| { | |
| "best_metric": 1.5478379726409912, | |
| "best_model_checkpoint": "/users/zyong2/data/zyong2/bigscience/data/processed/024/bloom-560m_az_continual-pretrain_100000samples_-1vocab_original-frozen/checkpoint-25000", | |
| "epoch": 2.194518203164432, | |
| "global_step": 25000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 9e-05, | |
| "loss": 2.2274, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 8e-05, | |
| "loss": 1.8241, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "eval_loss": 1.8055051565170288, | |
| "eval_runtime": 114.2042, | |
| "eval_samples_per_second": 7.101, | |
| "eval_steps_per_second": 3.555, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 7e-05, | |
| "loss": 1.7274, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 6e-05, | |
| "loss": 1.6599, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "eval_loss": 1.666320562362671, | |
| "eval_runtime": 114.1448, | |
| "eval_samples_per_second": 7.105, | |
| "eval_steps_per_second": 3.557, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 5e-05, | |
| "loss": 1.5609, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 4e-05, | |
| "loss": 1.4793, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "eval_loss": 1.6010019779205322, | |
| "eval_runtime": 114.2262, | |
| "eval_samples_per_second": 7.1, | |
| "eval_steps_per_second": 3.554, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4565, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4307, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "eval_loss": 1.549907922744751, | |
| "eval_runtime": 114.2215, | |
| "eval_samples_per_second": 7.1, | |
| "eval_steps_per_second": 3.554, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "learning_rate": 1e-05, | |
| "loss": 1.4124, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 0.0, | |
| "loss": 1.2636, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "eval_loss": 1.5478379726409912, | |
| "eval_runtime": 114.1947, | |
| "eval_samples_per_second": 7.102, | |
| "eval_steps_per_second": 3.555, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "step": 25000, | |
| "total_flos": 3.714902240494879e+17, | |
| "train_loss": 1.60421380859375, | |
| "train_runtime": 76890.5903, | |
| "train_samples_per_second": 2.601, | |
| "train_steps_per_second": 0.325 | |
| } | |
| ], | |
| "max_steps": 25000, | |
| "num_train_epochs": 3, | |
| "total_flos": 3.714902240494879e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |