| { | |
| "best_metric": 0.8997673516867003, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-22/checkpoint-1290", | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 2580, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.1603200435638428, | |
| "learning_rate": 0.00015804016249699036, | |
| "loss": 0.3253, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8945327646374563, | |
| "eval_f1": 0.19047619047619047, | |
| "eval_loss": 0.29357558488845825, | |
| "eval_mcc": 0.23285387927696083, | |
| "eval_precision": 0.6213592233009708, | |
| "eval_recall": 0.11247803163444639, | |
| "eval_runtime": 9.9704, | |
| "eval_samples_per_second": 517.33, | |
| "eval_steps_per_second": 16.248, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.3251060247421265, | |
| "learning_rate": 0.00011853012187274275, | |
| "loss": 0.2915, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8997673516867003, | |
| "eval_f1": 0.2385861561119293, | |
| "eval_loss": 0.2915143072605133, | |
| "eval_mcc": 0.2949928981799802, | |
| "eval_precision": 0.7363636363636363, | |
| "eval_recall": 0.14235500878734622, | |
| "eval_runtime": 9.9347, | |
| "eval_samples_per_second": 519.192, | |
| "eval_steps_per_second": 16.307, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.1212515830993652, | |
| "learning_rate": 7.902008124849518e-05, | |
| "loss": 0.278, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8993796044978674, | |
| "eval_f1": 0.23111111111111113, | |
| "eval_loss": 0.27244821190834045, | |
| "eval_mcc": 0.28922713720674353, | |
| "eval_precision": 0.7358490566037735, | |
| "eval_recall": 0.13708260105448156, | |
| "eval_runtime": 9.947, | |
| "eval_samples_per_second": 518.549, | |
| "eval_steps_per_second": 16.286, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.8891263008117676, | |
| "learning_rate": 3.951004062424759e-05, | |
| "loss": 0.2694, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8984102365257852, | |
| "eval_f1": 0.34825870646766166, | |
| "eval_loss": 0.26875007152557373, | |
| "eval_mcc": 0.338541834655946, | |
| "eval_precision": 0.5957446808510638, | |
| "eval_recall": 0.2460456942003515, | |
| "eval_runtime": 9.9622, | |
| "eval_samples_per_second": 517.756, | |
| "eval_steps_per_second": 16.261, | |
| "step": 2580 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 3225, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 6347770130400.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.9879592289455201, | |
| "learning_rate": 0.00019755020312123793, | |
| "num_train_epochs": 5, | |
| "temperature": 26 | |
| } | |
| } | |