{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 7.207207207207207,
  "eval_steps": 500,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.36036036036036034,
      "grad_norm": 4.098762512207031,
      "learning_rate": 9.259259259259259e-07,
      "logits/chosen": -2.332144260406494,
      "logits/rejected": -2.3385167121887207,
      "logps/chosen": -80.89369201660156,
      "logps/rejected": -70.11573791503906,
      "loss": 0.6929,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 0.00037987352698110044,
      "rewards/margins": 0.004227532539516687,
      "rewards/rejected": -0.003847658634185791,
      "step": 10
    },
    {
      "epoch": 0.7207207207207207,
      "grad_norm": 3.641324281692505,
      "learning_rate": 1.8518518518518519e-06,
      "logits/chosen": -2.323789119720459,
      "logits/rejected": -2.351041793823242,
      "logps/chosen": -73.2725601196289,
      "logps/rejected": -81.80250549316406,
      "loss": 0.6932,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.0012983012711629272,
      "rewards/margins": 0.004207946360111237,
      "rewards/rejected": -0.0029096449725329876,
      "step": 20
    },
    {
      "epoch": 1.0810810810810811,
      "grad_norm": 3.9276015758514404,
      "learning_rate": 2.7777777777777783e-06,
      "logits/chosen": -2.3353028297424316,
      "logits/rejected": -2.3445916175842285,
      "logps/chosen": -69.34381103515625,
      "logps/rejected": -74.37530517578125,
      "loss": 0.6941,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": -0.01035231165587902,
      "rewards/margins": -0.006389107555150986,
      "rewards/rejected": -0.00396320316940546,
      "step": 30
    },
    {
      "epoch": 1.4414414414414414,
      "grad_norm": 4.809000492095947,
      "learning_rate": 3.7037037037037037e-06,
      "logits/chosen": -2.343184232711792,
      "logits/rejected": -2.360262393951416,
      "logps/chosen": -77.91002655029297,
      "logps/rejected": -76.27156066894531,
      "loss": 0.6902,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.008437180891633034,
      "rewards/margins": 0.008391124196350574,
      "rewards/rejected": -0.016828304156661034,
      "step": 40
    },
    {
      "epoch": 1.8018018018018018,
      "grad_norm": 4.504117012023926,
      "learning_rate": 4.62962962962963e-06,
      "logits/chosen": -2.3394973278045654,
      "logits/rejected": -2.3635268211364746,
      "logps/chosen": -83.62376403808594,
      "logps/rejected": -267.64569091796875,
      "loss": 0.6851,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 0.01695835217833519,
      "rewards/margins": 0.14291717112064362,
      "rewards/rejected": -0.12595881521701813,
      "step": 50
    },
    {
      "epoch": 2.1621621621621623,
      "grad_norm": 4.033559322357178,
      "learning_rate": 4.998119881260576e-06,
      "logits/chosen": -2.32966685295105,
      "logits/rejected": -2.3370490074157715,
      "logps/chosen": -78.54629516601562,
      "logps/rejected": -82.67992401123047,
      "loss": 0.6767,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.03485359251499176,
      "rewards/margins": 0.035757362842559814,
      "rewards/rejected": -0.07061095535755157,
      "step": 60
    },
    {
      "epoch": 2.5225225225225225,
      "grad_norm": 4.979142189025879,
      "learning_rate": 4.9866405060165044e-06,
      "logits/chosen": -2.364291191101074,
      "logits/rejected": -2.376107931137085,
      "logps/chosen": -70.61842346191406,
      "logps/rejected": -81.80282592773438,
      "loss": 0.6636,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.06025966256856918,
      "rewards/margins": 0.03742799907922745,
      "rewards/rejected": -0.09768766909837723,
      "step": 70
    },
    {
      "epoch": 2.8828828828828827,
      "grad_norm": 4.0428690910339355,
      "learning_rate": 4.964774158361991e-06,
      "logits/chosen": -2.3341965675354004,
      "logits/rejected": -2.3440909385681152,
      "logps/chosen": -86.3591537475586,
      "logps/rejected": -77.45347595214844,
      "loss": 0.6519,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": -0.09531830251216888,
      "rewards/margins": 0.09854079782962799,
      "rewards/rejected": -0.19385910034179688,
      "step": 80
    },
    {
      "epoch": 3.2432432432432434,
      "grad_norm": 4.31919527053833,
      "learning_rate": 4.93261217644956e-06,
      "logits/chosen": -2.351658821105957,
      "logits/rejected": -2.3437318801879883,
      "logps/chosen": -77.31346130371094,
      "logps/rejected": -80.43277740478516,
      "loss": 0.6243,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.1381106823682785,
      "rewards/margins": 0.16527524590492249,
      "rewards/rejected": -0.3033859133720398,
      "step": 90
    },
    {
      "epoch": 3.6036036036036037,
      "grad_norm": 5.029748439788818,
      "learning_rate": 4.8902889044347e-06,
      "logits/chosen": -2.3354241847991943,
      "logits/rejected": -2.358518600463867,
      "logps/chosen": -75.03588104248047,
      "logps/rejected": -86.44483947753906,
      "loss": 0.6025,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.22239580750465393,
      "rewards/margins": 0.1877792775630951,
      "rewards/rejected": -0.41017502546310425,
      "step": 100
    },
    {
      "epoch": 3.963963963963964,
      "grad_norm": 4.6208648681640625,
      "learning_rate": 4.837981131305475e-06,
      "logits/chosen": -2.3195366859436035,
      "logits/rejected": -2.3129196166992188,
      "logps/chosen": -72.09532928466797,
      "logps/rejected": -73.18878936767578,
      "loss": 0.5955,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.22240504622459412,
      "rewards/margins": 0.22803232073783875,
      "rewards/rejected": -0.45043739676475525,
      "step": 110
    },
    {
      "epoch": 4.324324324324325,
      "grad_norm": 4.163040637969971,
      "learning_rate": 4.775907352415367e-06,
      "logits/chosen": -2.3427720069885254,
      "logits/rejected": -2.3731276988983154,
      "logps/chosen": -85.9415283203125,
      "logps/rejected": -93.53765869140625,
      "loss": 0.5506,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.24449148774147034,
      "rewards/margins": 0.3563767373561859,
      "rewards/rejected": -0.6008682250976562,
      "step": 120
    },
    {
      "epoch": 4.684684684684685,
      "grad_norm": 4.228254795074463,
      "learning_rate": 4.70432685680402e-06,
      "logits/chosen": -2.336733341217041,
      "logits/rejected": -2.3446521759033203,
      "logps/chosen": -81.07231140136719,
      "logps/rejected": -90.82849884033203,
      "loss": 0.5248,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -0.005195322446525097,
      "rewards/margins": 0.6936509609222412,
      "rewards/rejected": -0.6988462209701538,
      "step": 130
    },
    {
      "epoch": 5.045045045045045,
      "grad_norm": 4.454960346221924,
      "learning_rate": 4.623538644118244e-06,
      "logits/chosen": -2.3331754207611084,
      "logits/rejected": -2.3434836864471436,
      "logps/chosen": -83.67604064941406,
      "logps/rejected": -82.92774200439453,
      "loss": 0.5288,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.2716079652309418,
      "rewards/margins": 0.4593236446380615,
      "rewards/rejected": -0.7309316396713257,
      "step": 140
    },
    {
      "epoch": 5.405405405405405,
      "grad_norm": 5.223482608795166,
      "learning_rate": 4.533880175657419e-06,
      "logits/chosen": -2.362809658050537,
      "logits/rejected": -2.3657679557800293,
      "logps/chosen": -73.20018768310547,
      "logps/rejected": -85.37998962402344,
      "loss": 0.4682,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.3221462368965149,
      "rewards/margins": 0.5968301892280579,
      "rewards/rejected": -0.9189764261245728,
      "step": 150
    },
    {
      "epoch": 5.7657657657657655,
      "grad_norm": 4.905521869659424,
      "learning_rate": 4.435725964760331e-06,
      "logits/chosen": -2.3808655738830566,
      "logits/rejected": -2.368286609649658,
      "logps/chosen": -68.88943481445312,
      "logps/rejected": -82.69029235839844,
      "loss": 0.4586,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.3172217011451721,
      "rewards/margins": 0.7665462493896484,
      "rewards/rejected": -1.0837678909301758,
      "step": 160
    },
    {
      "epoch": 6.126126126126126,
      "grad_norm": 5.399628162384033,
      "learning_rate": 4.329486012421531e-06,
      "logits/chosen": -2.365935802459717,
      "logits/rejected": -2.363004684448242,
      "logps/chosen": -70.47642517089844,
      "logps/rejected": -84.02542877197266,
      "loss": 0.4462,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": -0.45835933089256287,
      "rewards/margins": 0.8438631892204285,
      "rewards/rejected": -1.302222490310669,
      "step": 170
    },
    {
      "epoch": 6.486486486486487,
      "grad_norm": 4.843445777893066,
      "learning_rate": 4.215604094671835e-06,
      "logits/chosen": -2.357231855392456,
      "logits/rejected": -2.360239028930664,
      "logps/chosen": -78.67561340332031,
      "logps/rejected": -88.39659118652344,
      "loss": 0.3976,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.4842923581600189,
      "rewards/margins": 0.8027322888374329,
      "rewards/rejected": -1.2870244979858398,
      "step": 180
    },
    {
      "epoch": 6.846846846846847,
      "grad_norm": 4.972764015197754,
      "learning_rate": 4.094555908876765e-06,
      "logits/chosen": -2.3751468658447266,
      "logits/rejected": -2.3993237018585205,
      "logps/chosen": -73.63652038574219,
      "logps/rejected": -278.0970458984375,
      "loss": 0.3959,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.4291106164455414,
      "rewards/margins": 0.9967883229255676,
      "rewards/rejected": -1.4258991479873657,
      "step": 190
    },
    {
      "epoch": 7.207207207207207,
      "grad_norm": 5.071193218231201,
      "learning_rate": 3.966847086696045e-06,
      "logits/chosen": -2.3572330474853516,
      "logits/rejected": -2.357269763946533,
      "logps/chosen": -84.92713928222656,
      "logps/rejected": -98.15062713623047,
      "loss": 0.3544,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.5852295756340027,
      "rewards/margins": 1.2983506917953491,
      "rewards/rejected": -1.883580207824707,
      "step": 200
    }
  ],
  "logging_steps": 10,
  "max_steps": 540,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 7.302906796614615e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}