trainer_state.json

{
  "best_metric": 0.9660714285714286,
  "best_model_checkpoint": "swin-finetuned-food101/checkpoint-105",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 105,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 7.0990753173828125,
      "learning_rate": 4.545454545454546e-05,
      "loss": 1.8603,
      "step": 10
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 7.225792407989502,
      "learning_rate": 4.5212765957446815e-05,
      "loss": 0.806,
      "step": 20
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 10.552105903625488,
      "learning_rate": 3.9893617021276594e-05,
      "loss": 0.2786,
      "step": 30
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9535714285714286,
      "eval_loss": 0.14325273036956787,
      "eval_runtime": 11.2243,
      "eval_samples_per_second": 49.892,
      "eval_steps_per_second": 3.118,
      "step": 35
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 7.090443134307861,
      "learning_rate": 3.4574468085106386e-05,
      "loss": 0.182,
      "step": 40
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 10.5905122756958,
      "learning_rate": 2.925531914893617e-05,
      "loss": 0.1237,
      "step": 50
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 4.789523601531982,
      "learning_rate": 2.393617021276596e-05,
      "loss": 0.1119,
      "step": 60
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.802328109741211,
      "learning_rate": 1.8617021276595745e-05,
      "loss": 0.1035,
      "step": 70
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9625,
      "eval_loss": 0.11008361726999283,
      "eval_runtime": 11.1386,
      "eval_samples_per_second": 50.276,
      "eval_steps_per_second": 3.142,
      "step": 70
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 2.4177331924438477,
      "learning_rate": 1.3297872340425532e-05,
      "loss": 0.0647,
      "step": 80
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 3.3316473960876465,
      "learning_rate": 7.97872340425532e-06,
      "loss": 0.0381,
      "step": 90
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 2.1031341552734375,
      "learning_rate": 2.6595744680851065e-06,
      "loss": 0.0288,
      "step": 100
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9660714285714286,
      "eval_loss": 0.10036059468984604,
      "eval_runtime": 11.3867,
      "eval_samples_per_second": 49.18,
      "eval_steps_per_second": 3.074,
      "step": 105
    },
    {
      "epoch": 3.0,
      "step": 105,
      "total_flos": 5.2651323892629504e+17,
      "train_loss": 0.34459258970760165,
      "train_runtime": 349.853,
      "train_samples_per_second": 19.208,
      "train_steps_per_second": 0.3
    }
  ],
  "logging_steps": 10,
  "max_steps": 105,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 5.2651323892629504e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}