| { | |
| "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": 0.5058665871620178, | |
| "eval_loss": 0.6931472420692444, | |
| "eval_runtime": 9.016, | |
| "eval_samples_per_second": 330.857, | |
| "eval_steps_per_second": 20.741, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.4958095848560333, | |
| "eval_loss": 0.6931472420692444, | |
| "eval_runtime": 9.075, | |
| "eval_samples_per_second": 328.705, | |
| "eval_steps_per_second": 20.606, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 2.6737967914438503, | |
| "grad_norm": 0.6687777638435364, | |
| "learning_rate": 4.2187622155535154e-05, | |
| "loss": 0.6988, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.48105934262275696, | |
| "eval_loss": 0.6931472420692444, | |
| "eval_runtime": 8.941, | |
| "eval_samples_per_second": 333.63, | |
| "eval_steps_per_second": 20.915, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5035199522972107, | |
| "eval_loss": 0.6931472420692444, | |
| "eval_runtime": 9.1554, | |
| "eval_samples_per_second": 325.818, | |
| "eval_steps_per_second": 20.425, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.49547436833381653, | |
| "eval_loss": 0.6931472420692444, | |
| "eval_runtime": 9.1937, | |
| "eval_samples_per_second": 324.463, | |
| "eval_steps_per_second": 20.34, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 5.347593582887701, | |
| "grad_norm": 0.7034493684768677, | |
| "learning_rate": 2.6790679763004077e-05, | |
| "loss": 0.6966, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.4904458522796631, | |
| "eval_loss": 0.6931472420692444, | |
| "eval_runtime": 9.0228, | |
| "eval_samples_per_second": 330.605, | |
| "eval_steps_per_second": 20.725, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.525980532169342, | |
| "eval_loss": 0.6931472420692444, | |
| "eval_runtime": 9.194, | |
| "eval_samples_per_second": 324.452, | |
| "eval_steps_per_second": 20.339, | |
| "step": 1309 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5005028247833252, | |
| "eval_loss": 0.6931472420692444, | |
| "eval_runtime": 9.1651, | |
| "eval_samples_per_second": 325.475, | |
| "eval_steps_per_second": 20.404, | |
| "step": 1496 | |
| }, | |
| { | |
| "epoch": 8.02139037433155, | |
| "grad_norm": 0.7838780283927917, | |
| "learning_rate": 1.1393737370472999e-05, | |
| "loss": 0.6959, | |
| "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": 5.758456454806623e-05, | |
| "per_device_train_batch_size": 64 | |
| } | |
| } | |