| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 30, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 4.817715644836426, | |
| "learning_rate": 0.0, | |
| "loss": 2.4596, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 4.894565105438232, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 2.468, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 4.599151134490967, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 2.4369, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 4.634383678436279, | |
| "learning_rate": 0.0001, | |
| "loss": 2.4219, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 4.405484676361084, | |
| "learning_rate": 9.997377845227576e-05, | |
| "loss": 2.3884, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 4.227508068084717, | |
| "learning_rate": 9.989514131188559e-05, | |
| "loss": 2.3539, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 3.989192247390747, | |
| "learning_rate": 9.97641710583307e-05, | |
| "loss": 2.3132, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.904679775238037, | |
| "learning_rate": 9.958100506132127e-05, | |
| "loss": 2.2919, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 3.8267431259155273, | |
| "learning_rate": 9.934583543669453e-05, | |
| "loss": 2.2709, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 3.6127519607543945, | |
| "learning_rate": 9.905890884491195e-05, | |
| "loss": 2.2313, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.1856, | |
| "eval_loss": 2.2107839584350586, | |
| "eval_runtime": 18.3869, | |
| "eval_samples_per_second": 543.864, | |
| "eval_steps_per_second": 0.109, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 3.6324033737182617, | |
| "learning_rate": 9.872052623234632e-05, | |
| "loss": 2.2162, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.4379923343658447, | |
| "learning_rate": 9.833104251563056e-05, | |
| "loss": 2.1835, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 3.394665479660034, | |
| "learning_rate": 9.789086620939936e-05, | |
| "loss": 2.1677, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 3.3317911624908447, | |
| "learning_rate": 9.740045899781352e-05, | |
| "loss": 2.1358, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 3.2182583808898926, | |
| "learning_rate": 9.686033525031719e-05, | |
| "loss": 2.1124, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 3.1929690837860107, | |
| "learning_rate": 9.627106148213522e-05, | |
| "loss": 2.0794, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 3.1744496822357178, | |
| "learning_rate": 9.563325576007701e-05, | |
| "loss": 2.0664, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 3.2177734375, | |
| "learning_rate": 9.494758705426978e-05, | |
| "loss": 2.0374, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 3.143730640411377, | |
| "learning_rate": 9.421477453650118e-05, | |
| "loss": 2.0087, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.086007595062256, | |
| "learning_rate": 9.343558682590756e-05, | |
| "loss": 1.9995, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.3659, | |
| "eval_loss": 1.9728946685791016, | |
| "eval_runtime": 18.2892, | |
| "eval_samples_per_second": 546.77, | |
| "eval_steps_per_second": 0.109, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 3.093388795852661, | |
| "learning_rate": 9.261084118279847e-05, | |
| "loss": 1.9768, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 3.0235588550567627, | |
| "learning_rate": 9.174140265146356e-05, | |
| "loss": 1.9573, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 3.054340362548828, | |
| "learning_rate": 9.082818315286055e-05, | |
| "loss": 1.9279, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 2.977816104888916, | |
| "learning_rate": 8.987214052813604e-05, | |
| "loss": 1.9009, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 2.974107027053833, | |
| "learning_rate": 8.887427753398248e-05, | |
| "loss": 1.8879, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 2.926490545272827, | |
| "learning_rate": 8.783564079088477e-05, | |
| "loss": 1.8672, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 2.8999850749969482, | |
| "learning_rate": 8.675731968536002e-05, | |
| "loss": 1.8563, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 2.8928842544555664, | |
| "learning_rate": 8.564044522734147e-05, | |
| "loss": 1.832, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 2.8881161212921143, | |
| "learning_rate": 8.448618886390522e-05, | |
| "loss": 1.8051, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.852555274963379, | |
| "learning_rate": 8.329576125058406e-05, | |
| "loss": 1.7828, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5687, | |
| "eval_loss": 1.7709296941757202, | |
| "eval_runtime": 18.5173, | |
| "eval_samples_per_second": 540.036, | |
| "eval_steps_per_second": 0.108, | |
| "step": 30 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 100, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 1, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 512, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |