trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 53,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 8.333333333333333e-08,
      "logits/chosen": -2.8462421894073486,
      "logits/rejected": -2.8283610343933105,
      "logps/chosen": -274.7393798828125,
      "logps/rejected": -204.42575073242188,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.911172937635942e-07,
      "logits/chosen": -2.8527991771698,
      "logits/rejected": -2.8377315998077393,
      "logps/chosen": -305.9073181152344,
      "logps/rejected": -295.8478698730469,
      "loss": 0.6914,
      "rewards/accuracies": 0.4513888955116272,
      "rewards/chosen": 0.0023197412956506014,
      "rewards/margins": 0.0025084479711949825,
      "rewards/rejected": -0.00018870655912905931,
      "step": 10
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.982949361823388e-07,
      "logits/chosen": -2.859750270843506,
      "logits/rejected": -2.880180835723877,
      "logps/chosen": -295.7957458496094,
      "logps/rejected": -332.6015930175781,
      "loss": 0.6653,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.02228003740310669,
      "rewards/margins": 0.059415679425001144,
      "rewards/rejected": -0.037135638296604156,
      "step": 20
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.416462557480814e-07,
      "logits/chosen": -2.843632936477661,
      "logits/rejected": -2.8286781311035156,
      "logps/chosen": -310.9751892089844,
      "logps/rejected": -322.77532958984375,
      "loss": 0.6283,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.027804672718048096,
      "rewards/margins": 0.1940310001373291,
      "rewards/rejected": -0.1662263423204422,
      "step": 30
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.859303711029939e-08,
      "logits/chosen": -2.8175368309020996,
      "logits/rejected": -2.821326494216919,
      "logps/chosen": -274.8536682128906,
      "logps/rejected": -349.11505126953125,
      "loss": 0.6099,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.002673505572602153,
      "rewards/margins": 0.2130366563796997,
      "rewards/rejected": -0.21036314964294434,
      "step": 40
    },
    {
      "epoch": 0.94,
      "learning_rate": 5.009573740853313e-09,
      "logits/chosen": -2.8560073375701904,
      "logits/rejected": -2.867896556854248,
      "logps/chosen": -307.2721862792969,
      "logps/rejected": -350.257568359375,
      "loss": 0.5984,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.0023462946992367506,
      "rewards/margins": 0.33910489082336426,
      "rewards/rejected": -0.336758553981781,
      "step": 50
    },
    {
      "epoch": 1.0,
      "step": 53,
      "total_flos": 0.0,
      "train_loss": 0.6385756753525644,
      "train_runtime": 422.6241,
      "train_samples_per_second": 15.972,
      "train_steps_per_second": 0.125
    }
  ],
  "logging_steps": 10,
  "max_steps": 53,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "trial_name": null,
  "trial_params": null
}