| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1176, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 1.914965986394558e-05, | |
| "loss": 0.2489, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 1.8299319727891158e-05, | |
| "loss": 0.0172, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 1.7448979591836738e-05, | |
| "loss": 0.0369, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 1.6598639455782314e-05, | |
| "loss": 0.0152, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.5748299319727894e-05, | |
| "loss": 0.0152, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 1.4897959183673472e-05, | |
| "loss": 0.0234, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.4047619047619048e-05, | |
| "loss": 0.0064, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 1.3197278911564626e-05, | |
| "loss": 0.0154, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 1.2346938775510204e-05, | |
| "loss": 0.011, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 1.1496598639455783e-05, | |
| "loss": 0.0123, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "eval_accuracy": 0.9990436722983742, | |
| "eval_loss": 0.0037615273613482714, | |
| "eval_runtime": 12.2684, | |
| "eval_samples_per_second": 255.697, | |
| "eval_steps_per_second": 16.057, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 1.064625850340136e-05, | |
| "loss": 0.0086, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 9.795918367346939e-06, | |
| "loss": 0.0043, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 8.945578231292518e-06, | |
| "loss": 0.0015, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 8.095238095238097e-06, | |
| "loss": 0.0052, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 7.244897959183675e-06, | |
| "loss": 0.0005, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 6.394557823129253e-06, | |
| "loss": 0.0039, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 5.54421768707483e-06, | |
| "loss": 0.0054, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 4.693877551020409e-06, | |
| "loss": 0.0002, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 3.843537414965986e-06, | |
| "loss": 0.0003, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 2.993197278911565e-06, | |
| "loss": 0.0002, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "eval_accuracy": 0.9987248963978324, | |
| "eval_loss": 0.005804423708468676, | |
| "eval_runtime": 12.2849, | |
| "eval_samples_per_second": 255.355, | |
| "eval_steps_per_second": 16.036, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "learning_rate": 2.1428571428571427e-06, | |
| "loss": 0.0044, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 1.2925170068027212e-06, | |
| "loss": 0.0039, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 4.421768707482994e-07, | |
| "loss": 0.0027, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1176, | |
| "total_flos": 3649943044058136.0, | |
| "train_loss": 0.01887278282083571, | |
| "train_runtime": 511.3764, | |
| "train_samples_per_second": 73.484, | |
| "train_steps_per_second": 2.3 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 1176, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 3649943044058136.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |