{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9988221436984688,
  "eval_steps": 500,
  "global_step": 477,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0020939667582777124,
      "grad_norm": 4.153600215911865,
      "learning_rate": 2.083333333333333e-08,
      "logits/chosen": -0.385508269071579,
      "logits/rejected": -0.3815978169441223,
      "logps/chosen": -313.31878662109375,
      "logps/ref_chosen": -313.1247253417969,
      "logps/ref_rejected": -257.44195556640625,
      "logps/rejected": -257.5481262207031,
      "loss": 0.6932,
      "margin_dpo/margin_mean": -0.08785581588745117,
      "margin_dpo/margin_std": 0.37070298194885254,
      "step": 1
    },
    {
      "epoch": 0.020939667582777124,
      "grad_norm": 3.972473382949829,
      "learning_rate": 2.0833333333333333e-07,
      "logits/chosen": -0.4696977436542511,
      "logits/rejected": -0.4675690829753876,
      "logps/chosen": -301.630859375,
      "logps/ref_chosen": -301.46142578125,
      "logps/ref_rejected": -265.9523620605469,
      "logps/rejected": -266.04180908203125,
      "loss": 0.6933,
      "margin_dpo/margin_mean": -0.08001349866390228,
      "margin_dpo/margin_std": 0.32228612899780273,
      "step": 10
    },
    {
      "epoch": 0.04187933516555425,
      "grad_norm": 5.9276628494262695,
      "learning_rate": 4.1666666666666667e-07,
      "logits/chosen": -0.46030598878860474,
      "logits/rejected": -0.49390584230422974,
      "logps/chosen": -287.4084777832031,
      "logps/ref_chosen": -287.79296875,
      "logps/ref_rejected": -241.2596435546875,
      "logps/rejected": -241.0130157470703,
      "loss": 0.6924,
      "margin_dpo/margin_mean": 0.13785475492477417,
      "margin_dpo/margin_std": 0.439021497964859,
      "step": 20
    },
    {
      "epoch": 0.06281900274833137,
      "grad_norm": 3.928417444229126,
      "learning_rate": 4.997836020254328e-07,
      "logits/chosen": -0.4774892330169678,
      "logits/rejected": -0.48092085123062134,
      "logps/chosen": -264.83526611328125,
      "logps/ref_chosen": -267.33026123046875,
      "logps/ref_rejected": -244.6634063720703,
      "logps/rejected": -242.9745635986328,
      "loss": 0.6888,
      "margin_dpo/margin_mean": 0.8061596751213074,
      "margin_dpo/margin_std": 1.0586752891540527,
      "step": 30
    },
    {
      "epoch": 0.0837586703311085,
      "grad_norm": 4.087418079376221,
      "learning_rate": 4.984625263805177e-07,
      "logits/chosen": -0.47675904631614685,
      "logits/rejected": -0.49461251497268677,
      "logps/chosen": -283.3618469238281,
      "logps/ref_chosen": -290.2306823730469,
      "logps/ref_rejected": -270.0351257324219,
      "logps/rejected": -265.93634033203125,
      "loss": 0.6806,
      "margin_dpo/margin_mean": 2.7700228691101074,
      "margin_dpo/margin_std": 3.1351680755615234,
      "step": 40
    },
    {
      "epoch": 0.10469833791388562,
      "grad_norm": 5.013513088226318,
      "learning_rate": 4.9594693969485e-07,
      "logits/chosen": -0.46262645721435547,
      "logits/rejected": -0.4923989176750183,
      "logps/chosen": -260.5640869140625,
      "logps/ref_chosen": -266.95281982421875,
      "logps/ref_rejected": -271.7529602050781,
      "logps/rejected": -269.3885498046875,
      "loss": 0.6734,
      "margin_dpo/margin_mean": 4.024331569671631,
      "margin_dpo/margin_std": 5.195467472076416,
      "step": 50
    },
    {
      "epoch": 0.12563800549666274,
      "grad_norm": 4.223069667816162,
      "learning_rate": 4.922489359292927e-07,
      "logits/chosen": -0.47976773977279663,
      "logits/rejected": -0.4877360463142395,
      "logps/chosen": -289.54461669921875,
      "logps/ref_chosen": -288.6689758300781,
      "logps/ref_rejected": -254.53158569335938,
      "logps/rejected": -261.9698486328125,
      "loss": 0.6622,
      "margin_dpo/margin_mean": 6.56261682510376,
      "margin_dpo/margin_std": 8.391728401184082,
      "step": 60
    },
    {
      "epoch": 0.14657767307943986,
      "grad_norm": 19.633056640625,
      "learning_rate": 4.873862936454303e-07,
      "logits/chosen": -0.47945109009742737,
      "logits/rejected": -0.4913012981414795,
      "logps/chosen": -302.0457763671875,
      "logps/ref_chosen": -299.79888916015625,
      "logps/ref_rejected": -278.92218017578125,
      "logps/rejected": -291.8330078125,
      "loss": 0.6488,
      "margin_dpo/margin_mean": 10.66395092010498,
      "margin_dpo/margin_std": 12.481794357299805,
      "step": 70
    },
    {
      "epoch": 0.167517340662217,
      "grad_norm": 10.23404598236084,
      "learning_rate": 4.813823905331703e-07,
      "logits/chosen": -0.37753456830978394,
      "logits/rejected": -0.33191266655921936,
      "logps/chosen": -276.65997314453125,
      "logps/ref_chosen": -273.9964294433594,
      "logps/ref_rejected": -261.94219970703125,
      "logps/rejected": -278.61053466796875,
      "loss": 0.6335,
      "margin_dpo/margin_mean": 14.00488567352295,
      "margin_dpo/margin_std": 14.380203247070312,
      "step": 80
    },
    {
      "epoch": 0.18845700824499412,
      "grad_norm": 16.92243766784668,
      "learning_rate": 4.74266091019916e-07,
      "logits/chosen": -0.34219780564308167,
      "logits/rejected": -0.2503359913825989,
      "logps/chosen": -305.6651916503906,
      "logps/ref_chosen": -290.7003479003906,
      "logps/ref_rejected": -248.63436889648438,
      "logps/rejected": -281.1633605957031,
      "loss": 0.6279,
      "margin_dpo/margin_mean": 17.564119338989258,
      "margin_dpo/margin_std": 19.235614776611328,
      "step": 90
    },
    {
      "epoch": 0.20939667582777124,
      "grad_norm": 19.076921463012695,
      "learning_rate": 4.660716075016441e-07,
      "logits/chosen": -0.18392817676067352,
      "logits/rejected": -0.14755138754844666,
      "logps/chosen": -284.04473876953125,
      "logps/ref_chosen": -273.87774658203125,
      "logps/ref_rejected": -261.7587890625,
      "logps/rejected": -292.94903564453125,
      "loss": 0.617,
      "margin_dpo/margin_mean": 21.023273468017578,
      "margin_dpo/margin_std": 23.208118438720703,
      "step": 100
    },
    {
      "epoch": 0.23033634341054834,
      "grad_norm": 11.625693321228027,
      "learning_rate": 4.568383358630315e-07,
      "logits/chosen": -0.0882129892706871,
      "logits/rejected": 0.031399063766002655,
      "logps/chosen": -312.0311584472656,
      "logps/ref_chosen": -294.00189208984375,
      "logps/ref_rejected": -256.91058349609375,
      "logps/rejected": -298.51300048828125,
      "loss": 0.5953,
      "margin_dpo/margin_mean": 23.573087692260742,
      "margin_dpo/margin_std": 27.610767364501953,
      "step": 110
    },
    {
      "epoch": 0.25127601099332547,
      "grad_norm": 59.717994689941406,
      "learning_rate": 4.466106660773884e-07,
      "logits/chosen": -0.06424394994974136,
      "logits/rejected": 0.046889692544937134,
      "logps/chosen": -304.84332275390625,
      "logps/ref_chosen": -285.0730895996094,
      "logps/ref_rejected": -267.43023681640625,
      "logps/rejected": -312.349609375,
      "loss": 0.6073,
      "margin_dpo/margin_mean": 25.149160385131836,
      "margin_dpo/margin_std": 28.71588706970215,
      "step": 120
    },
    {
      "epoch": 0.2722156785761026,
      "grad_norm": 14.268017768859863,
      "learning_rate": 4.3543776879695804e-07,
      "logits/chosen": 0.049873873591423035,
      "logits/rejected": 0.10700122267007828,
      "logps/chosen": -326.23651123046875,
      "logps/ref_chosen": -293.5408020019531,
      "logps/ref_rejected": -288.2164001464844,
      "logps/rejected": -354.05877685546875,
      "loss": 0.5941,
      "margin_dpo/margin_mean": 33.14662170410156,
      "margin_dpo/margin_std": 31.353755950927734,
      "step": 130
    },
    {
      "epoch": 0.2931553461588797,
      "grad_norm": 42.10697937011719,
      "learning_rate": 4.233733589595746e-07,
      "logits/chosen": 0.044216498732566833,
      "logits/rejected": 0.12556368112564087,
      "logps/chosen": -332.8862609863281,
      "logps/ref_chosen": -293.3451232910156,
      "logps/ref_rejected": -279.8287658691406,
      "logps/rejected": -351.7282409667969,
      "loss": 0.597,
      "margin_dpo/margin_mean": 32.35832214355469,
      "margin_dpo/margin_std": 35.75043869018555,
      "step": 140
    },
    {
      "epoch": 0.3140950137416568,
      "grad_norm": 10.703039169311523,
      "learning_rate": 4.104754375481664e-07,
      "logits/chosen": 0.01981363818049431,
      "logits/rejected": 0.15257558226585388,
      "logps/chosen": -323.87200927734375,
      "logps/ref_chosen": -292.59295654296875,
      "logps/ref_rejected": -258.49530029296875,
      "logps/rejected": -323.07672119140625,
      "loss": 0.5844,
      "margin_dpo/margin_mean": 33.3023681640625,
      "margin_dpo/margin_std": 36.2221565246582,
      "step": 150
    },
    {
      "epoch": 0.335034681324434,
      "grad_norm": 26.41683006286621,
      "learning_rate": 3.9680601274462354e-07,
      "logits/chosen": 0.16788892447948456,
      "logits/rejected": 0.2648904025554657,
      "logps/chosen": -324.91375732421875,
      "logps/ref_chosen": -279.3182678222656,
      "logps/ref_rejected": -266.2412414550781,
      "logps/rejected": -357.54644775390625,
      "loss": 0.5572,
      "margin_dpo/margin_mean": 45.709693908691406,
      "margin_dpo/margin_std": 37.458805084228516,
      "step": 160
    },
    {
      "epoch": 0.3559743489072111,
      "grad_norm": 20.263166427612305,
      "learning_rate": 3.824308018186143e-07,
      "logits/chosen": -0.04693519324064255,
      "logits/rejected": 0.14640530943870544,
      "logps/chosen": -361.36346435546875,
      "logps/ref_chosen": -297.69024658203125,
      "logps/ref_rejected": -246.70840454101562,
      "logps/rejected": -353.06201171875,
      "loss": 0.5614,
      "margin_dpo/margin_mean": 42.680423736572266,
      "margin_dpo/margin_std": 40.208683013916016,
      "step": 170
    },
    {
      "epoch": 0.37691401648998824,
      "grad_norm": 21.05248260498047,
      "learning_rate": 3.6741891518455146e-07,
      "logits/chosen": 0.04265492409467697,
      "logits/rejected": 0.12863174080848694,
      "logps/chosen": -336.97955322265625,
      "logps/ref_chosen": -282.07220458984375,
      "logps/ref_rejected": -267.81048583984375,
      "logps/rejected": -365.7275390625,
      "loss": 0.579,
      "margin_dpo/margin_mean": 43.00971984863281,
      "margin_dpo/margin_std": 39.17428207397461,
      "step": 180
    },
    {
      "epoch": 0.39785368407276533,
      "grad_norm": 13.581904411315918,
      "learning_rate": 3.5184252414564197e-07,
      "logits/chosen": 0.019876545295119286,
      "logits/rejected": 0.14703558385372162,
      "logps/chosen": -334.26300048828125,
      "logps/ref_chosen": -277.98150634765625,
      "logps/ref_rejected": -251.03671264648438,
      "logps/rejected": -350.8268127441406,
      "loss": 0.555,
      "margin_dpo/margin_mean": 43.508628845214844,
      "margin_dpo/margin_std": 42.15697479248047,
      "step": 190
    },
    {
      "epoch": 0.4187933516555425,
      "grad_norm": 18.143381118774414,
      "learning_rate": 3.3577651392237394e-07,
      "logits/chosen": 0.07500011473894119,
      "logits/rejected": 0.24331322312355042,
      "logps/chosen": -381.325927734375,
      "logps/ref_chosen": -301.87799072265625,
      "logps/ref_rejected": -271.30010986328125,
      "logps/rejected": -394.4322814941406,
      "loss": 0.5642,
      "margin_dpo/margin_mean": 43.68426513671875,
      "margin_dpo/margin_std": 38.98915100097656,
      "step": 200
    },
    {
      "epoch": 0.4397330192383196,
      "grad_norm": 24.03816032409668,
      "learning_rate": 3.1929812363354764e-07,
      "logits/chosen": 0.11765547096729279,
      "logits/rejected": 0.18735218048095703,
      "logps/chosen": -365.60113525390625,
      "logps/ref_chosen": -271.31854248046875,
      "logps/ref_rejected": -255.88186645507812,
      "logps/rejected": -395.07745361328125,
      "loss": 0.5552,
      "margin_dpo/margin_mean": 44.91301345825195,
      "margin_dpo/margin_std": 47.19767379760742,
      "step": 210
    },
    {
      "epoch": 0.4606726868210967,
      "grad_norm": 22.125686645507812,
      "learning_rate": 3.024865749606803e-07,
      "logits/chosen": 0.07137051969766617,
      "logits/rejected": 0.18482640385627747,
      "logps/chosen": -391.6545104980469,
      "logps/ref_chosen": -293.5024108886719,
      "logps/ref_rejected": -263.943115234375,
      "logps/rejected": -409.07977294921875,
      "loss": 0.5519,
      "margin_dpo/margin_mean": 46.98453140258789,
      "margin_dpo/margin_std": 43.669212341308594,
      "step": 220
    },
    {
      "epoch": 0.48161235440387384,
      "grad_norm": 18.157793045043945,
      "learning_rate": 2.854226912810206e-07,
      "logits/chosen": 0.11615083366632462,
      "logits/rejected": 0.24443212151527405,
      "logps/chosen": -374.0119323730469,
      "logps/ref_chosen": -265.1332702636719,
      "logps/ref_rejected": -261.63671875,
      "logps/rejected": -415.58380126953125,
      "loss": 0.5424,
      "margin_dpo/margin_mean": 45.06837844848633,
      "margin_dpo/margin_std": 45.513736724853516,
      "step": 230
    },
    {
      "epoch": 0.5025520219866509,
      "grad_norm": 10.94947624206543,
      "learning_rate": 2.68188509100236e-07,
      "logits/chosen": 0.07515954971313477,
      "logits/rejected": 0.18595214188098907,
      "logps/chosen": -355.45574951171875,
      "logps/ref_chosen": -265.394287109375,
      "logps/ref_rejected": -263.04449462890625,
      "logps/rejected": -400.0995788574219,
      "loss": 0.5557,
      "margin_dpo/margin_mean": 46.99358367919922,
      "margin_dpo/margin_std": 43.50702667236328,
      "step": 240
    },
    {
      "epoch": 0.523491689569428,
      "grad_norm": 26.13774299621582,
      "learning_rate": 2.5086688365284857e-07,
      "logits/chosen": 0.057459376752376556,
      "logits/rejected": 0.23900671303272247,
      "logps/chosen": -364.48931884765625,
      "logps/ref_chosen": -276.13116455078125,
      "logps/ref_rejected": -233.1983642578125,
      "logps/rejected": -372.2765197753906,
      "loss": 0.5488,
      "margin_dpo/margin_mean": 50.719993591308594,
      "margin_dpo/margin_std": 46.66020965576172,
      "step": 250
    },
    {
      "epoch": 0.5444313571522053,
      "grad_norm": 14.217330932617188,
      "learning_rate": 2.3354109056654418e-07,
      "logits/chosen": 0.1635710746049881,
      "logits/rejected": 0.27944889664649963,
      "logps/chosen": -396.4654846191406,
      "logps/ref_chosen": -289.42364501953125,
      "logps/ref_rejected": -255.2952117919922,
      "logps/rejected": -411.4334411621094,
      "loss": 0.5539,
      "margin_dpo/margin_mean": 49.09638595581055,
      "margin_dpo/margin_std": 50.09044647216797,
      "step": 260
    },
    {
      "epoch": 0.5653710247349824,
      "grad_norm": 47.02643585205078,
      "learning_rate": 2.162944255053928e-07,
      "logits/chosen": 0.23213577270507812,
      "logits/rejected": 0.2836950123310089,
      "logps/chosen": -406.9697265625,
      "logps/ref_chosen": -295.31121826171875,
      "logps/ref_rejected": -283.03643798828125,
      "logps/rejected": -445.3810119628906,
      "loss": 0.537,
      "margin_dpo/margin_mean": 50.686119079589844,
      "margin_dpo/margin_std": 50.13835906982422,
      "step": 270
    },
    {
      "epoch": 0.5863106923177595,
      "grad_norm": 16.077573776245117,
      "learning_rate": 1.9920980371674297e-07,
      "logits/chosen": 0.04740264266729355,
      "logits/rejected": 0.20282933115959167,
      "logps/chosen": -392.91693115234375,
      "logps/ref_chosen": -277.9930725097656,
      "logps/ref_rejected": -258.11956787109375,
      "logps/rejected": -436.17108154296875,
      "loss": 0.53,
      "margin_dpo/margin_mean": 63.12766647338867,
      "margin_dpo/margin_std": 48.20869827270508,
      "step": 280
    },
    {
      "epoch": 0.6072503599005366,
      "grad_norm": 30.580135345458984,
      "learning_rate": 1.823693614070116e-07,
      "logits/chosen": 0.1650705635547638,
      "logits/rejected": 0.3033481240272522,
      "logps/chosen": -422.15130615234375,
      "logps/ref_chosen": -295.2901306152344,
      "logps/ref_rejected": -269.10552978515625,
      "logps/rejected": -446.4153747558594,
      "loss": 0.5597,
      "margin_dpo/margin_mean": 50.44862747192383,
      "margin_dpo/margin_std": 48.35343933105469,
      "step": 290
    },
    {
      "epoch": 0.6281900274833137,
      "grad_norm": 46.89568328857422,
      "learning_rate": 1.6585406086279846e-07,
      "logits/chosen": 0.2200140655040741,
      "logits/rejected": 0.34437426924705505,
      "logps/chosen": -378.20428466796875,
      "logps/ref_chosen": -261.69488525390625,
      "logps/ref_rejected": -253.68557739257812,
      "logps/rejected": -420.6162109375,
      "loss": 0.5471,
      "margin_dpo/margin_mean": 50.421165466308594,
      "margin_dpo/margin_std": 47.259437561035156,
      "step": 300
    },
    {
      "epoch": 0.6491296950660909,
      "grad_norm": 16.970218658447266,
      "learning_rate": 1.4974330121575046e-07,
      "logits/chosen": 0.2641686797142029,
      "logits/rejected": 0.383655846118927,
      "logps/chosen": -414.44512939453125,
      "logps/ref_chosen": -285.978515625,
      "logps/ref_rejected": -262.2162170410156,
      "logps/rejected": -446.66815185546875,
      "loss": 0.5278,
      "margin_dpo/margin_mean": 55.9853515625,
      "margin_dpo/margin_std": 45.96001434326172,
      "step": 310
    },
    {
      "epoch": 0.670069362648868,
      "grad_norm": 16.45661735534668,
      "learning_rate": 1.341145367224657e-07,
      "logits/chosen": 0.14594177901744843,
      "logits/rejected": 0.3130393326282501,
      "logps/chosen": -418.21820068359375,
      "logps/ref_chosen": -297.89361572265625,
      "logps/ref_rejected": -261.50360107421875,
      "logps/rejected": -444.85150146484375,
      "loss": 0.5262,
      "margin_dpo/margin_mean": 63.023284912109375,
      "margin_dpo/margin_std": 49.52865219116211,
      "step": 320
    },
    {
      "epoch": 0.6910090302316451,
      "grad_norm": 52.76103210449219,
      "learning_rate": 1.1904290439459971e-07,
      "logits/chosen": 0.14333733916282654,
      "logits/rejected": 0.2348686158657074,
      "logps/chosen": -395.7355651855469,
      "logps/ref_chosen": -282.3043518066406,
      "logps/ref_rejected": -272.1080322265625,
      "logps/rejected": -451.24932861328125,
      "loss": 0.5455,
      "margin_dpo/margin_mean": 65.71006774902344,
      "margin_dpo/margin_std": 54.01335525512695,
      "step": 330
    },
    {
      "epoch": 0.7119486978144222,
      "grad_norm": 26.729217529296875,
      "learning_rate": 1.0460086276938143e-07,
      "logits/chosen": 0.1563154011964798,
      "logits/rejected": 0.2881699502468109,
      "logps/chosen": -378.7862854003906,
      "logps/ref_chosen": -264.32440185546875,
      "logps/ref_rejected": -231.1798095703125,
      "logps/rejected": -404.14886474609375,
      "loss": 0.5356,
      "margin_dpo/margin_mean": 58.50718307495117,
      "margin_dpo/margin_std": 58.1771125793457,
      "step": 340
    },
    {
      "epoch": 0.7328883653971994,
      "grad_norm": 31.903892517089844,
      "learning_rate": 9.085784355719256e-08,
      "logits/chosen": 0.10641894489526749,
      "logits/rejected": 0.2847670614719391,
      "logps/chosen": -425.1806640625,
      "logps/ref_chosen": -309.6944274902344,
      "logps/ref_rejected": -275.2430114746094,
      "logps/rejected": -448.0032653808594,
      "loss": 0.5388,
      "margin_dpo/margin_mean": 57.274009704589844,
      "margin_dpo/margin_std": 47.37055587768555,
      "step": 350
    },
    {
      "epoch": 0.7538280329799765,
      "grad_norm": 29.521696090698242,
      "learning_rate": 7.787991784094999e-08,
      "logits/chosen": 0.15716208517551422,
      "logits/rejected": 0.26787060499191284,
      "logps/chosen": -378.3572998046875,
      "logps/ref_chosen": -256.606201171875,
      "logps/ref_rejected": -244.90805053710938,
      "logps/rejected": -420.41864013671875,
      "loss": 0.5474,
      "margin_dpo/margin_mean": 53.7595100402832,
      "margin_dpo/margin_std": 58.170265197753906,
      "step": 360
    },
    {
      "epoch": 0.7747677005627536,
      "grad_norm": 19.0128231048584,
      "learning_rate": 6.572947843207677e-08,
      "logits/chosen": 0.16855831444263458,
      "logits/rejected": 0.3704710900783539,
      "logps/chosen": -413.5511169433594,
      "logps/ref_chosen": -297.1283264160156,
      "logps/ref_rejected": -262.2713317871094,
      "logps/rejected": -444.85626220703125,
      "loss": 0.5201,
      "margin_dpo/margin_mean": 66.16206359863281,
      "margin_dpo/margin_std": 55.746543884277344,
      "step": 370
    },
    {
      "epoch": 0.7957073681455307,
      "grad_norm": 23.740262985229492,
      "learning_rate": 5.446493991016879e-08,
      "logits/chosen": 0.18822483718395233,
      "logits/rejected": 0.29204973578453064,
      "logps/chosen": -412.62615966796875,
      "logps/ref_chosen": -291.81365966796875,
      "logps/ref_rejected": -264.4927673339844,
      "logps/rejected": -443.88836669921875,
      "loss": 0.5438,
      "margin_dpo/margin_mean": 58.58308029174805,
      "margin_dpo/margin_std": 47.825599670410156,
      "step": 380
    },
    {
      "epoch": 0.8166470357283078,
      "grad_norm": 14.888653755187988,
      "learning_rate": 4.4140457788451434e-08,
      "logits/chosen": 0.1765083372592926,
      "logits/rejected": 0.3199649751186371,
      "logps/chosen": -421.967529296875,
      "logps/ref_chosen": -289.7127990722656,
      "logps/ref_rejected": -267.163330078125,
      "logps/rejected": -455.80517578125,
      "loss": 0.5362,
      "margin_dpo/margin_mean": 56.387115478515625,
      "margin_dpo/margin_std": 51.367897033691406,
      "step": 390
    },
    {
      "epoch": 0.837586703311085,
      "grad_norm": 27.52701759338379,
      "learning_rate": 3.4805668155167556e-08,
      "logits/chosen": 0.23016759753227234,
      "logits/rejected": 0.414517879486084,
      "logps/chosen": -432.7222595214844,
      "logps/ref_chosen": -298.2846984863281,
      "logps/ref_rejected": -244.9308624267578,
      "logps/rejected": -438.0635681152344,
      "loss": 0.5389,
      "margin_dpo/margin_mean": 58.69511795043945,
      "margin_dpo/margin_std": 53.363685607910156,
      "step": 400
    },
    {
      "epoch": 0.8585263708938621,
      "grad_norm": 31.697511672973633,
      "learning_rate": 2.650544904260024e-08,
      "logits/chosen": 0.18534260988235474,
      "logits/rejected": 0.3869909346103668,
      "logps/chosen": -443.33575439453125,
      "logps/ref_chosen": -308.6795959472656,
      "logps/ref_rejected": -275.4808349609375,
      "logps/rejected": -468.99200439453125,
      "loss": 0.5198,
      "margin_dpo/margin_mean": 58.855003356933594,
      "margin_dpo/margin_std": 53.85388946533203,
      "step": 410
    },
    {
      "epoch": 0.8794660384766392,
      "grad_norm": 21.451770782470703,
      "learning_rate": 1.9279704670975726e-08,
      "logits/chosen": 0.10304410755634308,
      "logits/rejected": 0.2722802758216858,
      "logps/chosen": -427.65203857421875,
      "logps/ref_chosen": -291.2632141113281,
      "logps/ref_rejected": -276.36553955078125,
      "logps/rejected": -470.96038818359375,
      "loss": 0.5276,
      "margin_dpo/margin_mean": 58.20595169067383,
      "margin_dpo/margin_std": 57.847412109375,
      "step": 420
    },
    {
      "epoch": 0.9004057060594163,
      "grad_norm": 27.079269409179688,
      "learning_rate": 1.3163173604516409e-08,
      "logits/chosen": 0.2293200045824051,
      "logits/rejected": 0.399575412273407,
      "logps/chosen": -426.8741760253906,
      "logps/ref_chosen": -296.31085205078125,
      "logps/ref_rejected": -259.3516845703125,
      "logps/rejected": -452.6231384277344,
      "loss": 0.5185,
      "margin_dpo/margin_mean": 62.70819091796875,
      "margin_dpo/margin_std": 54.348899841308594,
      "step": 430
    },
    {
      "epoch": 0.9213453736421934,
      "grad_norm": 26.594852447509766,
      "learning_rate": 8.18526174195655e-09,
      "logits/chosen": 0.277737557888031,
      "logits/rejected": 0.3488383889198303,
      "logps/chosen": -430.6412658691406,
      "logps/ref_chosen": -290.34857177734375,
      "logps/ref_rejected": -267.0846862792969,
      "logps/rejected": -453.26580810546875,
      "loss": 0.5468,
      "margin_dpo/margin_mean": 45.888404846191406,
      "margin_dpo/margin_std": 55.84312057495117,
      "step": 440
    },
    {
      "epoch": 0.9422850412249706,
      "grad_norm": 17.050395965576172,
      "learning_rate": 4.369900944435734e-09,
      "logits/chosen": 0.14949986338615417,
      "logits/rejected": 0.3868565857410431,
      "logps/chosen": -442.13580322265625,
      "logps/ref_chosen": -316.7445068359375,
      "logps/ref_rejected": -255.24594116210938,
      "logps/rejected": -452.7721252441406,
      "loss": 0.5221,
      "margin_dpo/margin_mean": 72.13482666015625,
      "margin_dpo/margin_std": 50.0427360534668,
      "step": 450
    },
    {
      "epoch": 0.9632247088077477,
      "grad_norm": 24.54353141784668,
      "learning_rate": 1.73543398043291e-09,
      "logits/chosen": 0.1670341193675995,
      "logits/rejected": 0.3386840224266052,
      "logps/chosen": -449.431884765625,
      "logps/ref_chosen": -316.5346984863281,
      "logps/ref_rejected": -293.4228820800781,
      "logps/rejected": -487.91412353515625,
      "loss": 0.5389,
      "margin_dpo/margin_mean": 61.59406280517578,
      "margin_dpo/margin_std": 53.5306282043457,
      "step": 460
    },
    {
      "epoch": 0.9841643763905248,
      "grad_norm": 23.415748596191406,
      "learning_rate": 2.94526340880813e-10,
      "logits/chosen": 0.23086392879486084,
      "logits/rejected": 0.39752504229545593,
      "logps/chosen": -410.21978759765625,
      "logps/ref_chosen": -284.00579833984375,
      "logps/ref_rejected": -252.9314727783203,
      "logps/rejected": -446.0233459472656,
      "loss": 0.5346,
      "margin_dpo/margin_mean": 66.87791442871094,
      "margin_dpo/margin_std": 52.97399139404297,
      "step": 470
    },
    {
      "epoch": 0.9988221436984688,
      "step": 477,
      "total_flos": 0.0,
      "train_loss": 0.5733983517942689,
      "train_runtime": 5703.1485,
      "train_samples_per_second": 10.72,
      "train_steps_per_second": 0.084
    }
  ],
  "logging_steps": 10,
  "max_steps": 477,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 250,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}