trainer_state.json

{
  "best_metric": 0.8906666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0031/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": 130.635009765625,
      "learning_rate": 6.222222222222222e-05,
      "loss": 3.2853,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6701333333333334,
      "eval_loss": 1.8377747535705566,
      "eval_runtime": 7.1573,
      "eval_samples_per_second": 523.942,
      "eval_steps_per_second": 8.243,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 82.53572082519531,
      "learning_rate": 5.444444444444444e-05,
      "loss": 0.9163,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8328,
      "eval_loss": 0.567460834980011,
      "eval_runtime": 7.486,
      "eval_samples_per_second": 500.936,
      "eval_steps_per_second": 7.881,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 73.57218170166016,
      "learning_rate": 4.666666666666666e-05,
      "loss": 0.4562,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8664,
      "eval_loss": 0.4417571425437927,
      "eval_runtime": 7.1196,
      "eval_samples_per_second": 526.715,
      "eval_steps_per_second": 8.287,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 206.50604248046875,
      "learning_rate": 3.8888888888888884e-05,
      "loss": 0.3153,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8778666666666667,
      "eval_loss": 0.3905840814113617,
      "eval_runtime": 7.205,
      "eval_samples_per_second": 520.469,
      "eval_steps_per_second": 8.189,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 89.42232513427734,
      "learning_rate": 3.111111111111111e-05,
      "loss": 0.2364,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8856,
      "eval_loss": 0.3671533167362213,
      "eval_runtime": 7.1666,
      "eval_samples_per_second": 523.26,
      "eval_steps_per_second": 8.233,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 141.12457275390625,
      "learning_rate": 2.333333333333333e-05,
      "loss": 0.1888,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8896,
      "eval_loss": 0.3568798899650574,
      "eval_runtime": 7.0987,
      "eval_samples_per_second": 528.263,
      "eval_steps_per_second": 8.311,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 92.40489959716797,
      "learning_rate": 1.5555555555555555e-05,
      "loss": 0.1488,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8906666666666667,
      "eval_loss": 0.3592965602874756,
      "eval_runtime": 7.0734,
      "eval_samples_per_second": 530.159,
      "eval_steps_per_second": 8.341,
      "step": 2331
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}