trainer_state.json

{
  "best_metric": 0.7586206896551724,
  "best_model_checkpoint": "segformer-class-classWeights-augmentation/checkpoint-15",
  "epoch": 13.333333333333334,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.67,
      "eval_accuracy": 0.3103448275862069,
      "eval_loss": 1.2027289867401123,
      "eval_runtime": 1.1652,
      "eval_samples_per_second": 24.889,
      "eval_steps_per_second": 0.858,
      "step": 1
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3448275862068966,
      "eval_loss": 1.121168851852417,
      "eval_runtime": 0.8076,
      "eval_samples_per_second": 35.908,
      "eval_steps_per_second": 1.238,
      "step": 3
    },
    {
      "epoch": 2.67,
      "eval_accuracy": 0.41379310344827586,
      "eval_loss": 1.0648072957992554,
      "eval_runtime": 0.7994,
      "eval_samples_per_second": 36.278,
      "eval_steps_per_second": 1.251,
      "step": 4
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5172413793103449,
      "eval_loss": 0.9779136776924133,
      "eval_runtime": 1.1072,
      "eval_samples_per_second": 26.193,
      "eval_steps_per_second": 0.903,
      "step": 6
    },
    {
      "epoch": 4.67,
      "eval_accuracy": 0.5517241379310345,
      "eval_loss": 0.9493536353111267,
      "eval_runtime": 0.8027,
      "eval_samples_per_second": 36.129,
      "eval_steps_per_second": 1.246,
      "step": 7
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5862068965517241,
      "eval_loss": 0.9168237447738647,
      "eval_runtime": 0.8186,
      "eval_samples_per_second": 35.426,
      "eval_steps_per_second": 1.222,
      "step": 9
    },
    {
      "epoch": 6.67,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.9535,
      "step": 10
    },
    {
      "epoch": 6.67,
      "eval_accuracy": 0.6551724137931034,
      "eval_loss": 0.8808135390281677,
      "eval_runtime": 1.1217,
      "eval_samples_per_second": 25.853,
      "eval_steps_per_second": 0.891,
      "step": 10
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7241379310344828,
      "eval_loss": 0.8135725855827332,
      "eval_runtime": 0.8159,
      "eval_samples_per_second": 35.544,
      "eval_steps_per_second": 1.226,
      "step": 12
    },
    {
      "epoch": 8.67,
      "eval_accuracy": 0.7241379310344828,
      "eval_loss": 0.8014845848083496,
      "eval_runtime": 0.8311,
      "eval_samples_per_second": 34.894,
      "eval_steps_per_second": 1.203,
      "step": 13
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.7586206896551724,
      "eval_loss": 0.7727243900299072,
      "eval_runtime": 0.8049,
      "eval_samples_per_second": 36.028,
      "eval_steps_per_second": 1.242,
      "step": 15
    },
    {
      "epoch": 10.67,
      "eval_accuracy": 0.7586206896551724,
      "eval_loss": 0.7510485053062439,
      "eval_runtime": 0.7895,
      "eval_samples_per_second": 36.733,
      "eval_steps_per_second": 1.267,
      "step": 16
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.7586206896551724,
      "eval_loss": 0.6996777653694153,
      "eval_runtime": 0.8164,
      "eval_samples_per_second": 35.52,
      "eval_steps_per_second": 1.225,
      "step": 18
    },
    {
      "epoch": 12.67,
      "eval_accuracy": 0.7586206896551724,
      "eval_loss": 0.6856328248977661,
      "eval_runtime": 1.1618,
      "eval_samples_per_second": 24.961,
      "eval_steps_per_second": 0.861,
      "step": 19
    },
    {
      "epoch": 13.33,
      "learning_rate": 0.0,
      "loss": 0.5181,
      "step": 20
    },
    {
      "epoch": 13.33,
      "eval_accuracy": 0.7586206896551724,
      "eval_loss": 0.6802691221237183,
      "eval_runtime": 1.6744,
      "eval_samples_per_second": 17.32,
      "eval_steps_per_second": 0.597,
      "step": 20
    },
    {
      "epoch": 13.33,
      "step": 20,
      "total_flos": 8.682439554251366e+16,
      "train_loss": 0.735786771774292,
      "train_runtime": 315.7979,
      "train_samples_per_second": 16.53,
      "train_steps_per_second": 0.063
    }
  ],
  "max_steps": 20,
  "num_train_epochs": 20,
  "total_flos": 8.682439554251366e+16,
  "trial_name": null,
  "trial_params": null
}