{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9968652037617555,
  "eval_steps": 500,
  "global_step": 159,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 3.125e-08,
      "logits/chosen": -2.8596811294555664,
      "logits/rejected": -2.878772258758545,
      "logps/chosen": -377.73638916015625,
      "logps/rejected": -331.222412109375,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.06,
      "learning_rate": 3.1249999999999997e-07,
      "logits/chosen": -2.7957372665405273,
      "logits/rejected": -2.782663583755493,
      "logps/chosen": -276.7703857421875,
      "logps/rejected": -262.0076904296875,
      "loss": 0.6929,
      "rewards/accuracies": 0.5277777910232544,
      "rewards/chosen": 0.0007769857766106725,
      "rewards/margins": 0.0016400194726884365,
      "rewards/rejected": -0.0008630338124930859,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.990353313429303e-07,
      "logits/chosen": -2.8224599361419678,
      "logits/rejected": -2.819226026535034,
      "logps/chosen": -257.95306396484375,
      "logps/rejected": -246.1923065185547,
      "loss": 0.688,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": 0.005768328905105591,
      "rewards/margins": 0.012648087926208973,
      "rewards/rejected": -0.006879759021103382,
      "step": 20
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.882681251368548e-07,
      "logits/chosen": -2.817078113555908,
      "logits/rejected": -2.7680933475494385,
      "logps/chosen": -256.79595947265625,
      "logps/rejected": -242.519775390625,
      "loss": 0.6718,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": 0.023607144132256508,
      "rewards/margins": 0.046295683830976486,
      "rewards/rejected": -0.022688541561365128,
      "step": 30
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.6604720940421207e-07,
      "logits/chosen": -2.7658379077911377,
      "logits/rejected": -2.779132127761841,
      "logps/chosen": -303.56890869140625,
      "logps/rejected": -275.5705261230469,
      "loss": 0.6489,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.03137209266424179,
      "rewards/margins": 0.10878098011016846,
      "rewards/rejected": -0.14015306532382965,
      "step": 40
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.3344075855595097e-07,
      "logits/chosen": -2.7775702476501465,
      "logits/rejected": -2.7892866134643555,
      "logps/chosen": -298.8876953125,
      "logps/rejected": -303.22705078125,
      "loss": 0.6336,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.12522175908088684,
      "rewards/margins": 0.1828564703464508,
      "rewards/rejected": -0.30807822942733765,
      "step": 50
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.920161866827889e-07,
      "logits/chosen": -2.729459285736084,
      "logits/rejected": -2.708754539489746,
      "logps/chosen": -259.3099060058594,
      "logps/rejected": -277.4466857910156,
      "loss": 0.6141,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.30067360401153564,
      "rewards/margins": 0.1461033821105957,
      "rewards/rejected": -0.44677701592445374,
      "step": 60
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4376480090239047e-07,
      "logits/chosen": -2.742418050765991,
      "logits/rejected": -2.747220516204834,
      "logps/chosen": -293.5316467285156,
      "logps/rejected": -323.1183166503906,
      "loss": 0.5805,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.404310405254364,
      "rewards/margins": 0.3498535454273224,
      "rewards/rejected": -0.754163920879364,
      "step": 70
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.910060778827554e-07,
      "logits/chosen": -2.7769148349761963,
      "logits/rejected": -2.749772071838379,
      "logps/chosen": -318.1655578613281,
      "logps/rejected": -307.8402404785156,
      "loss": 0.602,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.4084464907646179,
      "rewards/margins": 0.3972111642360687,
      "rewards/rejected": -0.805657684803009,
      "step": 80
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.3627616503391812e-07,
      "logits/chosen": -2.724620819091797,
      "logits/rejected": -2.711826801300049,
      "logps/chosen": -345.8966064453125,
      "logps/rejected": -358.5138244628906,
      "loss": 0.5744,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.48604559898376465,
      "rewards/margins": 0.4906246066093445,
      "rewards/rejected": -0.9766701459884644,
      "step": 90
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8220596619089573e-07,
      "logits/chosen": -2.7404863834381104,
      "logits/rejected": -2.6992406845092773,
      "logps/chosen": -330.0885314941406,
      "logps/rejected": -324.7499084472656,
      "loss": 0.5834,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.4633980393409729,
      "rewards/margins": 0.3974135220050812,
      "rewards/rejected": -0.860811710357666,
      "step": 100
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.3139467229135998e-07,
      "logits/chosen": -2.694833755493164,
      "logits/rejected": -2.659644365310669,
      "logps/chosen": -270.2969970703125,
      "logps/rejected": -323.013916015625,
      "loss": 0.5582,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.448250949382782,
      "rewards/margins": 0.43962574005126953,
      "rewards/rejected": -0.8878766894340515,
      "step": 110
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.628481651367875e-08,
      "logits/chosen": -2.7171640396118164,
      "logits/rejected": -2.7176883220672607,
      "logps/chosen": -309.8388671875,
      "logps/rejected": -334.4871520996094,
      "loss": 0.5727,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.4930656850337982,
      "rewards/margins": 0.5099568963050842,
      "rewards/rejected": -1.00302255153656,
      "step": 120
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.904486005914027e-08,
      "logits/chosen": -2.6858134269714355,
      "logits/rejected": -2.667487859725952,
      "logps/chosen": -363.0616149902344,
      "logps/rejected": -362.6529846191406,
      "loss": 0.5814,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": -0.7948101162910461,
      "rewards/margins": 0.3632834851741791,
      "rewards/rejected": -1.1580935716629028,
      "step": 130
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.1464952759020856e-08,
      "logits/chosen": -2.694016695022583,
      "logits/rejected": -2.6449930667877197,
      "logps/chosen": -343.0268859863281,
      "logps/rejected": -364.78936767578125,
      "loss": 0.5468,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.7075488567352295,
      "rewards/margins": 0.5151935815811157,
      "rewards/rejected": -1.2227424383163452,
      "step": 140
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.8708793644441086e-09,
      "logits/chosen": -2.6975722312927246,
      "logits/rejected": -2.6778318881988525,
      "logps/chosen": -323.04046630859375,
      "logps/rejected": -371.57696533203125,
      "loss": 0.5554,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.5652332305908203,
      "rewards/margins": 0.48478445410728455,
      "rewards/rejected": -1.0500177145004272,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 159,
      "total_flos": 0.0,
      "train_loss": 0.6063382550605438,
      "train_runtime": 2648.5083,
      "train_samples_per_second": 7.694,
      "train_steps_per_second": 0.06
    }
  ],
  "logging_steps": 10,
  "max_steps": 159,
  "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
}