trainer_state.json

{
  "best_metric": 0.3194666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0069/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": 0.0005,
      "loss": 3.5486,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.1616,
      "eval_loss": 3.1767990589141846,
      "eval_runtime": 6.7343,
      "eval_samples_per_second": 556.849,
      "eval_steps_per_second": 8.761,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 17.4770565032959,
      "learning_rate": 0.0005,
      "loss": 3.1051,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.1992,
      "eval_loss": 3.036022901535034,
      "eval_runtime": 7.0124,
      "eval_samples_per_second": 534.768,
      "eval_steps_per_second": 8.414,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 21.679386138916016,
      "learning_rate": 0.0005,
      "loss": 2.959,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.22453333333333333,
      "eval_loss": 2.915682554244995,
      "eval_runtime": 6.6903,
      "eval_samples_per_second": 560.513,
      "eval_steps_per_second": 8.819,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 17.911623001098633,
      "learning_rate": 0.0005,
      "loss": 2.8155,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.2344,
      "eval_loss": 2.9144773483276367,
      "eval_runtime": 6.8859,
      "eval_samples_per_second": 544.592,
      "eval_steps_per_second": 8.568,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 17.08316993713379,
      "learning_rate": 0.0005,
      "loss": 2.7041,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.27013333333333334,
      "eval_loss": 2.6821722984313965,
      "eval_runtime": 6.7976,
      "eval_samples_per_second": 551.662,
      "eval_steps_per_second": 8.679,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 15.424040794372559,
      "learning_rate": 0.0005,
      "loss": 2.5979,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.2888,
      "eval_loss": 2.638681411743164,
      "eval_runtime": 6.8655,
      "eval_samples_per_second": 546.208,
      "eval_steps_per_second": 8.594,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 13.946167945861816,
      "learning_rate": 0.0005,
      "loss": 2.526,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.3194666666666667,
      "eval_loss": 2.5414249897003174,
      "eval_runtime": 7.6305,
      "eval_samples_per_second": 491.447,
      "eval_steps_per_second": 7.732,
      "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
}