| { |
| "best_global_step": 2844, |
| "best_metric": 0.3881676495075226, |
| "best_model_checkpoint": "distilbert_km_5_v2_qqp/checkpoint-2844", |
| "epoch": 7.0, |
| "eval_steps": 500, |
| "global_step": 9954, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 1.0, |
| "grad_norm": 2.0553817749023438, |
| "learning_rate": 4.9e-05, |
| "loss": 0.466, |
| "step": 1422 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.7976007914914668, |
| "eval_combined_score": 0.7471678863879896, |
| "eval_f1": 0.6967349812845125, |
| "eval_loss": 0.42404308915138245, |
| "eval_runtime": 16.6887, |
| "eval_samples_per_second": 2422.598, |
| "eval_steps_per_second": 9.467, |
| "step": 1422 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 2.685347557067871, |
| "learning_rate": 4.8e-05, |
| "loss": 0.3554, |
| "step": 2844 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.8230521889685877, |
| "eval_combined_score": 0.7867267085377444, |
| "eval_f1": 0.7504012281069011, |
| "eval_loss": 0.3881676495075226, |
| "eval_runtime": 16.3426, |
| "eval_samples_per_second": 2473.896, |
| "eval_steps_per_second": 9.668, |
| "step": 2844 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 3.180206298828125, |
| "learning_rate": 4.7e-05, |
| "loss": 0.2749, |
| "step": 4266 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.8299282710858273, |
| "eval_combined_score": 0.7979935312968603, |
| "eval_f1": 0.7660587915078934, |
| "eval_loss": 0.40492284297943115, |
| "eval_runtime": 16.1677, |
| "eval_samples_per_second": 2500.66, |
| "eval_steps_per_second": 9.773, |
| "step": 4266 |
| }, |
| { |
| "epoch": 4.0, |
| "grad_norm": 2.9606895446777344, |
| "learning_rate": 4.600000000000001e-05, |
| "loss": 0.2101, |
| "step": 5688 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_accuracy": 0.8358891911946574, |
| "eval_combined_score": 0.7983993539844139, |
| "eval_f1": 0.7609095167741703, |
| "eval_loss": 0.46142277121543884, |
| "eval_runtime": 16.5023, |
| "eval_samples_per_second": 2449.966, |
| "eval_steps_per_second": 9.574, |
| "step": 5688 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 2.8550710678100586, |
| "learning_rate": 4.5e-05, |
| "loss": 0.1628, |
| "step": 7110 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_accuracy": 0.839500371011625, |
| "eval_combined_score": 0.8082821537608714, |
| "eval_f1": 0.7770639365101178, |
| "eval_loss": 0.49306216835975647, |
| "eval_runtime": 16.1553, |
| "eval_samples_per_second": 2502.585, |
| "eval_steps_per_second": 9.78, |
| "step": 7110 |
| }, |
| { |
| "epoch": 6.0, |
| "grad_norm": 3.561419725418091, |
| "learning_rate": 4.4000000000000006e-05, |
| "loss": 0.1295, |
| "step": 8532 |
| }, |
| { |
| "epoch": 6.0, |
| "eval_accuracy": 0.8379421221864952, |
| "eval_combined_score": 0.8099382447449817, |
| "eval_f1": 0.781934367303468, |
| "eval_loss": 0.5082396268844604, |
| "eval_runtime": 16.3692, |
| "eval_samples_per_second": 2469.876, |
| "eval_steps_per_second": 9.652, |
| "step": 8532 |
| }, |
| { |
| "epoch": 7.0, |
| "grad_norm": 4.763936519622803, |
| "learning_rate": 4.3e-05, |
| "loss": 0.1044, |
| "step": 9954 |
| }, |
| { |
| "epoch": 7.0, |
| "eval_accuracy": 0.8408112787534009, |
| "eval_combined_score": 0.8121866455644863, |
| "eval_f1": 0.7835620123755717, |
| "eval_loss": 0.5691788792610168, |
| "eval_runtime": 16.3851, |
| "eval_samples_per_second": 2467.488, |
| "eval_steps_per_second": 9.643, |
| "step": 9954 |
| }, |
| { |
| "epoch": 7.0, |
| "step": 9954, |
| "total_flos": 1.686920659598684e+17, |
| "train_loss": 0.24329570442983034, |
| "train_runtime": 2425.7777, |
| "train_samples_per_second": 7499.574, |
| "train_steps_per_second": 29.31 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 71100, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 50, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "EarlyStoppingCallback": { |
| "args": { |
| "early_stopping_patience": 5, |
| "early_stopping_threshold": 0.0 |
| }, |
| "attributes": { |
| "early_stopping_patience_counter": 5 |
| } |
| }, |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.686920659598684e+17, |
| "train_batch_size": 256, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
