{
  "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": 0.02,
      "grad_norm": 141.4031591270893,
      "learning_rate": 4.545454545454545e-08,
      "logits/chosen": 111.1229019165039,
      "logits/rejected": 86.78887939453125,
      "logps/chosen": -328.8410339355469,
      "logps/rejected": -329.2784118652344,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.19,
      "grad_norm": 137.1330948401959,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": 110.3634033203125,
      "logits/rejected": 133.23895263671875,
      "logps/chosen": -350.8841247558594,
      "logps/rejected": -434.155029296875,
      "loss": 0.7084,
      "rewards/accuracies": 0.4722222089767456,
      "rewards/chosen": 0.10759051889181137,
      "rewards/margins": 0.03753471374511719,
      "rewards/rejected": 0.07005578279495239,
      "step": 10
    },
    {
      "epoch": 0.38,
      "grad_norm": 118.64244526415024,
      "learning_rate": 4.885348141000122e-07,
      "logits/chosen": 117.46419525146484,
      "logits/rejected": 128.240234375,
      "logps/chosen": -331.4848937988281,
      "logps/rejected": -408.83056640625,
      "loss": 0.6233,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.20835945010185242,
      "rewards/margins": 0.7139766812324524,
      "rewards/rejected": -0.5056172609329224,
      "step": 20
    },
    {
      "epoch": 0.57,
      "grad_norm": 101.92099166540923,
      "learning_rate": 4.5025027361734613e-07,
      "logits/chosen": 115.0364761352539,
      "logits/rejected": 119.716064453125,
      "logps/chosen": -399.8323669433594,
      "logps/rejected": -474.94384765625,
      "loss": 0.6029,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -1.7643420696258545,
      "rewards/margins": 1.0587323904037476,
      "rewards/rejected": -2.8230743408203125,
      "step": 30
    },
    {
      "epoch": 0.76,
      "grad_norm": 107.76688802812863,
      "learning_rate": 3.893311157806091e-07,
      "logits/chosen": 117.44291687011719,
      "logits/rejected": 112.09379577636719,
      "logps/chosen": -424.4700622558594,
      "logps/rejected": -459.14361572265625,
      "loss": 0.5401,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -2.072272539138794,
      "rewards/margins": 0.9466603398323059,
      "rewards/rejected": -3.018933057785034,
      "step": 40
    },
    {
      "epoch": 0.95,
      "grad_norm": 127.24884718202802,
      "learning_rate": 3.126631330646801e-07,
      "logits/chosen": 124.3499984741211,
      "logits/rejected": 125.83009338378906,
      "logps/chosen": -425.4100646972656,
      "logps/rejected": -460.123046875,
      "loss": 0.5226,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -1.6563717126846313,
      "rewards/margins": 1.290310263633728,
      "rewards/rejected": -2.9466819763183594,
      "step": 50
    },
    {
      "epoch": 1.14,
      "grad_norm": 60.21781496264702,
      "learning_rate": 2.2891223348923882e-07,
      "logits/chosen": 122.9765625,
      "logits/rejected": 126.6246566772461,
      "logps/chosen": -393.0502014160156,
      "logps/rejected": -447.2841796875,
      "loss": 0.287,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": -1.311488389968872,
      "rewards/margins": 2.3555827140808105,
      "rewards/rejected": -3.6670711040496826,
      "step": 60
    },
    {
      "epoch": 1.33,
      "grad_norm": 56.930992725835345,
      "learning_rate": 1.4754491880085317e-07,
      "logits/chosen": 117.85044860839844,
      "logits/rejected": 119.93992614746094,
      "logps/chosen": -365.17333984375,
      "logps/rejected": -490.39080810546875,
      "loss": 0.1969,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -1.3022083044052124,
      "rewards/margins": 2.6266796588897705,
      "rewards/rejected": -3.9288878440856934,
      "step": 70
    },
    {
      "epoch": 1.52,
      "grad_norm": 55.076942962446836,
      "learning_rate": 7.775827023107834e-08,
      "logits/chosen": 125.15724182128906,
      "logits/rejected": 126.8897476196289,
      "logps/chosen": -429.802978515625,
      "logps/rejected": -526.1527099609375,
      "loss": 0.1797,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.4758400917053223,
      "rewards/margins": 2.972074270248413,
      "rewards/rejected": -4.4479146003723145,
      "step": 80
    },
    {
      "epoch": 1.71,
      "grad_norm": 59.283310271361565,
      "learning_rate": 2.7440387297912122e-08,
      "logits/chosen": 108.48063659667969,
      "logits/rejected": 113.1912612915039,
      "logps/chosen": -411.52783203125,
      "logps/rejected": -497.0039978027344,
      "loss": 0.1734,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": -2.0875043869018555,
      "rewards/margins": 3.0847535133361816,
      "rewards/rejected": -5.172257423400879,
      "step": 90
    },
    {
      "epoch": 1.9,
      "grad_norm": 60.59000761849891,
      "learning_rate": 2.27878296044029e-09,
      "logits/chosen": 124.4859619140625,
      "logits/rejected": 115.1314697265625,
      "logps/chosen": -430.19573974609375,
      "logps/rejected": -541.2686767578125,
      "loss": 0.1884,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -2.0935351848602295,
      "rewards/margins": 3.198591709136963,
      "rewards/rejected": -5.292126655578613,
      "step": 100
    },
    {
      "epoch": 1.9,
      "eval_logits/chosen": 94.0592269897461,
      "eval_logits/rejected": 88.1426773071289,
      "eval_logps/chosen": -419.4271240234375,
      "eval_logps/rejected": -448.1896057128906,
      "eval_loss": 0.4704847037792206,
      "eval_rewards/accuracies": 0.71875,
      "eval_rewards/chosen": -2.870262861251831,
      "eval_rewards/margins": 1.558273196220398,
      "eval_rewards/rejected": -4.4285359382629395,
      "eval_runtime": 61.9568,
      "eval_samples_per_second": 12.105,
      "eval_steps_per_second": 0.387,
      "step": 100
    },
    {
      "epoch": 1.97,
      "step": 104,
      "total_flos": 0.0,
      "train_loss": 0.3924295770434233,
      "train_runtime": 3265.4264,
      "train_samples_per_second": 4.134,
      "train_steps_per_second": 0.032
    }
  ],
  "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
}