| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 90, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 25.66914681799565, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.0048, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 12.156020748334926, | |
| "learning_rate": 9.986966157589751e-06, | |
| "loss": 0.9831, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.652240514755249, | |
| "eval_runtime": 24.1162, | |
| "eval_samples_per_second": 10.74, | |
| "eval_steps_per_second": 0.373, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 3.8285863774272446, | |
| "learning_rate": 9.841114703012817e-06, | |
| "loss": 0.7569, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 2.607435400324957, | |
| "learning_rate": 9.537877098354787e-06, | |
| "loss": 0.601, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.4056125581264496, | |
| "eval_runtime": 24.0279, | |
| "eval_samples_per_second": 10.779, | |
| "eval_steps_per_second": 0.375, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 3.788253489544647, | |
| "learning_rate": 9.08711169279446e-06, | |
| "loss": 0.4813, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.7777777777777777, | |
| "grad_norm": 2.392342802419934, | |
| "learning_rate": 8.503473010366713e-06, | |
| "loss": 0.3464, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.18872837722301483, | |
| "eval_runtime": 24.0055, | |
| "eval_samples_per_second": 10.789, | |
| "eval_steps_per_second": 0.375, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 5.610647726544286, | |
| "learning_rate": 7.805935326811913e-06, | |
| "loss": 0.2162, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.888888888888889, | |
| "grad_norm": 2.516026593621024, | |
| "learning_rate": 7.017175809949044e-06, | |
| "loss": 0.1559, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.07645933330059052, | |
| "eval_runtime": 24.0333, | |
| "eval_samples_per_second": 10.777, | |
| "eval_steps_per_second": 0.374, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "grad_norm": 3.496241200147731, | |
| "learning_rate": 6.162837277871553e-06, | |
| "loss": 0.0797, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.562094511651351, | |
| "learning_rate": 5.270694542927089e-06, | |
| "loss": 0.0714, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.047790270298719406, | |
| "eval_runtime": 24.1049, | |
| "eval_samples_per_second": 10.745, | |
| "eval_steps_per_second": 0.373, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 5.555555555555555, | |
| "grad_norm": 2.038884810074137, | |
| "learning_rate": 4.369751443898554e-06, | |
| "loss": 0.0477, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.035123735666275024, | |
| "eval_runtime": 24.0485, | |
| "eval_samples_per_second": 10.77, | |
| "eval_steps_per_second": 0.374, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 6.111111111111111, | |
| "grad_norm": 1.2084030161409793, | |
| "learning_rate": 3.489297922152136e-06, | |
| "loss": 0.0452, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 1.5950605501326323, | |
| "learning_rate": 2.65795779650105e-06, | |
| "loss": 0.0345, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.02559623494744301, | |
| "eval_runtime": 24.0003, | |
| "eval_samples_per_second": 10.792, | |
| "eval_steps_per_second": 0.375, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 7.222222222222222, | |
| "grad_norm": 0.7529091490914854, | |
| "learning_rate": 1.9027581939213852e-06, | |
| "loss": 0.0295, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 7.777777777777778, | |
| "grad_norm": 1.569990631190196, | |
| "learning_rate": 1.2482508892179884e-06, | |
| "loss": 0.0252, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.019161375239491463, | |
| "eval_runtime": 24.061, | |
| "eval_samples_per_second": 10.764, | |
| "eval_steps_per_second": 0.374, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 0.674086772310731, | |
| "learning_rate": 7.157141191620548e-07, | |
| "loss": 0.0198, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 8.88888888888889, | |
| "grad_norm": 0.5564084869849035, | |
| "learning_rate": 3.224608203719953e-07, | |
| "loss": 0.0181, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.015782972797751427, | |
| "eval_runtime": 23.9574, | |
| "eval_samples_per_second": 10.811, | |
| "eval_steps_per_second": 0.376, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 9.444444444444445, | |
| "grad_norm": 0.402959494753, | |
| "learning_rate": 8.127578033998663e-08, | |
| "loss": 0.0156, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.3564826816641896, | |
| "learning_rate": 0.0, | |
| "loss": 0.0153, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.015113583765923977, | |
| "eval_runtime": 23.9569, | |
| "eval_samples_per_second": 10.811, | |
| "eval_steps_per_second": 0.376, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 90, | |
| "total_flos": 18844169011200.0, | |
| "train_loss": 0.21927920257051786, | |
| "train_runtime": 1273.3046, | |
| "train_samples_per_second": 2.034, | |
| "train_steps_per_second": 0.071 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 90, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 18844169011200.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |