| { | |
| "best_metric": 3.9286768436431885, | |
| "best_model_checkpoint": "car_brands_image_detection/checkpoint-21765", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 21765, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 3.6166505813598633, | |
| "learning_rate": 2.937830992401566e-06, | |
| "loss": 4.5799, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 2.645850419998169, | |
| "learning_rate": 2.8687543172921943e-06, | |
| "loss": 4.5716, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 2.978064775466919, | |
| "learning_rate": 2.799677642182823e-06, | |
| "loss": 4.5286, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 2.759573459625244, | |
| "learning_rate": 2.7306009670734517e-06, | |
| "loss": 4.4846, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 2.4581897258758545, | |
| "learning_rate": 2.66152429196408e-06, | |
| "loss": 4.4705, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 2.911447286605835, | |
| "learning_rate": 2.592447616854709e-06, | |
| "loss": 4.4394, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.8367297649383545, | |
| "learning_rate": 2.5233709417453375e-06, | |
| "loss": 4.41, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 2.930182456970215, | |
| "learning_rate": 2.454294266635966e-06, | |
| "loss": 4.385, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.40094773118897187, | |
| "eval_loss": 4.410472869873047, | |
| "eval_runtime": 357.321, | |
| "eval_samples_per_second": 97.447, | |
| "eval_steps_per_second": 12.182, | |
| "step": 4353 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 3.3039791584014893, | |
| "learning_rate": 2.3852175915265945e-06, | |
| "loss": 4.3545, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 3.3890039920806885, | |
| "learning_rate": 2.316140916417223e-06, | |
| "loss": 4.3126, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 4.098603248596191, | |
| "learning_rate": 2.247064241307852e-06, | |
| "loss": 4.2967, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 3.3784146308898926, | |
| "learning_rate": 2.1779875661984803e-06, | |
| "loss": 4.2693, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 3.1738123893737793, | |
| "learning_rate": 2.1089108910891088e-06, | |
| "loss": 4.2466, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 3.7148706912994385, | |
| "learning_rate": 2.0398342159797372e-06, | |
| "loss": 4.2021, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 3.290153741836548, | |
| "learning_rate": 1.970757540870366e-06, | |
| "loss": 4.1814, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 3.142911672592163, | |
| "learning_rate": 1.9016808657609948e-06, | |
| "loss": 4.1626, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 4.143848896026611, | |
| "learning_rate": 1.8326041906516235e-06, | |
| "loss": 4.1376, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.43512349224583574, | |
| "eval_loss": 4.2049126625061035, | |
| "eval_runtime": 371.7808, | |
| "eval_samples_per_second": 93.657, | |
| "eval_steps_per_second": 11.709, | |
| "step": 8706 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "grad_norm": 3.016541004180908, | |
| "learning_rate": 1.763527515542252e-06, | |
| "loss": 4.1126, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 3.141395092010498, | |
| "learning_rate": 1.6944508404328806e-06, | |
| "loss": 4.0881, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 4.218649864196777, | |
| "learning_rate": 1.6253741653235091e-06, | |
| "loss": 4.0854, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 3.368730306625366, | |
| "learning_rate": 1.5562974902141378e-06, | |
| "loss": 4.0432, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 3.461085557937622, | |
| "learning_rate": 1.4872208151047663e-06, | |
| "loss": 4.0324, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 3.5780344009399414, | |
| "learning_rate": 1.418144139995395e-06, | |
| "loss": 4.0083, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 3.727916955947876, | |
| "learning_rate": 1.3490674648860236e-06, | |
| "loss": 4.0215, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 3.84293532371521, | |
| "learning_rate": 1.279990789776652e-06, | |
| "loss": 3.9905, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 4.721088409423828, | |
| "learning_rate": 1.2109141146672808e-06, | |
| "loss": 3.9723, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.4607696726019529, | |
| "eval_loss": 4.052371025085449, | |
| "eval_runtime": 385.9466, | |
| "eval_samples_per_second": 90.22, | |
| "eval_steps_per_second": 11.279, | |
| "step": 13059 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 4.341360092163086, | |
| "learning_rate": 1.1418374395579092e-06, | |
| "loss": 3.9604, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 4.207143306732178, | |
| "learning_rate": 1.0727607644485377e-06, | |
| "loss": 3.9293, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 4.3804850578308105, | |
| "learning_rate": 1.0036840893391666e-06, | |
| "loss": 3.9193, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 3.7782533168792725, | |
| "learning_rate": 9.346074142297951e-07, | |
| "loss": 3.9125, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 3.5943124294281006, | |
| "learning_rate": 8.655307391204237e-07, | |
| "loss": 3.8896, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 3.702357053756714, | |
| "learning_rate": 7.964540640110522e-07, | |
| "loss": 3.9014, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 3.796862840652466, | |
| "learning_rate": 7.273773889016809e-07, | |
| "loss": 3.891, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "grad_norm": 4.275099277496338, | |
| "learning_rate": 6.583007137923095e-07, | |
| "loss": 3.8704, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.4705054566341183, | |
| "eval_loss": 3.9602584838867188, | |
| "eval_runtime": 391.9958, | |
| "eval_samples_per_second": 88.827, | |
| "eval_steps_per_second": 11.105, | |
| "step": 17412 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 3.892676591873169, | |
| "learning_rate": 5.892240386829381e-07, | |
| "loss": 3.8528, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 4.14, | |
| "grad_norm": 3.8556807041168213, | |
| "learning_rate": 5.201473635735666e-07, | |
| "loss": 3.8603, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 4.647615909576416, | |
| "learning_rate": 4.5107068846419527e-07, | |
| "loss": 3.8373, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 4.36, | |
| "grad_norm": 5.085917949676514, | |
| "learning_rate": 3.819940133548239e-07, | |
| "loss": 3.8558, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 3.4918742179870605, | |
| "learning_rate": 3.1291733824545247e-07, | |
| "loss": 3.8305, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "grad_norm": 5.628491401672363, | |
| "learning_rate": 2.4384066313608105e-07, | |
| "loss": 3.8442, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 4.71, | |
| "grad_norm": 6.177685737609863, | |
| "learning_rate": 1.7476398802670965e-07, | |
| "loss": 3.821, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 4.82, | |
| "grad_norm": 3.602433681488037, | |
| "learning_rate": 1.0568731291733824e-07, | |
| "loss": 3.8144, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "grad_norm": 3.561356782913208, | |
| "learning_rate": 3.6610637807966844e-08, | |
| "loss": 3.8221, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.4763928776565192, | |
| "eval_loss": 3.9286768436431885, | |
| "eval_runtime": 356.658, | |
| "eval_samples_per_second": 97.629, | |
| "eval_steps_per_second": 12.205, | |
| "step": 21765 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 21765, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 5.411951551988482e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |