| { | |
| "best_metric": 0.22801454365253448, | |
| "best_model_checkpoint": "/workspace/training_output/perspectives/checkpoint-12168", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 15210, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 1.9342537804076265e-05, | |
| "loss": 1.492, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 1.8685075608152533e-05, | |
| "loss": 0.9689, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 1.80276134122288e-05, | |
| "loss": 0.795, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.7370151216305063e-05, | |
| "loss": 0.6899, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 1.671268902038133e-05, | |
| "loss": 0.6365, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 1.6055226824457594e-05, | |
| "loss": 0.5677, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.883791336748952, | |
| "eval_loss": 0.39963653683662415, | |
| "eval_runtime": 331.6914, | |
| "eval_samples_per_second": 12.946, | |
| "eval_steps_per_second": 1.619, | |
| "step": 3042 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 1.539776462853386e-05, | |
| "loss": 0.53, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 1.4740302432610125e-05, | |
| "loss": 0.4905, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 1.4082840236686392e-05, | |
| "loss": 0.4689, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 1.3425378040762658e-05, | |
| "loss": 0.4579, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 1.2767915844838923e-05, | |
| "loss": 0.4237, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 1.2110453648915189e-05, | |
| "loss": 0.4259, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8747088961341407, | |
| "eval_loss": 0.3984226882457733, | |
| "eval_runtime": 367.0798, | |
| "eval_samples_per_second": 11.698, | |
| "eval_steps_per_second": 1.463, | |
| "step": 6084 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 1.1452991452991454e-05, | |
| "loss": 0.3722, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 1.0795529257067721e-05, | |
| "loss": 0.3872, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "learning_rate": 1.0138067061143987e-05, | |
| "loss": 0.3768, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "learning_rate": 9.48060486522025e-06, | |
| "loss": 0.3478, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 8.823142669296516e-06, | |
| "loss": 0.3586, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 8.165680473372781e-06, | |
| "loss": 0.3448, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.9189566837447601, | |
| "eval_loss": 0.2590956389904022, | |
| "eval_runtime": 345.333, | |
| "eval_samples_per_second": 12.434, | |
| "eval_steps_per_second": 1.555, | |
| "step": 9126 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 7.508218277449047e-06, | |
| "loss": 0.3242, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "learning_rate": 6.850756081525313e-06, | |
| "loss": 0.3168, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "learning_rate": 6.193293885601579e-06, | |
| "loss": 0.3266, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "learning_rate": 5.535831689677844e-06, | |
| "loss": 0.3492, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "learning_rate": 4.878369493754109e-06, | |
| "loss": 0.3307, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "learning_rate": 4.220907297830375e-06, | |
| "loss": 0.3094, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.9268747088961341, | |
| "eval_loss": 0.22801454365253448, | |
| "eval_runtime": 333.8049, | |
| "eval_samples_per_second": 12.864, | |
| "eval_steps_per_second": 1.609, | |
| "step": 12168 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "learning_rate": 3.563445101906641e-06, | |
| "loss": 0.2925, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "learning_rate": 2.9059829059829063e-06, | |
| "loss": 0.2796, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "learning_rate": 2.2485207100591717e-06, | |
| "loss": 0.2927, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "learning_rate": 1.5910585141354374e-06, | |
| "loss": 0.29, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 4.77, | |
| "learning_rate": 9.335963182117029e-07, | |
| "loss": 0.3004, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "learning_rate": 2.7613412228796843e-07, | |
| "loss": 0.2449, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.9229156963204471, | |
| "eval_loss": 0.2583230435848236, | |
| "eval_runtime": 324.7812, | |
| "eval_samples_per_second": 13.221, | |
| "eval_steps_per_second": 1.653, | |
| "step": 15210 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 15210, | |
| "total_flos": 9.427914420548198e+18, | |
| "train_loss": 0.4568727389203334, | |
| "train_runtime": 10636.8544, | |
| "train_samples_per_second": 11.437, | |
| "train_steps_per_second": 1.43 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 15210, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 9.427914420548198e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |