{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 385,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 1.282051282051282e-07,
      "logits/chosen": -1.7278180122375488,
      "logits/rejected": -1.7377450466156006,
      "logps/chosen": -29.553977966308594,
      "logps/rejected": -42.813133239746094,
      "loss": 6.25,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.282051282051282e-06,
      "logits/chosen": -1.866089105606079,
      "logits/rejected": -1.8704073429107666,
      "logps/chosen": -36.98554229736328,
      "logps/rejected": -33.6707763671875,
      "loss": 5.979,
      "rewards/accuracies": 0.5972222089767456,
      "rewards/chosen": 0.004210897721350193,
      "rewards/margins": 0.01235075294971466,
      "rewards/rejected": -0.008139855228364468,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.564102564102564e-06,
      "logits/chosen": -1.9978349208831787,
      "logits/rejected": -2.0004687309265137,
      "logps/chosen": -29.640878677368164,
      "logps/rejected": -29.042272567749023,
      "loss": 6.3859,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": 0.00026315837749280035,
      "rewards/margins": -0.003971050027757883,
      "rewards/rejected": 0.004234207794070244,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.846153846153847e-06,
      "logits/chosen": -1.9206383228302002,
      "logits/rejected": -1.9179503917694092,
      "logps/chosen": -31.404415130615234,
      "logps/rejected": -33.228981018066406,
      "loss": 6.1874,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.00233639357611537,
      "rewards/margins": 0.004440182354301214,
      "rewards/rejected": -0.002103788312524557,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.999896948438434e-06,
      "logits/chosen": -2.017291784286499,
      "logits/rejected": -2.008547067642212,
      "logps/chosen": -32.58599090576172,
      "logps/rejected": -32.512664794921875,
      "loss": 6.3474,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": -0.001837015151977539,
      "rewards/margins": -0.0018385002622380853,
      "rewards/rejected": 1.4854595065116882e-06,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.987541037542187e-06,
      "logits/chosen": -1.8622735738754272,
      "logits/rejected": -1.8514816761016846,
      "logps/chosen": -33.54685592651367,
      "logps/rejected": -35.447818756103516,
      "loss": 6.2748,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.0021248008124530315,
      "rewards/margins": 0.0012201189529150724,
      "rewards/rejected": 0.0009046817431226373,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.954691471941119e-06,
      "logits/chosen": -1.9413617849349976,
      "logits/rejected": -1.943302869796753,
      "logps/chosen": -32.52573013305664,
      "logps/rejected": -33.22004318237305,
      "loss": 5.8451,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.014890777878463268,
      "rewards/margins": 0.024623576551675797,
      "rewards/rejected": -0.009732798673212528,
      "step": 60
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.901618883413549e-06,
      "logits/chosen": -2.0721487998962402,
      "logits/rejected": -2.07711124420166,
      "logps/chosen": -33.97162628173828,
      "logps/rejected": -36.63127517700195,
      "loss": 6.0458,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.0033715851604938507,
      "rewards/margins": 0.015852421522140503,
      "rewards/rejected": -0.012480835430324078,
      "step": 70
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.828760511501322e-06,
      "logits/chosen": -1.9334779977798462,
      "logits/rejected": -1.936608910560608,
      "logps/chosen": -34.305667877197266,
      "logps/rejected": -34.637855529785156,
      "loss": 5.6523,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.02442934736609459,
      "rewards/margins": 0.033519335091114044,
      "rewards/rejected": -0.009089985862374306,
      "step": 80
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.7367166013034295e-06,
      "logits/chosen": -1.9406198263168335,
      "logits/rejected": -1.9451316595077515,
      "logps/chosen": -32.37959289550781,
      "logps/rejected": -32.313934326171875,
      "loss": 6.1848,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.017566002905368805,
      "rewards/margins": 0.008571788668632507,
      "rewards/rejected": 0.008994214236736298,
      "step": 90
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.626245458345211e-06,
      "logits/chosen": -2.0386805534362793,
      "logits/rejected": -2.036684513092041,
      "logps/chosen": -32.129981994628906,
      "logps/rejected": -31.296749114990234,
      "loss": 5.6745,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.022171173244714737,
      "rewards/margins": 0.028278371319174767,
      "rewards/rejected": -0.006107199937105179,
      "step": 100
    },
    {
      "epoch": 0.26,
      "eval_logits/chosen": -2.2338244915008545,
      "eval_logits/rejected": -2.2289819717407227,
      "eval_logps/chosen": -34.01533508300781,
      "eval_logps/rejected": -37.518131256103516,
      "eval_loss": 6.2508440017700195,
      "eval_rewards/accuracies": 0.5460963845252991,
      "eval_rewards/chosen": 0.0038431365974247456,
      "eval_rewards/margins": 0.0041458746418356895,
      "eval_rewards/rejected": -0.0003027375496458262,
      "eval_runtime": 145.7849,
      "eval_samples_per_second": 2.353,
      "eval_steps_per_second": 0.295,
      "step": 100
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.498257201263691e-06,
      "logits/chosen": -1.9937114715576172,
      "logits/rejected": -1.9913352727890015,
      "logps/chosen": -33.10139465332031,
      "logps/rejected": -33.98957061767578,
      "loss": 6.1578,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.028484445065259933,
      "rewards/margins": 0.02052464708685875,
      "rewards/rejected": 0.007959800772368908,
      "step": 110
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.353806263777678e-06,
      "logits/chosen": -2.0058560371398926,
      "logits/rejected": -1.9975417852401733,
      "logps/chosen": -32.32807159423828,
      "logps/rejected": -32.11988067626953,
      "loss": 6.0258,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.02336641401052475,
      "rewards/margins": 0.015041453763842583,
      "rewards/rejected": 0.008324960246682167,
      "step": 120
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.1940827077152755e-06,
      "logits/chosen": -2.034005641937256,
      "logits/rejected": -2.026031494140625,
      "logps/chosen": -30.320354461669922,
      "logps/rejected": -32.04728698730469,
      "loss": 5.7894,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 0.031072016805410385,
      "rewards/margins": 0.029927905648946762,
      "rewards/rejected": 0.0011441137176007032,
      "step": 130
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.0204024186666215e-06,
      "logits/chosen": -1.9651100635528564,
      "logits/rejected": -1.9753506183624268,
      "logps/chosen": -31.207500457763672,
      "logps/rejected": -32.54130172729492,
      "loss": 5.4157,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.042432788759469986,
      "rewards/margins": 0.0437164343893528,
      "rewards/rejected": -0.0012836471432819963,
      "step": 140
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.834196265035119e-06,
      "logits/chosen": -1.8767722845077515,
      "logits/rejected": -1.8779083490371704,
      "logps/chosen": -33.90843963623047,
      "logps/rejected": -34.76317596435547,
      "loss": 5.2936,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.05685017257928848,
      "rewards/margins": 0.058033354580402374,
      "rewards/rejected": -0.0011831853771582246,
      "step": 150
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.636998309800573e-06,
      "logits/chosen": -1.9284919500350952,
      "logits/rejected": -1.9250543117523193,
      "logps/chosen": -36.01051712036133,
      "logps/rejected": -32.69367980957031,
      "loss": 5.7928,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.032231830060482025,
      "rewards/margins": 0.026417434215545654,
      "rewards/rejected": 0.005814394913613796,
      "step": 160
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4304331721118078e-06,
      "logits/chosen": -2.029017925262451,
      "logits/rejected": -2.0216681957244873,
      "logps/chosen": -33.47245407104492,
      "logps/rejected": -31.359905242919922,
      "loss": 5.0567,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.0650816410779953,
      "rewards/margins": 0.06669269502162933,
      "rewards/rejected": -0.0016110436990857124,
      "step": 170
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.2162026428305436e-06,
      "logits/chosen": -2.0359654426574707,
      "logits/rejected": -2.041189670562744,
      "logps/chosen": -32.1973876953125,
      "logps/rejected": -32.39836883544922,
      "loss": 5.2683,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.0694805309176445,
      "rewards/margins": 0.05212607979774475,
      "rewards/rejected": 0.017354462295770645,
      "step": 180
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.996071664294641e-06,
      "logits/chosen": -2.0371925830841064,
      "logits/rejected": -2.0344390869140625,
      "logps/chosen": -31.23935317993164,
      "logps/rejected": -31.283512115478516,
      "loss": 5.5783,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.049920208752155304,
      "rewards/margins": 0.04297895357012749,
      "rewards/rejected": 0.006941256113350391,
      "step": 190
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.7718537898066833e-06,
      "logits/chosen": -1.9059925079345703,
      "logits/rejected": -1.9106495380401611,
      "logps/chosen": -31.285137176513672,
      "logps/rejected": -32.777244567871094,
      "loss": 5.2135,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.064228855073452,
      "rewards/margins": 0.061491239815950394,
      "rewards/rejected": 0.002737621311098337,
      "step": 200
    },
    {
      "epoch": 0.52,
      "eval_logits/chosen": -2.231491804122925,
      "eval_logits/rejected": -2.2266557216644287,
      "eval_logps/chosen": -34.00144577026367,
      "eval_logps/rejected": -37.5041618347168,
      "eval_loss": 6.2880659103393555,
      "eval_rewards/accuracies": 0.5402824282646179,
      "eval_rewards/chosen": 0.006621644366532564,
      "eval_rewards/margins": 0.00412956066429615,
      "eval_rewards/rejected": 0.0024920827709138393,
      "eval_runtime": 145.8338,
      "eval_samples_per_second": 2.352,
      "eval_steps_per_second": 0.295,
      "step": 200
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.5453962426402006e-06,
      "logits/chosen": -2.017942428588867,
      "logits/rejected": -2.0285964012145996,
      "logps/chosen": -31.721837997436523,
      "logps/rejected": -33.87845993041992,
      "loss": 5.312,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.0498933307826519,
      "rewards/margins": 0.05304562300443649,
      "rewards/rejected": -0.003152288496494293,
      "step": 210
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.3185646976551794e-06,
      "logits/chosen": -1.9104692935943604,
      "logits/rejected": -1.9252119064331055,
      "logps/chosen": -29.818248748779297,
      "logps/rejected": -31.569311141967773,
      "loss": 5.1276,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.05865820124745369,
      "rewards/margins": 0.060884904116392136,
      "rewards/rejected": -0.002226702868938446,
      "step": 220
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.0932279108998323e-06,
      "logits/chosen": -1.9671802520751953,
      "logits/rejected": -1.9711711406707764,
      "logps/chosen": -33.06322479248047,
      "logps/rejected": -31.57196617126465,
      "loss": 4.9986,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.0708485022187233,
      "rewards/margins": 0.075381800532341,
      "rewards/rejected": -0.004533302970230579,
      "step": 230
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8712423238279358e-06,
      "logits/chosen": -1.9659277200698853,
      "logits/rejected": -1.9440826177597046,
      "logps/chosen": -33.80937957763672,
      "logps/rejected": -35.049232482910156,
      "loss": 4.6987,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.06887368112802505,
      "rewards/margins": 0.08856189250946045,
      "rewards/rejected": -0.019688209518790245,
      "step": 240
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.6544367689701824e-06,
      "logits/chosen": -2.0067532062530518,
      "logits/rejected": -2.0034396648406982,
      "logps/chosen": -32.64842987060547,
      "logps/rejected": -36.244728088378906,
      "loss": 5.1569,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.05532795190811157,
      "rewards/margins": 0.05928860232234001,
      "rewards/rejected": -0.003960648085922003,
      "step": 250
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.4445974030621963e-06,
      "logits/chosen": -1.8740017414093018,
      "logits/rejected": -1.8715832233428955,
      "logps/chosen": -33.948036193847656,
      "logps/rejected": -35.48664093017578,
      "loss": 5.4652,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.04807313531637192,
      "rewards/margins": 0.0444767102599144,
      "rewards/rejected": 0.0035964243579655886,
      "step": 260
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.243452991757889e-06,
      "logits/chosen": -1.8596279621124268,
      "logits/rejected": -1.8571679592132568,
      "logps/chosen": -34.14976119995117,
      "logps/rejected": -31.774593353271484,
      "loss": 5.3924,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.049567773938179016,
      "rewards/margins": 0.05022105574607849,
      "rewards/rejected": -0.0006532802362926304,
      "step": 270
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.0526606671603523e-06,
      "logits/chosen": -1.9632982015609741,
      "logits/rejected": -1.9527626037597656,
      "logps/chosen": -34.97089767456055,
      "logps/rejected": -31.810266494750977,
      "loss": 4.9049,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.07714872062206268,
      "rewards/margins": 0.07100304216146469,
      "rewards/rejected": 0.0061456747353076935,
      "step": 280
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.737922755071455e-07,
      "logits/chosen": -2.0585777759552,
      "logits/rejected": -2.043663740158081,
      "logps/chosen": -30.67743492126465,
      "logps/rejected": -32.60033416748047,
      "loss": 5.7039,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.04933839291334152,
      "rewards/margins": 0.039037786424160004,
      "rewards/rejected": 0.01030060462653637,
      "step": 290
    },
    {
      "epoch": 0.78,
      "learning_rate": 7.08321427484816e-07,
      "logits/chosen": -1.9290869235992432,
      "logits/rejected": -1.9265540838241577,
      "logps/chosen": -32.41482925415039,
      "logps/rejected": -30.851070404052734,
      "loss": 4.3883,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.1024569720029831,
      "rewards/margins": 0.10930945724248886,
      "rewards/rejected": -0.006852488033473492,
      "step": 300
    },
    {
      "epoch": 0.78,
      "eval_logits/chosen": -2.2290894985198975,
      "eval_logits/rejected": -2.2242588996887207,
      "eval_logps/chosen": -34.0192756652832,
      "eval_logps/rejected": -37.53245544433594,
      "eval_loss": 6.2381510734558105,
      "eval_rewards/accuracies": 0.5166113376617432,
      "eval_rewards/chosen": 0.0030557620339095592,
      "eval_rewards/margins": 0.0062218476086854935,
      "eval_rewards/rejected": -0.003166085807606578,
      "eval_runtime": 145.7657,
      "eval_samples_per_second": 2.353,
      "eval_steps_per_second": 0.295,
      "step": 300
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.576113578589035e-07,
      "logits/chosen": -1.9141037464141846,
      "logits/rejected": -1.9108575582504272,
      "logps/chosen": -31.29278564453125,
      "logps/rejected": -33.755470275878906,
      "loss": 5.1987,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.06211275979876518,
      "rewards/margins": 0.06477634608745575,
      "rewards/rejected": -0.002663586288690567,
      "step": 310
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.229036944380913e-07,
      "logits/chosen": -1.9642059803009033,
      "logits/rejected": -1.9519942998886108,
      "logps/chosen": -34.297203063964844,
      "logps/rejected": -33.615535736083984,
      "loss": 4.8567,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.05977395921945572,
      "rewards/margins": 0.07511334121227264,
      "rewards/rejected": -0.015339391306042671,
      "step": 320
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.053082288996112e-07,
      "logits/chosen": -1.9994945526123047,
      "logits/rejected": -1.9980732202529907,
      "logps/chosen": -33.15888214111328,
      "logps/rejected": -32.49894332885742,
      "loss": 5.0065,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.06538230925798416,
      "rewards/margins": 0.06925411522388458,
      "rewards/rejected": -0.0038718082942068577,
      "step": 330
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.0579377374915805e-07,
      "logits/chosen": -2.0863518714904785,
      "logits/rejected": -2.0706593990325928,
      "logps/chosen": -33.7581901550293,
      "logps/rejected": -33.056785583496094,
      "loss": 5.0199,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.08424069732427597,
      "rewards/margins": 0.07199026644229889,
      "rewards/rejected": 0.012250429019331932,
      "step": 340
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.2518018074041684e-07,
      "logits/chosen": -1.9585072994232178,
      "logits/rejected": -1.9576387405395508,
      "logps/chosen": -32.82634735107422,
      "logps/rejected": -32.51306915283203,
      "loss": 4.6899,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.0911494642496109,
      "rewards/margins": 0.10073409974575043,
      "rewards/rejected": -0.009584645740687847,
      "step": 350
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.41315865106129e-08,
      "logits/chosen": -1.9139328002929688,
      "logits/rejected": -1.9242264032363892,
      "logps/chosen": -31.86139488220215,
      "logps/rejected": -35.2977180480957,
      "loss": 4.9538,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.06865896284580231,
      "rewards/margins": 0.0704503208398819,
      "rewards/rejected": -0.001791359856724739,
      "step": 360
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.3150941078050325e-08,
      "logits/chosen": -2.0540366172790527,
      "logits/rejected": -2.047560214996338,
      "logps/chosen": -33.293235778808594,
      "logps/rejected": -29.21136474609375,
      "loss": 4.9289,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.0717458575963974,
      "rewards/margins": 0.06873348355293274,
      "rewards/rejected": 0.0030123672913759947,
      "step": 370
    },
    {
      "epoch": 0.99,
      "learning_rate": 2.575864278703266e-09,
      "logits/chosen": -1.913193702697754,
      "logits/rejected": -1.9154123067855835,
      "logps/chosen": -33.81034469604492,
      "logps/rejected": -30.896728515625,
      "loss": 4.6897,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.08574088662862778,
      "rewards/margins": 0.09056078642606735,
      "rewards/rejected": -0.004819901194423437,
      "step": 380
    },
    {
      "epoch": 1.0,
      "step": 385,
      "total_flos": 0.0,
      "train_loss": 5.437272009911475,
      "train_runtime": 3253.1823,
      "train_samples_per_second": 0.946,
      "train_steps_per_second": 0.118
    }
  ],
  "logging_steps": 10,
  "max_steps": 385,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}