{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.971563981042654,
  "eval_steps": 100,
  "global_step": 104,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.018957345971563982,
      "grad_norm": 264.3919199015695,
      "learning_rate": 4.545454545454545e-08,
      "logits/chosen": 117.53560638427734,
      "logits/rejected": 126.8960952758789,
      "logps/chosen": -335.40118408203125,
      "logps/rejected": -439.16552734375,
      "loss": 1.0,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.1895734597156398,
      "grad_norm": 268.0892149105686,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": 135.0012969970703,
      "logits/rejected": 138.34600830078125,
      "logps/chosen": -395.8360595703125,
      "logps/rejected": -439.23095703125,
      "loss": 0.9693,
      "rewards/accuracies": 0.4513888955116272,
      "rewards/chosen": 0.008023276925086975,
      "rewards/margins": 0.0031940473709255457,
      "rewards/rejected": 0.004829231183975935,
      "step": 10
    },
    {
      "epoch": 0.3791469194312796,
      "grad_norm": 204.23414477809496,
      "learning_rate": 4.885348141000122e-07,
      "logits/chosen": 121.4811019897461,
      "logits/rejected": 125.18589782714844,
      "logps/chosen": -370.91253662109375,
      "logps/rejected": -425.2193298339844,
      "loss": 0.8003,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": 0.07497303187847137,
      "rewards/margins": 0.33636990189552307,
      "rewards/rejected": -0.2613968253135681,
      "step": 20
    },
    {
      "epoch": 0.5687203791469194,
      "grad_norm": 173.7304699078016,
      "learning_rate": 4.5025027361734613e-07,
      "logits/chosen": 141.61241149902344,
      "logits/rejected": 135.17759704589844,
      "logps/chosen": -426.39410400390625,
      "logps/rejected": -472.140625,
      "loss": 0.7258,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": -1.6964428424835205,
      "rewards/margins": 0.9104781150817871,
      "rewards/rejected": -2.6069209575653076,
      "step": 30
    },
    {
      "epoch": 0.7582938388625592,
      "grad_norm": 189.807566233107,
      "learning_rate": 3.893311157806091e-07,
      "logits/chosen": 125.9510498046875,
      "logits/rejected": 114.42036437988281,
      "logps/chosen": -409.6640625,
      "logps/rejected": -436.3352966308594,
      "loss": 0.6779,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -2.7734901905059814,
      "rewards/margins": 1.1495102643966675,
      "rewards/rejected": -3.9230003356933594,
      "step": 40
    },
    {
      "epoch": 0.9478672985781991,
      "grad_norm": 208.75706608415575,
      "learning_rate": 3.126631330646801e-07,
      "logits/chosen": 141.94976806640625,
      "logits/rejected": 145.94911193847656,
      "logps/chosen": -465.7303161621094,
      "logps/rejected": -547.1525268554688,
      "loss": 0.5763,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -2.3293235301971436,
      "rewards/margins": 1.2119842767715454,
      "rewards/rejected": -3.5413079261779785,
      "step": 50
    },
    {
      "epoch": 1.1374407582938388,
      "grad_norm": 137.79002373576745,
      "learning_rate": 2.2891223348923882e-07,
      "logits/chosen": 134.63436889648438,
      "logits/rejected": 138.12661743164062,
      "logps/chosen": -450.57525634765625,
      "logps/rejected": -530.6395263671875,
      "loss": 0.3625,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": -2.628477096557617,
      "rewards/margins": 2.0852205753326416,
      "rewards/rejected": -4.713697910308838,
      "step": 60
    },
    {
      "epoch": 1.3270142180094786,
      "grad_norm": 111.85723665841337,
      "learning_rate": 1.4754491880085317e-07,
      "logits/chosen": 128.9261932373047,
      "logits/rejected": 130.80894470214844,
      "logps/chosen": -414.4986877441406,
      "logps/rejected": -507.4915466308594,
      "loss": 0.2163,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -2.367741823196411,
      "rewards/margins": 2.3311409950256348,
      "rewards/rejected": -4.698883056640625,
      "step": 70
    },
    {
      "epoch": 1.5165876777251186,
      "grad_norm": 115.1507773079353,
      "learning_rate": 7.775827023107834e-08,
      "logits/chosen": 116.55960845947266,
      "logits/rejected": 133.2828369140625,
      "logps/chosen": -403.3376770019531,
      "logps/rejected": -510.32763671875,
      "loss": 0.1811,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -2.4843320846557617,
      "rewards/margins": 2.370164394378662,
      "rewards/rejected": -4.854496955871582,
      "step": 80
    },
    {
      "epoch": 1.7061611374407581,
      "grad_norm": 106.89250321890712,
      "learning_rate": 2.7440387297912122e-08,
      "logits/chosen": 116.5929183959961,
      "logits/rejected": 129.53289794921875,
      "logps/chosen": -430.5577087402344,
      "logps/rejected": -534.07861328125,
      "loss": 0.1604,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -2.206333637237549,
      "rewards/margins": 2.613506555557251,
      "rewards/rejected": -4.819840431213379,
      "step": 90
    },
    {
      "epoch": 1.8957345971563981,
      "grad_norm": 92.45827774262504,
      "learning_rate": 2.27878296044029e-09,
      "logits/chosen": 122.9133529663086,
      "logits/rejected": 122.8787612915039,
      "logps/chosen": -420.15228271484375,
      "logps/rejected": -503.9156188964844,
      "loss": 0.1516,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -1.9088605642318726,
      "rewards/margins": 2.448718547821045,
      "rewards/rejected": -4.357579231262207,
      "step": 100
    },
    {
      "epoch": 1.8957345971563981,
      "eval_logits/chosen": 101.81689453125,
      "eval_logits/rejected": 95.97393798828125,
      "eval_logps/chosen": -416.4397888183594,
      "eval_logps/rejected": -440.0738830566406,
      "eval_loss": 0.5250489711761475,
      "eval_rewards/accuracies": 0.71875,
      "eval_rewards/chosen": -2.638577699661255,
      "eval_rewards/margins": 1.2615679502487183,
      "eval_rewards/rejected": -3.9001457691192627,
      "eval_runtime": 123.0399,
      "eval_samples_per_second": 6.096,
      "eval_steps_per_second": 0.195,
      "step": 100
    },
    {
      "epoch": 1.971563981042654,
      "step": 104,
      "total_flos": 0.0,
      "train_loss": 0.4740314678503917,
      "train_runtime": 2291.7376,
      "train_samples_per_second": 5.891,
      "train_steps_per_second": 0.045
    }
  ],
  "logging_steps": 10,
  "max_steps": 104,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}