trainer_state.json

{
  "best_metric": 0.5298666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0500/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": 3.2565,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.20613333333333334,
      "eval_loss": 2.939532518386841,
      "eval_runtime": 12.1432,
      "eval_samples_per_second": 308.815,
      "eval_steps_per_second": 4.859,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.674,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3218666666666667,
      "eval_loss": 2.506107807159424,
      "eval_runtime": 12.2383,
      "eval_samples_per_second": 306.415,
      "eval_steps_per_second": 4.821,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 8.53410816192627,
      "learning_rate": 0.0005,
      "loss": 2.2594,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3586666666666667,
      "eval_loss": 2.3375556468963623,
      "eval_runtime": 12.0955,
      "eval_samples_per_second": 310.033,
      "eval_steps_per_second": 4.878,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 12.522454261779785,
      "learning_rate": 0.0005,
      "loss": 1.9483,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.43733333333333335,
      "eval_loss": 2.031986713409424,
      "eval_runtime": 11.8594,
      "eval_samples_per_second": 316.206,
      "eval_steps_per_second": 4.975,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 1.6697,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.48346666666666666,
      "eval_loss": 1.8123379945755005,
      "eval_runtime": 12.1646,
      "eval_samples_per_second": 308.271,
      "eval_steps_per_second": 4.85,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 20.4299259185791,
      "learning_rate": 0.0005,
      "loss": 1.4148,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4901333333333333,
      "eval_loss": 1.8386976718902588,
      "eval_runtime": 12.0838,
      "eval_samples_per_second": 310.332,
      "eval_steps_per_second": 4.883,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 20.94927406311035,
      "learning_rate": 0.0005,
      "loss": 1.1505,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.508,
      "eval_loss": 1.8421908617019653,
      "eval_runtime": 11.9992,
      "eval_samples_per_second": 312.521,
      "eval_steps_per_second": 4.917,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 15.200113296508789,
      "learning_rate": 0.0005,
      "loss": 0.8885,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.5061333333333333,
      "eval_loss": 1.8651669025421143,
      "eval_runtime": 12.2064,
      "eval_samples_per_second": 307.215,
      "eval_steps_per_second": 4.834,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 24.23003578186035,
      "learning_rate": 0.0005,
      "loss": 0.6529,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.5298666666666667,
      "eval_loss": 1.932701587677002,
      "eval_runtime": 12.2412,
      "eval_samples_per_second": 306.341,
      "eval_steps_per_second": 4.82,
      "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
}