{
  "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": 7707.786720275403,
      "learning_rate": 4.545454545454545e-08,
      "logits/chosen": 117.53560638427734,
      "logits/rejected": 126.8960952758789,
      "logps/chosen": -335.40118408203125,
      "logps/rejected": -439.16552734375,
      "loss": 1.0441,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.1895734597156398,
      "grad_norm": 5401.612875785957,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": 135.22718811035156,
      "logits/rejected": 138.51141357421875,
      "logps/chosen": -397.9040222167969,
      "logps/rejected": -441.5177001953125,
      "loss": 4.4651,
      "rewards/accuracies": 0.4583333432674408,
      "rewards/chosen": -0.09537385404109955,
      "rewards/margins": 0.01413203775882721,
      "rewards/rejected": -0.10950590670108795,
      "step": 10
    },
    {
      "epoch": 0.3791469194312796,
      "grad_norm": 3264.226616028638,
      "learning_rate": 4.885348141000122e-07,
      "logits/chosen": 125.89383697509766,
      "logits/rejected": 129.39395141601562,
      "logps/chosen": -378.54852294921875,
      "logps/rejected": -427.40765380859375,
      "loss": 4.7546,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": -0.30682721734046936,
      "rewards/margins": 0.06398675590753555,
      "rewards/rejected": -0.3708139657974243,
      "step": 20
    },
    {
      "epoch": 0.5687203791469194,
      "grad_norm": 4931.31560217951,
      "learning_rate": 4.5025027361734613e-07,
      "logits/chosen": 149.55715942382812,
      "logits/rejected": 142.95787048339844,
      "logps/chosen": -387.85272216796875,
      "logps/rejected": -423.25390625,
      "loss": 4.6409,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": 0.23062637448310852,
      "rewards/margins": 0.39320963621139526,
      "rewards/rejected": -0.16258326172828674,
      "step": 30
    },
    {
      "epoch": 0.7582938388625592,
      "grad_norm": 3460.158543068679,
      "learning_rate": 3.893311157806091e-07,
      "logits/chosen": 132.449462890625,
      "logits/rejected": 120.46671295166016,
      "logps/chosen": -360.9330139160156,
      "logps/rejected": -373.2904968261719,
      "loss": 4.3411,
      "rewards/accuracies": 0.606249988079071,
      "rewards/chosen": -0.33694085478782654,
      "rewards/margins": 0.4338204264640808,
      "rewards/rejected": -0.7707613110542297,
      "step": 40
    },
    {
      "epoch": 0.9478672985781991,
      "grad_norm": 3177.2465625206323,
      "learning_rate": 3.126631330646801e-07,
      "logits/chosen": 147.91026306152344,
      "logits/rejected": 152.0844268798828,
      "logps/chosen": -425.0814514160156,
      "logps/rejected": -489.4386291503906,
      "loss": 3.941,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.2968811094760895,
      "rewards/margins": 0.3587321639060974,
      "rewards/rejected": -0.6556132435798645,
      "step": 50
    },
    {
      "epoch": 1.1374407582938388,
      "grad_norm": 2157.319676771431,
      "learning_rate": 2.2891223348923882e-07,
      "logits/chosen": 137.85789489746094,
      "logits/rejected": 141.48818969726562,
      "logps/chosen": -409.8984375,
      "logps/rejected": -465.19024658203125,
      "loss": 2.3787,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.5946376919746399,
      "rewards/margins": 0.846598744392395,
      "rewards/rejected": -1.4412362575531006,
      "step": 60
    },
    {
      "epoch": 1.3270142180094786,
      "grad_norm": 2309.104547581061,
      "learning_rate": 1.4754491880085317e-07,
      "logits/chosen": 131.42678833007812,
      "logits/rejected": 133.21844482421875,
      "logps/chosen": -372.8443298339844,
      "logps/rejected": -443.36358642578125,
      "loss": 1.8286,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.28502368927001953,
      "rewards/margins": 1.2074631452560425,
      "rewards/rejected": -1.492486834526062,
      "step": 70
    },
    {
      "epoch": 1.5165876777251186,
      "grad_norm": 1891.7921596452784,
      "learning_rate": 7.775827023107834e-08,
      "logits/chosen": 119.9256820678711,
      "logits/rejected": 136.48963928222656,
      "logps/chosen": -362.90484619140625,
      "logps/rejected": -442.4962463378906,
      "loss": 1.3789,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.46269264817237854,
      "rewards/margins": 1.0002342462539673,
      "rewards/rejected": -1.4629267454147339,
      "step": 80
    },
    {
      "epoch": 1.7061611374407581,
      "grad_norm": 1690.6090366045441,
      "learning_rate": 2.7440387297912122e-08,
      "logits/chosen": 121.07257080078125,
      "logits/rejected": 134.4269561767578,
      "logps/chosen": -393.3634033203125,
      "logps/rejected": -468.45050048828125,
      "loss": 1.3089,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.346617192029953,
      "rewards/margins": 1.1918184757232666,
      "rewards/rejected": -1.538435697555542,
      "step": 90
    },
    {
      "epoch": 1.8957345971563981,
      "grad_norm": 1998.5413209764397,
      "learning_rate": 2.27878296044029e-09,
      "logits/chosen": 128.20663452148438,
      "logits/rejected": 128.03355407714844,
      "logps/chosen": -385.4119873046875,
      "logps/rejected": -443.74346923828125,
      "loss": 1.0587,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": -0.17184433341026306,
      "rewards/margins": 1.1771280765533447,
      "rewards/rejected": -1.3489725589752197,
      "step": 100
    },
    {
      "epoch": 1.971563981042654,
      "step": 104,
      "total_flos": 0.0,
      "train_loss": 2.9736096469255595,
      "train_runtime": 2166.7366,
      "train_samples_per_second": 6.231,
      "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
}