trainer_state.json

{
  "best_metric": 0.8973333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0928/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": Infinity,
      "learning_rate": 8.55698766265509e-05,
      "loss": 2.3687,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7493333333333333,
      "eval_loss": 0.9358195066452026,
      "eval_runtime": 5.4694,
      "eval_samples_per_second": 685.632,
      "eval_steps_per_second": 10.787,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9080297350883484,
      "learning_rate": 7.310443249967126e-05,
      "loss": 0.6448,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8232,
      "eval_loss": 0.5982705950737,
      "eval_runtime": 5.9549,
      "eval_samples_per_second": 629.733,
      "eval_steps_per_second": 9.908,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 5.5131661351986275e-05,
      "loss": 0.322,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8533333333333334,
      "eval_loss": 0.47934049367904663,
      "eval_runtime": 6.2188,
      "eval_samples_per_second": 603.007,
      "eval_steps_per_second": 9.487,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.3052771985530853,
      "learning_rate": 3.5104850050159014e-05,
      "loss": 0.1581,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8746666666666667,
      "eval_loss": 0.436244934797287,
      "eval_runtime": 6.0194,
      "eval_samples_per_second": 622.986,
      "eval_steps_per_second": 9.802,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.39884477853775024,
      "learning_rate": 1.703789455194791e-05,
      "loss": 0.0665,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8869333333333334,
      "eval_loss": 0.4047512412071228,
      "eval_runtime": 6.2093,
      "eval_samples_per_second": 603.929,
      "eval_steps_per_second": 9.502,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.062203384935855865,
      "learning_rate": 4.509176970441849e-06,
      "loss": 0.0264,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8914666666666666,
      "eval_loss": 0.3869095742702484,
      "eval_runtime": 5.8647,
      "eval_samples_per_second": 639.417,
      "eval_steps_per_second": 10.06,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.10593663156032562,
      "learning_rate": 1.6347729569543557e-10,
      "loss": 0.0148,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8973333333333333,
      "eval_loss": 0.3792930245399475,
      "eval_runtime": 5.9852,
      "eval_samples_per_second": 626.549,
      "eval_steps_per_second": 9.858,
      "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
}