{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 939,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003194888178913738,
      "grad_norm": 34.47548268439031,
      "learning_rate": 5.3191489361702125e-09,
      "logits/chosen": -1.6953125,
      "logits/rejected": -1.640625,
      "logps/chosen": -235.0,
      "logps/rejected": -262.0,
      "loss": 0.6914,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03194888178913738,
      "grad_norm": 33.08858843822245,
      "learning_rate": 5.3191489361702123e-08,
      "logits/chosen": -1.7890625,
      "logits/rejected": -1.7734375,
      "logps/chosen": -245.0,
      "logps/rejected": -272.0,
      "loss": 0.6967,
      "rewards/accuracies": 0.1805555522441864,
      "rewards/chosen": -0.004180908203125,
      "rewards/margins": -0.006988525390625,
      "rewards/rejected": 0.002777099609375,
      "step": 10
    },
    {
      "epoch": 0.06389776357827476,
      "grad_norm": 35.17419870896503,
      "learning_rate": 1.0638297872340425e-07,
      "logits/chosen": -1.7578125,
      "logits/rejected": -1.765625,
      "logps/chosen": -247.0,
      "logps/rejected": -256.0,
      "loss": 0.6896,
      "rewards/accuracies": 0.22499999403953552,
      "rewards/chosen": 0.003753662109375,
      "rewards/margins": 0.01129150390625,
      "rewards/rejected": -0.00750732421875,
      "step": 20
    },
    {
      "epoch": 0.09584664536741214,
      "grad_norm": 36.64639475774933,
      "learning_rate": 1.5957446808510638e-07,
      "logits/chosen": -1.796875,
      "logits/rejected": -1.8046875,
      "logps/chosen": -242.0,
      "logps/rejected": -264.0,
      "loss": 0.6837,
      "rewards/accuracies": 0.3125,
      "rewards/chosen": 0.0012664794921875,
      "rewards/margins": -0.0012054443359375,
      "rewards/rejected": 0.00250244140625,
      "step": 30
    },
    {
      "epoch": 0.12779552715654952,
      "grad_norm": 35.2711796102512,
      "learning_rate": 2.127659574468085e-07,
      "logits/chosen": -1.7421875,
      "logits/rejected": -1.7421875,
      "logps/chosen": -246.0,
      "logps/rejected": -264.0,
      "loss": 0.663,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.038818359375,
      "rewards/margins": 0.08154296875,
      "rewards/rejected": -0.04248046875,
      "step": 40
    },
    {
      "epoch": 0.1597444089456869,
      "grad_norm": 31.56733318399803,
      "learning_rate": 2.659574468085106e-07,
      "logits/chosen": -1.75,
      "logits/rejected": -1.71875,
      "logps/chosen": -238.0,
      "logps/rejected": -266.0,
      "loss": 0.65,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.04248046875,
      "rewards/margins": 0.1826171875,
      "rewards/rejected": -0.140625,
      "step": 50
    },
    {
      "epoch": 0.19169329073482427,
      "grad_norm": 32.09814766896921,
      "learning_rate": 3.1914893617021275e-07,
      "logits/chosen": -1.75,
      "logits/rejected": -1.7265625,
      "logps/chosen": -236.0,
      "logps/rejected": -268.0,
      "loss": 0.619,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.053955078125,
      "rewards/margins": 0.2099609375,
      "rewards/rejected": -0.15625,
      "step": 60
    },
    {
      "epoch": 0.22364217252396165,
      "grad_norm": 32.03938813135862,
      "learning_rate": 3.7234042553191484e-07,
      "logits/chosen": -1.765625,
      "logits/rejected": -1.7578125,
      "logps/chosen": -242.0,
      "logps/rejected": -264.0,
      "loss": 0.611,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.0277099609375,
      "rewards/margins": 0.255859375,
      "rewards/rejected": -0.2275390625,
      "step": 70
    },
    {
      "epoch": 0.25559105431309903,
      "grad_norm": 43.96835624603879,
      "learning_rate": 4.25531914893617e-07,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.75,
      "logps/chosen": -243.0,
      "logps/rejected": -262.0,
      "loss": 0.6006,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.103515625,
      "rewards/margins": 0.392578125,
      "rewards/rejected": -0.49609375,
      "step": 80
    },
    {
      "epoch": 0.28753993610223644,
      "grad_norm": 28.877852885313988,
      "learning_rate": 4.787234042553192e-07,
      "logits/chosen": -1.7421875,
      "logits/rejected": -1.71875,
      "logps/chosen": -245.0,
      "logps/rejected": -272.0,
      "loss": 0.5725,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.03369140625,
      "rewards/margins": 0.34765625,
      "rewards/rejected": -0.380859375,
      "step": 90
    },
    {
      "epoch": 0.3194888178913738,
      "grad_norm": 33.04150137611911,
      "learning_rate": 4.964497041420119e-07,
      "logits/chosen": -1.75,
      "logits/rejected": -1.734375,
      "logps/chosen": -240.0,
      "logps/rejected": -268.0,
      "loss": 0.6132,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.08544921875,
      "rewards/margins": 0.5078125,
      "rewards/rejected": -0.423828125,
      "step": 100
    },
    {
      "epoch": 0.3514376996805112,
      "grad_norm": 31.292790431153033,
      "learning_rate": 4.905325443786982e-07,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.7421875,
      "logps/chosen": -241.0,
      "logps/rejected": -260.0,
      "loss": 0.5318,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.07421875,
      "rewards/margins": 0.51953125,
      "rewards/rejected": -0.443359375,
      "step": 110
    },
    {
      "epoch": 0.38338658146964855,
      "grad_norm": 28.540124264789696,
      "learning_rate": 4.846153846153846e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.703125,
      "logps/chosen": -238.0,
      "logps/rejected": -268.0,
      "loss": 0.5829,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.0263671875,
      "rewards/margins": 0.57421875,
      "rewards/rejected": -0.546875,
      "step": 120
    },
    {
      "epoch": 0.41533546325878595,
      "grad_norm": 29.08477368681588,
      "learning_rate": 4.78698224852071e-07,
      "logits/chosen": -1.765625,
      "logits/rejected": -1.7265625,
      "logps/chosen": -245.0,
      "logps/rejected": -274.0,
      "loss": 0.5456,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.0712890625,
      "rewards/margins": 0.52734375,
      "rewards/rejected": -0.59765625,
      "step": 130
    },
    {
      "epoch": 0.4472843450479233,
      "grad_norm": 24.895122773023758,
      "learning_rate": 4.727810650887574e-07,
      "logits/chosen": -1.6796875,
      "logits/rejected": -1.7109375,
      "logps/chosen": -240.0,
      "logps/rejected": -268.0,
      "loss": 0.5276,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.0252685546875,
      "rewards/margins": 0.67578125,
      "rewards/rejected": -0.69921875,
      "step": 140
    },
    {
      "epoch": 0.4792332268370607,
      "grad_norm": 25.086960468247135,
      "learning_rate": 4.668639053254438e-07,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.7265625,
      "logps/chosen": -236.0,
      "logps/rejected": -264.0,
      "loss": 0.555,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.0751953125,
      "rewards/margins": 0.66796875,
      "rewards/rejected": -0.59375,
      "step": 150
    },
    {
      "epoch": 0.5111821086261981,
      "grad_norm": 28.17100030607696,
      "learning_rate": 4.6094674556213014e-07,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.75,
      "logps/chosen": -244.0,
      "logps/rejected": -264.0,
      "loss": 0.5135,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.0162353515625,
      "rewards/margins": 0.609375,
      "rewards/rejected": -0.625,
      "step": 160
    },
    {
      "epoch": 0.5431309904153354,
      "grad_norm": 26.942036465103737,
      "learning_rate": 4.5502958579881655e-07,
      "logits/chosen": -1.75,
      "logits/rejected": -1.7265625,
      "logps/chosen": -238.0,
      "logps/rejected": -268.0,
      "loss": 0.5144,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.1201171875,
      "rewards/margins": 0.875,
      "rewards/rejected": -0.75390625,
      "step": 170
    },
    {
      "epoch": 0.5750798722044729,
      "grad_norm": 27.057109325150893,
      "learning_rate": 4.491124260355029e-07,
      "logits/chosen": -1.703125,
      "logits/rejected": -1.7265625,
      "logps/chosen": -237.0,
      "logps/rejected": -284.0,
      "loss": 0.4916,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.09130859375,
      "rewards/margins": 0.94140625,
      "rewards/rejected": -0.8515625,
      "step": 180
    },
    {
      "epoch": 0.6070287539936102,
      "grad_norm": 30.82037140778774,
      "learning_rate": 4.4319526627218936e-07,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.703125,
      "logps/chosen": -242.0,
      "logps/rejected": -266.0,
      "loss": 0.54,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.09130859375,
      "rewards/margins": 0.69140625,
      "rewards/rejected": -0.59765625,
      "step": 190
    },
    {
      "epoch": 0.6389776357827476,
      "grad_norm": 24.497963916643677,
      "learning_rate": 4.372781065088757e-07,
      "logits/chosen": -1.78125,
      "logits/rejected": -1.7265625,
      "logps/chosen": -245.0,
      "logps/rejected": -268.0,
      "loss": 0.5101,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.06884765625,
      "rewards/margins": 0.73828125,
      "rewards/rejected": -0.671875,
      "step": 200
    },
    {
      "epoch": 0.670926517571885,
      "grad_norm": 30.284876833786186,
      "learning_rate": 4.313609467455621e-07,
      "logits/chosen": -1.7265625,
      "logits/rejected": -1.75,
      "logps/chosen": -242.0,
      "logps/rejected": -270.0,
      "loss": 0.551,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.107421875,
      "rewards/margins": 0.78515625,
      "rewards/rejected": -0.89453125,
      "step": 210
    },
    {
      "epoch": 0.7028753993610224,
      "grad_norm": 31.01459795666554,
      "learning_rate": 4.2544378698224847e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.6953125,
      "logps/chosen": -237.0,
      "logps/rejected": -266.0,
      "loss": 0.5036,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -0.10888671875,
      "rewards/margins": 0.68359375,
      "rewards/rejected": -0.7890625,
      "step": 220
    },
    {
      "epoch": 0.7348242811501597,
      "grad_norm": 31.681398259902394,
      "learning_rate": 4.195266272189349e-07,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.703125,
      "logps/chosen": -243.0,
      "logps/rejected": -264.0,
      "loss": 0.5146,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.0986328125,
      "rewards/margins": 0.7421875,
      "rewards/rejected": -0.83984375,
      "step": 230
    },
    {
      "epoch": 0.7667731629392971,
      "grad_norm": 30.3843807531397,
      "learning_rate": 4.1360946745562133e-07,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.7265625,
      "logps/chosen": -242.0,
      "logps/rejected": -272.0,
      "loss": 0.4829,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.0751953125,
      "rewards/margins": 0.75,
      "rewards/rejected": -0.828125,
      "step": 240
    },
    {
      "epoch": 0.7987220447284346,
      "grad_norm": 27.627369180135428,
      "learning_rate": 4.076923076923077e-07,
      "logits/chosen": -1.7421875,
      "logits/rejected": -1.734375,
      "logps/chosen": -240.0,
      "logps/rejected": -266.0,
      "loss": 0.4946,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.26171875,
      "rewards/margins": 0.6328125,
      "rewards/rejected": -0.890625,
      "step": 250
    },
    {
      "epoch": 0.8306709265175719,
      "grad_norm": 28.61228168078618,
      "learning_rate": 4.017751479289941e-07,
      "logits/chosen": -1.7421875,
      "logits/rejected": -1.765625,
      "logps/chosen": -240.0,
      "logps/rejected": -270.0,
      "loss": 0.4946,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.111328125,
      "rewards/margins": 0.8671875,
      "rewards/rejected": -0.98046875,
      "step": 260
    },
    {
      "epoch": 0.8626198083067093,
      "grad_norm": 33.147902615790635,
      "learning_rate": 3.9585798816568044e-07,
      "logits/chosen": -1.765625,
      "logits/rejected": -1.71875,
      "logps/chosen": -237.0,
      "logps/rejected": -276.0,
      "loss": 0.5384,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.099609375,
      "rewards/margins": 0.65234375,
      "rewards/rejected": -0.75,
      "step": 270
    },
    {
      "epoch": 0.8945686900958466,
      "grad_norm": 27.58369117700387,
      "learning_rate": 3.8994082840236685e-07,
      "logits/chosen": -1.703125,
      "logits/rejected": -1.734375,
      "logps/chosen": -241.0,
      "logps/rejected": -260.0,
      "loss": 0.5174,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.09912109375,
      "rewards/margins": 0.7109375,
      "rewards/rejected": -0.609375,
      "step": 280
    },
    {
      "epoch": 0.9265175718849841,
      "grad_norm": 27.845142754279863,
      "learning_rate": 3.840236686390532e-07,
      "logits/chosen": -1.671875,
      "logits/rejected": -1.71875,
      "logps/chosen": -238.0,
      "logps/rejected": -264.0,
      "loss": 0.4757,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.0177001953125,
      "rewards/margins": 0.984375,
      "rewards/rejected": -0.96875,
      "step": 290
    },
    {
      "epoch": 0.9584664536741214,
      "grad_norm": 32.815897540130955,
      "learning_rate": 3.7810650887573966e-07,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.7421875,
      "logps/chosen": -238.0,
      "logps/rejected": -276.0,
      "loss": 0.4941,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.228515625,
      "rewards/margins": 1.0,
      "rewards/rejected": -1.2265625,
      "step": 300
    },
    {
      "epoch": 0.9904153354632588,
      "grad_norm": 28.360981127394723,
      "learning_rate": 3.72189349112426e-07,
      "logits/chosen": -1.7421875,
      "logits/rejected": -1.7578125,
      "logps/chosen": -248.0,
      "logps/rejected": -274.0,
      "loss": 0.4374,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.2099609375,
      "rewards/margins": 1.3359375,
      "rewards/rejected": -1.5390625,
      "step": 310
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -1.7109375,
      "eval_logits/rejected": -1.7421875,
      "eval_logps/chosen": -240.0,
      "eval_logps/rejected": -278.0,
      "eval_loss": 0.5235937237739563,
      "eval_rewards/accuracies": 0.7142857313156128,
      "eval_rewards/chosen": -0.326171875,
      "eval_rewards/margins": 0.63671875,
      "eval_rewards/rejected": -0.9609375,
      "eval_runtime": 12.6819,
      "eval_samples_per_second": 15.771,
      "eval_steps_per_second": 0.552,
      "step": 313
    },
    {
      "epoch": 1.0223642172523961,
      "grad_norm": 23.23800693350808,
      "learning_rate": 3.662721893491124e-07,
      "logits/chosen": -1.7265625,
      "logits/rejected": -1.6875,
      "logps/chosen": -238.0,
      "logps/rejected": -266.0,
      "loss": 0.365,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.012451171875,
      "rewards/margins": 1.2734375,
      "rewards/rejected": -1.265625,
      "step": 320
    },
    {
      "epoch": 1.0543130990415335,
      "grad_norm": 28.085201782352296,
      "learning_rate": 3.603550295857988e-07,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.734375,
      "logps/chosen": -242.0,
      "logps/rejected": -276.0,
      "loss": 0.3465,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.00872802734375,
      "rewards/margins": 1.46875,
      "rewards/rejected": -1.4765625,
      "step": 330
    },
    {
      "epoch": 1.0862619808306708,
      "grad_norm": 24.247027348577692,
      "learning_rate": 3.544378698224852e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.6953125,
      "logps/chosen": -238.0,
      "logps/rejected": -278.0,
      "loss": 0.2981,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 0.1435546875,
      "rewards/margins": 1.4453125,
      "rewards/rejected": -1.3046875,
      "step": 340
    },
    {
      "epoch": 1.1182108626198084,
      "grad_norm": 21.346684851723847,
      "learning_rate": 3.485207100591716e-07,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.6953125,
      "logps/chosen": -239.0,
      "logps/rejected": -274.0,
      "loss": 0.2997,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.12158203125,
      "rewards/margins": 1.453125,
      "rewards/rejected": -1.5703125,
      "step": 350
    },
    {
      "epoch": 1.1501597444089458,
      "grad_norm": 22.92923404174748,
      "learning_rate": 3.42603550295858e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.7109375,
      "logps/chosen": -239.0,
      "logps/rejected": -278.0,
      "loss": 0.2966,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.197265625,
      "rewards/margins": 1.6328125,
      "rewards/rejected": -1.4296875,
      "step": 360
    },
    {
      "epoch": 1.182108626198083,
      "grad_norm": 19.074914011714277,
      "learning_rate": 3.366863905325444e-07,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.734375,
      "logps/chosen": -240.0,
      "logps/rejected": -278.0,
      "loss": 0.3032,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 0.2451171875,
      "rewards/margins": 1.78125,
      "rewards/rejected": -1.5390625,
      "step": 370
    },
    {
      "epoch": 1.2140575079872205,
      "grad_norm": 24.568535159721062,
      "learning_rate": 3.3076923076923075e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.7109375,
      "logps/chosen": -239.0,
      "logps/rejected": -292.0,
      "loss": 0.2915,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 0.047607421875,
      "rewards/margins": 1.71875,
      "rewards/rejected": -1.671875,
      "step": 380
    },
    {
      "epoch": 1.2460063897763578,
      "grad_norm": 27.96077981193079,
      "learning_rate": 3.2485207100591716e-07,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.6953125,
      "logps/chosen": -239.0,
      "logps/rejected": -280.0,
      "loss": 0.3166,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.12158203125,
      "rewards/margins": 1.6796875,
      "rewards/rejected": -1.796875,
      "step": 390
    },
    {
      "epoch": 1.2779552715654952,
      "grad_norm": 23.939263929977898,
      "learning_rate": 3.189349112426035e-07,
      "logits/chosen": -1.7421875,
      "logits/rejected": -1.703125,
      "logps/chosen": -247.0,
      "logps/rejected": -278.0,
      "loss": 0.3313,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.291015625,
      "rewards/margins": 1.5,
      "rewards/rejected": -1.7890625,
      "step": 400
    },
    {
      "epoch": 1.3099041533546325,
      "grad_norm": 23.445460332455205,
      "learning_rate": 3.130177514792899e-07,
      "logits/chosen": -1.765625,
      "logits/rejected": -1.7265625,
      "logps/chosen": -244.0,
      "logps/rejected": -276.0,
      "loss": 0.29,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.322265625,
      "rewards/margins": 1.59375,
      "rewards/rejected": -1.9140625,
      "step": 410
    },
    {
      "epoch": 1.34185303514377,
      "grad_norm": 15.133152781473944,
      "learning_rate": 3.071005917159763e-07,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.734375,
      "logps/chosen": -250.0,
      "logps/rejected": -278.0,
      "loss": 0.2755,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.322265625,
      "rewards/margins": 1.546875,
      "rewards/rejected": -1.8671875,
      "step": 420
    },
    {
      "epoch": 1.3738019169329074,
      "grad_norm": 19.11123594038131,
      "learning_rate": 3.011834319526627e-07,
      "logits/chosen": -1.765625,
      "logits/rejected": -1.7421875,
      "logps/chosen": -245.0,
      "logps/rejected": -274.0,
      "loss": 0.2506,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.10009765625,
      "rewards/margins": 1.8203125,
      "rewards/rejected": -1.921875,
      "step": 430
    },
    {
      "epoch": 1.4057507987220448,
      "grad_norm": 17.795524020323455,
      "learning_rate": 2.952662721893491e-07,
      "logits/chosen": -1.7578125,
      "logits/rejected": -1.7421875,
      "logps/chosen": -243.0,
      "logps/rejected": -286.0,
      "loss": 0.2583,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.0123291015625,
      "rewards/margins": 1.8046875,
      "rewards/rejected": -1.8125,
      "step": 440
    },
    {
      "epoch": 1.4376996805111821,
      "grad_norm": 23.74400717983665,
      "learning_rate": 2.893491124260355e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.7109375,
      "logps/chosen": -241.0,
      "logps/rejected": -288.0,
      "loss": 0.2642,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.03369140625,
      "rewards/margins": 1.9453125,
      "rewards/rejected": -1.984375,
      "step": 450
    },
    {
      "epoch": 1.4696485623003195,
      "grad_norm": 28.078711225527382,
      "learning_rate": 2.834319526627219e-07,
      "logits/chosen": -1.7265625,
      "logits/rejected": -1.734375,
      "logps/chosen": -240.0,
      "logps/rejected": -274.0,
      "loss": 0.2679,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 0.1669921875,
      "rewards/margins": 1.734375,
      "rewards/rejected": -1.5703125,
      "step": 460
    },
    {
      "epoch": 1.5015974440894568,
      "grad_norm": 22.708510750866107,
      "learning_rate": 2.7751479289940824e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.7265625,
      "logps/chosen": -242.0,
      "logps/rejected": -268.0,
      "loss": 0.2787,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.0673828125,
      "rewards/margins": 1.5703125,
      "rewards/rejected": -1.5,
      "step": 470
    },
    {
      "epoch": 1.5335463258785942,
      "grad_norm": 18.86680359426442,
      "learning_rate": 2.715976331360947e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.7421875,
      "logps/chosen": -241.0,
      "logps/rejected": -288.0,
      "loss": 0.2716,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.1298828125,
      "rewards/margins": 1.7578125,
      "rewards/rejected": -1.625,
      "step": 480
    },
    {
      "epoch": 1.5654952076677318,
      "grad_norm": 17.156329445818553,
      "learning_rate": 2.6568047337278106e-07,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.7578125,
      "logps/chosen": -237.0,
      "logps/rejected": -276.0,
      "loss": 0.3079,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 0.0712890625,
      "rewards/margins": 1.75,
      "rewards/rejected": -1.6796875,
      "step": 490
    },
    {
      "epoch": 1.5974440894568689,
      "grad_norm": 19.332118400424566,
      "learning_rate": 2.5976331360946746e-07,
      "logits/chosen": -1.7265625,
      "logits/rejected": -1.6953125,
      "logps/chosen": -243.0,
      "logps/rejected": -280.0,
      "loss": 0.2851,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.080078125,
      "rewards/margins": 1.7734375,
      "rewards/rejected": -1.6953125,
      "step": 500
    },
    {
      "epoch": 1.6293929712460065,
      "grad_norm": 19.62634703131097,
      "learning_rate": 2.538461538461538e-07,
      "logits/chosen": -1.7421875,
      "logits/rejected": -1.734375,
      "logps/chosen": -239.0,
      "logps/rejected": -272.0,
      "loss": 0.2776,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.030029296875,
      "rewards/margins": 1.5546875,
      "rewards/rejected": -1.5234375,
      "step": 510
    },
    {
      "epoch": 1.6613418530351438,
      "grad_norm": 30.286972397278998,
      "learning_rate": 2.479289940828402e-07,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.6875,
      "logps/chosen": -239.0,
      "logps/rejected": -272.0,
      "loss": 0.2678,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.14453125,
      "rewards/margins": 1.953125,
      "rewards/rejected": -1.8125,
      "step": 520
    },
    {
      "epoch": 1.6932907348242812,
      "grad_norm": 26.68217050843353,
      "learning_rate": 2.420118343195266e-07,
      "logits/chosen": -1.75,
      "logits/rejected": -1.7734375,
      "logps/chosen": -234.0,
      "logps/rejected": -272.0,
      "loss": 0.2765,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.404296875,
      "rewards/margins": 1.9140625,
      "rewards/rejected": -1.5078125,
      "step": 530
    },
    {
      "epoch": 1.7252396166134185,
      "grad_norm": 22.03681852867734,
      "learning_rate": 2.36094674556213e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.7421875,
      "logps/chosen": -243.0,
      "logps/rejected": -276.0,
      "loss": 0.2823,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": -0.09716796875,
      "rewards/margins": 1.7421875,
      "rewards/rejected": -1.8359375,
      "step": 540
    },
    {
      "epoch": 1.7571884984025559,
      "grad_norm": 19.88206815145969,
      "learning_rate": 2.301775147928994e-07,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.7265625,
      "logps/chosen": -237.0,
      "logps/rejected": -274.0,
      "loss": 0.258,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.0025634765625,
      "rewards/margins": 1.796875,
      "rewards/rejected": -1.796875,
      "step": 550
    },
    {
      "epoch": 1.7891373801916934,
      "grad_norm": 24.819444792106694,
      "learning_rate": 2.242603550295858e-07,
      "logits/chosen": -1.7578125,
      "logits/rejected": -1.7109375,
      "logps/chosen": -242.0,
      "logps/rejected": -282.0,
      "loss": 0.292,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 0.2470703125,
      "rewards/margins": 1.9765625,
      "rewards/rejected": -1.7265625,
      "step": 560
    },
    {
      "epoch": 1.8210862619808306,
      "grad_norm": 25.1506153366333,
      "learning_rate": 2.1834319526627217e-07,
      "logits/chosen": -1.7265625,
      "logits/rejected": -1.71875,
      "logps/chosen": -235.0,
      "logps/rejected": -280.0,
      "loss": 0.3011,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.24609375,
      "rewards/margins": 1.921875,
      "rewards/rejected": -1.6796875,
      "step": 570
    },
    {
      "epoch": 1.8530351437699681,
      "grad_norm": 25.408870033012057,
      "learning_rate": 2.1242603550295858e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.7421875,
      "logps/chosen": -243.0,
      "logps/rejected": -284.0,
      "loss": 0.2759,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.060302734375,
      "rewards/margins": 1.953125,
      "rewards/rejected": -1.8984375,
      "step": 580
    },
    {
      "epoch": 1.8849840255591053,
      "grad_norm": 30.243845189152044,
      "learning_rate": 2.0650887573964496e-07,
      "logits/chosen": -1.78125,
      "logits/rejected": -1.75,
      "logps/chosen": -248.0,
      "logps/rejected": -288.0,
      "loss": 0.3033,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.072265625,
      "rewards/margins": 1.8203125,
      "rewards/rejected": -1.890625,
      "step": 590
    },
    {
      "epoch": 1.9169329073482428,
      "grad_norm": 25.522451943807717,
      "learning_rate": 2.0059171597633133e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.765625,
      "logps/chosen": -236.0,
      "logps/rejected": -282.0,
      "loss": 0.2878,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.2021484375,
      "rewards/margins": 1.9296875,
      "rewards/rejected": -1.7265625,
      "step": 600
    },
    {
      "epoch": 1.9488817891373802,
      "grad_norm": 25.361411149259393,
      "learning_rate": 1.9467455621301774e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.734375,
      "logps/chosen": -241.0,
      "logps/rejected": -282.0,
      "loss": 0.2881,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.11767578125,
      "rewards/margins": 1.8046875,
      "rewards/rejected": -1.6953125,
      "step": 610
    },
    {
      "epoch": 1.9808306709265175,
      "grad_norm": 18.620086897639467,
      "learning_rate": 1.8875739644970412e-07,
      "logits/chosen": -1.703125,
      "logits/rejected": -1.703125,
      "logps/chosen": -243.0,
      "logps/rejected": -278.0,
      "loss": 0.275,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.072265625,
      "rewards/margins": 1.5859375,
      "rewards/rejected": -1.6640625,
      "step": 620
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -1.703125,
      "eval_logits/rejected": -1.7421875,
      "eval_logps/chosen": -240.0,
      "eval_logps/rejected": -280.0,
      "eval_loss": 0.5277343988418579,
      "eval_rewards/accuracies": 0.7678571343421936,
      "eval_rewards/chosen": -0.37109375,
      "eval_rewards/margins": 0.859375,
      "eval_rewards/rejected": -1.234375,
      "eval_runtime": 12.6,
      "eval_samples_per_second": 15.873,
      "eval_steps_per_second": 0.556,
      "step": 626
    },
    {
      "epoch": 2.012779552715655,
      "grad_norm": 15.955204221662331,
      "learning_rate": 1.8284023668639053e-07,
      "logits/chosen": -1.7421875,
      "logits/rejected": -1.734375,
      "logps/chosen": -244.0,
      "logps/rejected": -278.0,
      "loss": 0.2415,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.1376953125,
      "rewards/margins": 1.9296875,
      "rewards/rejected": -1.796875,
      "step": 630
    },
    {
      "epoch": 2.0447284345047922,
      "grad_norm": 14.058307718181862,
      "learning_rate": 1.7692307692307693e-07,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.7265625,
      "logps/chosen": -237.0,
      "logps/rejected": -280.0,
      "loss": 0.1544,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.21484375,
      "rewards/margins": 2.53125,
      "rewards/rejected": -2.3125,
      "step": 640
    },
    {
      "epoch": 2.07667731629393,
      "grad_norm": 14.300114625147318,
      "learning_rate": 1.710059171597633e-07,
      "logits/chosen": -1.75,
      "logits/rejected": -1.75,
      "logps/chosen": -241.0,
      "logps/rejected": -282.0,
      "loss": 0.1729,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.33203125,
      "rewards/margins": 2.234375,
      "rewards/rejected": -1.8984375,
      "step": 650
    },
    {
      "epoch": 2.108626198083067,
      "grad_norm": 15.65574607331697,
      "learning_rate": 1.650887573964497e-07,
      "logits/chosen": -1.7578125,
      "logits/rejected": -1.734375,
      "logps/chosen": -238.0,
      "logps/rejected": -290.0,
      "loss": 0.1595,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.015869140625,
      "rewards/margins": 2.421875,
      "rewards/rejected": -2.40625,
      "step": 660
    },
    {
      "epoch": 2.1405750798722045,
      "grad_norm": 17.250707654114066,
      "learning_rate": 1.591715976331361e-07,
      "logits/chosen": -1.6796875,
      "logits/rejected": -1.7265625,
      "logps/chosen": -240.0,
      "logps/rejected": -288.0,
      "loss": 0.1731,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.1396484375,
      "rewards/margins": 2.53125,
      "rewards/rejected": -2.390625,
      "step": 670
    },
    {
      "epoch": 2.1725239616613417,
      "grad_norm": 18.514625043103106,
      "learning_rate": 1.5325443786982248e-07,
      "logits/chosen": -1.75,
      "logits/rejected": -1.703125,
      "logps/chosen": -238.0,
      "logps/rejected": -286.0,
      "loss": 0.1608,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.36328125,
      "rewards/margins": 2.65625,
      "rewards/rejected": -2.296875,
      "step": 680
    },
    {
      "epoch": 2.2044728434504792,
      "grad_norm": 24.10831456497685,
      "learning_rate": 1.4733727810650885e-07,
      "logits/chosen": -1.6953125,
      "logits/rejected": -1.703125,
      "logps/chosen": -246.0,
      "logps/rejected": -282.0,
      "loss": 0.1953,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.068359375,
      "rewards/margins": 2.28125,
      "rewards/rejected": -2.21875,
      "step": 690
    },
    {
      "epoch": 2.236421725239617,
      "grad_norm": 17.181045102958997,
      "learning_rate": 1.4142011834319526e-07,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.71875,
      "logps/chosen": -235.0,
      "logps/rejected": -284.0,
      "loss": 0.1535,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.259765625,
      "rewards/margins": 2.5,
      "rewards/rejected": -2.234375,
      "step": 700
    },
    {
      "epoch": 2.268370607028754,
      "grad_norm": 14.135092040427887,
      "learning_rate": 1.3550295857988164e-07,
      "logits/chosen": -1.6875,
      "logits/rejected": -1.75,
      "logps/chosen": -235.0,
      "logps/rejected": -280.0,
      "loss": 0.154,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.23046875,
      "rewards/margins": 2.59375,
      "rewards/rejected": -2.359375,
      "step": 710
    },
    {
      "epoch": 2.3003194888178915,
      "grad_norm": 14.760186737903501,
      "learning_rate": 1.2958579881656802e-07,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.71875,
      "logps/chosen": -234.0,
      "logps/rejected": -290.0,
      "loss": 0.1606,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.265625,
      "rewards/margins": 2.65625,
      "rewards/rejected": -2.390625,
      "step": 720
    },
    {
      "epoch": 2.3322683706070286,
      "grad_norm": 19.868109224808542,
      "learning_rate": 1.2366863905325443e-07,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.7578125,
      "logps/chosen": -235.0,
      "logps/rejected": -282.0,
      "loss": 0.1487,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.267578125,
      "rewards/margins": 2.625,
      "rewards/rejected": -2.359375,
      "step": 730
    },
    {
      "epoch": 2.364217252396166,
      "grad_norm": 23.28653194497905,
      "learning_rate": 1.1775147928994082e-07,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.7265625,
      "logps/chosen": -244.0,
      "logps/rejected": -286.0,
      "loss": 0.1478,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.10595703125,
      "rewards/margins": 2.59375,
      "rewards/rejected": -2.484375,
      "step": 740
    },
    {
      "epoch": 2.3961661341853033,
      "grad_norm": 17.120801675488444,
      "learning_rate": 1.1183431952662721e-07,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.6796875,
      "logps/chosen": -236.0,
      "logps/rejected": -288.0,
      "loss": 0.1446,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.1015625,
      "rewards/margins": 2.75,
      "rewards/rejected": -2.65625,
      "step": 750
    },
    {
      "epoch": 2.428115015974441,
      "grad_norm": 15.22339588033004,
      "learning_rate": 1.059171597633136e-07,
      "logits/chosen": -1.7265625,
      "logits/rejected": -1.71875,
      "logps/chosen": -239.0,
      "logps/rejected": -280.0,
      "loss": 0.1655,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.1962890625,
      "rewards/margins": 2.453125,
      "rewards/rejected": -2.25,
      "step": 760
    },
    {
      "epoch": 2.460063897763578,
      "grad_norm": 11.99984881863295,
      "learning_rate": 1e-07,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.7578125,
      "logps/chosen": -232.0,
      "logps/rejected": -296.0,
      "loss": 0.1466,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.10009765625,
      "rewards/margins": 2.8125,
      "rewards/rejected": -2.71875,
      "step": 770
    },
    {
      "epoch": 2.4920127795527156,
      "grad_norm": 13.471810930026255,
      "learning_rate": 9.408284023668639e-08,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.734375,
      "logps/chosen": -242.0,
      "logps/rejected": -286.0,
      "loss": 0.1451,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.2119140625,
      "rewards/margins": 2.578125,
      "rewards/rejected": -2.359375,
      "step": 780
    },
    {
      "epoch": 2.523961661341853,
      "grad_norm": 25.459720363700047,
      "learning_rate": 8.816568047337278e-08,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.6875,
      "logps/chosen": -241.0,
      "logps/rejected": -300.0,
      "loss": 0.1488,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.142578125,
      "rewards/margins": 2.796875,
      "rewards/rejected": -2.640625,
      "step": 790
    },
    {
      "epoch": 2.5559105431309903,
      "grad_norm": 18.20570925745819,
      "learning_rate": 8.224852071005916e-08,
      "logits/chosen": -1.765625,
      "logits/rejected": -1.7265625,
      "logps/chosen": -237.0,
      "logps/rejected": -292.0,
      "loss": 0.1676,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.06298828125,
      "rewards/margins": 2.796875,
      "rewards/rejected": -2.734375,
      "step": 800
    },
    {
      "epoch": 2.587859424920128,
      "grad_norm": 22.009878869215555,
      "learning_rate": 7.633136094674555e-08,
      "logits/chosen": -1.7421875,
      "logits/rejected": -1.7265625,
      "logps/chosen": -244.0,
      "logps/rejected": -288.0,
      "loss": 0.1575,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.03759765625,
      "rewards/margins": 2.78125,
      "rewards/rejected": -2.828125,
      "step": 810
    },
    {
      "epoch": 2.619808306709265,
      "grad_norm": 11.203003537220331,
      "learning_rate": 7.041420118343195e-08,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.75,
      "logps/chosen": -236.0,
      "logps/rejected": -280.0,
      "loss": 0.1623,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.032470703125,
      "rewards/margins": 2.46875,
      "rewards/rejected": -2.4375,
      "step": 820
    },
    {
      "epoch": 2.6517571884984026,
      "grad_norm": 11.09178228981755,
      "learning_rate": 6.449704142011835e-08,
      "logits/chosen": -1.734375,
      "logits/rejected": -1.7265625,
      "logps/chosen": -245.0,
      "logps/rejected": -300.0,
      "loss": 0.1548,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.0235595703125,
      "rewards/margins": 3.078125,
      "rewards/rejected": -3.046875,
      "step": 830
    },
    {
      "epoch": 2.68370607028754,
      "grad_norm": 18.04056695002969,
      "learning_rate": 5.857988165680473e-08,
      "logits/chosen": -1.71875,
      "logits/rejected": -1.71875,
      "logps/chosen": -244.0,
      "logps/rejected": -286.0,
      "loss": 0.1733,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.169921875,
      "rewards/margins": 2.703125,
      "rewards/rejected": -2.53125,
      "step": 840
    },
    {
      "epoch": 2.7156549520766773,
      "grad_norm": 20.112810364895005,
      "learning_rate": 5.266272189349112e-08,
      "logits/chosen": -1.6953125,
      "logits/rejected": -1.7265625,
      "logps/chosen": -241.0,
      "logps/rejected": -288.0,
      "loss": 0.1534,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.1796875,
      "rewards/margins": 2.65625,
      "rewards/rejected": -2.484375,
      "step": 850
    },
    {
      "epoch": 2.747603833865815,
      "grad_norm": 15.351717469363905,
      "learning_rate": 4.674556213017751e-08,
      "logits/chosen": -1.703125,
      "logits/rejected": -1.703125,
      "logps/chosen": -240.0,
      "logps/rejected": -284.0,
      "loss": 0.1448,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.1796875,
      "rewards/margins": 2.671875,
      "rewards/rejected": -2.484375,
      "step": 860
    },
    {
      "epoch": 2.779552715654952,
      "grad_norm": 19.628185598163178,
      "learning_rate": 4.082840236686391e-08,
      "logits/chosen": -1.7421875,
      "logits/rejected": -1.71875,
      "logps/chosen": -236.0,
      "logps/rejected": -282.0,
      "loss": 0.1584,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.2373046875,
      "rewards/margins": 2.703125,
      "rewards/rejected": -2.46875,
      "step": 870
    },
    {
      "epoch": 2.8115015974440896,
      "grad_norm": 18.902685630993297,
      "learning_rate": 3.4911242603550294e-08,
      "logits/chosen": -1.7578125,
      "logits/rejected": -1.71875,
      "logps/chosen": -242.0,
      "logps/rejected": -282.0,
      "loss": 0.1584,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.1396484375,
      "rewards/margins": 2.625,
      "rewards/rejected": -2.484375,
      "step": 880
    },
    {
      "epoch": 2.8434504792332267,
      "grad_norm": 14.629776336594917,
      "learning_rate": 2.8994082840236687e-08,
      "logits/chosen": -1.6953125,
      "logits/rejected": -1.6953125,
      "logps/chosen": -238.0,
      "logps/rejected": -284.0,
      "loss": 0.1549,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.283203125,
      "rewards/margins": 2.734375,
      "rewards/rejected": -2.453125,
      "step": 890
    },
    {
      "epoch": 2.8753993610223643,
      "grad_norm": 23.27853134409678,
      "learning_rate": 2.3076923076923076e-08,
      "logits/chosen": -1.7421875,
      "logits/rejected": -1.71875,
      "logps/chosen": -241.0,
      "logps/rejected": -284.0,
      "loss": 0.1677,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.08251953125,
      "rewards/margins": 2.5,
      "rewards/rejected": -2.40625,
      "step": 900
    },
    {
      "epoch": 2.9073482428115014,
      "grad_norm": 23.121981374950852,
      "learning_rate": 1.7159763313609465e-08,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.7265625,
      "logps/chosen": -237.0,
      "logps/rejected": -282.0,
      "loss": 0.1647,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.18359375,
      "rewards/margins": 2.5625,
      "rewards/rejected": -2.375,
      "step": 910
    },
    {
      "epoch": 2.939297124600639,
      "grad_norm": 17.364099329669724,
      "learning_rate": 1.1242603550295858e-08,
      "logits/chosen": -1.7109375,
      "logits/rejected": -1.6796875,
      "logps/chosen": -242.0,
      "logps/rejected": -298.0,
      "loss": 0.1599,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.10888671875,
      "rewards/margins": 2.765625,
      "rewards/rejected": -2.65625,
      "step": 920
    },
    {
      "epoch": 2.9712460063897765,
      "grad_norm": 17.00765803212867,
      "learning_rate": 5.325443786982248e-09,
      "logits/chosen": -1.703125,
      "logits/rejected": -1.71875,
      "logps/chosen": -236.0,
      "logps/rejected": -286.0,
      "loss": 0.1526,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.058837890625,
      "rewards/margins": 2.640625,
      "rewards/rejected": -2.578125,
      "step": 930
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -1.6953125,
      "eval_logits/rejected": -1.7265625,
      "eval_logps/chosen": -243.0,
      "eval_logps/rejected": -284.0,
      "eval_loss": 0.5425000190734863,
      "eval_rewards/accuracies": 0.75,
      "eval_rewards/chosen": -0.65234375,
      "eval_rewards/margins": 1.078125,
      "eval_rewards/rejected": -1.734375,
      "eval_runtime": 15.5349,
      "eval_samples_per_second": 12.874,
      "eval_steps_per_second": 0.451,
      "step": 939
    },
    {
      "epoch": 3.0,
      "step": 939,
      "total_flos": 0.0,
      "train_loss": 0.33416016795002995,
      "train_runtime": 4941.0894,
      "train_samples_per_second": 6.064,
      "train_steps_per_second": 0.19
    }
  ],
  "logging_steps": 10,
  "max_steps": 939,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "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
}