{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9988571428571429,
  "eval_steps": 100,
  "global_step": 437,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.022857142857142857,
      "grad_norm": 9.822878263417993,
      "learning_rate": 1.1363636363636363e-07,
      "logits/chosen": -2.700991630554199,
      "logits/rejected": -2.625051975250244,
      "logps/chosen": -301.26373291015625,
      "logps/rejected": -281.7487487792969,
      "loss": 0.693,
      "rewards/accuracies": 0.40625,
      "rewards/chosen": 0.00022422037727665156,
      "rewards/margins": 0.0002149858046323061,
      "rewards/rejected": 9.234552635462023e-06,
      "step": 10
    },
    {
      "epoch": 0.045714285714285714,
      "grad_norm": 7.790697115653487,
      "learning_rate": 2.2727272727272726e-07,
      "logits/chosen": -2.6405527591705322,
      "logits/rejected": -2.6053428649902344,
      "logps/chosen": -278.93023681640625,
      "logps/rejected": -254.70883178710938,
      "loss": 0.6923,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.002409698674455285,
      "rewards/margins": 0.0018069592770189047,
      "rewards/rejected": 0.0006027388735674322,
      "step": 20
    },
    {
      "epoch": 0.06857142857142857,
      "grad_norm": 8.37723659427994,
      "learning_rate": 3.4090909090909085e-07,
      "logits/chosen": -2.637849807739258,
      "logits/rejected": -2.616870403289795,
      "logps/chosen": -263.4436950683594,
      "logps/rejected": -263.60205078125,
      "loss": 0.6871,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.012393913231790066,
      "rewards/margins": 0.00848361849784851,
      "rewards/rejected": 0.003910294268280268,
      "step": 30
    },
    {
      "epoch": 0.09142857142857143,
      "grad_norm": 9.39434892949384,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -2.64766263961792,
      "logits/rejected": -2.585386276245117,
      "logps/chosen": -290.52587890625,
      "logps/rejected": -268.34405517578125,
      "loss": 0.6756,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.03471367806196213,
      "rewards/margins": 0.041840970516204834,
      "rewards/rejected": -0.007127293851226568,
      "step": 40
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 13.922334228887747,
      "learning_rate": 4.997124959943201e-07,
      "logits/chosen": -2.676335573196411,
      "logits/rejected": -2.5971837043762207,
      "logps/chosen": -294.0897216796875,
      "logps/rejected": -253.9338836669922,
      "loss": 0.6628,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.022205885499715805,
      "rewards/margins": 0.09290142357349396,
      "rewards/rejected": -0.07069554179906845,
      "step": 50
    },
    {
      "epoch": 0.13714285714285715,
      "grad_norm": 11.361649612146103,
      "learning_rate": 4.979579212164186e-07,
      "logits/chosen": -2.5759811401367188,
      "logits/rejected": -2.472777843475342,
      "logps/chosen": -290.1505432128906,
      "logps/rejected": -270.30242919921875,
      "loss": 0.6374,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -0.09644584357738495,
      "rewards/margins": 0.1210569366812706,
      "rewards/rejected": -0.21750275790691376,
      "step": 60
    },
    {
      "epoch": 0.16,
      "grad_norm": 14.922787112175543,
      "learning_rate": 4.946196886175515e-07,
      "logits/chosen": -2.5715270042419434,
      "logits/rejected": -2.5131115913391113,
      "logps/chosen": -284.8800964355469,
      "logps/rejected": -291.2018737792969,
      "loss": 0.6102,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.09101851284503937,
      "rewards/margins": 0.22473247349262238,
      "rewards/rejected": -0.31575098633766174,
      "step": 70
    },
    {
      "epoch": 0.18285714285714286,
      "grad_norm": 11.602482470889782,
      "learning_rate": 4.897191188239667e-07,
      "logits/chosen": -2.5607805252075195,
      "logits/rejected": -2.50270938873291,
      "logps/chosen": -291.697998046875,
      "logps/rejected": -307.26983642578125,
      "loss": 0.6078,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.2466896027326584,
      "rewards/margins": 0.27291998267173767,
      "rewards/rejected": -0.5196096301078796,
      "step": 80
    },
    {
      "epoch": 0.2057142857142857,
      "grad_norm": 17.533063770924706,
      "learning_rate": 4.832875107981763e-07,
      "logits/chosen": -2.6349377632141113,
      "logits/rejected": -2.5697710514068604,
      "logps/chosen": -295.2223815917969,
      "logps/rejected": -310.2930603027344,
      "loss": 0.6076,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.18937139213085175,
      "rewards/margins": 0.3580796718597412,
      "rewards/rejected": -0.5474510192871094,
      "step": 90
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 16.194420237420413,
      "learning_rate": 4.753659419387223e-07,
      "logits/chosen": -2.6082775592803955,
      "logits/rejected": -2.501986026763916,
      "logps/chosen": -317.41748046875,
      "logps/rejected": -301.41156005859375,
      "loss": 0.5982,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.31031134724617004,
      "rewards/margins": 0.3692900538444519,
      "rewards/rejected": -0.6796014904975891,
      "step": 100
    },
    {
      "epoch": 0.22857142857142856,
      "eval_logits/chosen": -2.428199529647827,
      "eval_logits/rejected": -2.290175437927246,
      "eval_logps/chosen": -292.6412048339844,
      "eval_logps/rejected": -277.6054382324219,
      "eval_loss": 0.5954251885414124,
      "eval_rewards/accuracies": 0.7155172228813171,
      "eval_rewards/chosen": -0.17037244141101837,
      "eval_rewards/margins": 0.41492074728012085,
      "eval_rewards/rejected": -0.5852932333946228,
      "eval_runtime": 92.2211,
      "eval_samples_per_second": 19.854,
      "eval_steps_per_second": 0.314,
      "step": 100
    },
    {
      "epoch": 0.25142857142857145,
      "grad_norm": 17.063109893518277,
      "learning_rate": 4.660050057270191e-07,
      "logits/chosen": -2.2657978534698486,
      "logits/rejected": -2.190037965774536,
      "logps/chosen": -358.4273376464844,
      "logps/rejected": -371.21563720703125,
      "loss": 0.591,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": -0.4212372303009033,
      "rewards/margins": 0.31182247400283813,
      "rewards/rejected": -0.7330597639083862,
      "step": 110
    },
    {
      "epoch": 0.2742857142857143,
      "grad_norm": 18.104917147072054,
      "learning_rate": 4.5526448859687144e-07,
      "logits/chosen": -1.961683988571167,
      "logits/rejected": -1.7426531314849854,
      "logps/chosen": -373.987060546875,
      "logps/rejected": -336.79095458984375,
      "loss": 0.5711,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.6308258175849915,
      "rewards/margins": 0.4199417233467102,
      "rewards/rejected": -1.0507675409317017,
      "step": 120
    },
    {
      "epoch": 0.29714285714285715,
      "grad_norm": 23.635801722109512,
      "learning_rate": 4.432129880904388e-07,
      "logits/chosen": -1.0558098554611206,
      "logits/rejected": -0.7218537330627441,
      "logps/chosen": -395.3338317871094,
      "logps/rejected": -383.4067077636719,
      "loss": 0.551,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": -0.901779055595398,
      "rewards/margins": 0.42758530378341675,
      "rewards/rejected": -1.3293644189834595,
      "step": 130
    },
    {
      "epoch": 0.32,
      "grad_norm": 23.352536101734486,
      "learning_rate": 4.299274747394055e-07,
      "logits/chosen": -0.8035767674446106,
      "logits/rejected": -0.4614217281341553,
      "logps/chosen": -391.3183288574219,
      "logps/rejected": -408.2334289550781,
      "loss": 0.5542,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.7587070465087891,
      "rewards/margins": 0.6045126914978027,
      "rewards/rejected": -1.3632197380065918,
      "step": 140
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 20.925582545841166,
      "learning_rate": 4.1549280046953653e-07,
      "logits/chosen": -0.7857555150985718,
      "logits/rejected": -0.3248134255409241,
      "logps/chosen": -371.37091064453125,
      "logps/rejected": -420.19720458984375,
      "loss": 0.5294,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.8438502550125122,
      "rewards/margins": 0.5951322317123413,
      "rewards/rejected": -1.438982367515564,
      "step": 150
    },
    {
      "epoch": 0.3657142857142857,
      "grad_norm": 24.471284191884987,
      "learning_rate": 4.000011566683401e-07,
      "logits/chosen": -0.6328128576278687,
      "logits/rejected": -0.07759309560060501,
      "logps/chosen": -406.115966796875,
      "logps/rejected": -437.8179626464844,
      "loss": 0.5424,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -1.0677564144134521,
      "rewards/margins": 0.7539954781532288,
      "rewards/rejected": -1.8217518329620361,
      "step": 160
    },
    {
      "epoch": 0.38857142857142857,
      "grad_norm": 24.225836916738423,
      "learning_rate": 3.8355148537705047e-07,
      "logits/chosen": -1.5064557790756226,
      "logits/rejected": -1.0156046152114868,
      "logps/chosen": -380.225830078125,
      "logps/rejected": -391.9797668457031,
      "loss": 0.5359,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.7612096667289734,
      "rewards/margins": 0.5012774467468262,
      "rewards/rejected": -1.2624871730804443,
      "step": 170
    },
    {
      "epoch": 0.4114285714285714,
      "grad_norm": 21.58523060279369,
      "learning_rate": 3.662488473675315e-07,
      "logits/chosen": -1.3468199968338013,
      "logits/rejected": -0.5501200556755066,
      "logps/chosen": -415.23846435546875,
      "logps/rejected": -432.3148498535156,
      "loss": 0.5511,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": -0.8505870699882507,
      "rewards/margins": 0.873055636882782,
      "rewards/rejected": -1.7236427068710327,
      "step": 180
    },
    {
      "epoch": 0.4342857142857143,
      "grad_norm": 22.105812847980097,
      "learning_rate": 3.48203751140067e-07,
      "logits/chosen": -1.2279399633407593,
      "logits/rejected": -0.5397592782974243,
      "logps/chosen": -367.1853942871094,
      "logps/rejected": -383.40606689453125,
      "loss": 0.5454,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.9092782735824585,
      "rewards/margins": 0.5798004269599915,
      "rewards/rejected": -1.4890786409378052,
      "step": 190
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 18.556629832547383,
      "learning_rate": 3.2953144712759537e-07,
      "logits/chosen": -1.407790184020996,
      "logits/rejected": -0.6917704343795776,
      "logps/chosen": -349.32354736328125,
      "logps/rejected": -378.60113525390625,
      "loss": 0.5439,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.8098081350326538,
      "rewards/margins": 0.687397301197052,
      "rewards/rejected": -1.4972054958343506,
      "step": 200
    },
    {
      "epoch": 0.45714285714285713,
      "eval_logits/chosen": -1.037967562675476,
      "eval_logits/rejected": -0.04150526225566864,
      "eval_logps/chosen": -356.8392333984375,
      "eval_logps/rejected": -386.5412902832031,
      "eval_loss": 0.5353143215179443,
      "eval_rewards/accuracies": 0.7586206793785095,
      "eval_rewards/chosen": -0.8123530149459839,
      "eval_rewards/margins": 0.8622987270355225,
      "eval_rewards/rejected": -1.674651861190796,
      "eval_runtime": 90.9527,
      "eval_samples_per_second": 20.131,
      "eval_steps_per_second": 0.319,
      "step": 200
    },
    {
      "epoch": 0.48,
      "grad_norm": 20.355837183026274,
      "learning_rate": 3.103511916141658e-07,
      "logits/chosen": -0.7288470268249512,
      "logits/rejected": -0.05566452816128731,
      "logps/chosen": -357.4866943359375,
      "logps/rejected": -403.9605407714844,
      "loss": 0.5321,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.9403315782546997,
      "rewards/margins": 0.6464741230010986,
      "rewards/rejected": -1.5868055820465088,
      "step": 210
    },
    {
      "epoch": 0.5028571428571429,
      "grad_norm": 22.193503671674048,
      "learning_rate": 2.9078548506882117e-07,
      "logits/chosen": -0.47817331552505493,
      "logits/rejected": 0.3535853326320648,
      "logps/chosen": -388.7688293457031,
      "logps/rejected": -414.6498107910156,
      "loss": 0.5487,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.0370280742645264,
      "rewards/margins": 0.6476354002952576,
      "rewards/rejected": -1.6846634149551392,
      "step": 220
    },
    {
      "epoch": 0.5257142857142857,
      "grad_norm": 20.666846900072,
      "learning_rate": 2.709592897595191e-07,
      "logits/chosen": -0.1568877398967743,
      "logits/rejected": 0.7903480529785156,
      "logps/chosen": -379.3858642578125,
      "logps/rejected": -402.08740234375,
      "loss": 0.531,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.983267605304718,
      "rewards/margins": 0.6447398662567139,
      "rewards/rejected": -1.6280075311660767,
      "step": 230
    },
    {
      "epoch": 0.5485714285714286,
      "grad_norm": 26.65066022502421,
      "learning_rate": 2.509992316440332e-07,
      "logits/chosen": -0.4197590947151184,
      "logits/rejected": 0.6013392210006714,
      "logps/chosen": -385.48638916015625,
      "logps/rejected": -448.73175048828125,
      "loss": 0.5245,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.8677653074264526,
      "rewards/margins": 0.9053336977958679,
      "rewards/rejected": -1.7730991840362549,
      "step": 240
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 21.858480788312963,
      "learning_rate": 2.3103279163519918e-07,
      "logits/chosen": -0.17110395431518555,
      "logits/rejected": 0.49592718482017517,
      "logps/chosen": -368.26263427734375,
      "logps/rejected": -430.48468017578125,
      "loss": 0.5351,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.8763637542724609,
      "rewards/margins": 0.7264786958694458,
      "rewards/rejected": -1.6028425693511963,
      "step": 250
    },
    {
      "epoch": 0.5942857142857143,
      "grad_norm": 23.021393487481784,
      "learning_rate": 2.1118749140573358e-07,
      "logits/chosen": -0.3194964528083801,
      "logits/rejected": 0.2826068699359894,
      "logps/chosen": -373.82830810546875,
      "logps/rejected": -423.619384765625,
      "loss": 0.5387,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.9404159784317017,
      "rewards/margins": 0.5623483657836914,
      "rewards/rejected": -1.502764344215393,
      "step": 260
    },
    {
      "epoch": 0.6171428571428571,
      "grad_norm": 28.56553316601309,
      "learning_rate": 1.9159007893272703e-07,
      "logits/chosen": 0.10294933617115021,
      "logits/rejected": 1.3178448677062988,
      "logps/chosen": -370.7629089355469,
      "logps/rejected": -405.086181640625,
      "loss": 0.5159,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.9823705554008484,
      "rewards/margins": 0.7634164690971375,
      "rewards/rejected": -1.7457869052886963,
      "step": 270
    },
    {
      "epoch": 0.64,
      "grad_norm": 24.30460588422491,
      "learning_rate": 1.7236571898357766e-07,
      "logits/chosen": 0.1169511079788208,
      "logits/rejected": 1.0327664613723755,
      "logps/chosen": -367.6756591796875,
      "logps/rejected": -442.3148498535156,
      "loss": 0.5251,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.9463755488395691,
      "rewards/margins": 0.8457048535346985,
      "rewards/rejected": -1.792080283164978,
      "step": 280
    },
    {
      "epoch": 0.6628571428571428,
      "grad_norm": 28.120774422152078,
      "learning_rate": 1.5363719371356882e-07,
      "logits/chosen": 0.10812599956989288,
      "logits/rejected": 1.0026741027832031,
      "logps/chosen": -394.53759765625,
      "logps/rejected": -435.6455078125,
      "loss": 0.5215,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.9363173246383667,
      "rewards/margins": 0.764653205871582,
      "rewards/rejected": -1.7009706497192383,
      "step": 290
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 20.243539038486684,
      "learning_rate": 1.3552411848071565e-07,
      "logits/chosen": -0.2318529188632965,
      "logits/rejected": 1.2198255062103271,
      "logps/chosen": -389.1872863769531,
      "logps/rejected": -424.034912109375,
      "loss": 0.5109,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.8727809190750122,
      "rewards/margins": 0.8411470651626587,
      "rewards/rejected": -1.713927984237671,
      "step": 300
    },
    {
      "epoch": 0.6857142857142857,
      "eval_logits/chosen": -0.3942234218120575,
      "eval_logits/rejected": 1.1016443967819214,
      "eval_logps/chosen": -345.452880859375,
      "eval_logps/rejected": -383.7008361816406,
      "eval_loss": 0.5189641118049622,
      "eval_rewards/accuracies": 0.7715517282485962,
      "eval_rewards/chosen": -0.6984895467758179,
      "eval_rewards/margins": 0.9477577805519104,
      "eval_rewards/rejected": -1.6462475061416626,
      "eval_runtime": 91.163,
      "eval_samples_per_second": 20.085,
      "eval_steps_per_second": 0.318,
      "step": 300
    },
    {
      "epoch": 0.7085714285714285,
      "grad_norm": 21.712761933059525,
      "learning_rate": 1.1814217788631473e-07,
      "logits/chosen": -0.18514932692050934,
      "logits/rejected": 0.6901147961616516,
      "logps/chosen": -342.0721740722656,
      "logps/rejected": -390.0445861816406,
      "loss": 0.5324,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.8576802015304565,
      "rewards/margins": 0.6708122491836548,
      "rewards/rejected": -1.5284925699234009,
      "step": 310
    },
    {
      "epoch": 0.7314285714285714,
      "grad_norm": 22.89579586434415,
      "learning_rate": 1.0160238692045331e-07,
      "logits/chosen": 0.14465376734733582,
      "logits/rejected": 0.9812980890274048,
      "logps/chosen": -342.1743469238281,
      "logps/rejected": -399.4386901855469,
      "loss": 0.5259,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.9789196848869324,
      "rewards/margins": 0.6055852174758911,
      "rewards/rejected": -1.5845048427581787,
      "step": 320
    },
    {
      "epoch": 0.7542857142857143,
      "grad_norm": 21.156935890608,
      "learning_rate": 8.601038193139438e-08,
      "logits/chosen": -0.23074492812156677,
      "logits/rejected": 0.9273271560668945,
      "logps/chosen": -383.02691650390625,
      "logps/rejected": -410.26519775390625,
      "loss": 0.5203,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": -0.8380874395370483,
      "rewards/margins": 0.7866854071617126,
      "rewards/rejected": -1.6247727870941162,
      "step": 330
    },
    {
      "epoch": 0.7771428571428571,
      "grad_norm": 21.402992880624122,
      "learning_rate": 7.146574594727572e-08,
      "logits/chosen": 0.4683713912963867,
      "logits/rejected": 1.2558484077453613,
      "logps/chosen": -365.50213623046875,
      "logps/rejected": -427.18682861328125,
      "loss": 0.5174,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.9955114126205444,
      "rewards/margins": 0.8561191558837891,
      "rewards/rejected": -1.8516305685043335,
      "step": 340
    },
    {
      "epoch": 0.8,
      "grad_norm": 23.419375804789023,
      "learning_rate": 5.8061372659157306e-08,
      "logits/chosen": -0.03433036804199219,
      "logits/rejected": 1.043505072593689,
      "logps/chosen": -382.8368225097656,
      "logps/rejected": -416.30145263671875,
      "loss": 0.5252,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.9312959909439087,
      "rewards/margins": 0.6887425780296326,
      "rewards/rejected": -1.620038628578186,
      "step": 350
    },
    {
      "epoch": 0.8228571428571428,
      "grad_norm": 21.8208411590094,
      "learning_rate": 4.5882873127531614e-08,
      "logits/chosen": -0.20763058960437775,
      "logits/rejected": 1.1433674097061157,
      "logps/chosen": -370.57281494140625,
      "logps/rejected": -423.702392578125,
      "loss": 0.5099,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.8516775369644165,
      "rewards/margins": 0.8744813799858093,
      "rewards/rejected": -1.726158857345581,
      "step": 360
    },
    {
      "epoch": 0.8457142857142858,
      "grad_norm": 21.479692671582267,
      "learning_rate": 3.500802900154412e-08,
      "logits/chosen": -0.02174173854291439,
      "logits/rejected": 1.3574512004852295,
      "logps/chosen": -351.3383483886719,
      "logps/rejected": -407.7884826660156,
      "loss": 0.515,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.8027693033218384,
      "rewards/margins": 0.8909598588943481,
      "rewards/rejected": -1.6937291622161865,
      "step": 370
    },
    {
      "epoch": 0.8685714285714285,
      "grad_norm": 23.836354879990434,
      "learning_rate": 2.550629574310309e-08,
      "logits/chosen": -0.2868313491344452,
      "logits/rejected": 1.0226128101348877,
      "logps/chosen": -414.9786682128906,
      "logps/rejected": -421.4689025878906,
      "loss": 0.5176,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.969223141670227,
      "rewards/margins": 0.6714719533920288,
      "rewards/rejected": -1.6406952142715454,
      "step": 380
    },
    {
      "epoch": 0.8914285714285715,
      "grad_norm": 24.654106172641406,
      "learning_rate": 1.7438359028687983e-08,
      "logits/chosen": 0.04682580381631851,
      "logits/rejected": 0.8462721109390259,
      "logps/chosen": -395.02130126953125,
      "logps/rejected": -450.59375,
      "loss": 0.525,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.836483359336853,
      "rewards/margins": 0.7217278480529785,
      "rewards/rejected": -1.5582112073898315,
      "step": 390
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 35.84442600986327,
      "learning_rate": 1.0855747162029361e-08,
      "logits/chosen": 0.2935855984687805,
      "logits/rejected": 0.8280227780342102,
      "logps/chosen": -379.0578918457031,
      "logps/rejected": -430.62261962890625,
      "loss": 0.5457,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -1.0032463073730469,
      "rewards/margins": 0.6372436881065369,
      "rewards/rejected": -1.640489935874939,
      "step": 400
    },
    {
      "epoch": 0.9142857142857143,
      "eval_logits/chosen": 0.24073126912117004,
      "eval_logits/rejected": 1.7997794151306152,
      "eval_logps/chosen": -362.5745849609375,
      "eval_logps/rejected": -405.11981201171875,
      "eval_loss": 0.5175118446350098,
      "eval_rewards/accuracies": 0.7715517282485962,
      "eval_rewards/chosen": -0.8697066307067871,
      "eval_rewards/margins": 0.9907302856445312,
      "eval_rewards/rejected": -1.8604369163513184,
      "eval_runtime": 90.8335,
      "eval_samples_per_second": 20.158,
      "eval_steps_per_second": 0.319,
      "step": 400
    },
    {
      "epoch": 0.9371428571428572,
      "grad_norm": 27.47931239793458,
      "learning_rate": 5.8005019731033615e-09,
      "logits/chosen": 0.10173013061285019,
      "logits/rejected": 1.089665174484253,
      "logps/chosen": -389.411865234375,
      "logps/rejected": -432.0257873535156,
      "loss": 0.5148,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -1.0123193264007568,
      "rewards/margins": 0.710638165473938,
      "rewards/rejected": -1.7229573726654053,
      "step": 410
    },
    {
      "epoch": 0.96,
      "grad_norm": 21.72575201877371,
      "learning_rate": 2.3049103053431886e-09,
      "logits/chosen": -0.029790764674544334,
      "logits/rejected": 1.4268020391464233,
      "logps/chosen": -357.87603759765625,
      "logps/rejected": -413.14227294921875,
      "loss": 0.5143,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.7512796521186829,
      "rewards/margins": 1.0402743816375732,
      "rewards/rejected": -1.7915542125701904,
      "step": 420
    },
    {
      "epoch": 0.9828571428571429,
      "grad_norm": 22.231222051131574,
      "learning_rate": 3.9129780600541397e-10,
      "logits/chosen": 0.29859107732772827,
      "logits/rejected": 1.3065626621246338,
      "logps/chosen": -373.48834228515625,
      "logps/rejected": -435.72515869140625,
      "loss": 0.5136,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.9125651121139526,
      "rewards/margins": 0.8079904317855835,
      "rewards/rejected": -1.7205555438995361,
      "step": 430
    },
    {
      "epoch": 0.9988571428571429,
      "step": 437,
      "total_flos": 0.0,
      "train_loss": 0.5586249282758351,
      "train_runtime": 11006.8714,
      "train_samples_per_second": 5.088,
      "train_steps_per_second": 0.04
    }
  ],
  "logging_steps": 10,
  "max_steps": 437,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}