{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 252,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 16.79678467461835,
      "learning_rate": 1.923076923076923e-08,
      "logits/chosen": -0.5216625928878784,
      "logits/rejected": -1.6251907348632812,
      "logps/chosen": -339.42877197265625,
      "logps/rejected": -263.98431396484375,
      "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": 17.217350690399922,
      "learning_rate": 1.9230769230769231e-07,
      "logits/chosen": -1.7395856380462646,
      "logits/rejected": -1.1197137832641602,
      "logps/chosen": -194.7418212890625,
      "logps/rejected": -322.9896240234375,
      "loss": 0.693,
      "rewards/accuracies": 0.5069444179534912,
      "rewards/chosen": 0.0007130156154744327,
      "rewards/margins": 0.0009009492350742221,
      "rewards/rejected": -0.00018793345952872187,
      "step": 10
    },
    {
      "epoch": 0.08,
      "grad_norm": 17.755504451917034,
      "learning_rate": 3.8461538461538463e-07,
      "logits/chosen": -1.3529917001724243,
      "logits/rejected": -1.1581436395645142,
      "logps/chosen": -250.0107879638672,
      "logps/rejected": -309.7054138183594,
      "loss": 0.69,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.003061536233872175,
      "rewards/margins": 0.0077318595722317696,
      "rewards/rejected": -0.004670322872698307,
      "step": 20
    },
    {
      "epoch": 0.12,
      "grad_norm": 15.809698317809115,
      "learning_rate": 4.99613632163459e-07,
      "logits/chosen": -1.445340633392334,
      "logits/rejected": -0.8237818479537964,
      "logps/chosen": -262.8525695800781,
      "logps/rejected": -319.0481872558594,
      "loss": 0.6756,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": 0.011689678765833378,
      "rewards/margins": 0.03540501371026039,
      "rewards/rejected": -0.02371532842516899,
      "step": 30
    },
    {
      "epoch": 0.16,
      "grad_norm": 17.829217597278124,
      "learning_rate": 4.952806974561517e-07,
      "logits/chosen": -1.3351449966430664,
      "logits/rejected": -0.8461858034133911,
      "logps/chosen": -231.9071807861328,
      "logps/rejected": -339.33740234375,
      "loss": 0.6256,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 0.006707000080496073,
      "rewards/margins": 0.18326039612293243,
      "rewards/rejected": -0.1765533834695816,
      "step": 40
    },
    {
      "epoch": 0.2,
      "grad_norm": 20.372527502778215,
      "learning_rate": 4.862157403595598e-07,
      "logits/chosen": -1.0431455373764038,
      "logits/rejected": -0.7218812704086304,
      "logps/chosen": -242.2100372314453,
      "logps/rejected": -352.3864440917969,
      "loss": 0.5443,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -0.04869229719042778,
      "rewards/margins": 0.5344938039779663,
      "rewards/rejected": -0.5831860303878784,
      "step": 50
    },
    {
      "epoch": 0.24,
      "grad_norm": 24.684374727803654,
      "learning_rate": 4.725936445085709e-07,
      "logits/chosen": -0.7200717329978943,
      "logits/rejected": -0.5803043842315674,
      "logps/chosen": -277.9634094238281,
      "logps/rejected": -517.9424438476562,
      "loss": 0.4741,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": -0.17333514988422394,
      "rewards/margins": 1.385983943939209,
      "rewards/rejected": -1.559319019317627,
      "step": 60
    },
    {
      "epoch": 0.28,
      "grad_norm": 22.14152988454074,
      "learning_rate": 4.5467721110696685e-07,
      "logits/chosen": -0.515616774559021,
      "logits/rejected": -0.04089225083589554,
      "logps/chosen": -258.40606689453125,
      "logps/rejected": -408.2468566894531,
      "loss": 0.4328,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": -0.24228188395500183,
      "rewards/margins": 1.2516155242919922,
      "rewards/rejected": -1.493897557258606,
      "step": 70
    },
    {
      "epoch": 0.32,
      "grad_norm": 18.065752819849052,
      "learning_rate": 4.328120888946271e-07,
      "logits/chosen": -0.7557204961776733,
      "logits/rejected": -0.010405841283500195,
      "logps/chosen": -258.38092041015625,
      "logps/rejected": -523.7661743164062,
      "loss": 0.4252,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": -0.33308929204940796,
      "rewards/margins": 2.278934955596924,
      "rewards/rejected": -2.6120240688323975,
      "step": 80
    },
    {
      "epoch": 0.36,
      "grad_norm": 15.734543008245188,
      "learning_rate": 4.074201057973785e-07,
      "logits/chosen": -1.038962721824646,
      "logits/rejected": 0.15530693531036377,
      "logps/chosen": -266.2573547363281,
      "logps/rejected": -538.8092041015625,
      "loss": 0.3708,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.19907574355602264,
      "rewards/margins": 2.0345568656921387,
      "rewards/rejected": -2.2336325645446777,
      "step": 90
    },
    {
      "epoch": 0.4,
      "grad_norm": 24.698688306800403,
      "learning_rate": 3.789911309071252e-07,
      "logits/chosen": -0.2497592717409134,
      "logits/rejected": 0.24183444678783417,
      "logps/chosen": -277.8612365722656,
      "logps/rejected": -582.8795776367188,
      "loss": 0.34,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": -0.30463331937789917,
      "rewards/margins": 2.518612861633301,
      "rewards/rejected": -2.823246479034424,
      "step": 100
    },
    {
      "epoch": 0.4,
      "eval_logits/chosen": -1.5214157104492188,
      "eval_logits/rejected": 0.016294823959469795,
      "eval_logps/chosen": -229.1130828857422,
      "eval_logps/rejected": -291.0631408691406,
      "eval_loss": 0.5715546011924744,
      "eval_rewards/accuracies": 0.7202380895614624,
      "eval_rewards/chosen": -0.2893332839012146,
      "eval_rewards/margins": 0.37313312292099,
      "eval_rewards/rejected": -0.6624664068222046,
      "eval_runtime": 410.8502,
      "eval_samples_per_second": 9.789,
      "eval_steps_per_second": 0.307,
      "step": 100
    },
    {
      "epoch": 0.44,
      "grad_norm": 20.355882133570365,
      "learning_rate": 3.4807362379317026e-07,
      "logits/chosen": -0.35275131464004517,
      "logits/rejected": 0.4792235493659973,
      "logps/chosen": -295.2680358886719,
      "logps/rejected": -693.8294677734375,
      "loss": 0.3238,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": -0.3359094262123108,
      "rewards/margins": 3.903926372528076,
      "rewards/rejected": -4.239835739135742,
      "step": 110
    },
    {
      "epoch": 0.48,
      "grad_norm": 21.727867963945446,
      "learning_rate": 3.152640534699994e-07,
      "logits/chosen": -0.5010538697242737,
      "logits/rejected": 0.6018115878105164,
      "logps/chosen": -291.17840576171875,
      "logps/rejected": -668.4251098632812,
      "loss": 0.3106,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.3278064429759979,
      "rewards/margins": 3.2680602073669434,
      "rewards/rejected": -3.595867156982422,
      "step": 120
    },
    {
      "epoch": 0.52,
      "grad_norm": 37.26634759168863,
      "learning_rate": 2.811953911537022e-07,
      "logits/chosen": -0.57627934217453,
      "logits/rejected": 0.6331204175949097,
      "logps/chosen": -227.9058837890625,
      "logps/rejected": -711.4694213867188,
      "loss": 0.299,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": -0.33088651299476624,
      "rewards/margins": 3.8740921020507812,
      "rewards/rejected": -4.204977989196777,
      "step": 130
    },
    {
      "epoch": 0.56,
      "grad_norm": 38.01185078661505,
      "learning_rate": 2.4652489880792125e-07,
      "logits/chosen": 0.14344072341918945,
      "logits/rejected": 0.4568979740142822,
      "logps/chosen": -297.4254455566406,
      "logps/rejected": -691.5115356445312,
      "loss": 0.2841,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.29703259468078613,
      "rewards/margins": 3.58467173576355,
      "rewards/rejected": -3.881704330444336,
      "step": 140
    },
    {
      "epoch": 0.6,
      "grad_norm": 23.175417086747803,
      "learning_rate": 2.1192144906604874e-07,
      "logits/chosen": -0.35931748151779175,
      "logits/rejected": 0.38914966583251953,
      "logps/chosen": -250.97146606445312,
      "logps/rejected": -848.0675659179688,
      "loss": 0.2553,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": -0.24952073395252228,
      "rewards/margins": 5.0541534423828125,
      "rewards/rejected": -5.303674221038818,
      "step": 150
    },
    {
      "epoch": 0.63,
      "grad_norm": 20.4769087360566,
      "learning_rate": 1.780526211572016e-07,
      "logits/chosen": -0.17432162165641785,
      "logits/rejected": 0.5454779267311096,
      "logps/chosen": -261.4326477050781,
      "logps/rejected": -773.8187866210938,
      "loss": 0.2334,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": -0.21171124279499054,
      "rewards/margins": 4.594513416290283,
      "rewards/rejected": -4.806224822998047,
      "step": 160
    },
    {
      "epoch": 0.67,
      "grad_norm": 37.16207244417959,
      "learning_rate": 1.4557182178490635e-07,
      "logits/chosen": -0.05513007566332817,
      "logits/rejected": 0.6765660047531128,
      "logps/chosen": -286.18804931640625,
      "logps/rejected": -864.0691528320312,
      "loss": 0.2465,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.3841165602207184,
      "rewards/margins": 5.005141258239746,
      "rewards/rejected": -5.389257431030273,
      "step": 170
    },
    {
      "epoch": 0.71,
      "grad_norm": 18.25462627863088,
      "learning_rate": 1.1510567942602889e-07,
      "logits/chosen": -0.39599961042404175,
      "logits/rejected": 0.6781516075134277,
      "logps/chosen": -263.38983154296875,
      "logps/rejected": -744.1962280273438,
      "loss": 0.2155,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.31235271692276,
      "rewards/margins": 4.253737926483154,
      "rewards/rejected": -4.5660905838012695,
      "step": 180
    },
    {
      "epoch": 0.75,
      "grad_norm": 41.13316289752075,
      "learning_rate": 8.724195524258688e-08,
      "logits/chosen": -0.2813408672809601,
      "logits/rejected": 0.8199517130851746,
      "logps/chosen": -277.7645263671875,
      "logps/rejected": -847.7828369140625,
      "loss": 0.2412,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": -0.3673214316368103,
      "rewards/margins": 5.06696081161499,
      "rewards/rejected": -5.434282302856445,
      "step": 190
    },
    {
      "epoch": 0.79,
      "grad_norm": 29.643198990740647,
      "learning_rate": 6.251820383244468e-08,
      "logits/chosen": -0.14953655004501343,
      "logits/rejected": 0.8468856811523438,
      "logps/chosen": -270.6844177246094,
      "logps/rejected": -735.6868286132812,
      "loss": 0.2189,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.2686000466346741,
      "rewards/margins": 4.150389194488525,
      "rewards/rejected": -4.418989658355713,
      "step": 200
    },
    {
      "epoch": 0.79,
      "eval_logits/chosen": -1.359634280204773,
      "eval_logits/rejected": 0.4307384192943573,
      "eval_logps/chosen": -248.32972717285156,
      "eval_logps/rejected": -347.9086608886719,
      "eval_loss": 0.5082926750183105,
      "eval_rewards/accuracies": 0.75,
      "eval_rewards/chosen": -0.48149970173835754,
      "eval_rewards/margins": 0.749422013759613,
      "eval_rewards/rejected": -1.2309216260910034,
      "eval_runtime": 409.9823,
      "eval_samples_per_second": 9.81,
      "eval_steps_per_second": 0.307,
      "step": 200
    },
    {
      "epoch": 0.83,
      "grad_norm": 18.630491734171713,
      "learning_rate": 4.141140257879319e-08,
      "logits/chosen": 0.14813140034675598,
      "logits/rejected": 0.9779118299484253,
      "logps/chosen": -279.84320068359375,
      "logps/rejected": -633.3928833007812,
      "loss": 0.2247,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": -0.1574528068304062,
      "rewards/margins": 3.489206314086914,
      "rewards/rejected": -3.6466591358184814,
      "step": 210
    },
    {
      "epoch": 0.87,
      "grad_norm": 28.497946746927017,
      "learning_rate": 2.4328749671846117e-08,
      "logits/chosen": -0.18863503634929657,
      "logits/rejected": 0.7909995317459106,
      "logps/chosen": -280.1761779785156,
      "logps/rejected": -809.5506591796875,
      "loss": 0.2107,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.3041425049304962,
      "rewards/margins": 4.856934547424316,
      "rewards/rejected": -5.161076545715332,
      "step": 220
    },
    {
      "epoch": 0.91,
      "grad_norm": 23.671408143492002,
      "learning_rate": 1.1599808329836174e-08,
      "logits/chosen": 0.07416832447052002,
      "logits/rejected": 0.9995840787887573,
      "logps/chosen": -275.8365783691406,
      "logps/rejected": -751.8892822265625,
      "loss": 0.2143,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": -0.24566996097564697,
      "rewards/margins": 4.163148403167725,
      "rewards/rejected": -4.408819198608398,
      "step": 230
    },
    {
      "epoch": 0.95,
      "grad_norm": 27.856145151796408,
      "learning_rate": 3.4701487751534475e-09,
      "logits/chosen": -0.11121706664562225,
      "logits/rejected": 0.7320507168769836,
      "logps/chosen": -341.3576965332031,
      "logps/rejected": -890.82421875,
      "loss": 0.215,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": -0.3052319586277008,
      "rewards/margins": 5.738637924194336,
      "rewards/rejected": -6.043869972229004,
      "step": 240
    },
    {
      "epoch": 0.99,
      "grad_norm": 25.41636406288576,
      "learning_rate": 9.661062636148743e-11,
      "logits/chosen": -0.6346914172172546,
      "logits/rejected": 0.8026138544082642,
      "logps/chosen": -268.1880187988281,
      "logps/rejected": -801.5145263671875,
      "loss": 0.186,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -0.3312497138977051,
      "rewards/margins": 3.949364185333252,
      "rewards/rejected": -4.280614376068115,
      "step": 250
    },
    {
      "epoch": 1.0,
      "step": 252,
      "total_flos": 0.0,
      "train_loss": 0.35703677506673903,
      "train_runtime": 4562.5722,
      "train_samples_per_second": 3.533,
      "train_steps_per_second": 0.055
    }
  ],
  "logging_steps": 10,
  "max_steps": 252,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}