trainer_state.json

{
  "best_metric": 0.9474666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0810/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": 1.1267286539077759,
      "learning_rate": 2.8514533018536286e-05,
      "loss": 1.2279,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9210666666666667,
      "eval_loss": 0.3281845450401306,
      "eval_runtime": 6.2682,
      "eval_samples_per_second": 598.259,
      "eval_steps_per_second": 9.413,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.510447025299072,
      "learning_rate": 2.4352347027881003e-05,
      "loss": 0.1899,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9357333333333333,
      "eval_loss": 0.22560670971870422,
      "eval_runtime": 6.1899,
      "eval_samples_per_second": 605.829,
      "eval_steps_per_second": 9.532,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.1775122582912445,
      "learning_rate": 1.8337814009344716e-05,
      "loss": 0.0842,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9434666666666667,
      "eval_loss": 0.20121514797210693,
      "eval_runtime": 6.0967,
      "eval_samples_per_second": 615.084,
      "eval_steps_per_second": 9.677,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.07871516793966293,
      "learning_rate": 1.1662185990655285e-05,
      "loss": 0.0396,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9421333333333334,
      "eval_loss": 0.19765575230121613,
      "eval_runtime": 6.137,
      "eval_samples_per_second": 611.043,
      "eval_steps_per_second": 9.614,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.24383527040481567,
      "learning_rate": 5.647652972118998e-06,
      "loss": 0.0203,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9424,
      "eval_loss": 0.20603416860103607,
      "eval_runtime": 6.0935,
      "eval_samples_per_second": 615.409,
      "eval_steps_per_second": 9.682,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.166597843170166,
      "learning_rate": 1.4854669814637145e-06,
      "loss": 0.0129,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9461333333333334,
      "eval_loss": 0.19443659484386444,
      "eval_runtime": 6.1617,
      "eval_samples_per_second": 608.596,
      "eval_steps_per_second": 9.575,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.362311416175377e-11,
      "loss": 0.0099,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9474666666666667,
      "eval_loss": 0.1978539079427719,
      "eval_runtime": 6.0929,
      "eval_samples_per_second": 615.467,
      "eval_steps_per_second": 9.683,
      "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
}