trainer_state.json

{
  "best_metric": 0.9152,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0008/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": 6.7914265912926e-05,
      "loss": 0.8328,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8434666666666667,
      "eval_loss": 0.501545786857605,
      "eval_runtime": 7.6148,
      "eval_samples_per_second": 492.461,
      "eval_steps_per_second": 7.748,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.7411800622940063,
      "learning_rate": 6.185866251124284e-05,
      "loss": 0.2945,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8602666666666666,
      "eval_loss": 0.48235276341438293,
      "eval_runtime": 6.4604,
      "eval_samples_per_second": 580.457,
      "eval_steps_per_second": 9.133,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.11396384984254837,
      "learning_rate": 5.2563508019062023e-05,
      "loss": 0.1582,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8658666666666667,
      "eval_loss": 0.48657160997390747,
      "eval_runtime": 6.2702,
      "eval_samples_per_second": 598.069,
      "eval_steps_per_second": 9.41,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 48.15888214111328,
      "learning_rate": 4.1149935250010576e-05,
      "loss": 0.0887,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8917333333333334,
      "eval_loss": 0.43846946954727173,
      "eval_runtime": 5.9416,
      "eval_samples_per_second": 631.146,
      "eval_steps_per_second": 9.93,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.01761585846543312,
      "learning_rate": 2.8994589526430138e-05,
      "loss": 0.0452,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9008,
      "eval_loss": 0.4194916784763336,
      "eval_runtime": 5.8547,
      "eval_samples_per_second": 640.51,
      "eval_steps_per_second": 10.077,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.47807466983795166,
      "learning_rate": 1.756358493638142e-05,
      "loss": 0.0168,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9064,
      "eval_loss": 0.41272637248039246,
      "eval_runtime": 7.7129,
      "eval_samples_per_second": 486.197,
      "eval_steps_per_second": 7.649,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.1321132928133011,
      "learning_rate": 8.235669337084578e-06,
      "loss": 0.005,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9133333333333333,
      "eval_loss": 0.377856582403183,
      "eval_runtime": 6.9301,
      "eval_samples_per_second": 541.116,
      "eval_steps_per_second": 8.514,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.03976298123598099,
      "learning_rate": 2.135927015187301e-06,
      "loss": 0.002,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9152,
      "eval_loss": 0.36619389057159424,
      "eval_runtime": 6.6194,
      "eval_samples_per_second": 566.516,
      "eval_steps_per_second": 8.913,
      "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
}