| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 3180, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.9968553459119497, | |
| "grad_norm": 0.316487580537796, | |
| "learning_rate": 0.0003250399324920357, | |
| "loss": 0.2047, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.867741935483871, | |
| "eval_loss": 0.06543365120887756, | |
| "eval_runtime": 5.3433, | |
| "eval_samples_per_second": 580.166, | |
| "eval_steps_per_second": 12.165, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 0.5043902397155762, | |
| "learning_rate": 0.00030152270079867525, | |
| "loss": 0.0555, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.9038709677419355, | |
| "eval_loss": 0.05041499063372612, | |
| "eval_runtime": 5.404, | |
| "eval_samples_per_second": 573.648, | |
| "eval_steps_per_second": 12.028, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 0.3954188823699951, | |
| "learning_rate": 0.0002648777146860182, | |
| "loss": 0.038, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9180645161290323, | |
| "eval_loss": 0.04048455134034157, | |
| "eval_runtime": 5.3969, | |
| "eval_samples_per_second": 574.404, | |
| "eval_steps_per_second": 12.044, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.5388094186782837, | |
| "learning_rate": 0.00021866970042254042, | |
| "loss": 0.029, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9296774193548387, | |
| "eval_loss": 0.03541301190853119, | |
| "eval_runtime": 5.4403, | |
| "eval_samples_per_second": 569.818, | |
| "eval_steps_per_second": 11.948, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.07852072268724442, | |
| "learning_rate": 0.00016739365008581398, | |
| "loss": 0.0235, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9309677419354838, | |
| "eval_loss": 0.03130786865949631, | |
| "eval_runtime": 5.3733, | |
| "eval_samples_per_second": 576.931, | |
| "eval_steps_per_second": 12.097, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.08189109712839127, | |
| "learning_rate": 0.0001160375607518124, | |
| "loss": 0.0202, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9364516129032258, | |
| "eval_loss": 0.02977406606078148, | |
| "eval_runtime": 5.37, | |
| "eval_samples_per_second": 577.282, | |
| "eval_steps_per_second": 12.104, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.07439889013767242, | |
| "learning_rate": 6.959721547615756e-05, | |
| "loss": 0.0181, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.94, | |
| "eval_loss": 0.02745823562145233, | |
| "eval_runtime": 5.4422, | |
| "eval_samples_per_second": 569.621, | |
| "eval_steps_per_second": 11.944, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.06082445755600929, | |
| "learning_rate": 3.25902068760846e-05, | |
| "loss": 0.0164, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.9409677419354838, | |
| "eval_loss": 0.0263341274112463, | |
| "eval_runtime": 5.4487, | |
| "eval_samples_per_second": 568.945, | |
| "eval_steps_per_second": 11.929, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.060345038771629333, | |
| "learning_rate": 8.616477799677371e-06, | |
| "loss": 0.0157, | |
| "step": 2853 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.9403225806451613, | |
| "eval_loss": 0.025894558057188988, | |
| "eval_runtime": 5.3968, | |
| "eval_samples_per_second": 574.416, | |
| "eval_steps_per_second": 12.044, | |
| "step": 2862 | |
| }, | |
| { | |
| "epoch": 9.968553459119496, | |
| "grad_norm": 0.05505794286727905, | |
| "learning_rate": 8.128511055690263e-09, | |
| "loss": 0.0154, | |
| "step": 3170 | |
| } | |
| ], | |
| "logging_steps": 317, | |
| "max_steps": 3180, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 1000000000.0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 825254092458012.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.7646743801285832, | |
| "learning_rate": 0.00033314171634682974, | |
| "lr_scheduler_type": "cosine", | |
| "num_train_epochs": 10, | |
| "temperature": 5.485237170675724, | |
| "weight_decay": 0.1816702846280333 | |
| } | |
| } | |