trainer_state.json

{
  "best_metric": 0.924,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0526/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": 3.0430328845977783,
      "learning_rate": 9.621650897886543e-05,
      "loss": 0.867,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8485333333333334,
      "eval_loss": 0.5144333243370056,
      "eval_runtime": 5.4095,
      "eval_samples_per_second": 693.221,
      "eval_steps_per_second": 10.907,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.26578715443611145,
      "learning_rate": 8.543862808774192e-05,
      "loss": 0.2893,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8664,
      "eval_loss": 0.48409733176231384,
      "eval_runtime": 5.4462,
      "eval_samples_per_second": 688.549,
      "eval_steps_per_second": 10.833,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.0873289704322815,
      "learning_rate": 6.924306924189557e-05,
      "loss": 0.1505,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8698666666666667,
      "eval_loss": 0.5146114826202393,
      "eval_runtime": 5.5925,
      "eval_samples_per_second": 670.546,
      "eval_steps_per_second": 10.55,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.1440804898738861,
      "learning_rate": 5.011792754282769e-05,
      "loss": 0.0701,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8864,
      "eval_loss": 0.4485776722431183,
      "eval_runtime": 5.3675,
      "eval_samples_per_second": 698.643,
      "eval_steps_per_second": 10.992,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.038634125143289566,
      "learning_rate": 3.0974832444380136e-05,
      "loss": 0.0235,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9088,
      "eval_loss": 0.39442166686058044,
      "eval_runtime": 5.4057,
      "eval_samples_per_second": 693.714,
      "eval_steps_per_second": 10.914,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.00032873344025574625,
      "learning_rate": 1.472814664270234e-05,
      "loss": 0.0046,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9157333333333333,
      "eval_loss": 0.38658085465431213,
      "eval_runtime": 5.3545,
      "eval_samples_per_second": 700.344,
      "eval_steps_per_second": 11.019,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.020434336736798286,
      "learning_rate": 3.851280774526222e-06,
      "loss": 0.0003,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.924,
      "eval_loss": 0.3673185408115387,
      "eval_runtime": 5.3384,
      "eval_samples_per_second": 702.464,
      "eval_steps_per_second": 11.052,
      "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
}