| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 45, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 1.7660930008866629, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.9294, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.9354323826192693, | |
| "learning_rate": 2e-05, | |
| "loss": 1.9738, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.7674917347322091, | |
| "learning_rate": 1.9238795325112867e-05, | |
| "loss": 2.0451, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.7260043665794342, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 1.9906, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.1539851427078247, | |
| "eval_runtime": 37.0305, | |
| "eval_samples_per_second": 9.884, | |
| "eval_steps_per_second": 0.621, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.8223509374284563, | |
| "learning_rate": 1.3826834323650899e-05, | |
| "loss": 1.8117, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 0.7687537061346872, | |
| "learning_rate": 1e-05, | |
| "loss": 1.7551, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.6174446103331057, | |
| "learning_rate": 6.173165676349103e-06, | |
| "loss": 1.7061, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.179561734199524, | |
| "eval_runtime": 38.0097, | |
| "eval_samples_per_second": 9.629, | |
| "eval_steps_per_second": 0.605, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 0.6930337012100228, | |
| "learning_rate": 2.9289321881345257e-06, | |
| "loss": 1.5538, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 0.6429180596496781, | |
| "learning_rate": 7.612046748871327e-07, | |
| "loss": 1.6341, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.5782810761819465, | |
| "learning_rate": 0.0, | |
| "loss": 1.6448, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.2119849920272827, | |
| "eval_runtime": 37.7435, | |
| "eval_samples_per_second": 9.697, | |
| "eval_steps_per_second": 0.609, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 45, | |
| "total_flos": 9369739591680.0, | |
| "train_loss": 1.789585214191013, | |
| "train_runtime": 1008.3081, | |
| "train_samples_per_second": 2.853, | |
| "train_steps_per_second": 0.045 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 45, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 9369739591680.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |