trainer_state.json

{
  "best_metric": 0.38853333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0674/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": 13.38124942779541,
      "learning_rate": 0.0004294294294294294,
      "loss": 3.6173,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.1648,
      "eval_loss": 3.2028000354766846,
      "eval_runtime": 11.931,
      "eval_samples_per_second": 314.307,
      "eval_steps_per_second": 4.945,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.876428604125977,
      "learning_rate": 0.000358000858000858,
      "loss": 3.0389,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2344,
      "eval_loss": 2.9296176433563232,
      "eval_runtime": 12.0127,
      "eval_samples_per_second": 312.169,
      "eval_steps_per_second": 4.911,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 12.877269744873047,
      "learning_rate": 0.0002865722865722866,
      "loss": 2.7901,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.2786666666666667,
      "eval_loss": 2.7114906311035156,
      "eval_runtime": 11.8645,
      "eval_samples_per_second": 316.069,
      "eval_steps_per_second": 4.973,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 18.108457565307617,
      "learning_rate": 0.00021514371514371513,
      "loss": 2.581,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.31066666666666665,
      "eval_loss": 2.560798168182373,
      "eval_runtime": 12.0875,
      "eval_samples_per_second": 310.237,
      "eval_steps_per_second": 4.881,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 16.452150344848633,
      "learning_rate": 0.00014371514371514372,
      "loss": 2.3639,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3488,
      "eval_loss": 2.42657732963562,
      "eval_runtime": 13.6381,
      "eval_samples_per_second": 274.964,
      "eval_steps_per_second": 4.326,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 21.635099411010742,
      "learning_rate": 7.228657228657229e-05,
      "loss": 2.1167,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.3757333333333333,
      "eval_loss": 2.315603733062744,
      "eval_runtime": 12.0885,
      "eval_samples_per_second": 310.213,
      "eval_steps_per_second": 4.881,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 26.68172264099121,
      "learning_rate": 8.580008580008581e-07,
      "loss": 1.8185,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.38853333333333334,
      "eval_loss": 2.333320379257202,
      "eval_runtime": 12.2341,
      "eval_samples_per_second": 306.521,
      "eval_steps_per_second": 4.823,
      "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
}