| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 160, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 2.1026898827355764, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 1.6621, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.6413522165034502, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 1.7318, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 0.6085035314989437, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.6973, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.6097206000487838, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.7155, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 0.6181111292344054, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 1.7199, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.6095519945061052, | |
| "learning_rate": 5e-06, | |
| "loss": 1.7225, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 0.5993146111892984, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 1.6788, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.6168594833390011, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.6941, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 0.649510684690767, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 1.7237, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.6127108674374603, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 1.6575, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 0.6444847916730025, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 1.6477, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.6274944995465307, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6443, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 0.6543139619848919, | |
| "learning_rate": 9.999836101895782e-06, | |
| "loss": 1.6202, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.8235053540937074, | |
| "learning_rate": 9.999344418328161e-06, | |
| "loss": 1.6082, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 0.9775219557515483, | |
| "learning_rate": 9.99852498153154e-06, | |
| "loss": 1.5583, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.1030636654770976, | |
| "learning_rate": 9.997377845227577e-06, | |
| "loss": 1.5679, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 1.533566842443126, | |
| "learning_rate": 9.995903084621651e-06, | |
| "loss": 1.5359, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.718620417025083, | |
| "learning_rate": 9.994100796397954e-06, | |
| "loss": 1.4736, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.475, | |
| "grad_norm": 1.848832331677319, | |
| "learning_rate": 9.991971098713136e-06, | |
| "loss": 1.4416, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.1378030725614972, | |
| "learning_rate": 9.98951413118856e-06, | |
| "loss": 1.4385, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.525, | |
| "grad_norm": 0.7491289081750082, | |
| "learning_rate": 9.986730054901153e-06, | |
| "loss": 1.3793, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.6818788464479123, | |
| "learning_rate": 9.983619052372847e-06, | |
| "loss": 1.368, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.575, | |
| "grad_norm": 0.6341413849629475, | |
| "learning_rate": 9.98018132755861e-06, | |
| "loss": 1.3467, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.5923369379348106, | |
| "learning_rate": 9.97641710583307e-06, | |
| "loss": 1.3555, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 0.5340471860637729, | |
| "learning_rate": 9.972326633975754e-06, | |
| "loss": 1.349, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.4785970184408438, | |
| "learning_rate": 9.96791018015489e-06, | |
| "loss": 1.3337, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.675, | |
| "grad_norm": 0.4414536685738152, | |
| "learning_rate": 9.963168033909843e-06, | |
| "loss": 1.3334, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.4860929708817065, | |
| "learning_rate": 9.958100506132127e-06, | |
| "loss": 1.3159, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.725, | |
| "grad_norm": 0.45181979183711657, | |
| "learning_rate": 9.952707929045018e-06, | |
| "loss": 1.3231, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.4400284453917975, | |
| "learning_rate": 9.946990656181782e-06, | |
| "loss": 1.2993, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.775, | |
| "grad_norm": 0.38337739876412114, | |
| "learning_rate": 9.94094906236249e-06, | |
| "loss": 1.2952, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.37181863514378, | |
| "learning_rate": 9.934583543669454e-06, | |
| "loss": 1.3109, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.825, | |
| "grad_norm": 0.3401881416689356, | |
| "learning_rate": 9.927894517421252e-06, | |
| "loss": 1.3157, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.3239891568707606, | |
| "learning_rate": 9.920882422145372e-06, | |
| "loss": 1.34, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 0.3033475087643797, | |
| "learning_rate": 9.913547717549463e-06, | |
| "loss": 1.283, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.2888152197430936, | |
| "learning_rate": 9.905890884491196e-06, | |
| "loss": 1.2892, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.925, | |
| "grad_norm": 0.28458390089996766, | |
| "learning_rate": 9.897912424946741e-06, | |
| "loss": 1.2646, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.27339498391168426, | |
| "learning_rate": 9.889612861977855e-06, | |
| "loss": 1.2832, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.975, | |
| "grad_norm": 0.27940221076287475, | |
| "learning_rate": 9.88099273969759e-06, | |
| "loss": 1.2422, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.27720461090670556, | |
| "learning_rate": 9.872052623234632e-06, | |
| "loss": 1.2919, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.025, | |
| "grad_norm": 0.25894847092459033, | |
| "learning_rate": 9.86279309869623e-06, | |
| "loss": 1.2494, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.2824528364474668, | |
| "learning_rate": 9.853214773129796e-06, | |
| "loss": 1.2404, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.075, | |
| "grad_norm": 0.2731217503169999, | |
| "learning_rate": 9.843318274483089e-06, | |
| "loss": 1.2762, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.282165604906485, | |
| "learning_rate": 9.833104251563058e-06, | |
| "loss": 1.2373, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.125, | |
| "grad_norm": 0.27542846366926443, | |
| "learning_rate": 9.822573373993296e-06, | |
| "loss": 1.2467, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.2851918766199106, | |
| "learning_rate": 9.811726332170153e-06, | |
| "loss": 1.267, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.175, | |
| "grad_norm": 0.2599062629698892, | |
| "learning_rate": 9.800563837217465e-06, | |
| "loss": 1.2585, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.2586420496872339, | |
| "learning_rate": 9.789086620939936e-06, | |
| "loss": 1.2173, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.225, | |
| "grad_norm": 0.27264444944219546, | |
| "learning_rate": 9.777295435775165e-06, | |
| "loss": 1.1856, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.2648452730214859, | |
| "learning_rate": 9.765191054744305e-06, | |
| "loss": 1.2354, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.275, | |
| "grad_norm": 0.27625402001557403, | |
| "learning_rate": 9.752774271401403e-06, | |
| "loss": 1.1857, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.29532784079485663, | |
| "learning_rate": 9.740045899781353e-06, | |
| "loss": 1.2058, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.325, | |
| "grad_norm": 0.2701426980802868, | |
| "learning_rate": 9.727006774346551e-06, | |
| "loss": 1.1724, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.26735542553762237, | |
| "learning_rate": 9.713657749932172e-06, | |
| "loss": 1.194, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.375, | |
| "grad_norm": 0.24830074105443878, | |
| "learning_rate": 9.699999701690133e-06, | |
| "loss": 1.1969, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.2511171777018325, | |
| "learning_rate": 9.68603352503172e-06, | |
| "loss": 1.1822, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.425, | |
| "grad_norm": 0.25711859387852454, | |
| "learning_rate": 9.671760135568882e-06, | |
| "loss": 1.1941, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 0.24983954826654237, | |
| "learning_rate": 9.657180469054213e-06, | |
| "loss": 1.1597, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.475, | |
| "grad_norm": 0.2507740474880207, | |
| "learning_rate": 9.64229548131959e-06, | |
| "loss": 1.1631, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.23964630893140215, | |
| "learning_rate": 9.627106148213521e-06, | |
| "loss": 1.173, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.525, | |
| "grad_norm": 0.22971666963865991, | |
| "learning_rate": 9.61161346553717e-06, | |
| "loss": 1.1446, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.22847059454081425, | |
| "learning_rate": 9.595818448979061e-06, | |
| "loss": 1.1692, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.575, | |
| "grad_norm": 0.2544697716980138, | |
| "learning_rate": 9.579722134048506e-06, | |
| "loss": 1.1712, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.2248726099006599, | |
| "learning_rate": 9.563325576007702e-06, | |
| "loss": 1.1523, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.625, | |
| "grad_norm": 0.23257460453519807, | |
| "learning_rate": 9.546629849802562e-06, | |
| "loss": 1.146, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.2400345523124048, | |
| "learning_rate": 9.529636049992235e-06, | |
| "loss": 1.1238, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.675, | |
| "grad_norm": 0.2410972327939822, | |
| "learning_rate": 9.51234529067735e-06, | |
| "loss": 1.1745, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.2395032264738681, | |
| "learning_rate": 9.494758705426978e-06, | |
| "loss": 1.156, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.725, | |
| "grad_norm": 0.23423115235391723, | |
| "learning_rate": 9.476877447204308e-06, | |
| "loss": 1.1413, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.24379065472217767, | |
| "learning_rate": 9.458702688291072e-06, | |
| "loss": 1.1677, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.775, | |
| "grad_norm": 0.24016354424050318, | |
| "learning_rate": 9.440235620210682e-06, | |
| "loss": 1.1235, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.25909751239656786, | |
| "learning_rate": 9.421477453650118e-06, | |
| "loss": 1.1139, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.825, | |
| "grad_norm": 0.24033129150490448, | |
| "learning_rate": 9.402429418380555e-06, | |
| "loss": 1.1582, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 0.2413776856942996, | |
| "learning_rate": 9.38309276317674e-06, | |
| "loss": 1.1259, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 0.24663210297989538, | |
| "learning_rate": 9.363468755735123e-06, | |
| "loss": 1.0843, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.2469034876268094, | |
| "learning_rate": 9.343558682590757e-06, | |
| "loss": 1.1077, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.925, | |
| "grad_norm": 0.25810333557343196, | |
| "learning_rate": 9.323363849032933e-06, | |
| "loss": 1.1145, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.2457214465581943, | |
| "learning_rate": 9.302885579019626e-06, | |
| "loss": 1.119, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.975, | |
| "grad_norm": 0.2530172029554143, | |
| "learning_rate": 9.282125215090694e-06, | |
| "loss": 1.1324, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.26171895830403796, | |
| "learning_rate": 9.261084118279846e-06, | |
| "loss": 1.0914, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.025, | |
| "grad_norm": 0.2596448526704054, | |
| "learning_rate": 9.23976366802544e-06, | |
| "loss": 1.1202, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 0.2575246492390129, | |
| "learning_rate": 9.218165262080024e-06, | |
| "loss": 1.1191, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.075, | |
| "grad_norm": 0.24809186590262497, | |
| "learning_rate": 9.196290316418712e-06, | |
| "loss": 1.0852, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.25529825764524233, | |
| "learning_rate": 9.174140265146356e-06, | |
| "loss": 1.1091, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.125, | |
| "grad_norm": 0.2598333152137317, | |
| "learning_rate": 9.151716560403521e-06, | |
| "loss": 1.0512, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 0.2533838531424551, | |
| "learning_rate": 9.129020672271283e-06, | |
| "loss": 1.0862, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 2.175, | |
| "grad_norm": 0.2504337844095705, | |
| "learning_rate": 9.106054088674861e-06, | |
| "loss": 1.0717, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.25443455111995733, | |
| "learning_rate": 9.082818315286054e-06, | |
| "loss": 1.0658, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 2.225, | |
| "grad_norm": 0.26694514174541256, | |
| "learning_rate": 9.059314875424553e-06, | |
| "loss": 1.0228, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 0.26148960189209175, | |
| "learning_rate": 9.035545309958048e-06, | |
| "loss": 1.052, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.275, | |
| "grad_norm": 0.27532917589486194, | |
| "learning_rate": 9.011511177201226e-06, | |
| "loss": 1.07, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.28730068505165196, | |
| "learning_rate": 8.987214052813605e-06, | |
| "loss": 1.0609, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 2.325, | |
| "grad_norm": 0.2698216580724319, | |
| "learning_rate": 8.962655529696236e-06, | |
| "loss": 1.0579, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 0.24786365341013725, | |
| "learning_rate": 8.937837217887273e-06, | |
| "loss": 1.0452, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 2.375, | |
| "grad_norm": 0.24197792160968903, | |
| "learning_rate": 8.912760744456416e-06, | |
| "loss": 1.0346, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.25871618240180544, | |
| "learning_rate": 8.887427753398249e-06, | |
| "loss": 1.0452, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.425, | |
| "grad_norm": 0.2670995555409698, | |
| "learning_rate": 8.861839905524452e-06, | |
| "loss": 1.004, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 0.26240119221560043, | |
| "learning_rate": 8.83599887835493e-06, | |
| "loss": 1.0381, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 2.475, | |
| "grad_norm": 0.25007005933407084, | |
| "learning_rate": 8.809906366007832e-06, | |
| "loss": 1.0237, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.27210733980893814, | |
| "learning_rate": 8.783564079088478e-06, | |
| "loss": 1.0326, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.525, | |
| "grad_norm": 0.2778485282051736, | |
| "learning_rate": 8.756973744577221e-06, | |
| "loss": 1.0213, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 0.2709595058080484, | |
| "learning_rate": 8.730137105716231e-06, | |
| "loss": 1.0046, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.575, | |
| "grad_norm": 0.2723123534955665, | |
| "learning_rate": 8.7030559218952e-06, | |
| "loss": 1.0572, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.26318374858426913, | |
| "learning_rate": 8.675731968536004e-06, | |
| "loss": 1.0142, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.625, | |
| "grad_norm": 0.26423478733108197, | |
| "learning_rate": 8.648167036976303e-06, | |
| "loss": 1.0083, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 0.2527490323367728, | |
| "learning_rate": 8.620362934352109e-06, | |
| "loss": 1.0152, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.675, | |
| "grad_norm": 0.2817733909459089, | |
| "learning_rate": 8.592321483479305e-06, | |
| "loss": 0.9994, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.25088178573538616, | |
| "learning_rate": 8.564044522734147e-06, | |
| "loss": 0.966, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.725, | |
| "grad_norm": 0.25165839780166765, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.9897, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 0.27123004373609994, | |
| "learning_rate": 8.506791502209497e-06, | |
| "loss": 1.0011, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.775, | |
| "grad_norm": 0.25885979274531185, | |
| "learning_rate": 8.477819195894615e-06, | |
| "loss": 1.0375, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.25413119834372133, | |
| "learning_rate": 8.448618886390523e-06, | |
| "loss": 0.974, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.825, | |
| "grad_norm": 0.26932379526063627, | |
| "learning_rate": 8.419192488047369e-06, | |
| "loss": 0.9948, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 0.26954131129733855, | |
| "learning_rate": 8.389541930037516e-06, | |
| "loss": 1.0171, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.875, | |
| "grad_norm": 0.2667631834386252, | |
| "learning_rate": 8.359669156229062e-06, | |
| "loss": 0.9917, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.2875114899558342, | |
| "learning_rate": 8.329576125058406e-06, | |
| "loss": 0.9758, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.925, | |
| "grad_norm": 0.28675332351961746, | |
| "learning_rate": 8.29926480940185e-06, | |
| "loss": 0.963, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 0.3013638054236219, | |
| "learning_rate": 8.268737196446264e-06, | |
| "loss": 0.9859, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.975, | |
| "grad_norm": 0.30601615118396874, | |
| "learning_rate": 8.237995287558801e-06, | |
| "loss": 0.9685, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.30346948632781146, | |
| "learning_rate": 8.207041098155701e-06, | |
| "loss": 0.9872, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 3.025, | |
| "grad_norm": 0.31531631458437354, | |
| "learning_rate": 8.175876657570142e-06, | |
| "loss": 0.964, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 0.29114779511016237, | |
| "learning_rate": 8.144504008919224e-06, | |
| "loss": 0.9655, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 3.075, | |
| "grad_norm": 0.32729676853704537, | |
| "learning_rate": 8.112925208969996e-06, | |
| "loss": 0.9484, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.2968428193830138, | |
| "learning_rate": 8.081142328004638e-06, | |
| "loss": 0.9885, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 0.31850537244671423, | |
| "learning_rate": 8.049157449684723e-06, | |
| "loss": 0.9593, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 0.30145179934590266, | |
| "learning_rate": 8.016972670914624e-06, | |
| "loss": 0.9443, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 3.175, | |
| "grad_norm": 0.32941902019125735, | |
| "learning_rate": 7.984590101704026e-06, | |
| "loss": 0.9411, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.33230400688084355, | |
| "learning_rate": 7.952011865029614e-06, | |
| "loss": 0.9353, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 3.225, | |
| "grad_norm": 0.31073576739762815, | |
| "learning_rate": 7.919240096695877e-06, | |
| "loss": 0.9403, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 0.316466603610876, | |
| "learning_rate": 7.886276945195098e-06, | |
| "loss": 0.9575, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.275, | |
| "grad_norm": 0.3252873314387951, | |
| "learning_rate": 7.853124571566491e-06, | |
| "loss": 0.9476, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.34237602645794485, | |
| "learning_rate": 7.819785149254534e-06, | |
| "loss": 0.8938, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 3.325, | |
| "grad_norm": 0.3386834141148311, | |
| "learning_rate": 7.786260863966468e-06, | |
| "loss": 0.8937, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 0.32892426523618096, | |
| "learning_rate": 7.752553913529019e-06, | |
| "loss": 0.9226, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 3.375, | |
| "grad_norm": 0.3663645071149253, | |
| "learning_rate": 7.718666507744293e-06, | |
| "loss": 0.9331, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.35492405802228927, | |
| "learning_rate": 7.68460086824492e-06, | |
| "loss": 0.9024, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 3.425, | |
| "grad_norm": 0.3461059129442488, | |
| "learning_rate": 7.65035922834839e-06, | |
| "loss": 0.9118, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 0.3674770015752456, | |
| "learning_rate": 7.61594383291065e-06, | |
| "loss": 0.885, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 3.475, | |
| "grad_norm": 0.29868653659173194, | |
| "learning_rate": 7.58135693817893e-06, | |
| "loss": 0.8795, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.396103206356086, | |
| "learning_rate": 7.546600811643816e-06, | |
| "loss": 0.9057, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.525, | |
| "grad_norm": 0.4091265454754814, | |
| "learning_rate": 7.5116777318906105e-06, | |
| "loss": 0.9045, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 0.34738555176640296, | |
| "learning_rate": 7.476589988449939e-06, | |
| "loss": 0.897, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 3.575, | |
| "grad_norm": 0.3381371740638481, | |
| "learning_rate": 7.441339881647654e-06, | |
| "loss": 0.9232, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.35909644765143234, | |
| "learning_rate": 7.405929722454026e-06, | |
| "loss": 0.8787, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 3.625, | |
| "grad_norm": 0.3825665161255983, | |
| "learning_rate": 7.370361832332241e-06, | |
| "loss": 0.9113, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 0.30851922400853754, | |
| "learning_rate": 7.334638543086203e-06, | |
| "loss": 0.88, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 3.675, | |
| "grad_norm": 0.37247312676947136, | |
| "learning_rate": 7.298762196707668e-06, | |
| "loss": 0.874, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 0.38511911071863114, | |
| "learning_rate": 7.262735145222696e-06, | |
| "loss": 0.8783, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 3.725, | |
| "grad_norm": 0.45755114413180425, | |
| "learning_rate": 7.226559750537462e-06, | |
| "loss": 0.884, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 0.30880788937024484, | |
| "learning_rate": 7.190238384283413e-06, | |
| "loss": 0.9085, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.775, | |
| "grad_norm": 0.33118326548212185, | |
| "learning_rate": 7.1537734276617744e-06, | |
| "loss": 0.8671, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.41391364228528094, | |
| "learning_rate": 7.117167271287453e-06, | |
| "loss": 0.8695, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 3.825, | |
| "grad_norm": 0.41995379444939823, | |
| "learning_rate": 7.080422315032298e-06, | |
| "loss": 0.8989, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 0.33781129667429866, | |
| "learning_rate": 7.043540967867782e-06, | |
| "loss": 0.8927, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 3.875, | |
| "grad_norm": 0.37184734206990244, | |
| "learning_rate": 7.006525647707054e-06, | |
| "loss": 0.8986, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.42831001354188236, | |
| "learning_rate": 6.969378781246436e-06, | |
| "loss": 0.8899, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 3.925, | |
| "grad_norm": 0.37314784171511645, | |
| "learning_rate": 6.932102803806324e-06, | |
| "loss": 0.857, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 0.38265810168742836, | |
| "learning_rate": 6.894700159171535e-06, | |
| "loss": 0.8423, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 3.975, | |
| "grad_norm": 0.4345836191641771, | |
| "learning_rate": 6.857173299431084e-06, | |
| "loss": 0.8487, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.28712745976233395, | |
| "learning_rate": 6.819524684817439e-06, | |
| "loss": 0.8494, | |
| "step": 160 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 200.0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 182397787570176.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |