| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 22.857142857142858, | |
| "eval_steps": 100, | |
| "global_step": 800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 0.8547727465629578, | |
| "learning_rate": 4.962019382530521e-05, | |
| "loss": 0.2973, | |
| "num_input_tokens_seen": 1240440, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 16.666666666666668, | |
| "grad_norm": 0.5775834321975708, | |
| "learning_rate": 4.849231551964771e-05, | |
| "loss": 0.0165, | |
| "num_input_tokens_seen": 2467560, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 16.666666666666668, | |
| "eval_loss": 1.8753944635391235, | |
| "eval_runtime": 0.0881, | |
| "eval_samples_per_second": 669.817, | |
| "eval_steps_per_second": 22.706, | |
| "num_input_tokens_seen": 2467560, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 2.726205825805664, | |
| "learning_rate": 4.665063509461097e-05, | |
| "loss": 0.0117, | |
| "num_input_tokens_seen": 3714840, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 33.333333333333336, | |
| "grad_norm": 0.17272137105464935, | |
| "learning_rate": 4.415111107797445e-05, | |
| "loss": 0.0053, | |
| "num_input_tokens_seen": 4957560, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 33.333333333333336, | |
| "eval_loss": 2.176852226257324, | |
| "eval_runtime": 0.0894, | |
| "eval_samples_per_second": 660.304, | |
| "eval_steps_per_second": 22.383, | |
| "num_input_tokens_seen": 4957560, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 41.666666666666664, | |
| "grad_norm": 0.2948448061943054, | |
| "learning_rate": 4.1069690242163484e-05, | |
| "loss": 0.0054, | |
| "num_input_tokens_seen": 6181800, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 0.24040719866752625, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.0033, | |
| "num_input_tokens_seen": 7418880, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_loss": 2.383711814880371, | |
| "eval_runtime": 0.1033, | |
| "eval_samples_per_second": 571.316, | |
| "eval_steps_per_second": 19.367, | |
| "num_input_tokens_seen": 7418880, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 38.888888888888886, | |
| "grad_norm": 0.5190167427062988, | |
| "learning_rate": 4.215604094671835e-05, | |
| "loss": 0.3195, | |
| "num_input_tokens_seen": 13014840, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 44.44444444444444, | |
| "grad_norm": 0.37368133664131165, | |
| "learning_rate": 3.9928964792569655e-05, | |
| "loss": 0.0144, | |
| "num_input_tokens_seen": 18598920, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 44.44444444444444, | |
| "eval_loss": 0.9897297024726868, | |
| "eval_runtime": 1.0997, | |
| "eval_samples_per_second": 75.477, | |
| "eval_steps_per_second": 2.728, | |
| "num_input_tokens_seen": 18598920, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 0.15123282372951508, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 0.0082, | |
| "num_input_tokens_seen": 24251520, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 55.55555555555556, | |
| "grad_norm": 0.056949373334646225, | |
| "learning_rate": 3.490199415097892e-05, | |
| "loss": 0.0012, | |
| "num_input_tokens_seen": 29856960, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 55.55555555555556, | |
| "eval_loss": 1.1057275533676147, | |
| "eval_runtime": 1.1105, | |
| "eval_samples_per_second": 74.743, | |
| "eval_steps_per_second": 2.702, | |
| "num_input_tokens_seen": 29856960, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 15.714285714285714, | |
| "grad_norm": 1.0327025651931763, | |
| "learning_rate": 4.865818459497911e-05, | |
| "loss": 1.2219, | |
| "num_input_tokens_seen": 40827720, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 17.142857142857142, | |
| "grad_norm": 0.8191797137260437, | |
| "learning_rate": 4.8405871765993433e-05, | |
| "loss": 0.4354, | |
| "num_input_tokens_seen": 51641040, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 17.142857142857142, | |
| "eval_loss": 1.7303279638290405, | |
| "eval_runtime": 7.2384, | |
| "eval_samples_per_second": 48.63, | |
| "eval_steps_per_second": 1.52, | |
| "num_input_tokens_seen": 51641040, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 18.571428571428573, | |
| "grad_norm": 0.4633180797100067, | |
| "learning_rate": 4.813260751184992e-05, | |
| "loss": 0.0876, | |
| "num_input_tokens_seen": 62482320, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.23285113275051117, | |
| "learning_rate": 4.783863644106502e-05, | |
| "loss": 0.0268, | |
| "num_input_tokens_seen": 73404120, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 1.9282101392745972, | |
| "eval_runtime": 7.2488, | |
| "eval_samples_per_second": 48.56, | |
| "eval_steps_per_second": 1.517, | |
| "num_input_tokens_seen": 73404120, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 21.428571428571427, | |
| "grad_norm": 0.15514932572841644, | |
| "learning_rate": 4.752422169756048e-05, | |
| "loss": 0.0085, | |
| "num_input_tokens_seen": 84276120, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "grad_norm": 0.12508106231689453, | |
| "learning_rate": 4.718964472511386e-05, | |
| "loss": 0.0051, | |
| "num_input_tokens_seen": 95276640, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "eval_loss": 2.1112966537475586, | |
| "eval_runtime": 7.2618, | |
| "eval_samples_per_second": 48.473, | |
| "eval_steps_per_second": 1.515, | |
| "num_input_tokens_seen": 95276640, | |
| "step": 800 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 5250, | |
| "num_input_tokens_seen": 95276640, | |
| "num_train_epochs": 150, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.314993447114375e+18, | |
| "train_batch_size": 15, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |