{
  "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": 1316.2617480695828,
      "learning_rate": 1.0416666666666666e-08,
      "logits/chosen": -1.689455509185791,
      "logits/rejected": -1.4794573783874512,
      "logps/chosen": -126.21005249023438,
      "logps/rejected": -98.13133239746094,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.02,
      "grad_norm": 1084.7724692148897,
      "learning_rate": 1.0416666666666667e-07,
      "logits/chosen": -1.707624912261963,
      "logits/rejected": -1.6101186275482178,
      "logps/chosen": -139.66224670410156,
      "logps/rejected": -91.32621002197266,
      "loss": 0.686,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.01818913221359253,
      "rewards/margins": 0.027222516015172005,
      "rewards/rejected": -0.009033381938934326,
      "step": 10
    },
    {
      "epoch": 0.04,
      "grad_norm": 372.5890585979663,
      "learning_rate": 2.0833333333333333e-07,
      "logits/chosen": -1.6384038925170898,
      "logits/rejected": -1.6487312316894531,
      "logps/chosen": -130.37515258789062,
      "logps/rejected": -93.99095153808594,
      "loss": 0.4495,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 0.6898423433303833,
      "rewards/margins": 0.9020320177078247,
      "rewards/rejected": -0.21218962967395782,
      "step": 20
    },
    {
      "epoch": 0.06,
      "grad_norm": 363.4016752055454,
      "learning_rate": 3.1249999999999997e-07,
      "logits/chosen": -1.6896642446517944,
      "logits/rejected": -1.6273959875106812,
      "logps/chosen": -130.80935668945312,
      "logps/rejected": -106.37054443359375,
      "loss": 0.2556,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.216484785079956,
      "rewards/margins": 3.602745771408081,
      "rewards/rejected": -1.386260986328125,
      "step": 30
    },
    {
      "epoch": 0.08,
      "grad_norm": 254.07955635631413,
      "learning_rate": 4.1666666666666667e-07,
      "logits/chosen": -1.6251739263534546,
      "logits/rejected": -1.5512189865112305,
      "logps/chosen": -142.7510528564453,
      "logps/rejected": -113.31219482421875,
      "loss": 0.1925,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.6909842491149902,
      "rewards/margins": 7.113295078277588,
      "rewards/rejected": -4.422310829162598,
      "step": 40
    },
    {
      "epoch": 0.1,
      "grad_norm": 450.1083647442556,
      "learning_rate": 4.999733114418725e-07,
      "logits/chosen": -1.5827839374542236,
      "logits/rejected": -1.6095731258392334,
      "logps/chosen": -127.47169494628906,
      "logps/rejected": -124.7472152709961,
      "loss": 0.187,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": 2.2808213233947754,
      "rewards/margins": 9.9701509475708,
      "rewards/rejected": -7.6893310546875,
      "step": 50
    },
    {
      "epoch": 0.13,
      "grad_norm": 256.12553340364644,
      "learning_rate": 4.990398100856366e-07,
      "logits/chosen": -1.6854372024536133,
      "logits/rejected": -1.6347172260284424,
      "logps/chosen": -143.6377410888672,
      "logps/rejected": -138.22506713867188,
      "loss": 0.1745,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.208740711212158,
      "rewards/margins": 13.148699760437012,
      "rewards/rejected": -10.939959526062012,
      "step": 60
    },
    {
      "epoch": 0.15,
      "grad_norm": 266.6931182384945,
      "learning_rate": 4.967775735898179e-07,
      "logits/chosen": -1.6533622741699219,
      "logits/rejected": -1.6828343868255615,
      "logps/chosen": -136.77774047851562,
      "logps/rejected": -139.76748657226562,
      "loss": 0.1719,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.496314525604248,
      "rewards/margins": 15.036949157714844,
      "rewards/rejected": -12.540634155273438,
      "step": 70
    },
    {
      "epoch": 0.17,
      "grad_norm": 311.4093156505733,
      "learning_rate": 4.931986719649298e-07,
      "logits/chosen": -1.6512008905410767,
      "logits/rejected": -1.642218828201294,
      "logps/chosen": -129.7277374267578,
      "logps/rejected": -130.1470489501953,
      "loss": 0.1638,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 3.2196338176727295,
      "rewards/margins": 14.145663261413574,
      "rewards/rejected": -10.926031112670898,
      "step": 80
    },
    {
      "epoch": 0.19,
      "grad_norm": 353.08827705610673,
      "learning_rate": 4.883222001996351e-07,
      "logits/chosen": -1.7057621479034424,
      "logits/rejected": -1.7253615856170654,
      "logps/chosen": -136.57843017578125,
      "logps/rejected": -141.9429931640625,
      "loss": 0.1501,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 3.7895398139953613,
      "rewards/margins": 15.804595947265625,
      "rewards/rejected": -12.015056610107422,
      "step": 90
    },
    {
      "epoch": 0.21,
      "grad_norm": 173.53718581228586,
      "learning_rate": 4.821741763807186e-07,
      "logits/chosen": -1.7514231204986572,
      "logits/rejected": -1.7408148050308228,
      "logps/chosen": -116.7274169921875,
      "logps/rejected": -125.6600570678711,
      "loss": 0.1496,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 3.375108242034912,
      "rewards/margins": 13.876774787902832,
      "rewards/rejected": -10.501666069030762,
      "step": 100
    },
    {
      "epoch": 0.21,
      "eval_logits/chosen": -1.8009788990020752,
      "eval_logits/rejected": -1.790999174118042,
      "eval_logps/chosen": -123.99901580810547,
      "eval_logps/rejected": -130.8439178466797,
      "eval_loss": 0.13556738197803497,
      "eval_rewards/accuracies": 0.94140625,
      "eval_rewards/chosen": 4.131972789764404,
      "eval_rewards/margins": 15.41292953491211,
      "eval_rewards/rejected": -11.280956268310547,
      "eval_runtime": 97.6442,
      "eval_samples_per_second": 20.483,
      "eval_steps_per_second": 0.328,
      "step": 100
    },
    {
      "epoch": 0.23,
      "grad_norm": 220.3790872760113,
      "learning_rate": 4.747874028753375e-07,
      "logits/chosen": -1.6608690023422241,
      "logits/rejected": -1.7199954986572266,
      "logps/chosen": -120.6917724609375,
      "logps/rejected": -133.25762939453125,
      "loss": 0.1546,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 3.4064812660217285,
      "rewards/margins": 12.731483459472656,
      "rewards/rejected": -9.325002670288086,
      "step": 110
    },
    {
      "epoch": 0.25,
      "grad_norm": 179.12119430014053,
      "learning_rate": 4.662012913161997e-07,
      "logits/chosen": -1.7208821773529053,
      "logits/rejected": -1.7148889303207397,
      "logps/chosen": -118.548583984375,
      "logps/rejected": -133.46463012695312,
      "loss": 0.1456,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 4.286909580230713,
      "rewards/margins": 13.92347240447998,
      "rewards/rejected": -9.636563301086426,
      "step": 120
    },
    {
      "epoch": 0.27,
      "grad_norm": 119.5811123757762,
      "learning_rate": 4.5646165232345103e-07,
      "logits/chosen": -1.688997507095337,
      "logits/rejected": -1.7153244018554688,
      "logps/chosen": -124.4625244140625,
      "logps/rejected": -129.4587860107422,
      "loss": 0.1477,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 4.8756022453308105,
      "rewards/margins": 15.727206230163574,
      "rewards/rejected": -10.851605415344238,
      "step": 130
    },
    {
      "epoch": 0.29,
      "grad_norm": 293.67492501764275,
      "learning_rate": 4.456204510851956e-07,
      "logits/chosen": -1.6008217334747314,
      "logits/rejected": -1.5764399766921997,
      "logps/chosen": -121.84040832519531,
      "logps/rejected": -125.70499420166016,
      "loss": 0.1534,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 4.718628406524658,
      "rewards/margins": 14.510149955749512,
      "rewards/rejected": -9.791521072387695,
      "step": 140
    },
    {
      "epoch": 0.31,
      "grad_norm": 250.47040288737202,
      "learning_rate": 4.337355301007335e-07,
      "logits/chosen": -1.7394781112670898,
      "logits/rejected": -1.7636123895645142,
      "logps/chosen": -119.86863708496094,
      "logps/rejected": -122.49459075927734,
      "loss": 0.1403,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 4.755049705505371,
      "rewards/margins": 14.378946304321289,
      "rewards/rejected": -9.623896598815918,
      "step": 150
    },
    {
      "epoch": 0.33,
      "grad_norm": 394.54185792168863,
      "learning_rate": 4.2087030056579986e-07,
      "logits/chosen": -1.6095482110977173,
      "logits/rejected": -1.5893223285675049,
      "logps/chosen": -127.0114517211914,
      "logps/rejected": -128.70870971679688,
      "loss": 0.1577,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 4.486311435699463,
      "rewards/margins": 13.415287971496582,
      "rewards/rejected": -8.928976058959961,
      "step": 160
    },
    {
      "epoch": 0.36,
      "grad_norm": 336.28954028277553,
      "learning_rate": 4.070934040463998e-07,
      "logits/chosen": -1.7846260070800781,
      "logits/rejected": -1.7679131031036377,
      "logps/chosen": -121.9030990600586,
      "logps/rejected": -130.14794921875,
      "loss": 0.1516,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 5.171528339385986,
      "rewards/margins": 15.514431953430176,
      "rewards/rejected": -10.342904090881348,
      "step": 170
    },
    {
      "epoch": 0.38,
      "grad_norm": 266.7294156199543,
      "learning_rate": 3.9247834624635404e-07,
      "logits/chosen": -1.7311102151870728,
      "logits/rejected": -1.725064992904663,
      "logps/chosen": -125.84059143066406,
      "logps/rejected": -125.44111633300781,
      "loss": 0.1616,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 5.19378137588501,
      "rewards/margins": 15.565200805664062,
      "rewards/rejected": -10.371419906616211,
      "step": 180
    },
    {
      "epoch": 0.4,
      "grad_norm": 243.44642169675112,
      "learning_rate": 3.7710310482256523e-07,
      "logits/chosen": -1.764722466468811,
      "logits/rejected": -1.7476263046264648,
      "logps/chosen": -114.77116394042969,
      "logps/rejected": -139.33917236328125,
      "loss": 0.1684,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 5.108515739440918,
      "rewards/margins": 16.28169822692871,
      "rewards/rejected": -11.173181533813477,
      "step": 190
    },
    {
      "epoch": 0.42,
      "grad_norm": 258.83830985573707,
      "learning_rate": 3.610497133404795e-07,
      "logits/chosen": -1.7498699426651,
      "logits/rejected": -1.7545902729034424,
      "logps/chosen": -120.42120361328125,
      "logps/rejected": -123.0103988647461,
      "loss": 0.1795,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 5.614555835723877,
      "rewards/margins": 16.115243911743164,
      "rewards/rejected": -10.500688552856445,
      "step": 200
    },
    {
      "epoch": 0.42,
      "eval_logits/chosen": -1.86138117313385,
      "eval_logits/rejected": -1.8606913089752197,
      "eval_logps/chosen": -120.21395111083984,
      "eval_logps/rejected": -130.0475616455078,
      "eval_loss": 0.13641399145126343,
      "eval_rewards/accuracies": 0.93359375,
      "eval_rewards/chosen": 5.267488479614258,
      "eval_rewards/margins": 16.30953025817871,
      "eval_rewards/rejected": -11.042043685913086,
      "eval_runtime": 97.5652,
      "eval_samples_per_second": 20.499,
      "eval_steps_per_second": 0.328,
      "step": 200
    },
    {
      "epoch": 0.44,
      "grad_norm": 258.76531038458563,
      "learning_rate": 3.4440382358952115e-07,
      "logits/chosen": -1.689432144165039,
      "logits/rejected": -1.632364273071289,
      "logps/chosen": -117.7408447265625,
      "logps/rejected": -117.53926086425781,
      "loss": 0.1494,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 4.45497989654541,
      "rewards/margins": 12.836206436157227,
      "rewards/rejected": -8.381224632263184,
      "step": 210
    },
    {
      "epoch": 0.46,
      "grad_norm": 163.97448768931915,
      "learning_rate": 3.272542485937368e-07,
      "logits/chosen": -1.8162240982055664,
      "logits/rejected": -1.8604834079742432,
      "logps/chosen": -118.64457702636719,
      "logps/rejected": -118.34297180175781,
      "loss": 0.1524,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 5.34921407699585,
      "rewards/margins": 14.593510627746582,
      "rewards/rejected": -9.244296073913574,
      "step": 220
    },
    {
      "epoch": 0.48,
      "grad_norm": 122.96583033580265,
      "learning_rate": 3.096924887558854e-07,
      "logits/chosen": -1.7572132349014282,
      "logits/rejected": -1.708581566810608,
      "logps/chosen": -131.16842651367188,
      "logps/rejected": -137.84829711914062,
      "loss": 0.1815,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 4.962122440338135,
      "rewards/margins": 16.372663497924805,
      "rewards/rejected": -11.410540580749512,
      "step": 230
    },
    {
      "epoch": 0.5,
      "grad_norm": 300.9054437339817,
      "learning_rate": 2.9181224366319943e-07,
      "logits/chosen": -1.768376111984253,
      "logits/rejected": -1.7853620052337646,
      "logps/chosen": -118.08064270019531,
      "logps/rejected": -129.52337646484375,
      "loss": 0.1638,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 4.15334415435791,
      "rewards/margins": 13.721521377563477,
      "rewards/rejected": -9.568175315856934,
      "step": 240
    },
    {
      "epoch": 0.52,
      "grad_norm": 279.7087004166346,
      "learning_rate": 2.7370891215954565e-07,
      "logits/chosen": -1.7569881677627563,
      "logits/rejected": -1.7871322631835938,
      "logps/chosen": -121.43367767333984,
      "logps/rejected": -130.93099975585938,
      "loss": 0.1737,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 4.7944416999816895,
      "rewards/margins": 14.978727340698242,
      "rewards/rejected": -10.184286117553711,
      "step": 250
    },
    {
      "epoch": 0.54,
      "grad_norm": 163.84527157544522,
      "learning_rate": 2.55479083351317e-07,
      "logits/chosen": -1.7968614101409912,
      "logits/rejected": -1.8222767114639282,
      "logps/chosen": -125.540283203125,
      "logps/rejected": -119.97703552246094,
      "loss": 0.1484,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 5.486014366149902,
      "rewards/margins": 16.016239166259766,
      "rewards/rejected": -10.530224800109863,
      "step": 260
    },
    {
      "epoch": 0.56,
      "grad_norm": 150.71761898213634,
      "learning_rate": 2.3722002126275822e-07,
      "logits/chosen": -1.787398099899292,
      "logits/rejected": -1.723350167274475,
      "logps/chosen": -115.22535705566406,
      "logps/rejected": -123.31414794921875,
      "loss": 0.139,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 5.049933433532715,
      "rewards/margins": 14.674034118652344,
      "rewards/rejected": -9.624099731445312,
      "step": 270
    },
    {
      "epoch": 0.59,
      "grad_norm": 267.05255152770263,
      "learning_rate": 2.19029145890313e-07,
      "logits/chosen": -1.6403396129608154,
      "logits/rejected": -1.728877067565918,
      "logps/chosen": -121.6760482788086,
      "logps/rejected": -137.2602996826172,
      "loss": 0.1837,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 4.614184379577637,
      "rewards/margins": 15.387414932250977,
      "rewards/rejected": -10.773229598999023,
      "step": 280
    },
    {
      "epoch": 0.61,
      "grad_norm": 355.7773679008215,
      "learning_rate": 2.0100351342479216e-07,
      "logits/chosen": -1.7196556329727173,
      "logits/rejected": -1.7061948776245117,
      "logps/chosen": -112.95915222167969,
      "logps/rejected": -124.29833984375,
      "loss": 0.1648,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 4.5611090660095215,
      "rewards/margins": 15.115681648254395,
      "rewards/rejected": -10.554571151733398,
      "step": 290
    },
    {
      "epoch": 0.63,
      "grad_norm": 387.9028451533635,
      "learning_rate": 1.8323929841460178e-07,
      "logits/chosen": -1.6907150745391846,
      "logits/rejected": -1.6335220336914062,
      "logps/chosen": -130.05313110351562,
      "logps/rejected": -141.36476135253906,
      "loss": 0.1585,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 4.916111946105957,
      "rewards/margins": 15.453518867492676,
      "rewards/rejected": -10.537405967712402,
      "step": 300
    },
    {
      "epoch": 0.63,
      "eval_logits/chosen": -1.7980220317840576,
      "eval_logits/rejected": -1.7959610223770142,
      "eval_logps/chosen": -120.6431655883789,
      "eval_logps/rejected": -132.25042724609375,
      "eval_loss": 0.14247241616249084,
      "eval_rewards/accuracies": 0.92578125,
      "eval_rewards/chosen": 5.138728141784668,
      "eval_rewards/margins": 16.841632843017578,
      "eval_rewards/rejected": -11.702906608581543,
      "eval_runtime": 97.7011,
      "eval_samples_per_second": 20.471,
      "eval_steps_per_second": 0.328,
      "step": 300
    },
    {
      "epoch": 0.65,
      "grad_norm": 308.7683125090126,
      "learning_rate": 1.6583128063291573e-07,
      "logits/chosen": -1.6941922903060913,
      "logits/rejected": -1.7201515436172485,
      "logps/chosen": -121.898193359375,
      "logps/rejected": -132.00962829589844,
      "loss": 0.1619,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 5.231846809387207,
      "rewards/margins": 15.2730131149292,
      "rewards/rejected": -10.041168212890625,
      "step": 310
    },
    {
      "epoch": 0.67,
      "grad_norm": 208.88097429038612,
      "learning_rate": 1.488723393865766e-07,
      "logits/chosen": -1.7199640274047852,
      "logits/rejected": -1.6923316717147827,
      "logps/chosen": -111.40040588378906,
      "logps/rejected": -131.6649627685547,
      "loss": 0.154,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 5.818942070007324,
      "rewards/margins": 16.441822052001953,
      "rewards/rejected": -10.62287712097168,
      "step": 320
    },
    {
      "epoch": 0.69,
      "grad_norm": 167.39144709959334,
      "learning_rate": 1.3245295796480788e-07,
      "logits/chosen": -1.65103018283844,
      "logits/rejected": -1.7277615070343018,
      "logps/chosen": -121.6876449584961,
      "logps/rejected": -127.0235366821289,
      "loss": 0.1507,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 4.948590278625488,
      "rewards/margins": 14.931228637695312,
      "rewards/rejected": -9.982640266418457,
      "step": 330
    },
    {
      "epoch": 0.71,
      "grad_norm": 156.90849982717606,
      "learning_rate": 1.1666074087171627e-07,
      "logits/chosen": -1.602086067199707,
      "logits/rejected": -1.6703577041625977,
      "logps/chosen": -115.71211242675781,
      "logps/rejected": -116.56599426269531,
      "loss": 0.1908,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 4.433460235595703,
      "rewards/margins": 13.112295150756836,
      "rewards/rejected": -8.678834915161133,
      "step": 340
    },
    {
      "epoch": 0.73,
      "grad_norm": 295.768622683131,
      "learning_rate": 1.0157994641835734e-07,
      "logits/chosen": -1.6467043161392212,
      "logits/rejected": -1.659854531288147,
      "logps/chosen": -115.77754974365234,
      "logps/rejected": -137.26531982421875,
      "loss": 0.1326,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 4.292226314544678,
      "rewards/margins": 15.004185676574707,
      "rewards/rejected": -10.711957931518555,
      "step": 350
    },
    {
      "epoch": 0.75,
      "grad_norm": 201.50715466732544,
      "learning_rate": 8.729103716819111e-08,
      "logits/chosen": -1.6420570611953735,
      "logits/rejected": -1.663745641708374,
      "logps/chosen": -122.47066497802734,
      "logps/rejected": -131.9073486328125,
      "loss": 0.1543,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": 4.703906059265137,
      "rewards/margins": 14.480855941772461,
      "rewards/rejected": -9.776951789855957,
      "step": 360
    },
    {
      "epoch": 0.77,
      "grad_norm": 398.4959635020424,
      "learning_rate": 7.387025063449081e-08,
      "logits/chosen": -1.7505977153778076,
      "logits/rejected": -1.6930338144302368,
      "logps/chosen": -112.33251953125,
      "logps/rejected": -121.57230377197266,
      "loss": 0.1443,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 4.217165470123291,
      "rewards/margins": 12.93463134765625,
      "rewards/rejected": -8.7174654006958,
      "step": 370
    },
    {
      "epoch": 0.79,
      "grad_norm": 239.63863129657855,
      "learning_rate": 6.138919252022435e-08,
      "logits/chosen": -1.7603282928466797,
      "logits/rejected": -1.7833961248397827,
      "logps/chosen": -119.14958190917969,
      "logps/rejected": -136.85977172851562,
      "loss": 0.1576,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 5.251246452331543,
      "rewards/margins": 16.546478271484375,
      "rewards/rejected": -11.295232772827148,
      "step": 380
    },
    {
      "epoch": 0.82,
      "grad_norm": 197.58750581294214,
      "learning_rate": 4.991445467064689e-08,
      "logits/chosen": -1.656961441040039,
      "logits/rejected": -1.6608669757843018,
      "logps/chosen": -115.59519958496094,
      "logps/rejected": -124.886474609375,
      "loss": 0.1242,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 5.070186614990234,
      "rewards/margins": 15.444877624511719,
      "rewards/rejected": -10.3746919631958,
      "step": 390
    },
    {
      "epoch": 0.84,
      "grad_norm": 266.2162184394133,
      "learning_rate": 3.9507259776993954e-08,
      "logits/chosen": -1.636885643005371,
      "logits/rejected": -1.713200330734253,
      "logps/chosen": -117.88565826416016,
      "logps/rejected": -124.86863708496094,
      "loss": 0.2005,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 4.46274471282959,
      "rewards/margins": 14.449310302734375,
      "rewards/rejected": -9.986566543579102,
      "step": 400
    },
    {
      "epoch": 0.84,
      "eval_logits/chosen": -1.8133598566055298,
      "eval_logits/rejected": -1.8130455017089844,
      "eval_logps/chosen": -119.97606658935547,
      "eval_logps/rejected": -131.71229553222656,
      "eval_loss": 0.14202240109443665,
      "eval_rewards/accuracies": 0.92578125,
      "eval_rewards/chosen": 5.338851451873779,
      "eval_rewards/margins": 16.880319595336914,
      "eval_rewards/rejected": -11.541468620300293,
      "eval_runtime": 97.6019,
      "eval_samples_per_second": 20.491,
      "eval_steps_per_second": 0.328,
      "step": 400
    },
    {
      "epoch": 0.86,
      "grad_norm": 300.6588760835356,
      "learning_rate": 3.022313472693447e-08,
      "logits/chosen": -1.735870599746704,
      "logits/rejected": -1.7638452053070068,
      "logps/chosen": -128.670166015625,
      "logps/rejected": -129.5255126953125,
      "loss": 0.1508,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 5.593460559844971,
      "rewards/margins": 16.819358825683594,
      "rewards/rejected": -11.225897789001465,
      "step": 410
    },
    {
      "epoch": 0.88,
      "grad_norm": 184.26802651594184,
      "learning_rate": 2.2111614344599684e-08,
      "logits/chosen": -1.7033697366714478,
      "logits/rejected": -1.704993486404419,
      "logps/chosen": -113.72819519042969,
      "logps/rejected": -124.34715270996094,
      "loss": 0.1419,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 4.3744401931762695,
      "rewards/margins": 14.479223251342773,
      "rewards/rejected": -10.104782104492188,
      "step": 420
    },
    {
      "epoch": 0.9,
      "grad_norm": 269.33105520831003,
      "learning_rate": 1.521597710086439e-08,
      "logits/chosen": -1.6633046865463257,
      "logits/rejected": -1.6779390573501587,
      "logps/chosen": -129.4523162841797,
      "logps/rejected": -124.72953796386719,
      "loss": 0.154,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 4.997079372406006,
      "rewards/margins": 14.370283126831055,
      "rewards/rejected": -9.373201370239258,
      "step": 430
    },
    {
      "epoch": 0.92,
      "grad_norm": 181.5022378731684,
      "learning_rate": 9.57301420397924e-09,
      "logits/chosen": -1.7851531505584717,
      "logits/rejected": -1.790157675743103,
      "logps/chosen": -120.8554458618164,
      "logps/rejected": -131.34410095214844,
      "loss": 0.144,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 5.435299873352051,
      "rewards/margins": 16.81759262084961,
      "rewards/rejected": -11.382290840148926,
      "step": 440
    },
    {
      "epoch": 0.94,
      "grad_norm": 227.09412853715097,
      "learning_rate": 5.212833302556258e-09,
      "logits/chosen": -1.7998348474502563,
      "logits/rejected": -1.7618439197540283,
      "logps/chosen": -116.46590423583984,
      "logps/rejected": -127.42796325683594,
      "loss": 0.159,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 5.652300834655762,
      "rewards/margins": 17.369140625,
      "rewards/rejected": -11.716839790344238,
      "step": 450
    },
    {
      "epoch": 0.96,
      "grad_norm": 225.69247261851913,
      "learning_rate": 2.158697848236607e-09,
      "logits/chosen": -1.6827681064605713,
      "logits/rejected": -1.7046699523925781,
      "logps/chosen": -120.79095458984375,
      "logps/rejected": -130.98043823242188,
      "loss": 0.1335,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 5.568859100341797,
      "rewards/margins": 16.174407958984375,
      "rewards/rejected": -10.605547904968262,
      "step": 460
    },
    {
      "epoch": 0.98,
      "grad_norm": 199.48903785359678,
      "learning_rate": 4.269029751107489e-10,
      "logits/chosen": -1.6656444072723389,
      "logits/rejected": -1.6796722412109375,
      "logps/chosen": -114.7574462890625,
      "logps/rejected": -139.92117309570312,
      "loss": 0.1514,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 4.486257553100586,
      "rewards/margins": 15.330111503601074,
      "rewards/rejected": -10.843853950500488,
      "step": 470
    },
    {
      "epoch": 1.0,
      "step": 478,
      "total_flos": 0.0,
      "train_loss": 0.17749474911510196,
      "train_runtime": 7645.2484,
      "train_samples_per_second": 7.996,
      "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
}