, + "step": 6395 + }, + { + "epoch": 0.7690735285276258, + "flos": 18402965689200.0, + "grad_norm": 3.6205854172257466, + "learning_rate": 5.336439790446503e-07, + "loss": 0.707, + "num_input_tokens_seen": 135868520, + "step": 6396 + }, + { + "epoch": 0.769193771418265, + "flos": 40327196960640.0, + "grad_norm": 4.14233708523461, + "learning_rate": 5.331143609196711e-07, + "loss": 0.6012, + "num_input_tokens_seen": 135892055, + "step": 6397 + }, + { + "epoch": 0.769314014308904, + "flos": 27448674573120.0, + "grad_norm": 4.607229373495528, + "learning_rate": 5.325849653200758e-07, + "loss": 0.7611, + "num_input_tokens_seen": 135915725, + "step": 6398 + }, + { + "epoch": 0.7694342571995431, + "flos": 15082040062920.0, + "grad_norm": 2.8240195978419416, + "learning_rate": 5.32055792326175e-07, + "loss": 0.7472, + "num_input_tokens_seen": 135933870, + "step": 6399 + }, + { + "epoch": 0.7695545000901821, + "flos": 17728594185360.0, + "grad_norm": 5.32219627965145, + "learning_rate": 5.315268420182437e-07, + "loss": 0.7117, + "num_input_tokens_seen": 135952265, + "step": 6400 + }, + { + "epoch": 0.7696747429808213, + "flos": 20535953956440.0, + "grad_norm": 6.763103905188286, + "learning_rate": 5.309981144765221e-07, + "loss": 0.7427, + "num_input_tokens_seen": 135972130, + "step": 6401 + }, + { + "epoch": 0.7697949858714603, + "flos": 8330948253840.0, + "grad_norm": 5.154455624094723, + "learning_rate": 5.304696097812196e-07, + "loss": 0.7348, + "num_input_tokens_seen": 135988450, + "step": 6402 + }, + { + "epoch": 0.7699152287620994, + "flos": 19564313173200.0, + "grad_norm": 6.466071269885318, + "learning_rate": 5.299413280125078e-07, + "loss": 0.5732, + "num_input_tokens_seen": 136006480, + "step": 6403 + }, + { + "epoch": 0.7700354716527386, + "flos": 12057528897120.0, + "grad_norm": 4.384518323691456, + "learning_rate": 5.294132692505284e-07, + "loss": 0.7136, + "num_input_tokens_seen": 136024610, + "step": 6404 + }, + { + "epoch": 0.7701557145433776, + "flos": 14053768593960.0, + "grad_norm": 4.270436685642815, + "learning_rate": 5.288854335753861e-07, + "loss": 0.7643, + "num_input_tokens_seen": 136042590, + "step": 6405 + }, + { + "epoch": 0.7702759574340167, + "flos": 23262625963560.0, + "grad_norm": 3.7784004508576134, + "learning_rate": 5.283578210671551e-07, + "loss": 0.7485, + "num_input_tokens_seen": 136064550, + "step": 6406 + }, + { + "epoch": 0.7703962003246558, + "flos": 12246317459040.0, + "grad_norm": 4.093815460846497, + "learning_rate": 5.278304318058719e-07, + "loss": 0.7474, + "num_input_tokens_seen": 136082125, + "step": 6407 + }, + { + "epoch": 0.7705164432152949, + "flos": 26260198994760.0, + "grad_norm": 4.366880457257619, + "learning_rate": 5.273032658715411e-07, + "loss": 0.7769, + "num_input_tokens_seen": 136104655, + "step": 6408 + }, + { + "epoch": 0.7706366861059339, + "flos": 17105281214280.0, + "grad_norm": 6.525644917972398, + "learning_rate": 5.267763233441347e-07, + "loss": 0.757, + "num_input_tokens_seen": 136125005, + "step": 6409 + }, + { + "epoch": 0.7707569289965731, + "flos": 16782815098080.0, + "grad_norm": 9.153886196983455, + "learning_rate": 5.26249604303588e-07, + "loss": 0.6793, + "num_input_tokens_seen": 136143230, + "step": 6410 + }, + { + "epoch": 0.7708771718872122, + "flos": 12705627125520.0, + "grad_norm": 4.878144448319155, + "learning_rate": 5.257231088298057e-07, + "loss": 0.7636, + "num_input_tokens_seen": 136161360, + "step": 6411 + }, + { + "epoch": 0.7709974147778512, + "flos": 53279288112480.0, + "grad_norm": 0.840504940492723, + "learning_rate": 5.25196837002655e-07, + "loss": 0.5601, + "num_input_tokens_seen": 136220790, + "step": 6412 + }, + { + "epoch": 0.7711176576684904, + "flos": 29312059874640.0, + "grad_norm": 4.5355553529440185, + "learning_rate": 5.24670788901971e-07, + "loss": 0.6821, + "num_input_tokens_seen": 136243600, + "step": 6413 + }, + { + "epoch": 0.7712379005591294, + "flos": 27178090148640.0, + "grad_norm": 4.200906501308885, + "learning_rate": 5.241449646075557e-07, + "loss": 0.6547, + "num_input_tokens_seen": 136266545, + "step": 6414 + }, + { + "epoch": 0.7713581434497685, + "flos": 16668477288000.0, + "grad_norm": 7.081022296958225, + "learning_rate": 5.236193641991762e-07, + "loss": 0.7085, + "num_input_tokens_seen": 136284195, + "step": 6415 + }, + { + "epoch": 0.7714783863404077, + "flos": 17646893302680.0, + "grad_norm": 5.744419209178701, + "learning_rate": 5.23093987756565e-07, + "loss": 0.6815, + "num_input_tokens_seen": 136302610, + "step": 6416 + }, + { + "epoch": 0.7715986292310467, + "flos": 15400896943680.0, + "grad_norm": 3.587805566833352, + "learning_rate": 5.225688353594217e-07, + "loss": 0.745, + "num_input_tokens_seen": 136321960, + "step": 6417 + }, + { + "epoch": 0.7717188721216858, + "flos": 15054753668760.0, + "grad_norm": 5.105451004476142, + "learning_rate": 5.220439070874108e-07, + "loss": 0.7567, + "num_input_tokens_seen": 136340920, + "step": 6418 + }, + { + "epoch": 0.7718391150123249, + "flos": 19240738958400.0, + "grad_norm": 4.325822679191398, + "learning_rate": 5.215192030201652e-07, + "loss": 0.6971, + "num_input_tokens_seen": 136361630, + "step": 6419 + }, + { + "epoch": 0.771959357902964, + "flos": 16130822694600.0, + "grad_norm": 3.2365607648260863, + "learning_rate": 5.209947232372798e-07, + "loss": 0.844, + "num_input_tokens_seen": 136378840, + "step": 6420 + }, + { + "epoch": 0.772079600793603, + "flos": 22345241369040.0, + "grad_norm": 8.912404072420507, + "learning_rate": 5.204704678183196e-07, + "loss": 0.7898, + "num_input_tokens_seen": 136397295, + "step": 6421 + }, + { + "epoch": 0.7721998436842422, + "flos": 9411797933640.0, + "grad_norm": 4.897106616513569, + "learning_rate": 5.19946436842813e-07, + "loss": 0.8353, + "num_input_tokens_seen": 136414145, + "step": 6422 + }, + { + "epoch": 0.7723200865748813, + "flos": 23965676899800.0, + "grad_norm": 4.449341140148831, + "learning_rate": 5.194226303902546e-07, + "loss": 0.6779, + "num_input_tokens_seen": 136433600, + "step": 6423 + }, + { + "epoch": 0.7724403294655203, + "flos": 15432805692000.0, + "grad_norm": 5.760458018848333, + "learning_rate": 5.188990485401072e-07, + "loss": 0.6887, + "num_input_tokens_seen": 136452525, + "step": 6424 + }, + { + "epoch": 0.7725605723561595, + "flos": 16161781644120.0, + "grad_norm": 3.4273435886123025, + "learning_rate": 5.183756913717954e-07, + "loss": 0.8369, + "num_input_tokens_seen": 136472020, + "step": 6425 + }, + { + "epoch": 0.7726808152467985, + "flos": 25342181201040.0, + "grad_norm": 3.7563403888306306, + "learning_rate": 5.178525589647136e-07, + "loss": 0.7169, + "num_input_tokens_seen": 136493380, + "step": 6426 + }, + { + "epoch": 0.7728010581374376, + "flos": 16320877654920.0, + "grad_norm": 4.456560489490115, + "learning_rate": 5.173296513982197e-07, + "loss": 0.7691, + "num_input_tokens_seen": 136511625, + "step": 6427 + }, + { + "epoch": 0.7729213010280768, + "flos": 19888647227040.0, + "grad_norm": 9.272031373755075, + "learning_rate": 5.168069687516398e-07, + "loss": 0.6406, + "num_input_tokens_seen": 136531115, + "step": 6428 + }, + { + "epoch": 0.7730415439187158, + "flos": 13245529576080.0, + "grad_norm": 3.3056930563259503, + "learning_rate": 5.16284511104263e-07, + "loss": 0.6969, + "num_input_tokens_seen": 136549970, + "step": 6429 + }, + { + "epoch": 0.7731617868093549, + "flos": 8655060687960.0, + "grad_norm": 5.589831636790247, + "learning_rate": 5.157622785353457e-07, + "loss": 0.7632, + "num_input_tokens_seen": 136567805, + "step": 6430 + }, + { + "epoch": 0.7732820296999939, + "flos": 47332197378000.0, + "grad_norm": 0.6707757340654712, + "learning_rate": 5.152402711241113e-07, + "loss": 0.6129, + "num_input_tokens_seen": 136635430, + "step": 6431 + }, + { + "epoch": 0.7734022725906331, + "flos": 18538574501040.0, + "grad_norm": 4.93181969043723, + "learning_rate": 5.147184889497465e-07, + "loss": 0.8228, + "num_input_tokens_seen": 136654620, + "step": 6432 + }, + { + "epoch": 0.7735225154812722, + "flos": 12651402596760.0, + "grad_norm": 5.768909707550222, + "learning_rate": 5.141969320914072e-07, + "loss": 0.7839, + "num_input_tokens_seen": 136671845, + "step": 6433 + }, + { + "epoch": 0.7736427583719112, + "flos": 23960896245840.0, + "grad_norm": 8.184464452487006, + "learning_rate": 5.136756006282113e-07, + "loss": 0.6045, + "num_input_tokens_seen": 136690230, + "step": 6434 + }, + { + "epoch": 0.7737630012625504, + "flos": 14510228864040.0, + "grad_norm": 4.260270891128392, + "learning_rate": 5.131544946392446e-07, + "loss": 0.8482, + "num_input_tokens_seen": 136705230, + "step": 6435 + }, + { + "epoch": 0.7738832441531894, + "flos": 26471524956840.0, + "grad_norm": 15.802920609959651, + "learning_rate": 5.126336142035592e-07, + "loss": 0.6271, + "num_input_tokens_seen": 136724985, + "step": 6436 + }, + { + "epoch": 0.7740034870438285, + "flos": 9733694170560.0, + "grad_norm": 5.631239926819254, + "learning_rate": 5.121129594001721e-07, + "loss": 0.7134, + "num_input_tokens_seen": 136738970, + "step": 6437 + }, + { + "epoch": 0.7741237299344677, + "flos": 16158235728600.0, + "grad_norm": 3.331202413995621, + "learning_rate": 5.115925303080661e-07, + "loss": 0.8002, + "num_input_tokens_seen": 136758400, + "step": 6438 + }, + { + "epoch": 0.7742439728251067, + "flos": 14514376318800.0, + "grad_norm": 3.148646042498258, + "learning_rate": 5.110723270061899e-07, + "loss": 0.7699, + "num_input_tokens_seen": 136774610, + "step": 6439 + }, + { + "epoch": 0.7743642157157458, + "flos": 12164901516000.0, + "grad_norm": 5.574819402805921, + "learning_rate": 5.105523495734572e-07, + "loss": 0.7833, + "num_input_tokens_seen": 136791730, + "step": 6440 + }, + { + "epoch": 0.7744844586063849, + "flos": 14839945111080.0, + "grad_norm": 3.1108428710769487, + "learning_rate": 5.100325980887499e-07, + "loss": 0.728, + "num_input_tokens_seen": 136811375, + "step": 6441 + }, + { + "epoch": 0.774604701497024, + "flos": 16810101492240.0, + "grad_norm": 3.62665112480405, + "learning_rate": 5.095130726309116e-07, + "loss": 0.8142, + "num_input_tokens_seen": 136831270, + "step": 6442 + }, + { + "epoch": 0.774724944387663, + "flos": 44435475013920.0, + "grad_norm": 0.8601702738354086, + "learning_rate": 5.089937732787559e-07, + "loss": 0.6572, + "num_input_tokens_seen": 136895550, + "step": 6443 + }, + { + "epoch": 0.7748451872783022, + "flos": 19618126122480.0, + "grad_norm": 5.13065751589503, + "learning_rate": 5.084747001110592e-07, + "loss": 0.6437, + "num_input_tokens_seen": 136914895, + "step": 6444 + }, + { + "epoch": 0.7749654301689413, + "flos": 22266073283160.0, + "grad_norm": 2.7587559628821894, + "learning_rate": 5.07955853206564e-07, + "loss": 0.6843, + "num_input_tokens_seen": 136939320, + "step": 6445 + }, + { + "epoch": 0.7750856730595803, + "flos": 31769192235960.0, + "grad_norm": 5.2712218652797125, + "learning_rate": 5.074372326439807e-07, + "loss": 0.6901, + "num_input_tokens_seen": 136962050, + "step": 6446 + }, + { + "epoch": 0.7752059159502195, + "flos": 12868047432120.0, + "grad_norm": 3.3105831718142955, + "learning_rate": 5.069188385019814e-07, + "loss": 0.7201, + "num_input_tokens_seen": 136979470, + "step": 6447 + }, + { + "epoch": 0.7753261588408585, + "flos": 9195533017800.0, + "grad_norm": 12.82436807425248, + "learning_rate": 5.064006708592077e-07, + "loss": 0.577, + "num_input_tokens_seen": 136995435, + "step": 6448 + }, + { + "epoch": 0.7754464017314976, + "flos": 12165408075360.0, + "grad_norm": 3.6504706980667367, + "learning_rate": 5.058827297942641e-07, + "loss": 0.7388, + "num_input_tokens_seen": 137010260, + "step": 6449 + }, + { + "epoch": 0.7755666446221368, + "flos": 14406465480600.0, + "grad_norm": 8.731445694780357, + "learning_rate": 5.053650153857237e-07, + "loss": 0.7152, + "num_input_tokens_seen": 137028990, + "step": 6450 + }, + { + "epoch": 0.7756868875127758, + "flos": 13647132118200.0, + "grad_norm": 5.675601124120439, + "learning_rate": 5.048475277121214e-07, + "loss": 0.6888, + "num_input_tokens_seen": 137045925, + "step": 6451 + }, + { + "epoch": 0.7758071304034149, + "flos": 20834362994400.0, + "grad_norm": 4.403611495660342, + "learning_rate": 5.043302668519598e-07, + "loss": 0.7587, + "num_input_tokens_seen": 137064980, + "step": 6452 + }, + { + "epoch": 0.775927373294054, + "flos": 15055165248240.0, + "grad_norm": 4.569605272136713, + "learning_rate": 5.038132328837079e-07, + "loss": 0.7104, + "num_input_tokens_seen": 137083090, + "step": 6453 + }, + { + "epoch": 0.7760476161846931, + "flos": 16483867840800.0, + "grad_norm": 10.298110818289986, + "learning_rate": 5.032964258857993e-07, + "loss": 0.7238, + "num_input_tokens_seen": 137102905, + "step": 6454 + }, + { + "epoch": 0.7761678590753321, + "flos": 35818397295240.0, + "grad_norm": 3.285768431460237, + "learning_rate": 5.027798459366329e-07, + "loss": 0.6675, + "num_input_tokens_seen": 137127990, + "step": 6455 + }, + { + "epoch": 0.7762881019659713, + "flos": 19185912890400.0, + "grad_norm": 5.3478552681648255, + "learning_rate": 5.02263493114573e-07, + "loss": 0.6247, + "num_input_tokens_seen": 137149505, + "step": 6456 + }, + { + "epoch": 0.7764083448566104, + "flos": 15048991556040.0, + "grad_norm": 7.6606843524135915, + "learning_rate": 5.017473674979502e-07, + "loss": 0.7482, + "num_input_tokens_seen": 137165250, + "step": 6457 + }, + { + "epoch": 0.7765285877472494, + "flos": 49619216062440.0, + "grad_norm": 5.527302858450773, + "learning_rate": 5.01231469165061e-07, + "loss": 0.5916, + "num_input_tokens_seen": 137220795, + "step": 6458 + }, + { + "epoch": 0.7766488306378886, + "flos": 45216649257360.0, + "grad_norm": 0.948907275174996, + "learning_rate": 5.007157981941663e-07, + "loss": 0.6072, + "num_input_tokens_seen": 137285875, + "step": 6459 + }, + { + "epoch": 0.7767690735285276, + "flos": 46401515600280.0, + "grad_norm": 1.1664000217810577, + "learning_rate": 5.002003546634928e-07, + "loss": 0.6915, + "num_input_tokens_seen": 137341695, + "step": 6460 + }, + { + "epoch": 0.7768893164191667, + "flos": 15270448705320.0, + "grad_norm": 3.6041954495083695, + "learning_rate": 4.996851386512331e-07, + "loss": 0.7514, + "num_input_tokens_seen": 137360120, + "step": 6461 + }, + { + "epoch": 0.7770095593098058, + "flos": 15135979652040.0, + "grad_norm": 3.625637129993661, + "learning_rate": 4.991701502355444e-07, + "loss": 0.8193, + "num_input_tokens_seen": 137380305, + "step": 6462 + }, + { + "epoch": 0.7771298022004449, + "flos": 18105918029520.0, + "grad_norm": 2.457018150209763, + "learning_rate": 4.986553894945518e-07, + "loss": 0.7549, + "num_input_tokens_seen": 137401235, + "step": 6463 + }, + { + "epoch": 0.777250045091084, + "flos": 18321613066080.0, + "grad_norm": 9.89935775369023, + "learning_rate": 4.981408565063416e-07, + "loss": 0.8619, + "num_input_tokens_seen": 137420900, + "step": 6464 + }, + { + "epoch": 0.777370287981723, + "flos": 14703133220760.0, + "grad_norm": 4.447682906834309, + "learning_rate": 4.976265513489701e-07, + "loss": 0.7418, + "num_input_tokens_seen": 137440590, + "step": 6465 + }, + { + "epoch": 0.7774905308723622, + "flos": 15885530086800.0, + "grad_norm": 4.474281781916429, + "learning_rate": 4.971124741004562e-07, + "loss": 0.8018, + "num_input_tokens_seen": 137459310, + "step": 6466 + }, + { + "epoch": 0.7776107737630013, + "flos": 11679476873880.0, + "grad_norm": 5.176729902743808, + "learning_rate": 4.965986248387846e-07, + "loss": 0.7495, + "num_input_tokens_seen": 137477345, + "step": 6467 + }, + { + "epoch": 0.7777310166536403, + "flos": 18160110898320.0, + "grad_norm": 2.78030207906963, + "learning_rate": 4.960850036419073e-07, + "loss": 0.7597, + "num_input_tokens_seen": 137496165, + "step": 6468 + }, + { + "epoch": 0.7778512595442795, + "flos": 12595690049880.0, + "grad_norm": 3.330878895885465, + "learning_rate": 4.955716105877378e-07, + "loss": 0.774, + "num_input_tokens_seen": 137514655, + "step": 6469 + }, + { + "epoch": 0.7779715024349185, + "flos": 12948101996880.0, + "grad_norm": 2.9543201107174144, + "learning_rate": 4.950584457541598e-07, + "loss": 0.8249, + "num_input_tokens_seen": 137532840, + "step": 6470 + }, + { + "epoch": 0.7780917453255576, + "flos": 17809345269240.0, + "grad_norm": 4.477806038421563, + "learning_rate": 4.945455092190183e-07, + "loss": 0.8117, + "num_input_tokens_seen": 137553815, + "step": 6471 + }, + { + "epoch": 0.7782119882161967, + "flos": 41533849847400.0, + "grad_norm": 0.7870195209718998, + "learning_rate": 4.940328010601271e-07, + "loss": 0.5697, + "num_input_tokens_seen": 137618450, + "step": 6472 + }, + { + "epoch": 0.7783322311068358, + "flos": 34440658255560.0, + "grad_norm": 3.379548151862079, + "learning_rate": 4.935203213552621e-07, + "loss": 0.7461, + "num_input_tokens_seen": 137641910, + "step": 6473 + }, + { + "epoch": 0.7784524739974749, + "flos": 13916608444080.0, + "grad_norm": 5.080531656273819, + "learning_rate": 4.930080701821662e-07, + "loss": 0.6444, + "num_input_tokens_seen": 137659095, + "step": 6474 + }, + { + "epoch": 0.778572716888114, + "flos": 18160300858080.0, + "grad_norm": 4.012172060485983, + "learning_rate": 4.92496047618548e-07, + "loss": 0.7528, + "num_input_tokens_seen": 137678575, + "step": 6475 + }, + { + "epoch": 0.7786929597787531, + "flos": 14671446092160.0, + "grad_norm": 4.396955928315279, + "learning_rate": 4.919842537420811e-07, + "loss": 0.7672, + "num_input_tokens_seen": 137695410, + "step": 6476 + }, + { + "epoch": 0.7788132026693921, + "flos": 16000121176560.0, + "grad_norm": 4.509551402109372, + "learning_rate": 4.91472688630404e-07, + "loss": 0.7845, + "num_input_tokens_seen": 137715870, + "step": 6477 + }, + { + "epoch": 0.7789334455600313, + "flos": 8088600022320.0, + "grad_norm": 4.342195081618404, + "learning_rate": 4.909613523611202e-07, + "loss": 0.7305, + "num_input_tokens_seen": 137732470, + "step": 6478 + }, + { + "epoch": 0.7790536884506704, + "flos": 21049804751280.0, + "grad_norm": 4.117019794675039, + "learning_rate": 4.904502450117991e-07, + "loss": 0.7235, + "num_input_tokens_seen": 137753150, + "step": 6479 + }, + { + "epoch": 0.7791739313413094, + "flos": 8007437358960.0, + "grad_norm": 4.641957306908386, + "learning_rate": 4.899393666599762e-07, + "loss": 0.7097, + "num_input_tokens_seen": 137769445, + "step": 6480 + }, + { + "epoch": 0.7792941742319486, + "flos": 10676117302080.0, + "grad_norm": 3.895419556102578, + "learning_rate": 4.894287173831506e-07, + "loss": 0.7073, + "num_input_tokens_seen": 137785125, + "step": 6481 + }, + { + "epoch": 0.7794144171225876, + "flos": 17026366408080.0, + "grad_norm": 3.3654350092541785, + "learning_rate": 4.889182972587877e-07, + "loss": 0.8262, + "num_input_tokens_seen": 137804140, + "step": 6482 + }, + { + "epoch": 0.7795346600132267, + "flos": 15729853351680.0, + "grad_norm": 6.291932999319043, + "learning_rate": 4.884081063643177e-07, + "loss": 0.6572, + "num_input_tokens_seen": 137822520, + "step": 6483 + }, + { + "epoch": 0.7796549029038659, + "flos": 51659264161200.0, + "grad_norm": 0.8827295517563898, + "learning_rate": 4.878981447771353e-07, + "loss": 0.5589, + "num_input_tokens_seen": 137876620, + "step": 6484 + }, + { + "epoch": 0.7797751457945049, + "flos": 17567123677560.0, + "grad_norm": 4.963182316914975, + "learning_rate": 4.873884125746035e-07, + "loss": 0.7193, + "num_input_tokens_seen": 137898015, + "step": 6485 + }, + { + "epoch": 0.779895388685144, + "flos": 16594026536160.0, + "grad_norm": 4.249275370138378, + "learning_rate": 4.868789098340456e-07, + "loss": 0.7042, + "num_input_tokens_seen": 137915640, + "step": 6486 + }, + { + "epoch": 0.7800156315757831, + "flos": 17403531952440.0, + "grad_norm": 18.99320701834729, + "learning_rate": 4.863696366327543e-07, + "loss": 0.7002, + "num_input_tokens_seen": 137934530, + "step": 6487 + }, + { + "epoch": 0.7801358744664222, + "flos": 19372516915080.0, + "grad_norm": 4.112153365069396, + "learning_rate": 4.85860593047986e-07, + "loss": 0.7759, + "num_input_tokens_seen": 137954315, + "step": 6488 + }, + { + "epoch": 0.7802561173570612, + "flos": 19293190529400.0, + "grad_norm": 4.333760142703713, + "learning_rate": 4.853517791569613e-07, + "loss": 0.7297, + "num_input_tokens_seen": 137976215, + "step": 6489 + }, + { + "epoch": 0.7803763602477004, + "flos": 29662857163680.0, + "grad_norm": 7.396509388573363, + "learning_rate": 4.848431950368684e-07, + "loss": 0.6515, + "num_input_tokens_seen": 137998495, + "step": 6490 + }, + { + "epoch": 0.7804966031383395, + "flos": 49403426046000.0, + "grad_norm": 0.7506962564332612, + "learning_rate": 4.843348407648569e-07, + "loss": 0.5771, + "num_input_tokens_seen": 138059495, + "step": 6491 + }, + { + "epoch": 0.7806168460289785, + "flos": 12941326765440.0, + "grad_norm": 4.402009479115605, + "learning_rate": 4.838267164180457e-07, + "loss": 0.8122, + "num_input_tokens_seen": 138074885, + "step": 6492 + }, + { + "epoch": 0.7807370889196176, + "flos": 17534233470480.0, + "grad_norm": 5.349618110217859, + "learning_rate": 4.833188220735156e-07, + "loss": 0.8242, + "num_input_tokens_seen": 138094275, + "step": 6493 + }, + { + "epoch": 0.7808573318102567, + "flos": 13866246430440.0, + "grad_norm": 4.850242742382507, + "learning_rate": 4.828111578083152e-07, + "loss": 0.7366, + "num_input_tokens_seen": 138110900, + "step": 6494 + }, + { + "epoch": 0.7809775747008958, + "flos": 17559493627200.0, + "grad_norm": 5.191643955595368, + "learning_rate": 4.823037236994556e-07, + "loss": 0.797, + "num_input_tokens_seen": 138128785, + "step": 6495 + }, + { + "epoch": 0.7810978175915348, + "flos": 50539291753560.0, + "grad_norm": 0.7803693767285734, + "learning_rate": 4.817965198239136e-07, + "loss": 0.5837, + "num_input_tokens_seen": 138194965, + "step": 6496 + }, + { + "epoch": 0.781218060482174, + "flos": 14349074955840.0, + "grad_norm": 5.463628321549257, + "learning_rate": 4.812895462586331e-07, + "loss": 0.7063, + "num_input_tokens_seen": 138212510, + "step": 6497 + }, + { + "epoch": 0.7813383033728131, + "flos": 18779213094720.0, + "grad_norm": 5.0625162006636995, + "learning_rate": 4.807828030805207e-07, + "loss": 0.8033, + "num_input_tokens_seen": 138231220, + "step": 6498 + }, + { + "epoch": 0.7814585462634521, + "flos": 14974192544640.0, + "grad_norm": 11.256836911775029, + "learning_rate": 4.802762903664495e-07, + "loss": 0.6651, + "num_input_tokens_seen": 138250120, + "step": 6499 + }, + { + "epoch": 0.7815787891540913, + "flos": 16319073037200.0, + "grad_norm": 5.9145626929947435, + "learning_rate": 4.797700081932565e-07, + "loss": 0.705, + "num_input_tokens_seen": 138267705, + "step": 6500 + }, + { + "epoch": 0.7816990320447303, + "flos": 16539042168360.0, + "grad_norm": 4.669906354532971, + "learning_rate": 4.792639566377442e-07, + "loss": 0.7996, + "num_input_tokens_seen": 138284835, + "step": 6501 + }, + { + "epoch": 0.7818192749353694, + "flos": 18267356877360.0, + "grad_norm": 3.2964173491971387, + "learning_rate": 4.78758135776681e-07, + "loss": 0.7604, + "num_input_tokens_seen": 138304410, + "step": 6502 + }, + { + "epoch": 0.7819395178260086, + "flos": 17377258677000.0, + "grad_norm": 5.881095159359139, + "learning_rate": 4.782525456867989e-07, + "loss": 0.7682, + "num_input_tokens_seen": 138322985, + "step": 6503 + }, + { + "epoch": 0.7820597607166476, + "flos": 16998731754360.0, + "grad_norm": 2.8604379147941823, + "learning_rate": 4.777471864447959e-07, + "loss": 0.8173, + "num_input_tokens_seen": 138343445, + "step": 6504 + }, + { + "epoch": 0.7821800036072867, + "flos": 16323695391360.0, + "grad_norm": 5.60043481092416, + "learning_rate": 4.772420581273344e-07, + "loss": 0.7754, + "num_input_tokens_seen": 138360650, + "step": 6505 + }, + { + "epoch": 0.7823002464979258, + "flos": 15757393025520.0, + "grad_norm": 7.97192176078382, + "learning_rate": 4.7673716081104134e-07, + "loss": 0.7563, + "num_input_tokens_seen": 138380545, + "step": 6506 + }, + { + "epoch": 0.7824204893885649, + "flos": 17971227356520.0, + "grad_norm": 3.1752944329176427, + "learning_rate": 4.762324945725109e-07, + "loss": 0.8288, + "num_input_tokens_seen": 138399710, + "step": 6507 + }, + { + "epoch": 0.782540732279204, + "flos": 20101904446680.0, + "grad_norm": 7.708844582039455, + "learning_rate": 4.7572805948829844e-07, + "loss": 0.7499, + "num_input_tokens_seen": 138419690, + "step": 6508 + }, + { + "epoch": 0.7826609751698431, + "flos": 17836220083920.0, + "grad_norm": 3.5073201061183776, + "learning_rate": 4.7522385563492795e-07, + "loss": 0.6933, + "num_input_tokens_seen": 138439710, + "step": 6509 + }, + { + "epoch": 0.7827812180604822, + "flos": 17566363838520.0, + "grad_norm": 3.167500151985725, + "learning_rate": 4.747198830888863e-07, + "loss": 0.6763, + "num_input_tokens_seen": 138459300, + "step": 6510 + }, + { + "epoch": 0.7829014609511212, + "flos": 20131217078280.0, + "grad_norm": 6.495228703490319, + "learning_rate": 4.742161419266251e-07, + "loss": 0.6731, + "num_input_tokens_seen": 138478180, + "step": 6511 + }, + { + "epoch": 0.7830217038417604, + "flos": 21426337096440.0, + "grad_norm": 4.585978675150545, + "learning_rate": 4.7371263222456304e-07, + "loss": 0.6203, + "num_input_tokens_seen": 138495220, + "step": 6512 + }, + { + "epoch": 0.7831419467323995, + "flos": 44926471809000.0, + "grad_norm": 0.8268031601034785, + "learning_rate": 4.7320935405908004e-07, + "loss": 0.6339, + "num_input_tokens_seen": 138555810, + "step": 6513 + }, + { + "epoch": 0.7832621896230385, + "flos": 14379527346000.0, + "grad_norm": 4.911260876642321, + "learning_rate": 4.7270630750652475e-07, + "loss": 0.8114, + "num_input_tokens_seen": 138571485, + "step": 6514 + }, + { + "epoch": 0.7833824325136777, + "flos": 18321613066080.0, + "grad_norm": 5.524654443213525, + "learning_rate": 4.7220349264320746e-07, + "loss": 0.7966, + "num_input_tokens_seen": 138590290, + "step": 6515 + }, + { + "epoch": 0.7835026754043167, + "flos": 50736121945320.0, + "grad_norm": 1.0450831208641136, + "learning_rate": 4.71700909545407e-07, + "loss": 0.584, + "num_input_tokens_seen": 138652955, + "step": 6516 + }, + { + "epoch": 0.7836229182949558, + "flos": 14514218019000.0, + "grad_norm": 4.883192953928092, + "learning_rate": 4.711985582893627e-07, + "loss": 0.7497, + "num_input_tokens_seen": 138671195, + "step": 6517 + }, + { + "epoch": 0.783743161185595, + "flos": 16809341653200.0, + "grad_norm": 6.251510132501393, + "learning_rate": 4.706964389512811e-07, + "loss": 0.7021, + "num_input_tokens_seen": 138690950, + "step": 6518 + }, + { + "epoch": 0.783863404076234, + "flos": 9033271011000.0, + "grad_norm": 4.409673564705853, + "learning_rate": 4.701945516073345e-07, + "loss": 0.8568, + "num_input_tokens_seen": 138708145, + "step": 6519 + }, + { + "epoch": 0.7839836469668731, + "flos": 17755374020160.0, + "grad_norm": 4.408645952433823, + "learning_rate": 4.696928963336577e-07, + "loss": 0.7412, + "num_input_tokens_seen": 138727295, + "step": 6520 + }, + { + "epoch": 0.7841038898575122, + "flos": 45793652689680.0, + "grad_norm": 0.8737611174524214, + "learning_rate": 4.6919147320635224e-07, + "loss": 0.6252, + "num_input_tokens_seen": 138789725, + "step": 6521 + }, + { + "epoch": 0.7842241327481513, + "flos": 14758054268640.0, + "grad_norm": 3.8350554343622396, + "learning_rate": 4.6869028230148286e-07, + "loss": 0.7014, + "num_input_tokens_seen": 138807240, + "step": 6522 + }, + { + "epoch": 0.7843443756387903, + "flos": 20590210145160.0, + "grad_norm": 7.263959139923256, + "learning_rate": 4.6818932369507957e-07, + "loss": 0.583, + "num_input_tokens_seen": 138826460, + "step": 6523 + }, + { + "epoch": 0.7844646185294295, + "flos": 15593737980480.0, + "grad_norm": 7.814962746597893, + "learning_rate": 4.676885974631386e-07, + "loss": 0.8674, + "num_input_tokens_seen": 138844540, + "step": 6524 + }, + { + "epoch": 0.7845848614200686, + "flos": 17320184751840.0, + "grad_norm": 3.9210024163127555, + "learning_rate": 4.67188103681619e-07, + "loss": 0.7937, + "num_input_tokens_seen": 138864045, + "step": 6525 + }, + { + "epoch": 0.7847051043107076, + "flos": 17132061049080.0, + "grad_norm": 4.0835488450665, + "learning_rate": 4.666878424264453e-07, + "loss": 0.6859, + "num_input_tokens_seen": 138883720, + "step": 6526 + }, + { + "epoch": 0.7848253472013467, + "flos": 13890398488560.0, + "grad_norm": 5.979694657711453, + "learning_rate": 4.661878137735069e-07, + "loss": 0.7169, + "num_input_tokens_seen": 138901630, + "step": 6527 + }, + { + "epoch": 0.7849455900919858, + "flos": 15487251840480.0, + "grad_norm": 4.67383730704541, + "learning_rate": 4.656880177986571e-07, + "loss": 0.7351, + "num_input_tokens_seen": 138919895, + "step": 6528 + }, + { + "epoch": 0.7850658329826249, + "flos": 14271426548040.0, + "grad_norm": 5.142033253630835, + "learning_rate": 4.6518845457771607e-07, + "loss": 0.7942, + "num_input_tokens_seen": 138938475, + "step": 6529 + }, + { + "epoch": 0.7851860758732639, + "flos": 9060589065120.0, + "grad_norm": 6.1236165878299476, + "learning_rate": 4.646891241864652e-07, + "loss": 0.7733, + "num_input_tokens_seen": 138956760, + "step": 6530 + }, + { + "epoch": 0.7853063187639031, + "flos": 16805035898640.0, + "grad_norm": 4.398325618859036, + "learning_rate": 4.6419002670065397e-07, + "loss": 0.7149, + "num_input_tokens_seen": 138976060, + "step": 6531 + }, + { + "epoch": 0.7854265616545422, + "flos": 12651244296960.0, + "grad_norm": 5.700627864589752, + "learning_rate": 4.6369116219599445e-07, + "loss": 0.828, + "num_input_tokens_seen": 138991765, + "step": 6532 + }, + { + "epoch": 0.7855468045451812, + "flos": 17454558825240.0, + "grad_norm": 3.8940720818367036, + "learning_rate": 4.631925307481637e-07, + "loss": 0.7677, + "num_input_tokens_seen": 139011300, + "step": 6533 + }, + { + "epoch": 0.7856670474358204, + "flos": 18808367426520.0, + "grad_norm": 9.487570813383416, + "learning_rate": 4.6269413243280533e-07, + "loss": 0.7246, + "num_input_tokens_seen": 139030440, + "step": 6534 + }, + { + "epoch": 0.7857872903264594, + "flos": 13240622282280.0, + "grad_norm": 3.408011444735227, + "learning_rate": 4.621959673255236e-07, + "loss": 0.7111, + "num_input_tokens_seen": 139046460, + "step": 6535 + }, + { + "epoch": 0.7859075332170985, + "flos": 10458997567320.0, + "grad_norm": 7.2989295174587125, + "learning_rate": 4.6169803550189135e-07, + "loss": 0.8782, + "num_input_tokens_seen": 139061875, + "step": 6536 + }, + { + "epoch": 0.7860277761077377, + "flos": 14513743119600.0, + "grad_norm": 3.519105807214817, + "learning_rate": 4.6120033703744355e-07, + "loss": 0.7477, + "num_input_tokens_seen": 139080490, + "step": 6537 + }, + { + "epoch": 0.7861480189983767, + "flos": 19348111577280.0, + "grad_norm": 7.453818490895444, + "learning_rate": 4.607028720076822e-07, + "loss": 0.7653, + "num_input_tokens_seen": 139096890, + "step": 6538 + }, + { + "epoch": 0.7862682618890158, + "flos": 17748883728360.0, + "grad_norm": 4.609662994181014, + "learning_rate": 4.6020564048807074e-07, + "loss": 0.7146, + "num_input_tokens_seen": 139114285, + "step": 6539 + }, + { + "epoch": 0.7863885047796549, + "flos": 35003573005680.0, + "grad_norm": 3.8488082477935546, + "learning_rate": 4.5970864255403883e-07, + "loss": 0.6993, + "num_input_tokens_seen": 139135530, + "step": 6540 + }, + { + "epoch": 0.786508747670294, + "flos": 17862809958960.0, + "grad_norm": 3.306255716993152, + "learning_rate": 4.59211878280982e-07, + "loss": 0.8079, + "num_input_tokens_seen": 139154765, + "step": 6541 + }, + { + "epoch": 0.786628990560933, + "flos": 13164936792000.0, + "grad_norm": 6.722427851927102, + "learning_rate": 4.587153477442578e-07, + "loss": 0.6765, + "num_input_tokens_seen": 139170800, + "step": 6542 + }, + { + "epoch": 0.7867492334515722, + "flos": 18943058099520.0, + "grad_norm": 4.202765807505726, + "learning_rate": 4.582190510191899e-07, + "loss": 0.7924, + "num_input_tokens_seen": 139189180, + "step": 6543 + }, + { + "epoch": 0.7868694763422113, + "flos": 12083992132320.0, + "grad_norm": 3.8931178184414867, + "learning_rate": 4.5772298818106625e-07, + "loss": 0.8527, + "num_input_tokens_seen": 139204690, + "step": 6544 + }, + { + "epoch": 0.7869897192328503, + "flos": 21560742829800.0, + "grad_norm": 5.287569317013873, + "learning_rate": 4.572271593051384e-07, + "loss": 0.7029, + "num_input_tokens_seen": 139221765, + "step": 6545 + }, + { + "epoch": 0.7871099621234895, + "flos": 12489742129200.0, + "grad_norm": 3.1794508487449464, + "learning_rate": 4.567315644666245e-07, + "loss": 0.7616, + "num_input_tokens_seen": 139240280, + "step": 6546 + }, + { + "epoch": 0.7872302050141285, + "flos": 17160392221920.0, + "grad_norm": 7.993466570424874, + "learning_rate": 4.5623620374070507e-07, + "loss": 0.8384, + "num_input_tokens_seen": 139259315, + "step": 6547 + }, + { + "epoch": 0.7873504479047676, + "flos": 48634820806680.0, + "grad_norm": 0.8279854638924562, + "learning_rate": 4.557410772025263e-07, + "loss": 0.6202, + "num_input_tokens_seen": 139320985, + "step": 6548 + }, + { + "epoch": 0.7874706907954068, + "flos": 17025891508680.0, + "grad_norm": 4.38758152725681, + "learning_rate": 4.5524618492719803e-07, + "loss": 0.6326, + "num_input_tokens_seen": 139339925, + "step": 6549 + }, + { + "epoch": 0.7875909336860458, + "flos": 21103269441000.0, + "grad_norm": 3.0621993248146047, + "learning_rate": 4.54751526989795e-07, + "loss": 0.7709, + "num_input_tokens_seen": 139361485, + "step": 6550 + }, + { + "epoch": 0.7877111765766849, + "flos": 13650361434120.0, + "grad_norm": 6.285006638115994, + "learning_rate": 4.5425710346535775e-07, + "loss": 0.7907, + "num_input_tokens_seen": 139379150, + "step": 6551 + }, + { + "epoch": 0.787831419467324, + "flos": 20232795924480.0, + "grad_norm": 6.125791537462667, + "learning_rate": 4.537629144288877e-07, + "loss": 0.803, + "num_input_tokens_seen": 139396325, + "step": 6552 + }, + { + "epoch": 0.7879516623579631, + "flos": 13245497916120.0, + "grad_norm": 4.574025893604693, + "learning_rate": 4.5326895995535477e-07, + "loss": 0.7277, + "num_input_tokens_seen": 139414945, + "step": 6553 + }, + { + "epoch": 0.7880719052486022, + "flos": 14865490207440.0, + "grad_norm": 5.110180899148433, + "learning_rate": 4.527752401196907e-07, + "loss": 0.8246, + "num_input_tokens_seen": 139432680, + "step": 6554 + }, + { + "epoch": 0.7881921481392413, + "flos": 15838049129520.0, + "grad_norm": 4.68342677845615, + "learning_rate": 4.5228175499679254e-07, + "loss": 0.6554, + "num_input_tokens_seen": 139451985, + "step": 6555 + }, + { + "epoch": 0.7883123910298804, + "flos": 50560562755320.0, + "grad_norm": 0.8668468252937661, + "learning_rate": 4.5178850466152174e-07, + "loss": 0.5563, + "num_input_tokens_seen": 139510535, + "step": 6556 + }, + { + "epoch": 0.7884326339205194, + "flos": 14109987700200.0, + "grad_norm": 9.075711426815642, + "learning_rate": 4.512954891887031e-07, + "loss": 0.801, + "num_input_tokens_seen": 139528555, + "step": 6557 + }, + { + "epoch": 0.7885528768111585, + "flos": 12975103451400.0, + "grad_norm": 5.6736502903499, + "learning_rate": 4.5080270865312806e-07, + "loss": 0.8221, + "num_input_tokens_seen": 139545470, + "step": 6558 + }, + { + "epoch": 0.7886731197017977, + "flos": 13732157296680.0, + "grad_norm": 8.781360340443326, + "learning_rate": 4.5031016312954985e-07, + "loss": 0.6934, + "num_input_tokens_seen": 139563505, + "step": 6559 + }, + { + "epoch": 0.7887933625924367, + "flos": 24505832630040.0, + "grad_norm": 7.580387619293636, + "learning_rate": 4.498178526926886e-07, + "loss": 0.7326, + "num_input_tokens_seen": 139584090, + "step": 6560 + }, + { + "epoch": 0.7889136054830758, + "flos": 12408642785760.0, + "grad_norm": 5.48738489399582, + "learning_rate": 4.4932577741722635e-07, + "loss": 0.7195, + "num_input_tokens_seen": 139602340, + "step": 6561 + }, + { + "epoch": 0.7890338483737149, + "flos": 21589390602240.0, + "grad_norm": 6.594703930737302, + "learning_rate": 4.4883393737780985e-07, + "loss": 0.7407, + "num_input_tokens_seen": 139623010, + "step": 6562 + }, + { + "epoch": 0.789154091264354, + "flos": 14593037845320.0, + "grad_norm": 8.04467174403271, + "learning_rate": 4.4834233264905254e-07, + "loss": 0.7718, + "num_input_tokens_seen": 139639745, + "step": 6563 + }, + { + "epoch": 0.789274334154993, + "flos": 10572068979000.0, + "grad_norm": 4.954156714007768, + "learning_rate": 4.478509633055294e-07, + "loss": 0.6909, + "num_input_tokens_seen": 139657175, + "step": 6564 + }, + { + "epoch": 0.7893945770456322, + "flos": 15965774611320.0, + "grad_norm": 6.111292545438136, + "learning_rate": 4.473598294217813e-07, + "loss": 0.7707, + "num_input_tokens_seen": 139672320, + "step": 6565 + }, + { + "epoch": 0.7895148199362713, + "flos": 15162601187040.0, + "grad_norm": 3.734830820523812, + "learning_rate": 4.468689310723124e-07, + "loss": 0.7085, + "num_input_tokens_seen": 139689855, + "step": 6566 + }, + { + "epoch": 0.7896350628269103, + "flos": 12165344755440.0, + "grad_norm": 5.057972142973573, + "learning_rate": 4.463782683315913e-07, + "loss": 0.7673, + "num_input_tokens_seen": 139708580, + "step": 6567 + }, + { + "epoch": 0.7897553057175495, + "flos": 16567120061520.0, + "grad_norm": 5.065507160489338, + "learning_rate": 4.458878412740523e-07, + "loss": 0.7146, + "num_input_tokens_seen": 139727080, + "step": 6568 + }, + { + "epoch": 0.7898755486081885, + "flos": 10571847359280.0, + "grad_norm": 45.63223936572969, + "learning_rate": 4.453976499740919e-07, + "loss": 0.7481, + "num_input_tokens_seen": 139744445, + "step": 6569 + }, + { + "epoch": 0.7899957914988276, + "flos": 12570841472640.0, + "grad_norm": 3.03466840085727, + "learning_rate": 4.4490769450607215e-07, + "loss": 0.7705, + "num_input_tokens_seen": 139761790, + "step": 6570 + }, + { + "epoch": 0.7901160343894668, + "flos": 30362235544560.0, + "grad_norm": 11.594797585757481, + "learning_rate": 4.4441797494431845e-07, + "loss": 0.7155, + "num_input_tokens_seen": 139783315, + "step": 6571 + }, + { + "epoch": 0.7902362772801058, + "flos": 12273698833080.0, + "grad_norm": 4.52281677520008, + "learning_rate": 4.439284913631207e-07, + "loss": 0.7618, + "num_input_tokens_seen": 139800245, + "step": 6572 + }, + { + "epoch": 0.7903565201707449, + "flos": 19888235647560.0, + "grad_norm": 3.0424609954264668, + "learning_rate": 4.434392438367347e-07, + "loss": 0.8167, + "num_input_tokens_seen": 139819390, + "step": 6573 + }, + { + "epoch": 0.790476763061384, + "flos": 22773212166480.0, + "grad_norm": 4.039613077501108, + "learning_rate": 4.4295023243937677e-07, + "loss": 0.721, + "num_input_tokens_seen": 139839315, + "step": 6574 + }, + { + "epoch": 0.7905970059520231, + "flos": 16160325285960.0, + "grad_norm": 2.5181820591199386, + "learning_rate": 4.4246145724523123e-07, + "loss": 0.7948, + "num_input_tokens_seen": 139856780, + "step": 6575 + }, + { + "epoch": 0.7907172488426621, + "flos": 15027530594520.0, + "grad_norm": 5.605304399267471, + "learning_rate": 4.41972918328444e-07, + "loss": 0.7452, + "num_input_tokens_seen": 139873935, + "step": 6576 + }, + { + "epoch": 0.7908374917333013, + "flos": 22075575083400.0, + "grad_norm": 3.988100830540535, + "learning_rate": 4.4148461576312646e-07, + "loss": 0.7715, + "num_input_tokens_seen": 139893320, + "step": 6577 + }, + { + "epoch": 0.7909577346239404, + "flos": 15351896308320.0, + "grad_norm": 6.103663435710211, + "learning_rate": 4.4099654962335343e-07, + "loss": 0.7323, + "num_input_tokens_seen": 139913490, + "step": 6578 + }, + { + "epoch": 0.7910779775145794, + "flos": 19238079521760.0, + "grad_norm": 3.7766601196349034, + "learning_rate": 4.405087199831636e-07, + "loss": 0.7341, + "num_input_tokens_seen": 139933450, + "step": 6579 + }, + { + "epoch": 0.7911982204052186, + "flos": 16512262333560.0, + "grad_norm": 3.605532760299884, + "learning_rate": 4.400211269165619e-07, + "loss": 0.6603, + "num_input_tokens_seen": 139949625, + "step": 6580 + }, + { + "epoch": 0.7913184632958576, + "flos": 16918012330440.0, + "grad_norm": 4.580363380642464, + "learning_rate": 4.3953377049751416e-07, + "loss": 0.7611, + "num_input_tokens_seen": 139969770, + "step": 6581 + }, + { + "epoch": 0.7914387061864967, + "flos": 8924568673800.0, + "grad_norm": 5.234849230791295, + "learning_rate": 4.390466507999537e-07, + "loss": 0.758, + "num_input_tokens_seen": 139985240, + "step": 6582 + }, + { + "epoch": 0.7915589490771359, + "flos": 12840792697920.0, + "grad_norm": 7.487750922633213, + "learning_rate": 4.385597678977748e-07, + "loss": 0.7476, + "num_input_tokens_seen": 140003795, + "step": 6583 + }, + { + "epoch": 0.7916791919677749, + "flos": 18752116660320.0, + "grad_norm": 3.0011109838101615, + "learning_rate": 4.3807312186483726e-07, + "loss": 0.7312, + "num_input_tokens_seen": 140024235, + "step": 6584 + }, + { + "epoch": 0.791799434858414, + "flos": 13759348710960.0, + "grad_norm": 7.395356871251977, + "learning_rate": 4.375867127749655e-07, + "loss": 0.7695, + "num_input_tokens_seen": 140042230, + "step": 6585 + }, + { + "epoch": 0.7919196777490531, + "flos": 18915740045400.0, + "grad_norm": 7.047688304040425, + "learning_rate": 4.3710054070194744e-07, + "loss": 0.6611, + "num_input_tokens_seen": 140061645, + "step": 6586 + }, + { + "epoch": 0.7920399206396922, + "flos": 8655187327800.0, + "grad_norm": 7.926200934551358, + "learning_rate": 4.3661460571953455e-07, + "loss": 0.6367, + "num_input_tokens_seen": 140078100, + "step": 6587 + }, + { + "epoch": 0.7921601635303313, + "flos": 15783729620880.0, + "grad_norm": 3.7396534919355955, + "learning_rate": 4.36128907901443e-07, + "loss": 0.6733, + "num_input_tokens_seen": 140097415, + "step": 6588 + }, + { + "epoch": 0.7922804064209703, + "flos": 13218749741280.0, + "grad_norm": 6.534559661576526, + "learning_rate": 4.356434473213519e-07, + "loss": 0.7098, + "num_input_tokens_seen": 140114585, + "step": 6589 + }, + { + "epoch": 0.7924006493116095, + "flos": 15837669210000.0, + "grad_norm": 4.0040110504577155, + "learning_rate": 4.351582240529068e-07, + "loss": 0.7681, + "num_input_tokens_seen": 140135135, + "step": 6590 + }, + { + "epoch": 0.7925208922022485, + "flos": 47363821186680.0, + "grad_norm": 0.6984997254373151, + "learning_rate": 4.346732381697149e-07, + "loss": 0.601, + "num_input_tokens_seen": 140198985, + "step": 6591 + }, + { + "epoch": 0.7926411350928876, + "flos": 12350365782120.0, + "grad_norm": 3.4171748519963403, + "learning_rate": 4.3418848974534825e-07, + "loss": 0.797, + "num_input_tokens_seen": 140215645, + "step": 6592 + }, + { + "epoch": 0.7927613779835267, + "flos": 25315939585560.0, + "grad_norm": 5.823264050798509, + "learning_rate": 4.3370397885334276e-07, + "loss": 0.6693, + "num_input_tokens_seen": 140235995, + "step": 6593 + }, + { + "epoch": 0.7928816208741658, + "flos": 13838041897440.0, + "grad_norm": 4.9721276660834075, + "learning_rate": 4.3321970556719777e-07, + "loss": 0.7423, + "num_input_tokens_seen": 140254010, + "step": 6594 + }, + { + "epoch": 0.7930018637648049, + "flos": 13595218766520.0, + "grad_norm": 7.576201330424179, + "learning_rate": 4.3273566996037856e-07, + "loss": 0.6874, + "num_input_tokens_seen": 140270425, + "step": 6595 + }, + { + "epoch": 0.793122106655444, + "flos": 17966826622080.0, + "grad_norm": 4.003640902030564, + "learning_rate": 4.322518721063113e-07, + "loss": 0.7871, + "num_input_tokens_seen": 140288695, + "step": 6596 + }, + { + "epoch": 0.7932423495460831, + "flos": 25286911893600.0, + "grad_norm": 3.8619323165154245, + "learning_rate": 4.3176831207838906e-07, + "loss": 0.6834, + "num_input_tokens_seen": 140311825, + "step": 6597 + }, + { + "epoch": 0.7933625924367221, + "flos": 19780483109160.0, + "grad_norm": 4.2881285270673075, + "learning_rate": 4.3128498994996685e-07, + "loss": 0.7272, + "num_input_tokens_seen": 140331020, + "step": 6598 + }, + { + "epoch": 0.7934828353273613, + "flos": 21695686782480.0, + "grad_norm": 12.232341884867397, + "learning_rate": 4.308019057943646e-07, + "loss": 0.6917, + "num_input_tokens_seen": 140352465, + "step": 6599 + }, + { + "epoch": 0.7936030782180004, + "flos": 20989564830120.0, + "grad_norm": 3.357832159567928, + "learning_rate": 4.3031905968486535e-07, + "loss": 0.7236, + "num_input_tokens_seen": 140373015, + "step": 6600 + }, + { + "epoch": 0.7937233211086394, + "flos": 11943855946200.0, + "grad_norm": 4.628492943705504, + "learning_rate": 4.298364516947162e-07, + "loss": 0.6639, + "num_input_tokens_seen": 140389965, + "step": 6601 + }, + { + "epoch": 0.7938435639992786, + "flos": 16103947879920.0, + "grad_norm": 3.8159956618470297, + "learning_rate": 4.293540818971295e-07, + "loss": 0.6449, + "num_input_tokens_seen": 140407490, + "step": 6602 + }, + { + "epoch": 0.7939638068899176, + "flos": 16239999931200.0, + "grad_norm": 4.50165018978067, + "learning_rate": 4.2887195036527934e-07, + "loss": 0.7476, + "num_input_tokens_seen": 140426015, + "step": 6603 + }, + { + "epoch": 0.7940840497805567, + "flos": 12945252600480.0, + "grad_norm": 4.615904654035488, + "learning_rate": 4.28390057172306e-07, + "loss": 0.6968, + "num_input_tokens_seen": 140442240, + "step": 6604 + }, + { + "epoch": 0.7942042926711959, + "flos": 17430755026680.0, + "grad_norm": 3.2590606274509346, + "learning_rate": 4.279084023913111e-07, + "loss": 0.7026, + "num_input_tokens_seen": 140459835, + "step": 6605 + }, + { + "epoch": 0.7943245355618349, + "flos": 14055731511480.0, + "grad_norm": 17.974606722162413, + "learning_rate": 4.2742698609536096e-07, + "loss": 0.6775, + "num_input_tokens_seen": 140477865, + "step": 6606 + }, + { + "epoch": 0.794444778452474, + "flos": 18320188367880.0, + "grad_norm": 8.065234762843867, + "learning_rate": 4.2694580835748706e-07, + "loss": 0.7665, + "num_input_tokens_seen": 140497445, + "step": 6607 + }, + { + "epoch": 0.7945650213431131, + "flos": 16998573454560.0, + "grad_norm": 6.55662055263907, + "learning_rate": 4.264648692506836e-07, + "loss": 0.7245, + "num_input_tokens_seen": 140515955, + "step": 6608 + }, + { + "epoch": 0.7946852642337522, + "flos": 19100761072080.0, + "grad_norm": 3.479206031683217, + "learning_rate": 4.2598416884790824e-07, + "loss": 0.6987, + "num_input_tokens_seen": 140534725, + "step": 6609 + }, + { + "epoch": 0.7948055071243912, + "flos": 17404418431320.0, + "grad_norm": 5.692207819326209, + "learning_rate": 4.255037072220828e-07, + "loss": 0.7897, + "num_input_tokens_seen": 140555815, + "step": 6610 + }, + { + "epoch": 0.7949257500150304, + "flos": 16080365701080.0, + "grad_norm": 6.783688998759065, + "learning_rate": 4.2502348444609293e-07, + "loss": 0.7035, + "num_input_tokens_seen": 140575155, + "step": 6611 + }, + { + "epoch": 0.7950459929056695, + "flos": 18887503852440.0, + "grad_norm": 5.179394559789922, + "learning_rate": 4.2454350059278844e-07, + "loss": 0.6681, + "num_input_tokens_seen": 140595935, + "step": 6612 + }, + { + "epoch": 0.7951662357963085, + "flos": 16210813939440.0, + "grad_norm": 8.869328365594448, + "learning_rate": 4.240637557349824e-07, + "loss": 0.8319, + "num_input_tokens_seen": 140612870, + "step": 6613 + }, + { + "epoch": 0.7952864786869477, + "flos": 18049445643600.0, + "grad_norm": 3.0542560087271178, + "learning_rate": 4.235842499454516e-07, + "loss": 0.6404, + "num_input_tokens_seen": 140632505, + "step": 6614 + }, + { + "epoch": 0.7954067215775867, + "flos": 15968877287400.0, + "grad_norm": 4.027015768879931, + "learning_rate": 4.2310498329693687e-07, + "loss": 0.8054, + "num_input_tokens_seen": 140653125, + "step": 6615 + }, + { + "epoch": 0.7955269644682258, + "flos": 17619606908520.0, + "grad_norm": 4.576989146904207, + "learning_rate": 4.2262595586214164e-07, + "loss": 0.7967, + "num_input_tokens_seen": 140673940, + "step": 6616 + }, + { + "epoch": 0.795647207358865, + "flos": 18322942784400.0, + "grad_norm": 6.130937773272237, + "learning_rate": 4.221471677137358e-07, + "loss": 0.7549, + "num_input_tokens_seen": 140694475, + "step": 6617 + }, + { + "epoch": 0.795767450249504, + "flos": 10654023141360.0, + "grad_norm": 18.937945571542418, + "learning_rate": 4.216686189243492e-07, + "loss": 0.7001, + "num_input_tokens_seen": 140712985, + "step": 6618 + }, + { + "epoch": 0.7958876931401431, + "flos": 13539316259880.0, + "grad_norm": 4.502812036761397, + "learning_rate": 4.211903095665785e-07, + "loss": 0.7148, + "num_input_tokens_seen": 140732090, + "step": 6619 + }, + { + "epoch": 0.7960079360307821, + "flos": 15757203065760.0, + "grad_norm": 6.493743820114059, + "learning_rate": 4.2071223971298277e-07, + "loss": 0.7377, + "num_input_tokens_seen": 140748995, + "step": 6620 + }, + { + "epoch": 0.7961281789214213, + "flos": 18671048976840.0, + "grad_norm": 25.57532699838301, + "learning_rate": 4.2023440943608433e-07, + "loss": 0.5909, + "num_input_tokens_seen": 140768680, + "step": 6621 + }, + { + "epoch": 0.7962484218120603, + "flos": 16053459226440.0, + "grad_norm": 7.741221963396494, + "learning_rate": 4.1975681880837023e-07, + "loss": 0.7694, + "num_input_tokens_seen": 140788405, + "step": 6622 + }, + { + "epoch": 0.7963686647026994, + "flos": 13782329350560.0, + "grad_norm": 11.307657628028664, + "learning_rate": 4.192794679022895e-07, + "loss": 0.8141, + "num_input_tokens_seen": 140806450, + "step": 6623 + }, + { + "epoch": 0.7964889075933386, + "flos": 21805465558320.0, + "grad_norm": 3.189890993222969, + "learning_rate": 4.1880235679025743e-07, + "loss": 0.7123, + "num_input_tokens_seen": 140826265, + "step": 6624 + }, + { + "epoch": 0.7966091504839776, + "flos": 21637979658120.0, + "grad_norm": 3.8400615579916213, + "learning_rate": 4.1832548554464986e-07, + "loss": 0.6111, + "num_input_tokens_seen": 140844280, + "step": 6625 + }, + { + "epoch": 0.7967293933746167, + "flos": 49615353547320.0, + "grad_norm": 0.7738015060044356, + "learning_rate": 4.178488542378098e-07, + "loss": 0.5977, + "num_input_tokens_seen": 140901580, + "step": 6626 + }, + { + "epoch": 0.7968496362652558, + "flos": 18725020225920.0, + "grad_norm": 4.94891910503976, + "learning_rate": 4.173724629420401e-07, + "loss": 0.8795, + "num_input_tokens_seen": 140922660, + "step": 6627 + }, + { + "epoch": 0.7969698791558949, + "flos": 10544845904760.0, + "grad_norm": 9.425700307878737, + "learning_rate": 4.168963117296087e-07, + "loss": 0.6724, + "num_input_tokens_seen": 140939715, + "step": 6628 + }, + { + "epoch": 0.797090122046534, + "flos": 16189068038280.0, + "grad_norm": 4.111874592542302, + "learning_rate": 4.1642040067274876e-07, + "loss": 0.7308, + "num_input_tokens_seen": 140959105, + "step": 6629 + }, + { + "epoch": 0.7972103649371731, + "flos": 14538211777320.0, + "grad_norm": 3.4151092949687616, + "learning_rate": 4.1594472984365493e-07, + "loss": 0.7141, + "num_input_tokens_seen": 140977510, + "step": 6630 + }, + { + "epoch": 0.7973306078278122, + "flos": 26498083171920.0, + "grad_norm": 4.44221525448593, + "learning_rate": 4.154692993144862e-07, + "loss": 0.762, + "num_input_tokens_seen": 140997000, + "step": 6631 + }, + { + "epoch": 0.7974508507184512, + "flos": 15621372634200.0, + "grad_norm": 3.7761569441814578, + "learning_rate": 4.1499410915736476e-07, + "loss": 0.6919, + "num_input_tokens_seen": 141015650, + "step": 6632 + }, + { + "epoch": 0.7975710936090904, + "flos": 50331321747240.0, + "grad_norm": 0.8161744514191703, + "learning_rate": 4.145191594443762e-07, + "loss": 0.6998, + "num_input_tokens_seen": 141079725, + "step": 6633 + }, + { + "epoch": 0.7976913364997295, + "flos": 16459114243440.0, + "grad_norm": 4.818873359100632, + "learning_rate": 4.140444502475713e-07, + "loss": 0.6945, + "num_input_tokens_seen": 141098995, + "step": 6634 + }, + { + "epoch": 0.7978115793903685, + "flos": 11108298874200.0, + "grad_norm": 10.349144764823867, + "learning_rate": 4.1356998163896216e-07, + "loss": 0.6871, + "num_input_tokens_seen": 141115765, + "step": 6635 + }, + { + "epoch": 0.7979318222810077, + "flos": 14406718760280.0, + "grad_norm": 5.158173358431246, + "learning_rate": 4.130957536905255e-07, + "loss": 0.7233, + "num_input_tokens_seen": 141133500, + "step": 6636 + }, + { + "epoch": 0.7980520651716467, + "flos": 11328869544600.0, + "grad_norm": 4.503617210410458, + "learning_rate": 4.1262176647420134e-07, + "loss": 0.6865, + "num_input_tokens_seen": 141151385, + "step": 6637 + }, + { + "epoch": 0.7981723080622858, + "flos": 16323600411480.0, + "grad_norm": 6.672110274229182, + "learning_rate": 4.121480200618923e-07, + "loss": 0.7821, + "num_input_tokens_seen": 141170760, + "step": 6638 + }, + { + "epoch": 0.798292550952925, + "flos": 16782530158440.0, + "grad_norm": 3.3557282281985112, + "learning_rate": 4.116745145254674e-07, + "loss": 0.7887, + "num_input_tokens_seen": 141190015, + "step": 6639 + }, + { + "epoch": 0.798412793843564, + "flos": 47552641408560.0, + "grad_norm": 0.7899227551830852, + "learning_rate": 4.1120124993675476e-07, + "loss": 0.5969, + "num_input_tokens_seen": 141254165, + "step": 6640 + }, + { + "epoch": 0.7985330367342031, + "flos": 9868004924040.0, + "grad_norm": 3.9225616461904154, + "learning_rate": 4.107282263675498e-07, + "loss": 0.6022, + "num_input_tokens_seen": 141271555, + "step": 6641 + }, + { + "epoch": 0.7986532796248422, + "flos": 49920094577280.0, + "grad_norm": 0.7489090773753727, + "learning_rate": 4.1025544388960907e-07, + "loss": 0.5332, + "num_input_tokens_seen": 141332315, + "step": 6642 + }, + { + "epoch": 0.7987735225154813, + "flos": 16431606229560.0, + "grad_norm": 4.720158246830509, + "learning_rate": 4.097829025746538e-07, + "loss": 0.6926, + "num_input_tokens_seen": 141353580, + "step": 6643 + }, + { + "epoch": 0.7988937654061203, + "flos": 50782178204400.0, + "grad_norm": 0.7286374396785288, + "learning_rate": 4.0931060249436757e-07, + "loss": 0.6165, + "num_input_tokens_seen": 141417140, + "step": 6644 + }, + { + "epoch": 0.7990140082967595, + "flos": 15210683683560.0, + "grad_norm": 8.282983217061416, + "learning_rate": 4.088385437203978e-07, + "loss": 0.6838, + "num_input_tokens_seen": 141433870, + "step": 6645 + }, + { + "epoch": 0.7991342511873986, + "flos": 13862953794600.0, + "grad_norm": 4.764605703942151, + "learning_rate": 4.083667263243564e-07, + "loss": 0.7522, + "num_input_tokens_seen": 141451935, + "step": 6646 + }, + { + "epoch": 0.7992544940780376, + "flos": 15219421832520.0, + "grad_norm": 3.6731557357512123, + "learning_rate": 4.0789515037781653e-07, + "loss": 0.704, + "num_input_tokens_seen": 141472380, + "step": 6647 + }, + { + "epoch": 0.7993747369686768, + "flos": 9168151643760.0, + "grad_norm": 3.77393282528009, + "learning_rate": 4.0742381595231755e-07, + "loss": 0.8091, + "num_input_tokens_seen": 141488825, + "step": 6648 + }, + { + "epoch": 0.7994949798593158, + "flos": 14672047631400.0, + "grad_norm": 9.396382678791294, + "learning_rate": 4.06952723119359e-07, + "loss": 0.7741, + "num_input_tokens_seen": 141508420, + "step": 6649 + }, + { + "epoch": 0.7996152227499549, + "flos": 28419713817120.0, + "grad_norm": 4.4841691345385755, + "learning_rate": 4.0648187195040504e-07, + "loss": 0.6519, + "num_input_tokens_seen": 141530345, + "step": 6650 + }, + { + "epoch": 0.799735465640594, + "flos": 51802313063640.0, + "grad_norm": 0.9465538575617581, + "learning_rate": 4.060112625168848e-07, + "loss": 0.7176, + "num_input_tokens_seen": 141595175, + "step": 6651 + }, + { + "epoch": 0.7998557085312331, + "flos": 17752777903440.0, + "grad_norm": 5.047647800860635, + "learning_rate": 4.055408948901886e-07, + "loss": 0.7214, + "num_input_tokens_seen": 141616295, + "step": 6652 + }, + { + "epoch": 0.7999759514218722, + "flos": 20212443061560.0, + "grad_norm": 2.9641398723922374, + "learning_rate": 4.050707691416708e-07, + "loss": 0.6994, + "num_input_tokens_seen": 141637325, + "step": 6653 + }, + { + "epoch": 0.8000961943125112, + "flos": 49653151048200.0, + "grad_norm": 0.7222878831783198, + "learning_rate": 4.046008853426495e-07, + "loss": 0.6144, + "num_input_tokens_seen": 141700360, + "step": 6654 + }, + { + "epoch": 0.8002164372031504, + "flos": 21078484183680.0, + "grad_norm": 5.895773987308522, + "learning_rate": 4.0413124356440464e-07, + "loss": 0.613, + "num_input_tokens_seen": 141724125, + "step": 6655 + }, + { + "epoch": 0.8003366800937894, + "flos": 12867509212800.0, + "grad_norm": 11.639749337869674, + "learning_rate": 4.0366184387818223e-07, + "loss": 0.8052, + "num_input_tokens_seen": 141742305, + "step": 6656 + }, + { + "epoch": 0.8004569229844285, + "flos": 19051507157040.0, + "grad_norm": 4.258746403456665, + "learning_rate": 4.0319268635518797e-07, + "loss": 0.8331, + "num_input_tokens_seen": 141762600, + "step": 6657 + }, + { + "epoch": 0.8005771658750677, + "flos": 15215495997480.0, + "grad_norm": 4.482962675028995, + "learning_rate": 4.027237710665943e-07, + "loss": 0.7453, + "num_input_tokens_seen": 141780785, + "step": 6658 + }, + { + "epoch": 0.8006974087657067, + "flos": 18916278264720.0, + "grad_norm": 5.298306702079614, + "learning_rate": 4.022550980835344e-07, + "loss": 0.6833, + "num_input_tokens_seen": 141802750, + "step": 6659 + }, + { + "epoch": 0.8008176516563458, + "flos": 12516110384520.0, + "grad_norm": 4.339607074371656, + "learning_rate": 4.017866674771051e-07, + "loss": 0.7942, + "num_input_tokens_seen": 141819955, + "step": 6660 + }, + { + "epoch": 0.8009378945469849, + "flos": 17728245925800.0, + "grad_norm": 4.626599215767962, + "learning_rate": 4.013184793183688e-07, + "loss": 0.7269, + "num_input_tokens_seen": 141841770, + "step": 6661 + }, + { + "epoch": 0.801058137437624, + "flos": 14456415914760.0, + "grad_norm": 3.0450260127417508, + "learning_rate": 4.008505336783472e-07, + "loss": 0.7168, + "num_input_tokens_seen": 141859215, + "step": 6662 + }, + { + "epoch": 0.801178380328263, + "flos": 13622125241160.0, + "grad_norm": 9.591043792659631, + "learning_rate": 4.003828306280284e-07, + "loss": 0.8045, + "num_input_tokens_seen": 141876610, + "step": 6663 + }, + { + "epoch": 0.8012986232189022, + "flos": 11437191962280.0, + "grad_norm": 3.6994120981348066, + "learning_rate": 3.999153702383626e-07, + "loss": 0.7665, + "num_input_tokens_seen": 141894220, + "step": 6664 + }, + { + "epoch": 0.8014188661095413, + "flos": 20967185729760.0, + "grad_norm": 4.271707857736831, + "learning_rate": 3.9944815258026263e-07, + "loss": 0.7148, + "num_input_tokens_seen": 141915760, + "step": 6665 + }, + { + "epoch": 0.8015391090001803, + "flos": 21504143804040.0, + "grad_norm": 16.694464326840688, + "learning_rate": 3.989811777246057e-07, + "loss": 0.8198, + "num_input_tokens_seen": 141935650, + "step": 6666 + }, + { + "epoch": 0.8016593518908195, + "flos": 51916999133280.0, + "grad_norm": 0.8974731909379249, + "learning_rate": 3.985144457422305e-07, + "loss": 0.6923, + "num_input_tokens_seen": 141989655, + "step": 6667 + }, + { + "epoch": 0.8017795947814585, + "flos": 19073506337880.0, + "grad_norm": 4.818268724446216, + "learning_rate": 3.9804795670394096e-07, + "loss": 0.7406, + "num_input_tokens_seen": 142009500, + "step": 6668 + }, + { + "epoch": 0.8018998376720976, + "flos": 16158647308080.0, + "grad_norm": 4.546138743144981, + "learning_rate": 3.975817106805022e-07, + "loss": 0.7, + "num_input_tokens_seen": 142027920, + "step": 6669 + }, + { + "epoch": 0.8020200805627368, + "flos": 25396152450120.0, + "grad_norm": 5.524859859018326, + "learning_rate": 3.97115707742645e-07, + "loss": 0.6197, + "num_input_tokens_seen": 142048315, + "step": 6670 + }, + { + "epoch": 0.8021403234533758, + "flos": 14703544800240.0, + "grad_norm": 11.897915543787885, + "learning_rate": 3.966499479610599e-07, + "loss": 0.6394, + "num_input_tokens_seen": 142066130, + "step": 6671 + }, + { + "epoch": 0.8022605663440149, + "flos": 20347228714440.0, + "grad_norm": 5.149376559252063, + "learning_rate": 3.9618443140640225e-07, + "loss": 0.6393, + "num_input_tokens_seen": 142084760, + "step": 6672 + }, + { + "epoch": 0.802380809234654, + "flos": 44402268207240.0, + "grad_norm": 0.7422804924781129, + "learning_rate": 3.957191581492918e-07, + "loss": 0.5383, + "num_input_tokens_seen": 142145240, + "step": 6673 + }, + { + "epoch": 0.8025010521252931, + "flos": 10973228281680.0, + "grad_norm": 4.957441789417096, + "learning_rate": 3.952541282603097e-07, + "loss": 0.6841, + "num_input_tokens_seen": 142160065, + "step": 6674 + }, + { + "epoch": 0.8026212950159322, + "flos": 16215626253360.0, + "grad_norm": 4.166963922705581, + "learning_rate": 3.9478934181000013e-07, + "loss": 0.8253, + "num_input_tokens_seen": 142179810, + "step": 6675 + }, + { + "epoch": 0.8027415379065713, + "flos": 12894573987240.0, + "grad_norm": 5.595832650115236, + "learning_rate": 3.943247988688714e-07, + "loss": 0.8279, + "num_input_tokens_seen": 142198225, + "step": 6676 + }, + { + "epoch": 0.8028617807972104, + "flos": 16079289262440.0, + "grad_norm": 11.755192684669654, + "learning_rate": 3.938604995073933e-07, + "loss": 0.7131, + "num_input_tokens_seen": 142216415, + "step": 6677 + }, + { + "epoch": 0.8029820236878494, + "flos": 19371408816480.0, + "grad_norm": 3.090239411495706, + "learning_rate": 3.9339644379600157e-07, + "loss": 0.6471, + "num_input_tokens_seen": 142235965, + "step": 6678 + }, + { + "epoch": 0.8031022665784886, + "flos": 12895017226680.0, + "grad_norm": 5.689214091807574, + "learning_rate": 3.929326318050907e-07, + "loss": 0.7006, + "num_input_tokens_seen": 142253355, + "step": 6679 + }, + { + "epoch": 0.8032225094691277, + "flos": 11247390281640.0, + "grad_norm": 3.4176057284896504, + "learning_rate": 3.924690636050225e-07, + "loss": 0.7733, + "num_input_tokens_seen": 142270485, + "step": 6680 + }, + { + "epoch": 0.8033427523597667, + "flos": 19187812488000.0, + "grad_norm": 3.7551479608106746, + "learning_rate": 3.9200573926611915e-07, + "loss": 0.7075, + "num_input_tokens_seen": 142291620, + "step": 6681 + }, + { + "epoch": 0.8034629952504058, + "flos": 15595067698800.0, + "grad_norm": 3.306861252699377, + "learning_rate": 3.9154265885866613e-07, + "loss": 0.712, + "num_input_tokens_seen": 142310650, + "step": 6682 + }, + { + "epoch": 0.8035832381410449, + "flos": 15836592771360.0, + "grad_norm": 6.530271027856778, + "learning_rate": 3.9107982245291394e-07, + "loss": 0.7262, + "num_input_tokens_seen": 142328495, + "step": 6683 + }, + { + "epoch": 0.803703481031684, + "flos": 14998028003160.0, + "grad_norm": 8.210976525240277, + "learning_rate": 3.9061723011907245e-07, + "loss": 0.7468, + "num_input_tokens_seen": 142347570, + "step": 6684 + }, + { + "epoch": 0.803823723922323, + "flos": 16726944251400.0, + "grad_norm": 4.465038972793457, + "learning_rate": 3.901548819273179e-07, + "loss": 0.7639, + "num_input_tokens_seen": 142367305, + "step": 6685 + }, + { + "epoch": 0.8039439668129622, + "flos": 15622734012480.0, + "grad_norm": 3.887345858642762, + "learning_rate": 3.896927779477881e-07, + "loss": 0.6769, + "num_input_tokens_seen": 142386285, + "step": 6686 + }, + { + "epoch": 0.8040642097036013, + "flos": 17428063930080.0, + "grad_norm": 7.587814460283765, + "learning_rate": 3.892309182505833e-07, + "loss": 0.6673, + "num_input_tokens_seen": 142403820, + "step": 6687 + }, + { + "epoch": 0.8041844525942403, + "flos": 18997250968320.0, + "grad_norm": 5.018476679712127, + "learning_rate": 3.887693029057675e-07, + "loss": 0.8438, + "num_input_tokens_seen": 142423050, + "step": 6688 + }, + { + "epoch": 0.8043046954848795, + "flos": 18456145439280.0, + "grad_norm": 2.6172009823579967, + "learning_rate": 3.8830793198336684e-07, + "loss": 0.8041, + "num_input_tokens_seen": 142442360, + "step": 6689 + }, + { + "epoch": 0.8044249383755185, + "flos": 30688025956560.0, + "grad_norm": 21.032091597202715, + "learning_rate": 3.878468055533721e-07, + "loss": 0.6919, + "num_input_tokens_seen": 142464620, + "step": 6690 + }, + { + "epoch": 0.8045451812661576, + "flos": 15081818443200.0, + "grad_norm": 8.91164601475972, + "learning_rate": 3.8738592368573464e-07, + "loss": 0.8253, + "num_input_tokens_seen": 142481895, + "step": 6691 + }, + { + "epoch": 0.8046654241567968, + "flos": 21801254783640.0, + "grad_norm": 4.815896557200768, + "learning_rate": 3.8692528645037137e-07, + "loss": 0.8652, + "num_input_tokens_seen": 142500795, + "step": 6692 + }, + { + "epoch": 0.8047856670474358, + "flos": 12890521512360.0, + "grad_norm": 4.953960552799286, + "learning_rate": 3.8646489391715907e-07, + "loss": 0.7765, + "num_input_tokens_seen": 142514810, + "step": 6693 + }, + { + "epoch": 0.8049059099380749, + "flos": 12482491998360.0, + "grad_norm": 9.388381156761207, + "learning_rate": 3.8600474615593903e-07, + "loss": 0.8669, + "num_input_tokens_seen": 142529145, + "step": 6694 + }, + { + "epoch": 0.805026152828714, + "flos": 45858320513880.0, + "grad_norm": 0.8328297325197919, + "learning_rate": 3.8554484323651605e-07, + "loss": 0.6495, + "num_input_tokens_seen": 142590735, + "step": 6695 + }, + { + "epoch": 0.8051463957193531, + "flos": 15864069125280.0, + "grad_norm": 3.57287932766606, + "learning_rate": 3.85085185228657e-07, + "loss": 0.776, + "num_input_tokens_seen": 142609425, + "step": 6696 + }, + { + "epoch": 0.8052666386099921, + "flos": 23533400347800.0, + "grad_norm": 4.139527373310829, + "learning_rate": 3.8462577220209114e-07, + "loss": 0.7287, + "num_input_tokens_seen": 142629520, + "step": 6697 + }, + { + "epoch": 0.8053868815006313, + "flos": 49520676572400.0, + "grad_norm": 0.6779195219314678, + "learning_rate": 3.8416660422651127e-07, + "loss": 0.6001, + "num_input_tokens_seen": 142698890, + "step": 6698 + }, + { + "epoch": 0.8055071243912704, + "flos": 17453419066680.0, + "grad_norm": 6.570788571486064, + "learning_rate": 3.837076813715723e-07, + "loss": 0.672, + "num_input_tokens_seen": 142718495, + "step": 6699 + }, + { + "epoch": 0.8056273672819094, + "flos": 15594719439240.0, + "grad_norm": 3.7826439676160626, + "learning_rate": 3.832490037068941e-07, + "loss": 0.739, + "num_input_tokens_seen": 142737005, + "step": 6700 + }, + { + "epoch": 0.8057476101725486, + "flos": 18882438258840.0, + "grad_norm": 5.9326386477235316, + "learning_rate": 3.827905713020554e-07, + "loss": 0.7433, + "num_input_tokens_seen": 142754370, + "step": 6701 + }, + { + "epoch": 0.8058678530631876, + "flos": 17970910756920.0, + "grad_norm": 17.539182558052342, + "learning_rate": 3.823323842266017e-07, + "loss": 0.6668, + "num_input_tokens_seen": 142773485, + "step": 6702 + }, + { + "epoch": 0.8059880959538267, + "flos": 18132507904560.0, + "grad_norm": 4.987342835820666, + "learning_rate": 3.818744425500393e-07, + "loss": 0.7101, + "num_input_tokens_seen": 142791220, + "step": 6703 + }, + { + "epoch": 0.8061083388444659, + "flos": 16239398391960.0, + "grad_norm": 3.0963156494277375, + "learning_rate": 3.8141674634183675e-07, + "loss": 0.7971, + "num_input_tokens_seen": 142809970, + "step": 6704 + }, + { + "epoch": 0.8062285817351049, + "flos": 22047813789840.0, + "grad_norm": 3.127875569066251, + "learning_rate": 3.809592956714278e-07, + "loss": 0.6316, + "num_input_tokens_seen": 142832925, + "step": 6705 + }, + { + "epoch": 0.806348824625744, + "flos": 16673606201520.0, + "grad_norm": 3.277684045159041, + "learning_rate": 3.805020906082057e-07, + "loss": 0.7262, + "num_input_tokens_seen": 142851220, + "step": 6706 + }, + { + "epoch": 0.8064690675163831, + "flos": 17134182266400.0, + "grad_norm": 5.290219967745094, + "learning_rate": 3.8004513122152917e-07, + "loss": 0.8033, + "num_input_tokens_seen": 142869250, + "step": 6707 + }, + { + "epoch": 0.8065893104070222, + "flos": 17619828528240.0, + "grad_norm": 4.497950405025531, + "learning_rate": 3.79588417580718e-07, + "loss": 0.662, + "num_input_tokens_seen": 142887080, + "step": 6708 + }, + { + "epoch": 0.8067095532976613, + "flos": 16320054495960.0, + "grad_norm": 3.5946926041760605, + "learning_rate": 3.791319497550558e-07, + "loss": 0.7385, + "num_input_tokens_seen": 142904630, + "step": 6709 + }, + { + "epoch": 0.8068297961883004, + "flos": 12490248688560.0, + "grad_norm": 4.950848269750134, + "learning_rate": 3.78675727813788e-07, + "loss": 0.703, + "num_input_tokens_seen": 142921915, + "step": 6710 + }, + { + "epoch": 0.8069500390789395, + "flos": 16108728533880.0, + "grad_norm": 11.26128008435025, + "learning_rate": 3.782197518261225e-07, + "loss": 0.7133, + "num_input_tokens_seen": 142941075, + "step": 6711 + }, + { + "epoch": 0.8070702819695785, + "flos": 14055098312280.0, + "grad_norm": 5.285064944587329, + "learning_rate": 3.777640218612319e-07, + "loss": 0.9555, + "num_input_tokens_seen": 142958780, + "step": 6712 + }, + { + "epoch": 0.8071905248602176, + "flos": 15757519665360.0, + "grad_norm": 4.460433797908825, + "learning_rate": 3.773085379882488e-07, + "loss": 0.6991, + "num_input_tokens_seen": 142977555, + "step": 6713 + }, + { + "epoch": 0.8073107677508568, + "flos": 27394133444760.0, + "grad_norm": 3.4203843305703288, + "learning_rate": 3.768533002762715e-07, + "loss": 0.7475, + "num_input_tokens_seen": 143000810, + "step": 6714 + }, + { + "epoch": 0.8074310106414958, + "flos": 20806095141480.0, + "grad_norm": 4.26500320094571, + "learning_rate": 3.763983087943572e-07, + "loss": 0.7486, + "num_input_tokens_seen": 143019920, + "step": 6715 + }, + { + "epoch": 0.8075512535321349, + "flos": 17783451913320.0, + "grad_norm": 3.2016364383677116, + "learning_rate": 3.759435636115282e-07, + "loss": 0.7854, + "num_input_tokens_seen": 143040425, + "step": 6716 + }, + { + "epoch": 0.807671496422774, + "flos": 19077685452600.0, + "grad_norm": 4.167007635788391, + "learning_rate": 3.7548906479676967e-07, + "loss": 0.7227, + "num_input_tokens_seen": 143059740, + "step": 6717 + }, + { + "epoch": 0.8077917393134131, + "flos": 17375232439560.0, + "grad_norm": 3.220796249459097, + "learning_rate": 3.7503481241902855e-07, + "loss": 0.6943, + "num_input_tokens_seen": 143079435, + "step": 6718 + }, + { + "epoch": 0.8079119822040521, + "flos": 13431373761720.0, + "grad_norm": 3.8325665301317002, + "learning_rate": 3.745808065472145e-07, + "loss": 0.7832, + "num_input_tokens_seen": 143096450, + "step": 6719 + }, + { + "epoch": 0.8080322250946913, + "flos": 17294671315440.0, + "grad_norm": 5.130019916101653, + "learning_rate": 3.741270472501994e-07, + "loss": 0.7542, + "num_input_tokens_seen": 143116810, + "step": 6720 + }, + { + "epoch": 0.8081524679853304, + "flos": 16702032354240.0, + "grad_norm": 8.947279907828046, + "learning_rate": 3.736735345968183e-07, + "loss": 0.7189, + "num_input_tokens_seen": 143136140, + "step": 6721 + }, + { + "epoch": 0.8082727108759694, + "flos": 12868015772160.0, + "grad_norm": 2.7850387141825963, + "learning_rate": 3.7322026865586986e-07, + "loss": 0.7754, + "num_input_tokens_seen": 143154895, + "step": 6722 + }, + { + "epoch": 0.8083929537666086, + "flos": 19023935823240.0, + "grad_norm": 4.406126196454435, + "learning_rate": 3.7276724949611206e-07, + "loss": 0.7129, + "num_input_tokens_seen": 143174725, + "step": 6723 + }, + { + "epoch": 0.8085131966572476, + "flos": 19860949253400.0, + "grad_norm": 4.47557261975026, + "learning_rate": 3.723144771862694e-07, + "loss": 0.7269, + "num_input_tokens_seen": 143195085, + "step": 6724 + }, + { + "epoch": 0.8086334395478867, + "flos": 17566363838520.0, + "grad_norm": 4.230191912975754, + "learning_rate": 3.718619517950263e-07, + "loss": 0.7459, + "num_input_tokens_seen": 143215400, + "step": 6725 + }, + { + "epoch": 0.8087536824385259, + "flos": 14913952623480.0, + "grad_norm": 4.173207273884879, + "learning_rate": 3.714096733910301e-07, + "loss": 0.7501, + "num_input_tokens_seen": 143232645, + "step": 6726 + }, + { + "epoch": 0.8088739253291649, + "flos": 18995129751000.0, + "grad_norm": 7.965892084791443, + "learning_rate": 3.709576420428926e-07, + "loss": 0.6815, + "num_input_tokens_seen": 143253165, + "step": 6727 + }, + { + "epoch": 0.808994168219804, + "flos": 20832495056760.0, + "grad_norm": 6.951068428300924, + "learning_rate": 3.7050585781918463e-07, + "loss": 0.7175, + "num_input_tokens_seen": 143273185, + "step": 6728 + }, + { + "epoch": 0.8091144111104431, + "flos": 12705975385080.0, + "grad_norm": 5.014618674601306, + "learning_rate": 3.700543207884428e-07, + "loss": 0.6611, + "num_input_tokens_seen": 143289815, + "step": 6729 + }, + { + "epoch": 0.8092346540010822, + "flos": 23608136039280.0, + "grad_norm": 4.880389242439054, + "learning_rate": 3.6960303101916466e-07, + "loss": 0.6973, + "num_input_tokens_seen": 143309450, + "step": 6730 + }, + { + "epoch": 0.8093548968917212, + "flos": 42769014884040.0, + "grad_norm": 0.7868364869162074, + "learning_rate": 3.6915198857981047e-07, + "loss": 0.5767, + "num_input_tokens_seen": 143374370, + "step": 6731 + }, + { + "epoch": 0.8094751397823604, + "flos": 20077435788960.0, + "grad_norm": 8.460242245499774, + "learning_rate": 3.687011935388027e-07, + "loss": 0.666, + "num_input_tokens_seen": 143396985, + "step": 6732 + }, + { + "epoch": 0.8095953826729995, + "flos": 17619986828040.0, + "grad_norm": 7.227271660727787, + "learning_rate": 3.6825064596452646e-07, + "loss": 0.7087, + "num_input_tokens_seen": 143417050, + "step": 6733 + }, + { + "epoch": 0.8097156255636385, + "flos": 17538792504720.0, + "grad_norm": 7.317874856775948, + "learning_rate": 3.678003459253305e-07, + "loss": 0.687, + "num_input_tokens_seen": 143437620, + "step": 6734 + }, + { + "epoch": 0.8098358684542777, + "flos": 15946023287640.0, + "grad_norm": 3.1279952381517306, + "learning_rate": 3.673502934895236e-07, + "loss": 0.7242, + "num_input_tokens_seen": 143456845, + "step": 6735 + }, + { + "epoch": 0.8099561113449167, + "flos": 50744385194880.0, + "grad_norm": 0.6747349925285608, + "learning_rate": 3.669004887253802e-07, + "loss": 0.5822, + "num_input_tokens_seen": 143522855, + "step": 6736 + }, + { + "epoch": 0.8100763542355558, + "flos": 17268334720080.0, + "grad_norm": 8.472610325774571, + "learning_rate": 3.664509317011335e-07, + "loss": 0.7774, + "num_input_tokens_seen": 143542910, + "step": 6737 + }, + { + "epoch": 0.810196597126195, + "flos": 23236637627760.0, + "grad_norm": 4.465570132089995, + "learning_rate": 3.6600162248498134e-07, + "loss": 0.7132, + "num_input_tokens_seen": 143566260, + "step": 6738 + }, + { + "epoch": 0.810316840016834, + "flos": 18241210241760.0, + "grad_norm": 3.562988783115029, + "learning_rate": 3.6555256114508426e-07, + "loss": 0.7492, + "num_input_tokens_seen": 143585775, + "step": 6739 + }, + { + "epoch": 0.8104370829074731, + "flos": 20509142461680.0, + "grad_norm": 4.081213238537419, + "learning_rate": 3.651037477495642e-07, + "loss": 0.7142, + "num_input_tokens_seen": 143606945, + "step": 6740 + }, + { + "epoch": 0.8105573257981122, + "flos": 18048400864920.0, + "grad_norm": 38.06781219344386, + "learning_rate": 3.6465518236650584e-07, + "loss": 0.6543, + "num_input_tokens_seen": 143626810, + "step": 6741 + }, + { + "epoch": 0.8106775686887513, + "flos": 19320445263600.0, + "grad_norm": 3.559339359362895, + "learning_rate": 3.642068650639558e-07, + "loss": 0.7743, + "num_input_tokens_seen": 143646275, + "step": 6742 + }, + { + "epoch": 0.8107978115793903, + "flos": 19996209805680.0, + "grad_norm": 4.199943160626133, + "learning_rate": 3.6375879590992334e-07, + "loss": 0.6355, + "num_input_tokens_seen": 143666340, + "step": 6743 + }, + { + "epoch": 0.8109180544700295, + "flos": 18265900519200.0, + "grad_norm": 5.990678221655702, + "learning_rate": 3.6331097497238173e-07, + "loss": 0.7925, + "num_input_tokens_seen": 143685505, + "step": 6744 + }, + { + "epoch": 0.8110382973606686, + "flos": 15432520752360.0, + "grad_norm": 4.151053715412773, + "learning_rate": 3.628634023192627e-07, + "loss": 0.7792, + "num_input_tokens_seen": 143705470, + "step": 6745 + }, + { + "epoch": 0.8111585402513076, + "flos": 11220642106800.0, + "grad_norm": 6.606057202314699, + "learning_rate": 3.624160780184644e-07, + "loss": 0.7366, + "num_input_tokens_seen": 143722405, + "step": 6746 + }, + { + "epoch": 0.8112787831419467, + "flos": 17645373624600.0, + "grad_norm": 3.5385348366637857, + "learning_rate": 3.6196900213784496e-07, + "loss": 0.7359, + "num_input_tokens_seen": 143741440, + "step": 6747 + }, + { + "epoch": 0.8113990260325858, + "flos": 14971944687480.0, + "grad_norm": 7.877072742706931, + "learning_rate": 3.6152217474522527e-07, + "loss": 0.8523, + "num_input_tokens_seen": 143757975, + "step": 6748 + }, + { + "epoch": 0.8115192689232249, + "flos": 18241811781000.0, + "grad_norm": 3.9055697649906325, + "learning_rate": 3.6107559590838975e-07, + "loss": 0.7175, + "num_input_tokens_seen": 143776680, + "step": 6749 + }, + { + "epoch": 0.811639511813864, + "flos": 17616567552360.0, + "grad_norm": 6.732243671941654, + "learning_rate": 3.606292656950822e-07, + "loss": 0.6234, + "num_input_tokens_seen": 143794810, + "step": 6750 + }, + { + "epoch": 0.8117597547045031, + "flos": 16971698639880.0, + "grad_norm": 5.1942761757172695, + "learning_rate": 3.601831841730121e-07, + "loss": 0.8424, + "num_input_tokens_seen": 143812450, + "step": 6751 + }, + { + "epoch": 0.8118799975951422, + "flos": 16864610960640.0, + "grad_norm": 2.8467021332734297, + "learning_rate": 3.5973735140984916e-07, + "loss": 0.7247, + "num_input_tokens_seen": 143832340, + "step": 6752 + }, + { + "epoch": 0.8120002404857812, + "flos": 18047704345800.0, + "grad_norm": 5.678278899219746, + "learning_rate": 3.5929176747322607e-07, + "loss": 0.7574, + "num_input_tokens_seen": 143851165, + "step": 6753 + }, + { + "epoch": 0.8121204833764204, + "flos": 42310781656200.0, + "grad_norm": 0.9493668680049946, + "learning_rate": 3.588464324307372e-07, + "loss": 0.5659, + "num_input_tokens_seen": 143914510, + "step": 6754 + }, + { + "epoch": 0.8122407262670595, + "flos": 14218373437800.0, + "grad_norm": 2.8415911518105226, + "learning_rate": 3.584013463499391e-07, + "loss": 0.74, + "num_input_tokens_seen": 143932850, + "step": 6755 + }, + { + "epoch": 0.8123609691576985, + "flos": 41577056710080.0, + "grad_norm": 0.7237075993532003, + "learning_rate": 3.579565092983521e-07, + "loss": 0.6446, + "num_input_tokens_seen": 143993690, + "step": 6756 + }, + { + "epoch": 0.8124812120483377, + "flos": 15082293342600.0, + "grad_norm": 5.164146157191614, + "learning_rate": 3.575119213434565e-07, + "loss": 0.8294, + "num_input_tokens_seen": 144011925, + "step": 6757 + }, + { + "epoch": 0.8126014549389767, + "flos": 16459082583480.0, + "grad_norm": 3.7212141832416195, + "learning_rate": 3.5706758255269765e-07, + "loss": 0.8, + "num_input_tokens_seen": 144030100, + "step": 6758 + }, + { + "epoch": 0.8127216978296158, + "flos": 17046845910840.0, + "grad_norm": 2.72180249963988, + "learning_rate": 3.566234929934795e-07, + "loss": 0.687, + "num_input_tokens_seen": 144049020, + "step": 6759 + }, + { + "epoch": 0.812841940720255, + "flos": 18429460584360.0, + "grad_norm": 2.9570960739468304, + "learning_rate": 3.561796527331706e-07, + "loss": 0.71, + "num_input_tokens_seen": 144070415, + "step": 6760 + }, + { + "epoch": 0.812962183610894, + "flos": 19534367342400.0, + "grad_norm": 4.320240537516627, + "learning_rate": 3.5573606183910163e-07, + "loss": 0.7538, + "num_input_tokens_seen": 144090140, + "step": 6761 + }, + { + "epoch": 0.8130824265015331, + "flos": 18289514358000.0, + "grad_norm": 2.821570203168348, + "learning_rate": 3.5529272037856493e-07, + "loss": 0.7628, + "num_input_tokens_seen": 144108075, + "step": 6762 + }, + { + "epoch": 0.8132026693921722, + "flos": 49863527211480.0, + "grad_norm": 0.7544264110754878, + "learning_rate": 3.548496284188149e-07, + "loss": 0.5651, + "num_input_tokens_seen": 144168000, + "step": 6763 + }, + { + "epoch": 0.8133229122828113, + "flos": 14240150998920.0, + "grad_norm": 7.678328835339971, + "learning_rate": 3.544067860270681e-07, + "loss": 0.7773, + "num_input_tokens_seen": 144185295, + "step": 6764 + }, + { + "epoch": 0.8134431551734503, + "flos": 15109009857480.0, + "grad_norm": 4.042738755912825, + "learning_rate": 3.539641932705029e-07, + "loss": 0.6859, + "num_input_tokens_seen": 144203495, + "step": 6765 + }, + { + "epoch": 0.8135633980640895, + "flos": 15730074971400.0, + "grad_norm": 3.8281075613442086, + "learning_rate": 3.53521850216262e-07, + "loss": 0.7265, + "num_input_tokens_seen": 144222785, + "step": 6766 + }, + { + "epoch": 0.8136836409547286, + "flos": 15024269618640.0, + "grad_norm": 6.339473391542872, + "learning_rate": 3.530797569314461e-07, + "loss": 0.7561, + "num_input_tokens_seen": 144241530, + "step": 6767 + }, + { + "epoch": 0.8138038838453676, + "flos": 14835259437000.0, + "grad_norm": 5.033443944023025, + "learning_rate": 3.5263791348312235e-07, + "loss": 0.7535, + "num_input_tokens_seen": 144260445, + "step": 6768 + }, + { + "epoch": 0.8139241267360068, + "flos": 21859721747040.0, + "grad_norm": 3.785110646574486, + "learning_rate": 3.521963199383171e-07, + "loss": 0.688, + "num_input_tokens_seen": 144283120, + "step": 6769 + }, + { + "epoch": 0.8140443696266458, + "flos": 14401463206920.0, + "grad_norm": 4.153005839662697, + "learning_rate": 3.517549763640197e-07, + "loss": 0.7371, + "num_input_tokens_seen": 144300480, + "step": 6770 + }, + { + "epoch": 0.8141646125172849, + "flos": 19913559124200.0, + "grad_norm": 5.209584119017106, + "learning_rate": 3.513138828271829e-07, + "loss": 0.6962, + "num_input_tokens_seen": 144320070, + "step": 6771 + }, + { + "epoch": 0.8142848554079241, + "flos": 29173380046680.0, + "grad_norm": 3.7765514091104726, + "learning_rate": 3.508730393947179e-07, + "loss": 0.6963, + "num_input_tokens_seen": 144343045, + "step": 6772 + }, + { + "epoch": 0.8144050982985631, + "flos": 16270325681520.0, + "grad_norm": 3.165221804881402, + "learning_rate": 3.504324461335024e-07, + "loss": 0.6937, + "num_input_tokens_seen": 144362875, + "step": 6773 + }, + { + "epoch": 0.8145253411892022, + "flos": 16862838002880.0, + "grad_norm": 3.2975635261576706, + "learning_rate": 3.499921031103732e-07, + "loss": 0.8644, + "num_input_tokens_seen": 144383365, + "step": 6774 + }, + { + "epoch": 0.8146455840798413, + "flos": 18187365632520.0, + "grad_norm": 5.809615208921536, + "learning_rate": 3.4955201039212987e-07, + "loss": 0.7631, + "num_input_tokens_seen": 144404005, + "step": 6775 + }, + { + "epoch": 0.8147658269704804, + "flos": 14595633962040.0, + "grad_norm": 5.169476370644833, + "learning_rate": 3.4911216804553465e-07, + "loss": 0.6291, + "num_input_tokens_seen": 144422625, + "step": 6776 + }, + { + "epoch": 0.8148860698611194, + "flos": 15486238721760.0, + "grad_norm": 3.737332118930227, + "learning_rate": 3.4867257613731017e-07, + "loss": 0.688, + "num_input_tokens_seen": 144441540, + "step": 6777 + }, + { + "epoch": 0.8150063127517585, + "flos": 14323181599920.0, + "grad_norm": 3.8716049087816473, + "learning_rate": 3.4823323473414343e-07, + "loss": 0.8344, + "num_input_tokens_seen": 144460780, + "step": 6778 + }, + { + "epoch": 0.8151265556423977, + "flos": 16567120061520.0, + "grad_norm": 12.909994916730215, + "learning_rate": 3.477941439026812e-07, + "loss": 0.7307, + "num_input_tokens_seen": 144478720, + "step": 6779 + }, + { + "epoch": 0.8152467985330367, + "flos": 13110743923200.0, + "grad_norm": 3.2994866174113455, + "learning_rate": 3.473553037095349e-07, + "loss": 0.7054, + "num_input_tokens_seen": 144497465, + "step": 6780 + }, + { + "epoch": 0.8153670414236758, + "flos": 18291318975720.0, + "grad_norm": 5.21682191969138, + "learning_rate": 3.469167142212743e-07, + "loss": 0.8156, + "num_input_tokens_seen": 144519030, + "step": 6781 + }, + { + "epoch": 0.8154872843143149, + "flos": 22804107796080.0, + "grad_norm": 6.728854649477786, + "learning_rate": 3.4647837550443337e-07, + "loss": 0.6141, + "num_input_tokens_seen": 144537315, + "step": 6782 + }, + { + "epoch": 0.815607527204954, + "flos": 14164402188720.0, + "grad_norm": 5.3624050816669655, + "learning_rate": 3.460402876255086e-07, + "loss": 0.7288, + "num_input_tokens_seen": 144554425, + "step": 6783 + }, + { + "epoch": 0.815727770095593, + "flos": 19158784796040.0, + "grad_norm": 4.482767807321217, + "learning_rate": 3.456024506509574e-07, + "loss": 0.6914, + "num_input_tokens_seen": 144575065, + "step": 6784 + }, + { + "epoch": 0.8158480129862322, + "flos": 18564562836840.0, + "grad_norm": 5.253504355562677, + "learning_rate": 3.4516486464719873e-07, + "loss": 0.7287, + "num_input_tokens_seen": 144594175, + "step": 6785 + }, + { + "epoch": 0.8159682558768713, + "flos": 25289064770880.0, + "grad_norm": 5.795694318357028, + "learning_rate": 3.4472752968061445e-07, + "loss": 0.6146, + "num_input_tokens_seen": 144618325, + "step": 6786 + }, + { + "epoch": 0.8160884987675103, + "flos": 13616679728040.0, + "grad_norm": 7.792516189475157, + "learning_rate": 3.442904458175475e-07, + "loss": 0.7247, + "num_input_tokens_seen": 144635365, + "step": 6787 + }, + { + "epoch": 0.8162087416581495, + "flos": 23073774081720.0, + "grad_norm": 4.028550064470869, + "learning_rate": 3.438536131243044e-07, + "loss": 0.7417, + "num_input_tokens_seen": 144656245, + "step": 6788 + }, + { + "epoch": 0.8163289845487885, + "flos": 27634708718520.0, + "grad_norm": 3.421173859969581, + "learning_rate": 3.434170316671503e-07, + "loss": 0.6048, + "num_input_tokens_seen": 144680995, + "step": 6789 + }, + { + "epoch": 0.8164492274394276, + "flos": 10139982386760.0, + "grad_norm": 6.39208997429651, + "learning_rate": 3.4298070151231583e-07, + "loss": 0.875, + "num_input_tokens_seen": 144696115, + "step": 6790 + }, + { + "epoch": 0.8165694703300668, + "flos": 21265151528280.0, + "grad_norm": 4.901135542724186, + "learning_rate": 3.425446227259916e-07, + "loss": 0.586, + "num_input_tokens_seen": 144716800, + "step": 6791 + }, + { + "epoch": 0.8166897132207058, + "flos": 18402332490000.0, + "grad_norm": 12.085743700975296, + "learning_rate": 3.421087953743296e-07, + "loss": 0.8069, + "num_input_tokens_seen": 144736285, + "step": 6792 + }, + { + "epoch": 0.8168099561113449, + "flos": 16944633865440.0, + "grad_norm": 4.0497638303868015, + "learning_rate": 3.416732195234464e-07, + "loss": 0.7784, + "num_input_tokens_seen": 144756060, + "step": 6793 + }, + { + "epoch": 0.816930199001984, + "flos": 13436344375440.0, + "grad_norm": 2.654985553249559, + "learning_rate": 3.4123789523941613e-07, + "loss": 0.7784, + "num_input_tokens_seen": 144775605, + "step": 6794 + }, + { + "epoch": 0.8170504418926231, + "flos": 15540526570440.0, + "grad_norm": 3.6870360001704263, + "learning_rate": 3.4080282258827884e-07, + "loss": 0.6242, + "num_input_tokens_seen": 144793700, + "step": 6795 + }, + { + "epoch": 0.8171706847832622, + "flos": 13947915653160.0, + "grad_norm": 5.667869141095791, + "learning_rate": 3.403680016360342e-07, + "loss": 0.6959, + "num_input_tokens_seen": 144812025, + "step": 6796 + }, + { + "epoch": 0.8172909276739013, + "flos": 15702440317680.0, + "grad_norm": 4.733643327331563, + "learning_rate": 3.3993343244864403e-07, + "loss": 0.656, + "num_input_tokens_seen": 144831335, + "step": 6797 + }, + { + "epoch": 0.8174111705645404, + "flos": 20238811316880.0, + "grad_norm": 3.650401858847749, + "learning_rate": 3.394991150920323e-07, + "loss": 0.7089, + "num_input_tokens_seen": 144854175, + "step": 6798 + }, + { + "epoch": 0.8175314134551794, + "flos": 10221999869040.0, + "grad_norm": 4.001170225626307, + "learning_rate": 3.3906504963208396e-07, + "loss": 0.7278, + "num_input_tokens_seen": 144870590, + "step": 6799 + }, + { + "epoch": 0.8176516563458186, + "flos": 16618241914200.0, + "grad_norm": 4.592382318258756, + "learning_rate": 3.3863123613464774e-07, + "loss": 0.6417, + "num_input_tokens_seen": 144889210, + "step": 6800 + }, + { + "epoch": 0.8177718992364577, + "flos": 16054314045360.0, + "grad_norm": 5.131363988021458, + "learning_rate": 3.381976746655317e-07, + "loss": 0.7249, + "num_input_tokens_seen": 144908685, + "step": 6801 + }, + { + "epoch": 0.8178921421270967, + "flos": 16107493795440.0, + "grad_norm": 5.231498269373321, + "learning_rate": 3.3776436529050756e-07, + "loss": 0.6583, + "num_input_tokens_seen": 144927955, + "step": 6802 + }, + { + "epoch": 0.8180123850177359, + "flos": 24370983657240.0, + "grad_norm": 7.849483021034555, + "learning_rate": 3.373313080753073e-07, + "loss": 0.7095, + "num_input_tokens_seen": 144951735, + "step": 6803 + }, + { + "epoch": 0.8181326279083749, + "flos": 16162161563640.0, + "grad_norm": 6.650151427416398, + "learning_rate": 3.3689850308562527e-07, + "loss": 0.7671, + "num_input_tokens_seen": 144971900, + "step": 6804 + }, + { + "epoch": 0.818252870799014, + "flos": 11436463783200.0, + "grad_norm": 5.096981965701998, + "learning_rate": 3.364659503871183e-07, + "loss": 0.7608, + "num_input_tokens_seen": 144989555, + "step": 6805 + }, + { + "epoch": 0.8183731136896532, + "flos": 13704744262680.0, + "grad_norm": 4.013007472381365, + "learning_rate": 3.3603365004540417e-07, + "loss": 0.8177, + "num_input_tokens_seen": 145007570, + "step": 6806 + }, + { + "epoch": 0.8184933565802922, + "flos": 19455832455720.0, + "grad_norm": 5.599351859050749, + "learning_rate": 3.356016021260624e-07, + "loss": 0.7486, + "num_input_tokens_seen": 145027620, + "step": 6807 + }, + { + "epoch": 0.8186135994709313, + "flos": 12786631489080.0, + "grad_norm": 11.625210011543878, + "learning_rate": 3.35169806694634e-07, + "loss": 0.6259, + "num_input_tokens_seen": 145045590, + "step": 6808 + }, + { + "epoch": 0.8187338423615703, + "flos": 46800083277600.0, + "grad_norm": 0.7422497371879978, + "learning_rate": 3.3473826381662186e-07, + "loss": 0.6141, + "num_input_tokens_seen": 145116450, + "step": 6809 + }, + { + "epoch": 0.8188540852522095, + "flos": 12786504849240.0, + "grad_norm": 12.789406111755396, + "learning_rate": 3.3430697355749216e-07, + "loss": 0.8024, + "num_input_tokens_seen": 145133860, + "step": 6810 + }, + { + "epoch": 0.8189743281428485, + "flos": 10464633040200.0, + "grad_norm": 4.3504356692272745, + "learning_rate": 3.3387593598266907e-07, + "loss": 0.7244, + "num_input_tokens_seen": 145150190, + "step": 6811 + }, + { + "epoch": 0.8190945710334876, + "flos": 18374096297040.0, + "grad_norm": 3.3997574270893676, + "learning_rate": 3.3344515115754225e-07, + "loss": 0.7698, + "num_input_tokens_seen": 145168890, + "step": 6812 + }, + { + "epoch": 0.8192148139241268, + "flos": 15729916671600.0, + "grad_norm": 5.051452261715692, + "learning_rate": 3.33014619147461e-07, + "loss": 0.7755, + "num_input_tokens_seen": 145186635, + "step": 6813 + }, + { + "epoch": 0.8193350568147658, + "flos": 17539457363880.0, + "grad_norm": 6.637879430123556, + "learning_rate": 3.325843400177362e-07, + "loss": 0.6896, + "num_input_tokens_seen": 145207695, + "step": 6814 + }, + { + "epoch": 0.8194552997054049, + "flos": 15028702013040.0, + "grad_norm": 3.817789254712208, + "learning_rate": 3.32154313833642e-07, + "loss": 0.7164, + "num_input_tokens_seen": 145227570, + "step": 6815 + }, + { + "epoch": 0.819575542596044, + "flos": 19078476951600.0, + "grad_norm": 4.093319003880506, + "learning_rate": 3.3172454066041164e-07, + "loss": 0.5726, + "num_input_tokens_seen": 145246795, + "step": 6816 + }, + { + "epoch": 0.8196957854866831, + "flos": 21346979050800.0, + "grad_norm": 6.037955711968041, + "learning_rate": 3.3129502056324234e-07, + "loss": 0.7371, + "num_input_tokens_seen": 145267880, + "step": 6817 + }, + { + "epoch": 0.8198160283773221, + "flos": 50905819551360.0, + "grad_norm": 0.7885394109360259, + "learning_rate": 3.3086575360729165e-07, + "loss": 0.6155, + "num_input_tokens_seen": 145325135, + "step": 6818 + }, + { + "epoch": 0.8199362712679613, + "flos": 12111658446000.0, + "grad_norm": 4.608425236498949, + "learning_rate": 3.3043673985767906e-07, + "loss": 0.6972, + "num_input_tokens_seen": 145343920, + "step": 6819 + }, + { + "epoch": 0.8200565141586004, + "flos": 15914621098680.0, + "grad_norm": 3.198298246169008, + "learning_rate": 3.3000797937948564e-07, + "loss": 0.7541, + "num_input_tokens_seen": 145361935, + "step": 6820 + }, + { + "epoch": 0.8201767570492394, + "flos": 51479277068160.0, + "grad_norm": 0.9332300858497173, + "learning_rate": 3.295794722377534e-07, + "loss": 0.6701, + "num_input_tokens_seen": 145425260, + "step": 6821 + }, + { + "epoch": 0.8202969999398786, + "flos": 16917442451160.0, + "grad_norm": 3.3396779872515725, + "learning_rate": 3.291512184974876e-07, + "loss": 0.7818, + "num_input_tokens_seen": 145445370, + "step": 6822 + }, + { + "epoch": 0.8204172428305176, + "flos": 20698342603080.0, + "grad_norm": 6.162859316984082, + "learning_rate": 3.2872321822365346e-07, + "loss": 0.6526, + "num_input_tokens_seen": 145465305, + "step": 6823 + }, + { + "epoch": 0.8205374857211567, + "flos": 15270986924640.0, + "grad_norm": 8.791014106240148, + "learning_rate": 3.282954714811783e-07, + "loss": 0.7175, + "num_input_tokens_seen": 145483930, + "step": 6824 + }, + { + "epoch": 0.8206577286117959, + "flos": 9546678566400.0, + "grad_norm": 5.13692222743377, + "learning_rate": 3.2786797833495093e-07, + "loss": 0.6933, + "num_input_tokens_seen": 145499005, + "step": 6825 + }, + { + "epoch": 0.8207779715024349, + "flos": 18511509726600.0, + "grad_norm": 4.326603480868574, + "learning_rate": 3.274407388498213e-07, + "loss": 0.7142, + "num_input_tokens_seen": 145516855, + "step": 6826 + }, + { + "epoch": 0.820898214393074, + "flos": 14326125976200.0, + "grad_norm": 3.718393132372832, + "learning_rate": 3.270137530906021e-07, + "loss": 0.7316, + "num_input_tokens_seen": 145535810, + "step": 6827 + }, + { + "epoch": 0.8210184572837131, + "flos": 11355744359280.0, + "grad_norm": 3.872287662366047, + "learning_rate": 3.265870211220665e-07, + "loss": 0.8148, + "num_input_tokens_seen": 145553365, + "step": 6828 + }, + { + "epoch": 0.8211387001743522, + "flos": 15216889035720.0, + "grad_norm": 6.743003893033493, + "learning_rate": 3.2616054300894934e-07, + "loss": 0.7943, + "num_input_tokens_seen": 145572535, + "step": 6829 + }, + { + "epoch": 0.8212589430649913, + "flos": 20315699885640.0, + "grad_norm": 7.295709096161956, + "learning_rate": 3.2573431881594693e-07, + "loss": 0.8387, + "num_input_tokens_seen": 145591800, + "step": 6830 + }, + { + "epoch": 0.8213791859556304, + "flos": 16431827849280.0, + "grad_norm": 5.086741862876572, + "learning_rate": 3.2530834860771663e-07, + "loss": 0.6328, + "num_input_tokens_seen": 145610900, + "step": 6831 + }, + { + "epoch": 0.8214994288462695, + "flos": 12138596580600.0, + "grad_norm": 3.5928594398003657, + "learning_rate": 3.248826324488794e-07, + "loss": 0.7188, + "num_input_tokens_seen": 145627915, + "step": 6832 + }, + { + "epoch": 0.8216196717369085, + "flos": 18478017980280.0, + "grad_norm": 6.1584527581817055, + "learning_rate": 3.244571704040138e-07, + "loss": 0.8614, + "num_input_tokens_seen": 145647795, + "step": 6833 + }, + { + "epoch": 0.8217399146275477, + "flos": 18589728013680.0, + "grad_norm": 5.228928741675723, + "learning_rate": 3.2403196253766374e-07, + "loss": 0.7403, + "num_input_tokens_seen": 145666595, + "step": 6834 + }, + { + "epoch": 0.8218601575181868, + "flos": 18780764432760.0, + "grad_norm": 6.788112841736534, + "learning_rate": 3.2360700891433254e-07, + "loss": 0.7724, + "num_input_tokens_seen": 145685340, + "step": 6835 + }, + { + "epoch": 0.8219804004088258, + "flos": 49891256845080.0, + "grad_norm": 0.8264599741932342, + "learning_rate": 3.231823095984847e-07, + "loss": 0.5848, + "num_input_tokens_seen": 145739700, + "step": 6836 + }, + { + "epoch": 0.822100643299465, + "flos": 14218088498160.0, + "grad_norm": 13.94124481116251, + "learning_rate": 3.2275786465454814e-07, + "loss": 0.7419, + "num_input_tokens_seen": 145756070, + "step": 6837 + }, + { + "epoch": 0.822220886190104, + "flos": 18079043214840.0, + "grad_norm": 4.334080093791463, + "learning_rate": 3.2233367414690917e-07, + "loss": 0.7447, + "num_input_tokens_seen": 145777980, + "step": 6838 + }, + { + "epoch": 0.8223411290807431, + "flos": 20401136643600.0, + "grad_norm": 6.863849544716854, + "learning_rate": 3.219097381399183e-07, + "loss": 0.8205, + "num_input_tokens_seen": 145794875, + "step": 6839 + }, + { + "epoch": 0.8224613719713821, + "flos": 16995977337840.0, + "grad_norm": 3.913583570101507, + "learning_rate": 3.2148605669788584e-07, + "loss": 0.7957, + "num_input_tokens_seen": 145814485, + "step": 6840 + }, + { + "epoch": 0.8225816148620213, + "flos": 11436907022640.0, + "grad_norm": 8.220645816428302, + "learning_rate": 3.2106262988508405e-07, + "loss": 0.7501, + "num_input_tokens_seen": 145832255, + "step": 6841 + }, + { + "epoch": 0.8227018577526604, + "flos": 13515575781240.0, + "grad_norm": 4.437366322397824, + "learning_rate": 3.206394577657465e-07, + "loss": 0.721, + "num_input_tokens_seen": 145849755, + "step": 6842 + }, + { + "epoch": 0.8228221006432994, + "flos": 16269882442080.0, + "grad_norm": 4.3337641037281545, + "learning_rate": 3.202165404040675e-07, + "loss": 0.6945, + "num_input_tokens_seen": 145867395, + "step": 6843 + }, + { + "epoch": 0.8229423435339386, + "flos": 17646956622600.0, + "grad_norm": 5.2792710428173955, + "learning_rate": 3.1979387786420396e-07, + "loss": 0.7227, + "num_input_tokens_seen": 145887355, + "step": 6844 + }, + { + "epoch": 0.8230625864245776, + "flos": 17484884575560.0, + "grad_norm": 4.552638644893909, + "learning_rate": 3.1937147021027346e-07, + "loss": 0.8088, + "num_input_tokens_seen": 145905530, + "step": 6845 + }, + { + "epoch": 0.8231828293152167, + "flos": 12084245412000.0, + "grad_norm": 6.14443595215753, + "learning_rate": 3.189493175063547e-07, + "loss": 0.7519, + "num_input_tokens_seen": 145922485, + "step": 6846 + }, + { + "epoch": 0.8233030722058559, + "flos": 13785685306320.0, + "grad_norm": 4.126322678714906, + "learning_rate": 3.1852741981648776e-07, + "loss": 0.6592, + "num_input_tokens_seen": 145940855, + "step": 6847 + }, + { + "epoch": 0.8234233150964949, + "flos": 20859908090760.0, + "grad_norm": 8.42206342553593, + "learning_rate": 3.1810577720467404e-07, + "loss": 0.6769, + "num_input_tokens_seen": 145962305, + "step": 6848 + }, + { + "epoch": 0.823543557987134, + "flos": 24802183770600.0, + "grad_norm": 3.709124377171074, + "learning_rate": 3.176843897348769e-07, + "loss": 0.5574, + "num_input_tokens_seen": 145985220, + "step": 6849 + }, + { + "epoch": 0.8236638008777731, + "flos": 12462677354760.0, + "grad_norm": 7.411968872186112, + "learning_rate": 3.1726325747102034e-07, + "loss": 0.7379, + "num_input_tokens_seen": 146003315, + "step": 6850 + }, + { + "epoch": 0.8237840437684122, + "flos": 45430693484640.0, + "grad_norm": 3.9704281120176694, + "learning_rate": 3.1684238047698974e-07, + "loss": 0.6306, + "num_input_tokens_seen": 146031305, + "step": 6851 + }, + { + "epoch": 0.8239042866590512, + "flos": 20023781139480.0, + "grad_norm": 8.885637382696716, + "learning_rate": 3.1642175881663155e-07, + "loss": 0.5067, + "num_input_tokens_seen": 146050755, + "step": 6852 + }, + { + "epoch": 0.8240245295496904, + "flos": 15892178678400.0, + "grad_norm": 7.667305325430124, + "learning_rate": 3.160013925537537e-07, + "loss": 0.8312, + "num_input_tokens_seen": 146071310, + "step": 6853 + }, + { + "epoch": 0.8241447724403295, + "flos": 14621147398440.0, + "grad_norm": 5.008552675104056, + "learning_rate": 3.155812817521266e-07, + "loss": 0.7346, + "num_input_tokens_seen": 146091405, + "step": 6854 + }, + { + "epoch": 0.8242650153309685, + "flos": 16296630616920.0, + "grad_norm": 8.055763920102537, + "learning_rate": 3.151614264754787e-07, + "loss": 0.7593, + "num_input_tokens_seen": 146109070, + "step": 6855 + }, + { + "epoch": 0.8243852582216077, + "flos": 16324075310880.0, + "grad_norm": 13.657528745372058, + "learning_rate": 3.147418267875035e-07, + "loss": 0.7825, + "num_input_tokens_seen": 146126920, + "step": 6856 + }, + { + "epoch": 0.8245055011122467, + "flos": 18052516659720.0, + "grad_norm": 6.405173606693636, + "learning_rate": 3.1432248275185315e-07, + "loss": 0.6328, + "num_input_tokens_seen": 146147150, + "step": 6857 + }, + { + "epoch": 0.8246257440028858, + "flos": 12754026221640.0, + "grad_norm": 5.113424860115657, + "learning_rate": 3.139033944321412e-07, + "loss": 0.7648, + "num_input_tokens_seen": 146164230, + "step": 6858 + }, + { + "epoch": 0.824745986893525, + "flos": 18322404565080.0, + "grad_norm": 4.076232003105276, + "learning_rate": 3.1348456189194507e-07, + "loss": 0.779, + "num_input_tokens_seen": 146184410, + "step": 6859 + }, + { + "epoch": 0.824866229784164, + "flos": 13706517220440.0, + "grad_norm": 4.1050652361017805, + "learning_rate": 3.1306598519479876e-07, + "loss": 0.8196, + "num_input_tokens_seen": 146203950, + "step": 6860 + }, + { + "epoch": 0.8249864726748031, + "flos": 17457978100920.0, + "grad_norm": 2.979617656015214, + "learning_rate": 3.1264766440420177e-07, + "loss": 0.7672, + "num_input_tokens_seen": 146226140, + "step": 6861 + }, + { + "epoch": 0.8251067155654422, + "flos": 14838837012480.0, + "grad_norm": 5.552391725984775, + "learning_rate": 3.122295995836124e-07, + "loss": 0.6672, + "num_input_tokens_seen": 146245730, + "step": 6862 + }, + { + "epoch": 0.8252269584560813, + "flos": 18322024645560.0, + "grad_norm": 3.4008230671676656, + "learning_rate": 3.118117907964508e-07, + "loss": 0.7538, + "num_input_tokens_seen": 146267395, + "step": 6863 + }, + { + "epoch": 0.8253472013467203, + "flos": 12489425529600.0, + "grad_norm": 4.016872108450122, + "learning_rate": 3.1139423810609856e-07, + "loss": 0.7999, + "num_input_tokens_seen": 146283810, + "step": 6864 + }, + { + "epoch": 0.8254674442373595, + "flos": 16400995539600.0, + "grad_norm": 3.612194858850118, + "learning_rate": 3.1097694157589714e-07, + "loss": 0.7355, + "num_input_tokens_seen": 146303415, + "step": 6865 + }, + { + "epoch": 0.8255876871279986, + "flos": 18156470002920.0, + "grad_norm": 7.240310508012858, + "learning_rate": 3.105599012691511e-07, + "loss": 0.7323, + "num_input_tokens_seen": 146321565, + "step": 6866 + }, + { + "epoch": 0.8257079300186376, + "flos": 20132198537040.0, + "grad_norm": 2.4234550982093954, + "learning_rate": 3.101431172491249e-07, + "loss": 0.8086, + "num_input_tokens_seen": 146342830, + "step": 6867 + }, + { + "epoch": 0.8258281729092768, + "flos": 12003431008200.0, + "grad_norm": 4.723098872971768, + "learning_rate": 3.097265895790444e-07, + "loss": 0.7063, + "num_input_tokens_seen": 146360760, + "step": 6868 + }, + { + "epoch": 0.8259484157999158, + "flos": 15675597162960.0, + "grad_norm": 3.2870681769180012, + "learning_rate": 3.093103183220962e-07, + "loss": 0.8177, + "num_input_tokens_seen": 146380525, + "step": 6869 + }, + { + "epoch": 0.8260686586905549, + "flos": 42980942385360.0, + "grad_norm": 0.9343758024894815, + "learning_rate": 3.0889430354142796e-07, + "loss": 0.6384, + "num_input_tokens_seen": 146441755, + "step": 6870 + }, + { + "epoch": 0.826188901581194, + "flos": 20185219987320.0, + "grad_norm": 6.938655722638549, + "learning_rate": 3.084785453001497e-07, + "loss": 0.6893, + "num_input_tokens_seen": 146462390, + "step": 6871 + }, + { + "epoch": 0.8263091444718331, + "flos": 17349813983040.0, + "grad_norm": 4.428041720902781, + "learning_rate": 3.080630436613314e-07, + "loss": 0.7967, + "num_input_tokens_seen": 146479880, + "step": 6872 + }, + { + "epoch": 0.8264293873624722, + "flos": 12516521964000.0, + "grad_norm": 4.822351651913402, + "learning_rate": 3.076477986880039e-07, + "loss": 0.8351, + "num_input_tokens_seen": 146497395, + "step": 6873 + }, + { + "epoch": 0.8265496302531112, + "flos": 17647938081360.0, + "grad_norm": 3.474115569270467, + "learning_rate": 3.0723281044315986e-07, + "loss": 0.6815, + "num_input_tokens_seen": 146519070, + "step": 6874 + }, + { + "epoch": 0.8266698731437504, + "flos": 10247956544880.0, + "grad_norm": 5.676678097695299, + "learning_rate": 3.068180789897521e-07, + "loss": 0.7421, + "num_input_tokens_seen": 146537200, + "step": 6875 + }, + { + "epoch": 0.8267901160343895, + "flos": 22591072196160.0, + "grad_norm": 3.4248680367768825, + "learning_rate": 3.064036043906966e-07, + "loss": 0.8074, + "num_input_tokens_seen": 146560360, + "step": 6876 + }, + { + "epoch": 0.8269103589250285, + "flos": 29878235600640.0, + "grad_norm": 4.208887170788814, + "learning_rate": 3.059893867088668e-07, + "loss": 0.6665, + "num_input_tokens_seen": 146584225, + "step": 6877 + }, + { + "epoch": 0.8270306018156677, + "flos": 22210297416360.0, + "grad_norm": 5.830188375737274, + "learning_rate": 3.055754260071004e-07, + "loss": 0.6579, + "num_input_tokens_seen": 146606240, + "step": 6878 + }, + { + "epoch": 0.8271508447063067, + "flos": 18482260414920.0, + "grad_norm": 4.301822879225563, + "learning_rate": 3.051617223481948e-07, + "loss": 0.7274, + "num_input_tokens_seen": 146627280, + "step": 6879 + }, + { + "epoch": 0.8272710875969458, + "flos": 12813632943600.0, + "grad_norm": 5.012574976790793, + "learning_rate": 3.047482757949078e-07, + "loss": 0.7408, + "num_input_tokens_seen": 146644630, + "step": 6880 + }, + { + "epoch": 0.827391330487585, + "flos": 14433498595080.0, + "grad_norm": 11.347540608274787, + "learning_rate": 3.043350864099605e-07, + "loss": 0.8366, + "num_input_tokens_seen": 146662910, + "step": 6881 + }, + { + "epoch": 0.827511573378224, + "flos": 12272084175120.0, + "grad_norm": 4.637568125262968, + "learning_rate": 3.039221542560315e-07, + "loss": 0.796, + "num_input_tokens_seen": 146679195, + "step": 6882 + }, + { + "epoch": 0.8276318162688631, + "flos": 13407633283080.0, + "grad_norm": 5.1408674652521515, + "learning_rate": 3.0350947939576356e-07, + "loss": 0.7323, + "num_input_tokens_seen": 146698070, + "step": 6883 + }, + { + "epoch": 0.8277520591595022, + "flos": 14135216196960.0, + "grad_norm": 5.018431071673953, + "learning_rate": 3.0309706189175876e-07, + "loss": 0.7085, + "num_input_tokens_seen": 146717625, + "step": 6884 + }, + { + "epoch": 0.8278723020501413, + "flos": 50083116423120.0, + "grad_norm": 0.7876317110876067, + "learning_rate": 3.0268490180658045e-07, + "loss": 0.5863, + "num_input_tokens_seen": 146780125, + "step": 6885 + }, + { + "epoch": 0.8279925449407803, + "flos": 13271676211680.0, + "grad_norm": 4.759476294037597, + "learning_rate": 3.0227299920275305e-07, + "loss": 0.7651, + "num_input_tokens_seen": 146796160, + "step": 6886 + }, + { + "epoch": 0.8281127878314195, + "flos": 15081818443200.0, + "grad_norm": 3.9675949698741557, + "learning_rate": 3.018613541427613e-07, + "loss": 0.8365, + "num_input_tokens_seen": 146815400, + "step": 6887 + }, + { + "epoch": 0.8282330307220586, + "flos": 13137523758000.0, + "grad_norm": 3.868294112502346, + "learning_rate": 3.0144996668905243e-07, + "loss": 0.722, + "num_input_tokens_seen": 146832500, + "step": 6888 + }, + { + "epoch": 0.8283532736126976, + "flos": 14676764965440.0, + "grad_norm": 4.663437179997937, + "learning_rate": 3.010388369040331e-07, + "loss": 0.8126, + "num_input_tokens_seen": 146850880, + "step": 6889 + }, + { + "epoch": 0.8284735165033368, + "flos": 23397854855880.0, + "grad_norm": 3.015807151234558, + "learning_rate": 3.0062796485007156e-07, + "loss": 0.8211, + "num_input_tokens_seen": 146871540, + "step": 6890 + }, + { + "epoch": 0.8285937593939758, + "flos": 19536900139200.0, + "grad_norm": 11.184028299931644, + "learning_rate": 3.002173505894965e-07, + "loss": 0.6391, + "num_input_tokens_seen": 146890410, + "step": 6891 + }, + { + "epoch": 0.8287140022846149, + "flos": 14757864308880.0, + "grad_norm": 5.195754504224007, + "learning_rate": 2.998069941845973e-07, + "loss": 0.6007, + "num_input_tokens_seen": 146909200, + "step": 6892 + }, + { + "epoch": 0.8288342451752541, + "flos": 52181536505400.0, + "grad_norm": 0.7624022327343369, + "learning_rate": 2.993968956976258e-07, + "loss": 0.6102, + "num_input_tokens_seen": 146976665, + "step": 6893 + }, + { + "epoch": 0.8289544880658931, + "flos": 17998323790920.0, + "grad_norm": 5.021509444219802, + "learning_rate": 2.9898705519079313e-07, + "loss": 0.682, + "num_input_tokens_seen": 146995490, + "step": 6894 + }, + { + "epoch": 0.8290747309565322, + "flos": 16296978876480.0, + "grad_norm": 3.2511789677716845, + "learning_rate": 2.985774727262715e-07, + "loss": 0.729, + "num_input_tokens_seen": 147014055, + "step": 6895 + }, + { + "epoch": 0.8291949738471713, + "flos": 17023453691760.0, + "grad_norm": 5.2181107363664125, + "learning_rate": 2.981681483661949e-07, + "loss": 0.7944, + "num_input_tokens_seen": 147033360, + "step": 6896 + }, + { + "epoch": 0.8293152167378104, + "flos": 38707869528480.0, + "grad_norm": 3.3615126794646573, + "learning_rate": 2.9775908217265633e-07, + "loss": 0.6905, + "num_input_tokens_seen": 147058315, + "step": 6897 + }, + { + "epoch": 0.8294354596284494, + "flos": 46706129990400.0, + "grad_norm": 0.8169394764047827, + "learning_rate": 2.9735027420771253e-07, + "loss": 0.5142, + "num_input_tokens_seen": 147118370, + "step": 6898 + }, + { + "epoch": 0.8295557025190886, + "flos": 18186795753240.0, + "grad_norm": 6.597033221119675, + "learning_rate": 2.969417245333774e-07, + "loss": 0.6961, + "num_input_tokens_seen": 147137470, + "step": 6899 + }, + { + "epoch": 0.8296759454097277, + "flos": 18402522449760.0, + "grad_norm": 8.513900162186522, + "learning_rate": 2.9653343321162915e-07, + "loss": 0.7643, + "num_input_tokens_seen": 147156700, + "step": 6900 + }, + { + "epoch": 0.8297961883003667, + "flos": 17672691678720.0, + "grad_norm": 4.024272469110366, + "learning_rate": 2.9612540030440446e-07, + "loss": 0.6328, + "num_input_tokens_seen": 147176965, + "step": 6901 + }, + { + "epoch": 0.8299164311910058, + "flos": 49733870472120.0, + "grad_norm": 0.8367147664635128, + "learning_rate": 2.9571762587360206e-07, + "loss": 0.6364, + "num_input_tokens_seen": 147233070, + "step": 6902 + }, + { + "epoch": 0.8300366740816449, + "flos": 18484096692600.0, + "grad_norm": 3.6169929204886806, + "learning_rate": 2.953101099810806e-07, + "loss": 0.7255, + "num_input_tokens_seen": 147252395, + "step": 6903 + }, + { + "epoch": 0.830156916972284, + "flos": 13164746832240.0, + "grad_norm": 4.944370559730894, + "learning_rate": 2.9490285268865965e-07, + "loss": 0.8246, + "num_input_tokens_seen": 147269605, + "step": 6904 + }, + { + "epoch": 0.830277159862923, + "flos": 19293475469040.0, + "grad_norm": 5.530487870119866, + "learning_rate": 2.9449585405812085e-07, + "loss": 0.7873, + "num_input_tokens_seen": 147286705, + "step": 6905 + }, + { + "epoch": 0.8303974027535622, + "flos": 14569360686600.0, + "grad_norm": 3.5010799376735844, + "learning_rate": 2.940891141512043e-07, + "loss": 0.7275, + "num_input_tokens_seen": 147304445, + "step": 6906 + }, + { + "epoch": 0.8305176456442013, + "flos": 12516870223560.0, + "grad_norm": 4.352962467311347, + "learning_rate": 2.9368263302961385e-07, + "loss": 0.6962, + "num_input_tokens_seen": 147322865, + "step": 6907 + }, + { + "epoch": 0.8306378885348403, + "flos": 18778991475000.0, + "grad_norm": 5.988657084177971, + "learning_rate": 2.9327641075501075e-07, + "loss": 0.7849, + "num_input_tokens_seen": 147341575, + "step": 6908 + }, + { + "epoch": 0.8307581314254795, + "flos": 24936747803760.0, + "grad_norm": 5.415161021366687, + "learning_rate": 2.9287044738901866e-07, + "loss": 0.6387, + "num_input_tokens_seen": 147359280, + "step": 6909 + }, + { + "epoch": 0.8308783743161186, + "flos": 12810213667920.0, + "grad_norm": 3.172054642724981, + "learning_rate": 2.9246474299322274e-07, + "loss": 0.9051, + "num_input_tokens_seen": 147374560, + "step": 6910 + }, + { + "epoch": 0.8309986172067576, + "flos": 51188529740520.0, + "grad_norm": 0.9045361064412161, + "learning_rate": 2.920592976291678e-07, + "loss": 0.6554, + "num_input_tokens_seen": 147431610, + "step": 6911 + }, + { + "epoch": 0.8311188600973968, + "flos": 16323885351120.0, + "grad_norm": 5.011056691278598, + "learning_rate": 2.916541113583595e-07, + "loss": 0.7924, + "num_input_tokens_seen": 147449830, + "step": 6912 + }, + { + "epoch": 0.8312391029880358, + "flos": 13704206043360.0, + "grad_norm": 5.546347665569493, + "learning_rate": 2.912491842422642e-07, + "loss": 0.6452, + "num_input_tokens_seen": 147467255, + "step": 6913 + }, + { + "epoch": 0.8313593458786749, + "flos": 14893346480880.0, + "grad_norm": 5.30977264406078, + "learning_rate": 2.9084451634230857e-07, + "loss": 0.6876, + "num_input_tokens_seen": 147486275, + "step": 6914 + }, + { + "epoch": 0.831479588769314, + "flos": 23587086657240.0, + "grad_norm": 9.51987105723019, + "learning_rate": 2.9044010771988125e-07, + "loss": 0.7092, + "num_input_tokens_seen": 147505810, + "step": 6915 + }, + { + "epoch": 0.8315998316599531, + "flos": 33253734015240.0, + "grad_norm": 3.286647752884097, + "learning_rate": 2.900359584363303e-07, + "loss": 0.7226, + "num_input_tokens_seen": 147528635, + "step": 6916 + }, + { + "epoch": 0.8317200745505922, + "flos": 13402789309200.0, + "grad_norm": 5.919141153463812, + "learning_rate": 2.8963206855296494e-07, + "loss": 0.824, + "num_input_tokens_seen": 147544595, + "step": 6917 + }, + { + "epoch": 0.8318403174412313, + "flos": 17727581066640.0, + "grad_norm": 3.254882958999022, + "learning_rate": 2.892284381310548e-07, + "loss": 0.761, + "num_input_tokens_seen": 147565730, + "step": 6918 + }, + { + "epoch": 0.8319605603318704, + "flos": 16404509795160.0, + "grad_norm": 5.312939075949945, + "learning_rate": 2.888250672318302e-07, + "loss": 0.6976, + "num_input_tokens_seen": 147582850, + "step": 6919 + }, + { + "epoch": 0.8320808032225094, + "flos": 27502360882560.0, + "grad_norm": 5.071396722060345, + "learning_rate": 2.884219559164831e-07, + "loss": 0.6813, + "num_input_tokens_seen": 147605715, + "step": 6920 + }, + { + "epoch": 0.8322010461131486, + "flos": 9276759001080.0, + "grad_norm": 3.613020603985903, + "learning_rate": 2.880191042461635e-07, + "loss": 0.7986, + "num_input_tokens_seen": 147621375, + "step": 6921 + }, + { + "epoch": 0.8323212890037877, + "flos": 11517689766480.0, + "grad_norm": 3.5799928812137343, + "learning_rate": 2.876165122819849e-07, + "loss": 0.7911, + "num_input_tokens_seen": 147639075, + "step": 6922 + }, + { + "epoch": 0.8324415318944267, + "flos": 15885941666280.0, + "grad_norm": 5.038939860426217, + "learning_rate": 2.872141800850201e-07, + "loss": 0.7776, + "num_input_tokens_seen": 147655970, + "step": 6923 + }, + { + "epoch": 0.8325617747850659, + "flos": 25121737170480.0, + "grad_norm": 4.186090716170493, + "learning_rate": 2.868121077163024e-07, + "loss": 0.7207, + "num_input_tokens_seen": 147675245, + "step": 6924 + }, + { + "epoch": 0.8326820176757049, + "flos": 13407823242840.0, + "grad_norm": 8.65336257757213, + "learning_rate": 2.864102952368257e-07, + "loss": 0.7093, + "num_input_tokens_seen": 147692890, + "step": 6925 + }, + { + "epoch": 0.832802260566344, + "flos": 26449082536560.0, + "grad_norm": 6.16505816677402, + "learning_rate": 2.860087427075444e-07, + "loss": 0.5766, + "num_input_tokens_seen": 147716860, + "step": 6926 + }, + { + "epoch": 0.8329225034569832, + "flos": 10354727624520.0, + "grad_norm": 5.3068562766726375, + "learning_rate": 2.856074501893744e-07, + "loss": 0.8398, + "num_input_tokens_seen": 147731780, + "step": 6927 + }, + { + "epoch": 0.8330427463476222, + "flos": 13191716626800.0, + "grad_norm": 3.1735937413455146, + "learning_rate": 2.8520641774319054e-07, + "loss": 0.8079, + "num_input_tokens_seen": 147749590, + "step": 6928 + }, + { + "epoch": 0.8331629892382613, + "flos": 13164588532440.0, + "grad_norm": 3.7535754642565986, + "learning_rate": 2.848056454298309e-07, + "loss": 0.7323, + "num_input_tokens_seen": 147766635, + "step": 6929 + }, + { + "epoch": 0.8332832321289004, + "flos": 12732850199760.0, + "grad_norm": 5.981066075350203, + "learning_rate": 2.844051333100905e-07, + "loss": 0.6323, + "num_input_tokens_seen": 147783900, + "step": 6930 + }, + { + "epoch": 0.8334034750195395, + "flos": 10976869177080.0, + "grad_norm": 3.5242573868957634, + "learning_rate": 2.840048814447269e-07, + "loss": 0.8245, + "num_input_tokens_seen": 147801785, + "step": 6931 + }, + { + "epoch": 0.8335237179101785, + "flos": 14190548824320.0, + "grad_norm": 7.236987571166929, + "learning_rate": 2.836048898944587e-07, + "loss": 0.7178, + "num_input_tokens_seen": 147819930, + "step": 6932 + }, + { + "epoch": 0.8336439608008177, + "flos": 15918483613800.0, + "grad_norm": 4.044574805547057, + "learning_rate": 2.832051587199642e-07, + "loss": 0.705, + "num_input_tokens_seen": 147836905, + "step": 6933 + }, + { + "epoch": 0.8337642036914568, + "flos": 44001710443800.0, + "grad_norm": 0.79840016735014, + "learning_rate": 2.828056879818821e-07, + "loss": 0.5915, + "num_input_tokens_seen": 147895700, + "step": 6934 + }, + { + "epoch": 0.8338844465820958, + "flos": 19914983822400.0, + "grad_norm": 4.851747272799882, + "learning_rate": 2.824064777408117e-07, + "loss": 0.8199, + "num_input_tokens_seen": 147915210, + "step": 6935 + }, + { + "epoch": 0.8340046894727349, + "flos": 22371166384920.0, + "grad_norm": 2.8913537009080854, + "learning_rate": 2.8200752805731263e-07, + "loss": 0.7493, + "num_input_tokens_seen": 147937920, + "step": 6936 + }, + { + "epoch": 0.834124932363374, + "flos": 19888393947360.0, + "grad_norm": 9.475709478946394, + "learning_rate": 2.8160883899190625e-07, + "loss": 0.7925, + "num_input_tokens_seen": 147960910, + "step": 6937 + }, + { + "epoch": 0.8342451752540131, + "flos": 17996234233560.0, + "grad_norm": 3.808372890230803, + "learning_rate": 2.8121041060507234e-07, + "loss": 0.7295, + "num_input_tokens_seen": 147979660, + "step": 6938 + }, + { + "epoch": 0.8343654181446521, + "flos": 19510120304400.0, + "grad_norm": 4.55577323104059, + "learning_rate": 2.808122429572528e-07, + "loss": 0.7013, + "num_input_tokens_seen": 147999585, + "step": 6939 + }, + { + "epoch": 0.8344856610352913, + "flos": 15189729281400.0, + "grad_norm": 8.974998503873483, + "learning_rate": 2.804143361088489e-07, + "loss": 0.7366, + "num_input_tokens_seen": 148018485, + "step": 6940 + }, + { + "epoch": 0.8346059039259304, + "flos": 19124754830400.0, + "grad_norm": 5.6522182194941175, + "learning_rate": 2.8001669012022277e-07, + "loss": 0.7405, + "num_input_tokens_seen": 148036175, + "step": 6941 + }, + { + "epoch": 0.8347261468165694, + "flos": 21296268777600.0, + "grad_norm": 3.4800297624038787, + "learning_rate": 2.7961930505169795e-07, + "loss": 0.6792, + "num_input_tokens_seen": 148060140, + "step": 6942 + }, + { + "epoch": 0.8348463897072086, + "flos": 19348364856960.0, + "grad_norm": 5.895602847425385, + "learning_rate": 2.792221809635558e-07, + "loss": 0.7367, + "num_input_tokens_seen": 148080490, + "step": 6943 + }, + { + "epoch": 0.8349666325978476, + "flos": 17106832552320.0, + "grad_norm": 2.608659931815962, + "learning_rate": 2.788253179160411e-07, + "loss": 0.7361, + "num_input_tokens_seen": 148101370, + "step": 6944 + }, + { + "epoch": 0.8350868754884867, + "flos": 9357478425000.0, + "grad_norm": 5.74709080617098, + "learning_rate": 2.7842871596935725e-07, + "loss": 0.6348, + "num_input_tokens_seen": 148119605, + "step": 6945 + }, + { + "epoch": 0.8352071183791259, + "flos": 19429020960960.0, + "grad_norm": 3.584283850524449, + "learning_rate": 2.780323751836682e-07, + "loss": 0.6749, + "num_input_tokens_seen": 148140540, + "step": 6946 + }, + { + "epoch": 0.8353273612697649, + "flos": 15108946537560.0, + "grad_norm": 2.774164926473573, + "learning_rate": 2.7763629561909876e-07, + "loss": 0.7727, + "num_input_tokens_seen": 148161090, + "step": 6947 + }, + { + "epoch": 0.835447604160404, + "flos": 14432327176560.0, + "grad_norm": 4.72862292061249, + "learning_rate": 2.772404773357335e-07, + "loss": 0.7548, + "num_input_tokens_seen": 148180215, + "step": 6948 + }, + { + "epoch": 0.8355678470510431, + "flos": 17156561366760.0, + "grad_norm": 3.3523129065505346, + "learning_rate": 2.7684492039361853e-07, + "loss": 0.7579, + "num_input_tokens_seen": 148199160, + "step": 6949 + }, + { + "epoch": 0.8356880899416822, + "flos": 15513303496200.0, + "grad_norm": 3.5444023487244722, + "learning_rate": 2.764496248527586e-07, + "loss": 0.826, + "num_input_tokens_seen": 148217855, + "step": 6950 + }, + { + "epoch": 0.8358083328323213, + "flos": 20562765451200.0, + "grad_norm": 4.2108127101320445, + "learning_rate": 2.760545907731211e-07, + "loss": 0.7498, + "num_input_tokens_seen": 148238150, + "step": 6951 + }, + { + "epoch": 0.8359285757229604, + "flos": 20375084987880.0, + "grad_norm": 6.504707392474987, + "learning_rate": 2.75659818214631e-07, + "loss": 0.6593, + "num_input_tokens_seen": 148258975, + "step": 6952 + }, + { + "epoch": 0.8360488186135995, + "flos": 15676736921520.0, + "grad_norm": 3.5189883271674436, + "learning_rate": 2.752653072371749e-07, + "loss": 0.7675, + "num_input_tokens_seen": 148278130, + "step": 6953 + }, + { + "epoch": 0.8361690615042385, + "flos": 20262330175800.0, + "grad_norm": 6.090602590129213, + "learning_rate": 2.7487105790060105e-07, + "loss": 0.7368, + "num_input_tokens_seen": 148297260, + "step": 6954 + }, + { + "epoch": 0.8362893043948777, + "flos": 28825653773760.0, + "grad_norm": 4.198383138153478, + "learning_rate": 2.7447707026471587e-07, + "loss": 0.6817, + "num_input_tokens_seen": 148319955, + "step": 6955 + }, + { + "epoch": 0.8364095472855168, + "flos": 18155203604520.0, + "grad_norm": 3.115777559667982, + "learning_rate": 2.740833443892874e-07, + "loss": 0.792, + "num_input_tokens_seen": 148337845, + "step": 6956 + }, + { + "epoch": 0.8365297901761558, + "flos": 16644230250000.0, + "grad_norm": 4.533185874186263, + "learning_rate": 2.7368988033404327e-07, + "loss": 0.7805, + "num_input_tokens_seen": 148355080, + "step": 6957 + }, + { + "epoch": 0.836650033066795, + "flos": 20827144523520.0, + "grad_norm": 3.29114558928285, + "learning_rate": 2.732966781586712e-07, + "loss": 0.8373, + "num_input_tokens_seen": 148374545, + "step": 6958 + }, + { + "epoch": 0.836770275957434, + "flos": 16269945762000.0, + "grad_norm": 5.083774276536616, + "learning_rate": 2.729037379228205e-07, + "loss": 0.6527, + "num_input_tokens_seen": 148394450, + "step": 6959 + }, + { + "epoch": 0.8368905188480731, + "flos": 16458671004000.0, + "grad_norm": 2.863562014606812, + "learning_rate": 2.725110596860998e-07, + "loss": 0.7932, + "num_input_tokens_seen": 148414850, + "step": 6960 + }, + { + "epoch": 0.8370107617387123, + "flos": 9708117414240.0, + "grad_norm": 4.569019623822596, + "learning_rate": 2.7211864350807776e-07, + "loss": 0.6933, + "num_input_tokens_seen": 148432770, + "step": 6961 + }, + { + "epoch": 0.8371310046293513, + "flos": 18507678871440.0, + "grad_norm": 4.129161469539345, + "learning_rate": 2.717264894482836e-07, + "loss": 0.7201, + "num_input_tokens_seen": 148452830, + "step": 6962 + }, + { + "epoch": 0.8372512475199904, + "flos": 14461164908760.0, + "grad_norm": 4.2347060633667635, + "learning_rate": 2.7133459756620646e-07, + "loss": 0.7911, + "num_input_tokens_seen": 148469745, + "step": 6963 + }, + { + "epoch": 0.8373714904106295, + "flos": 14164022269200.0, + "grad_norm": 2.8684538245804134, + "learning_rate": 2.7094296792129733e-07, + "loss": 0.7121, + "num_input_tokens_seen": 148489065, + "step": 6964 + }, + { + "epoch": 0.8374917333012686, + "flos": 10896339712920.0, + "grad_norm": 4.128961452667281, + "learning_rate": 2.7055160057296424e-07, + "loss": 0.7447, + "num_input_tokens_seen": 148506025, + "step": 6965 + }, + { + "epoch": 0.8376119761919076, + "flos": 22260089550720.0, + "grad_norm": 4.35732841989932, + "learning_rate": 2.7016049558057896e-07, + "loss": 0.7058, + "num_input_tokens_seen": 148527705, + "step": 6966 + }, + { + "epoch": 0.8377322190825467, + "flos": 21588250843680.0, + "grad_norm": 4.008887907010691, + "learning_rate": 2.6976965300347074e-07, + "loss": 0.7018, + "num_input_tokens_seen": 148550035, + "step": 6967 + }, + { + "epoch": 0.8378524619731859, + "flos": 19563268394520.0, + "grad_norm": 4.958653971244916, + "learning_rate": 2.693790729009309e-07, + "loss": 0.6711, + "num_input_tokens_seen": 148571365, + "step": 6968 + }, + { + "epoch": 0.8379727048638249, + "flos": 15135346452840.0, + "grad_norm": 4.775896253906294, + "learning_rate": 2.6898875533220946e-07, + "loss": 0.8607, + "num_input_tokens_seen": 148590390, + "step": 6969 + }, + { + "epoch": 0.838092947754464, + "flos": 14676733305480.0, + "grad_norm": 3.5715647698726545, + "learning_rate": 2.685987003565171e-07, + "loss": 0.808, + "num_input_tokens_seen": 148608150, + "step": 6970 + }, + { + "epoch": 0.8382131906451031, + "flos": 13218116542080.0, + "grad_norm": 7.199182718303755, + "learning_rate": 2.6820890803302566e-07, + "loss": 0.7325, + "num_input_tokens_seen": 148623395, + "step": 6971 + }, + { + "epoch": 0.8383334335357422, + "flos": 12462930634440.0, + "grad_norm": 9.071605659189549, + "learning_rate": 2.6781937842086557e-07, + "loss": 0.8043, + "num_input_tokens_seen": 148641905, + "step": 6972 + }, + { + "epoch": 0.8384536764263812, + "flos": 15136106291880.0, + "grad_norm": 6.911625629883049, + "learning_rate": 2.6743011157912933e-07, + "loss": 0.6696, + "num_input_tokens_seen": 148661345, + "step": 6973 + }, + { + "epoch": 0.8385739193170204, + "flos": 21265404807960.0, + "grad_norm": 2.71224416414052, + "learning_rate": 2.6704110756686725e-07, + "loss": 0.629, + "num_input_tokens_seen": 148681890, + "step": 6974 + }, + { + "epoch": 0.8386941622076595, + "flos": 17158524284280.0, + "grad_norm": 5.049081042685922, + "learning_rate": 2.6665236644309085e-07, + "loss": 0.8254, + "num_input_tokens_seen": 148701920, + "step": 6975 + }, + { + "epoch": 0.8388144050982985, + "flos": 16836691367280.0, + "grad_norm": 3.507127452813182, + "learning_rate": 2.662638882667727e-07, + "loss": 0.7751, + "num_input_tokens_seen": 148720580, + "step": 6976 + }, + { + "epoch": 0.8389346479889377, + "flos": 17782312154760.0, + "grad_norm": 4.0868376402278335, + "learning_rate": 2.658756730968443e-07, + "loss": 0.7154, + "num_input_tokens_seen": 148738765, + "step": 6977 + }, + { + "epoch": 0.8390548908795767, + "flos": 15513683415720.0, + "grad_norm": 3.8811739067935864, + "learning_rate": 2.654877209921975e-07, + "loss": 0.8769, + "num_input_tokens_seen": 148756020, + "step": 6978 + }, + { + "epoch": 0.8391751337702158, + "flos": 26179796170440.0, + "grad_norm": 6.01410257470329, + "learning_rate": 2.651000320116843e-07, + "loss": 0.6138, + "num_input_tokens_seen": 148776625, + "step": 6979 + }, + { + "epoch": 0.839295376660855, + "flos": 15595352638440.0, + "grad_norm": 3.5852198994336346, + "learning_rate": 2.647126062141163e-07, + "loss": 0.7339, + "num_input_tokens_seen": 148795420, + "step": 6980 + }, + { + "epoch": 0.839415619551494, + "flos": 13461921131760.0, + "grad_norm": 3.676330719923856, + "learning_rate": 2.643254436582669e-07, + "loss": 0.8206, + "num_input_tokens_seen": 148814630, + "step": 6981 + }, + { + "epoch": 0.8395358624421331, + "flos": 16998985034040.0, + "grad_norm": 6.105565905636936, + "learning_rate": 2.6393854440286743e-07, + "loss": 0.7985, + "num_input_tokens_seen": 148833520, + "step": 6982 + }, + { + "epoch": 0.8396561053327722, + "flos": 17856129707400.0, + "grad_norm": 4.404844235665833, + "learning_rate": 2.6355190850661045e-07, + "loss": 0.693, + "num_input_tokens_seen": 148850075, + "step": 6983 + }, + { + "epoch": 0.8397763482234113, + "flos": 16270737261000.0, + "grad_norm": 3.4425271467968273, + "learning_rate": 2.631655360281486e-07, + "loss": 0.8485, + "num_input_tokens_seen": 148869470, + "step": 6984 + }, + { + "epoch": 0.8398965911140504, + "flos": 16215816213120.0, + "grad_norm": 3.827944879770799, + "learning_rate": 2.6277942702609323e-07, + "loss": 0.6468, + "num_input_tokens_seen": 148888670, + "step": 6985 + }, + { + "epoch": 0.8400168340046895, + "flos": 15756158287080.0, + "grad_norm": 5.056646471546929, + "learning_rate": 2.623935815590186e-07, + "loss": 0.8543, + "num_input_tokens_seen": 148906770, + "step": 6986 + }, + { + "epoch": 0.8401370768953286, + "flos": 16458702663960.0, + "grad_norm": 3.695658435718666, + "learning_rate": 2.6200799968545516e-07, + "loss": 0.7901, + "num_input_tokens_seen": 148926785, + "step": 6987 + }, + { + "epoch": 0.8402573197859676, + "flos": 43659872923440.0, + "grad_norm": 1.339952959331002, + "learning_rate": 2.616226814638969e-07, + "loss": 0.586, + "num_input_tokens_seen": 148991610, + "step": 6988 + }, + { + "epoch": 0.8403775626766068, + "flos": 16593899896320.0, + "grad_norm": 16.397557220894214, + "learning_rate": 2.612376269527954e-07, + "loss": 0.749, + "num_input_tokens_seen": 149011035, + "step": 6989 + }, + { + "epoch": 0.8404978055672458, + "flos": 14325651076800.0, + "grad_norm": 2.864679804246781, + "learning_rate": 2.608528362105635e-07, + "loss": 0.6561, + "num_input_tokens_seen": 149030125, + "step": 6990 + }, + { + "epoch": 0.8406180484578849, + "flos": 20184903387720.0, + "grad_norm": 3.3084320425174303, + "learning_rate": 2.6046830929557374e-07, + "loss": 0.7131, + "num_input_tokens_seen": 149049495, + "step": 6991 + }, + { + "epoch": 0.8407382913485241, + "flos": 16189004718360.0, + "grad_norm": 5.938166209540527, + "learning_rate": 2.6008404626615776e-07, + "loss": 0.8213, + "num_input_tokens_seen": 149067715, + "step": 6992 + }, + { + "epoch": 0.8408585342391631, + "flos": 10113835751160.0, + "grad_norm": 4.758424816599448, + "learning_rate": 2.597000471806092e-07, + "loss": 0.7167, + "num_input_tokens_seen": 149084000, + "step": 6993 + }, + { + "epoch": 0.8409787771298022, + "flos": 14752735395360.0, + "grad_norm": 4.61738330406521, + "learning_rate": 2.593163120971793e-07, + "loss": 0.7098, + "num_input_tokens_seen": 149102585, + "step": 6994 + }, + { + "epoch": 0.8410990200204413, + "flos": 16939283332200.0, + "grad_norm": 3.481559950558234, + "learning_rate": 2.5893284107408165e-07, + "loss": 0.674, + "num_input_tokens_seen": 149119675, + "step": 6995 + }, + { + "epoch": 0.8412192629110804, + "flos": 17595296550600.0, + "grad_norm": 4.030385242535392, + "learning_rate": 2.5854963416948726e-07, + "loss": 0.7703, + "num_input_tokens_seen": 149141660, + "step": 6996 + }, + { + "epoch": 0.8413395058017195, + "flos": 18752496579840.0, + "grad_norm": 3.776605827894231, + "learning_rate": 2.5816669144152816e-07, + "loss": 0.6751, + "num_input_tokens_seen": 149162560, + "step": 6997 + }, + { + "epoch": 0.8414597486923585, + "flos": 46910997320640.0, + "grad_norm": 0.9142953299012502, + "learning_rate": 2.5778401294829777e-07, + "loss": 0.7239, + "num_input_tokens_seen": 149221020, + "step": 6998 + }, + { + "epoch": 0.8415799915829977, + "flos": 13947472413720.0, + "grad_norm": 4.711728967381353, + "learning_rate": 2.574015987478473e-07, + "loss": 0.6416, + "num_input_tokens_seen": 149238870, + "step": 6999 + }, + { + "epoch": 0.8417002344736367, + "flos": 14481486111720.0, + "grad_norm": 4.512335295062134, + "learning_rate": 2.570194488981887e-07, + "loss": 0.8402, + "num_input_tokens_seen": 149255135, + "step": 7000 + }, + { + "epoch": 0.8418204773642758, + "flos": 45823345240800.0, + "grad_norm": 0.8579633041706421, + "learning_rate": 2.566375634572939e-07, + "loss": 0.628, + "num_input_tokens_seen": 149315495, + "step": 7001 + }, + { + "epoch": 0.841940720254915, + "flos": 12463152254160.0, + "grad_norm": 8.159616294515967, + "learning_rate": 2.562559424830943e-07, + "loss": 0.7291, + "num_input_tokens_seen": 149333175, + "step": 7002 + }, + { + "epoch": 0.842060963145554, + "flos": 11863263162120.0, + "grad_norm": 3.906131101706114, + "learning_rate": 2.5587458603348256e-07, + "loss": 0.6743, + "num_input_tokens_seen": 149350185, + "step": 7003 + }, + { + "epoch": 0.8421812060361931, + "flos": 16026711051600.0, + "grad_norm": 5.313299679874944, + "learning_rate": 2.554934941663085e-07, + "loss": 0.83, + "num_input_tokens_seen": 149367440, + "step": 7004 + }, + { + "epoch": 0.8423014489268322, + "flos": 20369829434520.0, + "grad_norm": 5.661670090108441, + "learning_rate": 2.5511266693938484e-07, + "loss": 0.707, + "num_input_tokens_seen": 149385620, + "step": 7005 + }, + { + "epoch": 0.8424216918174713, + "flos": 18401667630840.0, + "grad_norm": 3.1612635913344027, + "learning_rate": 2.547321044104822e-07, + "loss": 0.7575, + "num_input_tokens_seen": 149406835, + "step": 7006 + }, + { + "epoch": 0.8425419347081103, + "flos": 18127853890440.0, + "grad_norm": 3.55998087374232, + "learning_rate": 2.5435180663733113e-07, + "loss": 0.7515, + "num_input_tokens_seen": 149426855, + "step": 7007 + }, + { + "epoch": 0.8426621775987495, + "flos": 18181413560040.0, + "grad_norm": 5.735313899954026, + "learning_rate": 2.539717736776241e-07, + "loss": 0.6926, + "num_input_tokens_seen": 149442800, + "step": 7008 + }, + { + "epoch": 0.8427824204893886, + "flos": 17241776505000.0, + "grad_norm": 2.929274222152459, + "learning_rate": 2.535920055890097e-07, + "loss": 0.7485, + "num_input_tokens_seen": 149463815, + "step": 7009 + }, + { + "epoch": 0.8429026633800276, + "flos": 11760734517120.0, + "grad_norm": 4.997450560057571, + "learning_rate": 2.5321250242910006e-07, + "loss": 0.6325, + "num_input_tokens_seen": 149481450, + "step": 7010 + }, + { + "epoch": 0.8430229062706668, + "flos": 16241329649520.0, + "grad_norm": 4.315256536997834, + "learning_rate": 2.5283326425546493e-07, + "loss": 0.8468, + "num_input_tokens_seen": 149500280, + "step": 7011 + }, + { + "epoch": 0.8431431491613058, + "flos": 25959288819960.0, + "grad_norm": 5.306442217720812, + "learning_rate": 2.5245429112563443e-07, + "loss": 0.6738, + "num_input_tokens_seen": 149520675, + "step": 7012 + }, + { + "epoch": 0.8432633920519449, + "flos": 18916056645000.0, + "grad_norm": 4.599335473200878, + "learning_rate": 2.5207558309709865e-07, + "loss": 0.8109, + "num_input_tokens_seen": 149540130, + "step": 7013 + }, + { + "epoch": 0.8433836349425841, + "flos": 47892452691480.0, + "grad_norm": 0.68413404638637, + "learning_rate": 2.516971402273065e-07, + "loss": 0.5737, + "num_input_tokens_seen": 149605915, + "step": 7014 + }, + { + "epoch": 0.8435038778332231, + "flos": 14784707463600.0, + "grad_norm": 4.601236069504538, + "learning_rate": 2.513189625736687e-07, + "loss": 0.6547, + "num_input_tokens_seen": 149622530, + "step": 7015 + }, + { + "epoch": 0.8436241207238622, + "flos": 15348255412920.0, + "grad_norm": 7.12534853434131, + "learning_rate": 2.509410501935534e-07, + "loss": 0.6996, + "num_input_tokens_seen": 149637885, + "step": 7016 + }, + { + "epoch": 0.8437443636145013, + "flos": 10678745078760.0, + "grad_norm": 4.86349812832178, + "learning_rate": 2.5056340314429116e-07, + "loss": 0.7309, + "num_input_tokens_seen": 149655070, + "step": 7017 + }, + { + "epoch": 0.8438646065051404, + "flos": 15804050823840.0, + "grad_norm": 5.076651491445969, + "learning_rate": 2.5018602148316904e-07, + "loss": 0.7772, + "num_input_tokens_seen": 149670825, + "step": 7018 + }, + { + "epoch": 0.8439848493957794, + "flos": 17048587208640.0, + "grad_norm": 3.2218768292781697, + "learning_rate": 2.498089052674359e-07, + "loss": 0.7808, + "num_input_tokens_seen": 149688520, + "step": 7019 + }, + { + "epoch": 0.8441050922864186, + "flos": 14406813740160.0, + "grad_norm": 5.0820507662000365, + "learning_rate": 2.494320545543007e-07, + "loss": 0.7368, + "num_input_tokens_seen": 149707810, + "step": 7020 + }, + { + "epoch": 0.8442253351770577, + "flos": 15972708142560.0, + "grad_norm": 3.8662869108000164, + "learning_rate": 2.490554694009308e-07, + "loss": 0.6517, + "num_input_tokens_seen": 149728395, + "step": 7021 + }, + { + "epoch": 0.8443455780676967, + "flos": 25231262666640.0, + "grad_norm": 3.3666877110141433, + "learning_rate": 2.4867914986445426e-07, + "loss": 0.7838, + "num_input_tokens_seen": 149750505, + "step": 7022 + }, + { + "epoch": 0.8444658209583359, + "flos": 35495202999960.0, + "grad_norm": 3.856731420603642, + "learning_rate": 2.483030960019581e-07, + "loss": 0.6783, + "num_input_tokens_seen": 149774155, + "step": 7023 + }, + { + "epoch": 0.8445860638489749, + "flos": 50499404695320.0, + "grad_norm": 0.7591323370626859, + "learning_rate": 2.479273078704891e-07, + "loss": 0.5598, + "num_input_tokens_seen": 149827240, + "step": 7024 + }, + { + "epoch": 0.844706306739614, + "flos": 46317503540520.0, + "grad_norm": 0.7921048800330278, + "learning_rate": 2.475517855270552e-07, + "loss": 0.658, + "num_input_tokens_seen": 149887040, + "step": 7025 + }, + { + "epoch": 0.8448265496302532, + "flos": 10896498012720.0, + "grad_norm": 4.416163121914809, + "learning_rate": 2.4717652902862143e-07, + "loss": 0.7164, + "num_input_tokens_seen": 149905735, + "step": 7026 + }, + { + "epoch": 0.8449467925208922, + "flos": 17161342020720.0, + "grad_norm": 2.756963755948077, + "learning_rate": 2.4680153843211495e-07, + "loss": 0.8073, + "num_input_tokens_seen": 149925385, + "step": 7027 + }, + { + "epoch": 0.8450670354115313, + "flos": 16648314384840.0, + "grad_norm": 6.283343673144399, + "learning_rate": 2.464268137944212e-07, + "loss": 0.7071, + "num_input_tokens_seen": 149946400, + "step": 7028 + }, + { + "epoch": 0.8451872783021703, + "flos": 21886058342400.0, + "grad_norm": 4.310315670835603, + "learning_rate": 2.46052355172385e-07, + "loss": 0.7698, + "num_input_tokens_seen": 149964160, + "step": 7029 + }, + { + "epoch": 0.8453075211928095, + "flos": 15998189919000.0, + "grad_norm": 9.96757361666875, + "learning_rate": 2.456781626228128e-07, + "loss": 0.7207, + "num_input_tokens_seen": 149983385, + "step": 7030 + }, + { + "epoch": 0.8454277640834486, + "flos": 43297646388840.0, + "grad_norm": 1.0557208050379618, + "learning_rate": 2.453042362024675e-07, + "loss": 0.7369, + "num_input_tokens_seen": 150036350, + "step": 7031 + }, + { + "epoch": 0.8455480069740876, + "flos": 19861740752400.0, + "grad_norm": 3.206809404438699, + "learning_rate": 2.449305759680751e-07, + "loss": 0.7222, + "num_input_tokens_seen": 150057395, + "step": 7032 + }, + { + "epoch": 0.8456682498647268, + "flos": 19941985276920.0, + "grad_norm": 3.938611675946097, + "learning_rate": 2.445571819763188e-07, + "loss": 0.7317, + "num_input_tokens_seen": 150079415, + "step": 7033 + }, + { + "epoch": 0.8457884927553658, + "flos": 15082230022680.0, + "grad_norm": 3.6941785947775054, + "learning_rate": 2.4418405428384227e-07, + "loss": 0.575, + "num_input_tokens_seen": 150099345, + "step": 7034 + }, + { + "epoch": 0.8459087356460049, + "flos": 11134698789480.0, + "grad_norm": 12.34206129667088, + "learning_rate": 2.4381119294724864e-07, + "loss": 0.7085, + "num_input_tokens_seen": 150116510, + "step": 7035 + }, + { + "epoch": 0.846028978536644, + "flos": 13758303932280.0, + "grad_norm": 10.573296621226026, + "learning_rate": 2.434385980231004e-07, + "loss": 0.5243, + "num_input_tokens_seen": 150135070, + "step": 7036 + }, + { + "epoch": 0.8461492214272831, + "flos": 38785802875920.0, + "grad_norm": 4.014792270630992, + "learning_rate": 2.4306626956792043e-07, + "loss": 0.6395, + "num_input_tokens_seen": 150159735, + "step": 7037 + }, + { + "epoch": 0.8462694643179222, + "flos": 13569863629920.0, + "grad_norm": 3.871725113671548, + "learning_rate": 2.4269420763819017e-07, + "loss": 0.739, + "num_input_tokens_seen": 150177500, + "step": 7038 + }, + { + "epoch": 0.8463897072085613, + "flos": 17860910361360.0, + "grad_norm": 4.789132595327154, + "learning_rate": 2.4232241229035223e-07, + "loss": 0.8133, + "num_input_tokens_seen": 150194975, + "step": 7039 + }, + { + "epoch": 0.8465099500992004, + "flos": 50660052044160.0, + "grad_norm": 0.8229581041808127, + "learning_rate": 2.419508835808064e-07, + "loss": 0.5858, + "num_input_tokens_seen": 150251250, + "step": 7040 + }, + { + "epoch": 0.8466301929898394, + "flos": 9978543538920.0, + "grad_norm": 4.106600328710553, + "learning_rate": 2.415796215659134e-07, + "loss": 0.6033, + "num_input_tokens_seen": 150267675, + "step": 7041 + }, + { + "epoch": 0.8467504358804786, + "flos": 14052470535600.0, + "grad_norm": 7.378318686489782, + "learning_rate": 2.412086263019939e-07, + "loss": 0.7585, + "num_input_tokens_seen": 150285420, + "step": 7042 + }, + { + "epoch": 0.8468706787711177, + "flos": 15594371179680.0, + "grad_norm": 3.0928728806143404, + "learning_rate": 2.408378978453276e-07, + "loss": 0.7934, + "num_input_tokens_seen": 150305260, + "step": 7043 + }, + { + "epoch": 0.8469909216617567, + "flos": 47830408152600.0, + "grad_norm": 0.8453536864451839, + "learning_rate": 2.404674362521533e-07, + "loss": 0.6533, + "num_input_tokens_seen": 150363475, + "step": 7044 + }, + { + "epoch": 0.8471111645523959, + "flos": 14081973126960.0, + "grad_norm": 4.789776349032894, + "learning_rate": 2.4009724157866997e-07, + "loss": 0.7281, + "num_input_tokens_seen": 150380255, + "step": 7045 + }, + { + "epoch": 0.8472314074430349, + "flos": 16106069097240.0, + "grad_norm": 5.4607095957339125, + "learning_rate": 2.3972731388103564e-07, + "loss": 0.7529, + "num_input_tokens_seen": 150398455, + "step": 7046 + }, + { + "epoch": 0.847351650333674, + "flos": 42655215293280.0, + "grad_norm": 0.8076143460759984, + "learning_rate": 2.393576532153687e-07, + "loss": 0.6406, + "num_input_tokens_seen": 150461960, + "step": 7047 + }, + { + "epoch": 0.8474718932243132, + "flos": 30371706363960.0, + "grad_norm": 0.9587492178624033, + "learning_rate": 2.389882596377453e-07, + "loss": 0.6167, + "num_input_tokens_seen": 150515945, + "step": 7048 + }, + { + "epoch": 0.8475921361149522, + "flos": 28069581387240.0, + "grad_norm": 3.939417296564888, + "learning_rate": 2.386191332042031e-07, + "loss": 0.7506, + "num_input_tokens_seen": 150537560, + "step": 7049 + }, + { + "epoch": 0.8477123790055913, + "flos": 18592482430200.0, + "grad_norm": 3.1964667364400916, + "learning_rate": 2.3825027397073794e-07, + "loss": 0.7161, + "num_input_tokens_seen": 150557755, + "step": 7050 + }, + { + "epoch": 0.8478326218962304, + "flos": 22181554664040.0, + "grad_norm": 5.919179499187663, + "learning_rate": 2.3788168199330515e-07, + "loss": 0.6619, + "num_input_tokens_seen": 150579035, + "step": 7051 + }, + { + "epoch": 0.8479528647868695, + "flos": 28093416845760.0, + "grad_norm": 5.100583974278054, + "learning_rate": 2.3751335732782074e-07, + "loss": 0.7268, + "num_input_tokens_seen": 150600015, + "step": 7052 + }, + { + "epoch": 0.8480731076775085, + "flos": 15323058576120.0, + "grad_norm": 3.924333143987651, + "learning_rate": 2.371453000301582e-07, + "loss": 0.7867, + "num_input_tokens_seen": 150618420, + "step": 7053 + }, + { + "epoch": 0.8481933505681477, + "flos": 23936459248080.0, + "grad_norm": 3.0113270228991693, + "learning_rate": 2.3677751015615222e-07, + "loss": 0.7297, + "num_input_tokens_seen": 150640215, + "step": 7054 + }, + { + "epoch": 0.8483135934587868, + "flos": 15163076086440.0, + "grad_norm": 7.024342581508434, + "learning_rate": 2.3640998776159593e-07, + "loss": 0.8396, + "num_input_tokens_seen": 150657440, + "step": 7055 + }, + { + "epoch": 0.8484338363494258, + "flos": 15837890829720.0, + "grad_norm": 3.409896159003313, + "learning_rate": 2.3604273290224253e-07, + "loss": 0.798, + "num_input_tokens_seen": 150677875, + "step": 7056 + }, + { + "epoch": 0.848554079240065, + "flos": 10924702545720.0, + "grad_norm": 3.3620211802762303, + "learning_rate": 2.356757456338039e-07, + "loss": 0.7268, + "num_input_tokens_seen": 150695080, + "step": 7057 + }, + { + "epoch": 0.848674322130704, + "flos": 50187291386040.0, + "grad_norm": 0.828473359738088, + "learning_rate": 2.3530902601195147e-07, + "loss": 0.6428, + "num_input_tokens_seen": 150763290, + "step": 7058 + }, + { + "epoch": 0.8487945650213431, + "flos": 13486073189880.0, + "grad_norm": 3.4483132648083785, + "learning_rate": 2.34942574092317e-07, + "loss": 0.7708, + "num_input_tokens_seen": 150778260, + "step": 7059 + }, + { + "epoch": 0.8489148079119821, + "flos": 17184322660320.0, + "grad_norm": 6.422398009895507, + "learning_rate": 2.3457638993049045e-07, + "loss": 0.7416, + "num_input_tokens_seen": 150795970, + "step": 7060 + }, + { + "epoch": 0.8490350508026213, + "flos": 14568252588000.0, + "grad_norm": 4.349587449428349, + "learning_rate": 2.3421047358202252e-07, + "loss": 0.6213, + "num_input_tokens_seen": 150814540, + "step": 7061 + }, + { + "epoch": 0.8491552936932604, + "flos": 17781963895200.0, + "grad_norm": 19.265503820330572, + "learning_rate": 2.3384482510242144e-07, + "loss": 0.8006, + "num_input_tokens_seen": 150832120, + "step": 7062 + }, + { + "epoch": 0.8492755365838994, + "flos": 16480480225080.0, + "grad_norm": 3.068247326238265, + "learning_rate": 2.3347944454715575e-07, + "loss": 0.7578, + "num_input_tokens_seen": 150848230, + "step": 7063 + }, + { + "epoch": 0.8493957794745386, + "flos": 19780578089040.0, + "grad_norm": 3.5560706497314345, + "learning_rate": 2.331143319716542e-07, + "loss": 0.6463, + "num_input_tokens_seen": 150867480, + "step": 7064 + }, + { + "epoch": 0.8495160223651776, + "flos": 21912996477000.0, + "grad_norm": 7.110420435239336, + "learning_rate": 2.3274948743130363e-07, + "loss": 0.6369, + "num_input_tokens_seen": 150887035, + "step": 7065 + }, + { + "epoch": 0.8496362652558167, + "flos": 16890472656600.0, + "grad_norm": 4.342716011643875, + "learning_rate": 2.3238491098145085e-07, + "loss": 0.7866, + "num_input_tokens_seen": 150906285, + "step": 7066 + }, + { + "epoch": 0.8497565081464559, + "flos": 10624583869920.0, + "grad_norm": 6.515430035864562, + "learning_rate": 2.3202060267740141e-07, + "loss": 0.7018, + "num_input_tokens_seen": 150923530, + "step": 7067 + }, + { + "epoch": 0.8498767510370949, + "flos": 15454741552920.0, + "grad_norm": 3.4478611936841492, + "learning_rate": 2.3165656257442044e-07, + "loss": 0.7548, + "num_input_tokens_seen": 150941770, + "step": 7068 + }, + { + "epoch": 0.849996993927734, + "flos": 17318380134120.0, + "grad_norm": 8.655931818103031, + "learning_rate": 2.31292790727734e-07, + "loss": 0.8897, + "num_input_tokens_seen": 150959055, + "step": 7069 + }, + { + "epoch": 0.8501172368183731, + "flos": 15027467274600.0, + "grad_norm": 3.983842402745996, + "learning_rate": 2.3092928719252392e-07, + "loss": 0.7841, + "num_input_tokens_seen": 150977175, + "step": 7070 + }, + { + "epoch": 0.8502374797090122, + "flos": 16296314017320.0, + "grad_norm": 5.836246344425693, + "learning_rate": 2.3056605202393475e-07, + "loss": 0.7636, + "num_input_tokens_seen": 150994455, + "step": 7071 + }, + { + "epoch": 0.8503577225996513, + "flos": 17323034148240.0, + "grad_norm": 6.828968266632224, + "learning_rate": 2.3020308527706888e-07, + "loss": 0.6598, + "num_input_tokens_seen": 151013590, + "step": 7072 + }, + { + "epoch": 0.8504779654902904, + "flos": 19615909925280.0, + "grad_norm": 3.1445638524610358, + "learning_rate": 2.2984038700698715e-07, + "loss": 0.8684, + "num_input_tokens_seen": 151032620, + "step": 7073 + }, + { + "epoch": 0.8505982083809295, + "flos": 19401322987320.0, + "grad_norm": 5.746379873697181, + "learning_rate": 2.2947795726871222e-07, + "loss": 0.7805, + "num_input_tokens_seen": 151053365, + "step": 7074 + }, + { + "epoch": 0.8507184512715685, + "flos": 14839090292160.0, + "grad_norm": 4.441919194832662, + "learning_rate": 2.2911579611722253e-07, + "loss": 0.8429, + "num_input_tokens_seen": 151072230, + "step": 7075 + }, + { + "epoch": 0.8508386941622077, + "flos": 13892962945320.0, + "grad_norm": 5.8338460787717725, + "learning_rate": 2.2875390360745905e-07, + "loss": 0.8586, + "num_input_tokens_seen": 151091355, + "step": 7076 + }, + { + "epoch": 0.8509589370528468, + "flos": 11974688255880.0, + "grad_norm": 3.6740330624242645, + "learning_rate": 2.2839227979432008e-07, + "loss": 0.7638, + "num_input_tokens_seen": 151108725, + "step": 7077 + }, + { + "epoch": 0.8510791799434858, + "flos": 13326565599600.0, + "grad_norm": 6.596552104266792, + "learning_rate": 2.2803092473266373e-07, + "loss": 0.833, + "num_input_tokens_seen": 151125970, + "step": 7078 + }, + { + "epoch": 0.851199422834125, + "flos": 17161437000600.0, + "grad_norm": 5.536206282449296, + "learning_rate": 2.2766983847730724e-07, + "loss": 0.8399, + "num_input_tokens_seen": 151145360, + "step": 7079 + }, + { + "epoch": 0.851319665724764, + "flos": 11868392075640.0, + "grad_norm": 4.315301178889646, + "learning_rate": 2.2730902108302663e-07, + "loss": 0.6518, + "num_input_tokens_seen": 151161995, + "step": 7080 + }, + { + "epoch": 0.8514399086154031, + "flos": 13866531370080.0, + "grad_norm": 4.494462314649299, + "learning_rate": 2.269484726045583e-07, + "loss": 0.6778, + "num_input_tokens_seen": 151180630, + "step": 7081 + }, + { + "epoch": 0.8515601515060423, + "flos": 17997405652080.0, + "grad_norm": 4.388701428193775, + "learning_rate": 2.2658819309659672e-07, + "loss": 0.7767, + "num_input_tokens_seen": 151200550, + "step": 7082 + }, + { + "epoch": 0.8516803943966813, + "flos": 14265221195880.0, + "grad_norm": 3.2905107875033295, + "learning_rate": 2.2622818261379706e-07, + "loss": 0.8336, + "num_input_tokens_seen": 151217290, + "step": 7083 + }, + { + "epoch": 0.8518006372873204, + "flos": 14811107378880.0, + "grad_norm": 4.390823305477773, + "learning_rate": 2.2586844121077142e-07, + "loss": 0.7451, + "num_input_tokens_seen": 151235520, + "step": 7084 + }, + { + "epoch": 0.8519208801779595, + "flos": 17673261558000.0, + "grad_norm": 3.6798912040550387, + "learning_rate": 2.2550896894209215e-07, + "loss": 0.7023, + "num_input_tokens_seen": 151254755, + "step": 7085 + }, + { + "epoch": 0.8520411230685986, + "flos": 46469670511440.0, + "grad_norm": 0.6922990885530054, + "learning_rate": 2.2514976586229184e-07, + "loss": 0.5818, + "num_input_tokens_seen": 151322420, + "step": 7086 + }, + { + "epoch": 0.8521613659592376, + "flos": 48541975618080.0, + "grad_norm": 0.9550055727870979, + "learning_rate": 2.247908320258609e-07, + "loss": 0.6181, + "num_input_tokens_seen": 151382230, + "step": 7087 + }, + { + "epoch": 0.8522816088498768, + "flos": 16917822370680.0, + "grad_norm": 5.053277551616562, + "learning_rate": 2.2443216748724914e-07, + "loss": 0.7675, + "num_input_tokens_seen": 151402660, + "step": 7088 + }, + { + "epoch": 0.8524018517405159, + "flos": 23316122313240.0, + "grad_norm": 4.512987046097858, + "learning_rate": 2.2407377230086588e-07, + "loss": 0.7309, + "num_input_tokens_seen": 151424735, + "step": 7089 + }, + { + "epoch": 0.8525220946311549, + "flos": 13644219401880.0, + "grad_norm": 5.654775859461801, + "learning_rate": 2.23715646521079e-07, + "loss": 0.8233, + "num_input_tokens_seen": 151441975, + "step": 7090 + }, + { + "epoch": 0.852642337521794, + "flos": 15940799394240.0, + "grad_norm": 6.969903003778466, + "learning_rate": 2.2335779020221724e-07, + "loss": 0.8229, + "num_input_tokens_seen": 151458315, + "step": 7091 + }, + { + "epoch": 0.8527625804124331, + "flos": 50911550004120.0, + "grad_norm": 2.8251164066333714, + "learning_rate": 2.2300020339856497e-07, + "loss": 0.6154, + "num_input_tokens_seen": 151520720, + "step": 7092 + }, + { + "epoch": 0.8528828233030722, + "flos": 19779659950200.0, + "grad_norm": 5.751117100215455, + "learning_rate": 2.2264288616436966e-07, + "loss": 0.7627, + "num_input_tokens_seen": 151540695, + "step": 7093 + }, + { + "epoch": 0.8530030661937112, + "flos": 12754311161280.0, + "grad_norm": 8.319848046749177, + "learning_rate": 2.222858385538351e-07, + "loss": 0.7247, + "num_input_tokens_seen": 151557215, + "step": 7094 + }, + { + "epoch": 0.8531233090843504, + "flos": 16213220096400.0, + "grad_norm": 3.1236325132968186, + "learning_rate": 2.2192906062112527e-07, + "loss": 0.6605, + "num_input_tokens_seen": 151576810, + "step": 7095 + }, + { + "epoch": 0.8532435519749895, + "flos": 27666110907480.0, + "grad_norm": 2.854440068053946, + "learning_rate": 2.2157255242036377e-07, + "loss": 0.69, + "num_input_tokens_seen": 151600195, + "step": 7096 + }, + { + "epoch": 0.8533637948656285, + "flos": 15649323887520.0, + "grad_norm": 4.50705420398471, + "learning_rate": 2.2121631400563135e-07, + "loss": 0.7236, + "num_input_tokens_seen": 151619745, + "step": 7097 + }, + { + "epoch": 0.8534840377562677, + "flos": 39295320747600.0, + "grad_norm": 0.8772111950518727, + "learning_rate": 2.208603454309701e-07, + "loss": 0.5972, + "num_input_tokens_seen": 151677555, + "step": 7098 + }, + { + "epoch": 0.8536042806469067, + "flos": 15217078995480.0, + "grad_norm": 4.405735066944282, + "learning_rate": 2.2050464675037994e-07, + "loss": 0.6885, + "num_input_tokens_seen": 151695900, + "step": 7099 + }, + { + "epoch": 0.8537245235375458, + "flos": 18079043214840.0, + "grad_norm": 4.650320284077951, + "learning_rate": 2.2014921801782016e-07, + "loss": 0.7109, + "num_input_tokens_seen": 151715110, + "step": 7100 + }, + { + "epoch": 0.853844766428185, + "flos": 17858915783880.0, + "grad_norm": 6.355449242728975, + "learning_rate": 2.1979405928720872e-07, + "loss": 0.7334, + "num_input_tokens_seen": 151734485, + "step": 7101 + }, + { + "epoch": 0.853965009318824, + "flos": 15321222298440.0, + "grad_norm": 2.8810217446298085, + "learning_rate": 2.1943917061242257e-07, + "loss": 0.7794, + "num_input_tokens_seen": 151754060, + "step": 7102 + }, + { + "epoch": 0.8540852522094631, + "flos": 17723401951920.0, + "grad_norm": 3.723501423724453, + "learning_rate": 2.1908455204729903e-07, + "loss": 0.6455, + "num_input_tokens_seen": 151772930, + "step": 7103 + }, + { + "epoch": 0.8542054951001022, + "flos": 18376249174320.0, + "grad_norm": 4.294702931123591, + "learning_rate": 2.1873020364563265e-07, + "loss": 0.763, + "num_input_tokens_seen": 151791715, + "step": 7104 + }, + { + "epoch": 0.8543257379907413, + "flos": 17809060329600.0, + "grad_norm": 6.014638844696239, + "learning_rate": 2.183761254611789e-07, + "loss": 0.7389, + "num_input_tokens_seen": 151811760, + "step": 7105 + }, + { + "epoch": 0.8544459808813804, + "flos": 40728102983640.0, + "grad_norm": 7.151205492099523, + "learning_rate": 2.1802231754764987e-07, + "loss": 0.6781, + "num_input_tokens_seen": 151836920, + "step": 7106 + }, + { + "epoch": 0.8545662237720195, + "flos": 18889371790080.0, + "grad_norm": 5.87906262769594, + "learning_rate": 2.17668779958718e-07, + "loss": 0.7494, + "num_input_tokens_seen": 151859220, + "step": 7107 + }, + { + "epoch": 0.8546864666626586, + "flos": 8034027234000.0, + "grad_norm": 9.459866930164809, + "learning_rate": 2.1731551274801553e-07, + "loss": 0.7732, + "num_input_tokens_seen": 151875380, + "step": 7108 + }, + { + "epoch": 0.8548067095532976, + "flos": 18699760069200.0, + "grad_norm": 5.149049482311183, + "learning_rate": 2.169625159691324e-07, + "loss": 0.5859, + "num_input_tokens_seen": 151894975, + "step": 7109 + }, + { + "epoch": 0.8549269524439368, + "flos": 18106044669360.0, + "grad_norm": 4.713682776037262, + "learning_rate": 2.1660978967561784e-07, + "loss": 0.7238, + "num_input_tokens_seen": 151914030, + "step": 7110 + }, + { + "epoch": 0.8550471953345758, + "flos": 14485000367280.0, + "grad_norm": 5.213176560804936, + "learning_rate": 2.1625733392098035e-07, + "loss": 0.777, + "num_input_tokens_seen": 151929360, + "step": 7111 + }, + { + "epoch": 0.8551674382252149, + "flos": 16701810734520.0, + "grad_norm": 3.1883038273941784, + "learning_rate": 2.159051487586867e-07, + "loss": 0.7848, + "num_input_tokens_seen": 151949210, + "step": 7112 + }, + { + "epoch": 0.8552876811158541, + "flos": 15081660143400.0, + "grad_norm": 5.75855514312625, + "learning_rate": 2.155532342421642e-07, + "loss": 0.716, + "num_input_tokens_seen": 151966930, + "step": 7113 + }, + { + "epoch": 0.8554079240064931, + "flos": 16917854030640.0, + "grad_norm": 3.405840507881206, + "learning_rate": 2.1520159042479636e-07, + "loss": 0.7733, + "num_input_tokens_seen": 151984940, + "step": 7114 + }, + { + "epoch": 0.8555281668971322, + "flos": 16189511277720.0, + "grad_norm": 4.573793829671151, + "learning_rate": 2.148502173599287e-07, + "loss": 0.6965, + "num_input_tokens_seen": 152002800, + "step": 7115 + }, + { + "epoch": 0.8556484097877713, + "flos": 22858110705120.0, + "grad_norm": 4.394853286811479, + "learning_rate": 2.1449911510086372e-07, + "loss": 0.6402, + "num_input_tokens_seen": 152021990, + "step": 7116 + }, + { + "epoch": 0.8557686526784104, + "flos": 17809060329600.0, + "grad_norm": 4.6927013785268015, + "learning_rate": 2.141482837008628e-07, + "loss": 0.7626, + "num_input_tokens_seen": 152042250, + "step": 7117 + }, + { + "epoch": 0.8558888955690495, + "flos": 12921797061480.0, + "grad_norm": 5.137416777667638, + "learning_rate": 2.1379772321314826e-07, + "loss": 0.7013, + "num_input_tokens_seen": 152060015, + "step": 7118 + }, + { + "epoch": 0.8560091384596886, + "flos": 14001475322760.0, + "grad_norm": 4.5953098684937, + "learning_rate": 2.1344743369089802e-07, + "loss": 0.8058, + "num_input_tokens_seen": 152075515, + "step": 7119 + }, + { + "epoch": 0.8561293813503277, + "flos": 17511727730280.0, + "grad_norm": 3.3782457356575293, + "learning_rate": 2.130974151872522e-07, + "loss": 0.8114, + "num_input_tokens_seen": 152095570, + "step": 7120 + }, + { + "epoch": 0.8562496242409667, + "flos": 16486368977640.0, + "grad_norm": 3.290318047823791, + "learning_rate": 2.1274766775530773e-07, + "loss": 0.7688, + "num_input_tokens_seen": 152115155, + "step": 7121 + }, + { + "epoch": 0.8563698671316058, + "flos": 10700459319960.0, + "grad_norm": 4.521708999887442, + "learning_rate": 2.1239819144812077e-07, + "loss": 0.7841, + "num_input_tokens_seen": 152129335, + "step": 7122 + }, + { + "epoch": 0.856490110022245, + "flos": 28800171997320.0, + "grad_norm": 3.256657750232351, + "learning_rate": 2.1204898631870716e-07, + "loss": 0.6875, + "num_input_tokens_seen": 152153945, + "step": 7123 + }, + { + "epoch": 0.856610352912884, + "flos": 21318996137520.0, + "grad_norm": 4.022941129120477, + "learning_rate": 2.1170005242004006e-07, + "loss": 0.7406, + "num_input_tokens_seen": 152175015, + "step": 7124 + }, + { + "epoch": 0.8567305958035231, + "flos": 17484694615800.0, + "grad_norm": 9.99546253258119, + "learning_rate": 2.1135138980505384e-07, + "loss": 0.7547, + "num_input_tokens_seen": 152195405, + "step": 7125 + }, + { + "epoch": 0.8568508386941622, + "flos": 16242722687760.0, + "grad_norm": 3.5042619641590935, + "learning_rate": 2.110029985266395e-07, + "loss": 0.7106, + "num_input_tokens_seen": 152214830, + "step": 7126 + }, + { + "epoch": 0.8569710815848013, + "flos": 12621361786080.0, + "grad_norm": 3.1312952339118305, + "learning_rate": 2.1065487863764787e-07, + "loss": 0.7285, + "num_input_tokens_seen": 152232895, + "step": 7127 + }, + { + "epoch": 0.8570913244754403, + "flos": 17376593817840.0, + "grad_norm": 2.976271887982731, + "learning_rate": 2.1030703019088846e-07, + "loss": 0.8461, + "num_input_tokens_seen": 152253245, + "step": 7128 + }, + { + "epoch": 0.8572115673660795, + "flos": 14650744969680.0, + "grad_norm": 3.7695770419012664, + "learning_rate": 2.099594532391291e-07, + "loss": 0.6865, + "num_input_tokens_seen": 152271650, + "step": 7129 + }, + { + "epoch": 0.8573318102567186, + "flos": 19913400824400.0, + "grad_norm": 4.015262572561268, + "learning_rate": 2.0961214783509806e-07, + "loss": 0.7767, + "num_input_tokens_seen": 152294250, + "step": 7130 + }, + { + "epoch": 0.8574520531473576, + "flos": 18267895096680.0, + "grad_norm": 6.4622906117855665, + "learning_rate": 2.0926511403148051e-07, + "loss": 0.7403, + "num_input_tokens_seen": 152312935, + "step": 7131 + }, + { + "epoch": 0.8575722960379968, + "flos": 13705155842160.0, + "grad_norm": 4.189150356920915, + "learning_rate": 2.0891835188092143e-07, + "loss": 0.7355, + "num_input_tokens_seen": 152329655, + "step": 7132 + }, + { + "epoch": 0.8576925389286358, + "flos": 16242880987560.0, + "grad_norm": 4.879362257575206, + "learning_rate": 2.0857186143602434e-07, + "loss": 0.796, + "num_input_tokens_seen": 152348020, + "step": 7133 + }, + { + "epoch": 0.8578127818192749, + "flos": 16756636802520.0, + "grad_norm": 3.4913003976737484, + "learning_rate": 2.0822564274935094e-07, + "loss": 0.6629, + "num_input_tokens_seen": 152367165, + "step": 7134 + }, + { + "epoch": 0.8579330247099141, + "flos": 25585954130760.0, + "grad_norm": 5.461580994420769, + "learning_rate": 2.078796958734239e-07, + "loss": 0.6573, + "num_input_tokens_seen": 152389605, + "step": 7135 + }, + { + "epoch": 0.8580532676005531, + "flos": 14433688554840.0, + "grad_norm": 13.475000475598613, + "learning_rate": 2.0753402086072124e-07, + "loss": 0.7337, + "num_input_tokens_seen": 152407955, + "step": 7136 + }, + { + "epoch": 0.8581735104911922, + "flos": 16640621014560.0, + "grad_norm": 5.2800405347597525, + "learning_rate": 2.071886177636828e-07, + "loss": 0.7447, + "num_input_tokens_seen": 152424460, + "step": 7137 + }, + { + "epoch": 0.8582937533818313, + "flos": 16945108764840.0, + "grad_norm": 3.563301531653048, + "learning_rate": 2.0684348663470575e-07, + "loss": 0.8089, + "num_input_tokens_seen": 152444360, + "step": 7138 + }, + { + "epoch": 0.8584139962724704, + "flos": 14243253675000.0, + "grad_norm": 3.467807126130011, + "learning_rate": 2.0649862752614555e-07, + "loss": 0.6027, + "num_input_tokens_seen": 152462790, + "step": 7139 + }, + { + "epoch": 0.8585342391631094, + "flos": 52568200037760.0, + "grad_norm": 0.7731090965886502, + "learning_rate": 2.0615404049031838e-07, + "loss": 0.5857, + "num_input_tokens_seen": 152519480, + "step": 7140 + }, + { + "epoch": 0.8586544820537486, + "flos": 7817129118960.0, + "grad_norm": 5.438598171141041, + "learning_rate": 2.0580972557949616e-07, + "loss": 0.763, + "num_input_tokens_seen": 152534290, + "step": 7141 + }, + { + "epoch": 0.8587747249443877, + "flos": 47782325656080.0, + "grad_norm": 1.1023303122346875, + "learning_rate": 2.054656828459125e-07, + "loss": 0.541, + "num_input_tokens_seen": 152598120, + "step": 7142 + }, + { + "epoch": 0.8588949678350267, + "flos": 19672667250840.0, + "grad_norm": 4.676895964791019, + "learning_rate": 2.051219123417578e-07, + "loss": 0.7478, + "num_input_tokens_seen": 152617900, + "step": 7143 + }, + { + "epoch": 0.8590152107256659, + "flos": 19132068281160.0, + "grad_norm": 4.920231426657705, + "learning_rate": 2.0477841411918196e-07, + "loss": 0.582, + "num_input_tokens_seen": 152637145, + "step": 7144 + }, + { + "epoch": 0.859135453616305, + "flos": 19159196375520.0, + "grad_norm": 4.842617763964169, + "learning_rate": 2.0443518823029326e-07, + "loss": 0.7105, + "num_input_tokens_seen": 152657405, + "step": 7145 + }, + { + "epoch": 0.859255696506944, + "flos": 9411323034240.0, + "grad_norm": 4.155422042179689, + "learning_rate": 2.0409223472715854e-07, + "loss": 0.7386, + "num_input_tokens_seen": 152674270, + "step": 7146 + }, + { + "epoch": 0.8593759393975832, + "flos": 13486041529920.0, + "grad_norm": 4.526616535911076, + "learning_rate": 2.0374955366180434e-07, + "loss": 0.7252, + "num_input_tokens_seen": 152691630, + "step": 7147 + }, + { + "epoch": 0.8594961822882222, + "flos": 16242754347720.0, + "grad_norm": 3.416399751487359, + "learning_rate": 2.034071450862147e-07, + "loss": 0.7071, + "num_input_tokens_seen": 152708820, + "step": 7148 + }, + { + "epoch": 0.8596164251788613, + "flos": 17052829643280.0, + "grad_norm": 6.656267479528839, + "learning_rate": 2.030650090523327e-07, + "loss": 0.7549, + "num_input_tokens_seen": 152727730, + "step": 7149 + }, + { + "epoch": 0.8597366680695004, + "flos": 23236194388320.0, + "grad_norm": 4.9275427745895986, + "learning_rate": 2.0272314561205995e-07, + "loss": 0.5928, + "num_input_tokens_seen": 152747845, + "step": 7150 + }, + { + "epoch": 0.8598569109601395, + "flos": 15567464705040.0, + "grad_norm": 4.765749269960179, + "learning_rate": 2.023815548172567e-07, + "loss": 0.7179, + "num_input_tokens_seen": 152767635, + "step": 7151 + }, + { + "epoch": 0.8599771538507786, + "flos": 18646295379480.0, + "grad_norm": 4.436443579205052, + "learning_rate": 2.0204023671974267e-07, + "loss": 0.6553, + "num_input_tokens_seen": 152786740, + "step": 7152 + }, + { + "epoch": 0.8600973967414177, + "flos": 12189528473520.0, + "grad_norm": 3.9655555424420514, + "learning_rate": 2.0169919137129532e-07, + "loss": 0.7966, + "num_input_tokens_seen": 152804900, + "step": 7153 + }, + { + "epoch": 0.8602176396320568, + "flos": 18483431833440.0, + "grad_norm": 9.634012337836502, + "learning_rate": 2.013584188236508e-07, + "loss": 0.6771, + "num_input_tokens_seen": 152822525, + "step": 7154 + }, + { + "epoch": 0.8603378825226958, + "flos": 14919556436400.0, + "grad_norm": 2.911139855796998, + "learning_rate": 2.0101791912850396e-07, + "loss": 0.7772, + "num_input_tokens_seen": 152841785, + "step": 7155 + }, + { + "epoch": 0.8604581254133349, + "flos": 25663349258880.0, + "grad_norm": 9.317243967839953, + "learning_rate": 2.006776923375082e-07, + "loss": 0.6244, + "num_input_tokens_seen": 152863160, + "step": 7156 + }, + { + "epoch": 0.860578368303974, + "flos": 16535401272960.0, + "grad_norm": 3.586898774598172, + "learning_rate": 2.003377385022764e-07, + "loss": 0.6962, + "num_input_tokens_seen": 152881705, + "step": 7157 + }, + { + "epoch": 0.8606986111946131, + "flos": 15594561139440.0, + "grad_norm": 8.289107600313216, + "learning_rate": 1.9999805767437826e-07, + "loss": 0.7561, + "num_input_tokens_seen": 152900315, + "step": 7158 + }, + { + "epoch": 0.8608188540852522, + "flos": 21184242144600.0, + "grad_norm": 3.8478643215868864, + "learning_rate": 1.9965864990534386e-07, + "loss": 0.7016, + "num_input_tokens_seen": 152920560, + "step": 7159 + }, + { + "epoch": 0.8609390969758913, + "flos": 21805117298760.0, + "grad_norm": 3.5306175813412968, + "learning_rate": 1.9931951524666092e-07, + "loss": 0.7611, + "num_input_tokens_seen": 152941370, + "step": 7160 + }, + { + "epoch": 0.8610593398665304, + "flos": 15538785272640.0, + "grad_norm": 6.370513534087207, + "learning_rate": 1.9898065374977534e-07, + "loss": 0.7935, + "num_input_tokens_seen": 152961295, + "step": 7161 + }, + { + "epoch": 0.8611795827571694, + "flos": 10788428874720.0, + "grad_norm": 3.615947337771408, + "learning_rate": 1.9864206546609342e-07, + "loss": 0.7163, + "num_input_tokens_seen": 152979855, + "step": 7162 + }, + { + "epoch": 0.8612998256478086, + "flos": 17754772480920.0, + "grad_norm": 3.404751369232635, + "learning_rate": 1.983037504469771e-07, + "loss": 0.8293, + "num_input_tokens_seen": 152998285, + "step": 7163 + }, + { + "epoch": 0.8614200685384477, + "flos": 15541697988960.0, + "grad_norm": 3.250895450184048, + "learning_rate": 1.9796570874374984e-07, + "loss": 0.6539, + "num_input_tokens_seen": 153018110, + "step": 7164 + }, + { + "epoch": 0.8615403114290867, + "flos": 14619754360200.0, + "grad_norm": 3.526705222565061, + "learning_rate": 1.976279404076917e-07, + "loss": 0.7653, + "num_input_tokens_seen": 153037230, + "step": 7165 + }, + { + "epoch": 0.8616605543197259, + "flos": 21859310167560.0, + "grad_norm": 3.5411116370825897, + "learning_rate": 1.9729044549004193e-07, + "loss": 0.7451, + "num_input_tokens_seen": 153058335, + "step": 7166 + }, + { + "epoch": 0.8617807972103649, + "flos": 21209407321440.0, + "grad_norm": 3.630061726321173, + "learning_rate": 1.9695322404199822e-07, + "loss": 0.6894, + "num_input_tokens_seen": 153080100, + "step": 7167 + }, + { + "epoch": 0.861901040101004, + "flos": 20293732364760.0, + "grad_norm": 4.491173246573473, + "learning_rate": 1.9661627611471654e-07, + "loss": 0.813, + "num_input_tokens_seen": 153099615, + "step": 7168 + }, + { + "epoch": 0.8620212829916432, + "flos": 36630213888600.0, + "grad_norm": 3.7110169091010334, + "learning_rate": 1.9627960175931246e-07, + "loss": 0.6864, + "num_input_tokens_seen": 153124035, + "step": 7169 + }, + { + "epoch": 0.8621415258822822, + "flos": 15456704470440.0, + "grad_norm": 3.5945673800839204, + "learning_rate": 1.9594320102685847e-07, + "loss": 0.728, + "num_input_tokens_seen": 153143025, + "step": 7170 + }, + { + "epoch": 0.8622617687729213, + "flos": 15864449044800.0, + "grad_norm": 4.32419493133684, + "learning_rate": 1.956070739683864e-07, + "loss": 0.6143, + "num_input_tokens_seen": 153162080, + "step": 7171 + }, + { + "epoch": 0.8623820116635604, + "flos": 19240422358800.0, + "grad_norm": 4.206951682525775, + "learning_rate": 1.9527122063488678e-07, + "loss": 0.7248, + "num_input_tokens_seen": 153182915, + "step": 7172 + }, + { + "epoch": 0.8625022545541995, + "flos": 14353855609800.0, + "grad_norm": 4.741667214624321, + "learning_rate": 1.9493564107730755e-07, + "loss": 0.7876, + "num_input_tokens_seen": 153202635, + "step": 7173 + }, + { + "epoch": 0.8626224974448385, + "flos": 16026837691440.0, + "grad_norm": 4.277244977812205, + "learning_rate": 1.9460033534655684e-07, + "loss": 0.6029, + "num_input_tokens_seen": 153221715, + "step": 7174 + }, + { + "epoch": 0.8627427403354777, + "flos": 17079799437840.0, + "grad_norm": 2.226230743856112, + "learning_rate": 1.9426530349349978e-07, + "loss": 0.8232, + "num_input_tokens_seen": 153241885, + "step": 7175 + }, + { + "epoch": 0.8628629832261168, + "flos": 11922838224120.0, + "grad_norm": 3.9013156090144947, + "learning_rate": 1.9393054556896038e-07, + "loss": 0.6428, + "num_input_tokens_seen": 153259305, + "step": 7176 + }, + { + "epoch": 0.8629832261167558, + "flos": 20610974587560.0, + "grad_norm": 9.572669113550281, + "learning_rate": 1.9359606162372133e-07, + "loss": 0.6693, + "num_input_tokens_seen": 153280630, + "step": 7177 + }, + { + "epoch": 0.863103469007395, + "flos": 14784770783520.0, + "grad_norm": 4.310437580031482, + "learning_rate": 1.9326185170852293e-07, + "loss": 0.6978, + "num_input_tokens_seen": 153299315, + "step": 7178 + }, + { + "epoch": 0.863223711898034, + "flos": 17943687682680.0, + "grad_norm": 4.718772029193715, + "learning_rate": 1.9292791587406598e-07, + "loss": 0.711, + "num_input_tokens_seen": 153317895, + "step": 7179 + }, + { + "epoch": 0.8633439547886731, + "flos": 12894035767920.0, + "grad_norm": 4.23708467951974, + "learning_rate": 1.9259425417100661e-07, + "loss": 0.8497, + "num_input_tokens_seen": 153333730, + "step": 7180 + }, + { + "epoch": 0.8634641976793123, + "flos": 9356180366640.0, + "grad_norm": 4.5734366408784854, + "learning_rate": 1.9226086664996234e-07, + "loss": 0.7253, + "num_input_tokens_seen": 153351695, + "step": 7181 + }, + { + "epoch": 0.8635844405699513, + "flos": 17485264495080.0, + "grad_norm": 5.088088557810206, + "learning_rate": 1.9192775336150712e-07, + "loss": 0.7294, + "num_input_tokens_seen": 153371715, + "step": 7182 + }, + { + "epoch": 0.8637046834605904, + "flos": 44555000566080.0, + "grad_norm": 0.7862492635023006, + "learning_rate": 1.915949143561739e-07, + "loss": 0.5665, + "num_input_tokens_seen": 153426110, + "step": 7183 + }, + { + "epoch": 0.8638249263512295, + "flos": 15028068813840.0, + "grad_norm": 6.2525626460143515, + "learning_rate": 1.9126234968445498e-07, + "loss": 0.751, + "num_input_tokens_seen": 153445520, + "step": 7184 + }, + { + "epoch": 0.8639451692418686, + "flos": 19510721843640.0, + "grad_norm": 2.19763408368998, + "learning_rate": 1.9093005939679884e-07, + "loss": 0.6603, + "num_input_tokens_seen": 153467195, + "step": 7185 + }, + { + "epoch": 0.8640654121325076, + "flos": 11004852090360.0, + "grad_norm": 16.53494815528754, + "learning_rate": 1.9059804354361452e-07, + "loss": 0.746, + "num_input_tokens_seen": 153484690, + "step": 7186 + }, + { + "epoch": 0.8641856550231467, + "flos": 23398424735160.0, + "grad_norm": 3.4845275845497037, + "learning_rate": 1.902663021752684e-07, + "loss": 0.6938, + "num_input_tokens_seen": 153505840, + "step": 7187 + }, + { + "epoch": 0.8643058979137859, + "flos": 10896877932240.0, + "grad_norm": 4.08526477417884, + "learning_rate": 1.8993483534208556e-07, + "loss": 0.8128, + "num_input_tokens_seen": 153524470, + "step": 7188 + }, + { + "epoch": 0.8644261408044249, + "flos": 9519740431800.0, + "grad_norm": 8.507988709865982, + "learning_rate": 1.8960364309434884e-07, + "loss": 0.7274, + "num_input_tokens_seen": 153541685, + "step": 7189 + }, + { + "epoch": 0.864546383695064, + "flos": 15244333729680.0, + "grad_norm": 8.478676021792525, + "learning_rate": 1.8927272548229967e-07, + "loss": 0.7638, + "num_input_tokens_seen": 153561095, + "step": 7190 + }, + { + "epoch": 0.8646666265857031, + "flos": 15594782759160.0, + "grad_norm": 6.970393376243559, + "learning_rate": 1.8894208255613876e-07, + "loss": 0.8152, + "num_input_tokens_seen": 153580130, + "step": 7191 + }, + { + "epoch": 0.8647868694763422, + "flos": 14429414460240.0, + "grad_norm": 3.613941557064926, + "learning_rate": 1.8861171436602397e-07, + "loss": 0.7527, + "num_input_tokens_seen": 153596965, + "step": 7192 + }, + { + "epoch": 0.8649071123669813, + "flos": 19185659610720.0, + "grad_norm": 6.0037874347791265, + "learning_rate": 1.882816209620719e-07, + "loss": 0.7881, + "num_input_tokens_seen": 153613395, + "step": 7193 + }, + { + "epoch": 0.8650273552576204, + "flos": 14838647052720.0, + "grad_norm": 3.977316434545762, + "learning_rate": 1.8795180239435738e-07, + "loss": 0.7468, + "num_input_tokens_seen": 153631970, + "step": 7194 + }, + { + "epoch": 0.8651475981482595, + "flos": 17538507565080.0, + "grad_norm": 7.590940858528393, + "learning_rate": 1.8762225871291348e-07, + "loss": 0.736, + "num_input_tokens_seen": 153647565, + "step": 7195 + }, + { + "epoch": 0.8652678410388985, + "flos": 15860554869720.0, + "grad_norm": 3.74149209332475, + "learning_rate": 1.8729298996773201e-07, + "loss": 0.7993, + "num_input_tokens_seen": 153666035, + "step": 7196 + }, + { + "epoch": 0.8653880839295377, + "flos": 48090042722280.0, + "grad_norm": 0.8499340270189072, + "learning_rate": 1.8696399620876301e-07, + "loss": 0.6251, + "num_input_tokens_seen": 153722785, + "step": 7197 + }, + { + "epoch": 0.8655083268201768, + "flos": 12948766856040.0, + "grad_norm": 4.811446830150493, + "learning_rate": 1.866352774859141e-07, + "loss": 0.7771, + "num_input_tokens_seen": 153737730, + "step": 7198 + }, + { + "epoch": 0.8656285697108158, + "flos": 15135441432720.0, + "grad_norm": 11.05704928668687, + "learning_rate": 1.8630683384905188e-07, + "loss": 0.6685, + "num_input_tokens_seen": 153756780, + "step": 7199 + }, + { + "epoch": 0.865748812601455, + "flos": 13617217947360.0, + "grad_norm": 16.512267205375487, + "learning_rate": 1.8597866534800045e-07, + "loss": 0.8803, + "num_input_tokens_seen": 153771615, + "step": 7200 + }, + { + "epoch": 0.865869055492094, + "flos": 52100685950280.0, + "grad_norm": 4.676354343664492, + "learning_rate": 1.8565077203254398e-07, + "loss": 0.7325, + "num_input_tokens_seen": 153796795, + "step": 7201 + }, + { + "epoch": 0.8659892983827331, + "flos": 12678435711240.0, + "grad_norm": 51.159495162799445, + "learning_rate": 1.8532315395242203e-07, + "loss": 0.7022, + "num_input_tokens_seen": 153812965, + "step": 7202 + }, + { + "epoch": 0.8661095412733723, + "flos": 13056867654000.0, + "grad_norm": 5.041795755210226, + "learning_rate": 1.849958111573353e-07, + "loss": 0.7036, + "num_input_tokens_seen": 153831290, + "step": 7203 + }, + { + "epoch": 0.8662297841640113, + "flos": 13300387304040.0, + "grad_norm": 3.724457664180511, + "learning_rate": 1.8466874369694074e-07, + "loss": 0.6269, + "num_input_tokens_seen": 153848705, + "step": 7204 + }, + { + "epoch": 0.8663500270546504, + "flos": 11922205024920.0, + "grad_norm": 5.262845264987669, + "learning_rate": 1.843419516208542e-07, + "loss": 0.6941, + "num_input_tokens_seen": 153865350, + "step": 7205 + }, + { + "epoch": 0.8664702699452895, + "flos": 13056171134880.0, + "grad_norm": 4.1487740417868855, + "learning_rate": 1.8401543497865047e-07, + "loss": 0.7694, + "num_input_tokens_seen": 153883070, + "step": 7206 + }, + { + "epoch": 0.8665905128359286, + "flos": 22559923286880.0, + "grad_norm": 11.154371258342861, + "learning_rate": 1.836891938198608e-07, + "loss": 0.6181, + "num_input_tokens_seen": 153903215, + "step": 7207 + }, + { + "epoch": 0.8667107557265676, + "flos": 13619497464480.0, + "grad_norm": 6.472346830828456, + "learning_rate": 1.8336322819397677e-07, + "loss": 0.6949, + "num_input_tokens_seen": 153920470, + "step": 7208 + }, + { + "epoch": 0.8668309986172068, + "flos": 14676416705880.0, + "grad_norm": 5.95906855530921, + "learning_rate": 1.8303753815044654e-07, + "loss": 0.6033, + "num_input_tokens_seen": 153939495, + "step": 7209 + }, + { + "epoch": 0.8669512415078459, + "flos": 15809591316840.0, + "grad_norm": 4.4484354590628, + "learning_rate": 1.827121237386773e-07, + "loss": 0.6815, + "num_input_tokens_seen": 153956660, + "step": 7210 + }, + { + "epoch": 0.8670714843984849, + "flos": 12913755431640.0, + "grad_norm": 7.195850596566432, + "learning_rate": 1.8238698500803374e-07, + "loss": 0.7365, + "num_input_tokens_seen": 153969145, + "step": 7211 + }, + { + "epoch": 0.8671917272891241, + "flos": 44746263096240.0, + "grad_norm": 0.8149739528074365, + "learning_rate": 1.820621220078391e-07, + "loss": 0.6127, + "num_input_tokens_seen": 154032775, + "step": 7212 + }, + { + "epoch": 0.8673119701797631, + "flos": 14949343967400.0, + "grad_norm": 3.2544194030955995, + "learning_rate": 1.8173753478737553e-07, + "loss": 0.6588, + "num_input_tokens_seen": 154052930, + "step": 7213 + }, + { + "epoch": 0.8674322130704022, + "flos": 14353728969960.0, + "grad_norm": 3.880649509140573, + "learning_rate": 1.8141322339588205e-07, + "loss": 0.7734, + "num_input_tokens_seen": 154069990, + "step": 7214 + }, + { + "epoch": 0.8675524559610414, + "flos": 19072524879120.0, + "grad_norm": 5.38494255099029, + "learning_rate": 1.810891878825569e-07, + "loss": 0.6845, + "num_input_tokens_seen": 154089685, + "step": 7215 + }, + { + "epoch": 0.8676726988516804, + "flos": 10950817521360.0, + "grad_norm": 4.54733179688345, + "learning_rate": 1.8076542829655561e-07, + "loss": 0.6921, + "num_input_tokens_seen": 154108210, + "step": 7216 + }, + { + "epoch": 0.8677929417423195, + "flos": 11867252317080.0, + "grad_norm": 6.929441978810077, + "learning_rate": 1.8044194468699203e-07, + "loss": 0.7888, + "num_input_tokens_seen": 154125240, + "step": 7217 + }, + { + "epoch": 0.8679131846329585, + "flos": 13759190411160.0, + "grad_norm": 5.372786665023241, + "learning_rate": 1.8011873710293912e-07, + "loss": 0.7207, + "num_input_tokens_seen": 154143465, + "step": 7218 + }, + { + "epoch": 0.8680334275235977, + "flos": 24694811151720.0, + "grad_norm": 3.2813450390812196, + "learning_rate": 1.7979580559342677e-07, + "loss": 0.671, + "num_input_tokens_seen": 154163915, + "step": 7219 + }, + { + "epoch": 0.8681536704142367, + "flos": 18078853255080.0, + "grad_norm": 3.9014865976328545, + "learning_rate": 1.7947315020744358e-07, + "loss": 0.6555, + "num_input_tokens_seen": 154184730, + "step": 7220 + }, + { + "epoch": 0.8682739133048758, + "flos": 14622667076520.0, + "grad_norm": 4.95464371463673, + "learning_rate": 1.7915077099393594e-07, + "loss": 0.781, + "num_input_tokens_seen": 154201050, + "step": 7221 + }, + { + "epoch": 0.868394156195515, + "flos": 12138849860280.0, + "grad_norm": 17.707604726125968, + "learning_rate": 1.788286680018083e-07, + "loss": 0.722, + "num_input_tokens_seen": 154219480, + "step": 7222 + }, + { + "epoch": 0.868514399086154, + "flos": 20536048936320.0, + "grad_norm": 4.151419909135053, + "learning_rate": 1.7850684127992443e-07, + "loss": 0.7086, + "num_input_tokens_seen": 154238945, + "step": 7223 + }, + { + "epoch": 0.8686346419767931, + "flos": 14677176544920.0, + "grad_norm": 3.250481936594848, + "learning_rate": 1.7818529087710378e-07, + "loss": 0.6899, + "num_input_tokens_seen": 154259020, + "step": 7224 + }, + { + "epoch": 0.8687548848674322, + "flos": 13299817424760.0, + "grad_norm": 3.8409872903809723, + "learning_rate": 1.7786401684212637e-07, + "loss": 0.8315, + "num_input_tokens_seen": 154277570, + "step": 7225 + }, + { + "epoch": 0.8688751277580713, + "flos": 51960047696160.0, + "grad_norm": 0.7516812712469269, + "learning_rate": 1.7754301922372883e-07, + "loss": 0.5642, + "num_input_tokens_seen": 154326935, + "step": 7226 + }, + { + "epoch": 0.8689953706487104, + "flos": 19725973640760.0, + "grad_norm": 4.291929193757002, + "learning_rate": 1.7722229807060617e-07, + "loss": 0.795, + "num_input_tokens_seen": 154345235, + "step": 7227 + }, + { + "epoch": 0.8691156135393495, + "flos": 25446514463760.0, + "grad_norm": 4.281550041878338, + "learning_rate": 1.7690185343141172e-07, + "loss": 0.806, + "num_input_tokens_seen": 154364870, + "step": 7228 + }, + { + "epoch": 0.8692358564299886, + "flos": 13866942949560.0, + "grad_norm": 6.061438398715752, + "learning_rate": 1.7658168535475615e-07, + "loss": 0.6886, + "num_input_tokens_seen": 154382375, + "step": 7229 + }, + { + "epoch": 0.8693560993206276, + "flos": 22289307202440.0, + "grad_norm": 3.341715826749493, + "learning_rate": 1.7626179388920948e-07, + "loss": 0.6317, + "num_input_tokens_seen": 154403375, + "step": 7230 + }, + { + "epoch": 0.8694763422112668, + "flos": 19943536614960.0, + "grad_norm": 3.047115531621556, + "learning_rate": 1.7594217908329866e-07, + "loss": 0.7918, + "num_input_tokens_seen": 154425280, + "step": 7231 + }, + { + "epoch": 0.8695965851019059, + "flos": 19157676697440.0, + "grad_norm": 2.8955239512619375, + "learning_rate": 1.7562284098550895e-07, + "loss": 0.7198, + "num_input_tokens_seen": 154444710, + "step": 7232 + }, + { + "epoch": 0.8697168279925449, + "flos": 49648908613560.0, + "grad_norm": 0.9025305022160012, + "learning_rate": 1.753037796442838e-07, + "loss": 0.6533, + "num_input_tokens_seen": 154503870, + "step": 7233 + }, + { + "epoch": 0.8698370708831841, + "flos": 14406243860880.0, + "grad_norm": 3.5224407672436784, + "learning_rate": 1.74984995108024e-07, + "loss": 0.7374, + "num_input_tokens_seen": 154521520, + "step": 7234 + }, + { + "epoch": 0.8699573137738231, + "flos": 9330413650560.0, + "grad_norm": 6.942929729147267, + "learning_rate": 1.7466648742508981e-07, + "loss": 0.8141, + "num_input_tokens_seen": 154537425, + "step": 7235 + }, + { + "epoch": 0.8700775566644622, + "flos": 12705690445440.0, + "grad_norm": 5.381478951398556, + "learning_rate": 1.7434825664379837e-07, + "loss": 0.828, + "num_input_tokens_seen": 154555650, + "step": 7236 + }, + { + "epoch": 0.8701977995551013, + "flos": 9978100299480.0, + "grad_norm": 7.804235863433997, + "learning_rate": 1.740303028124246e-07, + "loss": 0.8333, + "num_input_tokens_seen": 154571430, + "step": 7237 + }, + { + "epoch": 0.8703180424457404, + "flos": 22426467352320.0, + "grad_norm": 5.05594691251094, + "learning_rate": 1.7371262597920212e-07, + "loss": 0.7444, + "num_input_tokens_seen": 154593210, + "step": 7238 + }, + { + "epoch": 0.8704382853363795, + "flos": 14325017877600.0, + "grad_norm": 5.290402579321723, + "learning_rate": 1.7339522619232195e-07, + "loss": 0.7426, + "num_input_tokens_seen": 154611310, + "step": 7239 + }, + { + "epoch": 0.8705585282270186, + "flos": 19509075525720.0, + "grad_norm": 4.124407407792555, + "learning_rate": 1.730781034999338e-07, + "loss": 0.7193, + "num_input_tokens_seen": 154632610, + "step": 7240 + }, + { + "epoch": 0.8706787711176577, + "flos": 25041809245560.0, + "grad_norm": 3.796213100179907, + "learning_rate": 1.7276125795014497e-07, + "loss": 0.7206, + "num_input_tokens_seen": 154650780, + "step": 7241 + }, + { + "epoch": 0.8707990140082967, + "flos": 10626705087240.0, + "grad_norm": 3.2811803901191476, + "learning_rate": 1.7244468959102054e-07, + "loss": 0.6524, + "num_input_tokens_seen": 154667555, + "step": 7242 + }, + { + "epoch": 0.8709192568989359, + "flos": 14676796625400.0, + "grad_norm": 5.87280232332966, + "learning_rate": 1.7212839847058348e-07, + "loss": 0.8425, + "num_input_tokens_seen": 154682405, + "step": 7243 + }, + { + "epoch": 0.871039499789575, + "flos": 12192282890040.0, + "grad_norm": 4.184690073585246, + "learning_rate": 1.718123846368147e-07, + "loss": 0.723, + "num_input_tokens_seen": 154701170, + "step": 7244 + }, + { + "epoch": 0.871159742680214, + "flos": 15405424317960.0, + "grad_norm": 3.489324227957336, + "learning_rate": 1.714966481376543e-07, + "loss": 0.7088, + "num_input_tokens_seen": 154717570, + "step": 7245 + }, + { + "epoch": 0.8712799855708532, + "flos": 20724837498240.0, + "grad_norm": 3.429156077582201, + "learning_rate": 1.7118118902099797e-07, + "loss": 0.7902, + "num_input_tokens_seen": 154735375, + "step": 7246 + }, + { + "epoch": 0.8714002284614922, + "flos": 16269787462200.0, + "grad_norm": 3.999618154419813, + "learning_rate": 1.7086600733470146e-07, + "loss": 0.7856, + "num_input_tokens_seen": 154755765, + "step": 7247 + }, + { + "epoch": 0.8715204713521313, + "flos": 15673919185080.0, + "grad_norm": 3.657252247698523, + "learning_rate": 1.7055110312657738e-07, + "loss": 0.7524, + "num_input_tokens_seen": 154774980, + "step": 7248 + }, + { + "epoch": 0.8716407142427703, + "flos": 17160613841640.0, + "grad_norm": 4.884248031166727, + "learning_rate": 1.702364764443962e-07, + "loss": 0.7332, + "num_input_tokens_seen": 154793775, + "step": 7249 + }, + { + "epoch": 0.8717609571334095, + "flos": 20503570308720.0, + "grad_norm": 5.631100885277743, + "learning_rate": 1.6992212733588685e-07, + "loss": 0.7128, + "num_input_tokens_seen": 154813160, + "step": 7250 + }, + { + "epoch": 0.8718812000240486, + "flos": 18669054399360.0, + "grad_norm": 4.274345454919755, + "learning_rate": 1.6960805584873538e-07, + "loss": 0.7339, + "num_input_tokens_seen": 154830880, + "step": 7251 + }, + { + "epoch": 0.8720014429146876, + "flos": 17132820888120.0, + "grad_norm": 2.969842328340041, + "learning_rate": 1.6929426203058684e-07, + "loss": 0.7721, + "num_input_tokens_seen": 154851025, + "step": 7252 + }, + { + "epoch": 0.8721216858053268, + "flos": 17835871824360.0, + "grad_norm": 4.660702413576202, + "learning_rate": 1.689807459290431e-07, + "loss": 0.799, + "num_input_tokens_seen": 154869400, + "step": 7253 + }, + { + "epoch": 0.8722419286959658, + "flos": 24878154200520.0, + "grad_norm": 5.9208490695874, + "learning_rate": 1.6866750759166437e-07, + "loss": 0.6955, + "num_input_tokens_seen": 154889100, + "step": 7254 + }, + { + "epoch": 0.8723621715866049, + "flos": 13407759922920.0, + "grad_norm": 6.222958468218149, + "learning_rate": 1.6835454706596865e-07, + "loss": 0.7575, + "num_input_tokens_seen": 154906650, + "step": 7255 + }, + { + "epoch": 0.8724824144772441, + "flos": 16103884560000.0, + "grad_norm": 4.794035925662528, + "learning_rate": 1.680418643994317e-07, + "loss": 0.7272, + "num_input_tokens_seen": 154924940, + "step": 7256 + }, + { + "epoch": 0.8726026573678831, + "flos": 49177853101920.0, + "grad_norm": 0.9456989664823797, + "learning_rate": 1.6772945963948738e-07, + "loss": 0.6795, + "num_input_tokens_seen": 154982825, + "step": 7257 + }, + { + "epoch": 0.8727229002585222, + "flos": 9708560653680.0, + "grad_norm": 10.055881384903369, + "learning_rate": 1.6741733283352733e-07, + "loss": 0.7435, + "num_input_tokens_seen": 155000150, + "step": 7258 + }, + { + "epoch": 0.8728431431491613, + "flos": 15944123690040.0, + "grad_norm": 3.414580351708169, + "learning_rate": 1.6710548402890102e-07, + "loss": 0.8252, + "num_input_tokens_seen": 155020395, + "step": 7259 + }, + { + "epoch": 0.8729633860398004, + "flos": 26584153129080.0, + "grad_norm": 6.14101760036305, + "learning_rate": 1.6679391327291527e-07, + "loss": 0.6511, + "num_input_tokens_seen": 155041320, + "step": 7260 + }, + { + "epoch": 0.8730836289304394, + "flos": 11920020487680.0, + "grad_norm": 16.910724266418743, + "learning_rate": 1.6648262061283492e-07, + "loss": 0.6533, + "num_input_tokens_seen": 155056340, + "step": 7261 + }, + { + "epoch": 0.8732038718210786, + "flos": 15513778395600.0, + "grad_norm": 4.870916759515092, + "learning_rate": 1.6617160609588353e-07, + "loss": 0.7231, + "num_input_tokens_seen": 155075235, + "step": 7262 + }, + { + "epoch": 0.8733241147117177, + "flos": 12105168154200.0, + "grad_norm": 4.945282030539033, + "learning_rate": 1.6586086976924163e-07, + "loss": 0.6983, + "num_input_tokens_seen": 155090455, + "step": 7263 + }, + { + "epoch": 0.8734443576023567, + "flos": 14758085928600.0, + "grad_norm": 6.002242622879073, + "learning_rate": 1.6555041168004747e-07, + "loss": 0.7738, + "num_input_tokens_seen": 155109495, + "step": 7264 + }, + { + "epoch": 0.8735646004929959, + "flos": 13164873472080.0, + "grad_norm": 5.654884974543459, + "learning_rate": 1.6524023187539715e-07, + "loss": 0.6714, + "num_input_tokens_seen": 155127500, + "step": 7265 + }, + { + "epoch": 0.873684843383635, + "flos": 14808068022720.0, + "grad_norm": 3.42455044322904, + "learning_rate": 1.649303304023446e-07, + "loss": 0.7253, + "num_input_tokens_seen": 155146975, + "step": 7266 + }, + { + "epoch": 0.873805086274274, + "flos": 12355431375720.0, + "grad_norm": 3.2323016464139984, + "learning_rate": 1.6462070730790246e-07, + "loss": 0.7739, + "num_input_tokens_seen": 155165855, + "step": 7267 + }, + { + "epoch": 0.8739253291649132, + "flos": 13164968451960.0, + "grad_norm": 6.886958870512504, + "learning_rate": 1.6431136263903912e-07, + "loss": 0.7629, + "num_input_tokens_seen": 155184575, + "step": 7268 + }, + { + "epoch": 0.8740455720555522, + "flos": 15595225998600.0, + "grad_norm": 4.6015527592347025, + "learning_rate": 1.6400229644268282e-07, + "loss": 0.7178, + "num_input_tokens_seen": 155202650, + "step": 7269 + }, + { + "epoch": 0.8741658149461913, + "flos": 11031631925160.0, + "grad_norm": 6.413677097130008, + "learning_rate": 1.6369350876571852e-07, + "loss": 0.7932, + "num_input_tokens_seen": 155220525, + "step": 7270 + }, + { + "epoch": 0.8742860578368304, + "flos": 16864262701080.0, + "grad_norm": 5.00136814472718, + "learning_rate": 1.6338499965498874e-07, + "loss": 0.7829, + "num_input_tokens_seen": 155238975, + "step": 7271 + }, + { + "epoch": 0.8744063007274695, + "flos": 20642313456600.0, + "grad_norm": 3.0133505036137755, + "learning_rate": 1.630767691572943e-07, + "loss": 0.7611, + "num_input_tokens_seen": 155258715, + "step": 7272 + }, + { + "epoch": 0.8745265436181086, + "flos": 47205670483320.0, + "grad_norm": 0.7519884373683949, + "learning_rate": 1.6276881731939306e-07, + "loss": 0.5528, + "num_input_tokens_seen": 155320325, + "step": 7273 + }, + { + "epoch": 0.8746467865087477, + "flos": 21022518357120.0, + "grad_norm": 5.098591194793995, + "learning_rate": 1.6246114418800193e-07, + "loss": 0.7394, + "num_input_tokens_seen": 155340325, + "step": 7274 + }, + { + "epoch": 0.8747670293993868, + "flos": 17563989341520.0, + "grad_norm": 3.0473308515664477, + "learning_rate": 1.6215374980979423e-07, + "loss": 0.7513, + "num_input_tokens_seen": 155360455, + "step": 7275 + }, + { + "epoch": 0.8748872722900258, + "flos": 33280038950640.0, + "grad_norm": 5.233734911597828, + "learning_rate": 1.6184663423140133e-07, + "loss": 0.6646, + "num_input_tokens_seen": 155380475, + "step": 7276 + }, + { + "epoch": 0.875007515180665, + "flos": 14432453816400.0, + "grad_norm": 10.952658998423269, + "learning_rate": 1.615397974994126e-07, + "loss": 0.6312, + "num_input_tokens_seen": 155398000, + "step": 7277 + }, + { + "epoch": 0.875127758071304, + "flos": 16620584751240.0, + "grad_norm": 3.406448078520751, + "learning_rate": 1.6123323966037438e-07, + "loss": 0.7922, + "num_input_tokens_seen": 155416240, + "step": 7278 + }, + { + "epoch": 0.8752480009619431, + "flos": 17133517407240.0, + "grad_norm": 5.000604592963336, + "learning_rate": 1.6092696076079216e-07, + "loss": 0.772, + "num_input_tokens_seen": 155434335, + "step": 7279 + }, + { + "epoch": 0.8753682438525822, + "flos": 19212534425400.0, + "grad_norm": 4.080461205613, + "learning_rate": 1.6062096084712785e-07, + "loss": 0.7275, + "num_input_tokens_seen": 155455405, + "step": 7280 + }, + { + "epoch": 0.8754884867432213, + "flos": 17075873602800.0, + "grad_norm": 27.035070178034495, + "learning_rate": 1.6031523996580098e-07, + "loss": 0.6952, + "num_input_tokens_seen": 155472685, + "step": 7281 + }, + { + "epoch": 0.8756087296338604, + "flos": 9060494085240.0, + "grad_norm": 9.147236941533262, + "learning_rate": 1.6000979816318981e-07, + "loss": 0.6441, + "num_input_tokens_seen": 155490870, + "step": 7282 + }, + { + "epoch": 0.8757289725244994, + "flos": 13839561575520.0, + "grad_norm": 4.8678989043571965, + "learning_rate": 1.5970463548562886e-07, + "loss": 0.7341, + "num_input_tokens_seen": 155509745, + "step": 7283 + }, + { + "epoch": 0.8758492154151386, + "flos": 18511794666240.0, + "grad_norm": 3.563203976325843, + "learning_rate": 1.5939975197941192e-07, + "loss": 0.7, + "num_input_tokens_seen": 155531120, + "step": 7284 + }, + { + "epoch": 0.8759694583057777, + "flos": 48344322267360.0, + "grad_norm": 0.8436252007200878, + "learning_rate": 1.5909514769078892e-07, + "loss": 0.5532, + "num_input_tokens_seen": 155595945, + "step": 7285 + }, + { + "epoch": 0.8760897011964167, + "flos": 18644870681280.0, + "grad_norm": 3.4568931767536353, + "learning_rate": 1.5879082266596867e-07, + "loss": 0.7557, + "num_input_tokens_seen": 155617005, + "step": 7286 + }, + { + "epoch": 0.8762099440870559, + "flos": 21263631850200.0, + "grad_norm": 2.711802056169606, + "learning_rate": 1.5848677695111645e-07, + "loss": 0.7221, + "num_input_tokens_seen": 155638325, + "step": 7287 + }, + { + "epoch": 0.8763301869776949, + "flos": 15804620703120.0, + "grad_norm": 4.178895451019534, + "learning_rate": 1.5818301059235562e-07, + "loss": 0.688, + "num_input_tokens_seen": 155653220, + "step": 7288 + }, + { + "epoch": 0.876450429868334, + "flos": 18051630180840.0, + "grad_norm": 3.0559338614811464, + "learning_rate": 1.578795236357684e-07, + "loss": 0.7983, + "num_input_tokens_seen": 155674405, + "step": 7289 + }, + { + "epoch": 0.8765706727589732, + "flos": 14806041785280.0, + "grad_norm": 6.156482982583786, + "learning_rate": 1.5757631612739218e-07, + "loss": 0.8415, + "num_input_tokens_seen": 155687670, + "step": 7290 + }, + { + "epoch": 0.8766909156496122, + "flos": 52636097177880.0, + "grad_norm": 0.8439182537590415, + "learning_rate": 1.572733881132242e-07, + "loss": 0.6632, + "num_input_tokens_seen": 155748035, + "step": 7291 + }, + { + "epoch": 0.8768111585402513, + "flos": 51270547222800.0, + "grad_norm": 0.7618151090268848, + "learning_rate": 1.5697073963921814e-07, + "loss": 0.6073, + "num_input_tokens_seen": 155806995, + "step": 7292 + }, + { + "epoch": 0.8769314014308904, + "flos": 13754251457400.0, + "grad_norm": 6.328337187800599, + "learning_rate": 1.566683707512857e-07, + "loss": 0.8267, + "num_input_tokens_seen": 155824390, + "step": 7293 + }, + { + "epoch": 0.8770516443215295, + "flos": 10896751292400.0, + "grad_norm": 5.622161827382884, + "learning_rate": 1.5636628149529553e-07, + "loss": 0.7828, + "num_input_tokens_seen": 155841900, + "step": 7294 + }, + { + "epoch": 0.8771718872121685, + "flos": 23235972768600.0, + "grad_norm": 17.908033458577066, + "learning_rate": 1.560644719170743e-07, + "loss": 0.7822, + "num_input_tokens_seen": 155862490, + "step": 7295 + }, + { + "epoch": 0.8772921301028077, + "flos": 26525401226040.0, + "grad_norm": 6.032066487601065, + "learning_rate": 1.5576294206240692e-07, + "loss": 0.7072, + "num_input_tokens_seen": 155881735, + "step": 7296 + }, + { + "epoch": 0.8774123729934468, + "flos": 42083526242880.0, + "grad_norm": 7.238688095689694, + "learning_rate": 1.5546169197703507e-07, + "loss": 0.6734, + "num_input_tokens_seen": 155907730, + "step": 7297 + }, + { + "epoch": 0.8775326158840858, + "flos": 17511411130680.0, + "grad_norm": 6.693127412452729, + "learning_rate": 1.5516072170665774e-07, + "loss": 0.7537, + "num_input_tokens_seen": 155925420, + "step": 7298 + }, + { + "epoch": 0.877652858774725, + "flos": 12484771515480.0, + "grad_norm": 3.448683700449647, + "learning_rate": 1.5486003129693214e-07, + "loss": 0.8432, + "num_input_tokens_seen": 155942505, + "step": 7299 + }, + { + "epoch": 0.877773101665364, + "flos": 12030970682040.0, + "grad_norm": 5.830863575126611, + "learning_rate": 1.545596207934725e-07, + "loss": 0.7589, + "num_input_tokens_seen": 155960710, + "step": 7300 + }, + { + "epoch": 0.8778933445560031, + "flos": 16133862050760.0, + "grad_norm": 3.580968268532182, + "learning_rate": 1.5425949024185147e-07, + "loss": 0.7562, + "num_input_tokens_seen": 155980455, + "step": 7301 + }, + { + "epoch": 0.8780135874466423, + "flos": 16512547273200.0, + "grad_norm": 4.502182960849533, + "learning_rate": 1.5395963968759818e-07, + "loss": 0.659, + "num_input_tokens_seen": 156000450, + "step": 7302 + }, + { + "epoch": 0.8781338303372813, + "flos": 45350987179440.0, + "grad_norm": 5.183229090060445, + "learning_rate": 1.536600691761998e-07, + "loss": 0.631, + "num_input_tokens_seen": 156026000, + "step": 7303 + }, + { + "epoch": 0.8782540732279204, + "flos": 16594026536160.0, + "grad_norm": 4.244675386236442, + "learning_rate": 1.5336077875310084e-07, + "loss": 0.6972, + "num_input_tokens_seen": 156044945, + "step": 7304 + }, + { + "epoch": 0.8783743161185595, + "flos": 11706256708680.0, + "grad_norm": 7.271674019233274, + "learning_rate": 1.5306176846370321e-07, + "loss": 0.7245, + "num_input_tokens_seen": 156062810, + "step": 7305 + }, + { + "epoch": 0.8784945590091986, + "flos": 19105161806520.0, + "grad_norm": 3.5426046737465193, + "learning_rate": 1.5276303835336712e-07, + "loss": 0.7157, + "num_input_tokens_seen": 156083070, + "step": 7306 + }, + { + "epoch": 0.8786148018998376, + "flos": 46233649780560.0, + "grad_norm": 0.7981908443357866, + "learning_rate": 1.524645884674094e-07, + "loss": 0.5494, + "num_input_tokens_seen": 156139720, + "step": 7307 + }, + { + "epoch": 0.8787350447904768, + "flos": 15837605890080.0, + "grad_norm": 5.106315980295872, + "learning_rate": 1.521664188511047e-07, + "loss": 0.7805, + "num_input_tokens_seen": 156159465, + "step": 7308 + }, + { + "epoch": 0.8788552876811159, + "flos": 18668167920480.0, + "grad_norm": 10.313339291722706, + "learning_rate": 1.518685295496851e-07, + "loss": 0.7918, + "num_input_tokens_seen": 156177045, + "step": 7309 + }, + { + "epoch": 0.8789755305717549, + "flos": 16324296930600.0, + "grad_norm": 2.6282751450984514, + "learning_rate": 1.5157092060833975e-07, + "loss": 0.8329, + "num_input_tokens_seen": 156196415, + "step": 7310 + }, + { + "epoch": 0.879095773462394, + "flos": 21504682023360.0, + "grad_norm": 8.776771788218088, + "learning_rate": 1.5127359207221658e-07, + "loss": 0.6377, + "num_input_tokens_seen": 156215615, + "step": 7311 + }, + { + "epoch": 0.8792160163530331, + "flos": 12164774876160.0, + "grad_norm": 5.70471002161057, + "learning_rate": 1.5097654398641923e-07, + "loss": 0.7091, + "num_input_tokens_seen": 156233240, + "step": 7312 + }, + { + "epoch": 0.8793362592436722, + "flos": 17944352541840.0, + "grad_norm": 2.5359050191607797, + "learning_rate": 1.5067977639601014e-07, + "loss": 0.7145, + "num_input_tokens_seen": 156255720, + "step": 7313 + }, + { + "epoch": 0.8794565021343113, + "flos": 10572575538360.0, + "grad_norm": 5.397132550197771, + "learning_rate": 1.5038328934600864e-07, + "loss": 0.6919, + "num_input_tokens_seen": 156272075, + "step": 7314 + }, + { + "epoch": 0.8795767450249504, + "flos": 29068128645120.0, + "grad_norm": 8.857991859072271, + "learning_rate": 1.5008708288139161e-07, + "loss": 0.687, + "num_input_tokens_seen": 156294155, + "step": 7315 + }, + { + "epoch": 0.8796969879155895, + "flos": 16804782618960.0, + "grad_norm": 6.514520418867516, + "learning_rate": 1.497911570470931e-07, + "loss": 0.7187, + "num_input_tokens_seen": 156313880, + "step": 7316 + }, + { + "epoch": 0.8798172308062285, + "flos": 20802707525760.0, + "grad_norm": 3.381160905549138, + "learning_rate": 1.494955118880048e-07, + "loss": 0.8422, + "num_input_tokens_seen": 156334585, + "step": 7317 + }, + { + "epoch": 0.8799374736968677, + "flos": 17566743758040.0, + "grad_norm": 5.387721067349399, + "learning_rate": 1.4920014744897634e-07, + "loss": 0.7185, + "num_input_tokens_seen": 156353720, + "step": 7318 + }, + { + "epoch": 0.8800577165875068, + "flos": 18781334312040.0, + "grad_norm": 3.6201592055239566, + "learning_rate": 1.4890506377481392e-07, + "loss": 0.8444, + "num_input_tokens_seen": 156372530, + "step": 7319 + }, + { + "epoch": 0.8801779594781458, + "flos": 17160518861760.0, + "grad_norm": 2.591670110113508, + "learning_rate": 1.486102609102815e-07, + "loss": 0.6209, + "num_input_tokens_seen": 156392800, + "step": 7320 + }, + { + "epoch": 0.880298202368785, + "flos": 8330093434920.0, + "grad_norm": 7.698700475643397, + "learning_rate": 1.483157389001004e-07, + "loss": 0.829, + "num_input_tokens_seen": 156410080, + "step": 7321 + }, + { + "epoch": 0.880418445259424, + "flos": 16591113819840.0, + "grad_norm": 7.354607789495283, + "learning_rate": 1.4802149778894933e-07, + "loss": 0.7596, + "num_input_tokens_seen": 156428590, + "step": 7322 + }, + { + "epoch": 0.8805386881500631, + "flos": 15000972379440.0, + "grad_norm": 4.144719240920858, + "learning_rate": 1.4772753762146484e-07, + "loss": 0.8575, + "num_input_tokens_seen": 156447565, + "step": 7323 + }, + { + "epoch": 0.8806589310407023, + "flos": 26854357634040.0, + "grad_norm": 3.518083992463418, + "learning_rate": 1.474338584422401e-07, + "loss": 0.6805, + "num_input_tokens_seen": 156472495, + "step": 7324 + }, + { + "epoch": 0.8807791739313413, + "flos": 17160392221920.0, + "grad_norm": 3.2454355906182397, + "learning_rate": 1.4714046029582595e-07, + "loss": 0.7378, + "num_input_tokens_seen": 156491280, + "step": 7325 + }, + { + "epoch": 0.8808994168219804, + "flos": 19022764404720.0, + "grad_norm": 5.088104114702809, + "learning_rate": 1.46847343226731e-07, + "loss": 0.7512, + "num_input_tokens_seen": 156512040, + "step": 7326 + }, + { + "epoch": 0.8810196597126195, + "flos": 12462424075080.0, + "grad_norm": 4.74871085893607, + "learning_rate": 1.465545072794203e-07, + "loss": 0.6828, + "num_input_tokens_seen": 156529635, + "step": 7327 + }, + { + "epoch": 0.8811399026032586, + "flos": 16836786347160.0, + "grad_norm": 5.080066861290858, + "learning_rate": 1.4626195249831774e-07, + "loss": 0.7426, + "num_input_tokens_seen": 156550255, + "step": 7328 + }, + { + "epoch": 0.8812601454938976, + "flos": 10517147931120.0, + "grad_norm": 4.904438048670019, + "learning_rate": 1.4596967892780244e-07, + "loss": 0.7065, + "num_input_tokens_seen": 156566305, + "step": 7329 + }, + { + "epoch": 0.8813803883845368, + "flos": 16459810762560.0, + "grad_norm": 69.619280650415, + "learning_rate": 1.4567768661221314e-07, + "loss": 0.7369, + "num_input_tokens_seen": 156586595, + "step": 7330 + }, + { + "epoch": 0.8815006312751759, + "flos": 15729695051880.0, + "grad_norm": 4.4671696355995545, + "learning_rate": 1.4538597559584442e-07, + "loss": 0.7221, + "num_input_tokens_seen": 156604105, + "step": 7331 + }, + { + "epoch": 0.8816208741658149, + "flos": 16675347499320.0, + "grad_norm": 6.235823154763168, + "learning_rate": 1.4509454592294823e-07, + "loss": 0.7602, + "num_input_tokens_seen": 156624310, + "step": 7332 + }, + { + "epoch": 0.8817411170564541, + "flos": 12970259477520.0, + "grad_norm": 5.438827527983434, + "learning_rate": 1.448033976377354e-07, + "loss": 0.7703, + "num_input_tokens_seen": 156639015, + "step": 7333 + }, + { + "epoch": 0.8818613599470931, + "flos": 13542767195520.0, + "grad_norm": 4.058063373294558, + "learning_rate": 1.445125307843713e-07, + "loss": 0.7207, + "num_input_tokens_seen": 156656960, + "step": 7334 + }, + { + "epoch": 0.8819816028377322, + "flos": 20238811316880.0, + "grad_norm": 3.5703385433524297, + "learning_rate": 1.442219454069813e-07, + "loss": 0.7307, + "num_input_tokens_seen": 156677705, + "step": 7335 + }, + { + "epoch": 0.8821018457283714, + "flos": 17134182266400.0, + "grad_norm": 3.8975621159896923, + "learning_rate": 1.4393164154964676e-07, + "loss": 0.6503, + "num_input_tokens_seen": 156696955, + "step": 7336 + }, + { + "epoch": 0.8822220886190104, + "flos": 21373125686400.0, + "grad_norm": 3.6705526365126095, + "learning_rate": 1.4364161925640649e-07, + "loss": 0.9233, + "num_input_tokens_seen": 156718075, + "step": 7337 + }, + { + "epoch": 0.8823423315096495, + "flos": 14973749305200.0, + "grad_norm": 5.252405069865074, + "learning_rate": 1.4335187857125663e-07, + "loss": 0.8374, + "num_input_tokens_seen": 156736495, + "step": 7338 + }, + { + "epoch": 0.8824625744002886, + "flos": 17727422766840.0, + "grad_norm": 4.976171169734623, + "learning_rate": 1.4306241953815023e-07, + "loss": 0.7261, + "num_input_tokens_seen": 156757275, + "step": 7339 + }, + { + "epoch": 0.8825828172909277, + "flos": 18077396896920.0, + "grad_norm": 6.327695478768039, + "learning_rate": 1.4277324220099862e-07, + "loss": 0.6854, + "num_input_tokens_seen": 156778905, + "step": 7340 + }, + { + "epoch": 0.8827030601815667, + "flos": 16432302748680.0, + "grad_norm": 4.725031227937067, + "learning_rate": 1.4248434660366938e-07, + "loss": 0.7283, + "num_input_tokens_seen": 156798100, + "step": 7341 + }, + { + "epoch": 0.8828233030722058, + "flos": 14515136157840.0, + "grad_norm": 3.8993704067569417, + "learning_rate": 1.4219573278998808e-07, + "loss": 0.6891, + "num_input_tokens_seen": 156816280, + "step": 7342 + }, + { + "epoch": 0.882943545962845, + "flos": 28932551493240.0, + "grad_norm": 4.794576187258858, + "learning_rate": 1.4190740080373685e-07, + "loss": 0.645, + "num_input_tokens_seen": 156836280, + "step": 7343 + }, + { + "epoch": 0.883063788853484, + "flos": 13913664067800.0, + "grad_norm": 4.107985577105765, + "learning_rate": 1.4161935068865538e-07, + "loss": 0.8355, + "num_input_tokens_seen": 156851145, + "step": 7344 + }, + { + "epoch": 0.8831840317441231, + "flos": 13676856329280.0, + "grad_norm": 6.041223373812205, + "learning_rate": 1.4133158248844113e-07, + "loss": 0.7412, + "num_input_tokens_seen": 156869770, + "step": 7345 + }, + { + "epoch": 0.8833042746347622, + "flos": 19666588538520.0, + "grad_norm": 3.035536419442166, + "learning_rate": 1.4104409624674785e-07, + "loss": 0.7159, + "num_input_tokens_seen": 156889275, + "step": 7346 + }, + { + "epoch": 0.8834245175254013, + "flos": 19131909981360.0, + "grad_norm": 3.887603689693706, + "learning_rate": 1.407568920071873e-07, + "loss": 0.774, + "num_input_tokens_seen": 156907860, + "step": 7347 + }, + { + "epoch": 0.8835447604160404, + "flos": 22480786860960.0, + "grad_norm": 7.481048035664125, + "learning_rate": 1.4046996981332782e-07, + "loss": 0.6525, + "num_input_tokens_seen": 156927465, + "step": 7348 + }, + { + "epoch": 0.8836650033066795, + "flos": 18106487908800.0, + "grad_norm": 3.389061070499497, + "learning_rate": 1.4018332970869516e-07, + "loss": 0.7629, + "num_input_tokens_seen": 156945125, + "step": 7349 + }, + { + "epoch": 0.8837852461973186, + "flos": 18621098542680.0, + "grad_norm": 3.8115837371927594, + "learning_rate": 1.398969717367733e-07, + "loss": 0.8254, + "num_input_tokens_seen": 156966170, + "step": 7350 + }, + { + "epoch": 0.8839054890879576, + "flos": 13003149684600.0, + "grad_norm": 4.0996019309722795, + "learning_rate": 1.396108959410014e-07, + "loss": 0.7475, + "num_input_tokens_seen": 156985105, + "step": 7351 + }, + { + "epoch": 0.8840257319785968, + "flos": 17403816892080.0, + "grad_norm": 4.517449139222695, + "learning_rate": 1.3932510236477745e-07, + "loss": 0.7839, + "num_input_tokens_seen": 157005495, + "step": 7352 + }, + { + "epoch": 0.8841459748692359, + "flos": 21319376057040.0, + "grad_norm": 5.4921104024193905, + "learning_rate": 1.3903959105145636e-07, + "loss": 0.5329, + "num_input_tokens_seen": 157025705, + "step": 7353 + }, + { + "epoch": 0.8842662177598749, + "flos": 17804627935200.0, + "grad_norm": 3.1078432230248305, + "learning_rate": 1.387543620443492e-07, + "loss": 0.8164, + "num_input_tokens_seen": 157042270, + "step": 7354 + }, + { + "epoch": 0.8843864606505141, + "flos": 18320409987600.0, + "grad_norm": 4.867960811798916, + "learning_rate": 1.3846941538672606e-07, + "loss": 0.827, + "num_input_tokens_seen": 157060695, + "step": 7355 + }, + { + "epoch": 0.8845067035411531, + "flos": 20671151188800.0, + "grad_norm": 3.6199060362533872, + "learning_rate": 1.3818475112181193e-07, + "loss": 0.7931, + "num_input_tokens_seen": 157079210, + "step": 7356 + }, + { + "epoch": 0.8846269464317922, + "flos": 9324461578080.0, + "grad_norm": 11.624382412820571, + "learning_rate": 1.3790036929279091e-07, + "loss": 0.7803, + "num_input_tokens_seen": 157096085, + "step": 7357 + }, + { + "epoch": 0.8847471893224313, + "flos": 13596390185040.0, + "grad_norm": 7.7478100543590855, + "learning_rate": 1.3761626994280363e-07, + "loss": 0.5732, + "num_input_tokens_seen": 157113275, + "step": 7358 + }, + { + "epoch": 0.8848674322130704, + "flos": 26284446032760.0, + "grad_norm": 5.764909074369013, + "learning_rate": 1.3733245311494735e-07, + "loss": 0.7217, + "num_input_tokens_seen": 157135650, + "step": 7359 + }, + { + "epoch": 0.8849876751037095, + "flos": 17755215720360.0, + "grad_norm": 5.999909400255915, + "learning_rate": 1.3704891885227676e-07, + "loss": 0.701, + "num_input_tokens_seen": 157155415, + "step": 7360 + }, + { + "epoch": 0.8851079179943486, + "flos": 15724122898920.0, + "grad_norm": 4.831144802038809, + "learning_rate": 1.367656671978037e-07, + "loss": 0.7606, + "num_input_tokens_seen": 157172600, + "step": 7361 + }, + { + "epoch": 0.8852281608849877, + "flos": 11136186807600.0, + "grad_norm": 3.307203124455571, + "learning_rate": 1.36482698194498e-07, + "loss": 0.7265, + "num_input_tokens_seen": 157188865, + "step": 7362 + }, + { + "epoch": 0.8853484037756267, + "flos": 17053684462200.0, + "grad_norm": 5.125121324545264, + "learning_rate": 1.3620001188528506e-07, + "loss": 0.6962, + "num_input_tokens_seen": 157209305, + "step": 7363 + }, + { + "epoch": 0.8854686466662659, + "flos": 18399166494000.0, + "grad_norm": 3.5529307289660945, + "learning_rate": 1.3591760831304865e-07, + "loss": 0.711, + "num_input_tokens_seen": 157226715, + "step": 7364 + }, + { + "epoch": 0.885588889556905, + "flos": 15642896915640.0, + "grad_norm": 4.794537599163752, + "learning_rate": 1.356354875206287e-07, + "loss": 0.7821, + "num_input_tokens_seen": 157244270, + "step": 7365 + }, + { + "epoch": 0.885709132447544, + "flos": 19725941980800.0, + "grad_norm": 3.6379468505735013, + "learning_rate": 1.3535364955082296e-07, + "loss": 0.6744, + "num_input_tokens_seen": 157263840, + "step": 7366 + }, + { + "epoch": 0.8858293753381832, + "flos": 19131308442120.0, + "grad_norm": 3.405681037630818, + "learning_rate": 1.3507209444638613e-07, + "loss": 0.6225, + "num_input_tokens_seen": 157285560, + "step": 7367 + }, + { + "epoch": 0.8859496182288222, + "flos": 17051120005440.0, + "grad_norm": 4.023942578658743, + "learning_rate": 1.347908222500298e-07, + "loss": 0.7224, + "num_input_tokens_seen": 157305355, + "step": 7368 + }, + { + "epoch": 0.8860698611194613, + "flos": 12300193728240.0, + "grad_norm": 4.866750143371633, + "learning_rate": 1.3450983300442276e-07, + "loss": 0.6756, + "num_input_tokens_seen": 157324305, + "step": 7369 + }, + { + "epoch": 0.8861901040101005, + "flos": 18079074874800.0, + "grad_norm": 3.1284473477944874, + "learning_rate": 1.3422912675219068e-07, + "loss": 0.7203, + "num_input_tokens_seen": 157343780, + "step": 7370 + }, + { + "epoch": 0.8863103469007395, + "flos": 17887436916480.0, + "grad_norm": 5.747701102410118, + "learning_rate": 1.339487035359166e-07, + "loss": 0.7684, + "num_input_tokens_seen": 157363870, + "step": 7371 + }, + { + "epoch": 0.8864305897913786, + "flos": 16134526909920.0, + "grad_norm": 8.944978159690708, + "learning_rate": 1.336685633981409e-07, + "loss": 0.8395, + "num_input_tokens_seen": 157384675, + "step": 7372 + }, + { + "epoch": 0.8865508326820177, + "flos": 13948105612920.0, + "grad_norm": 9.560373107141473, + "learning_rate": 1.333887063813597e-07, + "loss": 0.7267, + "num_input_tokens_seen": 157402500, + "step": 7373 + }, + { + "epoch": 0.8866710755726568, + "flos": 11220737086680.0, + "grad_norm": 10.665443548966513, + "learning_rate": 1.331091325280278e-07, + "loss": 0.6432, + "num_input_tokens_seen": 157421190, + "step": 7374 + }, + { + "epoch": 0.8867913184632958, + "flos": 14676385045920.0, + "grad_norm": 5.46878086626334, + "learning_rate": 1.3282984188055625e-07, + "loss": 0.7706, + "num_input_tokens_seen": 157440700, + "step": 7375 + }, + { + "epoch": 0.8869115613539349, + "flos": 17105186234400.0, + "grad_norm": 6.207587417066903, + "learning_rate": 1.3255083448131288e-07, + "loss": 0.7758, + "num_input_tokens_seen": 157459465, + "step": 7376 + }, + { + "epoch": 0.8870318042445741, + "flos": 15566673206040.0, + "grad_norm": 11.190063864736207, + "learning_rate": 1.3227211037262365e-07, + "loss": 0.7668, + "num_input_tokens_seen": 157476425, + "step": 7377 + }, + { + "epoch": 0.8871520471352131, + "flos": 14622508776720.0, + "grad_norm": 7.493499990942671, + "learning_rate": 1.319936695967696e-07, + "loss": 0.8426, + "num_input_tokens_seen": 157493970, + "step": 7378 + }, + { + "epoch": 0.8872722900258522, + "flos": 16539707027520.0, + "grad_norm": 10.151079663511707, + "learning_rate": 1.3171551219599097e-07, + "loss": 0.8061, + "num_input_tokens_seen": 157512215, + "step": 7379 + }, + { + "epoch": 0.8873925329164913, + "flos": 16215404633640.0, + "grad_norm": 4.846106171519024, + "learning_rate": 1.3143763821248377e-07, + "loss": 0.7569, + "num_input_tokens_seen": 157529020, + "step": 7380 + }, + { + "epoch": 0.8875127758071304, + "flos": 14028508437240.0, + "grad_norm": 5.8031350443336605, + "learning_rate": 1.3116004768840118e-07, + "loss": 0.7133, + "num_input_tokens_seen": 157547115, + "step": 7381 + }, + { + "epoch": 0.8876330186977694, + "flos": 13216026984720.0, + "grad_norm": 3.490120295115548, + "learning_rate": 1.3088274066585348e-07, + "loss": 0.7192, + "num_input_tokens_seen": 157564445, + "step": 7382 + }, + { + "epoch": 0.8877532615884086, + "flos": 16100433624360.0, + "grad_norm": 5.870855614660725, + "learning_rate": 1.3060571718690749e-07, + "loss": 0.8913, + "num_input_tokens_seen": 157581660, + "step": 7383 + }, + { + "epoch": 0.8878735044790477, + "flos": 53202747803280.0, + "grad_norm": 0.7721418066333972, + "learning_rate": 1.3032897729358805e-07, + "loss": 0.5917, + "num_input_tokens_seen": 157642335, + "step": 7384 + }, + { + "epoch": 0.8879937473696867, + "flos": 20184396828360.0, + "grad_norm": 7.947262984992803, + "learning_rate": 1.3005252102787645e-07, + "loss": 0.7852, + "num_input_tokens_seen": 157660995, + "step": 7385 + }, + { + "epoch": 0.8881139902603259, + "flos": 16267507945080.0, + "grad_norm": 18.644302504227415, + "learning_rate": 1.297763484317105e-07, + "loss": 0.7153, + "num_input_tokens_seen": 157679010, + "step": 7386 + }, + { + "epoch": 0.888234233150965, + "flos": 14835702676440.0, + "grad_norm": 8.969137469172717, + "learning_rate": 1.2950045954698551e-07, + "loss": 0.6806, + "num_input_tokens_seen": 157696565, + "step": 7387 + }, + { + "epoch": 0.888354476041604, + "flos": 13243566658560.0, + "grad_norm": 3.3275269163491856, + "learning_rate": 1.2922485441555343e-07, + "loss": 0.7473, + "num_input_tokens_seen": 157715365, + "step": 7388 + }, + { + "epoch": 0.8884747189322432, + "flos": 16104707718960.0, + "grad_norm": 70.34298297560497, + "learning_rate": 1.2894953307922363e-07, + "loss": 0.8076, + "num_input_tokens_seen": 157734045, + "step": 7389 + }, + { + "epoch": 0.8885949618228822, + "flos": 14455941015360.0, + "grad_norm": 3.812887953733953, + "learning_rate": 1.2867449557976208e-07, + "loss": 0.8323, + "num_input_tokens_seen": 157751865, + "step": 7390 + }, + { + "epoch": 0.8887152047135213, + "flos": 14649731850960.0, + "grad_norm": 9.728207221634946, + "learning_rate": 1.283997419588916e-07, + "loss": 0.7378, + "num_input_tokens_seen": 157771055, + "step": 7391 + }, + { + "epoch": 0.8888354476041604, + "flos": 13569610350240.0, + "grad_norm": 4.457087951887898, + "learning_rate": 1.2812527225829216e-07, + "loss": 0.6059, + "num_input_tokens_seen": 157789000, + "step": 7392 + }, + { + "epoch": 0.8889556904947995, + "flos": 15865113903960.0, + "grad_norm": 4.40303907344409, + "learning_rate": 1.2785108651960052e-07, + "loss": 0.7398, + "num_input_tokens_seen": 157810355, + "step": 7393 + }, + { + "epoch": 0.8890759333854386, + "flos": 20077087529400.0, + "grad_norm": 3.667222732225168, + "learning_rate": 1.2757718478441094e-07, + "loss": 0.7855, + "num_input_tokens_seen": 157830820, + "step": 7394 + }, + { + "epoch": 0.8891961762760777, + "flos": 17943940962360.0, + "grad_norm": 3.1927438125174032, + "learning_rate": 1.2730356709427302e-07, + "loss": 0.7502, + "num_input_tokens_seen": 157849220, + "step": 7395 + }, + { + "epoch": 0.8893164191667168, + "flos": 30525573990000.0, + "grad_norm": 4.335268811049956, + "learning_rate": 1.2703023349069542e-07, + "loss": 0.5754, + "num_input_tokens_seen": 157873790, + "step": 7396 + }, + { + "epoch": 0.8894366620573558, + "flos": 24667271477880.0, + "grad_norm": 3.4638680134619824, + "learning_rate": 1.2675718401514223e-07, + "loss": 0.6022, + "num_input_tokens_seen": 157897690, + "step": 7397 + }, + { + "epoch": 0.889556904947995, + "flos": 12327575102280.0, + "grad_norm": 4.270689699812545, + "learning_rate": 1.264844187090346e-07, + "loss": 0.7271, + "num_input_tokens_seen": 157914535, + "step": 7398 + }, + { + "epoch": 0.889677147838634, + "flos": 19077685452600.0, + "grad_norm": 6.3669011151085355, + "learning_rate": 1.262119376137516e-07, + "loss": 0.7357, + "num_input_tokens_seen": 157935315, + "step": 7399 + }, + { + "epoch": 0.8897973907292731, + "flos": 19401354647280.0, + "grad_norm": 3.0702865638316905, + "learning_rate": 1.2593974077062707e-07, + "loss": 0.8246, + "num_input_tokens_seen": 157956655, + "step": 7400 + }, + { + "epoch": 0.8899176336199123, + "flos": 19239630859800.0, + "grad_norm": 4.293621593259983, + "learning_rate": 1.2566782822095423e-07, + "loss": 0.6206, + "num_input_tokens_seen": 157976630, + "step": 7401 + }, + { + "epoch": 0.8900378765105513, + "flos": 15214957778160.0, + "grad_norm": 3.4753485799816355, + "learning_rate": 1.2539620000598162e-07, + "loss": 0.6975, + "num_input_tokens_seen": 157995685, + "step": 7402 + }, + { + "epoch": 0.8901581194011904, + "flos": 12003842587680.0, + "grad_norm": 4.861002266285927, + "learning_rate": 1.2512485616691492e-07, + "loss": 0.7775, + "num_input_tokens_seen": 158012460, + "step": 7403 + }, + { + "epoch": 0.8902783622918296, + "flos": 25830771839160.0, + "grad_norm": 2.6289983906122494, + "learning_rate": 1.2485379674491681e-07, + "loss": 0.794, + "num_input_tokens_seen": 158038375, + "step": 7404 + }, + { + "epoch": 0.8903986051824686, + "flos": 12543016859160.0, + "grad_norm": 5.388604623085233, + "learning_rate": 1.2458302178110657e-07, + "loss": 0.7597, + "num_input_tokens_seen": 158056460, + "step": 7405 + }, + { + "epoch": 0.8905188480731077, + "flos": 18672916914480.0, + "grad_norm": 4.306333915878325, + "learning_rate": 1.2431253131656118e-07, + "loss": 0.8067, + "num_input_tokens_seen": 158075655, + "step": 7406 + }, + { + "epoch": 0.8906390909637467, + "flos": 17105217894360.0, + "grad_norm": 6.166691788630821, + "learning_rate": 1.240423253923133e-07, + "loss": 0.751, + "num_input_tokens_seen": 158094980, + "step": 7407 + }, + { + "epoch": 0.8907593338543859, + "flos": 15405424317960.0, + "grad_norm": 4.969445329873011, + "learning_rate": 1.237724040493533e-07, + "loss": 0.6779, + "num_input_tokens_seen": 158113325, + "step": 7408 + }, + { + "epoch": 0.8908795767450249, + "flos": 15997778339520.0, + "grad_norm": 4.412231713803012, + "learning_rate": 1.2350276732862773e-07, + "loss": 0.6967, + "num_input_tokens_seen": 158134070, + "step": 7409 + }, + { + "epoch": 0.890999819635664, + "flos": 48890081810520.0, + "grad_norm": 1.1625462840721066, + "learning_rate": 1.2323341527103993e-07, + "loss": 0.6116, + "num_input_tokens_seen": 158188990, + "step": 7410 + }, + { + "epoch": 0.8911200625263032, + "flos": 19698623926680.0, + "grad_norm": 7.404221655348603, + "learning_rate": 1.2296434791745135e-07, + "loss": 0.8247, + "num_input_tokens_seen": 158207160, + "step": 7411 + }, + { + "epoch": 0.8912403054169422, + "flos": 15269942145960.0, + "grad_norm": 4.558505465217302, + "learning_rate": 1.2269556530867875e-07, + "loss": 0.7507, + "num_input_tokens_seen": 158225435, + "step": 7412 + }, + { + "epoch": 0.8913605483075813, + "flos": 19807262943960.0, + "grad_norm": 6.751949735137997, + "learning_rate": 1.2242706748549614e-07, + "loss": 0.8029, + "num_input_tokens_seen": 158243150, + "step": 7413 + }, + { + "epoch": 0.8914807911982204, + "flos": 17294418035760.0, + "grad_norm": 2.9077925331377212, + "learning_rate": 1.2215885448863473e-07, + "loss": 0.7996, + "num_input_tokens_seen": 158263745, + "step": 7414 + }, + { + "epoch": 0.8916010340888595, + "flos": 17916876187920.0, + "grad_norm": 3.915286544700377, + "learning_rate": 1.2189092635878152e-07, + "loss": 0.783, + "num_input_tokens_seen": 158284915, + "step": 7415 + }, + { + "epoch": 0.8917212769794985, + "flos": 15514348274880.0, + "grad_norm": 3.977713247896977, + "learning_rate": 1.216232831365822e-07, + "loss": 0.7601, + "num_input_tokens_seen": 158303580, + "step": 7416 + }, + { + "epoch": 0.8918415198701377, + "flos": 18694314556080.0, + "grad_norm": 5.82933608657153, + "learning_rate": 1.2135592486263678e-07, + "loss": 0.795, + "num_input_tokens_seen": 158322550, + "step": 7417 + }, + { + "epoch": 0.8919617627607768, + "flos": 27826979876040.0, + "grad_norm": 3.4535827333341658, + "learning_rate": 1.2108885157750415e-07, + "loss": 0.5911, + "num_input_tokens_seen": 158344630, + "step": 7418 + }, + { + "epoch": 0.8920820056514158, + "flos": 19213040984760.0, + "grad_norm": 3.871843960124206, + "learning_rate": 1.2082206332169897e-07, + "loss": 0.7828, + "num_input_tokens_seen": 158364445, + "step": 7419 + }, + { + "epoch": 0.892202248542055, + "flos": 12677707532160.0, + "grad_norm": 6.635168656805043, + "learning_rate": 1.2055556013569225e-07, + "loss": 0.7219, + "num_input_tokens_seen": 158379675, + "step": 7420 + }, + { + "epoch": 0.892322491432694, + "flos": 15594466159560.0, + "grad_norm": 3.2874756380636216, + "learning_rate": 1.2028934205991315e-07, + "loss": 0.8008, + "num_input_tokens_seen": 158398715, + "step": 7421 + }, + { + "epoch": 0.8924427343233331, + "flos": 17597196148200.0, + "grad_norm": 2.900392597852864, + "learning_rate": 1.2002340913474607e-07, + "loss": 0.7538, + "num_input_tokens_seen": 158422070, + "step": 7422 + }, + { + "epoch": 0.8925629772139723, + "flos": 22021128934920.0, + "grad_norm": 5.06521194722516, + "learning_rate": 1.1975776140053317e-07, + "loss": 0.7308, + "num_input_tokens_seen": 158441760, + "step": 7423 + }, + { + "epoch": 0.8926832201046113, + "flos": 16539960307200.0, + "grad_norm": 5.407500546861428, + "learning_rate": 1.194923988975729e-07, + "loss": 0.706, + "num_input_tokens_seen": 158461080, + "step": 7424 + }, + { + "epoch": 0.8928034629952504, + "flos": 9654367784880.0, + "grad_norm": 8.069706498000233, + "learning_rate": 1.192273216661206e-07, + "loss": 0.7118, + "num_input_tokens_seen": 158478890, + "step": 7425 + }, + { + "epoch": 0.8929237058858895, + "flos": 40414316187840.0, + "grad_norm": 0.8555594966606369, + "learning_rate": 1.189625297463881e-07, + "loss": 0.6, + "num_input_tokens_seen": 158540300, + "step": 7426 + }, + { + "epoch": 0.8930439487765286, + "flos": 21190099237200.0, + "grad_norm": 3.4421952519825187, + "learning_rate": 1.1869802317854394e-07, + "loss": 0.7754, + "num_input_tokens_seen": 158563805, + "step": 7427 + }, + { + "epoch": 0.8931641916671677, + "flos": 16404953034600.0, + "grad_norm": 8.642576580155229, + "learning_rate": 1.1843380200271425e-07, + "loss": 0.7082, + "num_input_tokens_seen": 158582725, + "step": 7428 + }, + { + "epoch": 0.8932844345578068, + "flos": 18938024165880.0, + "grad_norm": 4.317133087135607, + "learning_rate": 1.181698662589805e-07, + "loss": 0.7897, + "num_input_tokens_seen": 158602030, + "step": 7429 + }, + { + "epoch": 0.8934046774484459, + "flos": 16779205862640.0, + "grad_norm": 5.503490727417243, + "learning_rate": 1.1790621598738249e-07, + "loss": 0.743, + "num_input_tokens_seen": 158620065, + "step": 7430 + }, + { + "epoch": 0.8935249203390849, + "flos": 17916717888120.0, + "grad_norm": 4.214882214983194, + "learning_rate": 1.1764285122791461e-07, + "loss": 0.7399, + "num_input_tokens_seen": 158640505, + "step": 7431 + }, + { + "epoch": 0.8936451632297241, + "flos": 11463908477160.0, + "grad_norm": 3.547512719720964, + "learning_rate": 1.173797720205294e-07, + "loss": 0.7567, + "num_input_tokens_seen": 158658260, + "step": 7432 + }, + { + "epoch": 0.8937654061203631, + "flos": 25800889328280.0, + "grad_norm": 5.4618226563232515, + "learning_rate": 1.1711697840513602e-07, + "loss": 0.7023, + "num_input_tokens_seen": 158677415, + "step": 7433 + }, + { + "epoch": 0.8938856490110022, + "flos": 11733796382520.0, + "grad_norm": 10.362218130072119, + "learning_rate": 1.1685447042160012e-07, + "loss": 0.6912, + "num_input_tokens_seen": 158695170, + "step": 7434 + }, + { + "epoch": 0.8940058919016414, + "flos": 15135631392480.0, + "grad_norm": 4.071510534011235, + "learning_rate": 1.1659224810974367e-07, + "loss": 0.7002, + "num_input_tokens_seen": 158714850, + "step": 7435 + }, + { + "epoch": 0.8941261347922804, + "flos": 18484413292200.0, + "grad_norm": 4.750087737226813, + "learning_rate": 1.1633031150934591e-07, + "loss": 0.6696, + "num_input_tokens_seen": 158737600, + "step": 7436 + }, + { + "epoch": 0.8942463776829195, + "flos": 14272091407200.0, + "grad_norm": 5.9594548732643835, + "learning_rate": 1.1606866066014176e-07, + "loss": 0.777, + "num_input_tokens_seen": 158756370, + "step": 7437 + }, + { + "epoch": 0.8943666205735585, + "flos": 16316666880240.0, + "grad_norm": 4.354737142520124, + "learning_rate": 1.1580729560182434e-07, + "loss": 0.7338, + "num_input_tokens_seen": 158771945, + "step": 7438 + }, + { + "epoch": 0.8944868634641977, + "flos": 13809457444920.0, + "grad_norm": 6.1396958907671895, + "learning_rate": 1.1554621637404171e-07, + "loss": 0.7084, + "num_input_tokens_seen": 158789755, + "step": 7439 + }, + { + "epoch": 0.8946071063548368, + "flos": 10514583474360.0, + "grad_norm": 5.102823052055199, + "learning_rate": 1.1528542301639999e-07, + "loss": 0.5881, + "num_input_tokens_seen": 158806265, + "step": 7440 + }, + { + "epoch": 0.8947273492454758, + "flos": 14676733305480.0, + "grad_norm": 4.662849818138439, + "learning_rate": 1.1502491556846105e-07, + "loss": 0.802, + "num_input_tokens_seen": 158824480, + "step": 7441 + }, + { + "epoch": 0.894847592136115, + "flos": 13541659096920.0, + "grad_norm": 4.754852025309807, + "learning_rate": 1.1476469406974331e-07, + "loss": 0.7964, + "num_input_tokens_seen": 158839800, + "step": 7442 + }, + { + "epoch": 0.894967835026754, + "flos": 17188976674440.0, + "grad_norm": 3.6229621157538343, + "learning_rate": 1.1450475855972341e-07, + "loss": 0.7611, + "num_input_tokens_seen": 158860310, + "step": 7443 + }, + { + "epoch": 0.8950880779173931, + "flos": 11052586327320.0, + "grad_norm": 5.658183236394973, + "learning_rate": 1.1424510907783158e-07, + "loss": 0.6874, + "num_input_tokens_seen": 158877310, + "step": 7444 + }, + { + "epoch": 0.8952083208080323, + "flos": 16162858082760.0, + "grad_norm": 3.7489216843608766, + "learning_rate": 1.1398574566345787e-07, + "loss": 0.8103, + "num_input_tokens_seen": 158897665, + "step": 7445 + }, + { + "epoch": 0.8953285636986713, + "flos": 17022092313480.0, + "grad_norm": 4.152382334604362, + "learning_rate": 1.1372666835594702e-07, + "loss": 0.8017, + "num_input_tokens_seen": 158915710, + "step": 7446 + }, + { + "epoch": 0.8954488065893104, + "flos": 11922648264360.0, + "grad_norm": 5.141732789309471, + "learning_rate": 1.1346787719460071e-07, + "loss": 0.6958, + "num_input_tokens_seen": 158934315, + "step": 7447 + }, + { + "epoch": 0.8955690494799495, + "flos": 13324950941640.0, + "grad_norm": 2.959239140486686, + "learning_rate": 1.1320937221867732e-07, + "loss": 0.7176, + "num_input_tokens_seen": 158951615, + "step": 7448 + }, + { + "epoch": 0.8956892923705886, + "flos": 18646010439840.0, + "grad_norm": 21.413719756241047, + "learning_rate": 1.1295115346739192e-07, + "loss": 0.7841, + "num_input_tokens_seen": 158971335, + "step": 7449 + }, + { + "epoch": 0.8958095352612276, + "flos": 38785929515760.0, + "grad_norm": 4.120052491751778, + "learning_rate": 1.1269322097991629e-07, + "loss": 0.7191, + "num_input_tokens_seen": 158994340, + "step": 7450 + }, + { + "epoch": 0.8959297781518668, + "flos": 16972743418560.0, + "grad_norm": 4.259900622216113, + "learning_rate": 1.1243557479537846e-07, + "loss": 0.6683, + "num_input_tokens_seen": 159013950, + "step": 7451 + }, + { + "epoch": 0.8960500210425059, + "flos": 14861184452880.0, + "grad_norm": 4.243803555989152, + "learning_rate": 1.121782149528634e-07, + "loss": 0.6885, + "num_input_tokens_seen": 159030770, + "step": 7452 + }, + { + "epoch": 0.8961702639331449, + "flos": 14541757692840.0, + "grad_norm": 13.52331846152837, + "learning_rate": 1.1192114149141208e-07, + "loss": 0.7683, + "num_input_tokens_seen": 159050125, + "step": 7453 + }, + { + "epoch": 0.8962905068237841, + "flos": 9357320125200.0, + "grad_norm": 5.044572334423548, + "learning_rate": 1.1166435445002197e-07, + "loss": 0.6403, + "num_input_tokens_seen": 159067515, + "step": 7454 + }, + { + "epoch": 0.8964107497144231, + "flos": 17160518861760.0, + "grad_norm": 4.972799893901828, + "learning_rate": 1.1140785386764818e-07, + "loss": 0.6654, + "num_input_tokens_seen": 159085935, + "step": 7455 + }, + { + "epoch": 0.8965309926050622, + "flos": 14244583393320.0, + "grad_norm": 5.5111271400799655, + "learning_rate": 1.1115163978320153e-07, + "loss": 0.6799, + "num_input_tokens_seen": 159104385, + "step": 7456 + }, + { + "epoch": 0.8966512354957014, + "flos": 21021980137800.0, + "grad_norm": 3.77540464443045, + "learning_rate": 1.1089571223554917e-07, + "loss": 0.8016, + "num_input_tokens_seen": 159124990, + "step": 7457 + }, + { + "epoch": 0.8967714783863404, + "flos": 17109650288760.0, + "grad_norm": 3.768748906832864, + "learning_rate": 1.1064007126351537e-07, + "loss": 0.8339, + "num_input_tokens_seen": 159145425, + "step": 7458 + }, + { + "epoch": 0.8968917212769795, + "flos": 17970815777040.0, + "grad_norm": 4.002671779970702, + "learning_rate": 1.1038471690588003e-07, + "loss": 0.7385, + "num_input_tokens_seen": 159164290, + "step": 7459 + }, + { + "epoch": 0.8970119641676186, + "flos": 17186285577840.0, + "grad_norm": 4.497913724690132, + "learning_rate": 1.1012964920138145e-07, + "loss": 0.7833, + "num_input_tokens_seen": 159183595, + "step": 7460 + }, + { + "epoch": 0.8971322070582577, + "flos": 17973475213680.0, + "grad_norm": 3.4912461755214657, + "learning_rate": 1.0987486818871205e-07, + "loss": 0.7413, + "num_input_tokens_seen": 159206905, + "step": 7461 + }, + { + "epoch": 0.8972524499488967, + "flos": 15944535269520.0, + "grad_norm": 5.879962205311244, + "learning_rate": 1.0962037390652245e-07, + "loss": 0.697, + "num_input_tokens_seen": 159225645, + "step": 7462 + }, + { + "epoch": 0.8973726928395359, + "flos": 15892083698520.0, + "grad_norm": 3.849044750656156, + "learning_rate": 1.0936616639341911e-07, + "loss": 0.7021, + "num_input_tokens_seen": 159245655, + "step": 7463 + }, + { + "epoch": 0.897492935730175, + "flos": 39661916356680.0, + "grad_norm": 0.7435252531261365, + "learning_rate": 1.0911224568796473e-07, + "loss": 0.5561, + "num_input_tokens_seen": 159303570, + "step": 7464 + }, + { + "epoch": 0.897613178620814, + "flos": 13348406480640.0, + "grad_norm": 4.356136965277138, + "learning_rate": 1.0885861182867984e-07, + "loss": 0.6871, + "num_input_tokens_seen": 159321395, + "step": 7465 + }, + { + "epoch": 0.8977334215114532, + "flos": 24229834352400.0, + "grad_norm": 12.86207005381855, + "learning_rate": 1.0860526485403942e-07, + "loss": 0.6897, + "num_input_tokens_seen": 159342390, + "step": 7466 + }, + { + "epoch": 0.8978536644020922, + "flos": 11059044959160.0, + "grad_norm": 4.149522657403109, + "learning_rate": 1.0835220480247675e-07, + "loss": 0.7618, + "num_input_tokens_seen": 159360605, + "step": 7467 + }, + { + "epoch": 0.8979739072927313, + "flos": 13137618737880.0, + "grad_norm": 3.64629776675872, + "learning_rate": 1.0809943171238067e-07, + "loss": 0.8174, + "num_input_tokens_seen": 159378250, + "step": 7468 + }, + { + "epoch": 0.8980941501833704, + "flos": 16295364218520.0, + "grad_norm": 13.75558358310959, + "learning_rate": 1.078469456220965e-07, + "loss": 0.618, + "num_input_tokens_seen": 159398125, + "step": 7469 + }, + { + "epoch": 0.8982143930740095, + "flos": 27610619980320.0, + "grad_norm": 4.735934060604265, + "learning_rate": 1.0759474656992606e-07, + "loss": 0.6746, + "num_input_tokens_seen": 159420615, + "step": 7470 + }, + { + "epoch": 0.8983346359646486, + "flos": 13191210067440.0, + "grad_norm": 6.149469194533112, + "learning_rate": 1.0734283459412785e-07, + "loss": 0.765, + "num_input_tokens_seen": 159437185, + "step": 7471 + }, + { + "epoch": 0.8984548788552876, + "flos": 15028195453680.0, + "grad_norm": 5.154037093207061, + "learning_rate": 1.0709120973291707e-07, + "loss": 0.7793, + "num_input_tokens_seen": 159456685, + "step": 7472 + }, + { + "epoch": 0.8985751217459268, + "flos": 12975578350800.0, + "grad_norm": 4.831919934803106, + "learning_rate": 1.0683987202446475e-07, + "loss": 0.751, + "num_input_tokens_seen": 159474590, + "step": 7473 + }, + { + "epoch": 0.8986953646365659, + "flos": 15811269294720.0, + "grad_norm": 4.0556209129281156, + "learning_rate": 1.0658882150689862e-07, + "loss": 0.6913, + "num_input_tokens_seen": 159493170, + "step": 7474 + }, + { + "epoch": 0.8988156075272049, + "flos": 10194333555360.0, + "grad_norm": 5.660772549371057, + "learning_rate": 1.0633805821830288e-07, + "loss": 0.7639, + "num_input_tokens_seen": 159509575, + "step": 7475 + }, + { + "epoch": 0.8989358504178441, + "flos": 21318869497680.0, + "grad_norm": 5.331236347570078, + "learning_rate": 1.0608758219671753e-07, + "loss": 0.8154, + "num_input_tokens_seen": 159528335, + "step": 7476 + }, + { + "epoch": 0.8990560933084831, + "flos": 14785087383120.0, + "grad_norm": 5.8722017807345575, + "learning_rate": 1.0583739348014065e-07, + "loss": 0.689, + "num_input_tokens_seen": 159549140, + "step": 7477 + }, + { + "epoch": 0.8991763361991222, + "flos": 18700203308640.0, + "grad_norm": 3.4231558018319475, + "learning_rate": 1.0558749210652518e-07, + "loss": 0.8349, + "num_input_tokens_seen": 159568790, + "step": 7478 + }, + { + "epoch": 0.8992965790897613, + "flos": 18403060669080.0, + "grad_norm": 2.735337442227478, + "learning_rate": 1.053378781137808e-07, + "loss": 0.8378, + "num_input_tokens_seen": 159589430, + "step": 7479 + }, + { + "epoch": 0.8994168219804004, + "flos": 11706668288160.0, + "grad_norm": 6.003703881491063, + "learning_rate": 1.0508855153977392e-07, + "loss": 0.766, + "num_input_tokens_seen": 159605615, + "step": 7480 + }, + { + "epoch": 0.8995370648710395, + "flos": 18185687654640.0, + "grad_norm": 7.473062076506284, + "learning_rate": 1.0483951242232669e-07, + "loss": 0.6403, + "num_input_tokens_seen": 159625810, + "step": 7481 + }, + { + "epoch": 0.8996573077616786, + "flos": 46530765251520.0, + "grad_norm": 1.0222739091420165, + "learning_rate": 1.0459076079921936e-07, + "loss": 0.5994, + "num_input_tokens_seen": 159678190, + "step": 7482 + }, + { + "epoch": 0.8997775506523177, + "flos": 13296429809040.0, + "grad_norm": 4.007921005475634, + "learning_rate": 1.0434229670818618e-07, + "loss": 0.8326, + "num_input_tokens_seen": 159694585, + "step": 7483 + }, + { + "epoch": 0.8998977935429567, + "flos": 17697666895800.0, + "grad_norm": 2.389443099128876, + "learning_rate": 1.0409412018691944e-07, + "loss": 0.7871, + "num_input_tokens_seen": 159714770, + "step": 7484 + }, + { + "epoch": 0.9000180364335959, + "flos": 15185993406120.0, + "grad_norm": 3.3956108878516655, + "learning_rate": 1.0384623127306724e-07, + "loss": 0.7425, + "num_input_tokens_seen": 159731835, + "step": 7485 + }, + { + "epoch": 0.900138279324235, + "flos": 14026323900000.0, + "grad_norm": 3.4693784402216212, + "learning_rate": 1.0359863000423397e-07, + "loss": 0.776, + "num_input_tokens_seen": 159749690, + "step": 7486 + }, + { + "epoch": 0.900258522214874, + "flos": 21076141346640.0, + "grad_norm": 4.491224633561146, + "learning_rate": 1.0335131641798112e-07, + "loss": 0.6968, + "num_input_tokens_seen": 159771370, + "step": 7487 + }, + { + "epoch": 0.9003787651055132, + "flos": 42949381896600.0, + "grad_norm": 0.8659167903101826, + "learning_rate": 1.0310429055182512e-07, + "loss": 0.6175, + "num_input_tokens_seen": 159825410, + "step": 7488 + }, + { + "epoch": 0.9004990079961522, + "flos": 18726603223920.0, + "grad_norm": 4.169411378294912, + "learning_rate": 1.0285755244324024e-07, + "loss": 0.722, + "num_input_tokens_seen": 159845875, + "step": 7489 + }, + { + "epoch": 0.9006192508867913, + "flos": 17083313693400.0, + "grad_norm": 2.6435838182674143, + "learning_rate": 1.0261110212965629e-07, + "loss": 0.6748, + "num_input_tokens_seen": 159867390, + "step": 7490 + }, + { + "epoch": 0.9007394937774305, + "flos": 13164113633040.0, + "grad_norm": 3.7762160407834084, + "learning_rate": 1.023649396484596e-07, + "loss": 0.7642, + "num_input_tokens_seen": 159886165, + "step": 7491 + }, + { + "epoch": 0.9008597366680695, + "flos": 31686478234560.0, + "grad_norm": 6.275890588063167, + "learning_rate": 1.0211906503699275e-07, + "loss": 0.6604, + "num_input_tokens_seen": 159908860, + "step": 7492 + }, + { + "epoch": 0.9009799795587086, + "flos": 10868989998840.0, + "grad_norm": 4.874321327897123, + "learning_rate": 1.0187347833255455e-07, + "loss": 0.8006, + "num_input_tokens_seen": 159924485, + "step": 7493 + }, + { + "epoch": 0.9011002224493477, + "flos": 15783444681240.0, + "grad_norm": 4.405504528860297, + "learning_rate": 1.0162817957240056e-07, + "loss": 0.7808, + "num_input_tokens_seen": 159944100, + "step": 7494 + }, + { + "epoch": 0.9012204653399868, + "flos": 52486530815040.0, + "grad_norm": 0.9951875641170022, + "learning_rate": 1.0138316879374253e-07, + "loss": 0.6746, + "num_input_tokens_seen": 160013110, + "step": 7495 + }, + { + "epoch": 0.9013407082306258, + "flos": 11354288001120.0, + "grad_norm": 6.189525234275439, + "learning_rate": 1.0113844603374833e-07, + "loss": 0.7228, + "num_input_tokens_seen": 160029355, + "step": 7496 + }, + { + "epoch": 0.901460951121265, + "flos": 10950564241680.0, + "grad_norm": 8.17222416498314, + "learning_rate": 1.0089401132954178e-07, + "loss": 0.7105, + "num_input_tokens_seen": 160047055, + "step": 7497 + }, + { + "epoch": 0.9015811940119041, + "flos": 16269945762000.0, + "grad_norm": 5.4879953981941885, + "learning_rate": 1.006498647182037e-07, + "loss": 0.7046, + "num_input_tokens_seen": 160065430, + "step": 7498 + }, + { + "epoch": 0.9017014369025431, + "flos": 18295561410360.0, + "grad_norm": 3.115832500377733, + "learning_rate": 1.004060062367713e-07, + "loss": 0.7078, + "num_input_tokens_seen": 160086245, + "step": 7499 + }, + { + "epoch": 0.9018216797931822, + "flos": 13218908041080.0, + "grad_norm": 7.300709927671284, + "learning_rate": 1.0016243592223728e-07, + "loss": 0.6774, + "num_input_tokens_seen": 160106365, + "step": 7500 + }, + { + "epoch": 0.9019419226838213, + "flos": 27394260084600.0, + "grad_norm": 5.145903037315997, + "learning_rate": 9.991915381155114e-08, + "loss": 0.6513, + "num_input_tokens_seen": 160129065, + "step": 7501 + }, + { + "epoch": 0.9020621655744604, + "flos": 17161215380880.0, + "grad_norm": 3.8996753753148368, + "learning_rate": 9.967615994161871e-08, + "loss": 0.736, + "num_input_tokens_seen": 160148445, + "step": 7502 + }, + { + "epoch": 0.9021824084650995, + "flos": 16729287088440.0, + "grad_norm": 10.085051293003895, + "learning_rate": 9.943345434930161e-08, + "loss": 0.7664, + "num_input_tokens_seen": 160168415, + "step": 7503 + }, + { + "epoch": 0.9023026513557386, + "flos": 16189099698240.0, + "grad_norm": 6.452287005308811, + "learning_rate": 9.919103707141885e-08, + "loss": 0.6753, + "num_input_tokens_seen": 160187015, + "step": 7504 + }, + { + "epoch": 0.9024228942463777, + "flos": 17720995794960.0, + "grad_norm": 4.938916353297862, + "learning_rate": 9.89489081447441e-08, + "loss": 0.7474, + "num_input_tokens_seen": 160203935, + "step": 7505 + }, + { + "epoch": 0.9025431371370167, + "flos": 18320758247160.0, + "grad_norm": 4.028647122159956, + "learning_rate": 9.870706760600844e-08, + "loss": 0.8083, + "num_input_tokens_seen": 160223605, + "step": 7506 + }, + { + "epoch": 0.9026633800276559, + "flos": 13839529915560.0, + "grad_norm": 6.685006220378849, + "learning_rate": 9.846551549189918e-08, + "loss": 0.7, + "num_input_tokens_seen": 160242930, + "step": 7507 + }, + { + "epoch": 0.902783622918295, + "flos": 23803193273280.0, + "grad_norm": 5.955682800761892, + "learning_rate": 9.822425183905902e-08, + "loss": 0.6515, + "num_input_tokens_seen": 160263175, + "step": 7508 + }, + { + "epoch": 0.902903865808934, + "flos": 46973010199560.0, + "grad_norm": 0.9567444975081464, + "learning_rate": 9.798327668408823e-08, + "loss": 0.7613, + "num_input_tokens_seen": 160324530, + "step": 7509 + }, + { + "epoch": 0.9030241086995732, + "flos": 17429520288240.0, + "grad_norm": 5.412699474156512, + "learning_rate": 9.774259006354158e-08, + "loss": 0.6678, + "num_input_tokens_seen": 160344320, + "step": 7510 + }, + { + "epoch": 0.9031443515902122, + "flos": 19347858297600.0, + "grad_norm": 3.451499452787906, + "learning_rate": 9.750219201393184e-08, + "loss": 0.7435, + "num_input_tokens_seen": 160364005, + "step": 7511 + }, + { + "epoch": 0.9032645944808513, + "flos": 18270332913600.0, + "grad_norm": 3.6342965981054007, + "learning_rate": 9.726208257172697e-08, + "loss": 0.7704, + "num_input_tokens_seen": 160385420, + "step": 7512 + }, + { + "epoch": 0.9033848373714904, + "flos": 15486935240880.0, + "grad_norm": 4.463636126533252, + "learning_rate": 9.702226177335115e-08, + "loss": 0.7369, + "num_input_tokens_seen": 160403635, + "step": 7513 + }, + { + "epoch": 0.9035050802621295, + "flos": 19266790614120.0, + "grad_norm": 4.061314806082085, + "learning_rate": 9.67827296551853e-08, + "loss": 0.7123, + "num_input_tokens_seen": 160424640, + "step": 7514 + }, + { + "epoch": 0.9036253231527686, + "flos": 17725808108880.0, + "grad_norm": 3.3803638643867697, + "learning_rate": 9.65434862535659e-08, + "loss": 0.656, + "num_input_tokens_seen": 160443730, + "step": 7515 + }, + { + "epoch": 0.9037455660434077, + "flos": 13187474192160.0, + "grad_norm": 7.064257917799945, + "learning_rate": 9.630453160478635e-08, + "loss": 0.6281, + "num_input_tokens_seen": 160458805, + "step": 7516 + }, + { + "epoch": 0.9038658089340468, + "flos": 17619733548360.0, + "grad_norm": 4.570173734204798, + "learning_rate": 9.60658657450959e-08, + "loss": 0.7976, + "num_input_tokens_seen": 160478825, + "step": 7517 + }, + { + "epoch": 0.9039860518246858, + "flos": 15971853323640.0, + "grad_norm": 2.905431324491331, + "learning_rate": 9.582748871069979e-08, + "loss": 0.7755, + "num_input_tokens_seen": 160497985, + "step": 7518 + }, + { + "epoch": 0.904106294715325, + "flos": 19509930344640.0, + "grad_norm": 5.100893967011597, + "learning_rate": 9.558940053775954e-08, + "loss": 0.8112, + "num_input_tokens_seen": 160516345, + "step": 7519 + }, + { + "epoch": 0.904226537605964, + "flos": 12975799970520.0, + "grad_norm": 4.162264494777729, + "learning_rate": 9.535160126239294e-08, + "loss": 0.6712, + "num_input_tokens_seen": 160532690, + "step": 7520 + }, + { + "epoch": 0.9043467804966031, + "flos": 18159699318840.0, + "grad_norm": 2.1870698778442654, + "learning_rate": 9.511409092067424e-08, + "loss": 0.6991, + "num_input_tokens_seen": 160552765, + "step": 7521 + }, + { + "epoch": 0.9044670233872423, + "flos": 16561611228480.0, + "grad_norm": 4.71591950464509, + "learning_rate": 9.487686954863327e-08, + "loss": 0.6645, + "num_input_tokens_seen": 160572205, + "step": 7522 + }, + { + "epoch": 0.9045872662778813, + "flos": 17405843129520.0, + "grad_norm": 4.107817496104084, + "learning_rate": 9.46399371822566e-08, + "loss": 0.7512, + "num_input_tokens_seen": 160591700, + "step": 7523 + }, + { + "epoch": 0.9047075091685204, + "flos": 11054992484280.0, + "grad_norm": 11.57013644046502, + "learning_rate": 9.440329385748657e-08, + "loss": 0.6994, + "num_input_tokens_seen": 160608490, + "step": 7524 + }, + { + "epoch": 0.9048277520591596, + "flos": 13137017198640.0, + "grad_norm": 9.48816787319047, + "learning_rate": 9.416693961022137e-08, + "loss": 0.6968, + "num_input_tokens_seen": 160626460, + "step": 7525 + }, + { + "epoch": 0.9049479949497986, + "flos": 16000342796280.0, + "grad_norm": 21.78023768607931, + "learning_rate": 9.393087447631654e-08, + "loss": 0.7593, + "num_input_tokens_seen": 160644460, + "step": 7526 + }, + { + "epoch": 0.9050682378404377, + "flos": 15186784905120.0, + "grad_norm": 4.032393655527651, + "learning_rate": 9.36950984915823e-08, + "loss": 0.7117, + "num_input_tokens_seen": 160662535, + "step": 7527 + }, + { + "epoch": 0.9051884807310768, + "flos": 15783824600760.0, + "grad_norm": 4.677126787262759, + "learning_rate": 9.345961169178607e-08, + "loss": 0.6863, + "num_input_tokens_seen": 160681940, + "step": 7528 + }, + { + "epoch": 0.9053087236217159, + "flos": 16026552751800.0, + "grad_norm": 2.7370094904987674, + "learning_rate": 9.322441411265081e-08, + "loss": 0.7123, + "num_input_tokens_seen": 160702645, + "step": 7529 + }, + { + "epoch": 0.9054289665123549, + "flos": 12435612580320.0, + "grad_norm": 5.444887076223633, + "learning_rate": 9.298950578985554e-08, + "loss": 0.7153, + "num_input_tokens_seen": 160719440, + "step": 7530 + }, + { + "epoch": 0.905549209402994, + "flos": 15189729281400.0, + "grad_norm": 2.9916330326778837, + "learning_rate": 9.275488675903665e-08, + "loss": 0.6935, + "num_input_tokens_seen": 160738105, + "step": 7531 + }, + { + "epoch": 0.9056694522936332, + "flos": 15862486127280.0, + "grad_norm": 5.3686681312651565, + "learning_rate": 9.252055705578454e-08, + "loss": 0.7207, + "num_input_tokens_seen": 160757325, + "step": 7532 + }, + { + "epoch": 0.9057896951842722, + "flos": 21696130021920.0, + "grad_norm": 3.7066899776640527, + "learning_rate": 9.228651671564747e-08, + "loss": 0.7004, + "num_input_tokens_seen": 160779075, + "step": 7533 + }, + { + "epoch": 0.9059099380749113, + "flos": 20455329512400.0, + "grad_norm": 5.576875434489855, + "learning_rate": 9.205276577412901e-08, + "loss": 0.7743, + "num_input_tokens_seen": 160801575, + "step": 7534 + }, + { + "epoch": 0.9060301809655504, + "flos": 12948608556240.0, + "grad_norm": 5.336024360452433, + "learning_rate": 9.181930426668905e-08, + "loss": 0.7524, + "num_input_tokens_seen": 160818090, + "step": 7535 + }, + { + "epoch": 0.9061504238561895, + "flos": 23317262071800.0, + "grad_norm": 4.418247851553689, + "learning_rate": 9.158613222874346e-08, + "loss": 0.6763, + "num_input_tokens_seen": 160839435, + "step": 7536 + }, + { + "epoch": 0.9062706667468285, + "flos": 14650143430440.0, + "grad_norm": 2.8608239676898832, + "learning_rate": 9.135324969566394e-08, + "loss": 0.8089, + "num_input_tokens_seen": 160858655, + "step": 7537 + }, + { + "epoch": 0.9063909096374677, + "flos": 13457298777600.0, + "grad_norm": 72.47623196068315, + "learning_rate": 9.112065670277913e-08, + "loss": 0.7355, + "num_input_tokens_seen": 160874740, + "step": 7538 + }, + { + "epoch": 0.9065111525281068, + "flos": 24640459983120.0, + "grad_norm": 4.344920926795725, + "learning_rate": 9.088835328537303e-08, + "loss": 0.7071, + "num_input_tokens_seen": 160896050, + "step": 7539 + }, + { + "epoch": 0.9066313954187458, + "flos": 17106959192160.0, + "grad_norm": 4.1178079966340775, + "learning_rate": 9.065633947868568e-08, + "loss": 0.6927, + "num_input_tokens_seen": 160915375, + "step": 7540 + }, + { + "epoch": 0.906751638309385, + "flos": 19239567539880.0, + "grad_norm": 7.801321592609075, + "learning_rate": 9.042461531791379e-08, + "loss": 0.7776, + "num_input_tokens_seen": 160933515, + "step": 7541 + }, + { + "epoch": 0.906871881200024, + "flos": 12057117317640.0, + "grad_norm": 3.259210739312435, + "learning_rate": 9.019318083820903e-08, + "loss": 0.7503, + "num_input_tokens_seen": 160951815, + "step": 7542 + }, + { + "epoch": 0.9069921240906631, + "flos": 18022159249440.0, + "grad_norm": 4.765306005589179, + "learning_rate": 8.996203607468045e-08, + "loss": 0.8312, + "num_input_tokens_seen": 160970535, + "step": 7543 + }, + { + "epoch": 0.9071123669813023, + "flos": 18592197490560.0, + "grad_norm": 3.0358364369155497, + "learning_rate": 8.973118106239241e-08, + "loss": 0.7443, + "num_input_tokens_seen": 160992860, + "step": 7544 + }, + { + "epoch": 0.9072326098719413, + "flos": 19592232766560.0, + "grad_norm": 6.8260031050309795, + "learning_rate": 8.95006158363656e-08, + "loss": 0.9215, + "num_input_tokens_seen": 161012765, + "step": 7545 + }, + { + "epoch": 0.9073528527625804, + "flos": 17485106195280.0, + "grad_norm": 6.408065315399476, + "learning_rate": 8.9270340431576e-08, + "loss": 0.7576, + "num_input_tokens_seen": 161031575, + "step": 7546 + }, + { + "epoch": 0.9074730956532195, + "flos": 27826220037000.0, + "grad_norm": 5.054240706683226, + "learning_rate": 8.904035488295658e-08, + "loss": 0.7141, + "num_input_tokens_seen": 161050795, + "step": 7547 + }, + { + "epoch": 0.9075933385438586, + "flos": 48049934044320.0, + "grad_norm": 0.6924860437481061, + "learning_rate": 8.881065922539632e-08, + "loss": 0.5563, + "num_input_tokens_seen": 161110955, + "step": 7548 + }, + { + "epoch": 0.9077135814344977, + "flos": 14562933714720.0, + "grad_norm": 2.955810133487442, + "learning_rate": 8.85812534937389e-08, + "loss": 0.7293, + "num_input_tokens_seen": 161128775, + "step": 7549 + }, + { + "epoch": 0.9078338243251368, + "flos": 12786219909600.0, + "grad_norm": 6.386187740731029, + "learning_rate": 8.835213772278583e-08, + "loss": 0.6464, + "num_input_tokens_seen": 161145350, + "step": 7550 + }, + { + "epoch": 0.9079540672157759, + "flos": 21129511056480.0, + "grad_norm": 4.040496630595128, + "learning_rate": 8.812331194729373e-08, + "loss": 0.7734, + "num_input_tokens_seen": 161164715, + "step": 7551 + }, + { + "epoch": 0.9080743101064149, + "flos": 17214996670200.0, + "grad_norm": 3.736964393435883, + "learning_rate": 8.789477620197461e-08, + "loss": 0.7064, + "num_input_tokens_seen": 161183960, + "step": 7552 + }, + { + "epoch": 0.9081945529970541, + "flos": 16669648706520.0, + "grad_norm": 6.068645076453627, + "learning_rate": 8.766653052149831e-08, + "loss": 0.7775, + "num_input_tokens_seen": 161198865, + "step": 7553 + }, + { + "epoch": 0.9083147958876931, + "flos": 13780081493400.0, + "grad_norm": 37.39889038398232, + "learning_rate": 8.743857494048823e-08, + "loss": 0.7372, + "num_input_tokens_seen": 161215400, + "step": 7554 + }, + { + "epoch": 0.9084350387783322, + "flos": 13806354768840.0, + "grad_norm": 4.132815038323299, + "learning_rate": 8.721090949352605e-08, + "loss": 0.6263, + "num_input_tokens_seen": 161231360, + "step": 7555 + }, + { + "epoch": 0.9085552816689714, + "flos": 15055291888080.0, + "grad_norm": 14.890117240326353, + "learning_rate": 8.698353421514793e-08, + "loss": 0.7052, + "num_input_tokens_seen": 161249455, + "step": 7556 + }, + { + "epoch": 0.9086755245596104, + "flos": 13218021562200.0, + "grad_norm": 5.037265190918351, + "learning_rate": 8.67564491398467e-08, + "loss": 0.7857, + "num_input_tokens_seen": 161266180, + "step": 7557 + }, + { + "epoch": 0.9087957674502495, + "flos": 13969693214280.0, + "grad_norm": 6.3672486097527585, + "learning_rate": 8.652965430207104e-08, + "loss": 0.7294, + "num_input_tokens_seen": 161283805, + "step": 7558 + }, + { + "epoch": 0.9089160103408886, + "flos": 13215172165800.0, + "grad_norm": 8.285045286709785, + "learning_rate": 8.630314973622521e-08, + "loss": 0.6508, + "num_input_tokens_seen": 161301070, + "step": 7559 + }, + { + "epoch": 0.9090362532315277, + "flos": 24505389390600.0, + "grad_norm": 6.344418044219991, + "learning_rate": 8.607693547666995e-08, + "loss": 0.6951, + "num_input_tokens_seen": 161323330, + "step": 7560 + }, + { + "epoch": 0.9091564961221668, + "flos": 52717069881480.0, + "grad_norm": 0.9211782958594953, + "learning_rate": 8.585101155772201e-08, + "loss": 0.6174, + "num_input_tokens_seen": 161385170, + "step": 7561 + }, + { + "epoch": 0.9092767390128058, + "flos": 18101390655240.0, + "grad_norm": 3.1395369585574158, + "learning_rate": 8.562537801365377e-08, + "loss": 0.6766, + "num_input_tokens_seen": 161404625, + "step": 7562 + }, + { + "epoch": 0.909396981903445, + "flos": 17268714639600.0, + "grad_norm": 3.2143235370893244, + "learning_rate": 8.540003487869362e-08, + "loss": 0.6714, + "num_input_tokens_seen": 161426015, + "step": 7563 + }, + { + "epoch": 0.909517224794084, + "flos": 17133802346880.0, + "grad_norm": 3.84737268942432, + "learning_rate": 8.517498218702557e-08, + "loss": 0.7761, + "num_input_tokens_seen": 161443665, + "step": 7564 + }, + { + "epoch": 0.9096374676847231, + "flos": 14028856696800.0, + "grad_norm": 5.194574885960314, + "learning_rate": 8.49502199727905e-08, + "loss": 0.6888, + "num_input_tokens_seen": 161461410, + "step": 7565 + }, + { + "epoch": 0.9097577105753623, + "flos": 24451924700880.0, + "grad_norm": 6.005155573599636, + "learning_rate": 8.472574827008428e-08, + "loss": 0.631, + "num_input_tokens_seen": 161482015, + "step": 7566 + }, + { + "epoch": 0.9098779534660013, + "flos": 16025128053600.0, + "grad_norm": 4.129835589002001, + "learning_rate": 8.450156711295942e-08, + "loss": 0.8263, + "num_input_tokens_seen": 161501905, + "step": 7567 + }, + { + "epoch": 0.9099981963566404, + "flos": 18854075426040.0, + "grad_norm": 4.839501862230878, + "learning_rate": 8.427767653542383e-08, + "loss": 0.8454, + "num_input_tokens_seen": 161516795, + "step": 7568 + }, + { + "epoch": 0.9101184392472795, + "flos": 15405994197240.0, + "grad_norm": 3.3845369731716795, + "learning_rate": 8.405407657144125e-08, + "loss": 0.6883, + "num_input_tokens_seen": 161535675, + "step": 7569 + }, + { + "epoch": 0.9102386821379186, + "flos": 18131526445800.0, + "grad_norm": 5.920102943865655, + "learning_rate": 8.383076725493232e-08, + "loss": 0.6992, + "num_input_tokens_seen": 161552715, + "step": 7570 + }, + { + "epoch": 0.9103589250285576, + "flos": 16510837635360.0, + "grad_norm": 3.7027523320029547, + "learning_rate": 8.360774861977216e-08, + "loss": 0.6703, + "num_input_tokens_seen": 161571555, + "step": 7571 + }, + { + "epoch": 0.9104791679191968, + "flos": 18591089391960.0, + "grad_norm": 3.638565987455547, + "learning_rate": 8.338502069979281e-08, + "loss": 0.7306, + "num_input_tokens_seen": 161591585, + "step": 7572 + }, + { + "epoch": 0.9105994108098359, + "flos": 10489576597320.0, + "grad_norm": 4.112744244682371, + "learning_rate": 8.316258352878214e-08, + "loss": 0.7746, + "num_input_tokens_seen": 161607725, + "step": 7573 + }, + { + "epoch": 0.9107196537004749, + "flos": 19585615834920.0, + "grad_norm": 3.8820264546408296, + "learning_rate": 8.294043714048338e-08, + "loss": 0.6929, + "num_input_tokens_seen": 161626525, + "step": 7574 + }, + { + "epoch": 0.9108398965911141, + "flos": 49057251111120.0, + "grad_norm": 0.7874160055749256, + "learning_rate": 8.271858156859624e-08, + "loss": 0.6359, + "num_input_tokens_seen": 161691615, + "step": 7575 + }, + { + "epoch": 0.9109601394817531, + "flos": 18619072305240.0, + "grad_norm": 4.632525327379601, + "learning_rate": 8.249701684677557e-08, + "loss": 0.7253, + "num_input_tokens_seen": 161712660, + "step": 7576 + }, + { + "epoch": 0.9110803823723922, + "flos": 16648092765120.0, + "grad_norm": 4.362492147618963, + "learning_rate": 8.227574300863294e-08, + "loss": 0.7937, + "num_input_tokens_seen": 161732550, + "step": 7577 + }, + { + "epoch": 0.9112006252630314, + "flos": 35684149861680.0, + "grad_norm": 3.2026469548476646, + "learning_rate": 8.205476008773548e-08, + "loss": 0.6839, + "num_input_tokens_seen": 161756270, + "step": 7578 + }, + { + "epoch": 0.9113208681536704, + "flos": 22022521973160.0, + "grad_norm": 3.7004736571377466, + "learning_rate": 8.183406811760596e-08, + "loss": 0.8151, + "num_input_tokens_seen": 161775720, + "step": 7579 + }, + { + "epoch": 0.9114411110443095, + "flos": 18755377636200.0, + "grad_norm": 6.076772034206392, + "learning_rate": 8.161366713172313e-08, + "loss": 0.7316, + "num_input_tokens_seen": 161797830, + "step": 7580 + }, + { + "epoch": 0.9115613539349486, + "flos": 13429347524280.0, + "grad_norm": 6.862909401388026, + "learning_rate": 8.139355716352137e-08, + "loss": 0.8354, + "num_input_tokens_seen": 161812390, + "step": 7581 + }, + { + "epoch": 0.9116815968255877, + "flos": 15892020378600.0, + "grad_norm": 3.704731524337416, + "learning_rate": 8.117373824639196e-08, + "loss": 0.6862, + "num_input_tokens_seen": 161832375, + "step": 7582 + }, + { + "epoch": 0.9118018397162267, + "flos": 48191811528240.0, + "grad_norm": 0.7380036735469083, + "learning_rate": 8.095421041368067e-08, + "loss": 0.6019, + "num_input_tokens_seen": 161891510, + "step": 7583 + }, + { + "epoch": 0.9119220826068659, + "flos": 15297450159840.0, + "grad_norm": 4.322437722511481, + "learning_rate": 8.073497369868999e-08, + "loss": 0.691, + "num_input_tokens_seen": 161909690, + "step": 7584 + }, + { + "epoch": 0.912042325497505, + "flos": 20806253441280.0, + "grad_norm": 3.639532490145702, + "learning_rate": 8.051602813467772e-08, + "loss": 0.7347, + "num_input_tokens_seen": 161931265, + "step": 7585 + }, + { + "epoch": 0.912162568388144, + "flos": 12813791243400.0, + "grad_norm": 4.131956765435207, + "learning_rate": 8.029737375485756e-08, + "loss": 0.6951, + "num_input_tokens_seen": 161950215, + "step": 7586 + }, + { + "epoch": 0.9122828112787832, + "flos": 14487026604720.0, + "grad_norm": 7.221240650215191, + "learning_rate": 8.007901059239986e-08, + "loss": 0.7171, + "num_input_tokens_seen": 161969215, + "step": 7587 + }, + { + "epoch": 0.9124030541694222, + "flos": 15216762395880.0, + "grad_norm": 6.731823502020349, + "learning_rate": 7.986093868042964e-08, + "loss": 0.7901, + "num_input_tokens_seen": 161989180, + "step": 7588 + }, + { + "epoch": 0.9125232970600613, + "flos": 18457190217960.0, + "grad_norm": 5.059321406400439, + "learning_rate": 7.964315805202826e-08, + "loss": 0.6675, + "num_input_tokens_seen": 162009480, + "step": 7589 + }, + { + "epoch": 0.9126435399507005, + "flos": 14407035359880.0, + "grad_norm": 6.659253226446307, + "learning_rate": 7.942566874023304e-08, + "loss": 0.7177, + "num_input_tokens_seen": 162028385, + "step": 7590 + }, + { + "epoch": 0.9127637828413395, + "flos": 14298997881840.0, + "grad_norm": 4.100136703445201, + "learning_rate": 7.920847077803649e-08, + "loss": 0.6739, + "num_input_tokens_seen": 162045895, + "step": 7591 + }, + { + "epoch": 0.9128840257319786, + "flos": 14785277342880.0, + "grad_norm": 6.6025332126554845, + "learning_rate": 7.899156419838826e-08, + "loss": 0.788, + "num_input_tokens_seen": 162064585, + "step": 7592 + }, + { + "epoch": 0.9130042686226177, + "flos": 18209523113160.0, + "grad_norm": 5.323712711817183, + "learning_rate": 7.87749490341918e-08, + "loss": 0.6433, + "num_input_tokens_seen": 162084580, + "step": 7593 + }, + { + "epoch": 0.9131245115132568, + "flos": 17264535524880.0, + "grad_norm": 6.027204510705372, + "learning_rate": 7.855862531830836e-08, + "loss": 0.8326, + "num_input_tokens_seen": 162100410, + "step": 7594 + }, + { + "epoch": 0.9132447544038959, + "flos": 14562965374680.0, + "grad_norm": 12.81173361370258, + "learning_rate": 7.834259308355373e-08, + "loss": 0.715, + "num_input_tokens_seen": 162118895, + "step": 7595 + }, + { + "epoch": 0.9133649972945349, + "flos": 16081188860040.0, + "grad_norm": 5.632771866184923, + "learning_rate": 7.812685236269989e-08, + "loss": 0.7334, + "num_input_tokens_seen": 162137275, + "step": 7596 + }, + { + "epoch": 0.9134852401851741, + "flos": 52541578502760.0, + "grad_norm": 0.8413529785634006, + "learning_rate": 7.791140318847445e-08, + "loss": 0.6091, + "num_input_tokens_seen": 162195130, + "step": 7597 + }, + { + "epoch": 0.9136054830758131, + "flos": 17299198689720.0, + "grad_norm": 4.3232268563018765, + "learning_rate": 7.769624559356081e-08, + "loss": 0.7847, + "num_input_tokens_seen": 162218245, + "step": 7598 + }, + { + "epoch": 0.9137257259664522, + "flos": 17159220803400.0, + "grad_norm": 7.785576165450271, + "learning_rate": 7.748137961059842e-08, + "loss": 0.7009, + "num_input_tokens_seen": 162231945, + "step": 7599 + }, + { + "epoch": 0.9138459688570914, + "flos": 13967983576440.0, + "grad_norm": 4.479750957378849, + "learning_rate": 7.726680527218211e-08, + "loss": 0.646, + "num_input_tokens_seen": 162248705, + "step": 7600 + }, + { + "epoch": 0.9139662117477304, + "flos": 34064442510000.0, + "grad_norm": 2.9313367753513657, + "learning_rate": 7.70525226108627e-08, + "loss": 0.7386, + "num_input_tokens_seen": 162272095, + "step": 7601 + }, + { + "epoch": 0.9140864546383695, + "flos": 16296820576680.0, + "grad_norm": 9.868480993256146, + "learning_rate": 7.683853165914666e-08, + "loss": 0.786, + "num_input_tokens_seen": 162289585, + "step": 7602 + }, + { + "epoch": 0.9142066975290086, + "flos": 12786536509200.0, + "grad_norm": 5.318403666210632, + "learning_rate": 7.662483244949602e-08, + "loss": 0.7548, + "num_input_tokens_seen": 162306565, + "step": 7603 + }, + { + "epoch": 0.9143269404196477, + "flos": 12921290502120.0, + "grad_norm": 8.293708383761924, + "learning_rate": 7.641142501432951e-08, + "loss": 0.799, + "num_input_tokens_seen": 162322480, + "step": 7604 + }, + { + "epoch": 0.9144471833102867, + "flos": 24474018861600.0, + "grad_norm": 3.604524598328614, + "learning_rate": 7.619830938602013e-08, + "loss": 0.7295, + "num_input_tokens_seen": 162343425, + "step": 7605 + }, + { + "epoch": 0.9145674262009259, + "flos": 15405582617760.0, + "grad_norm": 5.084743470929594, + "learning_rate": 7.598548559689777e-08, + "loss": 0.7951, + "num_input_tokens_seen": 162361545, + "step": 7606 + }, + { + "epoch": 0.914687669091565, + "flos": 12246475758840.0, + "grad_norm": 6.037949078209, + "learning_rate": 7.577295367924751e-08, + "loss": 0.793, + "num_input_tokens_seen": 162377665, + "step": 7607 + }, + { + "epoch": 0.914807911982204, + "flos": 18887060613000.0, + "grad_norm": 5.798027150285622, + "learning_rate": 7.556071366531002e-08, + "loss": 0.8202, + "num_input_tokens_seen": 162398355, + "step": 7608 + }, + { + "epoch": 0.9149281548728432, + "flos": 14028350137440.0, + "grad_norm": 3.2192263945087123, + "learning_rate": 7.53487655872822e-08, + "loss": 0.7626, + "num_input_tokens_seen": 162417245, + "step": 7609 + }, + { + "epoch": 0.9150483977634822, + "flos": 19699193805960.0, + "grad_norm": 4.084864634181877, + "learning_rate": 7.513710947731656e-08, + "loss": 0.7335, + "num_input_tokens_seen": 162438175, + "step": 7610 + }, + { + "epoch": 0.9151686406541213, + "flos": 16027344250800.0, + "grad_norm": 4.420879777717689, + "learning_rate": 7.492574536752095e-08, + "loss": 0.8307, + "num_input_tokens_seen": 162457885, + "step": 7611 + }, + { + "epoch": 0.9152888835447605, + "flos": 20023369560000.0, + "grad_norm": 4.260770054021582, + "learning_rate": 7.471467328995907e-08, + "loss": 0.7708, + "num_input_tokens_seen": 162476415, + "step": 7612 + }, + { + "epoch": 0.9154091264353995, + "flos": 9708402353880.0, + "grad_norm": 5.506099266734285, + "learning_rate": 7.450389327665018e-08, + "loss": 0.5826, + "num_input_tokens_seen": 162493970, + "step": 7613 + }, + { + "epoch": 0.9155293693260386, + "flos": 14757294429600.0, + "grad_norm": 5.083473933208889, + "learning_rate": 7.429340535957029e-08, + "loss": 0.6465, + "num_input_tokens_seen": 162508885, + "step": 7614 + }, + { + "epoch": 0.9156496122166777, + "flos": 14137242434400.0, + "grad_norm": 11.458721880249396, + "learning_rate": 7.40832095706494e-08, + "loss": 0.6983, + "num_input_tokens_seen": 162525300, + "step": 7615 + }, + { + "epoch": 0.9157698551073168, + "flos": 15434198730240.0, + "grad_norm": 3.6856762530285634, + "learning_rate": 7.387330594177443e-08, + "loss": 0.7879, + "num_input_tokens_seen": 162547095, + "step": 7616 + }, + { + "epoch": 0.9158900979979558, + "flos": 18457950057000.0, + "grad_norm": 3.3991417867565645, + "learning_rate": 7.366369450478749e-08, + "loss": 0.7838, + "num_input_tokens_seen": 162567925, + "step": 7617 + }, + { + "epoch": 0.916010340888595, + "flos": 22122644461200.0, + "grad_norm": 4.452470697888057, + "learning_rate": 7.345437529148646e-08, + "loss": 0.6472, + "num_input_tokens_seen": 162586655, + "step": 7618 + }, + { + "epoch": 0.9161305837792341, + "flos": 12462202455360.0, + "grad_norm": 13.964503383089859, + "learning_rate": 7.324534833362483e-08, + "loss": 0.7244, + "num_input_tokens_seen": 162603950, + "step": 7619 + }, + { + "epoch": 0.9162508266698731, + "flos": 16756066923240.0, + "grad_norm": 4.1862324176828025, + "learning_rate": 7.303661366291192e-08, + "loss": 0.6674, + "num_input_tokens_seen": 162624340, + "step": 7620 + }, + { + "epoch": 0.9163710695605123, + "flos": 14595887241720.0, + "grad_norm": 5.061442135145551, + "learning_rate": 7.28281713110126e-08, + "loss": 0.8091, + "num_input_tokens_seen": 162642135, + "step": 7621 + }, + { + "epoch": 0.9164913124511513, + "flos": 16674682640160.0, + "grad_norm": 5.58161966396136, + "learning_rate": 7.262002130954759e-08, + "loss": 0.7586, + "num_input_tokens_seen": 162660310, + "step": 7622 + }, + { + "epoch": 0.9166115553417904, + "flos": 18241558501320.0, + "grad_norm": 3.9125895822037777, + "learning_rate": 7.241216369009296e-08, + "loss": 0.7731, + "num_input_tokens_seen": 162680215, + "step": 7623 + }, + { + "epoch": 0.9167317982324296, + "flos": 18835052281440.0, + "grad_norm": 4.104217193344134, + "learning_rate": 7.220459848418037e-08, + "loss": 0.6559, + "num_input_tokens_seen": 162700010, + "step": 7624 + }, + { + "epoch": 0.9168520411230686, + "flos": 11382492534120.0, + "grad_norm": 4.80648784867424, + "learning_rate": 7.199732572329708e-08, + "loss": 0.7761, + "num_input_tokens_seen": 162717630, + "step": 7625 + }, + { + "epoch": 0.9169722840137077, + "flos": 22205453442480.0, + "grad_norm": 4.1611308644166405, + "learning_rate": 7.179034543888684e-08, + "loss": 0.7354, + "num_input_tokens_seen": 162736855, + "step": 7626 + }, + { + "epoch": 0.9170925269043467, + "flos": 16458417724320.0, + "grad_norm": 4.420109804265353, + "learning_rate": 7.158365766234808e-08, + "loss": 0.7625, + "num_input_tokens_seen": 162755425, + "step": 7627 + }, + { + "epoch": 0.9172127697949859, + "flos": 16756256883000.0, + "grad_norm": 4.677235873424539, + "learning_rate": 7.137726242503527e-08, + "loss": 0.7077, + "num_input_tokens_seen": 162774065, + "step": 7628 + }, + { + "epoch": 0.917333012685625, + "flos": 12727974565920.0, + "grad_norm": 11.767485249892829, + "learning_rate": 7.11711597582585e-08, + "loss": 0.7661, + "num_input_tokens_seen": 162791145, + "step": 7629 + }, + { + "epoch": 0.917453255576264, + "flos": 10410091911840.0, + "grad_norm": 3.6832033114261806, + "learning_rate": 7.096534969328271e-08, + "loss": 0.7891, + "num_input_tokens_seen": 162808310, + "step": 7630 + }, + { + "epoch": 0.9175734984669032, + "flos": 15163076086440.0, + "grad_norm": 4.42466275928994, + "learning_rate": 7.075983226132987e-08, + "loss": 0.8242, + "num_input_tokens_seen": 162826960, + "step": 7631 + }, + { + "epoch": 0.9176937413575422, + "flos": 10788777134280.0, + "grad_norm": 6.85431096660374, + "learning_rate": 7.055460749357656e-08, + "loss": 0.7695, + "num_input_tokens_seen": 162842960, + "step": 7632 + }, + { + "epoch": 0.9178139842481813, + "flos": 13485028411200.0, + "grad_norm": 5.672027669906379, + "learning_rate": 7.034967542115521e-08, + "loss": 0.694, + "num_input_tokens_seen": 162860945, + "step": 7633 + }, + { + "epoch": 0.9179342271388204, + "flos": 14649605211120.0, + "grad_norm": 5.561203194659411, + "learning_rate": 7.014503607515388e-08, + "loss": 0.7392, + "num_input_tokens_seen": 162879970, + "step": 7634 + }, + { + "epoch": 0.9180544700294595, + "flos": 18074515840560.0, + "grad_norm": 25.70062321601183, + "learning_rate": 6.994068948661592e-08, + "loss": 0.662, + "num_input_tokens_seen": 162897845, + "step": 7635 + }, + { + "epoch": 0.9181747129200986, + "flos": 12192631149600.0, + "grad_norm": 3.994932421708296, + "learning_rate": 6.973663568654142e-08, + "loss": 0.745, + "num_input_tokens_seen": 162915270, + "step": 7636 + }, + { + "epoch": 0.9182949558107377, + "flos": 17774618784480.0, + "grad_norm": 16.55574928756341, + "learning_rate": 6.953287470588386e-08, + "loss": 0.6393, + "num_input_tokens_seen": 162932945, + "step": 7637 + }, + { + "epoch": 0.9184151987013768, + "flos": 16154183253720.0, + "grad_norm": 5.288344995241545, + "learning_rate": 6.932940657555452e-08, + "loss": 0.8367, + "num_input_tokens_seen": 162948795, + "step": 7638 + }, + { + "epoch": 0.9185354415920158, + "flos": 23619533624880.0, + "grad_norm": 4.372485719006415, + "learning_rate": 6.912623132641938e-08, + "loss": 0.7492, + "num_input_tokens_seen": 162973605, + "step": 7639 + }, + { + "epoch": 0.918655684482655, + "flos": 15353194366680.0, + "grad_norm": 4.3076139773069535, + "learning_rate": 6.892334898929952e-08, + "loss": 0.7456, + "num_input_tokens_seen": 162993570, + "step": 7640 + }, + { + "epoch": 0.918775927373294, + "flos": 11329249464120.0, + "grad_norm": 4.989153732674683, + "learning_rate": 6.872075959497236e-08, + "loss": 0.8276, + "num_input_tokens_seen": 163012065, + "step": 7641 + }, + { + "epoch": 0.9188961702639331, + "flos": 21966207887040.0, + "grad_norm": 5.083699426861397, + "learning_rate": 6.85184631741702e-08, + "loss": 0.8087, + "num_input_tokens_seen": 163032350, + "step": 7642 + }, + { + "epoch": 0.9190164131545723, + "flos": 15133035275760.0, + "grad_norm": 5.072617769411164, + "learning_rate": 6.831645975758161e-08, + "loss": 0.7617, + "num_input_tokens_seen": 163050010, + "step": 7643 + }, + { + "epoch": 0.9191366560452113, + "flos": 18781017712440.0, + "grad_norm": 4.956134062634077, + "learning_rate": 6.811474937585026e-08, + "loss": 0.6549, + "num_input_tokens_seen": 163069520, + "step": 7644 + }, + { + "epoch": 0.9192568989358504, + "flos": 15676040402400.0, + "grad_norm": 3.673795417019473, + "learning_rate": 6.79133320595755e-08, + "loss": 0.7703, + "num_input_tokens_seen": 163089160, + "step": 7645 + }, + { + "epoch": 0.9193771418264896, + "flos": 16971635319960.0, + "grad_norm": 3.511265064504594, + "learning_rate": 6.771220783931198e-08, + "loss": 0.722, + "num_input_tokens_seen": 163109040, + "step": 7646 + }, + { + "epoch": 0.9194973847171286, + "flos": 52054824142320.0, + "grad_norm": 0.863323469553633, + "learning_rate": 6.751137674556994e-08, + "loss": 0.6667, + "num_input_tokens_seen": 163169145, + "step": 7647 + }, + { + "epoch": 0.9196176276077677, + "flos": 10707456171120.0, + "grad_norm": 5.74736659243194, + "learning_rate": 6.731083880881572e-08, + "loss": 0.7696, + "num_input_tokens_seen": 163185085, + "step": 7648 + }, + { + "epoch": 0.9197378704984068, + "flos": 17052956283120.0, + "grad_norm": 3.5600486486304135, + "learning_rate": 6.711059405947072e-08, + "loss": 0.7922, + "num_input_tokens_seen": 163202995, + "step": 7649 + }, + { + "epoch": 0.9198581133890459, + "flos": 14838362113080.0, + "grad_norm": 4.602218607060918, + "learning_rate": 6.691064252791156e-08, + "loss": 0.7561, + "num_input_tokens_seen": 163222190, + "step": 7650 + }, + { + "epoch": 0.9199783562796849, + "flos": 12894289047600.0, + "grad_norm": 3.5211879544111993, + "learning_rate": 6.67109842444713e-08, + "loss": 0.7657, + "num_input_tokens_seen": 163240840, + "step": 7651 + }, + { + "epoch": 0.9200985991703241, + "flos": 12894605647200.0, + "grad_norm": 4.864864866726391, + "learning_rate": 6.651161923943704e-08, + "loss": 0.7543, + "num_input_tokens_seen": 163258465, + "step": 7652 + }, + { + "epoch": 0.9202188420609632, + "flos": 15352086268080.0, + "grad_norm": 3.5178237481914643, + "learning_rate": 6.631254754305326e-08, + "loss": 0.7533, + "num_input_tokens_seen": 163277645, + "step": 7653 + }, + { + "epoch": 0.9203390849516022, + "flos": 10113614131440.0, + "grad_norm": 4.17293440836758, + "learning_rate": 6.611376918551848e-08, + "loss": 0.7725, + "num_input_tokens_seen": 163296150, + "step": 7654 + }, + { + "epoch": 0.9204593278422414, + "flos": 15484370784120.0, + "grad_norm": 5.51857827633319, + "learning_rate": 6.591528419698744e-08, + "loss": 0.7807, + "num_input_tokens_seen": 163315655, + "step": 7655 + }, + { + "epoch": 0.9205795707328804, + "flos": 10545099184440.0, + "grad_norm": 4.308023091600563, + "learning_rate": 6.571709260756986e-08, + "loss": 0.8024, + "num_input_tokens_seen": 163332020, + "step": 7656 + }, + { + "epoch": 0.9206998136235195, + "flos": 16403433356520.0, + "grad_norm": 5.383512985648905, + "learning_rate": 6.551919444733122e-08, + "loss": 0.7266, + "num_input_tokens_seen": 163349555, + "step": 7657 + }, + { + "epoch": 0.9208200565141585, + "flos": 39301869867960.0, + "grad_norm": 6.280035859995708, + "learning_rate": 6.53215897462931e-08, + "loss": 0.6448, + "num_input_tokens_seen": 163373030, + "step": 7658 + }, + { + "epoch": 0.9209402994047977, + "flos": 22450777710240.0, + "grad_norm": 4.394422307677414, + "learning_rate": 6.512427853443103e-08, + "loss": 0.7379, + "num_input_tokens_seen": 163394830, + "step": 7659 + }, + { + "epoch": 0.9210605422954368, + "flos": 21373347306120.0, + "grad_norm": 4.7526265963369125, + "learning_rate": 6.492726084167799e-08, + "loss": 0.7456, + "num_input_tokens_seen": 163416665, + "step": 7660 + }, + { + "epoch": 0.9211807851860758, + "flos": 40413682988640.0, + "grad_norm": 0.7857959475747985, + "learning_rate": 6.473053669792072e-08, + "loss": 0.5734, + "num_input_tokens_seen": 163471075, + "step": 7661 + }, + { + "epoch": 0.921301028076715, + "flos": 14022999604200.0, + "grad_norm": 5.127252781421522, + "learning_rate": 6.453410613300248e-08, + "loss": 0.7103, + "num_input_tokens_seen": 163488725, + "step": 7662 + }, + { + "epoch": 0.921421270967354, + "flos": 20184871727760.0, + "grad_norm": 3.3824789750601014, + "learning_rate": 6.43379691767214e-08, + "loss": 0.5648, + "num_input_tokens_seen": 163507650, + "step": 7663 + }, + { + "epoch": 0.9215415138579931, + "flos": 51777052906920.0, + "grad_norm": 0.9315504434311481, + "learning_rate": 6.414212585883105e-08, + "loss": 0.6015, + "num_input_tokens_seen": 163570000, + "step": 7664 + }, + { + "epoch": 0.9216617567486323, + "flos": 26124716822760.0, + "grad_norm": 4.727221633728384, + "learning_rate": 6.394657620904143e-08, + "loss": 0.6791, + "num_input_tokens_seen": 163592830, + "step": 7665 + }, + { + "epoch": 0.9217819996392713, + "flos": 21669730106640.0, + "grad_norm": 3.999157220747557, + "learning_rate": 6.375132025701657e-08, + "loss": 0.7053, + "num_input_tokens_seen": 163614850, + "step": 7666 + }, + { + "epoch": 0.9219022425299104, + "flos": 10599387033120.0, + "grad_norm": 9.447574316269007, + "learning_rate": 6.355635803237724e-08, + "loss": 0.6729, + "num_input_tokens_seen": 163630270, + "step": 7667 + }, + { + "epoch": 0.9220224854205495, + "flos": 13191748286760.0, + "grad_norm": 33.118104382506935, + "learning_rate": 6.336168956469867e-08, + "loss": 0.7806, + "num_input_tokens_seen": 163648465, + "step": 7668 + }, + { + "epoch": 0.9221427283111886, + "flos": 18159730978800.0, + "grad_norm": 3.7290409571310605, + "learning_rate": 6.316731488351168e-08, + "loss": 0.7045, + "num_input_tokens_seen": 163669375, + "step": 7669 + }, + { + "epoch": 0.9222629712018277, + "flos": 10059421262640.0, + "grad_norm": 4.925625005266325, + "learning_rate": 6.297323401830334e-08, + "loss": 0.6287, + "num_input_tokens_seen": 163687880, + "step": 7670 + }, + { + "epoch": 0.9223832140924668, + "flos": 15810541115640.0, + "grad_norm": 4.709777385528985, + "learning_rate": 6.277944699851523e-08, + "loss": 0.668, + "num_input_tokens_seen": 163707120, + "step": 7671 + }, + { + "epoch": 0.9225034569831059, + "flos": 15459712166640.0, + "grad_norm": 4.724807186330258, + "learning_rate": 6.25859538535447e-08, + "loss": 0.7264, + "num_input_tokens_seen": 163727635, + "step": 7672 + }, + { + "epoch": 0.9226236998737449, + "flos": 9060335785440.0, + "grad_norm": 6.589628983792863, + "learning_rate": 6.239275461274474e-08, + "loss": 0.7574, + "num_input_tokens_seen": 163743730, + "step": 7673 + }, + { + "epoch": 0.9227439427643841, + "flos": 19213389244320.0, + "grad_norm": 6.877470370263158, + "learning_rate": 6.219984930542299e-08, + "loss": 0.8522, + "num_input_tokens_seen": 163764190, + "step": 7674 + }, + { + "epoch": 0.9228641856550232, + "flos": 13110332343720.0, + "grad_norm": 4.799467597681164, + "learning_rate": 6.200723796084383e-08, + "loss": 0.7275, + "num_input_tokens_seen": 163782005, + "step": 7675 + }, + { + "epoch": 0.9229844285456622, + "flos": 51932349722520.0, + "grad_norm": 0.7361092887312045, + "learning_rate": 6.181492060822546e-08, + "loss": 0.6357, + "num_input_tokens_seen": 163841900, + "step": 7676 + }, + { + "epoch": 0.9231046714363014, + "flos": 13110553963440.0, + "grad_norm": 7.371100995293206, + "learning_rate": 6.162289727674274e-08, + "loss": 0.7983, + "num_input_tokens_seen": 163859300, + "step": 7677 + }, + { + "epoch": 0.9232249143269404, + "flos": 13029422960040.0, + "grad_norm": 6.66633465258048, + "learning_rate": 6.143116799552527e-08, + "loss": 0.8713, + "num_input_tokens_seen": 163875265, + "step": 7678 + }, + { + "epoch": 0.9233451572175795, + "flos": 17133992306640.0, + "grad_norm": 16.076930113823522, + "learning_rate": 6.123973279365802e-08, + "loss": 0.5398, + "num_input_tokens_seen": 163893960, + "step": 7679 + }, + { + "epoch": 0.9234654001082186, + "flos": 13133154683520.0, + "grad_norm": 3.7941178723225177, + "learning_rate": 6.10485917001824e-08, + "loss": 0.7641, + "num_input_tokens_seen": 163911535, + "step": 7680 + }, + { + "epoch": 0.9235856429988577, + "flos": 18129880127880.0, + "grad_norm": 3.0859874093570845, + "learning_rate": 6.085774474409322e-08, + "loss": 0.7999, + "num_input_tokens_seen": 163931815, + "step": 7681 + }, + { + "epoch": 0.9237058858894968, + "flos": 10247956544880.0, + "grad_norm": 6.5055579851759315, + "learning_rate": 6.066719195434267e-08, + "loss": 0.6948, + "num_input_tokens_seen": 163949335, + "step": 7682 + }, + { + "epoch": 0.9238261287801359, + "flos": 21049013252280.0, + "grad_norm": 4.163701311947377, + "learning_rate": 6.047693335983717e-08, + "loss": 0.6424, + "num_input_tokens_seen": 163971400, + "step": 7683 + }, + { + "epoch": 0.923946371670775, + "flos": 16917410791200.0, + "grad_norm": 4.881286089540281, + "learning_rate": 6.028696898943853e-08, + "loss": 0.792, + "num_input_tokens_seen": 163990180, + "step": 7684 + }, + { + "epoch": 0.924066614561414, + "flos": 15995625462240.0, + "grad_norm": 3.7257703662984514, + "learning_rate": 6.00972988719648e-08, + "loss": 0.6824, + "num_input_tokens_seen": 164008135, + "step": 7685 + }, + { + "epoch": 0.9241868574520532, + "flos": 20913816019920.0, + "grad_norm": 4.323927182892768, + "learning_rate": 5.990792303618807e-08, + "loss": 0.6869, + "num_input_tokens_seen": 164027435, + "step": 7686 + }, + { + "epoch": 0.9243071003426923, + "flos": 22398674398800.0, + "grad_norm": 3.2091980466560694, + "learning_rate": 5.971884151083695e-08, + "loss": 0.6776, + "num_input_tokens_seen": 164049565, + "step": 7687 + }, + { + "epoch": 0.9244273432333313, + "flos": 21022265077440.0, + "grad_norm": 4.3236629470041805, + "learning_rate": 5.9530054324595124e-08, + "loss": 0.7334, + "num_input_tokens_seen": 164069400, + "step": 7688 + }, + { + "epoch": 0.9245475861239704, + "flos": 48832279706280.0, + "grad_norm": 0.7554223820560781, + "learning_rate": 5.934156150610103e-08, + "loss": 0.5889, + "num_input_tokens_seen": 164126485, + "step": 7689 + }, + { + "epoch": 0.9246678290146095, + "flos": 17751828104640.0, + "grad_norm": 5.166292894851261, + "learning_rate": 5.915336308394914e-08, + "loss": 0.7601, + "num_input_tokens_seen": 164145040, + "step": 7690 + }, + { + "epoch": 0.9247880719052486, + "flos": 13866056470680.0, + "grad_norm": 2.9190469202790363, + "learning_rate": 5.89654590866886e-08, + "loss": 0.7629, + "num_input_tokens_seen": 164164260, + "step": 7691 + }, + { + "epoch": 0.9249083147958876, + "flos": 17592542134080.0, + "grad_norm": 5.689031373829126, + "learning_rate": 5.877784954282483e-08, + "loss": 0.8668, + "num_input_tokens_seen": 164183320, + "step": 7692 + }, + { + "epoch": 0.9250285576865268, + "flos": 22587906200160.0, + "grad_norm": 4.091248191882047, + "learning_rate": 5.8590534480817963e-08, + "loss": 0.7103, + "num_input_tokens_seen": 164204765, + "step": 7693 + }, + { + "epoch": 0.9251488005771659, + "flos": 7682628405720.0, + "grad_norm": 5.7160771695287504, + "learning_rate": 5.840351392908349e-08, + "loss": 0.7084, + "num_input_tokens_seen": 164220205, + "step": 7694 + }, + { + "epoch": 0.9252690434678049, + "flos": 17268651319680.0, + "grad_norm": 5.010953349703067, + "learning_rate": 5.821678791599205e-08, + "loss": 0.696, + "num_input_tokens_seen": 164239370, + "step": 7695 + }, + { + "epoch": 0.9253892863584441, + "flos": 15701585498760.0, + "grad_norm": 5.211821382472996, + "learning_rate": 5.803035646986965e-08, + "loss": 0.7876, + "num_input_tokens_seen": 164258425, + "step": 7696 + }, + { + "epoch": 0.9255095292490831, + "flos": 12732217000560.0, + "grad_norm": 5.382228515128515, + "learning_rate": 5.7844219618998766e-08, + "loss": 0.6555, + "num_input_tokens_seen": 164272470, + "step": 7697 + }, + { + "epoch": 0.9256297721397222, + "flos": 18129816807960.0, + "grad_norm": 3.578874667731339, + "learning_rate": 5.765837739161505e-08, + "loss": 0.6858, + "num_input_tokens_seen": 164291310, + "step": 7698 + }, + { + "epoch": 0.9257500150303614, + "flos": 16917727390800.0, + "grad_norm": 96.96901184302385, + "learning_rate": 5.7472829815911504e-08, + "loss": 0.7374, + "num_input_tokens_seen": 164309855, + "step": 7699 + }, + { + "epoch": 0.9258702579210004, + "flos": 16512483953280.0, + "grad_norm": 6.8282657844468, + "learning_rate": 5.7287576920035164e-08, + "loss": 0.799, + "num_input_tokens_seen": 164328590, + "step": 7700 + }, + { + "epoch": 0.9259905008116395, + "flos": 22017804639120.0, + "grad_norm": 3.8994604079780926, + "learning_rate": 5.7102618732088435e-08, + "loss": 0.759, + "num_input_tokens_seen": 164347640, + "step": 7701 + }, + { + "epoch": 0.9261107437022786, + "flos": 17998228811040.0, + "grad_norm": 3.9077520305556765, + "learning_rate": 5.6917955280130216e-08, + "loss": 0.7325, + "num_input_tokens_seen": 164368840, + "step": 7702 + }, + { + "epoch": 0.9262309865929177, + "flos": 16108380274320.0, + "grad_norm": 6.143960682544472, + "learning_rate": 5.6733586592172755e-08, + "loss": 0.6983, + "num_input_tokens_seen": 164388055, + "step": 7703 + }, + { + "epoch": 0.9263512294835567, + "flos": 14865743487120.0, + "grad_norm": 3.9176260131745044, + "learning_rate": 5.6549512696185244e-08, + "loss": 0.7965, + "num_input_tokens_seen": 164406275, + "step": 7704 + }, + { + "epoch": 0.9264714723741959, + "flos": 15514000015320.0, + "grad_norm": 3.348658728756522, + "learning_rate": 5.636573362009156e-08, + "loss": 0.6594, + "num_input_tokens_seen": 164426055, + "step": 7705 + }, + { + "epoch": 0.926591715264835, + "flos": 13138125297240.0, + "grad_norm": 4.442980109157165, + "learning_rate": 5.618224939177074e-08, + "loss": 0.7456, + "num_input_tokens_seen": 164443680, + "step": 7706 + }, + { + "epoch": 0.926711958155474, + "flos": 26578992555600.0, + "grad_norm": 5.263034081257551, + "learning_rate": 5.599906003905719e-08, + "loss": 0.6898, + "num_input_tokens_seen": 164465945, + "step": 7707 + }, + { + "epoch": 0.9268322010461132, + "flos": 15378802782960.0, + "grad_norm": 3.9786213339693703, + "learning_rate": 5.581616558974023e-08, + "loss": 0.8001, + "num_input_tokens_seen": 164484230, + "step": 7708 + }, + { + "epoch": 0.9269524439367522, + "flos": 16808455174320.0, + "grad_norm": 28.01863241143349, + "learning_rate": 5.5633566071565444e-08, + "loss": 0.7736, + "num_input_tokens_seen": 164503190, + "step": 7709 + }, + { + "epoch": 0.9270726868273913, + "flos": 30877890957120.0, + "grad_norm": 4.604062707800405, + "learning_rate": 5.5451261512232896e-08, + "loss": 0.6888, + "num_input_tokens_seen": 164525590, + "step": 7710 + }, + { + "epoch": 0.9271929297180305, + "flos": 14460341749800.0, + "grad_norm": 4.991491708040348, + "learning_rate": 5.5269251939397576e-08, + "loss": 0.611, + "num_input_tokens_seen": 164544825, + "step": 7711 + }, + { + "epoch": 0.9273131726086695, + "flos": 14590188448920.0, + "grad_norm": 5.460451118082083, + "learning_rate": 5.508753738067073e-08, + "loss": 0.7542, + "num_input_tokens_seen": 164564085, + "step": 7712 + }, + { + "epoch": 0.9274334154993086, + "flos": 17026113128400.0, + "grad_norm": 8.433467551307993, + "learning_rate": 5.4906117863617875e-08, + "loss": 0.7733, + "num_input_tokens_seen": 164583190, + "step": 7713 + }, + { + "epoch": 0.9275536583899477, + "flos": 23343756966960.0, + "grad_norm": 5.025120573352782, + "learning_rate": 5.4724993415760533e-08, + "loss": 0.7651, + "num_input_tokens_seen": 164601265, + "step": 7714 + }, + { + "epoch": 0.9276739012805868, + "flos": 13650108154440.0, + "grad_norm": 5.332444395051707, + "learning_rate": 5.454416406457496e-08, + "loss": 0.7251, + "num_input_tokens_seen": 164620080, + "step": 7715 + }, + { + "epoch": 0.9277941441712259, + "flos": 10084269839880.0, + "grad_norm": 5.407416598187846, + "learning_rate": 5.436362983749299e-08, + "loss": 0.7214, + "num_input_tokens_seen": 164634970, + "step": 7716 + }, + { + "epoch": 0.927914387061865, + "flos": 17026176448320.0, + "grad_norm": 6.430787966844477, + "learning_rate": 5.418339076190137e-08, + "loss": 0.6226, + "num_input_tokens_seen": 164654200, + "step": 7717 + }, + { + "epoch": 0.9280346299525041, + "flos": 13188423990960.0, + "grad_norm": 4.361162591672199, + "learning_rate": 5.400344686514202e-08, + "loss": 0.8764, + "num_input_tokens_seen": 164671505, + "step": 7718 + }, + { + "epoch": 0.9281548728431431, + "flos": 16347942429360.0, + "grad_norm": 4.016173091547416, + "learning_rate": 5.38237981745131e-08, + "loss": 0.6547, + "num_input_tokens_seen": 164689340, + "step": 7719 + }, + { + "epoch": 0.9282751157337822, + "flos": 13758842151600.0, + "grad_norm": 5.565251216513525, + "learning_rate": 5.364444471726592e-08, + "loss": 0.7888, + "num_input_tokens_seen": 164708265, + "step": 7720 + }, + { + "epoch": 0.9283953586244214, + "flos": 18726919823520.0, + "grad_norm": 3.15761967013147, + "learning_rate": 5.346538652060939e-08, + "loss": 0.778, + "num_input_tokens_seen": 164729340, + "step": 7721 + }, + { + "epoch": 0.9285156015150604, + "flos": 13299785764800.0, + "grad_norm": 11.79143668625569, + "learning_rate": 5.3286623611705994e-08, + "loss": 0.6949, + "num_input_tokens_seen": 164747105, + "step": 7722 + }, + { + "epoch": 0.9286358444056995, + "flos": 48960991138200.0, + "grad_norm": 0.8942814440681868, + "learning_rate": 5.3108156017673824e-08, + "loss": 0.6432, + "num_input_tokens_seen": 164808585, + "step": 7723 + }, + { + "epoch": 0.9287560872963386, + "flos": 16350348586320.0, + "grad_norm": 2.7947546737635913, + "learning_rate": 5.2929983765586775e-08, + "loss": 0.7074, + "num_input_tokens_seen": 164827085, + "step": 7724 + }, + { + "epoch": 0.9288763301869777, + "flos": 18832677784440.0, + "grad_norm": 3.236333871213592, + "learning_rate": 5.275210688247278e-08, + "loss": 0.6142, + "num_input_tokens_seen": 164847130, + "step": 7725 + }, + { + "epoch": 0.9289965730776167, + "flos": 8924821953480.0, + "grad_norm": 7.5432241387694985, + "learning_rate": 5.257452539531604e-08, + "loss": 0.8268, + "num_input_tokens_seen": 164863920, + "step": 7726 + }, + { + "epoch": 0.9291168159682559, + "flos": 19562286935760.0, + "grad_norm": 3.6042126485430863, + "learning_rate": 5.2397239331055445e-08, + "loss": 0.6775, + "num_input_tokens_seen": 164885640, + "step": 7727 + }, + { + "epoch": 0.929237058858895, + "flos": 10572448898520.0, + "grad_norm": 7.226560916384962, + "learning_rate": 5.2220248716585036e-08, + "loss": 0.7906, + "num_input_tokens_seen": 164903040, + "step": 7728 + }, + { + "epoch": 0.929357301749534, + "flos": 17452057688400.0, + "grad_norm": 7.525718842196862, + "learning_rate": 5.204355357875445e-08, + "loss": 0.7384, + "num_input_tokens_seen": 164921105, + "step": 7729 + }, + { + "epoch": 0.9294775446401732, + "flos": 9411323034240.0, + "grad_norm": 4.254743061130172, + "learning_rate": 5.1867153944367584e-08, + "loss": 0.6935, + "num_input_tokens_seen": 164937215, + "step": 7730 + }, + { + "epoch": 0.9295977875308122, + "flos": 19211394666840.0, + "grad_norm": 3.046616763395575, + "learning_rate": 5.16910498401848e-08, + "loss": 0.7269, + "num_input_tokens_seen": 164956385, + "step": 7731 + }, + { + "epoch": 0.9297180304214513, + "flos": 12003810927720.0, + "grad_norm": 9.338128583756884, + "learning_rate": 5.151524129292073e-08, + "loss": 0.7999, + "num_input_tokens_seen": 164974000, + "step": 7732 + }, + { + "epoch": 0.9298382733120905, + "flos": 17619416948760.0, + "grad_norm": 5.05673247158524, + "learning_rate": 5.1339728329245155e-08, + "loss": 0.6643, + "num_input_tokens_seen": 164994285, + "step": 7733 + }, + { + "epoch": 0.9299585162027295, + "flos": 16189163018160.0, + "grad_norm": 8.161700652679276, + "learning_rate": 5.116451097578367e-08, + "loss": 0.7658, + "num_input_tokens_seen": 165013045, + "step": 7734 + }, + { + "epoch": 0.9300787590933686, + "flos": 15703896675840.0, + "grad_norm": 4.635013154143892, + "learning_rate": 5.0989589259115895e-08, + "loss": 0.7327, + "num_input_tokens_seen": 165033650, + "step": 7735 + }, + { + "epoch": 0.9301990019840077, + "flos": 12970417777320.0, + "grad_norm": 4.889476117114093, + "learning_rate": 5.081496320577816e-08, + "loss": 0.6858, + "num_input_tokens_seen": 165050490, + "step": 7736 + }, + { + "epoch": 0.9303192448746468, + "flos": 43404164188800.0, + "grad_norm": 0.941181900073617, + "learning_rate": 5.0640632842260835e-08, + "loss": 0.6518, + "num_input_tokens_seen": 165110470, + "step": 7737 + }, + { + "epoch": 0.9304394877652858, + "flos": 42487914861480.0, + "grad_norm": 2.7819902099393237, + "learning_rate": 5.0466598195009426e-08, + "loss": 0.7137, + "num_input_tokens_seen": 165137060, + "step": 7738 + }, + { + "epoch": 0.930559730655925, + "flos": 15351706348560.0, + "grad_norm": 14.39658016286306, + "learning_rate": 5.0292859290425036e-08, + "loss": 0.6807, + "num_input_tokens_seen": 165154650, + "step": 7739 + }, + { + "epoch": 0.9306799735465641, + "flos": 17025828188760.0, + "grad_norm": 5.546015971114352, + "learning_rate": 5.011941615486348e-08, + "loss": 0.7591, + "num_input_tokens_seen": 165173485, + "step": 7740 + }, + { + "epoch": 0.9308002164372031, + "flos": 11085539854320.0, + "grad_norm": 12.890854638106088, + "learning_rate": 4.994626881463659e-08, + "loss": 0.8425, + "num_input_tokens_seen": 165189460, + "step": 7741 + }, + { + "epoch": 0.9309204593278423, + "flos": 22642384008600.0, + "grad_norm": 3.5151332935221715, + "learning_rate": 4.9773417296009814e-08, + "loss": 0.6977, + "num_input_tokens_seen": 165210700, + "step": 7742 + }, + { + "epoch": 0.9310407022184813, + "flos": 16861983183960.0, + "grad_norm": 6.902190935896816, + "learning_rate": 4.960086162520527e-08, + "loss": 0.638, + "num_input_tokens_seen": 165230510, + "step": 7743 + }, + { + "epoch": 0.9311609451091204, + "flos": 16189447957800.0, + "grad_norm": 4.015454715428288, + "learning_rate": 4.942860182839936e-08, + "loss": 0.8185, + "num_input_tokens_seen": 165248575, + "step": 7744 + }, + { + "epoch": 0.9312811879997596, + "flos": 15427075239240.0, + "grad_norm": 4.431463514599391, + "learning_rate": 4.925663793172341e-08, + "loss": 0.7748, + "num_input_tokens_seen": 165266255, + "step": 7745 + }, + { + "epoch": 0.9314014308903986, + "flos": 49512128383200.0, + "grad_norm": 0.8454894397173109, + "learning_rate": 4.908496996126477e-08, + "loss": 0.6045, + "num_input_tokens_seen": 165329435, + "step": 7746 + }, + { + "epoch": 0.9315216737810377, + "flos": 16513275452280.0, + "grad_norm": 3.3952707554041166, + "learning_rate": 4.89135979430646e-08, + "loss": 0.7467, + "num_input_tokens_seen": 165349200, + "step": 7747 + }, + { + "epoch": 0.9316419166716768, + "flos": 17562406343520.0, + "grad_norm": 2.9303929814647507, + "learning_rate": 4.874252190312078e-08, + "loss": 0.8282, + "num_input_tokens_seen": 165369305, + "step": 7748 + }, + { + "epoch": 0.9317621595623159, + "flos": 22186208678160.0, + "grad_norm": 9.079299404322992, + "learning_rate": 4.857174186738477e-08, + "loss": 0.6332, + "num_input_tokens_seen": 165392375, + "step": 7749 + }, + { + "epoch": 0.931882402452955, + "flos": 11463908477160.0, + "grad_norm": 5.7191949056359626, + "learning_rate": 4.840125786176408e-08, + "loss": 0.7232, + "num_input_tokens_seen": 165408300, + "step": 7750 + }, + { + "epoch": 0.932002645343594, + "flos": 20806316761200.0, + "grad_norm": 5.231760584845094, + "learning_rate": 4.823106991212067e-08, + "loss": 0.7615, + "num_input_tokens_seen": 165427260, + "step": 7751 + }, + { + "epoch": 0.9321228882342332, + "flos": 11166670857720.0, + "grad_norm": 7.564180297816867, + "learning_rate": 4.806117804427212e-08, + "loss": 0.8267, + "num_input_tokens_seen": 165444915, + "step": 7752 + }, + { + "epoch": 0.9322431311248722, + "flos": 13056994293840.0, + "grad_norm": 3.648130481553122, + "learning_rate": 4.7891582283990926e-08, + "loss": 0.6358, + "num_input_tokens_seen": 165463360, + "step": 7753 + }, + { + "epoch": 0.9323633740155113, + "flos": 17700263012520.0, + "grad_norm": 3.209973861255167, + "learning_rate": 4.772228265700473e-08, + "loss": 0.7107, + "num_input_tokens_seen": 165483940, + "step": 7754 + }, + { + "epoch": 0.9324836169061504, + "flos": 10945561968000.0, + "grad_norm": 4.037027875855525, + "learning_rate": 4.75532791889961e-08, + "loss": 0.7368, + "num_input_tokens_seen": 165500360, + "step": 7755 + }, + { + "epoch": 0.9326038597967895, + "flos": 13596865084440.0, + "grad_norm": 3.5630650657774785, + "learning_rate": 4.738457190560252e-08, + "loss": 0.6376, + "num_input_tokens_seen": 165519190, + "step": 7756 + }, + { + "epoch": 0.9327241026874286, + "flos": 13839339955800.0, + "grad_norm": 6.1161494245120425, + "learning_rate": 4.721616083241664e-08, + "loss": 0.7731, + "num_input_tokens_seen": 165537165, + "step": 7757 + }, + { + "epoch": 0.9328443455780677, + "flos": 21697143140640.0, + "grad_norm": 5.46587727980296, + "learning_rate": 4.7048045994986684e-08, + "loss": 0.7594, + "num_input_tokens_seen": 165557745, + "step": 7758 + }, + { + "epoch": 0.9329645884687068, + "flos": 22075733383200.0, + "grad_norm": 3.4520792348526954, + "learning_rate": 4.688022741881559e-08, + "loss": 0.8856, + "num_input_tokens_seen": 165577990, + "step": 7759 + }, + { + "epoch": 0.9330848313593458, + "flos": 15996385301280.0, + "grad_norm": 2.5867729566142916, + "learning_rate": 4.671270512936076e-08, + "loss": 0.7439, + "num_input_tokens_seen": 165596870, + "step": 7760 + }, + { + "epoch": 0.933205074249985, + "flos": 16188783098640.0, + "grad_norm": 5.302449311801446, + "learning_rate": 4.6545479152035884e-08, + "loss": 0.8121, + "num_input_tokens_seen": 165615760, + "step": 7761 + }, + { + "epoch": 0.9333253171406241, + "flos": 11166765837600.0, + "grad_norm": 4.360171884420881, + "learning_rate": 4.637854951220821e-08, + "loss": 0.7602, + "num_input_tokens_seen": 165632265, + "step": 7762 + }, + { + "epoch": 0.9334455600312631, + "flos": 11436812042760.0, + "grad_norm": 3.6167535047756076, + "learning_rate": 4.621191623520171e-08, + "loss": 0.7312, + "num_input_tokens_seen": 165650415, + "step": 7763 + }, + { + "epoch": 0.9335658029219023, + "flos": 16215657913320.0, + "grad_norm": 15.856687541694258, + "learning_rate": 4.604557934629372e-08, + "loss": 0.8326, + "num_input_tokens_seen": 165669210, + "step": 7764 + }, + { + "epoch": 0.9336860458125413, + "flos": 14812025517720.0, + "grad_norm": 4.567874199332523, + "learning_rate": 4.587953887071805e-08, + "loss": 0.8007, + "num_input_tokens_seen": 165688750, + "step": 7765 + }, + { + "epoch": 0.9338062887031804, + "flos": 15294632423400.0, + "grad_norm": 3.0184012959954467, + "learning_rate": 4.5713794833662554e-08, + "loss": 0.8474, + "num_input_tokens_seen": 165707685, + "step": 7766 + }, + { + "epoch": 0.9339265315938196, + "flos": 16998700094400.0, + "grad_norm": 5.905680634035395, + "learning_rate": 4.5548347260270236e-08, + "loss": 0.6154, + "num_input_tokens_seen": 165726695, + "step": 7767 + }, + { + "epoch": 0.9340467744844586, + "flos": 16701652434720.0, + "grad_norm": 6.473013979062963, + "learning_rate": 4.538319617564012e-08, + "loss": 0.678, + "num_input_tokens_seen": 165745435, + "step": 7768 + }, + { + "epoch": 0.9341670173750977, + "flos": 17323572367560.0, + "grad_norm": 4.227296212020179, + "learning_rate": 4.521834160482485e-08, + "loss": 0.736, + "num_input_tokens_seen": 165763895, + "step": 7769 + }, + { + "epoch": 0.9342872602657368, + "flos": 18184357936320.0, + "grad_norm": 3.2060938033042543, + "learning_rate": 4.5053783572832846e-08, + "loss": 0.8062, + "num_input_tokens_seen": 165783795, + "step": 7770 + }, + { + "epoch": 0.9344075031563759, + "flos": 18885477615000.0, + "grad_norm": 3.996459944158844, + "learning_rate": 4.488952210462771e-08, + "loss": 0.7536, + "num_input_tokens_seen": 165803720, + "step": 7771 + }, + { + "epoch": 0.9345277460470149, + "flos": 18722044189680.0, + "grad_norm": 3.139173181613298, + "learning_rate": 4.4725557225127495e-08, + "loss": 0.8427, + "num_input_tokens_seen": 165821780, + "step": 7772 + }, + { + "epoch": 0.9346479889376541, + "flos": 25207110608520.0, + "grad_norm": 4.386932014587227, + "learning_rate": 4.456188895920565e-08, + "loss": 0.7864, + "num_input_tokens_seen": 165843255, + "step": 7773 + }, + { + "epoch": 0.9347682318282932, + "flos": 13943071679280.0, + "grad_norm": 5.909716360963613, + "learning_rate": 4.439851733169031e-08, + "loss": 0.8267, + "num_input_tokens_seen": 165860765, + "step": 7774 + }, + { + "epoch": 0.9348884747189322, + "flos": 19239535879920.0, + "grad_norm": 4.2440208562110175, + "learning_rate": 4.4235442367365204e-08, + "loss": 0.6669, + "num_input_tokens_seen": 165880795, + "step": 7775 + }, + { + "epoch": 0.9350087176095714, + "flos": 13380916768200.0, + "grad_norm": 6.658775459414177, + "learning_rate": 4.4072664090968545e-08, + "loss": 0.7634, + "num_input_tokens_seen": 165898900, + "step": 7776 + }, + { + "epoch": 0.9351289605002104, + "flos": 14110019360160.0, + "grad_norm": 4.987028844468081, + "learning_rate": 4.391018252719347e-08, + "loss": 0.8198, + "num_input_tokens_seen": 165918415, + "step": 7777 + }, + { + "epoch": 0.9352492033908495, + "flos": 13725413725200.0, + "grad_norm": 4.261696148470659, + "learning_rate": 4.374799770068849e-08, + "loss": 0.6693, + "num_input_tokens_seen": 165934810, + "step": 7778 + }, + { + "epoch": 0.9353694462814887, + "flos": 21666912370200.0, + "grad_norm": 14.252274181194295, + "learning_rate": 4.358610963605658e-08, + "loss": 0.7375, + "num_input_tokens_seen": 165954980, + "step": 7779 + }, + { + "epoch": 0.9354896891721277, + "flos": 22507724995560.0, + "grad_norm": 5.189676035212975, + "learning_rate": 4.342451835785677e-08, + "loss": 0.6638, + "num_input_tokens_seen": 165975610, + "step": 7780 + }, + { + "epoch": 0.9356099320627668, + "flos": 14217771898560.0, + "grad_norm": 3.2207737458619614, + "learning_rate": 4.3263223890601665e-08, + "loss": 0.7423, + "num_input_tokens_seen": 165994040, + "step": 7781 + }, + { + "epoch": 0.9357301749534058, + "flos": 13946965854360.0, + "grad_norm": 4.036749821464509, + "learning_rate": 4.31022262587597e-08, + "loss": 0.7764, + "num_input_tokens_seen": 166012435, + "step": 7782 + }, + { + "epoch": 0.935850417844045, + "flos": 17242029784680.0, + "grad_norm": 3.591449830722501, + "learning_rate": 4.2941525486754225e-08, + "loss": 0.6452, + "num_input_tokens_seen": 166032475, + "step": 7783 + }, + { + "epoch": 0.935970660734684, + "flos": 13569800310000.0, + "grad_norm": 4.279269686272596, + "learning_rate": 4.278112159896286e-08, + "loss": 0.7776, + "num_input_tokens_seen": 166050035, + "step": 7784 + }, + { + "epoch": 0.9360909036253231, + "flos": 15081660143400.0, + "grad_norm": 3.06934914265971, + "learning_rate": 4.2621014619719896e-08, + "loss": 0.6637, + "num_input_tokens_seen": 166067520, + "step": 7785 + }, + { + "epoch": 0.9362111465159623, + "flos": 45548708341440.0, + "grad_norm": 0.8205345943069108, + "learning_rate": 4.246120457331215e-08, + "loss": 0.6206, + "num_input_tokens_seen": 166129415, + "step": 7786 + }, + { + "epoch": 0.9363313894066013, + "flos": 17700864551760.0, + "grad_norm": 4.167496229878295, + "learning_rate": 4.2301691483983325e-08, + "loss": 0.7084, + "num_input_tokens_seen": 166149255, + "step": 7787 + }, + { + "epoch": 0.9364516322972404, + "flos": 14703544800240.0, + "grad_norm": 4.669165396435168, + "learning_rate": 4.214247537593163e-08, + "loss": 0.7432, + "num_input_tokens_seen": 166168225, + "step": 7788 + }, + { + "epoch": 0.9365718751878795, + "flos": 15135251472960.0, + "grad_norm": 4.552566584142593, + "learning_rate": 4.1983556273309293e-08, + "loss": 0.7984, + "num_input_tokens_seen": 166186695, + "step": 7789 + }, + { + "epoch": 0.9366921180785186, + "flos": 13618484345760.0, + "grad_norm": 7.275591440181572, + "learning_rate": 4.182493420022526e-08, + "loss": 0.6836, + "num_input_tokens_seen": 166202085, + "step": 7790 + }, + { + "epoch": 0.9368123609691577, + "flos": 18888232031520.0, + "grad_norm": 14.27372239470057, + "learning_rate": 4.166660918074139e-08, + "loss": 0.7706, + "num_input_tokens_seen": 166221710, + "step": 7791 + }, + { + "epoch": 0.9369326038597968, + "flos": 18723880467360.0, + "grad_norm": 2.9125671640927977, + "learning_rate": 4.15085812388758e-08, + "loss": 0.7208, + "num_input_tokens_seen": 166243650, + "step": 7792 + }, + { + "epoch": 0.9370528467504359, + "flos": 16997275396200.0, + "grad_norm": 3.992399146634129, + "learning_rate": 4.135085039860153e-08, + "loss": 0.7815, + "num_input_tokens_seen": 166262770, + "step": 7793 + }, + { + "epoch": 0.9371730896410749, + "flos": 18289957597440.0, + "grad_norm": 25.053373201807826, + "learning_rate": 4.1193416683845906e-08, + "loss": 0.7719, + "num_input_tokens_seen": 166281420, + "step": 7794 + }, + { + "epoch": 0.9372933325317141, + "flos": 11322537552600.0, + "grad_norm": 5.985837434008458, + "learning_rate": 4.103628011849136e-08, + "loss": 0.816, + "num_input_tokens_seen": 166296500, + "step": 7795 + }, + { + "epoch": 0.9374135754223532, + "flos": 15999804576960.0, + "grad_norm": 3.96228124935303, + "learning_rate": 4.0879440726375506e-08, + "loss": 0.7345, + "num_input_tokens_seen": 166314005, + "step": 7796 + }, + { + "epoch": 0.9375338183129922, + "flos": 16560661429680.0, + "grad_norm": 19.59165658673355, + "learning_rate": 4.0722898531291074e-08, + "loss": 0.5389, + "num_input_tokens_seen": 166330965, + "step": 7797 + }, + { + "epoch": 0.9376540612036314, + "flos": 19132099941120.0, + "grad_norm": 3.4563102013309956, + "learning_rate": 4.0566653556985295e-08, + "loss": 0.7483, + "num_input_tokens_seen": 166351230, + "step": 7798 + }, + { + "epoch": 0.9377743040942704, + "flos": 14405262402120.0, + "grad_norm": 5.7613545965431925, + "learning_rate": 4.0410705827159886e-08, + "loss": 0.7745, + "num_input_tokens_seen": 166368245, + "step": 7799 + }, + { + "epoch": 0.9378945469849095, + "flos": 11112098069400.0, + "grad_norm": 6.542395030816911, + "learning_rate": 4.0255055365472356e-08, + "loss": 0.6917, + "num_input_tokens_seen": 166386060, + "step": 7800 + }, + { + "epoch": 0.9380147898755486, + "flos": 15051682652640.0, + "grad_norm": 7.548166863421004, + "learning_rate": 4.009970219553471e-08, + "loss": 0.7288, + "num_input_tokens_seen": 166402730, + "step": 7801 + }, + { + "epoch": 0.9381350327661877, + "flos": 19262168259960.0, + "grad_norm": 5.3143077691670895, + "learning_rate": 3.99446463409141e-08, + "loss": 0.7336, + "num_input_tokens_seen": 166420305, + "step": 7802 + }, + { + "epoch": 0.9382552756568268, + "flos": 17268587999760.0, + "grad_norm": 10.002607005757104, + "learning_rate": 3.978988782513215e-08, + "loss": 0.6648, + "num_input_tokens_seen": 166437520, + "step": 7803 + }, + { + "epoch": 0.9383755185474659, + "flos": 20859718131000.0, + "grad_norm": 4.440772985238144, + "learning_rate": 3.963542667166586e-08, + "loss": 0.7475, + "num_input_tokens_seen": 166457345, + "step": 7804 + }, + { + "epoch": 0.938495761438105, + "flos": 14947317729960.0, + "grad_norm": 3.163323511551247, + "learning_rate": 3.9481262903946486e-08, + "loss": 0.6651, + "num_input_tokens_seen": 166476510, + "step": 7805 + }, + { + "epoch": 0.938616004328744, + "flos": 51106892177760.0, + "grad_norm": 0.7843668751086836, + "learning_rate": 3.932739654536066e-08, + "loss": 0.5636, + "num_input_tokens_seen": 166538930, + "step": 7806 + }, + { + "epoch": 0.9387362472193832, + "flos": 13808349346320.0, + "grad_norm": 4.233862814243511, + "learning_rate": 3.917382761925014e-08, + "loss": 0.7301, + "num_input_tokens_seen": 166554485, + "step": 7807 + }, + { + "epoch": 0.9388564901100223, + "flos": 19425886624920.0, + "grad_norm": 4.126141939178304, + "learning_rate": 3.9020556148910754e-08, + "loss": 0.7718, + "num_input_tokens_seen": 166573560, + "step": 7808 + }, + { + "epoch": 0.9389767330006613, + "flos": 43438637393880.0, + "grad_norm": 0.7213836599625697, + "learning_rate": 3.8867582157593895e-08, + "loss": 0.5797, + "num_input_tokens_seen": 166627485, + "step": 7809 + }, + { + "epoch": 0.9390969758913005, + "flos": 22831109250600.0, + "grad_norm": 2.772260905242037, + "learning_rate": 3.871490566850544e-08, + "loss": 0.7398, + "num_input_tokens_seen": 166651415, + "step": 7810 + }, + { + "epoch": 0.9392172187819395, + "flos": 16405048014480.0, + "grad_norm": 11.958833017221583, + "learning_rate": 3.856252670480642e-08, + "loss": 0.6923, + "num_input_tokens_seen": 166669795, + "step": 7811 + }, + { + "epoch": 0.9393374616725786, + "flos": 14406940380000.0, + "grad_norm": 5.747425332200389, + "learning_rate": 3.841044528961279e-08, + "loss": 0.7984, + "num_input_tokens_seen": 166687310, + "step": 7812 + }, + { + "epoch": 0.9394577045632178, + "flos": 17700706251960.0, + "grad_norm": 3.6948388987769194, + "learning_rate": 3.825866144599477e-08, + "loss": 0.7766, + "num_input_tokens_seen": 166706085, + "step": 7813 + }, + { + "epoch": 0.9395779474538568, + "flos": 13891854846720.0, + "grad_norm": 4.438476240202133, + "learning_rate": 3.8107175196978145e-08, + "loss": 0.7456, + "num_input_tokens_seen": 166722110, + "step": 7814 + }, + { + "epoch": 0.9396981903444959, + "flos": 10410250211640.0, + "grad_norm": 5.091351277790829, + "learning_rate": 3.7955986565542996e-08, + "loss": 0.731, + "num_input_tokens_seen": 166739910, + "step": 7815 + }, + { + "epoch": 0.9398184332351349, + "flos": 25558256157120.0, + "grad_norm": 5.67337966170859, + "learning_rate": 3.780509557462497e-08, + "loss": 0.6789, + "num_input_tokens_seen": 166759830, + "step": 7816 + }, + { + "epoch": 0.9399386761257741, + "flos": 18590012953320.0, + "grad_norm": 4.138019706609088, + "learning_rate": 3.765450224711375e-08, + "loss": 0.7393, + "num_input_tokens_seen": 166780055, + "step": 7817 + }, + { + "epoch": 0.9400589190164131, + "flos": 20020203564000.0, + "grad_norm": 6.963187559411913, + "learning_rate": 3.750420660585396e-08, + "loss": 0.7788, + "num_input_tokens_seen": 166801715, + "step": 7818 + }, + { + "epoch": 0.9401791619070522, + "flos": 17130003151680.0, + "grad_norm": 2.9008400719737972, + "learning_rate": 3.735420867364603e-08, + "loss": 0.7862, + "num_input_tokens_seen": 166822415, + "step": 7819 + }, + { + "epoch": 0.9402994047976914, + "flos": 26367666593520.0, + "grad_norm": 4.920576620659936, + "learning_rate": 3.7204508473244186e-08, + "loss": 0.6084, + "num_input_tokens_seen": 166845760, + "step": 7820 + }, + { + "epoch": 0.9404196476883304, + "flos": 16269914102040.0, + "grad_norm": 3.5310775579697977, + "learning_rate": 3.7055106027357395e-08, + "loss": 0.6843, + "num_input_tokens_seen": 166865345, + "step": 7821 + }, + { + "epoch": 0.9405398905789695, + "flos": 13812021901680.0, + "grad_norm": 3.862485057221677, + "learning_rate": 3.690600135865063e-08, + "loss": 0.705, + "num_input_tokens_seen": 166881990, + "step": 7822 + }, + { + "epoch": 0.9406601334696086, + "flos": 51823869005040.0, + "grad_norm": 0.7703328097066691, + "learning_rate": 3.675719448974246e-08, + "loss": 0.5925, + "num_input_tokens_seen": 166946800, + "step": 7823 + }, + { + "epoch": 0.9407803763602477, + "flos": 16216987631640.0, + "grad_norm": 5.54964788196238, + "learning_rate": 3.6608685443207054e-08, + "loss": 0.5837, + "num_input_tokens_seen": 166965670, + "step": 7824 + }, + { + "epoch": 0.9409006192508867, + "flos": 13785178746960.0, + "grad_norm": 8.510297035923507, + "learning_rate": 3.646047424157306e-08, + "loss": 0.6542, + "num_input_tokens_seen": 166982365, + "step": 7825 + }, + { + "epoch": 0.9410208621415259, + "flos": 17107307451720.0, + "grad_norm": 5.196046889177977, + "learning_rate": 3.631256090732382e-08, + "loss": 0.6548, + "num_input_tokens_seen": 167002545, + "step": 7826 + }, + { + "epoch": 0.941141105032165, + "flos": 16643185471320.0, + "grad_norm": 7.099330236517309, + "learning_rate": 3.6164945462897833e-08, + "loss": 0.814, + "num_input_tokens_seen": 167021555, + "step": 7827 + }, + { + "epoch": 0.941261347922804, + "flos": 15136264591680.0, + "grad_norm": 3.882172007857997, + "learning_rate": 3.6017627930687856e-08, + "loss": 0.743, + "num_input_tokens_seen": 167041100, + "step": 7828 + }, + { + "epoch": 0.9413815908134432, + "flos": 14185989790080.0, + "grad_norm": 9.388236099011424, + "learning_rate": 3.587060833304267e-08, + "loss": 0.7528, + "num_input_tokens_seen": 167059010, + "step": 7829 + }, + { + "epoch": 0.9415018337040822, + "flos": 12759471734760.0, + "grad_norm": 12.042343455020367, + "learning_rate": 3.5723886692264225e-08, + "loss": 0.6244, + "num_input_tokens_seen": 167076270, + "step": 7830 + }, + { + "epoch": 0.9416220765947213, + "flos": 23370568461720.0, + "grad_norm": 8.342026207410115, + "learning_rate": 3.557746303061071e-08, + "loss": 0.6041, + "num_input_tokens_seen": 167097745, + "step": 7831 + }, + { + "epoch": 0.9417423194853605, + "flos": 17213350352280.0, + "grad_norm": 7.714743373555871, + "learning_rate": 3.543133737029391e-08, + "loss": 0.7145, + "num_input_tokens_seen": 167117975, + "step": 7832 + }, + { + "epoch": 0.9418625623759995, + "flos": 17511949350000.0, + "grad_norm": 3.580729906486666, + "learning_rate": 3.5285509733481214e-08, + "loss": 0.666, + "num_input_tokens_seen": 167137420, + "step": 7833 + }, + { + "epoch": 0.9419828052666386, + "flos": 13191273387360.0, + "grad_norm": 3.4818524864891818, + "learning_rate": 3.513998014229469e-08, + "loss": 0.7459, + "num_input_tokens_seen": 167156090, + "step": 7834 + }, + { + "epoch": 0.9421030481572777, + "flos": 12921448801920.0, + "grad_norm": 5.38931656590011, + "learning_rate": 3.499474861881069e-08, + "loss": 0.8483, + "num_input_tokens_seen": 167173035, + "step": 7835 + }, + { + "epoch": 0.9422232910479168, + "flos": 14758244228400.0, + "grad_norm": 3.9255491273493077, + "learning_rate": 3.4849815185061136e-08, + "loss": 0.6572, + "num_input_tokens_seen": 167192645, + "step": 7836 + }, + { + "epoch": 0.9423435339385559, + "flos": 13461952791720.0, + "grad_norm": 5.191018255817115, + "learning_rate": 3.470517986303223e-08, + "loss": 0.7552, + "num_input_tokens_seen": 167211350, + "step": 7837 + }, + { + "epoch": 0.942463776829195, + "flos": 14673693949320.0, + "grad_norm": 9.405692353304968, + "learning_rate": 3.4560842674664856e-08, + "loss": 0.7731, + "num_input_tokens_seen": 167229585, + "step": 7838 + }, + { + "epoch": 0.9425840197198341, + "flos": 16565252123880.0, + "grad_norm": 5.461510274861879, + "learning_rate": 3.441680364185506e-08, + "loss": 0.7413, + "num_input_tokens_seen": 167249175, + "step": 7839 + }, + { + "epoch": 0.9427042626104731, + "flos": 14568505867680.0, + "grad_norm": 7.4773158029758005, + "learning_rate": 3.427306278645314e-08, + "loss": 0.7364, + "num_input_tokens_seen": 167267350, + "step": 7840 + }, + { + "epoch": 0.9428245055011123, + "flos": 16728875508960.0, + "grad_norm": 4.207945827860663, + "learning_rate": 3.4129620130264767e-08, + "loss": 0.7029, + "num_input_tokens_seen": 167285430, + "step": 7841 + }, + { + "epoch": 0.9429447483917514, + "flos": 15317739702840.0, + "grad_norm": 4.969117157044556, + "learning_rate": 3.398647569505009e-08, + "loss": 0.7748, + "num_input_tokens_seen": 167302575, + "step": 7842 + }, + { + "epoch": 0.9430649912823904, + "flos": 13758747171720.0, + "grad_norm": 6.077787649800251, + "learning_rate": 3.384362950252373e-08, + "loss": 0.7224, + "num_input_tokens_seen": 167319265, + "step": 7843 + }, + { + "epoch": 0.9431852341730296, + "flos": 23905943538000.0, + "grad_norm": 4.888269021493115, + "learning_rate": 3.3701081574355473e-08, + "loss": 0.5521, + "num_input_tokens_seen": 167340945, + "step": 7844 + }, + { + "epoch": 0.9433054770636686, + "flos": 49333439348520.0, + "grad_norm": 0.6557257692846429, + "learning_rate": 3.3558831932169796e-08, + "loss": 0.5276, + "num_input_tokens_seen": 167409335, + "step": 7845 + }, + { + "epoch": 0.9434257199543077, + "flos": 19158721476120.0, + "grad_norm": 5.04573088630398, + "learning_rate": 3.341688059754588e-08, + "loss": 0.8694, + "num_input_tokens_seen": 167424710, + "step": 7846 + }, + { + "epoch": 0.9435459628449467, + "flos": 18316547472480.0, + "grad_norm": 4.28587859610794, + "learning_rate": 3.327522759201762e-08, + "loss": 0.7554, + "num_input_tokens_seen": 167444300, + "step": 7847 + }, + { + "epoch": 0.9436662057355859, + "flos": 16215372973680.0, + "grad_norm": 11.418491264130452, + "learning_rate": 3.313387293707359e-08, + "loss": 0.65, + "num_input_tokens_seen": 167462725, + "step": 7848 + }, + { + "epoch": 0.943786448626225, + "flos": 14701930142280.0, + "grad_norm": 5.521225460689733, + "learning_rate": 3.29928166541571e-08, + "loss": 0.6641, + "num_input_tokens_seen": 167481400, + "step": 7849 + }, + { + "epoch": 0.943906691516864, + "flos": 16161844964040.0, + "grad_norm": 4.5716022986371305, + "learning_rate": 3.2852058764666346e-08, + "loss": 0.7786, + "num_input_tokens_seen": 167500220, + "step": 7850 + }, + { + "epoch": 0.9440269344075032, + "flos": 25936403160240.0, + "grad_norm": 4.326929912072676, + "learning_rate": 3.2711599289954264e-08, + "loss": 0.6747, + "num_input_tokens_seen": 167523975, + "step": 7851 + }, + { + "epoch": 0.9441471772981422, + "flos": 14049716119080.0, + "grad_norm": 3.600762466643214, + "learning_rate": 3.257143825132847e-08, + "loss": 0.7509, + "num_input_tokens_seen": 167541865, + "step": 7852 + }, + { + "epoch": 0.9442674201887813, + "flos": 18862401995520.0, + "grad_norm": 3.1980820570599757, + "learning_rate": 3.243157567005106e-08, + "loss": 0.7375, + "num_input_tokens_seen": 167559765, + "step": 7853 + }, + { + "epoch": 0.9443876630794205, + "flos": 11301456510600.0, + "grad_norm": 4.038576550645032, + "learning_rate": 3.2292011567339296e-08, + "loss": 0.6171, + "num_input_tokens_seen": 167577290, + "step": 7854 + }, + { + "epoch": 0.9445079059700595, + "flos": 9729768335520.0, + "grad_norm": 4.678822942231931, + "learning_rate": 3.21527459643649e-08, + "loss": 0.5291, + "num_input_tokens_seen": 167593895, + "step": 7855 + }, + { + "epoch": 0.9446281488606986, + "flos": 17322527588880.0, + "grad_norm": 3.8406097486300235, + "learning_rate": 3.2013778882254536e-08, + "loss": 0.7257, + "num_input_tokens_seen": 167612410, + "step": 7856 + }, + { + "epoch": 0.9447483917513377, + "flos": 18727363062960.0, + "grad_norm": 3.7119226693944545, + "learning_rate": 3.1875110342088676e-08, + "loss": 0.7483, + "num_input_tokens_seen": 167633580, + "step": 7857 + }, + { + "epoch": 0.9448686346419768, + "flos": 17971069056720.0, + "grad_norm": 3.8061559455663683, + "learning_rate": 3.1736740364904035e-08, + "loss": 0.6366, + "num_input_tokens_seen": 167653830, + "step": 7858 + }, + { + "epoch": 0.9449888775326158, + "flos": 10707962730480.0, + "grad_norm": 4.621898951800471, + "learning_rate": 3.159866897169094e-08, + "loss": 0.7387, + "num_input_tokens_seen": 167671750, + "step": 7859 + }, + { + "epoch": 0.945109120423255, + "flos": 11245269064320.0, + "grad_norm": 5.118445880877674, + "learning_rate": 3.146089618339487e-08, + "loss": 0.7432, + "num_input_tokens_seen": 167688325, + "step": 7860 + }, + { + "epoch": 0.9452293633138941, + "flos": 18646548659160.0, + "grad_norm": 5.0113897297371945, + "learning_rate": 3.132342202091554e-08, + "loss": 0.6656, + "num_input_tokens_seen": 167708270, + "step": 7861 + }, + { + "epoch": 0.9453496062045331, + "flos": 15513968355360.0, + "grad_norm": 6.640955586899955, + "learning_rate": 3.1186246505107595e-08, + "loss": 0.6558, + "num_input_tokens_seen": 167727130, + "step": 7862 + }, + { + "epoch": 0.9454698490951723, + "flos": 14622540436680.0, + "grad_norm": 4.755471140467815, + "learning_rate": 3.104936965678084e-08, + "loss": 0.822, + "num_input_tokens_seen": 167745180, + "step": 7863 + }, + { + "epoch": 0.9455900919858113, + "flos": 15406184157000.0, + "grad_norm": 3.611731350443762, + "learning_rate": 3.091279149669956e-08, + "loss": 0.7865, + "num_input_tokens_seen": 167763690, + "step": 7864 + }, + { + "epoch": 0.9457103348764504, + "flos": 15162791146800.0, + "grad_norm": 4.463343410770206, + "learning_rate": 3.0776512045581624e-08, + "loss": 0.7212, + "num_input_tokens_seen": 167782200, + "step": 7865 + }, + { + "epoch": 0.9458305777670896, + "flos": 15670784849040.0, + "grad_norm": 3.3485524993102924, + "learning_rate": 3.0640531324101384e-08, + "loss": 0.7549, + "num_input_tokens_seen": 167799685, + "step": 7866 + }, + { + "epoch": 0.9459508206577286, + "flos": 14622793716360.0, + "grad_norm": 3.746374157543278, + "learning_rate": 3.0504849352886554e-08, + "loss": 0.7382, + "num_input_tokens_seen": 167817550, + "step": 7867 + }, + { + "epoch": 0.9460710635483677, + "flos": 8817101075040.0, + "grad_norm": 6.930241266980154, + "learning_rate": 3.036946615252023e-08, + "loss": 0.6994, + "num_input_tokens_seen": 167832800, + "step": 7868 + }, + { + "epoch": 0.9461913064390068, + "flos": 25179729234480.0, + "grad_norm": 30.975658088940882, + "learning_rate": 3.0234381743539984e-08, + "loss": 0.6559, + "num_input_tokens_seen": 167850135, + "step": 7869 + }, + { + "epoch": 0.9463115493296459, + "flos": 14217645258720.0, + "grad_norm": 4.708349627351277, + "learning_rate": 3.0099596146437863e-08, + "loss": 0.79, + "num_input_tokens_seen": 167866960, + "step": 7870 + }, + { + "epoch": 0.946431792220285, + "flos": 52044123075840.0, + "grad_norm": 0.8402369089057344, + "learning_rate": 2.996510938166086e-08, + "loss": 0.6201, + "num_input_tokens_seen": 167929655, + "step": 7871 + }, + { + "epoch": 0.9465520351109241, + "flos": 13835160841080.0, + "grad_norm": 5.0976727130960215, + "learning_rate": 2.983092146960997e-08, + "loss": 0.7267, + "num_input_tokens_seen": 167946720, + "step": 7872 + }, + { + "epoch": 0.9466722780015632, + "flos": 13974853787760.0, + "grad_norm": 11.108933409164798, + "learning_rate": 2.9697032430642256e-08, + "loss": 0.7748, + "num_input_tokens_seen": 167964655, + "step": 7873 + }, + { + "epoch": 0.9467925208922022, + "flos": 12570334913280.0, + "grad_norm": 4.580813298872371, + "learning_rate": 2.9563442285067906e-08, + "loss": 0.7055, + "num_input_tokens_seen": 167981420, + "step": 7874 + }, + { + "epoch": 0.9469127637828414, + "flos": 21400190460840.0, + "grad_norm": 4.330928183926186, + "learning_rate": 2.943015105315294e-08, + "loss": 0.785, + "num_input_tokens_seen": 168001335, + "step": 7875 + }, + { + "epoch": 0.9470330066734804, + "flos": 19077337193040.0, + "grad_norm": 9.020380045460989, + "learning_rate": 2.929715875511718e-08, + "loss": 0.6597, + "num_input_tokens_seen": 168020090, + "step": 7876 + }, + { + "epoch": 0.9471532495641195, + "flos": 17161025421120.0, + "grad_norm": 6.96324645172501, + "learning_rate": 2.9164465411135375e-08, + "loss": 0.6856, + "num_input_tokens_seen": 168039580, + "step": 7877 + }, + { + "epoch": 0.9472734924547586, + "flos": 11518038026040.0, + "grad_norm": 6.211929645107991, + "learning_rate": 2.9032071041337426e-08, + "loss": 0.7886, + "num_input_tokens_seen": 168057535, + "step": 7878 + }, + { + "epoch": 0.9473937353453977, + "flos": 8088473382480.0, + "grad_norm": 4.504213268075419, + "learning_rate": 2.889997566580704e-08, + "loss": 0.7141, + "num_input_tokens_seen": 168075410, + "step": 7879 + }, + { + "epoch": 0.9475139782360368, + "flos": 18888738590880.0, + "grad_norm": 5.099725506445309, + "learning_rate": 2.8768179304583086e-08, + "loss": 0.6924, + "num_input_tokens_seen": 168097185, + "step": 7880 + }, + { + "epoch": 0.9476342211266758, + "flos": 16701462474960.0, + "grad_norm": 5.39719828226606, + "learning_rate": 2.8636681977659117e-08, + "loss": 0.7252, + "num_input_tokens_seen": 168116555, + "step": 7881 + }, + { + "epoch": 0.947754464017315, + "flos": 14757800988960.0, + "grad_norm": 11.999858219580725, + "learning_rate": 2.850548370498318e-08, + "loss": 0.7661, + "num_input_tokens_seen": 168134115, + "step": 7882 + }, + { + "epoch": 0.9478747069079541, + "flos": 18106107989280.0, + "grad_norm": 3.5606899013031432, + "learning_rate": 2.8374584506457798e-08, + "loss": 0.694, + "num_input_tokens_seen": 168155110, + "step": 7883 + }, + { + "epoch": 0.9479949497985931, + "flos": 15622037493360.0, + "grad_norm": 4.360834786528881, + "learning_rate": 2.824398440193998e-08, + "loss": 0.6501, + "num_input_tokens_seen": 168173630, + "step": 7884 + }, + { + "epoch": 0.9481151926892323, + "flos": 13244104877880.0, + "grad_norm": 4.917998613019936, + "learning_rate": 2.811368341124232e-08, + "loss": 0.6879, + "num_input_tokens_seen": 168192420, + "step": 7885 + }, + { + "epoch": 0.9482354355798713, + "flos": 16189479617760.0, + "grad_norm": 9.502457150365439, + "learning_rate": 2.7983681554131222e-08, + "loss": 0.6639, + "num_input_tokens_seen": 168212400, + "step": 7886 + }, + { + "epoch": 0.9483556784705104, + "flos": 13921199138280.0, + "grad_norm": 6.386309894692311, + "learning_rate": 2.7853978850327365e-08, + "loss": 0.6629, + "num_input_tokens_seen": 168231290, + "step": 7887 + }, + { + "epoch": 0.9484759213611496, + "flos": 18890131629120.0, + "grad_norm": 3.6569625290621786, + "learning_rate": 2.7724575319507225e-08, + "loss": 0.8577, + "num_input_tokens_seen": 168250720, + "step": 7888 + }, + { + "epoch": 0.9485961642517886, + "flos": 15108661597920.0, + "grad_norm": 5.658266161303163, + "learning_rate": 2.759547098130044e-08, + "loss": 0.7684, + "num_input_tokens_seen": 168269170, + "step": 7889 + }, + { + "epoch": 0.9487164071424277, + "flos": 16593424996920.0, + "grad_norm": 3.911045396377968, + "learning_rate": 2.746666585529267e-08, + "loss": 0.7516, + "num_input_tokens_seen": 168289165, + "step": 7890 + }, + { + "epoch": 0.9488366500330668, + "flos": 28473083526960.0, + "grad_norm": 5.159108057631414, + "learning_rate": 2.73381599610234e-08, + "loss": 0.7213, + "num_input_tokens_seen": 168309285, + "step": 7891 + }, + { + "epoch": 0.9489568929237059, + "flos": 20454158093880.0, + "grad_norm": 3.341900750869497, + "learning_rate": 2.7209953317987033e-08, + "loss": 0.6998, + "num_input_tokens_seen": 168330045, + "step": 7892 + }, + { + "epoch": 0.9490771358143449, + "flos": 24664643701200.0, + "grad_norm": 16.003646435070664, + "learning_rate": 2.7082045945631793e-08, + "loss": 0.7566, + "num_input_tokens_seen": 168351980, + "step": 7893 + }, + { + "epoch": 0.9491973787049841, + "flos": 10761585720000.0, + "grad_norm": 7.626211101468419, + "learning_rate": 2.6954437863361712e-08, + "loss": 0.675, + "num_input_tokens_seen": 168369615, + "step": 7894 + }, + { + "epoch": 0.9493176215956232, + "flos": 18560162102400.0, + "grad_norm": 4.283699170432452, + "learning_rate": 2.6827129090534862e-08, + "loss": 0.6968, + "num_input_tokens_seen": 168389635, + "step": 7895 + }, + { + "epoch": 0.9494378644862622, + "flos": 15376364966040.0, + "grad_norm": 3.4117576098996976, + "learning_rate": 2.670011964646335e-08, + "loss": 0.7692, + "num_input_tokens_seen": 168408035, + "step": 7896 + }, + { + "epoch": 0.9495581073769014, + "flos": 11058601719720.0, + "grad_norm": 7.343347936134635, + "learning_rate": 2.657340955041487e-08, + "loss": 0.6617, + "num_input_tokens_seen": 168426530, + "step": 7897 + }, + { + "epoch": 0.9496783502675404, + "flos": 20990609608800.0, + "grad_norm": 6.410131850768616, + "learning_rate": 2.6446998821611167e-08, + "loss": 0.6962, + "num_input_tokens_seen": 168446445, + "step": 7898 + }, + { + "epoch": 0.9497985931581795, + "flos": 10815683608920.0, + "grad_norm": 6.653148232881288, + "learning_rate": 2.6320887479228228e-08, + "loss": 0.6939, + "num_input_tokens_seen": 168462765, + "step": 7899 + }, + { + "epoch": 0.9499188360488187, + "flos": 19937267942880.0, + "grad_norm": 7.293291212491455, + "learning_rate": 2.619507554239786e-08, + "loss": 0.7006, + "num_input_tokens_seen": 168481045, + "step": 7900 + }, + { + "epoch": 0.9500390789394577, + "flos": 17808997009680.0, + "grad_norm": 4.890989052382193, + "learning_rate": 2.606956303020502e-08, + "loss": 0.6982, + "num_input_tokens_seen": 168501570, + "step": 7901 + }, + { + "epoch": 0.9501593218300968, + "flos": 10436491827120.0, + "grad_norm": 3.4079448438130098, + "learning_rate": 2.5944349961690036e-08, + "loss": 0.8234, + "num_input_tokens_seen": 168518310, + "step": 7902 + }, + { + "epoch": 0.9502795647207359, + "flos": 28474539885120.0, + "grad_norm": 12.677990586544068, + "learning_rate": 2.581943635584749e-08, + "loss": 0.7118, + "num_input_tokens_seen": 168540860, + "step": 7903 + }, + { + "epoch": 0.950399807611375, + "flos": 30013844412480.0, + "grad_norm": 4.615088280921421, + "learning_rate": 2.569482223162689e-08, + "loss": 0.6426, + "num_input_tokens_seen": 168564555, + "step": 7904 + }, + { + "epoch": 0.950520050502014, + "flos": 17160582181680.0, + "grad_norm": 3.3747349387740693, + "learning_rate": 2.5570507607932e-08, + "loss": 0.7171, + "num_input_tokens_seen": 168584190, + "step": 7905 + }, + { + "epoch": 0.9506402933926532, + "flos": 12974058672720.0, + "grad_norm": 6.833409409758569, + "learning_rate": 2.54464925036213e-08, + "loss": 0.5929, + "num_input_tokens_seen": 168601200, + "step": 7906 + }, + { + "epoch": 0.9507605362832923, + "flos": 23911135771440.0, + "grad_norm": 3.5895229524185996, + "learning_rate": 2.532277693750773e-08, + "loss": 0.5854, + "num_input_tokens_seen": 168621845, + "step": 7907 + }, + { + "epoch": 0.9508807791739313, + "flos": 14318907505320.0, + "grad_norm": 5.284463841319189, + "learning_rate": 2.5199360928358948e-08, + "loss": 0.7392, + "num_input_tokens_seen": 168638800, + "step": 7908 + }, + { + "epoch": 0.9510010220645704, + "flos": 15703959995760.0, + "grad_norm": 3.4821521186006787, + "learning_rate": 2.507624449489665e-08, + "loss": 0.8578, + "num_input_tokens_seen": 168657150, + "step": 7909 + }, + { + "epoch": 0.9511212649552095, + "flos": 13785621986400.0, + "grad_norm": 5.452337082392667, + "learning_rate": 2.495342765579811e-08, + "loss": 0.6384, + "num_input_tokens_seen": 168675530, + "step": 7910 + }, + { + "epoch": 0.9512415078458486, + "flos": 15213881339520.0, + "grad_norm": 4.852905677216784, + "learning_rate": 2.4830910429693984e-08, + "loss": 0.6935, + "num_input_tokens_seen": 168695210, + "step": 7911 + }, + { + "epoch": 0.9513617507364877, + "flos": 13407918222720.0, + "grad_norm": 3.339794064871658, + "learning_rate": 2.470869283517052e-08, + "loss": 0.7792, + "num_input_tokens_seen": 168712965, + "step": 7912 + }, + { + "epoch": 0.9514819936271268, + "flos": 18322436225040.0, + "grad_norm": 3.379261999200524, + "learning_rate": 2.458677489076777e-08, + "loss": 0.7582, + "num_input_tokens_seen": 168733695, + "step": 7913 + }, + { + "epoch": 0.9516022365177659, + "flos": 13624531398120.0, + "grad_norm": 3.557733526010914, + "learning_rate": 2.446515661498072e-08, + "loss": 0.8185, + "num_input_tokens_seen": 168752745, + "step": 7914 + }, + { + "epoch": 0.9517224794084049, + "flos": 18589886313480.0, + "grad_norm": 4.703505592535081, + "learning_rate": 2.434383802625861e-08, + "loss": 0.7214, + "num_input_tokens_seen": 168771420, + "step": 7915 + }, + { + "epoch": 0.9518427222990441, + "flos": 15703136836800.0, + "grad_norm": 5.328446454058587, + "learning_rate": 2.4222819143005168e-08, + "loss": 0.7298, + "num_input_tokens_seen": 168790735, + "step": 7916 + }, + { + "epoch": 0.9519629651896832, + "flos": 15379594281960.0, + "grad_norm": 4.258409594744419, + "learning_rate": 2.4102099983579706e-08, + "loss": 0.801, + "num_input_tokens_seen": 168809605, + "step": 7917 + }, + { + "epoch": 0.9520832080803222, + "flos": 15864607344600.0, + "grad_norm": 3.674275623995019, + "learning_rate": 2.3981680566294236e-08, + "loss": 0.7498, + "num_input_tokens_seen": 168828925, + "step": 7918 + }, + { + "epoch": 0.9522034509709614, + "flos": 16941721149120.0, + "grad_norm": 3.416822771057334, + "learning_rate": 2.3861560909416822e-08, + "loss": 0.7308, + "num_input_tokens_seen": 168848195, + "step": 7919 + }, + { + "epoch": 0.9523236938616004, + "flos": 18077333577000.0, + "grad_norm": 3.4088322362153063, + "learning_rate": 2.3741741031169325e-08, + "loss": 0.8176, + "num_input_tokens_seen": 168867485, + "step": 7920 + }, + { + "epoch": 0.9524439367522395, + "flos": 16591873658880.0, + "grad_norm": 3.6412127761169124, + "learning_rate": 2.3622220949728544e-08, + "loss": 0.683, + "num_input_tokens_seen": 168886090, + "step": 7921 + }, + { + "epoch": 0.9525641796428787, + "flos": 25016169169320.0, + "grad_norm": 5.803161995666605, + "learning_rate": 2.3503000683225526e-08, + "loss": 0.5929, + "num_input_tokens_seen": 168903525, + "step": 7922 + }, + { + "epoch": 0.9526844225335177, + "flos": 12192282890040.0, + "grad_norm": 4.374949645706316, + "learning_rate": 2.3384080249745585e-08, + "loss": 0.8188, + "num_input_tokens_seen": 168921135, + "step": 7923 + }, + { + "epoch": 0.9528046654241568, + "flos": 27150613794720.0, + "grad_norm": 10.312196299386947, + "learning_rate": 2.3265459667329178e-08, + "loss": 0.8114, + "num_input_tokens_seen": 168940345, + "step": 7924 + }, + { + "epoch": 0.9529249083147959, + "flos": 13322418144840.0, + "grad_norm": 5.038292167984178, + "learning_rate": 2.31471389539708e-08, + "loss": 0.8342, + "num_input_tokens_seen": 168957190, + "step": 7925 + }, + { + "epoch": 0.953045151205435, + "flos": 21022360057320.0, + "grad_norm": 7.062490007511598, + "learning_rate": 2.3029118127619872e-08, + "loss": 0.7047, + "num_input_tokens_seen": 168976625, + "step": 7926 + }, + { + "epoch": 0.953165394096074, + "flos": 15972676482600.0, + "grad_norm": 7.800633810652849, + "learning_rate": 2.2911397206179628e-08, + "loss": 0.8529, + "num_input_tokens_seen": 168993095, + "step": 7927 + }, + { + "epoch": 0.9532856369867132, + "flos": 14596108861440.0, + "grad_norm": 8.828010490822619, + "learning_rate": 2.279397620750845e-08, + "loss": 0.6079, + "num_input_tokens_seen": 169011860, + "step": 7928 + }, + { + "epoch": 0.9534058798773523, + "flos": 10950944161200.0, + "grad_norm": 3.3965672049093656, + "learning_rate": 2.2676855149419195e-08, + "loss": 0.7611, + "num_input_tokens_seen": 169028750, + "step": 7929 + }, + { + "epoch": 0.9535261227679913, + "flos": 12840381118440.0, + "grad_norm": 14.335647455342475, + "learning_rate": 2.2560034049678988e-08, + "loss": 0.7234, + "num_input_tokens_seen": 169042820, + "step": 7930 + }, + { + "epoch": 0.9536463656586305, + "flos": 16939663251720.0, + "grad_norm": 3.2681559560815767, + "learning_rate": 2.2443512926008988e-08, + "loss": 0.737, + "num_input_tokens_seen": 169061870, + "step": 7931 + }, + { + "epoch": 0.9537666085492695, + "flos": 13596991724280.0, + "grad_norm": 5.5092492792993095, + "learning_rate": 2.2327291796085946e-08, + "loss": 0.6858, + "num_input_tokens_seen": 169079950, + "step": 7932 + }, + { + "epoch": 0.9538868514399086, + "flos": 13866531370080.0, + "grad_norm": 4.852480594177592, + "learning_rate": 2.2211370677540197e-08, + "loss": 0.7554, + "num_input_tokens_seen": 169096195, + "step": 7933 + }, + { + "epoch": 0.9540070943305478, + "flos": 12246729038520.0, + "grad_norm": 4.6209383476485115, + "learning_rate": 2.2095749587957012e-08, + "loss": 0.7556, + "num_input_tokens_seen": 169113820, + "step": 7934 + }, + { + "epoch": 0.9541273372211868, + "flos": 14730609574680.0, + "grad_norm": 4.029821808238219, + "learning_rate": 2.1980428544876138e-08, + "loss": 0.6704, + "num_input_tokens_seen": 169132180, + "step": 7935 + }, + { + "epoch": 0.9542475801118259, + "flos": 19591979486880.0, + "grad_norm": 3.918200348144628, + "learning_rate": 2.1865407565791584e-08, + "loss": 0.7219, + "num_input_tokens_seen": 169153470, + "step": 7936 + }, + { + "epoch": 0.954367823002465, + "flos": 17079546158160.0, + "grad_norm": 11.208941664134954, + "learning_rate": 2.175068666815183e-08, + "loss": 0.7566, + "num_input_tokens_seen": 169174030, + "step": 7937 + }, + { + "epoch": 0.9544880658931041, + "flos": 10842115184160.0, + "grad_norm": 5.153453831680621, + "learning_rate": 2.163626586935985e-08, + "loss": 0.7744, + "num_input_tokens_seen": 169190290, + "step": 7938 + }, + { + "epoch": 0.9546083087837431, + "flos": 21346060911960.0, + "grad_norm": 5.773694849537527, + "learning_rate": 2.1522145186773755e-08, + "loss": 0.6053, + "num_input_tokens_seen": 169208930, + "step": 7939 + }, + { + "epoch": 0.9547285516743822, + "flos": 15459933786360.0, + "grad_norm": 3.2332508142172824, + "learning_rate": 2.140832463770481e-08, + "loss": 0.8388, + "num_input_tokens_seen": 169227845, + "step": 7940 + }, + { + "epoch": 0.9548487945650214, + "flos": 20157648653520.0, + "grad_norm": 4.423354693210048, + "learning_rate": 2.129480423941987e-08, + "loss": 0.7418, + "num_input_tokens_seen": 169244235, + "step": 7941 + }, + { + "epoch": 0.9549690374556604, + "flos": 16296820576680.0, + "grad_norm": 2.9223043360936014, + "learning_rate": 2.1181584009140052e-08, + "loss": 0.7928, + "num_input_tokens_seen": 169263495, + "step": 7942 + }, + { + "epoch": 0.9550892803462995, + "flos": 12833764186800.0, + "grad_norm": 6.38899609388255, + "learning_rate": 2.10686639640405e-08, + "loss": 0.8176, + "num_input_tokens_seen": 169277305, + "step": 7943 + }, + { + "epoch": 0.9552095232369386, + "flos": 17836156764000.0, + "grad_norm": 2.9550376228738777, + "learning_rate": 2.0956044121251294e-08, + "loss": 0.8018, + "num_input_tokens_seen": 169295810, + "step": 7944 + }, + { + "epoch": 0.9553297661275777, + "flos": 16647269606160.0, + "grad_norm": 5.721896280596009, + "learning_rate": 2.084372449785654e-08, + "loss": 0.7953, + "num_input_tokens_seen": 169315365, + "step": 7945 + }, + { + "epoch": 0.9554500090182168, + "flos": 11220452147040.0, + "grad_norm": 5.069606173416239, + "learning_rate": 2.0731705110895282e-08, + "loss": 0.6744, + "num_input_tokens_seen": 169332575, + "step": 7946 + }, + { + "epoch": 0.9555702519088559, + "flos": 17214711730560.0, + "grad_norm": 3.728132193239472, + "learning_rate": 2.0619985977360587e-08, + "loss": 0.8458, + "num_input_tokens_seen": 169350615, + "step": 7947 + }, + { + "epoch": 0.955690494799495, + "flos": 16806017357400.0, + "grad_norm": 3.9216218539908003, + "learning_rate": 2.0508567114200237e-08, + "loss": 0.768, + "num_input_tokens_seen": 169370250, + "step": 7948 + }, + { + "epoch": 0.955810737690134, + "flos": 19078033712160.0, + "grad_norm": 4.010607958438354, + "learning_rate": 2.0397448538316485e-08, + "loss": 0.7677, + "num_input_tokens_seen": 169391010, + "step": 7949 + }, + { + "epoch": 0.9559309805807732, + "flos": 15243478910760.0, + "grad_norm": 4.28567788244863, + "learning_rate": 2.028663026656563e-08, + "loss": 0.6474, + "num_input_tokens_seen": 169409585, + "step": 7950 + }, + { + "epoch": 0.9560512234714122, + "flos": 15782241602760.0, + "grad_norm": 5.821147761480233, + "learning_rate": 2.0176112315758885e-08, + "loss": 0.7089, + "num_input_tokens_seen": 169427095, + "step": 7951 + }, + { + "epoch": 0.9561714663620513, + "flos": 12726454887840.0, + "grad_norm": 6.236664239396004, + "learning_rate": 2.0065894702661957e-08, + "loss": 0.6807, + "num_input_tokens_seen": 169443490, + "step": 7952 + }, + { + "epoch": 0.9562917092526905, + "flos": 19127002687560.0, + "grad_norm": 21.554870140128717, + "learning_rate": 1.9955977443994577e-08, + "loss": 0.7658, + "num_input_tokens_seen": 169463200, + "step": 7953 + }, + { + "epoch": 0.9564119521433295, + "flos": 17646260103480.0, + "grad_norm": 9.144824506057075, + "learning_rate": 1.9846360556430965e-08, + "loss": 0.6141, + "num_input_tokens_seen": 169481220, + "step": 7954 + }, + { + "epoch": 0.9565321950339686, + "flos": 23501048360040.0, + "grad_norm": 4.442909274936568, + "learning_rate": 1.973704405660004e-08, + "loss": 0.5879, + "num_input_tokens_seen": 169502055, + "step": 7955 + }, + { + "epoch": 0.9566524379246077, + "flos": 17270614237200.0, + "grad_norm": 2.2780583668868353, + "learning_rate": 1.9628027961085203e-08, + "loss": 0.7581, + "num_input_tokens_seen": 169525005, + "step": 7956 + }, + { + "epoch": 0.9567726808152468, + "flos": 27985284387840.0, + "grad_norm": 3.4670086863111593, + "learning_rate": 1.9519312286423894e-08, + "loss": 0.8208, + "num_input_tokens_seen": 169547920, + "step": 7957 + }, + { + "epoch": 0.9568929237058859, + "flos": 16644800129280.0, + "grad_norm": 3.5677530476165042, + "learning_rate": 1.9410897049108255e-08, + "loss": 0.7595, + "num_input_tokens_seen": 169566920, + "step": 7958 + }, + { + "epoch": 0.957013166596525, + "flos": 17457503201520.0, + "grad_norm": 3.537503267428234, + "learning_rate": 1.9302782265584905e-08, + "loss": 0.8874, + "num_input_tokens_seen": 169587305, + "step": 7959 + }, + { + "epoch": 0.9571334094871641, + "flos": 12867382572960.0, + "grad_norm": 5.489522490866976, + "learning_rate": 1.9194967952254282e-08, + "loss": 0.8416, + "num_input_tokens_seen": 169605600, + "step": 7960 + }, + { + "epoch": 0.9572536523778031, + "flos": 11186612141160.0, + "grad_norm": 5.901403628698715, + "learning_rate": 1.9087454125472635e-08, + "loss": 0.7795, + "num_input_tokens_seen": 169619795, + "step": 7961 + }, + { + "epoch": 0.9573738952684423, + "flos": 18292047154800.0, + "grad_norm": 6.8110075549491365, + "learning_rate": 1.8980240801548696e-08, + "loss": 0.7648, + "num_input_tokens_seen": 169638705, + "step": 7962 + }, + { + "epoch": 0.9574941381590814, + "flos": 18883388057640.0, + "grad_norm": 3.4946085412398262, + "learning_rate": 1.8873327996747458e-08, + "loss": 0.7218, + "num_input_tokens_seen": 169656925, + "step": 7963 + }, + { + "epoch": 0.9576143810497204, + "flos": 23722695469080.0, + "grad_norm": 4.600082703551818, + "learning_rate": 1.8766715727287053e-08, + "loss": 0.6473, + "num_input_tokens_seen": 169678350, + "step": 7964 + }, + { + "epoch": 0.9577346239403596, + "flos": 20346817134960.0, + "grad_norm": 6.9760870520558536, + "learning_rate": 1.8660404009340546e-08, + "loss": 0.7776, + "num_input_tokens_seen": 169698520, + "step": 7965 + }, + { + "epoch": 0.9578548668309986, + "flos": 42346801707960.0, + "grad_norm": 0.9191769055729743, + "learning_rate": 1.8554392859035485e-08, + "loss": 0.6586, + "num_input_tokens_seen": 169755990, + "step": 7966 + }, + { + "epoch": 0.9579751097216377, + "flos": 14434005154440.0, + "grad_norm": 5.500159694780764, + "learning_rate": 1.8448682292453444e-08, + "loss": 0.7767, + "num_input_tokens_seen": 169774785, + "step": 7967 + }, + { + "epoch": 0.9580953526122769, + "flos": 13164968451960.0, + "grad_norm": 3.5747456813599157, + "learning_rate": 1.8343272325631154e-08, + "loss": 0.6483, + "num_input_tokens_seen": 169793450, + "step": 7968 + }, + { + "epoch": 0.9582155955029159, + "flos": 17773320726120.0, + "grad_norm": 6.176604262543008, + "learning_rate": 1.8238162974558492e-08, + "loss": 0.7741, + "num_input_tokens_seen": 169807100, + "step": 7969 + }, + { + "epoch": 0.958335838393555, + "flos": 16674239400720.0, + "grad_norm": 5.714643828205915, + "learning_rate": 1.8133354255181144e-08, + "loss": 0.7339, + "num_input_tokens_seen": 169827135, + "step": 7970 + }, + { + "epoch": 0.958456081284194, + "flos": 12329031460440.0, + "grad_norm": 3.681727659780258, + "learning_rate": 1.802884618339795e-08, + "loss": 0.7313, + "num_input_tokens_seen": 169845660, + "step": 7971 + }, + { + "epoch": 0.9585763241748332, + "flos": 14596013881560.0, + "grad_norm": 5.808603929358094, + "learning_rate": 1.7924638775062894e-08, + "loss": 0.7874, + "num_input_tokens_seen": 169864500, + "step": 7972 + }, + { + "epoch": 0.9586965670654722, + "flos": 15647107690320.0, + "grad_norm": 3.219116634487341, + "learning_rate": 1.7820732045984444e-08, + "loss": 0.8045, + "num_input_tokens_seen": 169884365, + "step": 7973 + }, + { + "epoch": 0.9588168099561113, + "flos": 15676736921520.0, + "grad_norm": 6.465415205577874, + "learning_rate": 1.7717126011924655e-08, + "loss": 0.7248, + "num_input_tokens_seen": 169905670, + "step": 7974 + }, + { + "epoch": 0.9589370528467505, + "flos": 8518913656800.0, + "grad_norm": 5.806918838461958, + "learning_rate": 1.7613820688600957e-08, + "loss": 0.7467, + "num_input_tokens_seen": 169921295, + "step": 7975 + }, + { + "epoch": 0.9590572957373895, + "flos": 17161563640440.0, + "grad_norm": 5.854817563544792, + "learning_rate": 1.7510816091684588e-08, + "loss": 0.7806, + "num_input_tokens_seen": 169940940, + "step": 7976 + }, + { + "epoch": 0.9591775386280286, + "flos": 16485545818680.0, + "grad_norm": 24.178041644308298, + "learning_rate": 1.740811223680083e-08, + "loss": 0.7623, + "num_input_tokens_seen": 169957515, + "step": 7977 + }, + { + "epoch": 0.9592977815186677, + "flos": 13272752650320.0, + "grad_norm": 6.187619294837215, + "learning_rate": 1.7305709139530334e-08, + "loss": 0.7144, + "num_input_tokens_seen": 169976015, + "step": 7978 + }, + { + "epoch": 0.9594180244093068, + "flos": 12051228565080.0, + "grad_norm": 4.739463765519045, + "learning_rate": 1.7203606815407334e-08, + "loss": 0.7274, + "num_input_tokens_seen": 169990330, + "step": 7979 + }, + { + "epoch": 0.9595382672999458, + "flos": 15024174638760.0, + "grad_norm": 3.895341947576593, + "learning_rate": 1.7101805279920557e-08, + "loss": 0.7788, + "num_input_tokens_seen": 170008210, + "step": 7980 + }, + { + "epoch": 0.959658510190585, + "flos": 16567025081640.0, + "grad_norm": 6.5504845476375175, + "learning_rate": 1.7000304548513643e-08, + "loss": 0.7989, + "num_input_tokens_seen": 170028035, + "step": 7981 + }, + { + "epoch": 0.9597787530812241, + "flos": 13974505528200.0, + "grad_norm": 4.886567306306391, + "learning_rate": 1.6899104636583394e-08, + "loss": 0.8085, + "num_input_tokens_seen": 170045805, + "step": 7982 + }, + { + "epoch": 0.9598989959718631, + "flos": 44293439230200.0, + "grad_norm": 0.7305343516874215, + "learning_rate": 1.6798205559482638e-08, + "loss": 0.6434, + "num_input_tokens_seen": 170107905, + "step": 7983 + }, + { + "epoch": 0.9600192388625023, + "flos": 15270796964880.0, + "grad_norm": 4.612754074583631, + "learning_rate": 1.669760733251713e-08, + "loss": 0.7464, + "num_input_tokens_seen": 170126500, + "step": 7984 + }, + { + "epoch": 0.9601394817531413, + "flos": 14943930114240.0, + "grad_norm": 8.671663408876954, + "learning_rate": 1.659730997094755e-08, + "loss": 0.8213, + "num_input_tokens_seen": 170144710, + "step": 7985 + }, + { + "epoch": 0.9602597246437804, + "flos": 15729916671600.0, + "grad_norm": 3.237692490676743, + "learning_rate": 1.6497313489989283e-08, + "loss": 0.6154, + "num_input_tokens_seen": 170164255, + "step": 7986 + }, + { + "epoch": 0.9603799675344196, + "flos": 21967126025880.0, + "grad_norm": 4.295758022433839, + "learning_rate": 1.639761790481131e-08, + "loss": 0.6647, + "num_input_tokens_seen": 170184855, + "step": 7987 + }, + { + "epoch": 0.9605002104250586, + "flos": 20536175576160.0, + "grad_norm": 4.933550063518972, + "learning_rate": 1.6298223230537754e-08, + "loss": 0.7768, + "num_input_tokens_seen": 170202375, + "step": 7988 + }, + { + "epoch": 0.9606204533156977, + "flos": 26152509776280.0, + "grad_norm": 6.154552244979454, + "learning_rate": 1.619912948224611e-08, + "loss": 0.6766, + "num_input_tokens_seen": 170223300, + "step": 7989 + }, + { + "epoch": 0.9607406962063368, + "flos": 19479604594320.0, + "grad_norm": 5.402736514385286, + "learning_rate": 1.6100336674969682e-08, + "loss": 0.5796, + "num_input_tokens_seen": 170241860, + "step": 7990 + }, + { + "epoch": 0.9608609390969759, + "flos": 18559497243240.0, + "grad_norm": 3.791633913227066, + "learning_rate": 1.600184482369449e-08, + "loss": 0.7508, + "num_input_tokens_seen": 170261495, + "step": 7991 + }, + { + "epoch": 0.960981181987615, + "flos": 15405709257600.0, + "grad_norm": 8.981372205397934, + "learning_rate": 1.5903653943362126e-08, + "loss": 0.878, + "num_input_tokens_seen": 170280210, + "step": 7992 + }, + { + "epoch": 0.9611014248782541, + "flos": 13003751223840.0, + "grad_norm": 3.9691958439825292, + "learning_rate": 1.580576404886802e-08, + "loss": 0.7455, + "num_input_tokens_seen": 170298460, + "step": 7993 + }, + { + "epoch": 0.9612216677688932, + "flos": 14136925834800.0, + "grad_norm": 8.868839328522997, + "learning_rate": 1.570817515506162e-08, + "loss": 0.7838, + "num_input_tokens_seen": 170316870, + "step": 7994 + }, + { + "epoch": 0.9613419106595322, + "flos": 11511769353960.0, + "grad_norm": 6.295674477000961, + "learning_rate": 1.561088727674753e-08, + "loss": 0.7997, + "num_input_tokens_seen": 170330800, + "step": 7995 + }, + { + "epoch": 0.9614621535501714, + "flos": 18834672361920.0, + "grad_norm": 4.932418139513612, + "learning_rate": 1.551390042868417e-08, + "loss": 0.6838, + "num_input_tokens_seen": 170352290, + "step": 7996 + }, + { + "epoch": 0.9615823964408104, + "flos": 13000047008520.0, + "grad_norm": 6.90022919868498, + "learning_rate": 1.5417214625584207e-08, + "loss": 0.7067, + "num_input_tokens_seen": 170369665, + "step": 7997 + }, + { + "epoch": 0.9617026393314495, + "flos": 14754634992960.0, + "grad_norm": 3.3298508731568415, + "learning_rate": 1.5320829882114806e-08, + "loss": 0.8389, + "num_input_tokens_seen": 170387460, + "step": 7998 + }, + { + "epoch": 0.9618228822220887, + "flos": 14812183817520.0, + "grad_norm": 3.97779907905184, + "learning_rate": 1.5224746212897378e-08, + "loss": 0.765, + "num_input_tokens_seen": 170406475, + "step": 7999 + }, + { + "epoch": 0.9619431251127277, + "flos": 15377536384560.0, + "grad_norm": 3.6952752077184736, + "learning_rate": 1.512896363250804e-08, + "loss": 0.7576, + "num_input_tokens_seen": 170426305, + "step": 8000 + }, + { + "epoch": 0.9620633680033668, + "flos": 16377824940240.0, + "grad_norm": 3.3328922138794677, + "learning_rate": 1.503348215547673e-08, + "loss": 0.7455, + "num_input_tokens_seen": 170447115, + "step": 8001 + }, + { + "epoch": 0.962183610894006, + "flos": 13482495614400.0, + "grad_norm": 4.696361253246613, + "learning_rate": 1.4938301796288078e-08, + "loss": 0.788, + "num_input_tokens_seen": 170463405, + "step": 8002 + }, + { + "epoch": 0.962303853784645, + "flos": 13455494159880.0, + "grad_norm": 11.440707397997798, + "learning_rate": 1.4843422569380537e-08, + "loss": 0.8042, + "num_input_tokens_seen": 170479880, + "step": 8003 + }, + { + "epoch": 0.9624240966752841, + "flos": 19343615862960.0, + "grad_norm": 5.447209601047507, + "learning_rate": 1.4748844489147483e-08, + "loss": 0.8119, + "num_input_tokens_seen": 170496590, + "step": 8004 + }, + { + "epoch": 0.9625443395659231, + "flos": 10653421602120.0, + "grad_norm": 4.343570842753186, + "learning_rate": 1.4654567569936326e-08, + "loss": 0.6914, + "num_input_tokens_seen": 170513885, + "step": 8005 + }, + { + "epoch": 0.9626645824565623, + "flos": 13406208584880.0, + "grad_norm": 3.069454005146432, + "learning_rate": 1.456059182604874e-08, + "loss": 0.8086, + "num_input_tokens_seen": 170532410, + "step": 8006 + }, + { + "epoch": 0.9627848253472013, + "flos": 12083643872760.0, + "grad_norm": 6.0304552256253, + "learning_rate": 1.4466917271740653e-08, + "loss": 0.7543, + "num_input_tokens_seen": 170550330, + "step": 8007 + }, + { + "epoch": 0.9629050682378404, + "flos": 15270860284800.0, + "grad_norm": 5.54704297308388, + "learning_rate": 1.4373543921222697e-08, + "loss": 0.6539, + "num_input_tokens_seen": 170569635, + "step": 8008 + }, + { + "epoch": 0.9630253111284796, + "flos": 12408547805880.0, + "grad_norm": 6.610345719569179, + "learning_rate": 1.428047178865932e-08, + "loss": 0.7562, + "num_input_tokens_seen": 170586145, + "step": 8009 + }, + { + "epoch": 0.9631455540191186, + "flos": 14865268587720.0, + "grad_norm": 4.271564526497091, + "learning_rate": 1.4187700888169451e-08, + "loss": 0.7276, + "num_input_tokens_seen": 170605040, + "step": 8010 + }, + { + "epoch": 0.9632657969097577, + "flos": 48632129710080.0, + "grad_norm": 0.8239066088838626, + "learning_rate": 1.40952312338265e-08, + "loss": 0.6385, + "num_input_tokens_seen": 170669405, + "step": 8011 + }, + { + "epoch": 0.9633860398003968, + "flos": 32686513510560.0, + "grad_norm": 4.474735333165514, + "learning_rate": 1.4003062839657909e-08, + "loss": 0.6699, + "num_input_tokens_seen": 170691605, + "step": 8012 + }, + { + "epoch": 0.9635062826910359, + "flos": 18187397292480.0, + "grad_norm": 2.8408297121922526, + "learning_rate": 1.391119571964583e-08, + "loss": 0.7874, + "num_input_tokens_seen": 170712265, + "step": 8013 + }, + { + "epoch": 0.9636265255816749, + "flos": 11625663924600.0, + "grad_norm": 3.5734827836890815, + "learning_rate": 1.3819629887726225e-08, + "loss": 0.7186, + "num_input_tokens_seen": 170730075, + "step": 8014 + }, + { + "epoch": 0.9637467684723141, + "flos": 16540023627120.0, + "grad_norm": 7.80618910143442, + "learning_rate": 1.3728365357789317e-08, + "loss": 0.7522, + "num_input_tokens_seen": 170749160, + "step": 8015 + }, + { + "epoch": 0.9638670113629532, + "flos": 12812936424480.0, + "grad_norm": 5.45012785735335, + "learning_rate": 1.3637402143680254e-08, + "loss": 0.7472, + "num_input_tokens_seen": 170763780, + "step": 8016 + }, + { + "epoch": 0.9639872542535922, + "flos": 40575153496440.0, + "grad_norm": 0.7540308399723413, + "learning_rate": 1.3546740259197998e-08, + "loss": 0.5655, + "num_input_tokens_seen": 170816310, + "step": 8017 + }, + { + "epoch": 0.9641074971442314, + "flos": 17674369656600.0, + "grad_norm": 6.57877359341452, + "learning_rate": 1.3456379718095989e-08, + "loss": 0.6706, + "num_input_tokens_seen": 170836445, + "step": 8018 + }, + { + "epoch": 0.9642277400348704, + "flos": 49215618942840.0, + "grad_norm": 0.8804907185911771, + "learning_rate": 1.3366320534081487e-08, + "loss": 0.6419, + "num_input_tokens_seen": 170898845, + "step": 8019 + }, + { + "epoch": 0.9643479829255095, + "flos": 22696228617840.0, + "grad_norm": 4.329158287786772, + "learning_rate": 1.3276562720816675e-08, + "loss": 0.7308, + "num_input_tokens_seen": 170920075, + "step": 8020 + }, + { + "epoch": 0.9644682258161487, + "flos": 14650301730240.0, + "grad_norm": 3.883202270348085, + "learning_rate": 1.3187106291917549e-08, + "loss": 0.8114, + "num_input_tokens_seen": 170936785, + "step": 8021 + }, + { + "epoch": 0.9645884687067877, + "flos": 15486840261000.0, + "grad_norm": 4.013631956918285, + "learning_rate": 1.309795126095503e-08, + "loss": 0.6934, + "num_input_tokens_seen": 170954805, + "step": 8022 + }, + { + "epoch": 0.9647087115974268, + "flos": 13833451203240.0, + "grad_norm": 5.290042577907499, + "learning_rate": 1.3009097641453192e-08, + "loss": 0.7938, + "num_input_tokens_seen": 170972375, + "step": 8023 + }, + { + "epoch": 0.9648289544880659, + "flos": 12057655536960.0, + "grad_norm": 3.7605593470224647, + "learning_rate": 1.2920545446891474e-08, + "loss": 0.7464, + "num_input_tokens_seen": 170988815, + "step": 8024 + }, + { + "epoch": 0.964949197378705, + "flos": 17616852492000.0, + "grad_norm": 3.375721183300726, + "learning_rate": 1.2832294690703127e-08, + "loss": 0.6887, + "num_input_tokens_seen": 171007510, + "step": 8025 + }, + { + "epoch": 0.965069440269344, + "flos": 17104932954720.0, + "grad_norm": 9.443711415684746, + "learning_rate": 1.2744345386275668e-08, + "loss": 0.7599, + "num_input_tokens_seen": 171026770, + "step": 8026 + }, + { + "epoch": 0.9651896831599832, + "flos": 18592324130400.0, + "grad_norm": 3.5825923035171354, + "learning_rate": 1.265669754695109e-08, + "loss": 0.7609, + "num_input_tokens_seen": 171046060, + "step": 8027 + }, + { + "epoch": 0.9653099260506223, + "flos": 16296535637040.0, + "grad_norm": 5.7923414902943655, + "learning_rate": 1.2569351186025201e-08, + "loss": 0.8115, + "num_input_tokens_seen": 171064235, + "step": 8028 + }, + { + "epoch": 0.9654301689412613, + "flos": 19617967822680.0, + "grad_norm": 2.81365527691355, + "learning_rate": 1.2482306316748737e-08, + "loss": 0.7458, + "num_input_tokens_seen": 171084400, + "step": 8029 + }, + { + "epoch": 0.9655504118319005, + "flos": 12698598614400.0, + "grad_norm": 3.72814600667607, + "learning_rate": 1.2395562952326021e-08, + "loss": 0.7657, + "num_input_tokens_seen": 171101280, + "step": 8030 + }, + { + "epoch": 0.9656706547225395, + "flos": 16188561478920.0, + "grad_norm": 4.019839201790534, + "learning_rate": 1.2309121105916309e-08, + "loss": 0.7934, + "num_input_tokens_seen": 171119290, + "step": 8031 + }, + { + "epoch": 0.9657908976131786, + "flos": 27232029737760.0, + "grad_norm": 3.5228949084626096, + "learning_rate": 1.222298079063222e-08, + "loss": 0.6766, + "num_input_tokens_seen": 171140150, + "step": 8032 + }, + { + "epoch": 0.9659111405038178, + "flos": 17863316518320.0, + "grad_norm": 3.342741912327609, + "learning_rate": 1.2137142019541524e-08, + "loss": 0.7125, + "num_input_tokens_seen": 171158425, + "step": 8033 + }, + { + "epoch": 0.9660313833944568, + "flos": 18321771365880.0, + "grad_norm": 9.390206986198566, + "learning_rate": 1.2051604805666027e-08, + "loss": 0.7279, + "num_input_tokens_seen": 171175270, + "step": 8034 + }, + { + "epoch": 0.9661516262850959, + "flos": 8547086529840.0, + "grad_norm": 5.661184950623501, + "learning_rate": 1.196636916198135e-08, + "loss": 0.764, + "num_input_tokens_seen": 171192530, + "step": 8035 + }, + { + "epoch": 0.9662718691757349, + "flos": 14649890150760.0, + "grad_norm": 3.5055492550859966, + "learning_rate": 1.1881435101418036e-08, + "loss": 0.7626, + "num_input_tokens_seen": 171211665, + "step": 8036 + }, + { + "epoch": 0.9663921120663741, + "flos": 50163582567360.0, + "grad_norm": 0.7796068459422093, + "learning_rate": 1.1796802636860003e-08, + "loss": 0.6946, + "num_input_tokens_seen": 171279915, + "step": 8037 + }, + { + "epoch": 0.9665123549570132, + "flos": 19293855388560.0, + "grad_norm": 4.7220928149474695, + "learning_rate": 1.1712471781146316e-08, + "loss": 0.7036, + "num_input_tokens_seen": 171298970, + "step": 8038 + }, + { + "epoch": 0.9666325978476522, + "flos": 32331885366360.0, + "grad_norm": 5.954651484184083, + "learning_rate": 1.1628442547069628e-08, + "loss": 0.6625, + "num_input_tokens_seen": 171320890, + "step": 8039 + }, + { + "epoch": 0.9667528407382914, + "flos": 15756728166360.0, + "grad_norm": 3.753090030402435, + "learning_rate": 1.1544714947377521e-08, + "loss": 0.754, + "num_input_tokens_seen": 171338295, + "step": 8040 + }, + { + "epoch": 0.9668730836289304, + "flos": 17484916235520.0, + "grad_norm": 5.532811693322361, + "learning_rate": 1.1461288994770945e-08, + "loss": 0.6868, + "num_input_tokens_seen": 171357090, + "step": 8041 + }, + { + "epoch": 0.9669933265195695, + "flos": 20751838952760.0, + "grad_norm": 2.8669539935994814, + "learning_rate": 1.1378164701906002e-08, + "loss": 0.7524, + "num_input_tokens_seen": 171378575, + "step": 8042 + }, + { + "epoch": 0.9671135694102087, + "flos": 16431068010240.0, + "grad_norm": 3.447648759250728, + "learning_rate": 1.1295342081392156e-08, + "loss": 0.6503, + "num_input_tokens_seen": 171397655, + "step": 8043 + }, + { + "epoch": 0.9672338123008477, + "flos": 14729754755760.0, + "grad_norm": 5.898904097335415, + "learning_rate": 1.1212821145793804e-08, + "loss": 0.6753, + "num_input_tokens_seen": 171416990, + "step": 8044 + }, + { + "epoch": 0.9673540551914868, + "flos": 12376892337240.0, + "grad_norm": 3.7083435629942882, + "learning_rate": 1.1130601907629156e-08, + "loss": 0.7677, + "num_input_tokens_seen": 171434440, + "step": 8045 + }, + { + "epoch": 0.9674742980821259, + "flos": 45622684193880.0, + "grad_norm": 0.8558596062148464, + "learning_rate": 1.1048684379370899e-08, + "loss": 0.6636, + "num_input_tokens_seen": 171494845, + "step": 8046 + }, + { + "epoch": 0.967594540972765, + "flos": 13650836333520.0, + "grad_norm": 6.045859554697492, + "learning_rate": 1.0967068573445759e-08, + "loss": 0.7243, + "num_input_tokens_seen": 171512050, + "step": 8047 + }, + { + "epoch": 0.967714783863404, + "flos": 15189855921240.0, + "grad_norm": 9.127478694008017, + "learning_rate": 1.0885754502234945e-08, + "loss": 0.6288, + "num_input_tokens_seen": 171531430, + "step": 8048 + }, + { + "epoch": 0.9678350267540432, + "flos": 16971983579520.0, + "grad_norm": 3.7628018888601575, + "learning_rate": 1.08047421780737e-08, + "loss": 0.7711, + "num_input_tokens_seen": 171550340, + "step": 8049 + }, + { + "epoch": 0.9679552696446823, + "flos": 15891925398720.0, + "grad_norm": 6.538135750020288, + "learning_rate": 1.0724031613251305e-08, + "loss": 0.7192, + "num_input_tokens_seen": 171567960, + "step": 8050 + }, + { + "epoch": 0.9680755125353213, + "flos": 19697864087640.0, + "grad_norm": 3.8508132917758653, + "learning_rate": 1.0643622820011744e-08, + "loss": 0.6398, + "num_input_tokens_seen": 171588735, + "step": 8051 + }, + { + "epoch": 0.9681957554259605, + "flos": 20774882912280.0, + "grad_norm": 29.217887270017275, + "learning_rate": 1.0563515810552814e-08, + "loss": 0.6711, + "num_input_tokens_seen": 171605425, + "step": 8052 + }, + { + "epoch": 0.9683159983165995, + "flos": 15027119015040.0, + "grad_norm": 4.98384000484041, + "learning_rate": 1.0483710597026795e-08, + "loss": 0.7251, + "num_input_tokens_seen": 171625005, + "step": 8053 + }, + { + "epoch": 0.9684362412072386, + "flos": 17728815805080.0, + "grad_norm": 5.292124911113597, + "learning_rate": 1.0404207191540227e-08, + "loss": 0.7203, + "num_input_tokens_seen": 171645180, + "step": 8054 + }, + { + "epoch": 0.9685564840978778, + "flos": 16350886805640.0, + "grad_norm": 3.935359797042417, + "learning_rate": 1.0325005606153236e-08, + "loss": 0.7255, + "num_input_tokens_seen": 171664360, + "step": 8055 + }, + { + "epoch": 0.9686767269885168, + "flos": 10459219187040.0, + "grad_norm": 6.856109467477113, + "learning_rate": 1.0246105852881104e-08, + "loss": 0.7658, + "num_input_tokens_seen": 171679180, + "step": 8056 + }, + { + "epoch": 0.9687969698791559, + "flos": 15567369725160.0, + "grad_norm": 7.28220187665539, + "learning_rate": 1.0167507943692476e-08, + "loss": 0.7671, + "num_input_tokens_seen": 171697985, + "step": 8057 + }, + { + "epoch": 0.968917212769795, + "flos": 14487406524240.0, + "grad_norm": 6.471337432123316, + "learning_rate": 1.008921189051093e-08, + "loss": 0.7006, + "num_input_tokens_seen": 171715050, + "step": 8058 + }, + { + "epoch": 0.9690374556604341, + "flos": 15858022072920.0, + "grad_norm": 8.485936367180232, + "learning_rate": 1.0011217705213848e-08, + "loss": 0.7554, + "num_input_tokens_seen": 171732645, + "step": 8059 + }, + { + "epoch": 0.9691576985510731, + "flos": 23965613579880.0, + "grad_norm": 3.528626944666158, + "learning_rate": 9.933525399632658e-09, + "loss": 0.7427, + "num_input_tokens_seen": 171750600, + "step": 8060 + }, + { + "epoch": 0.9692779414417123, + "flos": 26206259405640.0, + "grad_norm": 10.79966088221748, + "learning_rate": 9.856134985553488e-09, + "loss": 0.6402, + "num_input_tokens_seen": 171770045, + "step": 8061 + }, + { + "epoch": 0.9693981843323514, + "flos": 20806696680720.0, + "grad_norm": 3.370820742765536, + "learning_rate": 9.77904647471628e-09, + "loss": 0.7136, + "num_input_tokens_seen": 171792945, + "step": 8062 + }, + { + "epoch": 0.9695184272229904, + "flos": 17295621114240.0, + "grad_norm": 2.919988829186379, + "learning_rate": 9.702259878815454e-09, + "loss": 0.7341, + "num_input_tokens_seen": 171812990, + "step": 8063 + }, + { + "epoch": 0.9696386701136296, + "flos": 17052766323360.0, + "grad_norm": 4.708206625164633, + "learning_rate": 9.625775209499254e-09, + "loss": 0.7235, + "num_input_tokens_seen": 171832715, + "step": 8064 + }, + { + "epoch": 0.9697589130042686, + "flos": 11004345531000.0, + "grad_norm": 4.08306912456208, + "learning_rate": 9.549592478370172e-09, + "loss": 0.7224, + "num_input_tokens_seen": 171850615, + "step": 8065 + }, + { + "epoch": 0.9698791558949077, + "flos": 13785210406920.0, + "grad_norm": 5.121409577408383, + "learning_rate": 9.473711696985632e-09, + "loss": 0.779, + "num_input_tokens_seen": 171869665, + "step": 8066 + }, + { + "epoch": 0.9699993987855468, + "flos": 13083647488800.0, + "grad_norm": 6.119264992851215, + "learning_rate": 9.398132876856201e-09, + "loss": 0.7338, + "num_input_tokens_seen": 171888350, + "step": 8067 + }, + { + "epoch": 0.9701196416761859, + "flos": 49536945300480.0, + "grad_norm": 0.7756502115730701, + "learning_rate": 9.322856029447379e-09, + "loss": 0.636, + "num_input_tokens_seen": 171949255, + "step": 8068 + }, + { + "epoch": 0.970239884566825, + "flos": 17780064297600.0, + "grad_norm": 3.6134673199556815, + "learning_rate": 9.247881166178695e-09, + "loss": 0.7857, + "num_input_tokens_seen": 171967685, + "step": 8069 + }, + { + "epoch": 0.970360127457464, + "flos": 18538131261600.0, + "grad_norm": 6.618342959645877, + "learning_rate": 9.173208298423274e-09, + "loss": 0.7553, + "num_input_tokens_seen": 171988610, + "step": 8070 + }, + { + "epoch": 0.9704803703481032, + "flos": 21699074398200.0, + "grad_norm": 3.443853021593456, + "learning_rate": 9.09883743750961e-09, + "loss": 0.7553, + "num_input_tokens_seen": 172011220, + "step": 8071 + }, + { + "epoch": 0.9706006132387422, + "flos": 12674984775600.0, + "grad_norm": 3.1279897502428793, + "learning_rate": 9.024768594719124e-09, + "loss": 0.8173, + "num_input_tokens_seen": 172029320, + "step": 8072 + }, + { + "epoch": 0.9707208561293813, + "flos": 13272277750920.0, + "grad_norm": 4.61219210126854, + "learning_rate": 8.95100178128816e-09, + "loss": 0.7067, + "num_input_tokens_seen": 172048180, + "step": 8073 + }, + { + "epoch": 0.9708410990200205, + "flos": 22939020088800.0, + "grad_norm": 4.115836166027634, + "learning_rate": 8.877537008407321e-09, + "loss": 0.6943, + "num_input_tokens_seen": 172067950, + "step": 8074 + }, + { + "epoch": 0.9709613419106595, + "flos": 22425359253720.0, + "grad_norm": 3.9306294883960797, + "learning_rate": 8.804374287221028e-09, + "loss": 0.6664, + "num_input_tokens_seen": 172088905, + "step": 8075 + }, + { + "epoch": 0.9710815848012986, + "flos": 17374725880200.0, + "grad_norm": 4.422245724507493, + "learning_rate": 8.731513628827958e-09, + "loss": 0.8323, + "num_input_tokens_seen": 172107990, + "step": 8076 + }, + { + "epoch": 0.9712018276919377, + "flos": 17397516560040.0, + "grad_norm": 3.6760718203430565, + "learning_rate": 8.658955044280825e-09, + "loss": 0.8115, + "num_input_tokens_seen": 172126635, + "step": 8077 + }, + { + "epoch": 0.9713220705825768, + "flos": 17079609478080.0, + "grad_norm": 3.4611307955597646, + "learning_rate": 8.586698544587268e-09, + "loss": 0.7507, + "num_input_tokens_seen": 172147965, + "step": 8078 + }, + { + "epoch": 0.9714423134732159, + "flos": 16243450866840.0, + "grad_norm": 3.9995561591906696, + "learning_rate": 8.514744140707853e-09, + "loss": 0.733, + "num_input_tokens_seen": 172166825, + "step": 8079 + }, + { + "epoch": 0.971562556363855, + "flos": 14784644143680.0, + "grad_norm": 6.459882620108215, + "learning_rate": 8.443091843558515e-09, + "loss": 0.7469, + "num_input_tokens_seen": 172185630, + "step": 8080 + }, + { + "epoch": 0.9716827992544941, + "flos": 18292363754400.0, + "grad_norm": 4.783282477521647, + "learning_rate": 8.37174166400878e-09, + "loss": 0.6338, + "num_input_tokens_seen": 172200925, + "step": 8081 + }, + { + "epoch": 0.9718030421451331, + "flos": 18079138194720.0, + "grad_norm": 6.261346512056569, + "learning_rate": 8.300693612881992e-09, + "loss": 0.8283, + "num_input_tokens_seen": 172220710, + "step": 8082 + }, + { + "epoch": 0.9719232850357723, + "flos": 16161749984160.0, + "grad_norm": 7.853137717064041, + "learning_rate": 8.22994770095664e-09, + "loss": 0.7843, + "num_input_tokens_seen": 172239005, + "step": 8083 + }, + { + "epoch": 0.9720435279264114, + "flos": 17320469691480.0, + "grad_norm": 5.264555780086816, + "learning_rate": 8.159503938964585e-09, + "loss": 0.7352, + "num_input_tokens_seen": 172256045, + "step": 8084 + }, + { + "epoch": 0.9721637708170504, + "flos": 20805778541880.0, + "grad_norm": 4.023883565230634, + "learning_rate": 8.089362337592164e-09, + "loss": 0.6967, + "num_input_tokens_seen": 172279390, + "step": 8085 + }, + { + "epoch": 0.9722840137076896, + "flos": 21371669328240.0, + "grad_norm": 3.1721905330567925, + "learning_rate": 8.019522907479536e-09, + "loss": 0.7025, + "num_input_tokens_seen": 172299470, + "step": 8086 + }, + { + "epoch": 0.9724042565983286, + "flos": 14054845032600.0, + "grad_norm": 4.380081277728955, + "learning_rate": 7.949985659221558e-09, + "loss": 0.7478, + "num_input_tokens_seen": 172316455, + "step": 8087 + }, + { + "epoch": 0.9725244994889677, + "flos": 16864104401280.0, + "grad_norm": 3.5868228658771537, + "learning_rate": 7.880750603366904e-09, + "loss": 0.7684, + "num_input_tokens_seen": 172335045, + "step": 8088 + }, + { + "epoch": 0.9726447423796069, + "flos": 17106610932600.0, + "grad_norm": 5.470057042467799, + "learning_rate": 7.811817750418282e-09, + "loss": 0.7729, + "num_input_tokens_seen": 172353525, + "step": 8089 + }, + { + "epoch": 0.9727649852702459, + "flos": 19455705815880.0, + "grad_norm": 2.635445837128324, + "learning_rate": 7.743187110833105e-09, + "loss": 0.7825, + "num_input_tokens_seen": 172376005, + "step": 8090 + }, + { + "epoch": 0.972885228160885, + "flos": 15001130679240.0, + "grad_norm": 2.9852520517742, + "learning_rate": 7.674858695022602e-09, + "loss": 0.7978, + "num_input_tokens_seen": 172394080, + "step": 8091 + }, + { + "epoch": 0.9730054710515241, + "flos": 12813158044200.0, + "grad_norm": 5.843240481108779, + "learning_rate": 7.606832513351591e-09, + "loss": 0.7389, + "num_input_tokens_seen": 172411750, + "step": 8092 + }, + { + "epoch": 0.9731257139421632, + "flos": 50861821189680.0, + "grad_norm": 0.7951050343494802, + "learning_rate": 7.539108576140264e-09, + "loss": 0.6674, + "num_input_tokens_seen": 172475580, + "step": 8093 + }, + { + "epoch": 0.9732459568328022, + "flos": 13488764286480.0, + "grad_norm": 7.200463587353438, + "learning_rate": 7.471686893661732e-09, + "loss": 0.6767, + "num_input_tokens_seen": 172493595, + "step": 8094 + }, + { + "epoch": 0.9733661997234414, + "flos": 15268232508120.0, + "grad_norm": 4.2386180831734395, + "learning_rate": 7.4045674761442636e-09, + "loss": 0.6289, + "num_input_tokens_seen": 172510645, + "step": 8095 + }, + { + "epoch": 0.9734864426140805, + "flos": 17401853974560.0, + "grad_norm": 4.429217493597579, + "learning_rate": 7.337750333769488e-09, + "loss": 0.7183, + "num_input_tokens_seen": 172530170, + "step": 8096 + }, + { + "epoch": 0.9736066855047195, + "flos": 25746791439360.0, + "grad_norm": 9.289267009155395, + "learning_rate": 7.2712354766737425e-09, + "loss": 0.7156, + "num_input_tokens_seen": 172550220, + "step": 8097 + }, + { + "epoch": 0.9737269283953586, + "flos": 14918131738200.0, + "grad_norm": 3.339058981265694, + "learning_rate": 7.2050229149469565e-09, + "loss": 0.7928, + "num_input_tokens_seen": 172569950, + "step": 8098 + }, + { + "epoch": 0.9738471712859977, + "flos": 21209470641360.0, + "grad_norm": 3.8143278649268244, + "learning_rate": 7.139112658633984e-09, + "loss": 0.6137, + "num_input_tokens_seen": 172589820, + "step": 8099 + }, + { + "epoch": 0.9739674141766368, + "flos": 20373913569360.0, + "grad_norm": 4.115387090412112, + "learning_rate": 7.073504717733048e-09, + "loss": 0.6807, + "num_input_tokens_seen": 172609105, + "step": 8100 + }, + { + "epoch": 0.9740876570672758, + "flos": 50781260065560.0, + "grad_norm": 0.8851467768405467, + "learning_rate": 7.008199102196855e-09, + "loss": 0.5835, + "num_input_tokens_seen": 172670250, + "step": 8101 + }, + { + "epoch": 0.974207899957915, + "flos": 42915700190520.0, + "grad_norm": 0.8126562408464868, + "learning_rate": 6.9431958219321464e-09, + "loss": 0.6112, + "num_input_tokens_seen": 172726135, + "step": 8102 + }, + { + "epoch": 0.9743281428485541, + "flos": 16538503949040.0, + "grad_norm": 2.9059434954721595, + "learning_rate": 6.878494886800146e-09, + "loss": 0.7712, + "num_input_tokens_seen": 172746630, + "step": 8103 + }, + { + "epoch": 0.9744483857391931, + "flos": 14620704159000.0, + "grad_norm": 4.963898023186489, + "learning_rate": 6.814096306615669e-09, + "loss": 0.7415, + "num_input_tokens_seen": 172764490, + "step": 8104 + }, + { + "epoch": 0.9745686286298323, + "flos": 12893940788040.0, + "grad_norm": 4.4472536163100935, + "learning_rate": 6.750000091148011e-09, + "loss": 0.6248, + "num_input_tokens_seen": 172781505, + "step": 8105 + }, + { + "epoch": 0.9746888715204713, + "flos": 21615632217720.0, + "grad_norm": 7.701398489643882, + "learning_rate": 6.686206250120729e-09, + "loss": 0.7101, + "num_input_tokens_seen": 172802720, + "step": 8106 + }, + { + "epoch": 0.9748091144111104, + "flos": 13489175865960.0, + "grad_norm": 6.971299051649505, + "learning_rate": 6.622714793210749e-09, + "loss": 0.725, + "num_input_tokens_seen": 172821360, + "step": 8107 + }, + { + "epoch": 0.9749293573017496, + "flos": 15106382080800.0, + "grad_norm": 4.2229534542148155, + "learning_rate": 6.559525730050364e-09, + "loss": 0.7711, + "num_input_tokens_seen": 172841180, + "step": 8108 + }, + { + "epoch": 0.9750496001923886, + "flos": 13488542666760.0, + "grad_norm": 4.380891375482211, + "learning_rate": 6.496639070224574e-09, + "loss": 0.7462, + "num_input_tokens_seen": 172859385, + "step": 8109 + }, + { + "epoch": 0.9751698430830277, + "flos": 14298428002560.0, + "grad_norm": 7.1680725256452416, + "learning_rate": 6.4340548232739714e-09, + "loss": 0.8268, + "num_input_tokens_seen": 172875305, + "step": 8110 + }, + { + "epoch": 0.9752900859736668, + "flos": 17241998124720.0, + "grad_norm": 3.3502839323963785, + "learning_rate": 6.371772998692071e-09, + "loss": 0.7732, + "num_input_tokens_seen": 172894280, + "step": 8111 + }, + { + "epoch": 0.9754103288643059, + "flos": 14838931992360.0, + "grad_norm": 7.677887900716368, + "learning_rate": 6.309793605927094e-09, + "loss": 0.6305, + "num_input_tokens_seen": 172912320, + "step": 8112 + }, + { + "epoch": 0.975530571754945, + "flos": 14133443239200.0, + "grad_norm": 3.958926199968789, + "learning_rate": 6.248116654381297e-09, + "loss": 0.7813, + "num_input_tokens_seen": 172930510, + "step": 8113 + }, + { + "epoch": 0.9756508146455841, + "flos": 17265991883040.0, + "grad_norm": 4.636116483632037, + "learning_rate": 6.186742153410751e-09, + "loss": 0.7075, + "num_input_tokens_seen": 172949725, + "step": 8114 + }, + { + "epoch": 0.9757710575362232, + "flos": 16810291452000.0, + "grad_norm": 9.11286683985976, + "learning_rate": 6.125670112326453e-09, + "loss": 0.8483, + "num_input_tokens_seen": 172968705, + "step": 8115 + }, + { + "epoch": 0.9758913004268622, + "flos": 20510028940560.0, + "grad_norm": 3.5626901943901346, + "learning_rate": 6.064900540392548e-09, + "loss": 0.682, + "num_input_tokens_seen": 172990520, + "step": 8116 + }, + { + "epoch": 0.9760115433175014, + "flos": 16242754347720.0, + "grad_norm": 3.2087836176645936, + "learning_rate": 6.0044334468278835e-09, + "loss": 0.775, + "num_input_tokens_seen": 173009585, + "step": 8117 + }, + { + "epoch": 0.9761317862081405, + "flos": 19240548998640.0, + "grad_norm": 3.9564130173976273, + "learning_rate": 5.944268840805345e-09, + "loss": 0.7057, + "num_input_tokens_seen": 173030050, + "step": 8118 + }, + { + "epoch": 0.9762520290987795, + "flos": 19480332773400.0, + "grad_norm": 9.71839045506556, + "learning_rate": 5.88440673145163e-09, + "loss": 0.6309, + "num_input_tokens_seen": 173050820, + "step": 8119 + }, + { + "epoch": 0.9763722719894187, + "flos": 13138188617160.0, + "grad_norm": 3.868227547170597, + "learning_rate": 5.824847127848142e-09, + "loss": 0.8122, + "num_input_tokens_seen": 173069065, + "step": 8120 + }, + { + "epoch": 0.9764925148800577, + "flos": 16431859509240.0, + "grad_norm": 3.951715512101855, + "learning_rate": 5.765590039029433e-09, + "loss": 0.7655, + "num_input_tokens_seen": 173088105, + "step": 8121 + }, + { + "epoch": 0.9766127577706968, + "flos": 27015733161960.0, + "grad_norm": 8.209010630385684, + "learning_rate": 5.706635473985422e-09, + "loss": 0.7023, + "num_input_tokens_seen": 173111695, + "step": 8122 + }, + { + "epoch": 0.976733000661336, + "flos": 16324075310880.0, + "grad_norm": 5.45078120101091, + "learning_rate": 5.6479834416591764e-09, + "loss": 0.84, + "num_input_tokens_seen": 173130775, + "step": 8123 + }, + { + "epoch": 0.976853243551975, + "flos": 18914790246600.0, + "grad_norm": 4.314451135623693, + "learning_rate": 5.589633950947803e-09, + "loss": 0.6677, + "num_input_tokens_seen": 173147995, + "step": 8124 + }, + { + "epoch": 0.9769734864426141, + "flos": 15648690688320.0, + "grad_norm": 5.109523946008472, + "learning_rate": 5.5315870107035535e-09, + "loss": 0.6728, + "num_input_tokens_seen": 173165765, + "step": 8125 + }, + { + "epoch": 0.9770937293332532, + "flos": 10167427080720.0, + "grad_norm": 3.916722899229424, + "learning_rate": 5.473842629731607e-09, + "loss": 0.7658, + "num_input_tokens_seen": 173183985, + "step": 8126 + }, + { + "epoch": 0.9772139722238923, + "flos": 13084122388200.0, + "grad_norm": 5.125392619655126, + "learning_rate": 5.416400816792066e-09, + "loss": 0.7661, + "num_input_tokens_seen": 173201220, + "step": 8127 + }, + { + "epoch": 0.9773342151145313, + "flos": 14944310033760.0, + "grad_norm": 4.894914105200421, + "learning_rate": 5.359261580598407e-09, + "loss": 0.7487, + "num_input_tokens_seen": 173216780, + "step": 8128 + }, + { + "epoch": 0.9774544580051704, + "flos": 8573423125200.0, + "grad_norm": 4.554708228873821, + "learning_rate": 5.302424929819027e-09, + "loss": 0.7607, + "num_input_tokens_seen": 173230510, + "step": 8129 + }, + { + "epoch": 0.9775747008958096, + "flos": 9789185097720.0, + "grad_norm": 4.409974167936274, + "learning_rate": 5.24589087307592e-09, + "loss": 0.7054, + "num_input_tokens_seen": 173247850, + "step": 8130 + }, + { + "epoch": 0.9776949437864486, + "flos": 43872080873040.0, + "grad_norm": 3.5307381858247178, + "learning_rate": 5.189659418944891e-09, + "loss": 0.6396, + "num_input_tokens_seen": 173277745, + "step": 8131 + }, + { + "epoch": 0.9778151866770877, + "flos": 15486745281120.0, + "grad_norm": 4.475332680382282, + "learning_rate": 5.133730575956674e-09, + "loss": 0.7601, + "num_input_tokens_seen": 173297135, + "step": 8132 + }, + { + "epoch": 0.9779354295677268, + "flos": 15270796964880.0, + "grad_norm": 4.136469530013426, + "learning_rate": 5.0781043525953696e-09, + "loss": 0.7054, + "num_input_tokens_seen": 173314920, + "step": 8133 + }, + { + "epoch": 0.9780556724583659, + "flos": 17160772141440.0, + "grad_norm": 3.974689514701903, + "learning_rate": 5.0227807572995605e-09, + "loss": 0.7158, + "num_input_tokens_seen": 173336615, + "step": 8134 + }, + { + "epoch": 0.9781759153490049, + "flos": 15297766759440.0, + "grad_norm": 5.492893656538955, + "learning_rate": 4.967759798461646e-09, + "loss": 0.6591, + "num_input_tokens_seen": 173354680, + "step": 8135 + }, + { + "epoch": 0.9782961582396441, + "flos": 20751997252560.0, + "grad_norm": 5.364146430239281, + "learning_rate": 4.913041484428282e-09, + "loss": 0.7306, + "num_input_tokens_seen": 173374875, + "step": 8136 + }, + { + "epoch": 0.9784164011302832, + "flos": 18723500547840.0, + "grad_norm": 12.030191587982076, + "learning_rate": 4.858625823500384e-09, + "loss": 0.739, + "num_input_tokens_seen": 173392295, + "step": 8137 + }, + { + "epoch": 0.9785366440209222, + "flos": 21989156866680.0, + "grad_norm": 4.1355280761060875, + "learning_rate": 4.80451282393246e-09, + "loss": 0.7282, + "num_input_tokens_seen": 173412000, + "step": 8138 + }, + { + "epoch": 0.9786568869115614, + "flos": 23749633603680.0, + "grad_norm": 3.6391348347204557, + "learning_rate": 4.750702493933722e-09, + "loss": 0.6637, + "num_input_tokens_seen": 173431605, + "step": 8139 + }, + { + "epoch": 0.9787771298022004, + "flos": 17079989397600.0, + "grad_norm": 3.9712935579604958, + "learning_rate": 4.697194841666974e-09, + "loss": 0.8339, + "num_input_tokens_seen": 173450250, + "step": 8140 + }, + { + "epoch": 0.9788973726928395, + "flos": 15701047279440.0, + "grad_norm": 4.396634696244581, + "learning_rate": 4.6439898752492764e-09, + "loss": 0.8023, + "num_input_tokens_seen": 173470110, + "step": 8141 + }, + { + "epoch": 0.9790176155834787, + "flos": 50806425242400.0, + "grad_norm": 0.7593198905382317, + "learning_rate": 4.591087602751731e-09, + "loss": 0.6308, + "num_input_tokens_seen": 173531690, + "step": 8142 + }, + { + "epoch": 0.9791378584741177, + "flos": 15673001046240.0, + "grad_norm": 14.339405873013554, + "learning_rate": 4.538488032199916e-09, + "loss": 0.7073, + "num_input_tokens_seen": 173549510, + "step": 8143 + }, + { + "epoch": 0.9792581013647568, + "flos": 14727190299000.0, + "grad_norm": 10.101463936946644, + "learning_rate": 4.486191171572784e-09, + "loss": 0.6652, + "num_input_tokens_seen": 173566500, + "step": 8144 + }, + { + "epoch": 0.9793783442553959, + "flos": 17372984582400.0, + "grad_norm": 4.480535159163468, + "learning_rate": 4.434197028803766e-09, + "loss": 0.768, + "num_input_tokens_seen": 173585445, + "step": 8145 + }, + { + "epoch": 0.979498587146035, + "flos": 16863281242320.0, + "grad_norm": 4.481873987825742, + "learning_rate": 4.3825056117805514e-09, + "loss": 0.7934, + "num_input_tokens_seen": 173601050, + "step": 8146 + }, + { + "epoch": 0.979618830036674, + "flos": 10409996931960.0, + "grad_norm": 7.427055654879906, + "learning_rate": 4.331116928344425e-09, + "loss": 0.7784, + "num_input_tokens_seen": 173617085, + "step": 8147 + }, + { + "epoch": 0.9797390729273132, + "flos": 12192662809560.0, + "grad_norm": 5.910120104813464, + "learning_rate": 4.28003098629115e-09, + "loss": 0.6093, + "num_input_tokens_seen": 173632940, + "step": 8148 + }, + { + "epoch": 0.9798593158179523, + "flos": 17968599579840.0, + "grad_norm": 4.617012343335853, + "learning_rate": 4.229247793370305e-09, + "loss": 0.7801, + "num_input_tokens_seen": 173651785, + "step": 8149 + }, + { + "epoch": 0.9799795587085913, + "flos": 20023591179720.0, + "grad_norm": 4.225857095854057, + "learning_rate": 4.178767357285951e-09, + "loss": 0.6846, + "num_input_tokens_seen": 173673135, + "step": 8150 + }, + { + "epoch": 0.9800998015992305, + "flos": 19266980573880.0, + "grad_norm": 4.803519849859897, + "learning_rate": 4.128589685695516e-09, + "loss": 0.681, + "num_input_tokens_seen": 173693280, + "step": 8151 + }, + { + "epoch": 0.9802200444898695, + "flos": 12189243533880.0, + "grad_norm": 5.1039975300486935, + "learning_rate": 4.078714786211135e-09, + "loss": 0.8354, + "num_input_tokens_seen": 173708850, + "step": 8152 + }, + { + "epoch": 0.9803402873805086, + "flos": 18241210241760.0, + "grad_norm": 28.11818819419592, + "learning_rate": 4.029142666398977e-09, + "loss": 0.756, + "num_input_tokens_seen": 173728735, + "step": 8153 + }, + { + "epoch": 0.9804605302711478, + "flos": 16512705573000.0, + "grad_norm": 3.3552290061417245, + "learning_rate": 3.979873333778805e-09, + "loss": 0.7905, + "num_input_tokens_seen": 173746630, + "step": 8154 + }, + { + "epoch": 0.9805807731617868, + "flos": 28604988123480.0, + "grad_norm": 3.4367602852897554, + "learning_rate": 3.930906795824862e-09, + "loss": 0.7237, + "num_input_tokens_seen": 173767025, + "step": 8155 + }, + { + "epoch": 0.9807010160524259, + "flos": 13002643125240.0, + "grad_norm": 7.337539969080343, + "learning_rate": 3.882243059965207e-09, + "loss": 0.7647, + "num_input_tokens_seen": 173784460, + "step": 8156 + }, + { + "epoch": 0.980821258943065, + "flos": 9843251326680.0, + "grad_norm": 5.785558796026136, + "learning_rate": 3.833882133582156e-09, + "loss": 0.6509, + "num_input_tokens_seen": 173799840, + "step": 8157 + }, + { + "epoch": 0.9809415018337041, + "flos": 15865113903960.0, + "grad_norm": 3.5173408898519924, + "learning_rate": 3.785824024012285e-09, + "loss": 0.7624, + "num_input_tokens_seen": 173818560, + "step": 8158 + }, + { + "epoch": 0.9810617447243432, + "flos": 17052924623160.0, + "grad_norm": 3.9243819898265757, + "learning_rate": 3.738068738545541e-09, + "loss": 0.7675, + "num_input_tokens_seen": 173837365, + "step": 8159 + }, + { + "epoch": 0.9811819876149822, + "flos": 13379998629360.0, + "grad_norm": 4.965815031185737, + "learning_rate": 3.6906162844265733e-09, + "loss": 0.7635, + "num_input_tokens_seen": 173854170, + "step": 8160 + }, + { + "epoch": 0.9813022305056214, + "flos": 16540181926920.0, + "grad_norm": 3.771677305639492, + "learning_rate": 3.643466668853845e-09, + "loss": 0.7034, + "num_input_tokens_seen": 173871915, + "step": 8161 + }, + { + "epoch": 0.9814224733962604, + "flos": 18621066882720.0, + "grad_norm": 3.6413982045073103, + "learning_rate": 3.59661989898008e-09, + "loss": 0.7276, + "num_input_tokens_seen": 173892690, + "step": 8162 + }, + { + "epoch": 0.9815427162868995, + "flos": 18319808448360.0, + "grad_norm": 5.098007762964383, + "learning_rate": 3.5500759819115934e-09, + "loss": 0.7572, + "num_input_tokens_seen": 173912775, + "step": 8163 + }, + { + "epoch": 0.9816629591775387, + "flos": 15104134223640.0, + "grad_norm": 4.796878578607175, + "learning_rate": 3.5038349247094034e-09, + "loss": 0.8026, + "num_input_tokens_seen": 173929755, + "step": 8164 + }, + { + "epoch": 0.9817832020681777, + "flos": 12921353822040.0, + "grad_norm": 6.675774107204341, + "learning_rate": 3.4578967343878994e-09, + "loss": 0.7616, + "num_input_tokens_seen": 173945680, + "step": 8165 + }, + { + "epoch": 0.9819034449588168, + "flos": 16487097156720.0, + "grad_norm": 2.64453125, + "learning_rate": 3.4122614179161733e-09, + "loss": 0.7909, + "num_input_tokens_seen": 173965360, + "step": 8166 + }, + { + "epoch": 0.9820236878494559, + "flos": 14622825376320.0, + "grad_norm": 3.131706975001095, + "learning_rate": 3.36692898221691e-09, + "loss": 0.7548, + "num_input_tokens_seen": 173983445, + "step": 8167 + }, + { + "epoch": 0.982143930740095, + "flos": 13731872357040.0, + "grad_norm": 5.49812284860919, + "learning_rate": 3.3218994341668305e-09, + "loss": 0.7179, + "num_input_tokens_seen": 174002095, + "step": 8168 + }, + { + "epoch": 0.982264173630734, + "flos": 19482612290520.0, + "grad_norm": 3.797624553272694, + "learning_rate": 3.2771727805971373e-09, + "loss": 0.7386, + "num_input_tokens_seen": 174023200, + "step": 8169 + }, + { + "epoch": 0.9823844165213732, + "flos": 16104866018760.0, + "grad_norm": 6.933317831217724, + "learning_rate": 3.232749028292847e-09, + "loss": 0.7525, + "num_input_tokens_seen": 174039885, + "step": 8170 + }, + { + "epoch": 0.9825046594120123, + "flos": 16026647731680.0, + "grad_norm": 4.243604447982576, + "learning_rate": 3.188628183992792e-09, + "loss": 0.8696, + "num_input_tokens_seen": 174059870, + "step": 8171 + }, + { + "epoch": 0.9826249023026513, + "flos": 43849168044720.0, + "grad_norm": 0.7823916676518372, + "learning_rate": 3.1448102543902844e-09, + "loss": 0.6484, + "num_input_tokens_seen": 174123505, + "step": 8172 + }, + { + "epoch": 0.9827451451932905, + "flos": 11706858247920.0, + "grad_norm": 5.1341391232360625, + "learning_rate": 3.1012952461324515e-09, + "loss": 0.654, + "num_input_tokens_seen": 174142200, + "step": 8173 + }, + { + "epoch": 0.9828653880839295, + "flos": 14808447942240.0, + "grad_norm": 5.256905691501496, + "learning_rate": 3.0580831658204575e-09, + "loss": 0.7311, + "num_input_tokens_seen": 174159500, + "step": 8174 + }, + { + "epoch": 0.9829856309745686, + "flos": 15810509455680.0, + "grad_norm": 3.8120944948763023, + "learning_rate": 3.015174020009281e-09, + "loss": 0.7677, + "num_input_tokens_seen": 174178545, + "step": 8175 + }, + { + "epoch": 0.9831058738652078, + "flos": 17399574457440.0, + "grad_norm": 5.864016716135719, + "learning_rate": 2.9725678152086043e-09, + "loss": 0.7341, + "num_input_tokens_seen": 174196835, + "step": 8176 + }, + { + "epoch": 0.9832261167558468, + "flos": 8190463808160.0, + "grad_norm": 7.063168654395114, + "learning_rate": 2.930264557881257e-09, + "loss": 0.8015, + "num_input_tokens_seen": 174211740, + "step": 8177 + }, + { + "epoch": 0.9833463596464859, + "flos": 44222692693680.0, + "grad_norm": 0.8218002633464555, + "learning_rate": 2.8882642544452163e-09, + "loss": 0.6208, + "num_input_tokens_seen": 174276185, + "step": 8178 + }, + { + "epoch": 0.983466602537125, + "flos": 9897919094880.0, + "grad_norm": 5.324408792978895, + "learning_rate": 2.8465669112716083e-09, + "loss": 0.7293, + "num_input_tokens_seen": 174293430, + "step": 8179 + }, + { + "epoch": 0.9835868454277641, + "flos": 16780124001480.0, + "grad_norm": 4.177781627756965, + "learning_rate": 2.8051725346858177e-09, + "loss": 0.7321, + "num_input_tokens_seen": 174313410, + "step": 8180 + }, + { + "epoch": 0.9837070883184031, + "flos": 20293384105200.0, + "grad_norm": 4.032262630583116, + "learning_rate": 2.7640811309674883e-09, + "loss": 0.6859, + "num_input_tokens_seen": 174332630, + "step": 8181 + }, + { + "epoch": 0.9838273312090423, + "flos": 21454161709920.0, + "grad_norm": 4.95876859576328, + "learning_rate": 2.7232927063498557e-09, + "loss": 0.7982, + "num_input_tokens_seen": 174352725, + "step": 8182 + }, + { + "epoch": 0.9839475740996814, + "flos": 29496764301720.0, + "grad_norm": 3.618932579059713, + "learning_rate": 2.682807267020859e-09, + "loss": 0.667, + "num_input_tokens_seen": 174375205, + "step": 8183 + }, + { + "epoch": 0.9840678169903204, + "flos": 17699978072880.0, + "grad_norm": 3.5891928264254167, + "learning_rate": 2.642624819121808e-09, + "loss": 0.6179, + "num_input_tokens_seen": 174395075, + "step": 8184 + }, + { + "epoch": 0.9841880598809596, + "flos": 10680707996280.0, + "grad_norm": 9.174662021991516, + "learning_rate": 2.6027453687487154e-09, + "loss": 0.6014, + "num_input_tokens_seen": 174411885, + "step": 8185 + }, + { + "epoch": 0.9843083027715986, + "flos": 16349588747280.0, + "grad_norm": 7.093298809474213, + "learning_rate": 2.5631689219509643e-09, + "loss": 0.4902, + "num_input_tokens_seen": 174430285, + "step": 8186 + }, + { + "epoch": 0.9844285456622377, + "flos": 15946878106560.0, + "grad_norm": 9.506898082183248, + "learning_rate": 2.523895484732197e-09, + "loss": 0.8251, + "num_input_tokens_seen": 174449460, + "step": 8187 + }, + { + "epoch": 0.9845487885528769, + "flos": 13159776218520.0, + "grad_norm": 4.697543155457723, + "learning_rate": 2.4849250630505357e-09, + "loss": 0.7404, + "num_input_tokens_seen": 174467425, + "step": 8188 + }, + { + "epoch": 0.9846690314435159, + "flos": 18484033372680.0, + "grad_norm": 4.2570179609153245, + "learning_rate": 2.4462576628172528e-09, + "loss": 0.7182, + "num_input_tokens_seen": 174485775, + "step": 8189 + }, + { + "epoch": 0.984789274334155, + "flos": 13543083795120.0, + "grad_norm": 13.714636253236677, + "learning_rate": 2.407893289898766e-09, + "loss": 0.7267, + "num_input_tokens_seen": 174504525, + "step": 8190 + }, + { + "epoch": 0.984909517224794, + "flos": 20050719274080.0, + "grad_norm": 4.488390038933956, + "learning_rate": 2.3698319501144202e-09, + "loss": 0.8263, + "num_input_tokens_seen": 174525230, + "step": 8191 + }, + { + "epoch": 0.9850297601154332, + "flos": 13677616168320.0, + "grad_norm": 3.322190128569738, + "learning_rate": 2.3320736492382644e-09, + "loss": 0.7142, + "num_input_tokens_seen": 174543785, + "step": 8192 + }, + { + "epoch": 0.9851500030060723, + "flos": 16322555632800.0, + "grad_norm": 3.893129454221457, + "learning_rate": 2.29461839299816e-09, + "loss": 0.6736, + "num_input_tokens_seen": 174563220, + "step": 8193 + }, + { + "epoch": 0.9852702458967113, + "flos": 19315822909440.0, + "grad_norm": 3.5388085145737715, + "learning_rate": 2.257466187076229e-09, + "loss": 0.7876, + "num_input_tokens_seen": 174582145, + "step": 8194 + }, + { + "epoch": 0.9853904887873505, + "flos": 15267979228440.0, + "grad_norm": 6.524683521532086, + "learning_rate": 2.2206170371081854e-09, + "loss": 0.6829, + "num_input_tokens_seen": 174600450, + "step": 8195 + }, + { + "epoch": 0.9855107316779895, + "flos": 18509578469040.0, + "grad_norm": 4.225716497398025, + "learning_rate": 2.1840709486842247e-09, + "loss": 0.8332, + "num_input_tokens_seen": 174619790, + "step": 8196 + }, + { + "epoch": 0.9856309745686286, + "flos": 13920344319360.0, + "grad_norm": 3.9876123777237407, + "learning_rate": 2.1478279273481335e-09, + "loss": 0.7686, + "num_input_tokens_seen": 174637995, + "step": 8197 + }, + { + "epoch": 0.9857512174592677, + "flos": 25072166655840.0, + "grad_norm": 5.622938668527297, + "learning_rate": 2.1118879785981815e-09, + "loss": 0.7756, + "num_input_tokens_seen": 174657855, + "step": 8198 + }, + { + "epoch": 0.9858714603499068, + "flos": 19050968937720.0, + "grad_norm": 4.3688743213433545, + "learning_rate": 2.0762511078862288e-09, + "loss": 0.7829, + "num_input_tokens_seen": 174677920, + "step": 8199 + }, + { + "epoch": 0.9859917032405459, + "flos": 17349972282840.0, + "grad_norm": 5.154105925555441, + "learning_rate": 2.0409173206186183e-09, + "loss": 0.6441, + "num_input_tokens_seen": 174696880, + "step": 8200 + }, + { + "epoch": 0.986111946131185, + "flos": 14568854127240.0, + "grad_norm": 3.4661521462550904, + "learning_rate": 2.0058866221550617e-09, + "loss": 0.8517, + "num_input_tokens_seen": 174714840, + "step": 8201 + }, + { + "epoch": 0.9862321890218241, + "flos": 14487849763680.0, + "grad_norm": 3.3809523192289923, + "learning_rate": 1.971159017809976e-09, + "loss": 0.7405, + "num_input_tokens_seen": 174732850, + "step": 8202 + }, + { + "epoch": 0.9863524319124631, + "flos": 15837004350840.0, + "grad_norm": 4.707433001203037, + "learning_rate": 1.93673451285159e-09, + "loss": 0.7562, + "num_input_tokens_seen": 174751620, + "step": 8203 + }, + { + "epoch": 0.9864726748031023, + "flos": 38870895865680.0, + "grad_norm": 0.7497482672070624, + "learning_rate": 1.9026131125019495e-09, + "loss": 0.59, + "num_input_tokens_seen": 174808710, + "step": 8204 + }, + { + "epoch": 0.9865929176937414, + "flos": 17104109795760.0, + "grad_norm": 2.1741175428898765, + "learning_rate": 1.8687948219371363e-09, + "loss": 0.846, + "num_input_tokens_seen": 174827655, + "step": 8205 + }, + { + "epoch": 0.9867131605843804, + "flos": 15810984355080.0, + "grad_norm": 5.372920144020702, + "learning_rate": 1.835279646287491e-09, + "loss": 0.8682, + "num_input_tokens_seen": 174845385, + "step": 8206 + }, + { + "epoch": 0.9868334034750196, + "flos": 16295237578680.0, + "grad_norm": 3.222933522636328, + "learning_rate": 1.8020675906371685e-09, + "loss": 0.7559, + "num_input_tokens_seen": 174864500, + "step": 8207 + }, + { + "epoch": 0.9869536463656586, + "flos": 18913587168120.0, + "grad_norm": 3.722215208043817, + "learning_rate": 1.7691586600243612e-09, + "loss": 0.7277, + "num_input_tokens_seen": 174883120, + "step": 8208 + }, + { + "epoch": 0.9870738892562977, + "flos": 11949079839600.0, + "grad_norm": 6.932241560489545, + "learning_rate": 1.7365528594415202e-09, + "loss": 0.8525, + "num_input_tokens_seen": 174896910, + "step": 8209 + }, + { + "epoch": 0.9871941321469369, + "flos": 26071790352360.0, + "grad_norm": 4.278621562843265, + "learning_rate": 1.7042501938346888e-09, + "loss": 0.6646, + "num_input_tokens_seen": 174919360, + "step": 8210 + }, + { + "epoch": 0.9873143750375759, + "flos": 15675945422520.0, + "grad_norm": 3.922064666417242, + "learning_rate": 1.6722506681043913e-09, + "loss": 0.757, + "num_input_tokens_seen": 174938040, + "step": 8211 + }, + { + "epoch": 0.987434617928215, + "flos": 11895931749480.0, + "grad_norm": 5.900069297771498, + "learning_rate": 1.640554287104745e-09, + "loss": 0.6696, + "num_input_tokens_seen": 174956035, + "step": 8212 + }, + { + "epoch": 0.9875548608188541, + "flos": 13023914127000.0, + "grad_norm": 3.8937089945438776, + "learning_rate": 1.609161055644348e-09, + "loss": 0.764, + "num_input_tokens_seen": 174971680, + "step": 8213 + }, + { + "epoch": 0.9876751037094932, + "flos": 19152136204440.0, + "grad_norm": 5.208713222636561, + "learning_rate": 1.5780709784849467e-09, + "loss": 0.6617, + "num_input_tokens_seen": 174988420, + "step": 8214 + }, + { + "epoch": 0.9877953466001322, + "flos": 11646934926360.0, + "grad_norm": 3.5818941345928503, + "learning_rate": 1.5472840603436565e-09, + "loss": 0.806, + "num_input_tokens_seen": 175005370, + "step": 8215 + }, + { + "epoch": 0.9879155894907714, + "flos": 13731397457640.0, + "grad_norm": 4.950880245972522, + "learning_rate": 1.5168003058900757e-09, + "loss": 0.7806, + "num_input_tokens_seen": 175023090, + "step": 8216 + }, + { + "epoch": 0.9880358323814105, + "flos": 16377286720920.0, + "grad_norm": 3.0183556403759897, + "learning_rate": 1.4866197197491715e-09, + "loss": 0.9063, + "num_input_tokens_seen": 175042170, + "step": 8217 + }, + { + "epoch": 0.9881560752720495, + "flos": 11409082409160.0, + "grad_norm": 5.321963150808586, + "learning_rate": 1.4567423064988371e-09, + "loss": 0.7413, + "num_input_tokens_seen": 175059240, + "step": 8218 + }, + { + "epoch": 0.9882763181626887, + "flos": 15723711319440.0, + "grad_norm": 6.6465495171102456, + "learning_rate": 1.4271680706718913e-09, + "loss": 0.7649, + "num_input_tokens_seen": 175076635, + "step": 8219 + }, + { + "epoch": 0.9883965610533277, + "flos": 20560169334480.0, + "grad_norm": 4.0949332587424045, + "learning_rate": 1.3978970167543013e-09, + "loss": 0.8076, + "num_input_tokens_seen": 175096535, + "step": 8220 + }, + { + "epoch": 0.9885168039439668, + "flos": 10246816786320.0, + "grad_norm": 3.443124089070392, + "learning_rate": 1.3689291491867372e-09, + "loss": 0.7508, + "num_input_tokens_seen": 175114570, + "step": 8221 + }, + { + "epoch": 0.988637046834606, + "flos": 19374543152520.0, + "grad_norm": 4.782083476105004, + "learning_rate": 1.3402644723636836e-09, + "loss": 0.7227, + "num_input_tokens_seen": 175136320, + "step": 8222 + }, + { + "epoch": 0.988757289725245, + "flos": 18484476612120.0, + "grad_norm": 6.185367101014815, + "learning_rate": 1.311902990633218e-09, + "loss": 0.8114, + "num_input_tokens_seen": 175155005, + "step": 8223 + }, + { + "epoch": 0.9888775326158841, + "flos": 19320540243480.0, + "grad_norm": 3.2793825194016524, + "learning_rate": 1.2838447082978987e-09, + "loss": 0.6977, + "num_input_tokens_seen": 175175880, + "step": 8224 + }, + { + "epoch": 0.9889977755065231, + "flos": 17808680410080.0, + "grad_norm": 11.800322214673512, + "learning_rate": 1.2560896296143208e-09, + "loss": 0.8063, + "num_input_tokens_seen": 175194065, + "step": 8225 + }, + { + "epoch": 0.9891180183971623, + "flos": 13837946917560.0, + "grad_norm": 5.011939289952243, + "learning_rate": 1.2286377587926722e-09, + "loss": 0.7884, + "num_input_tokens_seen": 175210575, + "step": 8226 + }, + { + "epoch": 0.9892382612878013, + "flos": 19185501310920.0, + "grad_norm": 4.199754235479468, + "learning_rate": 1.2014890999973992e-09, + "loss": 0.7413, + "num_input_tokens_seen": 175227215, + "step": 8227 + }, + { + "epoch": 0.9893585041784404, + "flos": 18647878377480.0, + "grad_norm": 11.612380530786824, + "learning_rate": 1.1746436573472073e-09, + "loss": 0.7644, + "num_input_tokens_seen": 175248670, + "step": 8228 + }, + { + "epoch": 0.9894787470690796, + "flos": 14754001793760.0, + "grad_norm": 6.620365177005291, + "learning_rate": 1.1481014349141726e-09, + "loss": 0.6735, + "num_input_tokens_seen": 175265610, + "step": 8229 + }, + { + "epoch": 0.9895989899597186, + "flos": 18051535200960.0, + "grad_norm": 3.8451451196555557, + "learning_rate": 1.121862436724852e-09, + "loss": 0.8235, + "num_input_tokens_seen": 175284170, + "step": 8230 + }, + { + "epoch": 0.9897192328503577, + "flos": 15945928307760.0, + "grad_norm": 6.715972468216919, + "learning_rate": 1.0959266667598388e-09, + "loss": 0.7011, + "num_input_tokens_seen": 175302705, + "step": 8231 + }, + { + "epoch": 0.9898394757409968, + "flos": 15595352638440.0, + "grad_norm": 6.9865653680538795, + "learning_rate": 1.0702941289533196e-09, + "loss": 0.7243, + "num_input_tokens_seen": 175321100, + "step": 8232 + }, + { + "epoch": 0.9899597186316359, + "flos": 13536783463080.0, + "grad_norm": 5.3453141984934085, + "learning_rate": 1.0449648271939615e-09, + "loss": 0.8695, + "num_input_tokens_seen": 175337165, + "step": 8233 + }, + { + "epoch": 0.990079961522275, + "flos": 17401790654640.0, + "grad_norm": 3.40597267071819, + "learning_rate": 1.0199387653240243e-09, + "loss": 0.7209, + "num_input_tokens_seen": 175356575, + "step": 8234 + }, + { + "epoch": 0.9902002044129141, + "flos": 11949681378840.0, + "grad_norm": 5.610932625343799, + "learning_rate": 9.952159471400267e-10, + "loss": 0.6876, + "num_input_tokens_seen": 175373335, + "step": 8235 + }, + { + "epoch": 0.9903204473035532, + "flos": 16508051558880.0, + "grad_norm": 4.0202856662653845, + "learning_rate": 9.707963763923022e-10, + "loss": 0.8251, + "num_input_tokens_seen": 175392105, + "step": 8236 + }, + { + "epoch": 0.9904406901941922, + "flos": 11760861156960.0, + "grad_norm": 3.809909878763769, + "learning_rate": 9.466800567854427e-10, + "loss": 0.7718, + "num_input_tokens_seen": 175410425, + "step": 8237 + }, + { + "epoch": 0.9905609330848314, + "flos": 19537058439000.0, + "grad_norm": 7.3315325462224035, + "learning_rate": 9.228669919778553e-10, + "loss": 0.6778, + "num_input_tokens_seen": 175429070, + "step": 8238 + }, + { + "epoch": 0.9906811759754705, + "flos": 16917379131240.0, + "grad_norm": 8.34375960103029, + "learning_rate": 8.993571855817617e-10, + "loss": 0.7796, + "num_input_tokens_seen": 175447620, + "step": 8239 + }, + { + "epoch": 0.9908014188661095, + "flos": 16161655004280.0, + "grad_norm": 3.0660994260807644, + "learning_rate": 8.761506411638642e-10, + "loss": 0.7219, + "num_input_tokens_seen": 175466805, + "step": 8240 + }, + { + "epoch": 0.9909216617567487, + "flos": 14053958553720.0, + "grad_norm": 4.203634578630057, + "learning_rate": 8.53247362244236e-10, + "loss": 0.7198, + "num_input_tokens_seen": 175485335, + "step": 8241 + }, + { + "epoch": 0.9910419046473877, + "flos": 17294164756080.0, + "grad_norm": 3.5606466458268486, + "learning_rate": 8.306473522976532e-10, + "loss": 0.6728, + "num_input_tokens_seen": 175504460, + "step": 8242 + }, + { + "epoch": 0.9911621475380268, + "flos": 16621154630520.0, + "grad_norm": 9.477767226210943, + "learning_rate": 8.083506147522623e-10, + "loss": 0.7075, + "num_input_tokens_seen": 175523575, + "step": 8243 + }, + { + "epoch": 0.991282390428666, + "flos": 9816661451640.0, + "grad_norm": 4.561731143078088, + "learning_rate": 7.863571529906909e-10, + "loss": 0.8402, + "num_input_tokens_seen": 175538880, + "step": 8244 + }, + { + "epoch": 0.991402633319305, + "flos": 45577989313080.0, + "grad_norm": 0.8172791482506535, + "learning_rate": 7.646669703489372e-10, + "loss": 0.6425, + "num_input_tokens_seen": 175602910, + "step": 8245 + }, + { + "epoch": 0.9915228762099441, + "flos": 13704269363280.0, + "grad_norm": 2.8962578405176163, + "learning_rate": 7.432800701177023e-10, + "loss": 0.5621, + "num_input_tokens_seen": 175620630, + "step": 8246 + }, + { + "epoch": 0.9916431191005832, + "flos": 48614463452400.0, + "grad_norm": 0.8085887374353993, + "learning_rate": 7.221964555415017e-10, + "loss": 0.5933, + "num_input_tokens_seen": 175680010, + "step": 8247 + }, + { + "epoch": 0.9917633619912223, + "flos": 12084656991480.0, + "grad_norm": 3.782719728450199, + "learning_rate": 7.01416129818222e-10, + "loss": 0.7356, + "num_input_tokens_seen": 175697350, + "step": 8248 + }, + { + "epoch": 0.9918836048818613, + "flos": 18619103965200.0, + "grad_norm": 4.663037364766656, + "learning_rate": 6.809390961006745e-10, + "loss": 0.5688, + "num_input_tokens_seen": 175717200, + "step": 8249 + }, + { + "epoch": 0.9920038477725005, + "flos": 18349627639320.0, + "grad_norm": 4.709607089879158, + "learning_rate": 6.607653574948191e-10, + "loss": 0.6783, + "num_input_tokens_seen": 175737700, + "step": 8250 + }, + { + "epoch": 0.9921240906631396, + "flos": 15967864168680.0, + "grad_norm": 3.9943547944566746, + "learning_rate": 6.408949170613187e-10, + "loss": 0.8057, + "num_input_tokens_seen": 175756685, + "step": 8251 + }, + { + "epoch": 0.9922443335537786, + "flos": 18214430406960.0, + "grad_norm": 2.757084909912999, + "learning_rate": 6.213277778144288e-10, + "loss": 0.802, + "num_input_tokens_seen": 175778050, + "step": 8252 + }, + { + "epoch": 0.9923645764444178, + "flos": 15807818359080.0, + "grad_norm": 4.143766773126639, + "learning_rate": 6.020639427224416e-10, + "loss": 0.6607, + "num_input_tokens_seen": 175795415, + "step": 8253 + }, + { + "epoch": 0.9924848193350568, + "flos": 18315059454360.0, + "grad_norm": 5.00826362565584, + "learning_rate": 5.831034147076864e-10, + "loss": 0.7069, + "num_input_tokens_seen": 175812385, + "step": 8254 + }, + { + "epoch": 0.9926050622256959, + "flos": 50819722425600.0, + "grad_norm": 0.7112540494002396, + "learning_rate": 5.644461966463065e-10, + "loss": 0.5669, + "num_input_tokens_seen": 175879715, + "step": 8255 + }, + { + "epoch": 0.9927253051163349, + "flos": 15298146678960.0, + "grad_norm": 8.921529946945103, + "learning_rate": 5.460922913687049e-10, + "loss": 0.7382, + "num_input_tokens_seen": 175898525, + "step": 8256 + }, + { + "epoch": 0.9928455480069741, + "flos": 16322523972840.0, + "grad_norm": 3.678591943755065, + "learning_rate": 5.280417016593208e-10, + "loss": 0.7281, + "num_input_tokens_seen": 175918035, + "step": 8257 + }, + { + "epoch": 0.9929657908976132, + "flos": 12678372391320.0, + "grad_norm": 4.276548602604612, + "learning_rate": 5.102944302559642e-10, + "loss": 0.7392, + "num_input_tokens_seen": 175935250, + "step": 8258 + }, + { + "epoch": 0.9930860337882522, + "flos": 16593963216240.0, + "grad_norm": 5.060716479693542, + "learning_rate": 4.9285047985137e-10, + "loss": 0.7664, + "num_input_tokens_seen": 175954390, + "step": 8259 + }, + { + "epoch": 0.9932062766788914, + "flos": 20644466333880.0, + "grad_norm": 4.36566342969731, + "learning_rate": 4.757098530916436e-10, + "loss": 0.7386, + "num_input_tokens_seen": 175974555, + "step": 8260 + }, + { + "epoch": 0.9933265195695304, + "flos": 14730387954960.0, + "grad_norm": 5.771111583772533, + "learning_rate": 4.5887255257670563e-10, + "loss": 0.77, + "num_input_tokens_seen": 175991315, + "step": 8261 + }, + { + "epoch": 0.9934467624601695, + "flos": 15622639032600.0, + "grad_norm": 4.057890639289049, + "learning_rate": 4.4233858086117906e-10, + "loss": 0.7499, + "num_input_tokens_seen": 176009560, + "step": 8262 + }, + { + "epoch": 0.9935670053508087, + "flos": 14434005154440.0, + "grad_norm": 4.419587423133822, + "learning_rate": 4.261079404528356e-10, + "loss": 0.6548, + "num_input_tokens_seen": 176028760, + "step": 8263 + }, + { + "epoch": 0.9936872482414477, + "flos": 16078624403280.0, + "grad_norm": 4.08067224686425, + "learning_rate": 4.1018063381437205e-10, + "loss": 0.6687, + "num_input_tokens_seen": 176048865, + "step": 8264 + }, + { + "epoch": 0.9938074911320868, + "flos": 51482411404200.0, + "grad_norm": 0.9282482450711618, + "learning_rate": 3.9455666336141167e-10, + "loss": 0.6529, + "num_input_tokens_seen": 176112365, + "step": 8265 + }, + { + "epoch": 0.9939277340227259, + "flos": 10976267637840.0, + "grad_norm": 5.768813494728607, + "learning_rate": 3.7923603146450267e-10, + "loss": 0.8033, + "num_input_tokens_seen": 176128145, + "step": 8266 + }, + { + "epoch": 0.994047976913365, + "flos": 12921480461880.0, + "grad_norm": 4.069601102959236, + "learning_rate": 3.642187404473418e-10, + "loss": 0.7805, + "num_input_tokens_seen": 176146025, + "step": 8267 + }, + { + "epoch": 0.994168219804004, + "flos": 14001348682920.0, + "grad_norm": 6.377623971700908, + "learning_rate": 3.495047925885508e-10, + "loss": 0.8341, + "num_input_tokens_seen": 176164080, + "step": 8268 + }, + { + "epoch": 0.9942884626946432, + "flos": 13023565867440.0, + "grad_norm": 5.056049145833918, + "learning_rate": 3.350941901199e-10, + "loss": 0.8239, + "num_input_tokens_seen": 176180720, + "step": 8269 + }, + { + "epoch": 0.9944087055852823, + "flos": 13722406029000.0, + "grad_norm": 4.053297682990184, + "learning_rate": 3.2098693522764066e-10, + "loss": 0.821, + "num_input_tokens_seen": 176193640, + "step": 8270 + }, + { + "epoch": 0.9945289484759213, + "flos": 15352434527640.0, + "grad_norm": 4.355616326675833, + "learning_rate": 3.071830300516165e-10, + "loss": 0.8052, + "num_input_tokens_seen": 176211190, + "step": 8271 + }, + { + "epoch": 0.9946491913665605, + "flos": 10734267665880.0, + "grad_norm": 4.246376737062599, + "learning_rate": 2.9368247668615234e-10, + "loss": 0.6841, + "num_input_tokens_seen": 176229500, + "step": 8272 + }, + { + "epoch": 0.9947694342571995, + "flos": 9188694466440.0, + "grad_norm": 4.530914083723287, + "learning_rate": 2.804852771789434e-10, + "loss": 0.5926, + "num_input_tokens_seen": 176242520, + "step": 8273 + }, + { + "epoch": 0.9948896771478386, + "flos": 13758557211960.0, + "grad_norm": 5.737581193469959, + "learning_rate": 2.675914335321661e-10, + "loss": 0.5504, + "num_input_tokens_seen": 176260995, + "step": 8274 + }, + { + "epoch": 0.9950099200384778, + "flos": 18240988622040.0, + "grad_norm": 6.239741030967946, + "learning_rate": 2.550009477018111e-10, + "loss": 0.7625, + "num_input_tokens_seen": 176279485, + "step": 8275 + }, + { + "epoch": 0.9951301629291168, + "flos": 17377765236360.0, + "grad_norm": 3.1289947916330725, + "learning_rate": 2.4271382159790634e-10, + "loss": 0.6179, + "num_input_tokens_seen": 176296635, + "step": 8276 + }, + { + "epoch": 0.9952504058197559, + "flos": 16269977421960.0, + "grad_norm": 2.7584597770851897, + "learning_rate": 2.3073005708429406e-10, + "loss": 0.8426, + "num_input_tokens_seen": 176316000, + "step": 8277 + }, + { + "epoch": 0.995370648710395, + "flos": 15510169160160.0, + "grad_norm": 4.051616703940467, + "learning_rate": 2.190496559788535e-10, + "loss": 0.7099, + "num_input_tokens_seen": 176334005, + "step": 8278 + }, + { + "epoch": 0.9954908916010341, + "flos": 10815556969080.0, + "grad_norm": 5.68828468359325, + "learning_rate": 2.0767262005372265e-10, + "loss": 0.7428, + "num_input_tokens_seen": 176351240, + "step": 8279 + }, + { + "epoch": 0.9956111344916732, + "flos": 14028540097200.0, + "grad_norm": 15.866392670731301, + "learning_rate": 1.965989510346322e-10, + "loss": 0.7347, + "num_input_tokens_seen": 176370080, + "step": 8280 + }, + { + "epoch": 0.9957313773823123, + "flos": 14649256951560.0, + "grad_norm": 5.2245429318205, + "learning_rate": 1.8582865060134955e-10, + "loss": 0.6925, + "num_input_tokens_seen": 176387990, + "step": 8281 + }, + { + "epoch": 0.9958516202729514, + "flos": 42359972251320.0, + "grad_norm": 0.8092250478867352, + "learning_rate": 1.7536172038790098e-10, + "loss": 0.5803, + "num_input_tokens_seen": 176448020, + "step": 8282 + }, + { + "epoch": 0.9959718631635904, + "flos": 20374420128720.0, + "grad_norm": 10.04973403806488, + "learning_rate": 1.651981619819054e-10, + "loss": 0.6708, + "num_input_tokens_seen": 176464890, + "step": 8283 + }, + { + "epoch": 0.9960921060542296, + "flos": 17592542134080.0, + "grad_norm": 3.6291703537675164, + "learning_rate": 1.5533797692546257e-10, + "loss": 0.6748, + "num_input_tokens_seen": 176483345, + "step": 8284 + }, + { + "epoch": 0.9962123489448687, + "flos": 13650456414000.0, + "grad_norm": 4.219909282631286, + "learning_rate": 1.4578116671404296e-10, + "loss": 0.8172, + "num_input_tokens_seen": 176501345, + "step": 8285 + }, + { + "epoch": 0.9963325918355077, + "flos": 14622413796840.0, + "grad_norm": 5.353985893473176, + "learning_rate": 1.3652773279759777e-10, + "loss": 0.7038, + "num_input_tokens_seen": 176517715, + "step": 8286 + }, + { + "epoch": 0.9964528347261468, + "flos": 24315524390040.0, + "grad_norm": 4.326027923324558, + "learning_rate": 1.2757767657989305e-10, + "loss": 0.6116, + "num_input_tokens_seen": 176541225, + "step": 8287 + }, + { + "epoch": 0.9965730776167859, + "flos": 16915194594000.0, + "grad_norm": 5.303924280917853, + "learning_rate": 1.1893099941850948e-10, + "loss": 0.8528, + "num_input_tokens_seen": 176559840, + "step": 8288 + }, + { + "epoch": 0.996693320507425, + "flos": 16809341653200.0, + "grad_norm": 6.312868013135514, + "learning_rate": 1.105877026252866e-10, + "loss": 0.7523, + "num_input_tokens_seen": 176577890, + "step": 8289 + }, + { + "epoch": 0.996813563398064, + "flos": 9598591918080.0, + "grad_norm": 3.812950232764527, + "learning_rate": 1.0254778746565663e-10, + "loss": 0.7109, + "num_input_tokens_seen": 176592885, + "step": 8290 + }, + { + "epoch": 0.9969338062887032, + "flos": 10653579901920.0, + "grad_norm": 3.8258726383672683, + "learning_rate": 9.481125515953259e-11, + "loss": 0.7196, + "num_input_tokens_seen": 176610665, + "step": 8291 + }, + { + "epoch": 0.9970540491793423, + "flos": 18858824420040.0, + "grad_norm": 3.340333423076106, + "learning_rate": 8.737810688064228e-11, + "loss": 0.7872, + "num_input_tokens_seen": 176630220, + "step": 8292 + }, + { + "epoch": 0.9971742920699813, + "flos": 15702535297560.0, + "grad_norm": 3.50168868607345, + "learning_rate": 8.024834375608414e-11, + "loss": 0.7735, + "num_input_tokens_seen": 176648530, + "step": 8293 + }, + { + "epoch": 0.9972945349606205, + "flos": 53257922130840.0, + "grad_norm": 0.8343543275136708, + "learning_rate": 7.342196686788149e-11, + "loss": 0.6542, + "num_input_tokens_seen": 176701415, + "step": 8294 + }, + { + "epoch": 0.9974147778512595, + "flos": 14375791470720.0, + "grad_norm": 3.630300921112743, + "learning_rate": 6.689897725142834e-11, + "loss": 0.6619, + "num_input_tokens_seen": 176720610, + "step": 8295 + }, + { + "epoch": 0.9975350207418986, + "flos": 11625189025200.0, + "grad_norm": 3.422513579960946, + "learning_rate": 6.067937589615545e-11, + "loss": 0.8573, + "num_input_tokens_seen": 176738405, + "step": 8296 + }, + { + "epoch": 0.9976552636325378, + "flos": 44193981601320.0, + "grad_norm": 0.764050207058674, + "learning_rate": 5.476316374575241e-11, + "loss": 0.5747, + "num_input_tokens_seen": 176801610, + "step": 8297 + }, + { + "epoch": 0.9977755065231768, + "flos": 16455093428520.0, + "grad_norm": 6.43916136404933, + "learning_rate": 4.9150341697723476e-11, + "loss": 0.7206, + "num_input_tokens_seen": 176821220, + "step": 8298 + }, + { + "epoch": 0.9978957494138159, + "flos": 19077178893240.0, + "grad_norm": 3.271379720807409, + "learning_rate": 4.384091060338768e-11, + "loss": 0.6474, + "num_input_tokens_seen": 176841410, + "step": 8299 + }, + { + "epoch": 0.998015992304455, + "flos": 16566645162120.0, + "grad_norm": 5.216474391006006, + "learning_rate": 3.883487126810081e-11, + "loss": 0.7142, + "num_input_tokens_seen": 176860390, + "step": 8300 + }, + { + "epoch": 0.9981362351950941, + "flos": 13299754104840.0, + "grad_norm": 6.744372247054275, + "learning_rate": 3.41322244516995e-11, + "loss": 0.7871, + "num_input_tokens_seen": 176878055, + "step": 8301 + }, + { + "epoch": 0.9982564780857331, + "flos": 24586203794400.0, + "grad_norm": 3.0528747955848257, + "learning_rate": 2.9732970866946925e-11, + "loss": 0.6089, + "num_input_tokens_seen": 176897655, + "step": 8302 + }, + { + "epoch": 0.9983767209763723, + "flos": 11301583150440.0, + "grad_norm": 3.9762702639656133, + "learning_rate": 2.563711118175327e-11, + "loss": 0.7646, + "num_input_tokens_seen": 176914260, + "step": 8303 + }, + { + "epoch": 0.9984969638670114, + "flos": 14595507322200.0, + "grad_norm": 3.440611384864637, + "learning_rate": 2.184464601717728e-11, + "loss": 0.8147, + "num_input_tokens_seen": 176932295, + "step": 8304 + }, + { + "epoch": 0.9986172067576504, + "flos": 14892554981880.0, + "grad_norm": 4.033156070019091, + "learning_rate": 1.8355575948758585e-11, + "loss": 0.7533, + "num_input_tokens_seen": 176950000, + "step": 8305 + }, + { + "epoch": 0.9987374496482896, + "flos": 17372731302720.0, + "grad_norm": 8.449315818761646, + "learning_rate": 1.5169901505407424e-11, + "loss": 0.7209, + "num_input_tokens_seen": 176966785, + "step": 8306 + }, + { + "epoch": 0.9988576925389286, + "flos": 18347126502480.0, + "grad_norm": 3.3201519377538293, + "learning_rate": 1.228762317073695e-11, + "loss": 0.7301, + "num_input_tokens_seen": 176985335, + "step": 8307 + }, + { + "epoch": 0.9989779354295677, + "flos": 22966686402480.0, + "grad_norm": 6.265713888474845, + "learning_rate": 9.70874138195299e-12, + "loss": 0.7722, + "num_input_tokens_seen": 177006965, + "step": 8308 + }, + { + "epoch": 0.9990981783202069, + "flos": 14298807922080.0, + "grad_norm": 3.5834148900413307, + "learning_rate": 7.433256530076093e-12, + "loss": 0.7302, + "num_input_tokens_seen": 177026640, + "step": 8309 + }, + { + "epoch": 0.9992184212108459, + "flos": 12541782120720.0, + "grad_norm": 4.276149858352018, + "learning_rate": 5.46116896038562e-12, + "loss": 0.7428, + "num_input_tokens_seen": 177040770, + "step": 8310 + }, + { + "epoch": 0.999338664101485, + "flos": 34224424999680.0, + "grad_norm": 3.8161113707087146, + "learning_rate": 3.792478972197699e-12, + "loss": 0.612, + "num_input_tokens_seen": 177061075, + "step": 8311 + }, + { + "epoch": 0.9994589069921241, + "flos": 11031252005640.0, + "grad_norm": 4.530000080167588, + "learning_rate": 2.4271868181990895e-12, + "loss": 0.6652, + "num_input_tokens_seen": 177077960, + "step": 8312 + }, + { + "epoch": 0.9995791498827632, + "flos": 9087273920040.0, + "grad_norm": 3.7507020928854367, + "learning_rate": 1.3652927060014973e-12, + "loss": 0.7908, + "num_input_tokens_seen": 177093275, + "step": 8313 + }, + { + "epoch": 0.9996993927734023, + "flos": 14055921471240.0, + "grad_norm": 4.346244699399764, + "learning_rate": 6.067967965872612e-13, + "loss": 0.6298, + "num_input_tokens_seen": 177112605, + "step": 8314 + }, + { + "epoch": 0.9998196356640414, + "flos": 46402175968080.0, + "grad_norm": 3.445108628242103, + "learning_rate": 1.5169920497548615e-13, + "loss": 0.7604, + "num_input_tokens_seen": 177136945, + "step": 8315 + }, + { + "epoch": 0.9999398785546805, + "flos": 37505409230520.0, + "grad_norm": 1.7315440399887767, + "learning_rate": 0.0, + "loss": 0.5654, + "num_input_tokens_seen": 177185545, + "step": 8316 + } + ], + "logging_steps": 1.0, + "max_steps": 8316, + "num_input_tokens_seen": 177185545, + "num_train_epochs": 1, + "save_steps": 832, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 5.1159586152277606e+17, + "train_batch_size": 5, + "trial_name": null, + "trial_params": null +}