{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 285,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 2.671875,
      "learning_rate": 1.7241379310344828e-07,
      "logits/chosen": -2.715578079223633,
      "logits/rejected": -3.198270320892334,
      "logps/chosen": -181.673828125,
      "logps/rejected": -250.83145141601562,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.75,
      "learning_rate": 1.724137931034483e-06,
      "logits/chosen": -2.6499929428100586,
      "logits/rejected": -3.116462230682373,
      "logps/chosen": -179.62359619140625,
      "logps/rejected": -252.53973388671875,
      "loss": 0.6925,
      "rewards/accuracies": 0.6111111044883728,
      "rewards/chosen": 0.0010814073029905558,
      "rewards/margins": 0.0013816155260428786,
      "rewards/rejected": -0.00030020822305232286,
      "step": 10
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.5625,
      "learning_rate": 3.448275862068966e-06,
      "logits/chosen": -2.6711201667785645,
      "logits/rejected": -3.1086363792419434,
      "logps/chosen": -176.72164916992188,
      "logps/rejected": -252.67080688476562,
      "loss": 0.6743,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.023805765435099602,
      "rewards/margins": 0.03818682208657265,
      "rewards/rejected": -0.014381052926182747,
      "step": 20
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.78125,
      "learning_rate": 4.999811754597862e-06,
      "logits/chosen": -2.642437696456909,
      "logits/rejected": -3.1196322441101074,
      "logps/chosen": -167.26516723632812,
      "logps/rejected": -260.6128845214844,
      "loss": 0.6003,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.1062479019165039,
      "rewards/margins": 0.19668832421302795,
      "rewards/rejected": -0.09044040739536285,
      "step": 30
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4921875,
      "learning_rate": 4.97725658856945e-06,
      "logits/chosen": -2.6548495292663574,
      "logits/rejected": -3.105985403060913,
      "logps/chosen": -157.27615356445312,
      "logps/rejected": -282.587646484375,
      "loss": 0.4653,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.22060635685920715,
      "rewards/margins": 0.5277924537658691,
      "rewards/rejected": -0.30718618631362915,
      "step": 40
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.15625,
      "learning_rate": 4.917441177612131e-06,
      "logits/chosen": -2.6114723682403564,
      "logits/rejected": -3.054443120956421,
      "logps/chosen": -146.25564575195312,
      "logps/rejected": -317.78265380859375,
      "loss": 0.3246,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.314788281917572,
      "rewards/margins": 0.967236340045929,
      "rewards/rejected": -0.6524480581283569,
      "step": 50
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.97265625,
      "learning_rate": 4.82126520118304e-06,
      "logits/chosen": -2.6283130645751953,
      "logits/rejected": -3.036774158477783,
      "logps/chosen": -138.53135681152344,
      "logps/rejected": -358.7559509277344,
      "loss": 0.2051,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.41738367080688477,
      "rewards/margins": 1.4921326637268066,
      "rewards/rejected": -1.0747489929199219,
      "step": 60
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.69921875,
      "learning_rate": 4.6901752354885166e-06,
      "logits/chosen": -2.6115269660949707,
      "logits/rejected": -3.0476138591766357,
      "logps/chosen": -129.70721435546875,
      "logps/rejected": -413.200927734375,
      "loss": 0.1157,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.5094510316848755,
      "rewards/margins": 2.1198441982269287,
      "rewards/rejected": -1.6103931665420532,
      "step": 70
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.439453125,
      "learning_rate": 4.526142995631488e-06,
      "logits/chosen": -2.642693281173706,
      "logits/rejected": -3.0582263469696045,
      "logps/chosen": -123.13417053222656,
      "logps/rejected": -475.8509216308594,
      "loss": 0.0593,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.572167694568634,
      "rewards/margins": 2.8161401748657227,
      "rewards/rejected": -2.2439725399017334,
      "step": 80
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.306640625,
      "learning_rate": 4.331635679181032e-06,
      "logits/chosen": -2.616441249847412,
      "logits/rejected": -3.0742900371551514,
      "logps/chosen": -119.6812744140625,
      "logps/rejected": -535.9256591796875,
      "loss": 0.0314,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.61542809009552,
      "rewards/margins": 3.4634997844696045,
      "rewards/rejected": -2.848071813583374,
      "step": 90
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.224609375,
      "learning_rate": 4.109578857224478e-06,
      "logits/chosen": -2.6391441822052,
      "logits/rejected": -3.0866036415100098,
      "logps/chosen": -115.92496490478516,
      "logps/rejected": -598.8836059570312,
      "loss": 0.0166,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6349981427192688,
      "rewards/margins": 4.107515811920166,
      "rewards/rejected": -3.472517728805542,
      "step": 100
    },
    {
      "epoch": 0.35,
      "eval_logits/chosen": -2.680793523788452,
      "eval_logits/rejected": -3.105761766433716,
      "eval_logps/chosen": -96.58148956298828,
      "eval_logps/rejected": -579.1800537109375,
      "eval_loss": 0.01978849433362484,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 0.4884725511074066,
      "eval_rewards/margins": 3.9327571392059326,
      "eval_rewards/rejected": -3.444284439086914,
      "eval_runtime": 1.2795,
      "eval_samples_per_second": 3.908,
      "eval_steps_per_second": 2.345,
      "step": 100
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1416015625,
      "learning_rate": 3.863312471055116e-06,
      "logits/chosen": -2.5967559814453125,
      "logits/rejected": -3.022758722305298,
      "logps/chosen": -115.83799743652344,
      "logps/rejected": -666.7772827148438,
      "loss": 0.0084,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6381937265396118,
      "rewards/margins": 4.796027183532715,
      "rewards/rejected": -4.157833099365234,
      "step": 110
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.10400390625,
      "learning_rate": 3.5965405963463197e-06,
      "logits/chosen": -2.6488442420959473,
      "logits/rejected": -3.06384015083313,
      "logps/chosen": -113.5273208618164,
      "logps/rejected": -722.9757080078125,
      "loss": 0.0047,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6482743620872498,
      "rewards/margins": 5.373518943786621,
      "rewards/rejected": -4.725245475769043,
      "step": 120
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.07958984375,
      "learning_rate": 3.313275730405658e-06,
      "logits/chosen": -2.6456308364868164,
      "logits/rejected": -3.099287509918213,
      "logps/chosen": -114.421630859375,
      "logps/rejected": -765.133544921875,
      "loss": 0.0031,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6431270837783813,
      "rewards/margins": 5.790853977203369,
      "rewards/rejected": -5.147727012634277,
      "step": 130
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.056396484375,
      "learning_rate": 3.0177784404805466e-06,
      "logits/chosen": -2.6465344429016113,
      "logits/rejected": -3.0662121772766113,
      "logps/chosen": -115.8419418334961,
      "logps/rejected": -792.1008911132812,
      "loss": 0.0024,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.642734169960022,
      "rewards/margins": 6.044482231140137,
      "rewards/rejected": -5.401747703552246,
      "step": 140
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.06591796875,
      "learning_rate": 2.7144932808611002e-06,
      "logits/chosen": -2.597626209259033,
      "logits/rejected": -3.0344064235687256,
      "logps/chosen": -115.4371337890625,
      "logps/rejected": -800.051025390625,
      "loss": 0.0022,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6369476914405823,
      "rewards/margins": 6.132936000823975,
      "rewards/rejected": -5.495988368988037,
      "step": 150
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.057373046875,
      "learning_rate": 2.407981942646603e-06,
      "logits/chosen": -2.659702777862549,
      "logits/rejected": -3.115243434906006,
      "logps/chosen": -115.50006103515625,
      "logps/rejected": -816.8475341796875,
      "loss": 0.0019,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6319659948348999,
      "rewards/margins": 6.285677909851074,
      "rewards/rejected": -5.653711795806885,
      "step": 160
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.05029296875,
      "learning_rate": 2.102854641665347e-06,
      "logits/chosen": -2.6373775005340576,
      "logits/rejected": -3.08939790725708,
      "logps/chosen": -115.76806640625,
      "logps/rejected": -823.41650390625,
      "loss": 0.0018,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6388428807258606,
      "rewards/margins": 6.353511810302734,
      "rewards/rejected": -5.7146687507629395,
      "step": 170
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.052734375,
      "learning_rate": 1.8037007765373677e-06,
      "logits/chosen": -2.663148880004883,
      "logits/rejected": -3.0792253017425537,
      "logps/chosen": -116.19815826416016,
      "logps/rejected": -830.9599609375,
      "loss": 0.0016,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6323127150535583,
      "rewards/margins": 6.430493354797363,
      "rewards/rejected": -5.798180103302002,
      "step": 180
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.042724609375,
      "learning_rate": 1.5150198998473802e-06,
      "logits/chosen": -2.6086456775665283,
      "logits/rejected": -3.0565147399902344,
      "logps/chosen": -116.57249450683594,
      "logps/rejected": -834.7310791015625,
      "loss": 0.0016,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6343868970870972,
      "rewards/margins": 6.452385902404785,
      "rewards/rejected": -5.817999362945557,
      "step": 190
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.04443359375,
      "learning_rate": 1.2411540406857064e-06,
      "logits/chosen": -2.620572566986084,
      "logits/rejected": -3.073166608810425,
      "logps/chosen": -114.90836334228516,
      "logps/rejected": -838.9874877929688,
      "loss": 0.0015,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6308305263519287,
      "rewards/margins": 6.5031633377075195,
      "rewards/rejected": -5.872332572937012,
      "step": 200
    },
    {
      "epoch": 0.7,
      "eval_logits/chosen": -2.6815271377563477,
      "eval_logits/rejected": -3.1155121326446533,
      "eval_logps/chosen": -96.9991683959961,
      "eval_logps/rejected": -740.9360961914062,
      "eval_loss": 0.005451836623251438,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 0.4842957556247711,
      "eval_rewards/margins": 5.546140670776367,
      "eval_rewards/rejected": -5.061844825744629,
      "eval_runtime": 1.2858,
      "eval_samples_per_second": 3.889,
      "eval_steps_per_second": 2.333,
      "step": 200
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.044921875,
      "learning_rate": 9.862223964891864e-07,
      "logits/chosen": -2.659987449645996,
      "logits/rejected": -3.083799362182617,
      "logps/chosen": -114.6219482421875,
      "logps/rejected": -828.9215087890625,
      "loss": 0.002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6267526149749756,
      "rewards/margins": 6.408988952636719,
      "rewards/rejected": -5.782236576080322,
      "step": 210
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.041748046875,
      "learning_rate": 7.54059376477568e-07,
      "logits/chosen": -2.6489715576171875,
      "logits/rejected": -3.0990288257598877,
      "logps/chosen": -116.3112564086914,
      "logps/rejected": -840.1389770507812,
      "loss": 0.0015,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6445231437683105,
      "rewards/margins": 6.526413917541504,
      "rewards/rejected": -5.881890773773193,
      "step": 220
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.03955078125,
      "learning_rate": 5.481569285697638e-07,
      "logits/chosen": -2.6626932621002197,
      "logits/rejected": -3.06781005859375,
      "logps/chosen": -115.39896392822266,
      "logps/rejected": -836.1140747070312,
      "loss": 0.0016,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6213386058807373,
      "rewards/margins": 6.4759931564331055,
      "rewards/rejected": -5.854653835296631,
      "step": 230
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.038818359375,
      "learning_rate": 3.71612017236837e-07,
      "logits/chosen": -2.6279659271240234,
      "logits/rejected": -3.0500528812408447,
      "logps/chosen": -116.8084945678711,
      "logps/rejected": -845.3555908203125,
      "loss": 0.0014,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6305276155471802,
      "rewards/margins": 6.566590309143066,
      "rewards/rejected": -5.936062335968018,
      "step": 240
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.04443359375,
      "learning_rate": 2.2708004227369407e-07,
      "logits/chosen": -2.6384856700897217,
      "logits/rejected": -3.0602006912231445,
      "logps/chosen": -115.619140625,
      "logps/rejected": -842.5126953125,
      "loss": 0.0015,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6297703981399536,
      "rewards/margins": 6.527515411376953,
      "rewards/rejected": -5.897744655609131,
      "step": 250
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.041015625,
      "learning_rate": 1.1673489911451536e-07,
      "logits/chosen": -2.635697603225708,
      "logits/rejected": -3.0929272174835205,
      "logps/chosen": -116.23138427734375,
      "logps/rejected": -843.4679565429688,
      "loss": 0.0015,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6368875503540039,
      "rewards/margins": 6.544775485992432,
      "rewards/rejected": -5.907887935638428,
      "step": 260
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.037353515625,
      "learning_rate": 4.223628142195929e-08,
      "logits/chosen": -2.6346564292907715,
      "logits/rejected": -3.0773911476135254,
      "logps/chosen": -115.94686126708984,
      "logps/rejected": -848.7951049804688,
      "loss": 0.0014,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6426523923873901,
      "rewards/margins": 6.600114345550537,
      "rewards/rejected": -5.957463264465332,
      "step": 270
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.044677734375,
      "learning_rate": 4.704717749627052e-09,
      "logits/chosen": -2.6303486824035645,
      "logits/rejected": -3.084428310394287,
      "logps/chosen": -115.82963562011719,
      "logps/rejected": -837.4007568359375,
      "loss": 0.0015,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6384350061416626,
      "rewards/margins": 6.50278377532959,
      "rewards/rejected": -5.864348888397217,
      "step": 280
    },
    {
      "epoch": 1.0,
      "step": 285,
      "total_flos": 0.0,
      "train_loss": 0.11324987118704277,
      "train_runtime": 748.1588,
      "train_samples_per_second": 1.522,
      "train_steps_per_second": 0.381
    }
  ],
  "logging_steps": 10,
  "max_steps": 285,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 200,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}