{
  "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": 61.0959203382154,
      "learning_rate": 5e-07,
      "logits/chosen": -2.7374632358551025,
      "logits/rejected": -2.7177176475524902,
      "logps/chosen": -269.4803466796875,
      "logps/rejected": -218.5151824951172,
      "loss": 0.6899,
      "rewards/accuracies": 0.32499998807907104,
      "rewards/chosen": 0.015100441873073578,
      "rewards/margins": 0.00384025601670146,
      "rewards/rejected": 0.01126018911600113,
      "step": 5
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 47.476570306913814,
      "learning_rate": 1e-06,
      "logits/chosen": -2.707982063293457,
      "logits/rejected": -2.699347972869873,
      "logps/chosen": -259.8061218261719,
      "logps/rejected": -198.84274291992188,
      "loss": 0.6424,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.5450258255004883,
      "rewards/margins": 0.22254791855812073,
      "rewards/rejected": 0.32247787714004517,
      "step": 10
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 49.6551549631888,
      "learning_rate": 9.983100718730718e-07,
      "logits/chosen": -2.5320260524749756,
      "logits/rejected": -2.5290334224700928,
      "logps/chosen": -251.0294189453125,
      "logps/rejected": -215.8909912109375,
      "loss": 0.6323,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 1.2386159896850586,
      "rewards/margins": 0.6581863164901733,
      "rewards/rejected": 0.5804295539855957,
      "step": 15
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 45.55094865985364,
      "learning_rate": 9.932517109205849e-07,
      "logits/chosen": -2.3432915210723877,
      "logits/rejected": -2.327422857284546,
      "logps/chosen": -261.6549377441406,
      "logps/rejected": -206.00390625,
      "loss": 0.5745,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 1.5904659032821655,
      "rewards/margins": 0.9221122860908508,
      "rewards/rejected": 0.6683536767959595,
      "step": 20
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 66.0842546576943,
      "learning_rate": 9.848591102083375e-07,
      "logits/chosen": -2.217444896697998,
      "logits/rejected": -2.2028653621673584,
      "logps/chosen": -256.9215087890625,
      "logps/rejected": -198.91497802734375,
      "loss": 0.5856,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 1.46311354637146,
      "rewards/margins": 0.9280970692634583,
      "rewards/rejected": 0.535016655921936,
      "step": 25
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 39.48000022183853,
      "learning_rate": 9.731890013043367e-07,
      "logits/chosen": -2.1940948963165283,
      "logits/rejected": -2.1792752742767334,
      "logps/chosen": -220.172119140625,
      "logps/rejected": -174.2005615234375,
      "loss": 0.5741,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 1.1630795001983643,
      "rewards/margins": 1.0479199886322021,
      "rewards/rejected": 0.11515955626964569,
      "step": 30
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 39.73516854243651,
      "learning_rate": 9.583202707897073e-07,
      "logits/chosen": -2.151578426361084,
      "logits/rejected": -2.138328790664673,
      "logps/chosen": -271.12274169921875,
      "logps/rejected": -222.46035766601562,
      "loss": 0.557,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 1.6392730474472046,
      "rewards/margins": 1.4565646648406982,
      "rewards/rejected": 0.18270835280418396,
      "step": 35
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 48.22042549381445,
      "learning_rate": 9.403534270080829e-07,
      "logits/chosen": -2.090759515762329,
      "logits/rejected": -2.0454089641571045,
      "logps/chosen": -246.8798370361328,
      "logps/rejected": -208.13858032226562,
      "loss": 0.5337,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 1.0543644428253174,
      "rewards/margins": 1.0213631391525269,
      "rewards/rejected": 0.033001236617565155,
      "step": 40
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 33.38243910231056,
      "learning_rate": 9.19409920658098e-07,
      "logits/chosen": -2.1067168712615967,
      "logits/rejected": -2.0509848594665527,
      "logps/chosen": -238.5055694580078,
      "logps/rejected": -232.6929473876953,
      "loss": 0.56,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 1.017447590827942,
      "rewards/margins": 1.3171851634979248,
      "rewards/rejected": -0.2997375428676605,
      "step": 45
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 36.138653661809,
      "learning_rate": 8.956313238215823e-07,
      "logits/chosen": -2.1871049404144287,
      "logits/rejected": -2.1779439449310303,
      "logps/chosen": -246.9051513671875,
      "logps/rejected": -219.9314727783203,
      "loss": 0.5248,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.8862677812576294,
      "rewards/margins": 1.3551105260849,
      "rewards/rejected": -0.46884289383888245,
      "step": 50
    },
    {
      "epoch": 0.746268656716418,
      "eval_logits/chosen": -2.2122514247894287,
      "eval_logits/rejected": -2.189319372177124,
      "eval_logps/chosen": -256.70928955078125,
      "eval_logps/rejected": -219.54171752929688,
      "eval_loss": 0.5635779500007629,
      "eval_rewards/accuracies": 0.762499988079071,
      "eval_rewards/chosen": 0.7980793118476868,
      "eval_rewards/margins": 1.5356768369674683,
      "eval_rewards/rejected": -0.7375972867012024,
      "eval_runtime": 126.5339,
      "eval_samples_per_second": 15.016,
      "eval_steps_per_second": 0.237,
      "step": 50
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 33.94437450648354,
      "learning_rate": 8.691783729769873e-07,
      "logits/chosen": -2.2479443550109863,
      "logits/rejected": -2.2048230171203613,
      "logps/chosen": -273.5711669921875,
      "logps/rejected": -232.6655731201172,
      "loss": 0.5358,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 1.0500261783599854,
      "rewards/margins": 1.8445498943328857,
      "rewards/rejected": -0.7945237755775452,
      "step": 55
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 37.74564809329687,
      "learning_rate": 8.402298824670029e-07,
      "logits/chosen": -2.271031379699707,
      "logits/rejected": -2.2530617713928223,
      "logps/chosen": -274.867431640625,
      "logps/rejected": -254.64004516601562,
      "loss": 0.5812,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.9566941261291504,
      "rewards/margins": 1.3896009922027588,
      "rewards/rejected": -0.4329068064689636,
      "step": 60
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 38.80874148817025,
      "learning_rate": 8.089815357650089e-07,
      "logits/chosen": -2.2410566806793213,
      "logits/rejected": -2.2488720417022705,
      "logps/chosen": -264.5858154296875,
      "logps/rejected": -241.58132934570312,
      "loss": 0.5166,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 1.0656400918960571,
      "rewards/margins": 1.8931087255477905,
      "rewards/rejected": -0.8274686932563782,
      "step": 65
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 21.639463733660858,
      "learning_rate": 7.756445627110522e-07,
      "logits/chosen": -2.2948756217956543,
      "logits/rejected": -2.271857738494873,
      "logps/chosen": -264.6029052734375,
      "logps/rejected": -239.9524688720703,
      "loss": 0.3443,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 2.131361484527588,
      "rewards/margins": 2.913156747817993,
      "rewards/rejected": -0.78179532289505,
      "step": 70
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 20.321072984977437,
      "learning_rate": 7.404443116588547e-07,
      "logits/chosen": -2.4182841777801514,
      "logits/rejected": -2.384591579437256,
      "logps/chosen": -246.2628173828125,
      "logps/rejected": -195.36721801757812,
      "loss": 0.228,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.294548988342285,
      "rewards/margins": 2.9579689502716064,
      "rewards/rejected": -0.663419783115387,
      "step": 75
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 20.59550122688431,
      "learning_rate": 7.036187261857288e-07,
      "logits/chosen": -2.4467270374298096,
      "logits/rejected": -2.423452854156494,
      "logps/chosen": -260.4637145996094,
      "logps/rejected": -222.98629760742188,
      "loss": 0.2314,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.8386921882629395,
      "rewards/margins": 3.558907985687256,
      "rewards/rejected": -0.720215916633606,
      "step": 80
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 24.122507472196027,
      "learning_rate": 6.654167366624008e-07,
      "logits/chosen": -2.408604145050049,
      "logits/rejected": -2.39865779876709,
      "logps/chosen": -232.1356201171875,
      "logps/rejected": -221.2658233642578,
      "loss": 0.2303,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 1.9124815464019775,
      "rewards/margins": 3.2088208198547363,
      "rewards/rejected": -1.2963393926620483,
      "step": 85
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 42.3904683299003,
      "learning_rate": 6.260965775552713e-07,
      "logits/chosen": -2.4137942790985107,
      "logits/rejected": -2.3900933265686035,
      "logps/chosen": -248.0010986328125,
      "logps/rejected": -238.64633178710938,
      "loss": 0.256,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 1.8915069103240967,
      "rewards/margins": 3.8078389167785645,
      "rewards/rejected": -1.9163316488265991,
      "step": 90
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 21.496325022000395,
      "learning_rate": 5.859240418356614e-07,
      "logits/chosen": -2.409926176071167,
      "logits/rejected": -2.3968422412872314,
      "logps/chosen": -257.95928955078125,
      "logps/rejected": -229.03427124023438,
      "loss": 0.2499,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.9667186737060547,
      "rewards/margins": 3.4216525554656982,
      "rewards/rejected": -1.4549338817596436,
      "step": 95
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 18.487359146042973,
      "learning_rate": 5.451706842957421e-07,
      "logits/chosen": -2.389950752258301,
      "logits/rejected": -2.3891632556915283,
      "logps/chosen": -239.16134643554688,
      "logps/rejected": -224.28585815429688,
      "loss": 0.2521,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.022460460662842,
      "rewards/margins": 3.3933162689208984,
      "rewards/rejected": -1.3708559274673462,
      "step": 100
    },
    {
      "epoch": 1.4925373134328357,
      "eval_logits/chosen": -2.3766472339630127,
      "eval_logits/rejected": -2.3607330322265625,
      "eval_logps/chosen": -251.21282958984375,
      "eval_logps/rejected": -224.6766815185547,
      "eval_loss": 0.588749885559082,
      "eval_rewards/accuracies": 0.8041666746139526,
      "eval_rewards/chosen": 1.3477253913879395,
      "eval_rewards/margins": 2.5988175868988037,
      "eval_rewards/rejected": -1.251092553138733,
      "eval_runtime": 125.127,
      "eval_samples_per_second": 15.185,
      "eval_steps_per_second": 0.24,
      "step": 100
    },
    {
      "epoch": 1.5671641791044775,
      "grad_norm": 23.780577010359053,
      "learning_rate": 5.041119859162068e-07,
      "logits/chosen": -2.3645894527435303,
      "logits/rejected": -2.3296666145324707,
      "logps/chosen": -224.6007537841797,
      "logps/rejected": -210.8166961669922,
      "loss": 0.3064,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.5829575061798096,
      "rewards/margins": 3.0166916847229004,
      "rewards/rejected": -1.4337340593338013,
      "step": 105
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 21.34094409707377,
      "learning_rate": 4.630254916940423e-07,
      "logits/chosen": -2.327462673187256,
      "logits/rejected": -2.322633743286133,
      "logps/chosen": -242.1077117919922,
      "logps/rejected": -230.09719848632812,
      "loss": 0.2489,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 1.909563422203064,
      "rewards/margins": 3.9303386211395264,
      "rewards/rejected": -2.020775318145752,
      "step": 110
    },
    {
      "epoch": 1.716417910447761,
      "grad_norm": 22.148504337429173,
      "learning_rate": 4.2218893451814e-07,
      "logits/chosen": -2.3157925605773926,
      "logits/rejected": -2.328068733215332,
      "logps/chosen": -259.54437255859375,
      "logps/rejected": -237.518798828125,
      "loss": 0.2599,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 1.7707016468048096,
      "rewards/margins": 3.8864364624023438,
      "rewards/rejected": -2.1157352924346924,
      "step": 115
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 25.842103699055016,
      "learning_rate": 3.8187835777481375e-07,
      "logits/chosen": -2.306485652923584,
      "logits/rejected": -2.307807445526123,
      "logps/chosen": -252.6112518310547,
      "logps/rejected": -246.40322875976562,
      "loss": 0.2738,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 1.641487717628479,
      "rewards/margins": 3.6894028186798096,
      "rewards/rejected": -2.047914981842041,
      "step": 120
    },
    {
      "epoch": 1.8656716417910446,
      "grad_norm": 22.252666568025052,
      "learning_rate": 3.423662493738687e-07,
      "logits/chosen": -2.279599189758301,
      "logits/rejected": -2.2538650035858154,
      "logps/chosen": -243.33535766601562,
      "logps/rejected": -206.66940307617188,
      "loss": 0.2635,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.8119112253189087,
      "rewards/margins": 3.5522758960723877,
      "rewards/rejected": -1.740364670753479,
      "step": 125
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 25.3950797198185,
      "learning_rate": 3.039196998086687e-07,
      "logits/chosen": -2.2437617778778076,
      "logits/rejected": -2.2337818145751953,
      "logps/chosen": -241.3092498779297,
      "logps/rejected": -242.51657104492188,
      "loss": 0.278,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.186619281768799,
      "rewards/margins": 3.923696517944336,
      "rewards/rejected": -1.7370771169662476,
      "step": 130
    },
    {
      "epoch": 2.014925373134328,
      "grad_norm": 17.8657699205854,
      "learning_rate": 2.667985967011878e-07,
      "logits/chosen": -2.255903720855713,
      "logits/rejected": -2.2353529930114746,
      "logps/chosen": -257.59307861328125,
      "logps/rejected": -241.2735137939453,
      "loss": 0.2614,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.9635776281356812,
      "rewards/margins": 3.8186099529266357,
      "rewards/rejected": -1.8550323247909546,
      "step": 135
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 16.448108998699187,
      "learning_rate": 2.3125386803640183e-07,
      "logits/chosen": -2.245354175567627,
      "logits/rejected": -2.2395224571228027,
      "logps/chosen": -241.65097045898438,
      "logps/rejected": -234.07504272460938,
      "loss": 0.1341,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 1.9018363952636719,
      "rewards/margins": 3.9239959716796875,
      "rewards/rejected": -2.0221593379974365,
      "step": 140
    },
    {
      "epoch": 2.1641791044776117,
      "grad_norm": 16.872186227206235,
      "learning_rate": 1.9752578596124952e-07,
      "logits/chosen": -2.273176431655884,
      "logits/rejected": -2.252638578414917,
      "logps/chosen": -252.98904418945312,
      "logps/rejected": -219.36355590820312,
      "loss": 0.1518,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.0040693283081055,
      "rewards/margins": 3.9979896545410156,
      "rewards/rejected": -1.9939203262329102,
      "step": 145
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 14.753431701417778,
      "learning_rate": 1.6584234261399532e-07,
      "logits/chosen": -2.278550148010254,
      "logits/rejected": -2.28568696975708,
      "logps/chosen": -227.6191864013672,
      "logps/rejected": -248.12637329101562,
      "loss": 0.1273,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 1.9932959079742432,
      "rewards/margins": 4.387091636657715,
      "rewards/rejected": -2.3937957286834717,
      "step": 150
    },
    {
      "epoch": 2.2388059701492535,
      "eval_logits/chosen": -2.3105554580688477,
      "eval_logits/rejected": -2.2852249145507812,
      "eval_logps/chosen": -252.7459259033203,
      "eval_logps/rejected": -227.99476623535156,
      "eval_loss": 0.5529021620750427,
      "eval_rewards/accuracies": 0.824999988079071,
      "eval_rewards/chosen": 1.1944137811660767,
      "eval_rewards/margins": 2.777313709259033,
      "eval_rewards/rejected": -1.5828994512557983,
      "eval_runtime": 125.2259,
      "eval_samples_per_second": 15.173,
      "eval_steps_per_second": 0.24,
      "step": 150
    },
    {
      "epoch": 2.3134328358208958,
      "grad_norm": 16.834571942257156,
      "learning_rate": 1.3641770896292082e-07,
      "logits/chosen": -2.3284482955932617,
      "logits/rejected": -2.2805988788604736,
      "logps/chosen": -243.53829956054688,
      "logps/rejected": -229.0917205810547,
      "loss": 0.1588,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 2.213130235671997,
      "rewards/margins": 4.087383270263672,
      "rewards/rejected": -1.874253273010254,
      "step": 155
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 19.56642516410395,
      "learning_rate": 1.0945078707215221e-07,
      "logits/chosen": -2.3081579208374023,
      "logits/rejected": -2.2797722816467285,
      "logps/chosen": -247.8934783935547,
      "logps/rejected": -233.1248779296875,
      "loss": 0.1638,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.279715061187744,
      "rewards/margins": 4.036622524261475,
      "rewards/rejected": -1.7569072246551514,
      "step": 160
    },
    {
      "epoch": 2.4626865671641793,
      "grad_norm": 18.300870643833345,
      "learning_rate": 8.512386558088919e-08,
      "logits/chosen": -2.3260269165039062,
      "logits/rejected": -2.2975800037384033,
      "logps/chosen": -242.68701171875,
      "logps/rejected": -268.1483154296875,
      "loss": 0.1669,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.3186392784118652,
      "rewards/margins": 4.955061912536621,
      "rewards/rejected": -2.636422634124756,
      "step": 165
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 21.6227394122481,
      "learning_rate": 6.360138748461013e-08,
      "logits/chosen": -2.3063161373138428,
      "logits/rejected": -2.2875423431396484,
      "logps/chosen": -255.9014129638672,
      "logps/rejected": -227.6916961669922,
      "loss": 0.1735,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 2.3107035160064697,
      "rewards/margins": 4.324804782867432,
      "rewards/rejected": -2.01410174369812,
      "step": 170
    },
    {
      "epoch": 2.611940298507463,
      "grad_norm": 14.773801488331259,
      "learning_rate": 4.5028838547699346e-08,
      "logits/chosen": -2.3168740272521973,
      "logits/rejected": -2.2947440147399902,
      "logps/chosen": -243.5282745361328,
      "logps/rejected": -236.17990112304688,
      "loss": 0.1397,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.2061328887939453,
      "rewards/margins": 4.2199482917785645,
      "rewards/rejected": -2.013815402984619,
      "step": 175
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 13.973141137191872,
      "learning_rate": 2.9531763861505964e-08,
      "logits/chosen": -2.3316142559051514,
      "logits/rejected": -2.291159152984619,
      "logps/chosen": -252.21511840820312,
      "logps/rejected": -246.81796264648438,
      "loss": 0.1557,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 2.47810697555542,
      "rewards/margins": 4.51055908203125,
      "rewards/rejected": -2.032452344894409,
      "step": 180
    },
    {
      "epoch": 2.7611940298507465,
      "grad_norm": 19.964297927354572,
      "learning_rate": 1.7214919195619125e-08,
      "logits/chosen": -2.3097987174987793,
      "logits/rejected": -2.27604079246521,
      "logps/chosen": -237.06045532226562,
      "logps/rejected": -232.19808959960938,
      "loss": 0.1517,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 1.585829496383667,
      "rewards/margins": 4.088876724243164,
      "rewards/rejected": -2.503047466278076,
      "step": 185
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 21.11350475489601,
      "learning_rate": 8.161562878982398e-09,
      "logits/chosen": -2.3078198432922363,
      "logits/rejected": -2.282045602798462,
      "logps/chosen": -253.4175262451172,
      "logps/rejected": -224.1904296875,
      "loss": 0.1267,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 2.3276572227478027,
      "rewards/margins": 4.33654260635376,
      "rewards/rejected": -2.0088844299316406,
      "step": 190
    },
    {
      "epoch": 2.91044776119403,
      "grad_norm": 18.60016260779278,
      "learning_rate": 2.432892997526026e-09,
      "logits/chosen": -2.325418710708618,
      "logits/rejected": -2.2867703437805176,
      "logps/chosen": -226.9629669189453,
      "logps/rejected": -230.8715362548828,
      "loss": 0.1148,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.127265214920044,
      "rewards/margins": 4.185291290283203,
      "rewards/rejected": -2.0580265522003174,
      "step": 195
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 17.114004265650976,
      "learning_rate": 6.763371270035457e-11,
      "logits/chosen": -2.3209595680236816,
      "logits/rejected": -2.300201416015625,
      "logps/chosen": -249.88442993164062,
      "logps/rejected": -224.8131103515625,
      "loss": 0.1289,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 2.2298994064331055,
      "rewards/margins": 4.393718719482422,
      "rewards/rejected": -2.163818836212158,
      "step": 200
    },
    {
      "epoch": 2.9850746268656714,
      "eval_logits/chosen": -2.3139846324920654,
      "eval_logits/rejected": -2.28657865524292,
      "eval_logps/chosen": -253.50108337402344,
      "eval_logps/rejected": -229.3605499267578,
      "eval_loss": 0.5585500001907349,
      "eval_rewards/accuracies": 0.8125,
      "eval_rewards/chosen": 1.1188981533050537,
      "eval_rewards/margins": 2.8383777141571045,
      "eval_rewards/rejected": -1.7194799184799194,
      "eval_runtime": 125.2826,
      "eval_samples_per_second": 15.166,
      "eval_steps_per_second": 0.239,
      "step": 200
    },
    {
      "epoch": 3.0,
      "step": 201,
      "total_flos": 2369906314051584.0,
      "train_loss": 0.3258563337560317,
      "train_runtime": 7340.6492,
      "train_samples_per_second": 6.986,
      "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
}