trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9925925925925926,
  "eval_steps": 500,
  "global_step": 67,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.014814814814814815,
      "grad_norm": 0.13149102014428649,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 0.0831,
      "step": 1
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 0.0648243838618961,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 0.0773,
      "step": 5
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 0.12634669899371723,
      "learning_rate": 9.938441702975689e-06,
      "loss": 0.0692,
      "step": 10
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.11103203868763059,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.0623,
      "step": 15
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 0.08559439689483059,
      "learning_rate": 8.885729807284855e-06,
      "loss": 0.0589,
      "step": 20
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 0.05908430257437008,
      "learning_rate": 7.938926261462366e-06,
      "loss": 0.0427,
      "step": 25
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.043413213669282774,
      "learning_rate": 6.7918397477265e-06,
      "loss": 0.0427,
      "step": 30
    },
    {
      "epoch": 0.5185185185185185,
      "grad_norm": 0.04867781195090588,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.0437,
      "step": 35
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 0.04647862489770088,
      "learning_rate": 4.217827674798845e-06,
      "loss": 0.0428,
      "step": 40
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.03906049277069991,
      "learning_rate": 2.966316784621e-06,
      "loss": 0.0367,
      "step": 45
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 0.03242620502990518,
      "learning_rate": 1.8533980447508138e-06,
      "loss": 0.0318,
      "step": 50
    },
    {
      "epoch": 0.8148148148148148,
      "grad_norm": 0.03928624653755023,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.0326,
      "step": 55
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.04074437315257584,
      "learning_rate": 3.320978675139919e-07,
      "loss": 0.0339,
      "step": 60
    },
    {
      "epoch": 0.9629629629629629,
      "grad_norm": 0.04399843384947704,
      "learning_rate": 2.7390523158633552e-08,
      "loss": 0.0393,
      "step": 65
    },
    {
      "epoch": 0.9925925925925926,
      "step": 67,
      "total_flos": 1.64379520235733e+17,
      "train_loss": 0.0467690047043473,
      "train_runtime": 743.4434,
      "train_samples_per_second": 2.179,
      "train_steps_per_second": 0.09
    }
  ],
  "logging_steps": 5,
  "max_steps": 67,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.64379520235733e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}