diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,72814 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 10396, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 9.619084263178146e-05, + "grad_norm": 49.04870658248125, + "learning_rate": 6.41025641025641e-08, + "loss": 2.1796, + "step": 1 + }, + { + "epoch": 0.00019238168526356292, + "grad_norm": 68.81559431096773, + "learning_rate": 1.282051282051282e-07, + "loss": 2.2164, + "step": 2 + }, + { + "epoch": 0.00028857252789534437, + "grad_norm": 5.1379328101118205, + "learning_rate": 1.9230769230769234e-07, + "loss": 0.8158, + "step": 3 + }, + { + "epoch": 0.00038476337052712584, + "grad_norm": 79.76885727093025, + "learning_rate": 2.564102564102564e-07, + "loss": 2.2401, + "step": 4 + }, + { + "epoch": 0.00048095421315890726, + "grad_norm": 43.49271564692491, + "learning_rate": 3.205128205128205e-07, + "loss": 1.9641, + "step": 5 + }, + { + "epoch": 0.0005771450557906887, + "grad_norm": 47.973660254484095, + "learning_rate": 3.846153846153847e-07, + "loss": 2.0503, + "step": 6 + }, + { + "epoch": 0.0006733358984224702, + "grad_norm": 57.92907413977099, + "learning_rate": 4.4871794871794876e-07, + "loss": 1.9863, + "step": 7 + }, + { + "epoch": 0.0007695267410542517, + "grad_norm": 47.96212001394869, + "learning_rate": 5.128205128205128e-07, + "loss": 2.0875, + "step": 8 + }, + { + "epoch": 0.0008657175836860331, + "grad_norm": 55.124860858908335, + "learning_rate": 5.76923076923077e-07, + "loss": 1.9165, + "step": 9 + }, + { + "epoch": 0.0009619084263178145, + "grad_norm": 38.64245118612051, + "learning_rate": 6.41025641025641e-07, + "loss": 1.763, + "step": 10 + }, + { + "epoch": 0.001058099268949596, + "grad_norm": 51.42912179370831, + "learning_rate": 7.051282051282052e-07, + "loss": 1.8213, + "step": 11 + }, + { + "epoch": 0.0011542901115813775, + "grad_norm": 48.62763622334891, + "learning_rate": 7.692307692307694e-07, + "loss": 1.9155, + "step": 12 + }, + { + "epoch": 0.001250480954213159, + "grad_norm": 29.347600792254788, + "learning_rate": 8.333333333333333e-07, + "loss": 1.7073, + "step": 13 + }, + { + "epoch": 0.0013466717968449403, + "grad_norm": 22.90257148323482, + "learning_rate": 8.974358974358975e-07, + "loss": 1.6735, + "step": 14 + }, + { + "epoch": 0.0014428626394767218, + "grad_norm": 18.875980375606012, + "learning_rate": 9.615384615384617e-07, + "loss": 1.6251, + "step": 15 + }, + { + "epoch": 0.0015390534821085034, + "grad_norm": 3.9706038369465047, + "learning_rate": 1.0256410256410257e-06, + "loss": 0.788, + "step": 16 + }, + { + "epoch": 0.0016352443247402847, + "grad_norm": 14.446788500831698, + "learning_rate": 1.0897435897435899e-06, + "loss": 1.5176, + "step": 17 + }, + { + "epoch": 0.0017314351673720662, + "grad_norm": 8.347509504196825, + "learning_rate": 1.153846153846154e-06, + "loss": 1.4072, + "step": 18 + }, + { + "epoch": 0.0018276260100038477, + "grad_norm": 8.876912318224301, + "learning_rate": 1.217948717948718e-06, + "loss": 1.202, + "step": 19 + }, + { + "epoch": 0.001923816852635629, + "grad_norm": 9.290018477169843, + "learning_rate": 1.282051282051282e-06, + "loss": 1.3971, + "step": 20 + }, + { + "epoch": 0.0020200076952674104, + "grad_norm": 7.699156125320154, + "learning_rate": 1.3461538461538462e-06, + "loss": 1.2182, + "step": 21 + }, + { + "epoch": 0.002116198537899192, + "grad_norm": 7.626627147051183, + "learning_rate": 1.4102564102564104e-06, + "loss": 1.3193, + "step": 22 + }, + { + "epoch": 0.0022123893805309734, + "grad_norm": 7.471506356400602, + "learning_rate": 1.4743589743589745e-06, + "loss": 1.2706, + "step": 23 + }, + { + "epoch": 0.002308580223162755, + "grad_norm": 7.731920376411174, + "learning_rate": 1.5384615384615387e-06, + "loss": 1.3713, + "step": 24 + }, + { + "epoch": 0.0024047710657945365, + "grad_norm": 3.5072702572652843, + "learning_rate": 1.602564102564103e-06, + "loss": 0.7535, + "step": 25 + }, + { + "epoch": 0.002500961908426318, + "grad_norm": 6.193501413731751, + "learning_rate": 1.6666666666666667e-06, + "loss": 1.2039, + "step": 26 + }, + { + "epoch": 0.002597152751058099, + "grad_norm": 6.260883512719869, + "learning_rate": 1.7307692307692308e-06, + "loss": 1.2581, + "step": 27 + }, + { + "epoch": 0.0026933435936898806, + "grad_norm": 6.939957381242252, + "learning_rate": 1.794871794871795e-06, + "loss": 1.296, + "step": 28 + }, + { + "epoch": 0.002789534436321662, + "grad_norm": 5.821004973750919, + "learning_rate": 1.8589743589743592e-06, + "loss": 1.281, + "step": 29 + }, + { + "epoch": 0.0028857252789534437, + "grad_norm": 5.665131213413133, + "learning_rate": 1.9230769230769234e-06, + "loss": 1.176, + "step": 30 + }, + { + "epoch": 0.002981916121585225, + "grad_norm": 6.164914535164599, + "learning_rate": 1.987179487179487e-06, + "loss": 1.1102, + "step": 31 + }, + { + "epoch": 0.0030781069642170067, + "grad_norm": 5.3559324300665265, + "learning_rate": 2.0512820512820513e-06, + "loss": 1.1936, + "step": 32 + }, + { + "epoch": 0.003174297806848788, + "grad_norm": 5.915862198725912, + "learning_rate": 2.1153846153846155e-06, + "loss": 1.1601, + "step": 33 + }, + { + "epoch": 0.0032704886494805694, + "grad_norm": 4.308502898496419, + "learning_rate": 2.1794871794871797e-06, + "loss": 1.0742, + "step": 34 + }, + { + "epoch": 0.003366679492112351, + "grad_norm": 4.980253593736551, + "learning_rate": 2.243589743589744e-06, + "loss": 1.2159, + "step": 35 + }, + { + "epoch": 0.0034628703347441324, + "grad_norm": 4.388930625964937, + "learning_rate": 2.307692307692308e-06, + "loss": 1.0887, + "step": 36 + }, + { + "epoch": 0.003559061177375914, + "grad_norm": 4.915423332665316, + "learning_rate": 2.371794871794872e-06, + "loss": 1.2252, + "step": 37 + }, + { + "epoch": 0.0036552520200076955, + "grad_norm": 4.7633389718132575, + "learning_rate": 2.435897435897436e-06, + "loss": 1.2092, + "step": 38 + }, + { + "epoch": 0.0037514428626394766, + "grad_norm": 4.644639031578185, + "learning_rate": 2.5e-06, + "loss": 1.1453, + "step": 39 + }, + { + "epoch": 0.003847633705271258, + "grad_norm": 4.621961267488057, + "learning_rate": 2.564102564102564e-06, + "loss": 1.1972, + "step": 40 + }, + { + "epoch": 0.003943824547903039, + "grad_norm": 4.422768060923217, + "learning_rate": 2.6282051282051286e-06, + "loss": 1.1601, + "step": 41 + }, + { + "epoch": 0.004040015390534821, + "grad_norm": 4.095102817920099, + "learning_rate": 2.6923076923076923e-06, + "loss": 1.1654, + "step": 42 + }, + { + "epoch": 0.004136206233166602, + "grad_norm": 4.227863161721752, + "learning_rate": 2.756410256410257e-06, + "loss": 1.1372, + "step": 43 + }, + { + "epoch": 0.004232397075798384, + "grad_norm": 4.22085827402136, + "learning_rate": 2.8205128205128207e-06, + "loss": 1.1189, + "step": 44 + }, + { + "epoch": 0.004328587918430165, + "grad_norm": 4.116512060145265, + "learning_rate": 2.8846153846153845e-06, + "loss": 1.092, + "step": 45 + }, + { + "epoch": 0.004424778761061947, + "grad_norm": 4.908185327741023, + "learning_rate": 2.948717948717949e-06, + "loss": 1.096, + "step": 46 + }, + { + "epoch": 0.004520969603693728, + "grad_norm": 4.652432726997697, + "learning_rate": 3.012820512820513e-06, + "loss": 1.0852, + "step": 47 + }, + { + "epoch": 0.00461716044632551, + "grad_norm": 4.126827425588134, + "learning_rate": 3.0769230769230774e-06, + "loss": 0.758, + "step": 48 + }, + { + "epoch": 0.004713351288957291, + "grad_norm": 5.160137150626202, + "learning_rate": 3.141025641025641e-06, + "loss": 1.2273, + "step": 49 + }, + { + "epoch": 0.004809542131589073, + "grad_norm": 4.215095642785857, + "learning_rate": 3.205128205128206e-06, + "loss": 1.0572, + "step": 50 + }, + { + "epoch": 0.0049057329742208545, + "grad_norm": 4.0958626223172825, + "learning_rate": 3.2692307692307696e-06, + "loss": 1.1677, + "step": 51 + }, + { + "epoch": 0.005001923816852636, + "grad_norm": 4.776141183393512, + "learning_rate": 3.3333333333333333e-06, + "loss": 1.1975, + "step": 52 + }, + { + "epoch": 0.005098114659484417, + "grad_norm": 3.584451863502573, + "learning_rate": 3.397435897435898e-06, + "loss": 1.0982, + "step": 53 + }, + { + "epoch": 0.005194305502116198, + "grad_norm": 4.176329715100037, + "learning_rate": 3.4615384615384617e-06, + "loss": 1.232, + "step": 54 + }, + { + "epoch": 0.00529049634474798, + "grad_norm": 4.524821924554185, + "learning_rate": 3.5256410256410263e-06, + "loss": 1.0346, + "step": 55 + }, + { + "epoch": 0.005386687187379761, + "grad_norm": 2.9749311229880733, + "learning_rate": 3.58974358974359e-06, + "loss": 0.704, + "step": 56 + }, + { + "epoch": 0.005482878030011543, + "grad_norm": 4.169220943349724, + "learning_rate": 3.653846153846154e-06, + "loss": 1.0101, + "step": 57 + }, + { + "epoch": 0.005579068872643324, + "grad_norm": 4.130156808525423, + "learning_rate": 3.7179487179487184e-06, + "loss": 1.1372, + "step": 58 + }, + { + "epoch": 0.005675259715275106, + "grad_norm": 4.970848504822241, + "learning_rate": 3.782051282051282e-06, + "loss": 1.1756, + "step": 59 + }, + { + "epoch": 0.005771450557906887, + "grad_norm": 4.0618152888887185, + "learning_rate": 3.846153846153847e-06, + "loss": 1.1245, + "step": 60 + }, + { + "epoch": 0.005867641400538669, + "grad_norm": 4.570654948485955, + "learning_rate": 3.910256410256411e-06, + "loss": 1.0749, + "step": 61 + }, + { + "epoch": 0.00596383224317045, + "grad_norm": 4.271290675355735, + "learning_rate": 3.974358974358974e-06, + "loss": 1.0226, + "step": 62 + }, + { + "epoch": 0.006060023085802232, + "grad_norm": 4.289025626046844, + "learning_rate": 4.0384615384615385e-06, + "loss": 1.0814, + "step": 63 + }, + { + "epoch": 0.0061562139284340135, + "grad_norm": 4.432221785466514, + "learning_rate": 4.102564102564103e-06, + "loss": 1.0824, + "step": 64 + }, + { + "epoch": 0.006252404771065794, + "grad_norm": 4.340469414860295, + "learning_rate": 4.166666666666667e-06, + "loss": 1.1142, + "step": 65 + }, + { + "epoch": 0.006348595613697576, + "grad_norm": 3.0522544601998773, + "learning_rate": 4.230769230769231e-06, + "loss": 0.7246, + "step": 66 + }, + { + "epoch": 0.006444786456329357, + "grad_norm": 4.276086348302299, + "learning_rate": 4.294871794871795e-06, + "loss": 1.0613, + "step": 67 + }, + { + "epoch": 0.006540977298961139, + "grad_norm": 4.236183715027317, + "learning_rate": 4.358974358974359e-06, + "loss": 1.189, + "step": 68 + }, + { + "epoch": 0.00663716814159292, + "grad_norm": 2.689868158851896, + "learning_rate": 4.423076923076924e-06, + "loss": 0.6673, + "step": 69 + }, + { + "epoch": 0.006733358984224702, + "grad_norm": 3.040475607047146, + "learning_rate": 4.487179487179488e-06, + "loss": 0.7457, + "step": 70 + }, + { + "epoch": 0.006829549826856483, + "grad_norm": 3.7814074010339023, + "learning_rate": 4.551282051282052e-06, + "loss": 0.9952, + "step": 71 + }, + { + "epoch": 0.006925740669488265, + "grad_norm": 4.17018797765707, + "learning_rate": 4.615384615384616e-06, + "loss": 1.0251, + "step": 72 + }, + { + "epoch": 0.007021931512120046, + "grad_norm": 4.172639258176662, + "learning_rate": 4.6794871794871795e-06, + "loss": 1.0727, + "step": 73 + }, + { + "epoch": 0.007118122354751828, + "grad_norm": 4.999639955079181, + "learning_rate": 4.743589743589744e-06, + "loss": 1.1197, + "step": 74 + }, + { + "epoch": 0.007214313197383609, + "grad_norm": 3.9673963227394817, + "learning_rate": 4.807692307692308e-06, + "loss": 1.0344, + "step": 75 + }, + { + "epoch": 0.007310504040015391, + "grad_norm": 4.948504389012292, + "learning_rate": 4.871794871794872e-06, + "loss": 1.1739, + "step": 76 + }, + { + "epoch": 0.007406694882647172, + "grad_norm": 4.54936846994975, + "learning_rate": 4.935897435897436e-06, + "loss": 1.108, + "step": 77 + }, + { + "epoch": 0.007502885725278953, + "grad_norm": 2.4747647140875877, + "learning_rate": 5e-06, + "loss": 0.7042, + "step": 78 + }, + { + "epoch": 0.007599076567910735, + "grad_norm": 4.541172743732707, + "learning_rate": 5.064102564102565e-06, + "loss": 1.1328, + "step": 79 + }, + { + "epoch": 0.007695267410542516, + "grad_norm": 4.916515377201624, + "learning_rate": 5.128205128205128e-06, + "loss": 1.1361, + "step": 80 + }, + { + "epoch": 0.007791458253174298, + "grad_norm": 4.06996447164567, + "learning_rate": 5.192307692307693e-06, + "loss": 1.136, + "step": 81 + }, + { + "epoch": 0.007887649095806078, + "grad_norm": 4.201291525477582, + "learning_rate": 5.256410256410257e-06, + "loss": 1.1634, + "step": 82 + }, + { + "epoch": 0.00798383993843786, + "grad_norm": 4.406244621861785, + "learning_rate": 5.320512820512821e-06, + "loss": 1.01, + "step": 83 + }, + { + "epoch": 0.008080030781069641, + "grad_norm": 4.051168318502434, + "learning_rate": 5.384615384615385e-06, + "loss": 0.9813, + "step": 84 + }, + { + "epoch": 0.008176221623701423, + "grad_norm": 4.66015800551269, + "learning_rate": 5.448717948717949e-06, + "loss": 1.1142, + "step": 85 + }, + { + "epoch": 0.008272412466333205, + "grad_norm": 4.122189748459255, + "learning_rate": 5.512820512820514e-06, + "loss": 1.006, + "step": 86 + }, + { + "epoch": 0.008368603308964986, + "grad_norm": 2.499981568588016, + "learning_rate": 5.576923076923077e-06, + "loss": 0.7179, + "step": 87 + }, + { + "epoch": 0.008464794151596768, + "grad_norm": 4.616638878168988, + "learning_rate": 5.641025641025641e-06, + "loss": 1.1523, + "step": 88 + }, + { + "epoch": 0.008560984994228549, + "grad_norm": 4.668032238558048, + "learning_rate": 5.705128205128206e-06, + "loss": 1.0737, + "step": 89 + }, + { + "epoch": 0.00865717583686033, + "grad_norm": 4.18963869536402, + "learning_rate": 5.769230769230769e-06, + "loss": 1.0933, + "step": 90 + }, + { + "epoch": 0.008753366679492112, + "grad_norm": 4.318632494281906, + "learning_rate": 5.833333333333334e-06, + "loss": 1.0531, + "step": 91 + }, + { + "epoch": 0.008849557522123894, + "grad_norm": 3.965370254568588, + "learning_rate": 5.897435897435898e-06, + "loss": 1.0645, + "step": 92 + }, + { + "epoch": 0.008945748364755675, + "grad_norm": 4.833866815413111, + "learning_rate": 5.961538461538462e-06, + "loss": 1.0998, + "step": 93 + }, + { + "epoch": 0.009041939207387457, + "grad_norm": 3.716539441275778, + "learning_rate": 6.025641025641026e-06, + "loss": 1.0971, + "step": 94 + }, + { + "epoch": 0.009138130050019238, + "grad_norm": 4.272502728523228, + "learning_rate": 6.08974358974359e-06, + "loss": 1.0945, + "step": 95 + }, + { + "epoch": 0.00923432089265102, + "grad_norm": 3.634727830198063, + "learning_rate": 6.153846153846155e-06, + "loss": 1.0518, + "step": 96 + }, + { + "epoch": 0.009330511735282801, + "grad_norm": 4.246675137565664, + "learning_rate": 6.217948717948718e-06, + "loss": 1.1727, + "step": 97 + }, + { + "epoch": 0.009426702577914583, + "grad_norm": 4.12620836219417, + "learning_rate": 6.282051282051282e-06, + "loss": 1.0679, + "step": 98 + }, + { + "epoch": 0.009522893420546364, + "grad_norm": 4.128956690502437, + "learning_rate": 6.3461538461538466e-06, + "loss": 1.0245, + "step": 99 + }, + { + "epoch": 0.009619084263178146, + "grad_norm": 4.555285417611483, + "learning_rate": 6.410256410256412e-06, + "loss": 1.0512, + "step": 100 + }, + { + "epoch": 0.009715275105809927, + "grad_norm": 4.362640276195904, + "learning_rate": 6.474358974358975e-06, + "loss": 1.0604, + "step": 101 + }, + { + "epoch": 0.009811465948441709, + "grad_norm": 4.067047785120322, + "learning_rate": 6.538461538461539e-06, + "loss": 1.1247, + "step": 102 + }, + { + "epoch": 0.00990765679107349, + "grad_norm": 4.275953431091554, + "learning_rate": 6.602564102564103e-06, + "loss": 1.1214, + "step": 103 + }, + { + "epoch": 0.010003847633705272, + "grad_norm": 5.044851123843794, + "learning_rate": 6.666666666666667e-06, + "loss": 1.0362, + "step": 104 + }, + { + "epoch": 0.010100038476337054, + "grad_norm": 4.881989889972896, + "learning_rate": 6.730769230769232e-06, + "loss": 1.1309, + "step": 105 + }, + { + "epoch": 0.010196229318968833, + "grad_norm": 4.089131412616388, + "learning_rate": 6.794871794871796e-06, + "loss": 1.0617, + "step": 106 + }, + { + "epoch": 0.010292420161600615, + "grad_norm": 4.01106535132138, + "learning_rate": 6.858974358974359e-06, + "loss": 1.0733, + "step": 107 + }, + { + "epoch": 0.010388611004232396, + "grad_norm": 5.01197063313018, + "learning_rate": 6.923076923076923e-06, + "loss": 1.0674, + "step": 108 + }, + { + "epoch": 0.010484801846864178, + "grad_norm": 4.4090611850546235, + "learning_rate": 6.9871794871794876e-06, + "loss": 1.1629, + "step": 109 + }, + { + "epoch": 0.01058099268949596, + "grad_norm": 3.846927492951968, + "learning_rate": 7.051282051282053e-06, + "loss": 1.0756, + "step": 110 + }, + { + "epoch": 0.010677183532127741, + "grad_norm": 3.7039729268696546, + "learning_rate": 7.115384615384616e-06, + "loss": 1.0737, + "step": 111 + }, + { + "epoch": 0.010773374374759523, + "grad_norm": 4.565971233334966, + "learning_rate": 7.17948717948718e-06, + "loss": 0.9554, + "step": 112 + }, + { + "epoch": 0.010869565217391304, + "grad_norm": 4.5513060499447215, + "learning_rate": 7.243589743589744e-06, + "loss": 1.117, + "step": 113 + }, + { + "epoch": 0.010965756060023086, + "grad_norm": 4.116367056167208, + "learning_rate": 7.307692307692308e-06, + "loss": 1.208, + "step": 114 + }, + { + "epoch": 0.011061946902654867, + "grad_norm": 4.164482088220883, + "learning_rate": 7.371794871794873e-06, + "loss": 1.0807, + "step": 115 + }, + { + "epoch": 0.011158137745286649, + "grad_norm": 4.259161282348704, + "learning_rate": 7.435897435897437e-06, + "loss": 1.1026, + "step": 116 + }, + { + "epoch": 0.01125432858791843, + "grad_norm": 4.450182609851108, + "learning_rate": 7.500000000000001e-06, + "loss": 0.996, + "step": 117 + }, + { + "epoch": 0.011350519430550212, + "grad_norm": 3.9756196698426987, + "learning_rate": 7.564102564102564e-06, + "loss": 1.1022, + "step": 118 + }, + { + "epoch": 0.011446710273181993, + "grad_norm": 4.484408503463425, + "learning_rate": 7.6282051282051286e-06, + "loss": 1.0395, + "step": 119 + }, + { + "epoch": 0.011542901115813775, + "grad_norm": 4.581984920904069, + "learning_rate": 7.692307692307694e-06, + "loss": 1.0297, + "step": 120 + }, + { + "epoch": 0.011639091958445556, + "grad_norm": 2.7125916343782146, + "learning_rate": 7.756410256410258e-06, + "loss": 0.768, + "step": 121 + }, + { + "epoch": 0.011735282801077338, + "grad_norm": 2.3392950785995548, + "learning_rate": 7.820512820512822e-06, + "loss": 0.699, + "step": 122 + }, + { + "epoch": 0.01183147364370912, + "grad_norm": 4.608898277982076, + "learning_rate": 7.884615384615384e-06, + "loss": 1.1524, + "step": 123 + }, + { + "epoch": 0.0119276644863409, + "grad_norm": 3.69650267063383, + "learning_rate": 7.948717948717949e-06, + "loss": 1.0705, + "step": 124 + }, + { + "epoch": 0.012023855328972682, + "grad_norm": 4.732360070850073, + "learning_rate": 8.012820512820515e-06, + "loss": 1.0885, + "step": 125 + }, + { + "epoch": 0.012120046171604464, + "grad_norm": 4.276102015492886, + "learning_rate": 8.076923076923077e-06, + "loss": 1.2161, + "step": 126 + }, + { + "epoch": 0.012216237014236245, + "grad_norm": 3.765268946310117, + "learning_rate": 8.141025641025641e-06, + "loss": 1.0242, + "step": 127 + }, + { + "epoch": 0.012312427856868027, + "grad_norm": 3.8539577363973025, + "learning_rate": 8.205128205128205e-06, + "loss": 1.0794, + "step": 128 + }, + { + "epoch": 0.012408618699499807, + "grad_norm": 4.272391962490061, + "learning_rate": 8.26923076923077e-06, + "loss": 1.1562, + "step": 129 + }, + { + "epoch": 0.012504809542131588, + "grad_norm": 4.283344496271799, + "learning_rate": 8.333333333333334e-06, + "loss": 1.1757, + "step": 130 + }, + { + "epoch": 0.01260100038476337, + "grad_norm": 4.481605805426497, + "learning_rate": 8.397435897435898e-06, + "loss": 1.1615, + "step": 131 + }, + { + "epoch": 0.012697191227395151, + "grad_norm": 4.248410574944953, + "learning_rate": 8.461538461538462e-06, + "loss": 1.0139, + "step": 132 + }, + { + "epoch": 0.012793382070026933, + "grad_norm": 4.182670952528412, + "learning_rate": 8.525641025641026e-06, + "loss": 1.0841, + "step": 133 + }, + { + "epoch": 0.012889572912658714, + "grad_norm": 3.487699572173298, + "learning_rate": 8.58974358974359e-06, + "loss": 1.0704, + "step": 134 + }, + { + "epoch": 0.012985763755290496, + "grad_norm": 4.255222063585537, + "learning_rate": 8.653846153846155e-06, + "loss": 1.1001, + "step": 135 + }, + { + "epoch": 0.013081954597922277, + "grad_norm": 4.411041504767902, + "learning_rate": 8.717948717948719e-06, + "loss": 0.9594, + "step": 136 + }, + { + "epoch": 0.013178145440554059, + "grad_norm": 4.189532393873252, + "learning_rate": 8.782051282051283e-06, + "loss": 1.1804, + "step": 137 + }, + { + "epoch": 0.01327433628318584, + "grad_norm": 4.477709583390508, + "learning_rate": 8.846153846153847e-06, + "loss": 1.1134, + "step": 138 + }, + { + "epoch": 0.013370527125817622, + "grad_norm": 3.884345432292478, + "learning_rate": 8.910256410256411e-06, + "loss": 1.055, + "step": 139 + }, + { + "epoch": 0.013466717968449404, + "grad_norm": 4.135521959397555, + "learning_rate": 8.974358974358976e-06, + "loss": 0.9748, + "step": 140 + }, + { + "epoch": 0.013562908811081185, + "grad_norm": 4.093335382002577, + "learning_rate": 9.03846153846154e-06, + "loss": 1.096, + "step": 141 + }, + { + "epoch": 0.013659099653712967, + "grad_norm": 4.179651129096266, + "learning_rate": 9.102564102564104e-06, + "loss": 1.1242, + "step": 142 + }, + { + "epoch": 0.013755290496344748, + "grad_norm": 4.206907653878133, + "learning_rate": 9.166666666666666e-06, + "loss": 1.1062, + "step": 143 + }, + { + "epoch": 0.01385148133897653, + "grad_norm": 3.7257669850017585, + "learning_rate": 9.230769230769232e-06, + "loss": 1.0448, + "step": 144 + }, + { + "epoch": 0.013947672181608311, + "grad_norm": 4.337118745154926, + "learning_rate": 9.294871794871796e-06, + "loss": 1.1279, + "step": 145 + }, + { + "epoch": 0.014043863024240093, + "grad_norm": 3.872976644150853, + "learning_rate": 9.358974358974359e-06, + "loss": 1.0676, + "step": 146 + }, + { + "epoch": 0.014140053866871874, + "grad_norm": 4.47121182074263, + "learning_rate": 9.423076923076923e-06, + "loss": 1.1646, + "step": 147 + }, + { + "epoch": 0.014236244709503656, + "grad_norm": 3.503824645877157, + "learning_rate": 9.487179487179487e-06, + "loss": 1.0757, + "step": 148 + }, + { + "epoch": 0.014332435552135437, + "grad_norm": 4.383317957879249, + "learning_rate": 9.551282051282053e-06, + "loss": 1.115, + "step": 149 + }, + { + "epoch": 0.014428626394767219, + "grad_norm": 4.0808049971787455, + "learning_rate": 9.615384615384616e-06, + "loss": 1.0367, + "step": 150 + }, + { + "epoch": 0.014524817237399, + "grad_norm": 3.2148521200522464, + "learning_rate": 9.67948717948718e-06, + "loss": 1.1344, + "step": 151 + }, + { + "epoch": 0.014621008080030782, + "grad_norm": 4.772326275049353, + "learning_rate": 9.743589743589744e-06, + "loss": 1.0607, + "step": 152 + }, + { + "epoch": 0.014717198922662562, + "grad_norm": 3.909262362145375, + "learning_rate": 9.807692307692308e-06, + "loss": 1.0771, + "step": 153 + }, + { + "epoch": 0.014813389765294343, + "grad_norm": 4.223896761882109, + "learning_rate": 9.871794871794872e-06, + "loss": 1.083, + "step": 154 + }, + { + "epoch": 0.014909580607926125, + "grad_norm": 4.699590799741427, + "learning_rate": 9.935897435897437e-06, + "loss": 1.1197, + "step": 155 + }, + { + "epoch": 0.015005771450557906, + "grad_norm": 2.3074028728740936, + "learning_rate": 1e-05, + "loss": 0.747, + "step": 156 + }, + { + "epoch": 0.015101962293189688, + "grad_norm": 4.033823917089736, + "learning_rate": 1.0064102564102565e-05, + "loss": 1.0552, + "step": 157 + }, + { + "epoch": 0.01519815313582147, + "grad_norm": 4.652600463142672, + "learning_rate": 1.012820512820513e-05, + "loss": 1.1075, + "step": 158 + }, + { + "epoch": 0.015294343978453251, + "grad_norm": 3.632231880196277, + "learning_rate": 1.0192307692307692e-05, + "loss": 1.0913, + "step": 159 + }, + { + "epoch": 0.015390534821085032, + "grad_norm": 4.022417551505229, + "learning_rate": 1.0256410256410256e-05, + "loss": 1.0631, + "step": 160 + }, + { + "epoch": 0.015486725663716814, + "grad_norm": 4.005775930052858, + "learning_rate": 1.0320512820512822e-05, + "loss": 1.0269, + "step": 161 + }, + { + "epoch": 0.015582916506348595, + "grad_norm": 4.1045421075737885, + "learning_rate": 1.0384615384615386e-05, + "loss": 1.2105, + "step": 162 + }, + { + "epoch": 0.015679107348980377, + "grad_norm": 4.147816634505284, + "learning_rate": 1.044871794871795e-05, + "loss": 1.0727, + "step": 163 + }, + { + "epoch": 0.015775298191612157, + "grad_norm": 2.364180379253167, + "learning_rate": 1.0512820512820514e-05, + "loss": 0.7049, + "step": 164 + }, + { + "epoch": 0.01587148903424394, + "grad_norm": 4.572263372147381, + "learning_rate": 1.0576923076923078e-05, + "loss": 1.1445, + "step": 165 + }, + { + "epoch": 0.01596767987687572, + "grad_norm": 3.9899316910749167, + "learning_rate": 1.0641025641025643e-05, + "loss": 1.0756, + "step": 166 + }, + { + "epoch": 0.016063870719507503, + "grad_norm": 3.9540042547910756, + "learning_rate": 1.0705128205128205e-05, + "loss": 1.1292, + "step": 167 + }, + { + "epoch": 0.016160061562139283, + "grad_norm": 4.07604946456486, + "learning_rate": 1.076923076923077e-05, + "loss": 1.177, + "step": 168 + }, + { + "epoch": 0.016256252404771066, + "grad_norm": 3.672758607360489, + "learning_rate": 1.0833333333333334e-05, + "loss": 1.023, + "step": 169 + }, + { + "epoch": 0.016352443247402846, + "grad_norm": 3.9159305470375814, + "learning_rate": 1.0897435897435898e-05, + "loss": 1.142, + "step": 170 + }, + { + "epoch": 0.01644863409003463, + "grad_norm": 3.437148573074714, + "learning_rate": 1.0961538461538464e-05, + "loss": 1.0394, + "step": 171 + }, + { + "epoch": 0.01654482493266641, + "grad_norm": 3.7734853892731275, + "learning_rate": 1.1025641025641028e-05, + "loss": 1.0419, + "step": 172 + }, + { + "epoch": 0.016641015775298192, + "grad_norm": 3.5997634653856694, + "learning_rate": 1.1089743589743592e-05, + "loss": 1.0427, + "step": 173 + }, + { + "epoch": 0.016737206617929972, + "grad_norm": 3.4084313807726527, + "learning_rate": 1.1153846153846154e-05, + "loss": 0.9996, + "step": 174 + }, + { + "epoch": 0.016833397460561755, + "grad_norm": 3.736680101314529, + "learning_rate": 1.1217948717948719e-05, + "loss": 1.1714, + "step": 175 + }, + { + "epoch": 0.016929588303193535, + "grad_norm": 3.824357931364307, + "learning_rate": 1.1282051282051283e-05, + "loss": 1.142, + "step": 176 + }, + { + "epoch": 0.01702577914582532, + "grad_norm": 3.917043916475778, + "learning_rate": 1.1346153846153847e-05, + "loss": 1.1174, + "step": 177 + }, + { + "epoch": 0.017121969988457098, + "grad_norm": 4.089633134410509, + "learning_rate": 1.1410256410256411e-05, + "loss": 1.0247, + "step": 178 + }, + { + "epoch": 0.01721816083108888, + "grad_norm": 3.4266077478488053, + "learning_rate": 1.1474358974358974e-05, + "loss": 1.0688, + "step": 179 + }, + { + "epoch": 0.01731435167372066, + "grad_norm": 3.5526835869834033, + "learning_rate": 1.1538461538461538e-05, + "loss": 1.0875, + "step": 180 + }, + { + "epoch": 0.017410542516352445, + "grad_norm": 3.913953931799239, + "learning_rate": 1.1602564102564104e-05, + "loss": 1.1398, + "step": 181 + }, + { + "epoch": 0.017506733358984224, + "grad_norm": 3.705953776761441, + "learning_rate": 1.1666666666666668e-05, + "loss": 1.1686, + "step": 182 + }, + { + "epoch": 0.017602924201616008, + "grad_norm": 3.6566590103200904, + "learning_rate": 1.1730769230769232e-05, + "loss": 1.1232, + "step": 183 + }, + { + "epoch": 0.017699115044247787, + "grad_norm": 3.5250945411067445, + "learning_rate": 1.1794871794871796e-05, + "loss": 1.0819, + "step": 184 + }, + { + "epoch": 0.01779530588687957, + "grad_norm": 3.562053018148263, + "learning_rate": 1.185897435897436e-05, + "loss": 1.0744, + "step": 185 + }, + { + "epoch": 0.01789149672951135, + "grad_norm": 4.091926149124595, + "learning_rate": 1.1923076923076925e-05, + "loss": 1.1033, + "step": 186 + }, + { + "epoch": 0.017987687572143134, + "grad_norm": 3.584725918928567, + "learning_rate": 1.1987179487179487e-05, + "loss": 1.1384, + "step": 187 + }, + { + "epoch": 0.018083878414774913, + "grad_norm": 3.596221350229456, + "learning_rate": 1.2051282051282051e-05, + "loss": 1.1042, + "step": 188 + }, + { + "epoch": 0.018180069257406693, + "grad_norm": 3.392521920076525, + "learning_rate": 1.2115384615384615e-05, + "loss": 1.1363, + "step": 189 + }, + { + "epoch": 0.018276260100038477, + "grad_norm": 3.532132461032909, + "learning_rate": 1.217948717948718e-05, + "loss": 1.0739, + "step": 190 + }, + { + "epoch": 0.018372450942670256, + "grad_norm": 3.2494698569956606, + "learning_rate": 1.2243589743589746e-05, + "loss": 1.0763, + "step": 191 + }, + { + "epoch": 0.01846864178530204, + "grad_norm": 3.596694483193632, + "learning_rate": 1.230769230769231e-05, + "loss": 1.0909, + "step": 192 + }, + { + "epoch": 0.01856483262793382, + "grad_norm": 3.8843816697331137, + "learning_rate": 1.2371794871794874e-05, + "loss": 1.1496, + "step": 193 + }, + { + "epoch": 0.018661023470565603, + "grad_norm": 2.7591617070865224, + "learning_rate": 1.2435897435897436e-05, + "loss": 0.7908, + "step": 194 + }, + { + "epoch": 0.018757214313197382, + "grad_norm": 4.56810191638542, + "learning_rate": 1.25e-05, + "loss": 1.1048, + "step": 195 + }, + { + "epoch": 0.018853405155829166, + "grad_norm": 3.653477868354496, + "learning_rate": 1.2564102564102565e-05, + "loss": 1.0468, + "step": 196 + }, + { + "epoch": 0.018949595998460946, + "grad_norm": 4.278608867254806, + "learning_rate": 1.2628205128205129e-05, + "loss": 1.0295, + "step": 197 + }, + { + "epoch": 0.01904578684109273, + "grad_norm": 3.7149834659673364, + "learning_rate": 1.2692307692307693e-05, + "loss": 1.0181, + "step": 198 + }, + { + "epoch": 0.01914197768372451, + "grad_norm": 3.9599758606928224, + "learning_rate": 1.2756410256410257e-05, + "loss": 1.0888, + "step": 199 + }, + { + "epoch": 0.019238168526356292, + "grad_norm": 3.6729190891695658, + "learning_rate": 1.2820512820512823e-05, + "loss": 0.9898, + "step": 200 + }, + { + "epoch": 0.01933435936898807, + "grad_norm": 3.425600218804166, + "learning_rate": 1.2884615384615386e-05, + "loss": 1.1343, + "step": 201 + }, + { + "epoch": 0.019430550211619855, + "grad_norm": 3.646759646107961, + "learning_rate": 1.294871794871795e-05, + "loss": 1.1364, + "step": 202 + }, + { + "epoch": 0.019526741054251635, + "grad_norm": 3.4550506462192496, + "learning_rate": 1.3012820512820514e-05, + "loss": 1.0989, + "step": 203 + }, + { + "epoch": 0.019622931896883418, + "grad_norm": 2.161328706901354, + "learning_rate": 1.3076923076923078e-05, + "loss": 0.7577, + "step": 204 + }, + { + "epoch": 0.019719122739515198, + "grad_norm": 3.6246056987594573, + "learning_rate": 1.3141025641025642e-05, + "loss": 1.1493, + "step": 205 + }, + { + "epoch": 0.01981531358214698, + "grad_norm": 4.521182085727683, + "learning_rate": 1.3205128205128207e-05, + "loss": 1.1793, + "step": 206 + }, + { + "epoch": 0.01991150442477876, + "grad_norm": 3.9808818783555364, + "learning_rate": 1.3269230769230769e-05, + "loss": 1.0018, + "step": 207 + }, + { + "epoch": 0.020007695267410544, + "grad_norm": 3.528886305695303, + "learning_rate": 1.3333333333333333e-05, + "loss": 1.0367, + "step": 208 + }, + { + "epoch": 0.020103886110042324, + "grad_norm": 4.188121873679507, + "learning_rate": 1.3397435897435897e-05, + "loss": 1.101, + "step": 209 + }, + { + "epoch": 0.020200076952674107, + "grad_norm": 3.4288722568408407, + "learning_rate": 1.3461538461538463e-05, + "loss": 1.1611, + "step": 210 + }, + { + "epoch": 0.020296267795305887, + "grad_norm": 3.671750054929106, + "learning_rate": 1.3525641025641028e-05, + "loss": 1.097, + "step": 211 + }, + { + "epoch": 0.020392458637937667, + "grad_norm": 3.9139397471560176, + "learning_rate": 1.3589743589743592e-05, + "loss": 1.1417, + "step": 212 + }, + { + "epoch": 0.02048864948056945, + "grad_norm": 4.068266689598962, + "learning_rate": 1.3653846153846156e-05, + "loss": 1.1404, + "step": 213 + }, + { + "epoch": 0.02058484032320123, + "grad_norm": 3.5608296811997233, + "learning_rate": 1.3717948717948718e-05, + "loss": 1.1843, + "step": 214 + }, + { + "epoch": 0.020681031165833013, + "grad_norm": 3.762790037558702, + "learning_rate": 1.3782051282051283e-05, + "loss": 1.1559, + "step": 215 + }, + { + "epoch": 0.020777222008464793, + "grad_norm": 3.602593751216178, + "learning_rate": 1.3846153846153847e-05, + "loss": 1.1762, + "step": 216 + }, + { + "epoch": 0.020873412851096576, + "grad_norm": 4.9777154062897, + "learning_rate": 1.3910256410256411e-05, + "loss": 1.0401, + "step": 217 + }, + { + "epoch": 0.020969603693728356, + "grad_norm": 3.8592184882197587, + "learning_rate": 1.3974358974358975e-05, + "loss": 1.0881, + "step": 218 + }, + { + "epoch": 0.02106579453636014, + "grad_norm": 3.62364227637628, + "learning_rate": 1.403846153846154e-05, + "loss": 1.0886, + "step": 219 + }, + { + "epoch": 0.02116198537899192, + "grad_norm": 4.25628070324627, + "learning_rate": 1.4102564102564105e-05, + "loss": 1.0322, + "step": 220 + }, + { + "epoch": 0.021258176221623702, + "grad_norm": 4.2104925547049605, + "learning_rate": 1.416666666666667e-05, + "loss": 1.0183, + "step": 221 + }, + { + "epoch": 0.021354367064255482, + "grad_norm": 3.5093990767069796, + "learning_rate": 1.4230769230769232e-05, + "loss": 1.1409, + "step": 222 + }, + { + "epoch": 0.021450557906887265, + "grad_norm": 4.064890621146463, + "learning_rate": 1.4294871794871796e-05, + "loss": 1.0327, + "step": 223 + }, + { + "epoch": 0.021546748749519045, + "grad_norm": 3.428812405318278, + "learning_rate": 1.435897435897436e-05, + "loss": 1.0753, + "step": 224 + }, + { + "epoch": 0.02164293959215083, + "grad_norm": 4.028302552683022, + "learning_rate": 1.4423076923076924e-05, + "loss": 1.1448, + "step": 225 + }, + { + "epoch": 0.021739130434782608, + "grad_norm": 3.9733025781938083, + "learning_rate": 1.4487179487179489e-05, + "loss": 1.0289, + "step": 226 + }, + { + "epoch": 0.02183532127741439, + "grad_norm": 3.9283023123207315, + "learning_rate": 1.4551282051282051e-05, + "loss": 1.1066, + "step": 227 + }, + { + "epoch": 0.02193151212004617, + "grad_norm": 4.154430027682187, + "learning_rate": 1.4615384615384615e-05, + "loss": 0.9997, + "step": 228 + }, + { + "epoch": 0.022027702962677954, + "grad_norm": 4.32579244003894, + "learning_rate": 1.467948717948718e-05, + "loss": 1.0563, + "step": 229 + }, + { + "epoch": 0.022123893805309734, + "grad_norm": 4.439389852323162, + "learning_rate": 1.4743589743589745e-05, + "loss": 1.0966, + "step": 230 + }, + { + "epoch": 0.022220084647941517, + "grad_norm": 3.6202345907522537, + "learning_rate": 1.480769230769231e-05, + "loss": 1.0308, + "step": 231 + }, + { + "epoch": 0.022316275490573297, + "grad_norm": 2.8284475799696214, + "learning_rate": 1.4871794871794874e-05, + "loss": 0.77, + "step": 232 + }, + { + "epoch": 0.02241246633320508, + "grad_norm": 3.746103077668565, + "learning_rate": 1.4935897435897438e-05, + "loss": 1.0713, + "step": 233 + }, + { + "epoch": 0.02250865717583686, + "grad_norm": 4.108056904926606, + "learning_rate": 1.5000000000000002e-05, + "loss": 1.0922, + "step": 234 + }, + { + "epoch": 0.02260484801846864, + "grad_norm": 3.982888482986168, + "learning_rate": 1.5064102564102565e-05, + "loss": 1.1128, + "step": 235 + }, + { + "epoch": 0.022701038861100423, + "grad_norm": 3.7319776650293814, + "learning_rate": 1.5128205128205129e-05, + "loss": 1.0298, + "step": 236 + }, + { + "epoch": 0.022797229703732203, + "grad_norm": 3.8889026525662636, + "learning_rate": 1.5192307692307693e-05, + "loss": 1.0491, + "step": 237 + }, + { + "epoch": 0.022893420546363986, + "grad_norm": 4.244965147698684, + "learning_rate": 1.5256410256410257e-05, + "loss": 1.0754, + "step": 238 + }, + { + "epoch": 0.022989611388995766, + "grad_norm": 4.1780251790426455, + "learning_rate": 1.5320512820512823e-05, + "loss": 1.084, + "step": 239 + }, + { + "epoch": 0.02308580223162755, + "grad_norm": 3.752349876156193, + "learning_rate": 1.5384615384615387e-05, + "loss": 1.2086, + "step": 240 + }, + { + "epoch": 0.02318199307425933, + "grad_norm": 3.5397474264818, + "learning_rate": 1.544871794871795e-05, + "loss": 1.0037, + "step": 241 + }, + { + "epoch": 0.023278183916891113, + "grad_norm": 3.3298057335318805, + "learning_rate": 1.5512820512820516e-05, + "loss": 1.1693, + "step": 242 + }, + { + "epoch": 0.023374374759522892, + "grad_norm": 3.6071421571017703, + "learning_rate": 1.557692307692308e-05, + "loss": 1.112, + "step": 243 + }, + { + "epoch": 0.023470565602154676, + "grad_norm": 3.898374974784951, + "learning_rate": 1.5641025641025644e-05, + "loss": 1.1559, + "step": 244 + }, + { + "epoch": 0.023566756444786455, + "grad_norm": 3.698255740621529, + "learning_rate": 1.5705128205128205e-05, + "loss": 1.1832, + "step": 245 + }, + { + "epoch": 0.02366294728741824, + "grad_norm": 3.632567474236792, + "learning_rate": 1.576923076923077e-05, + "loss": 1.2394, + "step": 246 + }, + { + "epoch": 0.02375913813005002, + "grad_norm": 3.7525270715010395, + "learning_rate": 1.5833333333333333e-05, + "loss": 1.1683, + "step": 247 + }, + { + "epoch": 0.0238553289726818, + "grad_norm": 3.3187811098352737, + "learning_rate": 1.5897435897435897e-05, + "loss": 1.0901, + "step": 248 + }, + { + "epoch": 0.02395151981531358, + "grad_norm": 3.0331037590857317, + "learning_rate": 1.5961538461538465e-05, + "loss": 1.1156, + "step": 249 + }, + { + "epoch": 0.024047710657945365, + "grad_norm": 3.996058185643904, + "learning_rate": 1.602564102564103e-05, + "loss": 1.2007, + "step": 250 + }, + { + "epoch": 0.024143901500577145, + "grad_norm": 4.479217369856826, + "learning_rate": 1.6089743589743593e-05, + "loss": 1.1264, + "step": 251 + }, + { + "epoch": 0.024240092343208928, + "grad_norm": 4.1097801546354695, + "learning_rate": 1.6153846153846154e-05, + "loss": 0.9967, + "step": 252 + }, + { + "epoch": 0.024336283185840708, + "grad_norm": 3.46703982236511, + "learning_rate": 1.6217948717948718e-05, + "loss": 1.1303, + "step": 253 + }, + { + "epoch": 0.02443247402847249, + "grad_norm": 3.3864363803933744, + "learning_rate": 1.6282051282051282e-05, + "loss": 1.1418, + "step": 254 + }, + { + "epoch": 0.02452866487110427, + "grad_norm": 3.375755476891812, + "learning_rate": 1.6346153846153847e-05, + "loss": 1.0998, + "step": 255 + }, + { + "epoch": 0.024624855713736054, + "grad_norm": 3.9370247236886873, + "learning_rate": 1.641025641025641e-05, + "loss": 1.1952, + "step": 256 + }, + { + "epoch": 0.024721046556367834, + "grad_norm": 3.485785199032767, + "learning_rate": 1.6474358974358975e-05, + "loss": 1.1077, + "step": 257 + }, + { + "epoch": 0.024817237398999614, + "grad_norm": 3.6336374906178635, + "learning_rate": 1.653846153846154e-05, + "loss": 1.1108, + "step": 258 + }, + { + "epoch": 0.024913428241631397, + "grad_norm": 3.3859108358658476, + "learning_rate": 1.6602564102564103e-05, + "loss": 1.162, + "step": 259 + }, + { + "epoch": 0.025009619084263177, + "grad_norm": 3.780966552967039, + "learning_rate": 1.6666666666666667e-05, + "loss": 1.1797, + "step": 260 + }, + { + "epoch": 0.02510580992689496, + "grad_norm": 3.787551718325095, + "learning_rate": 1.673076923076923e-05, + "loss": 1.0978, + "step": 261 + }, + { + "epoch": 0.02520200076952674, + "grad_norm": 4.255909347265725, + "learning_rate": 1.6794871794871796e-05, + "loss": 1.03, + "step": 262 + }, + { + "epoch": 0.025298191612158523, + "grad_norm": 3.6455681623929324, + "learning_rate": 1.685897435897436e-05, + "loss": 1.0996, + "step": 263 + }, + { + "epoch": 0.025394382454790303, + "grad_norm": 4.346681151613543, + "learning_rate": 1.6923076923076924e-05, + "loss": 1.2734, + "step": 264 + }, + { + "epoch": 0.025490573297422086, + "grad_norm": 3.427805158024887, + "learning_rate": 1.698717948717949e-05, + "loss": 1.1385, + "step": 265 + }, + { + "epoch": 0.025586764140053866, + "grad_norm": 3.676966982490396, + "learning_rate": 1.7051282051282053e-05, + "loss": 1.1217, + "step": 266 + }, + { + "epoch": 0.02568295498268565, + "grad_norm": 3.368788566705791, + "learning_rate": 1.7115384615384617e-05, + "loss": 1.1388, + "step": 267 + }, + { + "epoch": 0.02577914582531743, + "grad_norm": 2.5245421077145487, + "learning_rate": 1.717948717948718e-05, + "loss": 0.8648, + "step": 268 + }, + { + "epoch": 0.025875336667949212, + "grad_norm": 3.799348336262582, + "learning_rate": 1.7243589743589745e-05, + "loss": 1.1368, + "step": 269 + }, + { + "epoch": 0.025971527510580992, + "grad_norm": 3.463574404081878, + "learning_rate": 1.730769230769231e-05, + "loss": 1.0765, + "step": 270 + }, + { + "epoch": 0.026067718353212775, + "grad_norm": 3.1487228165295966, + "learning_rate": 1.7371794871794873e-05, + "loss": 1.1033, + "step": 271 + }, + { + "epoch": 0.026163909195844555, + "grad_norm": 3.3436080806887585, + "learning_rate": 1.7435897435897438e-05, + "loss": 1.0236, + "step": 272 + }, + { + "epoch": 0.026260100038476338, + "grad_norm": 3.805013075363612, + "learning_rate": 1.7500000000000002e-05, + "loss": 1.1807, + "step": 273 + }, + { + "epoch": 0.026356290881108118, + "grad_norm": 3.5079198085643264, + "learning_rate": 1.7564102564102566e-05, + "loss": 1.208, + "step": 274 + }, + { + "epoch": 0.0264524817237399, + "grad_norm": 3.464319609355637, + "learning_rate": 1.762820512820513e-05, + "loss": 1.1546, + "step": 275 + }, + { + "epoch": 0.02654867256637168, + "grad_norm": 3.8435784769276027, + "learning_rate": 1.7692307692307694e-05, + "loss": 1.2144, + "step": 276 + }, + { + "epoch": 0.026644863409003464, + "grad_norm": 3.630790638473205, + "learning_rate": 1.775641025641026e-05, + "loss": 1.026, + "step": 277 + }, + { + "epoch": 0.026741054251635244, + "grad_norm": 3.330255121453628, + "learning_rate": 1.7820512820512823e-05, + "loss": 1.1194, + "step": 278 + }, + { + "epoch": 0.026837245094267027, + "grad_norm": 3.5218613073924563, + "learning_rate": 1.7884615384615387e-05, + "loss": 1.0425, + "step": 279 + }, + { + "epoch": 0.026933435936898807, + "grad_norm": 3.4251897711423775, + "learning_rate": 1.794871794871795e-05, + "loss": 1.1352, + "step": 280 + }, + { + "epoch": 0.027029626779530587, + "grad_norm": 3.1021170451409894, + "learning_rate": 1.8012820512820515e-05, + "loss": 1.0205, + "step": 281 + }, + { + "epoch": 0.02712581762216237, + "grad_norm": 3.7939770009115423, + "learning_rate": 1.807692307692308e-05, + "loss": 1.1547, + "step": 282 + }, + { + "epoch": 0.02722200846479415, + "grad_norm": 3.982003184372175, + "learning_rate": 1.8141025641025644e-05, + "loss": 1.1285, + "step": 283 + }, + { + "epoch": 0.027318199307425933, + "grad_norm": 3.5597129486845698, + "learning_rate": 1.8205128205128208e-05, + "loss": 1.1584, + "step": 284 + }, + { + "epoch": 0.027414390150057713, + "grad_norm": 3.6117399072349397, + "learning_rate": 1.826923076923077e-05, + "loss": 1.1079, + "step": 285 + }, + { + "epoch": 0.027510580992689496, + "grad_norm": 3.847865910480509, + "learning_rate": 1.8333333333333333e-05, + "loss": 1.1822, + "step": 286 + }, + { + "epoch": 0.027606771835321276, + "grad_norm": 3.603575872577993, + "learning_rate": 1.8397435897435897e-05, + "loss": 1.1067, + "step": 287 + }, + { + "epoch": 0.02770296267795306, + "grad_norm": 2.4038532016265015, + "learning_rate": 1.8461538461538465e-05, + "loss": 0.7685, + "step": 288 + }, + { + "epoch": 0.02779915352058484, + "grad_norm": 3.3619924493004603, + "learning_rate": 1.852564102564103e-05, + "loss": 1.1037, + "step": 289 + }, + { + "epoch": 0.027895344363216622, + "grad_norm": 3.438305266932585, + "learning_rate": 1.8589743589743593e-05, + "loss": 1.1183, + "step": 290 + }, + { + "epoch": 0.027991535205848402, + "grad_norm": 3.0419799287272613, + "learning_rate": 1.8653846153846157e-05, + "loss": 1.0813, + "step": 291 + }, + { + "epoch": 0.028087726048480186, + "grad_norm": 3.7382355352947383, + "learning_rate": 1.8717948717948718e-05, + "loss": 1.1112, + "step": 292 + }, + { + "epoch": 0.028183916891111965, + "grad_norm": 3.667581294934287, + "learning_rate": 1.8782051282051282e-05, + "loss": 1.2172, + "step": 293 + }, + { + "epoch": 0.02828010773374375, + "grad_norm": 4.175006327149602, + "learning_rate": 1.8846153846153846e-05, + "loss": 1.0242, + "step": 294 + }, + { + "epoch": 0.02837629857637553, + "grad_norm": 3.600312473213805, + "learning_rate": 1.891025641025641e-05, + "loss": 1.2832, + "step": 295 + }, + { + "epoch": 0.02847248941900731, + "grad_norm": 3.513152716541509, + "learning_rate": 1.8974358974358975e-05, + "loss": 1.1644, + "step": 296 + }, + { + "epoch": 0.02856868026163909, + "grad_norm": 3.267024209727086, + "learning_rate": 1.903846153846154e-05, + "loss": 1.0402, + "step": 297 + }, + { + "epoch": 0.028664871104270875, + "grad_norm": 3.426829034449214, + "learning_rate": 1.9102564102564106e-05, + "loss": 1.0972, + "step": 298 + }, + { + "epoch": 0.028761061946902654, + "grad_norm": 3.035817702344706, + "learning_rate": 1.916666666666667e-05, + "loss": 1.0514, + "step": 299 + }, + { + "epoch": 0.028857252789534438, + "grad_norm": 3.331573015916635, + "learning_rate": 1.923076923076923e-05, + "loss": 1.1477, + "step": 300 + }, + { + "epoch": 0.028953443632166218, + "grad_norm": 3.3718985584987164, + "learning_rate": 1.9294871794871796e-05, + "loss": 1.1854, + "step": 301 + }, + { + "epoch": 0.029049634474798, + "grad_norm": 3.1732601628694206, + "learning_rate": 1.935897435897436e-05, + "loss": 1.0278, + "step": 302 + }, + { + "epoch": 0.02914582531742978, + "grad_norm": 2.1548228084315832, + "learning_rate": 1.9423076923076924e-05, + "loss": 0.7648, + "step": 303 + }, + { + "epoch": 0.029242016160061564, + "grad_norm": 3.276668286689841, + "learning_rate": 1.9487179487179488e-05, + "loss": 1.1764, + "step": 304 + }, + { + "epoch": 0.029338207002693344, + "grad_norm": 3.033789626962792, + "learning_rate": 1.9551282051282052e-05, + "loss": 1.1443, + "step": 305 + }, + { + "epoch": 0.029434397845325123, + "grad_norm": 3.036278157415502, + "learning_rate": 1.9615384615384617e-05, + "loss": 1.0949, + "step": 306 + }, + { + "epoch": 0.029530588687956907, + "grad_norm": 3.2170276536086466, + "learning_rate": 1.967948717948718e-05, + "loss": 1.1836, + "step": 307 + }, + { + "epoch": 0.029626779530588686, + "grad_norm": 3.3102968476888903, + "learning_rate": 1.9743589743589745e-05, + "loss": 1.1404, + "step": 308 + }, + { + "epoch": 0.02972297037322047, + "grad_norm": 3.3488605976043955, + "learning_rate": 1.980769230769231e-05, + "loss": 1.1356, + "step": 309 + }, + { + "epoch": 0.02981916121585225, + "grad_norm": 3.53392756463014, + "learning_rate": 1.9871794871794873e-05, + "loss": 1.0454, + "step": 310 + }, + { + "epoch": 0.029915352058484033, + "grad_norm": 3.04431260711794, + "learning_rate": 1.9935897435897437e-05, + "loss": 1.155, + "step": 311 + }, + { + "epoch": 0.030011542901115813, + "grad_norm": 3.3667757613223888, + "learning_rate": 2e-05, + "loss": 1.1753, + "step": 312 + }, + { + "epoch": 0.030107733743747596, + "grad_norm": 3.523432368169292, + "learning_rate": 1.9999999514706952e-05, + "loss": 1.0744, + "step": 313 + }, + { + "epoch": 0.030203924586379376, + "grad_norm": 3.3516075318076974, + "learning_rate": 1.9999998058827844e-05, + "loss": 1.1195, + "step": 314 + }, + { + "epoch": 0.03030011542901116, + "grad_norm": 3.6003114918853676, + "learning_rate": 1.999999563236283e-05, + "loss": 1.1311, + "step": 315 + }, + { + "epoch": 0.03039630627164294, + "grad_norm": 3.7233842002329975, + "learning_rate": 1.9999992235312136e-05, + "loss": 1.1312, + "step": 316 + }, + { + "epoch": 0.030492497114274722, + "grad_norm": 3.533441788505297, + "learning_rate": 1.999998786767609e-05, + "loss": 1.1362, + "step": 317 + }, + { + "epoch": 0.030588687956906502, + "grad_norm": 3.24516365783465, + "learning_rate": 1.9999982529455127e-05, + "loss": 1.1124, + "step": 318 + }, + { + "epoch": 0.030684878799538285, + "grad_norm": 3.437544763725269, + "learning_rate": 1.9999976220649758e-05, + "loss": 1.1755, + "step": 319 + }, + { + "epoch": 0.030781069642170065, + "grad_norm": 3.6547164069625278, + "learning_rate": 1.9999968941260596e-05, + "loss": 0.9803, + "step": 320 + }, + { + "epoch": 0.030877260484801848, + "grad_norm": 3.738510595262846, + "learning_rate": 1.9999960691288344e-05, + "loss": 1.2045, + "step": 321 + }, + { + "epoch": 0.030973451327433628, + "grad_norm": 3.189850524752794, + "learning_rate": 1.9999951470733808e-05, + "loss": 0.9904, + "step": 322 + }, + { + "epoch": 0.03106964217006541, + "grad_norm": 3.0882658763650697, + "learning_rate": 1.9999941279597878e-05, + "loss": 1.1355, + "step": 323 + }, + { + "epoch": 0.03116583301269719, + "grad_norm": 3.2733470550382826, + "learning_rate": 1.9999930117881548e-05, + "loss": 1.1319, + "step": 324 + }, + { + "epoch": 0.031262023855328974, + "grad_norm": 4.100511779423645, + "learning_rate": 1.9999917985585905e-05, + "loss": 1.2194, + "step": 325 + }, + { + "epoch": 0.031358214697960754, + "grad_norm": 3.6089259608806135, + "learning_rate": 1.9999904882712115e-05, + "loss": 1.1797, + "step": 326 + }, + { + "epoch": 0.031454405540592534, + "grad_norm": 3.212761947368369, + "learning_rate": 1.999989080926146e-05, + "loss": 1.0256, + "step": 327 + }, + { + "epoch": 0.031550596383224314, + "grad_norm": 3.6266673005815666, + "learning_rate": 1.99998757652353e-05, + "loss": 1.222, + "step": 328 + }, + { + "epoch": 0.0316467872258561, + "grad_norm": 3.2136934191641284, + "learning_rate": 1.9999859750635095e-05, + "loss": 1.1652, + "step": 329 + }, + { + "epoch": 0.03174297806848788, + "grad_norm": 3.2657315092653665, + "learning_rate": 1.9999842765462403e-05, + "loss": 1.1765, + "step": 330 + }, + { + "epoch": 0.03183916891111966, + "grad_norm": 2.8434468270535795, + "learning_rate": 1.999982480971887e-05, + "loss": 0.9779, + "step": 331 + }, + { + "epoch": 0.03193535975375144, + "grad_norm": 3.290170123104568, + "learning_rate": 1.999980588340624e-05, + "loss": 1.2182, + "step": 332 + }, + { + "epoch": 0.032031550596383226, + "grad_norm": 2.9018995062329007, + "learning_rate": 1.999978598652635e-05, + "loss": 1.2162, + "step": 333 + }, + { + "epoch": 0.032127741439015006, + "grad_norm": 3.441549525812686, + "learning_rate": 1.9999765119081132e-05, + "loss": 1.027, + "step": 334 + }, + { + "epoch": 0.032223932281646786, + "grad_norm": 3.0591080391718437, + "learning_rate": 1.999974328107261e-05, + "loss": 1.0879, + "step": 335 + }, + { + "epoch": 0.032320123124278566, + "grad_norm": 4.069883382122382, + "learning_rate": 1.9999720472502902e-05, + "loss": 1.1761, + "step": 336 + }, + { + "epoch": 0.03241631396691035, + "grad_norm": 3.419451930494688, + "learning_rate": 1.9999696693374225e-05, + "loss": 1.2179, + "step": 337 + }, + { + "epoch": 0.03251250480954213, + "grad_norm": 3.537543256478108, + "learning_rate": 1.9999671943688885e-05, + "loss": 1.2446, + "step": 338 + }, + { + "epoch": 0.03260869565217391, + "grad_norm": 3.0262543053740734, + "learning_rate": 1.9999646223449284e-05, + "loss": 1.1474, + "step": 339 + }, + { + "epoch": 0.03270488649480569, + "grad_norm": 2.807547940467855, + "learning_rate": 1.9999619532657915e-05, + "loss": 0.7486, + "step": 340 + }, + { + "epoch": 0.03280107733743748, + "grad_norm": 3.7078084572052665, + "learning_rate": 1.9999591871317378e-05, + "loss": 1.0994, + "step": 341 + }, + { + "epoch": 0.03289726818006926, + "grad_norm": 3.504806492469408, + "learning_rate": 1.9999563239430352e-05, + "loss": 1.1995, + "step": 342 + }, + { + "epoch": 0.03299345902270104, + "grad_norm": 2.853410723607112, + "learning_rate": 1.9999533636999616e-05, + "loss": 1.0151, + "step": 343 + }, + { + "epoch": 0.03308964986533282, + "grad_norm": 3.0211244430141884, + "learning_rate": 1.9999503064028043e-05, + "loss": 1.0692, + "step": 344 + }, + { + "epoch": 0.033185840707964605, + "grad_norm": 3.0527497654937443, + "learning_rate": 1.99994715205186e-05, + "loss": 1.1449, + "step": 345 + }, + { + "epoch": 0.033282031550596385, + "grad_norm": 3.0454200150379034, + "learning_rate": 1.999943900647435e-05, + "loss": 1.1556, + "step": 346 + }, + { + "epoch": 0.033378222393228164, + "grad_norm": 3.2740011603220625, + "learning_rate": 1.999940552189845e-05, + "loss": 1.2304, + "step": 347 + }, + { + "epoch": 0.033474413235859944, + "grad_norm": 2.972393494793233, + "learning_rate": 1.9999371066794146e-05, + "loss": 0.9811, + "step": 348 + }, + { + "epoch": 0.03357060407849173, + "grad_norm": 3.197882159704637, + "learning_rate": 1.9999335641164787e-05, + "loss": 1.1227, + "step": 349 + }, + { + "epoch": 0.03366679492112351, + "grad_norm": 3.7211019173614495, + "learning_rate": 1.9999299245013805e-05, + "loss": 1.1927, + "step": 350 + }, + { + "epoch": 0.03376298576375529, + "grad_norm": 3.0470041143240594, + "learning_rate": 1.9999261878344737e-05, + "loss": 1.1095, + "step": 351 + }, + { + "epoch": 0.03385917660638707, + "grad_norm": 3.0319056439593783, + "learning_rate": 1.999922354116121e-05, + "loss": 1.0497, + "step": 352 + }, + { + "epoch": 0.03395536744901885, + "grad_norm": 3.0793654886980124, + "learning_rate": 1.9999184233466945e-05, + "loss": 1.172, + "step": 353 + }, + { + "epoch": 0.03405155829165064, + "grad_norm": 3.2400258015856274, + "learning_rate": 1.999914395526575e-05, + "loss": 1.1199, + "step": 354 + }, + { + "epoch": 0.03414774913428242, + "grad_norm": 2.9528327416471347, + "learning_rate": 1.9999102706561547e-05, + "loss": 1.1438, + "step": 355 + }, + { + "epoch": 0.034243939976914196, + "grad_norm": 3.0697220028996117, + "learning_rate": 1.9999060487358333e-05, + "loss": 1.2267, + "step": 356 + }, + { + "epoch": 0.034340130819545976, + "grad_norm": 3.170304158158417, + "learning_rate": 1.9999017297660204e-05, + "loss": 1.0517, + "step": 357 + }, + { + "epoch": 0.03443632166217776, + "grad_norm": 3.229218629692789, + "learning_rate": 1.9998973137471352e-05, + "loss": 1.1605, + "step": 358 + }, + { + "epoch": 0.03453251250480954, + "grad_norm": 3.4577021027277746, + "learning_rate": 1.9998928006796066e-05, + "loss": 1.0733, + "step": 359 + }, + { + "epoch": 0.03462870334744132, + "grad_norm": 3.1732395396028372, + "learning_rate": 1.9998881905638727e-05, + "loss": 1.1792, + "step": 360 + }, + { + "epoch": 0.0347248941900731, + "grad_norm": 3.760086818954158, + "learning_rate": 1.9998834834003804e-05, + "loss": 1.1387, + "step": 361 + }, + { + "epoch": 0.03482108503270489, + "grad_norm": 3.408611293262036, + "learning_rate": 1.9998786791895874e-05, + "loss": 1.0977, + "step": 362 + }, + { + "epoch": 0.03491727587533667, + "grad_norm": 3.549820527246876, + "learning_rate": 1.9998737779319592e-05, + "loss": 1.1824, + "step": 363 + }, + { + "epoch": 0.03501346671796845, + "grad_norm": 3.1791686488733264, + "learning_rate": 1.999868779627972e-05, + "loss": 1.1103, + "step": 364 + }, + { + "epoch": 0.03510965756060023, + "grad_norm": 3.2262028917095558, + "learning_rate": 1.9998636842781107e-05, + "loss": 1.1113, + "step": 365 + }, + { + "epoch": 0.035205848403232015, + "grad_norm": 3.443777373499206, + "learning_rate": 1.9998584918828695e-05, + "loss": 1.0672, + "step": 366 + }, + { + "epoch": 0.035302039245863795, + "grad_norm": 3.69344618672116, + "learning_rate": 1.999853202442753e-05, + "loss": 1.0975, + "step": 367 + }, + { + "epoch": 0.035398230088495575, + "grad_norm": 3.3953204712210376, + "learning_rate": 1.9998478159582747e-05, + "loss": 0.9894, + "step": 368 + }, + { + "epoch": 0.035494420931127355, + "grad_norm": 3.05237018737838, + "learning_rate": 1.999842332429957e-05, + "loss": 1.1388, + "step": 369 + }, + { + "epoch": 0.03559061177375914, + "grad_norm": 4.20434598608597, + "learning_rate": 1.999836751858332e-05, + "loss": 0.9997, + "step": 370 + }, + { + "epoch": 0.03568680261639092, + "grad_norm": 3.370459182161457, + "learning_rate": 1.9998310742439417e-05, + "loss": 1.2681, + "step": 371 + }, + { + "epoch": 0.0357829934590227, + "grad_norm": 2.9982084814352423, + "learning_rate": 1.9998252995873367e-05, + "loss": 1.1364, + "step": 372 + }, + { + "epoch": 0.03587918430165448, + "grad_norm": 2.802203698315694, + "learning_rate": 1.999819427889078e-05, + "loss": 1.1262, + "step": 373 + }, + { + "epoch": 0.03597537514428627, + "grad_norm": 3.121182660599843, + "learning_rate": 1.999813459149735e-05, + "loss": 1.0404, + "step": 374 + }, + { + "epoch": 0.03607156598691805, + "grad_norm": 3.3247987004614434, + "learning_rate": 1.9998073933698875e-05, + "loss": 1.1698, + "step": 375 + }, + { + "epoch": 0.03616775682954983, + "grad_norm": 3.1720976442151088, + "learning_rate": 1.9998012305501243e-05, + "loss": 1.2276, + "step": 376 + }, + { + "epoch": 0.03626394767218161, + "grad_norm": 3.453842328883157, + "learning_rate": 1.999794970691043e-05, + "loss": 1.2002, + "step": 377 + }, + { + "epoch": 0.03636013851481339, + "grad_norm": 3.9255319134010076, + "learning_rate": 1.999788613793251e-05, + "loss": 1.0455, + "step": 378 + }, + { + "epoch": 0.03645632935744517, + "grad_norm": 3.234854530274309, + "learning_rate": 1.9997821598573665e-05, + "loss": 1.1394, + "step": 379 + }, + { + "epoch": 0.03655252020007695, + "grad_norm": 3.6905714284439473, + "learning_rate": 1.999775608884015e-05, + "loss": 1.1825, + "step": 380 + }, + { + "epoch": 0.03664871104270873, + "grad_norm": 3.450812974415194, + "learning_rate": 1.9997689608738323e-05, + "loss": 1.1955, + "step": 381 + }, + { + "epoch": 0.03674490188534051, + "grad_norm": 2.9576699119075864, + "learning_rate": 1.9997622158274635e-05, + "loss": 1.058, + "step": 382 + }, + { + "epoch": 0.0368410927279723, + "grad_norm": 3.3963022689306004, + "learning_rate": 1.999755373745564e-05, + "loss": 1.049, + "step": 383 + }, + { + "epoch": 0.03693728357060408, + "grad_norm": 3.1805559454329027, + "learning_rate": 1.9997484346287973e-05, + "loss": 1.036, + "step": 384 + }, + { + "epoch": 0.03703347441323586, + "grad_norm": 3.411891051478741, + "learning_rate": 1.999741398477837e-05, + "loss": 1.073, + "step": 385 + }, + { + "epoch": 0.03712966525586764, + "grad_norm": 3.447509115060823, + "learning_rate": 1.9997342652933668e-05, + "loss": 1.2202, + "step": 386 + }, + { + "epoch": 0.037225856098499426, + "grad_norm": 3.0625311054360242, + "learning_rate": 1.999727035076078e-05, + "loss": 1.0651, + "step": 387 + }, + { + "epoch": 0.037322046941131205, + "grad_norm": 3.468400845005519, + "learning_rate": 1.9997197078266723e-05, + "loss": 1.1829, + "step": 388 + }, + { + "epoch": 0.037418237783762985, + "grad_norm": 2.9989083747846874, + "learning_rate": 1.9997122835458617e-05, + "loss": 1.1584, + "step": 389 + }, + { + "epoch": 0.037514428626394765, + "grad_norm": 3.4240081093432635, + "learning_rate": 1.999704762234366e-05, + "loss": 1.2238, + "step": 390 + }, + { + "epoch": 0.03761061946902655, + "grad_norm": 3.0974129802655836, + "learning_rate": 1.999697143892916e-05, + "loss": 1.1084, + "step": 391 + }, + { + "epoch": 0.03770681031165833, + "grad_norm": 3.1898090510873938, + "learning_rate": 1.99968942852225e-05, + "loss": 1.1884, + "step": 392 + }, + { + "epoch": 0.03780300115429011, + "grad_norm": 4.042586114168573, + "learning_rate": 1.999681616123118e-05, + "loss": 1.1562, + "step": 393 + }, + { + "epoch": 0.03789919199692189, + "grad_norm": 3.2317689999257975, + "learning_rate": 1.9996737066962778e-05, + "loss": 1.1351, + "step": 394 + }, + { + "epoch": 0.03799538283955368, + "grad_norm": 3.133508241735913, + "learning_rate": 1.9996657002424967e-05, + "loss": 1.0682, + "step": 395 + }, + { + "epoch": 0.03809157368218546, + "grad_norm": 3.1005067333032974, + "learning_rate": 1.9996575967625525e-05, + "loss": 1.2677, + "step": 396 + }, + { + "epoch": 0.03818776452481724, + "grad_norm": 3.4170469161551167, + "learning_rate": 1.9996493962572315e-05, + "loss": 1.1318, + "step": 397 + }, + { + "epoch": 0.03828395536744902, + "grad_norm": 3.563777504826418, + "learning_rate": 1.999641098727329e-05, + "loss": 1.2233, + "step": 398 + }, + { + "epoch": 0.0383801462100808, + "grad_norm": 2.98021724109124, + "learning_rate": 1.9996327041736513e-05, + "loss": 1.1017, + "step": 399 + }, + { + "epoch": 0.038476337052712584, + "grad_norm": 3.094157041849158, + "learning_rate": 1.999624212597013e-05, + "loss": 1.185, + "step": 400 + }, + { + "epoch": 0.03857252789534436, + "grad_norm": 3.212465455186619, + "learning_rate": 1.999615623998237e-05, + "loss": 1.2039, + "step": 401 + }, + { + "epoch": 0.03866871873797614, + "grad_norm": 3.1789356602339423, + "learning_rate": 1.9996069383781587e-05, + "loss": 1.1553, + "step": 402 + }, + { + "epoch": 0.03876490958060792, + "grad_norm": 3.041456722236429, + "learning_rate": 1.99959815573762e-05, + "loss": 1.1443, + "step": 403 + }, + { + "epoch": 0.03886110042323971, + "grad_norm": 4.297737542383364, + "learning_rate": 1.9995892760774738e-05, + "loss": 1.0926, + "step": 404 + }, + { + "epoch": 0.03895729126587149, + "grad_norm": 2.2166198196612164, + "learning_rate": 1.9995802993985816e-05, + "loss": 0.7669, + "step": 405 + }, + { + "epoch": 0.03905348210850327, + "grad_norm": 3.1460829356629896, + "learning_rate": 1.9995712257018153e-05, + "loss": 1.1009, + "step": 406 + }, + { + "epoch": 0.03914967295113505, + "grad_norm": 4.438953929086321, + "learning_rate": 1.9995620549880545e-05, + "loss": 1.0815, + "step": 407 + }, + { + "epoch": 0.039245863793766836, + "grad_norm": 3.180470482584508, + "learning_rate": 1.9995527872581903e-05, + "loss": 1.0557, + "step": 408 + }, + { + "epoch": 0.039342054636398616, + "grad_norm": 3.2402896561140424, + "learning_rate": 1.9995434225131215e-05, + "loss": 1.067, + "step": 409 + }, + { + "epoch": 0.039438245479030395, + "grad_norm": 2.9187170045099817, + "learning_rate": 1.9995339607537578e-05, + "loss": 1.14, + "step": 410 + }, + { + "epoch": 0.039534436321662175, + "grad_norm": 3.2015083034453773, + "learning_rate": 1.999524401981017e-05, + "loss": 1.046, + "step": 411 + }, + { + "epoch": 0.03963062716429396, + "grad_norm": 3.078524275654608, + "learning_rate": 1.9995147461958267e-05, + "loss": 1.1294, + "step": 412 + }, + { + "epoch": 0.03972681800692574, + "grad_norm": 3.4946930874834616, + "learning_rate": 1.9995049933991245e-05, + "loss": 1.1248, + "step": 413 + }, + { + "epoch": 0.03982300884955752, + "grad_norm": 3.3616402484362804, + "learning_rate": 1.999495143591857e-05, + "loss": 1.0804, + "step": 414 + }, + { + "epoch": 0.0399191996921893, + "grad_norm": 3.0894241655284937, + "learning_rate": 1.99948519677498e-05, + "loss": 1.2405, + "step": 415 + }, + { + "epoch": 0.04001539053482109, + "grad_norm": 3.4840159616950257, + "learning_rate": 1.999475152949459e-05, + "loss": 1.0279, + "step": 416 + }, + { + "epoch": 0.04011158137745287, + "grad_norm": 3.566249611952745, + "learning_rate": 1.999465012116269e-05, + "loss": 1.1031, + "step": 417 + }, + { + "epoch": 0.04020777222008465, + "grad_norm": 3.24804538959389, + "learning_rate": 1.9994547742763935e-05, + "loss": 1.2502, + "step": 418 + }, + { + "epoch": 0.04030396306271643, + "grad_norm": 3.0864449456829557, + "learning_rate": 1.999444439430827e-05, + "loss": 1.1117, + "step": 419 + }, + { + "epoch": 0.040400153905348214, + "grad_norm": 3.5218765011981583, + "learning_rate": 1.9994340075805724e-05, + "loss": 1.0519, + "step": 420 + }, + { + "epoch": 0.040496344747979994, + "grad_norm": 2.976151090508982, + "learning_rate": 1.9994234787266423e-05, + "loss": 1.0958, + "step": 421 + }, + { + "epoch": 0.040592535590611774, + "grad_norm": 2.6860549509253135, + "learning_rate": 1.9994128528700583e-05, + "loss": 0.8524, + "step": 422 + }, + { + "epoch": 0.040688726433243554, + "grad_norm": 3.57865005955668, + "learning_rate": 1.9994021300118518e-05, + "loss": 1.1263, + "step": 423 + }, + { + "epoch": 0.04078491727587533, + "grad_norm": 3.0654210367558297, + "learning_rate": 1.9993913101530635e-05, + "loss": 1.0972, + "step": 424 + }, + { + "epoch": 0.04088110811850712, + "grad_norm": 3.0713941655894357, + "learning_rate": 1.999380393294744e-05, + "loss": 1.1559, + "step": 425 + }, + { + "epoch": 0.0409772989611389, + "grad_norm": 3.154639849223427, + "learning_rate": 1.9993693794379525e-05, + "loss": 1.1231, + "step": 426 + }, + { + "epoch": 0.04107348980377068, + "grad_norm": 3.536244704480314, + "learning_rate": 1.9993582685837582e-05, + "loss": 1.1078, + "step": 427 + }, + { + "epoch": 0.04116968064640246, + "grad_norm": 3.3543812914225457, + "learning_rate": 1.9993470607332387e-05, + "loss": 1.2264, + "step": 428 + }, + { + "epoch": 0.041265871489034246, + "grad_norm": 2.858830925867057, + "learning_rate": 1.999335755887483e-05, + "loss": 1.1281, + "step": 429 + }, + { + "epoch": 0.041362062331666026, + "grad_norm": 3.0912335324949174, + "learning_rate": 1.999324354047588e-05, + "loss": 1.0933, + "step": 430 + }, + { + "epoch": 0.041458253174297806, + "grad_norm": 3.1028904144466387, + "learning_rate": 1.9993128552146595e-05, + "loss": 1.1532, + "step": 431 + }, + { + "epoch": 0.041554444016929586, + "grad_norm": 2.131179290760393, + "learning_rate": 1.9993012593898146e-05, + "loss": 0.7633, + "step": 432 + }, + { + "epoch": 0.04165063485956137, + "grad_norm": 2.9651473091689167, + "learning_rate": 1.9992895665741783e-05, + "loss": 1.1305, + "step": 433 + }, + { + "epoch": 0.04174682570219315, + "grad_norm": 3.3197186023986065, + "learning_rate": 1.9992777767688857e-05, + "loss": 1.08, + "step": 434 + }, + { + "epoch": 0.04184301654482493, + "grad_norm": 3.194949091334714, + "learning_rate": 1.999265889975081e-05, + "loss": 1.1336, + "step": 435 + }, + { + "epoch": 0.04193920738745671, + "grad_norm": 2.6755290169931336, + "learning_rate": 1.9992539061939175e-05, + "loss": 1.0662, + "step": 436 + }, + { + "epoch": 0.0420353982300885, + "grad_norm": 3.224997796929354, + "learning_rate": 1.999241825426559e-05, + "loss": 1.2036, + "step": 437 + }, + { + "epoch": 0.04213158907272028, + "grad_norm": 3.181456862536907, + "learning_rate": 1.999229647674178e-05, + "loss": 1.0023, + "step": 438 + }, + { + "epoch": 0.04222777991535206, + "grad_norm": 3.250674964818327, + "learning_rate": 1.9992173729379557e-05, + "loss": 1.1853, + "step": 439 + }, + { + "epoch": 0.04232397075798384, + "grad_norm": 3.1907923255250563, + "learning_rate": 1.9992050012190845e-05, + "loss": 1.1656, + "step": 440 + }, + { + "epoch": 0.042420161600615625, + "grad_norm": 3.05451122446422, + "learning_rate": 1.9991925325187643e-05, + "loss": 1.0571, + "step": 441 + }, + { + "epoch": 0.042516352443247404, + "grad_norm": 3.343164159692236, + "learning_rate": 1.9991799668382058e-05, + "loss": 1.1009, + "step": 442 + }, + { + "epoch": 0.042612543285879184, + "grad_norm": 2.8370977315387065, + "learning_rate": 1.9991673041786287e-05, + "loss": 1.1303, + "step": 443 + }, + { + "epoch": 0.042708734128510964, + "grad_norm": 2.999667003572657, + "learning_rate": 1.9991545445412614e-05, + "loss": 1.0809, + "step": 444 + }, + { + "epoch": 0.042804924971142744, + "grad_norm": 2.6209104689579035, + "learning_rate": 1.999141687927343e-05, + "loss": 0.8406, + "step": 445 + }, + { + "epoch": 0.04290111581377453, + "grad_norm": 3.3358300436538237, + "learning_rate": 1.9991287343381208e-05, + "loss": 1.1984, + "step": 446 + }, + { + "epoch": 0.04299730665640631, + "grad_norm": 2.69417096409408, + "learning_rate": 1.9991156837748527e-05, + "loss": 1.0984, + "step": 447 + }, + { + "epoch": 0.04309349749903809, + "grad_norm": 3.435243831754951, + "learning_rate": 1.9991025362388044e-05, + "loss": 1.1552, + "step": 448 + }, + { + "epoch": 0.04318968834166987, + "grad_norm": 3.026036579851048, + "learning_rate": 1.999089291731253e-05, + "loss": 1.1839, + "step": 449 + }, + { + "epoch": 0.04328587918430166, + "grad_norm": 2.72046762848266, + "learning_rate": 1.9990759502534835e-05, + "loss": 1.1124, + "step": 450 + }, + { + "epoch": 0.043382070026933436, + "grad_norm": 2.9861256134274377, + "learning_rate": 1.9990625118067912e-05, + "loss": 1.1388, + "step": 451 + }, + { + "epoch": 0.043478260869565216, + "grad_norm": 2.8068066469584836, + "learning_rate": 1.9990489763924796e-05, + "loss": 0.772, + "step": 452 + }, + { + "epoch": 0.043574451712196996, + "grad_norm": 3.502675003110032, + "learning_rate": 1.9990353440118633e-05, + "loss": 1.0339, + "step": 453 + }, + { + "epoch": 0.04367064255482878, + "grad_norm": 2.2319528216104496, + "learning_rate": 1.9990216146662648e-05, + "loss": 0.805, + "step": 454 + }, + { + "epoch": 0.04376683339746056, + "grad_norm": 3.283148605344532, + "learning_rate": 1.9990077883570168e-05, + "loss": 1.1524, + "step": 455 + }, + { + "epoch": 0.04386302424009234, + "grad_norm": 3.295415343743631, + "learning_rate": 1.9989938650854618e-05, + "loss": 1.0856, + "step": 456 + }, + { + "epoch": 0.04395921508272412, + "grad_norm": 3.2689104785301177, + "learning_rate": 1.9989798448529506e-05, + "loss": 1.2202, + "step": 457 + }, + { + "epoch": 0.04405540592535591, + "grad_norm": 3.577780890599133, + "learning_rate": 1.998965727660844e-05, + "loss": 1.1267, + "step": 458 + }, + { + "epoch": 0.04415159676798769, + "grad_norm": 3.0581813246013176, + "learning_rate": 1.9989515135105123e-05, + "loss": 1.143, + "step": 459 + }, + { + "epoch": 0.04424778761061947, + "grad_norm": 3.0049519065114088, + "learning_rate": 1.9989372024033352e-05, + "loss": 1.1966, + "step": 460 + }, + { + "epoch": 0.04434397845325125, + "grad_norm": 2.7977540876175406, + "learning_rate": 1.9989227943407016e-05, + "loss": 1.0512, + "step": 461 + }, + { + "epoch": 0.044440169295883035, + "grad_norm": 3.3387302539440653, + "learning_rate": 1.99890828932401e-05, + "loss": 0.795, + "step": 462 + }, + { + "epoch": 0.044536360138514815, + "grad_norm": 3.011889396737117, + "learning_rate": 1.9988936873546684e-05, + "loss": 1.0818, + "step": 463 + }, + { + "epoch": 0.044632550981146595, + "grad_norm": 3.156103838264388, + "learning_rate": 1.9988789884340938e-05, + "loss": 1.1202, + "step": 464 + }, + { + "epoch": 0.044728741823778374, + "grad_norm": 2.8245398643002426, + "learning_rate": 1.998864192563713e-05, + "loss": 1.1476, + "step": 465 + }, + { + "epoch": 0.04482493266641016, + "grad_norm": 2.271459832517004, + "learning_rate": 1.9988492997449615e-05, + "loss": 0.8199, + "step": 466 + }, + { + "epoch": 0.04492112350904194, + "grad_norm": 3.0830085765110837, + "learning_rate": 1.998834309979286e-05, + "loss": 1.1591, + "step": 467 + }, + { + "epoch": 0.04501731435167372, + "grad_norm": 3.5526918449117786, + "learning_rate": 1.9988192232681398e-05, + "loss": 1.1973, + "step": 468 + }, + { + "epoch": 0.0451135051943055, + "grad_norm": 1.976079453459768, + "learning_rate": 1.9988040396129888e-05, + "loss": 0.8135, + "step": 469 + }, + { + "epoch": 0.04520969603693728, + "grad_norm": 3.399060091828394, + "learning_rate": 1.9987887590153055e-05, + "loss": 0.9895, + "step": 470 + }, + { + "epoch": 0.04530588687956907, + "grad_norm": 3.3103336690152405, + "learning_rate": 1.9987733814765734e-05, + "loss": 1.0454, + "step": 471 + }, + { + "epoch": 0.04540207772220085, + "grad_norm": 3.1684817445749514, + "learning_rate": 1.9987579069982856e-05, + "loss": 1.1311, + "step": 472 + }, + { + "epoch": 0.04549826856483263, + "grad_norm": 3.299134065343487, + "learning_rate": 1.998742335581943e-05, + "loss": 1.1481, + "step": 473 + }, + { + "epoch": 0.045594459407464406, + "grad_norm": 3.7225861513063765, + "learning_rate": 1.9987266672290577e-05, + "loss": 1.2232, + "step": 474 + }, + { + "epoch": 0.04569065025009619, + "grad_norm": 2.323856080798524, + "learning_rate": 1.99871090194115e-05, + "loss": 0.7817, + "step": 475 + }, + { + "epoch": 0.04578684109272797, + "grad_norm": 2.890754862758368, + "learning_rate": 1.9986950397197503e-05, + "loss": 1.1388, + "step": 476 + }, + { + "epoch": 0.04588303193535975, + "grad_norm": 3.2214001640925836, + "learning_rate": 1.9986790805663983e-05, + "loss": 1.0986, + "step": 477 + }, + { + "epoch": 0.04597922277799153, + "grad_norm": 3.920485357561589, + "learning_rate": 1.9986630244826425e-05, + "loss": 1.1049, + "step": 478 + }, + { + "epoch": 0.04607541362062332, + "grad_norm": 3.5815365454458363, + "learning_rate": 1.998646871470042e-05, + "loss": 1.1504, + "step": 479 + }, + { + "epoch": 0.0461716044632551, + "grad_norm": 2.948656398856783, + "learning_rate": 1.998630621530164e-05, + "loss": 1.1768, + "step": 480 + }, + { + "epoch": 0.04626779530588688, + "grad_norm": 3.042164654461808, + "learning_rate": 1.9986142746645858e-05, + "loss": 1.1644, + "step": 481 + }, + { + "epoch": 0.04636398614851866, + "grad_norm": 2.8301943760588366, + "learning_rate": 1.998597830874894e-05, + "loss": 1.0719, + "step": 482 + }, + { + "epoch": 0.046460176991150445, + "grad_norm": 2.94096443244728, + "learning_rate": 1.998581290162685e-05, + "loss": 1.0913, + "step": 483 + }, + { + "epoch": 0.046556367833782225, + "grad_norm": 3.0870950357863585, + "learning_rate": 1.9985646525295634e-05, + "loss": 1.0201, + "step": 484 + }, + { + "epoch": 0.046652558676414005, + "grad_norm": 2.555143899009505, + "learning_rate": 1.9985479179771447e-05, + "loss": 1.0109, + "step": 485 + }, + { + "epoch": 0.046748749519045785, + "grad_norm": 2.927864480037982, + "learning_rate": 1.998531086507053e-05, + "loss": 1.0477, + "step": 486 + }, + { + "epoch": 0.04684494036167757, + "grad_norm": 2.7488896536919625, + "learning_rate": 1.998514158120922e-05, + "loss": 1.1192, + "step": 487 + }, + { + "epoch": 0.04694113120430935, + "grad_norm": 3.0265654022457253, + "learning_rate": 1.9984971328203945e-05, + "loss": 1.1243, + "step": 488 + }, + { + "epoch": 0.04703732204694113, + "grad_norm": 3.500925980622524, + "learning_rate": 1.998480010607123e-05, + "loss": 0.9454, + "step": 489 + }, + { + "epoch": 0.04713351288957291, + "grad_norm": 2.716440036004909, + "learning_rate": 1.9984627914827698e-05, + "loss": 1.1372, + "step": 490 + }, + { + "epoch": 0.0472297037322047, + "grad_norm": 3.030905190738157, + "learning_rate": 1.9984454754490057e-05, + "loss": 1.1586, + "step": 491 + }, + { + "epoch": 0.04732589457483648, + "grad_norm": 2.6225113229458805, + "learning_rate": 1.9984280625075115e-05, + "loss": 1.1274, + "step": 492 + }, + { + "epoch": 0.04742208541746826, + "grad_norm": 3.216361229620014, + "learning_rate": 1.998410552659977e-05, + "loss": 0.9366, + "step": 493 + }, + { + "epoch": 0.04751827626010004, + "grad_norm": 3.1271072486516385, + "learning_rate": 1.9983929459081022e-05, + "loss": 1.0776, + "step": 494 + }, + { + "epoch": 0.04761446710273182, + "grad_norm": 3.2348572664233437, + "learning_rate": 1.9983752422535957e-05, + "loss": 1.1579, + "step": 495 + }, + { + "epoch": 0.0477106579453636, + "grad_norm": 2.7482297508621407, + "learning_rate": 1.998357441698176e-05, + "loss": 1.1578, + "step": 496 + }, + { + "epoch": 0.04780684878799538, + "grad_norm": 3.1643800943683833, + "learning_rate": 1.99833954424357e-05, + "loss": 1.1726, + "step": 497 + }, + { + "epoch": 0.04790303963062716, + "grad_norm": 3.083919782380232, + "learning_rate": 1.998321549891516e-05, + "loss": 1.1505, + "step": 498 + }, + { + "epoch": 0.04799923047325894, + "grad_norm": 3.114186125952616, + "learning_rate": 1.9983034586437594e-05, + "loss": 1.1801, + "step": 499 + }, + { + "epoch": 0.04809542131589073, + "grad_norm": 3.250940127846583, + "learning_rate": 1.9982852705020572e-05, + "loss": 1.1287, + "step": 500 + }, + { + "epoch": 0.04819161215852251, + "grad_norm": 2.9303558023934597, + "learning_rate": 1.998266985468174e-05, + "loss": 1.1324, + "step": 501 + }, + { + "epoch": 0.04828780300115429, + "grad_norm": 2.8751273909601713, + "learning_rate": 1.9982486035438848e-05, + "loss": 1.1384, + "step": 502 + }, + { + "epoch": 0.04838399384378607, + "grad_norm": 3.2482242015967975, + "learning_rate": 1.9982301247309734e-05, + "loss": 1.1245, + "step": 503 + }, + { + "epoch": 0.048480184686417856, + "grad_norm": 3.3461579504331103, + "learning_rate": 1.9982115490312334e-05, + "loss": 0.9927, + "step": 504 + }, + { + "epoch": 0.048576375529049635, + "grad_norm": 2.925827678425482, + "learning_rate": 1.998192876446468e-05, + "loss": 1.1178, + "step": 505 + }, + { + "epoch": 0.048672566371681415, + "grad_norm": 3.24447675002664, + "learning_rate": 1.9981741069784894e-05, + "loss": 1.1847, + "step": 506 + }, + { + "epoch": 0.048768757214313195, + "grad_norm": 2.7443320943066127, + "learning_rate": 1.9981552406291192e-05, + "loss": 1.0946, + "step": 507 + }, + { + "epoch": 0.04886494805694498, + "grad_norm": 2.764162294528297, + "learning_rate": 1.9981362774001886e-05, + "loss": 1.099, + "step": 508 + }, + { + "epoch": 0.04896113889957676, + "grad_norm": 3.4990009333714407, + "learning_rate": 1.9981172172935387e-05, + "loss": 0.9557, + "step": 509 + }, + { + "epoch": 0.04905732974220854, + "grad_norm": 3.498886724960836, + "learning_rate": 1.9980980603110185e-05, + "loss": 1.1304, + "step": 510 + }, + { + "epoch": 0.04915352058484032, + "grad_norm": 3.0963152200769612, + "learning_rate": 1.9980788064544877e-05, + "loss": 1.1824, + "step": 511 + }, + { + "epoch": 0.04924971142747211, + "grad_norm": 3.631987119560739, + "learning_rate": 1.9980594557258158e-05, + "loss": 1.1874, + "step": 512 + }, + { + "epoch": 0.04934590227010389, + "grad_norm": 3.0349610976781127, + "learning_rate": 1.99804000812688e-05, + "loss": 1.0302, + "step": 513 + }, + { + "epoch": 0.04944209311273567, + "grad_norm": 2.952420176214481, + "learning_rate": 1.9980204636595682e-05, + "loss": 1.1883, + "step": 514 + }, + { + "epoch": 0.04953828395536745, + "grad_norm": 3.0881906092213605, + "learning_rate": 1.9980008223257773e-05, + "loss": 1.1261, + "step": 515 + }, + { + "epoch": 0.04963447479799923, + "grad_norm": 2.9611270786475563, + "learning_rate": 1.9979810841274135e-05, + "loss": 1.0345, + "step": 516 + }, + { + "epoch": 0.049730665640631014, + "grad_norm": 2.5977112326839316, + "learning_rate": 1.997961249066393e-05, + "loss": 1.135, + "step": 517 + }, + { + "epoch": 0.049826856483262794, + "grad_norm": 3.2569151196316235, + "learning_rate": 1.9979413171446403e-05, + "loss": 1.0955, + "step": 518 + }, + { + "epoch": 0.04992304732589457, + "grad_norm": 3.150605988237454, + "learning_rate": 1.997921288364091e-05, + "loss": 1.1347, + "step": 519 + }, + { + "epoch": 0.05001923816852635, + "grad_norm": 2.9458139164861428, + "learning_rate": 1.9979011627266884e-05, + "loss": 1.0285, + "step": 520 + }, + { + "epoch": 0.05011542901115814, + "grad_norm": 3.352587244113385, + "learning_rate": 1.9978809402343856e-05, + "loss": 1.1378, + "step": 521 + }, + { + "epoch": 0.05021161985378992, + "grad_norm": 3.3244806643960785, + "learning_rate": 1.997860620889146e-05, + "loss": 1.09, + "step": 522 + }, + { + "epoch": 0.0503078106964217, + "grad_norm": 3.0649672059537276, + "learning_rate": 1.997840204692941e-05, + "loss": 1.0646, + "step": 523 + }, + { + "epoch": 0.05040400153905348, + "grad_norm": 3.5469795216926343, + "learning_rate": 1.997819691647753e-05, + "loss": 1.1154, + "step": 524 + }, + { + "epoch": 0.050500192381685266, + "grad_norm": 3.2264942274214046, + "learning_rate": 1.9977990817555726e-05, + "loss": 1.1523, + "step": 525 + }, + { + "epoch": 0.050596383224317046, + "grad_norm": 3.090557948477923, + "learning_rate": 1.9977783750184e-05, + "loss": 1.0491, + "step": 526 + }, + { + "epoch": 0.050692574066948826, + "grad_norm": 3.039141598203861, + "learning_rate": 1.9977575714382453e-05, + "loss": 1.1626, + "step": 527 + }, + { + "epoch": 0.050788764909580605, + "grad_norm": 2.4632472829565524, + "learning_rate": 1.9977366710171274e-05, + "loss": 0.7873, + "step": 528 + }, + { + "epoch": 0.05088495575221239, + "grad_norm": 3.0125346146308947, + "learning_rate": 1.9977156737570746e-05, + "loss": 1.0904, + "step": 529 + }, + { + "epoch": 0.05098114659484417, + "grad_norm": 3.345106135501577, + "learning_rate": 1.9976945796601258e-05, + "loss": 1.1497, + "step": 530 + }, + { + "epoch": 0.05107733743747595, + "grad_norm": 3.045565980629826, + "learning_rate": 1.9976733887283274e-05, + "loss": 1.0754, + "step": 531 + }, + { + "epoch": 0.05117352828010773, + "grad_norm": 3.213561287240602, + "learning_rate": 1.9976521009637366e-05, + "loss": 1.1625, + "step": 532 + }, + { + "epoch": 0.05126971912273952, + "grad_norm": 3.4507998844373358, + "learning_rate": 1.9976307163684197e-05, + "loss": 1.1996, + "step": 533 + }, + { + "epoch": 0.0513659099653713, + "grad_norm": 3.1552597000347, + "learning_rate": 1.997609234944452e-05, + "loss": 1.1196, + "step": 534 + }, + { + "epoch": 0.05146210080800308, + "grad_norm": 3.1704486320456913, + "learning_rate": 1.9975876566939186e-05, + "loss": 1.1654, + "step": 535 + }, + { + "epoch": 0.05155829165063486, + "grad_norm": 3.103248087090796, + "learning_rate": 1.9975659816189137e-05, + "loss": 1.0146, + "step": 536 + }, + { + "epoch": 0.051654482493266644, + "grad_norm": 2.857675392250267, + "learning_rate": 1.997544209721541e-05, + "loss": 1.0927, + "step": 537 + }, + { + "epoch": 0.051750673335898424, + "grad_norm": 2.6692344536815775, + "learning_rate": 1.997522341003914e-05, + "loss": 1.1778, + "step": 538 + }, + { + "epoch": 0.051846864178530204, + "grad_norm": 3.3180547158071065, + "learning_rate": 1.997500375468155e-05, + "loss": 1.1197, + "step": 539 + }, + { + "epoch": 0.051943055021161984, + "grad_norm": 3.169278901066095, + "learning_rate": 1.9974783131163957e-05, + "loss": 1.2064, + "step": 540 + }, + { + "epoch": 0.052039245863793764, + "grad_norm": 2.9931794760136268, + "learning_rate": 1.997456153950778e-05, + "loss": 1.2103, + "step": 541 + }, + { + "epoch": 0.05213543670642555, + "grad_norm": 3.108376080368456, + "learning_rate": 1.9974338979734523e-05, + "loss": 1.118, + "step": 542 + }, + { + "epoch": 0.05223162754905733, + "grad_norm": 3.1204823975006106, + "learning_rate": 1.997411545186579e-05, + "loss": 1.1709, + "step": 543 + }, + { + "epoch": 0.05232781839168911, + "grad_norm": 2.8852674025701814, + "learning_rate": 1.997389095592327e-05, + "loss": 1.105, + "step": 544 + }, + { + "epoch": 0.05242400923432089, + "grad_norm": 2.921407865820021, + "learning_rate": 1.997366549192876e-05, + "loss": 1.08, + "step": 545 + }, + { + "epoch": 0.052520200076952676, + "grad_norm": 3.123394492914511, + "learning_rate": 1.9973439059904133e-05, + "loss": 1.2615, + "step": 546 + }, + { + "epoch": 0.052616390919584456, + "grad_norm": 3.1110431954721127, + "learning_rate": 1.997321165987138e-05, + "loss": 0.9751, + "step": 547 + }, + { + "epoch": 0.052712581762216236, + "grad_norm": 3.652792052584072, + "learning_rate": 1.9972983291852565e-05, + "loss": 1.1534, + "step": 548 + }, + { + "epoch": 0.052808772604848016, + "grad_norm": 3.236379736600093, + "learning_rate": 1.9972753955869848e-05, + "loss": 1.0862, + "step": 549 + }, + { + "epoch": 0.0529049634474798, + "grad_norm": 2.732332301586399, + "learning_rate": 1.9972523651945496e-05, + "loss": 1.1689, + "step": 550 + }, + { + "epoch": 0.05300115429011158, + "grad_norm": 2.72772334143674, + "learning_rate": 1.9972292380101857e-05, + "loss": 0.9587, + "step": 551 + }, + { + "epoch": 0.05309734513274336, + "grad_norm": 3.2146770422074256, + "learning_rate": 1.9972060140361384e-05, + "loss": 1.2157, + "step": 552 + }, + { + "epoch": 0.05319353597537514, + "grad_norm": 3.4762953440214464, + "learning_rate": 1.9971826932746615e-05, + "loss": 1.2298, + "step": 553 + }, + { + "epoch": 0.05328972681800693, + "grad_norm": 3.410772057621862, + "learning_rate": 1.997159275728018e-05, + "loss": 1.1403, + "step": 554 + }, + { + "epoch": 0.05338591766063871, + "grad_norm": 2.7534192114865585, + "learning_rate": 1.997135761398481e-05, + "loss": 1.0822, + "step": 555 + }, + { + "epoch": 0.05348210850327049, + "grad_norm": 2.826327363903914, + "learning_rate": 1.9971121502883332e-05, + "loss": 1.0971, + "step": 556 + }, + { + "epoch": 0.05357829934590227, + "grad_norm": 2.698343316661684, + "learning_rate": 1.9970884423998657e-05, + "loss": 0.7476, + "step": 557 + }, + { + "epoch": 0.053674490188534055, + "grad_norm": 3.1254382395856695, + "learning_rate": 1.9970646377353802e-05, + "loss": 1.177, + "step": 558 + }, + { + "epoch": 0.053770681031165835, + "grad_norm": 2.8490896779609787, + "learning_rate": 1.9970407362971867e-05, + "loss": 1.0806, + "step": 559 + }, + { + "epoch": 0.053866871873797614, + "grad_norm": 2.633044481155465, + "learning_rate": 1.997016738087605e-05, + "loss": 0.9836, + "step": 560 + }, + { + "epoch": 0.053963062716429394, + "grad_norm": 3.1036708119400918, + "learning_rate": 1.996992643108964e-05, + "loss": 1.081, + "step": 561 + }, + { + "epoch": 0.054059253559061174, + "grad_norm": 2.708580644433421, + "learning_rate": 1.9969684513636035e-05, + "loss": 1.1136, + "step": 562 + }, + { + "epoch": 0.05415544440169296, + "grad_norm": 3.3306774275533475, + "learning_rate": 1.9969441628538706e-05, + "loss": 1.121, + "step": 563 + }, + { + "epoch": 0.05425163524432474, + "grad_norm": 2.598094291872369, + "learning_rate": 1.9969197775821227e-05, + "loss": 1.0417, + "step": 564 + }, + { + "epoch": 0.05434782608695652, + "grad_norm": 3.121197357195202, + "learning_rate": 1.996895295550727e-05, + "loss": 1.2003, + "step": 565 + }, + { + "epoch": 0.0544440169295883, + "grad_norm": 3.2822145223118664, + "learning_rate": 1.9968707167620593e-05, + "loss": 1.0985, + "step": 566 + }, + { + "epoch": 0.05454020777222009, + "grad_norm": 2.878988981388173, + "learning_rate": 1.9968460412185054e-05, + "loss": 0.8806, + "step": 567 + }, + { + "epoch": 0.05463639861485187, + "grad_norm": 3.1772776394073694, + "learning_rate": 1.9968212689224603e-05, + "loss": 1.0856, + "step": 568 + }, + { + "epoch": 0.054732589457483646, + "grad_norm": 2.6501842678392498, + "learning_rate": 1.9967963998763285e-05, + "loss": 1.1224, + "step": 569 + }, + { + "epoch": 0.054828780300115426, + "grad_norm": 2.8769042755597676, + "learning_rate": 1.996771434082523e-05, + "loss": 1.0568, + "step": 570 + }, + { + "epoch": 0.05492497114274721, + "grad_norm": 2.959446810365323, + "learning_rate": 1.9967463715434683e-05, + "loss": 1.2569, + "step": 571 + }, + { + "epoch": 0.05502116198537899, + "grad_norm": 3.0729130089521735, + "learning_rate": 1.9967212122615958e-05, + "loss": 1.1186, + "step": 572 + }, + { + "epoch": 0.05511735282801077, + "grad_norm": 2.7066400847774355, + "learning_rate": 1.9966959562393477e-05, + "loss": 1.0437, + "step": 573 + }, + { + "epoch": 0.05521354367064255, + "grad_norm": 2.851023921120213, + "learning_rate": 1.9966706034791752e-05, + "loss": 0.846, + "step": 574 + }, + { + "epoch": 0.05530973451327434, + "grad_norm": 2.7535134075996637, + "learning_rate": 1.9966451539835397e-05, + "loss": 1.1503, + "step": 575 + }, + { + "epoch": 0.05540592535590612, + "grad_norm": 2.98279565538089, + "learning_rate": 1.9966196077549106e-05, + "loss": 1.2241, + "step": 576 + }, + { + "epoch": 0.0555021161985379, + "grad_norm": 3.672076857299595, + "learning_rate": 1.996593964795767e-05, + "loss": 1.1014, + "step": 577 + }, + { + "epoch": 0.05559830704116968, + "grad_norm": 2.7330565522050123, + "learning_rate": 1.996568225108599e-05, + "loss": 1.225, + "step": 578 + }, + { + "epoch": 0.055694497883801465, + "grad_norm": 3.630031664966837, + "learning_rate": 1.996542388695904e-05, + "loss": 1.1307, + "step": 579 + }, + { + "epoch": 0.055790688726433245, + "grad_norm": 2.857511592677229, + "learning_rate": 1.99651645556019e-05, + "loss": 1.1007, + "step": 580 + }, + { + "epoch": 0.055886879569065025, + "grad_norm": 3.230148915944173, + "learning_rate": 1.9964904257039737e-05, + "loss": 1.1202, + "step": 581 + }, + { + "epoch": 0.055983070411696804, + "grad_norm": 2.8125028026928383, + "learning_rate": 1.9964642991297817e-05, + "loss": 1.2225, + "step": 582 + }, + { + "epoch": 0.05607926125432859, + "grad_norm": 3.1063408670532215, + "learning_rate": 1.9964380758401497e-05, + "loss": 1.009, + "step": 583 + }, + { + "epoch": 0.05617545209696037, + "grad_norm": 2.659369728612584, + "learning_rate": 1.996411755837623e-05, + "loss": 1.0817, + "step": 584 + }, + { + "epoch": 0.05627164293959215, + "grad_norm": 2.990524619517537, + "learning_rate": 1.9963853391247566e-05, + "loss": 1.0986, + "step": 585 + }, + { + "epoch": 0.05636783378222393, + "grad_norm": 2.50939916449769, + "learning_rate": 1.9963588257041137e-05, + "loss": 1.1341, + "step": 586 + }, + { + "epoch": 0.05646402462485571, + "grad_norm": 3.0033675117604335, + "learning_rate": 1.996332215578268e-05, + "loss": 1.076, + "step": 587 + }, + { + "epoch": 0.0565602154674875, + "grad_norm": 3.2369486809105936, + "learning_rate": 1.996305508749802e-05, + "loss": 1.0551, + "step": 588 + }, + { + "epoch": 0.05665640631011928, + "grad_norm": 3.0859964906448885, + "learning_rate": 1.9962787052213085e-05, + "loss": 1.0919, + "step": 589 + }, + { + "epoch": 0.05675259715275106, + "grad_norm": 2.851304153251838, + "learning_rate": 1.9962518049953887e-05, + "loss": 1.0631, + "step": 590 + }, + { + "epoch": 0.056848787995382837, + "grad_norm": 3.5871574597561584, + "learning_rate": 1.996224808074653e-05, + "loss": 1.1938, + "step": 591 + }, + { + "epoch": 0.05694497883801462, + "grad_norm": 3.1493707667156876, + "learning_rate": 1.9961977144617225e-05, + "loss": 1.0774, + "step": 592 + }, + { + "epoch": 0.0570411696806464, + "grad_norm": 3.2451421050844487, + "learning_rate": 1.996170524159226e-05, + "loss": 1.2243, + "step": 593 + }, + { + "epoch": 0.05713736052327818, + "grad_norm": 2.9718976596367765, + "learning_rate": 1.996143237169803e-05, + "loss": 1.1004, + "step": 594 + }, + { + "epoch": 0.05723355136590996, + "grad_norm": 3.445925729912563, + "learning_rate": 1.996115853496102e-05, + "loss": 1.0325, + "step": 595 + }, + { + "epoch": 0.05732974220854175, + "grad_norm": 2.831797028889002, + "learning_rate": 1.996088373140781e-05, + "loss": 1.1493, + "step": 596 + }, + { + "epoch": 0.05742593305117353, + "grad_norm": 2.7279721341849816, + "learning_rate": 1.9960607961065072e-05, + "loss": 1.1067, + "step": 597 + }, + { + "epoch": 0.05752212389380531, + "grad_norm": 2.789667712744687, + "learning_rate": 1.9960331223959564e-05, + "loss": 1.0729, + "step": 598 + }, + { + "epoch": 0.05761831473643709, + "grad_norm": 3.0742157606952243, + "learning_rate": 1.9960053520118152e-05, + "loss": 1.2279, + "step": 599 + }, + { + "epoch": 0.057714505579068875, + "grad_norm": 2.5391873808689347, + "learning_rate": 1.995977484956779e-05, + "loss": 0.8366, + "step": 600 + }, + { + "epoch": 0.057810696421700655, + "grad_norm": 3.140690921176633, + "learning_rate": 1.9959495212335524e-05, + "loss": 1.187, + "step": 601 + }, + { + "epoch": 0.057906887264332435, + "grad_norm": 3.0226633420129376, + "learning_rate": 1.9959214608448495e-05, + "loss": 1.101, + "step": 602 + }, + { + "epoch": 0.058003078106964215, + "grad_norm": 3.0521374462620203, + "learning_rate": 1.995893303793394e-05, + "loss": 1.0499, + "step": 603 + }, + { + "epoch": 0.058099268949596, + "grad_norm": 3.3403439569416298, + "learning_rate": 1.9958650500819183e-05, + "loss": 1.1934, + "step": 604 + }, + { + "epoch": 0.05819545979222778, + "grad_norm": 2.643872627003572, + "learning_rate": 1.995836699713165e-05, + "loss": 1.0581, + "step": 605 + }, + { + "epoch": 0.05829165063485956, + "grad_norm": 3.0764147269551425, + "learning_rate": 1.995808252689886e-05, + "loss": 1.0575, + "step": 606 + }, + { + "epoch": 0.05838784147749134, + "grad_norm": 2.831062760177175, + "learning_rate": 1.995779709014842e-05, + "loss": 1.1842, + "step": 607 + }, + { + "epoch": 0.05848403232012313, + "grad_norm": 2.5917424888172698, + "learning_rate": 1.9957510686908034e-05, + "loss": 0.9523, + "step": 608 + }, + { + "epoch": 0.05858022316275491, + "grad_norm": 2.7734240078861476, + "learning_rate": 1.99572233172055e-05, + "loss": 1.1694, + "step": 609 + }, + { + "epoch": 0.05867641400538669, + "grad_norm": 2.7232716571545237, + "learning_rate": 1.9956934981068713e-05, + "loss": 1.0399, + "step": 610 + }, + { + "epoch": 0.05877260484801847, + "grad_norm": 2.8381133085794974, + "learning_rate": 1.9956645678525654e-05, + "loss": 1.1574, + "step": 611 + }, + { + "epoch": 0.05886879569065025, + "grad_norm": 2.9834957768385446, + "learning_rate": 1.9956355409604402e-05, + "loss": 1.1605, + "step": 612 + }, + { + "epoch": 0.058964986533282034, + "grad_norm": 2.9636418359092396, + "learning_rate": 1.9956064174333135e-05, + "loss": 1.1274, + "step": 613 + }, + { + "epoch": 0.05906117737591381, + "grad_norm": 3.0873131055838394, + "learning_rate": 1.9955771972740118e-05, + "loss": 1.2135, + "step": 614 + }, + { + "epoch": 0.05915736821854559, + "grad_norm": 2.797199126471606, + "learning_rate": 1.9955478804853707e-05, + "loss": 1.1736, + "step": 615 + }, + { + "epoch": 0.05925355906117737, + "grad_norm": 2.825522954706494, + "learning_rate": 1.9955184670702363e-05, + "loss": 1.1871, + "step": 616 + }, + { + "epoch": 0.05934974990380916, + "grad_norm": 3.2386218963834237, + "learning_rate": 1.995488957031463e-05, + "loss": 1.2171, + "step": 617 + }, + { + "epoch": 0.05944594074644094, + "grad_norm": 3.2342661158141155, + "learning_rate": 1.995459350371915e-05, + "loss": 1.2275, + "step": 618 + }, + { + "epoch": 0.05954213158907272, + "grad_norm": 2.624918455428919, + "learning_rate": 1.9954296470944666e-05, + "loss": 1.0808, + "step": 619 + }, + { + "epoch": 0.0596383224317045, + "grad_norm": 2.6457961645085684, + "learning_rate": 1.9953998472019996e-05, + "loss": 1.0193, + "step": 620 + }, + { + "epoch": 0.059734513274336286, + "grad_norm": 6.293400118325079, + "learning_rate": 1.9953699506974073e-05, + "loss": 1.2115, + "step": 621 + }, + { + "epoch": 0.059830704116968066, + "grad_norm": 2.5853848622081506, + "learning_rate": 1.995339957583591e-05, + "loss": 1.1463, + "step": 622 + }, + { + "epoch": 0.059926894959599845, + "grad_norm": 3.606148993186475, + "learning_rate": 1.995309867863462e-05, + "loss": 1.2341, + "step": 623 + }, + { + "epoch": 0.060023085802231625, + "grad_norm": 3.0407721983071956, + "learning_rate": 1.9952796815399403e-05, + "loss": 1.1892, + "step": 624 + }, + { + "epoch": 0.06011927664486341, + "grad_norm": 2.528843305816138, + "learning_rate": 1.995249398615956e-05, + "loss": 0.7696, + "step": 625 + }, + { + "epoch": 0.06021546748749519, + "grad_norm": 2.8741561294070697, + "learning_rate": 1.9952190190944484e-05, + "loss": 1.0855, + "step": 626 + }, + { + "epoch": 0.06031165833012697, + "grad_norm": 2.835034471781129, + "learning_rate": 1.995188542978366e-05, + "loss": 1.1286, + "step": 627 + }, + { + "epoch": 0.06040784917275875, + "grad_norm": 1.9374724980064657, + "learning_rate": 1.9951579702706668e-05, + "loss": 0.7571, + "step": 628 + }, + { + "epoch": 0.06050404001539054, + "grad_norm": 3.100441621063448, + "learning_rate": 1.9951273009743186e-05, + "loss": 1.1086, + "step": 629 + }, + { + "epoch": 0.06060023085802232, + "grad_norm": 2.892024105372368, + "learning_rate": 1.9950965350922975e-05, + "loss": 1.0767, + "step": 630 + }, + { + "epoch": 0.0606964217006541, + "grad_norm": 2.8841947567418478, + "learning_rate": 1.9950656726275897e-05, + "loss": 1.2358, + "step": 631 + }, + { + "epoch": 0.06079261254328588, + "grad_norm": 2.7808021022888525, + "learning_rate": 1.9950347135831907e-05, + "loss": 1.1206, + "step": 632 + }, + { + "epoch": 0.06088880338591766, + "grad_norm": 3.2224917330142797, + "learning_rate": 1.9950036579621053e-05, + "loss": 1.0806, + "step": 633 + }, + { + "epoch": 0.060984994228549444, + "grad_norm": 2.6879509181814427, + "learning_rate": 1.994972505767348e-05, + "loss": 1.1409, + "step": 634 + }, + { + "epoch": 0.061081185071181224, + "grad_norm": 3.344678034412936, + "learning_rate": 1.994941257001942e-05, + "loss": 1.1652, + "step": 635 + }, + { + "epoch": 0.061177375913813004, + "grad_norm": 2.803833683449268, + "learning_rate": 1.994909911668921e-05, + "loss": 1.0415, + "step": 636 + }, + { + "epoch": 0.06127356675644478, + "grad_norm": 3.1467227767199875, + "learning_rate": 1.994878469771326e-05, + "loss": 1.1809, + "step": 637 + }, + { + "epoch": 0.06136975759907657, + "grad_norm": 2.974339354291046, + "learning_rate": 1.99484693131221e-05, + "loss": 0.9751, + "step": 638 + }, + { + "epoch": 0.06146594844170835, + "grad_norm": 2.906214412121132, + "learning_rate": 1.9948152962946335e-05, + "loss": 1.1149, + "step": 639 + }, + { + "epoch": 0.06156213928434013, + "grad_norm": 2.7962192818132325, + "learning_rate": 1.994783564721667e-05, + "loss": 1.0802, + "step": 640 + }, + { + "epoch": 0.06165833012697191, + "grad_norm": 2.843002707440342, + "learning_rate": 1.9947517365963903e-05, + "loss": 1.1716, + "step": 641 + }, + { + "epoch": 0.061754520969603696, + "grad_norm": 2.805023572469057, + "learning_rate": 1.9947198119218924e-05, + "loss": 1.0613, + "step": 642 + }, + { + "epoch": 0.061850711812235476, + "grad_norm": 2.6213253392111913, + "learning_rate": 1.9946877907012727e-05, + "loss": 1.1188, + "step": 643 + }, + { + "epoch": 0.061946902654867256, + "grad_norm": 2.7591750516935707, + "learning_rate": 1.994655672937638e-05, + "loss": 1.1073, + "step": 644 + }, + { + "epoch": 0.062043093497499036, + "grad_norm": 2.941790909334639, + "learning_rate": 1.9946234586341063e-05, + "loss": 0.8642, + "step": 645 + }, + { + "epoch": 0.06213928434013082, + "grad_norm": 3.3345134769374765, + "learning_rate": 1.9945911477938044e-05, + "loss": 1.0473, + "step": 646 + }, + { + "epoch": 0.0622354751827626, + "grad_norm": 2.803889519265455, + "learning_rate": 1.994558740419868e-05, + "loss": 1.1685, + "step": 647 + }, + { + "epoch": 0.06233166602539438, + "grad_norm": 3.6804023579393523, + "learning_rate": 1.994526236515442e-05, + "loss": 0.9806, + "step": 648 + }, + { + "epoch": 0.06242785686802616, + "grad_norm": 2.8621640809694364, + "learning_rate": 1.9944936360836826e-05, + "loss": 0.8768, + "step": 649 + }, + { + "epoch": 0.06252404771065795, + "grad_norm": 2.6288242093288954, + "learning_rate": 1.994460939127753e-05, + "loss": 1.0851, + "step": 650 + }, + { + "epoch": 0.06262023855328973, + "grad_norm": 2.7880449753061445, + "learning_rate": 1.9944281456508264e-05, + "loss": 1.1124, + "step": 651 + }, + { + "epoch": 0.06271642939592151, + "grad_norm": 2.830232235757219, + "learning_rate": 1.9943952556560863e-05, + "loss": 1.1859, + "step": 652 + }, + { + "epoch": 0.06281262023855329, + "grad_norm": 2.225920770001663, + "learning_rate": 1.9943622691467246e-05, + "loss": 0.774, + "step": 653 + }, + { + "epoch": 0.06290881108118507, + "grad_norm": 3.1659893561604484, + "learning_rate": 1.9943291861259433e-05, + "loss": 1.218, + "step": 654 + }, + { + "epoch": 0.06300500192381685, + "grad_norm": 3.0019492000767274, + "learning_rate": 1.994296006596953e-05, + "loss": 1.1181, + "step": 655 + }, + { + "epoch": 0.06310119276644863, + "grad_norm": 3.147921119776967, + "learning_rate": 1.9942627305629747e-05, + "loss": 1.1714, + "step": 656 + }, + { + "epoch": 0.06319738360908042, + "grad_norm": 3.032299301203868, + "learning_rate": 1.9942293580272377e-05, + "loss": 1.1005, + "step": 657 + }, + { + "epoch": 0.0632935744517122, + "grad_norm": 2.9858181614558865, + "learning_rate": 1.9941958889929808e-05, + "loss": 1.2725, + "step": 658 + }, + { + "epoch": 0.06338976529434398, + "grad_norm": 2.688617086612652, + "learning_rate": 1.994162323463453e-05, + "loss": 1.2161, + "step": 659 + }, + { + "epoch": 0.06348595613697576, + "grad_norm": 3.0063912267752175, + "learning_rate": 1.9941286614419113e-05, + "loss": 1.133, + "step": 660 + }, + { + "epoch": 0.06358214697960754, + "grad_norm": 3.421987624811044, + "learning_rate": 1.994094902931624e-05, + "loss": 1.1723, + "step": 661 + }, + { + "epoch": 0.06367833782223932, + "grad_norm": 2.628362648730503, + "learning_rate": 1.994061047935867e-05, + "loss": 1.1782, + "step": 662 + }, + { + "epoch": 0.0637745286648711, + "grad_norm": 2.4921104524222693, + "learning_rate": 1.994027096457926e-05, + "loss": 1.023, + "step": 663 + }, + { + "epoch": 0.06387071950750288, + "grad_norm": 2.4603454014149637, + "learning_rate": 1.9939930485010968e-05, + "loss": 0.9562, + "step": 664 + }, + { + "epoch": 0.06396691035013467, + "grad_norm": 2.777054846800436, + "learning_rate": 1.993958904068684e-05, + "loss": 1.0503, + "step": 665 + }, + { + "epoch": 0.06406310119276645, + "grad_norm": 3.263456533155639, + "learning_rate": 1.9939246631640014e-05, + "loss": 1.1253, + "step": 666 + }, + { + "epoch": 0.06415929203539823, + "grad_norm": 2.915101321887973, + "learning_rate": 1.9938903257903726e-05, + "loss": 1.2372, + "step": 667 + }, + { + "epoch": 0.06425548287803001, + "grad_norm": 2.7507052773342116, + "learning_rate": 1.99385589195113e-05, + "loss": 1.1663, + "step": 668 + }, + { + "epoch": 0.06435167372066179, + "grad_norm": 2.5122155667444392, + "learning_rate": 1.9938213616496157e-05, + "loss": 1.0944, + "step": 669 + }, + { + "epoch": 0.06444786456329357, + "grad_norm": 2.6784294409720446, + "learning_rate": 1.9937867348891815e-05, + "loss": 1.1147, + "step": 670 + }, + { + "epoch": 0.06454405540592535, + "grad_norm": 3.12097801102859, + "learning_rate": 1.9937520116731882e-05, + "loss": 1.093, + "step": 671 + }, + { + "epoch": 0.06464024624855713, + "grad_norm": 2.784943264725122, + "learning_rate": 1.9937171920050057e-05, + "loss": 1.1592, + "step": 672 + }, + { + "epoch": 0.06473643709118893, + "grad_norm": 2.664410072917954, + "learning_rate": 1.9936822758880137e-05, + "loss": 1.058, + "step": 673 + }, + { + "epoch": 0.0648326279338207, + "grad_norm": 2.427521201434085, + "learning_rate": 1.9936472633256012e-05, + "loss": 0.9254, + "step": 674 + }, + { + "epoch": 0.06492881877645248, + "grad_norm": 3.0614421145603266, + "learning_rate": 1.9936121543211663e-05, + "loss": 1.1234, + "step": 675 + }, + { + "epoch": 0.06502500961908426, + "grad_norm": 3.0465449042859016, + "learning_rate": 1.9935769488781167e-05, + "loss": 1.1972, + "step": 676 + }, + { + "epoch": 0.06512120046171604, + "grad_norm": 2.5348150741068585, + "learning_rate": 1.9935416469998697e-05, + "loss": 1.0615, + "step": 677 + }, + { + "epoch": 0.06521739130434782, + "grad_norm": 2.8117278557823573, + "learning_rate": 1.993506248689851e-05, + "loss": 1.1761, + "step": 678 + }, + { + "epoch": 0.0653135821469796, + "grad_norm": 2.6591945503732406, + "learning_rate": 1.993470753951497e-05, + "loss": 1.1534, + "step": 679 + }, + { + "epoch": 0.06540977298961138, + "grad_norm": 3.0421501032554774, + "learning_rate": 1.993435162788252e-05, + "loss": 1.139, + "step": 680 + }, + { + "epoch": 0.06550596383224316, + "grad_norm": 2.8800450553464083, + "learning_rate": 1.993399475203571e-05, + "loss": 1.0934, + "step": 681 + }, + { + "epoch": 0.06560215467487496, + "grad_norm": 2.864062457184436, + "learning_rate": 1.993363691200918e-05, + "loss": 1.1328, + "step": 682 + }, + { + "epoch": 0.06569834551750674, + "grad_norm": 2.875567065599544, + "learning_rate": 1.9933278107837656e-05, + "loss": 1.1361, + "step": 683 + }, + { + "epoch": 0.06579453636013852, + "grad_norm": 2.762643578362577, + "learning_rate": 1.9932918339555965e-05, + "loss": 1.0857, + "step": 684 + }, + { + "epoch": 0.0658907272027703, + "grad_norm": 2.635230641881222, + "learning_rate": 1.9932557607199023e-05, + "loss": 1.0632, + "step": 685 + }, + { + "epoch": 0.06598691804540208, + "grad_norm": 3.4161326802785923, + "learning_rate": 1.9932195910801848e-05, + "loss": 1.0847, + "step": 686 + }, + { + "epoch": 0.06608310888803386, + "grad_norm": 2.9982744036295386, + "learning_rate": 1.9931833250399543e-05, + "loss": 1.0352, + "step": 687 + }, + { + "epoch": 0.06617929973066564, + "grad_norm": 2.9315587854908034, + "learning_rate": 1.9931469626027305e-05, + "loss": 1.1824, + "step": 688 + }, + { + "epoch": 0.06627549057329742, + "grad_norm": 2.4644918809909315, + "learning_rate": 1.993110503772043e-05, + "loss": 1.1435, + "step": 689 + }, + { + "epoch": 0.06637168141592921, + "grad_norm": 2.95711636188166, + "learning_rate": 1.9930739485514304e-05, + "loss": 1.06, + "step": 690 + }, + { + "epoch": 0.06646787225856099, + "grad_norm": 2.896544630401872, + "learning_rate": 1.9930372969444405e-05, + "loss": 0.9995, + "step": 691 + }, + { + "epoch": 0.06656406310119277, + "grad_norm": 2.7348152232479688, + "learning_rate": 1.9930005489546308e-05, + "loss": 1.1544, + "step": 692 + }, + { + "epoch": 0.06666025394382455, + "grad_norm": 2.548764254703571, + "learning_rate": 1.9929637045855678e-05, + "loss": 1.0427, + "step": 693 + }, + { + "epoch": 0.06675644478645633, + "grad_norm": 2.30705365539651, + "learning_rate": 1.9929267638408277e-05, + "loss": 1.01, + "step": 694 + }, + { + "epoch": 0.06685263562908811, + "grad_norm": 3.312931497741069, + "learning_rate": 1.9928897267239962e-05, + "loss": 1.0638, + "step": 695 + }, + { + "epoch": 0.06694882647171989, + "grad_norm": 3.302466741822662, + "learning_rate": 1.9928525932386678e-05, + "loss": 1.0628, + "step": 696 + }, + { + "epoch": 0.06704501731435167, + "grad_norm": 3.1489330825590316, + "learning_rate": 1.992815363388447e-05, + "loss": 0.8733, + "step": 697 + }, + { + "epoch": 0.06714120815698346, + "grad_norm": 3.184415648360582, + "learning_rate": 1.9927780371769463e-05, + "loss": 1.0413, + "step": 698 + }, + { + "epoch": 0.06723739899961524, + "grad_norm": 3.169888633533238, + "learning_rate": 1.9927406146077894e-05, + "loss": 1.1985, + "step": 699 + }, + { + "epoch": 0.06733358984224702, + "grad_norm": 2.239702799489659, + "learning_rate": 1.9927030956846083e-05, + "loss": 0.8026, + "step": 700 + }, + { + "epoch": 0.0674297806848788, + "grad_norm": 2.412514691943076, + "learning_rate": 1.9926654804110445e-05, + "loss": 1.1162, + "step": 701 + }, + { + "epoch": 0.06752597152751058, + "grad_norm": 3.1066270336879445, + "learning_rate": 1.992627768790749e-05, + "loss": 1.288, + "step": 702 + }, + { + "epoch": 0.06762216237014236, + "grad_norm": 3.5902331571520567, + "learning_rate": 1.9925899608273816e-05, + "loss": 1.1107, + "step": 703 + }, + { + "epoch": 0.06771835321277414, + "grad_norm": 2.3720608871161155, + "learning_rate": 1.9925520565246125e-05, + "loss": 0.8482, + "step": 704 + }, + { + "epoch": 0.06781454405540592, + "grad_norm": 3.003041113561195, + "learning_rate": 1.99251405588612e-05, + "loss": 1.1524, + "step": 705 + }, + { + "epoch": 0.0679107348980377, + "grad_norm": 2.687073111483425, + "learning_rate": 1.9924759589155932e-05, + "loss": 1.1381, + "step": 706 + }, + { + "epoch": 0.0680069257406695, + "grad_norm": 2.244658115431135, + "learning_rate": 1.9924377656167285e-05, + "loss": 0.8251, + "step": 707 + }, + { + "epoch": 0.06810311658330127, + "grad_norm": 2.9540583701348995, + "learning_rate": 1.9923994759932344e-05, + "loss": 1.1091, + "step": 708 + }, + { + "epoch": 0.06819930742593305, + "grad_norm": 2.9126767427385594, + "learning_rate": 1.9923610900488262e-05, + "loss": 1.2649, + "step": 709 + }, + { + "epoch": 0.06829549826856483, + "grad_norm": 2.570633345325202, + "learning_rate": 1.9923226077872296e-05, + "loss": 1.0878, + "step": 710 + }, + { + "epoch": 0.06839168911119661, + "grad_norm": 3.5021726765473074, + "learning_rate": 1.9922840292121807e-05, + "loss": 1.1493, + "step": 711 + }, + { + "epoch": 0.06848787995382839, + "grad_norm": 2.18652429143542, + "learning_rate": 1.9922453543274223e-05, + "loss": 0.8244, + "step": 712 + }, + { + "epoch": 0.06858407079646017, + "grad_norm": 3.012099595965582, + "learning_rate": 1.992206583136709e-05, + "loss": 1.2001, + "step": 713 + }, + { + "epoch": 0.06868026163909195, + "grad_norm": 2.8858309246722516, + "learning_rate": 1.9921677156438044e-05, + "loss": 1.1363, + "step": 714 + }, + { + "epoch": 0.06877645248172375, + "grad_norm": 3.2011645678483975, + "learning_rate": 1.9921287518524798e-05, + "loss": 1.135, + "step": 715 + }, + { + "epoch": 0.06887264332435553, + "grad_norm": 2.380167823083304, + "learning_rate": 1.9920896917665178e-05, + "loss": 1.026, + "step": 716 + }, + { + "epoch": 0.0689688341669873, + "grad_norm": 2.8602605116865174, + "learning_rate": 1.992050535389709e-05, + "loss": 1.1897, + "step": 717 + }, + { + "epoch": 0.06906502500961909, + "grad_norm": 2.8830226830900907, + "learning_rate": 1.992011282725854e-05, + "loss": 1.06, + "step": 718 + }, + { + "epoch": 0.06916121585225087, + "grad_norm": 2.671675008864062, + "learning_rate": 1.991971933778763e-05, + "loss": 1.0453, + "step": 719 + }, + { + "epoch": 0.06925740669488265, + "grad_norm": 2.5227230169228676, + "learning_rate": 1.9919324885522548e-05, + "loss": 1.1059, + "step": 720 + }, + { + "epoch": 0.06935359753751442, + "grad_norm": 3.0744891645774266, + "learning_rate": 1.9918929470501577e-05, + "loss": 1.0817, + "step": 721 + }, + { + "epoch": 0.0694497883801462, + "grad_norm": 2.3228054560235294, + "learning_rate": 1.99185330927631e-05, + "loss": 1.1124, + "step": 722 + }, + { + "epoch": 0.069545979222778, + "grad_norm": 3.106499557475855, + "learning_rate": 1.9918135752345584e-05, + "loss": 1.1908, + "step": 723 + }, + { + "epoch": 0.06964217006540978, + "grad_norm": 2.8311677615430115, + "learning_rate": 1.99177374492876e-05, + "loss": 1.054, + "step": 724 + }, + { + "epoch": 0.06973836090804156, + "grad_norm": 2.5115089712611263, + "learning_rate": 1.9917338183627802e-05, + "loss": 1.139, + "step": 725 + }, + { + "epoch": 0.06983455175067334, + "grad_norm": 2.4670232904583718, + "learning_rate": 1.991693795540494e-05, + "loss": 1.1071, + "step": 726 + }, + { + "epoch": 0.06993074259330512, + "grad_norm": 3.02289447084436, + "learning_rate": 1.991653676465787e-05, + "loss": 1.1461, + "step": 727 + }, + { + "epoch": 0.0700269334359369, + "grad_norm": 3.2940354216638164, + "learning_rate": 1.9916134611425522e-05, + "loss": 1.0963, + "step": 728 + }, + { + "epoch": 0.07012312427856868, + "grad_norm": 3.279697762135789, + "learning_rate": 1.991573149574693e-05, + "loss": 1.2579, + "step": 729 + }, + { + "epoch": 0.07021931512120046, + "grad_norm": 2.3843522628063347, + "learning_rate": 1.9915327417661226e-05, + "loss": 0.9165, + "step": 730 + }, + { + "epoch": 0.07031550596383224, + "grad_norm": 2.833389376285401, + "learning_rate": 1.991492237720762e-05, + "loss": 1.0284, + "step": 731 + }, + { + "epoch": 0.07041169680646403, + "grad_norm": 2.6661573022464253, + "learning_rate": 1.991451637442543e-05, + "loss": 1.0998, + "step": 732 + }, + { + "epoch": 0.07050788764909581, + "grad_norm": 3.7132508076286697, + "learning_rate": 1.991410940935406e-05, + "loss": 1.202, + "step": 733 + }, + { + "epoch": 0.07060407849172759, + "grad_norm": 2.462629629230357, + "learning_rate": 1.9913701482033008e-05, + "loss": 1.0708, + "step": 734 + }, + { + "epoch": 0.07070026933435937, + "grad_norm": 2.7328792060136013, + "learning_rate": 1.991329259250187e-05, + "loss": 1.0478, + "step": 735 + }, + { + "epoch": 0.07079646017699115, + "grad_norm": 3.1380079340814633, + "learning_rate": 1.9912882740800336e-05, + "loss": 1.0817, + "step": 736 + }, + { + "epoch": 0.07089265101962293, + "grad_norm": 2.8374204498052187, + "learning_rate": 1.991247192696818e-05, + "loss": 1.2339, + "step": 737 + }, + { + "epoch": 0.07098884186225471, + "grad_norm": 2.6242922557152055, + "learning_rate": 1.9912060151045273e-05, + "loss": 1.0042, + "step": 738 + }, + { + "epoch": 0.07108503270488649, + "grad_norm": 2.8630280883838353, + "learning_rate": 1.9911647413071584e-05, + "loss": 1.0752, + "step": 739 + }, + { + "epoch": 0.07118122354751828, + "grad_norm": 2.1798521613632094, + "learning_rate": 1.9911233713087172e-05, + "loss": 0.7969, + "step": 740 + }, + { + "epoch": 0.07127741439015006, + "grad_norm": 2.8362457454485646, + "learning_rate": 1.9910819051132195e-05, + "loss": 1.1407, + "step": 741 + }, + { + "epoch": 0.07137360523278184, + "grad_norm": 2.6752185963314923, + "learning_rate": 1.9910403427246895e-05, + "loss": 0.9935, + "step": 742 + }, + { + "epoch": 0.07146979607541362, + "grad_norm": 2.6160862401340945, + "learning_rate": 1.9909986841471613e-05, + "loss": 1.1909, + "step": 743 + }, + { + "epoch": 0.0715659869180454, + "grad_norm": 3.1323650247626738, + "learning_rate": 1.990956929384678e-05, + "loss": 1.141, + "step": 744 + }, + { + "epoch": 0.07166217776067718, + "grad_norm": 2.691662171131115, + "learning_rate": 1.9909150784412925e-05, + "loss": 1.2299, + "step": 745 + }, + { + "epoch": 0.07175836860330896, + "grad_norm": 2.9649176691888, + "learning_rate": 1.990873131321067e-05, + "loss": 1.1183, + "step": 746 + }, + { + "epoch": 0.07185455944594074, + "grad_norm": 2.693757262933104, + "learning_rate": 1.9908310880280723e-05, + "loss": 1.2287, + "step": 747 + }, + { + "epoch": 0.07195075028857253, + "grad_norm": 2.35007801112357, + "learning_rate": 1.9907889485663897e-05, + "loss": 1.0238, + "step": 748 + }, + { + "epoch": 0.07204694113120431, + "grad_norm": 2.686738321643009, + "learning_rate": 1.9907467129401086e-05, + "loss": 1.0794, + "step": 749 + }, + { + "epoch": 0.0721431319738361, + "grad_norm": 2.7049922324991167, + "learning_rate": 1.9907043811533283e-05, + "loss": 1.0088, + "step": 750 + }, + { + "epoch": 0.07223932281646787, + "grad_norm": 2.961008548895737, + "learning_rate": 1.9906619532101583e-05, + "loss": 1.1788, + "step": 751 + }, + { + "epoch": 0.07233551365909965, + "grad_norm": 2.638899467145543, + "learning_rate": 1.9906194291147155e-05, + "loss": 1.0734, + "step": 752 + }, + { + "epoch": 0.07243170450173143, + "grad_norm": 2.5464230027526025, + "learning_rate": 1.9905768088711283e-05, + "loss": 1.1047, + "step": 753 + }, + { + "epoch": 0.07252789534436321, + "grad_norm": 2.7089573094198105, + "learning_rate": 1.9905340924835322e-05, + "loss": 1.1283, + "step": 754 + }, + { + "epoch": 0.072624086186995, + "grad_norm": 2.516193353886799, + "learning_rate": 1.9904912799560744e-05, + "loss": 1.024, + "step": 755 + }, + { + "epoch": 0.07272027702962677, + "grad_norm": 2.9117800705396713, + "learning_rate": 1.9904483712929094e-05, + "loss": 1.0937, + "step": 756 + }, + { + "epoch": 0.07281646787225857, + "grad_norm": 2.9220258872370612, + "learning_rate": 1.9904053664982022e-05, + "loss": 1.0376, + "step": 757 + }, + { + "epoch": 0.07291265871489035, + "grad_norm": 3.037250744140794, + "learning_rate": 1.9903622655761267e-05, + "loss": 1.0983, + "step": 758 + }, + { + "epoch": 0.07300884955752213, + "grad_norm": 2.40307276694605, + "learning_rate": 1.990319068530866e-05, + "loss": 0.8843, + "step": 759 + }, + { + "epoch": 0.0731050404001539, + "grad_norm": 2.699306770901902, + "learning_rate": 1.990275775366613e-05, + "loss": 0.943, + "step": 760 + }, + { + "epoch": 0.07320123124278569, + "grad_norm": 2.601715963998991, + "learning_rate": 1.9902323860875695e-05, + "loss": 1.2266, + "step": 761 + }, + { + "epoch": 0.07329742208541747, + "grad_norm": 2.4900138222309387, + "learning_rate": 1.9901889006979473e-05, + "loss": 1.0909, + "step": 762 + }, + { + "epoch": 0.07339361292804925, + "grad_norm": 2.5205089766974282, + "learning_rate": 1.9901453192019663e-05, + "loss": 1.0674, + "step": 763 + }, + { + "epoch": 0.07348980377068103, + "grad_norm": 2.8543989356304444, + "learning_rate": 1.990101641603857e-05, + "loss": 1.1135, + "step": 764 + }, + { + "epoch": 0.07358599461331282, + "grad_norm": 2.8546684856479643, + "learning_rate": 1.9900578679078583e-05, + "loss": 1.1634, + "step": 765 + }, + { + "epoch": 0.0736821854559446, + "grad_norm": 2.7208415920508284, + "learning_rate": 1.9900139981182193e-05, + "loss": 1.0711, + "step": 766 + }, + { + "epoch": 0.07377837629857638, + "grad_norm": 2.595994433052512, + "learning_rate": 1.9899700322391977e-05, + "loss": 1.1261, + "step": 767 + }, + { + "epoch": 0.07387456714120816, + "grad_norm": 2.841848668693105, + "learning_rate": 1.9899259702750604e-05, + "loss": 1.1289, + "step": 768 + }, + { + "epoch": 0.07397075798383994, + "grad_norm": 2.3917071452916554, + "learning_rate": 1.9898818122300847e-05, + "loss": 1.1342, + "step": 769 + }, + { + "epoch": 0.07406694882647172, + "grad_norm": 2.5365985751321363, + "learning_rate": 1.9898375581085555e-05, + "loss": 1.0689, + "step": 770 + }, + { + "epoch": 0.0741631396691035, + "grad_norm": 3.7434049731046297, + "learning_rate": 1.9897932079147693e-05, + "loss": 1.0681, + "step": 771 + }, + { + "epoch": 0.07425933051173528, + "grad_norm": 2.4179502786077136, + "learning_rate": 1.9897487616530296e-05, + "loss": 1.027, + "step": 772 + }, + { + "epoch": 0.07435552135436706, + "grad_norm": 2.7774257249247354, + "learning_rate": 1.989704219327651e-05, + "loss": 1.1573, + "step": 773 + }, + { + "epoch": 0.07445171219699885, + "grad_norm": 2.9528085656247502, + "learning_rate": 1.9896595809429565e-05, + "loss": 1.0572, + "step": 774 + }, + { + "epoch": 0.07454790303963063, + "grad_norm": 3.2625556708895425, + "learning_rate": 1.9896148465032786e-05, + "loss": 0.8643, + "step": 775 + }, + { + "epoch": 0.07464409388226241, + "grad_norm": 2.8629964744496963, + "learning_rate": 1.9895700160129593e-05, + "loss": 1.189, + "step": 776 + }, + { + "epoch": 0.07474028472489419, + "grad_norm": 3.104957897108114, + "learning_rate": 1.989525089476349e-05, + "loss": 1.0264, + "step": 777 + }, + { + "epoch": 0.07483647556752597, + "grad_norm": 2.9014719758870338, + "learning_rate": 1.9894800668978095e-05, + "loss": 1.0973, + "step": 778 + }, + { + "epoch": 0.07493266641015775, + "grad_norm": 2.5845190242319265, + "learning_rate": 1.98943494828171e-05, + "loss": 1.0842, + "step": 779 + }, + { + "epoch": 0.07502885725278953, + "grad_norm": 2.8124512745287267, + "learning_rate": 1.9893897336324292e-05, + "loss": 1.0188, + "step": 780 + }, + { + "epoch": 0.07512504809542131, + "grad_norm": 2.7724372269864315, + "learning_rate": 1.9893444229543562e-05, + "loss": 1.0836, + "step": 781 + }, + { + "epoch": 0.0752212389380531, + "grad_norm": 2.8572681303932983, + "learning_rate": 1.9892990162518884e-05, + "loss": 1.1668, + "step": 782 + }, + { + "epoch": 0.07531742978068488, + "grad_norm": 2.9546153160956092, + "learning_rate": 1.9892535135294332e-05, + "loss": 1.0982, + "step": 783 + }, + { + "epoch": 0.07541362062331666, + "grad_norm": 2.6425089679581952, + "learning_rate": 1.9892079147914072e-05, + "loss": 1.1483, + "step": 784 + }, + { + "epoch": 0.07550981146594844, + "grad_norm": 2.8899918676330367, + "learning_rate": 1.9891622200422356e-05, + "loss": 1.2306, + "step": 785 + }, + { + "epoch": 0.07560600230858022, + "grad_norm": 2.9659153589775795, + "learning_rate": 1.9891164292863537e-05, + "loss": 1.1152, + "step": 786 + }, + { + "epoch": 0.075702193151212, + "grad_norm": 2.774964547034143, + "learning_rate": 1.9890705425282062e-05, + "loss": 1.1976, + "step": 787 + }, + { + "epoch": 0.07579838399384378, + "grad_norm": 2.7248768382389836, + "learning_rate": 1.9890245597722465e-05, + "loss": 1.0388, + "step": 788 + }, + { + "epoch": 0.07589457483647556, + "grad_norm": 2.195264234392313, + "learning_rate": 1.9889784810229376e-05, + "loss": 0.8031, + "step": 789 + }, + { + "epoch": 0.07599076567910736, + "grad_norm": 2.8059605979686926, + "learning_rate": 1.9889323062847516e-05, + "loss": 1.1215, + "step": 790 + }, + { + "epoch": 0.07608695652173914, + "grad_norm": 2.748470343639403, + "learning_rate": 1.988886035562171e-05, + "loss": 1.1171, + "step": 791 + }, + { + "epoch": 0.07618314736437092, + "grad_norm": 2.61142684792966, + "learning_rate": 1.988839668859686e-05, + "loss": 1.1192, + "step": 792 + }, + { + "epoch": 0.0762793382070027, + "grad_norm": 2.948369124149331, + "learning_rate": 1.988793206181797e-05, + "loss": 1.0547, + "step": 793 + }, + { + "epoch": 0.07637552904963447, + "grad_norm": 2.6786545820412186, + "learning_rate": 1.988746647533014e-05, + "loss": 1.1413, + "step": 794 + }, + { + "epoch": 0.07647171989226625, + "grad_norm": 2.59350341136655, + "learning_rate": 1.9886999929178553e-05, + "loss": 1.0201, + "step": 795 + }, + { + "epoch": 0.07656791073489803, + "grad_norm": 2.9035352559104766, + "learning_rate": 1.9886532423408495e-05, + "loss": 1.0975, + "step": 796 + }, + { + "epoch": 0.07666410157752981, + "grad_norm": 2.5999529449196945, + "learning_rate": 1.988606395806534e-05, + "loss": 1.0955, + "step": 797 + }, + { + "epoch": 0.0767602924201616, + "grad_norm": 2.757421521755794, + "learning_rate": 1.9885594533194564e-05, + "loss": 1.1283, + "step": 798 + }, + { + "epoch": 0.07685648326279339, + "grad_norm": 2.6377622720038283, + "learning_rate": 1.988512414884172e-05, + "loss": 1.1374, + "step": 799 + }, + { + "epoch": 0.07695267410542517, + "grad_norm": 2.7781608844552257, + "learning_rate": 1.9884652805052465e-05, + "loss": 1.2116, + "step": 800 + }, + { + "epoch": 0.07704886494805695, + "grad_norm": 2.736718525316454, + "learning_rate": 1.9884180501872544e-05, + "loss": 0.9545, + "step": 801 + }, + { + "epoch": 0.07714505579068873, + "grad_norm": 2.612821043288467, + "learning_rate": 1.9883707239347804e-05, + "loss": 1.2156, + "step": 802 + }, + { + "epoch": 0.0772412466333205, + "grad_norm": 2.325974429457091, + "learning_rate": 1.9883233017524176e-05, + "loss": 1.1012, + "step": 803 + }, + { + "epoch": 0.07733743747595229, + "grad_norm": 2.6577478978687212, + "learning_rate": 1.988275783644769e-05, + "loss": 1.1223, + "step": 804 + }, + { + "epoch": 0.07743362831858407, + "grad_norm": 2.4388307763854855, + "learning_rate": 1.9882281696164466e-05, + "loss": 1.0802, + "step": 805 + }, + { + "epoch": 0.07752981916121585, + "grad_norm": 2.9036875524434786, + "learning_rate": 1.988180459672071e-05, + "loss": 1.0604, + "step": 806 + }, + { + "epoch": 0.07762601000384764, + "grad_norm": 2.4950942153154045, + "learning_rate": 1.988132653816274e-05, + "loss": 1.0706, + "step": 807 + }, + { + "epoch": 0.07772220084647942, + "grad_norm": 2.7278877995165924, + "learning_rate": 1.988084752053695e-05, + "loss": 1.1847, + "step": 808 + }, + { + "epoch": 0.0778183916891112, + "grad_norm": 2.5482296612644344, + "learning_rate": 1.9880367543889827e-05, + "loss": 1.0699, + "step": 809 + }, + { + "epoch": 0.07791458253174298, + "grad_norm": 2.7627779033878435, + "learning_rate": 1.9879886608267967e-05, + "loss": 1.0319, + "step": 810 + }, + { + "epoch": 0.07801077337437476, + "grad_norm": 2.970314555149886, + "learning_rate": 1.9879404713718044e-05, + "loss": 1.1287, + "step": 811 + }, + { + "epoch": 0.07810696421700654, + "grad_norm": 2.7689651980283996, + "learning_rate": 1.9878921860286832e-05, + "loss": 1.1372, + "step": 812 + }, + { + "epoch": 0.07820315505963832, + "grad_norm": 2.6957457998086967, + "learning_rate": 1.987843804802119e-05, + "loss": 1.0968, + "step": 813 + }, + { + "epoch": 0.0782993459022701, + "grad_norm": 2.690682048163864, + "learning_rate": 1.9877953276968088e-05, + "loss": 1.1551, + "step": 814 + }, + { + "epoch": 0.07839553674490189, + "grad_norm": 2.505554607716093, + "learning_rate": 1.9877467547174567e-05, + "loss": 1.0146, + "step": 815 + }, + { + "epoch": 0.07849172758753367, + "grad_norm": 2.8346506059943346, + "learning_rate": 1.9876980858687777e-05, + "loss": 1.1629, + "step": 816 + }, + { + "epoch": 0.07858791843016545, + "grad_norm": 2.5592375034015005, + "learning_rate": 1.9876493211554948e-05, + "loss": 1.1289, + "step": 817 + }, + { + "epoch": 0.07868410927279723, + "grad_norm": 2.6132855292434103, + "learning_rate": 1.9876004605823417e-05, + "loss": 1.0628, + "step": 818 + }, + { + "epoch": 0.07878030011542901, + "grad_norm": 2.8352401344443274, + "learning_rate": 1.9875515041540607e-05, + "loss": 1.0216, + "step": 819 + }, + { + "epoch": 0.07887649095806079, + "grad_norm": 3.00508827733907, + "learning_rate": 1.987502451875403e-05, + "loss": 1.0673, + "step": 820 + }, + { + "epoch": 0.07897268180069257, + "grad_norm": 2.6639942843411526, + "learning_rate": 1.9874533037511302e-05, + "loss": 1.0697, + "step": 821 + }, + { + "epoch": 0.07906887264332435, + "grad_norm": 2.5067475101341796, + "learning_rate": 1.987404059786012e-05, + "loss": 1.198, + "step": 822 + }, + { + "epoch": 0.07916506348595613, + "grad_norm": 2.6755709156739385, + "learning_rate": 1.9873547199848283e-05, + "loss": 1.096, + "step": 823 + }, + { + "epoch": 0.07926125432858792, + "grad_norm": 2.837881074651402, + "learning_rate": 1.9873052843523676e-05, + "loss": 1.1014, + "step": 824 + }, + { + "epoch": 0.0793574451712197, + "grad_norm": 2.5446312155887525, + "learning_rate": 1.9872557528934283e-05, + "loss": 1.1553, + "step": 825 + }, + { + "epoch": 0.07945363601385148, + "grad_norm": 2.4460552661114856, + "learning_rate": 1.987206125612818e-05, + "loss": 1.0745, + "step": 826 + }, + { + "epoch": 0.07954982685648326, + "grad_norm": 2.8476520414467266, + "learning_rate": 1.987156402515353e-05, + "loss": 1.123, + "step": 827 + }, + { + "epoch": 0.07964601769911504, + "grad_norm": 2.887251759076321, + "learning_rate": 1.98710658360586e-05, + "loss": 1.1575, + "step": 828 + }, + { + "epoch": 0.07974220854174682, + "grad_norm": 2.4868013922231924, + "learning_rate": 1.9870566688891736e-05, + "loss": 1.1734, + "step": 829 + }, + { + "epoch": 0.0798383993843786, + "grad_norm": 2.8464455559662585, + "learning_rate": 1.987006658370139e-05, + "loss": 1.0487, + "step": 830 + }, + { + "epoch": 0.07993459022701038, + "grad_norm": 2.9948679855066764, + "learning_rate": 1.98695655205361e-05, + "loss": 1.0615, + "step": 831 + }, + { + "epoch": 0.08003078106964218, + "grad_norm": 2.7579444489912235, + "learning_rate": 1.9869063499444495e-05, + "loss": 1.0592, + "step": 832 + }, + { + "epoch": 0.08012697191227396, + "grad_norm": 3.103096152321995, + "learning_rate": 1.9868560520475304e-05, + "loss": 1.2178, + "step": 833 + }, + { + "epoch": 0.08022316275490574, + "grad_norm": 2.1330091702832368, + "learning_rate": 1.9868056583677346e-05, + "loss": 0.8007, + "step": 834 + }, + { + "epoch": 0.08031935359753752, + "grad_norm": 2.53094701529311, + "learning_rate": 1.9867551689099533e-05, + "loss": 1.09, + "step": 835 + }, + { + "epoch": 0.0804155444401693, + "grad_norm": 2.862644757714171, + "learning_rate": 1.9867045836790867e-05, + "loss": 1.034, + "step": 836 + }, + { + "epoch": 0.08051173528280108, + "grad_norm": 2.706645693400606, + "learning_rate": 1.9866539026800448e-05, + "loss": 1.0714, + "step": 837 + }, + { + "epoch": 0.08060792612543285, + "grad_norm": 2.7541308757415144, + "learning_rate": 1.9866031259177463e-05, + "loss": 1.0542, + "step": 838 + }, + { + "epoch": 0.08070411696806463, + "grad_norm": 2.520367197920566, + "learning_rate": 1.9865522533971195e-05, + "loss": 1.1634, + "step": 839 + }, + { + "epoch": 0.08080030781069643, + "grad_norm": 2.1879682739992825, + "learning_rate": 1.9865012851231022e-05, + "loss": 0.7823, + "step": 840 + }, + { + "epoch": 0.08089649865332821, + "grad_norm": 2.705738355670816, + "learning_rate": 1.9864502211006415e-05, + "loss": 1.0821, + "step": 841 + }, + { + "epoch": 0.08099268949595999, + "grad_norm": 2.8387722705918526, + "learning_rate": 1.9863990613346936e-05, + "loss": 1.0073, + "step": 842 + }, + { + "epoch": 0.08108888033859177, + "grad_norm": 2.759405444952385, + "learning_rate": 1.9863478058302234e-05, + "loss": 1.1116, + "step": 843 + }, + { + "epoch": 0.08118507118122355, + "grad_norm": 3.176396196528279, + "learning_rate": 1.986296454592206e-05, + "loss": 1.1187, + "step": 844 + }, + { + "epoch": 0.08128126202385533, + "grad_norm": 2.662067163042779, + "learning_rate": 1.986245007625626e-05, + "loss": 1.1403, + "step": 845 + }, + { + "epoch": 0.08137745286648711, + "grad_norm": 2.8614186425632777, + "learning_rate": 1.9861934649354763e-05, + "loss": 1.1984, + "step": 846 + }, + { + "epoch": 0.08147364370911889, + "grad_norm": 2.9735265002896685, + "learning_rate": 1.986141826526759e-05, + "loss": 1.1094, + "step": 847 + }, + { + "epoch": 0.08156983455175067, + "grad_norm": 2.794280974316224, + "learning_rate": 1.9860900924044873e-05, + "loss": 1.0787, + "step": 848 + }, + { + "epoch": 0.08166602539438246, + "grad_norm": 2.5444544504014734, + "learning_rate": 1.9860382625736816e-05, + "loss": 1.1125, + "step": 849 + }, + { + "epoch": 0.08176221623701424, + "grad_norm": 2.796440744383389, + "learning_rate": 1.9859863370393726e-05, + "loss": 1.1002, + "step": 850 + }, + { + "epoch": 0.08185840707964602, + "grad_norm": 2.1040126102832084, + "learning_rate": 1.9859343158066e-05, + "loss": 0.8209, + "step": 851 + }, + { + "epoch": 0.0819545979222778, + "grad_norm": 2.7963018504355577, + "learning_rate": 1.9858821988804132e-05, + "loss": 1.0992, + "step": 852 + }, + { + "epoch": 0.08205078876490958, + "grad_norm": 2.7279472914364393, + "learning_rate": 1.9858299862658706e-05, + "loss": 1.0138, + "step": 853 + }, + { + "epoch": 0.08214697960754136, + "grad_norm": 3.036584287151629, + "learning_rate": 1.9857776779680393e-05, + "loss": 1.1666, + "step": 854 + }, + { + "epoch": 0.08224317045017314, + "grad_norm": 3.2406627267667703, + "learning_rate": 1.9857252739919972e-05, + "loss": 1.0995, + "step": 855 + }, + { + "epoch": 0.08233936129280492, + "grad_norm": 2.76899629729338, + "learning_rate": 1.98567277434283e-05, + "loss": 1.0826, + "step": 856 + }, + { + "epoch": 0.08243555213543671, + "grad_norm": 2.4261153436538, + "learning_rate": 1.985620179025633e-05, + "loss": 1.1257, + "step": 857 + }, + { + "epoch": 0.08253174297806849, + "grad_norm": 2.5573210925677063, + "learning_rate": 1.9855674880455115e-05, + "loss": 1.1376, + "step": 858 + }, + { + "epoch": 0.08262793382070027, + "grad_norm": 2.80969583479713, + "learning_rate": 1.985514701407579e-05, + "loss": 1.2248, + "step": 859 + }, + { + "epoch": 0.08272412466333205, + "grad_norm": 3.0918813565713887, + "learning_rate": 1.98546181911696e-05, + "loss": 1.1107, + "step": 860 + }, + { + "epoch": 0.08282031550596383, + "grad_norm": 2.8478402665027738, + "learning_rate": 1.9854088411787863e-05, + "loss": 1.1362, + "step": 861 + }, + { + "epoch": 0.08291650634859561, + "grad_norm": 2.6600658143003217, + "learning_rate": 1.9853557675982e-05, + "loss": 1.1353, + "step": 862 + }, + { + "epoch": 0.08301269719122739, + "grad_norm": 2.793188871070118, + "learning_rate": 1.9853025983803523e-05, + "loss": 1.0934, + "step": 863 + }, + { + "epoch": 0.08310888803385917, + "grad_norm": 2.7051996975138413, + "learning_rate": 1.985249333530404e-05, + "loss": 1.1573, + "step": 864 + }, + { + "epoch": 0.08320507887649096, + "grad_norm": 3.0739240837454074, + "learning_rate": 1.985195973053525e-05, + "loss": 1.2235, + "step": 865 + }, + { + "epoch": 0.08330126971912274, + "grad_norm": 2.2450388497443643, + "learning_rate": 1.9851425169548938e-05, + "loss": 0.8647, + "step": 866 + }, + { + "epoch": 0.08339746056175452, + "grad_norm": 2.5483581765524255, + "learning_rate": 1.9850889652396993e-05, + "loss": 1.0992, + "step": 867 + }, + { + "epoch": 0.0834936514043863, + "grad_norm": 2.6803261899868014, + "learning_rate": 1.9850353179131392e-05, + "loss": 1.0731, + "step": 868 + }, + { + "epoch": 0.08358984224701808, + "grad_norm": 2.484552867774405, + "learning_rate": 1.9849815749804202e-05, + "loss": 1.0851, + "step": 869 + }, + { + "epoch": 0.08368603308964986, + "grad_norm": 2.6417687802431704, + "learning_rate": 1.9849277364467585e-05, + "loss": 1.1286, + "step": 870 + }, + { + "epoch": 0.08378222393228164, + "grad_norm": 2.374536304603201, + "learning_rate": 1.9848738023173794e-05, + "loss": 1.151, + "step": 871 + }, + { + "epoch": 0.08387841477491342, + "grad_norm": 2.4917952227438565, + "learning_rate": 1.984819772597518e-05, + "loss": 1.1719, + "step": 872 + }, + { + "epoch": 0.0839746056175452, + "grad_norm": 2.454765495875368, + "learning_rate": 1.9847656472924183e-05, + "loss": 1.1087, + "step": 873 + }, + { + "epoch": 0.084070796460177, + "grad_norm": 2.783952444707475, + "learning_rate": 1.9847114264073336e-05, + "loss": 1.0694, + "step": 874 + }, + { + "epoch": 0.08416698730280878, + "grad_norm": 3.328271095715426, + "learning_rate": 1.9846571099475266e-05, + "loss": 1.0485, + "step": 875 + }, + { + "epoch": 0.08426317814544056, + "grad_norm": 3.1921333404492027, + "learning_rate": 1.984602697918269e-05, + "loss": 1.1793, + "step": 876 + }, + { + "epoch": 0.08435936898807234, + "grad_norm": 2.335832594451702, + "learning_rate": 1.984548190324842e-05, + "loss": 1.1049, + "step": 877 + }, + { + "epoch": 0.08445555983070412, + "grad_norm": 2.5095942140203196, + "learning_rate": 1.9844935871725363e-05, + "loss": 1.1749, + "step": 878 + }, + { + "epoch": 0.0845517506733359, + "grad_norm": 2.691788495284311, + "learning_rate": 1.984438888466651e-05, + "loss": 1.0738, + "step": 879 + }, + { + "epoch": 0.08464794151596768, + "grad_norm": 2.627691761396076, + "learning_rate": 1.9843840942124956e-05, + "loss": 1.1043, + "step": 880 + }, + { + "epoch": 0.08474413235859946, + "grad_norm": 2.543997320687101, + "learning_rate": 1.9843292044153884e-05, + "loss": 1.0549, + "step": 881 + }, + { + "epoch": 0.08484032320123125, + "grad_norm": 2.5545501090444667, + "learning_rate": 1.9842742190806566e-05, + "loss": 1.1682, + "step": 882 + }, + { + "epoch": 0.08493651404386303, + "grad_norm": 2.6211765272066914, + "learning_rate": 1.984219138213637e-05, + "loss": 1.1256, + "step": 883 + }, + { + "epoch": 0.08503270488649481, + "grad_norm": 2.439354837394179, + "learning_rate": 1.984163961819676e-05, + "loss": 1.0435, + "step": 884 + }, + { + "epoch": 0.08512889572912659, + "grad_norm": 2.502318587070632, + "learning_rate": 1.9841086899041286e-05, + "loss": 0.9932, + "step": 885 + }, + { + "epoch": 0.08522508657175837, + "grad_norm": 2.287650055273837, + "learning_rate": 1.9840533224723595e-05, + "loss": 1.1045, + "step": 886 + }, + { + "epoch": 0.08532127741439015, + "grad_norm": 2.55447648454598, + "learning_rate": 1.9839978595297428e-05, + "loss": 1.0761, + "step": 887 + }, + { + "epoch": 0.08541746825702193, + "grad_norm": 2.7590438636946333, + "learning_rate": 1.9839423010816616e-05, + "loss": 1.1398, + "step": 888 + }, + { + "epoch": 0.08551365909965371, + "grad_norm": 2.791554067669648, + "learning_rate": 1.983886647133508e-05, + "loss": 1.1183, + "step": 889 + }, + { + "epoch": 0.08560984994228549, + "grad_norm": 2.9529128541807625, + "learning_rate": 1.983830897690684e-05, + "loss": 1.1738, + "step": 890 + }, + { + "epoch": 0.08570604078491728, + "grad_norm": 2.659934993622678, + "learning_rate": 1.9837750527586006e-05, + "loss": 1.183, + "step": 891 + }, + { + "epoch": 0.08580223162754906, + "grad_norm": 2.786403352499133, + "learning_rate": 1.9837191123426777e-05, + "loss": 0.9602, + "step": 892 + }, + { + "epoch": 0.08589842247018084, + "grad_norm": 2.636597588585077, + "learning_rate": 1.9836630764483452e-05, + "loss": 1.0856, + "step": 893 + }, + { + "epoch": 0.08599461331281262, + "grad_norm": 2.5935002041408204, + "learning_rate": 1.983606945081042e-05, + "loss": 1.0793, + "step": 894 + }, + { + "epoch": 0.0860908041554444, + "grad_norm": 2.878802144358279, + "learning_rate": 1.983550718246215e-05, + "loss": 1.1522, + "step": 895 + }, + { + "epoch": 0.08618699499807618, + "grad_norm": 2.515201307935657, + "learning_rate": 1.983494395949323e-05, + "loss": 1.0784, + "step": 896 + }, + { + "epoch": 0.08628318584070796, + "grad_norm": 2.5983702032247327, + "learning_rate": 1.9834379781958317e-05, + "loss": 1.184, + "step": 897 + }, + { + "epoch": 0.08637937668333974, + "grad_norm": 2.792545756259017, + "learning_rate": 1.983381464991217e-05, + "loss": 1.0513, + "step": 898 + }, + { + "epoch": 0.08647556752597153, + "grad_norm": 3.4372911540873887, + "learning_rate": 1.9833248563409643e-05, + "loss": 1.2035, + "step": 899 + }, + { + "epoch": 0.08657175836860331, + "grad_norm": 2.4063430054780754, + "learning_rate": 1.9832681522505676e-05, + "loss": 1.1342, + "step": 900 + }, + { + "epoch": 0.08666794921123509, + "grad_norm": 2.970831879917955, + "learning_rate": 1.9832113527255304e-05, + "loss": 1.0772, + "step": 901 + }, + { + "epoch": 0.08676414005386687, + "grad_norm": 2.586734782328061, + "learning_rate": 1.9831544577713663e-05, + "loss": 1.0611, + "step": 902 + }, + { + "epoch": 0.08686033089649865, + "grad_norm": 3.06657198925786, + "learning_rate": 1.983097467393597e-05, + "loss": 1.1618, + "step": 903 + }, + { + "epoch": 0.08695652173913043, + "grad_norm": 3.30692768667578, + "learning_rate": 1.983040381597754e-05, + "loss": 1.1404, + "step": 904 + }, + { + "epoch": 0.08705271258176221, + "grad_norm": 2.469510061186353, + "learning_rate": 1.9829832003893775e-05, + "loss": 1.02, + "step": 905 + }, + { + "epoch": 0.08714890342439399, + "grad_norm": 2.539805742720845, + "learning_rate": 1.982925923774018e-05, + "loss": 1.0618, + "step": 906 + }, + { + "epoch": 0.08724509426702579, + "grad_norm": 2.877598575041587, + "learning_rate": 1.9828685517572345e-05, + "loss": 0.9241, + "step": 907 + }, + { + "epoch": 0.08734128510965757, + "grad_norm": 2.5411177399402076, + "learning_rate": 1.9828110843445954e-05, + "loss": 1.0232, + "step": 908 + }, + { + "epoch": 0.08743747595228935, + "grad_norm": 2.20342865966654, + "learning_rate": 1.9827535215416787e-05, + "loss": 0.8466, + "step": 909 + }, + { + "epoch": 0.08753366679492113, + "grad_norm": 2.6521292888068744, + "learning_rate": 1.982695863354071e-05, + "loss": 1.1404, + "step": 910 + }, + { + "epoch": 0.0876298576375529, + "grad_norm": 2.242239180636538, + "learning_rate": 1.9826381097873683e-05, + "loss": 1.0174, + "step": 911 + }, + { + "epoch": 0.08772604848018468, + "grad_norm": 2.372013876903139, + "learning_rate": 1.9825802608471767e-05, + "loss": 0.8569, + "step": 912 + }, + { + "epoch": 0.08782223932281646, + "grad_norm": 2.7928312754664435, + "learning_rate": 1.9825223165391105e-05, + "loss": 1.2436, + "step": 913 + }, + { + "epoch": 0.08791843016544824, + "grad_norm": 3.1039226957817667, + "learning_rate": 1.982464276868794e-05, + "loss": 1.2331, + "step": 914 + }, + { + "epoch": 0.08801462100808002, + "grad_norm": 2.4454732729236293, + "learning_rate": 1.9824061418418605e-05, + "loss": 1.0916, + "step": 915 + }, + { + "epoch": 0.08811081185071182, + "grad_norm": 2.754647937327079, + "learning_rate": 1.982347911463952e-05, + "loss": 1.0464, + "step": 916 + }, + { + "epoch": 0.0882070026933436, + "grad_norm": 2.9330370056112844, + "learning_rate": 1.9822895857407206e-05, + "loss": 1.109, + "step": 917 + }, + { + "epoch": 0.08830319353597538, + "grad_norm": 2.2637576918771107, + "learning_rate": 1.9822311646778277e-05, + "loss": 1.0497, + "step": 918 + }, + { + "epoch": 0.08839938437860716, + "grad_norm": 2.528272059464997, + "learning_rate": 1.982172648280943e-05, + "loss": 1.1677, + "step": 919 + }, + { + "epoch": 0.08849557522123894, + "grad_norm": 2.3243045572805805, + "learning_rate": 1.982114036555746e-05, + "loss": 1.1636, + "step": 920 + }, + { + "epoch": 0.08859176606387072, + "grad_norm": 2.5904628985481515, + "learning_rate": 1.9820553295079254e-05, + "loss": 1.1379, + "step": 921 + }, + { + "epoch": 0.0886879569065025, + "grad_norm": 2.8459342352028076, + "learning_rate": 1.9819965271431797e-05, + "loss": 1.0777, + "step": 922 + }, + { + "epoch": 0.08878414774913428, + "grad_norm": 3.0613661740087155, + "learning_rate": 1.981937629467216e-05, + "loss": 1.1163, + "step": 923 + }, + { + "epoch": 0.08888033859176607, + "grad_norm": 2.689801511719151, + "learning_rate": 1.9818786364857506e-05, + "loss": 1.0504, + "step": 924 + }, + { + "epoch": 0.08897652943439785, + "grad_norm": 2.7589062831599525, + "learning_rate": 1.9818195482045097e-05, + "loss": 1.1022, + "step": 925 + }, + { + "epoch": 0.08907272027702963, + "grad_norm": 2.8040342218563916, + "learning_rate": 1.9817603646292278e-05, + "loss": 1.166, + "step": 926 + }, + { + "epoch": 0.08916891111966141, + "grad_norm": 2.86359695293593, + "learning_rate": 1.9817010857656496e-05, + "loss": 1.1219, + "step": 927 + }, + { + "epoch": 0.08926510196229319, + "grad_norm": 3.1883760846487164, + "learning_rate": 1.9816417116195287e-05, + "loss": 1.0902, + "step": 928 + }, + { + "epoch": 0.08936129280492497, + "grad_norm": 2.589557088428122, + "learning_rate": 1.9815822421966272e-05, + "loss": 1.0017, + "step": 929 + }, + { + "epoch": 0.08945748364755675, + "grad_norm": 3.1215368439151567, + "learning_rate": 1.9815226775027182e-05, + "loss": 1.0822, + "step": 930 + }, + { + "epoch": 0.08955367449018853, + "grad_norm": 2.864785519533366, + "learning_rate": 1.9814630175435818e-05, + "loss": 1.1276, + "step": 931 + }, + { + "epoch": 0.08964986533282032, + "grad_norm": 3.0658782425342945, + "learning_rate": 1.9814032623250093e-05, + "loss": 1.1143, + "step": 932 + }, + { + "epoch": 0.0897460561754521, + "grad_norm": 2.7480419028528766, + "learning_rate": 1.9813434118528e-05, + "loss": 1.0685, + "step": 933 + }, + { + "epoch": 0.08984224701808388, + "grad_norm": 2.5633117060589155, + "learning_rate": 1.9812834661327632e-05, + "loss": 0.9702, + "step": 934 + }, + { + "epoch": 0.08993843786071566, + "grad_norm": 2.7451090504605014, + "learning_rate": 1.981223425170717e-05, + "loss": 1.2593, + "step": 935 + }, + { + "epoch": 0.09003462870334744, + "grad_norm": 2.5087879074888004, + "learning_rate": 1.9811632889724888e-05, + "loss": 1.143, + "step": 936 + }, + { + "epoch": 0.09013081954597922, + "grad_norm": 2.5147396350192626, + "learning_rate": 1.9811030575439157e-05, + "loss": 1.1349, + "step": 937 + }, + { + "epoch": 0.090227010388611, + "grad_norm": 2.555879225917095, + "learning_rate": 1.9810427308908437e-05, + "loss": 1.0591, + "step": 938 + }, + { + "epoch": 0.09032320123124278, + "grad_norm": 2.6281082290448867, + "learning_rate": 1.9809823090191273e-05, + "loss": 1.1133, + "step": 939 + }, + { + "epoch": 0.09041939207387456, + "grad_norm": 2.4791833577857174, + "learning_rate": 1.9809217919346318e-05, + "loss": 1.089, + "step": 940 + }, + { + "epoch": 0.09051558291650635, + "grad_norm": 2.62746171220058, + "learning_rate": 1.9808611796432302e-05, + "loss": 1.103, + "step": 941 + }, + { + "epoch": 0.09061177375913813, + "grad_norm": 2.547917957368258, + "learning_rate": 1.980800472150806e-05, + "loss": 1.077, + "step": 942 + }, + { + "epoch": 0.09070796460176991, + "grad_norm": 2.488862311583742, + "learning_rate": 1.9807396694632515e-05, + "loss": 1.135, + "step": 943 + }, + { + "epoch": 0.0908041554444017, + "grad_norm": 2.8046541861589422, + "learning_rate": 1.9806787715864674e-05, + "loss": 1.1393, + "step": 944 + }, + { + "epoch": 0.09090034628703347, + "grad_norm": 2.931765190758823, + "learning_rate": 1.980617778526365e-05, + "loss": 1.0958, + "step": 945 + }, + { + "epoch": 0.09099653712966525, + "grad_norm": 2.666037865042453, + "learning_rate": 1.9805566902888637e-05, + "loss": 1.0915, + "step": 946 + }, + { + "epoch": 0.09109272797229703, + "grad_norm": 2.751425499659249, + "learning_rate": 1.980495506879893e-05, + "loss": 1.0949, + "step": 947 + }, + { + "epoch": 0.09118891881492881, + "grad_norm": 2.6498597851026027, + "learning_rate": 1.9804342283053916e-05, + "loss": 1.0956, + "step": 948 + }, + { + "epoch": 0.0912851096575606, + "grad_norm": 2.9436036983277685, + "learning_rate": 1.980372854571306e-05, + "loss": 1.0826, + "step": 949 + }, + { + "epoch": 0.09138130050019239, + "grad_norm": 2.8763844356314103, + "learning_rate": 1.980311385683594e-05, + "loss": 1.1532, + "step": 950 + }, + { + "epoch": 0.09147749134282417, + "grad_norm": 2.550906007508479, + "learning_rate": 1.9802498216482218e-05, + "loss": 1.0799, + "step": 951 + }, + { + "epoch": 0.09157368218545595, + "grad_norm": 2.891200125149618, + "learning_rate": 1.980188162471164e-05, + "loss": 1.1321, + "step": 952 + }, + { + "epoch": 0.09166987302808773, + "grad_norm": 2.7758802324195058, + "learning_rate": 1.9801264081584056e-05, + "loss": 1.1447, + "step": 953 + }, + { + "epoch": 0.0917660638707195, + "grad_norm": 2.866265702332936, + "learning_rate": 1.98006455871594e-05, + "loss": 1.012, + "step": 954 + }, + { + "epoch": 0.09186225471335129, + "grad_norm": 2.675391532143187, + "learning_rate": 1.980002614149771e-05, + "loss": 0.9343, + "step": 955 + }, + { + "epoch": 0.09195844555598306, + "grad_norm": 3.397969650028721, + "learning_rate": 1.97994057446591e-05, + "loss": 1.1532, + "step": 956 + }, + { + "epoch": 0.09205463639861486, + "grad_norm": 2.0907889664653685, + "learning_rate": 1.9798784396703792e-05, + "loss": 0.8323, + "step": 957 + }, + { + "epoch": 0.09215082724124664, + "grad_norm": 2.6096097548305326, + "learning_rate": 1.979816209769209e-05, + "loss": 1.147, + "step": 958 + }, + { + "epoch": 0.09224701808387842, + "grad_norm": 2.641676477273754, + "learning_rate": 1.979753884768439e-05, + "loss": 1.0387, + "step": 959 + }, + { + "epoch": 0.0923432089265102, + "grad_norm": 2.717980138168435, + "learning_rate": 1.9796914646741187e-05, + "loss": 1.1351, + "step": 960 + }, + { + "epoch": 0.09243939976914198, + "grad_norm": 2.2939913854892415, + "learning_rate": 1.9796289494923068e-05, + "loss": 0.8193, + "step": 961 + }, + { + "epoch": 0.09253559061177376, + "grad_norm": 2.798704538687323, + "learning_rate": 1.9795663392290702e-05, + "loss": 1.1221, + "step": 962 + }, + { + "epoch": 0.09263178145440554, + "grad_norm": 2.5239202581402935, + "learning_rate": 1.9795036338904863e-05, + "loss": 1.0291, + "step": 963 + }, + { + "epoch": 0.09272797229703732, + "grad_norm": 2.574310850680747, + "learning_rate": 1.9794408334826415e-05, + "loss": 1.0793, + "step": 964 + }, + { + "epoch": 0.0928241631396691, + "grad_norm": 2.422500507024019, + "learning_rate": 1.9793779380116304e-05, + "loss": 1.0436, + "step": 965 + }, + { + "epoch": 0.09292035398230089, + "grad_norm": 2.9204689769597363, + "learning_rate": 1.979314947483558e-05, + "loss": 1.2175, + "step": 966 + }, + { + "epoch": 0.09301654482493267, + "grad_norm": 2.6429968287259635, + "learning_rate": 1.9792518619045376e-05, + "loss": 1.0228, + "step": 967 + }, + { + "epoch": 0.09311273566756445, + "grad_norm": 2.8338084162941373, + "learning_rate": 1.9791886812806932e-05, + "loss": 1.0299, + "step": 968 + }, + { + "epoch": 0.09320892651019623, + "grad_norm": 3.05584507511113, + "learning_rate": 1.979125405618156e-05, + "loss": 1.1655, + "step": 969 + }, + { + "epoch": 0.09330511735282801, + "grad_norm": 2.416841316471866, + "learning_rate": 1.9790620349230676e-05, + "loss": 1.1054, + "step": 970 + }, + { + "epoch": 0.09340130819545979, + "grad_norm": 2.6226853975201827, + "learning_rate": 1.978998569201579e-05, + "loss": 1.1119, + "step": 971 + }, + { + "epoch": 0.09349749903809157, + "grad_norm": 2.6726844160994876, + "learning_rate": 1.9789350084598504e-05, + "loss": 1.1367, + "step": 972 + }, + { + "epoch": 0.09359368988072335, + "grad_norm": 2.4605063212555978, + "learning_rate": 1.97887135270405e-05, + "loss": 1.1573, + "step": 973 + }, + { + "epoch": 0.09368988072335514, + "grad_norm": 2.7398808368666474, + "learning_rate": 1.9788076019403565e-05, + "loss": 1.0922, + "step": 974 + }, + { + "epoch": 0.09378607156598692, + "grad_norm": 2.6395926485771426, + "learning_rate": 1.978743756174958e-05, + "loss": 1.1942, + "step": 975 + }, + { + "epoch": 0.0938822624086187, + "grad_norm": 2.677394640849634, + "learning_rate": 1.9786798154140507e-05, + "loss": 1.1499, + "step": 976 + }, + { + "epoch": 0.09397845325125048, + "grad_norm": 2.730632411112157, + "learning_rate": 1.9786157796638406e-05, + "loss": 1.0819, + "step": 977 + }, + { + "epoch": 0.09407464409388226, + "grad_norm": 2.481957952849485, + "learning_rate": 1.9785516489305437e-05, + "loss": 1.1522, + "step": 978 + }, + { + "epoch": 0.09417083493651404, + "grad_norm": 2.3684642706385883, + "learning_rate": 1.9784874232203833e-05, + "loss": 1.0828, + "step": 979 + }, + { + "epoch": 0.09426702577914582, + "grad_norm": 2.739252431741938, + "learning_rate": 1.9784231025395936e-05, + "loss": 1.0339, + "step": 980 + }, + { + "epoch": 0.0943632166217776, + "grad_norm": 2.735853637142669, + "learning_rate": 1.978358686894418e-05, + "loss": 1.1611, + "step": 981 + }, + { + "epoch": 0.0944594074644094, + "grad_norm": 2.5159735290238086, + "learning_rate": 1.9782941762911075e-05, + "loss": 1.1497, + "step": 982 + }, + { + "epoch": 0.09455559830704117, + "grad_norm": 2.960200806213733, + "learning_rate": 1.978229570735924e-05, + "loss": 1.0125, + "step": 983 + }, + { + "epoch": 0.09465178914967295, + "grad_norm": 2.3815779086869737, + "learning_rate": 1.9781648702351383e-05, + "loss": 1.1512, + "step": 984 + }, + { + "epoch": 0.09474797999230473, + "grad_norm": 2.825260213029565, + "learning_rate": 1.9781000747950295e-05, + "loss": 0.9831, + "step": 985 + }, + { + "epoch": 0.09484417083493651, + "grad_norm": 2.6531010868921463, + "learning_rate": 1.9780351844218874e-05, + "loss": 1.1712, + "step": 986 + }, + { + "epoch": 0.0949403616775683, + "grad_norm": 2.3314661355649324, + "learning_rate": 1.9779701991220092e-05, + "loss": 1.0771, + "step": 987 + }, + { + "epoch": 0.09503655252020007, + "grad_norm": 2.6162406469320367, + "learning_rate": 1.977905118901703e-05, + "loss": 1.073, + "step": 988 + }, + { + "epoch": 0.09513274336283185, + "grad_norm": 2.50865626372431, + "learning_rate": 1.9778399437672848e-05, + "loss": 1.2082, + "step": 989 + }, + { + "epoch": 0.09522893420546363, + "grad_norm": 2.726950334463593, + "learning_rate": 1.977774673725081e-05, + "loss": 1.1183, + "step": 990 + }, + { + "epoch": 0.09532512504809543, + "grad_norm": 2.7601940336944937, + "learning_rate": 1.9777093087814262e-05, + "loss": 1.054, + "step": 991 + }, + { + "epoch": 0.0954213158907272, + "grad_norm": 2.630032472173663, + "learning_rate": 1.977643848942665e-05, + "loss": 1.1297, + "step": 992 + }, + { + "epoch": 0.09551750673335899, + "grad_norm": 2.7561333016666443, + "learning_rate": 1.977578294215151e-05, + "loss": 1.0838, + "step": 993 + }, + { + "epoch": 0.09561369757599077, + "grad_norm": 2.69346259943373, + "learning_rate": 1.977512644605246e-05, + "loss": 1.0847, + "step": 994 + }, + { + "epoch": 0.09570988841862255, + "grad_norm": 2.9013064782039923, + "learning_rate": 1.9774469001193222e-05, + "loss": 0.9924, + "step": 995 + }, + { + "epoch": 0.09580607926125433, + "grad_norm": 2.677346505689912, + "learning_rate": 1.9773810607637612e-05, + "loss": 1.1175, + "step": 996 + }, + { + "epoch": 0.0959022701038861, + "grad_norm": 2.34438303835815, + "learning_rate": 1.977315126544953e-05, + "loss": 1.0722, + "step": 997 + }, + { + "epoch": 0.09599846094651789, + "grad_norm": 2.438671591019639, + "learning_rate": 1.9772490974692962e-05, + "loss": 1.0287, + "step": 998 + }, + { + "epoch": 0.09609465178914968, + "grad_norm": 2.6618748224482647, + "learning_rate": 1.977182973543201e-05, + "loss": 1.1817, + "step": 999 + }, + { + "epoch": 0.09619084263178146, + "grad_norm": 3.009464874145482, + "learning_rate": 1.9771167547730844e-05, + "loss": 1.1953, + "step": 1000 + }, + { + "epoch": 0.09628703347441324, + "grad_norm": 2.727471050900601, + "learning_rate": 1.977050441165374e-05, + "loss": 1.1435, + "step": 1001 + }, + { + "epoch": 0.09638322431704502, + "grad_norm": 2.321667825886464, + "learning_rate": 1.976984032726505e-05, + "loss": 1.0312, + "step": 1002 + }, + { + "epoch": 0.0964794151596768, + "grad_norm": 2.755612601388198, + "learning_rate": 1.9769175294629242e-05, + "loss": 1.1596, + "step": 1003 + }, + { + "epoch": 0.09657560600230858, + "grad_norm": 2.5881004391631883, + "learning_rate": 1.976850931381086e-05, + "loss": 1.0742, + "step": 1004 + }, + { + "epoch": 0.09667179684494036, + "grad_norm": 2.880420237617393, + "learning_rate": 1.9767842384874538e-05, + "loss": 1.1649, + "step": 1005 + }, + { + "epoch": 0.09676798768757214, + "grad_norm": 2.4492813401974507, + "learning_rate": 1.976717450788501e-05, + "loss": 1.0118, + "step": 1006 + }, + { + "epoch": 0.09686417853020392, + "grad_norm": 2.690671936852724, + "learning_rate": 1.97665056829071e-05, + "loss": 1.1278, + "step": 1007 + }, + { + "epoch": 0.09696036937283571, + "grad_norm": 2.585929014629608, + "learning_rate": 1.9765835910005726e-05, + "loss": 1.0771, + "step": 1008 + }, + { + "epoch": 0.09705656021546749, + "grad_norm": 2.7533079210879525, + "learning_rate": 1.9765165189245886e-05, + "loss": 0.9955, + "step": 1009 + }, + { + "epoch": 0.09715275105809927, + "grad_norm": 2.389119741675629, + "learning_rate": 1.9764493520692685e-05, + "loss": 0.8934, + "step": 1010 + }, + { + "epoch": 0.09724894190073105, + "grad_norm": 2.8723195215433344, + "learning_rate": 1.9763820904411322e-05, + "loss": 1.0943, + "step": 1011 + }, + { + "epoch": 0.09734513274336283, + "grad_norm": 2.6531187373904572, + "learning_rate": 1.9763147340467067e-05, + "loss": 1.035, + "step": 1012 + }, + { + "epoch": 0.09744132358599461, + "grad_norm": 2.7767340133259726, + "learning_rate": 1.9762472828925303e-05, + "loss": 1.1053, + "step": 1013 + }, + { + "epoch": 0.09753751442862639, + "grad_norm": 3.245374378839449, + "learning_rate": 1.9761797369851498e-05, + "loss": 1.0629, + "step": 1014 + }, + { + "epoch": 0.09763370527125817, + "grad_norm": 2.530715435144945, + "learning_rate": 1.9761120963311202e-05, + "loss": 1.1188, + "step": 1015 + }, + { + "epoch": 0.09772989611388996, + "grad_norm": 2.2957046216810753, + "learning_rate": 1.9760443609370074e-05, + "loss": 1.0311, + "step": 1016 + }, + { + "epoch": 0.09782608695652174, + "grad_norm": 2.695679510895763, + "learning_rate": 1.9759765308093856e-05, + "loss": 0.9864, + "step": 1017 + }, + { + "epoch": 0.09792227779915352, + "grad_norm": 2.6695989282632104, + "learning_rate": 1.975908605954838e-05, + "loss": 1.1048, + "step": 1018 + }, + { + "epoch": 0.0980184686417853, + "grad_norm": 2.638616488066361, + "learning_rate": 1.975840586379958e-05, + "loss": 1.2235, + "step": 1019 + }, + { + "epoch": 0.09811465948441708, + "grad_norm": 2.54631575549353, + "learning_rate": 1.9757724720913466e-05, + "loss": 1.1297, + "step": 1020 + }, + { + "epoch": 0.09821085032704886, + "grad_norm": 3.35416080539777, + "learning_rate": 1.9757042630956156e-05, + "loss": 1.0846, + "step": 1021 + }, + { + "epoch": 0.09830704116968064, + "grad_norm": 3.0030856074505823, + "learning_rate": 1.9756359593993845e-05, + "loss": 1.1003, + "step": 1022 + }, + { + "epoch": 0.09840323201231242, + "grad_norm": 2.9409558325264897, + "learning_rate": 1.9755675610092837e-05, + "loss": 1.2847, + "step": 1023 + }, + { + "epoch": 0.09849942285494422, + "grad_norm": 3.012652413277277, + "learning_rate": 1.975499067931951e-05, + "loss": 1.1397, + "step": 1024 + }, + { + "epoch": 0.098595613697576, + "grad_norm": 2.6157579857200197, + "learning_rate": 1.975430480174035e-05, + "loss": 1.1092, + "step": 1025 + }, + { + "epoch": 0.09869180454020778, + "grad_norm": 2.493813810661598, + "learning_rate": 1.975361797742192e-05, + "loss": 1.0857, + "step": 1026 + }, + { + "epoch": 0.09878799538283956, + "grad_norm": 2.5915416744640902, + "learning_rate": 1.9752930206430885e-05, + "loss": 1.0889, + "step": 1027 + }, + { + "epoch": 0.09888418622547133, + "grad_norm": 2.7174522475363845, + "learning_rate": 1.9752241488834002e-05, + "loss": 1.0661, + "step": 1028 + }, + { + "epoch": 0.09898037706810311, + "grad_norm": 2.873756529253318, + "learning_rate": 1.9751551824698113e-05, + "loss": 1.1358, + "step": 1029 + }, + { + "epoch": 0.0990765679107349, + "grad_norm": 2.563486042664789, + "learning_rate": 1.975086121409016e-05, + "loss": 1.1006, + "step": 1030 + }, + { + "epoch": 0.09917275875336667, + "grad_norm": 2.3717626212558933, + "learning_rate": 1.9750169657077167e-05, + "loss": 1.1795, + "step": 1031 + }, + { + "epoch": 0.09926894959599845, + "grad_norm": 2.7814767849254394, + "learning_rate": 1.974947715372626e-05, + "loss": 1.228, + "step": 1032 + }, + { + "epoch": 0.09936514043863025, + "grad_norm": 2.5098759185437145, + "learning_rate": 1.974878370410465e-05, + "loss": 1.0186, + "step": 1033 + }, + { + "epoch": 0.09946133128126203, + "grad_norm": 2.7309521692418257, + "learning_rate": 1.974808930827965e-05, + "loss": 1.1473, + "step": 1034 + }, + { + "epoch": 0.09955752212389381, + "grad_norm": 2.749696432043476, + "learning_rate": 1.9747393966318643e-05, + "loss": 1.062, + "step": 1035 + }, + { + "epoch": 0.09965371296652559, + "grad_norm": 2.648909411392473, + "learning_rate": 1.9746697678289128e-05, + "loss": 0.9203, + "step": 1036 + }, + { + "epoch": 0.09974990380915737, + "grad_norm": 2.520092019267388, + "learning_rate": 1.974600044425868e-05, + "loss": 1.0205, + "step": 1037 + }, + { + "epoch": 0.09984609465178915, + "grad_norm": 2.2824962624715504, + "learning_rate": 1.9745302264294982e-05, + "loss": 0.8551, + "step": 1038 + }, + { + "epoch": 0.09994228549442093, + "grad_norm": 2.2368038807238224, + "learning_rate": 1.9744603138465786e-05, + "loss": 1.1083, + "step": 1039 + }, + { + "epoch": 0.1000384763370527, + "grad_norm": 2.519944052118764, + "learning_rate": 1.9743903066838954e-05, + "loss": 1.091, + "step": 1040 + }, + { + "epoch": 0.1001346671796845, + "grad_norm": 2.562153357796346, + "learning_rate": 1.9743202049482435e-05, + "loss": 1.0864, + "step": 1041 + }, + { + "epoch": 0.10023085802231628, + "grad_norm": 2.4046766240905164, + "learning_rate": 1.9742500086464266e-05, + "loss": 1.0083, + "step": 1042 + }, + { + "epoch": 0.10032704886494806, + "grad_norm": 2.8853877606878675, + "learning_rate": 1.974179717785258e-05, + "loss": 1.1532, + "step": 1043 + }, + { + "epoch": 0.10042323970757984, + "grad_norm": 2.7726313401601783, + "learning_rate": 1.9741093323715597e-05, + "loss": 1.0166, + "step": 1044 + }, + { + "epoch": 0.10051943055021162, + "grad_norm": 2.292055990368359, + "learning_rate": 1.974038852412164e-05, + "loss": 0.8485, + "step": 1045 + }, + { + "epoch": 0.1006156213928434, + "grad_norm": 2.8053080428489228, + "learning_rate": 1.9739682779139107e-05, + "loss": 1.1021, + "step": 1046 + }, + { + "epoch": 0.10071181223547518, + "grad_norm": 2.7736925811861766, + "learning_rate": 1.97389760888365e-05, + "loss": 1.0345, + "step": 1047 + }, + { + "epoch": 0.10080800307810696, + "grad_norm": 2.657600618205276, + "learning_rate": 1.9738268453282414e-05, + "loss": 1.0856, + "step": 1048 + }, + { + "epoch": 0.10090419392073875, + "grad_norm": 3.1066015285930035, + "learning_rate": 1.9737559872545525e-05, + "loss": 1.133, + "step": 1049 + }, + { + "epoch": 0.10100038476337053, + "grad_norm": 1.9502884843377535, + "learning_rate": 1.9736850346694608e-05, + "loss": 0.8216, + "step": 1050 + }, + { + "epoch": 0.10109657560600231, + "grad_norm": 2.3011331103296864, + "learning_rate": 1.973613987579853e-05, + "loss": 1.0996, + "step": 1051 + }, + { + "epoch": 0.10119276644863409, + "grad_norm": 2.5041577361128824, + "learning_rate": 1.973542845992625e-05, + "loss": 1.0946, + "step": 1052 + }, + { + "epoch": 0.10128895729126587, + "grad_norm": 2.7907040030850205, + "learning_rate": 1.9734716099146812e-05, + "loss": 1.1437, + "step": 1053 + }, + { + "epoch": 0.10138514813389765, + "grad_norm": 2.586503758679151, + "learning_rate": 1.9734002793529362e-05, + "loss": 1.0783, + "step": 1054 + }, + { + "epoch": 0.10148133897652943, + "grad_norm": 2.9338276832214403, + "learning_rate": 1.9733288543143127e-05, + "loss": 1.0485, + "step": 1055 + }, + { + "epoch": 0.10157752981916121, + "grad_norm": 2.7400731044884634, + "learning_rate": 1.9732573348057437e-05, + "loss": 1.1489, + "step": 1056 + }, + { + "epoch": 0.10167372066179299, + "grad_norm": 2.0686141089761056, + "learning_rate": 1.97318572083417e-05, + "loss": 0.7937, + "step": 1057 + }, + { + "epoch": 0.10176991150442478, + "grad_norm": 2.798016980619357, + "learning_rate": 1.973114012406544e-05, + "loss": 1.1643, + "step": 1058 + }, + { + "epoch": 0.10186610234705656, + "grad_norm": 2.738774453489567, + "learning_rate": 1.9730422095298234e-05, + "loss": 1.1648, + "step": 1059 + }, + { + "epoch": 0.10196229318968834, + "grad_norm": 2.605538430326287, + "learning_rate": 1.9729703122109788e-05, + "loss": 1.1516, + "step": 1060 + }, + { + "epoch": 0.10205848403232012, + "grad_norm": 2.767558310134297, + "learning_rate": 1.9728983204569882e-05, + "loss": 1.1225, + "step": 1061 + }, + { + "epoch": 0.1021546748749519, + "grad_norm": 2.7671811739162746, + "learning_rate": 1.9728262342748384e-05, + "loss": 1.2591, + "step": 1062 + }, + { + "epoch": 0.10225086571758368, + "grad_norm": 2.1360489443651844, + "learning_rate": 1.9727540536715268e-05, + "loss": 0.9235, + "step": 1063 + }, + { + "epoch": 0.10234705656021546, + "grad_norm": 2.6543309601358733, + "learning_rate": 1.9726817786540584e-05, + "loss": 1.0865, + "step": 1064 + }, + { + "epoch": 0.10244324740284724, + "grad_norm": 2.6679775431888775, + "learning_rate": 1.9726094092294486e-05, + "loss": 1.1331, + "step": 1065 + }, + { + "epoch": 0.10253943824547904, + "grad_norm": 2.4665467463458537, + "learning_rate": 1.9725369454047215e-05, + "loss": 1.1828, + "step": 1066 + }, + { + "epoch": 0.10263562908811082, + "grad_norm": 2.4059777841849876, + "learning_rate": 1.97246438718691e-05, + "loss": 0.9639, + "step": 1067 + }, + { + "epoch": 0.1027318199307426, + "grad_norm": 2.7917125136226306, + "learning_rate": 1.9723917345830568e-05, + "loss": 1.0483, + "step": 1068 + }, + { + "epoch": 0.10282801077337438, + "grad_norm": 2.3695566830025565, + "learning_rate": 1.9723189876002132e-05, + "loss": 1.0215, + "step": 1069 + }, + { + "epoch": 0.10292420161600616, + "grad_norm": 2.617856887666521, + "learning_rate": 1.9722461462454405e-05, + "loss": 1.1059, + "step": 1070 + }, + { + "epoch": 0.10302039245863794, + "grad_norm": 3.0006489017360085, + "learning_rate": 1.972173210525808e-05, + "loss": 1.1086, + "step": 1071 + }, + { + "epoch": 0.10311658330126972, + "grad_norm": 2.395366761146141, + "learning_rate": 1.9721001804483947e-05, + "loss": 1.0203, + "step": 1072 + }, + { + "epoch": 0.1032127741439015, + "grad_norm": 2.4285383412025787, + "learning_rate": 1.972027056020289e-05, + "loss": 1.0554, + "step": 1073 + }, + { + "epoch": 0.10330896498653329, + "grad_norm": 2.6295851280518585, + "learning_rate": 1.9719538372485887e-05, + "loss": 1.1179, + "step": 1074 + }, + { + "epoch": 0.10340515582916507, + "grad_norm": 2.8003447942625863, + "learning_rate": 1.9718805241404e-05, + "loss": 1.044, + "step": 1075 + }, + { + "epoch": 0.10350134667179685, + "grad_norm": 2.4229353045173676, + "learning_rate": 1.9718071167028376e-05, + "loss": 1.0258, + "step": 1076 + }, + { + "epoch": 0.10359753751442863, + "grad_norm": 2.3464220158451954, + "learning_rate": 1.9717336149430276e-05, + "loss": 1.0847, + "step": 1077 + }, + { + "epoch": 0.10369372835706041, + "grad_norm": 2.5037798689856996, + "learning_rate": 1.9716600188681038e-05, + "loss": 1.0769, + "step": 1078 + }, + { + "epoch": 0.10378991919969219, + "grad_norm": 2.395745174957134, + "learning_rate": 1.9715863284852085e-05, + "loss": 1.1694, + "step": 1079 + }, + { + "epoch": 0.10388611004232397, + "grad_norm": 2.6094141744974237, + "learning_rate": 1.971512543801495e-05, + "loss": 1.2064, + "step": 1080 + }, + { + "epoch": 0.10398230088495575, + "grad_norm": 2.4558512246352118, + "learning_rate": 1.971438664824124e-05, + "loss": 0.9539, + "step": 1081 + }, + { + "epoch": 0.10407849172758753, + "grad_norm": 2.6807090044116846, + "learning_rate": 1.9713646915602663e-05, + "loss": 1.0377, + "step": 1082 + }, + { + "epoch": 0.10417468257021932, + "grad_norm": 2.287096556452081, + "learning_rate": 1.971290624017102e-05, + "loss": 1.0937, + "step": 1083 + }, + { + "epoch": 0.1042708734128511, + "grad_norm": 2.5295040868859804, + "learning_rate": 1.9712164622018197e-05, + "loss": 1.0922, + "step": 1084 + }, + { + "epoch": 0.10436706425548288, + "grad_norm": 2.4420753071795516, + "learning_rate": 1.9711422061216178e-05, + "loss": 1.0636, + "step": 1085 + }, + { + "epoch": 0.10446325509811466, + "grad_norm": 2.507513609938945, + "learning_rate": 1.9710678557837024e-05, + "loss": 1.0466, + "step": 1086 + }, + { + "epoch": 0.10455944594074644, + "grad_norm": 2.5442837648916132, + "learning_rate": 1.9709934111952913e-05, + "loss": 1.0309, + "step": 1087 + }, + { + "epoch": 0.10465563678337822, + "grad_norm": 2.701814899424353, + "learning_rate": 1.9709188723636088e-05, + "loss": 1.1138, + "step": 1088 + }, + { + "epoch": 0.10475182762601, + "grad_norm": 2.301420879494918, + "learning_rate": 1.97084423929589e-05, + "loss": 1.0094, + "step": 1089 + }, + { + "epoch": 0.10484801846864178, + "grad_norm": 2.3965304207435656, + "learning_rate": 1.970769511999379e-05, + "loss": 1.1392, + "step": 1090 + }, + { + "epoch": 0.10494420931127357, + "grad_norm": 2.5787412136871946, + "learning_rate": 1.9706946904813285e-05, + "loss": 1.1183, + "step": 1091 + }, + { + "epoch": 0.10504040015390535, + "grad_norm": 2.440440732419794, + "learning_rate": 1.9706197747490004e-05, + "loss": 1.0744, + "step": 1092 + }, + { + "epoch": 0.10513659099653713, + "grad_norm": 2.3855954976730778, + "learning_rate": 1.970544764809666e-05, + "loss": 1.0494, + "step": 1093 + }, + { + "epoch": 0.10523278183916891, + "grad_norm": 2.697303168548141, + "learning_rate": 1.9704696606706055e-05, + "loss": 1.0863, + "step": 1094 + }, + { + "epoch": 0.10532897268180069, + "grad_norm": 2.8288983881676044, + "learning_rate": 1.9703944623391085e-05, + "loss": 1.2456, + "step": 1095 + }, + { + "epoch": 0.10542516352443247, + "grad_norm": 2.707313381610882, + "learning_rate": 1.9703191698224742e-05, + "loss": 1.1062, + "step": 1096 + }, + { + "epoch": 0.10552135436706425, + "grad_norm": 2.517855764032878, + "learning_rate": 1.9702437831280097e-05, + "loss": 1.1003, + "step": 1097 + }, + { + "epoch": 0.10561754520969603, + "grad_norm": 2.6541933414623946, + "learning_rate": 1.9701683022630323e-05, + "loss": 1.0545, + "step": 1098 + }, + { + "epoch": 0.10571373605232783, + "grad_norm": 2.4773957960352737, + "learning_rate": 1.9700927272348676e-05, + "loss": 1.1403, + "step": 1099 + }, + { + "epoch": 0.1058099268949596, + "grad_norm": 2.634942745900511, + "learning_rate": 1.9700170580508514e-05, + "loss": 1.1406, + "step": 1100 + }, + { + "epoch": 0.10590611773759138, + "grad_norm": 2.8377956211160136, + "learning_rate": 1.9699412947183273e-05, + "loss": 1.2195, + "step": 1101 + }, + { + "epoch": 0.10600230858022316, + "grad_norm": 2.398183358082121, + "learning_rate": 1.9698654372446495e-05, + "loss": 1.1312, + "step": 1102 + }, + { + "epoch": 0.10609849942285494, + "grad_norm": 2.446188607602268, + "learning_rate": 1.9697894856371808e-05, + "loss": 1.1102, + "step": 1103 + }, + { + "epoch": 0.10619469026548672, + "grad_norm": 2.5508847080189763, + "learning_rate": 1.969713439903292e-05, + "loss": 1.1237, + "step": 1104 + }, + { + "epoch": 0.1062908811081185, + "grad_norm": 2.1937969361891154, + "learning_rate": 1.969637300050365e-05, + "loss": 1.1666, + "step": 1105 + }, + { + "epoch": 0.10638707195075028, + "grad_norm": 2.520651146406654, + "learning_rate": 1.9695610660857886e-05, + "loss": 0.942, + "step": 1106 + }, + { + "epoch": 0.10648326279338206, + "grad_norm": 2.693062509949498, + "learning_rate": 1.9694847380169632e-05, + "loss": 0.9981, + "step": 1107 + }, + { + "epoch": 0.10657945363601386, + "grad_norm": 2.374898085749084, + "learning_rate": 1.9694083158512965e-05, + "loss": 1.0747, + "step": 1108 + }, + { + "epoch": 0.10667564447864564, + "grad_norm": 2.4604846057189897, + "learning_rate": 1.969331799596206e-05, + "loss": 0.8805, + "step": 1109 + }, + { + "epoch": 0.10677183532127742, + "grad_norm": 2.377381489354385, + "learning_rate": 1.9692551892591185e-05, + "loss": 1.013, + "step": 1110 + }, + { + "epoch": 0.1068680261639092, + "grad_norm": 2.800484889669349, + "learning_rate": 1.9691784848474692e-05, + "loss": 1.1399, + "step": 1111 + }, + { + "epoch": 0.10696421700654098, + "grad_norm": 2.60237927315697, + "learning_rate": 1.9691016863687037e-05, + "loss": 1.0479, + "step": 1112 + }, + { + "epoch": 0.10706040784917276, + "grad_norm": 2.755744985925418, + "learning_rate": 1.969024793830275e-05, + "loss": 1.0877, + "step": 1113 + }, + { + "epoch": 0.10715659869180454, + "grad_norm": 2.263500536045772, + "learning_rate": 1.968947807239647e-05, + "loss": 1.1201, + "step": 1114 + }, + { + "epoch": 0.10725278953443632, + "grad_norm": 2.575618227575495, + "learning_rate": 1.9688707266042914e-05, + "loss": 1.0629, + "step": 1115 + }, + { + "epoch": 0.10734898037706811, + "grad_norm": 1.9118926440881143, + "learning_rate": 1.9687935519316897e-05, + "loss": 0.865, + "step": 1116 + }, + { + "epoch": 0.10744517121969989, + "grad_norm": 2.3746447210811668, + "learning_rate": 1.9687162832293323e-05, + "loss": 1.0278, + "step": 1117 + }, + { + "epoch": 0.10754136206233167, + "grad_norm": 2.6142542576657126, + "learning_rate": 1.9686389205047186e-05, + "loss": 1.0914, + "step": 1118 + }, + { + "epoch": 0.10763755290496345, + "grad_norm": 2.5202406465242113, + "learning_rate": 1.968561463765358e-05, + "loss": 1.1309, + "step": 1119 + }, + { + "epoch": 0.10773374374759523, + "grad_norm": 2.9278216365616445, + "learning_rate": 1.9684839130187678e-05, + "loss": 0.8959, + "step": 1120 + }, + { + "epoch": 0.10782993459022701, + "grad_norm": 2.2054670741287974, + "learning_rate": 1.968406268272475e-05, + "loss": 1.0362, + "step": 1121 + }, + { + "epoch": 0.10792612543285879, + "grad_norm": 2.6366317020887386, + "learning_rate": 1.968328529534016e-05, + "loss": 1.0306, + "step": 1122 + }, + { + "epoch": 0.10802231627549057, + "grad_norm": 2.525040791469544, + "learning_rate": 1.9682506968109358e-05, + "loss": 1.028, + "step": 1123 + }, + { + "epoch": 0.10811850711812235, + "grad_norm": 2.2374312309462514, + "learning_rate": 1.9681727701107885e-05, + "loss": 1.0085, + "step": 1124 + }, + { + "epoch": 0.10821469796075414, + "grad_norm": 2.2776980860383658, + "learning_rate": 1.968094749441138e-05, + "loss": 1.1005, + "step": 1125 + }, + { + "epoch": 0.10831088880338592, + "grad_norm": 2.2656015827514717, + "learning_rate": 1.9680166348095568e-05, + "loss": 1.0323, + "step": 1126 + }, + { + "epoch": 0.1084070796460177, + "grad_norm": 2.791084337320309, + "learning_rate": 1.967938426223626e-05, + "loss": 1.1532, + "step": 1127 + }, + { + "epoch": 0.10850327048864948, + "grad_norm": 2.984156422049629, + "learning_rate": 1.967860123690937e-05, + "loss": 1.1438, + "step": 1128 + }, + { + "epoch": 0.10859946133128126, + "grad_norm": 2.3351853960910853, + "learning_rate": 1.9677817272190897e-05, + "loss": 1.0801, + "step": 1129 + }, + { + "epoch": 0.10869565217391304, + "grad_norm": 2.7121600243857564, + "learning_rate": 1.9677032368156934e-05, + "loss": 1.1313, + "step": 1130 + }, + { + "epoch": 0.10879184301654482, + "grad_norm": 2.4933655991745525, + "learning_rate": 1.9676246524883656e-05, + "loss": 0.9877, + "step": 1131 + }, + { + "epoch": 0.1088880338591766, + "grad_norm": 2.89579573546109, + "learning_rate": 1.967545974244734e-05, + "loss": 1.0975, + "step": 1132 + }, + { + "epoch": 0.1089842247018084, + "grad_norm": 2.573827840645843, + "learning_rate": 1.9674672020924346e-05, + "loss": 1.157, + "step": 1133 + }, + { + "epoch": 0.10908041554444017, + "grad_norm": 2.3955307824961998, + "learning_rate": 1.9673883360391138e-05, + "loss": 1.1841, + "step": 1134 + }, + { + "epoch": 0.10917660638707195, + "grad_norm": 2.4139381887744134, + "learning_rate": 1.9673093760924253e-05, + "loss": 1.0724, + "step": 1135 + }, + { + "epoch": 0.10927279722970373, + "grad_norm": 2.380515454141817, + "learning_rate": 1.9672303222600333e-05, + "loss": 1.0723, + "step": 1136 + }, + { + "epoch": 0.10936898807233551, + "grad_norm": 3.0942477045252716, + "learning_rate": 1.9671511745496102e-05, + "loss": 1.0392, + "step": 1137 + }, + { + "epoch": 0.10946517891496729, + "grad_norm": 2.7131184926397554, + "learning_rate": 1.967071932968839e-05, + "loss": 1.0785, + "step": 1138 + }, + { + "epoch": 0.10956136975759907, + "grad_norm": 2.5208209146730445, + "learning_rate": 1.96699259752541e-05, + "loss": 1.1037, + "step": 1139 + }, + { + "epoch": 0.10965756060023085, + "grad_norm": 2.3681173427510185, + "learning_rate": 1.9669131682270232e-05, + "loss": 1.1164, + "step": 1140 + }, + { + "epoch": 0.10975375144286265, + "grad_norm": 2.388495569435449, + "learning_rate": 1.9668336450813884e-05, + "loss": 1.0397, + "step": 1141 + }, + { + "epoch": 0.10984994228549443, + "grad_norm": 2.597912370355133, + "learning_rate": 1.9667540280962235e-05, + "loss": 1.1607, + "step": 1142 + }, + { + "epoch": 0.1099461331281262, + "grad_norm": 2.457669620892253, + "learning_rate": 1.9666743172792567e-05, + "loss": 1.0942, + "step": 1143 + }, + { + "epoch": 0.11004232397075799, + "grad_norm": 2.8467576310471046, + "learning_rate": 1.966594512638224e-05, + "loss": 1.1624, + "step": 1144 + }, + { + "epoch": 0.11013851481338977, + "grad_norm": 2.4275747944815302, + "learning_rate": 1.9665146141808713e-05, + "loss": 1.0125, + "step": 1145 + }, + { + "epoch": 0.11023470565602154, + "grad_norm": 2.416253338693964, + "learning_rate": 1.9664346219149538e-05, + "loss": 1.1293, + "step": 1146 + }, + { + "epoch": 0.11033089649865332, + "grad_norm": 3.053411782095631, + "learning_rate": 1.966354535848235e-05, + "loss": 1.1272, + "step": 1147 + }, + { + "epoch": 0.1104270873412851, + "grad_norm": 2.216014478194383, + "learning_rate": 1.966274355988488e-05, + "loss": 0.8499, + "step": 1148 + }, + { + "epoch": 0.11052327818391688, + "grad_norm": 2.362729662134667, + "learning_rate": 1.966194082343495e-05, + "loss": 1.0917, + "step": 1149 + }, + { + "epoch": 0.11061946902654868, + "grad_norm": 2.477831144116438, + "learning_rate": 1.9661137149210473e-05, + "loss": 1.1896, + "step": 1150 + }, + { + "epoch": 0.11071565986918046, + "grad_norm": 2.8430967610815254, + "learning_rate": 1.9660332537289454e-05, + "loss": 1.2291, + "step": 1151 + }, + { + "epoch": 0.11081185071181224, + "grad_norm": 2.435210035558518, + "learning_rate": 1.9659526987749987e-05, + "loss": 1.0791, + "step": 1152 + }, + { + "epoch": 0.11090804155444402, + "grad_norm": 2.270252989941342, + "learning_rate": 1.9658720500670252e-05, + "loss": 1.1538, + "step": 1153 + }, + { + "epoch": 0.1110042323970758, + "grad_norm": 2.337636438231488, + "learning_rate": 1.9657913076128532e-05, + "loss": 1.1213, + "step": 1154 + }, + { + "epoch": 0.11110042323970758, + "grad_norm": 2.9832981725288565, + "learning_rate": 1.965710471420319e-05, + "loss": 1.1155, + "step": 1155 + }, + { + "epoch": 0.11119661408233936, + "grad_norm": 1.9404868850032364, + "learning_rate": 1.965629541497269e-05, + "loss": 0.7669, + "step": 1156 + }, + { + "epoch": 0.11129280492497114, + "grad_norm": 2.889155066731007, + "learning_rate": 1.965548517851558e-05, + "loss": 1.0236, + "step": 1157 + }, + { + "epoch": 0.11138899576760293, + "grad_norm": 2.4968764594911734, + "learning_rate": 1.9654674004910493e-05, + "loss": 1.1002, + "step": 1158 + }, + { + "epoch": 0.11148518661023471, + "grad_norm": 3.0981385819417917, + "learning_rate": 1.965386189423617e-05, + "loss": 0.982, + "step": 1159 + }, + { + "epoch": 0.11158137745286649, + "grad_norm": 2.9204466244663654, + "learning_rate": 1.9653048846571427e-05, + "loss": 1.1903, + "step": 1160 + }, + { + "epoch": 0.11167756829549827, + "grad_norm": 2.366645013191173, + "learning_rate": 1.9652234861995182e-05, + "loss": 1.1054, + "step": 1161 + }, + { + "epoch": 0.11177375913813005, + "grad_norm": 2.8391218375983565, + "learning_rate": 1.9651419940586437e-05, + "loss": 1.2024, + "step": 1162 + }, + { + "epoch": 0.11186994998076183, + "grad_norm": 2.500978733918135, + "learning_rate": 1.9650604082424284e-05, + "loss": 1.0199, + "step": 1163 + }, + { + "epoch": 0.11196614082339361, + "grad_norm": 2.590955142187799, + "learning_rate": 1.964978728758791e-05, + "loss": 1.0406, + "step": 1164 + }, + { + "epoch": 0.11206233166602539, + "grad_norm": 3.1103543015017507, + "learning_rate": 1.96489695561566e-05, + "loss": 1.0738, + "step": 1165 + }, + { + "epoch": 0.11215852250865718, + "grad_norm": 2.7189508040889834, + "learning_rate": 1.9648150888209715e-05, + "loss": 0.9461, + "step": 1166 + }, + { + "epoch": 0.11225471335128896, + "grad_norm": 2.790315137643281, + "learning_rate": 1.9647331283826713e-05, + "loss": 1.1414, + "step": 1167 + }, + { + "epoch": 0.11235090419392074, + "grad_norm": 2.647500421838687, + "learning_rate": 1.9646510743087144e-05, + "loss": 1.1953, + "step": 1168 + }, + { + "epoch": 0.11244709503655252, + "grad_norm": 2.382198391504249, + "learning_rate": 1.9645689266070654e-05, + "loss": 1.0741, + "step": 1169 + }, + { + "epoch": 0.1125432858791843, + "grad_norm": 2.522791532208661, + "learning_rate": 1.964486685285697e-05, + "loss": 1.077, + "step": 1170 + }, + { + "epoch": 0.11263947672181608, + "grad_norm": 2.569857704865992, + "learning_rate": 1.964404350352591e-05, + "loss": 1.16, + "step": 1171 + }, + { + "epoch": 0.11273566756444786, + "grad_norm": 2.4567415456091872, + "learning_rate": 1.9643219218157395e-05, + "loss": 1.0052, + "step": 1172 + }, + { + "epoch": 0.11283185840707964, + "grad_norm": 2.381417585084485, + "learning_rate": 1.9642393996831423e-05, + "loss": 0.9674, + "step": 1173 + }, + { + "epoch": 0.11292804924971142, + "grad_norm": 2.6478863581674053, + "learning_rate": 1.9641567839628092e-05, + "loss": 0.98, + "step": 1174 + }, + { + "epoch": 0.11302424009234321, + "grad_norm": 2.780617524645946, + "learning_rate": 1.964074074662759e-05, + "loss": 1.1677, + "step": 1175 + }, + { + "epoch": 0.113120430934975, + "grad_norm": 2.8323384463709766, + "learning_rate": 1.963991271791019e-05, + "loss": 1.098, + "step": 1176 + }, + { + "epoch": 0.11321662177760677, + "grad_norm": 2.2553584434613896, + "learning_rate": 1.9639083753556258e-05, + "loss": 1.0191, + "step": 1177 + }, + { + "epoch": 0.11331281262023855, + "grad_norm": 2.5985926773346324, + "learning_rate": 1.9638253853646255e-05, + "loss": 1.1144, + "step": 1178 + }, + { + "epoch": 0.11340900346287033, + "grad_norm": 2.635319316937903, + "learning_rate": 1.963742301826073e-05, + "loss": 1.1436, + "step": 1179 + }, + { + "epoch": 0.11350519430550211, + "grad_norm": 2.4727884225044447, + "learning_rate": 1.9636591247480323e-05, + "loss": 1.1027, + "step": 1180 + }, + { + "epoch": 0.1136013851481339, + "grad_norm": 2.418049288847308, + "learning_rate": 1.963575854138576e-05, + "loss": 1.035, + "step": 1181 + }, + { + "epoch": 0.11369757599076567, + "grad_norm": 2.4821900780422745, + "learning_rate": 1.9634924900057867e-05, + "loss": 1.1318, + "step": 1182 + }, + { + "epoch": 0.11379376683339747, + "grad_norm": 2.3675408869267174, + "learning_rate": 1.9634090323577558e-05, + "loss": 0.829, + "step": 1183 + }, + { + "epoch": 0.11388995767602925, + "grad_norm": 2.53138020174949, + "learning_rate": 1.963325481202583e-05, + "loss": 1.0388, + "step": 1184 + }, + { + "epoch": 0.11398614851866103, + "grad_norm": 2.5779479268429775, + "learning_rate": 1.9632418365483778e-05, + "loss": 1.0832, + "step": 1185 + }, + { + "epoch": 0.1140823393612928, + "grad_norm": 2.484456098955069, + "learning_rate": 1.963158098403259e-05, + "loss": 1.1178, + "step": 1186 + }, + { + "epoch": 0.11417853020392459, + "grad_norm": 2.276653388658507, + "learning_rate": 1.9630742667753538e-05, + "loss": 0.8284, + "step": 1187 + }, + { + "epoch": 0.11427472104655637, + "grad_norm": 2.849262163656551, + "learning_rate": 1.9629903416727987e-05, + "loss": 1.0791, + "step": 1188 + }, + { + "epoch": 0.11437091188918815, + "grad_norm": 2.4026094356773693, + "learning_rate": 1.962906323103739e-05, + "loss": 1.0614, + "step": 1189 + }, + { + "epoch": 0.11446710273181993, + "grad_norm": 2.724273983976748, + "learning_rate": 1.962822211076331e-05, + "loss": 1.1707, + "step": 1190 + }, + { + "epoch": 0.11456329357445172, + "grad_norm": 2.5269773834132065, + "learning_rate": 1.9627380055987366e-05, + "loss": 1.0313, + "step": 1191 + }, + { + "epoch": 0.1146594844170835, + "grad_norm": 2.3412883958725828, + "learning_rate": 1.96265370667913e-05, + "loss": 1.1409, + "step": 1192 + }, + { + "epoch": 0.11475567525971528, + "grad_norm": 2.329900118390609, + "learning_rate": 1.9625693143256922e-05, + "loss": 1.034, + "step": 1193 + }, + { + "epoch": 0.11485186610234706, + "grad_norm": 2.976663180533762, + "learning_rate": 1.9624848285466146e-05, + "loss": 1.1142, + "step": 1194 + }, + { + "epoch": 0.11494805694497884, + "grad_norm": 2.4350231908478626, + "learning_rate": 1.9624002493500974e-05, + "loss": 1.0444, + "step": 1195 + }, + { + "epoch": 0.11504424778761062, + "grad_norm": 2.315829135726338, + "learning_rate": 1.9623155767443498e-05, + "loss": 1.1016, + "step": 1196 + }, + { + "epoch": 0.1151404386302424, + "grad_norm": 2.5555752942941172, + "learning_rate": 1.96223081073759e-05, + "loss": 1.1144, + "step": 1197 + }, + { + "epoch": 0.11523662947287418, + "grad_norm": 2.432199501316201, + "learning_rate": 1.9621459513380445e-05, + "loss": 1.1034, + "step": 1198 + }, + { + "epoch": 0.11533282031550596, + "grad_norm": 2.725716439841287, + "learning_rate": 1.9620609985539508e-05, + "loss": 1.203, + "step": 1199 + }, + { + "epoch": 0.11542901115813775, + "grad_norm": 2.487150084298904, + "learning_rate": 1.9619759523935532e-05, + "loss": 1.2032, + "step": 1200 + }, + { + "epoch": 0.11552520200076953, + "grad_norm": 2.535352160168558, + "learning_rate": 1.9618908128651072e-05, + "loss": 1.1189, + "step": 1201 + }, + { + "epoch": 0.11562139284340131, + "grad_norm": 2.841078442538935, + "learning_rate": 1.9618055799768757e-05, + "loss": 1.1536, + "step": 1202 + }, + { + "epoch": 0.11571758368603309, + "grad_norm": 2.884956453671765, + "learning_rate": 1.9617202537371313e-05, + "loss": 1.2045, + "step": 1203 + }, + { + "epoch": 0.11581377452866487, + "grad_norm": 2.6264611173559618, + "learning_rate": 1.961634834154156e-05, + "loss": 1.0065, + "step": 1204 + }, + { + "epoch": 0.11590996537129665, + "grad_norm": 2.47045123563961, + "learning_rate": 1.96154932123624e-05, + "loss": 1.1522, + "step": 1205 + }, + { + "epoch": 0.11600615621392843, + "grad_norm": 2.478337087639889, + "learning_rate": 1.9614637149916834e-05, + "loss": 1.1374, + "step": 1206 + }, + { + "epoch": 0.11610234705656021, + "grad_norm": 2.3872810726148437, + "learning_rate": 1.9613780154287953e-05, + "loss": 1.1513, + "step": 1207 + }, + { + "epoch": 0.116198537899192, + "grad_norm": 2.2620794162890316, + "learning_rate": 1.9612922225558924e-05, + "loss": 0.9958, + "step": 1208 + }, + { + "epoch": 0.11629472874182378, + "grad_norm": 2.4282026070210176, + "learning_rate": 1.961206336381303e-05, + "loss": 1.0256, + "step": 1209 + }, + { + "epoch": 0.11639091958445556, + "grad_norm": 2.158680938565515, + "learning_rate": 1.961120356913363e-05, + "loss": 0.8327, + "step": 1210 + }, + { + "epoch": 0.11648711042708734, + "grad_norm": 2.7782531417627094, + "learning_rate": 1.9610342841604163e-05, + "loss": 1.0138, + "step": 1211 + }, + { + "epoch": 0.11658330126971912, + "grad_norm": 2.6934320033413095, + "learning_rate": 1.960948118130818e-05, + "loss": 1.1542, + "step": 1212 + }, + { + "epoch": 0.1166794921123509, + "grad_norm": 2.478637750613058, + "learning_rate": 1.960861858832931e-05, + "loss": 1.1082, + "step": 1213 + }, + { + "epoch": 0.11677568295498268, + "grad_norm": 2.62981566089326, + "learning_rate": 1.9607755062751273e-05, + "loss": 1.1214, + "step": 1214 + }, + { + "epoch": 0.11687187379761446, + "grad_norm": 2.5723166855873534, + "learning_rate": 1.9606890604657884e-05, + "loss": 1.1359, + "step": 1215 + }, + { + "epoch": 0.11696806464024626, + "grad_norm": 2.392916953897404, + "learning_rate": 1.9606025214133046e-05, + "loss": 1.1474, + "step": 1216 + }, + { + "epoch": 0.11706425548287804, + "grad_norm": 2.342795595056559, + "learning_rate": 1.960515889126075e-05, + "loss": 1.0628, + "step": 1217 + }, + { + "epoch": 0.11716044632550981, + "grad_norm": 2.1946470912698413, + "learning_rate": 1.9604291636125084e-05, + "loss": 1.0913, + "step": 1218 + }, + { + "epoch": 0.1172566371681416, + "grad_norm": 1.9978991029257476, + "learning_rate": 1.9603423448810225e-05, + "loss": 0.8592, + "step": 1219 + }, + { + "epoch": 0.11735282801077337, + "grad_norm": 2.2608243018455076, + "learning_rate": 1.960255432940043e-05, + "loss": 1.0541, + "step": 1220 + }, + { + "epoch": 0.11744901885340515, + "grad_norm": 2.129162096831206, + "learning_rate": 1.9601684277980054e-05, + "loss": 1.0035, + "step": 1221 + }, + { + "epoch": 0.11754520969603693, + "grad_norm": 2.444171599569276, + "learning_rate": 1.9600813294633552e-05, + "loss": 1.0815, + "step": 1222 + }, + { + "epoch": 0.11764140053866871, + "grad_norm": 2.3058561582305317, + "learning_rate": 1.959994137944546e-05, + "loss": 1.067, + "step": 1223 + }, + { + "epoch": 0.1177375913813005, + "grad_norm": 2.485171946604205, + "learning_rate": 1.9599068532500394e-05, + "loss": 1.1379, + "step": 1224 + }, + { + "epoch": 0.11783378222393229, + "grad_norm": 2.6779935074909402, + "learning_rate": 1.959819475388308e-05, + "loss": 1.1243, + "step": 1225 + }, + { + "epoch": 0.11792997306656407, + "grad_norm": 2.5037734427030527, + "learning_rate": 1.9597320043678322e-05, + "loss": 1.1027, + "step": 1226 + }, + { + "epoch": 0.11802616390919585, + "grad_norm": 2.509166631532743, + "learning_rate": 1.959644440197102e-05, + "loss": 1.0435, + "step": 1227 + }, + { + "epoch": 0.11812235475182763, + "grad_norm": 2.443940824778464, + "learning_rate": 1.9595567828846166e-05, + "loss": 1.1534, + "step": 1228 + }, + { + "epoch": 0.1182185455944594, + "grad_norm": 2.5076722775467046, + "learning_rate": 1.959469032438883e-05, + "loss": 1.0985, + "step": 1229 + }, + { + "epoch": 0.11831473643709119, + "grad_norm": 2.322127265766094, + "learning_rate": 1.9593811888684192e-05, + "loss": 1.1534, + "step": 1230 + }, + { + "epoch": 0.11841092727972297, + "grad_norm": 2.5237248213238703, + "learning_rate": 1.9592932521817504e-05, + "loss": 1.1841, + "step": 1231 + }, + { + "epoch": 0.11850711812235475, + "grad_norm": 3.1186413159332727, + "learning_rate": 1.9592052223874115e-05, + "loss": 1.072, + "step": 1232 + }, + { + "epoch": 0.11860330896498654, + "grad_norm": 2.340290715517549, + "learning_rate": 1.9591170994939473e-05, + "loss": 1.034, + "step": 1233 + }, + { + "epoch": 0.11869949980761832, + "grad_norm": 2.5565251042104524, + "learning_rate": 1.959028883509911e-05, + "loss": 1.1314, + "step": 1234 + }, + { + "epoch": 0.1187956906502501, + "grad_norm": 2.223872530109356, + "learning_rate": 1.9589405744438636e-05, + "loss": 1.0017, + "step": 1235 + }, + { + "epoch": 0.11889188149288188, + "grad_norm": 2.169950217812695, + "learning_rate": 1.9588521723043764e-05, + "loss": 0.9852, + "step": 1236 + }, + { + "epoch": 0.11898807233551366, + "grad_norm": 2.4132954683244114, + "learning_rate": 1.9587636771000306e-05, + "loss": 1.0889, + "step": 1237 + }, + { + "epoch": 0.11908426317814544, + "grad_norm": 2.471626380328651, + "learning_rate": 1.958675088839415e-05, + "loss": 1.0373, + "step": 1238 + }, + { + "epoch": 0.11918045402077722, + "grad_norm": 2.4740548405003886, + "learning_rate": 1.9585864075311275e-05, + "loss": 1.1017, + "step": 1239 + }, + { + "epoch": 0.119276644863409, + "grad_norm": 2.527612350867435, + "learning_rate": 1.9584976331837758e-05, + "loss": 0.8908, + "step": 1240 + }, + { + "epoch": 0.11937283570604078, + "grad_norm": 2.3751113673381967, + "learning_rate": 1.958408765805976e-05, + "loss": 1.052, + "step": 1241 + }, + { + "epoch": 0.11946902654867257, + "grad_norm": 2.4731770055333704, + "learning_rate": 1.9583198054063535e-05, + "loss": 1.031, + "step": 1242 + }, + { + "epoch": 0.11956521739130435, + "grad_norm": 2.0848491665556033, + "learning_rate": 1.9582307519935426e-05, + "loss": 0.857, + "step": 1243 + }, + { + "epoch": 0.11966140823393613, + "grad_norm": 2.3116737741609583, + "learning_rate": 1.9581416055761865e-05, + "loss": 1.0242, + "step": 1244 + }, + { + "epoch": 0.11975759907656791, + "grad_norm": 2.679433501013352, + "learning_rate": 1.958052366162938e-05, + "loss": 1.2115, + "step": 1245 + }, + { + "epoch": 0.11985378991919969, + "grad_norm": 2.606732610255057, + "learning_rate": 1.9579630337624585e-05, + "loss": 0.9986, + "step": 1246 + }, + { + "epoch": 0.11994998076183147, + "grad_norm": 2.3346682621805517, + "learning_rate": 1.9578736083834186e-05, + "loss": 1.0274, + "step": 1247 + }, + { + "epoch": 0.12004617160446325, + "grad_norm": 2.2365415070247074, + "learning_rate": 1.9577840900344974e-05, + "loss": 1.057, + "step": 1248 + }, + { + "epoch": 0.12014236244709503, + "grad_norm": 2.963878480655606, + "learning_rate": 1.957694478724384e-05, + "loss": 1.2181, + "step": 1249 + }, + { + "epoch": 0.12023855328972682, + "grad_norm": 2.283613092059061, + "learning_rate": 1.9576047744617752e-05, + "loss": 1.1413, + "step": 1250 + }, + { + "epoch": 0.1203347441323586, + "grad_norm": 2.3546371222911433, + "learning_rate": 1.957514977255378e-05, + "loss": 1.0545, + "step": 1251 + }, + { + "epoch": 0.12043093497499038, + "grad_norm": 2.2109218215474318, + "learning_rate": 1.957425087113908e-05, + "loss": 0.7962, + "step": 1252 + }, + { + "epoch": 0.12052712581762216, + "grad_norm": 2.4256783295609408, + "learning_rate": 1.9573351040460898e-05, + "loss": 1.1027, + "step": 1253 + }, + { + "epoch": 0.12062331666025394, + "grad_norm": 2.894058491136261, + "learning_rate": 1.9572450280606568e-05, + "loss": 1.1405, + "step": 1254 + }, + { + "epoch": 0.12071950750288572, + "grad_norm": 2.684570154541078, + "learning_rate": 1.957154859166352e-05, + "loss": 1.1254, + "step": 1255 + }, + { + "epoch": 0.1208156983455175, + "grad_norm": 2.7767491762706515, + "learning_rate": 1.9570645973719273e-05, + "loss": 1.0088, + "step": 1256 + }, + { + "epoch": 0.12091188918814928, + "grad_norm": 2.839368043895409, + "learning_rate": 1.9569742426861422e-05, + "loss": 1.0486, + "step": 1257 + }, + { + "epoch": 0.12100808003078108, + "grad_norm": 2.403173888297713, + "learning_rate": 1.9568837951177677e-05, + "loss": 1.1479, + "step": 1258 + }, + { + "epoch": 0.12110427087341286, + "grad_norm": 2.9042200133226608, + "learning_rate": 1.956793254675582e-05, + "loss": 1.0543, + "step": 1259 + }, + { + "epoch": 0.12120046171604464, + "grad_norm": 2.828981012013069, + "learning_rate": 1.9567026213683728e-05, + "loss": 1.0727, + "step": 1260 + }, + { + "epoch": 0.12129665255867642, + "grad_norm": 2.2423783045237538, + "learning_rate": 1.9566118952049368e-05, + "loss": 1.1193, + "step": 1261 + }, + { + "epoch": 0.1213928434013082, + "grad_norm": 2.626206435690357, + "learning_rate": 1.9565210761940798e-05, + "loss": 1.2029, + "step": 1262 + }, + { + "epoch": 0.12148903424393998, + "grad_norm": 2.3797932207273065, + "learning_rate": 1.9564301643446167e-05, + "loss": 0.9817, + "step": 1263 + }, + { + "epoch": 0.12158522508657175, + "grad_norm": 2.6412022792682555, + "learning_rate": 1.956339159665371e-05, + "loss": 1.1214, + "step": 1264 + }, + { + "epoch": 0.12168141592920353, + "grad_norm": 2.538410626269724, + "learning_rate": 1.9562480621651757e-05, + "loss": 1.1414, + "step": 1265 + }, + { + "epoch": 0.12177760677183531, + "grad_norm": 2.648131959114367, + "learning_rate": 1.956156871852873e-05, + "loss": 1.1662, + "step": 1266 + }, + { + "epoch": 0.12187379761446711, + "grad_norm": 2.618811688156537, + "learning_rate": 1.956065588737313e-05, + "loss": 1.1247, + "step": 1267 + }, + { + "epoch": 0.12196998845709889, + "grad_norm": 2.2514989405002046, + "learning_rate": 1.9559742128273558e-05, + "loss": 1.0666, + "step": 1268 + }, + { + "epoch": 0.12206617929973067, + "grad_norm": 2.7418972233166294, + "learning_rate": 1.95588274413187e-05, + "loss": 1.1043, + "step": 1269 + }, + { + "epoch": 0.12216237014236245, + "grad_norm": 2.683224278340472, + "learning_rate": 1.9557911826597337e-05, + "loss": 1.1452, + "step": 1270 + }, + { + "epoch": 0.12225856098499423, + "grad_norm": 2.54914649389246, + "learning_rate": 1.955699528419834e-05, + "loss": 1.0564, + "step": 1271 + }, + { + "epoch": 0.12235475182762601, + "grad_norm": 2.521187585554359, + "learning_rate": 1.9556077814210662e-05, + "loss": 1.0314, + "step": 1272 + }, + { + "epoch": 0.12245094267025779, + "grad_norm": 2.3502599061470737, + "learning_rate": 1.955515941672335e-05, + "loss": 0.8175, + "step": 1273 + }, + { + "epoch": 0.12254713351288957, + "grad_norm": 2.4436053800328255, + "learning_rate": 1.955424009182555e-05, + "loss": 1.1417, + "step": 1274 + }, + { + "epoch": 0.12264332435552136, + "grad_norm": 2.8686031850706195, + "learning_rate": 1.9553319839606484e-05, + "loss": 1.2306, + "step": 1275 + }, + { + "epoch": 0.12273951519815314, + "grad_norm": 2.3745047036896745, + "learning_rate": 1.955239866015547e-05, + "loss": 1.1013, + "step": 1276 + }, + { + "epoch": 0.12283570604078492, + "grad_norm": 2.544127539565482, + "learning_rate": 1.9551476553561925e-05, + "loss": 0.9991, + "step": 1277 + }, + { + "epoch": 0.1229318968834167, + "grad_norm": 2.2125485143362655, + "learning_rate": 1.9550553519915335e-05, + "loss": 0.9052, + "step": 1278 + }, + { + "epoch": 0.12302808772604848, + "grad_norm": 2.413878430437774, + "learning_rate": 1.95496295593053e-05, + "loss": 1.1643, + "step": 1279 + }, + { + "epoch": 0.12312427856868026, + "grad_norm": 2.6459754137743063, + "learning_rate": 1.954870467182149e-05, + "loss": 1.1492, + "step": 1280 + }, + { + "epoch": 0.12322046941131204, + "grad_norm": 2.424714500089933, + "learning_rate": 1.954777885755368e-05, + "loss": 1.1179, + "step": 1281 + }, + { + "epoch": 0.12331666025394382, + "grad_norm": 2.445943200963176, + "learning_rate": 1.954685211659172e-05, + "loss": 1.0595, + "step": 1282 + }, + { + "epoch": 0.12341285109657561, + "grad_norm": 2.4890022595540913, + "learning_rate": 1.9545924449025563e-05, + "loss": 1.0942, + "step": 1283 + }, + { + "epoch": 0.12350904193920739, + "grad_norm": 1.9230310602821366, + "learning_rate": 1.9544995854945248e-05, + "loss": 0.822, + "step": 1284 + }, + { + "epoch": 0.12360523278183917, + "grad_norm": 2.452211366950683, + "learning_rate": 1.95440663344409e-05, + "loss": 1.0474, + "step": 1285 + }, + { + "epoch": 0.12370142362447095, + "grad_norm": 2.6049751580245175, + "learning_rate": 1.954313588760274e-05, + "loss": 1.1396, + "step": 1286 + }, + { + "epoch": 0.12379761446710273, + "grad_norm": 2.5983702139244014, + "learning_rate": 1.954220451452108e-05, + "loss": 1.1661, + "step": 1287 + }, + { + "epoch": 0.12389380530973451, + "grad_norm": 2.3740399906287055, + "learning_rate": 1.9541272215286304e-05, + "loss": 1.086, + "step": 1288 + }, + { + "epoch": 0.12398999615236629, + "grad_norm": 2.447000216960373, + "learning_rate": 1.954033898998891e-05, + "loss": 1.0772, + "step": 1289 + }, + { + "epoch": 0.12408618699499807, + "grad_norm": 2.48540374368655, + "learning_rate": 1.9539404838719477e-05, + "loss": 1.1558, + "step": 1290 + }, + { + "epoch": 0.12418237783762985, + "grad_norm": 2.520365215758472, + "learning_rate": 1.9538469761568666e-05, + "loss": 1.067, + "step": 1291 + }, + { + "epoch": 0.12427856868026164, + "grad_norm": 2.6433349942826507, + "learning_rate": 1.9537533758627242e-05, + "loss": 1.0969, + "step": 1292 + }, + { + "epoch": 0.12437475952289342, + "grad_norm": 2.935528034979957, + "learning_rate": 1.953659682998604e-05, + "loss": 1.0694, + "step": 1293 + }, + { + "epoch": 0.1244709503655252, + "grad_norm": 2.5484345280572347, + "learning_rate": 1.953565897573601e-05, + "loss": 1.085, + "step": 1294 + }, + { + "epoch": 0.12456714120815698, + "grad_norm": 3.029232508888138, + "learning_rate": 1.9534720195968173e-05, + "loss": 1.0617, + "step": 1295 + }, + { + "epoch": 0.12466333205078876, + "grad_norm": 2.3060734894900183, + "learning_rate": 1.9533780490773645e-05, + "loss": 1.1176, + "step": 1296 + }, + { + "epoch": 0.12475952289342054, + "grad_norm": 2.552166136401685, + "learning_rate": 1.9532839860243636e-05, + "loss": 1.0975, + "step": 1297 + }, + { + "epoch": 0.12485571373605232, + "grad_norm": 2.875086526038745, + "learning_rate": 1.9531898304469435e-05, + "loss": 1.0758, + "step": 1298 + }, + { + "epoch": 0.1249519045786841, + "grad_norm": 2.2491092136680257, + "learning_rate": 1.9530955823542433e-05, + "loss": 1.135, + "step": 1299 + }, + { + "epoch": 0.1250480954213159, + "grad_norm": 2.3941963469167886, + "learning_rate": 1.953001241755411e-05, + "loss": 0.9539, + "step": 1300 + }, + { + "epoch": 0.12514428626394766, + "grad_norm": 2.515663785652908, + "learning_rate": 1.9529068086596025e-05, + "loss": 0.9966, + "step": 1301 + }, + { + "epoch": 0.12524047710657946, + "grad_norm": 2.2325121467196123, + "learning_rate": 1.952812283075984e-05, + "loss": 0.905, + "step": 1302 + }, + { + "epoch": 0.12533666794921122, + "grad_norm": 2.3954723064672185, + "learning_rate": 1.952717665013729e-05, + "loss": 1.2203, + "step": 1303 + }, + { + "epoch": 0.12543285879184302, + "grad_norm": 2.3940069859757434, + "learning_rate": 1.952622954482022e-05, + "loss": 1.0837, + "step": 1304 + }, + { + "epoch": 0.1255290496344748, + "grad_norm": 2.316246520049281, + "learning_rate": 1.952528151490055e-05, + "loss": 1.0667, + "step": 1305 + }, + { + "epoch": 0.12562524047710658, + "grad_norm": 2.4196973089214877, + "learning_rate": 1.9524332560470293e-05, + "loss": 1.121, + "step": 1306 + }, + { + "epoch": 0.12572143131973837, + "grad_norm": 2.6284721199606844, + "learning_rate": 1.952338268162156e-05, + "loss": 0.9841, + "step": 1307 + }, + { + "epoch": 0.12581762216237014, + "grad_norm": 2.074403001401577, + "learning_rate": 1.9522431878446536e-05, + "loss": 1.0332, + "step": 1308 + }, + { + "epoch": 0.12591381300500193, + "grad_norm": 2.3399791997503385, + "learning_rate": 1.952148015103751e-05, + "loss": 1.0448, + "step": 1309 + }, + { + "epoch": 0.1260100038476337, + "grad_norm": 2.3964422481063474, + "learning_rate": 1.9520527499486856e-05, + "loss": 1.0893, + "step": 1310 + }, + { + "epoch": 0.1261061946902655, + "grad_norm": 2.3767028949653013, + "learning_rate": 1.9519573923887033e-05, + "loss": 1.1036, + "step": 1311 + }, + { + "epoch": 0.12620238553289725, + "grad_norm": 2.753365811616902, + "learning_rate": 1.95186194243306e-05, + "loss": 1.1693, + "step": 1312 + }, + { + "epoch": 0.12629857637552905, + "grad_norm": 2.720847387745736, + "learning_rate": 1.9517664000910194e-05, + "loss": 1.1372, + "step": 1313 + }, + { + "epoch": 0.12639476721816084, + "grad_norm": 2.4220627880052206, + "learning_rate": 1.9516707653718546e-05, + "loss": 0.946, + "step": 1314 + }, + { + "epoch": 0.1264909580607926, + "grad_norm": 2.9942408935235774, + "learning_rate": 1.951575038284848e-05, + "loss": 1.127, + "step": 1315 + }, + { + "epoch": 0.1265871489034244, + "grad_norm": 2.674805920819459, + "learning_rate": 1.9514792188392914e-05, + "loss": 1.0641, + "step": 1316 + }, + { + "epoch": 0.12668333974605617, + "grad_norm": 2.4001512874633155, + "learning_rate": 1.9513833070444836e-05, + "loss": 1.0893, + "step": 1317 + }, + { + "epoch": 0.12677953058868796, + "grad_norm": 2.413508132405932, + "learning_rate": 1.9512873029097347e-05, + "loss": 1.124, + "step": 1318 + }, + { + "epoch": 0.12687572143131973, + "grad_norm": 2.4838908602637253, + "learning_rate": 1.9511912064443623e-05, + "loss": 1.1018, + "step": 1319 + }, + { + "epoch": 0.12697191227395152, + "grad_norm": 2.5012930641282196, + "learning_rate": 1.9510950176576933e-05, + "loss": 1.0802, + "step": 1320 + }, + { + "epoch": 0.12706810311658331, + "grad_norm": 2.46800143442793, + "learning_rate": 1.950998736559064e-05, + "loss": 0.9893, + "step": 1321 + }, + { + "epoch": 0.12716429395921508, + "grad_norm": 2.498189651843317, + "learning_rate": 1.950902363157819e-05, + "loss": 1.1371, + "step": 1322 + }, + { + "epoch": 0.12726048480184687, + "grad_norm": 2.6463351519671043, + "learning_rate": 1.9508058974633123e-05, + "loss": 1.0646, + "step": 1323 + }, + { + "epoch": 0.12735667564447864, + "grad_norm": 2.555736190640619, + "learning_rate": 1.950709339484907e-05, + "loss": 1.1047, + "step": 1324 + }, + { + "epoch": 0.12745286648711043, + "grad_norm": 2.5758981894590174, + "learning_rate": 1.950612689231974e-05, + "loss": 1.1758, + "step": 1325 + }, + { + "epoch": 0.1275490573297422, + "grad_norm": 2.3375207193642873, + "learning_rate": 1.9505159467138954e-05, + "loss": 1.0941, + "step": 1326 + }, + { + "epoch": 0.127645248172374, + "grad_norm": 2.373860638810503, + "learning_rate": 1.9504191119400594e-05, + "loss": 1.1464, + "step": 1327 + }, + { + "epoch": 0.12774143901500576, + "grad_norm": 3.017164171019496, + "learning_rate": 1.9503221849198655e-05, + "loss": 1.1309, + "step": 1328 + }, + { + "epoch": 0.12783762985763755, + "grad_norm": 2.083615900146798, + "learning_rate": 1.9502251656627216e-05, + "loss": 0.8175, + "step": 1329 + }, + { + "epoch": 0.12793382070026935, + "grad_norm": 2.8005559744497726, + "learning_rate": 1.9501280541780435e-05, + "loss": 1.0936, + "step": 1330 + }, + { + "epoch": 0.1280300115429011, + "grad_norm": 2.648232095654516, + "learning_rate": 1.950030850475257e-05, + "loss": 1.1124, + "step": 1331 + }, + { + "epoch": 0.1281262023855329, + "grad_norm": 2.5527903438910347, + "learning_rate": 1.9499335545637968e-05, + "loss": 1.129, + "step": 1332 + }, + { + "epoch": 0.12822239322816467, + "grad_norm": 2.4645330538222043, + "learning_rate": 1.9498361664531062e-05, + "loss": 1.1658, + "step": 1333 + }, + { + "epoch": 0.12831858407079647, + "grad_norm": 2.485679809700953, + "learning_rate": 1.949738686152637e-05, + "loss": 1.1704, + "step": 1334 + }, + { + "epoch": 0.12841477491342823, + "grad_norm": 2.308975437424871, + "learning_rate": 1.9496411136718513e-05, + "loss": 1.0736, + "step": 1335 + }, + { + "epoch": 0.12851096575606002, + "grad_norm": 2.457448907027355, + "learning_rate": 1.9495434490202188e-05, + "loss": 1.1417, + "step": 1336 + }, + { + "epoch": 0.1286071565986918, + "grad_norm": 2.4427364439448143, + "learning_rate": 1.949445692207219e-05, + "loss": 1.079, + "step": 1337 + }, + { + "epoch": 0.12870334744132358, + "grad_norm": 2.540361992840401, + "learning_rate": 1.94934784324234e-05, + "loss": 1.1382, + "step": 1338 + }, + { + "epoch": 0.12879953828395538, + "grad_norm": 2.303070286680491, + "learning_rate": 1.9492499021350788e-05, + "loss": 1.0896, + "step": 1339 + }, + { + "epoch": 0.12889572912658714, + "grad_norm": 2.8348726443789625, + "learning_rate": 1.9491518688949417e-05, + "loss": 1.0303, + "step": 1340 + }, + { + "epoch": 0.12899191996921894, + "grad_norm": 2.2056567937329996, + "learning_rate": 1.949053743531443e-05, + "loss": 0.9938, + "step": 1341 + }, + { + "epoch": 0.1290881108118507, + "grad_norm": 2.665661715032593, + "learning_rate": 1.9489555260541074e-05, + "loss": 1.0857, + "step": 1342 + }, + { + "epoch": 0.1291843016544825, + "grad_norm": 2.5929830656780277, + "learning_rate": 1.948857216472467e-05, + "loss": 0.9632, + "step": 1343 + }, + { + "epoch": 0.12928049249711426, + "grad_norm": 2.5378099455849967, + "learning_rate": 1.948758814796064e-05, + "loss": 1.1782, + "step": 1344 + }, + { + "epoch": 0.12937668333974606, + "grad_norm": 2.4495619286954913, + "learning_rate": 1.9486603210344494e-05, + "loss": 1.02, + "step": 1345 + }, + { + "epoch": 0.12947287418237785, + "grad_norm": 2.828350165918948, + "learning_rate": 1.9485617351971827e-05, + "loss": 1.037, + "step": 1346 + }, + { + "epoch": 0.12956906502500962, + "grad_norm": 2.438086405686982, + "learning_rate": 1.948463057293832e-05, + "loss": 1.2098, + "step": 1347 + }, + { + "epoch": 0.1296652558676414, + "grad_norm": 2.610674091171775, + "learning_rate": 1.9483642873339753e-05, + "loss": 1.1324, + "step": 1348 + }, + { + "epoch": 0.12976144671027318, + "grad_norm": 2.4115031779403866, + "learning_rate": 1.9482654253271992e-05, + "loss": 1.0028, + "step": 1349 + }, + { + "epoch": 0.12985763755290497, + "grad_norm": 2.2004984011207633, + "learning_rate": 1.9481664712830987e-05, + "loss": 1.099, + "step": 1350 + }, + { + "epoch": 0.12995382839553674, + "grad_norm": 2.34416916947838, + "learning_rate": 1.9480674252112784e-05, + "loss": 1.1014, + "step": 1351 + }, + { + "epoch": 0.13005001923816853, + "grad_norm": 2.380356721460329, + "learning_rate": 1.9479682871213515e-05, + "loss": 1.16, + "step": 1352 + }, + { + "epoch": 0.1301462100808003, + "grad_norm": 2.7215574864042957, + "learning_rate": 1.94786905702294e-05, + "loss": 1.0896, + "step": 1353 + }, + { + "epoch": 0.1302424009234321, + "grad_norm": 2.3163431611805225, + "learning_rate": 1.9477697349256756e-05, + "loss": 1.0138, + "step": 1354 + }, + { + "epoch": 0.13033859176606388, + "grad_norm": 2.5526554418999745, + "learning_rate": 1.947670320839198e-05, + "loss": 1.1752, + "step": 1355 + }, + { + "epoch": 0.13043478260869565, + "grad_norm": 2.5399166434818796, + "learning_rate": 1.947570814773156e-05, + "loss": 1.1203, + "step": 1356 + }, + { + "epoch": 0.13053097345132744, + "grad_norm": 2.4809097730109895, + "learning_rate": 1.9474712167372082e-05, + "loss": 1.1034, + "step": 1357 + }, + { + "epoch": 0.1306271642939592, + "grad_norm": 2.2062070482753033, + "learning_rate": 1.9473715267410206e-05, + "loss": 1.0872, + "step": 1358 + }, + { + "epoch": 0.130723355136591, + "grad_norm": 2.416274364177665, + "learning_rate": 1.9472717447942695e-05, + "loss": 1.0998, + "step": 1359 + }, + { + "epoch": 0.13081954597922277, + "grad_norm": 2.628620425556929, + "learning_rate": 1.9471718709066392e-05, + "loss": 1.2359, + "step": 1360 + }, + { + "epoch": 0.13091573682185456, + "grad_norm": 2.548782451967717, + "learning_rate": 1.947071905087824e-05, + "loss": 1.2363, + "step": 1361 + }, + { + "epoch": 0.13101192766448633, + "grad_norm": 2.384944285241692, + "learning_rate": 1.9469718473475256e-05, + "loss": 1.1213, + "step": 1362 + }, + { + "epoch": 0.13110811850711812, + "grad_norm": 2.477764469962014, + "learning_rate": 1.9468716976954563e-05, + "loss": 1.0677, + "step": 1363 + }, + { + "epoch": 0.13120430934974991, + "grad_norm": 1.9883997850533617, + "learning_rate": 1.9467714561413358e-05, + "loss": 0.9121, + "step": 1364 + }, + { + "epoch": 0.13130050019238168, + "grad_norm": 2.383803204296887, + "learning_rate": 1.9466711226948937e-05, + "loss": 1.0108, + "step": 1365 + }, + { + "epoch": 0.13139669103501347, + "grad_norm": 2.594844484250487, + "learning_rate": 1.9465706973658683e-05, + "loss": 1.1045, + "step": 1366 + }, + { + "epoch": 0.13149288187764524, + "grad_norm": 2.577041777043703, + "learning_rate": 1.946470180164007e-05, + "loss": 1.119, + "step": 1367 + }, + { + "epoch": 0.13158907272027703, + "grad_norm": 2.3940435237970017, + "learning_rate": 1.9463695710990648e-05, + "loss": 1.1168, + "step": 1368 + }, + { + "epoch": 0.1316852635629088, + "grad_norm": 2.225611697748096, + "learning_rate": 1.9462688701808074e-05, + "loss": 1.0816, + "step": 1369 + }, + { + "epoch": 0.1317814544055406, + "grad_norm": 2.3512626947588204, + "learning_rate": 1.946168077419009e-05, + "loss": 1.0573, + "step": 1370 + }, + { + "epoch": 0.1318776452481724, + "grad_norm": 2.204017021588997, + "learning_rate": 1.9460671928234522e-05, + "loss": 1.1309, + "step": 1371 + }, + { + "epoch": 0.13197383609080415, + "grad_norm": 2.2925030284681904, + "learning_rate": 1.9459662164039283e-05, + "loss": 0.9887, + "step": 1372 + }, + { + "epoch": 0.13207002693343595, + "grad_norm": 2.3545386710564316, + "learning_rate": 1.9458651481702385e-05, + "loss": 1.0065, + "step": 1373 + }, + { + "epoch": 0.1321662177760677, + "grad_norm": 2.3680548875079928, + "learning_rate": 1.9457639881321917e-05, + "loss": 1.0824, + "step": 1374 + }, + { + "epoch": 0.1322624086186995, + "grad_norm": 2.2696373525233273, + "learning_rate": 1.945662736299607e-05, + "loss": 1.1288, + "step": 1375 + }, + { + "epoch": 0.13235859946133127, + "grad_norm": 2.477372468695076, + "learning_rate": 1.9455613926823115e-05, + "loss": 0.9843, + "step": 1376 + }, + { + "epoch": 0.13245479030396307, + "grad_norm": 3.083293775124587, + "learning_rate": 1.9454599572901412e-05, + "loss": 1.2147, + "step": 1377 + }, + { + "epoch": 0.13255098114659483, + "grad_norm": 2.7612685386456164, + "learning_rate": 1.945358430132942e-05, + "loss": 1.1063, + "step": 1378 + }, + { + "epoch": 0.13264717198922663, + "grad_norm": 2.3748416365123455, + "learning_rate": 1.945256811220567e-05, + "loss": 1.1404, + "step": 1379 + }, + { + "epoch": 0.13274336283185842, + "grad_norm": 2.616959813064128, + "learning_rate": 1.9451551005628803e-05, + "loss": 1.0984, + "step": 1380 + }, + { + "epoch": 0.13283955367449019, + "grad_norm": 2.2350506553735587, + "learning_rate": 1.9450532981697525e-05, + "loss": 0.9831, + "step": 1381 + }, + { + "epoch": 0.13293574451712198, + "grad_norm": 2.4942205029830165, + "learning_rate": 1.9449514040510654e-05, + "loss": 1.0944, + "step": 1382 + }, + { + "epoch": 0.13303193535975374, + "grad_norm": 2.868192529123299, + "learning_rate": 1.9448494182167085e-05, + "loss": 1.1182, + "step": 1383 + }, + { + "epoch": 0.13312812620238554, + "grad_norm": 2.594408384940818, + "learning_rate": 1.9447473406765803e-05, + "loss": 1.015, + "step": 1384 + }, + { + "epoch": 0.1332243170450173, + "grad_norm": 2.503009295742886, + "learning_rate": 1.9446451714405884e-05, + "loss": 1.1887, + "step": 1385 + }, + { + "epoch": 0.1333205078876491, + "grad_norm": 2.2313356058745577, + "learning_rate": 1.9445429105186487e-05, + "loss": 0.8556, + "step": 1386 + }, + { + "epoch": 0.13341669873028086, + "grad_norm": 2.4763779748660433, + "learning_rate": 1.9444405579206872e-05, + "loss": 1.1008, + "step": 1387 + }, + { + "epoch": 0.13351288957291266, + "grad_norm": 2.1682222570772463, + "learning_rate": 1.9443381136566382e-05, + "loss": 1.0895, + "step": 1388 + }, + { + "epoch": 0.13360908041554445, + "grad_norm": 2.461812402233483, + "learning_rate": 1.944235577736444e-05, + "loss": 1.1364, + "step": 1389 + }, + { + "epoch": 0.13370527125817622, + "grad_norm": 2.570951808854439, + "learning_rate": 1.9441329501700568e-05, + "loss": 1.0198, + "step": 1390 + }, + { + "epoch": 0.133801462100808, + "grad_norm": 2.3448617198691024, + "learning_rate": 1.9440302309674378e-05, + "loss": 1.0542, + "step": 1391 + }, + { + "epoch": 0.13389765294343978, + "grad_norm": 2.1688975707677063, + "learning_rate": 1.943927420138557e-05, + "loss": 1.0946, + "step": 1392 + }, + { + "epoch": 0.13399384378607157, + "grad_norm": 2.7239452953360335, + "learning_rate": 1.9438245176933927e-05, + "loss": 1.1477, + "step": 1393 + }, + { + "epoch": 0.13409003462870334, + "grad_norm": 2.7284801034748236, + "learning_rate": 1.9437215236419322e-05, + "loss": 1.1117, + "step": 1394 + }, + { + "epoch": 0.13418622547133513, + "grad_norm": 2.4308381913862243, + "learning_rate": 1.9436184379941726e-05, + "loss": 1.0655, + "step": 1395 + }, + { + "epoch": 0.13428241631396692, + "grad_norm": 2.428016912019055, + "learning_rate": 1.9435152607601187e-05, + "loss": 1.0174, + "step": 1396 + }, + { + "epoch": 0.1343786071565987, + "grad_norm": 2.269664280020579, + "learning_rate": 1.9434119919497852e-05, + "loss": 1.0131, + "step": 1397 + }, + { + "epoch": 0.13447479799923048, + "grad_norm": 2.592150581869254, + "learning_rate": 1.943308631573195e-05, + "loss": 1.277, + "step": 1398 + }, + { + "epoch": 0.13457098884186225, + "grad_norm": 2.5065318317802205, + "learning_rate": 1.94320517964038e-05, + "loss": 1.0231, + "step": 1399 + }, + { + "epoch": 0.13466717968449404, + "grad_norm": 2.264102116892928, + "learning_rate": 1.9431016361613816e-05, + "loss": 1.0427, + "step": 1400 + }, + { + "epoch": 0.1347633705271258, + "grad_norm": 2.5504745034637963, + "learning_rate": 1.9429980011462488e-05, + "loss": 1.059, + "step": 1401 + }, + { + "epoch": 0.1348595613697576, + "grad_norm": 2.654201395858837, + "learning_rate": 1.9428942746050406e-05, + "loss": 1.1603, + "step": 1402 + }, + { + "epoch": 0.13495575221238937, + "grad_norm": 2.497185584519983, + "learning_rate": 1.9427904565478245e-05, + "loss": 1.0609, + "step": 1403 + }, + { + "epoch": 0.13505194305502116, + "grad_norm": 2.3016787252812474, + "learning_rate": 1.9426865469846773e-05, + "loss": 1.0953, + "step": 1404 + }, + { + "epoch": 0.13514813389765296, + "grad_norm": 2.4497387633507266, + "learning_rate": 1.9425825459256844e-05, + "loss": 1.0527, + "step": 1405 + }, + { + "epoch": 0.13524432474028472, + "grad_norm": 2.5718529484089396, + "learning_rate": 1.9424784533809393e-05, + "loss": 1.0529, + "step": 1406 + }, + { + "epoch": 0.13534051558291652, + "grad_norm": 2.202209848862309, + "learning_rate": 1.9423742693605455e-05, + "loss": 1.0055, + "step": 1407 + }, + { + "epoch": 0.13543670642554828, + "grad_norm": 2.3872996212722453, + "learning_rate": 1.942269993874615e-05, + "loss": 1.2077, + "step": 1408 + }, + { + "epoch": 0.13553289726818007, + "grad_norm": 2.6517191792846404, + "learning_rate": 1.9421656269332686e-05, + "loss": 1.0249, + "step": 1409 + }, + { + "epoch": 0.13562908811081184, + "grad_norm": 2.5060156212065112, + "learning_rate": 1.9420611685466358e-05, + "loss": 1.1325, + "step": 1410 + }, + { + "epoch": 0.13572527895344363, + "grad_norm": 2.5591029795803157, + "learning_rate": 1.9419566187248556e-05, + "loss": 1.1562, + "step": 1411 + }, + { + "epoch": 0.1358214697960754, + "grad_norm": 2.2372902360471, + "learning_rate": 1.9418519774780748e-05, + "loss": 1.1373, + "step": 1412 + }, + { + "epoch": 0.1359176606387072, + "grad_norm": 2.882002700317971, + "learning_rate": 1.9417472448164504e-05, + "loss": 0.8626, + "step": 1413 + }, + { + "epoch": 0.136013851481339, + "grad_norm": 2.3832546953135765, + "learning_rate": 1.9416424207501474e-05, + "loss": 1.0633, + "step": 1414 + }, + { + "epoch": 0.13611004232397075, + "grad_norm": 2.777082067306482, + "learning_rate": 1.9415375052893398e-05, + "loss": 1.1909, + "step": 1415 + }, + { + "epoch": 0.13620623316660255, + "grad_norm": 2.6359668962141076, + "learning_rate": 1.9414324984442102e-05, + "loss": 1.1662, + "step": 1416 + }, + { + "epoch": 0.1363024240092343, + "grad_norm": 2.5833602152580037, + "learning_rate": 1.9413274002249514e-05, + "loss": 1.0671, + "step": 1417 + }, + { + "epoch": 0.1363986148518661, + "grad_norm": 2.4647415121739833, + "learning_rate": 1.9412222106417632e-05, + "loss": 0.9774, + "step": 1418 + }, + { + "epoch": 0.13649480569449787, + "grad_norm": 2.325932495257822, + "learning_rate": 1.9411169297048552e-05, + "loss": 1.1464, + "step": 1419 + }, + { + "epoch": 0.13659099653712967, + "grad_norm": 2.6045238978107306, + "learning_rate": 1.9410115574244462e-05, + "loss": 0.9819, + "step": 1420 + }, + { + "epoch": 0.13668718737976146, + "grad_norm": 2.1609574215277716, + "learning_rate": 1.9409060938107635e-05, + "loss": 0.9704, + "step": 1421 + }, + { + "epoch": 0.13678337822239323, + "grad_norm": 2.57036285992292, + "learning_rate": 1.9408005388740433e-05, + "loss": 1.0821, + "step": 1422 + }, + { + "epoch": 0.13687956906502502, + "grad_norm": 2.455150558719291, + "learning_rate": 1.94069489262453e-05, + "loss": 0.9772, + "step": 1423 + }, + { + "epoch": 0.13697575990765679, + "grad_norm": 2.658253854074357, + "learning_rate": 1.9405891550724778e-05, + "loss": 0.9515, + "step": 1424 + }, + { + "epoch": 0.13707195075028858, + "grad_norm": 2.656103198136339, + "learning_rate": 1.9404833262281497e-05, + "loss": 0.9819, + "step": 1425 + }, + { + "epoch": 0.13716814159292035, + "grad_norm": 2.2136731910048404, + "learning_rate": 1.940377406101817e-05, + "loss": 1.0477, + "step": 1426 + }, + { + "epoch": 0.13726433243555214, + "grad_norm": 2.3276970665103045, + "learning_rate": 1.9402713947037606e-05, + "loss": 1.1645, + "step": 1427 + }, + { + "epoch": 0.1373605232781839, + "grad_norm": 2.4413259726278884, + "learning_rate": 1.9401652920442694e-05, + "loss": 0.964, + "step": 1428 + }, + { + "epoch": 0.1374567141208157, + "grad_norm": 2.70119860510585, + "learning_rate": 1.9400590981336416e-05, + "loss": 0.9649, + "step": 1429 + }, + { + "epoch": 0.1375529049634475, + "grad_norm": 2.2863879544849004, + "learning_rate": 1.9399528129821842e-05, + "loss": 0.9967, + "step": 1430 + }, + { + "epoch": 0.13764909580607926, + "grad_norm": 2.395233191320371, + "learning_rate": 1.9398464366002136e-05, + "loss": 1.0202, + "step": 1431 + }, + { + "epoch": 0.13774528664871105, + "grad_norm": 2.147941622645407, + "learning_rate": 1.939739968998054e-05, + "loss": 1.076, + "step": 1432 + }, + { + "epoch": 0.13784147749134282, + "grad_norm": 2.7074891268300276, + "learning_rate": 1.9396334101860386e-05, + "loss": 1.1633, + "step": 1433 + }, + { + "epoch": 0.1379376683339746, + "grad_norm": 2.4745714232794684, + "learning_rate": 1.939526760174511e-05, + "loss": 1.1439, + "step": 1434 + }, + { + "epoch": 0.13803385917660638, + "grad_norm": 2.4179501913548807, + "learning_rate": 1.939420018973822e-05, + "loss": 1.0092, + "step": 1435 + }, + { + "epoch": 0.13813005001923817, + "grad_norm": 2.4040867515773203, + "learning_rate": 1.939313186594331e-05, + "loss": 1.0636, + "step": 1436 + }, + { + "epoch": 0.13822624086186994, + "grad_norm": 2.5099088613883107, + "learning_rate": 1.9392062630464085e-05, + "loss": 1.0002, + "step": 1437 + }, + { + "epoch": 0.13832243170450173, + "grad_norm": 2.8906295674834497, + "learning_rate": 1.9390992483404308e-05, + "loss": 1.1872, + "step": 1438 + }, + { + "epoch": 0.13841862254713352, + "grad_norm": 2.900434831562919, + "learning_rate": 1.9389921424867853e-05, + "loss": 0.9068, + "step": 1439 + }, + { + "epoch": 0.1385148133897653, + "grad_norm": 2.5561854119758185, + "learning_rate": 1.938884945495868e-05, + "loss": 1.0984, + "step": 1440 + }, + { + "epoch": 0.13861100423239708, + "grad_norm": 2.7234452503993056, + "learning_rate": 1.938777657378083e-05, + "loss": 1.1976, + "step": 1441 + }, + { + "epoch": 0.13870719507502885, + "grad_norm": 2.2928986044490243, + "learning_rate": 1.9386702781438425e-05, + "loss": 1.0838, + "step": 1442 + }, + { + "epoch": 0.13880338591766064, + "grad_norm": 2.3047872049118925, + "learning_rate": 1.93856280780357e-05, + "loss": 1.1067, + "step": 1443 + }, + { + "epoch": 0.1388995767602924, + "grad_norm": 2.6615896641969092, + "learning_rate": 1.938455246367696e-05, + "loss": 1.0255, + "step": 1444 + }, + { + "epoch": 0.1389957676029242, + "grad_norm": 2.9186185487578244, + "learning_rate": 1.93834759384666e-05, + "loss": 1.1969, + "step": 1445 + }, + { + "epoch": 0.139091958445556, + "grad_norm": 2.3414197731662676, + "learning_rate": 1.9382398502509107e-05, + "loss": 0.9832, + "step": 1446 + }, + { + "epoch": 0.13918814928818776, + "grad_norm": 2.3665743018630105, + "learning_rate": 1.9381320155909057e-05, + "loss": 1.0656, + "step": 1447 + }, + { + "epoch": 0.13928434013081956, + "grad_norm": 2.3947661972613856, + "learning_rate": 1.938024089877111e-05, + "loss": 1.1006, + "step": 1448 + }, + { + "epoch": 0.13938053097345132, + "grad_norm": 2.5239171263790263, + "learning_rate": 1.937916073120002e-05, + "loss": 1.1159, + "step": 1449 + }, + { + "epoch": 0.13947672181608312, + "grad_norm": 2.378515022117774, + "learning_rate": 1.9378079653300624e-05, + "loss": 1.0181, + "step": 1450 + }, + { + "epoch": 0.13957291265871488, + "grad_norm": 2.6185450212750188, + "learning_rate": 1.9376997665177853e-05, + "loss": 1.1626, + "step": 1451 + }, + { + "epoch": 0.13966910350134668, + "grad_norm": 2.4237273346841004, + "learning_rate": 1.9375914766936723e-05, + "loss": 1.1192, + "step": 1452 + }, + { + "epoch": 0.13976529434397844, + "grad_norm": 2.5435441303551976, + "learning_rate": 1.9374830958682334e-05, + "loss": 1.1623, + "step": 1453 + }, + { + "epoch": 0.13986148518661023, + "grad_norm": 2.4895421912657283, + "learning_rate": 1.9373746240519884e-05, + "loss": 1.0913, + "step": 1454 + }, + { + "epoch": 0.13995767602924203, + "grad_norm": 2.3233700447866537, + "learning_rate": 1.9372660612554653e-05, + "loss": 1.0937, + "step": 1455 + }, + { + "epoch": 0.1400538668718738, + "grad_norm": 3.0105589597663713, + "learning_rate": 1.937157407489201e-05, + "loss": 0.9909, + "step": 1456 + }, + { + "epoch": 0.1401500577145056, + "grad_norm": 2.414563621905318, + "learning_rate": 1.937048662763741e-05, + "loss": 1.1042, + "step": 1457 + }, + { + "epoch": 0.14024624855713735, + "grad_norm": 2.1255259736448635, + "learning_rate": 1.9369398270896403e-05, + "loss": 1.1623, + "step": 1458 + }, + { + "epoch": 0.14034243939976915, + "grad_norm": 2.6065559289704274, + "learning_rate": 1.9368309004774623e-05, + "loss": 1.1713, + "step": 1459 + }, + { + "epoch": 0.1404386302424009, + "grad_norm": 2.3039433704074743, + "learning_rate": 1.936721882937779e-05, + "loss": 1.1195, + "step": 1460 + }, + { + "epoch": 0.1405348210850327, + "grad_norm": 2.5341572050382886, + "learning_rate": 1.936612774481172e-05, + "loss": 1.1296, + "step": 1461 + }, + { + "epoch": 0.14063101192766447, + "grad_norm": 2.43902378414851, + "learning_rate": 1.9365035751182307e-05, + "loss": 1.0122, + "step": 1462 + }, + { + "epoch": 0.14072720277029627, + "grad_norm": 2.354047187541459, + "learning_rate": 1.9363942848595544e-05, + "loss": 1.0023, + "step": 1463 + }, + { + "epoch": 0.14082339361292806, + "grad_norm": 2.4701238482859647, + "learning_rate": 1.93628490371575e-05, + "loss": 1.0586, + "step": 1464 + }, + { + "epoch": 0.14091958445555983, + "grad_norm": 1.9918059855571395, + "learning_rate": 1.936175431697434e-05, + "loss": 0.8168, + "step": 1465 + }, + { + "epoch": 0.14101577529819162, + "grad_norm": 2.1776629149958167, + "learning_rate": 1.9360658688152322e-05, + "loss": 1.1543, + "step": 1466 + }, + { + "epoch": 0.14111196614082339, + "grad_norm": 2.4181522205018458, + "learning_rate": 1.9359562150797777e-05, + "loss": 1.0777, + "step": 1467 + }, + { + "epoch": 0.14120815698345518, + "grad_norm": 2.3367095179032105, + "learning_rate": 1.9358464705017143e-05, + "loss": 1.0541, + "step": 1468 + }, + { + "epoch": 0.14130434782608695, + "grad_norm": 2.389922596708334, + "learning_rate": 1.935736635091693e-05, + "loss": 1.0231, + "step": 1469 + }, + { + "epoch": 0.14140053866871874, + "grad_norm": 2.339043101197681, + "learning_rate": 1.9356267088603745e-05, + "loss": 1.105, + "step": 1470 + }, + { + "epoch": 0.14149672951135053, + "grad_norm": 2.2448024210407214, + "learning_rate": 1.9355166918184278e-05, + "loss": 1.0905, + "step": 1471 + }, + { + "epoch": 0.1415929203539823, + "grad_norm": 2.093893934092994, + "learning_rate": 1.9354065839765316e-05, + "loss": 0.8154, + "step": 1472 + }, + { + "epoch": 0.1416891111966141, + "grad_norm": 2.224279032608599, + "learning_rate": 1.9352963853453723e-05, + "loss": 1.0378, + "step": 1473 + }, + { + "epoch": 0.14178530203924586, + "grad_norm": 2.7897788021023247, + "learning_rate": 1.9351860959356462e-05, + "loss": 1.0637, + "step": 1474 + }, + { + "epoch": 0.14188149288187765, + "grad_norm": 2.3943246492451333, + "learning_rate": 1.935075715758057e-05, + "loss": 0.9705, + "step": 1475 + }, + { + "epoch": 0.14197768372450942, + "grad_norm": 2.2197754270616668, + "learning_rate": 1.9349652448233187e-05, + "loss": 1.0658, + "step": 1476 + }, + { + "epoch": 0.1420738745671412, + "grad_norm": 1.8460038586985543, + "learning_rate": 1.934854683142153e-05, + "loss": 0.8355, + "step": 1477 + }, + { + "epoch": 0.14217006540977298, + "grad_norm": 2.3531936389202954, + "learning_rate": 1.934744030725291e-05, + "loss": 1.1375, + "step": 1478 + }, + { + "epoch": 0.14226625625240477, + "grad_norm": 2.2763051896554063, + "learning_rate": 1.934633287583473e-05, + "loss": 1.0246, + "step": 1479 + }, + { + "epoch": 0.14236244709503657, + "grad_norm": 2.1778991352419363, + "learning_rate": 1.934522453727447e-05, + "loss": 1.0958, + "step": 1480 + }, + { + "epoch": 0.14245863793766833, + "grad_norm": 2.068560575618041, + "learning_rate": 1.9344115291679703e-05, + "loss": 1.0809, + "step": 1481 + }, + { + "epoch": 0.14255482878030012, + "grad_norm": 2.4107970463463455, + "learning_rate": 1.93430051391581e-05, + "loss": 1.0255, + "step": 1482 + }, + { + "epoch": 0.1426510196229319, + "grad_norm": 2.5728643943300438, + "learning_rate": 1.9341894079817397e-05, + "loss": 1.1451, + "step": 1483 + }, + { + "epoch": 0.14274721046556368, + "grad_norm": 2.1914784361122397, + "learning_rate": 1.934078211376544e-05, + "loss": 0.9679, + "step": 1484 + }, + { + "epoch": 0.14284340130819545, + "grad_norm": 2.5434275867634346, + "learning_rate": 1.9339669241110153e-05, + "loss": 1.1274, + "step": 1485 + }, + { + "epoch": 0.14293959215082724, + "grad_norm": 2.311991336002969, + "learning_rate": 1.9338555461959554e-05, + "loss": 1.1432, + "step": 1486 + }, + { + "epoch": 0.143035782993459, + "grad_norm": 2.2519386789637577, + "learning_rate": 1.9337440776421738e-05, + "loss": 1.0817, + "step": 1487 + }, + { + "epoch": 0.1431319738360908, + "grad_norm": 2.172070279900757, + "learning_rate": 1.93363251846049e-05, + "loss": 1.0187, + "step": 1488 + }, + { + "epoch": 0.1432281646787226, + "grad_norm": 2.6397263601940253, + "learning_rate": 1.9335208686617317e-05, + "loss": 1.1202, + "step": 1489 + }, + { + "epoch": 0.14332435552135436, + "grad_norm": 2.2873959909461443, + "learning_rate": 1.9334091282567352e-05, + "loss": 1.0496, + "step": 1490 + }, + { + "epoch": 0.14342054636398616, + "grad_norm": 2.3362087446205178, + "learning_rate": 1.9332972972563463e-05, + "loss": 1.0789, + "step": 1491 + }, + { + "epoch": 0.14351673720661792, + "grad_norm": 2.210821664882981, + "learning_rate": 1.9331853756714185e-05, + "loss": 0.9559, + "step": 1492 + }, + { + "epoch": 0.14361292804924972, + "grad_norm": 2.305972497354605, + "learning_rate": 1.9330733635128155e-05, + "loss": 1.0936, + "step": 1493 + }, + { + "epoch": 0.14370911889188148, + "grad_norm": 3.3664600276655765, + "learning_rate": 1.9329612607914088e-05, + "loss": 1.062, + "step": 1494 + }, + { + "epoch": 0.14380530973451328, + "grad_norm": 2.4941063858261785, + "learning_rate": 1.9328490675180783e-05, + "loss": 1.0519, + "step": 1495 + }, + { + "epoch": 0.14390150057714507, + "grad_norm": 2.150964956114842, + "learning_rate": 1.9327367837037142e-05, + "loss": 1.1419, + "step": 1496 + }, + { + "epoch": 0.14399769141977684, + "grad_norm": 2.148655347861453, + "learning_rate": 1.9326244093592145e-05, + "loss": 1.0652, + "step": 1497 + }, + { + "epoch": 0.14409388226240863, + "grad_norm": 2.807718452632195, + "learning_rate": 1.9325119444954855e-05, + "loss": 1.1543, + "step": 1498 + }, + { + "epoch": 0.1441900731050404, + "grad_norm": 2.472302117477317, + "learning_rate": 1.9323993891234437e-05, + "loss": 1.179, + "step": 1499 + }, + { + "epoch": 0.1442862639476722, + "grad_norm": 2.5850935151188135, + "learning_rate": 1.9322867432540126e-05, + "loss": 1.0518, + "step": 1500 + }, + { + "epoch": 0.14438245479030395, + "grad_norm": 2.200551340401236, + "learning_rate": 1.9321740068981267e-05, + "loss": 0.9931, + "step": 1501 + }, + { + "epoch": 0.14447864563293575, + "grad_norm": 2.254833489902627, + "learning_rate": 1.9320611800667268e-05, + "loss": 1.193, + "step": 1502 + }, + { + "epoch": 0.14457483647556751, + "grad_norm": 2.5238686068986325, + "learning_rate": 1.931948262770764e-05, + "loss": 1.0591, + "step": 1503 + }, + { + "epoch": 0.1446710273181993, + "grad_norm": 2.41145396629823, + "learning_rate": 1.9318352550211986e-05, + "loss": 1.0719, + "step": 1504 + }, + { + "epoch": 0.1447672181608311, + "grad_norm": 2.381224864970082, + "learning_rate": 1.9317221568289982e-05, + "loss": 1.1662, + "step": 1505 + }, + { + "epoch": 0.14486340900346287, + "grad_norm": 2.2948457250668786, + "learning_rate": 1.9316089682051403e-05, + "loss": 0.8755, + "step": 1506 + }, + { + "epoch": 0.14495959984609466, + "grad_norm": 2.3176940354848523, + "learning_rate": 1.9314956891606108e-05, + "loss": 1.0166, + "step": 1507 + }, + { + "epoch": 0.14505579068872643, + "grad_norm": 2.2608682695247886, + "learning_rate": 1.9313823197064042e-05, + "loss": 0.8953, + "step": 1508 + }, + { + "epoch": 0.14515198153135822, + "grad_norm": 2.44815651958387, + "learning_rate": 1.931268859853525e-05, + "loss": 1.0725, + "step": 1509 + }, + { + "epoch": 0.14524817237399, + "grad_norm": 2.303910337875593, + "learning_rate": 1.9311553096129835e-05, + "loss": 1.1039, + "step": 1510 + }, + { + "epoch": 0.14534436321662178, + "grad_norm": 2.4440741708617293, + "learning_rate": 1.9310416689958024e-05, + "loss": 1.0532, + "step": 1511 + }, + { + "epoch": 0.14544055405925355, + "grad_norm": 2.254973121675276, + "learning_rate": 1.9309279380130112e-05, + "loss": 1.0958, + "step": 1512 + }, + { + "epoch": 0.14553674490188534, + "grad_norm": 2.212442125985351, + "learning_rate": 1.930814116675648e-05, + "loss": 1.1026, + "step": 1513 + }, + { + "epoch": 0.14563293574451713, + "grad_norm": 2.3153915751662826, + "learning_rate": 1.93070020499476e-05, + "loss": 1.0087, + "step": 1514 + }, + { + "epoch": 0.1457291265871489, + "grad_norm": 2.36239143440175, + "learning_rate": 1.9305862029814042e-05, + "loss": 1.0244, + "step": 1515 + }, + { + "epoch": 0.1458253174297807, + "grad_norm": 2.2962470891127227, + "learning_rate": 1.930472110646645e-05, + "loss": 1.0692, + "step": 1516 + }, + { + "epoch": 0.14592150827241246, + "grad_norm": 2.3958362819496286, + "learning_rate": 1.9303579280015555e-05, + "loss": 1.1256, + "step": 1517 + }, + { + "epoch": 0.14601769911504425, + "grad_norm": 2.860707572219863, + "learning_rate": 1.9302436550572187e-05, + "loss": 1.1262, + "step": 1518 + }, + { + "epoch": 0.14611388995767602, + "grad_norm": 2.161421993529861, + "learning_rate": 1.930129291824726e-05, + "loss": 1.1829, + "step": 1519 + }, + { + "epoch": 0.1462100808003078, + "grad_norm": 2.3559625260899053, + "learning_rate": 1.930014838315177e-05, + "loss": 1.0149, + "step": 1520 + }, + { + "epoch": 0.1463062716429396, + "grad_norm": 2.5835685001965674, + "learning_rate": 1.92990029453968e-05, + "loss": 0.9675, + "step": 1521 + }, + { + "epoch": 0.14640246248557137, + "grad_norm": 2.5886607190176933, + "learning_rate": 1.9297856605093534e-05, + "loss": 1.0901, + "step": 1522 + }, + { + "epoch": 0.14649865332820317, + "grad_norm": 2.6118940577593355, + "learning_rate": 1.9296709362353226e-05, + "loss": 1.0801, + "step": 1523 + }, + { + "epoch": 0.14659484417083493, + "grad_norm": 2.598402074631037, + "learning_rate": 1.9295561217287226e-05, + "loss": 1.093, + "step": 1524 + }, + { + "epoch": 0.14669103501346673, + "grad_norm": 2.1530016216769, + "learning_rate": 1.929441217000698e-05, + "loss": 1.0982, + "step": 1525 + }, + { + "epoch": 0.1467872258560985, + "grad_norm": 2.4486982683299656, + "learning_rate": 1.9293262220624002e-05, + "loss": 1.0015, + "step": 1526 + }, + { + "epoch": 0.14688341669873028, + "grad_norm": 2.1675160146436396, + "learning_rate": 1.929211136924991e-05, + "loss": 1.1274, + "step": 1527 + }, + { + "epoch": 0.14697960754136205, + "grad_norm": 2.280063071884801, + "learning_rate": 1.9290959615996407e-05, + "loss": 1.0702, + "step": 1528 + }, + { + "epoch": 0.14707579838399384, + "grad_norm": 2.3459039178759786, + "learning_rate": 1.9289806960975273e-05, + "loss": 1.163, + "step": 1529 + }, + { + "epoch": 0.14717198922662564, + "grad_norm": 2.415607526044039, + "learning_rate": 1.9288653404298392e-05, + "loss": 1.0934, + "step": 1530 + }, + { + "epoch": 0.1472681800692574, + "grad_norm": 2.504653179633396, + "learning_rate": 1.928749894607772e-05, + "loss": 1.0039, + "step": 1531 + }, + { + "epoch": 0.1473643709118892, + "grad_norm": 2.2549835219031267, + "learning_rate": 1.9286343586425307e-05, + "loss": 1.0149, + "step": 1532 + }, + { + "epoch": 0.14746056175452096, + "grad_norm": 2.74953481971864, + "learning_rate": 1.9285187325453295e-05, + "loss": 1.0084, + "step": 1533 + }, + { + "epoch": 0.14755675259715276, + "grad_norm": 2.22583599648608, + "learning_rate": 1.9284030163273907e-05, + "loss": 1.0317, + "step": 1534 + }, + { + "epoch": 0.14765294343978452, + "grad_norm": 2.303578683305244, + "learning_rate": 1.928287209999945e-05, + "loss": 1.0929, + "step": 1535 + }, + { + "epoch": 0.14774913428241632, + "grad_norm": 2.3365247026722424, + "learning_rate": 1.9281713135742333e-05, + "loss": 1.0764, + "step": 1536 + }, + { + "epoch": 0.14784532512504808, + "grad_norm": 2.4168754601789617, + "learning_rate": 1.928055327061504e-05, + "loss": 1.1187, + "step": 1537 + }, + { + "epoch": 0.14794151596767988, + "grad_norm": 2.0720473776093384, + "learning_rate": 1.9279392504730147e-05, + "loss": 0.8545, + "step": 1538 + }, + { + "epoch": 0.14803770681031167, + "grad_norm": 2.7922728511641286, + "learning_rate": 1.927823083820031e-05, + "loss": 1.0803, + "step": 1539 + }, + { + "epoch": 0.14813389765294344, + "grad_norm": 2.476700047635903, + "learning_rate": 1.9277068271138287e-05, + "loss": 1.0153, + "step": 1540 + }, + { + "epoch": 0.14823008849557523, + "grad_norm": 2.276584684603541, + "learning_rate": 1.9275904803656912e-05, + "loss": 1.0403, + "step": 1541 + }, + { + "epoch": 0.148326279338207, + "grad_norm": 2.371829014592917, + "learning_rate": 1.9274740435869107e-05, + "loss": 1.0171, + "step": 1542 + }, + { + "epoch": 0.1484224701808388, + "grad_norm": 2.3567441151152297, + "learning_rate": 1.927357516788789e-05, + "loss": 1.0432, + "step": 1543 + }, + { + "epoch": 0.14851866102347056, + "grad_norm": 2.445433185986829, + "learning_rate": 1.927240899982635e-05, + "loss": 1.2203, + "step": 1544 + }, + { + "epoch": 0.14861485186610235, + "grad_norm": 2.6979481214543135, + "learning_rate": 1.9271241931797687e-05, + "loss": 1.1349, + "step": 1545 + }, + { + "epoch": 0.14871104270873411, + "grad_norm": 2.2943747571324025, + "learning_rate": 1.9270073963915162e-05, + "loss": 1.0349, + "step": 1546 + }, + { + "epoch": 0.1488072335513659, + "grad_norm": 2.9078652811490584, + "learning_rate": 1.9268905096292147e-05, + "loss": 1.1114, + "step": 1547 + }, + { + "epoch": 0.1489034243939977, + "grad_norm": 2.8131042801844584, + "learning_rate": 1.9267735329042086e-05, + "loss": 1.1275, + "step": 1548 + }, + { + "epoch": 0.14899961523662947, + "grad_norm": 2.0011999094206767, + "learning_rate": 1.9266564662278515e-05, + "loss": 0.8842, + "step": 1549 + }, + { + "epoch": 0.14909580607926126, + "grad_norm": 2.335339743038342, + "learning_rate": 1.9265393096115056e-05, + "loss": 1.1127, + "step": 1550 + }, + { + "epoch": 0.14919199692189303, + "grad_norm": 2.02745186711538, + "learning_rate": 1.926422063066542e-05, + "loss": 0.9886, + "step": 1551 + }, + { + "epoch": 0.14928818776452482, + "grad_norm": 2.298790345150413, + "learning_rate": 1.926304726604341e-05, + "loss": 1.1144, + "step": 1552 + }, + { + "epoch": 0.1493843786071566, + "grad_norm": 2.4105817109376004, + "learning_rate": 1.9261873002362902e-05, + "loss": 1.0827, + "step": 1553 + }, + { + "epoch": 0.14948056944978838, + "grad_norm": 2.2749831500459137, + "learning_rate": 1.9260697839737875e-05, + "loss": 1.1894, + "step": 1554 + }, + { + "epoch": 0.14957676029242017, + "grad_norm": 2.525869130683203, + "learning_rate": 1.9259521778282388e-05, + "loss": 1.1326, + "step": 1555 + }, + { + "epoch": 0.14967295113505194, + "grad_norm": 2.270984175487789, + "learning_rate": 1.925834481811059e-05, + "loss": 1.1684, + "step": 1556 + }, + { + "epoch": 0.14976914197768373, + "grad_norm": 2.212782866258513, + "learning_rate": 1.9257166959336705e-05, + "loss": 1.0955, + "step": 1557 + }, + { + "epoch": 0.1498653328203155, + "grad_norm": 2.4073162528019254, + "learning_rate": 1.9255988202075065e-05, + "loss": 1.1352, + "step": 1558 + }, + { + "epoch": 0.1499615236629473, + "grad_norm": 2.1841602077949718, + "learning_rate": 1.9254808546440076e-05, + "loss": 0.9832, + "step": 1559 + }, + { + "epoch": 0.15005771450557906, + "grad_norm": 2.1597725227514264, + "learning_rate": 1.925362799254623e-05, + "loss": 0.8888, + "step": 1560 + }, + { + "epoch": 0.15015390534821085, + "grad_norm": 2.527737252470057, + "learning_rate": 1.9252446540508117e-05, + "loss": 1.0904, + "step": 1561 + }, + { + "epoch": 0.15025009619084262, + "grad_norm": 2.745815760469839, + "learning_rate": 1.9251264190440398e-05, + "loss": 1.1074, + "step": 1562 + }, + { + "epoch": 0.1503462870334744, + "grad_norm": 2.2167994609525334, + "learning_rate": 1.925008094245784e-05, + "loss": 1.0978, + "step": 1563 + }, + { + "epoch": 0.1504424778761062, + "grad_norm": 2.2672130475935406, + "learning_rate": 1.9248896796675277e-05, + "loss": 1.1384, + "step": 1564 + }, + { + "epoch": 0.15053866871873797, + "grad_norm": 2.7255194993718344, + "learning_rate": 1.924771175320765e-05, + "loss": 1.0912, + "step": 1565 + }, + { + "epoch": 0.15063485956136977, + "grad_norm": 2.211465390931834, + "learning_rate": 1.924652581216997e-05, + "loss": 1.011, + "step": 1566 + }, + { + "epoch": 0.15073105040400153, + "grad_norm": 2.2853659756386513, + "learning_rate": 1.9245338973677347e-05, + "loss": 1.0541, + "step": 1567 + }, + { + "epoch": 0.15082724124663333, + "grad_norm": 2.7391754034983493, + "learning_rate": 1.9244151237844975e-05, + "loss": 1.0762, + "step": 1568 + }, + { + "epoch": 0.1509234320892651, + "grad_norm": 2.3567166005918025, + "learning_rate": 1.9242962604788133e-05, + "loss": 0.9242, + "step": 1569 + }, + { + "epoch": 0.15101962293189689, + "grad_norm": 2.3944704685832434, + "learning_rate": 1.9241773074622182e-05, + "loss": 1.1627, + "step": 1570 + }, + { + "epoch": 0.15111581377452865, + "grad_norm": 2.5750864063080705, + "learning_rate": 1.9240582647462587e-05, + "loss": 1.0994, + "step": 1571 + }, + { + "epoch": 0.15121200461716044, + "grad_norm": 2.6828718675153866, + "learning_rate": 1.923939132342488e-05, + "loss": 1.0553, + "step": 1572 + }, + { + "epoch": 0.15130819545979224, + "grad_norm": 2.5767211227345577, + "learning_rate": 1.9238199102624695e-05, + "loss": 1.0301, + "step": 1573 + }, + { + "epoch": 0.151404386302424, + "grad_norm": 2.2510001426780653, + "learning_rate": 1.923700598517775e-05, + "loss": 1.1196, + "step": 1574 + }, + { + "epoch": 0.1515005771450558, + "grad_norm": 2.576690605640057, + "learning_rate": 1.9235811971199835e-05, + "loss": 1.0976, + "step": 1575 + }, + { + "epoch": 0.15159676798768756, + "grad_norm": 2.325179175510063, + "learning_rate": 1.923461706080685e-05, + "loss": 1.1088, + "step": 1576 + }, + { + "epoch": 0.15169295883031936, + "grad_norm": 2.615149903324325, + "learning_rate": 1.9233421254114765e-05, + "loss": 1.1176, + "step": 1577 + }, + { + "epoch": 0.15178914967295112, + "grad_norm": 2.5649974320517974, + "learning_rate": 1.923222455123965e-05, + "loss": 1.1126, + "step": 1578 + }, + { + "epoch": 0.15188534051558292, + "grad_norm": 2.3360431246753928, + "learning_rate": 1.9231026952297652e-05, + "loss": 1.1488, + "step": 1579 + }, + { + "epoch": 0.1519815313582147, + "grad_norm": 2.307382366338224, + "learning_rate": 1.9229828457405005e-05, + "loss": 1.0386, + "step": 1580 + }, + { + "epoch": 0.15207772220084648, + "grad_norm": 3.2965690804587413, + "learning_rate": 1.922862906667804e-05, + "loss": 0.9454, + "step": 1581 + }, + { + "epoch": 0.15217391304347827, + "grad_norm": 2.2701153270451706, + "learning_rate": 1.9227428780233162e-05, + "loss": 1.1051, + "step": 1582 + }, + { + "epoch": 0.15227010388611004, + "grad_norm": 2.3354418415639513, + "learning_rate": 1.9226227598186874e-05, + "loss": 1.0955, + "step": 1583 + }, + { + "epoch": 0.15236629472874183, + "grad_norm": 2.3963205273124513, + "learning_rate": 1.922502552065576e-05, + "loss": 0.9964, + "step": 1584 + }, + { + "epoch": 0.1524624855713736, + "grad_norm": 2.2611857930154207, + "learning_rate": 1.9223822547756485e-05, + "loss": 0.9791, + "step": 1585 + }, + { + "epoch": 0.1525586764140054, + "grad_norm": 2.407351790601184, + "learning_rate": 1.922261867960582e-05, + "loss": 1.0167, + "step": 1586 + }, + { + "epoch": 0.15265486725663716, + "grad_norm": 2.9473651327846055, + "learning_rate": 1.9221413916320602e-05, + "loss": 1.1732, + "step": 1587 + }, + { + "epoch": 0.15275105809926895, + "grad_norm": 2.2862511717752154, + "learning_rate": 1.9220208258017763e-05, + "loss": 1.1342, + "step": 1588 + }, + { + "epoch": 0.15284724894190074, + "grad_norm": 2.3382341695269218, + "learning_rate": 1.921900170481433e-05, + "loss": 0.9999, + "step": 1589 + }, + { + "epoch": 0.1529434397845325, + "grad_norm": 2.3897333558254337, + "learning_rate": 1.92177942568274e-05, + "loss": 1.0986, + "step": 1590 + }, + { + "epoch": 0.1530396306271643, + "grad_norm": 2.393270039300857, + "learning_rate": 1.9216585914174177e-05, + "loss": 1.0461, + "step": 1591 + }, + { + "epoch": 0.15313582146979607, + "grad_norm": 2.632238438641255, + "learning_rate": 1.921537667697193e-05, + "loss": 1.0354, + "step": 1592 + }, + { + "epoch": 0.15323201231242786, + "grad_norm": 2.3602277367371474, + "learning_rate": 1.9214166545338036e-05, + "loss": 1.0374, + "step": 1593 + }, + { + "epoch": 0.15332820315505963, + "grad_norm": 2.371038192172158, + "learning_rate": 1.9212955519389938e-05, + "loss": 1.1258, + "step": 1594 + }, + { + "epoch": 0.15342439399769142, + "grad_norm": 2.354745661984924, + "learning_rate": 1.921174359924519e-05, + "loss": 1.1663, + "step": 1595 + }, + { + "epoch": 0.1535205848403232, + "grad_norm": 2.039428100574102, + "learning_rate": 1.9210530785021405e-05, + "loss": 1.0871, + "step": 1596 + }, + { + "epoch": 0.15361677568295498, + "grad_norm": 2.3754703776066437, + "learning_rate": 1.9209317076836306e-05, + "loss": 1.0928, + "step": 1597 + }, + { + "epoch": 0.15371296652558677, + "grad_norm": 2.7488004193576403, + "learning_rate": 1.9208102474807692e-05, + "loss": 1.0528, + "step": 1598 + }, + { + "epoch": 0.15380915736821854, + "grad_norm": 2.526858581780953, + "learning_rate": 1.920688697905345e-05, + "loss": 1.183, + "step": 1599 + }, + { + "epoch": 0.15390534821085033, + "grad_norm": 2.472042300119515, + "learning_rate": 1.920567058969155e-05, + "loss": 1.1303, + "step": 1600 + }, + { + "epoch": 0.1540015390534821, + "grad_norm": 2.378751390095203, + "learning_rate": 1.9204453306840064e-05, + "loss": 1.1024, + "step": 1601 + }, + { + "epoch": 0.1540977298961139, + "grad_norm": 2.224234633764299, + "learning_rate": 1.920323513061713e-05, + "loss": 0.9535, + "step": 1602 + }, + { + "epoch": 0.15419392073874566, + "grad_norm": 2.7357996584443125, + "learning_rate": 1.9202016061140987e-05, + "loss": 1.0762, + "step": 1603 + }, + { + "epoch": 0.15429011158137745, + "grad_norm": 2.166952623877119, + "learning_rate": 1.9200796098529956e-05, + "loss": 0.9979, + "step": 1604 + }, + { + "epoch": 0.15438630242400925, + "grad_norm": 2.3850676259003785, + "learning_rate": 1.9199575242902442e-05, + "loss": 1.114, + "step": 1605 + }, + { + "epoch": 0.154482493266641, + "grad_norm": 2.690432796269543, + "learning_rate": 1.919835349437694e-05, + "loss": 1.121, + "step": 1606 + }, + { + "epoch": 0.1545786841092728, + "grad_norm": 2.258792004885156, + "learning_rate": 1.9197130853072033e-05, + "loss": 0.9606, + "step": 1607 + }, + { + "epoch": 0.15467487495190457, + "grad_norm": 2.3600759639124114, + "learning_rate": 1.9195907319106394e-05, + "loss": 1.1212, + "step": 1608 + }, + { + "epoch": 0.15477106579453637, + "grad_norm": 2.188496228739694, + "learning_rate": 1.9194682892598765e-05, + "loss": 1.0109, + "step": 1609 + }, + { + "epoch": 0.15486725663716813, + "grad_norm": 2.439309526473548, + "learning_rate": 1.9193457573667996e-05, + "loss": 0.9378, + "step": 1610 + }, + { + "epoch": 0.15496344747979993, + "grad_norm": 2.6535969642420145, + "learning_rate": 1.9192231362433015e-05, + "loss": 1.1348, + "step": 1611 + }, + { + "epoch": 0.1550596383224317, + "grad_norm": 2.592434191696967, + "learning_rate": 1.919100425901283e-05, + "loss": 1.0699, + "step": 1612 + }, + { + "epoch": 0.15515582916506349, + "grad_norm": 2.126024559398445, + "learning_rate": 1.918977626352655e-05, + "loss": 1.0606, + "step": 1613 + }, + { + "epoch": 0.15525202000769528, + "grad_norm": 2.3022004279623767, + "learning_rate": 1.9188547376093355e-05, + "loss": 1.0386, + "step": 1614 + }, + { + "epoch": 0.15534821085032705, + "grad_norm": 2.758245357602137, + "learning_rate": 1.9187317596832525e-05, + "loss": 1.15, + "step": 1615 + }, + { + "epoch": 0.15544440169295884, + "grad_norm": 2.053156133284418, + "learning_rate": 1.918608692586342e-05, + "loss": 0.8101, + "step": 1616 + }, + { + "epoch": 0.1555405925355906, + "grad_norm": 2.688718266777718, + "learning_rate": 1.9184855363305482e-05, + "loss": 1.1561, + "step": 1617 + }, + { + "epoch": 0.1556367833782224, + "grad_norm": 2.362106701300897, + "learning_rate": 1.918362290927825e-05, + "loss": 1.1449, + "step": 1618 + }, + { + "epoch": 0.15573297422085416, + "grad_norm": 2.336102973415394, + "learning_rate": 1.9182389563901344e-05, + "loss": 1.0664, + "step": 1619 + }, + { + "epoch": 0.15582916506348596, + "grad_norm": 2.2812069579707477, + "learning_rate": 1.9181155327294468e-05, + "loss": 1.086, + "step": 1620 + }, + { + "epoch": 0.15592535590611772, + "grad_norm": 2.8543827836326656, + "learning_rate": 1.917992019957742e-05, + "loss": 1.209, + "step": 1621 + }, + { + "epoch": 0.15602154674874952, + "grad_norm": 2.9021857683777186, + "learning_rate": 1.9178684180870072e-05, + "loss": 1.1218, + "step": 1622 + }, + { + "epoch": 0.1561177375913813, + "grad_norm": 2.270213045563028, + "learning_rate": 1.9177447271292397e-05, + "loss": 1.0429, + "step": 1623 + }, + { + "epoch": 0.15621392843401308, + "grad_norm": 2.2250365357728414, + "learning_rate": 1.9176209470964446e-05, + "loss": 1.0036, + "step": 1624 + }, + { + "epoch": 0.15631011927664487, + "grad_norm": 2.45110096867353, + "learning_rate": 1.9174970780006356e-05, + "loss": 0.9767, + "step": 1625 + }, + { + "epoch": 0.15640631011927664, + "grad_norm": 2.3009897057586715, + "learning_rate": 1.9173731198538354e-05, + "loss": 1.0274, + "step": 1626 + }, + { + "epoch": 0.15650250096190843, + "grad_norm": 2.325428040082247, + "learning_rate": 1.9172490726680756e-05, + "loss": 1.0005, + "step": 1627 + }, + { + "epoch": 0.1565986918045402, + "grad_norm": 2.283057039351701, + "learning_rate": 1.9171249364553956e-05, + "loss": 1.171, + "step": 1628 + }, + { + "epoch": 0.156694882647172, + "grad_norm": 2.74087790225205, + "learning_rate": 1.9170007112278436e-05, + "loss": 1.042, + "step": 1629 + }, + { + "epoch": 0.15679107348980378, + "grad_norm": 2.433312675641809, + "learning_rate": 1.9168763969974773e-05, + "loss": 1.1902, + "step": 1630 + }, + { + "epoch": 0.15688726433243555, + "grad_norm": 2.3146130378785323, + "learning_rate": 1.9167519937763624e-05, + "loss": 1.0642, + "step": 1631 + }, + { + "epoch": 0.15698345517506734, + "grad_norm": 2.7669184186351963, + "learning_rate": 1.916627501576573e-05, + "loss": 1.104, + "step": 1632 + }, + { + "epoch": 0.1570796460176991, + "grad_norm": 2.2695330005705303, + "learning_rate": 1.9165029204101923e-05, + "loss": 1.1059, + "step": 1633 + }, + { + "epoch": 0.1571758368603309, + "grad_norm": 2.2385506236330155, + "learning_rate": 1.916378250289312e-05, + "loss": 0.9899, + "step": 1634 + }, + { + "epoch": 0.15727202770296267, + "grad_norm": 2.3250383722374552, + "learning_rate": 1.9162534912260327e-05, + "loss": 1.0911, + "step": 1635 + }, + { + "epoch": 0.15736821854559446, + "grad_norm": 2.395115901706066, + "learning_rate": 1.9161286432324628e-05, + "loss": 1.14, + "step": 1636 + }, + { + "epoch": 0.15746440938822623, + "grad_norm": 2.640256990606303, + "learning_rate": 1.91600370632072e-05, + "loss": 0.9985, + "step": 1637 + }, + { + "epoch": 0.15756060023085802, + "grad_norm": 2.471902180946513, + "learning_rate": 1.9158786805029307e-05, + "loss": 1.0453, + "step": 1638 + }, + { + "epoch": 0.15765679107348982, + "grad_norm": 2.353389693115305, + "learning_rate": 1.9157535657912296e-05, + "loss": 1.0237, + "step": 1639 + }, + { + "epoch": 0.15775298191612158, + "grad_norm": 2.3277377071587213, + "learning_rate": 1.9156283621977603e-05, + "loss": 1.1325, + "step": 1640 + }, + { + "epoch": 0.15784917275875338, + "grad_norm": 2.3946922864605793, + "learning_rate": 1.9155030697346746e-05, + "loss": 1.0637, + "step": 1641 + }, + { + "epoch": 0.15794536360138514, + "grad_norm": 2.6327577627347925, + "learning_rate": 1.9153776884141336e-05, + "loss": 1.2346, + "step": 1642 + }, + { + "epoch": 0.15804155444401694, + "grad_norm": 2.4152829598230356, + "learning_rate": 1.9152522182483064e-05, + "loss": 1.1233, + "step": 1643 + }, + { + "epoch": 0.1581377452866487, + "grad_norm": 2.3060135723842206, + "learning_rate": 1.915126659249371e-05, + "loss": 1.0322, + "step": 1644 + }, + { + "epoch": 0.1582339361292805, + "grad_norm": 1.9911566841131927, + "learning_rate": 1.9150010114295138e-05, + "loss": 0.9594, + "step": 1645 + }, + { + "epoch": 0.15833012697191226, + "grad_norm": 2.1802847984775697, + "learning_rate": 1.9148752748009304e-05, + "loss": 0.9338, + "step": 1646 + }, + { + "epoch": 0.15842631781454405, + "grad_norm": 2.307077289760488, + "learning_rate": 1.9147494493758242e-05, + "loss": 0.9942, + "step": 1647 + }, + { + "epoch": 0.15852250865717585, + "grad_norm": 2.407673354280498, + "learning_rate": 1.914623535166408e-05, + "loss": 1.0623, + "step": 1648 + }, + { + "epoch": 0.1586186994998076, + "grad_norm": 2.3413770230374755, + "learning_rate": 1.914497532184903e-05, + "loss": 1.0744, + "step": 1649 + }, + { + "epoch": 0.1587148903424394, + "grad_norm": 2.5060182685355277, + "learning_rate": 1.9143714404435382e-05, + "loss": 0.9755, + "step": 1650 + }, + { + "epoch": 0.15881108118507117, + "grad_norm": 2.2105934605734094, + "learning_rate": 1.9142452599545526e-05, + "loss": 0.9895, + "step": 1651 + }, + { + "epoch": 0.15890727202770297, + "grad_norm": 2.139626486391277, + "learning_rate": 1.9141189907301922e-05, + "loss": 1.09, + "step": 1652 + }, + { + "epoch": 0.15900346287033473, + "grad_norm": 1.936102709148605, + "learning_rate": 1.9139926327827135e-05, + "loss": 1.0005, + "step": 1653 + }, + { + "epoch": 0.15909965371296653, + "grad_norm": 2.4282001824098245, + "learning_rate": 1.9138661861243802e-05, + "loss": 1.0973, + "step": 1654 + }, + { + "epoch": 0.15919584455559832, + "grad_norm": 2.277448689606596, + "learning_rate": 1.9137396507674648e-05, + "loss": 1.0058, + "step": 1655 + }, + { + "epoch": 0.1592920353982301, + "grad_norm": 2.2204333317881066, + "learning_rate": 1.913613026724249e-05, + "loss": 1.0907, + "step": 1656 + }, + { + "epoch": 0.15938822624086188, + "grad_norm": 2.5870168805357654, + "learning_rate": 1.913486314007023e-05, + "loss": 1.2493, + "step": 1657 + }, + { + "epoch": 0.15948441708349365, + "grad_norm": 2.402949786266253, + "learning_rate": 1.9133595126280848e-05, + "loss": 1.0616, + "step": 1658 + }, + { + "epoch": 0.15958060792612544, + "grad_norm": 2.2614447095323373, + "learning_rate": 1.9132326225997414e-05, + "loss": 1.0424, + "step": 1659 + }, + { + "epoch": 0.1596767987687572, + "grad_norm": 2.4188365302437846, + "learning_rate": 1.9131056439343095e-05, + "loss": 1.0121, + "step": 1660 + }, + { + "epoch": 0.159772989611389, + "grad_norm": 2.3349505788104272, + "learning_rate": 1.9129785766441123e-05, + "loss": 1.0205, + "step": 1661 + }, + { + "epoch": 0.15986918045402077, + "grad_norm": 2.40971132226489, + "learning_rate": 1.9128514207414838e-05, + "loss": 1.0902, + "step": 1662 + }, + { + "epoch": 0.15996537129665256, + "grad_norm": 2.6009501669036013, + "learning_rate": 1.9127241762387655e-05, + "loss": 1.0934, + "step": 1663 + }, + { + "epoch": 0.16006156213928435, + "grad_norm": 2.199327179250019, + "learning_rate": 1.9125968431483068e-05, + "loss": 0.9466, + "step": 1664 + }, + { + "epoch": 0.16015775298191612, + "grad_norm": 2.7087790414863586, + "learning_rate": 1.912469421482467e-05, + "loss": 1.1776, + "step": 1665 + }, + { + "epoch": 0.1602539438245479, + "grad_norm": 3.186471439351323, + "learning_rate": 1.9123419112536132e-05, + "loss": 1.2518, + "step": 1666 + }, + { + "epoch": 0.16035013466717968, + "grad_norm": 2.210023407167383, + "learning_rate": 1.912214312474122e-05, + "loss": 1.0539, + "step": 1667 + }, + { + "epoch": 0.16044632550981147, + "grad_norm": 2.153340853756053, + "learning_rate": 1.912086625156377e-05, + "loss": 1.238, + "step": 1668 + }, + { + "epoch": 0.16054251635244324, + "grad_norm": 2.322207994528599, + "learning_rate": 1.911958849312772e-05, + "loss": 1.1265, + "step": 1669 + }, + { + "epoch": 0.16063870719507503, + "grad_norm": 2.453111978606264, + "learning_rate": 1.911830984955709e-05, + "loss": 1.1336, + "step": 1670 + }, + { + "epoch": 0.1607348980377068, + "grad_norm": 2.2603018397469157, + "learning_rate": 1.9117030320975975e-05, + "loss": 1.1534, + "step": 1671 + }, + { + "epoch": 0.1608310888803386, + "grad_norm": 2.3252596318778567, + "learning_rate": 1.911574990750857e-05, + "loss": 1.1111, + "step": 1672 + }, + { + "epoch": 0.16092727972297038, + "grad_norm": 2.392653361471519, + "learning_rate": 1.911446860927915e-05, + "loss": 0.844, + "step": 1673 + }, + { + "epoch": 0.16102347056560215, + "grad_norm": 2.583205822561515, + "learning_rate": 1.9113186426412073e-05, + "loss": 1.131, + "step": 1674 + }, + { + "epoch": 0.16111966140823394, + "grad_norm": 2.3219159061967463, + "learning_rate": 1.911190335903179e-05, + "loss": 1.1122, + "step": 1675 + }, + { + "epoch": 0.1612158522508657, + "grad_norm": 2.2052095895555017, + "learning_rate": 1.9110619407262828e-05, + "loss": 1.1201, + "step": 1676 + }, + { + "epoch": 0.1613120430934975, + "grad_norm": 2.410904185593837, + "learning_rate": 1.9109334571229814e-05, + "loss": 1.0854, + "step": 1677 + }, + { + "epoch": 0.16140823393612927, + "grad_norm": 2.205305538390027, + "learning_rate": 1.9108048851057447e-05, + "loss": 1.0465, + "step": 1678 + }, + { + "epoch": 0.16150442477876106, + "grad_norm": 2.481830203341947, + "learning_rate": 1.9106762246870515e-05, + "loss": 1.0857, + "step": 1679 + }, + { + "epoch": 0.16160061562139286, + "grad_norm": 2.0510513126057677, + "learning_rate": 1.9105474758793897e-05, + "loss": 1.1245, + "step": 1680 + }, + { + "epoch": 0.16169680646402462, + "grad_norm": 2.174975603431563, + "learning_rate": 1.9104186386952556e-05, + "loss": 1.0713, + "step": 1681 + }, + { + "epoch": 0.16179299730665642, + "grad_norm": 2.2811936315889465, + "learning_rate": 1.9102897131471536e-05, + "loss": 1.0664, + "step": 1682 + }, + { + "epoch": 0.16188918814928818, + "grad_norm": 2.4645151672754597, + "learning_rate": 1.9101606992475975e-05, + "loss": 1.1068, + "step": 1683 + }, + { + "epoch": 0.16198537899191998, + "grad_norm": 2.47944823539284, + "learning_rate": 1.9100315970091088e-05, + "loss": 1.0195, + "step": 1684 + }, + { + "epoch": 0.16208156983455174, + "grad_norm": 2.3170787535440773, + "learning_rate": 1.9099024064442182e-05, + "loss": 1.1157, + "step": 1685 + }, + { + "epoch": 0.16217776067718354, + "grad_norm": 2.1791036567582713, + "learning_rate": 1.9097731275654645e-05, + "loss": 1.1003, + "step": 1686 + }, + { + "epoch": 0.1622739515198153, + "grad_norm": 2.1303595886282514, + "learning_rate": 1.9096437603853956e-05, + "loss": 1.0092, + "step": 1687 + }, + { + "epoch": 0.1623701423624471, + "grad_norm": 2.2113992275282177, + "learning_rate": 1.909514304916568e-05, + "loss": 1.1287, + "step": 1688 + }, + { + "epoch": 0.1624663332050789, + "grad_norm": 2.788933138729248, + "learning_rate": 1.9093847611715457e-05, + "loss": 1.175, + "step": 1689 + }, + { + "epoch": 0.16256252404771065, + "grad_norm": 2.257648931387764, + "learning_rate": 1.9092551291629026e-05, + "loss": 1.1099, + "step": 1690 + }, + { + "epoch": 0.16265871489034245, + "grad_norm": 2.4846410828362253, + "learning_rate": 1.9091254089032204e-05, + "loss": 1.1872, + "step": 1691 + }, + { + "epoch": 0.16275490573297421, + "grad_norm": 2.2128580468198646, + "learning_rate": 1.9089956004050893e-05, + "loss": 1.136, + "step": 1692 + }, + { + "epoch": 0.162851096575606, + "grad_norm": 2.358714615625096, + "learning_rate": 1.908865703681109e-05, + "loss": 1.1471, + "step": 1693 + }, + { + "epoch": 0.16294728741823777, + "grad_norm": 2.4614286308539435, + "learning_rate": 1.908735718743887e-05, + "loss": 1.0624, + "step": 1694 + }, + { + "epoch": 0.16304347826086957, + "grad_norm": 2.156030040976048, + "learning_rate": 1.9086056456060387e-05, + "loss": 1.057, + "step": 1695 + }, + { + "epoch": 0.16313966910350133, + "grad_norm": 2.306379733763104, + "learning_rate": 1.908475484280189e-05, + "loss": 1.062, + "step": 1696 + }, + { + "epoch": 0.16323585994613313, + "grad_norm": 2.4682799366117414, + "learning_rate": 1.9083452347789722e-05, + "loss": 1.0828, + "step": 1697 + }, + { + "epoch": 0.16333205078876492, + "grad_norm": 2.9000055493792916, + "learning_rate": 1.908214897115029e-05, + "loss": 1.0725, + "step": 1698 + }, + { + "epoch": 0.1634282416313967, + "grad_norm": 2.2627673413100067, + "learning_rate": 1.9080844713010104e-05, + "loss": 1.0167, + "step": 1699 + }, + { + "epoch": 0.16352443247402848, + "grad_norm": 2.2192098458173004, + "learning_rate": 1.907953957349575e-05, + "loss": 1.0902, + "step": 1700 + }, + { + "epoch": 0.16362062331666025, + "grad_norm": 3.121742148881612, + "learning_rate": 1.9078233552733904e-05, + "loss": 1.1316, + "step": 1701 + }, + { + "epoch": 0.16371681415929204, + "grad_norm": 2.4401611268078467, + "learning_rate": 1.907692665085133e-05, + "loss": 1.0294, + "step": 1702 + }, + { + "epoch": 0.1638130050019238, + "grad_norm": 2.284376409930131, + "learning_rate": 1.907561886797487e-05, + "loss": 1.1275, + "step": 1703 + }, + { + "epoch": 0.1639091958445556, + "grad_norm": 2.183288683954079, + "learning_rate": 1.9074310204231457e-05, + "loss": 1.1199, + "step": 1704 + }, + { + "epoch": 0.1640053866871874, + "grad_norm": 2.278695947417087, + "learning_rate": 1.9073000659748106e-05, + "loss": 1.0222, + "step": 1705 + }, + { + "epoch": 0.16410157752981916, + "grad_norm": 2.566900899255472, + "learning_rate": 1.9071690234651923e-05, + "loss": 1.0895, + "step": 1706 + }, + { + "epoch": 0.16419776837245095, + "grad_norm": 2.422410756607004, + "learning_rate": 1.9070378929070093e-05, + "loss": 0.957, + "step": 1707 + }, + { + "epoch": 0.16429395921508272, + "grad_norm": 2.2447014693186937, + "learning_rate": 1.9069066743129893e-05, + "loss": 0.9927, + "step": 1708 + }, + { + "epoch": 0.1643901500577145, + "grad_norm": 2.18684267634914, + "learning_rate": 1.906775367695868e-05, + "loss": 1.0791, + "step": 1709 + }, + { + "epoch": 0.16448634090034628, + "grad_norm": 2.7373974625207254, + "learning_rate": 1.90664397306839e-05, + "loss": 1.0273, + "step": 1710 + }, + { + "epoch": 0.16458253174297807, + "grad_norm": 2.4875359477824253, + "learning_rate": 1.9065124904433077e-05, + "loss": 1.0695, + "step": 1711 + }, + { + "epoch": 0.16467872258560984, + "grad_norm": 2.595295124734375, + "learning_rate": 1.9063809198333832e-05, + "loss": 0.9937, + "step": 1712 + }, + { + "epoch": 0.16477491342824163, + "grad_norm": 2.44396427270913, + "learning_rate": 1.9062492612513862e-05, + "loss": 1.1612, + "step": 1713 + }, + { + "epoch": 0.16487110427087343, + "grad_norm": 2.6162189215542626, + "learning_rate": 1.9061175147100957e-05, + "loss": 1.0594, + "step": 1714 + }, + { + "epoch": 0.1649672951135052, + "grad_norm": 2.4512920379501235, + "learning_rate": 1.905985680222299e-05, + "loss": 1.1115, + "step": 1715 + }, + { + "epoch": 0.16506348595613698, + "grad_norm": 2.2733188276940055, + "learning_rate": 1.905853757800791e-05, + "loss": 1.1053, + "step": 1716 + }, + { + "epoch": 0.16515967679876875, + "grad_norm": 2.260416144570957, + "learning_rate": 1.905721747458376e-05, + "loss": 1.1165, + "step": 1717 + }, + { + "epoch": 0.16525586764140054, + "grad_norm": 2.4688246653542607, + "learning_rate": 1.9055896492078675e-05, + "loss": 1.0471, + "step": 1718 + }, + { + "epoch": 0.1653520584840323, + "grad_norm": 2.275468452675372, + "learning_rate": 1.9054574630620862e-05, + "loss": 1.1268, + "step": 1719 + }, + { + "epoch": 0.1654482493266641, + "grad_norm": 2.4476070808130883, + "learning_rate": 1.905325189033862e-05, + "loss": 1.1117, + "step": 1720 + }, + { + "epoch": 0.16554444016929587, + "grad_norm": 2.2407052310327904, + "learning_rate": 1.9051928271360333e-05, + "loss": 1.0901, + "step": 1721 + }, + { + "epoch": 0.16564063101192766, + "grad_norm": 2.603899281200649, + "learning_rate": 1.905060377381447e-05, + "loss": 0.998, + "step": 1722 + }, + { + "epoch": 0.16573682185455946, + "grad_norm": 2.3579187560881194, + "learning_rate": 1.9049278397829583e-05, + "loss": 1.1492, + "step": 1723 + }, + { + "epoch": 0.16583301269719122, + "grad_norm": 2.427741166228005, + "learning_rate": 1.904795214353431e-05, + "loss": 1.1794, + "step": 1724 + }, + { + "epoch": 0.16592920353982302, + "grad_norm": 2.131325935056503, + "learning_rate": 1.904662501105738e-05, + "loss": 1.0498, + "step": 1725 + }, + { + "epoch": 0.16602539438245478, + "grad_norm": 2.0730146740842015, + "learning_rate": 1.90452970005276e-05, + "loss": 0.9939, + "step": 1726 + }, + { + "epoch": 0.16612158522508658, + "grad_norm": 2.3477347942409703, + "learning_rate": 1.9043968112073865e-05, + "loss": 1.0508, + "step": 1727 + }, + { + "epoch": 0.16621777606771834, + "grad_norm": 2.266959415592684, + "learning_rate": 1.9042638345825155e-05, + "loss": 1.1576, + "step": 1728 + }, + { + "epoch": 0.16631396691035014, + "grad_norm": 2.560980981732119, + "learning_rate": 1.9041307701910538e-05, + "loss": 1.0462, + "step": 1729 + }, + { + "epoch": 0.16641015775298193, + "grad_norm": 2.355381632951238, + "learning_rate": 1.9039976180459158e-05, + "loss": 1.1191, + "step": 1730 + }, + { + "epoch": 0.1665063485956137, + "grad_norm": 2.1856750423839975, + "learning_rate": 1.9038643781600257e-05, + "loss": 0.9987, + "step": 1731 + }, + { + "epoch": 0.1666025394382455, + "grad_norm": 2.0182841712436805, + "learning_rate": 1.9037310505463153e-05, + "loss": 1.0592, + "step": 1732 + }, + { + "epoch": 0.16669873028087726, + "grad_norm": 2.25348034036519, + "learning_rate": 1.903597635217725e-05, + "loss": 1.1012, + "step": 1733 + }, + { + "epoch": 0.16679492112350905, + "grad_norm": 2.3822732568085008, + "learning_rate": 1.9034641321872043e-05, + "loss": 1.1289, + "step": 1734 + }, + { + "epoch": 0.16689111196614081, + "grad_norm": 2.2858620195467148, + "learning_rate": 1.903330541467711e-05, + "loss": 1.0814, + "step": 1735 + }, + { + "epoch": 0.1669873028087726, + "grad_norm": 2.5255630095425587, + "learning_rate": 1.9031968630722104e-05, + "loss": 1.0831, + "step": 1736 + }, + { + "epoch": 0.16708349365140437, + "grad_norm": 2.2325462009612282, + "learning_rate": 1.9030630970136778e-05, + "loss": 1.1492, + "step": 1737 + }, + { + "epoch": 0.16717968449403617, + "grad_norm": 2.051715275112847, + "learning_rate": 1.902929243305096e-05, + "loss": 0.7736, + "step": 1738 + }, + { + "epoch": 0.16727587533666796, + "grad_norm": 2.0094974672896933, + "learning_rate": 1.902795301959457e-05, + "loss": 0.8609, + "step": 1739 + }, + { + "epoch": 0.16737206617929973, + "grad_norm": 2.630526420770888, + "learning_rate": 1.902661272989761e-05, + "loss": 1.0812, + "step": 1740 + }, + { + "epoch": 0.16746825702193152, + "grad_norm": 2.3162713518061375, + "learning_rate": 1.902527156409016e-05, + "loss": 1.0675, + "step": 1741 + }, + { + "epoch": 0.1675644478645633, + "grad_norm": 2.747020725885377, + "learning_rate": 1.9023929522302394e-05, + "loss": 1.1747, + "step": 1742 + }, + { + "epoch": 0.16766063870719508, + "grad_norm": 2.5528151249615156, + "learning_rate": 1.9022586604664578e-05, + "loss": 1.0236, + "step": 1743 + }, + { + "epoch": 0.16775682954982685, + "grad_norm": 2.733298478424683, + "learning_rate": 1.9021242811307044e-05, + "loss": 1.1745, + "step": 1744 + }, + { + "epoch": 0.16785302039245864, + "grad_norm": 2.1951438818279527, + "learning_rate": 1.901989814236022e-05, + "loss": 1.0582, + "step": 1745 + }, + { + "epoch": 0.1679492112350904, + "grad_norm": 2.337342165907147, + "learning_rate": 1.901855259795462e-05, + "loss": 1.0607, + "step": 1746 + }, + { + "epoch": 0.1680454020777222, + "grad_norm": 2.231080645924521, + "learning_rate": 1.901720617822084e-05, + "loss": 1.0469, + "step": 1747 + }, + { + "epoch": 0.168141592920354, + "grad_norm": 2.451230061180992, + "learning_rate": 1.9015858883289556e-05, + "loss": 1.1272, + "step": 1748 + }, + { + "epoch": 0.16823778376298576, + "grad_norm": 2.2760500110798842, + "learning_rate": 1.9014510713291547e-05, + "loss": 1.0451, + "step": 1749 + }, + { + "epoch": 0.16833397460561755, + "grad_norm": 2.1668040289871144, + "learning_rate": 1.9013161668357655e-05, + "loss": 0.9996, + "step": 1750 + }, + { + "epoch": 0.16843016544824932, + "grad_norm": 2.5071375439711248, + "learning_rate": 1.9011811748618818e-05, + "loss": 1.0966, + "step": 1751 + }, + { + "epoch": 0.1685263562908811, + "grad_norm": 2.9079878966267843, + "learning_rate": 1.901046095420606e-05, + "loss": 1.1199, + "step": 1752 + }, + { + "epoch": 0.16862254713351288, + "grad_norm": 2.5911192653059105, + "learning_rate": 1.9009109285250483e-05, + "loss": 1.0695, + "step": 1753 + }, + { + "epoch": 0.16871873797614467, + "grad_norm": 2.6452944413148223, + "learning_rate": 1.9007756741883284e-05, + "loss": 1.1376, + "step": 1754 + }, + { + "epoch": 0.16881492881877647, + "grad_norm": 2.5216394216124747, + "learning_rate": 1.9006403324235728e-05, + "loss": 1.0444, + "step": 1755 + }, + { + "epoch": 0.16891111966140823, + "grad_norm": 2.294246178082679, + "learning_rate": 1.9005049032439193e-05, + "loss": 1.1701, + "step": 1756 + }, + { + "epoch": 0.16900731050404003, + "grad_norm": 2.130735256961621, + "learning_rate": 1.9003693866625107e-05, + "loss": 1.0781, + "step": 1757 + }, + { + "epoch": 0.1691035013466718, + "grad_norm": 1.947919224727901, + "learning_rate": 1.9002337826925012e-05, + "loss": 0.9915, + "step": 1758 + }, + { + "epoch": 0.16919969218930359, + "grad_norm": 2.3612941362887683, + "learning_rate": 1.900098091347052e-05, + "loss": 1.039, + "step": 1759 + }, + { + "epoch": 0.16929588303193535, + "grad_norm": 2.4847948185465563, + "learning_rate": 1.899962312639333e-05, + "loss": 1.078, + "step": 1760 + }, + { + "epoch": 0.16939207387456715, + "grad_norm": 2.35188269505103, + "learning_rate": 1.899826446582523e-05, + "loss": 0.9781, + "step": 1761 + }, + { + "epoch": 0.1694882647171989, + "grad_norm": 2.1619449885384583, + "learning_rate": 1.8996904931898085e-05, + "loss": 0.9745, + "step": 1762 + }, + { + "epoch": 0.1695844555598307, + "grad_norm": 2.122770783799061, + "learning_rate": 1.8995544524743855e-05, + "loss": 1.0276, + "step": 1763 + }, + { + "epoch": 0.1696806464024625, + "grad_norm": 2.762081144434978, + "learning_rate": 1.899418324449457e-05, + "loss": 0.9607, + "step": 1764 + }, + { + "epoch": 0.16977683724509426, + "grad_norm": 2.6526117985747195, + "learning_rate": 1.899282109128237e-05, + "loss": 1.0719, + "step": 1765 + }, + { + "epoch": 0.16987302808772606, + "grad_norm": 2.367738816794845, + "learning_rate": 1.8991458065239444e-05, + "loss": 1.1311, + "step": 1766 + }, + { + "epoch": 0.16996921893035782, + "grad_norm": 2.4052226870310682, + "learning_rate": 1.89900941664981e-05, + "loss": 1.0356, + "step": 1767 + }, + { + "epoch": 0.17006540977298962, + "grad_norm": 2.266855311291164, + "learning_rate": 1.8988729395190712e-05, + "loss": 1.0953, + "step": 1768 + }, + { + "epoch": 0.17016160061562138, + "grad_norm": 2.285810038942163, + "learning_rate": 1.898736375144974e-05, + "loss": 0.9018, + "step": 1769 + }, + { + "epoch": 0.17025779145825318, + "grad_norm": 2.1582780127547285, + "learning_rate": 1.8985997235407735e-05, + "loss": 0.8882, + "step": 1770 + }, + { + "epoch": 0.17035398230088494, + "grad_norm": 2.3402931123916866, + "learning_rate": 1.898462984719733e-05, + "loss": 1.0162, + "step": 1771 + }, + { + "epoch": 0.17045017314351674, + "grad_norm": 2.288997717806108, + "learning_rate": 1.898326158695124e-05, + "loss": 1.0946, + "step": 1772 + }, + { + "epoch": 0.17054636398614853, + "grad_norm": 3.633646663256611, + "learning_rate": 1.898189245480226e-05, + "loss": 1.077, + "step": 1773 + }, + { + "epoch": 0.1706425548287803, + "grad_norm": 2.3644759272927804, + "learning_rate": 1.8980522450883287e-05, + "loss": 1.0957, + "step": 1774 + }, + { + "epoch": 0.1707387456714121, + "grad_norm": 2.232244316741146, + "learning_rate": 1.8979151575327287e-05, + "loss": 1.0608, + "step": 1775 + }, + { + "epoch": 0.17083493651404386, + "grad_norm": 2.3884064620486423, + "learning_rate": 1.8977779828267314e-05, + "loss": 1.0215, + "step": 1776 + }, + { + "epoch": 0.17093112735667565, + "grad_norm": 2.0675054624033318, + "learning_rate": 1.8976407209836508e-05, + "loss": 0.9404, + "step": 1777 + }, + { + "epoch": 0.17102731819930742, + "grad_norm": 2.2349374618933133, + "learning_rate": 1.8975033720168094e-05, + "loss": 0.9288, + "step": 1778 + }, + { + "epoch": 0.1711235090419392, + "grad_norm": 2.7390418650454578, + "learning_rate": 1.8973659359395384e-05, + "loss": 1.1602, + "step": 1779 + }, + { + "epoch": 0.17121969988457098, + "grad_norm": 2.1869861471065293, + "learning_rate": 1.897228412765177e-05, + "loss": 1.1597, + "step": 1780 + }, + { + "epoch": 0.17131589072720277, + "grad_norm": 2.376304293116555, + "learning_rate": 1.8970908025070725e-05, + "loss": 1.1151, + "step": 1781 + }, + { + "epoch": 0.17141208156983456, + "grad_norm": 2.0042268062852733, + "learning_rate": 1.896953105178582e-05, + "loss": 1.0756, + "step": 1782 + }, + { + "epoch": 0.17150827241246633, + "grad_norm": 2.238639375087131, + "learning_rate": 1.8968153207930692e-05, + "loss": 1.0404, + "step": 1783 + }, + { + "epoch": 0.17160446325509812, + "grad_norm": 2.324307823230136, + "learning_rate": 1.8966774493639084e-05, + "loss": 0.9403, + "step": 1784 + }, + { + "epoch": 0.1717006540977299, + "grad_norm": 2.5069976523473896, + "learning_rate": 1.8965394909044805e-05, + "loss": 0.9221, + "step": 1785 + }, + { + "epoch": 0.17179684494036168, + "grad_norm": 2.4384769066769887, + "learning_rate": 1.896401445428176e-05, + "loss": 1.0558, + "step": 1786 + }, + { + "epoch": 0.17189303578299345, + "grad_norm": 2.3156876927017476, + "learning_rate": 1.8962633129483925e-05, + "loss": 1.0698, + "step": 1787 + }, + { + "epoch": 0.17198922662562524, + "grad_norm": 2.4502549306381214, + "learning_rate": 1.896125093478538e-05, + "loss": 1.1307, + "step": 1788 + }, + { + "epoch": 0.17208541746825703, + "grad_norm": 2.422035389707054, + "learning_rate": 1.895986787032027e-05, + "loss": 1.0924, + "step": 1789 + }, + { + "epoch": 0.1721816083108888, + "grad_norm": 2.483431896825452, + "learning_rate": 1.895848393622284e-05, + "loss": 1.0856, + "step": 1790 + }, + { + "epoch": 0.1722777991535206, + "grad_norm": 1.9847197588360017, + "learning_rate": 1.895709913262741e-05, + "loss": 1.0422, + "step": 1791 + }, + { + "epoch": 0.17237398999615236, + "grad_norm": 2.7912000083318187, + "learning_rate": 1.895571345966839e-05, + "loss": 1.1131, + "step": 1792 + }, + { + "epoch": 0.17247018083878415, + "grad_norm": 2.4489931612683438, + "learning_rate": 1.8954326917480268e-05, + "loss": 0.9478, + "step": 1793 + }, + { + "epoch": 0.17256637168141592, + "grad_norm": 2.284751347300213, + "learning_rate": 1.8952939506197622e-05, + "loss": 1.0139, + "step": 1794 + }, + { + "epoch": 0.1726625625240477, + "grad_norm": 2.308423987371414, + "learning_rate": 1.8951551225955108e-05, + "loss": 1.0635, + "step": 1795 + }, + { + "epoch": 0.17275875336667948, + "grad_norm": 2.2045437733356956, + "learning_rate": 1.8950162076887477e-05, + "loss": 1.0878, + "step": 1796 + }, + { + "epoch": 0.17285494420931127, + "grad_norm": 2.5095913005932475, + "learning_rate": 1.894877205912955e-05, + "loss": 1.0637, + "step": 1797 + }, + { + "epoch": 0.17295113505194307, + "grad_norm": 2.745493671812561, + "learning_rate": 1.894738117281625e-05, + "loss": 0.9906, + "step": 1798 + }, + { + "epoch": 0.17304732589457483, + "grad_norm": 2.2973975226980903, + "learning_rate": 1.8945989418082565e-05, + "loss": 1.0775, + "step": 1799 + }, + { + "epoch": 0.17314351673720663, + "grad_norm": 2.3542693714141634, + "learning_rate": 1.8944596795063584e-05, + "loss": 1.1268, + "step": 1800 + }, + { + "epoch": 0.1732397075798384, + "grad_norm": 2.50315269796649, + "learning_rate": 1.894320330389447e-05, + "loss": 1.0903, + "step": 1801 + }, + { + "epoch": 0.17333589842247019, + "grad_norm": 2.2060900970481425, + "learning_rate": 1.894180894471047e-05, + "loss": 1.0042, + "step": 1802 + }, + { + "epoch": 0.17343208926510195, + "grad_norm": 2.2161076308616985, + "learning_rate": 1.8940413717646924e-05, + "loss": 1.1032, + "step": 1803 + }, + { + "epoch": 0.17352828010773375, + "grad_norm": 2.21694823047975, + "learning_rate": 1.8939017622839253e-05, + "loss": 1.0339, + "step": 1804 + }, + { + "epoch": 0.1736244709503655, + "grad_norm": 2.473289471677121, + "learning_rate": 1.893762066042295e-05, + "loss": 1.139, + "step": 1805 + }, + { + "epoch": 0.1737206617929973, + "grad_norm": 2.4008648154821044, + "learning_rate": 1.8936222830533613e-05, + "loss": 1.0694, + "step": 1806 + }, + { + "epoch": 0.1738168526356291, + "grad_norm": 2.3327660264845496, + "learning_rate": 1.8934824133306905e-05, + "loss": 0.88, + "step": 1807 + }, + { + "epoch": 0.17391304347826086, + "grad_norm": 2.382359058633383, + "learning_rate": 1.8933424568878586e-05, + "loss": 1.0879, + "step": 1808 + }, + { + "epoch": 0.17400923432089266, + "grad_norm": 2.412794042736415, + "learning_rate": 1.8932024137384495e-05, + "loss": 1.0617, + "step": 1809 + }, + { + "epoch": 0.17410542516352442, + "grad_norm": 2.2536354261810945, + "learning_rate": 1.8930622838960555e-05, + "loss": 1.1618, + "step": 1810 + }, + { + "epoch": 0.17420161600615622, + "grad_norm": 2.1649492631169953, + "learning_rate": 1.8929220673742776e-05, + "loss": 1.0665, + "step": 1811 + }, + { + "epoch": 0.17429780684878798, + "grad_norm": 2.656726273027108, + "learning_rate": 1.8927817641867244e-05, + "loss": 1.1789, + "step": 1812 + }, + { + "epoch": 0.17439399769141978, + "grad_norm": 2.211676556563137, + "learning_rate": 1.8926413743470147e-05, + "loss": 1.1053, + "step": 1813 + }, + { + "epoch": 0.17449018853405157, + "grad_norm": 2.3462267406901294, + "learning_rate": 1.8925008978687737e-05, + "loss": 1.1348, + "step": 1814 + }, + { + "epoch": 0.17458637937668334, + "grad_norm": 2.3175872189860782, + "learning_rate": 1.892360334765636e-05, + "loss": 1.094, + "step": 1815 + }, + { + "epoch": 0.17468257021931513, + "grad_norm": 2.198932351579036, + "learning_rate": 1.8922196850512446e-05, + "loss": 1.0836, + "step": 1816 + }, + { + "epoch": 0.1747787610619469, + "grad_norm": 2.107780811938873, + "learning_rate": 1.8920789487392506e-05, + "loss": 1.0817, + "step": 1817 + }, + { + "epoch": 0.1748749519045787, + "grad_norm": 2.532975479453458, + "learning_rate": 1.8919381258433135e-05, + "loss": 1.1773, + "step": 1818 + }, + { + "epoch": 0.17497114274721046, + "grad_norm": 2.1766055290524093, + "learning_rate": 1.8917972163771016e-05, + "loss": 1.0575, + "step": 1819 + }, + { + "epoch": 0.17506733358984225, + "grad_norm": 2.23314744927023, + "learning_rate": 1.8916562203542916e-05, + "loss": 1.1399, + "step": 1820 + }, + { + "epoch": 0.17516352443247402, + "grad_norm": 2.446504605991098, + "learning_rate": 1.8915151377885685e-05, + "loss": 1.0937, + "step": 1821 + }, + { + "epoch": 0.1752597152751058, + "grad_norm": 2.321184290041674, + "learning_rate": 1.8913739686936244e-05, + "loss": 1.0266, + "step": 1822 + }, + { + "epoch": 0.1753559061177376, + "grad_norm": 2.319449909006931, + "learning_rate": 1.8912327130831624e-05, + "loss": 1.0571, + "step": 1823 + }, + { + "epoch": 0.17545209696036937, + "grad_norm": 2.6007288082884847, + "learning_rate": 1.8910913709708918e-05, + "loss": 1.1216, + "step": 1824 + }, + { + "epoch": 0.17554828780300116, + "grad_norm": 2.3188509316410397, + "learning_rate": 1.8909499423705314e-05, + "loss": 0.9394, + "step": 1825 + }, + { + "epoch": 0.17564447864563293, + "grad_norm": 2.4296514226691994, + "learning_rate": 1.8908084272958077e-05, + "loss": 1.0371, + "step": 1826 + }, + { + "epoch": 0.17574066948826472, + "grad_norm": 2.450306058274327, + "learning_rate": 1.8906668257604565e-05, + "loss": 1.0987, + "step": 1827 + }, + { + "epoch": 0.1758368603308965, + "grad_norm": 2.206131187207759, + "learning_rate": 1.8905251377782206e-05, + "loss": 1.1148, + "step": 1828 + }, + { + "epoch": 0.17593305117352828, + "grad_norm": 2.4630609139078987, + "learning_rate": 1.8903833633628533e-05, + "loss": 1.0881, + "step": 1829 + }, + { + "epoch": 0.17602924201616005, + "grad_norm": 2.744795060905529, + "learning_rate": 1.8902415025281136e-05, + "loss": 1.1679, + "step": 1830 + }, + { + "epoch": 0.17612543285879184, + "grad_norm": 2.3206236815289, + "learning_rate": 1.890099555287771e-05, + "loss": 1.0645, + "step": 1831 + }, + { + "epoch": 0.17622162370142364, + "grad_norm": 2.296553967181465, + "learning_rate": 1.889957521655603e-05, + "loss": 1.1045, + "step": 1832 + }, + { + "epoch": 0.1763178145440554, + "grad_norm": 2.404332012978164, + "learning_rate": 1.8898154016453953e-05, + "loss": 1.132, + "step": 1833 + }, + { + "epoch": 0.1764140053866872, + "grad_norm": 2.3046704724354177, + "learning_rate": 1.8896731952709408e-05, + "loss": 0.9064, + "step": 1834 + }, + { + "epoch": 0.17651019622931896, + "grad_norm": 2.3540725037513703, + "learning_rate": 1.8895309025460428e-05, + "loss": 1.0601, + "step": 1835 + }, + { + "epoch": 0.17660638707195075, + "grad_norm": 2.124528856845999, + "learning_rate": 1.8893885234845117e-05, + "loss": 1.0094, + "step": 1836 + }, + { + "epoch": 0.17670257791458252, + "grad_norm": 2.288298753282544, + "learning_rate": 1.8892460581001667e-05, + "loss": 1.0296, + "step": 1837 + }, + { + "epoch": 0.17679876875721431, + "grad_norm": 2.1088279263570344, + "learning_rate": 1.8891035064068354e-05, + "loss": 1.097, + "step": 1838 + }, + { + "epoch": 0.1768949595998461, + "grad_norm": 2.2441807849123863, + "learning_rate": 1.8889608684183533e-05, + "loss": 1.2076, + "step": 1839 + }, + { + "epoch": 0.17699115044247787, + "grad_norm": 2.5996446762962577, + "learning_rate": 1.888818144148565e-05, + "loss": 1.1909, + "step": 1840 + }, + { + "epoch": 0.17708734128510967, + "grad_norm": 2.4653737457154965, + "learning_rate": 1.888675333611323e-05, + "loss": 1.2006, + "step": 1841 + }, + { + "epoch": 0.17718353212774143, + "grad_norm": 2.5706033089724074, + "learning_rate": 1.888532436820488e-05, + "loss": 0.8344, + "step": 1842 + }, + { + "epoch": 0.17727972297037323, + "grad_norm": 2.4926818298298206, + "learning_rate": 1.88838945378993e-05, + "loss": 1.0661, + "step": 1843 + }, + { + "epoch": 0.177375913813005, + "grad_norm": 2.6591862815090117, + "learning_rate": 1.8882463845335263e-05, + "loss": 1.0653, + "step": 1844 + }, + { + "epoch": 0.1774721046556368, + "grad_norm": 2.570150874737113, + "learning_rate": 1.8881032290651626e-05, + "loss": 1.1081, + "step": 1845 + }, + { + "epoch": 0.17756829549826855, + "grad_norm": 2.5231365318963417, + "learning_rate": 1.8879599873987343e-05, + "loss": 1.0753, + "step": 1846 + }, + { + "epoch": 0.17766448634090035, + "grad_norm": 2.4041708117117744, + "learning_rate": 1.8878166595481437e-05, + "loss": 1.1051, + "step": 1847 + }, + { + "epoch": 0.17776067718353214, + "grad_norm": 2.1363407087064377, + "learning_rate": 1.8876732455273022e-05, + "loss": 0.9597, + "step": 1848 + }, + { + "epoch": 0.1778568680261639, + "grad_norm": 2.051957844822382, + "learning_rate": 1.8875297453501295e-05, + "loss": 1.0787, + "step": 1849 + }, + { + "epoch": 0.1779530588687957, + "grad_norm": 2.1292838230007103, + "learning_rate": 1.8873861590305527e-05, + "loss": 1.0699, + "step": 1850 + }, + { + "epoch": 0.17804924971142747, + "grad_norm": 2.481514984931377, + "learning_rate": 1.8872424865825093e-05, + "loss": 1.0889, + "step": 1851 + }, + { + "epoch": 0.17814544055405926, + "grad_norm": 2.120903289356733, + "learning_rate": 1.8870987280199428e-05, + "loss": 0.979, + "step": 1852 + }, + { + "epoch": 0.17824163139669102, + "grad_norm": 2.0794556823809507, + "learning_rate": 1.886954883356807e-05, + "loss": 1.0908, + "step": 1853 + }, + { + "epoch": 0.17833782223932282, + "grad_norm": 2.904799655696239, + "learning_rate": 1.886810952607063e-05, + "loss": 1.0377, + "step": 1854 + }, + { + "epoch": 0.17843401308195458, + "grad_norm": 2.466616242868207, + "learning_rate": 1.8866669357846804e-05, + "loss": 1.046, + "step": 1855 + }, + { + "epoch": 0.17853020392458638, + "grad_norm": 2.196647998460462, + "learning_rate": 1.8865228329036372e-05, + "loss": 0.9856, + "step": 1856 + }, + { + "epoch": 0.17862639476721817, + "grad_norm": 2.675601987636162, + "learning_rate": 1.8863786439779204e-05, + "loss": 0.9074, + "step": 1857 + }, + { + "epoch": 0.17872258560984994, + "grad_norm": 2.4332546226985206, + "learning_rate": 1.886234369021524e-05, + "loss": 1.0134, + "step": 1858 + }, + { + "epoch": 0.17881877645248173, + "grad_norm": 2.6759099211967072, + "learning_rate": 1.886090008048452e-05, + "loss": 1.1267, + "step": 1859 + }, + { + "epoch": 0.1789149672951135, + "grad_norm": 2.611867196445604, + "learning_rate": 1.885945561072715e-05, + "loss": 1.0554, + "step": 1860 + }, + { + "epoch": 0.1790111581377453, + "grad_norm": 2.5231548624426074, + "learning_rate": 1.8858010281083335e-05, + "loss": 1.0522, + "step": 1861 + }, + { + "epoch": 0.17910734898037706, + "grad_norm": 2.313491692402073, + "learning_rate": 1.885656409169335e-05, + "loss": 1.0352, + "step": 1862 + }, + { + "epoch": 0.17920353982300885, + "grad_norm": 2.7434573327713205, + "learning_rate": 1.8855117042697567e-05, + "loss": 1.1818, + "step": 1863 + }, + { + "epoch": 0.17929973066564064, + "grad_norm": 2.337770853288953, + "learning_rate": 1.885366913423643e-05, + "loss": 1.1085, + "step": 1864 + }, + { + "epoch": 0.1793959215082724, + "grad_norm": 2.305458053911818, + "learning_rate": 1.885222036645047e-05, + "loss": 0.8884, + "step": 1865 + }, + { + "epoch": 0.1794921123509042, + "grad_norm": 1.961641507529729, + "learning_rate": 1.8850770739480312e-05, + "loss": 0.8128, + "step": 1866 + }, + { + "epoch": 0.17958830319353597, + "grad_norm": 2.1985650568807444, + "learning_rate": 1.884932025346664e-05, + "loss": 1.0316, + "step": 1867 + }, + { + "epoch": 0.17968449403616776, + "grad_norm": 2.2307113217227075, + "learning_rate": 1.8847868908550252e-05, + "loss": 1.0837, + "step": 1868 + }, + { + "epoch": 0.17978068487879953, + "grad_norm": 2.2346782065289634, + "learning_rate": 1.8846416704872e-05, + "loss": 1.0415, + "step": 1869 + }, + { + "epoch": 0.17987687572143132, + "grad_norm": 2.2630412908932933, + "learning_rate": 1.8844963642572837e-05, + "loss": 1.0919, + "step": 1870 + }, + { + "epoch": 0.1799730665640631, + "grad_norm": 2.324611011851236, + "learning_rate": 1.8843509721793798e-05, + "loss": 1.018, + "step": 1871 + }, + { + "epoch": 0.18006925740669488, + "grad_norm": 2.3278153886493436, + "learning_rate": 1.8842054942676e-05, + "loss": 1.1019, + "step": 1872 + }, + { + "epoch": 0.18016544824932668, + "grad_norm": 2.1934612123670987, + "learning_rate": 1.8840599305360634e-05, + "loss": 1.0975, + "step": 1873 + }, + { + "epoch": 0.18026163909195844, + "grad_norm": 2.186481062029247, + "learning_rate": 1.8839142809988987e-05, + "loss": 0.8019, + "step": 1874 + }, + { + "epoch": 0.18035782993459024, + "grad_norm": 2.5472850957575695, + "learning_rate": 1.8837685456702428e-05, + "loss": 1.042, + "step": 1875 + }, + { + "epoch": 0.180454020777222, + "grad_norm": 2.2606570552127, + "learning_rate": 1.88362272456424e-05, + "loss": 1.0279, + "step": 1876 + }, + { + "epoch": 0.1805502116198538, + "grad_norm": 2.319339538881481, + "learning_rate": 1.8834768176950438e-05, + "loss": 1.1814, + "step": 1877 + }, + { + "epoch": 0.18064640246248556, + "grad_norm": 2.3998469796670534, + "learning_rate": 1.8833308250768153e-05, + "loss": 1.0553, + "step": 1878 + }, + { + "epoch": 0.18074259330511735, + "grad_norm": 1.9267999723435885, + "learning_rate": 1.883184746723725e-05, + "loss": 0.8532, + "step": 1879 + }, + { + "epoch": 0.18083878414774912, + "grad_norm": 2.440686490209758, + "learning_rate": 1.8830385826499507e-05, + "loss": 1.2073, + "step": 1880 + }, + { + "epoch": 0.18093497499038091, + "grad_norm": 1.9561295062451742, + "learning_rate": 1.8828923328696788e-05, + "loss": 0.8653, + "step": 1881 + }, + { + "epoch": 0.1810311658330127, + "grad_norm": 2.5801295051030206, + "learning_rate": 1.882745997397104e-05, + "loss": 1.1057, + "step": 1882 + }, + { + "epoch": 0.18112735667564447, + "grad_norm": 2.244622245310317, + "learning_rate": 1.88259957624643e-05, + "loss": 1.1126, + "step": 1883 + }, + { + "epoch": 0.18122354751827627, + "grad_norm": 2.3216728022366198, + "learning_rate": 1.8824530694318675e-05, + "loss": 1.0764, + "step": 1884 + }, + { + "epoch": 0.18131973836090803, + "grad_norm": 2.032233989690227, + "learning_rate": 1.882306476967637e-05, + "loss": 1.0752, + "step": 1885 + }, + { + "epoch": 0.18141592920353983, + "grad_norm": 2.5331983337979818, + "learning_rate": 1.882159798867966e-05, + "loss": 1.054, + "step": 1886 + }, + { + "epoch": 0.1815121200461716, + "grad_norm": 2.106945682655599, + "learning_rate": 1.8820130351470906e-05, + "loss": 1.0509, + "step": 1887 + }, + { + "epoch": 0.1816083108888034, + "grad_norm": 2.39462182431405, + "learning_rate": 1.8818661858192562e-05, + "loss": 1.0269, + "step": 1888 + }, + { + "epoch": 0.18170450173143518, + "grad_norm": 2.2922686419439002, + "learning_rate": 1.8817192508987158e-05, + "loss": 1.0705, + "step": 1889 + }, + { + "epoch": 0.18180069257406695, + "grad_norm": 2.095266306868984, + "learning_rate": 1.88157223039973e-05, + "loss": 1.1752, + "step": 1890 + }, + { + "epoch": 0.18189688341669874, + "grad_norm": 2.4297658467231615, + "learning_rate": 1.8814251243365688e-05, + "loss": 1.0788, + "step": 1891 + }, + { + "epoch": 0.1819930742593305, + "grad_norm": 2.1097578166979374, + "learning_rate": 1.8812779327235106e-05, + "loss": 1.0008, + "step": 1892 + }, + { + "epoch": 0.1820892651019623, + "grad_norm": 2.2970036408752366, + "learning_rate": 1.8811306555748407e-05, + "loss": 1.1444, + "step": 1893 + }, + { + "epoch": 0.18218545594459407, + "grad_norm": 2.174260738855069, + "learning_rate": 1.880983292904854e-05, + "loss": 0.8613, + "step": 1894 + }, + { + "epoch": 0.18228164678722586, + "grad_norm": 2.4823830949854964, + "learning_rate": 1.8808358447278536e-05, + "loss": 1.1046, + "step": 1895 + }, + { + "epoch": 0.18237783762985763, + "grad_norm": 2.316117201468291, + "learning_rate": 1.88068831105815e-05, + "loss": 1.1101, + "step": 1896 + }, + { + "epoch": 0.18247402847248942, + "grad_norm": 2.3434026298739767, + "learning_rate": 1.8805406919100636e-05, + "loss": 1.0617, + "step": 1897 + }, + { + "epoch": 0.1825702193151212, + "grad_norm": 2.455373105810121, + "learning_rate": 1.8803929872979214e-05, + "loss": 1.1454, + "step": 1898 + }, + { + "epoch": 0.18266641015775298, + "grad_norm": 2.37422807529231, + "learning_rate": 1.880245197236059e-05, + "loss": 0.9969, + "step": 1899 + }, + { + "epoch": 0.18276260100038477, + "grad_norm": 2.428727221482773, + "learning_rate": 1.8800973217388215e-05, + "loss": 1.189, + "step": 1900 + }, + { + "epoch": 0.18285879184301654, + "grad_norm": 2.096704008829554, + "learning_rate": 1.8799493608205614e-05, + "loss": 1.0701, + "step": 1901 + }, + { + "epoch": 0.18295498268564833, + "grad_norm": 2.1302973349801477, + "learning_rate": 1.879801314495639e-05, + "loss": 1.1316, + "step": 1902 + }, + { + "epoch": 0.1830511735282801, + "grad_norm": 2.852044030252968, + "learning_rate": 1.8796531827784242e-05, + "loss": 1.1898, + "step": 1903 + }, + { + "epoch": 0.1831473643709119, + "grad_norm": 2.445764500697258, + "learning_rate": 1.879504965683294e-05, + "loss": 1.0916, + "step": 1904 + }, + { + "epoch": 0.18324355521354366, + "grad_norm": 2.3255564038762486, + "learning_rate": 1.879356663224634e-05, + "loss": 1.141, + "step": 1905 + }, + { + "epoch": 0.18333974605617545, + "grad_norm": 2.2248834061535487, + "learning_rate": 1.8792082754168385e-05, + "loss": 1.1046, + "step": 1906 + }, + { + "epoch": 0.18343593689880724, + "grad_norm": 2.343397719553691, + "learning_rate": 1.87905980227431e-05, + "loss": 1.1113, + "step": 1907 + }, + { + "epoch": 0.183532127741439, + "grad_norm": 2.272177272564819, + "learning_rate": 1.878911243811459e-05, + "loss": 0.9975, + "step": 1908 + }, + { + "epoch": 0.1836283185840708, + "grad_norm": 2.493324918527672, + "learning_rate": 1.878762600042704e-05, + "loss": 1.0818, + "step": 1909 + }, + { + "epoch": 0.18372450942670257, + "grad_norm": 2.7613636602105633, + "learning_rate": 1.8786138709824726e-05, + "loss": 1.0219, + "step": 1910 + }, + { + "epoch": 0.18382070026933436, + "grad_norm": 2.809388398774116, + "learning_rate": 1.8784650566452e-05, + "loss": 1.1266, + "step": 1911 + }, + { + "epoch": 0.18391689111196613, + "grad_norm": 2.464528113216418, + "learning_rate": 1.8783161570453295e-05, + "loss": 1.0901, + "step": 1912 + }, + { + "epoch": 0.18401308195459792, + "grad_norm": 2.05220368227022, + "learning_rate": 1.878167172197314e-05, + "loss": 1.0242, + "step": 1913 + }, + { + "epoch": 0.18410927279722972, + "grad_norm": 2.5340403311758455, + "learning_rate": 1.878018102115614e-05, + "loss": 0.9278, + "step": 1914 + }, + { + "epoch": 0.18420546363986148, + "grad_norm": 2.2886820851288046, + "learning_rate": 1.8778689468146965e-05, + "loss": 1.1382, + "step": 1915 + }, + { + "epoch": 0.18430165448249328, + "grad_norm": 2.1644642546681143, + "learning_rate": 1.8777197063090394e-05, + "loss": 0.9864, + "step": 1916 + }, + { + "epoch": 0.18439784532512504, + "grad_norm": 2.610276974836029, + "learning_rate": 1.8775703806131273e-05, + "loss": 1.0767, + "step": 1917 + }, + { + "epoch": 0.18449403616775684, + "grad_norm": 2.189663153799843, + "learning_rate": 1.877420969741454e-05, + "loss": 1.1313, + "step": 1918 + }, + { + "epoch": 0.1845902270103886, + "grad_norm": 2.2263106307304006, + "learning_rate": 1.877271473708521e-05, + "loss": 1.0282, + "step": 1919 + }, + { + "epoch": 0.1846864178530204, + "grad_norm": 2.326228268654555, + "learning_rate": 1.877121892528838e-05, + "loss": 1.1265, + "step": 1920 + }, + { + "epoch": 0.18478260869565216, + "grad_norm": 2.2666137102303545, + "learning_rate": 1.876972226216923e-05, + "loss": 0.8789, + "step": 1921 + }, + { + "epoch": 0.18487879953828396, + "grad_norm": 2.627696922145256, + "learning_rate": 1.876822474787303e-05, + "loss": 1.0489, + "step": 1922 + }, + { + "epoch": 0.18497499038091575, + "grad_norm": 2.4112735305790975, + "learning_rate": 1.8766726382545123e-05, + "loss": 1.0156, + "step": 1923 + }, + { + "epoch": 0.18507118122354752, + "grad_norm": 2.467463915786531, + "learning_rate": 1.8765227166330933e-05, + "loss": 1.1345, + "step": 1924 + }, + { + "epoch": 0.1851673720661793, + "grad_norm": 2.316027406682091, + "learning_rate": 1.876372709937598e-05, + "loss": 1.1609, + "step": 1925 + }, + { + "epoch": 0.18526356290881107, + "grad_norm": 2.4461850727605943, + "learning_rate": 1.8762226181825857e-05, + "loss": 1.0701, + "step": 1926 + }, + { + "epoch": 0.18535975375144287, + "grad_norm": 2.5289875839189557, + "learning_rate": 1.8760724413826238e-05, + "loss": 1.0834, + "step": 1927 + }, + { + "epoch": 0.18545594459407463, + "grad_norm": 2.285351130540244, + "learning_rate": 1.875922179552288e-05, + "loss": 1.172, + "step": 1928 + }, + { + "epoch": 0.18555213543670643, + "grad_norm": 2.1691413003619107, + "learning_rate": 1.8757718327061634e-05, + "loss": 1.0872, + "step": 1929 + }, + { + "epoch": 0.1856483262793382, + "grad_norm": 2.3519601240113226, + "learning_rate": 1.875621400858842e-05, + "loss": 0.9195, + "step": 1930 + }, + { + "epoch": 0.18574451712197, + "grad_norm": 2.144191922361123, + "learning_rate": 1.875470884024924e-05, + "loss": 1.0922, + "step": 1931 + }, + { + "epoch": 0.18584070796460178, + "grad_norm": 2.3819091085633515, + "learning_rate": 1.875320282219019e-05, + "loss": 1.066, + "step": 1932 + }, + { + "epoch": 0.18593689880723355, + "grad_norm": 2.314571771525246, + "learning_rate": 1.875169595455744e-05, + "loss": 1.0604, + "step": 1933 + }, + { + "epoch": 0.18603308964986534, + "grad_norm": 1.9592397629656584, + "learning_rate": 1.8750188237497247e-05, + "loss": 0.8302, + "step": 1934 + }, + { + "epoch": 0.1861292804924971, + "grad_norm": 2.1301926780163982, + "learning_rate": 1.8748679671155946e-05, + "loss": 1.0194, + "step": 1935 + }, + { + "epoch": 0.1862254713351289, + "grad_norm": 2.2392834500085192, + "learning_rate": 1.874717025567995e-05, + "loss": 1.0616, + "step": 1936 + }, + { + "epoch": 0.18632166217776067, + "grad_norm": 2.271678803478854, + "learning_rate": 1.8745659991215773e-05, + "loss": 1.0593, + "step": 1937 + }, + { + "epoch": 0.18641785302039246, + "grad_norm": 2.2945329084947623, + "learning_rate": 1.874414887790999e-05, + "loss": 1.1296, + "step": 1938 + }, + { + "epoch": 0.18651404386302425, + "grad_norm": 1.8201607501436001, + "learning_rate": 1.874263691590927e-05, + "loss": 0.8056, + "step": 1939 + }, + { + "epoch": 0.18661023470565602, + "grad_norm": 2.32970810228447, + "learning_rate": 1.8741124105360363e-05, + "loss": 1.0239, + "step": 1940 + }, + { + "epoch": 0.1867064255482878, + "grad_norm": 2.725270171690526, + "learning_rate": 1.8739610446410097e-05, + "loss": 1.0315, + "step": 1941 + }, + { + "epoch": 0.18680261639091958, + "grad_norm": 2.13428404945904, + "learning_rate": 1.873809593920539e-05, + "loss": 1.1136, + "step": 1942 + }, + { + "epoch": 0.18689880723355137, + "grad_norm": 2.1824283436538043, + "learning_rate": 1.8736580583893236e-05, + "loss": 1.0463, + "step": 1943 + }, + { + "epoch": 0.18699499807618314, + "grad_norm": 2.412683806460741, + "learning_rate": 1.8735064380620717e-05, + "loss": 1.131, + "step": 1944 + }, + { + "epoch": 0.18709118891881493, + "grad_norm": 4.671212905561489, + "learning_rate": 1.8733547329534985e-05, + "loss": 1.1462, + "step": 1945 + }, + { + "epoch": 0.1871873797614467, + "grad_norm": 2.1354896593316006, + "learning_rate": 1.873202943078329e-05, + "loss": 1.1189, + "step": 1946 + }, + { + "epoch": 0.1872835706040785, + "grad_norm": 2.17449155798944, + "learning_rate": 1.8730510684512956e-05, + "loss": 1.0684, + "step": 1947 + }, + { + "epoch": 0.18737976144671029, + "grad_norm": 2.2675805858044513, + "learning_rate": 1.8728991090871387e-05, + "loss": 1.1212, + "step": 1948 + }, + { + "epoch": 0.18747595228934205, + "grad_norm": 2.106710015634515, + "learning_rate": 1.8727470650006078e-05, + "loss": 0.8905, + "step": 1949 + }, + { + "epoch": 0.18757214313197385, + "grad_norm": 2.4255981217714635, + "learning_rate": 1.8725949362064596e-05, + "loss": 1.0139, + "step": 1950 + }, + { + "epoch": 0.1876683339746056, + "grad_norm": 2.086595554651129, + "learning_rate": 1.8724427227194596e-05, + "loss": 1.0289, + "step": 1951 + }, + { + "epoch": 0.1877645248172374, + "grad_norm": 2.123185021347397, + "learning_rate": 1.8722904245543817e-05, + "loss": 0.9998, + "step": 1952 + }, + { + "epoch": 0.18786071565986917, + "grad_norm": 2.6038510639820602, + "learning_rate": 1.8721380417260073e-05, + "loss": 1.1201, + "step": 1953 + }, + { + "epoch": 0.18795690650250096, + "grad_norm": 2.449020956664015, + "learning_rate": 1.871985574249127e-05, + "loss": 1.0229, + "step": 1954 + }, + { + "epoch": 0.18805309734513273, + "grad_norm": 2.1718434850173605, + "learning_rate": 1.8718330221385392e-05, + "loss": 1.0872, + "step": 1955 + }, + { + "epoch": 0.18814928818776452, + "grad_norm": 2.2323029625163318, + "learning_rate": 1.8716803854090495e-05, + "loss": 1.1121, + "step": 1956 + }, + { + "epoch": 0.18824547903039632, + "grad_norm": 2.3210719674514566, + "learning_rate": 1.8715276640754734e-05, + "loss": 1.0868, + "step": 1957 + }, + { + "epoch": 0.18834166987302808, + "grad_norm": 2.3468447119457214, + "learning_rate": 1.8713748581526334e-05, + "loss": 0.9782, + "step": 1958 + }, + { + "epoch": 0.18843786071565988, + "grad_norm": 2.4178527078649883, + "learning_rate": 1.8712219676553606e-05, + "loss": 1.0455, + "step": 1959 + }, + { + "epoch": 0.18853405155829164, + "grad_norm": 2.704582903383797, + "learning_rate": 1.871068992598495e-05, + "loss": 1.2548, + "step": 1960 + }, + { + "epoch": 0.18863024240092344, + "grad_norm": 2.2687788661757784, + "learning_rate": 1.8709159329968835e-05, + "loss": 0.9691, + "step": 1961 + }, + { + "epoch": 0.1887264332435552, + "grad_norm": 2.2903858877935206, + "learning_rate": 1.8707627888653816e-05, + "loss": 1.1277, + "step": 1962 + }, + { + "epoch": 0.188822624086187, + "grad_norm": 2.168890860596265, + "learning_rate": 1.870609560218854e-05, + "loss": 1.1145, + "step": 1963 + }, + { + "epoch": 0.1889188149288188, + "grad_norm": 2.1863591471536865, + "learning_rate": 1.8704562470721728e-05, + "loss": 0.9183, + "step": 1964 + }, + { + "epoch": 0.18901500577145056, + "grad_norm": 2.5324309774678246, + "learning_rate": 1.8703028494402178e-05, + "loss": 1.1024, + "step": 1965 + }, + { + "epoch": 0.18911119661408235, + "grad_norm": 2.149099617669097, + "learning_rate": 1.870149367337878e-05, + "loss": 1.0764, + "step": 1966 + }, + { + "epoch": 0.18920738745671412, + "grad_norm": 2.2782812294657635, + "learning_rate": 1.8699958007800498e-05, + "loss": 0.9805, + "step": 1967 + }, + { + "epoch": 0.1893035782993459, + "grad_norm": 2.0929975660799194, + "learning_rate": 1.8698421497816386e-05, + "loss": 1.1085, + "step": 1968 + }, + { + "epoch": 0.18939976914197768, + "grad_norm": 2.0608086865828534, + "learning_rate": 1.8696884143575574e-05, + "loss": 1.1887, + "step": 1969 + }, + { + "epoch": 0.18949595998460947, + "grad_norm": 2.180269505852477, + "learning_rate": 1.869534594522727e-05, + "loss": 1.1055, + "step": 1970 + }, + { + "epoch": 0.18959215082724123, + "grad_norm": 2.2696937969253765, + "learning_rate": 1.8693806902920778e-05, + "loss": 0.9359, + "step": 1971 + }, + { + "epoch": 0.18968834166987303, + "grad_norm": 2.2330893569545163, + "learning_rate": 1.8692267016805473e-05, + "loss": 1.1441, + "step": 1972 + }, + { + "epoch": 0.18978453251250482, + "grad_norm": 2.2173205677058925, + "learning_rate": 1.869072628703081e-05, + "loss": 1.0841, + "step": 1973 + }, + { + "epoch": 0.1898807233551366, + "grad_norm": 2.473481111151139, + "learning_rate": 1.8689184713746333e-05, + "loss": 1.1415, + "step": 1974 + }, + { + "epoch": 0.18997691419776838, + "grad_norm": 2.052371222443128, + "learning_rate": 1.8687642297101667e-05, + "loss": 0.9768, + "step": 1975 + }, + { + "epoch": 0.19007310504040015, + "grad_norm": 2.2808600793893583, + "learning_rate": 1.868609903724651e-05, + "loss": 1.1219, + "step": 1976 + }, + { + "epoch": 0.19016929588303194, + "grad_norm": 2.2165351199178067, + "learning_rate": 1.8684554934330658e-05, + "loss": 1.0349, + "step": 1977 + }, + { + "epoch": 0.1902654867256637, + "grad_norm": 2.7046300642830756, + "learning_rate": 1.8683009988503972e-05, + "loss": 1.0783, + "step": 1978 + }, + { + "epoch": 0.1903616775682955, + "grad_norm": 2.3621931708500035, + "learning_rate": 1.8681464199916407e-05, + "loss": 1.0595, + "step": 1979 + }, + { + "epoch": 0.19045786841092727, + "grad_norm": 2.337838847221424, + "learning_rate": 1.867991756871799e-05, + "loss": 1.1115, + "step": 1980 + }, + { + "epoch": 0.19055405925355906, + "grad_norm": 2.8385130574203736, + "learning_rate": 1.867837009505884e-05, + "loss": 1.1932, + "step": 1981 + }, + { + "epoch": 0.19065025009619085, + "grad_norm": 2.4387928539411408, + "learning_rate": 1.867682177908915e-05, + "loss": 1.0783, + "step": 1982 + }, + { + "epoch": 0.19074644093882262, + "grad_norm": 2.244522498184186, + "learning_rate": 1.8675272620959197e-05, + "loss": 1.014, + "step": 1983 + }, + { + "epoch": 0.1908426317814544, + "grad_norm": 2.394782243484656, + "learning_rate": 1.867372262081934e-05, + "loss": 0.887, + "step": 1984 + }, + { + "epoch": 0.19093882262408618, + "grad_norm": 2.2473445096902376, + "learning_rate": 1.8672171778820023e-05, + "loss": 1.1052, + "step": 1985 + }, + { + "epoch": 0.19103501346671797, + "grad_norm": 2.228422419015579, + "learning_rate": 1.8670620095111766e-05, + "loss": 1.0442, + "step": 1986 + }, + { + "epoch": 0.19113120430934974, + "grad_norm": 2.4448745759641346, + "learning_rate": 1.8669067569845176e-05, + "loss": 1.1295, + "step": 1987 + }, + { + "epoch": 0.19122739515198153, + "grad_norm": 2.3466830778634558, + "learning_rate": 1.8667514203170934e-05, + "loss": 1.1668, + "step": 1988 + }, + { + "epoch": 0.19132358599461333, + "grad_norm": 2.2949916957574144, + "learning_rate": 1.866595999523981e-05, + "loss": 0.9988, + "step": 1989 + }, + { + "epoch": 0.1914197768372451, + "grad_norm": 2.0445778513756485, + "learning_rate": 1.8664404946202658e-05, + "loss": 1.0338, + "step": 1990 + }, + { + "epoch": 0.1915159676798769, + "grad_norm": 2.1117075636517924, + "learning_rate": 1.86628490562104e-05, + "loss": 1.022, + "step": 1991 + }, + { + "epoch": 0.19161215852250865, + "grad_norm": 2.038531900313352, + "learning_rate": 1.8661292325414058e-05, + "loss": 0.8779, + "step": 1992 + }, + { + "epoch": 0.19170834936514045, + "grad_norm": 2.219814092191726, + "learning_rate": 1.8659734753964714e-05, + "loss": 1.096, + "step": 1993 + }, + { + "epoch": 0.1918045402077722, + "grad_norm": 2.4368827264827715, + "learning_rate": 1.865817634201356e-05, + "loss": 1.1213, + "step": 1994 + }, + { + "epoch": 0.191900731050404, + "grad_norm": 2.1670519146903704, + "learning_rate": 1.8656617089711834e-05, + "loss": 1.1327, + "step": 1995 + }, + { + "epoch": 0.19199692189303577, + "grad_norm": 2.2598264634196825, + "learning_rate": 1.8655056997210893e-05, + "loss": 1.0787, + "step": 1996 + }, + { + "epoch": 0.19209311273566756, + "grad_norm": 2.713378183636638, + "learning_rate": 1.8653496064662145e-05, + "loss": 1.1223, + "step": 1997 + }, + { + "epoch": 0.19218930357829936, + "grad_norm": 2.4731000040470468, + "learning_rate": 1.8651934292217097e-05, + "loss": 1.05, + "step": 1998 + }, + { + "epoch": 0.19228549442093112, + "grad_norm": 2.4636673105847073, + "learning_rate": 1.865037168002733e-05, + "loss": 1.1057, + "step": 1999 + }, + { + "epoch": 0.19238168526356292, + "grad_norm": 2.105485988906898, + "learning_rate": 1.864880822824452e-05, + "loss": 1.0282, + "step": 2000 + }, + { + "epoch": 0.19247787610619468, + "grad_norm": 2.4337846704711956, + "learning_rate": 1.8647243937020394e-05, + "loss": 1.0613, + "step": 2001 + }, + { + "epoch": 0.19257406694882648, + "grad_norm": 2.0272422472980565, + "learning_rate": 1.8645678806506795e-05, + "loss": 0.9852, + "step": 2002 + }, + { + "epoch": 0.19267025779145824, + "grad_norm": 2.2882734628893995, + "learning_rate": 1.864411283685563e-05, + "loss": 1.0249, + "step": 2003 + }, + { + "epoch": 0.19276644863409004, + "grad_norm": 2.26903397754255, + "learning_rate": 1.864254602821888e-05, + "loss": 1.0118, + "step": 2004 + }, + { + "epoch": 0.1928626394767218, + "grad_norm": 2.364005990194935, + "learning_rate": 1.8640978380748628e-05, + "loss": 1.054, + "step": 2005 + }, + { + "epoch": 0.1929588303193536, + "grad_norm": 2.152845041600155, + "learning_rate": 1.8639409894597026e-05, + "loss": 1.0284, + "step": 2006 + }, + { + "epoch": 0.1930550211619854, + "grad_norm": 2.3814917828565605, + "learning_rate": 1.8637840569916306e-05, + "loss": 1.1274, + "step": 2007 + }, + { + "epoch": 0.19315121200461716, + "grad_norm": 2.200801942938711, + "learning_rate": 1.8636270406858786e-05, + "loss": 0.9684, + "step": 2008 + }, + { + "epoch": 0.19324740284724895, + "grad_norm": 2.374913776557737, + "learning_rate": 1.8634699405576865e-05, + "loss": 1.0573, + "step": 2009 + }, + { + "epoch": 0.19334359368988072, + "grad_norm": 2.3280815449195833, + "learning_rate": 1.8633127566223023e-05, + "loss": 1.1366, + "step": 2010 + }, + { + "epoch": 0.1934397845325125, + "grad_norm": 2.3092998627561028, + "learning_rate": 1.8631554888949813e-05, + "loss": 1.1453, + "step": 2011 + }, + { + "epoch": 0.19353597537514428, + "grad_norm": 2.228286263154161, + "learning_rate": 1.862998137390989e-05, + "loss": 1.0959, + "step": 2012 + }, + { + "epoch": 0.19363216621777607, + "grad_norm": 2.217264008347923, + "learning_rate": 1.862840702125596e-05, + "loss": 1.1161, + "step": 2013 + }, + { + "epoch": 0.19372835706040784, + "grad_norm": 2.4007900582255566, + "learning_rate": 1.8626831831140845e-05, + "loss": 1.1424, + "step": 2014 + }, + { + "epoch": 0.19382454790303963, + "grad_norm": 2.3387610479481546, + "learning_rate": 1.8625255803717417e-05, + "loss": 1.1042, + "step": 2015 + }, + { + "epoch": 0.19392073874567142, + "grad_norm": 2.127546429578171, + "learning_rate": 1.8623678939138652e-05, + "loss": 0.9698, + "step": 2016 + }, + { + "epoch": 0.1940169295883032, + "grad_norm": 2.3290522255395647, + "learning_rate": 1.86221012375576e-05, + "loss": 1.0764, + "step": 2017 + }, + { + "epoch": 0.19411312043093498, + "grad_norm": 2.5668218623970063, + "learning_rate": 1.8620522699127374e-05, + "loss": 0.944, + "step": 2018 + }, + { + "epoch": 0.19420931127356675, + "grad_norm": 2.6125765847286275, + "learning_rate": 1.8618943324001205e-05, + "loss": 0.9296, + "step": 2019 + }, + { + "epoch": 0.19430550211619854, + "grad_norm": 2.4900371574936786, + "learning_rate": 1.8617363112332376e-05, + "loss": 0.9771, + "step": 2020 + }, + { + "epoch": 0.1944016929588303, + "grad_norm": 2.3812550984732077, + "learning_rate": 1.8615782064274256e-05, + "loss": 0.9174, + "step": 2021 + }, + { + "epoch": 0.1944978838014621, + "grad_norm": 2.3789354435868777, + "learning_rate": 1.8614200179980307e-05, + "loss": 0.9945, + "step": 2022 + }, + { + "epoch": 0.1945940746440939, + "grad_norm": 2.548640950084275, + "learning_rate": 1.8612617459604063e-05, + "loss": 1.1173, + "step": 2023 + }, + { + "epoch": 0.19469026548672566, + "grad_norm": 2.244707339983637, + "learning_rate": 1.8611033903299136e-05, + "loss": 1.0834, + "step": 2024 + }, + { + "epoch": 0.19478645632935745, + "grad_norm": 2.2433987945726943, + "learning_rate": 1.8609449511219226e-05, + "loss": 1.0477, + "step": 2025 + }, + { + "epoch": 0.19488264717198922, + "grad_norm": 2.634286764979884, + "learning_rate": 1.8607864283518116e-05, + "loss": 1.1116, + "step": 2026 + }, + { + "epoch": 0.19497883801462101, + "grad_norm": 2.263555918463014, + "learning_rate": 1.860627822034966e-05, + "loss": 0.9509, + "step": 2027 + }, + { + "epoch": 0.19507502885725278, + "grad_norm": 2.3183146677669653, + "learning_rate": 1.8604691321867804e-05, + "loss": 1.0774, + "step": 2028 + }, + { + "epoch": 0.19517121969988457, + "grad_norm": 2.210053999613424, + "learning_rate": 1.8603103588226567e-05, + "loss": 0.8849, + "step": 2029 + }, + { + "epoch": 0.19526741054251634, + "grad_norm": 2.4210720181237906, + "learning_rate": 1.8601515019580053e-05, + "loss": 1.1131, + "step": 2030 + }, + { + "epoch": 0.19536360138514813, + "grad_norm": 2.441305663126275, + "learning_rate": 1.8599925616082444e-05, + "loss": 1.0201, + "step": 2031 + }, + { + "epoch": 0.19545979222777993, + "grad_norm": 2.086411627901295, + "learning_rate": 1.8598335377888012e-05, + "loss": 1.0732, + "step": 2032 + }, + { + "epoch": 0.1955559830704117, + "grad_norm": 2.2542234272164516, + "learning_rate": 1.8596744305151095e-05, + "loss": 1.1575, + "step": 2033 + }, + { + "epoch": 0.1956521739130435, + "grad_norm": 2.1421806537636554, + "learning_rate": 1.8595152398026128e-05, + "loss": 1.0178, + "step": 2034 + }, + { + "epoch": 0.19574836475567525, + "grad_norm": 2.6924731426509094, + "learning_rate": 1.8593559656667614e-05, + "loss": 1.1154, + "step": 2035 + }, + { + "epoch": 0.19584455559830705, + "grad_norm": 2.2672893811354715, + "learning_rate": 1.8591966081230142e-05, + "loss": 1.0354, + "step": 2036 + }, + { + "epoch": 0.1959407464409388, + "grad_norm": 2.2476804106146653, + "learning_rate": 1.8590371671868385e-05, + "loss": 1.078, + "step": 2037 + }, + { + "epoch": 0.1960369372835706, + "grad_norm": 2.7410916711221667, + "learning_rate": 1.8588776428737095e-05, + "loss": 1.0817, + "step": 2038 + }, + { + "epoch": 0.19613312812620237, + "grad_norm": 2.1824477114345804, + "learning_rate": 1.85871803519911e-05, + "loss": 1.1, + "step": 2039 + }, + { + "epoch": 0.19622931896883417, + "grad_norm": 2.3713834605413493, + "learning_rate": 1.858558344178532e-05, + "loss": 0.9086, + "step": 2040 + }, + { + "epoch": 0.19632550981146596, + "grad_norm": 2.7965139781104846, + "learning_rate": 1.858398569827474e-05, + "loss": 1.0301, + "step": 2041 + }, + { + "epoch": 0.19642170065409773, + "grad_norm": 2.394993455298904, + "learning_rate": 1.8582387121614437e-05, + "loss": 1.1287, + "step": 2042 + }, + { + "epoch": 0.19651789149672952, + "grad_norm": 2.1349560561251706, + "learning_rate": 1.8580787711959572e-05, + "loss": 0.9815, + "step": 2043 + }, + { + "epoch": 0.19661408233936128, + "grad_norm": 2.555971376756515, + "learning_rate": 1.857918746946538e-05, + "loss": 1.0324, + "step": 2044 + }, + { + "epoch": 0.19671027318199308, + "grad_norm": 2.264700211241404, + "learning_rate": 1.8577586394287174e-05, + "loss": 1.0707, + "step": 2045 + }, + { + "epoch": 0.19680646402462484, + "grad_norm": 2.0741654922268897, + "learning_rate": 1.8575984486580353e-05, + "loss": 0.8722, + "step": 2046 + }, + { + "epoch": 0.19690265486725664, + "grad_norm": 2.6505918140149483, + "learning_rate": 1.85743817465004e-05, + "loss": 1.1411, + "step": 2047 + }, + { + "epoch": 0.19699884570988843, + "grad_norm": 2.4933856019917684, + "learning_rate": 1.857277817420287e-05, + "loss": 1.0727, + "step": 2048 + }, + { + "epoch": 0.1970950365525202, + "grad_norm": 2.2939122574978046, + "learning_rate": 1.8571173769843406e-05, + "loss": 1.11, + "step": 2049 + }, + { + "epoch": 0.197191227395152, + "grad_norm": 2.6277530223149834, + "learning_rate": 1.8569568533577727e-05, + "loss": 1.1893, + "step": 2050 + }, + { + "epoch": 0.19728741823778376, + "grad_norm": 2.205226819984941, + "learning_rate": 1.8567962465561643e-05, + "loss": 1.0772, + "step": 2051 + }, + { + "epoch": 0.19738360908041555, + "grad_norm": 2.1453888022362095, + "learning_rate": 1.8566355565951023e-05, + "loss": 0.9287, + "step": 2052 + }, + { + "epoch": 0.19747979992304732, + "grad_norm": 2.296110047237231, + "learning_rate": 1.8564747834901844e-05, + "loss": 1.1141, + "step": 2053 + }, + { + "epoch": 0.1975759907656791, + "grad_norm": 2.614279905527157, + "learning_rate": 1.8563139272570142e-05, + "loss": 1.1515, + "step": 2054 + }, + { + "epoch": 0.19767218160831088, + "grad_norm": 1.9182408778232227, + "learning_rate": 1.8561529879112044e-05, + "loss": 1.0202, + "step": 2055 + }, + { + "epoch": 0.19776837245094267, + "grad_norm": 2.1116130041315286, + "learning_rate": 1.8559919654683756e-05, + "loss": 1.0139, + "step": 2056 + }, + { + "epoch": 0.19786456329357446, + "grad_norm": 2.4898774077446895, + "learning_rate": 1.855830859944156e-05, + "loss": 1.0186, + "step": 2057 + }, + { + "epoch": 0.19796075413620623, + "grad_norm": 2.050150128546578, + "learning_rate": 1.8556696713541833e-05, + "loss": 1.038, + "step": 2058 + }, + { + "epoch": 0.19805694497883802, + "grad_norm": 2.5810529913808478, + "learning_rate": 1.8555083997141014e-05, + "loss": 1.1446, + "step": 2059 + }, + { + "epoch": 0.1981531358214698, + "grad_norm": 2.1277582478580883, + "learning_rate": 1.855347045039563e-05, + "loss": 1.1495, + "step": 2060 + }, + { + "epoch": 0.19824932666410158, + "grad_norm": 2.0624250917329046, + "learning_rate": 1.8551856073462294e-05, + "loss": 1.1081, + "step": 2061 + }, + { + "epoch": 0.19834551750673335, + "grad_norm": 2.4050551802801907, + "learning_rate": 1.8550240866497697e-05, + "loss": 1.1699, + "step": 2062 + }, + { + "epoch": 0.19844170834936514, + "grad_norm": 2.255404235537074, + "learning_rate": 1.85486248296586e-05, + "loss": 1.0872, + "step": 2063 + }, + { + "epoch": 0.1985378991919969, + "grad_norm": 2.2624509434265097, + "learning_rate": 1.854700796310186e-05, + "loss": 1.0628, + "step": 2064 + }, + { + "epoch": 0.1986340900346287, + "grad_norm": 2.7393174999529335, + "learning_rate": 1.854539026698441e-05, + "loss": 1.1024, + "step": 2065 + }, + { + "epoch": 0.1987302808772605, + "grad_norm": 2.236669588154934, + "learning_rate": 1.8543771741463254e-05, + "loss": 0.941, + "step": 2066 + }, + { + "epoch": 0.19882647171989226, + "grad_norm": 2.2306260407026826, + "learning_rate": 1.8542152386695486e-05, + "loss": 1.0046, + "step": 2067 + }, + { + "epoch": 0.19892266256252406, + "grad_norm": 2.1357170352251287, + "learning_rate": 1.8540532202838286e-05, + "loss": 1.0433, + "step": 2068 + }, + { + "epoch": 0.19901885340515582, + "grad_norm": 2.3111317350752483, + "learning_rate": 1.8538911190048897e-05, + "loss": 1.0952, + "step": 2069 + }, + { + "epoch": 0.19911504424778761, + "grad_norm": 2.2250690935537554, + "learning_rate": 1.8537289348484658e-05, + "loss": 1.0958, + "step": 2070 + }, + { + "epoch": 0.19921123509041938, + "grad_norm": 2.217844537397232, + "learning_rate": 1.8535666678302978e-05, + "loss": 1.1333, + "step": 2071 + }, + { + "epoch": 0.19930742593305117, + "grad_norm": 2.220740737168344, + "learning_rate": 1.8534043179661357e-05, + "loss": 0.9419, + "step": 2072 + }, + { + "epoch": 0.19940361677568297, + "grad_norm": 2.39324315262601, + "learning_rate": 1.8532418852717363e-05, + "loss": 1.1748, + "step": 2073 + }, + { + "epoch": 0.19949980761831473, + "grad_norm": 2.5246112535043443, + "learning_rate": 1.8530793697628658e-05, + "loss": 1.1624, + "step": 2074 + }, + { + "epoch": 0.19959599846094653, + "grad_norm": 2.062860020747521, + "learning_rate": 1.852916771455297e-05, + "loss": 1.0932, + "step": 2075 + }, + { + "epoch": 0.1996921893035783, + "grad_norm": 2.995483024966093, + "learning_rate": 1.8527540903648122e-05, + "loss": 1.1773, + "step": 2076 + }, + { + "epoch": 0.1997883801462101, + "grad_norm": 2.2705433028191395, + "learning_rate": 1.8525913265072006e-05, + "loss": 1.0232, + "step": 2077 + }, + { + "epoch": 0.19988457098884185, + "grad_norm": 2.3562074567121285, + "learning_rate": 1.8524284798982595e-05, + "loss": 0.9208, + "step": 2078 + }, + { + "epoch": 0.19998076183147365, + "grad_norm": 2.0345034123016057, + "learning_rate": 1.852265550553795e-05, + "loss": 1.0588, + "step": 2079 + }, + { + "epoch": 0.2000769526741054, + "grad_norm": 2.138400195549276, + "learning_rate": 1.852102538489621e-05, + "loss": 0.9661, + "step": 2080 + }, + { + "epoch": 0.2001731435167372, + "grad_norm": 2.3111814239497264, + "learning_rate": 1.8519394437215586e-05, + "loss": 1.1073, + "step": 2081 + }, + { + "epoch": 0.200269334359369, + "grad_norm": 2.4662386984811606, + "learning_rate": 1.8517762662654383e-05, + "loss": 1.0847, + "step": 2082 + }, + { + "epoch": 0.20036552520200077, + "grad_norm": 2.1683374043158636, + "learning_rate": 1.8516130061370972e-05, + "loss": 0.8468, + "step": 2083 + }, + { + "epoch": 0.20046171604463256, + "grad_norm": 2.2668193601816995, + "learning_rate": 1.851449663352381e-05, + "loss": 1.1213, + "step": 2084 + }, + { + "epoch": 0.20055790688726433, + "grad_norm": 2.529150465125981, + "learning_rate": 1.8512862379271446e-05, + "loss": 1.1365, + "step": 2085 + }, + { + "epoch": 0.20065409772989612, + "grad_norm": 2.186002409463522, + "learning_rate": 1.851122729877249e-05, + "loss": 1.1331, + "step": 2086 + }, + { + "epoch": 0.20075028857252789, + "grad_norm": 2.0160685587866736, + "learning_rate": 1.8509591392185638e-05, + "loss": 1.0806, + "step": 2087 + }, + { + "epoch": 0.20084647941515968, + "grad_norm": 2.0911304165489195, + "learning_rate": 1.8507954659669677e-05, + "loss": 0.9901, + "step": 2088 + }, + { + "epoch": 0.20094267025779144, + "grad_norm": 1.8371114170758105, + "learning_rate": 1.8506317101383463e-05, + "loss": 0.8155, + "step": 2089 + }, + { + "epoch": 0.20103886110042324, + "grad_norm": 2.4269531518734335, + "learning_rate": 1.850467871748593e-05, + "loss": 1.0441, + "step": 2090 + }, + { + "epoch": 0.20113505194305503, + "grad_norm": 2.3412887036377974, + "learning_rate": 1.8503039508136104e-05, + "loss": 1.0455, + "step": 2091 + }, + { + "epoch": 0.2012312427856868, + "grad_norm": 2.0385512711511136, + "learning_rate": 1.850139947349308e-05, + "loss": 1.0773, + "step": 2092 + }, + { + "epoch": 0.2013274336283186, + "grad_norm": 1.8990210723499419, + "learning_rate": 1.849975861371604e-05, + "loss": 0.89, + "step": 2093 + }, + { + "epoch": 0.20142362447095036, + "grad_norm": 2.2147274556023353, + "learning_rate": 1.8498116928964244e-05, + "loss": 1.0441, + "step": 2094 + }, + { + "epoch": 0.20151981531358215, + "grad_norm": 2.406983293010951, + "learning_rate": 1.8496474419397024e-05, + "loss": 1.1224, + "step": 2095 + }, + { + "epoch": 0.20161600615621392, + "grad_norm": 2.4172658454275284, + "learning_rate": 1.849483108517381e-05, + "loss": 1.1588, + "step": 2096 + }, + { + "epoch": 0.2017121969988457, + "grad_norm": 1.8516125805055805, + "learning_rate": 1.8493186926454103e-05, + "loss": 1.032, + "step": 2097 + }, + { + "epoch": 0.2018083878414775, + "grad_norm": 2.198984742913749, + "learning_rate": 1.849154194339747e-05, + "loss": 0.9757, + "step": 2098 + }, + { + "epoch": 0.20190457868410927, + "grad_norm": 2.2754524680751387, + "learning_rate": 1.8489896136163582e-05, + "loss": 1.0994, + "step": 2099 + }, + { + "epoch": 0.20200076952674106, + "grad_norm": 2.5796269487901164, + "learning_rate": 1.8488249504912173e-05, + "loss": 1.0218, + "step": 2100 + }, + { + "epoch": 0.20209696036937283, + "grad_norm": 2.370206482818414, + "learning_rate": 1.848660204980307e-05, + "loss": 1.0253, + "step": 2101 + }, + { + "epoch": 0.20219315121200462, + "grad_norm": 2.220312215545756, + "learning_rate": 1.8484953770996163e-05, + "loss": 1.0584, + "step": 2102 + }, + { + "epoch": 0.2022893420546364, + "grad_norm": 2.288817313208056, + "learning_rate": 1.8483304668651435e-05, + "loss": 1.1029, + "step": 2103 + }, + { + "epoch": 0.20238553289726818, + "grad_norm": 2.515551927235117, + "learning_rate": 1.848165474292895e-05, + "loss": 1.0167, + "step": 2104 + }, + { + "epoch": 0.20248172373989995, + "grad_norm": 2.4913901746133655, + "learning_rate": 1.848000399398884e-05, + "loss": 1.0615, + "step": 2105 + }, + { + "epoch": 0.20257791458253174, + "grad_norm": 2.1301410334596693, + "learning_rate": 1.8478352421991334e-05, + "loss": 1.0707, + "step": 2106 + }, + { + "epoch": 0.20267410542516354, + "grad_norm": 1.9727642611357943, + "learning_rate": 1.8476700027096725e-05, + "loss": 1.0005, + "step": 2107 + }, + { + "epoch": 0.2027702962677953, + "grad_norm": 2.448790235344974, + "learning_rate": 1.847504680946539e-05, + "loss": 0.9915, + "step": 2108 + }, + { + "epoch": 0.2028664871104271, + "grad_norm": 2.2162438518917216, + "learning_rate": 1.8473392769257792e-05, + "loss": 1.0221, + "step": 2109 + }, + { + "epoch": 0.20296267795305886, + "grad_norm": 2.275630265234729, + "learning_rate": 1.847173790663447e-05, + "loss": 1.1004, + "step": 2110 + }, + { + "epoch": 0.20305886879569066, + "grad_norm": 2.2138822254647303, + "learning_rate": 1.847008222175604e-05, + "loss": 1.1042, + "step": 2111 + }, + { + "epoch": 0.20315505963832242, + "grad_norm": 2.9932193891601204, + "learning_rate": 1.8468425714783206e-05, + "loss": 1.0278, + "step": 2112 + }, + { + "epoch": 0.20325125048095422, + "grad_norm": 2.5567906763200834, + "learning_rate": 1.846676838587674e-05, + "loss": 1.1023, + "step": 2113 + }, + { + "epoch": 0.20334744132358598, + "grad_norm": 1.9305930505790245, + "learning_rate": 1.84651102351975e-05, + "loss": 0.995, + "step": 2114 + }, + { + "epoch": 0.20344363216621777, + "grad_norm": 2.333533663107226, + "learning_rate": 1.846345126290643e-05, + "loss": 1.0274, + "step": 2115 + }, + { + "epoch": 0.20353982300884957, + "grad_norm": 2.589489577307867, + "learning_rate": 1.846179146916454e-05, + "loss": 1.1391, + "step": 2116 + }, + { + "epoch": 0.20363601385148133, + "grad_norm": 2.1909687853384265, + "learning_rate": 1.846013085413293e-05, + "loss": 0.8436, + "step": 2117 + }, + { + "epoch": 0.20373220469411313, + "grad_norm": 2.089266541440252, + "learning_rate": 1.8458469417972783e-05, + "loss": 1.0884, + "step": 2118 + }, + { + "epoch": 0.2038283955367449, + "grad_norm": 2.320160650337959, + "learning_rate": 1.8456807160845348e-05, + "loss": 0.972, + "step": 2119 + }, + { + "epoch": 0.2039245863793767, + "grad_norm": 1.7325357543962723, + "learning_rate": 1.8455144082911965e-05, + "loss": 0.8607, + "step": 2120 + }, + { + "epoch": 0.20402077722200845, + "grad_norm": 2.218623880429561, + "learning_rate": 1.845348018433405e-05, + "loss": 1.131, + "step": 2121 + }, + { + "epoch": 0.20411696806464025, + "grad_norm": 2.191786739664546, + "learning_rate": 1.8451815465273097e-05, + "loss": 1.0032, + "step": 2122 + }, + { + "epoch": 0.20421315890727204, + "grad_norm": 2.0085997850908544, + "learning_rate": 1.845014992589068e-05, + "loss": 1.0232, + "step": 2123 + }, + { + "epoch": 0.2043093497499038, + "grad_norm": 2.442638244342296, + "learning_rate": 1.8448483566348456e-05, + "loss": 1.0096, + "step": 2124 + }, + { + "epoch": 0.2044055405925356, + "grad_norm": 2.0649491581379427, + "learning_rate": 1.844681638680816e-05, + "loss": 1.0254, + "step": 2125 + }, + { + "epoch": 0.20450173143516737, + "grad_norm": 2.2421280329950917, + "learning_rate": 1.8445148387431605e-05, + "loss": 1.1025, + "step": 2126 + }, + { + "epoch": 0.20459792227779916, + "grad_norm": 2.3597388307010982, + "learning_rate": 1.844347956838069e-05, + "loss": 0.977, + "step": 2127 + }, + { + "epoch": 0.20469411312043093, + "grad_norm": 2.363725964007312, + "learning_rate": 1.8441809929817382e-05, + "loss": 1.0514, + "step": 2128 + }, + { + "epoch": 0.20479030396306272, + "grad_norm": 2.274251846720306, + "learning_rate": 1.8440139471903732e-05, + "loss": 1.1507, + "step": 2129 + }, + { + "epoch": 0.20488649480569449, + "grad_norm": 2.597895384121739, + "learning_rate": 1.8438468194801876e-05, + "loss": 1.0637, + "step": 2130 + }, + { + "epoch": 0.20498268564832628, + "grad_norm": 2.499088532003136, + "learning_rate": 1.8436796098674028e-05, + "loss": 0.9611, + "step": 2131 + }, + { + "epoch": 0.20507887649095807, + "grad_norm": 2.437928001050751, + "learning_rate": 1.8435123183682475e-05, + "loss": 1.1101, + "step": 2132 + }, + { + "epoch": 0.20517506733358984, + "grad_norm": 2.156937772261369, + "learning_rate": 1.8433449449989593e-05, + "loss": 1.1166, + "step": 2133 + }, + { + "epoch": 0.20527125817622163, + "grad_norm": 2.3720698897565478, + "learning_rate": 1.8431774897757824e-05, + "loss": 1.1371, + "step": 2134 + }, + { + "epoch": 0.2053674490188534, + "grad_norm": 2.250547988526807, + "learning_rate": 1.8430099527149707e-05, + "loss": 1.023, + "step": 2135 + }, + { + "epoch": 0.2054636398614852, + "grad_norm": 2.4416201098454726, + "learning_rate": 1.8428423338327847e-05, + "loss": 1.1882, + "step": 2136 + }, + { + "epoch": 0.20555983070411696, + "grad_norm": 2.3741838851453325, + "learning_rate": 1.842674633145493e-05, + "loss": 0.8692, + "step": 2137 + }, + { + "epoch": 0.20565602154674875, + "grad_norm": 2.3739887153197374, + "learning_rate": 1.8425068506693727e-05, + "loss": 1.106, + "step": 2138 + }, + { + "epoch": 0.20575221238938052, + "grad_norm": 2.260174918236695, + "learning_rate": 1.8423389864207083e-05, + "loss": 1.1415, + "step": 2139 + }, + { + "epoch": 0.2058484032320123, + "grad_norm": 2.2831917278502765, + "learning_rate": 1.842171040415793e-05, + "loss": 1.1503, + "step": 2140 + }, + { + "epoch": 0.2059445940746441, + "grad_norm": 2.388765863775207, + "learning_rate": 1.8420030126709266e-05, + "loss": 1.1014, + "step": 2141 + }, + { + "epoch": 0.20604078491727587, + "grad_norm": 2.183297488047451, + "learning_rate": 1.8418349032024185e-05, + "loss": 1.0468, + "step": 2142 + }, + { + "epoch": 0.20613697575990766, + "grad_norm": 2.6298994119799075, + "learning_rate": 1.8416667120265847e-05, + "loss": 1.1673, + "step": 2143 + }, + { + "epoch": 0.20623316660253943, + "grad_norm": 2.3027770284876605, + "learning_rate": 1.8414984391597492e-05, + "loss": 1.066, + "step": 2144 + }, + { + "epoch": 0.20632935744517122, + "grad_norm": 2.3355455133693592, + "learning_rate": 1.841330084618245e-05, + "loss": 1.0769, + "step": 2145 + }, + { + "epoch": 0.206425548287803, + "grad_norm": 2.2579961971597773, + "learning_rate": 1.8411616484184126e-05, + "loss": 1.0529, + "step": 2146 + }, + { + "epoch": 0.20652173913043478, + "grad_norm": 2.1452464725794513, + "learning_rate": 1.8409931305765995e-05, + "loss": 1.0865, + "step": 2147 + }, + { + "epoch": 0.20661792997306658, + "grad_norm": 2.328549957338583, + "learning_rate": 1.8408245311091618e-05, + "loss": 1.1114, + "step": 2148 + }, + { + "epoch": 0.20671412081569834, + "grad_norm": 2.2104121039215765, + "learning_rate": 1.840655850032464e-05, + "loss": 1.1478, + "step": 2149 + }, + { + "epoch": 0.20681031165833014, + "grad_norm": 2.058471114883506, + "learning_rate": 1.8404870873628774e-05, + "loss": 1.0404, + "step": 2150 + }, + { + "epoch": 0.2069065025009619, + "grad_norm": 2.232026871321177, + "learning_rate": 1.8403182431167828e-05, + "loss": 0.9584, + "step": 2151 + }, + { + "epoch": 0.2070026933435937, + "grad_norm": 2.164040011858402, + "learning_rate": 1.8401493173105675e-05, + "loss": 1.1276, + "step": 2152 + }, + { + "epoch": 0.20709888418622546, + "grad_norm": 2.369729026074161, + "learning_rate": 1.8399803099606268e-05, + "loss": 1.1339, + "step": 2153 + }, + { + "epoch": 0.20719507502885726, + "grad_norm": 2.1673256390774007, + "learning_rate": 1.8398112210833648e-05, + "loss": 0.853, + "step": 2154 + }, + { + "epoch": 0.20729126587148902, + "grad_norm": 2.0335014111213456, + "learning_rate": 1.839642050695193e-05, + "loss": 0.9662, + "step": 2155 + }, + { + "epoch": 0.20738745671412082, + "grad_norm": 2.871283823502028, + "learning_rate": 1.8394727988125308e-05, + "loss": 1.164, + "step": 2156 + }, + { + "epoch": 0.2074836475567526, + "grad_norm": 2.5140463024059536, + "learning_rate": 1.8393034654518057e-05, + "loss": 1.0914, + "step": 2157 + }, + { + "epoch": 0.20757983839938438, + "grad_norm": 2.3450356933217127, + "learning_rate": 1.8391340506294524e-05, + "loss": 1.0925, + "step": 2158 + }, + { + "epoch": 0.20767602924201617, + "grad_norm": 2.351303101367869, + "learning_rate": 1.838964554361915e-05, + "loss": 1.0718, + "step": 2159 + }, + { + "epoch": 0.20777222008464794, + "grad_norm": 2.3718318906303337, + "learning_rate": 1.8387949766656434e-05, + "loss": 1.0692, + "step": 2160 + }, + { + "epoch": 0.20786841092727973, + "grad_norm": 1.9971950313662612, + "learning_rate": 1.8386253175570975e-05, + "loss": 1.0924, + "step": 2161 + }, + { + "epoch": 0.2079646017699115, + "grad_norm": 2.4678921850759576, + "learning_rate": 1.8384555770527438e-05, + "loss": 1.1204, + "step": 2162 + }, + { + "epoch": 0.2080607926125433, + "grad_norm": 2.568036451927626, + "learning_rate": 1.8382857551690572e-05, + "loss": 1.0963, + "step": 2163 + }, + { + "epoch": 0.20815698345517505, + "grad_norm": 2.3006627706038034, + "learning_rate": 1.8381158519225204e-05, + "loss": 0.893, + "step": 2164 + }, + { + "epoch": 0.20825317429780685, + "grad_norm": 2.5790330777659043, + "learning_rate": 1.8379458673296238e-05, + "loss": 0.9699, + "step": 2165 + }, + { + "epoch": 0.20834936514043864, + "grad_norm": 2.3269369883794075, + "learning_rate": 1.8377758014068662e-05, + "loss": 1.1593, + "step": 2166 + }, + { + "epoch": 0.2084455559830704, + "grad_norm": 2.3911757455231006, + "learning_rate": 1.8376056541707532e-05, + "loss": 1.1287, + "step": 2167 + }, + { + "epoch": 0.2085417468257022, + "grad_norm": 2.1272098180202583, + "learning_rate": 1.8374354256378e-05, + "loss": 1.0723, + "step": 2168 + }, + { + "epoch": 0.20863793766833397, + "grad_norm": 2.5246513589823314, + "learning_rate": 1.8372651158245284e-05, + "loss": 1.1189, + "step": 2169 + }, + { + "epoch": 0.20873412851096576, + "grad_norm": 2.3963857569759615, + "learning_rate": 1.837094724747468e-05, + "loss": 1.0557, + "step": 2170 + }, + { + "epoch": 0.20883031935359753, + "grad_norm": 2.4859162678637223, + "learning_rate": 1.8369242524231573e-05, + "loss": 1.0034, + "step": 2171 + }, + { + "epoch": 0.20892651019622932, + "grad_norm": 2.2118052042869327, + "learning_rate": 1.8367536988681422e-05, + "loss": 0.938, + "step": 2172 + }, + { + "epoch": 0.20902270103886111, + "grad_norm": 2.3756283099130115, + "learning_rate": 1.8365830640989755e-05, + "loss": 1.009, + "step": 2173 + }, + { + "epoch": 0.20911889188149288, + "grad_norm": 2.5689763873728944, + "learning_rate": 1.83641234813222e-05, + "loss": 1.1701, + "step": 2174 + }, + { + "epoch": 0.20921508272412467, + "grad_norm": 2.018972442221367, + "learning_rate": 1.836241550984444e-05, + "loss": 0.8242, + "step": 2175 + }, + { + "epoch": 0.20931127356675644, + "grad_norm": 2.1978765342725715, + "learning_rate": 1.8360706726722253e-05, + "loss": 1.0924, + "step": 2176 + }, + { + "epoch": 0.20940746440938823, + "grad_norm": 2.0874600238697347, + "learning_rate": 1.8358997132121493e-05, + "loss": 1.0358, + "step": 2177 + }, + { + "epoch": 0.20950365525202, + "grad_norm": 2.3018443978506657, + "learning_rate": 1.835728672620809e-05, + "loss": 1.1663, + "step": 2178 + }, + { + "epoch": 0.2095998460946518, + "grad_norm": 2.348287363499548, + "learning_rate": 1.8355575509148055e-05, + "loss": 1.073, + "step": 2179 + }, + { + "epoch": 0.20969603693728356, + "grad_norm": 2.1198683757982444, + "learning_rate": 1.8353863481107473e-05, + "loss": 1.0183, + "step": 2180 + }, + { + "epoch": 0.20979222777991535, + "grad_norm": 2.2251766705266487, + "learning_rate": 1.835215064225251e-05, + "loss": 0.9803, + "step": 2181 + }, + { + "epoch": 0.20988841862254715, + "grad_norm": 2.394363382476245, + "learning_rate": 1.835043699274942e-05, + "loss": 1.1526, + "step": 2182 + }, + { + "epoch": 0.2099846094651789, + "grad_norm": 2.315907002597464, + "learning_rate": 1.8348722532764513e-05, + "loss": 1.0614, + "step": 2183 + }, + { + "epoch": 0.2100808003078107, + "grad_norm": 1.9907363626578545, + "learning_rate": 1.8347007262464206e-05, + "loss": 0.9741, + "step": 2184 + }, + { + "epoch": 0.21017699115044247, + "grad_norm": 2.0927400819295987, + "learning_rate": 1.8345291182014977e-05, + "loss": 1.0412, + "step": 2185 + }, + { + "epoch": 0.21027318199307427, + "grad_norm": 2.540496983588246, + "learning_rate": 1.8343574291583385e-05, + "loss": 1.0999, + "step": 2186 + }, + { + "epoch": 0.21036937283570603, + "grad_norm": 1.954740928787225, + "learning_rate": 1.8341856591336067e-05, + "loss": 1.0896, + "step": 2187 + }, + { + "epoch": 0.21046556367833782, + "grad_norm": 2.293616221891646, + "learning_rate": 1.8340138081439743e-05, + "loss": 1.1189, + "step": 2188 + }, + { + "epoch": 0.2105617545209696, + "grad_norm": 2.1272792235807443, + "learning_rate": 1.8338418762061207e-05, + "loss": 0.999, + "step": 2189 + }, + { + "epoch": 0.21065794536360138, + "grad_norm": 2.0694948513037565, + "learning_rate": 1.833669863336734e-05, + "loss": 1.0612, + "step": 2190 + }, + { + "epoch": 0.21075413620623318, + "grad_norm": 2.3895048938628864, + "learning_rate": 1.8334977695525086e-05, + "loss": 1.1241, + "step": 2191 + }, + { + "epoch": 0.21085032704886494, + "grad_norm": 2.3273705727468883, + "learning_rate": 1.833325594870148e-05, + "loss": 0.8797, + "step": 2192 + }, + { + "epoch": 0.21094651789149674, + "grad_norm": 2.646897957252055, + "learning_rate": 1.833153339306364e-05, + "loss": 1.1382, + "step": 2193 + }, + { + "epoch": 0.2110427087341285, + "grad_norm": 1.9918010699860682, + "learning_rate": 1.8329810028778747e-05, + "loss": 1.0996, + "step": 2194 + }, + { + "epoch": 0.2111388995767603, + "grad_norm": 2.1682377012249496, + "learning_rate": 1.832808585601407e-05, + "loss": 1.0881, + "step": 2195 + }, + { + "epoch": 0.21123509041939206, + "grad_norm": 2.3442067767828347, + "learning_rate": 1.8326360874936952e-05, + "loss": 1.1052, + "step": 2196 + }, + { + "epoch": 0.21133128126202386, + "grad_norm": 1.976596391650244, + "learning_rate": 1.832463508571482e-05, + "loss": 1.0815, + "step": 2197 + }, + { + "epoch": 0.21142747210465565, + "grad_norm": 2.113972481321962, + "learning_rate": 1.8322908488515182e-05, + "loss": 1.0454, + "step": 2198 + }, + { + "epoch": 0.21152366294728742, + "grad_norm": 2.156685245555923, + "learning_rate": 1.8321181083505612e-05, + "loss": 1.0324, + "step": 2199 + }, + { + "epoch": 0.2116198537899192, + "grad_norm": 2.2573126857755983, + "learning_rate": 1.8319452870853772e-05, + "loss": 1.0903, + "step": 2200 + }, + { + "epoch": 0.21171604463255098, + "grad_norm": 2.005952815705245, + "learning_rate": 1.8317723850727395e-05, + "loss": 1.0848, + "step": 2201 + }, + { + "epoch": 0.21181223547518277, + "grad_norm": 2.115578365970604, + "learning_rate": 1.8315994023294306e-05, + "loss": 0.9294, + "step": 2202 + }, + { + "epoch": 0.21190842631781454, + "grad_norm": 1.9578608654304988, + "learning_rate": 1.8314263388722397e-05, + "loss": 1.0589, + "step": 2203 + }, + { + "epoch": 0.21200461716044633, + "grad_norm": 2.72570378945444, + "learning_rate": 1.8312531947179634e-05, + "loss": 1.1396, + "step": 2204 + }, + { + "epoch": 0.2121008080030781, + "grad_norm": 2.2453365819164124, + "learning_rate": 1.831079969883408e-05, + "loss": 0.9657, + "step": 2205 + }, + { + "epoch": 0.2121969988457099, + "grad_norm": 2.0337209152678715, + "learning_rate": 1.8309066643853854e-05, + "loss": 1.0424, + "step": 2206 + }, + { + "epoch": 0.21229318968834168, + "grad_norm": 2.4293083864815985, + "learning_rate": 1.830733278240717e-05, + "loss": 1.0297, + "step": 2207 + }, + { + "epoch": 0.21238938053097345, + "grad_norm": 2.292025046999471, + "learning_rate": 1.8305598114662312e-05, + "loss": 1.1114, + "step": 2208 + }, + { + "epoch": 0.21248557137360524, + "grad_norm": 2.096823132337366, + "learning_rate": 1.8303862640787647e-05, + "loss": 1.0551, + "step": 2209 + }, + { + "epoch": 0.212581762216237, + "grad_norm": 2.461003859135117, + "learning_rate": 1.830212636095161e-05, + "loss": 1.0302, + "step": 2210 + }, + { + "epoch": 0.2126779530588688, + "grad_norm": 2.5726066847504723, + "learning_rate": 1.8300389275322732e-05, + "loss": 0.9806, + "step": 2211 + }, + { + "epoch": 0.21277414390150057, + "grad_norm": 2.3094047154539625, + "learning_rate": 1.8298651384069605e-05, + "loss": 0.9876, + "step": 2212 + }, + { + "epoch": 0.21287033474413236, + "grad_norm": 2.193974185529885, + "learning_rate": 1.829691268736091e-05, + "loss": 1.1446, + "step": 2213 + }, + { + "epoch": 0.21296652558676413, + "grad_norm": 2.3247536728599267, + "learning_rate": 1.8295173185365405e-05, + "loss": 0.9804, + "step": 2214 + }, + { + "epoch": 0.21306271642939592, + "grad_norm": 2.436722070952997, + "learning_rate": 1.8293432878251916e-05, + "loss": 1.0903, + "step": 2215 + }, + { + "epoch": 0.21315890727202771, + "grad_norm": 2.5044063050955865, + "learning_rate": 1.829169176618936e-05, + "loss": 0.9433, + "step": 2216 + }, + { + "epoch": 0.21325509811465948, + "grad_norm": 2.0695507050000286, + "learning_rate": 1.8289949849346723e-05, + "loss": 1.0255, + "step": 2217 + }, + { + "epoch": 0.21335128895729127, + "grad_norm": 2.1456375403870345, + "learning_rate": 1.828820712789308e-05, + "loss": 1.0584, + "step": 2218 + }, + { + "epoch": 0.21344747979992304, + "grad_norm": 2.0731584333652013, + "learning_rate": 1.828646360199757e-05, + "loss": 1.0191, + "step": 2219 + }, + { + "epoch": 0.21354367064255483, + "grad_norm": 2.4610413898557626, + "learning_rate": 1.828471927182942e-05, + "loss": 0.9622, + "step": 2220 + }, + { + "epoch": 0.2136398614851866, + "grad_norm": 1.9694514847650522, + "learning_rate": 1.8282974137557934e-05, + "loss": 1.14, + "step": 2221 + }, + { + "epoch": 0.2137360523278184, + "grad_norm": 2.444664970690522, + "learning_rate": 1.828122819935249e-05, + "loss": 1.1438, + "step": 2222 + }, + { + "epoch": 0.2138322431704502, + "grad_norm": 2.4031414490383933, + "learning_rate": 1.8279481457382546e-05, + "loss": 1.1062, + "step": 2223 + }, + { + "epoch": 0.21392843401308195, + "grad_norm": 2.1661240898548604, + "learning_rate": 1.8277733911817642e-05, + "loss": 0.8982, + "step": 2224 + }, + { + "epoch": 0.21402462485571375, + "grad_norm": 2.504507781063018, + "learning_rate": 1.8275985562827388e-05, + "loss": 1.1996, + "step": 2225 + }, + { + "epoch": 0.2141208156983455, + "grad_norm": 2.543173850520047, + "learning_rate": 1.8274236410581478e-05, + "loss": 1.0817, + "step": 2226 + }, + { + "epoch": 0.2142170065409773, + "grad_norm": 2.357775664362809, + "learning_rate": 1.8272486455249682e-05, + "loss": 1.071, + "step": 2227 + }, + { + "epoch": 0.21431319738360907, + "grad_norm": 2.0200676823755868, + "learning_rate": 1.827073569700185e-05, + "loss": 0.9782, + "step": 2228 + }, + { + "epoch": 0.21440938822624087, + "grad_norm": 2.1040386170992877, + "learning_rate": 1.8268984136007904e-05, + "loss": 1.03, + "step": 2229 + }, + { + "epoch": 0.21450557906887263, + "grad_norm": 2.344534202526496, + "learning_rate": 1.8267231772437854e-05, + "loss": 1.1524, + "step": 2230 + }, + { + "epoch": 0.21460176991150443, + "grad_norm": 2.0871602224891395, + "learning_rate": 1.8265478606461776e-05, + "loss": 0.9797, + "step": 2231 + }, + { + "epoch": 0.21469796075413622, + "grad_norm": 2.355580513993654, + "learning_rate": 1.8263724638249834e-05, + "loss": 1.1638, + "step": 2232 + }, + { + "epoch": 0.21479415159676798, + "grad_norm": 2.0715098682266353, + "learning_rate": 1.8261969867972263e-05, + "loss": 0.9004, + "step": 2233 + }, + { + "epoch": 0.21489034243939978, + "grad_norm": 2.282071286597523, + "learning_rate": 1.8260214295799382e-05, + "loss": 1.1089, + "step": 2234 + }, + { + "epoch": 0.21498653328203154, + "grad_norm": 2.6956465083930397, + "learning_rate": 1.8258457921901583e-05, + "loss": 1.108, + "step": 2235 + }, + { + "epoch": 0.21508272412466334, + "grad_norm": 2.317121807505584, + "learning_rate": 1.825670074644933e-05, + "loss": 1.0183, + "step": 2236 + }, + { + "epoch": 0.2151789149672951, + "grad_norm": 2.371078421879368, + "learning_rate": 1.8254942769613186e-05, + "loss": 1.1214, + "step": 2237 + }, + { + "epoch": 0.2152751058099269, + "grad_norm": 2.0156729724189737, + "learning_rate": 1.8253183991563768e-05, + "loss": 1.046, + "step": 2238 + }, + { + "epoch": 0.21537129665255866, + "grad_norm": 2.2987771683943072, + "learning_rate": 1.8251424412471782e-05, + "loss": 0.9773, + "step": 2239 + }, + { + "epoch": 0.21546748749519046, + "grad_norm": 2.177370338393166, + "learning_rate": 1.824966403250801e-05, + "loss": 1.0083, + "step": 2240 + }, + { + "epoch": 0.21556367833782225, + "grad_norm": 2.197293280892706, + "learning_rate": 1.8247902851843316e-05, + "loss": 1.0095, + "step": 2241 + }, + { + "epoch": 0.21565986918045402, + "grad_norm": 2.2605647869298284, + "learning_rate": 1.8246140870648633e-05, + "loss": 1.0837, + "step": 2242 + }, + { + "epoch": 0.2157560600230858, + "grad_norm": 2.2330278701318758, + "learning_rate": 1.824437808909498e-05, + "loss": 1.0842, + "step": 2243 + }, + { + "epoch": 0.21585225086571758, + "grad_norm": 2.1566260402547077, + "learning_rate": 1.8242614507353446e-05, + "loss": 1.0046, + "step": 2244 + }, + { + "epoch": 0.21594844170834937, + "grad_norm": 1.9805021470618231, + "learning_rate": 1.8240850125595203e-05, + "loss": 1.037, + "step": 2245 + }, + { + "epoch": 0.21604463255098114, + "grad_norm": 2.320668334611073, + "learning_rate": 1.8239084943991507e-05, + "loss": 1.1585, + "step": 2246 + }, + { + "epoch": 0.21614082339361293, + "grad_norm": 2.4693704071623506, + "learning_rate": 1.823731896271367e-05, + "loss": 1.182, + "step": 2247 + }, + { + "epoch": 0.2162370142362447, + "grad_norm": 2.079361994566818, + "learning_rate": 1.823555218193311e-05, + "loss": 1.0746, + "step": 2248 + }, + { + "epoch": 0.2163332050788765, + "grad_norm": 2.0536457202321894, + "learning_rate": 1.8233784601821298e-05, + "loss": 1.0434, + "step": 2249 + }, + { + "epoch": 0.21642939592150828, + "grad_norm": 2.3959996038013283, + "learning_rate": 1.8232016222549797e-05, + "loss": 1.0498, + "step": 2250 + }, + { + "epoch": 0.21652558676414005, + "grad_norm": 2.218192707685922, + "learning_rate": 1.8230247044290244e-05, + "loss": 1.0964, + "step": 2251 + }, + { + "epoch": 0.21662177760677184, + "grad_norm": 2.44706029863568, + "learning_rate": 1.8228477067214352e-05, + "loss": 0.982, + "step": 2252 + }, + { + "epoch": 0.2167179684494036, + "grad_norm": 2.418341440590914, + "learning_rate": 1.8226706291493913e-05, + "loss": 1.0729, + "step": 2253 + }, + { + "epoch": 0.2168141592920354, + "grad_norm": 2.588921785210921, + "learning_rate": 1.8224934717300794e-05, + "loss": 1.2071, + "step": 2254 + }, + { + "epoch": 0.21691035013466717, + "grad_norm": 2.3597380779335455, + "learning_rate": 1.8223162344806945e-05, + "loss": 1.0356, + "step": 2255 + }, + { + "epoch": 0.21700654097729896, + "grad_norm": 2.4485494625768127, + "learning_rate": 1.8221389174184385e-05, + "loss": 1.0081, + "step": 2256 + }, + { + "epoch": 0.21710273181993076, + "grad_norm": 2.2516895400758923, + "learning_rate": 1.8219615205605222e-05, + "loss": 1.0094, + "step": 2257 + }, + { + "epoch": 0.21719892266256252, + "grad_norm": 2.312850398256954, + "learning_rate": 1.8217840439241633e-05, + "loss": 0.9188, + "step": 2258 + }, + { + "epoch": 0.21729511350519432, + "grad_norm": 2.3105046018129602, + "learning_rate": 1.8216064875265868e-05, + "loss": 1.129, + "step": 2259 + }, + { + "epoch": 0.21739130434782608, + "grad_norm": 2.0837997123622594, + "learning_rate": 1.8214288513850267e-05, + "loss": 1.077, + "step": 2260 + }, + { + "epoch": 0.21748749519045787, + "grad_norm": 2.151858851408048, + "learning_rate": 1.821251135516724e-05, + "loss": 1.02, + "step": 2261 + }, + { + "epoch": 0.21758368603308964, + "grad_norm": 2.513352328834387, + "learning_rate": 1.8210733399389277e-05, + "loss": 1.0746, + "step": 2262 + }, + { + "epoch": 0.21767987687572143, + "grad_norm": 2.3437345111462258, + "learning_rate": 1.8208954646688942e-05, + "loss": 1.0853, + "step": 2263 + }, + { + "epoch": 0.2177760677183532, + "grad_norm": 2.2690695450573903, + "learning_rate": 1.820717509723888e-05, + "loss": 1.012, + "step": 2264 + }, + { + "epoch": 0.217872258560985, + "grad_norm": 2.343651858154551, + "learning_rate": 1.8205394751211803e-05, + "loss": 1.2269, + "step": 2265 + }, + { + "epoch": 0.2179684494036168, + "grad_norm": 2.2934714135191188, + "learning_rate": 1.8203613608780525e-05, + "loss": 0.939, + "step": 2266 + }, + { + "epoch": 0.21806464024624855, + "grad_norm": 2.417970627420252, + "learning_rate": 1.820183167011791e-05, + "loss": 0.8625, + "step": 2267 + }, + { + "epoch": 0.21816083108888035, + "grad_norm": 2.0593938907091287, + "learning_rate": 1.8200048935396908e-05, + "loss": 0.9441, + "step": 2268 + }, + { + "epoch": 0.2182570219315121, + "grad_norm": 2.24459571596215, + "learning_rate": 1.819826540479056e-05, + "loss": 1.0367, + "step": 2269 + }, + { + "epoch": 0.2183532127741439, + "grad_norm": 3.090538802810976, + "learning_rate": 1.819648107847196e-05, + "loss": 1.0135, + "step": 2270 + }, + { + "epoch": 0.21844940361677567, + "grad_norm": 2.4280154884090965, + "learning_rate": 1.8194695956614304e-05, + "loss": 1.1427, + "step": 2271 + }, + { + "epoch": 0.21854559445940747, + "grad_norm": 2.3033792922993235, + "learning_rate": 1.8192910039390844e-05, + "loss": 1.0774, + "step": 2272 + }, + { + "epoch": 0.21864178530203923, + "grad_norm": 2.5230948022907382, + "learning_rate": 1.8191123326974923e-05, + "loss": 0.9832, + "step": 2273 + }, + { + "epoch": 0.21873797614467103, + "grad_norm": 2.2137032705495514, + "learning_rate": 1.8189335819539963e-05, + "loss": 0.8489, + "step": 2274 + }, + { + "epoch": 0.21883416698730282, + "grad_norm": 1.9675847709473724, + "learning_rate": 1.8187547517259446e-05, + "loss": 0.8525, + "step": 2275 + }, + { + "epoch": 0.21893035782993459, + "grad_norm": 2.4624226990288602, + "learning_rate": 1.8185758420306947e-05, + "loss": 0.8618, + "step": 2276 + }, + { + "epoch": 0.21902654867256638, + "grad_norm": 2.757994771866779, + "learning_rate": 1.818396852885611e-05, + "loss": 1.0709, + "step": 2277 + }, + { + "epoch": 0.21912273951519814, + "grad_norm": 2.1889311789651114, + "learning_rate": 1.818217784308067e-05, + "loss": 1.1123, + "step": 2278 + }, + { + "epoch": 0.21921893035782994, + "grad_norm": 2.3459101880562896, + "learning_rate": 1.8180386363154413e-05, + "loss": 1.0265, + "step": 2279 + }, + { + "epoch": 0.2193151212004617, + "grad_norm": 2.2688203935241904, + "learning_rate": 1.817859408925123e-05, + "loss": 0.918, + "step": 2280 + }, + { + "epoch": 0.2194113120430935, + "grad_norm": 2.0067980637722136, + "learning_rate": 1.817680102154507e-05, + "loss": 1.0889, + "step": 2281 + }, + { + "epoch": 0.2195075028857253, + "grad_norm": 2.1898152155830886, + "learning_rate": 1.817500716020997e-05, + "loss": 1.0643, + "step": 2282 + }, + { + "epoch": 0.21960369372835706, + "grad_norm": 2.325619377645398, + "learning_rate": 1.8173212505420033e-05, + "loss": 1.1074, + "step": 2283 + }, + { + "epoch": 0.21969988457098885, + "grad_norm": 2.3558970671205843, + "learning_rate": 1.8171417057349457e-05, + "loss": 1.0846, + "step": 2284 + }, + { + "epoch": 0.21979607541362062, + "grad_norm": 2.225204657934813, + "learning_rate": 1.8169620816172492e-05, + "loss": 0.9798, + "step": 2285 + }, + { + "epoch": 0.2198922662562524, + "grad_norm": 2.1447357179134987, + "learning_rate": 1.816782378206349e-05, + "loss": 1.0004, + "step": 2286 + }, + { + "epoch": 0.21998845709888418, + "grad_norm": 1.8979200602690105, + "learning_rate": 1.8166025955196863e-05, + "loss": 0.9002, + "step": 2287 + }, + { + "epoch": 0.22008464794151597, + "grad_norm": 1.960120148958014, + "learning_rate": 1.8164227335747108e-05, + "loss": 1.0143, + "step": 2288 + }, + { + "epoch": 0.22018083878414774, + "grad_norm": 2.0702217184455294, + "learning_rate": 1.816242792388879e-05, + "loss": 0.8789, + "step": 2289 + }, + { + "epoch": 0.22027702962677953, + "grad_norm": 2.0761136576520265, + "learning_rate": 1.8160627719796568e-05, + "loss": 1.1109, + "step": 2290 + }, + { + "epoch": 0.22037322046941132, + "grad_norm": 1.986102898447727, + "learning_rate": 1.815882672364516e-05, + "loss": 0.9036, + "step": 2291 + }, + { + "epoch": 0.2204694113120431, + "grad_norm": 2.142566723194837, + "learning_rate": 1.815702493560937e-05, + "loss": 1.0646, + "step": 2292 + }, + { + "epoch": 0.22056560215467488, + "grad_norm": 2.362184351876668, + "learning_rate": 1.815522235586408e-05, + "loss": 1.0343, + "step": 2293 + }, + { + "epoch": 0.22066179299730665, + "grad_norm": 2.2172760727626364, + "learning_rate": 1.8153418984584238e-05, + "loss": 1.0608, + "step": 2294 + }, + { + "epoch": 0.22075798383993844, + "grad_norm": 1.899836717790623, + "learning_rate": 1.8151614821944884e-05, + "loss": 0.9469, + "step": 2295 + }, + { + "epoch": 0.2208541746825702, + "grad_norm": 2.344128729116692, + "learning_rate": 1.8149809868121125e-05, + "loss": 0.9879, + "step": 2296 + }, + { + "epoch": 0.220950365525202, + "grad_norm": 1.971511207660929, + "learning_rate": 1.8148004123288147e-05, + "loss": 0.8173, + "step": 2297 + }, + { + "epoch": 0.22104655636783377, + "grad_norm": 2.4324376087158543, + "learning_rate": 1.8146197587621217e-05, + "loss": 1.0402, + "step": 2298 + }, + { + "epoch": 0.22114274721046556, + "grad_norm": 2.1738318406115003, + "learning_rate": 1.814439026129567e-05, + "loss": 1.1162, + "step": 2299 + }, + { + "epoch": 0.22123893805309736, + "grad_norm": 2.276365781067651, + "learning_rate": 1.814258214448692e-05, + "loss": 1.1286, + "step": 2300 + }, + { + "epoch": 0.22133512889572912, + "grad_norm": 2.3233815908347477, + "learning_rate": 1.8140773237370467e-05, + "loss": 1.1448, + "step": 2301 + }, + { + "epoch": 0.22143131973836092, + "grad_norm": 2.31012461160092, + "learning_rate": 1.8138963540121878e-05, + "loss": 1.0585, + "step": 2302 + }, + { + "epoch": 0.22152751058099268, + "grad_norm": 1.9465736754789464, + "learning_rate": 1.8137153052916802e-05, + "loss": 1.0663, + "step": 2303 + }, + { + "epoch": 0.22162370142362448, + "grad_norm": 2.094679785742686, + "learning_rate": 1.813534177593096e-05, + "loss": 0.989, + "step": 2304 + }, + { + "epoch": 0.22171989226625624, + "grad_norm": 1.976518722459748, + "learning_rate": 1.8133529709340153e-05, + "loss": 1.1264, + "step": 2305 + }, + { + "epoch": 0.22181608310888803, + "grad_norm": 2.2402047021376017, + "learning_rate": 1.8131716853320254e-05, + "loss": 1.0143, + "step": 2306 + }, + { + "epoch": 0.22191227395151983, + "grad_norm": 2.028600102140142, + "learning_rate": 1.8129903208047222e-05, + "loss": 1.0574, + "step": 2307 + }, + { + "epoch": 0.2220084647941516, + "grad_norm": 2.0853830973318215, + "learning_rate": 1.8128088773697086e-05, + "loss": 0.9793, + "step": 2308 + }, + { + "epoch": 0.2221046556367834, + "grad_norm": 2.389087310545524, + "learning_rate": 1.812627355044595e-05, + "loss": 1.0941, + "step": 2309 + }, + { + "epoch": 0.22220084647941515, + "grad_norm": 2.1925681515733926, + "learning_rate": 1.8124457538469996e-05, + "loss": 1.025, + "step": 2310 + }, + { + "epoch": 0.22229703732204695, + "grad_norm": 2.4494329633450445, + "learning_rate": 1.8122640737945483e-05, + "loss": 1.1245, + "step": 2311 + }, + { + "epoch": 0.2223932281646787, + "grad_norm": 2.176129601956615, + "learning_rate": 1.8120823149048753e-05, + "loss": 0.9035, + "step": 2312 + }, + { + "epoch": 0.2224894190073105, + "grad_norm": 2.288189310528669, + "learning_rate": 1.8119004771956214e-05, + "loss": 1.118, + "step": 2313 + }, + { + "epoch": 0.22258560984994227, + "grad_norm": 2.0699147936775035, + "learning_rate": 1.811718560684436e-05, + "loss": 1.0444, + "step": 2314 + }, + { + "epoch": 0.22268180069257407, + "grad_norm": 2.219811711174476, + "learning_rate": 1.8115365653889747e-05, + "loss": 1.0531, + "step": 2315 + }, + { + "epoch": 0.22277799153520586, + "grad_norm": 2.121532782682548, + "learning_rate": 1.8113544913269025e-05, + "loss": 0.9688, + "step": 2316 + }, + { + "epoch": 0.22287418237783763, + "grad_norm": 1.8358722146636173, + "learning_rate": 1.8111723385158916e-05, + "loss": 1.0558, + "step": 2317 + }, + { + "epoch": 0.22297037322046942, + "grad_norm": 2.2628454485178287, + "learning_rate": 1.8109901069736202e-05, + "loss": 1.1553, + "step": 2318 + }, + { + "epoch": 0.22306656406310119, + "grad_norm": 2.1664625825396926, + "learning_rate": 1.8108077967177765e-05, + "loss": 1.0569, + "step": 2319 + }, + { + "epoch": 0.22316275490573298, + "grad_norm": 2.2717777148068725, + "learning_rate": 1.8106254077660552e-05, + "loss": 1.0893, + "step": 2320 + }, + { + "epoch": 0.22325894574836475, + "grad_norm": 2.3322608486260172, + "learning_rate": 1.8104429401361583e-05, + "loss": 1.1739, + "step": 2321 + }, + { + "epoch": 0.22335513659099654, + "grad_norm": 2.148015043660487, + "learning_rate": 1.810260393845796e-05, + "loss": 1.0727, + "step": 2322 + }, + { + "epoch": 0.2234513274336283, + "grad_norm": 2.023443755058333, + "learning_rate": 1.810077768912686e-05, + "loss": 0.8684, + "step": 2323 + }, + { + "epoch": 0.2235475182762601, + "grad_norm": 2.3899040213673204, + "learning_rate": 1.809895065354554e-05, + "loss": 1.0361, + "step": 2324 + }, + { + "epoch": 0.2236437091188919, + "grad_norm": 2.2247543620993024, + "learning_rate": 1.8097122831891325e-05, + "loss": 1.0938, + "step": 2325 + }, + { + "epoch": 0.22373989996152366, + "grad_norm": 2.5825276669886166, + "learning_rate": 1.8095294224341622e-05, + "loss": 1.0848, + "step": 2326 + }, + { + "epoch": 0.22383609080415545, + "grad_norm": 2.5854579806180196, + "learning_rate": 1.8093464831073912e-05, + "loss": 1.0516, + "step": 2327 + }, + { + "epoch": 0.22393228164678722, + "grad_norm": 2.154966591999431, + "learning_rate": 1.8091634652265755e-05, + "loss": 1.1485, + "step": 2328 + }, + { + "epoch": 0.224028472489419, + "grad_norm": 2.234385453554007, + "learning_rate": 1.8089803688094786e-05, + "loss": 1.0879, + "step": 2329 + }, + { + "epoch": 0.22412466333205078, + "grad_norm": 2.5338388100086564, + "learning_rate": 1.8087971938738715e-05, + "loss": 1.0265, + "step": 2330 + }, + { + "epoch": 0.22422085417468257, + "grad_norm": 1.9364741511307904, + "learning_rate": 1.8086139404375328e-05, + "loss": 1.1364, + "step": 2331 + }, + { + "epoch": 0.22431704501731436, + "grad_norm": 2.1272837851420614, + "learning_rate": 1.808430608518249e-05, + "loss": 1.1382, + "step": 2332 + }, + { + "epoch": 0.22441323585994613, + "grad_norm": 2.0755712368685924, + "learning_rate": 1.8082471981338138e-05, + "loss": 0.9874, + "step": 2333 + }, + { + "epoch": 0.22450942670257792, + "grad_norm": 2.1169786884322406, + "learning_rate": 1.808063709302029e-05, + "loss": 1.0964, + "step": 2334 + }, + { + "epoch": 0.2246056175452097, + "grad_norm": 2.440016579009818, + "learning_rate": 1.807880142040704e-05, + "loss": 1.0571, + "step": 2335 + }, + { + "epoch": 0.22470180838784148, + "grad_norm": 2.0576592217653524, + "learning_rate": 1.807696496367655e-05, + "loss": 0.9796, + "step": 2336 + }, + { + "epoch": 0.22479799923047325, + "grad_norm": 2.198633217740797, + "learning_rate": 1.8075127723007066e-05, + "loss": 1.0058, + "step": 2337 + }, + { + "epoch": 0.22489419007310504, + "grad_norm": 2.170216626092998, + "learning_rate": 1.8073289698576913e-05, + "loss": 1.1839, + "step": 2338 + }, + { + "epoch": 0.2249903809157368, + "grad_norm": 2.152534448904729, + "learning_rate": 1.8071450890564476e-05, + "loss": 0.9149, + "step": 2339 + }, + { + "epoch": 0.2250865717583686, + "grad_norm": 2.210818757625676, + "learning_rate": 1.8069611299148236e-05, + "loss": 1.04, + "step": 2340 + }, + { + "epoch": 0.2251827626010004, + "grad_norm": 2.229019890297196, + "learning_rate": 1.8067770924506746e-05, + "loss": 0.9903, + "step": 2341 + }, + { + "epoch": 0.22527895344363216, + "grad_norm": 2.3535602316046274, + "learning_rate": 1.8065929766818617e-05, + "loss": 1.1784, + "step": 2342 + }, + { + "epoch": 0.22537514428626396, + "grad_norm": 2.2809771575329028, + "learning_rate": 1.8064087826262557e-05, + "loss": 1.0276, + "step": 2343 + }, + { + "epoch": 0.22547133512889572, + "grad_norm": 1.9610802533442673, + "learning_rate": 1.806224510301734e-05, + "loss": 0.93, + "step": 2344 + }, + { + "epoch": 0.22556752597152752, + "grad_norm": 2.3389981895136644, + "learning_rate": 1.806040159726182e-05, + "loss": 1.0145, + "step": 2345 + }, + { + "epoch": 0.22566371681415928, + "grad_norm": 2.158410680299598, + "learning_rate": 1.8058557309174926e-05, + "loss": 1.0641, + "step": 2346 + }, + { + "epoch": 0.22575990765679108, + "grad_norm": 2.229137469720704, + "learning_rate": 1.8056712238935657e-05, + "loss": 1.0996, + "step": 2347 + }, + { + "epoch": 0.22585609849942284, + "grad_norm": 2.039331917952335, + "learning_rate": 1.8054866386723096e-05, + "loss": 0.9913, + "step": 2348 + }, + { + "epoch": 0.22595228934205464, + "grad_norm": 2.2947519283457423, + "learning_rate": 1.80530197527164e-05, + "loss": 1.0858, + "step": 2349 + }, + { + "epoch": 0.22604848018468643, + "grad_norm": 2.1280192724197016, + "learning_rate": 1.80511723370948e-05, + "loss": 1.1265, + "step": 2350 + }, + { + "epoch": 0.2261446710273182, + "grad_norm": 2.023272257019241, + "learning_rate": 1.80493241400376e-05, + "loss": 1.0241, + "step": 2351 + }, + { + "epoch": 0.22624086186995, + "grad_norm": 2.3566775623334646, + "learning_rate": 1.804747516172419e-05, + "loss": 1.2064, + "step": 2352 + }, + { + "epoch": 0.22633705271258175, + "grad_norm": 2.1175583983206274, + "learning_rate": 1.8045625402334027e-05, + "loss": 0.9696, + "step": 2353 + }, + { + "epoch": 0.22643324355521355, + "grad_norm": 2.4440948926268504, + "learning_rate": 1.8043774862046644e-05, + "loss": 0.9819, + "step": 2354 + }, + { + "epoch": 0.22652943439784531, + "grad_norm": 2.14039952457623, + "learning_rate": 1.8041923541041652e-05, + "loss": 1.2018, + "step": 2355 + }, + { + "epoch": 0.2266256252404771, + "grad_norm": 2.1284279663629246, + "learning_rate": 1.804007143949874e-05, + "loss": 1.0673, + "step": 2356 + }, + { + "epoch": 0.2267218160831089, + "grad_norm": 2.206898066679002, + "learning_rate": 1.8038218557597668e-05, + "loss": 1.04, + "step": 2357 + }, + { + "epoch": 0.22681800692574067, + "grad_norm": 2.0747281183279833, + "learning_rate": 1.8036364895518272e-05, + "loss": 1.0418, + "step": 2358 + }, + { + "epoch": 0.22691419776837246, + "grad_norm": 1.9426824605828978, + "learning_rate": 1.8034510453440473e-05, + "loss": 1.0382, + "step": 2359 + }, + { + "epoch": 0.22701038861100423, + "grad_norm": 2.196198108091476, + "learning_rate": 1.8032655231544253e-05, + "loss": 1.1807, + "step": 2360 + }, + { + "epoch": 0.22710657945363602, + "grad_norm": 2.4238257327192776, + "learning_rate": 1.8030799230009687e-05, + "loss": 1.0507, + "step": 2361 + }, + { + "epoch": 0.2272027702962678, + "grad_norm": 2.3679182454603733, + "learning_rate": 1.8028942449016903e-05, + "loss": 1.0038, + "step": 2362 + }, + { + "epoch": 0.22729896113889958, + "grad_norm": 1.8325631142086023, + "learning_rate": 1.8027084888746127e-05, + "loss": 0.9601, + "step": 2363 + }, + { + "epoch": 0.22739515198153135, + "grad_norm": 2.4403922111551326, + "learning_rate": 1.8025226549377647e-05, + "loss": 1.1356, + "step": 2364 + }, + { + "epoch": 0.22749134282416314, + "grad_norm": 2.62385571493866, + "learning_rate": 1.8023367431091836e-05, + "loss": 1.1066, + "step": 2365 + }, + { + "epoch": 0.22758753366679493, + "grad_norm": 2.2094048641037674, + "learning_rate": 1.8021507534069133e-05, + "loss": 1.1016, + "step": 2366 + }, + { + "epoch": 0.2276837245094267, + "grad_norm": 2.5565546261289684, + "learning_rate": 1.8019646858490056e-05, + "loss": 1.0408, + "step": 2367 + }, + { + "epoch": 0.2277799153520585, + "grad_norm": 2.1813512930570123, + "learning_rate": 1.8017785404535198e-05, + "loss": 1.1078, + "step": 2368 + }, + { + "epoch": 0.22787610619469026, + "grad_norm": 2.353136572248357, + "learning_rate": 1.801592317238524e-05, + "loss": 1.0938, + "step": 2369 + }, + { + "epoch": 0.22797229703732205, + "grad_norm": 2.1098295644029723, + "learning_rate": 1.8014060162220916e-05, + "loss": 0.872, + "step": 2370 + }, + { + "epoch": 0.22806848787995382, + "grad_norm": 1.9795426426559308, + "learning_rate": 1.8012196374223048e-05, + "loss": 0.9351, + "step": 2371 + }, + { + "epoch": 0.2281646787225856, + "grad_norm": 2.2614036676071545, + "learning_rate": 1.801033180857254e-05, + "loss": 1.1034, + "step": 2372 + }, + { + "epoch": 0.22826086956521738, + "grad_norm": 2.2095795782191625, + "learning_rate": 1.800846646545036e-05, + "loss": 0.9005, + "step": 2373 + }, + { + "epoch": 0.22835706040784917, + "grad_norm": 2.342487720297989, + "learning_rate": 1.8006600345037558e-05, + "loss": 0.9981, + "step": 2374 + }, + { + "epoch": 0.22845325125048097, + "grad_norm": 2.2989152751765465, + "learning_rate": 1.800473344751525e-05, + "loss": 1.0244, + "step": 2375 + }, + { + "epoch": 0.22854944209311273, + "grad_norm": 2.1485769543610767, + "learning_rate": 1.8002865773064644e-05, + "loss": 1.0805, + "step": 2376 + }, + { + "epoch": 0.22864563293574452, + "grad_norm": 2.123417973401048, + "learning_rate": 1.8000997321867005e-05, + "loss": 1.038, + "step": 2377 + }, + { + "epoch": 0.2287418237783763, + "grad_norm": 2.223758222597847, + "learning_rate": 1.799912809410369e-05, + "loss": 0.8893, + "step": 2378 + }, + { + "epoch": 0.22883801462100808, + "grad_norm": 2.2085170525458775, + "learning_rate": 1.7997258089956115e-05, + "loss": 0.9818, + "step": 2379 + }, + { + "epoch": 0.22893420546363985, + "grad_norm": 2.246367822259525, + "learning_rate": 1.799538730960579e-05, + "loss": 1.008, + "step": 2380 + }, + { + "epoch": 0.22903039630627164, + "grad_norm": 2.129768436406726, + "learning_rate": 1.7993515753234285e-05, + "loss": 1.0468, + "step": 2381 + }, + { + "epoch": 0.22912658714890344, + "grad_norm": 2.125215961715066, + "learning_rate": 1.799164342102325e-05, + "loss": 1.0883, + "step": 2382 + }, + { + "epoch": 0.2292227779915352, + "grad_norm": 2.6207933629074764, + "learning_rate": 1.7989770313154413e-05, + "loss": 0.9988, + "step": 2383 + }, + { + "epoch": 0.229318968834167, + "grad_norm": 2.3879136941107246, + "learning_rate": 1.7987896429809573e-05, + "loss": 1.1108, + "step": 2384 + }, + { + "epoch": 0.22941515967679876, + "grad_norm": 2.5447931666548276, + "learning_rate": 1.7986021771170607e-05, + "loss": 1.1355, + "step": 2385 + }, + { + "epoch": 0.22951135051943056, + "grad_norm": 2.3682223317766233, + "learning_rate": 1.798414633741947e-05, + "loss": 1.0437, + "step": 2386 + }, + { + "epoch": 0.22960754136206232, + "grad_norm": 2.091040462998426, + "learning_rate": 1.7982270128738183e-05, + "loss": 0.9901, + "step": 2387 + }, + { + "epoch": 0.22970373220469412, + "grad_norm": 2.2651478892028045, + "learning_rate": 1.7980393145308857e-05, + "loss": 1.1124, + "step": 2388 + }, + { + "epoch": 0.22979992304732588, + "grad_norm": 2.4455984639909203, + "learning_rate": 1.797851538731366e-05, + "loss": 1.1688, + "step": 2389 + }, + { + "epoch": 0.22989611388995768, + "grad_norm": 2.369736731127367, + "learning_rate": 1.797663685493485e-05, + "loss": 0.9202, + "step": 2390 + }, + { + "epoch": 0.22999230473258947, + "grad_norm": 2.1969150590240853, + "learning_rate": 1.7974757548354756e-05, + "loss": 1.0917, + "step": 2391 + }, + { + "epoch": 0.23008849557522124, + "grad_norm": 2.7534111541841786, + "learning_rate": 1.7972877467755777e-05, + "loss": 1.0848, + "step": 2392 + }, + { + "epoch": 0.23018468641785303, + "grad_norm": 2.0835417976326394, + "learning_rate": 1.797099661332039e-05, + "loss": 1.0407, + "step": 2393 + }, + { + "epoch": 0.2302808772604848, + "grad_norm": 2.395944786366268, + "learning_rate": 1.7969114985231152e-05, + "loss": 1.0508, + "step": 2394 + }, + { + "epoch": 0.2303770681031166, + "grad_norm": 2.4471702487129896, + "learning_rate": 1.796723258367069e-05, + "loss": 1.0769, + "step": 2395 + }, + { + "epoch": 0.23047325894574835, + "grad_norm": 2.355677397117836, + "learning_rate": 1.796534940882171e-05, + "loss": 1.1888, + "step": 2396 + }, + { + "epoch": 0.23056944978838015, + "grad_norm": 2.069691674095843, + "learning_rate": 1.7963465460866984e-05, + "loss": 1.0289, + "step": 2397 + }, + { + "epoch": 0.23066564063101191, + "grad_norm": 2.846673538492259, + "learning_rate": 1.7961580739989365e-05, + "loss": 1.1296, + "step": 2398 + }, + { + "epoch": 0.2307618314736437, + "grad_norm": 2.288710489740693, + "learning_rate": 1.795969524637179e-05, + "loss": 1.0453, + "step": 2399 + }, + { + "epoch": 0.2308580223162755, + "grad_norm": 2.1980488809767733, + "learning_rate": 1.795780898019726e-05, + "loss": 1.0922, + "step": 2400 + }, + { + "epoch": 0.23095421315890727, + "grad_norm": 2.4546753638288337, + "learning_rate": 1.7955921941648848e-05, + "loss": 1.0952, + "step": 2401 + }, + { + "epoch": 0.23105040400153906, + "grad_norm": 2.1979362737924295, + "learning_rate": 1.795403413090971e-05, + "loss": 1.0983, + "step": 2402 + }, + { + "epoch": 0.23114659484417083, + "grad_norm": 2.3743542667877318, + "learning_rate": 1.7952145548163074e-05, + "loss": 1.1369, + "step": 2403 + }, + { + "epoch": 0.23124278568680262, + "grad_norm": 2.5100402935163317, + "learning_rate": 1.7950256193592243e-05, + "loss": 0.8065, + "step": 2404 + }, + { + "epoch": 0.2313389765294344, + "grad_norm": 2.4117952586215963, + "learning_rate": 1.7948366067380596e-05, + "loss": 1.0196, + "step": 2405 + }, + { + "epoch": 0.23143516737206618, + "grad_norm": 2.4903856383489065, + "learning_rate": 1.794647516971159e-05, + "loss": 1.1103, + "step": 2406 + }, + { + "epoch": 0.23153135821469797, + "grad_norm": 2.3212197194303723, + "learning_rate": 1.7944583500768743e-05, + "loss": 0.9618, + "step": 2407 + }, + { + "epoch": 0.23162754905732974, + "grad_norm": 2.410285075778254, + "learning_rate": 1.7942691060735666e-05, + "loss": 1.1172, + "step": 2408 + }, + { + "epoch": 0.23172373989996153, + "grad_norm": 2.235718466249348, + "learning_rate": 1.7940797849796034e-05, + "loss": 1.0725, + "step": 2409 + }, + { + "epoch": 0.2318199307425933, + "grad_norm": 2.330592867938315, + "learning_rate": 1.79389038681336e-05, + "loss": 1.0403, + "step": 2410 + }, + { + "epoch": 0.2319161215852251, + "grad_norm": 2.202126286960448, + "learning_rate": 1.7937009115932192e-05, + "loss": 1.011, + "step": 2411 + }, + { + "epoch": 0.23201231242785686, + "grad_norm": 2.259674040523997, + "learning_rate": 1.7935113593375707e-05, + "loss": 1.0182, + "step": 2412 + }, + { + "epoch": 0.23210850327048865, + "grad_norm": 2.3325658347026472, + "learning_rate": 1.793321730064813e-05, + "loss": 1.0229, + "step": 2413 + }, + { + "epoch": 0.23220469411312042, + "grad_norm": 2.2534815832299424, + "learning_rate": 1.7931320237933503e-05, + "loss": 1.0773, + "step": 2414 + }, + { + "epoch": 0.2323008849557522, + "grad_norm": 2.371997669274316, + "learning_rate": 1.7929422405415964e-05, + "loss": 0.824, + "step": 2415 + }, + { + "epoch": 0.232397075798384, + "grad_norm": 2.319727582100772, + "learning_rate": 1.79275238032797e-05, + "loss": 1.1005, + "step": 2416 + }, + { + "epoch": 0.23249326664101577, + "grad_norm": 2.477291825407672, + "learning_rate": 1.7925624431709002e-05, + "loss": 1.0352, + "step": 2417 + }, + { + "epoch": 0.23258945748364757, + "grad_norm": 2.577036943429713, + "learning_rate": 1.7923724290888205e-05, + "loss": 1.2021, + "step": 2418 + }, + { + "epoch": 0.23268564832627933, + "grad_norm": 1.685515020272015, + "learning_rate": 1.7921823381001747e-05, + "loss": 0.9335, + "step": 2419 + }, + { + "epoch": 0.23278183916891113, + "grad_norm": 2.222418942386996, + "learning_rate": 1.791992170223412e-05, + "loss": 1.1068, + "step": 2420 + }, + { + "epoch": 0.2328780300115429, + "grad_norm": 2.071016767653497, + "learning_rate": 1.79180192547699e-05, + "loss": 0.9295, + "step": 2421 + }, + { + "epoch": 0.23297422085417469, + "grad_norm": 2.327125587600783, + "learning_rate": 1.791611603879374e-05, + "loss": 1.0917, + "step": 2422 + }, + { + "epoch": 0.23307041169680645, + "grad_norm": 2.111156344853162, + "learning_rate": 1.7914212054490353e-05, + "loss": 0.97, + "step": 2423 + }, + { + "epoch": 0.23316660253943824, + "grad_norm": 1.8835251329119915, + "learning_rate": 1.791230730204455e-05, + "loss": 0.8422, + "step": 2424 + }, + { + "epoch": 0.23326279338207004, + "grad_norm": 2.8281515665838928, + "learning_rate": 1.7910401781641194e-05, + "loss": 0.9946, + "step": 2425 + }, + { + "epoch": 0.2333589842247018, + "grad_norm": 3.467756376973237, + "learning_rate": 1.7908495493465236e-05, + "loss": 1.1455, + "step": 2426 + }, + { + "epoch": 0.2334551750673336, + "grad_norm": 2.2282994032160532, + "learning_rate": 1.7906588437701697e-05, + "loss": 1.1044, + "step": 2427 + }, + { + "epoch": 0.23355136590996536, + "grad_norm": 2.1670330394302884, + "learning_rate": 1.7904680614535675e-05, + "loss": 1.104, + "step": 2428 + }, + { + "epoch": 0.23364755675259716, + "grad_norm": 2.3072300688021645, + "learning_rate": 1.790277202415234e-05, + "loss": 1.0544, + "step": 2429 + }, + { + "epoch": 0.23374374759522892, + "grad_norm": 2.246587009671983, + "learning_rate": 1.7900862666736935e-05, + "loss": 1.1027, + "step": 2430 + }, + { + "epoch": 0.23383993843786072, + "grad_norm": 2.320523066019714, + "learning_rate": 1.7898952542474778e-05, + "loss": 1.1351, + "step": 2431 + }, + { + "epoch": 0.2339361292804925, + "grad_norm": 2.301611827881962, + "learning_rate": 1.789704165155127e-05, + "loss": 1.0888, + "step": 2432 + }, + { + "epoch": 0.23403232012312428, + "grad_norm": 2.1540900456281684, + "learning_rate": 1.7895129994151874e-05, + "loss": 0.9737, + "step": 2433 + }, + { + "epoch": 0.23412851096575607, + "grad_norm": 2.807624825650472, + "learning_rate": 1.7893217570462134e-05, + "loss": 1.0205, + "step": 2434 + }, + { + "epoch": 0.23422470180838784, + "grad_norm": 2.4538569634154257, + "learning_rate": 1.7891304380667672e-05, + "loss": 1.2005, + "step": 2435 + }, + { + "epoch": 0.23432089265101963, + "grad_norm": 2.287584055177564, + "learning_rate": 1.7889390424954168e-05, + "loss": 1.0966, + "step": 2436 + }, + { + "epoch": 0.2344170834936514, + "grad_norm": 2.267936129831169, + "learning_rate": 1.7887475703507398e-05, + "loss": 1.119, + "step": 2437 + }, + { + "epoch": 0.2345132743362832, + "grad_norm": 2.1295709382342203, + "learning_rate": 1.78855602165132e-05, + "loss": 0.9759, + "step": 2438 + }, + { + "epoch": 0.23460946517891496, + "grad_norm": 2.3870925019004656, + "learning_rate": 1.7883643964157485e-05, + "loss": 1.0366, + "step": 2439 + }, + { + "epoch": 0.23470565602154675, + "grad_norm": 2.2660854456135224, + "learning_rate": 1.7881726946626244e-05, + "loss": 1.1481, + "step": 2440 + }, + { + "epoch": 0.23480184686417854, + "grad_norm": 2.3745670915106434, + "learning_rate": 1.787980916410554e-05, + "loss": 1.0681, + "step": 2441 + }, + { + "epoch": 0.2348980377068103, + "grad_norm": 2.0894771105264214, + "learning_rate": 1.787789061678151e-05, + "loss": 1.0766, + "step": 2442 + }, + { + "epoch": 0.2349942285494421, + "grad_norm": 2.105007270015778, + "learning_rate": 1.7875971304840375e-05, + "loss": 1.0713, + "step": 2443 + }, + { + "epoch": 0.23509041939207387, + "grad_norm": 2.089579945567592, + "learning_rate": 1.78740512284684e-05, + "loss": 0.9287, + "step": 2444 + }, + { + "epoch": 0.23518661023470566, + "grad_norm": 1.9908066470848957, + "learning_rate": 1.7872130387851965e-05, + "loss": 1.0334, + "step": 2445 + }, + { + "epoch": 0.23528280107733743, + "grad_norm": 2.331210173894743, + "learning_rate": 1.787020878317749e-05, + "loss": 1.0664, + "step": 2446 + }, + { + "epoch": 0.23537899191996922, + "grad_norm": 2.1746382348217668, + "learning_rate": 1.7868286414631492e-05, + "loss": 1.0667, + "step": 2447 + }, + { + "epoch": 0.235475182762601, + "grad_norm": 2.385603379813618, + "learning_rate": 1.7866363282400555e-05, + "loss": 0.867, + "step": 2448 + }, + { + "epoch": 0.23557137360523278, + "grad_norm": 2.7801279522878857, + "learning_rate": 1.7864439386671324e-05, + "loss": 1.0016, + "step": 2449 + }, + { + "epoch": 0.23566756444786457, + "grad_norm": 2.447406410013494, + "learning_rate": 1.7862514727630543e-05, + "loss": 0.996, + "step": 2450 + }, + { + "epoch": 0.23576375529049634, + "grad_norm": 2.2299861219389254, + "learning_rate": 1.7860589305465007e-05, + "loss": 1.157, + "step": 2451 + }, + { + "epoch": 0.23585994613312813, + "grad_norm": 2.3620843967786653, + "learning_rate": 1.7858663120361597e-05, + "loss": 1.0576, + "step": 2452 + }, + { + "epoch": 0.2359561369757599, + "grad_norm": 2.2557191766883236, + "learning_rate": 1.785673617250727e-05, + "loss": 1.116, + "step": 2453 + }, + { + "epoch": 0.2360523278183917, + "grad_norm": 2.321635524663293, + "learning_rate": 1.785480846208905e-05, + "loss": 1.2003, + "step": 2454 + }, + { + "epoch": 0.23614851866102346, + "grad_norm": 2.0498611873650945, + "learning_rate": 1.7852879989294037e-05, + "loss": 1.1279, + "step": 2455 + }, + { + "epoch": 0.23624470950365525, + "grad_norm": 2.274021275943611, + "learning_rate": 1.7850950754309405e-05, + "loss": 0.9469, + "step": 2456 + }, + { + "epoch": 0.23634090034628705, + "grad_norm": 2.176456990257133, + "learning_rate": 1.784902075732241e-05, + "loss": 0.9366, + "step": 2457 + }, + { + "epoch": 0.2364370911889188, + "grad_norm": 2.087609675640316, + "learning_rate": 1.7847089998520365e-05, + "loss": 0.9659, + "step": 2458 + }, + { + "epoch": 0.2365332820315506, + "grad_norm": 2.4362275428898332, + "learning_rate": 1.7845158478090673e-05, + "loss": 1.1698, + "step": 2459 + }, + { + "epoch": 0.23662947287418237, + "grad_norm": 2.226645379891131, + "learning_rate": 1.7843226196220803e-05, + "loss": 1.0651, + "step": 2460 + }, + { + "epoch": 0.23672566371681417, + "grad_norm": 2.216301717116165, + "learning_rate": 1.7841293153098297e-05, + "loss": 1.0656, + "step": 2461 + }, + { + "epoch": 0.23682185455944593, + "grad_norm": 2.3694127868452193, + "learning_rate": 1.783935934891078e-05, + "loss": 1.0226, + "step": 2462 + }, + { + "epoch": 0.23691804540207773, + "grad_norm": 2.1050775524229652, + "learning_rate": 1.783742478384594e-05, + "loss": 1.0672, + "step": 2463 + }, + { + "epoch": 0.2370142362447095, + "grad_norm": 2.132247827524899, + "learning_rate": 1.7835489458091544e-05, + "loss": 1.1723, + "step": 2464 + }, + { + "epoch": 0.23711042708734129, + "grad_norm": 2.3112391941598305, + "learning_rate": 1.783355337183543e-05, + "loss": 0.9056, + "step": 2465 + }, + { + "epoch": 0.23720661792997308, + "grad_norm": 2.255730304884364, + "learning_rate": 1.7831616525265515e-05, + "loss": 1.0994, + "step": 2466 + }, + { + "epoch": 0.23730280877260485, + "grad_norm": 2.1936176263534857, + "learning_rate": 1.7829678918569782e-05, + "loss": 1.0006, + "step": 2467 + }, + { + "epoch": 0.23739899961523664, + "grad_norm": 2.2157174956913837, + "learning_rate": 1.7827740551936296e-05, + "loss": 1.0628, + "step": 2468 + }, + { + "epoch": 0.2374951904578684, + "grad_norm": 2.361073238701512, + "learning_rate": 1.7825801425553196e-05, + "loss": 0.9871, + "step": 2469 + }, + { + "epoch": 0.2375913813005002, + "grad_norm": 2.324838085045198, + "learning_rate": 1.7823861539608686e-05, + "loss": 1.0034, + "step": 2470 + }, + { + "epoch": 0.23768757214313196, + "grad_norm": 2.331025958912172, + "learning_rate": 1.7821920894291044e-05, + "loss": 1.0447, + "step": 2471 + }, + { + "epoch": 0.23778376298576376, + "grad_norm": 2.0380685442378095, + "learning_rate": 1.7819979489788638e-05, + "loss": 1.1106, + "step": 2472 + }, + { + "epoch": 0.23787995382839552, + "grad_norm": 1.9945600713403897, + "learning_rate": 1.7818037326289887e-05, + "loss": 1.0264, + "step": 2473 + }, + { + "epoch": 0.23797614467102732, + "grad_norm": 2.0815614629155146, + "learning_rate": 1.7816094403983298e-05, + "loss": 1.0251, + "step": 2474 + }, + { + "epoch": 0.2380723355136591, + "grad_norm": 2.1834941823831424, + "learning_rate": 1.7814150723057454e-05, + "loss": 1.0967, + "step": 2475 + }, + { + "epoch": 0.23816852635629088, + "grad_norm": 2.2209517988014538, + "learning_rate": 1.7812206283701002e-05, + "loss": 1.0312, + "step": 2476 + }, + { + "epoch": 0.23826471719892267, + "grad_norm": 1.9747298919394494, + "learning_rate": 1.7810261086102666e-05, + "loss": 1.0249, + "step": 2477 + }, + { + "epoch": 0.23836090804155444, + "grad_norm": 1.9482017266732983, + "learning_rate": 1.7808315130451244e-05, + "loss": 1.0433, + "step": 2478 + }, + { + "epoch": 0.23845709888418623, + "grad_norm": 1.9383002869974193, + "learning_rate": 1.7806368416935606e-05, + "loss": 1.0394, + "step": 2479 + }, + { + "epoch": 0.238553289726818, + "grad_norm": 2.064674849121263, + "learning_rate": 1.78044209457447e-05, + "loss": 1.0764, + "step": 2480 + }, + { + "epoch": 0.2386494805694498, + "grad_norm": 2.2363072260890036, + "learning_rate": 1.7802472717067544e-05, + "loss": 1.0175, + "step": 2481 + }, + { + "epoch": 0.23874567141208156, + "grad_norm": 2.0692085115252943, + "learning_rate": 1.7800523731093232e-05, + "loss": 1.0526, + "step": 2482 + }, + { + "epoch": 0.23884186225471335, + "grad_norm": 2.447081301460264, + "learning_rate": 1.779857398801093e-05, + "loss": 1.1654, + "step": 2483 + }, + { + "epoch": 0.23893805309734514, + "grad_norm": 2.2650609770787025, + "learning_rate": 1.7796623488009875e-05, + "loss": 1.0532, + "step": 2484 + }, + { + "epoch": 0.2390342439399769, + "grad_norm": 2.5540491193397923, + "learning_rate": 1.7794672231279376e-05, + "loss": 1.0578, + "step": 2485 + }, + { + "epoch": 0.2391304347826087, + "grad_norm": 2.2949441280657648, + "learning_rate": 1.7792720218008826e-05, + "loss": 1.2055, + "step": 2486 + }, + { + "epoch": 0.23922662562524047, + "grad_norm": 2.071676378637438, + "learning_rate": 1.7790767448387687e-05, + "loss": 0.9483, + "step": 2487 + }, + { + "epoch": 0.23932281646787226, + "grad_norm": 2.341857352453635, + "learning_rate": 1.7788813922605488e-05, + "loss": 1.0737, + "step": 2488 + }, + { + "epoch": 0.23941900731050403, + "grad_norm": 2.186056172878368, + "learning_rate": 1.778685964085183e-05, + "loss": 1.1204, + "step": 2489 + }, + { + "epoch": 0.23951519815313582, + "grad_norm": 2.029742600510513, + "learning_rate": 1.7784904603316402e-05, + "loss": 1.1071, + "step": 2490 + }, + { + "epoch": 0.23961138899576762, + "grad_norm": 2.530802738923262, + "learning_rate": 1.7782948810188952e-05, + "loss": 1.1552, + "step": 2491 + }, + { + "epoch": 0.23970757983839938, + "grad_norm": 2.2374725859875837, + "learning_rate": 1.7780992261659305e-05, + "loss": 1.1051, + "step": 2492 + }, + { + "epoch": 0.23980377068103118, + "grad_norm": 2.2912739693567206, + "learning_rate": 1.777903495791737e-05, + "loss": 1.0417, + "step": 2493 + }, + { + "epoch": 0.23989996152366294, + "grad_norm": 2.1881572737931783, + "learning_rate": 1.777707689915311e-05, + "loss": 1.155, + "step": 2494 + }, + { + "epoch": 0.23999615236629473, + "grad_norm": 2.2711587142379424, + "learning_rate": 1.7775118085556577e-05, + "loss": 1.0735, + "step": 2495 + }, + { + "epoch": 0.2400923432089265, + "grad_norm": 2.370112147331206, + "learning_rate": 1.777315851731789e-05, + "loss": 1.1594, + "step": 2496 + }, + { + "epoch": 0.2401885340515583, + "grad_norm": 2.308410051299662, + "learning_rate": 1.777119819462724e-05, + "loss": 1.0717, + "step": 2497 + }, + { + "epoch": 0.24028472489419006, + "grad_norm": 2.2163454672415233, + "learning_rate": 1.7769237117674893e-05, + "loss": 0.9069, + "step": 2498 + }, + { + "epoch": 0.24038091573682185, + "grad_norm": 2.1842365679748124, + "learning_rate": 1.7767275286651193e-05, + "loss": 1.019, + "step": 2499 + }, + { + "epoch": 0.24047710657945365, + "grad_norm": 2.699967796112051, + "learning_rate": 1.7765312701746543e-05, + "loss": 1.0073, + "step": 2500 + }, + { + "epoch": 0.2405732974220854, + "grad_norm": 2.137301216527156, + "learning_rate": 1.776334936315144e-05, + "loss": 0.9475, + "step": 2501 + }, + { + "epoch": 0.2406694882647172, + "grad_norm": 2.2164558007651736, + "learning_rate": 1.7761385271056436e-05, + "loss": 1.0907, + "step": 2502 + }, + { + "epoch": 0.24076567910734897, + "grad_norm": 2.5534481608080797, + "learning_rate": 1.775942042565217e-05, + "loss": 0.951, + "step": 2503 + }, + { + "epoch": 0.24086186994998077, + "grad_norm": 2.432168086506746, + "learning_rate": 1.7757454827129338e-05, + "loss": 1.0888, + "step": 2504 + }, + { + "epoch": 0.24095806079261253, + "grad_norm": 1.9190874500698816, + "learning_rate": 1.7755488475678724e-05, + "loss": 1.0661, + "step": 2505 + }, + { + "epoch": 0.24105425163524433, + "grad_norm": 2.503918182545742, + "learning_rate": 1.7753521371491174e-05, + "loss": 1.1138, + "step": 2506 + }, + { + "epoch": 0.2411504424778761, + "grad_norm": 2.140104595148045, + "learning_rate": 1.7751553514757622e-05, + "loss": 1.0445, + "step": 2507 + }, + { + "epoch": 0.24124663332050789, + "grad_norm": 2.458258545214171, + "learning_rate": 1.7749584905669057e-05, + "loss": 1.0842, + "step": 2508 + }, + { + "epoch": 0.24134282416313968, + "grad_norm": 2.522106517461456, + "learning_rate": 1.7747615544416553e-05, + "loss": 0.9848, + "step": 2509 + }, + { + "epoch": 0.24143901500577145, + "grad_norm": 2.030703335568784, + "learning_rate": 1.774564543119125e-05, + "loss": 1.0097, + "step": 2510 + }, + { + "epoch": 0.24153520584840324, + "grad_norm": 2.291151558829214, + "learning_rate": 1.774367456618437e-05, + "loss": 1.0719, + "step": 2511 + }, + { + "epoch": 0.241631396691035, + "grad_norm": 2.312042227017663, + "learning_rate": 1.7741702949587196e-05, + "loss": 1.025, + "step": 2512 + }, + { + "epoch": 0.2417275875336668, + "grad_norm": 2.1128632810742465, + "learning_rate": 1.7739730581591098e-05, + "loss": 1.0575, + "step": 2513 + }, + { + "epoch": 0.24182377837629856, + "grad_norm": 2.442997038858689, + "learning_rate": 1.7737757462387507e-05, + "loss": 1.0757, + "step": 2514 + }, + { + "epoch": 0.24191996921893036, + "grad_norm": 1.9986554635801717, + "learning_rate": 1.773578359216793e-05, + "loss": 0.9825, + "step": 2515 + }, + { + "epoch": 0.24201616006156215, + "grad_norm": 2.28807527758672, + "learning_rate": 1.7733808971123946e-05, + "loss": 1.0527, + "step": 2516 + }, + { + "epoch": 0.24211235090419392, + "grad_norm": 2.181290132688022, + "learning_rate": 1.7731833599447218e-05, + "loss": 1.1065, + "step": 2517 + }, + { + "epoch": 0.2422085417468257, + "grad_norm": 2.5931775023686114, + "learning_rate": 1.7729857477329463e-05, + "loss": 1.0248, + "step": 2518 + }, + { + "epoch": 0.24230473258945748, + "grad_norm": 2.1593854107328596, + "learning_rate": 1.772788060496249e-05, + "loss": 1.0698, + "step": 2519 + }, + { + "epoch": 0.24240092343208927, + "grad_norm": 2.5922363580465584, + "learning_rate": 1.7725902982538162e-05, + "loss": 1.1069, + "step": 2520 + }, + { + "epoch": 0.24249711427472104, + "grad_norm": 2.5576395252142685, + "learning_rate": 1.7723924610248428e-05, + "loss": 1.1731, + "step": 2521 + }, + { + "epoch": 0.24259330511735283, + "grad_norm": 2.1728626648479787, + "learning_rate": 1.772194548828531e-05, + "loss": 1.0786, + "step": 2522 + }, + { + "epoch": 0.2426894959599846, + "grad_norm": 1.961878713858308, + "learning_rate": 1.7719965616840892e-05, + "loss": 0.8633, + "step": 2523 + }, + { + "epoch": 0.2427856868026164, + "grad_norm": 2.147914404973155, + "learning_rate": 1.7717984996107346e-05, + "loss": 0.9749, + "step": 2524 + }, + { + "epoch": 0.24288187764524818, + "grad_norm": 2.28171862608566, + "learning_rate": 1.77160036262769e-05, + "loss": 1.0399, + "step": 2525 + }, + { + "epoch": 0.24297806848787995, + "grad_norm": 2.0744565720360804, + "learning_rate": 1.771402150754187e-05, + "loss": 1.0898, + "step": 2526 + }, + { + "epoch": 0.24307425933051174, + "grad_norm": 2.2460558818677256, + "learning_rate": 1.7712038640094626e-05, + "loss": 1.133, + "step": 2527 + }, + { + "epoch": 0.2431704501731435, + "grad_norm": 2.0148287877143605, + "learning_rate": 1.7710055024127637e-05, + "loss": 1.0055, + "step": 2528 + }, + { + "epoch": 0.2432666410157753, + "grad_norm": 2.1693236502124007, + "learning_rate": 1.7708070659833423e-05, + "loss": 0.9973, + "step": 2529 + }, + { + "epoch": 0.24336283185840707, + "grad_norm": 2.2110804556654853, + "learning_rate": 1.7706085547404582e-05, + "loss": 1.1693, + "step": 2530 + }, + { + "epoch": 0.24345902270103886, + "grad_norm": 2.5697377874182, + "learning_rate": 1.7704099687033793e-05, + "loss": 0.9514, + "step": 2531 + }, + { + "epoch": 0.24355521354367063, + "grad_norm": 2.070943308279074, + "learning_rate": 1.770211307891379e-05, + "loss": 0.9998, + "step": 2532 + }, + { + "epoch": 0.24365140438630242, + "grad_norm": 2.254146332637643, + "learning_rate": 1.77001257232374e-05, + "loss": 1.1786, + "step": 2533 + }, + { + "epoch": 0.24374759522893422, + "grad_norm": 2.29349770763289, + "learning_rate": 1.769813762019751e-05, + "loss": 1.0992, + "step": 2534 + }, + { + "epoch": 0.24384378607156598, + "grad_norm": 2.055733515692627, + "learning_rate": 1.7696148769987084e-05, + "loss": 1.0709, + "step": 2535 + }, + { + "epoch": 0.24393997691419778, + "grad_norm": 2.5377901705231674, + "learning_rate": 1.769415917279915e-05, + "loss": 1.0872, + "step": 2536 + }, + { + "epoch": 0.24403616775682954, + "grad_norm": 2.0947797122879765, + "learning_rate": 1.7692168828826827e-05, + "loss": 1.0737, + "step": 2537 + }, + { + "epoch": 0.24413235859946134, + "grad_norm": 2.265895797839112, + "learning_rate": 1.7690177738263284e-05, + "loss": 1.1034, + "step": 2538 + }, + { + "epoch": 0.2442285494420931, + "grad_norm": 1.9328848722807774, + "learning_rate": 1.768818590130178e-05, + "loss": 1.0499, + "step": 2539 + }, + { + "epoch": 0.2443247402847249, + "grad_norm": 2.168112347124213, + "learning_rate": 1.7686193318135635e-05, + "loss": 1.1158, + "step": 2540 + }, + { + "epoch": 0.2444209311273567, + "grad_norm": 2.0781236887950287, + "learning_rate": 1.7684199988958254e-05, + "loss": 1.1676, + "step": 2541 + }, + { + "epoch": 0.24451712196998845, + "grad_norm": 1.8724426890968882, + "learning_rate": 1.76822059139631e-05, + "loss": 0.8928, + "step": 2542 + }, + { + "epoch": 0.24461331281262025, + "grad_norm": 2.0698690759485965, + "learning_rate": 1.768021109334372e-05, + "loss": 1.0191, + "step": 2543 + }, + { + "epoch": 0.24470950365525201, + "grad_norm": 1.8925215584541812, + "learning_rate": 1.7678215527293724e-05, + "loss": 1.0433, + "step": 2544 + }, + { + "epoch": 0.2448056944978838, + "grad_norm": 2.278553144909306, + "learning_rate": 1.76762192160068e-05, + "loss": 0.9449, + "step": 2545 + }, + { + "epoch": 0.24490188534051557, + "grad_norm": 2.007123756317103, + "learning_rate": 1.767422215967671e-05, + "loss": 1.0314, + "step": 2546 + }, + { + "epoch": 0.24499807618314737, + "grad_norm": 2.111305929000632, + "learning_rate": 1.767222435849728e-05, + "loss": 0.9059, + "step": 2547 + }, + { + "epoch": 0.24509426702577913, + "grad_norm": 2.3009165554787234, + "learning_rate": 1.767022581266242e-05, + "loss": 1.025, + "step": 2548 + }, + { + "epoch": 0.24519045786841093, + "grad_norm": 2.068271846937556, + "learning_rate": 1.7668226522366105e-05, + "loss": 1.0063, + "step": 2549 + }, + { + "epoch": 0.24528664871104272, + "grad_norm": 2.100754306243669, + "learning_rate": 1.766622648780238e-05, + "loss": 0.9909, + "step": 2550 + }, + { + "epoch": 0.2453828395536745, + "grad_norm": 2.2318048610657484, + "learning_rate": 1.7664225709165366e-05, + "loss": 1.0488, + "step": 2551 + }, + { + "epoch": 0.24547903039630628, + "grad_norm": 2.3855732497962485, + "learning_rate": 1.766222418664926e-05, + "loss": 1.1064, + "step": 2552 + }, + { + "epoch": 0.24557522123893805, + "grad_norm": 2.4258616281663716, + "learning_rate": 1.766022192044832e-05, + "loss": 1.1946, + "step": 2553 + }, + { + "epoch": 0.24567141208156984, + "grad_norm": 2.2195383412139345, + "learning_rate": 1.765821891075689e-05, + "loss": 1.0912, + "step": 2554 + }, + { + "epoch": 0.2457676029242016, + "grad_norm": 2.2593786043051542, + "learning_rate": 1.7656215157769376e-05, + "loss": 1.149, + "step": 2555 + }, + { + "epoch": 0.2458637937668334, + "grad_norm": 1.9904066536939307, + "learning_rate": 1.7654210661680263e-05, + "loss": 1.081, + "step": 2556 + }, + { + "epoch": 0.24595998460946517, + "grad_norm": 2.2535838402413804, + "learning_rate": 1.7652205422684098e-05, + "loss": 1.1248, + "step": 2557 + }, + { + "epoch": 0.24605617545209696, + "grad_norm": 1.7135814538161913, + "learning_rate": 1.765019944097551e-05, + "loss": 0.9223, + "step": 2558 + }, + { + "epoch": 0.24615236629472875, + "grad_norm": 2.4238342744661288, + "learning_rate": 1.76481927167492e-05, + "loss": 1.0967, + "step": 2559 + }, + { + "epoch": 0.24624855713736052, + "grad_norm": 1.9524705673332632, + "learning_rate": 1.7646185250199936e-05, + "loss": 1.0772, + "step": 2560 + }, + { + "epoch": 0.2463447479799923, + "grad_norm": 2.4168333622156672, + "learning_rate": 1.7644177041522555e-05, + "loss": 0.9219, + "step": 2561 + }, + { + "epoch": 0.24644093882262408, + "grad_norm": 2.069804050373068, + "learning_rate": 1.7642168090911976e-05, + "loss": 1.0612, + "step": 2562 + }, + { + "epoch": 0.24653712966525587, + "grad_norm": 2.62948623657237, + "learning_rate": 1.7640158398563184e-05, + "loss": 1.1309, + "step": 2563 + }, + { + "epoch": 0.24663332050788764, + "grad_norm": 2.017705882295122, + "learning_rate": 1.763814796467124e-05, + "loss": 1.0851, + "step": 2564 + }, + { + "epoch": 0.24672951135051943, + "grad_norm": 2.367743639543081, + "learning_rate": 1.7636136789431265e-05, + "loss": 1.0164, + "step": 2565 + }, + { + "epoch": 0.24682570219315123, + "grad_norm": 2.411591190189917, + "learning_rate": 1.763412487303847e-05, + "loss": 1.0586, + "step": 2566 + }, + { + "epoch": 0.246921893035783, + "grad_norm": 2.0581164969514916, + "learning_rate": 1.7632112215688127e-05, + "loss": 0.9158, + "step": 2567 + }, + { + "epoch": 0.24701808387841478, + "grad_norm": 2.294757219402664, + "learning_rate": 1.7630098817575578e-05, + "loss": 1.0197, + "step": 2568 + }, + { + "epoch": 0.24711427472104655, + "grad_norm": 2.395893710205248, + "learning_rate": 1.762808467889624e-05, + "loss": 1.1617, + "step": 2569 + }, + { + "epoch": 0.24721046556367834, + "grad_norm": 2.0706207287886573, + "learning_rate": 1.762606979984561e-05, + "loss": 0.9716, + "step": 2570 + }, + { + "epoch": 0.2473066564063101, + "grad_norm": 2.1349439573915117, + "learning_rate": 1.7624054180619246e-05, + "loss": 1.0796, + "step": 2571 + }, + { + "epoch": 0.2474028472489419, + "grad_norm": 2.0732038603137273, + "learning_rate": 1.7622037821412775e-05, + "loss": 1.0246, + "step": 2572 + }, + { + "epoch": 0.24749903809157367, + "grad_norm": 2.2505457691396016, + "learning_rate": 1.7620020722421907e-05, + "loss": 1.017, + "step": 2573 + }, + { + "epoch": 0.24759522893420546, + "grad_norm": 2.192848233422174, + "learning_rate": 1.7618002883842426e-05, + "loss": 1.0352, + "step": 2574 + }, + { + "epoch": 0.24769141977683726, + "grad_norm": 2.301829012554004, + "learning_rate": 1.7615984305870165e-05, + "loss": 1.0723, + "step": 2575 + }, + { + "epoch": 0.24778761061946902, + "grad_norm": 2.2711714886596344, + "learning_rate": 1.7613964988701057e-05, + "loss": 1.054, + "step": 2576 + }, + { + "epoch": 0.24788380146210082, + "grad_norm": 2.3534110700834776, + "learning_rate": 1.7611944932531088e-05, + "loss": 1.1225, + "step": 2577 + }, + { + "epoch": 0.24797999230473258, + "grad_norm": 2.2218877268001265, + "learning_rate": 1.7609924137556326e-05, + "loss": 1.0887, + "step": 2578 + }, + { + "epoch": 0.24807618314736438, + "grad_norm": 2.5784781535556878, + "learning_rate": 1.76079026039729e-05, + "loss": 1.1578, + "step": 2579 + }, + { + "epoch": 0.24817237398999614, + "grad_norm": 2.5445995120425926, + "learning_rate": 1.7605880331977022e-05, + "loss": 1.1113, + "step": 2580 + }, + { + "epoch": 0.24826856483262794, + "grad_norm": 2.0592908635281586, + "learning_rate": 1.7603857321764972e-05, + "loss": 1.0167, + "step": 2581 + }, + { + "epoch": 0.2483647556752597, + "grad_norm": 2.3020687587850084, + "learning_rate": 1.76018335735331e-05, + "loss": 0.9933, + "step": 2582 + }, + { + "epoch": 0.2484609465178915, + "grad_norm": 2.0867282923457693, + "learning_rate": 1.7599809087477827e-05, + "loss": 1.1213, + "step": 2583 + }, + { + "epoch": 0.2485571373605233, + "grad_norm": 2.2922143055470783, + "learning_rate": 1.7597783863795644e-05, + "loss": 0.9441, + "step": 2584 + }, + { + "epoch": 0.24865332820315506, + "grad_norm": 1.7202908105591501, + "learning_rate": 1.759575790268312e-05, + "loss": 0.8547, + "step": 2585 + }, + { + "epoch": 0.24874951904578685, + "grad_norm": 2.0222329845434053, + "learning_rate": 1.7593731204336895e-05, + "loss": 0.9874, + "step": 2586 + }, + { + "epoch": 0.24884570988841861, + "grad_norm": 2.3761807971662154, + "learning_rate": 1.759170376895367e-05, + "loss": 1.0506, + "step": 2587 + }, + { + "epoch": 0.2489419007310504, + "grad_norm": 2.3495689138842573, + "learning_rate": 1.7589675596730233e-05, + "loss": 0.9655, + "step": 2588 + }, + { + "epoch": 0.24903809157368217, + "grad_norm": 1.9777326116499354, + "learning_rate": 1.7587646687863425e-05, + "loss": 1.1518, + "step": 2589 + }, + { + "epoch": 0.24913428241631397, + "grad_norm": 2.360304053262648, + "learning_rate": 1.758561704255018e-05, + "loss": 1.1509, + "step": 2590 + }, + { + "epoch": 0.24923047325894576, + "grad_norm": 2.7963680854097634, + "learning_rate": 1.7583586660987487e-05, + "loss": 1.0322, + "step": 2591 + }, + { + "epoch": 0.24932666410157753, + "grad_norm": 2.4789904104927567, + "learning_rate": 1.7581555543372413e-05, + "loss": 1.1276, + "step": 2592 + }, + { + "epoch": 0.24942285494420932, + "grad_norm": 2.2997981369642013, + "learning_rate": 1.7579523689902098e-05, + "loss": 1.1579, + "step": 2593 + }, + { + "epoch": 0.2495190457868411, + "grad_norm": 2.1379141688572667, + "learning_rate": 1.7577491100773744e-05, + "loss": 1.0794, + "step": 2594 + }, + { + "epoch": 0.24961523662947288, + "grad_norm": 2.449748143040644, + "learning_rate": 1.7575457776184635e-05, + "loss": 1.038, + "step": 2595 + }, + { + "epoch": 0.24971142747210465, + "grad_norm": 2.441789766100565, + "learning_rate": 1.7573423716332128e-05, + "loss": 1.0204, + "step": 2596 + }, + { + "epoch": 0.24980761831473644, + "grad_norm": 2.1077708848631787, + "learning_rate": 1.7571388921413637e-05, + "loss": 1.0602, + "step": 2597 + }, + { + "epoch": 0.2499038091573682, + "grad_norm": 2.1278247275503195, + "learning_rate": 1.7569353391626665e-05, + "loss": 1.0796, + "step": 2598 + }, + { + "epoch": 0.25, + "grad_norm": 2.058514554822889, + "learning_rate": 1.756731712716877e-05, + "loss": 1.0996, + "step": 2599 + }, + { + "epoch": 0.2500961908426318, + "grad_norm": 2.280149484568255, + "learning_rate": 1.7565280128237595e-05, + "loss": 1.0219, + "step": 2600 + }, + { + "epoch": 0.2501923816852636, + "grad_norm": 2.3932208926492278, + "learning_rate": 1.7563242395030843e-05, + "loss": 1.1169, + "step": 2601 + }, + { + "epoch": 0.2502885725278953, + "grad_norm": 2.2832172599330947, + "learning_rate": 1.75612039277463e-05, + "loss": 1.0661, + "step": 2602 + }, + { + "epoch": 0.2503847633705271, + "grad_norm": 2.0886525740042043, + "learning_rate": 1.755916472658181e-05, + "loss": 1.051, + "step": 2603 + }, + { + "epoch": 0.2504809542131589, + "grad_norm": 2.455718128828126, + "learning_rate": 1.75571247917353e-05, + "loss": 0.8908, + "step": 2604 + }, + { + "epoch": 0.2505771450557907, + "grad_norm": 2.2218588912139143, + "learning_rate": 1.755508412340476e-05, + "loss": 0.9476, + "step": 2605 + }, + { + "epoch": 0.25067333589842244, + "grad_norm": 2.2140926124466294, + "learning_rate": 1.7553042721788255e-05, + "loss": 1.0135, + "step": 2606 + }, + { + "epoch": 0.25076952674105424, + "grad_norm": 2.1875597257794284, + "learning_rate": 1.755100058708392e-05, + "loss": 1.133, + "step": 2607 + }, + { + "epoch": 0.25086571758368603, + "grad_norm": 2.3652638564329806, + "learning_rate": 1.754895771948997e-05, + "loss": 1.172, + "step": 2608 + }, + { + "epoch": 0.2509619084263178, + "grad_norm": 2.452598213411314, + "learning_rate": 1.754691411920467e-05, + "loss": 1.0412, + "step": 2609 + }, + { + "epoch": 0.2510580992689496, + "grad_norm": 2.3488930006664406, + "learning_rate": 1.754486978642637e-05, + "loss": 1.1317, + "step": 2610 + }, + { + "epoch": 0.25115429011158136, + "grad_norm": 2.191415421815261, + "learning_rate": 1.7542824721353505e-05, + "loss": 1.1393, + "step": 2611 + }, + { + "epoch": 0.25125048095421315, + "grad_norm": 2.1867550808585703, + "learning_rate": 1.7540778924184553e-05, + "loss": 1.0833, + "step": 2612 + }, + { + "epoch": 0.25134667179684494, + "grad_norm": 2.402694784181303, + "learning_rate": 1.7538732395118077e-05, + "loss": 1.0046, + "step": 2613 + }, + { + "epoch": 0.25144286263947674, + "grad_norm": 2.136867513472493, + "learning_rate": 1.7536685134352717e-05, + "loss": 0.9753, + "step": 2614 + }, + { + "epoch": 0.2515390534821085, + "grad_norm": 2.959013869086275, + "learning_rate": 1.7534637142087172e-05, + "loss": 0.962, + "step": 2615 + }, + { + "epoch": 0.25163524432474027, + "grad_norm": 2.645278883674903, + "learning_rate": 1.7532588418520215e-05, + "loss": 1.0904, + "step": 2616 + }, + { + "epoch": 0.25173143516737206, + "grad_norm": 2.2953918217439315, + "learning_rate": 1.7530538963850698e-05, + "loss": 1.0796, + "step": 2617 + }, + { + "epoch": 0.25182762601000386, + "grad_norm": 2.0914974126875863, + "learning_rate": 1.7528488778277535e-05, + "loss": 0.9625, + "step": 2618 + }, + { + "epoch": 0.25192381685263565, + "grad_norm": 2.0518690418760768, + "learning_rate": 1.7526437861999718e-05, + "loss": 1.0816, + "step": 2619 + }, + { + "epoch": 0.2520200076952674, + "grad_norm": 2.107476006608443, + "learning_rate": 1.75243862152163e-05, + "loss": 1.1057, + "step": 2620 + }, + { + "epoch": 0.2521161985378992, + "grad_norm": 2.219018295889806, + "learning_rate": 1.7522333838126413e-05, + "loss": 1.1049, + "step": 2621 + }, + { + "epoch": 0.252212389380531, + "grad_norm": 2.3478347055971254, + "learning_rate": 1.752028073092926e-05, + "loss": 0.9364, + "step": 2622 + }, + { + "epoch": 0.25230858022316277, + "grad_norm": 2.199843529158002, + "learning_rate": 1.7518226893824113e-05, + "loss": 1.1054, + "step": 2623 + }, + { + "epoch": 0.2524047710657945, + "grad_norm": 2.410912055740433, + "learning_rate": 1.7516172327010314e-05, + "loss": 1.1179, + "step": 2624 + }, + { + "epoch": 0.2525009619084263, + "grad_norm": 2.749723742903441, + "learning_rate": 1.7514117030687274e-05, + "loss": 1.1075, + "step": 2625 + }, + { + "epoch": 0.2525971527510581, + "grad_norm": 2.1637305270439455, + "learning_rate": 1.751206100505448e-05, + "loss": 1.0584, + "step": 2626 + }, + { + "epoch": 0.2526933435936899, + "grad_norm": 2.278179724274798, + "learning_rate": 1.751000425031148e-05, + "loss": 1.1345, + "step": 2627 + }, + { + "epoch": 0.2527895344363217, + "grad_norm": 2.1114784188380114, + "learning_rate": 1.7507946766657914e-05, + "loss": 1.0858, + "step": 2628 + }, + { + "epoch": 0.2528857252789534, + "grad_norm": 1.9970167906193415, + "learning_rate": 1.7505888554293467e-05, + "loss": 0.9627, + "step": 2629 + }, + { + "epoch": 0.2529819161215852, + "grad_norm": 2.0089680206784593, + "learning_rate": 1.7503829613417905e-05, + "loss": 1.0537, + "step": 2630 + }, + { + "epoch": 0.253078106964217, + "grad_norm": 2.5004553176371336, + "learning_rate": 1.7501769944231077e-05, + "loss": 0.9751, + "step": 2631 + }, + { + "epoch": 0.2531742978068488, + "grad_norm": 2.4722128618821118, + "learning_rate": 1.749970954693288e-05, + "loss": 1.1574, + "step": 2632 + }, + { + "epoch": 0.2532704886494806, + "grad_norm": 2.5299932850216313, + "learning_rate": 1.7497648421723303e-05, + "loss": 1.0116, + "step": 2633 + }, + { + "epoch": 0.25336667949211233, + "grad_norm": 2.2077061047256277, + "learning_rate": 1.7495586568802384e-05, + "loss": 1.0952, + "step": 2634 + }, + { + "epoch": 0.25346287033474413, + "grad_norm": 2.395920803277921, + "learning_rate": 1.7493523988370255e-05, + "loss": 1.171, + "step": 2635 + }, + { + "epoch": 0.2535590611773759, + "grad_norm": 2.2798846243680906, + "learning_rate": 1.7491460680627105e-05, + "loss": 1.0876, + "step": 2636 + }, + { + "epoch": 0.2536552520200077, + "grad_norm": 2.2316666903594107, + "learning_rate": 1.748939664577319e-05, + "loss": 1.1527, + "step": 2637 + }, + { + "epoch": 0.25375144286263945, + "grad_norm": 2.2985120749249726, + "learning_rate": 1.7487331884008845e-05, + "loss": 1.005, + "step": 2638 + }, + { + "epoch": 0.25384763370527125, + "grad_norm": 2.155887896977632, + "learning_rate": 1.7485266395534476e-05, + "loss": 1.0618, + "step": 2639 + }, + { + "epoch": 0.25394382454790304, + "grad_norm": 2.7801245999167623, + "learning_rate": 1.7483200180550554e-05, + "loss": 1.1192, + "step": 2640 + }, + { + "epoch": 0.25404001539053483, + "grad_norm": 2.2059055623867767, + "learning_rate": 1.7481133239257627e-05, + "loss": 1.0422, + "step": 2641 + }, + { + "epoch": 0.25413620623316663, + "grad_norm": 2.115847772635462, + "learning_rate": 1.74790655718563e-05, + "loss": 1.0797, + "step": 2642 + }, + { + "epoch": 0.25423239707579837, + "grad_norm": 2.184296745184354, + "learning_rate": 1.7476997178547268e-05, + "loss": 1.0501, + "step": 2643 + }, + { + "epoch": 0.25432858791843016, + "grad_norm": 2.2089084979028892, + "learning_rate": 1.747492805953128e-05, + "loss": 1.1034, + "step": 2644 + }, + { + "epoch": 0.25442477876106195, + "grad_norm": 2.0034546618919538, + "learning_rate": 1.7472858215009165e-05, + "loss": 1.0592, + "step": 2645 + }, + { + "epoch": 0.25452096960369375, + "grad_norm": 2.00934318011009, + "learning_rate": 1.7470787645181818e-05, + "loss": 1.0189, + "step": 2646 + }, + { + "epoch": 0.2546171604463255, + "grad_norm": 2.4567840417396978, + "learning_rate": 1.7468716350250202e-05, + "loss": 1.1532, + "step": 2647 + }, + { + "epoch": 0.2547133512889573, + "grad_norm": 2.1490899699879633, + "learning_rate": 1.7466644330415362e-05, + "loss": 1.0432, + "step": 2648 + }, + { + "epoch": 0.2548095421315891, + "grad_norm": 2.465944657970076, + "learning_rate": 1.74645715858784e-05, + "loss": 0.9988, + "step": 2649 + }, + { + "epoch": 0.25490573297422087, + "grad_norm": 2.056020328098462, + "learning_rate": 1.7462498116840496e-05, + "loss": 0.9421, + "step": 2650 + }, + { + "epoch": 0.25500192381685266, + "grad_norm": 2.1488088165559724, + "learning_rate": 1.7460423923502895e-05, + "loss": 0.9892, + "step": 2651 + }, + { + "epoch": 0.2550981146594844, + "grad_norm": 2.0493804478475837, + "learning_rate": 1.745834900606692e-05, + "loss": 1.008, + "step": 2652 + }, + { + "epoch": 0.2551943055021162, + "grad_norm": 2.08886291812587, + "learning_rate": 1.7456273364733953e-05, + "loss": 1.0334, + "step": 2653 + }, + { + "epoch": 0.255290496344748, + "grad_norm": 2.0290148982917855, + "learning_rate": 1.7454196999705458e-05, + "loss": 1.0565, + "step": 2654 + }, + { + "epoch": 0.2553866871873798, + "grad_norm": 1.959761156701766, + "learning_rate": 1.7452119911182964e-05, + "loss": 1.1088, + "step": 2655 + }, + { + "epoch": 0.2554828780300115, + "grad_norm": 2.04251984949894, + "learning_rate": 1.7450042099368066e-05, + "loss": 0.9871, + "step": 2656 + }, + { + "epoch": 0.2555790688726433, + "grad_norm": 2.0752859418974903, + "learning_rate": 1.744796356446244e-05, + "loss": 1.001, + "step": 2657 + }, + { + "epoch": 0.2556752597152751, + "grad_norm": 2.2669116988408127, + "learning_rate": 1.7445884306667823e-05, + "loss": 0.9718, + "step": 2658 + }, + { + "epoch": 0.2557714505579069, + "grad_norm": 1.9194178960574182, + "learning_rate": 1.744380432618602e-05, + "loss": 1.0564, + "step": 2659 + }, + { + "epoch": 0.2558676414005387, + "grad_norm": 2.1900363671422918, + "learning_rate": 1.7441723623218917e-05, + "loss": 0.9708, + "step": 2660 + }, + { + "epoch": 0.25596383224317043, + "grad_norm": 2.066624935680156, + "learning_rate": 1.7439642197968462e-05, + "loss": 1.0579, + "step": 2661 + }, + { + "epoch": 0.2560600230858022, + "grad_norm": 1.8974563860938305, + "learning_rate": 1.7437560050636678e-05, + "loss": 0.8052, + "step": 2662 + }, + { + "epoch": 0.256156213928434, + "grad_norm": 2.3012116604263055, + "learning_rate": 1.743547718142565e-05, + "loss": 1.1795, + "step": 2663 + }, + { + "epoch": 0.2562524047710658, + "grad_norm": 1.793766834339894, + "learning_rate": 1.7433393590537543e-05, + "loss": 0.828, + "step": 2664 + }, + { + "epoch": 0.25634859561369755, + "grad_norm": 1.9050149001746466, + "learning_rate": 1.7431309278174583e-05, + "loss": 0.9229, + "step": 2665 + }, + { + "epoch": 0.25644478645632934, + "grad_norm": 2.149082924718563, + "learning_rate": 1.7429224244539077e-05, + "loss": 1.0634, + "step": 2666 + }, + { + "epoch": 0.25654097729896114, + "grad_norm": 2.1263888331985354, + "learning_rate": 1.7427138489833392e-05, + "loss": 1.092, + "step": 2667 + }, + { + "epoch": 0.25663716814159293, + "grad_norm": 2.1218101691312237, + "learning_rate": 1.7425052014259965e-05, + "loss": 1.0319, + "step": 2668 + }, + { + "epoch": 0.2567333589842247, + "grad_norm": 2.3188482685501426, + "learning_rate": 1.7422964818021307e-05, + "loss": 1.0552, + "step": 2669 + }, + { + "epoch": 0.25682954982685646, + "grad_norm": 1.9487660627899797, + "learning_rate": 1.7420876901320006e-05, + "loss": 0.9324, + "step": 2670 + }, + { + "epoch": 0.25692574066948826, + "grad_norm": 2.1241593155264344, + "learning_rate": 1.741878826435871e-05, + "loss": 0.825, + "step": 2671 + }, + { + "epoch": 0.25702193151212005, + "grad_norm": 2.1072915976765314, + "learning_rate": 1.7416698907340128e-05, + "loss": 1.0885, + "step": 2672 + }, + { + "epoch": 0.25711812235475184, + "grad_norm": 1.94830969938343, + "learning_rate": 1.7414608830467063e-05, + "loss": 1.0706, + "step": 2673 + }, + { + "epoch": 0.2572143131973836, + "grad_norm": 1.8983055312492192, + "learning_rate": 1.741251803394237e-05, + "loss": 1.0523, + "step": 2674 + }, + { + "epoch": 0.2573105040400154, + "grad_norm": 2.2170210349773867, + "learning_rate": 1.741042651796898e-05, + "loss": 1.096, + "step": 2675 + }, + { + "epoch": 0.25740669488264717, + "grad_norm": 2.1640215220232193, + "learning_rate": 1.740833428274989e-05, + "loss": 0.9709, + "step": 2676 + }, + { + "epoch": 0.25750288572527896, + "grad_norm": 2.277224731876814, + "learning_rate": 1.7406241328488174e-05, + "loss": 0.9597, + "step": 2677 + }, + { + "epoch": 0.25759907656791076, + "grad_norm": 2.2444429243077533, + "learning_rate": 1.7404147655386966e-05, + "loss": 1.1107, + "step": 2678 + }, + { + "epoch": 0.2576952674105425, + "grad_norm": 2.1225577712810586, + "learning_rate": 1.740205326364948e-05, + "loss": 1.0776, + "step": 2679 + }, + { + "epoch": 0.2577914582531743, + "grad_norm": 2.131218168430351, + "learning_rate": 1.739995815347899e-05, + "loss": 0.9922, + "step": 2680 + }, + { + "epoch": 0.2578876490958061, + "grad_norm": 2.2298448291790804, + "learning_rate": 1.7397862325078846e-05, + "loss": 1.1446, + "step": 2681 + }, + { + "epoch": 0.2579838399384379, + "grad_norm": 2.5107534994644083, + "learning_rate": 1.739576577865247e-05, + "loss": 1.1279, + "step": 2682 + }, + { + "epoch": 0.25808003078106967, + "grad_norm": 2.1630314916246243, + "learning_rate": 1.7393668514403344e-05, + "loss": 1.0295, + "step": 2683 + }, + { + "epoch": 0.2581762216237014, + "grad_norm": 2.3277775672493, + "learning_rate": 1.739157053253503e-05, + "loss": 1.074, + "step": 2684 + }, + { + "epoch": 0.2582724124663332, + "grad_norm": 1.8841741207836455, + "learning_rate": 1.7389471833251153e-05, + "loss": 0.9507, + "step": 2685 + }, + { + "epoch": 0.258368603308965, + "grad_norm": 1.8946467713848976, + "learning_rate": 1.738737241675541e-05, + "loss": 1.0257, + "step": 2686 + }, + { + "epoch": 0.2584647941515968, + "grad_norm": 1.7789802497059377, + "learning_rate": 1.738527228325157e-05, + "loss": 1.034, + "step": 2687 + }, + { + "epoch": 0.2585609849942285, + "grad_norm": 2.4372216892323553, + "learning_rate": 1.7383171432943466e-05, + "loss": 0.8433, + "step": 2688 + }, + { + "epoch": 0.2586571758368603, + "grad_norm": 2.141941961687257, + "learning_rate": 1.7381069866035004e-05, + "loss": 1.1001, + "step": 2689 + }, + { + "epoch": 0.2587533666794921, + "grad_norm": 1.9019157726681468, + "learning_rate": 1.737896758273016e-05, + "loss": 0.9987, + "step": 2690 + }, + { + "epoch": 0.2588495575221239, + "grad_norm": 2.231924992511915, + "learning_rate": 1.7376864583232977e-05, + "loss": 1.0081, + "step": 2691 + }, + { + "epoch": 0.2589457483647557, + "grad_norm": 2.1062218666363086, + "learning_rate": 1.7374760867747574e-05, + "loss": 1.0037, + "step": 2692 + }, + { + "epoch": 0.25904193920738744, + "grad_norm": 2.1542410969778545, + "learning_rate": 1.7372656436478128e-05, + "loss": 0.9744, + "step": 2693 + }, + { + "epoch": 0.25913813005001923, + "grad_norm": 2.3041834397460357, + "learning_rate": 1.7370551289628895e-05, + "loss": 1.0645, + "step": 2694 + }, + { + "epoch": 0.259234320892651, + "grad_norm": 2.3666158089637928, + "learning_rate": 1.73684454274042e-05, + "loss": 1.0222, + "step": 2695 + }, + { + "epoch": 0.2593305117352828, + "grad_norm": 2.1441690923438346, + "learning_rate": 1.7366338850008432e-05, + "loss": 1.0494, + "step": 2696 + }, + { + "epoch": 0.25942670257791456, + "grad_norm": 1.8738990612155386, + "learning_rate": 1.7364231557646055e-05, + "loss": 0.9263, + "step": 2697 + }, + { + "epoch": 0.25952289342054635, + "grad_norm": 2.506516153340056, + "learning_rate": 1.73621235505216e-05, + "loss": 0.8583, + "step": 2698 + }, + { + "epoch": 0.25961908426317815, + "grad_norm": 1.8942020559775923, + "learning_rate": 1.7360014828839664e-05, + "loss": 0.9846, + "step": 2699 + }, + { + "epoch": 0.25971527510580994, + "grad_norm": 2.18010287413935, + "learning_rate": 1.7357905392804918e-05, + "loss": 0.9825, + "step": 2700 + }, + { + "epoch": 0.25981146594844173, + "grad_norm": 2.4693397594128754, + "learning_rate": 1.73557952426221e-05, + "loss": 1.0757, + "step": 2701 + }, + { + "epoch": 0.25990765679107347, + "grad_norm": 2.044999146852894, + "learning_rate": 1.735368437849602e-05, + "loss": 0.9777, + "step": 2702 + }, + { + "epoch": 0.26000384763370527, + "grad_norm": 2.285063691625122, + "learning_rate": 1.7351572800631556e-05, + "loss": 1.1348, + "step": 2703 + }, + { + "epoch": 0.26010003847633706, + "grad_norm": 2.0538230389772107, + "learning_rate": 1.7349460509233654e-05, + "loss": 1.0776, + "step": 2704 + }, + { + "epoch": 0.26019622931896885, + "grad_norm": 2.3226792704699744, + "learning_rate": 1.734734750450733e-05, + "loss": 1.0274, + "step": 2705 + }, + { + "epoch": 0.2602924201616006, + "grad_norm": 2.2930191302762197, + "learning_rate": 1.734523378665767e-05, + "loss": 1.1158, + "step": 2706 + }, + { + "epoch": 0.2603886110042324, + "grad_norm": 2.1801526790068033, + "learning_rate": 1.7343119355889826e-05, + "loss": 1.0013, + "step": 2707 + }, + { + "epoch": 0.2604848018468642, + "grad_norm": 2.214015165862299, + "learning_rate": 1.7341004212409026e-05, + "loss": 1.1064, + "step": 2708 + }, + { + "epoch": 0.26058099268949597, + "grad_norm": 2.0069446186322706, + "learning_rate": 1.7338888356420556e-05, + "loss": 1.0185, + "step": 2709 + }, + { + "epoch": 0.26067718353212777, + "grad_norm": 1.9502625747741476, + "learning_rate": 1.7336771788129785e-05, + "loss": 1.105, + "step": 2710 + }, + { + "epoch": 0.2607733743747595, + "grad_norm": 1.9697813414051208, + "learning_rate": 1.7334654507742143e-05, + "loss": 0.8096, + "step": 2711 + }, + { + "epoch": 0.2608695652173913, + "grad_norm": 2.1715732009256867, + "learning_rate": 1.7332536515463126e-05, + "loss": 1.0868, + "step": 2712 + }, + { + "epoch": 0.2609657560600231, + "grad_norm": 2.139117200075562, + "learning_rate": 1.7330417811498308e-05, + "loss": 1.1178, + "step": 2713 + }, + { + "epoch": 0.2610619469026549, + "grad_norm": 2.2511610779199382, + "learning_rate": 1.7328298396053324e-05, + "loss": 0.9513, + "step": 2714 + }, + { + "epoch": 0.2611581377452866, + "grad_norm": 2.42326491451699, + "learning_rate": 1.7326178269333885e-05, + "loss": 0.9852, + "step": 2715 + }, + { + "epoch": 0.2612543285879184, + "grad_norm": 2.1897451842869065, + "learning_rate": 1.7324057431545768e-05, + "loss": 1.07, + "step": 2716 + }, + { + "epoch": 0.2613505194305502, + "grad_norm": 2.150217362313851, + "learning_rate": 1.732193588289481e-05, + "loss": 0.9621, + "step": 2717 + }, + { + "epoch": 0.261446710273182, + "grad_norm": 2.440380137740623, + "learning_rate": 1.7319813623586935e-05, + "loss": 1.1079, + "step": 2718 + }, + { + "epoch": 0.2615429011158138, + "grad_norm": 2.2899804664524592, + "learning_rate": 1.7317690653828125e-05, + "loss": 1.008, + "step": 2719 + }, + { + "epoch": 0.26163909195844554, + "grad_norm": 2.1572477818621705, + "learning_rate": 1.7315566973824433e-05, + "loss": 1.0632, + "step": 2720 + }, + { + "epoch": 0.26173528280107733, + "grad_norm": 2.251625709186668, + "learning_rate": 1.7313442583781972e-05, + "loss": 1.0923, + "step": 2721 + }, + { + "epoch": 0.2618314736437091, + "grad_norm": 2.516100634140713, + "learning_rate": 1.7311317483906946e-05, + "loss": 1.125, + "step": 2722 + }, + { + "epoch": 0.2619276644863409, + "grad_norm": 2.1296703431883968, + "learning_rate": 1.7309191674405602e-05, + "loss": 1.0581, + "step": 2723 + }, + { + "epoch": 0.26202385532897265, + "grad_norm": 2.3943600750004186, + "learning_rate": 1.730706515548427e-05, + "loss": 1.0518, + "step": 2724 + }, + { + "epoch": 0.26212004617160445, + "grad_norm": 2.174340927332054, + "learning_rate": 1.7304937927349356e-05, + "loss": 1.0705, + "step": 2725 + }, + { + "epoch": 0.26221623701423624, + "grad_norm": 2.2402211479544962, + "learning_rate": 1.730280999020732e-05, + "loss": 1.0092, + "step": 2726 + }, + { + "epoch": 0.26231242785686804, + "grad_norm": 2.06287129401727, + "learning_rate": 1.7300681344264693e-05, + "loss": 1.1183, + "step": 2727 + }, + { + "epoch": 0.26240861869949983, + "grad_norm": 2.108544026619804, + "learning_rate": 1.729855198972808e-05, + "loss": 1.1198, + "step": 2728 + }, + { + "epoch": 0.26250480954213157, + "grad_norm": 2.2536393975667877, + "learning_rate": 1.7296421926804162e-05, + "loss": 1.0717, + "step": 2729 + }, + { + "epoch": 0.26260100038476336, + "grad_norm": 2.0935209978820937, + "learning_rate": 1.729429115569967e-05, + "loss": 0.8602, + "step": 2730 + }, + { + "epoch": 0.26269719122739515, + "grad_norm": 2.2071577213075573, + "learning_rate": 1.7292159676621416e-05, + "loss": 1.0994, + "step": 2731 + }, + { + "epoch": 0.26279338207002695, + "grad_norm": 2.25978829525274, + "learning_rate": 1.729002748977628e-05, + "loss": 0.9976, + "step": 2732 + }, + { + "epoch": 0.26288957291265874, + "grad_norm": 1.9952136003904657, + "learning_rate": 1.728789459537121e-05, + "loss": 1.1129, + "step": 2733 + }, + { + "epoch": 0.2629857637552905, + "grad_norm": 2.42754688726308, + "learning_rate": 1.7285760993613215e-05, + "loss": 0.9698, + "step": 2734 + }, + { + "epoch": 0.2630819545979223, + "grad_norm": 2.268415444148457, + "learning_rate": 1.7283626684709386e-05, + "loss": 1.1509, + "step": 2735 + }, + { + "epoch": 0.26317814544055407, + "grad_norm": 2.2626536837962323, + "learning_rate": 1.7281491668866874e-05, + "loss": 1.0793, + "step": 2736 + }, + { + "epoch": 0.26327433628318586, + "grad_norm": 1.9944402071185274, + "learning_rate": 1.7279355946292902e-05, + "loss": 1.0419, + "step": 2737 + }, + { + "epoch": 0.2633705271258176, + "grad_norm": 1.9856146675708448, + "learning_rate": 1.727721951719476e-05, + "loss": 1.0809, + "step": 2738 + }, + { + "epoch": 0.2634667179684494, + "grad_norm": 2.3601910914546904, + "learning_rate": 1.727508238177981e-05, + "loss": 1.1036, + "step": 2739 + }, + { + "epoch": 0.2635629088110812, + "grad_norm": 2.2092805430869484, + "learning_rate": 1.7272944540255468e-05, + "loss": 1.1073, + "step": 2740 + }, + { + "epoch": 0.263659099653713, + "grad_norm": 2.4489135041380057, + "learning_rate": 1.7270805992829243e-05, + "loss": 1.0776, + "step": 2741 + }, + { + "epoch": 0.2637552904963448, + "grad_norm": 2.0707938012192697, + "learning_rate": 1.726866673970869e-05, + "loss": 1.0118, + "step": 2742 + }, + { + "epoch": 0.2638514813389765, + "grad_norm": 2.027949041334135, + "learning_rate": 1.726652678110145e-05, + "loss": 1.0999, + "step": 2743 + }, + { + "epoch": 0.2639476721816083, + "grad_norm": 1.887111159975123, + "learning_rate": 1.7264386117215216e-05, + "loss": 1.0597, + "step": 2744 + }, + { + "epoch": 0.2640438630242401, + "grad_norm": 2.481142046736733, + "learning_rate": 1.7262244748257765e-05, + "loss": 0.9983, + "step": 2745 + }, + { + "epoch": 0.2641400538668719, + "grad_norm": 1.9738846909380114, + "learning_rate": 1.7260102674436933e-05, + "loss": 0.9461, + "step": 2746 + }, + { + "epoch": 0.26423624470950363, + "grad_norm": 1.904690962762321, + "learning_rate": 1.7257959895960624e-05, + "loss": 0.9095, + "step": 2747 + }, + { + "epoch": 0.2643324355521354, + "grad_norm": 2.1482286152035193, + "learning_rate": 1.7255816413036818e-05, + "loss": 1.1772, + "step": 2748 + }, + { + "epoch": 0.2644286263947672, + "grad_norm": 2.3119380656343766, + "learning_rate": 1.7253672225873554e-05, + "loss": 1.0127, + "step": 2749 + }, + { + "epoch": 0.264524817237399, + "grad_norm": 2.285671130804495, + "learning_rate": 1.7251527334678946e-05, + "loss": 0.9064, + "step": 2750 + }, + { + "epoch": 0.2646210080800308, + "grad_norm": 2.2606234491837025, + "learning_rate": 1.7249381739661173e-05, + "loss": 0.994, + "step": 2751 + }, + { + "epoch": 0.26471719892266254, + "grad_norm": 2.1564506391039524, + "learning_rate": 1.7247235441028486e-05, + "loss": 1.003, + "step": 2752 + }, + { + "epoch": 0.26481338976529434, + "grad_norm": 2.0348467343404346, + "learning_rate": 1.72450884389892e-05, + "loss": 1.0355, + "step": 2753 + }, + { + "epoch": 0.26490958060792613, + "grad_norm": 2.1136632772465127, + "learning_rate": 1.7242940733751696e-05, + "loss": 1.0814, + "step": 2754 + }, + { + "epoch": 0.2650057714505579, + "grad_norm": 2.158149499093015, + "learning_rate": 1.7240792325524432e-05, + "loss": 1.1681, + "step": 2755 + }, + { + "epoch": 0.26510196229318966, + "grad_norm": 2.4515584048301107, + "learning_rate": 1.7238643214515934e-05, + "loss": 0.9609, + "step": 2756 + }, + { + "epoch": 0.26519815313582146, + "grad_norm": 2.177425956926368, + "learning_rate": 1.7236493400934783e-05, + "loss": 0.9478, + "step": 2757 + }, + { + "epoch": 0.26529434397845325, + "grad_norm": 2.4816447273422804, + "learning_rate": 1.7234342884989642e-05, + "loss": 1.0588, + "step": 2758 + }, + { + "epoch": 0.26539053482108504, + "grad_norm": 2.6622605704890634, + "learning_rate": 1.7232191666889232e-05, + "loss": 1.1008, + "step": 2759 + }, + { + "epoch": 0.26548672566371684, + "grad_norm": 2.0192786442509174, + "learning_rate": 1.7230039746842352e-05, + "loss": 1.0738, + "step": 2760 + }, + { + "epoch": 0.2655829165063486, + "grad_norm": 2.5439331345497855, + "learning_rate": 1.7227887125057865e-05, + "loss": 1.0414, + "step": 2761 + }, + { + "epoch": 0.26567910734898037, + "grad_norm": 2.4054211842759354, + "learning_rate": 1.7225733801744698e-05, + "loss": 1.1645, + "step": 2762 + }, + { + "epoch": 0.26577529819161216, + "grad_norm": 2.124496058447132, + "learning_rate": 1.7223579777111853e-05, + "loss": 1.0179, + "step": 2763 + }, + { + "epoch": 0.26587148903424396, + "grad_norm": 2.0701455359367547, + "learning_rate": 1.7221425051368394e-05, + "loss": 1.0935, + "step": 2764 + }, + { + "epoch": 0.2659676798768757, + "grad_norm": 2.1726203832968993, + "learning_rate": 1.721926962472346e-05, + "loss": 1.1447, + "step": 2765 + }, + { + "epoch": 0.2660638707195075, + "grad_norm": 2.1267258310236223, + "learning_rate": 1.7217113497386245e-05, + "loss": 0.9176, + "step": 2766 + }, + { + "epoch": 0.2661600615621393, + "grad_norm": 2.1933918940562784, + "learning_rate": 1.7214956669566026e-05, + "loss": 1.0629, + "step": 2767 + }, + { + "epoch": 0.2662562524047711, + "grad_norm": 2.025123304054611, + "learning_rate": 1.721279914147214e-05, + "loss": 1.077, + "step": 2768 + }, + { + "epoch": 0.26635244324740287, + "grad_norm": 2.258746243735254, + "learning_rate": 1.7210640913313997e-05, + "loss": 1.1267, + "step": 2769 + }, + { + "epoch": 0.2664486340900346, + "grad_norm": 2.25509072208741, + "learning_rate": 1.7208481985301065e-05, + "loss": 1.0767, + "step": 2770 + }, + { + "epoch": 0.2665448249326664, + "grad_norm": 2.2586176934980835, + "learning_rate": 1.7206322357642896e-05, + "loss": 1.1664, + "step": 2771 + }, + { + "epoch": 0.2666410157752982, + "grad_norm": 2.0597304045795872, + "learning_rate": 1.7204162030549093e-05, + "loss": 0.8579, + "step": 2772 + }, + { + "epoch": 0.26673720661793, + "grad_norm": 2.1986510512111463, + "learning_rate": 1.7202001004229335e-05, + "loss": 1.0545, + "step": 2773 + }, + { + "epoch": 0.2668333974605617, + "grad_norm": 2.435936664129976, + "learning_rate": 1.7199839278893368e-05, + "loss": 1.1019, + "step": 2774 + }, + { + "epoch": 0.2669295883031935, + "grad_norm": 2.4662259002113447, + "learning_rate": 1.719767685475101e-05, + "loss": 0.9696, + "step": 2775 + }, + { + "epoch": 0.2670257791458253, + "grad_norm": 2.3703686939808457, + "learning_rate": 1.719551373201214e-05, + "loss": 1.082, + "step": 2776 + }, + { + "epoch": 0.2671219699884571, + "grad_norm": 1.7188882105806316, + "learning_rate": 1.719334991088671e-05, + "loss": 0.9031, + "step": 2777 + }, + { + "epoch": 0.2672181608310889, + "grad_norm": 2.353832537245355, + "learning_rate": 1.7191185391584736e-05, + "loss": 1.1649, + "step": 2778 + }, + { + "epoch": 0.26731435167372064, + "grad_norm": 2.068892052185437, + "learning_rate": 1.7189020174316296e-05, + "loss": 1.087, + "step": 2779 + }, + { + "epoch": 0.26741054251635243, + "grad_norm": 2.403984952532964, + "learning_rate": 1.7186854259291558e-05, + "loss": 1.0803, + "step": 2780 + }, + { + "epoch": 0.26750673335898423, + "grad_norm": 2.09353513318997, + "learning_rate": 1.718468764672073e-05, + "loss": 0.9551, + "step": 2781 + }, + { + "epoch": 0.267602924201616, + "grad_norm": 1.8488583905618685, + "learning_rate": 1.7182520336814105e-05, + "loss": 1.0741, + "step": 2782 + }, + { + "epoch": 0.2676991150442478, + "grad_norm": 2.1610457805282914, + "learning_rate": 1.718035232978204e-05, + "loss": 1.1296, + "step": 2783 + }, + { + "epoch": 0.26779530588687955, + "grad_norm": 2.1539898325161455, + "learning_rate": 1.717818362583496e-05, + "loss": 1.0362, + "step": 2784 + }, + { + "epoch": 0.26789149672951135, + "grad_norm": 2.046529986901739, + "learning_rate": 1.7176014225183354e-05, + "loss": 0.9615, + "step": 2785 + }, + { + "epoch": 0.26798768757214314, + "grad_norm": 2.497932791066088, + "learning_rate": 1.7173844128037777e-05, + "loss": 1.0844, + "step": 2786 + }, + { + "epoch": 0.26808387841477493, + "grad_norm": 1.8976734495593834, + "learning_rate": 1.717167333460886e-05, + "loss": 0.8217, + "step": 2787 + }, + { + "epoch": 0.2681800692574067, + "grad_norm": 1.99571228971705, + "learning_rate": 1.71695018451073e-05, + "loss": 1.0111, + "step": 2788 + }, + { + "epoch": 0.26827626010003847, + "grad_norm": 1.9720664257500893, + "learning_rate": 1.7167329659743856e-05, + "loss": 0.9842, + "step": 2789 + }, + { + "epoch": 0.26837245094267026, + "grad_norm": 1.9848577883782874, + "learning_rate": 1.7165156778729355e-05, + "loss": 0.912, + "step": 2790 + }, + { + "epoch": 0.26846864178530205, + "grad_norm": 2.268703582798459, + "learning_rate": 1.71629832022747e-05, + "loss": 1.1568, + "step": 2791 + }, + { + "epoch": 0.26856483262793385, + "grad_norm": 2.2920887268156713, + "learning_rate": 1.7160808930590845e-05, + "loss": 1.1777, + "step": 2792 + }, + { + "epoch": 0.2686610234705656, + "grad_norm": 2.139480761387448, + "learning_rate": 1.7158633963888832e-05, + "loss": 1.1082, + "step": 2793 + }, + { + "epoch": 0.2687572143131974, + "grad_norm": 2.3513148680306446, + "learning_rate": 1.7156458302379753e-05, + "loss": 1.1394, + "step": 2794 + }, + { + "epoch": 0.2688534051558292, + "grad_norm": 2.13516996163998, + "learning_rate": 1.715428194627478e-05, + "loss": 1.1158, + "step": 2795 + }, + { + "epoch": 0.26894959599846097, + "grad_norm": 2.2347090484569714, + "learning_rate": 1.7152104895785147e-05, + "loss": 0.9984, + "step": 2796 + }, + { + "epoch": 0.2690457868410927, + "grad_norm": 2.4645269761248114, + "learning_rate": 1.714992715112215e-05, + "loss": 1.0115, + "step": 2797 + }, + { + "epoch": 0.2691419776837245, + "grad_norm": 2.251218353947275, + "learning_rate": 1.7147748712497162e-05, + "loss": 1.1004, + "step": 2798 + }, + { + "epoch": 0.2692381685263563, + "grad_norm": 2.1455855562711816, + "learning_rate": 1.7145569580121623e-05, + "loss": 1.0209, + "step": 2799 + }, + { + "epoch": 0.2693343593689881, + "grad_norm": 2.3812133977616727, + "learning_rate": 1.7143389754207026e-05, + "loss": 1.0313, + "step": 2800 + }, + { + "epoch": 0.2694305502116199, + "grad_norm": 2.1206856657359707, + "learning_rate": 1.714120923496495e-05, + "loss": 0.7311, + "step": 2801 + }, + { + "epoch": 0.2695267410542516, + "grad_norm": 2.0166107879673936, + "learning_rate": 1.713902802260703e-05, + "loss": 1.0628, + "step": 2802 + }, + { + "epoch": 0.2696229318968834, + "grad_norm": 1.8128700387497803, + "learning_rate": 1.7136846117344977e-05, + "loss": 0.7833, + "step": 2803 + }, + { + "epoch": 0.2697191227395152, + "grad_norm": 2.207727710021498, + "learning_rate": 1.7134663519390557e-05, + "loss": 0.9555, + "step": 2804 + }, + { + "epoch": 0.269815313582147, + "grad_norm": 2.058302218608701, + "learning_rate": 1.7132480228955612e-05, + "loss": 1.0712, + "step": 2805 + }, + { + "epoch": 0.26991150442477874, + "grad_norm": 2.3684019471544175, + "learning_rate": 1.7130296246252048e-05, + "loss": 1.0716, + "step": 2806 + }, + { + "epoch": 0.27000769526741053, + "grad_norm": 2.0532264806444145, + "learning_rate": 1.712811157149184e-05, + "loss": 1.0419, + "step": 2807 + }, + { + "epoch": 0.2701038861100423, + "grad_norm": 2.331732264417224, + "learning_rate": 1.7125926204887034e-05, + "loss": 1.0945, + "step": 2808 + }, + { + "epoch": 0.2702000769526741, + "grad_norm": 2.1920911458543655, + "learning_rate": 1.712374014664973e-05, + "loss": 1.0441, + "step": 2809 + }, + { + "epoch": 0.2702962677953059, + "grad_norm": 2.3043256423729495, + "learning_rate": 1.712155339699211e-05, + "loss": 0.8605, + "step": 2810 + }, + { + "epoch": 0.27039245863793765, + "grad_norm": 2.2073603074176225, + "learning_rate": 1.7119365956126422e-05, + "loss": 1.0124, + "step": 2811 + }, + { + "epoch": 0.27048864948056944, + "grad_norm": 2.513692141516718, + "learning_rate": 1.7117177824264962e-05, + "loss": 1.0982, + "step": 2812 + }, + { + "epoch": 0.27058484032320124, + "grad_norm": 1.9705536483923265, + "learning_rate": 1.7114989001620118e-05, + "loss": 1.0076, + "step": 2813 + }, + { + "epoch": 0.27068103116583303, + "grad_norm": 2.191063351754774, + "learning_rate": 1.7112799488404327e-05, + "loss": 1.0194, + "step": 2814 + }, + { + "epoch": 0.27077722200846477, + "grad_norm": 2.207549866092178, + "learning_rate": 1.711060928483011e-05, + "loss": 1.107, + "step": 2815 + }, + { + "epoch": 0.27087341285109656, + "grad_norm": 2.2908630781998838, + "learning_rate": 1.7108418391110033e-05, + "loss": 1.0365, + "step": 2816 + }, + { + "epoch": 0.27096960369372836, + "grad_norm": 2.4041384879281167, + "learning_rate": 1.710622680745675e-05, + "loss": 0.8659, + "step": 2817 + }, + { + "epoch": 0.27106579453636015, + "grad_norm": 1.9465278068886378, + "learning_rate": 1.7104034534082968e-05, + "loss": 1.0593, + "step": 2818 + }, + { + "epoch": 0.27116198537899194, + "grad_norm": 2.0105114553006587, + "learning_rate": 1.7101841571201467e-05, + "loss": 0.8558, + "step": 2819 + }, + { + "epoch": 0.2712581762216237, + "grad_norm": 2.0059546443564074, + "learning_rate": 1.7099647919025096e-05, + "loss": 1.0767, + "step": 2820 + }, + { + "epoch": 0.2713543670642555, + "grad_norm": 2.343483131058143, + "learning_rate": 1.709745357776677e-05, + "loss": 1.1711, + "step": 2821 + }, + { + "epoch": 0.27145055790688727, + "grad_norm": 2.090911437845029, + "learning_rate": 1.7095258547639456e-05, + "loss": 0.9497, + "step": 2822 + }, + { + "epoch": 0.27154674874951906, + "grad_norm": 2.0994875949005865, + "learning_rate": 1.7093062828856213e-05, + "loss": 1.1028, + "step": 2823 + }, + { + "epoch": 0.2716429395921508, + "grad_norm": 2.104914570740781, + "learning_rate": 1.709086642163015e-05, + "loss": 1.0311, + "step": 2824 + }, + { + "epoch": 0.2717391304347826, + "grad_norm": 2.292609140503484, + "learning_rate": 1.7088669326174448e-05, + "loss": 1.0389, + "step": 2825 + }, + { + "epoch": 0.2718353212774144, + "grad_norm": 2.1956774378030635, + "learning_rate": 1.7086471542702355e-05, + "loss": 0.901, + "step": 2826 + }, + { + "epoch": 0.2719315121200462, + "grad_norm": 2.2885720410751147, + "learning_rate": 1.708427307142718e-05, + "loss": 1.1183, + "step": 2827 + }, + { + "epoch": 0.272027702962678, + "grad_norm": 2.152503550147727, + "learning_rate": 1.708207391256231e-05, + "loss": 1.0834, + "step": 2828 + }, + { + "epoch": 0.2721238938053097, + "grad_norm": 2.349491522260492, + "learning_rate": 1.707987406632119e-05, + "loss": 0.9659, + "step": 2829 + }, + { + "epoch": 0.2722200846479415, + "grad_norm": 2.0796257287528914, + "learning_rate": 1.707767353291733e-05, + "loss": 1.0359, + "step": 2830 + }, + { + "epoch": 0.2723162754905733, + "grad_norm": 2.1163465970954647, + "learning_rate": 1.707547231256432e-05, + "loss": 1.0624, + "step": 2831 + }, + { + "epoch": 0.2724124663332051, + "grad_norm": 2.2473598789638256, + "learning_rate": 1.7073270405475796e-05, + "loss": 0.9989, + "step": 2832 + }, + { + "epoch": 0.2725086571758369, + "grad_norm": 2.261531989306434, + "learning_rate": 1.7071067811865477e-05, + "loss": 1.0805, + "step": 2833 + }, + { + "epoch": 0.2726048480184686, + "grad_norm": 2.1446593043663698, + "learning_rate": 1.7068864531947147e-05, + "loss": 1.0326, + "step": 2834 + }, + { + "epoch": 0.2727010388611004, + "grad_norm": 2.178604629304556, + "learning_rate": 1.706666056593465e-05, + "loss": 0.9848, + "step": 2835 + }, + { + "epoch": 0.2727972297037322, + "grad_norm": 1.9525034273991946, + "learning_rate": 1.70644559140419e-05, + "loss": 1.0125, + "step": 2836 + }, + { + "epoch": 0.272893420546364, + "grad_norm": 2.372476084644405, + "learning_rate": 1.7062250576482876e-05, + "loss": 1.0933, + "step": 2837 + }, + { + "epoch": 0.27298961138899575, + "grad_norm": 1.9923621656335513, + "learning_rate": 1.706004455347163e-05, + "loss": 1.0226, + "step": 2838 + }, + { + "epoch": 0.27308580223162754, + "grad_norm": 2.143060676737996, + "learning_rate": 1.705783784522227e-05, + "loss": 1.1167, + "step": 2839 + }, + { + "epoch": 0.27318199307425933, + "grad_norm": 1.9440147374843757, + "learning_rate": 1.705563045194898e-05, + "loss": 0.8535, + "step": 2840 + }, + { + "epoch": 0.2732781839168911, + "grad_norm": 2.170825229008026, + "learning_rate": 1.7053422373866003e-05, + "loss": 1.0222, + "step": 2841 + }, + { + "epoch": 0.2733743747595229, + "grad_norm": 1.8983562998892198, + "learning_rate": 1.7051213611187657e-05, + "loss": 0.9281, + "step": 2842 + }, + { + "epoch": 0.27347056560215466, + "grad_norm": 2.072683685667925, + "learning_rate": 1.7049004164128317e-05, + "loss": 1.0681, + "step": 2843 + }, + { + "epoch": 0.27356675644478645, + "grad_norm": 1.9285894537612212, + "learning_rate": 1.704679403290243e-05, + "loss": 1.0402, + "step": 2844 + }, + { + "epoch": 0.27366294728741825, + "grad_norm": 1.93635920943151, + "learning_rate": 1.704458321772451e-05, + "loss": 0.9866, + "step": 2845 + }, + { + "epoch": 0.27375913813005004, + "grad_norm": 2.290278681813826, + "learning_rate": 1.7042371718809132e-05, + "loss": 0.9599, + "step": 2846 + }, + { + "epoch": 0.2738553289726818, + "grad_norm": 2.03901291751668, + "learning_rate": 1.704015953637094e-05, + "loss": 0.9006, + "step": 2847 + }, + { + "epoch": 0.27395151981531357, + "grad_norm": 1.9419650491096063, + "learning_rate": 1.7037946670624652e-05, + "loss": 1.0333, + "step": 2848 + }, + { + "epoch": 0.27404771065794536, + "grad_norm": 2.0909896075770003, + "learning_rate": 1.7035733121785042e-05, + "loss": 0.935, + "step": 2849 + }, + { + "epoch": 0.27414390150057716, + "grad_norm": 2.137437234884752, + "learning_rate": 1.7033518890066956e-05, + "loss": 1.0221, + "step": 2850 + }, + { + "epoch": 0.27424009234320895, + "grad_norm": 1.936588224277323, + "learning_rate": 1.70313039756853e-05, + "loss": 1.005, + "step": 2851 + }, + { + "epoch": 0.2743362831858407, + "grad_norm": 2.1576524705372897, + "learning_rate": 1.7029088378855055e-05, + "loss": 1.0174, + "step": 2852 + }, + { + "epoch": 0.2744324740284725, + "grad_norm": 2.0393918329322975, + "learning_rate": 1.7026872099791257e-05, + "loss": 0.962, + "step": 2853 + }, + { + "epoch": 0.2745286648711043, + "grad_norm": 2.2820647507673573, + "learning_rate": 1.7024655138709025e-05, + "loss": 1.057, + "step": 2854 + }, + { + "epoch": 0.27462485571373607, + "grad_norm": 2.166332506407864, + "learning_rate": 1.7022437495823525e-05, + "loss": 1.067, + "step": 2855 + }, + { + "epoch": 0.2747210465563678, + "grad_norm": 2.4352931264256825, + "learning_rate": 1.7020219171350004e-05, + "loss": 0.9989, + "step": 2856 + }, + { + "epoch": 0.2748172373989996, + "grad_norm": 1.9973014174051473, + "learning_rate": 1.701800016550377e-05, + "loss": 1.0352, + "step": 2857 + }, + { + "epoch": 0.2749134282416314, + "grad_norm": 2.1893175094990673, + "learning_rate": 1.7015780478500187e-05, + "loss": 1.0029, + "step": 2858 + }, + { + "epoch": 0.2750096190842632, + "grad_norm": 2.2052347034894675, + "learning_rate": 1.7013560110554706e-05, + "loss": 0.9849, + "step": 2859 + }, + { + "epoch": 0.275105809926895, + "grad_norm": 2.2109943993441195, + "learning_rate": 1.701133906188283e-05, + "loss": 1.0447, + "step": 2860 + }, + { + "epoch": 0.2752020007695267, + "grad_norm": 2.4539598285965605, + "learning_rate": 1.7009117332700128e-05, + "loss": 1.0349, + "step": 2861 + }, + { + "epoch": 0.2752981916121585, + "grad_norm": 2.2855917803681782, + "learning_rate": 1.700689492322224e-05, + "loss": 1.1057, + "step": 2862 + }, + { + "epoch": 0.2753943824547903, + "grad_norm": 2.2000558260176772, + "learning_rate": 1.700467183366487e-05, + "loss": 1.0883, + "step": 2863 + }, + { + "epoch": 0.2754905732974221, + "grad_norm": 2.325579810037006, + "learning_rate": 1.700244806424379e-05, + "loss": 1.0207, + "step": 2864 + }, + { + "epoch": 0.27558676414005384, + "grad_norm": 2.426997451776293, + "learning_rate": 1.700022361517483e-05, + "loss": 1.0698, + "step": 2865 + }, + { + "epoch": 0.27568295498268564, + "grad_norm": 2.244725975291221, + "learning_rate": 1.6997998486673893e-05, + "loss": 1.0805, + "step": 2866 + }, + { + "epoch": 0.27577914582531743, + "grad_norm": 1.8773593452819257, + "learning_rate": 1.6995772678956956e-05, + "loss": 1.0927, + "step": 2867 + }, + { + "epoch": 0.2758753366679492, + "grad_norm": 2.2135727583674116, + "learning_rate": 1.699354619224004e-05, + "loss": 1.0464, + "step": 2868 + }, + { + "epoch": 0.275971527510581, + "grad_norm": 2.6902806650154294, + "learning_rate": 1.6991319026739254e-05, + "loss": 1.0923, + "step": 2869 + }, + { + "epoch": 0.27606771835321275, + "grad_norm": 1.9625894712697673, + "learning_rate": 1.698909118267076e-05, + "loss": 0.8569, + "step": 2870 + }, + { + "epoch": 0.27616390919584455, + "grad_norm": 2.254231090219753, + "learning_rate": 1.6986862660250792e-05, + "loss": 1.0724, + "step": 2871 + }, + { + "epoch": 0.27626010003847634, + "grad_norm": 2.4737252590960868, + "learning_rate": 1.6984633459695646e-05, + "loss": 1.0721, + "step": 2872 + }, + { + "epoch": 0.27635629088110814, + "grad_norm": 2.274659703077461, + "learning_rate": 1.698240358122168e-05, + "loss": 1.1565, + "step": 2873 + }, + { + "epoch": 0.2764524817237399, + "grad_norm": 2.087262260350707, + "learning_rate": 1.6980173025045328e-05, + "loss": 0.9868, + "step": 2874 + }, + { + "epoch": 0.27654867256637167, + "grad_norm": 1.9613827570996092, + "learning_rate": 1.6977941791383087e-05, + "loss": 1.0294, + "step": 2875 + }, + { + "epoch": 0.27664486340900346, + "grad_norm": 1.8842444141941779, + "learning_rate": 1.697570988045151e-05, + "loss": 0.9925, + "step": 2876 + }, + { + "epoch": 0.27674105425163525, + "grad_norm": 2.0502077372407785, + "learning_rate": 1.697347729246723e-05, + "loss": 1.0727, + "step": 2877 + }, + { + "epoch": 0.27683724509426705, + "grad_norm": 2.1439632710929843, + "learning_rate": 1.6971244027646937e-05, + "loss": 1.0649, + "step": 2878 + }, + { + "epoch": 0.2769334359368988, + "grad_norm": 2.2075761255918946, + "learning_rate": 1.6969010086207385e-05, + "loss": 1.0332, + "step": 2879 + }, + { + "epoch": 0.2770296267795306, + "grad_norm": 2.0262860648527012, + "learning_rate": 1.69667754683654e-05, + "loss": 1.0243, + "step": 2880 + }, + { + "epoch": 0.2771258176221624, + "grad_norm": 1.99629662961434, + "learning_rate": 1.6964540174337876e-05, + "loss": 1.091, + "step": 2881 + }, + { + "epoch": 0.27722200846479417, + "grad_norm": 2.4521279720597615, + "learning_rate": 1.6962304204341758e-05, + "loss": 0.9333, + "step": 2882 + }, + { + "epoch": 0.2773181993074259, + "grad_norm": 2.215825669887065, + "learning_rate": 1.6960067558594075e-05, + "loss": 1.0272, + "step": 2883 + }, + { + "epoch": 0.2774143901500577, + "grad_norm": 2.3746278984695413, + "learning_rate": 1.6957830237311904e-05, + "loss": 1.0632, + "step": 2884 + }, + { + "epoch": 0.2775105809926895, + "grad_norm": 2.1249964132906145, + "learning_rate": 1.69555922407124e-05, + "loss": 0.9955, + "step": 2885 + }, + { + "epoch": 0.2776067718353213, + "grad_norm": 2.2405348236269647, + "learning_rate": 1.6953353569012784e-05, + "loss": 1.1437, + "step": 2886 + }, + { + "epoch": 0.2777029626779531, + "grad_norm": 1.9694176904209506, + "learning_rate": 1.695111422243034e-05, + "loss": 1.0592, + "step": 2887 + }, + { + "epoch": 0.2777991535205848, + "grad_norm": 2.1147718220683505, + "learning_rate": 1.6948874201182402e-05, + "loss": 1.1343, + "step": 2888 + }, + { + "epoch": 0.2778953443632166, + "grad_norm": 2.083193897816004, + "learning_rate": 1.6946633505486395e-05, + "loss": 0.9739, + "step": 2889 + }, + { + "epoch": 0.2779915352058484, + "grad_norm": 2.043909287448012, + "learning_rate": 1.6944392135559798e-05, + "loss": 0.9552, + "step": 2890 + }, + { + "epoch": 0.2780877260484802, + "grad_norm": 2.2023489038231596, + "learning_rate": 1.6942150091620147e-05, + "loss": 0.9718, + "step": 2891 + }, + { + "epoch": 0.278183916891112, + "grad_norm": 2.1389305812425032, + "learning_rate": 1.6939907373885062e-05, + "loss": 0.9917, + "step": 2892 + }, + { + "epoch": 0.27828010773374373, + "grad_norm": 2.112769936600907, + "learning_rate": 1.6937663982572213e-05, + "loss": 1.0134, + "step": 2893 + }, + { + "epoch": 0.2783762985763755, + "grad_norm": 2.0425279779912535, + "learning_rate": 1.6935419917899335e-05, + "loss": 1.0847, + "step": 2894 + }, + { + "epoch": 0.2784724894190073, + "grad_norm": 2.10005655850215, + "learning_rate": 1.6933175180084243e-05, + "loss": 0.9102, + "step": 2895 + }, + { + "epoch": 0.2785686802616391, + "grad_norm": 2.2344111612408657, + "learning_rate": 1.6930929769344807e-05, + "loss": 1.0161, + "step": 2896 + }, + { + "epoch": 0.27866487110427085, + "grad_norm": 2.116170745451169, + "learning_rate": 1.6928683685898955e-05, + "loss": 1.0313, + "step": 2897 + }, + { + "epoch": 0.27876106194690264, + "grad_norm": 1.8499043218695561, + "learning_rate": 1.69264369299647e-05, + "loss": 0.9133, + "step": 2898 + }, + { + "epoch": 0.27885725278953444, + "grad_norm": 2.2238303735071527, + "learning_rate": 1.69241895017601e-05, + "loss": 1.1763, + "step": 2899 + }, + { + "epoch": 0.27895344363216623, + "grad_norm": 2.1833172940312937, + "learning_rate": 1.692194140150329e-05, + "loss": 1.025, + "step": 2900 + }, + { + "epoch": 0.279049634474798, + "grad_norm": 2.0258183813543393, + "learning_rate": 1.691969262941247e-05, + "loss": 1.1387, + "step": 2901 + }, + { + "epoch": 0.27914582531742976, + "grad_norm": 2.5600348290389587, + "learning_rate": 1.69174431857059e-05, + "loss": 1.1877, + "step": 2902 + }, + { + "epoch": 0.27924201616006156, + "grad_norm": 2.2612246254390223, + "learning_rate": 1.691519307060191e-05, + "loss": 1.1289, + "step": 2903 + }, + { + "epoch": 0.27933820700269335, + "grad_norm": 2.160284449963364, + "learning_rate": 1.6912942284318898e-05, + "loss": 1.0392, + "step": 2904 + }, + { + "epoch": 0.27943439784532514, + "grad_norm": 2.1785607948995427, + "learning_rate": 1.6910690827075306e-05, + "loss": 1.0858, + "step": 2905 + }, + { + "epoch": 0.2795305886879569, + "grad_norm": 1.7692230887854659, + "learning_rate": 1.6908438699089674e-05, + "loss": 0.8272, + "step": 2906 + }, + { + "epoch": 0.2796267795305887, + "grad_norm": 1.6526696632668303, + "learning_rate": 1.6906185900580578e-05, + "loss": 0.8873, + "step": 2907 + }, + { + "epoch": 0.27972297037322047, + "grad_norm": 1.8870018702690168, + "learning_rate": 1.690393243176668e-05, + "loss": 0.8768, + "step": 2908 + }, + { + "epoch": 0.27981916121585226, + "grad_norm": 2.074630953359199, + "learning_rate": 1.6901678292866698e-05, + "loss": 1.0449, + "step": 2909 + }, + { + "epoch": 0.27991535205848406, + "grad_norm": 2.2115697288595024, + "learning_rate": 1.6899423484099413e-05, + "loss": 1.0259, + "step": 2910 + }, + { + "epoch": 0.2800115429011158, + "grad_norm": 2.1324743502211816, + "learning_rate": 1.6897168005683676e-05, + "loss": 0.959, + "step": 2911 + }, + { + "epoch": 0.2801077337437476, + "grad_norm": 2.039097285996013, + "learning_rate": 1.6894911857838394e-05, + "loss": 0.9462, + "step": 2912 + }, + { + "epoch": 0.2802039245863794, + "grad_norm": 2.233458190622792, + "learning_rate": 1.689265504078255e-05, + "loss": 1.1106, + "step": 2913 + }, + { + "epoch": 0.2803001154290112, + "grad_norm": 2.011380131388258, + "learning_rate": 1.689039755473519e-05, + "loss": 1.1109, + "step": 2914 + }, + { + "epoch": 0.2803963062716429, + "grad_norm": 2.8863866867364334, + "learning_rate": 1.688813939991542e-05, + "loss": 0.936, + "step": 2915 + }, + { + "epoch": 0.2804924971142747, + "grad_norm": 2.1916909072618416, + "learning_rate": 1.6885880576542417e-05, + "loss": 1.1192, + "step": 2916 + }, + { + "epoch": 0.2805886879569065, + "grad_norm": 2.093353355301099, + "learning_rate": 1.688362108483541e-05, + "loss": 1.0439, + "step": 2917 + }, + { + "epoch": 0.2806848787995383, + "grad_norm": 1.938794496672704, + "learning_rate": 1.6881360925013712e-05, + "loss": 1.0418, + "step": 2918 + }, + { + "epoch": 0.2807810696421701, + "grad_norm": 2.1572760942161855, + "learning_rate": 1.6879100097296688e-05, + "loss": 0.9873, + "step": 2919 + }, + { + "epoch": 0.2808772604848018, + "grad_norm": 3.0210174001808414, + "learning_rate": 1.6876838601903765e-05, + "loss": 0.9496, + "step": 2920 + }, + { + "epoch": 0.2809734513274336, + "grad_norm": 1.920640937747419, + "learning_rate": 1.687457643905445e-05, + "loss": 1.0393, + "step": 2921 + }, + { + "epoch": 0.2810696421700654, + "grad_norm": 2.3594507515289713, + "learning_rate": 1.6872313608968296e-05, + "loss": 1.0926, + "step": 2922 + }, + { + "epoch": 0.2811658330126972, + "grad_norm": 2.006118469017907, + "learning_rate": 1.6870050111864937e-05, + "loss": 0.9059, + "step": 2923 + }, + { + "epoch": 0.28126202385532895, + "grad_norm": 2.2318561414667415, + "learning_rate": 1.6867785947964065e-05, + "loss": 1.0892, + "step": 2924 + }, + { + "epoch": 0.28135821469796074, + "grad_norm": 1.9037503816567953, + "learning_rate": 1.6865521117485432e-05, + "loss": 0.9456, + "step": 2925 + }, + { + "epoch": 0.28145440554059253, + "grad_norm": 1.999324477882311, + "learning_rate": 1.6863255620648866e-05, + "loss": 1.0302, + "step": 2926 + }, + { + "epoch": 0.2815505963832243, + "grad_norm": 2.0370073622731786, + "learning_rate": 1.6860989457674243e-05, + "loss": 1.067, + "step": 2927 + }, + { + "epoch": 0.2816467872258561, + "grad_norm": 2.2542546625282274, + "learning_rate": 1.685872262878152e-05, + "loss": 1.1391, + "step": 2928 + }, + { + "epoch": 0.28174297806848786, + "grad_norm": 2.1826864430842527, + "learning_rate": 1.6856455134190714e-05, + "loss": 0.9132, + "step": 2929 + }, + { + "epoch": 0.28183916891111965, + "grad_norm": 2.2283432812630455, + "learning_rate": 1.6854186974121903e-05, + "loss": 1.1048, + "step": 2930 + }, + { + "epoch": 0.28193535975375145, + "grad_norm": 2.185964450992616, + "learning_rate": 1.6851918148795228e-05, + "loss": 0.9209, + "step": 2931 + }, + { + "epoch": 0.28203155059638324, + "grad_norm": 1.8561338308039232, + "learning_rate": 1.68496486584309e-05, + "loss": 0.9844, + "step": 2932 + }, + { + "epoch": 0.282127741439015, + "grad_norm": 2.055376534922826, + "learning_rate": 1.6847378503249197e-05, + "loss": 1.0684, + "step": 2933 + }, + { + "epoch": 0.28222393228164677, + "grad_norm": 1.9865232982622059, + "learning_rate": 1.6845107683470453e-05, + "loss": 1.0611, + "step": 2934 + }, + { + "epoch": 0.28232012312427857, + "grad_norm": 2.13864435908066, + "learning_rate": 1.6842836199315068e-05, + "loss": 0.9129, + "step": 2935 + }, + { + "epoch": 0.28241631396691036, + "grad_norm": 2.329050966846699, + "learning_rate": 1.6840564051003517e-05, + "loss": 0.9814, + "step": 2936 + }, + { + "epoch": 0.28251250480954215, + "grad_norm": 1.9265596463526968, + "learning_rate": 1.6838291238756324e-05, + "loss": 1.0708, + "step": 2937 + }, + { + "epoch": 0.2826086956521739, + "grad_norm": 2.508119504310265, + "learning_rate": 1.6836017762794087e-05, + "loss": 0.9432, + "step": 2938 + }, + { + "epoch": 0.2827048864948057, + "grad_norm": 1.8282904090114562, + "learning_rate": 1.6833743623337467e-05, + "loss": 0.8206, + "step": 2939 + }, + { + "epoch": 0.2828010773374375, + "grad_norm": 2.0356462895317713, + "learning_rate": 1.6831468820607192e-05, + "loss": 1.019, + "step": 2940 + }, + { + "epoch": 0.2828972681800693, + "grad_norm": 2.0455124417714208, + "learning_rate": 1.682919335482405e-05, + "loss": 1.034, + "step": 2941 + }, + { + "epoch": 0.28299345902270107, + "grad_norm": 2.5864040870345226, + "learning_rate": 1.6826917226208886e-05, + "loss": 1.0328, + "step": 2942 + }, + { + "epoch": 0.2830896498653328, + "grad_norm": 2.225608879025457, + "learning_rate": 1.6824640434982627e-05, + "loss": 1.0576, + "step": 2943 + }, + { + "epoch": 0.2831858407079646, + "grad_norm": 1.9884694312296132, + "learning_rate": 1.6822362981366257e-05, + "loss": 1.0487, + "step": 2944 + }, + { + "epoch": 0.2832820315505964, + "grad_norm": 2.2640207624004374, + "learning_rate": 1.6820084865580815e-05, + "loss": 1.175, + "step": 2945 + }, + { + "epoch": 0.2833782223932282, + "grad_norm": 2.120714345351045, + "learning_rate": 1.6817806087847417e-05, + "loss": 0.9727, + "step": 2946 + }, + { + "epoch": 0.2834744132358599, + "grad_norm": 2.0643988329380885, + "learning_rate": 1.6815526648387238e-05, + "loss": 0.9514, + "step": 2947 + }, + { + "epoch": 0.2835706040784917, + "grad_norm": 2.2217874730003957, + "learning_rate": 1.681324654742151e-05, + "loss": 1.0336, + "step": 2948 + }, + { + "epoch": 0.2836667949211235, + "grad_norm": 2.1932504878838417, + "learning_rate": 1.6810965785171546e-05, + "loss": 1.0293, + "step": 2949 + }, + { + "epoch": 0.2837629857637553, + "grad_norm": 1.8118413358030612, + "learning_rate": 1.6808684361858706e-05, + "loss": 1.0968, + "step": 2950 + }, + { + "epoch": 0.2838591766063871, + "grad_norm": 2.0399370560532635, + "learning_rate": 1.6806402277704425e-05, + "loss": 1.0864, + "step": 2951 + }, + { + "epoch": 0.28395536744901884, + "grad_norm": 1.9534539586312418, + "learning_rate": 1.6804119532930202e-05, + "loss": 0.7894, + "step": 2952 + }, + { + "epoch": 0.28405155829165063, + "grad_norm": 2.0410273079505843, + "learning_rate": 1.6801836127757594e-05, + "loss": 0.9331, + "step": 2953 + }, + { + "epoch": 0.2841477491342824, + "grad_norm": 2.158566938329928, + "learning_rate": 1.6799552062408225e-05, + "loss": 1.0659, + "step": 2954 + }, + { + "epoch": 0.2842439399769142, + "grad_norm": 2.118258265673171, + "learning_rate": 1.679726733710378e-05, + "loss": 1.1053, + "step": 2955 + }, + { + "epoch": 0.28434013081954596, + "grad_norm": 2.128537945465461, + "learning_rate": 1.6794981952066018e-05, + "loss": 1.0779, + "step": 2956 + }, + { + "epoch": 0.28443632166217775, + "grad_norm": 2.5069354879324783, + "learning_rate": 1.679269590751675e-05, + "loss": 1.0205, + "step": 2957 + }, + { + "epoch": 0.28453251250480954, + "grad_norm": 2.28367563228548, + "learning_rate": 1.6790409203677863e-05, + "loss": 1.0783, + "step": 2958 + }, + { + "epoch": 0.28462870334744134, + "grad_norm": 2.0126006827386527, + "learning_rate": 1.6788121840771294e-05, + "loss": 1.0295, + "step": 2959 + }, + { + "epoch": 0.28472489419007313, + "grad_norm": 1.9561409769538638, + "learning_rate": 1.6785833819019052e-05, + "loss": 1.081, + "step": 2960 + }, + { + "epoch": 0.28482108503270487, + "grad_norm": 2.2657273822288717, + "learning_rate": 1.6783545138643213e-05, + "loss": 1.1056, + "step": 2961 + }, + { + "epoch": 0.28491727587533666, + "grad_norm": 2.1039610049688293, + "learning_rate": 1.678125579986591e-05, + "loss": 1.0643, + "step": 2962 + }, + { + "epoch": 0.28501346671796846, + "grad_norm": 2.2336701319303827, + "learning_rate": 1.6778965802909345e-05, + "loss": 0.9709, + "step": 2963 + }, + { + "epoch": 0.28510965756060025, + "grad_norm": 2.1066207791639036, + "learning_rate": 1.677667514799578e-05, + "loss": 1.0939, + "step": 2964 + }, + { + "epoch": 0.285205848403232, + "grad_norm": 2.11774062925401, + "learning_rate": 1.677438383534755e-05, + "loss": 1.0196, + "step": 2965 + }, + { + "epoch": 0.2853020392458638, + "grad_norm": 2.0444595722979035, + "learning_rate": 1.6772091865187032e-05, + "loss": 1.1107, + "step": 2966 + }, + { + "epoch": 0.2853982300884956, + "grad_norm": 2.4089759861932794, + "learning_rate": 1.6769799237736694e-05, + "loss": 1.1794, + "step": 2967 + }, + { + "epoch": 0.28549442093112737, + "grad_norm": 1.9805803541236633, + "learning_rate": 1.676750595321905e-05, + "loss": 0.9715, + "step": 2968 + }, + { + "epoch": 0.28559061177375916, + "grad_norm": 1.9707101467772354, + "learning_rate": 1.6765212011856685e-05, + "loss": 0.8265, + "step": 2969 + }, + { + "epoch": 0.2856868026163909, + "grad_norm": 2.0255670392106553, + "learning_rate": 1.6762917413872246e-05, + "loss": 1.0909, + "step": 2970 + }, + { + "epoch": 0.2857829934590227, + "grad_norm": 2.2848274678740332, + "learning_rate": 1.6760622159488444e-05, + "loss": 0.9644, + "step": 2971 + }, + { + "epoch": 0.2858791843016545, + "grad_norm": 2.0248905259990586, + "learning_rate": 1.675832624892805e-05, + "loss": 1.0993, + "step": 2972 + }, + { + "epoch": 0.2859753751442863, + "grad_norm": 2.1458786497326763, + "learning_rate": 1.6756029682413903e-05, + "loss": 1.0405, + "step": 2973 + }, + { + "epoch": 0.286071565986918, + "grad_norm": 1.949321756473326, + "learning_rate": 1.6753732460168907e-05, + "loss": 1.04, + "step": 2974 + }, + { + "epoch": 0.2861677568295498, + "grad_norm": 2.2430953583734397, + "learning_rate": 1.675143458241602e-05, + "loss": 1.1742, + "step": 2975 + }, + { + "epoch": 0.2862639476721816, + "grad_norm": 2.1260432884929803, + "learning_rate": 1.674913604937828e-05, + "loss": 1.1481, + "step": 2976 + }, + { + "epoch": 0.2863601385148134, + "grad_norm": 2.1690858004402527, + "learning_rate": 1.6746836861278776e-05, + "loss": 0.9677, + "step": 2977 + }, + { + "epoch": 0.2864563293574452, + "grad_norm": 2.2500001131000067, + "learning_rate": 1.6744537018340662e-05, + "loss": 1.1466, + "step": 2978 + }, + { + "epoch": 0.28655252020007693, + "grad_norm": 2.208066353513775, + "learning_rate": 1.6742236520787165e-05, + "loss": 1.185, + "step": 2979 + }, + { + "epoch": 0.2866487110427087, + "grad_norm": 2.2173790712943258, + "learning_rate": 1.6739935368841555e-05, + "loss": 1.0098, + "step": 2980 + }, + { + "epoch": 0.2867449018853405, + "grad_norm": 2.371702465380246, + "learning_rate": 1.673763356272719e-05, + "loss": 1.064, + "step": 2981 + }, + { + "epoch": 0.2868410927279723, + "grad_norm": 2.292969140315606, + "learning_rate": 1.6735331102667475e-05, + "loss": 1.1133, + "step": 2982 + }, + { + "epoch": 0.28693728357060405, + "grad_norm": 2.0384478752855553, + "learning_rate": 1.6733027988885885e-05, + "loss": 0.8164, + "step": 2983 + }, + { + "epoch": 0.28703347441323585, + "grad_norm": 2.2288224279015405, + "learning_rate": 1.6730724221605955e-05, + "loss": 0.9825, + "step": 2984 + }, + { + "epoch": 0.28712966525586764, + "grad_norm": 2.144636263359917, + "learning_rate": 1.672841980105129e-05, + "loss": 1.0105, + "step": 2985 + }, + { + "epoch": 0.28722585609849943, + "grad_norm": 2.2890195566755924, + "learning_rate": 1.6726114727445547e-05, + "loss": 0.9879, + "step": 2986 + }, + { + "epoch": 0.2873220469411312, + "grad_norm": 2.6455929382992305, + "learning_rate": 1.6723809001012457e-05, + "loss": 1.1316, + "step": 2987 + }, + { + "epoch": 0.28741823778376296, + "grad_norm": 1.9961253502962504, + "learning_rate": 1.6721502621975813e-05, + "loss": 1.0402, + "step": 2988 + }, + { + "epoch": 0.28751442862639476, + "grad_norm": 2.135453997993856, + "learning_rate": 1.6719195590559466e-05, + "loss": 1.1211, + "step": 2989 + }, + { + "epoch": 0.28761061946902655, + "grad_norm": 2.214614463150351, + "learning_rate": 1.6716887906987332e-05, + "loss": 0.9512, + "step": 2990 + }, + { + "epoch": 0.28770681031165835, + "grad_norm": 2.1111835124707476, + "learning_rate": 1.67145795714834e-05, + "loss": 1.1245, + "step": 2991 + }, + { + "epoch": 0.28780300115429014, + "grad_norm": 2.275006395785719, + "learning_rate": 1.6712270584271703e-05, + "loss": 1.0886, + "step": 2992 + }, + { + "epoch": 0.2878991919969219, + "grad_norm": 2.323396612984043, + "learning_rate": 1.6709960945576352e-05, + "loss": 1.0317, + "step": 2993 + }, + { + "epoch": 0.28799538283955367, + "grad_norm": 2.146800053599642, + "learning_rate": 1.670765065562152e-05, + "loss": 1.111, + "step": 2994 + }, + { + "epoch": 0.28809157368218546, + "grad_norm": 1.9916811285306597, + "learning_rate": 1.6705339714631436e-05, + "loss": 1.0157, + "step": 2995 + }, + { + "epoch": 0.28818776452481726, + "grad_norm": 2.140113551823018, + "learning_rate": 1.67030281228304e-05, + "loss": 1.0298, + "step": 2996 + }, + { + "epoch": 0.288283955367449, + "grad_norm": 2.2647381009214254, + "learning_rate": 1.670071588044277e-05, + "loss": 0.9919, + "step": 2997 + }, + { + "epoch": 0.2883801462100808, + "grad_norm": 1.8865709762544145, + "learning_rate": 1.6698402987692968e-05, + "loss": 1.0067, + "step": 2998 + }, + { + "epoch": 0.2884763370527126, + "grad_norm": 2.0686237682426567, + "learning_rate": 1.6696089444805485e-05, + "loss": 1.0664, + "step": 2999 + }, + { + "epoch": 0.2885725278953444, + "grad_norm": 2.3731186754571283, + "learning_rate": 1.6693775252004866e-05, + "loss": 1.1253, + "step": 3000 + }, + { + "epoch": 0.28866871873797617, + "grad_norm": 2.1778951646531906, + "learning_rate": 1.6691460409515725e-05, + "loss": 1.1379, + "step": 3001 + }, + { + "epoch": 0.2887649095806079, + "grad_norm": 2.058563097275455, + "learning_rate": 1.668914491756274e-05, + "loss": 1.0259, + "step": 3002 + }, + { + "epoch": 0.2888611004232397, + "grad_norm": 2.5102665399789204, + "learning_rate": 1.6686828776370644e-05, + "loss": 1.0023, + "step": 3003 + }, + { + "epoch": 0.2889572912658715, + "grad_norm": 2.1971125998955685, + "learning_rate": 1.668451198616424e-05, + "loss": 1.0139, + "step": 3004 + }, + { + "epoch": 0.2890534821085033, + "grad_norm": 2.4369833791436855, + "learning_rate": 1.6682194547168394e-05, + "loss": 1.1067, + "step": 3005 + }, + { + "epoch": 0.28914967295113503, + "grad_norm": 2.128054332970666, + "learning_rate": 1.6679876459608033e-05, + "loss": 1.0753, + "step": 3006 + }, + { + "epoch": 0.2892458637937668, + "grad_norm": 2.306930902905139, + "learning_rate": 1.6677557723708145e-05, + "loss": 1.0534, + "step": 3007 + }, + { + "epoch": 0.2893420546363986, + "grad_norm": 2.2119141097172528, + "learning_rate": 1.667523833969379e-05, + "loss": 0.9609, + "step": 3008 + }, + { + "epoch": 0.2894382454790304, + "grad_norm": 2.470848309884198, + "learning_rate": 1.667291830779008e-05, + "loss": 1.0038, + "step": 3009 + }, + { + "epoch": 0.2895344363216622, + "grad_norm": 2.123759903495446, + "learning_rate": 1.667059762822219e-05, + "loss": 1.0259, + "step": 3010 + }, + { + "epoch": 0.28963062716429394, + "grad_norm": 2.180101202607538, + "learning_rate": 1.6668276301215364e-05, + "loss": 1.0401, + "step": 3011 + }, + { + "epoch": 0.28972681800692573, + "grad_norm": 2.871018478042207, + "learning_rate": 1.666595432699491e-05, + "loss": 1.0752, + "step": 3012 + }, + { + "epoch": 0.28982300884955753, + "grad_norm": 2.0708749275828704, + "learning_rate": 1.6663631705786197e-05, + "loss": 1.0567, + "step": 3013 + }, + { + "epoch": 0.2899191996921893, + "grad_norm": 2.035282568438115, + "learning_rate": 1.6661308437814652e-05, + "loss": 0.9847, + "step": 3014 + }, + { + "epoch": 0.29001539053482106, + "grad_norm": 2.15815526239796, + "learning_rate": 1.6658984523305764e-05, + "loss": 1.0704, + "step": 3015 + }, + { + "epoch": 0.29011158137745285, + "grad_norm": 2.176623984295083, + "learning_rate": 1.6656659962485097e-05, + "loss": 1.0037, + "step": 3016 + }, + { + "epoch": 0.29020777222008465, + "grad_norm": 2.172526181036289, + "learning_rate": 1.665433475557827e-05, + "loss": 1.0568, + "step": 3017 + }, + { + "epoch": 0.29030396306271644, + "grad_norm": 2.204750503079038, + "learning_rate": 1.6652008902810952e-05, + "loss": 1.1469, + "step": 3018 + }, + { + "epoch": 0.29040015390534824, + "grad_norm": 1.9904774632165678, + "learning_rate": 1.66496824044089e-05, + "loss": 1.0568, + "step": 3019 + }, + { + "epoch": 0.29049634474798, + "grad_norm": 1.9700562241435284, + "learning_rate": 1.6647355260597915e-05, + "loss": 1.0128, + "step": 3020 + }, + { + "epoch": 0.29059253559061177, + "grad_norm": 1.9200624104160924, + "learning_rate": 1.6645027471603867e-05, + "loss": 0.9162, + "step": 3021 + }, + { + "epoch": 0.29068872643324356, + "grad_norm": 2.2023417210396343, + "learning_rate": 1.664269903765269e-05, + "loss": 0.8974, + "step": 3022 + }, + { + "epoch": 0.29078491727587535, + "grad_norm": 2.3179476379369772, + "learning_rate": 1.6640369958970377e-05, + "loss": 1.0224, + "step": 3023 + }, + { + "epoch": 0.2908811081185071, + "grad_norm": 2.0120566937456914, + "learning_rate": 1.6638040235782983e-05, + "loss": 1.0329, + "step": 3024 + }, + { + "epoch": 0.2909772989611389, + "grad_norm": 2.304804249778472, + "learning_rate": 1.6635709868316634e-05, + "loss": 1.0375, + "step": 3025 + }, + { + "epoch": 0.2910734898037707, + "grad_norm": 2.0823263411930277, + "learning_rate": 1.6633378856797505e-05, + "loss": 1.0171, + "step": 3026 + }, + { + "epoch": 0.2911696806464025, + "grad_norm": 2.2822020721316805, + "learning_rate": 1.6631047201451844e-05, + "loss": 1.0266, + "step": 3027 + }, + { + "epoch": 0.29126587148903427, + "grad_norm": 2.1673434914045315, + "learning_rate": 1.662871490250596e-05, + "loss": 1.1164, + "step": 3028 + }, + { + "epoch": 0.291362062331666, + "grad_norm": 2.095437054168291, + "learning_rate": 1.6626381960186217e-05, + "loss": 1.1522, + "step": 3029 + }, + { + "epoch": 0.2914582531742978, + "grad_norm": 2.210973069827804, + "learning_rate": 1.662404837471905e-05, + "loss": 1.049, + "step": 3030 + }, + { + "epoch": 0.2915544440169296, + "grad_norm": 2.0782562562815645, + "learning_rate": 1.6621714146330964e-05, + "loss": 1.1198, + "step": 3031 + }, + { + "epoch": 0.2916506348595614, + "grad_norm": 2.094761076902606, + "learning_rate": 1.66193792752485e-05, + "loss": 0.9208, + "step": 3032 + }, + { + "epoch": 0.2917468257021931, + "grad_norm": 1.8099469756104587, + "learning_rate": 1.6617043761698282e-05, + "loss": 0.963, + "step": 3033 + }, + { + "epoch": 0.2918430165448249, + "grad_norm": 1.811984432591442, + "learning_rate": 1.6614707605906995e-05, + "loss": 1.1265, + "step": 3034 + }, + { + "epoch": 0.2919392073874567, + "grad_norm": 1.899819282760927, + "learning_rate": 1.6612370808101383e-05, + "loss": 1.0372, + "step": 3035 + }, + { + "epoch": 0.2920353982300885, + "grad_norm": 1.9441333634544755, + "learning_rate": 1.661003336850825e-05, + "loss": 1.0402, + "step": 3036 + }, + { + "epoch": 0.2921315890727203, + "grad_norm": 1.9643906210492066, + "learning_rate": 1.6607695287354467e-05, + "loss": 1.0204, + "step": 3037 + }, + { + "epoch": 0.29222777991535204, + "grad_norm": 2.250229428008739, + "learning_rate": 1.660535656486696e-05, + "loss": 0.8894, + "step": 3038 + }, + { + "epoch": 0.29232397075798383, + "grad_norm": 2.091160841057856, + "learning_rate": 1.6603017201272728e-05, + "loss": 0.9138, + "step": 3039 + }, + { + "epoch": 0.2924201616006156, + "grad_norm": 2.0597048627111207, + "learning_rate": 1.660067719679882e-05, + "loss": 0.97, + "step": 3040 + }, + { + "epoch": 0.2925163524432474, + "grad_norm": 2.1755994147437967, + "learning_rate": 1.6598336551672364e-05, + "loss": 1.0346, + "step": 3041 + }, + { + "epoch": 0.2926125432858792, + "grad_norm": 2.0712919809798263, + "learning_rate": 1.6595995266120528e-05, + "loss": 1.0379, + "step": 3042 + }, + { + "epoch": 0.29270873412851095, + "grad_norm": 2.049887520576137, + "learning_rate": 1.6593653340370562e-05, + "loss": 1.0674, + "step": 3043 + }, + { + "epoch": 0.29280492497114274, + "grad_norm": 1.8703649413924115, + "learning_rate": 1.6591310774649766e-05, + "loss": 0.9407, + "step": 3044 + }, + { + "epoch": 0.29290111581377454, + "grad_norm": 1.968528329408905, + "learning_rate": 1.6588967569185506e-05, + "loss": 1.0766, + "step": 3045 + }, + { + "epoch": 0.29299730665640633, + "grad_norm": 2.113844197440714, + "learning_rate": 1.6586623724205216e-05, + "loss": 1.0532, + "step": 3046 + }, + { + "epoch": 0.29309349749903807, + "grad_norm": 2.010490729344051, + "learning_rate": 1.658427923993638e-05, + "loss": 0.9841, + "step": 3047 + }, + { + "epoch": 0.29318968834166986, + "grad_norm": 2.1201508372642457, + "learning_rate": 1.6581934116606554e-05, + "loss": 1.1425, + "step": 3048 + }, + { + "epoch": 0.29328587918430166, + "grad_norm": 1.9177429427258523, + "learning_rate": 1.6579588354443346e-05, + "loss": 1.0051, + "step": 3049 + }, + { + "epoch": 0.29338207002693345, + "grad_norm": 2.103080062981982, + "learning_rate": 1.657724195367444e-05, + "loss": 1.0617, + "step": 3050 + }, + { + "epoch": 0.29347826086956524, + "grad_norm": 2.0326122163255227, + "learning_rate": 1.657489491452757e-05, + "loss": 0.9307, + "step": 3051 + }, + { + "epoch": 0.293574451712197, + "grad_norm": 2.017013224654033, + "learning_rate": 1.657254723723054e-05, + "loss": 0.9948, + "step": 3052 + }, + { + "epoch": 0.2936706425548288, + "grad_norm": 2.208203997670322, + "learning_rate": 1.657019892201121e-05, + "loss": 1.1557, + "step": 3053 + }, + { + "epoch": 0.29376683339746057, + "grad_norm": 2.487661764915718, + "learning_rate": 1.6567849969097505e-05, + "loss": 1.0803, + "step": 3054 + }, + { + "epoch": 0.29386302424009236, + "grad_norm": 2.0635690719040625, + "learning_rate": 1.6565500378717412e-05, + "loss": 0.9549, + "step": 3055 + }, + { + "epoch": 0.2939592150827241, + "grad_norm": 2.4313620939651313, + "learning_rate": 1.6563150151098973e-05, + "loss": 1.0194, + "step": 3056 + }, + { + "epoch": 0.2940554059253559, + "grad_norm": 1.9973929077028953, + "learning_rate": 1.6560799286470307e-05, + "loss": 1.0118, + "step": 3057 + }, + { + "epoch": 0.2941515967679877, + "grad_norm": 2.424003568233184, + "learning_rate": 1.6558447785059577e-05, + "loss": 1.1124, + "step": 3058 + }, + { + "epoch": 0.2942477876106195, + "grad_norm": 2.3537501330888064, + "learning_rate": 1.6556095647095023e-05, + "loss": 0.9906, + "step": 3059 + }, + { + "epoch": 0.2943439784532513, + "grad_norm": 2.2910399197067557, + "learning_rate": 1.655374287280494e-05, + "loss": 1.0662, + "step": 3060 + }, + { + "epoch": 0.294440169295883, + "grad_norm": 2.1913884785774407, + "learning_rate": 1.6551389462417677e-05, + "loss": 1.1195, + "step": 3061 + }, + { + "epoch": 0.2945363601385148, + "grad_norm": 2.200180719221942, + "learning_rate": 1.6549035416161662e-05, + "loss": 1.0514, + "step": 3062 + }, + { + "epoch": 0.2946325509811466, + "grad_norm": 2.2020957248132382, + "learning_rate": 1.6546680734265375e-05, + "loss": 1.0716, + "step": 3063 + }, + { + "epoch": 0.2947287418237784, + "grad_norm": 2.1320346008858966, + "learning_rate": 1.654432541695735e-05, + "loss": 1.0158, + "step": 3064 + }, + { + "epoch": 0.29482493266641013, + "grad_norm": 2.3839182447096374, + "learning_rate": 1.6541969464466197e-05, + "loss": 1.014, + "step": 3065 + }, + { + "epoch": 0.2949211235090419, + "grad_norm": 1.8217025453241085, + "learning_rate": 1.653961287702058e-05, + "loss": 0.9879, + "step": 3066 + }, + { + "epoch": 0.2950173143516737, + "grad_norm": 2.1480053180613408, + "learning_rate": 1.653725565484923e-05, + "loss": 1.0414, + "step": 3067 + }, + { + "epoch": 0.2951135051943055, + "grad_norm": 2.2846927978353726, + "learning_rate": 1.653489779818093e-05, + "loss": 1.1538, + "step": 3068 + }, + { + "epoch": 0.2952096960369373, + "grad_norm": 2.3029471070243317, + "learning_rate": 1.653253930724453e-05, + "loss": 1.0712, + "step": 3069 + }, + { + "epoch": 0.29530588687956905, + "grad_norm": 2.3327807794226274, + "learning_rate": 1.6530180182268946e-05, + "loss": 1.1645, + "step": 3070 + }, + { + "epoch": 0.29540207772220084, + "grad_norm": 2.3763940645084465, + "learning_rate": 1.6527820423483154e-05, + "loss": 0.9703, + "step": 3071 + }, + { + "epoch": 0.29549826856483263, + "grad_norm": 2.2690522775705646, + "learning_rate": 1.652546003111618e-05, + "loss": 0.984, + "step": 3072 + }, + { + "epoch": 0.2955944594074644, + "grad_norm": 2.0639580396669555, + "learning_rate": 1.6523099005397126e-05, + "loss": 1.0036, + "step": 3073 + }, + { + "epoch": 0.29569065025009617, + "grad_norm": 2.028988345220215, + "learning_rate": 1.652073734655515e-05, + "loss": 1.1321, + "step": 3074 + }, + { + "epoch": 0.29578684109272796, + "grad_norm": 2.189158807672227, + "learning_rate": 1.651837505481947e-05, + "loss": 1.0011, + "step": 3075 + }, + { + "epoch": 0.29588303193535975, + "grad_norm": 2.558634501079498, + "learning_rate": 1.6516012130419366e-05, + "loss": 1.1692, + "step": 3076 + }, + { + "epoch": 0.29597922277799155, + "grad_norm": 2.1998237106778267, + "learning_rate": 1.651364857358418e-05, + "loss": 1.0357, + "step": 3077 + }, + { + "epoch": 0.29607541362062334, + "grad_norm": 2.106755700006773, + "learning_rate": 1.6511284384543317e-05, + "loss": 0.8949, + "step": 3078 + }, + { + "epoch": 0.2961716044632551, + "grad_norm": 1.946460864115865, + "learning_rate": 1.6508919563526245e-05, + "loss": 1.0937, + "step": 3079 + }, + { + "epoch": 0.29626779530588687, + "grad_norm": 2.0495901853024048, + "learning_rate": 1.6506554110762483e-05, + "loss": 1.0192, + "step": 3080 + }, + { + "epoch": 0.29636398614851867, + "grad_norm": 1.9598907472299287, + "learning_rate": 1.6504188026481625e-05, + "loss": 0.9281, + "step": 3081 + }, + { + "epoch": 0.29646017699115046, + "grad_norm": 2.415374434402126, + "learning_rate": 1.650182131091332e-05, + "loss": 1.122, + "step": 3082 + }, + { + "epoch": 0.2965563678337822, + "grad_norm": 2.132543700187211, + "learning_rate": 1.6499453964287273e-05, + "loss": 1.0339, + "step": 3083 + }, + { + "epoch": 0.296652558676414, + "grad_norm": 2.007495989733691, + "learning_rate": 1.6497085986833252e-05, + "loss": 1.014, + "step": 3084 + }, + { + "epoch": 0.2967487495190458, + "grad_norm": 1.9387536380136463, + "learning_rate": 1.64947173787811e-05, + "loss": 0.9934, + "step": 3085 + }, + { + "epoch": 0.2968449403616776, + "grad_norm": 2.190197619839409, + "learning_rate": 1.6492348140360704e-05, + "loss": 1.0153, + "step": 3086 + }, + { + "epoch": 0.29694113120430937, + "grad_norm": 2.2673836661342976, + "learning_rate": 1.6489978271802027e-05, + "loss": 0.984, + "step": 3087 + }, + { + "epoch": 0.2970373220469411, + "grad_norm": 2.0316261559628175, + "learning_rate": 1.6487607773335074e-05, + "loss": 1.047, + "step": 3088 + }, + { + "epoch": 0.2971335128895729, + "grad_norm": 2.0438122436411423, + "learning_rate": 1.648523664518993e-05, + "loss": 0.9875, + "step": 3089 + }, + { + "epoch": 0.2972297037322047, + "grad_norm": 1.9099468261448276, + "learning_rate": 1.648286488759673e-05, + "loss": 1.0007, + "step": 3090 + }, + { + "epoch": 0.2973258945748365, + "grad_norm": 2.431092948255771, + "learning_rate": 1.6480492500785673e-05, + "loss": 0.9985, + "step": 3091 + }, + { + "epoch": 0.29742208541746823, + "grad_norm": 2.0813287986433253, + "learning_rate": 1.6478119484987026e-05, + "loss": 0.9908, + "step": 3092 + }, + { + "epoch": 0.2975182762601, + "grad_norm": 1.9035172001921457, + "learning_rate": 1.64757458404311e-05, + "loss": 0.8567, + "step": 3093 + }, + { + "epoch": 0.2976144671027318, + "grad_norm": 2.3264399575386507, + "learning_rate": 1.6473371567348287e-05, + "loss": 1.037, + "step": 3094 + }, + { + "epoch": 0.2977106579453636, + "grad_norm": 1.924183578139566, + "learning_rate": 1.6470996665969027e-05, + "loss": 1.0349, + "step": 3095 + }, + { + "epoch": 0.2978068487879954, + "grad_norm": 3.038541719609843, + "learning_rate": 1.6468621136523823e-05, + "loss": 1.0813, + "step": 3096 + }, + { + "epoch": 0.29790303963062714, + "grad_norm": 2.1592700201246426, + "learning_rate": 1.6466244979243242e-05, + "loss": 0.9652, + "step": 3097 + }, + { + "epoch": 0.29799923047325894, + "grad_norm": 2.327781229041272, + "learning_rate": 1.646386819435791e-05, + "loss": 1.083, + "step": 3098 + }, + { + "epoch": 0.29809542131589073, + "grad_norm": 2.0386797243231967, + "learning_rate": 1.6461490782098518e-05, + "loss": 0.9355, + "step": 3099 + }, + { + "epoch": 0.2981916121585225, + "grad_norm": 2.1972850790875236, + "learning_rate": 1.6459112742695807e-05, + "loss": 0.9653, + "step": 3100 + }, + { + "epoch": 0.2982878030011543, + "grad_norm": 1.9334498872119068, + "learning_rate": 1.6456734076380597e-05, + "loss": 0.9956, + "step": 3101 + }, + { + "epoch": 0.29838399384378606, + "grad_norm": 1.9254533284346889, + "learning_rate": 1.6454354783383748e-05, + "loss": 0.8861, + "step": 3102 + }, + { + "epoch": 0.29848018468641785, + "grad_norm": 2.121022944695441, + "learning_rate": 1.6451974863936194e-05, + "loss": 0.9381, + "step": 3103 + }, + { + "epoch": 0.29857637552904964, + "grad_norm": 1.8764840046347848, + "learning_rate": 1.644959431826893e-05, + "loss": 1.0377, + "step": 3104 + }, + { + "epoch": 0.29867256637168144, + "grad_norm": 2.1092378612891802, + "learning_rate": 1.6447213146613e-05, + "loss": 1.0172, + "step": 3105 + }, + { + "epoch": 0.2987687572143132, + "grad_norm": 2.1219097602856123, + "learning_rate": 1.6444831349199528e-05, + "loss": 1.1041, + "step": 3106 + }, + { + "epoch": 0.29886494805694497, + "grad_norm": 2.4295668446796794, + "learning_rate": 1.6442448926259684e-05, + "loss": 0.9713, + "step": 3107 + }, + { + "epoch": 0.29896113889957676, + "grad_norm": 2.1606083436009293, + "learning_rate": 1.6440065878024697e-05, + "loss": 0.9971, + "step": 3108 + }, + { + "epoch": 0.29905732974220856, + "grad_norm": 2.1860648268846465, + "learning_rate": 1.6437682204725872e-05, + "loss": 1.0022, + "step": 3109 + }, + { + "epoch": 0.29915352058484035, + "grad_norm": 2.114764339444055, + "learning_rate": 1.6435297906594553e-05, + "loss": 1.0102, + "step": 3110 + }, + { + "epoch": 0.2992497114274721, + "grad_norm": 2.2831211598743373, + "learning_rate": 1.6432912983862168e-05, + "loss": 1.0057, + "step": 3111 + }, + { + "epoch": 0.2993459022701039, + "grad_norm": 1.9836492853798093, + "learning_rate": 1.643052743676019e-05, + "loss": 0.995, + "step": 3112 + }, + { + "epoch": 0.2994420931127357, + "grad_norm": 2.450888633730417, + "learning_rate": 1.6428141265520158e-05, + "loss": 1.0327, + "step": 3113 + }, + { + "epoch": 0.29953828395536747, + "grad_norm": 2.3380053090296267, + "learning_rate": 1.6425754470373667e-05, + "loss": 0.9849, + "step": 3114 + }, + { + "epoch": 0.2996344747979992, + "grad_norm": 9.847444009747212, + "learning_rate": 1.6423367051552376e-05, + "loss": 0.8846, + "step": 3115 + }, + { + "epoch": 0.299730665640631, + "grad_norm": 2.0244272744583025, + "learning_rate": 1.642097900928801e-05, + "loss": 1.0026, + "step": 3116 + }, + { + "epoch": 0.2998268564832628, + "grad_norm": 5.352020031098635, + "learning_rate": 1.6418590343812345e-05, + "loss": 0.9671, + "step": 3117 + }, + { + "epoch": 0.2999230473258946, + "grad_norm": 2.029517598247488, + "learning_rate": 1.6416201055357225e-05, + "loss": 0.9117, + "step": 3118 + }, + { + "epoch": 0.3000192381685264, + "grad_norm": 2.197860399391281, + "learning_rate": 1.6413811144154547e-05, + "loss": 1.0849, + "step": 3119 + }, + { + "epoch": 0.3001154290111581, + "grad_norm": 2.1261987836184546, + "learning_rate": 1.641142061043627e-05, + "loss": 1.0296, + "step": 3120 + }, + { + "epoch": 0.3002116198537899, + "grad_norm": 7.302632706408644, + "learning_rate": 1.6409029454434424e-05, + "loss": 0.8725, + "step": 3121 + }, + { + "epoch": 0.3003078106964217, + "grad_norm": 2.1274692989272004, + "learning_rate": 1.640663767638108e-05, + "loss": 1.1139, + "step": 3122 + }, + { + "epoch": 0.3004040015390535, + "grad_norm": 2.2871476773852217, + "learning_rate": 1.64042452765084e-05, + "loss": 0.9649, + "step": 3123 + }, + { + "epoch": 0.30050019238168524, + "grad_norm": 2.105575775158263, + "learning_rate": 1.6401852255048564e-05, + "loss": 1.0372, + "step": 3124 + }, + { + "epoch": 0.30059638322431703, + "grad_norm": 4.236727777337895, + "learning_rate": 1.6399458612233852e-05, + "loss": 0.8946, + "step": 3125 + }, + { + "epoch": 0.3006925740669488, + "grad_norm": 2.3449652877331895, + "learning_rate": 1.6397064348296578e-05, + "loss": 1.1208, + "step": 3126 + }, + { + "epoch": 0.3007887649095806, + "grad_norm": 3.1026101157270363, + "learning_rate": 1.639466946346913e-05, + "loss": 0.8879, + "step": 3127 + }, + { + "epoch": 0.3008849557522124, + "grad_norm": 2.275624800965071, + "learning_rate": 1.6392273957983955e-05, + "loss": 1.0895, + "step": 3128 + }, + { + "epoch": 0.30098114659484415, + "grad_norm": 2.3308060768790213, + "learning_rate": 1.6389877832073553e-05, + "loss": 1.0119, + "step": 3129 + }, + { + "epoch": 0.30107733743747594, + "grad_norm": 2.153590237472881, + "learning_rate": 1.638748108597049e-05, + "loss": 0.9964, + "step": 3130 + }, + { + "epoch": 0.30117352828010774, + "grad_norm": 2.200271794824055, + "learning_rate": 1.6385083719907387e-05, + "loss": 1.1308, + "step": 3131 + }, + { + "epoch": 0.30126971912273953, + "grad_norm": 2.044170838688308, + "learning_rate": 1.6382685734116934e-05, + "loss": 1.0597, + "step": 3132 + }, + { + "epoch": 0.30136590996537127, + "grad_norm": 2.1692852782807663, + "learning_rate": 1.638028712883188e-05, + "loss": 1.1745, + "step": 3133 + }, + { + "epoch": 0.30146210080800306, + "grad_norm": 1.9405106944686985, + "learning_rate": 1.6377887904285018e-05, + "loss": 1.0263, + "step": 3134 + }, + { + "epoch": 0.30155829165063486, + "grad_norm": 1.9911280121066033, + "learning_rate": 1.6375488060709226e-05, + "loss": 0.9998, + "step": 3135 + }, + { + "epoch": 0.30165448249326665, + "grad_norm": 1.8238202703725461, + "learning_rate": 1.637308759833742e-05, + "loss": 1.0685, + "step": 3136 + }, + { + "epoch": 0.30175067333589844, + "grad_norm": 2.1752563474644404, + "learning_rate": 1.6370686517402592e-05, + "loss": 1.0561, + "step": 3137 + }, + { + "epoch": 0.3018468641785302, + "grad_norm": 1.9236059298333084, + "learning_rate": 1.6368284818137787e-05, + "loss": 1.0836, + "step": 3138 + }, + { + "epoch": 0.301943055021162, + "grad_norm": 1.9613555419183835, + "learning_rate": 1.6365882500776103e-05, + "loss": 0.9344, + "step": 3139 + }, + { + "epoch": 0.30203924586379377, + "grad_norm": 38.45888504116879, + "learning_rate": 1.636347956555072e-05, + "loss": 1.1887, + "step": 3140 + }, + { + "epoch": 0.30213543670642556, + "grad_norm": 2.138584636677699, + "learning_rate": 1.636107601269485e-05, + "loss": 1.1142, + "step": 3141 + }, + { + "epoch": 0.3022316275490573, + "grad_norm": 2.3777216057796284, + "learning_rate": 1.635867184244178e-05, + "loss": 1.1361, + "step": 3142 + }, + { + "epoch": 0.3023278183916891, + "grad_norm": 29.204331375856654, + "learning_rate": 1.635626705502486e-05, + "loss": 1.1939, + "step": 3143 + }, + { + "epoch": 0.3024240092343209, + "grad_norm": 2.529637310638677, + "learning_rate": 1.63538616506775e-05, + "loss": 1.1386, + "step": 3144 + }, + { + "epoch": 0.3025202000769527, + "grad_norm": 2.2964555659638686, + "learning_rate": 1.6351455629633154e-05, + "loss": 1.0129, + "step": 3145 + }, + { + "epoch": 0.3026163909195845, + "grad_norm": 2.5092326388026036, + "learning_rate": 1.6349048992125358e-05, + "loss": 1.0457, + "step": 3146 + }, + { + "epoch": 0.3027125817622162, + "grad_norm": 3.301843773988406, + "learning_rate": 1.6346641738387686e-05, + "loss": 1.1102, + "step": 3147 + }, + { + "epoch": 0.302808772604848, + "grad_norm": 2.378427416155516, + "learning_rate": 1.634423386865379e-05, + "loss": 1.0584, + "step": 3148 + }, + { + "epoch": 0.3029049634474798, + "grad_norm": 2.420984975366674, + "learning_rate": 1.634182538315737e-05, + "loss": 1.1317, + "step": 3149 + }, + { + "epoch": 0.3030011542901116, + "grad_norm": 2.3101079803129907, + "learning_rate": 1.6339416282132196e-05, + "loss": 1.0751, + "step": 3150 + }, + { + "epoch": 0.3030973451327434, + "grad_norm": 2.2380299456341843, + "learning_rate": 1.633700656581209e-05, + "loss": 1.0742, + "step": 3151 + }, + { + "epoch": 0.30319353597537513, + "grad_norm": 2.418727861474331, + "learning_rate": 1.633459623443093e-05, + "loss": 1.1345, + "step": 3152 + }, + { + "epoch": 0.3032897268180069, + "grad_norm": 9.887355150935788, + "learning_rate": 1.6332185288222667e-05, + "loss": 1.2389, + "step": 3153 + }, + { + "epoch": 0.3033859176606387, + "grad_norm": 2.2837174999295127, + "learning_rate": 1.6329773727421297e-05, + "loss": 1.0576, + "step": 3154 + }, + { + "epoch": 0.3034821085032705, + "grad_norm": 2.3153775903325755, + "learning_rate": 1.6327361552260893e-05, + "loss": 1.108, + "step": 3155 + }, + { + "epoch": 0.30357829934590225, + "grad_norm": 2.2515317415415192, + "learning_rate": 1.6324948762975567e-05, + "loss": 0.963, + "step": 3156 + }, + { + "epoch": 0.30367449018853404, + "grad_norm": 2.1718225948515313, + "learning_rate": 1.6322535359799505e-05, + "loss": 1.0361, + "step": 3157 + }, + { + "epoch": 0.30377068103116583, + "grad_norm": 2.402148942432617, + "learning_rate": 1.632012134296695e-05, + "loss": 1.0894, + "step": 3158 + }, + { + "epoch": 0.30386687187379763, + "grad_norm": 2.5183039411432717, + "learning_rate": 1.63177067127122e-05, + "loss": 1.0619, + "step": 3159 + }, + { + "epoch": 0.3039630627164294, + "grad_norm": 2.267341538567709, + "learning_rate": 1.6315291469269617e-05, + "loss": 1.0624, + "step": 3160 + }, + { + "epoch": 0.30405925355906116, + "grad_norm": 2.0673771732798394, + "learning_rate": 1.6312875612873626e-05, + "loss": 1.0302, + "step": 3161 + }, + { + "epoch": 0.30415544440169295, + "grad_norm": 1.933772901780693, + "learning_rate": 1.63104591437587e-05, + "loss": 1.0181, + "step": 3162 + }, + { + "epoch": 0.30425163524432475, + "grad_norm": 2.484661668113306, + "learning_rate": 1.6308042062159376e-05, + "loss": 0.9809, + "step": 3163 + }, + { + "epoch": 0.30434782608695654, + "grad_norm": 2.4398046898028114, + "learning_rate": 1.6305624368310265e-05, + "loss": 1.0635, + "step": 3164 + }, + { + "epoch": 0.3044440169295883, + "grad_norm": 1.9208123433933968, + "learning_rate": 1.630320606244601e-05, + "loss": 1.0261, + "step": 3165 + }, + { + "epoch": 0.3045402077722201, + "grad_norm": 2.1874800568110344, + "learning_rate": 1.630078714480134e-05, + "loss": 1.0482, + "step": 3166 + }, + { + "epoch": 0.30463639861485187, + "grad_norm": 2.1887637090745398, + "learning_rate": 1.6298367615611026e-05, + "loss": 1.1186, + "step": 3167 + }, + { + "epoch": 0.30473258945748366, + "grad_norm": 2.1579469561107065, + "learning_rate": 1.6295947475109904e-05, + "loss": 1.1217, + "step": 3168 + }, + { + "epoch": 0.30482878030011545, + "grad_norm": 2.399032423881041, + "learning_rate": 1.6293526723532867e-05, + "loss": 1.0139, + "step": 3169 + }, + { + "epoch": 0.3049249711427472, + "grad_norm": 2.265800178195505, + "learning_rate": 1.629110536111488e-05, + "loss": 1.0705, + "step": 3170 + }, + { + "epoch": 0.305021161985379, + "grad_norm": 2.0929275428498375, + "learning_rate": 1.628868338809095e-05, + "loss": 1.0651, + "step": 3171 + }, + { + "epoch": 0.3051173528280108, + "grad_norm": 2.018112247350248, + "learning_rate": 1.628626080469615e-05, + "loss": 0.9959, + "step": 3172 + }, + { + "epoch": 0.3052135436706426, + "grad_norm": 2.160356207776233, + "learning_rate": 1.6283837611165614e-05, + "loss": 1.1158, + "step": 3173 + }, + { + "epoch": 0.3053097345132743, + "grad_norm": 2.104689737364837, + "learning_rate": 1.628141380773453e-05, + "loss": 1.1826, + "step": 3174 + }, + { + "epoch": 0.3054059253559061, + "grad_norm": 2.115800464008481, + "learning_rate": 1.6278989394638157e-05, + "loss": 0.9967, + "step": 3175 + }, + { + "epoch": 0.3055021161985379, + "grad_norm": 2.1899303999153847, + "learning_rate": 1.6276564372111797e-05, + "loss": 1.0436, + "step": 3176 + }, + { + "epoch": 0.3055983070411697, + "grad_norm": 2.476827122037033, + "learning_rate": 1.6274138740390823e-05, + "loss": 1.1351, + "step": 3177 + }, + { + "epoch": 0.3056944978838015, + "grad_norm": 2.216451482663654, + "learning_rate": 1.6271712499710663e-05, + "loss": 1.1186, + "step": 3178 + }, + { + "epoch": 0.3057906887264332, + "grad_norm": 2.06099052769235, + "learning_rate": 1.6269285650306807e-05, + "loss": 1.0607, + "step": 3179 + }, + { + "epoch": 0.305886879569065, + "grad_norm": 2.1555572102017257, + "learning_rate": 1.62668581924148e-05, + "loss": 0.9586, + "step": 3180 + }, + { + "epoch": 0.3059830704116968, + "grad_norm": 1.8864927266215714, + "learning_rate": 1.6264430126270246e-05, + "loss": 0.941, + "step": 3181 + }, + { + "epoch": 0.3060792612543286, + "grad_norm": 2.194670765830389, + "learning_rate": 1.6262001452108807e-05, + "loss": 0.9515, + "step": 3182 + }, + { + "epoch": 0.30617545209696034, + "grad_norm": 2.099374767830327, + "learning_rate": 1.6259572170166215e-05, + "loss": 1.0625, + "step": 3183 + }, + { + "epoch": 0.30627164293959214, + "grad_norm": 2.221123291533065, + "learning_rate": 1.6257142280678247e-05, + "loss": 0.9263, + "step": 3184 + }, + { + "epoch": 0.30636783378222393, + "grad_norm": 2.167412639188574, + "learning_rate": 1.6254711783880744e-05, + "loss": 1.0564, + "step": 3185 + }, + { + "epoch": 0.3064640246248557, + "grad_norm": 2.0030527594961685, + "learning_rate": 1.6252280680009613e-05, + "loss": 0.9796, + "step": 3186 + }, + { + "epoch": 0.3065602154674875, + "grad_norm": 2.4311798904776003, + "learning_rate": 1.6249848969300804e-05, + "loss": 1.1375, + "step": 3187 + }, + { + "epoch": 0.30665640631011926, + "grad_norm": 2.221077607588961, + "learning_rate": 1.6247416651990343e-05, + "loss": 1.0341, + "step": 3188 + }, + { + "epoch": 0.30675259715275105, + "grad_norm": 2.050588547981524, + "learning_rate": 1.6244983728314308e-05, + "loss": 1.0072, + "step": 3189 + }, + { + "epoch": 0.30684878799538284, + "grad_norm": 2.1101907136262286, + "learning_rate": 1.624255019850883e-05, + "loss": 1.1181, + "step": 3190 + }, + { + "epoch": 0.30694497883801464, + "grad_norm": 2.2469182035986757, + "learning_rate": 1.6240116062810103e-05, + "loss": 1.0716, + "step": 3191 + }, + { + "epoch": 0.3070411696806464, + "grad_norm": 2.202144731951469, + "learning_rate": 1.6237681321454387e-05, + "loss": 1.0545, + "step": 3192 + }, + { + "epoch": 0.30713736052327817, + "grad_norm": 2.003711719817428, + "learning_rate": 1.6235245974677997e-05, + "loss": 1.025, + "step": 3193 + }, + { + "epoch": 0.30723355136590996, + "grad_norm": 2.898134815745158, + "learning_rate": 1.623281002271729e-05, + "loss": 0.9967, + "step": 3194 + }, + { + "epoch": 0.30732974220854176, + "grad_norm": 2.0737763766863475, + "learning_rate": 1.623037346580871e-05, + "loss": 1.0043, + "step": 3195 + }, + { + "epoch": 0.30742593305117355, + "grad_norm": 2.2946007685816916, + "learning_rate": 1.6227936304188738e-05, + "loss": 1.0385, + "step": 3196 + }, + { + "epoch": 0.3075221238938053, + "grad_norm": 2.33717128451612, + "learning_rate": 1.6225498538093926e-05, + "loss": 1.1356, + "step": 3197 + }, + { + "epoch": 0.3076183147364371, + "grad_norm": 1.9538551526914703, + "learning_rate": 1.622306016776088e-05, + "loss": 1.0986, + "step": 3198 + }, + { + "epoch": 0.3077145055790689, + "grad_norm": 2.1816755870952513, + "learning_rate": 1.622062119342626e-05, + "loss": 1.061, + "step": 3199 + }, + { + "epoch": 0.30781069642170067, + "grad_norm": 2.133138214982213, + "learning_rate": 1.6218181615326795e-05, + "loss": 1.0946, + "step": 3200 + }, + { + "epoch": 0.30790688726433246, + "grad_norm": 2.1366039291270664, + "learning_rate": 1.6215741433699267e-05, + "loss": 1.1028, + "step": 3201 + }, + { + "epoch": 0.3080030781069642, + "grad_norm": 1.9114636610534865, + "learning_rate": 1.6213300648780515e-05, + "loss": 0.9565, + "step": 3202 + }, + { + "epoch": 0.308099268949596, + "grad_norm": 2.0080132406663727, + "learning_rate": 1.6210859260807433e-05, + "loss": 0.9727, + "step": 3203 + }, + { + "epoch": 0.3081954597922278, + "grad_norm": 2.191978599568679, + "learning_rate": 1.620841727001699e-05, + "loss": 0.9854, + "step": 3204 + }, + { + "epoch": 0.3082916506348596, + "grad_norm": 2.1135565187663263, + "learning_rate": 1.620597467664619e-05, + "loss": 0.9716, + "step": 3205 + }, + { + "epoch": 0.3083878414774913, + "grad_norm": 2.0058890549776707, + "learning_rate": 1.6203531480932114e-05, + "loss": 1.0234, + "step": 3206 + }, + { + "epoch": 0.3084840323201231, + "grad_norm": 2.2202040284113083, + "learning_rate": 1.62010876831119e-05, + "loss": 1.0197, + "step": 3207 + }, + { + "epoch": 0.3085802231627549, + "grad_norm": 1.9060833300237043, + "learning_rate": 1.619864328342273e-05, + "loss": 1.0589, + "step": 3208 + }, + { + "epoch": 0.3086764140053867, + "grad_norm": 2.1472813529082546, + "learning_rate": 1.6196198282101857e-05, + "loss": 1.0239, + "step": 3209 + }, + { + "epoch": 0.3087726048480185, + "grad_norm": 2.0146840437931877, + "learning_rate": 1.6193752679386593e-05, + "loss": 0.987, + "step": 3210 + }, + { + "epoch": 0.30886879569065023, + "grad_norm": 1.9466968650653451, + "learning_rate": 1.6191306475514303e-05, + "loss": 1.0331, + "step": 3211 + }, + { + "epoch": 0.308964986533282, + "grad_norm": 2.2801642663645216, + "learning_rate": 1.6188859670722414e-05, + "loss": 0.9269, + "step": 3212 + }, + { + "epoch": 0.3090611773759138, + "grad_norm": 1.999001945852578, + "learning_rate": 1.6186412265248402e-05, + "loss": 0.9963, + "step": 3213 + }, + { + "epoch": 0.3091573682185456, + "grad_norm": 2.420384519048038, + "learning_rate": 1.6183964259329817e-05, + "loss": 1.0204, + "step": 3214 + }, + { + "epoch": 0.30925355906117735, + "grad_norm": 2.042670471439237, + "learning_rate": 1.6181515653204255e-05, + "loss": 1.0399, + "step": 3215 + }, + { + "epoch": 0.30934974990380915, + "grad_norm": 2.0422477155739176, + "learning_rate": 1.6179066447109376e-05, + "loss": 0.993, + "step": 3216 + }, + { + "epoch": 0.30944594074644094, + "grad_norm": 2.138640716542524, + "learning_rate": 1.6176616641282895e-05, + "loss": 0.9274, + "step": 3217 + }, + { + "epoch": 0.30954213158907273, + "grad_norm": 2.101333097624342, + "learning_rate": 1.6174166235962588e-05, + "loss": 1.0557, + "step": 3218 + }, + { + "epoch": 0.3096383224317045, + "grad_norm": 2.128589872677615, + "learning_rate": 1.6171715231386288e-05, + "loss": 1.046, + "step": 3219 + }, + { + "epoch": 0.30973451327433627, + "grad_norm": 2.283574889388088, + "learning_rate": 1.6169263627791886e-05, + "loss": 1.0911, + "step": 3220 + }, + { + "epoch": 0.30983070411696806, + "grad_norm": 2.1787349016047424, + "learning_rate": 1.616681142541733e-05, + "loss": 1.0137, + "step": 3221 + }, + { + "epoch": 0.30992689495959985, + "grad_norm": 20.017593727776347, + "learning_rate": 1.616435862450063e-05, + "loss": 1.8532, + "step": 3222 + }, + { + "epoch": 0.31002308580223165, + "grad_norm": 1.9626552981651533, + "learning_rate": 1.6161905225279846e-05, + "loss": 1.1192, + "step": 3223 + }, + { + "epoch": 0.3101192766448634, + "grad_norm": 2.2155601199539174, + "learning_rate": 1.615945122799311e-05, + "loss": 1.0557, + "step": 3224 + }, + { + "epoch": 0.3102154674874952, + "grad_norm": 2.032220632389221, + "learning_rate": 1.6156996632878597e-05, + "loss": 1.0647, + "step": 3225 + }, + { + "epoch": 0.31031165833012697, + "grad_norm": 2.0788324188512335, + "learning_rate": 1.6154541440174547e-05, + "loss": 1.0492, + "step": 3226 + }, + { + "epoch": 0.31040784917275877, + "grad_norm": 24.110876706968977, + "learning_rate": 1.615208565011926e-05, + "loss": 2.8816, + "step": 3227 + }, + { + "epoch": 0.31050404001539056, + "grad_norm": 12.330458054497445, + "learning_rate": 1.614962926295109e-05, + "loss": 1.9509, + "step": 3228 + }, + { + "epoch": 0.3106002308580223, + "grad_norm": 20.77888566571293, + "learning_rate": 1.614717227890845e-05, + "loss": 1.8198, + "step": 3229 + }, + { + "epoch": 0.3106964217006541, + "grad_norm": 3.007095962710836, + "learning_rate": 1.6144714698229814e-05, + "loss": 1.0309, + "step": 3230 + }, + { + "epoch": 0.3107926125432859, + "grad_norm": 2.0867178595080516, + "learning_rate": 1.614225652115371e-05, + "loss": 0.9459, + "step": 3231 + }, + { + "epoch": 0.3108888033859177, + "grad_norm": 2.122837999238009, + "learning_rate": 1.6139797747918725e-05, + "loss": 1.0269, + "step": 3232 + }, + { + "epoch": 0.3109849942285494, + "grad_norm": 2.7826404563326386, + "learning_rate": 1.6137338378763504e-05, + "loss": 1.0558, + "step": 3233 + }, + { + "epoch": 0.3110811850711812, + "grad_norm": 3.075296518686262, + "learning_rate": 1.613487841392675e-05, + "loss": 1.0482, + "step": 3234 + }, + { + "epoch": 0.311177375913813, + "grad_norm": 3.4933887410776365, + "learning_rate": 1.6132417853647226e-05, + "loss": 1.0681, + "step": 3235 + }, + { + "epoch": 0.3112735667564448, + "grad_norm": 2.5993806361686485, + "learning_rate": 1.612995669816375e-05, + "loss": 0.9443, + "step": 3236 + }, + { + "epoch": 0.3113697575990766, + "grad_norm": 2.396229586104887, + "learning_rate": 1.612749494771519e-05, + "loss": 1.0368, + "step": 3237 + }, + { + "epoch": 0.31146594844170833, + "grad_norm": 2.2260710079467216, + "learning_rate": 1.6125032602540492e-05, + "loss": 1.0058, + "step": 3238 + }, + { + "epoch": 0.3115621392843401, + "grad_norm": 2.128173362791873, + "learning_rate": 1.612256966287864e-05, + "loss": 0.8532, + "step": 3239 + }, + { + "epoch": 0.3116583301269719, + "grad_norm": 2.8115493928055506, + "learning_rate": 1.6120106128968686e-05, + "loss": 1.1043, + "step": 3240 + }, + { + "epoch": 0.3117545209696037, + "grad_norm": 2.5024065809833123, + "learning_rate": 1.611764200104974e-05, + "loss": 1.038, + "step": 3241 + }, + { + "epoch": 0.31185071181223545, + "grad_norm": 2.1430316628556647, + "learning_rate": 1.6115177279360965e-05, + "loss": 0.9842, + "step": 3242 + }, + { + "epoch": 0.31194690265486724, + "grad_norm": 1.9950766180856063, + "learning_rate": 1.6112711964141577e-05, + "loss": 1.087, + "step": 3243 + }, + { + "epoch": 0.31204309349749904, + "grad_norm": 2.2294480777881716, + "learning_rate": 1.611024605563087e-05, + "loss": 0.9256, + "step": 3244 + }, + { + "epoch": 0.31213928434013083, + "grad_norm": 2.7282219244687544, + "learning_rate": 1.610777955406817e-05, + "loss": 0.9935, + "step": 3245 + }, + { + "epoch": 0.3122354751827626, + "grad_norm": 2.6661278570948572, + "learning_rate": 1.610531245969287e-05, + "loss": 1.0536, + "step": 3246 + }, + { + "epoch": 0.31233166602539436, + "grad_norm": 1.8852790410757394, + "learning_rate": 1.6102844772744436e-05, + "loss": 0.9773, + "step": 3247 + }, + { + "epoch": 0.31242785686802615, + "grad_norm": 2.246178647622235, + "learning_rate": 1.6100376493462368e-05, + "loss": 1.0263, + "step": 3248 + }, + { + "epoch": 0.31252404771065795, + "grad_norm": 2.0523839635576557, + "learning_rate": 1.6097907622086238e-05, + "loss": 1.1102, + "step": 3249 + }, + { + "epoch": 0.31262023855328974, + "grad_norm": 2.2936213138084964, + "learning_rate": 1.6095438158855668e-05, + "loss": 1.1349, + "step": 3250 + }, + { + "epoch": 0.31271642939592154, + "grad_norm": 1.849264711615071, + "learning_rate": 1.609296810401034e-05, + "loss": 0.934, + "step": 3251 + }, + { + "epoch": 0.3128126202385533, + "grad_norm": 2.326775144032454, + "learning_rate": 1.609049745779e-05, + "loss": 1.0678, + "step": 3252 + }, + { + "epoch": 0.31290881108118507, + "grad_norm": 2.500419864341823, + "learning_rate": 1.608802622043444e-05, + "loss": 0.994, + "step": 3253 + }, + { + "epoch": 0.31300500192381686, + "grad_norm": 2.533791514045582, + "learning_rate": 1.6085554392183517e-05, + "loss": 1.0718, + "step": 3254 + }, + { + "epoch": 0.31310119276644865, + "grad_norm": 2.2547621960040494, + "learning_rate": 1.6083081973277143e-05, + "loss": 0.9563, + "step": 3255 + }, + { + "epoch": 0.3131973836090804, + "grad_norm": 2.463727102486478, + "learning_rate": 1.608060896395529e-05, + "loss": 1.0059, + "step": 3256 + }, + { + "epoch": 0.3132935744517122, + "grad_norm": 2.0519811392042246, + "learning_rate": 1.607813536445798e-05, + "loss": 0.9701, + "step": 3257 + }, + { + "epoch": 0.313389765294344, + "grad_norm": 2.0546694132842376, + "learning_rate": 1.60756611750253e-05, + "loss": 0.9963, + "step": 3258 + }, + { + "epoch": 0.3134859561369758, + "grad_norm": 2.46186478263281, + "learning_rate": 1.6073186395897388e-05, + "loss": 1.1641, + "step": 3259 + }, + { + "epoch": 0.31358214697960757, + "grad_norm": 2.0455459194986885, + "learning_rate": 1.6070711027314446e-05, + "loss": 1.0371, + "step": 3260 + }, + { + "epoch": 0.3136783378222393, + "grad_norm": 1.96349462469952, + "learning_rate": 1.6068235069516735e-05, + "loss": 1.0743, + "step": 3261 + }, + { + "epoch": 0.3137745286648711, + "grad_norm": 1.9877532806121103, + "learning_rate": 1.606575852274456e-05, + "loss": 1.0335, + "step": 3262 + }, + { + "epoch": 0.3138707195075029, + "grad_norm": 1.9794257160011577, + "learning_rate": 1.606328138723829e-05, + "loss": 1.1301, + "step": 3263 + }, + { + "epoch": 0.3139669103501347, + "grad_norm": 1.9843871267853834, + "learning_rate": 1.6060803663238357e-05, + "loss": 1.0249, + "step": 3264 + }, + { + "epoch": 0.3140631011927664, + "grad_norm": 1.9742171729052018, + "learning_rate": 1.6058325350985246e-05, + "loss": 1.0497, + "step": 3265 + }, + { + "epoch": 0.3141592920353982, + "grad_norm": 1.8912835807006347, + "learning_rate": 1.6055846450719498e-05, + "loss": 1.0515, + "step": 3266 + }, + { + "epoch": 0.31425548287803, + "grad_norm": 2.0100878157018363, + "learning_rate": 1.6053366962681708e-05, + "loss": 1.0067, + "step": 3267 + }, + { + "epoch": 0.3143516737206618, + "grad_norm": 2.4021663409071636, + "learning_rate": 1.6050886887112535e-05, + "loss": 1.0946, + "step": 3268 + }, + { + "epoch": 0.3144478645632936, + "grad_norm": 2.3395832566996684, + "learning_rate": 1.6048406224252688e-05, + "loss": 1.0851, + "step": 3269 + }, + { + "epoch": 0.31454405540592534, + "grad_norm": 2.314426801062467, + "learning_rate": 1.6045924974342945e-05, + "loss": 1.0143, + "step": 3270 + }, + { + "epoch": 0.31464024624855713, + "grad_norm": 2.0187406756787847, + "learning_rate": 1.604344313762412e-05, + "loss": 0.8891, + "step": 3271 + }, + { + "epoch": 0.3147364370911889, + "grad_norm": 1.9979126653256971, + "learning_rate": 1.604096071433711e-05, + "loss": 1.1476, + "step": 3272 + }, + { + "epoch": 0.3148326279338207, + "grad_norm": 2.060408068045893, + "learning_rate": 1.6038477704722846e-05, + "loss": 0.9652, + "step": 3273 + }, + { + "epoch": 0.31492881877645246, + "grad_norm": 2.0702639143548387, + "learning_rate": 1.6035994109022333e-05, + "loss": 1.011, + "step": 3274 + }, + { + "epoch": 0.31502500961908425, + "grad_norm": 1.9575548189421232, + "learning_rate": 1.6033509927476618e-05, + "loss": 0.9938, + "step": 3275 + }, + { + "epoch": 0.31512120046171604, + "grad_norm": 2.413255246030864, + "learning_rate": 1.6031025160326814e-05, + "loss": 1.0913, + "step": 3276 + }, + { + "epoch": 0.31521739130434784, + "grad_norm": 1.922887482131897, + "learning_rate": 1.6028539807814093e-05, + "loss": 1.0096, + "step": 3277 + }, + { + "epoch": 0.31531358214697963, + "grad_norm": 1.8920203532929192, + "learning_rate": 1.6026053870179678e-05, + "loss": 0.9671, + "step": 3278 + }, + { + "epoch": 0.31540977298961137, + "grad_norm": 2.0087592953753766, + "learning_rate": 1.602356734766485e-05, + "loss": 1.0227, + "step": 3279 + }, + { + "epoch": 0.31550596383224316, + "grad_norm": 14.305715560455033, + "learning_rate": 1.6021080240510943e-05, + "loss": 1.5883, + "step": 3280 + }, + { + "epoch": 0.31560215467487496, + "grad_norm": 2.0421319286828177, + "learning_rate": 1.6018592548959357e-05, + "loss": 1.1149, + "step": 3281 + }, + { + "epoch": 0.31569834551750675, + "grad_norm": 2.3307625589269274, + "learning_rate": 1.601610427325155e-05, + "loss": 1.111, + "step": 3282 + }, + { + "epoch": 0.3157945363601385, + "grad_norm": 2.6450301044948565, + "learning_rate": 1.6013615413629017e-05, + "loss": 0.9882, + "step": 3283 + }, + { + "epoch": 0.3158907272027703, + "grad_norm": 3.468239860239652, + "learning_rate": 1.6011125970333333e-05, + "loss": 1.0153, + "step": 3284 + }, + { + "epoch": 0.3159869180454021, + "grad_norm": 19.36057658015786, + "learning_rate": 1.6008635943606118e-05, + "loss": 1.2238, + "step": 3285 + }, + { + "epoch": 0.31608310888803387, + "grad_norm": 2.3370145921283174, + "learning_rate": 1.600614533368905e-05, + "loss": 1.0437, + "step": 3286 + }, + { + "epoch": 0.31617929973066566, + "grad_norm": 3.0805915655580827, + "learning_rate": 1.6003654140823857e-05, + "loss": 1.0533, + "step": 3287 + }, + { + "epoch": 0.3162754905732974, + "grad_norm": 2.2317130881898652, + "learning_rate": 1.6001162365252348e-05, + "loss": 0.8081, + "step": 3288 + }, + { + "epoch": 0.3163716814159292, + "grad_norm": 10.25056643688534, + "learning_rate": 1.5998670007216356e-05, + "loss": 1.388, + "step": 3289 + }, + { + "epoch": 0.316467872258561, + "grad_norm": 2.605292587641024, + "learning_rate": 1.5996177066957787e-05, + "loss": 1.0473, + "step": 3290 + }, + { + "epoch": 0.3165640631011928, + "grad_norm": 2.042116792671027, + "learning_rate": 1.599368354471861e-05, + "loss": 0.9044, + "step": 3291 + }, + { + "epoch": 0.3166602539438245, + "grad_norm": 2.4683816596433377, + "learning_rate": 1.5991189440740838e-05, + "loss": 1.0323, + "step": 3292 + }, + { + "epoch": 0.3167564447864563, + "grad_norm": 2.547494495486264, + "learning_rate": 1.5988694755266548e-05, + "loss": 1.0885, + "step": 3293 + }, + { + "epoch": 0.3168526356290881, + "grad_norm": 2.578005948186227, + "learning_rate": 1.5986199488537867e-05, + "loss": 1.0033, + "step": 3294 + }, + { + "epoch": 0.3169488264717199, + "grad_norm": 2.5247108510870766, + "learning_rate": 1.598370364079698e-05, + "loss": 1.0243, + "step": 3295 + }, + { + "epoch": 0.3170450173143517, + "grad_norm": 2.2514458236436345, + "learning_rate": 1.598120721228614e-05, + "loss": 1.0082, + "step": 3296 + }, + { + "epoch": 0.31714120815698343, + "grad_norm": 16.838127934704332, + "learning_rate": 1.5978710203247637e-05, + "loss": 1.1411, + "step": 3297 + }, + { + "epoch": 0.3172373989996152, + "grad_norm": 2.3312741143330684, + "learning_rate": 1.5976212613923836e-05, + "loss": 1.0511, + "step": 3298 + }, + { + "epoch": 0.317333589842247, + "grad_norm": 2.404580940074039, + "learning_rate": 1.5973714444557146e-05, + "loss": 1.1219, + "step": 3299 + }, + { + "epoch": 0.3174297806848788, + "grad_norm": 2.45666080015282, + "learning_rate": 1.5971215695390026e-05, + "loss": 1.0987, + "step": 3300 + }, + { + "epoch": 0.3175259715275106, + "grad_norm": 2.458149665496898, + "learning_rate": 1.5968716366665017e-05, + "loss": 1.1079, + "step": 3301 + }, + { + "epoch": 0.31762216237014235, + "grad_norm": 2.4020159048837106, + "learning_rate": 1.5966216458624692e-05, + "loss": 0.9795, + "step": 3302 + }, + { + "epoch": 0.31771835321277414, + "grad_norm": 2.3991901084248974, + "learning_rate": 1.596371597151169e-05, + "loss": 1.0247, + "step": 3303 + }, + { + "epoch": 0.31781454405540593, + "grad_norm": 2.2623621408363217, + "learning_rate": 1.5961214905568705e-05, + "loss": 1.0932, + "step": 3304 + }, + { + "epoch": 0.31791073489803773, + "grad_norm": 2.2404748991348127, + "learning_rate": 1.5958713261038487e-05, + "loss": 1.0357, + "step": 3305 + }, + { + "epoch": 0.31800692574066947, + "grad_norm": 2.3988686286299585, + "learning_rate": 1.595621103816384e-05, + "loss": 0.9507, + "step": 3306 + }, + { + "epoch": 0.31810311658330126, + "grad_norm": 2.4845666321838786, + "learning_rate": 1.5953708237187627e-05, + "loss": 0.8923, + "step": 3307 + }, + { + "epoch": 0.31819930742593305, + "grad_norm": 2.0303690265493177, + "learning_rate": 1.5951204858352772e-05, + "loss": 0.916, + "step": 3308 + }, + { + "epoch": 0.31829549826856485, + "grad_norm": 1.889634538799101, + "learning_rate": 1.5948700901902243e-05, + "loss": 0.9469, + "step": 3309 + }, + { + "epoch": 0.31839168911119664, + "grad_norm": 2.3031345290496534, + "learning_rate": 1.594619636807907e-05, + "loss": 1.1092, + "step": 3310 + }, + { + "epoch": 0.3184878799538284, + "grad_norm": 2.311355076576151, + "learning_rate": 1.5943691257126345e-05, + "loss": 0.8819, + "step": 3311 + }, + { + "epoch": 0.3185840707964602, + "grad_norm": 2.342590773585347, + "learning_rate": 1.5941185569287206e-05, + "loss": 1.0824, + "step": 3312 + }, + { + "epoch": 0.31868026163909197, + "grad_norm": 2.7170609269221586, + "learning_rate": 1.5938679304804853e-05, + "loss": 1.1456, + "step": 3313 + }, + { + "epoch": 0.31877645248172376, + "grad_norm": 2.1912548988537104, + "learning_rate": 1.5936172463922542e-05, + "loss": 1.0123, + "step": 3314 + }, + { + "epoch": 0.3188726433243555, + "grad_norm": 2.371885717512359, + "learning_rate": 1.5933665046883582e-05, + "loss": 0.9803, + "step": 3315 + }, + { + "epoch": 0.3189688341669873, + "grad_norm": 15.958475461827545, + "learning_rate": 1.593115705393134e-05, + "loss": 1.6313, + "step": 3316 + }, + { + "epoch": 0.3190650250096191, + "grad_norm": 2.1297468156085, + "learning_rate": 1.5928648485309238e-05, + "loss": 0.9661, + "step": 3317 + }, + { + "epoch": 0.3191612158522509, + "grad_norm": 1.957618834242923, + "learning_rate": 1.5926139341260755e-05, + "loss": 1.0516, + "step": 3318 + }, + { + "epoch": 0.3192574066948827, + "grad_norm": 2.0561701299933826, + "learning_rate": 1.5923629622029423e-05, + "loss": 1.0094, + "step": 3319 + }, + { + "epoch": 0.3193535975375144, + "grad_norm": 2.1967691432926197, + "learning_rate": 1.5921119327858835e-05, + "loss": 1.0569, + "step": 3320 + }, + { + "epoch": 0.3194497883801462, + "grad_norm": 2.031774929643997, + "learning_rate": 1.591860845899263e-05, + "loss": 1.0606, + "step": 3321 + }, + { + "epoch": 0.319545979222778, + "grad_norm": 2.1121386480054745, + "learning_rate": 1.5916097015674518e-05, + "loss": 1.042, + "step": 3322 + }, + { + "epoch": 0.3196421700654098, + "grad_norm": 2.109854400251898, + "learning_rate": 1.5913584998148252e-05, + "loss": 1.0016, + "step": 3323 + }, + { + "epoch": 0.31973836090804153, + "grad_norm": 2.070202799915352, + "learning_rate": 1.5911072406657646e-05, + "loss": 0.9515, + "step": 3324 + }, + { + "epoch": 0.3198345517506733, + "grad_norm": 2.3387968809161235, + "learning_rate": 1.590855924144657e-05, + "loss": 1.1641, + "step": 3325 + }, + { + "epoch": 0.3199307425933051, + "grad_norm": 2.0349758193260787, + "learning_rate": 1.5906045502758943e-05, + "loss": 1.0743, + "step": 3326 + }, + { + "epoch": 0.3200269334359369, + "grad_norm": 2.354455664236324, + "learning_rate": 1.5903531190838745e-05, + "loss": 1.0685, + "step": 3327 + }, + { + "epoch": 0.3201231242785687, + "grad_norm": 2.1722686321041746, + "learning_rate": 1.590101630593002e-05, + "loss": 1.1003, + "step": 3328 + }, + { + "epoch": 0.32021931512120044, + "grad_norm": 2.267795294623865, + "learning_rate": 1.5898500848276853e-05, + "loss": 0.8971, + "step": 3329 + }, + { + "epoch": 0.32031550596383224, + "grad_norm": 15.86829460496691, + "learning_rate": 1.5895984818123392e-05, + "loss": 1.3478, + "step": 3330 + }, + { + "epoch": 0.32041169680646403, + "grad_norm": 1.9311493279480751, + "learning_rate": 1.589346821571384e-05, + "loss": 0.9552, + "step": 3331 + }, + { + "epoch": 0.3205078876490958, + "grad_norm": 1.9596425630443683, + "learning_rate": 1.5890951041292453e-05, + "loss": 1.1288, + "step": 3332 + }, + { + "epoch": 0.32060407849172756, + "grad_norm": 2.1883975990962172, + "learning_rate": 1.5888433295103543e-05, + "loss": 1.0709, + "step": 3333 + }, + { + "epoch": 0.32070026933435936, + "grad_norm": 2.2574805323898866, + "learning_rate": 1.588591497739149e-05, + "loss": 0.9746, + "step": 3334 + }, + { + "epoch": 0.32079646017699115, + "grad_norm": 2.0435911987718764, + "learning_rate": 1.58833960884007e-05, + "loss": 0.9854, + "step": 3335 + }, + { + "epoch": 0.32089265101962294, + "grad_norm": 2.1744877312694135, + "learning_rate": 1.5880876628375668e-05, + "loss": 1.0636, + "step": 3336 + }, + { + "epoch": 0.32098884186225474, + "grad_norm": 2.2698504235745083, + "learning_rate": 1.5878356597560924e-05, + "loss": 0.991, + "step": 3337 + }, + { + "epoch": 0.3210850327048865, + "grad_norm": 2.0978082525537673, + "learning_rate": 1.587583599620106e-05, + "loss": 0.8711, + "step": 3338 + }, + { + "epoch": 0.32118122354751827, + "grad_norm": 1.8280233112084436, + "learning_rate": 1.5873314824540717e-05, + "loss": 0.8848, + "step": 3339 + }, + { + "epoch": 0.32127741439015006, + "grad_norm": 2.5573267729317575, + "learning_rate": 1.5870793082824604e-05, + "loss": 1.0585, + "step": 3340 + }, + { + "epoch": 0.32137360523278186, + "grad_norm": 2.008094072958577, + "learning_rate": 1.5868270771297475e-05, + "loss": 1.059, + "step": 3341 + }, + { + "epoch": 0.3214697960754136, + "grad_norm": 2.5123796044971045, + "learning_rate": 1.5865747890204138e-05, + "loss": 1.0784, + "step": 3342 + }, + { + "epoch": 0.3215659869180454, + "grad_norm": 2.087565900057572, + "learning_rate": 1.5863224439789467e-05, + "loss": 1.0669, + "step": 3343 + }, + { + "epoch": 0.3216621777606772, + "grad_norm": 2.2661893281917562, + "learning_rate": 1.5860700420298377e-05, + "loss": 1.0923, + "step": 3344 + }, + { + "epoch": 0.321758368603309, + "grad_norm": 2.31409764509966, + "learning_rate": 1.5858175831975854e-05, + "loss": 1.067, + "step": 3345 + }, + { + "epoch": 0.32185455944594077, + "grad_norm": 2.6633040605443927, + "learning_rate": 1.5855650675066924e-05, + "loss": 0.9271, + "step": 3346 + }, + { + "epoch": 0.3219507502885725, + "grad_norm": 2.056268213157848, + "learning_rate": 1.585312494981668e-05, + "loss": 1.0376, + "step": 3347 + }, + { + "epoch": 0.3220469411312043, + "grad_norm": 1.9228093575123941, + "learning_rate": 1.5850598656470265e-05, + "loss": 0.9863, + "step": 3348 + }, + { + "epoch": 0.3221431319738361, + "grad_norm": 1.985256216985392, + "learning_rate": 1.584807179527287e-05, + "loss": 1.0197, + "step": 3349 + }, + { + "epoch": 0.3222393228164679, + "grad_norm": 2.041564620177297, + "learning_rate": 1.584554436646976e-05, + "loss": 1.0707, + "step": 3350 + }, + { + "epoch": 0.3223355136590996, + "grad_norm": 2.0209690477038276, + "learning_rate": 1.5843016370306238e-05, + "loss": 1.0545, + "step": 3351 + }, + { + "epoch": 0.3224317045017314, + "grad_norm": 1.871878007130569, + "learning_rate": 1.5840487807027665e-05, + "loss": 0.9932, + "step": 3352 + }, + { + "epoch": 0.3225278953443632, + "grad_norm": 1.9356751722981995, + "learning_rate": 1.583795867687947e-05, + "loss": 1.0251, + "step": 3353 + }, + { + "epoch": 0.322624086186995, + "grad_norm": 1.9053058327860537, + "learning_rate": 1.5835428980107113e-05, + "loss": 1.0633, + "step": 3354 + }, + { + "epoch": 0.3227202770296268, + "grad_norm": 2.062101251912309, + "learning_rate": 1.5832898716956134e-05, + "loss": 1.0552, + "step": 3355 + }, + { + "epoch": 0.32281646787225854, + "grad_norm": 42.07315405471486, + "learning_rate": 1.583036788767211e-05, + "loss": 1.7989, + "step": 3356 + }, + { + "epoch": 0.32291265871489033, + "grad_norm": 2.0729380216668827, + "learning_rate": 1.5827836492500683e-05, + "loss": 0.8994, + "step": 3357 + }, + { + "epoch": 0.3230088495575221, + "grad_norm": 4.312518632375639, + "learning_rate": 1.5825304531687548e-05, + "loss": 0.9545, + "step": 3358 + }, + { + "epoch": 0.3231050404001539, + "grad_norm": 2.5786060257684453, + "learning_rate": 1.5822772005478447e-05, + "loss": 1.0106, + "step": 3359 + }, + { + "epoch": 0.3232012312427857, + "grad_norm": 2.6968684369540874, + "learning_rate": 1.5820238914119195e-05, + "loss": 1.0576, + "step": 3360 + }, + { + "epoch": 0.32329742208541745, + "grad_norm": 2.6798508988985965, + "learning_rate": 1.5817705257855636e-05, + "loss": 1.0727, + "step": 3361 + }, + { + "epoch": 0.32339361292804925, + "grad_norm": 6.433034741034344, + "learning_rate": 1.5815171036933697e-05, + "loss": 1.1123, + "step": 3362 + }, + { + "epoch": 0.32348980377068104, + "grad_norm": 4.263287777268199, + "learning_rate": 1.5812636251599336e-05, + "loss": 0.9789, + "step": 3363 + }, + { + "epoch": 0.32358599461331283, + "grad_norm": 2.5880042880787695, + "learning_rate": 1.5810100902098582e-05, + "loss": 1.1217, + "step": 3364 + }, + { + "epoch": 0.32368218545594457, + "grad_norm": 2.135704085367862, + "learning_rate": 1.580756498867751e-05, + "loss": 1.0636, + "step": 3365 + }, + { + "epoch": 0.32377837629857636, + "grad_norm": 2.3042317253468605, + "learning_rate": 1.580502851158225e-05, + "loss": 1.0576, + "step": 3366 + }, + { + "epoch": 0.32387456714120816, + "grad_norm": 34.91582178761314, + "learning_rate": 1.5802491471058988e-05, + "loss": 1.7899, + "step": 3367 + }, + { + "epoch": 0.32397075798383995, + "grad_norm": 2.3534743262632665, + "learning_rate": 1.5799953867353975e-05, + "loss": 0.9246, + "step": 3368 + }, + { + "epoch": 0.32406694882647175, + "grad_norm": 2.3110559119835017, + "learning_rate": 1.5797415700713498e-05, + "loss": 1.0619, + "step": 3369 + }, + { + "epoch": 0.3241631396691035, + "grad_norm": 2.2356229789397686, + "learning_rate": 1.579487697138391e-05, + "loss": 1.1579, + "step": 3370 + }, + { + "epoch": 0.3242593305117353, + "grad_norm": 2.0077208880353794, + "learning_rate": 1.579233767961162e-05, + "loss": 1.0498, + "step": 3371 + }, + { + "epoch": 0.32435552135436707, + "grad_norm": 2.1355551951189375, + "learning_rate": 1.5789797825643086e-05, + "loss": 1.0225, + "step": 3372 + }, + { + "epoch": 0.32445171219699886, + "grad_norm": 14.753030041112252, + "learning_rate": 1.5787257409724815e-05, + "loss": 1.4712, + "step": 3373 + }, + { + "epoch": 0.3245479030396306, + "grad_norm": 2.442182945374188, + "learning_rate": 1.5784716432103394e-05, + "loss": 0.9342, + "step": 3374 + }, + { + "epoch": 0.3246440938822624, + "grad_norm": 18.760943300130638, + "learning_rate": 1.578217489302543e-05, + "loss": 1.5038, + "step": 3375 + }, + { + "epoch": 0.3247402847248942, + "grad_norm": 2.510121987004665, + "learning_rate": 1.5779632792737608e-05, + "loss": 1.0219, + "step": 3376 + }, + { + "epoch": 0.324836475567526, + "grad_norm": 2.8802749201607343, + "learning_rate": 1.577709013148666e-05, + "loss": 0.9665, + "step": 3377 + }, + { + "epoch": 0.3249326664101578, + "grad_norm": 2.2095125515988645, + "learning_rate": 1.5774546909519376e-05, + "loss": 1.0286, + "step": 3378 + }, + { + "epoch": 0.3250288572527895, + "grad_norm": 2.158319077683593, + "learning_rate": 1.5772003127082588e-05, + "loss": 1.0285, + "step": 3379 + }, + { + "epoch": 0.3251250480954213, + "grad_norm": 2.0614740941374046, + "learning_rate": 1.5769458784423206e-05, + "loss": 1.0198, + "step": 3380 + }, + { + "epoch": 0.3252212389380531, + "grad_norm": 2.190228398779842, + "learning_rate": 1.576691388178817e-05, + "loss": 1.0794, + "step": 3381 + }, + { + "epoch": 0.3253174297806849, + "grad_norm": 2.9168051940542776, + "learning_rate": 1.5764368419424488e-05, + "loss": 1.2067, + "step": 3382 + }, + { + "epoch": 0.32541362062331664, + "grad_norm": 2.579741548459194, + "learning_rate": 1.5761822397579223e-05, + "loss": 1.0611, + "step": 3383 + }, + { + "epoch": 0.32550981146594843, + "grad_norm": 2.529829958920162, + "learning_rate": 1.575927581649948e-05, + "loss": 1.0938, + "step": 3384 + }, + { + "epoch": 0.3256060023085802, + "grad_norm": 2.8713765438909085, + "learning_rate": 1.5756728676432435e-05, + "loss": 0.9695, + "step": 3385 + }, + { + "epoch": 0.325702193151212, + "grad_norm": 2.163236943439513, + "learning_rate": 1.5754180977625303e-05, + "loss": 0.9788, + "step": 3386 + }, + { + "epoch": 0.3257983839938438, + "grad_norm": 2.5750931321768156, + "learning_rate": 1.5751632720325364e-05, + "loss": 1.2018, + "step": 3387 + }, + { + "epoch": 0.32589457483647555, + "grad_norm": 2.0962080652084856, + "learning_rate": 1.574908390477995e-05, + "loss": 1.09, + "step": 3388 + }, + { + "epoch": 0.32599076567910734, + "grad_norm": 2.367657086937463, + "learning_rate": 1.5746534531236437e-05, + "loss": 0.9542, + "step": 3389 + }, + { + "epoch": 0.32608695652173914, + "grad_norm": 2.345043476599637, + "learning_rate": 1.5743984599942273e-05, + "loss": 0.9864, + "step": 3390 + }, + { + "epoch": 0.32618314736437093, + "grad_norm": 2.168202531140876, + "learning_rate": 1.5741434111144944e-05, + "loss": 1.0531, + "step": 3391 + }, + { + "epoch": 0.32627933820700267, + "grad_norm": 2.052066304681745, + "learning_rate": 1.5738883065092005e-05, + "loss": 1.0137, + "step": 3392 + }, + { + "epoch": 0.32637552904963446, + "grad_norm": 2.166521138404668, + "learning_rate": 1.573633146203105e-05, + "loss": 1.0216, + "step": 3393 + }, + { + "epoch": 0.32647171989226625, + "grad_norm": 2.169239819484317, + "learning_rate": 1.5733779302209735e-05, + "loss": 1.1305, + "step": 3394 + }, + { + "epoch": 0.32656791073489805, + "grad_norm": 2.1327108346547035, + "learning_rate": 1.5731226585875773e-05, + "loss": 0.9877, + "step": 3395 + }, + { + "epoch": 0.32666410157752984, + "grad_norm": 1.9948967876121992, + "learning_rate": 1.572867331327692e-05, + "loss": 0.9636, + "step": 3396 + }, + { + "epoch": 0.3267602924201616, + "grad_norm": 2.2906728388915947, + "learning_rate": 1.5726119484661e-05, + "loss": 0.9883, + "step": 3397 + }, + { + "epoch": 0.3268564832627934, + "grad_norm": 2.1651141422183358, + "learning_rate": 1.5723565100275884e-05, + "loss": 1.0501, + "step": 3398 + }, + { + "epoch": 0.32695267410542517, + "grad_norm": 2.1907925540498776, + "learning_rate": 1.572101016036949e-05, + "loss": 0.9962, + "step": 3399 + }, + { + "epoch": 0.32704886494805696, + "grad_norm": 2.25137664781002, + "learning_rate": 1.5718454665189806e-05, + "loss": 0.9827, + "step": 3400 + }, + { + "epoch": 0.3271450557906887, + "grad_norm": 1.9083961844877397, + "learning_rate": 1.5715898614984854e-05, + "loss": 0.9374, + "step": 3401 + }, + { + "epoch": 0.3272412466333205, + "grad_norm": 2.0996318064557826, + "learning_rate": 1.5713342010002733e-05, + "loss": 1.0467, + "step": 3402 + }, + { + "epoch": 0.3273374374759523, + "grad_norm": 2.0985426662480937, + "learning_rate": 1.5710784850491574e-05, + "loss": 1.1283, + "step": 3403 + }, + { + "epoch": 0.3274336283185841, + "grad_norm": 2.173727228323782, + "learning_rate": 1.5708227136699578e-05, + "loss": 1.011, + "step": 3404 + }, + { + "epoch": 0.3275298191612159, + "grad_norm": 2.465488710362461, + "learning_rate": 1.5705668868874986e-05, + "loss": 1.0383, + "step": 3405 + }, + { + "epoch": 0.3276260100038476, + "grad_norm": 2.133177053504206, + "learning_rate": 1.5703110047266105e-05, + "loss": 0.8838, + "step": 3406 + }, + { + "epoch": 0.3277222008464794, + "grad_norm": 2.1036906599695024, + "learning_rate": 1.570055067212129e-05, + "loss": 0.9432, + "step": 3407 + }, + { + "epoch": 0.3278183916891112, + "grad_norm": 2.079008934004058, + "learning_rate": 1.569799074368895e-05, + "loss": 0.97, + "step": 3408 + }, + { + "epoch": 0.327914582531743, + "grad_norm": 1.8905367844602174, + "learning_rate": 1.569543026221755e-05, + "loss": 0.9895, + "step": 3409 + }, + { + "epoch": 0.3280107733743748, + "grad_norm": 2.2574037486206606, + "learning_rate": 1.5692869227955603e-05, + "loss": 1.2183, + "step": 3410 + }, + { + "epoch": 0.3281069642170065, + "grad_norm": 2.2393144756144934, + "learning_rate": 1.569030764115168e-05, + "loss": 1.1383, + "step": 3411 + }, + { + "epoch": 0.3282031550596383, + "grad_norm": 2.223384499283326, + "learning_rate": 1.5687745502054407e-05, + "loss": 1.0005, + "step": 3412 + }, + { + "epoch": 0.3282993459022701, + "grad_norm": 2.2936314250055183, + "learning_rate": 1.568518281091246e-05, + "loss": 1.0322, + "step": 3413 + }, + { + "epoch": 0.3283955367449019, + "grad_norm": 10.269392256833747, + "learning_rate": 1.5682619567974575e-05, + "loss": 1.2991, + "step": 3414 + }, + { + "epoch": 0.32849172758753364, + "grad_norm": 2.267274809680023, + "learning_rate": 1.568005577348953e-05, + "loss": 0.9912, + "step": 3415 + }, + { + "epoch": 0.32858791843016544, + "grad_norm": 2.1692742893004726, + "learning_rate": 1.567749142770617e-05, + "loss": 0.889, + "step": 3416 + }, + { + "epoch": 0.32868410927279723, + "grad_norm": 2.2442658789493426, + "learning_rate": 1.567492653087338e-05, + "loss": 0.9029, + "step": 3417 + }, + { + "epoch": 0.328780300115429, + "grad_norm": 1.991416114986474, + "learning_rate": 1.5672361083240106e-05, + "loss": 0.8822, + "step": 3418 + }, + { + "epoch": 0.3288764909580608, + "grad_norm": 2.0743243852064426, + "learning_rate": 1.5669795085055352e-05, + "loss": 1.0658, + "step": 3419 + }, + { + "epoch": 0.32897268180069256, + "grad_norm": 1.9649460009657709, + "learning_rate": 1.5667228536568167e-05, + "loss": 1.0001, + "step": 3420 + }, + { + "epoch": 0.32906887264332435, + "grad_norm": 1.907842932875488, + "learning_rate": 1.5664661438027655e-05, + "loss": 1.0803, + "step": 3421 + }, + { + "epoch": 0.32916506348595614, + "grad_norm": 2.1762633428707594, + "learning_rate": 1.566209378968298e-05, + "loss": 0.9849, + "step": 3422 + }, + { + "epoch": 0.32926125432858794, + "grad_norm": 1.9943808653572688, + "learning_rate": 1.565952559178335e-05, + "loss": 0.9959, + "step": 3423 + }, + { + "epoch": 0.3293574451712197, + "grad_norm": 1.9902967336862176, + "learning_rate": 1.565695684457803e-05, + "loss": 0.8848, + "step": 3424 + }, + { + "epoch": 0.32945363601385147, + "grad_norm": 2.3815976069728264, + "learning_rate": 1.5654387548316348e-05, + "loss": 1.1044, + "step": 3425 + }, + { + "epoch": 0.32954982685648326, + "grad_norm": 2.1248729197380847, + "learning_rate": 1.5651817703247666e-05, + "loss": 1.1181, + "step": 3426 + }, + { + "epoch": 0.32964601769911506, + "grad_norm": 1.9829386154746016, + "learning_rate": 1.5649247309621413e-05, + "loss": 1.0566, + "step": 3427 + }, + { + "epoch": 0.32974220854174685, + "grad_norm": 2.3276123784769003, + "learning_rate": 1.5646676367687067e-05, + "loss": 1.1182, + "step": 3428 + }, + { + "epoch": 0.3298383993843786, + "grad_norm": 2.370828129875056, + "learning_rate": 1.564410487769416e-05, + "loss": 1.0864, + "step": 3429 + }, + { + "epoch": 0.3299345902270104, + "grad_norm": 1.8779298546413454, + "learning_rate": 1.564153283989228e-05, + "loss": 1.1334, + "step": 3430 + }, + { + "epoch": 0.3300307810696422, + "grad_norm": 2.0936099672398374, + "learning_rate": 1.5638960254531067e-05, + "loss": 1.0872, + "step": 3431 + }, + { + "epoch": 0.33012697191227397, + "grad_norm": 2.032240644488037, + "learning_rate": 1.5636387121860207e-05, + "loss": 1.0597, + "step": 3432 + }, + { + "epoch": 0.3302231627549057, + "grad_norm": 2.0121267231157853, + "learning_rate": 1.563381344212944e-05, + "loss": 1.1014, + "step": 3433 + }, + { + "epoch": 0.3303193535975375, + "grad_norm": 2.2395850237667996, + "learning_rate": 1.5631239215588578e-05, + "loss": 1.0888, + "step": 3434 + }, + { + "epoch": 0.3304155444401693, + "grad_norm": 1.9418879017550312, + "learning_rate": 1.5628664442487464e-05, + "loss": 1.0219, + "step": 3435 + }, + { + "epoch": 0.3305117352828011, + "grad_norm": 2.127501032131168, + "learning_rate": 1.5626089123076004e-05, + "loss": 1.0656, + "step": 3436 + }, + { + "epoch": 0.3306079261254329, + "grad_norm": 2.2497477489523683, + "learning_rate": 1.562351325760415e-05, + "loss": 1.0579, + "step": 3437 + }, + { + "epoch": 0.3307041169680646, + "grad_norm": 2.228906360904693, + "learning_rate": 1.5620936846321917e-05, + "loss": 0.9846, + "step": 3438 + }, + { + "epoch": 0.3308003078106964, + "grad_norm": 2.0845679178324934, + "learning_rate": 1.5618359889479365e-05, + "loss": 0.9376, + "step": 3439 + }, + { + "epoch": 0.3308964986533282, + "grad_norm": 2.7196007041184687, + "learning_rate": 1.561578238732661e-05, + "loss": 0.9778, + "step": 3440 + }, + { + "epoch": 0.33099268949596, + "grad_norm": 2.1135275435510494, + "learning_rate": 1.5613204340113824e-05, + "loss": 1.0061, + "step": 3441 + }, + { + "epoch": 0.33108888033859174, + "grad_norm": 2.2773523499696924, + "learning_rate": 1.561062574809123e-05, + "loss": 1.0016, + "step": 3442 + }, + { + "epoch": 0.33118507118122353, + "grad_norm": 7.629665785895361, + "learning_rate": 1.560804661150909e-05, + "loss": 1.4212, + "step": 3443 + }, + { + "epoch": 0.3312812620238553, + "grad_norm": 2.312992082592267, + "learning_rate": 1.5605466930617747e-05, + "loss": 1.0268, + "step": 3444 + }, + { + "epoch": 0.3313774528664871, + "grad_norm": 2.0699658837475288, + "learning_rate": 1.560288670566757e-05, + "loss": 0.9619, + "step": 3445 + }, + { + "epoch": 0.3314736437091189, + "grad_norm": 2.29663478446103, + "learning_rate": 1.5600305936909005e-05, + "loss": 1.0333, + "step": 3446 + }, + { + "epoch": 0.33156983455175065, + "grad_norm": 2.080403685527939, + "learning_rate": 1.559772462459252e-05, + "loss": 0.9332, + "step": 3447 + }, + { + "epoch": 0.33166602539438245, + "grad_norm": 2.196538421248554, + "learning_rate": 1.559514276896867e-05, + "loss": 1.1418, + "step": 3448 + }, + { + "epoch": 0.33176221623701424, + "grad_norm": 2.1795407913178995, + "learning_rate": 1.5592560370288037e-05, + "loss": 1.0265, + "step": 3449 + }, + { + "epoch": 0.33185840707964603, + "grad_norm": 5.1114035301680945, + "learning_rate": 1.558997742880127e-05, + "loss": 1.1151, + "step": 3450 + }, + { + "epoch": 0.33195459792227777, + "grad_norm": 2.634335931221359, + "learning_rate": 1.558739394475906e-05, + "loss": 1.0952, + "step": 3451 + }, + { + "epoch": 0.33205078876490957, + "grad_norm": 2.817595923523426, + "learning_rate": 1.5584809918412158e-05, + "loss": 1.0606, + "step": 3452 + }, + { + "epoch": 0.33214697960754136, + "grad_norm": 2.3121581051127458, + "learning_rate": 1.558222535001137e-05, + "loss": 1.0438, + "step": 3453 + }, + { + "epoch": 0.33224317045017315, + "grad_norm": 2.4558304991573108, + "learning_rate": 1.557964023980755e-05, + "loss": 1.0075, + "step": 3454 + }, + { + "epoch": 0.33233936129280495, + "grad_norm": 2.3221436957868855, + "learning_rate": 1.55770545880516e-05, + "loss": 1.0858, + "step": 3455 + }, + { + "epoch": 0.3324355521354367, + "grad_norm": 2.0466776530570026, + "learning_rate": 1.5574468394994486e-05, + "loss": 0.9006, + "step": 3456 + }, + { + "epoch": 0.3325317429780685, + "grad_norm": 2.12014112607077, + "learning_rate": 1.5571881660887215e-05, + "loss": 1.0379, + "step": 3457 + }, + { + "epoch": 0.3326279338207003, + "grad_norm": 2.2599165962930527, + "learning_rate": 1.5569294385980856e-05, + "loss": 1.1623, + "step": 3458 + }, + { + "epoch": 0.33272412466333207, + "grad_norm": 2.087784844127195, + "learning_rate": 1.5566706570526524e-05, + "loss": 1.0676, + "step": 3459 + }, + { + "epoch": 0.33282031550596386, + "grad_norm": 2.3055752593130254, + "learning_rate": 1.556411821477539e-05, + "loss": 1.0136, + "step": 3460 + }, + { + "epoch": 0.3329165063485956, + "grad_norm": 2.2207329042581776, + "learning_rate": 1.5561529318978675e-05, + "loss": 0.9592, + "step": 3461 + }, + { + "epoch": 0.3330126971912274, + "grad_norm": 2.1126566872113792, + "learning_rate": 1.5558939883387657e-05, + "loss": 1.0902, + "step": 3462 + }, + { + "epoch": 0.3331088880338592, + "grad_norm": 2.123431917950585, + "learning_rate": 1.5556349908253656e-05, + "loss": 1.0717, + "step": 3463 + }, + { + "epoch": 0.333205078876491, + "grad_norm": 2.2400539618491595, + "learning_rate": 1.5553759393828058e-05, + "loss": 1.1394, + "step": 3464 + }, + { + "epoch": 0.3333012697191227, + "grad_norm": 1.9872071196712287, + "learning_rate": 1.555116834036229e-05, + "loss": 1.1588, + "step": 3465 + }, + { + "epoch": 0.3333974605617545, + "grad_norm": 2.4809650056026498, + "learning_rate": 1.554857674810784e-05, + "loss": 1.0086, + "step": 3466 + }, + { + "epoch": 0.3334936514043863, + "grad_norm": 1.987697611186048, + "learning_rate": 1.5545984617316246e-05, + "loss": 1.0684, + "step": 3467 + }, + { + "epoch": 0.3335898422470181, + "grad_norm": 2.198656329657026, + "learning_rate": 1.554339194823909e-05, + "loss": 1.0778, + "step": 3468 + }, + { + "epoch": 0.3336860330896499, + "grad_norm": 2.0420859443971917, + "learning_rate": 1.5540798741128015e-05, + "loss": 0.9163, + "step": 3469 + }, + { + "epoch": 0.33378222393228163, + "grad_norm": 2.1488322027766884, + "learning_rate": 1.553820499623472e-05, + "loss": 1.0503, + "step": 3470 + }, + { + "epoch": 0.3338784147749134, + "grad_norm": 2.397803625021861, + "learning_rate": 1.553561071381094e-05, + "loss": 1.0953, + "step": 3471 + }, + { + "epoch": 0.3339746056175452, + "grad_norm": 2.1452495541898293, + "learning_rate": 1.553301589410848e-05, + "loss": 1.0638, + "step": 3472 + }, + { + "epoch": 0.334070796460177, + "grad_norm": 1.9229060086367293, + "learning_rate": 1.553042053737919e-05, + "loss": 1.0602, + "step": 3473 + }, + { + "epoch": 0.33416698730280875, + "grad_norm": 2.319808284160632, + "learning_rate": 1.5527824643874968e-05, + "loss": 1.0822, + "step": 3474 + }, + { + "epoch": 0.33426317814544054, + "grad_norm": 2.164724425766077, + "learning_rate": 1.5525228213847767e-05, + "loss": 1.1073, + "step": 3475 + }, + { + "epoch": 0.33435936898807234, + "grad_norm": 1.958063963021809, + "learning_rate": 1.5522631247549598e-05, + "loss": 1.008, + "step": 3476 + }, + { + "epoch": 0.33445555983070413, + "grad_norm": 2.2705982647015586, + "learning_rate": 1.5520033745232513e-05, + "loss": 0.9574, + "step": 3477 + }, + { + "epoch": 0.3345517506733359, + "grad_norm": 2.283289183103875, + "learning_rate": 1.5517435707148628e-05, + "loss": 1.0806, + "step": 3478 + }, + { + "epoch": 0.33464794151596766, + "grad_norm": 2.2221656103506304, + "learning_rate": 1.55148371335501e-05, + "loss": 1.0571, + "step": 3479 + }, + { + "epoch": 0.33474413235859946, + "grad_norm": 2.162855512289493, + "learning_rate": 1.5512238024689144e-05, + "loss": 1.0079, + "step": 3480 + }, + { + "epoch": 0.33484032320123125, + "grad_norm": 2.3834645550851232, + "learning_rate": 1.550963838081803e-05, + "loss": 1.111, + "step": 3481 + }, + { + "epoch": 0.33493651404386304, + "grad_norm": 2.2078105377009556, + "learning_rate": 1.550703820218907e-05, + "loss": 1.0545, + "step": 3482 + }, + { + "epoch": 0.3350327048864948, + "grad_norm": 2.1826244857106314, + "learning_rate": 1.5504437489054636e-05, + "loss": 0.9998, + "step": 3483 + }, + { + "epoch": 0.3351288957291266, + "grad_norm": 2.025337770815498, + "learning_rate": 1.550183624166715e-05, + "loss": 0.9176, + "step": 3484 + }, + { + "epoch": 0.33522508657175837, + "grad_norm": 2.2178887663725404, + "learning_rate": 1.5499234460279088e-05, + "loss": 1.1126, + "step": 3485 + }, + { + "epoch": 0.33532127741439016, + "grad_norm": 2.0626759055175556, + "learning_rate": 1.549663214514297e-05, + "loss": 1.131, + "step": 3486 + }, + { + "epoch": 0.33541746825702196, + "grad_norm": 2.47050149103599, + "learning_rate": 1.5494029296511378e-05, + "loss": 1.0818, + "step": 3487 + }, + { + "epoch": 0.3355136590996537, + "grad_norm": 2.3608718740196064, + "learning_rate": 1.5491425914636934e-05, + "loss": 1.1604, + "step": 3488 + }, + { + "epoch": 0.3356098499422855, + "grad_norm": 2.108171020898088, + "learning_rate": 1.5488821999772324e-05, + "loss": 1.1557, + "step": 3489 + }, + { + "epoch": 0.3357060407849173, + "grad_norm": 2.2662802393842254, + "learning_rate": 1.5486217552170283e-05, + "loss": 1.065, + "step": 3490 + }, + { + "epoch": 0.3358022316275491, + "grad_norm": 2.608426572432713, + "learning_rate": 1.548361257208359e-05, + "loss": 1.1185, + "step": 3491 + }, + { + "epoch": 0.3358984224701808, + "grad_norm": 2.108772630849692, + "learning_rate": 1.548100705976508e-05, + "loss": 0.9649, + "step": 3492 + }, + { + "epoch": 0.3359946133128126, + "grad_norm": 2.0851560353343657, + "learning_rate": 1.5478401015467652e-05, + "loss": 1.003, + "step": 3493 + }, + { + "epoch": 0.3360908041554444, + "grad_norm": 2.312705201123972, + "learning_rate": 1.5475794439444226e-05, + "loss": 1.0657, + "step": 3494 + }, + { + "epoch": 0.3361869949980762, + "grad_norm": 1.994690276136198, + "learning_rate": 1.5473187331947808e-05, + "loss": 1.0725, + "step": 3495 + }, + { + "epoch": 0.336283185840708, + "grad_norm": 1.998440334584037, + "learning_rate": 1.5470579693231432e-05, + "loss": 1.0327, + "step": 3496 + }, + { + "epoch": 0.3363793766833397, + "grad_norm": 1.9339194971449318, + "learning_rate": 1.5467971523548197e-05, + "loss": 1.0995, + "step": 3497 + }, + { + "epoch": 0.3364755675259715, + "grad_norm": 2.02113772856826, + "learning_rate": 1.5465362823151245e-05, + "loss": 0.9773, + "step": 3498 + }, + { + "epoch": 0.3365717583686033, + "grad_norm": 3.577556184668743, + "learning_rate": 1.546275359229377e-05, + "loss": 1.0777, + "step": 3499 + }, + { + "epoch": 0.3366679492112351, + "grad_norm": 2.0907278995265606, + "learning_rate": 1.5460143831229026e-05, + "loss": 1.0808, + "step": 3500 + }, + { + "epoch": 0.33676414005386685, + "grad_norm": 2.164036659751721, + "learning_rate": 1.545753354021031e-05, + "loss": 1.1303, + "step": 3501 + }, + { + "epoch": 0.33686033089649864, + "grad_norm": 2.0165238954172198, + "learning_rate": 1.545492271949098e-05, + "loss": 0.9396, + "step": 3502 + }, + { + "epoch": 0.33695652173913043, + "grad_norm": 2.281233412401555, + "learning_rate": 1.5452311369324425e-05, + "loss": 1.0953, + "step": 3503 + }, + { + "epoch": 0.3370527125817622, + "grad_norm": 2.077896592066381, + "learning_rate": 1.544969948996411e-05, + "loss": 0.9898, + "step": 3504 + }, + { + "epoch": 0.337148903424394, + "grad_norm": 17.74024861961895, + "learning_rate": 1.5447087081663535e-05, + "loss": 1.1069, + "step": 3505 + }, + { + "epoch": 0.33724509426702576, + "grad_norm": 2.0202840952905703, + "learning_rate": 1.544447414467626e-05, + "loss": 0.9811, + "step": 3506 + }, + { + "epoch": 0.33734128510965755, + "grad_norm": 2.045345548465862, + "learning_rate": 1.5441860679255896e-05, + "loss": 0.99, + "step": 3507 + }, + { + "epoch": 0.33743747595228935, + "grad_norm": 2.295371338606732, + "learning_rate": 1.5439246685656093e-05, + "loss": 1.0212, + "step": 3508 + }, + { + "epoch": 0.33753366679492114, + "grad_norm": 2.026451557120518, + "learning_rate": 1.543663216413057e-05, + "loss": 1.0561, + "step": 3509 + }, + { + "epoch": 0.33762985763755293, + "grad_norm": 2.0678491494096747, + "learning_rate": 1.5434017114933082e-05, + "loss": 1.0226, + "step": 3510 + }, + { + "epoch": 0.33772604848018467, + "grad_norm": 2.0071910240521307, + "learning_rate": 1.5431401538317446e-05, + "loss": 1.0345, + "step": 3511 + }, + { + "epoch": 0.33782223932281646, + "grad_norm": 2.031143166137785, + "learning_rate": 1.5428785434537527e-05, + "loss": 0.9822, + "step": 3512 + }, + { + "epoch": 0.33791843016544826, + "grad_norm": 1.9030409182565342, + "learning_rate": 1.5426168803847237e-05, + "loss": 1.0384, + "step": 3513 + }, + { + "epoch": 0.33801462100808005, + "grad_norm": 2.114224038918024, + "learning_rate": 1.542355164650055e-05, + "loss": 0.9668, + "step": 3514 + }, + { + "epoch": 0.3381108118507118, + "grad_norm": 1.981104249467409, + "learning_rate": 1.542093396275147e-05, + "loss": 1.0124, + "step": 3515 + }, + { + "epoch": 0.3382070026933436, + "grad_norm": 2.8959512731489636, + "learning_rate": 1.541831575285408e-05, + "loss": 1.0405, + "step": 3516 + }, + { + "epoch": 0.3383031935359754, + "grad_norm": 1.8761426716963778, + "learning_rate": 1.5415697017062493e-05, + "loss": 1.0439, + "step": 3517 + }, + { + "epoch": 0.33839938437860717, + "grad_norm": 2.111838176697001, + "learning_rate": 1.541307775563088e-05, + "loss": 1.1669, + "step": 3518 + }, + { + "epoch": 0.33849557522123896, + "grad_norm": 2.323819763184905, + "learning_rate": 1.5410457968813463e-05, + "loss": 1.0268, + "step": 3519 + }, + { + "epoch": 0.3385917660638707, + "grad_norm": 2.1168499350426986, + "learning_rate": 1.540783765686452e-05, + "loss": 0.9741, + "step": 3520 + }, + { + "epoch": 0.3386879569065025, + "grad_norm": 2.2309100343423216, + "learning_rate": 1.5405216820038366e-05, + "loss": 1.1435, + "step": 3521 + }, + { + "epoch": 0.3387841477491343, + "grad_norm": 16.956092192519648, + "learning_rate": 1.540259545858938e-05, + "loss": 1.3266, + "step": 3522 + }, + { + "epoch": 0.3388803385917661, + "grad_norm": 1.9403471749968746, + "learning_rate": 1.5399973572771988e-05, + "loss": 1.0318, + "step": 3523 + }, + { + "epoch": 0.3389765294343978, + "grad_norm": 1.8599336767845882, + "learning_rate": 1.539735116284067e-05, + "loss": 1.0066, + "step": 3524 + }, + { + "epoch": 0.3390727202770296, + "grad_norm": 2.3312294353380976, + "learning_rate": 1.5394728229049947e-05, + "loss": 1.0783, + "step": 3525 + }, + { + "epoch": 0.3391689111196614, + "grad_norm": 2.0404686311815263, + "learning_rate": 1.53921047716544e-05, + "loss": 1.0845, + "step": 3526 + }, + { + "epoch": 0.3392651019622932, + "grad_norm": 2.0204631933203157, + "learning_rate": 1.538948079090866e-05, + "loss": 1.1413, + "step": 3527 + }, + { + "epoch": 0.339361292804925, + "grad_norm": 2.2173441453262748, + "learning_rate": 1.53868562870674e-05, + "loss": 0.9424, + "step": 3528 + }, + { + "epoch": 0.33945748364755673, + "grad_norm": 1.8928144373671039, + "learning_rate": 1.538423126038536e-05, + "loss": 0.9728, + "step": 3529 + }, + { + "epoch": 0.33955367449018853, + "grad_norm": 1.949899967319384, + "learning_rate": 1.5381605711117318e-05, + "loss": 1.0121, + "step": 3530 + }, + { + "epoch": 0.3396498653328203, + "grad_norm": 2.0016612772010487, + "learning_rate": 1.5378979639518107e-05, + "loss": 1.0051, + "step": 3531 + }, + { + "epoch": 0.3397460561754521, + "grad_norm": 2.1191749144764165, + "learning_rate": 1.5376353045842604e-05, + "loss": 1.0068, + "step": 3532 + }, + { + "epoch": 0.33984224701808385, + "grad_norm": 2.1134868201559174, + "learning_rate": 1.537372593034575e-05, + "loss": 1.0074, + "step": 3533 + }, + { + "epoch": 0.33993843786071565, + "grad_norm": 2.3754850599003516, + "learning_rate": 1.5371098293282526e-05, + "loss": 1.1719, + "step": 3534 + }, + { + "epoch": 0.34003462870334744, + "grad_norm": 2.054426805915552, + "learning_rate": 1.5368470134907966e-05, + "loss": 1.0645, + "step": 3535 + }, + { + "epoch": 0.34013081954597923, + "grad_norm": 2.0490498506160066, + "learning_rate": 1.5365841455477158e-05, + "loss": 1.1937, + "step": 3536 + }, + { + "epoch": 0.34022701038861103, + "grad_norm": 1.9688998173052206, + "learning_rate": 1.5363212255245232e-05, + "loss": 1.0885, + "step": 3537 + }, + { + "epoch": 0.34032320123124277, + "grad_norm": 2.172566475589607, + "learning_rate": 1.5360582534467382e-05, + "loss": 1.0242, + "step": 3538 + }, + { + "epoch": 0.34041939207387456, + "grad_norm": 1.9893822150201788, + "learning_rate": 1.5357952293398843e-05, + "loss": 0.9966, + "step": 3539 + }, + { + "epoch": 0.34051558291650635, + "grad_norm": 2.0632795735284732, + "learning_rate": 1.5355321532294897e-05, + "loss": 0.9585, + "step": 3540 + }, + { + "epoch": 0.34061177375913815, + "grad_norm": 2.2750641169597428, + "learning_rate": 1.535269025141089e-05, + "loss": 0.9986, + "step": 3541 + }, + { + "epoch": 0.3407079646017699, + "grad_norm": 1.815339175776888, + "learning_rate": 1.5350058451002204e-05, + "loss": 1.0033, + "step": 3542 + }, + { + "epoch": 0.3408041554444017, + "grad_norm": 2.1727102973057715, + "learning_rate": 1.5347426131324286e-05, + "loss": 1.0323, + "step": 3543 + }, + { + "epoch": 0.3409003462870335, + "grad_norm": 1.7933598358248788, + "learning_rate": 1.5344793292632614e-05, + "loss": 1.0142, + "step": 3544 + }, + { + "epoch": 0.34099653712966527, + "grad_norm": 1.7830140955694294, + "learning_rate": 1.5342159935182736e-05, + "loss": 1.0176, + "step": 3545 + }, + { + "epoch": 0.34109272797229706, + "grad_norm": 2.019374400694966, + "learning_rate": 1.533952605923024e-05, + "loss": 1.0405, + "step": 3546 + }, + { + "epoch": 0.3411889188149288, + "grad_norm": 1.8539886779370576, + "learning_rate": 1.533689166503077e-05, + "loss": 0.937, + "step": 3547 + }, + { + "epoch": 0.3412851096575606, + "grad_norm": 1.9718196107497814, + "learning_rate": 1.5334256752840007e-05, + "loss": 0.9751, + "step": 3548 + }, + { + "epoch": 0.3413813005001924, + "grad_norm": 2.1089744246143702, + "learning_rate": 1.5331621322913698e-05, + "loss": 1.1187, + "step": 3549 + }, + { + "epoch": 0.3414774913428242, + "grad_norm": 1.9592743612278452, + "learning_rate": 1.532898537550764e-05, + "loss": 0.9506, + "step": 3550 + }, + { + "epoch": 0.3415736821854559, + "grad_norm": 1.9745207343300852, + "learning_rate": 1.5326348910877663e-05, + "loss": 1.1084, + "step": 3551 + }, + { + "epoch": 0.3416698730280877, + "grad_norm": 2.1737040973908606, + "learning_rate": 1.532371192927966e-05, + "loss": 1.0018, + "step": 3552 + }, + { + "epoch": 0.3417660638707195, + "grad_norm": 2.2638382412585067, + "learning_rate": 1.532107443096958e-05, + "loss": 1.0835, + "step": 3553 + }, + { + "epoch": 0.3418622547133513, + "grad_norm": 1.9808992446884934, + "learning_rate": 1.5318436416203412e-05, + "loss": 1.0678, + "step": 3554 + }, + { + "epoch": 0.3419584455559831, + "grad_norm": 2.03658817304759, + "learning_rate": 1.53157978852372e-05, + "loss": 0.8295, + "step": 3555 + }, + { + "epoch": 0.34205463639861483, + "grad_norm": 2.41139235393124, + "learning_rate": 1.531315883832703e-05, + "loss": 1.0648, + "step": 3556 + }, + { + "epoch": 0.3421508272412466, + "grad_norm": 2.0774303634011186, + "learning_rate": 1.531051927572905e-05, + "loss": 0.9201, + "step": 3557 + }, + { + "epoch": 0.3422470180838784, + "grad_norm": 2.1586627370777554, + "learning_rate": 1.530787919769945e-05, + "loss": 0.9924, + "step": 3558 + }, + { + "epoch": 0.3423432089265102, + "grad_norm": 2.0214967307548126, + "learning_rate": 1.5305238604494473e-05, + "loss": 1.047, + "step": 3559 + }, + { + "epoch": 0.34243939976914195, + "grad_norm": 1.9517585479577237, + "learning_rate": 1.5302597496370408e-05, + "loss": 1.0589, + "step": 3560 + }, + { + "epoch": 0.34253559061177374, + "grad_norm": 2.380037824738837, + "learning_rate": 1.52999558735836e-05, + "loss": 1.0612, + "step": 3561 + }, + { + "epoch": 0.34263178145440554, + "grad_norm": 2.2556546996323346, + "learning_rate": 1.5297313736390447e-05, + "loss": 1.107, + "step": 3562 + }, + { + "epoch": 0.34272797229703733, + "grad_norm": 1.9718510065953483, + "learning_rate": 1.5294671085047378e-05, + "loss": 0.996, + "step": 3563 + }, + { + "epoch": 0.3428241631396691, + "grad_norm": 2.1666896586521407, + "learning_rate": 1.5292027919810898e-05, + "loss": 1.128, + "step": 3564 + }, + { + "epoch": 0.34292035398230086, + "grad_norm": 2.3241367605814927, + "learning_rate": 1.5289384240937542e-05, + "loss": 0.9626, + "step": 3565 + }, + { + "epoch": 0.34301654482493266, + "grad_norm": 2.047236475946252, + "learning_rate": 1.52867400486839e-05, + "loss": 1.0407, + "step": 3566 + }, + { + "epoch": 0.34311273566756445, + "grad_norm": 1.9288257424271318, + "learning_rate": 1.528409534330662e-05, + "loss": 1.0667, + "step": 3567 + }, + { + "epoch": 0.34320892651019624, + "grad_norm": 2.1164248204375773, + "learning_rate": 1.528145012506239e-05, + "loss": 1.0721, + "step": 3568 + }, + { + "epoch": 0.34330511735282804, + "grad_norm": 2.1654160378383795, + "learning_rate": 1.5278804394207952e-05, + "loss": 1.0598, + "step": 3569 + }, + { + "epoch": 0.3434013081954598, + "grad_norm": 2.3506704595459067, + "learning_rate": 1.5276158151000096e-05, + "loss": 0.9711, + "step": 3570 + }, + { + "epoch": 0.34349749903809157, + "grad_norm": 1.8196429847249167, + "learning_rate": 1.5273511395695662e-05, + "loss": 0.9745, + "step": 3571 + }, + { + "epoch": 0.34359368988072336, + "grad_norm": 2.0945503184712586, + "learning_rate": 1.5270864128551542e-05, + "loss": 1.0455, + "step": 3572 + }, + { + "epoch": 0.34368988072335516, + "grad_norm": 1.8533655214600497, + "learning_rate": 1.5268216349824676e-05, + "loss": 1.0126, + "step": 3573 + }, + { + "epoch": 0.3437860715659869, + "grad_norm": 1.8768493916323248, + "learning_rate": 1.5265568059772053e-05, + "loss": 0.9831, + "step": 3574 + }, + { + "epoch": 0.3438822624086187, + "grad_norm": 2.1795112010605844, + "learning_rate": 1.5262919258650714e-05, + "loss": 1.0445, + "step": 3575 + }, + { + "epoch": 0.3439784532512505, + "grad_norm": 2.1183288028016287, + "learning_rate": 1.5260269946717746e-05, + "loss": 1.0071, + "step": 3576 + }, + { + "epoch": 0.3440746440938823, + "grad_norm": 1.823501146813506, + "learning_rate": 1.5257620124230288e-05, + "loss": 0.9336, + "step": 3577 + }, + { + "epoch": 0.34417083493651407, + "grad_norm": 2.2017834367834093, + "learning_rate": 1.5254969791445526e-05, + "loss": 1.1825, + "step": 3578 + }, + { + "epoch": 0.3442670257791458, + "grad_norm": 2.232435871065602, + "learning_rate": 1.5252318948620706e-05, + "loss": 1.1102, + "step": 3579 + }, + { + "epoch": 0.3443632166217776, + "grad_norm": 1.810775907125353, + "learning_rate": 1.5249667596013102e-05, + "loss": 0.8924, + "step": 3580 + }, + { + "epoch": 0.3444594074644094, + "grad_norm": 2.1729659871537756, + "learning_rate": 1.5247015733880061e-05, + "loss": 0.9707, + "step": 3581 + }, + { + "epoch": 0.3445555983070412, + "grad_norm": 1.9605075747043077, + "learning_rate": 1.5244363362478967e-05, + "loss": 1.0567, + "step": 3582 + }, + { + "epoch": 0.3446517891496729, + "grad_norm": 2.046673507442641, + "learning_rate": 1.5241710482067254e-05, + "loss": 1.0665, + "step": 3583 + }, + { + "epoch": 0.3447479799923047, + "grad_norm": 2.114169135787687, + "learning_rate": 1.5239057092902404e-05, + "loss": 0.9239, + "step": 3584 + }, + { + "epoch": 0.3448441708349365, + "grad_norm": 2.1531924312423, + "learning_rate": 1.5236403195241955e-05, + "loss": 1.1396, + "step": 3585 + }, + { + "epoch": 0.3449403616775683, + "grad_norm": 2.1142631100464278, + "learning_rate": 1.523374878934349e-05, + "loss": 1.1115, + "step": 3586 + }, + { + "epoch": 0.3450365525202001, + "grad_norm": 3.896530398998254, + "learning_rate": 1.5231093875464641e-05, + "loss": 1.0402, + "step": 3587 + }, + { + "epoch": 0.34513274336283184, + "grad_norm": 2.158618007960049, + "learning_rate": 1.5228438453863095e-05, + "loss": 1.157, + "step": 3588 + }, + { + "epoch": 0.34522893420546363, + "grad_norm": 2.29718323742562, + "learning_rate": 1.5225782524796576e-05, + "loss": 1.0573, + "step": 3589 + }, + { + "epoch": 0.3453251250480954, + "grad_norm": 2.226494055930443, + "learning_rate": 1.522312608852287e-05, + "loss": 1.0439, + "step": 3590 + }, + { + "epoch": 0.3454213158907272, + "grad_norm": 2.245742035537366, + "learning_rate": 1.5220469145299808e-05, + "loss": 1.0293, + "step": 3591 + }, + { + "epoch": 0.34551750673335896, + "grad_norm": 2.2203584420968543, + "learning_rate": 1.5217811695385263e-05, + "loss": 1.0916, + "step": 3592 + }, + { + "epoch": 0.34561369757599075, + "grad_norm": 2.2074110842162664, + "learning_rate": 1.5215153739037167e-05, + "loss": 1.0227, + "step": 3593 + }, + { + "epoch": 0.34570988841862255, + "grad_norm": 2.329127597907507, + "learning_rate": 1.52124952765135e-05, + "loss": 1.0609, + "step": 3594 + }, + { + "epoch": 0.34580607926125434, + "grad_norm": 2.016654575257005, + "learning_rate": 1.5209836308072283e-05, + "loss": 1.0981, + "step": 3595 + }, + { + "epoch": 0.34590227010388613, + "grad_norm": 1.917434607724708, + "learning_rate": 1.5207176833971598e-05, + "loss": 1.0153, + "step": 3596 + }, + { + "epoch": 0.34599846094651787, + "grad_norm": 2.2778730476879288, + "learning_rate": 1.5204516854469566e-05, + "loss": 1.0833, + "step": 3597 + }, + { + "epoch": 0.34609465178914967, + "grad_norm": 2.216656016158791, + "learning_rate": 1.520185636982436e-05, + "loss": 0.969, + "step": 3598 + }, + { + "epoch": 0.34619084263178146, + "grad_norm": 1.9700223201312914, + "learning_rate": 1.5199195380294207e-05, + "loss": 1.0755, + "step": 3599 + }, + { + "epoch": 0.34628703347441325, + "grad_norm": 2.031775454984139, + "learning_rate": 1.5196533886137376e-05, + "loss": 0.9907, + "step": 3600 + }, + { + "epoch": 0.346383224317045, + "grad_norm": 1.9889315565356325, + "learning_rate": 1.5193871887612188e-05, + "loss": 1.1083, + "step": 3601 + }, + { + "epoch": 0.3464794151596768, + "grad_norm": 2.076404854400022, + "learning_rate": 1.5191209384977014e-05, + "loss": 0.8928, + "step": 3602 + }, + { + "epoch": 0.3465756060023086, + "grad_norm": 2.081496390207823, + "learning_rate": 1.5188546378490272e-05, + "loss": 1.0052, + "step": 3603 + }, + { + "epoch": 0.34667179684494037, + "grad_norm": 2.0261684329086527, + "learning_rate": 1.5185882868410431e-05, + "loss": 1.086, + "step": 3604 + }, + { + "epoch": 0.34676798768757217, + "grad_norm": 3.337359534814992, + "learning_rate": 1.5183218854996007e-05, + "loss": 1.0595, + "step": 3605 + }, + { + "epoch": 0.3468641785302039, + "grad_norm": 2.0932546296196675, + "learning_rate": 1.5180554338505564e-05, + "loss": 1.1172, + "step": 3606 + }, + { + "epoch": 0.3469603693728357, + "grad_norm": 3.305198311361733, + "learning_rate": 1.5177889319197716e-05, + "loss": 1.1121, + "step": 3607 + }, + { + "epoch": 0.3470565602154675, + "grad_norm": 1.7465289338587184, + "learning_rate": 1.517522379733113e-05, + "loss": 1.0772, + "step": 3608 + }, + { + "epoch": 0.3471527510580993, + "grad_norm": 2.0647187958627033, + "learning_rate": 1.5172557773164514e-05, + "loss": 0.8687, + "step": 3609 + }, + { + "epoch": 0.347248941900731, + "grad_norm": 1.8195542872549546, + "learning_rate": 1.5169891246956629e-05, + "loss": 1.033, + "step": 3610 + }, + { + "epoch": 0.3473451327433628, + "grad_norm": 2.2440925264272997, + "learning_rate": 1.5167224218966287e-05, + "loss": 0.9477, + "step": 3611 + }, + { + "epoch": 0.3474413235859946, + "grad_norm": 1.9957132224955692, + "learning_rate": 1.5164556689452346e-05, + "loss": 1.0125, + "step": 3612 + }, + { + "epoch": 0.3475375144286264, + "grad_norm": 1.9938977322875027, + "learning_rate": 1.5161888658673704e-05, + "loss": 1.027, + "step": 3613 + }, + { + "epoch": 0.3476337052712582, + "grad_norm": 2.1210268076810515, + "learning_rate": 1.5159220126889329e-05, + "loss": 1.0872, + "step": 3614 + }, + { + "epoch": 0.34772989611388994, + "grad_norm": 3.0365374413216064, + "learning_rate": 1.5156551094358216e-05, + "loss": 0.9952, + "step": 3615 + }, + { + "epoch": 0.34782608695652173, + "grad_norm": 2.026820125466472, + "learning_rate": 1.5153881561339426e-05, + "loss": 1.0178, + "step": 3616 + }, + { + "epoch": 0.3479222777991535, + "grad_norm": 2.16338723777089, + "learning_rate": 1.515121152809205e-05, + "loss": 1.0213, + "step": 3617 + }, + { + "epoch": 0.3480184686417853, + "grad_norm": 2.1878057939737188, + "learning_rate": 1.5148540994875242e-05, + "loss": 1.0731, + "step": 3618 + }, + { + "epoch": 0.3481146594844171, + "grad_norm": 1.8542112916406186, + "learning_rate": 1.5145869961948205e-05, + "loss": 0.8682, + "step": 3619 + }, + { + "epoch": 0.34821085032704885, + "grad_norm": 2.2564258308248735, + "learning_rate": 1.5143198429570181e-05, + "loss": 1.0014, + "step": 3620 + }, + { + "epoch": 0.34830704116968064, + "grad_norm": 2.179810572495606, + "learning_rate": 1.5140526398000463e-05, + "loss": 1.0066, + "step": 3621 + }, + { + "epoch": 0.34840323201231244, + "grad_norm": 2.325976302900072, + "learning_rate": 1.5137853867498403e-05, + "loss": 1.0351, + "step": 3622 + }, + { + "epoch": 0.34849942285494423, + "grad_norm": 2.210591076825397, + "learning_rate": 1.5135180838323382e-05, + "loss": 1.0792, + "step": 3623 + }, + { + "epoch": 0.34859561369757597, + "grad_norm": 2.2039623574351443, + "learning_rate": 1.5132507310734847e-05, + "loss": 1.0232, + "step": 3624 + }, + { + "epoch": 0.34869180454020776, + "grad_norm": 2.1337079529671197, + "learning_rate": 1.512983328499229e-05, + "loss": 1.0133, + "step": 3625 + }, + { + "epoch": 0.34878799538283956, + "grad_norm": 2.1484322189086207, + "learning_rate": 1.5127158761355241e-05, + "loss": 1.041, + "step": 3626 + }, + { + "epoch": 0.34888418622547135, + "grad_norm": 2.024368078545174, + "learning_rate": 1.5124483740083288e-05, + "loss": 1.0709, + "step": 3627 + }, + { + "epoch": 0.34898037706810314, + "grad_norm": 2.3644868041364075, + "learning_rate": 1.512180822143607e-05, + "loss": 1.0185, + "step": 3628 + }, + { + "epoch": 0.3490765679107349, + "grad_norm": 2.352766154430316, + "learning_rate": 1.5119132205673259e-05, + "loss": 1.0298, + "step": 3629 + }, + { + "epoch": 0.3491727587533667, + "grad_norm": 2.5264205530895834, + "learning_rate": 1.5116455693054594e-05, + "loss": 1.0799, + "step": 3630 + }, + { + "epoch": 0.34926894959599847, + "grad_norm": 2.56937169253593, + "learning_rate": 1.5113778683839853e-05, + "loss": 1.0242, + "step": 3631 + }, + { + "epoch": 0.34936514043863026, + "grad_norm": 1.938892595252592, + "learning_rate": 1.5111101178288858e-05, + "loss": 1.0637, + "step": 3632 + }, + { + "epoch": 0.349461331281262, + "grad_norm": 1.905407637895601, + "learning_rate": 1.5108423176661485e-05, + "loss": 1.1175, + "step": 3633 + }, + { + "epoch": 0.3495575221238938, + "grad_norm": 1.9941110158892341, + "learning_rate": 1.510574467921766e-05, + "loss": 0.9541, + "step": 3634 + }, + { + "epoch": 0.3496537129665256, + "grad_norm": 2.0778702137988176, + "learning_rate": 1.5103065686217356e-05, + "loss": 0.9372, + "step": 3635 + }, + { + "epoch": 0.3497499038091574, + "grad_norm": 1.9071343944876522, + "learning_rate": 1.5100386197920585e-05, + "loss": 0.9985, + "step": 3636 + }, + { + "epoch": 0.3498460946517892, + "grad_norm": 2.0644716242414556, + "learning_rate": 1.509770621458742e-05, + "loss": 0.9697, + "step": 3637 + }, + { + "epoch": 0.3499422854944209, + "grad_norm": 2.396736112582529, + "learning_rate": 1.5095025736477977e-05, + "loss": 0.953, + "step": 3638 + }, + { + "epoch": 0.3500384763370527, + "grad_norm": 2.244504179764334, + "learning_rate": 1.5092344763852413e-05, + "loss": 1.0226, + "step": 3639 + }, + { + "epoch": 0.3501346671796845, + "grad_norm": 2.195184206811099, + "learning_rate": 1.5089663296970952e-05, + "loss": 1.0532, + "step": 3640 + }, + { + "epoch": 0.3502308580223163, + "grad_norm": 2.454887783364361, + "learning_rate": 1.5086981336093835e-05, + "loss": 1.1288, + "step": 3641 + }, + { + "epoch": 0.35032704886494803, + "grad_norm": 2.2024943537657826, + "learning_rate": 1.5084298881481388e-05, + "loss": 0.9464, + "step": 3642 + }, + { + "epoch": 0.3504232397075798, + "grad_norm": 2.513964449667441, + "learning_rate": 1.5081615933393954e-05, + "loss": 1.0907, + "step": 3643 + }, + { + "epoch": 0.3505194305502116, + "grad_norm": 2.048010527064041, + "learning_rate": 1.5078932492091942e-05, + "loss": 0.9968, + "step": 3644 + }, + { + "epoch": 0.3506156213928434, + "grad_norm": 2.2148776043832856, + "learning_rate": 1.5076248557835803e-05, + "loss": 0.9411, + "step": 3645 + }, + { + "epoch": 0.3507118122354752, + "grad_norm": 2.0191609527105974, + "learning_rate": 1.5073564130886032e-05, + "loss": 0.9314, + "step": 3646 + }, + { + "epoch": 0.35080800307810694, + "grad_norm": 1.9736261014112741, + "learning_rate": 1.5070879211503177e-05, + "loss": 1.0973, + "step": 3647 + }, + { + "epoch": 0.35090419392073874, + "grad_norm": 2.104506067409004, + "learning_rate": 1.506819379994784e-05, + "loss": 1.104, + "step": 3648 + }, + { + "epoch": 0.35100038476337053, + "grad_norm": 1.9771106852672853, + "learning_rate": 1.5065507896480651e-05, + "loss": 0.9888, + "step": 3649 + }, + { + "epoch": 0.3510965756060023, + "grad_norm": 2.1820995881579885, + "learning_rate": 1.5062821501362308e-05, + "loss": 1.111, + "step": 3650 + }, + { + "epoch": 0.35119276644863406, + "grad_norm": 2.198081316778613, + "learning_rate": 1.5060134614853547e-05, + "loss": 1.0918, + "step": 3651 + }, + { + "epoch": 0.35128895729126586, + "grad_norm": 2.2131595126935193, + "learning_rate": 1.5057447237215152e-05, + "loss": 1.0756, + "step": 3652 + }, + { + "epoch": 0.35138514813389765, + "grad_norm": 2.455563751574842, + "learning_rate": 1.5054759368707956e-05, + "loss": 0.9706, + "step": 3653 + }, + { + "epoch": 0.35148133897652944, + "grad_norm": 1.958163450686244, + "learning_rate": 1.5052071009592846e-05, + "loss": 1.0354, + "step": 3654 + }, + { + "epoch": 0.35157752981916124, + "grad_norm": 2.334897078576552, + "learning_rate": 1.5049382160130743e-05, + "loss": 1.0691, + "step": 3655 + }, + { + "epoch": 0.351673720661793, + "grad_norm": 2.1015459284557227, + "learning_rate": 1.5046692820582625e-05, + "loss": 0.9152, + "step": 3656 + }, + { + "epoch": 0.35176991150442477, + "grad_norm": 2.1545792345016146, + "learning_rate": 1.5044002991209521e-05, + "loss": 1.0013, + "step": 3657 + }, + { + "epoch": 0.35186610234705656, + "grad_norm": 2.0927782926476954, + "learning_rate": 1.504131267227249e-05, + "loss": 0.9774, + "step": 3658 + }, + { + "epoch": 0.35196229318968836, + "grad_norm": 2.010593204797717, + "learning_rate": 1.503862186403266e-05, + "loss": 1.023, + "step": 3659 + }, + { + "epoch": 0.3520584840323201, + "grad_norm": 2.1262632505906005, + "learning_rate": 1.5035930566751198e-05, + "loss": 1.0189, + "step": 3660 + }, + { + "epoch": 0.3521546748749519, + "grad_norm": 2.116768882152646, + "learning_rate": 1.5033238780689315e-05, + "loss": 1.0094, + "step": 3661 + }, + { + "epoch": 0.3522508657175837, + "grad_norm": 1.970151196481008, + "learning_rate": 1.5030546506108268e-05, + "loss": 0.9945, + "step": 3662 + }, + { + "epoch": 0.3523470565602155, + "grad_norm": 2.104959581769643, + "learning_rate": 1.502785374326937e-05, + "loss": 1.0744, + "step": 3663 + }, + { + "epoch": 0.35244324740284727, + "grad_norm": 2.1774162549588505, + "learning_rate": 1.5025160492433976e-05, + "loss": 0.9973, + "step": 3664 + }, + { + "epoch": 0.352539438245479, + "grad_norm": 1.9725827777224756, + "learning_rate": 1.5022466753863488e-05, + "loss": 1.1049, + "step": 3665 + }, + { + "epoch": 0.3526356290881108, + "grad_norm": 1.845402693820414, + "learning_rate": 1.501977252781936e-05, + "loss": 0.9588, + "step": 3666 + }, + { + "epoch": 0.3527318199307426, + "grad_norm": 2.142097262230541, + "learning_rate": 1.5017077814563089e-05, + "loss": 1.1449, + "step": 3667 + }, + { + "epoch": 0.3528280107733744, + "grad_norm": 2.1782837081722404, + "learning_rate": 1.5014382614356213e-05, + "loss": 1.0459, + "step": 3668 + }, + { + "epoch": 0.3529242016160062, + "grad_norm": 2.2407569785866754, + "learning_rate": 1.5011686927460334e-05, + "loss": 1.0937, + "step": 3669 + }, + { + "epoch": 0.3530203924586379, + "grad_norm": 1.956448549994343, + "learning_rate": 1.5008990754137088e-05, + "loss": 1.0426, + "step": 3670 + }, + { + "epoch": 0.3531165833012697, + "grad_norm": 1.9617008755831566, + "learning_rate": 1.5006294094648159e-05, + "loss": 0.9251, + "step": 3671 + }, + { + "epoch": 0.3532127741439015, + "grad_norm": 2.2603510663182775, + "learning_rate": 1.5003596949255284e-05, + "loss": 0.9819, + "step": 3672 + }, + { + "epoch": 0.3533089649865333, + "grad_norm": 1.9604869890965697, + "learning_rate": 1.5000899318220244e-05, + "loss": 1.0341, + "step": 3673 + }, + { + "epoch": 0.35340515582916504, + "grad_norm": 2.181466575765584, + "learning_rate": 1.4998201201804867e-05, + "loss": 1.062, + "step": 3674 + }, + { + "epoch": 0.35350134667179683, + "grad_norm": 2.037364215514659, + "learning_rate": 1.4995502600271028e-05, + "loss": 1.0846, + "step": 3675 + }, + { + "epoch": 0.35359753751442863, + "grad_norm": 2.1209467154068267, + "learning_rate": 1.499280351388065e-05, + "loss": 1.0213, + "step": 3676 + }, + { + "epoch": 0.3536937283570604, + "grad_norm": 1.9689592070768398, + "learning_rate": 1.4990103942895702e-05, + "loss": 1.1034, + "step": 3677 + }, + { + "epoch": 0.3537899191996922, + "grad_norm": 11.001047781392662, + "learning_rate": 1.49874038875782e-05, + "loss": 0.8667, + "step": 3678 + }, + { + "epoch": 0.35388611004232395, + "grad_norm": 1.9862090018141303, + "learning_rate": 1.4984703348190209e-05, + "loss": 1.0905, + "step": 3679 + }, + { + "epoch": 0.35398230088495575, + "grad_norm": 2.1807216009158967, + "learning_rate": 1.498200232499384e-05, + "loss": 1.0385, + "step": 3680 + }, + { + "epoch": 0.35407849172758754, + "grad_norm": 2.002905669412871, + "learning_rate": 1.497930081825125e-05, + "loss": 0.9353, + "step": 3681 + }, + { + "epoch": 0.35417468257021933, + "grad_norm": 13.788881161400067, + "learning_rate": 1.4976598828224643e-05, + "loss": 1.3285, + "step": 3682 + }, + { + "epoch": 0.3542708734128511, + "grad_norm": 2.8342845234867595, + "learning_rate": 1.4973896355176271e-05, + "loss": 0.9618, + "step": 3683 + }, + { + "epoch": 0.35436706425548287, + "grad_norm": 2.06233385627462, + "learning_rate": 1.497119339936843e-05, + "loss": 1.1092, + "step": 3684 + }, + { + "epoch": 0.35446325509811466, + "grad_norm": 2.0485641616016927, + "learning_rate": 1.4968489961063471e-05, + "loss": 1.023, + "step": 3685 + }, + { + "epoch": 0.35455944594074645, + "grad_norm": 1.9097282875856838, + "learning_rate": 1.4965786040523779e-05, + "loss": 1.0494, + "step": 3686 + }, + { + "epoch": 0.35465563678337825, + "grad_norm": 6.155883250834932, + "learning_rate": 1.4963081638011798e-05, + "loss": 1.1981, + "step": 3687 + }, + { + "epoch": 0.35475182762601, + "grad_norm": 2.6042269197615355, + "learning_rate": 1.496037675379001e-05, + "loss": 1.0575, + "step": 3688 + }, + { + "epoch": 0.3548480184686418, + "grad_norm": 1.9841077211017644, + "learning_rate": 1.4957671388120949e-05, + "loss": 1.0141, + "step": 3689 + }, + { + "epoch": 0.3549442093112736, + "grad_norm": 1.9772137557216996, + "learning_rate": 1.4954965541267192e-05, + "loss": 1.104, + "step": 3690 + }, + { + "epoch": 0.35504040015390537, + "grad_norm": 1.9209284886929883, + "learning_rate": 1.4952259213491366e-05, + "loss": 1.057, + "step": 3691 + }, + { + "epoch": 0.3551365909965371, + "grad_norm": 1.8851140313989667, + "learning_rate": 1.494955240505615e-05, + "loss": 0.9994, + "step": 3692 + }, + { + "epoch": 0.3552327818391689, + "grad_norm": 2.045002651069991, + "learning_rate": 1.4946845116224249e-05, + "loss": 0.9875, + "step": 3693 + }, + { + "epoch": 0.3553289726818007, + "grad_norm": 1.9322837884773445, + "learning_rate": 1.494413734725844e-05, + "loss": 0.9627, + "step": 3694 + }, + { + "epoch": 0.3554251635244325, + "grad_norm": 2.0794464185084682, + "learning_rate": 1.4941429098421534e-05, + "loss": 1.136, + "step": 3695 + }, + { + "epoch": 0.3555213543670643, + "grad_norm": 1.9748028067753938, + "learning_rate": 1.4938720369976385e-05, + "loss": 1.0663, + "step": 3696 + }, + { + "epoch": 0.355617545209696, + "grad_norm": 2.40007455738116, + "learning_rate": 1.4936011162185904e-05, + "loss": 1.0454, + "step": 3697 + }, + { + "epoch": 0.3557137360523278, + "grad_norm": 2.197113167494122, + "learning_rate": 1.4933301475313036e-05, + "loss": 1.0297, + "step": 3698 + }, + { + "epoch": 0.3558099268949596, + "grad_norm": 2.2726966345556097, + "learning_rate": 1.4930591309620786e-05, + "loss": 1.0632, + "step": 3699 + }, + { + "epoch": 0.3559061177375914, + "grad_norm": 1.9584166806191565, + "learning_rate": 1.4927880665372197e-05, + "loss": 1.0154, + "step": 3700 + }, + { + "epoch": 0.35600230858022314, + "grad_norm": 2.0706063244574966, + "learning_rate": 1.492516954283036e-05, + "loss": 1.1067, + "step": 3701 + }, + { + "epoch": 0.35609849942285493, + "grad_norm": 2.0785730062473626, + "learning_rate": 1.4922457942258411e-05, + "loss": 1.0082, + "step": 3702 + }, + { + "epoch": 0.3561946902654867, + "grad_norm": 1.979490606929219, + "learning_rate": 1.4919745863919537e-05, + "loss": 0.932, + "step": 3703 + }, + { + "epoch": 0.3562908811081185, + "grad_norm": 2.3246527302268696, + "learning_rate": 1.4917033308076967e-05, + "loss": 1.078, + "step": 3704 + }, + { + "epoch": 0.3563870719507503, + "grad_norm": 2.7644342530458452, + "learning_rate": 1.4914320274993976e-05, + "loss": 0.9923, + "step": 3705 + }, + { + "epoch": 0.35648326279338205, + "grad_norm": 5.799496637218058, + "learning_rate": 1.4911606764933892e-05, + "loss": 1.0613, + "step": 3706 + }, + { + "epoch": 0.35657945363601384, + "grad_norm": 2.2462826318323876, + "learning_rate": 1.490889277816008e-05, + "loss": 0.9874, + "step": 3707 + }, + { + "epoch": 0.35667564447864564, + "grad_norm": 2.0471687730654162, + "learning_rate": 1.490617831493596e-05, + "loss": 0.9801, + "step": 3708 + }, + { + "epoch": 0.35677183532127743, + "grad_norm": 2.208363770251166, + "learning_rate": 1.490346337552499e-05, + "loss": 1.0484, + "step": 3709 + }, + { + "epoch": 0.35686802616390917, + "grad_norm": 2.052587339007546, + "learning_rate": 1.4900747960190682e-05, + "loss": 0.936, + "step": 3710 + }, + { + "epoch": 0.35696421700654096, + "grad_norm": 1.9897647426492913, + "learning_rate": 1.489803206919659e-05, + "loss": 1.0288, + "step": 3711 + }, + { + "epoch": 0.35706040784917276, + "grad_norm": 2.320175467956998, + "learning_rate": 1.489531570280631e-05, + "loss": 1.0221, + "step": 3712 + }, + { + "epoch": 0.35715659869180455, + "grad_norm": 2.0127868589040308, + "learning_rate": 1.4892598861283492e-05, + "loss": 1.0522, + "step": 3713 + }, + { + "epoch": 0.35725278953443634, + "grad_norm": 2.058604498648885, + "learning_rate": 1.488988154489183e-05, + "loss": 1.0436, + "step": 3714 + }, + { + "epoch": 0.3573489803770681, + "grad_norm": 2.027184935802777, + "learning_rate": 1.4887163753895062e-05, + "loss": 1.026, + "step": 3715 + }, + { + "epoch": 0.3574451712196999, + "grad_norm": 1.9025792307564107, + "learning_rate": 1.4884445488556972e-05, + "loss": 1.0556, + "step": 3716 + }, + { + "epoch": 0.35754136206233167, + "grad_norm": 2.135087443174667, + "learning_rate": 1.488172674914139e-05, + "loss": 1.0321, + "step": 3717 + }, + { + "epoch": 0.35763755290496346, + "grad_norm": 1.7888301735360452, + "learning_rate": 1.4879007535912198e-05, + "loss": 0.823, + "step": 3718 + }, + { + "epoch": 0.35773374374759526, + "grad_norm": 1.897039699222452, + "learning_rate": 1.4876287849133312e-05, + "loss": 0.9385, + "step": 3719 + }, + { + "epoch": 0.357829934590227, + "grad_norm": 2.0019953316569055, + "learning_rate": 1.4873567689068708e-05, + "loss": 1.0353, + "step": 3720 + }, + { + "epoch": 0.3579261254328588, + "grad_norm": 2.255374438509242, + "learning_rate": 1.4870847055982397e-05, + "loss": 0.9251, + "step": 3721 + }, + { + "epoch": 0.3580223162754906, + "grad_norm": 2.1120380205243405, + "learning_rate": 1.4868125950138442e-05, + "loss": 0.8934, + "step": 3722 + }, + { + "epoch": 0.3581185071181224, + "grad_norm": 1.9414262068967862, + "learning_rate": 1.4865404371800947e-05, + "loss": 1.0252, + "step": 3723 + }, + { + "epoch": 0.3582146979607541, + "grad_norm": 10.232349771030451, + "learning_rate": 1.4862682321234064e-05, + "loss": 1.1066, + "step": 3724 + }, + { + "epoch": 0.3583108888033859, + "grad_norm": 1.929491527165463, + "learning_rate": 1.4859959798701997e-05, + "loss": 0.9901, + "step": 3725 + }, + { + "epoch": 0.3584070796460177, + "grad_norm": 2.417926972493095, + "learning_rate": 1.4857236804468983e-05, + "loss": 1.0739, + "step": 3726 + }, + { + "epoch": 0.3585032704886495, + "grad_norm": 2.203825494478811, + "learning_rate": 1.4854513338799322e-05, + "loss": 1.047, + "step": 3727 + }, + { + "epoch": 0.3585994613312813, + "grad_norm": 1.6884525232010272, + "learning_rate": 1.4851789401957338e-05, + "loss": 1.011, + "step": 3728 + }, + { + "epoch": 0.358695652173913, + "grad_norm": 2.1279076032628588, + "learning_rate": 1.484906499420742e-05, + "loss": 1.012, + "step": 3729 + }, + { + "epoch": 0.3587918430165448, + "grad_norm": 1.9875642465465913, + "learning_rate": 1.4846340115813993e-05, + "loss": 1.1011, + "step": 3730 + }, + { + "epoch": 0.3588880338591766, + "grad_norm": 1.9606759453040024, + "learning_rate": 1.484361476704153e-05, + "loss": 1.1195, + "step": 3731 + }, + { + "epoch": 0.3589842247018084, + "grad_norm": 2.0019583564236045, + "learning_rate": 1.484088894815455e-05, + "loss": 1.0029, + "step": 3732 + }, + { + "epoch": 0.35908041554444015, + "grad_norm": 2.2474180064415297, + "learning_rate": 1.4838162659417617e-05, + "loss": 0.9836, + "step": 3733 + }, + { + "epoch": 0.35917660638707194, + "grad_norm": 12.682847954985931, + "learning_rate": 1.4835435901095341e-05, + "loss": 1.3289, + "step": 3734 + }, + { + "epoch": 0.35927279722970373, + "grad_norm": 2.1153593147320713, + "learning_rate": 1.4832708673452376e-05, + "loss": 1.1331, + "step": 3735 + }, + { + "epoch": 0.3593689880723355, + "grad_norm": 2.0598112003699818, + "learning_rate": 1.4829980976753426e-05, + "loss": 1.0848, + "step": 3736 + }, + { + "epoch": 0.3594651789149673, + "grad_norm": 2.544443357937647, + "learning_rate": 1.4827252811263235e-05, + "loss": 0.9675, + "step": 3737 + }, + { + "epoch": 0.35956136975759906, + "grad_norm": 1.8042761492879502, + "learning_rate": 1.4824524177246597e-05, + "loss": 1.0012, + "step": 3738 + }, + { + "epoch": 0.35965756060023085, + "grad_norm": 1.97914820613731, + "learning_rate": 1.4821795074968346e-05, + "loss": 1.0816, + "step": 3739 + }, + { + "epoch": 0.35975375144286265, + "grad_norm": 2.0679571127004106, + "learning_rate": 1.4819065504693365e-05, + "loss": 1.0794, + "step": 3740 + }, + { + "epoch": 0.35984994228549444, + "grad_norm": 2.146302862780628, + "learning_rate": 1.4816335466686588e-05, + "loss": 0.9094, + "step": 3741 + }, + { + "epoch": 0.3599461331281262, + "grad_norm": 2.111848256915086, + "learning_rate": 1.4813604961212984e-05, + "loss": 1.067, + "step": 3742 + }, + { + "epoch": 0.36004232397075797, + "grad_norm": 2.0930716966508514, + "learning_rate": 1.4810873988537569e-05, + "loss": 1.0609, + "step": 3743 + }, + { + "epoch": 0.36013851481338977, + "grad_norm": 1.8768317826573073, + "learning_rate": 1.4808142548925417e-05, + "loss": 0.9758, + "step": 3744 + }, + { + "epoch": 0.36023470565602156, + "grad_norm": 2.039918084523893, + "learning_rate": 1.4805410642641627e-05, + "loss": 0.928, + "step": 3745 + }, + { + "epoch": 0.36033089649865335, + "grad_norm": 1.8718282708639473, + "learning_rate": 1.4802678269951365e-05, + "loss": 0.8514, + "step": 3746 + }, + { + "epoch": 0.3604270873412851, + "grad_norm": 1.9538451830317263, + "learning_rate": 1.4799945431119818e-05, + "loss": 1.0189, + "step": 3747 + }, + { + "epoch": 0.3605232781839169, + "grad_norm": 2.063474742801012, + "learning_rate": 1.4797212126412243e-05, + "loss": 1.0368, + "step": 3748 + }, + { + "epoch": 0.3606194690265487, + "grad_norm": 2.0513403551502223, + "learning_rate": 1.4794478356093927e-05, + "loss": 1.0157, + "step": 3749 + }, + { + "epoch": 0.36071565986918047, + "grad_norm": 2.193396769145583, + "learning_rate": 1.4791744120430202e-05, + "loss": 1.0572, + "step": 3750 + }, + { + "epoch": 0.3608118507118122, + "grad_norm": 2.0734311557071203, + "learning_rate": 1.4789009419686458e-05, + "loss": 0.9823, + "step": 3751 + }, + { + "epoch": 0.360908041554444, + "grad_norm": 2.2800471476351674, + "learning_rate": 1.4786274254128112e-05, + "loss": 1.0129, + "step": 3752 + }, + { + "epoch": 0.3610042323970758, + "grad_norm": 2.0365423135494147, + "learning_rate": 1.4783538624020642e-05, + "loss": 0.9483, + "step": 3753 + }, + { + "epoch": 0.3611004232397076, + "grad_norm": 2.274648041455947, + "learning_rate": 1.4780802529629559e-05, + "loss": 1.0778, + "step": 3754 + }, + { + "epoch": 0.3611966140823394, + "grad_norm": 1.912978750948554, + "learning_rate": 1.477806597122043e-05, + "loss": 0.9101, + "step": 3755 + }, + { + "epoch": 0.3612928049249711, + "grad_norm": 2.1915627547594942, + "learning_rate": 1.4775328949058856e-05, + "loss": 1.0421, + "step": 3756 + }, + { + "epoch": 0.3613889957676029, + "grad_norm": 3.4547331637777785, + "learning_rate": 1.4772591463410492e-05, + "loss": 0.9974, + "step": 3757 + }, + { + "epoch": 0.3614851866102347, + "grad_norm": 2.152811121534182, + "learning_rate": 1.4769853514541037e-05, + "loss": 1.0248, + "step": 3758 + }, + { + "epoch": 0.3615813774528665, + "grad_norm": 2.292492260770183, + "learning_rate": 1.4767115102716225e-05, + "loss": 1.0083, + "step": 3759 + }, + { + "epoch": 0.36167756829549824, + "grad_norm": 2.1437915532165173, + "learning_rate": 1.4764376228201848e-05, + "loss": 1.0958, + "step": 3760 + }, + { + "epoch": 0.36177375913813004, + "grad_norm": 2.2735951144192468, + "learning_rate": 1.476163689126374e-05, + "loss": 0.9571, + "step": 3761 + }, + { + "epoch": 0.36186994998076183, + "grad_norm": 2.0141322596236484, + "learning_rate": 1.475889709216777e-05, + "loss": 0.9564, + "step": 3762 + }, + { + "epoch": 0.3619661408233936, + "grad_norm": 2.242971897128811, + "learning_rate": 1.4756156831179864e-05, + "loss": 1.0356, + "step": 3763 + }, + { + "epoch": 0.3620623316660254, + "grad_norm": 2.044439341881917, + "learning_rate": 1.4753416108565985e-05, + "loss": 1.0425, + "step": 3764 + }, + { + "epoch": 0.36215852250865715, + "grad_norm": 1.8193714840480253, + "learning_rate": 1.4750674924592146e-05, + "loss": 0.9995, + "step": 3765 + }, + { + "epoch": 0.36225471335128895, + "grad_norm": 2.2089055877150354, + "learning_rate": 1.47479332795244e-05, + "loss": 0.8819, + "step": 3766 + }, + { + "epoch": 0.36235090419392074, + "grad_norm": 6.079841238711645, + "learning_rate": 1.4745191173628855e-05, + "loss": 1.2928, + "step": 3767 + }, + { + "epoch": 0.36244709503655254, + "grad_norm": 2.1903695964255965, + "learning_rate": 1.4742448607171644e-05, + "loss": 0.9928, + "step": 3768 + }, + { + "epoch": 0.36254328587918433, + "grad_norm": 2.4459529972462564, + "learning_rate": 1.4739705580418964e-05, + "loss": 1.0661, + "step": 3769 + }, + { + "epoch": 0.36263947672181607, + "grad_norm": 2.1478143914586565, + "learning_rate": 1.473696209363705e-05, + "loss": 1.0801, + "step": 3770 + }, + { + "epoch": 0.36273566756444786, + "grad_norm": 2.1634142741003815, + "learning_rate": 1.4734218147092177e-05, + "loss": 0.9409, + "step": 3771 + }, + { + "epoch": 0.36283185840707965, + "grad_norm": 2.5634209866750886, + "learning_rate": 1.4731473741050673e-05, + "loss": 0.9504, + "step": 3772 + }, + { + "epoch": 0.36292804924971145, + "grad_norm": 2.140315063507992, + "learning_rate": 1.4728728875778901e-05, + "loss": 1.0491, + "step": 3773 + }, + { + "epoch": 0.3630242400923432, + "grad_norm": 2.0937252961842936, + "learning_rate": 1.4725983551543279e-05, + "loss": 1.0726, + "step": 3774 + }, + { + "epoch": 0.363120430934975, + "grad_norm": 2.0266126003363922, + "learning_rate": 1.472323776861026e-05, + "loss": 1.074, + "step": 3775 + }, + { + "epoch": 0.3632166217776068, + "grad_norm": 1.917870422071144, + "learning_rate": 1.472049152724635e-05, + "loss": 1.0446, + "step": 3776 + }, + { + "epoch": 0.36331281262023857, + "grad_norm": 2.071422078676856, + "learning_rate": 1.4717744827718092e-05, + "loss": 1.0048, + "step": 3777 + }, + { + "epoch": 0.36340900346287036, + "grad_norm": 2.177173020303273, + "learning_rate": 1.471499767029208e-05, + "loss": 0.9453, + "step": 3778 + }, + { + "epoch": 0.3635051943055021, + "grad_norm": 1.8646567978561213, + "learning_rate": 1.4712250055234947e-05, + "loss": 0.9998, + "step": 3779 + }, + { + "epoch": 0.3636013851481339, + "grad_norm": 2.0478814552003253, + "learning_rate": 1.470950198281337e-05, + "loss": 0.9778, + "step": 3780 + }, + { + "epoch": 0.3636975759907657, + "grad_norm": 2.103809060409356, + "learning_rate": 1.470675345329408e-05, + "loss": 1.0191, + "step": 3781 + }, + { + "epoch": 0.3637937668333975, + "grad_norm": 2.256985659146981, + "learning_rate": 1.470400446694384e-05, + "loss": 1.0249, + "step": 3782 + }, + { + "epoch": 0.3638899576760292, + "grad_norm": 1.8890320413237558, + "learning_rate": 1.4701255024029464e-05, + "loss": 0.9547, + "step": 3783 + }, + { + "epoch": 0.363986148518661, + "grad_norm": 2.179467952888104, + "learning_rate": 1.4698505124817811e-05, + "loss": 0.9521, + "step": 3784 + }, + { + "epoch": 0.3640823393612928, + "grad_norm": 2.573091362981731, + "learning_rate": 1.4695754769575779e-05, + "loss": 0.9891, + "step": 3785 + }, + { + "epoch": 0.3641785302039246, + "grad_norm": 2.1999515936809098, + "learning_rate": 1.4693003958570318e-05, + "loss": 1.0091, + "step": 3786 + }, + { + "epoch": 0.3642747210465564, + "grad_norm": 1.905528944501368, + "learning_rate": 1.4690252692068415e-05, + "loss": 0.9945, + "step": 3787 + }, + { + "epoch": 0.36437091188918813, + "grad_norm": 2.0610992882084647, + "learning_rate": 1.4687500970337103e-05, + "loss": 1.0013, + "step": 3788 + }, + { + "epoch": 0.3644671027318199, + "grad_norm": 1.8632670707803318, + "learning_rate": 1.4684748793643464e-05, + "loss": 1.0662, + "step": 3789 + }, + { + "epoch": 0.3645632935744517, + "grad_norm": 2.153720171284905, + "learning_rate": 1.4681996162254618e-05, + "loss": 1.0031, + "step": 3790 + }, + { + "epoch": 0.3646594844170835, + "grad_norm": 2.0796531053297853, + "learning_rate": 1.4679243076437733e-05, + "loss": 1.0645, + "step": 3791 + }, + { + "epoch": 0.36475567525971525, + "grad_norm": 2.1829399911837615, + "learning_rate": 1.4676489536460015e-05, + "loss": 1.0522, + "step": 3792 + }, + { + "epoch": 0.36485186610234704, + "grad_norm": 2.1125720662696397, + "learning_rate": 1.4673735542588725e-05, + "loss": 0.9576, + "step": 3793 + }, + { + "epoch": 0.36494805694497884, + "grad_norm": 2.1767352959851727, + "learning_rate": 1.467098109509116e-05, + "loss": 1.059, + "step": 3794 + }, + { + "epoch": 0.36504424778761063, + "grad_norm": 2.0717516458941834, + "learning_rate": 1.466822619423466e-05, + "loss": 1.0353, + "step": 3795 + }, + { + "epoch": 0.3651404386302424, + "grad_norm": 2.137125351184283, + "learning_rate": 1.4665470840286614e-05, + "loss": 0.948, + "step": 3796 + }, + { + "epoch": 0.36523662947287416, + "grad_norm": 2.024650045313637, + "learning_rate": 1.4662715033514455e-05, + "loss": 1.0297, + "step": 3797 + }, + { + "epoch": 0.36533282031550596, + "grad_norm": 2.0298535088499046, + "learning_rate": 1.4659958774185654e-05, + "loss": 1.045, + "step": 3798 + }, + { + "epoch": 0.36542901115813775, + "grad_norm": 2.076614232084985, + "learning_rate": 1.465720206256773e-05, + "loss": 0.971, + "step": 3799 + }, + { + "epoch": 0.36552520200076954, + "grad_norm": 1.7440291725374168, + "learning_rate": 1.4654444898928249e-05, + "loss": 0.8994, + "step": 3800 + }, + { + "epoch": 0.3656213928434013, + "grad_norm": 1.991526693407607, + "learning_rate": 1.4651687283534814e-05, + "loss": 1.008, + "step": 3801 + }, + { + "epoch": 0.3657175836860331, + "grad_norm": 2.0576855545615618, + "learning_rate": 1.4648929216655077e-05, + "loss": 0.9924, + "step": 3802 + }, + { + "epoch": 0.36581377452866487, + "grad_norm": 1.9404886417359046, + "learning_rate": 1.4646170698556732e-05, + "loss": 0.9839, + "step": 3803 + }, + { + "epoch": 0.36590996537129666, + "grad_norm": 2.1613577933455432, + "learning_rate": 1.4643411729507517e-05, + "loss": 1.031, + "step": 3804 + }, + { + "epoch": 0.36600615621392846, + "grad_norm": 2.088409880537927, + "learning_rate": 1.4640652309775211e-05, + "loss": 1.0794, + "step": 3805 + }, + { + "epoch": 0.3661023470565602, + "grad_norm": 1.994713107542709, + "learning_rate": 1.4637892439627644e-05, + "loss": 1.0009, + "step": 3806 + }, + { + "epoch": 0.366198537899192, + "grad_norm": 2.040648179605054, + "learning_rate": 1.4635132119332684e-05, + "loss": 1.0026, + "step": 3807 + }, + { + "epoch": 0.3662947287418238, + "grad_norm": 2.052599788251669, + "learning_rate": 1.4632371349158241e-05, + "loss": 0.9597, + "step": 3808 + }, + { + "epoch": 0.3663909195844556, + "grad_norm": 1.8997331154117587, + "learning_rate": 1.4629610129372274e-05, + "loss": 0.9901, + "step": 3809 + }, + { + "epoch": 0.3664871104270873, + "grad_norm": 2.1530513858233915, + "learning_rate": 1.4626848460242782e-05, + "loss": 1.0892, + "step": 3810 + }, + { + "epoch": 0.3665833012697191, + "grad_norm": 2.0506804050765712, + "learning_rate": 1.4624086342037809e-05, + "loss": 1.0514, + "step": 3811 + }, + { + "epoch": 0.3666794921123509, + "grad_norm": 2.0975166531403513, + "learning_rate": 1.4621323775025444e-05, + "loss": 1.0689, + "step": 3812 + }, + { + "epoch": 0.3667756829549827, + "grad_norm": 4.638827078763731, + "learning_rate": 1.4618560759473815e-05, + "loss": 1.1002, + "step": 3813 + }, + { + "epoch": 0.3668718737976145, + "grad_norm": 2.2025708441081133, + "learning_rate": 1.4615797295651099e-05, + "loss": 0.9484, + "step": 3814 + }, + { + "epoch": 0.3669680646402462, + "grad_norm": 2.0515658889584447, + "learning_rate": 1.4613033383825512e-05, + "loss": 1.0289, + "step": 3815 + }, + { + "epoch": 0.367064255482878, + "grad_norm": 2.069562129548872, + "learning_rate": 1.4610269024265317e-05, + "loss": 0.9749, + "step": 3816 + }, + { + "epoch": 0.3671604463255098, + "grad_norm": 2.228843877374464, + "learning_rate": 1.4607504217238819e-05, + "loss": 1.0232, + "step": 3817 + }, + { + "epoch": 0.3672566371681416, + "grad_norm": 2.087064048709252, + "learning_rate": 1.4604738963014365e-05, + "loss": 0.9651, + "step": 3818 + }, + { + "epoch": 0.36735282801077335, + "grad_norm": 1.9356602675637709, + "learning_rate": 1.4601973261860347e-05, + "loss": 0.9147, + "step": 3819 + }, + { + "epoch": 0.36744901885340514, + "grad_norm": 2.3483848552090523, + "learning_rate": 1.4599207114045202e-05, + "loss": 1.1016, + "step": 3820 + }, + { + "epoch": 0.36754520969603693, + "grad_norm": 2.1249292770572477, + "learning_rate": 1.4596440519837405e-05, + "loss": 0.9927, + "step": 3821 + }, + { + "epoch": 0.36764140053866873, + "grad_norm": 1.897370627404638, + "learning_rate": 1.4593673479505482e-05, + "loss": 1.0014, + "step": 3822 + }, + { + "epoch": 0.3677375913813005, + "grad_norm": 2.2774430133524985, + "learning_rate": 1.4590905993317992e-05, + "loss": 1.0665, + "step": 3823 + }, + { + "epoch": 0.36783378222393226, + "grad_norm": 2.056963684010702, + "learning_rate": 1.4588138061543551e-05, + "loss": 0.8785, + "step": 3824 + }, + { + "epoch": 0.36792997306656405, + "grad_norm": 2.4636491443100397, + "learning_rate": 1.4585369684450804e-05, + "loss": 1.0371, + "step": 3825 + }, + { + "epoch": 0.36802616390919585, + "grad_norm": 2.1112063211576433, + "learning_rate": 1.458260086230845e-05, + "loss": 1.0168, + "step": 3826 + }, + { + "epoch": 0.36812235475182764, + "grad_norm": 2.4748804955248165, + "learning_rate": 1.4579831595385226e-05, + "loss": 0.9823, + "step": 3827 + }, + { + "epoch": 0.36821854559445943, + "grad_norm": 1.94786785747286, + "learning_rate": 1.4577061883949912e-05, + "loss": 0.9597, + "step": 3828 + }, + { + "epoch": 0.3683147364370912, + "grad_norm": 2.1591684864283875, + "learning_rate": 1.4574291728271333e-05, + "loss": 0.9861, + "step": 3829 + }, + { + "epoch": 0.36841092727972297, + "grad_norm": 1.866668894297865, + "learning_rate": 1.4571521128618358e-05, + "loss": 1.052, + "step": 3830 + }, + { + "epoch": 0.36850711812235476, + "grad_norm": 2.1685538113119747, + "learning_rate": 1.4568750085259895e-05, + "loss": 0.944, + "step": 3831 + }, + { + "epoch": 0.36860330896498655, + "grad_norm": 2.1061532900520312, + "learning_rate": 1.4565978598464895e-05, + "loss": 1.047, + "step": 3832 + }, + { + "epoch": 0.3686994998076183, + "grad_norm": 1.9222907971169354, + "learning_rate": 1.4563206668502366e-05, + "loss": 1.023, + "step": 3833 + }, + { + "epoch": 0.3687956906502501, + "grad_norm": 2.497016594687687, + "learning_rate": 1.4560434295641338e-05, + "loss": 1.1669, + "step": 3834 + }, + { + "epoch": 0.3688918814928819, + "grad_norm": 3.1585605651080373, + "learning_rate": 1.455766148015089e-05, + "loss": 1.4177, + "step": 3835 + }, + { + "epoch": 0.3689880723355137, + "grad_norm": 2.0125378129501152, + "learning_rate": 1.455488822230016e-05, + "loss": 0.9719, + "step": 3836 + }, + { + "epoch": 0.36908426317814547, + "grad_norm": 2.04888971183944, + "learning_rate": 1.4552114522358308e-05, + "loss": 0.9948, + "step": 3837 + }, + { + "epoch": 0.3691804540207772, + "grad_norm": 1.828466881501544, + "learning_rate": 1.4549340380594545e-05, + "loss": 1.0329, + "step": 3838 + }, + { + "epoch": 0.369276644863409, + "grad_norm": 2.1914616263677953, + "learning_rate": 1.4546565797278132e-05, + "loss": 1.0199, + "step": 3839 + }, + { + "epoch": 0.3693728357060408, + "grad_norm": 1.8657162569460632, + "learning_rate": 1.454379077267836e-05, + "loss": 0.9501, + "step": 3840 + }, + { + "epoch": 0.3694690265486726, + "grad_norm": 6.7591789467573005, + "learning_rate": 1.454101530706457e-05, + "loss": 1.1299, + "step": 3841 + }, + { + "epoch": 0.3695652173913043, + "grad_norm": 1.885121915625411, + "learning_rate": 1.4538239400706147e-05, + "loss": 0.8907, + "step": 3842 + }, + { + "epoch": 0.3696614082339361, + "grad_norm": 1.9093953623929085, + "learning_rate": 1.4535463053872514e-05, + "loss": 1.0189, + "step": 3843 + }, + { + "epoch": 0.3697575990765679, + "grad_norm": 2.057220118437633, + "learning_rate": 1.4532686266833143e-05, + "loss": 0.9281, + "step": 3844 + }, + { + "epoch": 0.3698537899191997, + "grad_norm": 1.9161527095683042, + "learning_rate": 1.4529909039857541e-05, + "loss": 0.8165, + "step": 3845 + }, + { + "epoch": 0.3699499807618315, + "grad_norm": 2.171860880715705, + "learning_rate": 1.4527131373215265e-05, + "loss": 0.9356, + "step": 3846 + }, + { + "epoch": 0.37004617160446324, + "grad_norm": 1.9466429734731248, + "learning_rate": 1.452435326717591e-05, + "loss": 0.9513, + "step": 3847 + }, + { + "epoch": 0.37014236244709503, + "grad_norm": 2.253715943632336, + "learning_rate": 1.4521574722009115e-05, + "loss": 1.0746, + "step": 3848 + }, + { + "epoch": 0.3702385532897268, + "grad_norm": 1.8318202944609026, + "learning_rate": 1.4518795737984559e-05, + "loss": 0.9873, + "step": 3849 + }, + { + "epoch": 0.3703347441323586, + "grad_norm": 2.12089011778784, + "learning_rate": 1.4516016315371974e-05, + "loss": 1.0058, + "step": 3850 + }, + { + "epoch": 0.37043093497499036, + "grad_norm": 2.263518687750939, + "learning_rate": 1.451323645444112e-05, + "loss": 0.9479, + "step": 3851 + }, + { + "epoch": 0.37052712581762215, + "grad_norm": 1.835296563849416, + "learning_rate": 1.4510456155461807e-05, + "loss": 1.0258, + "step": 3852 + }, + { + "epoch": 0.37062331666025394, + "grad_norm": 1.9535989946210186, + "learning_rate": 1.450767541870389e-05, + "loss": 1.0649, + "step": 3853 + }, + { + "epoch": 0.37071950750288574, + "grad_norm": 2.127204900715754, + "learning_rate": 1.4504894244437264e-05, + "loss": 1.045, + "step": 3854 + }, + { + "epoch": 0.37081569834551753, + "grad_norm": 2.1517884739217856, + "learning_rate": 1.4502112632931864e-05, + "loss": 1.0726, + "step": 3855 + }, + { + "epoch": 0.37091188918814927, + "grad_norm": 2.0071593849107354, + "learning_rate": 1.4499330584457667e-05, + "loss": 1.0626, + "step": 3856 + }, + { + "epoch": 0.37100808003078106, + "grad_norm": 2.085343475177728, + "learning_rate": 1.4496548099284698e-05, + "loss": 0.9709, + "step": 3857 + }, + { + "epoch": 0.37110427087341286, + "grad_norm": 2.2910263221477822, + "learning_rate": 1.4493765177683017e-05, + "loss": 0.9871, + "step": 3858 + }, + { + "epoch": 0.37120046171604465, + "grad_norm": 2.2014916312519106, + "learning_rate": 1.449098181992274e-05, + "loss": 0.9173, + "step": 3859 + }, + { + "epoch": 0.3712966525586764, + "grad_norm": 2.161693969672884, + "learning_rate": 1.4488198026274007e-05, + "loss": 1.0224, + "step": 3860 + }, + { + "epoch": 0.3713928434013082, + "grad_norm": 2.5705116726668886, + "learning_rate": 1.4485413797007008e-05, + "loss": 0.9772, + "step": 3861 + }, + { + "epoch": 0.37148903424394, + "grad_norm": 2.024942183116106, + "learning_rate": 1.4482629132391985e-05, + "loss": 1.0404, + "step": 3862 + }, + { + "epoch": 0.37158522508657177, + "grad_norm": 2.101950004975818, + "learning_rate": 1.4479844032699206e-05, + "loss": 0.9795, + "step": 3863 + }, + { + "epoch": 0.37168141592920356, + "grad_norm": 2.045390824354521, + "learning_rate": 1.4477058498198993e-05, + "loss": 0.9822, + "step": 3864 + }, + { + "epoch": 0.3717776067718353, + "grad_norm": 1.9177655791508312, + "learning_rate": 1.4474272529161704e-05, + "loss": 1.0098, + "step": 3865 + }, + { + "epoch": 0.3718737976144671, + "grad_norm": 2.2930892270247525, + "learning_rate": 1.4471486125857743e-05, + "loss": 0.9979, + "step": 3866 + }, + { + "epoch": 0.3719699884570989, + "grad_norm": 2.0197821061205277, + "learning_rate": 1.4468699288557554e-05, + "loss": 1.0704, + "step": 3867 + }, + { + "epoch": 0.3720661792997307, + "grad_norm": 2.4021937424860917, + "learning_rate": 1.446591201753162e-05, + "loss": 1.0473, + "step": 3868 + }, + { + "epoch": 0.3721623701423624, + "grad_norm": 2.21574337200679, + "learning_rate": 1.4463124313050476e-05, + "loss": 1.0843, + "step": 3869 + }, + { + "epoch": 0.3722585609849942, + "grad_norm": 1.9535339336304753, + "learning_rate": 1.4460336175384688e-05, + "loss": 0.9672, + "step": 3870 + }, + { + "epoch": 0.372354751827626, + "grad_norm": 2.591694642111291, + "learning_rate": 1.445754760480487e-05, + "loss": 0.9717, + "step": 3871 + }, + { + "epoch": 0.3724509426702578, + "grad_norm": 1.8782830148875589, + "learning_rate": 1.4454758601581675e-05, + "loss": 1.0621, + "step": 3872 + }, + { + "epoch": 0.3725471335128896, + "grad_norm": 2.170872740692677, + "learning_rate": 1.4451969165985808e-05, + "loss": 0.9988, + "step": 3873 + }, + { + "epoch": 0.37264332435552133, + "grad_norm": 1.7646451279620552, + "learning_rate": 1.4449179298287999e-05, + "loss": 0.8654, + "step": 3874 + }, + { + "epoch": 0.3727395151981531, + "grad_norm": 2.076742071948776, + "learning_rate": 1.4446388998759027e-05, + "loss": 0.9619, + "step": 3875 + }, + { + "epoch": 0.3728357060407849, + "grad_norm": 2.184911047980085, + "learning_rate": 1.4443598267669723e-05, + "loss": 1.0113, + "step": 3876 + }, + { + "epoch": 0.3729318968834167, + "grad_norm": 1.9373707262275341, + "learning_rate": 1.4440807105290944e-05, + "loss": 1.0595, + "step": 3877 + }, + { + "epoch": 0.3730280877260485, + "grad_norm": 2.171647438871899, + "learning_rate": 1.4438015511893602e-05, + "loss": 0.9723, + "step": 3878 + }, + { + "epoch": 0.37312427856868025, + "grad_norm": 2.0886519757774757, + "learning_rate": 1.4435223487748644e-05, + "loss": 1.1482, + "step": 3879 + }, + { + "epoch": 0.37322046941131204, + "grad_norm": 2.135511930074801, + "learning_rate": 1.4432431033127056e-05, + "loss": 1.0572, + "step": 3880 + }, + { + "epoch": 0.37331666025394383, + "grad_norm": 2.00867750299333, + "learning_rate": 1.4429638148299874e-05, + "loss": 1.1209, + "step": 3881 + }, + { + "epoch": 0.3734128510965756, + "grad_norm": 1.8619806572686537, + "learning_rate": 1.442684483353817e-05, + "loss": 1.0506, + "step": 3882 + }, + { + "epoch": 0.37350904193920736, + "grad_norm": 1.9827942753847279, + "learning_rate": 1.4424051089113057e-05, + "loss": 1.0994, + "step": 3883 + }, + { + "epoch": 0.37360523278183916, + "grad_norm": 2.0261735857861747, + "learning_rate": 1.4421256915295697e-05, + "loss": 1.1212, + "step": 3884 + }, + { + "epoch": 0.37370142362447095, + "grad_norm": 2.0434389077185635, + "learning_rate": 1.4418462312357286e-05, + "loss": 1.0443, + "step": 3885 + }, + { + "epoch": 0.37379761446710275, + "grad_norm": 2.1484140836898575, + "learning_rate": 1.4415667280569064e-05, + "loss": 1.0035, + "step": 3886 + }, + { + "epoch": 0.37389380530973454, + "grad_norm": 1.7887649186879475, + "learning_rate": 1.4412871820202309e-05, + "loss": 0.9948, + "step": 3887 + }, + { + "epoch": 0.3739899961523663, + "grad_norm": 1.9805894467425593, + "learning_rate": 1.4410075931528356e-05, + "loss": 0.9169, + "step": 3888 + }, + { + "epoch": 0.37408618699499807, + "grad_norm": 2.0302370063446147, + "learning_rate": 1.4407279614818554e-05, + "loss": 0.9865, + "step": 3889 + }, + { + "epoch": 0.37418237783762986, + "grad_norm": 5.711516288753443, + "learning_rate": 1.4404482870344322e-05, + "loss": 0.9303, + "step": 3890 + }, + { + "epoch": 0.37427856868026166, + "grad_norm": 1.9990448250954485, + "learning_rate": 1.4401685698377108e-05, + "loss": 0.9965, + "step": 3891 + }, + { + "epoch": 0.3743747595228934, + "grad_norm": 1.9212960619273076, + "learning_rate": 1.4398888099188396e-05, + "loss": 1.0203, + "step": 3892 + }, + { + "epoch": 0.3744709503655252, + "grad_norm": 2.1481632206821297, + "learning_rate": 1.4396090073049717e-05, + "loss": 0.9943, + "step": 3893 + }, + { + "epoch": 0.374567141208157, + "grad_norm": 1.931167797736005, + "learning_rate": 1.4393291620232646e-05, + "loss": 1.1246, + "step": 3894 + }, + { + "epoch": 0.3746633320507888, + "grad_norm": 1.9542076074590338, + "learning_rate": 1.4390492741008797e-05, + "loss": 1.0043, + "step": 3895 + }, + { + "epoch": 0.37475952289342057, + "grad_norm": 2.059633905513498, + "learning_rate": 1.4387693435649826e-05, + "loss": 0.9847, + "step": 3896 + }, + { + "epoch": 0.3748557137360523, + "grad_norm": 2.2279177426598964, + "learning_rate": 1.4384893704427427e-05, + "loss": 1.1438, + "step": 3897 + }, + { + "epoch": 0.3749519045786841, + "grad_norm": 1.9453037998373373, + "learning_rate": 1.4382093547613338e-05, + "loss": 0.9972, + "step": 3898 + }, + { + "epoch": 0.3750480954213159, + "grad_norm": 2.1286481280501572, + "learning_rate": 1.4379292965479346e-05, + "loss": 1.0367, + "step": 3899 + }, + { + "epoch": 0.3751442862639477, + "grad_norm": 5.5440538040178495, + "learning_rate": 1.4376491958297263e-05, + "loss": 1.0231, + "step": 3900 + }, + { + "epoch": 0.37524047710657943, + "grad_norm": 2.0104178685557934, + "learning_rate": 1.4373690526338949e-05, + "loss": 0.9989, + "step": 3901 + }, + { + "epoch": 0.3753366679492112, + "grad_norm": 2.097637634166994, + "learning_rate": 1.4370888669876317e-05, + "loss": 1.0756, + "step": 3902 + }, + { + "epoch": 0.375432858791843, + "grad_norm": 2.1277265053404286, + "learning_rate": 1.43680863891813e-05, + "loss": 0.972, + "step": 3903 + }, + { + "epoch": 0.3755290496344748, + "grad_norm": 2.1530707245109983, + "learning_rate": 1.4365283684525895e-05, + "loss": 1.0672, + "step": 3904 + }, + { + "epoch": 0.3756252404771066, + "grad_norm": 2.123653254732934, + "learning_rate": 1.4362480556182118e-05, + "loss": 1.0514, + "step": 3905 + }, + { + "epoch": 0.37572143131973834, + "grad_norm": 1.8559544475975458, + "learning_rate": 1.4359677004422045e-05, + "loss": 1.0271, + "step": 3906 + }, + { + "epoch": 0.37581762216237014, + "grad_norm": 1.880870347323274, + "learning_rate": 1.4356873029517781e-05, + "loss": 0.9356, + "step": 3907 + }, + { + "epoch": 0.37591381300500193, + "grad_norm": 2.0399690746072365, + "learning_rate": 1.4354068631741476e-05, + "loss": 1.0105, + "step": 3908 + }, + { + "epoch": 0.3760100038476337, + "grad_norm": 2.1381285166166073, + "learning_rate": 1.4351263811365321e-05, + "loss": 0.9717, + "step": 3909 + }, + { + "epoch": 0.37610619469026546, + "grad_norm": 2.0401258434747542, + "learning_rate": 1.4348458568661548e-05, + "loss": 1.107, + "step": 3910 + }, + { + "epoch": 0.37620238553289725, + "grad_norm": 1.9408862441243917, + "learning_rate": 1.4345652903902432e-05, + "loss": 0.9596, + "step": 3911 + }, + { + "epoch": 0.37629857637552905, + "grad_norm": 1.9009983993913644, + "learning_rate": 1.434284681736028e-05, + "loss": 1.039, + "step": 3912 + }, + { + "epoch": 0.37639476721816084, + "grad_norm": 1.9502725872485556, + "learning_rate": 1.4340040309307456e-05, + "loss": 1.0929, + "step": 3913 + }, + { + "epoch": 0.37649095806079264, + "grad_norm": 2.7617862955858135, + "learning_rate": 1.4337233380016354e-05, + "loss": 0.9847, + "step": 3914 + }, + { + "epoch": 0.3765871489034244, + "grad_norm": 2.114501199357822, + "learning_rate": 1.4334426029759402e-05, + "loss": 0.9657, + "step": 3915 + }, + { + "epoch": 0.37668333974605617, + "grad_norm": 2.161448202578013, + "learning_rate": 1.433161825880909e-05, + "loss": 1.0094, + "step": 3916 + }, + { + "epoch": 0.37677953058868796, + "grad_norm": 1.9772394365792814, + "learning_rate": 1.4328810067437929e-05, + "loss": 0.9931, + "step": 3917 + }, + { + "epoch": 0.37687572143131975, + "grad_norm": 1.8067177730486923, + "learning_rate": 1.432600145591848e-05, + "loss": 1.0345, + "step": 3918 + }, + { + "epoch": 0.3769719122739515, + "grad_norm": 1.9469772757820922, + "learning_rate": 1.432319242452334e-05, + "loss": 1.0162, + "step": 3919 + }, + { + "epoch": 0.3770681031165833, + "grad_norm": 2.020572176674278, + "learning_rate": 1.4320382973525151e-05, + "loss": 0.9627, + "step": 3920 + }, + { + "epoch": 0.3771642939592151, + "grad_norm": 2.4284021771844237, + "learning_rate": 1.4317573103196598e-05, + "loss": 1.1121, + "step": 3921 + }, + { + "epoch": 0.3772604848018469, + "grad_norm": 2.042067332606612, + "learning_rate": 1.43147628138104e-05, + "loss": 0.9989, + "step": 3922 + }, + { + "epoch": 0.37735667564447867, + "grad_norm": 1.8351491681920857, + "learning_rate": 1.4311952105639322e-05, + "loss": 0.995, + "step": 3923 + }, + { + "epoch": 0.3774528664871104, + "grad_norm": 2.0640578154761102, + "learning_rate": 1.4309140978956161e-05, + "loss": 1.1084, + "step": 3924 + }, + { + "epoch": 0.3775490573297422, + "grad_norm": 1.9427275222145313, + "learning_rate": 1.4306329434033772e-05, + "loss": 1.0603, + "step": 3925 + }, + { + "epoch": 0.377645248172374, + "grad_norm": 2.0736777983793884, + "learning_rate": 1.430351747114503e-05, + "loss": 0.9848, + "step": 3926 + }, + { + "epoch": 0.3777414390150058, + "grad_norm": 1.7308349546226702, + "learning_rate": 1.4300705090562862e-05, + "loss": 0.9417, + "step": 3927 + }, + { + "epoch": 0.3778376298576376, + "grad_norm": 1.97272608368643, + "learning_rate": 1.429789229256024e-05, + "loss": 1.0179, + "step": 3928 + }, + { + "epoch": 0.3779338207002693, + "grad_norm": 2.102640233942354, + "learning_rate": 1.429507907741016e-05, + "loss": 0.9515, + "step": 3929 + }, + { + "epoch": 0.3780300115429011, + "grad_norm": 1.9929172586555794, + "learning_rate": 1.429226544538568e-05, + "loss": 0.9763, + "step": 3930 + }, + { + "epoch": 0.3781262023855329, + "grad_norm": 2.1500084286800405, + "learning_rate": 1.4289451396759879e-05, + "loss": 1.0337, + "step": 3931 + }, + { + "epoch": 0.3782223932281647, + "grad_norm": 2.1249206681581514, + "learning_rate": 1.4286636931805887e-05, + "loss": 1.0004, + "step": 3932 + }, + { + "epoch": 0.37831858407079644, + "grad_norm": 2.0266157957990703, + "learning_rate": 1.4283822050796875e-05, + "loss": 1.0009, + "step": 3933 + }, + { + "epoch": 0.37841477491342823, + "grad_norm": 1.9900381603201889, + "learning_rate": 1.4281006754006045e-05, + "loss": 1.0096, + "step": 3934 + }, + { + "epoch": 0.37851096575606, + "grad_norm": 2.1017281249328725, + "learning_rate": 1.427819104170665e-05, + "loss": 1.0506, + "step": 3935 + }, + { + "epoch": 0.3786071565986918, + "grad_norm": 1.8352495841603134, + "learning_rate": 1.427537491417198e-05, + "loss": 0.9601, + "step": 3936 + }, + { + "epoch": 0.3787033474413236, + "grad_norm": 1.9642870142358124, + "learning_rate": 1.4272558371675365e-05, + "loss": 0.9937, + "step": 3937 + }, + { + "epoch": 0.37879953828395535, + "grad_norm": 1.8242394088785723, + "learning_rate": 1.426974141449017e-05, + "loss": 0.9262, + "step": 3938 + }, + { + "epoch": 0.37889572912658714, + "grad_norm": 2.0164283538413583, + "learning_rate": 1.4266924042889808e-05, + "loss": 0.8984, + "step": 3939 + }, + { + "epoch": 0.37899191996921894, + "grad_norm": 1.9169738212784675, + "learning_rate": 1.4264106257147732e-05, + "loss": 0.9826, + "step": 3940 + }, + { + "epoch": 0.37908811081185073, + "grad_norm": 2.10300640049495, + "learning_rate": 1.4261288057537426e-05, + "loss": 1.0515, + "step": 3941 + }, + { + "epoch": 0.37918430165448247, + "grad_norm": 2.0969338194042955, + "learning_rate": 1.4258469444332423e-05, + "loss": 1.0583, + "step": 3942 + }, + { + "epoch": 0.37928049249711426, + "grad_norm": 1.8539487850537726, + "learning_rate": 1.4255650417806299e-05, + "loss": 0.9751, + "step": 3943 + }, + { + "epoch": 0.37937668333974606, + "grad_norm": 1.9662077085088192, + "learning_rate": 1.4252830978232658e-05, + "loss": 0.9304, + "step": 3944 + }, + { + "epoch": 0.37947287418237785, + "grad_norm": 1.9707499144708918, + "learning_rate": 1.4250011125885153e-05, + "loss": 0.9695, + "step": 3945 + }, + { + "epoch": 0.37956906502500964, + "grad_norm": 2.334506227142793, + "learning_rate": 1.4247190861037474e-05, + "loss": 0.9609, + "step": 3946 + }, + { + "epoch": 0.3796652558676414, + "grad_norm": 1.918090790867311, + "learning_rate": 1.4244370183963357e-05, + "loss": 1.0146, + "step": 3947 + }, + { + "epoch": 0.3797614467102732, + "grad_norm": 2.030782684131641, + "learning_rate": 1.4241549094936567e-05, + "loss": 0.9944, + "step": 3948 + }, + { + "epoch": 0.37985763755290497, + "grad_norm": 2.0935222667435025, + "learning_rate": 1.4238727594230914e-05, + "loss": 0.9442, + "step": 3949 + }, + { + "epoch": 0.37995382839553676, + "grad_norm": 1.906396711415354, + "learning_rate": 1.4235905682120255e-05, + "loss": 0.9973, + "step": 3950 + }, + { + "epoch": 0.3800500192381685, + "grad_norm": 2.116923729293029, + "learning_rate": 1.423308335887848e-05, + "loss": 1.1225, + "step": 3951 + }, + { + "epoch": 0.3801462100808003, + "grad_norm": 2.0194975713295387, + "learning_rate": 1.4230260624779512e-05, + "loss": 0.9372, + "step": 3952 + }, + { + "epoch": 0.3802424009234321, + "grad_norm": 2.1464633884300692, + "learning_rate": 1.4227437480097332e-05, + "loss": 1.016, + "step": 3953 + }, + { + "epoch": 0.3803385917660639, + "grad_norm": 2.28531740124822, + "learning_rate": 1.4224613925105947e-05, + "loss": 1.0332, + "step": 3954 + }, + { + "epoch": 0.3804347826086957, + "grad_norm": 1.930019335181937, + "learning_rate": 1.4221789960079403e-05, + "loss": 1.0033, + "step": 3955 + }, + { + "epoch": 0.3805309734513274, + "grad_norm": 2.1886830578313656, + "learning_rate": 1.4218965585291792e-05, + "loss": 1.116, + "step": 3956 + }, + { + "epoch": 0.3806271642939592, + "grad_norm": 2.289676577442772, + "learning_rate": 1.421614080101725e-05, + "loss": 1.0726, + "step": 3957 + }, + { + "epoch": 0.380723355136591, + "grad_norm": 2.032241595752511, + "learning_rate": 1.4213315607529939e-05, + "loss": 1.0423, + "step": 3958 + }, + { + "epoch": 0.3808195459792228, + "grad_norm": 2.142082175342106, + "learning_rate": 1.4210490005104071e-05, + "loss": 1.0186, + "step": 3959 + }, + { + "epoch": 0.38091573682185453, + "grad_norm": 3.5317616649184016, + "learning_rate": 1.4207663994013896e-05, + "loss": 1.127, + "step": 3960 + }, + { + "epoch": 0.3810119276644863, + "grad_norm": 2.554935274821424, + "learning_rate": 1.4204837574533703e-05, + "loss": 1.137, + "step": 3961 + }, + { + "epoch": 0.3811081185071181, + "grad_norm": 2.166930754735197, + "learning_rate": 1.4202010746937815e-05, + "loss": 1.043, + "step": 3962 + }, + { + "epoch": 0.3812043093497499, + "grad_norm": 2.3341542801989887, + "learning_rate": 1.419918351150061e-05, + "loss": 0.9874, + "step": 3963 + }, + { + "epoch": 0.3813005001923817, + "grad_norm": 1.9759601543659466, + "learning_rate": 1.4196355868496485e-05, + "loss": 0.9941, + "step": 3964 + }, + { + "epoch": 0.38139669103501345, + "grad_norm": 2.1014402062825157, + "learning_rate": 1.4193527818199893e-05, + "loss": 1.0821, + "step": 3965 + }, + { + "epoch": 0.38149288187764524, + "grad_norm": 1.8146733833803939, + "learning_rate": 1.4190699360885323e-05, + "loss": 0.9716, + "step": 3966 + }, + { + "epoch": 0.38158907272027703, + "grad_norm": 2.5647076628053553, + "learning_rate": 1.4187870496827294e-05, + "loss": 0.9671, + "step": 3967 + }, + { + "epoch": 0.3816852635629088, + "grad_norm": 2.337679187430277, + "learning_rate": 1.4185041226300376e-05, + "loss": 1.0351, + "step": 3968 + }, + { + "epoch": 0.38178145440554057, + "grad_norm": 1.8197080848005178, + "learning_rate": 1.4182211549579174e-05, + "loss": 0.9933, + "step": 3969 + }, + { + "epoch": 0.38187764524817236, + "grad_norm": 1.9606603024677889, + "learning_rate": 1.4179381466938332e-05, + "loss": 1.1281, + "step": 3970 + }, + { + "epoch": 0.38197383609080415, + "grad_norm": 2.0770634923753195, + "learning_rate": 1.4176550978652532e-05, + "loss": 0.9273, + "step": 3971 + }, + { + "epoch": 0.38207002693343595, + "grad_norm": 2.001794464642446, + "learning_rate": 1.4173720084996501e-05, + "loss": 0.9628, + "step": 3972 + }, + { + "epoch": 0.38216621777606774, + "grad_norm": 1.983138392344239, + "learning_rate": 1.4170888786244998e-05, + "loss": 0.9645, + "step": 3973 + }, + { + "epoch": 0.3822624086186995, + "grad_norm": 2.2492464899428812, + "learning_rate": 1.4168057082672828e-05, + "loss": 0.9797, + "step": 3974 + }, + { + "epoch": 0.38235859946133127, + "grad_norm": 2.1453915995062687, + "learning_rate": 1.416522497455483e-05, + "loss": 0.9994, + "step": 3975 + }, + { + "epoch": 0.38245479030396307, + "grad_norm": 2.0741223326932445, + "learning_rate": 1.4162392462165884e-05, + "loss": 0.9823, + "step": 3976 + }, + { + "epoch": 0.38255098114659486, + "grad_norm": 2.3843906413944427, + "learning_rate": 1.4159559545780914e-05, + "loss": 1.0505, + "step": 3977 + }, + { + "epoch": 0.38264717198922665, + "grad_norm": 2.1012624140593394, + "learning_rate": 1.4156726225674874e-05, + "loss": 1.0512, + "step": 3978 + }, + { + "epoch": 0.3827433628318584, + "grad_norm": 1.7653430761792885, + "learning_rate": 1.415389250212276e-05, + "loss": 1.0052, + "step": 3979 + }, + { + "epoch": 0.3828395536744902, + "grad_norm": 2.113948068268183, + "learning_rate": 1.415105837539962e-05, + "loss": 1.0044, + "step": 3980 + }, + { + "epoch": 0.382935744517122, + "grad_norm": 1.9132469149694333, + "learning_rate": 1.414822384578052e-05, + "loss": 1.0674, + "step": 3981 + }, + { + "epoch": 0.3830319353597538, + "grad_norm": 2.0761617196196696, + "learning_rate": 1.414538891354058e-05, + "loss": 1.0444, + "step": 3982 + }, + { + "epoch": 0.3831281262023855, + "grad_norm": 2.117038216805687, + "learning_rate": 1.4142553578954953e-05, + "loss": 0.9708, + "step": 3983 + }, + { + "epoch": 0.3832243170450173, + "grad_norm": 1.854543124338176, + "learning_rate": 1.4139717842298835e-05, + "loss": 1.0121, + "step": 3984 + }, + { + "epoch": 0.3833205078876491, + "grad_norm": 1.8983977580611489, + "learning_rate": 1.4136881703847455e-05, + "loss": 1.0063, + "step": 3985 + }, + { + "epoch": 0.3834166987302809, + "grad_norm": 2.1648744259998, + "learning_rate": 1.4134045163876086e-05, + "loss": 1.0815, + "step": 3986 + }, + { + "epoch": 0.3835128895729127, + "grad_norm": 2.028363462319873, + "learning_rate": 1.4131208222660042e-05, + "loss": 1.0549, + "step": 3987 + }, + { + "epoch": 0.3836090804155444, + "grad_norm": 1.9615050188698622, + "learning_rate": 1.4128370880474667e-05, + "loss": 1.0372, + "step": 3988 + }, + { + "epoch": 0.3837052712581762, + "grad_norm": 2.067809680316338, + "learning_rate": 1.4125533137595358e-05, + "loss": 0.981, + "step": 3989 + }, + { + "epoch": 0.383801462100808, + "grad_norm": 2.3734509018273173, + "learning_rate": 1.412269499429753e-05, + "loss": 1.1161, + "step": 3990 + }, + { + "epoch": 0.3838976529434398, + "grad_norm": 1.8601989110089463, + "learning_rate": 1.4119856450856657e-05, + "loss": 0.8376, + "step": 3991 + }, + { + "epoch": 0.38399384378607154, + "grad_norm": 1.9184007511177683, + "learning_rate": 1.4117017507548244e-05, + "loss": 0.9673, + "step": 3992 + }, + { + "epoch": 0.38409003462870334, + "grad_norm": 2.162706513767995, + "learning_rate": 1.4114178164647836e-05, + "loss": 1.0193, + "step": 3993 + }, + { + "epoch": 0.38418622547133513, + "grad_norm": 1.9545159750025034, + "learning_rate": 1.4111338422431013e-05, + "loss": 0.9319, + "step": 3994 + }, + { + "epoch": 0.3842824163139669, + "grad_norm": 2.90151962936319, + "learning_rate": 1.4108498281173395e-05, + "loss": 0.8218, + "step": 3995 + }, + { + "epoch": 0.3843786071565987, + "grad_norm": 2.1976971928696365, + "learning_rate": 1.4105657741150648e-05, + "loss": 0.9169, + "step": 3996 + }, + { + "epoch": 0.38447479799923046, + "grad_norm": 2.120631174872339, + "learning_rate": 1.4102816802638462e-05, + "loss": 0.9022, + "step": 3997 + }, + { + "epoch": 0.38457098884186225, + "grad_norm": 2.0909113213337247, + "learning_rate": 1.4099975465912584e-05, + "loss": 1.0183, + "step": 3998 + }, + { + "epoch": 0.38466717968449404, + "grad_norm": 1.906055247963866, + "learning_rate": 1.4097133731248783e-05, + "loss": 0.9186, + "step": 3999 + }, + { + "epoch": 0.38476337052712584, + "grad_norm": 2.1127539670466957, + "learning_rate": 1.4094291598922877e-05, + "loss": 1.0527, + "step": 4000 + }, + { + "epoch": 0.3848595613697576, + "grad_norm": 2.0924288627444265, + "learning_rate": 1.4091449069210721e-05, + "loss": 0.967, + "step": 4001 + }, + { + "epoch": 0.38495575221238937, + "grad_norm": 1.8089441614941852, + "learning_rate": 1.40886061423882e-05, + "loss": 1.0599, + "step": 4002 + }, + { + "epoch": 0.38505194305502116, + "grad_norm": 1.8286842515827202, + "learning_rate": 1.4085762818731255e-05, + "loss": 0.9156, + "step": 4003 + }, + { + "epoch": 0.38514813389765296, + "grad_norm": 2.013867548267657, + "learning_rate": 1.4082919098515846e-05, + "loss": 0.8521, + "step": 4004 + }, + { + "epoch": 0.38524432474028475, + "grad_norm": 1.8639502434227082, + "learning_rate": 1.4080074982017986e-05, + "loss": 0.943, + "step": 4005 + }, + { + "epoch": 0.3853405155829165, + "grad_norm": 2.0080227919132563, + "learning_rate": 1.407723046951372e-05, + "loss": 1.106, + "step": 4006 + }, + { + "epoch": 0.3854367064255483, + "grad_norm": 2.0646591277561335, + "learning_rate": 1.407438556127913e-05, + "loss": 1.0093, + "step": 4007 + }, + { + "epoch": 0.3855328972681801, + "grad_norm": 2.007934302931956, + "learning_rate": 1.4071540257590341e-05, + "loss": 1.051, + "step": 4008 + }, + { + "epoch": 0.38562908811081187, + "grad_norm": 2.3867116132789072, + "learning_rate": 1.4068694558723515e-05, + "loss": 1.0758, + "step": 4009 + }, + { + "epoch": 0.3857252789534436, + "grad_norm": 4.361413366993087, + "learning_rate": 1.4065848464954848e-05, + "loss": 0.8773, + "step": 4010 + }, + { + "epoch": 0.3858214697960754, + "grad_norm": 2.0223047518023427, + "learning_rate": 1.406300197656058e-05, + "loss": 1.0065, + "step": 4011 + }, + { + "epoch": 0.3859176606387072, + "grad_norm": 2.020804733470196, + "learning_rate": 1.4060155093816988e-05, + "loss": 0.8865, + "step": 4012 + }, + { + "epoch": 0.386013851481339, + "grad_norm": 2.0515607444419497, + "learning_rate": 1.4057307817000385e-05, + "loss": 1.0343, + "step": 4013 + }, + { + "epoch": 0.3861100423239708, + "grad_norm": 2.0625946088674687, + "learning_rate": 1.4054460146387124e-05, + "loss": 1.045, + "step": 4014 + }, + { + "epoch": 0.3862062331666025, + "grad_norm": 2.1680279415113124, + "learning_rate": 1.40516120822536e-05, + "loss": 1.0304, + "step": 4015 + }, + { + "epoch": 0.3863024240092343, + "grad_norm": 1.9541683269527497, + "learning_rate": 1.4048763624876233e-05, + "loss": 1.0037, + "step": 4016 + }, + { + "epoch": 0.3863986148518661, + "grad_norm": 1.8335986500964996, + "learning_rate": 1.40459147745315e-05, + "loss": 0.9524, + "step": 4017 + }, + { + "epoch": 0.3864948056944979, + "grad_norm": 2.0801420605484915, + "learning_rate": 1.4043065531495904e-05, + "loss": 1.0726, + "step": 4018 + }, + { + "epoch": 0.38659099653712964, + "grad_norm": 1.9995504525184802, + "learning_rate": 1.4040215896045983e-05, + "loss": 0.923, + "step": 4019 + }, + { + "epoch": 0.38668718737976143, + "grad_norm": 2.0499524533269473, + "learning_rate": 1.4037365868458325e-05, + "loss": 1.0513, + "step": 4020 + }, + { + "epoch": 0.3867833782223932, + "grad_norm": 1.9134862910269235, + "learning_rate": 1.4034515449009547e-05, + "loss": 1.0007, + "step": 4021 + }, + { + "epoch": 0.386879569065025, + "grad_norm": 2.1591041402677815, + "learning_rate": 1.4031664637976305e-05, + "loss": 0.9628, + "step": 4022 + }, + { + "epoch": 0.3869757599076568, + "grad_norm": 1.938629787883233, + "learning_rate": 1.40288134356353e-05, + "loss": 0.9661, + "step": 4023 + }, + { + "epoch": 0.38707195075028855, + "grad_norm": 1.936787132936258, + "learning_rate": 1.402596184226326e-05, + "loss": 0.9935, + "step": 4024 + }, + { + "epoch": 0.38716814159292035, + "grad_norm": 1.8301739539784725, + "learning_rate": 1.402310985813696e-05, + "loss": 0.8918, + "step": 4025 + }, + { + "epoch": 0.38726433243555214, + "grad_norm": 2.074579410078967, + "learning_rate": 1.4020257483533208e-05, + "loss": 1.0473, + "step": 4026 + }, + { + "epoch": 0.38736052327818393, + "grad_norm": 1.9132312629979593, + "learning_rate": 1.4017404718728855e-05, + "loss": 0.9715, + "step": 4027 + }, + { + "epoch": 0.38745671412081567, + "grad_norm": 2.1711124786067733, + "learning_rate": 1.401455156400078e-05, + "loss": 0.9977, + "step": 4028 + }, + { + "epoch": 0.38755290496344746, + "grad_norm": 2.214492183188635, + "learning_rate": 1.4011698019625914e-05, + "loss": 1.0781, + "step": 4029 + }, + { + "epoch": 0.38764909580607926, + "grad_norm": 2.462019087958773, + "learning_rate": 1.400884408588121e-05, + "loss": 0.9549, + "step": 4030 + }, + { + "epoch": 0.38774528664871105, + "grad_norm": 2.0448087678035844, + "learning_rate": 1.4005989763043672e-05, + "loss": 0.9676, + "step": 4031 + }, + { + "epoch": 0.38784147749134285, + "grad_norm": 2.2279991400136554, + "learning_rate": 1.400313505139034e-05, + "loss": 1.0342, + "step": 4032 + }, + { + "epoch": 0.3879376683339746, + "grad_norm": 8.222987980517473, + "learning_rate": 1.4000279951198279e-05, + "loss": 0.9265, + "step": 4033 + }, + { + "epoch": 0.3880338591766064, + "grad_norm": 2.263448074260342, + "learning_rate": 1.3997424462744607e-05, + "loss": 1.0216, + "step": 4034 + }, + { + "epoch": 0.38813005001923817, + "grad_norm": 1.9961450017917732, + "learning_rate": 1.3994568586306474e-05, + "loss": 1.008, + "step": 4035 + }, + { + "epoch": 0.38822624086186996, + "grad_norm": 2.068727456908949, + "learning_rate": 1.3991712322161065e-05, + "loss": 0.9961, + "step": 4036 + }, + { + "epoch": 0.38832243170450176, + "grad_norm": 1.8366253923587645, + "learning_rate": 1.3988855670585606e-05, + "loss": 1.0163, + "step": 4037 + }, + { + "epoch": 0.3884186225471335, + "grad_norm": 2.043899386591496, + "learning_rate": 1.3985998631857359e-05, + "loss": 1.0476, + "step": 4038 + }, + { + "epoch": 0.3885148133897653, + "grad_norm": 2.53862388961861, + "learning_rate": 1.3983141206253626e-05, + "loss": 1.0416, + "step": 4039 + }, + { + "epoch": 0.3886110042323971, + "grad_norm": 1.9233498377344624, + "learning_rate": 1.398028339405174e-05, + "loss": 0.9131, + "step": 4040 + }, + { + "epoch": 0.3887071950750289, + "grad_norm": 1.699178152474687, + "learning_rate": 1.3977425195529087e-05, + "loss": 0.929, + "step": 4041 + }, + { + "epoch": 0.3888033859176606, + "grad_norm": 2.261456658308898, + "learning_rate": 1.3974566610963068e-05, + "loss": 0.9624, + "step": 4042 + }, + { + "epoch": 0.3888995767602924, + "grad_norm": 2.231435924174746, + "learning_rate": 1.3971707640631138e-05, + "loss": 1.0328, + "step": 4043 + }, + { + "epoch": 0.3889957676029242, + "grad_norm": 1.90318974830293, + "learning_rate": 1.3968848284810785e-05, + "loss": 1.0621, + "step": 4044 + }, + { + "epoch": 0.389091958445556, + "grad_norm": 1.9612680777455895, + "learning_rate": 1.3965988543779533e-05, + "loss": 1.0277, + "step": 4045 + }, + { + "epoch": 0.3891881492881878, + "grad_norm": 2.040880696403343, + "learning_rate": 1.3963128417814951e-05, + "loss": 1.0257, + "step": 4046 + }, + { + "epoch": 0.38928434013081953, + "grad_norm": 2.0003608951327068, + "learning_rate": 1.3960267907194625e-05, + "loss": 1.0997, + "step": 4047 + }, + { + "epoch": 0.3893805309734513, + "grad_norm": 1.9772192402739432, + "learning_rate": 1.3957407012196204e-05, + "loss": 1.0335, + "step": 4048 + }, + { + "epoch": 0.3894767218160831, + "grad_norm": 2.0711617523967765, + "learning_rate": 1.395454573309736e-05, + "loss": 1.0719, + "step": 4049 + }, + { + "epoch": 0.3895729126587149, + "grad_norm": 1.896856268770319, + "learning_rate": 1.3951684070175802e-05, + "loss": 0.9086, + "step": 4050 + }, + { + "epoch": 0.38966910350134665, + "grad_norm": 2.250231184724206, + "learning_rate": 1.394882202370928e-05, + "loss": 0.9257, + "step": 4051 + }, + { + "epoch": 0.38976529434397844, + "grad_norm": 2.126619294357842, + "learning_rate": 1.3945959593975582e-05, + "loss": 1.0637, + "step": 4052 + }, + { + "epoch": 0.38986148518661023, + "grad_norm": 1.7130301873571196, + "learning_rate": 1.3943096781252529e-05, + "loss": 0.8743, + "step": 4053 + }, + { + "epoch": 0.38995767602924203, + "grad_norm": 2.026510911769511, + "learning_rate": 1.3940233585817984e-05, + "loss": 0.9577, + "step": 4054 + }, + { + "epoch": 0.3900538668718738, + "grad_norm": 1.9506925181561583, + "learning_rate": 1.3937370007949844e-05, + "loss": 0.9295, + "step": 4055 + }, + { + "epoch": 0.39015005771450556, + "grad_norm": 2.085126579438011, + "learning_rate": 1.3934506047926042e-05, + "loss": 1.0551, + "step": 4056 + }, + { + "epoch": 0.39024624855713735, + "grad_norm": 2.101498772963978, + "learning_rate": 1.3931641706024553e-05, + "loss": 1.0419, + "step": 4057 + }, + { + "epoch": 0.39034243939976915, + "grad_norm": 2.1875190255611563, + "learning_rate": 1.3928776982523384e-05, + "loss": 0.8813, + "step": 4058 + }, + { + "epoch": 0.39043863024240094, + "grad_norm": 2.2381492961167586, + "learning_rate": 1.3925911877700582e-05, + "loss": 1.1181, + "step": 4059 + }, + { + "epoch": 0.3905348210850327, + "grad_norm": 1.9924440563339734, + "learning_rate": 1.3923046391834229e-05, + "loss": 0.8799, + "step": 4060 + }, + { + "epoch": 0.3906310119276645, + "grad_norm": 1.9568579540431381, + "learning_rate": 1.3920180525202447e-05, + "loss": 1.0636, + "step": 4061 + }, + { + "epoch": 0.39072720277029627, + "grad_norm": 2.0562048432481714, + "learning_rate": 1.3917314278083391e-05, + "loss": 0.826, + "step": 4062 + }, + { + "epoch": 0.39082339361292806, + "grad_norm": 1.9812706503716038, + "learning_rate": 1.3914447650755256e-05, + "loss": 0.9113, + "step": 4063 + }, + { + "epoch": 0.39091958445555985, + "grad_norm": 1.9627854451291045, + "learning_rate": 1.3911580643496272e-05, + "loss": 0.9977, + "step": 4064 + }, + { + "epoch": 0.3910157752981916, + "grad_norm": 2.0290545940869205, + "learning_rate": 1.390871325658471e-05, + "loss": 0.9032, + "step": 4065 + }, + { + "epoch": 0.3911119661408234, + "grad_norm": 1.743877062190088, + "learning_rate": 1.3905845490298867e-05, + "loss": 0.8419, + "step": 4066 + }, + { + "epoch": 0.3912081569834552, + "grad_norm": 2.1051047146375006, + "learning_rate": 1.3902977344917094e-05, + "loss": 1.0555, + "step": 4067 + }, + { + "epoch": 0.391304347826087, + "grad_norm": 2.1850575851289626, + "learning_rate": 1.390010882071776e-05, + "loss": 0.9442, + "step": 4068 + }, + { + "epoch": 0.3914005386687187, + "grad_norm": 1.7706835028745882, + "learning_rate": 1.3897239917979293e-05, + "loss": 0.9628, + "step": 4069 + }, + { + "epoch": 0.3914967295113505, + "grad_norm": 2.1162269496783446, + "learning_rate": 1.3894370636980128e-05, + "loss": 0.9509, + "step": 4070 + }, + { + "epoch": 0.3915929203539823, + "grad_norm": 2.289051416520777, + "learning_rate": 1.3891500977998765e-05, + "loss": 1.06, + "step": 4071 + }, + { + "epoch": 0.3916891111966141, + "grad_norm": 2.1389681148658415, + "learning_rate": 1.3888630941313728e-05, + "loss": 0.9543, + "step": 4072 + }, + { + "epoch": 0.3917853020392459, + "grad_norm": 2.1320785414801007, + "learning_rate": 1.3885760527203571e-05, + "loss": 1.0557, + "step": 4073 + }, + { + "epoch": 0.3918814928818776, + "grad_norm": 2.049065919927566, + "learning_rate": 1.3882889735946901e-05, + "loss": 1.044, + "step": 4074 + }, + { + "epoch": 0.3919776837245094, + "grad_norm": 2.112892376581695, + "learning_rate": 1.3880018567822351e-05, + "loss": 0.9686, + "step": 4075 + }, + { + "epoch": 0.3920738745671412, + "grad_norm": 2.4218678823725988, + "learning_rate": 1.3877147023108592e-05, + "loss": 1.1066, + "step": 4076 + }, + { + "epoch": 0.392170065409773, + "grad_norm": 1.9710969568069234, + "learning_rate": 1.387427510208433e-05, + "loss": 1.0315, + "step": 4077 + }, + { + "epoch": 0.39226625625240474, + "grad_norm": 2.2717163427712137, + "learning_rate": 1.3871402805028314e-05, + "loss": 1.0743, + "step": 4078 + }, + { + "epoch": 0.39236244709503654, + "grad_norm": 2.1405379054146336, + "learning_rate": 1.386853013221932e-05, + "loss": 1.1036, + "step": 4079 + }, + { + "epoch": 0.39245863793766833, + "grad_norm": 2.1374889233319396, + "learning_rate": 1.3865657083936167e-05, + "loss": 1.0058, + "step": 4080 + }, + { + "epoch": 0.3925548287803001, + "grad_norm": 2.058075229149637, + "learning_rate": 1.3862783660457712e-05, + "loss": 0.8882, + "step": 4081 + }, + { + "epoch": 0.3926510196229319, + "grad_norm": 1.9986848976938525, + "learning_rate": 1.3859909862062844e-05, + "loss": 0.9219, + "step": 4082 + }, + { + "epoch": 0.39274721046556366, + "grad_norm": 2.046642794731875, + "learning_rate": 1.385703568903049e-05, + "loss": 1.0749, + "step": 4083 + }, + { + "epoch": 0.39284340130819545, + "grad_norm": 2.1833954032109424, + "learning_rate": 1.385416114163961e-05, + "loss": 1.0083, + "step": 4084 + }, + { + "epoch": 0.39293959215082724, + "grad_norm": 2.417411357566989, + "learning_rate": 1.385128622016921e-05, + "loss": 0.8732, + "step": 4085 + }, + { + "epoch": 0.39303578299345904, + "grad_norm": 2.000997821008849, + "learning_rate": 1.3848410924898321e-05, + "loss": 0.9805, + "step": 4086 + }, + { + "epoch": 0.39313197383609083, + "grad_norm": 1.8836263933698105, + "learning_rate": 1.3845535256106018e-05, + "loss": 1.0844, + "step": 4087 + }, + { + "epoch": 0.39322816467872257, + "grad_norm": 1.858534028636421, + "learning_rate": 1.3842659214071406e-05, + "loss": 1.1333, + "step": 4088 + }, + { + "epoch": 0.39332435552135436, + "grad_norm": 2.097007577037917, + "learning_rate": 1.3839782799073632e-05, + "loss": 1.0188, + "step": 4089 + }, + { + "epoch": 0.39342054636398616, + "grad_norm": 1.743279568251556, + "learning_rate": 1.3836906011391878e-05, + "loss": 0.8906, + "step": 4090 + }, + { + "epoch": 0.39351673720661795, + "grad_norm": 1.9426614047680482, + "learning_rate": 1.3834028851305356e-05, + "loss": 1.1038, + "step": 4091 + }, + { + "epoch": 0.3936129280492497, + "grad_norm": 2.1083386336342813, + "learning_rate": 1.3831151319093323e-05, + "loss": 1.1035, + "step": 4092 + }, + { + "epoch": 0.3937091188918815, + "grad_norm": 2.01788603193679, + "learning_rate": 1.3828273415035072e-05, + "loss": 0.9532, + "step": 4093 + }, + { + "epoch": 0.3938053097345133, + "grad_norm": 2.2625555737751686, + "learning_rate": 1.382539513940992e-05, + "loss": 1.0921, + "step": 4094 + }, + { + "epoch": 0.39390150057714507, + "grad_norm": 2.224497982014692, + "learning_rate": 1.3822516492497235e-05, + "loss": 0.9584, + "step": 4095 + }, + { + "epoch": 0.39399769141977686, + "grad_norm": 2.3578731951893293, + "learning_rate": 1.3819637474576411e-05, + "loss": 1.0306, + "step": 4096 + }, + { + "epoch": 0.3940938822624086, + "grad_norm": 1.9513958840583547, + "learning_rate": 1.3816758085926884e-05, + "loss": 0.8476, + "step": 4097 + }, + { + "epoch": 0.3941900731050404, + "grad_norm": 1.8466738271263237, + "learning_rate": 1.381387832682812e-05, + "loss": 0.9987, + "step": 4098 + }, + { + "epoch": 0.3942862639476722, + "grad_norm": 2.0897090313442908, + "learning_rate": 1.3810998197559628e-05, + "loss": 1.055, + "step": 4099 + }, + { + "epoch": 0.394382454790304, + "grad_norm": 1.9405793460021616, + "learning_rate": 1.380811769840095e-05, + "loss": 1.0225, + "step": 4100 + }, + { + "epoch": 0.3944786456329357, + "grad_norm": 2.419674581559365, + "learning_rate": 1.3805236829631656e-05, + "loss": 0.9917, + "step": 4101 + }, + { + "epoch": 0.3945748364755675, + "grad_norm": 2.0743037859776394, + "learning_rate": 1.3802355591531366e-05, + "loss": 1.0754, + "step": 4102 + }, + { + "epoch": 0.3946710273181993, + "grad_norm": 1.977125204529584, + "learning_rate": 1.379947398437973e-05, + "loss": 0.9633, + "step": 4103 + }, + { + "epoch": 0.3947672181608311, + "grad_norm": 1.9695873872581866, + "learning_rate": 1.3796592008456427e-05, + "loss": 1.0061, + "step": 4104 + }, + { + "epoch": 0.3948634090034629, + "grad_norm": 2.056286245304845, + "learning_rate": 1.3793709664041182e-05, + "loss": 1.0225, + "step": 4105 + }, + { + "epoch": 0.39495959984609463, + "grad_norm": 2.0045649486512986, + "learning_rate": 1.3790826951413747e-05, + "loss": 1.0517, + "step": 4106 + }, + { + "epoch": 0.3950557906887264, + "grad_norm": 2.6217384146014706, + "learning_rate": 1.3787943870853921e-05, + "loss": 0.903, + "step": 4107 + }, + { + "epoch": 0.3951519815313582, + "grad_norm": 2.072725770432475, + "learning_rate": 1.3785060422641526e-05, + "loss": 1.0526, + "step": 4108 + }, + { + "epoch": 0.39524817237399, + "grad_norm": 1.9784851120684683, + "learning_rate": 1.3782176607056425e-05, + "loss": 0.9701, + "step": 4109 + }, + { + "epoch": 0.39534436321662175, + "grad_norm": 1.9940801100483794, + "learning_rate": 1.3779292424378521e-05, + "loss": 0.9036, + "step": 4110 + }, + { + "epoch": 0.39544055405925355, + "grad_norm": 1.9373343149586457, + "learning_rate": 1.377640787488775e-05, + "loss": 1.0272, + "step": 4111 + }, + { + "epoch": 0.39553674490188534, + "grad_norm": 1.886051854907147, + "learning_rate": 1.3773522958864076e-05, + "loss": 1.0311, + "step": 4112 + }, + { + "epoch": 0.39563293574451713, + "grad_norm": 2.240412454687543, + "learning_rate": 1.377063767658751e-05, + "loss": 0.9796, + "step": 4113 + }, + { + "epoch": 0.3957291265871489, + "grad_norm": 1.9439454371421105, + "learning_rate": 1.3767752028338091e-05, + "loss": 0.9952, + "step": 4114 + }, + { + "epoch": 0.39582531742978067, + "grad_norm": 2.3194016122027246, + "learning_rate": 1.3764866014395894e-05, + "loss": 1.0573, + "step": 4115 + }, + { + "epoch": 0.39592150827241246, + "grad_norm": 1.9345195891181686, + "learning_rate": 1.376197963504104e-05, + "loss": 1.0876, + "step": 4116 + }, + { + "epoch": 0.39601769911504425, + "grad_norm": 2.091574396189734, + "learning_rate": 1.375909289055367e-05, + "loss": 1.0652, + "step": 4117 + }, + { + "epoch": 0.39611388995767605, + "grad_norm": 2.0379354489575863, + "learning_rate": 1.3756205781213965e-05, + "loss": 1.0165, + "step": 4118 + }, + { + "epoch": 0.3962100808003078, + "grad_norm": 2.028729178773316, + "learning_rate": 1.3753318307302152e-05, + "loss": 0.9917, + "step": 4119 + }, + { + "epoch": 0.3963062716429396, + "grad_norm": 1.9656329851699572, + "learning_rate": 1.375043046909848e-05, + "loss": 1.0902, + "step": 4120 + }, + { + "epoch": 0.39640246248557137, + "grad_norm": 2.293114515998912, + "learning_rate": 1.3747542266883238e-05, + "loss": 1.014, + "step": 4121 + }, + { + "epoch": 0.39649865332820317, + "grad_norm": 2.242994987606146, + "learning_rate": 1.3744653700936752e-05, + "loss": 1.0372, + "step": 4122 + }, + { + "epoch": 0.39659484417083496, + "grad_norm": 1.9289609444892917, + "learning_rate": 1.3741764771539383e-05, + "loss": 1.0928, + "step": 4123 + }, + { + "epoch": 0.3966910350134667, + "grad_norm": 1.9561843706357094, + "learning_rate": 1.3738875478971526e-05, + "loss": 0.9692, + "step": 4124 + }, + { + "epoch": 0.3967872258560985, + "grad_norm": 2.008780000253285, + "learning_rate": 1.3735985823513613e-05, + "loss": 0.8511, + "step": 4125 + }, + { + "epoch": 0.3968834166987303, + "grad_norm": 2.1739775825201306, + "learning_rate": 1.3733095805446107e-05, + "loss": 0.9258, + "step": 4126 + }, + { + "epoch": 0.3969796075413621, + "grad_norm": 2.1877757918140235, + "learning_rate": 1.3730205425049512e-05, + "loss": 1.0474, + "step": 4127 + }, + { + "epoch": 0.3970757983839938, + "grad_norm": 1.9391661500117485, + "learning_rate": 1.372731468260436e-05, + "loss": 1.0115, + "step": 4128 + }, + { + "epoch": 0.3971719892266256, + "grad_norm": 1.932465085257942, + "learning_rate": 1.3724423578391228e-05, + "loss": 0.9518, + "step": 4129 + }, + { + "epoch": 0.3972681800692574, + "grad_norm": 2.1545746381221393, + "learning_rate": 1.372153211269072e-05, + "loss": 0.9299, + "step": 4130 + }, + { + "epoch": 0.3973643709118892, + "grad_norm": 2.0384954989986253, + "learning_rate": 1.3718640285783477e-05, + "loss": 0.9966, + "step": 4131 + }, + { + "epoch": 0.397460561754521, + "grad_norm": 1.9940325692225693, + "learning_rate": 1.3715748097950176e-05, + "loss": 1.101, + "step": 4132 + }, + { + "epoch": 0.39755675259715273, + "grad_norm": 2.205929454635862, + "learning_rate": 1.371285554947153e-05, + "loss": 1.0219, + "step": 4133 + }, + { + "epoch": 0.3976529434397845, + "grad_norm": 2.026898600564301, + "learning_rate": 1.3709962640628284e-05, + "loss": 0.972, + "step": 4134 + }, + { + "epoch": 0.3977491342824163, + "grad_norm": 2.1751100822383744, + "learning_rate": 1.3707069371701223e-05, + "loss": 1.0312, + "step": 4135 + }, + { + "epoch": 0.3978453251250481, + "grad_norm": 2.0704068057884175, + "learning_rate": 1.3704175742971158e-05, + "loss": 1.0035, + "step": 4136 + }, + { + "epoch": 0.3979415159676799, + "grad_norm": 2.022586203317941, + "learning_rate": 1.3701281754718945e-05, + "loss": 1.0849, + "step": 4137 + }, + { + "epoch": 0.39803770681031164, + "grad_norm": 2.1997235045222663, + "learning_rate": 1.369838740722547e-05, + "loss": 1.0304, + "step": 4138 + }, + { + "epoch": 0.39813389765294344, + "grad_norm": 2.080959766163983, + "learning_rate": 1.3695492700771653e-05, + "loss": 0.9714, + "step": 4139 + }, + { + "epoch": 0.39823008849557523, + "grad_norm": 2.1743086566932917, + "learning_rate": 1.3692597635638452e-05, + "loss": 0.9743, + "step": 4140 + }, + { + "epoch": 0.398326279338207, + "grad_norm": 2.3348157859635132, + "learning_rate": 1.3689702212106853e-05, + "loss": 1.0019, + "step": 4141 + }, + { + "epoch": 0.39842247018083876, + "grad_norm": 1.9711446866773399, + "learning_rate": 1.368680643045789e-05, + "loss": 1.1113, + "step": 4142 + }, + { + "epoch": 0.39851866102347056, + "grad_norm": 2.054485440245032, + "learning_rate": 1.3683910290972618e-05, + "loss": 0.9137, + "step": 4143 + }, + { + "epoch": 0.39861485186610235, + "grad_norm": 2.0499641458059177, + "learning_rate": 1.3681013793932132e-05, + "loss": 1.0607, + "step": 4144 + }, + { + "epoch": 0.39871104270873414, + "grad_norm": 2.1191426251070524, + "learning_rate": 1.3678116939617564e-05, + "loss": 0.9865, + "step": 4145 + }, + { + "epoch": 0.39880723355136594, + "grad_norm": 1.9065003238909155, + "learning_rate": 1.3675219728310076e-05, + "loss": 1.0436, + "step": 4146 + }, + { + "epoch": 0.3989034243939977, + "grad_norm": 2.2357075972882825, + "learning_rate": 1.3672322160290875e-05, + "loss": 1.0921, + "step": 4147 + }, + { + "epoch": 0.39899961523662947, + "grad_norm": 2.1477291483565666, + "learning_rate": 1.3669424235841185e-05, + "loss": 1.0219, + "step": 4148 + }, + { + "epoch": 0.39909580607926126, + "grad_norm": 1.9705974216901914, + "learning_rate": 1.3666525955242277e-05, + "loss": 0.9758, + "step": 4149 + }, + { + "epoch": 0.39919199692189306, + "grad_norm": 2.0103051406389714, + "learning_rate": 1.3663627318775459e-05, + "loss": 1.0822, + "step": 4150 + }, + { + "epoch": 0.3992881877645248, + "grad_norm": 1.9118460499266632, + "learning_rate": 1.3660728326722066e-05, + "loss": 0.8994, + "step": 4151 + }, + { + "epoch": 0.3993843786071566, + "grad_norm": 2.1662173250740637, + "learning_rate": 1.3657828979363468e-05, + "loss": 1.0185, + "step": 4152 + }, + { + "epoch": 0.3994805694497884, + "grad_norm": 2.2819687801037696, + "learning_rate": 1.3654929276981074e-05, + "loss": 1.0554, + "step": 4153 + }, + { + "epoch": 0.3995767602924202, + "grad_norm": 1.8022387195997651, + "learning_rate": 1.3652029219856324e-05, + "loss": 0.818, + "step": 4154 + }, + { + "epoch": 0.39967295113505197, + "grad_norm": 1.9585741480974217, + "learning_rate": 1.3649128808270691e-05, + "loss": 1.0966, + "step": 4155 + }, + { + "epoch": 0.3997691419776837, + "grad_norm": 1.9356339196927086, + "learning_rate": 1.3646228042505694e-05, + "loss": 1.0442, + "step": 4156 + }, + { + "epoch": 0.3998653328203155, + "grad_norm": 2.2126310928413586, + "learning_rate": 1.3643326922842868e-05, + "loss": 1.0872, + "step": 4157 + }, + { + "epoch": 0.3999615236629473, + "grad_norm": 2.154108990400732, + "learning_rate": 1.3640425449563793e-05, + "loss": 0.9048, + "step": 4158 + }, + { + "epoch": 0.4000577145055791, + "grad_norm": 1.9279236290899417, + "learning_rate": 1.3637523622950087e-05, + "loss": 1.0833, + "step": 4159 + }, + { + "epoch": 0.4001539053482108, + "grad_norm": 1.9774246624601117, + "learning_rate": 1.3634621443283389e-05, + "loss": 1.0396, + "step": 4160 + }, + { + "epoch": 0.4002500961908426, + "grad_norm": 2.068009730031748, + "learning_rate": 1.363171891084539e-05, + "loss": 0.8762, + "step": 4161 + }, + { + "epoch": 0.4003462870334744, + "grad_norm": 1.9056581388307723, + "learning_rate": 1.36288160259178e-05, + "loss": 0.9891, + "step": 4162 + }, + { + "epoch": 0.4004424778761062, + "grad_norm": 2.0093809287124027, + "learning_rate": 1.362591278878237e-05, + "loss": 0.9833, + "step": 4163 + }, + { + "epoch": 0.400538668718738, + "grad_norm": 1.810172887196961, + "learning_rate": 1.3623009199720882e-05, + "loss": 0.9446, + "step": 4164 + }, + { + "epoch": 0.40063485956136974, + "grad_norm": 2.112466010593526, + "learning_rate": 1.362010525901516e-05, + "loss": 1.0244, + "step": 4165 + }, + { + "epoch": 0.40073105040400153, + "grad_norm": 1.992799655697273, + "learning_rate": 1.3617200966947053e-05, + "loss": 1.0914, + "step": 4166 + }, + { + "epoch": 0.4008272412466333, + "grad_norm": 2.023943045025389, + "learning_rate": 1.3614296323798444e-05, + "loss": 0.992, + "step": 4167 + }, + { + "epoch": 0.4009234320892651, + "grad_norm": 2.308366954072252, + "learning_rate": 1.3611391329851262e-05, + "loss": 0.9873, + "step": 4168 + }, + { + "epoch": 0.40101962293189686, + "grad_norm": 1.9370225722123322, + "learning_rate": 1.3608485985387455e-05, + "loss": 0.9906, + "step": 4169 + }, + { + "epoch": 0.40111581377452865, + "grad_norm": 1.9325488131668849, + "learning_rate": 1.3605580290689013e-05, + "loss": 1.0572, + "step": 4170 + }, + { + "epoch": 0.40121200461716044, + "grad_norm": 1.979976910259472, + "learning_rate": 1.3602674246037962e-05, + "loss": 0.9394, + "step": 4171 + }, + { + "epoch": 0.40130819545979224, + "grad_norm": 2.038355847078855, + "learning_rate": 1.3599767851716353e-05, + "loss": 1.0922, + "step": 4172 + }, + { + "epoch": 0.40140438630242403, + "grad_norm": 2.106264454747121, + "learning_rate": 1.3596861108006283e-05, + "loss": 1.0386, + "step": 4173 + }, + { + "epoch": 0.40150057714505577, + "grad_norm": 2.089427479415232, + "learning_rate": 1.3593954015189867e-05, + "loss": 1.0447, + "step": 4174 + }, + { + "epoch": 0.40159676798768756, + "grad_norm": 2.0384137821021624, + "learning_rate": 1.3591046573549273e-05, + "loss": 1.0662, + "step": 4175 + }, + { + "epoch": 0.40169295883031936, + "grad_norm": 2.0204907920821413, + "learning_rate": 1.3588138783366692e-05, + "loss": 0.9379, + "step": 4176 + }, + { + "epoch": 0.40178914967295115, + "grad_norm": 1.736748650215178, + "learning_rate": 1.3585230644924346e-05, + "loss": 0.9815, + "step": 4177 + }, + { + "epoch": 0.4018853405155829, + "grad_norm": 2.0571800399016453, + "learning_rate": 1.3582322158504495e-05, + "loss": 1.0974, + "step": 4178 + }, + { + "epoch": 0.4019815313582147, + "grad_norm": 2.24035625085091, + "learning_rate": 1.3579413324389437e-05, + "loss": 1.0549, + "step": 4179 + }, + { + "epoch": 0.4020777222008465, + "grad_norm": 2.097171194557185, + "learning_rate": 1.3576504142861496e-05, + "loss": 1.0636, + "step": 4180 + }, + { + "epoch": 0.40217391304347827, + "grad_norm": 2.1747811961789605, + "learning_rate": 1.357359461420303e-05, + "loss": 1.021, + "step": 4181 + }, + { + "epoch": 0.40227010388611006, + "grad_norm": 1.8676235088056443, + "learning_rate": 1.3570684738696444e-05, + "loss": 0.9405, + "step": 4182 + }, + { + "epoch": 0.4023662947287418, + "grad_norm": 2.006689034808893, + "learning_rate": 1.3567774516624158e-05, + "loss": 0.9413, + "step": 4183 + }, + { + "epoch": 0.4024624855713736, + "grad_norm": 2.1139553225069543, + "learning_rate": 1.3564863948268631e-05, + "loss": 0.9433, + "step": 4184 + }, + { + "epoch": 0.4025586764140054, + "grad_norm": 2.110594931113313, + "learning_rate": 1.3561953033912371e-05, + "loss": 0.8606, + "step": 4185 + }, + { + "epoch": 0.4026548672566372, + "grad_norm": 1.9717654849694917, + "learning_rate": 1.3559041773837898e-05, + "loss": 1.1199, + "step": 4186 + }, + { + "epoch": 0.402751058099269, + "grad_norm": 1.7594496063587717, + "learning_rate": 1.3556130168327775e-05, + "loss": 0.8811, + "step": 4187 + }, + { + "epoch": 0.4028472489419007, + "grad_norm": 2.0517088251082933, + "learning_rate": 1.3553218217664603e-05, + "loss": 1.1315, + "step": 4188 + }, + { + "epoch": 0.4029434397845325, + "grad_norm": 2.1702295853187614, + "learning_rate": 1.3550305922131009e-05, + "loss": 1.04, + "step": 4189 + }, + { + "epoch": 0.4030396306271643, + "grad_norm": 1.7702381629824075, + "learning_rate": 1.3547393282009656e-05, + "loss": 0.9424, + "step": 4190 + }, + { + "epoch": 0.4031358214697961, + "grad_norm": 2.087218841137555, + "learning_rate": 1.3544480297583243e-05, + "loss": 1.0767, + "step": 4191 + }, + { + "epoch": 0.40323201231242783, + "grad_norm": 2.0226436845032962, + "learning_rate": 1.3541566969134496e-05, + "loss": 0.9548, + "step": 4192 + }, + { + "epoch": 0.40332820315505963, + "grad_norm": 1.957640369292656, + "learning_rate": 1.3538653296946183e-05, + "loss": 1.0274, + "step": 4193 + }, + { + "epoch": 0.4034243939976914, + "grad_norm": 2.3042434728980923, + "learning_rate": 1.3535739281301102e-05, + "loss": 1.0355, + "step": 4194 + }, + { + "epoch": 0.4035205848403232, + "grad_norm": 2.1199283075684443, + "learning_rate": 1.3532824922482078e-05, + "loss": 1.0364, + "step": 4195 + }, + { + "epoch": 0.403616775682955, + "grad_norm": 2.066102121650769, + "learning_rate": 1.3529910220771975e-05, + "loss": 1.0148, + "step": 4196 + }, + { + "epoch": 0.40371296652558675, + "grad_norm": 1.9025280166952743, + "learning_rate": 1.3526995176453696e-05, + "loss": 0.9743, + "step": 4197 + }, + { + "epoch": 0.40380915736821854, + "grad_norm": 2.0404652518732562, + "learning_rate": 1.3524079789810163e-05, + "loss": 1.0128, + "step": 4198 + }, + { + "epoch": 0.40390534821085033, + "grad_norm": 2.033815802872566, + "learning_rate": 1.352116406112435e-05, + "loss": 1.1207, + "step": 4199 + }, + { + "epoch": 0.40400153905348213, + "grad_norm": 2.043870944728902, + "learning_rate": 1.3518247990679241e-05, + "loss": 0.9776, + "step": 4200 + }, + { + "epoch": 0.40409772989611387, + "grad_norm": 2.2408666526359293, + "learning_rate": 1.3515331578757876e-05, + "loss": 1.1193, + "step": 4201 + }, + { + "epoch": 0.40419392073874566, + "grad_norm": 2.408035740486178, + "learning_rate": 1.3512414825643312e-05, + "loss": 0.9408, + "step": 4202 + }, + { + "epoch": 0.40429011158137745, + "grad_norm": 1.927636227081207, + "learning_rate": 1.3509497731618649e-05, + "loss": 1.0749, + "step": 4203 + }, + { + "epoch": 0.40438630242400925, + "grad_norm": 2.2631974070987857, + "learning_rate": 1.3506580296967011e-05, + "loss": 0.9615, + "step": 4204 + }, + { + "epoch": 0.40448249326664104, + "grad_norm": 2.005808173770803, + "learning_rate": 1.3503662521971565e-05, + "loss": 1.0227, + "step": 4205 + }, + { + "epoch": 0.4045786841092728, + "grad_norm": 2.105115987289468, + "learning_rate": 1.3500744406915505e-05, + "loss": 0.9292, + "step": 4206 + }, + { + "epoch": 0.4046748749519046, + "grad_norm": 2.164090405085205, + "learning_rate": 1.3497825952082056e-05, + "loss": 0.9225, + "step": 4207 + }, + { + "epoch": 0.40477106579453637, + "grad_norm": 2.290692340148268, + "learning_rate": 1.3494907157754485e-05, + "loss": 0.8722, + "step": 4208 + }, + { + "epoch": 0.40486725663716816, + "grad_norm": 2.005937774269653, + "learning_rate": 1.349198802421608e-05, + "loss": 0.8874, + "step": 4209 + }, + { + "epoch": 0.4049634474797999, + "grad_norm": 1.8866942419977488, + "learning_rate": 1.348906855175017e-05, + "loss": 0.9362, + "step": 4210 + }, + { + "epoch": 0.4050596383224317, + "grad_norm": 1.9875011106948497, + "learning_rate": 1.3486148740640121e-05, + "loss": 0.9838, + "step": 4211 + }, + { + "epoch": 0.4051558291650635, + "grad_norm": 2.0337761171137503, + "learning_rate": 1.3483228591169315e-05, + "loss": 0.981, + "step": 4212 + }, + { + "epoch": 0.4052520200076953, + "grad_norm": 1.990764402882747, + "learning_rate": 1.3480308103621187e-05, + "loss": 1.0218, + "step": 4213 + }, + { + "epoch": 0.4053482108503271, + "grad_norm": 1.892556470152748, + "learning_rate": 1.347738727827919e-05, + "loss": 1.0293, + "step": 4214 + }, + { + "epoch": 0.4054444016929588, + "grad_norm": 2.7289734603188043, + "learning_rate": 1.3474466115426816e-05, + "loss": 0.9611, + "step": 4215 + }, + { + "epoch": 0.4055405925355906, + "grad_norm": 1.9385856373419867, + "learning_rate": 1.3471544615347591e-05, + "loss": 0.9892, + "step": 4216 + }, + { + "epoch": 0.4056367833782224, + "grad_norm": 1.9264366134137292, + "learning_rate": 1.3468622778325074e-05, + "loss": 0.9427, + "step": 4217 + }, + { + "epoch": 0.4057329742208542, + "grad_norm": 1.957986453387599, + "learning_rate": 1.3465700604642847e-05, + "loss": 1.0462, + "step": 4218 + }, + { + "epoch": 0.40582916506348593, + "grad_norm": 1.9137432508056926, + "learning_rate": 1.346277809458454e-05, + "loss": 0.944, + "step": 4219 + }, + { + "epoch": 0.4059253559061177, + "grad_norm": 1.9319587044246034, + "learning_rate": 1.34598552484338e-05, + "loss": 0.9963, + "step": 4220 + }, + { + "epoch": 0.4060215467487495, + "grad_norm": 2.0854243193930517, + "learning_rate": 1.3456932066474318e-05, + "loss": 1.0193, + "step": 4221 + }, + { + "epoch": 0.4061177375913813, + "grad_norm": 2.0899667667760617, + "learning_rate": 1.3454008548989816e-05, + "loss": 1.0049, + "step": 4222 + }, + { + "epoch": 0.4062139284340131, + "grad_norm": 2.0049347954453967, + "learning_rate": 1.3451084696264048e-05, + "loss": 1.033, + "step": 4223 + }, + { + "epoch": 0.40631011927664484, + "grad_norm": 2.0175758486377027, + "learning_rate": 1.3448160508580789e-05, + "loss": 1.0491, + "step": 4224 + }, + { + "epoch": 0.40640631011927664, + "grad_norm": 1.941925003530211, + "learning_rate": 1.344523598622387e-05, + "loss": 0.9824, + "step": 4225 + }, + { + "epoch": 0.40650250096190843, + "grad_norm": 2.1181626309046555, + "learning_rate": 1.3442311129477133e-05, + "loss": 0.9635, + "step": 4226 + }, + { + "epoch": 0.4065986918045402, + "grad_norm": 2.1818395756455677, + "learning_rate": 1.343938593862446e-05, + "loss": 1.006, + "step": 4227 + }, + { + "epoch": 0.40669488264717196, + "grad_norm": 1.9863666527234713, + "learning_rate": 1.343646041394977e-05, + "loss": 1.0144, + "step": 4228 + }, + { + "epoch": 0.40679107348980376, + "grad_norm": 2.4706338800862646, + "learning_rate": 1.343353455573701e-05, + "loss": 0.9972, + "step": 4229 + }, + { + "epoch": 0.40688726433243555, + "grad_norm": 2.232735975722647, + "learning_rate": 1.3430608364270156e-05, + "loss": 1.0084, + "step": 4230 + }, + { + "epoch": 0.40698345517506734, + "grad_norm": 3.133111937122784, + "learning_rate": 1.3427681839833221e-05, + "loss": 0.8695, + "step": 4231 + }, + { + "epoch": 0.40707964601769914, + "grad_norm": 1.9185287115348282, + "learning_rate": 1.3424754982710256e-05, + "loss": 1.0509, + "step": 4232 + }, + { + "epoch": 0.4071758368603309, + "grad_norm": 2.0652366052497633, + "learning_rate": 1.3421827793185328e-05, + "loss": 1.0642, + "step": 4233 + }, + { + "epoch": 0.40727202770296267, + "grad_norm": 2.2048807996047586, + "learning_rate": 1.3418900271542552e-05, + "loss": 0.9983, + "step": 4234 + }, + { + "epoch": 0.40736821854559446, + "grad_norm": 2.2375888835328954, + "learning_rate": 1.3415972418066066e-05, + "loss": 1.0028, + "step": 4235 + }, + { + "epoch": 0.40746440938822626, + "grad_norm": 2.1432251530348805, + "learning_rate": 1.3413044233040045e-05, + "loss": 0.964, + "step": 4236 + }, + { + "epoch": 0.40756060023085805, + "grad_norm": 2.063708872209188, + "learning_rate": 1.3410115716748697e-05, + "loss": 1.0549, + "step": 4237 + }, + { + "epoch": 0.4076567910734898, + "grad_norm": 2.103720211207092, + "learning_rate": 1.3407186869476253e-05, + "loss": 1.0738, + "step": 4238 + }, + { + "epoch": 0.4077529819161216, + "grad_norm": 2.3178785900615098, + "learning_rate": 1.340425769150699e-05, + "loss": 1.0293, + "step": 4239 + }, + { + "epoch": 0.4078491727587534, + "grad_norm": 2.1492046265599813, + "learning_rate": 1.3401328183125208e-05, + "loss": 1.0128, + "step": 4240 + }, + { + "epoch": 0.40794536360138517, + "grad_norm": 2.0647579504098372, + "learning_rate": 1.339839834461524e-05, + "loss": 0.9516, + "step": 4241 + }, + { + "epoch": 0.4080415544440169, + "grad_norm": 2.0597861292495447, + "learning_rate": 1.339546817626145e-05, + "loss": 1.0099, + "step": 4242 + }, + { + "epoch": 0.4081377452866487, + "grad_norm": 2.1161100704839573, + "learning_rate": 1.339253767834824e-05, + "loss": 0.9968, + "step": 4243 + }, + { + "epoch": 0.4082339361292805, + "grad_norm": 1.842510783822968, + "learning_rate": 1.3389606851160037e-05, + "loss": 1.0025, + "step": 4244 + }, + { + "epoch": 0.4083301269719123, + "grad_norm": 2.0458004781939874, + "learning_rate": 1.3386675694981306e-05, + "loss": 0.9404, + "step": 4245 + }, + { + "epoch": 0.4084263178145441, + "grad_norm": 1.9972776303514082, + "learning_rate": 1.3383744210096537e-05, + "loss": 0.9053, + "step": 4246 + }, + { + "epoch": 0.4085225086571758, + "grad_norm": 1.868382297539071, + "learning_rate": 1.3380812396790257e-05, + "loss": 1.079, + "step": 4247 + }, + { + "epoch": 0.4086186994998076, + "grad_norm": 2.1587991791849395, + "learning_rate": 1.3377880255347026e-05, + "loss": 1.073, + "step": 4248 + }, + { + "epoch": 0.4087148903424394, + "grad_norm": 1.9656760248615828, + "learning_rate": 1.3374947786051435e-05, + "loss": 1.0185, + "step": 4249 + }, + { + "epoch": 0.4088110811850712, + "grad_norm": 2.095822740058794, + "learning_rate": 1.3372014989188098e-05, + "loss": 1.0084, + "step": 4250 + }, + { + "epoch": 0.40890727202770294, + "grad_norm": 1.9856633896217744, + "learning_rate": 1.3369081865041678e-05, + "loss": 1.001, + "step": 4251 + }, + { + "epoch": 0.40900346287033473, + "grad_norm": 1.9440117537984674, + "learning_rate": 1.3366148413896851e-05, + "loss": 0.9643, + "step": 4252 + }, + { + "epoch": 0.4090996537129665, + "grad_norm": 1.974453291513518, + "learning_rate": 1.3363214636038337e-05, + "loss": 0.9422, + "step": 4253 + }, + { + "epoch": 0.4091958445555983, + "grad_norm": 1.9066466134308215, + "learning_rate": 1.3360280531750886e-05, + "loss": 0.9553, + "step": 4254 + }, + { + "epoch": 0.4092920353982301, + "grad_norm": 2.1356552159117523, + "learning_rate": 1.335734610131928e-05, + "loss": 1.0485, + "step": 4255 + }, + { + "epoch": 0.40938822624086185, + "grad_norm": 1.9360473605168114, + "learning_rate": 1.3354411345028324e-05, + "loss": 1.0296, + "step": 4256 + }, + { + "epoch": 0.40948441708349365, + "grad_norm": 1.9581141853321298, + "learning_rate": 1.3351476263162867e-05, + "loss": 1.0971, + "step": 4257 + }, + { + "epoch": 0.40958060792612544, + "grad_norm": 1.8736861635651239, + "learning_rate": 1.3348540856007782e-05, + "loss": 1.0352, + "step": 4258 + }, + { + "epoch": 0.40967679876875723, + "grad_norm": 2.0567406297286306, + "learning_rate": 1.3345605123847974e-05, + "loss": 0.9844, + "step": 4259 + }, + { + "epoch": 0.40977298961138897, + "grad_norm": 1.8361687441111192, + "learning_rate": 1.3342669066968385e-05, + "loss": 0.8873, + "step": 4260 + }, + { + "epoch": 0.40986918045402077, + "grad_norm": 1.8948467200333723, + "learning_rate": 1.3339732685653978e-05, + "loss": 0.9423, + "step": 4261 + }, + { + "epoch": 0.40996537129665256, + "grad_norm": 1.7994277870919888, + "learning_rate": 1.3336795980189763e-05, + "loss": 0.8765, + "step": 4262 + }, + { + "epoch": 0.41006156213928435, + "grad_norm": 2.3493094274452413, + "learning_rate": 1.333385895086077e-05, + "loss": 1.1166, + "step": 4263 + }, + { + "epoch": 0.41015775298191615, + "grad_norm": 2.031005125374534, + "learning_rate": 1.3330921597952056e-05, + "loss": 1.0466, + "step": 4264 + }, + { + "epoch": 0.4102539438245479, + "grad_norm": 2.530205236511711, + "learning_rate": 1.3327983921748721e-05, + "loss": 0.9097, + "step": 4265 + }, + { + "epoch": 0.4103501346671797, + "grad_norm": 2.1404981952381705, + "learning_rate": 1.3325045922535896e-05, + "loss": 1.1006, + "step": 4266 + }, + { + "epoch": 0.41044632550981147, + "grad_norm": 2.2607662266463304, + "learning_rate": 1.3322107600598733e-05, + "loss": 0.9604, + "step": 4267 + }, + { + "epoch": 0.41054251635244327, + "grad_norm": 2.015770972356204, + "learning_rate": 1.3319168956222423e-05, + "loss": 0.9527, + "step": 4268 + }, + { + "epoch": 0.410638707195075, + "grad_norm": 2.1516084816062184, + "learning_rate": 1.3316229989692188e-05, + "loss": 1.0085, + "step": 4269 + }, + { + "epoch": 0.4107348980377068, + "grad_norm": 1.9686293111739483, + "learning_rate": 1.331329070129328e-05, + "loss": 1.001, + "step": 4270 + }, + { + "epoch": 0.4108310888803386, + "grad_norm": 1.832896241703136, + "learning_rate": 1.3310351091310982e-05, + "loss": 0.9631, + "step": 4271 + }, + { + "epoch": 0.4109272797229704, + "grad_norm": 2.02580686316329, + "learning_rate": 1.3307411160030608e-05, + "loss": 1.0083, + "step": 4272 + }, + { + "epoch": 0.4110234705656022, + "grad_norm": 2.2669142958830224, + "learning_rate": 1.33044709077375e-05, + "loss": 0.997, + "step": 4273 + }, + { + "epoch": 0.4111196614082339, + "grad_norm": 2.1219250019654807, + "learning_rate": 1.3301530334717046e-05, + "loss": 0.9626, + "step": 4274 + }, + { + "epoch": 0.4112158522508657, + "grad_norm": 1.9630619360784232, + "learning_rate": 1.3298589441254641e-05, + "loss": 0.957, + "step": 4275 + }, + { + "epoch": 0.4113120430934975, + "grad_norm": 2.4280571332465035, + "learning_rate": 1.3295648227635729e-05, + "loss": 1.0703, + "step": 4276 + }, + { + "epoch": 0.4114082339361293, + "grad_norm": 2.1779538799468723, + "learning_rate": 1.3292706694145784e-05, + "loss": 1.0227, + "step": 4277 + }, + { + "epoch": 0.41150442477876104, + "grad_norm": 2.303869295659606, + "learning_rate": 1.32897648410703e-05, + "loss": 0.9856, + "step": 4278 + }, + { + "epoch": 0.41160061562139283, + "grad_norm": 1.918672044616703, + "learning_rate": 1.3286822668694817e-05, + "loss": 0.9554, + "step": 4279 + }, + { + "epoch": 0.4116968064640246, + "grad_norm": 2.037710850899571, + "learning_rate": 1.328388017730489e-05, + "loss": 1.0969, + "step": 4280 + }, + { + "epoch": 0.4117929973066564, + "grad_norm": 1.9874251485644991, + "learning_rate": 1.328093736718612e-05, + "loss": 0.9987, + "step": 4281 + }, + { + "epoch": 0.4118891881492882, + "grad_norm": 1.8698364209120562, + "learning_rate": 1.327799423862413e-05, + "loss": 1.0002, + "step": 4282 + }, + { + "epoch": 0.41198537899191995, + "grad_norm": 2.3546382796750724, + "learning_rate": 1.3275050791904573e-05, + "loss": 0.9858, + "step": 4283 + }, + { + "epoch": 0.41208156983455174, + "grad_norm": 1.8344873612485895, + "learning_rate": 1.3272107027313142e-05, + "loss": 0.9224, + "step": 4284 + }, + { + "epoch": 0.41217776067718354, + "grad_norm": 2.0771758727890655, + "learning_rate": 1.3269162945135544e-05, + "loss": 0.9924, + "step": 4285 + }, + { + "epoch": 0.41227395151981533, + "grad_norm": 2.206832787294158, + "learning_rate": 1.3266218545657541e-05, + "loss": 0.9985, + "step": 4286 + }, + { + "epoch": 0.41237014236244707, + "grad_norm": 2.315013963733881, + "learning_rate": 1.32632738291649e-05, + "loss": 1.0761, + "step": 4287 + }, + { + "epoch": 0.41246633320507886, + "grad_norm": 1.8565323472900737, + "learning_rate": 1.326032879594344e-05, + "loss": 0.846, + "step": 4288 + }, + { + "epoch": 0.41256252404771065, + "grad_norm": 2.0988249954158213, + "learning_rate": 1.3257383446279e-05, + "loss": 1.007, + "step": 4289 + }, + { + "epoch": 0.41265871489034245, + "grad_norm": 2.366905305015421, + "learning_rate": 1.3254437780457448e-05, + "loss": 1.0296, + "step": 4290 + }, + { + "epoch": 0.41275490573297424, + "grad_norm": 2.128348804314874, + "learning_rate": 1.3251491798764685e-05, + "loss": 0.9225, + "step": 4291 + }, + { + "epoch": 0.412851096575606, + "grad_norm": 1.9689933265492956, + "learning_rate": 1.3248545501486654e-05, + "loss": 1.0199, + "step": 4292 + }, + { + "epoch": 0.4129472874182378, + "grad_norm": 1.8985063444389587, + "learning_rate": 1.3245598888909308e-05, + "loss": 1.0737, + "step": 4293 + }, + { + "epoch": 0.41304347826086957, + "grad_norm": 2.1958045681416034, + "learning_rate": 1.3242651961318646e-05, + "loss": 0.9786, + "step": 4294 + }, + { + "epoch": 0.41313966910350136, + "grad_norm": 1.9487711154784668, + "learning_rate": 1.323970471900069e-05, + "loss": 0.9755, + "step": 4295 + }, + { + "epoch": 0.41323585994613315, + "grad_norm": 2.1824650918063004, + "learning_rate": 1.32367571622415e-05, + "loss": 1.0357, + "step": 4296 + }, + { + "epoch": 0.4133320507887649, + "grad_norm": 2.020384172677645, + "learning_rate": 1.3233809291327155e-05, + "loss": 1.0085, + "step": 4297 + }, + { + "epoch": 0.4134282416313967, + "grad_norm": 1.9509192295351883, + "learning_rate": 1.3230861106543777e-05, + "loss": 0.98, + "step": 4298 + }, + { + "epoch": 0.4135244324740285, + "grad_norm": 2.050814362617783, + "learning_rate": 1.3227912608177508e-05, + "loss": 0.9317, + "step": 4299 + }, + { + "epoch": 0.4136206233166603, + "grad_norm": 1.8798451435778876, + "learning_rate": 1.3224963796514532e-05, + "loss": 0.95, + "step": 4300 + }, + { + "epoch": 0.413716814159292, + "grad_norm": 2.043944730508104, + "learning_rate": 1.3222014671841048e-05, + "loss": 0.9855, + "step": 4301 + }, + { + "epoch": 0.4138130050019238, + "grad_norm": 1.877161418230952, + "learning_rate": 1.32190652344433e-05, + "loss": 0.9556, + "step": 4302 + }, + { + "epoch": 0.4139091958445556, + "grad_norm": 2.163931484098275, + "learning_rate": 1.3216115484607557e-05, + "loss": 0.9871, + "step": 4303 + }, + { + "epoch": 0.4140053866871874, + "grad_norm": 2.0818684307703963, + "learning_rate": 1.3213165422620111e-05, + "loss": 0.9777, + "step": 4304 + }, + { + "epoch": 0.4141015775298192, + "grad_norm": 2.134410249612343, + "learning_rate": 1.3210215048767297e-05, + "loss": 1.0403, + "step": 4305 + }, + { + "epoch": 0.4141977683724509, + "grad_norm": 2.3452569958514675, + "learning_rate": 1.3207264363335472e-05, + "loss": 0.8853, + "step": 4306 + }, + { + "epoch": 0.4142939592150827, + "grad_norm": 1.9087331893350084, + "learning_rate": 1.3204313366611025e-05, + "loss": 0.989, + "step": 4307 + }, + { + "epoch": 0.4143901500577145, + "grad_norm": 1.9071269356415517, + "learning_rate": 1.3201362058880375e-05, + "loss": 0.931, + "step": 4308 + }, + { + "epoch": 0.4144863409003463, + "grad_norm": 2.295828044954043, + "learning_rate": 1.3198410440429976e-05, + "loss": 0.9794, + "step": 4309 + }, + { + "epoch": 0.41458253174297804, + "grad_norm": 2.2790906541079727, + "learning_rate": 1.3195458511546307e-05, + "loss": 0.9614, + "step": 4310 + }, + { + "epoch": 0.41467872258560984, + "grad_norm": 1.9610878919948571, + "learning_rate": 1.319250627251587e-05, + "loss": 1.0125, + "step": 4311 + }, + { + "epoch": 0.41477491342824163, + "grad_norm": 2.133338917425296, + "learning_rate": 1.3189553723625217e-05, + "loss": 1.0533, + "step": 4312 + }, + { + "epoch": 0.4148711042708734, + "grad_norm": 2.151862936982474, + "learning_rate": 1.3186600865160908e-05, + "loss": 1.0586, + "step": 4313 + }, + { + "epoch": 0.4149672951135052, + "grad_norm": 2.1660779824240146, + "learning_rate": 1.318364769740955e-05, + "loss": 1.022, + "step": 4314 + }, + { + "epoch": 0.41506348595613696, + "grad_norm": 2.1155979564781022, + "learning_rate": 1.3180694220657774e-05, + "loss": 0.9177, + "step": 4315 + }, + { + "epoch": 0.41515967679876875, + "grad_norm": 1.7844584851350365, + "learning_rate": 1.3177740435192235e-05, + "loss": 0.9238, + "step": 4316 + }, + { + "epoch": 0.41525586764140054, + "grad_norm": 2.1430791410898777, + "learning_rate": 1.3174786341299625e-05, + "loss": 0.9825, + "step": 4317 + }, + { + "epoch": 0.41535205848403234, + "grad_norm": 2.0560197545621994, + "learning_rate": 1.3171831939266668e-05, + "loss": 0.9781, + "step": 4318 + }, + { + "epoch": 0.4154482493266641, + "grad_norm": 2.0586810364067714, + "learning_rate": 1.316887722938011e-05, + "loss": 1.1901, + "step": 4319 + }, + { + "epoch": 0.41554444016929587, + "grad_norm": 2.0546392321517892, + "learning_rate": 1.3165922211926734e-05, + "loss": 1.0386, + "step": 4320 + }, + { + "epoch": 0.41564063101192766, + "grad_norm": 1.8593281622568734, + "learning_rate": 1.3162966887193349e-05, + "loss": 0.9265, + "step": 4321 + }, + { + "epoch": 0.41573682185455946, + "grad_norm": 2.0430531865513144, + "learning_rate": 1.3160011255466791e-05, + "loss": 1.0846, + "step": 4322 + }, + { + "epoch": 0.41583301269719125, + "grad_norm": 2.127561457413482, + "learning_rate": 1.3157055317033936e-05, + "loss": 0.9618, + "step": 4323 + }, + { + "epoch": 0.415929203539823, + "grad_norm": 2.3403956017116823, + "learning_rate": 1.3154099072181677e-05, + "loss": 0.8879, + "step": 4324 + }, + { + "epoch": 0.4160253943824548, + "grad_norm": 1.8441745432858634, + "learning_rate": 1.3151142521196945e-05, + "loss": 0.8748, + "step": 4325 + }, + { + "epoch": 0.4161215852250866, + "grad_norm": 2.027374816216007, + "learning_rate": 1.3148185664366704e-05, + "loss": 0.9329, + "step": 4326 + }, + { + "epoch": 0.41621777606771837, + "grad_norm": 2.003382582579115, + "learning_rate": 1.3145228501977934e-05, + "loss": 1.0059, + "step": 4327 + }, + { + "epoch": 0.4163139669103501, + "grad_norm": 2.0255282824605745, + "learning_rate": 1.314227103431766e-05, + "loss": 0.9876, + "step": 4328 + }, + { + "epoch": 0.4164101577529819, + "grad_norm": 2.1502935760485955, + "learning_rate": 1.3139313261672926e-05, + "loss": 0.9548, + "step": 4329 + }, + { + "epoch": 0.4165063485956137, + "grad_norm": 1.9065192417340027, + "learning_rate": 1.3136355184330809e-05, + "loss": 1.0671, + "step": 4330 + }, + { + "epoch": 0.4166025394382455, + "grad_norm": 2.0279715064057524, + "learning_rate": 1.3133396802578418e-05, + "loss": 0.9734, + "step": 4331 + }, + { + "epoch": 0.4166987302808773, + "grad_norm": 1.9800376301737415, + "learning_rate": 1.3130438116702888e-05, + "loss": 0.8945, + "step": 4332 + }, + { + "epoch": 0.416794921123509, + "grad_norm": 1.8883368659227784, + "learning_rate": 1.3127479126991384e-05, + "loss": 0.9874, + "step": 4333 + }, + { + "epoch": 0.4168911119661408, + "grad_norm": 2.094123800642038, + "learning_rate": 1.3124519833731106e-05, + "loss": 0.9264, + "step": 4334 + }, + { + "epoch": 0.4169873028087726, + "grad_norm": 2.216006437609481, + "learning_rate": 1.3121560237209275e-05, + "loss": 1.0291, + "step": 4335 + }, + { + "epoch": 0.4170834936514044, + "grad_norm": 1.7895554465081964, + "learning_rate": 1.3118600337713146e-05, + "loss": 0.9594, + "step": 4336 + }, + { + "epoch": 0.41717968449403614, + "grad_norm": 1.9379266696892887, + "learning_rate": 1.311564013553e-05, + "loss": 1.0135, + "step": 4337 + }, + { + "epoch": 0.41727587533666793, + "grad_norm": 2.304629845832514, + "learning_rate": 1.3112679630947156e-05, + "loss": 0.9959, + "step": 4338 + }, + { + "epoch": 0.4173720661792997, + "grad_norm": 2.1343367281821592, + "learning_rate": 1.3109718824251949e-05, + "loss": 1.1296, + "step": 4339 + }, + { + "epoch": 0.4174682570219315, + "grad_norm": 2.2286420789539325, + "learning_rate": 1.310675771573176e-05, + "loss": 0.984, + "step": 4340 + }, + { + "epoch": 0.4175644478645633, + "grad_norm": 2.2366534469658843, + "learning_rate": 1.3103796305673984e-05, + "loss": 1.0845, + "step": 4341 + }, + { + "epoch": 0.41766063870719505, + "grad_norm": 1.9669510858943686, + "learning_rate": 1.310083459436605e-05, + "loss": 1.0131, + "step": 4342 + }, + { + "epoch": 0.41775682954982685, + "grad_norm": 1.9008821412197954, + "learning_rate": 1.3097872582095423e-05, + "loss": 1.0205, + "step": 4343 + }, + { + "epoch": 0.41785302039245864, + "grad_norm": 1.984928546044024, + "learning_rate": 1.3094910269149587e-05, + "loss": 0.9863, + "step": 4344 + }, + { + "epoch": 0.41794921123509043, + "grad_norm": 2.0850999757162603, + "learning_rate": 1.3091947655816063e-05, + "loss": 0.9459, + "step": 4345 + }, + { + "epoch": 0.41804540207772223, + "grad_norm": 1.7135865989450993, + "learning_rate": 1.3088984742382395e-05, + "loss": 1.0367, + "step": 4346 + }, + { + "epoch": 0.41814159292035397, + "grad_norm": 2.1786107520318247, + "learning_rate": 1.3086021529136167e-05, + "loss": 0.8627, + "step": 4347 + }, + { + "epoch": 0.41823778376298576, + "grad_norm": 2.5482295339993892, + "learning_rate": 1.3083058016364972e-05, + "loss": 0.952, + "step": 4348 + }, + { + "epoch": 0.41833397460561755, + "grad_norm": 2.548332015862751, + "learning_rate": 1.3080094204356454e-05, + "loss": 1.0534, + "step": 4349 + }, + { + "epoch": 0.41843016544824935, + "grad_norm": 2.0124970610767408, + "learning_rate": 1.3077130093398274e-05, + "loss": 0.9025, + "step": 4350 + }, + { + "epoch": 0.4185263562908811, + "grad_norm": 1.9655981114855121, + "learning_rate": 1.307416568377812e-05, + "loss": 0.868, + "step": 4351 + }, + { + "epoch": 0.4186225471335129, + "grad_norm": 2.2243562855020116, + "learning_rate": 1.3071200975783725e-05, + "loss": 1.0314, + "step": 4352 + }, + { + "epoch": 0.4187187379761447, + "grad_norm": 2.0690122108266054, + "learning_rate": 1.3068235969702822e-05, + "loss": 0.9638, + "step": 4353 + }, + { + "epoch": 0.41881492881877647, + "grad_norm": 1.7544621827111995, + "learning_rate": 1.3065270665823206e-05, + "loss": 0.8551, + "step": 4354 + }, + { + "epoch": 0.41891111966140826, + "grad_norm": 1.966360044996326, + "learning_rate": 1.3062305064432684e-05, + "loss": 0.9469, + "step": 4355 + }, + { + "epoch": 0.41900731050404, + "grad_norm": 2.483485981288788, + "learning_rate": 1.3059339165819082e-05, + "loss": 1.0727, + "step": 4356 + }, + { + "epoch": 0.4191035013466718, + "grad_norm": 2.060452960685288, + "learning_rate": 1.3056372970270274e-05, + "loss": 1.029, + "step": 4357 + }, + { + "epoch": 0.4191996921893036, + "grad_norm": 2.071863885174645, + "learning_rate": 1.3053406478074155e-05, + "loss": 0.9339, + "step": 4358 + }, + { + "epoch": 0.4192958830319354, + "grad_norm": 2.0718707625422907, + "learning_rate": 1.3050439689518646e-05, + "loss": 1.0608, + "step": 4359 + }, + { + "epoch": 0.4193920738745671, + "grad_norm": 2.0820384950633604, + "learning_rate": 1.3047472604891701e-05, + "loss": 1.0175, + "step": 4360 + }, + { + "epoch": 0.4194882647171989, + "grad_norm": 2.088966703520462, + "learning_rate": 1.30445052244813e-05, + "loss": 0.9674, + "step": 4361 + }, + { + "epoch": 0.4195844555598307, + "grad_norm": 2.001827673999608, + "learning_rate": 1.3041537548575455e-05, + "loss": 1.0712, + "step": 4362 + }, + { + "epoch": 0.4196806464024625, + "grad_norm": 2.3276090749600016, + "learning_rate": 1.30385695774622e-05, + "loss": 1.0368, + "step": 4363 + }, + { + "epoch": 0.4197768372450943, + "grad_norm": 1.9771526430524478, + "learning_rate": 1.303560131142961e-05, + "loss": 0.9485, + "step": 4364 + }, + { + "epoch": 0.41987302808772603, + "grad_norm": 2.019006479006924, + "learning_rate": 1.303263275076577e-05, + "loss": 1.0852, + "step": 4365 + }, + { + "epoch": 0.4199692189303578, + "grad_norm": 2.3163062312100564, + "learning_rate": 1.3029663895758814e-05, + "loss": 1.1665, + "step": 4366 + }, + { + "epoch": 0.4200654097729896, + "grad_norm": 1.8394914488066962, + "learning_rate": 1.302669474669689e-05, + "loss": 0.9938, + "step": 4367 + }, + { + "epoch": 0.4201616006156214, + "grad_norm": 2.082534692574384, + "learning_rate": 1.3023725303868183e-05, + "loss": 0.9152, + "step": 4368 + }, + { + "epoch": 0.42025779145825315, + "grad_norm": 2.0878684177202578, + "learning_rate": 1.30207555675609e-05, + "loss": 1.0577, + "step": 4369 + }, + { + "epoch": 0.42035398230088494, + "grad_norm": 1.8658813776294283, + "learning_rate": 1.3017785538063277e-05, + "loss": 1.0567, + "step": 4370 + }, + { + "epoch": 0.42045017314351674, + "grad_norm": 1.8866231593364164, + "learning_rate": 1.3014815215663588e-05, + "loss": 0.9918, + "step": 4371 + }, + { + "epoch": 0.42054636398614853, + "grad_norm": 1.945426728233761, + "learning_rate": 1.3011844600650121e-05, + "loss": 0.9594, + "step": 4372 + }, + { + "epoch": 0.4206425548287803, + "grad_norm": 1.942087689974858, + "learning_rate": 1.3008873693311205e-05, + "loss": 1.1038, + "step": 4373 + }, + { + "epoch": 0.42073874567141206, + "grad_norm": 2.418619962987544, + "learning_rate": 1.300590249393519e-05, + "loss": 0.957, + "step": 4374 + }, + { + "epoch": 0.42083493651404386, + "grad_norm": 1.994883877760787, + "learning_rate": 1.3002931002810457e-05, + "loss": 1.0305, + "step": 4375 + }, + { + "epoch": 0.42093112735667565, + "grad_norm": 2.0643895032000863, + "learning_rate": 1.2999959220225416e-05, + "loss": 1.065, + "step": 4376 + }, + { + "epoch": 0.42102731819930744, + "grad_norm": 1.803430227876067, + "learning_rate": 1.2996987146468499e-05, + "loss": 0.9056, + "step": 4377 + }, + { + "epoch": 0.4211235090419392, + "grad_norm": 2.142800689371254, + "learning_rate": 1.299401478182818e-05, + "loss": 0.9844, + "step": 4378 + }, + { + "epoch": 0.421219699884571, + "grad_norm": 2.028028794361931, + "learning_rate": 1.2991042126592941e-05, + "loss": 1.0816, + "step": 4379 + }, + { + "epoch": 0.42131589072720277, + "grad_norm": 1.8452752938304413, + "learning_rate": 1.2988069181051314e-05, + "loss": 0.8286, + "step": 4380 + }, + { + "epoch": 0.42141208156983456, + "grad_norm": 1.9549199483970001, + "learning_rate": 1.2985095945491849e-05, + "loss": 0.9991, + "step": 4381 + }, + { + "epoch": 0.42150827241246636, + "grad_norm": 2.4410940152943126, + "learning_rate": 1.2982122420203114e-05, + "loss": 0.9876, + "step": 4382 + }, + { + "epoch": 0.4216044632550981, + "grad_norm": 1.7450989995167556, + "learning_rate": 1.2979148605473725e-05, + "loss": 1.0129, + "step": 4383 + }, + { + "epoch": 0.4217006540977299, + "grad_norm": 2.3354477905927817, + "learning_rate": 1.2976174501592313e-05, + "loss": 1.0226, + "step": 4384 + }, + { + "epoch": 0.4217968449403617, + "grad_norm": 1.9215174023941508, + "learning_rate": 1.2973200108847542e-05, + "loss": 0.8778, + "step": 4385 + }, + { + "epoch": 0.4218930357829935, + "grad_norm": 2.214251612570898, + "learning_rate": 1.2970225427528098e-05, + "loss": 1.0719, + "step": 4386 + }, + { + "epoch": 0.4219892266256252, + "grad_norm": 1.8779467219126857, + "learning_rate": 1.2967250457922705e-05, + "loss": 0.8856, + "step": 4387 + }, + { + "epoch": 0.422085417468257, + "grad_norm": 2.6538609583984387, + "learning_rate": 1.2964275200320104e-05, + "loss": 0.8966, + "step": 4388 + }, + { + "epoch": 0.4221816083108888, + "grad_norm": 2.001728196422684, + "learning_rate": 1.2961299655009072e-05, + "loss": 1.0408, + "step": 4389 + }, + { + "epoch": 0.4222777991535206, + "grad_norm": 2.227946121613281, + "learning_rate": 1.2958323822278413e-05, + "loss": 0.9609, + "step": 4390 + }, + { + "epoch": 0.4223739899961524, + "grad_norm": 2.3219583024501915, + "learning_rate": 1.2955347702416954e-05, + "loss": 0.9799, + "step": 4391 + }, + { + "epoch": 0.4224701808387841, + "grad_norm": 1.9565686634146995, + "learning_rate": 1.2952371295713558e-05, + "loss": 0.9783, + "step": 4392 + }, + { + "epoch": 0.4225663716814159, + "grad_norm": 2.6068413656836102, + "learning_rate": 1.2949394602457104e-05, + "loss": 1.0729, + "step": 4393 + }, + { + "epoch": 0.4226625625240477, + "grad_norm": 2.716507744936942, + "learning_rate": 1.2946417622936512e-05, + "loss": 1.1541, + "step": 4394 + }, + { + "epoch": 0.4227587533666795, + "grad_norm": 2.1888468209428966, + "learning_rate": 1.2943440357440719e-05, + "loss": 0.9479, + "step": 4395 + }, + { + "epoch": 0.4228549442093113, + "grad_norm": 1.9565204684501716, + "learning_rate": 1.2940462806258696e-05, + "loss": 1.0448, + "step": 4396 + }, + { + "epoch": 0.42295113505194304, + "grad_norm": 2.3727275865955626, + "learning_rate": 1.2937484969679441e-05, + "loss": 1.0331, + "step": 4397 + }, + { + "epoch": 0.42304732589457483, + "grad_norm": 2.0296700813713446, + "learning_rate": 1.2934506847991976e-05, + "loss": 1.0533, + "step": 4398 + }, + { + "epoch": 0.4231435167372066, + "grad_norm": 2.0039688409939806, + "learning_rate": 1.2931528441485357e-05, + "loss": 0.9352, + "step": 4399 + }, + { + "epoch": 0.4232397075798384, + "grad_norm": 1.9119375115178079, + "learning_rate": 1.2928549750448661e-05, + "loss": 0.9279, + "step": 4400 + }, + { + "epoch": 0.42333589842247016, + "grad_norm": 2.152248330889395, + "learning_rate": 1.2925570775170994e-05, + "loss": 0.8491, + "step": 4401 + }, + { + "epoch": 0.42343208926510195, + "grad_norm": 2.477840961754527, + "learning_rate": 1.2922591515941498e-05, + "loss": 0.8814, + "step": 4402 + }, + { + "epoch": 0.42352828010773375, + "grad_norm": 1.8593300371205497, + "learning_rate": 1.2919611973049328e-05, + "loss": 0.9539, + "step": 4403 + }, + { + "epoch": 0.42362447095036554, + "grad_norm": 1.8994937041187028, + "learning_rate": 1.2916632146783683e-05, + "loss": 1.0076, + "step": 4404 + }, + { + "epoch": 0.42372066179299733, + "grad_norm": 2.2005219779328598, + "learning_rate": 1.2913652037433767e-05, + "loss": 0.957, + "step": 4405 + }, + { + "epoch": 0.42381685263562907, + "grad_norm": 1.9671294254576204, + "learning_rate": 1.2910671645288841e-05, + "loss": 0.9399, + "step": 4406 + }, + { + "epoch": 0.42391304347826086, + "grad_norm": 1.869740632161371, + "learning_rate": 1.290769097063817e-05, + "loss": 0.9046, + "step": 4407 + }, + { + "epoch": 0.42400923432089266, + "grad_norm": 2.1834618491262683, + "learning_rate": 1.2904710013771054e-05, + "loss": 1.0831, + "step": 4408 + }, + { + "epoch": 0.42410542516352445, + "grad_norm": 1.929247914890025, + "learning_rate": 1.290172877497682e-05, + "loss": 0.9966, + "step": 4409 + }, + { + "epoch": 0.4242016160061562, + "grad_norm": 1.7994635893464919, + "learning_rate": 1.2898747254544826e-05, + "loss": 0.8801, + "step": 4410 + }, + { + "epoch": 0.424297806848788, + "grad_norm": 2.1481707137910697, + "learning_rate": 1.289576545276445e-05, + "loss": 1.0153, + "step": 4411 + }, + { + "epoch": 0.4243939976914198, + "grad_norm": 1.8900225416458114, + "learning_rate": 1.2892783369925105e-05, + "loss": 0.9414, + "step": 4412 + }, + { + "epoch": 0.42449018853405157, + "grad_norm": 1.921254769861582, + "learning_rate": 1.2889801006316229e-05, + "loss": 1.0733, + "step": 4413 + }, + { + "epoch": 0.42458637937668336, + "grad_norm": 2.012766301458718, + "learning_rate": 1.2886818362227283e-05, + "loss": 1.035, + "step": 4414 + }, + { + "epoch": 0.4246825702193151, + "grad_norm": 1.9700191679817807, + "learning_rate": 1.2883835437947756e-05, + "loss": 0.9898, + "step": 4415 + }, + { + "epoch": 0.4247787610619469, + "grad_norm": 2.6270679564484327, + "learning_rate": 1.2880852233767174e-05, + "loss": 0.9259, + "step": 4416 + }, + { + "epoch": 0.4248749519045787, + "grad_norm": 2.193013498641088, + "learning_rate": 1.2877868749975074e-05, + "loss": 0.9652, + "step": 4417 + }, + { + "epoch": 0.4249711427472105, + "grad_norm": 2.281171752396599, + "learning_rate": 1.2874884986861038e-05, + "loss": 1.0884, + "step": 4418 + }, + { + "epoch": 0.4250673335898422, + "grad_norm": 1.7290129085639014, + "learning_rate": 1.2871900944714655e-05, + "loss": 0.9558, + "step": 4419 + }, + { + "epoch": 0.425163524432474, + "grad_norm": 2.225676532813948, + "learning_rate": 1.2868916623825561e-05, + "loss": 0.9629, + "step": 4420 + }, + { + "epoch": 0.4252597152751058, + "grad_norm": 1.9966527252506274, + "learning_rate": 1.286593202448341e-05, + "loss": 1.0935, + "step": 4421 + }, + { + "epoch": 0.4253559061177376, + "grad_norm": 2.1524690099691868, + "learning_rate": 1.2862947146977876e-05, + "loss": 1.0952, + "step": 4422 + }, + { + "epoch": 0.4254520969603694, + "grad_norm": 2.0700774242704205, + "learning_rate": 1.2859961991598671e-05, + "loss": 0.9945, + "step": 4423 + }, + { + "epoch": 0.42554828780300114, + "grad_norm": 2.2756179430829118, + "learning_rate": 1.2856976558635532e-05, + "loss": 1.1099, + "step": 4424 + }, + { + "epoch": 0.42564447864563293, + "grad_norm": 2.162662152515934, + "learning_rate": 1.2853990848378217e-05, + "loss": 0.8765, + "step": 4425 + }, + { + "epoch": 0.4257406694882647, + "grad_norm": 2.0810535003650728, + "learning_rate": 1.2851004861116519e-05, + "loss": 1.088, + "step": 4426 + }, + { + "epoch": 0.4258368603308965, + "grad_norm": 1.9890489405269736, + "learning_rate": 1.2848018597140253e-05, + "loss": 1.0437, + "step": 4427 + }, + { + "epoch": 0.42593305117352825, + "grad_norm": 1.940073780545283, + "learning_rate": 1.2845032056739257e-05, + "loss": 0.9998, + "step": 4428 + }, + { + "epoch": 0.42602924201616005, + "grad_norm": 2.2095006775897685, + "learning_rate": 1.2842045240203404e-05, + "loss": 1.0181, + "step": 4429 + }, + { + "epoch": 0.42612543285879184, + "grad_norm": 1.9753675481440447, + "learning_rate": 1.2839058147822595e-05, + "loss": 1.0288, + "step": 4430 + }, + { + "epoch": 0.42622162370142364, + "grad_norm": 1.987515630892843, + "learning_rate": 1.283607077988674e-05, + "loss": 1.0515, + "step": 4431 + }, + { + "epoch": 0.42631781454405543, + "grad_norm": 2.1595927028449884, + "learning_rate": 1.2833083136685803e-05, + "loss": 0.9841, + "step": 4432 + }, + { + "epoch": 0.42641400538668717, + "grad_norm": 1.993230222831204, + "learning_rate": 1.2830095218509752e-05, + "loss": 1.0144, + "step": 4433 + }, + { + "epoch": 0.42651019622931896, + "grad_norm": 2.0747639578630674, + "learning_rate": 1.2827107025648595e-05, + "loss": 0.9088, + "step": 4434 + }, + { + "epoch": 0.42660638707195075, + "grad_norm": 1.9440294521878239, + "learning_rate": 1.2824118558392358e-05, + "loss": 1.0339, + "step": 4435 + }, + { + "epoch": 0.42670257791458255, + "grad_norm": 1.9873025920270753, + "learning_rate": 1.2821129817031099e-05, + "loss": 1.1487, + "step": 4436 + }, + { + "epoch": 0.4267987687572143, + "grad_norm": 1.97212652203668, + "learning_rate": 1.2818140801854901e-05, + "loss": 1.0617, + "step": 4437 + }, + { + "epoch": 0.4268949595998461, + "grad_norm": 2.1279071537574397, + "learning_rate": 1.2815151513153874e-05, + "loss": 1.0562, + "step": 4438 + }, + { + "epoch": 0.4269911504424779, + "grad_norm": 1.9283174130304515, + "learning_rate": 1.2812161951218155e-05, + "loss": 1.0788, + "step": 4439 + }, + { + "epoch": 0.42708734128510967, + "grad_norm": 1.7730007281504123, + "learning_rate": 1.2809172116337903e-05, + "loss": 1.0317, + "step": 4440 + }, + { + "epoch": 0.42718353212774146, + "grad_norm": 2.029322317500773, + "learning_rate": 1.280618200880331e-05, + "loss": 1.0982, + "step": 4441 + }, + { + "epoch": 0.4272797229703732, + "grad_norm": 2.240661826102564, + "learning_rate": 1.2803191628904594e-05, + "loss": 0.9585, + "step": 4442 + }, + { + "epoch": 0.427375913813005, + "grad_norm": 1.8484909488695138, + "learning_rate": 1.2800200976931994e-05, + "loss": 0.9813, + "step": 4443 + }, + { + "epoch": 0.4274721046556368, + "grad_norm": 2.154147200108876, + "learning_rate": 1.2797210053175779e-05, + "loss": 1.0485, + "step": 4444 + }, + { + "epoch": 0.4275682954982686, + "grad_norm": 1.8920769732527356, + "learning_rate": 1.2794218857926241e-05, + "loss": 0.9656, + "step": 4445 + }, + { + "epoch": 0.4276644863409004, + "grad_norm": 2.0807375343201984, + "learning_rate": 1.2791227391473706e-05, + "loss": 0.8876, + "step": 4446 + }, + { + "epoch": 0.4277606771835321, + "grad_norm": 2.1474233320943616, + "learning_rate": 1.2788235654108525e-05, + "loss": 0.9615, + "step": 4447 + }, + { + "epoch": 0.4278568680261639, + "grad_norm": 2.2511703188937053, + "learning_rate": 1.2785243646121059e-05, + "loss": 1.0499, + "step": 4448 + }, + { + "epoch": 0.4279530588687957, + "grad_norm": 1.9384089577442112, + "learning_rate": 1.2782251367801722e-05, + "loss": 0.986, + "step": 4449 + }, + { + "epoch": 0.4280492497114275, + "grad_norm": 2.1279723055556605, + "learning_rate": 1.277925881944093e-05, + "loss": 1.0529, + "step": 4450 + }, + { + "epoch": 0.42814544055405923, + "grad_norm": 2.0495383443740565, + "learning_rate": 1.2776266001329142e-05, + "loss": 0.9603, + "step": 4451 + }, + { + "epoch": 0.428241631396691, + "grad_norm": 2.084720827436572, + "learning_rate": 1.2773272913756833e-05, + "loss": 0.9084, + "step": 4452 + }, + { + "epoch": 0.4283378222393228, + "grad_norm": 1.8749390856989299, + "learning_rate": 1.2770279557014514e-05, + "loss": 1.0194, + "step": 4453 + }, + { + "epoch": 0.4284340130819546, + "grad_norm": 2.088266905152646, + "learning_rate": 1.2767285931392705e-05, + "loss": 0.8961, + "step": 4454 + }, + { + "epoch": 0.4285302039245864, + "grad_norm": 2.1123092727330994, + "learning_rate": 1.2764292037181974e-05, + "loss": 1.0178, + "step": 4455 + }, + { + "epoch": 0.42862639476721814, + "grad_norm": 2.0084319884541517, + "learning_rate": 1.27612978746729e-05, + "loss": 1.0292, + "step": 4456 + }, + { + "epoch": 0.42872258560984994, + "grad_norm": 1.707991557698537, + "learning_rate": 1.2758303444156088e-05, + "loss": 0.9109, + "step": 4457 + }, + { + "epoch": 0.42881877645248173, + "grad_norm": 2.238811177266488, + "learning_rate": 1.2755308745922182e-05, + "loss": 1.0081, + "step": 4458 + }, + { + "epoch": 0.4289149672951135, + "grad_norm": 2.034003473226225, + "learning_rate": 1.2752313780261835e-05, + "loss": 1.0882, + "step": 4459 + }, + { + "epoch": 0.42901115813774526, + "grad_norm": 1.918937005562645, + "learning_rate": 1.2749318547465742e-05, + "loss": 1.007, + "step": 4460 + }, + { + "epoch": 0.42910734898037706, + "grad_norm": 2.0303805665991606, + "learning_rate": 1.2746323047824612e-05, + "loss": 0.9522, + "step": 4461 + }, + { + "epoch": 0.42920353982300885, + "grad_norm": 2.3444867947188186, + "learning_rate": 1.2743327281629181e-05, + "loss": 1.1237, + "step": 4462 + }, + { + "epoch": 0.42929973066564064, + "grad_norm": 1.9135156649294043, + "learning_rate": 1.2740331249170219e-05, + "loss": 0.867, + "step": 4463 + }, + { + "epoch": 0.42939592150827244, + "grad_norm": 1.867968271206877, + "learning_rate": 1.2737334950738512e-05, + "loss": 0.887, + "step": 4464 + }, + { + "epoch": 0.4294921123509042, + "grad_norm": 2.0796363705456926, + "learning_rate": 1.2734338386624882e-05, + "loss": 0.993, + "step": 4465 + }, + { + "epoch": 0.42958830319353597, + "grad_norm": 2.245293816598467, + "learning_rate": 1.273134155712017e-05, + "loss": 1.0307, + "step": 4466 + }, + { + "epoch": 0.42968449403616776, + "grad_norm": 2.077429149157342, + "learning_rate": 1.2728344462515238e-05, + "loss": 1.1281, + "step": 4467 + }, + { + "epoch": 0.42978068487879956, + "grad_norm": 1.9229203863032756, + "learning_rate": 1.272534710310099e-05, + "loss": 0.9225, + "step": 4468 + }, + { + "epoch": 0.4298768757214313, + "grad_norm": 1.758068722729123, + "learning_rate": 1.2722349479168335e-05, + "loss": 0.8932, + "step": 4469 + }, + { + "epoch": 0.4299730665640631, + "grad_norm": 2.5429384711516083, + "learning_rate": 1.2719351591008228e-05, + "loss": 0.984, + "step": 4470 + }, + { + "epoch": 0.4300692574066949, + "grad_norm": 2.0430770927754276, + "learning_rate": 1.271635343891163e-05, + "loss": 0.9148, + "step": 4471 + }, + { + "epoch": 0.4301654482493267, + "grad_norm": 2.2434710643390443, + "learning_rate": 1.2713355023169547e-05, + "loss": 1.0909, + "step": 4472 + }, + { + "epoch": 0.43026163909195847, + "grad_norm": 2.174849104868877, + "learning_rate": 1.2710356344072993e-05, + "loss": 1.0499, + "step": 4473 + }, + { + "epoch": 0.4303578299345902, + "grad_norm": 1.9810454448237047, + "learning_rate": 1.2707357401913022e-05, + "loss": 0.9903, + "step": 4474 + }, + { + "epoch": 0.430454020777222, + "grad_norm": 2.0633479397598413, + "learning_rate": 1.2704358196980707e-05, + "loss": 1.0145, + "step": 4475 + }, + { + "epoch": 0.4305502116198538, + "grad_norm": 2.0489747218084817, + "learning_rate": 1.270135872956714e-05, + "loss": 0.9345, + "step": 4476 + }, + { + "epoch": 0.4306464024624856, + "grad_norm": 2.1456041338733955, + "learning_rate": 1.2698358999963452e-05, + "loss": 0.9348, + "step": 4477 + }, + { + "epoch": 0.4307425933051173, + "grad_norm": 2.0668074492608697, + "learning_rate": 1.2695359008460785e-05, + "loss": 1.0285, + "step": 4478 + }, + { + "epoch": 0.4308387841477491, + "grad_norm": 2.114250894247961, + "learning_rate": 1.2692358755350325e-05, + "loss": 1.126, + "step": 4479 + }, + { + "epoch": 0.4309349749903809, + "grad_norm": 2.2569483465604825, + "learning_rate": 1.2689358240923264e-05, + "loss": 1.0948, + "step": 4480 + }, + { + "epoch": 0.4310311658330127, + "grad_norm": 1.9156256456587168, + "learning_rate": 1.2686357465470827e-05, + "loss": 0.9484, + "step": 4481 + }, + { + "epoch": 0.4311273566756445, + "grad_norm": 1.9385554979345636, + "learning_rate": 1.2683356429284273e-05, + "loss": 0.9397, + "step": 4482 + }, + { + "epoch": 0.43122354751827624, + "grad_norm": 1.8908134790964872, + "learning_rate": 1.2680355132654868e-05, + "loss": 0.9502, + "step": 4483 + }, + { + "epoch": 0.43131973836090803, + "grad_norm": 2.12401884949432, + "learning_rate": 1.2677353575873926e-05, + "loss": 1.0411, + "step": 4484 + }, + { + "epoch": 0.4314159292035398, + "grad_norm": 2.124676884999901, + "learning_rate": 1.2674351759232764e-05, + "loss": 1.1007, + "step": 4485 + }, + { + "epoch": 0.4315121200461716, + "grad_norm": 2.0866795094981034, + "learning_rate": 1.2671349683022736e-05, + "loss": 1.0356, + "step": 4486 + }, + { + "epoch": 0.43160831088880336, + "grad_norm": 2.4255411293937224, + "learning_rate": 1.2668347347535222e-05, + "loss": 0.9202, + "step": 4487 + }, + { + "epoch": 0.43170450173143515, + "grad_norm": 2.269549905251864, + "learning_rate": 1.2665344753061622e-05, + "loss": 1.0644, + "step": 4488 + }, + { + "epoch": 0.43180069257406695, + "grad_norm": 2.3542605481946093, + "learning_rate": 1.2662341899893366e-05, + "loss": 0.9685, + "step": 4489 + }, + { + "epoch": 0.43189688341669874, + "grad_norm": 1.8980856763748324, + "learning_rate": 1.2659338788321904e-05, + "loss": 0.9394, + "step": 4490 + }, + { + "epoch": 0.43199307425933053, + "grad_norm": 1.9516408703893333, + "learning_rate": 1.265633541863872e-05, + "loss": 0.9804, + "step": 4491 + }, + { + "epoch": 0.43208926510196227, + "grad_norm": 2.107175123503912, + "learning_rate": 1.2653331791135308e-05, + "loss": 1.0157, + "step": 4492 + }, + { + "epoch": 0.43218545594459407, + "grad_norm": 1.840767607776085, + "learning_rate": 1.26503279061032e-05, + "loss": 1.0786, + "step": 4493 + }, + { + "epoch": 0.43228164678722586, + "grad_norm": 2.1420660521745916, + "learning_rate": 1.2647323763833952e-05, + "loss": 0.9868, + "step": 4494 + }, + { + "epoch": 0.43237783762985765, + "grad_norm": 1.9164474331163965, + "learning_rate": 1.2644319364619136e-05, + "loss": 0.9817, + "step": 4495 + }, + { + "epoch": 0.4324740284724894, + "grad_norm": 1.8239926151008494, + "learning_rate": 1.264131470875036e-05, + "loss": 0.9709, + "step": 4496 + }, + { + "epoch": 0.4325702193151212, + "grad_norm": 2.2927340830723, + "learning_rate": 1.2638309796519247e-05, + "loss": 0.9163, + "step": 4497 + }, + { + "epoch": 0.432666410157753, + "grad_norm": 2.1976431895017714, + "learning_rate": 1.2635304628217452e-05, + "loss": 1.1779, + "step": 4498 + }, + { + "epoch": 0.4327626010003848, + "grad_norm": 2.051574848197481, + "learning_rate": 1.2632299204136658e-05, + "loss": 1.0276, + "step": 4499 + }, + { + "epoch": 0.43285879184301657, + "grad_norm": 2.1531872467931685, + "learning_rate": 1.2629293524568555e-05, + "loss": 1.0237, + "step": 4500 + }, + { + "epoch": 0.4329549826856483, + "grad_norm": 1.9832952380646984, + "learning_rate": 1.2626287589804882e-05, + "loss": 0.861, + "step": 4501 + }, + { + "epoch": 0.4330511735282801, + "grad_norm": 2.023303467278798, + "learning_rate": 1.2623281400137383e-05, + "loss": 1.0029, + "step": 4502 + }, + { + "epoch": 0.4331473643709119, + "grad_norm": 1.9111676701158251, + "learning_rate": 1.2620274955857839e-05, + "loss": 1.0107, + "step": 4503 + }, + { + "epoch": 0.4332435552135437, + "grad_norm": 1.9931904659174804, + "learning_rate": 1.2617268257258051e-05, + "loss": 0.9155, + "step": 4504 + }, + { + "epoch": 0.4333397460561755, + "grad_norm": 1.8101759968873783, + "learning_rate": 1.261426130462984e-05, + "loss": 0.9555, + "step": 4505 + }, + { + "epoch": 0.4334359368988072, + "grad_norm": 2.266182872968236, + "learning_rate": 1.2611254098265063e-05, + "loss": 0.8703, + "step": 4506 + }, + { + "epoch": 0.433532127741439, + "grad_norm": 2.0785753407447753, + "learning_rate": 1.2608246638455588e-05, + "loss": 0.9663, + "step": 4507 + }, + { + "epoch": 0.4336283185840708, + "grad_norm": 2.0411068576568088, + "learning_rate": 1.2605238925493326e-05, + "loss": 1.1177, + "step": 4508 + }, + { + "epoch": 0.4337245094267026, + "grad_norm": 2.1883004769193146, + "learning_rate": 1.2602230959670189e-05, + "loss": 0.887, + "step": 4509 + }, + { + "epoch": 0.43382070026933434, + "grad_norm": 2.0211873999705903, + "learning_rate": 1.2599222741278136e-05, + "loss": 1.0361, + "step": 4510 + }, + { + "epoch": 0.43391689111196613, + "grad_norm": 2.1198453349570086, + "learning_rate": 1.2596214270609135e-05, + "loss": 1.0269, + "step": 4511 + }, + { + "epoch": 0.4340130819545979, + "grad_norm": 2.0397545967656634, + "learning_rate": 1.2593205547955185e-05, + "loss": 0.9707, + "step": 4512 + }, + { + "epoch": 0.4341092727972297, + "grad_norm": 1.9262899391699302, + "learning_rate": 1.259019657360831e-05, + "loss": 0.8855, + "step": 4513 + }, + { + "epoch": 0.4342054636398615, + "grad_norm": 2.0408848261421078, + "learning_rate": 1.2587187347860554e-05, + "loss": 0.8765, + "step": 4514 + }, + { + "epoch": 0.43430165448249325, + "grad_norm": 1.8794512664071075, + "learning_rate": 1.2584177871003993e-05, + "loss": 0.9165, + "step": 4515 + }, + { + "epoch": 0.43439784532512504, + "grad_norm": 2.43331805078555, + "learning_rate": 1.2581168143330716e-05, + "loss": 0.9849, + "step": 4516 + }, + { + "epoch": 0.43449403616775684, + "grad_norm": 1.7939388056225867, + "learning_rate": 1.257815816513285e-05, + "loss": 0.9769, + "step": 4517 + }, + { + "epoch": 0.43459022701038863, + "grad_norm": 2.2132679711970775, + "learning_rate": 1.2575147936702531e-05, + "loss": 0.9473, + "step": 4518 + }, + { + "epoch": 0.43468641785302037, + "grad_norm": 1.933378156448785, + "learning_rate": 1.2572137458331937e-05, + "loss": 0.9181, + "step": 4519 + }, + { + "epoch": 0.43478260869565216, + "grad_norm": 2.1893638401450493, + "learning_rate": 1.2569126730313255e-05, + "loss": 1.0182, + "step": 4520 + }, + { + "epoch": 0.43487879953828396, + "grad_norm": 1.8562908169216652, + "learning_rate": 1.25661157529387e-05, + "loss": 1.0932, + "step": 4521 + }, + { + "epoch": 0.43497499038091575, + "grad_norm": 1.7989943211942956, + "learning_rate": 1.2563104526500523e-05, + "loss": 0.908, + "step": 4522 + }, + { + "epoch": 0.43507118122354754, + "grad_norm": 1.9149455738423617, + "learning_rate": 1.2560093051290978e-05, + "loss": 0.9484, + "step": 4523 + }, + { + "epoch": 0.4351673720661793, + "grad_norm": 1.7482302166147268, + "learning_rate": 1.2557081327602361e-05, + "loss": 1.0307, + "step": 4524 + }, + { + "epoch": 0.4352635629088111, + "grad_norm": 2.302178582789786, + "learning_rate": 1.2554069355726986e-05, + "loss": 0.96, + "step": 4525 + }, + { + "epoch": 0.43535975375144287, + "grad_norm": 1.7776610401588175, + "learning_rate": 1.2551057135957187e-05, + "loss": 0.9376, + "step": 4526 + }, + { + "epoch": 0.43545594459407466, + "grad_norm": 1.8245823195887532, + "learning_rate": 1.254804466858533e-05, + "loss": 0.8994, + "step": 4527 + }, + { + "epoch": 0.4355521354367064, + "grad_norm": 1.9904601836883613, + "learning_rate": 1.2545031953903796e-05, + "loss": 0.9901, + "step": 4528 + }, + { + "epoch": 0.4356483262793382, + "grad_norm": 2.1098119377488054, + "learning_rate": 1.2542018992205001e-05, + "loss": 1.0351, + "step": 4529 + }, + { + "epoch": 0.43574451712197, + "grad_norm": 1.7848352523133915, + "learning_rate": 1.2539005783781374e-05, + "loss": 1.0399, + "step": 4530 + }, + { + "epoch": 0.4358407079646018, + "grad_norm": 1.9721846752936365, + "learning_rate": 1.2535992328925377e-05, + "loss": 1.1195, + "step": 4531 + }, + { + "epoch": 0.4359368988072336, + "grad_norm": 2.1022920065553525, + "learning_rate": 1.2532978627929486e-05, + "loss": 1.0821, + "step": 4532 + }, + { + "epoch": 0.4360330896498653, + "grad_norm": 2.158935766067678, + "learning_rate": 1.2529964681086211e-05, + "loss": 1.0199, + "step": 4533 + }, + { + "epoch": 0.4361292804924971, + "grad_norm": 2.009106420692235, + "learning_rate": 1.2526950488688083e-05, + "loss": 1.0865, + "step": 4534 + }, + { + "epoch": 0.4362254713351289, + "grad_norm": 2.0383279875539966, + "learning_rate": 1.2523936051027649e-05, + "loss": 0.9704, + "step": 4535 + }, + { + "epoch": 0.4363216621777607, + "grad_norm": 2.179659297152683, + "learning_rate": 1.2520921368397492e-05, + "loss": 0.9616, + "step": 4536 + }, + { + "epoch": 0.43641785302039243, + "grad_norm": 2.1560081090677006, + "learning_rate": 1.251790644109021e-05, + "loss": 1.0302, + "step": 4537 + }, + { + "epoch": 0.4365140438630242, + "grad_norm": 2.2123120596648578, + "learning_rate": 1.2514891269398429e-05, + "loss": 1.0241, + "step": 4538 + }, + { + "epoch": 0.436610234705656, + "grad_norm": 2.1344876661931846, + "learning_rate": 1.2511875853614796e-05, + "loss": 1.0023, + "step": 4539 + }, + { + "epoch": 0.4367064255482878, + "grad_norm": 1.9243085988838395, + "learning_rate": 1.2508860194031986e-05, + "loss": 0.9474, + "step": 4540 + }, + { + "epoch": 0.4368026163909196, + "grad_norm": 2.0935214870376706, + "learning_rate": 1.250584429094269e-05, + "loss": 0.8245, + "step": 4541 + }, + { + "epoch": 0.43689880723355135, + "grad_norm": 2.0592655798240433, + "learning_rate": 1.2502828144639629e-05, + "loss": 1.0147, + "step": 4542 + }, + { + "epoch": 0.43699499807618314, + "grad_norm": 1.91760914586063, + "learning_rate": 1.2499811755415552e-05, + "loss": 1.0439, + "step": 4543 + }, + { + "epoch": 0.43709118891881493, + "grad_norm": 2.1284748927078447, + "learning_rate": 1.2496795123563218e-05, + "loss": 1.0216, + "step": 4544 + }, + { + "epoch": 0.4371873797614467, + "grad_norm": 2.1124087149259414, + "learning_rate": 1.249377824937542e-05, + "loss": 1.0442, + "step": 4545 + }, + { + "epoch": 0.43728357060407846, + "grad_norm": 2.2045085977917647, + "learning_rate": 1.249076113314497e-05, + "loss": 1.0229, + "step": 4546 + }, + { + "epoch": 0.43737976144671026, + "grad_norm": 1.944535870971776, + "learning_rate": 1.2487743775164704e-05, + "loss": 0.8481, + "step": 4547 + }, + { + "epoch": 0.43747595228934205, + "grad_norm": 1.9122702660286712, + "learning_rate": 1.248472617572749e-05, + "loss": 1.0044, + "step": 4548 + }, + { + "epoch": 0.43757214313197385, + "grad_norm": 1.9304735523178662, + "learning_rate": 1.2481708335126203e-05, + "loss": 1.0685, + "step": 4549 + }, + { + "epoch": 0.43766833397460564, + "grad_norm": 1.8884020202399572, + "learning_rate": 1.2478690253653756e-05, + "loss": 0.9641, + "step": 4550 + }, + { + "epoch": 0.4377645248172374, + "grad_norm": 2.1365899442280765, + "learning_rate": 1.2475671931603077e-05, + "loss": 1.0494, + "step": 4551 + }, + { + "epoch": 0.43786071565986917, + "grad_norm": 2.1446956174743796, + "learning_rate": 1.2472653369267122e-05, + "loss": 0.9112, + "step": 4552 + }, + { + "epoch": 0.43795690650250096, + "grad_norm": 2.1616111911108336, + "learning_rate": 1.2469634566938869e-05, + "loss": 1.0988, + "step": 4553 + }, + { + "epoch": 0.43805309734513276, + "grad_norm": 1.9406119023696409, + "learning_rate": 1.2466615524911316e-05, + "loss": 0.9673, + "step": 4554 + }, + { + "epoch": 0.43814928818776455, + "grad_norm": 2.041275371797504, + "learning_rate": 1.2463596243477489e-05, + "loss": 0.9661, + "step": 4555 + }, + { + "epoch": 0.4382454790303963, + "grad_norm": 2.1374419158591014, + "learning_rate": 1.2460576722930432e-05, + "loss": 0.9957, + "step": 4556 + }, + { + "epoch": 0.4383416698730281, + "grad_norm": 2.1078487671661805, + "learning_rate": 1.2457556963563222e-05, + "loss": 0.981, + "step": 4557 + }, + { + "epoch": 0.4384378607156599, + "grad_norm": 2.070189823683848, + "learning_rate": 1.2454536965668949e-05, + "loss": 0.9992, + "step": 4558 + }, + { + "epoch": 0.43853405155829167, + "grad_norm": 2.069413720211236, + "learning_rate": 1.2451516729540724e-05, + "loss": 1.0166, + "step": 4559 + }, + { + "epoch": 0.4386302424009234, + "grad_norm": 2.5804485263983, + "learning_rate": 1.24484962554717e-05, + "loss": 0.9746, + "step": 4560 + }, + { + "epoch": 0.4387264332435552, + "grad_norm": 1.9324600390932494, + "learning_rate": 1.2445475543755028e-05, + "loss": 0.9984, + "step": 4561 + }, + { + "epoch": 0.438822624086187, + "grad_norm": 2.0039548415023645, + "learning_rate": 1.24424545946839e-05, + "loss": 1.0646, + "step": 4562 + }, + { + "epoch": 0.4389188149288188, + "grad_norm": 1.8512636821619293, + "learning_rate": 1.2439433408551524e-05, + "loss": 0.9933, + "step": 4563 + }, + { + "epoch": 0.4390150057714506, + "grad_norm": 1.8041760844069352, + "learning_rate": 1.2436411985651131e-05, + "loss": 1.0166, + "step": 4564 + }, + { + "epoch": 0.4391111966140823, + "grad_norm": 1.9045294705757856, + "learning_rate": 1.2433390326275978e-05, + "loss": 0.9222, + "step": 4565 + }, + { + "epoch": 0.4392073874567141, + "grad_norm": 1.968150428524994, + "learning_rate": 1.2430368430719342e-05, + "loss": 1.0329, + "step": 4566 + }, + { + "epoch": 0.4393035782993459, + "grad_norm": 2.003748276487947, + "learning_rate": 1.2427346299274526e-05, + "loss": 0.987, + "step": 4567 + }, + { + "epoch": 0.4393997691419777, + "grad_norm": 2.0102958345835336, + "learning_rate": 1.242432393223485e-05, + "loss": 1.0603, + "step": 4568 + }, + { + "epoch": 0.43949595998460944, + "grad_norm": 1.9811784704122997, + "learning_rate": 1.2421301329893663e-05, + "loss": 0.8687, + "step": 4569 + }, + { + "epoch": 0.43959215082724123, + "grad_norm": 2.155229503144703, + "learning_rate": 1.2418278492544328e-05, + "loss": 0.801, + "step": 4570 + }, + { + "epoch": 0.43968834166987303, + "grad_norm": 2.0904177377926962, + "learning_rate": 1.2415255420480255e-05, + "loss": 1.1319, + "step": 4571 + }, + { + "epoch": 0.4397845325125048, + "grad_norm": 1.995567257744878, + "learning_rate": 1.2412232113994841e-05, + "loss": 0.9672, + "step": 4572 + }, + { + "epoch": 0.4398807233551366, + "grad_norm": 1.8825719883598204, + "learning_rate": 1.240920857338153e-05, + "loss": 0.9936, + "step": 4573 + }, + { + "epoch": 0.43997691419776835, + "grad_norm": 1.9187469491193538, + "learning_rate": 1.2406184798933786e-05, + "loss": 0.9628, + "step": 4574 + }, + { + "epoch": 0.44007310504040015, + "grad_norm": 2.1816165560797702, + "learning_rate": 1.2403160790945088e-05, + "loss": 1.099, + "step": 4575 + }, + { + "epoch": 0.44016929588303194, + "grad_norm": 1.9097500376520704, + "learning_rate": 1.2400136549708945e-05, + "loss": 1.0183, + "step": 4576 + }, + { + "epoch": 0.44026548672566373, + "grad_norm": 1.9273177565020458, + "learning_rate": 1.2397112075518884e-05, + "loss": 0.9926, + "step": 4577 + }, + { + "epoch": 0.4403616775682955, + "grad_norm": 2.048922569717011, + "learning_rate": 1.239408736866846e-05, + "loss": 0.9026, + "step": 4578 + }, + { + "epoch": 0.44045786841092727, + "grad_norm": 2.080117371862769, + "learning_rate": 1.2391062429451239e-05, + "loss": 0.9984, + "step": 4579 + }, + { + "epoch": 0.44055405925355906, + "grad_norm": 1.7518887493121467, + "learning_rate": 1.2388037258160823e-05, + "loss": 1.0038, + "step": 4580 + }, + { + "epoch": 0.44065025009619085, + "grad_norm": 2.150547158662364, + "learning_rate": 1.2385011855090832e-05, + "loss": 0.9785, + "step": 4581 + }, + { + "epoch": 0.44074644093882265, + "grad_norm": 2.145820683600345, + "learning_rate": 1.23819862205349e-05, + "loss": 0.9875, + "step": 4582 + }, + { + "epoch": 0.4408426317814544, + "grad_norm": 1.9747812239900837, + "learning_rate": 1.2378960354786704e-05, + "loss": 1.0563, + "step": 4583 + }, + { + "epoch": 0.4409388226240862, + "grad_norm": 2.1765492687868817, + "learning_rate": 1.2375934258139917e-05, + "loss": 0.8711, + "step": 4584 + }, + { + "epoch": 0.441035013466718, + "grad_norm": 2.3620299062532446, + "learning_rate": 1.2372907930888254e-05, + "loss": 0.9793, + "step": 4585 + }, + { + "epoch": 0.44113120430934977, + "grad_norm": 2.0300125920671976, + "learning_rate": 1.2369881373325448e-05, + "loss": 0.9728, + "step": 4586 + }, + { + "epoch": 0.4412273951519815, + "grad_norm": 2.2618694747401187, + "learning_rate": 1.2366854585745244e-05, + "loss": 1.0811, + "step": 4587 + }, + { + "epoch": 0.4413235859946133, + "grad_norm": 2.0642082449293744, + "learning_rate": 1.236382756844143e-05, + "loss": 0.9713, + "step": 4588 + }, + { + "epoch": 0.4414197768372451, + "grad_norm": 2.1260900322212053, + "learning_rate": 1.2360800321707793e-05, + "loss": 1.0058, + "step": 4589 + }, + { + "epoch": 0.4415159676798769, + "grad_norm": 2.0817341899540494, + "learning_rate": 1.2357772845838159e-05, + "loss": 1.0475, + "step": 4590 + }, + { + "epoch": 0.4416121585225087, + "grad_norm": 1.9782877068833988, + "learning_rate": 1.2354745141126372e-05, + "loss": 1.0687, + "step": 4591 + }, + { + "epoch": 0.4417083493651404, + "grad_norm": 2.0921016262118903, + "learning_rate": 1.2351717207866292e-05, + "loss": 1.1335, + "step": 4592 + }, + { + "epoch": 0.4418045402077722, + "grad_norm": 1.9631337770103094, + "learning_rate": 1.2348689046351809e-05, + "loss": 0.9898, + "step": 4593 + }, + { + "epoch": 0.441900731050404, + "grad_norm": 2.112654085549621, + "learning_rate": 1.2345660656876832e-05, + "loss": 0.997, + "step": 4594 + }, + { + "epoch": 0.4419969218930358, + "grad_norm": 1.8674097118988588, + "learning_rate": 1.2342632039735292e-05, + "loss": 0.9664, + "step": 4595 + }, + { + "epoch": 0.44209311273566754, + "grad_norm": 2.2295764615445317, + "learning_rate": 1.233960319522114e-05, + "loss": 0.9361, + "step": 4596 + }, + { + "epoch": 0.44218930357829933, + "grad_norm": 2.0769335776281475, + "learning_rate": 1.2336574123628357e-05, + "loss": 0.8903, + "step": 4597 + }, + { + "epoch": 0.4422854944209311, + "grad_norm": 2.1052749442796936, + "learning_rate": 1.2333544825250938e-05, + "loss": 1.0915, + "step": 4598 + }, + { + "epoch": 0.4423816852635629, + "grad_norm": 1.8725096657499205, + "learning_rate": 1.2330515300382897e-05, + "loss": 1.0333, + "step": 4599 + }, + { + "epoch": 0.4424778761061947, + "grad_norm": 2.0661638817211463, + "learning_rate": 1.2327485549318285e-05, + "loss": 0.9847, + "step": 4600 + }, + { + "epoch": 0.44257406694882645, + "grad_norm": 2.117340323582186, + "learning_rate": 1.2324455572351154e-05, + "loss": 1.0763, + "step": 4601 + }, + { + "epoch": 0.44267025779145824, + "grad_norm": 2.3403155448119444, + "learning_rate": 1.2321425369775601e-05, + "loss": 1.0948, + "step": 4602 + }, + { + "epoch": 0.44276644863409004, + "grad_norm": 1.842888338042567, + "learning_rate": 1.2318394941885727e-05, + "loss": 0.8964, + "step": 4603 + }, + { + "epoch": 0.44286263947672183, + "grad_norm": 2.137555479258762, + "learning_rate": 1.2315364288975665e-05, + "loss": 1.0239, + "step": 4604 + }, + { + "epoch": 0.4429588303193536, + "grad_norm": 2.0107651470952974, + "learning_rate": 1.231233341133956e-05, + "loss": 0.9613, + "step": 4605 + }, + { + "epoch": 0.44305502116198536, + "grad_norm": 1.9288296025179577, + "learning_rate": 1.2309302309271587e-05, + "loss": 1.0195, + "step": 4606 + }, + { + "epoch": 0.44315121200461716, + "grad_norm": 1.9414862555141, + "learning_rate": 1.2306270983065945e-05, + "loss": 1.0482, + "step": 4607 + }, + { + "epoch": 0.44324740284724895, + "grad_norm": 2.0097951032363217, + "learning_rate": 1.2303239433016842e-05, + "loss": 0.994, + "step": 4608 + }, + { + "epoch": 0.44334359368988074, + "grad_norm": 1.9861059152961682, + "learning_rate": 1.2300207659418526e-05, + "loss": 1.0625, + "step": 4609 + }, + { + "epoch": 0.4434397845325125, + "grad_norm": 2.053817816831179, + "learning_rate": 1.2297175662565248e-05, + "loss": 1.1753, + "step": 4610 + }, + { + "epoch": 0.4435359753751443, + "grad_norm": 2.2604938587309795, + "learning_rate": 1.2294143442751293e-05, + "loss": 0.9479, + "step": 4611 + }, + { + "epoch": 0.44363216621777607, + "grad_norm": 1.9286098584093643, + "learning_rate": 1.229111100027097e-05, + "loss": 0.9068, + "step": 4612 + }, + { + "epoch": 0.44372835706040786, + "grad_norm": 1.8595965707782076, + "learning_rate": 1.228807833541859e-05, + "loss": 1.0168, + "step": 4613 + }, + { + "epoch": 0.44382454790303966, + "grad_norm": 1.8676191592413747, + "learning_rate": 1.228504544848851e-05, + "loss": 1.0028, + "step": 4614 + }, + { + "epoch": 0.4439207387456714, + "grad_norm": 1.8595306246155894, + "learning_rate": 1.2282012339775096e-05, + "loss": 1.0453, + "step": 4615 + }, + { + "epoch": 0.4440169295883032, + "grad_norm": 2.2348156609281693, + "learning_rate": 1.2278979009572736e-05, + "loss": 0.9562, + "step": 4616 + }, + { + "epoch": 0.444113120430935, + "grad_norm": 2.247196260790933, + "learning_rate": 1.2275945458175839e-05, + "loss": 1.0525, + "step": 4617 + }, + { + "epoch": 0.4442093112735668, + "grad_norm": 2.0488372588422643, + "learning_rate": 1.2272911685878841e-05, + "loss": 0.9639, + "step": 4618 + }, + { + "epoch": 0.4443055021161985, + "grad_norm": 2.2946167689485515, + "learning_rate": 1.2269877692976194e-05, + "loss": 1.0096, + "step": 4619 + }, + { + "epoch": 0.4444016929588303, + "grad_norm": 2.082945922361367, + "learning_rate": 1.2266843479762372e-05, + "loss": 1.1067, + "step": 4620 + }, + { + "epoch": 0.4444978838014621, + "grad_norm": 1.8990219929009025, + "learning_rate": 1.226380904653187e-05, + "loss": 1.0304, + "step": 4621 + }, + { + "epoch": 0.4445940746440939, + "grad_norm": 2.261133063640885, + "learning_rate": 1.2260774393579209e-05, + "loss": 0.9344, + "step": 4622 + }, + { + "epoch": 0.4446902654867257, + "grad_norm": 2.073856327437953, + "learning_rate": 1.2257739521198932e-05, + "loss": 0.9447, + "step": 4623 + }, + { + "epoch": 0.4447864563293574, + "grad_norm": 2.0056558076586106, + "learning_rate": 1.2254704429685593e-05, + "loss": 1.0573, + "step": 4624 + }, + { + "epoch": 0.4448826471719892, + "grad_norm": 2.3181054940332055, + "learning_rate": 1.2251669119333773e-05, + "loss": 1.0963, + "step": 4625 + }, + { + "epoch": 0.444978838014621, + "grad_norm": 2.1165642973718692, + "learning_rate": 1.2248633590438084e-05, + "loss": 1.0557, + "step": 4626 + }, + { + "epoch": 0.4450750288572528, + "grad_norm": 2.051856976578165, + "learning_rate": 1.2245597843293138e-05, + "loss": 0.9002, + "step": 4627 + }, + { + "epoch": 0.44517121969988455, + "grad_norm": 1.8382644489689162, + "learning_rate": 1.2242561878193589e-05, + "loss": 0.8696, + "step": 4628 + }, + { + "epoch": 0.44526741054251634, + "grad_norm": 2.11133433153012, + "learning_rate": 1.2239525695434101e-05, + "loss": 1.0237, + "step": 4629 + }, + { + "epoch": 0.44536360138514813, + "grad_norm": 2.0558295052454842, + "learning_rate": 1.2236489295309362e-05, + "loss": 1.12, + "step": 4630 + }, + { + "epoch": 0.4454597922277799, + "grad_norm": 2.6820930951452824, + "learning_rate": 1.2233452678114079e-05, + "loss": 1.0409, + "step": 4631 + }, + { + "epoch": 0.4455559830704117, + "grad_norm": 2.0459259705614015, + "learning_rate": 1.2230415844142984e-05, + "loss": 1.0064, + "step": 4632 + }, + { + "epoch": 0.44565217391304346, + "grad_norm": 2.011431395570863, + "learning_rate": 1.2227378793690828e-05, + "loss": 1.0217, + "step": 4633 + }, + { + "epoch": 0.44574836475567525, + "grad_norm": 2.153952829011952, + "learning_rate": 1.2224341527052378e-05, + "loss": 0.8897, + "step": 4634 + }, + { + "epoch": 0.44584455559830705, + "grad_norm": 2.3048375451740646, + "learning_rate": 1.2221304044522437e-05, + "loss": 1.0008, + "step": 4635 + }, + { + "epoch": 0.44594074644093884, + "grad_norm": 1.8887441113262708, + "learning_rate": 1.2218266346395811e-05, + "loss": 0.9299, + "step": 4636 + }, + { + "epoch": 0.4460369372835706, + "grad_norm": 1.7672229778866846, + "learning_rate": 1.2215228432967335e-05, + "loss": 1.0238, + "step": 4637 + }, + { + "epoch": 0.44613312812620237, + "grad_norm": 1.9006053491663921, + "learning_rate": 1.221219030453187e-05, + "loss": 0.9491, + "step": 4638 + }, + { + "epoch": 0.44622931896883417, + "grad_norm": 2.133912133254871, + "learning_rate": 1.2209151961384285e-05, + "loss": 1.0047, + "step": 4639 + }, + { + "epoch": 0.44632550981146596, + "grad_norm": 2.0762714361124956, + "learning_rate": 1.220611340381948e-05, + "loss": 0.9686, + "step": 4640 + }, + { + "epoch": 0.44642170065409775, + "grad_norm": 1.9195628902644182, + "learning_rate": 1.2203074632132376e-05, + "loss": 1.042, + "step": 4641 + }, + { + "epoch": 0.4465178914967295, + "grad_norm": 2.3116691353618095, + "learning_rate": 1.2200035646617912e-05, + "loss": 1.1567, + "step": 4642 + }, + { + "epoch": 0.4466140823393613, + "grad_norm": 1.633877588971109, + "learning_rate": 1.2196996447571045e-05, + "loss": 0.8517, + "step": 4643 + }, + { + "epoch": 0.4467102731819931, + "grad_norm": 2.293871921978619, + "learning_rate": 1.2193957035286757e-05, + "loss": 1.0661, + "step": 4644 + }, + { + "epoch": 0.44680646402462487, + "grad_norm": 1.9482925799210742, + "learning_rate": 1.2190917410060046e-05, + "loss": 0.9563, + "step": 4645 + }, + { + "epoch": 0.4469026548672566, + "grad_norm": 1.969806060640816, + "learning_rate": 1.2187877572185937e-05, + "loss": 0.9001, + "step": 4646 + }, + { + "epoch": 0.4469988457098884, + "grad_norm": 2.1536458707256902, + "learning_rate": 1.2184837521959475e-05, + "loss": 0.9402, + "step": 4647 + }, + { + "epoch": 0.4470950365525202, + "grad_norm": 1.923587106620294, + "learning_rate": 1.2181797259675713e-05, + "loss": 0.9916, + "step": 4648 + }, + { + "epoch": 0.447191227395152, + "grad_norm": 1.8157864392594487, + "learning_rate": 1.2178756785629749e-05, + "loss": 0.8472, + "step": 4649 + }, + { + "epoch": 0.4472874182377838, + "grad_norm": 2.163395086758489, + "learning_rate": 1.2175716100116677e-05, + "loss": 0.9388, + "step": 4650 + }, + { + "epoch": 0.4473836090804155, + "grad_norm": 2.216227134550447, + "learning_rate": 1.2172675203431621e-05, + "loss": 1.0802, + "step": 4651 + }, + { + "epoch": 0.4474797999230473, + "grad_norm": 1.9423305395313086, + "learning_rate": 1.2169634095869736e-05, + "loss": 0.956, + "step": 4652 + }, + { + "epoch": 0.4475759907656791, + "grad_norm": 2.0671081785456886, + "learning_rate": 1.2166592777726173e-05, + "loss": 1.0094, + "step": 4653 + }, + { + "epoch": 0.4476721816083109, + "grad_norm": 2.018681637343893, + "learning_rate": 1.2163551249296132e-05, + "loss": 1.016, + "step": 4654 + }, + { + "epoch": 0.4477683724509427, + "grad_norm": 1.9645026359422393, + "learning_rate": 1.2160509510874815e-05, + "loss": 1.0193, + "step": 4655 + }, + { + "epoch": 0.44786456329357444, + "grad_norm": 2.011676926538131, + "learning_rate": 1.2157467562757443e-05, + "loss": 1.0446, + "step": 4656 + }, + { + "epoch": 0.44796075413620623, + "grad_norm": 2.0628183471737644, + "learning_rate": 1.2154425405239271e-05, + "loss": 1.0298, + "step": 4657 + }, + { + "epoch": 0.448056944978838, + "grad_norm": 2.123724722334256, + "learning_rate": 1.2151383038615563e-05, + "loss": 1.0337, + "step": 4658 + }, + { + "epoch": 0.4481531358214698, + "grad_norm": 2.470764742653538, + "learning_rate": 1.2148340463181607e-05, + "loss": 0.9462, + "step": 4659 + }, + { + "epoch": 0.44824932666410156, + "grad_norm": 1.8586633926225442, + "learning_rate": 1.214529767923271e-05, + "loss": 1.0152, + "step": 4660 + }, + { + "epoch": 0.44834551750673335, + "grad_norm": 1.8744363994296356, + "learning_rate": 1.2142254687064207e-05, + "loss": 1.0303, + "step": 4661 + }, + { + "epoch": 0.44844170834936514, + "grad_norm": 1.8682556078445272, + "learning_rate": 1.2139211486971436e-05, + "loss": 0.9618, + "step": 4662 + }, + { + "epoch": 0.44853789919199694, + "grad_norm": 1.9141338998839827, + "learning_rate": 1.2136168079249775e-05, + "loss": 1.0473, + "step": 4663 + }, + { + "epoch": 0.44863409003462873, + "grad_norm": 2.1818016042722084, + "learning_rate": 1.213312446419461e-05, + "loss": 1.1016, + "step": 4664 + }, + { + "epoch": 0.44873028087726047, + "grad_norm": 1.7523006467847344, + "learning_rate": 1.2130080642101345e-05, + "loss": 0.9505, + "step": 4665 + }, + { + "epoch": 0.44882647171989226, + "grad_norm": 1.9788495955049614, + "learning_rate": 1.2127036613265418e-05, + "loss": 0.9759, + "step": 4666 + }, + { + "epoch": 0.44892266256252406, + "grad_norm": 2.012012511834009, + "learning_rate": 1.2123992377982272e-05, + "loss": 0.9956, + "step": 4667 + }, + { + "epoch": 0.44901885340515585, + "grad_norm": 2.2190391898758133, + "learning_rate": 1.2120947936547375e-05, + "loss": 1.0637, + "step": 4668 + }, + { + "epoch": 0.4491150442477876, + "grad_norm": 1.9314552169248973, + "learning_rate": 1.2117903289256223e-05, + "loss": 0.9733, + "step": 4669 + }, + { + "epoch": 0.4492112350904194, + "grad_norm": 1.9359148067824592, + "learning_rate": 1.2114858436404322e-05, + "loss": 0.9779, + "step": 4670 + }, + { + "epoch": 0.4493074259330512, + "grad_norm": 1.9464303312994655, + "learning_rate": 1.2111813378287198e-05, + "loss": 0.9642, + "step": 4671 + }, + { + "epoch": 0.44940361677568297, + "grad_norm": 1.7080041542960478, + "learning_rate": 1.2108768115200405e-05, + "loss": 0.839, + "step": 4672 + }, + { + "epoch": 0.44949980761831476, + "grad_norm": 2.178264699708914, + "learning_rate": 1.2105722647439506e-05, + "loss": 0.9217, + "step": 4673 + }, + { + "epoch": 0.4495959984609465, + "grad_norm": 2.1512605342951194, + "learning_rate": 1.2102676975300095e-05, + "loss": 0.992, + "step": 4674 + }, + { + "epoch": 0.4496921893035783, + "grad_norm": 1.9321335613709707, + "learning_rate": 1.2099631099077783e-05, + "loss": 1.014, + "step": 4675 + }, + { + "epoch": 0.4497883801462101, + "grad_norm": 2.61048396892342, + "learning_rate": 1.209658501906819e-05, + "loss": 0.9262, + "step": 4676 + }, + { + "epoch": 0.4498845709888419, + "grad_norm": 2.142469258423318, + "learning_rate": 1.209353873556697e-05, + "loss": 1.0543, + "step": 4677 + }, + { + "epoch": 0.4499807618314736, + "grad_norm": 1.9446074414628494, + "learning_rate": 1.2090492248869795e-05, + "loss": 1.0147, + "step": 4678 + }, + { + "epoch": 0.4500769526741054, + "grad_norm": 1.9447246755914749, + "learning_rate": 1.208744555927234e-05, + "loss": 0.9035, + "step": 4679 + }, + { + "epoch": 0.4501731435167372, + "grad_norm": 1.8396792061570009, + "learning_rate": 1.2084398667070325e-05, + "loss": 0.9547, + "step": 4680 + }, + { + "epoch": 0.450269334359369, + "grad_norm": 2.018507004298174, + "learning_rate": 1.2081351572559471e-05, + "loss": 0.9678, + "step": 4681 + }, + { + "epoch": 0.4503655252020008, + "grad_norm": 2.326262757331146, + "learning_rate": 1.2078304276035527e-05, + "loss": 1.047, + "step": 4682 + }, + { + "epoch": 0.45046171604463253, + "grad_norm": 2.2364603546148443, + "learning_rate": 1.2075256777794259e-05, + "loss": 1.0315, + "step": 4683 + }, + { + "epoch": 0.4505579068872643, + "grad_norm": 1.8305345385336436, + "learning_rate": 1.2072209078131451e-05, + "loss": 0.8646, + "step": 4684 + }, + { + "epoch": 0.4506540977298961, + "grad_norm": 2.0684741309709644, + "learning_rate": 1.2069161177342912e-05, + "loss": 1.0617, + "step": 4685 + }, + { + "epoch": 0.4507502885725279, + "grad_norm": 1.9367388187034422, + "learning_rate": 1.2066113075724461e-05, + "loss": 0.9243, + "step": 4686 + }, + { + "epoch": 0.45084647941515965, + "grad_norm": 2.170761492539027, + "learning_rate": 1.2063064773571953e-05, + "loss": 1.0011, + "step": 4687 + }, + { + "epoch": 0.45094267025779144, + "grad_norm": 1.9117483355537686, + "learning_rate": 1.206001627118124e-05, + "loss": 0.9493, + "step": 4688 + }, + { + "epoch": 0.45103886110042324, + "grad_norm": 2.1292381209466518, + "learning_rate": 1.2056967568848215e-05, + "loss": 1.0157, + "step": 4689 + }, + { + "epoch": 0.45113505194305503, + "grad_norm": 2.10169007112341, + "learning_rate": 1.2053918666868776e-05, + "loss": 0.967, + "step": 4690 + }, + { + "epoch": 0.4512312427856868, + "grad_norm": 2.4969517778924724, + "learning_rate": 1.2050869565538842e-05, + "loss": 1.0234, + "step": 4691 + }, + { + "epoch": 0.45132743362831856, + "grad_norm": 2.0583807665482463, + "learning_rate": 1.2047820265154362e-05, + "loss": 0.9961, + "step": 4692 + }, + { + "epoch": 0.45142362447095036, + "grad_norm": 1.99777957865735, + "learning_rate": 1.2044770766011293e-05, + "loss": 1.0111, + "step": 4693 + }, + { + "epoch": 0.45151981531358215, + "grad_norm": 2.0331261149513806, + "learning_rate": 1.2041721068405614e-05, + "loss": 0.9993, + "step": 4694 + }, + { + "epoch": 0.45161600615621394, + "grad_norm": 2.143058014488448, + "learning_rate": 1.2038671172633327e-05, + "loss": 1.0348, + "step": 4695 + }, + { + "epoch": 0.4517121969988457, + "grad_norm": 2.1911867858832625, + "learning_rate": 1.203562107899045e-05, + "loss": 1.1332, + "step": 4696 + }, + { + "epoch": 0.4518083878414775, + "grad_norm": 2.5733697488420333, + "learning_rate": 1.203257078777302e-05, + "loss": 1.0836, + "step": 4697 + }, + { + "epoch": 0.45190457868410927, + "grad_norm": 1.8253141957550136, + "learning_rate": 1.2029520299277095e-05, + "loss": 0.9639, + "step": 4698 + }, + { + "epoch": 0.45200076952674106, + "grad_norm": 1.682669672846911, + "learning_rate": 1.202646961379875e-05, + "loss": 0.8565, + "step": 4699 + }, + { + "epoch": 0.45209696036937286, + "grad_norm": 2.0855150002660703, + "learning_rate": 1.2023418731634078e-05, + "loss": 1.0386, + "step": 4700 + }, + { + "epoch": 0.4521931512120046, + "grad_norm": 1.9475241445545448, + "learning_rate": 1.2020367653079201e-05, + "loss": 0.9575, + "step": 4701 + }, + { + "epoch": 0.4522893420546364, + "grad_norm": 1.9951069547067344, + "learning_rate": 1.2017316378430244e-05, + "loss": 0.9553, + "step": 4702 + }, + { + "epoch": 0.4523855328972682, + "grad_norm": 2.0062252161251206, + "learning_rate": 1.2014264907983364e-05, + "loss": 0.9551, + "step": 4703 + }, + { + "epoch": 0.4524817237399, + "grad_norm": 2.060293902640883, + "learning_rate": 1.2011213242034733e-05, + "loss": 0.9644, + "step": 4704 + }, + { + "epoch": 0.45257791458253177, + "grad_norm": 2.0176782433979406, + "learning_rate": 1.2008161380880538e-05, + "loss": 0.9332, + "step": 4705 + }, + { + "epoch": 0.4526741054251635, + "grad_norm": 1.9315659718128864, + "learning_rate": 1.2005109324816992e-05, + "loss": 0.9775, + "step": 4706 + }, + { + "epoch": 0.4527702962677953, + "grad_norm": 1.7396178546996004, + "learning_rate": 1.2002057074140323e-05, + "loss": 0.8978, + "step": 4707 + }, + { + "epoch": 0.4528664871104271, + "grad_norm": 2.043243324193625, + "learning_rate": 1.1999004629146775e-05, + "loss": 1.0037, + "step": 4708 + }, + { + "epoch": 0.4529626779530589, + "grad_norm": 1.9563729859783434, + "learning_rate": 1.1995951990132617e-05, + "loss": 1.1002, + "step": 4709 + }, + { + "epoch": 0.45305886879569063, + "grad_norm": 2.020381798004692, + "learning_rate": 1.1992899157394133e-05, + "loss": 1.0365, + "step": 4710 + }, + { + "epoch": 0.4531550596383224, + "grad_norm": 1.8659064699263768, + "learning_rate": 1.1989846131227627e-05, + "loss": 0.8845, + "step": 4711 + }, + { + "epoch": 0.4532512504809542, + "grad_norm": 2.4640529275422622, + "learning_rate": 1.1986792911929418e-05, + "loss": 1.0419, + "step": 4712 + }, + { + "epoch": 0.453347441323586, + "grad_norm": 2.0957602818966805, + "learning_rate": 1.1983739499795856e-05, + "loss": 1.0235, + "step": 4713 + }, + { + "epoch": 0.4534436321662178, + "grad_norm": 2.485077820985735, + "learning_rate": 1.198068589512329e-05, + "loss": 1.0524, + "step": 4714 + }, + { + "epoch": 0.45353982300884954, + "grad_norm": 2.0546114495626915, + "learning_rate": 1.1977632098208108e-05, + "loss": 0.9336, + "step": 4715 + }, + { + "epoch": 0.45363601385148133, + "grad_norm": 2.178818397622712, + "learning_rate": 1.1974578109346702e-05, + "loss": 0.9061, + "step": 4716 + }, + { + "epoch": 0.45373220469411313, + "grad_norm": 2.1229850593555577, + "learning_rate": 1.197152392883549e-05, + "loss": 1.0739, + "step": 4717 + }, + { + "epoch": 0.4538283955367449, + "grad_norm": 2.310092255130532, + "learning_rate": 1.1968469556970905e-05, + "loss": 0.9987, + "step": 4718 + }, + { + "epoch": 0.45392458637937666, + "grad_norm": 2.3556416522095787, + "learning_rate": 1.1965414994049399e-05, + "loss": 0.9466, + "step": 4719 + }, + { + "epoch": 0.45402077722200845, + "grad_norm": 1.7946012729282355, + "learning_rate": 1.1962360240367445e-05, + "loss": 0.8562, + "step": 4720 + }, + { + "epoch": 0.45411696806464025, + "grad_norm": 2.03952544890581, + "learning_rate": 1.1959305296221534e-05, + "loss": 0.9749, + "step": 4721 + }, + { + "epoch": 0.45421315890727204, + "grad_norm": 1.9770225840154483, + "learning_rate": 1.1956250161908179e-05, + "loss": 0.9714, + "step": 4722 + }, + { + "epoch": 0.45430934974990383, + "grad_norm": 1.9923363202294118, + "learning_rate": 1.1953194837723899e-05, + "loss": 1.0562, + "step": 4723 + }, + { + "epoch": 0.4544055405925356, + "grad_norm": 1.9952574757696746, + "learning_rate": 1.195013932396524e-05, + "loss": 1.0259, + "step": 4724 + }, + { + "epoch": 0.45450173143516737, + "grad_norm": 2.1812887364778226, + "learning_rate": 1.1947083620928772e-05, + "loss": 1.0626, + "step": 4725 + }, + { + "epoch": 0.45459792227779916, + "grad_norm": 2.1277882062193774, + "learning_rate": 1.1944027728911072e-05, + "loss": 1.0457, + "step": 4726 + }, + { + "epoch": 0.45469411312043095, + "grad_norm": 1.9684492319005407, + "learning_rate": 1.1940971648208746e-05, + "loss": 0.9403, + "step": 4727 + }, + { + "epoch": 0.4547903039630627, + "grad_norm": 1.9029238239975002, + "learning_rate": 1.1937915379118406e-05, + "loss": 0.9576, + "step": 4728 + }, + { + "epoch": 0.4548864948056945, + "grad_norm": 2.345009787947025, + "learning_rate": 1.1934858921936692e-05, + "loss": 0.8985, + "step": 4729 + }, + { + "epoch": 0.4549826856483263, + "grad_norm": 2.2049675186805615, + "learning_rate": 1.1931802276960265e-05, + "loss": 1.0411, + "step": 4730 + }, + { + "epoch": 0.4550788764909581, + "grad_norm": 2.113908310256375, + "learning_rate": 1.192874544448579e-05, + "loss": 1.0225, + "step": 4731 + }, + { + "epoch": 0.45517506733358987, + "grad_norm": 2.017024164083717, + "learning_rate": 1.1925688424809965e-05, + "loss": 0.9824, + "step": 4732 + }, + { + "epoch": 0.4552712581762216, + "grad_norm": 1.9876163998207512, + "learning_rate": 1.1922631218229497e-05, + "loss": 0.8999, + "step": 4733 + }, + { + "epoch": 0.4553674490188534, + "grad_norm": 2.139250408693713, + "learning_rate": 1.1919573825041115e-05, + "loss": 0.9805, + "step": 4734 + }, + { + "epoch": 0.4554636398614852, + "grad_norm": 2.0608970361835186, + "learning_rate": 1.1916516245541567e-05, + "loss": 1.0549, + "step": 4735 + }, + { + "epoch": 0.455559830704117, + "grad_norm": 2.2016801458663853, + "learning_rate": 1.1913458480027614e-05, + "loss": 0.9611, + "step": 4736 + }, + { + "epoch": 0.4556560215467487, + "grad_norm": 1.9318006622926538, + "learning_rate": 1.1910400528796042e-05, + "loss": 1.1174, + "step": 4737 + }, + { + "epoch": 0.4557522123893805, + "grad_norm": 2.085939483323767, + "learning_rate": 1.1907342392143646e-05, + "loss": 1.0034, + "step": 4738 + }, + { + "epoch": 0.4558484032320123, + "grad_norm": 2.1709919061698755, + "learning_rate": 1.1904284070367255e-05, + "loss": 1.0294, + "step": 4739 + }, + { + "epoch": 0.4559445940746441, + "grad_norm": 1.8676781193078256, + "learning_rate": 1.1901225563763694e-05, + "loss": 1.0125, + "step": 4740 + }, + { + "epoch": 0.4560407849172759, + "grad_norm": 1.9966979136854863, + "learning_rate": 1.1898166872629823e-05, + "loss": 1.0319, + "step": 4741 + }, + { + "epoch": 0.45613697575990764, + "grad_norm": 2.1161859167267565, + "learning_rate": 1.1895107997262516e-05, + "loss": 1.0264, + "step": 4742 + }, + { + "epoch": 0.45623316660253943, + "grad_norm": 1.8179931222698353, + "learning_rate": 1.1892048937958658e-05, + "loss": 0.9878, + "step": 4743 + }, + { + "epoch": 0.4563293574451712, + "grad_norm": 1.9747382275476728, + "learning_rate": 1.1888989695015166e-05, + "loss": 0.9283, + "step": 4744 + }, + { + "epoch": 0.456425548287803, + "grad_norm": 1.8150510469436558, + "learning_rate": 1.188593026872895e-05, + "loss": 0.9715, + "step": 4745 + }, + { + "epoch": 0.45652173913043476, + "grad_norm": 1.9624538637155555, + "learning_rate": 1.1882870659396968e-05, + "loss": 0.9646, + "step": 4746 + }, + { + "epoch": 0.45661792997306655, + "grad_norm": 2.162818255224631, + "learning_rate": 1.1879810867316178e-05, + "loss": 1.0391, + "step": 4747 + }, + { + "epoch": 0.45671412081569834, + "grad_norm": 2.2521337789771536, + "learning_rate": 1.1876750892783558e-05, + "loss": 0.9743, + "step": 4748 + }, + { + "epoch": 0.45681031165833014, + "grad_norm": 2.2943467509043063, + "learning_rate": 1.1873690736096104e-05, + "loss": 0.9418, + "step": 4749 + }, + { + "epoch": 0.45690650250096193, + "grad_norm": 2.2498982186426604, + "learning_rate": 1.1870630397550831e-05, + "loss": 1.0977, + "step": 4750 + }, + { + "epoch": 0.45700269334359367, + "grad_norm": 2.0900746857523465, + "learning_rate": 1.186756987744477e-05, + "loss": 1.0451, + "step": 4751 + }, + { + "epoch": 0.45709888418622546, + "grad_norm": 2.2066706254724617, + "learning_rate": 1.1864509176074974e-05, + "loss": 0.9364, + "step": 4752 + }, + { + "epoch": 0.45719507502885726, + "grad_norm": 2.0657032144569953, + "learning_rate": 1.1861448293738511e-05, + "loss": 0.9441, + "step": 4753 + }, + { + "epoch": 0.45729126587148905, + "grad_norm": 2.007432452539004, + "learning_rate": 1.185838723073246e-05, + "loss": 0.8557, + "step": 4754 + }, + { + "epoch": 0.4573874567141208, + "grad_norm": 2.0711662666103914, + "learning_rate": 1.1855325987353928e-05, + "loss": 0.9197, + "step": 4755 + }, + { + "epoch": 0.4574836475567526, + "grad_norm": 2.1213731212523714, + "learning_rate": 1.1852264563900038e-05, + "loss": 0.9608, + "step": 4756 + }, + { + "epoch": 0.4575798383993844, + "grad_norm": 2.047141168348332, + "learning_rate": 1.184920296066792e-05, + "loss": 1.0874, + "step": 4757 + }, + { + "epoch": 0.45767602924201617, + "grad_norm": 1.7846079177595064, + "learning_rate": 1.1846141177954733e-05, + "loss": 1.0502, + "step": 4758 + }, + { + "epoch": 0.45777222008464796, + "grad_norm": 2.120621424892378, + "learning_rate": 1.1843079216057652e-05, + "loss": 1.0521, + "step": 4759 + }, + { + "epoch": 0.4578684109272797, + "grad_norm": 2.0127579719688895, + "learning_rate": 1.1840017075273861e-05, + "loss": 0.9271, + "step": 4760 + }, + { + "epoch": 0.4579646017699115, + "grad_norm": 1.9283500911151834, + "learning_rate": 1.1836954755900572e-05, + "loss": 0.9826, + "step": 4761 + }, + { + "epoch": 0.4580607926125433, + "grad_norm": 2.179336521189267, + "learning_rate": 1.1833892258235008e-05, + "loss": 1.1189, + "step": 4762 + }, + { + "epoch": 0.4581569834551751, + "grad_norm": 2.07803930424581, + "learning_rate": 1.1830829582574407e-05, + "loss": 0.9885, + "step": 4763 + }, + { + "epoch": 0.4582531742978069, + "grad_norm": 2.163317110376232, + "learning_rate": 1.1827766729216035e-05, + "loss": 0.9848, + "step": 4764 + }, + { + "epoch": 0.4583493651404386, + "grad_norm": 1.9274798532253747, + "learning_rate": 1.1824703698457165e-05, + "loss": 1.0064, + "step": 4765 + }, + { + "epoch": 0.4584455559830704, + "grad_norm": 2.0023912797213277, + "learning_rate": 1.1821640490595086e-05, + "loss": 1.0894, + "step": 4766 + }, + { + "epoch": 0.4585417468257022, + "grad_norm": 2.0396222257609122, + "learning_rate": 1.1818577105927116e-05, + "loss": 0.9345, + "step": 4767 + }, + { + "epoch": 0.458637937668334, + "grad_norm": 2.09141277202505, + "learning_rate": 1.181551354475058e-05, + "loss": 1.0138, + "step": 4768 + }, + { + "epoch": 0.45873412851096573, + "grad_norm": 1.8410931620575766, + "learning_rate": 1.181244980736282e-05, + "loss": 0.9163, + "step": 4769 + }, + { + "epoch": 0.4588303193535975, + "grad_norm": 1.8770372260962311, + "learning_rate": 1.1809385894061206e-05, + "loss": 1.0821, + "step": 4770 + }, + { + "epoch": 0.4589265101962293, + "grad_norm": 3.255227449633506, + "learning_rate": 1.1806321805143106e-05, + "loss": 0.9405, + "step": 4771 + }, + { + "epoch": 0.4590227010388611, + "grad_norm": 2.347287448325613, + "learning_rate": 1.1803257540905926e-05, + "loss": 1.0009, + "step": 4772 + }, + { + "epoch": 0.4591188918814929, + "grad_norm": 2.158147074268914, + "learning_rate": 1.1800193101647074e-05, + "loss": 1.1063, + "step": 4773 + }, + { + "epoch": 0.45921508272412465, + "grad_norm": 1.993125026612633, + "learning_rate": 1.1797128487663982e-05, + "loss": 1.0359, + "step": 4774 + }, + { + "epoch": 0.45931127356675644, + "grad_norm": 1.7485116958724645, + "learning_rate": 1.1794063699254097e-05, + "loss": 0.9045, + "step": 4775 + }, + { + "epoch": 0.45940746440938823, + "grad_norm": 1.7625401462078572, + "learning_rate": 1.1790998736714882e-05, + "loss": 0.9852, + "step": 4776 + }, + { + "epoch": 0.45950365525202, + "grad_norm": 1.8627548202639002, + "learning_rate": 1.178793360034382e-05, + "loss": 0.9137, + "step": 4777 + }, + { + "epoch": 0.45959984609465176, + "grad_norm": 1.9132418073915158, + "learning_rate": 1.1784868290438404e-05, + "loss": 1.0027, + "step": 4778 + }, + { + "epoch": 0.45969603693728356, + "grad_norm": 2.157234115494852, + "learning_rate": 1.1781802807296158e-05, + "loss": 1.0419, + "step": 4779 + }, + { + "epoch": 0.45979222777991535, + "grad_norm": 2.017829530022445, + "learning_rate": 1.1778737151214606e-05, + "loss": 1.0669, + "step": 4780 + }, + { + "epoch": 0.45988841862254715, + "grad_norm": 2.098294893573202, + "learning_rate": 1.1775671322491298e-05, + "loss": 1.0169, + "step": 4781 + }, + { + "epoch": 0.45998460946517894, + "grad_norm": 3.349756218913346, + "learning_rate": 1.17726053214238e-05, + "loss": 0.9433, + "step": 4782 + }, + { + "epoch": 0.4600808003078107, + "grad_norm": 2.2842088403071963, + "learning_rate": 1.1769539148309692e-05, + "loss": 1.1876, + "step": 4783 + }, + { + "epoch": 0.46017699115044247, + "grad_norm": 1.9068292838853806, + "learning_rate": 1.1766472803446577e-05, + "loss": 1.0015, + "step": 4784 + }, + { + "epoch": 0.46027318199307427, + "grad_norm": 2.0971526216361025, + "learning_rate": 1.1763406287132061e-05, + "loss": 1.0821, + "step": 4785 + }, + { + "epoch": 0.46036937283570606, + "grad_norm": 2.0449644192493333, + "learning_rate": 1.1760339599663788e-05, + "loss": 1.0208, + "step": 4786 + }, + { + "epoch": 0.4604655636783378, + "grad_norm": 2.025916671188749, + "learning_rate": 1.1757272741339395e-05, + "loss": 1.0671, + "step": 4787 + }, + { + "epoch": 0.4605617545209696, + "grad_norm": 1.7881114959550741, + "learning_rate": 1.1754205712456556e-05, + "loss": 0.9423, + "step": 4788 + }, + { + "epoch": 0.4606579453636014, + "grad_norm": 1.9661052979324165, + "learning_rate": 1.1751138513312948e-05, + "loss": 1.0092, + "step": 4789 + }, + { + "epoch": 0.4607541362062332, + "grad_norm": 1.9233708269454408, + "learning_rate": 1.1748071144206266e-05, + "loss": 1.0027, + "step": 4790 + }, + { + "epoch": 0.46085032704886497, + "grad_norm": 2.133789191236924, + "learning_rate": 1.1745003605434233e-05, + "loss": 0.9475, + "step": 4791 + }, + { + "epoch": 0.4609465178914967, + "grad_norm": 2.1594492416130056, + "learning_rate": 1.1741935897294572e-05, + "loss": 1.021, + "step": 4792 + }, + { + "epoch": 0.4610427087341285, + "grad_norm": 1.7324196126887803, + "learning_rate": 1.173886802008504e-05, + "loss": 0.9713, + "step": 4793 + }, + { + "epoch": 0.4611388995767603, + "grad_norm": 2.0515095142907134, + "learning_rate": 1.1735799974103388e-05, + "loss": 0.9245, + "step": 4794 + }, + { + "epoch": 0.4612350904193921, + "grad_norm": 1.8549236325345773, + "learning_rate": 1.1732731759647404e-05, + "loss": 1.0362, + "step": 4795 + }, + { + "epoch": 0.46133128126202383, + "grad_norm": 1.9192585700816118, + "learning_rate": 1.1729663377014888e-05, + "loss": 0.95, + "step": 4796 + }, + { + "epoch": 0.4614274721046556, + "grad_norm": 1.9568173868506118, + "learning_rate": 1.1726594826503646e-05, + "loss": 0.9857, + "step": 4797 + }, + { + "epoch": 0.4615236629472874, + "grad_norm": 2.2053820510386632, + "learning_rate": 1.172352610841151e-05, + "loss": 1.029, + "step": 4798 + }, + { + "epoch": 0.4616198537899192, + "grad_norm": 2.273338473793562, + "learning_rate": 1.1720457223036323e-05, + "loss": 1.0843, + "step": 4799 + }, + { + "epoch": 0.461716044632551, + "grad_norm": 2.150558166986021, + "learning_rate": 1.1717388170675954e-05, + "loss": 0.9202, + "step": 4800 + }, + { + "epoch": 0.46181223547518274, + "grad_norm": 1.955525854975695, + "learning_rate": 1.1714318951628272e-05, + "loss": 1.0534, + "step": 4801 + }, + { + "epoch": 0.46190842631781454, + "grad_norm": 2.2509040297172627, + "learning_rate": 1.1711249566191179e-05, + "loss": 1.0397, + "step": 4802 + }, + { + "epoch": 0.46200461716044633, + "grad_norm": 2.1640702242955268, + "learning_rate": 1.1708180014662579e-05, + "loss": 0.9185, + "step": 4803 + }, + { + "epoch": 0.4621008080030781, + "grad_norm": 2.0314272396697843, + "learning_rate": 1.17051102973404e-05, + "loss": 1.0712, + "step": 4804 + }, + { + "epoch": 0.46219699884570986, + "grad_norm": 2.0837161462699667, + "learning_rate": 1.1702040414522588e-05, + "loss": 1.0617, + "step": 4805 + }, + { + "epoch": 0.46229318968834165, + "grad_norm": 1.8466964490995472, + "learning_rate": 1.1698970366507096e-05, + "loss": 1.0578, + "step": 4806 + }, + { + "epoch": 0.46238938053097345, + "grad_norm": 2.042352396905683, + "learning_rate": 1.1695900153591904e-05, + "loss": 0.854, + "step": 4807 + }, + { + "epoch": 0.46248557137360524, + "grad_norm": 1.8876456323755595, + "learning_rate": 1.1692829776074999e-05, + "loss": 0.9586, + "step": 4808 + }, + { + "epoch": 0.46258176221623704, + "grad_norm": 1.940779643405636, + "learning_rate": 1.1689759234254389e-05, + "loss": 0.9165, + "step": 4809 + }, + { + "epoch": 0.4626779530588688, + "grad_norm": 1.9258442967872684, + "learning_rate": 1.1686688528428099e-05, + "loss": 1.0796, + "step": 4810 + }, + { + "epoch": 0.46277414390150057, + "grad_norm": 1.9385137990888281, + "learning_rate": 1.1683617658894162e-05, + "loss": 1.0098, + "step": 4811 + }, + { + "epoch": 0.46287033474413236, + "grad_norm": 2.115169292356426, + "learning_rate": 1.1680546625950635e-05, + "loss": 1.0095, + "step": 4812 + }, + { + "epoch": 0.46296652558676415, + "grad_norm": 1.9026296749334848, + "learning_rate": 1.1677475429895588e-05, + "loss": 0.8981, + "step": 4813 + }, + { + "epoch": 0.46306271642939595, + "grad_norm": 2.0244082985785967, + "learning_rate": 1.167440407102711e-05, + "loss": 1.0095, + "step": 4814 + }, + { + "epoch": 0.4631589072720277, + "grad_norm": 1.91911015094894, + "learning_rate": 1.16713325496433e-05, + "loss": 0.9206, + "step": 4815 + }, + { + "epoch": 0.4632550981146595, + "grad_norm": 2.3703515130089485, + "learning_rate": 1.1668260866042271e-05, + "loss": 0.9872, + "step": 4816 + }, + { + "epoch": 0.4633512889572913, + "grad_norm": 2.1123884065682432, + "learning_rate": 1.1665189020522168e-05, + "loss": 0.869, + "step": 4817 + }, + { + "epoch": 0.46344747979992307, + "grad_norm": 1.9510232316076075, + "learning_rate": 1.1662117013381126e-05, + "loss": 0.9437, + "step": 4818 + }, + { + "epoch": 0.4635436706425548, + "grad_norm": 2.079292094831337, + "learning_rate": 1.1659044844917322e-05, + "loss": 1.0376, + "step": 4819 + }, + { + "epoch": 0.4636398614851866, + "grad_norm": 2.043240254005357, + "learning_rate": 1.1655972515428928e-05, + "loss": 1.0343, + "step": 4820 + }, + { + "epoch": 0.4637360523278184, + "grad_norm": 2.1042521125960483, + "learning_rate": 1.1652900025214145e-05, + "loss": 1.0626, + "step": 4821 + }, + { + "epoch": 0.4638322431704502, + "grad_norm": 2.292082658070195, + "learning_rate": 1.1649827374571182e-05, + "loss": 0.9882, + "step": 4822 + }, + { + "epoch": 0.463928434013082, + "grad_norm": 2.1199231073082783, + "learning_rate": 1.1646754563798268e-05, + "loss": 1.0715, + "step": 4823 + }, + { + "epoch": 0.4640246248557137, + "grad_norm": 1.9241141950668186, + "learning_rate": 1.1643681593193642e-05, + "loss": 0.9051, + "step": 4824 + }, + { + "epoch": 0.4641208156983455, + "grad_norm": 2.1096369366127465, + "learning_rate": 1.1640608463055569e-05, + "loss": 0.9693, + "step": 4825 + }, + { + "epoch": 0.4642170065409773, + "grad_norm": 1.9920422756291691, + "learning_rate": 1.1637535173682318e-05, + "loss": 0.9834, + "step": 4826 + }, + { + "epoch": 0.4643131973836091, + "grad_norm": 2.0581153846551037, + "learning_rate": 1.1634461725372179e-05, + "loss": 1.0542, + "step": 4827 + }, + { + "epoch": 0.46440938822624084, + "grad_norm": 1.9623195889532086, + "learning_rate": 1.1631388118423457e-05, + "loss": 0.9539, + "step": 4828 + }, + { + "epoch": 0.46450557906887263, + "grad_norm": 2.058930592826399, + "learning_rate": 1.1628314353134468e-05, + "loss": 1.0735, + "step": 4829 + }, + { + "epoch": 0.4646017699115044, + "grad_norm": 1.8895403766386174, + "learning_rate": 1.1625240429803553e-05, + "loss": 0.9353, + "step": 4830 + }, + { + "epoch": 0.4646979607541362, + "grad_norm": 2.2085262127567096, + "learning_rate": 1.1622166348729066e-05, + "loss": 0.9011, + "step": 4831 + }, + { + "epoch": 0.464794151596768, + "grad_norm": 1.9559516554339247, + "learning_rate": 1.1619092110209361e-05, + "loss": 0.9885, + "step": 4832 + }, + { + "epoch": 0.46489034243939975, + "grad_norm": 4.960469428846224, + "learning_rate": 1.1616017714542829e-05, + "loss": 0.9385, + "step": 4833 + }, + { + "epoch": 0.46498653328203154, + "grad_norm": 2.058476387283652, + "learning_rate": 1.1612943162027863e-05, + "loss": 1.0206, + "step": 4834 + }, + { + "epoch": 0.46508272412466334, + "grad_norm": 2.131556214541961, + "learning_rate": 1.1609868452962876e-05, + "loss": 0.9533, + "step": 4835 + }, + { + "epoch": 0.46517891496729513, + "grad_norm": 2.01170927763449, + "learning_rate": 1.1606793587646295e-05, + "loss": 0.9179, + "step": 4836 + }, + { + "epoch": 0.46527510580992687, + "grad_norm": 2.0727773813533448, + "learning_rate": 1.160371856637656e-05, + "loss": 1.0025, + "step": 4837 + }, + { + "epoch": 0.46537129665255866, + "grad_norm": 2.1007901753425786, + "learning_rate": 1.160064338945213e-05, + "loss": 0.9518, + "step": 4838 + }, + { + "epoch": 0.46546748749519046, + "grad_norm": 1.8918787106684871, + "learning_rate": 1.1597568057171478e-05, + "loss": 0.9823, + "step": 4839 + }, + { + "epoch": 0.46556367833782225, + "grad_norm": 2.370979731296428, + "learning_rate": 1.1594492569833093e-05, + "loss": 1.0292, + "step": 4840 + }, + { + "epoch": 0.46565986918045404, + "grad_norm": 2.061346986575668, + "learning_rate": 1.1591416927735473e-05, + "loss": 0.9762, + "step": 4841 + }, + { + "epoch": 0.4657560600230858, + "grad_norm": 2.31134666009213, + "learning_rate": 1.1588341131177137e-05, + "loss": 1.0684, + "step": 4842 + }, + { + "epoch": 0.4658522508657176, + "grad_norm": 1.8908057424725708, + "learning_rate": 1.1585265180456622e-05, + "loss": 0.8788, + "step": 4843 + }, + { + "epoch": 0.46594844170834937, + "grad_norm": 1.9349321796395, + "learning_rate": 1.1582189075872467e-05, + "loss": 1.0659, + "step": 4844 + }, + { + "epoch": 0.46604463255098116, + "grad_norm": 1.7224593826283643, + "learning_rate": 1.1579112817723246e-05, + "loss": 0.9203, + "step": 4845 + }, + { + "epoch": 0.4661408233936129, + "grad_norm": 2.1312862026495885, + "learning_rate": 1.1576036406307523e-05, + "loss": 0.9682, + "step": 4846 + }, + { + "epoch": 0.4662370142362447, + "grad_norm": 2.0129755656014647, + "learning_rate": 1.1572959841923899e-05, + "loss": 0.8782, + "step": 4847 + }, + { + "epoch": 0.4663332050788765, + "grad_norm": 2.0377678101011942, + "learning_rate": 1.156988312487098e-05, + "loss": 0.9685, + "step": 4848 + }, + { + "epoch": 0.4664293959215083, + "grad_norm": 2.190222047784055, + "learning_rate": 1.1566806255447387e-05, + "loss": 1.0487, + "step": 4849 + }, + { + "epoch": 0.4665255867641401, + "grad_norm": 1.8842090463887853, + "learning_rate": 1.1563729233951757e-05, + "loss": 0.8818, + "step": 4850 + }, + { + "epoch": 0.4666217776067718, + "grad_norm": 2.156902361132796, + "learning_rate": 1.156065206068274e-05, + "loss": 0.9676, + "step": 4851 + }, + { + "epoch": 0.4667179684494036, + "grad_norm": 1.8582051082084183, + "learning_rate": 1.1557574735939003e-05, + "loss": 0.9544, + "step": 4852 + }, + { + "epoch": 0.4668141592920354, + "grad_norm": 2.023460357879254, + "learning_rate": 1.1554497260019226e-05, + "loss": 0.9743, + "step": 4853 + }, + { + "epoch": 0.4669103501346672, + "grad_norm": 1.7422868850889788, + "learning_rate": 1.1551419633222107e-05, + "loss": 0.9151, + "step": 4854 + }, + { + "epoch": 0.46700654097729893, + "grad_norm": 2.1217757080578785, + "learning_rate": 1.1548341855846355e-05, + "loss": 1.0173, + "step": 4855 + }, + { + "epoch": 0.4671027318199307, + "grad_norm": 2.0854874632407703, + "learning_rate": 1.1545263928190692e-05, + "loss": 0.9689, + "step": 4856 + }, + { + "epoch": 0.4671989226625625, + "grad_norm": 1.8762714469070503, + "learning_rate": 1.1542185850553865e-05, + "loss": 0.9871, + "step": 4857 + }, + { + "epoch": 0.4672951135051943, + "grad_norm": 1.7560442341287041, + "learning_rate": 1.1539107623234618e-05, + "loss": 0.9545, + "step": 4858 + }, + { + "epoch": 0.4673913043478261, + "grad_norm": 2.1153447854199037, + "learning_rate": 1.1536029246531726e-05, + "loss": 0.9603, + "step": 4859 + }, + { + "epoch": 0.46748749519045785, + "grad_norm": 1.8301149286935203, + "learning_rate": 1.153295072074397e-05, + "loss": 1.0102, + "step": 4860 + }, + { + "epoch": 0.46758368603308964, + "grad_norm": 1.692074425744785, + "learning_rate": 1.1529872046170148e-05, + "loss": 0.8482, + "step": 4861 + }, + { + "epoch": 0.46767987687572143, + "grad_norm": 2.1863323385683175, + "learning_rate": 1.1526793223109072e-05, + "loss": 0.982, + "step": 4862 + }, + { + "epoch": 0.46777606771835323, + "grad_norm": 2.0955400177112287, + "learning_rate": 1.1523714251859566e-05, + "loss": 1.0399, + "step": 4863 + }, + { + "epoch": 0.467872258560985, + "grad_norm": 2.072608958568706, + "learning_rate": 1.1520635132720475e-05, + "loss": 0.9586, + "step": 4864 + }, + { + "epoch": 0.46796844940361676, + "grad_norm": 2.3343717592746867, + "learning_rate": 1.1517555865990646e-05, + "loss": 0.8543, + "step": 4865 + }, + { + "epoch": 0.46806464024624855, + "grad_norm": 1.8756529976867502, + "learning_rate": 1.1514476451968961e-05, + "loss": 0.9775, + "step": 4866 + }, + { + "epoch": 0.46816083108888035, + "grad_norm": 2.076575553714538, + "learning_rate": 1.1511396890954294e-05, + "loss": 1.0478, + "step": 4867 + }, + { + "epoch": 0.46825702193151214, + "grad_norm": 1.9425869046893536, + "learning_rate": 1.1508317183245545e-05, + "loss": 0.9873, + "step": 4868 + }, + { + "epoch": 0.4683532127741439, + "grad_norm": 2.2164955236745976, + "learning_rate": 1.1505237329141627e-05, + "loss": 1.0687, + "step": 4869 + }, + { + "epoch": 0.4684494036167757, + "grad_norm": 1.9224712035811198, + "learning_rate": 1.1502157328941466e-05, + "loss": 1.0287, + "step": 4870 + }, + { + "epoch": 0.46854559445940747, + "grad_norm": 2.226283693806502, + "learning_rate": 1.1499077182944002e-05, + "loss": 0.9681, + "step": 4871 + }, + { + "epoch": 0.46864178530203926, + "grad_norm": 2.199811981371519, + "learning_rate": 1.149599689144819e-05, + "loss": 1.0795, + "step": 4872 + }, + { + "epoch": 0.46873797614467105, + "grad_norm": 2.010333926038434, + "learning_rate": 1.1492916454752997e-05, + "loss": 0.9462, + "step": 4873 + }, + { + "epoch": 0.4688341669873028, + "grad_norm": 1.7030188400448583, + "learning_rate": 1.1489835873157414e-05, + "loss": 0.9347, + "step": 4874 + }, + { + "epoch": 0.4689303578299346, + "grad_norm": 2.125147431238822, + "learning_rate": 1.1486755146960427e-05, + "loss": 1.016, + "step": 4875 + }, + { + "epoch": 0.4690265486725664, + "grad_norm": 2.1683709107298315, + "learning_rate": 1.1483674276461053e-05, + "loss": 1.0188, + "step": 4876 + }, + { + "epoch": 0.4691227395151982, + "grad_norm": 2.0740285712229873, + "learning_rate": 1.1480593261958314e-05, + "loss": 0.9073, + "step": 4877 + }, + { + "epoch": 0.4692189303578299, + "grad_norm": 2.4680268846403486, + "learning_rate": 1.1477512103751254e-05, + "loss": 1.0116, + "step": 4878 + }, + { + "epoch": 0.4693151212004617, + "grad_norm": 1.8021450591511536, + "learning_rate": 1.147443080213892e-05, + "loss": 0.9612, + "step": 4879 + }, + { + "epoch": 0.4694113120430935, + "grad_norm": 1.901529478994135, + "learning_rate": 1.1471349357420384e-05, + "loss": 0.9087, + "step": 4880 + }, + { + "epoch": 0.4695075028857253, + "grad_norm": 1.8497405034116206, + "learning_rate": 1.1468267769894725e-05, + "loss": 1.0117, + "step": 4881 + }, + { + "epoch": 0.4696036937283571, + "grad_norm": 2.2130879212073507, + "learning_rate": 1.1465186039861033e-05, + "loss": 1.0406, + "step": 4882 + }, + { + "epoch": 0.4696998845709888, + "grad_norm": 1.8514118352094184, + "learning_rate": 1.1462104167618426e-05, + "loss": 1.0121, + "step": 4883 + }, + { + "epoch": 0.4697960754136206, + "grad_norm": 1.840055373349649, + "learning_rate": 1.1459022153466016e-05, + "loss": 1.0809, + "step": 4884 + }, + { + "epoch": 0.4698922662562524, + "grad_norm": 1.8776879237330482, + "learning_rate": 1.1455939997702946e-05, + "loss": 1.0639, + "step": 4885 + }, + { + "epoch": 0.4699884570988842, + "grad_norm": 2.033979715670483, + "learning_rate": 1.1452857700628362e-05, + "loss": 1.0324, + "step": 4886 + }, + { + "epoch": 0.47008464794151594, + "grad_norm": 2.0774446789658736, + "learning_rate": 1.144977526254143e-05, + "loss": 0.9807, + "step": 4887 + }, + { + "epoch": 0.47018083878414774, + "grad_norm": 1.9581193494538245, + "learning_rate": 1.1446692683741326e-05, + "loss": 0.9116, + "step": 4888 + }, + { + "epoch": 0.47027702962677953, + "grad_norm": 1.880924712754078, + "learning_rate": 1.144360996452724e-05, + "loss": 0.9252, + "step": 4889 + }, + { + "epoch": 0.4703732204694113, + "grad_norm": 1.9354969261105957, + "learning_rate": 1.1440527105198377e-05, + "loss": 0.9277, + "step": 4890 + }, + { + "epoch": 0.4704694113120431, + "grad_norm": 2.219510426479481, + "learning_rate": 1.1437444106053955e-05, + "loss": 0.9198, + "step": 4891 + }, + { + "epoch": 0.47056560215467486, + "grad_norm": 1.8567232527891884, + "learning_rate": 1.143436096739321e-05, + "loss": 0.8728, + "step": 4892 + }, + { + "epoch": 0.47066179299730665, + "grad_norm": 2.0945551696617013, + "learning_rate": 1.1431277689515379e-05, + "loss": 0.894, + "step": 4893 + }, + { + "epoch": 0.47075798383993844, + "grad_norm": 1.8672523432377408, + "learning_rate": 1.1428194272719729e-05, + "loss": 0.9723, + "step": 4894 + }, + { + "epoch": 0.47085417468257024, + "grad_norm": 2.2714240348847468, + "learning_rate": 1.1425110717305524e-05, + "loss": 1.0827, + "step": 4895 + }, + { + "epoch": 0.470950365525202, + "grad_norm": 1.8035259974804827, + "learning_rate": 1.1422027023572052e-05, + "loss": 0.8749, + "step": 4896 + }, + { + "epoch": 0.47104655636783377, + "grad_norm": 1.9170219987698278, + "learning_rate": 1.1418943191818616e-05, + "loss": 0.9991, + "step": 4897 + }, + { + "epoch": 0.47114274721046556, + "grad_norm": 1.9101877937229146, + "learning_rate": 1.1415859222344525e-05, + "loss": 1.0177, + "step": 4898 + }, + { + "epoch": 0.47123893805309736, + "grad_norm": 1.8595554984164462, + "learning_rate": 1.1412775115449109e-05, + "loss": 1.0113, + "step": 4899 + }, + { + "epoch": 0.47133512889572915, + "grad_norm": 1.8351275773593718, + "learning_rate": 1.14096908714317e-05, + "loss": 0.9961, + "step": 4900 + }, + { + "epoch": 0.4714313197383609, + "grad_norm": 1.8829848580639368, + "learning_rate": 1.1406606490591656e-05, + "loss": 1.0001, + "step": 4901 + }, + { + "epoch": 0.4715275105809927, + "grad_norm": 1.9496134704769092, + "learning_rate": 1.1403521973228342e-05, + "loss": 0.9095, + "step": 4902 + }, + { + "epoch": 0.4716237014236245, + "grad_norm": 1.9983983694385576, + "learning_rate": 1.1400437319641135e-05, + "loss": 1.1302, + "step": 4903 + }, + { + "epoch": 0.47171989226625627, + "grad_norm": 2.085720018888879, + "learning_rate": 1.1397352530129428e-05, + "loss": 1.0121, + "step": 4904 + }, + { + "epoch": 0.471816083108888, + "grad_norm": 2.09215039325497, + "learning_rate": 1.1394267604992625e-05, + "loss": 0.9844, + "step": 4905 + }, + { + "epoch": 0.4719122739515198, + "grad_norm": 2.028923800122762, + "learning_rate": 1.139118254453015e-05, + "loss": 1.0267, + "step": 4906 + }, + { + "epoch": 0.4720084647941516, + "grad_norm": 2.2186440774594933, + "learning_rate": 1.1388097349041429e-05, + "loss": 1.0576, + "step": 4907 + }, + { + "epoch": 0.4721046556367834, + "grad_norm": 2.16241251878298, + "learning_rate": 1.1385012018825907e-05, + "loss": 0.8972, + "step": 4908 + }, + { + "epoch": 0.4722008464794152, + "grad_norm": 1.9586782872479218, + "learning_rate": 1.1381926554183047e-05, + "loss": 1.013, + "step": 4909 + }, + { + "epoch": 0.4722970373220469, + "grad_norm": 1.9159290954736359, + "learning_rate": 1.1378840955412313e-05, + "loss": 0.9623, + "step": 4910 + }, + { + "epoch": 0.4723932281646787, + "grad_norm": 2.0580775363205737, + "learning_rate": 1.1375755222813195e-05, + "loss": 1.0579, + "step": 4911 + }, + { + "epoch": 0.4724894190073105, + "grad_norm": 2.080677294217094, + "learning_rate": 1.1372669356685185e-05, + "loss": 1.0305, + "step": 4912 + }, + { + "epoch": 0.4725856098499423, + "grad_norm": 2.012439640345742, + "learning_rate": 1.1369583357327795e-05, + "loss": 0.972, + "step": 4913 + }, + { + "epoch": 0.4726818006925741, + "grad_norm": 1.883401440615278, + "learning_rate": 1.1366497225040549e-05, + "loss": 0.9733, + "step": 4914 + }, + { + "epoch": 0.47277799153520583, + "grad_norm": 1.7695943805062153, + "learning_rate": 1.1363410960122982e-05, + "loss": 0.9034, + "step": 4915 + }, + { + "epoch": 0.4728741823778376, + "grad_norm": 2.4815751779151967, + "learning_rate": 1.1360324562874643e-05, + "loss": 0.9202, + "step": 4916 + }, + { + "epoch": 0.4729703732204694, + "grad_norm": 1.95736736499373, + "learning_rate": 1.1357238033595088e-05, + "loss": 0.9336, + "step": 4917 + }, + { + "epoch": 0.4730665640631012, + "grad_norm": 2.1438310073703386, + "learning_rate": 1.1354151372583901e-05, + "loss": 0.9117, + "step": 4918 + }, + { + "epoch": 0.47316275490573295, + "grad_norm": 1.8246098732402436, + "learning_rate": 1.135106458014066e-05, + "loss": 0.9738, + "step": 4919 + }, + { + "epoch": 0.47325894574836475, + "grad_norm": 1.9315835973280606, + "learning_rate": 1.1347977656564974e-05, + "loss": 0.9491, + "step": 4920 + }, + { + "epoch": 0.47335513659099654, + "grad_norm": 1.8609138797501388, + "learning_rate": 1.1344890602156449e-05, + "loss": 0.9605, + "step": 4921 + }, + { + "epoch": 0.47345132743362833, + "grad_norm": 2.117999204976607, + "learning_rate": 1.1341803417214705e-05, + "loss": 1.1106, + "step": 4922 + }, + { + "epoch": 0.4735475182762601, + "grad_norm": 2.2118626987052936, + "learning_rate": 1.1338716102039394e-05, + "loss": 0.9285, + "step": 4923 + }, + { + "epoch": 0.47364370911889186, + "grad_norm": 1.953455191287686, + "learning_rate": 1.1335628656930153e-05, + "loss": 0.9163, + "step": 4924 + }, + { + "epoch": 0.47373989996152366, + "grad_norm": 2.0229637678432213, + "learning_rate": 1.1332541082186654e-05, + "loss": 1.0778, + "step": 4925 + }, + { + "epoch": 0.47383609080415545, + "grad_norm": 2.1545602276575493, + "learning_rate": 1.132945337810857e-05, + "loss": 0.9127, + "step": 4926 + }, + { + "epoch": 0.47393228164678725, + "grad_norm": 2.020593508076279, + "learning_rate": 1.1326365544995587e-05, + "loss": 0.9883, + "step": 4927 + }, + { + "epoch": 0.474028472489419, + "grad_norm": 1.876984435366489, + "learning_rate": 1.132327758314741e-05, + "loss": 0.9894, + "step": 4928 + }, + { + "epoch": 0.4741246633320508, + "grad_norm": 2.3932929006494104, + "learning_rate": 1.132018949286375e-05, + "loss": 0.9576, + "step": 4929 + }, + { + "epoch": 0.47422085417468257, + "grad_norm": 2.065781364925584, + "learning_rate": 1.131710127444433e-05, + "loss": 1.0108, + "step": 4930 + }, + { + "epoch": 0.47431704501731436, + "grad_norm": 2.308708496977275, + "learning_rate": 1.1314012928188892e-05, + "loss": 1.0115, + "step": 4931 + }, + { + "epoch": 0.47441323585994616, + "grad_norm": 1.945933687499086, + "learning_rate": 1.1310924454397187e-05, + "loss": 0.9104, + "step": 4932 + }, + { + "epoch": 0.4745094267025779, + "grad_norm": 2.131905968075027, + "learning_rate": 1.1307835853368973e-05, + "loss": 1.0419, + "step": 4933 + }, + { + "epoch": 0.4746056175452097, + "grad_norm": 1.983387464444571, + "learning_rate": 1.1304747125404031e-05, + "loss": 1.0213, + "step": 4934 + }, + { + "epoch": 0.4747018083878415, + "grad_norm": 2.22063879166823, + "learning_rate": 1.130165827080215e-05, + "loss": 0.8543, + "step": 4935 + }, + { + "epoch": 0.4747979992304733, + "grad_norm": 2.0019824475682846, + "learning_rate": 1.129856928986312e-05, + "loss": 1.0362, + "step": 4936 + }, + { + "epoch": 0.474894190073105, + "grad_norm": 2.1733690874335307, + "learning_rate": 1.1295480182886762e-05, + "loss": 1.0883, + "step": 4937 + }, + { + "epoch": 0.4749903809157368, + "grad_norm": 1.881488746656745, + "learning_rate": 1.12923909501729e-05, + "loss": 1.028, + "step": 4938 + }, + { + "epoch": 0.4750865717583686, + "grad_norm": 1.9285112056710008, + "learning_rate": 1.1289301592021366e-05, + "loss": 0.9813, + "step": 4939 + }, + { + "epoch": 0.4751827626010004, + "grad_norm": 2.0684344938502273, + "learning_rate": 1.1286212108732015e-05, + "loss": 0.9231, + "step": 4940 + }, + { + "epoch": 0.4752789534436322, + "grad_norm": 2.0321549280830826, + "learning_rate": 1.1283122500604702e-05, + "loss": 0.9876, + "step": 4941 + }, + { + "epoch": 0.47537514428626393, + "grad_norm": 1.870929775515831, + "learning_rate": 1.1280032767939302e-05, + "loss": 0.982, + "step": 4942 + }, + { + "epoch": 0.4754713351288957, + "grad_norm": 1.846977077970949, + "learning_rate": 1.1276942911035702e-05, + "loss": 1.017, + "step": 4943 + }, + { + "epoch": 0.4755675259715275, + "grad_norm": 2.2104480150732346, + "learning_rate": 1.1273852930193798e-05, + "loss": 1.0876, + "step": 4944 + }, + { + "epoch": 0.4756637168141593, + "grad_norm": 1.85366406501718, + "learning_rate": 1.1270762825713497e-05, + "loss": 1.0587, + "step": 4945 + }, + { + "epoch": 0.47575990765679105, + "grad_norm": 2.009727374926086, + "learning_rate": 1.1267672597894725e-05, + "loss": 1.1491, + "step": 4946 + }, + { + "epoch": 0.47585609849942284, + "grad_norm": 2.154525416084689, + "learning_rate": 1.1264582247037413e-05, + "loss": 0.9163, + "step": 4947 + }, + { + "epoch": 0.47595228934205464, + "grad_norm": 2.5457397099402304, + "learning_rate": 1.12614917734415e-05, + "loss": 0.996, + "step": 4948 + }, + { + "epoch": 0.47604848018468643, + "grad_norm": 2.256698078960757, + "learning_rate": 1.1258401177406956e-05, + "loss": 1.0185, + "step": 4949 + }, + { + "epoch": 0.4761446710273182, + "grad_norm": 1.7953189677135104, + "learning_rate": 1.1255310459233737e-05, + "loss": 1.0147, + "step": 4950 + }, + { + "epoch": 0.47624086186994996, + "grad_norm": 1.9876283259747347, + "learning_rate": 1.1252219619221834e-05, + "loss": 0.9434, + "step": 4951 + }, + { + "epoch": 0.47633705271258175, + "grad_norm": 1.8530302688353286, + "learning_rate": 1.1249128657671233e-05, + "loss": 1.0229, + "step": 4952 + }, + { + "epoch": 0.47643324355521355, + "grad_norm": 2.0168610287218947, + "learning_rate": 1.124603757488194e-05, + "loss": 0.9317, + "step": 4953 + }, + { + "epoch": 0.47652943439784534, + "grad_norm": 2.1423211266330804, + "learning_rate": 1.1242946371153974e-05, + "loss": 1.0597, + "step": 4954 + }, + { + "epoch": 0.4766256252404771, + "grad_norm": 2.243645398821531, + "learning_rate": 1.123985504678736e-05, + "loss": 1.0383, + "step": 4955 + }, + { + "epoch": 0.4767218160831089, + "grad_norm": 1.7880805948040404, + "learning_rate": 1.1236763602082136e-05, + "loss": 0.9693, + "step": 4956 + }, + { + "epoch": 0.47681800692574067, + "grad_norm": 1.9935034545940857, + "learning_rate": 1.1233672037338356e-05, + "loss": 0.983, + "step": 4957 + }, + { + "epoch": 0.47691419776837246, + "grad_norm": 2.2301578929052193, + "learning_rate": 1.1230580352856088e-05, + "loss": 0.9173, + "step": 4958 + }, + { + "epoch": 0.47701038861100425, + "grad_norm": 1.8022769020141882, + "learning_rate": 1.12274885489354e-05, + "loss": 0.9532, + "step": 4959 + }, + { + "epoch": 0.477106579453636, + "grad_norm": 1.98546558193556, + "learning_rate": 1.1224396625876375e-05, + "loss": 0.833, + "step": 4960 + }, + { + "epoch": 0.4772027702962678, + "grad_norm": 2.0241243940225435, + "learning_rate": 1.122130458397912e-05, + "loss": 1.0318, + "step": 4961 + }, + { + "epoch": 0.4772989611388996, + "grad_norm": 2.0474985173453004, + "learning_rate": 1.1218212423543734e-05, + "loss": 1.0986, + "step": 4962 + }, + { + "epoch": 0.4773951519815314, + "grad_norm": 1.818289473350269, + "learning_rate": 1.1215120144870349e-05, + "loss": 0.9749, + "step": 4963 + }, + { + "epoch": 0.4774913428241631, + "grad_norm": 2.186424789465782, + "learning_rate": 1.1212027748259086e-05, + "loss": 1.0282, + "step": 4964 + }, + { + "epoch": 0.4775875336667949, + "grad_norm": 1.8904141908893146, + "learning_rate": 1.1208935234010098e-05, + "loss": 1.036, + "step": 4965 + }, + { + "epoch": 0.4776837245094267, + "grad_norm": 2.034719976039686, + "learning_rate": 1.1205842602423537e-05, + "loss": 1.0391, + "step": 4966 + }, + { + "epoch": 0.4777799153520585, + "grad_norm": 1.8327947250398782, + "learning_rate": 1.1202749853799568e-05, + "loss": 0.9199, + "step": 4967 + }, + { + "epoch": 0.4778761061946903, + "grad_norm": 1.9136000714958956, + "learning_rate": 1.1199656988438373e-05, + "loss": 1.034, + "step": 4968 + }, + { + "epoch": 0.477972297037322, + "grad_norm": 1.8566374832818784, + "learning_rate": 1.1196564006640134e-05, + "loss": 1.0138, + "step": 4969 + }, + { + "epoch": 0.4780684878799538, + "grad_norm": 2.2698631094605237, + "learning_rate": 1.1193470908705055e-05, + "loss": 1.1035, + "step": 4970 + }, + { + "epoch": 0.4781646787225856, + "grad_norm": 2.0645207994051558, + "learning_rate": 1.119037769493335e-05, + "loss": 0.9152, + "step": 4971 + }, + { + "epoch": 0.4782608695652174, + "grad_norm": 2.0703943821622803, + "learning_rate": 1.1187284365625241e-05, + "loss": 1.027, + "step": 4972 + }, + { + "epoch": 0.4783570604078492, + "grad_norm": 1.966010616617947, + "learning_rate": 1.1184190921080962e-05, + "loss": 1.0523, + "step": 4973 + }, + { + "epoch": 0.47845325125048094, + "grad_norm": 1.9105616965415682, + "learning_rate": 1.1181097361600754e-05, + "loss": 0.9478, + "step": 4974 + }, + { + "epoch": 0.47854944209311273, + "grad_norm": 1.8735329622924854, + "learning_rate": 1.1178003687484885e-05, + "loss": 1.0283, + "step": 4975 + }, + { + "epoch": 0.4786456329357445, + "grad_norm": 2.067541535965363, + "learning_rate": 1.1174909899033608e-05, + "loss": 0.9479, + "step": 4976 + }, + { + "epoch": 0.4787418237783763, + "grad_norm": 1.910073746233887, + "learning_rate": 1.1171815996547213e-05, + "loss": 1.0462, + "step": 4977 + }, + { + "epoch": 0.47883801462100806, + "grad_norm": 1.8397966686268306, + "learning_rate": 1.1168721980325987e-05, + "loss": 0.9874, + "step": 4978 + }, + { + "epoch": 0.47893420546363985, + "grad_norm": 1.979595752012595, + "learning_rate": 1.116562785067023e-05, + "loss": 1.0365, + "step": 4979 + }, + { + "epoch": 0.47903039630627164, + "grad_norm": 2.0944746364475297, + "learning_rate": 1.1162533607880251e-05, + "loss": 0.8897, + "step": 4980 + }, + { + "epoch": 0.47912658714890344, + "grad_norm": 1.8388786928896004, + "learning_rate": 1.1159439252256376e-05, + "loss": 1.0789, + "step": 4981 + }, + { + "epoch": 0.47922277799153523, + "grad_norm": 2.089029641145658, + "learning_rate": 1.1156344784098942e-05, + "loss": 1.0281, + "step": 4982 + }, + { + "epoch": 0.47931896883416697, + "grad_norm": 2.0418539927066606, + "learning_rate": 1.1153250203708285e-05, + "loss": 0.9428, + "step": 4983 + }, + { + "epoch": 0.47941515967679876, + "grad_norm": 2.2717148467579396, + "learning_rate": 1.1150155511384772e-05, + "loss": 1.0507, + "step": 4984 + }, + { + "epoch": 0.47951135051943056, + "grad_norm": 3.194489411583233, + "learning_rate": 1.114706070742876e-05, + "loss": 0.9022, + "step": 4985 + }, + { + "epoch": 0.47960754136206235, + "grad_norm": 2.08112942713465, + "learning_rate": 1.1143965792140631e-05, + "loss": 0.945, + "step": 4986 + }, + { + "epoch": 0.4797037322046941, + "grad_norm": 1.762516362683892, + "learning_rate": 1.1140870765820775e-05, + "loss": 0.8948, + "step": 4987 + }, + { + "epoch": 0.4797999230473259, + "grad_norm": 2.177118815460261, + "learning_rate": 1.1137775628769584e-05, + "loss": 1.0236, + "step": 4988 + }, + { + "epoch": 0.4798961138899577, + "grad_norm": 1.9185914281442884, + "learning_rate": 1.1134680381287472e-05, + "loss": 1.092, + "step": 4989 + }, + { + "epoch": 0.47999230473258947, + "grad_norm": 2.0257730402390064, + "learning_rate": 1.1131585023674863e-05, + "loss": 1.0378, + "step": 4990 + }, + { + "epoch": 0.48008849557522126, + "grad_norm": 2.044078636868884, + "learning_rate": 1.1128489556232183e-05, + "loss": 0.9356, + "step": 4991 + }, + { + "epoch": 0.480184686417853, + "grad_norm": 2.3801752903261177, + "learning_rate": 1.1125393979259874e-05, + "loss": 1.0024, + "step": 4992 + }, + { + "epoch": 0.4802808772604848, + "grad_norm": 1.7442854974124786, + "learning_rate": 1.1122298293058388e-05, + "loss": 0.9234, + "step": 4993 + }, + { + "epoch": 0.4803770681031166, + "grad_norm": 2.0717930875492154, + "learning_rate": 1.1119202497928192e-05, + "loss": 0.9979, + "step": 4994 + }, + { + "epoch": 0.4804732589457484, + "grad_norm": 1.8411536109130957, + "learning_rate": 1.1116106594169758e-05, + "loss": 0.9552, + "step": 4995 + }, + { + "epoch": 0.4805694497883801, + "grad_norm": 2.01021338912377, + "learning_rate": 1.1113010582083568e-05, + "loss": 0.9572, + "step": 4996 + }, + { + "epoch": 0.4806656406310119, + "grad_norm": 2.121292691607287, + "learning_rate": 1.1109914461970114e-05, + "loss": 1.0085, + "step": 4997 + }, + { + "epoch": 0.4807618314736437, + "grad_norm": 1.8999663528827953, + "learning_rate": 1.1106818234129913e-05, + "loss": 0.8782, + "step": 4998 + }, + { + "epoch": 0.4808580223162755, + "grad_norm": 2.1419667405121, + "learning_rate": 1.1103721898863467e-05, + "loss": 0.9123, + "step": 4999 + }, + { + "epoch": 0.4809542131589073, + "grad_norm": 1.9187832839500416, + "learning_rate": 1.1100625456471307e-05, + "loss": 0.9468, + "step": 5000 + }, + { + "epoch": 0.48105040400153903, + "grad_norm": 2.2289708603320997, + "learning_rate": 1.1097528907253974e-05, + "loss": 0.9817, + "step": 5001 + }, + { + "epoch": 0.4811465948441708, + "grad_norm": 2.040914800626172, + "learning_rate": 1.1094432251512006e-05, + "loss": 0.9665, + "step": 5002 + }, + { + "epoch": 0.4812427856868026, + "grad_norm": 2.026306239499146, + "learning_rate": 1.1091335489545967e-05, + "loss": 0.9168, + "step": 5003 + }, + { + "epoch": 0.4813389765294344, + "grad_norm": 2.2614861689875476, + "learning_rate": 1.1088238621656422e-05, + "loss": 0.9917, + "step": 5004 + }, + { + "epoch": 0.48143516737206615, + "grad_norm": 2.062002926977596, + "learning_rate": 1.1085141648143952e-05, + "loss": 0.8978, + "step": 5005 + }, + { + "epoch": 0.48153135821469795, + "grad_norm": 2.1526136589382854, + "learning_rate": 1.1082044569309138e-05, + "loss": 0.9462, + "step": 5006 + }, + { + "epoch": 0.48162754905732974, + "grad_norm": 1.9122580871995798, + "learning_rate": 1.1078947385452583e-05, + "loss": 1.007, + "step": 5007 + }, + { + "epoch": 0.48172373989996153, + "grad_norm": 2.068749218292341, + "learning_rate": 1.1075850096874894e-05, + "loss": 1.0541, + "step": 5008 + }, + { + "epoch": 0.4818199307425933, + "grad_norm": 1.959389419823863, + "learning_rate": 1.107275270387669e-05, + "loss": 1.0504, + "step": 5009 + }, + { + "epoch": 0.48191612158522507, + "grad_norm": 1.8841399349403538, + "learning_rate": 1.1069655206758603e-05, + "loss": 0.8685, + "step": 5010 + }, + { + "epoch": 0.48201231242785686, + "grad_norm": 2.2345079946313167, + "learning_rate": 1.1066557605821261e-05, + "loss": 0.9506, + "step": 5011 + }, + { + "epoch": 0.48210850327048865, + "grad_norm": 1.9653525143408286, + "learning_rate": 1.1063459901365325e-05, + "loss": 0.9129, + "step": 5012 + }, + { + "epoch": 0.48220469411312045, + "grad_norm": 2.0966401274347604, + "learning_rate": 1.1060362093691452e-05, + "loss": 0.9746, + "step": 5013 + }, + { + "epoch": 0.4823008849557522, + "grad_norm": 1.8913128605808291, + "learning_rate": 1.1057264183100303e-05, + "loss": 0.964, + "step": 5014 + }, + { + "epoch": 0.482397075798384, + "grad_norm": 2.016131655307724, + "learning_rate": 1.1054166169892565e-05, + "loss": 1.0684, + "step": 5015 + }, + { + "epoch": 0.48249326664101577, + "grad_norm": 1.893758577391111, + "learning_rate": 1.1051068054368921e-05, + "loss": 0.9297, + "step": 5016 + }, + { + "epoch": 0.48258945748364757, + "grad_norm": 2.031264568683315, + "learning_rate": 1.1047969836830073e-05, + "loss": 1.0306, + "step": 5017 + }, + { + "epoch": 0.48268564832627936, + "grad_norm": 1.84262588097307, + "learning_rate": 1.104487151757673e-05, + "loss": 1.0574, + "step": 5018 + }, + { + "epoch": 0.4827818391689111, + "grad_norm": 1.9108508043340113, + "learning_rate": 1.104177309690961e-05, + "loss": 1.0156, + "step": 5019 + }, + { + "epoch": 0.4828780300115429, + "grad_norm": 1.938542232435125, + "learning_rate": 1.1038674575129442e-05, + "loss": 0.9135, + "step": 5020 + }, + { + "epoch": 0.4829742208541747, + "grad_norm": 2.0389268894903005, + "learning_rate": 1.103557595253696e-05, + "loss": 0.9766, + "step": 5021 + }, + { + "epoch": 0.4830704116968065, + "grad_norm": 2.0072302744284847, + "learning_rate": 1.1032477229432921e-05, + "loss": 1.1237, + "step": 5022 + }, + { + "epoch": 0.4831666025394383, + "grad_norm": 1.9402377490797906, + "learning_rate": 1.102937840611807e-05, + "loss": 0.9744, + "step": 5023 + }, + { + "epoch": 0.48326279338207, + "grad_norm": 2.1534062511617478, + "learning_rate": 1.1026279482893187e-05, + "loss": 1.0396, + "step": 5024 + }, + { + "epoch": 0.4833589842247018, + "grad_norm": 2.224441723669427, + "learning_rate": 1.1023180460059042e-05, + "loss": 1.0354, + "step": 5025 + }, + { + "epoch": 0.4834551750673336, + "grad_norm": 1.9234123366727014, + "learning_rate": 1.1020081337916425e-05, + "loss": 0.818, + "step": 5026 + }, + { + "epoch": 0.4835513659099654, + "grad_norm": 1.939569740672499, + "learning_rate": 1.101698211676613e-05, + "loss": 0.926, + "step": 5027 + }, + { + "epoch": 0.48364755675259713, + "grad_norm": 2.0140101936859107, + "learning_rate": 1.1013882796908963e-05, + "loss": 0.9169, + "step": 5028 + }, + { + "epoch": 0.4837437475952289, + "grad_norm": 2.1760851253157116, + "learning_rate": 1.1010783378645742e-05, + "loss": 0.8729, + "step": 5029 + }, + { + "epoch": 0.4838399384378607, + "grad_norm": 1.9116729217454802, + "learning_rate": 1.1007683862277292e-05, + "loss": 0.9171, + "step": 5030 + }, + { + "epoch": 0.4839361292804925, + "grad_norm": 1.8897796772865911, + "learning_rate": 1.1004584248104447e-05, + "loss": 0.8369, + "step": 5031 + }, + { + "epoch": 0.4840323201231243, + "grad_norm": 1.7503435215317853, + "learning_rate": 1.1001484536428052e-05, + "loss": 1.0045, + "step": 5032 + }, + { + "epoch": 0.48412851096575604, + "grad_norm": 2.1003458936417467, + "learning_rate": 1.0998384727548956e-05, + "loss": 1.0463, + "step": 5033 + }, + { + "epoch": 0.48422470180838784, + "grad_norm": 2.3769331823099633, + "learning_rate": 1.0995284821768029e-05, + "loss": 1.0387, + "step": 5034 + }, + { + "epoch": 0.48432089265101963, + "grad_norm": 2.0757148868083264, + "learning_rate": 1.0992184819386138e-05, + "loss": 0.9897, + "step": 5035 + }, + { + "epoch": 0.4844170834936514, + "grad_norm": 2.0863497203014343, + "learning_rate": 1.098908472070417e-05, + "loss": 0.9858, + "step": 5036 + }, + { + "epoch": 0.48451327433628316, + "grad_norm": 1.8495780116702711, + "learning_rate": 1.0985984526023013e-05, + "loss": 0.9939, + "step": 5037 + }, + { + "epoch": 0.48460946517891496, + "grad_norm": 1.873722291886999, + "learning_rate": 1.0982884235643567e-05, + "loss": 0.8386, + "step": 5038 + }, + { + "epoch": 0.48470565602154675, + "grad_norm": 1.7856182675091883, + "learning_rate": 1.0979783849866748e-05, + "loss": 1.0128, + "step": 5039 + }, + { + "epoch": 0.48480184686417854, + "grad_norm": 1.9984676345108736, + "learning_rate": 1.0976683368993464e-05, + "loss": 1.0483, + "step": 5040 + }, + { + "epoch": 0.48489803770681034, + "grad_norm": 2.022371997017358, + "learning_rate": 1.0973582793324653e-05, + "loss": 0.9994, + "step": 5041 + }, + { + "epoch": 0.4849942285494421, + "grad_norm": 2.0437357955981366, + "learning_rate": 1.0970482123161249e-05, + "loss": 1.0114, + "step": 5042 + }, + { + "epoch": 0.48509041939207387, + "grad_norm": 2.187481584434219, + "learning_rate": 1.0967381358804199e-05, + "loss": 1.0524, + "step": 5043 + }, + { + "epoch": 0.48518661023470566, + "grad_norm": 1.9947741522666893, + "learning_rate": 1.0964280500554459e-05, + "loss": 0.9923, + "step": 5044 + }, + { + "epoch": 0.48528280107733746, + "grad_norm": 1.7858371818029672, + "learning_rate": 1.0961179548712992e-05, + "loss": 0.941, + "step": 5045 + }, + { + "epoch": 0.4853789919199692, + "grad_norm": 2.258816338676785, + "learning_rate": 1.0958078503580776e-05, + "loss": 0.998, + "step": 5046 + }, + { + "epoch": 0.485475182762601, + "grad_norm": 1.8781520988380318, + "learning_rate": 1.0954977365458792e-05, + "loss": 1.0259, + "step": 5047 + }, + { + "epoch": 0.4855713736052328, + "grad_norm": 1.9729798224395962, + "learning_rate": 1.0951876134648032e-05, + "loss": 1.0329, + "step": 5048 + }, + { + "epoch": 0.4856675644478646, + "grad_norm": 1.999226356906015, + "learning_rate": 1.0948774811449494e-05, + "loss": 1.0417, + "step": 5049 + }, + { + "epoch": 0.48576375529049637, + "grad_norm": 1.834207235457478, + "learning_rate": 1.0945673396164198e-05, + "loss": 1.0044, + "step": 5050 + }, + { + "epoch": 0.4858599461331281, + "grad_norm": 2.1998224277285523, + "learning_rate": 1.0942571889093152e-05, + "loss": 0.9707, + "step": 5051 + }, + { + "epoch": 0.4859561369757599, + "grad_norm": 2.2500384174968553, + "learning_rate": 1.0939470290537389e-05, + "loss": 1.0815, + "step": 5052 + }, + { + "epoch": 0.4860523278183917, + "grad_norm": 2.3442030395019247, + "learning_rate": 1.0936368600797949e-05, + "loss": 0.9972, + "step": 5053 + }, + { + "epoch": 0.4861485186610235, + "grad_norm": 1.9925728476820908, + "learning_rate": 1.0933266820175868e-05, + "loss": 0.8986, + "step": 5054 + }, + { + "epoch": 0.4862447095036552, + "grad_norm": 1.9661950364961847, + "learning_rate": 1.0930164948972211e-05, + "loss": 1.0362, + "step": 5055 + }, + { + "epoch": 0.486340900346287, + "grad_norm": 2.152933128079554, + "learning_rate": 1.0927062987488035e-05, + "loss": 0.947, + "step": 5056 + }, + { + "epoch": 0.4864370911889188, + "grad_norm": 2.0837181888646454, + "learning_rate": 1.0923960936024414e-05, + "loss": 1.0006, + "step": 5057 + }, + { + "epoch": 0.4865332820315506, + "grad_norm": 1.9536129973159495, + "learning_rate": 1.0920858794882429e-05, + "loss": 0.986, + "step": 5058 + }, + { + "epoch": 0.4866294728741824, + "grad_norm": 1.832457136031206, + "learning_rate": 1.091775656436317e-05, + "loss": 0.9623, + "step": 5059 + }, + { + "epoch": 0.48672566371681414, + "grad_norm": 2.069807305666751, + "learning_rate": 1.0914654244767736e-05, + "loss": 1.0084, + "step": 5060 + }, + { + "epoch": 0.48682185455944593, + "grad_norm": 2.284297552314281, + "learning_rate": 1.0911551836397227e-05, + "loss": 0.9777, + "step": 5061 + }, + { + "epoch": 0.4869180454020777, + "grad_norm": 2.1555182293731545, + "learning_rate": 1.0908449339552769e-05, + "loss": 0.8681, + "step": 5062 + }, + { + "epoch": 0.4870142362447095, + "grad_norm": 1.9087158617171287, + "learning_rate": 1.0905346754535477e-05, + "loss": 0.8542, + "step": 5063 + }, + { + "epoch": 0.48711042708734126, + "grad_norm": 1.876331808438281, + "learning_rate": 1.0902244081646489e-05, + "loss": 0.9305, + "step": 5064 + }, + { + "epoch": 0.48720661792997305, + "grad_norm": 2.4327733316696394, + "learning_rate": 1.0899141321186946e-05, + "loss": 1.0886, + "step": 5065 + }, + { + "epoch": 0.48730280877260485, + "grad_norm": 1.9614333785634064, + "learning_rate": 1.0896038473457993e-05, + "loss": 0.9521, + "step": 5066 + }, + { + "epoch": 0.48739899961523664, + "grad_norm": 2.1633366982554825, + "learning_rate": 1.0892935538760796e-05, + "loss": 0.924, + "step": 5067 + }, + { + "epoch": 0.48749519045786843, + "grad_norm": 1.928835104492705, + "learning_rate": 1.0889832517396511e-05, + "loss": 0.921, + "step": 5068 + }, + { + "epoch": 0.48759138130050017, + "grad_norm": 1.8448567254444754, + "learning_rate": 1.088672940966632e-05, + "loss": 0.9962, + "step": 5069 + }, + { + "epoch": 0.48768757214313196, + "grad_norm": 1.9211126903430673, + "learning_rate": 1.0883626215871408e-05, + "loss": 0.9326, + "step": 5070 + }, + { + "epoch": 0.48778376298576376, + "grad_norm": 1.9597890012366468, + "learning_rate": 1.0880522936312961e-05, + "loss": 0.7858, + "step": 5071 + }, + { + "epoch": 0.48787995382839555, + "grad_norm": 1.894909239505472, + "learning_rate": 1.0877419571292183e-05, + "loss": 0.9273, + "step": 5072 + }, + { + "epoch": 0.48797614467102735, + "grad_norm": 1.9992038540399657, + "learning_rate": 1.087431612111028e-05, + "loss": 0.9348, + "step": 5073 + }, + { + "epoch": 0.4880723355136591, + "grad_norm": 2.0419287605633807, + "learning_rate": 1.0871212586068469e-05, + "loss": 1.0311, + "step": 5074 + }, + { + "epoch": 0.4881685263562909, + "grad_norm": 1.8293672026543737, + "learning_rate": 1.0868108966467975e-05, + "loss": 0.9927, + "step": 5075 + }, + { + "epoch": 0.48826471719892267, + "grad_norm": 1.8706262407968484, + "learning_rate": 1.0865005262610033e-05, + "loss": 0.902, + "step": 5076 + }, + { + "epoch": 0.48836090804155446, + "grad_norm": 2.2558914151707703, + "learning_rate": 1.086190147479588e-05, + "loss": 1.0825, + "step": 5077 + }, + { + "epoch": 0.4884570988841862, + "grad_norm": 1.9816202798664482, + "learning_rate": 1.085879760332677e-05, + "loss": 1.035, + "step": 5078 + }, + { + "epoch": 0.488553289726818, + "grad_norm": 2.1360131453210194, + "learning_rate": 1.085569364850396e-05, + "loss": 0.9823, + "step": 5079 + }, + { + "epoch": 0.4886494805694498, + "grad_norm": 2.078161210075355, + "learning_rate": 1.085258961062871e-05, + "loss": 0.9174, + "step": 5080 + }, + { + "epoch": 0.4887456714120816, + "grad_norm": 2.2269976980769677, + "learning_rate": 1.0849485490002298e-05, + "loss": 1.0094, + "step": 5081 + }, + { + "epoch": 0.4888418622547134, + "grad_norm": 2.083731730609748, + "learning_rate": 1.0846381286926007e-05, + "loss": 0.8696, + "step": 5082 + }, + { + "epoch": 0.4889380530973451, + "grad_norm": 1.9866874363246478, + "learning_rate": 1.0843277001701124e-05, + "loss": 0.9103, + "step": 5083 + }, + { + "epoch": 0.4890342439399769, + "grad_norm": 2.1835100438409722, + "learning_rate": 1.0840172634628948e-05, + "loss": 0.9394, + "step": 5084 + }, + { + "epoch": 0.4891304347826087, + "grad_norm": 2.3397088467959075, + "learning_rate": 1.0837068186010782e-05, + "loss": 0.9629, + "step": 5085 + }, + { + "epoch": 0.4892266256252405, + "grad_norm": 1.974395924807058, + "learning_rate": 1.0833963656147944e-05, + "loss": 0.941, + "step": 5086 + }, + { + "epoch": 0.48932281646787223, + "grad_norm": 2.0970150914452614, + "learning_rate": 1.0830859045341748e-05, + "loss": 1.1146, + "step": 5087 + }, + { + "epoch": 0.48941900731050403, + "grad_norm": 1.904270970601564, + "learning_rate": 1.082775435389353e-05, + "loss": 0.9032, + "step": 5088 + }, + { + "epoch": 0.4895151981531358, + "grad_norm": 2.370688371858579, + "learning_rate": 1.0824649582104625e-05, + "loss": 1.0161, + "step": 5089 + }, + { + "epoch": 0.4896113889957676, + "grad_norm": 2.1731332731472075, + "learning_rate": 1.0821544730276379e-05, + "loss": 0.9112, + "step": 5090 + }, + { + "epoch": 0.4897075798383994, + "grad_norm": 1.8485929754958492, + "learning_rate": 1.0818439798710142e-05, + "loss": 1.0726, + "step": 5091 + }, + { + "epoch": 0.48980377068103115, + "grad_norm": 1.9829324284948142, + "learning_rate": 1.0815334787707277e-05, + "loss": 1.0, + "step": 5092 + }, + { + "epoch": 0.48989996152366294, + "grad_norm": 1.876816799337412, + "learning_rate": 1.0812229697569153e-05, + "loss": 1.0041, + "step": 5093 + }, + { + "epoch": 0.48999615236629473, + "grad_norm": 2.0637504061748086, + "learning_rate": 1.0809124528597138e-05, + "loss": 0.9891, + "step": 5094 + }, + { + "epoch": 0.49009234320892653, + "grad_norm": 1.6856472361149153, + "learning_rate": 1.0806019281092623e-05, + "loss": 1.0105, + "step": 5095 + }, + { + "epoch": 0.49018853405155827, + "grad_norm": 1.8490003932742933, + "learning_rate": 1.0802913955356998e-05, + "loss": 0.8274, + "step": 5096 + }, + { + "epoch": 0.49028472489419006, + "grad_norm": 1.708639340305721, + "learning_rate": 1.0799808551691659e-05, + "loss": 0.8876, + "step": 5097 + }, + { + "epoch": 0.49038091573682185, + "grad_norm": 2.1320240058277147, + "learning_rate": 1.0796703070398016e-05, + "loss": 0.8782, + "step": 5098 + }, + { + "epoch": 0.49047710657945365, + "grad_norm": 2.2239465700319143, + "learning_rate": 1.0793597511777479e-05, + "loss": 1.0302, + "step": 5099 + }, + { + "epoch": 0.49057329742208544, + "grad_norm": 1.9294207085564998, + "learning_rate": 1.079049187613147e-05, + "loss": 0.897, + "step": 5100 + }, + { + "epoch": 0.4906694882647172, + "grad_norm": 2.3865731253486766, + "learning_rate": 1.0787386163761416e-05, + "loss": 0.9273, + "step": 5101 + }, + { + "epoch": 0.490765679107349, + "grad_norm": 2.081417846096606, + "learning_rate": 1.0784280374968761e-05, + "loss": 0.982, + "step": 5102 + }, + { + "epoch": 0.49086186994998077, + "grad_norm": 1.876150913166332, + "learning_rate": 1.0781174510054936e-05, + "loss": 0.961, + "step": 5103 + }, + { + "epoch": 0.49095806079261256, + "grad_norm": 2.064872326171445, + "learning_rate": 1.0778068569321403e-05, + "loss": 0.9898, + "step": 5104 + }, + { + "epoch": 0.4910542516352443, + "grad_norm": 2.206382915078761, + "learning_rate": 1.077496255306962e-05, + "loss": 0.9776, + "step": 5105 + }, + { + "epoch": 0.4911504424778761, + "grad_norm": 2.254380469334719, + "learning_rate": 1.077185646160104e-05, + "loss": 0.9148, + "step": 5106 + }, + { + "epoch": 0.4912466333205079, + "grad_norm": 1.926630057418251, + "learning_rate": 1.076875029521715e-05, + "loss": 0.8656, + "step": 5107 + }, + { + "epoch": 0.4913428241631397, + "grad_norm": 1.8827924805500282, + "learning_rate": 1.0765644054219422e-05, + "loss": 0.9466, + "step": 5108 + }, + { + "epoch": 0.4914390150057715, + "grad_norm": 2.1849543462865944, + "learning_rate": 1.076253773890935e-05, + "loss": 1.0474, + "step": 5109 + }, + { + "epoch": 0.4915352058484032, + "grad_norm": 3.4724428651023707, + "learning_rate": 1.0759431349588421e-05, + "loss": 0.8793, + "step": 5110 + }, + { + "epoch": 0.491631396691035, + "grad_norm": 1.8663818140648647, + "learning_rate": 1.0756324886558142e-05, + "loss": 0.8622, + "step": 5111 + }, + { + "epoch": 0.4917275875336668, + "grad_norm": 2.4108000141889585, + "learning_rate": 1.0753218350120023e-05, + "loss": 0.9107, + "step": 5112 + }, + { + "epoch": 0.4918237783762986, + "grad_norm": 2.0623135201362994, + "learning_rate": 1.0750111740575572e-05, + "loss": 0.9639, + "step": 5113 + }, + { + "epoch": 0.49191996921893033, + "grad_norm": 2.017723091701873, + "learning_rate": 1.0747005058226325e-05, + "loss": 1.0651, + "step": 5114 + }, + { + "epoch": 0.4920161600615621, + "grad_norm": 2.2808473575189696, + "learning_rate": 1.0743898303373802e-05, + "loss": 1.1167, + "step": 5115 + }, + { + "epoch": 0.4921123509041939, + "grad_norm": 2.154158108403636, + "learning_rate": 1.0740791476319543e-05, + "loss": 0.9865, + "step": 5116 + }, + { + "epoch": 0.4922085417468257, + "grad_norm": 1.9816378530270342, + "learning_rate": 1.0737684577365096e-05, + "loss": 0.9676, + "step": 5117 + }, + { + "epoch": 0.4923047325894575, + "grad_norm": 1.8646973819103028, + "learning_rate": 1.0734577606812007e-05, + "loss": 0.9783, + "step": 5118 + }, + { + "epoch": 0.49240092343208924, + "grad_norm": 2.056810731133637, + "learning_rate": 1.0731470564961838e-05, + "loss": 0.9446, + "step": 5119 + }, + { + "epoch": 0.49249711427472104, + "grad_norm": 1.9742116056553891, + "learning_rate": 1.0728363452116149e-05, + "loss": 0.8901, + "step": 5120 + }, + { + "epoch": 0.49259330511735283, + "grad_norm": 2.1331216766997114, + "learning_rate": 1.0725256268576518e-05, + "loss": 1.1129, + "step": 5121 + }, + { + "epoch": 0.4926894959599846, + "grad_norm": 1.9939252687636624, + "learning_rate": 1.0722149014644523e-05, + "loss": 0.9403, + "step": 5122 + }, + { + "epoch": 0.4927856868026164, + "grad_norm": 1.845909886008212, + "learning_rate": 1.0719041690621745e-05, + "loss": 0.8626, + "step": 5123 + }, + { + "epoch": 0.49288187764524816, + "grad_norm": 2.069710534597736, + "learning_rate": 1.0715934296809782e-05, + "loss": 1.0344, + "step": 5124 + }, + { + "epoch": 0.49297806848787995, + "grad_norm": 1.98339398507883, + "learning_rate": 1.0712826833510232e-05, + "loss": 0.8612, + "step": 5125 + }, + { + "epoch": 0.49307425933051174, + "grad_norm": 2.0175555324427794, + "learning_rate": 1.0709719301024698e-05, + "loss": 0.9229, + "step": 5126 + }, + { + "epoch": 0.49317045017314354, + "grad_norm": 2.0709038183444926, + "learning_rate": 1.0706611699654797e-05, + "loss": 1.0524, + "step": 5127 + }, + { + "epoch": 0.4932666410157753, + "grad_norm": 2.3639606399621034, + "learning_rate": 1.0703504029702148e-05, + "loss": 1.0242, + "step": 5128 + }, + { + "epoch": 0.49336283185840707, + "grad_norm": 1.8243295915747162, + "learning_rate": 1.0700396291468371e-05, + "loss": 0.8777, + "step": 5129 + }, + { + "epoch": 0.49345902270103886, + "grad_norm": 2.2282252069411226, + "learning_rate": 1.0697288485255107e-05, + "loss": 1.002, + "step": 5130 + }, + { + "epoch": 0.49355521354367066, + "grad_norm": 2.029896144368045, + "learning_rate": 1.0694180611363989e-05, + "loss": 0.9762, + "step": 5131 + }, + { + "epoch": 0.49365140438630245, + "grad_norm": 1.8169630953913032, + "learning_rate": 1.0691072670096669e-05, + "loss": 0.994, + "step": 5132 + }, + { + "epoch": 0.4937475952289342, + "grad_norm": 2.1290481825491367, + "learning_rate": 1.0687964661754795e-05, + "loss": 0.9951, + "step": 5133 + }, + { + "epoch": 0.493843786071566, + "grad_norm": 2.3286889677656917, + "learning_rate": 1.0684856586640026e-05, + "loss": 0.9531, + "step": 5134 + }, + { + "epoch": 0.4939399769141978, + "grad_norm": 2.0716142052335638, + "learning_rate": 1.068174844505403e-05, + "loss": 0.9348, + "step": 5135 + }, + { + "epoch": 0.49403616775682957, + "grad_norm": 2.1638639134945654, + "learning_rate": 1.0678640237298476e-05, + "loss": 1.013, + "step": 5136 + }, + { + "epoch": 0.4941323585994613, + "grad_norm": 2.0243347509233347, + "learning_rate": 1.0675531963675045e-05, + "loss": 0.952, + "step": 5137 + }, + { + "epoch": 0.4942285494420931, + "grad_norm": 2.300412195389057, + "learning_rate": 1.0672423624485423e-05, + "loss": 0.9802, + "step": 5138 + }, + { + "epoch": 0.4943247402847249, + "grad_norm": 2.114116664452061, + "learning_rate": 1.0669315220031291e-05, + "loss": 1.0738, + "step": 5139 + }, + { + "epoch": 0.4944209311273567, + "grad_norm": 1.9120769357521714, + "learning_rate": 1.0666206750614363e-05, + "loss": 0.9816, + "step": 5140 + }, + { + "epoch": 0.4945171219699885, + "grad_norm": 2.132358984546464, + "learning_rate": 1.0663098216536327e-05, + "loss": 1.0755, + "step": 5141 + }, + { + "epoch": 0.4946133128126202, + "grad_norm": 1.8962107406781412, + "learning_rate": 1.0659989618098904e-05, + "loss": 0.9018, + "step": 5142 + }, + { + "epoch": 0.494709503655252, + "grad_norm": 2.013771272270337, + "learning_rate": 1.0656880955603804e-05, + "loss": 1.0401, + "step": 5143 + }, + { + "epoch": 0.4948056944978838, + "grad_norm": 2.1417529916346534, + "learning_rate": 1.065377222935275e-05, + "loss": 1.0368, + "step": 5144 + }, + { + "epoch": 0.4949018853405156, + "grad_norm": 1.63598819959575, + "learning_rate": 1.0650663439647477e-05, + "loss": 0.8095, + "step": 5145 + }, + { + "epoch": 0.49499807618314734, + "grad_norm": 2.1316437265488424, + "learning_rate": 1.0647554586789708e-05, + "loss": 0.939, + "step": 5146 + }, + { + "epoch": 0.49509426702577913, + "grad_norm": 1.899895380216439, + "learning_rate": 1.0644445671081196e-05, + "loss": 0.9555, + "step": 5147 + }, + { + "epoch": 0.4951904578684109, + "grad_norm": 2.1129549858871566, + "learning_rate": 1.064133669282368e-05, + "loss": 0.956, + "step": 5148 + }, + { + "epoch": 0.4952866487110427, + "grad_norm": 1.8400805663063673, + "learning_rate": 1.0638227652318917e-05, + "loss": 0.9609, + "step": 5149 + }, + { + "epoch": 0.4953828395536745, + "grad_norm": 2.37505968445819, + "learning_rate": 1.0635118549868668e-05, + "loss": 0.9821, + "step": 5150 + }, + { + "epoch": 0.49547903039630625, + "grad_norm": 2.324610909343112, + "learning_rate": 1.0632009385774693e-05, + "loss": 1.004, + "step": 5151 + }, + { + "epoch": 0.49557522123893805, + "grad_norm": 1.99827807605493, + "learning_rate": 1.0628900160338764e-05, + "loss": 0.9287, + "step": 5152 + }, + { + "epoch": 0.49567141208156984, + "grad_norm": 1.9501528303948685, + "learning_rate": 1.062579087386266e-05, + "loss": 1.0197, + "step": 5153 + }, + { + "epoch": 0.49576760292420163, + "grad_norm": 1.9561310679912431, + "learning_rate": 1.0622681526648167e-05, + "loss": 0.9379, + "step": 5154 + }, + { + "epoch": 0.49586379376683337, + "grad_norm": 1.991288333640989, + "learning_rate": 1.0619572118997066e-05, + "loss": 0.9091, + "step": 5155 + }, + { + "epoch": 0.49595998460946517, + "grad_norm": 1.9993692759081216, + "learning_rate": 1.0616462651211156e-05, + "loss": 0.9718, + "step": 5156 + }, + { + "epoch": 0.49605617545209696, + "grad_norm": 1.973547959313998, + "learning_rate": 1.061335312359224e-05, + "loss": 0.9864, + "step": 5157 + }, + { + "epoch": 0.49615236629472875, + "grad_norm": 2.156550738683227, + "learning_rate": 1.0610243536442125e-05, + "loss": 1.0408, + "step": 5158 + }, + { + "epoch": 0.49624855713736055, + "grad_norm": 2.032421462524198, + "learning_rate": 1.0607133890062618e-05, + "loss": 0.8033, + "step": 5159 + }, + { + "epoch": 0.4963447479799923, + "grad_norm": 1.9036526036073877, + "learning_rate": 1.0604024184755539e-05, + "loss": 1.0161, + "step": 5160 + }, + { + "epoch": 0.4964409388226241, + "grad_norm": 1.81813638511458, + "learning_rate": 1.0600914420822713e-05, + "loss": 1.0044, + "step": 5161 + }, + { + "epoch": 0.49653712966525587, + "grad_norm": 6.62485629700568, + "learning_rate": 1.0597804598565969e-05, + "loss": 0.8968, + "step": 5162 + }, + { + "epoch": 0.49663332050788767, + "grad_norm": 2.1536781130883127, + "learning_rate": 1.059469471828714e-05, + "loss": 0.9807, + "step": 5163 + }, + { + "epoch": 0.4967295113505194, + "grad_norm": 1.8612557791424673, + "learning_rate": 1.0591584780288069e-05, + "loss": 0.9389, + "step": 5164 + }, + { + "epoch": 0.4968257021931512, + "grad_norm": 2.0759566480109477, + "learning_rate": 1.05884747848706e-05, + "loss": 1.0407, + "step": 5165 + }, + { + "epoch": 0.496921893035783, + "grad_norm": 1.8760979926279937, + "learning_rate": 1.0585364732336587e-05, + "loss": 0.9198, + "step": 5166 + }, + { + "epoch": 0.4970180838784148, + "grad_norm": 1.9278396435758633, + "learning_rate": 1.0582254622987887e-05, + "loss": 1.0686, + "step": 5167 + }, + { + "epoch": 0.4971142747210466, + "grad_norm": 1.797815272951351, + "learning_rate": 1.0579144457126365e-05, + "loss": 0.9186, + "step": 5168 + }, + { + "epoch": 0.4972104655636783, + "grad_norm": 1.8439694719005617, + "learning_rate": 1.0576034235053882e-05, + "loss": 0.9929, + "step": 5169 + }, + { + "epoch": 0.4973066564063101, + "grad_norm": 1.841278519436763, + "learning_rate": 1.057292395707232e-05, + "loss": 0.9949, + "step": 5170 + }, + { + "epoch": 0.4974028472489419, + "grad_norm": 1.8772230501398526, + "learning_rate": 1.0569813623483555e-05, + "loss": 0.9986, + "step": 5171 + }, + { + "epoch": 0.4974990380915737, + "grad_norm": 1.8855848866192415, + "learning_rate": 1.0566703234589471e-05, + "loss": 0.9156, + "step": 5172 + }, + { + "epoch": 0.4975952289342055, + "grad_norm": 1.8491786315495455, + "learning_rate": 1.056359279069196e-05, + "loss": 0.9523, + "step": 5173 + }, + { + "epoch": 0.49769141977683723, + "grad_norm": 2.163910302049965, + "learning_rate": 1.0560482292092912e-05, + "loss": 0.959, + "step": 5174 + }, + { + "epoch": 0.497787610619469, + "grad_norm": 1.6730458540005917, + "learning_rate": 1.0557371739094238e-05, + "loss": 0.9044, + "step": 5175 + }, + { + "epoch": 0.4978838014621008, + "grad_norm": 1.8484440593389748, + "learning_rate": 1.0554261131997833e-05, + "loss": 0.9077, + "step": 5176 + }, + { + "epoch": 0.4979799923047326, + "grad_norm": 1.737423530111402, + "learning_rate": 1.0551150471105615e-05, + "loss": 1.0029, + "step": 5177 + }, + { + "epoch": 0.49807618314736435, + "grad_norm": 2.201688461606284, + "learning_rate": 1.0548039756719497e-05, + "loss": 0.9226, + "step": 5178 + }, + { + "epoch": 0.49817237398999614, + "grad_norm": 1.8115300579898916, + "learning_rate": 1.05449289891414e-05, + "loss": 0.9511, + "step": 5179 + }, + { + "epoch": 0.49826856483262794, + "grad_norm": 2.193944555627895, + "learning_rate": 1.054181816867326e-05, + "loss": 1.0341, + "step": 5180 + }, + { + "epoch": 0.49836475567525973, + "grad_norm": 4.829964493648491, + "learning_rate": 1.0538707295616994e-05, + "loss": 0.831, + "step": 5181 + }, + { + "epoch": 0.4984609465178915, + "grad_norm": 1.8786406969551346, + "learning_rate": 1.053559637027455e-05, + "loss": 1.0231, + "step": 5182 + }, + { + "epoch": 0.49855713736052326, + "grad_norm": 2.2274305008984827, + "learning_rate": 1.0532485392947866e-05, + "loss": 0.9117, + "step": 5183 + }, + { + "epoch": 0.49865332820315506, + "grad_norm": 2.206757089199066, + "learning_rate": 1.0529374363938888e-05, + "loss": 0.9979, + "step": 5184 + }, + { + "epoch": 0.49874951904578685, + "grad_norm": 2.1450615824159387, + "learning_rate": 1.0526263283549573e-05, + "loss": 0.9701, + "step": 5185 + }, + { + "epoch": 0.49884570988841864, + "grad_norm": 1.8707734388300536, + "learning_rate": 1.0523152152081875e-05, + "loss": 0.9674, + "step": 5186 + }, + { + "epoch": 0.4989419007310504, + "grad_norm": 1.7765842514150678, + "learning_rate": 1.0520040969837758e-05, + "loss": 0.8751, + "step": 5187 + }, + { + "epoch": 0.4990380915736822, + "grad_norm": 2.0055463156116984, + "learning_rate": 1.051692973711918e-05, + "loss": 0.988, + "step": 5188 + }, + { + "epoch": 0.49913428241631397, + "grad_norm": 2.1575774354613535, + "learning_rate": 1.0513818454228128e-05, + "loss": 0.8965, + "step": 5189 + }, + { + "epoch": 0.49923047325894576, + "grad_norm": 2.3673257054440753, + "learning_rate": 1.0510707121466568e-05, + "loss": 1.0818, + "step": 5190 + }, + { + "epoch": 0.49932666410157756, + "grad_norm": 1.813384172824016, + "learning_rate": 1.0507595739136483e-05, + "loss": 0.9782, + "step": 5191 + }, + { + "epoch": 0.4994228549442093, + "grad_norm": 1.956943644894123, + "learning_rate": 1.0504484307539864e-05, + "loss": 0.9487, + "step": 5192 + }, + { + "epoch": 0.4995190457868411, + "grad_norm": 1.9569604849035356, + "learning_rate": 1.0501372826978698e-05, + "loss": 0.9354, + "step": 5193 + }, + { + "epoch": 0.4996152366294729, + "grad_norm": 2.0444336397673792, + "learning_rate": 1.0498261297754984e-05, + "loss": 1.0341, + "step": 5194 + }, + { + "epoch": 0.4997114274721047, + "grad_norm": 1.7658821890814496, + "learning_rate": 1.0495149720170717e-05, + "loss": 0.8859, + "step": 5195 + }, + { + "epoch": 0.4998076183147364, + "grad_norm": 1.9200417401491292, + "learning_rate": 1.0492038094527907e-05, + "loss": 0.9909, + "step": 5196 + }, + { + "epoch": 0.4999038091573682, + "grad_norm": 1.9406385763110112, + "learning_rate": 1.0488926421128567e-05, + "loss": 0.8802, + "step": 5197 + }, + { + "epoch": 0.5, + "grad_norm": 1.8400303332722514, + "learning_rate": 1.0485814700274706e-05, + "loss": 0.8661, + "step": 5198 + }, + { + "epoch": 0.5000961908426318, + "grad_norm": 2.080576903052638, + "learning_rate": 1.0482702932268346e-05, + "loss": 1.0372, + "step": 5199 + }, + { + "epoch": 0.5001923816852636, + "grad_norm": 1.930866791742584, + "learning_rate": 1.047959111741151e-05, + "loss": 0.8975, + "step": 5200 + }, + { + "epoch": 0.5002885725278954, + "grad_norm": 2.0015927058596397, + "learning_rate": 1.0476479256006225e-05, + "loss": 1.1045, + "step": 5201 + }, + { + "epoch": 0.5003847633705272, + "grad_norm": 2.301153520484906, + "learning_rate": 1.0473367348354529e-05, + "loss": 1.0458, + "step": 5202 + }, + { + "epoch": 0.5004809542131589, + "grad_norm": 2.0537531158790268, + "learning_rate": 1.0470255394758452e-05, + "loss": 0.8614, + "step": 5203 + }, + { + "epoch": 0.5005771450557907, + "grad_norm": 2.132182993271762, + "learning_rate": 1.0467143395520044e-05, + "loss": 0.9591, + "step": 5204 + }, + { + "epoch": 0.5006733358984224, + "grad_norm": 1.8722284406415224, + "learning_rate": 1.0464031350941344e-05, + "loss": 1.0711, + "step": 5205 + }, + { + "epoch": 0.5007695267410542, + "grad_norm": 1.8754063566758732, + "learning_rate": 1.046091926132441e-05, + "loss": 1.0401, + "step": 5206 + }, + { + "epoch": 0.500865717583686, + "grad_norm": 2.103256178924945, + "learning_rate": 1.0457807126971287e-05, + "loss": 0.9516, + "step": 5207 + }, + { + "epoch": 0.5009619084263178, + "grad_norm": 1.8251325034481212, + "learning_rate": 1.0454694948184045e-05, + "loss": 0.9768, + "step": 5208 + }, + { + "epoch": 0.5010580992689496, + "grad_norm": 2.149338444583277, + "learning_rate": 1.0451582725264742e-05, + "loss": 0.9223, + "step": 5209 + }, + { + "epoch": 0.5011542901115814, + "grad_norm": 1.8639387351755565, + "learning_rate": 1.044847045851545e-05, + "loss": 1.028, + "step": 5210 + }, + { + "epoch": 0.5012504809542132, + "grad_norm": 2.096624246670216, + "learning_rate": 1.0445358148238237e-05, + "loss": 0.9983, + "step": 5211 + }, + { + "epoch": 0.5013466717968449, + "grad_norm": 2.143019330790679, + "learning_rate": 1.044224579473518e-05, + "loss": 1.0494, + "step": 5212 + }, + { + "epoch": 0.5014428626394767, + "grad_norm": 2.0615772462077113, + "learning_rate": 1.0439133398308361e-05, + "loss": 1.0337, + "step": 5213 + }, + { + "epoch": 0.5015390534821085, + "grad_norm": 2.166839471526494, + "learning_rate": 1.0436020959259862e-05, + "loss": 1.0287, + "step": 5214 + }, + { + "epoch": 0.5016352443247403, + "grad_norm": 1.8148373811367204, + "learning_rate": 1.043290847789178e-05, + "loss": 1.0035, + "step": 5215 + }, + { + "epoch": 0.5017314351673721, + "grad_norm": 1.7669744986738432, + "learning_rate": 1.0429795954506203e-05, + "loss": 1.0005, + "step": 5216 + }, + { + "epoch": 0.5018276260100039, + "grad_norm": 2.5723901996077463, + "learning_rate": 1.0426683389405225e-05, + "loss": 0.7776, + "step": 5217 + }, + { + "epoch": 0.5019238168526357, + "grad_norm": 2.0803446697545285, + "learning_rate": 1.0423570782890951e-05, + "loss": 0.9314, + "step": 5218 + }, + { + "epoch": 0.5020200076952674, + "grad_norm": 1.9650544347986358, + "learning_rate": 1.0420458135265483e-05, + "loss": 1.0273, + "step": 5219 + }, + { + "epoch": 0.5021161985378992, + "grad_norm": 2.320978741660231, + "learning_rate": 1.0417345446830938e-05, + "loss": 1.0091, + "step": 5220 + }, + { + "epoch": 0.5022123893805309, + "grad_norm": 1.9866245532766231, + "learning_rate": 1.041423271788942e-05, + "loss": 0.9792, + "step": 5221 + }, + { + "epoch": 0.5023085802231627, + "grad_norm": 1.7571542143386605, + "learning_rate": 1.0411119948743052e-05, + "loss": 0.9927, + "step": 5222 + }, + { + "epoch": 0.5024047710657945, + "grad_norm": 2.0080664424851635, + "learning_rate": 1.0408007139693951e-05, + "loss": 1.0739, + "step": 5223 + }, + { + "epoch": 0.5025009619084263, + "grad_norm": 2.252737812421911, + "learning_rate": 1.0404894291044247e-05, + "loss": 1.032, + "step": 5224 + }, + { + "epoch": 0.5025971527510581, + "grad_norm": 2.27579829038379, + "learning_rate": 1.0401781403096065e-05, + "loss": 1.0667, + "step": 5225 + }, + { + "epoch": 0.5026933435936899, + "grad_norm": 1.922242010957432, + "learning_rate": 1.0398668476151538e-05, + "loss": 0.9317, + "step": 5226 + }, + { + "epoch": 0.5027895344363217, + "grad_norm": 2.3246809354662923, + "learning_rate": 1.0395555510512803e-05, + "loss": 0.9837, + "step": 5227 + }, + { + "epoch": 0.5028857252789535, + "grad_norm": 1.9579292230496426, + "learning_rate": 1.0392442506482e-05, + "loss": 0.8359, + "step": 5228 + }, + { + "epoch": 0.5029819161215853, + "grad_norm": 1.7551747283204968, + "learning_rate": 1.0389329464361273e-05, + "loss": 0.9714, + "step": 5229 + }, + { + "epoch": 0.503078106964217, + "grad_norm": 1.9478969278403495, + "learning_rate": 1.038621638445277e-05, + "loss": 0.9026, + "step": 5230 + }, + { + "epoch": 0.5031742978068487, + "grad_norm": 2.001892850003267, + "learning_rate": 1.0383103267058638e-05, + "loss": 1.0325, + "step": 5231 + }, + { + "epoch": 0.5032704886494805, + "grad_norm": 2.1166032362590173, + "learning_rate": 1.037999011248104e-05, + "loss": 1.0787, + "step": 5232 + }, + { + "epoch": 0.5033666794921123, + "grad_norm": 2.2762393426378935, + "learning_rate": 1.0376876921022126e-05, + "loss": 0.9818, + "step": 5233 + }, + { + "epoch": 0.5034628703347441, + "grad_norm": 1.7618988445959063, + "learning_rate": 1.0373763692984062e-05, + "loss": 0.8459, + "step": 5234 + }, + { + "epoch": 0.5035590611773759, + "grad_norm": 2.05413452407222, + "learning_rate": 1.0370650428669016e-05, + "loss": 1.0506, + "step": 5235 + }, + { + "epoch": 0.5036552520200077, + "grad_norm": 1.7028696597981692, + "learning_rate": 1.0367537128379154e-05, + "loss": 0.8914, + "step": 5236 + }, + { + "epoch": 0.5037514428626395, + "grad_norm": 2.044107634121608, + "learning_rate": 1.0364423792416649e-05, + "loss": 0.8519, + "step": 5237 + }, + { + "epoch": 0.5038476337052713, + "grad_norm": 1.9894997503562464, + "learning_rate": 1.0361310421083677e-05, + "loss": 0.9161, + "step": 5238 + }, + { + "epoch": 0.503943824547903, + "grad_norm": 1.9580568297720968, + "learning_rate": 1.0358197014682419e-05, + "loss": 0.893, + "step": 5239 + }, + { + "epoch": 0.5040400153905348, + "grad_norm": 1.8173073328855034, + "learning_rate": 1.0355083573515052e-05, + "loss": 0.9544, + "step": 5240 + }, + { + "epoch": 0.5041362062331666, + "grad_norm": 1.9537052127066448, + "learning_rate": 1.0351970097883773e-05, + "loss": 0.958, + "step": 5241 + }, + { + "epoch": 0.5042323970757984, + "grad_norm": 1.9997162057183744, + "learning_rate": 1.0348856588090764e-05, + "loss": 1.0797, + "step": 5242 + }, + { + "epoch": 0.5043285879184302, + "grad_norm": 1.9309277639436775, + "learning_rate": 1.0345743044438217e-05, + "loss": 1.0145, + "step": 5243 + }, + { + "epoch": 0.504424778761062, + "grad_norm": 1.954677014472462, + "learning_rate": 1.0342629467228331e-05, + "loss": 1.0909, + "step": 5244 + }, + { + "epoch": 0.5045209696036937, + "grad_norm": 2.1983768428913586, + "learning_rate": 1.0339515856763305e-05, + "loss": 0.9307, + "step": 5245 + }, + { + "epoch": 0.5046171604463255, + "grad_norm": 1.7593325553919483, + "learning_rate": 1.0336402213345345e-05, + "loss": 0.8462, + "step": 5246 + }, + { + "epoch": 0.5047133512889573, + "grad_norm": 2.1995008024795712, + "learning_rate": 1.033328853727665e-05, + "loss": 1.0241, + "step": 5247 + }, + { + "epoch": 0.504809542131589, + "grad_norm": 1.7303241178867166, + "learning_rate": 1.0330174828859434e-05, + "loss": 0.9952, + "step": 5248 + }, + { + "epoch": 0.5049057329742208, + "grad_norm": 2.518659745312762, + "learning_rate": 1.0327061088395906e-05, + "loss": 0.8302, + "step": 5249 + }, + { + "epoch": 0.5050019238168526, + "grad_norm": 2.1459107625210083, + "learning_rate": 1.0323947316188288e-05, + "loss": 1.0286, + "step": 5250 + }, + { + "epoch": 0.5050981146594844, + "grad_norm": 1.9434462075337449, + "learning_rate": 1.032083351253879e-05, + "loss": 0.8962, + "step": 5251 + }, + { + "epoch": 0.5051943055021162, + "grad_norm": 1.8379811380089004, + "learning_rate": 1.031771967774964e-05, + "loss": 0.9627, + "step": 5252 + }, + { + "epoch": 0.505290496344748, + "grad_norm": 1.7375982236538499, + "learning_rate": 1.0314605812123057e-05, + "loss": 0.9914, + "step": 5253 + }, + { + "epoch": 0.5053866871873798, + "grad_norm": 2.123357963209762, + "learning_rate": 1.0311491915961271e-05, + "loss": 1.0264, + "step": 5254 + }, + { + "epoch": 0.5054828780300116, + "grad_norm": 1.939455947383127, + "learning_rate": 1.0308377989566517e-05, + "loss": 0.9927, + "step": 5255 + }, + { + "epoch": 0.5055790688726434, + "grad_norm": 2.1024279309343625, + "learning_rate": 1.030526403324102e-05, + "loss": 1.0346, + "step": 5256 + }, + { + "epoch": 0.5056752597152752, + "grad_norm": 1.9894405137197155, + "learning_rate": 1.0302150047287018e-05, + "loss": 1.0502, + "step": 5257 + }, + { + "epoch": 0.5057714505579068, + "grad_norm": 1.9212099035895958, + "learning_rate": 1.0299036032006759e-05, + "loss": 0.8731, + "step": 5258 + }, + { + "epoch": 0.5058676414005386, + "grad_norm": 1.868509391488821, + "learning_rate": 1.0295921987702473e-05, + "loss": 1.0052, + "step": 5259 + }, + { + "epoch": 0.5059638322431704, + "grad_norm": 2.3037298517477836, + "learning_rate": 1.0292807914676412e-05, + "loss": 0.8929, + "step": 5260 + }, + { + "epoch": 0.5060600230858022, + "grad_norm": 1.9816649254515657, + "learning_rate": 1.0289693813230822e-05, + "loss": 1.0468, + "step": 5261 + }, + { + "epoch": 0.506156213928434, + "grad_norm": 2.0428774265067426, + "learning_rate": 1.0286579683667952e-05, + "loss": 0.9338, + "step": 5262 + }, + { + "epoch": 0.5062524047710658, + "grad_norm": 1.8923369826317686, + "learning_rate": 1.0283465526290058e-05, + "loss": 0.955, + "step": 5263 + }, + { + "epoch": 0.5063485956136976, + "grad_norm": 1.9810838164365876, + "learning_rate": 1.0280351341399392e-05, + "loss": 1.0095, + "step": 5264 + }, + { + "epoch": 0.5064447864563294, + "grad_norm": 2.6604179541396484, + "learning_rate": 1.0277237129298216e-05, + "loss": 0.9553, + "step": 5265 + }, + { + "epoch": 0.5065409772989612, + "grad_norm": 2.017350376848706, + "learning_rate": 1.027412289028879e-05, + "loss": 1.0387, + "step": 5266 + }, + { + "epoch": 0.5066371681415929, + "grad_norm": 2.066365328476905, + "learning_rate": 1.0271008624673377e-05, + "loss": 0.9554, + "step": 5267 + }, + { + "epoch": 0.5067333589842247, + "grad_norm": 1.8971573485487352, + "learning_rate": 1.0267894332754243e-05, + "loss": 1.0059, + "step": 5268 + }, + { + "epoch": 0.5068295498268565, + "grad_norm": 1.8103760273837601, + "learning_rate": 1.0264780014833657e-05, + "loss": 0.9407, + "step": 5269 + }, + { + "epoch": 0.5069257406694883, + "grad_norm": 1.708951287072917, + "learning_rate": 1.0261665671213891e-05, + "loss": 0.8819, + "step": 5270 + }, + { + "epoch": 0.50702193151212, + "grad_norm": 1.9512697732679158, + "learning_rate": 1.0258551302197216e-05, + "loss": 0.8874, + "step": 5271 + }, + { + "epoch": 0.5071181223547518, + "grad_norm": 1.8663062181321677, + "learning_rate": 1.0255436908085919e-05, + "loss": 0.9777, + "step": 5272 + }, + { + "epoch": 0.5072143131973836, + "grad_norm": 2.211144101625023, + "learning_rate": 1.0252322489182263e-05, + "loss": 1.0054, + "step": 5273 + }, + { + "epoch": 0.5073105040400154, + "grad_norm": 1.9390044960454538, + "learning_rate": 1.024920804578854e-05, + "loss": 0.9666, + "step": 5274 + }, + { + "epoch": 0.5074066948826472, + "grad_norm": 2.2555657072238096, + "learning_rate": 1.0246093578207029e-05, + "loss": 1.0004, + "step": 5275 + }, + { + "epoch": 0.5075028857252789, + "grad_norm": 2.0103649137265625, + "learning_rate": 1.0242979086740019e-05, + "loss": 0.9302, + "step": 5276 + }, + { + "epoch": 0.5075990765679107, + "grad_norm": 1.9250356237007131, + "learning_rate": 1.0239864571689795e-05, + "loss": 0.9064, + "step": 5277 + }, + { + "epoch": 0.5076952674105425, + "grad_norm": 2.1708036341700994, + "learning_rate": 1.023675003335865e-05, + "loss": 0.9521, + "step": 5278 + }, + { + "epoch": 0.5077914582531743, + "grad_norm": 2.1354884645783785, + "learning_rate": 1.0233635472048875e-05, + "loss": 0.9901, + "step": 5279 + }, + { + "epoch": 0.5078876490958061, + "grad_norm": 2.1477660812887756, + "learning_rate": 1.0230520888062765e-05, + "loss": 1.003, + "step": 5280 + }, + { + "epoch": 0.5079838399384379, + "grad_norm": 2.1290815097852125, + "learning_rate": 1.022740628170262e-05, + "loss": 0.968, + "step": 5281 + }, + { + "epoch": 0.5080800307810697, + "grad_norm": 2.1206041846330037, + "learning_rate": 1.0224291653270739e-05, + "loss": 0.8809, + "step": 5282 + }, + { + "epoch": 0.5081762216237015, + "grad_norm": 2.8654380193637916, + "learning_rate": 1.0221177003069419e-05, + "loss": 1.0557, + "step": 5283 + }, + { + "epoch": 0.5082724124663333, + "grad_norm": 1.9475509844349488, + "learning_rate": 1.0218062331400969e-05, + "loss": 0.9349, + "step": 5284 + }, + { + "epoch": 0.5083686033089649, + "grad_norm": 2.117342931578037, + "learning_rate": 1.0214947638567688e-05, + "loss": 0.9628, + "step": 5285 + }, + { + "epoch": 0.5084647941515967, + "grad_norm": 2.056169791391271, + "learning_rate": 1.0211832924871889e-05, + "loss": 0.957, + "step": 5286 + }, + { + "epoch": 0.5085609849942285, + "grad_norm": 1.9838766317335068, + "learning_rate": 1.0208718190615883e-05, + "loss": 0.9746, + "step": 5287 + }, + { + "epoch": 0.5086571758368603, + "grad_norm": 2.139617718069099, + "learning_rate": 1.0205603436101978e-05, + "loss": 1.0122, + "step": 5288 + }, + { + "epoch": 0.5087533666794921, + "grad_norm": 1.7781244557367444, + "learning_rate": 1.020248866163249e-05, + "loss": 0.9355, + "step": 5289 + }, + { + "epoch": 0.5088495575221239, + "grad_norm": 1.9471529283872375, + "learning_rate": 1.0199373867509734e-05, + "loss": 1.0159, + "step": 5290 + }, + { + "epoch": 0.5089457483647557, + "grad_norm": 1.788039607120968, + "learning_rate": 1.0196259054036029e-05, + "loss": 1.0174, + "step": 5291 + }, + { + "epoch": 0.5090419392073875, + "grad_norm": 2.1681349016918134, + "learning_rate": 1.019314422151369e-05, + "loss": 1.0269, + "step": 5292 + }, + { + "epoch": 0.5091381300500193, + "grad_norm": 1.9041526264067492, + "learning_rate": 1.0190029370245042e-05, + "loss": 0.7979, + "step": 5293 + }, + { + "epoch": 0.509234320892651, + "grad_norm": 2.2146666058520252, + "learning_rate": 1.0186914500532408e-05, + "loss": 1.0244, + "step": 5294 + }, + { + "epoch": 0.5093305117352828, + "grad_norm": 1.8018916510462704, + "learning_rate": 1.0183799612678114e-05, + "loss": 1.0217, + "step": 5295 + }, + { + "epoch": 0.5094267025779146, + "grad_norm": 2.0957854917014735, + "learning_rate": 1.0180684706984483e-05, + "loss": 1.0466, + "step": 5296 + }, + { + "epoch": 0.5095228934205464, + "grad_norm": 2.0409249949133863, + "learning_rate": 1.0177569783753844e-05, + "loss": 0.9563, + "step": 5297 + }, + { + "epoch": 0.5096190842631781, + "grad_norm": 2.196988164417973, + "learning_rate": 1.0174454843288533e-05, + "loss": 0.9621, + "step": 5298 + }, + { + "epoch": 0.5097152751058099, + "grad_norm": 1.8803962311071358, + "learning_rate": 1.0171339885890874e-05, + "loss": 0.9619, + "step": 5299 + }, + { + "epoch": 0.5098114659484417, + "grad_norm": 2.094353022808376, + "learning_rate": 1.0168224911863205e-05, + "loss": 1.0476, + "step": 5300 + }, + { + "epoch": 0.5099076567910735, + "grad_norm": 1.7378997568640997, + "learning_rate": 1.0165109921507858e-05, + "loss": 0.8431, + "step": 5301 + }, + { + "epoch": 0.5100038476337053, + "grad_norm": 1.738876613124847, + "learning_rate": 1.0161994915127173e-05, + "loss": 0.9875, + "step": 5302 + }, + { + "epoch": 0.510100038476337, + "grad_norm": 1.9696880275238202, + "learning_rate": 1.0158879893023488e-05, + "loss": 1.0165, + "step": 5303 + }, + { + "epoch": 0.5101962293189688, + "grad_norm": 1.9926884225229669, + "learning_rate": 1.015576485549914e-05, + "loss": 0.8653, + "step": 5304 + }, + { + "epoch": 0.5102924201616006, + "grad_norm": 2.0594195989554196, + "learning_rate": 1.0152649802856471e-05, + "loss": 1.0636, + "step": 5305 + }, + { + "epoch": 0.5103886110042324, + "grad_norm": 2.1012019414455105, + "learning_rate": 1.0149534735397823e-05, + "loss": 0.9625, + "step": 5306 + }, + { + "epoch": 0.5104848018468642, + "grad_norm": 2.38687525670669, + "learning_rate": 1.0146419653425548e-05, + "loss": 0.9493, + "step": 5307 + }, + { + "epoch": 0.510580992689496, + "grad_norm": 1.9924555860944317, + "learning_rate": 1.0143304557241979e-05, + "loss": 1.0275, + "step": 5308 + }, + { + "epoch": 0.5106771835321278, + "grad_norm": 2.121334671229157, + "learning_rate": 1.0140189447149469e-05, + "loss": 1.0322, + "step": 5309 + }, + { + "epoch": 0.5107733743747596, + "grad_norm": 1.9248668894434164, + "learning_rate": 1.0137074323450372e-05, + "loss": 0.9976, + "step": 5310 + }, + { + "epoch": 0.5108695652173914, + "grad_norm": 2.001633753693386, + "learning_rate": 1.0133959186447021e-05, + "loss": 0.8188, + "step": 5311 + }, + { + "epoch": 0.510965756060023, + "grad_norm": 2.1064723542734645, + "learning_rate": 1.0130844036441787e-05, + "loss": 1.0766, + "step": 5312 + }, + { + "epoch": 0.5110619469026548, + "grad_norm": 2.0253318486005294, + "learning_rate": 1.0127728873737009e-05, + "loss": 0.9175, + "step": 5313 + }, + { + "epoch": 0.5111581377452866, + "grad_norm": 1.9758097016635778, + "learning_rate": 1.0124613698635043e-05, + "loss": 1.0076, + "step": 5314 + }, + { + "epoch": 0.5112543285879184, + "grad_norm": 2.2365745923668183, + "learning_rate": 1.0121498511438249e-05, + "loss": 1.029, + "step": 5315 + }, + { + "epoch": 0.5113505194305502, + "grad_norm": 1.8643187384561866, + "learning_rate": 1.0118383312448973e-05, + "loss": 0.9406, + "step": 5316 + }, + { + "epoch": 0.511446710273182, + "grad_norm": 1.9049059390339398, + "learning_rate": 1.011526810196958e-05, + "loss": 0.977, + "step": 5317 + }, + { + "epoch": 0.5115429011158138, + "grad_norm": 1.9655785521667843, + "learning_rate": 1.0112152880302426e-05, + "loss": 0.9616, + "step": 5318 + }, + { + "epoch": 0.5116390919584456, + "grad_norm": 1.7528258341606435, + "learning_rate": 1.0109037647749868e-05, + "loss": 1.0247, + "step": 5319 + }, + { + "epoch": 0.5117352828010774, + "grad_norm": 1.9943289007153853, + "learning_rate": 1.0105922404614265e-05, + "loss": 1.007, + "step": 5320 + }, + { + "epoch": 0.5118314736437091, + "grad_norm": 1.782206303502714, + "learning_rate": 1.0102807151197987e-05, + "loss": 0.8785, + "step": 5321 + }, + { + "epoch": 0.5119276644863409, + "grad_norm": 2.0222599536456305, + "learning_rate": 1.0099691887803385e-05, + "loss": 0.9914, + "step": 5322 + }, + { + "epoch": 0.5120238553289727, + "grad_norm": 2.472908612725462, + "learning_rate": 1.0096576614732827e-05, + "loss": 0.9481, + "step": 5323 + }, + { + "epoch": 0.5121200461716044, + "grad_norm": 2.1577288770840806, + "learning_rate": 1.0093461332288678e-05, + "loss": 1.0588, + "step": 5324 + }, + { + "epoch": 0.5122162370142362, + "grad_norm": 2.0030579522526235, + "learning_rate": 1.0090346040773303e-05, + "loss": 1.0195, + "step": 5325 + }, + { + "epoch": 0.512312427856868, + "grad_norm": 1.847014415591592, + "learning_rate": 1.0087230740489065e-05, + "loss": 0.8366, + "step": 5326 + }, + { + "epoch": 0.5124086186994998, + "grad_norm": 2.0350681630080105, + "learning_rate": 1.0084115431738333e-05, + "loss": 0.9339, + "step": 5327 + }, + { + "epoch": 0.5125048095421316, + "grad_norm": 2.346570628349163, + "learning_rate": 1.0081000114823473e-05, + "loss": 0.9947, + "step": 5328 + }, + { + "epoch": 0.5126010003847634, + "grad_norm": 2.0278523890076174, + "learning_rate": 1.0077884790046858e-05, + "loss": 0.881, + "step": 5329 + }, + { + "epoch": 0.5126971912273951, + "grad_norm": 1.9131283518372446, + "learning_rate": 1.007476945771085e-05, + "loss": 0.9933, + "step": 5330 + }, + { + "epoch": 0.5127933820700269, + "grad_norm": 2.1999142453027556, + "learning_rate": 1.0071654118117826e-05, + "loss": 0.9725, + "step": 5331 + }, + { + "epoch": 0.5128895729126587, + "grad_norm": 1.756751553766363, + "learning_rate": 1.006853877157015e-05, + "loss": 0.89, + "step": 5332 + }, + { + "epoch": 0.5129857637552905, + "grad_norm": 2.144954867774544, + "learning_rate": 1.0065423418370198e-05, + "loss": 0.9817, + "step": 5333 + }, + { + "epoch": 0.5130819545979223, + "grad_norm": 1.8424474641462227, + "learning_rate": 1.0062308058820337e-05, + "loss": 0.9283, + "step": 5334 + }, + { + "epoch": 0.5131781454405541, + "grad_norm": 1.99773528945835, + "learning_rate": 1.0059192693222943e-05, + "loss": 0.9127, + "step": 5335 + }, + { + "epoch": 0.5132743362831859, + "grad_norm": 1.9862006097458773, + "learning_rate": 1.0056077321880393e-05, + "loss": 0.8854, + "step": 5336 + }, + { + "epoch": 0.5133705271258177, + "grad_norm": 2.176299450107123, + "learning_rate": 1.005296194509505e-05, + "loss": 0.8522, + "step": 5337 + }, + { + "epoch": 0.5134667179684494, + "grad_norm": 1.7781871066112311, + "learning_rate": 1.0049846563169297e-05, + "loss": 0.925, + "step": 5338 + }, + { + "epoch": 0.5135629088110811, + "grad_norm": 1.911967439695259, + "learning_rate": 1.0046731176405507e-05, + "loss": 0.9761, + "step": 5339 + }, + { + "epoch": 0.5136590996537129, + "grad_norm": 2.012949984772879, + "learning_rate": 1.0043615785106051e-05, + "loss": 0.8975, + "step": 5340 + }, + { + "epoch": 0.5137552904963447, + "grad_norm": 1.8708484947695325, + "learning_rate": 1.004050038957331e-05, + "loss": 0.9529, + "step": 5341 + }, + { + "epoch": 0.5138514813389765, + "grad_norm": 2.0346617352403067, + "learning_rate": 1.0037384990109658e-05, + "loss": 0.9281, + "step": 5342 + }, + { + "epoch": 0.5139476721816083, + "grad_norm": 1.7267640241592517, + "learning_rate": 1.0034269587017468e-05, + "loss": 0.9572, + "step": 5343 + }, + { + "epoch": 0.5140438630242401, + "grad_norm": 1.8632215173420523, + "learning_rate": 1.0031154180599123e-05, + "loss": 0.9019, + "step": 5344 + }, + { + "epoch": 0.5141400538668719, + "grad_norm": 1.9477100082898084, + "learning_rate": 1.0028038771156993e-05, + "loss": 1.0628, + "step": 5345 + }, + { + "epoch": 0.5142362447095037, + "grad_norm": 2.014778085330994, + "learning_rate": 1.0024923358993458e-05, + "loss": 0.937, + "step": 5346 + }, + { + "epoch": 0.5143324355521355, + "grad_norm": 2.0036510596655437, + "learning_rate": 1.0021807944410901e-05, + "loss": 0.8344, + "step": 5347 + }, + { + "epoch": 0.5144286263947672, + "grad_norm": 1.9819220607537749, + "learning_rate": 1.0018692527711695e-05, + "loss": 0.98, + "step": 5348 + }, + { + "epoch": 0.514524817237399, + "grad_norm": 1.7444658432035922, + "learning_rate": 1.0015577109198213e-05, + "loss": 0.9278, + "step": 5349 + }, + { + "epoch": 0.5146210080800308, + "grad_norm": 2.1329735449903935, + "learning_rate": 1.0012461689172846e-05, + "loss": 0.9515, + "step": 5350 + }, + { + "epoch": 0.5147171989226625, + "grad_norm": 1.9476455426588597, + "learning_rate": 1.0009346267937958e-05, + "loss": 1.0955, + "step": 5351 + }, + { + "epoch": 0.5148133897652943, + "grad_norm": 2.023318038299039, + "learning_rate": 1.0006230845795937e-05, + "loss": 1.0129, + "step": 5352 + }, + { + "epoch": 0.5149095806079261, + "grad_norm": 2.344643244798358, + "learning_rate": 1.0003115423049159e-05, + "loss": 1.0009, + "step": 5353 + }, + { + "epoch": 0.5150057714505579, + "grad_norm": 2.3527859641066846, + "learning_rate": 1e-05, + "loss": 0.9357, + "step": 5354 + }, + { + "epoch": 0.5151019622931897, + "grad_norm": 1.9427994842828482, + "learning_rate": 9.996884576950844e-06, + "loss": 0.969, + "step": 5355 + }, + { + "epoch": 0.5151981531358215, + "grad_norm": 2.0505857158561573, + "learning_rate": 9.993769154204063e-06, + "loss": 0.8115, + "step": 5356 + }, + { + "epoch": 0.5152943439784533, + "grad_norm": 2.0216009739130936, + "learning_rate": 9.990653732062047e-06, + "loss": 1.0239, + "step": 5357 + }, + { + "epoch": 0.515390534821085, + "grad_norm": 1.931523348224552, + "learning_rate": 9.987538310827159e-06, + "loss": 0.9604, + "step": 5358 + }, + { + "epoch": 0.5154867256637168, + "grad_norm": 2.006234658944619, + "learning_rate": 9.98442289080179e-06, + "loss": 0.923, + "step": 5359 + }, + { + "epoch": 0.5155829165063486, + "grad_norm": 2.065109616805153, + "learning_rate": 9.981307472288308e-06, + "loss": 0.9924, + "step": 5360 + }, + { + "epoch": 0.5156791073489804, + "grad_norm": 2.186019504575707, + "learning_rate": 9.9781920555891e-06, + "loss": 0.9462, + "step": 5361 + }, + { + "epoch": 0.5157752981916122, + "grad_norm": 1.8456450767482344, + "learning_rate": 9.975076641006542e-06, + "loss": 0.9846, + "step": 5362 + }, + { + "epoch": 0.515871489034244, + "grad_norm": 1.9006332083394264, + "learning_rate": 9.971961228843012e-06, + "loss": 0.894, + "step": 5363 + }, + { + "epoch": 0.5159676798768758, + "grad_norm": 2.151524323906138, + "learning_rate": 9.968845819400883e-06, + "loss": 0.9341, + "step": 5364 + }, + { + "epoch": 0.5160638707195075, + "grad_norm": 1.9069100359191105, + "learning_rate": 9.965730412982534e-06, + "loss": 0.9277, + "step": 5365 + }, + { + "epoch": 0.5161600615621393, + "grad_norm": 2.0672048493107997, + "learning_rate": 9.962615009890346e-06, + "loss": 1.0085, + "step": 5366 + }, + { + "epoch": 0.516256252404771, + "grad_norm": 1.9904972875154847, + "learning_rate": 9.959499610426693e-06, + "loss": 0.9704, + "step": 5367 + }, + { + "epoch": 0.5163524432474028, + "grad_norm": 1.9483788350151316, + "learning_rate": 9.956384214893949e-06, + "loss": 0.896, + "step": 5368 + }, + { + "epoch": 0.5164486340900346, + "grad_norm": 1.9411898607383906, + "learning_rate": 9.953268823594498e-06, + "loss": 0.9533, + "step": 5369 + }, + { + "epoch": 0.5165448249326664, + "grad_norm": 1.7703410854495343, + "learning_rate": 9.950153436830707e-06, + "loss": 1.0186, + "step": 5370 + }, + { + "epoch": 0.5166410157752982, + "grad_norm": 1.9498888134191255, + "learning_rate": 9.947038054904952e-06, + "loss": 0.9607, + "step": 5371 + }, + { + "epoch": 0.51673720661793, + "grad_norm": 2.0428572830996217, + "learning_rate": 9.94392267811961e-06, + "loss": 0.9908, + "step": 5372 + }, + { + "epoch": 0.5168333974605618, + "grad_norm": 1.895843662977906, + "learning_rate": 9.940807306777057e-06, + "loss": 0.9831, + "step": 5373 + }, + { + "epoch": 0.5169295883031936, + "grad_norm": 2.011469153116394, + "learning_rate": 9.937691941179665e-06, + "loss": 1.0155, + "step": 5374 + }, + { + "epoch": 0.5170257791458254, + "grad_norm": 2.2072621486099036, + "learning_rate": 9.934576581629807e-06, + "loss": 0.9447, + "step": 5375 + }, + { + "epoch": 0.517121969988457, + "grad_norm": 2.220373301922955, + "learning_rate": 9.931461228429856e-06, + "loss": 1.0046, + "step": 5376 + }, + { + "epoch": 0.5172181608310888, + "grad_norm": 2.1394083491085043, + "learning_rate": 9.928345881882177e-06, + "loss": 0.9781, + "step": 5377 + }, + { + "epoch": 0.5173143516737206, + "grad_norm": 2.146431662579385, + "learning_rate": 9.925230542289151e-06, + "loss": 0.9978, + "step": 5378 + }, + { + "epoch": 0.5174105425163524, + "grad_norm": 2.05299468687482, + "learning_rate": 9.922115209953144e-06, + "loss": 1.0372, + "step": 5379 + }, + { + "epoch": 0.5175067333589842, + "grad_norm": 1.9397861952995834, + "learning_rate": 9.91899988517653e-06, + "loss": 0.9626, + "step": 5380 + }, + { + "epoch": 0.517602924201616, + "grad_norm": 2.0247129817844223, + "learning_rate": 9.91588456826167e-06, + "loss": 0.9271, + "step": 5381 + }, + { + "epoch": 0.5176991150442478, + "grad_norm": 2.227221777206951, + "learning_rate": 9.912769259510938e-06, + "loss": 0.9783, + "step": 5382 + }, + { + "epoch": 0.5177953058868796, + "grad_norm": 2.02604447758254, + "learning_rate": 9.909653959226702e-06, + "loss": 0.9421, + "step": 5383 + }, + { + "epoch": 0.5178914967295114, + "grad_norm": 1.872543268209078, + "learning_rate": 9.906538667711324e-06, + "loss": 0.9067, + "step": 5384 + }, + { + "epoch": 0.5179876875721431, + "grad_norm": 2.2820036288628174, + "learning_rate": 9.903423385267175e-06, + "loss": 1.1178, + "step": 5385 + }, + { + "epoch": 0.5180838784147749, + "grad_norm": 1.998978187793747, + "learning_rate": 9.90030811219662e-06, + "loss": 0.8963, + "step": 5386 + }, + { + "epoch": 0.5181800692574067, + "grad_norm": 1.9171575078380878, + "learning_rate": 9.897192848802018e-06, + "loss": 0.9524, + "step": 5387 + }, + { + "epoch": 0.5182762601000385, + "grad_norm": 2.018052097516678, + "learning_rate": 9.894077595385736e-06, + "loss": 0.8902, + "step": 5388 + }, + { + "epoch": 0.5183724509426703, + "grad_norm": 1.8990824313952972, + "learning_rate": 9.890962352250135e-06, + "loss": 0.9653, + "step": 5389 + }, + { + "epoch": 0.518468641785302, + "grad_norm": 2.24863816153331, + "learning_rate": 9.887847119697577e-06, + "loss": 1.0616, + "step": 5390 + }, + { + "epoch": 0.5185648326279338, + "grad_norm": 2.160363773271566, + "learning_rate": 9.88473189803042e-06, + "loss": 0.9816, + "step": 5391 + }, + { + "epoch": 0.5186610234705656, + "grad_norm": 2.178498422836843, + "learning_rate": 9.881616687551032e-06, + "loss": 1.1245, + "step": 5392 + }, + { + "epoch": 0.5187572143131974, + "grad_norm": 1.8674617935018059, + "learning_rate": 9.878501488561756e-06, + "loss": 0.9888, + "step": 5393 + }, + { + "epoch": 0.5188534051558291, + "grad_norm": 1.8666053839493646, + "learning_rate": 9.875386301364958e-06, + "loss": 0.9083, + "step": 5394 + }, + { + "epoch": 0.5189495959984609, + "grad_norm": 1.9547355338368595, + "learning_rate": 9.872271126262994e-06, + "loss": 0.9805, + "step": 5395 + }, + { + "epoch": 0.5190457868410927, + "grad_norm": 2.104789156056862, + "learning_rate": 9.869155963558215e-06, + "loss": 0.949, + "step": 5396 + }, + { + "epoch": 0.5191419776837245, + "grad_norm": 2.0696356143237256, + "learning_rate": 9.866040813552977e-06, + "loss": 0.9596, + "step": 5397 + }, + { + "epoch": 0.5192381685263563, + "grad_norm": 1.835711543024591, + "learning_rate": 9.862925676549635e-06, + "loss": 0.998, + "step": 5398 + }, + { + "epoch": 0.5193343593689881, + "grad_norm": 1.9816377789519717, + "learning_rate": 9.859810552850533e-06, + "loss": 0.9278, + "step": 5399 + }, + { + "epoch": 0.5194305502116199, + "grad_norm": 1.8599935964987309, + "learning_rate": 9.856695442758023e-06, + "loss": 0.9375, + "step": 5400 + }, + { + "epoch": 0.5195267410542517, + "grad_norm": 2.0554516427557608, + "learning_rate": 9.853580346574456e-06, + "loss": 0.8968, + "step": 5401 + }, + { + "epoch": 0.5196229318968835, + "grad_norm": 1.9422231802328103, + "learning_rate": 9.850465264602175e-06, + "loss": 0.813, + "step": 5402 + }, + { + "epoch": 0.5197191227395151, + "grad_norm": 2.1566427570562987, + "learning_rate": 9.847350197143528e-06, + "loss": 0.9616, + "step": 5403 + }, + { + "epoch": 0.5198153135821469, + "grad_norm": 2.1074960846471957, + "learning_rate": 9.844235144500865e-06, + "loss": 0.9723, + "step": 5404 + }, + { + "epoch": 0.5199115044247787, + "grad_norm": 2.1417330977978777, + "learning_rate": 9.841120106976515e-06, + "loss": 1.0105, + "step": 5405 + }, + { + "epoch": 0.5200076952674105, + "grad_norm": 2.403330791955167, + "learning_rate": 9.83800508487283e-06, + "loss": 1.0087, + "step": 5406 + }, + { + "epoch": 0.5201038861100423, + "grad_norm": 1.6788555617271226, + "learning_rate": 9.834890078492143e-06, + "loss": 0.8449, + "step": 5407 + }, + { + "epoch": 0.5202000769526741, + "grad_norm": 2.118577888458396, + "learning_rate": 9.831775088136797e-06, + "loss": 1.0874, + "step": 5408 + }, + { + "epoch": 0.5202962677953059, + "grad_norm": 1.6145359438756177, + "learning_rate": 9.82866011410913e-06, + "loss": 0.767, + "step": 5409 + }, + { + "epoch": 0.5203924586379377, + "grad_norm": 2.1221988848853788, + "learning_rate": 9.82554515671147e-06, + "loss": 1.0529, + "step": 5410 + }, + { + "epoch": 0.5204886494805695, + "grad_norm": 2.2128589160953314, + "learning_rate": 9.822430216246158e-06, + "loss": 0.9106, + "step": 5411 + }, + { + "epoch": 0.5205848403232012, + "grad_norm": 1.9589524989413793, + "learning_rate": 9.819315293015519e-06, + "loss": 0.9641, + "step": 5412 + }, + { + "epoch": 0.520681031165833, + "grad_norm": 1.9472160472130293, + "learning_rate": 9.81620038732189e-06, + "loss": 0.9765, + "step": 5413 + }, + { + "epoch": 0.5207772220084648, + "grad_norm": 1.9208107983113072, + "learning_rate": 9.813085499467594e-06, + "loss": 1.0729, + "step": 5414 + }, + { + "epoch": 0.5208734128510966, + "grad_norm": 2.1744248841040705, + "learning_rate": 9.809970629754963e-06, + "loss": 1.0435, + "step": 5415 + }, + { + "epoch": 0.5209696036937284, + "grad_norm": 2.076025461668634, + "learning_rate": 9.806855778486314e-06, + "loss": 1.0537, + "step": 5416 + }, + { + "epoch": 0.5210657945363601, + "grad_norm": 1.785581560287918, + "learning_rate": 9.803740945963975e-06, + "loss": 0.8596, + "step": 5417 + }, + { + "epoch": 0.5211619853789919, + "grad_norm": 1.9182214919693263, + "learning_rate": 9.800626132490268e-06, + "loss": 0.9843, + "step": 5418 + }, + { + "epoch": 0.5212581762216237, + "grad_norm": 1.9722116604298028, + "learning_rate": 9.797511338367513e-06, + "loss": 1.0546, + "step": 5419 + }, + { + "epoch": 0.5213543670642555, + "grad_norm": 1.8048717042565885, + "learning_rate": 9.794396563898022e-06, + "loss": 0.9473, + "step": 5420 + }, + { + "epoch": 0.5214505579068872, + "grad_norm": 1.9532801789510599, + "learning_rate": 9.79128180938412e-06, + "loss": 0.9561, + "step": 5421 + }, + { + "epoch": 0.521546748749519, + "grad_norm": 1.8402608754426442, + "learning_rate": 9.788167075128113e-06, + "loss": 0.9623, + "step": 5422 + }, + { + "epoch": 0.5216429395921508, + "grad_norm": 2.1374047464775914, + "learning_rate": 9.785052361432315e-06, + "loss": 1.0095, + "step": 5423 + }, + { + "epoch": 0.5217391304347826, + "grad_norm": 2.216889167030151, + "learning_rate": 9.781937668599035e-06, + "loss": 1.0369, + "step": 5424 + }, + { + "epoch": 0.5218353212774144, + "grad_norm": 1.8880191928022492, + "learning_rate": 9.778822996930583e-06, + "loss": 1.0162, + "step": 5425 + }, + { + "epoch": 0.5219315121200462, + "grad_norm": 2.011873124146862, + "learning_rate": 9.775708346729263e-06, + "loss": 1.0036, + "step": 5426 + }, + { + "epoch": 0.522027702962678, + "grad_norm": 1.9736869804493857, + "learning_rate": 9.772593718297383e-06, + "loss": 1.007, + "step": 5427 + }, + { + "epoch": 0.5221238938053098, + "grad_norm": 1.8734436751383092, + "learning_rate": 9.769479111937238e-06, + "loss": 1.0038, + "step": 5428 + }, + { + "epoch": 0.5222200846479416, + "grad_norm": 1.930683371924382, + "learning_rate": 9.766364527951126e-06, + "loss": 0.9062, + "step": 5429 + }, + { + "epoch": 0.5223162754905732, + "grad_norm": 2.0930551042862784, + "learning_rate": 9.763249966641352e-06, + "loss": 0.95, + "step": 5430 + }, + { + "epoch": 0.522412466333205, + "grad_norm": 2.0816036675053646, + "learning_rate": 9.760135428310206e-06, + "loss": 0.9873, + "step": 5431 + }, + { + "epoch": 0.5225086571758368, + "grad_norm": 1.7482479678254756, + "learning_rate": 9.757020913259986e-06, + "loss": 0.9308, + "step": 5432 + }, + { + "epoch": 0.5226048480184686, + "grad_norm": 1.9722394599852349, + "learning_rate": 9.753906421792973e-06, + "loss": 0.9165, + "step": 5433 + }, + { + "epoch": 0.5227010388611004, + "grad_norm": 1.8155493426872857, + "learning_rate": 9.750791954211464e-06, + "loss": 0.9874, + "step": 5434 + }, + { + "epoch": 0.5227972297037322, + "grad_norm": 1.9351866011488512, + "learning_rate": 9.74767751081774e-06, + "loss": 0.9868, + "step": 5435 + }, + { + "epoch": 0.522893420546364, + "grad_norm": 1.9745509184197532, + "learning_rate": 9.744563091914085e-06, + "loss": 0.8495, + "step": 5436 + }, + { + "epoch": 0.5229896113889958, + "grad_norm": 1.9076473912791552, + "learning_rate": 9.741448697802783e-06, + "loss": 0.9755, + "step": 5437 + }, + { + "epoch": 0.5230858022316276, + "grad_norm": 2.1831119156235683, + "learning_rate": 9.738334328786114e-06, + "loss": 0.9326, + "step": 5438 + }, + { + "epoch": 0.5231819930742593, + "grad_norm": 2.1160772858330854, + "learning_rate": 9.735219985166348e-06, + "loss": 0.919, + "step": 5439 + }, + { + "epoch": 0.5232781839168911, + "grad_norm": 1.7970902870699752, + "learning_rate": 9.732105667245759e-06, + "loss": 1.0666, + "step": 5440 + }, + { + "epoch": 0.5233743747595229, + "grad_norm": 1.7834126192358242, + "learning_rate": 9.728991375326626e-06, + "loss": 0.945, + "step": 5441 + }, + { + "epoch": 0.5234705656021547, + "grad_norm": 2.1032825383140654, + "learning_rate": 9.725877109711212e-06, + "loss": 0.9485, + "step": 5442 + }, + { + "epoch": 0.5235667564447865, + "grad_norm": 2.242632649497909, + "learning_rate": 9.722762870701783e-06, + "loss": 1.0069, + "step": 5443 + }, + { + "epoch": 0.5236629472874182, + "grad_norm": 1.828864590702526, + "learning_rate": 9.719648658600611e-06, + "loss": 0.9641, + "step": 5444 + }, + { + "epoch": 0.52375913813005, + "grad_norm": 2.1263123100722963, + "learning_rate": 9.716534473709944e-06, + "loss": 1.0063, + "step": 5445 + }, + { + "epoch": 0.5238553289726818, + "grad_norm": 2.106890360587902, + "learning_rate": 9.71342031633205e-06, + "loss": 0.9923, + "step": 5446 + }, + { + "epoch": 0.5239515198153136, + "grad_norm": 2.101212840753378, + "learning_rate": 9.71030618676918e-06, + "loss": 0.9695, + "step": 5447 + }, + { + "epoch": 0.5240477106579453, + "grad_norm": 2.3033914973642546, + "learning_rate": 9.70719208532359e-06, + "loss": 1.0406, + "step": 5448 + }, + { + "epoch": 0.5241439015005771, + "grad_norm": 1.8373084375462734, + "learning_rate": 9.704078012297528e-06, + "loss": 0.8659, + "step": 5449 + }, + { + "epoch": 0.5242400923432089, + "grad_norm": 1.9353286669421141, + "learning_rate": 9.700963967993246e-06, + "loss": 1.0636, + "step": 5450 + }, + { + "epoch": 0.5243362831858407, + "grad_norm": 1.9899166196176643, + "learning_rate": 9.697849952712984e-06, + "loss": 1.0702, + "step": 5451 + }, + { + "epoch": 0.5244324740284725, + "grad_norm": 1.8759832161568373, + "learning_rate": 9.694735966758982e-06, + "loss": 0.9857, + "step": 5452 + }, + { + "epoch": 0.5245286648711043, + "grad_norm": 2.139661319263725, + "learning_rate": 9.691622010433487e-06, + "loss": 1.0026, + "step": 5453 + }, + { + "epoch": 0.5246248557137361, + "grad_norm": 1.75698529418727, + "learning_rate": 9.688508084038729e-06, + "loss": 0.9609, + "step": 5454 + }, + { + "epoch": 0.5247210465563679, + "grad_norm": 1.7469299201077948, + "learning_rate": 9.685394187876948e-06, + "loss": 0.8357, + "step": 5455 + }, + { + "epoch": 0.5248172373989997, + "grad_norm": 2.063623934696085, + "learning_rate": 9.682280322250365e-06, + "loss": 1.0281, + "step": 5456 + }, + { + "epoch": 0.5249134282416313, + "grad_norm": 2.19020117731007, + "learning_rate": 9.679166487461213e-06, + "loss": 1.048, + "step": 5457 + }, + { + "epoch": 0.5250096190842631, + "grad_norm": 2.003631403749758, + "learning_rate": 9.676052683811715e-06, + "loss": 1.0178, + "step": 5458 + }, + { + "epoch": 0.5251058099268949, + "grad_norm": 2.0912220680953015, + "learning_rate": 9.672938911604096e-06, + "loss": 1.0327, + "step": 5459 + }, + { + "epoch": 0.5252020007695267, + "grad_norm": 1.98611804994526, + "learning_rate": 9.669825171140568e-06, + "loss": 0.8974, + "step": 5460 + }, + { + "epoch": 0.5252981916121585, + "grad_norm": 1.8424339341464147, + "learning_rate": 9.666711462723354e-06, + "loss": 0.8112, + "step": 5461 + }, + { + "epoch": 0.5253943824547903, + "grad_norm": 1.9950647277173796, + "learning_rate": 9.66359778665466e-06, + "loss": 0.9827, + "step": 5462 + }, + { + "epoch": 0.5254905732974221, + "grad_norm": 1.9931843081423304, + "learning_rate": 9.660484143236698e-06, + "loss": 1.0238, + "step": 5463 + }, + { + "epoch": 0.5255867641400539, + "grad_norm": 1.8240288592044245, + "learning_rate": 9.657370532771672e-06, + "loss": 0.8497, + "step": 5464 + }, + { + "epoch": 0.5256829549826857, + "grad_norm": 2.2330985044886917, + "learning_rate": 9.654256955561786e-06, + "loss": 1.0943, + "step": 5465 + }, + { + "epoch": 0.5257791458253175, + "grad_norm": 2.0673266533114805, + "learning_rate": 9.651143411909241e-06, + "loss": 1.0318, + "step": 5466 + }, + { + "epoch": 0.5258753366679492, + "grad_norm": 1.8775793869909412, + "learning_rate": 9.648029902116232e-06, + "loss": 0.9562, + "step": 5467 + }, + { + "epoch": 0.525971527510581, + "grad_norm": 2.2238100034637083, + "learning_rate": 9.64491642648495e-06, + "loss": 1.0481, + "step": 5468 + }, + { + "epoch": 0.5260677183532128, + "grad_norm": 2.108232243599161, + "learning_rate": 9.641802985317585e-06, + "loss": 0.9875, + "step": 5469 + }, + { + "epoch": 0.5261639091958445, + "grad_norm": 1.8036514523340075, + "learning_rate": 9.638689578916326e-06, + "loss": 1.0639, + "step": 5470 + }, + { + "epoch": 0.5262601000384763, + "grad_norm": 1.8819721161587948, + "learning_rate": 9.635576207583354e-06, + "loss": 0.9452, + "step": 5471 + }, + { + "epoch": 0.5263562908811081, + "grad_norm": 1.9975529092899995, + "learning_rate": 9.632462871620847e-06, + "loss": 0.9545, + "step": 5472 + }, + { + "epoch": 0.5264524817237399, + "grad_norm": 1.7967630918834303, + "learning_rate": 9.629349571330986e-06, + "loss": 0.9105, + "step": 5473 + }, + { + "epoch": 0.5265486725663717, + "grad_norm": 2.2251221187779784, + "learning_rate": 9.62623630701594e-06, + "loss": 0.8776, + "step": 5474 + }, + { + "epoch": 0.5266448634090035, + "grad_norm": 1.83743549639502, + "learning_rate": 9.623123078977878e-06, + "loss": 0.9397, + "step": 5475 + }, + { + "epoch": 0.5267410542516352, + "grad_norm": 1.8505222233566616, + "learning_rate": 9.620009887518963e-06, + "loss": 0.9582, + "step": 5476 + }, + { + "epoch": 0.526837245094267, + "grad_norm": 1.831529448425279, + "learning_rate": 9.616896732941363e-06, + "loss": 0.8482, + "step": 5477 + }, + { + "epoch": 0.5269334359368988, + "grad_norm": 1.8604116034148925, + "learning_rate": 9.613783615547233e-06, + "loss": 0.9154, + "step": 5478 + }, + { + "epoch": 0.5270296267795306, + "grad_norm": 1.937450575081648, + "learning_rate": 9.610670535638732e-06, + "loss": 0.9343, + "step": 5479 + }, + { + "epoch": 0.5271258176221624, + "grad_norm": 2.2845427718000657, + "learning_rate": 9.607557493518006e-06, + "loss": 1.0117, + "step": 5480 + }, + { + "epoch": 0.5272220084647942, + "grad_norm": 2.065515622030749, + "learning_rate": 9.604444489487199e-06, + "loss": 1.0095, + "step": 5481 + }, + { + "epoch": 0.527318199307426, + "grad_norm": 1.7753877432342355, + "learning_rate": 9.601331523848464e-06, + "loss": 0.919, + "step": 5482 + }, + { + "epoch": 0.5274143901500578, + "grad_norm": 1.9891817250630048, + "learning_rate": 9.598218596903935e-06, + "loss": 0.9542, + "step": 5483 + }, + { + "epoch": 0.5275105809926895, + "grad_norm": 1.9838166772106038, + "learning_rate": 9.595105708955758e-06, + "loss": 0.9613, + "step": 5484 + }, + { + "epoch": 0.5276067718353212, + "grad_norm": 2.3554658147405183, + "learning_rate": 9.59199286030605e-06, + "loss": 0.9513, + "step": 5485 + }, + { + "epoch": 0.527702962677953, + "grad_norm": 1.6563734722677557, + "learning_rate": 9.588880051256951e-06, + "loss": 0.9261, + "step": 5486 + }, + { + "epoch": 0.5277991535205848, + "grad_norm": 1.9000990793003472, + "learning_rate": 9.585767282110583e-06, + "loss": 1.0513, + "step": 5487 + }, + { + "epoch": 0.5278953443632166, + "grad_norm": 2.0208907370274365, + "learning_rate": 9.582654553169064e-06, + "loss": 0.9295, + "step": 5488 + }, + { + "epoch": 0.5279915352058484, + "grad_norm": 2.151237246804005, + "learning_rate": 9.579541864734515e-06, + "loss": 1.0397, + "step": 5489 + }, + { + "epoch": 0.5280877260484802, + "grad_norm": 2.010753098743075, + "learning_rate": 9.576429217109054e-06, + "loss": 1.0464, + "step": 5490 + }, + { + "epoch": 0.528183916891112, + "grad_norm": 2.0732023405064384, + "learning_rate": 9.57331661059478e-06, + "loss": 0.9109, + "step": 5491 + }, + { + "epoch": 0.5282801077337438, + "grad_norm": 2.084026081467257, + "learning_rate": 9.5702040454938e-06, + "loss": 1.0627, + "step": 5492 + }, + { + "epoch": 0.5283762985763756, + "grad_norm": 2.063081776991347, + "learning_rate": 9.567091522108221e-06, + "loss": 0.9228, + "step": 5493 + }, + { + "epoch": 0.5284724894190073, + "grad_norm": 1.949569206008309, + "learning_rate": 9.563979040740138e-06, + "loss": 1.0002, + "step": 5494 + }, + { + "epoch": 0.5285686802616391, + "grad_norm": 2.3441614940980116, + "learning_rate": 9.56086660169164e-06, + "loss": 0.9858, + "step": 5495 + }, + { + "epoch": 0.5286648711042709, + "grad_norm": 2.4435472616323572, + "learning_rate": 9.557754205264826e-06, + "loss": 0.9245, + "step": 5496 + }, + { + "epoch": 0.5287610619469026, + "grad_norm": 2.047188408279119, + "learning_rate": 9.554641851761768e-06, + "loss": 0.981, + "step": 5497 + }, + { + "epoch": 0.5288572527895344, + "grad_norm": 2.04392884565389, + "learning_rate": 9.551529541484554e-06, + "loss": 0.8329, + "step": 5498 + }, + { + "epoch": 0.5289534436321662, + "grad_norm": 2.1858611924455325, + "learning_rate": 9.54841727473526e-06, + "loss": 0.976, + "step": 5499 + }, + { + "epoch": 0.529049634474798, + "grad_norm": 2.0884038483958607, + "learning_rate": 9.545305051815957e-06, + "loss": 0.9935, + "step": 5500 + }, + { + "epoch": 0.5291458253174298, + "grad_norm": 2.0433125489210524, + "learning_rate": 9.542192873028713e-06, + "loss": 0.9981, + "step": 5501 + }, + { + "epoch": 0.5292420161600616, + "grad_norm": 1.9700300258295604, + "learning_rate": 9.539080738675597e-06, + "loss": 0.9188, + "step": 5502 + }, + { + "epoch": 0.5293382070026933, + "grad_norm": 1.7876376784802086, + "learning_rate": 9.53596864905866e-06, + "loss": 1.0052, + "step": 5503 + }, + { + "epoch": 0.5294343978453251, + "grad_norm": 2.1728976957338455, + "learning_rate": 9.53285660447996e-06, + "loss": 1.0397, + "step": 5504 + }, + { + "epoch": 0.5295305886879569, + "grad_norm": 2.0977624684009735, + "learning_rate": 9.52974460524155e-06, + "loss": 0.9518, + "step": 5505 + }, + { + "epoch": 0.5296267795305887, + "grad_norm": 2.1964446508170217, + "learning_rate": 9.526632651645476e-06, + "loss": 0.9123, + "step": 5506 + }, + { + "epoch": 0.5297229703732205, + "grad_norm": 1.926964103766058, + "learning_rate": 9.523520743993779e-06, + "loss": 1.0465, + "step": 5507 + }, + { + "epoch": 0.5298191612158523, + "grad_norm": 2.0365463925807705, + "learning_rate": 9.520408882588497e-06, + "loss": 0.9521, + "step": 5508 + }, + { + "epoch": 0.5299153520584841, + "grad_norm": 1.9064149839082216, + "learning_rate": 9.517297067731658e-06, + "loss": 0.9863, + "step": 5509 + }, + { + "epoch": 0.5300115429011159, + "grad_norm": 1.9421185815786328, + "learning_rate": 9.514185299725299e-06, + "loss": 0.8178, + "step": 5510 + }, + { + "epoch": 0.5301077337437476, + "grad_norm": 2.409580181149228, + "learning_rate": 9.511073578871438e-06, + "loss": 0.8748, + "step": 5511 + }, + { + "epoch": 0.5302039245863793, + "grad_norm": 1.8584092326568211, + "learning_rate": 9.507961905472093e-06, + "loss": 0.8705, + "step": 5512 + }, + { + "epoch": 0.5303001154290111, + "grad_norm": 1.8448563226021215, + "learning_rate": 9.504850279829287e-06, + "loss": 0.9031, + "step": 5513 + }, + { + "epoch": 0.5303963062716429, + "grad_norm": 2.3256410029679397, + "learning_rate": 9.501738702245023e-06, + "loss": 1.0242, + "step": 5514 + }, + { + "epoch": 0.5304924971142747, + "grad_norm": 2.23167820479621, + "learning_rate": 9.498627173021306e-06, + "loss": 0.9324, + "step": 5515 + }, + { + "epoch": 0.5305886879569065, + "grad_norm": 2.205185929552201, + "learning_rate": 9.495515692460138e-06, + "loss": 0.8463, + "step": 5516 + }, + { + "epoch": 0.5306848787995383, + "grad_norm": 1.9230418847268484, + "learning_rate": 9.492404260863518e-06, + "loss": 0.9613, + "step": 5517 + }, + { + "epoch": 0.5307810696421701, + "grad_norm": 2.0241067350649464, + "learning_rate": 9.489292878533436e-06, + "loss": 1.0774, + "step": 5518 + }, + { + "epoch": 0.5308772604848019, + "grad_norm": 1.8546747971068926, + "learning_rate": 9.486181545771877e-06, + "loss": 0.8619, + "step": 5519 + }, + { + "epoch": 0.5309734513274337, + "grad_norm": 2.2241089476412004, + "learning_rate": 9.483070262880823e-06, + "loss": 0.9281, + "step": 5520 + }, + { + "epoch": 0.5310696421700654, + "grad_norm": 1.8244264832305648, + "learning_rate": 9.479959030162247e-06, + "loss": 0.9918, + "step": 5521 + }, + { + "epoch": 0.5311658330126972, + "grad_norm": 2.030644165345185, + "learning_rate": 9.476847847918126e-06, + "loss": 0.9487, + "step": 5522 + }, + { + "epoch": 0.531262023855329, + "grad_norm": 2.331632037665147, + "learning_rate": 9.473736716450427e-06, + "loss": 0.9994, + "step": 5523 + }, + { + "epoch": 0.5313582146979607, + "grad_norm": 1.961588981139043, + "learning_rate": 9.47062563606111e-06, + "loss": 0.9275, + "step": 5524 + }, + { + "epoch": 0.5314544055405925, + "grad_norm": 2.0971472990987943, + "learning_rate": 9.467514607052137e-06, + "loss": 0.9892, + "step": 5525 + }, + { + "epoch": 0.5315505963832243, + "grad_norm": 1.67581447114505, + "learning_rate": 9.464403629725454e-06, + "loss": 0.7994, + "step": 5526 + }, + { + "epoch": 0.5316467872258561, + "grad_norm": 2.0517160004274184, + "learning_rate": 9.461292704383009e-06, + "loss": 1.0956, + "step": 5527 + }, + { + "epoch": 0.5317429780684879, + "grad_norm": 1.9223220343456155, + "learning_rate": 9.458181831326744e-06, + "loss": 1.0213, + "step": 5528 + }, + { + "epoch": 0.5318391689111197, + "grad_norm": 1.8088944991415004, + "learning_rate": 9.4550710108586e-06, + "loss": 0.8958, + "step": 5529 + }, + { + "epoch": 0.5319353597537514, + "grad_norm": 2.0001358088577263, + "learning_rate": 9.451960243280506e-06, + "loss": 1.0076, + "step": 5530 + }, + { + "epoch": 0.5320315505963832, + "grad_norm": 2.4694637274590385, + "learning_rate": 9.44884952889439e-06, + "loss": 1.0323, + "step": 5531 + }, + { + "epoch": 0.532127741439015, + "grad_norm": 2.1679613211891784, + "learning_rate": 9.44573886800217e-06, + "loss": 0.9701, + "step": 5532 + }, + { + "epoch": 0.5322239322816468, + "grad_norm": 2.050302182446461, + "learning_rate": 9.442628260905767e-06, + "loss": 1.0272, + "step": 5533 + }, + { + "epoch": 0.5323201231242786, + "grad_norm": 1.8607029243119422, + "learning_rate": 9.43951770790709e-06, + "loss": 0.8946, + "step": 5534 + }, + { + "epoch": 0.5324163139669104, + "grad_norm": 1.9740833208475892, + "learning_rate": 9.436407209308042e-06, + "loss": 1.0459, + "step": 5535 + }, + { + "epoch": 0.5325125048095422, + "grad_norm": 1.9118658266779631, + "learning_rate": 9.433296765410534e-06, + "loss": 1.0186, + "step": 5536 + }, + { + "epoch": 0.532608695652174, + "grad_norm": 2.1460121154259166, + "learning_rate": 9.430186376516448e-06, + "loss": 1.0354, + "step": 5537 + }, + { + "epoch": 0.5327048864948057, + "grad_norm": 1.9947915236851321, + "learning_rate": 9.427076042927683e-06, + "loss": 0.9894, + "step": 5538 + }, + { + "epoch": 0.5328010773374374, + "grad_norm": 2.6404448403167255, + "learning_rate": 9.42396576494612e-06, + "loss": 0.8006, + "step": 5539 + }, + { + "epoch": 0.5328972681800692, + "grad_norm": 2.1194983199931468, + "learning_rate": 9.420855542873638e-06, + "loss": 1.0551, + "step": 5540 + }, + { + "epoch": 0.532993459022701, + "grad_norm": 2.1901331434569324, + "learning_rate": 9.417745377012114e-06, + "loss": 0.9194, + "step": 5541 + }, + { + "epoch": 0.5330896498653328, + "grad_norm": 2.07944981047003, + "learning_rate": 9.414635267663416e-06, + "loss": 0.9544, + "step": 5542 + }, + { + "epoch": 0.5331858407079646, + "grad_norm": 1.8131469253004198, + "learning_rate": 9.411525215129404e-06, + "loss": 0.8976, + "step": 5543 + }, + { + "epoch": 0.5332820315505964, + "grad_norm": 1.8367617025650125, + "learning_rate": 9.408415219711934e-06, + "loss": 0.9228, + "step": 5544 + }, + { + "epoch": 0.5333782223932282, + "grad_norm": 2.18353507082447, + "learning_rate": 9.405305281712865e-06, + "loss": 1.0275, + "step": 5545 + }, + { + "epoch": 0.53347441323586, + "grad_norm": 2.4654782580914043, + "learning_rate": 9.402195401434036e-06, + "loss": 1.0514, + "step": 5546 + }, + { + "epoch": 0.5335706040784918, + "grad_norm": 1.7973237062420102, + "learning_rate": 9.399085579177289e-06, + "loss": 0.9498, + "step": 5547 + }, + { + "epoch": 0.5336667949211235, + "grad_norm": 2.092662294426833, + "learning_rate": 9.395975815244468e-06, + "loss": 0.8874, + "step": 5548 + }, + { + "epoch": 0.5337629857637552, + "grad_norm": 1.8740544320588912, + "learning_rate": 9.392866109937387e-06, + "loss": 0.8512, + "step": 5549 + }, + { + "epoch": 0.533859176606387, + "grad_norm": 2.394982844776059, + "learning_rate": 9.389756463557878e-06, + "loss": 1.0114, + "step": 5550 + }, + { + "epoch": 0.5339553674490188, + "grad_norm": 1.9785231234917686, + "learning_rate": 9.386646876407762e-06, + "loss": 1.0302, + "step": 5551 + }, + { + "epoch": 0.5340515582916506, + "grad_norm": 1.7964453366221007, + "learning_rate": 9.383537348788844e-06, + "loss": 1.0803, + "step": 5552 + }, + { + "epoch": 0.5341477491342824, + "grad_norm": 2.098091582881405, + "learning_rate": 9.380427881002936e-06, + "loss": 0.9586, + "step": 5553 + }, + { + "epoch": 0.5342439399769142, + "grad_norm": 2.2992220743545864, + "learning_rate": 9.377318473351838e-06, + "loss": 1.0171, + "step": 5554 + }, + { + "epoch": 0.534340130819546, + "grad_norm": 1.8365151950395664, + "learning_rate": 9.374209126137343e-06, + "loss": 0.9261, + "step": 5555 + }, + { + "epoch": 0.5344363216621778, + "grad_norm": 2.054874801078141, + "learning_rate": 9.371099839661238e-06, + "loss": 0.9207, + "step": 5556 + }, + { + "epoch": 0.5345325125048095, + "grad_norm": 2.3572004120306937, + "learning_rate": 9.367990614225312e-06, + "loss": 1.0704, + "step": 5557 + }, + { + "epoch": 0.5346287033474413, + "grad_norm": 1.7379573037557812, + "learning_rate": 9.364881450131335e-06, + "loss": 0.9109, + "step": 5558 + }, + { + "epoch": 0.5347248941900731, + "grad_norm": 1.722334461970924, + "learning_rate": 9.361772347681086e-06, + "loss": 0.9843, + "step": 5559 + }, + { + "epoch": 0.5348210850327049, + "grad_norm": 2.0763821027563707, + "learning_rate": 9.358663307176323e-06, + "loss": 0.9803, + "step": 5560 + }, + { + "epoch": 0.5349172758753367, + "grad_norm": 2.315727446465209, + "learning_rate": 9.355554328918807e-06, + "loss": 0.8671, + "step": 5561 + }, + { + "epoch": 0.5350134667179685, + "grad_norm": 2.069480950105315, + "learning_rate": 9.352445413210294e-06, + "loss": 0.9709, + "step": 5562 + }, + { + "epoch": 0.5351096575606002, + "grad_norm": 1.8904687747959759, + "learning_rate": 9.349336560352526e-06, + "loss": 0.9248, + "step": 5563 + }, + { + "epoch": 0.535205848403232, + "grad_norm": 1.897223670382873, + "learning_rate": 9.346227770647251e-06, + "loss": 1.0033, + "step": 5564 + }, + { + "epoch": 0.5353020392458638, + "grad_norm": 1.9651676051438334, + "learning_rate": 9.343119044396201e-06, + "loss": 0.9876, + "step": 5565 + }, + { + "epoch": 0.5353982300884956, + "grad_norm": 1.9413299590344548, + "learning_rate": 9.3400103819011e-06, + "loss": 0.8744, + "step": 5566 + }, + { + "epoch": 0.5354944209311273, + "grad_norm": 1.9930389022071213, + "learning_rate": 9.336901783463677e-06, + "loss": 0.8079, + "step": 5567 + }, + { + "epoch": 0.5355906117737591, + "grad_norm": 1.9999595058156372, + "learning_rate": 9.33379324938564e-06, + "loss": 0.8569, + "step": 5568 + }, + { + "epoch": 0.5356868026163909, + "grad_norm": 2.4476056579998793, + "learning_rate": 9.330684779968709e-06, + "loss": 1.0096, + "step": 5569 + }, + { + "epoch": 0.5357829934590227, + "grad_norm": 2.024557650475253, + "learning_rate": 9.327576375514582e-06, + "loss": 0.9173, + "step": 5570 + }, + { + "epoch": 0.5358791843016545, + "grad_norm": 2.1286327972742973, + "learning_rate": 9.32446803632496e-06, + "loss": 0.9414, + "step": 5571 + }, + { + "epoch": 0.5359753751442863, + "grad_norm": 2.1392319639464246, + "learning_rate": 9.321359762701527e-06, + "loss": 0.9805, + "step": 5572 + }, + { + "epoch": 0.5360715659869181, + "grad_norm": 2.313348028395674, + "learning_rate": 9.318251554945973e-06, + "loss": 0.999, + "step": 5573 + }, + { + "epoch": 0.5361677568295499, + "grad_norm": 1.7989776389325374, + "learning_rate": 9.315143413359975e-06, + "loss": 0.945, + "step": 5574 + }, + { + "epoch": 0.5362639476721817, + "grad_norm": 1.8522646784970354, + "learning_rate": 9.312035338245207e-06, + "loss": 0.991, + "step": 5575 + }, + { + "epoch": 0.5363601385148133, + "grad_norm": 2.0371582261498022, + "learning_rate": 9.308927329903333e-06, + "loss": 0.9528, + "step": 5576 + }, + { + "epoch": 0.5364563293574451, + "grad_norm": 1.908264470636764, + "learning_rate": 9.305819388636013e-06, + "loss": 0.894, + "step": 5577 + }, + { + "epoch": 0.5365525202000769, + "grad_norm": 2.6013492996016225, + "learning_rate": 9.302711514744897e-06, + "loss": 1.0697, + "step": 5578 + }, + { + "epoch": 0.5366487110427087, + "grad_norm": 2.0180479453021842, + "learning_rate": 9.299603708531632e-06, + "loss": 1.0342, + "step": 5579 + }, + { + "epoch": 0.5367449018853405, + "grad_norm": 1.9302612396623682, + "learning_rate": 9.296495970297855e-06, + "loss": 1.0114, + "step": 5580 + }, + { + "epoch": 0.5368410927279723, + "grad_norm": 2.044324310228146, + "learning_rate": 9.293388300345205e-06, + "loss": 0.9818, + "step": 5581 + }, + { + "epoch": 0.5369372835706041, + "grad_norm": 2.5329943331900133, + "learning_rate": 9.290280698975307e-06, + "loss": 0.8506, + "step": 5582 + }, + { + "epoch": 0.5370334744132359, + "grad_norm": 1.886786567739531, + "learning_rate": 9.287173166489773e-06, + "loss": 1.0313, + "step": 5583 + }, + { + "epoch": 0.5371296652558677, + "grad_norm": 1.9511331985203408, + "learning_rate": 9.284065703190221e-06, + "loss": 0.9285, + "step": 5584 + }, + { + "epoch": 0.5372258560984994, + "grad_norm": 1.8668314438757327, + "learning_rate": 9.280958309378257e-06, + "loss": 0.9404, + "step": 5585 + }, + { + "epoch": 0.5373220469411312, + "grad_norm": 1.8575697911977103, + "learning_rate": 9.27785098535548e-06, + "loss": 0.8487, + "step": 5586 + }, + { + "epoch": 0.537418237783763, + "grad_norm": 2.0623342974323893, + "learning_rate": 9.274743731423482e-06, + "loss": 1.0007, + "step": 5587 + }, + { + "epoch": 0.5375144286263948, + "grad_norm": 1.8883394397747066, + "learning_rate": 9.271636547883856e-06, + "loss": 0.9353, + "step": 5588 + }, + { + "epoch": 0.5376106194690266, + "grad_norm": 2.019592811960824, + "learning_rate": 9.268529435038167e-06, + "loss": 0.8896, + "step": 5589 + }, + { + "epoch": 0.5377068103116583, + "grad_norm": 1.962939684638706, + "learning_rate": 9.265422393187998e-06, + "loss": 0.9933, + "step": 5590 + }, + { + "epoch": 0.5378030011542901, + "grad_norm": 2.143226409131343, + "learning_rate": 9.26231542263491e-06, + "loss": 1.0318, + "step": 5591 + }, + { + "epoch": 0.5378991919969219, + "grad_norm": 2.157146790434487, + "learning_rate": 9.259208523680457e-06, + "loss": 1.0691, + "step": 5592 + }, + { + "epoch": 0.5379953828395537, + "grad_norm": 1.9239800119559507, + "learning_rate": 9.2561016966262e-06, + "loss": 0.8903, + "step": 5593 + }, + { + "epoch": 0.5380915736821854, + "grad_norm": 2.5368375900378277, + "learning_rate": 9.252994941773679e-06, + "loss": 0.8663, + "step": 5594 + }, + { + "epoch": 0.5381877645248172, + "grad_norm": 2.1734542857153674, + "learning_rate": 9.24988825942443e-06, + "loss": 1.0036, + "step": 5595 + }, + { + "epoch": 0.538283955367449, + "grad_norm": 2.0168855013708105, + "learning_rate": 9.24678164987998e-06, + "loss": 1.0612, + "step": 5596 + }, + { + "epoch": 0.5383801462100808, + "grad_norm": 2.1182093786380194, + "learning_rate": 9.24367511344186e-06, + "loss": 0.9621, + "step": 5597 + }, + { + "epoch": 0.5384763370527126, + "grad_norm": 1.8065032634166804, + "learning_rate": 9.24056865041158e-06, + "loss": 0.9442, + "step": 5598 + }, + { + "epoch": 0.5385725278953444, + "grad_norm": 2.160487415656165, + "learning_rate": 9.237462261090652e-06, + "loss": 1.0351, + "step": 5599 + }, + { + "epoch": 0.5386687187379762, + "grad_norm": 2.19417918309358, + "learning_rate": 9.234355945780581e-06, + "loss": 1.0858, + "step": 5600 + }, + { + "epoch": 0.538764909580608, + "grad_norm": 2.28516337591263, + "learning_rate": 9.231249704782854e-06, + "loss": 1.0241, + "step": 5601 + }, + { + "epoch": 0.5388611004232398, + "grad_norm": 1.6770565056011544, + "learning_rate": 9.228143538398963e-06, + "loss": 0.9719, + "step": 5602 + }, + { + "epoch": 0.5389572912658714, + "grad_norm": 1.899522799796236, + "learning_rate": 9.225037446930386e-06, + "loss": 1.0291, + "step": 5603 + }, + { + "epoch": 0.5390534821085032, + "grad_norm": 1.9775527015118686, + "learning_rate": 9.221931430678598e-06, + "loss": 1.0912, + "step": 5604 + }, + { + "epoch": 0.539149672951135, + "grad_norm": 2.0113845061656597, + "learning_rate": 9.218825489945067e-06, + "loss": 0.839, + "step": 5605 + }, + { + "epoch": 0.5392458637937668, + "grad_norm": 2.000712583273962, + "learning_rate": 9.215719625031245e-06, + "loss": 0.9382, + "step": 5606 + }, + { + "epoch": 0.5393420546363986, + "grad_norm": 1.9749922583103843, + "learning_rate": 9.212613836238586e-06, + "loss": 1.0038, + "step": 5607 + }, + { + "epoch": 0.5394382454790304, + "grad_norm": 2.1373202283722446, + "learning_rate": 9.209508123868534e-06, + "loss": 0.9994, + "step": 5608 + }, + { + "epoch": 0.5395344363216622, + "grad_norm": 1.8563373702297274, + "learning_rate": 9.206402488222524e-06, + "loss": 0.9517, + "step": 5609 + }, + { + "epoch": 0.539630627164294, + "grad_norm": 1.9496768706200804, + "learning_rate": 9.203296929601986e-06, + "loss": 0.9873, + "step": 5610 + }, + { + "epoch": 0.5397268180069258, + "grad_norm": 1.71279622835469, + "learning_rate": 9.200191448308344e-06, + "loss": 0.9556, + "step": 5611 + }, + { + "epoch": 0.5398230088495575, + "grad_norm": 1.9521245018418247, + "learning_rate": 9.197086044643004e-06, + "loss": 1.012, + "step": 5612 + }, + { + "epoch": 0.5399191996921893, + "grad_norm": 1.96098332417582, + "learning_rate": 9.19398071890738e-06, + "loss": 0.8508, + "step": 5613 + }, + { + "epoch": 0.5400153905348211, + "grad_norm": 1.7974150051737667, + "learning_rate": 9.190875471402865e-06, + "loss": 0.8496, + "step": 5614 + }, + { + "epoch": 0.5401115813774529, + "grad_norm": 2.036758278946662, + "learning_rate": 9.187770302430849e-06, + "loss": 0.9347, + "step": 5615 + }, + { + "epoch": 0.5402077722200846, + "grad_norm": 2.082355396983703, + "learning_rate": 9.184665212292723e-06, + "loss": 0.9281, + "step": 5616 + }, + { + "epoch": 0.5403039630627164, + "grad_norm": 2.086413271513012, + "learning_rate": 9.18156020128986e-06, + "loss": 0.8642, + "step": 5617 + }, + { + "epoch": 0.5404001539053482, + "grad_norm": 1.8050721297871497, + "learning_rate": 9.178455269723623e-06, + "loss": 0.9345, + "step": 5618 + }, + { + "epoch": 0.54049634474798, + "grad_norm": 1.907636937147771, + "learning_rate": 9.175350417895377e-06, + "loss": 1.0698, + "step": 5619 + }, + { + "epoch": 0.5405925355906118, + "grad_norm": 2.459644540334886, + "learning_rate": 9.172245646106471e-06, + "loss": 0.9159, + "step": 5620 + }, + { + "epoch": 0.5406887264332435, + "grad_norm": 1.985813069999203, + "learning_rate": 9.169140954658254e-06, + "loss": 0.9764, + "step": 5621 + }, + { + "epoch": 0.5407849172758753, + "grad_norm": 2.0429222511145917, + "learning_rate": 9.166036343852061e-06, + "loss": 0.9816, + "step": 5622 + }, + { + "epoch": 0.5408811081185071, + "grad_norm": 2.163859857039476, + "learning_rate": 9.162931813989223e-06, + "loss": 1.0551, + "step": 5623 + }, + { + "epoch": 0.5409772989611389, + "grad_norm": 1.7286542262120284, + "learning_rate": 9.159827365371055e-06, + "loss": 0.9679, + "step": 5624 + }, + { + "epoch": 0.5410734898037707, + "grad_norm": 1.6834940298261978, + "learning_rate": 9.156722998298878e-06, + "loss": 0.7728, + "step": 5625 + }, + { + "epoch": 0.5411696806464025, + "grad_norm": 2.3945020561701362, + "learning_rate": 9.153618713073995e-06, + "loss": 0.8958, + "step": 5626 + }, + { + "epoch": 0.5412658714890343, + "grad_norm": 2.2077127680480606, + "learning_rate": 9.1505145099977e-06, + "loss": 0.941, + "step": 5627 + }, + { + "epoch": 0.5413620623316661, + "grad_norm": 2.2587267574335415, + "learning_rate": 9.14741038937129e-06, + "loss": 0.9031, + "step": 5628 + }, + { + "epoch": 0.5414582531742979, + "grad_norm": 1.841663697712556, + "learning_rate": 9.144306351496045e-06, + "loss": 0.9302, + "step": 5629 + }, + { + "epoch": 0.5415544440169295, + "grad_norm": 2.0932171480743142, + "learning_rate": 9.141202396673232e-06, + "loss": 1.0498, + "step": 5630 + }, + { + "epoch": 0.5416506348595613, + "grad_norm": 2.0044261375487302, + "learning_rate": 9.138098525204122e-06, + "loss": 0.9556, + "step": 5631 + }, + { + "epoch": 0.5417468257021931, + "grad_norm": 2.1599408099376833, + "learning_rate": 9.13499473738997e-06, + "loss": 0.9535, + "step": 5632 + }, + { + "epoch": 0.5418430165448249, + "grad_norm": 2.288424309494425, + "learning_rate": 9.131891033532027e-06, + "loss": 0.9298, + "step": 5633 + }, + { + "epoch": 0.5419392073874567, + "grad_norm": 2.0173084918669475, + "learning_rate": 9.128787413931536e-06, + "loss": 0.9926, + "step": 5634 + }, + { + "epoch": 0.5420353982300885, + "grad_norm": 2.1900683748570278, + "learning_rate": 9.125683878889726e-06, + "loss": 0.8541, + "step": 5635 + }, + { + "epoch": 0.5421315890727203, + "grad_norm": 2.3730147254495657, + "learning_rate": 9.122580428707822e-06, + "loss": 1.1286, + "step": 5636 + }, + { + "epoch": 0.5422277799153521, + "grad_norm": 1.7997859569940422, + "learning_rate": 9.119477063687042e-06, + "loss": 0.8745, + "step": 5637 + }, + { + "epoch": 0.5423239707579839, + "grad_norm": 1.9839335026478704, + "learning_rate": 9.116373784128597e-06, + "loss": 0.942, + "step": 5638 + }, + { + "epoch": 0.5424201616006156, + "grad_norm": 2.0638273021040194, + "learning_rate": 9.113270590333681e-06, + "loss": 1.0377, + "step": 5639 + }, + { + "epoch": 0.5425163524432474, + "grad_norm": 1.9712335908142384, + "learning_rate": 9.110167482603494e-06, + "loss": 0.8924, + "step": 5640 + }, + { + "epoch": 0.5426125432858792, + "grad_norm": 2.0592684138314747, + "learning_rate": 9.10706446123921e-06, + "loss": 0.9257, + "step": 5641 + }, + { + "epoch": 0.542708734128511, + "grad_norm": 2.2559252171739317, + "learning_rate": 9.10396152654201e-06, + "loss": 1.0834, + "step": 5642 + }, + { + "epoch": 0.5428049249711427, + "grad_norm": 2.0153699414728616, + "learning_rate": 9.100858678813056e-06, + "loss": 1.06, + "step": 5643 + }, + { + "epoch": 0.5429011158137745, + "grad_norm": 1.831079239580315, + "learning_rate": 9.097755918353513e-06, + "loss": 0.9381, + "step": 5644 + }, + { + "epoch": 0.5429973066564063, + "grad_norm": 1.7633053318146417, + "learning_rate": 9.094653245464525e-06, + "loss": 0.8213, + "step": 5645 + }, + { + "epoch": 0.5430934974990381, + "grad_norm": 2.484284777309606, + "learning_rate": 9.091550660447236e-06, + "loss": 1.0579, + "step": 5646 + }, + { + "epoch": 0.5431896883416699, + "grad_norm": 2.0395827276542424, + "learning_rate": 9.088448163602776e-06, + "loss": 0.9503, + "step": 5647 + }, + { + "epoch": 0.5432858791843016, + "grad_norm": 2.1514414107937267, + "learning_rate": 9.08534575523227e-06, + "loss": 0.9995, + "step": 5648 + }, + { + "epoch": 0.5433820700269334, + "grad_norm": 1.8895514779270233, + "learning_rate": 9.082243435636833e-06, + "loss": 0.9217, + "step": 5649 + }, + { + "epoch": 0.5434782608695652, + "grad_norm": 1.9456910266807237, + "learning_rate": 9.079141205117573e-06, + "loss": 1.0057, + "step": 5650 + }, + { + "epoch": 0.543574451712197, + "grad_norm": 2.0509305558779127, + "learning_rate": 9.076039063975586e-06, + "loss": 0.9797, + "step": 5651 + }, + { + "epoch": 0.5436706425548288, + "grad_norm": 1.9384808542014333, + "learning_rate": 9.072937012511968e-06, + "loss": 0.9742, + "step": 5652 + }, + { + "epoch": 0.5437668333974606, + "grad_norm": 1.7365730023657153, + "learning_rate": 9.069835051027794e-06, + "loss": 0.8615, + "step": 5653 + }, + { + "epoch": 0.5438630242400924, + "grad_norm": 1.9655206741203808, + "learning_rate": 9.066733179824134e-06, + "loss": 0.8484, + "step": 5654 + }, + { + "epoch": 0.5439592150827242, + "grad_norm": 2.2787473396840947, + "learning_rate": 9.063631399202055e-06, + "loss": 0.9942, + "step": 5655 + }, + { + "epoch": 0.544055405925356, + "grad_norm": 2.024449391364958, + "learning_rate": 9.060529709462613e-06, + "loss": 0.9799, + "step": 5656 + }, + { + "epoch": 0.5441515967679876, + "grad_norm": 1.902724626720422, + "learning_rate": 9.057428110906853e-06, + "loss": 0.9429, + "step": 5657 + }, + { + "epoch": 0.5442477876106194, + "grad_norm": 1.9444530680785745, + "learning_rate": 9.054326603835807e-06, + "loss": 1.0025, + "step": 5658 + }, + { + "epoch": 0.5443439784532512, + "grad_norm": 2.156658391296779, + "learning_rate": 9.051225188550507e-06, + "loss": 0.9571, + "step": 5659 + }, + { + "epoch": 0.544440169295883, + "grad_norm": 2.211105230548213, + "learning_rate": 9.048123865351971e-06, + "loss": 1.0469, + "step": 5660 + }, + { + "epoch": 0.5445363601385148, + "grad_norm": 1.859317180256454, + "learning_rate": 9.045022634541211e-06, + "loss": 0.9023, + "step": 5661 + }, + { + "epoch": 0.5446325509811466, + "grad_norm": 2.046119332414595, + "learning_rate": 9.041921496419225e-06, + "loss": 1.0263, + "step": 5662 + }, + { + "epoch": 0.5447287418237784, + "grad_norm": 2.391189456411007, + "learning_rate": 9.038820451287013e-06, + "loss": 0.9708, + "step": 5663 + }, + { + "epoch": 0.5448249326664102, + "grad_norm": 2.1323401074145925, + "learning_rate": 9.035719499445545e-06, + "loss": 0.9678, + "step": 5664 + }, + { + "epoch": 0.544921123509042, + "grad_norm": 2.1584346096978932, + "learning_rate": 9.032618641195804e-06, + "loss": 0.9018, + "step": 5665 + }, + { + "epoch": 0.5450173143516738, + "grad_norm": 1.9500362003995872, + "learning_rate": 9.029517876838755e-06, + "loss": 0.8999, + "step": 5666 + }, + { + "epoch": 0.5451135051943055, + "grad_norm": 2.015098150909185, + "learning_rate": 9.026417206675348e-06, + "loss": 0.9681, + "step": 5667 + }, + { + "epoch": 0.5452096960369373, + "grad_norm": 2.0782608290779345, + "learning_rate": 9.023316631006536e-06, + "loss": 1.0249, + "step": 5668 + }, + { + "epoch": 0.545305886879569, + "grad_norm": 2.1773526191449823, + "learning_rate": 9.020216150133257e-06, + "loss": 1.0783, + "step": 5669 + }, + { + "epoch": 0.5454020777222008, + "grad_norm": 1.8668747626489282, + "learning_rate": 9.017115764356436e-06, + "loss": 0.8045, + "step": 5670 + }, + { + "epoch": 0.5454982685648326, + "grad_norm": 1.6917822384236492, + "learning_rate": 9.01401547397699e-06, + "loss": 0.957, + "step": 5671 + }, + { + "epoch": 0.5455944594074644, + "grad_norm": 2.0076045912995255, + "learning_rate": 9.010915279295833e-06, + "loss": 0.9081, + "step": 5672 + }, + { + "epoch": 0.5456906502500962, + "grad_norm": 3.945669665154974, + "learning_rate": 9.007815180613863e-06, + "loss": 0.8246, + "step": 5673 + }, + { + "epoch": 0.545786841092728, + "grad_norm": 1.9998285389990311, + "learning_rate": 9.004715178231975e-06, + "loss": 1.0839, + "step": 5674 + }, + { + "epoch": 0.5458830319353598, + "grad_norm": 1.7702771064158207, + "learning_rate": 9.001615272451049e-06, + "loss": 0.8823, + "step": 5675 + }, + { + "epoch": 0.5459792227779915, + "grad_norm": 2.2222204427941405, + "learning_rate": 8.998515463571953e-06, + "loss": 0.9761, + "step": 5676 + }, + { + "epoch": 0.5460754136206233, + "grad_norm": 2.323084009142863, + "learning_rate": 8.995415751895556e-06, + "loss": 1.0949, + "step": 5677 + }, + { + "epoch": 0.5461716044632551, + "grad_norm": 2.0675441892774176, + "learning_rate": 8.992316137722711e-06, + "loss": 1.0034, + "step": 5678 + }, + { + "epoch": 0.5462677953058869, + "grad_norm": 2.121751276969261, + "learning_rate": 8.989216621354258e-06, + "loss": 0.9529, + "step": 5679 + }, + { + "epoch": 0.5463639861485187, + "grad_norm": 2.0292366917017945, + "learning_rate": 8.986117203091042e-06, + "loss": 0.9993, + "step": 5680 + }, + { + "epoch": 0.5464601769911505, + "grad_norm": 1.77385233661843, + "learning_rate": 8.983017883233875e-06, + "loss": 0.9242, + "step": 5681 + }, + { + "epoch": 0.5465563678337823, + "grad_norm": 1.655356905084097, + "learning_rate": 8.97991866208358e-06, + "loss": 0.9687, + "step": 5682 + }, + { + "epoch": 0.546652558676414, + "grad_norm": 1.7232890453403502, + "learning_rate": 8.976819539940961e-06, + "loss": 0.8833, + "step": 5683 + }, + { + "epoch": 0.5467487495190458, + "grad_norm": 2.192943474415788, + "learning_rate": 8.973720517106814e-06, + "loss": 0.9018, + "step": 5684 + }, + { + "epoch": 0.5468449403616775, + "grad_norm": 2.0034891295965114, + "learning_rate": 8.970621593881929e-06, + "loss": 0.9191, + "step": 5685 + }, + { + "epoch": 0.5469411312043093, + "grad_norm": 1.8767814294407852, + "learning_rate": 8.967522770567086e-06, + "loss": 1.0068, + "step": 5686 + }, + { + "epoch": 0.5470373220469411, + "grad_norm": 1.980272801067979, + "learning_rate": 8.964424047463043e-06, + "loss": 0.9193, + "step": 5687 + }, + { + "epoch": 0.5471335128895729, + "grad_norm": 1.9483813936364798, + "learning_rate": 8.961325424870561e-06, + "loss": 0.8888, + "step": 5688 + }, + { + "epoch": 0.5472297037322047, + "grad_norm": 1.839063530015663, + "learning_rate": 8.958226903090392e-06, + "loss": 0.9622, + "step": 5689 + }, + { + "epoch": 0.5473258945748365, + "grad_norm": 1.7911582610506944, + "learning_rate": 8.955128482423271e-06, + "loss": 0.8644, + "step": 5690 + }, + { + "epoch": 0.5474220854174683, + "grad_norm": 1.5536564135431699, + "learning_rate": 8.952030163169927e-06, + "loss": 0.8588, + "step": 5691 + }, + { + "epoch": 0.5475182762601001, + "grad_norm": 2.077174735519367, + "learning_rate": 8.948931945631082e-06, + "loss": 1.0513, + "step": 5692 + }, + { + "epoch": 0.5476144671027319, + "grad_norm": 1.964964312121262, + "learning_rate": 8.94583383010744e-06, + "loss": 0.9837, + "step": 5693 + }, + { + "epoch": 0.5477106579453636, + "grad_norm": 1.8174287185013354, + "learning_rate": 8.9427358168997e-06, + "loss": 0.9536, + "step": 5694 + }, + { + "epoch": 0.5478068487879953, + "grad_norm": 1.9343150093961772, + "learning_rate": 8.939637906308551e-06, + "loss": 0.9729, + "step": 5695 + }, + { + "epoch": 0.5479030396306271, + "grad_norm": 1.7582635600671763, + "learning_rate": 8.936540098634675e-06, + "loss": 0.978, + "step": 5696 + }, + { + "epoch": 0.5479992304732589, + "grad_norm": 1.7055537964453376, + "learning_rate": 8.933442394178738e-06, + "loss": 0.867, + "step": 5697 + }, + { + "epoch": 0.5480954213158907, + "grad_norm": 2.1282659228169076, + "learning_rate": 8.930344793241404e-06, + "loss": 1.0451, + "step": 5698 + }, + { + "epoch": 0.5481916121585225, + "grad_norm": 2.196632997768425, + "learning_rate": 8.927247296123313e-06, + "loss": 1.0958, + "step": 5699 + }, + { + "epoch": 0.5482878030011543, + "grad_norm": 2.0001662961419653, + "learning_rate": 8.924149903125108e-06, + "loss": 0.9201, + "step": 5700 + }, + { + "epoch": 0.5483839938437861, + "grad_norm": 1.8668771874646672, + "learning_rate": 8.92105261454742e-06, + "loss": 0.9911, + "step": 5701 + }, + { + "epoch": 0.5484801846864179, + "grad_norm": 1.9464824360711537, + "learning_rate": 8.917955430690865e-06, + "loss": 0.8808, + "step": 5702 + }, + { + "epoch": 0.5485763755290496, + "grad_norm": 1.8557823065351673, + "learning_rate": 8.91485835185605e-06, + "loss": 1.0286, + "step": 5703 + }, + { + "epoch": 0.5486725663716814, + "grad_norm": 2.01129434377053, + "learning_rate": 8.91176137834358e-06, + "loss": 0.9611, + "step": 5704 + }, + { + "epoch": 0.5487687572143132, + "grad_norm": 1.8862247622810417, + "learning_rate": 8.908664510454036e-06, + "loss": 1.0531, + "step": 5705 + }, + { + "epoch": 0.548864948056945, + "grad_norm": 1.7496941220002364, + "learning_rate": 8.905567748487997e-06, + "loss": 0.9574, + "step": 5706 + }, + { + "epoch": 0.5489611388995768, + "grad_norm": 2.125067816652298, + "learning_rate": 8.902471092746029e-06, + "loss": 1.0024, + "step": 5707 + }, + { + "epoch": 0.5490573297422086, + "grad_norm": 2.1101750274524416, + "learning_rate": 8.899374543528695e-06, + "loss": 0.964, + "step": 5708 + }, + { + "epoch": 0.5491535205848403, + "grad_norm": 1.9293584210575252, + "learning_rate": 8.896278101136538e-06, + "loss": 0.9969, + "step": 5709 + }, + { + "epoch": 0.5492497114274721, + "grad_norm": 1.9258488616512814, + "learning_rate": 8.893181765870094e-06, + "loss": 0.9737, + "step": 5710 + }, + { + "epoch": 0.5493459022701039, + "grad_norm": 2.2984417311725447, + "learning_rate": 8.890085538029887e-06, + "loss": 1.0339, + "step": 5711 + }, + { + "epoch": 0.5494420931127356, + "grad_norm": 1.9788311073998894, + "learning_rate": 8.886989417916435e-06, + "loss": 0.9274, + "step": 5712 + }, + { + "epoch": 0.5495382839553674, + "grad_norm": 2.0543627748059228, + "learning_rate": 8.883893405830245e-06, + "loss": 1.027, + "step": 5713 + }, + { + "epoch": 0.5496344747979992, + "grad_norm": 1.9043310020461275, + "learning_rate": 8.88079750207181e-06, + "loss": 0.9616, + "step": 5714 + }, + { + "epoch": 0.549730665640631, + "grad_norm": 1.8687378277686473, + "learning_rate": 8.877701706941617e-06, + "loss": 0.9599, + "step": 5715 + }, + { + "epoch": 0.5498268564832628, + "grad_norm": 1.9672928609820952, + "learning_rate": 8.87460602074013e-06, + "loss": 1.0583, + "step": 5716 + }, + { + "epoch": 0.5499230473258946, + "grad_norm": 2.0501006598890705, + "learning_rate": 8.87151044376782e-06, + "loss": 0.9709, + "step": 5717 + }, + { + "epoch": 0.5500192381685264, + "grad_norm": 1.820815864454182, + "learning_rate": 8.86841497632514e-06, + "loss": 1.0215, + "step": 5718 + }, + { + "epoch": 0.5501154290111582, + "grad_norm": 1.868436099415058, + "learning_rate": 8.865319618712526e-06, + "loss": 0.9896, + "step": 5719 + }, + { + "epoch": 0.55021161985379, + "grad_norm": 2.2138042448590767, + "learning_rate": 8.862224371230418e-06, + "loss": 1.0564, + "step": 5720 + }, + { + "epoch": 0.5503078106964217, + "grad_norm": 1.9848729693291816, + "learning_rate": 8.85912923417923e-06, + "loss": 0.9984, + "step": 5721 + }, + { + "epoch": 0.5504040015390534, + "grad_norm": 1.942695728095249, + "learning_rate": 8.85603420785937e-06, + "loss": 0.9137, + "step": 5722 + }, + { + "epoch": 0.5505001923816852, + "grad_norm": 2.2835037224796815, + "learning_rate": 8.852939292571243e-06, + "loss": 1.1087, + "step": 5723 + }, + { + "epoch": 0.550596383224317, + "grad_norm": 1.8998902788438572, + "learning_rate": 8.84984448861523e-06, + "loss": 0.9496, + "step": 5724 + }, + { + "epoch": 0.5506925740669488, + "grad_norm": 2.062648541619988, + "learning_rate": 8.846749796291715e-06, + "loss": 0.9218, + "step": 5725 + }, + { + "epoch": 0.5507887649095806, + "grad_norm": 5.304602149591799, + "learning_rate": 8.84365521590106e-06, + "loss": 0.8218, + "step": 5726 + }, + { + "epoch": 0.5508849557522124, + "grad_norm": 1.8978781173104717, + "learning_rate": 8.840560747743627e-06, + "loss": 0.9753, + "step": 5727 + }, + { + "epoch": 0.5509811465948442, + "grad_norm": 2.0574921604476644, + "learning_rate": 8.837466392119752e-06, + "loss": 0.8847, + "step": 5728 + }, + { + "epoch": 0.551077337437476, + "grad_norm": 1.7972024072435637, + "learning_rate": 8.834372149329774e-06, + "loss": 0.8753, + "step": 5729 + }, + { + "epoch": 0.5511735282801077, + "grad_norm": 2.1867870125148983, + "learning_rate": 8.831278019674017e-06, + "loss": 0.943, + "step": 5730 + }, + { + "epoch": 0.5512697191227395, + "grad_norm": 1.9056488679366916, + "learning_rate": 8.828184003452787e-06, + "loss": 0.9854, + "step": 5731 + }, + { + "epoch": 0.5513659099653713, + "grad_norm": 2.004380341717103, + "learning_rate": 8.825090100966396e-06, + "loss": 0.994, + "step": 5732 + }, + { + "epoch": 0.5514621008080031, + "grad_norm": 1.9422597267394461, + "learning_rate": 8.82199631251512e-06, + "loss": 1.0656, + "step": 5733 + }, + { + "epoch": 0.5515582916506349, + "grad_norm": 2.064221711137433, + "learning_rate": 8.818902638399247e-06, + "loss": 0.8641, + "step": 5734 + }, + { + "epoch": 0.5516544824932667, + "grad_norm": 2.183865901908032, + "learning_rate": 8.815809078919041e-06, + "loss": 0.8234, + "step": 5735 + }, + { + "epoch": 0.5517506733358984, + "grad_norm": 1.9237942328206639, + "learning_rate": 8.81271563437476e-06, + "loss": 1.0352, + "step": 5736 + }, + { + "epoch": 0.5518468641785302, + "grad_norm": 2.081534017557748, + "learning_rate": 8.809622305066652e-06, + "loss": 0.9738, + "step": 5737 + }, + { + "epoch": 0.551943055021162, + "grad_norm": 2.1748786559011966, + "learning_rate": 8.806529091294948e-06, + "loss": 0.975, + "step": 5738 + }, + { + "epoch": 0.5520392458637937, + "grad_norm": 2.4625501310602993, + "learning_rate": 8.803435993359873e-06, + "loss": 0.8963, + "step": 5739 + }, + { + "epoch": 0.5521354367064255, + "grad_norm": 1.794294695018131, + "learning_rate": 8.800343011561633e-06, + "loss": 0.9841, + "step": 5740 + }, + { + "epoch": 0.5522316275490573, + "grad_norm": 2.237042175562181, + "learning_rate": 8.797250146200434e-06, + "loss": 0.9844, + "step": 5741 + }, + { + "epoch": 0.5523278183916891, + "grad_norm": 1.7941234315149925, + "learning_rate": 8.794157397576464e-06, + "loss": 0.9092, + "step": 5742 + }, + { + "epoch": 0.5524240092343209, + "grad_norm": 1.968623144446623, + "learning_rate": 8.7910647659899e-06, + "loss": 0.942, + "step": 5743 + }, + { + "epoch": 0.5525202000769527, + "grad_norm": 1.849851971797638, + "learning_rate": 8.787972251740916e-06, + "loss": 0.8962, + "step": 5744 + }, + { + "epoch": 0.5526163909195845, + "grad_norm": 2.01434553827672, + "learning_rate": 8.784879855129656e-06, + "loss": 0.9009, + "step": 5745 + }, + { + "epoch": 0.5527125817622163, + "grad_norm": 2.0515539578127218, + "learning_rate": 8.781787576456269e-06, + "loss": 0.973, + "step": 5746 + }, + { + "epoch": 0.5528087726048481, + "grad_norm": 2.0174898786141124, + "learning_rate": 8.778695416020883e-06, + "loss": 0.9041, + "step": 5747 + }, + { + "epoch": 0.5529049634474797, + "grad_norm": 2.121758071916497, + "learning_rate": 8.775603374123627e-06, + "loss": 1.0503, + "step": 5748 + }, + { + "epoch": 0.5530011542901115, + "grad_norm": 2.426302168760283, + "learning_rate": 8.772511451064604e-06, + "loss": 0.9479, + "step": 5749 + }, + { + "epoch": 0.5530973451327433, + "grad_norm": 7.992123213751276, + "learning_rate": 8.769419647143917e-06, + "loss": 0.9728, + "step": 5750 + }, + { + "epoch": 0.5531935359753751, + "grad_norm": 1.967143658771755, + "learning_rate": 8.766327962661645e-06, + "loss": 0.8968, + "step": 5751 + }, + { + "epoch": 0.5532897268180069, + "grad_norm": 2.2763519691947893, + "learning_rate": 8.763236397917865e-06, + "loss": 1.0548, + "step": 5752 + }, + { + "epoch": 0.5533859176606387, + "grad_norm": 1.7674626741818205, + "learning_rate": 8.760144953212643e-06, + "loss": 0.9433, + "step": 5753 + }, + { + "epoch": 0.5534821085032705, + "grad_norm": 1.8998390905209128, + "learning_rate": 8.757053628846028e-06, + "loss": 0.8614, + "step": 5754 + }, + { + "epoch": 0.5535782993459023, + "grad_norm": 1.9953081696021715, + "learning_rate": 8.753962425118064e-06, + "loss": 1.0535, + "step": 5755 + }, + { + "epoch": 0.5536744901885341, + "grad_norm": 2.206096687758582, + "learning_rate": 8.75087134232877e-06, + "loss": 0.9935, + "step": 5756 + }, + { + "epoch": 0.5537706810311658, + "grad_norm": 1.785939794434311, + "learning_rate": 8.74778038077817e-06, + "loss": 1.0629, + "step": 5757 + }, + { + "epoch": 0.5538668718737976, + "grad_norm": 2.188494441756408, + "learning_rate": 8.744689540766265e-06, + "loss": 1.0173, + "step": 5758 + }, + { + "epoch": 0.5539630627164294, + "grad_norm": 2.08921782395856, + "learning_rate": 8.741598822593047e-06, + "loss": 0.9254, + "step": 5759 + }, + { + "epoch": 0.5540592535590612, + "grad_norm": 2.0766224042135377, + "learning_rate": 8.738508226558499e-06, + "loss": 1.0674, + "step": 5760 + }, + { + "epoch": 0.554155444401693, + "grad_norm": 1.901368923904158, + "learning_rate": 8.735417752962594e-06, + "loss": 0.8896, + "step": 5761 + }, + { + "epoch": 0.5542516352443247, + "grad_norm": 2.0011440950102264, + "learning_rate": 8.73232740210528e-06, + "loss": 0.9525, + "step": 5762 + }, + { + "epoch": 0.5543478260869565, + "grad_norm": 1.913779388691494, + "learning_rate": 8.729237174286506e-06, + "loss": 1.0835, + "step": 5763 + }, + { + "epoch": 0.5544440169295883, + "grad_norm": 1.6887578352322965, + "learning_rate": 8.726147069806206e-06, + "loss": 0.9375, + "step": 5764 + }, + { + "epoch": 0.5545402077722201, + "grad_norm": 2.087396844946638, + "learning_rate": 8.723057088964301e-06, + "loss": 0.9091, + "step": 5765 + }, + { + "epoch": 0.5546363986148518, + "grad_norm": 2.2623275702286327, + "learning_rate": 8.719967232060698e-06, + "loss": 0.9522, + "step": 5766 + }, + { + "epoch": 0.5547325894574836, + "grad_norm": 1.8828695351725446, + "learning_rate": 8.716877499395303e-06, + "loss": 0.9626, + "step": 5767 + }, + { + "epoch": 0.5548287803001154, + "grad_norm": 4.672619339914033, + "learning_rate": 8.713787891267988e-06, + "loss": 0.8809, + "step": 5768 + }, + { + "epoch": 0.5549249711427472, + "grad_norm": 1.7198694061029027, + "learning_rate": 8.710698407978636e-06, + "loss": 1.0476, + "step": 5769 + }, + { + "epoch": 0.555021161985379, + "grad_norm": 2.051192482440112, + "learning_rate": 8.707609049827102e-06, + "loss": 0.8288, + "step": 5770 + }, + { + "epoch": 0.5551173528280108, + "grad_norm": 1.9194948313132716, + "learning_rate": 8.704519817113238e-06, + "loss": 0.988, + "step": 5771 + }, + { + "epoch": 0.5552135436706426, + "grad_norm": 2.175980131363841, + "learning_rate": 8.70143071013688e-06, + "loss": 1.0391, + "step": 5772 + }, + { + "epoch": 0.5553097345132744, + "grad_norm": 2.1491197120349774, + "learning_rate": 8.698341729197857e-06, + "loss": 1.0572, + "step": 5773 + }, + { + "epoch": 0.5554059253559062, + "grad_norm": 2.188017171439972, + "learning_rate": 8.695252874595972e-06, + "loss": 1.1006, + "step": 5774 + }, + { + "epoch": 0.555502116198538, + "grad_norm": 1.8944849062476756, + "learning_rate": 8.692164146631028e-06, + "loss": 1.0293, + "step": 5775 + }, + { + "epoch": 0.5555983070411696, + "grad_norm": 2.2310098597597237, + "learning_rate": 8.689075545602816e-06, + "loss": 0.9424, + "step": 5776 + }, + { + "epoch": 0.5556944978838014, + "grad_norm": 2.242948536661259, + "learning_rate": 8.68598707181111e-06, + "loss": 0.9483, + "step": 5777 + }, + { + "epoch": 0.5557906887264332, + "grad_norm": 1.7367736839971044, + "learning_rate": 8.68289872555567e-06, + "loss": 1.005, + "step": 5778 + }, + { + "epoch": 0.555886879569065, + "grad_norm": 1.995244605582053, + "learning_rate": 8.679810507136258e-06, + "loss": 1.0711, + "step": 5779 + }, + { + "epoch": 0.5559830704116968, + "grad_norm": 2.361882960125292, + "learning_rate": 8.676722416852594e-06, + "loss": 0.9822, + "step": 5780 + }, + { + "epoch": 0.5560792612543286, + "grad_norm": 2.128667631305734, + "learning_rate": 8.673634455004416e-06, + "loss": 0.9807, + "step": 5781 + }, + { + "epoch": 0.5561754520969604, + "grad_norm": 1.8736781356019911, + "learning_rate": 8.670546621891434e-06, + "loss": 0.9778, + "step": 5782 + }, + { + "epoch": 0.5562716429395922, + "grad_norm": 1.8047511380626338, + "learning_rate": 8.667458917813348e-06, + "loss": 0.9551, + "step": 5783 + }, + { + "epoch": 0.556367833782224, + "grad_norm": 2.009915679639374, + "learning_rate": 8.66437134306985e-06, + "loss": 1.0515, + "step": 5784 + }, + { + "epoch": 0.5564640246248557, + "grad_norm": 1.7832717155131916, + "learning_rate": 8.66128389796061e-06, + "loss": 0.9238, + "step": 5785 + }, + { + "epoch": 0.5565602154674875, + "grad_norm": 1.748439434307066, + "learning_rate": 8.658196582785297e-06, + "loss": 0.9001, + "step": 5786 + }, + { + "epoch": 0.5566564063101193, + "grad_norm": 1.9674364790336087, + "learning_rate": 8.655109397843555e-06, + "loss": 0.9762, + "step": 5787 + }, + { + "epoch": 0.556752597152751, + "grad_norm": 1.8853917393931394, + "learning_rate": 8.652022343435027e-06, + "loss": 1.0098, + "step": 5788 + }, + { + "epoch": 0.5568487879953828, + "grad_norm": 1.800399148965345, + "learning_rate": 8.64893541985934e-06, + "loss": 0.8801, + "step": 5789 + }, + { + "epoch": 0.5569449788380146, + "grad_norm": 1.7650360650591264, + "learning_rate": 8.645848627416102e-06, + "loss": 0.9678, + "step": 5790 + }, + { + "epoch": 0.5570411696806464, + "grad_norm": 2.0725781052342835, + "learning_rate": 8.642761966404914e-06, + "loss": 0.9646, + "step": 5791 + }, + { + "epoch": 0.5571373605232782, + "grad_norm": 1.7922716212885597, + "learning_rate": 8.63967543712536e-06, + "loss": 0.9771, + "step": 5792 + }, + { + "epoch": 0.55723355136591, + "grad_norm": 1.9889147522204833, + "learning_rate": 8.63658903987702e-06, + "loss": 0.996, + "step": 5793 + }, + { + "epoch": 0.5573297422085417, + "grad_norm": 1.9619242423790013, + "learning_rate": 8.633502774959453e-06, + "loss": 0.9376, + "step": 5794 + }, + { + "epoch": 0.5574259330511735, + "grad_norm": 2.0066291109099366, + "learning_rate": 8.630416642672205e-06, + "loss": 0.887, + "step": 5795 + }, + { + "epoch": 0.5575221238938053, + "grad_norm": 1.8653375289643963, + "learning_rate": 8.627330643314818e-06, + "loss": 0.9411, + "step": 5796 + }, + { + "epoch": 0.5576183147364371, + "grad_norm": 1.977246350597133, + "learning_rate": 8.62424477718681e-06, + "loss": 0.9432, + "step": 5797 + }, + { + "epoch": 0.5577145055790689, + "grad_norm": 2.1161302808191773, + "learning_rate": 8.62115904458769e-06, + "loss": 0.9371, + "step": 5798 + }, + { + "epoch": 0.5578106964217007, + "grad_norm": 2.0362268905628307, + "learning_rate": 8.618073445816955e-06, + "loss": 0.9799, + "step": 5799 + }, + { + "epoch": 0.5579068872643325, + "grad_norm": 1.9419607911558157, + "learning_rate": 8.614987981174093e-06, + "loss": 1.0209, + "step": 5800 + }, + { + "epoch": 0.5580030781069643, + "grad_norm": 2.0428043124500896, + "learning_rate": 8.611902650958573e-06, + "loss": 1.0536, + "step": 5801 + }, + { + "epoch": 0.558099268949596, + "grad_norm": 1.9686260588146287, + "learning_rate": 8.608817455469854e-06, + "loss": 0.9847, + "step": 5802 + }, + { + "epoch": 0.5581954597922277, + "grad_norm": 2.096212583281625, + "learning_rate": 8.605732395007378e-06, + "loss": 0.9615, + "step": 5803 + }, + { + "epoch": 0.5582916506348595, + "grad_norm": 2.117134684794037, + "learning_rate": 8.602647469870573e-06, + "loss": 0.967, + "step": 5804 + }, + { + "epoch": 0.5583878414774913, + "grad_norm": 1.868387589742587, + "learning_rate": 8.599562680358868e-06, + "loss": 0.8895, + "step": 5805 + }, + { + "epoch": 0.5584840323201231, + "grad_norm": 1.9704860099766557, + "learning_rate": 8.596478026771658e-06, + "loss": 0.8944, + "step": 5806 + }, + { + "epoch": 0.5585802231627549, + "grad_norm": 2.1914699604064123, + "learning_rate": 8.59339350940835e-06, + "loss": 0.8698, + "step": 5807 + }, + { + "epoch": 0.5586764140053867, + "grad_norm": 2.0594553741675217, + "learning_rate": 8.590309128568303e-06, + "loss": 0.9915, + "step": 5808 + }, + { + "epoch": 0.5587726048480185, + "grad_norm": 2.0106261037779602, + "learning_rate": 8.587224884550895e-06, + "loss": 0.7878, + "step": 5809 + }, + { + "epoch": 0.5588687956906503, + "grad_norm": 2.2474713345964537, + "learning_rate": 8.584140777655476e-06, + "loss": 0.9372, + "step": 5810 + }, + { + "epoch": 0.5589649865332821, + "grad_norm": 2.0622893540341756, + "learning_rate": 8.581056808181385e-06, + "loss": 1.0086, + "step": 5811 + }, + { + "epoch": 0.5590611773759138, + "grad_norm": 2.091945825460006, + "learning_rate": 8.57797297642795e-06, + "loss": 1.0187, + "step": 5812 + }, + { + "epoch": 0.5591573682185456, + "grad_norm": 1.6708864222021447, + "learning_rate": 8.574889282694481e-06, + "loss": 0.867, + "step": 5813 + }, + { + "epoch": 0.5592535590611774, + "grad_norm": 2.4671241123333676, + "learning_rate": 8.571805727280278e-06, + "loss": 0.886, + "step": 5814 + }, + { + "epoch": 0.5593497499038091, + "grad_norm": 7.607539752597827, + "learning_rate": 8.568722310484623e-06, + "loss": 1.0013, + "step": 5815 + }, + { + "epoch": 0.5594459407464409, + "grad_norm": 2.1330095066413177, + "learning_rate": 8.565639032606794e-06, + "loss": 1.0367, + "step": 5816 + }, + { + "epoch": 0.5595421315890727, + "grad_norm": 2.635427130976797, + "learning_rate": 8.562555893946045e-06, + "loss": 0.7722, + "step": 5817 + }, + { + "epoch": 0.5596383224317045, + "grad_norm": 2.0271330810787966, + "learning_rate": 8.559472894801623e-06, + "loss": 0.9637, + "step": 5818 + }, + { + "epoch": 0.5597345132743363, + "grad_norm": 1.9916891510885997, + "learning_rate": 8.556390035472765e-06, + "loss": 0.9381, + "step": 5819 + }, + { + "epoch": 0.5598307041169681, + "grad_norm": 1.8541637544299534, + "learning_rate": 8.553307316258678e-06, + "loss": 1.0085, + "step": 5820 + }, + { + "epoch": 0.5599268949595998, + "grad_norm": 2.2275773213213172, + "learning_rate": 8.550224737458574e-06, + "loss": 1.0245, + "step": 5821 + }, + { + "epoch": 0.5600230858022316, + "grad_norm": 2.1307651795613705, + "learning_rate": 8.547142299371642e-06, + "loss": 0.9549, + "step": 5822 + }, + { + "epoch": 0.5601192766448634, + "grad_norm": 2.252233504318925, + "learning_rate": 8.544060002297056e-06, + "loss": 1.0267, + "step": 5823 + }, + { + "epoch": 0.5602154674874952, + "grad_norm": 1.9447795098912757, + "learning_rate": 8.540977846533986e-06, + "loss": 1.0239, + "step": 5824 + }, + { + "epoch": 0.560311658330127, + "grad_norm": 1.8971167737117594, + "learning_rate": 8.53789583238158e-06, + "loss": 0.971, + "step": 5825 + }, + { + "epoch": 0.5604078491727588, + "grad_norm": 1.9522665053628454, + "learning_rate": 8.534813960138968e-06, + "loss": 0.926, + "step": 5826 + }, + { + "epoch": 0.5605040400153906, + "grad_norm": 5.196025019189655, + "learning_rate": 8.531732230105278e-06, + "loss": 0.8652, + "step": 5827 + }, + { + "epoch": 0.5606002308580224, + "grad_norm": 1.6954747072142045, + "learning_rate": 8.528650642579618e-06, + "loss": 0.9117, + "step": 5828 + }, + { + "epoch": 0.5606964217006541, + "grad_norm": 2.140849378060848, + "learning_rate": 8.525569197861081e-06, + "loss": 1.0163, + "step": 5829 + }, + { + "epoch": 0.5607926125432858, + "grad_norm": 2.182825504288981, + "learning_rate": 8.52248789624875e-06, + "loss": 1.0031, + "step": 5830 + }, + { + "epoch": 0.5608888033859176, + "grad_norm": 2.1138436043402367, + "learning_rate": 8.519406738041689e-06, + "loss": 1.0312, + "step": 5831 + }, + { + "epoch": 0.5609849942285494, + "grad_norm": 1.8050199810468686, + "learning_rate": 8.516325723538949e-06, + "loss": 0.904, + "step": 5832 + }, + { + "epoch": 0.5610811850711812, + "grad_norm": 1.9644177196888246, + "learning_rate": 8.513244853039577e-06, + "loss": 0.9252, + "step": 5833 + }, + { + "epoch": 0.561177375913813, + "grad_norm": 2.147973267600397, + "learning_rate": 8.510164126842591e-06, + "loss": 0.9973, + "step": 5834 + }, + { + "epoch": 0.5612735667564448, + "grad_norm": 2.017563641849335, + "learning_rate": 8.507083545247001e-06, + "loss": 1.0159, + "step": 5835 + }, + { + "epoch": 0.5613697575990766, + "grad_norm": 1.801961560347253, + "learning_rate": 8.504003108551814e-06, + "loss": 0.9928, + "step": 5836 + }, + { + "epoch": 0.5614659484417084, + "grad_norm": 2.3128072539821445, + "learning_rate": 8.500922817056003e-06, + "loss": 0.9652, + "step": 5837 + }, + { + "epoch": 0.5615621392843402, + "grad_norm": 1.7166017393668225, + "learning_rate": 8.497842671058539e-06, + "loss": 0.939, + "step": 5838 + }, + { + "epoch": 0.5616583301269719, + "grad_norm": 1.821297925124965, + "learning_rate": 8.494762670858375e-06, + "loss": 0.9689, + "step": 5839 + }, + { + "epoch": 0.5617545209696037, + "grad_norm": 1.9075529083611407, + "learning_rate": 8.491682816754456e-06, + "loss": 0.952, + "step": 5840 + }, + { + "epoch": 0.5618507118122354, + "grad_norm": 1.9214965245972033, + "learning_rate": 8.488603109045707e-06, + "loss": 0.9827, + "step": 5841 + }, + { + "epoch": 0.5619469026548672, + "grad_norm": 2.0998189802349465, + "learning_rate": 8.485523548031044e-06, + "loss": 0.8017, + "step": 5842 + }, + { + "epoch": 0.562043093497499, + "grad_norm": 1.8720368053299903, + "learning_rate": 8.482444134009356e-06, + "loss": 0.9096, + "step": 5843 + }, + { + "epoch": 0.5621392843401308, + "grad_norm": 1.9683569163230936, + "learning_rate": 8.479364867279529e-06, + "loss": 0.9013, + "step": 5844 + }, + { + "epoch": 0.5622354751827626, + "grad_norm": 1.8980136543531712, + "learning_rate": 8.476285748140437e-06, + "loss": 1.0227, + "step": 5845 + }, + { + "epoch": 0.5623316660253944, + "grad_norm": 1.6852282637697051, + "learning_rate": 8.47320677689093e-06, + "loss": 0.9593, + "step": 5846 + }, + { + "epoch": 0.5624278568680262, + "grad_norm": 1.6442334808251569, + "learning_rate": 8.470127953829852e-06, + "loss": 0.7935, + "step": 5847 + }, + { + "epoch": 0.5625240477106579, + "grad_norm": 1.8677586203242402, + "learning_rate": 8.467049279256034e-06, + "loss": 0.9146, + "step": 5848 + }, + { + "epoch": 0.5626202385532897, + "grad_norm": 2.1016036834058616, + "learning_rate": 8.463970753468279e-06, + "loss": 1.0218, + "step": 5849 + }, + { + "epoch": 0.5627164293959215, + "grad_norm": 1.9498135218467223, + "learning_rate": 8.460892376765387e-06, + "loss": 0.8533, + "step": 5850 + }, + { + "epoch": 0.5628126202385533, + "grad_norm": 2.315202048077853, + "learning_rate": 8.457814149446138e-06, + "loss": 0.8924, + "step": 5851 + }, + { + "epoch": 0.5629088110811851, + "grad_norm": 2.2433785318227537, + "learning_rate": 8.45473607180931e-06, + "loss": 0.9786, + "step": 5852 + }, + { + "epoch": 0.5630050019238169, + "grad_norm": 2.0352246059081125, + "learning_rate": 8.451658144153646e-06, + "loss": 1.0581, + "step": 5853 + }, + { + "epoch": 0.5631011927664487, + "grad_norm": 2.367756477024049, + "learning_rate": 8.448580366777898e-06, + "loss": 0.9098, + "step": 5854 + }, + { + "epoch": 0.5631973836090804, + "grad_norm": 2.123356833051372, + "learning_rate": 8.445502739980776e-06, + "loss": 0.84, + "step": 5855 + }, + { + "epoch": 0.5632935744517122, + "grad_norm": 1.891221279826769, + "learning_rate": 8.442425264061e-06, + "loss": 0.9518, + "step": 5856 + }, + { + "epoch": 0.5633897652943439, + "grad_norm": 2.1768604320222575, + "learning_rate": 8.439347939317264e-06, + "loss": 1.0432, + "step": 5857 + }, + { + "epoch": 0.5634859561369757, + "grad_norm": 1.9985913651575087, + "learning_rate": 8.436270766048245e-06, + "loss": 0.8953, + "step": 5858 + }, + { + "epoch": 0.5635821469796075, + "grad_norm": 2.073824677656413, + "learning_rate": 8.433193744552618e-06, + "loss": 0.9526, + "step": 5859 + }, + { + "epoch": 0.5636783378222393, + "grad_norm": 1.836015399407912, + "learning_rate": 8.430116875129023e-06, + "loss": 0.9449, + "step": 5860 + }, + { + "epoch": 0.5637745286648711, + "grad_norm": 3.1845650378417747, + "learning_rate": 8.427040158076103e-06, + "loss": 1.0023, + "step": 5861 + }, + { + "epoch": 0.5638707195075029, + "grad_norm": 2.1373071931204493, + "learning_rate": 8.42396359369248e-06, + "loss": 0.9477, + "step": 5862 + }, + { + "epoch": 0.5639669103501347, + "grad_norm": 2.004635537973823, + "learning_rate": 8.420887182276757e-06, + "loss": 0.9471, + "step": 5863 + }, + { + "epoch": 0.5640631011927665, + "grad_norm": 2.000742593049672, + "learning_rate": 8.417810924127533e-06, + "loss": 1.0167, + "step": 5864 + }, + { + "epoch": 0.5641592920353983, + "grad_norm": 2.0319779428475973, + "learning_rate": 8.414734819543383e-06, + "loss": 0.9216, + "step": 5865 + }, + { + "epoch": 0.56425548287803, + "grad_norm": 2.894914179324063, + "learning_rate": 8.411658868822866e-06, + "loss": 1.0344, + "step": 5866 + }, + { + "epoch": 0.5643516737206618, + "grad_norm": 1.8324191502530083, + "learning_rate": 8.408583072264529e-06, + "loss": 0.9746, + "step": 5867 + }, + { + "epoch": 0.5644478645632935, + "grad_norm": 1.949009203252653, + "learning_rate": 8.40550743016691e-06, + "loss": 0.9408, + "step": 5868 + }, + { + "epoch": 0.5645440554059253, + "grad_norm": 2.0079446066256503, + "learning_rate": 8.402431942828524e-06, + "loss": 0.9245, + "step": 5869 + }, + { + "epoch": 0.5646402462485571, + "grad_norm": 2.1155693249783933, + "learning_rate": 8.39935661054787e-06, + "loss": 0.9892, + "step": 5870 + }, + { + "epoch": 0.5647364370911889, + "grad_norm": 2.2149461231387266, + "learning_rate": 8.396281433623446e-06, + "loss": 0.8601, + "step": 5871 + }, + { + "epoch": 0.5648326279338207, + "grad_norm": 1.9717181300131739, + "learning_rate": 8.393206412353709e-06, + "loss": 0.9758, + "step": 5872 + }, + { + "epoch": 0.5649288187764525, + "grad_norm": 3.751791172819481, + "learning_rate": 8.390131547037128e-06, + "loss": 0.8363, + "step": 5873 + }, + { + "epoch": 0.5650250096190843, + "grad_norm": 6.547667928811636, + "learning_rate": 8.38705683797214e-06, + "loss": 1.0001, + "step": 5874 + }, + { + "epoch": 0.5651212004617161, + "grad_norm": 2.1740226787481194, + "learning_rate": 8.383982285457173e-06, + "loss": 0.8882, + "step": 5875 + }, + { + "epoch": 0.5652173913043478, + "grad_norm": 2.0178299838029212, + "learning_rate": 8.38090788979064e-06, + "loss": 1.0236, + "step": 5876 + }, + { + "epoch": 0.5653135821469796, + "grad_norm": 2.0826630124320205, + "learning_rate": 8.37783365127094e-06, + "loss": 0.9467, + "step": 5877 + }, + { + "epoch": 0.5654097729896114, + "grad_norm": 1.846603665455576, + "learning_rate": 8.374759570196448e-06, + "loss": 0.9452, + "step": 5878 + }, + { + "epoch": 0.5655059638322432, + "grad_norm": 2.16681333948721, + "learning_rate": 8.371685646865533e-06, + "loss": 0.9372, + "step": 5879 + }, + { + "epoch": 0.565602154674875, + "grad_norm": 2.0052989289026146, + "learning_rate": 8.368611881576547e-06, + "loss": 0.9378, + "step": 5880 + }, + { + "epoch": 0.5656983455175068, + "grad_norm": 1.9893051765604317, + "learning_rate": 8.365538274627825e-06, + "loss": 0.9725, + "step": 5881 + }, + { + "epoch": 0.5657945363601385, + "grad_norm": 2.0836579313351167, + "learning_rate": 8.362464826317687e-06, + "loss": 1.0096, + "step": 5882 + }, + { + "epoch": 0.5658907272027703, + "grad_norm": 2.0358217591414034, + "learning_rate": 8.359391536944436e-06, + "loss": 0.9687, + "step": 5883 + }, + { + "epoch": 0.5659869180454021, + "grad_norm": 1.8796271541897185, + "learning_rate": 8.35631840680636e-06, + "loss": 0.9524, + "step": 5884 + }, + { + "epoch": 0.5660831088880338, + "grad_norm": 1.9399117286742942, + "learning_rate": 8.353245436201735e-06, + "loss": 0.8561, + "step": 5885 + }, + { + "epoch": 0.5661792997306656, + "grad_norm": 1.9181120659091726, + "learning_rate": 8.35017262542882e-06, + "loss": 0.9441, + "step": 5886 + }, + { + "epoch": 0.5662754905732974, + "grad_norm": 2.2561662965023617, + "learning_rate": 8.347099974785857e-06, + "loss": 1.0304, + "step": 5887 + }, + { + "epoch": 0.5663716814159292, + "grad_norm": 1.9942282229588495, + "learning_rate": 8.344027484571075e-06, + "loss": 0.9702, + "step": 5888 + }, + { + "epoch": 0.566467872258561, + "grad_norm": 2.571728517214515, + "learning_rate": 8.340955155082683e-06, + "loss": 0.8852, + "step": 5889 + }, + { + "epoch": 0.5665640631011928, + "grad_norm": 1.9323553327787257, + "learning_rate": 8.337882986618877e-06, + "loss": 0.9103, + "step": 5890 + }, + { + "epoch": 0.5666602539438246, + "grad_norm": 2.214873176348412, + "learning_rate": 8.334810979477837e-06, + "loss": 1.0504, + "step": 5891 + }, + { + "epoch": 0.5667564447864564, + "grad_norm": 1.9874350616096634, + "learning_rate": 8.331739133957729e-06, + "loss": 0.9889, + "step": 5892 + }, + { + "epoch": 0.5668526356290882, + "grad_norm": 1.8155315512905092, + "learning_rate": 8.328667450356703e-06, + "loss": 0.93, + "step": 5893 + }, + { + "epoch": 0.5669488264717198, + "grad_norm": 1.9201400595799805, + "learning_rate": 8.325595928972894e-06, + "loss": 0.9717, + "step": 5894 + }, + { + "epoch": 0.5670450173143516, + "grad_norm": 1.8002886862282328, + "learning_rate": 8.322524570104415e-06, + "loss": 0.936, + "step": 5895 + }, + { + "epoch": 0.5671412081569834, + "grad_norm": 2.095698270706636, + "learning_rate": 8.319453374049367e-06, + "loss": 1.0577, + "step": 5896 + }, + { + "epoch": 0.5672373989996152, + "grad_norm": 1.9486978339019398, + "learning_rate": 8.316382341105841e-06, + "loss": 0.9081, + "step": 5897 + }, + { + "epoch": 0.567333589842247, + "grad_norm": 2.004390721153025, + "learning_rate": 8.313311471571903e-06, + "loss": 0.9946, + "step": 5898 + }, + { + "epoch": 0.5674297806848788, + "grad_norm": 2.0581105490960576, + "learning_rate": 8.31024076574561e-06, + "loss": 1.0316, + "step": 5899 + }, + { + "epoch": 0.5675259715275106, + "grad_norm": 1.9495810652060006, + "learning_rate": 8.307170223925003e-06, + "loss": 0.9521, + "step": 5900 + }, + { + "epoch": 0.5676221623701424, + "grad_norm": 1.9785162399022205, + "learning_rate": 8.304099846408099e-06, + "loss": 1.0216, + "step": 5901 + }, + { + "epoch": 0.5677183532127742, + "grad_norm": 2.5664268289614225, + "learning_rate": 8.301029633492907e-06, + "loss": 0.951, + "step": 5902 + }, + { + "epoch": 0.5678145440554059, + "grad_norm": 1.9463162197257076, + "learning_rate": 8.297959585477413e-06, + "loss": 1.0115, + "step": 5903 + }, + { + "epoch": 0.5679107348980377, + "grad_norm": 2.5265694460327737, + "learning_rate": 8.294889702659602e-06, + "loss": 0.9412, + "step": 5904 + }, + { + "epoch": 0.5680069257406695, + "grad_norm": 2.1302395958048894, + "learning_rate": 8.291819985337427e-06, + "loss": 1.0576, + "step": 5905 + }, + { + "epoch": 0.5681031165833013, + "grad_norm": 1.7413779681426187, + "learning_rate": 8.288750433808828e-06, + "loss": 0.9722, + "step": 5906 + }, + { + "epoch": 0.568199307425933, + "grad_norm": 1.8242190185309992, + "learning_rate": 8.28568104837173e-06, + "loss": 1.0299, + "step": 5907 + }, + { + "epoch": 0.5682954982685648, + "grad_norm": 1.9753583485310584, + "learning_rate": 8.282611829324049e-06, + "loss": 0.9217, + "step": 5908 + }, + { + "epoch": 0.5683916891111966, + "grad_norm": 1.8541507085530877, + "learning_rate": 8.279542776963679e-06, + "loss": 1.0264, + "step": 5909 + }, + { + "epoch": 0.5684878799538284, + "grad_norm": 1.9092259295157654, + "learning_rate": 8.276473891588492e-06, + "loss": 0.9426, + "step": 5910 + }, + { + "epoch": 0.5685840707964602, + "grad_norm": 1.7487120382244086, + "learning_rate": 8.27340517349636e-06, + "loss": 0.9716, + "step": 5911 + }, + { + "epoch": 0.5686802616390919, + "grad_norm": 2.0745381353653767, + "learning_rate": 8.270336622985116e-06, + "loss": 0.9602, + "step": 5912 + }, + { + "epoch": 0.5687764524817237, + "grad_norm": 2.2333723840459303, + "learning_rate": 8.267268240352597e-06, + "loss": 1.1087, + "step": 5913 + }, + { + "epoch": 0.5688726433243555, + "grad_norm": 2.240716141173985, + "learning_rate": 8.264200025896616e-06, + "loss": 1.0061, + "step": 5914 + }, + { + "epoch": 0.5689688341669873, + "grad_norm": 2.4485330169785797, + "learning_rate": 8.261131979914964e-06, + "loss": 1.025, + "step": 5915 + }, + { + "epoch": 0.5690650250096191, + "grad_norm": 1.9848683405868432, + "learning_rate": 8.258064102705428e-06, + "loss": 0.897, + "step": 5916 + }, + { + "epoch": 0.5691612158522509, + "grad_norm": 2.0170509273258777, + "learning_rate": 8.25499639456577e-06, + "loss": 0.9694, + "step": 5917 + }, + { + "epoch": 0.5692574066948827, + "grad_norm": 2.462815134658907, + "learning_rate": 8.251928855793736e-06, + "loss": 0.906, + "step": 5918 + }, + { + "epoch": 0.5693535975375145, + "grad_norm": 1.937330004434201, + "learning_rate": 8.248861486687056e-06, + "loss": 0.9024, + "step": 5919 + }, + { + "epoch": 0.5694497883801463, + "grad_norm": 1.9595842052945072, + "learning_rate": 8.245794287543447e-06, + "loss": 1.0587, + "step": 5920 + }, + { + "epoch": 0.5695459792227779, + "grad_norm": 2.01781100426055, + "learning_rate": 8.242727258660606e-06, + "loss": 0.9928, + "step": 5921 + }, + { + "epoch": 0.5696421700654097, + "grad_norm": 2.2132830140404156, + "learning_rate": 8.239660400336213e-06, + "loss": 0.986, + "step": 5922 + }, + { + "epoch": 0.5697383609080415, + "grad_norm": 2.0408269699385504, + "learning_rate": 8.236593712867942e-06, + "loss": 0.9207, + "step": 5923 + }, + { + "epoch": 0.5698345517506733, + "grad_norm": 1.9010459607709673, + "learning_rate": 8.233527196553428e-06, + "loss": 0.9267, + "step": 5924 + }, + { + "epoch": 0.5699307425933051, + "grad_norm": 2.0617539520044867, + "learning_rate": 8.230460851690311e-06, + "loss": 0.8986, + "step": 5925 + }, + { + "epoch": 0.5700269334359369, + "grad_norm": 2.0599204229969863, + "learning_rate": 8.227394678576204e-06, + "loss": 0.8845, + "step": 5926 + }, + { + "epoch": 0.5701231242785687, + "grad_norm": 2.122287479305512, + "learning_rate": 8.224328677508703e-06, + "loss": 0.9098, + "step": 5927 + }, + { + "epoch": 0.5702193151212005, + "grad_norm": 1.6939542226225532, + "learning_rate": 8.221262848785395e-06, + "loss": 0.895, + "step": 5928 + }, + { + "epoch": 0.5703155059638323, + "grad_norm": 2.1772857914228907, + "learning_rate": 8.218197192703845e-06, + "loss": 0.9411, + "step": 5929 + }, + { + "epoch": 0.570411696806464, + "grad_norm": 2.1248557539826085, + "learning_rate": 8.215131709561597e-06, + "loss": 0.9734, + "step": 5930 + }, + { + "epoch": 0.5705078876490958, + "grad_norm": 1.8794611875630944, + "learning_rate": 8.212066399656181e-06, + "loss": 0.9887, + "step": 5931 + }, + { + "epoch": 0.5706040784917276, + "grad_norm": 2.313221130527497, + "learning_rate": 8.20900126328512e-06, + "loss": 0.9911, + "step": 5932 + }, + { + "epoch": 0.5707002693343594, + "grad_norm": 2.087896449574467, + "learning_rate": 8.205936300745906e-06, + "loss": 0.9592, + "step": 5933 + }, + { + "epoch": 0.5707964601769911, + "grad_norm": 2.4783695042431395, + "learning_rate": 8.202871512336023e-06, + "loss": 1.0213, + "step": 5934 + }, + { + "epoch": 0.5708926510196229, + "grad_norm": 1.955061648677722, + "learning_rate": 8.19980689835293e-06, + "loss": 0.7751, + "step": 5935 + }, + { + "epoch": 0.5709888418622547, + "grad_norm": 1.8426808499400207, + "learning_rate": 8.196742459094079e-06, + "loss": 0.9967, + "step": 5936 + }, + { + "epoch": 0.5710850327048865, + "grad_norm": 2.2804899737534283, + "learning_rate": 8.193678194856896e-06, + "loss": 0.931, + "step": 5937 + }, + { + "epoch": 0.5711812235475183, + "grad_norm": 1.9090697438425916, + "learning_rate": 8.190614105938796e-06, + "loss": 0.9814, + "step": 5938 + }, + { + "epoch": 0.57127741439015, + "grad_norm": 1.8187799572536243, + "learning_rate": 8.187550192637179e-06, + "loss": 0.9729, + "step": 5939 + }, + { + "epoch": 0.5713736052327818, + "grad_norm": 2.0228680755755426, + "learning_rate": 8.184486455249424e-06, + "loss": 0.8891, + "step": 5940 + }, + { + "epoch": 0.5714697960754136, + "grad_norm": 2.2229389839689593, + "learning_rate": 8.181422894072888e-06, + "loss": 0.9782, + "step": 5941 + }, + { + "epoch": 0.5715659869180454, + "grad_norm": 1.813047451021729, + "learning_rate": 8.178359509404916e-06, + "loss": 0.8991, + "step": 5942 + }, + { + "epoch": 0.5716621777606772, + "grad_norm": 2.0959987220386753, + "learning_rate": 8.175296301542838e-06, + "loss": 0.9161, + "step": 5943 + }, + { + "epoch": 0.571758368603309, + "grad_norm": 2.0579099301442927, + "learning_rate": 8.172233270783966e-06, + "loss": 0.8279, + "step": 5944 + }, + { + "epoch": 0.5718545594459408, + "grad_norm": 2.0616741745212157, + "learning_rate": 8.169170417425594e-06, + "loss": 0.9175, + "step": 5945 + }, + { + "epoch": 0.5719507502885726, + "grad_norm": 2.1471793402280595, + "learning_rate": 8.166107741764997e-06, + "loss": 1.0225, + "step": 5946 + }, + { + "epoch": 0.5720469411312044, + "grad_norm": 1.87784023808214, + "learning_rate": 8.16304524409943e-06, + "loss": 0.9794, + "step": 5947 + }, + { + "epoch": 0.572143131973836, + "grad_norm": 1.9873509437368784, + "learning_rate": 8.15998292472614e-06, + "loss": 1.0828, + "step": 5948 + }, + { + "epoch": 0.5722393228164678, + "grad_norm": 1.836517259218951, + "learning_rate": 8.156920783942352e-06, + "loss": 1.0368, + "step": 5949 + }, + { + "epoch": 0.5723355136590996, + "grad_norm": 1.960796485287766, + "learning_rate": 8.153858822045267e-06, + "loss": 0.9663, + "step": 5950 + }, + { + "epoch": 0.5724317045017314, + "grad_norm": 2.3633656420933282, + "learning_rate": 8.150797039332082e-06, + "loss": 1.0778, + "step": 5951 + }, + { + "epoch": 0.5725278953443632, + "grad_norm": 1.775359912982485, + "learning_rate": 8.147735436099967e-06, + "loss": 0.8153, + "step": 5952 + }, + { + "epoch": 0.572624086186995, + "grad_norm": 1.8820750608568264, + "learning_rate": 8.144674012646074e-06, + "loss": 1.0552, + "step": 5953 + }, + { + "epoch": 0.5727202770296268, + "grad_norm": 2.106877192064276, + "learning_rate": 8.141612769267543e-06, + "loss": 1.0872, + "step": 5954 + }, + { + "epoch": 0.5728164678722586, + "grad_norm": 1.9950975297986238, + "learning_rate": 8.138551706261492e-06, + "loss": 1.0404, + "step": 5955 + }, + { + "epoch": 0.5729126587148904, + "grad_norm": 1.7732665236245653, + "learning_rate": 8.135490823925027e-06, + "loss": 0.8631, + "step": 5956 + }, + { + "epoch": 0.5730088495575221, + "grad_norm": 2.4262094404089454, + "learning_rate": 8.132430122555234e-06, + "loss": 1.0125, + "step": 5957 + }, + { + "epoch": 0.5731050404001539, + "grad_norm": 1.753313456401414, + "learning_rate": 8.129369602449176e-06, + "loss": 0.9673, + "step": 5958 + }, + { + "epoch": 0.5732012312427857, + "grad_norm": 2.3467119888051675, + "learning_rate": 8.1263092639039e-06, + "loss": 1.0874, + "step": 5959 + }, + { + "epoch": 0.5732974220854175, + "grad_norm": 1.8639660524867094, + "learning_rate": 8.123249107216446e-06, + "loss": 0.9779, + "step": 5960 + }, + { + "epoch": 0.5733936129280492, + "grad_norm": 2.0008848630759872, + "learning_rate": 8.120189132683824e-06, + "loss": 0.9277, + "step": 5961 + }, + { + "epoch": 0.573489803770681, + "grad_norm": 2.0556784522414926, + "learning_rate": 8.117129340603032e-06, + "loss": 0.9659, + "step": 5962 + }, + { + "epoch": 0.5735859946133128, + "grad_norm": 2.1819097100155638, + "learning_rate": 8.114069731271054e-06, + "loss": 0.8612, + "step": 5963 + }, + { + "epoch": 0.5736821854559446, + "grad_norm": 2.0051922875845087, + "learning_rate": 8.111010304984841e-06, + "loss": 0.9995, + "step": 5964 + }, + { + "epoch": 0.5737783762985764, + "grad_norm": 1.725311944670893, + "learning_rate": 8.107951062041344e-06, + "loss": 0.8998, + "step": 5965 + }, + { + "epoch": 0.5738745671412081, + "grad_norm": 1.950466543655226, + "learning_rate": 8.104892002737488e-06, + "loss": 0.9611, + "step": 5966 + }, + { + "epoch": 0.5739707579838399, + "grad_norm": 2.0528616327302505, + "learning_rate": 8.101833127370177e-06, + "loss": 0.9655, + "step": 5967 + }, + { + "epoch": 0.5740669488264717, + "grad_norm": 2.021569058403387, + "learning_rate": 8.098774436236308e-06, + "loss": 0.9281, + "step": 5968 + }, + { + "epoch": 0.5741631396691035, + "grad_norm": 2.1552127857964716, + "learning_rate": 8.09571592963275e-06, + "loss": 1.0036, + "step": 5969 + }, + { + "epoch": 0.5742593305117353, + "grad_norm": 2.026654527975133, + "learning_rate": 8.092657607856356e-06, + "loss": 1.0288, + "step": 5970 + }, + { + "epoch": 0.5743555213543671, + "grad_norm": 2.4792670513248147, + "learning_rate": 8.089599471203961e-06, + "loss": 1.0444, + "step": 5971 + }, + { + "epoch": 0.5744517121969989, + "grad_norm": 2.006016931971129, + "learning_rate": 8.086541519972388e-06, + "loss": 0.9303, + "step": 5972 + }, + { + "epoch": 0.5745479030396307, + "grad_norm": 1.974573837358113, + "learning_rate": 8.083483754458436e-06, + "loss": 0.8982, + "step": 5973 + }, + { + "epoch": 0.5746440938822625, + "grad_norm": 2.1917091606052357, + "learning_rate": 8.080426174958886e-06, + "loss": 1.0301, + "step": 5974 + }, + { + "epoch": 0.5747402847248941, + "grad_norm": 1.9820989606199138, + "learning_rate": 8.07736878177051e-06, + "loss": 0.9402, + "step": 5975 + }, + { + "epoch": 0.5748364755675259, + "grad_norm": 2.227938426852212, + "learning_rate": 8.074311575190039e-06, + "loss": 0.9358, + "step": 5976 + }, + { + "epoch": 0.5749326664101577, + "grad_norm": 1.9596166715476546, + "learning_rate": 8.071254555514213e-06, + "loss": 0.872, + "step": 5977 + }, + { + "epoch": 0.5750288572527895, + "grad_norm": 2.0151275974850127, + "learning_rate": 8.068197723039738e-06, + "loss": 0.9967, + "step": 5978 + }, + { + "epoch": 0.5751250480954213, + "grad_norm": 2.0174330482800333, + "learning_rate": 8.065141078063307e-06, + "loss": 0.9066, + "step": 5979 + }, + { + "epoch": 0.5752212389380531, + "grad_norm": 1.90401403326163, + "learning_rate": 8.062084620881598e-06, + "loss": 0.9015, + "step": 5980 + }, + { + "epoch": 0.5753174297806849, + "grad_norm": 1.9774398411570957, + "learning_rate": 8.05902835179126e-06, + "loss": 0.9266, + "step": 5981 + }, + { + "epoch": 0.5754136206233167, + "grad_norm": 2.097550164254468, + "learning_rate": 8.055972271088933e-06, + "loss": 0.9913, + "step": 5982 + }, + { + "epoch": 0.5755098114659485, + "grad_norm": 2.0736328964896598, + "learning_rate": 8.052916379071231e-06, + "loss": 0.8514, + "step": 5983 + }, + { + "epoch": 0.5756060023085803, + "grad_norm": 2.4505616645156483, + "learning_rate": 8.049860676034762e-06, + "loss": 0.9625, + "step": 5984 + }, + { + "epoch": 0.575702193151212, + "grad_norm": 2.136053267641486, + "learning_rate": 8.046805162276106e-06, + "loss": 0.9977, + "step": 5985 + }, + { + "epoch": 0.5757983839938438, + "grad_norm": 1.913848501574532, + "learning_rate": 8.043749838091828e-06, + "loss": 1.0118, + "step": 5986 + }, + { + "epoch": 0.5758945748364755, + "grad_norm": 1.8878112628849433, + "learning_rate": 8.040694703778467e-06, + "loss": 1.0855, + "step": 5987 + }, + { + "epoch": 0.5759907656791073, + "grad_norm": 1.8404831029130315, + "learning_rate": 8.037639759632558e-06, + "loss": 0.8935, + "step": 5988 + }, + { + "epoch": 0.5760869565217391, + "grad_norm": 2.119250367564298, + "learning_rate": 8.034585005950605e-06, + "loss": 0.9601, + "step": 5989 + }, + { + "epoch": 0.5761831473643709, + "grad_norm": 1.9244229885331239, + "learning_rate": 8.031530443029099e-06, + "loss": 0.9486, + "step": 5990 + }, + { + "epoch": 0.5762793382070027, + "grad_norm": 1.8757734653883644, + "learning_rate": 8.028476071164512e-06, + "loss": 1.0332, + "step": 5991 + }, + { + "epoch": 0.5763755290496345, + "grad_norm": 1.9523303597561208, + "learning_rate": 8.025421890653303e-06, + "loss": 0.9459, + "step": 5992 + }, + { + "epoch": 0.5764717198922663, + "grad_norm": 2.1892430759484207, + "learning_rate": 8.022367901791895e-06, + "loss": 0.8813, + "step": 5993 + }, + { + "epoch": 0.576567910734898, + "grad_norm": 1.9946812759222625, + "learning_rate": 8.019314104876712e-06, + "loss": 0.9751, + "step": 5994 + }, + { + "epoch": 0.5766641015775298, + "grad_norm": 1.9386454517952294, + "learning_rate": 8.016260500204147e-06, + "loss": 1.0564, + "step": 5995 + }, + { + "epoch": 0.5767602924201616, + "grad_norm": 1.7168875454182304, + "learning_rate": 8.013207088070582e-06, + "loss": 0.9534, + "step": 5996 + }, + { + "epoch": 0.5768564832627934, + "grad_norm": 1.8676824599439557, + "learning_rate": 8.010153868772376e-06, + "loss": 0.967, + "step": 5997 + }, + { + "epoch": 0.5769526741054252, + "grad_norm": 1.8344953316647181, + "learning_rate": 8.007100842605872e-06, + "loss": 0.9721, + "step": 5998 + }, + { + "epoch": 0.577048864948057, + "grad_norm": 2.248212672883676, + "learning_rate": 8.004048009867385e-06, + "loss": 1.0443, + "step": 5999 + }, + { + "epoch": 0.5771450557906888, + "grad_norm": 1.919587360561451, + "learning_rate": 8.000995370853227e-06, + "loss": 0.9173, + "step": 6000 + }, + { + "epoch": 0.5772412466333205, + "grad_norm": 1.8451530509268896, + "learning_rate": 7.99794292585968e-06, + "loss": 0.9582, + "step": 6001 + }, + { + "epoch": 0.5773374374759523, + "grad_norm": 1.9496282510171628, + "learning_rate": 7.994890675183008e-06, + "loss": 0.9549, + "step": 6002 + }, + { + "epoch": 0.577433628318584, + "grad_norm": 2.0575351889934814, + "learning_rate": 7.991838619119467e-06, + "loss": 0.9943, + "step": 6003 + }, + { + "epoch": 0.5775298191612158, + "grad_norm": 1.9401560825944444, + "learning_rate": 7.98878675796527e-06, + "loss": 0.9684, + "step": 6004 + }, + { + "epoch": 0.5776260100038476, + "grad_norm": 2.171595613642783, + "learning_rate": 7.98573509201664e-06, + "loss": 1.0656, + "step": 6005 + }, + { + "epoch": 0.5777222008464794, + "grad_norm": 1.75814379956528, + "learning_rate": 7.98268362156976e-06, + "loss": 0.96, + "step": 6006 + }, + { + "epoch": 0.5778183916891112, + "grad_norm": 1.6831077448706553, + "learning_rate": 7.979632346920802e-06, + "loss": 0.7877, + "step": 6007 + }, + { + "epoch": 0.577914582531743, + "grad_norm": 1.789017191136282, + "learning_rate": 7.976581268365924e-06, + "loss": 0.775, + "step": 6008 + }, + { + "epoch": 0.5780107733743748, + "grad_norm": 1.6832888255436322, + "learning_rate": 7.973530386201257e-06, + "loss": 0.9766, + "step": 6009 + }, + { + "epoch": 0.5781069642170066, + "grad_norm": 1.9630845777687767, + "learning_rate": 7.97047970072291e-06, + "loss": 1.0432, + "step": 6010 + }, + { + "epoch": 0.5782031550596384, + "grad_norm": 1.9673882386886443, + "learning_rate": 7.967429212226982e-06, + "loss": 0.9837, + "step": 6011 + }, + { + "epoch": 0.5782993459022701, + "grad_norm": 2.130162713068122, + "learning_rate": 7.964378921009552e-06, + "loss": 0.7628, + "step": 6012 + }, + { + "epoch": 0.5783955367449019, + "grad_norm": 2.1425684467183976, + "learning_rate": 7.961328827366676e-06, + "loss": 0.9901, + "step": 6013 + }, + { + "epoch": 0.5784917275875336, + "grad_norm": 2.023323472622882, + "learning_rate": 7.958278931594385e-06, + "loss": 0.9767, + "step": 6014 + }, + { + "epoch": 0.5785879184301654, + "grad_norm": 1.8291585805626653, + "learning_rate": 7.955229233988712e-06, + "loss": 0.8315, + "step": 6015 + }, + { + "epoch": 0.5786841092727972, + "grad_norm": 2.0461915887197244, + "learning_rate": 7.952179734845642e-06, + "loss": 0.8627, + "step": 6016 + }, + { + "epoch": 0.578780300115429, + "grad_norm": 2.289768399979209, + "learning_rate": 7.949130434461161e-06, + "loss": 0.9633, + "step": 6017 + }, + { + "epoch": 0.5788764909580608, + "grad_norm": 1.8795440852873608, + "learning_rate": 7.946081333131227e-06, + "loss": 1.0349, + "step": 6018 + }, + { + "epoch": 0.5789726818006926, + "grad_norm": 1.9524451279783797, + "learning_rate": 7.943032431151787e-06, + "loss": 1.0398, + "step": 6019 + }, + { + "epoch": 0.5790688726433244, + "grad_norm": 1.9362144787478734, + "learning_rate": 7.93998372881876e-06, + "loss": 0.8906, + "step": 6020 + }, + { + "epoch": 0.5791650634859561, + "grad_norm": 1.869380265109331, + "learning_rate": 7.936935226428052e-06, + "loss": 0.8452, + "step": 6021 + }, + { + "epoch": 0.5792612543285879, + "grad_norm": 1.907903121304303, + "learning_rate": 7.93388692427554e-06, + "loss": 0.9791, + "step": 6022 + }, + { + "epoch": 0.5793574451712197, + "grad_norm": 1.9414093806992423, + "learning_rate": 7.93083882265709e-06, + "loss": 0.9823, + "step": 6023 + }, + { + "epoch": 0.5794536360138515, + "grad_norm": 1.8108327690182948, + "learning_rate": 7.92779092186855e-06, + "loss": 0.9713, + "step": 6024 + }, + { + "epoch": 0.5795498268564833, + "grad_norm": 1.9848070627871295, + "learning_rate": 7.924743222205744e-06, + "loss": 0.9154, + "step": 6025 + }, + { + "epoch": 0.5796460176991151, + "grad_norm": 1.7325507856764772, + "learning_rate": 7.921695723964473e-06, + "loss": 0.9651, + "step": 6026 + }, + { + "epoch": 0.5797422085417469, + "grad_norm": 2.0064758666794815, + "learning_rate": 7.918648427440532e-06, + "loss": 1.0569, + "step": 6027 + }, + { + "epoch": 0.5798383993843786, + "grad_norm": 1.681975384723234, + "learning_rate": 7.915601332929678e-06, + "loss": 0.847, + "step": 6028 + }, + { + "epoch": 0.5799345902270104, + "grad_norm": 2.083582841760878, + "learning_rate": 7.912554440727662e-06, + "loss": 1.016, + "step": 6029 + }, + { + "epoch": 0.5800307810696421, + "grad_norm": 2.157870284537565, + "learning_rate": 7.90950775113021e-06, + "loss": 0.995, + "step": 6030 + }, + { + "epoch": 0.5801269719122739, + "grad_norm": 1.816418002806487, + "learning_rate": 7.90646126443303e-06, + "loss": 0.9872, + "step": 6031 + }, + { + "epoch": 0.5802231627549057, + "grad_norm": 1.7853413996206349, + "learning_rate": 7.903414980931813e-06, + "loss": 1.0281, + "step": 6032 + }, + { + "epoch": 0.5803193535975375, + "grad_norm": 1.7219339290334754, + "learning_rate": 7.900368900922222e-06, + "loss": 0.8379, + "step": 6033 + }, + { + "epoch": 0.5804155444401693, + "grad_norm": 2.313359415115661, + "learning_rate": 7.897323024699907e-06, + "loss": 0.8913, + "step": 6034 + }, + { + "epoch": 0.5805117352828011, + "grad_norm": 2.0494716274593228, + "learning_rate": 7.894277352560495e-06, + "loss": 0.9668, + "step": 6035 + }, + { + "epoch": 0.5806079261254329, + "grad_norm": 1.9841137365570236, + "learning_rate": 7.8912318847996e-06, + "loss": 0.9265, + "step": 6036 + }, + { + "epoch": 0.5807041169680647, + "grad_norm": 2.3826313308612974, + "learning_rate": 7.888186621712805e-06, + "loss": 0.8952, + "step": 6037 + }, + { + "epoch": 0.5808003078106965, + "grad_norm": 1.79032502963878, + "learning_rate": 7.885141563595685e-06, + "loss": 0.9017, + "step": 6038 + }, + { + "epoch": 0.5808964986533282, + "grad_norm": 2.075085801281545, + "learning_rate": 7.882096710743778e-06, + "loss": 0.9578, + "step": 6039 + }, + { + "epoch": 0.58099268949596, + "grad_norm": 1.9292994530434218, + "learning_rate": 7.879052063452626e-06, + "loss": 0.9857, + "step": 6040 + }, + { + "epoch": 0.5810888803385917, + "grad_norm": 1.8657778092721078, + "learning_rate": 7.876007622017731e-06, + "loss": 0.9856, + "step": 6041 + }, + { + "epoch": 0.5811850711812235, + "grad_norm": 1.5485468490832077, + "learning_rate": 7.872963386734584e-06, + "loss": 0.7455, + "step": 6042 + }, + { + "epoch": 0.5812812620238553, + "grad_norm": 1.9125006934370203, + "learning_rate": 7.869919357898655e-06, + "loss": 0.9221, + "step": 6043 + }, + { + "epoch": 0.5813774528664871, + "grad_norm": 1.9120046076387234, + "learning_rate": 7.866875535805394e-06, + "loss": 1.0261, + "step": 6044 + }, + { + "epoch": 0.5814736437091189, + "grad_norm": 2.090341774643499, + "learning_rate": 7.863831920750229e-06, + "loss": 0.9872, + "step": 6045 + }, + { + "epoch": 0.5815698345517507, + "grad_norm": 1.90914214101541, + "learning_rate": 7.860788513028566e-06, + "loss": 0.9735, + "step": 6046 + }, + { + "epoch": 0.5816660253943825, + "grad_norm": 1.7537198039118709, + "learning_rate": 7.857745312935796e-06, + "loss": 0.8197, + "step": 6047 + }, + { + "epoch": 0.5817622162370142, + "grad_norm": 2.09829994058122, + "learning_rate": 7.85470232076729e-06, + "loss": 1.0584, + "step": 6048 + }, + { + "epoch": 0.581858407079646, + "grad_norm": 1.8839831290827942, + "learning_rate": 7.851659536818392e-06, + "loss": 0.9784, + "step": 6049 + }, + { + "epoch": 0.5819545979222778, + "grad_norm": 1.7779834751748222, + "learning_rate": 7.848616961384442e-06, + "loss": 0.9233, + "step": 6050 + }, + { + "epoch": 0.5820507887649096, + "grad_norm": 2.2613006451306497, + "learning_rate": 7.845574594760732e-06, + "loss": 0.7984, + "step": 6051 + }, + { + "epoch": 0.5821469796075414, + "grad_norm": 2.089682191977548, + "learning_rate": 7.842532437242559e-06, + "loss": 0.9047, + "step": 6052 + }, + { + "epoch": 0.5822431704501732, + "grad_norm": 1.8669537915330126, + "learning_rate": 7.83949048912519e-06, + "loss": 0.9768, + "step": 6053 + }, + { + "epoch": 0.582339361292805, + "grad_norm": 2.2083072360598304, + "learning_rate": 7.83644875070387e-06, + "loss": 0.9707, + "step": 6054 + }, + { + "epoch": 0.5824355521354367, + "grad_norm": 2.0987880100246423, + "learning_rate": 7.83340722227383e-06, + "loss": 1.0596, + "step": 6055 + }, + { + "epoch": 0.5825317429780685, + "grad_norm": 1.9024175978297917, + "learning_rate": 7.83036590413027e-06, + "loss": 0.9149, + "step": 6056 + }, + { + "epoch": 0.5826279338207002, + "grad_norm": 1.7659161693652599, + "learning_rate": 7.827324796568382e-06, + "loss": 0.8341, + "step": 6057 + }, + { + "epoch": 0.582724124663332, + "grad_norm": 1.7210187123308938, + "learning_rate": 7.824283899883327e-06, + "loss": 0.9349, + "step": 6058 + }, + { + "epoch": 0.5828203155059638, + "grad_norm": 2.171185340018266, + "learning_rate": 7.821243214370253e-06, + "loss": 1.0097, + "step": 6059 + }, + { + "epoch": 0.5829165063485956, + "grad_norm": 1.9723069290200823, + "learning_rate": 7.818202740324287e-06, + "loss": 0.9611, + "step": 6060 + }, + { + "epoch": 0.5830126971912274, + "grad_norm": 2.0934617135262683, + "learning_rate": 7.81516247804053e-06, + "loss": 1.0464, + "step": 6061 + }, + { + "epoch": 0.5831088880338592, + "grad_norm": 2.0402392059820875, + "learning_rate": 7.812122427814068e-06, + "loss": 0.8698, + "step": 6062 + }, + { + "epoch": 0.583205078876491, + "grad_norm": 1.9218869634717246, + "learning_rate": 7.809082589939957e-06, + "loss": 0.9113, + "step": 6063 + }, + { + "epoch": 0.5833012697191228, + "grad_norm": 2.150672754631416, + "learning_rate": 7.806042964713248e-06, + "loss": 0.8516, + "step": 6064 + }, + { + "epoch": 0.5833974605617546, + "grad_norm": 1.6909359567342388, + "learning_rate": 7.803003552428959e-06, + "loss": 0.917, + "step": 6065 + }, + { + "epoch": 0.5834936514043862, + "grad_norm": 2.004121581077939, + "learning_rate": 7.79996435338209e-06, + "loss": 1.0078, + "step": 6066 + }, + { + "epoch": 0.583589842247018, + "grad_norm": 1.7747554374016223, + "learning_rate": 7.796925367867625e-06, + "loss": 0.9506, + "step": 6067 + }, + { + "epoch": 0.5836860330896498, + "grad_norm": 1.7928917084246268, + "learning_rate": 7.793886596180521e-06, + "loss": 0.8417, + "step": 6068 + }, + { + "epoch": 0.5837822239322816, + "grad_norm": 1.9927135818210961, + "learning_rate": 7.790848038615718e-06, + "loss": 0.9814, + "step": 6069 + }, + { + "epoch": 0.5838784147749134, + "grad_norm": 1.8100048612918758, + "learning_rate": 7.787809695468134e-06, + "loss": 0.8611, + "step": 6070 + }, + { + "epoch": 0.5839746056175452, + "grad_norm": 1.8089724728107701, + "learning_rate": 7.784771567032665e-06, + "loss": 0.8974, + "step": 6071 + }, + { + "epoch": 0.584070796460177, + "grad_norm": 1.9672276210672592, + "learning_rate": 7.78173365360419e-06, + "loss": 0.9624, + "step": 6072 + }, + { + "epoch": 0.5841669873028088, + "grad_norm": 2.045270814251043, + "learning_rate": 7.778695955477566e-06, + "loss": 1.0023, + "step": 6073 + }, + { + "epoch": 0.5842631781454406, + "grad_norm": 2.0726104890431025, + "learning_rate": 7.775658472947623e-06, + "loss": 0.991, + "step": 6074 + }, + { + "epoch": 0.5843593689880723, + "grad_norm": 2.1477338592037776, + "learning_rate": 7.772621206309174e-06, + "loss": 1.0397, + "step": 6075 + }, + { + "epoch": 0.5844555598307041, + "grad_norm": 1.9894337262116024, + "learning_rate": 7.769584155857019e-06, + "loss": 0.9564, + "step": 6076 + }, + { + "epoch": 0.5845517506733359, + "grad_norm": 2.711499951746927, + "learning_rate": 7.766547321885925e-06, + "loss": 1.0785, + "step": 6077 + }, + { + "epoch": 0.5846479415159677, + "grad_norm": 2.0567674186231297, + "learning_rate": 7.763510704690645e-06, + "loss": 0.9378, + "step": 6078 + }, + { + "epoch": 0.5847441323585995, + "grad_norm": 2.0193507967387245, + "learning_rate": 7.760474304565902e-06, + "loss": 0.9947, + "step": 6079 + }, + { + "epoch": 0.5848403232012312, + "grad_norm": 1.8069661187757735, + "learning_rate": 7.757438121806414e-06, + "loss": 0.9634, + "step": 6080 + }, + { + "epoch": 0.584936514043863, + "grad_norm": 2.076818947000926, + "learning_rate": 7.754402156706866e-06, + "loss": 0.9278, + "step": 6081 + }, + { + "epoch": 0.5850327048864948, + "grad_norm": 2.0411318245871106, + "learning_rate": 7.75136640956192e-06, + "loss": 0.9266, + "step": 6082 + }, + { + "epoch": 0.5851288957291266, + "grad_norm": 2.0093822657909, + "learning_rate": 7.748330880666227e-06, + "loss": 0.9889, + "step": 6083 + }, + { + "epoch": 0.5852250865717584, + "grad_norm": 1.9855602606548595, + "learning_rate": 7.745295570314412e-06, + "loss": 0.8519, + "step": 6084 + }, + { + "epoch": 0.5853212774143901, + "grad_norm": 2.077342609287016, + "learning_rate": 7.742260478801072e-06, + "loss": 0.8571, + "step": 6085 + }, + { + "epoch": 0.5854174682570219, + "grad_norm": 1.9355944315126974, + "learning_rate": 7.739225606420793e-06, + "loss": 0.8618, + "step": 6086 + }, + { + "epoch": 0.5855136590996537, + "grad_norm": 1.9167961754354241, + "learning_rate": 7.736190953468132e-06, + "loss": 0.941, + "step": 6087 + }, + { + "epoch": 0.5856098499422855, + "grad_norm": 2.0635686427846567, + "learning_rate": 7.733156520237633e-06, + "loss": 0.9186, + "step": 6088 + }, + { + "epoch": 0.5857060407849173, + "grad_norm": 1.9150420914267527, + "learning_rate": 7.730122307023811e-06, + "loss": 0.8551, + "step": 6089 + }, + { + "epoch": 0.5858022316275491, + "grad_norm": 1.7508559658004337, + "learning_rate": 7.727088314121165e-06, + "loss": 0.9725, + "step": 6090 + }, + { + "epoch": 0.5858984224701809, + "grad_norm": 2.0641267147465285, + "learning_rate": 7.724054541824164e-06, + "loss": 1.1007, + "step": 6091 + }, + { + "epoch": 0.5859946133128127, + "grad_norm": 2.2953021391955164, + "learning_rate": 7.721020990427268e-06, + "loss": 0.9781, + "step": 6092 + }, + { + "epoch": 0.5860908041554445, + "grad_norm": 1.9019117915875259, + "learning_rate": 7.717987660224907e-06, + "loss": 0.9211, + "step": 6093 + }, + { + "epoch": 0.5861869949980761, + "grad_norm": 1.993184109928366, + "learning_rate": 7.714954551511489e-06, + "loss": 0.9956, + "step": 6094 + }, + { + "epoch": 0.5862831858407079, + "grad_norm": 1.9051410827056579, + "learning_rate": 7.71192166458141e-06, + "loss": 0.8846, + "step": 6095 + }, + { + "epoch": 0.5863793766833397, + "grad_norm": 1.6996818695529303, + "learning_rate": 7.708888999729036e-06, + "loss": 0.831, + "step": 6096 + }, + { + "epoch": 0.5864755675259715, + "grad_norm": 1.9022565479605573, + "learning_rate": 7.705856557248708e-06, + "loss": 0.8395, + "step": 6097 + }, + { + "epoch": 0.5865717583686033, + "grad_norm": 2.0271946946378727, + "learning_rate": 7.702824337434756e-06, + "loss": 0.8963, + "step": 6098 + }, + { + "epoch": 0.5866679492112351, + "grad_norm": 2.1445912892462924, + "learning_rate": 7.699792340581477e-06, + "loss": 0.8884, + "step": 6099 + }, + { + "epoch": 0.5867641400538669, + "grad_norm": 2.0236875897079383, + "learning_rate": 7.69676056698316e-06, + "loss": 0.986, + "step": 6100 + }, + { + "epoch": 0.5868603308964987, + "grad_norm": 2.4200066922538075, + "learning_rate": 7.693729016934055e-06, + "loss": 0.9213, + "step": 6101 + }, + { + "epoch": 0.5869565217391305, + "grad_norm": 2.1013923324552164, + "learning_rate": 7.690697690728417e-06, + "loss": 1.0326, + "step": 6102 + }, + { + "epoch": 0.5870527125817622, + "grad_norm": 1.9382653662969271, + "learning_rate": 7.687666588660444e-06, + "loss": 0.9299, + "step": 6103 + }, + { + "epoch": 0.587148903424394, + "grad_norm": 1.9888027775530182, + "learning_rate": 7.68463571102434e-06, + "loss": 0.9064, + "step": 6104 + }, + { + "epoch": 0.5872450942670258, + "grad_norm": 2.131335703626778, + "learning_rate": 7.681605058114275e-06, + "loss": 0.984, + "step": 6105 + }, + { + "epoch": 0.5873412851096576, + "grad_norm": 1.9179615218007966, + "learning_rate": 7.678574630224399e-06, + "loss": 0.8344, + "step": 6106 + }, + { + "epoch": 0.5874374759522893, + "grad_norm": 2.07254803323473, + "learning_rate": 7.675544427648848e-06, + "loss": 0.8744, + "step": 6107 + }, + { + "epoch": 0.5875336667949211, + "grad_norm": 1.9524223601699258, + "learning_rate": 7.672514450681721e-06, + "loss": 1.022, + "step": 6108 + }, + { + "epoch": 0.5876298576375529, + "grad_norm": 2.0040832117839313, + "learning_rate": 7.669484699617106e-06, + "loss": 0.8573, + "step": 6109 + }, + { + "epoch": 0.5877260484801847, + "grad_norm": 1.9789403802746848, + "learning_rate": 7.666455174749066e-06, + "loss": 0.9199, + "step": 6110 + }, + { + "epoch": 0.5878222393228165, + "grad_norm": 1.9894757730644033, + "learning_rate": 7.663425876371645e-06, + "loss": 0.8786, + "step": 6111 + }, + { + "epoch": 0.5879184301654482, + "grad_norm": 1.9778922163635761, + "learning_rate": 7.66039680477886e-06, + "loss": 0.904, + "step": 6112 + }, + { + "epoch": 0.58801462100808, + "grad_norm": 1.8377051373427482, + "learning_rate": 7.657367960264712e-06, + "loss": 0.9131, + "step": 6113 + }, + { + "epoch": 0.5881108118507118, + "grad_norm": 2.139059449114341, + "learning_rate": 7.654339343123173e-06, + "loss": 0.9393, + "step": 6114 + }, + { + "epoch": 0.5882070026933436, + "grad_norm": 2.165401813154412, + "learning_rate": 7.651310953648193e-06, + "loss": 0.8402, + "step": 6115 + }, + { + "epoch": 0.5883031935359754, + "grad_norm": 1.9597964923773772, + "learning_rate": 7.648282792133711e-06, + "loss": 1.0105, + "step": 6116 + }, + { + "epoch": 0.5883993843786072, + "grad_norm": 1.7309315990557914, + "learning_rate": 7.645254858873631e-06, + "loss": 0.9465, + "step": 6117 + }, + { + "epoch": 0.588495575221239, + "grad_norm": 2.3095950776715917, + "learning_rate": 7.642227154161841e-06, + "loss": 0.8487, + "step": 6118 + }, + { + "epoch": 0.5885917660638708, + "grad_norm": 1.9457776098573385, + "learning_rate": 7.63919967829221e-06, + "loss": 0.9988, + "step": 6119 + }, + { + "epoch": 0.5886879569065026, + "grad_norm": 1.974475918672698, + "learning_rate": 7.636172431558575e-06, + "loss": 0.8916, + "step": 6120 + }, + { + "epoch": 0.5887841477491342, + "grad_norm": 1.9788743623581917, + "learning_rate": 7.633145414254759e-06, + "loss": 0.9349, + "step": 6121 + }, + { + "epoch": 0.588880338591766, + "grad_norm": 1.8278402830575564, + "learning_rate": 7.630118626674557e-06, + "loss": 0.8467, + "step": 6122 + }, + { + "epoch": 0.5889765294343978, + "grad_norm": 2.110080731329572, + "learning_rate": 7.6270920691117475e-06, + "loss": 0.8954, + "step": 6123 + }, + { + "epoch": 0.5890727202770296, + "grad_norm": 2.037912463838371, + "learning_rate": 7.6240657418600846e-06, + "loss": 0.8886, + "step": 6124 + }, + { + "epoch": 0.5891689111196614, + "grad_norm": 2.193019723415255, + "learning_rate": 7.621039645213302e-06, + "loss": 0.9876, + "step": 6125 + }, + { + "epoch": 0.5892651019622932, + "grad_norm": 2.0637016799814725, + "learning_rate": 7.618013779465101e-06, + "loss": 0.9578, + "step": 6126 + }, + { + "epoch": 0.589361292804925, + "grad_norm": 2.200300889244726, + "learning_rate": 7.614988144909171e-06, + "loss": 1.0543, + "step": 6127 + }, + { + "epoch": 0.5894574836475568, + "grad_norm": 1.960258644129958, + "learning_rate": 7.611962741839178e-06, + "loss": 0.8721, + "step": 6128 + }, + { + "epoch": 0.5895536744901886, + "grad_norm": 2.065985917482778, + "learning_rate": 7.608937570548764e-06, + "loss": 0.9414, + "step": 6129 + }, + { + "epoch": 0.5896498653328203, + "grad_norm": 2.190857950738911, + "learning_rate": 7.6059126313315466e-06, + "loss": 0.9289, + "step": 6130 + }, + { + "epoch": 0.5897460561754521, + "grad_norm": 1.9252844508362228, + "learning_rate": 7.602887924481117e-06, + "loss": 1.0121, + "step": 6131 + }, + { + "epoch": 0.5898422470180839, + "grad_norm": 2.1644698657802723, + "learning_rate": 7.599863450291056e-06, + "loss": 0.7978, + "step": 6132 + }, + { + "epoch": 0.5899384378607156, + "grad_norm": 2.399144961150317, + "learning_rate": 7.596839209054914e-06, + "loss": 0.9842, + "step": 6133 + }, + { + "epoch": 0.5900346287033474, + "grad_norm": 1.838403169289553, + "learning_rate": 7.593815201066215e-06, + "loss": 0.9019, + "step": 6134 + }, + { + "epoch": 0.5901308195459792, + "grad_norm": 1.9824028800513824, + "learning_rate": 7.59079142661847e-06, + "loss": 0.9478, + "step": 6135 + }, + { + "epoch": 0.590227010388611, + "grad_norm": 2.1447002020132455, + "learning_rate": 7.587767886005164e-06, + "loss": 0.9475, + "step": 6136 + }, + { + "epoch": 0.5903232012312428, + "grad_norm": 2.0553066617428666, + "learning_rate": 7.584744579519752e-06, + "loss": 0.9321, + "step": 6137 + }, + { + "epoch": 0.5904193920738746, + "grad_norm": 2.0967017265492878, + "learning_rate": 7.581721507455672e-06, + "loss": 0.9224, + "step": 6138 + }, + { + "epoch": 0.5905155829165063, + "grad_norm": 1.9887940099226609, + "learning_rate": 7.578698670106341e-06, + "loss": 1.0447, + "step": 6139 + }, + { + "epoch": 0.5906117737591381, + "grad_norm": 2.0535760754552923, + "learning_rate": 7.575676067765154e-06, + "loss": 0.9615, + "step": 6140 + }, + { + "epoch": 0.5907079646017699, + "grad_norm": 2.2478419612474667, + "learning_rate": 7.572653700725477e-06, + "loss": 0.9393, + "step": 6141 + }, + { + "epoch": 0.5908041554444017, + "grad_norm": 2.0008407206064804, + "learning_rate": 7.569631569280662e-06, + "loss": 0.9485, + "step": 6142 + }, + { + "epoch": 0.5909003462870335, + "grad_norm": 2.1312985533151094, + "learning_rate": 7.566609673724024e-06, + "loss": 1.0832, + "step": 6143 + }, + { + "epoch": 0.5909965371296653, + "grad_norm": 1.979686364055458, + "learning_rate": 7.563588014348871e-06, + "loss": 0.8995, + "step": 6144 + }, + { + "epoch": 0.5910927279722971, + "grad_norm": 2.046673538185124, + "learning_rate": 7.560566591448479e-06, + "loss": 1.0043, + "step": 6145 + }, + { + "epoch": 0.5911889188149289, + "grad_norm": 2.188520053405082, + "learning_rate": 7.5575454053161e-06, + "loss": 0.9277, + "step": 6146 + }, + { + "epoch": 0.5912851096575606, + "grad_norm": 1.957474726029068, + "learning_rate": 7.554524456244973e-06, + "loss": 0.9827, + "step": 6147 + }, + { + "epoch": 0.5913813005001923, + "grad_norm": 1.7881570610832118, + "learning_rate": 7.551503744528304e-06, + "loss": 1.0439, + "step": 6148 + }, + { + "epoch": 0.5914774913428241, + "grad_norm": 1.8708129579137613, + "learning_rate": 7.548483270459277e-06, + "loss": 0.9254, + "step": 6149 + }, + { + "epoch": 0.5915736821854559, + "grad_norm": 1.837353729881757, + "learning_rate": 7.545463034331054e-06, + "loss": 0.9876, + "step": 6150 + }, + { + "epoch": 0.5916698730280877, + "grad_norm": 1.843856985477198, + "learning_rate": 7.542443036436779e-06, + "loss": 0.951, + "step": 6151 + }, + { + "epoch": 0.5917660638707195, + "grad_norm": 2.057747276340816, + "learning_rate": 7.539423277069568e-06, + "loss": 0.8257, + "step": 6152 + }, + { + "epoch": 0.5918622547133513, + "grad_norm": 1.8426356041370524, + "learning_rate": 7.536403756522517e-06, + "loss": 1.0197, + "step": 6153 + }, + { + "epoch": 0.5919584455559831, + "grad_norm": 1.835599618286274, + "learning_rate": 7.53338447508869e-06, + "loss": 0.9297, + "step": 6154 + }, + { + "epoch": 0.5920546363986149, + "grad_norm": 2.0902162660869905, + "learning_rate": 7.530365433061134e-06, + "loss": 0.9639, + "step": 6155 + }, + { + "epoch": 0.5921508272412467, + "grad_norm": 2.102925887628142, + "learning_rate": 7.52734663073288e-06, + "loss": 0.948, + "step": 6156 + }, + { + "epoch": 0.5922470180838784, + "grad_norm": 2.172457947800006, + "learning_rate": 7.5243280683969254e-06, + "loss": 1.0782, + "step": 6157 + }, + { + "epoch": 0.5923432089265102, + "grad_norm": 1.8218262315724143, + "learning_rate": 7.521309746346246e-06, + "loss": 0.9344, + "step": 6158 + }, + { + "epoch": 0.592439399769142, + "grad_norm": 2.2361579535396103, + "learning_rate": 7.5182916648738e-06, + "loss": 1.0002, + "step": 6159 + }, + { + "epoch": 0.5925355906117737, + "grad_norm": 2.224452980594507, + "learning_rate": 7.515273824272516e-06, + "loss": 0.9926, + "step": 6160 + }, + { + "epoch": 0.5926317814544055, + "grad_norm": 1.7678314105562034, + "learning_rate": 7.5122562248352995e-06, + "loss": 1.0009, + "step": 6161 + }, + { + "epoch": 0.5927279722970373, + "grad_norm": 2.03552216072052, + "learning_rate": 7.509238866855033e-06, + "loss": 0.999, + "step": 6162 + }, + { + "epoch": 0.5928241631396691, + "grad_norm": 2.0744231764375636, + "learning_rate": 7.506221750624584e-06, + "loss": 0.9883, + "step": 6163 + }, + { + "epoch": 0.5929203539823009, + "grad_norm": 2.1479721112407524, + "learning_rate": 7.503204876436785e-06, + "loss": 0.8605, + "step": 6164 + }, + { + "epoch": 0.5930165448249327, + "grad_norm": 2.0302512652445697, + "learning_rate": 7.5001882445844535e-06, + "loss": 1.0086, + "step": 6165 + }, + { + "epoch": 0.5931127356675644, + "grad_norm": 2.0332581766942597, + "learning_rate": 7.497171855360372e-06, + "loss": 0.8576, + "step": 6166 + }, + { + "epoch": 0.5932089265101962, + "grad_norm": 2.194476503434704, + "learning_rate": 7.494155709057312e-06, + "loss": 1.0286, + "step": 6167 + }, + { + "epoch": 0.593305117352828, + "grad_norm": 1.9975993512250778, + "learning_rate": 7.491139805968018e-06, + "loss": 0.9538, + "step": 6168 + }, + { + "epoch": 0.5934013081954598, + "grad_norm": 2.071752805883625, + "learning_rate": 7.4881241463852065e-06, + "loss": 1.0105, + "step": 6169 + }, + { + "epoch": 0.5934974990380916, + "grad_norm": 1.715383848026076, + "learning_rate": 7.485108730601571e-06, + "loss": 0.7566, + "step": 6170 + }, + { + "epoch": 0.5935936898807234, + "grad_norm": 1.9018456762509215, + "learning_rate": 7.4820935589097935e-06, + "loss": 0.8967, + "step": 6171 + }, + { + "epoch": 0.5936898807233552, + "grad_norm": 1.848300953052826, + "learning_rate": 7.4790786316025125e-06, + "loss": 0.9349, + "step": 6172 + }, + { + "epoch": 0.593786071565987, + "grad_norm": 1.8915277701928064, + "learning_rate": 7.476063948972354e-06, + "loss": 0.8012, + "step": 6173 + }, + { + "epoch": 0.5938822624086187, + "grad_norm": 2.0407726734876244, + "learning_rate": 7.473049511311921e-06, + "loss": 1.0112, + "step": 6174 + }, + { + "epoch": 0.5939784532512504, + "grad_norm": 1.8206189532569244, + "learning_rate": 7.470035318913789e-06, + "loss": 0.8714, + "step": 6175 + }, + { + "epoch": 0.5940746440938822, + "grad_norm": 2.000315557584964, + "learning_rate": 7.467021372070515e-06, + "loss": 0.991, + "step": 6176 + }, + { + "epoch": 0.594170834936514, + "grad_norm": 1.9178263668777793, + "learning_rate": 7.464007671074629e-06, + "loss": 0.9247, + "step": 6177 + }, + { + "epoch": 0.5942670257791458, + "grad_norm": 1.9675302631587392, + "learning_rate": 7.46099421621863e-06, + "loss": 0.9852, + "step": 6178 + }, + { + "epoch": 0.5943632166217776, + "grad_norm": 2.4838807453100995, + "learning_rate": 7.457981007795001e-06, + "loss": 0.9029, + "step": 6179 + }, + { + "epoch": 0.5944594074644094, + "grad_norm": 2.062238794218959, + "learning_rate": 7.4549680460962044e-06, + "loss": 0.9528, + "step": 6180 + }, + { + "epoch": 0.5945555983070412, + "grad_norm": 2.2033739047023815, + "learning_rate": 7.451955331414671e-06, + "loss": 0.9648, + "step": 6181 + }, + { + "epoch": 0.594651789149673, + "grad_norm": 2.0611212806764883, + "learning_rate": 7.448942864042819e-06, + "loss": 1.0317, + "step": 6182 + }, + { + "epoch": 0.5947479799923048, + "grad_norm": 2.038047430922752, + "learning_rate": 7.445930644273018e-06, + "loss": 0.9054, + "step": 6183 + }, + { + "epoch": 0.5948441708349365, + "grad_norm": 2.1947638285269195, + "learning_rate": 7.4429186723976425e-06, + "loss": 1.0048, + "step": 6184 + }, + { + "epoch": 0.5949403616775683, + "grad_norm": 2.1926968157452555, + "learning_rate": 7.439906948709026e-06, + "loss": 1.1428, + "step": 6185 + }, + { + "epoch": 0.5950365525202, + "grad_norm": 2.315034042195352, + "learning_rate": 7.43689547349948e-06, + "loss": 1.0373, + "step": 6186 + }, + { + "epoch": 0.5951327433628318, + "grad_norm": 1.7694420689141623, + "learning_rate": 7.4338842470613e-06, + "loss": 1.0379, + "step": 6187 + }, + { + "epoch": 0.5952289342054636, + "grad_norm": 2.140929791367133, + "learning_rate": 7.43087326968675e-06, + "loss": 0.9327, + "step": 6188 + }, + { + "epoch": 0.5953251250480954, + "grad_norm": 1.977828650751382, + "learning_rate": 7.427862541668067e-06, + "loss": 0.958, + "step": 6189 + }, + { + "epoch": 0.5954213158907272, + "grad_norm": 1.891893449072317, + "learning_rate": 7.42485206329747e-06, + "loss": 0.9937, + "step": 6190 + }, + { + "epoch": 0.595517506733359, + "grad_norm": 2.1647017921891885, + "learning_rate": 7.421841834867154e-06, + "loss": 0.8423, + "step": 6191 + }, + { + "epoch": 0.5956136975759908, + "grad_norm": 2.0305117612250587, + "learning_rate": 7.418831856669286e-06, + "loss": 0.993, + "step": 6192 + }, + { + "epoch": 0.5957098884186226, + "grad_norm": 1.9953573259148585, + "learning_rate": 7.415822128996008e-06, + "loss": 0.9229, + "step": 6193 + }, + { + "epoch": 0.5958060792612543, + "grad_norm": 2.1448497490284897, + "learning_rate": 7.41281265213945e-06, + "loss": 0.9808, + "step": 6194 + }, + { + "epoch": 0.5959022701038861, + "grad_norm": 1.912241825813576, + "learning_rate": 7.409803426391693e-06, + "loss": 0.8878, + "step": 6195 + }, + { + "epoch": 0.5959984609465179, + "grad_norm": 1.813134861737907, + "learning_rate": 7.406794452044816e-06, + "loss": 0.8774, + "step": 6196 + }, + { + "epoch": 0.5960946517891497, + "grad_norm": 1.860720452531209, + "learning_rate": 7.403785729390868e-06, + "loss": 0.7909, + "step": 6197 + }, + { + "epoch": 0.5961908426317815, + "grad_norm": 1.8890110544803418, + "learning_rate": 7.400777258721865e-06, + "loss": 0.9206, + "step": 6198 + }, + { + "epoch": 0.5962870334744133, + "grad_norm": 2.102102201537754, + "learning_rate": 7.397769040329811e-06, + "loss": 0.9142, + "step": 6199 + }, + { + "epoch": 0.596383224317045, + "grad_norm": 1.7787946261454222, + "learning_rate": 7.394761074506679e-06, + "loss": 1.0059, + "step": 6200 + }, + { + "epoch": 0.5964794151596768, + "grad_norm": 1.9811741815099067, + "learning_rate": 7.391753361544413e-06, + "loss": 0.9351, + "step": 6201 + }, + { + "epoch": 0.5965756060023086, + "grad_norm": 2.3548781386219546, + "learning_rate": 7.3887459017349405e-06, + "loss": 0.9778, + "step": 6202 + }, + { + "epoch": 0.5966717968449403, + "grad_norm": 1.7946814504156208, + "learning_rate": 7.3857386953701624e-06, + "loss": 0.9483, + "step": 6203 + }, + { + "epoch": 0.5967679876875721, + "grad_norm": 2.132542738938019, + "learning_rate": 7.382731742741953e-06, + "loss": 1.052, + "step": 6204 + }, + { + "epoch": 0.5968641785302039, + "grad_norm": 1.9585623371599972, + "learning_rate": 7.379725044142166e-06, + "loss": 0.8687, + "step": 6205 + }, + { + "epoch": 0.5969603693728357, + "grad_norm": 1.9845629115692562, + "learning_rate": 7.376718599862621e-06, + "loss": 0.9919, + "step": 6206 + }, + { + "epoch": 0.5970565602154675, + "grad_norm": 1.9622884086083334, + "learning_rate": 7.373712410195121e-06, + "loss": 0.9715, + "step": 6207 + }, + { + "epoch": 0.5971527510580993, + "grad_norm": 2.1932701920977387, + "learning_rate": 7.370706475431446e-06, + "loss": 0.9351, + "step": 6208 + }, + { + "epoch": 0.5972489419007311, + "grad_norm": 1.8447726206879629, + "learning_rate": 7.367700795863346e-06, + "loss": 0.8757, + "step": 6209 + }, + { + "epoch": 0.5973451327433629, + "grad_norm": 1.9373530440496118, + "learning_rate": 7.364695371782547e-06, + "loss": 1.0072, + "step": 6210 + }, + { + "epoch": 0.5974413235859947, + "grad_norm": 1.9333767637169141, + "learning_rate": 7.361690203480756e-06, + "loss": 0.9961, + "step": 6211 + }, + { + "epoch": 0.5975375144286263, + "grad_norm": 1.95818548529336, + "learning_rate": 7.358685291249644e-06, + "loss": 0.9795, + "step": 6212 + }, + { + "epoch": 0.5976337052712581, + "grad_norm": 1.8928010081696658, + "learning_rate": 7.355680635380867e-06, + "loss": 0.9102, + "step": 6213 + }, + { + "epoch": 0.5977298961138899, + "grad_norm": 1.8581231377167038, + "learning_rate": 7.352676236166051e-06, + "loss": 1.0039, + "step": 6214 + }, + { + "epoch": 0.5978260869565217, + "grad_norm": 1.921964981544614, + "learning_rate": 7.349672093896802e-06, + "loss": 0.809, + "step": 6215 + }, + { + "epoch": 0.5979222777991535, + "grad_norm": 2.1957524851785357, + "learning_rate": 7.346668208864695e-06, + "loss": 0.8167, + "step": 6216 + }, + { + "epoch": 0.5980184686417853, + "grad_norm": 1.7287743158271542, + "learning_rate": 7.3436645813612865e-06, + "loss": 0.9076, + "step": 6217 + }, + { + "epoch": 0.5981146594844171, + "grad_norm": 2.0275875428640604, + "learning_rate": 7.3406612116781e-06, + "loss": 0.9189, + "step": 6218 + }, + { + "epoch": 0.5982108503270489, + "grad_norm": 2.18187932080458, + "learning_rate": 7.337658100106636e-06, + "loss": 0.9346, + "step": 6219 + }, + { + "epoch": 0.5983070411696807, + "grad_norm": 2.128569812209351, + "learning_rate": 7.33465524693838e-06, + "loss": 0.9216, + "step": 6220 + }, + { + "epoch": 0.5984032320123124, + "grad_norm": 1.9125022425613911, + "learning_rate": 7.331652652464779e-06, + "loss": 0.9414, + "step": 6221 + }, + { + "epoch": 0.5984994228549442, + "grad_norm": 1.9847069181568566, + "learning_rate": 7.328650316977265e-06, + "loss": 0.8952, + "step": 6222 + }, + { + "epoch": 0.598595613697576, + "grad_norm": 2.054232558480421, + "learning_rate": 7.32564824076724e-06, + "loss": 0.8835, + "step": 6223 + }, + { + "epoch": 0.5986918045402078, + "grad_norm": 1.808672264909347, + "learning_rate": 7.322646424126079e-06, + "loss": 0.7983, + "step": 6224 + }, + { + "epoch": 0.5987879953828396, + "grad_norm": 1.8727638423340986, + "learning_rate": 7.3196448673451325e-06, + "loss": 0.929, + "step": 6225 + }, + { + "epoch": 0.5988841862254713, + "grad_norm": 1.8947221133440075, + "learning_rate": 7.316643570715729e-06, + "loss": 0.8912, + "step": 6226 + }, + { + "epoch": 0.5989803770681031, + "grad_norm": 1.8312761391769534, + "learning_rate": 7.313642534529173e-06, + "loss": 0.9177, + "step": 6227 + }, + { + "epoch": 0.5990765679107349, + "grad_norm": 2.1065033440892, + "learning_rate": 7.310641759076742e-06, + "loss": 0.9854, + "step": 6228 + }, + { + "epoch": 0.5991727587533667, + "grad_norm": 2.404796512979238, + "learning_rate": 7.3076412446496804e-06, + "loss": 0.9152, + "step": 6229 + }, + { + "epoch": 0.5992689495959984, + "grad_norm": 2.079576866019174, + "learning_rate": 7.304640991539216e-06, + "loss": 0.8969, + "step": 6230 + }, + { + "epoch": 0.5993651404386302, + "grad_norm": 1.9038821127814558, + "learning_rate": 7.301641000036552e-06, + "loss": 0.9902, + "step": 6231 + }, + { + "epoch": 0.599461331281262, + "grad_norm": 2.205986472379369, + "learning_rate": 7.2986412704328625e-06, + "loss": 1.0022, + "step": 6232 + }, + { + "epoch": 0.5995575221238938, + "grad_norm": 2.0438846740539525, + "learning_rate": 7.295641803019296e-06, + "loss": 1.0332, + "step": 6233 + }, + { + "epoch": 0.5996537129665256, + "grad_norm": 2.1111638509932096, + "learning_rate": 7.292642598086982e-06, + "loss": 0.913, + "step": 6234 + }, + { + "epoch": 0.5997499038091574, + "grad_norm": 2.0797829795781397, + "learning_rate": 7.289643655927008e-06, + "loss": 1.0483, + "step": 6235 + }, + { + "epoch": 0.5998460946517892, + "grad_norm": 1.8200078220364821, + "learning_rate": 7.286644976830457e-06, + "loss": 0.9774, + "step": 6236 + }, + { + "epoch": 0.599942285494421, + "grad_norm": 2.0229931811121618, + "learning_rate": 7.2836465610883725e-06, + "loss": 0.9901, + "step": 6237 + }, + { + "epoch": 0.6000384763370528, + "grad_norm": 1.8867719082395076, + "learning_rate": 7.280648408991775e-06, + "loss": 0.8698, + "step": 6238 + }, + { + "epoch": 0.6001346671796844, + "grad_norm": 2.2432683773464754, + "learning_rate": 7.277650520831666e-06, + "loss": 1.0008, + "step": 6239 + }, + { + "epoch": 0.6002308580223162, + "grad_norm": 2.1221485182161115, + "learning_rate": 7.274652896899015e-06, + "loss": 0.9507, + "step": 6240 + }, + { + "epoch": 0.600327048864948, + "grad_norm": 1.6587338237300926, + "learning_rate": 7.271655537484766e-06, + "loss": 0.8061, + "step": 6241 + }, + { + "epoch": 0.6004232397075798, + "grad_norm": 2.0855982079876534, + "learning_rate": 7.268658442879834e-06, + "loss": 0.937, + "step": 6242 + }, + { + "epoch": 0.6005194305502116, + "grad_norm": 1.9689141953262852, + "learning_rate": 7.265661613375119e-06, + "loss": 0.9324, + "step": 6243 + }, + { + "epoch": 0.6006156213928434, + "grad_norm": 2.160282090021336, + "learning_rate": 7.262665049261489e-06, + "loss": 1.0106, + "step": 6244 + }, + { + "epoch": 0.6007118122354752, + "grad_norm": 2.0663611068663115, + "learning_rate": 7.259668750829783e-06, + "loss": 0.9987, + "step": 6245 + }, + { + "epoch": 0.600808003078107, + "grad_norm": 2.6744546962683895, + "learning_rate": 7.256672718370824e-06, + "loss": 0.9295, + "step": 6246 + }, + { + "epoch": 0.6009041939207388, + "grad_norm": 2.265924482666179, + "learning_rate": 7.2536769521753925e-06, + "loss": 1.0155, + "step": 6247 + }, + { + "epoch": 0.6010003847633705, + "grad_norm": 1.9057922340966467, + "learning_rate": 7.250681452534261e-06, + "loss": 0.9535, + "step": 6248 + }, + { + "epoch": 0.6010965756060023, + "grad_norm": 1.9865026584984844, + "learning_rate": 7.247686219738166e-06, + "loss": 0.9027, + "step": 6249 + }, + { + "epoch": 0.6011927664486341, + "grad_norm": 2.155459586555438, + "learning_rate": 7.2446912540778196e-06, + "loss": 0.9727, + "step": 6250 + }, + { + "epoch": 0.6012889572912659, + "grad_norm": 1.8133445740912544, + "learning_rate": 7.2416965558439116e-06, + "loss": 0.9388, + "step": 6251 + }, + { + "epoch": 0.6013851481338977, + "grad_norm": 1.877042139670687, + "learning_rate": 7.238702125327106e-06, + "loss": 0.9333, + "step": 6252 + }, + { + "epoch": 0.6014813389765294, + "grad_norm": 1.9051408025953003, + "learning_rate": 7.2357079628180304e-06, + "loss": 1.0038, + "step": 6253 + }, + { + "epoch": 0.6015775298191612, + "grad_norm": 1.9954682616216979, + "learning_rate": 7.232714068607296e-06, + "loss": 0.9682, + "step": 6254 + }, + { + "epoch": 0.601673720661793, + "grad_norm": 1.861905989369991, + "learning_rate": 7.2297204429854905e-06, + "loss": 0.8388, + "step": 6255 + }, + { + "epoch": 0.6017699115044248, + "grad_norm": 2.1678782636459615, + "learning_rate": 7.226727086243168e-06, + "loss": 1.0267, + "step": 6256 + }, + { + "epoch": 0.6018661023470565, + "grad_norm": 2.2517693653675965, + "learning_rate": 7.2237339986708635e-06, + "loss": 0.8948, + "step": 6257 + }, + { + "epoch": 0.6019622931896883, + "grad_norm": 2.117859388510199, + "learning_rate": 7.220741180559074e-06, + "loss": 1.0172, + "step": 6258 + }, + { + "epoch": 0.6020584840323201, + "grad_norm": 1.936155643231025, + "learning_rate": 7.2177486321982825e-06, + "loss": 0.9169, + "step": 6259 + }, + { + "epoch": 0.6021546748749519, + "grad_norm": 1.9393177147467078, + "learning_rate": 7.214756353878942e-06, + "loss": 0.8362, + "step": 6260 + }, + { + "epoch": 0.6022508657175837, + "grad_norm": 2.2137988529686834, + "learning_rate": 7.211764345891479e-06, + "loss": 0.8907, + "step": 6261 + }, + { + "epoch": 0.6023470565602155, + "grad_norm": 1.9211275282929, + "learning_rate": 7.208772608526293e-06, + "loss": 1.0259, + "step": 6262 + }, + { + "epoch": 0.6024432474028473, + "grad_norm": 1.864085308541365, + "learning_rate": 7.2057811420737614e-06, + "loss": 0.9368, + "step": 6263 + }, + { + "epoch": 0.6025394382454791, + "grad_norm": 2.0004521710248584, + "learning_rate": 7.202789946824227e-06, + "loss": 1.0159, + "step": 6264 + }, + { + "epoch": 0.6026356290881109, + "grad_norm": 2.077838969765707, + "learning_rate": 7.1997990230680104e-06, + "loss": 0.8508, + "step": 6265 + }, + { + "epoch": 0.6027318199307425, + "grad_norm": 1.8158058523044271, + "learning_rate": 7.1968083710954075e-06, + "loss": 0.9308, + "step": 6266 + }, + { + "epoch": 0.6028280107733743, + "grad_norm": 1.7229097761862664, + "learning_rate": 7.193817991196691e-06, + "loss": 0.861, + "step": 6267 + }, + { + "epoch": 0.6029242016160061, + "grad_norm": 1.9318114628955372, + "learning_rate": 7.1908278836621e-06, + "loss": 1.0218, + "step": 6268 + }, + { + "epoch": 0.6030203924586379, + "grad_norm": 1.8374972283306852, + "learning_rate": 7.18783804878185e-06, + "loss": 0.9282, + "step": 6269 + }, + { + "epoch": 0.6031165833012697, + "grad_norm": 2.040048701549786, + "learning_rate": 7.184848486846128e-06, + "loss": 0.9726, + "step": 6270 + }, + { + "epoch": 0.6032127741439015, + "grad_norm": 2.1208785670774315, + "learning_rate": 7.181859198145101e-06, + "loss": 0.8797, + "step": 6271 + }, + { + "epoch": 0.6033089649865333, + "grad_norm": 1.785002764545986, + "learning_rate": 7.178870182968904e-06, + "loss": 0.9149, + "step": 6272 + }, + { + "epoch": 0.6034051558291651, + "grad_norm": 2.005509023077572, + "learning_rate": 7.175881441607643e-06, + "loss": 0.909, + "step": 6273 + }, + { + "epoch": 0.6035013466717969, + "grad_norm": 1.9977895235949845, + "learning_rate": 7.1728929743514065e-06, + "loss": 0.9047, + "step": 6274 + }, + { + "epoch": 0.6035975375144286, + "grad_norm": 2.0792564448804542, + "learning_rate": 7.16990478149025e-06, + "loss": 0.9845, + "step": 6275 + }, + { + "epoch": 0.6036937283570604, + "grad_norm": 2.0648289204043264, + "learning_rate": 7.166916863314199e-06, + "loss": 0.9019, + "step": 6276 + }, + { + "epoch": 0.6037899191996922, + "grad_norm": 1.8737170510402381, + "learning_rate": 7.163929220113262e-06, + "loss": 0.954, + "step": 6277 + }, + { + "epoch": 0.603886110042324, + "grad_norm": 2.0527398530988945, + "learning_rate": 7.1609418521774095e-06, + "loss": 0.9221, + "step": 6278 + }, + { + "epoch": 0.6039823008849557, + "grad_norm": 2.1416929475466566, + "learning_rate": 7.157954759796597e-06, + "loss": 0.8732, + "step": 6279 + }, + { + "epoch": 0.6040784917275875, + "grad_norm": 1.9299293038288787, + "learning_rate": 7.154967943260748e-06, + "loss": 0.815, + "step": 6280 + }, + { + "epoch": 0.6041746825702193, + "grad_norm": 2.1349133441891714, + "learning_rate": 7.151981402859753e-06, + "loss": 1.0091, + "step": 6281 + }, + { + "epoch": 0.6042708734128511, + "grad_norm": 1.9191016093456208, + "learning_rate": 7.148995138883483e-06, + "loss": 0.779, + "step": 6282 + }, + { + "epoch": 0.6043670642554829, + "grad_norm": 2.16653552952168, + "learning_rate": 7.146009151621785e-06, + "loss": 0.9644, + "step": 6283 + }, + { + "epoch": 0.6044632550981146, + "grad_norm": 2.4213754918039063, + "learning_rate": 7.143023441364471e-06, + "loss": 0.9143, + "step": 6284 + }, + { + "epoch": 0.6045594459407464, + "grad_norm": 1.8997239688335985, + "learning_rate": 7.14003800840133e-06, + "loss": 1.001, + "step": 6285 + }, + { + "epoch": 0.6046556367833782, + "grad_norm": 2.050503886029238, + "learning_rate": 7.13705285302213e-06, + "loss": 0.9453, + "step": 6286 + }, + { + "epoch": 0.60475182762601, + "grad_norm": 2.263518351917789, + "learning_rate": 7.1340679755165965e-06, + "loss": 1.0059, + "step": 6287 + }, + { + "epoch": 0.6048480184686418, + "grad_norm": 1.8996468089739253, + "learning_rate": 7.131083376174441e-06, + "loss": 0.8829, + "step": 6288 + }, + { + "epoch": 0.6049442093112736, + "grad_norm": 2.3420633836775915, + "learning_rate": 7.128099055285348e-06, + "loss": 0.9398, + "step": 6289 + }, + { + "epoch": 0.6050404001539054, + "grad_norm": 1.6188850424903445, + "learning_rate": 7.125115013138966e-06, + "loss": 0.7488, + "step": 6290 + }, + { + "epoch": 0.6051365909965372, + "grad_norm": 2.1123285609014983, + "learning_rate": 7.122131250024927e-06, + "loss": 0.9782, + "step": 6291 + }, + { + "epoch": 0.605232781839169, + "grad_norm": 2.0079389076355083, + "learning_rate": 7.119147766232832e-06, + "loss": 0.9169, + "step": 6292 + }, + { + "epoch": 0.6053289726818007, + "grad_norm": 1.8197293359891915, + "learning_rate": 7.116164562052248e-06, + "loss": 0.8684, + "step": 6293 + }, + { + "epoch": 0.6054251635244324, + "grad_norm": 1.8793920534736273, + "learning_rate": 7.113181637772721e-06, + "loss": 0.93, + "step": 6294 + }, + { + "epoch": 0.6055213543670642, + "grad_norm": 2.1175212130902503, + "learning_rate": 7.110198993683774e-06, + "loss": 1.0784, + "step": 6295 + }, + { + "epoch": 0.605617545209696, + "grad_norm": 1.9321406994409578, + "learning_rate": 7.107216630074895e-06, + "loss": 0.9685, + "step": 6296 + }, + { + "epoch": 0.6057137360523278, + "grad_norm": 2.2885832428465376, + "learning_rate": 7.104234547235549e-06, + "loss": 0.899, + "step": 6297 + }, + { + "epoch": 0.6058099268949596, + "grad_norm": 2.0793279055199747, + "learning_rate": 7.1012527454551795e-06, + "loss": 0.9157, + "step": 6298 + }, + { + "epoch": 0.6059061177375914, + "grad_norm": 2.2103641430524554, + "learning_rate": 7.098271225023184e-06, + "loss": 0.9314, + "step": 6299 + }, + { + "epoch": 0.6060023085802232, + "grad_norm": 1.8555460959337544, + "learning_rate": 7.09528998622895e-06, + "loss": 0.9367, + "step": 6300 + }, + { + "epoch": 0.606098499422855, + "grad_norm": 2.0122518588547718, + "learning_rate": 7.092309029361833e-06, + "loss": 0.8927, + "step": 6301 + }, + { + "epoch": 0.6061946902654868, + "grad_norm": 1.8475671037668253, + "learning_rate": 7.089328354711159e-06, + "loss": 0.9871, + "step": 6302 + }, + { + "epoch": 0.6062908811081185, + "grad_norm": 2.0679987172956946, + "learning_rate": 7.086347962566233e-06, + "loss": 1.0813, + "step": 6303 + }, + { + "epoch": 0.6063870719507503, + "grad_norm": 2.140923535738934, + "learning_rate": 7.083367853216323e-06, + "loss": 0.9073, + "step": 6304 + }, + { + "epoch": 0.606483262793382, + "grad_norm": 2.0151550863734777, + "learning_rate": 7.0803880269506755e-06, + "loss": 0.9564, + "step": 6305 + }, + { + "epoch": 0.6065794536360138, + "grad_norm": 1.8827474128373578, + "learning_rate": 7.077408484058505e-06, + "loss": 0.9749, + "step": 6306 + }, + { + "epoch": 0.6066756444786456, + "grad_norm": 2.0346419929271433, + "learning_rate": 7.074429224829007e-06, + "loss": 1.0379, + "step": 6307 + }, + { + "epoch": 0.6067718353212774, + "grad_norm": 2.098872111836252, + "learning_rate": 7.071450249551342e-06, + "loss": 0.9216, + "step": 6308 + }, + { + "epoch": 0.6068680261639092, + "grad_norm": 1.8059701873838976, + "learning_rate": 7.068471558514649e-06, + "loss": 0.9891, + "step": 6309 + }, + { + "epoch": 0.606964217006541, + "grad_norm": 2.171479213249999, + "learning_rate": 7.065493152008026e-06, + "loss": 0.8715, + "step": 6310 + }, + { + "epoch": 0.6070604078491728, + "grad_norm": 2.009231950484062, + "learning_rate": 7.062515030320562e-06, + "loss": 1.0058, + "step": 6311 + }, + { + "epoch": 0.6071565986918045, + "grad_norm": 1.8781800926795063, + "learning_rate": 7.059537193741306e-06, + "loss": 0.9564, + "step": 6312 + }, + { + "epoch": 0.6072527895344363, + "grad_norm": 2.0652662807704916, + "learning_rate": 7.056559642559282e-06, + "loss": 0.887, + "step": 6313 + }, + { + "epoch": 0.6073489803770681, + "grad_norm": 2.2827723077223374, + "learning_rate": 7.053582377063489e-06, + "loss": 0.98, + "step": 6314 + }, + { + "epoch": 0.6074451712196999, + "grad_norm": 2.2303335988877677, + "learning_rate": 7.050605397542898e-06, + "loss": 0.9535, + "step": 6315 + }, + { + "epoch": 0.6075413620623317, + "grad_norm": 1.987486774592881, + "learning_rate": 7.047628704286446e-06, + "loss": 1.0193, + "step": 6316 + }, + { + "epoch": 0.6076375529049635, + "grad_norm": 1.8180517065615194, + "learning_rate": 7.044652297583048e-06, + "loss": 0.8656, + "step": 6317 + }, + { + "epoch": 0.6077337437475953, + "grad_norm": 1.8975353857677295, + "learning_rate": 7.041676177721588e-06, + "loss": 0.9339, + "step": 6318 + }, + { + "epoch": 0.607829934590227, + "grad_norm": 1.880475442388071, + "learning_rate": 7.038700344990928e-06, + "loss": 0.9716, + "step": 6319 + }, + { + "epoch": 0.6079261254328588, + "grad_norm": 2.254988990732961, + "learning_rate": 7.035724799679898e-06, + "loss": 1.0289, + "step": 6320 + }, + { + "epoch": 0.6080223162754905, + "grad_norm": 1.7220997285867363, + "learning_rate": 7.032749542077301e-06, + "loss": 0.9628, + "step": 6321 + }, + { + "epoch": 0.6081185071181223, + "grad_norm": 1.790126867416525, + "learning_rate": 7.029774572471904e-06, + "loss": 0.8527, + "step": 6322 + }, + { + "epoch": 0.6082146979607541, + "grad_norm": 1.9899091576532788, + "learning_rate": 7.026799891152462e-06, + "loss": 0.9232, + "step": 6323 + }, + { + "epoch": 0.6083108888033859, + "grad_norm": 2.2233049771578632, + "learning_rate": 7.023825498407689e-06, + "loss": 1.1049, + "step": 6324 + }, + { + "epoch": 0.6084070796460177, + "grad_norm": 1.7806066102868, + "learning_rate": 7.020851394526276e-06, + "loss": 0.8842, + "step": 6325 + }, + { + "epoch": 0.6085032704886495, + "grad_norm": 2.065336693919807, + "learning_rate": 7.0178775797968855e-06, + "loss": 0.8952, + "step": 6326 + }, + { + "epoch": 0.6085994613312813, + "grad_norm": 1.882432126566884, + "learning_rate": 7.014904054508156e-06, + "loss": 0.9213, + "step": 6327 + }, + { + "epoch": 0.6086956521739131, + "grad_norm": 2.2901163117446637, + "learning_rate": 7.011930818948688e-06, + "loss": 0.9394, + "step": 6328 + }, + { + "epoch": 0.6087918430165449, + "grad_norm": 2.066105339763021, + "learning_rate": 7.0089578734070604e-06, + "loss": 0.9369, + "step": 6329 + }, + { + "epoch": 0.6088880338591766, + "grad_norm": 2.193774000707313, + "learning_rate": 7.005985218171825e-06, + "loss": 0.9747, + "step": 6330 + }, + { + "epoch": 0.6089842247018084, + "grad_norm": 2.126511788195862, + "learning_rate": 7.0030128535315035e-06, + "loss": 0.9278, + "step": 6331 + }, + { + "epoch": 0.6090804155444401, + "grad_norm": 1.6544316284474478, + "learning_rate": 7.000040779774591e-06, + "loss": 0.8003, + "step": 6332 + }, + { + "epoch": 0.6091766063870719, + "grad_norm": 1.6533081223832313, + "learning_rate": 6.997068997189548e-06, + "loss": 0.8578, + "step": 6333 + }, + { + "epoch": 0.6092727972297037, + "grad_norm": 2.305432188724679, + "learning_rate": 6.994097506064812e-06, + "loss": 0.9225, + "step": 6334 + }, + { + "epoch": 0.6093689880723355, + "grad_norm": 1.8530105291795538, + "learning_rate": 6.991126306688798e-06, + "loss": 0.9948, + "step": 6335 + }, + { + "epoch": 0.6094651789149673, + "grad_norm": 1.8557835534812366, + "learning_rate": 6.9881553993498805e-06, + "loss": 0.8865, + "step": 6336 + }, + { + "epoch": 0.6095613697575991, + "grad_norm": 2.4133022195879312, + "learning_rate": 6.985184784336414e-06, + "loss": 0.94, + "step": 6337 + }, + { + "epoch": 0.6096575606002309, + "grad_norm": 1.8113094588437546, + "learning_rate": 6.9822144619367275e-06, + "loss": 0.962, + "step": 6338 + }, + { + "epoch": 0.6097537514428626, + "grad_norm": 1.7360854857522685, + "learning_rate": 6.979244432439104e-06, + "loss": 0.8343, + "step": 6339 + }, + { + "epoch": 0.6098499422854944, + "grad_norm": 2.101261445693936, + "learning_rate": 6.97627469613182e-06, + "loss": 0.8085, + "step": 6340 + }, + { + "epoch": 0.6099461331281262, + "grad_norm": 2.0127786488310084, + "learning_rate": 6.973305253303111e-06, + "loss": 0.9928, + "step": 6341 + }, + { + "epoch": 0.610042323970758, + "grad_norm": 1.9812176159314234, + "learning_rate": 6.970336104241186e-06, + "loss": 0.8864, + "step": 6342 + }, + { + "epoch": 0.6101385148133898, + "grad_norm": 2.105850576396237, + "learning_rate": 6.967367249234229e-06, + "loss": 0.9937, + "step": 6343 + }, + { + "epoch": 0.6102347056560216, + "grad_norm": 2.002618656153176, + "learning_rate": 6.9643986885703955e-06, + "loss": 0.9395, + "step": 6344 + }, + { + "epoch": 0.6103308964986534, + "grad_norm": 1.5425953121324638, + "learning_rate": 6.961430422537803e-06, + "loss": 0.7287, + "step": 6345 + }, + { + "epoch": 0.6104270873412851, + "grad_norm": 2.115679084784277, + "learning_rate": 6.958462451424547e-06, + "loss": 0.9877, + "step": 6346 + }, + { + "epoch": 0.6105232781839169, + "grad_norm": 2.126036507033292, + "learning_rate": 6.955494775518701e-06, + "loss": 0.961, + "step": 6347 + }, + { + "epoch": 0.6106194690265486, + "grad_norm": 2.093366387058406, + "learning_rate": 6.952527395108302e-06, + "loss": 0.9033, + "step": 6348 + }, + { + "epoch": 0.6107156598691804, + "grad_norm": 2.174639965477029, + "learning_rate": 6.949560310481353e-06, + "loss": 0.994, + "step": 6349 + }, + { + "epoch": 0.6108118507118122, + "grad_norm": 2.031538728052787, + "learning_rate": 6.9465935219258504e-06, + "loss": 0.9773, + "step": 6350 + }, + { + "epoch": 0.610908041554444, + "grad_norm": 1.7447237180922803, + "learning_rate": 6.943627029729728e-06, + "loss": 0.8761, + "step": 6351 + }, + { + "epoch": 0.6110042323970758, + "grad_norm": 2.0840893620737604, + "learning_rate": 6.9406608341809215e-06, + "loss": 1.0326, + "step": 6352 + }, + { + "epoch": 0.6111004232397076, + "grad_norm": 1.9838807048043625, + "learning_rate": 6.937694935567319e-06, + "loss": 1.0206, + "step": 6353 + }, + { + "epoch": 0.6111966140823394, + "grad_norm": 1.84800172338746, + "learning_rate": 6.934729334176793e-06, + "loss": 1.0091, + "step": 6354 + }, + { + "epoch": 0.6112928049249712, + "grad_norm": 2.0406415976061885, + "learning_rate": 6.9317640302971786e-06, + "loss": 0.8661, + "step": 6355 + }, + { + "epoch": 0.611388995767603, + "grad_norm": 1.8982101027923282, + "learning_rate": 6.928799024216282e-06, + "loss": 0.9039, + "step": 6356 + }, + { + "epoch": 0.6114851866102347, + "grad_norm": 2.186791450318259, + "learning_rate": 6.925834316221882e-06, + "loss": 0.928, + "step": 6357 + }, + { + "epoch": 0.6115813774528664, + "grad_norm": 2.155521232562027, + "learning_rate": 6.92286990660173e-06, + "loss": 0.9686, + "step": 6358 + }, + { + "epoch": 0.6116775682954982, + "grad_norm": 1.8955463357086273, + "learning_rate": 6.919905795643549e-06, + "loss": 0.9606, + "step": 6359 + }, + { + "epoch": 0.61177375913813, + "grad_norm": 1.884931856275791, + "learning_rate": 6.91694198363503e-06, + "loss": 0.8934, + "step": 6360 + }, + { + "epoch": 0.6118699499807618, + "grad_norm": 1.964701071636186, + "learning_rate": 6.91397847086384e-06, + "loss": 0.9396, + "step": 6361 + }, + { + "epoch": 0.6119661408233936, + "grad_norm": 1.830549475301447, + "learning_rate": 6.911015257617606e-06, + "loss": 0.7823, + "step": 6362 + }, + { + "epoch": 0.6120623316660254, + "grad_norm": 1.8023581238774244, + "learning_rate": 6.90805234418394e-06, + "loss": 0.9051, + "step": 6363 + }, + { + "epoch": 0.6121585225086572, + "grad_norm": 2.2943021372408454, + "learning_rate": 6.905089730850416e-06, + "loss": 0.9741, + "step": 6364 + }, + { + "epoch": 0.612254713351289, + "grad_norm": 1.723275928785974, + "learning_rate": 6.902127417904578e-06, + "loss": 0.8749, + "step": 6365 + }, + { + "epoch": 0.6123509041939207, + "grad_norm": 1.972487021361981, + "learning_rate": 6.8991654056339505e-06, + "loss": 0.9246, + "step": 6366 + }, + { + "epoch": 0.6124470950365525, + "grad_norm": 2.0619415186572136, + "learning_rate": 6.89620369432602e-06, + "loss": 0.9719, + "step": 6367 + }, + { + "epoch": 0.6125432858791843, + "grad_norm": 2.042800191497414, + "learning_rate": 6.893242284268244e-06, + "loss": 0.9872, + "step": 6368 + }, + { + "epoch": 0.6126394767218161, + "grad_norm": 1.8834316369034503, + "learning_rate": 6.890281175748053e-06, + "loss": 0.939, + "step": 6369 + }, + { + "epoch": 0.6127356675644479, + "grad_norm": 1.780726761900578, + "learning_rate": 6.887320369052848e-06, + "loss": 0.9099, + "step": 6370 + }, + { + "epoch": 0.6128318584070797, + "grad_norm": 2.293864104016348, + "learning_rate": 6.884359864470003e-06, + "loss": 0.9445, + "step": 6371 + }, + { + "epoch": 0.6129280492497114, + "grad_norm": 1.7919529120479287, + "learning_rate": 6.8813996622868584e-06, + "loss": 0.9501, + "step": 6372 + }, + { + "epoch": 0.6130242400923432, + "grad_norm": 1.861472043456133, + "learning_rate": 6.878439762790731e-06, + "loss": 0.8888, + "step": 6373 + }, + { + "epoch": 0.613120430934975, + "grad_norm": 2.1048766180653455, + "learning_rate": 6.8754801662688964e-06, + "loss": 0.9627, + "step": 6374 + }, + { + "epoch": 0.6132166217776067, + "grad_norm": 1.9969241283583028, + "learning_rate": 6.8725208730086165e-06, + "loss": 0.9566, + "step": 6375 + }, + { + "epoch": 0.6133128126202385, + "grad_norm": 1.9921724341968774, + "learning_rate": 6.869561883297116e-06, + "loss": 1.1284, + "step": 6376 + }, + { + "epoch": 0.6134090034628703, + "grad_norm": 1.8268167461392897, + "learning_rate": 6.866603197421583e-06, + "loss": 0.8397, + "step": 6377 + }, + { + "epoch": 0.6135051943055021, + "grad_norm": 2.2791215183452835, + "learning_rate": 6.863644815669197e-06, + "loss": 1.0166, + "step": 6378 + }, + { + "epoch": 0.6136013851481339, + "grad_norm": 1.7572584354569796, + "learning_rate": 6.860686738327078e-06, + "loss": 0.9656, + "step": 6379 + }, + { + "epoch": 0.6136975759907657, + "grad_norm": 1.9792829732786574, + "learning_rate": 6.857728965682344e-06, + "loss": 0.9666, + "step": 6380 + }, + { + "epoch": 0.6137937668333975, + "grad_norm": 1.9896395587260367, + "learning_rate": 6.854771498022068e-06, + "loss": 1.0613, + "step": 6381 + }, + { + "epoch": 0.6138899576760293, + "grad_norm": 2.1159095479011687, + "learning_rate": 6.851814335633298e-06, + "loss": 0.9755, + "step": 6382 + }, + { + "epoch": 0.6139861485186611, + "grad_norm": 1.7519600687734114, + "learning_rate": 6.848857478803055e-06, + "loss": 0.8772, + "step": 6383 + }, + { + "epoch": 0.6140823393612928, + "grad_norm": 2.095243568596136, + "learning_rate": 6.8459009278183275e-06, + "loss": 0.9536, + "step": 6384 + }, + { + "epoch": 0.6141785302039245, + "grad_norm": 1.860138925026349, + "learning_rate": 6.842944682966071e-06, + "loss": 0.9048, + "step": 6385 + }, + { + "epoch": 0.6142747210465563, + "grad_norm": 1.92410535164394, + "learning_rate": 6.839988744533211e-06, + "loss": 0.8361, + "step": 6386 + }, + { + "epoch": 0.6143709118891881, + "grad_norm": 1.8945740639319586, + "learning_rate": 6.837033112806655e-06, + "loss": 0.9507, + "step": 6387 + }, + { + "epoch": 0.6144671027318199, + "grad_norm": 2.042897961008772, + "learning_rate": 6.834077788073268e-06, + "loss": 0.9798, + "step": 6388 + }, + { + "epoch": 0.6145632935744517, + "grad_norm": 2.1664756156210365, + "learning_rate": 6.831122770619889e-06, + "loss": 0.9711, + "step": 6389 + }, + { + "epoch": 0.6146594844170835, + "grad_norm": 1.7656102653226955, + "learning_rate": 6.8281680607333364e-06, + "loss": 0.8808, + "step": 6390 + }, + { + "epoch": 0.6147556752597153, + "grad_norm": 1.9543530463603815, + "learning_rate": 6.825213658700376e-06, + "loss": 0.9959, + "step": 6391 + }, + { + "epoch": 0.6148518661023471, + "grad_norm": 1.9503614003873733, + "learning_rate": 6.822259564807768e-06, + "loss": 0.9709, + "step": 6392 + }, + { + "epoch": 0.6149480569449788, + "grad_norm": 2.352766126911951, + "learning_rate": 6.8193057793422275e-06, + "loss": 0.9548, + "step": 6393 + }, + { + "epoch": 0.6150442477876106, + "grad_norm": 1.626312703826161, + "learning_rate": 6.81635230259045e-06, + "loss": 0.7553, + "step": 6394 + }, + { + "epoch": 0.6151404386302424, + "grad_norm": 1.9102774934777729, + "learning_rate": 6.813399134839092e-06, + "loss": 0.8938, + "step": 6395 + }, + { + "epoch": 0.6152366294728742, + "grad_norm": 1.9219924965061375, + "learning_rate": 6.810446276374789e-06, + "loss": 0.8972, + "step": 6396 + }, + { + "epoch": 0.615332820315506, + "grad_norm": 2.045223618369766, + "learning_rate": 6.807493727484134e-06, + "loss": 0.843, + "step": 6397 + }, + { + "epoch": 0.6154290111581378, + "grad_norm": 2.084200853687975, + "learning_rate": 6.8045414884536975e-06, + "loss": 0.9584, + "step": 6398 + }, + { + "epoch": 0.6155252020007695, + "grad_norm": 1.622196761281758, + "learning_rate": 6.801589559570025e-06, + "loss": 0.9785, + "step": 6399 + }, + { + "epoch": 0.6156213928434013, + "grad_norm": 2.189270186841389, + "learning_rate": 6.7986379411196255e-06, + "loss": 0.9877, + "step": 6400 + }, + { + "epoch": 0.6157175836860331, + "grad_norm": 1.6216135332596981, + "learning_rate": 6.795686633388976e-06, + "loss": 0.7772, + "step": 6401 + }, + { + "epoch": 0.6158137745286649, + "grad_norm": 1.9021678323432527, + "learning_rate": 6.7927356366645315e-06, + "loss": 0.8972, + "step": 6402 + }, + { + "epoch": 0.6159099653712966, + "grad_norm": 2.0391601514845505, + "learning_rate": 6.789784951232706e-06, + "loss": 0.9525, + "step": 6403 + }, + { + "epoch": 0.6160061562139284, + "grad_norm": 2.1005775934406423, + "learning_rate": 6.786834577379893e-06, + "loss": 0.926, + "step": 6404 + }, + { + "epoch": 0.6161023470565602, + "grad_norm": 2.099469137992617, + "learning_rate": 6.783884515392446e-06, + "loss": 0.9025, + "step": 6405 + }, + { + "epoch": 0.616198537899192, + "grad_norm": 1.658474103619178, + "learning_rate": 6.780934765556702e-06, + "loss": 0.8004, + "step": 6406 + }, + { + "epoch": 0.6162947287418238, + "grad_norm": 1.9572465837200788, + "learning_rate": 6.7779853281589545e-06, + "loss": 0.9112, + "step": 6407 + }, + { + "epoch": 0.6163909195844556, + "grad_norm": 2.0767650639013886, + "learning_rate": 6.775036203485472e-06, + "loss": 0.9689, + "step": 6408 + }, + { + "epoch": 0.6164871104270874, + "grad_norm": 2.4836330989956386, + "learning_rate": 6.772087391822495e-06, + "loss": 0.9005, + "step": 6409 + }, + { + "epoch": 0.6165833012697192, + "grad_norm": 2.0656287827089552, + "learning_rate": 6.769138893456225e-06, + "loss": 0.9816, + "step": 6410 + }, + { + "epoch": 0.616679492112351, + "grad_norm": 1.579735423598241, + "learning_rate": 6.766190708672846e-06, + "loss": 0.8149, + "step": 6411 + }, + { + "epoch": 0.6167756829549826, + "grad_norm": 2.024158376873138, + "learning_rate": 6.763242837758504e-06, + "loss": 0.9472, + "step": 6412 + }, + { + "epoch": 0.6168718737976144, + "grad_norm": 1.9199119920829228, + "learning_rate": 6.7602952809993135e-06, + "loss": 0.9315, + "step": 6413 + }, + { + "epoch": 0.6169680646402462, + "grad_norm": 1.839648597481183, + "learning_rate": 6.757348038681357e-06, + "loss": 0.8836, + "step": 6414 + }, + { + "epoch": 0.617064255482878, + "grad_norm": 1.7462884072216494, + "learning_rate": 6.754401111090694e-06, + "loss": 0.9181, + "step": 6415 + }, + { + "epoch": 0.6171604463255098, + "grad_norm": 2.12738752770965, + "learning_rate": 6.751454498513349e-06, + "loss": 0.8914, + "step": 6416 + }, + { + "epoch": 0.6172566371681416, + "grad_norm": 1.696234468709365, + "learning_rate": 6.748508201235313e-06, + "loss": 0.9794, + "step": 6417 + }, + { + "epoch": 0.6173528280107734, + "grad_norm": 1.9059601565125448, + "learning_rate": 6.745562219542554e-06, + "loss": 0.9132, + "step": 6418 + }, + { + "epoch": 0.6174490188534052, + "grad_norm": 2.188024061348216, + "learning_rate": 6.7426165537210044e-06, + "loss": 0.9711, + "step": 6419 + }, + { + "epoch": 0.617545209696037, + "grad_norm": 1.9050040325634028, + "learning_rate": 6.7396712040565625e-06, + "loss": 0.9365, + "step": 6420 + }, + { + "epoch": 0.6176414005386687, + "grad_norm": 2.2919834007881774, + "learning_rate": 6.736726170835102e-06, + "loss": 0.8872, + "step": 6421 + }, + { + "epoch": 0.6177375913813005, + "grad_norm": 1.9647517096953226, + "learning_rate": 6.733781454342463e-06, + "loss": 0.892, + "step": 6422 + }, + { + "epoch": 0.6178337822239323, + "grad_norm": 2.338767691135069, + "learning_rate": 6.730837054864456e-06, + "loss": 0.9858, + "step": 6423 + }, + { + "epoch": 0.617929973066564, + "grad_norm": 2.0542587763761087, + "learning_rate": 6.727892972686861e-06, + "loss": 1.0048, + "step": 6424 + }, + { + "epoch": 0.6180261639091958, + "grad_norm": 1.9568076985243463, + "learning_rate": 6.7249492080954305e-06, + "loss": 0.9628, + "step": 6425 + }, + { + "epoch": 0.6181223547518276, + "grad_norm": 1.8448829435258645, + "learning_rate": 6.722005761375873e-06, + "loss": 0.8563, + "step": 6426 + }, + { + "epoch": 0.6182185455944594, + "grad_norm": 2.15814821917866, + "learning_rate": 6.719062632813882e-06, + "loss": 0.9089, + "step": 6427 + }, + { + "epoch": 0.6183147364370912, + "grad_norm": 2.0369223516066883, + "learning_rate": 6.716119822695111e-06, + "loss": 0.9021, + "step": 6428 + }, + { + "epoch": 0.618410927279723, + "grad_norm": 1.8838843375192316, + "learning_rate": 6.7131773313051855e-06, + "loss": 0.9309, + "step": 6429 + }, + { + "epoch": 0.6185071181223547, + "grad_norm": 1.9342430936879402, + "learning_rate": 6.710235158929703e-06, + "loss": 0.9328, + "step": 6430 + }, + { + "epoch": 0.6186033089649865, + "grad_norm": 1.9060417570568666, + "learning_rate": 6.70729330585422e-06, + "loss": 0.9389, + "step": 6431 + }, + { + "epoch": 0.6186994998076183, + "grad_norm": 2.0213566596907433, + "learning_rate": 6.704351772364274e-06, + "loss": 0.9011, + "step": 6432 + }, + { + "epoch": 0.6187956906502501, + "grad_norm": 1.8737363602106263, + "learning_rate": 6.701410558745361e-06, + "loss": 1.03, + "step": 6433 + }, + { + "epoch": 0.6188918814928819, + "grad_norm": 1.9570753791452877, + "learning_rate": 6.698469665282958e-06, + "loss": 0.9378, + "step": 6434 + }, + { + "epoch": 0.6189880723355137, + "grad_norm": 2.434703968878361, + "learning_rate": 6.6955290922624995e-06, + "loss": 0.9509, + "step": 6435 + }, + { + "epoch": 0.6190842631781455, + "grad_norm": 1.6953456033300773, + "learning_rate": 6.692588839969397e-06, + "loss": 0.9945, + "step": 6436 + }, + { + "epoch": 0.6191804540207773, + "grad_norm": 1.927973871797001, + "learning_rate": 6.689648908689023e-06, + "loss": 1.0256, + "step": 6437 + }, + { + "epoch": 0.619276644863409, + "grad_norm": 1.9808618196436947, + "learning_rate": 6.6867092987067214e-06, + "loss": 0.9621, + "step": 6438 + }, + { + "epoch": 0.6193728357060407, + "grad_norm": 1.906647990324173, + "learning_rate": 6.683770010307813e-06, + "loss": 0.9915, + "step": 6439 + }, + { + "epoch": 0.6194690265486725, + "grad_norm": 2.0002941193867128, + "learning_rate": 6.680831043777579e-06, + "loss": 0.986, + "step": 6440 + }, + { + "epoch": 0.6195652173913043, + "grad_norm": 2.0764641958207766, + "learning_rate": 6.6778923994012676e-06, + "loss": 0.9472, + "step": 6441 + }, + { + "epoch": 0.6196614082339361, + "grad_norm": 1.9184144751995957, + "learning_rate": 6.674954077464108e-06, + "loss": 0.8555, + "step": 6442 + }, + { + "epoch": 0.6197575990765679, + "grad_norm": 1.9332882084641243, + "learning_rate": 6.67201607825128e-06, + "loss": 1.0315, + "step": 6443 + }, + { + "epoch": 0.6198537899191997, + "grad_norm": 1.965602208326682, + "learning_rate": 6.6690784020479484e-06, + "loss": 0.8917, + "step": 6444 + }, + { + "epoch": 0.6199499807618315, + "grad_norm": 2.056003480387028, + "learning_rate": 6.666141049139233e-06, + "loss": 0.9011, + "step": 6445 + }, + { + "epoch": 0.6200461716044633, + "grad_norm": 1.8502152439018178, + "learning_rate": 6.6632040198102364e-06, + "loss": 0.9849, + "step": 6446 + }, + { + "epoch": 0.6201423624470951, + "grad_norm": 2.0669824839364157, + "learning_rate": 6.660267314346021e-06, + "loss": 0.9466, + "step": 6447 + }, + { + "epoch": 0.6202385532897268, + "grad_norm": 1.9315122760293322, + "learning_rate": 6.657330933031619e-06, + "loss": 0.981, + "step": 6448 + }, + { + "epoch": 0.6203347441323586, + "grad_norm": 2.073342359091989, + "learning_rate": 6.6543948761520295e-06, + "loss": 0.8182, + "step": 6449 + }, + { + "epoch": 0.6204309349749904, + "grad_norm": 1.9314358674033165, + "learning_rate": 6.651459143992221e-06, + "loss": 1.0296, + "step": 6450 + }, + { + "epoch": 0.6205271258176221, + "grad_norm": 2.0716813457948153, + "learning_rate": 6.648523736837136e-06, + "loss": 0.976, + "step": 6451 + }, + { + "epoch": 0.6206233166602539, + "grad_norm": 2.000305740817404, + "learning_rate": 6.645588654971677e-06, + "loss": 1.0828, + "step": 6452 + }, + { + "epoch": 0.6207195075028857, + "grad_norm": 1.7457260186991026, + "learning_rate": 6.642653898680725e-06, + "loss": 0.8137, + "step": 6453 + }, + { + "epoch": 0.6208156983455175, + "grad_norm": 1.877099602276022, + "learning_rate": 6.639719468249115e-06, + "loss": 0.9784, + "step": 6454 + }, + { + "epoch": 0.6209118891881493, + "grad_norm": 1.877201642851219, + "learning_rate": 6.6367853639616645e-06, + "loss": 0.9202, + "step": 6455 + }, + { + "epoch": 0.6210080800307811, + "grad_norm": 1.923944330179909, + "learning_rate": 6.633851586103153e-06, + "loss": 0.9579, + "step": 6456 + }, + { + "epoch": 0.6211042708734128, + "grad_norm": 2.0757463261226357, + "learning_rate": 6.630918134958325e-06, + "loss": 0.9792, + "step": 6457 + }, + { + "epoch": 0.6212004617160446, + "grad_norm": 2.1028844671008695, + "learning_rate": 6.627985010811903e-06, + "loss": 1.0031, + "step": 6458 + }, + { + "epoch": 0.6212966525586764, + "grad_norm": 2.3166520984504517, + "learning_rate": 6.625052213948571e-06, + "loss": 0.9276, + "step": 6459 + }, + { + "epoch": 0.6213928434013082, + "grad_norm": 2.2033441398218536, + "learning_rate": 6.622119744652977e-06, + "loss": 1.015, + "step": 6460 + }, + { + "epoch": 0.62148903424394, + "grad_norm": 2.035309951753383, + "learning_rate": 6.619187603209747e-06, + "loss": 0.9577, + "step": 6461 + }, + { + "epoch": 0.6215852250865718, + "grad_norm": 2.105523633007966, + "learning_rate": 6.616255789903467e-06, + "loss": 0.9407, + "step": 6462 + }, + { + "epoch": 0.6216814159292036, + "grad_norm": 1.758347240663497, + "learning_rate": 6.613324305018698e-06, + "loss": 0.8877, + "step": 6463 + }, + { + "epoch": 0.6217776067718354, + "grad_norm": 2.618641269160405, + "learning_rate": 6.610393148839964e-06, + "loss": 0.8014, + "step": 6464 + }, + { + "epoch": 0.6218737976144671, + "grad_norm": 2.134215977241512, + "learning_rate": 6.6074623216517654e-06, + "loss": 0.817, + "step": 6465 + }, + { + "epoch": 0.6219699884570988, + "grad_norm": 1.9201163011422582, + "learning_rate": 6.6045318237385526e-06, + "loss": 1.0459, + "step": 6466 + }, + { + "epoch": 0.6220661792997306, + "grad_norm": 1.9861803851033277, + "learning_rate": 6.601601655384764e-06, + "loss": 1.0195, + "step": 6467 + }, + { + "epoch": 0.6221623701423624, + "grad_norm": 1.8816512010736226, + "learning_rate": 6.598671816874794e-06, + "loss": 1.0279, + "step": 6468 + }, + { + "epoch": 0.6222585609849942, + "grad_norm": 2.0919317968070015, + "learning_rate": 6.5957423084930096e-06, + "loss": 0.8929, + "step": 6469 + }, + { + "epoch": 0.622354751827626, + "grad_norm": 2.4245073983956362, + "learning_rate": 6.5928131305237465e-06, + "loss": 0.9747, + "step": 6470 + }, + { + "epoch": 0.6224509426702578, + "grad_norm": 1.9526055560143623, + "learning_rate": 6.589884283251306e-06, + "loss": 0.9269, + "step": 6471 + }, + { + "epoch": 0.6225471335128896, + "grad_norm": 2.1346078424875286, + "learning_rate": 6.586955766959958e-06, + "loss": 0.9454, + "step": 6472 + }, + { + "epoch": 0.6226433243555214, + "grad_norm": 1.810590846770154, + "learning_rate": 6.584027581933936e-06, + "loss": 0.8919, + "step": 6473 + }, + { + "epoch": 0.6227395151981532, + "grad_norm": 2.322661426073338, + "learning_rate": 6.581099728457451e-06, + "loss": 1.0054, + "step": 6474 + }, + { + "epoch": 0.6228357060407849, + "grad_norm": 1.9471384756069459, + "learning_rate": 6.578172206814675e-06, + "loss": 0.9099, + "step": 6475 + }, + { + "epoch": 0.6229318968834167, + "grad_norm": 2.003223106321727, + "learning_rate": 6.5752450172897466e-06, + "loss": 0.9619, + "step": 6476 + }, + { + "epoch": 0.6230280877260485, + "grad_norm": 1.9045453015368703, + "learning_rate": 6.572318160166782e-06, + "loss": 0.9176, + "step": 6477 + }, + { + "epoch": 0.6231242785686802, + "grad_norm": 2.005525395276741, + "learning_rate": 6.569391635729847e-06, + "loss": 0.9143, + "step": 6478 + }, + { + "epoch": 0.623220469411312, + "grad_norm": 1.952572678850409, + "learning_rate": 6.566465444262994e-06, + "loss": 0.8858, + "step": 6479 + }, + { + "epoch": 0.6233166602539438, + "grad_norm": 1.7436522396384515, + "learning_rate": 6.563539586050233e-06, + "loss": 0.7902, + "step": 6480 + }, + { + "epoch": 0.6234128510965756, + "grad_norm": 2.017336799238422, + "learning_rate": 6.560614061375541e-06, + "loss": 0.9816, + "step": 6481 + }, + { + "epoch": 0.6235090419392074, + "grad_norm": 1.8725158869286613, + "learning_rate": 6.557688870522871e-06, + "loss": 0.9193, + "step": 6482 + }, + { + "epoch": 0.6236052327818392, + "grad_norm": 1.6915085687055085, + "learning_rate": 6.554764013776134e-06, + "loss": 0.8093, + "step": 6483 + }, + { + "epoch": 0.6237014236244709, + "grad_norm": 2.051185990831874, + "learning_rate": 6.551839491419213e-06, + "loss": 1.0004, + "step": 6484 + }, + { + "epoch": 0.6237976144671027, + "grad_norm": 1.957255515925775, + "learning_rate": 6.548915303735957e-06, + "loss": 1.0203, + "step": 6485 + }, + { + "epoch": 0.6238938053097345, + "grad_norm": 1.9811108614231965, + "learning_rate": 6.545991451010185e-06, + "loss": 0.894, + "step": 6486 + }, + { + "epoch": 0.6239899961523663, + "grad_norm": 2.181402974036737, + "learning_rate": 6.5430679335256844e-06, + "loss": 1.0228, + "step": 6487 + }, + { + "epoch": 0.6240861869949981, + "grad_norm": 2.109278080124145, + "learning_rate": 6.5401447515662065e-06, + "loss": 0.8593, + "step": 6488 + }, + { + "epoch": 0.6241823778376299, + "grad_norm": 2.0476189949615202, + "learning_rate": 6.537221905415467e-06, + "loss": 0.9355, + "step": 6489 + }, + { + "epoch": 0.6242785686802617, + "grad_norm": 1.9688750313548504, + "learning_rate": 6.5342993953571556e-06, + "loss": 1.0066, + "step": 6490 + }, + { + "epoch": 0.6243747595228935, + "grad_norm": 2.0769108242282757, + "learning_rate": 6.53137722167493e-06, + "loss": 0.9701, + "step": 6491 + }, + { + "epoch": 0.6244709503655252, + "grad_norm": 1.7689727867486573, + "learning_rate": 6.52845538465241e-06, + "loss": 0.9541, + "step": 6492 + }, + { + "epoch": 0.6245671412081569, + "grad_norm": 2.0152214271189357, + "learning_rate": 6.525533884573183e-06, + "loss": 0.8871, + "step": 6493 + }, + { + "epoch": 0.6246633320507887, + "grad_norm": 1.8716239840004767, + "learning_rate": 6.522612721720813e-06, + "loss": 0.9114, + "step": 6494 + }, + { + "epoch": 0.6247595228934205, + "grad_norm": 1.7743986492971673, + "learning_rate": 6.519691896378817e-06, + "loss": 0.9435, + "step": 6495 + }, + { + "epoch": 0.6248557137360523, + "grad_norm": 1.8918067131553464, + "learning_rate": 6.5167714088306865e-06, + "loss": 0.8647, + "step": 6496 + }, + { + "epoch": 0.6249519045786841, + "grad_norm": 1.9969161524304724, + "learning_rate": 6.513851259359882e-06, + "loss": 0.986, + "step": 6497 + }, + { + "epoch": 0.6250480954213159, + "grad_norm": 4.2662222523122555, + "learning_rate": 6.51093144824983e-06, + "loss": 0.8272, + "step": 6498 + }, + { + "epoch": 0.6251442862639477, + "grad_norm": 1.794258676067159, + "learning_rate": 6.5080119757839215e-06, + "loss": 0.9778, + "step": 6499 + }, + { + "epoch": 0.6252404771065795, + "grad_norm": 1.6742807098776067, + "learning_rate": 6.505092842245519e-06, + "loss": 0.8376, + "step": 6500 + }, + { + "epoch": 0.6253366679492113, + "grad_norm": 1.9425717039120607, + "learning_rate": 6.502174047917947e-06, + "loss": 0.8949, + "step": 6501 + }, + { + "epoch": 0.6254328587918431, + "grad_norm": 1.96772590041475, + "learning_rate": 6.499255593084498e-06, + "loss": 0.9703, + "step": 6502 + }, + { + "epoch": 0.6255290496344748, + "grad_norm": 1.9637174867702631, + "learning_rate": 6.496337478028437e-06, + "loss": 0.969, + "step": 6503 + }, + { + "epoch": 0.6256252404771065, + "grad_norm": 1.895759383012726, + "learning_rate": 6.493419703032991e-06, + "loss": 0.9596, + "step": 6504 + }, + { + "epoch": 0.6257214313197383, + "grad_norm": 1.73008462645521, + "learning_rate": 6.490502268381356e-06, + "loss": 0.9715, + "step": 6505 + }, + { + "epoch": 0.6258176221623701, + "grad_norm": 1.9947756666937708, + "learning_rate": 6.487585174356691e-06, + "loss": 0.882, + "step": 6506 + }, + { + "epoch": 0.6259138130050019, + "grad_norm": 1.9279109011869613, + "learning_rate": 6.484668421242127e-06, + "loss": 0.9128, + "step": 6507 + }, + { + "epoch": 0.6260100038476337, + "grad_norm": 1.9412759444456187, + "learning_rate": 6.481752009320761e-06, + "loss": 0.9222, + "step": 6508 + }, + { + "epoch": 0.6261061946902655, + "grad_norm": 2.239787720308772, + "learning_rate": 6.478835938875654e-06, + "loss": 1.0498, + "step": 6509 + }, + { + "epoch": 0.6262023855328973, + "grad_norm": 2.0483870973252847, + "learning_rate": 6.4759202101898366e-06, + "loss": 0.9238, + "step": 6510 + }, + { + "epoch": 0.6262985763755291, + "grad_norm": 2.2753242019707542, + "learning_rate": 6.473004823546309e-06, + "loss": 0.9979, + "step": 6511 + }, + { + "epoch": 0.6263947672181608, + "grad_norm": 2.213318940957391, + "learning_rate": 6.4700897792280285e-06, + "loss": 0.9721, + "step": 6512 + }, + { + "epoch": 0.6264909580607926, + "grad_norm": 2.0601877165970195, + "learning_rate": 6.4671750775179266e-06, + "loss": 1.0394, + "step": 6513 + }, + { + "epoch": 0.6265871489034244, + "grad_norm": 2.020827272239908, + "learning_rate": 6.464260718698902e-06, + "loss": 0.9325, + "step": 6514 + }, + { + "epoch": 0.6266833397460562, + "grad_norm": 2.0255785695950843, + "learning_rate": 6.461346703053819e-06, + "loss": 1.0561, + "step": 6515 + }, + { + "epoch": 0.626779530588688, + "grad_norm": 2.0619001778597057, + "learning_rate": 6.458433030865503e-06, + "loss": 0.8793, + "step": 6516 + }, + { + "epoch": 0.6268757214313198, + "grad_norm": 1.8588259776022904, + "learning_rate": 6.455519702416763e-06, + "loss": 0.9222, + "step": 6517 + }, + { + "epoch": 0.6269719122739515, + "grad_norm": 2.047409678940232, + "learning_rate": 6.452606717990346e-06, + "loss": 1.0262, + "step": 6518 + }, + { + "epoch": 0.6270681031165833, + "grad_norm": 2.341960746466321, + "learning_rate": 6.449694077868995e-06, + "loss": 0.9912, + "step": 6519 + }, + { + "epoch": 0.6271642939592151, + "grad_norm": 1.8377694030989349, + "learning_rate": 6.4467817823354005e-06, + "loss": 0.8695, + "step": 6520 + }, + { + "epoch": 0.6272604848018468, + "grad_norm": 1.9195495855423181, + "learning_rate": 6.443869831672225e-06, + "loss": 1.0001, + "step": 6521 + }, + { + "epoch": 0.6273566756444786, + "grad_norm": 2.0138694603761484, + "learning_rate": 6.440958226162104e-06, + "loss": 0.9467, + "step": 6522 + }, + { + "epoch": 0.6274528664871104, + "grad_norm": 1.8799669638176322, + "learning_rate": 6.438046966087632e-06, + "loss": 0.9316, + "step": 6523 + }, + { + "epoch": 0.6275490573297422, + "grad_norm": 2.002609247283148, + "learning_rate": 6.43513605173137e-06, + "loss": 0.9541, + "step": 6524 + }, + { + "epoch": 0.627645248172374, + "grad_norm": 2.0719041734471904, + "learning_rate": 6.432225483375846e-06, + "loss": 1.0365, + "step": 6525 + }, + { + "epoch": 0.6277414390150058, + "grad_norm": 2.104064569134316, + "learning_rate": 6.4293152613035594e-06, + "loss": 1.0674, + "step": 6526 + }, + { + "epoch": 0.6278376298576376, + "grad_norm": 1.7657052924572247, + "learning_rate": 6.42640538579697e-06, + "loss": 0.7787, + "step": 6527 + }, + { + "epoch": 0.6279338207002694, + "grad_norm": 2.3128739351979672, + "learning_rate": 6.4234958571385095e-06, + "loss": 0.9888, + "step": 6528 + }, + { + "epoch": 0.6280300115429012, + "grad_norm": 2.0009218154967865, + "learning_rate": 6.4205866756105685e-06, + "loss": 1.0058, + "step": 6529 + }, + { + "epoch": 0.6281262023855329, + "grad_norm": 2.0201145602318524, + "learning_rate": 6.4176778414955075e-06, + "loss": 0.9067, + "step": 6530 + }, + { + "epoch": 0.6282223932281646, + "grad_norm": 2.131670589527866, + "learning_rate": 6.4147693550756586e-06, + "loss": 0.8399, + "step": 6531 + }, + { + "epoch": 0.6283185840707964, + "grad_norm": 1.94936366161308, + "learning_rate": 6.4118612166333124e-06, + "loss": 0.8378, + "step": 6532 + }, + { + "epoch": 0.6284147749134282, + "grad_norm": 1.8893490711567806, + "learning_rate": 6.408953426450727e-06, + "loss": 0.8496, + "step": 6533 + }, + { + "epoch": 0.62851096575606, + "grad_norm": 2.1778296139701965, + "learning_rate": 6.4060459848101354e-06, + "loss": 0.9873, + "step": 6534 + }, + { + "epoch": 0.6286071565986918, + "grad_norm": 1.9061340346932234, + "learning_rate": 6.403138891993723e-06, + "loss": 0.9213, + "step": 6535 + }, + { + "epoch": 0.6287033474413236, + "grad_norm": 1.8225441976004633, + "learning_rate": 6.400232148283651e-06, + "loss": 0.8718, + "step": 6536 + }, + { + "epoch": 0.6287995382839554, + "grad_norm": 1.899557992701573, + "learning_rate": 6.39732575396204e-06, + "loss": 0.9917, + "step": 6537 + }, + { + "epoch": 0.6288957291265872, + "grad_norm": 2.035254707180953, + "learning_rate": 6.3944197093109885e-06, + "loss": 0.9505, + "step": 6538 + }, + { + "epoch": 0.6289919199692189, + "grad_norm": 2.231428821351853, + "learning_rate": 6.391514014612548e-06, + "loss": 1.0767, + "step": 6539 + }, + { + "epoch": 0.6290881108118507, + "grad_norm": 2.1256771253961833, + "learning_rate": 6.388608670148741e-06, + "loss": 0.9331, + "step": 6540 + }, + { + "epoch": 0.6291843016544825, + "grad_norm": 2.079026540628945, + "learning_rate": 6.385703676201558e-06, + "loss": 1.0235, + "step": 6541 + }, + { + "epoch": 0.6292804924971143, + "grad_norm": 2.018586235474117, + "learning_rate": 6.38279903305295e-06, + "loss": 0.9066, + "step": 6542 + }, + { + "epoch": 0.6293766833397461, + "grad_norm": 2.053803394934105, + "learning_rate": 6.379894740984842e-06, + "loss": 0.9068, + "step": 6543 + }, + { + "epoch": 0.6294728741823779, + "grad_norm": 1.956116309852617, + "learning_rate": 6.376990800279119e-06, + "loss": 1.0515, + "step": 6544 + }, + { + "epoch": 0.6295690650250096, + "grad_norm": 1.9379212926509484, + "learning_rate": 6.374087211217632e-06, + "loss": 0.9359, + "step": 6545 + }, + { + "epoch": 0.6296652558676414, + "grad_norm": 2.024135455654986, + "learning_rate": 6.3711839740822035e-06, + "loss": 0.948, + "step": 6546 + }, + { + "epoch": 0.6297614467102732, + "grad_norm": 2.1404530692276946, + "learning_rate": 6.368281089154613e-06, + "loss": 1.0833, + "step": 6547 + }, + { + "epoch": 0.6298576375529049, + "grad_norm": 2.077530037444529, + "learning_rate": 6.3653785567166125e-06, + "loss": 0.9552, + "step": 6548 + }, + { + "epoch": 0.6299538283955367, + "grad_norm": 1.9168865761858294, + "learning_rate": 6.3624763770499155e-06, + "loss": 0.9789, + "step": 6549 + }, + { + "epoch": 0.6300500192381685, + "grad_norm": 1.955626102632126, + "learning_rate": 6.359574550436209e-06, + "loss": 0.9863, + "step": 6550 + }, + { + "epoch": 0.6301462100808003, + "grad_norm": 2.05703562943821, + "learning_rate": 6.356673077157135e-06, + "loss": 0.957, + "step": 6551 + }, + { + "epoch": 0.6302424009234321, + "grad_norm": 2.357854832232745, + "learning_rate": 6.3537719574943105e-06, + "loss": 0.9494, + "step": 6552 + }, + { + "epoch": 0.6303385917660639, + "grad_norm": 1.972649578886453, + "learning_rate": 6.350871191729311e-06, + "loss": 0.9787, + "step": 6553 + }, + { + "epoch": 0.6304347826086957, + "grad_norm": 1.8204484089576531, + "learning_rate": 6.347970780143678e-06, + "loss": 0.876, + "step": 6554 + }, + { + "epoch": 0.6305309734513275, + "grad_norm": 1.8437086658029491, + "learning_rate": 6.345070723018929e-06, + "loss": 0.9623, + "step": 6555 + }, + { + "epoch": 0.6306271642939593, + "grad_norm": 1.8651048847722453, + "learning_rate": 6.342171020636533e-06, + "loss": 0.9569, + "step": 6556 + }, + { + "epoch": 0.630723355136591, + "grad_norm": 2.135519648322648, + "learning_rate": 6.33927167327794e-06, + "loss": 0.8788, + "step": 6557 + }, + { + "epoch": 0.6308195459792227, + "grad_norm": 1.9081101234457596, + "learning_rate": 6.336372681224543e-06, + "loss": 0.8806, + "step": 6558 + }, + { + "epoch": 0.6309157368218545, + "grad_norm": 2.3685838910893198, + "learning_rate": 6.333474044757725e-06, + "loss": 1.0304, + "step": 6559 + }, + { + "epoch": 0.6310119276644863, + "grad_norm": 2.364150705654083, + "learning_rate": 6.330575764158819e-06, + "loss": 0.8861, + "step": 6560 + }, + { + "epoch": 0.6311081185071181, + "grad_norm": 1.867372073997998, + "learning_rate": 6.327677839709128e-06, + "loss": 0.9376, + "step": 6561 + }, + { + "epoch": 0.6312043093497499, + "grad_norm": 1.7729578575220548, + "learning_rate": 6.324780271689923e-06, + "loss": 0.7909, + "step": 6562 + }, + { + "epoch": 0.6313005001923817, + "grad_norm": 2.1093187761501646, + "learning_rate": 6.321883060382441e-06, + "loss": 0.9395, + "step": 6563 + }, + { + "epoch": 0.6313966910350135, + "grad_norm": 1.718962312984291, + "learning_rate": 6.318986206067872e-06, + "loss": 0.9178, + "step": 6564 + }, + { + "epoch": 0.6314928818776453, + "grad_norm": 1.989304638808329, + "learning_rate": 6.316089709027386e-06, + "loss": 0.9055, + "step": 6565 + }, + { + "epoch": 0.631589072720277, + "grad_norm": 2.039036694446046, + "learning_rate": 6.313193569542113e-06, + "loss": 0.982, + "step": 6566 + }, + { + "epoch": 0.6316852635629088, + "grad_norm": 1.944092522277781, + "learning_rate": 6.310297787893148e-06, + "loss": 0.8884, + "step": 6567 + }, + { + "epoch": 0.6317814544055406, + "grad_norm": 1.818300001919966, + "learning_rate": 6.30740236436155e-06, + "loss": 0.9748, + "step": 6568 + }, + { + "epoch": 0.6318776452481724, + "grad_norm": 2.0269376545639575, + "learning_rate": 6.3045072992283515e-06, + "loss": 0.9587, + "step": 6569 + }, + { + "epoch": 0.6319738360908042, + "grad_norm": 1.7495144886645404, + "learning_rate": 6.301612592774533e-06, + "loss": 0.8846, + "step": 6570 + }, + { + "epoch": 0.632070026933436, + "grad_norm": 2.2444712288874737, + "learning_rate": 6.298718245281057e-06, + "loss": 1.021, + "step": 6571 + }, + { + "epoch": 0.6321662177760677, + "grad_norm": 1.751148918615231, + "learning_rate": 6.295824257028844e-06, + "loss": 0.8173, + "step": 6572 + }, + { + "epoch": 0.6322624086186995, + "grad_norm": 1.9158805708776547, + "learning_rate": 6.292930628298779e-06, + "loss": 0.8812, + "step": 6573 + }, + { + "epoch": 0.6323585994613313, + "grad_norm": 2.060299896457625, + "learning_rate": 6.290037359371717e-06, + "loss": 0.8737, + "step": 6574 + }, + { + "epoch": 0.632454790303963, + "grad_norm": 1.9347197621971677, + "learning_rate": 6.287144450528473e-06, + "loss": 1.0174, + "step": 6575 + }, + { + "epoch": 0.6325509811465948, + "grad_norm": 2.175346466246964, + "learning_rate": 6.284251902049827e-06, + "loss": 0.9554, + "step": 6576 + }, + { + "epoch": 0.6326471719892266, + "grad_norm": 1.8560042865741029, + "learning_rate": 6.281359714216525e-06, + "loss": 1.0268, + "step": 6577 + }, + { + "epoch": 0.6327433628318584, + "grad_norm": 1.8811271806785508, + "learning_rate": 6.278467887309283e-06, + "loss": 0.899, + "step": 6578 + }, + { + "epoch": 0.6328395536744902, + "grad_norm": 1.8163385828826029, + "learning_rate": 6.275576421608773e-06, + "loss": 0.8589, + "step": 6579 + }, + { + "epoch": 0.632935744517122, + "grad_norm": 1.9369014749287077, + "learning_rate": 6.272685317395644e-06, + "loss": 0.8526, + "step": 6580 + }, + { + "epoch": 0.6330319353597538, + "grad_norm": 1.8820795242355062, + "learning_rate": 6.269794574950494e-06, + "loss": 0.7825, + "step": 6581 + }, + { + "epoch": 0.6331281262023856, + "grad_norm": 2.192893566516969, + "learning_rate": 6.266904194553896e-06, + "loss": 0.9406, + "step": 6582 + }, + { + "epoch": 0.6332243170450174, + "grad_norm": 2.01956505771351, + "learning_rate": 6.26401417648639e-06, + "loss": 0.9154, + "step": 6583 + }, + { + "epoch": 0.633320507887649, + "grad_norm": 2.3313599107434197, + "learning_rate": 6.261124521028477e-06, + "loss": 0.8501, + "step": 6584 + }, + { + "epoch": 0.6334166987302808, + "grad_norm": 1.9946557139881254, + "learning_rate": 6.258235228460618e-06, + "loss": 0.947, + "step": 6585 + }, + { + "epoch": 0.6335128895729126, + "grad_norm": 1.8150642646661541, + "learning_rate": 6.255346299063252e-06, + "loss": 0.9852, + "step": 6586 + }, + { + "epoch": 0.6336090804155444, + "grad_norm": 2.1262229301011537, + "learning_rate": 6.252457733116766e-06, + "loss": 0.8961, + "step": 6587 + }, + { + "epoch": 0.6337052712581762, + "grad_norm": 1.7641226353991981, + "learning_rate": 6.249569530901525e-06, + "loss": 0.9191, + "step": 6588 + }, + { + "epoch": 0.633801462100808, + "grad_norm": 2.0043777697604943, + "learning_rate": 6.246681692697849e-06, + "loss": 1.0066, + "step": 6589 + }, + { + "epoch": 0.6338976529434398, + "grad_norm": 2.03381181735271, + "learning_rate": 6.243794218786034e-06, + "loss": 0.8926, + "step": 6590 + }, + { + "epoch": 0.6339938437860716, + "grad_norm": 2.1219603433321987, + "learning_rate": 6.240907109446332e-06, + "loss": 0.9071, + "step": 6591 + }, + { + "epoch": 0.6340900346287034, + "grad_norm": 1.9036264790774784, + "learning_rate": 6.238020364958964e-06, + "loss": 0.9619, + "step": 6592 + }, + { + "epoch": 0.6341862254713351, + "grad_norm": 2.220921547458059, + "learning_rate": 6.235133985604107e-06, + "loss": 0.9199, + "step": 6593 + }, + { + "epoch": 0.6342824163139669, + "grad_norm": 1.9515511192302657, + "learning_rate": 6.232247971661912e-06, + "loss": 0.8915, + "step": 6594 + }, + { + "epoch": 0.6343786071565987, + "grad_norm": 2.001047878220698, + "learning_rate": 6.229362323412493e-06, + "loss": 1.0327, + "step": 6595 + }, + { + "epoch": 0.6344747979992305, + "grad_norm": 1.921879679775871, + "learning_rate": 6.2264770411359256e-06, + "loss": 0.8735, + "step": 6596 + }, + { + "epoch": 0.6345709888418622, + "grad_norm": 2.1431320360660697, + "learning_rate": 6.223592125112252e-06, + "loss": 1.0295, + "step": 6597 + }, + { + "epoch": 0.634667179684494, + "grad_norm": 2.11536423083137, + "learning_rate": 6.22070757562148e-06, + "loss": 0.9378, + "step": 6598 + }, + { + "epoch": 0.6347633705271258, + "grad_norm": 1.9920953572413735, + "learning_rate": 6.217823392943577e-06, + "loss": 0.9698, + "step": 6599 + }, + { + "epoch": 0.6348595613697576, + "grad_norm": 2.198357312263764, + "learning_rate": 6.214939577358479e-06, + "loss": 0.967, + "step": 6600 + }, + { + "epoch": 0.6349557522123894, + "grad_norm": 2.0539120401815563, + "learning_rate": 6.212056129146082e-06, + "loss": 0.9064, + "step": 6601 + }, + { + "epoch": 0.6350519430550212, + "grad_norm": 1.95858505519301, + "learning_rate": 6.209173048586253e-06, + "loss": 0.9082, + "step": 6602 + }, + { + "epoch": 0.6351481338976529, + "grad_norm": 2.4106930022393174, + "learning_rate": 6.206290335958822e-06, + "loss": 0.9024, + "step": 6603 + }, + { + "epoch": 0.6352443247402847, + "grad_norm": 2.131453096033484, + "learning_rate": 6.203407991543577e-06, + "loss": 0.974, + "step": 6604 + }, + { + "epoch": 0.6353405155829165, + "grad_norm": 1.9486385635748078, + "learning_rate": 6.200526015620273e-06, + "loss": 0.9212, + "step": 6605 + }, + { + "epoch": 0.6354367064255483, + "grad_norm": 1.9177019517804008, + "learning_rate": 6.197644408468635e-06, + "loss": 0.9102, + "step": 6606 + }, + { + "epoch": 0.6355328972681801, + "grad_norm": 1.848853500418794, + "learning_rate": 6.194763170368345e-06, + "loss": 0.7816, + "step": 6607 + }, + { + "epoch": 0.6356290881108119, + "grad_norm": 1.9886599844037023, + "learning_rate": 6.191882301599052e-06, + "loss": 0.9582, + "step": 6608 + }, + { + "epoch": 0.6357252789534437, + "grad_norm": 2.0136512388449854, + "learning_rate": 6.189001802440377e-06, + "loss": 0.9869, + "step": 6609 + }, + { + "epoch": 0.6358214697960755, + "grad_norm": 2.3579477382090372, + "learning_rate": 6.186121673171882e-06, + "loss": 1.056, + "step": 6610 + }, + { + "epoch": 0.6359176606387072, + "grad_norm": 2.349086029717404, + "learning_rate": 6.183241914073119e-06, + "loss": 1.0259, + "step": 6611 + }, + { + "epoch": 0.6360138514813389, + "grad_norm": 1.975101829404891, + "learning_rate": 6.180362525423591e-06, + "loss": 0.988, + "step": 6612 + }, + { + "epoch": 0.6361100423239707, + "grad_norm": 1.735605989780002, + "learning_rate": 6.177483507502766e-06, + "loss": 0.944, + "step": 6613 + }, + { + "epoch": 0.6362062331666025, + "grad_norm": 2.0418783357013424, + "learning_rate": 6.174604860590081e-06, + "loss": 0.8736, + "step": 6614 + }, + { + "epoch": 0.6363024240092343, + "grad_norm": 1.867325601256191, + "learning_rate": 6.171726584964932e-06, + "loss": 0.9901, + "step": 6615 + }, + { + "epoch": 0.6363986148518661, + "grad_norm": 1.9809197711419793, + "learning_rate": 6.168848680906678e-06, + "loss": 0.9931, + "step": 6616 + }, + { + "epoch": 0.6364948056944979, + "grad_norm": 2.098349764888436, + "learning_rate": 6.165971148694645e-06, + "loss": 0.8907, + "step": 6617 + }, + { + "epoch": 0.6365909965371297, + "grad_norm": 1.8512655509685567, + "learning_rate": 6.163093988608127e-06, + "loss": 1.0023, + "step": 6618 + }, + { + "epoch": 0.6366871873797615, + "grad_norm": 2.001039586445718, + "learning_rate": 6.160217200926371e-06, + "loss": 0.9682, + "step": 6619 + }, + { + "epoch": 0.6367833782223933, + "grad_norm": 1.8242121219199567, + "learning_rate": 6.157340785928595e-06, + "loss": 0.9727, + "step": 6620 + }, + { + "epoch": 0.636879569065025, + "grad_norm": 2.3056293494863995, + "learning_rate": 6.154464743893987e-06, + "loss": 0.8075, + "step": 6621 + }, + { + "epoch": 0.6369757599076568, + "grad_norm": 1.764605913982995, + "learning_rate": 6.151589075101681e-06, + "loss": 1.0033, + "step": 6622 + }, + { + "epoch": 0.6370719507502886, + "grad_norm": 2.1470659072277747, + "learning_rate": 6.148713779830793e-06, + "loss": 0.9801, + "step": 6623 + }, + { + "epoch": 0.6371681415929203, + "grad_norm": 2.007495522129916, + "learning_rate": 6.145838858360391e-06, + "loss": 0.9209, + "step": 6624 + }, + { + "epoch": 0.6372643324355521, + "grad_norm": 1.5555529841693638, + "learning_rate": 6.142964310969513e-06, + "loss": 0.8061, + "step": 6625 + }, + { + "epoch": 0.6373605232781839, + "grad_norm": 1.922074420706846, + "learning_rate": 6.140090137937158e-06, + "loss": 1.0083, + "step": 6626 + }, + { + "epoch": 0.6374567141208157, + "grad_norm": 1.8717282228496537, + "learning_rate": 6.137216339542291e-06, + "loss": 0.8967, + "step": 6627 + }, + { + "epoch": 0.6375529049634475, + "grad_norm": 1.9248821670420673, + "learning_rate": 6.134342916063838e-06, + "loss": 0.8695, + "step": 6628 + }, + { + "epoch": 0.6376490958060793, + "grad_norm": 1.970595782901586, + "learning_rate": 6.131469867780684e-06, + "loss": 0.9915, + "step": 6629 + }, + { + "epoch": 0.637745286648711, + "grad_norm": 1.7757748646613667, + "learning_rate": 6.128597194971691e-06, + "loss": 1.0227, + "step": 6630 + }, + { + "epoch": 0.6378414774913428, + "grad_norm": 1.9656331525155217, + "learning_rate": 6.125724897915673e-06, + "loss": 0.9686, + "step": 6631 + }, + { + "epoch": 0.6379376683339746, + "grad_norm": 1.751301174148949, + "learning_rate": 6.122852976891413e-06, + "loss": 0.8702, + "step": 6632 + }, + { + "epoch": 0.6380338591766064, + "grad_norm": 2.332028509570936, + "learning_rate": 6.119981432177653e-06, + "loss": 0.8831, + "step": 6633 + }, + { + "epoch": 0.6381300500192382, + "grad_norm": 2.058649795618286, + "learning_rate": 6.117110264053101e-06, + "loss": 0.8627, + "step": 6634 + }, + { + "epoch": 0.63822624086187, + "grad_norm": 2.0469996083687794, + "learning_rate": 6.11423947279643e-06, + "loss": 0.8949, + "step": 6635 + }, + { + "epoch": 0.6383224317045018, + "grad_norm": 1.9786086272097783, + "learning_rate": 6.111369058686276e-06, + "loss": 0.9228, + "step": 6636 + }, + { + "epoch": 0.6384186225471336, + "grad_norm": 1.9721200237383003, + "learning_rate": 6.108499022001237e-06, + "loss": 0.9688, + "step": 6637 + }, + { + "epoch": 0.6385148133897653, + "grad_norm": 2.1612233062621216, + "learning_rate": 6.105629363019875e-06, + "loss": 0.9621, + "step": 6638 + }, + { + "epoch": 0.638611004232397, + "grad_norm": 2.2719902007734456, + "learning_rate": 6.1027600820207134e-06, + "loss": 0.9865, + "step": 6639 + }, + { + "epoch": 0.6387071950750288, + "grad_norm": 1.9602123319880538, + "learning_rate": 6.099891179282242e-06, + "loss": 0.9054, + "step": 6640 + }, + { + "epoch": 0.6388033859176606, + "grad_norm": 2.0353051439773466, + "learning_rate": 6.097022655082908e-06, + "loss": 0.8866, + "step": 6641 + }, + { + "epoch": 0.6388995767602924, + "grad_norm": 2.0577770935685535, + "learning_rate": 6.094154509701133e-06, + "loss": 0.9759, + "step": 6642 + }, + { + "epoch": 0.6389957676029242, + "grad_norm": 2.2125822671737665, + "learning_rate": 6.091286743415295e-06, + "loss": 0.9188, + "step": 6643 + }, + { + "epoch": 0.639091958445556, + "grad_norm": 1.6632505107865985, + "learning_rate": 6.088419356503732e-06, + "loss": 0.7609, + "step": 6644 + }, + { + "epoch": 0.6391881492881878, + "grad_norm": 2.167085056617668, + "learning_rate": 6.085552349244747e-06, + "loss": 0.9092, + "step": 6645 + }, + { + "epoch": 0.6392843401308196, + "grad_norm": 2.246940654121295, + "learning_rate": 6.082685721916612e-06, + "loss": 0.9758, + "step": 6646 + }, + { + "epoch": 0.6393805309734514, + "grad_norm": 2.0552670867211824, + "learning_rate": 6.079819474797557e-06, + "loss": 0.8854, + "step": 6647 + }, + { + "epoch": 0.6394767218160831, + "grad_norm": 2.0710403544173146, + "learning_rate": 6.076953608165772e-06, + "loss": 0.9717, + "step": 6648 + }, + { + "epoch": 0.6395729126587149, + "grad_norm": 2.0625137904137176, + "learning_rate": 6.074088122299419e-06, + "loss": 0.8395, + "step": 6649 + }, + { + "epoch": 0.6396691035013466, + "grad_norm": 1.7407025063372628, + "learning_rate": 6.07122301747662e-06, + "loss": 0.7737, + "step": 6650 + }, + { + "epoch": 0.6397652943439784, + "grad_norm": 2.102299832460372, + "learning_rate": 6.06835829397545e-06, + "loss": 1.027, + "step": 6651 + }, + { + "epoch": 0.6398614851866102, + "grad_norm": 2.0047647726792426, + "learning_rate": 6.065493952073961e-06, + "loss": 0.9353, + "step": 6652 + }, + { + "epoch": 0.639957676029242, + "grad_norm": 2.055725261500478, + "learning_rate": 6.062629992050158e-06, + "loss": 0.9826, + "step": 6653 + }, + { + "epoch": 0.6400538668718738, + "grad_norm": 1.912666664344225, + "learning_rate": 6.0597664141820176e-06, + "loss": 0.968, + "step": 6654 + }, + { + "epoch": 0.6401500577145056, + "grad_norm": 2.2288339326011797, + "learning_rate": 6.056903218747474e-06, + "loss": 0.9299, + "step": 6655 + }, + { + "epoch": 0.6402462485571374, + "grad_norm": 1.864815490483885, + "learning_rate": 6.054040406024422e-06, + "loss": 0.962, + "step": 6656 + }, + { + "epoch": 0.6403424393997691, + "grad_norm": 2.0263023435620098, + "learning_rate": 6.0511779762907215e-06, + "loss": 1.036, + "step": 6657 + }, + { + "epoch": 0.6404386302424009, + "grad_norm": 2.119701026295236, + "learning_rate": 6.0483159298242e-06, + "loss": 0.9881, + "step": 6658 + }, + { + "epoch": 0.6405348210850327, + "grad_norm": 1.7990446098543473, + "learning_rate": 6.045454266902643e-06, + "loss": 0.9786, + "step": 6659 + }, + { + "epoch": 0.6406310119276645, + "grad_norm": 1.8223200944336506, + "learning_rate": 6.042592987803796e-06, + "loss": 0.8071, + "step": 6660 + }, + { + "epoch": 0.6407272027702963, + "grad_norm": 1.9725195437219198, + "learning_rate": 6.039732092805379e-06, + "loss": 0.9364, + "step": 6661 + }, + { + "epoch": 0.6408233936129281, + "grad_norm": 2.039791764825053, + "learning_rate": 6.036871582185054e-06, + "loss": 0.8632, + "step": 6662 + }, + { + "epoch": 0.6409195844555599, + "grad_norm": 1.7400042015675745, + "learning_rate": 6.034011456220468e-06, + "loss": 0.9436, + "step": 6663 + }, + { + "epoch": 0.6410157752981916, + "grad_norm": 1.9534491465683268, + "learning_rate": 6.031151715189217e-06, + "loss": 0.9594, + "step": 6664 + }, + { + "epoch": 0.6411119661408234, + "grad_norm": 2.012288060856916, + "learning_rate": 6.028292359368864e-06, + "loss": 0.9567, + "step": 6665 + }, + { + "epoch": 0.6412081569834551, + "grad_norm": 2.151355151060917, + "learning_rate": 6.025433389036935e-06, + "loss": 1.0002, + "step": 6666 + }, + { + "epoch": 0.6413043478260869, + "grad_norm": 2.265991894889201, + "learning_rate": 6.0225748044709175e-06, + "loss": 1.0282, + "step": 6667 + }, + { + "epoch": 0.6414005386687187, + "grad_norm": 2.3847965643144673, + "learning_rate": 6.019716605948261e-06, + "loss": 0.9028, + "step": 6668 + }, + { + "epoch": 0.6414967295113505, + "grad_norm": 1.9558722186415054, + "learning_rate": 6.0168587937463765e-06, + "loss": 0.8578, + "step": 6669 + }, + { + "epoch": 0.6415929203539823, + "grad_norm": 2.0795126052384822, + "learning_rate": 6.014001368142643e-06, + "loss": 0.8779, + "step": 6670 + }, + { + "epoch": 0.6416891111966141, + "grad_norm": 1.7299372575290843, + "learning_rate": 6.0111443294143974e-06, + "loss": 0.8925, + "step": 6671 + }, + { + "epoch": 0.6417853020392459, + "grad_norm": 1.95472564476186, + "learning_rate": 6.008287677838937e-06, + "loss": 0.9674, + "step": 6672 + }, + { + "epoch": 0.6418814928818777, + "grad_norm": 1.8827620031648549, + "learning_rate": 6.005431413693532e-06, + "loss": 0.9102, + "step": 6673 + }, + { + "epoch": 0.6419776837245095, + "grad_norm": 2.284997684421834, + "learning_rate": 6.002575537255395e-06, + "loss": 0.9367, + "step": 6674 + }, + { + "epoch": 0.6420738745671412, + "grad_norm": 2.1876166782056776, + "learning_rate": 5.9997200488017245e-06, + "loss": 0.9257, + "step": 6675 + }, + { + "epoch": 0.642170065409773, + "grad_norm": 1.889714121268612, + "learning_rate": 5.996864948609662e-06, + "loss": 0.8588, + "step": 6676 + }, + { + "epoch": 0.6422662562524047, + "grad_norm": 2.1572712053133207, + "learning_rate": 5.994010236956327e-06, + "loss": 0.8174, + "step": 6677 + }, + { + "epoch": 0.6423624470950365, + "grad_norm": 1.863212295747941, + "learning_rate": 5.9911559141187924e-06, + "loss": 0.9566, + "step": 6678 + }, + { + "epoch": 0.6424586379376683, + "grad_norm": 2.4092033557878225, + "learning_rate": 5.98830198037409e-06, + "loss": 1.0258, + "step": 6679 + }, + { + "epoch": 0.6425548287803001, + "grad_norm": 2.0306390523837536, + "learning_rate": 5.9854484359992235e-06, + "loss": 1.1069, + "step": 6680 + }, + { + "epoch": 0.6426510196229319, + "grad_norm": 1.6452610104456291, + "learning_rate": 5.982595281271148e-06, + "loss": 0.7728, + "step": 6681 + }, + { + "epoch": 0.6427472104655637, + "grad_norm": 1.8085313552941167, + "learning_rate": 5.979742516466793e-06, + "loss": 0.7491, + "step": 6682 + }, + { + "epoch": 0.6428434013081955, + "grad_norm": 2.1628665197618466, + "learning_rate": 5.976890141863044e-06, + "loss": 0.9594, + "step": 6683 + }, + { + "epoch": 0.6429395921508272, + "grad_norm": 2.1637451272223776, + "learning_rate": 5.974038157736746e-06, + "loss": 0.963, + "step": 6684 + }, + { + "epoch": 0.643035782993459, + "grad_norm": 1.9788790165920778, + "learning_rate": 5.971186564364704e-06, + "loss": 0.9623, + "step": 6685 + }, + { + "epoch": 0.6431319738360908, + "grad_norm": 1.899374909439092, + "learning_rate": 5.968335362023697e-06, + "loss": 0.9396, + "step": 6686 + }, + { + "epoch": 0.6432281646787226, + "grad_norm": 1.8413292855716634, + "learning_rate": 5.965484550990457e-06, + "loss": 1.0336, + "step": 6687 + }, + { + "epoch": 0.6433243555213544, + "grad_norm": 1.988737635537535, + "learning_rate": 5.962634131541676e-06, + "loss": 0.8335, + "step": 6688 + }, + { + "epoch": 0.6434205463639862, + "grad_norm": 1.967217761401454, + "learning_rate": 5.959784103954018e-06, + "loss": 0.9859, + "step": 6689 + }, + { + "epoch": 0.643516737206618, + "grad_norm": 2.1233352723442245, + "learning_rate": 5.956934468504101e-06, + "loss": 0.9051, + "step": 6690 + }, + { + "epoch": 0.6436129280492497, + "grad_norm": 2.2190936275089017, + "learning_rate": 5.954085225468502e-06, + "loss": 0.8984, + "step": 6691 + }, + { + "epoch": 0.6437091188918815, + "grad_norm": 2.027096009958846, + "learning_rate": 5.951236375123768e-06, + "loss": 0.9883, + "step": 6692 + }, + { + "epoch": 0.6438053097345132, + "grad_norm": 2.0489075486385984, + "learning_rate": 5.9483879177464035e-06, + "loss": 1.0269, + "step": 6693 + }, + { + "epoch": 0.643901500577145, + "grad_norm": 1.9219592774399092, + "learning_rate": 5.945539853612876e-06, + "loss": 0.8474, + "step": 6694 + }, + { + "epoch": 0.6439976914197768, + "grad_norm": 2.021462412897448, + "learning_rate": 5.942692182999616e-06, + "loss": 0.8563, + "step": 6695 + }, + { + "epoch": 0.6440938822624086, + "grad_norm": 2.1646502469670503, + "learning_rate": 5.939844906183016e-06, + "loss": 1.0033, + "step": 6696 + }, + { + "epoch": 0.6441900731050404, + "grad_norm": 2.166894805440856, + "learning_rate": 5.936998023439423e-06, + "loss": 0.9077, + "step": 6697 + }, + { + "epoch": 0.6442862639476722, + "grad_norm": 2.0194505036579358, + "learning_rate": 5.934151535045156e-06, + "loss": 0.8479, + "step": 6698 + }, + { + "epoch": 0.644382454790304, + "grad_norm": 1.759964667955383, + "learning_rate": 5.931305441276489e-06, + "loss": 0.8994, + "step": 6699 + }, + { + "epoch": 0.6444786456329358, + "grad_norm": 2.274522838904976, + "learning_rate": 5.92845974240966e-06, + "loss": 0.9879, + "step": 6700 + }, + { + "epoch": 0.6445748364755676, + "grad_norm": 2.087333651401282, + "learning_rate": 5.9256144387208705e-06, + "loss": 1.0981, + "step": 6701 + }, + { + "epoch": 0.6446710273181993, + "grad_norm": 1.8784474228517527, + "learning_rate": 5.922769530486283e-06, + "loss": 0.9331, + "step": 6702 + }, + { + "epoch": 0.644767218160831, + "grad_norm": 1.9446899930940464, + "learning_rate": 5.919925017982016e-06, + "loss": 0.9495, + "step": 6703 + }, + { + "epoch": 0.6448634090034628, + "grad_norm": 1.9639299184742351, + "learning_rate": 5.917080901484156e-06, + "loss": 0.9373, + "step": 6704 + }, + { + "epoch": 0.6449595998460946, + "grad_norm": 1.8293734625277638, + "learning_rate": 5.914237181268747e-06, + "loss": 0.9456, + "step": 6705 + }, + { + "epoch": 0.6450557906887264, + "grad_norm": 1.9296477907503073, + "learning_rate": 5.9113938576118e-06, + "loss": 0.9643, + "step": 6706 + }, + { + "epoch": 0.6451519815313582, + "grad_norm": 2.0942135733353218, + "learning_rate": 5.908550930789285e-06, + "loss": 0.9191, + "step": 6707 + }, + { + "epoch": 0.64524817237399, + "grad_norm": 2.2228881990646205, + "learning_rate": 5.905708401077128e-06, + "loss": 0.9588, + "step": 6708 + }, + { + "epoch": 0.6453443632166218, + "grad_norm": 2.040192726505004, + "learning_rate": 5.90286626875122e-06, + "loss": 0.9503, + "step": 6709 + }, + { + "epoch": 0.6454405540592536, + "grad_norm": 2.134318516558896, + "learning_rate": 5.900024534087421e-06, + "loss": 0.8921, + "step": 6710 + }, + { + "epoch": 0.6455367449018854, + "grad_norm": 1.894165364020166, + "learning_rate": 5.89718319736154e-06, + "loss": 0.8732, + "step": 6711 + }, + { + "epoch": 0.6456329357445171, + "grad_norm": 2.24980935375773, + "learning_rate": 5.894342258849355e-06, + "loss": 0.8748, + "step": 6712 + }, + { + "epoch": 0.6457291265871489, + "grad_norm": 1.8187813726221065, + "learning_rate": 5.891501718826609e-06, + "loss": 0.8464, + "step": 6713 + }, + { + "epoch": 0.6458253174297807, + "grad_norm": 2.0594459212858203, + "learning_rate": 5.88866157756899e-06, + "loss": 0.9516, + "step": 6714 + }, + { + "epoch": 0.6459215082724125, + "grad_norm": 2.1734811827920018, + "learning_rate": 5.885821835352167e-06, + "loss": 0.8379, + "step": 6715 + }, + { + "epoch": 0.6460176991150443, + "grad_norm": 1.9494816651050504, + "learning_rate": 5.882982492451757e-06, + "loss": 0.9554, + "step": 6716 + }, + { + "epoch": 0.646113889957676, + "grad_norm": 2.0518240328807975, + "learning_rate": 5.8801435491433435e-06, + "loss": 0.885, + "step": 6717 + }, + { + "epoch": 0.6462100808003078, + "grad_norm": 2.114429129893731, + "learning_rate": 5.877305005702471e-06, + "loss": 0.9042, + "step": 6718 + }, + { + "epoch": 0.6463062716429396, + "grad_norm": 2.092852310327803, + "learning_rate": 5.8744668624046485e-06, + "loss": 0.929, + "step": 6719 + }, + { + "epoch": 0.6464024624855714, + "grad_norm": 2.0331961070649114, + "learning_rate": 5.871629119525335e-06, + "loss": 0.8902, + "step": 6720 + }, + { + "epoch": 0.6464986533282031, + "grad_norm": 1.9029165739030183, + "learning_rate": 5.86879177733996e-06, + "loss": 0.9683, + "step": 6721 + }, + { + "epoch": 0.6465948441708349, + "grad_norm": 2.1159490418992717, + "learning_rate": 5.865954836123915e-06, + "loss": 0.8553, + "step": 6722 + }, + { + "epoch": 0.6466910350134667, + "grad_norm": 2.051567665162532, + "learning_rate": 5.863118296152548e-06, + "loss": 0.9676, + "step": 6723 + }, + { + "epoch": 0.6467872258560985, + "grad_norm": 2.0630288307399036, + "learning_rate": 5.860282157701167e-06, + "loss": 0.8879, + "step": 6724 + }, + { + "epoch": 0.6468834166987303, + "grad_norm": 2.013416142349825, + "learning_rate": 5.85744642104505e-06, + "loss": 0.9671, + "step": 6725 + }, + { + "epoch": 0.6469796075413621, + "grad_norm": 2.0420904101261077, + "learning_rate": 5.854611086459423e-06, + "loss": 1.003, + "step": 6726 + }, + { + "epoch": 0.6470757983839939, + "grad_norm": 2.1030647012594463, + "learning_rate": 5.8517761542194825e-06, + "loss": 1.0708, + "step": 6727 + }, + { + "epoch": 0.6471719892266257, + "grad_norm": 1.6073153225002503, + "learning_rate": 5.8489416246003814e-06, + "loss": 0.7366, + "step": 6728 + }, + { + "epoch": 0.6472681800692575, + "grad_norm": 1.9188089957687995, + "learning_rate": 5.846107497877238e-06, + "loss": 0.8579, + "step": 6729 + }, + { + "epoch": 0.6473643709118891, + "grad_norm": 2.209599639621399, + "learning_rate": 5.8432737743251315e-06, + "loss": 0.8045, + "step": 6730 + }, + { + "epoch": 0.6474605617545209, + "grad_norm": 2.06296821570073, + "learning_rate": 5.840440454219091e-06, + "loss": 0.8433, + "step": 6731 + }, + { + "epoch": 0.6475567525971527, + "grad_norm": 1.9207543109141008, + "learning_rate": 5.8376075378341194e-06, + "loss": 0.9416, + "step": 6732 + }, + { + "epoch": 0.6476529434397845, + "grad_norm": 1.9752635345227116, + "learning_rate": 5.834775025445172e-06, + "loss": 0.8433, + "step": 6733 + }, + { + "epoch": 0.6477491342824163, + "grad_norm": 2.0706821121625, + "learning_rate": 5.831942917327172e-06, + "loss": 1.0293, + "step": 6734 + }, + { + "epoch": 0.6478453251250481, + "grad_norm": 2.1394603759496946, + "learning_rate": 5.829111213755003e-06, + "loss": 1.1299, + "step": 6735 + }, + { + "epoch": 0.6479415159676799, + "grad_norm": 1.8080875734572457, + "learning_rate": 5.826279915003503e-06, + "loss": 0.8922, + "step": 6736 + }, + { + "epoch": 0.6480377068103117, + "grad_norm": 1.8596634537799581, + "learning_rate": 5.823449021347469e-06, + "loss": 0.9612, + "step": 6737 + }, + { + "epoch": 0.6481338976529435, + "grad_norm": 2.063764587852409, + "learning_rate": 5.8206185330616725e-06, + "loss": 0.936, + "step": 6738 + }, + { + "epoch": 0.6482300884955752, + "grad_norm": 1.9355162002145452, + "learning_rate": 5.8177884504208295e-06, + "loss": 1.0094, + "step": 6739 + }, + { + "epoch": 0.648326279338207, + "grad_norm": 1.9720777211474405, + "learning_rate": 5.814958773699625e-06, + "loss": 0.8947, + "step": 6740 + }, + { + "epoch": 0.6484224701808388, + "grad_norm": 2.07103231782439, + "learning_rate": 5.81212950317271e-06, + "loss": 0.9329, + "step": 6741 + }, + { + "epoch": 0.6485186610234706, + "grad_norm": 2.1750404587254177, + "learning_rate": 5.809300639114683e-06, + "loss": 0.9893, + "step": 6742 + }, + { + "epoch": 0.6486148518661023, + "grad_norm": 2.1931612945878207, + "learning_rate": 5.806472181800109e-06, + "loss": 0.9287, + "step": 6743 + }, + { + "epoch": 0.6487110427087341, + "grad_norm": 2.077753203369903, + "learning_rate": 5.803644131503516e-06, + "loss": 0.9664, + "step": 6744 + }, + { + "epoch": 0.6488072335513659, + "grad_norm": 1.8783952836513684, + "learning_rate": 5.800816488499395e-06, + "loss": 0.8855, + "step": 6745 + }, + { + "epoch": 0.6489034243939977, + "grad_norm": 2.1852688148518427, + "learning_rate": 5.797989253062186e-06, + "loss": 1.0016, + "step": 6746 + }, + { + "epoch": 0.6489996152366295, + "grad_norm": 2.0262378407218353, + "learning_rate": 5.795162425466299e-06, + "loss": 0.9449, + "step": 6747 + }, + { + "epoch": 0.6490958060792612, + "grad_norm": 2.069330558671261, + "learning_rate": 5.792336005986105e-06, + "loss": 1.0203, + "step": 6748 + }, + { + "epoch": 0.649191996921893, + "grad_norm": 1.9595573905845474, + "learning_rate": 5.789509994895933e-06, + "loss": 0.9199, + "step": 6749 + }, + { + "epoch": 0.6492881877645248, + "grad_norm": 2.6477332537047484, + "learning_rate": 5.786684392470064e-06, + "loss": 0.8663, + "step": 6750 + }, + { + "epoch": 0.6493843786071566, + "grad_norm": 2.1625618315320208, + "learning_rate": 5.783859198982751e-06, + "loss": 1.0483, + "step": 6751 + }, + { + "epoch": 0.6494805694497884, + "grad_norm": 2.1054283551778195, + "learning_rate": 5.781034414708208e-06, + "loss": 0.9035, + "step": 6752 + }, + { + "epoch": 0.6495767602924202, + "grad_norm": 2.020414862818825, + "learning_rate": 5.778210039920602e-06, + "loss": 0.9603, + "step": 6753 + }, + { + "epoch": 0.649672951135052, + "grad_norm": 2.0384323960481914, + "learning_rate": 5.775386074894058e-06, + "loss": 1.0689, + "step": 6754 + }, + { + "epoch": 0.6497691419776838, + "grad_norm": 2.050708755500335, + "learning_rate": 5.772562519902669e-06, + "loss": 0.9374, + "step": 6755 + }, + { + "epoch": 0.6498653328203156, + "grad_norm": 2.106799041297636, + "learning_rate": 5.769739375220489e-06, + "loss": 0.9537, + "step": 6756 + }, + { + "epoch": 0.6499615236629472, + "grad_norm": 1.8913516966978758, + "learning_rate": 5.7669166411215225e-06, + "loss": 0.9301, + "step": 6757 + }, + { + "epoch": 0.650057714505579, + "grad_norm": 1.9166862625881989, + "learning_rate": 5.7640943178797445e-06, + "loss": 0.906, + "step": 6758 + }, + { + "epoch": 0.6501539053482108, + "grad_norm": 1.865136766876453, + "learning_rate": 5.7612724057690915e-06, + "loss": 0.9908, + "step": 6759 + }, + { + "epoch": 0.6502500961908426, + "grad_norm": 2.0571461120063286, + "learning_rate": 5.7584509050634395e-06, + "loss": 0.9812, + "step": 6760 + }, + { + "epoch": 0.6503462870334744, + "grad_norm": 2.0664214374511882, + "learning_rate": 5.755629816036648e-06, + "loss": 0.9028, + "step": 6761 + }, + { + "epoch": 0.6504424778761062, + "grad_norm": 1.7635357119842416, + "learning_rate": 5.752809138962525e-06, + "loss": 0.8896, + "step": 6762 + }, + { + "epoch": 0.650538668718738, + "grad_norm": 2.047599278246829, + "learning_rate": 5.74998887411485e-06, + "loss": 0.993, + "step": 6763 + }, + { + "epoch": 0.6506348595613698, + "grad_norm": 2.2357466960371455, + "learning_rate": 5.747169021767342e-06, + "loss": 0.9272, + "step": 6764 + }, + { + "epoch": 0.6507310504040016, + "grad_norm": 1.5863904845911816, + "learning_rate": 5.7443495821937044e-06, + "loss": 0.7862, + "step": 6765 + }, + { + "epoch": 0.6508272412466333, + "grad_norm": 1.8208449947111907, + "learning_rate": 5.7415305556675805e-06, + "loss": 0.9044, + "step": 6766 + }, + { + "epoch": 0.6509234320892651, + "grad_norm": 1.80201077719029, + "learning_rate": 5.738711942462577e-06, + "loss": 0.9507, + "step": 6767 + }, + { + "epoch": 0.6510196229318969, + "grad_norm": 1.9117095289550439, + "learning_rate": 5.73589374285227e-06, + "loss": 0.9802, + "step": 6768 + }, + { + "epoch": 0.6511158137745287, + "grad_norm": 2.2690735336833594, + "learning_rate": 5.733075957110195e-06, + "loss": 0.9644, + "step": 6769 + }, + { + "epoch": 0.6512120046171604, + "grad_norm": 2.133347107906313, + "learning_rate": 5.730258585509832e-06, + "loss": 0.9327, + "step": 6770 + }, + { + "epoch": 0.6513081954597922, + "grad_norm": 1.943754695057202, + "learning_rate": 5.727441628324639e-06, + "loss": 1.014, + "step": 6771 + }, + { + "epoch": 0.651404386302424, + "grad_norm": 1.9147503089015838, + "learning_rate": 5.724625085828022e-06, + "loss": 0.9723, + "step": 6772 + }, + { + "epoch": 0.6515005771450558, + "grad_norm": 1.8946950419355983, + "learning_rate": 5.7218089582933534e-06, + "loss": 0.8851, + "step": 6773 + }, + { + "epoch": 0.6515967679876876, + "grad_norm": 2.2100370069437965, + "learning_rate": 5.718993245993958e-06, + "loss": 0.8614, + "step": 6774 + }, + { + "epoch": 0.6516929588303193, + "grad_norm": 2.224765866609601, + "learning_rate": 5.716177949203128e-06, + "loss": 0.9705, + "step": 6775 + }, + { + "epoch": 0.6517891496729511, + "grad_norm": 2.0780882745789784, + "learning_rate": 5.713363068194115e-06, + "loss": 0.9049, + "step": 6776 + }, + { + "epoch": 0.6518853405155829, + "grad_norm": 1.7743287781291248, + "learning_rate": 5.710548603240127e-06, + "loss": 0.95, + "step": 6777 + }, + { + "epoch": 0.6519815313582147, + "grad_norm": 2.14376402863359, + "learning_rate": 5.7077345546143235e-06, + "loss": 0.9224, + "step": 6778 + }, + { + "epoch": 0.6520777222008465, + "grad_norm": 2.246841818054855, + "learning_rate": 5.7049209225898395e-06, + "loss": 0.8553, + "step": 6779 + }, + { + "epoch": 0.6521739130434783, + "grad_norm": 1.8263594613517529, + "learning_rate": 5.702107707439766e-06, + "loss": 0.9048, + "step": 6780 + }, + { + "epoch": 0.6522701038861101, + "grad_norm": 1.930226925617037, + "learning_rate": 5.699294909437138e-06, + "loss": 0.8299, + "step": 6781 + }, + { + "epoch": 0.6523662947287419, + "grad_norm": 2.1146450058165107, + "learning_rate": 5.6964825288549745e-06, + "loss": 0.8651, + "step": 6782 + }, + { + "epoch": 0.6524624855713737, + "grad_norm": 1.8724983975432452, + "learning_rate": 5.693670565966232e-06, + "loss": 0.8901, + "step": 6783 + }, + { + "epoch": 0.6525586764140053, + "grad_norm": 2.14527499395335, + "learning_rate": 5.690859021043842e-06, + "loss": 0.9188, + "step": 6784 + }, + { + "epoch": 0.6526548672566371, + "grad_norm": 2.0311282679984943, + "learning_rate": 5.688047894360682e-06, + "loss": 0.9654, + "step": 6785 + }, + { + "epoch": 0.6527510580992689, + "grad_norm": 2.416978812063671, + "learning_rate": 5.685237186189601e-06, + "loss": 0.9691, + "step": 6786 + }, + { + "epoch": 0.6528472489419007, + "grad_norm": 1.988199040657682, + "learning_rate": 5.682426896803403e-06, + "loss": 0.7919, + "step": 6787 + }, + { + "epoch": 0.6529434397845325, + "grad_norm": 2.1650552774698943, + "learning_rate": 5.679617026474853e-06, + "loss": 1.0366, + "step": 6788 + }, + { + "epoch": 0.6530396306271643, + "grad_norm": 1.8615780960761366, + "learning_rate": 5.676807575476664e-06, + "loss": 0.896, + "step": 6789 + }, + { + "epoch": 0.6531358214697961, + "grad_norm": 1.8147238571429174, + "learning_rate": 5.673998544081527e-06, + "loss": 0.9049, + "step": 6790 + }, + { + "epoch": 0.6532320123124279, + "grad_norm": 2.1178569733211043, + "learning_rate": 5.6711899325620745e-06, + "loss": 0.9116, + "step": 6791 + }, + { + "epoch": 0.6533282031550597, + "grad_norm": 2.0606708739140216, + "learning_rate": 5.6683817411909114e-06, + "loss": 0.9633, + "step": 6792 + }, + { + "epoch": 0.6534243939976914, + "grad_norm": 2.1060389490174174, + "learning_rate": 5.6655739702405945e-06, + "loss": 0.9883, + "step": 6793 + }, + { + "epoch": 0.6535205848403232, + "grad_norm": 2.2228885241427543, + "learning_rate": 5.662766619983653e-06, + "loss": 0.8797, + "step": 6794 + }, + { + "epoch": 0.653616775682955, + "grad_norm": 1.9105214966680935, + "learning_rate": 5.659959690692547e-06, + "loss": 0.9266, + "step": 6795 + }, + { + "epoch": 0.6537129665255867, + "grad_norm": 1.8893708417753132, + "learning_rate": 5.65715318263972e-06, + "loss": 0.9279, + "step": 6796 + }, + { + "epoch": 0.6538091573682185, + "grad_norm": 1.990662789713451, + "learning_rate": 5.654347096097575e-06, + "loss": 0.9545, + "step": 6797 + }, + { + "epoch": 0.6539053482108503, + "grad_norm": 2.378459482344992, + "learning_rate": 5.651541431338454e-06, + "loss": 0.9294, + "step": 6798 + }, + { + "epoch": 0.6540015390534821, + "grad_norm": 2.2343809371906325, + "learning_rate": 5.64873618863468e-06, + "loss": 1.0894, + "step": 6799 + }, + { + "epoch": 0.6540977298961139, + "grad_norm": 2.130119203261282, + "learning_rate": 5.645931368258527e-06, + "loss": 0.9795, + "step": 6800 + }, + { + "epoch": 0.6541939207387457, + "grad_norm": 2.317742072232144, + "learning_rate": 5.643126970482225e-06, + "loss": 0.8893, + "step": 6801 + }, + { + "epoch": 0.6542901115813774, + "grad_norm": 2.0223334717205037, + "learning_rate": 5.640322995577958e-06, + "loss": 0.9192, + "step": 6802 + }, + { + "epoch": 0.6543863024240092, + "grad_norm": 1.8537174530250007, + "learning_rate": 5.637519443817882e-06, + "loss": 0.8394, + "step": 6803 + }, + { + "epoch": 0.654482493266641, + "grad_norm": 2.100027438557215, + "learning_rate": 5.634716315474109e-06, + "loss": 1.0122, + "step": 6804 + }, + { + "epoch": 0.6545786841092728, + "grad_norm": 1.8193490004459003, + "learning_rate": 5.631913610818704e-06, + "loss": 0.8754, + "step": 6805 + }, + { + "epoch": 0.6546748749519046, + "grad_norm": 2.039185003038868, + "learning_rate": 5.629111330123689e-06, + "loss": 0.992, + "step": 6806 + }, + { + "epoch": 0.6547710657945364, + "grad_norm": 2.415117463066418, + "learning_rate": 5.626309473661051e-06, + "loss": 0.8371, + "step": 6807 + }, + { + "epoch": 0.6548672566371682, + "grad_norm": 1.7997148749946377, + "learning_rate": 5.623508041702743e-06, + "loss": 1.0017, + "step": 6808 + }, + { + "epoch": 0.6549634474798, + "grad_norm": 2.196073484767749, + "learning_rate": 5.620707034520657e-06, + "loss": 0.9929, + "step": 6809 + }, + { + "epoch": 0.6550596383224317, + "grad_norm": 1.986670831260111, + "learning_rate": 5.617906452386659e-06, + "loss": 0.8996, + "step": 6810 + }, + { + "epoch": 0.6551558291650635, + "grad_norm": 2.2172666278594377, + "learning_rate": 5.615106295572574e-06, + "loss": 0.8058, + "step": 6811 + }, + { + "epoch": 0.6552520200076952, + "grad_norm": 2.2500421058096225, + "learning_rate": 5.612306564350179e-06, + "loss": 0.8547, + "step": 6812 + }, + { + "epoch": 0.655348210850327, + "grad_norm": 1.9416347646487575, + "learning_rate": 5.609507258991205e-06, + "loss": 0.9646, + "step": 6813 + }, + { + "epoch": 0.6554444016929588, + "grad_norm": 2.1469208698207725, + "learning_rate": 5.6067083797673535e-06, + "loss": 1.0284, + "step": 6814 + }, + { + "epoch": 0.6555405925355906, + "grad_norm": 2.09343266971347, + "learning_rate": 5.603909926950286e-06, + "loss": 0.875, + "step": 6815 + }, + { + "epoch": 0.6556367833782224, + "grad_norm": 2.5064657076197823, + "learning_rate": 5.601111900811607e-06, + "loss": 0.9156, + "step": 6816 + }, + { + "epoch": 0.6557329742208542, + "grad_norm": 1.8189506002467006, + "learning_rate": 5.598314301622896e-06, + "loss": 0.9306, + "step": 6817 + }, + { + "epoch": 0.655829165063486, + "grad_norm": 1.8593933910098055, + "learning_rate": 5.595517129655681e-06, + "loss": 0.9611, + "step": 6818 + }, + { + "epoch": 0.6559253559061178, + "grad_norm": 1.9003372422502893, + "learning_rate": 5.592720385181447e-06, + "loss": 0.8559, + "step": 6819 + }, + { + "epoch": 0.6560215467487496, + "grad_norm": 1.7808381244469722, + "learning_rate": 5.589924068471648e-06, + "loss": 0.9777, + "step": 6820 + }, + { + "epoch": 0.6561177375913813, + "grad_norm": 2.0814460542180653, + "learning_rate": 5.5871281797976925e-06, + "loss": 0.9483, + "step": 6821 + }, + { + "epoch": 0.656213928434013, + "grad_norm": 1.8939330129679246, + "learning_rate": 5.58433271943094e-06, + "loss": 0.8964, + "step": 6822 + }, + { + "epoch": 0.6563101192766448, + "grad_norm": 2.004316157663629, + "learning_rate": 5.581537687642718e-06, + "loss": 0.9634, + "step": 6823 + }, + { + "epoch": 0.6564063101192766, + "grad_norm": 1.86394719797808, + "learning_rate": 5.578743084704306e-06, + "loss": 0.942, + "step": 6824 + }, + { + "epoch": 0.6565025009619084, + "grad_norm": 1.7395099421719487, + "learning_rate": 5.575948910886946e-06, + "loss": 0.9221, + "step": 6825 + }, + { + "epoch": 0.6565986918045402, + "grad_norm": 1.5369015858676205, + "learning_rate": 5.573155166461833e-06, + "loss": 0.8076, + "step": 6826 + }, + { + "epoch": 0.656694882647172, + "grad_norm": 1.9044154735303098, + "learning_rate": 5.570361851700127e-06, + "loss": 1.0077, + "step": 6827 + }, + { + "epoch": 0.6567910734898038, + "grad_norm": 2.065876771247427, + "learning_rate": 5.567568966872947e-06, + "loss": 0.9322, + "step": 6828 + }, + { + "epoch": 0.6568872643324356, + "grad_norm": 2.283254307929903, + "learning_rate": 5.564776512251362e-06, + "loss": 1.0151, + "step": 6829 + }, + { + "epoch": 0.6569834551750673, + "grad_norm": 1.9596129685347485, + "learning_rate": 5.5619844881064e-06, + "loss": 0.9569, + "step": 6830 + }, + { + "epoch": 0.6570796460176991, + "grad_norm": 2.025082139710384, + "learning_rate": 5.5591928947090554e-06, + "loss": 0.9626, + "step": 6831 + }, + { + "epoch": 0.6571758368603309, + "grad_norm": 2.002995581835142, + "learning_rate": 5.556401732330281e-06, + "loss": 0.9511, + "step": 6832 + }, + { + "epoch": 0.6572720277029627, + "grad_norm": 2.0285726640139266, + "learning_rate": 5.553611001240974e-06, + "loss": 1.0016, + "step": 6833 + }, + { + "epoch": 0.6573682185455945, + "grad_norm": 2.3101201774059277, + "learning_rate": 5.550820701712007e-06, + "loss": 0.9199, + "step": 6834 + }, + { + "epoch": 0.6574644093882263, + "grad_norm": 1.8175031965369024, + "learning_rate": 5.548030834014195e-06, + "loss": 0.9937, + "step": 6835 + }, + { + "epoch": 0.657560600230858, + "grad_norm": 1.9821465016999777, + "learning_rate": 5.545241398418326e-06, + "loss": 0.9212, + "step": 6836 + }, + { + "epoch": 0.6576567910734898, + "grad_norm": 2.2402149508867004, + "learning_rate": 5.542452395195131e-06, + "loss": 0.8959, + "step": 6837 + }, + { + "epoch": 0.6577529819161216, + "grad_norm": 2.0305709785552906, + "learning_rate": 5.539663824615312e-06, + "loss": 0.8793, + "step": 6838 + }, + { + "epoch": 0.6578491727587533, + "grad_norm": 1.9650898226612907, + "learning_rate": 5.536875686949527e-06, + "loss": 0.9753, + "step": 6839 + }, + { + "epoch": 0.6579453636013851, + "grad_norm": 1.8713794149547516, + "learning_rate": 5.534087982468384e-06, + "loss": 0.9451, + "step": 6840 + }, + { + "epoch": 0.6580415544440169, + "grad_norm": 1.6941759546344888, + "learning_rate": 5.531300711442449e-06, + "loss": 0.8602, + "step": 6841 + }, + { + "epoch": 0.6581377452866487, + "grad_norm": 1.7498766371482166, + "learning_rate": 5.5285138741422615e-06, + "loss": 0.9162, + "step": 6842 + }, + { + "epoch": 0.6582339361292805, + "grad_norm": 2.0409951838088984, + "learning_rate": 5.525727470838298e-06, + "loss": 0.8346, + "step": 6843 + }, + { + "epoch": 0.6583301269719123, + "grad_norm": 2.0006812910085263, + "learning_rate": 5.522941501801008e-06, + "loss": 0.9329, + "step": 6844 + }, + { + "epoch": 0.6584263178145441, + "grad_norm": 2.049905558052995, + "learning_rate": 5.520155967300793e-06, + "loss": 0.8817, + "step": 6845 + }, + { + "epoch": 0.6585225086571759, + "grad_norm": 2.014867100796319, + "learning_rate": 5.517370867608021e-06, + "loss": 0.9057, + "step": 6846 + }, + { + "epoch": 0.6586186994998077, + "grad_norm": 2.0757793451062443, + "learning_rate": 5.514586202992994e-06, + "loss": 0.9332, + "step": 6847 + }, + { + "epoch": 0.6587148903424394, + "grad_norm": 2.1650457521038433, + "learning_rate": 5.511801973725997e-06, + "loss": 0.948, + "step": 6848 + }, + { + "epoch": 0.6588110811850711, + "grad_norm": 1.8834714899318932, + "learning_rate": 5.509018180077264e-06, + "loss": 0.8261, + "step": 6849 + }, + { + "epoch": 0.6589072720277029, + "grad_norm": 1.9207206408502477, + "learning_rate": 5.506234822316983e-06, + "loss": 0.7717, + "step": 6850 + }, + { + "epoch": 0.6590034628703347, + "grad_norm": 2.3555056781455006, + "learning_rate": 5.503451900715302e-06, + "loss": 0.9197, + "step": 6851 + }, + { + "epoch": 0.6590996537129665, + "grad_norm": 2.5654703708928053, + "learning_rate": 5.500669415542336e-06, + "loss": 0.8067, + "step": 6852 + }, + { + "epoch": 0.6591958445555983, + "grad_norm": 2.0858855510312435, + "learning_rate": 5.497887367068142e-06, + "loss": 0.9732, + "step": 6853 + }, + { + "epoch": 0.6592920353982301, + "grad_norm": 2.042335555905552, + "learning_rate": 5.495105755562738e-06, + "loss": 1.0353, + "step": 6854 + }, + { + "epoch": 0.6593882262408619, + "grad_norm": 1.7072985142394408, + "learning_rate": 5.492324581296109e-06, + "loss": 0.8325, + "step": 6855 + }, + { + "epoch": 0.6594844170834937, + "grad_norm": 1.9716045405456113, + "learning_rate": 5.4895438445381945e-06, + "loss": 0.9629, + "step": 6856 + }, + { + "epoch": 0.6595806079261254, + "grad_norm": 1.9037565370547922, + "learning_rate": 5.486763545558887e-06, + "loss": 0.8523, + "step": 6857 + }, + { + "epoch": 0.6596767987687572, + "grad_norm": 1.6778184746367637, + "learning_rate": 5.48398368462803e-06, + "loss": 0.8268, + "step": 6858 + }, + { + "epoch": 0.659772989611389, + "grad_norm": 1.8106885821069716, + "learning_rate": 5.481204262015441e-06, + "loss": 0.9844, + "step": 6859 + }, + { + "epoch": 0.6598691804540208, + "grad_norm": 2.050947648331127, + "learning_rate": 5.4784252779908905e-06, + "loss": 0.916, + "step": 6860 + }, + { + "epoch": 0.6599653712966526, + "grad_norm": 2.069074195670621, + "learning_rate": 5.475646732824094e-06, + "loss": 0.9655, + "step": 6861 + }, + { + "epoch": 0.6600615621392844, + "grad_norm": 2.3237770435864684, + "learning_rate": 5.4728686267847354e-06, + "loss": 0.9593, + "step": 6862 + }, + { + "epoch": 0.6601577529819161, + "grad_norm": 1.8934695228348748, + "learning_rate": 5.470090960142462e-06, + "loss": 0.9279, + "step": 6863 + }, + { + "epoch": 0.6602539438245479, + "grad_norm": 1.716127037528629, + "learning_rate": 5.467313733166863e-06, + "loss": 0.8548, + "step": 6864 + }, + { + "epoch": 0.6603501346671797, + "grad_norm": 2.0978722345819287, + "learning_rate": 5.464536946127488e-06, + "loss": 0.9297, + "step": 6865 + }, + { + "epoch": 0.6604463255098114, + "grad_norm": 1.9492154822343564, + "learning_rate": 5.461760599293855e-06, + "loss": 0.9889, + "step": 6866 + }, + { + "epoch": 0.6605425163524432, + "grad_norm": 1.878931979235015, + "learning_rate": 5.458984692935434e-06, + "loss": 0.9819, + "step": 6867 + }, + { + "epoch": 0.660638707195075, + "grad_norm": 1.9234444942918247, + "learning_rate": 5.456209227321643e-06, + "loss": 0.8875, + "step": 6868 + }, + { + "epoch": 0.6607348980377068, + "grad_norm": 1.9261513859623434, + "learning_rate": 5.453434202721872e-06, + "loss": 1.0072, + "step": 6869 + }, + { + "epoch": 0.6608310888803386, + "grad_norm": 2.0820034980967277, + "learning_rate": 5.450659619405458e-06, + "loss": 0.9838, + "step": 6870 + }, + { + "epoch": 0.6609272797229704, + "grad_norm": 1.8766244746947438, + "learning_rate": 5.447885477641697e-06, + "loss": 1.0212, + "step": 6871 + }, + { + "epoch": 0.6610234705656022, + "grad_norm": 2.1108324749162057, + "learning_rate": 5.445111777699842e-06, + "loss": 0.9795, + "step": 6872 + }, + { + "epoch": 0.661119661408234, + "grad_norm": 2.0500362755728445, + "learning_rate": 5.442338519849112e-06, + "loss": 0.9958, + "step": 6873 + }, + { + "epoch": 0.6612158522508658, + "grad_norm": 1.7429859160247672, + "learning_rate": 5.439565704358667e-06, + "loss": 0.941, + "step": 6874 + }, + { + "epoch": 0.6613120430934974, + "grad_norm": 1.8113576397312237, + "learning_rate": 5.4367933314976385e-06, + "loss": 0.784, + "step": 6875 + }, + { + "epoch": 0.6614082339361292, + "grad_norm": 2.053319345643632, + "learning_rate": 5.434021401535105e-06, + "loss": 0.8593, + "step": 6876 + }, + { + "epoch": 0.661504424778761, + "grad_norm": 2.141411619806518, + "learning_rate": 5.431249914740111e-06, + "loss": 0.9711, + "step": 6877 + }, + { + "epoch": 0.6616006156213928, + "grad_norm": 1.806644882391706, + "learning_rate": 5.428478871381646e-06, + "loss": 0.9739, + "step": 6878 + }, + { + "epoch": 0.6616968064640246, + "grad_norm": 2.293280169651588, + "learning_rate": 5.425708271728668e-06, + "loss": 0.9023, + "step": 6879 + }, + { + "epoch": 0.6617929973066564, + "grad_norm": 2.1511183961529112, + "learning_rate": 5.422938116050092e-06, + "loss": 0.9161, + "step": 6880 + }, + { + "epoch": 0.6618891881492882, + "grad_norm": 1.7801920961161888, + "learning_rate": 5.4201684046147794e-06, + "loss": 0.853, + "step": 6881 + }, + { + "epoch": 0.66198537899192, + "grad_norm": 2.2470493042590753, + "learning_rate": 5.417399137691552e-06, + "loss": 0.9704, + "step": 6882 + }, + { + "epoch": 0.6620815698345518, + "grad_norm": 2.001407969031559, + "learning_rate": 5.4146303155491964e-06, + "loss": 0.9389, + "step": 6883 + }, + { + "epoch": 0.6621777606771835, + "grad_norm": 2.5394377517361444, + "learning_rate": 5.411861938456453e-06, + "loss": 0.8923, + "step": 6884 + }, + { + "epoch": 0.6622739515198153, + "grad_norm": 1.6941766103556324, + "learning_rate": 5.409094006682009e-06, + "loss": 0.8576, + "step": 6885 + }, + { + "epoch": 0.6623701423624471, + "grad_norm": 2.035688045312571, + "learning_rate": 5.406326520494522e-06, + "loss": 0.9836, + "step": 6886 + }, + { + "epoch": 0.6624663332050789, + "grad_norm": 1.9358374004388288, + "learning_rate": 5.403559480162597e-06, + "loss": 0.9165, + "step": 6887 + }, + { + "epoch": 0.6625625240477107, + "grad_norm": 2.179624333555326, + "learning_rate": 5.400792885954802e-06, + "loss": 0.8331, + "step": 6888 + }, + { + "epoch": 0.6626587148903424, + "grad_norm": 1.9048484603678646, + "learning_rate": 5.398026738139655e-06, + "loss": 1.0023, + "step": 6889 + }, + { + "epoch": 0.6627549057329742, + "grad_norm": 2.09115579699317, + "learning_rate": 5.395261036985635e-06, + "loss": 0.8998, + "step": 6890 + }, + { + "epoch": 0.662851096575606, + "grad_norm": 2.243186435749925, + "learning_rate": 5.3924957827611825e-06, + "loss": 0.9711, + "step": 6891 + }, + { + "epoch": 0.6629472874182378, + "grad_norm": 2.2498601612488147, + "learning_rate": 5.389730975734686e-06, + "loss": 0.9077, + "step": 6892 + }, + { + "epoch": 0.6630434782608695, + "grad_norm": 2.2208494965292336, + "learning_rate": 5.386966616174489e-06, + "loss": 0.9607, + "step": 6893 + }, + { + "epoch": 0.6631396691035013, + "grad_norm": 2.0106552168800174, + "learning_rate": 5.384202704348902e-06, + "loss": 0.8273, + "step": 6894 + }, + { + "epoch": 0.6632358599461331, + "grad_norm": 1.7710110499682599, + "learning_rate": 5.381439240526187e-06, + "loss": 0.9132, + "step": 6895 + }, + { + "epoch": 0.6633320507887649, + "grad_norm": 1.8051283405315737, + "learning_rate": 5.378676224974557e-06, + "loss": 0.8059, + "step": 6896 + }, + { + "epoch": 0.6634282416313967, + "grad_norm": 1.9638037157772865, + "learning_rate": 5.375913657962191e-06, + "loss": 0.9265, + "step": 6897 + }, + { + "epoch": 0.6635244324740285, + "grad_norm": 1.9957207896857612, + "learning_rate": 5.373151539757224e-06, + "loss": 1.0007, + "step": 6898 + }, + { + "epoch": 0.6636206233166603, + "grad_norm": 1.7313914624278242, + "learning_rate": 5.37038987062773e-06, + "loss": 0.9683, + "step": 6899 + }, + { + "epoch": 0.6637168141592921, + "grad_norm": 2.0167462300351158, + "learning_rate": 5.367628650841761e-06, + "loss": 0.9761, + "step": 6900 + }, + { + "epoch": 0.6638130050019239, + "grad_norm": 1.7856844015464786, + "learning_rate": 5.36486788066732e-06, + "loss": 0.9252, + "step": 6901 + }, + { + "epoch": 0.6639091958445555, + "grad_norm": 2.219899866136461, + "learning_rate": 5.362107560372358e-06, + "loss": 0.9723, + "step": 6902 + }, + { + "epoch": 0.6640053866871873, + "grad_norm": 2.1933146754874366, + "learning_rate": 5.359347690224792e-06, + "loss": 0.8608, + "step": 6903 + }, + { + "epoch": 0.6641015775298191, + "grad_norm": 1.944759846392225, + "learning_rate": 5.356588270492487e-06, + "loss": 1.0052, + "step": 6904 + }, + { + "epoch": 0.6641977683724509, + "grad_norm": 1.92748054273133, + "learning_rate": 5.353829301443273e-06, + "loss": 0.9035, + "step": 6905 + }, + { + "epoch": 0.6642939592150827, + "grad_norm": 2.125982200413693, + "learning_rate": 5.351070783344926e-06, + "loss": 1.057, + "step": 6906 + }, + { + "epoch": 0.6643901500577145, + "grad_norm": 2.1315967431569955, + "learning_rate": 5.348312716465187e-06, + "loss": 0.8904, + "step": 6907 + }, + { + "epoch": 0.6644863409003463, + "grad_norm": 1.9411147590136197, + "learning_rate": 5.3455551010717545e-06, + "loss": 0.9436, + "step": 6908 + }, + { + "epoch": 0.6645825317429781, + "grad_norm": 2.066211843815381, + "learning_rate": 5.342797937432274e-06, + "loss": 0.8651, + "step": 6909 + }, + { + "epoch": 0.6646787225856099, + "grad_norm": 2.0298049057168837, + "learning_rate": 5.34004122581435e-06, + "loss": 0.9785, + "step": 6910 + }, + { + "epoch": 0.6647749134282416, + "grad_norm": 1.8912974957912472, + "learning_rate": 5.337284966485548e-06, + "loss": 0.8499, + "step": 6911 + }, + { + "epoch": 0.6648711042708734, + "grad_norm": 1.7715883887298176, + "learning_rate": 5.334529159713389e-06, + "loss": 0.9994, + "step": 6912 + }, + { + "epoch": 0.6649672951135052, + "grad_norm": 1.9824253834813232, + "learning_rate": 5.331773805765343e-06, + "loss": 1.0028, + "step": 6913 + }, + { + "epoch": 0.665063485956137, + "grad_norm": 1.8343356070053531, + "learning_rate": 5.329018904908841e-06, + "loss": 0.7956, + "step": 6914 + }, + { + "epoch": 0.6651596767987688, + "grad_norm": 1.9734469757154274, + "learning_rate": 5.326264457411276e-06, + "loss": 1.0746, + "step": 6915 + }, + { + "epoch": 0.6652558676414005, + "grad_norm": 1.800657959749197, + "learning_rate": 5.323510463539989e-06, + "loss": 0.7422, + "step": 6916 + }, + { + "epoch": 0.6653520584840323, + "grad_norm": 1.7708093585406979, + "learning_rate": 5.320756923562271e-06, + "loss": 0.9318, + "step": 6917 + }, + { + "epoch": 0.6654482493266641, + "grad_norm": 1.934782651876076, + "learning_rate": 5.318003837745382e-06, + "loss": 0.9621, + "step": 6918 + }, + { + "epoch": 0.6655444401692959, + "grad_norm": 2.0177899449290884, + "learning_rate": 5.315251206356539e-06, + "loss": 0.9125, + "step": 6919 + }, + { + "epoch": 0.6656406310119277, + "grad_norm": 2.292448921349169, + "learning_rate": 5.3124990296628974e-06, + "loss": 0.957, + "step": 6920 + }, + { + "epoch": 0.6657368218545594, + "grad_norm": 1.9102086717626365, + "learning_rate": 5.309747307931589e-06, + "loss": 0.954, + "step": 6921 + }, + { + "epoch": 0.6658330126971912, + "grad_norm": 2.1063437816807618, + "learning_rate": 5.306996041429688e-06, + "loss": 0.8753, + "step": 6922 + }, + { + "epoch": 0.665929203539823, + "grad_norm": 2.418526952422416, + "learning_rate": 5.304245230424223e-06, + "loss": 0.942, + "step": 6923 + }, + { + "epoch": 0.6660253943824548, + "grad_norm": 1.921865260397202, + "learning_rate": 5.301494875182192e-06, + "loss": 0.957, + "step": 6924 + }, + { + "epoch": 0.6661215852250866, + "grad_norm": 1.7909665470049512, + "learning_rate": 5.298744975970535e-06, + "loss": 0.86, + "step": 6925 + }, + { + "epoch": 0.6662177760677184, + "grad_norm": 2.002365301876595, + "learning_rate": 5.295995533056162e-06, + "loss": 0.9025, + "step": 6926 + }, + { + "epoch": 0.6663139669103502, + "grad_norm": 1.929373423262317, + "learning_rate": 5.293246546705925e-06, + "loss": 1.0833, + "step": 6927 + }, + { + "epoch": 0.666410157752982, + "grad_norm": 1.9125281030636725, + "learning_rate": 5.290498017186631e-06, + "loss": 0.9788, + "step": 6928 + }, + { + "epoch": 0.6665063485956138, + "grad_norm": 1.6616657215261306, + "learning_rate": 5.287749944765059e-06, + "loss": 0.8587, + "step": 6929 + }, + { + "epoch": 0.6666025394382454, + "grad_norm": 2.2045213579150142, + "learning_rate": 5.2850023297079235e-06, + "loss": 0.9284, + "step": 6930 + }, + { + "epoch": 0.6666987302808772, + "grad_norm": 2.363905543820861, + "learning_rate": 5.2822551722819085e-06, + "loss": 0.9509, + "step": 6931 + }, + { + "epoch": 0.666794921123509, + "grad_norm": 1.8577139915472063, + "learning_rate": 5.279508472753654e-06, + "loss": 0.9725, + "step": 6932 + }, + { + "epoch": 0.6668911119661408, + "grad_norm": 1.8449571859528626, + "learning_rate": 5.276762231389745e-06, + "loss": 0.913, + "step": 6933 + }, + { + "epoch": 0.6669873028087726, + "grad_norm": 2.2091602482881214, + "learning_rate": 5.274016448456725e-06, + "loss": 0.9245, + "step": 6934 + }, + { + "epoch": 0.6670834936514044, + "grad_norm": 1.9931727208817294, + "learning_rate": 5.2712711242211e-06, + "loss": 0.9781, + "step": 6935 + }, + { + "epoch": 0.6671796844940362, + "grad_norm": 1.8814897355497815, + "learning_rate": 5.2685262589493314e-06, + "loss": 0.9822, + "step": 6936 + }, + { + "epoch": 0.667275875336668, + "grad_norm": 1.966383157564007, + "learning_rate": 5.265781852907824e-06, + "loss": 0.9308, + "step": 6937 + }, + { + "epoch": 0.6673720661792998, + "grad_norm": 2.0429215873140887, + "learning_rate": 5.263037906362953e-06, + "loss": 0.8518, + "step": 6938 + }, + { + "epoch": 0.6674682570219315, + "grad_norm": 1.9137298268269674, + "learning_rate": 5.260294419581037e-06, + "loss": 0.8632, + "step": 6939 + }, + { + "epoch": 0.6675644478645633, + "grad_norm": 2.13879841649124, + "learning_rate": 5.257551392828359e-06, + "loss": 1.0468, + "step": 6940 + }, + { + "epoch": 0.667660638707195, + "grad_norm": 2.170530977832475, + "learning_rate": 5.254808826371149e-06, + "loss": 0.9236, + "step": 6941 + }, + { + "epoch": 0.6677568295498268, + "grad_norm": 2.0121883409575365, + "learning_rate": 5.252066720475597e-06, + "loss": 0.9208, + "step": 6942 + }, + { + "epoch": 0.6678530203924586, + "grad_norm": 2.053023599791058, + "learning_rate": 5.249325075407856e-06, + "loss": 0.9414, + "step": 6943 + }, + { + "epoch": 0.6679492112350904, + "grad_norm": 2.1731390854858694, + "learning_rate": 5.246583891434018e-06, + "loss": 0.8963, + "step": 6944 + }, + { + "epoch": 0.6680454020777222, + "grad_norm": 1.8269717412805326, + "learning_rate": 5.24384316882014e-06, + "loss": 0.9395, + "step": 6945 + }, + { + "epoch": 0.668141592920354, + "grad_norm": 1.7747793756416208, + "learning_rate": 5.241102907832232e-06, + "loss": 0.8853, + "step": 6946 + }, + { + "epoch": 0.6682377837629858, + "grad_norm": 2.410815454212121, + "learning_rate": 5.238363108736264e-06, + "loss": 1.0061, + "step": 6947 + }, + { + "epoch": 0.6683339746056175, + "grad_norm": 2.0139506874227977, + "learning_rate": 5.235623771798151e-06, + "loss": 0.8181, + "step": 6948 + }, + { + "epoch": 0.6684301654482493, + "grad_norm": 1.9904292351755344, + "learning_rate": 5.232884897283774e-06, + "loss": 0.9628, + "step": 6949 + }, + { + "epoch": 0.6685263562908811, + "grad_norm": 2.045333771817236, + "learning_rate": 5.23014648545897e-06, + "loss": 0.9039, + "step": 6950 + }, + { + "epoch": 0.6686225471335129, + "grad_norm": 1.9184292328391135, + "learning_rate": 5.227408536589511e-06, + "loss": 0.9302, + "step": 6951 + }, + { + "epoch": 0.6687187379761447, + "grad_norm": 2.57093888515211, + "learning_rate": 5.224671050941146e-06, + "loss": 0.9635, + "step": 6952 + }, + { + "epoch": 0.6688149288187765, + "grad_norm": 2.0221040374268218, + "learning_rate": 5.221934028779575e-06, + "loss": 0.8363, + "step": 6953 + }, + { + "epoch": 0.6689111196614083, + "grad_norm": 1.781847432099931, + "learning_rate": 5.2191974703704425e-06, + "loss": 0.933, + "step": 6954 + }, + { + "epoch": 0.66900731050404, + "grad_norm": 1.9594408380662587, + "learning_rate": 5.2164613759793625e-06, + "loss": 1.0421, + "step": 6955 + }, + { + "epoch": 0.6691035013466718, + "grad_norm": 2.0967956183390353, + "learning_rate": 5.213725745871889e-06, + "loss": 0.9971, + "step": 6956 + }, + { + "epoch": 0.6691996921893035, + "grad_norm": 1.9528044366526291, + "learning_rate": 5.210990580313548e-06, + "loss": 0.9756, + "step": 6957 + }, + { + "epoch": 0.6692958830319353, + "grad_norm": 1.977904701666624, + "learning_rate": 5.208255879569799e-06, + "loss": 0.9113, + "step": 6958 + }, + { + "epoch": 0.6693920738745671, + "grad_norm": 2.4659319694158266, + "learning_rate": 5.205521643906075e-06, + "loss": 0.9967, + "step": 6959 + }, + { + "epoch": 0.6694882647171989, + "grad_norm": 1.9755945651508544, + "learning_rate": 5.20278787358776e-06, + "loss": 1.0029, + "step": 6960 + }, + { + "epoch": 0.6695844555598307, + "grad_norm": 1.9898850352671955, + "learning_rate": 5.200054568880187e-06, + "loss": 0.9231, + "step": 6961 + }, + { + "epoch": 0.6696806464024625, + "grad_norm": 1.863634596008416, + "learning_rate": 5.197321730048641e-06, + "loss": 0.9326, + "step": 6962 + }, + { + "epoch": 0.6697768372450943, + "grad_norm": 2.058544308813162, + "learning_rate": 5.1945893573583725e-06, + "loss": 0.8819, + "step": 6963 + }, + { + "epoch": 0.6698730280877261, + "grad_norm": 2.4827593062777074, + "learning_rate": 5.1918574510745865e-06, + "loss": 0.8971, + "step": 6964 + }, + { + "epoch": 0.6699692189303579, + "grad_norm": 1.8361921104563699, + "learning_rate": 5.189126011462431e-06, + "loss": 0.9336, + "step": 6965 + }, + { + "epoch": 0.6700654097729896, + "grad_norm": 1.9501125696357962, + "learning_rate": 5.186395038787017e-06, + "loss": 0.9189, + "step": 6966 + }, + { + "epoch": 0.6701616006156214, + "grad_norm": 2.48052235340551, + "learning_rate": 5.183664533313414e-06, + "loss": 0.9383, + "step": 6967 + }, + { + "epoch": 0.6702577914582531, + "grad_norm": 2.186611488047608, + "learning_rate": 5.180934495306638e-06, + "loss": 0.8581, + "step": 6968 + }, + { + "epoch": 0.6703539823008849, + "grad_norm": 1.6741722034265056, + "learning_rate": 5.178204925031657e-06, + "loss": 0.7848, + "step": 6969 + }, + { + "epoch": 0.6704501731435167, + "grad_norm": 2.1279326328453685, + "learning_rate": 5.175475822753404e-06, + "loss": 0.9709, + "step": 6970 + }, + { + "epoch": 0.6705463639861485, + "grad_norm": 2.1255734237420136, + "learning_rate": 5.172747188736766e-06, + "loss": 0.95, + "step": 6971 + }, + { + "epoch": 0.6706425548287803, + "grad_norm": 2.0806329516731736, + "learning_rate": 5.170019023246574e-06, + "loss": 1.0605, + "step": 6972 + }, + { + "epoch": 0.6707387456714121, + "grad_norm": 2.0453086635794873, + "learning_rate": 5.167291326547625e-06, + "loss": 0.915, + "step": 6973 + }, + { + "epoch": 0.6708349365140439, + "grad_norm": 1.9848996711486318, + "learning_rate": 5.16456409890466e-06, + "loss": 0.8975, + "step": 6974 + }, + { + "epoch": 0.6709311273566756, + "grad_norm": 1.9168628150213907, + "learning_rate": 5.161837340582387e-06, + "loss": 0.8776, + "step": 6975 + }, + { + "epoch": 0.6710273181993074, + "grad_norm": 1.9320683611147114, + "learning_rate": 5.159111051845451e-06, + "loss": 0.8708, + "step": 6976 + }, + { + "epoch": 0.6711235090419392, + "grad_norm": 1.9393018913735274, + "learning_rate": 5.1563852329584695e-06, + "loss": 0.9705, + "step": 6977 + }, + { + "epoch": 0.671219699884571, + "grad_norm": 2.06960508896212, + "learning_rate": 5.153659884186013e-06, + "loss": 0.9056, + "step": 6978 + }, + { + "epoch": 0.6713158907272028, + "grad_norm": 2.115403858593565, + "learning_rate": 5.150935005792586e-06, + "loss": 1.0067, + "step": 6979 + }, + { + "epoch": 0.6714120815698346, + "grad_norm": 1.831976149594577, + "learning_rate": 5.148210598042665e-06, + "loss": 0.8139, + "step": 6980 + }, + { + "epoch": 0.6715082724124664, + "grad_norm": 2.042720200332244, + "learning_rate": 5.145486661200685e-06, + "loss": 0.9453, + "step": 6981 + }, + { + "epoch": 0.6716044632550981, + "grad_norm": 1.903719095430354, + "learning_rate": 5.142763195531017e-06, + "loss": 0.8228, + "step": 6982 + }, + { + "epoch": 0.6717006540977299, + "grad_norm": 2.246396978557996, + "learning_rate": 5.140040201298004e-06, + "loss": 0.8198, + "step": 6983 + }, + { + "epoch": 0.6717968449403616, + "grad_norm": 1.9903932834595337, + "learning_rate": 5.137317678765939e-06, + "loss": 0.9415, + "step": 6984 + }, + { + "epoch": 0.6718930357829934, + "grad_norm": 2.0758048414153345, + "learning_rate": 5.134595628199059e-06, + "loss": 1.006, + "step": 6985 + }, + { + "epoch": 0.6719892266256252, + "grad_norm": 2.018152973836983, + "learning_rate": 5.131874049861563e-06, + "loss": 0.996, + "step": 6986 + }, + { + "epoch": 0.672085417468257, + "grad_norm": 2.0133559021382594, + "learning_rate": 5.129152944017604e-06, + "loss": 0.9428, + "step": 6987 + }, + { + "epoch": 0.6721816083108888, + "grad_norm": 2.0547707114923925, + "learning_rate": 5.126432310931295e-06, + "loss": 0.9032, + "step": 6988 + }, + { + "epoch": 0.6722777991535206, + "grad_norm": 1.7672526343033854, + "learning_rate": 5.123712150866688e-06, + "loss": 0.7665, + "step": 6989 + }, + { + "epoch": 0.6723739899961524, + "grad_norm": 2.2007115354045585, + "learning_rate": 5.120992464087807e-06, + "loss": 0.9859, + "step": 6990 + }, + { + "epoch": 0.6724701808387842, + "grad_norm": 2.0110824153007734, + "learning_rate": 5.118273250858612e-06, + "loss": 0.876, + "step": 6991 + }, + { + "epoch": 0.672566371681416, + "grad_norm": 2.0799974940015047, + "learning_rate": 5.115554511443033e-06, + "loss": 0.9483, + "step": 6992 + }, + { + "epoch": 0.6726625625240477, + "grad_norm": 2.3443876473471175, + "learning_rate": 5.1128362461049406e-06, + "loss": 1.0103, + "step": 6993 + }, + { + "epoch": 0.6727587533666795, + "grad_norm": 2.06148387655934, + "learning_rate": 5.1101184551081705e-06, + "loss": 0.9304, + "step": 6994 + }, + { + "epoch": 0.6728549442093112, + "grad_norm": 2.1196132625690387, + "learning_rate": 5.10740113871651e-06, + "loss": 0.9204, + "step": 6995 + }, + { + "epoch": 0.672951135051943, + "grad_norm": 1.8463391219844274, + "learning_rate": 5.104684297193694e-06, + "loss": 0.9138, + "step": 6996 + }, + { + "epoch": 0.6730473258945748, + "grad_norm": 1.9511541698556, + "learning_rate": 5.101967930803413e-06, + "loss": 0.9791, + "step": 6997 + }, + { + "epoch": 0.6731435167372066, + "grad_norm": 2.129287689511349, + "learning_rate": 5.099252039809317e-06, + "loss": 0.9954, + "step": 6998 + }, + { + "epoch": 0.6732397075798384, + "grad_norm": 2.0219301374786682, + "learning_rate": 5.09653662447501e-06, + "loss": 0.8534, + "step": 6999 + }, + { + "epoch": 0.6733358984224702, + "grad_norm": 2.0297852581931544, + "learning_rate": 5.09382168506404e-06, + "loss": 0.9707, + "step": 7000 + }, + { + "epoch": 0.673432089265102, + "grad_norm": 1.9450352297032774, + "learning_rate": 5.091107221839918e-06, + "loss": 1.0632, + "step": 7001 + }, + { + "epoch": 0.6735282801077337, + "grad_norm": 1.6940650896728593, + "learning_rate": 5.088393235066114e-06, + "loss": 0.7826, + "step": 7002 + }, + { + "epoch": 0.6736244709503655, + "grad_norm": 2.231580049952601, + "learning_rate": 5.085679725006028e-06, + "loss": 0.917, + "step": 7003 + }, + { + "epoch": 0.6737206617929973, + "grad_norm": 1.9813749703983736, + "learning_rate": 5.082966691923037e-06, + "loss": 0.8762, + "step": 7004 + }, + { + "epoch": 0.6738168526356291, + "grad_norm": 2.0531755453224587, + "learning_rate": 5.080254136080465e-06, + "loss": 0.9255, + "step": 7005 + }, + { + "epoch": 0.6739130434782609, + "grad_norm": 2.216963088690934, + "learning_rate": 5.077542057741592e-06, + "loss": 0.9604, + "step": 7006 + }, + { + "epoch": 0.6740092343208927, + "grad_norm": 1.8341737839499292, + "learning_rate": 5.074830457169646e-06, + "loss": 0.9503, + "step": 7007 + }, + { + "epoch": 0.6741054251635245, + "grad_norm": 1.5911155421139331, + "learning_rate": 5.0721193346278066e-06, + "loss": 0.7505, + "step": 7008 + }, + { + "epoch": 0.6742016160061562, + "grad_norm": 2.1149130153448166, + "learning_rate": 5.0694086903792184e-06, + "loss": 0.9585, + "step": 7009 + }, + { + "epoch": 0.674297806848788, + "grad_norm": 2.1929722742526594, + "learning_rate": 5.066698524686966e-06, + "loss": 0.9828, + "step": 7010 + }, + { + "epoch": 0.6743939976914197, + "grad_norm": 2.0330000178534484, + "learning_rate": 5.063988837814099e-06, + "loss": 0.9542, + "step": 7011 + }, + { + "epoch": 0.6744901885340515, + "grad_norm": 2.0476088022279813, + "learning_rate": 5.061279630023618e-06, + "loss": 0.9704, + "step": 7012 + }, + { + "epoch": 0.6745863793766833, + "grad_norm": 1.9428867726893602, + "learning_rate": 5.058570901578472e-06, + "loss": 0.8684, + "step": 7013 + }, + { + "epoch": 0.6746825702193151, + "grad_norm": 2.2668767573829727, + "learning_rate": 5.055862652741562e-06, + "loss": 0.9068, + "step": 7014 + }, + { + "epoch": 0.6747787610619469, + "grad_norm": 2.107789208239503, + "learning_rate": 5.053154883775752e-06, + "loss": 0.9941, + "step": 7015 + }, + { + "epoch": 0.6748749519045787, + "grad_norm": 2.040263944578326, + "learning_rate": 5.050447594943856e-06, + "loss": 0.8898, + "step": 7016 + }, + { + "epoch": 0.6749711427472105, + "grad_norm": 2.20552872236722, + "learning_rate": 5.047740786508635e-06, + "loss": 0.9809, + "step": 7017 + }, + { + "epoch": 0.6750673335898423, + "grad_norm": 2.11712275812853, + "learning_rate": 5.045034458732808e-06, + "loss": 0.9189, + "step": 7018 + }, + { + "epoch": 0.6751635244324741, + "grad_norm": 1.6381756874940103, + "learning_rate": 5.042328611879055e-06, + "loss": 0.8755, + "step": 7019 + }, + { + "epoch": 0.6752597152751059, + "grad_norm": 2.0120884838049586, + "learning_rate": 5.0396232462099945e-06, + "loss": 0.8632, + "step": 7020 + }, + { + "epoch": 0.6753559061177375, + "grad_norm": 1.980046920012394, + "learning_rate": 5.036918361988205e-06, + "loss": 0.9044, + "step": 7021 + }, + { + "epoch": 0.6754520969603693, + "grad_norm": 1.9751754810503133, + "learning_rate": 5.034213959476222e-06, + "loss": 1.0331, + "step": 7022 + }, + { + "epoch": 0.6755482878030011, + "grad_norm": 2.384305951310544, + "learning_rate": 5.031510038936532e-06, + "loss": 0.9212, + "step": 7023 + }, + { + "epoch": 0.6756444786456329, + "grad_norm": 2.939673828414853, + "learning_rate": 5.028806600631569e-06, + "loss": 1.0161, + "step": 7024 + }, + { + "epoch": 0.6757406694882647, + "grad_norm": 1.882455583534238, + "learning_rate": 5.026103644823732e-06, + "loss": 1.0019, + "step": 7025 + }, + { + "epoch": 0.6758368603308965, + "grad_norm": 1.8848912035287464, + "learning_rate": 5.023401171775357e-06, + "loss": 0.9424, + "step": 7026 + }, + { + "epoch": 0.6759330511735283, + "grad_norm": 1.9482576761890569, + "learning_rate": 5.020699181748754e-06, + "loss": 0.8736, + "step": 7027 + }, + { + "epoch": 0.6760292420161601, + "grad_norm": 1.9900601769444584, + "learning_rate": 5.017997675006161e-06, + "loss": 0.9233, + "step": 7028 + }, + { + "epoch": 0.6761254328587919, + "grad_norm": 1.8777068190923918, + "learning_rate": 5.0152966518097905e-06, + "loss": 0.9082, + "step": 7029 + }, + { + "epoch": 0.6762216237014236, + "grad_norm": 2.0310111286919827, + "learning_rate": 5.012596112421806e-06, + "loss": 0.9302, + "step": 7030 + }, + { + "epoch": 0.6763178145440554, + "grad_norm": 2.164490131383824, + "learning_rate": 5.009896057104303e-06, + "loss": 0.9159, + "step": 7031 + }, + { + "epoch": 0.6764140053866872, + "grad_norm": 1.8261852075189184, + "learning_rate": 5.007196486119355e-06, + "loss": 0.909, + "step": 7032 + }, + { + "epoch": 0.676510196229319, + "grad_norm": 1.840574920623825, + "learning_rate": 5.004497399728978e-06, + "loss": 0.8445, + "step": 7033 + }, + { + "epoch": 0.6766063870719508, + "grad_norm": 1.8030712951442784, + "learning_rate": 5.001798798195136e-06, + "loss": 0.8177, + "step": 7034 + }, + { + "epoch": 0.6767025779145825, + "grad_norm": 1.995004524283674, + "learning_rate": 4.999100681779757e-06, + "loss": 0.901, + "step": 7035 + }, + { + "epoch": 0.6767987687572143, + "grad_norm": 2.0694092958101713, + "learning_rate": 4.996403050744719e-06, + "loss": 0.9395, + "step": 7036 + }, + { + "epoch": 0.6768949595998461, + "grad_norm": 2.5551869786209775, + "learning_rate": 4.993705905351846e-06, + "loss": 0.9107, + "step": 7037 + }, + { + "epoch": 0.6769911504424779, + "grad_norm": 1.873227447243851, + "learning_rate": 4.991009245862917e-06, + "loss": 1.0007, + "step": 7038 + }, + { + "epoch": 0.6770873412851096, + "grad_norm": 2.1363466996778944, + "learning_rate": 4.988313072539667e-06, + "loss": 1.0544, + "step": 7039 + }, + { + "epoch": 0.6771835321277414, + "grad_norm": 1.9749548444120435, + "learning_rate": 4.985617385643789e-06, + "loss": 0.983, + "step": 7040 + }, + { + "epoch": 0.6772797229703732, + "grad_norm": 1.9338152377997755, + "learning_rate": 4.9829221854369145e-06, + "loss": 0.9162, + "step": 7041 + }, + { + "epoch": 0.677375913813005, + "grad_norm": 1.8589499886739842, + "learning_rate": 4.980227472180643e-06, + "loss": 1.0049, + "step": 7042 + }, + { + "epoch": 0.6774721046556368, + "grad_norm": 1.8549058031351011, + "learning_rate": 4.977533246136512e-06, + "loss": 0.913, + "step": 7043 + }, + { + "epoch": 0.6775682954982686, + "grad_norm": 2.0895054034293343, + "learning_rate": 4.974839507566027e-06, + "loss": 0.9478, + "step": 7044 + }, + { + "epoch": 0.6776644863409004, + "grad_norm": 2.224609773923293, + "learning_rate": 4.972146256730632e-06, + "loss": 0.9701, + "step": 7045 + }, + { + "epoch": 0.6777606771835322, + "grad_norm": 2.1809623478241384, + "learning_rate": 4.969453493891733e-06, + "loss": 1.104, + "step": 7046 + }, + { + "epoch": 0.677856868026164, + "grad_norm": 1.7751047650985263, + "learning_rate": 4.96676121931069e-06, + "loss": 0.9612, + "step": 7047 + }, + { + "epoch": 0.6779530588687956, + "grad_norm": 2.0073092069959744, + "learning_rate": 4.9640694332488075e-06, + "loss": 1.0644, + "step": 7048 + }, + { + "epoch": 0.6780492497114274, + "grad_norm": 2.140324063454761, + "learning_rate": 4.961378135967341e-06, + "loss": 0.8987, + "step": 7049 + }, + { + "epoch": 0.6781454405540592, + "grad_norm": 2.0264319436068114, + "learning_rate": 4.958687327727511e-06, + "loss": 0.8084, + "step": 7050 + }, + { + "epoch": 0.678241631396691, + "grad_norm": 2.3031704728524063, + "learning_rate": 4.955997008790485e-06, + "loss": 0.9604, + "step": 7051 + }, + { + "epoch": 0.6783378222393228, + "grad_norm": 1.8160170989576139, + "learning_rate": 4.953307179417376e-06, + "loss": 0.8517, + "step": 7052 + }, + { + "epoch": 0.6784340130819546, + "grad_norm": 1.8656687726178702, + "learning_rate": 4.9506178398692605e-06, + "loss": 0.9446, + "step": 7053 + }, + { + "epoch": 0.6785302039245864, + "grad_norm": 1.8638304903908334, + "learning_rate": 4.947928990407156e-06, + "loss": 0.9574, + "step": 7054 + }, + { + "epoch": 0.6786263947672182, + "grad_norm": 2.4296665702241835, + "learning_rate": 4.945240631292045e-06, + "loss": 0.9238, + "step": 7055 + }, + { + "epoch": 0.67872258560985, + "grad_norm": 2.0651475177260354, + "learning_rate": 4.94255276278485e-06, + "loss": 1.0419, + "step": 7056 + }, + { + "epoch": 0.6788187764524817, + "grad_norm": 2.272900381030875, + "learning_rate": 4.939865385146455e-06, + "loss": 1.0387, + "step": 7057 + }, + { + "epoch": 0.6789149672951135, + "grad_norm": 2.115189976943744, + "learning_rate": 4.937178498637696e-06, + "loss": 0.925, + "step": 7058 + }, + { + "epoch": 0.6790111581377453, + "grad_norm": 2.1415002704876422, + "learning_rate": 4.934492103519354e-06, + "loss": 0.9806, + "step": 7059 + }, + { + "epoch": 0.6791073489803771, + "grad_norm": 2.2500981249827294, + "learning_rate": 4.931806200052165e-06, + "loss": 0.847, + "step": 7060 + }, + { + "epoch": 0.6792035398230089, + "grad_norm": 1.9038405156036797, + "learning_rate": 4.929120788496826e-06, + "loss": 0.9616, + "step": 7061 + }, + { + "epoch": 0.6792997306656406, + "grad_norm": 2.315654613257656, + "learning_rate": 4.926435869113971e-06, + "loss": 0.9613, + "step": 7062 + }, + { + "epoch": 0.6793959215082724, + "grad_norm": 2.000519870416938, + "learning_rate": 4.923751442164198e-06, + "loss": 0.9093, + "step": 7063 + }, + { + "epoch": 0.6794921123509042, + "grad_norm": 2.529389911609761, + "learning_rate": 4.92106750790806e-06, + "loss": 1.0575, + "step": 7064 + }, + { + "epoch": 0.679588303193536, + "grad_norm": 2.0504416956792966, + "learning_rate": 4.91838406660605e-06, + "loss": 0.965, + "step": 7065 + }, + { + "epoch": 0.6796844940361677, + "grad_norm": 2.0832584983287887, + "learning_rate": 4.915701118518616e-06, + "loss": 0.8897, + "step": 7066 + }, + { + "epoch": 0.6797806848787995, + "grad_norm": 1.7399344118624118, + "learning_rate": 4.913018663906165e-06, + "loss": 0.8163, + "step": 7067 + }, + { + "epoch": 0.6798768757214313, + "grad_norm": 1.983314622934363, + "learning_rate": 4.910336703029055e-06, + "loss": 0.9349, + "step": 7068 + }, + { + "epoch": 0.6799730665640631, + "grad_norm": 1.8484546335805148, + "learning_rate": 4.907655236147587e-06, + "loss": 0.8794, + "step": 7069 + }, + { + "epoch": 0.6800692574066949, + "grad_norm": 2.086654947573459, + "learning_rate": 4.904974263522025e-06, + "loss": 1.0322, + "step": 7070 + }, + { + "epoch": 0.6801654482493267, + "grad_norm": 1.997926550624945, + "learning_rate": 4.9022937854125816e-06, + "loss": 0.8864, + "step": 7071 + }, + { + "epoch": 0.6802616390919585, + "grad_norm": 3.2316852100289335, + "learning_rate": 4.899613802079419e-06, + "loss": 1.0159, + "step": 7072 + }, + { + "epoch": 0.6803578299345903, + "grad_norm": 1.9289195475399399, + "learning_rate": 4.896934313782647e-06, + "loss": 0.9332, + "step": 7073 + }, + { + "epoch": 0.6804540207772221, + "grad_norm": 2.1166550152756223, + "learning_rate": 4.8942553207823395e-06, + "loss": 1.0758, + "step": 7074 + }, + { + "epoch": 0.6805502116198537, + "grad_norm": 1.8065335774038132, + "learning_rate": 4.891576823338518e-06, + "loss": 0.842, + "step": 7075 + }, + { + "epoch": 0.6806464024624855, + "grad_norm": 2.2266674014971763, + "learning_rate": 4.888898821711144e-06, + "loss": 0.9709, + "step": 7076 + }, + { + "epoch": 0.6807425933051173, + "grad_norm": 2.0533119210645965, + "learning_rate": 4.886221316160152e-06, + "loss": 0.8951, + "step": 7077 + }, + { + "epoch": 0.6808387841477491, + "grad_norm": 2.088482793290069, + "learning_rate": 4.883544306945407e-06, + "loss": 0.9369, + "step": 7078 + }, + { + "epoch": 0.6809349749903809, + "grad_norm": 1.647059926752232, + "learning_rate": 4.880867794326744e-06, + "loss": 0.7488, + "step": 7079 + }, + { + "epoch": 0.6810311658330127, + "grad_norm": 2.2173706046824395, + "learning_rate": 4.878191778563934e-06, + "loss": 0.9365, + "step": 7080 + }, + { + "epoch": 0.6811273566756445, + "grad_norm": 1.4773525453958396, + "learning_rate": 4.8755162599167115e-06, + "loss": 0.7446, + "step": 7081 + }, + { + "epoch": 0.6812235475182763, + "grad_norm": 2.1762485915640593, + "learning_rate": 4.872841238644766e-06, + "loss": 0.9385, + "step": 7082 + }, + { + "epoch": 0.6813197383609081, + "grad_norm": 2.3284992047029798, + "learning_rate": 4.870166715007716e-06, + "loss": 1.01, + "step": 7083 + }, + { + "epoch": 0.6814159292035398, + "grad_norm": 1.8944009437777434, + "learning_rate": 4.867492689265154e-06, + "loss": 0.953, + "step": 7084 + }, + { + "epoch": 0.6815121200461716, + "grad_norm": 1.9048012724412502, + "learning_rate": 4.8648191616766226e-06, + "loss": 1.0238, + "step": 7085 + }, + { + "epoch": 0.6816083108888034, + "grad_norm": 1.932972846007164, + "learning_rate": 4.8621461325016015e-06, + "loss": 0.9401, + "step": 7086 + }, + { + "epoch": 0.6817045017314352, + "grad_norm": 1.9987531937028038, + "learning_rate": 4.859473601999536e-06, + "loss": 0.967, + "step": 7087 + }, + { + "epoch": 0.681800692574067, + "grad_norm": 2.0828124232129293, + "learning_rate": 4.856801570429822e-06, + "loss": 0.8666, + "step": 7088 + }, + { + "epoch": 0.6818968834166987, + "grad_norm": 1.941402126472065, + "learning_rate": 4.8541300380518e-06, + "loss": 0.906, + "step": 7089 + }, + { + "epoch": 0.6819930742593305, + "grad_norm": 1.9535601279465657, + "learning_rate": 4.851459005124759e-06, + "loss": 0.9698, + "step": 7090 + }, + { + "epoch": 0.6820892651019623, + "grad_norm": 1.9820365536047977, + "learning_rate": 4.848788471907952e-06, + "loss": 0.8981, + "step": 7091 + }, + { + "epoch": 0.6821854559445941, + "grad_norm": 1.9379358046058006, + "learning_rate": 4.846118438660578e-06, + "loss": 1.0402, + "step": 7092 + }, + { + "epoch": 0.6822816467872258, + "grad_norm": 2.1259573026432217, + "learning_rate": 4.843448905641783e-06, + "loss": 0.9734, + "step": 7093 + }, + { + "epoch": 0.6823778376298576, + "grad_norm": 1.8860098408483867, + "learning_rate": 4.840779873110675e-06, + "loss": 0.8578, + "step": 7094 + }, + { + "epoch": 0.6824740284724894, + "grad_norm": 2.227133005195165, + "learning_rate": 4.8381113413262956e-06, + "loss": 0.9775, + "step": 7095 + }, + { + "epoch": 0.6825702193151212, + "grad_norm": 2.020039073207391, + "learning_rate": 4.83544331054766e-06, + "loss": 0.8901, + "step": 7096 + }, + { + "epoch": 0.682666410157753, + "grad_norm": 2.868502555651228, + "learning_rate": 4.832775781033715e-06, + "loss": 0.8037, + "step": 7097 + }, + { + "epoch": 0.6827626010003848, + "grad_norm": 1.7229871529714451, + "learning_rate": 4.83010875304337e-06, + "loss": 0.812, + "step": 7098 + }, + { + "epoch": 0.6828587918430166, + "grad_norm": 1.7761399605927028, + "learning_rate": 4.827442226835488e-06, + "loss": 0.9634, + "step": 7099 + }, + { + "epoch": 0.6829549826856484, + "grad_norm": 2.0094641411463017, + "learning_rate": 4.824776202668875e-06, + "loss": 1.0309, + "step": 7100 + }, + { + "epoch": 0.6830511735282802, + "grad_norm": 2.3464661922395247, + "learning_rate": 4.822110680802286e-06, + "loss": 0.936, + "step": 7101 + }, + { + "epoch": 0.6831473643709118, + "grad_norm": 2.0425848743703168, + "learning_rate": 4.819445661494437e-06, + "loss": 1.0674, + "step": 7102 + }, + { + "epoch": 0.6832435552135436, + "grad_norm": 1.940808057204083, + "learning_rate": 4.816781145003997e-06, + "loss": 0.9525, + "step": 7103 + }, + { + "epoch": 0.6833397460561754, + "grad_norm": 1.7215621764833187, + "learning_rate": 4.8141171315895694e-06, + "loss": 0.9423, + "step": 7104 + }, + { + "epoch": 0.6834359368988072, + "grad_norm": 1.7718470779934286, + "learning_rate": 4.811453621509731e-06, + "loss": 0.8311, + "step": 7105 + }, + { + "epoch": 0.683532127741439, + "grad_norm": 1.9485550870991613, + "learning_rate": 4.808790615022987e-06, + "loss": 0.9596, + "step": 7106 + }, + { + "epoch": 0.6836283185840708, + "grad_norm": 2.564446959622738, + "learning_rate": 4.806128112387816e-06, + "loss": 1.0252, + "step": 7107 + }, + { + "epoch": 0.6837245094267026, + "grad_norm": 1.7514664406225358, + "learning_rate": 4.803466113862626e-06, + "loss": 0.795, + "step": 7108 + }, + { + "epoch": 0.6838207002693344, + "grad_norm": 1.8605772220745227, + "learning_rate": 4.800804619705793e-06, + "loss": 0.8961, + "step": 7109 + }, + { + "epoch": 0.6839168911119662, + "grad_norm": 1.7837042195599557, + "learning_rate": 4.798143630175642e-06, + "loss": 0.8398, + "step": 7110 + }, + { + "epoch": 0.6840130819545979, + "grad_norm": 1.9755173444574337, + "learning_rate": 4.795483145530439e-06, + "loss": 0.8738, + "step": 7111 + }, + { + "epoch": 0.6841092727972297, + "grad_norm": 1.849667194375746, + "learning_rate": 4.792823166028405e-06, + "loss": 0.9127, + "step": 7112 + }, + { + "epoch": 0.6842054636398615, + "grad_norm": 2.009212145530881, + "learning_rate": 4.790163691927721e-06, + "loss": 0.9698, + "step": 7113 + }, + { + "epoch": 0.6843016544824932, + "grad_norm": 2.113208071642309, + "learning_rate": 4.787504723486505e-06, + "loss": 0.9502, + "step": 7114 + }, + { + "epoch": 0.684397845325125, + "grad_norm": 2.1791814350539975, + "learning_rate": 4.784846260962834e-06, + "loss": 0.9812, + "step": 7115 + }, + { + "epoch": 0.6844940361677568, + "grad_norm": 2.062685231080133, + "learning_rate": 4.7821883046147414e-06, + "loss": 0.9359, + "step": 7116 + }, + { + "epoch": 0.6845902270103886, + "grad_norm": 1.8697270451142212, + "learning_rate": 4.779530854700198e-06, + "loss": 0.9428, + "step": 7117 + }, + { + "epoch": 0.6846864178530204, + "grad_norm": 2.030065800680439, + "learning_rate": 4.776873911477133e-06, + "loss": 0.8171, + "step": 7118 + }, + { + "epoch": 0.6847826086956522, + "grad_norm": 1.8411083840751674, + "learning_rate": 4.774217475203424e-06, + "loss": 0.7848, + "step": 7119 + }, + { + "epoch": 0.6848787995382839, + "grad_norm": 1.957278034336153, + "learning_rate": 4.771561546136908e-06, + "loss": 0.9906, + "step": 7120 + }, + { + "epoch": 0.6849749903809157, + "grad_norm": 2.0720046224885666, + "learning_rate": 4.768906124535359e-06, + "loss": 1.0186, + "step": 7121 + }, + { + "epoch": 0.6850711812235475, + "grad_norm": 1.8591858202728981, + "learning_rate": 4.766251210656509e-06, + "loss": 0.9403, + "step": 7122 + }, + { + "epoch": 0.6851673720661793, + "grad_norm": 1.9055969463333662, + "learning_rate": 4.763596804758047e-06, + "loss": 0.8666, + "step": 7123 + }, + { + "epoch": 0.6852635629088111, + "grad_norm": 1.7027318809960805, + "learning_rate": 4.760942907097601e-06, + "loss": 0.8085, + "step": 7124 + }, + { + "epoch": 0.6853597537514429, + "grad_norm": 2.1517975323361807, + "learning_rate": 4.758289517932751e-06, + "loss": 0.885, + "step": 7125 + }, + { + "epoch": 0.6854559445940747, + "grad_norm": 1.8957743773612716, + "learning_rate": 4.755636637521035e-06, + "loss": 0.9961, + "step": 7126 + }, + { + "epoch": 0.6855521354367065, + "grad_norm": 2.129021030539592, + "learning_rate": 4.752984266119941e-06, + "loss": 0.9899, + "step": 7127 + }, + { + "epoch": 0.6856483262793382, + "grad_norm": 1.7919301642050507, + "learning_rate": 4.750332403986902e-06, + "loss": 0.8042, + "step": 7128 + }, + { + "epoch": 0.68574451712197, + "grad_norm": 2.1543589292061576, + "learning_rate": 4.747681051379299e-06, + "loss": 0.9555, + "step": 7129 + }, + { + "epoch": 0.6858407079646017, + "grad_norm": 2.1024490384896, + "learning_rate": 4.7450302085544735e-06, + "loss": 0.8048, + "step": 7130 + }, + { + "epoch": 0.6859368988072335, + "grad_norm": 2.2585266264502493, + "learning_rate": 4.742379875769716e-06, + "loss": 0.994, + "step": 7131 + }, + { + "epoch": 0.6860330896498653, + "grad_norm": 2.361825041658134, + "learning_rate": 4.739730053282255e-06, + "loss": 0.9801, + "step": 7132 + }, + { + "epoch": 0.6861292804924971, + "grad_norm": 2.0857521370085585, + "learning_rate": 4.737080741349286e-06, + "loss": 0.9614, + "step": 7133 + }, + { + "epoch": 0.6862254713351289, + "grad_norm": 2.1272595888900656, + "learning_rate": 4.734431940227951e-06, + "loss": 1.0012, + "step": 7134 + }, + { + "epoch": 0.6863216621777607, + "grad_norm": 2.2512751344624244, + "learning_rate": 4.731783650175328e-06, + "loss": 0.9827, + "step": 7135 + }, + { + "epoch": 0.6864178530203925, + "grad_norm": 1.951425558569047, + "learning_rate": 4.7291358714484594e-06, + "loss": 0.8476, + "step": 7136 + }, + { + "epoch": 0.6865140438630243, + "grad_norm": 2.13713703921426, + "learning_rate": 4.726488604304338e-06, + "loss": 1.041, + "step": 7137 + }, + { + "epoch": 0.6866102347056561, + "grad_norm": 2.026576635877115, + "learning_rate": 4.723841848999907e-06, + "loss": 0.874, + "step": 7138 + }, + { + "epoch": 0.6867064255482878, + "grad_norm": 1.85190507997625, + "learning_rate": 4.721195605792049e-06, + "loss": 0.9453, + "step": 7139 + }, + { + "epoch": 0.6868026163909196, + "grad_norm": 2.08243155315621, + "learning_rate": 4.718549874937612e-06, + "loss": 0.9737, + "step": 7140 + }, + { + "epoch": 0.6868988072335513, + "grad_norm": 1.6168234448275356, + "learning_rate": 4.7159046566933845e-06, + "loss": 0.7621, + "step": 7141 + }, + { + "epoch": 0.6869949980761831, + "grad_norm": 1.993945957535869, + "learning_rate": 4.713259951316103e-06, + "loss": 0.9484, + "step": 7142 + }, + { + "epoch": 0.6870911889188149, + "grad_norm": 2.0450205342961905, + "learning_rate": 4.710615759062461e-06, + "loss": 0.8822, + "step": 7143 + }, + { + "epoch": 0.6871873797614467, + "grad_norm": 1.8459547611768357, + "learning_rate": 4.707972080189106e-06, + "loss": 0.8443, + "step": 7144 + }, + { + "epoch": 0.6872835706040785, + "grad_norm": 1.9274418431417082, + "learning_rate": 4.7053289149526225e-06, + "loss": 0.9072, + "step": 7145 + }, + { + "epoch": 0.6873797614467103, + "grad_norm": 2.016886614067571, + "learning_rate": 4.702686263609559e-06, + "loss": 1.0116, + "step": 7146 + }, + { + "epoch": 0.6874759522893421, + "grad_norm": 2.0662493237747457, + "learning_rate": 4.7000441264164e-06, + "loss": 0.9089, + "step": 7147 + }, + { + "epoch": 0.6875721431319738, + "grad_norm": 2.125320375105112, + "learning_rate": 4.697402503629596e-06, + "loss": 0.9291, + "step": 7148 + }, + { + "epoch": 0.6876683339746056, + "grad_norm": 2.0206094883995047, + "learning_rate": 4.694761395505531e-06, + "loss": 0.9092, + "step": 7149 + }, + { + "epoch": 0.6877645248172374, + "grad_norm": 2.0300665065749, + "learning_rate": 4.69212080230055e-06, + "loss": 0.8943, + "step": 7150 + }, + { + "epoch": 0.6878607156598692, + "grad_norm": 2.0049354717879835, + "learning_rate": 4.689480724270952e-06, + "loss": 0.9465, + "step": 7151 + }, + { + "epoch": 0.687956906502501, + "grad_norm": 1.8639678340518802, + "learning_rate": 4.686841161672974e-06, + "loss": 0.8657, + "step": 7152 + }, + { + "epoch": 0.6880530973451328, + "grad_norm": 2.0745690905665257, + "learning_rate": 4.684202114762804e-06, + "loss": 0.9659, + "step": 7153 + }, + { + "epoch": 0.6881492881877646, + "grad_norm": 2.021197652413569, + "learning_rate": 4.681563583796587e-06, + "loss": 0.895, + "step": 7154 + }, + { + "epoch": 0.6882454790303963, + "grad_norm": 2.1097061399811983, + "learning_rate": 4.678925569030421e-06, + "loss": 0.9229, + "step": 7155 + }, + { + "epoch": 0.6883416698730281, + "grad_norm": 2.0011415072630547, + "learning_rate": 4.67628807072034e-06, + "loss": 0.9343, + "step": 7156 + }, + { + "epoch": 0.6884378607156598, + "grad_norm": 1.9381739001171487, + "learning_rate": 4.673651089122344e-06, + "loss": 0.9649, + "step": 7157 + }, + { + "epoch": 0.6885340515582916, + "grad_norm": 2.3204472002514893, + "learning_rate": 4.6710146244923645e-06, + "loss": 0.7457, + "step": 7158 + }, + { + "epoch": 0.6886302424009234, + "grad_norm": 1.920059138515663, + "learning_rate": 4.668378677086305e-06, + "loss": 0.9707, + "step": 7159 + }, + { + "epoch": 0.6887264332435552, + "grad_norm": 2.0745525541932492, + "learning_rate": 4.665743247159995e-06, + "loss": 0.9728, + "step": 7160 + }, + { + "epoch": 0.688822624086187, + "grad_norm": 2.3236153799963186, + "learning_rate": 4.663108334969233e-06, + "loss": 0.9843, + "step": 7161 + }, + { + "epoch": 0.6889188149288188, + "grad_norm": 2.096837617188926, + "learning_rate": 4.660473940769761e-06, + "loss": 0.9666, + "step": 7162 + }, + { + "epoch": 0.6890150057714506, + "grad_norm": 2.113175613101246, + "learning_rate": 4.6578400648172675e-06, + "loss": 0.943, + "step": 7163 + }, + { + "epoch": 0.6891111966140824, + "grad_norm": 1.9700904803060812, + "learning_rate": 4.655206707367388e-06, + "loss": 0.931, + "step": 7164 + }, + { + "epoch": 0.6892073874567142, + "grad_norm": 2.1902547300389537, + "learning_rate": 4.65257386867572e-06, + "loss": 0.918, + "step": 7165 + }, + { + "epoch": 0.6893035782993459, + "grad_norm": 1.743415492042311, + "learning_rate": 4.649941548997797e-06, + "loss": 0.8426, + "step": 7166 + }, + { + "epoch": 0.6893997691419776, + "grad_norm": 2.05398785481668, + "learning_rate": 4.647309748589112e-06, + "loss": 0.961, + "step": 7167 + }, + { + "epoch": 0.6894959599846094, + "grad_norm": 1.9766452371364012, + "learning_rate": 4.644678467705101e-06, + "loss": 1.0271, + "step": 7168 + }, + { + "epoch": 0.6895921508272412, + "grad_norm": 1.9332452551300718, + "learning_rate": 4.6420477066011645e-06, + "loss": 0.9459, + "step": 7169 + }, + { + "epoch": 0.689688341669873, + "grad_norm": 2.2342769936451803, + "learning_rate": 4.639417465532622e-06, + "loss": 0.9842, + "step": 7170 + }, + { + "epoch": 0.6897845325125048, + "grad_norm": 2.1649790851487496, + "learning_rate": 4.63678774475477e-06, + "loss": 0.9502, + "step": 7171 + }, + { + "epoch": 0.6898807233551366, + "grad_norm": 1.8479971081008253, + "learning_rate": 4.634158544522849e-06, + "loss": 0.8683, + "step": 7172 + }, + { + "epoch": 0.6899769141977684, + "grad_norm": 1.8780479172938953, + "learning_rate": 4.631529865092037e-06, + "loss": 0.8993, + "step": 7173 + }, + { + "epoch": 0.6900731050404002, + "grad_norm": 1.943499432029915, + "learning_rate": 4.628901706717476e-06, + "loss": 0.9414, + "step": 7174 + }, + { + "epoch": 0.6901692958830319, + "grad_norm": 1.832477801876679, + "learning_rate": 4.626274069654253e-06, + "loss": 1.0515, + "step": 7175 + }, + { + "epoch": 0.6902654867256637, + "grad_norm": 1.87116708266287, + "learning_rate": 4.623646954157399e-06, + "loss": 0.8339, + "step": 7176 + }, + { + "epoch": 0.6903616775682955, + "grad_norm": 2.222593484376215, + "learning_rate": 4.621020360481897e-06, + "loss": 0.9598, + "step": 7177 + }, + { + "epoch": 0.6904578684109273, + "grad_norm": 1.9203245272112086, + "learning_rate": 4.618394288882681e-06, + "loss": 0.8423, + "step": 7178 + }, + { + "epoch": 0.6905540592535591, + "grad_norm": 2.178806150191431, + "learning_rate": 4.615768739614641e-06, + "loss": 0.8746, + "step": 7179 + }, + { + "epoch": 0.6906502500961909, + "grad_norm": 2.5077961952227303, + "learning_rate": 4.613143712932603e-06, + "loss": 0.8459, + "step": 7180 + }, + { + "epoch": 0.6907464409388226, + "grad_norm": 1.8650143181987713, + "learning_rate": 4.610519209091346e-06, + "loss": 0.9109, + "step": 7181 + }, + { + "epoch": 0.6908426317814544, + "grad_norm": 2.0306503937281377, + "learning_rate": 4.607895228345603e-06, + "loss": 0.9587, + "step": 7182 + }, + { + "epoch": 0.6909388226240862, + "grad_norm": 1.7864651508924154, + "learning_rate": 4.605271770950057e-06, + "loss": 0.8689, + "step": 7183 + }, + { + "epoch": 0.6910350134667179, + "grad_norm": 2.051252596617164, + "learning_rate": 4.602648837159333e-06, + "loss": 1.0627, + "step": 7184 + }, + { + "epoch": 0.6911312043093497, + "grad_norm": 2.111070452258855, + "learning_rate": 4.6000264272280105e-06, + "loss": 1.1169, + "step": 7185 + }, + { + "epoch": 0.6912273951519815, + "grad_norm": 1.9372731414546498, + "learning_rate": 4.597404541410622e-06, + "loss": 0.8412, + "step": 7186 + }, + { + "epoch": 0.6913235859946133, + "grad_norm": 1.8048595208425473, + "learning_rate": 4.594783179961638e-06, + "loss": 0.9057, + "step": 7187 + }, + { + "epoch": 0.6914197768372451, + "grad_norm": 1.8892957366588405, + "learning_rate": 4.592162343135483e-06, + "loss": 0.9472, + "step": 7188 + }, + { + "epoch": 0.6915159676798769, + "grad_norm": 2.0753186328044584, + "learning_rate": 4.589542031186536e-06, + "loss": 0.9174, + "step": 7189 + }, + { + "epoch": 0.6916121585225087, + "grad_norm": 1.9117043177427613, + "learning_rate": 4.586922244369122e-06, + "loss": 0.8583, + "step": 7190 + }, + { + "epoch": 0.6917083493651405, + "grad_norm": 1.9694754889421962, + "learning_rate": 4.584302982937507e-06, + "loss": 0.9513, + "step": 7191 + }, + { + "epoch": 0.6918045402077723, + "grad_norm": 2.0336499712306955, + "learning_rate": 4.5816842471459224e-06, + "loss": 0.9213, + "step": 7192 + }, + { + "epoch": 0.691900731050404, + "grad_norm": 1.434845682332341, + "learning_rate": 4.579066037248533e-06, + "loss": 0.68, + "step": 7193 + }, + { + "epoch": 0.6919969218930357, + "grad_norm": 1.875405778650574, + "learning_rate": 4.576448353499457e-06, + "loss": 0.8053, + "step": 7194 + }, + { + "epoch": 0.6920931127356675, + "grad_norm": 1.8935724067024469, + "learning_rate": 4.5738311961527635e-06, + "loss": 0.875, + "step": 7195 + }, + { + "epoch": 0.6921893035782993, + "grad_norm": 1.8450715427991524, + "learning_rate": 4.571214565462477e-06, + "loss": 0.9063, + "step": 7196 + }, + { + "epoch": 0.6922854944209311, + "grad_norm": 1.8340880735471572, + "learning_rate": 4.568598461682556e-06, + "loss": 0.8259, + "step": 7197 + }, + { + "epoch": 0.6923816852635629, + "grad_norm": 1.8923822568330613, + "learning_rate": 4.565982885066923e-06, + "loss": 0.8446, + "step": 7198 + }, + { + "epoch": 0.6924778761061947, + "grad_norm": 1.7820525724501355, + "learning_rate": 4.563367835869434e-06, + "loss": 0.7452, + "step": 7199 + }, + { + "epoch": 0.6925740669488265, + "grad_norm": 1.9542275718639814, + "learning_rate": 4.560753314343912e-06, + "loss": 0.969, + "step": 7200 + }, + { + "epoch": 0.6926702577914583, + "grad_norm": 2.008733380014375, + "learning_rate": 4.558139320744108e-06, + "loss": 0.8903, + "step": 7201 + }, + { + "epoch": 0.69276644863409, + "grad_norm": 2.1404927361165815, + "learning_rate": 4.555525855323738e-06, + "loss": 0.9031, + "step": 7202 + }, + { + "epoch": 0.6928626394767218, + "grad_norm": 2.147798730485786, + "learning_rate": 4.552912918336465e-06, + "loss": 0.9133, + "step": 7203 + }, + { + "epoch": 0.6929588303193536, + "grad_norm": 2.0362349906090684, + "learning_rate": 4.5503005100358945e-06, + "loss": 0.9381, + "step": 7204 + }, + { + "epoch": 0.6930550211619854, + "grad_norm": 1.9147912848740656, + "learning_rate": 4.5476886306755775e-06, + "loss": 0.921, + "step": 7205 + }, + { + "epoch": 0.6931512120046172, + "grad_norm": 1.9926077674619396, + "learning_rate": 4.545077280509022e-06, + "loss": 0.8646, + "step": 7206 + }, + { + "epoch": 0.693247402847249, + "grad_norm": 1.8112387779766657, + "learning_rate": 4.54246645978969e-06, + "loss": 0.913, + "step": 7207 + }, + { + "epoch": 0.6933435936898807, + "grad_norm": 2.354702344268706, + "learning_rate": 4.539856168770974e-06, + "loss": 0.8444, + "step": 7208 + }, + { + "epoch": 0.6934397845325125, + "grad_norm": 1.8689616179702657, + "learning_rate": 4.537246407706233e-06, + "loss": 0.8863, + "step": 7209 + }, + { + "epoch": 0.6935359753751443, + "grad_norm": 1.9155277994748552, + "learning_rate": 4.534637176848758e-06, + "loss": 0.8271, + "step": 7210 + }, + { + "epoch": 0.693632166217776, + "grad_norm": 1.8148513349776005, + "learning_rate": 4.532028476451808e-06, + "loss": 0.8201, + "step": 7211 + }, + { + "epoch": 0.6937283570604078, + "grad_norm": 2.075138191686755, + "learning_rate": 4.52942030676857e-06, + "loss": 1.0238, + "step": 7212 + }, + { + "epoch": 0.6938245479030396, + "grad_norm": 1.9251304146779793, + "learning_rate": 4.526812668052192e-06, + "loss": 0.9911, + "step": 7213 + }, + { + "epoch": 0.6939207387456714, + "grad_norm": 2.060307714078898, + "learning_rate": 4.524205560555774e-06, + "loss": 0.8959, + "step": 7214 + }, + { + "epoch": 0.6940169295883032, + "grad_norm": 1.9984081569969019, + "learning_rate": 4.521598984532354e-06, + "loss": 1.0205, + "step": 7215 + }, + { + "epoch": 0.694113120430935, + "grad_norm": 2.06455874748666, + "learning_rate": 4.5189929402349175e-06, + "loss": 0.8855, + "step": 7216 + }, + { + "epoch": 0.6942093112735668, + "grad_norm": 1.8803049191701853, + "learning_rate": 4.516387427916409e-06, + "loss": 0.8026, + "step": 7217 + }, + { + "epoch": 0.6943055021161986, + "grad_norm": 2.004213452471472, + "learning_rate": 4.513782447829717e-06, + "loss": 1.0163, + "step": 7218 + }, + { + "epoch": 0.6944016929588304, + "grad_norm": 2.2234153274052035, + "learning_rate": 4.511178000227674e-06, + "loss": 0.9211, + "step": 7219 + }, + { + "epoch": 0.694497883801462, + "grad_norm": 1.732972836063388, + "learning_rate": 4.508574085363065e-06, + "loss": 0.8315, + "step": 7220 + }, + { + "epoch": 0.6945940746440938, + "grad_norm": 2.124249442601919, + "learning_rate": 4.5059707034886294e-06, + "loss": 0.9906, + "step": 7221 + }, + { + "epoch": 0.6946902654867256, + "grad_norm": 1.8523188026526736, + "learning_rate": 4.503367854857035e-06, + "loss": 0.8136, + "step": 7222 + }, + { + "epoch": 0.6947864563293574, + "grad_norm": 2.1897314651575988, + "learning_rate": 4.500765539720915e-06, + "loss": 0.9791, + "step": 7223 + }, + { + "epoch": 0.6948826471719892, + "grad_norm": 2.280495674868045, + "learning_rate": 4.498163758332853e-06, + "loss": 0.7739, + "step": 7224 + }, + { + "epoch": 0.694978838014621, + "grad_norm": 2.0821626272973837, + "learning_rate": 4.495562510945366e-06, + "loss": 0.9285, + "step": 7225 + }, + { + "epoch": 0.6950750288572528, + "grad_norm": 2.2115745719752695, + "learning_rate": 4.492961797810932e-06, + "loss": 0.8395, + "step": 7226 + }, + { + "epoch": 0.6951712196998846, + "grad_norm": 2.331050878682334, + "learning_rate": 4.490361619181974e-06, + "loss": 1.0251, + "step": 7227 + }, + { + "epoch": 0.6952674105425164, + "grad_norm": 2.2362464426677455, + "learning_rate": 4.4877619753108605e-06, + "loss": 0.9191, + "step": 7228 + }, + { + "epoch": 0.6953636013851482, + "grad_norm": 2.0578898673062023, + "learning_rate": 4.485162866449904e-06, + "loss": 0.9488, + "step": 7229 + }, + { + "epoch": 0.6954597922277799, + "grad_norm": 1.9555692294492555, + "learning_rate": 4.4825642928513746e-06, + "loss": 0.8172, + "step": 7230 + }, + { + "epoch": 0.6955559830704117, + "grad_norm": 2.1236576224392416, + "learning_rate": 4.479966254767491e-06, + "loss": 0.9559, + "step": 7231 + }, + { + "epoch": 0.6956521739130435, + "grad_norm": 2.0158977526457176, + "learning_rate": 4.477368752450409e-06, + "loss": 0.975, + "step": 7232 + }, + { + "epoch": 0.6957483647556753, + "grad_norm": 1.8695221160802407, + "learning_rate": 4.474771786152236e-06, + "loss": 0.9141, + "step": 7233 + }, + { + "epoch": 0.695844555598307, + "grad_norm": 2.107266721798604, + "learning_rate": 4.472175356125036e-06, + "loss": 1.0006, + "step": 7234 + }, + { + "epoch": 0.6959407464409388, + "grad_norm": 2.1424622004123566, + "learning_rate": 4.469579462620815e-06, + "loss": 0.8665, + "step": 7235 + }, + { + "epoch": 0.6960369372835706, + "grad_norm": 1.8688924803225917, + "learning_rate": 4.466984105891521e-06, + "loss": 0.9355, + "step": 7236 + }, + { + "epoch": 0.6961331281262024, + "grad_norm": 2.2260907780015025, + "learning_rate": 4.464389286189061e-06, + "loss": 0.9421, + "step": 7237 + }, + { + "epoch": 0.6962293189688342, + "grad_norm": 1.956536375094425, + "learning_rate": 4.461795003765285e-06, + "loss": 0.9863, + "step": 7238 + }, + { + "epoch": 0.6963255098114659, + "grad_norm": 2.0239209594975724, + "learning_rate": 4.459201258871988e-06, + "loss": 0.9937, + "step": 7239 + }, + { + "epoch": 0.6964217006540977, + "grad_norm": 2.0292912912641654, + "learning_rate": 4.456608051760914e-06, + "loss": 0.9019, + "step": 7240 + }, + { + "epoch": 0.6965178914967295, + "grad_norm": 2.192216276062042, + "learning_rate": 4.454015382683755e-06, + "loss": 1.018, + "step": 7241 + }, + { + "epoch": 0.6966140823393613, + "grad_norm": 1.8487151080454913, + "learning_rate": 4.45142325189216e-06, + "loss": 0.8911, + "step": 7242 + }, + { + "epoch": 0.6967102731819931, + "grad_norm": 2.0368066795532855, + "learning_rate": 4.448831659637709e-06, + "loss": 0.9519, + "step": 7243 + }, + { + "epoch": 0.6968064640246249, + "grad_norm": 2.032737456653422, + "learning_rate": 4.446240606171945e-06, + "loss": 0.9496, + "step": 7244 + }, + { + "epoch": 0.6969026548672567, + "grad_norm": 1.8282533736118078, + "learning_rate": 4.4436500917463485e-06, + "loss": 0.9288, + "step": 7245 + }, + { + "epoch": 0.6969988457098885, + "grad_norm": 2.2039848753970315, + "learning_rate": 4.4410601166123475e-06, + "loss": 0.9486, + "step": 7246 + }, + { + "epoch": 0.6970950365525203, + "grad_norm": 1.9428088534694028, + "learning_rate": 4.4384706810213255e-06, + "loss": 0.9263, + "step": 7247 + }, + { + "epoch": 0.6971912273951519, + "grad_norm": 1.9623410815710547, + "learning_rate": 4.4358817852246124e-06, + "loss": 0.9318, + "step": 7248 + }, + { + "epoch": 0.6972874182377837, + "grad_norm": 1.9490935141173482, + "learning_rate": 4.433293429473476e-06, + "loss": 0.9994, + "step": 7249 + }, + { + "epoch": 0.6973836090804155, + "grad_norm": 2.373371987029641, + "learning_rate": 4.430705614019147e-06, + "loss": 0.94, + "step": 7250 + }, + { + "epoch": 0.6974797999230473, + "grad_norm": 2.0597274374420365, + "learning_rate": 4.4281183391127855e-06, + "loss": 0.957, + "step": 7251 + }, + { + "epoch": 0.6975759907656791, + "grad_norm": 1.9933590768600318, + "learning_rate": 4.425531605005519e-06, + "loss": 0.922, + "step": 7252 + }, + { + "epoch": 0.6976721816083109, + "grad_norm": 2.037969775905642, + "learning_rate": 4.422945411948402e-06, + "loss": 0.9092, + "step": 7253 + }, + { + "epoch": 0.6977683724509427, + "grad_norm": 1.820251239112685, + "learning_rate": 4.420359760192452e-06, + "loss": 0.9163, + "step": 7254 + }, + { + "epoch": 0.6978645632935745, + "grad_norm": 2.068372842225035, + "learning_rate": 4.417774649988632e-06, + "loss": 0.9527, + "step": 7255 + }, + { + "epoch": 0.6979607541362063, + "grad_norm": 2.2199839304187514, + "learning_rate": 4.4151900815878455e-06, + "loss": 0.9899, + "step": 7256 + }, + { + "epoch": 0.698056944978838, + "grad_norm": 1.8631300828975972, + "learning_rate": 4.412606055240945e-06, + "loss": 0.9379, + "step": 7257 + }, + { + "epoch": 0.6981531358214698, + "grad_norm": 1.8100966762325954, + "learning_rate": 4.410022571198734e-06, + "loss": 0.9334, + "step": 7258 + }, + { + "epoch": 0.6982493266641016, + "grad_norm": 2.015946321549401, + "learning_rate": 4.407439629711966e-06, + "loss": 0.8634, + "step": 7259 + }, + { + "epoch": 0.6983455175067333, + "grad_norm": 1.8520730512989492, + "learning_rate": 4.404857231031332e-06, + "loss": 0.8755, + "step": 7260 + }, + { + "epoch": 0.6984417083493651, + "grad_norm": 1.796789041904657, + "learning_rate": 4.402275375407481e-06, + "loss": 0.8886, + "step": 7261 + }, + { + "epoch": 0.6985378991919969, + "grad_norm": 2.080222493580181, + "learning_rate": 4.399694063090999e-06, + "loss": 1.0007, + "step": 7262 + }, + { + "epoch": 0.6986340900346287, + "grad_norm": 2.0288071912576147, + "learning_rate": 4.397113294332432e-06, + "loss": 1.0153, + "step": 7263 + }, + { + "epoch": 0.6987302808772605, + "grad_norm": 2.172325351799537, + "learning_rate": 4.394533069382255e-06, + "loss": 0.9215, + "step": 7264 + }, + { + "epoch": 0.6988264717198923, + "grad_norm": 1.8926014969661396, + "learning_rate": 4.391953388490909e-06, + "loss": 0.9681, + "step": 7265 + }, + { + "epoch": 0.698922662562524, + "grad_norm": 1.6112842979192599, + "learning_rate": 4.3893742519087754e-06, + "loss": 0.8303, + "step": 7266 + }, + { + "epoch": 0.6990188534051558, + "grad_norm": 1.8209552465953551, + "learning_rate": 4.386795659886179e-06, + "loss": 0.8642, + "step": 7267 + }, + { + "epoch": 0.6991150442477876, + "grad_norm": 1.9692861529560999, + "learning_rate": 4.3842176126733914e-06, + "loss": 0.9435, + "step": 7268 + }, + { + "epoch": 0.6992112350904194, + "grad_norm": 2.034859331416446, + "learning_rate": 4.381640110520636e-06, + "loss": 0.9505, + "step": 7269 + }, + { + "epoch": 0.6993074259330512, + "grad_norm": 2.0475049985230114, + "learning_rate": 4.379063153678087e-06, + "loss": 0.9529, + "step": 7270 + }, + { + "epoch": 0.699403616775683, + "grad_norm": 2.0471557126793587, + "learning_rate": 4.3764867423958514e-06, + "loss": 0.9785, + "step": 7271 + }, + { + "epoch": 0.6994998076183148, + "grad_norm": 1.9735196128443093, + "learning_rate": 4.373910876923997e-06, + "loss": 0.9868, + "step": 7272 + }, + { + "epoch": 0.6995959984609466, + "grad_norm": 1.8637920812705993, + "learning_rate": 4.371335557512541e-06, + "loss": 0.9655, + "step": 7273 + }, + { + "epoch": 0.6996921893035783, + "grad_norm": 1.8743373351859869, + "learning_rate": 4.368760784411423e-06, + "loss": 0.8919, + "step": 7274 + }, + { + "epoch": 0.69978838014621, + "grad_norm": 1.8289599326179393, + "learning_rate": 4.3661865578705585e-06, + "loss": 0.8591, + "step": 7275 + }, + { + "epoch": 0.6998845709888418, + "grad_norm": 1.9437225448731954, + "learning_rate": 4.363612878139799e-06, + "loss": 0.959, + "step": 7276 + }, + { + "epoch": 0.6999807618314736, + "grad_norm": 1.994514845017122, + "learning_rate": 4.361039745468937e-06, + "loss": 0.9837, + "step": 7277 + }, + { + "epoch": 0.7000769526741054, + "grad_norm": 1.8748664045216603, + "learning_rate": 4.3584671601077224e-06, + "loss": 0.7801, + "step": 7278 + }, + { + "epoch": 0.7001731435167372, + "grad_norm": 2.1508209523730777, + "learning_rate": 4.355895122305841e-06, + "loss": 0.9258, + "step": 7279 + }, + { + "epoch": 0.700269334359369, + "grad_norm": 1.9019122293828516, + "learning_rate": 4.353323632312938e-06, + "loss": 0.8645, + "step": 7280 + }, + { + "epoch": 0.7003655252020008, + "grad_norm": 2.3229601077082633, + "learning_rate": 4.35075269037859e-06, + "loss": 0.9148, + "step": 7281 + }, + { + "epoch": 0.7004617160446326, + "grad_norm": 2.2435827645935706, + "learning_rate": 4.348182296752336e-06, + "loss": 0.9477, + "step": 7282 + }, + { + "epoch": 0.7005579068872644, + "grad_norm": 1.9406496490054082, + "learning_rate": 4.3456124516836554e-06, + "loss": 0.9143, + "step": 7283 + }, + { + "epoch": 0.7006540977298961, + "grad_norm": 1.891374139526653, + "learning_rate": 4.343043155421971e-06, + "loss": 0.8471, + "step": 7284 + }, + { + "epoch": 0.7007502885725279, + "grad_norm": 1.9717768874508723, + "learning_rate": 4.340474408216653e-06, + "loss": 0.8651, + "step": 7285 + }, + { + "epoch": 0.7008464794151597, + "grad_norm": 2.2627098133822128, + "learning_rate": 4.3379062103170214e-06, + "loss": 0.877, + "step": 7286 + }, + { + "epoch": 0.7009426702577914, + "grad_norm": 1.8624147330872522, + "learning_rate": 4.3353385619723475e-06, + "loss": 0.9827, + "step": 7287 + }, + { + "epoch": 0.7010388611004232, + "grad_norm": 2.0506059376884616, + "learning_rate": 4.332771463431837e-06, + "loss": 0.9633, + "step": 7288 + }, + { + "epoch": 0.701135051943055, + "grad_norm": 2.0073453773911374, + "learning_rate": 4.330204914944649e-06, + "loss": 0.9471, + "step": 7289 + }, + { + "epoch": 0.7012312427856868, + "grad_norm": 2.196822155355831, + "learning_rate": 4.327638916759898e-06, + "loss": 0.9696, + "step": 7290 + }, + { + "epoch": 0.7013274336283186, + "grad_norm": 2.0689200891434383, + "learning_rate": 4.325073469126627e-06, + "loss": 0.9232, + "step": 7291 + }, + { + "epoch": 0.7014236244709504, + "grad_norm": 2.097021860600131, + "learning_rate": 4.322508572293836e-06, + "loss": 0.8918, + "step": 7292 + }, + { + "epoch": 0.7015198153135821, + "grad_norm": 2.08071711677035, + "learning_rate": 4.31994422651047e-06, + "loss": 1.0162, + "step": 7293 + }, + { + "epoch": 0.7016160061562139, + "grad_norm": 2.1833674603114193, + "learning_rate": 4.317380432025428e-06, + "loss": 0.8633, + "step": 7294 + }, + { + "epoch": 0.7017121969988457, + "grad_norm": 2.217699578932874, + "learning_rate": 4.314817189087539e-06, + "loss": 0.9216, + "step": 7295 + }, + { + "epoch": 0.7018083878414775, + "grad_norm": 2.113841286533317, + "learning_rate": 4.312254497945595e-06, + "loss": 0.8181, + "step": 7296 + }, + { + "epoch": 0.7019045786841093, + "grad_norm": 2.201013326482761, + "learning_rate": 4.309692358848324e-06, + "loss": 0.9297, + "step": 7297 + }, + { + "epoch": 0.7020007695267411, + "grad_norm": 2.1476852220250384, + "learning_rate": 4.3071307720444015e-06, + "loss": 0.8698, + "step": 7298 + }, + { + "epoch": 0.7020969603693729, + "grad_norm": 1.989198323948642, + "learning_rate": 4.304569737782453e-06, + "loss": 0.8933, + "step": 7299 + }, + { + "epoch": 0.7021931512120047, + "grad_norm": 2.061542617889346, + "learning_rate": 4.3020092563110485e-06, + "loss": 0.9979, + "step": 7300 + }, + { + "epoch": 0.7022893420546364, + "grad_norm": 1.8676756199659967, + "learning_rate": 4.299449327878714e-06, + "loss": 0.9106, + "step": 7301 + }, + { + "epoch": 0.7023855328972681, + "grad_norm": 2.1002108147442318, + "learning_rate": 4.2968899527338984e-06, + "loss": 0.9515, + "step": 7302 + }, + { + "epoch": 0.7024817237398999, + "grad_norm": 2.0123070370856015, + "learning_rate": 4.294331131125015e-06, + "loss": 0.8934, + "step": 7303 + }, + { + "epoch": 0.7025779145825317, + "grad_norm": 2.2946680618087543, + "learning_rate": 4.291772863300428e-06, + "loss": 1.0114, + "step": 7304 + }, + { + "epoch": 0.7026741054251635, + "grad_norm": 2.173932916401254, + "learning_rate": 4.2892151495084275e-06, + "loss": 0.985, + "step": 7305 + }, + { + "epoch": 0.7027702962677953, + "grad_norm": 2.354202794200368, + "learning_rate": 4.2866579899972686e-06, + "loss": 1.0628, + "step": 7306 + }, + { + "epoch": 0.7028664871104271, + "grad_norm": 2.027531747268154, + "learning_rate": 4.284101385015147e-06, + "loss": 0.899, + "step": 7307 + }, + { + "epoch": 0.7029626779530589, + "grad_norm": 1.7362260998019665, + "learning_rate": 4.281545334810201e-06, + "loss": 0.899, + "step": 7308 + }, + { + "epoch": 0.7030588687956907, + "grad_norm": 2.029032145693454, + "learning_rate": 4.278989839630513e-06, + "loss": 0.8933, + "step": 7309 + }, + { + "epoch": 0.7031550596383225, + "grad_norm": 2.2740664006370674, + "learning_rate": 4.276434899724119e-06, + "loss": 0.798, + "step": 7310 + }, + { + "epoch": 0.7032512504809542, + "grad_norm": 2.1450691145127703, + "learning_rate": 4.2738805153390025e-06, + "loss": 1.0294, + "step": 7311 + }, + { + "epoch": 0.703347441323586, + "grad_norm": 2.522443717418391, + "learning_rate": 4.27132668672308e-06, + "loss": 1.0084, + "step": 7312 + }, + { + "epoch": 0.7034436321662177, + "grad_norm": 2.196452304608866, + "learning_rate": 4.268773414124232e-06, + "loss": 0.905, + "step": 7313 + }, + { + "epoch": 0.7035398230088495, + "grad_norm": 1.9123628977969969, + "learning_rate": 4.266220697790266e-06, + "loss": 0.9377, + "step": 7314 + }, + { + "epoch": 0.7036360138514813, + "grad_norm": 1.893951186003835, + "learning_rate": 4.263668537968955e-06, + "loss": 0.867, + "step": 7315 + }, + { + "epoch": 0.7037322046941131, + "grad_norm": 2.147201195947699, + "learning_rate": 4.2611169349079985e-06, + "loss": 0.9181, + "step": 7316 + }, + { + "epoch": 0.7038283955367449, + "grad_norm": 2.1004507936619388, + "learning_rate": 4.258565888855055e-06, + "loss": 0.8664, + "step": 7317 + }, + { + "epoch": 0.7039245863793767, + "grad_norm": 1.8408081585832263, + "learning_rate": 4.25601540005773e-06, + "loss": 0.9792, + "step": 7318 + }, + { + "epoch": 0.7040207772220085, + "grad_norm": 2.284343236031086, + "learning_rate": 4.253465468763568e-06, + "loss": 1.0136, + "step": 7319 + }, + { + "epoch": 0.7041169680646402, + "grad_norm": 2.1248517517932113, + "learning_rate": 4.250916095220056e-06, + "loss": 0.9733, + "step": 7320 + }, + { + "epoch": 0.704213158907272, + "grad_norm": 2.2397165284061726, + "learning_rate": 4.248367279674637e-06, + "loss": 0.9202, + "step": 7321 + }, + { + "epoch": 0.7043093497499038, + "grad_norm": 1.973880641860958, + "learning_rate": 4.2458190223747e-06, + "loss": 0.8872, + "step": 7322 + }, + { + "epoch": 0.7044055405925356, + "grad_norm": 1.9067776121647717, + "learning_rate": 4.243271323567567e-06, + "loss": 0.9387, + "step": 7323 + }, + { + "epoch": 0.7045017314351674, + "grad_norm": 2.0616462370099247, + "learning_rate": 4.240724183500518e-06, + "loss": 0.9425, + "step": 7324 + }, + { + "epoch": 0.7045979222777992, + "grad_norm": 1.8454702610355798, + "learning_rate": 4.2381776024207825e-06, + "loss": 0.9532, + "step": 7325 + }, + { + "epoch": 0.704694113120431, + "grad_norm": 2.1419833720008534, + "learning_rate": 4.2356315805755135e-06, + "loss": 0.9484, + "step": 7326 + }, + { + "epoch": 0.7047903039630627, + "grad_norm": 1.9550139002977942, + "learning_rate": 4.23308611821183e-06, + "loss": 0.8894, + "step": 7327 + }, + { + "epoch": 0.7048864948056945, + "grad_norm": 1.936133315827245, + "learning_rate": 4.230541215576798e-06, + "loss": 0.9403, + "step": 7328 + }, + { + "epoch": 0.7049826856483262, + "grad_norm": 2.0562361115652537, + "learning_rate": 4.227996872917412e-06, + "loss": 0.9502, + "step": 7329 + }, + { + "epoch": 0.705078876490958, + "grad_norm": 1.982186239161317, + "learning_rate": 4.225453090480631e-06, + "loss": 0.9256, + "step": 7330 + }, + { + "epoch": 0.7051750673335898, + "grad_norm": 1.9575654286470383, + "learning_rate": 4.222909868513343e-06, + "loss": 0.9716, + "step": 7331 + }, + { + "epoch": 0.7052712581762216, + "grad_norm": 1.7353983974744138, + "learning_rate": 4.220367207262398e-06, + "loss": 0.9101, + "step": 7332 + }, + { + "epoch": 0.7053674490188534, + "grad_norm": 1.9521763091378015, + "learning_rate": 4.217825106974575e-06, + "loss": 0.8585, + "step": 7333 + }, + { + "epoch": 0.7054636398614852, + "grad_norm": 1.999380111242874, + "learning_rate": 4.21528356789661e-06, + "loss": 0.8835, + "step": 7334 + }, + { + "epoch": 0.705559830704117, + "grad_norm": 1.9821875602744905, + "learning_rate": 4.212742590275185e-06, + "loss": 0.9393, + "step": 7335 + }, + { + "epoch": 0.7056560215467488, + "grad_norm": 2.0005203544318455, + "learning_rate": 4.210202174356922e-06, + "loss": 0.9685, + "step": 7336 + }, + { + "epoch": 0.7057522123893806, + "grad_norm": 1.9835035874542817, + "learning_rate": 4.207662320388383e-06, + "loss": 0.8599, + "step": 7337 + }, + { + "epoch": 0.7058484032320124, + "grad_norm": 2.045535095110443, + "learning_rate": 4.20512302861609e-06, + "loss": 0.9659, + "step": 7338 + }, + { + "epoch": 0.705944594074644, + "grad_norm": 2.2415952042769396, + "learning_rate": 4.2025842992865065e-06, + "loss": 0.9343, + "step": 7339 + }, + { + "epoch": 0.7060407849172758, + "grad_norm": 1.9477221440355803, + "learning_rate": 4.2000461326460274e-06, + "loss": 0.9678, + "step": 7340 + }, + { + "epoch": 0.7061369757599076, + "grad_norm": 2.0922520058322926, + "learning_rate": 4.19750852894101e-06, + "loss": 0.8934, + "step": 7341 + }, + { + "epoch": 0.7062331666025394, + "grad_norm": 2.193766802341114, + "learning_rate": 4.194971488417753e-06, + "loss": 0.8994, + "step": 7342 + }, + { + "epoch": 0.7063293574451712, + "grad_norm": 2.0211194382225957, + "learning_rate": 4.192435011322495e-06, + "loss": 1.0214, + "step": 7343 + }, + { + "epoch": 0.706425548287803, + "grad_norm": 1.970341579842377, + "learning_rate": 4.189899097901421e-06, + "loss": 0.8104, + "step": 7344 + }, + { + "epoch": 0.7065217391304348, + "grad_norm": 1.9033085299027936, + "learning_rate": 4.187363748400664e-06, + "loss": 0.7871, + "step": 7345 + }, + { + "epoch": 0.7066179299730666, + "grad_norm": 1.9596020133845593, + "learning_rate": 4.184828963066305e-06, + "loss": 0.925, + "step": 7346 + }, + { + "epoch": 0.7067141208156984, + "grad_norm": 1.9812185367133415, + "learning_rate": 4.182294742144363e-06, + "loss": 0.9272, + "step": 7347 + }, + { + "epoch": 0.7068103116583301, + "grad_norm": 1.8338477351172675, + "learning_rate": 4.179761085880809e-06, + "loss": 1.0147, + "step": 7348 + }, + { + "epoch": 0.7069065025009619, + "grad_norm": 2.0026539522705478, + "learning_rate": 4.177227994521552e-06, + "loss": 1.0188, + "step": 7349 + }, + { + "epoch": 0.7070026933435937, + "grad_norm": 2.1986356229831623, + "learning_rate": 4.174695468312456e-06, + "loss": 0.9684, + "step": 7350 + }, + { + "epoch": 0.7070988841862255, + "grad_norm": 2.095727605805695, + "learning_rate": 4.172163507499319e-06, + "loss": 0.9217, + "step": 7351 + }, + { + "epoch": 0.7071950750288573, + "grad_norm": 1.679449221074248, + "learning_rate": 4.16963211232789e-06, + "loss": 0.9775, + "step": 7352 + }, + { + "epoch": 0.707291265871489, + "grad_norm": 2.301581129653948, + "learning_rate": 4.167101283043873e-06, + "loss": 0.9141, + "step": 7353 + }, + { + "epoch": 0.7073874567141208, + "grad_norm": 2.399279674075043, + "learning_rate": 4.16457101989289e-06, + "loss": 0.9521, + "step": 7354 + }, + { + "epoch": 0.7074836475567526, + "grad_norm": 1.8843123297537132, + "learning_rate": 4.162041323120534e-06, + "loss": 0.9343, + "step": 7355 + }, + { + "epoch": 0.7075798383993844, + "grad_norm": 1.7573617909136645, + "learning_rate": 4.159512192972337e-06, + "loss": 0.9111, + "step": 7356 + }, + { + "epoch": 0.7076760292420161, + "grad_norm": 1.8756650674403188, + "learning_rate": 4.156983629693765e-06, + "loss": 0.9365, + "step": 7357 + }, + { + "epoch": 0.7077722200846479, + "grad_norm": 2.315467135997442, + "learning_rate": 4.15445563353024e-06, + "loss": 0.8059, + "step": 7358 + }, + { + "epoch": 0.7078684109272797, + "grad_norm": 1.9514152766827149, + "learning_rate": 4.1519282047271314e-06, + "loss": 0.9601, + "step": 7359 + }, + { + "epoch": 0.7079646017699115, + "grad_norm": 1.8198712479017518, + "learning_rate": 4.149401343529742e-06, + "loss": 0.9213, + "step": 7360 + }, + { + "epoch": 0.7080607926125433, + "grad_norm": 1.629888702687517, + "learning_rate": 4.146875050183323e-06, + "loss": 0.859, + "step": 7361 + }, + { + "epoch": 0.7081569834551751, + "grad_norm": 2.3567712138483934, + "learning_rate": 4.144349324933077e-06, + "loss": 0.9199, + "step": 7362 + }, + { + "epoch": 0.7082531742978069, + "grad_norm": 1.790863592848283, + "learning_rate": 4.14182416802415e-06, + "loss": 0.8962, + "step": 7363 + }, + { + "epoch": 0.7083493651404387, + "grad_norm": 2.110580795139393, + "learning_rate": 4.139299579701623e-06, + "loss": 0.9508, + "step": 7364 + }, + { + "epoch": 0.7084455559830705, + "grad_norm": 1.9825385549262358, + "learning_rate": 4.136775560210538e-06, + "loss": 0.8716, + "step": 7365 + }, + { + "epoch": 0.7085417468257021, + "grad_norm": 2.0879321091980527, + "learning_rate": 4.134252109795863e-06, + "loss": 0.9091, + "step": 7366 + }, + { + "epoch": 0.7086379376683339, + "grad_norm": 1.9616663286542981, + "learning_rate": 4.13172922870253e-06, + "loss": 0.8733, + "step": 7367 + }, + { + "epoch": 0.7087341285109657, + "grad_norm": 1.7879628580339362, + "learning_rate": 4.129206917175397e-06, + "loss": 0.7901, + "step": 7368 + }, + { + "epoch": 0.7088303193535975, + "grad_norm": 1.8760875368967886, + "learning_rate": 4.1266851754592815e-06, + "loss": 0.8245, + "step": 7369 + }, + { + "epoch": 0.7089265101962293, + "grad_norm": 1.858765253561096, + "learning_rate": 4.124164003798944e-06, + "loss": 0.8615, + "step": 7370 + }, + { + "epoch": 0.7090227010388611, + "grad_norm": 1.6995238785447682, + "learning_rate": 4.121643402439081e-06, + "loss": 0.8364, + "step": 7371 + }, + { + "epoch": 0.7091188918814929, + "grad_norm": 1.8410778243633363, + "learning_rate": 4.119123371624335e-06, + "loss": 0.9013, + "step": 7372 + }, + { + "epoch": 0.7092150827241247, + "grad_norm": 1.8296884388047143, + "learning_rate": 4.1166039115993e-06, + "loss": 0.8555, + "step": 7373 + }, + { + "epoch": 0.7093112735667565, + "grad_norm": 1.8064485868867968, + "learning_rate": 4.114085022608517e-06, + "loss": 0.8478, + "step": 7374 + }, + { + "epoch": 0.7094074644093882, + "grad_norm": 2.0151161336970675, + "learning_rate": 4.1115667048964566e-06, + "loss": 0.8771, + "step": 7375 + }, + { + "epoch": 0.70950365525202, + "grad_norm": 1.7003095482540103, + "learning_rate": 4.109048958707552e-06, + "loss": 0.7706, + "step": 7376 + }, + { + "epoch": 0.7095998460946518, + "grad_norm": 1.7314305415626985, + "learning_rate": 4.106531784286167e-06, + "loss": 0.8067, + "step": 7377 + }, + { + "epoch": 0.7096960369372836, + "grad_norm": 1.9527039837900708, + "learning_rate": 4.104015181876613e-06, + "loss": 0.9554, + "step": 7378 + }, + { + "epoch": 0.7097922277799154, + "grad_norm": 2.3146672770275143, + "learning_rate": 4.10149915172315e-06, + "loss": 0.8887, + "step": 7379 + }, + { + "epoch": 0.7098884186225471, + "grad_norm": 2.1375648489547046, + "learning_rate": 4.09898369406998e-06, + "loss": 0.9257, + "step": 7380 + }, + { + "epoch": 0.7099846094651789, + "grad_norm": 2.0265905460783697, + "learning_rate": 4.096468809161256e-06, + "loss": 0.8948, + "step": 7381 + }, + { + "epoch": 0.7100808003078107, + "grad_norm": 1.9390175949739108, + "learning_rate": 4.0939544972410636e-06, + "loss": 0.8911, + "step": 7382 + }, + { + "epoch": 0.7101769911504425, + "grad_norm": 2.0355943288652942, + "learning_rate": 4.091440758553435e-06, + "loss": 0.8457, + "step": 7383 + }, + { + "epoch": 0.7102731819930742, + "grad_norm": 2.4276196573767197, + "learning_rate": 4.0889275933423576e-06, + "loss": 0.8941, + "step": 7384 + }, + { + "epoch": 0.710369372835706, + "grad_norm": 2.1453467731026006, + "learning_rate": 4.086415001851749e-06, + "loss": 0.8761, + "step": 7385 + }, + { + "epoch": 0.7104655636783378, + "grad_norm": 1.8819084489831917, + "learning_rate": 4.0839029843254815e-06, + "loss": 1.0453, + "step": 7386 + }, + { + "epoch": 0.7105617545209696, + "grad_norm": 1.8762310806040594, + "learning_rate": 4.081391541007371e-06, + "loss": 0.9493, + "step": 7387 + }, + { + "epoch": 0.7106579453636014, + "grad_norm": 1.9054555534174575, + "learning_rate": 4.078880672141171e-06, + "loss": 0.9301, + "step": 7388 + }, + { + "epoch": 0.7107541362062332, + "grad_norm": 1.9716896442285141, + "learning_rate": 4.0763703779705795e-06, + "loss": 0.9908, + "step": 7389 + }, + { + "epoch": 0.710850327048865, + "grad_norm": 1.9550482822239188, + "learning_rate": 4.073860658739246e-06, + "loss": 0.9541, + "step": 7390 + }, + { + "epoch": 0.7109465178914968, + "grad_norm": 1.9116350261911577, + "learning_rate": 4.0713515146907645e-06, + "loss": 0.9449, + "step": 7391 + }, + { + "epoch": 0.7110427087341286, + "grad_norm": 2.4706806025096455, + "learning_rate": 4.068842946068661e-06, + "loss": 0.9313, + "step": 7392 + }, + { + "epoch": 0.7111388995767602, + "grad_norm": 1.634186753462492, + "learning_rate": 4.066334953116417e-06, + "loss": 0.7464, + "step": 7393 + }, + { + "epoch": 0.711235090419392, + "grad_norm": 2.1625648732594547, + "learning_rate": 4.063827536077459e-06, + "loss": 0.9635, + "step": 7394 + }, + { + "epoch": 0.7113312812620238, + "grad_norm": 2.3147443751208487, + "learning_rate": 4.06132069519515e-06, + "loss": 1.003, + "step": 7395 + }, + { + "epoch": 0.7114274721046556, + "grad_norm": 2.0018877713152237, + "learning_rate": 4.058814430712796e-06, + "loss": 0.9439, + "step": 7396 + }, + { + "epoch": 0.7115236629472874, + "grad_norm": 1.8695293147078313, + "learning_rate": 4.056308742873657e-06, + "loss": 0.9267, + "step": 7397 + }, + { + "epoch": 0.7116198537899192, + "grad_norm": 2.0306359678931143, + "learning_rate": 4.0538036319209325e-06, + "loss": 0.9384, + "step": 7398 + }, + { + "epoch": 0.711716044632551, + "grad_norm": 2.2342914841423775, + "learning_rate": 4.05129909809776e-06, + "loss": 0.9731, + "step": 7399 + }, + { + "epoch": 0.7118122354751828, + "grad_norm": 2.1212152454912916, + "learning_rate": 4.0487951416472324e-06, + "loss": 1.0022, + "step": 7400 + }, + { + "epoch": 0.7119084263178146, + "grad_norm": 2.182652074682824, + "learning_rate": 4.046291762812373e-06, + "loss": 0.9627, + "step": 7401 + }, + { + "epoch": 0.7120046171604463, + "grad_norm": 2.0918582514275803, + "learning_rate": 4.043788961836164e-06, + "loss": 0.9514, + "step": 7402 + }, + { + "epoch": 0.7121008080030781, + "grad_norm": 1.992998356199099, + "learning_rate": 4.041286738961516e-06, + "loss": 0.9074, + "step": 7403 + }, + { + "epoch": 0.7121969988457099, + "grad_norm": 2.2316138098935974, + "learning_rate": 4.038785094431295e-06, + "loss": 0.9442, + "step": 7404 + }, + { + "epoch": 0.7122931896883417, + "grad_norm": 1.8785355651781892, + "learning_rate": 4.036284028488315e-06, + "loss": 0.9195, + "step": 7405 + }, + { + "epoch": 0.7123893805309734, + "grad_norm": 2.0919494156277483, + "learning_rate": 4.0337835413753116e-06, + "loss": 0.9281, + "step": 7406 + }, + { + "epoch": 0.7124855713736052, + "grad_norm": 2.0070045010467217, + "learning_rate": 4.031283633334985e-06, + "loss": 0.9967, + "step": 7407 + }, + { + "epoch": 0.712581762216237, + "grad_norm": 2.0685952823958162, + "learning_rate": 4.0287843046099765e-06, + "loss": 0.9861, + "step": 7408 + }, + { + "epoch": 0.7126779530588688, + "grad_norm": 2.9045102683014243, + "learning_rate": 4.0262855554428594e-06, + "loss": 1.0262, + "step": 7409 + }, + { + "epoch": 0.7127741439015006, + "grad_norm": 2.0894583617303546, + "learning_rate": 4.0237873860761645e-06, + "loss": 0.8599, + "step": 7410 + }, + { + "epoch": 0.7128703347441323, + "grad_norm": 1.858583994178063, + "learning_rate": 4.021289796752363e-06, + "loss": 0.8415, + "step": 7411 + }, + { + "epoch": 0.7129665255867641, + "grad_norm": 1.896025594348371, + "learning_rate": 4.018792787713865e-06, + "loss": 1.0302, + "step": 7412 + }, + { + "epoch": 0.7130627164293959, + "grad_norm": 2.057453116855949, + "learning_rate": 4.01629635920302e-06, + "loss": 0.9192, + "step": 7413 + }, + { + "epoch": 0.7131589072720277, + "grad_norm": 2.062772343417142, + "learning_rate": 4.013800511462135e-06, + "loss": 0.9509, + "step": 7414 + }, + { + "epoch": 0.7132550981146595, + "grad_norm": 2.289746671126283, + "learning_rate": 4.011305244733455e-06, + "loss": 0.8462, + "step": 7415 + }, + { + "epoch": 0.7133512889572913, + "grad_norm": 1.7924619838992824, + "learning_rate": 4.008810559259162e-06, + "loss": 0.8504, + "step": 7416 + }, + { + "epoch": 0.7134474797999231, + "grad_norm": 1.7053403310481432, + "learning_rate": 4.006316455281393e-06, + "loss": 0.7839, + "step": 7417 + }, + { + "epoch": 0.7135436706425549, + "grad_norm": 1.9074131860866221, + "learning_rate": 4.003822933042213e-06, + "loss": 0.8332, + "step": 7418 + }, + { + "epoch": 0.7136398614851867, + "grad_norm": 2.2827441685310124, + "learning_rate": 4.001329992783649e-06, + "loss": 0.9848, + "step": 7419 + }, + { + "epoch": 0.7137360523278183, + "grad_norm": 2.0027561041586224, + "learning_rate": 3.998837634747655e-06, + "loss": 0.7832, + "step": 7420 + }, + { + "epoch": 0.7138322431704501, + "grad_norm": 2.379436246100145, + "learning_rate": 3.99634585917614e-06, + "loss": 0.8772, + "step": 7421 + }, + { + "epoch": 0.7139284340130819, + "grad_norm": 1.989423333266667, + "learning_rate": 3.993854666310955e-06, + "loss": 0.8551, + "step": 7422 + }, + { + "epoch": 0.7140246248557137, + "grad_norm": 2.1977519404668975, + "learning_rate": 3.991364056393887e-06, + "loss": 0.9218, + "step": 7423 + }, + { + "epoch": 0.7141208156983455, + "grad_norm": 2.1645358655283897, + "learning_rate": 3.98887402966667e-06, + "loss": 1.0035, + "step": 7424 + }, + { + "epoch": 0.7142170065409773, + "grad_norm": 1.952264977644941, + "learning_rate": 3.986384586370984e-06, + "loss": 1.045, + "step": 7425 + }, + { + "epoch": 0.7143131973836091, + "grad_norm": 2.089841128237915, + "learning_rate": 3.983895726748455e-06, + "loss": 0.8935, + "step": 7426 + }, + { + "epoch": 0.7144093882262409, + "grad_norm": 1.9610727681575544, + "learning_rate": 3.981407451040642e-06, + "loss": 0.9818, + "step": 7427 + }, + { + "epoch": 0.7145055790688727, + "grad_norm": 2.1123274449103735, + "learning_rate": 3.97891975948906e-06, + "loss": 0.9646, + "step": 7428 + }, + { + "epoch": 0.7146017699115044, + "grad_norm": 2.0832485965404697, + "learning_rate": 3.976432652335155e-06, + "loss": 0.9266, + "step": 7429 + }, + { + "epoch": 0.7146979607541362, + "grad_norm": 2.385233822363926, + "learning_rate": 3.973946129820326e-06, + "loss": 0.9954, + "step": 7430 + }, + { + "epoch": 0.714794151596768, + "grad_norm": 2.1557803100719912, + "learning_rate": 3.971460192185909e-06, + "loss": 0.8104, + "step": 7431 + }, + { + "epoch": 0.7148903424393998, + "grad_norm": 2.1635684548122898, + "learning_rate": 3.968974839673186e-06, + "loss": 0.8686, + "step": 7432 + }, + { + "epoch": 0.7149865332820315, + "grad_norm": 1.9454098792905692, + "learning_rate": 3.966490072523385e-06, + "loss": 0.9808, + "step": 7433 + }, + { + "epoch": 0.7150827241246633, + "grad_norm": 2.0731535219504247, + "learning_rate": 3.964005890977672e-06, + "loss": 0.9367, + "step": 7434 + }, + { + "epoch": 0.7151789149672951, + "grad_norm": 2.029965740994494, + "learning_rate": 3.9615222952771545e-06, + "loss": 0.8726, + "step": 7435 + }, + { + "epoch": 0.7152751058099269, + "grad_norm": 1.834891933200052, + "learning_rate": 3.9590392856628946e-06, + "loss": 0.9275, + "step": 7436 + }, + { + "epoch": 0.7153712966525587, + "grad_norm": 1.9700480375023055, + "learning_rate": 3.956556862375881e-06, + "loss": 0.9143, + "step": 7437 + }, + { + "epoch": 0.7154674874951905, + "grad_norm": 2.0595230820363546, + "learning_rate": 3.954075025657058e-06, + "loss": 0.9188, + "step": 7438 + }, + { + "epoch": 0.7155636783378222, + "grad_norm": 2.1196348735377293, + "learning_rate": 3.951593775747315e-06, + "loss": 0.9447, + "step": 7439 + }, + { + "epoch": 0.715659869180454, + "grad_norm": 2.0490697145737284, + "learning_rate": 3.949113112887471e-06, + "loss": 0.9934, + "step": 7440 + }, + { + "epoch": 0.7157560600230858, + "grad_norm": 1.9875606996123312, + "learning_rate": 3.946633037318297e-06, + "loss": 0.9693, + "step": 7441 + }, + { + "epoch": 0.7158522508657176, + "grad_norm": 2.237499123990056, + "learning_rate": 3.944153549280506e-06, + "loss": 1.0059, + "step": 7442 + }, + { + "epoch": 0.7159484417083494, + "grad_norm": 2.1885863874487597, + "learning_rate": 3.9416746490147584e-06, + "loss": 0.774, + "step": 7443 + }, + { + "epoch": 0.7160446325509812, + "grad_norm": 1.7905225520259498, + "learning_rate": 3.939196336761645e-06, + "loss": 0.9316, + "step": 7444 + }, + { + "epoch": 0.716140823393613, + "grad_norm": 1.932208936860802, + "learning_rate": 3.936718612761711e-06, + "loss": 1.033, + "step": 7445 + }, + { + "epoch": 0.7162370142362448, + "grad_norm": 1.8965827142200555, + "learning_rate": 3.934241477255445e-06, + "loss": 0.938, + "step": 7446 + }, + { + "epoch": 0.7163332050788765, + "grad_norm": 2.1576890550758754, + "learning_rate": 3.93176493048327e-06, + "loss": 0.8643, + "step": 7447 + }, + { + "epoch": 0.7164293959215082, + "grad_norm": 1.9334742078158447, + "learning_rate": 3.929288972685555e-06, + "loss": 0.9585, + "step": 7448 + }, + { + "epoch": 0.71652558676414, + "grad_norm": 2.402265200891068, + "learning_rate": 3.926813604102613e-06, + "loss": 0.9904, + "step": 7449 + }, + { + "epoch": 0.7166217776067718, + "grad_norm": 1.5702430527561961, + "learning_rate": 3.924338824974705e-06, + "loss": 0.8029, + "step": 7450 + }, + { + "epoch": 0.7167179684494036, + "grad_norm": 2.416375982890363, + "learning_rate": 3.921864635542027e-06, + "loss": 1.0157, + "step": 7451 + }, + { + "epoch": 0.7168141592920354, + "grad_norm": 1.9493120836849562, + "learning_rate": 3.919391036044715e-06, + "loss": 0.9089, + "step": 7452 + }, + { + "epoch": 0.7169103501346672, + "grad_norm": 2.273126891581307, + "learning_rate": 3.916918026722858e-06, + "loss": 0.8787, + "step": 7453 + }, + { + "epoch": 0.717006540977299, + "grad_norm": 1.832237590379681, + "learning_rate": 3.914445607816486e-06, + "loss": 0.8785, + "step": 7454 + }, + { + "epoch": 0.7171027318199308, + "grad_norm": 2.0921022989865423, + "learning_rate": 3.911973779565562e-06, + "loss": 0.9035, + "step": 7455 + }, + { + "epoch": 0.7171989226625626, + "grad_norm": 2.0881168588227266, + "learning_rate": 3.909502542210001e-06, + "loss": 0.8903, + "step": 7456 + }, + { + "epoch": 0.7172951135051943, + "grad_norm": 1.7596438717645595, + "learning_rate": 3.907031895989665e-06, + "loss": 0.8912, + "step": 7457 + }, + { + "epoch": 0.717391304347826, + "grad_norm": 2.14549833079571, + "learning_rate": 3.904561841144338e-06, + "loss": 0.9493, + "step": 7458 + }, + { + "epoch": 0.7174874951904578, + "grad_norm": 2.102966824734941, + "learning_rate": 3.902092377913766e-06, + "loss": 0.8989, + "step": 7459 + }, + { + "epoch": 0.7175836860330896, + "grad_norm": 1.9031163146748238, + "learning_rate": 3.899623506537635e-06, + "loss": 0.8416, + "step": 7460 + }, + { + "epoch": 0.7176798768757214, + "grad_norm": 1.913923141223475, + "learning_rate": 3.897155227255566e-06, + "loss": 0.959, + "step": 7461 + }, + { + "epoch": 0.7177760677183532, + "grad_norm": 1.7943394466003062, + "learning_rate": 3.894687540307127e-06, + "loss": 0.9613, + "step": 7462 + }, + { + "epoch": 0.717872258560985, + "grad_norm": 2.120570773844473, + "learning_rate": 3.892220445931835e-06, + "loss": 0.942, + "step": 7463 + }, + { + "epoch": 0.7179684494036168, + "grad_norm": 1.7295106785791163, + "learning_rate": 3.8897539443691355e-06, + "loss": 0.7958, + "step": 7464 + }, + { + "epoch": 0.7180646402462486, + "grad_norm": 2.023486788885807, + "learning_rate": 3.887288035858423e-06, + "loss": 0.9583, + "step": 7465 + }, + { + "epoch": 0.7181608310888803, + "grad_norm": 1.866083433410992, + "learning_rate": 3.884822720639036e-06, + "loss": 0.9037, + "step": 7466 + }, + { + "epoch": 0.7182570219315121, + "grad_norm": 2.0000349697602258, + "learning_rate": 3.882357998950262e-06, + "loss": 0.9364, + "step": 7467 + }, + { + "epoch": 0.7183532127741439, + "grad_norm": 1.7463131389302826, + "learning_rate": 3.879893871031314e-06, + "loss": 0.7902, + "step": 7468 + }, + { + "epoch": 0.7184494036167757, + "grad_norm": 2.0596317868141423, + "learning_rate": 3.877430337121363e-06, + "loss": 0.9648, + "step": 7469 + }, + { + "epoch": 0.7185455944594075, + "grad_norm": 2.181163605505284, + "learning_rate": 3.874967397459511e-06, + "loss": 0.8568, + "step": 7470 + }, + { + "epoch": 0.7186417853020393, + "grad_norm": 2.4626789407337757, + "learning_rate": 3.872505052284814e-06, + "loss": 1.0006, + "step": 7471 + }, + { + "epoch": 0.718737976144671, + "grad_norm": 1.8775964447334934, + "learning_rate": 3.870043301836256e-06, + "loss": 0.9427, + "step": 7472 + }, + { + "epoch": 0.7188341669873028, + "grad_norm": 1.8757853209466469, + "learning_rate": 3.867582146352775e-06, + "loss": 0.9325, + "step": 7473 + }, + { + "epoch": 0.7189303578299346, + "grad_norm": 1.8620055414471794, + "learning_rate": 3.86512158607325e-06, + "loss": 0.9317, + "step": 7474 + }, + { + "epoch": 0.7190265486725663, + "grad_norm": 2.5734255891075173, + "learning_rate": 3.8626616212365e-06, + "loss": 0.9441, + "step": 7475 + }, + { + "epoch": 0.7191227395151981, + "grad_norm": 1.91575467552774, + "learning_rate": 3.860202252081276e-06, + "loss": 0.8634, + "step": 7476 + }, + { + "epoch": 0.7192189303578299, + "grad_norm": 2.1330924490160785, + "learning_rate": 3.85774347884629e-06, + "loss": 0.8838, + "step": 7477 + }, + { + "epoch": 0.7193151212004617, + "grad_norm": 2.05484283327735, + "learning_rate": 3.855285301770188e-06, + "loss": 0.9733, + "step": 7478 + }, + { + "epoch": 0.7194113120430935, + "grad_norm": 2.3632152773543806, + "learning_rate": 3.85282772109155e-06, + "loss": 0.8254, + "step": 7479 + }, + { + "epoch": 0.7195075028857253, + "grad_norm": 2.176073408482039, + "learning_rate": 3.850370737048913e-06, + "loss": 0.9131, + "step": 7480 + }, + { + "epoch": 0.7196036937283571, + "grad_norm": 2.0973591019490634, + "learning_rate": 3.847914349880743e-06, + "loss": 0.9777, + "step": 7481 + }, + { + "epoch": 0.7196998845709889, + "grad_norm": 1.633005070762982, + "learning_rate": 3.8454585598254565e-06, + "loss": 0.8077, + "step": 7482 + }, + { + "epoch": 0.7197960754136207, + "grad_norm": 2.1881761618992703, + "learning_rate": 3.843003367121406e-06, + "loss": 0.9543, + "step": 7483 + }, + { + "epoch": 0.7198922662562524, + "grad_norm": 2.1945058359125267, + "learning_rate": 3.840548772006891e-06, + "loss": 0.7653, + "step": 7484 + }, + { + "epoch": 0.7199884570988841, + "grad_norm": 1.9686168737680252, + "learning_rate": 3.838094774720155e-06, + "loss": 0.9323, + "step": 7485 + }, + { + "epoch": 0.7200846479415159, + "grad_norm": 2.157096630321142, + "learning_rate": 3.835641375499375e-06, + "loss": 1.0235, + "step": 7486 + }, + { + "epoch": 0.7201808387841477, + "grad_norm": 1.9095021217239383, + "learning_rate": 3.833188574582672e-06, + "loss": 0.9895, + "step": 7487 + }, + { + "epoch": 0.7202770296267795, + "grad_norm": 1.9916546886075104, + "learning_rate": 3.830736372208118e-06, + "loss": 0.947, + "step": 7488 + }, + { + "epoch": 0.7203732204694113, + "grad_norm": 2.073327116799804, + "learning_rate": 3.8282847686137136e-06, + "loss": 0.9327, + "step": 7489 + }, + { + "epoch": 0.7204694113120431, + "grad_norm": 1.7669866456966918, + "learning_rate": 3.8258337640374125e-06, + "loss": 0.8929, + "step": 7490 + }, + { + "epoch": 0.7205656021546749, + "grad_norm": 2.219483269575287, + "learning_rate": 3.823383358717108e-06, + "loss": 0.8466, + "step": 7491 + }, + { + "epoch": 0.7206617929973067, + "grad_norm": 1.9897676580009676, + "learning_rate": 3.820933552890629e-06, + "loss": 0.8029, + "step": 7492 + }, + { + "epoch": 0.7207579838399384, + "grad_norm": 1.7364893735736637, + "learning_rate": 3.818484346795747e-06, + "loss": 0.863, + "step": 7493 + }, + { + "epoch": 0.7208541746825702, + "grad_norm": 2.1092010444682194, + "learning_rate": 3.816035740670185e-06, + "loss": 0.9589, + "step": 7494 + }, + { + "epoch": 0.720950365525202, + "grad_norm": 2.289309078417099, + "learning_rate": 3.8135877347516015e-06, + "loss": 0.9845, + "step": 7495 + }, + { + "epoch": 0.7210465563678338, + "grad_norm": 2.0635747445907335, + "learning_rate": 3.811140329277591e-06, + "loss": 0.9274, + "step": 7496 + }, + { + "epoch": 0.7211427472104656, + "grad_norm": 2.0932468390263748, + "learning_rate": 3.808693524485697e-06, + "loss": 0.876, + "step": 7497 + }, + { + "epoch": 0.7212389380530974, + "grad_norm": 2.036971180968727, + "learning_rate": 3.8062473206134088e-06, + "loss": 1.0106, + "step": 7498 + }, + { + "epoch": 0.7213351288957291, + "grad_norm": 1.9109402012444796, + "learning_rate": 3.8038017178981456e-06, + "loss": 0.9123, + "step": 7499 + }, + { + "epoch": 0.7214313197383609, + "grad_norm": 2.075800901397278, + "learning_rate": 3.8013567165772735e-06, + "loss": 0.9563, + "step": 7500 + }, + { + "epoch": 0.7215275105809927, + "grad_norm": 2.3443354814216244, + "learning_rate": 3.7989123168881026e-06, + "loss": 1.0956, + "step": 7501 + }, + { + "epoch": 0.7216237014236244, + "grad_norm": 2.088225750077387, + "learning_rate": 3.7964685190678874e-06, + "loss": 0.9495, + "step": 7502 + }, + { + "epoch": 0.7217198922662562, + "grad_norm": 1.9197897979798175, + "learning_rate": 3.794025323353815e-06, + "loss": 0.8466, + "step": 7503 + }, + { + "epoch": 0.721816083108888, + "grad_norm": 2.2958078828147253, + "learning_rate": 3.7915827299830154e-06, + "loss": 0.9214, + "step": 7504 + }, + { + "epoch": 0.7219122739515198, + "grad_norm": 1.869483366743815, + "learning_rate": 3.7891407391925674e-06, + "loss": 1.0182, + "step": 7505 + }, + { + "epoch": 0.7220084647941516, + "grad_norm": 2.3545935584264086, + "learning_rate": 3.7866993512194895e-06, + "loss": 0.9452, + "step": 7506 + }, + { + "epoch": 0.7221046556367834, + "grad_norm": 2.042838879955491, + "learning_rate": 3.7842585663007346e-06, + "loss": 0.8372, + "step": 7507 + }, + { + "epoch": 0.7222008464794152, + "grad_norm": 1.817118292182454, + "learning_rate": 3.7818183846732024e-06, + "loss": 0.839, + "step": 7508 + }, + { + "epoch": 0.722297037322047, + "grad_norm": 1.9804996612056036, + "learning_rate": 3.7793788065737434e-06, + "loss": 1.0624, + "step": 7509 + }, + { + "epoch": 0.7223932281646788, + "grad_norm": 1.8963321012533594, + "learning_rate": 3.776939832239125e-06, + "loss": 0.8749, + "step": 7510 + }, + { + "epoch": 0.7224894190073105, + "grad_norm": 1.965301535829867, + "learning_rate": 3.774501461906076e-06, + "loss": 0.7978, + "step": 7511 + }, + { + "epoch": 0.7225856098499422, + "grad_norm": 2.1047746221636863, + "learning_rate": 3.7720636958112623e-06, + "loss": 1.0074, + "step": 7512 + }, + { + "epoch": 0.722681800692574, + "grad_norm": 1.9344067006778265, + "learning_rate": 3.769626534191294e-06, + "loss": 0.9129, + "step": 7513 + }, + { + "epoch": 0.7227779915352058, + "grad_norm": 1.8242852968632584, + "learning_rate": 3.7671899772827113e-06, + "loss": 0.8186, + "step": 7514 + }, + { + "epoch": 0.7228741823778376, + "grad_norm": 1.9888507300928677, + "learning_rate": 3.7647540253220092e-06, + "loss": 0.9643, + "step": 7515 + }, + { + "epoch": 0.7229703732204694, + "grad_norm": 1.993429460236432, + "learning_rate": 3.7623186785456156e-06, + "loss": 1.0111, + "step": 7516 + }, + { + "epoch": 0.7230665640631012, + "grad_norm": 1.789086641886631, + "learning_rate": 3.7598839371898974e-06, + "loss": 0.8328, + "step": 7517 + }, + { + "epoch": 0.723162754905733, + "grad_norm": 1.9473887712561782, + "learning_rate": 3.757449801491172e-06, + "loss": 0.9416, + "step": 7518 + }, + { + "epoch": 0.7232589457483648, + "grad_norm": 2.580294428413491, + "learning_rate": 3.7550162716856963e-06, + "loss": 0.7996, + "step": 7519 + }, + { + "epoch": 0.7233551365909965, + "grad_norm": 1.9340484193369751, + "learning_rate": 3.7525833480096575e-06, + "loss": 0.9327, + "step": 7520 + }, + { + "epoch": 0.7234513274336283, + "grad_norm": 1.80553110396372, + "learning_rate": 3.7501510306991983e-06, + "loss": 0.9211, + "step": 7521 + }, + { + "epoch": 0.7235475182762601, + "grad_norm": 2.4565441996411264, + "learning_rate": 3.7477193199903903e-06, + "loss": 0.9399, + "step": 7522 + }, + { + "epoch": 0.7236437091188919, + "grad_norm": 2.037787621626827, + "learning_rate": 3.7452882161192593e-06, + "loss": 0.9892, + "step": 7523 + }, + { + "epoch": 0.7237398999615237, + "grad_norm": 1.9242176602685268, + "learning_rate": 3.7428577193217563e-06, + "loss": 0.9098, + "step": 7524 + }, + { + "epoch": 0.7238360908041555, + "grad_norm": 1.8902238811128012, + "learning_rate": 3.740427829833787e-06, + "loss": 0.9036, + "step": 7525 + }, + { + "epoch": 0.7239322816467872, + "grad_norm": 1.9569348249567189, + "learning_rate": 3.737998547891195e-06, + "loss": 0.9221, + "step": 7526 + }, + { + "epoch": 0.724028472489419, + "grad_norm": 1.9954210958618677, + "learning_rate": 3.735569873729762e-06, + "loss": 0.9379, + "step": 7527 + }, + { + "epoch": 0.7241246633320508, + "grad_norm": 1.9837690917556472, + "learning_rate": 3.7331418075852053e-06, + "loss": 0.9264, + "step": 7528 + }, + { + "epoch": 0.7242208541746825, + "grad_norm": 1.9699613519141461, + "learning_rate": 3.7307143496931952e-06, + "loss": 0.9677, + "step": 7529 + }, + { + "epoch": 0.7243170450173143, + "grad_norm": 2.6378131804206366, + "learning_rate": 3.728287500289339e-06, + "loss": 0.9444, + "step": 7530 + }, + { + "epoch": 0.7244132358599461, + "grad_norm": 1.7978031626193998, + "learning_rate": 3.7258612596091793e-06, + "loss": 0.7864, + "step": 7531 + }, + { + "epoch": 0.7245094267025779, + "grad_norm": 2.1664378149194805, + "learning_rate": 3.7234356278882076e-06, + "loss": 0.8778, + "step": 7532 + }, + { + "epoch": 0.7246056175452097, + "grad_norm": 2.2994402071400364, + "learning_rate": 3.721010605361847e-06, + "loss": 1.0738, + "step": 7533 + }, + { + "epoch": 0.7247018083878415, + "grad_norm": 1.806122733826827, + "learning_rate": 3.718586192265473e-06, + "loss": 0.8478, + "step": 7534 + }, + { + "epoch": 0.7247979992304733, + "grad_norm": 1.7288458452572029, + "learning_rate": 3.71616238883439e-06, + "loss": 0.8092, + "step": 7535 + }, + { + "epoch": 0.7248941900731051, + "grad_norm": 1.5609811158777858, + "learning_rate": 3.7137391953038516e-06, + "loss": 0.6767, + "step": 7536 + }, + { + "epoch": 0.7249903809157369, + "grad_norm": 1.6997190069682049, + "learning_rate": 3.711316611909053e-06, + "loss": 0.8701, + "step": 7537 + }, + { + "epoch": 0.7250865717583687, + "grad_norm": 1.91917648144073, + "learning_rate": 3.7088946388851223e-06, + "loss": 1.0127, + "step": 7538 + }, + { + "epoch": 0.7251827626010003, + "grad_norm": 2.093264409330805, + "learning_rate": 3.706473276467132e-06, + "loss": 0.8614, + "step": 7539 + }, + { + "epoch": 0.7252789534436321, + "grad_norm": 2.0054878744329834, + "learning_rate": 3.7040525248901003e-06, + "loss": 0.8654, + "step": 7540 + }, + { + "epoch": 0.7253751442862639, + "grad_norm": 1.9905459938411438, + "learning_rate": 3.701632384388978e-06, + "loss": 1.0438, + "step": 7541 + }, + { + "epoch": 0.7254713351288957, + "grad_norm": 1.749811363374732, + "learning_rate": 3.6992128551986617e-06, + "loss": 0.937, + "step": 7542 + }, + { + "epoch": 0.7255675259715275, + "grad_norm": 1.6598613232973505, + "learning_rate": 3.6967939375539886e-06, + "loss": 0.8247, + "step": 7543 + }, + { + "epoch": 0.7256637168141593, + "grad_norm": 2.0901725562022913, + "learning_rate": 3.6943756316897406e-06, + "loss": 0.9916, + "step": 7544 + }, + { + "epoch": 0.7257599076567911, + "grad_norm": 2.090632051013715, + "learning_rate": 3.6919579378406245e-06, + "loss": 0.8373, + "step": 7545 + }, + { + "epoch": 0.7258560984994229, + "grad_norm": 2.0451928771982004, + "learning_rate": 3.6895408562413027e-06, + "loss": 1.0347, + "step": 7546 + }, + { + "epoch": 0.7259522893420547, + "grad_norm": 1.9189013372762587, + "learning_rate": 3.687124387126377e-06, + "loss": 0.942, + "step": 7547 + }, + { + "epoch": 0.7260484801846864, + "grad_norm": 1.9580572386031567, + "learning_rate": 3.684708530730382e-06, + "loss": 0.9826, + "step": 7548 + }, + { + "epoch": 0.7261446710273182, + "grad_norm": 1.7919932212411627, + "learning_rate": 3.682293287287799e-06, + "loss": 0.8795, + "step": 7549 + }, + { + "epoch": 0.72624086186995, + "grad_norm": 2.070813588366758, + "learning_rate": 3.6798786570330526e-06, + "loss": 0.8504, + "step": 7550 + }, + { + "epoch": 0.7263370527125818, + "grad_norm": 2.0033993586992143, + "learning_rate": 3.6774646402004986e-06, + "loss": 0.9731, + "step": 7551 + }, + { + "epoch": 0.7264332435552135, + "grad_norm": 2.1175904532998, + "learning_rate": 3.6750512370244363e-06, + "loss": 1.0478, + "step": 7552 + }, + { + "epoch": 0.7265294343978453, + "grad_norm": 2.088077935532286, + "learning_rate": 3.6726384477391095e-06, + "loss": 0.9588, + "step": 7553 + }, + { + "epoch": 0.7266256252404771, + "grad_norm": 2.030758201318008, + "learning_rate": 3.670226272578704e-06, + "loss": 1.0044, + "step": 7554 + }, + { + "epoch": 0.7267218160831089, + "grad_norm": 1.777946509043547, + "learning_rate": 3.667814711777339e-06, + "loss": 0.819, + "step": 7555 + }, + { + "epoch": 0.7268180069257407, + "grad_norm": 2.137038223263769, + "learning_rate": 3.6654037655690732e-06, + "loss": 0.9072, + "step": 7556 + }, + { + "epoch": 0.7269141977683724, + "grad_norm": 1.7796237228237126, + "learning_rate": 3.6629934341879134e-06, + "loss": 0.8915, + "step": 7557 + }, + { + "epoch": 0.7270103886110042, + "grad_norm": 1.865021559906866, + "learning_rate": 3.660583717867807e-06, + "loss": 0.8414, + "step": 7558 + }, + { + "epoch": 0.727106579453636, + "grad_norm": 1.9396803398137665, + "learning_rate": 3.6581746168426303e-06, + "loss": 0.9884, + "step": 7559 + }, + { + "epoch": 0.7272027702962678, + "grad_norm": 1.9643349144513773, + "learning_rate": 3.655766131346211e-06, + "loss": 0.9257, + "step": 7560 + }, + { + "epoch": 0.7272989611388996, + "grad_norm": 2.36680713531363, + "learning_rate": 3.6533582616123164e-06, + "loss": 0.8796, + "step": 7561 + }, + { + "epoch": 0.7273951519815314, + "grad_norm": 2.0950174420584826, + "learning_rate": 3.650951007874648e-06, + "loss": 0.9491, + "step": 7562 + }, + { + "epoch": 0.7274913428241632, + "grad_norm": 1.7777247734672064, + "learning_rate": 3.6485443703668467e-06, + "loss": 0.8428, + "step": 7563 + }, + { + "epoch": 0.727587533666795, + "grad_norm": 2.1937338614078383, + "learning_rate": 3.6461383493225012e-06, + "loss": 0.9652, + "step": 7564 + }, + { + "epoch": 0.7276837245094268, + "grad_norm": 2.051722108632502, + "learning_rate": 3.643732944975139e-06, + "loss": 0.923, + "step": 7565 + }, + { + "epoch": 0.7277799153520584, + "grad_norm": 1.918953946610661, + "learning_rate": 3.6413281575582194e-06, + "loss": 0.9613, + "step": 7566 + }, + { + "epoch": 0.7278761061946902, + "grad_norm": 1.9818772949404457, + "learning_rate": 3.6389239873051554e-06, + "loss": 0.847, + "step": 7567 + }, + { + "epoch": 0.727972297037322, + "grad_norm": 1.8797122312704326, + "learning_rate": 3.6365204344492867e-06, + "loss": 0.8785, + "step": 7568 + }, + { + "epoch": 0.7280684878799538, + "grad_norm": 1.7720912466755088, + "learning_rate": 3.634117499223897e-06, + "loss": 0.7808, + "step": 7569 + }, + { + "epoch": 0.7281646787225856, + "grad_norm": 1.9179249491397166, + "learning_rate": 3.6317151818622154e-06, + "loss": 0.9276, + "step": 7570 + }, + { + "epoch": 0.7282608695652174, + "grad_norm": 2.112722380426048, + "learning_rate": 3.6293134825974095e-06, + "loss": 1.0174, + "step": 7571 + }, + { + "epoch": 0.7283570604078492, + "grad_norm": 2.0500380744294397, + "learning_rate": 3.62691240166258e-06, + "loss": 0.8863, + "step": 7572 + }, + { + "epoch": 0.728453251250481, + "grad_norm": 1.8339894451210705, + "learning_rate": 3.624511939290778e-06, + "loss": 0.8982, + "step": 7573 + }, + { + "epoch": 0.7285494420931128, + "grad_norm": 1.8579163763536124, + "learning_rate": 3.6221120957149826e-06, + "loss": 0.9026, + "step": 7574 + }, + { + "epoch": 0.7286456329357445, + "grad_norm": 1.9482325620862224, + "learning_rate": 3.6197128711681263e-06, + "loss": 0.9708, + "step": 7575 + }, + { + "epoch": 0.7287418237783763, + "grad_norm": 1.8360609313501643, + "learning_rate": 3.617314265883066e-06, + "loss": 0.9, + "step": 7576 + }, + { + "epoch": 0.7288380146210081, + "grad_norm": 2.2432412904450945, + "learning_rate": 3.6149162800926142e-06, + "loss": 0.9392, + "step": 7577 + }, + { + "epoch": 0.7289342054636399, + "grad_norm": 1.9847473878010777, + "learning_rate": 3.612518914029515e-06, + "loss": 1.0292, + "step": 7578 + }, + { + "epoch": 0.7290303963062716, + "grad_norm": 1.9114047428915035, + "learning_rate": 3.610122167926453e-06, + "loss": 0.8873, + "step": 7579 + }, + { + "epoch": 0.7291265871489034, + "grad_norm": 1.8961627879598568, + "learning_rate": 3.6077260420160487e-06, + "loss": 0.8883, + "step": 7580 + }, + { + "epoch": 0.7292227779915352, + "grad_norm": 2.2272081207391015, + "learning_rate": 3.60533053653087e-06, + "loss": 0.9099, + "step": 7581 + }, + { + "epoch": 0.729318968834167, + "grad_norm": 2.0300884867616027, + "learning_rate": 3.602935651703424e-06, + "loss": 0.8928, + "step": 7582 + }, + { + "epoch": 0.7294151596767988, + "grad_norm": 2.014528903336679, + "learning_rate": 3.6005413877661497e-06, + "loss": 0.9305, + "step": 7583 + }, + { + "epoch": 0.7295113505194305, + "grad_norm": 2.072146505662314, + "learning_rate": 3.598147744951438e-06, + "loss": 0.906, + "step": 7584 + }, + { + "epoch": 0.7296075413620623, + "grad_norm": 2.18750436491577, + "learning_rate": 3.5957547234916045e-06, + "loss": 0.9822, + "step": 7585 + }, + { + "epoch": 0.7297037322046941, + "grad_norm": 1.9691556684579126, + "learning_rate": 3.5933623236189198e-06, + "loss": 0.9338, + "step": 7586 + }, + { + "epoch": 0.7297999230473259, + "grad_norm": 2.1769300701933822, + "learning_rate": 3.5909705455655798e-06, + "loss": 0.736, + "step": 7587 + }, + { + "epoch": 0.7298961138899577, + "grad_norm": 2.088127933795111, + "learning_rate": 3.58857938956373e-06, + "loss": 0.9698, + "step": 7588 + }, + { + "epoch": 0.7299923047325895, + "grad_norm": 2.1262096270881927, + "learning_rate": 3.5861888558454583e-06, + "loss": 0.8515, + "step": 7589 + }, + { + "epoch": 0.7300884955752213, + "grad_norm": 1.8988007893546273, + "learning_rate": 3.58379894464278e-06, + "loss": 0.9605, + "step": 7590 + }, + { + "epoch": 0.7301846864178531, + "grad_norm": 1.9637792411074655, + "learning_rate": 3.5814096561876568e-06, + "loss": 0.8967, + "step": 7591 + }, + { + "epoch": 0.7302808772604849, + "grad_norm": 1.6871337401249604, + "learning_rate": 3.57902099071199e-06, + "loss": 0.7324, + "step": 7592 + }, + { + "epoch": 0.7303770681031165, + "grad_norm": 1.9458887667180693, + "learning_rate": 3.5766329484476246e-06, + "loss": 0.917, + "step": 7593 + }, + { + "epoch": 0.7304732589457483, + "grad_norm": 1.9969478825857143, + "learning_rate": 3.5742455296263346e-06, + "loss": 0.9371, + "step": 7594 + }, + { + "epoch": 0.7305694497883801, + "grad_norm": 1.8351949309337874, + "learning_rate": 3.571858734479843e-06, + "loss": 0.7832, + "step": 7595 + }, + { + "epoch": 0.7306656406310119, + "grad_norm": 1.896331860124418, + "learning_rate": 3.569472563239814e-06, + "loss": 0.9777, + "step": 7596 + }, + { + "epoch": 0.7307618314736437, + "grad_norm": 1.833622039501555, + "learning_rate": 3.5670870161378335e-06, + "loss": 1.0278, + "step": 7597 + }, + { + "epoch": 0.7308580223162755, + "grad_norm": 1.7763730835378049, + "learning_rate": 3.5647020934054465e-06, + "loss": 0.8262, + "step": 7598 + }, + { + "epoch": 0.7309542131589073, + "grad_norm": 1.7933156099006953, + "learning_rate": 3.5623177952741338e-06, + "loss": 0.932, + "step": 7599 + }, + { + "epoch": 0.7310504040015391, + "grad_norm": 2.054089732993917, + "learning_rate": 3.559934121975304e-06, + "loss": 0.8151, + "step": 7600 + }, + { + "epoch": 0.7311465948441709, + "grad_norm": 2.287612072682747, + "learning_rate": 3.5575510737403207e-06, + "loss": 1.1103, + "step": 7601 + }, + { + "epoch": 0.7312427856868026, + "grad_norm": 2.1004146626612386, + "learning_rate": 3.5551686508004735e-06, + "loss": 0.9307, + "step": 7602 + }, + { + "epoch": 0.7313389765294344, + "grad_norm": 1.9089310577867467, + "learning_rate": 3.5527868533870015e-06, + "loss": 0.9796, + "step": 7603 + }, + { + "epoch": 0.7314351673720662, + "grad_norm": 1.8317607622214163, + "learning_rate": 3.550405681731074e-06, + "loss": 0.883, + "step": 7604 + }, + { + "epoch": 0.731531358214698, + "grad_norm": 1.7725372497505547, + "learning_rate": 3.5480251360638064e-06, + "loss": 0.8242, + "step": 7605 + }, + { + "epoch": 0.7316275490573297, + "grad_norm": 2.1465395517806107, + "learning_rate": 3.5456452166162547e-06, + "loss": 0.9578, + "step": 7606 + }, + { + "epoch": 0.7317237398999615, + "grad_norm": 2.1613473178485325, + "learning_rate": 3.543265923619408e-06, + "loss": 1.0137, + "step": 7607 + }, + { + "epoch": 0.7318199307425933, + "grad_norm": 1.8960059072592972, + "learning_rate": 3.540887257304193e-06, + "loss": 0.76, + "step": 7608 + }, + { + "epoch": 0.7319161215852251, + "grad_norm": 1.9288518216011832, + "learning_rate": 3.5385092179014836e-06, + "loss": 1.0208, + "step": 7609 + }, + { + "epoch": 0.7320123124278569, + "grad_norm": 2.113165169934846, + "learning_rate": 3.5361318056420925e-06, + "loss": 1.0485, + "step": 7610 + }, + { + "epoch": 0.7321085032704886, + "grad_norm": 2.2739243298419454, + "learning_rate": 3.5337550207567594e-06, + "loss": 0.9556, + "step": 7611 + }, + { + "epoch": 0.7322046941131204, + "grad_norm": 2.1062416088906564, + "learning_rate": 3.531378863476178e-06, + "loss": 0.823, + "step": 7612 + }, + { + "epoch": 0.7323008849557522, + "grad_norm": 1.9413715884907325, + "learning_rate": 3.529003334030976e-06, + "loss": 0.8947, + "step": 7613 + }, + { + "epoch": 0.732397075798384, + "grad_norm": 2.005082447139988, + "learning_rate": 3.5266284326517165e-06, + "loss": 0.8942, + "step": 7614 + }, + { + "epoch": 0.7324932666410158, + "grad_norm": 2.0540950949150023, + "learning_rate": 3.5242541595689005e-06, + "loss": 0.8984, + "step": 7615 + }, + { + "epoch": 0.7325894574836476, + "grad_norm": 1.8949798536013072, + "learning_rate": 3.5218805150129755e-06, + "loss": 0.8718, + "step": 7616 + }, + { + "epoch": 0.7326856483262794, + "grad_norm": 2.0149043141891734, + "learning_rate": 3.5195074992143274e-06, + "loss": 0.8793, + "step": 7617 + }, + { + "epoch": 0.7327818391689112, + "grad_norm": 2.2215046188003567, + "learning_rate": 3.5171351124032703e-06, + "loss": 0.9946, + "step": 7618 + }, + { + "epoch": 0.732878030011543, + "grad_norm": 1.9346873613967672, + "learning_rate": 3.5147633548100723e-06, + "loss": 0.8735, + "step": 7619 + }, + { + "epoch": 0.7329742208541746, + "grad_norm": 2.312518992699883, + "learning_rate": 3.51239222666493e-06, + "loss": 0.923, + "step": 7620 + }, + { + "epoch": 0.7330704116968064, + "grad_norm": 1.8879391863746255, + "learning_rate": 3.5100217281979778e-06, + "loss": 0.8239, + "step": 7621 + }, + { + "epoch": 0.7331666025394382, + "grad_norm": 1.8860994579980659, + "learning_rate": 3.507651859639295e-06, + "loss": 0.9343, + "step": 7622 + }, + { + "epoch": 0.73326279338207, + "grad_norm": 2.002059758139966, + "learning_rate": 3.5052826212188997e-06, + "loss": 0.9652, + "step": 7623 + }, + { + "epoch": 0.7333589842247018, + "grad_norm": 1.7559566820186148, + "learning_rate": 3.5029140131667493e-06, + "loss": 0.7498, + "step": 7624 + }, + { + "epoch": 0.7334551750673336, + "grad_norm": 1.8165783213889881, + "learning_rate": 3.500546035712735e-06, + "loss": 0.8767, + "step": 7625 + }, + { + "epoch": 0.7335513659099654, + "grad_norm": 2.0016040346992288, + "learning_rate": 3.4981786890866853e-06, + "loss": 0.9947, + "step": 7626 + }, + { + "epoch": 0.7336475567525972, + "grad_norm": 1.7506875765980354, + "learning_rate": 3.495811973518378e-06, + "loss": 0.8829, + "step": 7627 + }, + { + "epoch": 0.733743747595229, + "grad_norm": 1.8593683366011584, + "learning_rate": 3.493445889237518e-06, + "loss": 0.8391, + "step": 7628 + }, + { + "epoch": 0.7338399384378607, + "grad_norm": 2.0718822803918875, + "learning_rate": 3.491080436473757e-06, + "loss": 0.9521, + "step": 7629 + }, + { + "epoch": 0.7339361292804925, + "grad_norm": 1.8891364457417013, + "learning_rate": 3.4887156154566847e-06, + "loss": 0.8589, + "step": 7630 + }, + { + "epoch": 0.7340323201231242, + "grad_norm": 2.0544787465234893, + "learning_rate": 3.486351426415825e-06, + "loss": 0.9686, + "step": 7631 + }, + { + "epoch": 0.734128510965756, + "grad_norm": 2.0607119984992512, + "learning_rate": 3.4839878695806385e-06, + "loss": 0.8853, + "step": 7632 + }, + { + "epoch": 0.7342247018083878, + "grad_norm": 2.0974323797951953, + "learning_rate": 3.481624945180534e-06, + "loss": 0.9204, + "step": 7633 + }, + { + "epoch": 0.7343208926510196, + "grad_norm": 2.4133563051279334, + "learning_rate": 3.4792626534448547e-06, + "loss": 0.9874, + "step": 7634 + }, + { + "epoch": 0.7344170834936514, + "grad_norm": 2.277510475785065, + "learning_rate": 3.4769009946028762e-06, + "loss": 0.9671, + "step": 7635 + }, + { + "epoch": 0.7345132743362832, + "grad_norm": 2.006844758298822, + "learning_rate": 3.4745399688838243e-06, + "loss": 0.8653, + "step": 7636 + }, + { + "epoch": 0.734609465178915, + "grad_norm": 2.046872675253851, + "learning_rate": 3.4721795765168498e-06, + "loss": 0.9819, + "step": 7637 + }, + { + "epoch": 0.7347056560215467, + "grad_norm": 1.7421266190926508, + "learning_rate": 3.469819817731056e-06, + "loss": 0.8487, + "step": 7638 + }, + { + "epoch": 0.7348018468641785, + "grad_norm": 1.920994789762515, + "learning_rate": 3.467460692755471e-06, + "loss": 0.9115, + "step": 7639 + }, + { + "epoch": 0.7348980377068103, + "grad_norm": 1.9135207641503493, + "learning_rate": 3.4651022018190715e-06, + "loss": 0.898, + "step": 7640 + }, + { + "epoch": 0.7349942285494421, + "grad_norm": 1.7880202820970827, + "learning_rate": 3.4627443451507737e-06, + "loss": 0.9159, + "step": 7641 + }, + { + "epoch": 0.7350904193920739, + "grad_norm": 2.0997656885528238, + "learning_rate": 3.460387122979423e-06, + "loss": 0.9119, + "step": 7642 + }, + { + "epoch": 0.7351866102347057, + "grad_norm": 1.8260101791955052, + "learning_rate": 3.458030535533806e-06, + "loss": 0.8781, + "step": 7643 + }, + { + "epoch": 0.7352828010773375, + "grad_norm": 1.9008787976577413, + "learning_rate": 3.455674583042652e-06, + "loss": 0.8349, + "step": 7644 + }, + { + "epoch": 0.7353789919199692, + "grad_norm": 1.8352188913468745, + "learning_rate": 3.4533192657346303e-06, + "loss": 0.8569, + "step": 7645 + }, + { + "epoch": 0.735475182762601, + "grad_norm": 1.9810421268184277, + "learning_rate": 3.4509645838383386e-06, + "loss": 0.9445, + "step": 7646 + }, + { + "epoch": 0.7355713736052328, + "grad_norm": 1.8469365867298422, + "learning_rate": 3.448610537582322e-06, + "loss": 0.9456, + "step": 7647 + }, + { + "epoch": 0.7356675644478645, + "grad_norm": 2.015862519902527, + "learning_rate": 3.4462571271950674e-06, + "loss": 0.8277, + "step": 7648 + }, + { + "epoch": 0.7357637552904963, + "grad_norm": 1.9070371495051008, + "learning_rate": 3.4439043529049796e-06, + "loss": 0.9299, + "step": 7649 + }, + { + "epoch": 0.7358599461331281, + "grad_norm": 1.7405714618385788, + "learning_rate": 3.4415522149404233e-06, + "loss": 0.8923, + "step": 7650 + }, + { + "epoch": 0.7359561369757599, + "grad_norm": 2.07490476314376, + "learning_rate": 3.439200713529698e-06, + "loss": 1.0017, + "step": 7651 + }, + { + "epoch": 0.7360523278183917, + "grad_norm": 1.880184060414193, + "learning_rate": 3.436849848901028e-06, + "loss": 0.7659, + "step": 7652 + }, + { + "epoch": 0.7361485186610235, + "grad_norm": 2.022906087831767, + "learning_rate": 3.434499621282593e-06, + "loss": 1.0086, + "step": 7653 + }, + { + "epoch": 0.7362447095036553, + "grad_norm": 2.0439581986336237, + "learning_rate": 3.432150030902497e-06, + "loss": 0.957, + "step": 7654 + }, + { + "epoch": 0.7363409003462871, + "grad_norm": 2.0591040218692824, + "learning_rate": 3.4298010779887934e-06, + "loss": 0.8985, + "step": 7655 + }, + { + "epoch": 0.7364370911889189, + "grad_norm": 1.9394984949276293, + "learning_rate": 3.427452762769462e-06, + "loss": 1.0064, + "step": 7656 + }, + { + "epoch": 0.7365332820315506, + "grad_norm": 1.9602635129744495, + "learning_rate": 3.42510508547243e-06, + "loss": 0.8774, + "step": 7657 + }, + { + "epoch": 0.7366294728741823, + "grad_norm": 1.9390684428577452, + "learning_rate": 3.4227580463255628e-06, + "loss": 0.9411, + "step": 7658 + }, + { + "epoch": 0.7367256637168141, + "grad_norm": 2.0685816195049123, + "learning_rate": 3.4204116455566596e-06, + "loss": 0.8864, + "step": 7659 + }, + { + "epoch": 0.7368218545594459, + "grad_norm": 1.8427371101526102, + "learning_rate": 3.4180658833934523e-06, + "loss": 1.0645, + "step": 7660 + }, + { + "epoch": 0.7369180454020777, + "grad_norm": 1.9081424899719173, + "learning_rate": 3.415720760063622e-06, + "loss": 0.8675, + "step": 7661 + }, + { + "epoch": 0.7370142362447095, + "grad_norm": 3.106013106257747, + "learning_rate": 3.4133762757947873e-06, + "loss": 0.9665, + "step": 7662 + }, + { + "epoch": 0.7371104270873413, + "grad_norm": 1.986730099135014, + "learning_rate": 3.411032430814494e-06, + "loss": 0.9224, + "step": 7663 + }, + { + "epoch": 0.7372066179299731, + "grad_norm": 2.0203377115587777, + "learning_rate": 3.4086892253502344e-06, + "loss": 0.9461, + "step": 7664 + }, + { + "epoch": 0.7373028087726049, + "grad_norm": 2.1088200904707755, + "learning_rate": 3.40634665962944e-06, + "loss": 0.9874, + "step": 7665 + }, + { + "epoch": 0.7373989996152366, + "grad_norm": 2.21292514650847, + "learning_rate": 3.4040047338794756e-06, + "loss": 0.8221, + "step": 7666 + }, + { + "epoch": 0.7374951904578684, + "grad_norm": 1.870924918548775, + "learning_rate": 3.40166344832764e-06, + "loss": 0.9428, + "step": 7667 + }, + { + "epoch": 0.7375913813005002, + "grad_norm": 1.742080493471946, + "learning_rate": 3.3993228032011784e-06, + "loss": 0.9133, + "step": 7668 + }, + { + "epoch": 0.737687572143132, + "grad_norm": 2.092209745930969, + "learning_rate": 3.396982798727275e-06, + "loss": 0.9828, + "step": 7669 + }, + { + "epoch": 0.7377837629857638, + "grad_norm": 1.845146739755079, + "learning_rate": 3.3946434351330415e-06, + "loss": 0.889, + "step": 7670 + }, + { + "epoch": 0.7378799538283956, + "grad_norm": 2.032599343475053, + "learning_rate": 3.3923047126455376e-06, + "loss": 0.9068, + "step": 7671 + }, + { + "epoch": 0.7379761446710273, + "grad_norm": 1.829968001220579, + "learning_rate": 3.3899666314917512e-06, + "loss": 0.8919, + "step": 7672 + }, + { + "epoch": 0.7380723355136591, + "grad_norm": 2.178411818475362, + "learning_rate": 3.3876291918986203e-06, + "loss": 0.9533, + "step": 7673 + }, + { + "epoch": 0.7381685263562909, + "grad_norm": 1.8828341796867656, + "learning_rate": 3.385292394093006e-06, + "loss": 0.8158, + "step": 7674 + }, + { + "epoch": 0.7382647171989226, + "grad_norm": 2.0129693067269225, + "learning_rate": 3.3829562383017177e-06, + "loss": 0.94, + "step": 7675 + }, + { + "epoch": 0.7383609080415544, + "grad_norm": 2.076595028337033, + "learning_rate": 3.3806207247515068e-06, + "loss": 0.8737, + "step": 7676 + }, + { + "epoch": 0.7384570988841862, + "grad_norm": 1.8100085082375357, + "learning_rate": 3.378285853669041e-06, + "loss": 0.896, + "step": 7677 + }, + { + "epoch": 0.738553289726818, + "grad_norm": 3.10029077998588, + "learning_rate": 3.375951625280948e-06, + "loss": 1.0215, + "step": 7678 + }, + { + "epoch": 0.7386494805694498, + "grad_norm": 1.9170680484092621, + "learning_rate": 3.3736180398137853e-06, + "loss": 0.8742, + "step": 7679 + }, + { + "epoch": 0.7387456714120816, + "grad_norm": 1.88479158474177, + "learning_rate": 3.3712850974940437e-06, + "loss": 0.872, + "step": 7680 + }, + { + "epoch": 0.7388418622547134, + "grad_norm": 1.7443861258770357, + "learning_rate": 3.3689527985481565e-06, + "loss": 0.8061, + "step": 7681 + }, + { + "epoch": 0.7389380530973452, + "grad_norm": 2.1164144335062027, + "learning_rate": 3.3666211432024974e-06, + "loss": 0.8876, + "step": 7682 + }, + { + "epoch": 0.739034243939977, + "grad_norm": 1.7876484122701686, + "learning_rate": 3.364290131683371e-06, + "loss": 0.8474, + "step": 7683 + }, + { + "epoch": 0.7391304347826086, + "grad_norm": 1.9250009154569456, + "learning_rate": 3.361959764217018e-06, + "loss": 0.807, + "step": 7684 + }, + { + "epoch": 0.7392266256252404, + "grad_norm": 1.7311198716936638, + "learning_rate": 3.3596300410296246e-06, + "loss": 0.9263, + "step": 7685 + }, + { + "epoch": 0.7393228164678722, + "grad_norm": 1.9776858145502365, + "learning_rate": 3.357300962347313e-06, + "loss": 0.9849, + "step": 7686 + }, + { + "epoch": 0.739419007310504, + "grad_norm": 1.7716313151251446, + "learning_rate": 3.354972528396133e-06, + "loss": 0.9101, + "step": 7687 + }, + { + "epoch": 0.7395151981531358, + "grad_norm": 2.3382177352140485, + "learning_rate": 3.3526447394020887e-06, + "loss": 0.93, + "step": 7688 + }, + { + "epoch": 0.7396113889957676, + "grad_norm": 2.307278418758446, + "learning_rate": 3.3503175955911027e-06, + "loss": 0.9462, + "step": 7689 + }, + { + "epoch": 0.7397075798383994, + "grad_norm": 2.3086106101579627, + "learning_rate": 3.3479910971890516e-06, + "loss": 0.9321, + "step": 7690 + }, + { + "epoch": 0.7398037706810312, + "grad_norm": 1.8158540885545076, + "learning_rate": 3.345665244421736e-06, + "loss": 0.9474, + "step": 7691 + }, + { + "epoch": 0.739899961523663, + "grad_norm": 1.9116596140229984, + "learning_rate": 3.343340037514903e-06, + "loss": 0.9283, + "step": 7692 + }, + { + "epoch": 0.7399961523662947, + "grad_norm": 1.9578042368095612, + "learning_rate": 3.341015476694237e-06, + "loss": 0.9533, + "step": 7693 + }, + { + "epoch": 0.7400923432089265, + "grad_norm": 2.1000765805575337, + "learning_rate": 3.3386915621853533e-06, + "loss": 0.9941, + "step": 7694 + }, + { + "epoch": 0.7401885340515583, + "grad_norm": 2.000632726292349, + "learning_rate": 3.3363682942138054e-06, + "loss": 0.9956, + "step": 7695 + }, + { + "epoch": 0.7402847248941901, + "grad_norm": 1.699162018292973, + "learning_rate": 3.3340456730050887e-06, + "loss": 0.9076, + "step": 7696 + }, + { + "epoch": 0.7403809157368219, + "grad_norm": 2.293758234458851, + "learning_rate": 3.331723698784638e-06, + "loss": 0.8551, + "step": 7697 + }, + { + "epoch": 0.7404771065794536, + "grad_norm": 2.0691783635204954, + "learning_rate": 3.3294023717778122e-06, + "loss": 0.9647, + "step": 7698 + }, + { + "epoch": 0.7405732974220854, + "grad_norm": 1.9492574900889585, + "learning_rate": 3.3270816922099227e-06, + "loss": 0.863, + "step": 7699 + }, + { + "epoch": 0.7406694882647172, + "grad_norm": 1.9381272125471825, + "learning_rate": 3.324761660306215e-06, + "loss": 0.9814, + "step": 7700 + }, + { + "epoch": 0.740765679107349, + "grad_norm": 1.9802695666850172, + "learning_rate": 3.322442276291855e-06, + "loss": 1.0027, + "step": 7701 + }, + { + "epoch": 0.7408618699499807, + "grad_norm": 2.0137975258318637, + "learning_rate": 3.3201235403919683e-06, + "loss": 0.8631, + "step": 7702 + }, + { + "epoch": 0.7409580607926125, + "grad_norm": 1.8266662648501855, + "learning_rate": 3.3178054528316094e-06, + "loss": 0.8872, + "step": 7703 + }, + { + "epoch": 0.7410542516352443, + "grad_norm": 1.6818652815761925, + "learning_rate": 3.3154880138357626e-06, + "loss": 0.8469, + "step": 7704 + }, + { + "epoch": 0.7411504424778761, + "grad_norm": 1.6230610404186907, + "learning_rate": 3.3131712236293614e-06, + "loss": 0.7604, + "step": 7705 + }, + { + "epoch": 0.7412466333205079, + "grad_norm": 2.0896932310934373, + "learning_rate": 3.3108550824372632e-06, + "loss": 0.8709, + "step": 7706 + }, + { + "epoch": 0.7413428241631397, + "grad_norm": 2.099612776114551, + "learning_rate": 3.3085395904842775e-06, + "loss": 0.8559, + "step": 7707 + }, + { + "epoch": 0.7414390150057715, + "grad_norm": 2.0426318747643357, + "learning_rate": 3.306224747995136e-06, + "loss": 0.9049, + "step": 7708 + }, + { + "epoch": 0.7415352058484033, + "grad_norm": 2.107261014024911, + "learning_rate": 3.303910555194516e-06, + "loss": 0.9174, + "step": 7709 + }, + { + "epoch": 0.7416313966910351, + "grad_norm": 2.0912108810798267, + "learning_rate": 3.301597012307034e-06, + "loss": 0.8947, + "step": 7710 + }, + { + "epoch": 0.7417275875336667, + "grad_norm": 1.9572903076520014, + "learning_rate": 3.2992841195572354e-06, + "loss": 0.7429, + "step": 7711 + }, + { + "epoch": 0.7418237783762985, + "grad_norm": 2.0090975714716297, + "learning_rate": 3.2969718771696047e-06, + "loss": 0.8643, + "step": 7712 + }, + { + "epoch": 0.7419199692189303, + "grad_norm": 2.118904854269835, + "learning_rate": 3.294660285368567e-06, + "loss": 0.847, + "step": 7713 + }, + { + "epoch": 0.7420161600615621, + "grad_norm": 1.749980623543267, + "learning_rate": 3.292349344378486e-06, + "loss": 0.8889, + "step": 7714 + }, + { + "epoch": 0.7421123509041939, + "grad_norm": 1.751455798110032, + "learning_rate": 3.2900390544236506e-06, + "loss": 0.7627, + "step": 7715 + }, + { + "epoch": 0.7422085417468257, + "grad_norm": 2.1050060001453414, + "learning_rate": 3.287729415728298e-06, + "loss": 1.0129, + "step": 7716 + }, + { + "epoch": 0.7423047325894575, + "grad_norm": 1.9445390574702595, + "learning_rate": 3.2854204285166036e-06, + "loss": 0.7973, + "step": 7717 + }, + { + "epoch": 0.7424009234320893, + "grad_norm": 2.152641822924677, + "learning_rate": 3.283112093012669e-06, + "loss": 0.8323, + "step": 7718 + }, + { + "epoch": 0.7424971142747211, + "grad_norm": 1.8052171333756246, + "learning_rate": 3.2808044094405357e-06, + "loss": 0.8792, + "step": 7719 + }, + { + "epoch": 0.7425933051173528, + "grad_norm": 2.1336173708350916, + "learning_rate": 3.278497378024187e-06, + "loss": 0.9733, + "step": 7720 + }, + { + "epoch": 0.7426894959599846, + "grad_norm": 1.9050067463991218, + "learning_rate": 3.276190998987545e-06, + "loss": 0.7892, + "step": 7721 + }, + { + "epoch": 0.7427856868026164, + "grad_norm": 2.1265026971980823, + "learning_rate": 3.2738852725544547e-06, + "loss": 0.9789, + "step": 7722 + }, + { + "epoch": 0.7428818776452482, + "grad_norm": 1.8961986033623908, + "learning_rate": 3.271580198948716e-06, + "loss": 0.9169, + "step": 7723 + }, + { + "epoch": 0.74297806848788, + "grad_norm": 1.9902442571413548, + "learning_rate": 3.2692757783940467e-06, + "loss": 0.8715, + "step": 7724 + }, + { + "epoch": 0.7430742593305117, + "grad_norm": 2.0861823953014462, + "learning_rate": 3.26697201111412e-06, + "loss": 1.0213, + "step": 7725 + }, + { + "epoch": 0.7431704501731435, + "grad_norm": 2.24271207571784, + "learning_rate": 3.264668897332527e-06, + "loss": 0.9314, + "step": 7726 + }, + { + "epoch": 0.7432666410157753, + "grad_norm": 2.17432037812732, + "learning_rate": 3.2623664372728103e-06, + "loss": 0.9415, + "step": 7727 + }, + { + "epoch": 0.7433628318584071, + "grad_norm": 2.0872021105073943, + "learning_rate": 3.2600646311584494e-06, + "loss": 0.9084, + "step": 7728 + }, + { + "epoch": 0.7434590227010388, + "grad_norm": 2.392197087180253, + "learning_rate": 3.257763479212841e-06, + "loss": 0.7532, + "step": 7729 + }, + { + "epoch": 0.7435552135436706, + "grad_norm": 2.0033543936205724, + "learning_rate": 3.2554629816593375e-06, + "loss": 0.9089, + "step": 7730 + }, + { + "epoch": 0.7436514043863024, + "grad_norm": 1.7458366968406727, + "learning_rate": 3.253163138721227e-06, + "loss": 0.8046, + "step": 7731 + }, + { + "epoch": 0.7437475952289342, + "grad_norm": 2.0269896356623374, + "learning_rate": 3.250863950621721e-06, + "loss": 0.9359, + "step": 7732 + }, + { + "epoch": 0.743843786071566, + "grad_norm": 2.1072002575592004, + "learning_rate": 3.24856541758398e-06, + "loss": 0.9553, + "step": 7733 + }, + { + "epoch": 0.7439399769141978, + "grad_norm": 2.1089460422639688, + "learning_rate": 3.2462675398310984e-06, + "loss": 0.9182, + "step": 7734 + }, + { + "epoch": 0.7440361677568296, + "grad_norm": 1.737278509263473, + "learning_rate": 3.2439703175861027e-06, + "loss": 0.8346, + "step": 7735 + }, + { + "epoch": 0.7441323585994614, + "grad_norm": 1.9922066985140947, + "learning_rate": 3.241673751071954e-06, + "loss": 0.9222, + "step": 7736 + }, + { + "epoch": 0.7442285494420932, + "grad_norm": 1.8282667191479427, + "learning_rate": 3.2393778405115583e-06, + "loss": 0.808, + "step": 7737 + }, + { + "epoch": 0.7443247402847248, + "grad_norm": 1.9111771759089309, + "learning_rate": 3.2370825861277567e-06, + "loss": 0.9074, + "step": 7738 + }, + { + "epoch": 0.7444209311273566, + "grad_norm": 2.050917223334157, + "learning_rate": 3.2347879881433154e-06, + "loss": 0.8949, + "step": 7739 + }, + { + "epoch": 0.7445171219699884, + "grad_norm": 2.0643743182771845, + "learning_rate": 3.2324940467809527e-06, + "loss": 0.8537, + "step": 7740 + }, + { + "epoch": 0.7446133128126202, + "grad_norm": 2.207965790796979, + "learning_rate": 3.230200762263308e-06, + "loss": 0.9882, + "step": 7741 + }, + { + "epoch": 0.744709503655252, + "grad_norm": 1.8441394568777358, + "learning_rate": 3.2279081348129713e-06, + "loss": 0.9524, + "step": 7742 + }, + { + "epoch": 0.7448056944978838, + "grad_norm": 2.098974607052195, + "learning_rate": 3.2256161646524566e-06, + "loss": 1.0028, + "step": 7743 + }, + { + "epoch": 0.7449018853405156, + "grad_norm": 1.8588403216007592, + "learning_rate": 3.223324852004219e-06, + "loss": 0.9294, + "step": 7744 + }, + { + "epoch": 0.7449980761831474, + "grad_norm": 2.189986753413178, + "learning_rate": 3.221034197090658e-06, + "loss": 0.8741, + "step": 7745 + }, + { + "epoch": 0.7450942670257792, + "grad_norm": 1.797530053780219, + "learning_rate": 3.2187442001340942e-06, + "loss": 0.8077, + "step": 7746 + }, + { + "epoch": 0.745190457868411, + "grad_norm": 2.4319606090534527, + "learning_rate": 3.2164548613567903e-06, + "loss": 1.0, + "step": 7747 + }, + { + "epoch": 0.7452866487110427, + "grad_norm": 2.0964634575958296, + "learning_rate": 3.21416618098095e-06, + "loss": 0.9605, + "step": 7748 + }, + { + "epoch": 0.7453828395536745, + "grad_norm": 1.9358324061145276, + "learning_rate": 3.2118781592287105e-06, + "loss": 0.8555, + "step": 7749 + }, + { + "epoch": 0.7454790303963063, + "grad_norm": 1.7584420155249492, + "learning_rate": 3.2095907963221396e-06, + "loss": 0.9272, + "step": 7750 + }, + { + "epoch": 0.745575221238938, + "grad_norm": 2.1332897108360025, + "learning_rate": 3.207304092483251e-06, + "loss": 0.9149, + "step": 7751 + }, + { + "epoch": 0.7456714120815698, + "grad_norm": 1.9640547794418315, + "learning_rate": 3.2050180479339865e-06, + "loss": 0.9636, + "step": 7752 + }, + { + "epoch": 0.7457676029242016, + "grad_norm": 1.914099755248394, + "learning_rate": 3.202732662896223e-06, + "loss": 0.8237, + "step": 7753 + }, + { + "epoch": 0.7458637937668334, + "grad_norm": 1.9196516050156214, + "learning_rate": 3.2004479375917783e-06, + "loss": 0.9178, + "step": 7754 + }, + { + "epoch": 0.7459599846094652, + "grad_norm": 2.042652351229686, + "learning_rate": 3.198163872242408e-06, + "loss": 0.8774, + "step": 7755 + }, + { + "epoch": 0.746056175452097, + "grad_norm": 1.888602395987869, + "learning_rate": 3.1958804670698008e-06, + "loss": 0.8705, + "step": 7756 + }, + { + "epoch": 0.7461523662947287, + "grad_norm": 2.067048466641504, + "learning_rate": 3.1935977222955783e-06, + "loss": 0.8787, + "step": 7757 + }, + { + "epoch": 0.7462485571373605, + "grad_norm": 2.2258646143686796, + "learning_rate": 3.191315638141297e-06, + "loss": 1.0563, + "step": 7758 + }, + { + "epoch": 0.7463447479799923, + "grad_norm": 2.242683333555599, + "learning_rate": 3.18903421482846e-06, + "loss": 0.8754, + "step": 7759 + }, + { + "epoch": 0.7464409388226241, + "grad_norm": 1.7398139282264005, + "learning_rate": 3.1867534525784937e-06, + "loss": 0.8772, + "step": 7760 + }, + { + "epoch": 0.7465371296652559, + "grad_norm": 2.068865311117451, + "learning_rate": 3.1844733516127667e-06, + "loss": 0.9568, + "step": 7761 + }, + { + "epoch": 0.7466333205078877, + "grad_norm": 1.883709173942402, + "learning_rate": 3.182193912152586e-06, + "loss": 0.9578, + "step": 7762 + }, + { + "epoch": 0.7467295113505195, + "grad_norm": 1.9011487339814384, + "learning_rate": 3.1799151344191893e-06, + "loss": 0.9745, + "step": 7763 + }, + { + "epoch": 0.7468257021931513, + "grad_norm": 1.9842807142412122, + "learning_rate": 3.177637018633746e-06, + "loss": 0.8691, + "step": 7764 + }, + { + "epoch": 0.746921893035783, + "grad_norm": 2.50632000595488, + "learning_rate": 3.1753595650173717e-06, + "loss": 0.879, + "step": 7765 + }, + { + "epoch": 0.7470180838784147, + "grad_norm": 2.1526824801370084, + "learning_rate": 3.1730827737911163e-06, + "loss": 0.9867, + "step": 7766 + }, + { + "epoch": 0.7471142747210465, + "grad_norm": 2.0053880267451674, + "learning_rate": 3.1708066451759546e-06, + "loss": 0.9392, + "step": 7767 + }, + { + "epoch": 0.7472104655636783, + "grad_norm": 2.4460353379763116, + "learning_rate": 3.1685311793928077e-06, + "loss": 1.0058, + "step": 7768 + }, + { + "epoch": 0.7473066564063101, + "grad_norm": 2.1989353657405, + "learning_rate": 3.1662563766625333e-06, + "loss": 0.8878, + "step": 7769 + }, + { + "epoch": 0.7474028472489419, + "grad_norm": 1.9119401736679595, + "learning_rate": 3.163982237205917e-06, + "loss": 0.8016, + "step": 7770 + }, + { + "epoch": 0.7474990380915737, + "grad_norm": 1.7276078055570594, + "learning_rate": 3.161708761243679e-06, + "loss": 0.8115, + "step": 7771 + }, + { + "epoch": 0.7475952289342055, + "grad_norm": 2.087221988909293, + "learning_rate": 3.1594359489964853e-06, + "loss": 0.938, + "step": 7772 + }, + { + "epoch": 0.7476914197768373, + "grad_norm": 2.224202584428127, + "learning_rate": 3.1571638006849337e-06, + "loss": 0.9015, + "step": 7773 + }, + { + "epoch": 0.7477876106194691, + "grad_norm": 1.8316391918173869, + "learning_rate": 3.15489231652955e-06, + "loss": 0.8676, + "step": 7774 + }, + { + "epoch": 0.7478838014621008, + "grad_norm": 2.042438280580578, + "learning_rate": 3.1526214967508063e-06, + "loss": 0.8406, + "step": 7775 + }, + { + "epoch": 0.7479799923047326, + "grad_norm": 2.5500639780254324, + "learning_rate": 3.150351341569101e-06, + "loss": 0.8928, + "step": 7776 + }, + { + "epoch": 0.7480761831473643, + "grad_norm": 1.887415788042384, + "learning_rate": 3.148081851204776e-06, + "loss": 0.9035, + "step": 7777 + }, + { + "epoch": 0.7481723739899961, + "grad_norm": 1.712385589918395, + "learning_rate": 3.1458130258781006e-06, + "loss": 0.9214, + "step": 7778 + }, + { + "epoch": 0.7482685648326279, + "grad_norm": 1.9107463615673004, + "learning_rate": 3.143544865809287e-06, + "loss": 0.8605, + "step": 7779 + }, + { + "epoch": 0.7483647556752597, + "grad_norm": 2.2491880005321665, + "learning_rate": 3.141277371218484e-06, + "loss": 0.9714, + "step": 7780 + }, + { + "epoch": 0.7484609465178915, + "grad_norm": 2.0318532216205805, + "learning_rate": 3.139010542325761e-06, + "loss": 0.9188, + "step": 7781 + }, + { + "epoch": 0.7485571373605233, + "grad_norm": 1.9503184182045579, + "learning_rate": 3.136744379351139e-06, + "loss": 0.8957, + "step": 7782 + }, + { + "epoch": 0.7486533282031551, + "grad_norm": 1.9682820895102375, + "learning_rate": 3.1344788825145713e-06, + "loss": 0.9333, + "step": 7783 + }, + { + "epoch": 0.7487495190457868, + "grad_norm": 1.667522367606763, + "learning_rate": 3.1322140520359366e-06, + "loss": 0.7869, + "step": 7784 + }, + { + "epoch": 0.7488457098884186, + "grad_norm": 1.7446266894713305, + "learning_rate": 3.129949888135062e-06, + "loss": 0.7816, + "step": 7785 + }, + { + "epoch": 0.7489419007310504, + "grad_norm": 1.9982004784460512, + "learning_rate": 3.1276863910317057e-06, + "loss": 0.9223, + "step": 7786 + }, + { + "epoch": 0.7490380915736822, + "grad_norm": 1.8316231983942344, + "learning_rate": 3.1254235609455553e-06, + "loss": 0.9066, + "step": 7787 + }, + { + "epoch": 0.749134282416314, + "grad_norm": 1.904443831735891, + "learning_rate": 3.1231613980962373e-06, + "loss": 0.8219, + "step": 7788 + }, + { + "epoch": 0.7492304732589458, + "grad_norm": 1.9259268934685916, + "learning_rate": 3.120899902703315e-06, + "loss": 0.8394, + "step": 7789 + }, + { + "epoch": 0.7493266641015776, + "grad_norm": 2.2766647130082043, + "learning_rate": 3.1186390749862904e-06, + "loss": 0.9782, + "step": 7790 + }, + { + "epoch": 0.7494228549442093, + "grad_norm": 1.9522713050887612, + "learning_rate": 3.1163789151645897e-06, + "loss": 0.9533, + "step": 7791 + }, + { + "epoch": 0.7495190457868411, + "grad_norm": 2.145252403770488, + "learning_rate": 3.1141194234575878e-06, + "loss": 1.0294, + "step": 7792 + }, + { + "epoch": 0.7496152366294728, + "grad_norm": 2.030418269827213, + "learning_rate": 3.1118606000845797e-06, + "loss": 0.7873, + "step": 7793 + }, + { + "epoch": 0.7497114274721046, + "grad_norm": 2.012591446852166, + "learning_rate": 3.1096024452648123e-06, + "loss": 0.958, + "step": 7794 + }, + { + "epoch": 0.7498076183147364, + "grad_norm": 2.097764828558568, + "learning_rate": 3.10734495921745e-06, + "loss": 0.899, + "step": 7795 + }, + { + "epoch": 0.7499038091573682, + "grad_norm": 1.9063193201231559, + "learning_rate": 3.1050881421616076e-06, + "loss": 0.8764, + "step": 7796 + }, + { + "epoch": 0.75, + "grad_norm": 2.022652234049953, + "learning_rate": 3.1028319943163287e-06, + "loss": 1.0053, + "step": 7797 + }, + { + "epoch": 0.7500961908426318, + "grad_norm": 1.9322247692475667, + "learning_rate": 3.100576515900591e-06, + "loss": 0.8756, + "step": 7798 + }, + { + "epoch": 0.7501923816852636, + "grad_norm": 2.10919706140481, + "learning_rate": 3.0983217071333036e-06, + "loss": 0.8028, + "step": 7799 + }, + { + "epoch": 0.7502885725278954, + "grad_norm": 2.2030885077724722, + "learning_rate": 3.0960675682333186e-06, + "loss": 0.896, + "step": 7800 + }, + { + "epoch": 0.7503847633705272, + "grad_norm": 1.984081375027549, + "learning_rate": 3.093814099419423e-06, + "loss": 0.7795, + "step": 7801 + }, + { + "epoch": 0.7504809542131589, + "grad_norm": 1.765527654782595, + "learning_rate": 3.0915613009103296e-06, + "loss": 0.8092, + "step": 7802 + }, + { + "epoch": 0.7505771450557907, + "grad_norm": 1.877045881119617, + "learning_rate": 3.089309172924697e-06, + "loss": 0.928, + "step": 7803 + }, + { + "epoch": 0.7506733358984224, + "grad_norm": 2.110623774176451, + "learning_rate": 3.0870577156811077e-06, + "loss": 0.9098, + "step": 7804 + }, + { + "epoch": 0.7507695267410542, + "grad_norm": 1.9385421628639525, + "learning_rate": 3.084806929398091e-06, + "loss": 0.9239, + "step": 7805 + }, + { + "epoch": 0.750865717583686, + "grad_norm": 2.084798002223567, + "learning_rate": 3.0825568142940998e-06, + "loss": 0.9311, + "step": 7806 + }, + { + "epoch": 0.7509619084263178, + "grad_norm": 1.9398568778055665, + "learning_rate": 3.08030737058753e-06, + "loss": 0.8027, + "step": 7807 + }, + { + "epoch": 0.7510580992689496, + "grad_norm": 2.4590753572678135, + "learning_rate": 3.0780585984967113e-06, + "loss": 0.8789, + "step": 7808 + }, + { + "epoch": 0.7511542901115814, + "grad_norm": 1.952307977919565, + "learning_rate": 3.075810498239905e-06, + "loss": 0.8579, + "step": 7809 + }, + { + "epoch": 0.7512504809542132, + "grad_norm": 2.0358500270549382, + "learning_rate": 3.073563070035305e-06, + "loss": 0.7754, + "step": 7810 + }, + { + "epoch": 0.7513466717968449, + "grad_norm": 1.884407168748181, + "learning_rate": 3.0713163141010483e-06, + "loss": 0.7768, + "step": 7811 + }, + { + "epoch": 0.7514428626394767, + "grad_norm": 2.071376768589055, + "learning_rate": 3.069070230655198e-06, + "loss": 0.87, + "step": 7812 + }, + { + "epoch": 0.7515390534821085, + "grad_norm": 1.9922841748382936, + "learning_rate": 3.066824819915758e-06, + "loss": 0.95, + "step": 7813 + }, + { + "epoch": 0.7516352443247403, + "grad_norm": 1.893496178565716, + "learning_rate": 3.0645800821006667e-06, + "loss": 0.8831, + "step": 7814 + }, + { + "epoch": 0.7517314351673721, + "grad_norm": 2.1096605754719833, + "learning_rate": 3.062336017427794e-06, + "loss": 0.9934, + "step": 7815 + }, + { + "epoch": 0.7518276260100039, + "grad_norm": 1.884168462198255, + "learning_rate": 3.060092626114941e-06, + "loss": 0.9386, + "step": 7816 + }, + { + "epoch": 0.7519238168526357, + "grad_norm": 1.8778851714120737, + "learning_rate": 3.057849908379853e-06, + "loss": 0.975, + "step": 7817 + }, + { + "epoch": 0.7520200076952674, + "grad_norm": 2.051971774895796, + "learning_rate": 3.0556078644402066e-06, + "loss": 0.918, + "step": 7818 + }, + { + "epoch": 0.7521161985378992, + "grad_norm": 2.010037206091895, + "learning_rate": 3.0533664945136053e-06, + "loss": 0.9509, + "step": 7819 + }, + { + "epoch": 0.7522123893805309, + "grad_norm": 1.942838610386814, + "learning_rate": 3.051125798817598e-06, + "loss": 1.0161, + "step": 7820 + }, + { + "epoch": 0.7523085802231627, + "grad_norm": 2.0344007534972124, + "learning_rate": 3.0488857775696645e-06, + "loss": 0.8198, + "step": 7821 + }, + { + "epoch": 0.7524047710657945, + "grad_norm": 1.9023393240365152, + "learning_rate": 3.0466464309872167e-06, + "loss": 0.959, + "step": 7822 + }, + { + "epoch": 0.7525009619084263, + "grad_norm": 2.1104966072577995, + "learning_rate": 3.0444077592875985e-06, + "loss": 0.95, + "step": 7823 + }, + { + "epoch": 0.7525971527510581, + "grad_norm": 2.0188921489366622, + "learning_rate": 3.042169762688096e-06, + "loss": 0.8708, + "step": 7824 + }, + { + "epoch": 0.7526933435936899, + "grad_norm": 2.1913793368726795, + "learning_rate": 3.0399324414059293e-06, + "loss": 0.9435, + "step": 7825 + }, + { + "epoch": 0.7527895344363217, + "grad_norm": 2.576495737277252, + "learning_rate": 3.0376957956582452e-06, + "loss": 0.7947, + "step": 7826 + }, + { + "epoch": 0.7528857252789535, + "grad_norm": 1.9037479330984377, + "learning_rate": 3.035459825662128e-06, + "loss": 0.9714, + "step": 7827 + }, + { + "epoch": 0.7529819161215853, + "grad_norm": 2.116506476874448, + "learning_rate": 3.0332245316346e-06, + "loss": 0.96, + "step": 7828 + }, + { + "epoch": 0.753078106964217, + "grad_norm": 2.070582024147193, + "learning_rate": 3.030989913792618e-06, + "loss": 0.968, + "step": 7829 + }, + { + "epoch": 0.7531742978068487, + "grad_norm": 1.891687198303632, + "learning_rate": 3.0287559723530667e-06, + "loss": 0.9555, + "step": 7830 + }, + { + "epoch": 0.7532704886494805, + "grad_norm": 1.9547925217908575, + "learning_rate": 3.0265227075327706e-06, + "loss": 0.9041, + "step": 7831 + }, + { + "epoch": 0.7533666794921123, + "grad_norm": 2.125563154860615, + "learning_rate": 3.024290119548495e-06, + "loss": 0.9455, + "step": 7832 + }, + { + "epoch": 0.7534628703347441, + "grad_norm": 2.047286610903115, + "learning_rate": 3.0220582086169194e-06, + "loss": 0.9293, + "step": 7833 + }, + { + "epoch": 0.7535590611773759, + "grad_norm": 2.1857848048030304, + "learning_rate": 3.019826974954674e-06, + "loss": 1.0031, + "step": 7834 + }, + { + "epoch": 0.7536552520200077, + "grad_norm": 2.0524806635806483, + "learning_rate": 3.0175964187783213e-06, + "loss": 0.8271, + "step": 7835 + }, + { + "epoch": 0.7537514428626395, + "grad_norm": 1.919866266217891, + "learning_rate": 3.0153665403043586e-06, + "loss": 0.8481, + "step": 7836 + }, + { + "epoch": 0.7538476337052713, + "grad_norm": 1.7726450734727115, + "learning_rate": 3.013137339749208e-06, + "loss": 0.881, + "step": 7837 + }, + { + "epoch": 0.753943824547903, + "grad_norm": 2.2860793956950705, + "learning_rate": 3.01090881732924e-06, + "loss": 0.9577, + "step": 7838 + }, + { + "epoch": 0.7540400153905348, + "grad_norm": 2.1733963102062703, + "learning_rate": 3.008680973260748e-06, + "loss": 0.9488, + "step": 7839 + }, + { + "epoch": 0.7541362062331666, + "grad_norm": 2.0910710177497163, + "learning_rate": 3.0064538077599603e-06, + "loss": 0.91, + "step": 7840 + }, + { + "epoch": 0.7542323970757984, + "grad_norm": 2.087041355985565, + "learning_rate": 3.004227321043046e-06, + "loss": 0.955, + "step": 7841 + }, + { + "epoch": 0.7543285879184302, + "grad_norm": 2.013660838956955, + "learning_rate": 3.002001513326107e-06, + "loss": 0.8625, + "step": 7842 + }, + { + "epoch": 0.754424778761062, + "grad_norm": 2.0573805693893505, + "learning_rate": 2.9997763848251727e-06, + "loss": 0.9457, + "step": 7843 + }, + { + "epoch": 0.7545209696036937, + "grad_norm": 2.0061131334791433, + "learning_rate": 2.9975519357562155e-06, + "loss": 0.9521, + "step": 7844 + }, + { + "epoch": 0.7546171604463255, + "grad_norm": 1.87223753614118, + "learning_rate": 2.995328166335131e-06, + "loss": 0.8906, + "step": 7845 + }, + { + "epoch": 0.7547133512889573, + "grad_norm": 1.6522531333464014, + "learning_rate": 2.9931050767777626e-06, + "loss": 0.7591, + "step": 7846 + }, + { + "epoch": 0.754809542131589, + "grad_norm": 2.097291448014548, + "learning_rate": 2.9908826672998737e-06, + "loss": 0.9465, + "step": 7847 + }, + { + "epoch": 0.7549057329742208, + "grad_norm": 1.9033241106435226, + "learning_rate": 2.9886609381171703e-06, + "loss": 0.8672, + "step": 7848 + }, + { + "epoch": 0.7550019238168526, + "grad_norm": 2.1403906206739327, + "learning_rate": 2.986439889445295e-06, + "loss": 0.794, + "step": 7849 + }, + { + "epoch": 0.7550981146594844, + "grad_norm": 1.9227152329131565, + "learning_rate": 2.984219521499816e-06, + "loss": 0.9343, + "step": 7850 + }, + { + "epoch": 0.7551943055021162, + "grad_norm": 2.1338289278848523, + "learning_rate": 2.981999834496235e-06, + "loss": 0.9523, + "step": 7851 + }, + { + "epoch": 0.755290496344748, + "grad_norm": 1.7547000463323805, + "learning_rate": 2.9797808286499976e-06, + "loss": 0.7873, + "step": 7852 + }, + { + "epoch": 0.7553866871873798, + "grad_norm": 1.960966759344265, + "learning_rate": 2.9775625041764776e-06, + "loss": 0.8169, + "step": 7853 + }, + { + "epoch": 0.7554828780300116, + "grad_norm": 1.9505285502951082, + "learning_rate": 2.9753448612909775e-06, + "loss": 0.9306, + "step": 7854 + }, + { + "epoch": 0.7555790688726434, + "grad_norm": 2.119953663546545, + "learning_rate": 2.9731279002087453e-06, + "loss": 0.9548, + "step": 7855 + }, + { + "epoch": 0.7556752597152752, + "grad_norm": 1.9528414718111238, + "learning_rate": 2.9709116211449484e-06, + "loss": 0.9568, + "step": 7856 + }, + { + "epoch": 0.7557714505579068, + "grad_norm": 2.1084248134116392, + "learning_rate": 2.9686960243147033e-06, + "loss": 0.9126, + "step": 7857 + }, + { + "epoch": 0.7558676414005386, + "grad_norm": 2.108672796481855, + "learning_rate": 2.966481109933047e-06, + "loss": 0.9678, + "step": 7858 + }, + { + "epoch": 0.7559638322431704, + "grad_norm": 2.071283232592033, + "learning_rate": 2.9642668782149575e-06, + "loss": 0.9197, + "step": 7859 + }, + { + "epoch": 0.7560600230858022, + "grad_norm": 1.8681045753667112, + "learning_rate": 2.9620533293753495e-06, + "loss": 0.9858, + "step": 7860 + }, + { + "epoch": 0.756156213928434, + "grad_norm": 1.762121644023336, + "learning_rate": 2.959840463629062e-06, + "loss": 0.7334, + "step": 7861 + }, + { + "epoch": 0.7562524047710658, + "grad_norm": 1.9006048955344526, + "learning_rate": 2.957628281190873e-06, + "loss": 0.8746, + "step": 7862 + }, + { + "epoch": 0.7563485956136976, + "grad_norm": 1.8779930100006395, + "learning_rate": 2.9554167822754964e-06, + "loss": 0.8917, + "step": 7863 + }, + { + "epoch": 0.7564447864563294, + "grad_norm": 2.142081884144201, + "learning_rate": 2.9532059670975732e-06, + "loss": 0.9828, + "step": 7864 + }, + { + "epoch": 0.7565409772989612, + "grad_norm": 1.9307367819953614, + "learning_rate": 2.950995835871685e-06, + "loss": 0.992, + "step": 7865 + }, + { + "epoch": 0.7566371681415929, + "grad_norm": 1.9157540801603112, + "learning_rate": 2.948786388812346e-06, + "loss": 0.9829, + "step": 7866 + }, + { + "epoch": 0.7567333589842247, + "grad_norm": 2.0808966780982296, + "learning_rate": 2.946577626134001e-06, + "loss": 0.9405, + "step": 7867 + }, + { + "epoch": 0.7568295498268565, + "grad_norm": 1.8895839437691622, + "learning_rate": 2.9443695480510225e-06, + "loss": 0.9316, + "step": 7868 + }, + { + "epoch": 0.7569257406694883, + "grad_norm": 1.8821323890639796, + "learning_rate": 2.9421621547777314e-06, + "loss": 0.9225, + "step": 7869 + }, + { + "epoch": 0.75702193151212, + "grad_norm": 2.146852354723736, + "learning_rate": 2.9399554465283742e-06, + "loss": 0.8804, + "step": 7870 + }, + { + "epoch": 0.7571181223547518, + "grad_norm": 6.471053873069071, + "learning_rate": 2.9377494235171256e-06, + "loss": 0.9295, + "step": 7871 + }, + { + "epoch": 0.7572143131973836, + "grad_norm": 2.1309178971404084, + "learning_rate": 2.935544085958102e-06, + "loss": 1.0212, + "step": 7872 + }, + { + "epoch": 0.7573105040400154, + "grad_norm": 2.01833637518646, + "learning_rate": 2.9333394340653533e-06, + "loss": 0.99, + "step": 7873 + }, + { + "epoch": 0.7574066948826472, + "grad_norm": 1.8783378517697968, + "learning_rate": 2.931135468052858e-06, + "loss": 0.9326, + "step": 7874 + }, + { + "epoch": 0.7575028857252789, + "grad_norm": 1.953955418964125, + "learning_rate": 2.9289321881345257e-06, + "loss": 0.9731, + "step": 7875 + }, + { + "epoch": 0.7575990765679107, + "grad_norm": 2.006977539506632, + "learning_rate": 2.926729594524207e-06, + "loss": 0.8697, + "step": 7876 + }, + { + "epoch": 0.7576952674105425, + "grad_norm": 2.0822595653926586, + "learning_rate": 2.9245276874356865e-06, + "loss": 0.9392, + "step": 7877 + }, + { + "epoch": 0.7577914582531743, + "grad_norm": 1.9329341667988718, + "learning_rate": 2.9223264670826746e-06, + "loss": 0.9178, + "step": 7878 + }, + { + "epoch": 0.7578876490958061, + "grad_norm": 1.5122903526416047, + "learning_rate": 2.9201259336788145e-06, + "loss": 0.7825, + "step": 7879 + }, + { + "epoch": 0.7579838399384379, + "grad_norm": 1.8951612581171378, + "learning_rate": 2.9179260874376915e-06, + "loss": 1.0276, + "step": 7880 + }, + { + "epoch": 0.7580800307810697, + "grad_norm": 2.1737268012169966, + "learning_rate": 2.9157269285728227e-06, + "loss": 0.9728, + "step": 7881 + }, + { + "epoch": 0.7581762216237015, + "grad_norm": 2.1776867087076757, + "learning_rate": 2.9135284572976486e-06, + "loss": 0.9375, + "step": 7882 + }, + { + "epoch": 0.7582724124663333, + "grad_norm": 1.8301472080611523, + "learning_rate": 2.911330673825552e-06, + "loss": 0.7831, + "step": 7883 + }, + { + "epoch": 0.7583686033089649, + "grad_norm": 1.9087839776622162, + "learning_rate": 2.9091335783698517e-06, + "loss": 0.8754, + "step": 7884 + }, + { + "epoch": 0.7584647941515967, + "grad_norm": 1.638512841438797, + "learning_rate": 2.906937171143791e-06, + "loss": 0.8154, + "step": 7885 + }, + { + "epoch": 0.7585609849942285, + "grad_norm": 1.7821483292157776, + "learning_rate": 2.9047414523605467e-06, + "loss": 0.9971, + "step": 7886 + }, + { + "epoch": 0.7586571758368603, + "grad_norm": 2.133302261847634, + "learning_rate": 2.9025464222332345e-06, + "loss": 0.9382, + "step": 7887 + }, + { + "epoch": 0.7587533666794921, + "grad_norm": 2.0941936634455067, + "learning_rate": 2.9003520809749053e-06, + "loss": 0.9181, + "step": 7888 + }, + { + "epoch": 0.7588495575221239, + "grad_norm": 2.172327809635813, + "learning_rate": 2.898158428798533e-06, + "loss": 0.9711, + "step": 7889 + }, + { + "epoch": 0.7589457483647557, + "grad_norm": 2.257911209349653, + "learning_rate": 2.8959654659170354e-06, + "loss": 1.0132, + "step": 7890 + }, + { + "epoch": 0.7590419392073875, + "grad_norm": 1.9409675441034577, + "learning_rate": 2.8937731925432555e-06, + "loss": 0.9469, + "step": 7891 + }, + { + "epoch": 0.7591381300500193, + "grad_norm": 2.158946454650605, + "learning_rate": 2.8915816088899696e-06, + "loss": 0.9252, + "step": 7892 + }, + { + "epoch": 0.759234320892651, + "grad_norm": 2.1074758373193943, + "learning_rate": 2.889390715169893e-06, + "loss": 0.9391, + "step": 7893 + }, + { + "epoch": 0.7593305117352828, + "grad_norm": 2.121821641446082, + "learning_rate": 2.8872005115956746e-06, + "loss": 0.9136, + "step": 7894 + }, + { + "epoch": 0.7594267025779146, + "grad_norm": 2.287692954366015, + "learning_rate": 2.8850109983798847e-06, + "loss": 0.9243, + "step": 7895 + }, + { + "epoch": 0.7595228934205464, + "grad_norm": 1.884338199368434, + "learning_rate": 2.8828221757350406e-06, + "loss": 0.9875, + "step": 7896 + }, + { + "epoch": 0.7596190842631781, + "grad_norm": 2.0730328254630543, + "learning_rate": 2.8806340438735814e-06, + "loss": 0.8681, + "step": 7897 + }, + { + "epoch": 0.7597152751058099, + "grad_norm": 2.316272055802184, + "learning_rate": 2.8784466030078905e-06, + "loss": 0.8363, + "step": 7898 + }, + { + "epoch": 0.7598114659484417, + "grad_norm": 1.9915806897264516, + "learning_rate": 2.8762598533502704e-06, + "loss": 0.8461, + "step": 7899 + }, + { + "epoch": 0.7599076567910735, + "grad_norm": 2.229706608820416, + "learning_rate": 2.874073795112967e-06, + "loss": 0.9126, + "step": 7900 + }, + { + "epoch": 0.7600038476337053, + "grad_norm": 2.6074745843142964, + "learning_rate": 2.8718884285081617e-06, + "loss": 0.912, + "step": 7901 + }, + { + "epoch": 0.760100038476337, + "grad_norm": 1.8958222196276253, + "learning_rate": 2.8697037537479565e-06, + "loss": 0.9232, + "step": 7902 + }, + { + "epoch": 0.7601962293189688, + "grad_norm": 2.2542475034782465, + "learning_rate": 2.8675197710443925e-06, + "loss": 0.8464, + "step": 7903 + }, + { + "epoch": 0.7602924201616006, + "grad_norm": 2.1056980139124306, + "learning_rate": 2.8653364806094454e-06, + "loss": 0.8408, + "step": 7904 + }, + { + "epoch": 0.7603886110042324, + "grad_norm": 1.9138474624847999, + "learning_rate": 2.8631538826550264e-06, + "loss": 0.8681, + "step": 7905 + }, + { + "epoch": 0.7604848018468642, + "grad_norm": 2.14874012666461, + "learning_rate": 2.86097197739297e-06, + "loss": 0.8754, + "step": 7906 + }, + { + "epoch": 0.760580992689496, + "grad_norm": 2.328156931634739, + "learning_rate": 2.858790765035053e-06, + "loss": 0.9915, + "step": 7907 + }, + { + "epoch": 0.7606771835321278, + "grad_norm": 1.8820662438549254, + "learning_rate": 2.856610245792976e-06, + "loss": 0.8081, + "step": 7908 + }, + { + "epoch": 0.7607733743747596, + "grad_norm": 2.364031113019645, + "learning_rate": 2.8544304198783834e-06, + "loss": 0.8631, + "step": 7909 + }, + { + "epoch": 0.7608695652173914, + "grad_norm": 1.7994499900467156, + "learning_rate": 2.8522512875028396e-06, + "loss": 0.8572, + "step": 7910 + }, + { + "epoch": 0.760965756060023, + "grad_norm": 1.972064401568258, + "learning_rate": 2.8500728488778508e-06, + "loss": 0.9206, + "step": 7911 + }, + { + "epoch": 0.7610619469026548, + "grad_norm": 2.0721736474378076, + "learning_rate": 2.847895104214856e-06, + "loss": 0.898, + "step": 7912 + }, + { + "epoch": 0.7611581377452866, + "grad_norm": 2.035449660147545, + "learning_rate": 2.8457180537252227e-06, + "loss": 0.9233, + "step": 7913 + }, + { + "epoch": 0.7612543285879184, + "grad_norm": 2.0871228168515503, + "learning_rate": 2.843541697620249e-06, + "loss": 0.8619, + "step": 7914 + }, + { + "epoch": 0.7613505194305502, + "grad_norm": 1.907312622057409, + "learning_rate": 2.841366036111174e-06, + "loss": 1.0619, + "step": 7915 + }, + { + "epoch": 0.761446710273182, + "grad_norm": 2.2810265575147732, + "learning_rate": 2.8391910694091584e-06, + "loss": 0.9735, + "step": 7916 + }, + { + "epoch": 0.7615429011158138, + "grad_norm": 1.9801630933171768, + "learning_rate": 2.837016797725305e-06, + "loss": 0.8705, + "step": 7917 + }, + { + "epoch": 0.7616390919584456, + "grad_norm": 2.175934841246276, + "learning_rate": 2.8348432212706443e-06, + "loss": 0.9692, + "step": 7918 + }, + { + "epoch": 0.7617352828010774, + "grad_norm": 1.7369822249343105, + "learning_rate": 2.8326703402561495e-06, + "loss": 0.9377, + "step": 7919 + }, + { + "epoch": 0.7618314736437091, + "grad_norm": 1.590910663497937, + "learning_rate": 2.8304981548927025e-06, + "loss": 0.6814, + "step": 7920 + }, + { + "epoch": 0.7619276644863409, + "grad_norm": 1.8656118085561841, + "learning_rate": 2.82832666539114e-06, + "loss": 0.9092, + "step": 7921 + }, + { + "epoch": 0.7620238553289727, + "grad_norm": 1.824774665274775, + "learning_rate": 2.826155871962227e-06, + "loss": 0.9874, + "step": 7922 + }, + { + "epoch": 0.7621200461716044, + "grad_norm": 1.9906742785144342, + "learning_rate": 2.823985774816651e-06, + "loss": 1.0215, + "step": 7923 + }, + { + "epoch": 0.7622162370142362, + "grad_norm": 2.1248920029958804, + "learning_rate": 2.8218163741650415e-06, + "loss": 0.828, + "step": 7924 + }, + { + "epoch": 0.762312427856868, + "grad_norm": 1.8941809921831805, + "learning_rate": 2.819647670217962e-06, + "loss": 0.8443, + "step": 7925 + }, + { + "epoch": 0.7624086186994998, + "grad_norm": 2.4701986756486694, + "learning_rate": 2.817479663185898e-06, + "loss": 0.8198, + "step": 7926 + }, + { + "epoch": 0.7625048095421316, + "grad_norm": 1.6285839044195438, + "learning_rate": 2.8153123532792725e-06, + "loss": 0.8084, + "step": 7927 + }, + { + "epoch": 0.7626010003847634, + "grad_norm": 1.8571318967259212, + "learning_rate": 2.813145740708445e-06, + "loss": 0.9282, + "step": 7928 + }, + { + "epoch": 0.7626971912273951, + "grad_norm": 1.8966498111980241, + "learning_rate": 2.810979825683705e-06, + "loss": 0.9625, + "step": 7929 + }, + { + "epoch": 0.7627933820700269, + "grad_norm": 2.1735530331597004, + "learning_rate": 2.808814608415271e-06, + "loss": 0.8721, + "step": 7930 + }, + { + "epoch": 0.7628895729126587, + "grad_norm": 1.919560926937871, + "learning_rate": 2.8066500891132943e-06, + "loss": 0.9088, + "step": 7931 + }, + { + "epoch": 0.7629857637552905, + "grad_norm": 2.0884514486424854, + "learning_rate": 2.8044862679878605e-06, + "loss": 0.9415, + "step": 7932 + }, + { + "epoch": 0.7630819545979223, + "grad_norm": 1.936395333860836, + "learning_rate": 2.802323145248993e-06, + "loss": 0.8993, + "step": 7933 + }, + { + "epoch": 0.7631781454405541, + "grad_norm": 4.530637208775659, + "learning_rate": 2.800160721106633e-06, + "loss": 0.8371, + "step": 7934 + }, + { + "epoch": 0.7632743362831859, + "grad_norm": 2.0252643883523147, + "learning_rate": 2.7979989957706666e-06, + "loss": 0.9098, + "step": 7935 + }, + { + "epoch": 0.7633705271258177, + "grad_norm": 2.053996921022645, + "learning_rate": 2.7958379694509108e-06, + "loss": 0.9344, + "step": 7936 + }, + { + "epoch": 0.7634667179684494, + "grad_norm": 2.1158452052301464, + "learning_rate": 2.793677642357108e-06, + "loss": 0.8944, + "step": 7937 + }, + { + "epoch": 0.7635629088110811, + "grad_norm": 1.7833925516693052, + "learning_rate": 2.791518014698935e-06, + "loss": 0.8947, + "step": 7938 + }, + { + "epoch": 0.7636590996537129, + "grad_norm": 2.1194569431594954, + "learning_rate": 2.7893590866860043e-06, + "loss": 0.8795, + "step": 7939 + }, + { + "epoch": 0.7637552904963447, + "grad_norm": 2.3612427978259283, + "learning_rate": 2.787200858527862e-06, + "loss": 0.9912, + "step": 7940 + }, + { + "epoch": 0.7638514813389765, + "grad_norm": 2.304327194466563, + "learning_rate": 2.785043330433975e-06, + "loss": 1.0017, + "step": 7941 + }, + { + "epoch": 0.7639476721816083, + "grad_norm": 1.799915448493635, + "learning_rate": 2.7828865026137584e-06, + "loss": 0.8481, + "step": 7942 + }, + { + "epoch": 0.7640438630242401, + "grad_norm": 2.345191388395854, + "learning_rate": 2.780730375276547e-06, + "loss": 0.8775, + "step": 7943 + }, + { + "epoch": 0.7641400538668719, + "grad_norm": 1.9493398489719773, + "learning_rate": 2.7785749486316085e-06, + "loss": 0.9228, + "step": 7944 + }, + { + "epoch": 0.7642362447095037, + "grad_norm": 2.000898329680106, + "learning_rate": 2.7764202228881476e-06, + "loss": 0.9194, + "step": 7945 + }, + { + "epoch": 0.7643324355521355, + "grad_norm": 2.370361038957896, + "learning_rate": 2.774266198255303e-06, + "loss": 0.8676, + "step": 7946 + }, + { + "epoch": 0.7644286263947672, + "grad_norm": 1.996261943176916, + "learning_rate": 2.7721128749421357e-06, + "loss": 0.907, + "step": 7947 + }, + { + "epoch": 0.764524817237399, + "grad_norm": 1.9871679947989427, + "learning_rate": 2.7699602531576496e-06, + "loss": 0.9138, + "step": 7948 + }, + { + "epoch": 0.7646210080800308, + "grad_norm": 2.0339770838456945, + "learning_rate": 2.7678083331107695e-06, + "loss": 0.9047, + "step": 7949 + }, + { + "epoch": 0.7647171989226625, + "grad_norm": 1.880068114705548, + "learning_rate": 2.765657115010364e-06, + "loss": 0.8763, + "step": 7950 + }, + { + "epoch": 0.7648133897652943, + "grad_norm": 2.0223980277945386, + "learning_rate": 2.76350659906522e-06, + "loss": 0.984, + "step": 7951 + }, + { + "epoch": 0.7649095806079261, + "grad_norm": 1.8263942228994814, + "learning_rate": 2.7613567854840685e-06, + "loss": 0.9324, + "step": 7952 + }, + { + "epoch": 0.7650057714505579, + "grad_norm": 2.023227717535428, + "learning_rate": 2.7592076744755682e-06, + "loss": 0.8719, + "step": 7953 + }, + { + "epoch": 0.7651019622931897, + "grad_norm": 2.0272335425356265, + "learning_rate": 2.7570592662483086e-06, + "loss": 0.9623, + "step": 7954 + }, + { + "epoch": 0.7651981531358215, + "grad_norm": 2.0969745619562064, + "learning_rate": 2.7549115610108056e-06, + "loss": 0.9466, + "step": 7955 + }, + { + "epoch": 0.7652943439784533, + "grad_norm": 2.305204942320966, + "learning_rate": 2.752764558971517e-06, + "loss": 0.9177, + "step": 7956 + }, + { + "epoch": 0.765390534821085, + "grad_norm": 2.1029679816917923, + "learning_rate": 2.75061826033883e-06, + "loss": 0.9016, + "step": 7957 + }, + { + "epoch": 0.7654867256637168, + "grad_norm": 1.8517295247457637, + "learning_rate": 2.748472665321056e-06, + "loss": 0.8683, + "step": 7958 + }, + { + "epoch": 0.7655829165063486, + "grad_norm": 2.129615569497362, + "learning_rate": 2.7463277741264493e-06, + "loss": 0.8621, + "step": 7959 + }, + { + "epoch": 0.7656791073489804, + "grad_norm": 1.9982457435261693, + "learning_rate": 2.744183586963185e-06, + "loss": 0.8866, + "step": 7960 + }, + { + "epoch": 0.7657752981916122, + "grad_norm": 1.805110279913455, + "learning_rate": 2.7420401040393785e-06, + "loss": 0.7257, + "step": 7961 + }, + { + "epoch": 0.765871489034244, + "grad_norm": 2.0579757285349327, + "learning_rate": 2.739897325563069e-06, + "loss": 0.9582, + "step": 7962 + }, + { + "epoch": 0.7659676798768758, + "grad_norm": 2.0264416672497916, + "learning_rate": 2.7377552517422345e-06, + "loss": 0.8724, + "step": 7963 + }, + { + "epoch": 0.7660638707195075, + "grad_norm": 2.122686164044951, + "learning_rate": 2.7356138827847856e-06, + "loss": 0.9436, + "step": 7964 + }, + { + "epoch": 0.7661600615621393, + "grad_norm": 2.01335922761843, + "learning_rate": 2.733473218898555e-06, + "loss": 0.9206, + "step": 7965 + }, + { + "epoch": 0.766256252404771, + "grad_norm": 1.9707845340191497, + "learning_rate": 2.731333260291311e-06, + "loss": 0.9552, + "step": 7966 + }, + { + "epoch": 0.7663524432474028, + "grad_norm": 2.01295423064342, + "learning_rate": 2.729194007170759e-06, + "loss": 0.8965, + "step": 7967 + }, + { + "epoch": 0.7664486340900346, + "grad_norm": 2.162865200208241, + "learning_rate": 2.7270554597445343e-06, + "loss": 0.8528, + "step": 7968 + }, + { + "epoch": 0.7665448249326664, + "grad_norm": 2.0964334964384883, + "learning_rate": 2.7249176182201944e-06, + "loss": 0.9274, + "step": 7969 + }, + { + "epoch": 0.7666410157752982, + "grad_norm": 2.088828567766474, + "learning_rate": 2.7227804828052384e-06, + "loss": 0.9759, + "step": 7970 + }, + { + "epoch": 0.76673720661793, + "grad_norm": 2.449926729499435, + "learning_rate": 2.720644053707101e-06, + "loss": 0.9934, + "step": 7971 + }, + { + "epoch": 0.7668333974605618, + "grad_norm": 2.0032129412300868, + "learning_rate": 2.7185083311331283e-06, + "loss": 0.9516, + "step": 7972 + }, + { + "epoch": 0.7669295883031936, + "grad_norm": 2.0455116996612617, + "learning_rate": 2.716373315290616e-06, + "loss": 0.9508, + "step": 7973 + }, + { + "epoch": 0.7670257791458254, + "grad_norm": 2.097028756184354, + "learning_rate": 2.7142390063867896e-06, + "loss": 1.0561, + "step": 7974 + }, + { + "epoch": 0.767121969988457, + "grad_norm": 1.9715414372490863, + "learning_rate": 2.7121054046287953e-06, + "loss": 0.8317, + "step": 7975 + }, + { + "epoch": 0.7672181608310888, + "grad_norm": 1.994309273900705, + "learning_rate": 2.709972510223725e-06, + "loss": 0.8526, + "step": 7976 + }, + { + "epoch": 0.7673143516737206, + "grad_norm": 2.2146274011837432, + "learning_rate": 2.7078403233785864e-06, + "loss": 0.9408, + "step": 7977 + }, + { + "epoch": 0.7674105425163524, + "grad_norm": 2.0216317456560957, + "learning_rate": 2.7057088443003343e-06, + "loss": 0.8149, + "step": 7978 + }, + { + "epoch": 0.7675067333589842, + "grad_norm": 1.8429432851604484, + "learning_rate": 2.7035780731958406e-06, + "loss": 0.94, + "step": 7979 + }, + { + "epoch": 0.767602924201616, + "grad_norm": 1.9920964176421778, + "learning_rate": 2.7014480102719174e-06, + "loss": 0.8921, + "step": 7980 + }, + { + "epoch": 0.7676991150442478, + "grad_norm": 1.9348179216745631, + "learning_rate": 2.699318655735309e-06, + "loss": 0.9319, + "step": 7981 + }, + { + "epoch": 0.7677953058868796, + "grad_norm": 1.916487994832898, + "learning_rate": 2.697190009792685e-06, + "loss": 0.9377, + "step": 7982 + }, + { + "epoch": 0.7678914967295114, + "grad_norm": 1.9921087313219368, + "learning_rate": 2.695062072650646e-06, + "loss": 0.7947, + "step": 7983 + }, + { + "epoch": 0.7679876875721431, + "grad_norm": 1.8033283417031154, + "learning_rate": 2.692934844515729e-06, + "loss": 0.8897, + "step": 7984 + }, + { + "epoch": 0.7680838784147749, + "grad_norm": 2.1716697019858455, + "learning_rate": 2.690808325594403e-06, + "loss": 0.851, + "step": 7985 + }, + { + "epoch": 0.7681800692574067, + "grad_norm": 1.8752576052179482, + "learning_rate": 2.6886825160930587e-06, + "loss": 0.8219, + "step": 7986 + }, + { + "epoch": 0.7682762601000385, + "grad_norm": 1.9580753400493982, + "learning_rate": 2.6865574162180262e-06, + "loss": 0.8606, + "step": 7987 + }, + { + "epoch": 0.7683724509426703, + "grad_norm": 2.0444779793493324, + "learning_rate": 2.6844330261755715e-06, + "loss": 0.9883, + "step": 7988 + }, + { + "epoch": 0.768468641785302, + "grad_norm": 1.939663521442891, + "learning_rate": 2.6823093461718773e-06, + "loss": 0.8431, + "step": 7989 + }, + { + "epoch": 0.7685648326279338, + "grad_norm": 1.9476913341072546, + "learning_rate": 2.6801863764130653e-06, + "loss": 0.7451, + "step": 7990 + }, + { + "epoch": 0.7686610234705656, + "grad_norm": 1.8976218356581214, + "learning_rate": 2.678064117105189e-06, + "loss": 0.9227, + "step": 7991 + }, + { + "epoch": 0.7687572143131974, + "grad_norm": 1.8955240613764262, + "learning_rate": 2.675942568454236e-06, + "loss": 0.886, + "step": 7992 + }, + { + "epoch": 0.7688534051558291, + "grad_norm": 2.1673465150485263, + "learning_rate": 2.673821730666115e-06, + "loss": 0.895, + "step": 7993 + }, + { + "epoch": 0.7689495959984609, + "grad_norm": 1.942971606428182, + "learning_rate": 2.671701603946678e-06, + "loss": 0.942, + "step": 7994 + }, + { + "epoch": 0.7690457868410927, + "grad_norm": 2.2571957834990335, + "learning_rate": 2.669582188501697e-06, + "loss": 0.9608, + "step": 7995 + }, + { + "epoch": 0.7691419776837245, + "grad_norm": 2.5214341695397073, + "learning_rate": 2.667463484536876e-06, + "loss": 0.907, + "step": 7996 + }, + { + "epoch": 0.7692381685263563, + "grad_norm": 2.262975041756148, + "learning_rate": 2.6653454922578593e-06, + "loss": 0.9152, + "step": 7997 + }, + { + "epoch": 0.7693343593689881, + "grad_norm": 2.2167625687554366, + "learning_rate": 2.6632282118702147e-06, + "loss": 0.9094, + "step": 7998 + }, + { + "epoch": 0.7694305502116199, + "grad_norm": 1.777180478576892, + "learning_rate": 2.661111643579445e-06, + "loss": 0.814, + "step": 7999 + }, + { + "epoch": 0.7695267410542517, + "grad_norm": 2.3041668479353707, + "learning_rate": 2.65899578759098e-06, + "loss": 0.9027, + "step": 8000 + }, + { + "epoch": 0.7696229318968835, + "grad_norm": 1.807815616769361, + "learning_rate": 2.6568806441101757e-06, + "loss": 0.9198, + "step": 8001 + }, + { + "epoch": 0.7697191227395151, + "grad_norm": 2.081313978092258, + "learning_rate": 2.654766213342335e-06, + "loss": 0.933, + "step": 8002 + }, + { + "epoch": 0.7698153135821469, + "grad_norm": 1.905184935187157, + "learning_rate": 2.6526524954926727e-06, + "loss": 0.8539, + "step": 8003 + }, + { + "epoch": 0.7699115044247787, + "grad_norm": 1.4535498099340167, + "learning_rate": 2.650539490766346e-06, + "loss": 0.7202, + "step": 8004 + }, + { + "epoch": 0.7700076952674105, + "grad_norm": 1.9370423113454944, + "learning_rate": 2.6484271993684462e-06, + "loss": 0.9741, + "step": 8005 + }, + { + "epoch": 0.7701038861100423, + "grad_norm": 1.9974976819355965, + "learning_rate": 2.646315621503983e-06, + "loss": 0.8967, + "step": 8006 + }, + { + "epoch": 0.7702000769526741, + "grad_norm": 1.9006020782860642, + "learning_rate": 2.6442047573779027e-06, + "loss": 0.9055, + "step": 8007 + }, + { + "epoch": 0.7702962677953059, + "grad_norm": 2.0208434785475387, + "learning_rate": 2.642094607195085e-06, + "loss": 0.9815, + "step": 8008 + }, + { + "epoch": 0.7703924586379377, + "grad_norm": 1.9951705358616234, + "learning_rate": 2.6399851711603406e-06, + "loss": 0.8677, + "step": 8009 + }, + { + "epoch": 0.7704886494805695, + "grad_norm": 1.9106670018796195, + "learning_rate": 2.6378764494784027e-06, + "loss": 0.8903, + "step": 8010 + }, + { + "epoch": 0.7705848403232012, + "grad_norm": 1.8273703099728633, + "learning_rate": 2.6357684423539474e-06, + "loss": 0.9006, + "step": 8011 + }, + { + "epoch": 0.770681031165833, + "grad_norm": 2.2135809119043297, + "learning_rate": 2.633661149991569e-06, + "loss": 0.9884, + "step": 8012 + }, + { + "epoch": 0.7707772220084648, + "grad_norm": 1.8319734864901835, + "learning_rate": 2.6315545725958024e-06, + "loss": 0.8388, + "step": 8013 + }, + { + "epoch": 0.7708734128510966, + "grad_norm": 1.969196923877019, + "learning_rate": 2.6294487103711064e-06, + "loss": 0.9006, + "step": 8014 + }, + { + "epoch": 0.7709696036937284, + "grad_norm": 2.0316441129043707, + "learning_rate": 2.6273435635218735e-06, + "loss": 0.8553, + "step": 8015 + }, + { + "epoch": 0.7710657945363601, + "grad_norm": 1.9210532568325258, + "learning_rate": 2.6252391322524297e-06, + "loss": 0.8796, + "step": 8016 + }, + { + "epoch": 0.7711619853789919, + "grad_norm": 2.3652332195462, + "learning_rate": 2.6231354167670265e-06, + "loss": 0.8408, + "step": 8017 + }, + { + "epoch": 0.7712581762216237, + "grad_norm": 1.975047708067743, + "learning_rate": 2.6210324172698432e-06, + "loss": 0.9831, + "step": 8018 + }, + { + "epoch": 0.7713543670642555, + "grad_norm": 1.9170265857714255, + "learning_rate": 2.6189301339649975e-06, + "loss": 0.8232, + "step": 8019 + }, + { + "epoch": 0.7714505579068872, + "grad_norm": 1.879123781292071, + "learning_rate": 2.6168285670565374e-06, + "loss": 0.9076, + "step": 8020 + }, + { + "epoch": 0.771546748749519, + "grad_norm": 2.0715064874648372, + "learning_rate": 2.614727716748432e-06, + "loss": 0.9275, + "step": 8021 + }, + { + "epoch": 0.7716429395921508, + "grad_norm": 2.2307117812386825, + "learning_rate": 2.6126275832445892e-06, + "loss": 0.8471, + "step": 8022 + }, + { + "epoch": 0.7717391304347826, + "grad_norm": 2.095360774021156, + "learning_rate": 2.6105281667488514e-06, + "loss": 0.9937, + "step": 8023 + }, + { + "epoch": 0.7718353212774144, + "grad_norm": 1.6558873107876426, + "learning_rate": 2.6084294674649734e-06, + "loss": 0.7818, + "step": 8024 + }, + { + "epoch": 0.7719315121200462, + "grad_norm": 2.0491959788884415, + "learning_rate": 2.606331485596657e-06, + "loss": 0.8988, + "step": 8025 + }, + { + "epoch": 0.772027702962678, + "grad_norm": 1.9007483157473422, + "learning_rate": 2.6042342213475346e-06, + "loss": 0.9526, + "step": 8026 + }, + { + "epoch": 0.7721238938053098, + "grad_norm": 2.106522401180427, + "learning_rate": 2.6021376749211556e-06, + "loss": 0.9308, + "step": 8027 + }, + { + "epoch": 0.7722200846479416, + "grad_norm": 2.1873799534920444, + "learning_rate": 2.6000418465210143e-06, + "loss": 0.9107, + "step": 8028 + }, + { + "epoch": 0.7723162754905732, + "grad_norm": 2.069467253235778, + "learning_rate": 2.597946736350524e-06, + "loss": 0.9072, + "step": 8029 + }, + { + "epoch": 0.772412466333205, + "grad_norm": 1.84626070169988, + "learning_rate": 2.595852344613038e-06, + "loss": 0.9384, + "step": 8030 + }, + { + "epoch": 0.7725086571758368, + "grad_norm": 1.9457590673567458, + "learning_rate": 2.593758671511829e-06, + "loss": 0.9892, + "step": 8031 + }, + { + "epoch": 0.7726048480184686, + "grad_norm": 1.9864374521851051, + "learning_rate": 2.5916657172501103e-06, + "loss": 0.9114, + "step": 8032 + }, + { + "epoch": 0.7727010388611004, + "grad_norm": 2.2937165472319516, + "learning_rate": 2.5895734820310225e-06, + "loss": 0.8836, + "step": 8033 + }, + { + "epoch": 0.7727972297037322, + "grad_norm": 2.3052407662458196, + "learning_rate": 2.587481966057633e-06, + "loss": 0.9419, + "step": 8034 + }, + { + "epoch": 0.772893420546364, + "grad_norm": 1.9188021623480707, + "learning_rate": 2.585391169532939e-06, + "loss": 0.9499, + "step": 8035 + }, + { + "epoch": 0.7729896113889958, + "grad_norm": 2.0053144919844508, + "learning_rate": 2.583301092659872e-06, + "loss": 0.7897, + "step": 8036 + }, + { + "epoch": 0.7730858022316276, + "grad_norm": 1.7066329033438088, + "learning_rate": 2.5812117356412957e-06, + "loss": 0.8587, + "step": 8037 + }, + { + "epoch": 0.7731819930742593, + "grad_norm": 2.0009605985058565, + "learning_rate": 2.5791230986799944e-06, + "loss": 0.9255, + "step": 8038 + }, + { + "epoch": 0.7732781839168911, + "grad_norm": 1.9083206703283664, + "learning_rate": 2.5770351819786908e-06, + "loss": 0.8862, + "step": 8039 + }, + { + "epoch": 0.7733743747595229, + "grad_norm": 1.8852039222745542, + "learning_rate": 2.5749479857400383e-06, + "loss": 0.9236, + "step": 8040 + }, + { + "epoch": 0.7734705656021547, + "grad_norm": 2.354865018564687, + "learning_rate": 2.5728615101666134e-06, + "loss": 0.8613, + "step": 8041 + }, + { + "epoch": 0.7735667564447865, + "grad_norm": 2.053816019842643, + "learning_rate": 2.5707757554609247e-06, + "loss": 0.9892, + "step": 8042 + }, + { + "epoch": 0.7736629472874182, + "grad_norm": 1.956648839014291, + "learning_rate": 2.5686907218254164e-06, + "loss": 0.9648, + "step": 8043 + }, + { + "epoch": 0.77375913813005, + "grad_norm": 2.0640197801821802, + "learning_rate": 2.56660640946246e-06, + "loss": 0.9242, + "step": 8044 + }, + { + "epoch": 0.7738553289726818, + "grad_norm": 2.060344781845763, + "learning_rate": 2.5645228185743507e-06, + "loss": 0.8713, + "step": 8045 + }, + { + "epoch": 0.7739515198153136, + "grad_norm": 1.8362820027315916, + "learning_rate": 2.5624399493633257e-06, + "loss": 0.894, + "step": 8046 + }, + { + "epoch": 0.7740477106579453, + "grad_norm": 1.87236437919069, + "learning_rate": 2.560357802031539e-06, + "loss": 0.9087, + "step": 8047 + }, + { + "epoch": 0.7741439015005771, + "grad_norm": 1.8961983804903417, + "learning_rate": 2.558276376781086e-06, + "loss": 0.8794, + "step": 8048 + }, + { + "epoch": 0.7742400923432089, + "grad_norm": 2.249719303398439, + "learning_rate": 2.5561956738139814e-06, + "loss": 0.9625, + "step": 8049 + }, + { + "epoch": 0.7743362831858407, + "grad_norm": 1.9846552947026028, + "learning_rate": 2.55411569333218e-06, + "loss": 0.861, + "step": 8050 + }, + { + "epoch": 0.7744324740284725, + "grad_norm": 2.0838002952663355, + "learning_rate": 2.552036435537565e-06, + "loss": 0.9448, + "step": 8051 + }, + { + "epoch": 0.7745286648711043, + "grad_norm": 2.1176814953223793, + "learning_rate": 2.5499579006319365e-06, + "loss": 0.9592, + "step": 8052 + }, + { + "epoch": 0.7746248557137361, + "grad_norm": 1.9924580413650679, + "learning_rate": 2.5478800888170395e-06, + "loss": 0.8266, + "step": 8053 + }, + { + "epoch": 0.7747210465563679, + "grad_norm": 1.8449993426828168, + "learning_rate": 2.5458030002945457e-06, + "loss": 0.9052, + "step": 8054 + }, + { + "epoch": 0.7748172373989997, + "grad_norm": 2.253265518999058, + "learning_rate": 2.5437266352660493e-06, + "loss": 0.9344, + "step": 8055 + }, + { + "epoch": 0.7749134282416313, + "grad_norm": 1.9908191285922205, + "learning_rate": 2.5416509939330836e-06, + "loss": 0.9775, + "step": 8056 + }, + { + "epoch": 0.7750096190842631, + "grad_norm": 1.7126376155638297, + "learning_rate": 2.539576076497108e-06, + "loss": 0.9124, + "step": 8057 + }, + { + "epoch": 0.7751058099268949, + "grad_norm": 2.001543037663691, + "learning_rate": 2.537501883159509e-06, + "loss": 0.8759, + "step": 8058 + }, + { + "epoch": 0.7752020007695267, + "grad_norm": 1.4693434846045965, + "learning_rate": 2.5354284141216024e-06, + "loss": 0.7267, + "step": 8059 + }, + { + "epoch": 0.7752981916121585, + "grad_norm": 1.979807284277813, + "learning_rate": 2.5333556695846384e-06, + "loss": 0.9681, + "step": 8060 + }, + { + "epoch": 0.7753943824547903, + "grad_norm": 2.2782014147646135, + "learning_rate": 2.5312836497497996e-06, + "loss": 0.8116, + "step": 8061 + }, + { + "epoch": 0.7754905732974221, + "grad_norm": 2.2080015392502204, + "learning_rate": 2.5292123548181847e-06, + "loss": 0.9388, + "step": 8062 + }, + { + "epoch": 0.7755867641400539, + "grad_norm": 1.7338335686272428, + "learning_rate": 2.5271417849908387e-06, + "loss": 0.7569, + "step": 8063 + }, + { + "epoch": 0.7756829549826857, + "grad_norm": 2.1844824938000627, + "learning_rate": 2.525071940468722e-06, + "loss": 0.872, + "step": 8064 + }, + { + "epoch": 0.7757791458253175, + "grad_norm": 2.1167638849115877, + "learning_rate": 2.523002821452736e-06, + "loss": 1.0267, + "step": 8065 + }, + { + "epoch": 0.7758753366679492, + "grad_norm": 2.2774862267293368, + "learning_rate": 2.520934428143701e-06, + "loss": 0.951, + "step": 8066 + }, + { + "epoch": 0.775971527510581, + "grad_norm": 2.123289553160881, + "learning_rate": 2.5188667607423755e-06, + "loss": 0.9735, + "step": 8067 + }, + { + "epoch": 0.7760677183532128, + "grad_norm": 1.933822677850637, + "learning_rate": 2.5167998194494468e-06, + "loss": 0.8984, + "step": 8068 + }, + { + "epoch": 0.7761639091958445, + "grad_norm": 1.9342179672258368, + "learning_rate": 2.514733604465527e-06, + "loss": 0.7894, + "step": 8069 + }, + { + "epoch": 0.7762601000384763, + "grad_norm": 2.0025501378998953, + "learning_rate": 2.5126681159911558e-06, + "loss": 0.929, + "step": 8070 + }, + { + "epoch": 0.7763562908811081, + "grad_norm": 1.7903726062490233, + "learning_rate": 2.510603354226813e-06, + "loss": 0.7933, + "step": 8071 + }, + { + "epoch": 0.7764524817237399, + "grad_norm": 1.9300031096516654, + "learning_rate": 2.5085393193729e-06, + "loss": 0.8564, + "step": 8072 + }, + { + "epoch": 0.7765486725663717, + "grad_norm": 2.0370688415658416, + "learning_rate": 2.506476011629746e-06, + "loss": 0.9424, + "step": 8073 + }, + { + "epoch": 0.7766448634090035, + "grad_norm": 1.7946864664607745, + "learning_rate": 2.5044134311976156e-06, + "loss": 0.848, + "step": 8074 + }, + { + "epoch": 0.7767410542516352, + "grad_norm": 2.404187632454082, + "learning_rate": 2.5023515782767048e-06, + "loss": 0.9133, + "step": 8075 + }, + { + "epoch": 0.776837245094267, + "grad_norm": 2.0964724534258607, + "learning_rate": 2.5002904530671236e-06, + "loss": 1.0169, + "step": 8076 + }, + { + "epoch": 0.7769334359368988, + "grad_norm": 2.188726763031533, + "learning_rate": 2.4982300557689267e-06, + "loss": 1.005, + "step": 8077 + }, + { + "epoch": 0.7770296267795306, + "grad_norm": 1.9309469333137566, + "learning_rate": 2.4961703865820974e-06, + "loss": 0.9203, + "step": 8078 + }, + { + "epoch": 0.7771258176221624, + "grad_norm": 1.772416166357763, + "learning_rate": 2.4941114457065376e-06, + "loss": 0.7819, + "step": 8079 + }, + { + "epoch": 0.7772220084647942, + "grad_norm": 2.017647597645238, + "learning_rate": 2.492053233342091e-06, + "loss": 0.9058, + "step": 8080 + }, + { + "epoch": 0.777318199307426, + "grad_norm": 1.9086098615689515, + "learning_rate": 2.4899957496885196e-06, + "loss": 0.7956, + "step": 8081 + }, + { + "epoch": 0.7774143901500578, + "grad_norm": 1.9558367330710464, + "learning_rate": 2.487938994945527e-06, + "loss": 0.8892, + "step": 8082 + }, + { + "epoch": 0.7775105809926895, + "grad_norm": 2.079781053566446, + "learning_rate": 2.4858829693127294e-06, + "loss": 0.8669, + "step": 8083 + }, + { + "epoch": 0.7776067718353212, + "grad_norm": 2.0561664821045356, + "learning_rate": 2.4838276729896884e-06, + "loss": 0.8877, + "step": 8084 + }, + { + "epoch": 0.777702962677953, + "grad_norm": 2.023839400121962, + "learning_rate": 2.48177310617589e-06, + "loss": 0.9132, + "step": 8085 + }, + { + "epoch": 0.7777991535205848, + "grad_norm": 1.8575141449962356, + "learning_rate": 2.479719269070743e-06, + "loss": 0.9225, + "step": 8086 + }, + { + "epoch": 0.7778953443632166, + "grad_norm": 1.4874933011304148, + "learning_rate": 2.47766616187359e-06, + "loss": 0.7488, + "step": 8087 + }, + { + "epoch": 0.7779915352058484, + "grad_norm": 2.3273750751743236, + "learning_rate": 2.4756137847837025e-06, + "loss": 0.9037, + "step": 8088 + }, + { + "epoch": 0.7780877260484802, + "grad_norm": 1.8252807131421986, + "learning_rate": 2.473562138000287e-06, + "loss": 0.9808, + "step": 8089 + }, + { + "epoch": 0.778183916891112, + "grad_norm": 1.9590774357390326, + "learning_rate": 2.4715112217224657e-06, + "loss": 0.9103, + "step": 8090 + }, + { + "epoch": 0.7782801077337438, + "grad_norm": 2.202499650859165, + "learning_rate": 2.469461036149302e-06, + "loss": 0.8877, + "step": 8091 + }, + { + "epoch": 0.7783762985763756, + "grad_norm": 2.0531480175819596, + "learning_rate": 2.467411581479786e-06, + "loss": 0.9167, + "step": 8092 + }, + { + "epoch": 0.7784724894190073, + "grad_norm": 1.9122679034131194, + "learning_rate": 2.465362857912833e-06, + "loss": 0.8249, + "step": 8093 + }, + { + "epoch": 0.7785686802616391, + "grad_norm": 2.040789033465442, + "learning_rate": 2.463314865647286e-06, + "loss": 0.9547, + "step": 8094 + }, + { + "epoch": 0.7786648711042709, + "grad_norm": 2.140419961976528, + "learning_rate": 2.4612676048819217e-06, + "loss": 0.9591, + "step": 8095 + }, + { + "epoch": 0.7787610619469026, + "grad_norm": 1.7570362081135233, + "learning_rate": 2.45922107581545e-06, + "loss": 0.8223, + "step": 8096 + }, + { + "epoch": 0.7788572527895344, + "grad_norm": 1.8315840978341615, + "learning_rate": 2.4571752786464965e-06, + "loss": 0.8725, + "step": 8097 + }, + { + "epoch": 0.7789534436321662, + "grad_norm": 2.4063759721251237, + "learning_rate": 2.4551302135736287e-06, + "loss": 0.8736, + "step": 8098 + }, + { + "epoch": 0.779049634474798, + "grad_norm": 1.9186648501049124, + "learning_rate": 2.4530858807953338e-06, + "loss": 0.9152, + "step": 8099 + }, + { + "epoch": 0.7791458253174298, + "grad_norm": 2.2314563536247363, + "learning_rate": 2.4510422805100366e-06, + "loss": 0.9743, + "step": 8100 + }, + { + "epoch": 0.7792420161600616, + "grad_norm": 2.282695095384544, + "learning_rate": 2.44899941291608e-06, + "loss": 0.8351, + "step": 8101 + }, + { + "epoch": 0.7793382070026933, + "grad_norm": 1.9813370494633575, + "learning_rate": 2.446957278211746e-06, + "loss": 0.9884, + "step": 8102 + }, + { + "epoch": 0.7794343978453251, + "grad_norm": 2.068020971579037, + "learning_rate": 2.444915876595246e-06, + "loss": 1.0014, + "step": 8103 + }, + { + "epoch": 0.7795305886879569, + "grad_norm": 2.160826990369233, + "learning_rate": 2.4428752082647044e-06, + "loss": 0.9851, + "step": 8104 + }, + { + "epoch": 0.7796267795305887, + "grad_norm": 2.2462470032203465, + "learning_rate": 2.440835273418193e-06, + "loss": 0.8526, + "step": 8105 + }, + { + "epoch": 0.7797229703732205, + "grad_norm": 2.0277713173222036, + "learning_rate": 2.438796072253704e-06, + "loss": 1.0172, + "step": 8106 + }, + { + "epoch": 0.7798191612158523, + "grad_norm": 1.936221838414029, + "learning_rate": 2.436757604969158e-06, + "loss": 0.9071, + "step": 8107 + }, + { + "epoch": 0.7799153520584841, + "grad_norm": 1.6216303127137697, + "learning_rate": 2.4347198717624054e-06, + "loss": 0.7941, + "step": 8108 + }, + { + "epoch": 0.7800115429011159, + "grad_norm": 2.395572807023616, + "learning_rate": 2.4326828728312313e-06, + "loss": 0.9414, + "step": 8109 + }, + { + "epoch": 0.7801077337437476, + "grad_norm": 2.2858623608000945, + "learning_rate": 2.4306466083733392e-06, + "loss": 0.8026, + "step": 8110 + }, + { + "epoch": 0.7802039245863793, + "grad_norm": 1.6331764483062605, + "learning_rate": 2.4286110785863637e-06, + "loss": 0.8585, + "step": 8111 + }, + { + "epoch": 0.7803001154290111, + "grad_norm": 2.0636973213215213, + "learning_rate": 2.426576283667873e-06, + "loss": 0.9415, + "step": 8112 + }, + { + "epoch": 0.7803963062716429, + "grad_norm": 2.0212185452268616, + "learning_rate": 2.424542223815366e-06, + "loss": 0.8042, + "step": 8113 + }, + { + "epoch": 0.7804924971142747, + "grad_norm": 2.158564174821542, + "learning_rate": 2.422508899226258e-06, + "loss": 0.9192, + "step": 8114 + }, + { + "epoch": 0.7805886879569065, + "grad_norm": 1.9885095365667642, + "learning_rate": 2.420476310097908e-06, + "loss": 0.8056, + "step": 8115 + }, + { + "epoch": 0.7806848787995383, + "grad_norm": 2.0502956687394667, + "learning_rate": 2.418444456627589e-06, + "loss": 0.8438, + "step": 8116 + }, + { + "epoch": 0.7807810696421701, + "grad_norm": 1.6152906292560474, + "learning_rate": 2.4164133390125167e-06, + "loss": 0.7812, + "step": 8117 + }, + { + "epoch": 0.7808772604848019, + "grad_norm": 2.2067916532970866, + "learning_rate": 2.4143829574498224e-06, + "loss": 0.9121, + "step": 8118 + }, + { + "epoch": 0.7809734513274337, + "grad_norm": 1.8924040347672213, + "learning_rate": 2.4123533121365748e-06, + "loss": 0.9387, + "step": 8119 + }, + { + "epoch": 0.7810696421700654, + "grad_norm": 1.969193159273031, + "learning_rate": 2.4103244032697717e-06, + "loss": 0.9761, + "step": 8120 + }, + { + "epoch": 0.7811658330126972, + "grad_norm": 2.101386910454979, + "learning_rate": 2.408296231046333e-06, + "loss": 0.9745, + "step": 8121 + }, + { + "epoch": 0.781262023855329, + "grad_norm": 2.3315858743774434, + "learning_rate": 2.406268795663108e-06, + "loss": 0.7368, + "step": 8122 + }, + { + "epoch": 0.7813582146979607, + "grad_norm": 1.8765676156469255, + "learning_rate": 2.4042420973168788e-06, + "loss": 0.8739, + "step": 8123 + }, + { + "epoch": 0.7814544055405925, + "grad_norm": 1.8195099366121947, + "learning_rate": 2.4022161362043574e-06, + "loss": 0.889, + "step": 8124 + }, + { + "epoch": 0.7815505963832243, + "grad_norm": 2.0775644318810187, + "learning_rate": 2.4001909125221746e-06, + "loss": 0.9167, + "step": 8125 + }, + { + "epoch": 0.7816467872258561, + "grad_norm": 1.7187688131438343, + "learning_rate": 2.3981664264669025e-06, + "loss": 0.8579, + "step": 8126 + }, + { + "epoch": 0.7817429780684879, + "grad_norm": 1.9110810197570471, + "learning_rate": 2.3961426782350307e-06, + "loss": 0.9781, + "step": 8127 + }, + { + "epoch": 0.7818391689111197, + "grad_norm": 1.7318682567791512, + "learning_rate": 2.3941196680229794e-06, + "loss": 0.824, + "step": 8128 + }, + { + "epoch": 0.7819353597537514, + "grad_norm": 2.2986222842171182, + "learning_rate": 2.3920973960271023e-06, + "loss": 0.8933, + "step": 8129 + }, + { + "epoch": 0.7820315505963832, + "grad_norm": 1.697702328052704, + "learning_rate": 2.3900758624436772e-06, + "loss": 0.8913, + "step": 8130 + }, + { + "epoch": 0.782127741439015, + "grad_norm": 1.899833355432147, + "learning_rate": 2.388055067468914e-06, + "loss": 0.908, + "step": 8131 + }, + { + "epoch": 0.7822239322816468, + "grad_norm": 1.7402898603482804, + "learning_rate": 2.3860350112989473e-06, + "loss": 0.8186, + "step": 8132 + }, + { + "epoch": 0.7823201231242786, + "grad_norm": 1.9888788747783281, + "learning_rate": 2.3840156941298365e-06, + "loss": 1.0212, + "step": 8133 + }, + { + "epoch": 0.7824163139669104, + "grad_norm": 2.2859827763682534, + "learning_rate": 2.3819971161575807e-06, + "loss": 0.9762, + "step": 8134 + }, + { + "epoch": 0.7825125048095422, + "grad_norm": 2.142195541148097, + "learning_rate": 2.379979277578093e-06, + "loss": 1.0219, + "step": 8135 + }, + { + "epoch": 0.782608695652174, + "grad_norm": 2.100641842161352, + "learning_rate": 2.3779621785872252e-06, + "loss": 0.8441, + "step": 8136 + }, + { + "epoch": 0.7827048864948057, + "grad_norm": 1.796527518879271, + "learning_rate": 2.3759458193807573e-06, + "loss": 0.8826, + "step": 8137 + }, + { + "epoch": 0.7828010773374374, + "grad_norm": 2.035849129254467, + "learning_rate": 2.3739302001543918e-06, + "loss": 0.9752, + "step": 8138 + }, + { + "epoch": 0.7828972681800692, + "grad_norm": 2.270668995691737, + "learning_rate": 2.3719153211037595e-06, + "loss": 0.8852, + "step": 8139 + }, + { + "epoch": 0.782993459022701, + "grad_norm": 1.9822064518796962, + "learning_rate": 2.3699011824244234e-06, + "loss": 0.8282, + "step": 8140 + }, + { + "epoch": 0.7830896498653328, + "grad_norm": 1.9430008803919419, + "learning_rate": 2.3678877843118762e-06, + "loss": 0.8291, + "step": 8141 + }, + { + "epoch": 0.7831858407079646, + "grad_norm": 1.9389434976741509, + "learning_rate": 2.365875126961531e-06, + "loss": 0.8512, + "step": 8142 + }, + { + "epoch": 0.7832820315505964, + "grad_norm": 2.2335636871270093, + "learning_rate": 2.3638632105687344e-06, + "loss": 1.0092, + "step": 8143 + }, + { + "epoch": 0.7833782223932282, + "grad_norm": 1.7929253017329376, + "learning_rate": 2.3618520353287644e-06, + "loss": 0.7918, + "step": 8144 + }, + { + "epoch": 0.78347441323586, + "grad_norm": 2.023984901304435, + "learning_rate": 2.3598416014368186e-06, + "loss": 0.8906, + "step": 8145 + }, + { + "epoch": 0.7835706040784918, + "grad_norm": 1.8857637500148696, + "learning_rate": 2.3578319090880263e-06, + "loss": 0.9889, + "step": 8146 + }, + { + "epoch": 0.7836667949211235, + "grad_norm": 1.8425403272240506, + "learning_rate": 2.355822958477447e-06, + "loss": 0.8642, + "step": 8147 + }, + { + "epoch": 0.7837629857637552, + "grad_norm": 2.0733425350379155, + "learning_rate": 2.3538147498000695e-06, + "loss": 0.9374, + "step": 8148 + }, + { + "epoch": 0.783859176606387, + "grad_norm": 2.094088826430377, + "learning_rate": 2.351807283250802e-06, + "loss": 0.8839, + "step": 8149 + }, + { + "epoch": 0.7839553674490188, + "grad_norm": 2.206257242092516, + "learning_rate": 2.349800559024492e-06, + "loss": 0.917, + "step": 8150 + }, + { + "epoch": 0.7840515582916506, + "grad_norm": 2.0503692608730617, + "learning_rate": 2.347794577315904e-06, + "loss": 0.8964, + "step": 8151 + }, + { + "epoch": 0.7841477491342824, + "grad_norm": 2.1186654343833333, + "learning_rate": 2.3457893383197415e-06, + "loss": 0.8822, + "step": 8152 + }, + { + "epoch": 0.7842439399769142, + "grad_norm": 1.861879304790957, + "learning_rate": 2.3437848422306252e-06, + "loss": 0.7496, + "step": 8153 + }, + { + "epoch": 0.784340130819546, + "grad_norm": 2.05032033555599, + "learning_rate": 2.3417810892431104e-06, + "loss": 0.898, + "step": 8154 + }, + { + "epoch": 0.7844363216621778, + "grad_norm": 1.7973633230570585, + "learning_rate": 2.3397780795516847e-06, + "loss": 0.8084, + "step": 8155 + }, + { + "epoch": 0.7845325125048095, + "grad_norm": 2.0869509249070686, + "learning_rate": 2.3377758133507455e-06, + "loss": 0.8526, + "step": 8156 + }, + { + "epoch": 0.7846287033474413, + "grad_norm": 1.7179353749528965, + "learning_rate": 2.335774290834636e-06, + "loss": 0.7617, + "step": 8157 + }, + { + "epoch": 0.7847248941900731, + "grad_norm": 2.0697532409790576, + "learning_rate": 2.3337735121976247e-06, + "loss": 0.8142, + "step": 8158 + }, + { + "epoch": 0.7848210850327049, + "grad_norm": 1.9675958937593754, + "learning_rate": 2.3317734776338984e-06, + "loss": 0.8913, + "step": 8159 + }, + { + "epoch": 0.7849172758753367, + "grad_norm": 1.8066022870625924, + "learning_rate": 2.32977418733758e-06, + "loss": 0.8928, + "step": 8160 + }, + { + "epoch": 0.7850134667179685, + "grad_norm": 2.109369615750123, + "learning_rate": 2.327775641502722e-06, + "loss": 1.0384, + "step": 8161 + }, + { + "epoch": 0.7851096575606002, + "grad_norm": 1.888084074354987, + "learning_rate": 2.3257778403232954e-06, + "loss": 0.9088, + "step": 8162 + }, + { + "epoch": 0.785205848403232, + "grad_norm": 1.8816810817738356, + "learning_rate": 2.3237807839932024e-06, + "loss": 0.9234, + "step": 8163 + }, + { + "epoch": 0.7853020392458638, + "grad_norm": 2.0179448506371647, + "learning_rate": 2.321784472706279e-06, + "loss": 0.9071, + "step": 8164 + }, + { + "epoch": 0.7853982300884956, + "grad_norm": 2.1483467672067706, + "learning_rate": 2.319788906656284e-06, + "loss": 0.862, + "step": 8165 + }, + { + "epoch": 0.7854944209311273, + "grad_norm": 1.913012686703025, + "learning_rate": 2.317794086036901e-06, + "loss": 0.8831, + "step": 8166 + }, + { + "epoch": 0.7855906117737591, + "grad_norm": 1.9454718452338542, + "learning_rate": 2.315800011041749e-06, + "loss": 0.9573, + "step": 8167 + }, + { + "epoch": 0.7856868026163909, + "grad_norm": 1.9461088755682379, + "learning_rate": 2.3138066818643647e-06, + "loss": 0.8638, + "step": 8168 + }, + { + "epoch": 0.7857829934590227, + "grad_norm": 1.9434664997288753, + "learning_rate": 2.3118140986982242e-06, + "loss": 0.9104, + "step": 8169 + }, + { + "epoch": 0.7858791843016545, + "grad_norm": 1.9154505146258927, + "learning_rate": 2.3098222617367184e-06, + "loss": 0.9084, + "step": 8170 + }, + { + "epoch": 0.7859753751442863, + "grad_norm": 2.110784137738965, + "learning_rate": 2.307831171173176e-06, + "loss": 0.8326, + "step": 8171 + }, + { + "epoch": 0.7860715659869181, + "grad_norm": 1.8883850629382528, + "learning_rate": 2.30584082720085e-06, + "loss": 0.8509, + "step": 8172 + }, + { + "epoch": 0.7861677568295499, + "grad_norm": 2.4102052495287976, + "learning_rate": 2.303851230012921e-06, + "loss": 0.9255, + "step": 8173 + }, + { + "epoch": 0.7862639476721817, + "grad_norm": 1.8925884492491036, + "learning_rate": 2.301862379802492e-06, + "loss": 0.7516, + "step": 8174 + }, + { + "epoch": 0.7863601385148133, + "grad_norm": 1.8636102058290394, + "learning_rate": 2.299874276762599e-06, + "loss": 0.9734, + "step": 8175 + }, + { + "epoch": 0.7864563293574451, + "grad_norm": 1.813413036973505, + "learning_rate": 2.297886921086211e-06, + "loss": 0.8804, + "step": 8176 + }, + { + "epoch": 0.7865525202000769, + "grad_norm": 1.9227185354651262, + "learning_rate": 2.2959003129662094e-06, + "loss": 0.9204, + "step": 8177 + }, + { + "epoch": 0.7866487110427087, + "grad_norm": 2.0503892836912443, + "learning_rate": 2.2939144525954194e-06, + "loss": 0.959, + "step": 8178 + }, + { + "epoch": 0.7867449018853405, + "grad_norm": 1.9109054623563728, + "learning_rate": 2.291929340166579e-06, + "loss": 0.8932, + "step": 8179 + }, + { + "epoch": 0.7868410927279723, + "grad_norm": 2.09172939406365, + "learning_rate": 2.2899449758723657e-06, + "loss": 0.9128, + "step": 8180 + }, + { + "epoch": 0.7869372835706041, + "grad_norm": 1.9402546794157933, + "learning_rate": 2.287961359905374e-06, + "loss": 0.81, + "step": 8181 + }, + { + "epoch": 0.7870334744132359, + "grad_norm": 1.8096507311635968, + "learning_rate": 2.285978492458134e-06, + "loss": 0.9751, + "step": 8182 + }, + { + "epoch": 0.7871296652558677, + "grad_norm": 1.997004493861007, + "learning_rate": 2.283996373723103e-06, + "loss": 0.8958, + "step": 8183 + }, + { + "epoch": 0.7872258560984994, + "grad_norm": 2.1440847302126986, + "learning_rate": 2.282015003892659e-06, + "loss": 0.8827, + "step": 8184 + }, + { + "epoch": 0.7873220469411312, + "grad_norm": 2.028118695101041, + "learning_rate": 2.280034383159109e-06, + "loss": 0.8763, + "step": 8185 + }, + { + "epoch": 0.787418237783763, + "grad_norm": 1.948886354593859, + "learning_rate": 2.2780545117146947e-06, + "loss": 0.934, + "step": 8186 + }, + { + "epoch": 0.7875144286263948, + "grad_norm": 1.8489613540084202, + "learning_rate": 2.276075389751574e-06, + "loss": 0.9095, + "step": 8187 + }, + { + "epoch": 0.7876106194690266, + "grad_norm": 1.7209579178491599, + "learning_rate": 2.2740970174618405e-06, + "loss": 1.014, + "step": 8188 + }, + { + "epoch": 0.7877068103116583, + "grad_norm": 1.838006152817569, + "learning_rate": 2.2721193950375155e-06, + "loss": 0.7706, + "step": 8189 + }, + { + "epoch": 0.7878030011542901, + "grad_norm": 1.819268900970195, + "learning_rate": 2.270142522670541e-06, + "loss": 0.8976, + "step": 8190 + }, + { + "epoch": 0.7878991919969219, + "grad_norm": 2.373324097467628, + "learning_rate": 2.268166400552786e-06, + "loss": 0.8972, + "step": 8191 + }, + { + "epoch": 0.7879953828395537, + "grad_norm": 2.0752208288009992, + "learning_rate": 2.2661910288760545e-06, + "loss": 0.9657, + "step": 8192 + }, + { + "epoch": 0.7880915736821854, + "grad_norm": 2.0587521267530997, + "learning_rate": 2.264216407832075e-06, + "loss": 0.8602, + "step": 8193 + }, + { + "epoch": 0.7881877645248172, + "grad_norm": 1.8549346922034413, + "learning_rate": 2.262242537612497e-06, + "loss": 0.8557, + "step": 8194 + }, + { + "epoch": 0.788283955367449, + "grad_norm": 2.124585375431867, + "learning_rate": 2.2602694184089026e-06, + "loss": 0.879, + "step": 8195 + }, + { + "epoch": 0.7883801462100808, + "grad_norm": 2.1144092822309695, + "learning_rate": 2.258297050412804e-06, + "loss": 1.0023, + "step": 8196 + }, + { + "epoch": 0.7884763370527126, + "grad_norm": 2.2288948102673998, + "learning_rate": 2.2563254338156336e-06, + "loss": 0.8387, + "step": 8197 + }, + { + "epoch": 0.7885725278953444, + "grad_norm": 1.8281371135617517, + "learning_rate": 2.254354568808752e-06, + "loss": 0.9237, + "step": 8198 + }, + { + "epoch": 0.7886687187379762, + "grad_norm": 1.9564765307888867, + "learning_rate": 2.2523844555834497e-06, + "loss": 0.8222, + "step": 8199 + }, + { + "epoch": 0.788764909580608, + "grad_norm": 2.7286630253473905, + "learning_rate": 2.2504150943309455e-06, + "loss": 1.0271, + "step": 8200 + }, + { + "epoch": 0.7888611004232398, + "grad_norm": 1.8756703459924173, + "learning_rate": 2.2484464852423826e-06, + "loss": 0.9181, + "step": 8201 + }, + { + "epoch": 0.7889572912658714, + "grad_norm": 1.9879861822373612, + "learning_rate": 2.246478628508827e-06, + "loss": 0.8737, + "step": 8202 + }, + { + "epoch": 0.7890534821085032, + "grad_norm": 2.019397254695081, + "learning_rate": 2.2445115243212776e-06, + "loss": 0.9453, + "step": 8203 + }, + { + "epoch": 0.789149672951135, + "grad_norm": 2.197910856287507, + "learning_rate": 2.242545172870665e-06, + "loss": 0.9917, + "step": 8204 + }, + { + "epoch": 0.7892458637937668, + "grad_norm": 1.5897230097077288, + "learning_rate": 2.240579574347832e-06, + "loss": 0.78, + "step": 8205 + }, + { + "epoch": 0.7893420546363986, + "grad_norm": 1.9217537192895118, + "learning_rate": 2.238614728943561e-06, + "loss": 0.8327, + "step": 8206 + }, + { + "epoch": 0.7894382454790304, + "grad_norm": 1.78365539619299, + "learning_rate": 2.236650636848563e-06, + "loss": 0.8381, + "step": 8207 + }, + { + "epoch": 0.7895344363216622, + "grad_norm": 1.9321357837209177, + "learning_rate": 2.2346872982534584e-06, + "loss": 0.9066, + "step": 8208 + }, + { + "epoch": 0.789630627164294, + "grad_norm": 1.938042625393243, + "learning_rate": 2.2327247133488116e-06, + "loss": 0.9615, + "step": 8209 + }, + { + "epoch": 0.7897268180069258, + "grad_norm": 1.7623488305859032, + "learning_rate": 2.2307628823251083e-06, + "loss": 0.9035, + "step": 8210 + }, + { + "epoch": 0.7898230088495575, + "grad_norm": 1.9943751626132666, + "learning_rate": 2.2288018053727634e-06, + "loss": 0.7437, + "step": 8211 + }, + { + "epoch": 0.7899191996921893, + "grad_norm": 1.940963305331633, + "learning_rate": 2.2268414826821117e-06, + "loss": 0.8906, + "step": 8212 + }, + { + "epoch": 0.7900153905348211, + "grad_norm": 1.9510894649158945, + "learning_rate": 2.2248819144434255e-06, + "loss": 0.9249, + "step": 8213 + }, + { + "epoch": 0.7901115813774529, + "grad_norm": 2.2535113763849615, + "learning_rate": 2.222923100846893e-06, + "loss": 0.9128, + "step": 8214 + }, + { + "epoch": 0.7902077722200846, + "grad_norm": 2.0170901441628604, + "learning_rate": 2.2209650420826323e-06, + "loss": 0.9351, + "step": 8215 + }, + { + "epoch": 0.7903039630627164, + "grad_norm": 2.242812318852286, + "learning_rate": 2.2190077383406938e-06, + "loss": 0.8825, + "step": 8216 + }, + { + "epoch": 0.7904001539053482, + "grad_norm": 2.0272736248600474, + "learning_rate": 2.217051189811051e-06, + "loss": 0.9073, + "step": 8217 + }, + { + "epoch": 0.79049634474798, + "grad_norm": 2.020508184145412, + "learning_rate": 2.2150953966835996e-06, + "loss": 0.9012, + "step": 8218 + }, + { + "epoch": 0.7905925355906118, + "grad_norm": 1.932308177176004, + "learning_rate": 2.2131403591481728e-06, + "loss": 0.8792, + "step": 8219 + }, + { + "epoch": 0.7906887264332435, + "grad_norm": 1.9301591091076715, + "learning_rate": 2.211186077394516e-06, + "loss": 0.9027, + "step": 8220 + }, + { + "epoch": 0.7907849172758753, + "grad_norm": 2.0595630365944464, + "learning_rate": 2.209232551612316e-06, + "loss": 0.7882, + "step": 8221 + }, + { + "epoch": 0.7908811081185071, + "grad_norm": 1.9916380522272723, + "learning_rate": 2.207279781991173e-06, + "loss": 1.0166, + "step": 8222 + }, + { + "epoch": 0.7909772989611389, + "grad_norm": 2.0422851101977537, + "learning_rate": 2.2053277687206233e-06, + "loss": 0.8571, + "step": 8223 + }, + { + "epoch": 0.7910734898037707, + "grad_norm": 2.0319122414874937, + "learning_rate": 2.2033765119901294e-06, + "loss": 0.9029, + "step": 8224 + }, + { + "epoch": 0.7911696806464025, + "grad_norm": 2.018863993823409, + "learning_rate": 2.201426011989075e-06, + "loss": 0.834, + "step": 8225 + }, + { + "epoch": 0.7912658714890343, + "grad_norm": 2.0856854642669016, + "learning_rate": 2.1994762689067705e-06, + "loss": 0.8804, + "step": 8226 + }, + { + "epoch": 0.7913620623316661, + "grad_norm": 1.7616656802436061, + "learning_rate": 2.197527282932457e-06, + "loss": 0.852, + "step": 8227 + }, + { + "epoch": 0.7914582531742979, + "grad_norm": 2.186807717151232, + "learning_rate": 2.1955790542553036e-06, + "loss": 0.8632, + "step": 8228 + }, + { + "epoch": 0.7915544440169295, + "grad_norm": 2.1072321531275686, + "learning_rate": 2.193631583064396e-06, + "loss": 0.818, + "step": 8229 + }, + { + "epoch": 0.7916506348595613, + "grad_norm": 1.881910976912469, + "learning_rate": 2.1916848695487615e-06, + "loss": 0.8326, + "step": 8230 + }, + { + "epoch": 0.7917468257021931, + "grad_norm": 2.124624607976205, + "learning_rate": 2.189738913897337e-06, + "loss": 0.9278, + "step": 8231 + }, + { + "epoch": 0.7918430165448249, + "grad_norm": 1.9835007625106407, + "learning_rate": 2.1877937162990015e-06, + "loss": 0.8817, + "step": 8232 + }, + { + "epoch": 0.7919392073874567, + "grad_norm": 2.271811655397712, + "learning_rate": 2.1858492769425464e-06, + "loss": 0.8845, + "step": 8233 + }, + { + "epoch": 0.7920353982300885, + "grad_norm": 2.0074504206882144, + "learning_rate": 2.1839055960167e-06, + "loss": 0.9509, + "step": 8234 + }, + { + "epoch": 0.7921315890727203, + "grad_norm": 2.0622670636804195, + "learning_rate": 2.181962673710115e-06, + "loss": 0.8846, + "step": 8235 + }, + { + "epoch": 0.7922277799153521, + "grad_norm": 1.9468251096497031, + "learning_rate": 2.180020510211367e-06, + "loss": 0.9174, + "step": 8236 + }, + { + "epoch": 0.7923239707579839, + "grad_norm": 2.1484190957738534, + "learning_rate": 2.178079105708958e-06, + "loss": 0.9581, + "step": 8237 + }, + { + "epoch": 0.7924201616006156, + "grad_norm": 1.8570863658306382, + "learning_rate": 2.1761384603913203e-06, + "loss": 0.8402, + "step": 8238 + }, + { + "epoch": 0.7925163524432474, + "grad_norm": 2.1965777617009588, + "learning_rate": 2.1741985744468073e-06, + "loss": 0.9325, + "step": 8239 + }, + { + "epoch": 0.7926125432858792, + "grad_norm": 2.0442773523933546, + "learning_rate": 2.172259448063704e-06, + "loss": 0.9317, + "step": 8240 + }, + { + "epoch": 0.792708734128511, + "grad_norm": 2.1391415364405484, + "learning_rate": 2.1703210814302177e-06, + "loss": 0.9232, + "step": 8241 + }, + { + "epoch": 0.7928049249711427, + "grad_norm": 2.2187813843793003, + "learning_rate": 2.1683834747344913e-06, + "loss": 0.9858, + "step": 8242 + }, + { + "epoch": 0.7929011158137745, + "grad_norm": 1.8753991424952534, + "learning_rate": 2.1664466281645737e-06, + "loss": 0.8227, + "step": 8243 + }, + { + "epoch": 0.7929973066564063, + "grad_norm": 1.9914521612450278, + "learning_rate": 2.1645105419084587e-06, + "loss": 0.9626, + "step": 8244 + }, + { + "epoch": 0.7930934974990381, + "grad_norm": 2.0375409368897, + "learning_rate": 2.1625752161540636e-06, + "loss": 0.8061, + "step": 8245 + }, + { + "epoch": 0.7931896883416699, + "grad_norm": 2.1644903483703404, + "learning_rate": 2.160640651089221e-06, + "loss": 0.8774, + "step": 8246 + }, + { + "epoch": 0.7932858791843016, + "grad_norm": 1.7005454338744221, + "learning_rate": 2.1587068469017015e-06, + "loss": 0.7676, + "step": 8247 + }, + { + "epoch": 0.7933820700269334, + "grad_norm": 1.8720815837558171, + "learning_rate": 2.1567738037791998e-06, + "loss": 0.8936, + "step": 8248 + }, + { + "epoch": 0.7934782608695652, + "grad_norm": 2.0830417525701526, + "learning_rate": 2.1548415219093312e-06, + "loss": 0.8942, + "step": 8249 + }, + { + "epoch": 0.793574451712197, + "grad_norm": 2.0210366433313536, + "learning_rate": 2.152910001479638e-06, + "loss": 0.7896, + "step": 8250 + }, + { + "epoch": 0.7936706425548288, + "grad_norm": 1.8169978623077738, + "learning_rate": 2.1509792426775934e-06, + "loss": 0.882, + "step": 8251 + }, + { + "epoch": 0.7937668333974606, + "grad_norm": 1.8950350940178649, + "learning_rate": 2.1490492456905964e-06, + "loss": 0.869, + "step": 8252 + }, + { + "epoch": 0.7938630242400924, + "grad_norm": 2.0266647112662977, + "learning_rate": 2.1471200107059686e-06, + "loss": 0.8922, + "step": 8253 + }, + { + "epoch": 0.7939592150827242, + "grad_norm": 1.8555122709473073, + "learning_rate": 2.1451915379109546e-06, + "loss": 0.8411, + "step": 8254 + }, + { + "epoch": 0.794055405925356, + "grad_norm": 2.001265022214711, + "learning_rate": 2.143263827492732e-06, + "loss": 0.9269, + "step": 8255 + }, + { + "epoch": 0.7941515967679876, + "grad_norm": 2.015190027892136, + "learning_rate": 2.141336879638406e-06, + "loss": 0.8452, + "step": 8256 + }, + { + "epoch": 0.7942477876106194, + "grad_norm": 2.284294699492564, + "learning_rate": 2.139410694534997e-06, + "loss": 1.0004, + "step": 8257 + }, + { + "epoch": 0.7943439784532512, + "grad_norm": 2.4061263488832805, + "learning_rate": 2.1374852723694595e-06, + "loss": 0.7902, + "step": 8258 + }, + { + "epoch": 0.794440169295883, + "grad_norm": 2.0528566014342453, + "learning_rate": 2.1355606133286764e-06, + "loss": 0.8676, + "step": 8259 + }, + { + "epoch": 0.7945363601385148, + "grad_norm": 1.9478409939914456, + "learning_rate": 2.133636717599451e-06, + "loss": 0.8517, + "step": 8260 + }, + { + "epoch": 0.7946325509811466, + "grad_norm": 2.0306058581659223, + "learning_rate": 2.131713585368508e-06, + "loss": 0.9078, + "step": 8261 + }, + { + "epoch": 0.7947287418237784, + "grad_norm": 1.9479821649797882, + "learning_rate": 2.1297912168225086e-06, + "loss": 0.815, + "step": 8262 + }, + { + "epoch": 0.7948249326664102, + "grad_norm": 2.0367176465127956, + "learning_rate": 2.127869612148038e-06, + "loss": 0.9367, + "step": 8263 + }, + { + "epoch": 0.794921123509042, + "grad_norm": 2.2691460141401825, + "learning_rate": 2.1259487715316e-06, + "loss": 1.0498, + "step": 8264 + }, + { + "epoch": 0.7950173143516738, + "grad_norm": 1.5868377595637313, + "learning_rate": 2.1240286951596313e-06, + "loss": 0.7676, + "step": 8265 + }, + { + "epoch": 0.7951135051943055, + "grad_norm": 1.8354362789663454, + "learning_rate": 2.1221093832184903e-06, + "loss": 0.9872, + "step": 8266 + }, + { + "epoch": 0.7952096960369373, + "grad_norm": 1.8630124677018598, + "learning_rate": 2.1201908358944612e-06, + "loss": 0.9991, + "step": 8267 + }, + { + "epoch": 0.795305886879569, + "grad_norm": 1.7408511446144026, + "learning_rate": 2.118273053373757e-06, + "loss": 0.8094, + "step": 8268 + }, + { + "epoch": 0.7954020777222008, + "grad_norm": 2.1196408823338113, + "learning_rate": 2.1163560358425196e-06, + "loss": 0.8671, + "step": 8269 + }, + { + "epoch": 0.7954982685648326, + "grad_norm": 2.079513424580162, + "learning_rate": 2.1144397834868034e-06, + "loss": 0.8536, + "step": 8270 + }, + { + "epoch": 0.7955944594074644, + "grad_norm": 1.934459220675783, + "learning_rate": 2.1125242964926053e-06, + "loss": 0.986, + "step": 8271 + }, + { + "epoch": 0.7956906502500962, + "grad_norm": 2.0918049353296913, + "learning_rate": 2.1106095750458332e-06, + "loss": 0.9299, + "step": 8272 + }, + { + "epoch": 0.795786841092728, + "grad_norm": 2.098746543693215, + "learning_rate": 2.108695619332334e-06, + "loss": 0.9258, + "step": 8273 + }, + { + "epoch": 0.7958830319353598, + "grad_norm": 1.4616863799248212, + "learning_rate": 2.106782429537866e-06, + "loss": 0.7433, + "step": 8274 + }, + { + "epoch": 0.7959792227779915, + "grad_norm": 2.0042013947144532, + "learning_rate": 2.104870005848125e-06, + "loss": 0.8621, + "step": 8275 + }, + { + "epoch": 0.7960754136206233, + "grad_norm": 2.1920380830760293, + "learning_rate": 2.1029583484487315e-06, + "loss": 0.9429, + "step": 8276 + }, + { + "epoch": 0.7961716044632551, + "grad_norm": 2.1723049298743007, + "learning_rate": 2.101047457525224e-06, + "loss": 0.9516, + "step": 8277 + }, + { + "epoch": 0.7962677953058869, + "grad_norm": 2.1473707646562072, + "learning_rate": 2.0991373332630683e-06, + "loss": 0.7422, + "step": 8278 + }, + { + "epoch": 0.7963639861485187, + "grad_norm": 1.9986397657227823, + "learning_rate": 2.0972279758476622e-06, + "loss": 0.8422, + "step": 8279 + }, + { + "epoch": 0.7964601769911505, + "grad_norm": 1.769168786891254, + "learning_rate": 2.0953193854643274e-06, + "loss": 0.8309, + "step": 8280 + }, + { + "epoch": 0.7965563678337823, + "grad_norm": 2.047165978377121, + "learning_rate": 2.093411562298304e-06, + "loss": 0.8969, + "step": 8281 + }, + { + "epoch": 0.796652558676414, + "grad_norm": 2.075711719259788, + "learning_rate": 2.0915045065347673e-06, + "loss": 0.9514, + "step": 8282 + }, + { + "epoch": 0.7967487495190458, + "grad_norm": 2.184371959333324, + "learning_rate": 2.089598218358808e-06, + "loss": 1.0028, + "step": 8283 + }, + { + "epoch": 0.7968449403616775, + "grad_norm": 2.2252841169960087, + "learning_rate": 2.0876926979554545e-06, + "loss": 0.832, + "step": 8284 + }, + { + "epoch": 0.7969411312043093, + "grad_norm": 1.9672094802867277, + "learning_rate": 2.0857879455096485e-06, + "loss": 0.8836, + "step": 8285 + }, + { + "epoch": 0.7970373220469411, + "grad_norm": 2.0400515726895647, + "learning_rate": 2.0838839612062633e-06, + "loss": 0.9062, + "step": 8286 + }, + { + "epoch": 0.7971335128895729, + "grad_norm": 2.051716430821874, + "learning_rate": 2.081980745230102e-06, + "loss": 0.9537, + "step": 8287 + }, + { + "epoch": 0.7972297037322047, + "grad_norm": 1.9569015254962363, + "learning_rate": 2.080078297765884e-06, + "loss": 0.8844, + "step": 8288 + }, + { + "epoch": 0.7973258945748365, + "grad_norm": 2.11652868134092, + "learning_rate": 2.0781766189982554e-06, + "loss": 0.9036, + "step": 8289 + }, + { + "epoch": 0.7974220854174683, + "grad_norm": 2.089611948132179, + "learning_rate": 2.0762757091117937e-06, + "loss": 0.8613, + "step": 8290 + }, + { + "epoch": 0.7975182762601001, + "grad_norm": 1.9220939445859269, + "learning_rate": 2.0743755682910014e-06, + "loss": 0.9649, + "step": 8291 + }, + { + "epoch": 0.7976144671027319, + "grad_norm": 2.007861472593318, + "learning_rate": 2.0724761967202987e-06, + "loss": 0.9237, + "step": 8292 + }, + { + "epoch": 0.7977106579453636, + "grad_norm": 2.014897977280439, + "learning_rate": 2.070577594584037e-06, + "loss": 0.9444, + "step": 8293 + }, + { + "epoch": 0.7978068487879953, + "grad_norm": 1.8225061864426868, + "learning_rate": 2.0686797620664987e-06, + "loss": 0.9006, + "step": 8294 + }, + { + "epoch": 0.7979030396306271, + "grad_norm": 2.237159604237435, + "learning_rate": 2.0667826993518735e-06, + "loss": 0.8626, + "step": 8295 + }, + { + "epoch": 0.7979992304732589, + "grad_norm": 1.9748838798935004, + "learning_rate": 2.0648864066242937e-06, + "loss": 0.9104, + "step": 8296 + }, + { + "epoch": 0.7980954213158907, + "grad_norm": 2.0006417177190676, + "learning_rate": 2.0629908840678126e-06, + "loss": 0.9362, + "step": 8297 + }, + { + "epoch": 0.7981916121585225, + "grad_norm": 2.2589987417519155, + "learning_rate": 2.0610961318664013e-06, + "loss": 1.0347, + "step": 8298 + }, + { + "epoch": 0.7982878030011543, + "grad_norm": 1.6504985904681209, + "learning_rate": 2.059202150203966e-06, + "loss": 0.7012, + "step": 8299 + }, + { + "epoch": 0.7983839938437861, + "grad_norm": 2.193275359875959, + "learning_rate": 2.0573089392643362e-06, + "loss": 0.8605, + "step": 8300 + }, + { + "epoch": 0.7984801846864179, + "grad_norm": 2.138077178708678, + "learning_rate": 2.055416499231261e-06, + "loss": 0.9304, + "step": 8301 + }, + { + "epoch": 0.7985763755290496, + "grad_norm": 1.7374979087814262, + "learning_rate": 2.0535248302884147e-06, + "loss": 0.92, + "step": 8302 + }, + { + "epoch": 0.7986725663716814, + "grad_norm": 1.6721722635259901, + "learning_rate": 2.0516339326194045e-06, + "loss": 0.699, + "step": 8303 + }, + { + "epoch": 0.7987687572143132, + "grad_norm": 1.9437528303742553, + "learning_rate": 2.0497438064077603e-06, + "loss": 0.96, + "step": 8304 + }, + { + "epoch": 0.798864948056945, + "grad_norm": 2.057850698318068, + "learning_rate": 2.0478544518369316e-06, + "loss": 0.9561, + "step": 8305 + }, + { + "epoch": 0.7989611388995768, + "grad_norm": 1.997378493484274, + "learning_rate": 2.045965869090295e-06, + "loss": 0.9102, + "step": 8306 + }, + { + "epoch": 0.7990573297422086, + "grad_norm": 1.9499493384781446, + "learning_rate": 2.0440780583511556e-06, + "loss": 0.8725, + "step": 8307 + }, + { + "epoch": 0.7991535205848403, + "grad_norm": 1.9497892764362186, + "learning_rate": 2.0421910198027452e-06, + "loss": 0.9207, + "step": 8308 + }, + { + "epoch": 0.7992497114274721, + "grad_norm": 1.8690671780147867, + "learning_rate": 2.0403047536282104e-06, + "loss": 0.8728, + "step": 8309 + }, + { + "epoch": 0.7993459022701039, + "grad_norm": 2.0115025135824336, + "learning_rate": 2.0384192600106335e-06, + "loss": 0.9005, + "step": 8310 + }, + { + "epoch": 0.7994420931127356, + "grad_norm": 2.071075998603577, + "learning_rate": 2.0365345391330193e-06, + "loss": 0.9009, + "step": 8311 + }, + { + "epoch": 0.7995382839553674, + "grad_norm": 1.9581913790647323, + "learning_rate": 2.0346505911782956e-06, + "loss": 0.9792, + "step": 8312 + }, + { + "epoch": 0.7996344747979992, + "grad_norm": 2.064952579780526, + "learning_rate": 2.032767416329311e-06, + "loss": 1.0026, + "step": 8313 + }, + { + "epoch": 0.799730665640631, + "grad_norm": 1.9842037206170826, + "learning_rate": 2.0308850147688484e-06, + "loss": 0.9376, + "step": 8314 + }, + { + "epoch": 0.7998268564832628, + "grad_norm": 1.8589447358512037, + "learning_rate": 2.029003386679612e-06, + "loss": 0.836, + "step": 8315 + }, + { + "epoch": 0.7999230473258946, + "grad_norm": 1.936980059378256, + "learning_rate": 2.0271225322442255e-06, + "loss": 0.9419, + "step": 8316 + }, + { + "epoch": 0.8000192381685264, + "grad_norm": 1.9987132601531121, + "learning_rate": 2.0252424516452473e-06, + "loss": 0.9515, + "step": 8317 + }, + { + "epoch": 0.8001154290111582, + "grad_norm": 1.9316048511955815, + "learning_rate": 2.0233631450651525e-06, + "loss": 0.9376, + "step": 8318 + }, + { + "epoch": 0.80021161985379, + "grad_norm": 1.83787217198301, + "learning_rate": 2.021484612686342e-06, + "loss": 0.9555, + "step": 8319 + }, + { + "epoch": 0.8003078106964217, + "grad_norm": 1.8351273785527442, + "learning_rate": 2.019606854691145e-06, + "loss": 0.9201, + "step": 8320 + }, + { + "epoch": 0.8004040015390534, + "grad_norm": 2.1509828557647834, + "learning_rate": 2.0177298712618177e-06, + "loss": 0.8394, + "step": 8321 + }, + { + "epoch": 0.8005001923816852, + "grad_norm": 2.085835586662869, + "learning_rate": 2.0158536625805325e-06, + "loss": 0.8541, + "step": 8322 + }, + { + "epoch": 0.800596383224317, + "grad_norm": 2.169878194067153, + "learning_rate": 2.013978228829395e-06, + "loss": 0.903, + "step": 8323 + }, + { + "epoch": 0.8006925740669488, + "grad_norm": 2.052231848406959, + "learning_rate": 2.01210357019043e-06, + "loss": 0.9321, + "step": 8324 + }, + { + "epoch": 0.8007887649095806, + "grad_norm": 2.057983184169466, + "learning_rate": 2.010229686845592e-06, + "loss": 0.8821, + "step": 8325 + }, + { + "epoch": 0.8008849557522124, + "grad_norm": 1.878431271300211, + "learning_rate": 2.008356578976752e-06, + "loss": 0.9433, + "step": 8326 + }, + { + "epoch": 0.8009811465948442, + "grad_norm": 1.9746252324461968, + "learning_rate": 2.0064842467657165e-06, + "loss": 0.9126, + "step": 8327 + }, + { + "epoch": 0.801077337437476, + "grad_norm": 1.9583572797881872, + "learning_rate": 2.004612690394212e-06, + "loss": 1.0111, + "step": 8328 + }, + { + "epoch": 0.8011735282801077, + "grad_norm": 2.037442486061321, + "learning_rate": 2.002741910043887e-06, + "loss": 0.8844, + "step": 8329 + }, + { + "epoch": 0.8012697191227395, + "grad_norm": 2.2210442150065814, + "learning_rate": 2.0008719058963144e-06, + "loss": 0.8537, + "step": 8330 + }, + { + "epoch": 0.8013659099653713, + "grad_norm": 1.9643112423982576, + "learning_rate": 1.9990026781329965e-06, + "loss": 0.974, + "step": 8331 + }, + { + "epoch": 0.8014621008080031, + "grad_norm": 2.092565142923105, + "learning_rate": 1.997134226935361e-06, + "loss": 0.965, + "step": 8332 + }, + { + "epoch": 0.8015582916506349, + "grad_norm": 1.7571274202742353, + "learning_rate": 1.9952665524847514e-06, + "loss": 0.8166, + "step": 8333 + }, + { + "epoch": 0.8016544824932667, + "grad_norm": 2.1421371506370592, + "learning_rate": 1.9933996549624468e-06, + "loss": 0.9615, + "step": 8334 + }, + { + "epoch": 0.8017506733358984, + "grad_norm": 2.4146344665315693, + "learning_rate": 1.991533534549641e-06, + "loss": 0.9199, + "step": 8335 + }, + { + "epoch": 0.8018468641785302, + "grad_norm": 1.9264736512748029, + "learning_rate": 1.9896681914274616e-06, + "loss": 0.9463, + "step": 8336 + }, + { + "epoch": 0.801943055021162, + "grad_norm": 2.2468317864365157, + "learning_rate": 1.987803625776953e-06, + "loss": 0.8787, + "step": 8337 + }, + { + "epoch": 0.8020392458637937, + "grad_norm": 2.110922833639807, + "learning_rate": 1.9859398377790872e-06, + "loss": 0.9604, + "step": 8338 + }, + { + "epoch": 0.8021354367064255, + "grad_norm": 1.9107316468294322, + "learning_rate": 1.9840768276147648e-06, + "loss": 0.9384, + "step": 8339 + }, + { + "epoch": 0.8022316275490573, + "grad_norm": 2.031379790885148, + "learning_rate": 1.982214595464804e-06, + "loss": 0.9903, + "step": 8340 + }, + { + "epoch": 0.8023278183916891, + "grad_norm": 1.974511906136445, + "learning_rate": 1.9803531415099484e-06, + "loss": 0.9581, + "step": 8341 + }, + { + "epoch": 0.8024240092343209, + "grad_norm": 2.4966859235224805, + "learning_rate": 1.97849246593087e-06, + "loss": 0.9953, + "step": 8342 + }, + { + "epoch": 0.8025202000769527, + "grad_norm": 1.8928578959269895, + "learning_rate": 1.976632568908168e-06, + "loss": 0.8251, + "step": 8343 + }, + { + "epoch": 0.8026163909195845, + "grad_norm": 2.340786548863918, + "learning_rate": 1.9747734506223525e-06, + "loss": 1.0123, + "step": 8344 + }, + { + "epoch": 0.8027125817622163, + "grad_norm": 1.9090407606791455, + "learning_rate": 1.9729151112538724e-06, + "loss": 0.9312, + "step": 8345 + }, + { + "epoch": 0.8028087726048481, + "grad_norm": 1.7523954809594124, + "learning_rate": 1.9710575509831008e-06, + "loss": 0.9091, + "step": 8346 + }, + { + "epoch": 0.8029049634474797, + "grad_norm": 2.0841039787000817, + "learning_rate": 1.969200769990318e-06, + "loss": 0.99, + "step": 8347 + }, + { + "epoch": 0.8030011542901115, + "grad_norm": 2.3437167688522313, + "learning_rate": 1.967344768455747e-06, + "loss": 0.9658, + "step": 8348 + }, + { + "epoch": 0.8030973451327433, + "grad_norm": 1.947293156755834, + "learning_rate": 1.96548954655953e-06, + "loss": 0.9152, + "step": 8349 + }, + { + "epoch": 0.8031935359753751, + "grad_norm": 1.8136253967648437, + "learning_rate": 1.9636351044817292e-06, + "loss": 0.8111, + "step": 8350 + }, + { + "epoch": 0.8032897268180069, + "grad_norm": 1.9629623349065268, + "learning_rate": 1.9617814424023375e-06, + "loss": 0.9519, + "step": 8351 + }, + { + "epoch": 0.8033859176606387, + "grad_norm": 2.0344578122728216, + "learning_rate": 1.9599285605012643e-06, + "loss": 0.9317, + "step": 8352 + }, + { + "epoch": 0.8034821085032705, + "grad_norm": 2.0274640943186952, + "learning_rate": 1.9580764589583524e-06, + "loss": 0.8746, + "step": 8353 + }, + { + "epoch": 0.8035782993459023, + "grad_norm": 2.293966703540545, + "learning_rate": 1.9562251379533593e-06, + "loss": 0.985, + "step": 8354 + }, + { + "epoch": 0.8036744901885341, + "grad_norm": 1.7967075964154644, + "learning_rate": 1.9543745976659745e-06, + "loss": 0.8807, + "step": 8355 + }, + { + "epoch": 0.8037706810311658, + "grad_norm": 1.8945128004000837, + "learning_rate": 1.952524838275811e-06, + "loss": 0.8788, + "step": 8356 + }, + { + "epoch": 0.8038668718737976, + "grad_norm": 2.0349669641335884, + "learning_rate": 1.9506758599624022e-06, + "loss": 1.0192, + "step": 8357 + }, + { + "epoch": 0.8039630627164294, + "grad_norm": 2.057653128598581, + "learning_rate": 1.9488276629052026e-06, + "loss": 0.8065, + "step": 8358 + }, + { + "epoch": 0.8040592535590612, + "grad_norm": 2.000911540535853, + "learning_rate": 1.946980247283602e-06, + "loss": 1.0717, + "step": 8359 + }, + { + "epoch": 0.804155444401693, + "grad_norm": 1.7288077522469314, + "learning_rate": 1.945133613276907e-06, + "loss": 0.8875, + "step": 8360 + }, + { + "epoch": 0.8042516352443247, + "grad_norm": 2.0768275132412706, + "learning_rate": 1.9432877610643453e-06, + "loss": 0.8434, + "step": 8361 + }, + { + "epoch": 0.8043478260869565, + "grad_norm": 1.9985410491322118, + "learning_rate": 1.941442690825076e-06, + "loss": 0.9822, + "step": 8362 + }, + { + "epoch": 0.8044440169295883, + "grad_norm": 1.949457328547113, + "learning_rate": 1.939598402738182e-06, + "loss": 0.8339, + "step": 8363 + }, + { + "epoch": 0.8045402077722201, + "grad_norm": 2.286958019332817, + "learning_rate": 1.937754896982663e-06, + "loss": 0.7785, + "step": 8364 + }, + { + "epoch": 0.8046363986148518, + "grad_norm": 1.8940136727166639, + "learning_rate": 1.9359121737374455e-06, + "loss": 0.914, + "step": 8365 + }, + { + "epoch": 0.8047325894574836, + "grad_norm": 1.985424682070563, + "learning_rate": 1.9340702331813842e-06, + "loss": 0.8294, + "step": 8366 + }, + { + "epoch": 0.8048287803001154, + "grad_norm": 2.0763301872143245, + "learning_rate": 1.9322290754932582e-06, + "loss": 0.8864, + "step": 8367 + }, + { + "epoch": 0.8049249711427472, + "grad_norm": 1.8619248952084404, + "learning_rate": 1.9303887008517618e-06, + "loss": 0.9332, + "step": 8368 + }, + { + "epoch": 0.805021161985379, + "grad_norm": 1.925327669437261, + "learning_rate": 1.928549109435525e-06, + "loss": 0.9255, + "step": 8369 + }, + { + "epoch": 0.8051173528280108, + "grad_norm": 2.1064025996582116, + "learning_rate": 1.9267103014230935e-06, + "loss": 0.9463, + "step": 8370 + }, + { + "epoch": 0.8052135436706426, + "grad_norm": 1.911422159919007, + "learning_rate": 1.9248722769929363e-06, + "loss": 0.8784, + "step": 8371 + }, + { + "epoch": 0.8053097345132744, + "grad_norm": 2.0383252401587475, + "learning_rate": 1.923035036323452e-06, + "loss": 0.8114, + "step": 8372 + }, + { + "epoch": 0.8054059253559062, + "grad_norm": 1.9918943594468945, + "learning_rate": 1.921198579592961e-06, + "loss": 0.9194, + "step": 8373 + }, + { + "epoch": 0.805502116198538, + "grad_norm": 2.017056328789721, + "learning_rate": 1.91936290697971e-06, + "loss": 0.8889, + "step": 8374 + }, + { + "epoch": 0.8055983070411696, + "grad_norm": 1.6722400555710184, + "learning_rate": 1.917528018661864e-06, + "loss": 0.8066, + "step": 8375 + }, + { + "epoch": 0.8056944978838014, + "grad_norm": 2.1441897373261525, + "learning_rate": 1.9156939148175125e-06, + "loss": 0.9439, + "step": 8376 + }, + { + "epoch": 0.8057906887264332, + "grad_norm": 1.906559048599058, + "learning_rate": 1.9138605956246757e-06, + "loss": 0.8577, + "step": 8377 + }, + { + "epoch": 0.805886879569065, + "grad_norm": 2.022439251161116, + "learning_rate": 1.9120280612612873e-06, + "loss": 1.015, + "step": 8378 + }, + { + "epoch": 0.8059830704116968, + "grad_norm": 2.1603951545382385, + "learning_rate": 1.9101963119052146e-06, + "loss": 0.8723, + "step": 8379 + }, + { + "epoch": 0.8060792612543286, + "grad_norm": 1.8286406626545995, + "learning_rate": 1.9083653477342467e-06, + "loss": 0.8746, + "step": 8380 + }, + { + "epoch": 0.8061754520969604, + "grad_norm": 1.759835838423859, + "learning_rate": 1.9065351689260914e-06, + "loss": 0.8469, + "step": 8381 + }, + { + "epoch": 0.8062716429395922, + "grad_norm": 1.9884407801981252, + "learning_rate": 1.904705775658381e-06, + "loss": 0.9267, + "step": 8382 + }, + { + "epoch": 0.806367833782224, + "grad_norm": 1.9999308615402618, + "learning_rate": 1.9028771681086767e-06, + "loss": 0.8531, + "step": 8383 + }, + { + "epoch": 0.8064640246248557, + "grad_norm": 2.012231250916898, + "learning_rate": 1.9010493464544621e-06, + "loss": 0.8729, + "step": 8384 + }, + { + "epoch": 0.8065602154674875, + "grad_norm": 2.0331242199377817, + "learning_rate": 1.899222310873139e-06, + "loss": 0.8605, + "step": 8385 + }, + { + "epoch": 0.8066564063101193, + "grad_norm": 2.032096314341096, + "learning_rate": 1.8973960615420416e-06, + "loss": 0.8734, + "step": 8386 + }, + { + "epoch": 0.806752597152751, + "grad_norm": 2.0257787257709734, + "learning_rate": 1.8955705986384187e-06, + "loss": 0.7249, + "step": 8387 + }, + { + "epoch": 0.8068487879953828, + "grad_norm": 1.8814603874236515, + "learning_rate": 1.8937459223394517e-06, + "loss": 0.9467, + "step": 8388 + }, + { + "epoch": 0.8069449788380146, + "grad_norm": 2.085496978388125, + "learning_rate": 1.8919220328222366e-06, + "loss": 0.9721, + "step": 8389 + }, + { + "epoch": 0.8070411696806464, + "grad_norm": 1.901146175479011, + "learning_rate": 1.8900989302637985e-06, + "loss": 0.9058, + "step": 8390 + }, + { + "epoch": 0.8071373605232782, + "grad_norm": 1.6521660176386197, + "learning_rate": 1.8882766148410892e-06, + "loss": 0.7085, + "step": 8391 + }, + { + "epoch": 0.80723355136591, + "grad_norm": 2.203508665423299, + "learning_rate": 1.8864550867309771e-06, + "loss": 0.8781, + "step": 8392 + }, + { + "epoch": 0.8073297422085417, + "grad_norm": 2.03040903426885, + "learning_rate": 1.8846343461102557e-06, + "loss": 0.9307, + "step": 8393 + }, + { + "epoch": 0.8074259330511735, + "grad_norm": 2.0736754206200056, + "learning_rate": 1.8828143931556442e-06, + "loss": 0.8598, + "step": 8394 + }, + { + "epoch": 0.8075221238938053, + "grad_norm": 1.8706580495964984, + "learning_rate": 1.8809952280437881e-06, + "loss": 0.8698, + "step": 8395 + }, + { + "epoch": 0.8076183147364371, + "grad_norm": 2.0156095449664546, + "learning_rate": 1.8791768509512487e-06, + "loss": 0.8925, + "step": 8396 + }, + { + "epoch": 0.8077145055790689, + "grad_norm": 1.826993292226288, + "learning_rate": 1.8773592620545168e-06, + "loss": 0.8387, + "step": 8397 + }, + { + "epoch": 0.8078106964217007, + "grad_norm": 1.962922265314669, + "learning_rate": 1.875542461530011e-06, + "loss": 0.9056, + "step": 8398 + }, + { + "epoch": 0.8079068872643325, + "grad_norm": 2.1323057472676794, + "learning_rate": 1.8737264495540564e-06, + "loss": 0.9278, + "step": 8399 + }, + { + "epoch": 0.8080030781069643, + "grad_norm": 1.7963178464519454, + "learning_rate": 1.871911226302917e-06, + "loss": 0.9856, + "step": 8400 + }, + { + "epoch": 0.808099268949596, + "grad_norm": 2.2158206930574478, + "learning_rate": 1.8700967919527801e-06, + "loss": 0.95, + "step": 8401 + }, + { + "epoch": 0.8081954597922277, + "grad_norm": 1.7579526115465665, + "learning_rate": 1.868283146679747e-06, + "loss": 0.7879, + "step": 8402 + }, + { + "epoch": 0.8082916506348595, + "grad_norm": 1.9277710113102644, + "learning_rate": 1.8664702906598508e-06, + "loss": 0.8639, + "step": 8403 + }, + { + "epoch": 0.8083878414774913, + "grad_norm": 1.7338053943297072, + "learning_rate": 1.8646582240690414e-06, + "loss": 0.8135, + "step": 8404 + }, + { + "epoch": 0.8084840323201231, + "grad_norm": 2.0083024188858847, + "learning_rate": 1.8628469470832e-06, + "loss": 0.9164, + "step": 8405 + }, + { + "epoch": 0.8085802231627549, + "grad_norm": 1.9753470955257122, + "learning_rate": 1.8610364598781227e-06, + "loss": 0.7725, + "step": 8406 + }, + { + "epoch": 0.8086764140053867, + "grad_norm": 1.85753247335716, + "learning_rate": 1.8592267626295325e-06, + "loss": 0.993, + "step": 8407 + }, + { + "epoch": 0.8087726048480185, + "grad_norm": 1.9352507930610143, + "learning_rate": 1.8574178555130818e-06, + "loss": 0.8985, + "step": 8408 + }, + { + "epoch": 0.8088687956906503, + "grad_norm": 2.0071412464377847, + "learning_rate": 1.8556097387043359e-06, + "loss": 0.9523, + "step": 8409 + }, + { + "epoch": 0.8089649865332821, + "grad_norm": 2.217363550359012, + "learning_rate": 1.8538024123787868e-06, + "loss": 0.9345, + "step": 8410 + }, + { + "epoch": 0.8090611773759138, + "grad_norm": 2.0382368590882756, + "learning_rate": 1.851995876711853e-06, + "loss": 0.8619, + "step": 8411 + }, + { + "epoch": 0.8091573682185456, + "grad_norm": 1.7660859183964286, + "learning_rate": 1.8501901318788773e-06, + "loss": 0.917, + "step": 8412 + }, + { + "epoch": 0.8092535590611774, + "grad_norm": 1.9511434302307693, + "learning_rate": 1.8483851780551177e-06, + "loss": 0.9775, + "step": 8413 + }, + { + "epoch": 0.8093497499038091, + "grad_norm": 2.004608001840495, + "learning_rate": 1.8465810154157626e-06, + "loss": 0.9678, + "step": 8414 + }, + { + "epoch": 0.8094459407464409, + "grad_norm": 2.0880351845789153, + "learning_rate": 1.8447776441359243e-06, + "loss": 0.9105, + "step": 8415 + }, + { + "epoch": 0.8095421315890727, + "grad_norm": 2.0096909988319154, + "learning_rate": 1.8429750643906331e-06, + "loss": 0.8719, + "step": 8416 + }, + { + "epoch": 0.8096383224317045, + "grad_norm": 2.2772703218701, + "learning_rate": 1.8411732763548417e-06, + "loss": 0.8662, + "step": 8417 + }, + { + "epoch": 0.8097345132743363, + "grad_norm": 1.9649079545995705, + "learning_rate": 1.8393722802034331e-06, + "loss": 0.9424, + "step": 8418 + }, + { + "epoch": 0.8098307041169681, + "grad_norm": 2.0609698223547213, + "learning_rate": 1.8375720761112103e-06, + "loss": 0.8125, + "step": 8419 + }, + { + "epoch": 0.8099268949595998, + "grad_norm": 2.0372335258951866, + "learning_rate": 1.835772664252895e-06, + "loss": 0.8818, + "step": 8420 + }, + { + "epoch": 0.8100230858022316, + "grad_norm": 1.9650818656445685, + "learning_rate": 1.8339740448031395e-06, + "loss": 0.9214, + "step": 8421 + }, + { + "epoch": 0.8101192766448634, + "grad_norm": 1.9302104584921376, + "learning_rate": 1.832176217936511e-06, + "loss": 0.9018, + "step": 8422 + }, + { + "epoch": 0.8102154674874952, + "grad_norm": 1.9968639582353085, + "learning_rate": 1.8303791838275087e-06, + "loss": 0.9678, + "step": 8423 + }, + { + "epoch": 0.810311658330127, + "grad_norm": 2.093974334776294, + "learning_rate": 1.8285829426505453e-06, + "loss": 1.0401, + "step": 8424 + }, + { + "epoch": 0.8104078491727588, + "grad_norm": 1.8983698259824684, + "learning_rate": 1.8267874945799647e-06, + "loss": 0.9273, + "step": 8425 + }, + { + "epoch": 0.8105040400153906, + "grad_norm": 2.0831421822141514, + "learning_rate": 1.8249928397900351e-06, + "loss": 0.8557, + "step": 8426 + }, + { + "epoch": 0.8106002308580224, + "grad_norm": 1.9748142840945249, + "learning_rate": 1.823198978454932e-06, + "loss": 0.8961, + "step": 8427 + }, + { + "epoch": 0.8106964217006541, + "grad_norm": 2.1966802228761004, + "learning_rate": 1.8214059107487726e-06, + "loss": 0.914, + "step": 8428 + }, + { + "epoch": 0.8107926125432858, + "grad_norm": 1.8277934512496212, + "learning_rate": 1.8196136368455896e-06, + "loss": 0.9474, + "step": 8429 + }, + { + "epoch": 0.8108888033859176, + "grad_norm": 2.073448298830218, + "learning_rate": 1.8178221569193343e-06, + "loss": 0.9272, + "step": 8430 + }, + { + "epoch": 0.8109849942285494, + "grad_norm": 1.9794527321531838, + "learning_rate": 1.8160314711438897e-06, + "loss": 0.856, + "step": 8431 + }, + { + "epoch": 0.8110811850711812, + "grad_norm": 2.0728747486129877, + "learning_rate": 1.8142415796930568e-06, + "loss": 0.9558, + "step": 8432 + }, + { + "epoch": 0.811177375913813, + "grad_norm": 1.9976208662853439, + "learning_rate": 1.8124524827405588e-06, + "loss": 0.8944, + "step": 8433 + }, + { + "epoch": 0.8112735667564448, + "grad_norm": 1.9645667813003853, + "learning_rate": 1.8106641804600411e-06, + "loss": 0.9784, + "step": 8434 + }, + { + "epoch": 0.8113697575990766, + "grad_norm": 2.1190735176995648, + "learning_rate": 1.8088766730250761e-06, + "loss": 0.9445, + "step": 8435 + }, + { + "epoch": 0.8114659484417084, + "grad_norm": 2.02615187905903, + "learning_rate": 1.8070899606091586e-06, + "loss": 0.8691, + "step": 8436 + }, + { + "epoch": 0.8115621392843402, + "grad_norm": 2.3678435070193222, + "learning_rate": 1.8053040433856995e-06, + "loss": 0.847, + "step": 8437 + }, + { + "epoch": 0.8116583301269719, + "grad_norm": 1.831720451238064, + "learning_rate": 1.8035189215280423e-06, + "loss": 0.9291, + "step": 8438 + }, + { + "epoch": 0.8117545209696037, + "grad_norm": 2.049672055511785, + "learning_rate": 1.8017345952094444e-06, + "loss": 0.992, + "step": 8439 + }, + { + "epoch": 0.8118507118122354, + "grad_norm": 1.7514234463525684, + "learning_rate": 1.799951064603095e-06, + "loss": 0.824, + "step": 8440 + }, + { + "epoch": 0.8119469026548672, + "grad_norm": 2.1108359839462447, + "learning_rate": 1.798168329882095e-06, + "loss": 0.731, + "step": 8441 + }, + { + "epoch": 0.812043093497499, + "grad_norm": 2.0451032967528255, + "learning_rate": 1.7963863912194768e-06, + "loss": 0.8755, + "step": 8442 + }, + { + "epoch": 0.8121392843401308, + "grad_norm": 2.0640071036412406, + "learning_rate": 1.794605248788196e-06, + "loss": 0.9051, + "step": 8443 + }, + { + "epoch": 0.8122354751827626, + "grad_norm": 1.9930349200776363, + "learning_rate": 1.7928249027611255e-06, + "loss": 0.9059, + "step": 8444 + }, + { + "epoch": 0.8123316660253944, + "grad_norm": 1.9805882953275882, + "learning_rate": 1.7910453533110605e-06, + "loss": 0.9732, + "step": 8445 + }, + { + "epoch": 0.8124278568680262, + "grad_norm": 2.1351518725304612, + "learning_rate": 1.789266600610724e-06, + "loss": 0.9809, + "step": 8446 + }, + { + "epoch": 0.8125240477106579, + "grad_norm": 1.9187860013355826, + "learning_rate": 1.787488644832761e-06, + "loss": 0.8362, + "step": 8447 + }, + { + "epoch": 0.8126202385532897, + "grad_norm": 1.8144724578813463, + "learning_rate": 1.7857114861497337e-06, + "loss": 0.766, + "step": 8448 + }, + { + "epoch": 0.8127164293959215, + "grad_norm": 2.1752367960348766, + "learning_rate": 1.783935124734133e-06, + "loss": 0.8441, + "step": 8449 + }, + { + "epoch": 0.8128126202385533, + "grad_norm": 1.9207418936508924, + "learning_rate": 1.782159560758373e-06, + "loss": 0.7981, + "step": 8450 + }, + { + "epoch": 0.8129088110811851, + "grad_norm": 2.233410828764646, + "learning_rate": 1.7803847943947805e-06, + "loss": 0.8729, + "step": 8451 + }, + { + "epoch": 0.8130050019238169, + "grad_norm": 2.0715796791230465, + "learning_rate": 1.7786108258156154e-06, + "loss": 0.8925, + "step": 8452 + }, + { + "epoch": 0.8131011927664487, + "grad_norm": 2.114726467545736, + "learning_rate": 1.7768376551930566e-06, + "loss": 0.9113, + "step": 8453 + }, + { + "epoch": 0.8131973836090804, + "grad_norm": 2.3760477594947247, + "learning_rate": 1.7750652826992077e-06, + "loss": 0.934, + "step": 8454 + }, + { + "epoch": 0.8132935744517122, + "grad_norm": 2.2765737541968702, + "learning_rate": 1.7732937085060908e-06, + "loss": 1.0003, + "step": 8455 + }, + { + "epoch": 0.8133897652943439, + "grad_norm": 2.338658683459913, + "learning_rate": 1.7715229327856498e-06, + "loss": 0.9016, + "step": 8456 + }, + { + "epoch": 0.8134859561369757, + "grad_norm": 2.1820079382225313, + "learning_rate": 1.7697529557097593e-06, + "loss": 0.8548, + "step": 8457 + }, + { + "epoch": 0.8135821469796075, + "grad_norm": 1.9241276416842819, + "learning_rate": 1.7679837774502052e-06, + "loss": 0.7539, + "step": 8458 + }, + { + "epoch": 0.8136783378222393, + "grad_norm": 1.9370804054319626, + "learning_rate": 1.7662153981787033e-06, + "loss": 0.8346, + "step": 8459 + }, + { + "epoch": 0.8137745286648711, + "grad_norm": 2.1643399389029616, + "learning_rate": 1.7644478180668945e-06, + "loss": 0.8832, + "step": 8460 + }, + { + "epoch": 0.8138707195075029, + "grad_norm": 1.840828382621656, + "learning_rate": 1.7626810372863323e-06, + "loss": 0.822, + "step": 8461 + }, + { + "epoch": 0.8139669103501347, + "grad_norm": 1.9972424295943925, + "learning_rate": 1.7609150560084986e-06, + "loss": 0.9248, + "step": 8462 + }, + { + "epoch": 0.8140631011927665, + "grad_norm": 2.2217626767815344, + "learning_rate": 1.7591498744047974e-06, + "loss": 0.9608, + "step": 8463 + }, + { + "epoch": 0.8141592920353983, + "grad_norm": 1.9760998915789967, + "learning_rate": 1.7573854926465582e-06, + "loss": 0.8748, + "step": 8464 + }, + { + "epoch": 0.81425548287803, + "grad_norm": 1.7647892644080343, + "learning_rate": 1.7556219109050243e-06, + "loss": 0.848, + "step": 8465 + }, + { + "epoch": 0.8143516737206618, + "grad_norm": 2.1915009901303186, + "learning_rate": 1.7538591293513685e-06, + "loss": 0.8951, + "step": 8466 + }, + { + "epoch": 0.8144478645632935, + "grad_norm": 1.9814326200502295, + "learning_rate": 1.7520971481566873e-06, + "loss": 0.9139, + "step": 8467 + }, + { + "epoch": 0.8145440554059253, + "grad_norm": 1.9049870346530577, + "learning_rate": 1.7503359674919929e-06, + "loss": 0.6943, + "step": 8468 + }, + { + "epoch": 0.8146402462485571, + "grad_norm": 2.177036111918073, + "learning_rate": 1.7485755875282207e-06, + "loss": 0.9036, + "step": 8469 + }, + { + "epoch": 0.8147364370911889, + "grad_norm": 2.023397243563379, + "learning_rate": 1.746816008436234e-06, + "loss": 0.9681, + "step": 8470 + }, + { + "epoch": 0.8148326279338207, + "grad_norm": 1.8397994881915671, + "learning_rate": 1.7450572303868163e-06, + "loss": 0.9288, + "step": 8471 + }, + { + "epoch": 0.8149288187764525, + "grad_norm": 1.778191413808346, + "learning_rate": 1.7432992535506687e-06, + "loss": 0.8832, + "step": 8472 + }, + { + "epoch": 0.8150250096190843, + "grad_norm": 1.9859845529593094, + "learning_rate": 1.7415420780984216e-06, + "loss": 0.9109, + "step": 8473 + }, + { + "epoch": 0.8151212004617161, + "grad_norm": 1.7954186871851243, + "learning_rate": 1.7397857042006194e-06, + "loss": 0.8251, + "step": 8474 + }, + { + "epoch": 0.8152173913043478, + "grad_norm": 1.8574320901153984, + "learning_rate": 1.7380301320277382e-06, + "loss": 0.9194, + "step": 8475 + }, + { + "epoch": 0.8153135821469796, + "grad_norm": 1.9987109446089495, + "learning_rate": 1.736275361750167e-06, + "loss": 0.9626, + "step": 8476 + }, + { + "epoch": 0.8154097729896114, + "grad_norm": 1.8494727228397319, + "learning_rate": 1.7345213935382232e-06, + "loss": 0.9363, + "step": 8477 + }, + { + "epoch": 0.8155059638322432, + "grad_norm": 2.0830954026789623, + "learning_rate": 1.7327682275621506e-06, + "loss": 0.9313, + "step": 8478 + }, + { + "epoch": 0.815602154674875, + "grad_norm": 1.9471689225532116, + "learning_rate": 1.7310158639920982e-06, + "loss": 0.861, + "step": 8479 + }, + { + "epoch": 0.8156983455175068, + "grad_norm": 1.9013023308638939, + "learning_rate": 1.7292643029981525e-06, + "loss": 0.7652, + "step": 8480 + }, + { + "epoch": 0.8157945363601385, + "grad_norm": 2.0210293121898846, + "learning_rate": 1.72751354475032e-06, + "loss": 0.9884, + "step": 8481 + }, + { + "epoch": 0.8158907272027703, + "grad_norm": 1.9509712203669547, + "learning_rate": 1.7257635894185232e-06, + "loss": 0.9449, + "step": 8482 + }, + { + "epoch": 0.8159869180454021, + "grad_norm": 2.5941670491011752, + "learning_rate": 1.7240144371726119e-06, + "loss": 0.9334, + "step": 8483 + }, + { + "epoch": 0.8160831088880338, + "grad_norm": 2.07173636285884, + "learning_rate": 1.7222660881823594e-06, + "loss": 0.958, + "step": 8484 + }, + { + "epoch": 0.8161792997306656, + "grad_norm": 2.066520178748421, + "learning_rate": 1.7205185426174553e-06, + "loss": 0.8525, + "step": 8485 + }, + { + "epoch": 0.8162754905732974, + "grad_norm": 1.9701843473638687, + "learning_rate": 1.7187718006475117e-06, + "loss": 0.8981, + "step": 8486 + }, + { + "epoch": 0.8163716814159292, + "grad_norm": 2.1831908776404654, + "learning_rate": 1.7170258624420665e-06, + "loss": 0.9607, + "step": 8487 + }, + { + "epoch": 0.816467872258561, + "grad_norm": 1.9557642977855028, + "learning_rate": 1.7152807281705809e-06, + "loss": 0.9299, + "step": 8488 + }, + { + "epoch": 0.8165640631011928, + "grad_norm": 1.9287450631158911, + "learning_rate": 1.713536398002431e-06, + "loss": 0.9429, + "step": 8489 + }, + { + "epoch": 0.8166602539438246, + "grad_norm": 2.077652950862333, + "learning_rate": 1.7117928721069233e-06, + "loss": 0.8905, + "step": 8490 + }, + { + "epoch": 0.8167564447864564, + "grad_norm": 2.4407096384339098, + "learning_rate": 1.7100501506532775e-06, + "loss": 0.938, + "step": 8491 + }, + { + "epoch": 0.8168526356290882, + "grad_norm": 2.144548898139231, + "learning_rate": 1.708308233810644e-06, + "loss": 0.8344, + "step": 8492 + }, + { + "epoch": 0.8169488264717198, + "grad_norm": 2.2372516936877087, + "learning_rate": 1.706567121748086e-06, + "loss": 0.8682, + "step": 8493 + }, + { + "epoch": 0.8170450173143516, + "grad_norm": 2.0798342645389276, + "learning_rate": 1.704826814634597e-06, + "loss": 0.7717, + "step": 8494 + }, + { + "epoch": 0.8171412081569834, + "grad_norm": 2.0544912504208948, + "learning_rate": 1.7030873126390901e-06, + "loss": 0.9603, + "step": 8495 + }, + { + "epoch": 0.8172373989996152, + "grad_norm": 1.7514974472501437, + "learning_rate": 1.701348615930397e-06, + "loss": 0.9007, + "step": 8496 + }, + { + "epoch": 0.817333589842247, + "grad_norm": 2.3502266370730664, + "learning_rate": 1.6996107246772697e-06, + "loss": 0.9681, + "step": 8497 + }, + { + "epoch": 0.8174297806848788, + "grad_norm": 2.221528318800539, + "learning_rate": 1.6978736390483896e-06, + "loss": 0.9455, + "step": 8498 + }, + { + "epoch": 0.8175259715275106, + "grad_norm": 2.1970846702287417, + "learning_rate": 1.6961373592123576e-06, + "loss": 0.8237, + "step": 8499 + }, + { + "epoch": 0.8176221623701424, + "grad_norm": 2.107872346795539, + "learning_rate": 1.6944018853376898e-06, + "loss": 0.9477, + "step": 8500 + }, + { + "epoch": 0.8177183532127742, + "grad_norm": 2.040786662201964, + "learning_rate": 1.6926672175928327e-06, + "loss": 0.892, + "step": 8501 + }, + { + "epoch": 0.8178145440554059, + "grad_norm": 1.8741227759742611, + "learning_rate": 1.6909333561461471e-06, + "loss": 0.8848, + "step": 8502 + }, + { + "epoch": 0.8179107348980377, + "grad_norm": 1.700833169001233, + "learning_rate": 1.6892003011659242e-06, + "loss": 0.7537, + "step": 8503 + }, + { + "epoch": 0.8180069257406695, + "grad_norm": 1.7818304478517197, + "learning_rate": 1.6874680528203657e-06, + "loss": 0.7702, + "step": 8504 + }, + { + "epoch": 0.8181031165833013, + "grad_norm": 2.1636544667522366, + "learning_rate": 1.6857366112776042e-06, + "loss": 0.9575, + "step": 8505 + }, + { + "epoch": 0.818199307425933, + "grad_norm": 2.085725753932172, + "learning_rate": 1.6840059767056949e-06, + "loss": 0.993, + "step": 8506 + }, + { + "epoch": 0.8182954982685648, + "grad_norm": 1.833260695652439, + "learning_rate": 1.6822761492726059e-06, + "loss": 0.8809, + "step": 8507 + }, + { + "epoch": 0.8183916891111966, + "grad_norm": 2.044695969586892, + "learning_rate": 1.6805471291462316e-06, + "loss": 0.8142, + "step": 8508 + }, + { + "epoch": 0.8184878799538284, + "grad_norm": 1.8338470415044117, + "learning_rate": 1.6788189164943912e-06, + "loss": 0.8825, + "step": 8509 + }, + { + "epoch": 0.8185840707964602, + "grad_norm": 2.0765867636863997, + "learning_rate": 1.6770915114848197e-06, + "loss": 0.8604, + "step": 8510 + }, + { + "epoch": 0.8186802616390919, + "grad_norm": 2.0916575008190557, + "learning_rate": 1.675364914285178e-06, + "loss": 0.8924, + "step": 8511 + }, + { + "epoch": 0.8187764524817237, + "grad_norm": 2.516262657928399, + "learning_rate": 1.67363912506305e-06, + "loss": 0.9669, + "step": 8512 + }, + { + "epoch": 0.8188726433243555, + "grad_norm": 1.8171283905054338, + "learning_rate": 1.6719141439859355e-06, + "loss": 0.8594, + "step": 8513 + }, + { + "epoch": 0.8189688341669873, + "grad_norm": 2.5057124091529364, + "learning_rate": 1.6701899712212565e-06, + "loss": 0.8941, + "step": 8514 + }, + { + "epoch": 0.8190650250096191, + "grad_norm": 1.907733658006722, + "learning_rate": 1.668466606936362e-06, + "loss": 0.904, + "step": 8515 + }, + { + "epoch": 0.8191612158522509, + "grad_norm": 1.834193036359602, + "learning_rate": 1.66674405129852e-06, + "loss": 0.7925, + "step": 8516 + }, + { + "epoch": 0.8192574066948827, + "grad_norm": 1.966447000944269, + "learning_rate": 1.665022304474917e-06, + "loss": 0.8771, + "step": 8517 + }, + { + "epoch": 0.8193535975375145, + "grad_norm": 1.722043423283059, + "learning_rate": 1.6633013666326636e-06, + "loss": 0.6884, + "step": 8518 + }, + { + "epoch": 0.8194497883801463, + "grad_norm": 2.5751192611708773, + "learning_rate": 1.661581237938794e-06, + "loss": 0.9519, + "step": 8519 + }, + { + "epoch": 0.8195459792227779, + "grad_norm": 2.0469489600830117, + "learning_rate": 1.6598619185602616e-06, + "loss": 0.7738, + "step": 8520 + }, + { + "epoch": 0.8196421700654097, + "grad_norm": 1.7021713358492538, + "learning_rate": 1.658143408663936e-06, + "loss": 0.7569, + "step": 8521 + }, + { + "epoch": 0.8197383609080415, + "grad_norm": 1.9848999757921837, + "learning_rate": 1.656425708416617e-06, + "loss": 0.8231, + "step": 8522 + }, + { + "epoch": 0.8198345517506733, + "grad_norm": 1.8549250324852484, + "learning_rate": 1.6547088179850245e-06, + "loss": 0.9042, + "step": 8523 + }, + { + "epoch": 0.8199307425933051, + "grad_norm": 1.869486099360875, + "learning_rate": 1.6529927375357957e-06, + "loss": 0.9165, + "step": 8524 + }, + { + "epoch": 0.8200269334359369, + "grad_norm": 1.875122260081714, + "learning_rate": 1.6512774672354869e-06, + "loss": 0.8259, + "step": 8525 + }, + { + "epoch": 0.8201231242785687, + "grad_norm": 1.6586536109205887, + "learning_rate": 1.6495630072505841e-06, + "loss": 0.7409, + "step": 8526 + }, + { + "epoch": 0.8202193151212005, + "grad_norm": 2.3548529000692673, + "learning_rate": 1.6478493577474919e-06, + "loss": 0.8722, + "step": 8527 + }, + { + "epoch": 0.8203155059638323, + "grad_norm": 2.463569050140471, + "learning_rate": 1.6461365188925304e-06, + "loss": 0.9789, + "step": 8528 + }, + { + "epoch": 0.820411696806464, + "grad_norm": 2.081951012084594, + "learning_rate": 1.6444244908519468e-06, + "loss": 0.9967, + "step": 8529 + }, + { + "epoch": 0.8205078876490958, + "grad_norm": 1.996015014821554, + "learning_rate": 1.642713273791914e-06, + "loss": 0.8354, + "step": 8530 + }, + { + "epoch": 0.8206040784917276, + "grad_norm": 1.8921133927639815, + "learning_rate": 1.6410028678785095e-06, + "loss": 0.8859, + "step": 8531 + }, + { + "epoch": 0.8207002693343594, + "grad_norm": 1.9186115280833091, + "learning_rate": 1.6392932732777489e-06, + "loss": 0.8842, + "step": 8532 + }, + { + "epoch": 0.8207964601769911, + "grad_norm": 1.9156061486595464, + "learning_rate": 1.6375844901555649e-06, + "loss": 0.866, + "step": 8533 + }, + { + "epoch": 0.8208926510196229, + "grad_norm": 2.1231666284384145, + "learning_rate": 1.6358765186778057e-06, + "loss": 0.9065, + "step": 8534 + }, + { + "epoch": 0.8209888418622547, + "grad_norm": 2.092612276326798, + "learning_rate": 1.6341693590102458e-06, + "loss": 0.8357, + "step": 8535 + }, + { + "epoch": 0.8210850327048865, + "grad_norm": 1.945915678181241, + "learning_rate": 1.6324630113185835e-06, + "loss": 0.9446, + "step": 8536 + }, + { + "epoch": 0.8211812235475183, + "grad_norm": 1.650538909647081, + "learning_rate": 1.6307574757684307e-06, + "loss": 0.757, + "step": 8537 + }, + { + "epoch": 0.82127741439015, + "grad_norm": 2.1747253727982785, + "learning_rate": 1.629052752525323e-06, + "loss": 0.9308, + "step": 8538 + }, + { + "epoch": 0.8213736052327818, + "grad_norm": 2.0269959164173685, + "learning_rate": 1.62734884175472e-06, + "loss": 0.8407, + "step": 8539 + }, + { + "epoch": 0.8214697960754136, + "grad_norm": 2.231623310612907, + "learning_rate": 1.625645743622003e-06, + "loss": 0.9543, + "step": 8540 + }, + { + "epoch": 0.8215659869180454, + "grad_norm": 1.8484200324868072, + "learning_rate": 1.6239434582924696e-06, + "loss": 0.8185, + "step": 8541 + }, + { + "epoch": 0.8216621777606772, + "grad_norm": 2.421504483792138, + "learning_rate": 1.6222419859313443e-06, + "loss": 0.8579, + "step": 8542 + }, + { + "epoch": 0.821758368603309, + "grad_norm": 2.211354695481937, + "learning_rate": 1.6205413267037651e-06, + "loss": 0.9766, + "step": 8543 + }, + { + "epoch": 0.8218545594459408, + "grad_norm": 2.124816405250101, + "learning_rate": 1.6188414807747999e-06, + "loss": 0.9399, + "step": 8544 + }, + { + "epoch": 0.8219507502885726, + "grad_norm": 1.9458896251420406, + "learning_rate": 1.61714244830943e-06, + "loss": 0.943, + "step": 8545 + }, + { + "epoch": 0.8220469411312044, + "grad_norm": 1.9139226679384242, + "learning_rate": 1.6154442294725636e-06, + "loss": 0.9545, + "step": 8546 + }, + { + "epoch": 0.822143131973836, + "grad_norm": 2.187538000108518, + "learning_rate": 1.6137468244290278e-06, + "loss": 0.9305, + "step": 8547 + }, + { + "epoch": 0.8222393228164678, + "grad_norm": 2.0060350000702654, + "learning_rate": 1.6120502333435695e-06, + "loss": 0.8338, + "step": 8548 + }, + { + "epoch": 0.8223355136590996, + "grad_norm": 2.129140562925111, + "learning_rate": 1.6103544563808548e-06, + "loss": 0.9829, + "step": 8549 + }, + { + "epoch": 0.8224317045017314, + "grad_norm": 2.112887395441764, + "learning_rate": 1.6086594937054767e-06, + "loss": 0.8435, + "step": 8550 + }, + { + "epoch": 0.8225278953443632, + "grad_norm": 1.8347654782303524, + "learning_rate": 1.6069653454819467e-06, + "loss": 0.861, + "step": 8551 + }, + { + "epoch": 0.822624086186995, + "grad_norm": 1.9336151339932872, + "learning_rate": 1.6052720118746923e-06, + "loss": 0.8878, + "step": 8552 + }, + { + "epoch": 0.8227202770296268, + "grad_norm": 2.064067647813545, + "learning_rate": 1.6035794930480708e-06, + "loss": 1.0045, + "step": 8553 + }, + { + "epoch": 0.8228164678722586, + "grad_norm": 1.9953671255023668, + "learning_rate": 1.6018877891663521e-06, + "loss": 0.9411, + "step": 8554 + }, + { + "epoch": 0.8229126587148904, + "grad_norm": 2.0766302214252694, + "learning_rate": 1.6001969003937345e-06, + "loss": 0.9858, + "step": 8555 + }, + { + "epoch": 0.8230088495575221, + "grad_norm": 1.9498307057887707, + "learning_rate": 1.5985068268943283e-06, + "loss": 0.884, + "step": 8556 + }, + { + "epoch": 0.8231050404001539, + "grad_norm": 1.9928999034485853, + "learning_rate": 1.5968175688321719e-06, + "loss": 0.865, + "step": 8557 + }, + { + "epoch": 0.8232012312427857, + "grad_norm": 1.7908717904155256, + "learning_rate": 1.5951291263712255e-06, + "loss": 0.8058, + "step": 8558 + }, + { + "epoch": 0.8232974220854175, + "grad_norm": 2.03136083457223, + "learning_rate": 1.5934414996753632e-06, + "loss": 0.9317, + "step": 8559 + }, + { + "epoch": 0.8233936129280492, + "grad_norm": 2.0830966866644216, + "learning_rate": 1.5917546889083834e-06, + "loss": 0.9797, + "step": 8560 + }, + { + "epoch": 0.823489803770681, + "grad_norm": 2.046146000179601, + "learning_rate": 1.5900686942340093e-06, + "loss": 0.9523, + "step": 8561 + }, + { + "epoch": 0.8235859946133128, + "grad_norm": 2.0245060641656885, + "learning_rate": 1.5883835158158767e-06, + "loss": 0.9064, + "step": 8562 + }, + { + "epoch": 0.8236821854559446, + "grad_norm": 2.0283984037248004, + "learning_rate": 1.586699153817548e-06, + "loss": 0.9114, + "step": 8563 + }, + { + "epoch": 0.8237783762985764, + "grad_norm": 1.99702138521711, + "learning_rate": 1.5850156084025091e-06, + "loss": 0.9243, + "step": 8564 + }, + { + "epoch": 0.8238745671412081, + "grad_norm": 2.0424289891809746, + "learning_rate": 1.5833328797341585e-06, + "loss": 0.7687, + "step": 8565 + }, + { + "epoch": 0.8239707579838399, + "grad_norm": 2.1139068935039935, + "learning_rate": 1.5816509679758185e-06, + "loss": 0.9368, + "step": 8566 + }, + { + "epoch": 0.8240669488264717, + "grad_norm": 2.0007289604510436, + "learning_rate": 1.5799698732907343e-06, + "loss": 0.8574, + "step": 8567 + }, + { + "epoch": 0.8241631396691035, + "grad_norm": 2.1220532218667056, + "learning_rate": 1.578289595842074e-06, + "loss": 0.8363, + "step": 8568 + }, + { + "epoch": 0.8242593305117353, + "grad_norm": 1.8220704395818592, + "learning_rate": 1.5766101357929187e-06, + "loss": 0.8882, + "step": 8569 + }, + { + "epoch": 0.8243555213543671, + "grad_norm": 2.080689042670111, + "learning_rate": 1.5749314933062754e-06, + "loss": 0.9978, + "step": 8570 + }, + { + "epoch": 0.8244517121969989, + "grad_norm": 1.976066894026698, + "learning_rate": 1.5732536685450728e-06, + "loss": 0.9557, + "step": 8571 + }, + { + "epoch": 0.8245479030396307, + "grad_norm": 2.11177947208273, + "learning_rate": 1.5715766616721584e-06, + "loss": 0.9339, + "step": 8572 + }, + { + "epoch": 0.8246440938822625, + "grad_norm": 2.1128262335433785, + "learning_rate": 1.5699004728502953e-06, + "loss": 0.8828, + "step": 8573 + }, + { + "epoch": 0.8247402847248941, + "grad_norm": 2.0442961916991504, + "learning_rate": 1.5682251022421757e-06, + "loss": 0.8681, + "step": 8574 + }, + { + "epoch": 0.8248364755675259, + "grad_norm": 1.7789269674093038, + "learning_rate": 1.5665505500104106e-06, + "loss": 0.8387, + "step": 8575 + }, + { + "epoch": 0.8249326664101577, + "grad_norm": 1.8231254529424161, + "learning_rate": 1.5648768163175277e-06, + "loss": 0.8783, + "step": 8576 + }, + { + "epoch": 0.8250288572527895, + "grad_norm": 2.066600930217257, + "learning_rate": 1.5632039013259747e-06, + "loss": 0.7279, + "step": 8577 + }, + { + "epoch": 0.8251250480954213, + "grad_norm": 2.0168330539023653, + "learning_rate": 1.5615318051981243e-06, + "loss": 0.9352, + "step": 8578 + }, + { + "epoch": 0.8252212389380531, + "grad_norm": 2.188515557621439, + "learning_rate": 1.559860528096272e-06, + "loss": 0.9551, + "step": 8579 + }, + { + "epoch": 0.8253174297806849, + "grad_norm": 2.0073866395223012, + "learning_rate": 1.5581900701826226e-06, + "loss": 0.9192, + "step": 8580 + }, + { + "epoch": 0.8254136206233167, + "grad_norm": 3.379353719816458, + "learning_rate": 1.5565204316193117e-06, + "loss": 0.8259, + "step": 8581 + }, + { + "epoch": 0.8255098114659485, + "grad_norm": 1.9103210636669163, + "learning_rate": 1.5548516125683976e-06, + "loss": 0.872, + "step": 8582 + }, + { + "epoch": 0.8256060023085803, + "grad_norm": 2.1336647850630537, + "learning_rate": 1.5531836131918421e-06, + "loss": 0.9475, + "step": 8583 + }, + { + "epoch": 0.825702193151212, + "grad_norm": 2.2774025405088643, + "learning_rate": 1.5515164336515465e-06, + "loss": 0.9211, + "step": 8584 + }, + { + "epoch": 0.8257983839938438, + "grad_norm": 2.053904193578837, + "learning_rate": 1.5498500741093224e-06, + "loss": 0.806, + "step": 8585 + }, + { + "epoch": 0.8258945748364755, + "grad_norm": 1.896020649871728, + "learning_rate": 1.5481845347269077e-06, + "loss": 0.9193, + "step": 8586 + }, + { + "epoch": 0.8259907656791073, + "grad_norm": 2.1147760368522546, + "learning_rate": 1.5465198156659533e-06, + "loss": 0.8251, + "step": 8587 + }, + { + "epoch": 0.8260869565217391, + "grad_norm": 1.6579136899162812, + "learning_rate": 1.5448559170880373e-06, + "loss": 0.7799, + "step": 8588 + }, + { + "epoch": 0.8261831473643709, + "grad_norm": 2.18215425001434, + "learning_rate": 1.5431928391546558e-06, + "loss": 0.9704, + "step": 8589 + }, + { + "epoch": 0.8262793382070027, + "grad_norm": 2.1432902093953046, + "learning_rate": 1.5415305820272198e-06, + "loss": 0.8424, + "step": 8590 + }, + { + "epoch": 0.8263755290496345, + "grad_norm": 2.0600768660277344, + "learning_rate": 1.539869145867069e-06, + "loss": 0.859, + "step": 8591 + }, + { + "epoch": 0.8264717198922663, + "grad_norm": 1.856689120523216, + "learning_rate": 1.5382085308354633e-06, + "loss": 0.925, + "step": 8592 + }, + { + "epoch": 0.826567910734898, + "grad_norm": 2.1955730454802236, + "learning_rate": 1.536548737093574e-06, + "loss": 0.9591, + "step": 8593 + }, + { + "epoch": 0.8266641015775298, + "grad_norm": 1.702137688724279, + "learning_rate": 1.534889764802503e-06, + "loss": 0.7937, + "step": 8594 + }, + { + "epoch": 0.8267602924201616, + "grad_norm": 1.8663051187942363, + "learning_rate": 1.533231614123264e-06, + "loss": 0.8954, + "step": 8595 + }, + { + "epoch": 0.8268564832627934, + "grad_norm": 1.9740687231755125, + "learning_rate": 1.5315742852167992e-06, + "loss": 0.9166, + "step": 8596 + }, + { + "epoch": 0.8269526741054252, + "grad_norm": 2.0057855359210937, + "learning_rate": 1.5299177782439612e-06, + "loss": 0.8768, + "step": 8597 + }, + { + "epoch": 0.827048864948057, + "grad_norm": 2.054048184619807, + "learning_rate": 1.5282620933655312e-06, + "loss": 0.9725, + "step": 8598 + }, + { + "epoch": 0.8271450557906888, + "grad_norm": 2.2400725623897086, + "learning_rate": 1.5266072307422098e-06, + "loss": 0.9898, + "step": 8599 + }, + { + "epoch": 0.8272412466333205, + "grad_norm": 2.012467357323519, + "learning_rate": 1.5249531905346138e-06, + "loss": 0.926, + "step": 8600 + }, + { + "epoch": 0.8273374374759523, + "grad_norm": 2.086028971545372, + "learning_rate": 1.5232999729032793e-06, + "loss": 0.9662, + "step": 8601 + }, + { + "epoch": 0.827433628318584, + "grad_norm": 1.7741863944416199, + "learning_rate": 1.521647578008667e-06, + "loss": 0.7887, + "step": 8602 + }, + { + "epoch": 0.8275298191612158, + "grad_norm": 1.9940939137572087, + "learning_rate": 1.5199960060111596e-06, + "loss": 0.9512, + "step": 8603 + }, + { + "epoch": 0.8276260100038476, + "grad_norm": 2.1043286225296702, + "learning_rate": 1.5183452570710522e-06, + "loss": 0.9099, + "step": 8604 + }, + { + "epoch": 0.8277222008464794, + "grad_norm": 2.184139607216336, + "learning_rate": 1.5166953313485667e-06, + "loss": 0.8802, + "step": 8605 + }, + { + "epoch": 0.8278183916891112, + "grad_norm": 2.138234026605065, + "learning_rate": 1.5150462290038392e-06, + "loss": 0.9538, + "step": 8606 + }, + { + "epoch": 0.827914582531743, + "grad_norm": 1.8954698788373865, + "learning_rate": 1.5133979501969343e-06, + "loss": 0.8819, + "step": 8607 + }, + { + "epoch": 0.8280107733743748, + "grad_norm": 1.9301918312351172, + "learning_rate": 1.511750495087827e-06, + "loss": 1.0016, + "step": 8608 + }, + { + "epoch": 0.8281069642170066, + "grad_norm": 2.0604607182342947, + "learning_rate": 1.5101038638364184e-06, + "loss": 0.9639, + "step": 8609 + }, + { + "epoch": 0.8282031550596384, + "grad_norm": 2.158954599699346, + "learning_rate": 1.5084580566025309e-06, + "loss": 0.8673, + "step": 8610 + }, + { + "epoch": 0.8282993459022701, + "grad_norm": 1.971728981198663, + "learning_rate": 1.506813073545902e-06, + "loss": 0.9596, + "step": 8611 + }, + { + "epoch": 0.8283955367449019, + "grad_norm": 2.119178891614609, + "learning_rate": 1.5051689148261895e-06, + "loss": 0.9179, + "step": 8612 + }, + { + "epoch": 0.8284917275875336, + "grad_norm": 2.115106239403763, + "learning_rate": 1.5035255806029768e-06, + "loss": 0.9118, + "step": 8613 + }, + { + "epoch": 0.8285879184301654, + "grad_norm": 1.854025150023163, + "learning_rate": 1.5018830710357612e-06, + "loss": 0.8513, + "step": 8614 + }, + { + "epoch": 0.8286841092727972, + "grad_norm": 1.910001421462307, + "learning_rate": 1.5002413862839628e-06, + "loss": 0.8148, + "step": 8615 + }, + { + "epoch": 0.828780300115429, + "grad_norm": 2.1549829723469673, + "learning_rate": 1.4986005265069204e-06, + "loss": 0.9449, + "step": 8616 + }, + { + "epoch": 0.8288764909580608, + "grad_norm": 2.0076136970462795, + "learning_rate": 1.4969604918639014e-06, + "loss": 0.9308, + "step": 8617 + }, + { + "epoch": 0.8289726818006926, + "grad_norm": 1.8920354339638132, + "learning_rate": 1.4953212825140728e-06, + "loss": 0.8896, + "step": 8618 + }, + { + "epoch": 0.8290688726433244, + "grad_norm": 2.2179803505279767, + "learning_rate": 1.4936828986165396e-06, + "loss": 0.8709, + "step": 8619 + }, + { + "epoch": 0.8291650634859561, + "grad_norm": 2.086992388688582, + "learning_rate": 1.4920453403303249e-06, + "loss": 0.9228, + "step": 8620 + }, + { + "epoch": 0.8292612543285879, + "grad_norm": 1.9208731801462235, + "learning_rate": 1.490408607814362e-06, + "loss": 0.8223, + "step": 8621 + }, + { + "epoch": 0.8293574451712197, + "grad_norm": 1.81492553522846, + "learning_rate": 1.4887727012275112e-06, + "loss": 0.8285, + "step": 8622 + }, + { + "epoch": 0.8294536360138515, + "grad_norm": 2.000717798490841, + "learning_rate": 1.4871376207285549e-06, + "loss": 0.8021, + "step": 8623 + }, + { + "epoch": 0.8295498268564833, + "grad_norm": 2.0533811807179743, + "learning_rate": 1.4855033664761898e-06, + "loss": 0.97, + "step": 8624 + }, + { + "epoch": 0.8296460176991151, + "grad_norm": 1.6797087613505988, + "learning_rate": 1.4838699386290311e-06, + "loss": 0.7248, + "step": 8625 + }, + { + "epoch": 0.8297422085417469, + "grad_norm": 2.0455607857203213, + "learning_rate": 1.48223733734562e-06, + "loss": 0.9503, + "step": 8626 + }, + { + "epoch": 0.8298383993843786, + "grad_norm": 1.7326966173050247, + "learning_rate": 1.480605562784415e-06, + "loss": 0.8586, + "step": 8627 + }, + { + "epoch": 0.8299345902270104, + "grad_norm": 2.084732836331876, + "learning_rate": 1.4789746151037942e-06, + "loss": 0.9551, + "step": 8628 + }, + { + "epoch": 0.8300307810696421, + "grad_norm": 1.9840756459664037, + "learning_rate": 1.4773444944620519e-06, + "loss": 0.8753, + "step": 8629 + }, + { + "epoch": 0.8301269719122739, + "grad_norm": 2.001440049689466, + "learning_rate": 1.475715201017407e-06, + "loss": 0.9095, + "step": 8630 + }, + { + "epoch": 0.8302231627549057, + "grad_norm": 2.149510892172494, + "learning_rate": 1.4740867349279997e-06, + "loss": 0.8918, + "step": 8631 + }, + { + "epoch": 0.8303193535975375, + "grad_norm": 1.944974095009438, + "learning_rate": 1.4724590963518803e-06, + "loss": 0.9113, + "step": 8632 + }, + { + "epoch": 0.8304155444401693, + "grad_norm": 2.022535782136591, + "learning_rate": 1.4708322854470291e-06, + "loss": 0.8691, + "step": 8633 + }, + { + "epoch": 0.8305117352828011, + "grad_norm": 1.7433063718254451, + "learning_rate": 1.4692063023713444e-06, + "loss": 0.7754, + "step": 8634 + }, + { + "epoch": 0.8306079261254329, + "grad_norm": 1.8411593421089758, + "learning_rate": 1.4675811472826395e-06, + "loss": 0.8523, + "step": 8635 + }, + { + "epoch": 0.8307041169680647, + "grad_norm": 1.7536222069345124, + "learning_rate": 1.4659568203386464e-06, + "loss": 0.8542, + "step": 8636 + }, + { + "epoch": 0.8308003078106965, + "grad_norm": 1.7513092472204643, + "learning_rate": 1.4643333216970223e-06, + "loss": 0.9062, + "step": 8637 + }, + { + "epoch": 0.8308964986533282, + "grad_norm": 1.6026223878204815, + "learning_rate": 1.4627106515153456e-06, + "loss": 0.7363, + "step": 8638 + }, + { + "epoch": 0.83099268949596, + "grad_norm": 1.962257019265362, + "learning_rate": 1.4610888099511045e-06, + "loss": 0.9183, + "step": 8639 + }, + { + "epoch": 0.8310888803385917, + "grad_norm": 1.914650369026646, + "learning_rate": 1.4594677971617178e-06, + "loss": 0.9613, + "step": 8640 + }, + { + "epoch": 0.8311850711812235, + "grad_norm": 1.9974498729646877, + "learning_rate": 1.457847613304516e-06, + "loss": 0.8255, + "step": 8641 + }, + { + "epoch": 0.8312812620238553, + "grad_norm": 2.0282422738177512, + "learning_rate": 1.4562282585367493e-06, + "loss": 0.8516, + "step": 8642 + }, + { + "epoch": 0.8313774528664871, + "grad_norm": 1.8996381594240306, + "learning_rate": 1.454609733015594e-06, + "loss": 0.8548, + "step": 8643 + }, + { + "epoch": 0.8314736437091189, + "grad_norm": 1.9619889117189366, + "learning_rate": 1.452992036898142e-06, + "loss": 0.8741, + "step": 8644 + }, + { + "epoch": 0.8315698345517507, + "grad_norm": 1.811643590864478, + "learning_rate": 1.4513751703414025e-06, + "loss": 0.9182, + "step": 8645 + }, + { + "epoch": 0.8316660253943825, + "grad_norm": 1.7665649216008983, + "learning_rate": 1.4497591335023087e-06, + "loss": 0.9883, + "step": 8646 + }, + { + "epoch": 0.8317622162370142, + "grad_norm": 2.074383066458232, + "learning_rate": 1.4481439265377074e-06, + "loss": 0.9337, + "step": 8647 + }, + { + "epoch": 0.831858407079646, + "grad_norm": 2.019158314563175, + "learning_rate": 1.446529549604373e-06, + "loss": 0.9398, + "step": 8648 + }, + { + "epoch": 0.8319545979222778, + "grad_norm": 2.192357840281267, + "learning_rate": 1.4449160028589892e-06, + "loss": 0.9453, + "step": 8649 + }, + { + "epoch": 0.8320507887649096, + "grad_norm": 1.9255708461709233, + "learning_rate": 1.4433032864581687e-06, + "loss": 0.9501, + "step": 8650 + }, + { + "epoch": 0.8321469796075414, + "grad_norm": 1.9535845028094554, + "learning_rate": 1.4416914005584393e-06, + "loss": 0.9026, + "step": 8651 + }, + { + "epoch": 0.8322431704501732, + "grad_norm": 2.0691112751942535, + "learning_rate": 1.4400803453162482e-06, + "loss": 0.9566, + "step": 8652 + }, + { + "epoch": 0.832339361292805, + "grad_norm": 1.9450576372743265, + "learning_rate": 1.4384701208879591e-06, + "loss": 0.9119, + "step": 8653 + }, + { + "epoch": 0.8324355521354367, + "grad_norm": 2.1414172980844586, + "learning_rate": 1.4368607274298596e-06, + "loss": 0.9773, + "step": 8654 + }, + { + "epoch": 0.8325317429780685, + "grad_norm": 1.8508295958039809, + "learning_rate": 1.4352521650981587e-06, + "loss": 0.9992, + "step": 8655 + }, + { + "epoch": 0.8326279338207002, + "grad_norm": 1.9448050455783852, + "learning_rate": 1.4336444340489775e-06, + "loss": 0.8476, + "step": 8656 + }, + { + "epoch": 0.832724124663332, + "grad_norm": 2.3977687810175037, + "learning_rate": 1.432037534438362e-06, + "loss": 0.9262, + "step": 8657 + }, + { + "epoch": 0.8328203155059638, + "grad_norm": 2.009641573156605, + "learning_rate": 1.430431466422273e-06, + "loss": 0.921, + "step": 8658 + }, + { + "epoch": 0.8329165063485956, + "grad_norm": 2.461942933422088, + "learning_rate": 1.428826230156598e-06, + "loss": 0.9229, + "step": 8659 + }, + { + "epoch": 0.8330126971912274, + "grad_norm": 2.3274653258448517, + "learning_rate": 1.4272218257971327e-06, + "loss": 0.9432, + "step": 8660 + }, + { + "epoch": 0.8331088880338592, + "grad_norm": 1.5383123252129183, + "learning_rate": 1.4256182534996033e-06, + "loss": 0.706, + "step": 8661 + }, + { + "epoch": 0.833205078876491, + "grad_norm": 1.8592983447645732, + "learning_rate": 1.4240155134196499e-06, + "loss": 0.8982, + "step": 8662 + }, + { + "epoch": 0.8333012697191228, + "grad_norm": 2.06633139030785, + "learning_rate": 1.4224136057128312e-06, + "loss": 1.0045, + "step": 8663 + }, + { + "epoch": 0.8333974605617546, + "grad_norm": 1.9571610954658551, + "learning_rate": 1.4208125305346232e-06, + "loss": 0.9071, + "step": 8664 + }, + { + "epoch": 0.8334936514043862, + "grad_norm": 2.033644604236915, + "learning_rate": 1.419212288040428e-06, + "loss": 0.8643, + "step": 8665 + }, + { + "epoch": 0.833589842247018, + "grad_norm": 1.8884716368371488, + "learning_rate": 1.4176128783855636e-06, + "loss": 0.8449, + "step": 8666 + }, + { + "epoch": 0.8336860330896498, + "grad_norm": 2.017120214158284, + "learning_rate": 1.4160143017252613e-06, + "loss": 0.9036, + "step": 8667 + }, + { + "epoch": 0.8337822239322816, + "grad_norm": 1.9471227610350819, + "learning_rate": 1.4144165582146819e-06, + "loss": 0.9127, + "step": 8668 + }, + { + "epoch": 0.8338784147749134, + "grad_norm": 2.4211833619854395, + "learning_rate": 1.4128196480089019e-06, + "loss": 0.8488, + "step": 8669 + }, + { + "epoch": 0.8339746056175452, + "grad_norm": 1.9096349261052337, + "learning_rate": 1.4112235712629063e-06, + "loss": 0.9276, + "step": 8670 + }, + { + "epoch": 0.834070796460177, + "grad_norm": 1.9403561527006576, + "learning_rate": 1.4096283281316148e-06, + "loss": 0.9674, + "step": 8671 + }, + { + "epoch": 0.8341669873028088, + "grad_norm": 2.1848137258743647, + "learning_rate": 1.40803391876986e-06, + "loss": 0.868, + "step": 8672 + }, + { + "epoch": 0.8342631781454406, + "grad_norm": 2.1484998005834846, + "learning_rate": 1.4064403433323881e-06, + "loss": 0.8973, + "step": 8673 + }, + { + "epoch": 0.8343593689880723, + "grad_norm": 2.149675431714583, + "learning_rate": 1.4048476019738756e-06, + "loss": 0.9297, + "step": 8674 + }, + { + "epoch": 0.8344555598307041, + "grad_norm": 2.2324792125699493, + "learning_rate": 1.403255694848905e-06, + "loss": 0.9523, + "step": 8675 + }, + { + "epoch": 0.8345517506733359, + "grad_norm": 1.9300830850250335, + "learning_rate": 1.4016646221119912e-06, + "loss": 0.8994, + "step": 8676 + }, + { + "epoch": 0.8346479415159677, + "grad_norm": 2.1251796217233045, + "learning_rate": 1.4000743839175557e-06, + "loss": 0.8834, + "step": 8677 + }, + { + "epoch": 0.8347441323585995, + "grad_norm": 2.27206444864006, + "learning_rate": 1.3984849804199485e-06, + "loss": 0.9478, + "step": 8678 + }, + { + "epoch": 0.8348403232012312, + "grad_norm": 1.9262103349935682, + "learning_rate": 1.3968964117734352e-06, + "loss": 0.7252, + "step": 8679 + }, + { + "epoch": 0.834936514043863, + "grad_norm": 1.9944297332523522, + "learning_rate": 1.395308678132199e-06, + "loss": 0.8678, + "step": 8680 + }, + { + "epoch": 0.8350327048864948, + "grad_norm": 2.0303589489426925, + "learning_rate": 1.3937217796503411e-06, + "loss": 0.9703, + "step": 8681 + }, + { + "epoch": 0.8351288957291266, + "grad_norm": 1.9987968921378476, + "learning_rate": 1.392135716481885e-06, + "loss": 0.9578, + "step": 8682 + }, + { + "epoch": 0.8352250865717584, + "grad_norm": 1.9924174534174799, + "learning_rate": 1.3905504887807753e-06, + "loss": 0.846, + "step": 8683 + }, + { + "epoch": 0.8353212774143901, + "grad_norm": 2.2492645130158073, + "learning_rate": 1.3889660967008656e-06, + "loss": 0.9135, + "step": 8684 + }, + { + "epoch": 0.8354174682570219, + "grad_norm": 1.983378323774453, + "learning_rate": 1.3873825403959384e-06, + "loss": 0.8734, + "step": 8685 + }, + { + "epoch": 0.8355136590996537, + "grad_norm": 1.9497385541817374, + "learning_rate": 1.3857998200196943e-06, + "loss": 0.9565, + "step": 8686 + }, + { + "epoch": 0.8356098499422855, + "grad_norm": 1.9133656864325477, + "learning_rate": 1.3842179357257456e-06, + "loss": 0.9099, + "step": 8687 + }, + { + "epoch": 0.8357060407849173, + "grad_norm": 2.000238914995273, + "learning_rate": 1.3826368876676278e-06, + "loss": 0.9595, + "step": 8688 + }, + { + "epoch": 0.8358022316275491, + "grad_norm": 2.1860351062427466, + "learning_rate": 1.3810566759987965e-06, + "loss": 0.9287, + "step": 8689 + }, + { + "epoch": 0.8358984224701809, + "grad_norm": 2.119289765686162, + "learning_rate": 1.379477300872626e-06, + "loss": 0.98, + "step": 8690 + }, + { + "epoch": 0.8359946133128127, + "grad_norm": 2.0429756605501614, + "learning_rate": 1.3778987624424057e-06, + "loss": 0.901, + "step": 8691 + }, + { + "epoch": 0.8360908041554445, + "grad_norm": 2.346329428771618, + "learning_rate": 1.3763210608613497e-06, + "loss": 0.9202, + "step": 8692 + }, + { + "epoch": 0.8361869949980761, + "grad_norm": 2.1813874084722777, + "learning_rate": 1.3747441962825858e-06, + "loss": 1.0079, + "step": 8693 + }, + { + "epoch": 0.8362831858407079, + "grad_norm": 1.890659094367112, + "learning_rate": 1.3731681688591593e-06, + "loss": 0.9216, + "step": 8694 + }, + { + "epoch": 0.8363793766833397, + "grad_norm": 1.8028921783276557, + "learning_rate": 1.3715929787440408e-06, + "loss": 0.9053, + "step": 8695 + }, + { + "epoch": 0.8364755675259715, + "grad_norm": 2.071378697897013, + "learning_rate": 1.370018626090116e-06, + "loss": 0.902, + "step": 8696 + }, + { + "epoch": 0.8365717583686033, + "grad_norm": 2.0714795248058935, + "learning_rate": 1.3684451110501873e-06, + "loss": 0.8419, + "step": 8697 + }, + { + "epoch": 0.8366679492112351, + "grad_norm": 1.9259181814747612, + "learning_rate": 1.3668724337769823e-06, + "loss": 0.8031, + "step": 8698 + }, + { + "epoch": 0.8367641400538669, + "grad_norm": 1.863874954107373, + "learning_rate": 1.3653005944231356e-06, + "loss": 0.9355, + "step": 8699 + }, + { + "epoch": 0.8368603308964987, + "grad_norm": 2.0369665874535903, + "learning_rate": 1.3637295931412153e-06, + "loss": 0.9076, + "step": 8700 + }, + { + "epoch": 0.8369565217391305, + "grad_norm": 1.865394068375341, + "learning_rate": 1.362159430083695e-06, + "loss": 0.9333, + "step": 8701 + }, + { + "epoch": 0.8370527125817622, + "grad_norm": 1.9898323342589905, + "learning_rate": 1.3605901054029746e-06, + "loss": 0.9669, + "step": 8702 + }, + { + "epoch": 0.837148903424394, + "grad_norm": 1.8386358736472066, + "learning_rate": 1.3590216192513727e-06, + "loss": 0.9897, + "step": 8703 + }, + { + "epoch": 0.8372450942670258, + "grad_norm": 1.9677791165958185, + "learning_rate": 1.3574539717811231e-06, + "loss": 0.8826, + "step": 8704 + }, + { + "epoch": 0.8373412851096576, + "grad_norm": 2.1401524177256115, + "learning_rate": 1.355887163144376e-06, + "loss": 0.8354, + "step": 8705 + }, + { + "epoch": 0.8374374759522893, + "grad_norm": 2.1045662039370465, + "learning_rate": 1.3543211934932065e-06, + "loss": 0.9206, + "step": 8706 + }, + { + "epoch": 0.8375336667949211, + "grad_norm": 2.087808460855201, + "learning_rate": 1.3527560629796076e-06, + "loss": 1.0046, + "step": 8707 + }, + { + "epoch": 0.8376298576375529, + "grad_norm": 1.917796400052326, + "learning_rate": 1.3511917717554846e-06, + "loss": 0.8939, + "step": 8708 + }, + { + "epoch": 0.8377260484801847, + "grad_norm": 1.6098591612991215, + "learning_rate": 1.349628319972669e-06, + "loss": 0.7239, + "step": 8709 + }, + { + "epoch": 0.8378222393228165, + "grad_norm": 1.9242029583321487, + "learning_rate": 1.348065707782904e-06, + "loss": 0.776, + "step": 8710 + }, + { + "epoch": 0.8379184301654482, + "grad_norm": 2.0049783221012434, + "learning_rate": 1.3465039353378572e-06, + "loss": 0.8795, + "step": 8711 + }, + { + "epoch": 0.83801462100808, + "grad_norm": 2.2362694530816682, + "learning_rate": 1.3449430027891096e-06, + "loss": 0.9502, + "step": 8712 + }, + { + "epoch": 0.8381108118507118, + "grad_norm": 2.201311236003835, + "learning_rate": 1.3433829102881646e-06, + "loss": 0.9907, + "step": 8713 + }, + { + "epoch": 0.8382070026933436, + "grad_norm": 1.9668306436991723, + "learning_rate": 1.3418236579864452e-06, + "loss": 0.9243, + "step": 8714 + }, + { + "epoch": 0.8383031935359754, + "grad_norm": 2.1914079604070698, + "learning_rate": 1.3402652460352872e-06, + "loss": 0.9093, + "step": 8715 + }, + { + "epoch": 0.8383993843786072, + "grad_norm": 1.9850681527891294, + "learning_rate": 1.338707674585945e-06, + "loss": 0.9613, + "step": 8716 + }, + { + "epoch": 0.838495575221239, + "grad_norm": 2.367675364574272, + "learning_rate": 1.3371509437895991e-06, + "loss": 0.8796, + "step": 8717 + }, + { + "epoch": 0.8385917660638708, + "grad_norm": 2.0186753460031324, + "learning_rate": 1.3355950537973438e-06, + "loss": 0.8972, + "step": 8718 + }, + { + "epoch": 0.8386879569065026, + "grad_norm": 2.023477718904297, + "learning_rate": 1.3340400047601876e-06, + "loss": 0.9615, + "step": 8719 + }, + { + "epoch": 0.8387841477491342, + "grad_norm": 1.9318231190689867, + "learning_rate": 1.332485796829065e-06, + "loss": 0.8198, + "step": 8720 + }, + { + "epoch": 0.838880338591766, + "grad_norm": 2.0242427509893335, + "learning_rate": 1.3309324301548276e-06, + "loss": 0.9349, + "step": 8721 + }, + { + "epoch": 0.8389765294343978, + "grad_norm": 2.1510424914610526, + "learning_rate": 1.329379904888235e-06, + "loss": 0.8833, + "step": 8722 + }, + { + "epoch": 0.8390727202770296, + "grad_norm": 1.915771432837434, + "learning_rate": 1.3278282211799776e-06, + "loss": 0.8203, + "step": 8723 + }, + { + "epoch": 0.8391689111196614, + "grad_norm": 1.6303909895212638, + "learning_rate": 1.3262773791806617e-06, + "loss": 0.7918, + "step": 8724 + }, + { + "epoch": 0.8392651019622932, + "grad_norm": 2.0826413295173816, + "learning_rate": 1.3247273790408055e-06, + "loss": 0.954, + "step": 8725 + }, + { + "epoch": 0.839361292804925, + "grad_norm": 2.0320175853227065, + "learning_rate": 1.3231782209108546e-06, + "loss": 0.8556, + "step": 8726 + }, + { + "epoch": 0.8394574836475568, + "grad_norm": 2.0636312822807565, + "learning_rate": 1.321629904941164e-06, + "loss": 0.9915, + "step": 8727 + }, + { + "epoch": 0.8395536744901886, + "grad_norm": 2.216624763201214, + "learning_rate": 1.3200824312820137e-06, + "loss": 0.9224, + "step": 8728 + }, + { + "epoch": 0.8396498653328203, + "grad_norm": 2.3027400015202995, + "learning_rate": 1.318535800083598e-06, + "loss": 0.9038, + "step": 8729 + }, + { + "epoch": 0.8397460561754521, + "grad_norm": 2.2303869159439227, + "learning_rate": 1.3169900114960298e-06, + "loss": 0.9285, + "step": 8730 + }, + { + "epoch": 0.8398422470180839, + "grad_norm": 1.8406874055645317, + "learning_rate": 1.3154450656693452e-06, + "loss": 0.8308, + "step": 8731 + }, + { + "epoch": 0.8399384378607156, + "grad_norm": 1.8660461531659798, + "learning_rate": 1.3139009627534927e-06, + "loss": 0.9505, + "step": 8732 + }, + { + "epoch": 0.8400346287033474, + "grad_norm": 1.9137844319452944, + "learning_rate": 1.312357702898337e-06, + "loss": 0.863, + "step": 8733 + }, + { + "epoch": 0.8401308195459792, + "grad_norm": 2.0107545367309525, + "learning_rate": 1.3108152862536683e-06, + "loss": 0.896, + "step": 8734 + }, + { + "epoch": 0.840227010388611, + "grad_norm": 1.9034069935349125, + "learning_rate": 1.3092737129691923e-06, + "loss": 0.7978, + "step": 8735 + }, + { + "epoch": 0.8403232012312428, + "grad_norm": 2.2643248296302034, + "learning_rate": 1.3077329831945295e-06, + "loss": 0.9455, + "step": 8736 + }, + { + "epoch": 0.8404193920738746, + "grad_norm": 2.0850974352308125, + "learning_rate": 1.3061930970792214e-06, + "loss": 0.9046, + "step": 8737 + }, + { + "epoch": 0.8405155829165063, + "grad_norm": 2.0198178632367956, + "learning_rate": 1.3046540547727305e-06, + "loss": 0.9137, + "step": 8738 + }, + { + "epoch": 0.8406117737591381, + "grad_norm": 1.7712824300454146, + "learning_rate": 1.3031158564244306e-06, + "loss": 0.7437, + "step": 8739 + }, + { + "epoch": 0.8407079646017699, + "grad_norm": 1.911927541750486, + "learning_rate": 1.3015785021836159e-06, + "loss": 0.7908, + "step": 8740 + }, + { + "epoch": 0.8408041554444017, + "grad_norm": 2.0498000663717457, + "learning_rate": 1.3000419921995022e-06, + "loss": 0.934, + "step": 8741 + }, + { + "epoch": 0.8409003462870335, + "grad_norm": 1.934105170273305, + "learning_rate": 1.2985063266212229e-06, + "loss": 0.9502, + "step": 8742 + }, + { + "epoch": 0.8409965371296653, + "grad_norm": 1.9310908673264349, + "learning_rate": 1.2969715055978226e-06, + "loss": 0.9383, + "step": 8743 + }, + { + "epoch": 0.8410927279722971, + "grad_norm": 2.0004910046723254, + "learning_rate": 1.295437529278275e-06, + "loss": 0.8928, + "step": 8744 + }, + { + "epoch": 0.8411889188149289, + "grad_norm": 2.2281936320502864, + "learning_rate": 1.2939043978114618e-06, + "loss": 0.9309, + "step": 8745 + }, + { + "epoch": 0.8412851096575606, + "grad_norm": 2.077147400186654, + "learning_rate": 1.2923721113461852e-06, + "loss": 0.8796, + "step": 8746 + }, + { + "epoch": 0.8413813005001923, + "grad_norm": 2.2178908199747727, + "learning_rate": 1.2908406700311682e-06, + "loss": 0.9846, + "step": 8747 + }, + { + "epoch": 0.8414774913428241, + "grad_norm": 2.137822041010853, + "learning_rate": 1.2893100740150522e-06, + "loss": 0.9989, + "step": 8748 + }, + { + "epoch": 0.8415736821854559, + "grad_norm": 2.1076555012024762, + "learning_rate": 1.287780323446397e-06, + "loss": 0.8171, + "step": 8749 + }, + { + "epoch": 0.8416698730280877, + "grad_norm": 2.2463023620102414, + "learning_rate": 1.2862514184736695e-06, + "loss": 0.8763, + "step": 8750 + }, + { + "epoch": 0.8417660638707195, + "grad_norm": 1.6528077561594672, + "learning_rate": 1.284723359245269e-06, + "loss": 0.7988, + "step": 8751 + }, + { + "epoch": 0.8418622547133513, + "grad_norm": 2.04404973240456, + "learning_rate": 1.2831961459095088e-06, + "loss": 0.7734, + "step": 8752 + }, + { + "epoch": 0.8419584455559831, + "grad_norm": 2.0353897738537814, + "learning_rate": 1.2816697786146126e-06, + "loss": 0.862, + "step": 8753 + }, + { + "epoch": 0.8420546363986149, + "grad_norm": 2.295010047230991, + "learning_rate": 1.2801442575087296e-06, + "loss": 0.9792, + "step": 8754 + }, + { + "epoch": 0.8421508272412467, + "grad_norm": 1.9549123785577425, + "learning_rate": 1.2786195827399273e-06, + "loss": 0.9248, + "step": 8755 + }, + { + "epoch": 0.8422470180838784, + "grad_norm": 2.000050151403622, + "learning_rate": 1.2770957544561868e-06, + "loss": 0.8966, + "step": 8756 + }, + { + "epoch": 0.8423432089265102, + "grad_norm": 2.0158678276546333, + "learning_rate": 1.2755727728054068e-06, + "loss": 0.9026, + "step": 8757 + }, + { + "epoch": 0.842439399769142, + "grad_norm": 2.21511088811811, + "learning_rate": 1.274050637935408e-06, + "loss": 0.9103, + "step": 8758 + }, + { + "epoch": 0.8425355906117737, + "grad_norm": 2.2676791417162394, + "learning_rate": 1.2725293499939272e-06, + "loss": 0.9439, + "step": 8759 + }, + { + "epoch": 0.8426317814544055, + "grad_norm": 2.2135690059709052, + "learning_rate": 1.2710089091286148e-06, + "loss": 0.8547, + "step": 8760 + }, + { + "epoch": 0.8427279722970373, + "grad_norm": 2.0724008035949706, + "learning_rate": 1.2694893154870492e-06, + "loss": 0.8789, + "step": 8761 + }, + { + "epoch": 0.8428241631396691, + "grad_norm": 1.913557080399493, + "learning_rate": 1.2679705692167122e-06, + "loss": 0.9528, + "step": 8762 + }, + { + "epoch": 0.8429203539823009, + "grad_norm": 1.7609276585537073, + "learning_rate": 1.2664526704650182e-06, + "loss": 0.7513, + "step": 8763 + }, + { + "epoch": 0.8430165448249327, + "grad_norm": 2.0260821640059428, + "learning_rate": 1.2649356193792873e-06, + "loss": 0.9422, + "step": 8764 + }, + { + "epoch": 0.8431127356675644, + "grad_norm": 1.6326098858365536, + "learning_rate": 1.2634194161067636e-06, + "loss": 0.7396, + "step": 8765 + }, + { + "epoch": 0.8432089265101962, + "grad_norm": 2.293394356065538, + "learning_rate": 1.261904060794612e-06, + "loss": 0.9099, + "step": 8766 + }, + { + "epoch": 0.843305117352828, + "grad_norm": 2.0104553818193476, + "learning_rate": 1.2603895535899059e-06, + "loss": 0.8795, + "step": 8767 + }, + { + "epoch": 0.8434013081954598, + "grad_norm": 2.3397141208709624, + "learning_rate": 1.2588758946396417e-06, + "loss": 0.9666, + "step": 8768 + }, + { + "epoch": 0.8434974990380916, + "grad_norm": 2.1191737902880816, + "learning_rate": 1.2573630840907324e-06, + "loss": 0.9337, + "step": 8769 + }, + { + "epoch": 0.8435936898807234, + "grad_norm": 2.2800939689721056, + "learning_rate": 1.2558511220900138e-06, + "loss": 0.9576, + "step": 8770 + }, + { + "epoch": 0.8436898807233552, + "grad_norm": 2.349717362923563, + "learning_rate": 1.2543400087842306e-06, + "loss": 0.9039, + "step": 8771 + }, + { + "epoch": 0.843786071565987, + "grad_norm": 1.8892551587051956, + "learning_rate": 1.2528297443200489e-06, + "loss": 0.8322, + "step": 8772 + }, + { + "epoch": 0.8438822624086187, + "grad_norm": 1.7715830608788121, + "learning_rate": 1.2513203288440601e-06, + "loss": 0.8103, + "step": 8773 + }, + { + "epoch": 0.8439784532512504, + "grad_norm": 1.8050424206639373, + "learning_rate": 1.2498117625027562e-06, + "loss": 0.8816, + "step": 8774 + }, + { + "epoch": 0.8440746440938822, + "grad_norm": 2.2163068286102776, + "learning_rate": 1.2483040454425598e-06, + "loss": 0.8756, + "step": 8775 + }, + { + "epoch": 0.844170834936514, + "grad_norm": 1.5554440840419723, + "learning_rate": 1.246797177809812e-06, + "loss": 0.7224, + "step": 8776 + }, + { + "epoch": 0.8442670257791458, + "grad_norm": 1.9041851902422742, + "learning_rate": 1.245291159750761e-06, + "loss": 0.8833, + "step": 8777 + }, + { + "epoch": 0.8443632166217776, + "grad_norm": 1.8598159076796326, + "learning_rate": 1.2437859914115847e-06, + "loss": 0.74, + "step": 8778 + }, + { + "epoch": 0.8444594074644094, + "grad_norm": 2.251500998049379, + "learning_rate": 1.242281672938368e-06, + "loss": 0.9118, + "step": 8779 + }, + { + "epoch": 0.8445555983070412, + "grad_norm": 2.084675314713605, + "learning_rate": 1.2407782044771222e-06, + "loss": 0.9348, + "step": 8780 + }, + { + "epoch": 0.844651789149673, + "grad_norm": 1.8994323240520725, + "learning_rate": 1.239275586173766e-06, + "loss": 0.8608, + "step": 8781 + }, + { + "epoch": 0.8447479799923048, + "grad_norm": 2.1895320309409594, + "learning_rate": 1.237773818174146e-06, + "loss": 0.9504, + "step": 8782 + }, + { + "epoch": 0.8448441708349365, + "grad_norm": 2.2349228409798103, + "learning_rate": 1.2362729006240226e-06, + "loss": 0.8325, + "step": 8783 + }, + { + "epoch": 0.8449403616775683, + "grad_norm": 2.2863370563799132, + "learning_rate": 1.23477283366907e-06, + "loss": 0.9176, + "step": 8784 + }, + { + "epoch": 0.8450365525202, + "grad_norm": 2.172950706698506, + "learning_rate": 1.2332736174548832e-06, + "loss": 0.7644, + "step": 8785 + }, + { + "epoch": 0.8451327433628318, + "grad_norm": 2.0346734582173007, + "learning_rate": 1.2317752521269722e-06, + "loss": 0.8955, + "step": 8786 + }, + { + "epoch": 0.8452289342054636, + "grad_norm": 2.1497772823023, + "learning_rate": 1.2302777378307717e-06, + "loss": 0.8521, + "step": 8787 + }, + { + "epoch": 0.8453251250480954, + "grad_norm": 2.008698875342654, + "learning_rate": 1.2287810747116224e-06, + "loss": 0.9585, + "step": 8788 + }, + { + "epoch": 0.8454213158907272, + "grad_norm": 1.8713316109521134, + "learning_rate": 1.2272852629147913e-06, + "loss": 0.9173, + "step": 8789 + }, + { + "epoch": 0.845517506733359, + "grad_norm": 2.095174902573671, + "learning_rate": 1.225790302585461e-06, + "loss": 0.9508, + "step": 8790 + }, + { + "epoch": 0.8456136975759908, + "grad_norm": 1.9705704716295522, + "learning_rate": 1.2242961938687292e-06, + "loss": 0.9041, + "step": 8791 + }, + { + "epoch": 0.8457098884186226, + "grad_norm": 1.9040727688495944, + "learning_rate": 1.2228029369096094e-06, + "loss": 0.9079, + "step": 8792 + }, + { + "epoch": 0.8458060792612543, + "grad_norm": 1.885119182245373, + "learning_rate": 1.2213105318530372e-06, + "loss": 0.8311, + "step": 8793 + }, + { + "epoch": 0.8459022701038861, + "grad_norm": 1.9194260991175744, + "learning_rate": 1.2198189788438652e-06, + "loss": 0.9366, + "step": 8794 + }, + { + "epoch": 0.8459984609465179, + "grad_norm": 1.8147723256775197, + "learning_rate": 1.2183282780268579e-06, + "loss": 0.9171, + "step": 8795 + }, + { + "epoch": 0.8460946517891497, + "grad_norm": 2.0278908917444864, + "learning_rate": 1.216838429546704e-06, + "loss": 0.8478, + "step": 8796 + }, + { + "epoch": 0.8461908426317815, + "grad_norm": 1.8204450974818565, + "learning_rate": 1.215349433548003e-06, + "loss": 0.8969, + "step": 8797 + }, + { + "epoch": 0.8462870334744133, + "grad_norm": 1.6858953809498367, + "learning_rate": 1.2138612901752777e-06, + "loss": 0.7829, + "step": 8798 + }, + { + "epoch": 0.846383224317045, + "grad_norm": 1.9644887282158832, + "learning_rate": 1.2123739995729621e-06, + "loss": 0.9576, + "step": 8799 + }, + { + "epoch": 0.8464794151596768, + "grad_norm": 1.9864757013162646, + "learning_rate": 1.2108875618854122e-06, + "loss": 0.9151, + "step": 8800 + }, + { + "epoch": 0.8465756060023086, + "grad_norm": 1.9674915714941439, + "learning_rate": 1.2094019772569033e-06, + "loss": 0.8704, + "step": 8801 + }, + { + "epoch": 0.8466717968449403, + "grad_norm": 3.1432277553209023, + "learning_rate": 1.2079172458316168e-06, + "loss": 0.8887, + "step": 8802 + }, + { + "epoch": 0.8467679876875721, + "grad_norm": 1.8964413392409218, + "learning_rate": 1.2064333677536622e-06, + "loss": 0.8922, + "step": 8803 + }, + { + "epoch": 0.8468641785302039, + "grad_norm": 2.0275500053508044, + "learning_rate": 1.204950343167065e-06, + "loss": 0.9374, + "step": 8804 + }, + { + "epoch": 0.8469603693728357, + "grad_norm": 1.9304699603650792, + "learning_rate": 1.2034681722157615e-06, + "loss": 0.9262, + "step": 8805 + }, + { + "epoch": 0.8470565602154675, + "grad_norm": 1.7904501029707296, + "learning_rate": 1.2019868550436099e-06, + "loss": 0.9371, + "step": 8806 + }, + { + "epoch": 0.8471527510580993, + "grad_norm": 1.950806568198777, + "learning_rate": 1.2005063917943894e-06, + "loss": 0.868, + "step": 8807 + }, + { + "epoch": 0.8472489419007311, + "grad_norm": 2.0790247887951128, + "learning_rate": 1.1990267826117874e-06, + "loss": 0.9151, + "step": 8808 + }, + { + "epoch": 0.8473451327433629, + "grad_norm": 2.045425604023194, + "learning_rate": 1.1975480276394114e-06, + "loss": 0.7999, + "step": 8809 + }, + { + "epoch": 0.8474413235859947, + "grad_norm": 2.4896453959015155, + "learning_rate": 1.1960701270207885e-06, + "loss": 0.9202, + "step": 8810 + }, + { + "epoch": 0.8475375144286263, + "grad_norm": 1.8537641927068709, + "learning_rate": 1.1945930808993655e-06, + "loss": 0.9379, + "step": 8811 + }, + { + "epoch": 0.8476337052712581, + "grad_norm": 1.9339169148662798, + "learning_rate": 1.1931168894184974e-06, + "loss": 0.8876, + "step": 8812 + }, + { + "epoch": 0.8477298961138899, + "grad_norm": 2.1150589441215906, + "learning_rate": 1.1916415527214664e-06, + "loss": 0.9347, + "step": 8813 + }, + { + "epoch": 0.8478260869565217, + "grad_norm": 1.9787386302343761, + "learning_rate": 1.19016707095146e-06, + "loss": 0.9308, + "step": 8814 + }, + { + "epoch": 0.8479222777991535, + "grad_norm": 2.4793009287070893, + "learning_rate": 1.1886934442515962e-06, + "loss": 1.029, + "step": 8815 + }, + { + "epoch": 0.8480184686417853, + "grad_norm": 1.8605153040165394, + "learning_rate": 1.187220672764897e-06, + "loss": 0.8711, + "step": 8816 + }, + { + "epoch": 0.8481146594844171, + "grad_norm": 1.9298899094133537, + "learning_rate": 1.1857487566343108e-06, + "loss": 0.954, + "step": 8817 + }, + { + "epoch": 0.8482108503270489, + "grad_norm": 1.843352661165117, + "learning_rate": 1.1842776960027014e-06, + "loss": 0.8573, + "step": 8818 + }, + { + "epoch": 0.8483070411696807, + "grad_norm": 2.1179308910462282, + "learning_rate": 1.1828074910128462e-06, + "loss": 1.0172, + "step": 8819 + }, + { + "epoch": 0.8484032320123124, + "grad_norm": 2.327792130777586, + "learning_rate": 1.1813381418074388e-06, + "loss": 0.8606, + "step": 8820 + }, + { + "epoch": 0.8484994228549442, + "grad_norm": 1.7732389033152156, + "learning_rate": 1.179869648529095e-06, + "loss": 0.8128, + "step": 8821 + }, + { + "epoch": 0.848595613697576, + "grad_norm": 2.126372057913107, + "learning_rate": 1.1784020113203453e-06, + "loss": 0.972, + "step": 8822 + }, + { + "epoch": 0.8486918045402078, + "grad_norm": 2.0722070342258654, + "learning_rate": 1.1769352303236337e-06, + "loss": 0.8883, + "step": 8823 + }, + { + "epoch": 0.8487879953828396, + "grad_norm": 2.0212767228883544, + "learning_rate": 1.1754693056813272e-06, + "loss": 0.9054, + "step": 8824 + }, + { + "epoch": 0.8488841862254713, + "grad_norm": 2.0820288679236354, + "learning_rate": 1.1740042375357042e-06, + "loss": 0.9526, + "step": 8825 + }, + { + "epoch": 0.8489803770681031, + "grad_norm": 2.1417362837323846, + "learning_rate": 1.172540026028962e-06, + "loss": 0.8793, + "step": 8826 + }, + { + "epoch": 0.8490765679107349, + "grad_norm": 2.046984513915157, + "learning_rate": 1.171076671303214e-06, + "loss": 0.8549, + "step": 8827 + }, + { + "epoch": 0.8491727587533667, + "grad_norm": 1.8707108698516177, + "learning_rate": 1.169614173500494e-06, + "loss": 0.8447, + "step": 8828 + }, + { + "epoch": 0.8492689495959984, + "grad_norm": 1.9250466478124177, + "learning_rate": 1.168152532762752e-06, + "loss": 0.8474, + "step": 8829 + }, + { + "epoch": 0.8493651404386302, + "grad_norm": 1.971489591685738, + "learning_rate": 1.1666917492318486e-06, + "loss": 0.8473, + "step": 8830 + }, + { + "epoch": 0.849461331281262, + "grad_norm": 2.005397128404614, + "learning_rate": 1.1652318230495652e-06, + "loss": 0.9024, + "step": 8831 + }, + { + "epoch": 0.8495575221238938, + "grad_norm": 2.1491001516579407, + "learning_rate": 1.1637727543576027e-06, + "loss": 0.8198, + "step": 8832 + }, + { + "epoch": 0.8496537129665256, + "grad_norm": 1.8007008085182203, + "learning_rate": 1.1623145432975747e-06, + "loss": 0.8549, + "step": 8833 + }, + { + "epoch": 0.8497499038091574, + "grad_norm": 1.9439586316567241, + "learning_rate": 1.1608571900110122e-06, + "loss": 0.8649, + "step": 8834 + }, + { + "epoch": 0.8498460946517892, + "grad_norm": 1.9416321377041605, + "learning_rate": 1.1594006946393687e-06, + "loss": 0.8968, + "step": 8835 + }, + { + "epoch": 0.849942285494421, + "grad_norm": 2.0004593332983993, + "learning_rate": 1.1579450573240058e-06, + "loss": 0.9016, + "step": 8836 + }, + { + "epoch": 0.8500384763370528, + "grad_norm": 1.9938612074110407, + "learning_rate": 1.1564902782062038e-06, + "loss": 0.947, + "step": 8837 + }, + { + "epoch": 0.8501346671796844, + "grad_norm": 1.680300977104406, + "learning_rate": 1.1550363574271638e-06, + "loss": 0.8924, + "step": 8838 + }, + { + "epoch": 0.8502308580223162, + "grad_norm": 1.8257903757688203, + "learning_rate": 1.1535832951280046e-06, + "loss": 0.8758, + "step": 8839 + }, + { + "epoch": 0.850327048864948, + "grad_norm": 2.0918377773392334, + "learning_rate": 1.1521310914497518e-06, + "loss": 0.9596, + "step": 8840 + }, + { + "epoch": 0.8504232397075798, + "grad_norm": 2.031520380628256, + "learning_rate": 1.1506797465333585e-06, + "loss": 0.9551, + "step": 8841 + }, + { + "epoch": 0.8505194305502116, + "grad_norm": 2.007276055487443, + "learning_rate": 1.149229260519691e-06, + "loss": 0.818, + "step": 8842 + }, + { + "epoch": 0.8506156213928434, + "grad_norm": 2.2381547625287905, + "learning_rate": 1.1477796335495306e-06, + "loss": 0.8016, + "step": 8843 + }, + { + "epoch": 0.8507118122354752, + "grad_norm": 1.8498011430387107, + "learning_rate": 1.1463308657635718e-06, + "loss": 0.9153, + "step": 8844 + }, + { + "epoch": 0.850808003078107, + "grad_norm": 1.8942477688300572, + "learning_rate": 1.1448829573024345e-06, + "loss": 0.9337, + "step": 8845 + }, + { + "epoch": 0.8509041939207388, + "grad_norm": 2.001539028361862, + "learning_rate": 1.1434359083066515e-06, + "loss": 0.9986, + "step": 8846 + }, + { + "epoch": 0.8510003847633705, + "grad_norm": 1.7539928345417173, + "learning_rate": 1.1419897189166684e-06, + "loss": 0.8619, + "step": 8847 + }, + { + "epoch": 0.8510965756060023, + "grad_norm": 1.8955624891092306, + "learning_rate": 1.140544389272853e-06, + "loss": 0.915, + "step": 8848 + }, + { + "epoch": 0.8511927664486341, + "grad_norm": 1.8507555903856896, + "learning_rate": 1.139099919515483e-06, + "loss": 0.8133, + "step": 8849 + }, + { + "epoch": 0.8512889572912659, + "grad_norm": 2.1037910824240518, + "learning_rate": 1.1376563097847616e-06, + "loss": 0.8276, + "step": 8850 + }, + { + "epoch": 0.8513851481338977, + "grad_norm": 1.8465006237179908, + "learning_rate": 1.136213560220798e-06, + "loss": 0.9182, + "step": 8851 + }, + { + "epoch": 0.8514813389765294, + "grad_norm": 2.064130010680119, + "learning_rate": 1.1347716709636282e-06, + "loss": 0.8341, + "step": 8852 + }, + { + "epoch": 0.8515775298191612, + "grad_norm": 2.141685385135148, + "learning_rate": 1.133330642153201e-06, + "loss": 0.9002, + "step": 8853 + }, + { + "epoch": 0.851673720661793, + "grad_norm": 1.6319607927162292, + "learning_rate": 1.1318904739293745e-06, + "loss": 0.7701, + "step": 8854 + }, + { + "epoch": 0.8517699115044248, + "grad_norm": 1.9694240067834647, + "learning_rate": 1.130451166431933e-06, + "loss": 0.8747, + "step": 8855 + }, + { + "epoch": 0.8518661023470565, + "grad_norm": 1.8718409752326337, + "learning_rate": 1.129012719800575e-06, + "loss": 0.9289, + "step": 8856 + }, + { + "epoch": 0.8519622931896883, + "grad_norm": 2.259621712377048, + "learning_rate": 1.1275751341749108e-06, + "loss": 0.8537, + "step": 8857 + }, + { + "epoch": 0.8520584840323201, + "grad_norm": 1.9744504366628, + "learning_rate": 1.1261384096944728e-06, + "loss": 0.9363, + "step": 8858 + }, + { + "epoch": 0.8521546748749519, + "grad_norm": 1.745190550793436, + "learning_rate": 1.1247025464987094e-06, + "loss": 0.7998, + "step": 8859 + }, + { + "epoch": 0.8522508657175837, + "grad_norm": 2.0646928083937968, + "learning_rate": 1.1232675447269803e-06, + "loss": 0.6825, + "step": 8860 + }, + { + "epoch": 0.8523470565602155, + "grad_norm": 2.138072462717194, + "learning_rate": 1.1218334045185642e-06, + "loss": 0.9124, + "step": 8861 + }, + { + "epoch": 0.8524432474028473, + "grad_norm": 2.332898498007731, + "learning_rate": 1.1204001260126574e-06, + "loss": 0.9117, + "step": 8862 + }, + { + "epoch": 0.8525394382454791, + "grad_norm": 1.9636073624576478, + "learning_rate": 1.1189677093483752e-06, + "loss": 0.8552, + "step": 8863 + }, + { + "epoch": 0.8526356290881109, + "grad_norm": 2.02201082200275, + "learning_rate": 1.1175361546647413e-06, + "loss": 0.9027, + "step": 8864 + }, + { + "epoch": 0.8527318199307425, + "grad_norm": 1.9151668304094422, + "learning_rate": 1.1161054621007051e-06, + "loss": 0.9449, + "step": 8865 + }, + { + "epoch": 0.8528280107733743, + "grad_norm": 1.9804040491027037, + "learning_rate": 1.1146756317951224e-06, + "loss": 0.8763, + "step": 8866 + }, + { + "epoch": 0.8529242016160061, + "grad_norm": 2.299308378574845, + "learning_rate": 1.1132466638867757e-06, + "loss": 0.8423, + "step": 8867 + }, + { + "epoch": 0.8530203924586379, + "grad_norm": 1.86839052248357, + "learning_rate": 1.1118185585143536e-06, + "loss": 0.8639, + "step": 8868 + }, + { + "epoch": 0.8531165833012697, + "grad_norm": 1.927447748129872, + "learning_rate": 1.1103913158164692e-06, + "loss": 0.8951, + "step": 8869 + }, + { + "epoch": 0.8532127741439015, + "grad_norm": 1.9810610861412377, + "learning_rate": 1.1089649359316501e-06, + "loss": 0.8072, + "step": 8870 + }, + { + "epoch": 0.8533089649865333, + "grad_norm": 2.205636393632905, + "learning_rate": 1.1075394189983357e-06, + "loss": 1.0122, + "step": 8871 + }, + { + "epoch": 0.8534051558291651, + "grad_norm": 2.0275725223313428, + "learning_rate": 1.1061147651548855e-06, + "loss": 0.9389, + "step": 8872 + }, + { + "epoch": 0.8535013466717969, + "grad_norm": 1.8920005918762308, + "learning_rate": 1.1046909745395728e-06, + "loss": 0.7263, + "step": 8873 + }, + { + "epoch": 0.8535975375144286, + "grad_norm": 2.0721687499615022, + "learning_rate": 1.1032680472905932e-06, + "loss": 0.8383, + "step": 8874 + }, + { + "epoch": 0.8536937283570604, + "grad_norm": 2.196327375199569, + "learning_rate": 1.1018459835460504e-06, + "loss": 0.9331, + "step": 8875 + }, + { + "epoch": 0.8537899191996922, + "grad_norm": 2.1309836397214448, + "learning_rate": 1.1004247834439697e-06, + "loss": 1.0247, + "step": 8876 + }, + { + "epoch": 0.853886110042324, + "grad_norm": 1.8553504224042217, + "learning_rate": 1.0990044471222882e-06, + "loss": 0.8646, + "step": 8877 + }, + { + "epoch": 0.8539823008849557, + "grad_norm": 2.0007254946941155, + "learning_rate": 1.097584974718866e-06, + "loss": 0.8227, + "step": 8878 + }, + { + "epoch": 0.8540784917275875, + "grad_norm": 2.0448476950297665, + "learning_rate": 1.0961663663714718e-06, + "loss": 0.8648, + "step": 8879 + }, + { + "epoch": 0.8541746825702193, + "grad_norm": 2.110738549042147, + "learning_rate": 1.0947486222177928e-06, + "loss": 0.8498, + "step": 8880 + }, + { + "epoch": 0.8542708734128511, + "grad_norm": 2.168545911654912, + "learning_rate": 1.093331742395438e-06, + "loss": 0.9414, + "step": 8881 + }, + { + "epoch": 0.8543670642554829, + "grad_norm": 2.086311270580897, + "learning_rate": 1.0919157270419257e-06, + "loss": 0.8107, + "step": 8882 + }, + { + "epoch": 0.8544632550981146, + "grad_norm": 2.339633728468106, + "learning_rate": 1.0905005762946884e-06, + "loss": 0.8308, + "step": 8883 + }, + { + "epoch": 0.8545594459407464, + "grad_norm": 2.529509305984353, + "learning_rate": 1.0890862902910849e-06, + "loss": 0.8323, + "step": 8884 + }, + { + "epoch": 0.8546556367833782, + "grad_norm": 1.9889817276129225, + "learning_rate": 1.087672869168378e-06, + "loss": 0.943, + "step": 8885 + }, + { + "epoch": 0.85475182762601, + "grad_norm": 1.9789626008761019, + "learning_rate": 1.0862603130637562e-06, + "loss": 0.8931, + "step": 8886 + }, + { + "epoch": 0.8548480184686418, + "grad_norm": 2.0879169519050604, + "learning_rate": 1.0848486221143206e-06, + "loss": 0.8849, + "step": 8887 + }, + { + "epoch": 0.8549442093112736, + "grad_norm": 2.0387804168731565, + "learning_rate": 1.0834377964570863e-06, + "loss": 0.9057, + "step": 8888 + }, + { + "epoch": 0.8550404001539054, + "grad_norm": 2.0796885463984602, + "learning_rate": 1.082027836228985e-06, + "loss": 0.9144, + "step": 8889 + }, + { + "epoch": 0.8551365909965372, + "grad_norm": 2.1949260063546534, + "learning_rate": 1.0806187415668668e-06, + "loss": 0.9244, + "step": 8890 + }, + { + "epoch": 0.855232781839169, + "grad_norm": 2.226691563957958, + "learning_rate": 1.0792105126074982e-06, + "loss": 0.9844, + "step": 8891 + }, + { + "epoch": 0.8553289726818007, + "grad_norm": 2.2551353255065223, + "learning_rate": 1.0778031494875574e-06, + "loss": 0.9447, + "step": 8892 + }, + { + "epoch": 0.8554251635244324, + "grad_norm": 2.0613708942537565, + "learning_rate": 1.076396652343641e-06, + "loss": 0.9411, + "step": 8893 + }, + { + "epoch": 0.8555213543670642, + "grad_norm": 1.8447798425498383, + "learning_rate": 1.0749910213122649e-06, + "loss": 0.7864, + "step": 8894 + }, + { + "epoch": 0.855617545209696, + "grad_norm": 2.0696254588816867, + "learning_rate": 1.073586256529855e-06, + "loss": 1.0012, + "step": 8895 + }, + { + "epoch": 0.8557137360523278, + "grad_norm": 2.0273412567709452, + "learning_rate": 1.072182358132755e-06, + "loss": 0.886, + "step": 8896 + }, + { + "epoch": 0.8558099268949596, + "grad_norm": 1.8534626427848107, + "learning_rate": 1.0707793262572263e-06, + "loss": 0.8064, + "step": 8897 + }, + { + "epoch": 0.8559061177375914, + "grad_norm": 2.183778895964188, + "learning_rate": 1.0693771610394477e-06, + "loss": 0.9724, + "step": 8898 + }, + { + "epoch": 0.8560023085802232, + "grad_norm": 2.026323648842757, + "learning_rate": 1.067975862615509e-06, + "loss": 0.8522, + "step": 8899 + }, + { + "epoch": 0.856098499422855, + "grad_norm": 2.211465815139668, + "learning_rate": 1.066575431121417e-06, + "loss": 1.0211, + "step": 8900 + }, + { + "epoch": 0.8561946902654868, + "grad_norm": 1.9802401131766074, + "learning_rate": 1.0651758666930967e-06, + "loss": 0.8019, + "step": 8901 + }, + { + "epoch": 0.8562908811081185, + "grad_norm": 2.2193057044173403, + "learning_rate": 1.06377716946639e-06, + "loss": 0.9673, + "step": 8902 + }, + { + "epoch": 0.8563870719507503, + "grad_norm": 1.9153151535581188, + "learning_rate": 1.0623793395770498e-06, + "loss": 0.9246, + "step": 8903 + }, + { + "epoch": 0.856483262793382, + "grad_norm": 1.8209847480617396, + "learning_rate": 1.0609823771607487e-06, + "loss": 0.9627, + "step": 8904 + }, + { + "epoch": 0.8565794536360138, + "grad_norm": 1.9197350165744895, + "learning_rate": 1.0595862823530768e-06, + "loss": 0.8064, + "step": 8905 + }, + { + "epoch": 0.8566756444786456, + "grad_norm": 2.052229121025186, + "learning_rate": 1.0581910552895302e-06, + "loss": 0.9052, + "step": 8906 + }, + { + "epoch": 0.8567718353212774, + "grad_norm": 2.005757078368611, + "learning_rate": 1.0567966961055321e-06, + "loss": 0.9293, + "step": 8907 + }, + { + "epoch": 0.8568680261639092, + "grad_norm": 2.169788088468882, + "learning_rate": 1.055403204936416e-06, + "loss": 0.9044, + "step": 8908 + }, + { + "epoch": 0.856964217006541, + "grad_norm": 2.1372122070787016, + "learning_rate": 1.054010581917435e-06, + "loss": 0.9469, + "step": 8909 + }, + { + "epoch": 0.8570604078491728, + "grad_norm": 2.4804488870878565, + "learning_rate": 1.0526188271837512e-06, + "loss": 0.8857, + "step": 8910 + }, + { + "epoch": 0.8571565986918045, + "grad_norm": 1.9800628168271583, + "learning_rate": 1.0512279408704506e-06, + "loss": 0.9241, + "step": 8911 + }, + { + "epoch": 0.8572527895344363, + "grad_norm": 2.085386854708331, + "learning_rate": 1.0498379231125278e-06, + "loss": 0.8646, + "step": 8912 + }, + { + "epoch": 0.8573489803770681, + "grad_norm": 1.8609701233934064, + "learning_rate": 1.0484487740448945e-06, + "loss": 0.8337, + "step": 8913 + }, + { + "epoch": 0.8574451712196999, + "grad_norm": 1.9355503698204732, + "learning_rate": 1.047060493802381e-06, + "loss": 0.7669, + "step": 8914 + }, + { + "epoch": 0.8575413620623317, + "grad_norm": 2.026872183152867, + "learning_rate": 1.0456730825197347e-06, + "loss": 0.9201, + "step": 8915 + }, + { + "epoch": 0.8576375529049635, + "grad_norm": 1.545242051838363, + "learning_rate": 1.0442865403316117e-06, + "loss": 0.7422, + "step": 8916 + }, + { + "epoch": 0.8577337437475953, + "grad_norm": 1.7847762079350014, + "learning_rate": 1.0429008673725904e-06, + "loss": 0.8104, + "step": 8917 + }, + { + "epoch": 0.857829934590227, + "grad_norm": 2.170782664852275, + "learning_rate": 1.0415160637771604e-06, + "loss": 0.9524, + "step": 8918 + }, + { + "epoch": 0.8579261254328588, + "grad_norm": 1.8490290748341234, + "learning_rate": 1.040132129679733e-06, + "loss": 0.8466, + "step": 8919 + }, + { + "epoch": 0.8580223162754905, + "grad_norm": 1.8187817996146234, + "learning_rate": 1.0387490652146236e-06, + "loss": 0.8428, + "step": 8920 + }, + { + "epoch": 0.8581185071181223, + "grad_norm": 2.042999671265767, + "learning_rate": 1.0373668705160767e-06, + "loss": 1.0162, + "step": 8921 + }, + { + "epoch": 0.8582146979607541, + "grad_norm": 2.1149989565221157, + "learning_rate": 1.0359855457182455e-06, + "loss": 1.0332, + "step": 8922 + }, + { + "epoch": 0.8583108888033859, + "grad_norm": 2.078665637484803, + "learning_rate": 1.0346050909551985e-06, + "loss": 0.9276, + "step": 8923 + }, + { + "epoch": 0.8584070796460177, + "grad_norm": 2.128426570863344, + "learning_rate": 1.0332255063609177e-06, + "loss": 0.7918, + "step": 8924 + }, + { + "epoch": 0.8585032704886495, + "grad_norm": 2.0501606311581066, + "learning_rate": 1.0318467920693065e-06, + "loss": 0.9351, + "step": 8925 + }, + { + "epoch": 0.8585994613312813, + "grad_norm": 1.9628638367803828, + "learning_rate": 1.0304689482141839e-06, + "loss": 0.8765, + "step": 8926 + }, + { + "epoch": 0.8586956521739131, + "grad_norm": 2.0865616815903705, + "learning_rate": 1.029091974929276e-06, + "loss": 0.8631, + "step": 8927 + }, + { + "epoch": 0.8587918430165449, + "grad_norm": 1.7786031169934897, + "learning_rate": 1.027715872348234e-06, + "loss": 0.8374, + "step": 8928 + }, + { + "epoch": 0.8588880338591766, + "grad_norm": 1.6941766282254513, + "learning_rate": 1.0263406406046173e-06, + "loss": 0.901, + "step": 8929 + }, + { + "epoch": 0.8589842247018084, + "grad_norm": 2.1464473382328726, + "learning_rate": 1.0249662798319072e-06, + "loss": 1.0435, + "step": 8930 + }, + { + "epoch": 0.8590804155444401, + "grad_norm": 1.9872253904452695, + "learning_rate": 1.0235927901634934e-06, + "loss": 0.9996, + "step": 8931 + }, + { + "epoch": 0.8591766063870719, + "grad_norm": 1.9758271863361139, + "learning_rate": 1.0222201717326885e-06, + "loss": 0.8016, + "step": 8932 + }, + { + "epoch": 0.8592727972297037, + "grad_norm": 2.170335955042446, + "learning_rate": 1.0208484246727169e-06, + "loss": 0.8741, + "step": 8933 + }, + { + "epoch": 0.8593689880723355, + "grad_norm": 2.0719678426180095, + "learning_rate": 1.0194775491167164e-06, + "loss": 0.9017, + "step": 8934 + }, + { + "epoch": 0.8594651789149673, + "grad_norm": 1.879799934574783, + "learning_rate": 1.0181075451977417e-06, + "loss": 0.8557, + "step": 8935 + }, + { + "epoch": 0.8595613697575991, + "grad_norm": 1.968183935832445, + "learning_rate": 1.0167384130487667e-06, + "loss": 0.9842, + "step": 8936 + }, + { + "epoch": 0.8596575606002309, + "grad_norm": 2.012804183224893, + "learning_rate": 1.0153701528026738e-06, + "loss": 0.9023, + "step": 8937 + }, + { + "epoch": 0.8597537514428626, + "grad_norm": 2.3394996877389866, + "learning_rate": 1.0140027645922656e-06, + "loss": 0.962, + "step": 8938 + }, + { + "epoch": 0.8598499422854944, + "grad_norm": 1.7609109314582014, + "learning_rate": 1.0126362485502617e-06, + "loss": 0.7536, + "step": 8939 + }, + { + "epoch": 0.8599461331281262, + "grad_norm": 1.955607382993682, + "learning_rate": 1.0112706048092924e-06, + "loss": 0.9113, + "step": 8940 + }, + { + "epoch": 0.860042323970758, + "grad_norm": 1.8456914169267853, + "learning_rate": 1.009905833501903e-06, + "loss": 0.9186, + "step": 8941 + }, + { + "epoch": 0.8601385148133898, + "grad_norm": 2.2610609209840997, + "learning_rate": 1.0085419347605575e-06, + "loss": 0.9198, + "step": 8942 + }, + { + "epoch": 0.8602347056560216, + "grad_norm": 2.0687940050316502, + "learning_rate": 1.0071789087176364e-06, + "loss": 0.9219, + "step": 8943 + }, + { + "epoch": 0.8603308964986534, + "grad_norm": 1.9395949009270048, + "learning_rate": 1.00581675550543e-06, + "loss": 0.9656, + "step": 8944 + }, + { + "epoch": 0.8604270873412851, + "grad_norm": 2.0798482275635903, + "learning_rate": 1.0044554752561485e-06, + "loss": 0.911, + "step": 8945 + }, + { + "epoch": 0.8605232781839169, + "grad_norm": 1.979850532394952, + "learning_rate": 1.003095068101917e-06, + "loss": 0.9719, + "step": 8946 + }, + { + "epoch": 0.8606194690265486, + "grad_norm": 2.1512502380507925, + "learning_rate": 1.0017355341747736e-06, + "loss": 0.9357, + "step": 8947 + }, + { + "epoch": 0.8607156598691804, + "grad_norm": 2.15666054794705, + "learning_rate": 1.0003768736066722e-06, + "loss": 0.9315, + "step": 8948 + }, + { + "epoch": 0.8608118507118122, + "grad_norm": 1.8874189516739146, + "learning_rate": 9.990190865294812e-07, + "loss": 0.8321, + "step": 8949 + }, + { + "epoch": 0.860908041554444, + "grad_norm": 1.7389313941297644, + "learning_rate": 9.976621730749892e-07, + "loss": 0.9113, + "step": 8950 + }, + { + "epoch": 0.8610042323970758, + "grad_norm": 2.1488531167137728, + "learning_rate": 9.963061333748957e-07, + "loss": 0.9262, + "step": 8951 + }, + { + "epoch": 0.8611004232397076, + "grad_norm": 2.1813430278297155, + "learning_rate": 9.949509675608115e-07, + "loss": 0.9569, + "step": 8952 + }, + { + "epoch": 0.8611966140823394, + "grad_norm": 2.045326828860944, + "learning_rate": 9.935966757642713e-07, + "loss": 0.995, + "step": 8953 + }, + { + "epoch": 0.8612928049249712, + "grad_norm": 1.8071766261964681, + "learning_rate": 9.922432581167207e-07, + "loss": 0.9045, + "step": 8954 + }, + { + "epoch": 0.861388995767603, + "grad_norm": 2.139652574163629, + "learning_rate": 9.908907147495184e-07, + "loss": 0.9331, + "step": 8955 + }, + { + "epoch": 0.8614851866102347, + "grad_norm": 2.0230268810174823, + "learning_rate": 9.895390457939414e-07, + "loss": 0.9384, + "step": 8956 + }, + { + "epoch": 0.8615813774528664, + "grad_norm": 1.9649136221701244, + "learning_rate": 9.881882513811858e-07, + "loss": 0.8637, + "step": 8957 + }, + { + "epoch": 0.8616775682954982, + "grad_norm": 2.2316218763892848, + "learning_rate": 9.86838331642348e-07, + "loss": 1.0089, + "step": 8958 + }, + { + "epoch": 0.86177375913813, + "grad_norm": 2.0184233660345092, + "learning_rate": 9.854892867084554e-07, + "loss": 1.0192, + "step": 8959 + }, + { + "epoch": 0.8618699499807618, + "grad_norm": 2.0063247668116815, + "learning_rate": 9.84141116710442e-07, + "loss": 0.8406, + "step": 8960 + }, + { + "epoch": 0.8619661408233936, + "grad_norm": 1.9306449868790836, + "learning_rate": 9.827938217791643e-07, + "loss": 0.8568, + "step": 8961 + }, + { + "epoch": 0.8620623316660254, + "grad_norm": 1.937255597697611, + "learning_rate": 9.814474020453824e-07, + "loss": 0.8591, + "step": 8962 + }, + { + "epoch": 0.8621585225086572, + "grad_norm": 1.8010367807947192, + "learning_rate": 9.801018576397824e-07, + "loss": 0.86, + "step": 8963 + }, + { + "epoch": 0.862254713351289, + "grad_norm": 1.8695899390211377, + "learning_rate": 9.787571886929604e-07, + "loss": 0.9691, + "step": 8964 + }, + { + "epoch": 0.8623509041939207, + "grad_norm": 1.7709363608790758, + "learning_rate": 9.774133953354248e-07, + "loss": 0.809, + "step": 8965 + }, + { + "epoch": 0.8624470950365525, + "grad_norm": 2.330570863380311, + "learning_rate": 9.76070477697605e-07, + "loss": 0.9509, + "step": 8966 + }, + { + "epoch": 0.8625432858791843, + "grad_norm": 2.1944365308572387, + "learning_rate": 9.74728435909843e-07, + "loss": 0.8655, + "step": 8967 + }, + { + "epoch": 0.8626394767218161, + "grad_norm": 2.0226360411491395, + "learning_rate": 9.733872701023938e-07, + "loss": 0.9765, + "step": 8968 + }, + { + "epoch": 0.8627356675644479, + "grad_norm": 2.0827534717962157, + "learning_rate": 9.720469804054322e-07, + "loss": 0.8207, + "step": 8969 + }, + { + "epoch": 0.8628318584070797, + "grad_norm": 1.959824001080523, + "learning_rate": 9.707075669490407e-07, + "loss": 0.9238, + "step": 8970 + }, + { + "epoch": 0.8629280492497114, + "grad_norm": 2.91116351226782, + "learning_rate": 9.693690298632252e-07, + "loss": 0.818, + "step": 8971 + }, + { + "epoch": 0.8630242400923432, + "grad_norm": 1.8431286289160622, + "learning_rate": 9.680313692778976e-07, + "loss": 0.9732, + "step": 8972 + }, + { + "epoch": 0.863120430934975, + "grad_norm": 2.291111042508238, + "learning_rate": 9.666945853228927e-07, + "loss": 0.8561, + "step": 8973 + }, + { + "epoch": 0.8632166217776067, + "grad_norm": 2.029788584749035, + "learning_rate": 9.653586781279567e-07, + "loss": 0.9628, + "step": 8974 + }, + { + "epoch": 0.8633128126202385, + "grad_norm": 2.06979387520342, + "learning_rate": 9.640236478227516e-07, + "loss": 0.9842, + "step": 8975 + }, + { + "epoch": 0.8634090034628703, + "grad_norm": 1.7358910658682263, + "learning_rate": 9.626894945368492e-07, + "loss": 0.9987, + "step": 8976 + }, + { + "epoch": 0.8635051943055021, + "grad_norm": 1.867268362145197, + "learning_rate": 9.613562183997439e-07, + "loss": 0.8394, + "step": 8977 + }, + { + "epoch": 0.8636013851481339, + "grad_norm": 1.9594607949500022, + "learning_rate": 9.600238195408428e-07, + "loss": 0.9035, + "step": 8978 + }, + { + "epoch": 0.8636975759907657, + "grad_norm": 2.346258034832485, + "learning_rate": 9.58692298089464e-07, + "loss": 0.9193, + "step": 8979 + }, + { + "epoch": 0.8637937668333975, + "grad_norm": 1.8919810038564064, + "learning_rate": 9.573616541748464e-07, + "loss": 0.9065, + "step": 8980 + }, + { + "epoch": 0.8638899576760293, + "grad_norm": 2.0574932425532078, + "learning_rate": 9.560318879261354e-07, + "loss": 0.9172, + "step": 8981 + }, + { + "epoch": 0.8639861485186611, + "grad_norm": 2.1562732664752553, + "learning_rate": 9.547029994724023e-07, + "loss": 0.8166, + "step": 8982 + }, + { + "epoch": 0.8640823393612928, + "grad_norm": 2.0240861014959024, + "learning_rate": 9.533749889426214e-07, + "loss": 0.7754, + "step": 8983 + }, + { + "epoch": 0.8641785302039245, + "grad_norm": 2.0986511197478177, + "learning_rate": 9.520478564656898e-07, + "loss": 0.8751, + "step": 8984 + }, + { + "epoch": 0.8642747210465563, + "grad_norm": 2.227723949631069, + "learning_rate": 9.507216021704202e-07, + "loss": 0.8028, + "step": 8985 + }, + { + "epoch": 0.8643709118891881, + "grad_norm": 2.072747986961221, + "learning_rate": 9.49396226185535e-07, + "loss": 0.9408, + "step": 8986 + }, + { + "epoch": 0.8644671027318199, + "grad_norm": 2.2181082640028786, + "learning_rate": 9.48071728639669e-07, + "loss": 0.8223, + "step": 8987 + }, + { + "epoch": 0.8645632935744517, + "grad_norm": 2.0212821198526454, + "learning_rate": 9.467481096613829e-07, + "loss": 0.9342, + "step": 8988 + }, + { + "epoch": 0.8646594844170835, + "grad_norm": 2.2432114332518815, + "learning_rate": 9.454253693791404e-07, + "loss": 0.9251, + "step": 8989 + }, + { + "epoch": 0.8647556752597153, + "grad_norm": 1.925811100002851, + "learning_rate": 9.441035079213267e-07, + "loss": 0.9367, + "step": 8990 + }, + { + "epoch": 0.8648518661023471, + "grad_norm": 1.9074544575137733, + "learning_rate": 9.427825254162392e-07, + "loss": 0.9068, + "step": 8991 + }, + { + "epoch": 0.8649480569449788, + "grad_norm": 2.0502915486783824, + "learning_rate": 9.414624219920953e-07, + "loss": 0.9138, + "step": 8992 + }, + { + "epoch": 0.8650442477876106, + "grad_norm": 1.91393823840554, + "learning_rate": 9.401431977770148e-07, + "loss": 0.9699, + "step": 8993 + }, + { + "epoch": 0.8651404386302424, + "grad_norm": 2.0293712872934813, + "learning_rate": 9.38824852899043e-07, + "loss": 0.8198, + "step": 8994 + }, + { + "epoch": 0.8652366294728742, + "grad_norm": 2.296604190332723, + "learning_rate": 9.375073874861396e-07, + "loss": 0.9155, + "step": 8995 + }, + { + "epoch": 0.865332820315506, + "grad_norm": 2.233133866964427, + "learning_rate": 9.361908016661703e-07, + "loss": 0.9707, + "step": 8996 + }, + { + "epoch": 0.8654290111581378, + "grad_norm": 1.8015007508190248, + "learning_rate": 9.348750955669239e-07, + "loss": 0.8778, + "step": 8997 + }, + { + "epoch": 0.8655252020007695, + "grad_norm": 2.3933319813324543, + "learning_rate": 9.335602693161039e-07, + "loss": 0.8836, + "step": 8998 + }, + { + "epoch": 0.8656213928434013, + "grad_norm": 1.9435355449924303, + "learning_rate": 9.322463230413226e-07, + "loss": 0.9196, + "step": 8999 + }, + { + "epoch": 0.8657175836860331, + "grad_norm": 1.7872813011992683, + "learning_rate": 9.309332568701079e-07, + "loss": 0.8862, + "step": 9000 + }, + { + "epoch": 0.8658137745286649, + "grad_norm": 2.0427823275004813, + "learning_rate": 9.296210709299069e-07, + "loss": 0.9267, + "step": 9001 + }, + { + "epoch": 0.8659099653712966, + "grad_norm": 2.2417638306629133, + "learning_rate": 9.283097653480788e-07, + "loss": 0.7917, + "step": 9002 + }, + { + "epoch": 0.8660061562139284, + "grad_norm": 2.092099104354167, + "learning_rate": 9.269993402518962e-07, + "loss": 0.8717, + "step": 9003 + }, + { + "epoch": 0.8661023470565602, + "grad_norm": 1.924670077585621, + "learning_rate": 9.256897957685463e-07, + "loss": 0.8331, + "step": 9004 + }, + { + "epoch": 0.866198537899192, + "grad_norm": 1.9200104733571435, + "learning_rate": 9.243811320251317e-07, + "loss": 0.9054, + "step": 9005 + }, + { + "epoch": 0.8662947287418238, + "grad_norm": 1.7612269502436126, + "learning_rate": 9.230733491486721e-07, + "loss": 0.8252, + "step": 9006 + }, + { + "epoch": 0.8663909195844556, + "grad_norm": 1.9125246052154785, + "learning_rate": 9.217664472660959e-07, + "loss": 0.93, + "step": 9007 + }, + { + "epoch": 0.8664871104270874, + "grad_norm": 1.9286246044490845, + "learning_rate": 9.204604265042505e-07, + "loss": 0.9113, + "step": 9008 + }, + { + "epoch": 0.8665833012697192, + "grad_norm": 2.164195788363737, + "learning_rate": 9.191552869898979e-07, + "loss": 0.8839, + "step": 9009 + }, + { + "epoch": 0.866679492112351, + "grad_norm": 2.131952497716736, + "learning_rate": 9.178510288497123e-07, + "loss": 0.9792, + "step": 9010 + }, + { + "epoch": 0.8667756829549826, + "grad_norm": 2.2868827790581574, + "learning_rate": 9.1654765221028e-07, + "loss": 0.7869, + "step": 9011 + }, + { + "epoch": 0.8668718737976144, + "grad_norm": 2.065327188536704, + "learning_rate": 9.15245157198108e-07, + "loss": 0.901, + "step": 9012 + }, + { + "epoch": 0.8669680646402462, + "grad_norm": 2.260705746509806, + "learning_rate": 9.139435439396172e-07, + "loss": 0.9739, + "step": 9013 + }, + { + "epoch": 0.867064255482878, + "grad_norm": 1.8741488064250555, + "learning_rate": 9.126428125611342e-07, + "loss": 0.9632, + "step": 9014 + }, + { + "epoch": 0.8671604463255098, + "grad_norm": 1.9400299028038743, + "learning_rate": 9.113429631889115e-07, + "loss": 0.8437, + "step": 9015 + }, + { + "epoch": 0.8672566371681416, + "grad_norm": 1.8202816289716504, + "learning_rate": 9.10043995949108e-07, + "loss": 0.7853, + "step": 9016 + }, + { + "epoch": 0.8673528280107734, + "grad_norm": 2.0085320947087784, + "learning_rate": 9.087459109677987e-07, + "loss": 0.9296, + "step": 9017 + }, + { + "epoch": 0.8674490188534052, + "grad_norm": 1.8020407460259562, + "learning_rate": 9.074487083709759e-07, + "loss": 0.7517, + "step": 9018 + }, + { + "epoch": 0.867545209696037, + "grad_norm": 2.021135127734194, + "learning_rate": 9.061523882845458e-07, + "loss": 0.9217, + "step": 9019 + }, + { + "epoch": 0.8676414005386687, + "grad_norm": 1.8407787872732082, + "learning_rate": 9.04856950834323e-07, + "loss": 0.9067, + "step": 9020 + }, + { + "epoch": 0.8677375913813005, + "grad_norm": 2.2170721887155787, + "learning_rate": 9.035623961460449e-07, + "loss": 0.9785, + "step": 9021 + }, + { + "epoch": 0.8678337822239323, + "grad_norm": 2.0400541249013324, + "learning_rate": 9.022687243453554e-07, + "loss": 0.8487, + "step": 9022 + }, + { + "epoch": 0.867929973066564, + "grad_norm": 2.087175397807464, + "learning_rate": 9.009759355578218e-07, + "loss": 0.9328, + "step": 9023 + }, + { + "epoch": 0.8680261639091958, + "grad_norm": 1.778976899549927, + "learning_rate": 8.996840299089149e-07, + "loss": 0.8657, + "step": 9024 + }, + { + "epoch": 0.8681223547518276, + "grad_norm": 2.052677258663188, + "learning_rate": 8.983930075240266e-07, + "loss": 0.9525, + "step": 9025 + }, + { + "epoch": 0.8682185455944594, + "grad_norm": 1.8633023571686356, + "learning_rate": 8.971028685284655e-07, + "loss": 0.944, + "step": 9026 + }, + { + "epoch": 0.8683147364370912, + "grad_norm": 2.089641180554067, + "learning_rate": 8.958136130474481e-07, + "loss": 0.8664, + "step": 9027 + }, + { + "epoch": 0.868410927279723, + "grad_norm": 2.0417154593905638, + "learning_rate": 8.945252412061056e-07, + "loss": 0.9755, + "step": 9028 + }, + { + "epoch": 0.8685071181223547, + "grad_norm": 1.81313021391909, + "learning_rate": 8.932377531294867e-07, + "loss": 0.8517, + "step": 9029 + }, + { + "epoch": 0.8686033089649865, + "grad_norm": 1.7879036807329038, + "learning_rate": 8.91951148942557e-07, + "loss": 0.7698, + "step": 9030 + }, + { + "epoch": 0.8686994998076183, + "grad_norm": 2.208414884500028, + "learning_rate": 8.906654287701877e-07, + "loss": 0.9064, + "step": 9031 + }, + { + "epoch": 0.8687956906502501, + "grad_norm": 2.251931563312136, + "learning_rate": 8.893805927371724e-07, + "loss": 0.9059, + "step": 9032 + }, + { + "epoch": 0.8688918814928819, + "grad_norm": 1.7781712868694624, + "learning_rate": 8.880966409682113e-07, + "loss": 0.7687, + "step": 9033 + }, + { + "epoch": 0.8689880723355137, + "grad_norm": 2.0109888959110607, + "learning_rate": 8.868135735879291e-07, + "loss": 0.8709, + "step": 9034 + }, + { + "epoch": 0.8690842631781455, + "grad_norm": 2.0383144806569047, + "learning_rate": 8.855313907208518e-07, + "loss": 0.9273, + "step": 9035 + }, + { + "epoch": 0.8691804540207773, + "grad_norm": 1.9567648552639625, + "learning_rate": 8.842500924914299e-07, + "loss": 0.8562, + "step": 9036 + }, + { + "epoch": 0.869276644863409, + "grad_norm": 1.8743142809630322, + "learning_rate": 8.829696790240272e-07, + "loss": 0.8365, + "step": 9037 + }, + { + "epoch": 0.8693728357060407, + "grad_norm": 2.037312345503149, + "learning_rate": 8.816901504429143e-07, + "loss": 0.9526, + "step": 9038 + }, + { + "epoch": 0.8694690265486725, + "grad_norm": 1.8805970516276758, + "learning_rate": 8.804115068722808e-07, + "loss": 0.8206, + "step": 9039 + }, + { + "epoch": 0.8695652173913043, + "grad_norm": 1.959680263705289, + "learning_rate": 8.791337484362305e-07, + "loss": 0.8768, + "step": 9040 + }, + { + "epoch": 0.8696614082339361, + "grad_norm": 2.247643205314895, + "learning_rate": 8.778568752587846e-07, + "loss": 0.8605, + "step": 9041 + }, + { + "epoch": 0.8697575990765679, + "grad_norm": 1.9749573618731906, + "learning_rate": 8.765808874638682e-07, + "loss": 0.8858, + "step": 9042 + }, + { + "epoch": 0.8698537899191997, + "grad_norm": 1.8070498300099074, + "learning_rate": 8.753057851753321e-07, + "loss": 0.8094, + "step": 9043 + }, + { + "epoch": 0.8699499807618315, + "grad_norm": 2.0858422243793986, + "learning_rate": 8.740315685169364e-07, + "loss": 0.8493, + "step": 9044 + }, + { + "epoch": 0.8700461716044633, + "grad_norm": 1.8829269486498559, + "learning_rate": 8.727582376123489e-07, + "loss": 0.9682, + "step": 9045 + }, + { + "epoch": 0.8701423624470951, + "grad_norm": 1.932915193663648, + "learning_rate": 8.714857925851617e-07, + "loss": 0.8016, + "step": 9046 + }, + { + "epoch": 0.8702385532897268, + "grad_norm": 1.7424872889617078, + "learning_rate": 8.70214233558877e-07, + "loss": 0.7834, + "step": 9047 + }, + { + "epoch": 0.8703347441323586, + "grad_norm": 2.2944749419965684, + "learning_rate": 8.689435606569086e-07, + "loss": 0.9324, + "step": 9048 + }, + { + "epoch": 0.8704309349749904, + "grad_norm": 1.8870682589207202, + "learning_rate": 8.676737740025876e-07, + "loss": 0.9228, + "step": 9049 + }, + { + "epoch": 0.8705271258176221, + "grad_norm": 2.278923668552798, + "learning_rate": 8.664048737191566e-07, + "loss": 0.929, + "step": 9050 + }, + { + "epoch": 0.8706233166602539, + "grad_norm": 2.134053689446095, + "learning_rate": 8.651368599297738e-07, + "loss": 0.8253, + "step": 9051 + }, + { + "epoch": 0.8707195075028857, + "grad_norm": 2.0205875853541007, + "learning_rate": 8.638697327575108e-07, + "loss": 0.8581, + "step": 9052 + }, + { + "epoch": 0.8708156983455175, + "grad_norm": 1.867404677731671, + "learning_rate": 8.626034923253523e-07, + "loss": 0.9336, + "step": 9053 + }, + { + "epoch": 0.8709118891881493, + "grad_norm": 2.359702051064883, + "learning_rate": 8.613381387562015e-07, + "loss": 0.9002, + "step": 9054 + }, + { + "epoch": 0.8710080800307811, + "grad_norm": 2.2504318582201206, + "learning_rate": 8.600736721728687e-07, + "loss": 1.0256, + "step": 9055 + }, + { + "epoch": 0.8711042708734128, + "grad_norm": 2.1288415195603574, + "learning_rate": 8.588100926980802e-07, + "loss": 0.8861, + "step": 9056 + }, + { + "epoch": 0.8712004617160446, + "grad_norm": 2.1250627513737705, + "learning_rate": 8.575474004544781e-07, + "loss": 0.897, + "step": 9057 + }, + { + "epoch": 0.8712966525586764, + "grad_norm": 2.253968754739755, + "learning_rate": 8.56285595564621e-07, + "loss": 0.8764, + "step": 9058 + }, + { + "epoch": 0.8713928434013082, + "grad_norm": 1.8132018851303966, + "learning_rate": 8.550246781509731e-07, + "loss": 0.835, + "step": 9059 + }, + { + "epoch": 0.87148903424394, + "grad_norm": 1.5473647602688383, + "learning_rate": 8.537646483359185e-07, + "loss": 0.7627, + "step": 9060 + }, + { + "epoch": 0.8715852250865718, + "grad_norm": 1.8586734619486047, + "learning_rate": 8.525055062417576e-07, + "loss": 0.9885, + "step": 9061 + }, + { + "epoch": 0.8716814159292036, + "grad_norm": 1.7439705566084371, + "learning_rate": 8.512472519906978e-07, + "loss": 0.8193, + "step": 9062 + }, + { + "epoch": 0.8717776067718354, + "grad_norm": 2.067919133949697, + "learning_rate": 8.499898857048628e-07, + "loss": 0.9089, + "step": 9063 + }, + { + "epoch": 0.8718737976144671, + "grad_norm": 1.941835289295631, + "learning_rate": 8.487334075062914e-07, + "loss": 0.9916, + "step": 9064 + }, + { + "epoch": 0.8719699884570988, + "grad_norm": 2.5014394096188255, + "learning_rate": 8.474778175169384e-07, + "loss": 0.8059, + "step": 9065 + }, + { + "epoch": 0.8720661792997306, + "grad_norm": 2.2275574821065725, + "learning_rate": 8.462231158586654e-07, + "loss": 1.0286, + "step": 9066 + }, + { + "epoch": 0.8721623701423624, + "grad_norm": 1.975032503330523, + "learning_rate": 8.449693026532557e-07, + "loss": 0.9434, + "step": 9067 + }, + { + "epoch": 0.8722585609849942, + "grad_norm": 2.1785016546274045, + "learning_rate": 8.437163780224011e-07, + "loss": 0.9363, + "step": 9068 + }, + { + "epoch": 0.872354751827626, + "grad_norm": 2.112102867718652, + "learning_rate": 8.424643420877066e-07, + "loss": 0.8244, + "step": 9069 + }, + { + "epoch": 0.8724509426702578, + "grad_norm": 1.928823219091913, + "learning_rate": 8.412131949706958e-07, + "loss": 0.7948, + "step": 9070 + }, + { + "epoch": 0.8725471335128896, + "grad_norm": 2.0455129917806674, + "learning_rate": 8.399629367928019e-07, + "loss": 0.8829, + "step": 9071 + }, + { + "epoch": 0.8726433243555214, + "grad_norm": 2.1201185212925204, + "learning_rate": 8.387135676753755e-07, + "loss": 0.9668, + "step": 9072 + }, + { + "epoch": 0.8727395151981532, + "grad_norm": 2.069965038098699, + "learning_rate": 8.374650877396773e-07, + "loss": 0.7955, + "step": 9073 + }, + { + "epoch": 0.8728357060407849, + "grad_norm": 1.8086366369522162, + "learning_rate": 8.362174971068804e-07, + "loss": 0.9219, + "step": 9074 + }, + { + "epoch": 0.8729318968834167, + "grad_norm": 2.01611025089467, + "learning_rate": 8.349707958980791e-07, + "loss": 0.964, + "step": 9075 + }, + { + "epoch": 0.8730280877260485, + "grad_norm": 2.0263895700905294, + "learning_rate": 8.337249842342721e-07, + "loss": 0.9143, + "step": 9076 + }, + { + "epoch": 0.8731242785686802, + "grad_norm": 1.817370778622754, + "learning_rate": 8.324800622363783e-07, + "loss": 0.7673, + "step": 9077 + }, + { + "epoch": 0.873220469411312, + "grad_norm": 2.120569811262115, + "learning_rate": 8.312360300252287e-07, + "loss": 0.9666, + "step": 9078 + }, + { + "epoch": 0.8733166602539438, + "grad_norm": 2.008749975650139, + "learning_rate": 8.299928877215669e-07, + "loss": 1.005, + "step": 9079 + }, + { + "epoch": 0.8734128510965756, + "grad_norm": 1.600699242975325, + "learning_rate": 8.287506354460484e-07, + "loss": 0.7413, + "step": 9080 + }, + { + "epoch": 0.8735090419392074, + "grad_norm": 1.7995468914083919, + "learning_rate": 8.275092733192458e-07, + "loss": 0.9488, + "step": 9081 + }, + { + "epoch": 0.8736052327818392, + "grad_norm": 1.8991859116006324, + "learning_rate": 8.26268801461646e-07, + "loss": 0.9242, + "step": 9082 + }, + { + "epoch": 0.8737014236244709, + "grad_norm": 1.9467191865582338, + "learning_rate": 8.25029219993645e-07, + "loss": 0.7458, + "step": 9083 + }, + { + "epoch": 0.8737976144671027, + "grad_norm": 2.1537289424166604, + "learning_rate": 8.237905290355563e-07, + "loss": 0.9377, + "step": 9084 + }, + { + "epoch": 0.8738938053097345, + "grad_norm": 1.8407181514064643, + "learning_rate": 8.22552728707604e-07, + "loss": 0.8845, + "step": 9085 + }, + { + "epoch": 0.8739899961523663, + "grad_norm": 2.0196170095320425, + "learning_rate": 8.213158191299297e-07, + "loss": 0.973, + "step": 9086 + }, + { + "epoch": 0.8740861869949981, + "grad_norm": 2.1468330778542213, + "learning_rate": 8.200798004225829e-07, + "loss": 0.8226, + "step": 9087 + }, + { + "epoch": 0.8741823778376299, + "grad_norm": 2.005230146672427, + "learning_rate": 8.188446727055311e-07, + "loss": 0.9236, + "step": 9088 + }, + { + "epoch": 0.8742785686802617, + "grad_norm": 1.9197702258495304, + "learning_rate": 8.17610436098657e-07, + "loss": 0.8638, + "step": 9089 + }, + { + "epoch": 0.8743747595228935, + "grad_norm": 1.9507323935529048, + "learning_rate": 8.163770907217506e-07, + "loss": 0.8498, + "step": 9090 + }, + { + "epoch": 0.8744709503655252, + "grad_norm": 1.974833048326194, + "learning_rate": 8.151446366945182e-07, + "loss": 0.8637, + "step": 9091 + }, + { + "epoch": 0.8745671412081569, + "grad_norm": 1.87048108919086, + "learning_rate": 8.139130741365819e-07, + "loss": 0.9433, + "step": 9092 + }, + { + "epoch": 0.8746633320507887, + "grad_norm": 1.9837030151207575, + "learning_rate": 8.126824031674763e-07, + "loss": 0.9198, + "step": 9093 + }, + { + "epoch": 0.8747595228934205, + "grad_norm": 2.2389729622471943, + "learning_rate": 8.114526239066456e-07, + "loss": 0.8616, + "step": 9094 + }, + { + "epoch": 0.8748557137360523, + "grad_norm": 2.0519448068115334, + "learning_rate": 8.102237364734511e-07, + "loss": 0.8532, + "step": 9095 + }, + { + "epoch": 0.8749519045786841, + "grad_norm": 2.039956802685304, + "learning_rate": 8.08995740987173e-07, + "loss": 0.9696, + "step": 9096 + }, + { + "epoch": 0.8750480954213159, + "grad_norm": 2.08055015574849, + "learning_rate": 8.077686375669891e-07, + "loss": 0.947, + "step": 9097 + }, + { + "epoch": 0.8751442862639477, + "grad_norm": 2.2057198772671742, + "learning_rate": 8.065424263320054e-07, + "loss": 0.9384, + "step": 9098 + }, + { + "epoch": 0.8752404771065795, + "grad_norm": 1.9885528279151026, + "learning_rate": 8.053171074012378e-07, + "loss": 0.9884, + "step": 9099 + }, + { + "epoch": 0.8753366679492113, + "grad_norm": 2.142832758959495, + "learning_rate": 8.040926808936112e-07, + "loss": 0.8756, + "step": 9100 + }, + { + "epoch": 0.8754328587918431, + "grad_norm": 1.9640797523491542, + "learning_rate": 8.028691469279682e-07, + "loss": 0.8294, + "step": 9101 + }, + { + "epoch": 0.8755290496344748, + "grad_norm": 1.925065391219502, + "learning_rate": 8.016465056230616e-07, + "loss": 0.896, + "step": 9102 + }, + { + "epoch": 0.8756252404771065, + "grad_norm": 2.020670659970949, + "learning_rate": 8.004247570975621e-07, + "loss": 0.834, + "step": 9103 + }, + { + "epoch": 0.8757214313197383, + "grad_norm": 2.337942786758394, + "learning_rate": 7.99203901470047e-07, + "loss": 0.9584, + "step": 9104 + }, + { + "epoch": 0.8758176221623701, + "grad_norm": 1.7621635971532275, + "learning_rate": 7.979839388590138e-07, + "loss": 0.8309, + "step": 9105 + }, + { + "epoch": 0.8759138130050019, + "grad_norm": 1.784161982353027, + "learning_rate": 7.967648693828712e-07, + "loss": 0.8844, + "step": 9106 + }, + { + "epoch": 0.8760100038476337, + "grad_norm": 1.9898387514511964, + "learning_rate": 7.955466931599387e-07, + "loss": 0.8973, + "step": 9107 + }, + { + "epoch": 0.8761061946902655, + "grad_norm": 2.0700392890964228, + "learning_rate": 7.943294103084487e-07, + "loss": 1.012, + "step": 9108 + }, + { + "epoch": 0.8762023855328973, + "grad_norm": 1.6630509916363243, + "learning_rate": 7.931130209465521e-07, + "loss": 0.778, + "step": 9109 + }, + { + "epoch": 0.8762985763755291, + "grad_norm": 2.022318790116037, + "learning_rate": 7.9189752519231e-07, + "loss": 0.9209, + "step": 9110 + }, + { + "epoch": 0.8763947672181608, + "grad_norm": 1.879878420042895, + "learning_rate": 7.906829231636947e-07, + "loss": 0.8921, + "step": 9111 + }, + { + "epoch": 0.8764909580607926, + "grad_norm": 1.9635658000008507, + "learning_rate": 7.894692149785954e-07, + "loss": 0.9394, + "step": 9112 + }, + { + "epoch": 0.8765871489034244, + "grad_norm": 2.074506355410214, + "learning_rate": 7.882564007548133e-07, + "loss": 0.9279, + "step": 9113 + }, + { + "epoch": 0.8766833397460562, + "grad_norm": 1.9992735225013099, + "learning_rate": 7.870444806100619e-07, + "loss": 0.9349, + "step": 9114 + }, + { + "epoch": 0.876779530588688, + "grad_norm": 2.277258180377917, + "learning_rate": 7.858334546619672e-07, + "loss": 0.8408, + "step": 9115 + }, + { + "epoch": 0.8768757214313198, + "grad_norm": 1.9869432977352381, + "learning_rate": 7.846233230280698e-07, + "loss": 0.9058, + "step": 9116 + }, + { + "epoch": 0.8769719122739515, + "grad_norm": 1.9813673725867338, + "learning_rate": 7.834140858258266e-07, + "loss": 0.9181, + "step": 9117 + }, + { + "epoch": 0.8770681031165833, + "grad_norm": 2.0742443853295027, + "learning_rate": 7.822057431725994e-07, + "loss": 0.8821, + "step": 9118 + }, + { + "epoch": 0.8771642939592151, + "grad_norm": 2.2437001045364067, + "learning_rate": 7.809982951856732e-07, + "loss": 0.9147, + "step": 9119 + }, + { + "epoch": 0.8772604848018468, + "grad_norm": 1.9703336555990276, + "learning_rate": 7.797917419822377e-07, + "loss": 0.9591, + "step": 9120 + }, + { + "epoch": 0.8773566756444786, + "grad_norm": 2.0042189753296316, + "learning_rate": 7.785860836794023e-07, + "loss": 0.9016, + "step": 9121 + }, + { + "epoch": 0.8774528664871104, + "grad_norm": 1.9420827546227353, + "learning_rate": 7.773813203941827e-07, + "loss": 0.8478, + "step": 9122 + }, + { + "epoch": 0.8775490573297422, + "grad_norm": 1.696886182435857, + "learning_rate": 7.76177452243514e-07, + "loss": 0.7612, + "step": 9123 + }, + { + "epoch": 0.877645248172374, + "grad_norm": 2.1510835394906476, + "learning_rate": 7.749744793442448e-07, + "loss": 0.9192, + "step": 9124 + }, + { + "epoch": 0.8777414390150058, + "grad_norm": 2.06855245404924, + "learning_rate": 7.737724018131287e-07, + "loss": 0.8913, + "step": 9125 + }, + { + "epoch": 0.8778376298576376, + "grad_norm": 1.7783259773705953, + "learning_rate": 7.725712197668378e-07, + "loss": 0.9014, + "step": 9126 + }, + { + "epoch": 0.8779338207002694, + "grad_norm": 1.9607863812853217, + "learning_rate": 7.713709333219621e-07, + "loss": 0.9178, + "step": 9127 + }, + { + "epoch": 0.8780300115429012, + "grad_norm": 1.9092468679964074, + "learning_rate": 7.701715425949952e-07, + "loss": 0.7193, + "step": 9128 + }, + { + "epoch": 0.8781262023855329, + "grad_norm": 2.077465744716933, + "learning_rate": 7.689730477023494e-07, + "loss": 0.985, + "step": 9129 + }, + { + "epoch": 0.8782223932281646, + "grad_norm": 2.1885313144937624, + "learning_rate": 7.677754487603517e-07, + "loss": 0.9202, + "step": 9130 + }, + { + "epoch": 0.8783185840707964, + "grad_norm": 1.7220234527282372, + "learning_rate": 7.66578745885237e-07, + "loss": 0.8845, + "step": 9131 + }, + { + "epoch": 0.8784147749134282, + "grad_norm": 1.838784479884038, + "learning_rate": 7.653829391931533e-07, + "loss": 0.9432, + "step": 9132 + }, + { + "epoch": 0.87851096575606, + "grad_norm": 2.0221247537566835, + "learning_rate": 7.641880288001668e-07, + "loss": 0.9109, + "step": 9133 + }, + { + "epoch": 0.8786071565986918, + "grad_norm": 2.0028603794098885, + "learning_rate": 7.629940148222559e-07, + "loss": 0.8463, + "step": 9134 + }, + { + "epoch": 0.8787033474413236, + "grad_norm": 2.0566636115526418, + "learning_rate": 7.618008973753043e-07, + "loss": 0.879, + "step": 9135 + }, + { + "epoch": 0.8787995382839554, + "grad_norm": 2.0886494082622913, + "learning_rate": 7.606086765751209e-07, + "loss": 0.9923, + "step": 9136 + }, + { + "epoch": 0.8788957291265872, + "grad_norm": 2.139642952657427, + "learning_rate": 7.594173525374149e-07, + "loss": 0.9025, + "step": 9137 + }, + { + "epoch": 0.8789919199692189, + "grad_norm": 1.7882407927955197, + "learning_rate": 7.582269253778185e-07, + "loss": 0.8402, + "step": 9138 + }, + { + "epoch": 0.8790881108118507, + "grad_norm": 2.0759143346946765, + "learning_rate": 7.570373952118715e-07, + "loss": 1.019, + "step": 9139 + }, + { + "epoch": 0.8791843016544825, + "grad_norm": 2.0935424671555167, + "learning_rate": 7.55848762155027e-07, + "loss": 0.8921, + "step": 9140 + }, + { + "epoch": 0.8792804924971143, + "grad_norm": 1.8628321641259158, + "learning_rate": 7.54661026322655e-07, + "loss": 0.9472, + "step": 9141 + }, + { + "epoch": 0.8793766833397461, + "grad_norm": 1.6421153548782188, + "learning_rate": 7.534741878300333e-07, + "loss": 0.782, + "step": 9142 + }, + { + "epoch": 0.8794728741823779, + "grad_norm": 2.0723340881024357, + "learning_rate": 7.522882467923542e-07, + "loss": 0.9708, + "step": 9143 + }, + { + "epoch": 0.8795690650250096, + "grad_norm": 1.8477463831647487, + "learning_rate": 7.511032033247256e-07, + "loss": 0.8825, + "step": 9144 + }, + { + "epoch": 0.8796652558676414, + "grad_norm": 1.8908420462472588, + "learning_rate": 7.499190575421656e-07, + "loss": 0.971, + "step": 9145 + }, + { + "epoch": 0.8797614467102732, + "grad_norm": 2.034452320522671, + "learning_rate": 7.487358095596031e-07, + "loss": 0.9285, + "step": 9146 + }, + { + "epoch": 0.8798576375529049, + "grad_norm": 2.0335010395069673, + "learning_rate": 7.475534594918854e-07, + "loss": 0.9471, + "step": 9147 + }, + { + "epoch": 0.8799538283955367, + "grad_norm": 2.0988724292844148, + "learning_rate": 7.463720074537728e-07, + "loss": 0.9316, + "step": 9148 + }, + { + "epoch": 0.8800500192381685, + "grad_norm": 1.9283549246600173, + "learning_rate": 7.45191453559927e-07, + "loss": 0.8674, + "step": 9149 + }, + { + "epoch": 0.8801462100808003, + "grad_norm": 1.97453639130713, + "learning_rate": 7.440117979249362e-07, + "loss": 0.809, + "step": 9150 + }, + { + "epoch": 0.8802424009234321, + "grad_norm": 1.9749171783695953, + "learning_rate": 7.428330406632977e-07, + "loss": 0.8625, + "step": 9151 + }, + { + "epoch": 0.8803385917660639, + "grad_norm": 1.8130696849455392, + "learning_rate": 7.416551818894158e-07, + "loss": 0.8477, + "step": 9152 + }, + { + "epoch": 0.8804347826086957, + "grad_norm": 1.8682459061586114, + "learning_rate": 7.404782217176143e-07, + "loss": 0.9458, + "step": 9153 + }, + { + "epoch": 0.8805309734513275, + "grad_norm": 2.146793619813614, + "learning_rate": 7.393021602621264e-07, + "loss": 0.8608, + "step": 9154 + }, + { + "epoch": 0.8806271642939593, + "grad_norm": 1.9179714421049612, + "learning_rate": 7.381269976371009e-07, + "loss": 0.8611, + "step": 9155 + }, + { + "epoch": 0.880723355136591, + "grad_norm": 2.0652078946514494, + "learning_rate": 7.369527339565951e-07, + "loss": 0.8593, + "step": 9156 + }, + { + "epoch": 0.8808195459792227, + "grad_norm": 1.7985514229467272, + "learning_rate": 7.357793693345816e-07, + "loss": 0.6953, + "step": 9157 + }, + { + "epoch": 0.8809157368218545, + "grad_norm": 1.9864762196656849, + "learning_rate": 7.346069038849469e-07, + "loss": 0.895, + "step": 9158 + }, + { + "epoch": 0.8810119276644863, + "grad_norm": 1.9177572383556103, + "learning_rate": 7.334353377214898e-07, + "loss": 0.8684, + "step": 9159 + }, + { + "epoch": 0.8811081185071181, + "grad_norm": 2.03219717056584, + "learning_rate": 7.322646709579173e-07, + "loss": 0.8521, + "step": 9160 + }, + { + "epoch": 0.8812043093497499, + "grad_norm": 2.053544274799706, + "learning_rate": 7.310949037078541e-07, + "loss": 0.8577, + "step": 9161 + }, + { + "epoch": 0.8813005001923817, + "grad_norm": 2.04053195968409, + "learning_rate": 7.299260360848382e-07, + "loss": 0.7817, + "step": 9162 + }, + { + "epoch": 0.8813966910350135, + "grad_norm": 1.8848354267749865, + "learning_rate": 7.287580682023155e-07, + "loss": 0.9056, + "step": 9163 + }, + { + "epoch": 0.8814928818776453, + "grad_norm": 1.9516421726361552, + "learning_rate": 7.275910001736497e-07, + "loss": 0.7863, + "step": 9164 + }, + { + "epoch": 0.881589072720277, + "grad_norm": 1.9849670586563113, + "learning_rate": 7.264248321121137e-07, + "loss": 0.8093, + "step": 9165 + }, + { + "epoch": 0.8816852635629088, + "grad_norm": 2.470731685107478, + "learning_rate": 7.252595641308957e-07, + "loss": 0.9334, + "step": 9166 + }, + { + "epoch": 0.8817814544055406, + "grad_norm": 1.800940963639631, + "learning_rate": 7.240951963430909e-07, + "loss": 0.8109, + "step": 9167 + }, + { + "epoch": 0.8818776452481724, + "grad_norm": 2.1839866407012427, + "learning_rate": 7.229317288617144e-07, + "loss": 0.9701, + "step": 9168 + }, + { + "epoch": 0.8819738360908042, + "grad_norm": 2.1462004890996687, + "learning_rate": 7.217691617996914e-07, + "loss": 0.963, + "step": 9169 + }, + { + "epoch": 0.882070026933436, + "grad_norm": 1.461842030636379, + "learning_rate": 7.20607495269856e-07, + "loss": 0.7756, + "step": 9170 + }, + { + "epoch": 0.8821662177760677, + "grad_norm": 1.9445341394913356, + "learning_rate": 7.194467293849616e-07, + "loss": 0.8775, + "step": 9171 + }, + { + "epoch": 0.8822624086186995, + "grad_norm": 2.0089475153021743, + "learning_rate": 7.182868642576679e-07, + "loss": 0.7482, + "step": 9172 + }, + { + "epoch": 0.8823585994613313, + "grad_norm": 2.299043861679115, + "learning_rate": 7.171279000005516e-07, + "loss": 0.8941, + "step": 9173 + }, + { + "epoch": 0.882454790303963, + "grad_norm": 1.8252139760438708, + "learning_rate": 7.15969836726097e-07, + "loss": 0.7817, + "step": 9174 + }, + { + "epoch": 0.8825509811465948, + "grad_norm": 2.0305581278209903, + "learning_rate": 7.148126745467066e-07, + "loss": 0.8918, + "step": 9175 + }, + { + "epoch": 0.8826471719892266, + "grad_norm": 2.0494657250254082, + "learning_rate": 7.13656413574696e-07, + "loss": 0.9487, + "step": 9176 + }, + { + "epoch": 0.8827433628318584, + "grad_norm": 1.895750674757115, + "learning_rate": 7.125010539222832e-07, + "loss": 0.8546, + "step": 9177 + }, + { + "epoch": 0.8828395536744902, + "grad_norm": 1.9354357302424106, + "learning_rate": 7.113465957016097e-07, + "loss": 0.8113, + "step": 9178 + }, + { + "epoch": 0.882935744517122, + "grad_norm": 1.908156241935171, + "learning_rate": 7.10193039024727e-07, + "loss": 0.9336, + "step": 9179 + }, + { + "epoch": 0.8830319353597538, + "grad_norm": 1.74712863124145, + "learning_rate": 7.090403840035942e-07, + "loss": 0.8833, + "step": 9180 + }, + { + "epoch": 0.8831281262023856, + "grad_norm": 2.0762242650818377, + "learning_rate": 7.078886307500887e-07, + "loss": 0.9156, + "step": 9181 + }, + { + "epoch": 0.8832243170450174, + "grad_norm": 2.2721596331945135, + "learning_rate": 7.067377793759999e-07, + "loss": 0.9136, + "step": 9182 + }, + { + "epoch": 0.883320507887649, + "grad_norm": 1.9896962642105276, + "learning_rate": 7.055878299930252e-07, + "loss": 0.8764, + "step": 9183 + }, + { + "epoch": 0.8834166987302808, + "grad_norm": 1.9839000723492755, + "learning_rate": 7.044387827127752e-07, + "loss": 0.8693, + "step": 9184 + }, + { + "epoch": 0.8835128895729126, + "grad_norm": 2.1530838872419387, + "learning_rate": 7.032906376467774e-07, + "loss": 0.8501, + "step": 9185 + }, + { + "epoch": 0.8836090804155444, + "grad_norm": 2.0546123578001905, + "learning_rate": 7.021433949064704e-07, + "loss": 0.8847, + "step": 9186 + }, + { + "epoch": 0.8837052712581762, + "grad_norm": 1.990804711038259, + "learning_rate": 7.009970546032008e-07, + "loss": 0.8523, + "step": 9187 + }, + { + "epoch": 0.883801462100808, + "grad_norm": 2.3630482963951707, + "learning_rate": 6.99851616848235e-07, + "loss": 0.9058, + "step": 9188 + }, + { + "epoch": 0.8838976529434398, + "grad_norm": 2.2508438890792615, + "learning_rate": 6.987070817527431e-07, + "loss": 0.9235, + "step": 9189 + }, + { + "epoch": 0.8839938437860716, + "grad_norm": 1.9125904145201562, + "learning_rate": 6.975634494278149e-07, + "loss": 0.9257, + "step": 9190 + }, + { + "epoch": 0.8840900346287034, + "grad_norm": 2.017785287149117, + "learning_rate": 6.964207199844475e-07, + "loss": 0.9145, + "step": 9191 + }, + { + "epoch": 0.8841862254713351, + "grad_norm": 2.317626418456962, + "learning_rate": 6.952788935335541e-07, + "loss": 0.937, + "step": 9192 + }, + { + "epoch": 0.8842824163139669, + "grad_norm": 2.2163120393573283, + "learning_rate": 6.941379701859607e-07, + "loss": 0.9136, + "step": 9193 + }, + { + "epoch": 0.8843786071565987, + "grad_norm": 2.036618743606627, + "learning_rate": 6.92997950052402e-07, + "loss": 0.9874, + "step": 9194 + }, + { + "epoch": 0.8844747979992305, + "grad_norm": 2.2316753000999174, + "learning_rate": 6.918588332435238e-07, + "loss": 0.9439, + "step": 9195 + }, + { + "epoch": 0.8845709888418622, + "grad_norm": 1.94732488091564, + "learning_rate": 6.907206198698912e-07, + "loss": 0.8509, + "step": 9196 + }, + { + "epoch": 0.884667179684494, + "grad_norm": 2.0965842416217204, + "learning_rate": 6.895833100419769e-07, + "loss": 0.9857, + "step": 9197 + }, + { + "epoch": 0.8847633705271258, + "grad_norm": 1.9891695681505288, + "learning_rate": 6.884469038701646e-07, + "loss": 0.9628, + "step": 9198 + }, + { + "epoch": 0.8848595613697576, + "grad_norm": 2.0894579176412664, + "learning_rate": 6.873114014647564e-07, + "loss": 0.95, + "step": 9199 + }, + { + "epoch": 0.8849557522123894, + "grad_norm": 2.336290484178676, + "learning_rate": 6.861768029359595e-07, + "loss": 0.9522, + "step": 9200 + }, + { + "epoch": 0.8850519430550212, + "grad_norm": 2.2369918238334097, + "learning_rate": 6.850431083938947e-07, + "loss": 0.9599, + "step": 9201 + }, + { + "epoch": 0.8851481338976529, + "grad_norm": 2.195306369825476, + "learning_rate": 6.839103179485995e-07, + "loss": 0.9287, + "step": 9202 + }, + { + "epoch": 0.8852443247402847, + "grad_norm": 1.9837357031908016, + "learning_rate": 6.827784317100195e-07, + "loss": 0.8977, + "step": 9203 + }, + { + "epoch": 0.8853405155829165, + "grad_norm": 1.8741835049601645, + "learning_rate": 6.816474497880177e-07, + "loss": 0.8965, + "step": 9204 + }, + { + "epoch": 0.8854367064255483, + "grad_norm": 2.151968643115082, + "learning_rate": 6.805173722923631e-07, + "loss": 0.994, + "step": 9205 + }, + { + "epoch": 0.8855328972681801, + "grad_norm": 1.9686845282525967, + "learning_rate": 6.793881993327366e-07, + "loss": 0.9484, + "step": 9206 + }, + { + "epoch": 0.8856290881108119, + "grad_norm": 2.016716743975715, + "learning_rate": 6.782599310187388e-07, + "loss": 0.8443, + "step": 9207 + }, + { + "epoch": 0.8857252789534437, + "grad_norm": 1.9054019733331489, + "learning_rate": 6.77132567459875e-07, + "loss": 0.8645, + "step": 9208 + }, + { + "epoch": 0.8858214697960755, + "grad_norm": 2.412513775666417, + "learning_rate": 6.760061087655656e-07, + "loss": 0.8678, + "step": 9209 + }, + { + "epoch": 0.8859176606387072, + "grad_norm": 2.206861029157715, + "learning_rate": 6.748805550451453e-07, + "loss": 0.9361, + "step": 9210 + }, + { + "epoch": 0.8860138514813389, + "grad_norm": 1.8511758927639481, + "learning_rate": 6.737559064078581e-07, + "loss": 0.892, + "step": 9211 + }, + { + "epoch": 0.8861100423239707, + "grad_norm": 1.9674312251420456, + "learning_rate": 6.726321629628585e-07, + "loss": 0.9453, + "step": 9212 + }, + { + "epoch": 0.8862062331666025, + "grad_norm": 2.078255572401999, + "learning_rate": 6.715093248192172e-07, + "loss": 0.8773, + "step": 9213 + }, + { + "epoch": 0.8863024240092343, + "grad_norm": 2.066393128299769, + "learning_rate": 6.703873920859161e-07, + "loss": 0.9016, + "step": 9214 + }, + { + "epoch": 0.8863986148518661, + "grad_norm": 2.0500634491481713, + "learning_rate": 6.692663648718467e-07, + "loss": 0.8713, + "step": 9215 + }, + { + "epoch": 0.8864948056944979, + "grad_norm": 2.1077598850469843, + "learning_rate": 6.681462432858154e-07, + "loss": 0.9673, + "step": 9216 + }, + { + "epoch": 0.8865909965371297, + "grad_norm": 1.9591167824001452, + "learning_rate": 6.670270274365398e-07, + "loss": 0.8217, + "step": 9217 + }, + { + "epoch": 0.8866871873797615, + "grad_norm": 2.047432122885484, + "learning_rate": 6.659087174326506e-07, + "loss": 0.8676, + "step": 9218 + }, + { + "epoch": 0.8867833782223933, + "grad_norm": 2.1404747517143763, + "learning_rate": 6.647913133826856e-07, + "loss": 0.9703, + "step": 9219 + }, + { + "epoch": 0.886879569065025, + "grad_norm": 2.0354660235683664, + "learning_rate": 6.636748153951e-07, + "loss": 0.8986, + "step": 9220 + }, + { + "epoch": 0.8869757599076568, + "grad_norm": 1.8625489562806583, + "learning_rate": 6.625592235782629e-07, + "loss": 0.8906, + "step": 9221 + }, + { + "epoch": 0.8870719507502886, + "grad_norm": 2.06631533848724, + "learning_rate": 6.614445380404478e-07, + "loss": 0.8718, + "step": 9222 + }, + { + "epoch": 0.8871681415929203, + "grad_norm": 2.1893384921690933, + "learning_rate": 6.603307588898477e-07, + "loss": 0.9405, + "step": 9223 + }, + { + "epoch": 0.8872643324355521, + "grad_norm": 1.8351473332412327, + "learning_rate": 6.592178862345622e-07, + "loss": 0.9228, + "step": 9224 + }, + { + "epoch": 0.8873605232781839, + "grad_norm": 1.926220718205039, + "learning_rate": 6.581059201826067e-07, + "loss": 0.9501, + "step": 9225 + }, + { + "epoch": 0.8874567141208157, + "grad_norm": 1.7651407651826163, + "learning_rate": 6.569948608419041e-07, + "loss": 0.8355, + "step": 9226 + }, + { + "epoch": 0.8875529049634475, + "grad_norm": 1.9705768192383841, + "learning_rate": 6.558847083202957e-07, + "loss": 0.9216, + "step": 9227 + }, + { + "epoch": 0.8876490958060793, + "grad_norm": 2.3072885256416766, + "learning_rate": 6.547754627255332e-07, + "loss": 0.9763, + "step": 9228 + }, + { + "epoch": 0.887745286648711, + "grad_norm": 1.9620631901064582, + "learning_rate": 6.536671241652726e-07, + "loss": 0.9388, + "step": 9229 + }, + { + "epoch": 0.8878414774913428, + "grad_norm": 1.8792188446054692, + "learning_rate": 6.52559692747089e-07, + "loss": 0.8785, + "step": 9230 + }, + { + "epoch": 0.8879376683339746, + "grad_norm": 2.1965427342767807, + "learning_rate": 6.514531685784731e-07, + "loss": 0.8808, + "step": 9231 + }, + { + "epoch": 0.8880338591766064, + "grad_norm": 1.9869239302158226, + "learning_rate": 6.503475517668168e-07, + "loss": 0.9517, + "step": 9232 + }, + { + "epoch": 0.8881300500192382, + "grad_norm": 2.014528816895788, + "learning_rate": 6.492428424194319e-07, + "loss": 0.9304, + "step": 9233 + }, + { + "epoch": 0.88822624086187, + "grad_norm": 1.8820701912376232, + "learning_rate": 6.481390406435417e-07, + "loss": 0.8913, + "step": 9234 + }, + { + "epoch": 0.8883224317045018, + "grad_norm": 2.178756367710181, + "learning_rate": 6.470361465462782e-07, + "loss": 0.7828, + "step": 9235 + }, + { + "epoch": 0.8884186225471336, + "grad_norm": 2.252159731157635, + "learning_rate": 6.459341602346858e-07, + "loss": 0.996, + "step": 9236 + }, + { + "epoch": 0.8885148133897653, + "grad_norm": 2.2511504378800367, + "learning_rate": 6.448330818157223e-07, + "loss": 0.9139, + "step": 9237 + }, + { + "epoch": 0.888611004232397, + "grad_norm": 2.5175522064637814, + "learning_rate": 6.437329113962576e-07, + "loss": 0.9202, + "step": 9238 + }, + { + "epoch": 0.8887071950750288, + "grad_norm": 2.2059353819493492, + "learning_rate": 6.42633649083072e-07, + "loss": 0.7692, + "step": 9239 + }, + { + "epoch": 0.8888033859176606, + "grad_norm": 2.0483507549721245, + "learning_rate": 6.415352949828601e-07, + "loss": 0.9042, + "step": 9240 + }, + { + "epoch": 0.8888995767602924, + "grad_norm": 1.7756522120346907, + "learning_rate": 6.404378492022234e-07, + "loss": 0.7899, + "step": 9241 + }, + { + "epoch": 0.8889957676029242, + "grad_norm": 2.0313636187755444, + "learning_rate": 6.393413118476821e-07, + "loss": 0.9271, + "step": 9242 + }, + { + "epoch": 0.889091958445556, + "grad_norm": 1.998639275588839, + "learning_rate": 6.382456830256612e-07, + "loss": 0.8015, + "step": 9243 + }, + { + "epoch": 0.8891881492881878, + "grad_norm": 2.1865496976044856, + "learning_rate": 6.371509628425021e-07, + "loss": 0.9064, + "step": 9244 + }, + { + "epoch": 0.8892843401308196, + "grad_norm": 1.9139191067903407, + "learning_rate": 6.360571514044589e-07, + "loss": 0.8177, + "step": 9245 + }, + { + "epoch": 0.8893805309734514, + "grad_norm": 2.0378090938507656, + "learning_rate": 6.349642488176943e-07, + "loss": 0.9066, + "step": 9246 + }, + { + "epoch": 0.8894767218160831, + "grad_norm": 1.7641283023798193, + "learning_rate": 6.338722551882815e-07, + "loss": 0.7964, + "step": 9247 + }, + { + "epoch": 0.8895729126587149, + "grad_norm": 2.376130536151312, + "learning_rate": 6.327811706222097e-07, + "loss": 0.9293, + "step": 9248 + }, + { + "epoch": 0.8896691035013466, + "grad_norm": 1.9591100996047062, + "learning_rate": 6.316909952253791e-07, + "loss": 0.8764, + "step": 9249 + }, + { + "epoch": 0.8897652943439784, + "grad_norm": 1.8503596578906671, + "learning_rate": 6.306017291035981e-07, + "loss": 0.9144, + "step": 9250 + }, + { + "epoch": 0.8898614851866102, + "grad_norm": 2.219868694131462, + "learning_rate": 6.295133723625923e-07, + "loss": 0.9157, + "step": 9251 + }, + { + "epoch": 0.889957676029242, + "grad_norm": 2.082581366023959, + "learning_rate": 6.284259251079939e-07, + "loss": 0.9481, + "step": 9252 + }, + { + "epoch": 0.8900538668718738, + "grad_norm": 1.6263809682088424, + "learning_rate": 6.273393874453504e-07, + "loss": 0.7745, + "step": 9253 + }, + { + "epoch": 0.8901500577145056, + "grad_norm": 2.0076841115799837, + "learning_rate": 6.262537594801177e-07, + "loss": 1.0318, + "step": 9254 + }, + { + "epoch": 0.8902462485571374, + "grad_norm": 1.9868493707686552, + "learning_rate": 6.25169041317667e-07, + "loss": 0.8934, + "step": 9255 + }, + { + "epoch": 0.8903424393997691, + "grad_norm": 1.9569837457059645, + "learning_rate": 6.240852330632796e-07, + "loss": 0.9498, + "step": 9256 + }, + { + "epoch": 0.8904386302424009, + "grad_norm": 1.9762166236522343, + "learning_rate": 6.230023348221492e-07, + "loss": 0.8437, + "step": 9257 + }, + { + "epoch": 0.8905348210850327, + "grad_norm": 1.9783037672373445, + "learning_rate": 6.219203466993762e-07, + "loss": 0.8441, + "step": 9258 + }, + { + "epoch": 0.8906310119276645, + "grad_norm": 1.9936202436876664, + "learning_rate": 6.208392687999831e-07, + "loss": 0.8596, + "step": 9259 + }, + { + "epoch": 0.8907272027702963, + "grad_norm": 1.8399794567376022, + "learning_rate": 6.197591012288918e-07, + "loss": 0.9087, + "step": 9260 + }, + { + "epoch": 0.8908233936129281, + "grad_norm": 2.1822603770408686, + "learning_rate": 6.186798440909447e-07, + "loss": 0.8543, + "step": 9261 + }, + { + "epoch": 0.8909195844555599, + "grad_norm": 2.1035011996058848, + "learning_rate": 6.17601497490895e-07, + "loss": 0.7982, + "step": 9262 + }, + { + "epoch": 0.8910157752981916, + "grad_norm": 2.1926535827246623, + "learning_rate": 6.165240615334034e-07, + "loss": 0.8883, + "step": 9263 + }, + { + "epoch": 0.8911119661408234, + "grad_norm": 1.942105625685171, + "learning_rate": 6.154475363230417e-07, + "loss": 0.9049, + "step": 9264 + }, + { + "epoch": 0.8912081569834551, + "grad_norm": 1.948415202493371, + "learning_rate": 6.143719219642996e-07, + "loss": 0.9424, + "step": 9265 + }, + { + "epoch": 0.8913043478260869, + "grad_norm": 2.398046930904414, + "learning_rate": 6.132972185615749e-07, + "loss": 0.9562, + "step": 9266 + }, + { + "epoch": 0.8914005386687187, + "grad_norm": 1.7830476326626745, + "learning_rate": 6.122234262191751e-07, + "loss": 0.8379, + "step": 9267 + }, + { + "epoch": 0.8914967295113505, + "grad_norm": 2.0509336557641555, + "learning_rate": 6.111505450413202e-07, + "loss": 0.9154, + "step": 9268 + }, + { + "epoch": 0.8915929203539823, + "grad_norm": 2.0502538150891456, + "learning_rate": 6.100785751321458e-07, + "loss": 0.9264, + "step": 9269 + }, + { + "epoch": 0.8916891111966141, + "grad_norm": 1.9350576336986238, + "learning_rate": 6.090075165956943e-07, + "loss": 0.9097, + "step": 9270 + }, + { + "epoch": 0.8917853020392459, + "grad_norm": 2.3706158619532833, + "learning_rate": 6.079373695359192e-07, + "loss": 0.9357, + "step": 9271 + }, + { + "epoch": 0.8918814928818777, + "grad_norm": 2.2578302197610487, + "learning_rate": 6.068681340566896e-07, + "loss": 0.8938, + "step": 9272 + }, + { + "epoch": 0.8919776837245095, + "grad_norm": 1.613472198940949, + "learning_rate": 6.057998102617835e-07, + "loss": 0.7867, + "step": 9273 + }, + { + "epoch": 0.8920738745671412, + "grad_norm": 2.4069644673432298, + "learning_rate": 6.047323982548924e-07, + "loss": 0.9676, + "step": 9274 + }, + { + "epoch": 0.892170065409773, + "grad_norm": 1.9751680518877237, + "learning_rate": 6.036658981396137e-07, + "loss": 0.9856, + "step": 9275 + }, + { + "epoch": 0.8922662562524047, + "grad_norm": 2.0715244057860294, + "learning_rate": 6.026003100194633e-07, + "loss": 0.9053, + "step": 9276 + }, + { + "epoch": 0.8923624470950365, + "grad_norm": 2.1225506744897005, + "learning_rate": 6.015356339978673e-07, + "loss": 0.8827, + "step": 9277 + }, + { + "epoch": 0.8924586379376683, + "grad_norm": 2.1102237201412595, + "learning_rate": 6.004718701781575e-07, + "loss": 0.9927, + "step": 9278 + }, + { + "epoch": 0.8925548287803001, + "grad_norm": 2.4042224150525526, + "learning_rate": 5.994090186635848e-07, + "loss": 0.966, + "step": 9279 + }, + { + "epoch": 0.8926510196229319, + "grad_norm": 1.9180004345922081, + "learning_rate": 5.983470795573088e-07, + "loss": 0.929, + "step": 9280 + }, + { + "epoch": 0.8927472104655637, + "grad_norm": 2.7379898867072683, + "learning_rate": 5.972860529623959e-07, + "loss": 0.9117, + "step": 9281 + }, + { + "epoch": 0.8928434013081955, + "grad_norm": 2.1409271464428272, + "learning_rate": 5.962259389818292e-07, + "loss": 0.9498, + "step": 9282 + }, + { + "epoch": 0.8929395921508272, + "grad_norm": 1.9064876017081587, + "learning_rate": 5.951667377185033e-07, + "loss": 0.8485, + "step": 9283 + }, + { + "epoch": 0.893035782993459, + "grad_norm": 2.2768460513428663, + "learning_rate": 5.941084492752236e-07, + "loss": 0.8681, + "step": 9284 + }, + { + "epoch": 0.8931319738360908, + "grad_norm": 2.0935368027062737, + "learning_rate": 5.930510737547024e-07, + "loss": 0.8923, + "step": 9285 + }, + { + "epoch": 0.8932281646787226, + "grad_norm": 2.1228842650007675, + "learning_rate": 5.91994611259572e-07, + "loss": 0.9863, + "step": 9286 + }, + { + "epoch": 0.8933243555213544, + "grad_norm": 1.9388114325094628, + "learning_rate": 5.909390618923672e-07, + "loss": 0.7748, + "step": 9287 + }, + { + "epoch": 0.8934205463639862, + "grad_norm": 2.0024753612868147, + "learning_rate": 5.898844257555392e-07, + "loss": 0.8716, + "step": 9288 + }, + { + "epoch": 0.893516737206618, + "grad_norm": 1.828585002222719, + "learning_rate": 5.888307029514484e-07, + "loss": 0.8928, + "step": 9289 + }, + { + "epoch": 0.8936129280492497, + "grad_norm": 2.216831360165824, + "learning_rate": 5.87777893582372e-07, + "loss": 0.9713, + "step": 9290 + }, + { + "epoch": 0.8937091188918815, + "grad_norm": 2.1085280405922, + "learning_rate": 5.86725997750488e-07, + "loss": 0.8627, + "step": 9291 + }, + { + "epoch": 0.8938053097345132, + "grad_norm": 1.8774579054798466, + "learning_rate": 5.856750155578983e-07, + "loss": 0.8843, + "step": 9292 + }, + { + "epoch": 0.893901500577145, + "grad_norm": 2.3108033034719426, + "learning_rate": 5.846249471066046e-07, + "loss": 0.9805, + "step": 9293 + }, + { + "epoch": 0.8939976914197768, + "grad_norm": 1.8838036408684564, + "learning_rate": 5.835757924985286e-07, + "loss": 0.9031, + "step": 9294 + }, + { + "epoch": 0.8940938822624086, + "grad_norm": 2.170663442803962, + "learning_rate": 5.825275518354967e-07, + "loss": 0.9283, + "step": 9295 + }, + { + "epoch": 0.8941900731050404, + "grad_norm": 2.0798073957357404, + "learning_rate": 5.81480225219252e-07, + "loss": 0.9841, + "step": 9296 + }, + { + "epoch": 0.8942862639476722, + "grad_norm": 2.1604308480607295, + "learning_rate": 5.804338127514475e-07, + "loss": 0.8965, + "step": 9297 + }, + { + "epoch": 0.894382454790304, + "grad_norm": 2.1372726711673735, + "learning_rate": 5.793883145336443e-07, + "loss": 0.9384, + "step": 9298 + }, + { + "epoch": 0.8944786456329358, + "grad_norm": 1.9977328686396796, + "learning_rate": 5.783437306673168e-07, + "loss": 0.9045, + "step": 9299 + }, + { + "epoch": 0.8945748364755676, + "grad_norm": 1.9191253243645807, + "learning_rate": 5.773000612538505e-07, + "loss": 0.8755, + "step": 9300 + }, + { + "epoch": 0.8946710273181993, + "grad_norm": 1.7663947440291146, + "learning_rate": 5.762573063945465e-07, + "loss": 0.8447, + "step": 9301 + }, + { + "epoch": 0.894767218160831, + "grad_norm": 2.0817446691182098, + "learning_rate": 5.752154661906085e-07, + "loss": 0.8867, + "step": 9302 + }, + { + "epoch": 0.8948634090034628, + "grad_norm": 2.1221138003423774, + "learning_rate": 5.741745407431587e-07, + "loss": 0.8491, + "step": 9303 + }, + { + "epoch": 0.8949595998460946, + "grad_norm": 1.8608880620951351, + "learning_rate": 5.731345301532265e-07, + "loss": 0.9183, + "step": 9304 + }, + { + "epoch": 0.8950557906887264, + "grad_norm": 1.7001386554632494, + "learning_rate": 5.720954345217555e-07, + "loss": 0.7826, + "step": 9305 + }, + { + "epoch": 0.8951519815313582, + "grad_norm": 1.8790563517151369, + "learning_rate": 5.710572539495962e-07, + "loss": 0.8316, + "step": 9306 + }, + { + "epoch": 0.89524817237399, + "grad_norm": 2.123534723020043, + "learning_rate": 5.700199885375146e-07, + "loss": 0.9286, + "step": 9307 + }, + { + "epoch": 0.8953443632166218, + "grad_norm": 2.0554251987363754, + "learning_rate": 5.68983638386188e-07, + "loss": 0.847, + "step": 9308 + }, + { + "epoch": 0.8954405540592536, + "grad_norm": 1.8990770434190694, + "learning_rate": 5.679482035962014e-07, + "loss": 0.8671, + "step": 9309 + }, + { + "epoch": 0.8955367449018854, + "grad_norm": 2.062497175171352, + "learning_rate": 5.669136842680512e-07, + "loss": 0.8104, + "step": 9310 + }, + { + "epoch": 0.8956329357445171, + "grad_norm": 2.1283147144275953, + "learning_rate": 5.658800805021492e-07, + "loss": 0.9949, + "step": 9311 + }, + { + "epoch": 0.8957291265871489, + "grad_norm": 2.091657477071494, + "learning_rate": 5.648473923988129e-07, + "loss": 1.005, + "step": 9312 + }, + { + "epoch": 0.8958253174297807, + "grad_norm": 2.1233838728901473, + "learning_rate": 5.638156200582745e-07, + "loss": 0.9324, + "step": 9313 + }, + { + "epoch": 0.8959215082724125, + "grad_norm": 2.13305166195019, + "learning_rate": 5.627847635806771e-07, + "loss": 0.9604, + "step": 9314 + }, + { + "epoch": 0.8960176991150443, + "grad_norm": 2.1228941599506754, + "learning_rate": 5.617548230660763e-07, + "loss": 0.8977, + "step": 9315 + }, + { + "epoch": 0.896113889957676, + "grad_norm": 1.9935368049794568, + "learning_rate": 5.607257986144321e-07, + "loss": 0.8816, + "step": 9316 + }, + { + "epoch": 0.8962100808003078, + "grad_norm": 1.9184728977182872, + "learning_rate": 5.596976903256213e-07, + "loss": 0.8241, + "step": 9317 + }, + { + "epoch": 0.8963062716429396, + "grad_norm": 2.0250548907148325, + "learning_rate": 5.58670498299434e-07, + "loss": 0.7925, + "step": 9318 + }, + { + "epoch": 0.8964024624855714, + "grad_norm": 1.9779875690102109, + "learning_rate": 5.576442226355638e-07, + "loss": 0.9328, + "step": 9319 + }, + { + "epoch": 0.8964986533282031, + "grad_norm": 2.4285347098119536, + "learning_rate": 5.566188634336212e-07, + "loss": 0.9375, + "step": 9320 + }, + { + "epoch": 0.8965948441708349, + "grad_norm": 2.4053916537861775, + "learning_rate": 5.555944207931285e-07, + "loss": 1.0088, + "step": 9321 + }, + { + "epoch": 0.8966910350134667, + "grad_norm": 2.1014685226506464, + "learning_rate": 5.545708948135142e-07, + "loss": 0.8815, + "step": 9322 + }, + { + "epoch": 0.8967872258560985, + "grad_norm": 2.130525146535093, + "learning_rate": 5.535482855941188e-07, + "loss": 0.9414, + "step": 9323 + }, + { + "epoch": 0.8968834166987303, + "grad_norm": 2.506828833003266, + "learning_rate": 5.525265932341984e-07, + "loss": 0.938, + "step": 9324 + }, + { + "epoch": 0.8969796075413621, + "grad_norm": 2.1329848488019723, + "learning_rate": 5.515058178329169e-07, + "loss": 0.7673, + "step": 9325 + }, + { + "epoch": 0.8970757983839939, + "grad_norm": 2.0504588977120983, + "learning_rate": 5.504859594893475e-07, + "loss": 0.9129, + "step": 9326 + }, + { + "epoch": 0.8971719892266257, + "grad_norm": 1.8409548061349053, + "learning_rate": 5.494670183024764e-07, + "loss": 0.8995, + "step": 9327 + }, + { + "epoch": 0.8972681800692575, + "grad_norm": 2.0150544010861404, + "learning_rate": 5.484489943712013e-07, + "loss": 0.8955, + "step": 9328 + }, + { + "epoch": 0.8973643709118891, + "grad_norm": 2.045015134868283, + "learning_rate": 5.474318877943319e-07, + "loss": 0.8906, + "step": 9329 + }, + { + "epoch": 0.8974605617545209, + "grad_norm": 2.1286775656743013, + "learning_rate": 5.464156986705826e-07, + "loss": 0.8364, + "step": 9330 + }, + { + "epoch": 0.8975567525971527, + "grad_norm": 1.9044092029625903, + "learning_rate": 5.45400427098588e-07, + "loss": 0.9039, + "step": 9331 + }, + { + "epoch": 0.8976529434397845, + "grad_norm": 1.7530862044691076, + "learning_rate": 5.443860731768869e-07, + "loss": 0.7927, + "step": 9332 + }, + { + "epoch": 0.8977491342824163, + "grad_norm": 2.1709046791621294, + "learning_rate": 5.433726370039316e-07, + "loss": 0.936, + "step": 9333 + }, + { + "epoch": 0.8978453251250481, + "grad_norm": 1.9028523495237952, + "learning_rate": 5.423601186780836e-07, + "loss": 0.9232, + "step": 9334 + }, + { + "epoch": 0.8979415159676799, + "grad_norm": 2.0562033815393477, + "learning_rate": 5.413485182976175e-07, + "loss": 0.9071, + "step": 9335 + }, + { + "epoch": 0.8980377068103117, + "grad_norm": 1.8792448056485256, + "learning_rate": 5.403378359607181e-07, + "loss": 0.8612, + "step": 9336 + }, + { + "epoch": 0.8981338976529435, + "grad_norm": 2.0816026288996303, + "learning_rate": 5.393280717654803e-07, + "loss": 0.7938, + "step": 9337 + }, + { + "epoch": 0.8982300884955752, + "grad_norm": 2.02831523096087, + "learning_rate": 5.383192258099113e-07, + "loss": 0.9862, + "step": 9338 + }, + { + "epoch": 0.898326279338207, + "grad_norm": 1.9227463215139, + "learning_rate": 5.373112981919271e-07, + "loss": 0.8659, + "step": 9339 + }, + { + "epoch": 0.8984224701808388, + "grad_norm": 2.317826509854393, + "learning_rate": 5.36304289009355e-07, + "loss": 0.8888, + "step": 9340 + }, + { + "epoch": 0.8985186610234706, + "grad_norm": 2.0948952310410505, + "learning_rate": 5.352981983599359e-07, + "loss": 0.8864, + "step": 9341 + }, + { + "epoch": 0.8986148518661023, + "grad_norm": 2.0558758343065606, + "learning_rate": 5.342930263413193e-07, + "loss": 0.8754, + "step": 9342 + }, + { + "epoch": 0.8987110427087341, + "grad_norm": 2.213876448045447, + "learning_rate": 5.332887730510639e-07, + "loss": 0.9703, + "step": 9343 + }, + { + "epoch": 0.8988072335513659, + "grad_norm": 1.8533676220754927, + "learning_rate": 5.322854385866439e-07, + "loss": 0.8209, + "step": 9344 + }, + { + "epoch": 0.8989034243939977, + "grad_norm": 1.7655418602879775, + "learning_rate": 5.312830230454391e-07, + "loss": 0.7804, + "step": 9345 + }, + { + "epoch": 0.8989996152366295, + "grad_norm": 1.9541386164906889, + "learning_rate": 5.302815265247452e-07, + "loss": 0.8575, + "step": 9346 + }, + { + "epoch": 0.8990958060792612, + "grad_norm": 1.8683961748305575, + "learning_rate": 5.292809491217621e-07, + "loss": 0.8171, + "step": 9347 + }, + { + "epoch": 0.899191996921893, + "grad_norm": 2.1926950364760778, + "learning_rate": 5.282812909336077e-07, + "loss": 0.9726, + "step": 9348 + }, + { + "epoch": 0.8992881877645248, + "grad_norm": 2.0108497946339816, + "learning_rate": 5.272825520573077e-07, + "loss": 0.8856, + "step": 9349 + }, + { + "epoch": 0.8993843786071566, + "grad_norm": 2.1042198828707233, + "learning_rate": 5.262847325897968e-07, + "loss": 0.8156, + "step": 9350 + }, + { + "epoch": 0.8994805694497884, + "grad_norm": 2.3008120917564066, + "learning_rate": 5.25287832627921e-07, + "loss": 0.8591, + "step": 9351 + }, + { + "epoch": 0.8995767602924202, + "grad_norm": 2.0008485068629085, + "learning_rate": 5.242918522684392e-07, + "loss": 0.907, + "step": 9352 + }, + { + "epoch": 0.899672951135052, + "grad_norm": 2.166450504129422, + "learning_rate": 5.232967916080211e-07, + "loss": 0.9015, + "step": 9353 + }, + { + "epoch": 0.8997691419776838, + "grad_norm": 2.1058093127667643, + "learning_rate": 5.22302650743245e-07, + "loss": 0.9383, + "step": 9354 + }, + { + "epoch": 0.8998653328203156, + "grad_norm": 2.3091920313042023, + "learning_rate": 5.213094297706001e-07, + "loss": 0.92, + "step": 9355 + }, + { + "epoch": 0.8999615236629472, + "grad_norm": 1.970769643108414, + "learning_rate": 5.203171287864872e-07, + "loss": 0.9359, + "step": 9356 + }, + { + "epoch": 0.900057714505579, + "grad_norm": 1.8822698097492618, + "learning_rate": 5.193257478872193e-07, + "loss": 1.0356, + "step": 9357 + }, + { + "epoch": 0.9001539053482108, + "grad_norm": 1.9366187954229719, + "learning_rate": 5.183352871690162e-07, + "loss": 0.864, + "step": 9358 + }, + { + "epoch": 0.9002500961908426, + "grad_norm": 1.9766004911373978, + "learning_rate": 5.173457467280108e-07, + "loss": 0.889, + "step": 9359 + }, + { + "epoch": 0.9003462870334744, + "grad_norm": 1.9494453172307902, + "learning_rate": 5.163571266602485e-07, + "loss": 0.9722, + "step": 9360 + }, + { + "epoch": 0.9004424778761062, + "grad_norm": 2.0595259994651167, + "learning_rate": 5.153694270616838e-07, + "loss": 0.9498, + "step": 9361 + }, + { + "epoch": 0.900538668718738, + "grad_norm": 2.2320835977596363, + "learning_rate": 5.143826480281778e-07, + "loss": 0.9277, + "step": 9362 + }, + { + "epoch": 0.9006348595613698, + "grad_norm": 1.7430778798357902, + "learning_rate": 5.133967896555081e-07, + "loss": 0.7675, + "step": 9363 + }, + { + "epoch": 0.9007310504040016, + "grad_norm": 2.5438661555357047, + "learning_rate": 5.124118520393606e-07, + "loss": 0.8185, + "step": 9364 + }, + { + "epoch": 0.9008272412466333, + "grad_norm": 2.212830377426942, + "learning_rate": 5.114278352753321e-07, + "loss": 1.026, + "step": 9365 + }, + { + "epoch": 0.9009234320892651, + "grad_norm": 2.1210299386029763, + "learning_rate": 5.104447394589295e-07, + "loss": 0.8838, + "step": 9366 + }, + { + "epoch": 0.9010196229318969, + "grad_norm": 1.7137103378883207, + "learning_rate": 5.094625646855733e-07, + "loss": 0.7345, + "step": 9367 + }, + { + "epoch": 0.9011158137745287, + "grad_norm": 1.887303397821032, + "learning_rate": 5.084813110505871e-07, + "loss": 0.9425, + "step": 9368 + }, + { + "epoch": 0.9012120046171604, + "grad_norm": 2.0950471860196305, + "learning_rate": 5.075009786492135e-07, + "loss": 0.8129, + "step": 9369 + }, + { + "epoch": 0.9013081954597922, + "grad_norm": 2.0300820592243722, + "learning_rate": 5.065215675766023e-07, + "loss": 0.8244, + "step": 9370 + }, + { + "epoch": 0.901404386302424, + "grad_norm": 1.9918197915957243, + "learning_rate": 5.055430779278115e-07, + "loss": 0.8256, + "step": 9371 + }, + { + "epoch": 0.9015005771450558, + "grad_norm": 1.9446802058009924, + "learning_rate": 5.045655097978131e-07, + "loss": 0.828, + "step": 9372 + }, + { + "epoch": 0.9015967679876876, + "grad_norm": 1.8740739848574228, + "learning_rate": 5.0358886328149e-07, + "loss": 0.9981, + "step": 9373 + }, + { + "epoch": 0.9016929588303193, + "grad_norm": 1.685883878042805, + "learning_rate": 5.026131384736321e-07, + "loss": 0.9041, + "step": 9374 + }, + { + "epoch": 0.9017891496729511, + "grad_norm": 1.9587372453630385, + "learning_rate": 5.016383354689425e-07, + "loss": 0.904, + "step": 9375 + }, + { + "epoch": 0.9018853405155829, + "grad_norm": 1.7016613037334796, + "learning_rate": 5.006644543620342e-07, + "loss": 0.6888, + "step": 9376 + }, + { + "epoch": 0.9019815313582147, + "grad_norm": 1.9210589000854583, + "learning_rate": 4.996914952474319e-07, + "loss": 0.8364, + "step": 9377 + }, + { + "epoch": 0.9020777222008465, + "grad_norm": 1.721675808181308, + "learning_rate": 4.987194582195687e-07, + "loss": 0.8227, + "step": 9378 + }, + { + "epoch": 0.9021739130434783, + "grad_norm": 2.1575682406036654, + "learning_rate": 4.977483433727881e-07, + "loss": 0.8656, + "step": 9379 + }, + { + "epoch": 0.9022701038861101, + "grad_norm": 1.9186980215068834, + "learning_rate": 4.967781508013459e-07, + "loss": 0.8703, + "step": 9380 + }, + { + "epoch": 0.9023662947287419, + "grad_norm": 2.038692883839514, + "learning_rate": 4.958088805994088e-07, + "loss": 0.7969, + "step": 9381 + }, + { + "epoch": 0.9024624855713737, + "grad_norm": 2.0411798314813234, + "learning_rate": 4.948405328610506e-07, + "loss": 0.8993, + "step": 9382 + }, + { + "epoch": 0.9025586764140053, + "grad_norm": 1.9184907476403785, + "learning_rate": 4.938731076802605e-07, + "loss": 0.8912, + "step": 9383 + }, + { + "epoch": 0.9026548672566371, + "grad_norm": 2.1898265344742227, + "learning_rate": 4.929066051509346e-07, + "loss": 1.0124, + "step": 9384 + }, + { + "epoch": 0.9027510580992689, + "grad_norm": 1.8047018046495968, + "learning_rate": 4.919410253668788e-07, + "loss": 0.8592, + "step": 9385 + }, + { + "epoch": 0.9028472489419007, + "grad_norm": 2.4352325222419937, + "learning_rate": 4.909763684218116e-07, + "loss": 0.9632, + "step": 9386 + }, + { + "epoch": 0.9029434397845325, + "grad_norm": 2.196601960977354, + "learning_rate": 4.900126344093614e-07, + "loss": 0.8909, + "step": 9387 + }, + { + "epoch": 0.9030396306271643, + "grad_norm": 2.089963291061896, + "learning_rate": 4.890498234230689e-07, + "loss": 0.9613, + "step": 9388 + }, + { + "epoch": 0.9031358214697961, + "grad_norm": 1.83576316899071, + "learning_rate": 4.880879355563794e-07, + "loss": 0.8664, + "step": 9389 + }, + { + "epoch": 0.9032320123124279, + "grad_norm": 2.057264277003193, + "learning_rate": 4.871269709026561e-07, + "loss": 0.8991, + "step": 9390 + }, + { + "epoch": 0.9033282031550597, + "grad_norm": 1.8162030975355294, + "learning_rate": 4.861669295551664e-07, + "loss": 0.7645, + "step": 9391 + }, + { + "epoch": 0.9034243939976914, + "grad_norm": 1.9957953131278938, + "learning_rate": 4.852078116070902e-07, + "loss": 0.867, + "step": 9392 + }, + { + "epoch": 0.9035205848403232, + "grad_norm": 2.379515863410986, + "learning_rate": 4.842496171515198e-07, + "loss": 0.9423, + "step": 9393 + }, + { + "epoch": 0.903616775682955, + "grad_norm": 1.8389620892216438, + "learning_rate": 4.832923462814565e-07, + "loss": 0.9374, + "step": 9394 + }, + { + "epoch": 0.9037129665255867, + "grad_norm": 1.900609864949451, + "learning_rate": 4.823359990898091e-07, + "loss": 0.9196, + "step": 9395 + }, + { + "epoch": 0.9038091573682185, + "grad_norm": 1.9582482810189925, + "learning_rate": 4.813805756694035e-07, + "loss": 0.9596, + "step": 9396 + }, + { + "epoch": 0.9039053482108503, + "grad_norm": 1.9254602783975514, + "learning_rate": 4.804260761129675e-07, + "loss": 0.9029, + "step": 9397 + }, + { + "epoch": 0.9040015390534821, + "grad_norm": 2.2192426989018967, + "learning_rate": 4.794725005131462e-07, + "loss": 1.0079, + "step": 9398 + }, + { + "epoch": 0.9040977298961139, + "grad_norm": 1.8421554741889252, + "learning_rate": 4.785198489624909e-07, + "loss": 0.8711, + "step": 9399 + }, + { + "epoch": 0.9041939207387457, + "grad_norm": 1.9905292498180853, + "learning_rate": 4.775681215534656e-07, + "loss": 0.9241, + "step": 9400 + }, + { + "epoch": 0.9042901115813774, + "grad_norm": 2.025722803596472, + "learning_rate": 4.7661731837844395e-07, + "loss": 0.9379, + "step": 9401 + }, + { + "epoch": 0.9043863024240092, + "grad_norm": 2.031425586025061, + "learning_rate": 4.7566743952970894e-07, + "loss": 0.9366, + "step": 9402 + }, + { + "epoch": 0.904482493266641, + "grad_norm": 1.8502131659325707, + "learning_rate": 4.747184850994524e-07, + "loss": 0.7783, + "step": 9403 + }, + { + "epoch": 0.9045786841092728, + "grad_norm": 1.8994141212703826, + "learning_rate": 4.7377045517978173e-07, + "loss": 0.9936, + "step": 9404 + }, + { + "epoch": 0.9046748749519046, + "grad_norm": 1.9459868165786856, + "learning_rate": 4.7282334986271104e-07, + "loss": 0.9836, + "step": 9405 + }, + { + "epoch": 0.9047710657945364, + "grad_norm": 1.9223246320318566, + "learning_rate": 4.7187716924016355e-07, + "loss": 0.9665, + "step": 9406 + }, + { + "epoch": 0.9048672566371682, + "grad_norm": 1.876364238492994, + "learning_rate": 4.709319134039758e-07, + "loss": 0.8859, + "step": 9407 + }, + { + "epoch": 0.9049634474798, + "grad_norm": 1.8125702844606018, + "learning_rate": 4.6998758244588995e-07, + "loss": 0.8123, + "step": 9408 + }, + { + "epoch": 0.9050596383224317, + "grad_norm": 2.1700973857765318, + "learning_rate": 4.6904417645756596e-07, + "loss": 0.8678, + "step": 9409 + }, + { + "epoch": 0.9051558291650635, + "grad_norm": 1.8812762010453645, + "learning_rate": 4.6810169553056616e-07, + "loss": 0.8949, + "step": 9410 + }, + { + "epoch": 0.9052520200076952, + "grad_norm": 2.0031761225414417, + "learning_rate": 4.671601397563674e-07, + "loss": 0.9878, + "step": 9411 + }, + { + "epoch": 0.905348210850327, + "grad_norm": 2.066483845022001, + "learning_rate": 4.662195092263566e-07, + "loss": 0.9096, + "step": 9412 + }, + { + "epoch": 0.9054444016929588, + "grad_norm": 2.0695489087400403, + "learning_rate": 4.652798040318296e-07, + "loss": 0.9521, + "step": 9413 + }, + { + "epoch": 0.9055405925355906, + "grad_norm": 1.9209083959329567, + "learning_rate": 4.643410242639912e-07, + "loss": 0.8483, + "step": 9414 + }, + { + "epoch": 0.9056367833782224, + "grad_norm": 1.8301708680715914, + "learning_rate": 4.634031700139596e-07, + "loss": 0.8022, + "step": 9415 + }, + { + "epoch": 0.9057329742208542, + "grad_norm": 2.157724692704469, + "learning_rate": 4.6246624137276206e-07, + "loss": 0.8572, + "step": 9416 + }, + { + "epoch": 0.905829165063486, + "grad_norm": 1.9319571981636159, + "learning_rate": 4.615302384313347e-07, + "loss": 0.9466, + "step": 9417 + }, + { + "epoch": 0.9059253559061178, + "grad_norm": 2.0975503767651507, + "learning_rate": 4.605951612805237e-07, + "loss": 0.8961, + "step": 9418 + }, + { + "epoch": 0.9060215467487496, + "grad_norm": 1.9153023009977004, + "learning_rate": 4.59661010011091e-07, + "loss": 0.8862, + "step": 9419 + }, + { + "epoch": 0.9061177375913813, + "grad_norm": 2.244180704836283, + "learning_rate": 4.587277847136984e-07, + "loss": 0.8548, + "step": 9420 + }, + { + "epoch": 0.906213928434013, + "grad_norm": 2.0306030106398625, + "learning_rate": 4.5779548547892573e-07, + "loss": 0.8841, + "step": 9421 + }, + { + "epoch": 0.9063101192766448, + "grad_norm": 1.841676256716317, + "learning_rate": 4.568641123972606e-07, + "loss": 0.8711, + "step": 9422 + }, + { + "epoch": 0.9064063101192766, + "grad_norm": 2.2551062336647703, + "learning_rate": 4.5593366555910065e-07, + "loss": 0.907, + "step": 9423 + }, + { + "epoch": 0.9065025009619084, + "grad_norm": 2.236834902991878, + "learning_rate": 4.550041450547549e-07, + "loss": 0.945, + "step": 9424 + }, + { + "epoch": 0.9065986918045402, + "grad_norm": 1.9632081909006296, + "learning_rate": 4.540755509744388e-07, + "loss": 0.9006, + "step": 9425 + }, + { + "epoch": 0.906694882647172, + "grad_norm": 2.0777413083217455, + "learning_rate": 4.5314788340828365e-07, + "loss": 0.9254, + "step": 9426 + }, + { + "epoch": 0.9067910734898038, + "grad_norm": 2.5140963755648102, + "learning_rate": 4.522211424463241e-07, + "loss": 0.8645, + "step": 9427 + }, + { + "epoch": 0.9068872643324356, + "grad_norm": 2.102461559745254, + "learning_rate": 4.512953281785104e-07, + "loss": 0.91, + "step": 9428 + }, + { + "epoch": 0.9069834551750673, + "grad_norm": 1.832549987380974, + "learning_rate": 4.5037044069470184e-07, + "loss": 0.7552, + "step": 9429 + }, + { + "epoch": 0.9070796460176991, + "grad_norm": 2.546048670520564, + "learning_rate": 4.494464800846654e-07, + "loss": 0.9467, + "step": 9430 + }, + { + "epoch": 0.9071758368603309, + "grad_norm": 2.2607950964500354, + "learning_rate": 4.4852344643807724e-07, + "loss": 0.9075, + "step": 9431 + }, + { + "epoch": 0.9072720277029627, + "grad_norm": 1.94186067335933, + "learning_rate": 4.476013398445289e-07, + "loss": 0.8131, + "step": 9432 + }, + { + "epoch": 0.9073682185455945, + "grad_norm": 1.8450615088817262, + "learning_rate": 4.466801603935178e-07, + "loss": 0.9267, + "step": 9433 + }, + { + "epoch": 0.9074644093882263, + "grad_norm": 1.998885238207198, + "learning_rate": 4.4575990817445234e-07, + "loss": 0.9283, + "step": 9434 + }, + { + "epoch": 0.907560600230858, + "grad_norm": 2.0220850985123278, + "learning_rate": 4.4484058327665e-07, + "loss": 0.9016, + "step": 9435 + }, + { + "epoch": 0.9076567910734898, + "grad_norm": 1.844338700302622, + "learning_rate": 4.4392218578934164e-07, + "loss": 0.8772, + "step": 9436 + }, + { + "epoch": 0.9077529819161216, + "grad_norm": 1.5528078032843864, + "learning_rate": 4.4300471580166373e-07, + "loss": 0.6468, + "step": 9437 + }, + { + "epoch": 0.9078491727587533, + "grad_norm": 2.0153099322625945, + "learning_rate": 4.4208817340266385e-07, + "loss": 0.9314, + "step": 9438 + }, + { + "epoch": 0.9079453636013851, + "grad_norm": 1.949672541882256, + "learning_rate": 4.4117255868130093e-07, + "loss": 0.8672, + "step": 9439 + }, + { + "epoch": 0.9080415544440169, + "grad_norm": 2.1248810509092175, + "learning_rate": 4.4025787172644495e-07, + "loss": 0.9178, + "step": 9440 + }, + { + "epoch": 0.9081377452866487, + "grad_norm": 1.7685366026283895, + "learning_rate": 4.393441126268716e-07, + "loss": 0.9846, + "step": 9441 + }, + { + "epoch": 0.9082339361292805, + "grad_norm": 1.959168389411376, + "learning_rate": 4.384312814712721e-07, + "loss": 0.849, + "step": 9442 + }, + { + "epoch": 0.9083301269719123, + "grad_norm": 2.0558874477445035, + "learning_rate": 4.3751937834824344e-07, + "loss": 0.9071, + "step": 9443 + }, + { + "epoch": 0.9084263178145441, + "grad_norm": 2.253071574862059, + "learning_rate": 4.366084033462914e-07, + "loss": 0.9472, + "step": 9444 + }, + { + "epoch": 0.9085225086571759, + "grad_norm": 2.079841361822861, + "learning_rate": 4.356983565538353e-07, + "loss": 0.8905, + "step": 9445 + }, + { + "epoch": 0.9086186994998077, + "grad_norm": 2.161051659763665, + "learning_rate": 4.3478923805920335e-07, + "loss": 0.9386, + "step": 9446 + }, + { + "epoch": 0.9087148903424394, + "grad_norm": 2.156570305638997, + "learning_rate": 4.3388104795063503e-07, + "loss": 0.8836, + "step": 9447 + }, + { + "epoch": 0.9088110811850711, + "grad_norm": 2.1816521472806913, + "learning_rate": 4.329737863162753e-07, + "loss": 0.883, + "step": 9448 + }, + { + "epoch": 0.9089072720277029, + "grad_norm": 2.1204848628823783, + "learning_rate": 4.3206745324418266e-07, + "loss": 0.8231, + "step": 9449 + }, + { + "epoch": 0.9090034628703347, + "grad_norm": 2.0078008370250306, + "learning_rate": 4.311620488223256e-07, + "loss": 0.8931, + "step": 9450 + }, + { + "epoch": 0.9090996537129665, + "grad_norm": 1.9699561057790151, + "learning_rate": 4.302575731385783e-07, + "loss": 0.9088, + "step": 9451 + }, + { + "epoch": 0.9091958445555983, + "grad_norm": 2.0248179557640387, + "learning_rate": 4.2935402628073166e-07, + "loss": 0.9239, + "step": 9452 + }, + { + "epoch": 0.9092920353982301, + "grad_norm": 2.0257159957395228, + "learning_rate": 4.28451408336481e-07, + "loss": 0.9286, + "step": 9453 + }, + { + "epoch": 0.9093882262408619, + "grad_norm": 2.0604439934342973, + "learning_rate": 4.27549719393433e-07, + "loss": 0.8231, + "step": 9454 + }, + { + "epoch": 0.9094844170834937, + "grad_norm": 1.9958999834173647, + "learning_rate": 4.2664895953910434e-07, + "loss": 1.0141, + "step": 9455 + }, + { + "epoch": 0.9095806079261254, + "grad_norm": 1.9716476475046116, + "learning_rate": 4.2574912886092166e-07, + "loss": 0.908, + "step": 9456 + }, + { + "epoch": 0.9096767987687572, + "grad_norm": 1.8658655636352774, + "learning_rate": 4.2485022744622185e-07, + "loss": 0.7913, + "step": 9457 + }, + { + "epoch": 0.909772989611389, + "grad_norm": 2.015110127383828, + "learning_rate": 4.239522553822495e-07, + "loss": 0.8817, + "step": 9458 + }, + { + "epoch": 0.9098691804540208, + "grad_norm": 1.8706167317605664, + "learning_rate": 4.230552127561627e-07, + "loss": 0.8352, + "step": 9459 + }, + { + "epoch": 0.9099653712966526, + "grad_norm": 1.9821446807053775, + "learning_rate": 4.221590996550251e-07, + "loss": 0.895, + "step": 9460 + }, + { + "epoch": 0.9100615621392844, + "grad_norm": 2.266513407020008, + "learning_rate": 4.21263916165815e-07, + "loss": 0.7433, + "step": 9461 + }, + { + "epoch": 0.9101577529819161, + "grad_norm": 1.8252484184441087, + "learning_rate": 4.203696623754139e-07, + "loss": 0.8824, + "step": 9462 + }, + { + "epoch": 0.9102539438245479, + "grad_norm": 1.7077798491401903, + "learning_rate": 4.1947633837061794e-07, + "loss": 0.8237, + "step": 9463 + }, + { + "epoch": 0.9103501346671797, + "grad_norm": 2.061073668408351, + "learning_rate": 4.1858394423813563e-07, + "loss": 0.9005, + "step": 9464 + }, + { + "epoch": 0.9104463255098114, + "grad_norm": 2.221501188549029, + "learning_rate": 4.176924800645776e-07, + "loss": 0.9307, + "step": 9465 + }, + { + "epoch": 0.9105425163524432, + "grad_norm": 2.045647964021497, + "learning_rate": 4.1680194593646696e-07, + "loss": 0.9989, + "step": 9466 + }, + { + "epoch": 0.910638707195075, + "grad_norm": 2.237132182417587, + "learning_rate": 4.159123419402411e-07, + "loss": 0.9411, + "step": 9467 + }, + { + "epoch": 0.9107348980377068, + "grad_norm": 2.022605763673777, + "learning_rate": 4.1502366816224327e-07, + "loss": 0.8202, + "step": 9468 + }, + { + "epoch": 0.9108310888803386, + "grad_norm": 2.0163660179648124, + "learning_rate": 4.141359246887244e-07, + "loss": 0.9291, + "step": 9469 + }, + { + "epoch": 0.9109272797229704, + "grad_norm": 2.387719704926006, + "learning_rate": 4.1324911160585014e-07, + "loss": 0.9893, + "step": 9470 + }, + { + "epoch": 0.9110234705656022, + "grad_norm": 1.8561658312005598, + "learning_rate": 4.1236322899969484e-07, + "loss": 0.9137, + "step": 9471 + }, + { + "epoch": 0.911119661408234, + "grad_norm": 1.8863127043754695, + "learning_rate": 4.1147827695623643e-07, + "loss": 0.8223, + "step": 9472 + }, + { + "epoch": 0.9112158522508658, + "grad_norm": 2.1915396359921826, + "learning_rate": 4.1059425556136844e-07, + "loss": 0.9366, + "step": 9473 + }, + { + "epoch": 0.9113120430934974, + "grad_norm": 2.2207569069449415, + "learning_rate": 4.097111649008967e-07, + "loss": 0.9151, + "step": 9474 + }, + { + "epoch": 0.9114082339361292, + "grad_norm": 1.7879110259612045, + "learning_rate": 4.0882900506052705e-07, + "loss": 0.8735, + "step": 9475 + }, + { + "epoch": 0.911504424778761, + "grad_norm": 1.8630766336163187, + "learning_rate": 4.0794777612588543e-07, + "loss": 0.8474, + "step": 9476 + }, + { + "epoch": 0.9116006156213928, + "grad_norm": 1.9862410135345918, + "learning_rate": 4.0706747818249903e-07, + "loss": 0.8901, + "step": 9477 + }, + { + "epoch": 0.9116968064640246, + "grad_norm": 2.166563585402207, + "learning_rate": 4.061881113158117e-07, + "loss": 0.9025, + "step": 9478 + }, + { + "epoch": 0.9117929973066564, + "grad_norm": 2.2852258000119625, + "learning_rate": 4.0530967561117075e-07, + "loss": 0.9108, + "step": 9479 + }, + { + "epoch": 0.9118891881492882, + "grad_norm": 2.427365413731276, + "learning_rate": 4.044321711538368e-07, + "loss": 0.9484, + "step": 9480 + }, + { + "epoch": 0.91198537899192, + "grad_norm": 2.5127839535389085, + "learning_rate": 4.035555980289807e-07, + "loss": 0.9989, + "step": 9481 + }, + { + "epoch": 0.9120815698345518, + "grad_norm": 1.9360707855486885, + "learning_rate": 4.02679956321681e-07, + "loss": 0.9517, + "step": 9482 + }, + { + "epoch": 0.9121777606771835, + "grad_norm": 1.989088084822038, + "learning_rate": 4.018052461169231e-07, + "loss": 0.7967, + "step": 9483 + }, + { + "epoch": 0.9122739515198153, + "grad_norm": 2.3723072850133335, + "learning_rate": 4.00931467499609e-07, + "loss": 0.9166, + "step": 9484 + }, + { + "epoch": 0.9123701423624471, + "grad_norm": 1.7098019604891226, + "learning_rate": 4.0005862055454534e-07, + "loss": 0.9077, + "step": 9485 + }, + { + "epoch": 0.9124663332050789, + "grad_norm": 2.1373393438394173, + "learning_rate": 3.9918670536644776e-07, + "loss": 0.8901, + "step": 9486 + }, + { + "epoch": 0.9125625240477107, + "grad_norm": 1.9452049136092564, + "learning_rate": 3.9831572201994405e-07, + "loss": 0.99, + "step": 9487 + }, + { + "epoch": 0.9126587148903424, + "grad_norm": 2.0576587458425952, + "learning_rate": 3.974456705995733e-07, + "loss": 0.8319, + "step": 9488 + }, + { + "epoch": 0.9127549057329742, + "grad_norm": 1.8065203288495393, + "learning_rate": 3.965765511897779e-07, + "loss": 0.8413, + "step": 9489 + }, + { + "epoch": 0.912851096575606, + "grad_norm": 2.553895805478312, + "learning_rate": 3.9570836387491487e-07, + "loss": 0.8421, + "step": 9490 + }, + { + "epoch": 0.9129472874182378, + "grad_norm": 2.072356220411211, + "learning_rate": 3.9484110873924896e-07, + "loss": 0.9966, + "step": 9491 + }, + { + "epoch": 0.9130434782608695, + "grad_norm": 2.1200104724905997, + "learning_rate": 3.9397478586695513e-07, + "loss": 0.9289, + "step": 9492 + }, + { + "epoch": 0.9131396691035013, + "grad_norm": 1.8560515521039407, + "learning_rate": 3.931093953421172e-07, + "loss": 0.8582, + "step": 9493 + }, + { + "epoch": 0.9132358599461331, + "grad_norm": 1.9621381701251333, + "learning_rate": 3.9224493724872915e-07, + "loss": 1.0333, + "step": 9494 + }, + { + "epoch": 0.9133320507887649, + "grad_norm": 1.995066036220763, + "learning_rate": 3.9138141167069264e-07, + "loss": 0.7926, + "step": 9495 + }, + { + "epoch": 0.9134282416313967, + "grad_norm": 2.1622405985662296, + "learning_rate": 3.90518818691823e-07, + "loss": 0.9387, + "step": 9496 + }, + { + "epoch": 0.9135244324740285, + "grad_norm": 2.323334326300415, + "learning_rate": 3.8965715839583864e-07, + "loss": 0.9774, + "step": 9497 + }, + { + "epoch": 0.9136206233166603, + "grad_norm": 2.0768221945258, + "learning_rate": 3.8879643086637384e-07, + "loss": 0.9313, + "step": 9498 + }, + { + "epoch": 0.9137168141592921, + "grad_norm": 1.902616873665879, + "learning_rate": 3.8793663618697076e-07, + "loss": 0.9486, + "step": 9499 + }, + { + "epoch": 0.9138130050019239, + "grad_norm": 1.8407506742010864, + "learning_rate": 3.8707777444107697e-07, + "loss": 0.8871, + "step": 9500 + }, + { + "epoch": 0.9139091958445555, + "grad_norm": 1.9249543107743394, + "learning_rate": 3.862198457120525e-07, + "loss": 0.9281, + "step": 9501 + }, + { + "epoch": 0.9140053866871873, + "grad_norm": 1.8140254362077344, + "learning_rate": 3.8536285008316854e-07, + "loss": 0.8459, + "step": 9502 + }, + { + "epoch": 0.9141015775298191, + "grad_norm": 2.2011109429555256, + "learning_rate": 3.8450678763760187e-07, + "loss": 0.9201, + "step": 9503 + }, + { + "epoch": 0.9141977683724509, + "grad_norm": 2.1237652129733062, + "learning_rate": 3.8365165845844266e-07, + "loss": 0.964, + "step": 9504 + }, + { + "epoch": 0.9142939592150827, + "grad_norm": 2.0588053418149626, + "learning_rate": 3.82797462628689e-07, + "loss": 0.9034, + "step": 9505 + }, + { + "epoch": 0.9143901500577145, + "grad_norm": 1.9702796157428883, + "learning_rate": 3.819442002312457e-07, + "loss": 0.9002, + "step": 9506 + }, + { + "epoch": 0.9144863409003463, + "grad_norm": 2.1178312825100787, + "learning_rate": 3.8109187134892976e-07, + "loss": 0.9398, + "step": 9507 + }, + { + "epoch": 0.9145825317429781, + "grad_norm": 2.022022831961647, + "learning_rate": 3.8024047606446736e-07, + "loss": 0.8483, + "step": 9508 + }, + { + "epoch": 0.9146787225856099, + "grad_norm": 2.1136746213197513, + "learning_rate": 3.793900144604956e-07, + "loss": 0.8906, + "step": 9509 + }, + { + "epoch": 0.9147749134282416, + "grad_norm": 2.0764057537490928, + "learning_rate": 3.785404866195552e-07, + "loss": 0.9069, + "step": 9510 + }, + { + "epoch": 0.9148711042708734, + "grad_norm": 2.146685387332749, + "learning_rate": 3.7769189262410465e-07, + "loss": 0.9401, + "step": 9511 + }, + { + "epoch": 0.9149672951135052, + "grad_norm": 1.969111390124162, + "learning_rate": 3.768442325565036e-07, + "loss": 0.9256, + "step": 9512 + }, + { + "epoch": 0.915063485956137, + "grad_norm": 2.3327813750000965, + "learning_rate": 3.7599750649902733e-07, + "loss": 0.8528, + "step": 9513 + }, + { + "epoch": 0.9151596767987688, + "grad_norm": 2.1865884665770423, + "learning_rate": 3.751517145338546e-07, + "loss": 0.9504, + "step": 9514 + }, + { + "epoch": 0.9152558676414005, + "grad_norm": 1.9408392917059845, + "learning_rate": 3.7430685674307966e-07, + "loss": 0.8468, + "step": 9515 + }, + { + "epoch": 0.9153520584840323, + "grad_norm": 1.9067348676729412, + "learning_rate": 3.7346293320870363e-07, + "loss": 0.8831, + "step": 9516 + }, + { + "epoch": 0.9154482493266641, + "grad_norm": 2.20586243421729, + "learning_rate": 3.726199440126366e-07, + "loss": 0.8961, + "step": 9517 + }, + { + "epoch": 0.9155444401692959, + "grad_norm": 2.0957290321426836, + "learning_rate": 3.717778892366941e-07, + "loss": 0.8993, + "step": 9518 + }, + { + "epoch": 0.9156406310119277, + "grad_norm": 2.0239898598334856, + "learning_rate": 3.7093676896260756e-07, + "loss": 0.8695, + "step": 9519 + }, + { + "epoch": 0.9157368218545594, + "grad_norm": 2.2801162643495045, + "learning_rate": 3.700965832720171e-07, + "loss": 0.9166, + "step": 9520 + }, + { + "epoch": 0.9158330126971912, + "grad_norm": 1.8000042916968295, + "learning_rate": 3.6925733224646543e-07, + "loss": 0.9279, + "step": 9521 + }, + { + "epoch": 0.915929203539823, + "grad_norm": 1.8038391438725434, + "learning_rate": 3.684190159674117e-07, + "loss": 0.8778, + "step": 9522 + }, + { + "epoch": 0.9160253943824548, + "grad_norm": 1.8415288138221115, + "learning_rate": 3.675816345162242e-07, + "loss": 0.895, + "step": 9523 + }, + { + "epoch": 0.9161215852250866, + "grad_norm": 2.158619382182804, + "learning_rate": 3.6674518797417236e-07, + "loss": 0.919, + "step": 9524 + }, + { + "epoch": 0.9162177760677184, + "grad_norm": 2.2653685436725364, + "learning_rate": 3.6590967642244457e-07, + "loss": 0.954, + "step": 9525 + }, + { + "epoch": 0.9163139669103502, + "grad_norm": 1.6684132289885463, + "learning_rate": 3.6507509994213155e-07, + "loss": 0.7961, + "step": 9526 + }, + { + "epoch": 0.916410157752982, + "grad_norm": 2.005179020923869, + "learning_rate": 3.6424145861423953e-07, + "loss": 0.919, + "step": 9527 + }, + { + "epoch": 0.9165063485956138, + "grad_norm": 1.850316432642332, + "learning_rate": 3.6340875251967946e-07, + "loss": 0.972, + "step": 9528 + }, + { + "epoch": 0.9166025394382454, + "grad_norm": 1.9345607959146878, + "learning_rate": 3.6257698173927105e-07, + "loss": 0.8329, + "step": 9529 + }, + { + "epoch": 0.9166987302808772, + "grad_norm": 1.951806598577865, + "learning_rate": 3.617461463537464e-07, + "loss": 0.8017, + "step": 9530 + }, + { + "epoch": 0.916794921123509, + "grad_norm": 1.8845254632395656, + "learning_rate": 3.609162464437432e-07, + "loss": 0.9333, + "step": 9531 + }, + { + "epoch": 0.9168911119661408, + "grad_norm": 2.1794057631773245, + "learning_rate": 3.6008728208981157e-07, + "loss": 0.9248, + "step": 9532 + }, + { + "epoch": 0.9169873028087726, + "grad_norm": 1.8866118249939943, + "learning_rate": 3.5925925337241153e-07, + "loss": 0.8869, + "step": 9533 + }, + { + "epoch": 0.9170834936514044, + "grad_norm": 2.2120930197305873, + "learning_rate": 3.5843216037190873e-07, + "loss": 0.9436, + "step": 9534 + }, + { + "epoch": 0.9171796844940362, + "grad_norm": 2.4871587780421565, + "learning_rate": 3.576060031685791e-07, + "loss": 0.9025, + "step": 9535 + }, + { + "epoch": 0.917275875336668, + "grad_norm": 2.0130648139503515, + "learning_rate": 3.5678078184260834e-07, + "loss": 0.9042, + "step": 9536 + }, + { + "epoch": 0.9173720661792998, + "grad_norm": 2.146791841741117, + "learning_rate": 3.5595649647409357e-07, + "loss": 0.9058, + "step": 9537 + }, + { + "epoch": 0.9174682570219315, + "grad_norm": 2.020309936924498, + "learning_rate": 3.5513314714303524e-07, + "loss": 0.8472, + "step": 9538 + }, + { + "epoch": 0.9175644478645633, + "grad_norm": 2.161507631744035, + "learning_rate": 3.5431073392934836e-07, + "loss": 0.8842, + "step": 9539 + }, + { + "epoch": 0.917660638707195, + "grad_norm": 2.4174868608316475, + "learning_rate": 3.5348925691285675e-07, + "loss": 0.9113, + "step": 9540 + }, + { + "epoch": 0.9177568295498268, + "grad_norm": 2.0588423136765046, + "learning_rate": 3.5266871617329e-07, + "loss": 0.8914, + "step": 9541 + }, + { + "epoch": 0.9178530203924586, + "grad_norm": 1.9858559461794958, + "learning_rate": 3.518491117902878e-07, + "loss": 1.0049, + "step": 9542 + }, + { + "epoch": 0.9179492112350904, + "grad_norm": 1.8708335049718958, + "learning_rate": 3.5103044384340093e-07, + "loss": 0.8749, + "step": 9543 + }, + { + "epoch": 0.9180454020777222, + "grad_norm": 2.11184509660981, + "learning_rate": 3.502127124120891e-07, + "loss": 0.8262, + "step": 9544 + }, + { + "epoch": 0.918141592920354, + "grad_norm": 1.9309946255899437, + "learning_rate": 3.493959175757189e-07, + "loss": 0.8437, + "step": 9545 + }, + { + "epoch": 0.9182377837629858, + "grad_norm": 2.1009470227700042, + "learning_rate": 3.48580059413568e-07, + "loss": 0.8772, + "step": 9546 + }, + { + "epoch": 0.9183339746056175, + "grad_norm": 2.019578852916481, + "learning_rate": 3.477651380048208e-07, + "loss": 0.8664, + "step": 9547 + }, + { + "epoch": 0.9184301654482493, + "grad_norm": 2.3703490217707124, + "learning_rate": 3.4695115342857524e-07, + "loss": 0.9461, + "step": 9548 + }, + { + "epoch": 0.9185263562908811, + "grad_norm": 1.9058300064073939, + "learning_rate": 3.4613810576383246e-07, + "loss": 0.9148, + "step": 9549 + }, + { + "epoch": 0.9186225471335129, + "grad_norm": 1.9755969560190207, + "learning_rate": 3.4532599508950826e-07, + "loss": 0.9532, + "step": 9550 + }, + { + "epoch": 0.9187187379761447, + "grad_norm": 2.1870458360296134, + "learning_rate": 3.4451482148442515e-07, + "loss": 0.9339, + "step": 9551 + }, + { + "epoch": 0.9188149288187765, + "grad_norm": 2.011759280725956, + "learning_rate": 3.437045850273113e-07, + "loss": 0.7578, + "step": 9552 + }, + { + "epoch": 0.9189111196614083, + "grad_norm": 1.7058282981222408, + "learning_rate": 3.428952857968093e-07, + "loss": 0.751, + "step": 9553 + }, + { + "epoch": 0.91900731050404, + "grad_norm": 1.9300086483329, + "learning_rate": 3.420869238714708e-07, + "loss": 0.8069, + "step": 9554 + }, + { + "epoch": 0.9191035013466718, + "grad_norm": 1.9860095360744088, + "learning_rate": 3.412794993297497e-07, + "loss": 0.8678, + "step": 9555 + }, + { + "epoch": 0.9191996921893035, + "grad_norm": 1.774323887644503, + "learning_rate": 3.404730122500155e-07, + "loss": 0.868, + "step": 9556 + }, + { + "epoch": 0.9192958830319353, + "grad_norm": 1.84701937816644, + "learning_rate": 3.3966746271054675e-07, + "loss": 0.9061, + "step": 9557 + }, + { + "epoch": 0.9193920738745671, + "grad_norm": 2.1396481412916426, + "learning_rate": 3.3886285078952753e-07, + "loss": 0.8266, + "step": 9558 + }, + { + "epoch": 0.9194882647171989, + "grad_norm": 2.2745248172675745, + "learning_rate": 3.3805917656505094e-07, + "loss": 0.9054, + "step": 9559 + }, + { + "epoch": 0.9195844555598307, + "grad_norm": 1.9788880564463776, + "learning_rate": 3.3725644011512125e-07, + "loss": 0.9974, + "step": 9560 + }, + { + "epoch": 0.9196806464024625, + "grad_norm": 1.8475608414554947, + "learning_rate": 3.3645464151765283e-07, + "loss": 0.8897, + "step": 9561 + }, + { + "epoch": 0.9197768372450943, + "grad_norm": 1.9775809095259775, + "learning_rate": 3.356537808504634e-07, + "loss": 0.9009, + "step": 9562 + }, + { + "epoch": 0.9198730280877261, + "grad_norm": 2.151104994443056, + "learning_rate": 3.3485385819128744e-07, + "loss": 0.8416, + "step": 9563 + }, + { + "epoch": 0.9199692189303579, + "grad_norm": 1.9169534844811782, + "learning_rate": 3.3405487361776177e-07, + "loss": 0.8199, + "step": 9564 + }, + { + "epoch": 0.9200654097729896, + "grad_norm": 1.9511314139763574, + "learning_rate": 3.3325682720743655e-07, + "loss": 0.9026, + "step": 9565 + }, + { + "epoch": 0.9201616006156214, + "grad_norm": 2.295509262521845, + "learning_rate": 3.3245971903776654e-07, + "loss": 0.8237, + "step": 9566 + }, + { + "epoch": 0.9202577914582531, + "grad_norm": 1.99194437491437, + "learning_rate": 3.3166354918611866e-07, + "loss": 0.893, + "step": 9567 + }, + { + "epoch": 0.9203539823008849, + "grad_norm": 1.9884910514216583, + "learning_rate": 3.308683177297711e-07, + "loss": 0.8688, + "step": 9568 + }, + { + "epoch": 0.9204501731435167, + "grad_norm": 1.7786240494851877, + "learning_rate": 3.3007402474590444e-07, + "loss": 0.7607, + "step": 9569 + }, + { + "epoch": 0.9205463639861485, + "grad_norm": 1.9362356384341366, + "learning_rate": 3.292806703116125e-07, + "loss": 0.9254, + "step": 9570 + }, + { + "epoch": 0.9206425548287803, + "grad_norm": 1.8807855896687038, + "learning_rate": 3.28488254503897e-07, + "loss": 0.9109, + "step": 9571 + }, + { + "epoch": 0.9207387456714121, + "grad_norm": 2.044855424510194, + "learning_rate": 3.2769677739966975e-07, + "loss": 0.9491, + "step": 9572 + }, + { + "epoch": 0.9208349365140439, + "grad_norm": 2.1616133244767024, + "learning_rate": 3.269062390757494e-07, + "loss": 0.9284, + "step": 9573 + }, + { + "epoch": 0.9209311273566756, + "grad_norm": 1.908969605973677, + "learning_rate": 3.2611663960886665e-07, + "loss": 0.8245, + "step": 9574 + }, + { + "epoch": 0.9210273181993074, + "grad_norm": 1.8052076110389552, + "learning_rate": 3.253279790756547e-07, + "loss": 0.9391, + "step": 9575 + }, + { + "epoch": 0.9211235090419392, + "grad_norm": 2.1440277267851298, + "learning_rate": 3.245402575526646e-07, + "loss": 0.835, + "step": 9576 + }, + { + "epoch": 0.921219699884571, + "grad_norm": 2.1588460257681334, + "learning_rate": 3.2375347511634736e-07, + "loss": 0.9235, + "step": 9577 + }, + { + "epoch": 0.9213158907272028, + "grad_norm": 1.677912894765792, + "learning_rate": 3.2296763184306965e-07, + "loss": 0.7515, + "step": 9578 + }, + { + "epoch": 0.9214120815698346, + "grad_norm": 2.0530033099830804, + "learning_rate": 3.2218272780910385e-07, + "loss": 0.8536, + "step": 9579 + }, + { + "epoch": 0.9215082724124664, + "grad_norm": 2.169400210060168, + "learning_rate": 3.2139876309063233e-07, + "loss": 1.0509, + "step": 9580 + }, + { + "epoch": 0.9216044632550981, + "grad_norm": 2.1203721636730255, + "learning_rate": 3.20615737763742e-07, + "loss": 0.9044, + "step": 9581 + }, + { + "epoch": 0.9217006540977299, + "grad_norm": 1.835534154440209, + "learning_rate": 3.198336519044376e-07, + "loss": 0.924, + "step": 9582 + }, + { + "epoch": 0.9217968449403616, + "grad_norm": 2.316666732523616, + "learning_rate": 3.190525055886229e-07, + "loss": 0.8445, + "step": 9583 + }, + { + "epoch": 0.9218930357829934, + "grad_norm": 1.8501381341815166, + "learning_rate": 3.182722988921161e-07, + "loss": 0.9413, + "step": 9584 + }, + { + "epoch": 0.9219892266256252, + "grad_norm": 2.0072567866869355, + "learning_rate": 3.174930318906444e-07, + "loss": 0.9392, + "step": 9585 + }, + { + "epoch": 0.922085417468257, + "grad_norm": 1.8023564157235734, + "learning_rate": 3.167147046598418e-07, + "loss": 0.9218, + "step": 9586 + }, + { + "epoch": 0.9221816083108888, + "grad_norm": 2.0772474803504517, + "learning_rate": 3.1593731727525e-07, + "loss": 0.9045, + "step": 9587 + }, + { + "epoch": 0.9222777991535206, + "grad_norm": 2.345043382158773, + "learning_rate": 3.151608698123232e-07, + "loss": 0.8519, + "step": 9588 + }, + { + "epoch": 0.9223739899961524, + "grad_norm": 1.8308134107068457, + "learning_rate": 3.1438536234642214e-07, + "loss": 0.8858, + "step": 9589 + }, + { + "epoch": 0.9224701808387842, + "grad_norm": 1.907961132886943, + "learning_rate": 3.1361079495281443e-07, + "loss": 0.8628, + "step": 9590 + }, + { + "epoch": 0.922566371681416, + "grad_norm": 2.4827440645451047, + "learning_rate": 3.1283716770667994e-07, + "loss": 0.9628, + "step": 9591 + }, + { + "epoch": 0.9226625625240477, + "grad_norm": 2.0114806651259536, + "learning_rate": 3.1206448068310635e-07, + "loss": 0.9177, + "step": 9592 + }, + { + "epoch": 0.9227587533666795, + "grad_norm": 1.9438095341021873, + "learning_rate": 3.1129273395709036e-07, + "loss": 0.8091, + "step": 9593 + }, + { + "epoch": 0.9228549442093112, + "grad_norm": 2.1815743568085453, + "learning_rate": 3.1052192760353316e-07, + "loss": 0.8646, + "step": 9594 + }, + { + "epoch": 0.922951135051943, + "grad_norm": 2.118984751712038, + "learning_rate": 3.0975206169725046e-07, + "loss": 0.8921, + "step": 9595 + }, + { + "epoch": 0.9230473258945748, + "grad_norm": 2.1006405751150052, + "learning_rate": 3.0898313631296586e-07, + "loss": 0.7874, + "step": 9596 + }, + { + "epoch": 0.9231435167372066, + "grad_norm": 2.2401145859056726, + "learning_rate": 3.0821515152530736e-07, + "loss": 0.9398, + "step": 9597 + }, + { + "epoch": 0.9232397075798384, + "grad_norm": 2.0181649402078143, + "learning_rate": 3.0744810740881646e-07, + "loss": 0.9074, + "step": 9598 + }, + { + "epoch": 0.9233358984224702, + "grad_norm": 1.9112232566370226, + "learning_rate": 3.0668200403794036e-07, + "loss": 0.9877, + "step": 9599 + }, + { + "epoch": 0.923432089265102, + "grad_norm": 1.9464646729306376, + "learning_rate": 3.0591684148703617e-07, + "loss": 0.9677, + "step": 9600 + }, + { + "epoch": 0.9235282801077337, + "grad_norm": 2.0793130556827006, + "learning_rate": 3.0515261983037004e-07, + "loss": 0.8949, + "step": 9601 + }, + { + "epoch": 0.9236244709503655, + "grad_norm": 1.6567633227655594, + "learning_rate": 3.043893391421149e-07, + "loss": 0.6974, + "step": 9602 + }, + { + "epoch": 0.9237206617929973, + "grad_norm": 2.0011054207599823, + "learning_rate": 3.03626999496357e-07, + "loss": 0.9206, + "step": 9603 + }, + { + "epoch": 0.9238168526356291, + "grad_norm": 1.9433574105131552, + "learning_rate": 3.0286560096708275e-07, + "loss": 0.8647, + "step": 9604 + }, + { + "epoch": 0.9239130434782609, + "grad_norm": 1.9905801687966913, + "learning_rate": 3.0210514362819633e-07, + "loss": 0.8296, + "step": 9605 + }, + { + "epoch": 0.9240092343208927, + "grad_norm": 2.0941174927286483, + "learning_rate": 3.013456275535054e-07, + "loss": 0.7361, + "step": 9606 + }, + { + "epoch": 0.9241054251635245, + "grad_norm": 1.9227348976832863, + "learning_rate": 3.005870528167276e-07, + "loss": 0.8604, + "step": 9607 + }, + { + "epoch": 0.9242016160061562, + "grad_norm": 1.9580745844011174, + "learning_rate": 2.998294194914897e-07, + "loss": 0.9259, + "step": 9608 + }, + { + "epoch": 0.924297806848788, + "grad_norm": 2.17885767312909, + "learning_rate": 2.9907272765132613e-07, + "loss": 0.8404, + "step": 9609 + }, + { + "epoch": 0.9243939976914197, + "grad_norm": 1.9591564612508705, + "learning_rate": 2.983169773696815e-07, + "loss": 0.888, + "step": 9610 + }, + { + "epoch": 0.9244901885340515, + "grad_norm": 2.117228527428923, + "learning_rate": 2.9756216871990484e-07, + "loss": 0.921, + "step": 9611 + }, + { + "epoch": 0.9245863793766833, + "grad_norm": 2.334162391936128, + "learning_rate": 2.968083017752599e-07, + "loss": 0.892, + "step": 9612 + }, + { + "epoch": 0.9246825702193151, + "grad_norm": 1.998946198377449, + "learning_rate": 2.960553766089147e-07, + "loss": 0.9785, + "step": 9613 + }, + { + "epoch": 0.9247787610619469, + "grad_norm": 2.2313824811097063, + "learning_rate": 2.953033932939464e-07, + "loss": 0.845, + "step": 9614 + }, + { + "epoch": 0.9248749519045787, + "grad_norm": 2.0504008077137224, + "learning_rate": 2.945523519033433e-07, + "loss": 0.9296, + "step": 9615 + }, + { + "epoch": 0.9249711427472105, + "grad_norm": 1.7176569731194318, + "learning_rate": 2.938022525099982e-07, + "loss": 0.835, + "step": 9616 + }, + { + "epoch": 0.9250673335898423, + "grad_norm": 2.483581393611196, + "learning_rate": 2.930530951867172e-07, + "loss": 0.9235, + "step": 9617 + }, + { + "epoch": 0.9251635244324741, + "grad_norm": 2.04782008411944, + "learning_rate": 2.9230488000621003e-07, + "loss": 0.8303, + "step": 9618 + }, + { + "epoch": 0.9252597152751059, + "grad_norm": 2.306926584732548, + "learning_rate": 2.9155760704109857e-07, + "loss": 0.9075, + "step": 9619 + }, + { + "epoch": 0.9253559061177375, + "grad_norm": 2.096980715570762, + "learning_rate": 2.908112763639137e-07, + "loss": 1.0072, + "step": 9620 + }, + { + "epoch": 0.9254520969603693, + "grad_norm": 1.9035820933337637, + "learning_rate": 2.900658880470908e-07, + "loss": 0.8858, + "step": 9621 + }, + { + "epoch": 0.9255482878030011, + "grad_norm": 2.225693080989866, + "learning_rate": 2.8932144216297643e-07, + "loss": 0.954, + "step": 9622 + }, + { + "epoch": 0.9256444786456329, + "grad_norm": 1.901647962906975, + "learning_rate": 2.885779387838261e-07, + "loss": 0.8917, + "step": 9623 + }, + { + "epoch": 0.9257406694882647, + "grad_norm": 2.267269297615002, + "learning_rate": 2.878353779818044e-07, + "loss": 0.9403, + "step": 9624 + }, + { + "epoch": 0.9258368603308965, + "grad_norm": 1.9462090569120212, + "learning_rate": 2.8709375982898023e-07, + "loss": 0.8695, + "step": 9625 + }, + { + "epoch": 0.9259330511735283, + "grad_norm": 1.9985332177509387, + "learning_rate": 2.863530843973372e-07, + "loss": 0.8689, + "step": 9626 + }, + { + "epoch": 0.9260292420161601, + "grad_norm": 2.198125916028095, + "learning_rate": 2.856133517587623e-07, + "loss": 0.8486, + "step": 9627 + }, + { + "epoch": 0.9261254328587919, + "grad_norm": 1.4794602545193134, + "learning_rate": 2.848745619850546e-07, + "loss": 0.7276, + "step": 9628 + }, + { + "epoch": 0.9262216237014236, + "grad_norm": 2.0800179164490444, + "learning_rate": 2.8413671514791683e-07, + "loss": 0.7964, + "step": 9629 + }, + { + "epoch": 0.9263178145440554, + "grad_norm": 2.198213189283268, + "learning_rate": 2.833998113189662e-07, + "loss": 1.0077, + "step": 9630 + }, + { + "epoch": 0.9264140053866872, + "grad_norm": 1.909801527036412, + "learning_rate": 2.8266385056972543e-07, + "loss": 0.8181, + "step": 9631 + }, + { + "epoch": 0.926510196229319, + "grad_norm": 1.8475418860483872, + "learning_rate": 2.8192883297162634e-07, + "loss": 0.8713, + "step": 9632 + }, + { + "epoch": 0.9266063870719508, + "grad_norm": 2.2754053241837897, + "learning_rate": 2.811947585960062e-07, + "loss": 0.8068, + "step": 9633 + }, + { + "epoch": 0.9267025779145825, + "grad_norm": 1.9120168866916856, + "learning_rate": 2.804616275141148e-07, + "loss": 0.9226, + "step": 9634 + }, + { + "epoch": 0.9267987687572143, + "grad_norm": 2.379167461950954, + "learning_rate": 2.797294397971084e-07, + "loss": 0.8875, + "step": 9635 + }, + { + "epoch": 0.9268949595998461, + "grad_norm": 2.108911891300356, + "learning_rate": 2.7899819551605256e-07, + "loss": 0.7035, + "step": 9636 + }, + { + "epoch": 0.9269911504424779, + "grad_norm": 2.151856906185293, + "learning_rate": 2.782678947419215e-07, + "loss": 0.9571, + "step": 9637 + }, + { + "epoch": 0.9270873412851096, + "grad_norm": 2.006776914692016, + "learning_rate": 2.7753853754559634e-07, + "loss": 0.8537, + "step": 9638 + }, + { + "epoch": 0.9271835321277414, + "grad_norm": 1.6236312751270663, + "learning_rate": 2.768101239978671e-07, + "loss": 0.7782, + "step": 9639 + }, + { + "epoch": 0.9272797229703732, + "grad_norm": 2.033432683941518, + "learning_rate": 2.760826541694328e-07, + "loss": 0.9631, + "step": 9640 + }, + { + "epoch": 0.927375913813005, + "grad_norm": 1.7805620099010493, + "learning_rate": 2.753561281309014e-07, + "loss": 0.7446, + "step": 9641 + }, + { + "epoch": 0.9274721046556368, + "grad_norm": 2.139832569150566, + "learning_rate": 2.746305459527876e-07, + "loss": 0.9009, + "step": 9642 + }, + { + "epoch": 0.9275682954982686, + "grad_norm": 1.8939657560494945, + "learning_rate": 2.7390590770551505e-07, + "loss": 0.9484, + "step": 9643 + }, + { + "epoch": 0.9276644863409004, + "grad_norm": 2.11510111421425, + "learning_rate": 2.7318221345941865e-07, + "loss": 0.9316, + "step": 9644 + }, + { + "epoch": 0.9277606771835322, + "grad_norm": 1.5374890785195126, + "learning_rate": 2.724594632847377e-07, + "loss": 0.7392, + "step": 9645 + }, + { + "epoch": 0.927856868026164, + "grad_norm": 2.0810904604912768, + "learning_rate": 2.717376572516184e-07, + "loss": 0.9565, + "step": 9646 + }, + { + "epoch": 0.9279530588687956, + "grad_norm": 1.9299038099680381, + "learning_rate": 2.710167954301224e-07, + "loss": 0.7971, + "step": 9647 + }, + { + "epoch": 0.9280492497114274, + "grad_norm": 1.8605406948138454, + "learning_rate": 2.7029687789021377e-07, + "loss": 0.8346, + "step": 9648 + }, + { + "epoch": 0.9281454405540592, + "grad_norm": 2.0779947065093913, + "learning_rate": 2.695779047017677e-07, + "loss": 0.865, + "step": 9649 + }, + { + "epoch": 0.928241631396691, + "grad_norm": 1.8028996235493133, + "learning_rate": 2.688598759345651e-07, + "loss": 0.951, + "step": 9650 + }, + { + "epoch": 0.9283378222393228, + "grad_norm": 1.962341244381338, + "learning_rate": 2.6814279165829794e-07, + "loss": 0.7852, + "step": 9651 + }, + { + "epoch": 0.9284340130819546, + "grad_norm": 2.2812906753970377, + "learning_rate": 2.67426651942565e-07, + "loss": 0.8246, + "step": 9652 + }, + { + "epoch": 0.9285302039245864, + "grad_norm": 2.0034346782225247, + "learning_rate": 2.6671145685687407e-07, + "loss": 1.0396, + "step": 9653 + }, + { + "epoch": 0.9286263947672182, + "grad_norm": 1.9796090183849169, + "learning_rate": 2.659972064706406e-07, + "loss": 0.9231, + "step": 9654 + }, + { + "epoch": 0.92872258560985, + "grad_norm": 1.8167963506998122, + "learning_rate": 2.6528390085319047e-07, + "loss": 0.8361, + "step": 9655 + }, + { + "epoch": 0.9288187764524817, + "grad_norm": 2.142243327864443, + "learning_rate": 2.645715400737536e-07, + "loss": 0.9, + "step": 9656 + }, + { + "epoch": 0.9289149672951135, + "grad_norm": 2.177847967844901, + "learning_rate": 2.6386012420147156e-07, + "loss": 0.8912, + "step": 9657 + }, + { + "epoch": 0.9290111581377453, + "grad_norm": 1.9228845356945632, + "learning_rate": 2.631496533053934e-07, + "loss": 0.906, + "step": 9658 + }, + { + "epoch": 0.9291073489803771, + "grad_norm": 1.915437053349209, + "learning_rate": 2.6244012745447743e-07, + "loss": 1.0012, + "step": 9659 + }, + { + "epoch": 0.9292035398230089, + "grad_norm": 2.316647320976354, + "learning_rate": 2.6173154671758847e-07, + "loss": 0.857, + "step": 9660 + }, + { + "epoch": 0.9292997306656406, + "grad_norm": 2.0244365249176175, + "learning_rate": 2.6102391116350043e-07, + "loss": 0.8283, + "step": 9661 + }, + { + "epoch": 0.9293959215082724, + "grad_norm": 2.1330700690409756, + "learning_rate": 2.603172208608962e-07, + "loss": 0.9617, + "step": 9662 + }, + { + "epoch": 0.9294921123509042, + "grad_norm": 1.940980234411344, + "learning_rate": 2.596114758783641e-07, + "loss": 0.9731, + "step": 9663 + }, + { + "epoch": 0.929588303193536, + "grad_norm": 1.8613919658945444, + "learning_rate": 2.589066762844039e-07, + "loss": 0.9387, + "step": 9664 + }, + { + "epoch": 0.9296844940361677, + "grad_norm": 1.7138229347801797, + "learning_rate": 2.582028221474242e-07, + "loss": 0.7371, + "step": 9665 + }, + { + "epoch": 0.9297806848787995, + "grad_norm": 2.273020753917363, + "learning_rate": 2.57499913535737e-07, + "loss": 0.9577, + "step": 9666 + }, + { + "epoch": 0.9298768757214313, + "grad_norm": 2.29373024144504, + "learning_rate": 2.5679795051756773e-07, + "loss": 0.9301, + "step": 9667 + }, + { + "epoch": 0.9299730665640631, + "grad_norm": 2.0855103233116434, + "learning_rate": 2.5609693316104745e-07, + "loss": 0.8925, + "step": 9668 + }, + { + "epoch": 0.9300692574066949, + "grad_norm": 2.2568274532743198, + "learning_rate": 2.553968615342162e-07, + "loss": 0.875, + "step": 9669 + }, + { + "epoch": 0.9301654482493267, + "grad_norm": 1.973953752060842, + "learning_rate": 2.5469773570502063e-07, + "loss": 0.9186, + "step": 9670 + }, + { + "epoch": 0.9302616390919585, + "grad_norm": 2.2375936293383796, + "learning_rate": 2.5399955574131884e-07, + "loss": 0.9495, + "step": 9671 + }, + { + "epoch": 0.9303578299345903, + "grad_norm": 2.001141020645328, + "learning_rate": 2.5330232171087433e-07, + "loss": 0.851, + "step": 9672 + }, + { + "epoch": 0.9304540207772221, + "grad_norm": 1.9593974550201712, + "learning_rate": 2.526060336813596e-07, + "loss": 0.915, + "step": 9673 + }, + { + "epoch": 0.9305502116198537, + "grad_norm": 2.069315628996111, + "learning_rate": 2.51910691720354e-07, + "loss": 0.9132, + "step": 9674 + }, + { + "epoch": 0.9306464024624855, + "grad_norm": 2.095417874881965, + "learning_rate": 2.5121629589534903e-07, + "loss": 1.0051, + "step": 9675 + }, + { + "epoch": 0.9307425933051173, + "grad_norm": 2.2797685678484707, + "learning_rate": 2.5052284627374077e-07, + "loss": 0.9477, + "step": 9676 + }, + { + "epoch": 0.9308387841477491, + "grad_norm": 2.058747456120158, + "learning_rate": 2.4983034292283327e-07, + "loss": 0.9971, + "step": 9677 + }, + { + "epoch": 0.9309349749903809, + "grad_norm": 2.175979988000839, + "learning_rate": 2.491387859098426e-07, + "loss": 0.9144, + "step": 9678 + }, + { + "epoch": 0.9310311658330127, + "grad_norm": 2.0334967241017634, + "learning_rate": 2.484481753018875e-07, + "loss": 0.9407, + "step": 9679 + }, + { + "epoch": 0.9311273566756445, + "grad_norm": 1.9116663798413058, + "learning_rate": 2.477585111659997e-07, + "loss": 0.9159, + "step": 9680 + }, + { + "epoch": 0.9312235475182763, + "grad_norm": 2.1276670009066434, + "learning_rate": 2.4706979356911576e-07, + "loss": 0.9401, + "step": 9681 + }, + { + "epoch": 0.9313197383609081, + "grad_norm": 1.9337206574325105, + "learning_rate": 2.463820225780811e-07, + "loss": 0.8754, + "step": 9682 + }, + { + "epoch": 0.9314159292035398, + "grad_norm": 1.8612823720034695, + "learning_rate": 2.456951982596534e-07, + "loss": 0.822, + "step": 9683 + }, + { + "epoch": 0.9315121200461716, + "grad_norm": 2.3791981270645857, + "learning_rate": 2.4500932068049046e-07, + "loss": 0.8799, + "step": 9684 + }, + { + "epoch": 0.9316083108888034, + "grad_norm": 2.2454029065833443, + "learning_rate": 2.4432438990716455e-07, + "loss": 0.9487, + "step": 9685 + }, + { + "epoch": 0.9317045017314352, + "grad_norm": 2.066570295880389, + "learning_rate": 2.4364040600615477e-07, + "loss": 0.868, + "step": 9686 + }, + { + "epoch": 0.931800692574067, + "grad_norm": 1.9920024531449896, + "learning_rate": 2.429573690438458e-07, + "loss": 0.9399, + "step": 9687 + }, + { + "epoch": 0.9318968834166987, + "grad_norm": 1.8221371682317518, + "learning_rate": 2.422752790865346e-07, + "loss": 0.9001, + "step": 9688 + }, + { + "epoch": 0.9319930742593305, + "grad_norm": 1.906286596178336, + "learning_rate": 2.415941362004215e-07, + "loss": 0.9186, + "step": 9689 + }, + { + "epoch": 0.9320892651019623, + "grad_norm": 2.0666033976794695, + "learning_rate": 2.409139404516203e-07, + "loss": 0.9125, + "step": 9690 + }, + { + "epoch": 0.9321854559445941, + "grad_norm": 2.435950366720347, + "learning_rate": 2.40234691906146e-07, + "loss": 0.9139, + "step": 9691 + }, + { + "epoch": 0.9322816467872258, + "grad_norm": 2.1784440900803865, + "learning_rate": 2.3955639062992696e-07, + "loss": 0.9381, + "step": 9692 + }, + { + "epoch": 0.9323778376298576, + "grad_norm": 2.013276203973527, + "learning_rate": 2.3887903668880055e-07, + "loss": 0.9332, + "step": 9693 + }, + { + "epoch": 0.9324740284724894, + "grad_norm": 1.8966208843111034, + "learning_rate": 2.3820263014850741e-07, + "loss": 0.9112, + "step": 9694 + }, + { + "epoch": 0.9325702193151212, + "grad_norm": 1.9825864859523434, + "learning_rate": 2.3752717107469737e-07, + "loss": 0.8629, + "step": 9695 + }, + { + "epoch": 0.932666410157753, + "grad_norm": 2.1587237585860946, + "learning_rate": 2.3685265953293345e-07, + "loss": 0.843, + "step": 9696 + }, + { + "epoch": 0.9327626010003848, + "grad_norm": 1.7574486271422132, + "learning_rate": 2.3617909558868114e-07, + "loss": 0.9085, + "step": 9697 + }, + { + "epoch": 0.9328587918430166, + "grad_norm": 2.2816245797830557, + "learning_rate": 2.3550647930731362e-07, + "loss": 0.8958, + "step": 9698 + }, + { + "epoch": 0.9329549826856484, + "grad_norm": 1.8750679377148156, + "learning_rate": 2.3483481075411542e-07, + "loss": 0.9178, + "step": 9699 + }, + { + "epoch": 0.9330511735282802, + "grad_norm": 1.9927753658901324, + "learning_rate": 2.3416408999427876e-07, + "loss": 0.8734, + "step": 9700 + }, + { + "epoch": 0.9331473643709118, + "grad_norm": 1.8136520868465913, + "learning_rate": 2.334943170929027e-07, + "loss": 0.8784, + "step": 9701 + }, + { + "epoch": 0.9332435552135436, + "grad_norm": 1.9858986558191436, + "learning_rate": 2.3282549211499307e-07, + "loss": 0.9138, + "step": 9702 + }, + { + "epoch": 0.9333397460561754, + "grad_norm": 1.9161579073293662, + "learning_rate": 2.3215761512546452e-07, + "loss": 0.7893, + "step": 9703 + }, + { + "epoch": 0.9334359368988072, + "grad_norm": 2.2716484938678687, + "learning_rate": 2.3149068618914417e-07, + "loss": 0.8333, + "step": 9704 + }, + { + "epoch": 0.933532127741439, + "grad_norm": 1.8704743171778413, + "learning_rate": 2.3082470537075796e-07, + "loss": 0.9408, + "step": 9705 + }, + { + "epoch": 0.9336283185840708, + "grad_norm": 2.4305826174187324, + "learning_rate": 2.3015967273494867e-07, + "loss": 1.0106, + "step": 9706 + }, + { + "epoch": 0.9337245094267026, + "grad_norm": 2.152979804931901, + "learning_rate": 2.2949558834626352e-07, + "loss": 0.9478, + "step": 9707 + }, + { + "epoch": 0.9338207002693344, + "grad_norm": 1.6173790336327436, + "learning_rate": 2.2883245226915652e-07, + "loss": 0.8259, + "step": 9708 + }, + { + "epoch": 0.9339168911119662, + "grad_norm": 1.9090579315076728, + "learning_rate": 2.2817026456799064e-07, + "loss": 0.9462, + "step": 9709 + }, + { + "epoch": 0.9340130819545979, + "grad_norm": 2.116443241750745, + "learning_rate": 2.2750902530703667e-07, + "loss": 0.9049, + "step": 9710 + }, + { + "epoch": 0.9341092727972297, + "grad_norm": 2.1248326003456497, + "learning_rate": 2.2684873455047441e-07, + "loss": 0.8976, + "step": 9711 + }, + { + "epoch": 0.9342054636398615, + "grad_norm": 2.0167347220859657, + "learning_rate": 2.2618939236238924e-07, + "loss": 0.9786, + "step": 9712 + }, + { + "epoch": 0.9343016544824932, + "grad_norm": 1.7846359563993772, + "learning_rate": 2.2553099880677887e-07, + "loss": 0.8507, + "step": 9713 + }, + { + "epoch": 0.934397845325125, + "grad_norm": 1.9317305057230294, + "learning_rate": 2.2487355394754328e-07, + "loss": 0.9806, + "step": 9714 + }, + { + "epoch": 0.9344940361677568, + "grad_norm": 2.2475017499274332, + "learning_rate": 2.2421705784849368e-07, + "loss": 0.8365, + "step": 9715 + }, + { + "epoch": 0.9345902270103886, + "grad_norm": 1.9672142288642929, + "learning_rate": 2.2356151057334908e-07, + "loss": 0.9411, + "step": 9716 + }, + { + "epoch": 0.9346864178530204, + "grad_norm": 2.073640389668777, + "learning_rate": 2.2290691218573746e-07, + "loss": 0.9096, + "step": 9717 + }, + { + "epoch": 0.9347826086956522, + "grad_norm": 1.990921404633412, + "learning_rate": 2.2225326274919135e-07, + "loss": 0.8615, + "step": 9718 + }, + { + "epoch": 0.9348787995382839, + "grad_norm": 1.86368172240635, + "learning_rate": 2.2160056232715332e-07, + "loss": 0.8751, + "step": 9719 + }, + { + "epoch": 0.9349749903809157, + "grad_norm": 2.2175520702326548, + "learning_rate": 2.209488109829727e-07, + "loss": 0.931, + "step": 9720 + }, + { + "epoch": 0.9350711812235475, + "grad_norm": 2.107375012372047, + "learning_rate": 2.2029800877991115e-07, + "loss": 0.9725, + "step": 9721 + }, + { + "epoch": 0.9351673720661793, + "grad_norm": 1.7618433806682576, + "learning_rate": 2.196481557811303e-07, + "loss": 0.8129, + "step": 9722 + }, + { + "epoch": 0.9352635629088111, + "grad_norm": 1.8867444727258773, + "learning_rate": 2.189992520497053e-07, + "loss": 0.8867, + "step": 9723 + }, + { + "epoch": 0.9353597537514429, + "grad_norm": 1.9248812645978224, + "learning_rate": 2.1835129764861907e-07, + "loss": 0.8538, + "step": 9724 + }, + { + "epoch": 0.9354559445940747, + "grad_norm": 1.9291167103234772, + "learning_rate": 2.1770429264076132e-07, + "loss": 0.8952, + "step": 9725 + }, + { + "epoch": 0.9355521354367065, + "grad_norm": 2.185998575565238, + "learning_rate": 2.1705823708892737e-07, + "loss": 0.9001, + "step": 9726 + }, + { + "epoch": 0.9356483262793382, + "grad_norm": 1.9586400056278916, + "learning_rate": 2.1641313105582373e-07, + "loss": 0.8258, + "step": 9727 + }, + { + "epoch": 0.93574451712197, + "grad_norm": 2.1424872216914093, + "learning_rate": 2.1576897460406477e-07, + "loss": 1.0146, + "step": 9728 + }, + { + "epoch": 0.9358407079646017, + "grad_norm": 2.234460065606518, + "learning_rate": 2.1512576779616933e-07, + "loss": 0.978, + "step": 9729 + }, + { + "epoch": 0.9359368988072335, + "grad_norm": 2.1758051970903955, + "learning_rate": 2.144835106945664e-07, + "loss": 0.8494, + "step": 9730 + }, + { + "epoch": 0.9360330896498653, + "grad_norm": 1.9438649217373307, + "learning_rate": 2.138422033615939e-07, + "loss": 0.909, + "step": 9731 + }, + { + "epoch": 0.9361292804924971, + "grad_norm": 2.2154600889803633, + "learning_rate": 2.1320184585949532e-07, + "loss": 0.9703, + "step": 9732 + }, + { + "epoch": 0.9362254713351289, + "grad_norm": 1.8676189916787882, + "learning_rate": 2.125624382504221e-07, + "loss": 0.8967, + "step": 9733 + }, + { + "epoch": 0.9363216621777607, + "grad_norm": 2.201454171733821, + "learning_rate": 2.119239805964357e-07, + "loss": 0.9495, + "step": 9734 + }, + { + "epoch": 0.9364178530203925, + "grad_norm": 2.0946283370474155, + "learning_rate": 2.1128647295950322e-07, + "loss": 0.9247, + "step": 9735 + }, + { + "epoch": 0.9365140438630243, + "grad_norm": 2.0656157602256853, + "learning_rate": 2.106499154015018e-07, + "loss": 0.8564, + "step": 9736 + }, + { + "epoch": 0.9366102347056561, + "grad_norm": 1.8404461292493703, + "learning_rate": 2.1001430798421207e-07, + "loss": 0.8493, + "step": 9737 + }, + { + "epoch": 0.9367064255482878, + "grad_norm": 2.1590579698922916, + "learning_rate": 2.0937965076932576e-07, + "loss": 0.995, + "step": 9738 + }, + { + "epoch": 0.9368026163909196, + "grad_norm": 1.9563973225826468, + "learning_rate": 2.087459438184447e-07, + "loss": 0.9439, + "step": 9739 + }, + { + "epoch": 0.9368988072335513, + "grad_norm": 1.929742790272997, + "learning_rate": 2.0811318719307194e-07, + "loss": 0.8969, + "step": 9740 + }, + { + "epoch": 0.9369949980761831, + "grad_norm": 1.8595050105267583, + "learning_rate": 2.074813809546239e-07, + "loss": 0.9025, + "step": 9741 + }, + { + "epoch": 0.9370911889188149, + "grad_norm": 2.1049987209202454, + "learning_rate": 2.0685052516442373e-07, + "loss": 0.9043, + "step": 9742 + }, + { + "epoch": 0.9371873797614467, + "grad_norm": 1.984387646815511, + "learning_rate": 2.0622061988369912e-07, + "loss": 0.9011, + "step": 9743 + }, + { + "epoch": 0.9372835706040785, + "grad_norm": 1.8750016874775666, + "learning_rate": 2.0559166517358787e-07, + "loss": 0.9083, + "step": 9744 + }, + { + "epoch": 0.9373797614467103, + "grad_norm": 2.1874039178298186, + "learning_rate": 2.0496366109513776e-07, + "loss": 0.8275, + "step": 9745 + }, + { + "epoch": 0.9374759522893421, + "grad_norm": 2.0634997751495194, + "learning_rate": 2.0433660770930009e-07, + "loss": 0.8633, + "step": 9746 + }, + { + "epoch": 0.9375721431319738, + "grad_norm": 2.2221009417579403, + "learning_rate": 2.0371050507693613e-07, + "loss": 0.882, + "step": 9747 + }, + { + "epoch": 0.9376683339746056, + "grad_norm": 1.9454833338103488, + "learning_rate": 2.0308535325881616e-07, + "loss": 0.8164, + "step": 9748 + }, + { + "epoch": 0.9377645248172374, + "grad_norm": 2.0411936153861223, + "learning_rate": 2.0246115231561392e-07, + "loss": 0.8541, + "step": 9749 + }, + { + "epoch": 0.9378607156598692, + "grad_norm": 2.1499052583581135, + "learning_rate": 2.0183790230791532e-07, + "loss": 0.9253, + "step": 9750 + }, + { + "epoch": 0.937956906502501, + "grad_norm": 2.141383120356753, + "learning_rate": 2.012156032962098e-07, + "loss": 1.0035, + "step": 9751 + }, + { + "epoch": 0.9380530973451328, + "grad_norm": 2.101236948284896, + "learning_rate": 2.0059425534090128e-07, + "loss": 0.9672, + "step": 9752 + }, + { + "epoch": 0.9381492881877646, + "grad_norm": 2.190258692063918, + "learning_rate": 1.9997385850229255e-07, + "loss": 0.8954, + "step": 9753 + }, + { + "epoch": 0.9382454790303963, + "grad_norm": 2.1590388227819104, + "learning_rate": 1.9935441284059998e-07, + "loss": 0.9634, + "step": 9754 + }, + { + "epoch": 0.9383416698730281, + "grad_norm": 2.19105079854763, + "learning_rate": 1.9873591841594652e-07, + "loss": 0.8826, + "step": 9755 + }, + { + "epoch": 0.9384378607156598, + "grad_norm": 1.912465922886687, + "learning_rate": 1.981183752883631e-07, + "loss": 0.8511, + "step": 9756 + }, + { + "epoch": 0.9385340515582916, + "grad_norm": 1.935533988645202, + "learning_rate": 1.9750178351778504e-07, + "loss": 0.9395, + "step": 9757 + }, + { + "epoch": 0.9386302424009234, + "grad_norm": 2.2672586007661355, + "learning_rate": 1.9688614316406006e-07, + "loss": 0.8026, + "step": 9758 + }, + { + "epoch": 0.9387264332435552, + "grad_norm": 1.9423097377945717, + "learning_rate": 1.962714542869415e-07, + "loss": 0.9292, + "step": 9759 + }, + { + "epoch": 0.938822624086187, + "grad_norm": 1.8433893832524015, + "learning_rate": 1.9565771694608937e-07, + "loss": 0.8853, + "step": 9760 + }, + { + "epoch": 0.9389188149288188, + "grad_norm": 2.051723473016321, + "learning_rate": 1.950449312010716e-07, + "loss": 0.9147, + "step": 9761 + }, + { + "epoch": 0.9390150057714506, + "grad_norm": 2.3001572595241297, + "learning_rate": 1.9443309711136393e-07, + "loss": 0.79, + "step": 9762 + }, + { + "epoch": 0.9391111966140824, + "grad_norm": 2.0847657034362466, + "learning_rate": 1.9382221473635333e-07, + "loss": 0.7428, + "step": 9763 + }, + { + "epoch": 0.9392073874567142, + "grad_norm": 1.8920734567612978, + "learning_rate": 1.9321228413532788e-07, + "loss": 0.9127, + "step": 9764 + }, + { + "epoch": 0.9393035782993459, + "grad_norm": 2.083411028242113, + "learning_rate": 1.9260330536748807e-07, + "loss": 0.9039, + "step": 9765 + }, + { + "epoch": 0.9393997691419776, + "grad_norm": 1.7322232364792485, + "learning_rate": 1.9199527849194098e-07, + "loss": 0.7552, + "step": 9766 + }, + { + "epoch": 0.9394959599846094, + "grad_norm": 1.8281983647645088, + "learning_rate": 1.9138820356769838e-07, + "loss": 0.9154, + "step": 9767 + }, + { + "epoch": 0.9395921508272412, + "grad_norm": 2.166473752249959, + "learning_rate": 1.907820806536842e-07, + "loss": 0.9863, + "step": 9768 + }, + { + "epoch": 0.939688341669873, + "grad_norm": 2.199230153378922, + "learning_rate": 1.9017690980872804e-07, + "loss": 0.8974, + "step": 9769 + }, + { + "epoch": 0.9397845325125048, + "grad_norm": 2.0025867939944706, + "learning_rate": 1.895726910915663e-07, + "loss": 0.9037, + "step": 9770 + }, + { + "epoch": 0.9398807233551366, + "grad_norm": 2.2911211248620984, + "learning_rate": 1.889694245608442e-07, + "loss": 0.9823, + "step": 9771 + }, + { + "epoch": 0.9399769141977684, + "grad_norm": 2.176551760421022, + "learning_rate": 1.883671102751128e-07, + "loss": 0.908, + "step": 9772 + }, + { + "epoch": 0.9400731050404002, + "grad_norm": 2.2630858429108125, + "learning_rate": 1.87765748292833e-07, + "loss": 0.8522, + "step": 9773 + }, + { + "epoch": 0.9401692958830319, + "grad_norm": 2.007236829933018, + "learning_rate": 1.8716533867237153e-07, + "loss": 0.8014, + "step": 9774 + }, + { + "epoch": 0.9402654867256637, + "grad_norm": 2.1955758695579273, + "learning_rate": 1.8656588147200283e-07, + "loss": 0.8544, + "step": 9775 + }, + { + "epoch": 0.9403616775682955, + "grad_norm": 2.19960488023607, + "learning_rate": 1.859673767499115e-07, + "loss": 0.893, + "step": 9776 + }, + { + "epoch": 0.9404578684109273, + "grad_norm": 1.891571207770077, + "learning_rate": 1.853698245641855e-07, + "loss": 0.7602, + "step": 9777 + }, + { + "epoch": 0.9405540592535591, + "grad_norm": 1.9606610658615022, + "learning_rate": 1.847732249728218e-07, + "loss": 1.0058, + "step": 9778 + }, + { + "epoch": 0.9406502500961909, + "grad_norm": 2.087508037800622, + "learning_rate": 1.8417757803372737e-07, + "loss": 0.931, + "step": 9779 + }, + { + "epoch": 0.9407464409388226, + "grad_norm": 1.8171839666612453, + "learning_rate": 1.83582883804716e-07, + "loss": 0.8426, + "step": 9780 + }, + { + "epoch": 0.9408426317814544, + "grad_norm": 1.837454950233703, + "learning_rate": 1.8298914234350374e-07, + "loss": 0.8617, + "step": 9781 + }, + { + "epoch": 0.9409388226240862, + "grad_norm": 2.193356534011653, + "learning_rate": 1.8239635370772223e-07, + "loss": 0.7919, + "step": 9782 + }, + { + "epoch": 0.9410350134667179, + "grad_norm": 1.9660072672423519, + "learning_rate": 1.818045179549055e-07, + "loss": 0.9056, + "step": 9783 + }, + { + "epoch": 0.9411312043093497, + "grad_norm": 2.4089022229316743, + "learning_rate": 1.8121363514249534e-07, + "loss": 0.8103, + "step": 9784 + }, + { + "epoch": 0.9412273951519815, + "grad_norm": 1.9581151796609557, + "learning_rate": 1.8062370532784257e-07, + "loss": 0.7979, + "step": 9785 + }, + { + "epoch": 0.9413235859946133, + "grad_norm": 1.8476132419692304, + "learning_rate": 1.8003472856820469e-07, + "loss": 0.894, + "step": 9786 + }, + { + "epoch": 0.9414197768372451, + "grad_norm": 2.015542627858561, + "learning_rate": 1.7944670492074823e-07, + "loss": 0.9926, + "step": 9787 + }, + { + "epoch": 0.9415159676798769, + "grad_norm": 2.088599114454988, + "learning_rate": 1.7885963444254528e-07, + "loss": 0.856, + "step": 9788 + }, + { + "epoch": 0.9416121585225087, + "grad_norm": 1.7722843787059124, + "learning_rate": 1.7827351719057473e-07, + "loss": 0.8173, + "step": 9789 + }, + { + "epoch": 0.9417083493651405, + "grad_norm": 2.1619131849783884, + "learning_rate": 1.7768835322172552e-07, + "loss": 0.8322, + "step": 9790 + }, + { + "epoch": 0.9418045402077723, + "grad_norm": 2.208861355725129, + "learning_rate": 1.771041425927944e-07, + "loss": 0.8963, + "step": 9791 + }, + { + "epoch": 0.941900731050404, + "grad_norm": 1.9121311363427964, + "learning_rate": 1.7652088536048052e-07, + "loss": 0.9419, + "step": 9792 + }, + { + "epoch": 0.9419969218930357, + "grad_norm": 1.9943318675404853, + "learning_rate": 1.759385815813963e-07, + "loss": 0.8674, + "step": 9793 + }, + { + "epoch": 0.9420931127356675, + "grad_norm": 1.9724077087689589, + "learning_rate": 1.7535723131206106e-07, + "loss": 0.8529, + "step": 9794 + }, + { + "epoch": 0.9421893035782993, + "grad_norm": 1.8348907168756667, + "learning_rate": 1.7477683460889626e-07, + "loss": 0.8933, + "step": 9795 + }, + { + "epoch": 0.9422854944209311, + "grad_norm": 2.1509132607757224, + "learning_rate": 1.7419739152823468e-07, + "loss": 0.7957, + "step": 9796 + }, + { + "epoch": 0.9423816852635629, + "grad_norm": 1.9617660088535613, + "learning_rate": 1.7361890212631903e-07, + "loss": 0.9234, + "step": 9797 + }, + { + "epoch": 0.9424778761061947, + "grad_norm": 1.5311462488845098, + "learning_rate": 1.7304136645929448e-07, + "loss": 0.685, + "step": 9798 + }, + { + "epoch": 0.9425740669488265, + "grad_norm": 1.9457691892571671, + "learning_rate": 1.7246478458321724e-07, + "loss": 0.8448, + "step": 9799 + }, + { + "epoch": 0.9426702577914583, + "grad_norm": 1.977519376964785, + "learning_rate": 1.7188915655404814e-07, + "loss": 0.8619, + "step": 9800 + }, + { + "epoch": 0.94276644863409, + "grad_norm": 1.830598231610248, + "learning_rate": 1.7131448242765804e-07, + "loss": 0.8803, + "step": 9801 + }, + { + "epoch": 0.9428626394767218, + "grad_norm": 2.084923168709575, + "learning_rate": 1.707407622598223e-07, + "loss": 0.9633, + "step": 9802 + }, + { + "epoch": 0.9429588303193536, + "grad_norm": 1.9368141406319217, + "learning_rate": 1.7016799610622635e-07, + "loss": 0.8556, + "step": 9803 + }, + { + "epoch": 0.9430550211619854, + "grad_norm": 2.1584776472846783, + "learning_rate": 1.695961840224636e-07, + "loss": 0.898, + "step": 9804 + }, + { + "epoch": 0.9431512120046172, + "grad_norm": 2.070387496536384, + "learning_rate": 1.6902532606403176e-07, + "loss": 0.9567, + "step": 9805 + }, + { + "epoch": 0.943247402847249, + "grad_norm": 2.0089768194081796, + "learning_rate": 1.6845542228633772e-07, + "loss": 0.8213, + "step": 9806 + }, + { + "epoch": 0.9433435936898807, + "grad_norm": 1.8949494690188102, + "learning_rate": 1.6788647274469495e-07, + "loss": 0.8249, + "step": 9807 + }, + { + "epoch": 0.9434397845325125, + "grad_norm": 1.9008598730294919, + "learning_rate": 1.6731847749432705e-07, + "loss": 0.8072, + "step": 9808 + }, + { + "epoch": 0.9435359753751443, + "grad_norm": 2.064063814715632, + "learning_rate": 1.6675143659035997e-07, + "loss": 0.9844, + "step": 9809 + }, + { + "epoch": 0.943632166217776, + "grad_norm": 2.129876704850234, + "learning_rate": 1.6618535008783075e-07, + "loss": 0.9707, + "step": 9810 + }, + { + "epoch": 0.9437283570604078, + "grad_norm": 1.9177619470469365, + "learning_rate": 1.6562021804168548e-07, + "loss": 0.8996, + "step": 9811 + }, + { + "epoch": 0.9438245479030396, + "grad_norm": 2.098177341561943, + "learning_rate": 1.6505604050677249e-07, + "loss": 0.9389, + "step": 9812 + }, + { + "epoch": 0.9439207387456714, + "grad_norm": 2.0126734365728303, + "learning_rate": 1.6449281753784907e-07, + "loss": 0.8502, + "step": 9813 + }, + { + "epoch": 0.9440169295883032, + "grad_norm": 1.9240970614425168, + "learning_rate": 1.6393054918958373e-07, + "loss": 0.9842, + "step": 9814 + }, + { + "epoch": 0.944113120430935, + "grad_norm": 2.0163977403281406, + "learning_rate": 1.6336923551654837e-07, + "loss": 0.9635, + "step": 9815 + }, + { + "epoch": 0.9442093112735668, + "grad_norm": 2.0428787198423364, + "learning_rate": 1.6280887657322276e-07, + "loss": 0.9218, + "step": 9816 + }, + { + "epoch": 0.9443055021161986, + "grad_norm": 1.7619157079786627, + "learning_rate": 1.622494724139956e-07, + "loss": 0.8177, + "step": 9817 + }, + { + "epoch": 0.9444016929588304, + "grad_norm": 1.92793157461577, + "learning_rate": 1.616910230931612e-07, + "loss": 0.8763, + "step": 9818 + }, + { + "epoch": 0.944497883801462, + "grad_norm": 2.0188859788124214, + "learning_rate": 1.611335286649207e-07, + "loss": 0.9388, + "step": 9819 + }, + { + "epoch": 0.9445940746440938, + "grad_norm": 1.763777402647252, + "learning_rate": 1.6057698918338526e-07, + "loss": 0.7543, + "step": 9820 + }, + { + "epoch": 0.9446902654867256, + "grad_norm": 1.80100039176446, + "learning_rate": 1.6002140470257165e-07, + "loss": 0.8416, + "step": 9821 + }, + { + "epoch": 0.9447864563293574, + "grad_norm": 1.736864737123953, + "learning_rate": 1.5946677527640563e-07, + "loss": 0.9119, + "step": 9822 + }, + { + "epoch": 0.9448826471719892, + "grad_norm": 2.2133078692055776, + "learning_rate": 1.5891310095871525e-07, + "loss": 0.8895, + "step": 9823 + }, + { + "epoch": 0.944978838014621, + "grad_norm": 1.9389428680826428, + "learning_rate": 1.5836038180324198e-07, + "loss": 0.8618, + "step": 9824 + }, + { + "epoch": 0.9450750288572528, + "grad_norm": 2.0269155084198593, + "learning_rate": 1.5780861786363177e-07, + "loss": 0.9254, + "step": 9825 + }, + { + "epoch": 0.9451712196998846, + "grad_norm": 2.0717659865385074, + "learning_rate": 1.5725780919343624e-07, + "loss": 0.9305, + "step": 9826 + }, + { + "epoch": 0.9452674105425164, + "grad_norm": 1.991242059506686, + "learning_rate": 1.567079558461171e-07, + "loss": 0.8717, + "step": 9827 + }, + { + "epoch": 0.9453636013851482, + "grad_norm": 1.9838093444632967, + "learning_rate": 1.561590578750438e-07, + "loss": 0.8361, + "step": 9828 + }, + { + "epoch": 0.9454597922277799, + "grad_norm": 2.237270040088478, + "learning_rate": 1.556111153334905e-07, + "loss": 0.7955, + "step": 9829 + }, + { + "epoch": 0.9455559830704117, + "grad_norm": 2.092057826915608, + "learning_rate": 1.55064128274639e-07, + "loss": 0.9272, + "step": 9830 + }, + { + "epoch": 0.9456521739130435, + "grad_norm": 2.007180087552734, + "learning_rate": 1.5451809675158026e-07, + "loss": 0.824, + "step": 9831 + }, + { + "epoch": 0.9457483647556753, + "grad_norm": 2.4954870296283382, + "learning_rate": 1.5397302081731069e-07, + "loss": 0.8598, + "step": 9832 + }, + { + "epoch": 0.945844555598307, + "grad_norm": 2.177327161424608, + "learning_rate": 1.5342890052473469e-07, + "loss": 0.9027, + "step": 9833 + }, + { + "epoch": 0.9459407464409388, + "grad_norm": 1.9545141106080874, + "learning_rate": 1.5288573592666445e-07, + "loss": 0.8775, + "step": 9834 + }, + { + "epoch": 0.9460369372835706, + "grad_norm": 1.9451443172744756, + "learning_rate": 1.523435270758178e-07, + "loss": 0.7904, + "step": 9835 + }, + { + "epoch": 0.9461331281262024, + "grad_norm": 1.961914255684508, + "learning_rate": 1.518022740248215e-07, + "loss": 1.0128, + "step": 9836 + }, + { + "epoch": 0.9462293189688342, + "grad_norm": 1.6337123500408015, + "learning_rate": 1.5126197682620802e-07, + "loss": 0.7056, + "step": 9837 + }, + { + "epoch": 0.9463255098114659, + "grad_norm": 1.4217872964128855, + "learning_rate": 1.5072263553241872e-07, + "loss": 0.7047, + "step": 9838 + }, + { + "epoch": 0.9464217006540977, + "grad_norm": 2.3093951897554104, + "learning_rate": 1.5018425019580175e-07, + "loss": 0.9656, + "step": 9839 + }, + { + "epoch": 0.9465178914967295, + "grad_norm": 1.973168397662792, + "learning_rate": 1.4964682086861082e-07, + "loss": 0.8891, + "step": 9840 + }, + { + "epoch": 0.9466140823393613, + "grad_norm": 2.235739401910948, + "learning_rate": 1.4911034760300758e-07, + "loss": 0.8559, + "step": 9841 + }, + { + "epoch": 0.9467102731819931, + "grad_norm": 2.1750262333209207, + "learning_rate": 1.4857483045106258e-07, + "loss": 0.9781, + "step": 9842 + }, + { + "epoch": 0.9468064640246249, + "grad_norm": 2.3187495143780983, + "learning_rate": 1.4804026946475315e-07, + "loss": 0.8477, + "step": 9843 + }, + { + "epoch": 0.9469026548672567, + "grad_norm": 1.929919939778247, + "learning_rate": 1.475066646959611e-07, + "loss": 0.9333, + "step": 9844 + }, + { + "epoch": 0.9469988457098885, + "grad_norm": 3.606760033019361, + "learning_rate": 1.4697401619647721e-07, + "loss": 0.8082, + "step": 9845 + }, + { + "epoch": 0.9470950365525203, + "grad_norm": 1.7714832814911, + "learning_rate": 1.4644232401800352e-07, + "loss": 0.8074, + "step": 9846 + }, + { + "epoch": 0.9471912273951519, + "grad_norm": 2.01185614774025, + "learning_rate": 1.4591158821214091e-07, + "loss": 0.9078, + "step": 9847 + }, + { + "epoch": 0.9472874182377837, + "grad_norm": 2.1474223343096464, + "learning_rate": 1.4538180883040264e-07, + "loss": 0.9587, + "step": 9848 + }, + { + "epoch": 0.9473836090804155, + "grad_norm": 2.1198559741984284, + "learning_rate": 1.4485298592421094e-07, + "loss": 0.8395, + "step": 9849 + }, + { + "epoch": 0.9474797999230473, + "grad_norm": 2.275675484954276, + "learning_rate": 1.4432511954488915e-07, + "loss": 0.8682, + "step": 9850 + }, + { + "epoch": 0.9475759907656791, + "grad_norm": 1.510884351012286, + "learning_rate": 1.437982097436741e-07, + "loss": 0.6916, + "step": 9851 + }, + { + "epoch": 0.9476721816083109, + "grad_norm": 1.651883913423519, + "learning_rate": 1.4327225657170485e-07, + "loss": 0.7344, + "step": 9852 + }, + { + "epoch": 0.9477683724509427, + "grad_norm": 1.9852483962820258, + "learning_rate": 1.427472600800306e-07, + "loss": 0.9506, + "step": 9853 + }, + { + "epoch": 0.9478645632935745, + "grad_norm": 1.9301167547509248, + "learning_rate": 1.4222322031960723e-07, + "loss": 0.9616, + "step": 9854 + }, + { + "epoch": 0.9479607541362063, + "grad_norm": 2.110976660012355, + "learning_rate": 1.4170013734129628e-07, + "loss": 0.9225, + "step": 9855 + }, + { + "epoch": 0.948056944978838, + "grad_norm": 1.8840229016285461, + "learning_rate": 1.411780111958694e-07, + "loss": 0.9356, + "step": 9856 + }, + { + "epoch": 0.9481531358214698, + "grad_norm": 2.2428072689515797, + "learning_rate": 1.4065684193400152e-07, + "loss": 0.7245, + "step": 9857 + }, + { + "epoch": 0.9482493266641016, + "grad_norm": 2.224376200337507, + "learning_rate": 1.4013662960627562e-07, + "loss": 0.8491, + "step": 9858 + }, + { + "epoch": 0.9483455175067333, + "grad_norm": 2.066955987727841, + "learning_rate": 1.3961737426318566e-07, + "loss": 0.9, + "step": 9859 + }, + { + "epoch": 0.9484417083493651, + "grad_norm": 2.0577698191545677, + "learning_rate": 1.3909907595512806e-07, + "loss": 0.9352, + "step": 9860 + }, + { + "epoch": 0.9485378991919969, + "grad_norm": 2.1150149923629247, + "learning_rate": 1.3858173473240922e-07, + "loss": 0.9606, + "step": 9861 + }, + { + "epoch": 0.9486340900346287, + "grad_norm": 2.2241985172224332, + "learning_rate": 1.3806535064524006e-07, + "loss": 0.9286, + "step": 9862 + }, + { + "epoch": 0.9487302808772605, + "grad_norm": 2.621301905485671, + "learning_rate": 1.3754992374374164e-07, + "loss": 0.8897, + "step": 9863 + }, + { + "epoch": 0.9488264717198923, + "grad_norm": 2.071964122838833, + "learning_rate": 1.3703545407793951e-07, + "loss": 0.9336, + "step": 9864 + }, + { + "epoch": 0.948922662562524, + "grad_norm": 2.142784215053192, + "learning_rate": 1.36521941697767e-07, + "loss": 0.8985, + "step": 9865 + }, + { + "epoch": 0.9490188534051558, + "grad_norm": 1.8802411881716965, + "learning_rate": 1.360093866530665e-07, + "loss": 0.9723, + "step": 9866 + }, + { + "epoch": 0.9491150442477876, + "grad_norm": 2.161566387554341, + "learning_rate": 1.354977889935849e-07, + "loss": 0.9261, + "step": 9867 + }, + { + "epoch": 0.9492112350904194, + "grad_norm": 1.8549685258774706, + "learning_rate": 1.34987148768978e-07, + "loss": 0.8995, + "step": 9868 + }, + { + "epoch": 0.9493074259330512, + "grad_norm": 2.1526004344606617, + "learning_rate": 1.3447746602880618e-07, + "loss": 0.8576, + "step": 9869 + }, + { + "epoch": 0.949403616775683, + "grad_norm": 2.0054663387711815, + "learning_rate": 1.3396874082253986e-07, + "loss": 0.9233, + "step": 9870 + }, + { + "epoch": 0.9494998076183148, + "grad_norm": 1.9091758859481982, + "learning_rate": 1.334609731995551e-07, + "loss": 0.8847, + "step": 9871 + }, + { + "epoch": 0.9495959984609466, + "grad_norm": 2.011295805454815, + "learning_rate": 1.3295416320913357e-07, + "loss": 0.834, + "step": 9872 + }, + { + "epoch": 0.9496921893035783, + "grad_norm": 2.0942903089986586, + "learning_rate": 1.3244831090046816e-07, + "loss": 0.9739, + "step": 9873 + }, + { + "epoch": 0.94978838014621, + "grad_norm": 1.8784278973694297, + "learning_rate": 1.3194341632265518e-07, + "loss": 0.7668, + "step": 9874 + }, + { + "epoch": 0.9498845709888418, + "grad_norm": 1.9586712715437948, + "learning_rate": 1.314394795246976e-07, + "loss": 0.8849, + "step": 9875 + }, + { + "epoch": 0.9499807618314736, + "grad_norm": 2.333027727656628, + "learning_rate": 1.3093650055550855e-07, + "loss": 0.9508, + "step": 9876 + }, + { + "epoch": 0.9500769526741054, + "grad_norm": 2.0840168103791537, + "learning_rate": 1.3043447946390563e-07, + "loss": 0.9505, + "step": 9877 + }, + { + "epoch": 0.9501731435167372, + "grad_norm": 2.4251762360352562, + "learning_rate": 1.2993341629861432e-07, + "loss": 0.8859, + "step": 9878 + }, + { + "epoch": 0.950269334359369, + "grad_norm": 1.9818806349487423, + "learning_rate": 1.2943331110826683e-07, + "loss": 0.7724, + "step": 9879 + }, + { + "epoch": 0.9503655252020008, + "grad_norm": 2.2768421363452056, + "learning_rate": 1.2893416394140323e-07, + "loss": 0.9662, + "step": 9880 + }, + { + "epoch": 0.9504617160446326, + "grad_norm": 1.722664993217765, + "learning_rate": 1.284359748464714e-07, + "loss": 0.7137, + "step": 9881 + }, + { + "epoch": 0.9505579068872644, + "grad_norm": 1.919515197286295, + "learning_rate": 1.279387438718216e-07, + "loss": 0.9095, + "step": 9882 + }, + { + "epoch": 0.9506540977298961, + "grad_norm": 2.352323701949408, + "learning_rate": 1.274424710657174e-07, + "loss": 0.9022, + "step": 9883 + }, + { + "epoch": 0.9507502885725279, + "grad_norm": 1.6670080778297798, + "learning_rate": 1.269471564763247e-07, + "loss": 0.675, + "step": 9884 + }, + { + "epoch": 0.9508464794151597, + "grad_norm": 1.9340746154848585, + "learning_rate": 1.2645280015171847e-07, + "loss": 0.8946, + "step": 9885 + }, + { + "epoch": 0.9509426702577914, + "grad_norm": 1.9215396487274343, + "learning_rate": 1.2595940213988024e-07, + "loss": 0.9289, + "step": 9886 + }, + { + "epoch": 0.9510388611004232, + "grad_norm": 2.4569696123244853, + "learning_rate": 1.2546696248869837e-07, + "loss": 0.9298, + "step": 9887 + }, + { + "epoch": 0.951135051943055, + "grad_norm": 1.9653704322871397, + "learning_rate": 1.2497548124597026e-07, + "loss": 0.9239, + "step": 9888 + }, + { + "epoch": 0.9512312427856868, + "grad_norm": 1.9294588874805405, + "learning_rate": 1.2448495845939435e-07, + "loss": 0.8294, + "step": 9889 + }, + { + "epoch": 0.9513274336283186, + "grad_norm": 1.815391895249397, + "learning_rate": 1.2399539417658368e-07, + "loss": 0.8484, + "step": 9890 + }, + { + "epoch": 0.9514236244709504, + "grad_norm": 2.041649885892513, + "learning_rate": 1.235067884450536e-07, + "loss": 0.9587, + "step": 9891 + }, + { + "epoch": 0.9515198153135821, + "grad_norm": 2.0449610216356855, + "learning_rate": 1.2301914131222726e-07, + "loss": 0.9811, + "step": 9892 + }, + { + "epoch": 0.9516160061562139, + "grad_norm": 2.019305516766255, + "learning_rate": 1.2253245282543459e-07, + "loss": 0.9514, + "step": 9893 + }, + { + "epoch": 0.9517121969988457, + "grad_norm": 2.3440356103922055, + "learning_rate": 1.2204672303191335e-07, + "loss": 0.9626, + "step": 9894 + }, + { + "epoch": 0.9518083878414775, + "grad_norm": 2.0319068200012906, + "learning_rate": 1.2156195197880917e-07, + "loss": 0.8269, + "step": 9895 + }, + { + "epoch": 0.9519045786841093, + "grad_norm": 2.0974252515129064, + "learning_rate": 1.2107813971317106e-07, + "loss": 0.8557, + "step": 9896 + }, + { + "epoch": 0.9520007695267411, + "grad_norm": 2.032180054045493, + "learning_rate": 1.2059528628195816e-07, + "loss": 0.9541, + "step": 9897 + }, + { + "epoch": 0.9520969603693729, + "grad_norm": 2.1859195036959713, + "learning_rate": 1.201133917320363e-07, + "loss": 0.9057, + "step": 9898 + }, + { + "epoch": 0.9521931512120047, + "grad_norm": 2.3754400324885085, + "learning_rate": 1.1963245611017472e-07, + "loss": 0.9946, + "step": 9899 + }, + { + "epoch": 0.9522893420546364, + "grad_norm": 1.6627410867131716, + "learning_rate": 1.1915247946305498e-07, + "loss": 0.7473, + "step": 9900 + }, + { + "epoch": 0.9523855328972681, + "grad_norm": 1.9821193453793156, + "learning_rate": 1.1867346183726202e-07, + "loss": 0.8294, + "step": 9901 + }, + { + "epoch": 0.9524817237398999, + "grad_norm": 1.6913263671665217, + "learning_rate": 1.1819540327929092e-07, + "loss": 0.7521, + "step": 9902 + }, + { + "epoch": 0.9525779145825317, + "grad_norm": 1.9124698064967112, + "learning_rate": 1.1771830383553783e-07, + "loss": 0.8609, + "step": 9903 + }, + { + "epoch": 0.9526741054251635, + "grad_norm": 1.9848868881359543, + "learning_rate": 1.1724216355231022e-07, + "loss": 0.9464, + "step": 9904 + }, + { + "epoch": 0.9527702962677953, + "grad_norm": 2.5102186201413477, + "learning_rate": 1.167669824758233e-07, + "loss": 0.8656, + "step": 9905 + }, + { + "epoch": 0.9528664871104271, + "grad_norm": 2.175328794795736, + "learning_rate": 1.1629276065219575e-07, + "loss": 0.8235, + "step": 9906 + }, + { + "epoch": 0.9529626779530589, + "grad_norm": 2.0862236970609866, + "learning_rate": 1.158194981274563e-07, + "loss": 0.7964, + "step": 9907 + }, + { + "epoch": 0.9530588687956907, + "grad_norm": 2.390021771064962, + "learning_rate": 1.1534719494753821e-07, + "loss": 0.9425, + "step": 9908 + }, + { + "epoch": 0.9531550596383225, + "grad_norm": 2.231311411481707, + "learning_rate": 1.1487585115828259e-07, + "loss": 0.8017, + "step": 9909 + }, + { + "epoch": 0.9532512504809542, + "grad_norm": 1.9463035310988237, + "learning_rate": 1.144054668054373e-07, + "loss": 0.9303, + "step": 9910 + }, + { + "epoch": 0.953347441323586, + "grad_norm": 1.9401010106227718, + "learning_rate": 1.1393604193465801e-07, + "loss": 0.859, + "step": 9911 + }, + { + "epoch": 0.9534436321662177, + "grad_norm": 1.9431934718310355, + "learning_rate": 1.1346757659150498e-07, + "loss": 0.853, + "step": 9912 + }, + { + "epoch": 0.9535398230088495, + "grad_norm": 1.9412837294458736, + "learning_rate": 1.1300007082144848e-07, + "loss": 0.9251, + "step": 9913 + }, + { + "epoch": 0.9536360138514813, + "grad_norm": 2.141288607133379, + "learning_rate": 1.1253352466986334e-07, + "loss": 0.8456, + "step": 9914 + }, + { + "epoch": 0.9537322046941131, + "grad_norm": 2.111719641291855, + "learning_rate": 1.120679381820322e-07, + "loss": 0.9215, + "step": 9915 + }, + { + "epoch": 0.9538283955367449, + "grad_norm": 1.906049751022184, + "learning_rate": 1.116033114031434e-07, + "loss": 0.8154, + "step": 9916 + }, + { + "epoch": 0.9539245863793767, + "grad_norm": 1.8611780879132154, + "learning_rate": 1.1113964437829306e-07, + "loss": 0.9067, + "step": 9917 + }, + { + "epoch": 0.9540207772220085, + "grad_norm": 1.9809047681604577, + "learning_rate": 1.1067693715248406e-07, + "loss": 0.8958, + "step": 9918 + }, + { + "epoch": 0.9541169680646402, + "grad_norm": 1.5914289696526467, + "learning_rate": 1.1021518977062717e-07, + "loss": 0.7395, + "step": 9919 + }, + { + "epoch": 0.954213158907272, + "grad_norm": 1.9903626118537132, + "learning_rate": 1.0975440227753764e-07, + "loss": 0.8536, + "step": 9920 + }, + { + "epoch": 0.9543093497499038, + "grad_norm": 2.1730953453412507, + "learning_rate": 1.0929457471793969e-07, + "loss": 0.8942, + "step": 9921 + }, + { + "epoch": 0.9544055405925356, + "grad_norm": 1.9490694092494398, + "learning_rate": 1.0883570713646318e-07, + "loss": 0.8419, + "step": 9922 + }, + { + "epoch": 0.9545017314351674, + "grad_norm": 1.825720344062207, + "learning_rate": 1.0837779957764582e-07, + "loss": 0.8857, + "step": 9923 + }, + { + "epoch": 0.9545979222777992, + "grad_norm": 1.89314482916416, + "learning_rate": 1.0792085208593095e-07, + "loss": 0.8779, + "step": 9924 + }, + { + "epoch": 0.954694113120431, + "grad_norm": 2.420870221480448, + "learning_rate": 1.0746486470566864e-07, + "loss": 0.9542, + "step": 9925 + }, + { + "epoch": 0.9547903039630627, + "grad_norm": 1.7834702845059942, + "learning_rate": 1.0700983748111792e-07, + "loss": 0.7942, + "step": 9926 + }, + { + "epoch": 0.9548864948056945, + "grad_norm": 1.9973024715560115, + "learning_rate": 1.0655577045644127e-07, + "loss": 0.7791, + "step": 9927 + }, + { + "epoch": 0.9549826856483262, + "grad_norm": 2.06193810667361, + "learning_rate": 1.061026636757101e-07, + "loss": 0.9871, + "step": 9928 + }, + { + "epoch": 0.955078876490958, + "grad_norm": 2.076546661211416, + "learning_rate": 1.0565051718290475e-07, + "loss": 1.0011, + "step": 9929 + }, + { + "epoch": 0.9551750673335898, + "grad_norm": 2.277139378984437, + "learning_rate": 1.0519933102190682e-07, + "loss": 0.9295, + "step": 9930 + }, + { + "epoch": 0.9552712581762216, + "grad_norm": 1.7685294640977727, + "learning_rate": 1.0474910523650905e-07, + "loss": 0.7148, + "step": 9931 + }, + { + "epoch": 0.9553674490188534, + "grad_norm": 2.087470342970776, + "learning_rate": 1.0429983987041092e-07, + "loss": 0.8562, + "step": 9932 + }, + { + "epoch": 0.9554636398614852, + "grad_norm": 2.0954591870963006, + "learning_rate": 1.0385153496721534e-07, + "loss": 0.9502, + "step": 9933 + }, + { + "epoch": 0.955559830704117, + "grad_norm": 2.1806150900102, + "learning_rate": 1.0340419057043527e-07, + "loss": 0.951, + "step": 9934 + }, + { + "epoch": 0.9556560215467488, + "grad_norm": 2.118925378957631, + "learning_rate": 1.0295780672348931e-07, + "loss": 0.7888, + "step": 9935 + }, + { + "epoch": 0.9557522123893806, + "grad_norm": 2.0814577978152995, + "learning_rate": 1.0251238346970393e-07, + "loss": 0.8811, + "step": 9936 + }, + { + "epoch": 0.9558484032320124, + "grad_norm": 2.1094065943988705, + "learning_rate": 1.0206792085230788e-07, + "loss": 0.8598, + "step": 9937 + }, + { + "epoch": 0.955944594074644, + "grad_norm": 2.142675137669361, + "learning_rate": 1.0162441891444441e-07, + "loss": 0.7916, + "step": 9938 + }, + { + "epoch": 0.9560407849172758, + "grad_norm": 1.917863027756942, + "learning_rate": 1.0118187769915688e-07, + "loss": 0.7624, + "step": 9939 + }, + { + "epoch": 0.9561369757599076, + "grad_norm": 2.2126994967413, + "learning_rate": 1.007402972493976e-07, + "loss": 0.9108, + "step": 9940 + }, + { + "epoch": 0.9562331666025394, + "grad_norm": 2.3184606711157993, + "learning_rate": 1.0029967760802562e-07, + "loss": 0.9329, + "step": 9941 + }, + { + "epoch": 0.9563293574451712, + "grad_norm": 1.9964122978144176, + "learning_rate": 9.986001881780783e-08, + "loss": 0.9139, + "step": 9942 + }, + { + "epoch": 0.956425548287803, + "grad_norm": 2.3266508561155095, + "learning_rate": 9.942132092141677e-08, + "loss": 0.9388, + "step": 9943 + }, + { + "epoch": 0.9565217391304348, + "grad_norm": 2.762410424233107, + "learning_rate": 9.898358396143171e-08, + "loss": 0.9673, + "step": 9944 + }, + { + "epoch": 0.9566179299730666, + "grad_norm": 1.9656778353298343, + "learning_rate": 9.854680798033755e-08, + "loss": 0.9625, + "step": 9945 + }, + { + "epoch": 0.9567141208156984, + "grad_norm": 1.7289945217497205, + "learning_rate": 9.811099302052928e-08, + "loss": 0.7722, + "step": 9946 + }, + { + "epoch": 0.9568103116583301, + "grad_norm": 2.133184189858799, + "learning_rate": 9.76761391243053e-08, + "loss": 0.9609, + "step": 9947 + }, + { + "epoch": 0.9569065025009619, + "grad_norm": 2.1971521985471036, + "learning_rate": 9.72422463338718e-08, + "loss": 0.8314, + "step": 9948 + }, + { + "epoch": 0.9570026933435937, + "grad_norm": 1.8301259841596274, + "learning_rate": 9.680931469134291e-08, + "loss": 0.8014, + "step": 9949 + }, + { + "epoch": 0.9570988841862255, + "grad_norm": 1.9024482611705549, + "learning_rate": 9.637734423873612e-08, + "loss": 0.9267, + "step": 9950 + }, + { + "epoch": 0.9571950750288573, + "grad_norm": 1.8401695244680205, + "learning_rate": 9.594633501798012e-08, + "loss": 0.8571, + "step": 9951 + }, + { + "epoch": 0.957291265871489, + "grad_norm": 2.146971806650546, + "learning_rate": 9.55162870709081e-08, + "loss": 0.9069, + "step": 9952 + }, + { + "epoch": 0.9573874567141208, + "grad_norm": 1.8291098454950974, + "learning_rate": 9.508720043925778e-08, + "loss": 0.8832, + "step": 9953 + }, + { + "epoch": 0.9574836475567526, + "grad_norm": 1.814327131083917, + "learning_rate": 9.465907516467698e-08, + "loss": 0.8076, + "step": 9954 + }, + { + "epoch": 0.9575798383993844, + "grad_norm": 2.2903203961255385, + "learning_rate": 9.423191128872022e-08, + "loss": 0.8588, + "step": 9955 + }, + { + "epoch": 0.9576760292420161, + "grad_norm": 1.8368698757402342, + "learning_rate": 9.380570885284546e-08, + "loss": 0.6998, + "step": 9956 + }, + { + "epoch": 0.9577722200846479, + "grad_norm": 1.5893013266880167, + "learning_rate": 9.33804678984207e-08, + "loss": 0.6777, + "step": 9957 + }, + { + "epoch": 0.9578684109272797, + "grad_norm": 2.506166872831163, + "learning_rate": 9.295618846671739e-08, + "loss": 0.9496, + "step": 9958 + }, + { + "epoch": 0.9579646017699115, + "grad_norm": 1.9698018550726129, + "learning_rate": 9.253287059891703e-08, + "loss": 0.9086, + "step": 9959 + }, + { + "epoch": 0.9580607926125433, + "grad_norm": 2.1614677767062007, + "learning_rate": 9.211051433610674e-08, + "loss": 0.9637, + "step": 9960 + }, + { + "epoch": 0.9581569834551751, + "grad_norm": 1.9146707966071972, + "learning_rate": 9.168911971927819e-08, + "loss": 0.9076, + "step": 9961 + }, + { + "epoch": 0.9582531742978069, + "grad_norm": 1.9763072762523821, + "learning_rate": 9.126868678933198e-08, + "loss": 0.7608, + "step": 9962 + }, + { + "epoch": 0.9583493651404387, + "grad_norm": 1.847174268801456, + "learning_rate": 9.084921558707549e-08, + "loss": 0.8428, + "step": 9963 + }, + { + "epoch": 0.9584455559830705, + "grad_norm": 2.0849151069681398, + "learning_rate": 9.04307061532217e-08, + "loss": 0.9483, + "step": 9964 + }, + { + "epoch": 0.9585417468257021, + "grad_norm": 1.838632318998594, + "learning_rate": 9.001315852839033e-08, + "loss": 0.883, + "step": 9965 + }, + { + "epoch": 0.9586379376683339, + "grad_norm": 2.2567480607080137, + "learning_rate": 8.959657275310674e-08, + "loss": 0.9168, + "step": 9966 + }, + { + "epoch": 0.9587341285109657, + "grad_norm": 2.125625687255212, + "learning_rate": 8.918094886780748e-08, + "loss": 0.7506, + "step": 9967 + }, + { + "epoch": 0.9588303193535975, + "grad_norm": 2.1003482421122937, + "learning_rate": 8.876628691282918e-08, + "loss": 0.9056, + "step": 9968 + }, + { + "epoch": 0.9589265101962293, + "grad_norm": 2.207538720033849, + "learning_rate": 8.835258692841964e-08, + "loss": 0.9765, + "step": 9969 + }, + { + "epoch": 0.9590227010388611, + "grad_norm": 2.035040025165713, + "learning_rate": 8.793984895473117e-08, + "loss": 0.8874, + "step": 9970 + }, + { + "epoch": 0.9591188918814929, + "grad_norm": 2.009700125645916, + "learning_rate": 8.75280730318251e-08, + "loss": 0.9385, + "step": 9971 + }, + { + "epoch": 0.9592150827241247, + "grad_norm": 1.7191655435709368, + "learning_rate": 8.711725919966718e-08, + "loss": 0.7999, + "step": 9972 + }, + { + "epoch": 0.9593112735667565, + "grad_norm": 2.29615368110596, + "learning_rate": 8.670740749813001e-08, + "loss": 0.8875, + "step": 9973 + }, + { + "epoch": 0.9594074644093882, + "grad_norm": 2.03583894310717, + "learning_rate": 8.629851796699284e-08, + "loss": 0.9703, + "step": 9974 + }, + { + "epoch": 0.95950365525202, + "grad_norm": 2.075109083779865, + "learning_rate": 8.589059064594285e-08, + "loss": 0.8359, + "step": 9975 + }, + { + "epoch": 0.9595998460946518, + "grad_norm": 1.829506423899601, + "learning_rate": 8.54836255745728e-08, + "loss": 0.8914, + "step": 9976 + }, + { + "epoch": 0.9596960369372836, + "grad_norm": 1.7576657018188913, + "learning_rate": 8.507762279238108e-08, + "loss": 0.8556, + "step": 9977 + }, + { + "epoch": 0.9597922277799154, + "grad_norm": 1.8335970966195656, + "learning_rate": 8.467258233877728e-08, + "loss": 0.7269, + "step": 9978 + }, + { + "epoch": 0.9598884186225471, + "grad_norm": 1.974657907155207, + "learning_rate": 8.426850425306887e-08, + "loss": 0.8863, + "step": 9979 + }, + { + "epoch": 0.9599846094651789, + "grad_norm": 2.0542572271145185, + "learning_rate": 8.386538857447779e-08, + "loss": 0.825, + "step": 9980 + }, + { + "epoch": 0.9600808003078107, + "grad_norm": 2.0336144845462343, + "learning_rate": 8.346323534213052e-08, + "loss": 0.9428, + "step": 9981 + }, + { + "epoch": 0.9601769911504425, + "grad_norm": 2.021526199400331, + "learning_rate": 8.306204459505807e-08, + "loss": 0.9567, + "step": 9982 + }, + { + "epoch": 0.9602731819930742, + "grad_norm": 2.289521777067588, + "learning_rate": 8.266181637220038e-08, + "loss": 0.9331, + "step": 9983 + }, + { + "epoch": 0.960369372835706, + "grad_norm": 1.7507039587618036, + "learning_rate": 8.226255071240308e-08, + "loss": 0.8452, + "step": 9984 + }, + { + "epoch": 0.9604655636783378, + "grad_norm": 1.7455459794419963, + "learning_rate": 8.186424765441847e-08, + "loss": 0.8439, + "step": 9985 + }, + { + "epoch": 0.9605617545209696, + "grad_norm": 2.289638702829941, + "learning_rate": 8.146690723690342e-08, + "loss": 0.8124, + "step": 9986 + }, + { + "epoch": 0.9606579453636014, + "grad_norm": 1.9958717126942431, + "learning_rate": 8.107052949842486e-08, + "loss": 0.9357, + "step": 9987 + }, + { + "epoch": 0.9607541362062332, + "grad_norm": 2.276958441582239, + "learning_rate": 8.067511447745535e-08, + "loss": 0.9326, + "step": 9988 + }, + { + "epoch": 0.960850327048865, + "grad_norm": 1.8801864281050047, + "learning_rate": 8.028066221237196e-08, + "loss": 0.9231, + "step": 9989 + }, + { + "epoch": 0.9609465178914968, + "grad_norm": 1.919202156488657, + "learning_rate": 7.988717274146074e-08, + "loss": 0.8235, + "step": 9990 + }, + { + "epoch": 0.9610427087341286, + "grad_norm": 4.165927814540542, + "learning_rate": 7.949464610291224e-08, + "loss": 0.8733, + "step": 9991 + }, + { + "epoch": 0.9611388995767602, + "grad_norm": 2.9241528590521413, + "learning_rate": 7.910308233482488e-08, + "loss": 0.8332, + "step": 9992 + }, + { + "epoch": 0.961235090419392, + "grad_norm": 1.8712359970836818, + "learning_rate": 7.871248147520272e-08, + "loss": 0.8808, + "step": 9993 + }, + { + "epoch": 0.9613312812620238, + "grad_norm": 1.737940494913272, + "learning_rate": 7.832284356195764e-08, + "loss": 0.7476, + "step": 9994 + }, + { + "epoch": 0.9614274721046556, + "grad_norm": 2.231370583036418, + "learning_rate": 7.793416863290826e-08, + "loss": 0.7943, + "step": 9995 + }, + { + "epoch": 0.9615236629472874, + "grad_norm": 2.136224043870013, + "learning_rate": 7.754645672577776e-08, + "loss": 0.9726, + "step": 9996 + }, + { + "epoch": 0.9616198537899192, + "grad_norm": 1.7542718231595806, + "learning_rate": 7.715970787819715e-08, + "loss": 0.791, + "step": 9997 + }, + { + "epoch": 0.961716044632551, + "grad_norm": 2.0395070846393564, + "learning_rate": 7.677392212770196e-08, + "loss": 0.9077, + "step": 9998 + }, + { + "epoch": 0.9618122354751828, + "grad_norm": 2.0332596270927707, + "learning_rate": 7.638909951174001e-08, + "loss": 0.8546, + "step": 9999 + }, + { + "epoch": 0.9619084263178146, + "grad_norm": 2.0788845188501814, + "learning_rate": 7.600524006765808e-08, + "loss": 0.8781, + "step": 10000 + }, + { + "epoch": 0.9620046171604463, + "grad_norm": 2.1707630285130692, + "learning_rate": 7.562234383271416e-08, + "loss": 0.9366, + "step": 10001 + }, + { + "epoch": 0.9621008080030781, + "grad_norm": 2.09774678158012, + "learning_rate": 7.524041084407185e-08, + "loss": 0.9516, + "step": 10002 + }, + { + "epoch": 0.9621969988457099, + "grad_norm": 2.0029219449396227, + "learning_rate": 7.485944113880151e-08, + "loss": 0.8114, + "step": 10003 + }, + { + "epoch": 0.9622931896883417, + "grad_norm": 2.020585417206158, + "learning_rate": 7.447943475387797e-08, + "loss": 0.9236, + "step": 10004 + }, + { + "epoch": 0.9623893805309734, + "grad_norm": 1.9594877371261012, + "learning_rate": 7.41003917261851e-08, + "loss": 0.915, + "step": 10005 + }, + { + "epoch": 0.9624855713736052, + "grad_norm": 2.2169411303698614, + "learning_rate": 7.372231209251346e-08, + "loss": 0.9026, + "step": 10006 + }, + { + "epoch": 0.962581762216237, + "grad_norm": 1.9554145929646527, + "learning_rate": 7.334519588955702e-08, + "loss": 0.8442, + "step": 10007 + }, + { + "epoch": 0.9626779530588688, + "grad_norm": 1.909549095735018, + "learning_rate": 7.296904315391873e-08, + "loss": 0.9328, + "step": 10008 + }, + { + "epoch": 0.9627741439015006, + "grad_norm": 1.987576040279448, + "learning_rate": 7.259385392210716e-08, + "loss": 0.9243, + "step": 10009 + }, + { + "epoch": 0.9628703347441323, + "grad_norm": 2.106246041938445, + "learning_rate": 7.221962823053874e-08, + "loss": 0.9468, + "step": 10010 + }, + { + "epoch": 0.9629665255867641, + "grad_norm": 2.210849938418939, + "learning_rate": 7.184636611553442e-08, + "loss": 0.829, + "step": 10011 + }, + { + "epoch": 0.9630627164293959, + "grad_norm": 2.171050924273552, + "learning_rate": 7.147406761332298e-08, + "loss": 0.9185, + "step": 10012 + }, + { + "epoch": 0.9631589072720277, + "grad_norm": 2.2373513919761927, + "learning_rate": 7.110273276003998e-08, + "loss": 0.8211, + "step": 10013 + }, + { + "epoch": 0.9632550981146595, + "grad_norm": 2.0046885708455546, + "learning_rate": 7.073236159172325e-08, + "loss": 0.9074, + "step": 10014 + }, + { + "epoch": 0.9633512889572913, + "grad_norm": 1.9964228749490365, + "learning_rate": 7.036295414432404e-08, + "loss": 0.8626, + "step": 10015 + }, + { + "epoch": 0.9634474797999231, + "grad_norm": 2.1289031142372874, + "learning_rate": 6.999451045369587e-08, + "loss": 1.0182, + "step": 10016 + }, + { + "epoch": 0.9635436706425549, + "grad_norm": 1.891412687429106, + "learning_rate": 6.962703055559905e-08, + "loss": 0.8931, + "step": 10017 + }, + { + "epoch": 0.9636398614851867, + "grad_norm": 2.449029698280583, + "learning_rate": 6.926051448569948e-08, + "loss": 0.9692, + "step": 10018 + }, + { + "epoch": 0.9637360523278183, + "grad_norm": 1.8308166398723003, + "learning_rate": 6.889496227957204e-08, + "loss": 0.7515, + "step": 10019 + }, + { + "epoch": 0.9638322431704501, + "grad_norm": 2.0624403037885495, + "learning_rate": 6.853037397269724e-08, + "loss": 0.9709, + "step": 10020 + }, + { + "epoch": 0.9639284340130819, + "grad_norm": 1.8308174667108164, + "learning_rate": 6.816674960045899e-08, + "loss": 0.803, + "step": 10021 + }, + { + "epoch": 0.9640246248557137, + "grad_norm": 1.99174656732368, + "learning_rate": 6.78040891981524e-08, + "loss": 0.8085, + "step": 10022 + }, + { + "epoch": 0.9641208156983455, + "grad_norm": 2.1073477214174923, + "learning_rate": 6.744239280097709e-08, + "loss": 0.7254, + "step": 10023 + }, + { + "epoch": 0.9642170065409773, + "grad_norm": 1.935153521444864, + "learning_rate": 6.70816604440383e-08, + "loss": 0.86, + "step": 10024 + }, + { + "epoch": 0.9643131973836091, + "grad_norm": 2.018920365323206, + "learning_rate": 6.672189216234693e-08, + "loss": 0.9557, + "step": 10025 + }, + { + "epoch": 0.9644093882262409, + "grad_norm": 2.0362047764114886, + "learning_rate": 6.63630879908217e-08, + "loss": 0.8519, + "step": 10026 + }, + { + "epoch": 0.9645055790688727, + "grad_norm": 2.2521899154090304, + "learning_rate": 6.600524796429031e-08, + "loss": 0.8648, + "step": 10027 + }, + { + "epoch": 0.9646017699115044, + "grad_norm": 1.7367424224158492, + "learning_rate": 6.564837211748054e-08, + "loss": 0.7809, + "step": 10028 + }, + { + "epoch": 0.9646979607541362, + "grad_norm": 2.303251703177181, + "learning_rate": 6.529246048503357e-08, + "loss": 0.9854, + "step": 10029 + }, + { + "epoch": 0.964794151596768, + "grad_norm": 1.6583036562868485, + "learning_rate": 6.493751310149177e-08, + "loss": 0.7662, + "step": 10030 + }, + { + "epoch": 0.9648903424393998, + "grad_norm": 1.9578805824143026, + "learning_rate": 6.458353000130646e-08, + "loss": 0.9385, + "step": 10031 + }, + { + "epoch": 0.9649865332820315, + "grad_norm": 2.062216234615, + "learning_rate": 6.42305112188335e-08, + "loss": 0.867, + "step": 10032 + }, + { + "epoch": 0.9650827241246633, + "grad_norm": 1.8991253955142404, + "learning_rate": 6.38784567883377e-08, + "loss": 0.9287, + "step": 10033 + }, + { + "epoch": 0.9651789149672951, + "grad_norm": 2.3297062693851336, + "learning_rate": 6.352736674398951e-08, + "loss": 0.9991, + "step": 10034 + }, + { + "epoch": 0.9652751058099269, + "grad_norm": 2.1725512421833373, + "learning_rate": 6.31772411198639e-08, + "loss": 0.8624, + "step": 10035 + }, + { + "epoch": 0.9653712966525587, + "grad_norm": 2.14781031672789, + "learning_rate": 6.282807994994477e-08, + "loss": 0.9333, + "step": 10036 + }, + { + "epoch": 0.9654674874951905, + "grad_norm": 2.020997068806861, + "learning_rate": 6.24798832681206e-08, + "loss": 0.8978, + "step": 10037 + }, + { + "epoch": 0.9655636783378222, + "grad_norm": 14.352221208981284, + "learning_rate": 6.213265110818656e-08, + "loss": 0.7911, + "step": 10038 + }, + { + "epoch": 0.965659869180454, + "grad_norm": 1.9809012706115015, + "learning_rate": 6.178638350384459e-08, + "loss": 0.7436, + "step": 10039 + }, + { + "epoch": 0.9657560600230858, + "grad_norm": 1.7176282992622873, + "learning_rate": 6.144108048870335e-08, + "loss": 0.8128, + "step": 10040 + }, + { + "epoch": 0.9658522508657176, + "grad_norm": 1.6278307471866305, + "learning_rate": 6.109674209627714e-08, + "loss": 0.7601, + "step": 10041 + }, + { + "epoch": 0.9659484417083494, + "grad_norm": 2.111896783614166, + "learning_rate": 6.075336835998813e-08, + "loss": 0.9492, + "step": 10042 + }, + { + "epoch": 0.9660446325509812, + "grad_norm": 2.379323031607485, + "learning_rate": 6.041095931316188e-08, + "loss": 0.8032, + "step": 10043 + }, + { + "epoch": 0.966140823393613, + "grad_norm": 2.0695254027432246, + "learning_rate": 6.00695149890329e-08, + "loss": 0.9161, + "step": 10044 + }, + { + "epoch": 0.9662370142362448, + "grad_norm": 2.0177483446561544, + "learning_rate": 5.972903542074138e-08, + "loss": 0.8867, + "step": 10045 + }, + { + "epoch": 0.9663332050788765, + "grad_norm": 1.966469879306146, + "learning_rate": 5.938952064133419e-08, + "loss": 0.8356, + "step": 10046 + }, + { + "epoch": 0.9664293959215082, + "grad_norm": 1.9966611333837332, + "learning_rate": 5.9050970683762755e-08, + "loss": 0.8643, + "step": 10047 + }, + { + "epoch": 0.96652558676414, + "grad_norm": 1.8666916617044051, + "learning_rate": 5.871338558088857e-08, + "loss": 0.8398, + "step": 10048 + }, + { + "epoch": 0.9666217776067718, + "grad_norm": 2.1653044922918396, + "learning_rate": 5.837676536547543e-08, + "loss": 0.8924, + "step": 10049 + }, + { + "epoch": 0.9667179684494036, + "grad_norm": 2.167692153174894, + "learning_rate": 5.8041110070194976e-08, + "loss": 0.8871, + "step": 10050 + }, + { + "epoch": 0.9668141592920354, + "grad_norm": 1.99012710063018, + "learning_rate": 5.770641972762669e-08, + "loss": 1.0021, + "step": 10051 + }, + { + "epoch": 0.9669103501346672, + "grad_norm": 2.096166863714953, + "learning_rate": 5.7372694370254614e-08, + "loss": 0.8373, + "step": 10052 + }, + { + "epoch": 0.967006540977299, + "grad_norm": 2.0271979122550965, + "learning_rate": 5.703993403046948e-08, + "loss": 0.9557, + "step": 10053 + }, + { + "epoch": 0.9671027318199308, + "grad_norm": 1.9866122115298863, + "learning_rate": 5.67081387405688e-08, + "loss": 0.8611, + "step": 10054 + }, + { + "epoch": 0.9671989226625626, + "grad_norm": 1.9015728980271076, + "learning_rate": 5.6377308532755694e-08, + "loss": 0.9664, + "step": 10055 + }, + { + "epoch": 0.9672951135051943, + "grad_norm": 2.2427319808012056, + "learning_rate": 5.6047443439141146e-08, + "loss": 0.975, + "step": 10056 + }, + { + "epoch": 0.967391304347826, + "grad_norm": 2.0364546208367025, + "learning_rate": 5.571854349173955e-08, + "loss": 0.923, + "step": 10057 + }, + { + "epoch": 0.9674874951904578, + "grad_norm": 1.8136865049572621, + "learning_rate": 5.539060872247537e-08, + "loss": 0.8258, + "step": 10058 + }, + { + "epoch": 0.9675836860330896, + "grad_norm": 2.0326468171762975, + "learning_rate": 5.5063639163177585e-08, + "loss": 0.9317, + "step": 10059 + }, + { + "epoch": 0.9676798768757214, + "grad_norm": 2.167324550404644, + "learning_rate": 5.47376348455797e-08, + "loss": 0.9117, + "step": 10060 + }, + { + "epoch": 0.9677760677183532, + "grad_norm": 1.9233702541858926, + "learning_rate": 5.441259580132418e-08, + "loss": 0.8164, + "step": 10061 + }, + { + "epoch": 0.967872258560985, + "grad_norm": 1.9203696853533994, + "learning_rate": 5.408852206195914e-08, + "loss": 0.837, + "step": 10062 + }, + { + "epoch": 0.9679684494036168, + "grad_norm": 2.078261604281515, + "learning_rate": 5.376541365893717e-08, + "loss": 0.9291, + "step": 10063 + }, + { + "epoch": 0.9680646402462486, + "grad_norm": 1.9518905939575206, + "learning_rate": 5.344327062362098e-08, + "loss": 0.9333, + "step": 10064 + }, + { + "epoch": 0.9681608310888803, + "grad_norm": 1.9191350870337929, + "learning_rate": 5.312209298727666e-08, + "loss": 0.8228, + "step": 10065 + }, + { + "epoch": 0.9682570219315121, + "grad_norm": 2.1495525598910223, + "learning_rate": 5.2801880781075954e-08, + "loss": 0.8779, + "step": 10066 + }, + { + "epoch": 0.9683532127741439, + "grad_norm": 2.0280810958036057, + "learning_rate": 5.248263403609954e-08, + "loss": 0.8488, + "step": 10067 + }, + { + "epoch": 0.9684494036167757, + "grad_norm": 1.9427633914269062, + "learning_rate": 5.216435278333376e-08, + "loss": 0.885, + "step": 10068 + }, + { + "epoch": 0.9685455944594075, + "grad_norm": 1.8926115630174307, + "learning_rate": 5.1847037053668335e-08, + "loss": 0.8623, + "step": 10069 + }, + { + "epoch": 0.9686417853020393, + "grad_norm": 1.9532767286827597, + "learning_rate": 5.153068687790197e-08, + "loss": 0.9121, + "step": 10070 + }, + { + "epoch": 0.968737976144671, + "grad_norm": 2.2447335352410716, + "learning_rate": 5.121530228674121e-08, + "loss": 0.9087, + "step": 10071 + }, + { + "epoch": 0.9688341669873028, + "grad_norm": 1.900067726943489, + "learning_rate": 5.0900883310794903e-08, + "loss": 0.8989, + "step": 10072 + }, + { + "epoch": 0.9689303578299346, + "grad_norm": 2.074349494298663, + "learning_rate": 5.058742998058086e-08, + "loss": 0.9861, + "step": 10073 + }, + { + "epoch": 0.9690265486725663, + "grad_norm": 1.8760865200578902, + "learning_rate": 5.0274942326521414e-08, + "loss": 0.8178, + "step": 10074 + }, + { + "epoch": 0.9691227395151981, + "grad_norm": 2.083115589186306, + "learning_rate": 4.996342037894786e-08, + "loss": 0.9896, + "step": 10075 + }, + { + "epoch": 0.9692189303578299, + "grad_norm": 2.1987731871194027, + "learning_rate": 4.9652864168096e-08, + "loss": 0.937, + "step": 10076 + }, + { + "epoch": 0.9693151212004617, + "grad_norm": 1.8791684868040095, + "learning_rate": 4.934327372410619e-08, + "loss": 0.957, + "step": 10077 + }, + { + "epoch": 0.9694113120430935, + "grad_norm": 2.131107285744801, + "learning_rate": 4.9034649077027706e-08, + "loss": 0.8222, + "step": 10078 + }, + { + "epoch": 0.9695075028857253, + "grad_norm": 2.0945271500558564, + "learning_rate": 4.8726990256815486e-08, + "loss": 0.882, + "step": 10079 + }, + { + "epoch": 0.9696036937283571, + "grad_norm": 2.0813401858472713, + "learning_rate": 4.84202972933312e-08, + "loss": 0.8243, + "step": 10080 + }, + { + "epoch": 0.9696998845709889, + "grad_norm": 1.960064502098066, + "learning_rate": 4.811457021633992e-08, + "loss": 0.98, + "step": 10081 + }, + { + "epoch": 0.9697960754136207, + "grad_norm": 2.4507515117776637, + "learning_rate": 4.7809809055517906e-08, + "loss": 1.0243, + "step": 10082 + }, + { + "epoch": 0.9698922662562524, + "grad_norm": 1.600548090562407, + "learning_rate": 4.7506013840442623e-08, + "loss": 0.6954, + "step": 10083 + }, + { + "epoch": 0.9699884570988841, + "grad_norm": 2.2668153481739033, + "learning_rate": 4.720318460060047e-08, + "loss": 0.8867, + "step": 10084 + }, + { + "epoch": 0.9700846479415159, + "grad_norm": 2.1677988752620436, + "learning_rate": 4.69013213653835e-08, + "loss": 0.8663, + "step": 10085 + }, + { + "epoch": 0.9701808387841477, + "grad_norm": 2.0436050362796494, + "learning_rate": 4.6600424164091606e-08, + "loss": 0.901, + "step": 10086 + }, + { + "epoch": 0.9702770296267795, + "grad_norm": 2.3449669399979016, + "learning_rate": 4.6300493025928095e-08, + "loss": 0.7957, + "step": 10087 + }, + { + "epoch": 0.9703732204694113, + "grad_norm": 1.851456635885915, + "learning_rate": 4.6001527980004125e-08, + "loss": 0.933, + "step": 10088 + }, + { + "epoch": 0.9704694113120431, + "grad_norm": 2.014118353567169, + "learning_rate": 4.570352905533648e-08, + "loss": 0.8824, + "step": 10089 + }, + { + "epoch": 0.9705656021546749, + "grad_norm": 2.1736214776945446, + "learning_rate": 4.54064962808487e-08, + "loss": 0.8291, + "step": 10090 + }, + { + "epoch": 0.9706617929973067, + "grad_norm": 2.2322962578531134, + "learning_rate": 4.511042968537105e-08, + "loss": 1.0124, + "step": 10091 + }, + { + "epoch": 0.9707579838399384, + "grad_norm": 1.8359482953281616, + "learning_rate": 4.4815329297639434e-08, + "loss": 0.8656, + "step": 10092 + }, + { + "epoch": 0.9708541746825702, + "grad_norm": 1.9992613195095557, + "learning_rate": 4.4521195146294274e-08, + "loss": 0.7825, + "step": 10093 + }, + { + "epoch": 0.970950365525202, + "grad_norm": 2.430431760607962, + "learning_rate": 4.422802725988606e-08, + "loss": 0.9519, + "step": 10094 + }, + { + "epoch": 0.9710465563678338, + "grad_norm": 2.767040757778906, + "learning_rate": 4.3935825666866496e-08, + "loss": 0.9521, + "step": 10095 + }, + { + "epoch": 0.9711427472104656, + "grad_norm": 1.8976523560592555, + "learning_rate": 4.364459039559843e-08, + "loss": 0.9447, + "step": 10096 + }, + { + "epoch": 0.9712389380530974, + "grad_norm": 1.73755371650869, + "learning_rate": 4.335432147434926e-08, + "loss": 0.8835, + "step": 10097 + }, + { + "epoch": 0.9713351288957291, + "grad_norm": 1.944915948085115, + "learning_rate": 4.3065018931289784e-08, + "loss": 0.8845, + "step": 10098 + }, + { + "epoch": 0.9714313197383609, + "grad_norm": 2.1176426506984893, + "learning_rate": 4.277668279450087e-08, + "loss": 0.989, + "step": 10099 + }, + { + "epoch": 0.9715275105809927, + "grad_norm": 2.0770301347536364, + "learning_rate": 4.248931309196791e-08, + "loss": 0.9252, + "step": 10100 + }, + { + "epoch": 0.9716237014236244, + "grad_norm": 1.9649542943252876, + "learning_rate": 4.220290985158193e-08, + "loss": 0.976, + "step": 10101 + }, + { + "epoch": 0.9717198922662562, + "grad_norm": 2.120064658455997, + "learning_rate": 4.1917473101140696e-08, + "loss": 0.9571, + "step": 10102 + }, + { + "epoch": 0.971816083108888, + "grad_norm": 1.9148577083857858, + "learning_rate": 4.1633002868349815e-08, + "loss": 0.885, + "step": 10103 + }, + { + "epoch": 0.9719122739515198, + "grad_norm": 1.9190986869253701, + "learning_rate": 4.134949918081832e-08, + "loss": 0.9107, + "step": 10104 + }, + { + "epoch": 0.9720084647941516, + "grad_norm": 2.0108613283596553, + "learning_rate": 4.1066962066063085e-08, + "loss": 0.9882, + "step": 10105 + }, + { + "epoch": 0.9721046556367834, + "grad_norm": 1.8080026653421282, + "learning_rate": 4.0785391551506626e-08, + "loss": 0.8072, + "step": 10106 + }, + { + "epoch": 0.9722008464794152, + "grad_norm": 1.8104663137559063, + "learning_rate": 4.050478766447708e-08, + "loss": 0.9094, + "step": 10107 + }, + { + "epoch": 0.972297037322047, + "grad_norm": 2.202536874214832, + "learning_rate": 4.022515043221154e-08, + "loss": 0.8971, + "step": 10108 + }, + { + "epoch": 0.9723932281646788, + "grad_norm": 1.854871341261309, + "learning_rate": 3.99464798818483e-08, + "loss": 0.938, + "step": 10109 + }, + { + "epoch": 0.9724894190073105, + "grad_norm": 2.175227013943483, + "learning_rate": 3.966877604043795e-08, + "loss": 0.9111, + "step": 10110 + }, + { + "epoch": 0.9725856098499422, + "grad_norm": 2.1234067718935425, + "learning_rate": 3.939203893493226e-08, + "loss": 0.8891, + "step": 10111 + }, + { + "epoch": 0.972681800692574, + "grad_norm": 2.107393146172736, + "learning_rate": 3.9116268592189755e-08, + "loss": 0.9954, + "step": 10112 + }, + { + "epoch": 0.9727779915352058, + "grad_norm": 1.9996790119060726, + "learning_rate": 3.88414650389779e-08, + "loss": 0.8707, + "step": 10113 + }, + { + "epoch": 0.9728741823778376, + "grad_norm": 2.2282546287795117, + "learning_rate": 3.8567628301969806e-08, + "loss": 0.9173, + "step": 10114 + }, + { + "epoch": 0.9729703732204694, + "grad_norm": 2.2446480550671524, + "learning_rate": 3.8294758407740884e-08, + "loss": 0.907, + "step": 10115 + }, + { + "epoch": 0.9730665640631012, + "grad_norm": 2.206569773698811, + "learning_rate": 3.802285538277772e-08, + "loss": 0.8583, + "step": 10116 + }, + { + "epoch": 0.973162754905733, + "grad_norm": 2.0288679157833482, + "learning_rate": 3.7751919253471437e-08, + "loss": 0.905, + "step": 10117 + }, + { + "epoch": 0.9732589457483648, + "grad_norm": 2.058024070150561, + "learning_rate": 3.748195004611543e-08, + "loss": 0.9443, + "step": 10118 + }, + { + "epoch": 0.9733551365909965, + "grad_norm": 2.0648942721374297, + "learning_rate": 3.721294778691542e-08, + "loss": 0.9102, + "step": 10119 + }, + { + "epoch": 0.9734513274336283, + "grad_norm": 2.351386444172677, + "learning_rate": 3.69449125019794e-08, + "loss": 0.9553, + "step": 10120 + }, + { + "epoch": 0.9735475182762601, + "grad_norm": 1.5562784258310278, + "learning_rate": 3.667784421732212e-08, + "loss": 0.7281, + "step": 10121 + }, + { + "epoch": 0.9736437091188919, + "grad_norm": 1.8010023917087072, + "learning_rate": 3.6411742958866184e-08, + "loss": 0.9146, + "step": 10122 + }, + { + "epoch": 0.9737398999615237, + "grad_norm": 1.9465436170058943, + "learning_rate": 3.6146608752437585e-08, + "loss": 0.9033, + "step": 10123 + }, + { + "epoch": 0.9738360908041555, + "grad_norm": 2.0825184479719776, + "learning_rate": 3.588244162377019e-08, + "loss": 0.8714, + "step": 10124 + }, + { + "epoch": 0.9739322816467872, + "grad_norm": 2.2987644328356054, + "learning_rate": 3.5619241598504603e-08, + "loss": 0.8028, + "step": 10125 + }, + { + "epoch": 0.974028472489419, + "grad_norm": 1.9734308523413244, + "learning_rate": 3.5357008702185945e-08, + "loss": 0.9299, + "step": 10126 + }, + { + "epoch": 0.9741246633320508, + "grad_norm": 2.1484603816297745, + "learning_rate": 3.5095742960266074e-08, + "loss": 0.9103, + "step": 10127 + }, + { + "epoch": 0.9742208541746825, + "grad_norm": 2.1825975296909146, + "learning_rate": 3.483544439810249e-08, + "loss": 0.9096, + "step": 10128 + }, + { + "epoch": 0.9743170450173143, + "grad_norm": 2.227432632282664, + "learning_rate": 3.4576113040961644e-08, + "loss": 0.9199, + "step": 10129 + }, + { + "epoch": 0.9744132358599461, + "grad_norm": 1.9402388497906844, + "learning_rate": 3.4317748914011187e-08, + "loss": 0.8651, + "step": 10130 + }, + { + "epoch": 0.9745094267025779, + "grad_norm": 2.2700906041270152, + "learning_rate": 3.4060352042328825e-08, + "loss": 0.9605, + "step": 10131 + }, + { + "epoch": 0.9746056175452097, + "grad_norm": 2.163107909057661, + "learning_rate": 3.3803922450897917e-08, + "loss": 0.8082, + "step": 10132 + }, + { + "epoch": 0.9747018083878415, + "grad_norm": 1.7718367645190494, + "learning_rate": 3.354846016460522e-08, + "loss": 0.7488, + "step": 10133 + }, + { + "epoch": 0.9747979992304733, + "grad_norm": 2.216284258176509, + "learning_rate": 3.329396520824757e-08, + "loss": 0.9346, + "step": 10134 + }, + { + "epoch": 0.9748941900731051, + "grad_norm": 2.351358028802287, + "learning_rate": 3.30404376065252e-08, + "loss": 0.9427, + "step": 10135 + }, + { + "epoch": 0.9749903809157369, + "grad_norm": 2.183213996873459, + "learning_rate": 3.2787877384045095e-08, + "loss": 1.0762, + "step": 10136 + }, + { + "epoch": 0.9750865717583687, + "grad_norm": 2.0342183530268914, + "learning_rate": 3.253628456531877e-08, + "loss": 0.923, + "step": 10137 + }, + { + "epoch": 0.9751827626010003, + "grad_norm": 2.002240755076845, + "learning_rate": 3.228565917476889e-08, + "loss": 0.8838, + "step": 10138 + }, + { + "epoch": 0.9752789534436321, + "grad_norm": 1.886773516656198, + "learning_rate": 3.203600123671713e-08, + "loss": 0.9162, + "step": 10139 + }, + { + "epoch": 0.9753751442862639, + "grad_norm": 2.167072029094215, + "learning_rate": 3.178731077539743e-08, + "loss": 0.96, + "step": 10140 + }, + { + "epoch": 0.9754713351288957, + "grad_norm": 2.402997215777824, + "learning_rate": 3.153958781494715e-08, + "loss": 0.8147, + "step": 10141 + }, + { + "epoch": 0.9755675259715275, + "grad_norm": 2.1851805410446326, + "learning_rate": 3.129283237940928e-08, + "loss": 0.9477, + "step": 10142 + }, + { + "epoch": 0.9756637168141593, + "grad_norm": 1.9658612812616998, + "learning_rate": 3.1047044492732433e-08, + "loss": 0.9246, + "step": 10143 + }, + { + "epoch": 0.9757599076567911, + "grad_norm": 1.9175176413143593, + "learning_rate": 3.080222417877421e-08, + "loss": 0.878, + "step": 10144 + }, + { + "epoch": 0.9758560984994229, + "grad_norm": 1.8777035428436464, + "learning_rate": 3.05583714612967e-08, + "loss": 0.9192, + "step": 10145 + }, + { + "epoch": 0.9759522893420547, + "grad_norm": 2.015274184845531, + "learning_rate": 3.031548636396764e-08, + "loss": 0.9003, + "step": 10146 + }, + { + "epoch": 0.9760484801846864, + "grad_norm": 2.2919097766427723, + "learning_rate": 3.0073568910359284e-08, + "loss": 0.825, + "step": 10147 + }, + { + "epoch": 0.9761446710273182, + "grad_norm": 2.176831913866839, + "learning_rate": 2.983261912395397e-08, + "loss": 0.8163, + "step": 10148 + }, + { + "epoch": 0.97624086186995, + "grad_norm": 2.119543066828234, + "learning_rate": 2.9592637028136306e-08, + "loss": 0.9291, + "step": 10149 + }, + { + "epoch": 0.9763370527125818, + "grad_norm": 2.2242424471936757, + "learning_rate": 2.9353622646199898e-08, + "loss": 0.933, + "step": 10150 + }, + { + "epoch": 0.9764332435552135, + "grad_norm": 1.9217804410546606, + "learning_rate": 2.9115576001343958e-08, + "loss": 0.8649, + "step": 10151 + }, + { + "epoch": 0.9765294343978453, + "grad_norm": 1.9618933208461196, + "learning_rate": 2.8878497116671124e-08, + "loss": 0.8345, + "step": 10152 + }, + { + "epoch": 0.9766256252404771, + "grad_norm": 2.1064126707093243, + "learning_rate": 2.864238601519187e-08, + "loss": 0.9184, + "step": 10153 + }, + { + "epoch": 0.9767218160831089, + "grad_norm": 2.144326045366129, + "learning_rate": 2.8407242719823424e-08, + "loss": 0.8945, + "step": 10154 + }, + { + "epoch": 0.9768180069257407, + "grad_norm": 2.071078776206681, + "learning_rate": 2.817306725338864e-08, + "loss": 0.9205, + "step": 10155 + }, + { + "epoch": 0.9769141977683724, + "grad_norm": 2.1909672487798844, + "learning_rate": 2.7939859638617118e-08, + "loss": 0.9396, + "step": 10156 + }, + { + "epoch": 0.9770103886110042, + "grad_norm": 1.9915788070337441, + "learning_rate": 2.7707619898141857e-08, + "loss": 0.9866, + "step": 10157 + }, + { + "epoch": 0.977106579453636, + "grad_norm": 2.1241554414271313, + "learning_rate": 2.7476348054504832e-08, + "loss": 0.89, + "step": 10158 + }, + { + "epoch": 0.9772027702962678, + "grad_norm": 1.7636912379044436, + "learning_rate": 2.7246044130153635e-08, + "loss": 0.8789, + "step": 10159 + }, + { + "epoch": 0.9772989611388996, + "grad_norm": 1.973901371614129, + "learning_rate": 2.7016708147439285e-08, + "loss": 0.9581, + "step": 10160 + }, + { + "epoch": 0.9773951519815314, + "grad_norm": 2.065330157714372, + "learning_rate": 2.678834012862175e-08, + "loss": 0.9139, + "step": 10161 + }, + { + "epoch": 0.9774913428241632, + "grad_norm": 2.3451913102993385, + "learning_rate": 2.6560940095866626e-08, + "loss": 1.0552, + "step": 10162 + }, + { + "epoch": 0.977587533666795, + "grad_norm": 2.0379948512016677, + "learning_rate": 2.6334508071245158e-08, + "loss": 0.8311, + "step": 10163 + }, + { + "epoch": 0.9776837245094268, + "grad_norm": 2.101487986182619, + "learning_rate": 2.6109044076733092e-08, + "loss": 0.8128, + "step": 10164 + }, + { + "epoch": 0.9777799153520584, + "grad_norm": 1.9904626677324433, + "learning_rate": 2.588454813421404e-08, + "loss": 0.7685, + "step": 10165 + }, + { + "epoch": 0.9778761061946902, + "grad_norm": 1.9880443188139074, + "learning_rate": 2.5661020265479452e-08, + "loss": 0.9459, + "step": 10166 + }, + { + "epoch": 0.977972297037322, + "grad_norm": 1.7749936679103773, + "learning_rate": 2.5438460492221983e-08, + "loss": 0.7283, + "step": 10167 + }, + { + "epoch": 0.9780684878799538, + "grad_norm": 2.1497911866940393, + "learning_rate": 2.5216868836043242e-08, + "loss": 0.8711, + "step": 10168 + }, + { + "epoch": 0.9781646787225856, + "grad_norm": 1.9665529475222854, + "learning_rate": 2.4996245318453792e-08, + "loss": 0.8613, + "step": 10169 + }, + { + "epoch": 0.9782608695652174, + "grad_norm": 1.8829663793828566, + "learning_rate": 2.4776589960862074e-08, + "loss": 0.8043, + "step": 10170 + }, + { + "epoch": 0.9783570604078492, + "grad_norm": 1.9944182229535705, + "learning_rate": 2.4557902784591024e-08, + "loss": 0.951, + "step": 10171 + }, + { + "epoch": 0.978453251250481, + "grad_norm": 2.181748791210864, + "learning_rate": 2.434018381086589e-08, + "loss": 0.9121, + "step": 10172 + }, + { + "epoch": 0.9785494420931128, + "grad_norm": 1.9018226915998457, + "learning_rate": 2.4123433060816436e-08, + "loss": 0.8899, + "step": 10173 + }, + { + "epoch": 0.9786456329357445, + "grad_norm": 1.8496433033022053, + "learning_rate": 2.3907650555481387e-08, + "loss": 0.9301, + "step": 10174 + }, + { + "epoch": 0.9787418237783763, + "grad_norm": 1.9551947109454424, + "learning_rate": 2.3692836315803992e-08, + "loss": 0.8821, + "step": 10175 + }, + { + "epoch": 0.9788380146210081, + "grad_norm": 1.9972216797477913, + "learning_rate": 2.3478990362634235e-08, + "loss": 0.9597, + "step": 10176 + }, + { + "epoch": 0.9789342054636399, + "grad_norm": 2.12358474284022, + "learning_rate": 2.326611271672663e-08, + "loss": 0.9061, + "step": 10177 + }, + { + "epoch": 0.9790303963062716, + "grad_norm": 1.892569612064968, + "learning_rate": 2.3054203398743537e-08, + "loss": 0.8801, + "step": 10178 + }, + { + "epoch": 0.9791265871489034, + "grad_norm": 1.9694745776857852, + "learning_rate": 2.2843262429254055e-08, + "loss": 0.9145, + "step": 10179 + }, + { + "epoch": 0.9792227779915352, + "grad_norm": 1.8274926316087443, + "learning_rate": 2.263328982872959e-08, + "loss": 0.8648, + "step": 10180 + }, + { + "epoch": 0.979318968834167, + "grad_norm": 1.9742589049964914, + "learning_rate": 2.242428561754939e-08, + "loss": 0.8858, + "step": 10181 + }, + { + "epoch": 0.9794151596767988, + "grad_norm": 1.975085651555176, + "learning_rate": 2.221624981600168e-08, + "loss": 0.9753, + "step": 10182 + }, + { + "epoch": 0.9795113505194305, + "grad_norm": 2.2580448539406692, + "learning_rate": 2.2009182444275855e-08, + "loss": 0.924, + "step": 10183 + }, + { + "epoch": 0.9796075413620623, + "grad_norm": 1.9560750522299215, + "learning_rate": 2.1803083522471402e-08, + "loss": 0.9559, + "step": 10184 + }, + { + "epoch": 0.9797037322046941, + "grad_norm": 1.97996007774171, + "learning_rate": 2.1597953070590095e-08, + "loss": 0.8903, + "step": 10185 + }, + { + "epoch": 0.9797999230473259, + "grad_norm": 2.0027856214179933, + "learning_rate": 2.1393791108542672e-08, + "loss": 0.8883, + "step": 10186 + }, + { + "epoch": 0.9798961138899577, + "grad_norm": 2.169145216073615, + "learning_rate": 2.1190597656145508e-08, + "loss": 0.9392, + "step": 10187 + }, + { + "epoch": 0.9799923047325895, + "grad_norm": 2.0593854868838206, + "learning_rate": 2.098837273311838e-08, + "loss": 0.8487, + "step": 10188 + }, + { + "epoch": 0.9800884955752213, + "grad_norm": 2.152688723730182, + "learning_rate": 2.078711635909114e-08, + "loss": 0.9093, + "step": 10189 + }, + { + "epoch": 0.9801846864178531, + "grad_norm": 2.208271629536105, + "learning_rate": 2.058682855359595e-08, + "loss": 0.8683, + "step": 10190 + }, + { + "epoch": 0.9802808772604849, + "grad_norm": 1.85442066331143, + "learning_rate": 2.0387509336071697e-08, + "loss": 0.9014, + "step": 10191 + }, + { + "epoch": 0.9803770681031165, + "grad_norm": 2.5521373034457016, + "learning_rate": 2.0189158725867353e-08, + "loss": 0.8538, + "step": 10192 + }, + { + "epoch": 0.9804732589457483, + "grad_norm": 2.0263043813100596, + "learning_rate": 1.999177674223085e-08, + "loss": 0.8577, + "step": 10193 + }, + { + "epoch": 0.9805694497883801, + "grad_norm": 2.247733385976737, + "learning_rate": 1.979536340432131e-08, + "loss": 1.0552, + "step": 10194 + }, + { + "epoch": 0.9806656406310119, + "grad_norm": 1.8709521083378933, + "learning_rate": 1.9599918731203482e-08, + "loss": 0.8636, + "step": 10195 + }, + { + "epoch": 0.9807618314736437, + "grad_norm": 2.021303739808852, + "learning_rate": 1.9405442741844415e-08, + "loss": 0.9131, + "step": 10196 + }, + { + "epoch": 0.9808580223162755, + "grad_norm": 2.1347612293793654, + "learning_rate": 1.9211935455122343e-08, + "loss": 0.9012, + "step": 10197 + }, + { + "epoch": 0.9809542131589073, + "grad_norm": 1.8639711220203468, + "learning_rate": 1.9019396889816688e-08, + "loss": 0.8858, + "step": 10198 + }, + { + "epoch": 0.9810504040015391, + "grad_norm": 2.0110613634756054, + "learning_rate": 1.8827827064616943e-08, + "loss": 0.9048, + "step": 10199 + }, + { + "epoch": 0.9811465948441709, + "grad_norm": 1.8980657711301567, + "learning_rate": 1.8637225998114904e-08, + "loss": 0.7941, + "step": 10200 + }, + { + "epoch": 0.9812427856868026, + "grad_norm": 1.9669618764481247, + "learning_rate": 1.844759370881022e-08, + "loss": 0.9357, + "step": 10201 + }, + { + "epoch": 0.9813389765294344, + "grad_norm": 2.003095793914403, + "learning_rate": 1.825893021510927e-08, + "loss": 0.8694, + "step": 10202 + }, + { + "epoch": 0.9814351673720662, + "grad_norm": 1.9482950397897372, + "learning_rate": 1.8071235535322972e-08, + "loss": 0.9358, + "step": 10203 + }, + { + "epoch": 0.981531358214698, + "grad_norm": 2.065107673051166, + "learning_rate": 1.7884509687668972e-08, + "loss": 0.9511, + "step": 10204 + }, + { + "epoch": 0.9816275490573297, + "grad_norm": 1.8025024213305618, + "learning_rate": 1.769875269027055e-08, + "loss": 0.8129, + "step": 10205 + }, + { + "epoch": 0.9817237398999615, + "grad_norm": 2.0353270203356173, + "learning_rate": 1.7513964561156617e-08, + "loss": 0.8602, + "step": 10206 + }, + { + "epoch": 0.9818199307425933, + "grad_norm": 2.165079697152219, + "learning_rate": 1.7330145318262826e-08, + "loss": 0.8676, + "step": 10207 + }, + { + "epoch": 0.9819161215852251, + "grad_norm": 1.8526720704568413, + "learning_rate": 1.714729497942935e-08, + "loss": 0.9281, + "step": 10208 + }, + { + "epoch": 0.9820123124278569, + "grad_norm": 2.1825700212817507, + "learning_rate": 1.6965413562405332e-08, + "loss": 0.9429, + "step": 10209 + }, + { + "epoch": 0.9821085032704886, + "grad_norm": 2.343956883646239, + "learning_rate": 1.6784501084843307e-08, + "loss": 0.9948, + "step": 10210 + }, + { + "epoch": 0.9822046941131204, + "grad_norm": 1.9383034720305723, + "learning_rate": 1.660455756430146e-08, + "loss": 0.8956, + "step": 10211 + }, + { + "epoch": 0.9823008849557522, + "grad_norm": 1.9698058556913909, + "learning_rate": 1.6425583018244706e-08, + "loss": 0.8677, + "step": 10212 + }, + { + "epoch": 0.982397075798384, + "grad_norm": 1.826396022712644, + "learning_rate": 1.6247577464045817e-08, + "loss": 0.9542, + "step": 10213 + }, + { + "epoch": 0.9824932666410158, + "grad_norm": 2.0667516398258456, + "learning_rate": 1.607054091897986e-08, + "loss": 0.8795, + "step": 10214 + }, + { + "epoch": 0.9825894574836476, + "grad_norm": 2.0460748825676727, + "learning_rate": 1.589447340023087e-08, + "loss": 0.9093, + "step": 10215 + }, + { + "epoch": 0.9826856483262794, + "grad_norm": 1.9295303679907754, + "learning_rate": 1.57193749248874e-08, + "loss": 0.8576, + "step": 10216 + }, + { + "epoch": 0.9827818391689112, + "grad_norm": 1.9901580632121265, + "learning_rate": 1.554524550994474e-08, + "loss": 0.8932, + "step": 10217 + }, + { + "epoch": 0.982878030011543, + "grad_norm": 1.8211200711387265, + "learning_rate": 1.537208517230271e-08, + "loss": 0.8453, + "step": 10218 + }, + { + "epoch": 0.9829742208541746, + "grad_norm": 1.9620092205708648, + "learning_rate": 1.5199893928768972e-08, + "loss": 0.9046, + "step": 10219 + }, + { + "epoch": 0.9830704116968064, + "grad_norm": 1.9561403093083674, + "learning_rate": 1.5028671796055715e-08, + "loss": 0.8801, + "step": 10220 + }, + { + "epoch": 0.9831666025394382, + "grad_norm": 1.8037491748170162, + "learning_rate": 1.485841879078187e-08, + "loss": 0.8356, + "step": 10221 + }, + { + "epoch": 0.98326279338207, + "grad_norm": 1.947094100391303, + "learning_rate": 1.4689134929470884e-08, + "loss": 0.796, + "step": 10222 + }, + { + "epoch": 0.9833589842247018, + "grad_norm": 1.9331445882625737, + "learning_rate": 1.4520820228554055e-08, + "loss": 0.8206, + "step": 10223 + }, + { + "epoch": 0.9834551750673336, + "grad_norm": 1.6140856131699715, + "learning_rate": 1.435347470436832e-08, + "loss": 0.7688, + "step": 10224 + }, + { + "epoch": 0.9835513659099654, + "grad_norm": 1.9922737794888945, + "learning_rate": 1.4187098373155129e-08, + "loss": 0.9316, + "step": 10225 + }, + { + "epoch": 0.9836475567525972, + "grad_norm": 1.6915979057811306, + "learning_rate": 1.4021691251062675e-08, + "loss": 0.7691, + "step": 10226 + }, + { + "epoch": 0.983743747595229, + "grad_norm": 1.929932822099693, + "learning_rate": 1.3857253354144784e-08, + "loss": 0.9412, + "step": 10227 + }, + { + "epoch": 0.9838399384378607, + "grad_norm": 1.7101316837696163, + "learning_rate": 1.3693784698363133e-08, + "loss": 0.7769, + "step": 10228 + }, + { + "epoch": 0.9839361292804925, + "grad_norm": 2.083501748166912, + "learning_rate": 1.3531285299582809e-08, + "loss": 0.9247, + "step": 10229 + }, + { + "epoch": 0.9840323201231242, + "grad_norm": 2.026721819081889, + "learning_rate": 1.3369755173575639e-08, + "loss": 0.8786, + "step": 10230 + }, + { + "epoch": 0.984128510965756, + "grad_norm": 2.059904212602746, + "learning_rate": 1.3209194336019082e-08, + "loss": 0.7687, + "step": 10231 + }, + { + "epoch": 0.9842247018083878, + "grad_norm": 1.9985941686595816, + "learning_rate": 1.3049602802498451e-08, + "loss": 0.9371, + "step": 10232 + }, + { + "epoch": 0.9843208926510196, + "grad_norm": 2.2426407822413448, + "learning_rate": 1.2890980588502466e-08, + "loss": 0.8253, + "step": 10233 + }, + { + "epoch": 0.9844170834936514, + "grad_norm": 1.8042756722976794, + "learning_rate": 1.273332770942659e-08, + "loss": 0.9409, + "step": 10234 + }, + { + "epoch": 0.9845132743362832, + "grad_norm": 2.127656556822123, + "learning_rate": 1.2576644180573028e-08, + "loss": 0.8639, + "step": 10235 + }, + { + "epoch": 0.984609465178915, + "grad_norm": 2.159461346934955, + "learning_rate": 1.2420930017148503e-08, + "loss": 0.9708, + "step": 10236 + }, + { + "epoch": 0.9847056560215467, + "grad_norm": 2.0870850181777274, + "learning_rate": 1.2266185234266481e-08, + "loss": 0.8706, + "step": 10237 + }, + { + "epoch": 0.9848018468641785, + "grad_norm": 2.1497812448005504, + "learning_rate": 1.2112409846947171e-08, + "loss": 0.8264, + "step": 10238 + }, + { + "epoch": 0.9848980377068103, + "grad_norm": 2.3122582896710084, + "learning_rate": 1.195960387011641e-08, + "loss": 0.9043, + "step": 10239 + }, + { + "epoch": 0.9849942285494421, + "grad_norm": 2.036796832438141, + "learning_rate": 1.1807767318602337e-08, + "loss": 0.9951, + "step": 10240 + }, + { + "epoch": 0.9850904193920739, + "grad_norm": 1.9167679229286936, + "learning_rate": 1.1656900207144272e-08, + "loss": 0.929, + "step": 10241 + }, + { + "epoch": 0.9851866102347057, + "grad_norm": 2.041539230038121, + "learning_rate": 1.150700255038606e-08, + "loss": 0.8576, + "step": 10242 + }, + { + "epoch": 0.9852828010773375, + "grad_norm": 1.8806853607939022, + "learning_rate": 1.1358074362873838e-08, + "loss": 0.8966, + "step": 10243 + }, + { + "epoch": 0.9853789919199692, + "grad_norm": 1.8771061698734635, + "learning_rate": 1.1210115659063825e-08, + "loss": 0.9002, + "step": 10244 + }, + { + "epoch": 0.985475182762601, + "grad_norm": 1.8591553061401964, + "learning_rate": 1.1063126453317862e-08, + "loss": 0.7866, + "step": 10245 + }, + { + "epoch": 0.9855713736052328, + "grad_norm": 2.055573302573506, + "learning_rate": 1.0917106759900097e-08, + "loss": 0.9675, + "step": 10246 + }, + { + "epoch": 0.9856675644478645, + "grad_norm": 2.180625176809647, + "learning_rate": 1.0772056592984748e-08, + "loss": 0.8889, + "step": 10247 + }, + { + "epoch": 0.9857637552904963, + "grad_norm": 1.870690488874582, + "learning_rate": 1.0627975966649439e-08, + "loss": 0.8641, + "step": 10248 + }, + { + "epoch": 0.9858599461331281, + "grad_norm": 1.9753837912944872, + "learning_rate": 1.0484864894878544e-08, + "loss": 0.8387, + "step": 10249 + }, + { + "epoch": 0.9859561369757599, + "grad_norm": 1.893688759163754, + "learning_rate": 1.034272339156206e-08, + "loss": 0.9112, + "step": 10250 + }, + { + "epoch": 0.9860523278183917, + "grad_norm": 1.9686771147757467, + "learning_rate": 1.020155147049673e-08, + "loss": 0.8364, + "step": 10251 + }, + { + "epoch": 0.9861485186610235, + "grad_norm": 1.843641192781525, + "learning_rate": 1.0061349145383814e-08, + "loss": 0.8688, + "step": 10252 + }, + { + "epoch": 0.9862447095036553, + "grad_norm": 2.3066146028190473, + "learning_rate": 9.92211642983132e-09, + "loss": 1.0292, + "step": 10253 + }, + { + "epoch": 0.9863409003462871, + "grad_norm": 1.904992660747896, + "learning_rate": 9.783853337353987e-09, + "loss": 0.8804, + "step": 10254 + }, + { + "epoch": 0.9864370911889189, + "grad_norm": 2.0093790188932226, + "learning_rate": 9.646559881369977e-09, + "loss": 0.8106, + "step": 10255 + }, + { + "epoch": 0.9865332820315506, + "grad_norm": 2.125326679314012, + "learning_rate": 9.510236075205292e-09, + "loss": 0.8453, + "step": 10256 + }, + { + "epoch": 0.9866294728741823, + "grad_norm": 2.027201577057414, + "learning_rate": 9.374881932090462e-09, + "loss": 0.8132, + "step": 10257 + }, + { + "epoch": 0.9867256637168141, + "grad_norm": 2.1215208479141805, + "learning_rate": 9.240497465164978e-09, + "loss": 0.8969, + "step": 10258 + }, + { + "epoch": 0.9868218545594459, + "grad_norm": 2.16781106617069, + "learning_rate": 9.107082687470626e-09, + "loss": 0.8724, + "step": 10259 + }, + { + "epoch": 0.9869180454020777, + "grad_norm": 2.0095177055304854, + "learning_rate": 8.974637611955939e-09, + "loss": 0.9704, + "step": 10260 + }, + { + "epoch": 0.9870142362447095, + "grad_norm": 2.2400634598592317, + "learning_rate": 8.8431622514773e-09, + "loss": 0.9492, + "step": 10261 + }, + { + "epoch": 0.9871104270873413, + "grad_norm": 2.280219778808852, + "learning_rate": 8.712656618793391e-09, + "loss": 0.8797, + "step": 10262 + }, + { + "epoch": 0.9872066179299731, + "grad_norm": 1.7589634651324688, + "learning_rate": 8.583120726572969e-09, + "loss": 0.9316, + "step": 10263 + }, + { + "epoch": 0.9873028087726049, + "grad_norm": 1.8042860805170555, + "learning_rate": 8.454554587388198e-09, + "loss": 0.8653, + "step": 10264 + }, + { + "epoch": 0.9873989996152366, + "grad_norm": 2.061090271919015, + "learning_rate": 8.326958213716874e-09, + "loss": 0.9358, + "step": 10265 + }, + { + "epoch": 0.9874951904578684, + "grad_norm": 2.321630384442382, + "learning_rate": 8.200331617943535e-09, + "loss": 0.9768, + "step": 10266 + }, + { + "epoch": 0.9875913813005002, + "grad_norm": 2.09010548857701, + "learning_rate": 8.074674812358351e-09, + "loss": 0.8163, + "step": 10267 + }, + { + "epoch": 0.987687572143132, + "grad_norm": 1.925816544255212, + "learning_rate": 7.949987809158232e-09, + "loss": 0.9079, + "step": 10268 + }, + { + "epoch": 0.9877837629857638, + "grad_norm": 1.884599154035372, + "learning_rate": 7.8262706204435e-09, + "loss": 0.8682, + "step": 10269 + }, + { + "epoch": 0.9878799538283956, + "grad_norm": 1.9466538175410621, + "learning_rate": 7.703523258223433e-09, + "loss": 0.7963, + "step": 10270 + }, + { + "epoch": 0.9879761446710273, + "grad_norm": 2.380077137916529, + "learning_rate": 7.581745734410728e-09, + "loss": 0.8689, + "step": 10271 + }, + { + "epoch": 0.9880723355136591, + "grad_norm": 2.603677409003968, + "learning_rate": 7.460938060825929e-09, + "loss": 0.9027, + "step": 10272 + }, + { + "epoch": 0.9881685263562909, + "grad_norm": 1.8981775742315314, + "learning_rate": 7.341100249194099e-09, + "loss": 0.9339, + "step": 10273 + }, + { + "epoch": 0.9882647171989226, + "grad_norm": 1.8085135270221973, + "learning_rate": 7.222232311145938e-09, + "loss": 0.8607, + "step": 10274 + }, + { + "epoch": 0.9883609080415544, + "grad_norm": 1.8974074399879237, + "learning_rate": 7.1043342582188816e-09, + "loss": 0.9379, + "step": 10275 + }, + { + "epoch": 0.9884570988841862, + "grad_norm": 2.353552190423512, + "learning_rate": 6.987406101855998e-09, + "loss": 0.871, + "step": 10276 + }, + { + "epoch": 0.988553289726818, + "grad_norm": 2.0262049875064547, + "learning_rate": 6.871447853405988e-09, + "loss": 0.8679, + "step": 10277 + }, + { + "epoch": 0.9886494805694498, + "grad_norm": 2.167016559107659, + "learning_rate": 6.756459524125403e-09, + "loss": 0.9079, + "step": 10278 + }, + { + "epoch": 0.9887456714120816, + "grad_norm": 2.0141634660409755, + "learning_rate": 6.642441125171984e-09, + "loss": 0.9211, + "step": 10279 + }, + { + "epoch": 0.9888418622547134, + "grad_norm": 2.048128719620014, + "learning_rate": 6.5293926676135434e-09, + "loss": 0.9377, + "step": 10280 + }, + { + "epoch": 0.9889380530973452, + "grad_norm": 2.1782078480562808, + "learning_rate": 6.417314162422417e-09, + "loss": 0.8689, + "step": 10281 + }, + { + "epoch": 0.989034243939977, + "grad_norm": 2.0527878950345704, + "learning_rate": 6.306205620477679e-09, + "loss": 0.8846, + "step": 10282 + }, + { + "epoch": 0.9891304347826086, + "grad_norm": 2.0371972404926133, + "learning_rate": 6.196067052561816e-09, + "loss": 0.9875, + "step": 10283 + }, + { + "epoch": 0.9892266256252404, + "grad_norm": 2.107078955383078, + "learning_rate": 6.086898469365166e-09, + "loss": 0.8507, + "step": 10284 + }, + { + "epoch": 0.9893228164678722, + "grad_norm": 1.719077414013469, + "learning_rate": 5.9786998814836964e-09, + "loss": 0.7402, + "step": 10285 + }, + { + "epoch": 0.989419007310504, + "grad_norm": 2.163236340453398, + "learning_rate": 5.871471299419007e-09, + "loss": 0.9263, + "step": 10286 + }, + { + "epoch": 0.9895151981531358, + "grad_norm": 2.0892488686969286, + "learning_rate": 5.76521273357944e-09, + "loss": 0.8773, + "step": 10287 + }, + { + "epoch": 0.9896113889957676, + "grad_norm": 2.0489674012761196, + "learning_rate": 5.6599241942767445e-09, + "loss": 0.9193, + "step": 10288 + }, + { + "epoch": 0.9897075798383994, + "grad_norm": 2.1255040893292336, + "learning_rate": 5.555605691730526e-09, + "loss": 0.9532, + "step": 10289 + }, + { + "epoch": 0.9898037706810312, + "grad_norm": 2.3266018746780333, + "learning_rate": 5.452257236066017e-09, + "loss": 0.9751, + "step": 10290 + }, + { + "epoch": 0.989899961523663, + "grad_norm": 1.9684283840557744, + "learning_rate": 5.349878837314082e-09, + "loss": 0.7569, + "step": 10291 + }, + { + "epoch": 0.9899961523662947, + "grad_norm": 2.163069130049635, + "learning_rate": 5.248470505412328e-09, + "loss": 0.9737, + "step": 10292 + }, + { + "epoch": 0.9900923432089265, + "grad_norm": 2.1694041660759544, + "learning_rate": 5.148032250201773e-09, + "loss": 0.9141, + "step": 10293 + }, + { + "epoch": 0.9901885340515583, + "grad_norm": 1.7881232350427523, + "learning_rate": 5.0485640814312844e-09, + "loss": 0.7902, + "step": 10294 + }, + { + "epoch": 0.9902847248941901, + "grad_norm": 1.954022722339971, + "learning_rate": 4.950066008755361e-09, + "loss": 0.9837, + "step": 10295 + }, + { + "epoch": 0.9903809157368219, + "grad_norm": 1.956721071876201, + "learning_rate": 4.8525380417330234e-09, + "loss": 0.9834, + "step": 10296 + }, + { + "epoch": 0.9904771065794536, + "grad_norm": 1.8640944959456853, + "learning_rate": 4.7559801898322546e-09, + "loss": 0.9331, + "step": 10297 + }, + { + "epoch": 0.9905732974220854, + "grad_norm": 1.9780533335150745, + "learning_rate": 4.660392462424446e-09, + "loss": 0.9004, + "step": 10298 + }, + { + "epoch": 0.9906694882647172, + "grad_norm": 2.2484055954365556, + "learning_rate": 4.565774868785511e-09, + "loss": 0.807, + "step": 10299 + }, + { + "epoch": 0.990765679107349, + "grad_norm": 1.90640961600953, + "learning_rate": 4.472127418099215e-09, + "loss": 0.8543, + "step": 10300 + }, + { + "epoch": 0.9908618699499807, + "grad_norm": 2.2503400837992062, + "learning_rate": 4.379450119457173e-09, + "loss": 0.8217, + "step": 10301 + }, + { + "epoch": 0.9909580607926125, + "grad_norm": 1.7014240661799707, + "learning_rate": 4.287742981851084e-09, + "loss": 0.7872, + "step": 10302 + }, + { + "epoch": 0.9910542516352443, + "grad_norm": 1.8686521933947202, + "learning_rate": 4.1970060141849346e-09, + "loss": 0.873, + "step": 10303 + }, + { + "epoch": 0.9911504424778761, + "grad_norm": 2.022439469854263, + "learning_rate": 4.1072392252639034e-09, + "loss": 0.9111, + "step": 10304 + }, + { + "epoch": 0.9912466333205079, + "grad_norm": 1.623837605289439, + "learning_rate": 4.0184426238010224e-09, + "loss": 0.8557, + "step": 10305 + }, + { + "epoch": 0.9913428241631397, + "grad_norm": 1.8949220420104296, + "learning_rate": 3.930616218414951e-09, + "loss": 0.8869, + "step": 10306 + }, + { + "epoch": 0.9914390150057715, + "grad_norm": 1.9838545100349885, + "learning_rate": 3.8437600176299824e-09, + "loss": 0.8999, + "step": 10307 + }, + { + "epoch": 0.9915352058484033, + "grad_norm": 2.1478909325862188, + "learning_rate": 3.757874029874931e-09, + "loss": 0.9411, + "step": 10308 + }, + { + "epoch": 0.9916313966910351, + "grad_norm": 2.108344745350667, + "learning_rate": 3.6729582634875695e-09, + "loss": 0.8072, + "step": 10309 + }, + { + "epoch": 0.9917275875336667, + "grad_norm": 1.9444399235963477, + "learning_rate": 3.5890127267090844e-09, + "loss": 0.8656, + "step": 10310 + }, + { + "epoch": 0.9918237783762985, + "grad_norm": 1.8497956679331906, + "learning_rate": 3.506037427687403e-09, + "loss": 0.8866, + "step": 10311 + }, + { + "epoch": 0.9919199692189303, + "grad_norm": 2.066299299677996, + "learning_rate": 3.424032374476083e-09, + "loss": 0.7852, + "step": 10312 + }, + { + "epoch": 0.9920161600615621, + "grad_norm": 2.202488644940473, + "learning_rate": 3.3429975750332024e-09, + "loss": 0.8612, + "step": 10313 + }, + { + "epoch": 0.9921123509041939, + "grad_norm": 1.902263837931421, + "learning_rate": 3.2629330372246915e-09, + "loss": 0.8985, + "step": 10314 + }, + { + "epoch": 0.9922085417468257, + "grad_norm": 2.324251957926468, + "learning_rate": 3.183838768822112e-09, + "loss": 0.8558, + "step": 10315 + }, + { + "epoch": 0.9923047325894575, + "grad_norm": 1.7322058131672484, + "learning_rate": 3.105714777501545e-09, + "loss": 0.7303, + "step": 10316 + }, + { + "epoch": 0.9924009234320893, + "grad_norm": 2.0188240214965303, + "learning_rate": 3.0285610708447046e-09, + "loss": 0.8805, + "step": 10317 + }, + { + "epoch": 0.9924971142747211, + "grad_norm": 2.439550592511485, + "learning_rate": 2.9523776563422644e-09, + "loss": 0.9472, + "step": 10318 + }, + { + "epoch": 0.9925933051173528, + "grad_norm": 1.9656361812572598, + "learning_rate": 2.8771645413860904e-09, + "loss": 0.9528, + "step": 10319 + }, + { + "epoch": 0.9926894959599846, + "grad_norm": 2.0332726035629354, + "learning_rate": 2.802921733278119e-09, + "loss": 0.9884, + "step": 10320 + }, + { + "epoch": 0.9927856868026164, + "grad_norm": 1.5762090815005965, + "learning_rate": 2.729649239223697e-09, + "loss": 0.7635, + "step": 10321 + }, + { + "epoch": 0.9928818776452482, + "grad_norm": 2.016281276335049, + "learning_rate": 2.657347066333804e-09, + "loss": 0.9262, + "step": 10322 + }, + { + "epoch": 0.99297806848788, + "grad_norm": 1.7443401646552552, + "learning_rate": 2.5860152216272693e-09, + "loss": 0.7904, + "step": 10323 + }, + { + "epoch": 0.9930742593305117, + "grad_norm": 1.9280079777179362, + "learning_rate": 2.5156537120263335e-09, + "loss": 0.8883, + "step": 10324 + }, + { + "epoch": 0.9931704501731435, + "grad_norm": 2.0725352896896796, + "learning_rate": 2.4462625443599784e-09, + "loss": 0.928, + "step": 10325 + }, + { + "epoch": 0.9932666410157753, + "grad_norm": 2.218123525502165, + "learning_rate": 2.3778417253650376e-09, + "loss": 0.914, + "step": 10326 + }, + { + "epoch": 0.9933628318584071, + "grad_norm": 2.5725495644273884, + "learning_rate": 2.3103912616806446e-09, + "loss": 1.0951, + "step": 10327 + }, + { + "epoch": 0.9934590227010388, + "grad_norm": 1.6502113021190197, + "learning_rate": 2.2439111598537844e-09, + "loss": 0.7984, + "step": 10328 + }, + { + "epoch": 0.9935552135436706, + "grad_norm": 1.8368950959322776, + "learning_rate": 2.1784014263370732e-09, + "loss": 0.9086, + "step": 10329 + }, + { + "epoch": 0.9936514043863024, + "grad_norm": 2.3076235640863443, + "learning_rate": 2.113862067488759e-09, + "loss": 0.8972, + "step": 10330 + }, + { + "epoch": 0.9937475952289342, + "grad_norm": 1.911619903842417, + "learning_rate": 2.0502930895738294e-09, + "loss": 0.8761, + "step": 10331 + }, + { + "epoch": 0.993843786071566, + "grad_norm": 2.042592923520889, + "learning_rate": 1.987694498760684e-09, + "loss": 0.7848, + "step": 10332 + }, + { + "epoch": 0.9939399769141978, + "grad_norm": 2.1009650553847394, + "learning_rate": 1.926066301125573e-09, + "loss": 0.895, + "step": 10333 + }, + { + "epoch": 0.9940361677568296, + "grad_norm": 1.8826276827541122, + "learning_rate": 1.865408502650379e-09, + "loss": 0.8413, + "step": 10334 + }, + { + "epoch": 0.9941323585994614, + "grad_norm": 2.068855878533087, + "learning_rate": 1.8057211092226134e-09, + "loss": 0.7863, + "step": 10335 + }, + { + "epoch": 0.9942285494420932, + "grad_norm": 2.0059543747933675, + "learning_rate": 1.747004126635421e-09, + "loss": 0.8749, + "step": 10336 + }, + { + "epoch": 0.9943247402847248, + "grad_norm": 2.0465837319973685, + "learning_rate": 1.6892575605864658e-09, + "loss": 0.9369, + "step": 10337 + }, + { + "epoch": 0.9944209311273566, + "grad_norm": 2.1069673856132556, + "learning_rate": 1.6324814166823744e-09, + "loss": 0.8462, + "step": 10338 + }, + { + "epoch": 0.9945171219699884, + "grad_norm": 2.1996526137523214, + "learning_rate": 1.576675700432073e-09, + "loss": 0.8786, + "step": 10339 + }, + { + "epoch": 0.9946133128126202, + "grad_norm": 1.852553957040892, + "learning_rate": 1.5218404172545609e-09, + "loss": 0.7902, + "step": 10340 + }, + { + "epoch": 0.994709503655252, + "grad_norm": 2.0213305037121385, + "learning_rate": 1.467975572468916e-09, + "loss": 0.8466, + "step": 10341 + }, + { + "epoch": 0.9948056944978838, + "grad_norm": 2.112135132847389, + "learning_rate": 1.415081171305399e-09, + "loss": 0.8196, + "step": 10342 + }, + { + "epoch": 0.9949018853405156, + "grad_norm": 2.07773801172187, + "learning_rate": 1.3631572188965714e-09, + "loss": 0.8618, + "step": 10343 + }, + { + "epoch": 0.9949980761831474, + "grad_norm": 1.9364517334662246, + "learning_rate": 1.3122037202828452e-09, + "loss": 0.8597, + "step": 10344 + }, + { + "epoch": 0.9950942670257792, + "grad_norm": 2.027872239725166, + "learning_rate": 1.2622206804102644e-09, + "loss": 0.8796, + "step": 10345 + }, + { + "epoch": 0.995190457868411, + "grad_norm": 2.076628142821543, + "learning_rate": 1.2132081041282829e-09, + "loss": 0.8663, + "step": 10346 + }, + { + "epoch": 0.9952866487110427, + "grad_norm": 2.0183943158976185, + "learning_rate": 1.1651659961964269e-09, + "loss": 0.9238, + "step": 10347 + }, + { + "epoch": 0.9953828395536745, + "grad_norm": 2.0615382785018066, + "learning_rate": 1.1180943612754124e-09, + "loss": 0.9476, + "step": 10348 + }, + { + "epoch": 0.9954790303963063, + "grad_norm": 2.1577269724617554, + "learning_rate": 1.0719932039338077e-09, + "loss": 0.9247, + "step": 10349 + }, + { + "epoch": 0.995575221238938, + "grad_norm": 2.0205240564251365, + "learning_rate": 1.026862528649142e-09, + "loss": 0.8817, + "step": 10350 + }, + { + "epoch": 0.9956714120815698, + "grad_norm": 1.857007697148919, + "learning_rate": 9.82702339797914e-10, + "loss": 0.8876, + "step": 10351 + }, + { + "epoch": 0.9957676029242016, + "grad_norm": 2.013825650328713, + "learning_rate": 9.39512641668916e-10, + "loss": 0.9626, + "step": 10352 + }, + { + "epoch": 0.9958637937668334, + "grad_norm": 1.9776756498356731, + "learning_rate": 8.972934384532395e-10, + "loss": 0.8745, + "step": 10353 + }, + { + "epoch": 0.9959599846094652, + "grad_norm": 1.9447648773908586, + "learning_rate": 8.560447342487177e-10, + "loss": 0.8843, + "step": 10354 + }, + { + "epoch": 0.996056175452097, + "grad_norm": 1.903908239005716, + "learning_rate": 8.157665330577047e-10, + "loss": 0.8829, + "step": 10355 + }, + { + "epoch": 0.9961523662947287, + "grad_norm": 1.9367092278549278, + "learning_rate": 7.764588387915161e-10, + "loss": 0.9132, + "step": 10356 + }, + { + "epoch": 0.9962485571373605, + "grad_norm": 1.921333110523481, + "learning_rate": 7.381216552648785e-10, + "loss": 0.848, + "step": 10357 + }, + { + "epoch": 0.9963447479799923, + "grad_norm": 2.1634114952446057, + "learning_rate": 7.007549861970387e-10, + "loss": 0.7935, + "step": 10358 + }, + { + "epoch": 0.9964409388226241, + "grad_norm": 2.1780393766022677, + "learning_rate": 6.643588352162056e-10, + "loss": 1.0863, + "step": 10359 + }, + { + "epoch": 0.9965371296652559, + "grad_norm": 2.013210299842679, + "learning_rate": 6.289332058551089e-10, + "loss": 0.8967, + "step": 10360 + }, + { + "epoch": 0.9966333205078877, + "grad_norm": 2.134262996948331, + "learning_rate": 5.944781015521095e-10, + "loss": 0.9797, + "step": 10361 + }, + { + "epoch": 0.9967295113505195, + "grad_norm": 2.0597579579055507, + "learning_rate": 5.609935256500887e-10, + "loss": 0.8941, + "step": 10362 + }, + { + "epoch": 0.9968257021931513, + "grad_norm": 1.7950881523844542, + "learning_rate": 5.284794814008898e-10, + "loss": 0.8904, + "step": 10363 + }, + { + "epoch": 0.996921893035783, + "grad_norm": 1.7862948959578082, + "learning_rate": 4.969359719586563e-10, + "loss": 0.974, + "step": 10364 + }, + { + "epoch": 0.9970180838784147, + "grad_norm": 2.004804541879481, + "learning_rate": 4.663630003853836e-10, + "loss": 0.9005, + "step": 10365 + }, + { + "epoch": 0.9971142747210465, + "grad_norm": 1.9292481317733774, + "learning_rate": 4.3676056964869764e-10, + "loss": 0.8427, + "step": 10366 + }, + { + "epoch": 0.9972104655636783, + "grad_norm": 1.8905077033956459, + "learning_rate": 4.081286826218556e-10, + "loss": 0.8017, + "step": 10367 + }, + { + "epoch": 0.9973066564063101, + "grad_norm": 2.3355325951678405, + "learning_rate": 3.804673420837457e-10, + "loss": 0.9452, + "step": 10368 + }, + { + "epoch": 0.9974028472489419, + "grad_norm": 1.7640683548309242, + "learning_rate": 3.537765507188873e-10, + "loss": 0.9531, + "step": 10369 + }, + { + "epoch": 0.9974990380915737, + "grad_norm": 1.9918675881929706, + "learning_rate": 3.2805631111743064e-10, + "loss": 0.9034, + "step": 10370 + }, + { + "epoch": 0.9975952289342055, + "grad_norm": 1.9136152711514351, + "learning_rate": 3.033066257773776e-10, + "loss": 0.8983, + "step": 10371 + }, + { + "epoch": 0.9976914197768373, + "grad_norm": 2.1789845352099677, + "learning_rate": 2.795274971001405e-10, + "loss": 0.9282, + "step": 10372 + }, + { + "epoch": 0.9977876106194691, + "grad_norm": 2.217519249426482, + "learning_rate": 2.567189273927628e-10, + "loss": 0.928, + "step": 10373 + }, + { + "epoch": 0.9978838014621008, + "grad_norm": 2.0962008389543425, + "learning_rate": 2.3488091886902933e-10, + "loss": 0.8748, + "step": 10374 + }, + { + "epoch": 0.9979799923047326, + "grad_norm": 2.095064475832426, + "learning_rate": 2.1401347365057613e-10, + "loss": 0.8507, + "step": 10375 + }, + { + "epoch": 0.9980761831473643, + "grad_norm": 1.997127289961751, + "learning_rate": 1.941165937602296e-10, + "loss": 0.9199, + "step": 10376 + }, + { + "epoch": 0.9981723739899961, + "grad_norm": 1.890369996324005, + "learning_rate": 1.751902811308881e-10, + "loss": 0.9165, + "step": 10377 + }, + { + "epoch": 0.9982685648326279, + "grad_norm": 2.2422442313177235, + "learning_rate": 1.5723453759886042e-10, + "loss": 0.9054, + "step": 10378 + }, + { + "epoch": 0.9983647556752597, + "grad_norm": 2.0489986657440595, + "learning_rate": 1.4024936490719677e-10, + "loss": 0.8572, + "step": 10379 + }, + { + "epoch": 0.9984609465178915, + "grad_norm": 1.8050906192914915, + "learning_rate": 1.2423476470346808e-10, + "loss": 0.8322, + "step": 10380 + }, + { + "epoch": 0.9985571373605233, + "grad_norm": 1.8999125695930943, + "learning_rate": 1.091907385430968e-10, + "loss": 0.9483, + "step": 10381 + }, + { + "epoch": 0.9986533282031551, + "grad_norm": 1.7801949680824989, + "learning_rate": 9.511728788602625e-11, + "loss": 0.807, + "step": 10382 + }, + { + "epoch": 0.9987495190457868, + "grad_norm": 1.857271881363245, + "learning_rate": 8.201441409783073e-11, + "loss": 0.8925, + "step": 10383 + }, + { + "epoch": 0.9988457098884186, + "grad_norm": 2.0357384735416675, + "learning_rate": 6.988211845082582e-11, + "loss": 0.8314, + "step": 10384 + }, + { + "epoch": 0.9989419007310504, + "grad_norm": 2.1198413570870813, + "learning_rate": 5.872040212184794e-11, + "loss": 0.888, + "step": 10385 + }, + { + "epoch": 0.9990380915736822, + "grad_norm": 1.6399318341727054, + "learning_rate": 4.852926619447473e-11, + "loss": 0.8246, + "step": 10386 + }, + { + "epoch": 0.999134282416314, + "grad_norm": 2.3402110192601904, + "learning_rate": 3.9308711657914945e-11, + "loss": 0.9098, + "step": 10387 + }, + { + "epoch": 0.9992304732589458, + "grad_norm": 1.99148033913017, + "learning_rate": 3.105873940811854e-11, + "loss": 0.8518, + "step": 10388 + }, + { + "epoch": 0.9993266641015776, + "grad_norm": 2.101345938968807, + "learning_rate": 2.3779350244446108e-11, + "loss": 0.8366, + "step": 10389 + }, + { + "epoch": 0.9994228549442093, + "grad_norm": 1.988279001738194, + "learning_rate": 1.7470544874109706e-11, + "loss": 0.8833, + "step": 10390 + }, + { + "epoch": 0.9995190457868411, + "grad_norm": 1.868766441270856, + "learning_rate": 1.2132323908842225e-11, + "loss": 0.9512, + "step": 10391 + }, + { + "epoch": 0.9996152366294728, + "grad_norm": 1.8797513257442497, + "learning_rate": 7.764687866007592e-12, + "loss": 0.8102, + "step": 10392 + }, + { + "epoch": 0.9997114274721046, + "grad_norm": 2.205390481198828, + "learning_rate": 4.367637171931449e-12, + "loss": 1.0668, + "step": 10393 + }, + { + "epoch": 0.9998076183147364, + "grad_norm": 2.211713690889391, + "learning_rate": 1.9411721552398123e-12, + "loss": 0.9433, + "step": 10394 + }, + { + "epoch": 0.9999038091573682, + "grad_norm": 2.3271771015838136, + "learning_rate": 4.852930501897391e-13, + "loss": 0.7642, + "step": 10395 + }, + { + "epoch": 1.0, + "grad_norm": 2.2192596166220757, + "learning_rate": 0.0, + "loss": 0.8736, + "step": 10396 + }, + { + "epoch": 1.0, + "step": 10396, + "total_flos": 1355795318128640.0, + "train_loss": 0.9816701800296196, + "train_runtime": 36581.4439, + "train_samples_per_second": 18.187, + "train_steps_per_second": 0.284 + } + ], + "logging_steps": 1.0, + "max_steps": 10396, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 50000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": false, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1355795318128640.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +}