trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 57,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 5.16129094151242,
      "learning_rate": 4.9394909565685894e-05,
      "loss": 0.3072,
      "num_input_tokens_seen": 588672,
      "step": 5,
      "train_runtime": 43.9847,
      "train_tokens_per_second": 13383.571
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 1.1352466919383923,
      "learning_rate": 4.698684378016222e-05,
      "loss": 0.2503,
      "num_input_tokens_seen": 1191152,
      "step": 10,
      "train_runtime": 84.6669,
      "train_tokens_per_second": 14068.69
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 0.9235059881629116,
      "learning_rate": 4.2919562829211283e-05,
      "loss": 0.1434,
      "num_input_tokens_seen": 1792128,
      "step": 15,
      "train_runtime": 124.6102,
      "train_tokens_per_second": 14381.876
    },
    {
      "epoch": 1.054054054054054,
      "grad_norm": 1.1350634950914822,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.119,
      "num_input_tokens_seen": 2338720,
      "step": 20,
      "train_runtime": 160.5005,
      "train_tokens_per_second": 14571.423
    },
    {
      "epoch": 1.3243243243243243,
      "grad_norm": 0.6261741496657048,
      "learning_rate": 3.1137137178519985e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 2944992,
      "step": 25,
      "train_runtime": 201.2088,
      "train_tokens_per_second": 14636.494
    },
    {
      "epoch": 1.5945945945945947,
      "grad_norm": 0.6896041221746694,
      "learning_rate": 2.4311141440795953e-05,
      "loss": 0.0714,
      "num_input_tokens_seen": 3543936,
      "step": 30,
      "train_runtime": 240.9507,
      "train_tokens_per_second": 14708.137
    },
    {
      "epoch": 1.864864864864865,
      "grad_norm": 0.5393100198337258,
      "learning_rate": 1.7537129724957642e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 4133600,
      "step": 35,
      "train_runtime": 280.1915,
      "train_tokens_per_second": 14752.768
    },
    {
      "epoch": 2.108108108108108,
      "grad_norm": 0.4797863235648709,
      "learning_rate": 1.1326296046939333e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 4674192,
      "step": 40,
      "train_runtime": 316.3527,
      "train_tokens_per_second": 14775.256
    },
    {
      "epoch": 2.3783783783783785,
      "grad_norm": 0.28429321473085484,
      "learning_rate": 6.147334755577596e-06,
      "loss": 0.0349,
      "num_input_tokens_seen": 5271408,
      "step": 45,
      "train_runtime": 357.1448,
      "train_tokens_per_second": 14759.861
    },
    {
      "epoch": 2.6486486486486487,
      "grad_norm": 0.3086922220806531,
      "learning_rate": 2.391070982560564e-06,
      "loss": 0.028,
      "num_input_tokens_seen": 5854528,
      "step": 50,
      "train_runtime": 395.8893,
      "train_tokens_per_second": 14788.297
    },
    {
      "epoch": 2.918918918918919,
      "grad_norm": 0.3181234653618051,
      "learning_rate": 3.4096741493194197e-07,
      "loss": 0.0241,
      "num_input_tokens_seen": 6461216,
      "step": 55,
      "train_runtime": 437.0222,
      "train_tokens_per_second": 14784.639
    },
    {
      "epoch": 3.0,
      "num_input_tokens_seen": 6642400,
      "step": 57,
      "total_flos": 13254831570944.0,
      "train_loss": 0.10312069207429886,
      "train_runtime": 603.3131,
      "train_samples_per_second": 11.735,
      "train_steps_per_second": 0.094
    }
  ],
  "logging_steps": 5,
  "max_steps": 57,
  "num_input_tokens_seen": 6642400,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 13254831570944.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}