| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "global_step": 6762, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 4.9992605737947355e-05, | |
| "loss": 3.6847, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.852114758947057e-05, | |
| "loss": 3.4099, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.704229517894114e-05, | |
| "loss": 3.2337, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.5563442768411715e-05, | |
| "loss": 3.0788, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.408459035788228e-05, | |
| "loss": 3.1394, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 4.260573794735286e-05, | |
| "loss": 3.0528, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "eval_accuracy": 0.22231172025203705, | |
| "eval_loss": 3.0265257358551025, | |
| "eval_runtime": 7.2925, | |
| "eval_samples_per_second": 680.975, | |
| "eval_steps_per_second": 85.156, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 4.112688553682343e-05, | |
| "loss": 2.9704, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 3.9648033126293996e-05, | |
| "loss": 3.0544, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 3.816918071576457e-05, | |
| "loss": 2.9979, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 3.669032830523514e-05, | |
| "loss": 3.0237, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 3.521147589470571e-05, | |
| "loss": 2.9836, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "eval_accuracy": 0.23318566381931305, | |
| "eval_loss": 2.926269054412842, | |
| "eval_runtime": 7.3605, | |
| "eval_samples_per_second": 674.678, | |
| "eval_steps_per_second": 84.369, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 3.3732623484176276e-05, | |
| "loss": 2.9896, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 3.225377107364685e-05, | |
| "loss": 2.8578, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 3.077491866311743e-05, | |
| "loss": 2.7667, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 2.9296066252587996e-05, | |
| "loss": 2.8017, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 2.7817213842058564e-05, | |
| "loss": 2.7409, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "eval_accuracy": 0.2533226013183594, | |
| "eval_loss": 2.9040770530700684, | |
| "eval_runtime": 7.3851, | |
| "eval_samples_per_second": 672.437, | |
| "eval_steps_per_second": 84.088, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 2.6338361431529136e-05, | |
| "loss": 2.7605, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 2.4859509020999705e-05, | |
| "loss": 2.7569, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 2.3380656610470276e-05, | |
| "loss": 2.7154, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 2.190180419994085e-05, | |
| "loss": 2.788, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 2.0422951789411417e-05, | |
| "loss": 2.7905, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "eval_accuracy": 0.2605718970298767, | |
| "eval_loss": 2.876274824142456, | |
| "eval_runtime": 7.4268, | |
| "eval_samples_per_second": 668.66, | |
| "eval_steps_per_second": 83.616, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 1.894409937888199e-05, | |
| "loss": 2.7044, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 1.7465246968352557e-05, | |
| "loss": 2.7526, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 1.5986394557823133e-05, | |
| "loss": 2.6027, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 1.4507542147293701e-05, | |
| "loss": 2.4391, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 1.3028689736764271e-05, | |
| "loss": 2.4359, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "eval_accuracy": 0.2641965448856354, | |
| "eval_loss": 2.9071688652038574, | |
| "eval_runtime": 7.4381, | |
| "eval_samples_per_second": 667.644, | |
| "eval_steps_per_second": 83.489, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 1.1549837326234843e-05, | |
| "loss": 2.4514, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 1.0070984915705414e-05, | |
| "loss": 2.4754, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 8.592132505175984e-06, | |
| "loss": 2.4698, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 7.113280094646555e-06, | |
| "loss": 2.4732, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 5.634427684117125e-06, | |
| "loss": 2.4507, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "eval_accuracy": 0.26439791917800903, | |
| "eval_loss": 2.9230332374572754, | |
| "eval_runtime": 7.4558, | |
| "eval_samples_per_second": 666.058, | |
| "eval_steps_per_second": 83.291, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "learning_rate": 4.155575273587696e-06, | |
| "loss": 2.5628, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "learning_rate": 2.6767228630582667e-06, | |
| "loss": 2.3556, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 1.1978704525288378e-06, | |
| "loss": 2.4206, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 6762, | |
| "total_flos": 3412474671843072.0, | |
| "train_loss": 2.764526012873939, | |
| "train_runtime": 382.9395, | |
| "train_samples_per_second": 141.242, | |
| "train_steps_per_second": 17.658 | |
| } | |
| ], | |
| "max_steps": 6762, | |
| "num_train_epochs": 3, | |
| "total_flos": 3412474671843072.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |