| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 2440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.088006496429443, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.5634, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7192982456140351, | |
| "eval_f1": 0.6524390243902439, | |
| "eval_loss": 0.5107927322387695, | |
| "eval_precision": 0.6572301881961337, | |
| "eval_recall": 0.6488907074013457, | |
| "eval_runtime": 5.0968, | |
| "eval_samples_per_second": 78.284, | |
| "eval_steps_per_second": 9.81, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.8167991638183594, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.5081, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7218045112781954, | |
| "eval_f1": 0.6887653635603403, | |
| "eval_loss": 0.5049471855163574, | |
| "eval_precision": 0.6829453441295547, | |
| "eval_recall": 0.7081742134933624, | |
| "eval_runtime": 5.0562, | |
| "eval_samples_per_second": 78.913, | |
| "eval_steps_per_second": 9.889, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 4.013192176818848, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.4924, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7493734335839599, | |
| "eval_f1": 0.6976723040552828, | |
| "eval_loss": 0.46672317385673523, | |
| "eval_precision": 0.6976723040552828, | |
| "eval_recall": 0.6976723040552828, | |
| "eval_runtime": 5.0806, | |
| "eval_samples_per_second": 78.535, | |
| "eval_steps_per_second": 9.841, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.717806100845337, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4698, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7794486215538847, | |
| "eval_f1": 0.7206949217258496, | |
| "eval_loss": 0.43917685747146606, | |
| "eval_precision": 0.7348989898989899, | |
| "eval_recall": 0.711447535915621, | |
| "eval_runtime": 5.1289, | |
| "eval_samples_per_second": 77.794, | |
| "eval_steps_per_second": 9.749, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.4102962017059326, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.4519, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7468671679197995, | |
| "eval_f1": 0.7226057806810438, | |
| "eval_loss": 0.4547964334487915, | |
| "eval_precision": 0.7169434353918007, | |
| "eval_recall": 0.7534097108565194, | |
| "eval_runtime": 5.1149, | |
| "eval_samples_per_second": 78.007, | |
| "eval_steps_per_second": 9.775, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.93445885181427, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.4356, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8145363408521303, | |
| "eval_f1": 0.7739895897121861, | |
| "eval_loss": 0.41105952858924866, | |
| "eval_precision": 0.7769509251810136, | |
| "eval_recall": 0.7712765957446808, | |
| "eval_runtime": 5.085, | |
| "eval_samples_per_second": 78.466, | |
| "eval_steps_per_second": 9.833, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.6042848825454712, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.421, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7944862155388471, | |
| "eval_f1": 0.7611824817518249, | |
| "eval_loss": 0.41012144088745117, | |
| "eval_precision": 0.7537593984962405, | |
| "eval_recall": 0.7720949263502455, | |
| "eval_runtime": 5.0805, | |
| "eval_samples_per_second": 78.536, | |
| "eval_steps_per_second": 9.842, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 18.006345748901367, | |
| "learning_rate": 3e-05, | |
| "loss": 0.4039, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8295739348370927, | |
| "eval_f1": 0.7933776044839771, | |
| "eval_loss": 0.38294023275375366, | |
| "eval_precision": 0.7949020208205757, | |
| "eval_recall": 0.7919167121294781, | |
| "eval_runtime": 5.064, | |
| "eval_samples_per_second": 78.791, | |
| "eval_steps_per_second": 9.874, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 4.092437744140625, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.3887, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.8320802005012531, | |
| "eval_f1": 0.7979441442122369, | |
| "eval_loss": 0.3800281286239624, | |
| "eval_precision": 0.7972133421798662, | |
| "eval_recall": 0.7986906710310966, | |
| "eval_runtime": 5.0646, | |
| "eval_samples_per_second": 78.782, | |
| "eval_steps_per_second": 9.872, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 4.934494972229004, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.3797, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.8370927318295739, | |
| "eval_f1": 0.8019881353214686, | |
| "eval_loss": 0.37680280208587646, | |
| "eval_precision": 0.8043859649122806, | |
| "eval_recall": 0.7997363156937625, | |
| "eval_runtime": 5.0723, | |
| "eval_samples_per_second": 78.663, | |
| "eval_steps_per_second": 9.857, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 3.9598286151885986, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.368, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.8220551378446115, | |
| "eval_f1": 0.791846991484015, | |
| "eval_loss": 0.38417863845825195, | |
| "eval_precision": 0.7845581927366314, | |
| "eval_recall": 0.801600290961993, | |
| "eval_runtime": 5.0582, | |
| "eval_samples_per_second": 78.882, | |
| "eval_steps_per_second": 9.885, | |
| "step": 1342 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 11.679308891296387, | |
| "learning_rate": 2e-05, | |
| "loss": 0.3598, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.8270676691729323, | |
| "eval_f1": 0.7967966933608887, | |
| "eval_loss": 0.37781035900115967, | |
| "eval_precision": 0.7902444649446494, | |
| "eval_recall": 0.8051463902527732, | |
| "eval_runtime": 5.1525, | |
| "eval_samples_per_second": 77.438, | |
| "eval_steps_per_second": 9.704, | |
| "step": 1464 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 1.9093892574310303, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.3548, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.8471177944862155, | |
| "eval_f1": 0.8141734808401475, | |
| "eval_loss": 0.3623768091201782, | |
| "eval_precision": 0.8166666666666667, | |
| "eval_recall": 0.8118294235315512, | |
| "eval_runtime": 5.0613, | |
| "eval_samples_per_second": 78.833, | |
| "eval_steps_per_second": 9.879, | |
| "step": 1586 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 8.159423828125, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.3469, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.8446115288220551, | |
| "eval_f1": 0.8134839254478557, | |
| "eval_loss": 0.36370429396629333, | |
| "eval_precision": 0.8119747899159664, | |
| "eval_recall": 0.8150572831423895, | |
| "eval_runtime": 5.06, | |
| "eval_samples_per_second": 78.854, | |
| "eval_steps_per_second": 9.881, | |
| "step": 1708 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 1.1732066869735718, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.3431, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.8395989974937343, | |
| "eval_f1": 0.81021463820679, | |
| "eval_loss": 0.3684653341770172, | |
| "eval_precision": 0.8049051094890511, | |
| "eval_recall": 0.8165120931078378, | |
| "eval_runtime": 5.0473, | |
| "eval_samples_per_second": 79.053, | |
| "eval_steps_per_second": 9.906, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 6.608555316925049, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3275, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.8370927318295739, | |
| "eval_f1": 0.8085765951950401, | |
| "eval_loss": 0.3663918673992157, | |
| "eval_precision": 0.8017470018450185, | |
| "eval_recall": 0.817239498090562, | |
| "eval_runtime": 5.098, | |
| "eval_samples_per_second": 78.267, | |
| "eval_steps_per_second": 9.808, | |
| "step": 1952 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 3.14452862739563, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.3288, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.8395989974937343, | |
| "eval_f1": 0.8084033613445378, | |
| "eval_loss": 0.359037846326828, | |
| "eval_precision": 0.8055472976990309, | |
| "eval_recall": 0.8115111838516094, | |
| "eval_runtime": 5.0473, | |
| "eval_samples_per_second": 79.052, | |
| "eval_steps_per_second": 9.906, | |
| "step": 2074 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 2.8248753547668457, | |
| "learning_rate": 5e-06, | |
| "loss": 0.3335, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.8471177944862155, | |
| "eval_f1": 0.8186863532409097, | |
| "eval_loss": 0.36074599623680115, | |
| "eval_precision": 0.8138123167155425, | |
| "eval_recall": 0.8243316966721222, | |
| "eval_runtime": 5.0588, | |
| "eval_samples_per_second": 78.872, | |
| "eval_steps_per_second": 9.884, | |
| "step": 2196 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 4.17168664932251, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.3239, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.8446115288220551, | |
| "eval_f1": 0.8161454307628278, | |
| "eval_loss": 0.36129772663116455, | |
| "eval_precision": 0.8107299270072992, | |
| "eval_recall": 0.8225586470267321, | |
| "eval_runtime": 5.0523, | |
| "eval_samples_per_second": 78.975, | |
| "eval_steps_per_second": 9.897, | |
| "step": 2318 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 6.108452796936035, | |
| "learning_rate": 0.0, | |
| "loss": 0.327, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.8471177944862155, | |
| "eval_f1": 0.8186863532409097, | |
| "eval_loss": 0.3607771396636963, | |
| "eval_precision": 0.8138123167155425, | |
| "eval_recall": 0.8243316966721222, | |
| "eval_runtime": 5.1353, | |
| "eval_samples_per_second": 77.698, | |
| "eval_steps_per_second": 9.737, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 2440, | |
| "total_flos": 7590599775312000.0, | |
| "train_loss": 0.39638843536376955, | |
| "train_runtime": 1955.7644, | |
| "train_samples_per_second": 37.203, | |
| "train_steps_per_second": 1.248 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2440, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 7590599775312000.0, | |
| "train_batch_size": 30, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |