trainer_state.json

{
  "best_metric": 0.4848,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0504/checkpoints/checkpoint-2997",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 2997,
  "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": 0.0002911677661332091,
      "loss": 3.2776,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.21626666666666666,
      "eval_loss": 2.893810749053955,
      "eval_runtime": 8.4056,
      "eval_samples_per_second": 446.132,
      "eval_steps_per_second": 7.019,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 18.97915267944336,
      "learning_rate": 0.0002653099348375843,
      "loss": 2.7181,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2813333333333333,
      "eval_loss": 2.6460208892822266,
      "eval_runtime": 8.0341,
      "eval_samples_per_second": 466.761,
      "eval_steps_per_second": 7.344,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 25.910329818725586,
      "learning_rate": 0.00022554402360715464,
      "loss": 2.4644,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.34746666666666665,
      "eval_loss": 2.3927314281463623,
      "eval_runtime": 7.7607,
      "eval_samples_per_second": 483.204,
      "eval_steps_per_second": 7.602,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 31.87564468383789,
      "learning_rate": 0.00017666638821865508,
      "loss": 2.2336,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3685333333333333,
      "eval_loss": 2.2496750354766846,
      "eval_runtime": 7.4342,
      "eval_samples_per_second": 504.429,
      "eval_steps_per_second": 7.936,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 26.804018020629883,
      "learning_rate": 0.00012457239285701023,
      "loss": 2.0148,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4048,
      "eval_loss": 2.1030690670013428,
      "eval_runtime": 7.4431,
      "eval_samples_per_second": 503.826,
      "eval_steps_per_second": 7.927,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 25.961620330810547,
      "learning_rate": 7.554534218832392e-05,
      "loss": 1.7793,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4232,
      "eval_loss": 2.0560646057128906,
      "eval_runtime": 9.4591,
      "eval_samples_per_second": 396.442,
      "eval_steps_per_second": 6.237,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 16.118000030517578,
      "learning_rate": 3.5498622085445925e-05,
      "loss": 1.5128,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4728,
      "eval_loss": 1.9165841341018677,
      "eval_runtime": 7.5908,
      "eval_samples_per_second": 494.022,
      "eval_steps_per_second": 7.773,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": Infinity,
      "learning_rate": 9.31693256552663e-06,
      "loss": 1.2723,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.484,
      "eval_loss": 1.9250226020812988,
      "eval_runtime": 7.6048,
      "eval_samples_per_second": 493.109,
      "eval_steps_per_second": 7.758,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 94.65938568115234,
      "learning_rate": 2.060281379051565e-09,
      "loss": 1.1142,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.4848,
      "eval_loss": 2.0009613037109375,
      "eval_runtime": 7.5901,
      "eval_samples_per_second": 494.067,
      "eval_steps_per_second": 7.773,
      "step": 2997
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.482671885603328e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}