{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.99581589958159,
  "eval_steps": 500,
  "global_step": 119,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008368200836820083,
      "grad_norm": 13.7675517674306,
      "learning_rate": 4.166666666666666e-08,
      "logits/chosen": -2.6905035972595215,
      "logits/rejected": -2.6002821922302246,
      "logps/chosen": -238.50418090820312,
      "logps/pi_response": -60.972923278808594,
      "logps/ref_response": -60.972923278808594,
      "logps/rejected": -122.8811264038086,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.08368200836820083,
      "grad_norm": 11.556330314168122,
      "learning_rate": 4.1666666666666667e-07,
      "logits/chosen": -2.7245476245880127,
      "logits/rejected": -2.6954076290130615,
      "logps/chosen": -223.26068115234375,
      "logps/pi_response": -68.41806030273438,
      "logps/ref_response": -68.24886322021484,
      "logps/rejected": -108.38255310058594,
      "loss": 0.6909,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.00402547512203455,
      "rewards/margins": 0.00467924727126956,
      "rewards/rejected": -0.0006537723238579929,
      "step": 10
    },
    {
      "epoch": 0.16736401673640167,
      "grad_norm": 8.0051562247015,
      "learning_rate": 4.931352528237397e-07,
      "logits/chosen": -2.7007386684417725,
      "logits/rejected": -2.6637234687805176,
      "logps/chosen": -239.10507202148438,
      "logps/pi_response": -75.34381866455078,
      "logps/ref_response": -65.86756896972656,
      "logps/rejected": -120.88731384277344,
      "loss": 0.6538,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 0.02172889932990074,
      "rewards/margins": 0.08931762725114822,
      "rewards/rejected": -0.06758873164653778,
      "step": 20
    },
    {
      "epoch": 0.2510460251046025,
      "grad_norm": 6.2838771200093735,
      "learning_rate": 4.658920803689553e-07,
      "logits/chosen": -2.5834615230560303,
      "logits/rejected": -2.541572093963623,
      "logps/chosen": -234.6986541748047,
      "logps/pi_response": -112.6206283569336,
      "logps/ref_response": -70.35834503173828,
      "logps/rejected": -146.62071228027344,
      "loss": 0.6128,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.1661997139453888,
      "rewards/margins": 0.1909169852733612,
      "rewards/rejected": -0.35711669921875,
      "step": 30
    },
    {
      "epoch": 0.33472803347280333,
      "grad_norm": 8.1669760332498,
      "learning_rate": 4.201712553872657e-07,
      "logits/chosen": -2.544977903366089,
      "logits/rejected": -2.512420415878296,
      "logps/chosen": -266.4893798828125,
      "logps/pi_response": -148.94216918945312,
      "logps/ref_response": -70.82736206054688,
      "logps/rejected": -183.97122192382812,
      "loss": 0.5908,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.31368952989578247,
      "rewards/margins": 0.3473220467567444,
      "rewards/rejected": -0.6610115170478821,
      "step": 40
    },
    {
      "epoch": 0.41841004184100417,
      "grad_norm": 10.22318295333062,
      "learning_rate": 3.598859066780754e-07,
      "logits/chosen": -2.5687315464019775,
      "logits/rejected": -2.520003318786621,
      "logps/chosen": -306.9870300292969,
      "logps/pi_response": -178.8148956298828,
      "logps/ref_response": -73.63240814208984,
      "logps/rejected": -191.5988006591797,
      "loss": 0.5573,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.37883269786834717,
      "rewards/margins": 0.5206230282783508,
      "rewards/rejected": -0.899455726146698,
      "step": 50
    },
    {
      "epoch": 0.502092050209205,
      "grad_norm": 12.014191391493059,
      "learning_rate": 2.9019570347986706e-07,
      "logits/chosen": -2.542023181915283,
      "logits/rejected": -2.512960195541382,
      "logps/chosen": -290.0708923339844,
      "logps/pi_response": -188.24777221679688,
      "logps/ref_response": -76.79859924316406,
      "logps/rejected": -204.185791015625,
      "loss": 0.5334,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -0.3407396078109741,
      "rewards/margins": 0.7002220153808594,
      "rewards/rejected": -1.0409616231918335,
      "step": 60
    },
    {
      "epoch": 0.5857740585774058,
      "grad_norm": 11.306193009663613,
      "learning_rate": 2.1706525253979534e-07,
      "logits/chosen": -2.555372476577759,
      "logits/rejected": -2.517289400100708,
      "logps/chosen": -280.76043701171875,
      "logps/pi_response": -182.2036895751953,
      "logps/ref_response": -66.60467529296875,
      "logps/rejected": -207.8173828125,
      "loss": 0.549,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": -0.4452129900455475,
      "rewards/margins": 0.5477712750434875,
      "rewards/rejected": -0.9929842948913574,
      "step": 70
    },
    {
      "epoch": 0.6694560669456067,
      "grad_norm": 17.500798301996966,
      "learning_rate": 1.4675360263490295e-07,
      "logits/chosen": -2.516525983810425,
      "logits/rejected": -2.505002021789551,
      "logps/chosen": -231.9325714111328,
      "logps/pi_response": -168.99024963378906,
      "logps/ref_response": -66.72526550292969,
      "logps/rejected": -189.8736114501953,
      "loss": 0.5258,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.42112892866134644,
      "rewards/margins": 0.49120765924453735,
      "rewards/rejected": -0.912336528301239,
      "step": 80
    },
    {
      "epoch": 0.7531380753138075,
      "grad_norm": 16.641411155096407,
      "learning_rate": 8.527854855097224e-08,
      "logits/chosen": -2.596357822418213,
      "logits/rejected": -2.5533790588378906,
      "logps/chosen": -280.3979187011719,
      "logps/pi_response": -187.88308715820312,
      "logps/ref_response": -65.7259750366211,
      "logps/rejected": -200.82089233398438,
      "loss": 0.5428,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": -0.4766550064086914,
      "rewards/margins": 0.6347955465316772,
      "rewards/rejected": -1.1114505529403687,
      "step": 90
    },
    {
      "epoch": 0.8368200836820083,
      "grad_norm": 12.985065267871795,
      "learning_rate": 3.790158337517127e-08,
      "logits/chosen": -2.542086124420166,
      "logits/rejected": -2.512556552886963,
      "logps/chosen": -261.39459228515625,
      "logps/pi_response": -190.4388427734375,
      "logps/ref_response": -67.59146881103516,
      "logps/rejected": -229.26296997070312,
      "loss": 0.5035,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.5009726285934448,
      "rewards/margins": 0.5954475998878479,
      "rewards/rejected": -1.096420168876648,
      "step": 100
    },
    {
      "epoch": 0.9205020920502092,
      "grad_norm": 20.329401448237313,
      "learning_rate": 8.677580722139671e-09,
      "logits/chosen": -2.6092028617858887,
      "logits/rejected": -2.5456957817077637,
      "logps/chosen": -294.7940979003906,
      "logps/pi_response": -201.2043914794922,
      "logps/ref_response": -79.07398223876953,
      "logps/rejected": -223.47451782226562,
      "loss": 0.5098,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.3964020609855652,
      "rewards/margins": 0.7144461870193481,
      "rewards/rejected": -1.110848307609558,
      "step": 110
    },
    {
      "epoch": 0.99581589958159,
      "step": 119,
      "total_flos": 0.0,
      "train_loss": 0.5697346855612362,
      "train_runtime": 2586.0024,
      "train_samples_per_second": 5.91,
      "train_steps_per_second": 0.046
    }
  ],
  "logging_steps": 10,
  "max_steps": 119,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}