{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0695187165775402,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.053475935828877004,
      "grad_norm": 57.38362641680358,
      "learning_rate": 5e-07,
      "logits/chosen": -2.734001874923706,
      "logits/rejected": -2.714400053024292,
      "logps/chosen": -259.45416259765625,
      "logps/rejected": -213.60086059570312,
      "loss": 0.69,
      "rewards/accuracies": 0.3687500059604645,
      "rewards/chosen": 0.022585459053516388,
      "rewards/margins": 0.012519368901848793,
      "rewards/rejected": 0.010066090151667595,
      "step": 5
    },
    {
      "epoch": 0.10695187165775401,
      "grad_norm": 47.481455328039516,
      "learning_rate": 1e-06,
      "logits/chosen": -2.6481270790100098,
      "logits/rejected": -2.652583599090576,
      "logps/chosen": -257.953369140625,
      "logps/rejected": -188.35134887695312,
      "loss": 0.6387,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.550362229347229,
      "rewards/margins": 0.1817861795425415,
      "rewards/rejected": 0.3685761094093323,
      "step": 10
    },
    {
      "epoch": 0.16042780748663102,
      "grad_norm": 57.27168970604102,
      "learning_rate": 9.991477798614637e-07,
      "logits/chosen": -2.517893075942993,
      "logits/rejected": -2.512808322906494,
      "logps/chosen": -238.94729614257812,
      "logps/rejected": -195.82278442382812,
      "loss": 0.6343,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 1.433282732963562,
      "rewards/margins": 0.7576474547386169,
      "rewards/rejected": 0.6756354570388794,
      "step": 15
    },
    {
      "epoch": 0.21390374331550802,
      "grad_norm": 49.36931799258527,
      "learning_rate": 9.965940245625131e-07,
      "logits/chosen": -2.3814330101013184,
      "logits/rejected": -2.3650126457214355,
      "logps/chosen": -240.35299682617188,
      "logps/rejected": -224.2460174560547,
      "loss": 0.6444,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 1.3586658239364624,
      "rewards/margins": 0.8488560914993286,
      "rewards/rejected": 0.5098099112510681,
      "step": 20
    },
    {
      "epoch": 0.26737967914438504,
      "grad_norm": 38.23771695176387,
      "learning_rate": 9.923474395499264e-07,
      "logits/chosen": -2.3290395736694336,
      "logits/rejected": -2.300835371017456,
      "logps/chosen": -240.2759246826172,
      "logps/rejected": -190.5952911376953,
      "loss": 0.5839,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": 1.567830204963684,
      "rewards/margins": 1.0694384574890137,
      "rewards/rejected": 0.4983917772769928,
      "step": 25
    },
    {
      "epoch": 0.32085561497326204,
      "grad_norm": 38.55306482248171,
      "learning_rate": 9.86422500924775e-07,
      "logits/chosen": -2.385282039642334,
      "logits/rejected": -2.380516767501831,
      "logps/chosen": -237.0162353515625,
      "logps/rejected": -203.95974731445312,
      "loss": 0.5888,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 1.3996493816375732,
      "rewards/margins": 0.9638010263442993,
      "rewards/rejected": 0.43584829568862915,
      "step": 30
    },
    {
      "epoch": 0.37433155080213903,
      "grad_norm": 40.05177774617033,
      "learning_rate": 9.788394060951227e-07,
      "logits/chosen": -2.4351038932800293,
      "logits/rejected": -2.440431833267212,
      "logps/chosen": -247.30227661132812,
      "logps/rejected": -201.399658203125,
      "loss": 0.6124,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 1.2308628559112549,
      "rewards/margins": 1.0766284465789795,
      "rewards/rejected": 0.15423443913459778,
      "step": 35
    },
    {
      "epoch": 0.42780748663101603,
      "grad_norm": 38.64888093844348,
      "learning_rate": 9.696240049254742e-07,
      "logits/chosen": -2.4633519649505615,
      "logits/rejected": -2.466259241104126,
      "logps/chosen": -249.7180938720703,
      "logps/rejected": -194.3883056640625,
      "loss": 0.5959,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 1.1658284664154053,
      "rewards/margins": 1.0980390310287476,
      "rewards/rejected": 0.06778934597969055,
      "step": 40
    },
    {
      "epoch": 0.48128342245989303,
      "grad_norm": 41.079430965451316,
      "learning_rate": 9.588077116176756e-07,
      "logits/chosen": -2.4081149101257324,
      "logits/rejected": -2.406311273574829,
      "logps/chosen": -238.9816436767578,
      "logps/rejected": -201.0675048828125,
      "loss": 0.6076,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 1.1393609046936035,
      "rewards/margins": 1.4659656286239624,
      "rewards/rejected": -0.32660484313964844,
      "step": 45
    },
    {
      "epoch": 0.5347593582887701,
      "grad_norm": 46.70159372711324,
      "learning_rate": 9.464273976236516e-07,
      "logits/chosen": -2.4118716716766357,
      "logits/rejected": -2.408468246459961,
      "logps/chosen": -219.6370849609375,
      "logps/rejected": -186.2394561767578,
      "loss": 0.5857,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.8120909929275513,
      "rewards/margins": 0.906074047088623,
      "rewards/rejected": -0.09398309141397476,
      "step": 50
    },
    {
      "epoch": 0.5347593582887701,
      "eval_logits/chosen": -2.4070043563842773,
      "eval_logits/rejected": -2.395042896270752,
      "eval_logps/chosen": -246.7726287841797,
      "eval_logps/rejected": -220.63446044921875,
      "eval_loss": 0.5587548613548279,
      "eval_rewards/accuracies": 0.7202380895614624,
      "eval_rewards/chosen": 0.8626245856285095,
      "eval_rewards/margins": 1.329900860786438,
      "eval_rewards/rejected": -0.46727630496025085,
      "eval_runtime": 180.0372,
      "eval_samples_per_second": 14.775,
      "eval_steps_per_second": 0.233,
      "step": 50
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 38.079371712676014,
      "learning_rate": 9.325252659550308e-07,
      "logits/chosen": -2.387826919555664,
      "logits/rejected": -2.3770289421081543,
      "logps/chosen": -226.2966766357422,
      "logps/rejected": -208.65219116210938,
      "loss": 0.5977,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.6860524415969849,
      "rewards/margins": 0.8900126218795776,
      "rewards/rejected": -0.2039601057767868,
      "step": 55
    },
    {
      "epoch": 0.6417112299465241,
      "grad_norm": 37.176849832612625,
      "learning_rate": 9.171487073181197e-07,
      "logits/chosen": -2.309826374053955,
      "logits/rejected": -2.319997549057007,
      "logps/chosen": -230.22189331054688,
      "logps/rejected": -206.3903350830078,
      "loss": 0.5777,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.7191376686096191,
      "rewards/margins": 1.5079154968261719,
      "rewards/rejected": -0.7887779474258423,
      "step": 60
    },
    {
      "epoch": 0.6951871657754011,
      "grad_norm": 36.39927326888579,
      "learning_rate": 9.003501385646448e-07,
      "logits/chosen": -2.2107081413269043,
      "logits/rejected": -2.21248722076416,
      "logps/chosen": -238.040771484375,
      "logps/rejected": -207.3212127685547,
      "loss": 0.5854,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.5230123996734619,
      "rewards/margins": 1.2387675046920776,
      "rewards/rejected": -0.715755045413971,
      "step": 65
    },
    {
      "epoch": 0.7486631016042781,
      "grad_norm": 34.98626076335809,
      "learning_rate": 8.821868240089676e-07,
      "logits/chosen": -2.1564557552337646,
      "logits/rejected": -2.1244616508483887,
      "logps/chosen": -232.23452758789062,
      "logps/rejected": -215.52719116210938,
      "loss": 0.5572,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": 1.0822027921676636,
      "rewards/margins": 1.5664498805999756,
      "rewards/rejected": -0.48424673080444336,
      "step": 70
    },
    {
      "epoch": 0.8021390374331551,
      "grad_norm": 33.602724500864966,
      "learning_rate": 8.62720680220876e-07,
      "logits/chosen": -2.2200913429260254,
      "logits/rejected": -2.161189079284668,
      "logps/chosen": -232.7351837158203,
      "logps/rejected": -218.43045043945312,
      "loss": 0.5712,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.5824815630912781,
      "rewards/margins": 1.2068629264831543,
      "rewards/rejected": -0.6243813037872314,
      "step": 75
    },
    {
      "epoch": 0.8556149732620321,
      "grad_norm": 32.25811213479902,
      "learning_rate": 8.420180649593929e-07,
      "logits/chosen": -2.223334789276123,
      "logits/rejected": -2.206092119216919,
      "logps/chosen": -237.7117156982422,
      "logps/rejected": -213.7598114013672,
      "loss": 0.5111,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": 0.454035222530365,
      "rewards/margins": 1.7620937824249268,
      "rewards/rejected": -1.3080583810806274,
      "step": 80
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 28.217025801566006,
      "learning_rate": 8.201495509671036e-07,
      "logits/chosen": -2.205242872238159,
      "logits/rejected": -2.2124757766723633,
      "logps/chosen": -250.62673950195312,
      "logps/rejected": -233.39517211914062,
      "loss": 0.5286,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.5412265658378601,
      "rewards/margins": 1.507828712463379,
      "rewards/rejected": -0.9666021466255188,
      "step": 85
    },
    {
      "epoch": 0.9625668449197861,
      "grad_norm": 33.081462718842836,
      "learning_rate": 7.971896853961042e-07,
      "logits/chosen": -2.286748170852661,
      "logits/rejected": -2.2566237449645996,
      "logps/chosen": -235.38076782226562,
      "logps/rejected": -223.08413696289062,
      "loss": 0.5372,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 1.0189321041107178,
      "rewards/margins": 1.5707125663757324,
      "rewards/rejected": -0.5517805814743042,
      "step": 90
    },
    {
      "epoch": 1.0160427807486632,
      "grad_norm": 19.031150325303443,
      "learning_rate": 7.732167356856654e-07,
      "logits/chosen": -2.383920192718506,
      "logits/rejected": -2.3609161376953125,
      "logps/chosen": -249.6156768798828,
      "logps/rejected": -201.63851928710938,
      "loss": 0.3778,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.3106629848480225,
      "rewards/margins": 2.160562038421631,
      "rewards/rejected": -0.8498989939689636,
      "step": 95
    },
    {
      "epoch": 1.0695187165775402,
      "grad_norm": 21.9932361270153,
      "learning_rate": 7.48312422757881e-07,
      "logits/chosen": -2.432481050491333,
      "logits/rejected": -2.416440010070801,
      "logps/chosen": -221.5184326171875,
      "logps/rejected": -238.3723602294922,
      "loss": 0.2515,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 1.44422447681427,
      "rewards/margins": 3.0625545978546143,
      "rewards/rejected": -1.6183300018310547,
      "step": 100
    },
    {
      "epoch": 1.0695187165775402,
      "eval_logits/chosen": -2.3926949501037598,
      "eval_logits/rejected": -2.3801937103271484,
      "eval_logps/chosen": -243.76559448242188,
      "eval_logps/rejected": -223.52978515625,
      "eval_loss": 0.50016188621521,
      "eval_rewards/accuracies": 0.7827380895614624,
      "eval_rewards/chosen": 1.1633288860321045,
      "eval_rewards/margins": 1.9201369285583496,
      "eval_rewards/rejected": -0.7568081617355347,
      "eval_runtime": 179.117,
      "eval_samples_per_second": 14.851,
      "eval_steps_per_second": 0.234,
      "step": 100
    }
  ],
  "logging_steps": 5,
  "max_steps": 279,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1178822762299392.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}