{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 168,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 17.800336779303485,
      "learning_rate": 2.941176470588235e-08,
      "logits/chosen": -3.0399391651153564,
      "logits/rejected": -2.5624823570251465,
      "logps/chosen": -891.05517578125,
      "logps/rejected": -1084.880126953125,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.06,
      "grad_norm": 18.838509896307443,
      "learning_rate": 2.941176470588235e-07,
      "logits/chosen": -2.7760655879974365,
      "logits/rejected": -2.790731906890869,
      "logps/chosen": -503.9539489746094,
      "logps/rejected": -1057.63232421875,
      "loss": 0.6926,
      "rewards/accuracies": 0.5763888955116272,
      "rewards/chosen": 0.0020897421054542065,
      "rewards/margins": 0.001932685961946845,
      "rewards/rejected": 0.0001570563472341746,
      "step": 10
    },
    {
      "epoch": 0.12,
      "grad_norm": 20.774413071968993,
      "learning_rate": 4.995131923687487e-07,
      "logits/chosen": -2.7422165870666504,
      "logits/rejected": -2.696645498275757,
      "logps/chosen": -548.654296875,
      "logps/rejected": -999.8585815429688,
      "loss": 0.6741,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.03212609142065048,
      "rewards/margins": 0.042353663593530655,
      "rewards/rejected": -0.010227566584944725,
      "step": 20
    },
    {
      "epoch": 0.18,
      "grad_norm": 24.957524647561584,
      "learning_rate": 4.909114739839079e-07,
      "logits/chosen": -2.9761385917663574,
      "logits/rejected": -2.836198329925537,
      "logps/chosen": -549.5018310546875,
      "logps/rejected": -1060.82763671875,
      "loss": 0.6134,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 0.06840862333774567,
      "rewards/margins": 0.18814805150032043,
      "rewards/rejected": -0.11973947286605835,
      "step": 30
    },
    {
      "epoch": 0.24,
      "grad_norm": 29.2493335087329,
      "learning_rate": 4.719192614212969e-07,
      "logits/chosen": -3.023461103439331,
      "logits/rejected": -2.9939627647399902,
      "logps/chosen": -569.5960083007812,
      "logps/rejected": -1105.576416015625,
      "loss": 0.4716,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": -0.18537160754203796,
      "rewards/margins": 0.7153643369674683,
      "rewards/rejected": -0.9007358551025391,
      "step": 40
    },
    {
      "epoch": 0.3,
      "grad_norm": 34.52922011163671,
      "learning_rate": 4.4335568741374695e-07,
      "logits/chosen": -3.0847573280334473,
      "logits/rejected": -3.1392886638641357,
      "logps/chosen": -724.968505859375,
      "logps/rejected": -1334.053955078125,
      "loss": 0.3356,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -1.0850236415863037,
      "rewards/margins": 1.9837610721588135,
      "rewards/rejected": -3.068784713745117,
      "step": 50
    },
    {
      "epoch": 0.36,
      "grad_norm": 74.1970434298791,
      "learning_rate": 4.0645269681018434e-07,
      "logits/chosen": -3.01001238822937,
      "logits/rejected": -3.0855369567871094,
      "logps/chosen": -763.7335815429688,
      "logps/rejected": -1466.1148681640625,
      "loss": 0.2659,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -1.215951681137085,
      "rewards/margins": 3.7804553508758545,
      "rewards/rejected": -4.996407508850098,
      "step": 60
    },
    {
      "epoch": 0.42,
      "grad_norm": 24.166083003223434,
      "learning_rate": 3.6280191288478435e-07,
      "logits/chosen": -2.9843807220458984,
      "logits/rejected": -3.0654196739196777,
      "logps/chosen": -652.9537353515625,
      "logps/rejected": -1663.2164306640625,
      "loss": 0.2156,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": -0.9659550786018372,
      "rewards/margins": 4.433414459228516,
      "rewards/rejected": -5.399369716644287,
      "step": 70
    },
    {
      "epoch": 0.48,
      "grad_norm": 17.793411473003722,
      "learning_rate": 3.142859907420615e-07,
      "logits/chosen": -3.0192360877990723,
      "logits/rejected": -3.050020217895508,
      "logps/chosen": -661.2525634765625,
      "logps/rejected": -1447.4609375,
      "loss": 0.1989,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -0.9658153653144836,
      "rewards/margins": 3.9476540088653564,
      "rewards/rejected": -4.913470268249512,
      "step": 80
    },
    {
      "epoch": 0.54,
      "grad_norm": 55.47746105929639,
      "learning_rate": 2.629974185404951e-07,
      "logits/chosen": -2.978877305984497,
      "logits/rejected": -3.029723644256592,
      "logps/chosen": -613.0646362304688,
      "logps/rejected": -1531.10546875,
      "loss": 0.1844,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -1.270330786705017,
      "rewards/margins": 4.801483154296875,
      "rewards/rejected": -6.071813583374023,
      "step": 90
    },
    {
      "epoch": 0.6,
      "grad_norm": 33.71679055058231,
      "learning_rate": 2.1114826863194878e-07,
      "logits/chosen": -2.983506441116333,
      "logits/rejected": -3.02791690826416,
      "logps/chosen": -676.8287353515625,
      "logps/rejected": -1784.007568359375,
      "loss": 0.1684,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": -1.3490874767303467,
      "rewards/margins": 5.845183849334717,
      "rewards/rejected": -7.194271087646484,
      "step": 100
    },
    {
      "epoch": 0.6,
      "eval_logits/chosen": -3.1281025409698486,
      "eval_logits/rejected": -2.964585304260254,
      "eval_logps/chosen": -755.873291015625,
      "eval_logps/rejected": -1595.0167236328125,
      "eval_loss": 0.3332486152648926,
      "eval_rewards/accuracies": 0.8686440587043762,
      "eval_rewards/chosen": -1.8907400369644165,
      "eval_rewards/margins": 3.9363222122192383,
      "eval_rewards/rejected": -5.827062606811523,
      "eval_runtime": 195.0133,
      "eval_samples_per_second": 9.579,
      "eval_steps_per_second": 0.303,
      "step": 100
    },
    {
      "epoch": 0.65,
      "grad_norm": 36.49070034902309,
      "learning_rate": 1.6097479104361326e-07,
      "logits/chosen": -2.9343550205230713,
      "logits/rejected": -3.0102057456970215,
      "logps/chosen": -727.3587646484375,
      "logps/rejected": -1638.8382568359375,
      "loss": 0.1954,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -1.3601789474487305,
      "rewards/margins": 5.164590835571289,
      "rewards/rejected": -6.524770259857178,
      "step": 110
    },
    {
      "epoch": 0.71,
      "grad_norm": 23.81218846817336,
      "learning_rate": 1.146409641785882e-07,
      "logits/chosen": -2.8368687629699707,
      "logits/rejected": -3.0225136280059814,
      "logps/chosen": -666.7093505859375,
      "logps/rejected": -1738.1021728515625,
      "loss": 0.1519,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": -1.2082051038742065,
      "rewards/margins": 5.624727249145508,
      "rewards/rejected": -6.832932472229004,
      "step": 120
    },
    {
      "epoch": 0.77,
      "grad_norm": 20.21152127941066,
      "learning_rate": 7.414516258630244e-08,
      "logits/chosen": -2.9129068851470947,
      "logits/rejected": -3.0287842750549316,
      "logps/chosen": -708.4816284179688,
      "logps/rejected": -1789.521728515625,
      "loss": 0.1424,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -1.3833436965942383,
      "rewards/margins": 5.361766338348389,
      "rewards/rejected": -6.745110511779785,
      "step": 130
    },
    {
      "epoch": 0.83,
      "grad_norm": 19.81204521774453,
      "learning_rate": 4.1233967214979764e-08,
      "logits/chosen": -2.979017734527588,
      "logits/rejected": -3.0025482177734375,
      "logps/chosen": -664.9674072265625,
      "logps/rejected": -1679.622802734375,
      "loss": 0.1512,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": -1.3218904733657837,
      "rewards/margins": 5.435315132141113,
      "rewards/rejected": -6.757205963134766,
      "step": 140
    },
    {
      "epoch": 0.89,
      "grad_norm": 33.74454128402628,
      "learning_rate": 1.732683550362954e-08,
      "logits/chosen": -2.971137523651123,
      "logits/rejected": -3.0957703590393066,
      "logps/chosen": -662.1959228515625,
      "logps/rejected": -1967.736083984375,
      "loss": 0.1032,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.2782970666885376,
      "rewards/margins": 7.209610939025879,
      "rewards/rejected": -8.487907409667969,
      "step": 150
    },
    {
      "epoch": 0.95,
      "grad_norm": 35.57488951709773,
      "learning_rate": 3.4548802869627804e-09,
      "logits/chosen": -2.9537863731384277,
      "logits/rejected": -3.0035552978515625,
      "logps/chosen": -699.1692504882812,
      "logps/rejected": -1824.681884765625,
      "loss": 0.1233,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": -1.6038345098495483,
      "rewards/margins": 6.514244079589844,
      "rewards/rejected": -8.118078231811523,
      "step": 160
    },
    {
      "epoch": 1.0,
      "step": 168,
      "total_flos": 0.0,
      "train_loss": 0.05682134947606495,
      "train_runtime": 921.1998,
      "train_samples_per_second": 11.663,
      "train_steps_per_second": 0.182
    }
  ],
  "logging_steps": 10,
  "max_steps": 168,
  "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
}