trainer_state.json

{
  "best_metric": 0.8632,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0500/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 50.2782096862793,
      "learning_rate": 7e-05,
      "loss": 0.9314,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8418666666666667,
      "eval_loss": 0.5125667452812195,
      "eval_runtime": 11.0704,
      "eval_samples_per_second": 338.742,
      "eval_steps_per_second": 5.33,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 23.733840942382812,
      "learning_rate": 7e-05,
      "loss": 0.3025,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8626666666666667,
      "eval_loss": 0.46237635612487793,
      "eval_runtime": 11.0786,
      "eval_samples_per_second": 338.489,
      "eval_steps_per_second": 5.326,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.6017625331878662,
      "learning_rate": 7e-05,
      "loss": 0.1634,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.48798274993896484,
      "eval_runtime": 11.5644,
      "eval_samples_per_second": 324.27,
      "eval_steps_per_second": 5.102,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 7e-05,
      "loss": 0.1302,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8592,
      "eval_loss": 0.5303805470466614,
      "eval_runtime": 11.0332,
      "eval_samples_per_second": 339.883,
      "eval_steps_per_second": 5.347,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.01777472160756588,
      "learning_rate": 7e-05,
      "loss": 0.0998,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8530666666666666,
      "eval_loss": 0.60077303647995,
      "eval_runtime": 10.9949,
      "eval_samples_per_second": 341.067,
      "eval_steps_per_second": 5.366,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.2100156992673874,
      "learning_rate": 7e-05,
      "loss": 0.0961,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8632,
      "eval_loss": 0.570968747138977,
      "eval_runtime": 11.1744,
      "eval_samples_per_second": 335.588,
      "eval_steps_per_second": 5.28,
      "step": 1998
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}