| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 6880, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11627906976744186, | |
| "grad_norm": 1.6077862977981567, | |
| "learning_rate": 1.1511627906976744e-05, | |
| "loss": 7.0526, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23255813953488372, | |
| "grad_norm": 0.98894202709198, | |
| "learning_rate": 2.313953488372093e-05, | |
| "loss": 5.427, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3488372093023256, | |
| "grad_norm": 0.7544777989387512, | |
| "learning_rate": 3.4767441860465115e-05, | |
| "loss": 4.7742, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.46511627906976744, | |
| "grad_norm": 1.1207612752914429, | |
| "learning_rate": 4.639534883720931e-05, | |
| "loss": 4.5344, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5813953488372093, | |
| "grad_norm": 0.8379786610603333, | |
| "learning_rate": 5.802325581395349e-05, | |
| "loss": 4.3805, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6976744186046512, | |
| "grad_norm": 0.7775167226791382, | |
| "learning_rate": 6.965116279069767e-05, | |
| "loss": 4.1699, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.813953488372093, | |
| "grad_norm": 0.7881819009780884, | |
| "learning_rate": 8.127906976744186e-05, | |
| "loss": 4.0952, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9302325581395349, | |
| "grad_norm": 0.7651063799858093, | |
| "learning_rate": 9.290697674418604e-05, | |
| "loss": 4.0067, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 3.915109872817993, | |
| "eval_runtime": 7.8813, | |
| "eval_samples_per_second": 388.009, | |
| "eval_steps_per_second": 48.596, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.0465116279069768, | |
| "grad_norm": 0.6977295279502869, | |
| "learning_rate": 9.949612403100775e-05, | |
| "loss": 3.855, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.1627906976744187, | |
| "grad_norm": 0.8410159349441528, | |
| "learning_rate": 9.820413436692507e-05, | |
| "loss": 3.8458, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.2790697674418605, | |
| "grad_norm": 0.7741426825523376, | |
| "learning_rate": 9.691214470284238e-05, | |
| "loss": 3.7573, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.3953488372093024, | |
| "grad_norm": 0.7446125745773315, | |
| "learning_rate": 9.56201550387597e-05, | |
| "loss": 3.6336, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.5116279069767442, | |
| "grad_norm": 0.900859534740448, | |
| "learning_rate": 9.4328165374677e-05, | |
| "loss": 3.6147, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.627906976744186, | |
| "grad_norm": 0.8706986308097839, | |
| "learning_rate": 9.303617571059432e-05, | |
| "loss": 3.5227, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.744186046511628, | |
| "grad_norm": 0.9106377959251404, | |
| "learning_rate": 9.174418604651163e-05, | |
| "loss": 3.4545, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.8604651162790697, | |
| "grad_norm": 0.7554797530174255, | |
| "learning_rate": 9.045219638242895e-05, | |
| "loss": 3.4257, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.9767441860465116, | |
| "grad_norm": 0.8181759119033813, | |
| "learning_rate": 8.916020671834626e-05, | |
| "loss": 3.3851, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 3.351266860961914, | |
| "eval_runtime": 7.8508, | |
| "eval_samples_per_second": 389.516, | |
| "eval_steps_per_second": 48.785, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.0930232558139537, | |
| "grad_norm": 0.783263623714447, | |
| "learning_rate": 8.786821705426357e-05, | |
| "loss": 3.2934, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.2093023255813953, | |
| "grad_norm": 0.8410254716873169, | |
| "learning_rate": 8.657622739018088e-05, | |
| "loss": 3.2805, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.3255813953488373, | |
| "grad_norm": 0.829415500164032, | |
| "learning_rate": 8.52842377260982e-05, | |
| "loss": 3.2845, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.441860465116279, | |
| "grad_norm": 0.8594213128089905, | |
| "learning_rate": 8.399224806201551e-05, | |
| "loss": 3.1855, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.558139534883721, | |
| "grad_norm": 1.0678484439849854, | |
| "learning_rate": 8.270025839793282e-05, | |
| "loss": 3.1735, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.6744186046511627, | |
| "grad_norm": 0.7645172476768494, | |
| "learning_rate": 8.140826873385013e-05, | |
| "loss": 3.1673, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.7906976744186047, | |
| "grad_norm": 0.838912844657898, | |
| "learning_rate": 8.011627906976745e-05, | |
| "loss": 3.1312, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.9069767441860463, | |
| "grad_norm": 0.8251073360443115, | |
| "learning_rate": 7.882428940568476e-05, | |
| "loss": 3.0803, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 3.090207815170288, | |
| "eval_runtime": 7.8554, | |
| "eval_samples_per_second": 389.286, | |
| "eval_steps_per_second": 48.756, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 3.0232558139534884, | |
| "grad_norm": 0.8788666725158691, | |
| "learning_rate": 7.753229974160207e-05, | |
| "loss": 3.05, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.13953488372093, | |
| "grad_norm": 0.8054720163345337, | |
| "learning_rate": 7.624031007751938e-05, | |
| "loss": 2.9595, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 3.255813953488372, | |
| "grad_norm": 0.8272735476493835, | |
| "learning_rate": 7.49483204134367e-05, | |
| "loss": 2.9702, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.3720930232558137, | |
| "grad_norm": 0.8935261368751526, | |
| "learning_rate": 7.365633074935401e-05, | |
| "loss": 2.971, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 3.488372093023256, | |
| "grad_norm": 0.8290767073631287, | |
| "learning_rate": 7.236434108527133e-05, | |
| "loss": 2.9443, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.604651162790698, | |
| "grad_norm": 0.891353964805603, | |
| "learning_rate": 7.107235142118863e-05, | |
| "loss": 2.9608, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 3.7209302325581395, | |
| "grad_norm": 1.282278299331665, | |
| "learning_rate": 6.978036175710595e-05, | |
| "loss": 2.9595, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 3.8372093023255816, | |
| "grad_norm": 0.8332661986351013, | |
| "learning_rate": 6.848837209302326e-05, | |
| "loss": 2.9128, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 3.953488372093023, | |
| "grad_norm": 0.8648365139961243, | |
| "learning_rate": 6.719638242894058e-05, | |
| "loss": 2.8727, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.9329044818878174, | |
| "eval_runtime": 7.8343, | |
| "eval_samples_per_second": 390.333, | |
| "eval_steps_per_second": 48.887, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 4.069767441860465, | |
| "grad_norm": 0.8708928823471069, | |
| "learning_rate": 6.590439276485788e-05, | |
| "loss": 2.8547, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.186046511627907, | |
| "grad_norm": 0.9987735152244568, | |
| "learning_rate": 6.46124031007752e-05, | |
| "loss": 2.7684, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 4.3023255813953485, | |
| "grad_norm": 0.8593104481697083, | |
| "learning_rate": 6.332041343669251e-05, | |
| "loss": 2.7917, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 4.4186046511627906, | |
| "grad_norm": 0.9889882206916809, | |
| "learning_rate": 6.202842377260983e-05, | |
| "loss": 2.8232, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 4.534883720930233, | |
| "grad_norm": 0.9209152460098267, | |
| "learning_rate": 6.073643410852713e-05, | |
| "loss": 2.7823, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 4.651162790697675, | |
| "grad_norm": 0.9376141428947449, | |
| "learning_rate": 5.9444444444444445e-05, | |
| "loss": 2.7623, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.767441860465116, | |
| "grad_norm": 0.9320041537284851, | |
| "learning_rate": 5.815245478036177e-05, | |
| "loss": 2.7476, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 4.883720930232558, | |
| "grad_norm": 0.8268842101097107, | |
| "learning_rate": 5.686046511627907e-05, | |
| "loss": 2.7621, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.042984962463379, | |
| "learning_rate": 5.5568475452196386e-05, | |
| "loss": 2.7301, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 2.827605724334717, | |
| "eval_runtime": 7.8468, | |
| "eval_samples_per_second": 389.714, | |
| "eval_steps_per_second": 48.81, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 5.116279069767442, | |
| "grad_norm": 0.931705117225647, | |
| "learning_rate": 5.4276485788113695e-05, | |
| "loss": 2.6633, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 5.232558139534884, | |
| "grad_norm": 1.0284597873687744, | |
| "learning_rate": 5.298449612403101e-05, | |
| "loss": 2.6595, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 5.348837209302325, | |
| "grad_norm": 1.0262751579284668, | |
| "learning_rate": 5.169250645994832e-05, | |
| "loss": 2.6781, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 5.465116279069767, | |
| "grad_norm": 1.0172241926193237, | |
| "learning_rate": 5.0400516795865635e-05, | |
| "loss": 2.6229, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 5.5813953488372094, | |
| "grad_norm": 0.8998900055885315, | |
| "learning_rate": 4.910852713178295e-05, | |
| "loss": 2.6155, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 5.6976744186046515, | |
| "grad_norm": 0.9043959975242615, | |
| "learning_rate": 4.781653746770026e-05, | |
| "loss": 2.6454, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 5.813953488372093, | |
| "grad_norm": 1.0376083850860596, | |
| "learning_rate": 4.6524547803617576e-05, | |
| "loss": 2.6783, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.930232558139535, | |
| "grad_norm": 0.8958259224891663, | |
| "learning_rate": 4.5232558139534885e-05, | |
| "loss": 2.6266, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.757228136062622, | |
| "eval_runtime": 7.8611, | |
| "eval_samples_per_second": 389.006, | |
| "eval_steps_per_second": 48.721, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 6.046511627906977, | |
| "grad_norm": 1.0552242994308472, | |
| "learning_rate": 4.39405684754522e-05, | |
| "loss": 2.5892, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 6.162790697674419, | |
| "grad_norm": 1.0633459091186523, | |
| "learning_rate": 4.264857881136951e-05, | |
| "loss": 2.5359, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 6.27906976744186, | |
| "grad_norm": 1.0308552980422974, | |
| "learning_rate": 4.1356589147286825e-05, | |
| "loss": 2.5536, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 6.395348837209302, | |
| "grad_norm": 0.9943313598632812, | |
| "learning_rate": 4.0064599483204134e-05, | |
| "loss": 2.5282, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 6.511627906976744, | |
| "grad_norm": 1.0665168762207031, | |
| "learning_rate": 3.877260981912145e-05, | |
| "loss": 2.5605, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 6.627906976744186, | |
| "grad_norm": 0.9594394564628601, | |
| "learning_rate": 3.748062015503876e-05, | |
| "loss": 2.5349, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 6.7441860465116275, | |
| "grad_norm": 1.0906480550765991, | |
| "learning_rate": 3.6188630490956075e-05, | |
| "loss": 2.5282, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 6.8604651162790695, | |
| "grad_norm": 0.9688259363174438, | |
| "learning_rate": 3.489664082687339e-05, | |
| "loss": 2.5354, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 6.976744186046512, | |
| "grad_norm": 1.0499610900878906, | |
| "learning_rate": 3.36046511627907e-05, | |
| "loss": 2.5551, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 2.7097859382629395, | |
| "eval_runtime": 7.8861, | |
| "eval_samples_per_second": 387.773, | |
| "eval_steps_per_second": 48.567, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 7.093023255813954, | |
| "grad_norm": 1.0121361017227173, | |
| "learning_rate": 3.2312661498708015e-05, | |
| "loss": 2.473, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 7.209302325581396, | |
| "grad_norm": 1.0595309734344482, | |
| "learning_rate": 3.1020671834625324e-05, | |
| "loss": 2.4383, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 7.325581395348837, | |
| "grad_norm": 1.0951762199401855, | |
| "learning_rate": 2.9728682170542636e-05, | |
| "loss": 2.4518, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 7.441860465116279, | |
| "grad_norm": 1.131641149520874, | |
| "learning_rate": 2.843669250645995e-05, | |
| "loss": 2.4143, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 7.558139534883721, | |
| "grad_norm": 1.1722636222839355, | |
| "learning_rate": 2.714470284237726e-05, | |
| "loss": 2.5198, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 7.674418604651163, | |
| "grad_norm": 1.0916329622268677, | |
| "learning_rate": 2.5852713178294573e-05, | |
| "loss": 2.4656, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 7.790697674418604, | |
| "grad_norm": 1.102444052696228, | |
| "learning_rate": 2.4560723514211886e-05, | |
| "loss": 2.4779, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 7.906976744186046, | |
| "grad_norm": 1.101860761642456, | |
| "learning_rate": 2.32687338501292e-05, | |
| "loss": 2.4783, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.6801645755767822, | |
| "eval_runtime": 7.8846, | |
| "eval_samples_per_second": 387.844, | |
| "eval_steps_per_second": 48.576, | |
| "step": 6880 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 8600, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.7517692420096e+16, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |