trainer_state.json

{
  "best_metric": 0.8909333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0937/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": 73.03301239013672,
      "learning_rate": 5e-05,
      "loss": 0.7962,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8536,
      "eval_loss": 0.46861961483955383,
      "eval_runtime": 8.1188,
      "eval_samples_per_second": 461.888,
      "eval_steps_per_second": 7.267,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.0931453704833984,
      "learning_rate": 5e-05,
      "loss": 0.2589,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8597333333333333,
      "eval_loss": 0.44259724020957947,
      "eval_runtime": 6.8116,
      "eval_samples_per_second": 550.534,
      "eval_steps_per_second": 8.662,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 5e-05,
      "loss": 0.1724,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8864,
      "eval_loss": 0.3783620297908783,
      "eval_runtime": 6.3369,
      "eval_samples_per_second": 591.773,
      "eval_steps_per_second": 9.311,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.480880469083786,
      "learning_rate": 5e-05,
      "loss": 0.1195,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8808,
      "eval_loss": 0.437979519367218,
      "eval_runtime": 6.4316,
      "eval_samples_per_second": 583.063,
      "eval_steps_per_second": 9.174,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0006578485481441021,
      "learning_rate": 5e-05,
      "loss": 0.0875,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8805333333333333,
      "eval_loss": 0.4621821343898773,
      "eval_runtime": 6.4651,
      "eval_samples_per_second": 580.041,
      "eval_steps_per_second": 9.126,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.9557762742042542,
      "learning_rate": 5e-05,
      "loss": 0.0725,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8909333333333334,
      "eval_loss": 0.4704670011997223,
      "eval_runtime": 6.3532,
      "eval_samples_per_second": 590.255,
      "eval_steps_per_second": 9.287,
      "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
}