{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 478,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 579.7744748341399,
      "learning_rate": 1.0416666666666666e-08,
      "logits/chosen": -2.408252239227295,
      "logits/rejected": -2.408294677734375,
      "logps/chosen": -208.4792022705078,
      "logps/rejected": -178.0951690673828,
      "loss": 0.69,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.02,
      "grad_norm": 535.1486260819535,
      "learning_rate": 1.0416666666666667e-07,
      "logits/chosen": -2.544614553451538,
      "logits/rejected": -2.538756847381592,
      "logps/chosen": -261.5119323730469,
      "logps/rejected": -166.40280151367188,
      "loss": 0.7009,
      "rewards/accuracies": 0.4652777910232544,
      "rewards/chosen": 0.008291814476251602,
      "rewards/margins": 0.011442840099334717,
      "rewards/rejected": -0.003151026088744402,
      "step": 10
    },
    {
      "epoch": 0.04,
      "grad_norm": 384.83945719674614,
      "learning_rate": 2.0833333333333333e-07,
      "logits/chosen": -2.5195038318634033,
      "logits/rejected": -2.5265002250671387,
      "logps/chosen": -252.65768432617188,
      "logps/rejected": -178.087158203125,
      "loss": 0.5876,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.10371126234531403,
      "rewards/margins": 0.296107679605484,
      "rewards/rejected": -0.1923964023590088,
      "step": 20
    },
    {
      "epoch": 0.06,
      "grad_norm": 202.13874393663576,
      "learning_rate": 3.1249999999999997e-07,
      "logits/chosen": -2.4858596324920654,
      "logits/rejected": -2.47269606590271,
      "logps/chosen": -240.64242553710938,
      "logps/rejected": -181.0841522216797,
      "loss": 0.3862,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": 0.3985957205295563,
      "rewards/margins": 1.4220483303070068,
      "rewards/rejected": -1.023452639579773,
      "step": 30
    },
    {
      "epoch": 0.08,
      "grad_norm": 142.19955934070376,
      "learning_rate": 4.1666666666666667e-07,
      "logits/chosen": -2.4594264030456543,
      "logits/rejected": -2.4308407306671143,
      "logps/chosen": -283.8061828613281,
      "logps/rejected": -208.5924835205078,
      "loss": 0.3332,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": 1.9889726638793945,
      "rewards/margins": 3.029081344604492,
      "rewards/rejected": -1.0401084423065186,
      "step": 40
    },
    {
      "epoch": 0.1,
      "grad_norm": 713.6549875009309,
      "learning_rate": 4.999733114418725e-07,
      "logits/chosen": -2.442108631134033,
      "logits/rejected": -2.46040678024292,
      "logps/chosen": -246.4988555908203,
      "logps/rejected": -187.14138793945312,
      "loss": 0.4454,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": 2.2771620750427246,
      "rewards/margins": 4.51132345199585,
      "rewards/rejected": -2.234161853790283,
      "step": 50
    },
    {
      "epoch": 0.13,
      "grad_norm": 206.28301377298783,
      "learning_rate": 4.990398100856366e-07,
      "logits/chosen": -2.4876925945281982,
      "logits/rejected": -2.449016571044922,
      "logps/chosen": -271.1671142578125,
      "logps/rejected": -196.48915100097656,
      "loss": 0.282,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 1.714887261390686,
      "rewards/margins": 5.19891881942749,
      "rewards/rejected": -3.4840316772460938,
      "step": 60
    },
    {
      "epoch": 0.15,
      "grad_norm": 133.99852927659347,
      "learning_rate": 4.967775735898179e-07,
      "logits/chosen": -2.4287192821502686,
      "logits/rejected": -2.4621310234069824,
      "logps/chosen": -274.3385925292969,
      "logps/rejected": -198.64663696289062,
      "loss": 0.3118,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": 3.02872371673584,
      "rewards/margins": 6.160218238830566,
      "rewards/rejected": -3.1314942836761475,
      "step": 70
    },
    {
      "epoch": 0.17,
      "grad_norm": 274.89332519675594,
      "learning_rate": 4.931986719649298e-07,
      "logits/chosen": -2.42030668258667,
      "logits/rejected": -2.423424243927002,
      "logps/chosen": -263.48272705078125,
      "logps/rejected": -186.29991149902344,
      "loss": 0.3017,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 3.6749210357666016,
      "rewards/margins": 6.233966827392578,
      "rewards/rejected": -2.559046506881714,
      "step": 80
    },
    {
      "epoch": 0.19,
      "grad_norm": 202.07196994453352,
      "learning_rate": 4.883222001996351e-07,
      "logits/chosen": -2.4479732513427734,
      "logits/rejected": -2.4088387489318848,
      "logps/chosen": -267.5174255371094,
      "logps/rejected": -193.828125,
      "loss": 0.3603,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 3.906665802001953,
      "rewards/margins": 7.1938323974609375,
      "rewards/rejected": -3.2871665954589844,
      "step": 90
    },
    {
      "epoch": 0.21,
      "grad_norm": 260.799637557766,
      "learning_rate": 4.821741763807186e-07,
      "logits/chosen": -2.4610652923583984,
      "logits/rejected": -2.454207181930542,
      "logps/chosen": -231.1355743408203,
      "logps/rejected": -183.55123901367188,
      "loss": 0.3315,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.3220086097717285,
      "rewards/margins": 5.911337852478027,
      "rewards/rejected": -3.589329242706299,
      "step": 100
    },
    {
      "epoch": 0.21,
      "eval_logits/chosen": -2.5300614833831787,
      "eval_logits/rejected": -2.515779495239258,
      "eval_logps/chosen": -250.55865478515625,
      "eval_logps/rejected": -187.6100616455078,
      "eval_loss": 0.297485888004303,
      "eval_rewards/accuracies": 0.90234375,
      "eval_rewards/chosen": 3.437309980392456,
      "eval_rewards/margins": 7.265252113342285,
      "eval_rewards/rejected": -3.82794189453125,
      "eval_runtime": 97.5532,
      "eval_samples_per_second": 20.502,
      "eval_steps_per_second": 0.328,
      "step": 100
    },
    {
      "epoch": 0.23,
      "grad_norm": 164.61060726491735,
      "learning_rate": 4.747874028753375e-07,
      "logits/chosen": -2.36721134185791,
      "logits/rejected": -2.398892879486084,
      "logps/chosen": -238.1865692138672,
      "logps/rejected": -193.205810546875,
      "loss": 0.3602,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 2.924015522003174,
      "rewards/margins": 6.160517692565918,
      "rewards/rejected": -3.2365028858184814,
      "step": 110
    },
    {
      "epoch": 0.25,
      "grad_norm": 394.19842542709944,
      "learning_rate": 4.662012913161997e-07,
      "logits/chosen": -2.4796204566955566,
      "logits/rejected": -2.462428331375122,
      "logps/chosen": -237.8124237060547,
      "logps/rejected": -191.6023712158203,
      "loss": 0.2323,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": 3.661634922027588,
      "rewards/margins": 6.902056694030762,
      "rewards/rejected": -3.240421772003174,
      "step": 120
    },
    {
      "epoch": 0.27,
      "grad_norm": 409.3218341794219,
      "learning_rate": 4.5646165232345103e-07,
      "logits/chosen": -2.4416961669921875,
      "logits/rejected": -2.473330497741699,
      "logps/chosen": -259.2460632324219,
      "logps/rejected": -185.64683532714844,
      "loss": 0.3435,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 4.6567606925964355,
      "rewards/margins": 7.721456050872803,
      "rewards/rejected": -3.064694881439209,
      "step": 130
    },
    {
      "epoch": 0.29,
      "grad_norm": 328.4389039905761,
      "learning_rate": 4.456204510851956e-07,
      "logits/chosen": -2.326636791229248,
      "logits/rejected": -2.3331127166748047,
      "logps/chosen": -261.88006591796875,
      "logps/rejected": -197.84036254882812,
      "loss": 0.2942,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 3.5752804279327393,
      "rewards/margins": 6.8619065284729,
      "rewards/rejected": -3.2866263389587402,
      "step": 140
    },
    {
      "epoch": 0.31,
      "grad_norm": 664.6717326413732,
      "learning_rate": 4.337355301007335e-07,
      "logits/chosen": -2.4599616527557373,
      "logits/rejected": -2.4934587478637695,
      "logps/chosen": -239.1993865966797,
      "logps/rejected": -179.16249084472656,
      "loss": 0.4201,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 3.849210262298584,
      "rewards/margins": 7.315499782562256,
      "rewards/rejected": -3.46628999710083,
      "step": 150
    },
    {
      "epoch": 0.33,
      "grad_norm": 389.31781863733255,
      "learning_rate": 4.2087030056579986e-07,
      "logits/chosen": -2.4196219444274902,
      "logits/rejected": -2.3700897693634033,
      "logps/chosen": -260.79736328125,
      "logps/rejected": -198.62042236328125,
      "loss": 0.4699,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": 2.786912441253662,
      "rewards/margins": 6.555578708648682,
      "rewards/rejected": -3.768665313720703,
      "step": 160
    },
    {
      "epoch": 0.36,
      "grad_norm": 450.98983026990686,
      "learning_rate": 4.070934040463998e-07,
      "logits/chosen": -2.5323286056518555,
      "logits/rejected": -2.534219741821289,
      "logps/chosen": -251.9270782470703,
      "logps/rejected": -188.5828094482422,
      "loss": 0.3003,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 2.2955946922302246,
      "rewards/margins": 8.251626968383789,
      "rewards/rejected": -5.956032752990723,
      "step": 170
    },
    {
      "epoch": 0.38,
      "grad_norm": 222.8025881040315,
      "learning_rate": 3.9247834624635404e-07,
      "logits/chosen": -2.4943268299102783,
      "logits/rejected": -2.4797425270080566,
      "logps/chosen": -261.6493835449219,
      "logps/rejected": -186.13827514648438,
      "loss": 0.3893,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 3.1764323711395264,
      "rewards/margins": 8.824117660522461,
      "rewards/rejected": -5.6476850509643555,
      "step": 180
    },
    {
      "epoch": 0.4,
      "grad_norm": 170.38903025294675,
      "learning_rate": 3.7710310482256523e-07,
      "logits/chosen": -2.4573581218719482,
      "logits/rejected": -2.4471302032470703,
      "logps/chosen": -242.3362274169922,
      "logps/rejected": -206.1191864013672,
      "loss": 0.2873,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 2.354017496109009,
      "rewards/margins": 8.386995315551758,
      "rewards/rejected": -6.0329766273498535,
      "step": 190
    },
    {
      "epoch": 0.42,
      "grad_norm": 233.638489796621,
      "learning_rate": 3.610497133404795e-07,
      "logits/chosen": -2.4837393760681152,
      "logits/rejected": -2.4798505306243896,
      "logps/chosen": -252.32040405273438,
      "logps/rejected": -173.71713256835938,
      "loss": 0.2909,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 3.7421042919158936,
      "rewards/margins": 8.136415481567383,
      "rewards/rejected": -4.394310474395752,
      "step": 200
    },
    {
      "epoch": 0.42,
      "eval_logits/chosen": -2.5459139347076416,
      "eval_logits/rejected": -2.5309653282165527,
      "eval_logps/chosen": -247.70968627929688,
      "eval_logps/rejected": -188.19419860839844,
      "eval_loss": 0.2753676474094391,
      "eval_rewards/accuracies": 0.91796875,
      "eval_rewards/chosen": 4.861801624298096,
      "eval_rewards/margins": 8.981813430786133,
      "eval_rewards/rejected": -4.120011806488037,
      "eval_runtime": 97.2888,
      "eval_samples_per_second": 20.557,
      "eval_steps_per_second": 0.329,
      "step": 200
    },
    {
      "epoch": 0.44,
      "grad_norm": 337.30781066112155,
      "learning_rate": 3.4440382358952115e-07,
      "logits/chosen": -2.4054043292999268,
      "logits/rejected": -2.334240436553955,
      "logps/chosen": -229.5664825439453,
      "logps/rejected": -177.37989807128906,
      "loss": 0.3982,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 3.1289913654327393,
      "rewards/margins": 6.229460716247559,
      "rewards/rejected": -3.100468873977661,
      "step": 210
    },
    {
      "epoch": 0.46,
      "grad_norm": 196.23824528381684,
      "learning_rate": 3.272542485937368e-07,
      "logits/chosen": -2.4668684005737305,
      "logits/rejected": -2.4853501319885254,
      "logps/chosen": -247.2740478515625,
      "logps/rejected": -182.8220977783203,
      "loss": 0.564,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 4.549849510192871,
      "rewards/margins": 9.18285083770752,
      "rewards/rejected": -4.633000373840332,
      "step": 220
    },
    {
      "epoch": 0.48,
      "grad_norm": 188.59878410714853,
      "learning_rate": 3.096924887558854e-07,
      "logits/chosen": -2.4467105865478516,
      "logits/rejected": -2.394397735595703,
      "logps/chosen": -269.22100830078125,
      "logps/rejected": -207.54287719726562,
      "loss": 0.61,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 3.9587960243225098,
      "rewards/margins": 9.096400260925293,
      "rewards/rejected": -5.137604713439941,
      "step": 230
    },
    {
      "epoch": 0.5,
      "grad_norm": 248.56418253631352,
      "learning_rate": 2.9181224366319943e-07,
      "logits/chosen": -2.438028335571289,
      "logits/rejected": -2.4430088996887207,
      "logps/chosen": -228.5868377685547,
      "logps/rejected": -184.39010620117188,
      "loss": 0.2929,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 2.860379934310913,
      "rewards/margins": 8.014602661132812,
      "rewards/rejected": -5.15422248840332,
      "step": 240
    },
    {
      "epoch": 0.52,
      "grad_norm": 236.10656708079867,
      "learning_rate": 2.7370891215954565e-07,
      "logits/chosen": -2.468085527420044,
      "logits/rejected": -2.471389055252075,
      "logps/chosen": -245.74221801757812,
      "logps/rejected": -189.75096130371094,
      "loss": 0.2793,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 4.629996299743652,
      "rewards/margins": 8.119891166687012,
      "rewards/rejected": -3.489894151687622,
      "step": 250
    },
    {
      "epoch": 0.54,
      "grad_norm": 536.510359732396,
      "learning_rate": 2.55479083351317e-07,
      "logits/chosen": -2.4861550331115723,
      "logits/rejected": -2.4899330139160156,
      "logps/chosen": -254.806396484375,
      "logps/rejected": -172.7798614501953,
      "loss": 0.343,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 4.790152549743652,
      "rewards/margins": 9.49361515045166,
      "rewards/rejected": -4.703463077545166,
      "step": 260
    },
    {
      "epoch": 0.56,
      "grad_norm": 203.6794572887938,
      "learning_rate": 2.3722002126275822e-07,
      "logits/chosen": -2.456101179122925,
      "logits/rejected": -2.407947063446045,
      "logps/chosen": -236.4749755859375,
      "logps/rejected": -187.09112548828125,
      "loss": 0.2939,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": 3.9466958045959473,
      "rewards/margins": 8.208832740783691,
      "rewards/rejected": -4.262135982513428,
      "step": 270
    },
    {
      "epoch": 0.59,
      "grad_norm": 394.1475095150393,
      "learning_rate": 2.19029145890313e-07,
      "logits/chosen": -2.343445301055908,
      "logits/rejected": -2.3995354175567627,
      "logps/chosen": -257.370361328125,
      "logps/rejected": -201.5979461669922,
      "loss": 0.3501,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 3.717083692550659,
      "rewards/margins": 8.61104965209961,
      "rewards/rejected": -4.893965721130371,
      "step": 280
    },
    {
      "epoch": 0.61,
      "grad_norm": 206.90087071186883,
      "learning_rate": 2.0100351342479216e-07,
      "logits/chosen": -2.4418747425079346,
      "logits/rejected": -2.428316593170166,
      "logps/chosen": -234.2547607421875,
      "logps/rejected": -182.3361053466797,
      "loss": 0.3955,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": 3.4869391918182373,
      "rewards/margins": 8.724508285522461,
      "rewards/rejected": -5.2375688552856445,
      "step": 290
    },
    {
      "epoch": 0.63,
      "grad_norm": 232.39029956497873,
      "learning_rate": 1.8323929841460178e-07,
      "logits/chosen": -2.434248208999634,
      "logits/rejected": -2.357980251312256,
      "logps/chosen": -265.85308837890625,
      "logps/rejected": -211.96859741210938,
      "loss": 0.6445,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 3.5436806678771973,
      "rewards/margins": 8.844223976135254,
      "rewards/rejected": -5.300544738769531,
      "step": 300
    },
    {
      "epoch": 0.63,
      "eval_logits/chosen": -2.5059330463409424,
      "eval_logits/rejected": -2.491495132446289,
      "eval_logps/chosen": -249.0003662109375,
      "eval_logps/rejected": -190.69680786132812,
      "eval_loss": 0.22447091341018677,
      "eval_rewards/accuracies": 0.91015625,
      "eval_rewards/chosen": 4.216455936431885,
      "eval_rewards/margins": 9.587770462036133,
      "eval_rewards/rejected": -5.371314525604248,
      "eval_runtime": 97.2649,
      "eval_samples_per_second": 20.562,
      "eval_steps_per_second": 0.329,
      "step": 300
    },
    {
      "epoch": 0.65,
      "grad_norm": 216.5700858394941,
      "learning_rate": 1.6583128063291573e-07,
      "logits/chosen": -2.4202470779418945,
      "logits/rejected": -2.42097806930542,
      "logps/chosen": -255.98904418945312,
      "logps/rejected": -196.6630401611328,
      "loss": 0.3789,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 3.9936203956604004,
      "rewards/margins": 8.884687423706055,
      "rewards/rejected": -4.891066551208496,
      "step": 310
    },
    {
      "epoch": 0.67,
      "grad_norm": 232.75719945601446,
      "learning_rate": 1.488723393865766e-07,
      "logits/chosen": -2.493706464767456,
      "logits/rejected": -2.461465358734131,
      "logps/chosen": -238.67599487304688,
      "logps/rejected": -194.2203826904297,
      "loss": 0.3515,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": 4.206465244293213,
      "rewards/margins": 9.311426162719727,
      "rewards/rejected": -5.104961395263672,
      "step": 320
    },
    {
      "epoch": 0.69,
      "grad_norm": 281.239048485679,
      "learning_rate": 1.3245295796480788e-07,
      "logits/chosen": -2.4040043354034424,
      "logits/rejected": -2.4380762577056885,
      "logps/chosen": -247.7274169921875,
      "logps/rejected": -179.15911865234375,
      "loss": 0.4043,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 5.208529472351074,
      "rewards/margins": 9.102964401245117,
      "rewards/rejected": -3.8944339752197266,
      "step": 330
    },
    {
      "epoch": 0.71,
      "grad_norm": 183.00963279480595,
      "learning_rate": 1.1666074087171627e-07,
      "logits/chosen": -2.3740360736846924,
      "logits/rejected": -2.4260330200195312,
      "logps/chosen": -235.00357055664062,
      "logps/rejected": -166.9297637939453,
      "loss": 0.2751,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": 4.350711822509766,
      "rewards/margins": 7.54253625869751,
      "rewards/rejected": -3.1918249130249023,
      "step": 340
    },
    {
      "epoch": 0.73,
      "grad_norm": 258.2328231454923,
      "learning_rate": 1.0157994641835734e-07,
      "logits/chosen": -2.4078361988067627,
      "logits/rejected": -2.4156928062438965,
      "logps/chosen": -234.7695770263672,
      "logps/rejected": -214.1270294189453,
      "loss": 0.3005,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 4.440161228179932,
      "rewards/margins": 9.19445514678955,
      "rewards/rejected": -4.754293918609619,
      "step": 350
    },
    {
      "epoch": 0.75,
      "grad_norm": 260.2428729882813,
      "learning_rate": 8.729103716819111e-08,
      "logits/chosen": -2.372528076171875,
      "logits/rejected": -2.3916471004486084,
      "logps/chosen": -241.23934936523438,
      "logps/rejected": -187.92953491210938,
      "loss": 0.2667,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 4.416121482849121,
      "rewards/margins": 8.337173461914062,
      "rewards/rejected": -3.921051502227783,
      "step": 360
    },
    {
      "epoch": 0.77,
      "grad_norm": 316.09399625633,
      "learning_rate": 7.387025063449081e-08,
      "logits/chosen": -2.4834017753601074,
      "logits/rejected": -2.428774118423462,
      "logps/chosen": -223.4984893798828,
      "logps/rejected": -176.99790954589844,
      "loss": 0.3594,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 3.544004440307617,
      "rewards/margins": 8.096233367919922,
      "rewards/rejected": -4.552228927612305,
      "step": 370
    },
    {
      "epoch": 0.79,
      "grad_norm": 229.93737644129538,
      "learning_rate": 6.138919252022435e-08,
      "logits/chosen": -2.4690134525299072,
      "logits/rejected": -2.4799323081970215,
      "logps/chosen": -249.5747528076172,
      "logps/rejected": -198.73275756835938,
      "loss": 0.3121,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 3.8754029273986816,
      "rewards/margins": 9.315340042114258,
      "rewards/rejected": -5.439937591552734,
      "step": 380
    },
    {
      "epoch": 0.82,
      "grad_norm": 178.09073107787978,
      "learning_rate": 4.991445467064689e-08,
      "logits/chosen": -2.463972568511963,
      "logits/rejected": -2.4395322799682617,
      "logps/chosen": -239.1776123046875,
      "logps/rejected": -189.71803283691406,
      "loss": 0.222,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 3.557317018508911,
      "rewards/margins": 8.469549179077148,
      "rewards/rejected": -4.912230968475342,
      "step": 390
    },
    {
      "epoch": 0.84,
      "grad_norm": 211.56309254204024,
      "learning_rate": 3.9507259776993954e-08,
      "logits/chosen": -2.3866400718688965,
      "logits/rejected": -2.4389476776123047,
      "logps/chosen": -239.34860229492188,
      "logps/rejected": -181.77676391601562,
      "loss": 0.2653,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 3.949272632598877,
      "rewards/margins": 8.648920059204102,
      "rewards/rejected": -4.699648857116699,
      "step": 400
    },
    {
      "epoch": 0.84,
      "eval_logits/chosen": -2.512333631515503,
      "eval_logits/rejected": -2.4965250492095947,
      "eval_logps/chosen": -248.40835571289062,
      "eval_logps/rejected": -190.5157470703125,
      "eval_loss": 0.21030600368976593,
      "eval_rewards/accuracies": 0.92578125,
      "eval_rewards/chosen": 4.512471675872803,
      "eval_rewards/margins": 9.793259620666504,
      "eval_rewards/rejected": -5.280787944793701,
      "eval_runtime": 97.2069,
      "eval_samples_per_second": 20.575,
      "eval_steps_per_second": 0.329,
      "step": 400
    },
    {
      "epoch": 0.86,
      "grad_norm": 337.4045735200527,
      "learning_rate": 3.022313472693447e-08,
      "logits/chosen": -2.431990146636963,
      "logits/rejected": -2.4296875,
      "logps/chosen": -269.318603515625,
      "logps/rejected": -194.96946716308594,
      "loss": 0.6111,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 3.5342013835906982,
      "rewards/margins": 8.450251579284668,
      "rewards/rejected": -4.916050910949707,
      "step": 410
    },
    {
      "epoch": 0.88,
      "grad_norm": 403.24647604289254,
      "learning_rate": 2.2111614344599684e-08,
      "logits/chosen": -2.39760684967041,
      "logits/rejected": -2.376661777496338,
      "logps/chosen": -237.7588653564453,
      "logps/rejected": -190.60279846191406,
      "loss": 0.3801,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 3.0914576053619385,
      "rewards/margins": 7.803999423980713,
      "rewards/rejected": -4.712540626525879,
      "step": 420
    },
    {
      "epoch": 0.9,
      "grad_norm": 208.90618387532385,
      "learning_rate": 1.521597710086439e-08,
      "logits/chosen": -2.4121103286743164,
      "logits/rejected": -2.3859496116638184,
      "logps/chosen": -263.43292236328125,
      "logps/rejected": -191.71656799316406,
      "loss": 0.3449,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 3.8668575286865234,
      "rewards/margins": 8.32303524017334,
      "rewards/rejected": -4.456177234649658,
      "step": 430
    },
    {
      "epoch": 0.92,
      "grad_norm": 221.6697110810539,
      "learning_rate": 9.57301420397924e-09,
      "logits/chosen": -2.5127618312835693,
      "logits/rejected": -2.4956202507019043,
      "logps/chosen": -254.6200408935547,
      "logps/rejected": -190.78817749023438,
      "loss": 0.3624,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 4.447186470031738,
      "rewards/margins": 9.914505004882812,
      "rewards/rejected": -5.467317581176758,
      "step": 440
    },
    {
      "epoch": 0.94,
      "grad_norm": 304.27875154546507,
      "learning_rate": 5.212833302556258e-09,
      "logits/chosen": -2.5326547622680664,
      "logits/rejected": -2.4754910469055176,
      "logps/chosen": -242.94784545898438,
      "logps/rejected": -187.10147094726562,
      "loss": 0.2887,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 4.133586406707764,
      "rewards/margins": 9.706035614013672,
      "rewards/rejected": -5.572449207305908,
      "step": 450
    },
    {
      "epoch": 0.96,
      "grad_norm": 480.62571484354635,
      "learning_rate": 2.158697848236607e-09,
      "logits/chosen": -2.419916868209839,
      "logits/rejected": -2.4316189289093018,
      "logps/chosen": -255.69723510742188,
      "logps/rejected": -203.59130859375,
      "loss": 0.4414,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 3.5629382133483887,
      "rewards/margins": 8.139005661010742,
      "rewards/rejected": -4.576067924499512,
      "step": 460
    },
    {
      "epoch": 0.98,
      "grad_norm": 258.3454560672873,
      "learning_rate": 4.269029751107489e-10,
      "logits/chosen": -2.416884660720825,
      "logits/rejected": -2.3896336555480957,
      "logps/chosen": -230.7972412109375,
      "logps/rejected": -200.25830078125,
      "loss": 0.2781,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 3.390946865081787,
      "rewards/margins": 9.046823501586914,
      "rewards/rejected": -5.655877113342285,
      "step": 470
    },
    {
      "epoch": 1.0,
      "step": 478,
      "total_flos": 0.0,
      "train_loss": 0.4018711235732713,
      "train_runtime": 7633.33,
      "train_samples_per_second": 8.009,
      "train_steps_per_second": 0.063
    }
  ],
  "logging_steps": 10,
  "max_steps": 478,
  "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
}