trainer_state.json

{
  "best_metric": 0.9109333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_1000/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": 4.7538820348083834e-05,
      "loss": 1.7331,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.828,
      "eval_loss": 0.7055222392082214,
      "eval_runtime": 8.1774,
      "eval_samples_per_second": 458.582,
      "eval_steps_per_second": 7.215,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.2515413761138916,
      "learning_rate": 4.0613573610928476e-05,
      "loss": 0.4856,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8693333333333333,
      "eval_loss": 0.471589058637619,
      "eval_runtime": 6.9075,
      "eval_samples_per_second": 542.891,
      "eval_steps_per_second": 8.541,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.565695285797119,
      "learning_rate": 3.059586713221479e-05,
      "loss": 0.2601,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8877333333333334,
      "eval_loss": 0.4003779888153076,
      "eval_runtime": 6.9091,
      "eval_samples_per_second": 542.765,
      "eval_steps_per_second": 8.54,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 7.286423206329346,
      "learning_rate": 1.946983053915936e-05,
      "loss": 0.1405,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8944,
      "eval_loss": 0.3739269971847534,
      "eval_runtime": 7.1103,
      "eval_samples_per_second": 527.402,
      "eval_steps_per_second": 8.298,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.6013352870941162,
      "learning_rate": 9.43911183090096e-06,
      "loss": 0.0667,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9050666666666667,
      "eval_loss": 0.3527183532714844,
      "eval_runtime": 6.9408,
      "eval_samples_per_second": 540.282,
      "eval_steps_per_second": 8.5,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.3147793114185333,
      "learning_rate": 2.4904178663020224e-06,
      "loss": 0.0346,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9090666666666667,
      "eval_loss": 0.34101757407188416,
      "eval_runtime": 7.1571,
      "eval_samples_per_second": 523.953,
      "eval_steps_per_second": 8.244,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.556058406829834,
      "learning_rate": 2.2705190269589617e-11,
      "loss": 0.0225,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9109333333333334,
      "eval_loss": 0.32630035281181335,
      "eval_runtime": 6.8918,
      "eval_samples_per_second": 544.123,
      "eval_steps_per_second": 8.561,
      "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
}