{
  "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": -2.2583727836608887,
      "logits/rejected": -2.194669246673584,
      "logps/chosen": -344.6805114746094,
      "logps/rejected": -430.3193359375,
      "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": -2.2495357990264893,
      "logits/rejected": -2.220221996307373,
      "logps/chosen": -281.9364929199219,
      "logps/rejected": -305.81884765625,
      "loss": 0.689,
      "rewards/accuracies": 0.4652777910232544,
      "rewards/chosen": -0.01338761206716299,
      "rewards/margins": 0.006979748606681824,
      "rewards/rejected": -0.02036735974252224,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.990353313429303e-07,
      "logits/chosen": -2.1989874839782715,
      "logits/rejected": -2.167945146560669,
      "logps/chosen": -258.35382080078125,
      "logps/rejected": -364.7663269042969,
      "loss": 0.6641,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.2977336645126343,
      "rewards/margins": 0.21266262233257294,
      "rewards/rejected": -0.5103963613510132,
      "step": 20
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.882681251368548e-07,
      "logits/chosen": -2.0941319465637207,
      "logits/rejected": -2.063507080078125,
      "logps/chosen": -295.8477783203125,
      "logps/rejected": -372.314453125,
      "loss": 0.6929,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.3137749135494232,
      "rewards/margins": 0.2363557368516922,
      "rewards/rejected": -0.5501306056976318,
      "step": 30
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.6604720940421207e-07,
      "logits/chosen": -2.178748607635498,
      "logits/rejected": -2.1334617137908936,
      "logps/chosen": -285.15130615234375,
      "logps/rejected": -334.0563659667969,
      "loss": 0.6656,
      "rewards/accuracies": 0.543749988079071,
      "rewards/chosen": -0.13720116019248962,
      "rewards/margins": 0.07690997421741486,
      "rewards/rejected": -0.21411113440990448,
      "step": 40
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.3344075855595097e-07,
      "logits/chosen": -2.209764003753662,
      "logits/rejected": -2.192694664001465,
      "logps/chosen": -287.18182373046875,
      "logps/rejected": -383.46197509765625,
      "loss": 0.6532,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.21152925491333008,
      "rewards/margins": 0.1477278620004654,
      "rewards/rejected": -0.35925716161727905,
      "step": 50
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.920161866827889e-07,
      "logits/chosen": -2.0722219944000244,
      "logits/rejected": -2.036907196044922,
      "logps/chosen": -295.5084228515625,
      "logps/rejected": -374.8094787597656,
      "loss": 0.6568,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.31626835465431213,
      "rewards/margins": 0.2158782035112381,
      "rewards/rejected": -0.5321465730667114,
      "step": 60
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4376480090239047e-07,
      "logits/chosen": -2.0726826190948486,
      "logits/rejected": -2.032034397125244,
      "logps/chosen": -259.9215393066406,
      "logps/rejected": -369.0412292480469,
      "loss": 0.6325,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.19777385890483856,
      "rewards/margins": 0.27027979493141174,
      "rewards/rejected": -0.4680536389350891,
      "step": 70
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.910060778827554e-07,
      "logits/chosen": -2.107374906539917,
      "logits/rejected": -2.0704822540283203,
      "logps/chosen": -334.5880432128906,
      "logps/rejected": -371.534912109375,
      "loss": 0.6459,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.4631880223751068,
      "rewards/margins": 0.126481831073761,
      "rewards/rejected": -0.5896698236465454,
      "step": 80
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.3627616503391812e-07,
      "logits/chosen": -2.1149628162384033,
      "logits/rejected": -2.0265822410583496,
      "logps/chosen": -295.9393005371094,
      "logps/rejected": -402.9593200683594,
      "loss": 0.6123,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.31046614050865173,
      "rewards/margins": 0.31097108125686646,
      "rewards/rejected": -0.6214371919631958,
      "step": 90
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8220596619089573e-07,
      "logits/chosen": -2.065992593765259,
      "logits/rejected": -2.0325350761413574,
      "logps/chosen": -307.4365234375,
      "logps/rejected": -424.5880432128906,
      "loss": 0.6372,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -0.38495463132858276,
      "rewards/margins": 0.3768821358680725,
      "rewards/rejected": -0.7618367671966553,
      "step": 100
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.3139467229135998e-07,
      "logits/chosen": -2.0376338958740234,
      "logits/rejected": -2.006533145904541,
      "logps/chosen": -308.51556396484375,
      "logps/rejected": -427.515625,
      "loss": 0.645,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -0.45776548981666565,
      "rewards/margins": 0.3285262882709503,
      "rewards/rejected": -0.7862917184829712,
      "step": 110
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.628481651367875e-08,
      "logits/chosen": -2.0662786960601807,
      "logits/rejected": -2.0550084114074707,
      "logps/chosen": -303.14727783203125,
      "logps/rejected": -370.09716796875,
      "loss": 0.6298,
      "rewards/accuracies": 0.5687500238418579,
      "rewards/chosen": -0.4310707151889801,
      "rewards/margins": 0.14983342587947845,
      "rewards/rejected": -0.5809041261672974,
      "step": 120
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.904486005914027e-08,
      "logits/chosen": -2.055657148361206,
      "logits/rejected": -2.022127389907837,
      "logps/chosen": -311.9975891113281,
      "logps/rejected": -384.63360595703125,
      "loss": 0.6525,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": -0.4104999005794525,
      "rewards/margins": 0.20988547801971436,
      "rewards/rejected": -0.620385468006134,
      "step": 130
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.1464952759020856e-08,
      "logits/chosen": -2.042844533920288,
      "logits/rejected": -2.010148525238037,
      "logps/chosen": -300.5053405761719,
      "logps/rejected": -373.68572998046875,
      "loss": 0.6315,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": -0.37878933548927307,
      "rewards/margins": 0.2560186982154846,
      "rewards/rejected": -0.6348080039024353,
      "step": 140
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.8708793644441086e-09,
      "logits/chosen": -2.0472328662872314,
      "logits/rejected": -2.0771114826202393,
      "logps/chosen": -301.08148193359375,
      "logps/rejected": -429.6272888183594,
      "loss": 0.6246,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": -0.3863406181335449,
      "rewards/margins": 0.33363914489746094,
      "rewards/rejected": -0.7199797630310059,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 159,
      "total_flos": 0.0,
      "train_loss": 0.6471116527821283,
      "train_runtime": 2657.8372,
      "train_samples_per_second": 7.667,
      "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
}