{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9345794392523364,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04672897196261682,
      "grad_norm": 61.970988574529464,
      "learning_rate": 5e-07,
      "logits/chosen": -2.7293832302093506,
      "logits/rejected": -2.7098002433776855,
      "logps/chosen": -282.619384765625,
      "logps/rejected": -220.626708984375,
      "loss": 0.6898,
      "rewards/accuracies": 0.34375,
      "rewards/chosen": 0.016255810856819153,
      "rewards/margins": 0.00715771596878767,
      "rewards/rejected": 0.009098095819354057,
      "step": 5
    },
    {
      "epoch": 0.09345794392523364,
      "grad_norm": 59.149513121461624,
      "learning_rate": 1e-06,
      "logits/chosen": -2.6811907291412354,
      "logits/rejected": -2.6537957191467285,
      "logps/chosen": -256.4613037109375,
      "logps/rejected": -214.9097442626953,
      "loss": 0.6465,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.46442437171936035,
      "rewards/margins": 0.15177568793296814,
      "rewards/rejected": 0.3126486837863922,
      "step": 10
    },
    {
      "epoch": 0.14018691588785046,
      "grad_norm": 43.868507115118234,
      "learning_rate": 9.993623730611148e-07,
      "logits/chosen": -2.4990592002868652,
      "logits/rejected": -2.4942288398742676,
      "logps/chosen": -252.8370361328125,
      "logps/rejected": -212.9906768798828,
      "loss": 0.6365,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": 1.3768469095230103,
      "rewards/margins": 0.5569905042648315,
      "rewards/rejected": 0.8198563456535339,
      "step": 15
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 40.58644668089117,
      "learning_rate": 9.97451118516912e-07,
      "logits/chosen": -2.357009172439575,
      "logits/rejected": -2.3200223445892334,
      "logps/chosen": -245.38623046875,
      "logps/rejected": -190.89620971679688,
      "loss": 0.6273,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 1.5086402893066406,
      "rewards/margins": 0.8777653574943542,
      "rewards/rejected": 0.6308748722076416,
      "step": 20
    },
    {
      "epoch": 0.2336448598130841,
      "grad_norm": 46.4139582288858,
      "learning_rate": 9.94271111036929e-07,
      "logits/chosen": -2.2126636505126953,
      "logits/rejected": -2.2065536975860596,
      "logps/chosen": -242.53775024414062,
      "logps/rejected": -225.9445343017578,
      "loss": 0.6348,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 1.3302034139633179,
      "rewards/margins": 0.9245456457138062,
      "rewards/rejected": 0.4056577682495117,
      "step": 25
    },
    {
      "epoch": 0.2803738317757009,
      "grad_norm": 64.74245067517272,
      "learning_rate": 9.898304612549066e-07,
      "logits/chosen": -2.183245897293091,
      "logits/rejected": -2.1688549518585205,
      "logps/chosen": -233.8648223876953,
      "logps/rejected": -205.1978759765625,
      "loss": 0.5938,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 1.1376529932022095,
      "rewards/margins": 0.8598777651786804,
      "rewards/rejected": 0.27777519822120667,
      "step": 30
    },
    {
      "epoch": 0.32710280373831774,
      "grad_norm": 51.84730256605764,
      "learning_rate": 9.841404950825536e-07,
      "logits/chosen": -2.227497100830078,
      "logits/rejected": -2.1977126598358154,
      "logps/chosen": -237.3373565673828,
      "logps/rejected": -212.15512084960938,
      "loss": 0.5785,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 1.2570827007293701,
      "rewards/margins": 0.9708169102668762,
      "rewards/rejected": 0.2862659692764282,
      "step": 35
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 40.60340275031408,
      "learning_rate": 9.77215724822721e-07,
      "logits/chosen": -2.243427276611328,
      "logits/rejected": -2.2088351249694824,
      "logps/chosen": -248.1513671875,
      "logps/rejected": -229.47531127929688,
      "loss": 0.5829,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.9161311984062195,
      "rewards/margins": 1.0557386875152588,
      "rewards/rejected": -0.13960735499858856,
      "step": 40
    },
    {
      "epoch": 0.4205607476635514,
      "grad_norm": 34.21102418831377,
      "learning_rate": 9.69073812155662e-07,
      "logits/chosen": -2.2237961292266846,
      "logits/rejected": -2.2018837928771973,
      "logps/chosen": -269.6021423339844,
      "logps/rejected": -224.58645629882812,
      "loss": 0.5896,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 1.0438940525054932,
      "rewards/margins": 0.9452205896377563,
      "rewards/rejected": 0.09867370873689651,
      "step": 45
    },
    {
      "epoch": 0.4672897196261682,
      "grad_norm": 42.29092168681447,
      "learning_rate": 9.597355230927788e-07,
      "logits/chosen": -2.0892410278320312,
      "logits/rejected": -2.0624351501464844,
      "logps/chosen": -240.1529083251953,
      "logps/rejected": -198.04843139648438,
      "loss": 0.5529,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": 1.232414722442627,
      "rewards/margins": 1.3885271549224854,
      "rewards/rejected": -0.156112402677536,
      "step": 50
    },
    {
      "epoch": 0.4672897196261682,
      "eval_logits/chosen": -2.0593109130859375,
      "eval_logits/rejected": -2.02968692779541,
      "eval_logps/chosen": -239.67454528808594,
      "eval_logps/rejected": -216.6856689453125,
      "eval_loss": 0.5947180390357971,
      "eval_rewards/accuracies": 0.7317708134651184,
      "eval_rewards/chosen": 0.8738771080970764,
      "eval_rewards/margins": 1.1332703828811646,
      "eval_rewards/rejected": -0.2593933641910553,
      "eval_runtime": 202.7096,
      "eval_samples_per_second": 14.997,
      "eval_steps_per_second": 0.237,
      "step": 50
    },
    {
      "epoch": 0.514018691588785,
      "grad_norm": 35.93324426868974,
      "learning_rate": 9.4922467501275e-07,
      "logits/chosen": -2.0398037433624268,
      "logits/rejected": -2.0083518028259277,
      "logps/chosen": -261.76177978515625,
      "logps/rejected": -210.04684448242188,
      "loss": 0.5467,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.9814583659172058,
      "rewards/margins": 1.524784803390503,
      "rewards/rejected": -0.5433263778686523,
      "step": 55
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 50.252777089157334,
      "learning_rate": 9.375680759151206e-07,
      "logits/chosen": -2.092087984085083,
      "logits/rejected": -2.098419427871704,
      "logps/chosen": -252.8424530029297,
      "logps/rejected": -214.93258666992188,
      "loss": 0.5409,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 0.7585053443908691,
      "rewards/margins": 1.2850215435028076,
      "rewards/rejected": -0.5265161395072937,
      "step": 60
    },
    {
      "epoch": 0.6074766355140186,
      "grad_norm": 43.0211288247323,
      "learning_rate": 9.247954560462927e-07,
      "logits/chosen": -2.0893867015838623,
      "logits/rejected": -2.0975987911224365,
      "logps/chosen": -241.7879180908203,
      "logps/rejected": -239.7193145751953,
      "loss": 0.5337,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.8641234636306763,
      "rewards/margins": 1.493786096572876,
      "rewards/rejected": -0.6296626329421997,
      "step": 65
    },
    {
      "epoch": 0.6542056074766355,
      "grad_norm": 38.40101091003172,
      "learning_rate": 9.109393920723001e-07,
      "logits/chosen": -2.0060245990753174,
      "logits/rejected": -1.9721952676773071,
      "logps/chosen": -256.7455139160156,
      "logps/rejected": -214.79403686523438,
      "loss": 0.533,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.6738287210464478,
      "rewards/margins": 1.2377485036849976,
      "rewards/rejected": -0.563919723033905,
      "step": 70
    },
    {
      "epoch": 0.7009345794392523,
      "grad_norm": 40.40382308924294,
      "learning_rate": 8.960352239917699e-07,
      "logits/chosen": -1.9107002019882202,
      "logits/rejected": -1.8710416555404663,
      "logps/chosen": -246.01028442382812,
      "logps/rejected": -224.3345947265625,
      "loss": 0.5452,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.6804067492485046,
      "rewards/margins": 1.3526126146316528,
      "rewards/rejected": -0.6722058653831482,
      "step": 75
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 37.2632538141041,
      "learning_rate": 8.801209650009814e-07,
      "logits/chosen": -1.8342845439910889,
      "logits/rejected": -1.7722011804580688,
      "logps/chosen": -238.6627960205078,
      "logps/rejected": -204.77720642089844,
      "loss": 0.4954,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": 0.8959934115409851,
      "rewards/margins": 1.759472131729126,
      "rewards/rejected": -0.8634785413742065,
      "step": 80
    },
    {
      "epoch": 0.794392523364486,
      "grad_norm": 41.835599660746425,
      "learning_rate": 8.632372045409141e-07,
      "logits/chosen": -1.9798192977905273,
      "logits/rejected": -1.9717410802841187,
      "logps/chosen": -251.6102294921875,
      "logps/rejected": -240.1038360595703,
      "loss": 0.5641,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 1.3537020683288574,
      "rewards/margins": 1.3792067766189575,
      "rewards/rejected": -0.025504767894744873,
      "step": 85
    },
    {
      "epoch": 0.8411214953271028,
      "grad_norm": 33.80081107648868,
      "learning_rate": 8.454270047735642e-07,
      "logits/chosen": -2.012608051300049,
      "logits/rejected": -2.0120816230773926,
      "logps/chosen": -253.6774139404297,
      "logps/rejected": -194.4556884765625,
      "loss": 0.5553,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.8388773798942566,
      "rewards/margins": 1.2849448919296265,
      "rewards/rejected": -0.4460674822330475,
      "step": 90
    },
    {
      "epoch": 0.8878504672897196,
      "grad_norm": 29.38305905850477,
      "learning_rate": 8.267357907515661e-07,
      "logits/chosen": -1.9744971990585327,
      "logits/rejected": -1.9980430603027344,
      "logps/chosen": -264.6416320800781,
      "logps/rejected": -246.7568817138672,
      "loss": 0.538,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.25706252455711365,
      "rewards/margins": 2.062177896499634,
      "rewards/rejected": -1.8051154613494873,
      "step": 95
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 33.767105871104185,
      "learning_rate": 8.072112345612433e-07,
      "logits/chosen": -1.9495391845703125,
      "logits/rejected": -1.8999382257461548,
      "logps/chosen": -248.95272827148438,
      "logps/rejected": -255.79019165039062,
      "loss": 0.5159,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": -0.18092432618141174,
      "rewards/margins": 1.6218032836914062,
      "rewards/rejected": -1.8027276992797852,
      "step": 100
    },
    {
      "epoch": 0.9345794392523364,
      "eval_logits/chosen": -1.9645830392837524,
      "eval_logits/rejected": -1.9168211221694946,
      "eval_logps/chosen": -250.5727081298828,
      "eval_logps/rejected": -234.28236389160156,
      "eval_loss": 0.5285552144050598,
      "eval_rewards/accuracies": 0.78125,
      "eval_rewards/chosen": -0.21594171226024628,
      "eval_rewards/margins": 1.8031220436096191,
      "eval_rewards/rejected": -2.019063711166382,
      "eval_runtime": 202.4154,
      "eval_samples_per_second": 15.019,
      "eval_steps_per_second": 0.237,
      "step": 100
    }
  ],
  "logging_steps": 5,
  "max_steps": 321,
  "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
}