{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 100,
  "global_step": 288,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010416666666666666,
      "grad_norm": 617.2018982452129,
      "learning_rate": 1.426036352613182e-08,
      "logits/chosen": -2.590585231781006,
      "logits/rejected": -2.5664222240448,
      "logps/chosen": -80.29847717285156,
      "logps/rejected": -53.10200881958008,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.10416666666666667,
      "grad_norm": 527.5551545265153,
      "learning_rate": 1.426036352613182e-07,
      "logits/chosen": -2.5563437938690186,
      "logits/rejected": -2.5385162830352783,
      "logps/chosen": -87.91403198242188,
      "logps/rejected": -81.02542114257812,
      "loss": 0.7003,
      "rewards/accuracies": 0.1805555522441864,
      "rewards/chosen": 0.0010530749568715692,
      "rewards/margins": -0.01120604109019041,
      "rewards/rejected": 0.012259114533662796,
      "step": 10
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 438.4524613822673,
      "learning_rate": 2.852072705226364e-07,
      "logits/chosen": -2.6143155097961426,
      "logits/rejected": -2.5655746459960938,
      "logps/chosen": -103.23042297363281,
      "logps/rejected": -89.8201675415039,
      "loss": 0.6727,
      "rewards/accuracies": 0.2874999940395355,
      "rewards/chosen": 0.11292153596878052,
      "rewards/margins": 0.0527835413813591,
      "rewards/rejected": 0.06013798713684082,
      "step": 20
    },
    {
      "epoch": 0.3125,
      "grad_norm": 605.3404221140555,
      "learning_rate": 4.123399224035455e-07,
      "logits/chosen": -2.516937017440796,
      "logits/rejected": -2.5314974784851074,
      "logps/chosen": -67.30062103271484,
      "logps/rejected": -75.01806640625,
      "loss": 0.6755,
      "rewards/accuracies": 0.33125001192092896,
      "rewards/chosen": 0.12800100445747375,
      "rewards/margins": 0.21436181664466858,
      "rewards/rejected": -0.08636082708835602,
      "step": 30
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 444.59888459571647,
      "learning_rate": 4.002337238607724e-07,
      "logits/chosen": -2.58128023147583,
      "logits/rejected": -2.5694420337677,
      "logps/chosen": -71.9786376953125,
      "logps/rejected": -70.76583099365234,
      "loss": 0.707,
      "rewards/accuracies": 0.25,
      "rewards/chosen": 0.15727010369300842,
      "rewards/margins": 0.20649325847625732,
      "rewards/rejected": -0.049223143607378006,
      "step": 40
    },
    {
      "epoch": 0.5208333333333334,
      "grad_norm": 550.2437260335452,
      "learning_rate": 3.881275253179993e-07,
      "logits/chosen": -2.4956462383270264,
      "logits/rejected": -2.5077638626098633,
      "logps/chosen": -50.16924285888672,
      "logps/rejected": -58.39775466918945,
      "loss": 0.7248,
      "rewards/accuracies": 0.24375000596046448,
      "rewards/chosen": 0.685566782951355,
      "rewards/margins": 0.20654968917369843,
      "rewards/rejected": 0.47901707887649536,
      "step": 50
    },
    {
      "epoch": 0.625,
      "grad_norm": 398.6501309271306,
      "learning_rate": 3.7602132677522625e-07,
      "logits/chosen": -2.5804848670959473,
      "logits/rejected": -2.561837673187256,
      "logps/chosen": -77.69390869140625,
      "logps/rejected": -78.58424377441406,
      "loss": 0.6856,
      "rewards/accuracies": 0.34375,
      "rewards/chosen": 1.2882606983184814,
      "rewards/margins": 0.582649827003479,
      "rewards/rejected": 0.7056108713150024,
      "step": 60
    },
    {
      "epoch": 0.7291666666666666,
      "grad_norm": 522.2478868366107,
      "learning_rate": 3.6391512823245316e-07,
      "logits/chosen": -2.5641486644744873,
      "logits/rejected": -2.5553622245788574,
      "logps/chosen": -94.89775085449219,
      "logps/rejected": -84.7375259399414,
      "loss": 0.6908,
      "rewards/accuracies": 0.35624998807907104,
      "rewards/chosen": 1.3555562496185303,
      "rewards/margins": 0.6880232691764832,
      "rewards/rejected": 0.6675330400466919,
      "step": 70
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 784.7617891372116,
      "learning_rate": 3.5180892968968006e-07,
      "logits/chosen": -2.563314199447632,
      "logits/rejected": -2.5102317333221436,
      "logps/chosen": -84.65424346923828,
      "logps/rejected": -79.29447174072266,
      "loss": 0.6788,
      "rewards/accuracies": 0.40625,
      "rewards/chosen": 0.9282310605049133,
      "rewards/margins": 0.845805823802948,
      "rewards/rejected": 0.08242519199848175,
      "step": 80
    },
    {
      "epoch": 0.9375,
      "grad_norm": 653.0219382894146,
      "learning_rate": 3.39702731146907e-07,
      "logits/chosen": -2.4886748790740967,
      "logits/rejected": -2.4833056926727295,
      "logps/chosen": -53.71883010864258,
      "logps/rejected": -63.658111572265625,
      "loss": 0.7149,
      "rewards/accuracies": 0.2562499940395355,
      "rewards/chosen": 0.24449601769447327,
      "rewards/margins": 0.4224325716495514,
      "rewards/rejected": -0.17793650925159454,
      "step": 90
    },
    {
      "epoch": 1.0416666666666667,
      "grad_norm": 211.1512305376009,
      "learning_rate": 3.2759653260413393e-07,
      "logits/chosen": -2.486382246017456,
      "logits/rejected": -2.4644775390625,
      "logps/chosen": -70.5580062866211,
      "logps/rejected": -73.16108703613281,
      "loss": 0.5697,
      "rewards/accuracies": 0.39375001192092896,
      "rewards/chosen": 2.7764008045196533,
      "rewards/margins": 3.8405537605285645,
      "rewards/rejected": -1.0641529560089111,
      "step": 100
    },
    {
      "epoch": 1.0416666666666667,
      "eval_logits/chosen": -2.5235462188720703,
      "eval_logits/rejected": -2.507991313934326,
      "eval_logps/chosen": -73.52123260498047,
      "eval_logps/rejected": -80.83517456054688,
      "eval_loss": 0.731433093547821,
      "eval_rewards/accuracies": 0.3432539701461792,
      "eval_rewards/chosen": 0.9086403846740723,
      "eval_rewards/margins": 0.6340094804763794,
      "eval_rewards/rejected": 0.27463099360466003,
      "eval_runtime": 113.6041,
      "eval_samples_per_second": 17.605,
      "eval_steps_per_second": 0.555,
      "step": 100
    },
    {
      "epoch": 1.1458333333333333,
      "grad_norm": 53.69152144204709,
      "learning_rate": 3.1549033406136084e-07,
      "logits/chosen": -2.499805212020874,
      "logits/rejected": -2.531052350997925,
      "logps/chosen": -61.31892776489258,
      "logps/rejected": -85.37156677246094,
      "loss": 0.3654,
      "rewards/accuracies": 0.48124998807907104,
      "rewards/chosen": 3.7935471534729004,
      "rewards/margins": 7.435075283050537,
      "rewards/rejected": -3.6415278911590576,
      "step": 110
    },
    {
      "epoch": 1.25,
      "grad_norm": 20.68999344570945,
      "learning_rate": 3.0338413551858774e-07,
      "logits/chosen": -2.563140630722046,
      "logits/rejected": -2.531294345855713,
      "logps/chosen": -100.05778503417969,
      "logps/rejected": -99.29850769042969,
      "loss": 0.383,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": 7.445196628570557,
      "rewards/margins": 13.030279159545898,
      "rewards/rejected": -5.585082530975342,
      "step": 120
    },
    {
      "epoch": 1.3541666666666667,
      "grad_norm": 14.636781759163808,
      "learning_rate": 2.9127793697581465e-07,
      "logits/chosen": -2.4888994693756104,
      "logits/rejected": -2.491908073425293,
      "logps/chosen": -79.58487701416016,
      "logps/rejected": -93.6611328125,
      "loss": 0.3782,
      "rewards/accuracies": 0.46875,
      "rewards/chosen": 6.539540767669678,
      "rewards/margins": 10.590567588806152,
      "rewards/rejected": -4.051028251647949,
      "step": 130
    },
    {
      "epoch": 1.4583333333333333,
      "grad_norm": 244.07787693322098,
      "learning_rate": 2.7917173843304155e-07,
      "logits/chosen": -2.4806408882141113,
      "logits/rejected": -2.495673179626465,
      "logps/chosen": -43.72606658935547,
      "logps/rejected": -66.98627471923828,
      "loss": 0.3801,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": 5.209879398345947,
      "rewards/margins": 6.956570625305176,
      "rewards/rejected": -1.746691346168518,
      "step": 140
    },
    {
      "epoch": 1.5625,
      "grad_norm": 6.49560423904862,
      "learning_rate": 2.6706553989026857e-07,
      "logits/chosen": -2.5112829208374023,
      "logits/rejected": -2.486833333969116,
      "logps/chosen": -65.41731262207031,
      "logps/rejected": -75.73360443115234,
      "loss": 0.396,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 6.304823875427246,
      "rewards/margins": 8.325952529907227,
      "rewards/rejected": -2.021127939224243,
      "step": 150
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 6.063779849774731,
      "learning_rate": 2.5495934134749547e-07,
      "logits/chosen": -2.474635362625122,
      "logits/rejected": -2.482422113418579,
      "logps/chosen": -72.21180725097656,
      "logps/rejected": -92.7371826171875,
      "loss": 0.3721,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 8.162205696105957,
      "rewards/margins": 12.658352851867676,
      "rewards/rejected": -4.496148109436035,
      "step": 160
    },
    {
      "epoch": 1.7708333333333335,
      "grad_norm": 85.93105910386429,
      "learning_rate": 2.428531428047224e-07,
      "logits/chosen": -2.4428138732910156,
      "logits/rejected": -2.432603359222412,
      "logps/chosen": -69.3802261352539,
      "logps/rejected": -89.16145324707031,
      "loss": 0.3955,
      "rewards/accuracies": 0.53125,
      "rewards/chosen": 5.470473289489746,
      "rewards/margins": 11.255060195922852,
      "rewards/rejected": -5.784585952758789,
      "step": 170
    },
    {
      "epoch": 1.875,
      "grad_norm": 98.1345548828152,
      "learning_rate": 2.307469442619493e-07,
      "logits/chosen": -2.473292827606201,
      "logits/rejected": -2.4571070671081543,
      "logps/chosen": -66.85240936279297,
      "logps/rejected": -79.23042297363281,
      "loss": 0.3619,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 4.686768531799316,
      "rewards/margins": 9.949495315551758,
      "rewards/rejected": -5.262727737426758,
      "step": 180
    },
    {
      "epoch": 1.9791666666666665,
      "grad_norm": 12.502260702545252,
      "learning_rate": 2.186407457191762e-07,
      "logits/chosen": -2.5277750492095947,
      "logits/rejected": -2.5254924297332764,
      "logps/chosen": -81.32090759277344,
      "logps/rejected": -96.72193908691406,
      "loss": 0.3801,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 6.689125061035156,
      "rewards/margins": 12.403680801391602,
      "rewards/rejected": -5.714554309844971,
      "step": 190
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 0.8365271794834654,
      "learning_rate": 2.065345471764031e-07,
      "logits/chosen": -2.48215651512146,
      "logits/rejected": -2.5054686069488525,
      "logps/chosen": -81.12793731689453,
      "logps/rejected": -109.39041900634766,
      "loss": 0.3568,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 7.378162384033203,
      "rewards/margins": 14.7354154586792,
      "rewards/rejected": -7.3572540283203125,
      "step": 200
    },
    {
      "epoch": 2.0833333333333335,
      "eval_logits/chosen": -2.534412384033203,
      "eval_logits/rejected": -2.5182077884674072,
      "eval_logps/chosen": -72.82362365722656,
      "eval_logps/rejected": -80.80423736572266,
      "eval_loss": 0.9403718709945679,
      "eval_rewards/accuracies": 0.329365074634552,
      "eval_rewards/chosen": 1.5621305704116821,
      "eval_rewards/margins": 1.2585159540176392,
      "eval_rewards/rejected": 0.3036145269870758,
      "eval_runtime": 113.4863,
      "eval_samples_per_second": 17.623,
      "eval_steps_per_second": 0.555,
      "step": 200
    },
    {
      "epoch": 2.1875,
      "grad_norm": 0.17365138232640678,
      "learning_rate": 1.9442834863363e-07,
      "logits/chosen": -2.538646697998047,
      "logits/rejected": -2.531092405319214,
      "logps/chosen": -57.24335861206055,
      "logps/rejected": -73.53451538085938,
      "loss": 0.381,
      "rewards/accuracies": 0.42500001192092896,
      "rewards/chosen": 5.948960781097412,
      "rewards/margins": 10.486295700073242,
      "rewards/rejected": -4.537335395812988,
      "step": 210
    },
    {
      "epoch": 2.2916666666666665,
      "grad_norm": 2.449230292077233,
      "learning_rate": 1.8232215009085696e-07,
      "logits/chosen": -2.558448314666748,
      "logits/rejected": -2.511837959289551,
      "logps/chosen": -92.26642608642578,
      "logps/rejected": -99.34980010986328,
      "loss": 0.3282,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": 9.164371490478516,
      "rewards/margins": 15.474553108215332,
      "rewards/rejected": -6.310182094573975,
      "step": 220
    },
    {
      "epoch": 2.3958333333333335,
      "grad_norm": 0.8917188204265395,
      "learning_rate": 1.7021595154808387e-07,
      "logits/chosen": -2.5435986518859863,
      "logits/rejected": -2.512115001678467,
      "logps/chosen": -54.11479949951172,
      "logps/rejected": -59.78227615356445,
      "loss": 0.3796,
      "rewards/accuracies": 0.4312500059604645,
      "rewards/chosen": 5.355578422546387,
      "rewards/margins": 8.94926643371582,
      "rewards/rejected": -3.593688488006592,
      "step": 230
    },
    {
      "epoch": 2.5,
      "grad_norm": 164.121389909511,
      "learning_rate": 1.5810975300531077e-07,
      "logits/chosen": -2.5293872356414795,
      "logits/rejected": -2.483024835586548,
      "logps/chosen": -81.53751373291016,
      "logps/rejected": -85.67427062988281,
      "loss": 0.3814,
      "rewards/accuracies": 0.4937500059604645,
      "rewards/chosen": 5.955917835235596,
      "rewards/margins": 11.81001091003418,
      "rewards/rejected": -5.8540940284729,
      "step": 240
    },
    {
      "epoch": 2.6041666666666665,
      "grad_norm": 0.25571279122418833,
      "learning_rate": 1.4600355446253773e-07,
      "logits/chosen": -2.485600233078003,
      "logits/rejected": -2.473313570022583,
      "logps/chosen": -66.71253204345703,
      "logps/rejected": -76.95586395263672,
      "loss": 0.3582,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 5.579711437225342,
      "rewards/margins": 11.685328483581543,
      "rewards/rejected": -6.105617523193359,
      "step": 250
    },
    {
      "epoch": 2.7083333333333335,
      "grad_norm": 20.873017456890867,
      "learning_rate": 1.3389735591976464e-07,
      "logits/chosen": -2.487583637237549,
      "logits/rejected": -2.5043182373046875,
      "logps/chosen": -51.980262756347656,
      "logps/rejected": -80.04218292236328,
      "loss": 0.3794,
      "rewards/accuracies": 0.45625001192092896,
      "rewards/chosen": 5.761561393737793,
      "rewards/margins": 12.39246654510498,
      "rewards/rejected": -6.6309051513671875,
      "step": 260
    },
    {
      "epoch": 2.8125,
      "grad_norm": 39.42095791773263,
      "learning_rate": 1.2179115737699154e-07,
      "logits/chosen": -2.4937992095947266,
      "logits/rejected": -2.5243396759033203,
      "logps/chosen": -71.54154205322266,
      "logps/rejected": -85.11618041992188,
      "loss": 0.352,
      "rewards/accuracies": 0.4937500059604645,
      "rewards/chosen": 4.807980537414551,
      "rewards/margins": 11.203904151916504,
      "rewards/rejected": -6.395923614501953,
      "step": 270
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 0.222552372589121,
      "learning_rate": 1.0968495883421846e-07,
      "logits/chosen": -2.5492477416992188,
      "logits/rejected": -2.515148639678955,
      "logps/chosen": -78.6363525390625,
      "logps/rejected": -90.13154602050781,
      "loss": 0.3626,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 6.29600191116333,
      "rewards/margins": 13.563056945800781,
      "rewards/rejected": -7.267056465148926,
      "step": 280
    },
    {
      "epoch": 3.0,
      "step": 288,
      "total_flos": 0.0,
      "train_loss": 0.479405563738611,
      "train_runtime": 3137.8924,
      "train_samples_per_second": 5.844,
      "train_steps_per_second": 0.092
    }
  ],
  "logging_steps": 10,
  "max_steps": 288,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}