| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.98989898989899, | |
| "eval_steps": 500, | |
| "global_step": 1235, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.006182590499520302, | |
| "learning_rate": 9.206477732793523e-06, | |
| "loss": 0.0729, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 21.23886489868164, | |
| "learning_rate": 8.39676113360324e-06, | |
| "loss": 0.0698, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.006576939485967159, | |
| "learning_rate": 7.587044534412956e-06, | |
| "loss": 0.0464, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.009822617284953594, | |
| "learning_rate": 6.785425101214575e-06, | |
| "loss": 0.0689, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 0.00560363894328475, | |
| "learning_rate": 5.975708502024292e-06, | |
| "loss": 0.0228, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "eval_accuracy": 0.973737359046936, | |
| "eval_loss": 0.17728736996650696, | |
| "eval_runtime": 23.1344, | |
| "eval_samples_per_second": 21.397, | |
| "eval_steps_per_second": 5.36, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 0.010259171947836876, | |
| "learning_rate": 5.165991902834009e-06, | |
| "loss": 0.0253, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 0.5489621758460999, | |
| "learning_rate": 4.356275303643725e-06, | |
| "loss": 0.0227, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 0.003071287414059043, | |
| "learning_rate": 3.5465587044534415e-06, | |
| "loss": 0.0398, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 0.0034055381547659636, | |
| "learning_rate": 2.7368421052631583e-06, | |
| "loss": 0.0299, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 0.005517472513020039, | |
| "learning_rate": 1.9271255060728746e-06, | |
| "loss": 0.0385, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "eval_accuracy": 0.9696969985961914, | |
| "eval_loss": 0.190963014960289, | |
| "eval_runtime": 23.4329, | |
| "eval_samples_per_second": 21.124, | |
| "eval_steps_per_second": 5.292, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "grad_norm": 19.359519958496094, | |
| "learning_rate": 1.1174089068825912e-06, | |
| "loss": 0.0365, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "grad_norm": 1.323317527770996, | |
| "learning_rate": 3.0769230769230774e-07, | |
| "loss": 0.0064, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "step": 1235, | |
| "total_flos": 3.160649695892544e+17, | |
| "train_loss": 0.03992950008948322, | |
| "train_runtime": 666.1547, | |
| "train_samples_per_second": 14.854, | |
| "train_steps_per_second": 1.854 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 1235, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 3.160649695892544e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |