checkpoint-1100/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 39.00666666666667,
  "global_step": 1100,
  "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": 11.0688,
      "step": 1
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00025,
      "loss": 8.3635,
      "step": 50
    },
    {
      "epoch": 3.01,
      "learning_rate": 0.0005,
      "loss": 6.4596,
      "step": 100
    },
    {
      "epoch": 5.01,
      "learning_rate": 0.0005833333333333333,
      "loss": 5.8337,
      "step": 150
    },
    {
      "epoch": 7.0,
      "learning_rate": 0.0005555555555555556,
      "loss": 5.4472,
      "step": 200
    },
    {
      "epoch": 8.02,
      "learning_rate": 0.0005277777777777777,
      "loss": 5.1974,
      "step": 250
    },
    {
      "epoch": 10.02,
      "learning_rate": 0.0005,
      "loss": 5.1091,
      "step": 300
    },
    {
      "epoch": 12.01,
      "learning_rate": 0.00047222222222222224,
      "loss": 4.9724,
      "step": 350
    },
    {
      "epoch": 14.01,
      "learning_rate": 0.00044444444444444436,
      "loss": 4.8054,
      "step": 400
    },
    {
      "epoch": 16.0,
      "learning_rate": 0.00041666666666666664,
      "loss": 4.5929,
      "step": 450
    },
    {
      "epoch": 17.02,
      "learning_rate": 0.00038888888888888887,
      "loss": 4.3775,
      "step": 500
    },
    {
      "epoch": 19.02,
      "learning_rate": 0.0003611111111111111,
      "loss": 4.2825,
      "step": 550
    },
    {
      "epoch": 21.01,
      "learning_rate": 0.0003333333333333333,
      "loss": 4.1618,
      "step": 600
    },
    {
      "epoch": 23.0,
      "learning_rate": 0.00030555555555555555,
      "loss": 4.0581,
      "step": 650
    },
    {
      "epoch": 24.02,
      "learning_rate": 0.0002777777777777778,
      "loss": 3.9257,
      "step": 700
    },
    {
      "epoch": 26.02,
      "learning_rate": 0.00025,
      "loss": 3.8778,
      "step": 750
    },
    {
      "epoch": 28.01,
      "learning_rate": 0.00022222222222222218,
      "loss": 3.8028,
      "step": 800
    },
    {
      "epoch": 30.01,
      "learning_rate": 0.00019444444444444443,
      "loss": 3.7326,
      "step": 850
    },
    {
      "epoch": 32.0,
      "learning_rate": 0.00016666666666666666,
      "loss": 3.6708,
      "step": 900
    },
    {
      "epoch": 33.02,
      "learning_rate": 0.0001388888888888889,
      "loss": 3.5837,
      "step": 950
    },
    {
      "epoch": 35.02,
      "learning_rate": 0.00011111111111111109,
      "loss": 3.5664,
      "step": 1000
    },
    {
      "epoch": 35.02,
      "eval_loss": 4.406866073608398,
      "eval_runtime": 1.7783,
      "eval_samples_per_second": 52.296,
      "eval_steps_per_second": 3.374,
      "step": 1000
    },
    {
      "epoch": 35.02,
      "eval_/local/xiulyang/mission-impossible-language-models/training/babylm_dataset.py_loss": 4.406866073608398,
      "eval_/local/xiulyang/mission-impossible-language-models/training/babylm_dataset.py_ppl": 82.01204064414016,
      "eval_/local/xiulyang/mission-impossible-language-models/training/babylm_dataset.py_runtime": 1.7783,
      "eval_/local/xiulyang/mission-impossible-language-models/training/babylm_dataset.py_samples_per_second": 52.296,
      "step": 1000
    },
    {
      "epoch": 37.01,
      "learning_rate": 8.333333333333333e-05,
      "loss": 3.5206,
      "step": 1050
    },
    {
      "epoch": 39.01,
      "learning_rate": 5.5555555555555545e-05,
      "loss": 3.4829,
      "step": 1100
    }
  ],
  "max_steps": 1200,
  "num_train_epochs": 9223372036854775807,
  "total_flos": 2.98578927550464e+17,
  "trial_name": null,
  "trial_params": null
}