| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "global_step": 61360, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 2.877770534550196e-05, | |
| "loss": 0.5826, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 2.7555410691003913e-05, | |
| "loss": 0.4701, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.6333116036505867e-05, | |
| "loss": 0.4449, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 2.511082138200782e-05, | |
| "loss": 0.4274, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8523688232297504, | |
| "eval_loss": 0.38915878534317017, | |
| "eval_runtime": 26.5235, | |
| "eval_samples_per_second": 370.049, | |
| "eval_steps_per_second": 11.575, | |
| "step": 12272 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 2.388852672750978e-05, | |
| "loss": 0.4051, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 2.2666232073011734e-05, | |
| "loss": 0.2711, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "learning_rate": 2.144393741851369e-05, | |
| "loss": 0.2755, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 2.0221642764015646e-05, | |
| "loss": 0.2808, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 1.89993481095176e-05, | |
| "loss": 0.2844, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8565461029037188, | |
| "eval_loss": 0.4079381227493286, | |
| "eval_runtime": 26.505, | |
| "eval_samples_per_second": 370.308, | |
| "eval_steps_per_second": 11.583, | |
| "step": 24544 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 1.7777053455019558e-05, | |
| "loss": 0.2537, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 1.6554758800521512e-05, | |
| "loss": 0.1468, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 1.533246414602347e-05, | |
| "loss": 0.1536, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 1.4110169491525424e-05, | |
| "loss": 0.1574, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 1.288787483702738e-05, | |
| "loss": 0.1589, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8526744778400408, | |
| "eval_loss": 0.503338098526001, | |
| "eval_runtime": 26.5083, | |
| "eval_samples_per_second": 370.261, | |
| "eval_steps_per_second": 11.581, | |
| "step": 36816 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 1.1665580182529336e-05, | |
| "loss": 0.14, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 3.26, | |
| "learning_rate": 1.044328552803129e-05, | |
| "loss": 0.0868, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "learning_rate": 9.220990873533247e-06, | |
| "loss": 0.0901, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "learning_rate": 7.998696219035203e-06, | |
| "loss": 0.0904, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "learning_rate": 6.776401564537158e-06, | |
| "loss": 0.0877, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8575649516046867, | |
| "eval_loss": 0.6623563170433044, | |
| "eval_runtime": 26.5163, | |
| "eval_samples_per_second": 370.149, | |
| "eval_steps_per_second": 11.578, | |
| "step": 49088 | |
| }, | |
| { | |
| "epoch": 4.07, | |
| "learning_rate": 5.554106910039114e-06, | |
| "loss": 0.0698, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.28, | |
| "learning_rate": 4.331812255541069e-06, | |
| "loss": 0.0468, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "learning_rate": 3.109517601043025e-06, | |
| "loss": 0.0494, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "learning_rate": 1.8872229465449805e-06, | |
| "loss": 0.045, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "learning_rate": 6.649282920469362e-07, | |
| "loss": 0.0426, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8633723892002038, | |
| "eval_loss": 0.8447091579437256, | |
| "eval_runtime": 26.5083, | |
| "eval_samples_per_second": 370.261, | |
| "eval_steps_per_second": 11.581, | |
| "step": 61360 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 61360, | |
| "total_flos": 1.4676274702232822e+17, | |
| "train_loss": 0.2071122335486182, | |
| "train_runtime": 11012.4801, | |
| "train_samples_per_second": 178.299, | |
| "train_steps_per_second": 5.572 | |
| } | |
| ], | |
| "max_steps": 61360, | |
| "num_train_epochs": 5, | |
| "total_flos": 1.4676274702232822e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |