trainer_state.json

{
  "best_metric": 0.9056,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0444/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 22.917757034301758,
      "learning_rate": 6.653391037658466e-05,
      "loss": 1.677,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8274666666666667,
      "eval_loss": 0.6589551568031311,
      "eval_runtime": 6.2534,
      "eval_samples_per_second": 599.672,
      "eval_steps_per_second": 9.435,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 14.160438537597656,
      "learning_rate": 5.685900305529986e-05,
      "loss": 0.4805,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8528,
      "eval_loss": 0.4895952343940735,
      "eval_runtime": 6.1645,
      "eval_samples_per_second": 608.321,
      "eval_steps_per_second": 9.571,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.8385580778121948,
      "learning_rate": 4.28342139851007e-05,
      "loss": 0.2661,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8866666666666667,
      "eval_loss": 0.3979380130767822,
      "eval_runtime": 6.5472,
      "eval_samples_per_second": 572.765,
      "eval_steps_per_second": 9.012,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 17.137426376342773,
      "learning_rate": 2.7257762754823102e-05,
      "loss": 0.1474,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8922666666666667,
      "eval_loss": 0.3674142062664032,
      "eval_runtime": 6.4105,
      "eval_samples_per_second": 584.98,
      "eval_steps_per_second": 9.204,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.5376176834106445,
      "learning_rate": 1.3214756563261342e-05,
      "loss": 0.0722,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9037333333333334,
      "eval_loss": 0.33976253867149353,
      "eval_runtime": 6.0173,
      "eval_samples_per_second": 623.2,
      "eval_steps_per_second": 9.805,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.15913553535938263,
      "learning_rate": 3.4865850128228307e-06,
      "loss": 0.0334,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9037333333333334,
      "eval_loss": 0.34078437089920044,
      "eval_runtime": 6.4199,
      "eval_samples_per_second": 584.117,
      "eval_steps_per_second": 9.19,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.4457952380180359,
      "learning_rate": 3.178726637742546e-11,
      "loss": 0.0199,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9056,
      "eval_loss": 0.3380773961544037,
      "eval_runtime": 6.0093,
      "eval_samples_per_second": 624.028,
      "eval_steps_per_second": 9.818,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}