| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9991004497751125, | |
| "eval_steps": 500, | |
| "global_step": 833, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0011994002998500749, | |
| "grad_norm": 0.11149752140045166, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 10.3757, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0023988005997001498, | |
| "grad_norm": 0.1320822536945343, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 10.3748, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.003598200899550225, | |
| "grad_norm": 0.13625332713127136, | |
| "learning_rate": 3e-06, | |
| "loss": 10.3734, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0047976011994002995, | |
| "grad_norm": 0.1434432715177536, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 10.3739, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.005997001499250375, | |
| "grad_norm": 0.14035488665103912, | |
| "learning_rate": 5e-06, | |
| "loss": 10.3765, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00719640179910045, | |
| "grad_norm": 0.14915668964385986, | |
| "learning_rate": 6e-06, | |
| "loss": 10.3743, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.008395802098950524, | |
| "grad_norm": 0.15720613300800323, | |
| "learning_rate": 7.000000000000001e-06, | |
| "loss": 10.3724, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.009595202398800599, | |
| "grad_norm": 0.15728804469108582, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 10.3741, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.010794602698650674, | |
| "grad_norm": 0.16915467381477356, | |
| "learning_rate": 9e-06, | |
| "loss": 10.3754, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.01199400299850075, | |
| "grad_norm": 0.17764562368392944, | |
| "learning_rate": 1e-05, | |
| "loss": 10.3721, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.013193403298350824, | |
| "grad_norm": 0.19067725539207458, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 10.3733, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0143928035982009, | |
| "grad_norm": 0.20686227083206177, | |
| "learning_rate": 1.2e-05, | |
| "loss": 10.3718, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.015592203898050975, | |
| "grad_norm": 0.13864944875240326, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 10.3757, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.016791604197901048, | |
| "grad_norm": 0.13187511265277863, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 10.3746, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.017991004497751123, | |
| "grad_norm": 0.13840411603450775, | |
| "learning_rate": 1.5e-05, | |
| "loss": 10.3745, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.019190404797601198, | |
| "grad_norm": 0.13771533966064453, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 10.3757, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.020389805097451273, | |
| "grad_norm": 0.144022136926651, | |
| "learning_rate": 1.7000000000000003e-05, | |
| "loss": 10.3742, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.02158920539730135, | |
| "grad_norm": 0.14401449263095856, | |
| "learning_rate": 1.8e-05, | |
| "loss": 10.3742, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.022788605697151423, | |
| "grad_norm": 0.15697641670703888, | |
| "learning_rate": 1.9e-05, | |
| "loss": 10.3749, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0239880059970015, | |
| "grad_norm": 0.15877871215343475, | |
| "learning_rate": 2e-05, | |
| "loss": 10.3725, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.025187406296851574, | |
| "grad_norm": 0.16563433408737183, | |
| "learning_rate": 2.1e-05, | |
| "loss": 10.3743, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.02638680659670165, | |
| "grad_norm": 0.17025835812091827, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 10.3738, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.027586206896551724, | |
| "grad_norm": 0.17753835022449493, | |
| "learning_rate": 2.3000000000000003e-05, | |
| "loss": 10.3732, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0287856071964018, | |
| "grad_norm": 0.19433385133743286, | |
| "learning_rate": 2.4e-05, | |
| "loss": 10.373, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.029985007496251874, | |
| "grad_norm": 0.2346523553133011, | |
| "learning_rate": 2.5e-05, | |
| "loss": 10.3702, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03118440779610195, | |
| "grad_norm": 0.11483809351921082, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 10.3731, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.032383808095952024, | |
| "grad_norm": 0.12950515747070312, | |
| "learning_rate": 2.7000000000000002e-05, | |
| "loss": 10.3734, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.033583208395802096, | |
| "grad_norm": 0.13586033880710602, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 10.3725, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.034782608695652174, | |
| "grad_norm": 0.14205914735794067, | |
| "learning_rate": 2.9e-05, | |
| "loss": 10.375, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.035982008995502246, | |
| "grad_norm": 0.14829206466674805, | |
| "learning_rate": 3e-05, | |
| "loss": 10.3727, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.037181409295352325, | |
| "grad_norm": 0.15150436758995056, | |
| "learning_rate": 3.1e-05, | |
| "loss": 10.3721, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.038380809595202396, | |
| "grad_norm": 0.1555749624967575, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 10.3742, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.039580209895052475, | |
| "grad_norm": 0.1620868444442749, | |
| "learning_rate": 3.3e-05, | |
| "loss": 10.3729, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.040779610194902546, | |
| "grad_norm": 0.1725643426179886, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 10.3718, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.041979010494752625, | |
| "grad_norm": 0.18715962767601013, | |
| "learning_rate": 3.5e-05, | |
| "loss": 10.371, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0431784107946027, | |
| "grad_norm": 0.19183875620365143, | |
| "learning_rate": 3.6e-05, | |
| "loss": 10.3718, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.044377811094452775, | |
| "grad_norm": 0.21521849930286407, | |
| "learning_rate": 3.7e-05, | |
| "loss": 10.3719, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.04557721139430285, | |
| "grad_norm": 0.1310717910528183, | |
| "learning_rate": 3.8e-05, | |
| "loss": 10.3731, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.046776611694152925, | |
| "grad_norm": 0.12645111978054047, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "loss": 10.3728, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.047976011994003, | |
| "grad_norm": 0.1375029981136322, | |
| "learning_rate": 4e-05, | |
| "loss": 10.3719, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.049175412293853075, | |
| "grad_norm": 0.1415010392665863, | |
| "learning_rate": 4.1e-05, | |
| "loss": 10.3723, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.05037481259370315, | |
| "grad_norm": 0.1474965512752533, | |
| "learning_rate": 4.2e-05, | |
| "loss": 10.3708, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.051574212893553226, | |
| "grad_norm": 0.1504737138748169, | |
| "learning_rate": 4.3e-05, | |
| "loss": 10.3718, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.0527736131934033, | |
| "grad_norm": 0.15808707475662231, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 10.3712, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.053973013493253376, | |
| "grad_norm": 0.16332747042179108, | |
| "learning_rate": 4.5e-05, | |
| "loss": 10.3695, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05517241379310345, | |
| "grad_norm": 0.17212961614131927, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 10.3692, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.056371814092953526, | |
| "grad_norm": 0.17262108623981476, | |
| "learning_rate": 4.7e-05, | |
| "loss": 10.3705, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.0575712143928036, | |
| "grad_norm": 0.1849043071269989, | |
| "learning_rate": 4.8e-05, | |
| "loss": 10.3671, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.058770614692653676, | |
| "grad_norm": 0.19811047613620758, | |
| "learning_rate": 4.9e-05, | |
| "loss": 10.3671, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.05997001499250375, | |
| "grad_norm": 0.2461751103401184, | |
| "learning_rate": 5e-05, | |
| "loss": 10.3667, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.061169415292353826, | |
| "grad_norm": 0.12041845172643661, | |
| "learning_rate": 5.1000000000000006e-05, | |
| "loss": 10.371, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.0623688155922039, | |
| "grad_norm": 0.12933555245399475, | |
| "learning_rate": 5.2000000000000004e-05, | |
| "loss": 10.3694, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.06356821589205397, | |
| "grad_norm": 0.14095033705234528, | |
| "learning_rate": 5.300000000000001e-05, | |
| "loss": 10.3693, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.06476761619190405, | |
| "grad_norm": 0.14537349343299866, | |
| "learning_rate": 5.4000000000000005e-05, | |
| "loss": 10.3671, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.06596701649175413, | |
| "grad_norm": 0.1486896276473999, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 10.3671, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.06716641679160419, | |
| "grad_norm": 0.15299569070339203, | |
| "learning_rate": 5.6000000000000006e-05, | |
| "loss": 10.3668, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.06836581709145427, | |
| "grad_norm": 0.16295485198497772, | |
| "learning_rate": 5.6999999999999996e-05, | |
| "loss": 10.3653, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.06956521739130435, | |
| "grad_norm": 0.16358605027198792, | |
| "learning_rate": 5.8e-05, | |
| "loss": 10.3661, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.07076461769115443, | |
| "grad_norm": 0.17602834105491638, | |
| "learning_rate": 5.9e-05, | |
| "loss": 10.3633, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.07196401799100449, | |
| "grad_norm": 0.18307778239250183, | |
| "learning_rate": 6e-05, | |
| "loss": 10.3636, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07316341829085457, | |
| "grad_norm": 0.19053678214550018, | |
| "learning_rate": 6.1e-05, | |
| "loss": 10.3633, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.07436281859070465, | |
| "grad_norm": 0.20003724098205566, | |
| "learning_rate": 6.2e-05, | |
| "loss": 10.3622, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.07556221889055473, | |
| "grad_norm": 0.15672719478607178, | |
| "learning_rate": 6.3e-05, | |
| "loss": 10.3644, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.07676161919040479, | |
| "grad_norm": 0.14309260249137878, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 10.3649, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.07796101949025487, | |
| "grad_norm": 0.1525091975927353, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 10.3646, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07916041979010495, | |
| "grad_norm": 0.15806975960731506, | |
| "learning_rate": 6.6e-05, | |
| "loss": 10.3623, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.08035982008995503, | |
| "grad_norm": 0.16560155153274536, | |
| "learning_rate": 6.7e-05, | |
| "loss": 10.3609, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.08155922038980509, | |
| "grad_norm": 0.16659671068191528, | |
| "learning_rate": 6.800000000000001e-05, | |
| "loss": 10.3601, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.08275862068965517, | |
| "grad_norm": 0.18391086161136627, | |
| "learning_rate": 6.9e-05, | |
| "loss": 10.3617, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.08395802098950525, | |
| "grad_norm": 0.1868380904197693, | |
| "learning_rate": 7e-05, | |
| "loss": 10.3576, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08515742128935533, | |
| "grad_norm": 0.20636723935604095, | |
| "learning_rate": 7.1e-05, | |
| "loss": 10.357, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.0863568215892054, | |
| "grad_norm": 0.2090313583612442, | |
| "learning_rate": 7.2e-05, | |
| "loss": 10.3535, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.08755622188905547, | |
| "grad_norm": 0.23220857977867126, | |
| "learning_rate": 7.3e-05, | |
| "loss": 10.355, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.08875562218890555, | |
| "grad_norm": 0.23836413025856018, | |
| "learning_rate": 7.4e-05, | |
| "loss": 10.3502, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.08995502248875563, | |
| "grad_norm": 0.2619498670101166, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 10.3485, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.0911544227886057, | |
| "grad_norm": 0.20814213156700134, | |
| "learning_rate": 7.6e-05, | |
| "loss": 10.3572, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.09235382308845577, | |
| "grad_norm": 0.2424175888299942, | |
| "learning_rate": 7.7e-05, | |
| "loss": 10.3524, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.09355322338830585, | |
| "grad_norm": 0.2402586191892624, | |
| "learning_rate": 7.800000000000001e-05, | |
| "loss": 10.3501, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.09475262368815592, | |
| "grad_norm": 0.27135393023490906, | |
| "learning_rate": 7.900000000000001e-05, | |
| "loss": 10.3486, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.095952023988006, | |
| "grad_norm": 0.279787540435791, | |
| "learning_rate": 8e-05, | |
| "loss": 10.3447, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09715142428785607, | |
| "grad_norm": 0.28797608613967896, | |
| "learning_rate": 8.1e-05, | |
| "loss": 10.3438, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.09835082458770615, | |
| "grad_norm": 0.3241998851299286, | |
| "learning_rate": 8.2e-05, | |
| "loss": 10.3392, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.09955022488755622, | |
| "grad_norm": 0.3485376238822937, | |
| "learning_rate": 8.3e-05, | |
| "loss": 10.3364, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.1007496251874063, | |
| "grad_norm": 0.3528934419155121, | |
| "learning_rate": 8.4e-05, | |
| "loss": 10.3339, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.10194902548725637, | |
| "grad_norm": 0.361213743686676, | |
| "learning_rate": 8.5e-05, | |
| "loss": 10.3271, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.10314842578710645, | |
| "grad_norm": 0.3762340843677521, | |
| "learning_rate": 8.6e-05, | |
| "loss": 10.3242, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.10434782608695652, | |
| "grad_norm": 0.3962249159812927, | |
| "learning_rate": 8.7e-05, | |
| "loss": 10.3198, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.1055472263868066, | |
| "grad_norm": 0.4154474139213562, | |
| "learning_rate": 8.800000000000001e-05, | |
| "loss": 10.3241, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.10674662668665667, | |
| "grad_norm": 0.42189449071884155, | |
| "learning_rate": 8.900000000000001e-05, | |
| "loss": 10.319, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.10794602698650675, | |
| "grad_norm": 0.3983931541442871, | |
| "learning_rate": 9e-05, | |
| "loss": 10.3142, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.10914542728635682, | |
| "grad_norm": 0.39982685446739197, | |
| "learning_rate": 9.1e-05, | |
| "loss": 10.3076, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.1103448275862069, | |
| "grad_norm": 0.3935539424419403, | |
| "learning_rate": 9.200000000000001e-05, | |
| "loss": 10.3047, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.11154422788605697, | |
| "grad_norm": 0.3751447796821594, | |
| "learning_rate": 9.300000000000001e-05, | |
| "loss": 10.2953, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.11274362818590705, | |
| "grad_norm": 0.3766322135925293, | |
| "learning_rate": 9.4e-05, | |
| "loss": 10.2932, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.11394302848575712, | |
| "grad_norm": 0.3571270704269409, | |
| "learning_rate": 9.5e-05, | |
| "loss": 10.2875, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.1151424287856072, | |
| "grad_norm": 0.34838995337486267, | |
| "learning_rate": 9.6e-05, | |
| "loss": 10.2812, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.11634182908545727, | |
| "grad_norm": 0.33183571696281433, | |
| "learning_rate": 9.7e-05, | |
| "loss": 10.276, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.11754122938530735, | |
| "grad_norm": 0.3224335312843323, | |
| "learning_rate": 9.8e-05, | |
| "loss": 10.2705, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.11874062968515742, | |
| "grad_norm": 0.33488717675209045, | |
| "learning_rate": 9.900000000000001e-05, | |
| "loss": 10.2681, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.1199400299850075, | |
| "grad_norm": 0.3330170810222626, | |
| "learning_rate": 0.0001, | |
| "loss": 10.2595, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12113943028485757, | |
| "grad_norm": 0.29305723309516907, | |
| "learning_rate": 9.999954076906038e-05, | |
| "loss": 10.2593, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.12233883058470765, | |
| "grad_norm": 0.2937624454498291, | |
| "learning_rate": 9.999816308467719e-05, | |
| "loss": 10.2547, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.12353823088455772, | |
| "grad_norm": 0.29269203543663025, | |
| "learning_rate": 9.999586697215748e-05, | |
| "loss": 10.2482, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.1247376311844078, | |
| "grad_norm": 0.2853996157646179, | |
| "learning_rate": 9.999265247367908e-05, | |
| "loss": 10.2453, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.12593703148425786, | |
| "grad_norm": 0.28077051043510437, | |
| "learning_rate": 9.998851964828986e-05, | |
| "loss": 10.2412, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.12713643178410794, | |
| "grad_norm": 0.27728626132011414, | |
| "learning_rate": 9.99834685719067e-05, | |
| "loss": 10.2357, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.12833583208395802, | |
| "grad_norm": 0.28931453824043274, | |
| "learning_rate": 9.997749933731398e-05, | |
| "loss": 10.2308, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.1295352323838081, | |
| "grad_norm": 0.3001117408275604, | |
| "learning_rate": 9.997061205416203e-05, | |
| "loss": 10.2248, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.13073463268365818, | |
| "grad_norm": 0.33052486181259155, | |
| "learning_rate": 9.996280684896495e-05, | |
| "loss": 10.2211, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.13193403298350825, | |
| "grad_norm": 0.3081456124782562, | |
| "learning_rate": 9.995408386509846e-05, | |
| "loss": 10.2144, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.13313343328335833, | |
| "grad_norm": 0.29855111241340637, | |
| "learning_rate": 9.994444326279708e-05, | |
| "loss": 10.2106, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.13433283358320838, | |
| "grad_norm": 0.31099098920822144, | |
| "learning_rate": 9.993388521915134e-05, | |
| "loss": 10.208, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.13553223388305846, | |
| "grad_norm": 0.2797197103500366, | |
| "learning_rate": 9.992240992810444e-05, | |
| "loss": 10.2032, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.13673163418290854, | |
| "grad_norm": 0.2696700692176819, | |
| "learning_rate": 9.991001760044875e-05, | |
| "loss": 10.1995, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.13793103448275862, | |
| "grad_norm": 0.2778705954551697, | |
| "learning_rate": 9.989670846382188e-05, | |
| "loss": 10.1948, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1391304347826087, | |
| "grad_norm": 0.28072160482406616, | |
| "learning_rate": 9.988248276270248e-05, | |
| "loss": 10.1918, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.14032983508245878, | |
| "grad_norm": 0.2873566150665283, | |
| "learning_rate": 9.98673407584059e-05, | |
| "loss": 10.1859, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.14152923538230885, | |
| "grad_norm": 0.2754334807395935, | |
| "learning_rate": 9.985128272907918e-05, | |
| "loss": 10.1841, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.14272863568215893, | |
| "grad_norm": 0.27839502692222595, | |
| "learning_rate": 9.983430896969605e-05, | |
| "loss": 10.1797, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.14392803598200898, | |
| "grad_norm": 0.28016844391822815, | |
| "learning_rate": 9.981641979205158e-05, | |
| "loss": 10.1762, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.14512743628185906, | |
| "grad_norm": 0.2875809669494629, | |
| "learning_rate": 9.979761552475628e-05, | |
| "loss": 10.169, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.14632683658170914, | |
| "grad_norm": 0.29734012484550476, | |
| "learning_rate": 9.977789651323023e-05, | |
| "loss": 10.1667, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.14752623688155922, | |
| "grad_norm": 0.34327661991119385, | |
| "learning_rate": 9.975726311969664e-05, | |
| "loss": 10.1617, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.1487256371814093, | |
| "grad_norm": 0.32941052317619324, | |
| "learning_rate": 9.973571572317519e-05, | |
| "loss": 10.1619, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.14992503748125938, | |
| "grad_norm": 0.3153591752052307, | |
| "learning_rate": 9.971325471947517e-05, | |
| "loss": 10.1544, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.15112443778110946, | |
| "grad_norm": 0.26798173785209656, | |
| "learning_rate": 9.968988052118804e-05, | |
| "loss": 10.1517, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.15232383808095953, | |
| "grad_norm": 0.2733558714389801, | |
| "learning_rate": 9.966559355768005e-05, | |
| "loss": 10.1469, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.15352323838080958, | |
| "grad_norm": 0.275776207447052, | |
| "learning_rate": 9.964039427508418e-05, | |
| "loss": 10.1425, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.15472263868065966, | |
| "grad_norm": 0.2864780128002167, | |
| "learning_rate": 9.961428313629203e-05, | |
| "loss": 10.1389, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.15592203898050974, | |
| "grad_norm": 0.2879246473312378, | |
| "learning_rate": 9.958726062094534e-05, | |
| "loss": 10.1327, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.15712143928035982, | |
| "grad_norm": 0.28718408942222595, | |
| "learning_rate": 9.955932722542708e-05, | |
| "loss": 10.1328, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.1583208395802099, | |
| "grad_norm": 0.2976157069206238, | |
| "learning_rate": 9.953048346285245e-05, | |
| "loss": 10.1255, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.15952023988005998, | |
| "grad_norm": 0.2953495681285858, | |
| "learning_rate": 9.950072986305939e-05, | |
| "loss": 10.1214, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.16071964017991006, | |
| "grad_norm": 0.2883056700229645, | |
| "learning_rate": 9.947006697259882e-05, | |
| "loss": 10.1202, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.1619190404797601, | |
| "grad_norm": 0.3668458163738251, | |
| "learning_rate": 9.943849535472467e-05, | |
| "loss": 10.1184, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.16311844077961019, | |
| "grad_norm": 0.3660871982574463, | |
| "learning_rate": 9.940601558938348e-05, | |
| "loss": 10.1143, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.16431784107946026, | |
| "grad_norm": 0.32822495698928833, | |
| "learning_rate": 9.937262827320379e-05, | |
| "loss": 10.1094, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.16551724137931034, | |
| "grad_norm": 0.2712533175945282, | |
| "learning_rate": 9.933833401948513e-05, | |
| "loss": 10.1083, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.16671664167916042, | |
| "grad_norm": 0.27984705567359924, | |
| "learning_rate": 9.930313345818682e-05, | |
| "loss": 10.1001, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.1679160419790105, | |
| "grad_norm": 0.2825195789337158, | |
| "learning_rate": 9.92670272359163e-05, | |
| "loss": 10.0952, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.16911544227886058, | |
| "grad_norm": 0.27770835161209106, | |
| "learning_rate": 9.923001601591738e-05, | |
| "loss": 10.0949, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.17031484257871066, | |
| "grad_norm": 0.2809567451477051, | |
| "learning_rate": 9.919210047805792e-05, | |
| "loss": 10.0913, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.1715142428785607, | |
| "grad_norm": 0.2936723530292511, | |
| "learning_rate": 9.915328131881745e-05, | |
| "loss": 10.0867, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.1727136431784108, | |
| "grad_norm": 0.29460567235946655, | |
| "learning_rate": 9.911355925127433e-05, | |
| "loss": 10.0855, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 0.2977669835090637, | |
| "learning_rate": 9.907293500509268e-05, | |
| "loss": 10.08, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.17511244377811094, | |
| "grad_norm": 0.3061806559562683, | |
| "learning_rate": 9.903140932650891e-05, | |
| "loss": 10.0707, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.17631184407796102, | |
| "grad_norm": 0.2858673632144928, | |
| "learning_rate": 9.898898297831807e-05, | |
| "loss": 10.072, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.1775112443778111, | |
| "grad_norm": 0.3454626202583313, | |
| "learning_rate": 9.894565673985985e-05, | |
| "loss": 10.071, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.17871064467766118, | |
| "grad_norm": 0.5128405690193176, | |
| "learning_rate": 9.890143140700419e-05, | |
| "loss": 10.0703, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.17991004497751126, | |
| "grad_norm": 0.5272448658943176, | |
| "learning_rate": 9.885630779213677e-05, | |
| "loss": 10.0602, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1811094452773613, | |
| "grad_norm": 0.27561256289482117, | |
| "learning_rate": 9.881028672414396e-05, | |
| "loss": 10.0577, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.1823088455772114, | |
| "grad_norm": 0.2745487093925476, | |
| "learning_rate": 9.876336904839772e-05, | |
| "loss": 10.0557, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.18350824587706147, | |
| "grad_norm": 0.2808258831501007, | |
| "learning_rate": 9.871555562673995e-05, | |
| "loss": 10.05, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.18470764617691154, | |
| "grad_norm": 0.28502652049064636, | |
| "learning_rate": 9.866684733746679e-05, | |
| "loss": 10.0468, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.18590704647676162, | |
| "grad_norm": 0.28735095262527466, | |
| "learning_rate": 9.861724507531233e-05, | |
| "loss": 10.0436, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.1871064467766117, | |
| "grad_norm": 0.2824231684207916, | |
| "learning_rate": 9.856674975143236e-05, | |
| "loss": 10.0452, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.18830584707646178, | |
| "grad_norm": 0.2854205369949341, | |
| "learning_rate": 9.851536229338747e-05, | |
| "loss": 10.0425, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.18950524737631183, | |
| "grad_norm": 0.3004373610019684, | |
| "learning_rate": 9.846308364512606e-05, | |
| "loss": 10.0368, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.1907046476761619, | |
| "grad_norm": 0.3045557141304016, | |
| "learning_rate": 9.840991476696706e-05, | |
| "loss": 10.0328, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.191904047976012, | |
| "grad_norm": 0.2962295413017273, | |
| "learning_rate": 9.835585663558221e-05, | |
| "loss": 10.0301, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.19310344827586207, | |
| "grad_norm": 0.31199753284454346, | |
| "learning_rate": 9.830091024397818e-05, | |
| "loss": 10.0286, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.19430284857571214, | |
| "grad_norm": 0.33795246481895447, | |
| "learning_rate": 9.82450766014783e-05, | |
| "loss": 10.0224, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.19550224887556222, | |
| "grad_norm": 0.28454065322875977, | |
| "learning_rate": 9.818835673370401e-05, | |
| "loss": 10.0172, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.1967016491754123, | |
| "grad_norm": 0.2783207595348358, | |
| "learning_rate": 9.813075168255601e-05, | |
| "loss": 10.0135, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.19790104947526238, | |
| "grad_norm": 0.2809619605541229, | |
| "learning_rate": 9.807226250619521e-05, | |
| "loss": 10.0112, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.19910044977511243, | |
| "grad_norm": 0.2793205976486206, | |
| "learning_rate": 9.801289027902316e-05, | |
| "loss": 10.0068, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.2002998500749625, | |
| "grad_norm": 0.2848748564720154, | |
| "learning_rate": 9.795263609166243e-05, | |
| "loss": 10.0044, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.2014992503748126, | |
| "grad_norm": 0.3925122618675232, | |
| "learning_rate": 9.789150105093647e-05, | |
| "loss": 10.0025, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.20269865067466267, | |
| "grad_norm": 0.48036912083625793, | |
| "learning_rate": 9.78294862798494e-05, | |
| "loss": 9.9991, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.20389805097451275, | |
| "grad_norm": 0.32086822390556335, | |
| "learning_rate": 9.776659291756528e-05, | |
| "loss": 9.9968, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.20509745127436282, | |
| "grad_norm": 0.28707683086395264, | |
| "learning_rate": 9.770282211938721e-05, | |
| "loss": 9.9895, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.2062968515742129, | |
| "grad_norm": 0.2987452745437622, | |
| "learning_rate": 9.763817505673613e-05, | |
| "loss": 9.9897, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.20749625187406298, | |
| "grad_norm": 0.3029066324234009, | |
| "learning_rate": 9.75726529171293e-05, | |
| "loss": 9.9879, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.20869565217391303, | |
| "grad_norm": 0.321458637714386, | |
| "learning_rate": 9.750625690415848e-05, | |
| "loss": 9.9815, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.2098950524737631, | |
| "grad_norm": 0.35623157024383545, | |
| "learning_rate": 9.74389882374678e-05, | |
| "loss": 9.9831, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2110944527736132, | |
| "grad_norm": 0.27146583795547485, | |
| "learning_rate": 9.737084815273137e-05, | |
| "loss": 9.9741, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.21229385307346327, | |
| "grad_norm": 0.2866266071796417, | |
| "learning_rate": 9.730183790163062e-05, | |
| "loss": 9.9692, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.21349325337331335, | |
| "grad_norm": 0.28268909454345703, | |
| "learning_rate": 9.72319587518312e-05, | |
| "loss": 9.9681, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.21469265367316342, | |
| "grad_norm": 0.2824539244174957, | |
| "learning_rate": 9.716121198695986e-05, | |
| "loss": 9.9671, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.2158920539730135, | |
| "grad_norm": 0.2851243317127228, | |
| "learning_rate": 9.708959890658073e-05, | |
| "loss": 9.9606, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.21709145427286355, | |
| "grad_norm": 0.28162598609924316, | |
| "learning_rate": 9.701712082617149e-05, | |
| "loss": 9.9617, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.21829085457271363, | |
| "grad_norm": 0.28222355246543884, | |
| "learning_rate": 9.69437790770992e-05, | |
| "loss": 9.9595, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.2194902548725637, | |
| "grad_norm": 0.32921475172042847, | |
| "learning_rate": 9.68695750065959e-05, | |
| "loss": 9.9553, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.2206896551724138, | |
| "grad_norm": 0.4057653248310089, | |
| "learning_rate": 9.679450997773378e-05, | |
| "loss": 9.9576, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.22188905547226387, | |
| "grad_norm": 0.37460216879844666, | |
| "learning_rate": 9.67185853694002e-05, | |
| "loss": 9.9495, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.22308845577211395, | |
| "grad_norm": 0.28542500734329224, | |
| "learning_rate": 9.66418025762723e-05, | |
| "loss": 9.9535, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.22428785607196403, | |
| "grad_norm": 0.3168298304080963, | |
| "learning_rate": 9.656416300879148e-05, | |
| "loss": 9.9461, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.2254872563718141, | |
| "grad_norm": 0.2682456076145172, | |
| "learning_rate": 9.648566809313738e-05, | |
| "loss": 9.941, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.22668665667166416, | |
| "grad_norm": 0.27266478538513184, | |
| "learning_rate": 9.640631927120177e-05, | |
| "loss": 9.9355, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.22788605697151423, | |
| "grad_norm": 0.2777270972728729, | |
| "learning_rate": 9.632611800056201e-05, | |
| "loss": 9.9321, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2290854572713643, | |
| "grad_norm": 0.2846197485923767, | |
| "learning_rate": 9.624506575445429e-05, | |
| "loss": 9.93, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.2302848575712144, | |
| "grad_norm": 0.28478533029556274, | |
| "learning_rate": 9.616316402174656e-05, | |
| "loss": 9.9284, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.23148425787106447, | |
| "grad_norm": 0.2874702215194702, | |
| "learning_rate": 9.608041430691126e-05, | |
| "loss": 9.9276, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.23268365817091455, | |
| "grad_norm": 0.29689356684684753, | |
| "learning_rate": 9.59968181299975e-05, | |
| "loss": 9.9224, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.23388305847076463, | |
| "grad_norm": 0.2949802577495575, | |
| "learning_rate": 9.591237702660335e-05, | |
| "loss": 9.9178, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2350824587706147, | |
| "grad_norm": 0.29631638526916504, | |
| "learning_rate": 9.582709254784748e-05, | |
| "loss": 9.9202, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.23628185907046476, | |
| "grad_norm": 0.29446399211883545, | |
| "learning_rate": 9.574096626034077e-05, | |
| "loss": 9.9169, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.23748125937031483, | |
| "grad_norm": 0.29663321375846863, | |
| "learning_rate": 9.565399974615743e-05, | |
| "loss": 9.9164, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.2386806596701649, | |
| "grad_norm": 0.4233105182647705, | |
| "learning_rate": 9.556619460280605e-05, | |
| "loss": 9.9167, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.239880059970015, | |
| "grad_norm": 0.902298092842102, | |
| "learning_rate": 9.547755244320012e-05, | |
| "loss": 9.9114, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.24107946026986507, | |
| "grad_norm": 0.27147176861763, | |
| "learning_rate": 9.538807489562859e-05, | |
| "loss": 9.9017, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.24227886056971515, | |
| "grad_norm": 0.2803528308868408, | |
| "learning_rate": 9.529776360372575e-05, | |
| "loss": 9.8995, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.24347826086956523, | |
| "grad_norm": 0.26581478118896484, | |
| "learning_rate": 9.520662022644119e-05, | |
| "loss": 9.9054, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.2446776611694153, | |
| "grad_norm": 0.27326396107673645, | |
| "learning_rate": 9.511464643800925e-05, | |
| "loss": 9.8952, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.24587706146926536, | |
| "grad_norm": 0.277313768863678, | |
| "learning_rate": 9.502184392791834e-05, | |
| "loss": 9.8951, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.24707646176911544, | |
| "grad_norm": 0.27905702590942383, | |
| "learning_rate": 9.492821440087976e-05, | |
| "loss": 9.8936, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.2482758620689655, | |
| "grad_norm": 0.28943029046058655, | |
| "learning_rate": 9.48337595767966e-05, | |
| "loss": 9.8861, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.2494752623688156, | |
| "grad_norm": 0.2902354896068573, | |
| "learning_rate": 9.473848119073189e-05, | |
| "loss": 9.8864, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.25067466266866567, | |
| "grad_norm": 0.2864895462989807, | |
| "learning_rate": 9.4642380992877e-05, | |
| "loss": 9.8891, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.2518740629685157, | |
| "grad_norm": 0.29123008251190186, | |
| "learning_rate": 9.454546074851926e-05, | |
| "loss": 9.8855, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.25307346326836583, | |
| "grad_norm": 0.30264273285865784, | |
| "learning_rate": 9.44477222380097e-05, | |
| "loss": 9.88, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.2542728635682159, | |
| "grad_norm": 0.3108195662498474, | |
| "learning_rate": 9.434916725673024e-05, | |
| "loss": 9.8845, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.255472263868066, | |
| "grad_norm": 0.27868786454200745, | |
| "learning_rate": 9.42497976150607e-05, | |
| "loss": 9.8742, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.25667166416791604, | |
| "grad_norm": 0.27031615376472473, | |
| "learning_rate": 9.414961513834568e-05, | |
| "loss": 9.8714, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.25787106446776614, | |
| "grad_norm": 0.2734402120113373, | |
| "learning_rate": 9.404862166686088e-05, | |
| "loss": 9.8673, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.2590704647676162, | |
| "grad_norm": 0.28026947379112244, | |
| "learning_rate": 9.394681905577937e-05, | |
| "loss": 9.8689, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.26026986506746624, | |
| "grad_norm": 0.2765568196773529, | |
| "learning_rate": 9.384420917513752e-05, | |
| "loss": 9.871, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.26146926536731635, | |
| "grad_norm": 0.28846895694732666, | |
| "learning_rate": 9.374079390980058e-05, | |
| "loss": 9.8626, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.2626686656671664, | |
| "grad_norm": 0.28785768151283264, | |
| "learning_rate": 9.363657515942814e-05, | |
| "loss": 9.8594, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.2638680659670165, | |
| "grad_norm": 0.28602316975593567, | |
| "learning_rate": 9.353155483843919e-05, | |
| "loss": 9.8568, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.26506746626686656, | |
| "grad_norm": 0.2956307530403137, | |
| "learning_rate": 9.342573487597696e-05, | |
| "loss": 9.8599, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.26626686656671666, | |
| "grad_norm": 0.28586798906326294, | |
| "learning_rate": 9.331911721587345e-05, | |
| "loss": 9.8601, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.2674662668665667, | |
| "grad_norm": 0.30228516459465027, | |
| "learning_rate": 9.321170381661383e-05, | |
| "loss": 9.8549, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.26866566716641677, | |
| "grad_norm": 0.3037481904029846, | |
| "learning_rate": 9.310349665130035e-05, | |
| "loss": 9.8593, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.2698650674662669, | |
| "grad_norm": 0.3253049850463867, | |
| "learning_rate": 9.299449770761611e-05, | |
| "loss": 9.8551, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.2710644677661169, | |
| "grad_norm": 0.26533886790275574, | |
| "learning_rate": 9.288470898778863e-05, | |
| "loss": 9.8453, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.27226386806596703, | |
| "grad_norm": 0.2740223705768585, | |
| "learning_rate": 9.277413250855296e-05, | |
| "loss": 9.8406, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.2734632683658171, | |
| "grad_norm": 0.27240288257598877, | |
| "learning_rate": 9.266277030111474e-05, | |
| "loss": 9.8468, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.2746626686656672, | |
| "grad_norm": 0.33709290623664856, | |
| "learning_rate": 9.255062441111281e-05, | |
| "loss": 9.837, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 0.44585081934928894, | |
| "learning_rate": 9.243769689858166e-05, | |
| "loss": 9.8394, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.27706146926536734, | |
| "grad_norm": 0.3157913088798523, | |
| "learning_rate": 9.232398983791361e-05, | |
| "loss": 9.8386, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.2782608695652174, | |
| "grad_norm": 0.2746964693069458, | |
| "learning_rate": 9.220950531782069e-05, | |
| "loss": 9.8347, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.27946026986506745, | |
| "grad_norm": 0.28360387682914734, | |
| "learning_rate": 9.20942454412962e-05, | |
| "loss": 9.8367, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.28065967016491755, | |
| "grad_norm": 0.2914506494998932, | |
| "learning_rate": 9.197821232557624e-05, | |
| "loss": 9.8285, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.2818590704647676, | |
| "grad_norm": 0.29733598232269287, | |
| "learning_rate": 9.186140810210065e-05, | |
| "loss": 9.8322, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2830584707646177, | |
| "grad_norm": 0.30295151472091675, | |
| "learning_rate": 9.174383491647399e-05, | |
| "loss": 9.8292, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.28425787106446776, | |
| "grad_norm": 0.32045435905456543, | |
| "learning_rate": 9.162549492842602e-05, | |
| "loss": 9.8248, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.28545727136431787, | |
| "grad_norm": 0.2680381238460541, | |
| "learning_rate": 9.150639031177211e-05, | |
| "loss": 9.8168, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.2866566716641679, | |
| "grad_norm": 0.27739235758781433, | |
| "learning_rate": 9.138652325437324e-05, | |
| "loss": 9.8155, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.28785607196401797, | |
| "grad_norm": 0.271766722202301, | |
| "learning_rate": 9.12658959580959e-05, | |
| "loss": 9.8195, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2890554722638681, | |
| "grad_norm": 0.28111475706100464, | |
| "learning_rate": 9.114451063877151e-05, | |
| "loss": 9.8112, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.2902548725637181, | |
| "grad_norm": 0.27953851222991943, | |
| "learning_rate": 9.102236952615589e-05, | |
| "loss": 9.814, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.29145427286356823, | |
| "grad_norm": 0.27629002928733826, | |
| "learning_rate": 9.08994748638881e-05, | |
| "loss": 9.8131, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.2926536731634183, | |
| "grad_norm": 0.2811156213283539, | |
| "learning_rate": 9.077582890944945e-05, | |
| "loss": 9.8045, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.2938530734632684, | |
| "grad_norm": 0.3269944489002228, | |
| "learning_rate": 9.065143393412179e-05, | |
| "loss": 9.8066, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.29505247376311844, | |
| "grad_norm": 0.3400765657424927, | |
| "learning_rate": 9.052629222294604e-05, | |
| "loss": 9.8138, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.2962518740629685, | |
| "grad_norm": 0.28989219665527344, | |
| "learning_rate": 9.040040607467999e-05, | |
| "loss": 9.8014, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.2974512743628186, | |
| "grad_norm": 0.33483076095581055, | |
| "learning_rate": 9.02737778017562e-05, | |
| "loss": 9.8082, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.29865067466266865, | |
| "grad_norm": 0.2836906909942627, | |
| "learning_rate": 9.014640973023951e-05, | |
| "loss": 9.8131, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.29985007496251875, | |
| "grad_norm": 0.3271000385284424, | |
| "learning_rate": 9.00183041997843e-05, | |
| "loss": 9.7969, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.3010494752623688, | |
| "grad_norm": 0.2659075856208801, | |
| "learning_rate": 8.988946356359146e-05, | |
| "loss": 9.7947, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.3022488755622189, | |
| "grad_norm": 0.2756604552268982, | |
| "learning_rate": 8.97598901883653e-05, | |
| "loss": 9.7903, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.30344827586206896, | |
| "grad_norm": 0.2782731354236603, | |
| "learning_rate": 8.962958645426989e-05, | |
| "loss": 9.7927, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.30464767616191907, | |
| "grad_norm": 0.28496497869491577, | |
| "learning_rate": 8.949855475488549e-05, | |
| "loss": 9.788, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.3058470764617691, | |
| "grad_norm": 0.2814723253250122, | |
| "learning_rate": 8.936679749716452e-05, | |
| "loss": 9.7867, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.30704647676161917, | |
| "grad_norm": 0.27949169278144836, | |
| "learning_rate": 8.923431710138734e-05, | |
| "loss": 9.7937, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.3082458770614693, | |
| "grad_norm": 0.2898804843425751, | |
| "learning_rate": 8.910111600111785e-05, | |
| "loss": 9.783, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.3094452773613193, | |
| "grad_norm": 0.2944411337375641, | |
| "learning_rate": 8.896719664315867e-05, | |
| "loss": 9.7809, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.31064467766116943, | |
| "grad_norm": 0.29111993312835693, | |
| "learning_rate": 8.883256148750633e-05, | |
| "loss": 9.7834, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.3118440779610195, | |
| "grad_norm": 0.3007718324661255, | |
| "learning_rate": 8.869721300730596e-05, | |
| "loss": 9.7882, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3130434782608696, | |
| "grad_norm": 0.28827887773513794, | |
| "learning_rate": 8.856115368880598e-05, | |
| "loss": 9.7902, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.31424287856071964, | |
| "grad_norm": 0.31684109568595886, | |
| "learning_rate": 8.842438603131232e-05, | |
| "loss": 9.7778, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.3154422788605697, | |
| "grad_norm": 0.32087424397468567, | |
| "learning_rate": 8.828691254714259e-05, | |
| "loss": 9.7689, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.3166416791604198, | |
| "grad_norm": 0.27183249592781067, | |
| "learning_rate": 8.814873576157987e-05, | |
| "loss": 9.7738, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.31784107946026985, | |
| "grad_norm": 0.27800437808036804, | |
| "learning_rate": 8.800985821282637e-05, | |
| "loss": 9.7711, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.31904047976011995, | |
| "grad_norm": 0.28154927492141724, | |
| "learning_rate": 8.787028245195676e-05, | |
| "loss": 9.7662, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.32023988005997, | |
| "grad_norm": 0.2830137014389038, | |
| "learning_rate": 8.773001104287137e-05, | |
| "loss": 9.767, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.3214392803598201, | |
| "grad_norm": 0.27607715129852295, | |
| "learning_rate": 8.758904656224904e-05, | |
| "loss": 9.7658, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.32263868065967016, | |
| "grad_norm": 0.2993113696575165, | |
| "learning_rate": 8.744739159949981e-05, | |
| "loss": 9.7659, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.3238380809595202, | |
| "grad_norm": 0.316902220249176, | |
| "learning_rate": 8.730504875671732e-05, | |
| "loss": 9.7573, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.3250374812593703, | |
| "grad_norm": 0.3067699670791626, | |
| "learning_rate": 8.716202064863111e-05, | |
| "loss": 9.7598, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.32623688155922037, | |
| "grad_norm": 0.3003675937652588, | |
| "learning_rate": 8.701830990255843e-05, | |
| "loss": 9.7639, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.3274362818590705, | |
| "grad_norm": 0.2981228232383728, | |
| "learning_rate": 8.687391915835616e-05, | |
| "loss": 9.7576, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.32863568215892053, | |
| "grad_norm": 0.2995956242084503, | |
| "learning_rate": 8.672885106837216e-05, | |
| "loss": 9.7714, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.32983508245877063, | |
| "grad_norm": 0.30962345004081726, | |
| "learning_rate": 8.658310829739665e-05, | |
| "loss": 9.7645, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3310344827586207, | |
| "grad_norm": 0.26187556982040405, | |
| "learning_rate": 8.643669352261321e-05, | |
| "loss": 9.7506, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.3322338830584708, | |
| "grad_norm": 0.276991605758667, | |
| "learning_rate": 8.628960943354965e-05, | |
| "loss": 9.7492, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.33343328335832084, | |
| "grad_norm": 0.2857518196105957, | |
| "learning_rate": 8.614185873202851e-05, | |
| "loss": 9.7469, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.3346326836581709, | |
| "grad_norm": 0.28504660725593567, | |
| "learning_rate": 8.599344413211755e-05, | |
| "loss": 9.7518, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.335832083958021, | |
| "grad_norm": 0.27488988637924194, | |
| "learning_rate": 8.584436836007981e-05, | |
| "loss": 9.7501, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.33703148425787105, | |
| "grad_norm": 0.29049110412597656, | |
| "learning_rate": 8.569463415432356e-05, | |
| "loss": 9.7418, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.33823088455772116, | |
| "grad_norm": 0.2897820472717285, | |
| "learning_rate": 8.554424426535201e-05, | |
| "loss": 9.7481, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.3394302848575712, | |
| "grad_norm": 0.28421247005462646, | |
| "learning_rate": 8.539320145571276e-05, | |
| "loss": 9.7456, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.3406296851574213, | |
| "grad_norm": 0.28690865635871887, | |
| "learning_rate": 8.524150849994707e-05, | |
| "loss": 9.7501, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.34182908545727136, | |
| "grad_norm": 0.2818574905395508, | |
| "learning_rate": 8.50891681845389e-05, | |
| "loss": 9.7422, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.3430284857571214, | |
| "grad_norm": 0.5660536885261536, | |
| "learning_rate": 8.493618330786365e-05, | |
| "loss": 9.7497, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.3442278860569715, | |
| "grad_norm": 0.31119802594184875, | |
| "learning_rate": 8.47825566801369e-05, | |
| "loss": 9.7429, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.3454272863568216, | |
| "grad_norm": 0.2645992636680603, | |
| "learning_rate": 8.462829112336266e-05, | |
| "loss": 9.7354, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.3466266866566717, | |
| "grad_norm": 0.27782142162323, | |
| "learning_rate": 8.44733894712816e-05, | |
| "loss": 9.7309, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.27114003896713257, | |
| "learning_rate": 8.431785456931898e-05, | |
| "loss": 9.7329, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.34902548725637184, | |
| "grad_norm": 0.27776530385017395, | |
| "learning_rate": 8.416168927453236e-05, | |
| "loss": 9.7294, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.3502248875562219, | |
| "grad_norm": 0.2819390594959259, | |
| "learning_rate": 8.400489645555914e-05, | |
| "loss": 9.7324, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.35142428785607194, | |
| "grad_norm": 0.2786363363265991, | |
| "learning_rate": 8.384747899256386e-05, | |
| "loss": 9.7327, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.35262368815592204, | |
| "grad_norm": 0.29060226678848267, | |
| "learning_rate": 8.368943977718528e-05, | |
| "loss": 9.7265, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.3538230884557721, | |
| "grad_norm": 0.28789106011390686, | |
| "learning_rate": 8.353078171248335e-05, | |
| "loss": 9.7269, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3550224887556222, | |
| "grad_norm": 0.28383123874664307, | |
| "learning_rate": 8.337150771288572e-05, | |
| "loss": 9.7357, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.35622188905547225, | |
| "grad_norm": 0.28761202096939087, | |
| "learning_rate": 8.32116207041343e-05, | |
| "loss": 9.7277, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.35742128935532236, | |
| "grad_norm": 0.29686328768730164, | |
| "learning_rate": 8.30511236232316e-05, | |
| "loss": 9.7278, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.3586206896551724, | |
| "grad_norm": 0.3019421100616455, | |
| "learning_rate": 8.289001941838659e-05, | |
| "loss": 9.7348, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.3598200899550225, | |
| "grad_norm": 0.3201374411582947, | |
| "learning_rate": 8.27283110489607e-05, | |
| "loss": 9.7275, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.36101949025487257, | |
| "grad_norm": 0.2733359932899475, | |
| "learning_rate": 8.256600148541339e-05, | |
| "loss": 9.7121, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.3622188905547226, | |
| "grad_norm": 0.2780385911464691, | |
| "learning_rate": 8.240309370924759e-05, | |
| "loss": 9.7179, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.3634182908545727, | |
| "grad_norm": 0.27753978967666626, | |
| "learning_rate": 8.223959071295493e-05, | |
| "loss": 9.7121, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.3646176911544228, | |
| "grad_norm": 0.2738651633262634, | |
| "learning_rate": 8.207549549996083e-05, | |
| "loss": 9.7152, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.3658170914542729, | |
| "grad_norm": 0.4075029790401459, | |
| "learning_rate": 8.191081108456921e-05, | |
| "loss": 9.7168, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.36701649175412293, | |
| "grad_norm": 0.35438272356987, | |
| "learning_rate": 8.174554049190725e-05, | |
| "loss": 9.7143, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.36821589205397304, | |
| "grad_norm": 0.46225133538246155, | |
| "learning_rate": 8.157968675786972e-05, | |
| "loss": 9.7133, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.3694152923538231, | |
| "grad_norm": 0.2845197319984436, | |
| "learning_rate": 8.141325292906326e-05, | |
| "loss": 9.7149, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.37061469265367314, | |
| "grad_norm": 0.29232627153396606, | |
| "learning_rate": 8.12462420627504e-05, | |
| "loss": 9.7107, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.37181409295352325, | |
| "grad_norm": 0.28868958353996277, | |
| "learning_rate": 8.107865722679347e-05, | |
| "loss": 9.7176, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3730134932533733, | |
| "grad_norm": 0.3159126341342926, | |
| "learning_rate": 8.091050149959808e-05, | |
| "loss": 9.713, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.3742128935532234, | |
| "grad_norm": 0.3219504952430725, | |
| "learning_rate": 8.074177797005678e-05, | |
| "loss": 9.7166, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.37541229385307345, | |
| "grad_norm": 0.2772824168205261, | |
| "learning_rate": 8.057248973749215e-05, | |
| "loss": 9.7027, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.37661169415292356, | |
| "grad_norm": 0.2774364948272705, | |
| "learning_rate": 8.040263991159995e-05, | |
| "loss": 9.7026, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.3778110944527736, | |
| "grad_norm": 0.2747974693775177, | |
| "learning_rate": 8.0232231612392e-05, | |
| "loss": 9.702, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.37901049475262366, | |
| "grad_norm": 0.2756046652793884, | |
| "learning_rate": 8.006126797013883e-05, | |
| "loss": 9.7022, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.38020989505247377, | |
| "grad_norm": 0.269083172082901, | |
| "learning_rate": 7.98897521253122e-05, | |
| "loss": 9.7024, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.3814092953523238, | |
| "grad_norm": 0.2777722477912903, | |
| "learning_rate": 7.97176872285274e-05, | |
| "loss": 9.7029, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.3826086956521739, | |
| "grad_norm": 0.2875417172908783, | |
| "learning_rate": 7.954507644048544e-05, | |
| "loss": 9.7008, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.383808095952024, | |
| "grad_norm": 0.29414165019989014, | |
| "learning_rate": 7.937192293191485e-05, | |
| "loss": 9.7004, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.3850074962518741, | |
| "grad_norm": 0.2859031558036804, | |
| "learning_rate": 7.919822988351357e-05, | |
| "loss": 9.7048, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.38620689655172413, | |
| "grad_norm": 0.2967624068260193, | |
| "learning_rate": 7.902400048589051e-05, | |
| "loss": 9.7018, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.38740629685157424, | |
| "grad_norm": 0.40655517578125, | |
| "learning_rate": 7.884923793950685e-05, | |
| "loss": 9.693, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.3886056971514243, | |
| "grad_norm": 0.3629460632801056, | |
| "learning_rate": 7.86739454546173e-05, | |
| "loss": 9.7021, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.38980509745127434, | |
| "grad_norm": 0.3573906421661377, | |
| "learning_rate": 7.84981262512112e-05, | |
| "loss": 9.7026, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.39100449775112445, | |
| "grad_norm": 0.2747887969017029, | |
| "learning_rate": 7.832178355895326e-05, | |
| "loss": 9.6855, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.3922038980509745, | |
| "grad_norm": 0.27436476945877075, | |
| "learning_rate": 7.814492061712428e-05, | |
| "loss": 9.6864, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.3934032983508246, | |
| "grad_norm": 0.2805567681789398, | |
| "learning_rate": 7.796754067456168e-05, | |
| "loss": 9.6899, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.39460269865067465, | |
| "grad_norm": 0.2744491696357727, | |
| "learning_rate": 7.778964698959972e-05, | |
| "loss": 9.6882, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.39580209895052476, | |
| "grad_norm": 0.2762869894504547, | |
| "learning_rate": 7.761124283000983e-05, | |
| "loss": 9.6909, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3970014992503748, | |
| "grad_norm": 0.27481362223625183, | |
| "learning_rate": 7.743233147294035e-05, | |
| "loss": 9.6929, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.39820089955022486, | |
| "grad_norm": 0.28461942076683044, | |
| "learning_rate": 7.725291620485653e-05, | |
| "loss": 9.6901, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.39940029985007497, | |
| "grad_norm": 0.2874203026294708, | |
| "learning_rate": 7.707300032148004e-05, | |
| "loss": 9.6879, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.400599700149925, | |
| "grad_norm": 0.2960827052593231, | |
| "learning_rate": 7.689258712772851e-05, | |
| "loss": 9.6883, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.4017991004497751, | |
| "grad_norm": 0.2913392186164856, | |
| "learning_rate": 7.671167993765474e-05, | |
| "loss": 9.6886, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.4029985007496252, | |
| "grad_norm": 0.2986817955970764, | |
| "learning_rate": 7.653028207438589e-05, | |
| "loss": 9.6875, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.4041979010494753, | |
| "grad_norm": 0.31125518679618835, | |
| "learning_rate": 7.634839687006242e-05, | |
| "loss": 9.693, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.40539730134932533, | |
| "grad_norm": 0.27948254346847534, | |
| "learning_rate": 7.616602766577683e-05, | |
| "loss": 9.677, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.4065967016491754, | |
| "grad_norm": 0.2667854428291321, | |
| "learning_rate": 7.59831778115124e-05, | |
| "loss": 9.6728, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.4077961019490255, | |
| "grad_norm": 0.26580169796943665, | |
| "learning_rate": 7.579985066608153e-05, | |
| "loss": 9.6734, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.40899550224887554, | |
| "grad_norm": 0.27677300572395325, | |
| "learning_rate": 7.56160495970641e-05, | |
| "loss": 9.6744, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.41019490254872565, | |
| "grad_norm": 0.28340858221054077, | |
| "learning_rate": 7.543177798074564e-05, | |
| "loss": 9.6755, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.4113943028485757, | |
| "grad_norm": 0.28086498379707336, | |
| "learning_rate": 7.52470392020552e-05, | |
| "loss": 9.6741, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.4125937031484258, | |
| "grad_norm": 0.2807992100715637, | |
| "learning_rate": 7.506183665450336e-05, | |
| "loss": 9.6789, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.41379310344827586, | |
| "grad_norm": 0.27423274517059326, | |
| "learning_rate": 7.487617374011968e-05, | |
| "loss": 9.6791, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.41499250374812596, | |
| "grad_norm": 0.2901366353034973, | |
| "learning_rate": 7.469005386939036e-05, | |
| "loss": 9.6742, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.416191904047976, | |
| "grad_norm": 0.33871832489967346, | |
| "learning_rate": 7.45034804611955e-05, | |
| "loss": 9.6731, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.41739130434782606, | |
| "grad_norm": 0.3808429539203644, | |
| "learning_rate": 7.43164569427464e-05, | |
| "loss": 9.6811, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.41859070464767617, | |
| "grad_norm": 0.37340763211250305, | |
| "learning_rate": 7.412898674952248e-05, | |
| "loss": 9.6826, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.4197901049475262, | |
| "grad_norm": 0.31507405638694763, | |
| "learning_rate": 7.394107332520828e-05, | |
| "loss": 9.6792, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4209895052473763, | |
| "grad_norm": 0.2747836410999298, | |
| "learning_rate": 7.37527201216301e-05, | |
| "loss": 9.6618, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.4221889055472264, | |
| "grad_norm": 0.26785317063331604, | |
| "learning_rate": 7.356393059869272e-05, | |
| "loss": 9.668, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.4233883058470765, | |
| "grad_norm": 0.27754732966423035, | |
| "learning_rate": 7.337470822431572e-05, | |
| "loss": 9.6617, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.42458770614692654, | |
| "grad_norm": 0.2815973460674286, | |
| "learning_rate": 7.318505647436986e-05, | |
| "loss": 9.6655, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.4257871064467766, | |
| "grad_norm": 0.27644169330596924, | |
| "learning_rate": 7.299497883261319e-05, | |
| "loss": 9.6683, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.4269865067466267, | |
| "grad_norm": 0.27770036458969116, | |
| "learning_rate": 7.28044787906271e-05, | |
| "loss": 9.6686, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.42818590704647674, | |
| "grad_norm": 0.28763288259506226, | |
| "learning_rate": 7.261355984775208e-05, | |
| "loss": 9.6643, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.42938530734632685, | |
| "grad_norm": 0.28251275420188904, | |
| "learning_rate": 7.242222551102356e-05, | |
| "loss": 9.6609, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.4305847076461769, | |
| "grad_norm": 0.29022759199142456, | |
| "learning_rate": 7.223047929510743e-05, | |
| "loss": 9.6656, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.431784107946027, | |
| "grad_norm": 0.2947325110435486, | |
| "learning_rate": 7.20383247222355e-05, | |
| "loss": 9.666, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.43298350824587706, | |
| "grad_norm": 0.29398787021636963, | |
| "learning_rate": 7.184576532214077e-05, | |
| "loss": 9.6692, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.4341829085457271, | |
| "grad_norm": 0.30600595474243164, | |
| "learning_rate": 7.16528046319926e-05, | |
| "loss": 9.6675, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.4353823088455772, | |
| "grad_norm": 0.26775041222572327, | |
| "learning_rate": 7.145944619633176e-05, | |
| "loss": 9.6627, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.43658170914542727, | |
| "grad_norm": 0.2672569155693054, | |
| "learning_rate": 7.126569356700529e-05, | |
| "loss": 9.6575, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.43778110944527737, | |
| "grad_norm": 0.2710895538330078, | |
| "learning_rate": 7.107155030310126e-05, | |
| "loss": 9.6538, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.4389805097451274, | |
| "grad_norm": 0.2715386152267456, | |
| "learning_rate": 7.087701997088345e-05, | |
| "loss": 9.6533, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.44017991004497753, | |
| "grad_norm": 0.2757709324359894, | |
| "learning_rate": 7.068210614372568e-05, | |
| "loss": 9.6559, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.4413793103448276, | |
| "grad_norm": 0.27887001633644104, | |
| "learning_rate": 7.048681240204641e-05, | |
| "loss": 9.6604, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.4425787106446777, | |
| "grad_norm": 0.2874414920806885, | |
| "learning_rate": 7.029114233324276e-05, | |
| "loss": 9.6537, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.44377811094452774, | |
| "grad_norm": 0.2877376079559326, | |
| "learning_rate": 7.009509953162471e-05, | |
| "loss": 9.6594, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.4449775112443778, | |
| "grad_norm": 0.2886502146720886, | |
| "learning_rate": 6.989868759834908e-05, | |
| "loss": 9.6522, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.4461769115442279, | |
| "grad_norm": 0.28471824526786804, | |
| "learning_rate": 6.97019101413533e-05, | |
| "loss": 9.6611, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.44737631184407795, | |
| "grad_norm": 0.2849258780479431, | |
| "learning_rate": 6.950477077528926e-05, | |
| "loss": 9.6583, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.44857571214392805, | |
| "grad_norm": 0.3675067126750946, | |
| "learning_rate": 6.93072731214568e-05, | |
| "loss": 9.6677, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.4497751124437781, | |
| "grad_norm": 0.6856014728546143, | |
| "learning_rate": 6.910942080773724e-05, | |
| "loss": 9.6579, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.4509745127436282, | |
| "grad_norm": 0.27253374457359314, | |
| "learning_rate": 6.891121746852674e-05, | |
| "loss": 9.6466, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.45217391304347826, | |
| "grad_norm": 0.26753801107406616, | |
| "learning_rate": 6.871266674466955e-05, | |
| "loss": 9.6491, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.4533733133433283, | |
| "grad_norm": 0.2716609239578247, | |
| "learning_rate": 6.851377228339106e-05, | |
| "loss": 9.6484, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.4545727136431784, | |
| "grad_norm": 0.2831350266933441, | |
| "learning_rate": 6.831453773823091e-05, | |
| "loss": 9.6464, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.45577211394302847, | |
| "grad_norm": 0.28324607014656067, | |
| "learning_rate": 6.811496676897578e-05, | |
| "loss": 9.6475, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.4569715142428786, | |
| "grad_norm": 0.27571067214012146, | |
| "learning_rate": 6.791506304159221e-05, | |
| "loss": 9.645, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.4581709145427286, | |
| "grad_norm": 0.28332218527793884, | |
| "learning_rate": 6.771483022815925e-05, | |
| "loss": 9.6559, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.45937031484257873, | |
| "grad_norm": 0.2815491855144501, | |
| "learning_rate": 6.751427200680108e-05, | |
| "loss": 9.6518, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.4605697151424288, | |
| "grad_norm": 0.28399744629859924, | |
| "learning_rate": 6.731339206161928e-05, | |
| "loss": 9.6512, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.4617691154422789, | |
| "grad_norm": 0.288397878408432, | |
| "learning_rate": 6.711219408262527e-05, | |
| "loss": 9.6452, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.46296851574212894, | |
| "grad_norm": 0.29081466794013977, | |
| "learning_rate": 6.691068176567257e-05, | |
| "loss": 9.66, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.464167916041979, | |
| "grad_norm": 0.29959914088249207, | |
| "learning_rate": 6.670885881238877e-05, | |
| "loss": 9.6601, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.4653673163418291, | |
| "grad_norm": 0.27491918206214905, | |
| "learning_rate": 6.650672893010768e-05, | |
| "loss": 9.6448, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.46656671664167915, | |
| "grad_norm": 0.2780735194683075, | |
| "learning_rate": 6.630429583180112e-05, | |
| "loss": 9.6355, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.46776611694152925, | |
| "grad_norm": 0.2666113078594208, | |
| "learning_rate": 6.610156323601075e-05, | |
| "loss": 9.6384, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.4689655172413793, | |
| "grad_norm": 0.2784653902053833, | |
| "learning_rate": 6.589853486677981e-05, | |
| "loss": 9.6384, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.4701649175412294, | |
| "grad_norm": 0.28066155314445496, | |
| "learning_rate": 6.569521445358464e-05, | |
| "loss": 9.6417, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.47136431784107946, | |
| "grad_norm": 0.27688291668891907, | |
| "learning_rate": 6.549160573126623e-05, | |
| "loss": 9.6387, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.4725637181409295, | |
| "grad_norm": 0.279243141412735, | |
| "learning_rate": 6.528771243996157e-05, | |
| "loss": 9.645, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.4737631184407796, | |
| "grad_norm": 0.2818412184715271, | |
| "learning_rate": 6.508353832503494e-05, | |
| "loss": 9.6442, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.47496251874062967, | |
| "grad_norm": 0.2805486023426056, | |
| "learning_rate": 6.48790871370092e-05, | |
| "loss": 9.6417, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.4761619190404798, | |
| "grad_norm": 0.2866521179676056, | |
| "learning_rate": 6.467436263149678e-05, | |
| "loss": 9.6496, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.4773613193403298, | |
| "grad_norm": 0.29199638962745667, | |
| "learning_rate": 6.446936856913078e-05, | |
| "loss": 9.6433, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.47856071964017993, | |
| "grad_norm": 0.29411137104034424, | |
| "learning_rate": 6.426410871549581e-05, | |
| "loss": 9.6499, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.47976011994003, | |
| "grad_norm": 0.3119426369667053, | |
| "learning_rate": 6.405858684105892e-05, | |
| "loss": 9.655, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.48095952023988003, | |
| "grad_norm": 0.2667071521282196, | |
| "learning_rate": 6.385280672110024e-05, | |
| "loss": 9.6329, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.48215892053973014, | |
| "grad_norm": 0.2752053737640381, | |
| "learning_rate": 6.364677213564365e-05, | |
| "loss": 9.6306, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.4833583208395802, | |
| "grad_norm": 0.27557632327079773, | |
| "learning_rate": 6.344048686938745e-05, | |
| "loss": 9.6324, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.4845577211394303, | |
| "grad_norm": 0.2749324142932892, | |
| "learning_rate": 6.323395471163467e-05, | |
| "loss": 9.639, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.48575712143928035, | |
| "grad_norm": 0.27776116132736206, | |
| "learning_rate": 6.30271794562236e-05, | |
| "loss": 9.6358, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.48695652173913045, | |
| "grad_norm": 0.27685850858688354, | |
| "learning_rate": 6.282016490145803e-05, | |
| "loss": 9.6354, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.4881559220389805, | |
| "grad_norm": 0.2896622121334076, | |
| "learning_rate": 6.261291485003751e-05, | |
| "loss": 9.6398, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.4893553223388306, | |
| "grad_norm": 0.34289979934692383, | |
| "learning_rate": 6.240543310898746e-05, | |
| "loss": 9.6447, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.49055472263868066, | |
| "grad_norm": 0.30854368209838867, | |
| "learning_rate": 6.219772348958927e-05, | |
| "loss": 9.6312, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.4917541229385307, | |
| "grad_norm": 0.3343330919742584, | |
| "learning_rate": 6.198978980731034e-05, | |
| "loss": 9.6383, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4929535232383808, | |
| "grad_norm": 0.2979169487953186, | |
| "learning_rate": 6.178163588173381e-05, | |
| "loss": 9.6352, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.49415292353823087, | |
| "grad_norm": 0.29163146018981934, | |
| "learning_rate": 6.157326553648862e-05, | |
| "loss": 9.6349, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.495352323838081, | |
| "grad_norm": 0.32137271761894226, | |
| "learning_rate": 6.136468259917917e-05, | |
| "loss": 9.6287, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.496551724137931, | |
| "grad_norm": 0.26861318945884705, | |
| "learning_rate": 6.115589090131497e-05, | |
| "loss": 9.6261, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.49775112443778113, | |
| "grad_norm": 0.2670891582965851, | |
| "learning_rate": 6.094689427824031e-05, | |
| "loss": 9.6272, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.4989505247376312, | |
| "grad_norm": 0.2772047519683838, | |
| "learning_rate": 6.073769656906385e-05, | |
| "loss": 9.6257, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.5001499250374812, | |
| "grad_norm": 0.27719056606292725, | |
| "learning_rate": 6.052830161658799e-05, | |
| "loss": 9.6287, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.5013493253373313, | |
| "grad_norm": 0.27757909893989563, | |
| "learning_rate": 6.031871326723837e-05, | |
| "loss": 9.6331, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.5025487256371814, | |
| "grad_norm": 0.28167879581451416, | |
| "learning_rate": 6.010893537099316e-05, | |
| "loss": 9.6289, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.5037481259370314, | |
| "grad_norm": 0.28175151348114014, | |
| "learning_rate": 5.9898971781312384e-05, | |
| "loss": 9.6342, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.5049475262368815, | |
| "grad_norm": 0.2766707241535187, | |
| "learning_rate": 5.9688826355067105e-05, | |
| "loss": 9.6337, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.5061469265367317, | |
| "grad_norm": 0.29078051447868347, | |
| "learning_rate": 5.9478502952468595e-05, | |
| "loss": 9.6292, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.5073463268365818, | |
| "grad_norm": 0.2969301640987396, | |
| "learning_rate": 5.92680054369974e-05, | |
| "loss": 9.6297, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.5085457271364318, | |
| "grad_norm": 0.29746097326278687, | |
| "learning_rate": 5.905733767533238e-05, | |
| "loss": 9.6367, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.5097451274362819, | |
| "grad_norm": 0.32283881306648254, | |
| "learning_rate": 5.8846503537279715e-05, | |
| "loss": 9.6347, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.510944527736132, | |
| "grad_norm": 0.2667362689971924, | |
| "learning_rate": 5.863550689570179e-05, | |
| "loss": 9.6198, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.512143928035982, | |
| "grad_norm": 0.274853378534317, | |
| "learning_rate": 5.842435162644601e-05, | |
| "loss": 9.6217, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.5133433283358321, | |
| "grad_norm": 0.2751438319683075, | |
| "learning_rate": 5.821304160827371e-05, | |
| "loss": 9.6246, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.5145427286356822, | |
| "grad_norm": 0.279313325881958, | |
| "learning_rate": 5.8001580722788795e-05, | |
| "loss": 9.6222, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.5157421289355323, | |
| "grad_norm": 0.27348318696022034, | |
| "learning_rate": 5.7789972854366536e-05, | |
| "loss": 9.6226, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.5169415292353823, | |
| "grad_norm": 0.2807561457157135, | |
| "learning_rate": 5.757822189008214e-05, | |
| "loss": 9.6246, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.5181409295352324, | |
| "grad_norm": 0.28267139196395874, | |
| "learning_rate": 5.7366331719639366e-05, | |
| "loss": 9.6234, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.5193403298350825, | |
| "grad_norm": 0.28372693061828613, | |
| "learning_rate": 5.715430623529909e-05, | |
| "loss": 9.6304, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.5205397301349325, | |
| "grad_norm": 0.2916060984134674, | |
| "learning_rate": 5.6942149331807836e-05, | |
| "loss": 9.6256, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 0.30614909529685974, | |
| "learning_rate": 5.6729864906326136e-05, | |
| "loss": 9.6258, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.5229385307346327, | |
| "grad_norm": 0.32992836833000183, | |
| "learning_rate": 5.651745685835707e-05, | |
| "loss": 9.6317, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.5241379310344828, | |
| "grad_norm": 0.3490801751613617, | |
| "learning_rate": 5.630492908967451e-05, | |
| "loss": 9.6334, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.5253373313343328, | |
| "grad_norm": 0.2723431885242462, | |
| "learning_rate": 5.609228550425154e-05, | |
| "loss": 9.6261, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.5265367316341829, | |
| "grad_norm": 0.26900023221969604, | |
| "learning_rate": 5.5879530008188716e-05, | |
| "loss": 9.6217, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.527736131934033, | |
| "grad_norm": 0.26944032311439514, | |
| "learning_rate": 5.566666650964228e-05, | |
| "loss": 9.6232, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.528935532233883, | |
| "grad_norm": 0.2752003073692322, | |
| "learning_rate": 5.545369891875241e-05, | |
| "loss": 9.6213, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.5301349325337331, | |
| "grad_norm": 0.27395889163017273, | |
| "learning_rate": 5.524063114757139e-05, | |
| "loss": 9.6238, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.5313343328335832, | |
| "grad_norm": 0.2798815965652466, | |
| "learning_rate": 5.5027467109991705e-05, | |
| "loss": 9.6211, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.5325337331334333, | |
| "grad_norm": 0.2802503705024719, | |
| "learning_rate": 5.481421072167423e-05, | |
| "loss": 9.6214, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.5337331334332833, | |
| "grad_norm": 0.2843863368034363, | |
| "learning_rate": 5.4600865899976225e-05, | |
| "loss": 9.6235, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.5349325337331334, | |
| "grad_norm": 0.28727102279663086, | |
| "learning_rate": 5.43874365638794e-05, | |
| "loss": 9.6231, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.5361319340329835, | |
| "grad_norm": 0.2915343940258026, | |
| "learning_rate": 5.417392663391796e-05, | |
| "loss": 9.6246, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.5373313343328335, | |
| "grad_norm": 0.295317679643631, | |
| "learning_rate": 5.3960340032106515e-05, | |
| "loss": 9.6214, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.5385307346326836, | |
| "grad_norm": 0.29819992184638977, | |
| "learning_rate": 5.374668068186809e-05, | |
| "loss": 9.6253, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.5397301349325337, | |
| "grad_norm": 0.31375864148139954, | |
| "learning_rate": 5.3532952507962066e-05, | |
| "loss": 9.6318, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5409295352323839, | |
| "grad_norm": 0.27647456526756287, | |
| "learning_rate": 5.3319159436412046e-05, | |
| "loss": 9.616, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.5421289355322338, | |
| "grad_norm": 0.27085641026496887, | |
| "learning_rate": 5.310530539443375e-05, | |
| "loss": 9.6163, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.543328335832084, | |
| "grad_norm": 0.2696115970611572, | |
| "learning_rate": 5.28913943103629e-05, | |
| "loss": 9.6192, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.5445277361319341, | |
| "grad_norm": 0.281088650226593, | |
| "learning_rate": 5.2677430113583005e-05, | |
| "loss": 9.6158, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.545727136431784, | |
| "grad_norm": 0.27259501814842224, | |
| "learning_rate": 5.246341673445323e-05, | |
| "loss": 9.6236, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.5469265367316342, | |
| "grad_norm": 0.2712749242782593, | |
| "learning_rate": 5.22493581042362e-05, | |
| "loss": 9.6263, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.5481259370314843, | |
| "grad_norm": 0.2738138437271118, | |
| "learning_rate": 5.203525815502574e-05, | |
| "loss": 9.6257, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.5493253373313344, | |
| "grad_norm": 0.2845352292060852, | |
| "learning_rate": 5.182112081967466e-05, | |
| "loss": 9.6207, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.5505247376311844, | |
| "grad_norm": 0.2933952510356903, | |
| "learning_rate": 5.160695003172259e-05, | |
| "loss": 9.6218, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.5517241379310345, | |
| "grad_norm": 0.32854798436164856, | |
| "learning_rate": 5.13927497253236e-05, | |
| "loss": 9.6258, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.5529235382308846, | |
| "grad_norm": 0.32506442070007324, | |
| "learning_rate": 5.1178523835174e-05, | |
| "loss": 9.6305, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.5541229385307347, | |
| "grad_norm": 0.31355130672454834, | |
| "learning_rate": 5.0964276296440075e-05, | |
| "loss": 9.6294, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.5553223388305847, | |
| "grad_norm": 0.27378547191619873, | |
| "learning_rate": 5.075001104468576e-05, | |
| "loss": 9.6173, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.5565217391304348, | |
| "grad_norm": 0.2689533829689026, | |
| "learning_rate": 5.053573201580039e-05, | |
| "loss": 9.6162, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.5577211394302849, | |
| "grad_norm": 0.27102944254875183, | |
| "learning_rate": 5.032144314592633e-05, | |
| "loss": 9.6134, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.5589205397301349, | |
| "grad_norm": 0.2733670473098755, | |
| "learning_rate": 5.010714837138675e-05, | |
| "loss": 9.6183, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.560119940029985, | |
| "grad_norm": 0.2813307046890259, | |
| "learning_rate": 4.989285162861326e-05, | |
| "loss": 9.6184, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.5613193403298351, | |
| "grad_norm": 0.2758241593837738, | |
| "learning_rate": 4.967855685407368e-05, | |
| "loss": 9.6172, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.5625187406296852, | |
| "grad_norm": 0.28401094675064087, | |
| "learning_rate": 4.946426798419962e-05, | |
| "loss": 9.616, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.5637181409295352, | |
| "grad_norm": 0.28821876645088196, | |
| "learning_rate": 4.924998895531425e-05, | |
| "loss": 9.6195, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5649175412293853, | |
| "grad_norm": 0.2845361828804016, | |
| "learning_rate": 4.903572370355993e-05, | |
| "loss": 9.6186, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.5661169415292354, | |
| "grad_norm": 0.29093310236930847, | |
| "learning_rate": 4.882147616482602e-05, | |
| "loss": 9.619, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.5673163418290854, | |
| "grad_norm": 0.28855013847351074, | |
| "learning_rate": 4.8607250274676415e-05, | |
| "loss": 9.6224, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.5685157421289355, | |
| "grad_norm": 0.3042353689670563, | |
| "learning_rate": 4.839304996827741e-05, | |
| "loss": 9.6186, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.5697151424287856, | |
| "grad_norm": 0.32164472341537476, | |
| "learning_rate": 4.817887918032535e-05, | |
| "loss": 9.6202, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5709145427286357, | |
| "grad_norm": 0.26481005549430847, | |
| "learning_rate": 4.7964741844974275e-05, | |
| "loss": 9.6097, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.5721139430284857, | |
| "grad_norm": 0.2751154601573944, | |
| "learning_rate": 4.775064189576381e-05, | |
| "loss": 9.6077, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.5733133433283358, | |
| "grad_norm": 0.26990050077438354, | |
| "learning_rate": 4.7536583265546775e-05, | |
| "loss": 9.609, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.5745127436281859, | |
| "grad_norm": 0.2816186845302582, | |
| "learning_rate": 4.7322569886417006e-05, | |
| "loss": 9.6101, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.5757121439280359, | |
| "grad_norm": 0.2793320417404175, | |
| "learning_rate": 4.71086056896371e-05, | |
| "loss": 9.6206, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.576911544227886, | |
| "grad_norm": 0.2865123748779297, | |
| "learning_rate": 4.689469460556626e-05, | |
| "loss": 9.6109, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.5781109445277361, | |
| "grad_norm": 0.2744526267051697, | |
| "learning_rate": 4.6680840563587966e-05, | |
| "loss": 9.6222, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.5793103448275863, | |
| "grad_norm": 0.30048128962516785, | |
| "learning_rate": 4.646704749203793e-05, | |
| "loss": 9.6182, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.5805097451274362, | |
| "grad_norm": 0.29160621762275696, | |
| "learning_rate": 4.6253319318131926e-05, | |
| "loss": 9.618, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.5817091454272864, | |
| "grad_norm": 0.31267857551574707, | |
| "learning_rate": 4.60396599678935e-05, | |
| "loss": 9.622, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5829085457271365, | |
| "grad_norm": 0.3598839044570923, | |
| "learning_rate": 4.582607336608205e-05, | |
| "loss": 9.6176, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.5841079460269865, | |
| "grad_norm": 0.33458805084228516, | |
| "learning_rate": 4.561256343612061e-05, | |
| "loss": 9.6256, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.5853073463268366, | |
| "grad_norm": 0.27461162209510803, | |
| "learning_rate": 4.539913410002378e-05, | |
| "loss": 9.6119, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.5865067466266867, | |
| "grad_norm": 0.2723887264728546, | |
| "learning_rate": 4.518578927832577e-05, | |
| "loss": 9.6056, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.5877061469265368, | |
| "grad_norm": 0.2768537998199463, | |
| "learning_rate": 4.4972532890008313e-05, | |
| "loss": 9.6079, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5889055472263868, | |
| "grad_norm": 0.2774599492549896, | |
| "learning_rate": 4.4759368852428625e-05, | |
| "loss": 9.6092, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.5901049475262369, | |
| "grad_norm": 0.27346640825271606, | |
| "learning_rate": 4.45463010812476e-05, | |
| "loss": 9.6143, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.591304347826087, | |
| "grad_norm": 0.2797171175479889, | |
| "learning_rate": 4.433333349035773e-05, | |
| "loss": 9.6168, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.592503748125937, | |
| "grad_norm": 0.2800818085670471, | |
| "learning_rate": 4.4120469991811296e-05, | |
| "loss": 9.6165, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.5937031484257871, | |
| "grad_norm": 0.280519038438797, | |
| "learning_rate": 4.390771449574846e-05, | |
| "loss": 9.6195, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5949025487256372, | |
| "grad_norm": 0.2884778678417206, | |
| "learning_rate": 4.369507091032551e-05, | |
| "loss": 9.6132, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.5961019490254873, | |
| "grad_norm": 0.2894138693809509, | |
| "learning_rate": 4.3482543141642943e-05, | |
| "loss": 9.6147, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.5973013493253373, | |
| "grad_norm": 0.2868705093860626, | |
| "learning_rate": 4.327013509367386e-05, | |
| "loss": 9.6242, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.5985007496251874, | |
| "grad_norm": 0.2994021773338318, | |
| "learning_rate": 4.305785066819218e-05, | |
| "loss": 9.6189, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.5997001499250375, | |
| "grad_norm": 0.3168644607067108, | |
| "learning_rate": 4.2845693764700914e-05, | |
| "loss": 9.6247, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6008995502248875, | |
| "grad_norm": 0.26666632294654846, | |
| "learning_rate": 4.263366828036065e-05, | |
| "loss": 9.6057, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.6020989505247376, | |
| "grad_norm": 0.26327091455459595, | |
| "learning_rate": 4.242177810991789e-05, | |
| "loss": 9.6115, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.6032983508245877, | |
| "grad_norm": 0.27538183331489563, | |
| "learning_rate": 4.221002714563347e-05, | |
| "loss": 9.6082, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.6044977511244378, | |
| "grad_norm": 0.27597832679748535, | |
| "learning_rate": 4.19984192772112e-05, | |
| "loss": 9.6075, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.6056971514242878, | |
| "grad_norm": 0.28365880250930786, | |
| "learning_rate": 4.1786958391726314e-05, | |
| "loss": 9.6136, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.6068965517241379, | |
| "grad_norm": 0.2802659273147583, | |
| "learning_rate": 4.1575648373554e-05, | |
| "loss": 9.6158, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.608095952023988, | |
| "grad_norm": 0.2841864228248596, | |
| "learning_rate": 4.136449310429822e-05, | |
| "loss": 9.6115, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.6092953523238381, | |
| "grad_norm": 0.2928536832332611, | |
| "learning_rate": 4.115349646272029e-05, | |
| "loss": 9.6156, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.6104947526236881, | |
| "grad_norm": 0.2854699492454529, | |
| "learning_rate": 4.0942662324667627e-05, | |
| "loss": 9.6137, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.6116941529235382, | |
| "grad_norm": 0.29192522168159485, | |
| "learning_rate": 4.0731994563002606e-05, | |
| "loss": 9.6136, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.6128935532233883, | |
| "grad_norm": 0.3441016674041748, | |
| "learning_rate": 4.052149704753142e-05, | |
| "loss": 9.6224, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.6140929535232383, | |
| "grad_norm": 0.3597991466522217, | |
| "learning_rate": 4.03111736449329e-05, | |
| "loss": 9.6219, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.6152923538230884, | |
| "grad_norm": 0.2781412899494171, | |
| "learning_rate": 4.010102821868762e-05, | |
| "loss": 9.6056, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.6164917541229386, | |
| "grad_norm": 0.27280357480049133, | |
| "learning_rate": 3.989106462900686e-05, | |
| "loss": 9.6063, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.6176911544227887, | |
| "grad_norm": 0.27366748452186584, | |
| "learning_rate": 3.968128673276165e-05, | |
| "loss": 9.6104, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.6188905547226387, | |
| "grad_norm": 0.27588704228401184, | |
| "learning_rate": 3.947169838341202e-05, | |
| "loss": 9.605, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.6200899550224888, | |
| "grad_norm": 0.27753859758377075, | |
| "learning_rate": 3.9262303430936164e-05, | |
| "loss": 9.6033, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.6212893553223389, | |
| "grad_norm": 0.27255064249038696, | |
| "learning_rate": 3.9053105721759696e-05, | |
| "loss": 9.6098, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.6224887556221889, | |
| "grad_norm": 0.2782951295375824, | |
| "learning_rate": 3.8844109098685045e-05, | |
| "loss": 9.6184, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.623688155922039, | |
| "grad_norm": 0.28660768270492554, | |
| "learning_rate": 3.8635317400820855e-05, | |
| "loss": 9.6113, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.6248875562218891, | |
| "grad_norm": 0.28494128584861755, | |
| "learning_rate": 3.842673446351138e-05, | |
| "loss": 9.6105, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.6260869565217392, | |
| "grad_norm": 0.28198301792144775, | |
| "learning_rate": 3.82183641182662e-05, | |
| "loss": 9.626, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.6272863568215892, | |
| "grad_norm": 0.2875995337963104, | |
| "learning_rate": 3.801021019268969e-05, | |
| "loss": 9.6176, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.6284857571214393, | |
| "grad_norm": 0.2956449091434479, | |
| "learning_rate": 3.780227651041073e-05, | |
| "loss": 9.6229, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.6296851574212894, | |
| "grad_norm": 0.37847524881362915, | |
| "learning_rate": 3.7594566891012546e-05, | |
| "loss": 9.6214, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.6308845577211394, | |
| "grad_norm": 0.27030348777770996, | |
| "learning_rate": 3.7387085149962507e-05, | |
| "loss": 9.6011, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.6320839580209895, | |
| "grad_norm": 0.274962455034256, | |
| "learning_rate": 3.717983509854198e-05, | |
| "loss": 9.6023, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.6332833583208396, | |
| "grad_norm": 0.27726373076438904, | |
| "learning_rate": 3.69728205437764e-05, | |
| "loss": 9.6102, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.6344827586206897, | |
| "grad_norm": 0.27569401264190674, | |
| "learning_rate": 3.676604528836535e-05, | |
| "loss": 9.6077, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.6356821589205397, | |
| "grad_norm": 0.2719118893146515, | |
| "learning_rate": 3.6559513130612565e-05, | |
| "loss": 9.6078, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.6368815592203898, | |
| "grad_norm": 0.27930060029029846, | |
| "learning_rate": 3.635322786435635e-05, | |
| "loss": 9.6099, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.6380809595202399, | |
| "grad_norm": 0.2761722505092621, | |
| "learning_rate": 3.614719327889978e-05, | |
| "loss": 9.6161, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.6392803598200899, | |
| "grad_norm": 0.2825543284416199, | |
| "learning_rate": 3.594141315894108e-05, | |
| "loss": 9.616, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.64047976011994, | |
| "grad_norm": 0.28519946336746216, | |
| "learning_rate": 3.573589128450418e-05, | |
| "loss": 9.6134, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.6416791604197901, | |
| "grad_norm": 0.2859567105770111, | |
| "learning_rate": 3.5530631430869234e-05, | |
| "loss": 9.6181, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.6428785607196402, | |
| "grad_norm": 0.293560653924942, | |
| "learning_rate": 3.532563736850322e-05, | |
| "loss": 9.6141, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.6440779610194902, | |
| "grad_norm": 0.31543228030204773, | |
| "learning_rate": 3.512091286299081e-05, | |
| "loss": 9.6132, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.6452773613193403, | |
| "grad_norm": 0.28361520171165466, | |
| "learning_rate": 3.491646167496507e-05, | |
| "loss": 9.5993, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.6464767616191904, | |
| "grad_norm": 0.2670563757419586, | |
| "learning_rate": 3.4712287560038446e-05, | |
| "loss": 9.6042, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.6476761619190404, | |
| "grad_norm": 0.2657446265220642, | |
| "learning_rate": 3.450839426873378e-05, | |
| "loss": 9.6106, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.6488755622188905, | |
| "grad_norm": 0.271816611289978, | |
| "learning_rate": 3.4304785546415374e-05, | |
| "loss": 9.608, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.6500749625187406, | |
| "grad_norm": 0.27191296219825745, | |
| "learning_rate": 3.41014651332202e-05, | |
| "loss": 9.6103, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.6512743628185907, | |
| "grad_norm": 0.27644070982933044, | |
| "learning_rate": 3.3898436763989247e-05, | |
| "loss": 9.6039, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.6524737631184407, | |
| "grad_norm": 0.27742430567741394, | |
| "learning_rate": 3.369570416819889e-05, | |
| "loss": 9.6053, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.6536731634182908, | |
| "grad_norm": 0.2793113589286804, | |
| "learning_rate": 3.349327106989232e-05, | |
| "loss": 9.615, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.654872563718141, | |
| "grad_norm": 0.28077057003974915, | |
| "learning_rate": 3.329114118761123e-05, | |
| "loss": 9.6101, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.656071964017991, | |
| "grad_norm": 0.2894865870475769, | |
| "learning_rate": 3.308931823432744e-05, | |
| "loss": 9.6093, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.6572713643178411, | |
| "grad_norm": 0.2894723415374756, | |
| "learning_rate": 3.288780591737474e-05, | |
| "loss": 9.6141, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.6584707646176912, | |
| "grad_norm": 0.3010658323764801, | |
| "learning_rate": 3.268660793838074e-05, | |
| "loss": 9.6249, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.6596701649175413, | |
| "grad_norm": 0.3542385399341583, | |
| "learning_rate": 3.2485727993198945e-05, | |
| "loss": 9.6182, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.6608695652173913, | |
| "grad_norm": 0.2821604907512665, | |
| "learning_rate": 3.228516977184075e-05, | |
| "loss": 9.6229, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.6620689655172414, | |
| "grad_norm": 0.27113085985183716, | |
| "learning_rate": 3.2084936958407805e-05, | |
| "loss": 9.6041, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.6632683658170915, | |
| "grad_norm": 0.26982516050338745, | |
| "learning_rate": 3.188503323102425e-05, | |
| "loss": 9.6084, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.6644677661169416, | |
| "grad_norm": 0.2756569981575012, | |
| "learning_rate": 3.1685462261769105e-05, | |
| "loss": 9.6126, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.6656671664167916, | |
| "grad_norm": 0.27629488706588745, | |
| "learning_rate": 3.1486227716608946e-05, | |
| "loss": 9.6056, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.6668665667166417, | |
| "grad_norm": 0.28036460280418396, | |
| "learning_rate": 3.128733325533047e-05, | |
| "loss": 9.6054, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.6680659670164918, | |
| "grad_norm": 0.27844056487083435, | |
| "learning_rate": 3.1088782531473266e-05, | |
| "loss": 9.6111, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.6692653673163418, | |
| "grad_norm": 0.2862386405467987, | |
| "learning_rate": 3.089057919226277e-05, | |
| "loss": 9.612, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.6704647676161919, | |
| "grad_norm": 0.2859496474266052, | |
| "learning_rate": 3.069272687854322e-05, | |
| "loss": 9.6114, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.671664167916042, | |
| "grad_norm": 0.28554123640060425, | |
| "learning_rate": 3.049522922471075e-05, | |
| "loss": 9.6105, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6728635682158921, | |
| "grad_norm": 0.30089861154556274, | |
| "learning_rate": 3.02980898586467e-05, | |
| "loss": 9.6205, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.6740629685157421, | |
| "grad_norm": 0.30331140756607056, | |
| "learning_rate": 3.0101312401650937e-05, | |
| "loss": 9.6158, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.6752623688155922, | |
| "grad_norm": 0.2732248902320862, | |
| "learning_rate": 2.9904900468375297e-05, | |
| "loss": 9.6064, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.6764617691154423, | |
| "grad_norm": 0.27510005235671997, | |
| "learning_rate": 2.9708857666757246e-05, | |
| "loss": 9.6019, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.6776611694152923, | |
| "grad_norm": 0.27365824580192566, | |
| "learning_rate": 2.9513187597953607e-05, | |
| "loss": 9.5995, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.6788605697151424, | |
| "grad_norm": 0.2792357802391052, | |
| "learning_rate": 2.931789385627433e-05, | |
| "loss": 9.606, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.6800599700149925, | |
| "grad_norm": 0.2759556770324707, | |
| "learning_rate": 2.9122980029116586e-05, | |
| "loss": 9.6039, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.6812593703148426, | |
| "grad_norm": 0.2814030647277832, | |
| "learning_rate": 2.8928449696898763e-05, | |
| "loss": 9.602, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.6824587706146926, | |
| "grad_norm": 0.2769099771976471, | |
| "learning_rate": 2.8734306432994735e-05, | |
| "loss": 9.6079, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.6836581709145427, | |
| "grad_norm": 0.2809275686740875, | |
| "learning_rate": 2.8540553803668252e-05, | |
| "loss": 9.613, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6848575712143928, | |
| "grad_norm": 0.275016725063324, | |
| "learning_rate": 2.8347195368007418e-05, | |
| "loss": 9.6097, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.6860569715142428, | |
| "grad_norm": 0.2964610755443573, | |
| "learning_rate": 2.815423467785925e-05, | |
| "loss": 9.6111, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.6872563718140929, | |
| "grad_norm": 0.2884480059146881, | |
| "learning_rate": 2.7961675277764498e-05, | |
| "loss": 9.6089, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.688455772113943, | |
| "grad_norm": 0.30310893058776855, | |
| "learning_rate": 2.7769520704892566e-05, | |
| "loss": 9.6102, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 0.4733683466911316, | |
| "learning_rate": 2.757777448897646e-05, | |
| "loss": 9.6083, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6908545727136431, | |
| "grad_norm": 0.272512823343277, | |
| "learning_rate": 2.7386440152247933e-05, | |
| "loss": 9.5963, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.6920539730134933, | |
| "grad_norm": 0.2810138165950775, | |
| "learning_rate": 2.71955212093729e-05, | |
| "loss": 9.6012, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.6932533733133434, | |
| "grad_norm": 0.2755623161792755, | |
| "learning_rate": 2.7005021167386803e-05, | |
| "loss": 9.6022, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.6944527736131934, | |
| "grad_norm": 0.2718299329280853, | |
| "learning_rate": 2.681494352563013e-05, | |
| "loss": 9.6096, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.2746315896511078, | |
| "learning_rate": 2.6625291775684292e-05, | |
| "loss": 9.6124, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6968515742128936, | |
| "grad_norm": 0.2844776511192322, | |
| "learning_rate": 2.6436069401307284e-05, | |
| "loss": 9.6054, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.6980509745127437, | |
| "grad_norm": 0.2785060703754425, | |
| "learning_rate": 2.624727987836991e-05, | |
| "loss": 9.6112, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.6992503748125937, | |
| "grad_norm": 0.2840147316455841, | |
| "learning_rate": 2.6058926674791728e-05, | |
| "loss": 9.6061, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.7004497751124438, | |
| "grad_norm": 0.28523436188697815, | |
| "learning_rate": 2.5871013250477528e-05, | |
| "loss": 9.6057, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.7016491754122939, | |
| "grad_norm": 0.29284006357192993, | |
| "learning_rate": 2.56835430572536e-05, | |
| "loss": 9.6091, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.7028485757121439, | |
| "grad_norm": 0.29574641585350037, | |
| "learning_rate": 2.5496519538804486e-05, | |
| "loss": 9.6155, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.704047976011994, | |
| "grad_norm": 0.3032572269439697, | |
| "learning_rate": 2.530994613060965e-05, | |
| "loss": 9.6162, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.7052473763118441, | |
| "grad_norm": 0.2718828320503235, | |
| "learning_rate": 2.5123826259880323e-05, | |
| "loss": 9.6001, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.7064467766116942, | |
| "grad_norm": 0.27074381709098816, | |
| "learning_rate": 2.493816334549664e-05, | |
| "loss": 9.6014, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.7076461769115442, | |
| "grad_norm": 0.2791549265384674, | |
| "learning_rate": 2.4752960797944802e-05, | |
| "loss": 9.5998, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.7088455772113943, | |
| "grad_norm": 0.28340011835098267, | |
| "learning_rate": 2.4568222019254377e-05, | |
| "loss": 9.5979, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.7100449775112444, | |
| "grad_norm": 0.2762751579284668, | |
| "learning_rate": 2.43839504029359e-05, | |
| "loss": 9.6032, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.7112443778110945, | |
| "grad_norm": 0.2753763198852539, | |
| "learning_rate": 2.4200149333918487e-05, | |
| "loss": 9.6089, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.7124437781109445, | |
| "grad_norm": 0.27482444047927856, | |
| "learning_rate": 2.4016822188487603e-05, | |
| "loss": 9.6081, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.7136431784107946, | |
| "grad_norm": 0.28210797905921936, | |
| "learning_rate": 2.383397233422318e-05, | |
| "loss": 9.6041, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.7148425787106447, | |
| "grad_norm": 0.2853706479072571, | |
| "learning_rate": 2.3651603129937592e-05, | |
| "loss": 9.6042, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.7160419790104947, | |
| "grad_norm": 0.3066234886646271, | |
| "learning_rate": 2.346971792561413e-05, | |
| "loss": 9.6053, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.7172413793103448, | |
| "grad_norm": 0.2879929542541504, | |
| "learning_rate": 2.3288320062345277e-05, | |
| "loss": 9.6069, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.7184407796101949, | |
| "grad_norm": 0.35332369804382324, | |
| "learning_rate": 2.3107412872271518e-05, | |
| "loss": 9.6162, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.719640179910045, | |
| "grad_norm": 0.5152252316474915, | |
| "learning_rate": 2.2926999678519974e-05, | |
| "loss": 9.6182, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.720839580209895, | |
| "grad_norm": 0.2663346230983734, | |
| "learning_rate": 2.274708379514348e-05, | |
| "loss": 9.5986, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.7220389805097451, | |
| "grad_norm": 0.27524423599243164, | |
| "learning_rate": 2.256766852705967e-05, | |
| "loss": 9.5986, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.7232383808095952, | |
| "grad_norm": 0.2814219295978546, | |
| "learning_rate": 2.238875716999019e-05, | |
| "loss": 9.6037, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.7244377811094452, | |
| "grad_norm": 0.2859136760234833, | |
| "learning_rate": 2.221035301040027e-05, | |
| "loss": 9.6002, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.7256371814092953, | |
| "grad_norm": 0.27460747957229614, | |
| "learning_rate": 2.2032459325438336e-05, | |
| "loss": 9.6031, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.7268365817091454, | |
| "grad_norm": 0.2745445966720581, | |
| "learning_rate": 2.185507938287572e-05, | |
| "loss": 9.6072, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.7280359820089956, | |
| "grad_norm": 0.2816024124622345, | |
| "learning_rate": 2.1678216441046734e-05, | |
| "loss": 9.6128, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.7292353823088455, | |
| "grad_norm": 0.28734058141708374, | |
| "learning_rate": 2.1501873748788802e-05, | |
| "loss": 9.6127, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.7304347826086957, | |
| "grad_norm": 0.28445249795913696, | |
| "learning_rate": 2.1326054545382695e-05, | |
| "loss": 9.6118, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.7316341829085458, | |
| "grad_norm": 0.2825443148612976, | |
| "learning_rate": 2.1150762060493155e-05, | |
| "loss": 9.6182, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.7328335832083958, | |
| "grad_norm": 0.29409319162368774, | |
| "learning_rate": 2.09759995141095e-05, | |
| "loss": 9.611, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.7340329835082459, | |
| "grad_norm": 0.30348506569862366, | |
| "learning_rate": 2.0801770116486447e-05, | |
| "loss": 9.6193, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.735232383808096, | |
| "grad_norm": 0.2586905360221863, | |
| "learning_rate": 2.0628077068085173e-05, | |
| "loss": 9.6146, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.7364317841079461, | |
| "grad_norm": 0.27243587374687195, | |
| "learning_rate": 2.0454923559514595e-05, | |
| "loss": 9.6025, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.7376311844077961, | |
| "grad_norm": 0.27491042017936707, | |
| "learning_rate": 2.028231277147261e-05, | |
| "loss": 9.6013, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.7388305847076462, | |
| "grad_norm": 0.279153048992157, | |
| "learning_rate": 2.0110247874687815e-05, | |
| "loss": 9.5937, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.7400299850074963, | |
| "grad_norm": 0.27780649065971375, | |
| "learning_rate": 1.993873202986119e-05, | |
| "loss": 9.6022, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.7412293853073463, | |
| "grad_norm": 0.2798539698123932, | |
| "learning_rate": 1.976776838760801e-05, | |
| "loss": 9.6022, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.7424287856071964, | |
| "grad_norm": 0.27843162417411804, | |
| "learning_rate": 1.9597360088400052e-05, | |
| "loss": 9.6062, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.7436281859070465, | |
| "grad_norm": 0.27371302247047424, | |
| "learning_rate": 1.9427510262507864e-05, | |
| "loss": 9.6119, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.7448275862068966, | |
| "grad_norm": 0.2873663604259491, | |
| "learning_rate": 1.925822202994323e-05, | |
| "loss": 9.6004, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.7460269865067466, | |
| "grad_norm": 0.2875591218471527, | |
| "learning_rate": 1.9089498500401914e-05, | |
| "loss": 9.6119, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.7472263868065967, | |
| "grad_norm": 0.2853778004646301, | |
| "learning_rate": 1.892134277320655e-05, | |
| "loss": 9.6091, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.7484257871064468, | |
| "grad_norm": 0.2952004075050354, | |
| "learning_rate": 1.87537579372496e-05, | |
| "loss": 9.6182, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.7496251874062968, | |
| "grad_norm": 0.3686712980270386, | |
| "learning_rate": 1.858674707093675e-05, | |
| "loss": 9.614, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.7508245877061469, | |
| "grad_norm": 0.2664184868335724, | |
| "learning_rate": 1.8420313242130293e-05, | |
| "loss": 9.6005, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.752023988005997, | |
| "grad_norm": 0.2688407301902771, | |
| "learning_rate": 1.8254459508092768e-05, | |
| "loss": 9.5988, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.7532233883058471, | |
| "grad_norm": 0.2794104516506195, | |
| "learning_rate": 1.8089188915430793e-05, | |
| "loss": 9.5987, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.7544227886056971, | |
| "grad_norm": 0.26486334204673767, | |
| "learning_rate": 1.792450450003919e-05, | |
| "loss": 9.6129, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.7556221889055472, | |
| "grad_norm": 0.2762359082698822, | |
| "learning_rate": 1.7760409287045078e-05, | |
| "loss": 9.6052, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.7568215892053973, | |
| "grad_norm": 0.27764591574668884, | |
| "learning_rate": 1.7596906290752425e-05, | |
| "loss": 9.6056, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.7580209895052473, | |
| "grad_norm": 0.276153028011322, | |
| "learning_rate": 1.743399851458663e-05, | |
| "loss": 9.609, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.7592203898050974, | |
| "grad_norm": 0.2780199646949768, | |
| "learning_rate": 1.727168895103931e-05, | |
| "loss": 9.6081, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.7604197901049475, | |
| "grad_norm": 0.276457816362381, | |
| "learning_rate": 1.7109980581613417e-05, | |
| "loss": 9.6062, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.7616191904047976, | |
| "grad_norm": 0.2808220088481903, | |
| "learning_rate": 1.6948876376768418e-05, | |
| "loss": 9.6123, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.7628185907046476, | |
| "grad_norm": 0.29566583037376404, | |
| "learning_rate": 1.6788379295865704e-05, | |
| "loss": 9.6094, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.7640179910044977, | |
| "grad_norm": 0.33136534690856934, | |
| "learning_rate": 1.6628492287114296e-05, | |
| "loss": 9.614, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.7652173913043478, | |
| "grad_norm": 0.27251994609832764, | |
| "learning_rate": 1.6469218287516664e-05, | |
| "loss": 9.6011, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.766416791604198, | |
| "grad_norm": 0.2670121490955353, | |
| "learning_rate": 1.6310560222814714e-05, | |
| "loss": 9.6037, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.767616191904048, | |
| "grad_norm": 0.2792399227619171, | |
| "learning_rate": 1.6152521007436145e-05, | |
| "loss": 9.6036, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.7688155922038981, | |
| "grad_norm": 0.275511234998703, | |
| "learning_rate": 1.599510354444087e-05, | |
| "loss": 9.5973, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.7700149925037482, | |
| "grad_norm": 0.2751782536506653, | |
| "learning_rate": 1.5838310725467644e-05, | |
| "loss": 9.6005, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.7712143928035982, | |
| "grad_norm": 0.28111734986305237, | |
| "learning_rate": 1.5682145430681027e-05, | |
| "loss": 9.6015, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.7724137931034483, | |
| "grad_norm": 0.2826797068119049, | |
| "learning_rate": 1.5526610528718415e-05, | |
| "loss": 9.6054, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.7736131934032984, | |
| "grad_norm": 0.28505128622055054, | |
| "learning_rate": 1.5371708876637354e-05, | |
| "loss": 9.6042, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.7748125937031485, | |
| "grad_norm": 0.28200674057006836, | |
| "learning_rate": 1.5217443319863112e-05, | |
| "loss": 9.6051, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.7760119940029985, | |
| "grad_norm": 0.2859637439250946, | |
| "learning_rate": 1.5063816692136373e-05, | |
| "loss": 9.6004, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.7772113943028486, | |
| "grad_norm": 0.28504401445388794, | |
| "learning_rate": 1.4910831815461123e-05, | |
| "loss": 9.6177, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.7784107946026987, | |
| "grad_norm": 0.2949487268924713, | |
| "learning_rate": 1.4758491500052924e-05, | |
| "loss": 9.6204, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.7796101949025487, | |
| "grad_norm": 0.3952041268348694, | |
| "learning_rate": 1.4606798544287243e-05, | |
| "loss": 9.62, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.7808095952023988, | |
| "grad_norm": 0.2684868574142456, | |
| "learning_rate": 1.445575573464799e-05, | |
| "loss": 9.5986, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.7820089955022489, | |
| "grad_norm": 0.2751760184764862, | |
| "learning_rate": 1.4305365845676439e-05, | |
| "loss": 9.5993, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.783208395802099, | |
| "grad_norm": 0.27565452456474304, | |
| "learning_rate": 1.4155631639920209e-05, | |
| "loss": 9.5939, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.784407796101949, | |
| "grad_norm": 0.27967387437820435, | |
| "learning_rate": 1.4006555867882464e-05, | |
| "loss": 9.6024, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.7856071964017991, | |
| "grad_norm": 0.28178393840789795, | |
| "learning_rate": 1.3858141267971491e-05, | |
| "loss": 9.6057, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.7868065967016492, | |
| "grad_norm": 0.27983683347702026, | |
| "learning_rate": 1.3710390566450366e-05, | |
| "loss": 9.6059, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.7880059970014992, | |
| "grad_norm": 0.286726713180542, | |
| "learning_rate": 1.3563306477386784e-05, | |
| "loss": 9.6032, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.7892053973013493, | |
| "grad_norm": 0.2814926505088806, | |
| "learning_rate": 1.3416891702603358e-05, | |
| "loss": 9.6077, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.7904047976011994, | |
| "grad_norm": 0.291660875082016, | |
| "learning_rate": 1.3271148931627858e-05, | |
| "loss": 9.6055, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.7916041979010495, | |
| "grad_norm": 0.2863795757293701, | |
| "learning_rate": 1.3126080841643856e-05, | |
| "loss": 9.6111, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.7928035982008995, | |
| "grad_norm": 0.2854698896408081, | |
| "learning_rate": 1.2981690097441573e-05, | |
| "loss": 9.6172, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.7940029985007496, | |
| "grad_norm": 0.3119170367717743, | |
| "learning_rate": 1.2837979351368912e-05, | |
| "loss": 9.6102, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.7952023988005997, | |
| "grad_norm": 0.27526015043258667, | |
| "learning_rate": 1.2694951243282683e-05, | |
| "loss": 9.6006, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.7964017991004497, | |
| "grad_norm": 0.27086350321769714, | |
| "learning_rate": 1.2552608400500199e-05, | |
| "loss": 9.6, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.7976011994002998, | |
| "grad_norm": 0.2674426734447479, | |
| "learning_rate": 1.2410953437750966e-05, | |
| "loss": 9.599, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.7988005997001499, | |
| "grad_norm": 0.26960834860801697, | |
| "learning_rate": 1.2269988957128636e-05, | |
| "loss": 9.6059, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.27745890617370605, | |
| "learning_rate": 1.212971754804324e-05, | |
| "loss": 9.6046, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.80119940029985, | |
| "grad_norm": 0.2803892493247986, | |
| "learning_rate": 1.1990141787173648e-05, | |
| "loss": 9.6036, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.8023988005997001, | |
| "grad_norm": 0.2826705574989319, | |
| "learning_rate": 1.1851264238420135e-05, | |
| "loss": 9.6031, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.8035982008995503, | |
| "grad_norm": 0.28543218970298767, | |
| "learning_rate": 1.1713087452857408e-05, | |
| "loss": 9.6047, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.8047976011994002, | |
| "grad_norm": 0.2749161124229431, | |
| "learning_rate": 1.1575613968687682e-05, | |
| "loss": 9.6061, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.8059970014992504, | |
| "grad_norm": 0.2880239486694336, | |
| "learning_rate": 1.1438846311194024e-05, | |
| "loss": 9.607, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.8071964017991005, | |
| "grad_norm": 0.2794909179210663, | |
| "learning_rate": 1.1302786992694048e-05, | |
| "loss": 9.6098, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.8083958020989506, | |
| "grad_norm": 0.3027547299861908, | |
| "learning_rate": 1.1167438512493683e-05, | |
| "loss": 9.6116, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.8095952023988006, | |
| "grad_norm": 0.34445720911026, | |
| "learning_rate": 1.1032803356841342e-05, | |
| "loss": 9.6171, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.8107946026986507, | |
| "grad_norm": 0.2722347378730774, | |
| "learning_rate": 1.0898883998882158e-05, | |
| "loss": 9.601, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.8119940029985008, | |
| "grad_norm": 0.27299922704696655, | |
| "learning_rate": 1.0765682898612656e-05, | |
| "loss": 9.5976, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.8131934032983508, | |
| "grad_norm": 0.2737182080745697, | |
| "learning_rate": 1.0633202502835494e-05, | |
| "loss": 9.5965, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.8143928035982009, | |
| "grad_norm": 0.2752780020236969, | |
| "learning_rate": 1.0501445245114522e-05, | |
| "loss": 9.6009, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.815592203898051, | |
| "grad_norm": 0.2721465826034546, | |
| "learning_rate": 1.0370413545730118e-05, | |
| "loss": 9.6064, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.8167916041979011, | |
| "grad_norm": 0.2846396267414093, | |
| "learning_rate": 1.0240109811634712e-05, | |
| "loss": 9.5995, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.8179910044977511, | |
| "grad_norm": 0.28411293029785156, | |
| "learning_rate": 1.0110536436408535e-05, | |
| "loss": 9.5975, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.8191904047976012, | |
| "grad_norm": 0.2815098762512207, | |
| "learning_rate": 9.9816958002157e-06, | |
| "loss": 9.6078, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.8203898050974513, | |
| "grad_norm": 0.278131902217865, | |
| "learning_rate": 9.853590269760493e-06, | |
| "loss": 9.6143, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.8215892053973014, | |
| "grad_norm": 0.2930939197540283, | |
| "learning_rate": 9.726222198243806e-06, | |
| "loss": 9.6042, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.8227886056971514, | |
| "grad_norm": 0.2876308560371399, | |
| "learning_rate": 9.599593925320016e-06, | |
| "loss": 9.6187, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.8239880059970015, | |
| "grad_norm": 0.3398456573486328, | |
| "learning_rate": 9.47370777705397e-06, | |
| "loss": 9.6115, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.8251874062968516, | |
| "grad_norm": 0.28324592113494873, | |
| "learning_rate": 9.348566065878217e-06, | |
| "loss": 9.5972, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.8263868065967016, | |
| "grad_norm": 0.271178662776947, | |
| "learning_rate": 9.224171090550571e-06, | |
| "loss": 9.6004, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.8275862068965517, | |
| "grad_norm": 0.26743438839912415, | |
| "learning_rate": 9.100525136111915e-06, | |
| "loss": 9.604, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.8287856071964018, | |
| "grad_norm": 0.2741158604621887, | |
| "learning_rate": 8.97763047384414e-06, | |
| "loss": 9.6024, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.8299850074962519, | |
| "grad_norm": 0.2776412069797516, | |
| "learning_rate": 8.855489361228496e-06, | |
| "loss": 9.5996, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.8311844077961019, | |
| "grad_norm": 0.2762274742126465, | |
| "learning_rate": 8.734104041904129e-06, | |
| "loss": 9.6041, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.832383808095952, | |
| "grad_norm": 0.2758176624774933, | |
| "learning_rate": 8.61347674562677e-06, | |
| "loss": 9.6084, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.8335832083958021, | |
| "grad_norm": 0.28230342268943787, | |
| "learning_rate": 8.4936096882279e-06, | |
| "loss": 9.6047, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.8347826086956521, | |
| "grad_norm": 0.28801631927490234, | |
| "learning_rate": 8.37450507157399e-06, | |
| "loss": 9.6084, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.8359820089955022, | |
| "grad_norm": 0.289760559797287, | |
| "learning_rate": 8.256165083526019e-06, | |
| "loss": 9.6033, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.8371814092953523, | |
| "grad_norm": 0.29011571407318115, | |
| "learning_rate": 8.138591897899345e-06, | |
| "loss": 9.6161, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.8383808095952024, | |
| "grad_norm": 0.3083633780479431, | |
| "learning_rate": 8.021787674423775e-06, | |
| "loss": 9.6152, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.8395802098950524, | |
| "grad_norm": 0.36470380425453186, | |
| "learning_rate": 7.905754558703803e-06, | |
| "loss": 9.6132, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.8407796101949025, | |
| "grad_norm": 0.26850220561027527, | |
| "learning_rate": 7.790494682179317e-06, | |
| "loss": 9.5949, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.8419790104947527, | |
| "grad_norm": 0.2714633643627167, | |
| "learning_rate": 7.676010162086388e-06, | |
| "loss": 9.604, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.8431784107946027, | |
| "grad_norm": 0.2753824293613434, | |
| "learning_rate": 7.56230310141835e-06, | |
| "loss": 9.5993, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.8443778110944528, | |
| "grad_norm": 0.2757047116756439, | |
| "learning_rate": 7.449375588887203e-06, | |
| "loss": 9.5993, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.8455772113943029, | |
| "grad_norm": 0.27331098914146423, | |
| "learning_rate": 7.337229698885279e-06, | |
| "loss": 9.6088, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.846776611694153, | |
| "grad_norm": 0.2818980813026428, | |
| "learning_rate": 7.225867491447053e-06, | |
| "loss": 9.6, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.847976011994003, | |
| "grad_norm": 0.2784759998321533, | |
| "learning_rate": 7.115291012211383e-06, | |
| "loss": 9.6056, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.8491754122938531, | |
| "grad_norm": 0.2809768319129944, | |
| "learning_rate": 7.005502292383898e-06, | |
| "loss": 9.6092, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.8503748125937032, | |
| "grad_norm": 0.29430076479911804, | |
| "learning_rate": 6.896503348699657e-06, | |
| "loss": 9.6031, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.8515742128935532, | |
| "grad_norm": 0.28350192308425903, | |
| "learning_rate": 6.788296183386162e-06, | |
| "loss": 9.6105, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.8527736131934033, | |
| "grad_norm": 0.29121461510658264, | |
| "learning_rate": 6.680882784126552e-06, | |
| "loss": 9.6108, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.8539730134932534, | |
| "grad_norm": 0.3215639889240265, | |
| "learning_rate": 6.5742651240230545e-06, | |
| "loss": 9.6104, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.8551724137931035, | |
| "grad_norm": 0.27074047923088074, | |
| "learning_rate": 6.46844516156081e-06, | |
| "loss": 9.598, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.8563718140929535, | |
| "grad_norm": 0.2728975713253021, | |
| "learning_rate": 6.363424840571869e-06, | |
| "loss": 9.5965, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.8575712143928036, | |
| "grad_norm": 0.2756417393684387, | |
| "learning_rate": 6.259206090199426e-06, | |
| "loss": 9.6021, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.8587706146926537, | |
| "grad_norm": 0.28334730863571167, | |
| "learning_rate": 6.155790824862484e-06, | |
| "loss": 9.5923, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.8599700149925037, | |
| "grad_norm": 0.2780725359916687, | |
| "learning_rate": 6.053180944220627e-06, | |
| "loss": 9.5977, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.8611694152923538, | |
| "grad_norm": 0.2788305878639221, | |
| "learning_rate": 5.951378333139118e-06, | |
| "loss": 9.6104, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.8623688155922039, | |
| "grad_norm": 0.28093886375427246, | |
| "learning_rate": 5.850384861654329e-06, | |
| "loss": 9.5993, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.863568215892054, | |
| "grad_norm": 0.28586381673812866, | |
| "learning_rate": 5.750202384939313e-06, | |
| "loss": 9.6017, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.864767616191904, | |
| "grad_norm": 0.2862658202648163, | |
| "learning_rate": 5.650832743269779e-06, | |
| "loss": 9.6105, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.8659670164917541, | |
| "grad_norm": 0.29242414236068726, | |
| "learning_rate": 5.552277761990294e-06, | |
| "loss": 9.6003, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.8671664167916042, | |
| "grad_norm": 0.2869936525821686, | |
| "learning_rate": 5.454539251480739e-06, | |
| "loss": 9.6101, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.8683658170914542, | |
| "grad_norm": 0.30616676807403564, | |
| "learning_rate": 5.3576190071230106e-06, | |
| "loss": 9.6093, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 0.49281755089759827, | |
| "learning_rate": 5.2615188092681176e-06, | |
| "loss": 9.6174, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.8707646176911544, | |
| "grad_norm": 0.27237847447395325, | |
| "learning_rate": 5.166240423203428e-06, | |
| "loss": 9.5972, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.8719640179910045, | |
| "grad_norm": 0.26910293102264404, | |
| "learning_rate": 5.071785599120243e-06, | |
| "loss": 9.6002, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.8731634182908545, | |
| "grad_norm": 0.28134414553642273, | |
| "learning_rate": 4.978156072081669e-06, | |
| "loss": 9.5976, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.8743628185907046, | |
| "grad_norm": 0.2796195149421692, | |
| "learning_rate": 4.885353561990752e-06, | |
| "loss": 9.6045, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.8755622188905547, | |
| "grad_norm": 0.2702418863773346, | |
| "learning_rate": 4.793379773558815e-06, | |
| "loss": 9.611, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.8767616191904049, | |
| "grad_norm": 0.27752310037612915, | |
| "learning_rate": 4.7022363962742514e-06, | |
| "loss": 9.6131, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.8779610194902548, | |
| "grad_norm": 0.27505457401275635, | |
| "learning_rate": 4.6119251043714225e-06, | |
| "loss": 9.6062, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.879160419790105, | |
| "grad_norm": 0.2814129590988159, | |
| "learning_rate": 4.522447556799875e-06, | |
| "loss": 9.6059, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.8803598200899551, | |
| "grad_norm": 0.28014951944351196, | |
| "learning_rate": 4.433805397193969e-06, | |
| "loss": 9.6106, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.881559220389805, | |
| "grad_norm": 0.2873791456222534, | |
| "learning_rate": 4.3460002538425805e-06, | |
| "loss": 9.6109, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.8827586206896552, | |
| "grad_norm": 0.297184020280838, | |
| "learning_rate": 4.2590337396592406e-06, | |
| "loss": 9.614, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.8839580209895053, | |
| "grad_norm": 0.3112678527832031, | |
| "learning_rate": 4.172907452152519e-06, | |
| "loss": 9.6144, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.8851574212893554, | |
| "grad_norm": 0.27748996019363403, | |
| "learning_rate": 4.087622973396665e-06, | |
| "loss": 9.6036, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.8863568215892054, | |
| "grad_norm": 0.268793523311615, | |
| "learning_rate": 4.0031818700025095e-06, | |
| "loss": 9.6002, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.8875562218890555, | |
| "grad_norm": 0.27348071336746216, | |
| "learning_rate": 3.919585693088751e-06, | |
| "loss": 9.5986, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.8887556221889056, | |
| "grad_norm": 0.2737883925437927, | |
| "learning_rate": 3.836835978253433e-06, | |
| "loss": 9.5968, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.8899550224887556, | |
| "grad_norm": 0.2754174768924713, | |
| "learning_rate": 3.7549342455457216e-06, | |
| "loss": 9.6053, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.8911544227886057, | |
| "grad_norm": 0.27495238184928894, | |
| "learning_rate": 3.6738819994379945e-06, | |
| "loss": 9.607, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.8923538230884558, | |
| "grad_norm": 0.2809472680091858, | |
| "learning_rate": 3.593680728798238e-06, | |
| "loss": 9.6063, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.8935532233883059, | |
| "grad_norm": 0.2831871211528778, | |
| "learning_rate": 3.5143319068626225e-06, | |
| "loss": 9.6096, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.8947526236881559, | |
| "grad_norm": 0.28572776913642883, | |
| "learning_rate": 3.435836991208524e-06, | |
| "loss": 9.606, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.895952023988006, | |
| "grad_norm": 0.2837792634963989, | |
| "learning_rate": 3.35819742372771e-06, | |
| "loss": 9.608, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.8971514242878561, | |
| "grad_norm": 0.2817115783691406, | |
| "learning_rate": 3.2814146305998107e-06, | |
| "loss": 9.6116, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.8983508245877061, | |
| "grad_norm": 0.3011699914932251, | |
| "learning_rate": 3.2054900222662276e-06, | |
| "loss": 9.6132, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.8995502248875562, | |
| "grad_norm": 0.342312753200531, | |
| "learning_rate": 3.1304249934041017e-06, | |
| "loss": 9.61, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.9007496251874063, | |
| "grad_norm": 0.27040937542915344, | |
| "learning_rate": 3.0562209229008042e-06, | |
| "loss": 9.5958, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.9019490254872564, | |
| "grad_norm": 0.2634391188621521, | |
| "learning_rate": 2.982879173828523e-06, | |
| "loss": 9.6023, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.9031484257871064, | |
| "grad_norm": 0.275547057390213, | |
| "learning_rate": 2.9104010934192794e-06, | |
| "loss": 9.6, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.9043478260869565, | |
| "grad_norm": 0.2732800841331482, | |
| "learning_rate": 2.838788013040139e-06, | |
| "loss": 9.6007, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.9055472263868066, | |
| "grad_norm": 0.27795758843421936, | |
| "learning_rate": 2.768041248168801e-06, | |
| "loss": 9.6015, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.9067466266866566, | |
| "grad_norm": 0.2714845836162567, | |
| "learning_rate": 2.6981620983694057e-06, | |
| "loss": 9.6031, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.9079460269865067, | |
| "grad_norm": 0.29261884093284607, | |
| "learning_rate": 2.6291518472686404e-06, | |
| "loss": 9.6028, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.9091454272863568, | |
| "grad_norm": 0.2895471453666687, | |
| "learning_rate": 2.5610117625322118e-06, | |
| "loss": 9.6029, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.9103448275862069, | |
| "grad_norm": 0.2894986867904663, | |
| "learning_rate": 2.4937430958415278e-06, | |
| "loss": 9.6058, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.9115442278860569, | |
| "grad_norm": 0.2817150950431824, | |
| "learning_rate": 2.427347082870701e-06, | |
| "loss": 9.6065, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.912743628185907, | |
| "grad_norm": 0.2893367111682892, | |
| "learning_rate": 2.361824943263874e-06, | |
| "loss": 9.6136, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.9139430284857571, | |
| "grad_norm": 0.3113311529159546, | |
| "learning_rate": 2.2971778806127996e-06, | |
| "loss": 9.6116, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.9151424287856071, | |
| "grad_norm": 0.26471975445747375, | |
| "learning_rate": 2.233407082434724e-06, | |
| "loss": 9.608, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.9163418290854572, | |
| "grad_norm": 0.2689943015575409, | |
| "learning_rate": 2.1705137201505965e-06, | |
| "loss": 9.6016, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.9175412293853074, | |
| "grad_norm": 0.2714982032775879, | |
| "learning_rate": 2.1084989490635255e-06, | |
| "loss": 9.5975, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.9187406296851575, | |
| "grad_norm": 0.2796451151371002, | |
| "learning_rate": 2.0473639083375795e-06, | |
| "loss": 9.6013, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.9199400299850075, | |
| "grad_norm": 0.2697984278202057, | |
| "learning_rate": 1.9871097209768375e-06, | |
| "loss": 9.6081, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.9211394302848576, | |
| "grad_norm": 0.2762463092803955, | |
| "learning_rate": 1.9277374938047988e-06, | |
| "loss": 9.6034, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.9223388305847077, | |
| "grad_norm": 0.28663188219070435, | |
| "learning_rate": 1.8692483174439946e-06, | |
| "loss": 9.5996, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.9235382308845578, | |
| "grad_norm": 0.28348681330680847, | |
| "learning_rate": 1.8116432662960037e-06, | |
| "loss": 9.6014, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.9247376311844078, | |
| "grad_norm": 0.2859058976173401, | |
| "learning_rate": 1.7549233985217074e-06, | |
| "loss": 9.6014, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.9259370314842579, | |
| "grad_norm": 0.2842879295349121, | |
| "learning_rate": 1.6990897560218211e-06, | |
| "loss": 9.6047, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.927136431784108, | |
| "grad_norm": 0.289318323135376, | |
| "learning_rate": 1.644143364417794e-06, | |
| "loss": 9.6067, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.928335832083958, | |
| "grad_norm": 0.29014360904693604, | |
| "learning_rate": 1.5900852330329563e-06, | |
| "loss": 9.6226, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.9295352323838081, | |
| "grad_norm": 0.3719955384731293, | |
| "learning_rate": 1.5369163548739462e-06, | |
| "loss": 9.6146, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.9307346326836582, | |
| "grad_norm": 0.2645496428012848, | |
| "learning_rate": 1.484637706612535e-06, | |
| "loss": 9.6015, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.9319340329835083, | |
| "grad_norm": 0.27676111459732056, | |
| "learning_rate": 1.4332502485676358e-06, | |
| "loss": 9.6031, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.9331334332833583, | |
| "grad_norm": 0.26751890778541565, | |
| "learning_rate": 1.3827549246876625e-06, | |
| "loss": 9.6031, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.9343328335832084, | |
| "grad_norm": 0.26769477128982544, | |
| "learning_rate": 1.333152662533227e-06, | |
| "loss": 9.6071, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.9355322338830585, | |
| "grad_norm": 0.2727934420108795, | |
| "learning_rate": 1.2844443732600576e-06, | |
| "loss": 9.6092, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.9367316341829085, | |
| "grad_norm": 0.2800491452217102, | |
| "learning_rate": 1.2366309516022966e-06, | |
| "loss": 9.6069, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.9379310344827586, | |
| "grad_norm": 0.27642175555229187, | |
| "learning_rate": 1.189713275856047e-06, | |
| "loss": 9.6083, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.9391304347826087, | |
| "grad_norm": 0.2824404239654541, | |
| "learning_rate": 1.1436922078632394e-06, | |
| "loss": 9.6075, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.9403298350824588, | |
| "grad_norm": 0.28561386466026306, | |
| "learning_rate": 1.0985685929958134e-06, | |
| "loss": 9.607, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.9415292353823088, | |
| "grad_norm": 0.29175078868865967, | |
| "learning_rate": 1.0543432601401615e-06, | |
| "loss": 9.6059, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.9427286356821589, | |
| "grad_norm": 0.29381459951400757, | |
| "learning_rate": 1.0110170216819316e-06, | |
| "loss": 9.6138, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.943928035982009, | |
| "grad_norm": 0.30094021558761597, | |
| "learning_rate": 9.685906734910988e-07, | |
| "loss": 9.6111, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.945127436281859, | |
| "grad_norm": 0.27452078461647034, | |
| "learning_rate": 9.270649949073229e-07, | |
| "loss": 9.5987, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.9463268365817091, | |
| "grad_norm": 0.26897814869880676, | |
| "learning_rate": 8.864407487256699e-07, | |
| "loss": 9.6044, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.9475262368815592, | |
| "grad_norm": 0.27936410903930664, | |
| "learning_rate": 8.467186811825623e-07, | |
| "loss": 9.5946, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.9487256371814093, | |
| "grad_norm": 0.279367595911026, | |
| "learning_rate": 8.07899521942096e-07, | |
| "loss": 9.6049, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.9499250374812593, | |
| "grad_norm": 0.2792900502681732, | |
| "learning_rate": 7.69983984082634e-07, | |
| "loss": 9.6013, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.9511244377811094, | |
| "grad_norm": 0.27940914034843445, | |
| "learning_rate": 7.329727640837058e-07, | |
| "loss": 9.6057, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.9523238380809596, | |
| "grad_norm": 0.2867107391357422, | |
| "learning_rate": 6.968665418131848e-07, | |
| "loss": 9.6074, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.9535232383808095, | |
| "grad_norm": 0.28246691823005676, | |
| "learning_rate": 6.616659805148695e-07, | |
| "loss": 9.6092, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.9547226386806597, | |
| "grad_norm": 0.2778690755367279, | |
| "learning_rate": 6.273717267962164e-07, | |
| "loss": 9.612, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.9559220389805098, | |
| "grad_norm": 0.2856720983982086, | |
| "learning_rate": 5.93984410616527e-07, | |
| "loss": 9.6044, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.9571214392803599, | |
| "grad_norm": 0.30416610836982727, | |
| "learning_rate": 5.615046452753403e-07, | |
| "loss": 9.6137, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.9583208395802099, | |
| "grad_norm": 0.2967727780342102, | |
| "learning_rate": 5.299330274011916e-07, | |
| "loss": 9.6139, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.95952023988006, | |
| "grad_norm": 0.33488917350769043, | |
| "learning_rate": 4.992701369406161e-07, | |
| "loss": 9.6133, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.9607196401799101, | |
| "grad_norm": 0.2625320851802826, | |
| "learning_rate": 4.695165371475463e-07, | |
| "loss": 9.6024, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.9619190404797601, | |
| "grad_norm": 0.27328136563301086, | |
| "learning_rate": 4.4067277457292556e-07, | |
| "loss": 9.5989, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.9631184407796102, | |
| "grad_norm": 0.27370065450668335, | |
| "learning_rate": 4.1273937905467185e-07, | |
| "loss": 9.6009, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.9643178410794603, | |
| "grad_norm": 0.27158039808273315, | |
| "learning_rate": 3.8571686370797443e-07, | |
| "loss": 9.6003, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.9655172413793104, | |
| "grad_norm": 0.2797560691833496, | |
| "learning_rate": 3.5960572491583466e-07, | |
| "loss": 9.6009, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.9667166416791604, | |
| "grad_norm": 0.27934470772743225, | |
| "learning_rate": 3.3440644231995664e-07, | |
| "loss": 9.6051, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.9679160419790105, | |
| "grad_norm": 0.2808217406272888, | |
| "learning_rate": 3.101194788119599e-07, | |
| "loss": 9.606, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.9691154422788606, | |
| "grad_norm": 0.2805561423301697, | |
| "learning_rate": 2.867452805248416e-07, | |
| "loss": 9.6049, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.9703148425787106, | |
| "grad_norm": 0.2780965566635132, | |
| "learning_rate": 2.642842768248055e-07, | |
| "loss": 9.6102, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.9715142428785607, | |
| "grad_norm": 0.28727665543556213, | |
| "learning_rate": 2.4273688030336805e-07, | |
| "loss": 9.6085, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.9727136431784108, | |
| "grad_norm": 0.29147574305534363, | |
| "learning_rate": 2.2210348676977023e-07, | |
| "loss": 9.6056, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.9739130434782609, | |
| "grad_norm": 0.30791252851486206, | |
| "learning_rate": 2.0238447524372205e-07, | |
| "loss": 9.6104, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.9751124437781109, | |
| "grad_norm": 0.27840015292167664, | |
| "learning_rate": 1.8358020794843056e-07, | |
| "loss": 9.5996, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.976311844077961, | |
| "grad_norm": 0.26889273524284363, | |
| "learning_rate": 1.6569103030394938e-07, | |
| "loss": 9.6008, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.9775112443778111, | |
| "grad_norm": 0.2736169397830963, | |
| "learning_rate": 1.48717270920834e-07, | |
| "loss": 9.5996, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.9787106446776612, | |
| "grad_norm": 0.27559977769851685, | |
| "learning_rate": 1.3265924159410192e-07, | |
| "loss": 9.5988, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.9799100449775112, | |
| "grad_norm": 0.27969279885292053, | |
| "learning_rate": 1.1751723729750974e-07, | |
| "loss": 9.5987, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.9811094452773613, | |
| "grad_norm": 0.28211551904678345, | |
| "learning_rate": 1.0329153617812947e-07, | |
| "loss": 9.5975, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.9823088455772114, | |
| "grad_norm": 0.27674898505210876, | |
| "learning_rate": 8.998239955124721e-08, | |
| "loss": 9.6071, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.9835082458770614, | |
| "grad_norm": 0.27468326687812805, | |
| "learning_rate": 7.759007189555579e-08, | |
| "loss": 9.6077, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.9847076461769115, | |
| "grad_norm": 0.2778529226779938, | |
| "learning_rate": 6.611478084866951e-08, | |
| "loss": 9.6102, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.9859070464767616, | |
| "grad_norm": 0.27971234917640686, | |
| "learning_rate": 5.555673720292753e-08, | |
| "loss": 9.6097, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.9871064467766117, | |
| "grad_norm": 0.29513809084892273, | |
| "learning_rate": 4.5916134901552443e-08, | |
| "loss": 9.6067, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.9883058470764617, | |
| "grad_norm": 0.2978692352771759, | |
| "learning_rate": 3.7193151035047616e-08, | |
| "loss": 9.6096, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.9895052473763118, | |
| "grad_norm": 0.46041354537010193, | |
| "learning_rate": 2.93879458379831e-08, | |
| "loss": 9.615, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.990704647676162, | |
| "grad_norm": 0.270831823348999, | |
| "learning_rate": 2.2500662686025797e-08, | |
| "loss": 9.5977, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.991904047976012, | |
| "grad_norm": 0.2735693156719208, | |
| "learning_rate": 1.653142809331376e-08, | |
| "loss": 9.6035, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.993103448275862, | |
| "grad_norm": 0.27037888765335083, | |
| "learning_rate": 1.148035171014139e-08, | |
| "loss": 9.6053, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.9943028485757122, | |
| "grad_norm": 0.27673718333244324, | |
| "learning_rate": 7.347526320927723e-09, | |
| "loss": 9.6065, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.9955022488755623, | |
| "grad_norm": 0.29478275775909424, | |
| "learning_rate": 4.133027842517789e-09, | |
| "loss": 9.605, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.9967016491754123, | |
| "grad_norm": 0.2775980830192566, | |
| "learning_rate": 1.8369153228114944e-09, | |
| "loss": 9.6047, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.9979010494752624, | |
| "grad_norm": 0.27957436442375183, | |
| "learning_rate": 4.5923093963118335e-10, | |
| "loss": 9.6086, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.9991004497751125, | |
| "grad_norm": 0.2953091561794281, | |
| "learning_rate": 0.0, | |
| "loss": 9.6163, | |
| "step": 833 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 833, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 239, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 691668038713344.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |