{
  "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": 224.95185141260285,
      "learning_rate": 4.545454545454545e-08,
      "logits/chosen": 117.53560638427734,
      "logits/rejected": 126.8960952758789,
      "logps/chosen": -335.40118408203125,
      "logps/rejected": -439.16552734375,
      "loss": 1.675,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.1895734597156398,
      "grad_norm": 560.8547743946698,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": 135.12060546875,
      "logits/rejected": 138.44573974609375,
      "logps/chosen": -396.12554931640625,
      "logps/rejected": -439.31695556640625,
      "loss": 1.5409,
      "rewards/accuracies": 0.3888888955116272,
      "rewards/chosen": -0.006451513152569532,
      "rewards/margins": -0.006981005426496267,
      "rewards/rejected": 0.0005294906441122293,
      "step": 10
    },
    {
      "epoch": 0.3791469194312796,
      "grad_norm": 352.8580401134438,
      "learning_rate": 4.885348141000122e-07,
      "logits/chosen": 122.6643295288086,
      "logits/rejected": 126.45143127441406,
      "logps/chosen": -362.1632385253906,
      "logps/rejected": -410.5125427246094,
      "loss": 1.3475,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 0.5124378800392151,
      "rewards/margins": 0.038495950400829315,
      "rewards/rejected": 0.47394195199012756,
      "step": 20
    },
    {
      "epoch": 0.5687203791469194,
      "grad_norm": 182.31492393740325,
      "learning_rate": 4.5025027361734613e-07,
      "logits/chosen": 141.10702514648438,
      "logits/rejected": 134.88328552246094,
      "logps/chosen": -382.1487121582031,
      "logps/rejected": -422.1126403808594,
      "loss": 1.1829,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.5158268809318542,
      "rewards/margins": 0.6213487386703491,
      "rewards/rejected": -0.10552184283733368,
      "step": 30
    },
    {
      "epoch": 0.7582938388625592,
      "grad_norm": 224.33186118194806,
      "learning_rate": 3.893311157806091e-07,
      "logits/chosen": 123.0732421875,
      "logits/rejected": 112.23960876464844,
      "logps/chosen": -348.61285400390625,
      "logps/rejected": -374.5006408691406,
      "loss": 1.089,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.2790655493736267,
      "rewards/margins": 1.1103360652923584,
      "rewards/rejected": -0.8312705159187317,
      "step": 40
    },
    {
      "epoch": 0.9478672985781991,
      "grad_norm": 236.4441203380903,
      "learning_rate": 3.126631330646801e-07,
      "logits/chosen": 138.73300170898438,
      "logits/rejected": 142.71951293945312,
      "logps/chosen": -433.286376953125,
      "logps/rejected": -517.555419921875,
      "loss": 1.0317,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": -0.7071272134780884,
      "rewards/margins": 1.3543267250061035,
      "rewards/rejected": -2.0614538192749023,
      "step": 50
    },
    {
      "epoch": 1.1374407582938388,
      "grad_norm": 196.50177344791985,
      "learning_rate": 2.2891223348923882e-07,
      "logits/chosen": 130.94046020507812,
      "logits/rejected": 134.78746032714844,
      "logps/chosen": -420.18060302734375,
      "logps/rejected": -497.28265380859375,
      "loss": 0.8704,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": -1.1087467670440674,
      "rewards/margins": 1.9371082782745361,
      "rewards/rejected": -3.0458550453186035,
      "step": 60
    },
    {
      "epoch": 1.3270142180094786,
      "grad_norm": 153.06309754271552,
      "learning_rate": 1.4754491880085317e-07,
      "logits/chosen": 126.71126556396484,
      "logits/rejected": 128.40968322753906,
      "logps/chosen": -398.8050842285156,
      "logps/rejected": -479.12548828125,
      "loss": 0.7806,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -1.5830620527267456,
      "rewards/margins": 1.6975189447402954,
      "rewards/rejected": -3.280580997467041,
      "step": 70
    },
    {
      "epoch": 1.5165876777251186,
      "grad_norm": 124.94625668106731,
      "learning_rate": 7.775827023107834e-08,
      "logits/chosen": 114.69215393066406,
      "logits/rejected": 131.03684997558594,
      "logps/chosen": -391.1977233886719,
      "logps/rejected": -490.43597412109375,
      "loss": 0.7047,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": -1.8773349523544312,
      "rewards/margins": 1.9825798273086548,
      "rewards/rejected": -3.8599143028259277,
      "step": 80
    },
    {
      "epoch": 1.7061611374407581,
      "grad_norm": 145.93223382181054,
      "learning_rate": 2.7440387297912122e-08,
      "logits/chosen": 113.77287292480469,
      "logits/rejected": 126.99942779541016,
      "logps/chosen": -424.3692932128906,
      "logps/rejected": -520.7833251953125,
      "loss": 0.6871,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": -1.8969141244888306,
      "rewards/margins": 2.258162021636963,
      "rewards/rejected": -4.155076026916504,
      "step": 90
    },
    {
      "epoch": 1.8957345971563981,
      "grad_norm": 127.26184598743976,
      "learning_rate": 2.27878296044029e-09,
      "logits/chosen": 120.37916564941406,
      "logits/rejected": 120.5556640625,
      "logps/chosen": -417.59722900390625,
      "logps/rejected": -491.83563232421875,
      "loss": 0.6935,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -1.781106948852539,
      "rewards/margins": 1.9724727869033813,
      "rewards/rejected": -3.753579616546631,
      "step": 100
    },
    {
      "epoch": 1.971563981042654,
      "step": 104,
      "total_flos": 0.0,
      "train_loss": 0.9854127982488046,
      "train_runtime": 2188.3418,
      "train_samples_per_second": 6.169,
      "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
}