{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 50,
  "global_step": 201,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 60.46895766728989,
      "learning_rate": 5e-07,
      "logits/chosen": -2.737339735031128,
      "logits/rejected": -2.717646837234497,
      "logps/chosen": -269.46966552734375,
      "logps/rejected": -218.519287109375,
      "loss": 0.6897,
      "rewards/accuracies": 0.33125001192092896,
      "rewards/chosen": 0.016166506335139275,
      "rewards/margins": 0.005314634181559086,
      "rewards/rejected": 0.010851874016225338,
      "step": 5
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 47.55077346828281,
      "learning_rate": 1e-06,
      "logits/chosen": -2.707444190979004,
      "logits/rejected": -2.6988959312438965,
      "logps/chosen": -259.7918395996094,
      "logps/rejected": -198.8335723876953,
      "loss": 0.6419,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 0.5464529991149902,
      "rewards/margins": 0.22305770218372345,
      "rewards/rejected": 0.323395311832428,
      "step": 10
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 49.794624138997065,
      "learning_rate": 9.983100718730718e-07,
      "logits/chosen": -2.5312893390655518,
      "logits/rejected": -2.528350353240967,
      "logps/chosen": -250.9792022705078,
      "logps/rejected": -215.8843231201172,
      "loss": 0.6325,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 1.2436325550079346,
      "rewards/margins": 0.6625381708145142,
      "rewards/rejected": 0.5810943841934204,
      "step": 15
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 45.89015113492354,
      "learning_rate": 9.932517109205849e-07,
      "logits/chosen": -2.345860719680786,
      "logits/rejected": -2.330219030380249,
      "logps/chosen": -261.6305236816406,
      "logps/rejected": -205.9855499267578,
      "loss": 0.5741,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 1.5929079055786133,
      "rewards/margins": 0.922720730304718,
      "rewards/rejected": 0.6701873540878296,
      "step": 20
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 63.53673896186562,
      "learning_rate": 9.848591102083375e-07,
      "logits/chosen": -2.2251970767974854,
      "logits/rejected": -2.210585832595825,
      "logps/chosen": -256.94158935546875,
      "logps/rejected": -198.90554809570312,
      "loss": 0.5846,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 1.4611073732376099,
      "rewards/margins": 0.925146758556366,
      "rewards/rejected": 0.5359605550765991,
      "step": 25
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 40.6998829965568,
      "learning_rate": 9.731890013043367e-07,
      "logits/chosen": -2.203108310699463,
      "logits/rejected": -2.188232898712158,
      "logps/chosen": -220.20474243164062,
      "logps/rejected": -174.1934356689453,
      "loss": 0.5755,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 1.1598186492919922,
      "rewards/margins": 1.0439460277557373,
      "rewards/rejected": 0.11587252467870712,
      "step": 30
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 39.86530018922445,
      "learning_rate": 9.583202707897073e-07,
      "logits/chosen": -2.159541606903076,
      "logits/rejected": -2.146280288696289,
      "logps/chosen": -271.1134033203125,
      "logps/rejected": -222.50732421875,
      "loss": 0.5581,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": 1.6402053833007812,
      "rewards/margins": 1.4621944427490234,
      "rewards/rejected": 0.1780107319355011,
      "step": 35
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 47.764041993597665,
      "learning_rate": 9.403534270080829e-07,
      "logits/chosen": -2.097519874572754,
      "logits/rejected": -2.0528969764709473,
      "logps/chosen": -246.8362274169922,
      "logps/rejected": -208.1300048828125,
      "loss": 0.5329,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 1.0587269067764282,
      "rewards/margins": 1.0248701572418213,
      "rewards/rejected": 0.033856604248285294,
      "step": 40
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 33.33992129726437,
      "learning_rate": 9.19409920658098e-07,
      "logits/chosen": -2.1117985248565674,
      "logits/rejected": -2.0575757026672363,
      "logps/chosen": -238.47335815429688,
      "logps/rejected": -232.6268310546875,
      "loss": 0.5597,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 1.020668387413025,
      "rewards/margins": 1.3137915134429932,
      "rewards/rejected": -0.2931232154369354,
      "step": 45
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 36.75599018498795,
      "learning_rate": 8.956313238215823e-07,
      "logits/chosen": -2.192502021789551,
      "logits/rejected": -2.1830239295959473,
      "logps/chosen": -246.92733764648438,
      "logps/rejected": -219.84817504882812,
      "loss": 0.5242,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.8840469121932983,
      "rewards/margins": 1.3445603847503662,
      "rewards/rejected": -0.4605133533477783,
      "step": 50
    },
    {
      "epoch": 0.746268656716418,
      "eval_logits/chosen": -2.2182116508483887,
      "eval_logits/rejected": -2.195751190185547,
      "eval_logps/chosen": -256.5732727050781,
      "eval_logps/rejected": -219.44544982910156,
      "eval_loss": 0.5641670823097229,
      "eval_rewards/accuracies": 0.7583333253860474,
      "eval_rewards/chosen": 0.8116810321807861,
      "eval_rewards/margins": 1.5396502017974854,
      "eval_rewards/rejected": -0.7279691696166992,
      "eval_runtime": 125.8473,
      "eval_samples_per_second": 15.098,
      "eval_steps_per_second": 0.238,
      "step": 50
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 33.66480221016764,
      "learning_rate": 8.691783729769873e-07,
      "logits/chosen": -2.252897024154663,
      "logits/rejected": -2.210509777069092,
      "logps/chosen": -273.36102294921875,
      "logps/rejected": -232.4573974609375,
      "loss": 0.5359,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": 1.0710394382476807,
      "rewards/margins": 1.8447446823120117,
      "rewards/rejected": -0.7737051248550415,
      "step": 55
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 42.22753576523594,
      "learning_rate": 8.402298824670029e-07,
      "logits/chosen": -2.2703890800476074,
      "logits/rejected": -2.252363443374634,
      "logps/chosen": -274.8329162597656,
      "logps/rejected": -254.45663452148438,
      "loss": 0.5811,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.9601489901542664,
      "rewards/margins": 1.3747169971466064,
      "rewards/rejected": -0.4145679473876953,
      "step": 60
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 38.82616914526827,
      "learning_rate": 8.089815357650089e-07,
      "logits/chosen": -2.237731456756592,
      "logits/rejected": -2.245671510696411,
      "logps/chosen": -264.57989501953125,
      "logps/rejected": -241.2982940673828,
      "loss": 0.5166,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": 1.0662319660186768,
      "rewards/margins": 1.8653961420059204,
      "rewards/rejected": -0.799164354801178,
      "step": 65
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 21.744667598719886,
      "learning_rate": 7.756445627110522e-07,
      "logits/chosen": -2.2960398197174072,
      "logits/rejected": -2.2724010944366455,
      "logps/chosen": -264.32720947265625,
      "logps/rejected": -239.6345977783203,
      "loss": 0.346,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": 2.158931255340576,
      "rewards/margins": 2.908942222595215,
      "rewards/rejected": -0.7500110864639282,
      "step": 70
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 19.678947962162013,
      "learning_rate": 7.404443116588547e-07,
      "logits/chosen": -2.396239995956421,
      "logits/rejected": -2.362546443939209,
      "logps/chosen": -246.3358154296875,
      "logps/rejected": -195.57357788085938,
      "loss": 0.2283,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.2872486114501953,
      "rewards/margins": 2.971306562423706,
      "rewards/rejected": -0.6840583086013794,
      "step": 75
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 20.2993371431914,
      "learning_rate": 7.036187261857288e-07,
      "logits/chosen": -2.4099860191345215,
      "logits/rejected": -2.384648084640503,
      "logps/chosen": -260.5028381347656,
      "logps/rejected": -223.2133331298828,
      "loss": 0.2312,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.834780693054199,
      "rewards/margins": 3.5776984691619873,
      "rewards/rejected": -0.7429176568984985,
      "step": 80
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 25.4573275034557,
      "learning_rate": 6.654167366624008e-07,
      "logits/chosen": -2.36556339263916,
      "logits/rejected": -2.349255084991455,
      "logps/chosen": -231.6350860595703,
      "logps/rejected": -221.1610870361328,
      "loss": 0.2307,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 1.9625335931777954,
      "rewards/margins": 3.2484028339385986,
      "rewards/rejected": -1.2858693599700928,
      "step": 85
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 29.79365948091661,
      "learning_rate": 6.260965775552713e-07,
      "logits/chosen": -2.374537944793701,
      "logits/rejected": -2.348567485809326,
      "logps/chosen": -247.57614135742188,
      "logps/rejected": -238.37899780273438,
      "loss": 0.2523,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 1.934003233909607,
      "rewards/margins": 3.8235995769500732,
      "rewards/rejected": -1.8895965814590454,
      "step": 90
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 22.109779497676236,
      "learning_rate": 5.859240418356614e-07,
      "logits/chosen": -2.3896210193634033,
      "logits/rejected": -2.3736777305603027,
      "logps/chosen": -257.51898193359375,
      "logps/rejected": -228.7509765625,
      "loss": 0.2453,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.010754108428955,
      "rewards/margins": 3.4373581409454346,
      "rewards/rejected": -1.4266036748886108,
      "step": 95
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 18.59821752381129,
      "learning_rate": 5.451706842957421e-07,
      "logits/chosen": -2.3828063011169434,
      "logits/rejected": -2.3802871704101562,
      "logps/chosen": -238.89547729492188,
      "logps/rejected": -224.2349395751953,
      "loss": 0.2523,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.049046039581299,
      "rewards/margins": 3.4148106575012207,
      "rewards/rejected": -1.3657643795013428,
      "step": 100
    },
    {
      "epoch": 1.4925373134328357,
      "eval_logits/chosen": -2.3702290058135986,
      "eval_logits/rejected": -2.352924108505249,
      "eval_logps/chosen": -250.75942993164062,
      "eval_logps/rejected": -224.2054901123047,
      "eval_loss": 0.5878031253814697,
      "eval_rewards/accuracies": 0.8041666746139526,
      "eval_rewards/chosen": 1.393066644668579,
      "eval_rewards/margins": 2.597040891647339,
      "eval_rewards/rejected": -1.2039742469787598,
      "eval_runtime": 125.6186,
      "eval_samples_per_second": 15.125,
      "eval_steps_per_second": 0.239,
      "step": 100
    },
    {
      "epoch": 1.5671641791044775,
      "grad_norm": 24.221807777872822,
      "learning_rate": 5.041119859162068e-07,
      "logits/chosen": -2.3542027473449707,
      "logits/rejected": -2.3176567554473877,
      "logps/chosen": -224.2234649658203,
      "logps/rejected": -210.6978302001953,
      "loss": 0.3,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.6206858158111572,
      "rewards/margins": 3.042532444000244,
      "rewards/rejected": -1.4218467473983765,
      "step": 105
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 21.221389216973705,
      "learning_rate": 4.630254916940423e-07,
      "logits/chosen": -2.3093810081481934,
      "logits/rejected": -2.300978660583496,
      "logps/chosen": -242.2421875,
      "logps/rejected": -230.2930145263672,
      "loss": 0.2471,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 1.8961149454116821,
      "rewards/margins": 3.936471462249756,
      "rewards/rejected": -2.0403571128845215,
      "step": 110
    },
    {
      "epoch": 1.716417910447761,
      "grad_norm": 28.90188297206431,
      "learning_rate": 4.2218893451814e-07,
      "logits/chosen": -2.299715042114258,
      "logits/rejected": -2.3139047622680664,
      "logps/chosen": -260.2466735839844,
      "logps/rejected": -238.0985870361328,
      "loss": 0.2589,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 1.7004687786102295,
      "rewards/margins": 3.8741843700408936,
      "rewards/rejected": -2.173715353012085,
      "step": 115
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 25.74215758531903,
      "learning_rate": 3.8187835777481375e-07,
      "logits/chosen": -2.292194366455078,
      "logits/rejected": -2.290935754776001,
      "logps/chosen": -252.3697509765625,
      "logps/rejected": -246.54690551757812,
      "loss": 0.2766,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 1.6656386852264404,
      "rewards/margins": 3.7279200553894043,
      "rewards/rejected": -2.062281608581543,
      "step": 120
    },
    {
      "epoch": 1.8656716417910446,
      "grad_norm": 22.13418877428713,
      "learning_rate": 3.423662493738687e-07,
      "logits/chosen": -2.2597708702087402,
      "logits/rejected": -2.22979474067688,
      "logps/chosen": -242.34994506835938,
      "logps/rejected": -205.6407470703125,
      "loss": 0.2627,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 1.9104535579681396,
      "rewards/margins": 3.547952175140381,
      "rewards/rejected": -1.6374982595443726,
      "step": 125
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 25.182255638442037,
      "learning_rate": 3.039196998086687e-07,
      "logits/chosen": -2.216444730758667,
      "logits/rejected": -2.2051291465759277,
      "logps/chosen": -240.48526000976562,
      "logps/rejected": -241.6737518310547,
      "loss": 0.278,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.2690188884735107,
      "rewards/margins": 3.9218125343322754,
      "rewards/rejected": -1.6527936458587646,
      "step": 130
    },
    {
      "epoch": 2.014925373134328,
      "grad_norm": 17.639179913168814,
      "learning_rate": 2.667985967011878e-07,
      "logits/chosen": -2.2279441356658936,
      "logits/rejected": -2.202115297317505,
      "logps/chosen": -256.98162841796875,
      "logps/rejected": -240.4128875732422,
      "loss": 0.2636,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 2.024724245071411,
      "rewards/margins": 3.793694257736206,
      "rewards/rejected": -1.7689695358276367,
      "step": 135
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 16.436781206837193,
      "learning_rate": 2.3125386803640183e-07,
      "logits/chosen": -2.208228349685669,
      "logits/rejected": -2.1990818977355957,
      "logps/chosen": -241.15542602539062,
      "logps/rejected": -233.29031372070312,
      "loss": 0.1344,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 1.9513914585113525,
      "rewards/margins": 3.8950798511505127,
      "rewards/rejected": -1.9436886310577393,
      "step": 140
    },
    {
      "epoch": 2.1641791044776117,
      "grad_norm": 17.047112239762754,
      "learning_rate": 1.9752578596124952e-07,
      "logits/chosen": -2.233797550201416,
      "logits/rejected": -2.2078850269317627,
      "logps/chosen": -252.17111206054688,
      "logps/rejected": -218.3225555419922,
      "loss": 0.1554,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 2.0858609676361084,
      "rewards/margins": 3.9756836891174316,
      "rewards/rejected": -1.8898223638534546,
      "step": 145
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 14.963330285006984,
      "learning_rate": 1.6584234261399532e-07,
      "logits/chosen": -2.2354073524475098,
      "logits/rejected": -2.238706588745117,
      "logps/chosen": -226.289306640625,
      "logps/rejected": -246.685791015625,
      "loss": 0.1264,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 2.1262850761413574,
      "rewards/margins": 4.3760247230529785,
      "rewards/rejected": -2.249739408493042,
      "step": 150
    },
    {
      "epoch": 2.2388059701492535,
      "eval_logits/chosen": -2.265167713165283,
      "eval_logits/rejected": -2.2381813526153564,
      "eval_logps/chosen": -251.42041015625,
      "eval_logps/rejected": -226.3953399658203,
      "eval_loss": 0.5601586103439331,
      "eval_rewards/accuracies": 0.8166666626930237,
      "eval_rewards/chosen": 1.326967477798462,
      "eval_rewards/margins": 2.749927282333374,
      "eval_rewards/rejected": -1.422959566116333,
      "eval_runtime": 125.5687,
      "eval_samples_per_second": 15.131,
      "eval_steps_per_second": 0.239,
      "step": 150
    },
    {
      "epoch": 2.3134328358208958,
      "grad_norm": 17.27452729186339,
      "learning_rate": 1.3641770896292082e-07,
      "logits/chosen": -2.2842135429382324,
      "logits/rejected": -2.229306697845459,
      "logps/chosen": -242.30044555664062,
      "logps/rejected": -227.74948120117188,
      "loss": 0.1578,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": 2.336923360824585,
      "rewards/margins": 4.076951026916504,
      "rewards/rejected": -1.7400274276733398,
      "step": 155
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 19.880826026501197,
      "learning_rate": 1.0945078707215221e-07,
      "logits/chosen": -2.263514757156372,
      "logits/rejected": -2.2274537086486816,
      "logps/chosen": -247.376220703125,
      "logps/rejected": -231.4269256591797,
      "loss": 0.1686,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.331441640853882,
      "rewards/margins": 3.9185550212860107,
      "rewards/rejected": -1.5871132612228394,
      "step": 160
    },
    {
      "epoch": 2.4626865671641793,
      "grad_norm": 18.037005682288928,
      "learning_rate": 8.512386558088919e-08,
      "logits/chosen": -2.2772862911224365,
      "logits/rejected": -2.248198986053467,
      "logps/chosen": -241.83132934570312,
      "logps/rejected": -266.8996887207031,
      "loss": 0.168,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.4042108058929443,
      "rewards/margins": 4.915772914886475,
      "rewards/rejected": -2.511561870574951,
      "step": 165
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 19.957029846976177,
      "learning_rate": 6.360138748461013e-08,
      "logits/chosen": -2.2610926628112793,
      "logits/rejected": -2.2388763427734375,
      "logps/chosen": -255.074951171875,
      "logps/rejected": -226.93997192382812,
      "loss": 0.1771,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.393348455429077,
      "rewards/margins": 4.332278251647949,
      "rewards/rejected": -1.9389299154281616,
      "step": 170
    },
    {
      "epoch": 2.611940298507463,
      "grad_norm": 15.104058742033901,
      "learning_rate": 4.5028838547699346e-08,
      "logits/chosen": -2.271852970123291,
      "logits/rejected": -2.2440991401672363,
      "logps/chosen": -242.74697875976562,
      "logps/rejected": -235.72128295898438,
      "loss": 0.1392,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.28426456451416,
      "rewards/margins": 4.252217769622803,
      "rewards/rejected": -1.9679529666900635,
      "step": 175
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 14.425654092236623,
      "learning_rate": 2.9531763861505964e-08,
      "logits/chosen": -2.2854719161987305,
      "logits/rejected": -2.243256092071533,
      "logps/chosen": -251.6754150390625,
      "logps/rejected": -246.1281280517578,
      "loss": 0.1574,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 2.5320801734924316,
      "rewards/margins": 4.495547771453857,
      "rewards/rejected": -1.9634678363800049,
      "step": 180
    },
    {
      "epoch": 2.7611940298507465,
      "grad_norm": 18.97471020277948,
      "learning_rate": 1.7214919195619125e-08,
      "logits/chosen": -2.2592127323150635,
      "logits/rejected": -2.2206928730010986,
      "logps/chosen": -236.07418823242188,
      "logps/rejected": -231.7993621826172,
      "loss": 0.1479,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 1.6844587326049805,
      "rewards/margins": 4.147632122039795,
      "rewards/rejected": -2.4631736278533936,
      "step": 185
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 19.094777076672973,
      "learning_rate": 8.161562878982398e-09,
      "logits/chosen": -2.2626476287841797,
      "logits/rejected": -2.2289066314697266,
      "logps/chosen": -253.0904083251953,
      "logps/rejected": -223.6345672607422,
      "loss": 0.1273,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 2.360367774963379,
      "rewards/margins": 4.313665390014648,
      "rewards/rejected": -1.9532979726791382,
      "step": 190
    },
    {
      "epoch": 2.91044776119403,
      "grad_norm": 18.606332473099318,
      "learning_rate": 2.432892997526026e-09,
      "logits/chosen": -2.2822461128234863,
      "logits/rejected": -2.233109951019287,
      "logps/chosen": -226.60153198242188,
      "logps/rejected": -230.42422485351562,
      "loss": 0.1154,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 2.1634068489074707,
      "rewards/margins": 4.176701545715332,
      "rewards/rejected": -2.013294219970703,
      "step": 195
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 17.32503927255026,
      "learning_rate": 6.763371270035457e-11,
      "logits/chosen": -2.2751100063323975,
      "logits/rejected": -2.2504701614379883,
      "logps/chosen": -249.61959838867188,
      "logps/rejected": -224.1602020263672,
      "loss": 0.1301,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 2.2563796043395996,
      "rewards/margins": 4.354907512664795,
      "rewards/rejected": -2.0985279083251953,
      "step": 200
    },
    {
      "epoch": 2.9850746268656714,
      "eval_logits/chosen": -2.265265464782715,
      "eval_logits/rejected": -2.235792398452759,
      "eval_logps/chosen": -253.1015625,
      "eval_logps/rejected": -228.8603515625,
      "eval_loss": 0.5625271201133728,
      "eval_rewards/accuracies": 0.8208333253860474,
      "eval_rewards/chosen": 1.1588523387908936,
      "eval_rewards/margins": 2.8283116817474365,
      "eval_rewards/rejected": -1.6694589853286743,
      "eval_runtime": 125.4908,
      "eval_samples_per_second": 15.141,
      "eval_steps_per_second": 0.239,
      "step": 200
    },
    {
      "epoch": 3.0,
      "step": 201,
      "total_flos": 2369906314051584.0,
      "train_loss": 0.3258402518075497,
      "train_runtime": 7367.4174,
      "train_samples_per_second": 6.961,
      "train_steps_per_second": 0.027
    }
  ],
  "logging_steps": 5,
  "max_steps": 201,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2369906314051584.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}