{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.1076923076923078,
  "eval_steps": 5,
  "global_step": 90,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.024615384615384615,
      "grad_norm": 76.71143242842001,
      "learning_rate": 5.88235294117647e-08,
      "logits/chosen": -0.8526538610458374,
      "logits/rejected": -0.8570448756217957,
      "logps/chosen": -23.282678604125977,
      "logps/rejected": -30.289661407470703,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 2
    },
    {
      "epoch": 0.04923076923076923,
      "grad_norm": 83.48860140376064,
      "learning_rate": 1.176470588235294e-07,
      "logits/chosen": -0.8658735752105713,
      "logits/rejected": -0.8668463826179504,
      "logps/chosen": -30.029348373413086,
      "logps/rejected": -32.66902160644531,
      "loss": 0.6971,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": 0.0059752073138952255,
      "rewards/margins": -0.010342838242650032,
      "rewards/rejected": 0.016318045556545258,
      "step": 4
    },
    {
      "epoch": 0.06153846153846154,
      "eval_logits/chosen": -0.8142690658569336,
      "eval_logits/rejected": -0.8192334771156311,
      "eval_logps/chosen": -29.028133392333984,
      "eval_logps/rejected": -38.206329345703125,
      "eval_loss": 0.6946294903755188,
      "eval_rewards/accuracies": 0.47602739930152893,
      "eval_rewards/chosen": -0.0063199191354215145,
      "eval_rewards/margins": 0.0024889421183615923,
      "eval_rewards/rejected": -0.00880886148661375,
      "eval_runtime": 507.7462,
      "eval_samples_per_second": 3.415,
      "eval_steps_per_second": 0.144,
      "step": 5
    },
    {
      "epoch": 0.07384615384615385,
      "grad_norm": 69.27587032017023,
      "learning_rate": 1.764705882352941e-07,
      "logits/chosen": -0.8687959909439087,
      "logits/rejected": -0.8719169497489929,
      "logps/chosen": -32.97465133666992,
      "logps/rejected": -38.870269775390625,
      "loss": 0.6921,
      "rewards/accuracies": 0.453125,
      "rewards/chosen": 0.006213514134287834,
      "rewards/margins": 0.019537178799510002,
      "rewards/rejected": -0.013323664665222168,
      "step": 6
    },
    {
      "epoch": 0.09846153846153846,
      "grad_norm": 62.33035519033225,
      "learning_rate": 2.352941176470588e-07,
      "logits/chosen": -0.8589476346969604,
      "logits/rejected": -0.8617635369300842,
      "logps/chosen": -32.01108932495117,
      "logps/rejected": -34.94812774658203,
      "loss": 0.6928,
      "rewards/accuracies": 0.515625,
      "rewards/chosen": -0.01385729480534792,
      "rewards/margins": -8.035916835069656e-05,
      "rewards/rejected": -0.013776935636997223,
      "step": 8
    },
    {
      "epoch": 0.12307692307692308,
      "grad_norm": 57.3164290353471,
      "learning_rate": 2.941176470588235e-07,
      "logits/chosen": -0.8420453071594238,
      "logits/rejected": -0.8434039354324341,
      "logps/chosen": -30.150243759155273,
      "logps/rejected": -33.6817626953125,
      "loss": 0.6827,
      "rewards/accuracies": 0.484375,
      "rewards/chosen": -0.021684179082512856,
      "rewards/margins": -0.011505719274282455,
      "rewards/rejected": -0.0101784598082304,
      "step": 10
    },
    {
      "epoch": 0.12307692307692308,
      "eval_logits/chosen": -0.8148170709609985,
      "eval_logits/rejected": -0.8197983503341675,
      "eval_logps/chosen": -29.04338264465332,
      "eval_logps/rejected": -38.422237396240234,
      "eval_loss": 0.6693353056907654,
      "eval_rewards/accuracies": 0.5650684833526611,
      "eval_rewards/chosen": -0.013945668935775757,
      "eval_rewards/margins": 0.10281600803136826,
      "eval_rewards/rejected": -0.11676166206598282,
      "eval_runtime": 507.283,
      "eval_samples_per_second": 3.418,
      "eval_steps_per_second": 0.144,
      "step": 10
    },
    {
      "epoch": 0.1476923076923077,
      "grad_norm": 54.695358572076394,
      "learning_rate": 3.529411764705882e-07,
      "logits/chosen": -0.8680934906005859,
      "logits/rejected": -0.8719905614852905,
      "logps/chosen": -28.021072387695312,
      "logps/rejected": -35.96692657470703,
      "loss": 0.6666,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.036122702062129974,
      "rewards/margins": 0.10737781226634979,
      "rewards/rejected": -0.14350052177906036,
      "step": 12
    },
    {
      "epoch": 0.1723076923076923,
      "grad_norm": 52.51352144095569,
      "learning_rate": 4.117647058823529e-07,
      "logits/chosen": -0.8596158027648926,
      "logits/rejected": -0.8575820922851562,
      "logps/chosen": -25.149404525756836,
      "logps/rejected": -24.465974807739258,
      "loss": 0.6357,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": -0.029037628322839737,
      "rewards/margins": 0.0849144458770752,
      "rewards/rejected": -0.11395206302404404,
      "step": 14
    },
    {
      "epoch": 0.18461538461538463,
      "eval_logits/chosen": -0.8164225816726685,
      "eval_logits/rejected": -0.8213981986045837,
      "eval_logps/chosen": -29.094459533691406,
      "eval_logps/rejected": -39.09992599487305,
      "eval_loss": 0.6130890846252441,
      "eval_rewards/accuracies": 0.6575342416763306,
      "eval_rewards/chosen": -0.039482928812503815,
      "eval_rewards/margins": 0.4161252975463867,
      "eval_rewards/rejected": -0.4556082487106323,
      "eval_runtime": 506.4372,
      "eval_samples_per_second": 3.424,
      "eval_steps_per_second": 0.144,
      "step": 15
    },
    {
      "epoch": 0.19692307692307692,
      "grad_norm": 47.087567227271336,
      "learning_rate": 4.705882352941176e-07,
      "logits/chosen": -0.8515424728393555,
      "logits/rejected": -0.858619213104248,
      "logps/chosen": -21.42935562133789,
      "logps/rejected": -35.25102233886719,
      "loss": 0.6054,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.047799061983823776,
      "rewards/margins": 0.34381866455078125,
      "rewards/rejected": -0.39161768555641174,
      "step": 16
    },
    {
      "epoch": 0.22153846153846155,
      "grad_norm": 45.639809270904706,
      "learning_rate": 4.99941324504621e-07,
      "logits/chosen": -0.8649481534957886,
      "logits/rejected": -0.8680992126464844,
      "logps/chosen": -26.61380386352539,
      "logps/rejected": -37.828311920166016,
      "loss": 0.5893,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.06860450655221939,
      "rewards/margins": 0.5221824645996094,
      "rewards/rejected": -0.5907869338989258,
      "step": 18
    },
    {
      "epoch": 0.24615384615384617,
      "grad_norm": 42.712271689759746,
      "learning_rate": 4.99472085783721e-07,
      "logits/chosen": -0.8538618087768555,
      "logits/rejected": -0.8574154376983643,
      "logps/chosen": -26.50242805480957,
      "logps/rejected": -36.465850830078125,
      "loss": 0.5349,
      "rewards/accuracies": 0.796875,
      "rewards/chosen": -0.20561164617538452,
      "rewards/margins": 0.5988239645957947,
      "rewards/rejected": -0.8044356107711792,
      "step": 20
    },
    {
      "epoch": 0.24615384615384617,
      "eval_logits/chosen": -0.8193703889846802,
      "eval_logits/rejected": -0.8243635892868042,
      "eval_logps/chosen": -29.048322677612305,
      "eval_logps/rejected": -40.16524124145508,
      "eval_loss": 0.5301748514175415,
      "eval_rewards/accuracies": 0.7123287916183472,
      "eval_rewards/chosen": -0.016416184604167938,
      "eval_rewards/margins": 0.9718519449234009,
      "eval_rewards/rejected": -0.9882679581642151,
      "eval_runtime": 501.8283,
      "eval_samples_per_second": 3.455,
      "eval_steps_per_second": 0.145,
      "step": 20
    },
    {
      "epoch": 0.27076923076923076,
      "grad_norm": 34.55903445927571,
      "learning_rate": 4.985344892885899e-07,
      "logits/chosen": -0.8641754984855652,
      "logits/rejected": -0.8657775521278381,
      "logps/chosen": -29.697185516357422,
      "logps/rejected": -37.21875,
      "loss": 0.5547,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.11109799891710281,
      "rewards/margins": 0.7374492287635803,
      "rewards/rejected": -0.8485472202301025,
      "step": 22
    },
    {
      "epoch": 0.2953846153846154,
      "grad_norm": 34.94086237660157,
      "learning_rate": 4.971302952586796e-07,
      "logits/chosen": -0.8593113422393799,
      "logits/rejected": -0.8664268255233765,
      "logps/chosen": -28.04953384399414,
      "logps/rejected": -47.995513916015625,
      "loss": 0.4738,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": -0.05936732143163681,
      "rewards/margins": 1.61855947971344,
      "rewards/rejected": -1.6779268980026245,
      "step": 24
    },
    {
      "epoch": 0.3076923076923077,
      "eval_logits/chosen": -0.8194184899330139,
      "eval_logits/rejected": -0.8244072794914246,
      "eval_logps/chosen": -28.87408447265625,
      "eval_logps/rejected": -41.07496643066406,
      "eval_loss": 0.47070130705833435,
      "eval_rewards/accuracies": 0.715753436088562,
      "eval_rewards/chosen": 0.07070425897836685,
      "eval_rewards/margins": 1.513830304145813,
      "eval_rewards/rejected": -1.4431262016296387,
      "eval_runtime": 505.6197,
      "eval_samples_per_second": 3.429,
      "eval_steps_per_second": 0.144,
      "step": 25
    },
    {
      "epoch": 0.32,
      "grad_norm": 36.047711342996145,
      "learning_rate": 4.952621399215597e-07,
      "logits/chosen": -0.8540530204772949,
      "logits/rejected": -0.8563276529312134,
      "logps/chosen": -25.321256637573242,
      "logps/rejected": -26.884302139282227,
      "loss": 0.4982,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.013541080057621002,
      "rewards/margins": 0.5806317329406738,
      "rewards/rejected": -0.5670906901359558,
      "step": 26
    },
    {
      "epoch": 0.3446153846153846,
      "grad_norm": 34.93615246503423,
      "learning_rate": 4.929335305436764e-07,
      "logits/chosen": -0.8274993300437927,
      "logits/rejected": -0.8279544115066528,
      "logps/chosen": -20.29788589477539,
      "logps/rejected": -38.79467010498047,
      "loss": 0.4679,
      "rewards/accuracies": 0.859375,
      "rewards/chosen": -0.06851354986429214,
      "rewards/margins": 1.8159857988357544,
      "rewards/rejected": -1.8844993114471436,
      "step": 28
    },
    {
      "epoch": 0.36923076923076925,
      "grad_norm": 34.651174705720734,
      "learning_rate": 4.901488388458247e-07,
      "logits/chosen": -0.8323720693588257,
      "logits/rejected": -0.8362505435943604,
      "logps/chosen": -23.299272537231445,
      "logps/rejected": -31.706920623779297,
      "loss": 0.4411,
      "rewards/accuracies": 0.703125,
      "rewards/chosen": -0.14743047952651978,
      "rewards/margins": 1.210200548171997,
      "rewards/rejected": -1.3576310873031616,
      "step": 30
    },
    {
      "epoch": 0.36923076923076925,
      "eval_logits/chosen": -0.8136431574821472,
      "eval_logits/rejected": -0.8185814023017883,
      "eval_logps/chosen": -28.679155349731445,
      "eval_logps/rejected": -42.067020416259766,
      "eval_loss": 0.40313246846199036,
      "eval_rewards/accuracies": 0.7465753555297852,
      "eval_rewards/chosen": 0.16816774010658264,
      "eval_rewards/margins": 2.1073246002197266,
      "eval_rewards/rejected": -1.9391568899154663,
      "eval_runtime": 511.2322,
      "eval_samples_per_second": 3.392,
      "eval_steps_per_second": 0.143,
      "step": 30
    },
    {
      "epoch": 0.39384615384615385,
      "grad_norm": 28.222143999842437,
      "learning_rate": 4.869132927957006e-07,
      "logits/chosen": -0.8474501371383667,
      "logits/rejected": -0.8529994487762451,
      "logps/chosen": -27.735734939575195,
      "logps/rejected": -40.067569732666016,
      "loss": 0.3871,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 0.20432481169700623,
      "rewards/margins": 2.164912462234497,
      "rewards/rejected": -1.960587739944458,
      "step": 32
    },
    {
      "epoch": 0.41846153846153844,
      "grad_norm": 23.28549229020256,
      "learning_rate": 4.832329667929376e-07,
      "logits/chosen": -0.8629348278045654,
      "logits/rejected": -0.8702473640441895,
      "logps/chosen": -20.325801849365234,
      "logps/rejected": -44.433528900146484,
      "loss": 0.3497,
      "rewards/accuracies": 0.921875,
      "rewards/chosen": 0.10242755711078644,
      "rewards/margins": 2.8072259426116943,
      "rewards/rejected": -2.7047982215881348,
      "step": 34
    },
    {
      "epoch": 0.4307692307692308,
      "eval_logits/chosen": -0.8100441098213196,
      "eval_logits/rejected": -0.8149632811546326,
      "eval_logps/chosen": -28.504199981689453,
      "eval_logps/rejected": -42.80283737182617,
      "eval_loss": 0.354584664106369,
      "eval_rewards/accuracies": 0.7739726305007935,
      "eval_rewards/chosen": 0.25564688444137573,
      "eval_rewards/margins": 2.5627098083496094,
      "eval_rewards/rejected": -2.307062864303589,
      "eval_runtime": 507.4648,
      "eval_samples_per_second": 3.417,
      "eval_steps_per_second": 0.144,
      "step": 35
    },
    {
      "epoch": 0.4430769230769231,
      "grad_norm": 28.694201633510016,
      "learning_rate": 4.791147702650565e-07,
      "logits/chosen": -0.8724276423454285,
      "logits/rejected": -0.8785867691040039,
      "logps/chosen": -22.125022888183594,
      "logps/rejected": -39.55330276489258,
      "loss": 0.3713,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 0.26476141810417175,
      "rewards/margins": 2.4366393089294434,
      "rewards/rejected": -2.171877861022949,
      "step": 36
    },
    {
      "epoch": 0.4676923076923077,
      "grad_norm": 25.53720719902276,
      "learning_rate": 4.745664346957361e-07,
      "logits/chosen": -0.8554552793502808,
      "logits/rejected": -0.8553615808486938,
      "logps/chosen": -33.38023376464844,
      "logps/rejected": -28.513282775878906,
      "loss": 0.3246,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 0.2387702763080597,
      "rewards/margins": 1.5690217018127441,
      "rewards/rejected": -1.3302514553070068,
      "step": 38
    },
    {
      "epoch": 0.49230769230769234,
      "grad_norm": 21.696391543968662,
      "learning_rate": 4.695964991097616e-07,
      "logits/chosen": -0.8336246013641357,
      "logits/rejected": -0.8416473269462585,
      "logps/chosen": -22.365100860595703,
      "logps/rejected": -47.882694244384766,
      "loss": 0.346,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": 0.2181028425693512,
      "rewards/margins": 3.067445993423462,
      "rewards/rejected": -2.8493428230285645,
      "step": 40
    },
    {
      "epoch": 0.49230769230769234,
      "eval_logits/chosen": -0.8083286285400391,
      "eval_logits/rejected": -0.8132520914077759,
      "eval_logps/chosen": -28.318235397338867,
      "eval_logps/rejected": -43.32075881958008,
      "eval_loss": 0.3231227993965149,
      "eval_rewards/accuracies": 0.801369845867157,
      "eval_rewards/chosen": 0.3486267626285553,
      "eval_rewards/margins": 2.9146482944488525,
      "eval_rewards/rejected": -2.56602144241333,
      "eval_runtime": 507.9764,
      "eval_samples_per_second": 3.414,
      "eval_steps_per_second": 0.144,
      "step": 40
    },
    {
      "epoch": 0.5169230769230769,
      "grad_norm": 21.517342705892485,
      "learning_rate": 4.642142940418973e-07,
      "logits/chosen": -0.8719685077667236,
      "logits/rejected": -0.8759061694145203,
      "logps/chosen": -23.35956573486328,
      "logps/rejected": -28.740066528320312,
      "loss": 0.3379,
      "rewards/accuracies": 0.765625,
      "rewards/chosen": 0.2940492033958435,
      "rewards/margins": 1.586828589439392,
      "rewards/rejected": -1.2927793264389038,
      "step": 42
    },
    {
      "epoch": 0.5415384615384615,
      "grad_norm": 22.644901266797667,
      "learning_rate": 4.5842992401978256e-07,
      "logits/chosen": -0.8451048135757446,
      "logits/rejected": -0.847484827041626,
      "logps/chosen": -33.87168884277344,
      "logps/rejected": -46.649681091308594,
      "loss": 0.3148,
      "rewards/accuracies": 0.859375,
      "rewards/chosen": 0.3618454933166504,
      "rewards/margins": 3.3943986892700195,
      "rewards/rejected": -3.0325536727905273,
      "step": 44
    },
    {
      "epoch": 0.5538461538461539,
      "eval_logits/chosen": -0.8085425496101379,
      "eval_logits/rejected": -0.8135057687759399,
      "eval_logps/chosen": -28.0711727142334,
      "eval_logps/rejected": -43.83769226074219,
      "eval_loss": 0.2965641915798187,
      "eval_rewards/accuracies": 0.8082191944122314,
      "eval_rewards/chosen": 0.472160279750824,
      "eval_rewards/margins": 3.296651840209961,
      "eval_rewards/rejected": -2.824491500854492,
      "eval_runtime": 508.0249,
      "eval_samples_per_second": 3.413,
      "eval_steps_per_second": 0.144,
      "step": 45
    },
    {
      "epoch": 0.5661538461538461,
      "grad_norm": 20.742372238934433,
      "learning_rate": 4.5225424859373684e-07,
      "logits/chosen": -0.8579592108726501,
      "logits/rejected": -0.8616234064102173,
      "logps/chosen": -21.818204879760742,
      "logps/rejected": -30.142559051513672,
      "loss": 0.2855,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": 0.27730488777160645,
      "rewards/margins": 1.6132526397705078,
      "rewards/rejected": -1.335947871208191,
      "step": 46
    },
    {
      "epoch": 0.5907692307692308,
      "grad_norm": 23.021619175104203,
      "learning_rate": 4.456988619490889e-07,
      "logits/chosen": -0.8618453741073608,
      "logits/rejected": -0.8655754923820496,
      "logps/chosen": -22.56821060180664,
      "logps/rejected": -36.07129669189453,
      "loss": 0.2874,
      "rewards/accuracies": 0.859375,
      "rewards/chosen": 0.2860993444919586,
      "rewards/margins": 2.503146171569824,
      "rewards/rejected": -2.2170469760894775,
      "step": 48
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 19.42806196152356,
      "learning_rate": 4.3877607113930516e-07,
      "logits/chosen": -0.8602553009986877,
      "logits/rejected": -0.8559304475784302,
      "logps/chosen": -29.165634155273438,
      "logps/rejected": -24.462434768676758,
      "loss": 0.2644,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.5700158476829529,
      "rewards/margins": 2.1351168155670166,
      "rewards/rejected": -1.5651010274887085,
      "step": 50
    },
    {
      "epoch": 0.6153846153846154,
      "eval_logits/chosen": -0.812079131603241,
      "eval_logits/rejected": -0.8171290159225464,
      "eval_logps/chosen": -27.80472183227539,
      "eval_logps/rejected": -44.38835906982422,
      "eval_loss": 0.27530089020729065,
      "eval_rewards/accuracies": 0.8047945499420166,
      "eval_rewards/chosen": 0.6053856015205383,
      "eval_rewards/margins": 3.7052102088928223,
      "eval_rewards/rejected": -3.0998241901397705,
      "eval_runtime": 507.1773,
      "eval_samples_per_second": 3.419,
      "eval_steps_per_second": 0.144,
      "step": 50
    },
    {
      "epoch": 0.64,
      "grad_norm": 19.235430940038626,
      "learning_rate": 4.314988729807827e-07,
      "logits/chosen": -0.877213716506958,
      "logits/rejected": -0.8788630366325378,
      "logps/chosen": -35.67780685424805,
      "logps/rejected": -39.79027557373047,
      "loss": 0.261,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.6618055701255798,
      "rewards/margins": 2.99965763092041,
      "rewards/rejected": -2.3378520011901855,
      "step": 52
    },
    {
      "epoch": 0.6646153846153846,
      "grad_norm": 18.497275752898116,
      "learning_rate": 4.238809296526846e-07,
      "logits/chosen": -0.872668445110321,
      "logits/rejected": -0.8716113567352295,
      "logps/chosen": -30.384735107421875,
      "logps/rejected": -35.33108139038086,
      "loss": 0.2597,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.40745919942855835,
      "rewards/margins": 2.814540386199951,
      "rewards/rejected": -2.407081365585327,
      "step": 54
    },
    {
      "epoch": 0.676923076923077,
      "eval_logits/chosen": -0.817668616771698,
      "eval_logits/rejected": -0.8228325247764587,
      "eval_logps/chosen": -27.571548461914062,
      "eval_logps/rejected": -44.867122650146484,
      "eval_loss": 0.2588183581829071,
      "eval_rewards/accuracies": 0.8047945499420166,
      "eval_rewards/chosen": 0.7219717502593994,
      "eval_rewards/margins": 4.061178207397461,
      "eval_rewards/rejected": -3.3392062187194824,
      "eval_runtime": 512.6554,
      "eval_samples_per_second": 3.382,
      "eval_steps_per_second": 0.142,
      "step": 55
    },
    {
      "epoch": 0.6892307692307692,
      "grad_norm": 19.164588052012302,
      "learning_rate": 4.159365430476261e-07,
      "logits/chosen": -0.8461377024650574,
      "logits/rejected": -0.8481893539428711,
      "logps/chosen": -23.498355865478516,
      "logps/rejected": -27.572532653808594,
      "loss": 0.254,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 0.4778762459754944,
      "rewards/margins": 2.123842716217041,
      "rewards/rejected": -1.645966649055481,
      "step": 56
    },
    {
      "epoch": 0.7138461538461538,
      "grad_norm": 18.695688461758973,
      "learning_rate": 4.076806279213655e-07,
      "logits/chosen": -0.8788058757781982,
      "logits/rejected": -0.8795627951622009,
      "logps/chosen": -27.387178421020508,
      "logps/rejected": -24.94676971435547,
      "loss": 0.2539,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 0.39298391342163086,
      "rewards/margins": 2.3424487113952637,
      "rewards/rejected": -1.9494649171829224,
      "step": 58
    },
    {
      "epoch": 0.7384615384615385,
      "grad_norm": 13.869174391468711,
      "learning_rate": 3.991286838919086e-07,
      "logits/chosen": -0.8662111163139343,
      "logits/rejected": -0.8639529943466187,
      "logps/chosen": -29.126564025878906,
      "logps/rejected": -26.93923568725586,
      "loss": 0.2163,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 0.5371630787849426,
      "rewards/margins": 2.0503089427948,
      "rewards/rejected": -1.5131456851959229,
      "step": 60
    },
    {
      "epoch": 0.7384615384615385,
      "eval_logits/chosen": -0.8232029676437378,
      "eval_logits/rejected": -0.8284507989883423,
      "eval_logps/chosen": -27.39695167541504,
      "eval_logps/rejected": -45.318092346191406,
      "eval_loss": 0.24679508805274963,
      "eval_rewards/accuracies": 0.801369845867157,
      "eval_rewards/chosen": 0.8092703819274902,
      "eval_rewards/margins": 4.373960494995117,
      "eval_rewards/rejected": -3.5646896362304688,
      "eval_runtime": 504.2778,
      "eval_samples_per_second": 3.439,
      "eval_steps_per_second": 0.145,
      "step": 60
    },
    {
      "epoch": 0.7630769230769231,
      "grad_norm": 19.770687722416966,
      "learning_rate": 3.902967663405956e-07,
      "logits/chosen": -0.8548551201820374,
      "logits/rejected": -0.8548004031181335,
      "logps/chosen": -23.15505599975586,
      "logps/rejected": -25.987648010253906,
      "loss": 0.2459,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 0.9548823237419128,
      "rewards/margins": 2.6681408882141113,
      "rewards/rejected": -1.7132583856582642,
      "step": 62
    },
    {
      "epoch": 0.7876923076923077,
      "grad_norm": 17.375432493553873,
      "learning_rate": 3.8120145626980015e-07,
      "logits/chosen": -0.837457537651062,
      "logits/rejected": -0.8325684070587158,
      "logps/chosen": -28.196165084838867,
      "logps/rejected": -29.63174819946289,
      "loss": 0.226,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.5275837779045105,
      "rewards/margins": 2.420928716659546,
      "rewards/rejected": -1.8933448791503906,
      "step": 64
    },
    {
      "epoch": 0.8,
      "eval_logits/chosen": -0.8285362124443054,
      "eval_logits/rejected": -0.8338667750358582,
      "eval_logps/chosen": -27.281187057495117,
      "eval_logps/rejected": -45.72236251831055,
      "eval_loss": 0.2386258840560913,
      "eval_rewards/accuracies": 0.8082191944122314,
      "eval_rewards/chosen": 0.8671532869338989,
      "eval_rewards/margins": 4.633976936340332,
      "eval_rewards/rejected": -3.7668235301971436,
      "eval_runtime": 504.9865,
      "eval_samples_per_second": 3.434,
      "eval_steps_per_second": 0.145,
      "step": 65
    },
    {
      "epoch": 0.8123076923076923,
      "grad_norm": 16.23920440465993,
      "learning_rate": 3.718598291738298e-07,
      "logits/chosen": -0.8503263592720032,
      "logits/rejected": -0.8503541350364685,
      "logps/chosen": -26.526954650878906,
      "logps/rejected": -33.03715133666992,
      "loss": 0.2399,
      "rewards/accuracies": 0.828125,
      "rewards/chosen": 0.8604952096939087,
      "rewards/margins": 3.0948374271392822,
      "rewards/rejected": -2.234342336654663,
      "step": 66
    },
    {
      "epoch": 0.8369230769230769,
      "grad_norm": 16.54810979887679,
      "learning_rate": 3.622894229814698e-07,
      "logits/chosen": -0.8387259840965271,
      "logits/rejected": -0.8485996127128601,
      "logps/chosen": -19.66301155090332,
      "logps/rejected": -59.002899169921875,
      "loss": 0.2031,
      "rewards/accuracies": 0.796875,
      "rewards/chosen": 0.6494669318199158,
      "rewards/margins": 6.0868682861328125,
      "rewards/rejected": -5.43740177154541,
      "step": 68
    },
    {
      "epoch": 0.8615384615384616,
      "grad_norm": 14.045977696001755,
      "learning_rate": 3.52508205130354e-07,
      "logits/chosen": -0.8623999953269958,
      "logits/rejected": -0.8682447075843811,
      "logps/chosen": -26.10405158996582,
      "logps/rejected": -45.96481704711914,
      "loss": 0.1936,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.3354326486587524,
      "rewards/margins": 5.366407871246338,
      "rewards/rejected": -4.030975341796875,
      "step": 70
    },
    {
      "epoch": 0.8615384615384616,
      "eval_logits/chosen": -0.8313784003257751,
      "eval_logits/rejected": -0.83674556016922,
      "eval_logps/chosen": -27.231040954589844,
      "eval_logps/rejected": -46.05835723876953,
      "eval_loss": 0.23013119399547577,
      "eval_rewards/accuracies": 0.8116438388824463,
      "eval_rewards/chosen": 0.8922267556190491,
      "eval_rewards/margins": 4.827047824859619,
      "eval_rewards/rejected": -3.9348206520080566,
      "eval_runtime": 505.5211,
      "eval_samples_per_second": 3.43,
      "eval_steps_per_second": 0.144,
      "step": 70
    },
    {
      "epoch": 0.8861538461538462,
      "grad_norm": 20.76852884591744,
      "learning_rate": 3.4253453883497864e-07,
      "logits/chosen": -0.8721504211425781,
      "logits/rejected": -0.8774159550666809,
      "logps/chosen": -27.097929000854492,
      "logps/rejected": -44.38412094116211,
      "loss": 0.2111,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.6299474239349365,
      "rewards/margins": 4.718615531921387,
      "rewards/rejected": -4.088667869567871,
      "step": 72
    },
    {
      "epoch": 0.9107692307692308,
      "grad_norm": 21.511447954249494,
      "learning_rate": 3.323871486116851e-07,
      "logits/chosen": -0.8902064561843872,
      "logits/rejected": -0.8880136013031006,
      "logps/chosen": -26.3861141204834,
      "logps/rejected": -28.770051956176758,
      "loss": 0.2491,
      "rewards/accuracies": 0.890625,
      "rewards/chosen": 0.5782642364501953,
      "rewards/margins": 2.829862594604492,
      "rewards/rejected": -2.251598358154297,
      "step": 74
    },
    {
      "epoch": 0.9230769230769231,
      "eval_logits/chosen": -0.8325244188308716,
      "eval_logits/rejected": -0.8379253149032593,
      "eval_logps/chosen": -27.20527458190918,
      "eval_logps/rejected": -46.30801010131836,
      "eval_loss": 0.22474366426467896,
      "eval_rewards/accuracies": 0.8150684833526611,
      "eval_rewards/chosen": 0.9051090478897095,
      "eval_rewards/margins": 4.96475887298584,
      "eval_rewards/rejected": -4.05964994430542,
      "eval_runtime": 519.4473,
      "eval_samples_per_second": 3.338,
      "eval_steps_per_second": 0.141,
      "step": 75
    },
    {
      "epoch": 0.9353846153846154,
      "grad_norm": 15.058617576147917,
      "learning_rate": 3.220850851253377e-07,
      "logits/chosen": -0.8858978748321533,
      "logits/rejected": -0.8874486684799194,
      "logps/chosen": -26.12487030029297,
      "logps/rejected": -37.51970672607422,
      "loss": 0.2158,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.6246981024742126,
      "rewards/margins": 3.997086763381958,
      "rewards/rejected": -3.372389316558838,
      "step": 76
    },
    {
      "epoch": 0.96,
      "grad_norm": 18.385729369704148,
      "learning_rate": 3.1164768942369053e-07,
      "logits/chosen": -0.8896593451499939,
      "logits/rejected": -0.8969188332557678,
      "logps/chosen": -16.990447998046875,
      "logps/rejected": -45.557037353515625,
      "loss": 0.2207,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 0.7743335366249084,
      "rewards/margins": 4.943416118621826,
      "rewards/rejected": -4.1690826416015625,
      "step": 78
    },
    {
      "epoch": 0.9846153846153847,
      "grad_norm": 17.268106016496986,
      "learning_rate": 3.010945566265912e-07,
      "logits/chosen": -0.8736187219619751,
      "logits/rejected": -0.876487672328949,
      "logps/chosen": -28.458038330078125,
      "logps/rejected": -34.158485412597656,
      "loss": 0.2315,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.3357160687446594,
      "rewards/margins": 3.1702334880828857,
      "rewards/rejected": -2.834517478942871,
      "step": 80
    },
    {
      "epoch": 0.9846153846153847,
      "eval_logits/chosen": -0.8342627882957458,
      "eval_logits/rejected": -0.8397004008293152,
      "eval_logps/chosen": -27.177371978759766,
      "eval_logps/rejected": -46.54536437988281,
      "eval_loss": 0.21979747712612152,
      "eval_rewards/accuracies": 0.8287671208381653,
      "eval_rewards/chosen": 0.9190611243247986,
      "eval_rewards/margins": 5.097388744354248,
      "eval_rewards/rejected": -4.178327560424805,
      "eval_runtime": 503.3189,
      "eval_samples_per_second": 3.445,
      "eval_steps_per_second": 0.145,
      "step": 80
    },
    {
      "epoch": 1.0092307692307692,
      "grad_norm": 28.182829085656643,
      "learning_rate": 2.9044549913819124e-07,
      "logits/chosen": -0.8684386014938354,
      "logits/rejected": -0.8742809891700745,
      "logps/chosen": -25.187530517578125,
      "logps/rejected": -54.615386962890625,
      "loss": 0.2079,
      "rewards/accuracies": 0.890625,
      "rewards/chosen": 0.7747483849525452,
      "rewards/margins": 6.844926834106445,
      "rewards/rejected": -6.070178985595703,
      "step": 82
    },
    {
      "epoch": 1.0338461538461539,
      "grad_norm": 13.964979005113792,
      "learning_rate": 2.797205094512266e-07,
      "logits/chosen": -0.8816483020782471,
      "logits/rejected": -0.8857114315032959,
      "logps/chosen": -28.65145492553711,
      "logps/rejected": -51.351768493652344,
      "loss": 0.1902,
      "rewards/accuracies": 0.859375,
      "rewards/chosen": 1.2536072731018066,
      "rewards/margins": 6.389583587646484,
      "rewards/rejected": -5.1359758377075195,
      "step": 84
    },
    {
      "epoch": 1.0461538461538462,
      "eval_logits/chosen": -0.8372805714607239,
      "eval_logits/rejected": -0.8427817225456238,
      "eval_logps/chosen": -27.166318893432617,
      "eval_logps/rejected": -46.72504806518555,
      "eval_loss": 0.21650995314121246,
      "eval_rewards/accuracies": 0.818493127822876,
      "eval_rewards/chosen": 0.924587070941925,
      "eval_rewards/margins": 5.192756652832031,
      "eval_rewards/rejected": -4.268169403076172,
      "eval_runtime": 500.3676,
      "eval_samples_per_second": 3.465,
      "eval_steps_per_second": 0.146,
      "step": 85
    },
    {
      "epoch": 1.0584615384615386,
      "grad_norm": 7.945903874330553,
      "learning_rate": 2.6893972261320264e-07,
      "logits/chosen": -0.8936392068862915,
      "logits/rejected": -0.8996983170509338,
      "logps/chosen": -22.51564598083496,
      "logps/rejected": -45.285972595214844,
      "loss": 0.1474,
      "rewards/accuracies": 0.921875,
      "rewards/chosen": 1.0068135261535645,
      "rewards/margins": 4.962003231048584,
      "rewards/rejected": -3.9551897048950195,
      "step": 86
    },
    {
      "epoch": 1.083076923076923,
      "grad_norm": 12.2186980926015,
      "learning_rate": 2.5812337842494516e-07,
      "logits/chosen": -0.8699577450752258,
      "logits/rejected": -0.8725451827049255,
      "logps/chosen": -29.06435203552246,
      "logps/rejected": -37.989776611328125,
      "loss": 0.1718,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.28801125288009644,
      "rewards/margins": 3.8652138710021973,
      "rewards/rejected": -3.577202796936035,
      "step": 88
    },
    {
      "epoch": 1.1076923076923078,
      "grad_norm": 10.256960594018043,
      "learning_rate": 2.4729178344249006e-07,
      "logits/chosen": -0.9174846410751343,
      "logits/rejected": -0.9205011129379272,
      "logps/chosen": -29.678579330444336,
      "logps/rejected": -41.8637580871582,
      "loss": 0.168,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.9668111801147461,
      "rewards/margins": 5.104186058044434,
      "rewards/rejected": -4.1373748779296875,
      "step": 90
    },
    {
      "epoch": 1.1076923076923078,
      "eval_logits/chosen": -0.840661346912384,
      "eval_logits/rejected": -0.8462072014808655,
      "eval_logps/chosen": -27.118879318237305,
      "eval_logps/rejected": -46.88822555541992,
      "eval_loss": 0.21390603482723236,
      "eval_rewards/accuracies": 0.818493127822876,
      "eval_rewards/chosen": 0.9483062624931335,
      "eval_rewards/margins": 5.298061847686768,
      "eval_rewards/rejected": -4.349754810333252,
      "eval_runtime": 504.3162,
      "eval_samples_per_second": 3.438,
      "eval_steps_per_second": 0.145,
      "step": 90
    }
  ],
  "logging_steps": 2,
  "max_steps": 162,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}