{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9984,
  "eval_steps": 500,
  "global_step": 156,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 3.125e-07,
      "logits/chosen": -0.18967239558696747,
      "logits/rejected": -0.41899582743644714,
      "logps/chosen": -1382.499267578125,
      "logps/rejected": -2863.098388671875,
      "loss": 0.5,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.06,
      "learning_rate": 3.125e-06,
      "logits/chosen": -0.15270556509494781,
      "logits/rejected": -0.2915438413619995,
      "logps/chosen": -2153.170654296875,
      "logps/rejected": -3371.856689453125,
      "loss": 0.4997,
      "rewards/accuracies": 0.3993055522441864,
      "rewards/chosen": -0.007403078954666853,
      "rewards/margins": 0.0016438195016235113,
      "rewards/rejected": -0.00904689822345972,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.989935734988098e-06,
      "logits/chosen": -0.2100469321012497,
      "logits/rejected": -0.32501596212387085,
      "logps/chosen": -2614.16064453125,
      "logps/rejected": -3642.28076171875,
      "loss": 0.4963,
      "rewards/accuracies": 0.4937500059604645,
      "rewards/chosen": -0.1257179230451584,
      "rewards/margins": 0.020404411479830742,
      "rewards/rejected": -0.14612232148647308,
      "step": 20
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.8776412907378845e-06,
      "logits/chosen": -0.19536757469177246,
      "logits/rejected": -0.4390452802181244,
      "logps/chosen": -2770.899169921875,
      "logps/rejected": -4525.7744140625,
      "loss": 0.4889,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": -0.5548638701438904,
      "rewards/margins": 0.1145726665854454,
      "rewards/rejected": -0.6694365739822388,
      "step": 30
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.646121984004666e-06,
      "logits/chosen": -0.23629632592201233,
      "logits/rejected": -0.5095050930976868,
      "logps/chosen": -3372.510498046875,
      "logps/rejected": -5071.37451171875,
      "loss": 0.4868,
      "rewards/accuracies": 0.503125011920929,
      "rewards/chosen": -1.0432322025299072,
      "rewards/margins": 0.22232362627983093,
      "rewards/rejected": -1.26555597782135,
      "step": 40
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.3069871595684795e-06,
      "logits/chosen": -0.32125282287597656,
      "logits/rejected": -0.562100887298584,
      "logps/chosen": -3342.119873046875,
      "logps/rejected": -5389.53955078125,
      "loss": 0.4856,
      "rewards/accuracies": 0.546875,
      "rewards/chosen": -0.9807957410812378,
      "rewards/margins": 0.4719497263431549,
      "rewards/rejected": -1.4527455568313599,
      "step": 50
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.8772424536302565e-06,
      "logits/chosen": -0.3323759138584137,
      "logits/rejected": -0.5304074883460999,
      "logps/chosen": -3480.371826171875,
      "logps/rejected": -4698.51025390625,
      "loss": 0.4883,
      "rewards/accuracies": 0.53125,
      "rewards/chosen": -0.97138512134552,
      "rewards/margins": 0.20621006190776825,
      "rewards/rejected": -1.1775951385498047,
      "step": 60
    },
    {
      "epoch": 0.45,
      "learning_rate": 3.3784370602033572e-06,
      "logits/chosen": -0.3379088044166565,
      "logits/rejected": -0.5837884545326233,
      "logps/chosen": -3488.02587890625,
      "logps/rejected": -5069.7412109375,
      "loss": 0.4871,
      "rewards/accuracies": 0.4937500059604645,
      "rewards/chosen": -1.0387831926345825,
      "rewards/margins": 0.1595762073993683,
      "rewards/rejected": -1.198359489440918,
      "step": 70
    },
    {
      "epoch": 0.51,
      "learning_rate": 2.835583164544139e-06,
      "logits/chosen": -0.2774398624897003,
      "logits/rejected": -0.5058914422988892,
      "logps/chosen": -2817.748779296875,
      "logps/rejected": -4369.88720703125,
      "loss": 0.4891,
      "rewards/accuracies": 0.48124998807907104,
      "rewards/chosen": -0.8934494853019714,
      "rewards/margins": 0.030101608484983444,
      "rewards/rejected": -0.9235512018203735,
      "step": 80
    },
    {
      "epoch": 0.58,
      "learning_rate": 2.2759017277414165e-06,
      "logits/chosen": -0.21915356814861298,
      "logits/rejected": -0.47689905762672424,
      "logps/chosen": -2961.93994140625,
      "logps/rejected": -4903.95166015625,
      "loss": 0.4839,
      "rewards/accuracies": 0.5531250238418579,
      "rewards/chosen": -0.7773372530937195,
      "rewards/margins": 0.3540068566799164,
      "rewards/rejected": -1.1313440799713135,
      "step": 90
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.7274575140626318e-06,
      "logits/chosen": -0.3382042348384857,
      "logits/rejected": -0.6111919283866882,
      "logps/chosen": -2994.310302734375,
      "logps/rejected": -4617.9794921875,
      "loss": 0.4856,
      "rewards/accuracies": 0.4937500059604645,
      "rewards/chosen": -0.9103735089302063,
      "rewards/margins": 0.1410413533449173,
      "rewards/rejected": -1.051414966583252,
      "step": 100
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.217751806485235e-06,
      "logits/chosen": -0.28220734000205994,
      "logits/rejected": -0.5584867000579834,
      "logps/chosen": -2910.51953125,
      "logps/rejected": -4760.12158203125,
      "loss": 0.4865,
      "rewards/accuracies": 0.503125011920929,
      "rewards/chosen": -0.8143825531005859,
      "rewards/margins": 0.21515560150146484,
      "rewards/rejected": -1.0295381546020508,
      "step": 110
    },
    {
      "epoch": 0.77,
      "learning_rate": 7.723433775328385e-07,
      "logits/chosen": -0.281380295753479,
      "logits/rejected": -0.500409722328186,
      "logps/chosen": -2965.12841796875,
      "logps/rejected": -4710.0673828125,
      "loss": 0.486,
      "rewards/accuracies": 0.5531250238418579,
      "rewards/chosen": -0.8462094068527222,
      "rewards/margins": 0.2777535617351532,
      "rewards/rejected": -1.1239629983901978,
      "step": 120
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.1356686569674344e-07,
      "logits/chosen": -0.24006013572216034,
      "logits/rejected": -0.46842899918556213,
      "logps/chosen": -3051.799072265625,
      "logps/rejected": -4552.3515625,
      "loss": 0.4854,
      "rewards/accuracies": 0.546875,
      "rewards/chosen": -0.8304941058158875,
      "rewards/margins": 0.16177485883235931,
      "rewards/rejected": -0.9922689199447632,
      "step": 130
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.59412823400657e-07,
      "logits/chosen": -0.2925623953342438,
      "logits/rejected": -0.4558919072151184,
      "logps/chosen": -3453.52880859375,
      "logps/rejected": -4216.4990234375,
      "loss": 0.4876,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": -0.9790051579475403,
      "rewards/margins": -0.0291127972304821,
      "rewards/rejected": -0.9498924016952515,
      "step": 140
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.262559558016325e-08,
      "logits/chosen": -0.29268592596054077,
      "logits/rejected": -0.5145989656448364,
      "logps/chosen": -2635.45703125,
      "logps/rejected": -4756.46875,
      "loss": 0.4888,
      "rewards/accuracies": 0.546875,
      "rewards/chosen": -0.7879113554954529,
      "rewards/margins": 0.24760587513446808,
      "rewards/rejected": -1.0355170965194702,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 156,
      "total_flos": 0.0,
      "train_loss": 0.4880518607604198,
      "train_runtime": 4047.7642,
      "train_samples_per_second": 4.941,
      "train_steps_per_second": 0.039
    }
  ],
  "logging_steps": 10,
  "max_steps": 156,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}