| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 6936, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.639561707035756e-05, | |
| "loss": 0.8105, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 4.2791234140715114e-05, | |
| "loss": 0.5111, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 3.9186851211072664e-05, | |
| "loss": 0.4297, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 3.558246828143022e-05, | |
| "loss": 0.3814, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 3.1978085351787776e-05, | |
| "loss": 0.3232, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 2.8373702422145332e-05, | |
| "loss": 0.2746, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 2.4769319492502884e-05, | |
| "loss": 0.2602, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 2.116493656286044e-05, | |
| "loss": 0.239, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 1.7560553633217993e-05, | |
| "loss": 0.2268, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 1.395617070357555e-05, | |
| "loss": 0.1843, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 1.0351787773933102e-05, | |
| "loss": 0.1678, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 6.747404844290659e-06, | |
| "loss": 0.1586, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 3.143021914648212e-06, | |
| "loss": 0.1531, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 6936, | |
| "total_flos": 2.370754172808069e+17, | |
| "train_loss": 0.3064197311489365, | |
| "train_runtime": 5244.4956, | |
| "train_samples_per_second": 338.499, | |
| "train_steps_per_second": 1.323 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 6936, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 2.370754172808069e+17, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |