trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 274,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2191780821917808,
      "grad_norm": 5.121945381164551,
      "learning_rate": 9.997441533484527e-05,
      "log_odds_chosen": 7.0944390296936035,
      "log_odds_ratio": -0.043758511543273926,
      "logits/chosen": -11.323039054870605,
      "logits/rejected": -11.33891487121582,
      "logps/chosen": -0.5756699442863464,
      "logps/rejected": -6.692573547363281,
      "loss": 0.721,
      "nll_loss": 0.716595470905304,
      "rewards/accuracies": 0.9791666865348816,
      "rewards/chosen": -0.057566989213228226,
      "rewards/margins": 0.6116903424263,
      "rewards/rejected": -0.6692573428153992,
      "step": 60
    },
    {
      "epoch": 0.4383561643835616,
      "grad_norm": 6.375761032104492,
      "learning_rate": 9.573779290555039e-05,
      "log_odds_chosen": 7.421543598175049,
      "log_odds_ratio": -0.0334482379257679,
      "logits/chosen": -12.235654830932617,
      "logits/rejected": -12.264952659606934,
      "logps/chosen": -0.5262126326560974,
      "logps/rejected": -6.922983169555664,
      "loss": 0.6832,
      "nll_loss": 0.679860532283783,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -0.0526212677359581,
      "rewards/margins": 0.6396770477294922,
      "rewards/rejected": -0.6922982931137085,
      "step": 120
    },
    {
      "epoch": 0.6575342465753424,
      "grad_norm": 5.337998390197754,
      "learning_rate": 8.484270686572078e-05,
      "log_odds_chosen": 7.344568729400635,
      "log_odds_ratio": -0.047614481300115585,
      "logits/chosen": -11.1071138381958,
      "logits/rejected": -11.16305160522461,
      "logps/chosen": -0.6026201844215393,
      "logps/rejected": -7.023316860198975,
      "loss": 0.7666,
      "nll_loss": 0.7618839144706726,
      "rewards/accuracies": 0.9791666865348816,
      "rewards/chosen": -0.06026201695203781,
      "rewards/margins": 0.6420697569847107,
      "rewards/rejected": -0.7023317217826843,
      "step": 180
    },
    {
      "epoch": 0.8767123287671232,
      "grad_norm": 7.20003080368042,
      "learning_rate": 6.887525329304994e-05,
      "log_odds_chosen": 6.544267177581787,
      "log_odds_ratio": -0.08398038893938065,
      "logits/chosen": -9.961735725402832,
      "logits/rejected": -9.985981941223145,
      "logps/chosen": -0.7504870295524597,
      "logps/rejected": -6.561529636383057,
      "loss": 0.8312,
      "nll_loss": 0.8227914571762085,
      "rewards/accuracies": 0.9583333134651184,
      "rewards/chosen": -0.07504869997501373,
      "rewards/margins": 0.5811043381690979,
      "rewards/rejected": -0.6561529636383057,
      "step": 240
    }
  ],
  "logging_steps": 60,
  "max_steps": 546,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}