{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4444444444444444,
  "eval_steps": 60,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008888888888888889,
      "grad_norm": 8.760091781616211,
      "learning_rate": 4.4247787610619474e-07,
      "logits/chosen": -0.8248252868652344,
      "logits/rejected": -0.8263720273971558,
      "logps/chosen": -0.36086463928222656,
      "logps/rejected": -5.696224689483643,
      "loss": 1.1038,
      "rewards/accuracies": 0.5125000476837158,
      "rewards/chosen": 17.43745994567871,
      "rewards/margins": 0.5984855890274048,
      "rewards/rejected": 16.838973999023438,
      "step": 10
    },
    {
      "epoch": 0.017777777777777778,
      "grad_norm": 8.855981826782227,
      "learning_rate": 8.849557522123895e-07,
      "logits/chosen": -0.8169006109237671,
      "logits/rejected": -0.819770872592926,
      "logps/chosen": -0.12464660406112671,
      "logps/rejected": -7.139842987060547,
      "loss": 1.1887,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": 17.17649269104004,
      "rewards/margins": 0.19107049703598022,
      "rewards/rejected": 16.98542022705078,
      "step": 20
    },
    {
      "epoch": 0.02666666666666667,
      "grad_norm": 16.764184951782227,
      "learning_rate": 1.3274336283185843e-06,
      "logits/chosen": -0.8003113865852356,
      "logits/rejected": -0.8030117750167847,
      "logps/chosen": -0.34651467204093933,
      "logps/rejected": -6.967917442321777,
      "loss": 1.0563,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": 17.280975341796875,
      "rewards/margins": 0.40005987882614136,
      "rewards/rejected": 16.88091468811035,
      "step": 30
    },
    {
      "epoch": 0.035555555555555556,
      "grad_norm": 8.33682918548584,
      "learning_rate": 1.769911504424779e-06,
      "logits/chosen": -0.7695047855377197,
      "logits/rejected": -0.7739207148551941,
      "logps/chosen": -1.5993006229400635,
      "logps/rejected": -8.504932403564453,
      "loss": 0.7596,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 17.283912658691406,
      "rewards/margins": 0.6976072192192078,
      "rewards/rejected": 16.5863037109375,
      "step": 40
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 4.494723320007324,
      "learning_rate": 2.212389380530974e-06,
      "logits/chosen": -0.7154140472412109,
      "logits/rejected": -0.7225576043128967,
      "logps/chosen": -3.112199068069458,
      "logps/rejected": -12.212080001831055,
      "loss": 0.6083,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 17.03064727783203,
      "rewards/margins": 0.7148451805114746,
      "rewards/rejected": 16.3158016204834,
      "step": 50
    },
    {
      "epoch": 0.05333333333333334,
      "grad_norm": 5.110287666320801,
      "learning_rate": 2.6548672566371687e-06,
      "logits/chosen": -0.6322453022003174,
      "logits/rejected": -0.6387485265731812,
      "logps/chosen": -5.650620460510254,
      "logps/rejected": -12.759811401367188,
      "loss": 0.3835,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 17.101289749145508,
      "rewards/margins": 1.1824612617492676,
      "rewards/rejected": 15.918828964233398,
      "step": 60
    },
    {
      "epoch": 0.05333333333333334,
      "eval_logits/chosen": -0.5826543569564819,
      "eval_logits/rejected": -0.5914276838302612,
      "eval_logps/chosen": -3.5471787452697754,
      "eval_logps/rejected": -16.51181983947754,
      "eval_loss": 0.3286525011062622,
      "eval_rewards/accuracies": 0.9280000925064087,
      "eval_rewards/chosen": 17.148174285888672,
      "eval_rewards/margins": 1.4386365413665771,
      "eval_rewards/rejected": 15.709539413452148,
      "eval_runtime": 372.0227,
      "eval_samples_per_second": 2.688,
      "eval_steps_per_second": 0.336,
      "step": 60
    },
    {
      "epoch": 0.06222222222222222,
      "grad_norm": 5.098133563995361,
      "learning_rate": 3.097345132743363e-06,
      "logits/chosen": -0.5378152132034302,
      "logits/rejected": -0.5494933724403381,
      "logps/chosen": -1.5099802017211914,
      "logps/rejected": -21.206321716308594,
      "loss": 0.2931,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 17.083791732788086,
      "rewards/margins": 1.5844331979751587,
      "rewards/rejected": 15.499359130859375,
      "step": 70
    },
    {
      "epoch": 0.07111111111111111,
      "grad_norm": 29.787437438964844,
      "learning_rate": 3.539823008849558e-06,
      "logits/chosen": -0.443774938583374,
      "logits/rejected": -0.45571577548980713,
      "logps/chosen": -1.5804342031478882,
      "logps/rejected": -22.606929779052734,
      "loss": 0.202,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 17.302125930786133,
      "rewards/margins": 2.174014091491699,
      "rewards/rejected": 15.128110885620117,
      "step": 80
    },
    {
      "epoch": 0.08,
      "grad_norm": 23.14398193359375,
      "learning_rate": 3.982300884955752e-06,
      "logits/chosen": -0.3626072406768799,
      "logits/rejected": -0.3787815570831299,
      "logps/chosen": -2.203828811645508,
      "logps/rejected": -29.433551788330078,
      "loss": 0.2123,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 17.00284194946289,
      "rewards/margins": 2.320391893386841,
      "rewards/rejected": 14.682450294494629,
      "step": 90
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 29.672739028930664,
      "learning_rate": 4.424778761061948e-06,
      "logits/chosen": -0.3035663962364197,
      "logits/rejected": -0.31762221455574036,
      "logps/chosen": -3.433589458465576,
      "logps/rejected": -29.9322509765625,
      "loss": 0.2592,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 16.929956436157227,
      "rewards/margins": 2.31272029876709,
      "rewards/rejected": 14.617237091064453,
      "step": 100
    },
    {
      "epoch": 0.09777777777777778,
      "grad_norm": 1.873722791671753,
      "learning_rate": 4.867256637168142e-06,
      "logits/chosen": -0.2679600715637207,
      "logits/rejected": -0.2826440930366516,
      "logps/chosen": -0.9653514623641968,
      "logps/rejected": -30.235322952270508,
      "loss": 0.1336,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 17.462385177612305,
      "rewards/margins": 3.1994175910949707,
      "rewards/rejected": 14.26296615600586,
      "step": 110
    },
    {
      "epoch": 0.10666666666666667,
      "grad_norm": 1.6913721561431885,
      "learning_rate": 4.999409761242696e-06,
      "logits/chosen": -0.22222033143043518,
      "logits/rejected": -0.23720571398735046,
      "logps/chosen": -4.4953508377075195,
      "logps/rejected": -34.074745178222656,
      "loss": 0.2552,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 17.04866600036621,
      "rewards/margins": 3.1014418601989746,
      "rewards/rejected": 13.947224617004395,
      "step": 120
    },
    {
      "epoch": 0.10666666666666667,
      "eval_logits/chosen": -0.206527978181839,
      "eval_logits/rejected": -0.22178640961647034,
      "eval_logps/chosen": -3.69442081451416,
      "eval_logps/rejected": -36.072166442871094,
      "eval_loss": 0.18996010720729828,
      "eval_rewards/accuracies": 0.9320000410079956,
      "eval_rewards/chosen": 17.133451461791992,
      "eval_rewards/margins": 3.379946708679199,
      "eval_rewards/rejected": 13.753504753112793,
      "eval_runtime": 361.5279,
      "eval_samples_per_second": 2.766,
      "eval_steps_per_second": 0.346,
      "step": 120
    },
    {
      "epoch": 0.11555555555555555,
      "grad_norm": 61.80262756347656,
      "learning_rate": 4.996519466816778e-06,
      "logits/chosen": -0.18473535776138306,
      "logits/rejected": -0.1988501250743866,
      "logps/chosen": -3.7009687423706055,
      "logps/rejected": -39.289939880371094,
      "loss": 0.1394,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 17.106964111328125,
      "rewards/margins": 3.633338212966919,
      "rewards/rejected": 13.473625183105469,
      "step": 130
    },
    {
      "epoch": 0.12444444444444444,
      "grad_norm": 1.6732702255249023,
      "learning_rate": 4.9912234871722805e-06,
      "logits/chosen": -0.16134041547775269,
      "logits/rejected": -0.17547868192195892,
      "logps/chosen": -3.0637736320495605,
      "logps/rejected": -40.07548522949219,
      "loss": 0.1408,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 17.392223358154297,
      "rewards/margins": 4.242353439331055,
      "rewards/rejected": 13.149867057800293,
      "step": 140
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 0.346453994512558,
      "learning_rate": 4.98352692559805e-06,
      "logits/chosen": -0.13797929883003235,
      "logits/rejected": -0.15283086895942688,
      "logps/chosen": -5.14492130279541,
      "logps/rejected": -47.97212219238281,
      "loss": 0.2153,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 16.896778106689453,
      "rewards/margins": 4.227695465087891,
      "rewards/rejected": 12.669081687927246,
      "step": 150
    },
    {
      "epoch": 0.14222222222222222,
      "grad_norm": 0.21871662139892578,
      "learning_rate": 4.973437198621237e-06,
      "logits/chosen": -0.12396670132875443,
      "logits/rejected": -0.13780555129051208,
      "logps/chosen": -6.108860015869141,
      "logps/rejected": -54.90739440917969,
      "loss": 0.0388,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 16.75935935974121,
      "rewards/margins": 4.755282878875732,
      "rewards/rejected": 12.004077911376953,
      "step": 160
    },
    {
      "epoch": 0.1511111111111111,
      "grad_norm": 235.12429809570312,
      "learning_rate": 4.960964028860621e-06,
      "logits/chosen": -0.1140839159488678,
      "logits/rejected": -0.1263057291507721,
      "logps/chosen": -12.605452537536621,
      "logps/rejected": -53.81230926513672,
      "loss": 0.4651,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 16.101238250732422,
      "rewards/margins": 3.9864249229431152,
      "rewards/rejected": 12.114812850952148,
      "step": 170
    },
    {
      "epoch": 0.16,
      "grad_norm": 190.97048950195312,
      "learning_rate": 4.946119435657738e-06,
      "logits/chosen": -0.10746976733207703,
      "logits/rejected": -0.11878640949726105,
      "logps/chosen": -8.5105562210083,
      "logps/rejected": -51.314781188964844,
      "loss": 0.2362,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 16.719980239868164,
      "rewards/margins": 4.549674034118652,
      "rewards/rejected": 12.170306205749512,
      "step": 180
    },
    {
      "epoch": 0.16,
      "eval_logits/chosen": -0.10870806127786636,
      "eval_logits/rejected": -0.12223993986845016,
      "eval_logps/chosen": -4.414996147155762,
      "eval_logps/rejected": -53.885032653808594,
      "eval_loss": 0.20236633718013763,
      "eval_rewards/accuracies": 0.9510000944137573,
      "eval_rewards/chosen": 17.06139373779297,
      "eval_rewards/margins": 5.089176177978516,
      "eval_rewards/rejected": 11.97221851348877,
      "eval_runtime": 361.4355,
      "eval_samples_per_second": 2.767,
      "eval_steps_per_second": 0.346,
      "step": 180
    },
    {
      "epoch": 0.1688888888888889,
      "grad_norm": 56.81266784667969,
      "learning_rate": 4.928917723494854e-06,
      "logits/chosen": -0.10682469606399536,
      "logits/rejected": -0.12124393880367279,
      "logps/chosen": -3.058413028717041,
      "logps/rejected": -55.052528381347656,
      "loss": 0.2442,
      "rewards/accuracies": 0.9500000476837158,
      "rewards/chosen": 17.058589935302734,
      "rewards/margins": 5.056097984313965,
      "rewards/rejected": 12.002490043640137,
      "step": 190
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 175.06552124023438,
      "learning_rate": 4.909375468210947e-06,
      "logits/chosen": -0.10520349442958832,
      "logits/rejected": -0.12018950283527374,
      "logps/chosen": -4.114959716796875,
      "logps/rejected": -55.9394645690918,
      "loss": 0.1915,
      "rewards/accuracies": 0.9500000476837158,
      "rewards/chosen": 16.98603057861328,
      "rewards/margins": 5.105838775634766,
      "rewards/rejected": 11.880191802978516,
      "step": 200
    },
    {
      "epoch": 0.18666666666666668,
      "grad_norm": 78.06558990478516,
      "learning_rate": 4.8875115010289655e-06,
      "logits/chosen": -0.10475558042526245,
      "logits/rejected": -0.11949175596237183,
      "logps/chosen": -6.760301113128662,
      "logps/rejected": -53.91607666015625,
      "loss": 0.2843,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 16.857545852661133,
      "rewards/margins": 4.917357921600342,
      "rewards/rejected": 11.94018840789795,
      "step": 210
    },
    {
      "epoch": 0.19555555555555557,
      "grad_norm": 15.880486488342285,
      "learning_rate": 4.863346890409768e-06,
      "logits/chosen": -0.11213523149490356,
      "logits/rejected": -0.12581588327884674,
      "logps/chosen": -6.759585380554199,
      "logps/rejected": -51.10936737060547,
      "loss": 0.5104,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 16.859071731567383,
      "rewards/margins": 4.638372898101807,
      "rewards/rejected": 12.220698356628418,
      "step": 220
    },
    {
      "epoch": 0.20444444444444446,
      "grad_norm": 46.97845458984375,
      "learning_rate": 4.836904921750224e-06,
      "logits/chosen": -0.11947059631347656,
      "logits/rejected": -0.1329912692308426,
      "logps/chosen": -3.608184814453125,
      "logps/rejected": -48.794761657714844,
      "loss": 0.2134,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 17.235904693603516,
      "rewards/margins": 4.859888076782227,
      "rewards/rejected": 12.376014709472656,
      "step": 230
    },
    {
      "epoch": 0.21333333333333335,
      "grad_norm": 24.032859802246094,
      "learning_rate": 4.808211074945042e-06,
      "logits/chosen": -0.1200513243675232,
      "logits/rejected": -0.1333036869764328,
      "logps/chosen": -3.7552154064178467,
      "logps/rejected": -49.87453079223633,
      "loss": 0.1781,
      "rewards/accuracies": 0.9500000476837158,
      "rewards/chosen": 17.094650268554688,
      "rewards/margins": 4.68077278137207,
      "rewards/rejected": 12.41387939453125,
      "step": 240
    },
    {
      "epoch": 0.21333333333333335,
      "eval_logits/chosen": -0.12433278560638428,
      "eval_logits/rejected": -0.13808581233024597,
      "eval_logps/chosen": -4.408891201019287,
      "eval_logps/rejected": -50.744781494140625,
      "eval_loss": 0.1546352356672287,
      "eval_rewards/accuracies": 0.9500000476837158,
      "eval_rewards/chosen": 17.06200408935547,
      "eval_rewards/margins": 4.775761604309082,
      "eval_rewards/rejected": 12.286243438720703,
      "eval_runtime": 361.4974,
      "eval_samples_per_second": 2.766,
      "eval_steps_per_second": 0.346,
      "step": 240
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.25737640261650085,
      "learning_rate": 4.7772929998339485e-06,
      "logits/chosen": -0.12348780035972595,
      "logits/rejected": -0.13704943656921387,
      "logps/chosen": -4.4299187660217285,
      "logps/rejected": -53.074607849121094,
      "loss": 0.1373,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 17.087068557739258,
      "rewards/margins": 5.06691837310791,
      "rewards/rejected": 12.020149230957031,
      "step": 250
    },
    {
      "epoch": 0.2311111111111111,
      "grad_norm": 0.1839389204978943,
      "learning_rate": 4.744180489557859e-06,
      "logits/chosen": -0.12177034467458725,
      "logits/rejected": -0.1342695653438568,
      "logps/chosen": -3.775188446044922,
      "logps/rejected": -53.98720932006836,
      "loss": 0.1896,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 17.12021255493164,
      "rewards/margins": 5.148064613342285,
      "rewards/rejected": 11.972146987915039,
      "step": 260
    },
    {
      "epoch": 0.24,
      "grad_norm": 12.258485794067383,
      "learning_rate": 4.708905451849754e-06,
      "logits/chosen": -0.11067859083414078,
      "logits/rejected": -0.12377731502056122,
      "logps/chosen": -6.418317794799805,
      "logps/rejected": -56.57402801513672,
      "loss": 0.2315,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 16.738832473754883,
      "rewards/margins": 4.884931564331055,
      "rewards/rejected": 11.853900909423828,
      "step": 270
    },
    {
      "epoch": 0.24888888888888888,
      "grad_norm": 77.56194305419922,
      "learning_rate": 4.671501878287879e-06,
      "logits/chosen": -0.1184445172548294,
      "logits/rejected": -0.1339874565601349,
      "logps/chosen": -10.12116527557373,
      "logps/rejected": -53.403907775878906,
      "loss": 0.5343,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 16.458633422851562,
      "rewards/margins": 4.402472496032715,
      "rewards/rejected": 12.056160926818848,
      "step": 280
    },
    {
      "epoch": 0.2577777777777778,
      "grad_norm": 67.53883361816406,
      "learning_rate": 4.6320058115409295e-06,
      "logits/chosen": -0.1448262631893158,
      "logits/rejected": -0.15793387591838837,
      "logps/chosen": -3.4666190147399902,
      "logps/rejected": -48.79213333129883,
      "loss": 0.5017,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 16.945899963378906,
      "rewards/margins": 4.2686333656311035,
      "rewards/rejected": 12.677268028259277,
      "step": 290
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 0.17521341145038605,
      "learning_rate": 4.590455310636778e-06,
      "logits/chosen": -0.16128253936767578,
      "logits/rejected": -0.17375555634498596,
      "logps/chosen": -2.9032950401306152,
      "logps/rejected": -47.69734191894531,
      "loss": 0.265,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 17.18383026123047,
      "rewards/margins": 4.541309356689453,
      "rewards/rejected": 12.642518997192383,
      "step": 300
    },
    {
      "epoch": 0.26666666666666666,
      "eval_logits/chosen": -0.17444846034049988,
      "eval_logits/rejected": -0.18559777736663818,
      "eval_logps/chosen": -2.535512924194336,
      "eval_logps/rejected": -47.16367721557617,
      "eval_loss": 0.15360687673091888,
      "eval_rewards/accuracies": 0.9440000653266907,
      "eval_rewards/chosen": 17.249343872070312,
      "eval_rewards/margins": 4.604989051818848,
      "eval_rewards/rejected": 12.644353866577148,
      "eval_runtime": 361.4575,
      "eval_samples_per_second": 2.767,
      "eval_steps_per_second": 0.346,
      "step": 300
    },
    {
      "epoch": 0.27555555555555555,
      "grad_norm": 0.5040452480316162,
      "learning_rate": 4.54689041428819e-06,
      "logits/chosen": -0.16974106431007385,
      "logits/rejected": -0.1810058057308197,
      "logps/chosen": -1.233938217163086,
      "logps/rejected": -49.907745361328125,
      "loss": 0.1132,
      "rewards/accuracies": 0.9500000476837158,
      "rewards/chosen": 17.34117889404297,
      "rewards/margins": 4.934173583984375,
      "rewards/rejected": 12.407005310058594,
      "step": 310
    },
    {
      "epoch": 0.28444444444444444,
      "grad_norm": 100.02949523925781,
      "learning_rate": 4.501353102310901e-06,
      "logits/chosen": -0.15705889463424683,
      "logits/rejected": -0.1695334017276764,
      "logps/chosen": -1.0820492506027222,
      "logps/rejected": -52.577110290527344,
      "loss": 0.1194,
      "rewards/accuracies": 0.9500000476837158,
      "rewards/chosen": 17.33388900756836,
      "rewards/margins": 5.154760837554932,
      "rewards/rejected": 12.179126739501953,
      "step": 320
    },
    {
      "epoch": 0.29333333333333333,
      "grad_norm": 0.2689219117164612,
      "learning_rate": 4.453887255171206e-06,
      "logits/chosen": -0.13849371671676636,
      "logits/rejected": -0.14990833401679993,
      "logps/chosen": -1.8435032367706299,
      "logps/rejected": -54.79044723510742,
      "loss": 0.0926,
      "rewards/accuracies": 0.9500000476837158,
      "rewards/chosen": 17.2423095703125,
      "rewards/margins": 5.28987979888916,
      "rewards/rejected": 11.952428817749023,
      "step": 330
    },
    {
      "epoch": 0.3022222222222222,
      "grad_norm": 0.09305431693792343,
      "learning_rate": 4.404538611702055e-06,
      "logits/chosen": -0.12299702316522598,
      "logits/rejected": -0.13453055918216705,
      "logps/chosen": -2.9897143840789795,
      "logps/rejected": -52.954498291015625,
      "loss": 0.2873,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 17.17474365234375,
      "rewards/margins": 5.071004867553711,
      "rewards/rejected": 12.103739738464355,
      "step": 340
    },
    {
      "epoch": 0.3111111111111111,
      "grad_norm": 59.282073974609375,
      "learning_rate": 4.3533547250284015e-06,
      "logits/chosen": -0.11913029849529266,
      "logits/rejected": -0.12785324454307556,
      "logps/chosen": -3.9456872940063477,
      "logps/rejected": -48.68487548828125,
      "loss": 0.4332,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 17.12805938720703,
      "rewards/margins": 4.669450283050537,
      "rewards/rejected": 12.458610534667969,
      "step": 350
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.31101909279823303,
      "learning_rate": 4.300384916744261e-06,
      "logits/chosen": -0.11280188709497452,
      "logits/rejected": -0.12300585210323334,
      "logps/chosen": -2.1714723110198975,
      "logps/rejected": -54.74174118041992,
      "loss": 0.1605,
      "rewards/accuracies": 0.9500000476837158,
      "rewards/chosen": 17.326162338256836,
      "rewards/margins": 5.467062473297119,
      "rewards/rejected": 11.859098434448242,
      "step": 360
    },
    {
      "epoch": 0.32,
      "eval_logits/chosen": -0.10620756447315216,
      "eval_logits/rejected": -0.11727114766836166,
      "eval_logps/chosen": -1.4165427684783936,
      "eval_logps/rejected": -50.9525146484375,
      "eval_loss": 0.3194349706172943,
      "eval_rewards/accuracies": 0.9210000038146973,
      "eval_rewards/chosen": 17.36124038696289,
      "eval_rewards/margins": 5.095769882202148,
      "eval_rewards/rejected": 12.26546859741211,
      "eval_runtime": 361.5072,
      "eval_samples_per_second": 2.766,
      "eval_steps_per_second": 0.346,
      "step": 360
    },
    {
      "epoch": 0.3288888888888889,
      "grad_norm": 6.1126532554626465,
      "learning_rate": 4.24568022938566e-06,
      "logits/chosen": -0.10354311764240265,
      "logits/rejected": -0.11526636779308319,
      "logps/chosen": -1.2935255765914917,
      "logps/rejected": -55.57566833496094,
      "loss": 0.1711,
      "rewards/accuracies": 0.9500000476837158,
      "rewards/chosen": 17.439346313476562,
      "rewards/margins": 5.700921058654785,
      "rewards/rejected": 11.738424301147461,
      "step": 370
    },
    {
      "epoch": 0.3377777777777778,
      "grad_norm": 34.15927505493164,
      "learning_rate": 4.189293377245241e-06,
      "logits/chosen": -0.1029932051897049,
      "logits/rejected": -0.11382515728473663,
      "logps/chosen": -2.5132687091827393,
      "logps/rejected": -55.50346374511719,
      "loss": 0.4359,
      "rewards/accuracies": 0.8875000476837158,
      "rewards/chosen": 16.731037139892578,
      "rewards/margins": 4.368172645568848,
      "rewards/rejected": 12.362865447998047,
      "step": 380
    },
    {
      "epoch": 0.3466666666666667,
      "grad_norm": 2.8422904014587402,
      "learning_rate": 4.131278695575952e-06,
      "logits/chosen": -0.10793520510196686,
      "logits/rejected": -0.12109285593032837,
      "logps/chosen": -3.014652729034424,
      "logps/rejected": -53.98411560058594,
      "loss": 0.2161,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 17.137393951416016,
      "rewards/margins": 5.105995178222656,
      "rewards/rejected": 12.03139877319336,
      "step": 390
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 54.0329475402832,
      "learning_rate": 4.071692088232743e-06,
      "logits/chosen": -0.10393750667572021,
      "logits/rejected": -0.11834606528282166,
      "logps/chosen": -2.1508543491363525,
      "logps/rejected": -45.60733413696289,
      "loss": 0.2077,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 17.586124420166016,
      "rewards/margins": 5.077212333679199,
      "rewards/rejected": 12.5089111328125,
      "step": 400
    },
    {
      "epoch": 0.36444444444444446,
      "grad_norm": 81.61144256591797,
      "learning_rate": 4.010590973802737e-06,
      "logits/chosen": -0.09564584493637085,
      "logits/rejected": -0.10617707669734955,
      "logps/chosen": -3.4572842121124268,
      "logps/rejected": -50.92162322998047,
      "loss": 0.2478,
      "rewards/accuracies": 0.8875000476837158,
      "rewards/chosen": 17.010910034179688,
      "rewards/margins": 4.556198596954346,
      "rewards/rejected": 12.454713821411133,
      "step": 410
    },
    {
      "epoch": 0.37333333333333335,
      "grad_norm": 0.30974289774894714,
      "learning_rate": 3.948034230275781e-06,
      "logits/chosen": -0.09134417027235031,
      "logits/rejected": -0.1020016297698021,
      "logps/chosen": -5.046698570251465,
      "logps/rejected": -48.908958435058594,
      "loss": 0.2894,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 17.007888793945312,
      "rewards/margins": 4.53641414642334,
      "rewards/rejected": 12.471475601196289,
      "step": 420
    },
    {
      "epoch": 0.37333333333333335,
      "eval_logits/chosen": -0.09054450690746307,
      "eval_logits/rejected": -0.10264354199171066,
      "eval_logps/chosen": -1.913105845451355,
      "eval_logps/rejected": -51.11127471923828,
      "eval_loss": 0.16789735853672028,
      "eval_rewards/accuracies": 0.9450000524520874,
      "eval_rewards/chosen": 17.311582565307617,
      "eval_rewards/margins": 5.061989784240723,
      "eval_rewards/rejected": 12.249593734741211,
      "eval_runtime": 361.5337,
      "eval_samples_per_second": 2.766,
      "eval_steps_per_second": 0.346,
      "step": 420
    },
    {
      "epoch": 0.38222222222222224,
      "grad_norm": 12.824393272399902,
      "learning_rate": 3.884082138308699e-06,
      "logits/chosen": -0.08666776865720749,
      "logits/rejected": -0.0997733399271965,
      "logps/chosen": -1.7306327819824219,
      "logps/rejected": -54.273292541503906,
      "loss": 0.2298,
      "rewards/accuracies": 0.9500000476837158,
      "rewards/chosen": 17.167621612548828,
      "rewards/margins": 5.065673351287842,
      "rewards/rejected": 12.101947784423828,
      "step": 430
    },
    {
      "epoch": 0.39111111111111113,
      "grad_norm": 0.30713599920272827,
      "learning_rate": 3.818796323137896e-06,
      "logits/chosen": -0.09174907952547073,
      "logits/rejected": -0.10376611351966858,
      "logps/chosen": -1.489154577255249,
      "logps/rejected": -54.580726623535156,
      "loss": 0.2513,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 17.22280502319336,
      "rewards/margins": 5.175349235534668,
      "rewards/rejected": 12.047454833984375,
      "step": 440
    },
    {
      "epoch": 0.4,
      "grad_norm": 87.4791488647461,
      "learning_rate": 3.7522396951963303e-06,
      "logits/chosen": -0.09688778221607208,
      "logits/rejected": -0.10897806286811829,
      "logps/chosen": -3.157695770263672,
      "logps/rejected": -50.96417236328125,
      "loss": 0.1758,
      "rewards/accuracies": 0.9500000476837158,
      "rewards/chosen": 17.345651626586914,
      "rewards/margins": 5.245656967163086,
      "rewards/rejected": 12.099993705749512,
      "step": 450
    },
    {
      "epoch": 0.4088888888888889,
      "grad_norm": 146.2008056640625,
      "learning_rate": 3.684476389492026e-06,
      "logits/chosen": -0.09378582239151001,
      "logits/rejected": -0.10475654900074005,
      "logps/chosen": -0.5611928701400757,
      "logps/rejected": -56.518890380859375,
      "loss": 0.1981,
      "rewards/accuracies": 0.9500000476837158,
      "rewards/chosen": 17.113712310791016,
      "rewards/margins": 5.068872928619385,
      "rewards/rejected": 12.044839859008789,
      "step": 460
    },
    {
      "epoch": 0.4177777777777778,
      "grad_norm": 1.9137721061706543,
      "learning_rate": 3.6155717038065783e-06,
      "logits/chosen": -0.08695463836193085,
      "logits/rejected": -0.09596743434667587,
      "logps/chosen": -1.5298550128936768,
      "logps/rejected": -50.27445983886719,
      "loss": 0.2066,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 17.35186004638672,
      "rewards/margins": 5.014693260192871,
      "rewards/rejected": 12.337167739868164,
      "step": 470
    },
    {
      "epoch": 0.4266666666666667,
      "grad_norm": 84.80391693115234,
      "learning_rate": 3.545592035773192e-06,
      "logits/chosen": -0.0746893435716629,
      "logits/rejected": -0.08653923869132996,
      "logps/chosen": -2.0052125453948975,
      "logps/rejected": -57.502811431884766,
      "loss": 0.1149,
      "rewards/accuracies": 0.9500000476837158,
      "rewards/chosen": 17.14373016357422,
      "rewards/margins": 5.360415935516357,
      "rewards/rejected": 11.783313751220703,
      "step": 480
    },
    {
      "epoch": 0.4266666666666667,
      "eval_logits/chosen": -0.07700399309396744,
      "eval_logits/rejected": -0.08828537166118622,
      "eval_logps/chosen": -4.48896598815918,
      "eval_logps/rejected": -53.76282501220703,
      "eval_loss": 0.29511645436286926,
      "eval_rewards/accuracies": 0.9230000376701355,
      "eval_rewards/chosen": 17.053997039794922,
      "eval_rewards/margins": 5.069558143615723,
      "eval_rewards/rejected": 11.984437942504883,
      "eval_runtime": 361.5035,
      "eval_samples_per_second": 2.766,
      "eval_steps_per_second": 0.346,
      "step": 480
    },
    {
      "epoch": 0.43555555555555553,
      "grad_norm": 82.9616470336914,
      "learning_rate": 3.4746048188948806e-06,
      "logits/chosen": -0.06675051152706146,
      "logits/rejected": -0.07860895991325378,
      "logps/chosen": -4.162237167358398,
      "logps/rejected": -54.77789306640625,
      "loss": 0.2979,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 17.047603607177734,
      "rewards/margins": 5.138361930847168,
      "rewards/rejected": 11.909242630004883,
      "step": 490
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.04293210059404373,
      "learning_rate": 3.4026784575644887e-06,
      "logits/chosen": -0.06424491107463837,
      "logits/rejected": -0.07567107677459717,
      "logps/chosen": -2.05729603767395,
      "logps/rejected": -56.646087646484375,
      "loss": 0.4378,
      "rewards/accuracies": 0.8875000476837158,
      "rewards/chosen": 16.947803497314453,
      "rewards/margins": 4.919981956481934,
      "rewards/rejected": 12.02782154083252,
      "step": 500
    }
  ],
  "logging_steps": 10,
  "max_steps": 1125,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.372648598634496e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}