{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.984,
  "eval_steps": 100,
  "global_step": 124,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 3.846153846153847e-07,
      "logits/chosen": 0.48606717586517334,
      "logits/rejected": 0.39871031045913696,
      "logps/chosen": -212.65087890625,
      "logps/rejected": -202.6477508544922,
      "loss": 0.0102,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.16,
      "learning_rate": 3.846153846153847e-06,
      "logits/chosen": 0.15301527082920074,
      "logits/rejected": 0.20366798341274261,
      "logps/chosen": -161.48782348632812,
      "logps/rejected": -135.7443389892578,
      "loss": 0.0105,
      "rewards/accuracies": 0.3194444477558136,
      "rewards/chosen": -0.002633917611092329,
      "rewards/margins": -0.0009920704178512096,
      "rewards/rejected": -0.0016418471932411194,
      "step": 10
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.951096619903317e-06,
      "logits/chosen": 0.09449799358844757,
      "logits/rejected": 0.0444490909576416,
      "logps/chosen": -182.85354614257812,
      "logps/rejected": -149.98532104492188,
      "loss": 0.0104,
      "rewards/accuracies": 0.40625,
      "rewards/chosen": -0.0008511155610904098,
      "rewards/margins": -0.0002212001709267497,
      "rewards/rejected": -0.000629915448371321,
      "step": 20
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.716164218065246e-06,
      "logits/chosen": 0.19659022986888885,
      "logits/rejected": 0.24700064957141876,
      "logps/chosen": -175.32278442382812,
      "logps/rejected": -143.53839111328125,
      "loss": 0.0104,
      "rewards/accuracies": 0.38749998807907104,
      "rewards/chosen": -0.0011144612217321992,
      "rewards/margins": -0.000345538865076378,
      "rewards/rejected": -0.000768922152929008,
      "step": 30
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.3048902348863116e-06,
      "logits/chosen": 0.14261968433856964,
      "logits/rejected": 0.19110862910747528,
      "logps/chosen": -181.75755310058594,
      "logps/rejected": -164.41835021972656,
      "loss": 0.0103,
      "rewards/accuracies": 0.39375001192092896,
      "rewards/chosen": -0.0007939288625493646,
      "rewards/margins": 0.00022867057123221457,
      "rewards/rejected": -0.0010225994046777487,
      "step": 40
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.7500000000000005e-06,
      "logits/chosen": 0.07993963360786438,
      "logits/rejected": 0.1553264558315277,
      "logps/chosen": -170.30532836914062,
      "logps/rejected": -144.61050415039062,
      "loss": 0.0101,
      "rewards/accuracies": 0.42500001192092896,
      "rewards/chosen": 0.00015746071585454047,
      "rewards/margins": 0.0011130261700600386,
      "rewards/rejected": -0.0009555654833093286,
      "step": 50
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.0956464785579125e-06,
      "logits/chosen": 0.18755468726158142,
      "logits/rejected": 0.17403154075145721,
      "logps/chosen": -173.29977416992188,
      "logps/rejected": -144.7869873046875,
      "loss": 0.0105,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": -0.0005446333670988679,
      "rewards/margins": -0.000993421534076333,
      "rewards/rejected": 0.00044878822518512607,
      "step": 60
    },
    {
      "epoch": 1.12,
      "learning_rate": 2.39389699200963e-06,
      "logits/chosen": 0.2639772593975067,
      "logits/rejected": 0.2394874542951584,
      "logps/chosen": -188.4938201904297,
      "logps/rejected": -160.391357421875,
      "loss": 0.0102,
      "rewards/accuracies": 0.3812499940395355,
      "rewards/chosen": -0.00019804532348643988,
      "rewards/margins": 0.000605274923145771,
      "rewards/rejected": -0.0008033200865611434,
      "step": 70
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.700590188571887e-06,
      "logits/chosen": 0.1733393371105194,
      "logits/rejected": 0.2405407875776291,
      "logps/chosen": -166.1864471435547,
      "logps/rejected": -136.4402618408203,
      "loss": 0.0102,
      "rewards/accuracies": 0.41874998807907104,
      "rewards/chosen": 0.0004581899265758693,
      "rewards/margins": 0.0007379798917099833,
      "rewards/rejected": -0.0002797898487187922,
      "step": 80
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.0708929268538034e-06,
      "logits/chosen": 0.16160213947296143,
      "logits/rejected": 0.13504673540592194,
      "logps/chosen": -175.88116455078125,
      "logps/rejected": -145.18658447265625,
      "loss": 0.01,
      "rewards/accuracies": 0.4312500059604645,
      "rewards/chosen": 0.0007930597057566047,
      "rewards/margins": 0.001615246757864952,
      "rewards/rejected": -0.0008221869356930256,
      "step": 90
    },
    {
      "epoch": 1.6,
      "learning_rate": 5.549106142039018e-07,
      "logits/chosen": 0.18513503670692444,
      "logits/rejected": 0.14850696921348572,
      "logps/chosen": -177.84588623046875,
      "logps/rejected": -152.69290161132812,
      "loss": 0.0103,
      "rewards/accuracies": 0.4437499940395355,
      "rewards/chosen": 0.0004619792161975056,
      "rewards/margins": 0.00023423954553436488,
      "rewards/rejected": 0.00022773972887080163,
      "step": 100
    },
    {
      "epoch": 1.6,
      "eval_logits/chosen": -0.019983915612101555,
      "eval_logits/rejected": 0.0771104097366333,
      "eval_logps/chosen": -306.49530029296875,
      "eval_logps/rejected": -278.806396484375,
      "eval_loss": 0.010805144906044006,
      "eval_rewards/accuracies": 0.5009999871253967,
      "eval_rewards/chosen": -0.0014956285012885928,
      "eval_rewards/margins": -9.011628571897745e-05,
      "eval_rewards/rejected": -0.0014055122155696154,
      "eval_runtime": 470.6682,
      "eval_samples_per_second": 4.249,
      "eval_steps_per_second": 1.062,
      "step": 100
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.937002879188285e-07,
      "logits/chosen": 0.18757307529449463,
      "logits/rejected": 0.19283342361450195,
      "logps/chosen": -186.30393981933594,
      "logps/rejected": -159.83816528320312,
      "loss": 0.0104,
      "rewards/accuracies": 0.375,
      "rewards/chosen": 0.0009433077648282051,
      "rewards/margins": -0.0004140757955610752,
      "rewards/rejected": 0.0013573834439739585,
      "step": 110
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.6003680950742728e-08,
      "logits/chosen": 0.07835674285888672,
      "logits/rejected": 0.13737662136554718,
      "logps/chosen": -174.80299377441406,
      "logps/rejected": -154.0167694091797,
      "loss": 0.0104,
      "rewards/accuracies": 0.35624998807907104,
      "rewards/chosen": -5.471259282785468e-05,
      "rewards/margins": -0.0002534462546464056,
      "rewards/rejected": 0.0001987336145248264,
      "step": 120
    },
    {
      "epoch": 1.98,
      "step": 124,
      "total_flos": 0.0,
      "train_loss": 0.010314414998696696,
      "train_runtime": 1522.7247,
      "train_samples_per_second": 1.313,
      "train_steps_per_second": 0.081
    }
  ],
  "logging_steps": 10,
  "max_steps": 124,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}