| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 2862, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.9968553459119497, | |
| "grad_norm": 0.6297022104263306, | |
| "learning_rate": 1.7784765897973445e-05, | |
| "loss": 0.3821, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5680645161290323, | |
| "eval_loss": 0.19172954559326172, | |
| "eval_runtime": 1.3575, | |
| "eval_samples_per_second": 2283.627, | |
| "eval_steps_per_second": 47.883, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.9937106918238994, | |
| "grad_norm": 0.5253167152404785, | |
| "learning_rate": 1.556953179594689e-05, | |
| "loss": 0.1603, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8248387096774193, | |
| "eval_loss": 0.09447979182004929, | |
| "eval_runtime": 1.5605, | |
| "eval_samples_per_second": 1986.57, | |
| "eval_steps_per_second": 41.654, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.990566037735849, | |
| "grad_norm": 0.4476560950279236, | |
| "learning_rate": 1.3354297693920338e-05, | |
| "loss": 0.101, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8806451612903226, | |
| "eval_loss": 0.06346584856510162, | |
| "eval_runtime": 1.3904, | |
| "eval_samples_per_second": 2229.649, | |
| "eval_steps_per_second": 46.751, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 3.9874213836477987, | |
| "grad_norm": 0.3931984305381775, | |
| "learning_rate": 1.1139063591893781e-05, | |
| "loss": 0.0764, | |
| "step": 1268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8974193548387097, | |
| "eval_loss": 0.04843955859541893, | |
| "eval_runtime": 1.4033, | |
| "eval_samples_per_second": 2209.103, | |
| "eval_steps_per_second": 46.32, | |
| "step": 1272 | |
| }, | |
| { | |
| "epoch": 4.984276729559748, | |
| "grad_norm": 0.2905353307723999, | |
| "learning_rate": 8.923829489867226e-06, | |
| "loss": 0.0631, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9064516129032258, | |
| "eval_loss": 0.0406271256506443, | |
| "eval_runtime": 1.601, | |
| "eval_samples_per_second": 1936.266, | |
| "eval_steps_per_second": 40.599, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 5.981132075471698, | |
| "grad_norm": 0.3535286486148834, | |
| "learning_rate": 6.708595387840672e-06, | |
| "loss": 0.0553, | |
| "step": 1902 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.9129032258064517, | |
| "eval_loss": 0.0355432964861393, | |
| "eval_runtime": 1.6068, | |
| "eval_samples_per_second": 1929.261, | |
| "eval_steps_per_second": 40.452, | |
| "step": 1908 | |
| }, | |
| { | |
| "epoch": 6.977987421383648, | |
| "grad_norm": 0.23496131598949432, | |
| "learning_rate": 4.4933612858141165e-06, | |
| "loss": 0.0506, | |
| "step": 2219 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.9190322580645162, | |
| "eval_loss": 0.033058661967515945, | |
| "eval_runtime": 1.6082, | |
| "eval_samples_per_second": 1927.576, | |
| "eval_steps_per_second": 40.417, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 7.9748427672955975, | |
| "grad_norm": 0.26389679312705994, | |
| "learning_rate": 2.2781271837875614e-06, | |
| "loss": 0.0478, | |
| "step": 2536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.922258064516129, | |
| "eval_loss": 0.03128081187605858, | |
| "eval_runtime": 1.6078, | |
| "eval_samples_per_second": 1928.074, | |
| "eval_steps_per_second": 40.427, | |
| "step": 2544 | |
| }, | |
| { | |
| "epoch": 8.971698113207546, | |
| "grad_norm": 0.2600725591182709, | |
| "learning_rate": 6.289308176100629e-08, | |
| "loss": 0.0461, | |
| "step": 2853 | |
| } | |
| ], | |
| "logging_steps": 317, | |
| "max_steps": 2862, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "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": 742838622388212.0, | |
| "train_batch_size": 48, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.7443072695022369, | |
| "num_train_epochs": 9, | |
| "temperature": 18 | |
| } | |
| } | |