{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 356,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.056338028169014086,
      "grad_norm": 574.27978515625,
      "learning_rate": 6.25e-07,
      "logits/chosen": 2.945640802383423,
      "logits/rejected": 3.0515265464782715,
      "logps/chosen": -246.39541625976562,
      "logps/rejected": -195.46109008789062,
      "loss": 0.9304,
      "rewards/accuracies": 0.5218750238418579,
      "rewards/chosen": 0.03325582295656204,
      "rewards/margins": 0.014068150892853737,
      "rewards/rejected": 0.019187677651643753,
      "step": 10
    },
    {
      "epoch": 0.11267605633802817,
      "grad_norm": 267.4449462890625,
      "learning_rate": 1.3194444444444446e-06,
      "logits/chosen": 3.0365357398986816,
      "logits/rejected": 3.0756676197052,
      "logps/chosen": -231.8340606689453,
      "logps/rejected": -196.01499938964844,
      "loss": 0.5293,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 1.023498773574829,
      "rewards/margins": 1.6332218647003174,
      "rewards/rejected": -0.6097229719161987,
      "step": 20
    },
    {
      "epoch": 0.16901408450704225,
      "grad_norm": 168.65689086914062,
      "learning_rate": 2.0138888888888893e-06,
      "logits/chosen": 3.038414478302002,
      "logits/rejected": 3.0314807891845703,
      "logps/chosen": -265.27825927734375,
      "logps/rejected": -199.1730499267578,
      "loss": 0.1929,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 4.1361799240112305,
      "rewards/margins": 6.971987724304199,
      "rewards/rejected": -2.835808277130127,
      "step": 30
    },
    {
      "epoch": 0.22535211267605634,
      "grad_norm": 37.93267822265625,
      "learning_rate": 2.7083333333333334e-06,
      "logits/chosen": 3.0692856311798096,
      "logits/rejected": 3.1041903495788574,
      "logps/chosen": -245.3774871826172,
      "logps/rejected": -207.87789916992188,
      "loss": 0.1002,
      "rewards/accuracies": 0.9671875238418579,
      "rewards/chosen": 6.395417213439941,
      "rewards/margins": 13.0038480758667,
      "rewards/rejected": -6.608429908752441,
      "step": 40
    },
    {
      "epoch": 0.28169014084507044,
      "grad_norm": 129.71878051757812,
      "learning_rate": 3.4027777777777783e-06,
      "logits/chosen": 2.943250894546509,
      "logits/rejected": 2.9869542121887207,
      "logps/chosen": -245.9439239501953,
      "logps/rejected": -202.8123321533203,
      "loss": 0.1375,
      "rewards/accuracies": 0.973437488079071,
      "rewards/chosen": 7.993535041809082,
      "rewards/margins": 19.0964412689209,
      "rewards/rejected": -11.102909088134766,
      "step": 50
    },
    {
      "epoch": 0.3380281690140845,
      "grad_norm": 57.64502716064453,
      "learning_rate": 4.097222222222222e-06,
      "logits/chosen": 3.0049643516540527,
      "logits/rejected": 3.0770554542541504,
      "logps/chosen": -236.73410034179688,
      "logps/rejected": -210.151611328125,
      "loss": 0.1722,
      "rewards/accuracies": 0.965624988079071,
      "rewards/chosen": 9.21152114868164,
      "rewards/margins": 22.748985290527344,
      "rewards/rejected": -13.53746509552002,
      "step": 60
    },
    {
      "epoch": 0.39436619718309857,
      "grad_norm": 105.868408203125,
      "learning_rate": 4.791666666666668e-06,
      "logits/chosen": 3.0406033992767334,
      "logits/rejected": 3.0739083290100098,
      "logps/chosen": -256.1812744140625,
      "logps/rejected": -204.35446166992188,
      "loss": 0.1129,
      "rewards/accuracies": 0.979687511920929,
      "rewards/chosen": 8.951081275939941,
      "rewards/margins": 21.80655860900879,
      "rewards/rejected": -12.85547924041748,
      "step": 70
    },
    {
      "epoch": 0.4507042253521127,
      "grad_norm": 168.75357055664062,
      "learning_rate": 4.998524282731094e-06,
      "logits/chosen": 3.0504136085510254,
      "logits/rejected": 3.0491514205932617,
      "logps/chosen": -255.90316772460938,
      "logps/rejected": -206.5497283935547,
      "loss": 0.3368,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 8.498771667480469,
      "rewards/margins": 22.880569458007812,
      "rewards/rejected": -14.381797790527344,
      "step": 80
    },
    {
      "epoch": 0.5070422535211268,
      "grad_norm": 109.43980407714844,
      "learning_rate": 4.991300473502437e-06,
      "logits/chosen": 3.0595602989196777,
      "logits/rejected": 3.1263351440429688,
      "logps/chosen": -228.88070678710938,
      "logps/rejected": -218.99783325195312,
      "loss": 0.3032,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 6.379992485046387,
      "rewards/margins": 28.722070693969727,
      "rewards/rejected": -22.34208106994629,
      "step": 90
    },
    {
      "epoch": 0.5633802816901409,
      "grad_norm": 128.98760986328125,
      "learning_rate": 4.978074903220964e-06,
      "logits/chosen": 3.1774604320526123,
      "logits/rejected": 3.181511402130127,
      "logps/chosen": -228.3056640625,
      "logps/rejected": -222.18038940429688,
      "loss": 0.2029,
      "rewards/accuracies": 0.9781249761581421,
      "rewards/chosen": 5.856657028198242,
      "rewards/margins": 34.18266677856445,
      "rewards/rejected": -28.326007843017578,
      "step": 100
    },
    {
      "epoch": 0.6197183098591549,
      "grad_norm": 144.1073455810547,
      "learning_rate": 4.958879433443904e-06,
      "logits/chosen": 3.218785047531128,
      "logits/rejected": 3.1809394359588623,
      "logps/chosen": -243.4155731201172,
      "logps/rejected": -236.49868774414062,
      "loss": 0.2134,
      "rewards/accuracies": 0.9765625,
      "rewards/chosen": 6.719731330871582,
      "rewards/margins": 35.10533905029297,
      "rewards/rejected": -28.385604858398438,
      "step": 110
    },
    {
      "epoch": 0.676056338028169,
      "grad_norm": 149.68939208984375,
      "learning_rate": 4.933760307739277e-06,
      "logits/chosen": 3.1437926292419434,
      "logits/rejected": 3.1433277130126953,
      "logps/chosen": -255.87796020507812,
      "logps/rejected": -231.52182006835938,
      "loss": 0.2745,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": 6.636951446533203,
      "rewards/margins": 37.604583740234375,
      "rewards/rejected": -30.967632293701172,
      "step": 120
    },
    {
      "epoch": 0.7323943661971831,
      "grad_norm": 171.0680389404297,
      "learning_rate": 4.90277804028108e-06,
      "logits/chosen": 3.206712245941162,
      "logits/rejected": 3.2256839275360107,
      "logps/chosen": -261.3406982421875,
      "logps/rejected": -228.3511505126953,
      "loss": 0.5352,
      "rewards/accuracies": 0.964062511920929,
      "rewards/chosen": 4.778416633605957,
      "rewards/margins": 33.36115646362305,
      "rewards/rejected": -28.58274269104004,
      "step": 130
    },
    {
      "epoch": 0.7887323943661971,
      "grad_norm": 118.76382446289062,
      "learning_rate": 4.866007270065345e-06,
      "logits/chosen": 3.2129406929016113,
      "logits/rejected": 3.227287769317627,
      "logps/chosen": -246.7953643798828,
      "logps/rejected": -230.45651245117188,
      "loss": 0.2311,
      "rewards/accuracies": 0.9765625,
      "rewards/chosen": 0.6871539354324341,
      "rewards/margins": 36.90045166015625,
      "rewards/rejected": -36.213294982910156,
      "step": 140
    },
    {
      "epoch": 0.8450704225352113,
      "grad_norm": 198.268310546875,
      "learning_rate": 4.823536581098262e-06,
      "logits/chosen": 3.1713156700134277,
      "logits/rejected": 3.152228355407715,
      "logps/chosen": -238.29159545898438,
      "logps/rejected": -238.2996826171875,
      "loss": 0.2612,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 1.5464673042297363,
      "rewards/margins": 41.4525146484375,
      "rewards/rejected": -39.90605163574219,
      "step": 150
    },
    {
      "epoch": 0.9014084507042254,
      "grad_norm": 191.44915771484375,
      "learning_rate": 4.775468288989545e-06,
      "logits/chosen": 3.2888169288635254,
      "logits/rejected": 3.2170681953430176,
      "logps/chosen": -251.0401153564453,
      "logps/rejected": -236.00479125976562,
      "loss": 0.51,
      "rewards/accuracies": 0.971875011920929,
      "rewards/chosen": 2.7297756671905518,
      "rewards/margins": 42.41936492919922,
      "rewards/rejected": -39.68959426879883,
      "step": 160
    },
    {
      "epoch": 0.9577464788732394,
      "grad_norm": 73.10994720458984,
      "learning_rate": 4.7219181944651695e-06,
      "logits/chosen": 3.2599377632141113,
      "logits/rejected": 3.2560737133026123,
      "logps/chosen": -245.1952667236328,
      "logps/rejected": -241.4735565185547,
      "loss": 0.313,
      "rewards/accuracies": 0.973437488079071,
      "rewards/chosen": -0.8346878290176392,
      "rewards/margins": 44.581295013427734,
      "rewards/rejected": -45.415985107421875,
      "step": 170
    },
    {
      "epoch": 1.0112676056338028,
      "grad_norm": 71.17866516113281,
      "learning_rate": 4.663015304393279e-06,
      "logits/chosen": 3.3155758380889893,
      "logits/rejected": 3.3333241939544678,
      "logps/chosen": -270.6006774902344,
      "logps/rejected": -241.6360321044922,
      "loss": 0.338,
      "rewards/accuracies": 0.9736841917037964,
      "rewards/chosen": 0.171695277094841,
      "rewards/margins": 45.20180130004883,
      "rewards/rejected": -45.03010177612305,
      "step": 180
    },
    {
      "epoch": 1.0676056338028168,
      "grad_norm": 113.73174285888672,
      "learning_rate": 4.59890152099534e-06,
      "logits/chosen": 3.508065700531006,
      "logits/rejected": 3.5083301067352295,
      "logps/chosen": -266.78387451171875,
      "logps/rejected": -252.5698699951172,
      "loss": 0.8076,
      "rewards/accuracies": 0.964062511920929,
      "rewards/chosen": 10.96754264831543,
      "rewards/margins": 63.106483459472656,
      "rewards/rejected": -52.138938903808594,
      "step": 190
    },
    {
      "epoch": 1.123943661971831,
      "grad_norm": 0.890201985836029,
      "learning_rate": 4.5297312999912625e-06,
      "logits/chosen": 3.5750420093536377,
      "logits/rejected": 3.5598015785217285,
      "logps/chosen": -244.08615112304688,
      "logps/rejected": -249.61984252929688,
      "loss": 0.4625,
      "rewards/accuracies": 0.9765625,
      "rewards/chosen": 8.597976684570312,
      "rewards/margins": 65.22010803222656,
      "rewards/rejected": -56.62213897705078,
      "step": 200
    },
    {
      "epoch": 1.180281690140845,
      "grad_norm": 204.70803833007812,
      "learning_rate": 4.455671278502042e-06,
      "logits/chosen": 3.6519992351531982,
      "logits/rejected": 3.6381404399871826,
      "logps/chosen": -240.0857696533203,
      "logps/rejected": -238.0441131591797,
      "loss": 0.653,
      "rewards/accuracies": 0.9546874761581421,
      "rewards/chosen": 6.305617332458496,
      "rewards/margins": 53.15886688232422,
      "rewards/rejected": -46.8532600402832,
      "step": 210
    },
    {
      "epoch": 1.236619718309859,
      "grad_norm": 139.86141967773438,
      "learning_rate": 4.376899873606336e-06,
      "logits/chosen": 3.622997283935547,
      "logits/rejected": 3.6359076499938965,
      "logps/chosen": -259.12725830078125,
      "logps/rejected": -250.89501953125,
      "loss": 0.9057,
      "rewards/accuracies": 0.9453125,
      "rewards/chosen": 4.471914291381836,
      "rewards/margins": 51.15581130981445,
      "rewards/rejected": -46.683895111083984,
      "step": 220
    },
    {
      "epoch": 1.2929577464788733,
      "grad_norm": 94.85035705566406,
      "learning_rate": 4.293606852518101e-06,
      "logits/chosen": 3.6598830223083496,
      "logits/rejected": 3.594897508621216,
      "logps/chosen": -263.00079345703125,
      "logps/rejected": -248.59524536132812,
      "loss": 0.3918,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 2.2544591426849365,
      "rewards/margins": 56.7413444519043,
      "rewards/rejected": -54.48688507080078,
      "step": 230
    },
    {
      "epoch": 1.3492957746478873,
      "grad_norm": 37.788360595703125,
      "learning_rate": 4.205992875420742e-06,
      "logits/chosen": 3.57995343208313,
      "logits/rejected": 3.5664329528808594,
      "logps/chosen": -243.8499755859375,
      "logps/rejected": -252.65353393554688,
      "loss": 0.4873,
      "rewards/accuracies": 0.9781249761581421,
      "rewards/chosen": 7.1952104568481445,
      "rewards/margins": 64.06478118896484,
      "rewards/rejected": -56.86956787109375,
      "step": 240
    },
    {
      "epoch": 1.4056338028169013,
      "grad_norm": 159.25875854492188,
      "learning_rate": 4.114269012059169e-06,
      "logits/chosen": 3.5491080284118652,
      "logits/rejected": 3.510200023651123,
      "logps/chosen": -253.5936279296875,
      "logps/rejected": -254.4651336669922,
      "loss": 0.3683,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 4.560586452484131,
      "rewards/margins": 59.49503707885742,
      "rewards/rejected": -54.9344482421875,
      "step": 250
    },
    {
      "epoch": 1.4619718309859155,
      "grad_norm": 113.43865966796875,
      "learning_rate": 4.018656233254278e-06,
      "logits/chosen": 3.4692349433898926,
      "logits/rejected": 3.3723576068878174,
      "logps/chosen": -253.46249389648438,
      "logps/rejected": -268.2179870605469,
      "loss": 0.3905,
      "rewards/accuracies": 0.984375,
      "rewards/chosen": 3.9269375801086426,
      "rewards/margins": 74.34013366699219,
      "rewards/rejected": -70.41320037841797,
      "step": 260
    },
    {
      "epoch": 1.5183098591549296,
      "grad_norm": 36.17473602294922,
      "learning_rate": 3.919384878564902e-06,
      "logits/chosen": 3.488018035888672,
      "logits/rejected": 3.4679126739501953,
      "logps/chosen": -248.9741668701172,
      "logps/rejected": -282.92840576171875,
      "loss": 0.5658,
      "rewards/accuracies": 0.9781249761581421,
      "rewards/chosen": 4.083906650543213,
      "rewards/margins": 77.73112487792969,
      "rewards/rejected": -73.647216796875,
      "step": 270
    },
    {
      "epoch": 1.5746478873239438,
      "grad_norm": 177.54251098632812,
      "learning_rate": 3.816694101379631e-06,
      "logits/chosen": 3.6568634510040283,
      "logits/rejected": 3.6167445182800293,
      "logps/chosen": -240.26895141601562,
      "logps/rejected": -263.4219665527344,
      "loss": 0.3476,
      "rewards/accuracies": 0.979687511920929,
      "rewards/chosen": 4.189328193664551,
      "rewards/margins": 66.97871398925781,
      "rewards/rejected": -62.78938674926758,
      "step": 280
    },
    {
      "epoch": 1.6309859154929578,
      "grad_norm": 76.29850006103516,
      "learning_rate": 3.7108312927753533e-06,
      "logits/chosen": 3.5761330127716064,
      "logits/rejected": 3.5349583625793457,
      "logps/chosen": -256.83697509765625,
      "logps/rejected": -257.34063720703125,
      "loss": 0.2007,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 3.5510315895080566,
      "rewards/margins": 67.4579849243164,
      "rewards/rejected": -63.906944274902344,
      "step": 290
    },
    {
      "epoch": 1.6873239436619718,
      "grad_norm": 88.87206268310547,
      "learning_rate": 3.6020514855304856e-06,
      "logits/chosen": 3.4625442028045654,
      "logits/rejected": 3.4353599548339844,
      "logps/chosen": -246.8657684326172,
      "logps/rejected": -262.365966796875,
      "loss": 0.3359,
      "rewards/accuracies": 0.9765625,
      "rewards/chosen": -1.413503885269165,
      "rewards/margins": 65.82698059082031,
      "rewards/rejected": -67.240478515625,
      "step": 300
    },
    {
      "epoch": 1.7436619718309858,
      "grad_norm": 109.3917007446289,
      "learning_rate": 3.4906167397286643e-06,
      "logits/chosen": 3.4818034172058105,
      "logits/rejected": 3.438699722290039,
      "logps/chosen": -248.93112182617188,
      "logps/rejected": -257.42694091796875,
      "loss": 0.4265,
      "rewards/accuracies": 0.9828125238418579,
      "rewards/chosen": 0.5807734131813049,
      "rewards/margins": 62.88755416870117,
      "rewards/rejected": -62.306793212890625,
      "step": 310
    },
    {
      "epoch": 1.8,
      "grad_norm": 88.23355102539062,
      "learning_rate": 3.3767955114330586e-06,
      "logits/chosen": 3.3789494037628174,
      "logits/rejected": 3.329315662384033,
      "logps/chosen": -258.67291259765625,
      "logps/rejected": -259.5810546875,
      "loss": 0.4489,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -2.7831504344940186,
      "rewards/margins": 59.0428466796875,
      "rewards/rejected": -61.82598876953125,
      "step": 320
    },
    {
      "epoch": 1.856338028169014,
      "grad_norm": 1.0168672800064087,
      "learning_rate": 3.2608620059521935e-06,
      "logits/chosen": 3.38688325881958,
      "logits/rejected": 3.349194288253784,
      "logps/chosen": -233.11666870117188,
      "logps/rejected": -270.48590087890625,
      "loss": 0.5576,
      "rewards/accuracies": 0.979687511920929,
      "rewards/chosen": -4.500494480133057,
      "rewards/margins": 67.32182312011719,
      "rewards/rejected": -71.82231140136719,
      "step": 330
    },
    {
      "epoch": 1.9126760563380283,
      "grad_norm": 106.62237548828125,
      "learning_rate": 3.14309551725535e-06,
      "logits/chosen": 3.401437759399414,
      "logits/rejected": 3.3626091480255127,
      "logps/chosen": -247.9010009765625,
      "logps/rejected": -256.19097900390625,
      "loss": 0.4024,
      "rewards/accuracies": 0.979687511920929,
      "rewards/chosen": 0.5352304577827454,
      "rewards/margins": 63.979896545410156,
      "rewards/rejected": -63.444664001464844,
      "step": 340
    },
    {
      "epoch": 1.9690140845070423,
      "grad_norm": 159.5343475341797,
      "learning_rate": 3.0237797551289228e-06,
      "logits/chosen": 3.4421494007110596,
      "logits/rejected": 3.4059250354766846,
      "logps/chosen": -229.91134643554688,
      "logps/rejected": -252.29019165039062,
      "loss": 0.6959,
      "rewards/accuracies": 0.971875011920929,
      "rewards/chosen": 0.6608916521072388,
      "rewards/margins": 60.5278205871582,
      "rewards/rejected": -59.86692428588867,
      "step": 350
    }
  ],
  "logging_steps": 10,
  "max_steps": 712,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}