{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7246376811594203,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.036231884057971016,
      "grad_norm": 61.75757328159282,
      "learning_rate": 5e-07,
      "logits/chosen": -2.732090473175049,
      "logits/rejected": -2.7100460529327393,
      "logps/chosen": -182.59107971191406,
      "logps/rejected": -189.5584716796875,
      "loss": 0.6889,
      "rewards/accuracies": 0.30000001192092896,
      "rewards/chosen": -0.00281245238147676,
      "rewards/margins": 0.0058334446512162685,
      "rewards/rejected": -0.008645896799862385,
      "step": 5
    },
    {
      "epoch": 0.07246376811594203,
      "grad_norm": 44.951594498596215,
      "learning_rate": 1e-06,
      "logits/chosen": -2.754081964492798,
      "logits/rejected": -2.752152919769287,
      "logps/chosen": -197.337158203125,
      "logps/rejected": -184.00933837890625,
      "loss": 0.6274,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.03147688880562782,
      "rewards/margins": 0.1896156221628189,
      "rewards/rejected": -0.1581387221813202,
      "step": 10
    },
    {
      "epoch": 0.10869565217391304,
      "grad_norm": 51.34158391398985,
      "learning_rate": 9.996221126793764e-07,
      "logits/chosen": -2.694983959197998,
      "logits/rejected": -2.692361831665039,
      "logps/chosen": -203.20387268066406,
      "logps/rejected": -204.64244079589844,
      "loss": 0.5838,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.6150370836257935,
      "rewards/margins": 0.9413955807685852,
      "rewards/rejected": -0.32635849714279175,
      "step": 15
    },
    {
      "epoch": 0.14492753623188406,
      "grad_norm": 34.76477183019994,
      "learning_rate": 9.984890219128145e-07,
      "logits/chosen": -2.612672805786133,
      "logits/rejected": -2.5829074382781982,
      "logps/chosen": -188.62716674804688,
      "logps/rejected": -192.87452697753906,
      "loss": 0.5142,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.8545471429824829,
      "rewards/margins": 1.280996561050415,
      "rewards/rejected": -0.4264492094516754,
      "step": 20
    },
    {
      "epoch": 0.18115942028985507,
      "grad_norm": 36.75278346647978,
      "learning_rate": 9.966024404228493e-07,
      "logits/chosen": -2.450106143951416,
      "logits/rejected": -2.4297895431518555,
      "logps/chosen": -179.98348999023438,
      "logps/rejected": -179.38925170898438,
      "loss": 0.5032,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": 0.38695499300956726,
      "rewards/margins": 0.8900691866874695,
      "rewards/rejected": -0.5031141638755798,
      "step": 25
    },
    {
      "epoch": 0.21739130434782608,
      "grad_norm": 31.781918105397544,
      "learning_rate": 9.939652198703783e-07,
      "logits/chosen": -2.324214458465576,
      "logits/rejected": -2.325657367706299,
      "logps/chosen": -188.5428466796875,
      "logps/rejected": -193.8271942138672,
      "loss": 0.4995,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.6558719873428345,
      "rewards/margins": 1.2207121849060059,
      "rewards/rejected": -0.5648401975631714,
      "step": 30
    },
    {
      "epoch": 0.2536231884057971,
      "grad_norm": 39.36776247005876,
      "learning_rate": 9.905813465442354e-07,
      "logits/chosen": -2.236240863800049,
      "logits/rejected": -2.2105681896209717,
      "logps/chosen": -203.98277282714844,
      "logps/rejected": -194.84640502929688,
      "loss": 0.5091,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.8834564089775085,
      "rewards/margins": 1.2675695419311523,
      "rewards/rejected": -0.3841131329536438,
      "step": 35
    },
    {
      "epoch": 0.2898550724637681,
      "grad_norm": 30.817630358317576,
      "learning_rate": 9.864559353357187e-07,
      "logits/chosen": -2.068774700164795,
      "logits/rejected": -2.0603950023651123,
      "logps/chosen": -182.76817321777344,
      "logps/rejected": -185.9797821044922,
      "loss": 0.4873,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": 1.03325617313385,
      "rewards/margins": 1.0384714603424072,
      "rewards/rejected": -0.005215352866798639,
      "step": 40
    },
    {
      "epoch": 0.32608695652173914,
      "grad_norm": 29.09268118121073,
      "learning_rate": 9.815952220071804e-07,
      "logits/chosen": -1.8718488216400146,
      "logits/rejected": -1.8250553607940674,
      "logps/chosen": -195.60968017578125,
      "logps/rejected": -221.5565643310547,
      "loss": 0.4597,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": 1.3850222826004028,
      "rewards/margins": 1.8469291925430298,
      "rewards/rejected": -0.4619070589542389,
      "step": 45
    },
    {
      "epoch": 0.36231884057971014,
      "grad_norm": 29.526743630011346,
      "learning_rate": 9.76006553766365e-07,
      "logits/chosen": -1.653713583946228,
      "logits/rejected": -1.6171553134918213,
      "logps/chosen": -198.85989379882812,
      "logps/rejected": -203.60678100585938,
      "loss": 0.4516,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": 0.8292047381401062,
      "rewards/margins": 1.6851797103881836,
      "rewards/rejected": -0.8559748530387878,
      "step": 50
    },
    {
      "epoch": 0.36231884057971014,
      "eval_logits/chosen": -1.7065542936325073,
      "eval_logits/rejected": -1.630993127822876,
      "eval_logps/chosen": -192.20655822753906,
      "eval_logps/rejected": -206.51295471191406,
      "eval_loss": 0.4420754015445709,
      "eval_rewards/accuracies": 0.7903226017951965,
      "eval_rewards/chosen": 0.8112886548042297,
      "eval_rewards/margins": 1.641775369644165,
      "eval_rewards/rejected": -0.8304866552352905,
      "eval_runtime": 247.7543,
      "eval_samples_per_second": 15.83,
      "eval_steps_per_second": 0.25,
      "step": 50
    },
    {
      "epoch": 0.39855072463768115,
      "grad_norm": 30.94859785748943,
      "learning_rate": 9.696983781607415e-07,
      "logits/chosen": -1.7253024578094482,
      "logits/rejected": -1.6905288696289062,
      "logps/chosen": -182.9173126220703,
      "logps/rejected": -171.9159698486328,
      "loss": 0.4573,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": 0.6172864437103271,
      "rewards/margins": 1.648385763168335,
      "rewards/rejected": -1.031099557876587,
      "step": 55
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 40.75469044830845,
      "learning_rate": 9.626802303086209e-07,
      "logits/chosen": -1.87893807888031,
      "logits/rejected": -1.8299003839492798,
      "logps/chosen": -186.30145263671875,
      "logps/rejected": -193.9145965576172,
      "loss": 0.4264,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": 0.4657188057899475,
      "rewards/margins": 1.7288262844085693,
      "rewards/rejected": -1.2631075382232666,
      "step": 60
    },
    {
      "epoch": 0.47101449275362317,
      "grad_norm": 35.556274541495966,
      "learning_rate": 9.549627184863528e-07,
      "logits/chosen": -2.016784906387329,
      "logits/rejected": -1.9150521755218506,
      "logps/chosen": -191.3840789794922,
      "logps/rejected": -192.66639709472656,
      "loss": 0.4289,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": 0.0974711924791336,
      "rewards/margins": 1.6010549068450928,
      "rewards/rejected": -1.5035837888717651,
      "step": 65
    },
    {
      "epoch": 0.5072463768115942,
      "grad_norm": 26.46585227154451,
      "learning_rate": 9.465575080933957e-07,
      "logits/chosen": -1.853308916091919,
      "logits/rejected": -1.7947351932525635,
      "logps/chosen": -172.3099822998047,
      "logps/rejected": -208.057373046875,
      "loss": 0.3948,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 0.21238946914672852,
      "rewards/margins": 1.8403332233428955,
      "rewards/rejected": -1.627943754196167,
      "step": 70
    },
    {
      "epoch": 0.5434782608695652,
      "grad_norm": 31.533541728553253,
      "learning_rate": 9.374773040194878e-07,
      "logits/chosen": -1.8850362300872803,
      "logits/rejected": -1.8103622198104858,
      "logps/chosen": -205.5053253173828,
      "logps/rejected": -210.96981811523438,
      "loss": 0.4364,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": 0.17310531437397003,
      "rewards/margins": 1.8103282451629639,
      "rewards/rejected": -1.6372228860855103,
      "step": 75
    },
    {
      "epoch": 0.5797101449275363,
      "grad_norm": 29.780905727815526,
      "learning_rate": 9.277358314405818e-07,
      "logits/chosen": -1.7906593084335327,
      "logits/rejected": -1.742597222328186,
      "logps/chosen": -188.9757080078125,
      "logps/rejected": -205.398193359375,
      "loss": 0.3987,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.19728976488113403,
      "rewards/margins": 1.879663109779358,
      "rewards/rejected": -2.0769529342651367,
      "step": 80
    },
    {
      "epoch": 0.6159420289855072,
      "grad_norm": 34.4646468352745,
      "learning_rate": 9.173478150725651e-07,
      "logits/chosen": -1.7377640008926392,
      "logits/rejected": -1.6257518529891968,
      "logps/chosen": -210.00320434570312,
      "logps/rejected": -215.84835815429688,
      "loss": 0.4258,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 0.08612661063671112,
      "rewards/margins": 2.4435980319976807,
      "rewards/rejected": -2.357471227645874,
      "step": 85
    },
    {
      "epoch": 0.6521739130434783,
      "grad_norm": 29.12537980218493,
      "learning_rate": 9.063289569141251e-07,
      "logits/chosen": -1.7976572513580322,
      "logits/rejected": -1.739854097366333,
      "logps/chosen": -214.8435821533203,
      "logps/rejected": -224.52005004882812,
      "loss": 0.4147,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 0.46363013982772827,
      "rewards/margins": 2.330965518951416,
      "rewards/rejected": -1.867335557937622,
      "step": 90
    },
    {
      "epoch": 0.6884057971014492,
      "grad_norm": 35.00421638148543,
      "learning_rate": 8.946959125124051e-07,
      "logits/chosen": -1.861108422279358,
      "logits/rejected": -1.780923843383789,
      "logps/chosen": -207.5733184814453,
      "logps/rejected": -193.34400939941406,
      "loss": 0.4121,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 0.4294491708278656,
      "rewards/margins": 2.142913341522217,
      "rewards/rejected": -1.7134641408920288,
      "step": 95
    },
    {
      "epoch": 0.7246376811594203,
      "grad_norm": 31.611698501726103,
      "learning_rate": 8.824662657873238e-07,
      "logits/chosen": -1.8221423625946045,
      "logits/rejected": -1.802095651626587,
      "logps/chosen": -173.2090301513672,
      "logps/rejected": -206.5529327392578,
      "loss": 0.3759,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.04721298813819885,
      "rewards/margins": 1.9821780920028687,
      "rewards/rejected": -2.029391050338745,
      "step": 100
    },
    {
      "epoch": 0.7246376811594203,
      "eval_logits/chosen": -1.8523844480514526,
      "eval_logits/rejected": -1.7929590940475464,
      "eval_logps/chosen": -200.7910614013672,
      "eval_logps/rejected": -220.96961975097656,
      "eval_loss": 0.4121003746986389,
      "eval_rewards/accuracies": 0.8145161271095276,
      "eval_rewards/chosen": -0.047160252928733826,
      "eval_rewards/margins": 2.2289960384368896,
      "eval_rewards/rejected": -2.276156187057495,
      "eval_runtime": 247.371,
      "eval_samples_per_second": 15.855,
      "eval_steps_per_second": 0.251,
      "step": 100
    }
  ],
  "logging_steps": 5,
  "max_steps": 414,
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1178822762299392.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}