{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9984301412872841,
  "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": 0.24608471989631653,
      "logits/rejected": 0.3027251064777374,
      "logps/chosen": -328.4881286621094,
      "logps/pi_response": -187.7349853515625,
      "logps/ref_response": -187.7349853515625,
      "logps/rejected": -357.2791442871094,
      "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": 0.21224091947078705,
      "logits/rejected": 0.3317519724369049,
      "logps/chosen": -283.24053955078125,
      "logps/pi_response": -185.40020751953125,
      "logps/ref_response": -185.55345153808594,
      "logps/rejected": -298.3199462890625,
      "loss": 0.6931,
      "rewards/accuracies": 0.4583333432674408,
      "rewards/chosen": 0.0004613929195329547,
      "rewards/margins": 0.0005008090520277619,
      "rewards/rejected": -3.9416205254383385e-05,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.990353313429303e-07,
      "logits/chosen": 0.15066958963871002,
      "logits/rejected": 0.3227843642234802,
      "logps/chosen": -273.8597717285156,
      "logps/pi_response": -170.4561004638672,
      "logps/ref_response": -172.29898071289062,
      "logps/rejected": -288.9510498046875,
      "loss": 0.6878,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -0.0035666346084326506,
      "rewards/margins": 0.013525977730751038,
      "rewards/rejected": -0.017092611640691757,
      "step": 20
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.882681251368548e-07,
      "logits/chosen": 0.18726065754890442,
      "logits/rejected": 0.30163514614105225,
      "logps/chosen": -277.689453125,
      "logps/pi_response": -152.5356903076172,
      "logps/ref_response": -158.19248962402344,
      "logps/rejected": -310.6690979003906,
      "loss": 0.6673,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.022925913333892822,
      "rewards/margins": 0.057639896869659424,
      "rewards/rejected": -0.08056580275297165,
      "step": 30
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.6604720940421207e-07,
      "logits/chosen": 0.16672617197036743,
      "logits/rejected": 0.32053881883621216,
      "logps/chosen": -329.8152770996094,
      "logps/pi_response": -176.9584503173828,
      "logps/ref_response": -190.54837036132812,
      "logps/rejected": -333.39990234375,
      "loss": 0.6357,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.058331944048404694,
      "rewards/margins": 0.13682730495929718,
      "rewards/rejected": -0.19515924155712128,
      "step": 40
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.3344075855595097e-07,
      "logits/chosen": 0.2507833242416382,
      "logits/rejected": 0.41381892561912537,
      "logps/chosen": -286.53997802734375,
      "logps/pi_response": -155.46978759765625,
      "logps/ref_response": -172.08847045898438,
      "logps/rejected": -325.3523864746094,
      "loss": 0.6136,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.08929577469825745,
      "rewards/margins": 0.19610214233398438,
      "rewards/rejected": -0.2853979170322418,
      "step": 50
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.920161866827889e-07,
      "logits/chosen": 0.4524998664855957,
      "logits/rejected": 0.5539639592170715,
      "logps/chosen": -303.2442321777344,
      "logps/pi_response": -161.90357971191406,
      "logps/ref_response": -174.90908813476562,
      "logps/rejected": -359.1524658203125,
      "loss": 0.5928,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.1562139093875885,
      "rewards/margins": 0.29649364948272705,
      "rewards/rejected": -0.45270758867263794,
      "step": 60
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4376480090239047e-07,
      "logits/chosen": 0.5355798006057739,
      "logits/rejected": 0.6822081804275513,
      "logps/chosen": -266.3974609375,
      "logps/pi_response": -162.60568237304688,
      "logps/ref_response": -172.18724060058594,
      "logps/rejected": -373.76263427734375,
      "loss": 0.5701,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.18911810219287872,
      "rewards/margins": 0.4514671266078949,
      "rewards/rejected": -0.64058518409729,
      "step": 70
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.910060778827554e-07,
      "logits/chosen": 0.4139408469200134,
      "logits/rejected": 0.754637598991394,
      "logps/chosen": -344.97308349609375,
      "logps/pi_response": -186.01515197753906,
      "logps/ref_response": -194.66531372070312,
      "logps/rejected": -374.3211975097656,
      "loss": 0.5782,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.2587112784385681,
      "rewards/margins": 0.38729003071784973,
      "rewards/rejected": -0.6460012793540955,
      "step": 80
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.3627616503391812e-07,
      "logits/chosen": 0.4229448437690735,
      "logits/rejected": 0.7810887694358826,
      "logps/chosen": -333.3769836425781,
      "logps/pi_response": -182.8963165283203,
      "logps/ref_response": -185.94192504882812,
      "logps/rejected": -443.4452209472656,
      "loss": 0.551,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.24502289295196533,
      "rewards/margins": 0.5839612483978271,
      "rewards/rejected": -0.8289840817451477,
      "step": 90
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8220596619089573e-07,
      "logits/chosen": 0.5892740488052368,
      "logits/rejected": 0.810705304145813,
      "logps/chosen": -308.01953125,
      "logps/pi_response": -164.831298828125,
      "logps/ref_response": -163.5237274169922,
      "logps/rejected": -327.8468017578125,
      "loss": 0.5579,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.27581167221069336,
      "rewards/margins": 0.32737815380096436,
      "rewards/rejected": -0.6031898856163025,
      "step": 100
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.3139467229135998e-07,
      "logits/chosen": 0.5539606213569641,
      "logits/rejected": 0.8888662457466125,
      "logps/chosen": -314.2873840332031,
      "logps/pi_response": -189.80223083496094,
      "logps/ref_response": -190.3043212890625,
      "logps/rejected": -397.5984802246094,
      "loss": 0.5393,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.22623829543590546,
      "rewards/margins": 0.5061606764793396,
      "rewards/rejected": -0.732399046421051,
      "step": 110
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.628481651367875e-08,
      "logits/chosen": 0.540047287940979,
      "logits/rejected": 0.8741270303726196,
      "logps/chosen": -330.16851806640625,
      "logps/pi_response": -181.96485900878906,
      "logps/ref_response": -180.7099609375,
      "logps/rejected": -383.4455871582031,
      "loss": 0.5562,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.2805691361427307,
      "rewards/margins": 0.5478684306144714,
      "rewards/rejected": -0.8284376263618469,
      "step": 120
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.904486005914027e-08,
      "logits/chosen": 0.5884017944335938,
      "logits/rejected": 0.8417636156082153,
      "logps/chosen": -316.0371398925781,
      "logps/pi_response": -182.68063354492188,
      "logps/ref_response": -177.9791717529297,
      "logps/rejected": -384.3453674316406,
      "loss": 0.5297,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.2747536301612854,
      "rewards/margins": 0.5689390897750854,
      "rewards/rejected": -0.8436927795410156,
      "step": 130
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.1464952759020856e-08,
      "logits/chosen": 0.5350295305252075,
      "logits/rejected": 0.8135560750961304,
      "logps/chosen": -317.75628662109375,
      "logps/pi_response": -170.38433837890625,
      "logps/ref_response": -169.5862579345703,
      "logps/rejected": -415.2865295410156,
      "loss": 0.5428,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.2749920189380646,
      "rewards/margins": 0.7153186202049255,
      "rewards/rejected": -0.9903106689453125,
      "step": 140
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.8708793644441086e-09,
      "logits/chosen": 0.5133858919143677,
      "logits/rejected": 0.890362560749054,
      "logps/chosen": -354.5374755859375,
      "logps/pi_response": -193.73138427734375,
      "logps/ref_response": -188.74514770507812,
      "logps/rejected": -411.3604431152344,
      "loss": 0.5354,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.33739715814590454,
      "rewards/margins": 0.5403859615325928,
      "rewards/rejected": -0.8777831792831421,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 159,
      "total_flos": 0.0,
      "train_loss": 0.5850164965263702,
      "train_runtime": 4486.0111,
      "train_samples_per_second": 4.543,
      "train_steps_per_second": 0.035
    }
  ],
  "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": 4,
  "trial_name": null,
  "trial_params": null
}