| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 8.02139037433155, | |
| "eval_steps": 500, | |
| "global_step": 1500, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7683539986610413, | |
| "eval_loss": 0.4618770182132721, | |
| "eval_runtime": 9.2947, | |
| "eval_samples_per_second": 320.936, | |
| "eval_steps_per_second": 20.119, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7971840500831604, | |
| "eval_loss": 0.42974236607551575, | |
| "eval_runtime": 9.3349, | |
| "eval_samples_per_second": 319.554, | |
| "eval_steps_per_second": 20.032, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 2.6737967914438503, | |
| "grad_norm": 16.69818115234375, | |
| "learning_rate": 8.87708337266305e-06, | |
| "loss": 0.4407, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8159570693969727, | |
| "eval_loss": 0.4164794683456421, | |
| "eval_runtime": 9.288, | |
| "eval_samples_per_second": 321.167, | |
| "eval_steps_per_second": 20.134, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8129400014877319, | |
| "eval_loss": 0.5053750276565552, | |
| "eval_runtime": 9.4534, | |
| "eval_samples_per_second": 315.549, | |
| "eval_steps_per_second": 19.781, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8099228739738464, | |
| "eval_loss": 0.590611457824707, | |
| "eval_runtime": 9.4637, | |
| "eval_samples_per_second": 315.205, | |
| "eval_steps_per_second": 19.76, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 5.347593582887701, | |
| "grad_norm": 18.07274627685547, | |
| "learning_rate": 5.637271922786024e-06, | |
| "loss": 0.1629, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8132752180099487, | |
| "eval_loss": 0.6388338804244995, | |
| "eval_runtime": 9.4591, | |
| "eval_samples_per_second": 315.357, | |
| "eval_steps_per_second": 19.769, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.8105933666229248, | |
| "eval_loss": 0.7334861159324646, | |
| "eval_runtime": 9.4671, | |
| "eval_samples_per_second": 315.09, | |
| "eval_steps_per_second": 19.753, | |
| "step": 1309 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8162923455238342, | |
| "eval_loss": 0.7645613551139832, | |
| "eval_runtime": 9.4764, | |
| "eval_samples_per_second": 314.781, | |
| "eval_steps_per_second": 19.733, | |
| "step": 1496 | |
| }, | |
| { | |
| "epoch": 8.02139037433155, | |
| "grad_norm": 7.866405963897705, | |
| "learning_rate": 2.3974604729089986e-06, | |
| "loss": 0.0626, | |
| "step": 1500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 1870, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 1.7876643808636188e+16, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 1.2116894822540074e-05, | |
| "per_device_train_batch_size": 64 | |
| } | |
| } | |