| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 10560, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.2239537239074707, | |
| "learning_rate": 4.75e-05, | |
| "loss": 0.7885, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8599324257179768, | |
| "eval_f1": 0.12956964368347987, | |
| "eval_loss": 0.46156567335128784, | |
| "eval_precision": 0.3181818181818182, | |
| "eval_recall": 0.08134805345729228, | |
| "eval_runtime": 4.4353, | |
| "eval_samples_per_second": 210.81, | |
| "eval_steps_per_second": 3.382, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.8783295750617981, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.3921, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9273079598529266, | |
| "eval_f1": 0.6403940886699507, | |
| "eval_loss": 0.25241389870643616, | |
| "eval_precision": 0.6052767718572167, | |
| "eval_recall": 0.6798373038930854, | |
| "eval_runtime": 4.5545, | |
| "eval_samples_per_second": 205.29, | |
| "eval_steps_per_second": 3.293, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.2721015214920044, | |
| "learning_rate": 4.25e-05, | |
| "loss": 0.2392, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9382390937096293, | |
| "eval_f1": 0.7109004739336493, | |
| "eval_loss": 0.19324691593647003, | |
| "eval_precision": 0.6499759268175253, | |
| "eval_recall": 0.7844276583381755, | |
| "eval_runtime": 4.5398, | |
| "eval_samples_per_second": 205.957, | |
| "eval_steps_per_second": 3.304, | |
| "step": 1584 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.5632168054580688, | |
| "learning_rate": 4e-05, | |
| "loss": 0.1931, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9444002782470436, | |
| "eval_f1": 0.7511264245958124, | |
| "eval_loss": 0.16761070489883423, | |
| "eval_precision": 0.6905458089668616, | |
| "eval_recall": 0.8233585124927367, | |
| "eval_runtime": 4.6171, | |
| "eval_samples_per_second": 202.507, | |
| "eval_steps_per_second": 3.249, | |
| "step": 2112 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.1446164846420288, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.1719, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9477789923482063, | |
| "eval_f1": 0.766781639692226, | |
| "eval_loss": 0.15831433236598969, | |
| "eval_precision": 0.70556640625, | |
| "eval_recall": 0.8396281231841952, | |
| "eval_runtime": 4.5172, | |
| "eval_samples_per_second": 206.989, | |
| "eval_steps_per_second": 3.321, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.3869203329086304, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.1602, | |
| "step": 3168 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9501639670078506, | |
| "eval_f1": 0.7779826178562023, | |
| "eval_loss": 0.15387818217277527, | |
| "eval_precision": 0.7114643545279383, | |
| "eval_recall": 0.8582219639744335, | |
| "eval_runtime": 4.7566, | |
| "eval_samples_per_second": 196.569, | |
| "eval_steps_per_second": 3.154, | |
| "step": 3168 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 2.276439666748047, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.1533, | |
| "step": 3696 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.950561462784458, | |
| "eval_f1": 0.7748499869553874, | |
| "eval_loss": 0.152009978890419, | |
| "eval_precision": 0.703125, | |
| "eval_recall": 0.862870424171993, | |
| "eval_runtime": 4.636, | |
| "eval_samples_per_second": 201.68, | |
| "eval_steps_per_second": 3.236, | |
| "step": 3696 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.6530007719993591, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1455, | |
| "step": 4224 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9525489416674948, | |
| "eval_f1": 0.7858095492131236, | |
| "eval_loss": 0.14559316635131836, | |
| "eval_precision": 0.7263313609467456, | |
| "eval_recall": 0.8558977338756537, | |
| "eval_runtime": 4.525, | |
| "eval_samples_per_second": 206.629, | |
| "eval_steps_per_second": 3.315, | |
| "step": 4224 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 1.1390868425369263, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.1398, | |
| "step": 4752 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.953741428997317, | |
| "eval_f1": 0.7870345566568443, | |
| "eval_loss": 0.1424936205148697, | |
| "eval_precision": 0.7301192842942346, | |
| "eval_recall": 0.8535735037768739, | |
| "eval_runtime": 4.5767, | |
| "eval_samples_per_second": 204.296, | |
| "eval_steps_per_second": 3.277, | |
| "step": 4752 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.3200411796569824, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.1368, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.9533439332207095, | |
| "eval_f1": 0.7828403943511858, | |
| "eval_loss": 0.1394987255334854, | |
| "eval_precision": 0.7229330708661418, | |
| "eval_recall": 0.8535735037768739, | |
| "eval_runtime": 4.6067, | |
| "eval_samples_per_second": 202.964, | |
| "eval_steps_per_second": 3.256, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 1.8225971460342407, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.1331, | |
| "step": 5808 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.9551326642154427, | |
| "eval_f1": 0.7904223836427225, | |
| "eval_loss": 0.1364581286907196, | |
| "eval_precision": 0.7359719438877755, | |
| "eval_recall": 0.8535735037768739, | |
| "eval_runtime": 4.6028, | |
| "eval_samples_per_second": 203.138, | |
| "eval_steps_per_second": 3.259, | |
| "step": 5808 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 1.1310523748397827, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1305, | |
| "step": 6336 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.954933916327139, | |
| "eval_f1": 0.7917669072440523, | |
| "eval_loss": 0.13772083818912506, | |
| "eval_precision": 0.7331683168316832, | |
| "eval_recall": 0.8605461940732132, | |
| "eval_runtime": 4.5652, | |
| "eval_samples_per_second": 204.812, | |
| "eval_steps_per_second": 3.286, | |
| "step": 6336 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 1.616012692451477, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.1279, | |
| "step": 6864 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.9565238994335685, | |
| "eval_f1": 0.7955830864530029, | |
| "eval_loss": 0.13570919632911682, | |
| "eval_precision": 0.7414658634538153, | |
| "eval_recall": 0.8582219639744335, | |
| "eval_runtime": 4.5974, | |
| "eval_samples_per_second": 203.376, | |
| "eval_steps_per_second": 3.263, | |
| "step": 6864 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 1.0812599658966064, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.1251, | |
| "step": 7392 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.95553015999205, | |
| "eval_f1": 0.7960438385458433, | |
| "eval_loss": 0.1355232298374176, | |
| "eval_precision": 0.7371287128712871, | |
| "eval_recall": 0.8651946542707728, | |
| "eval_runtime": 4.5977, | |
| "eval_samples_per_second": 203.362, | |
| "eval_steps_per_second": 3.262, | |
| "step": 7392 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.9974623322486877, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.1239, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.954933916327139, | |
| "eval_f1": 0.7947551511907948, | |
| "eval_loss": 0.13592766225337982, | |
| "eval_precision": 0.7366071428571429, | |
| "eval_recall": 0.862870424171993, | |
| "eval_runtime": 4.8881, | |
| "eval_samples_per_second": 191.283, | |
| "eval_steps_per_second": 3.069, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 1.7211874723434448, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1231, | |
| "step": 8448 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.9551326642154427, | |
| "eval_f1": 0.7939053728949478, | |
| "eval_loss": 0.13472402095794678, | |
| "eval_precision": 0.7351485148514851, | |
| "eval_recall": 0.862870424171993, | |
| "eval_runtime": 4.5799, | |
| "eval_samples_per_second": 204.151, | |
| "eval_steps_per_second": 3.275, | |
| "step": 8448 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.8234815001487732, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.122, | |
| "step": 8976 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.95553015999205, | |
| "eval_f1": 0.7939053728949478, | |
| "eval_loss": 0.13527412712574005, | |
| "eval_precision": 0.7351485148514851, | |
| "eval_recall": 0.862870424171993, | |
| "eval_runtime": 4.617, | |
| "eval_samples_per_second": 202.512, | |
| "eval_steps_per_second": 3.249, | |
| "step": 8976 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 1.37420654296875, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1205, | |
| "step": 9504 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.954933916327139, | |
| "eval_f1": 0.7909212283044059, | |
| "eval_loss": 0.1356407105922699, | |
| "eval_precision": 0.7317193675889329, | |
| "eval_recall": 0.8605461940732132, | |
| "eval_runtime": 4.6112, | |
| "eval_samples_per_second": 202.766, | |
| "eval_steps_per_second": 3.253, | |
| "step": 9504 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.861274242401123, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1202, | |
| "step": 10032 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.9551326642154427, | |
| "eval_f1": 0.7939053728949478, | |
| "eval_loss": 0.13467516005039215, | |
| "eval_precision": 0.7351485148514851, | |
| "eval_recall": 0.862870424171993, | |
| "eval_runtime": 4.5519, | |
| "eval_samples_per_second": 205.408, | |
| "eval_steps_per_second": 3.295, | |
| "step": 10032 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 1.8209986686706543, | |
| "learning_rate": 0.0, | |
| "loss": 0.1204, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.95553015999205, | |
| "eval_f1": 0.7947551511907948, | |
| "eval_loss": 0.13456128537654877, | |
| "eval_precision": 0.7366071428571429, | |
| "eval_recall": 0.862870424171993, | |
| "eval_runtime": 4.5477, | |
| "eval_samples_per_second": 205.599, | |
| "eval_steps_per_second": 3.298, | |
| "step": 10560 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 10560, | |
| "total_flos": 4541164131293502.0, | |
| "train_loss": 0.18836262515096952, | |
| "train_runtime": 1256.9137, | |
| "train_samples_per_second": 134.249, | |
| "train_steps_per_second": 8.402 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10560, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 4541164131293502.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |