trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9874476987447699,
  "eval_steps": 500,
  "global_step": 59,
  "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": -2.809058427810669,
      "logits/rejected": -2.8124935626983643,
      "logps/chosen": -318.11346435546875,
      "logps/rejected": -229.77012634277344,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.930057285201027e-07,
      "logits/chosen": -2.7690165042877197,
      "logits/rejected": -2.7547106742858887,
      "logps/chosen": -277.6565246582031,
      "logps/rejected": -264.4493408203125,
      "loss": 0.6916,
      "rewards/accuracies": 0.4791666567325592,
      "rewards/chosen": 0.003391754813492298,
      "rewards/margins": 0.002254640683531761,
      "rewards/rejected": 0.0011371138971298933,
      "step": 10
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.187457503795526e-07,
      "logits/chosen": -2.79248046875,
      "logits/rejected": -2.789917230606079,
      "logps/chosen": -264.5072326660156,
      "logps/rejected": -252.63845825195312,
      "loss": 0.6752,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.026929911226034164,
      "rewards/margins": 0.03922198340296745,
      "rewards/rejected": -0.012292074970901012,
      "step": 20
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.8691164100062034e-07,
      "logits/chosen": -2.7963314056396484,
      "logits/rejected": -2.796976089477539,
      "logps/chosen": -298.3793640136719,
      "logps/rejected": -256.82110595703125,
      "loss": 0.6504,
      "rewards/accuracies": 0.6656249761581421,
      "rewards/chosen": -0.021613802760839462,
      "rewards/margins": 0.11929114162921906,
      "rewards/rejected": -0.14090493321418762,
      "step": 30
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.4248369943086995e-07,
      "logits/chosen": -2.7643628120422363,
      "logits/rejected": -2.7510218620300293,
      "logps/chosen": -265.435302734375,
      "logps/rejected": -256.5634765625,
      "loss": 0.6363,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.1126113086938858,
      "rewards/margins": 0.15844564139842987,
      "rewards/rejected": -0.2710569500923157,
      "step": 40
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.473909705816111e-08,
      "logits/chosen": -2.76975679397583,
      "logits/rejected": -2.7606043815612793,
      "logps/chosen": -280.33734130859375,
      "logps/rejected": -287.0810241699219,
      "loss": 0.6219,
      "rewards/accuracies": 0.640625,
      "rewards/chosen": -0.20978038012981415,
      "rewards/margins": 0.129803866147995,
      "rewards/rejected": -0.33958423137664795,
      "step": 50
    },
    {
      "epoch": 0.99,
      "step": 59,
      "total_flos": 0.0,
      "train_loss": 0.6501501697604939,
      "train_runtime": 1927.4076,
      "train_samples_per_second": 7.929,
      "train_steps_per_second": 0.031
    }
  ],
  "logging_steps": 10,
  "max_steps": 59,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}