checkpoint-1200/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 39.025,
  "global_step": 1200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 4.9999999999999996e-06,
      "loss": 10.9361,
      "step": 1
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00025,
      "loss": 8.2553,
      "step": 50
    },
    {
      "epoch": 3.01,
      "learning_rate": 0.0005,
      "loss": 6.7321,
      "step": 100
    },
    {
      "epoch": 4.03,
      "learning_rate": 0.0005833333333333333,
      "loss": 6.4515,
      "step": 150
    },
    {
      "epoch": 6.02,
      "learning_rate": 0.0005555555555555556,
      "loss": 6.2916,
      "step": 200
    },
    {
      "epoch": 8.01,
      "learning_rate": 0.0005277777777777777,
      "loss": 6.103,
      "step": 250
    },
    {
      "epoch": 9.03,
      "learning_rate": 0.0005,
      "loss": 5.7716,
      "step": 300
    },
    {
      "epoch": 11.02,
      "learning_rate": 0.00047222222222222224,
      "loss": 5.4902,
      "step": 350
    },
    {
      "epoch": 13.01,
      "learning_rate": 0.00044444444444444436,
      "loss": 5.1844,
      "step": 400
    },
    {
      "epoch": 14.03,
      "learning_rate": 0.00041666666666666664,
      "loss": 4.8669,
      "step": 450
    },
    {
      "epoch": 16.02,
      "learning_rate": 0.00038888888888888887,
      "loss": 4.6547,
      "step": 500
    },
    {
      "epoch": 18.01,
      "learning_rate": 0.0003611111111111111,
      "loss": 4.4311,
      "step": 550
    },
    {
      "epoch": 19.02,
      "learning_rate": 0.0003333333333333333,
      "loss": 4.1923,
      "step": 600
    },
    {
      "epoch": 21.02,
      "learning_rate": 0.00030555555555555555,
      "loss": 4.0473,
      "step": 650
    },
    {
      "epoch": 23.01,
      "learning_rate": 0.0002777777777777778,
      "loss": 3.9047,
      "step": 700
    },
    {
      "epoch": 24.02,
      "learning_rate": 0.00025,
      "loss": 3.7631,
      "step": 750
    },
    {
      "epoch": 26.02,
      "learning_rate": 0.00022222222222222218,
      "loss": 3.7075,
      "step": 800
    },
    {
      "epoch": 28.01,
      "learning_rate": 0.00019444444444444443,
      "loss": 3.6326,
      "step": 850
    },
    {
      "epoch": 29.02,
      "learning_rate": 0.00016666666666666666,
      "loss": 3.5357,
      "step": 900
    },
    {
      "epoch": 31.02,
      "learning_rate": 0.0001388888888888889,
      "loss": 3.5105,
      "step": 950
    },
    {
      "epoch": 33.01,
      "learning_rate": 0.00011111111111111109,
      "loss": 3.4606,
      "step": 1000
    },
    {
      "epoch": 33.01,
      "eval_loss": 4.271039009094238,
      "eval_runtime": 1.2465,
      "eval_samples_per_second": 89.851,
      "eval_steps_per_second": 5.616,
      "step": 1000
    },
    {
      "epoch": 33.01,
      "eval_/scratch/ykyao/projects/multilingual-LM/training/multilingual_dataset.py_loss": 4.271039009094238,
      "eval_/scratch/ykyao/projects/multilingual-LM/training/multilingual_dataset.py_ppl": 71.59598586766586,
      "eval_/scratch/ykyao/projects/multilingual-LM/training/multilingual_dataset.py_runtime": 1.2465,
      "eval_/scratch/ykyao/projects/multilingual-LM/training/multilingual_dataset.py_samples_per_second": 89.851,
      "step": 1000
    },
    {
      "epoch": 34.02,
      "learning_rate": 8.333333333333333e-05,
      "loss": 3.3886,
      "step": 1050
    },
    {
      "epoch": 36.02,
      "learning_rate": 5.5555555555555545e-05,
      "loss": 3.3823,
      "step": 1100
    },
    {
      "epoch": 38.01,
      "learning_rate": 2.7777777777777772e-05,
      "loss": 3.3532,
      "step": 1150
    },
    {
      "epoch": 39.02,
      "learning_rate": 0.0,
      "loss": 3.305,
      "step": 1200
    }
  ],
  "max_steps": 1200,
  "num_train_epochs": 9223372036854775807,
  "total_flos": 3.2545751629824e+17,
  "trial_name": null,
  "trial_params": null
}