{
  "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": 0.6931,
      "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": -1.8665436506271362,
      "logits/rejected": -1.8708542585372925,
      "logps/chosen": -36.982460021972656,
      "logps/rejected": -33.656585693359375,
      "loss": 0.6882,
      "rewards/accuracies": 0.5138888955116272,
      "rewards/chosen": 0.0048276386223733425,
      "rewards/margins": 0.010130541399121284,
      "rewards/rejected": -0.005302901845425367,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.564102564102564e-06,
      "logits/chosen": -1.9975639581680298,
      "logits/rejected": -2.000220537185669,
      "logps/chosen": -29.6324462890625,
      "logps/rejected": -29.057437896728516,
      "loss": 0.693,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.0019495869055390358,
      "rewards/margins": 0.0007479515625163913,
      "rewards/rejected": 0.0012016354594379663,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.846153846153847e-06,
      "logits/chosen": -1.9208921194076538,
      "logits/rejected": -1.918195128440857,
      "logps/chosen": -31.420969009399414,
      "logps/rejected": -33.22509765625,
      "loss": 0.6932,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.000974868715275079,
      "rewards/margins": 0.0003514128620736301,
      "rewards/rejected": -0.0013262818101793528,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.999896948438434e-06,
      "logits/chosen": -2.018131971359253,
      "logits/rejected": -2.0093750953674316,
      "logps/chosen": -32.56100082397461,
      "logps/rejected": -32.525699615478516,
      "loss": 0.6905,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.0031610552687197924,
      "rewards/margins": 0.0057663023471832275,
      "rewards/rejected": -0.002605247776955366,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.987541037542187e-06,
      "logits/chosen": -1.8627980947494507,
      "logits/rejected": -1.852043867111206,
      "logps/chosen": -33.57464599609375,
      "logps/rejected": -35.459083557128906,
      "loss": 0.6945,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": -0.003433463629335165,
      "rewards/margins": -0.0020852875895798206,
      "rewards/rejected": -0.0013481763890013099,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.954691471941119e-06,
      "logits/chosen": -1.9412128925323486,
      "logits/rejected": -1.9431606531143188,
      "logps/chosen": -32.51996612548828,
      "logps/rejected": -33.226768493652344,
      "loss": 0.6806,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.01604365184903145,
      "rewards/margins": 0.02712194062769413,
      "rewards/rejected": -0.011078287847340107,
      "step": 60
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.901618883413549e-06,
      "logits/chosen": -2.072383165359497,
      "logits/rejected": -2.0773634910583496,
      "logps/chosen": -33.964813232421875,
      "logps/rejected": -36.62574005126953,
      "loss": 0.6861,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.004735334776341915,
      "rewards/margins": 0.016109289601445198,
      "rewards/rejected": -0.011373954825103283,
      "step": 70
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.828760511501322e-06,
      "logits/chosen": -1.9332096576690674,
      "logits/rejected": -1.9363422393798828,
      "logps/chosen": -34.310150146484375,
      "logps/rejected": -34.669891357421875,
      "loss": 0.6749,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.023532329127192497,
      "rewards/margins": 0.03902902454137802,
      "rewards/rejected": -0.015496693551540375,
      "step": 80
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.7367166013034295e-06,
      "logits/chosen": -1.9401609897613525,
      "logits/rejected": -1.944685935974121,
      "logps/chosen": -32.39472961425781,
      "logps/rejected": -32.360809326171875,
      "loss": 0.6865,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.01453828252851963,
      "rewards/margins": 0.014918209984898567,
      "rewards/rejected": -0.000379929319024086,
      "step": 90
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.626245458345211e-06,
      "logits/chosen": -2.0377113819122314,
      "logits/rejected": -2.035726308822632,
      "logps/chosen": -32.14560317993164,
      "logps/rejected": -31.297948837280273,
      "loss": 0.6813,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.01904645375907421,
      "rewards/margins": 0.0253940187394619,
      "rewards/rejected": -0.006347564049065113,
      "step": 100
    },
    {
      "epoch": 0.26,
      "eval_logits/chosen": -2.231863498687744,
      "eval_logits/rejected": -2.227025270462036,
      "eval_logps/chosen": -34.02951431274414,
      "eval_logps/rejected": -37.534454345703125,
      "eval_loss": 0.6914929151535034,
      "eval_rewards/accuracies": 0.5220099687576294,
      "eval_rewards/chosen": 0.0010076783364638686,
      "eval_rewards/margins": 0.004574176389724016,
      "eval_rewards/rejected": -0.0035664979368448257,
      "eval_runtime": 145.7925,
      "eval_samples_per_second": 2.353,
      "eval_steps_per_second": 0.295,
      "step": 100
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.498257201263691e-06,
      "logits/chosen": -1.992661476135254,
      "logits/rejected": -1.9902756214141846,
      "logps/chosen": -33.13728713989258,
      "logps/rejected": -34.051490783691406,
      "loss": 0.6834,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.02130572497844696,
      "rewards/margins": 0.025729293003678322,
      "rewards/rejected": -0.004423565696924925,
      "step": 110
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.353806263777678e-06,
      "logits/chosen": -2.0037741661071777,
      "logits/rejected": -1.995439887046814,
      "logps/chosen": -32.33051681518555,
      "logps/rejected": -32.13713836669922,
      "loss": 0.6853,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.022878289222717285,
      "rewards/margins": 0.01800495944917202,
      "rewards/rejected": 0.004873327445238829,
      "step": 120
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.1940827077152755e-06,
      "logits/chosen": -2.032257318496704,
      "logits/rejected": -2.024292230606079,
      "logps/chosen": -30.306774139404297,
      "logps/rejected": -32.081024169921875,
      "loss": 0.6754,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.03378837928175926,
      "rewards/margins": 0.03939158096909523,
      "rewards/rejected": -0.00560319609940052,
      "step": 130
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.0204024186666215e-06,
      "logits/chosen": -1.9630043506622314,
      "logits/rejected": -1.973233938217163,
      "logps/chosen": -31.224035263061523,
      "logps/rejected": -32.56679916381836,
      "loss": 0.6718,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.0391254760324955,
      "rewards/margins": 0.045508723706007004,
      "rewards/rejected": -0.006383246276527643,
      "step": 140
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.834196265035119e-06,
      "logits/chosen": -1.874066710472107,
      "logits/rejected": -1.8752658367156982,
      "logps/chosen": -33.90449905395508,
      "logps/rejected": -34.78742218017578,
      "loss": 0.6642,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.05763882398605347,
      "rewards/margins": 0.06367169320583344,
      "rewards/rejected": -0.006032869219779968,
      "step": 150
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.636998309800573e-06,
      "logits/chosen": -1.9251247644424438,
      "logits/rejected": -1.9217197895050049,
      "logps/chosen": -36.002262115478516,
      "logps/rejected": -32.70859909057617,
      "loss": 0.6788,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.03388362005352974,
      "rewards/margins": 0.031053191050887108,
      "rewards/rejected": 0.0028304329607635736,
      "step": 160
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4304331721118078e-06,
      "logits/chosen": -2.025562286376953,
      "logits/rejected": -2.018228530883789,
      "logps/chosen": -33.46672439575195,
      "logps/rejected": -31.409042358398438,
      "loss": 0.657,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.06622789800167084,
      "rewards/margins": 0.07766623049974442,
      "rewards/rejected": -0.011438337154686451,
      "step": 170
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.2162026428305436e-06,
      "logits/chosen": -2.0323550701141357,
      "logits/rejected": -2.037605047225952,
      "logps/chosen": -32.23218536376953,
      "logps/rejected": -32.439491271972656,
      "loss": 0.6682,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.06252055615186691,
      "rewards/margins": 0.05339093878865242,
      "rewards/rejected": 0.009129621088504791,
      "step": 180
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.996071664294641e-06,
      "logits/chosen": -2.0334537029266357,
      "logits/rejected": -2.030702590942383,
      "logps/chosen": -31.266775131225586,
      "logps/rejected": -31.347997665405273,
      "loss": 0.67,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.04443611577153206,
      "rewards/margins": 0.050393205136060715,
      "rewards/rejected": -0.005957084707915783,
      "step": 190
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.7718537898066833e-06,
      "logits/chosen": -1.9037307500839233,
      "logits/rejected": -1.9083878993988037,
      "logps/chosen": -31.30868148803711,
      "logps/rejected": -32.823123931884766,
      "loss": 0.6632,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.05951983854174614,
      "rewards/margins": 0.0659579336643219,
      "rewards/rejected": -0.006438094191253185,
      "step": 200
    },
    {
      "epoch": 0.52,
      "eval_logits/chosen": -2.22880482673645,
      "eval_logits/rejected": -2.2239701747894287,
      "eval_logps/chosen": -34.05141067504883,
      "eval_logps/rejected": -37.58522033691406,
      "eval_loss": 0.6888246536254883,
      "eval_rewards/accuracies": 0.5660299062728882,
      "eval_rewards/chosen": -0.003371814964339137,
      "eval_rewards/margins": 0.01034836657345295,
      "eval_rewards/rejected": -0.013720180839300156,
      "eval_runtime": 145.665,
      "eval_samples_per_second": 2.355,
      "eval_steps_per_second": 0.295,
      "step": 200
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.5453962426402006e-06,
      "logits/chosen": -2.016369342803955,
      "logits/rejected": -2.027027130126953,
      "logps/chosen": -31.75009536743164,
      "logps/rejected": -33.965911865234375,
      "loss": 0.6631,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.04424174129962921,
      "rewards/margins": 0.06488426774740219,
      "rewards/rejected": -0.020642530173063278,
      "step": 210
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.3185646976551794e-06,
      "logits/chosen": -1.909157156944275,
      "logits/rejected": -1.9239362478256226,
      "logps/chosen": -29.8031005859375,
      "logps/rejected": -31.647253036499023,
      "loss": 0.6564,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.06168792396783829,
      "rewards/margins": 0.07950346171855927,
      "rewards/rejected": -0.017815548926591873,
      "step": 220
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.0932279108998323e-06,
      "logits/chosen": -1.9656283855438232,
      "logits/rejected": -1.9696086645126343,
      "logps/chosen": -33.09147644042969,
      "logps/rejected": -31.6539306640625,
      "loss": 0.6543,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.06519898772239685,
      "rewards/margins": 0.08612470328807831,
      "rewards/rejected": -0.02092570997774601,
      "step": 230
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8712423238279358e-06,
      "logits/chosen": -1.963721513748169,
      "logits/rejected": -1.9419059753417969,
      "logps/chosen": -33.84326934814453,
      "logps/rejected": -35.13850784301758,
      "loss": 0.6479,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.06209472939372063,
      "rewards/margins": 0.0996372401714325,
      "rewards/rejected": -0.03754251450300217,
      "step": 240
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.6544367689701824e-06,
      "logits/chosen": -2.004868268966675,
      "logits/rejected": -2.0015549659729004,
      "logps/chosen": -32.729896545410156,
      "logps/rejected": -36.27812957763672,
      "loss": 0.6706,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.03903444483876228,
      "rewards/margins": 0.04967557638883591,
      "rewards/rejected": -0.010641133412718773,
      "step": 250
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.4445974030621963e-06,
      "logits/chosen": -1.8725497722625732,
      "logits/rejected": -1.8701589107513428,
      "logps/chosen": -33.99323654174805,
      "logps/rejected": -35.557159423828125,
      "loss": 0.6706,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.039031852036714554,
      "rewards/margins": 0.04953894019126892,
      "rewards/rejected": -0.010507088154554367,
      "step": 260
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.243452991757889e-06,
      "logits/chosen": -1.8571383953094482,
      "logits/rejected": -1.854741096496582,
      "logps/chosen": -34.20746612548828,
      "logps/rejected": -31.83416748046875,
      "loss": 0.6708,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.0380266010761261,
      "rewards/margins": 0.05059366300702095,
      "rewards/rejected": -0.01256705541163683,
      "step": 270
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.0526606671603523e-06,
      "logits/chosen": -1.9600715637207031,
      "logits/rejected": -1.9495933055877686,
      "logps/chosen": -34.99934387207031,
      "logps/rejected": -31.900096893310547,
      "loss": 0.6544,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.07145937532186508,
      "rewards/margins": 0.08328022807836533,
      "rewards/rejected": -0.011820845305919647,
      "step": 280
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.737922755071455e-07,
      "logits/chosen": -2.0559241771698,
      "logits/rejected": -2.04101300239563,
      "logps/chosen": -30.744888305664062,
      "logps/rejected": -32.62079620361328,
      "loss": 0.6809,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.035847391933202744,
      "rewards/margins": 0.029639026150107384,
      "rewards/rejected": 0.006208371836692095,
      "step": 290
    },
    {
      "epoch": 0.78,
      "learning_rate": 7.08321427484816e-07,
      "logits/chosen": -1.9259971380233765,
      "logits/rejected": -1.9234883785247803,
      "logps/chosen": -32.35750961303711,
      "logps/rejected": -30.92331314086914,
      "loss": 0.6327,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.11392021179199219,
      "rewards/margins": 0.13522081077098846,
      "rewards/rejected": -0.021300604566931725,
      "step": 300
    },
    {
      "epoch": 0.78,
      "eval_logits/chosen": -2.2262163162231445,
      "eval_logits/rejected": -2.2213830947875977,
      "eval_logps/chosen": -34.093963623046875,
      "eval_logps/rejected": -37.60972213745117,
      "eval_loss": 0.6908519268035889,
      "eval_rewards/accuracies": 0.49501657485961914,
      "eval_rewards/chosen": -0.011881927028298378,
      "eval_rewards/margins": 0.006738076452165842,
      "eval_rewards/rejected": -0.018620004877448082,
      "eval_runtime": 145.4585,
      "eval_samples_per_second": 2.358,
      "eval_steps_per_second": 0.296,
      "step": 300
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.576113578589035e-07,
      "logits/chosen": -1.9112951755523682,
      "logits/rejected": -1.9080655574798584,
      "logps/chosen": -31.314626693725586,
      "logps/rejected": -33.85969543457031,
      "loss": 0.6566,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.05774437263607979,
      "rewards/margins": 0.08125325292348862,
      "rewards/rejected": -0.023508887737989426,
      "step": 310
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.229036944380913e-07,
      "logits/chosen": -1.960738182067871,
      "logits/rejected": -1.9485470056533813,
      "logps/chosen": -34.32622528076172,
      "logps/rejected": -33.65083312988281,
      "loss": 0.6586,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.0539703294634819,
      "rewards/margins": 0.0763682946562767,
      "rewards/rejected": -0.022397976368665695,
      "step": 320
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.053082288996112e-07,
      "logits/chosen": -1.9961421489715576,
      "logits/rejected": -1.9947240352630615,
      "logps/chosen": -33.134159088134766,
      "logps/rejected": -32.56267547607422,
      "loss": 0.6536,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.07032694667577744,
      "rewards/margins": 0.08694546669721603,
      "rewards/rejected": -0.0166185162961483,
      "step": 330
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.0579377374915805e-07,
      "logits/chosen": -2.083061933517456,
      "logits/rejected": -2.0673813819885254,
      "logps/chosen": -33.78687286376953,
      "logps/rejected": -33.12049102783203,
      "loss": 0.6568,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.07850398868322372,
      "rewards/margins": 0.07899539172649384,
      "rewards/rejected": -0.0004914018791168928,
      "step": 340
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.2518018074041684e-07,
      "logits/chosen": -1.9557991027832031,
      "logits/rejected": -1.9549639225006104,
      "logps/chosen": -32.84474563598633,
      "logps/rejected": -32.53619384765625,
      "loss": 0.648,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.0874704122543335,
      "rewards/margins": 0.10167930275201797,
      "rewards/rejected": -0.014208881184458733,
      "step": 350
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.41315865106129e-08,
      "logits/chosen": -1.9108169078826904,
      "logits/rejected": -1.9210844039916992,
      "logps/chosen": -31.8836612701416,
      "logps/rejected": -35.36156463623047,
      "loss": 0.6572,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.06420587003231049,
      "rewards/margins": 0.07876714318990707,
      "rewards/rejected": -0.014561265707015991,
      "step": 360
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.3150941078050325e-08,
      "logits/chosen": -2.0503978729248047,
      "logits/rejected": -2.043942928314209,
      "logps/chosen": -33.33507537841797,
      "logps/rejected": -29.28571128845215,
      "loss": 0.6582,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.06337906420230865,
      "rewards/margins": 0.07523626834154129,
      "rewards/rejected": -0.011857211589813232,
      "step": 370
    },
    {
      "epoch": 0.99,
      "learning_rate": 2.575864278703266e-09,
      "logits/chosen": -1.9104009866714478,
      "logits/rejected": -1.9126161336898804,
      "logps/chosen": -33.88220977783203,
      "logps/rejected": -30.987438201904297,
      "loss": 0.65,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.07136790454387665,
      "rewards/margins": 0.09433047473430634,
      "rewards/rejected": -0.02296256460249424,
      "step": 380
    },
    {
      "epoch": 1.0,
      "step": 385,
      "total_flos": 0.0,
      "train_loss": 0.6689951481757226,
      "train_runtime": 3249.3574,
      "train_samples_per_second": 0.948,
      "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
}