trainer_state.json

{
  "best_metric": 0.9005333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0607/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": 4.634542942047119,
      "learning_rate": 0.00047524221697560476,
      "loss": 0.7911,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7970666666666667,
      "eval_loss": 0.6831286549568176,
      "eval_runtime": 12.5625,
      "eval_samples_per_second": 298.509,
      "eval_steps_per_second": 4.697,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.1631980538368225,
      "learning_rate": 0.0004058724504646834,
      "loss": 0.4293,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8122666666666667,
      "eval_loss": 0.6867760419845581,
      "eval_runtime": 12.3341,
      "eval_samples_per_second": 304.035,
      "eval_steps_per_second": 4.783,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.0292050838470459,
      "learning_rate": 0.0003056302334890786,
      "loss": 0.253,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8352,
      "eval_loss": 0.5841264128684998,
      "eval_runtime": 12.4678,
      "eval_samples_per_second": 300.774,
      "eval_steps_per_second": 4.732,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.005561563652008772,
      "learning_rate": 0.00019436976651092142,
      "loss": 0.1266,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8570666666666666,
      "eval_loss": 0.5676218867301941,
      "eval_runtime": 12.9914,
      "eval_samples_per_second": 288.653,
      "eval_steps_per_second": 4.541,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.014790846966207027,
      "learning_rate": 9.412754953531663e-05,
      "loss": 0.0414,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8877333333333334,
      "eval_loss": 0.4373477101325989,
      "eval_runtime": 12.3028,
      "eval_samples_per_second": 304.809,
      "eval_steps_per_second": 4.796,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.010635070502758026,
      "learning_rate": 2.4757783024395242e-05,
      "loss": 0.0076,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8994666666666666,
      "eval_loss": 0.4087597727775574,
      "eval_runtime": 13.2044,
      "eval_samples_per_second": 283.997,
      "eval_steps_per_second": 4.468,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 2.2705190269589615e-10,
      "loss": 0.0029,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9005333333333333,
      "eval_loss": 0.4134864807128906,
      "eval_runtime": 12.1636,
      "eval_samples_per_second": 308.297,
      "eval_steps_per_second": 4.851,
      "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
}