trainer_state.json

{
  "best_metric": 0.33973333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0328/checkpoints/checkpoint-2997",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 2997,
  "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": 0.0005,
      "loss": 3.4697,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.15786666666666666,
      "eval_loss": 3.186244010925293,
      "eval_runtime": 5.4837,
      "eval_samples_per_second": 683.849,
      "eval_steps_per_second": 10.759,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 22.184337615966797,
      "learning_rate": 0.0005,
      "loss": 2.9685,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.22293333333333334,
      "eval_loss": 2.902283191680908,
      "eval_runtime": 5.5356,
      "eval_samples_per_second": 677.437,
      "eval_steps_per_second": 10.658,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.518645763397217,
      "learning_rate": 0.0005,
      "loss": 2.811,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.24053333333333332,
      "eval_loss": 2.8519999980926514,
      "eval_runtime": 6.2289,
      "eval_samples_per_second": 602.033,
      "eval_steps_per_second": 9.472,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 16.709171295166016,
      "learning_rate": 0.0005,
      "loss": 2.7041,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.26586666666666664,
      "eval_loss": 2.7456977367401123,
      "eval_runtime": 6.4859,
      "eval_samples_per_second": 578.177,
      "eval_steps_per_second": 9.097,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 17.95757484436035,
      "learning_rate": 0.0005,
      "loss": 2.59,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.2872,
      "eval_loss": 2.6285910606384277,
      "eval_runtime": 5.63,
      "eval_samples_per_second": 666.073,
      "eval_steps_per_second": 10.48,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 19.4842586517334,
      "learning_rate": 0.0005,
      "loss": 2.4828,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.32106666666666667,
      "eval_loss": 2.505120277404785,
      "eval_runtime": 5.2785,
      "eval_samples_per_second": 710.429,
      "eval_steps_per_second": 11.177,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 18.53711700439453,
      "learning_rate": 0.0005,
      "loss": 2.3735,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.33226666666666665,
      "eval_loss": 2.4737603664398193,
      "eval_runtime": 5.3142,
      "eval_samples_per_second": 705.652,
      "eval_steps_per_second": 11.102,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 6.7616868019104,
      "learning_rate": 0.0005,
      "loss": 2.3016,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.33413333333333334,
      "eval_loss": 2.4510879516601562,
      "eval_runtime": 5.2655,
      "eval_samples_per_second": 712.189,
      "eval_steps_per_second": 11.205,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 22.808273315429688,
      "learning_rate": 0.0005,
      "loss": 2.2126,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.33973333333333333,
      "eval_loss": 2.455998182296753,
      "eval_runtime": 7.1533,
      "eval_samples_per_second": 524.236,
      "eval_steps_per_second": 8.248,
      "step": 2997
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.482671885603328e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}