{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9961977186311787,
  "eval_steps": 100,
  "global_step": 131,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 7.4010863350398886,
      "learning_rate": 3.571428571428571e-08,
      "logits/chosen": -0.4023560881614685,
      "logits/rejected": -0.03399945795536041,
      "logps/chosen": -348.97918701171875,
      "logps/rejected": -524.31689453125,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.936227880426271,
      "learning_rate": 3.5714285714285716e-07,
      "logits/chosen": -0.5313352942466736,
      "logits/rejected": -0.32142913341522217,
      "logps/chosen": -339.4945068359375,
      "logps/rejected": -743.97900390625,
      "loss": 0.6925,
      "rewards/accuracies": 0.4652777910232544,
      "rewards/chosen": -0.00040899438317865133,
      "rewards/margins": 0.0011602669255807996,
      "rewards/rejected": -0.001569261308759451,
      "step": 10
    },
    {
      "epoch": 0.15,
      "grad_norm": 11.613933346354703,
      "learning_rate": 4.967625656594781e-07,
      "logits/chosen": -0.5781797170639038,
      "logits/rejected": -0.3319936990737915,
      "logps/chosen": -303.1843566894531,
      "logps/rejected": -763.7269897460938,
      "loss": 0.6663,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": -0.006486949510872364,
      "rewards/margins": 0.046987053006887436,
      "rewards/rejected": -0.053474001586437225,
      "step": 20
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.944322603367183,
      "learning_rate": 4.772810749506809e-07,
      "logits/chosen": -0.4766184687614441,
      "logits/rejected": -0.320529967546463,
      "logps/chosen": -367.22760009765625,
      "logps/rejected": -772.4141845703125,
      "loss": 0.5753,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": -0.0389162078499794,
      "rewards/margins": 0.31977471709251404,
      "rewards/rejected": -0.35869091749191284,
      "step": 30
    },
    {
      "epoch": 0.3,
      "grad_norm": 6.585270206448342,
      "learning_rate": 4.415111107797445e-07,
      "logits/chosen": -0.5007311105728149,
      "logits/rejected": -0.31793397665023804,
      "logps/chosen": -339.66522216796875,
      "logps/rejected": -744.0791015625,
      "loss": 0.4984,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": -0.06443960964679718,
      "rewards/margins": 0.832168459892273,
      "rewards/rejected": -0.8966080546379089,
      "step": 40
    },
    {
      "epoch": 0.38,
      "grad_norm": 5.431472038503189,
      "learning_rate": 3.920161866827889e-07,
      "logits/chosen": -0.5240360498428345,
      "logits/rejected": -0.3544757664203644,
      "logps/chosen": -341.50177001953125,
      "logps/rejected": -987.6541748046875,
      "loss": 0.3943,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -0.15628758072853088,
      "rewards/margins": 2.394836902618408,
      "rewards/rejected": -2.5511245727539062,
      "step": 50
    },
    {
      "epoch": 0.46,
      "grad_norm": 5.4415990048052185,
      "learning_rate": 3.323434381080199e-07,
      "logits/chosen": -0.5124906301498413,
      "logits/rejected": -0.3748542368412018,
      "logps/chosen": -387.4263916015625,
      "logps/rejected": -991.39697265625,
      "loss": 0.3611,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.2954562306404114,
      "rewards/margins": 2.682974338531494,
      "rewards/rejected": -2.97843074798584,
      "step": 60
    },
    {
      "epoch": 0.53,
      "grad_norm": 4.921715920068345,
      "learning_rate": 2.667694110977506e-07,
      "logits/chosen": -0.5369243025779724,
      "logits/rejected": -0.4064721167087555,
      "logps/chosen": -352.8229675292969,
      "logps/rejected": -1028.784912109375,
      "loss": 0.3188,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.32619190216064453,
      "rewards/margins": 3.4282054901123047,
      "rewards/rejected": -3.754396915435791,
      "step": 70
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.517789071403685,
      "learning_rate": 1.9999357655598891e-07,
      "logits/chosen": -0.5413460731506348,
      "logits/rejected": -0.48942112922668457,
      "logps/chosen": -381.2135925292969,
      "logps/rejected": -1212.373779296875,
      "loss": 0.2756,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": -0.3778177797794342,
      "rewards/margins": 4.809152126312256,
      "rewards/rejected": -5.186970233917236,
      "step": 80
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.895205453434471,
      "learning_rate": 1.3680153490759073e-07,
      "logits/chosen": -0.4575203061103821,
      "logits/rejected": -0.43544453382492065,
      "logps/chosen": -313.4459533691406,
      "logps/rejected": -1254.440673828125,
      "loss": 0.276,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": -0.2767553925514221,
      "rewards/margins": 5.2881364822387695,
      "rewards/rejected": -5.564891815185547,
      "step": 90
    },
    {
      "epoch": 0.76,
      "grad_norm": 4.400965922122272,
      "learning_rate": 8.17220481431074e-08,
      "logits/chosen": -0.4834328591823578,
      "logits/rejected": -0.4576188027858734,
      "logps/chosen": -366.92529296875,
      "logps/rejected": -1311.052734375,
      "loss": 0.2708,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.38915878534317017,
      "rewards/margins": 5.743791103363037,
      "rewards/rejected": -6.1329498291015625,
      "step": 100
    },
    {
      "epoch": 0.76,
      "eval_logits/chosen": -0.4292154312133789,
      "eval_logits/rejected": -0.47080573439598083,
      "eval_logps/chosen": -476.12548828125,
      "eval_logps/rejected": -635.9033203125,
      "eval_loss": 0.4682200849056244,
      "eval_rewards/accuracies": 0.925000011920929,
      "eval_rewards/chosen": -0.1608993411064148,
      "eval_rewards/margins": 0.9414912462234497,
      "eval_rewards/rejected": -1.1023906469345093,
      "eval_runtime": 39.9857,
      "eval_samples_per_second": 8.003,
      "eval_steps_per_second": 0.25,
      "step": 100
    },
    {
      "epoch": 0.84,
      "grad_norm": 4.121824084175862,
      "learning_rate": 3.8702478614051345e-08,
      "logits/chosen": -0.3803841173648834,
      "logits/rejected": -0.4126942753791809,
      "logps/chosen": -375.005126953125,
      "logps/rejected": -1269.2353515625,
      "loss": 0.2576,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": -0.4322739243507385,
      "rewards/margins": 5.517593860626221,
      "rewards/rejected": -5.949868202209473,
      "step": 110
    },
    {
      "epoch": 0.91,
      "grad_norm": 3.4792807330473705,
      "learning_rate": 1.082589479314372e-08,
      "logits/chosen": -0.4776890277862549,
      "logits/rejected": -0.5241965055465698,
      "logps/chosen": -376.96905517578125,
      "logps/rejected": -1401.2740478515625,
      "loss": 0.2385,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": -0.4306109845638275,
      "rewards/margins": 5.790417194366455,
      "rewards/rejected": -6.2210283279418945,
      "step": 120
    },
    {
      "epoch": 0.99,
      "grad_norm": 5.31082634895388,
      "learning_rate": 9.011808227865625e-11,
      "logits/chosen": -0.47645965218544006,
      "logits/rejected": -0.45030665397644043,
      "logps/chosen": -378.35400390625,
      "logps/rejected": -1342.3602294921875,
      "loss": 0.2593,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.346562922000885,
      "rewards/margins": 5.536257743835449,
      "rewards/rejected": -5.882820129394531,
      "step": 130
    },
    {
      "epoch": 1.0,
      "step": 131,
      "total_flos": 0.0,
      "train_loss": 0.05898992068894947,
      "train_runtime": 471.0672,
      "train_samples_per_second": 17.853,
      "train_steps_per_second": 0.278
    }
  ],
  "logging_steps": 10,
  "max_steps": 131,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}