| { | |
| "best_metric": 0.46745064854621887, | |
| "best_model_checkpoint": "secdisclosure-28l/checkpoint-1144", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 1144, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 6.856657028198242, | |
| "learning_rate": 6.993006993006993e-05, | |
| "loss": 3.041, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 6.033031940460205, | |
| "learning_rate": 9.953357444566039e-05, | |
| "loss": 1.4738, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7657342657342657, | |
| "eval_f1_macro": 0.7322810229739891, | |
| "eval_f1_micro": 0.7657342657342657, | |
| "eval_f1_weighted": 0.7319583902073576, | |
| "eval_loss": 0.7600058317184448, | |
| "eval_precision_macro": 0.7772121039163471, | |
| "eval_precision_micro": 0.7657342657342657, | |
| "eval_precision_weighted": 0.777820015465549, | |
| "eval_recall_macro": 0.7660714285714286, | |
| "eval_recall_micro": 0.7657342657342657, | |
| "eval_recall_weighted": 0.7657342657342657, | |
| "eval_runtime": 1.7527, | |
| "eval_samples_per_second": 326.349, | |
| "eval_steps_per_second": 20.539, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 10.619550704956055, | |
| "learning_rate": 9.641839665080363e-05, | |
| "loss": 0.9378, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 9.948221206665039, | |
| "learning_rate": 9.055104900071376e-05, | |
| "loss": 0.8129, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 7.1508331298828125, | |
| "learning_rate": 8.22794094596864e-05, | |
| "loss": 0.7151, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8479020979020979, | |
| "eval_f1_macro": 0.829473167087361, | |
| "eval_f1_micro": 0.8479020979020979, | |
| "eval_f1_weighted": 0.8300987833238631, | |
| "eval_loss": 0.5089119672775269, | |
| "eval_precision_macro": 0.836193234751419, | |
| "eval_precision_micro": 0.8479020979020979, | |
| "eval_precision_weighted": 0.8361691693898202, | |
| "eval_recall_macro": 0.8466836734693878, | |
| "eval_recall_micro": 0.8479020979020979, | |
| "eval_recall_weighted": 0.8479020979020979, | |
| "eval_runtime": 1.7558, | |
| "eval_samples_per_second": 325.77, | |
| "eval_steps_per_second": 20.503, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 2.9531726837158203, | |
| "learning_rate": 7.209390765564318e-05, | |
| "loss": 0.5656, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 10.758095741271973, | |
| "learning_rate": 6.0598447063242855e-05, | |
| "loss": 0.5336, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 4.116607666015625, | |
| "learning_rate": 4.847459926610619e-05, | |
| "loss": 0.5038, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8531468531468531, | |
| "eval_f1_macro": 0.846184381645929, | |
| "eval_f1_micro": 0.8531468531468531, | |
| "eval_f1_weighted": 0.8463781132711151, | |
| "eval_loss": 0.5155654549598694, | |
| "eval_precision_macro": 0.8566878178421709, | |
| "eval_precision_micro": 0.8531468531468531, | |
| "eval_precision_weighted": 0.8564943001399524, | |
| "eval_recall_macro": 0.8526360544217687, | |
| "eval_recall_micro": 0.8531468531468531, | |
| "eval_recall_weighted": 0.8531468531468531, | |
| "eval_runtime": 1.7505, | |
| "eval_samples_per_second": 326.761, | |
| "eval_steps_per_second": 20.565, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 1.0205471515655518, | |
| "learning_rate": 3.644119323817915e-05, | |
| "loss": 0.4418, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 3.8727076053619385, | |
| "learning_rate": 2.5211695615868458e-05, | |
| "loss": 0.4395, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 6.586015224456787, | |
| "learning_rate": 1.5540241344564915e-05, | |
| "loss": 0.3997, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8479020979020979, | |
| "eval_f1_macro": 0.8387454620589124, | |
| "eval_f1_micro": 0.8479020979020979, | |
| "eval_f1_weighted": 0.839048143374511, | |
| "eval_loss": 0.46745064854621887, | |
| "eval_precision_macro": 0.8497877221797597, | |
| "eval_precision_micro": 0.8479020979020979, | |
| "eval_precision_weighted": 0.849814874011776, | |
| "eval_recall_macro": 0.8473639455782312, | |
| "eval_recall_micro": 0.8479020979020979, | |
| "eval_recall_weighted": 0.8479020979020979, | |
| "eval_runtime": 1.7526, | |
| "eval_samples_per_second": 326.368, | |
| "eval_steps_per_second": 20.541, | |
| "step": 1144 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 1430, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 1210783301566464.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |