trainer_state.json

{
  "best_metric": 0.7586850523948669,
  "best_model_checkpoint": "experiments/checkpoint-42",
  "epoch": 78.0,
  "global_step": 98,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 4.0,
      "learning_rate": 4.9999999999999996e-05,
      "loss": 0.3998,
      "step": 5
    },
    {
      "epoch": 8.0,
      "learning_rate": 9.999999999999999e-05,
      "loss": 0.3916,
      "step": 10
    },
    {
      "epoch": 11.0,
      "eval_loss": 1.4091475009918213,
      "eval_runtime": 4.4251,
      "eval_samples_per_second": 2.26,
      "eval_steps_per_second": 0.452,
      "step": 14
    },
    {
      "epoch": 12.0,
      "learning_rate": 0.00015,
      "loss": 0.3745,
      "step": 15
    },
    {
      "epoch": 16.0,
      "learning_rate": 0.00019999999999999998,
      "loss": 0.3387,
      "step": 20
    },
    {
      "epoch": 20.0,
      "learning_rate": 0.00025,
      "loss": 0.2894,
      "step": 25
    },
    {
      "epoch": 22.4,
      "eval_loss": 0.8969688415527344,
      "eval_runtime": 4.4836,
      "eval_samples_per_second": 2.23,
      "eval_steps_per_second": 0.446,
      "step": 28
    },
    {
      "epoch": 24.0,
      "learning_rate": 0.0003,
      "loss": 0.2322,
      "step": 30
    },
    {
      "epoch": 28.0,
      "learning_rate": 0.00027857142857142854,
      "loss": 0.1964,
      "step": 35
    },
    {
      "epoch": 32.0,
      "learning_rate": 0.0002571428571428571,
      "loss": 0.1739,
      "step": 40
    },
    {
      "epoch": 34.0,
      "eval_loss": 0.7586850523948669,
      "eval_runtime": 4.4178,
      "eval_samples_per_second": 2.264,
      "eval_steps_per_second": 0.453,
      "step": 42
    },
    {
      "epoch": 36.0,
      "learning_rate": 0.00023571428571428569,
      "loss": 0.157,
      "step": 45
    },
    {
      "epoch": 40.0,
      "learning_rate": 0.00021428571428571427,
      "loss": 0.1408,
      "step": 50
    },
    {
      "epoch": 44.0,
      "learning_rate": 0.00019285714285714286,
      "loss": 0.1264,
      "step": 55
    },
    {
      "epoch": 44.8,
      "eval_loss": 0.7922688126564026,
      "eval_runtime": 4.4244,
      "eval_samples_per_second": 2.26,
      "eval_steps_per_second": 0.452,
      "step": 56
    },
    {
      "epoch": 48.0,
      "learning_rate": 0.0001714285714285714,
      "loss": 0.1134,
      "step": 60
    },
    {
      "epoch": 52.0,
      "learning_rate": 0.00015,
      "loss": 0.1021,
      "step": 65
    },
    {
      "epoch": 56.0,
      "learning_rate": 0.00012857142857142855,
      "loss": 0.0893,
      "step": 70
    },
    {
      "epoch": 56.0,
      "eval_loss": 0.8863828778266907,
      "eval_runtime": 4.4142,
      "eval_samples_per_second": 2.265,
      "eval_steps_per_second": 0.453,
      "step": 70
    },
    {
      "epoch": 60.0,
      "learning_rate": 0.00010714285714285714,
      "loss": 0.0778,
      "step": 75
    },
    {
      "epoch": 64.0,
      "learning_rate": 8.57142857142857e-05,
      "loss": 0.0677,
      "step": 80
    },
    {
      "epoch": 67.2,
      "eval_loss": 1.0279890298843384,
      "eval_runtime": 4.4181,
      "eval_samples_per_second": 2.263,
      "eval_steps_per_second": 0.453,
      "step": 84
    },
    {
      "epoch": 68.0,
      "learning_rate": 6.428571428571427e-05,
      "loss": 0.0598,
      "step": 85
    },
    {
      "epoch": 72.0,
      "learning_rate": 4.285714285714285e-05,
      "loss": 0.0524,
      "step": 90
    },
    {
      "epoch": 76.0,
      "learning_rate": 2.1428571428571425e-05,
      "loss": 0.0474,
      "step": 95
    },
    {
      "epoch": 78.0,
      "eval_loss": 1.1172549724578857,
      "eval_runtime": 4.4004,
      "eval_samples_per_second": 2.273,
      "eval_steps_per_second": 0.455,
      "step": 98
    }
  ],
  "max_steps": 100,
  "num_train_epochs": 100,
  "total_flos": 6.335270841090048e+16,
  "trial_name": null,
  "trial_params": null
}