{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.971563981042654,
  "eval_steps": 128,
  "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": 198.31413823442554,
      "learning_rate": 4.545454545454545e-08,
      "logits/chosen": 117.53560638427734,
      "logits/rejected": 126.8960952758789,
      "logps/chosen": -335.40118408203125,
      "logps/rejected": -439.16552734375,
      "loss": 0.8466,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.1895734597156398,
      "grad_norm": 207.70800418455437,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": 134.9920654296875,
      "logits/rejected": 138.34112548828125,
      "logps/chosen": -396.8204040527344,
      "logps/rejected": -440.0408020019531,
      "loss": 0.9152,
      "rewards/accuracies": 0.4305555522441864,
      "rewards/chosen": -0.04119214043021202,
      "rewards/margins": -0.00552941020578146,
      "rewards/rejected": -0.035662729293107986,
      "step": 10
    },
    {
      "epoch": 0.3791469194312796,
      "grad_norm": 375.1572055463285,
      "learning_rate": 4.885348141000122e-07,
      "logits/chosen": 122.67049407958984,
      "logits/rejected": 126.37223815917969,
      "logps/chosen": -373.7001953125,
      "logps/rejected": -425.6162109375,
      "loss": 0.906,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": -0.06441108137369156,
      "rewards/margins": 0.21682703495025635,
      "rewards/rejected": -0.2812381088733673,
      "step": 20
    },
    {
      "epoch": 0.5687203791469194,
      "grad_norm": 333.9265607541922,
      "learning_rate": 4.5025027361734613e-07,
      "logits/chosen": 144.07449340820312,
      "logits/rejected": 137.60853576660156,
      "logps/chosen": -403.1298828125,
      "logps/rejected": -441.2322692871094,
      "loss": 0.8322,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": -0.5332325100898743,
      "rewards/margins": 0.5282701253890991,
      "rewards/rejected": -1.0615026950836182,
      "step": 30
    },
    {
      "epoch": 0.7582938388625592,
      "grad_norm": 232.4211596454109,
      "learning_rate": 3.893311157806091e-07,
      "logits/chosen": 125.85306549072266,
      "logits/rejected": 114.46476745605469,
      "logps/chosen": -358.6048278808594,
      "logps/rejected": -375.98065185546875,
      "loss": 0.7672,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.2205326110124588,
      "rewards/margins": 0.6847385168075562,
      "rewards/rejected": -0.9052711725234985,
      "step": 40
    },
    {
      "epoch": 0.9478672985781991,
      "grad_norm": 327.7428496612311,
      "learning_rate": 3.126631330646801e-07,
      "logits/chosen": 142.5763397216797,
      "logits/rejected": 146.8212432861328,
      "logps/chosen": -427.7535705566406,
      "logps/rejected": -498.94049072265625,
      "loss": 0.7443,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.4304870665073395,
      "rewards/margins": 0.7002173066139221,
      "rewards/rejected": -1.130704402923584,
      "step": 50
    },
    {
      "epoch": 1.1374407582938388,
      "grad_norm": 97.15451854160752,
      "learning_rate": 2.2891223348923882e-07,
      "logits/chosen": 136.35397338867188,
      "logits/rejected": 140.20547485351562,
      "logps/chosen": -405.622802734375,
      "logps/rejected": -470.76971435546875,
      "loss": 41.4564,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.38085630536079407,
      "rewards/margins": 1.3393514156341553,
      "rewards/rejected": -1.7202078104019165,
      "step": 60
    },
    {
      "epoch": 1.3270142180094786,
      "grad_norm": 109.76902556974036,
      "learning_rate": 1.4754491880085317e-07,
      "logits/chosen": 128.60755920410156,
      "logits/rejected": 130.52413940429688,
      "logps/chosen": -360.59564208984375,
      "logps/rejected": -440.8587951660156,
      "loss": 0.4135,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.327410489320755,
      "rewards/margins": 1.6946580410003662,
      "rewards/rejected": -1.367247462272644,
      "step": 70
    },
    {
      "epoch": 1.5165876777251186,
      "grad_norm": 106.4939956277441,
      "learning_rate": 7.775827023107834e-08,
      "logits/chosen": 115.87138366699219,
      "logits/rejected": 132.57986450195312,
      "logps/chosen": -347.29681396484375,
      "logps/rejected": -440.5848693847656,
      "loss": 24152.7516,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.31770962476730347,
      "rewards/margins": 1.685066819190979,
      "rewards/rejected": -1.3673571348190308,
      "step": 80
    },
    {
      "epoch": 1.7061611374407581,
      "grad_norm": 89.69098645819697,
      "learning_rate": 2.7440387297912122e-08,
      "logits/chosen": 115.1356430053711,
      "logits/rejected": 128.1872100830078,
      "logps/chosen": -380.496337890625,
      "logps/rejected": -470.81854248046875,
      "loss": 0.3571,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.29673463106155396,
      "rewards/margins": 1.9535739421844482,
      "rewards/rejected": -1.65683913230896,
      "step": 90
    },
    {
      "epoch": 1.8957345971563981,
      "grad_norm": 81.51150335383699,
      "learning_rate": 2.27878296044029e-09,
      "logits/chosen": 121.6316909790039,
      "logits/rejected": 121.7571029663086,
      "logps/chosen": -375.3849792480469,
      "logps/rejected": -443.69586181640625,
      "loss": 0.3455,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.3295063376426697,
      "rewards/margins": 1.6760984659194946,
      "rewards/rejected": -1.3465921878814697,
      "step": 100
    },
    {
      "epoch": 1.971563981042654,
      "step": 104,
      "total_flos": 0.0,
      "train_loss": 3302.790195810107,
      "train_runtime": 2170.5017,
      "train_samples_per_second": 6.22,
      "train_steps_per_second": 0.048
    }
  ],
  "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
}