| { | |
| "best_metric": 0.09647562354803085, | |
| "best_model_checkpoint": "outputs/checkpoint-329", | |
| "epoch": 5.995444191343964, | |
| "eval_steps": 500, | |
| "global_step": 329, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.36446469248291574, | |
| "grad_norm": 2.546876907348633, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 0.373, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7289293849658315, | |
| "grad_norm": 1.2323638200759888, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 0.2883, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.9840546697038725, | |
| "eval_loss": 0.1884891539812088, | |
| "eval_runtime": 48.9142, | |
| "eval_samples_per_second": 3.435, | |
| "eval_steps_per_second": 0.429, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.0933940774487472, | |
| "grad_norm": 0.5503110289573669, | |
| "learning_rate": 5.8e-06, | |
| "loss": 0.1829, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.4578587699316627, | |
| "grad_norm": 0.3597412109375, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 0.1368, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.8223234624145785, | |
| "grad_norm": 1.4852937459945679, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 0.1263, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.9863325740318907, | |
| "eval_loss": 0.13872365653514862, | |
| "eval_runtime": 48.9202, | |
| "eval_samples_per_second": 3.434, | |
| "eval_steps_per_second": 0.429, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.1867881548974943, | |
| "grad_norm": 0.40878018736839294, | |
| "learning_rate": 9.984149663879994e-06, | |
| "loss": 0.1061, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.55125284738041, | |
| "grad_norm": 0.5748926997184753, | |
| "learning_rate": 9.929487384240103e-06, | |
| "loss": 0.0895, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.9157175398633255, | |
| "grad_norm": 0.3740793764591217, | |
| "learning_rate": 9.83624518217252e-06, | |
| "loss": 0.0879, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.988610478359909, | |
| "eval_loss": 0.11915451288223267, | |
| "eval_runtime": 48.9178, | |
| "eval_samples_per_second": 3.434, | |
| "eval_steps_per_second": 0.429, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 3.2801822323462413, | |
| "grad_norm": 0.5470688343048096, | |
| "learning_rate": 9.705152804330872e-06, | |
| "loss": 0.0851, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.644646924829157, | |
| "grad_norm": 0.5719650387763977, | |
| "learning_rate": 9.53723622631339e-06, | |
| "loss": 0.0773, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.990888382687927, | |
| "eval_loss": 0.1070408970117569, | |
| "eval_runtime": 48.9193, | |
| "eval_samples_per_second": 3.434, | |
| "eval_steps_per_second": 0.429, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 4.009111617312073, | |
| "grad_norm": 0.46819359064102173, | |
| "learning_rate": 9.333809623012763e-06, | |
| "loss": 0.0745, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.373576309794989, | |
| "grad_norm": 0.5022829174995422, | |
| "learning_rate": 9.096465083415809e-06, | |
| "loss": 0.0693, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 4.738041002277904, | |
| "grad_norm": 0.5556521415710449, | |
| "learning_rate": 8.82706015034849e-06, | |
| "loss": 0.0637, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.993166287015946, | |
| "eval_loss": 0.09954136610031128, | |
| "eval_runtime": 48.9147, | |
| "eval_samples_per_second": 3.435, | |
| "eval_steps_per_second": 0.429, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 5.10250569476082, | |
| "grad_norm": 0.46870681643486023, | |
| "learning_rate": 8.527703282684766e-06, | |
| "loss": 0.0634, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.466970387243736, | |
| "grad_norm": 0.615917980670929, | |
| "learning_rate": 8.20073735379715e-06, | |
| "loss": 0.0592, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.831435079726651, | |
| "grad_norm": 0.7252725958824158, | |
| "learning_rate": 7.848721315395975e-06, | |
| "loss": 0.0591, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 5.995444191343964, | |
| "eval_loss": 0.09647562354803085, | |
| "eval_runtime": 48.8589, | |
| "eval_samples_per_second": 3.438, | |
| "eval_steps_per_second": 0.43, | |
| "step": 329 | |
| } | |
| ], | |
| "logging_steps": 20, | |
| "max_steps": 810, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "save_steps": 500, | |
| "total_flos": 1.2319125158965248e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |