| { | |
| "best_metric": 0.686046511627907, | |
| "best_model_checkpoint": "deberta-transcript-classification/checkpoint-8126", | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 8126, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.12306177701206006, | |
| "grad_norm": 2.0186474323272705, | |
| "learning_rate": 1.9179588153252935e-05, | |
| "loss": 0.2543, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.24612355402412012, | |
| "grad_norm": 2.220743417739868, | |
| "learning_rate": 1.8359176306505868e-05, | |
| "loss": 0.1969, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.3691853310361802, | |
| "grad_norm": 2.504720449447632, | |
| "learning_rate": 1.75387644597588e-05, | |
| "loss": 0.1779, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.49224710804824023, | |
| "grad_norm": 0.3125078082084656, | |
| "learning_rate": 1.671835261301173e-05, | |
| "loss": 0.1607, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.6153088850603002, | |
| "grad_norm": 0.8753023147583008, | |
| "learning_rate": 1.5897940766264664e-05, | |
| "loss": 0.1519, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.7383706620723604, | |
| "grad_norm": 1.9338386058807373, | |
| "learning_rate": 1.50775289195176e-05, | |
| "loss": 0.1481, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.8614324390844204, | |
| "grad_norm": 4.1667304039001465, | |
| "learning_rate": 1.4257117072770533e-05, | |
| "loss": 0.1308, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.9844942160964805, | |
| "grad_norm": 1.195173978805542, | |
| "learning_rate": 1.3436705226023466e-05, | |
| "loss": 0.141, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5977653631284916, | |
| "eval_f1": 0.6445783132530121, | |
| "eval_loss": 0.12648048996925354, | |
| "eval_roc_auc": 0.7908519553072625, | |
| "eval_runtime": 8.0537, | |
| "eval_samples_per_second": 44.452, | |
| "eval_steps_per_second": 44.452, | |
| "step": 4063 | |
| }, | |
| { | |
| "epoch": 1.1075559931085406, | |
| "grad_norm": 5.743582248687744, | |
| "learning_rate": 1.26162933792764e-05, | |
| "loss": 0.1232, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.2306177701206005, | |
| "grad_norm": 1.529942274093628, | |
| "learning_rate": 1.1795881532529329e-05, | |
| "loss": 0.111, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.3536795471326606, | |
| "grad_norm": 0.2728979289531708, | |
| "learning_rate": 1.0975469685782262e-05, | |
| "loss": 0.1241, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.4767413241447207, | |
| "grad_norm": 0.21842879056930542, | |
| "learning_rate": 1.0155057839035196e-05, | |
| "loss": 0.1185, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.5998031011567808, | |
| "grad_norm": 0.08660631626844406, | |
| "learning_rate": 9.334645992288129e-06, | |
| "loss": 0.1276, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.7228648781688407, | |
| "grad_norm": 0.07522659748792648, | |
| "learning_rate": 8.514234145541062e-06, | |
| "loss": 0.1116, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.8459266551809008, | |
| "grad_norm": 0.2456115484237671, | |
| "learning_rate": 7.693822298793995e-06, | |
| "loss": 0.1065, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.968988432192961, | |
| "grad_norm": 0.16100919246673584, | |
| "learning_rate": 6.873410452046928e-06, | |
| "loss": 0.1113, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6564245810055865, | |
| "eval_f1": 0.686046511627907, | |
| "eval_loss": 0.1288219690322876, | |
| "eval_roc_auc": 0.8214036312849162, | |
| "eval_runtime": 7.9308, | |
| "eval_samples_per_second": 45.141, | |
| "eval_steps_per_second": 45.141, | |
| "step": 8126 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 12189, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 2138366728169472.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |