trainer_state.json

{
  "best_metric": 0.41546666666666665,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0667/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "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": 0.0004451117784451118,
      "loss": 3.4367,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.16693333333333332,
      "eval_loss": 3.1042072772979736,
      "eval_runtime": 10.2125,
      "eval_samples_per_second": 367.196,
      "eval_steps_per_second": 5.777,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 36.273136138916016,
      "learning_rate": 0.0003895562228895562,
      "loss": 2.9497,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.24853333333333333,
      "eval_loss": 2.8285257816314697,
      "eval_runtime": 10.2111,
      "eval_samples_per_second": 367.249,
      "eval_steps_per_second": 5.778,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 25.742374420166016,
      "learning_rate": 0.0003340006673340007,
      "loss": 2.7129,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.28,
      "eval_loss": 2.6997876167297363,
      "eval_runtime": 11.1082,
      "eval_samples_per_second": 337.589,
      "eval_steps_per_second": 5.311,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 25.348499298095703,
      "learning_rate": 0.0002784451117784451,
      "loss": 2.509,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.31733333333333336,
      "eval_loss": 2.499070644378662,
      "eval_runtime": 10.1845,
      "eval_samples_per_second": 368.207,
      "eval_steps_per_second": 5.793,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 30.843135833740234,
      "learning_rate": 0.00022288955622288957,
      "loss": 2.3278,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3602666666666667,
      "eval_loss": 2.3540921211242676,
      "eval_runtime": 10.2297,
      "eval_samples_per_second": 366.581,
      "eval_steps_per_second": 5.768,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 25.438068389892578,
      "learning_rate": 0.00016733400066733402,
      "loss": 2.1215,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.3768,
      "eval_loss": 2.2989134788513184,
      "eval_runtime": 10.1233,
      "eval_samples_per_second": 370.432,
      "eval_steps_per_second": 5.828,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 30.831300735473633,
      "learning_rate": 0.00011177844511177845,
      "loss": 1.8742,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4048,
      "eval_loss": 2.2484190464019775,
      "eval_runtime": 10.0973,
      "eval_samples_per_second": 371.387,
      "eval_steps_per_second": 5.843,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": Infinity,
      "learning_rate": 5.6389723056389724e-05,
      "loss": 1.5591,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.41546666666666665,
      "eval_loss": 2.2613067626953125,
      "eval_runtime": 10.501,
      "eval_samples_per_second": 357.107,
      "eval_steps_per_second": 5.618,
      "step": 2664
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}