| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "global_step": 36816, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 2.9569163602941176e-05, | |
| "loss": 0.6153, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 2.741498161764706e-05, | |
| "loss": 0.4236, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.5260799632352943e-05, | |
| "loss": 0.391, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 2.3106617647058826e-05, | |
| "loss": 0.3762, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8793683138053999, | |
| "eval_loss": 0.331181138753891, | |
| "eval_runtime": 14.8457, | |
| "eval_samples_per_second": 661.136, | |
| "eval_steps_per_second": 20.679, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 2.0952435661764706e-05, | |
| "loss": 0.3551, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 1.879825367647059e-05, | |
| "loss": 0.2603, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 1.6644071691176472e-05, | |
| "loss": 0.2538, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 1.4489889705882352e-05, | |
| "loss": 0.2529, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 1.2335707720588235e-05, | |
| "loss": 0.2542, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8842587875700458, | |
| "eval_loss": 0.3466881811618805, | |
| "eval_runtime": 14.8408, | |
| "eval_samples_per_second": 661.35, | |
| "eval_steps_per_second": 20.686, | |
| "step": 24544 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 1.0181525735294117e-05, | |
| "loss": 0.2337, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 8.02734375e-06, | |
| "loss": 0.1582, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 5.873161764705882e-06, | |
| "loss": 0.1561, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 3.7189797794117646e-06, | |
| "loss": 0.1564, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 1.5647977941176472e-06, | |
| "loss": 0.1503, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8879266428935303, | |
| "eval_loss": 0.42651689052581787, | |
| "eval_runtime": 14.833, | |
| "eval_samples_per_second": 661.699, | |
| "eval_steps_per_second": 20.697, | |
| "step": 36816 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 36816, | |
| "total_flos": 2.486132475548778e+16, | |
| "train_loss": 0.28156463129009596, | |
| "train_runtime": 4491.1198, | |
| "train_samples_per_second": 262.319, | |
| "train_steps_per_second": 8.198 | |
| } | |
| ], | |
| "max_steps": 36816, | |
| "num_train_epochs": 3, | |
| "total_flos": 2.486132475548778e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |