checkpoint-110/trainer_state.json

{
  "best_metric": 6.849837779998779,
  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/lm-training/src/outputs/2023-11-15/09-23-00/checkpoints/checkpoint-110",
  "epoch": 0.88,
  "eval_steps": 10,
  "global_step": 110,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "learning_rate": 4.866666666666667e-05,
      "loss": 9.9717,
      "step": 10
    },
    {
      "epoch": 0.08,
      "eval_loss": 9.324381828308105,
      "eval_runtime": 2.0042,
      "eval_samples_per_second": 498.949,
      "eval_steps_per_second": 62.369,
      "step": 10
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.7333333333333336e-05,
      "loss": 9.1002,
      "step": 20
    },
    {
      "epoch": 0.16,
      "eval_loss": 8.892333984375,
      "eval_runtime": 1.9424,
      "eval_samples_per_second": 514.829,
      "eval_steps_per_second": 64.354,
      "step": 20
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.600000000000001e-05,
      "loss": 8.7625,
      "step": 30
    },
    {
      "epoch": 0.24,
      "eval_loss": 8.54922103881836,
      "eval_runtime": 1.9052,
      "eval_samples_per_second": 524.878,
      "eval_steps_per_second": 65.61,
      "step": 30
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.466666666666667e-05,
      "loss": 8.4987,
      "step": 40
    },
    {
      "epoch": 0.32,
      "eval_loss": 8.234408378601074,
      "eval_runtime": 2.0895,
      "eval_samples_per_second": 478.585,
      "eval_steps_per_second": 59.823,
      "step": 40
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 8.2485,
      "step": 50
    },
    {
      "epoch": 0.4,
      "eval_loss": 7.941398620605469,
      "eval_runtime": 2.1108,
      "eval_samples_per_second": 473.76,
      "eval_steps_per_second": 59.22,
      "step": 50
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.2e-05,
      "loss": 8.0802,
      "step": 60
    },
    {
      "epoch": 0.48,
      "eval_loss": 7.663424968719482,
      "eval_runtime": 1.991,
      "eval_samples_per_second": 502.251,
      "eval_steps_per_second": 62.781,
      "step": 60
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.066666666666667e-05,
      "loss": 7.7924,
      "step": 70
    },
    {
      "epoch": 0.56,
      "eval_loss": 7.419986724853516,
      "eval_runtime": 1.9753,
      "eval_samples_per_second": 506.241,
      "eval_steps_per_second": 63.28,
      "step": 70
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.933333333333333e-05,
      "loss": 7.5139,
      "step": 80
    },
    {
      "epoch": 0.64,
      "eval_loss": 7.188984394073486,
      "eval_runtime": 1.9721,
      "eval_samples_per_second": 507.08,
      "eval_steps_per_second": 63.385,
      "step": 80
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.8e-05,
      "loss": 7.3932,
      "step": 90
    },
    {
      "epoch": 0.72,
      "eval_loss": 7.042924880981445,
      "eval_runtime": 2.0739,
      "eval_samples_per_second": 482.185,
      "eval_steps_per_second": 60.273,
      "step": 90
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 7.334,
      "step": 100
    },
    {
      "epoch": 0.8,
      "eval_loss": 6.937138557434082,
      "eval_runtime": 2.0003,
      "eval_samples_per_second": 499.914,
      "eval_steps_per_second": 62.489,
      "step": 100
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.5333333333333336e-05,
      "loss": 7.1699,
      "step": 110
    },
    {
      "epoch": 0.88,
      "eval_loss": 6.849837779998779,
      "eval_runtime": 2.0147,
      "eval_samples_per_second": 496.356,
      "eval_steps_per_second": 62.045,
      "step": 110
    }
  ],
  "logging_steps": 10,
  "max_steps": 375,
  "num_train_epochs": 3,
  "save_steps": 10,
  "total_flos": 13034123182080.0,
  "trial_name": null,
  "trial_params": null
}