| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.98989898989899, | |
| "eval_steps": 500, | |
| "global_step": 1235, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 4.1816935539245605, | |
| "learning_rate": 9.206477732793523e-06, | |
| "loss": 2.284, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 7.234564304351807, | |
| "learning_rate": 8.39676113360324e-06, | |
| "loss": 1.1378, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 4.295170783996582, | |
| "learning_rate": 7.587044534412956e-06, | |
| "loss": 0.7549, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 2.47021746635437, | |
| "learning_rate": 6.777327935222673e-06, | |
| "loss": 0.545, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 4.9068498611450195, | |
| "learning_rate": 5.9676113360323896e-06, | |
| "loss": 0.4039, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "eval_accuracy": 0.8949494957923889, | |
| "eval_loss": 0.3197959065437317, | |
| "eval_runtime": 43.5773, | |
| "eval_samples_per_second": 11.359, | |
| "eval_steps_per_second": 2.846, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 10.178821563720703, | |
| "learning_rate": 5.157894736842106e-06, | |
| "loss": 0.3572, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 22.03152847290039, | |
| "learning_rate": 4.348178137651822e-06, | |
| "loss": 0.3237, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 4.856521129608154, | |
| "learning_rate": 3.538461538461539e-06, | |
| "loss": 0.2635, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 4.2341766357421875, | |
| "learning_rate": 2.7368421052631583e-06, | |
| "loss": 0.2138, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 0.4156922996044159, | |
| "learning_rate": 1.9271255060728746e-06, | |
| "loss": 0.2195, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "eval_accuracy": 0.9333333373069763, | |
| "eval_loss": 0.2362980842590332, | |
| "eval_runtime": 43.7442, | |
| "eval_samples_per_second": 11.316, | |
| "eval_steps_per_second": 2.835, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "grad_norm": 5.4564337730407715, | |
| "learning_rate": 1.1174089068825912e-06, | |
| "loss": 0.2312, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "grad_norm": 2.0700230598449707, | |
| "learning_rate": 3.0769230769230774e-07, | |
| "loss": 0.182, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "step": 1235, | |
| "total_flos": 1.053348926143392e+18, | |
| "train_loss": 0.5674189640925481, | |
| "train_runtime": 1664.3033, | |
| "train_samples_per_second": 5.945, | |
| "train_steps_per_second": 0.742 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 1235, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 1.053348926143392e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |