| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.2797651464850307, | |
| "eval_steps": 500, | |
| "global_step": 600000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004662752441417178, | |
| "grad_norm": 1.3671598434448242, | |
| "learning_rate": 4.976690900545356e-05, | |
| "loss": 4.8243, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.009325504882834356, | |
| "grad_norm": 1.3568960428237915, | |
| "learning_rate": 4.953384132466932e-05, | |
| "loss": 3.7236, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.013988257324251536, | |
| "grad_norm": 1.4119852781295776, | |
| "learning_rate": 4.930079695764729e-05, | |
| "loss": 3.4332, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.018651009765668712, | |
| "grad_norm": 1.1216883659362793, | |
| "learning_rate": 4.906772927686305e-05, | |
| "loss": 3.2825, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.023313762207085892, | |
| "grad_norm": 1.2323102951049805, | |
| "learning_rate": 4.883468490984102e-05, | |
| "loss": 3.1866, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.02797651464850307, | |
| "grad_norm": 0.9565121531486511, | |
| "learning_rate": 4.860161722905678e-05, | |
| "loss": 3.1126, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 0.03263926708992025, | |
| "grad_norm": 1.1173287630081177, | |
| "learning_rate": 4.836861948955917e-05, | |
| "loss": 3.0577, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 0.037302019531337424, | |
| "grad_norm": 1.4626446962356567, | |
| "learning_rate": 4.813555180877493e-05, | |
| "loss": 3.0145, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 0.041964771972754604, | |
| "grad_norm": 1.0016125440597534, | |
| "learning_rate": 4.790246081422848e-05, | |
| "loss": 2.973, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 0.046627524414171784, | |
| "grad_norm": 1.3417049646377563, | |
| "learning_rate": 4.766941644720645e-05, | |
| "loss": 2.9423, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 0.051290276855588963, | |
| "grad_norm": 1.2270045280456543, | |
| "learning_rate": 4.7436372080184424e-05, | |
| "loss": 2.9127, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 0.05595302929700614, | |
| "grad_norm": 1.0300427675247192, | |
| "learning_rate": 4.7203327713162395e-05, | |
| "loss": 2.8823, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 0.060615781738423316, | |
| "grad_norm": 0.8841068148612976, | |
| "learning_rate": 4.6970283346140366e-05, | |
| "loss": 2.8588, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 0.0652785341798405, | |
| "grad_norm": 1.0524730682373047, | |
| "learning_rate": 4.673721566535613e-05, | |
| "loss": 2.8425, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 0.06994128662125768, | |
| "grad_norm": 0.9874018430709839, | |
| "learning_rate": 4.650417129833409e-05, | |
| "loss": 2.8257, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 0.07460403906267485, | |
| "grad_norm": 0.9634119272232056, | |
| "learning_rate": 4.6271126931312064e-05, | |
| "loss": 2.8114, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 0.07926679150409204, | |
| "grad_norm": 0.885671854019165, | |
| "learning_rate": 4.603808256429003e-05, | |
| "loss": 2.7911, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 0.08392954394550921, | |
| "grad_norm": 1.0135940313339233, | |
| "learning_rate": 4.58050148835058e-05, | |
| "loss": 2.7786, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 0.0885922963869264, | |
| "grad_norm": 1.011932611465454, | |
| "learning_rate": 4.557194720272156e-05, | |
| "loss": 2.7636, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 0.09325504882834357, | |
| "grad_norm": 0.7796096205711365, | |
| "learning_rate": 4.533892614946173e-05, | |
| "loss": 2.753, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 0.09791780126976074, | |
| "grad_norm": 1.1194034814834595, | |
| "learning_rate": 4.5105858468677495e-05, | |
| "loss": 2.7371, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 0.10258055371117793, | |
| "grad_norm": 1.1135520935058594, | |
| "learning_rate": 4.4872814101655467e-05, | |
| "loss": 2.7285, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 0.1072433061525951, | |
| "grad_norm": 0.7772097587585449, | |
| "learning_rate": 4.463976973463344e-05, | |
| "loss": 2.7173, | |
| "step": 230000 | |
| }, | |
| { | |
| "epoch": 0.11190605859401229, | |
| "grad_norm": 1.096358299255371, | |
| "learning_rate": 4.44067020538492e-05, | |
| "loss": 2.7141, | |
| "step": 240000 | |
| }, | |
| { | |
| "epoch": 0.11656881103542946, | |
| "grad_norm": 0.8112640380859375, | |
| "learning_rate": 4.417363437306496e-05, | |
| "loss": 2.7073, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 0.12123156347684663, | |
| "grad_norm": 1.0931545495986938, | |
| "learning_rate": 4.394059000604293e-05, | |
| "loss": 2.6931, | |
| "step": 260000 | |
| }, | |
| { | |
| "epoch": 0.1258943159182638, | |
| "grad_norm": 1.1369918584823608, | |
| "learning_rate": 4.37075456390209e-05, | |
| "loss": 2.6824, | |
| "step": 270000 | |
| }, | |
| { | |
| "epoch": 0.130557068359681, | |
| "grad_norm": 1.258300542831421, | |
| "learning_rate": 4.347450127199887e-05, | |
| "loss": 2.6741, | |
| "step": 280000 | |
| }, | |
| { | |
| "epoch": 0.13521982080109818, | |
| "grad_norm": 0.9752686023712158, | |
| "learning_rate": 4.324143359121463e-05, | |
| "loss": 2.6645, | |
| "step": 290000 | |
| }, | |
| { | |
| "epoch": 0.13988257324251535, | |
| "grad_norm": 1.0001367330551147, | |
| "learning_rate": 4.3008412537954805e-05, | |
| "loss": 2.6592, | |
| "step": 300000 | |
| }, | |
| { | |
| "epoch": 0.14454532568393252, | |
| "grad_norm": 1.0314422845840454, | |
| "learning_rate": 4.277534485717057e-05, | |
| "loss": 2.6523, | |
| "step": 310000 | |
| }, | |
| { | |
| "epoch": 0.1492080781253497, | |
| "grad_norm": 0.9287506937980652, | |
| "learning_rate": 4.254230049014854e-05, | |
| "loss": 2.6503, | |
| "step": 320000 | |
| }, | |
| { | |
| "epoch": 0.1538708305667669, | |
| "grad_norm": 0.8209073543548584, | |
| "learning_rate": 4.23092328093643e-05, | |
| "loss": 2.6379, | |
| "step": 330000 | |
| }, | |
| { | |
| "epoch": 0.15853358300818407, | |
| "grad_norm": 0.8727386593818665, | |
| "learning_rate": 4.207618844234227e-05, | |
| "loss": 2.6332, | |
| "step": 340000 | |
| }, | |
| { | |
| "epoch": 0.16319633544960124, | |
| "grad_norm": 0.9841961860656738, | |
| "learning_rate": 4.184314407532024e-05, | |
| "loss": 2.6279, | |
| "step": 350000 | |
| }, | |
| { | |
| "epoch": 0.16785908789101842, | |
| "grad_norm": 0.7831237316131592, | |
| "learning_rate": 4.1610099708298214e-05, | |
| "loss": 2.6237, | |
| "step": 360000 | |
| }, | |
| { | |
| "epoch": 0.1725218403324356, | |
| "grad_norm": 0.9184048175811768, | |
| "learning_rate": 4.137707865503839e-05, | |
| "loss": 2.6167, | |
| "step": 370000 | |
| }, | |
| { | |
| "epoch": 0.1771845927738528, | |
| "grad_norm": 0.9598727822303772, | |
| "learning_rate": 4.114401097425415e-05, | |
| "loss": 2.6082, | |
| "step": 380000 | |
| }, | |
| { | |
| "epoch": 0.18184734521526996, | |
| "grad_norm": 0.8814136981964111, | |
| "learning_rate": 4.0910966607232115e-05, | |
| "loss": 2.6027, | |
| "step": 390000 | |
| }, | |
| { | |
| "epoch": 0.18651009765668713, | |
| "grad_norm": 0.9080318212509155, | |
| "learning_rate": 4.067789892644788e-05, | |
| "loss": 2.6014, | |
| "step": 400000 | |
| }, | |
| { | |
| "epoch": 0.1911728500981043, | |
| "grad_norm": 0.8321977257728577, | |
| "learning_rate": 4.044485455942585e-05, | |
| "loss": 2.5964, | |
| "step": 410000 | |
| }, | |
| { | |
| "epoch": 0.19583560253952148, | |
| "grad_norm": 1.3375693559646606, | |
| "learning_rate": 4.021178687864161e-05, | |
| "loss": 2.5906, | |
| "step": 420000 | |
| }, | |
| { | |
| "epoch": 0.20049835498093868, | |
| "grad_norm": 0.8211286067962646, | |
| "learning_rate": 3.997874251161958e-05, | |
| "loss": 2.5866, | |
| "step": 430000 | |
| }, | |
| { | |
| "epoch": 0.20516110742235585, | |
| "grad_norm": 0.7890422940254211, | |
| "learning_rate": 3.9745698144597546e-05, | |
| "loss": 2.5817, | |
| "step": 440000 | |
| }, | |
| { | |
| "epoch": 0.20982385986377303, | |
| "grad_norm": 1.0580294132232666, | |
| "learning_rate": 3.951263046381331e-05, | |
| "loss": 2.5776, | |
| "step": 450000 | |
| }, | |
| { | |
| "epoch": 0.2144866123051902, | |
| "grad_norm": 1.0666168928146362, | |
| "learning_rate": 3.927956278302907e-05, | |
| "loss": 2.5729, | |
| "step": 460000 | |
| }, | |
| { | |
| "epoch": 0.21914936474660737, | |
| "grad_norm": 1.0440067052841187, | |
| "learning_rate": 3.904651841600704e-05, | |
| "loss": 2.5748, | |
| "step": 470000 | |
| }, | |
| { | |
| "epoch": 0.22381211718802457, | |
| "grad_norm": 0.8746099472045898, | |
| "learning_rate": 3.88134507352228e-05, | |
| "loss": 2.5704, | |
| "step": 480000 | |
| }, | |
| { | |
| "epoch": 0.22847486962944175, | |
| "grad_norm": 0.882897675037384, | |
| "learning_rate": 3.858040636820078e-05, | |
| "loss": 2.5623, | |
| "step": 490000 | |
| }, | |
| { | |
| "epoch": 0.23313762207085892, | |
| "grad_norm": 0.8458369970321655, | |
| "learning_rate": 3.834740862870316e-05, | |
| "loss": 2.5612, | |
| "step": 500000 | |
| }, | |
| { | |
| "epoch": 0.2378003745122761, | |
| "grad_norm": 0.9579658508300781, | |
| "learning_rate": 3.811434094791892e-05, | |
| "loss": 2.5551, | |
| "step": 510000 | |
| }, | |
| { | |
| "epoch": 0.24246312695369326, | |
| "grad_norm": 1.0498722791671753, | |
| "learning_rate": 3.78813198946591e-05, | |
| "loss": 2.5502, | |
| "step": 520000 | |
| }, | |
| { | |
| "epoch": 0.24712587939511046, | |
| "grad_norm": 1.032334804534912, | |
| "learning_rate": 3.764825221387486e-05, | |
| "loss": 2.5534, | |
| "step": 530000 | |
| }, | |
| { | |
| "epoch": 0.2517886318365276, | |
| "grad_norm": 0.9145790934562683, | |
| "learning_rate": 3.7415184533090624e-05, | |
| "loss": 2.547, | |
| "step": 540000 | |
| }, | |
| { | |
| "epoch": 0.2564513842779448, | |
| "grad_norm": 1.0633904933929443, | |
| "learning_rate": 3.71821634798308e-05, | |
| "loss": 2.543, | |
| "step": 550000 | |
| }, | |
| { | |
| "epoch": 0.261114136719362, | |
| "grad_norm": 0.9828123450279236, | |
| "learning_rate": 3.694911911280877e-05, | |
| "loss": 2.5398, | |
| "step": 560000 | |
| }, | |
| { | |
| "epoch": 0.2657768891607792, | |
| "grad_norm": 0.8735861778259277, | |
| "learning_rate": 3.671607474578674e-05, | |
| "loss": 2.5345, | |
| "step": 570000 | |
| }, | |
| { | |
| "epoch": 0.27043964160219636, | |
| "grad_norm": 1.1347264051437378, | |
| "learning_rate": 3.6483053692526915e-05, | |
| "loss": 2.5312, | |
| "step": 580000 | |
| }, | |
| { | |
| "epoch": 0.27510239404361353, | |
| "grad_norm": 0.8275557160377502, | |
| "learning_rate": 3.625000932550489e-05, | |
| "loss": 2.5299, | |
| "step": 590000 | |
| }, | |
| { | |
| "epoch": 0.2797651464850307, | |
| "grad_norm": 0.8891148567199707, | |
| "learning_rate": 3.601696495848285e-05, | |
| "loss": 2.5301, | |
| "step": 600000 | |
| } | |
| ], | |
| "logging_steps": 10000, | |
| "max_steps": 2144656, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 100000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4794966500590223e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |