| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 20, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 0.0001, | |
| "loss": 2.9957, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 3.1984338760375977, | |
| "eval_runtime": 0.465, | |
| "eval_samples_per_second": 21.504, | |
| "eval_steps_per_second": 2.15, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 3.0672874450683594, | |
| "eval_runtime": 0.4711, | |
| "eval_samples_per_second": 21.225, | |
| "eval_steps_per_second": 2.122, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 2.8525, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 2.9477875232696533, | |
| "eval_runtime": 0.4781, | |
| "eval_samples_per_second": 20.915, | |
| "eval_steps_per_second": 2.091, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.8806045055389404, | |
| "eval_runtime": 0.4817, | |
| "eval_samples_per_second": 20.76, | |
| "eval_steps_per_second": 2.076, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 2.3323, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 2.83823823928833, | |
| "eval_runtime": 0.4837, | |
| "eval_samples_per_second": 20.676, | |
| "eval_steps_per_second": 2.068, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.8050312995910645, | |
| "eval_runtime": 0.4968, | |
| "eval_samples_per_second": 20.128, | |
| "eval_steps_per_second": 2.013, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 2.7636163234710693, | |
| "eval_runtime": 0.5049, | |
| "eval_samples_per_second": 19.805, | |
| "eval_steps_per_second": 1.98, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 1.125, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 2.0887, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.7495064735412598, | |
| "eval_runtime": 0.5092, | |
| "eval_samples_per_second": 19.638, | |
| "eval_steps_per_second": 1.964, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 2.743269681930542, | |
| "eval_runtime": 0.5168, | |
| "eval_samples_per_second": 19.35, | |
| "eval_steps_per_second": 1.935, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.9375, | |
| "learning_rate": 0.0, | |
| "loss": 1.9997, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 2.742553472518921, | |
| "eval_runtime": 0.5167, | |
| "eval_samples_per_second": 19.353, | |
| "eval_steps_per_second": 1.935, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 20, | |
| "total_flos": 1.56632893882368e+16, | |
| "train_loss": 2.325469994544983, | |
| "train_runtime": 70.3098, | |
| "train_samples_per_second": 14.792, | |
| "train_steps_per_second": 0.284 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "total_flos": 1.56632893882368e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |