trainer_state.json

{
  "best_metric": 0.733502538071066,
  "best_model_checkpoint": "../results/elpv-vit/checkpoint-828",
  "epoch": 14.0,
  "eval_steps": 500,
  "global_step": 966,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5786802030456852,
      "eval_loss": 1.1146228313446045,
      "eval_runtime": 15.1595,
      "eval_samples_per_second": 25.99,
      "eval_steps_per_second": 1.649,
      "step": 69
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5786802030456852,
      "eval_loss": 0.9812496304512024,
      "eval_runtime": 4.4653,
      "eval_samples_per_second": 88.237,
      "eval_steps_per_second": 5.599,
      "step": 138
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6472081218274112,
      "eval_loss": 0.8884502053260803,
      "eval_runtime": 2.1388,
      "eval_samples_per_second": 184.219,
      "eval_steps_per_second": 11.689,
      "step": 207
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7081218274111675,
      "eval_loss": 0.7930060625076294,
      "eval_runtime": 2.1153,
      "eval_samples_per_second": 186.265,
      "eval_steps_per_second": 11.819,
      "step": 276
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6928934010152284,
      "eval_loss": 0.8019062280654907,
      "eval_runtime": 7.7818,
      "eval_samples_per_second": 50.631,
      "eval_steps_per_second": 3.213,
      "step": 345
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6878172588832487,
      "eval_loss": 0.8009192943572998,
      "eval_runtime": 2.0611,
      "eval_samples_per_second": 191.162,
      "eval_steps_per_second": 12.13,
      "step": 414
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6852791878172588,
      "eval_loss": 0.7984281182289124,
      "eval_runtime": 12.8592,
      "eval_samples_per_second": 30.639,
      "eval_steps_per_second": 1.944,
      "step": 483
    },
    {
      "epoch": 7.25,
      "learning_rate": 2.6812428078250863e-05,
      "loss": 0.8194,
      "step": 500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7106598984771574,
      "eval_loss": 0.7714311480522156,
      "eval_runtime": 2.123,
      "eval_samples_per_second": 185.586,
      "eval_steps_per_second": 11.776,
      "step": 552
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.7081218274111675,
      "eval_loss": 0.7666618824005127,
      "eval_runtime": 2.021,
      "eval_samples_per_second": 194.952,
      "eval_steps_per_second": 12.37,
      "step": 621
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.7233502538071066,
      "eval_loss": 0.7302789688110352,
      "eval_runtime": 14.4927,
      "eval_samples_per_second": 27.186,
      "eval_steps_per_second": 1.725,
      "step": 690
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.7284263959390863,
      "eval_loss": 0.7321494817733765,
      "eval_runtime": 2.2455,
      "eval_samples_per_second": 175.465,
      "eval_steps_per_second": 11.134,
      "step": 759
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.733502538071066,
      "eval_loss": 0.7373437285423279,
      "eval_runtime": 5.5421,
      "eval_samples_per_second": 71.092,
      "eval_steps_per_second": 4.511,
      "step": 828
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.6903553299492385,
      "eval_loss": 0.8051439523696899,
      "eval_runtime": 12.7177,
      "eval_samples_per_second": 30.98,
      "eval_steps_per_second": 1.966,
      "step": 897
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.7258883248730964,
      "eval_loss": 0.7687421441078186,
      "eval_runtime": 4.2192,
      "eval_samples_per_second": 93.382,
      "eval_steps_per_second": 5.925,
      "step": 966
    },
    {
      "epoch": 14.0,
      "step": 966,
      "total_flos": 4.790950643943604e+18,
      "train_loss": 0.63200661163646,
      "train_runtime": 1457.5958,
      "train_samples_per_second": 42.415,
      "train_steps_per_second": 0.663
    }
  ],
  "logging_steps": 500,
  "max_steps": 966,
  "num_train_epochs": 14,
  "save_steps": 500,
  "total_flos": 4.790950643943604e+18,
  "trial_name": null,
  "trial_params": null
}