trainer_state.json

{
  "best_metric": 0.9210666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0958/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": 17.784086227416992,
      "learning_rate": 6.735155628520579e-05,
      "loss": 0.7544,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.856,
      "eval_loss": 0.4751030504703522,
      "eval_runtime": 11.6007,
      "eval_samples_per_second": 323.255,
      "eval_steps_per_second": 5.086,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.36588549613952637,
      "learning_rate": 5.9777905730703695e-05,
      "loss": 0.2372,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.46749603748321533,
      "eval_runtime": 12.5101,
      "eval_samples_per_second": 299.757,
      "eval_steps_per_second": 4.716,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 24.4576473236084,
      "learning_rate": 4.843204364097674e-05,
      "loss": 0.1181,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8816,
      "eval_loss": 0.4303298890590668,
      "eval_runtime": 12.5031,
      "eval_samples_per_second": 299.926,
      "eval_steps_per_second": 4.719,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.027621794492006302,
      "learning_rate": 3.5041274668689884e-05,
      "loss": 0.0574,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8882666666666666,
      "eval_loss": 0.4303155541419983,
      "eval_runtime": 12.994,
      "eval_samples_per_second": 288.594,
      "eval_steps_per_second": 4.541,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0113552650436759,
      "learning_rate": 2.16442220022508e-05,
      "loss": 0.0249,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9021333333333333,
      "eval_loss": 0.42478588223457336,
      "eval_runtime": 12.4306,
      "eval_samples_per_second": 301.676,
      "eval_steps_per_second": 4.746,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0007456074235960841,
      "learning_rate": 1.0280465465539994e-05,
      "loss": 0.0058,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9128,
      "eval_loss": 0.36055806279182434,
      "eval_runtime": 12.5268,
      "eval_samples_per_second": 299.359,
      "eval_steps_per_second": 4.71,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.1283661425113678,
      "learning_rate": 2.6800339785621635e-06,
      "loss": 0.001,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9210666666666667,
      "eval_loss": 0.34702742099761963,
      "eval_runtime": 12.5085,
      "eval_samples_per_second": 299.797,
      "eval_steps_per_second": 4.717,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}