{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.999259807549963,
  "eval_steps": 500,
  "global_step": 675,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.050333086602516654,
      "grad_norm": 3.643551826477051,
      "learning_rate": 5e-07,
      "logits/chosen": -0.5447720289230347,
      "logits/rejected": -0.4665735065937042,
      "logps/chosen": -70.74549102783203,
      "logps/rejected": -10.341827392578125,
      "loss": 0.6933,
      "rewards/accuracies": 0.4852941334247589,
      "rewards/chosen": 0.0009019060526043177,
      "rewards/margins": 2.6691925086197443e-05,
      "rewards/rejected": 0.0008752142894081771,
      "step": 34
    },
    {
      "epoch": 0.10066617320503331,
      "grad_norm": 8.25416374206543,
      "learning_rate": 1e-06,
      "logits/chosen": -0.5530118346214294,
      "logits/rejected": -0.46221238374710083,
      "logps/chosen": -64.53274536132812,
      "logps/rejected": -8.40461254119873,
      "loss": 0.6902,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.004430013243108988,
      "rewards/margins": 0.006295821163803339,
      "rewards/rejected": -0.0018658083863556385,
      "step": 68
    },
    {
      "epoch": 0.15099925980754997,
      "grad_norm": 5.755356788635254,
      "learning_rate": 9.43986820428336e-07,
      "logits/chosen": -0.531723141670227,
      "logits/rejected": -0.4240272343158722,
      "logps/chosen": -74.01490783691406,
      "logps/rejected": -9.048775672912598,
      "loss": 0.6686,
      "rewards/accuracies": 0.8713235259056091,
      "rewards/chosen": 0.0434609092772007,
      "rewards/margins": 0.05061572045087814,
      "rewards/rejected": -0.007154808845371008,
      "step": 102
    },
    {
      "epoch": 0.20133234641006661,
      "grad_norm": 6.1793928146362305,
      "learning_rate": 8.879736408566722e-07,
      "logits/chosen": -0.5414551496505737,
      "logits/rejected": -0.44891485571861267,
      "logps/chosen": -74.40818786621094,
      "logps/rejected": -11.396381378173828,
      "loss": 0.6245,
      "rewards/accuracies": 0.9779411554336548,
      "rewards/chosen": 0.12970797717571259,
      "rewards/margins": 0.14640042185783386,
      "rewards/rejected": -0.016692442819476128,
      "step": 136
    },
    {
      "epoch": 0.25166543301258326,
      "grad_norm": 4.775036811828613,
      "learning_rate": 8.319604612850081e-07,
      "logits/chosen": -0.5510014891624451,
      "logits/rejected": -0.4638059437274933,
      "logps/chosen": -65.10930633544922,
      "logps/rejected": -9.851871490478516,
      "loss": 0.573,
      "rewards/accuracies": 0.9852941036224365,
      "rewards/chosen": 0.23002856969833374,
      "rewards/margins": 0.26800811290740967,
      "rewards/rejected": -0.037979576736688614,
      "step": 170
    },
    {
      "epoch": 0.30199851961509994,
      "grad_norm": 4.691457271575928,
      "learning_rate": 7.759472817133442e-07,
      "logits/chosen": -0.5684951543807983,
      "logits/rejected": -0.48908984661102295,
      "logps/chosen": -63.056541442871094,
      "logps/rejected": -9.765806198120117,
      "loss": 0.4989,
      "rewards/accuracies": 0.9742646813392639,
      "rewards/chosen": 0.39861220121383667,
      "rewards/margins": 0.48014047741889954,
      "rewards/rejected": -0.08152831345796585,
      "step": 204
    },
    {
      "epoch": 0.35233160621761656,
      "grad_norm": 3.0710372924804688,
      "learning_rate": 7.199341021416804e-07,
      "logits/chosen": -0.5603327751159668,
      "logits/rejected": -0.474366694688797,
      "logps/chosen": -68.22887420654297,
      "logps/rejected": -11.145505905151367,
      "loss": 0.4096,
      "rewards/accuracies": 0.9742646813392639,
      "rewards/chosen": 0.6239223480224609,
      "rewards/margins": 0.7788197994232178,
      "rewards/rejected": -0.15489746630191803,
      "step": 238
    },
    {
      "epoch": 0.40266469282013323,
      "grad_norm": 3.50852108001709,
      "learning_rate": 6.639209225700164e-07,
      "logits/chosen": -0.554121732711792,
      "logits/rejected": -0.47616252303123474,
      "logps/chosen": -59.60889434814453,
      "logps/rejected": -12.72043514251709,
      "loss": 0.3624,
      "rewards/accuracies": 0.9779411554336548,
      "rewards/chosen": 0.769890308380127,
      "rewards/margins": 1.017148494720459,
      "rewards/rejected": -0.24725815653800964,
      "step": 272
    },
    {
      "epoch": 0.4529977794226499,
      "grad_norm": 2.061458110809326,
      "learning_rate": 6.079077429983526e-07,
      "logits/chosen": -0.5355279445648193,
      "logits/rejected": -0.46640804409980774,
      "logps/chosen": -64.37019348144531,
      "logps/rejected": -15.296799659729004,
      "loss": 0.2901,
      "rewards/accuracies": 0.9779411554336548,
      "rewards/chosen": 0.9912306666374207,
      "rewards/margins": 1.3743596076965332,
      "rewards/rejected": -0.3831288516521454,
      "step": 306
    },
    {
      "epoch": 0.5033308660251665,
      "grad_norm": 2.341501474380493,
      "learning_rate": 5.518945634266886e-07,
      "logits/chosen": -0.5389317870140076,
      "logits/rejected": -0.4714396893978119,
      "logps/chosen": -62.30534744262695,
      "logps/rejected": -16.425796508789062,
      "loss": 0.2661,
      "rewards/accuracies": 0.9816176295280457,
      "rewards/chosen": 1.0894008874893188,
      "rewards/margins": 1.5544438362121582,
      "rewards/rejected": -0.46504291892051697,
      "step": 340
    },
    {
      "epoch": 0.5536639526276832,
      "grad_norm": 1.6648075580596924,
      "learning_rate": 4.958813838550247e-07,
      "logits/chosen": -0.5351024270057678,
      "logits/rejected": -0.48344936966896057,
      "logps/chosen": -53.63265609741211,
      "logps/rejected": -17.541196823120117,
      "loss": 0.2411,
      "rewards/accuracies": 0.9705882668495178,
      "rewards/chosen": 1.1532927751541138,
      "rewards/margins": 1.7396329641342163,
      "rewards/rejected": -0.5863401889801025,
      "step": 374
    },
    {
      "epoch": 0.6039970392301999,
      "grad_norm": 2.055748701095581,
      "learning_rate": 4.3986820428336077e-07,
      "logits/chosen": -0.538494884967804,
      "logits/rejected": -0.4838006794452667,
      "logps/chosen": -57.38086700439453,
      "logps/rejected": -18.22422981262207,
      "loss": 0.2084,
      "rewards/accuracies": 0.9705882668495178,
      "rewards/chosen": 1.300463318824768,
      "rewards/margins": 1.98318350315094,
      "rewards/rejected": -0.6827201843261719,
      "step": 408
    },
    {
      "epoch": 0.6543301258327165,
      "grad_norm": 3.2196602821350098,
      "learning_rate": 3.8385502471169683e-07,
      "logits/chosen": -0.5572548508644104,
      "logits/rejected": -0.5117458701133728,
      "logps/chosen": -54.00746154785156,
      "logps/rejected": -17.898473739624023,
      "loss": 0.1869,
      "rewards/accuracies": 0.9889705777168274,
      "rewards/chosen": 1.3422279357910156,
      "rewards/margins": 2.1278655529022217,
      "rewards/rejected": -0.7856374382972717,
      "step": 442
    },
    {
      "epoch": 0.7046632124352331,
      "grad_norm": 1.63054358959198,
      "learning_rate": 3.2784184514003294e-07,
      "logits/chosen": -0.5344655513763428,
      "logits/rejected": -0.47442930936813354,
      "logps/chosen": -56.39248275756836,
      "logps/rejected": -17.505578994750977,
      "loss": 0.1707,
      "rewards/accuracies": 0.9889705777168274,
      "rewards/chosen": 1.3500103950500488,
      "rewards/margins": 2.1952497959136963,
      "rewards/rejected": -0.8452394008636475,
      "step": 476
    },
    {
      "epoch": 0.7549962990377498,
      "grad_norm": 1.9224135875701904,
      "learning_rate": 2.7182866556836905e-07,
      "logits/chosen": -0.5313306450843811,
      "logits/rejected": -0.48571935296058655,
      "logps/chosen": -50.30500411987305,
      "logps/rejected": -18.38837242126465,
      "loss": 0.1665,
      "rewards/accuracies": 0.9816176295280457,
      "rewards/chosen": 1.3861879110336304,
      "rewards/margins": 2.343000888824463,
      "rewards/rejected": -0.9568131566047668,
      "step": 510
    },
    {
      "epoch": 0.8053293856402665,
      "grad_norm": 2.294525146484375,
      "learning_rate": 2.158154859967051e-07,
      "logits/chosen": -0.5495839715003967,
      "logits/rejected": -0.4935465455055237,
      "logps/chosen": -54.87118911743164,
      "logps/rejected": -20.08263397216797,
      "loss": 0.1558,
      "rewards/accuracies": 0.9889705777168274,
      "rewards/chosen": 1.3922532796859741,
      "rewards/margins": 2.4232118129730225,
      "rewards/rejected": -1.0309585332870483,
      "step": 544
    },
    {
      "epoch": 0.8556624722427831,
      "grad_norm": 1.459617257118225,
      "learning_rate": 1.5980230642504117e-07,
      "logits/chosen": -0.5833328366279602,
      "logits/rejected": -0.517787516117096,
      "logps/chosen": -62.456932067871094,
      "logps/rejected": -20.363330841064453,
      "loss": 0.1342,
      "rewards/accuracies": 0.9926470518112183,
      "rewards/chosen": 1.5634468793869019,
      "rewards/margins": 2.619420289993286,
      "rewards/rejected": -1.0559734106063843,
      "step": 578
    },
    {
      "epoch": 0.9059955588452998,
      "grad_norm": 1.3194255828857422,
      "learning_rate": 1.0378912685337726e-07,
      "logits/chosen": -0.5344783663749695,
      "logits/rejected": -0.4842905104160309,
      "logps/chosen": -54.44826126098633,
      "logps/rejected": -20.549943923950195,
      "loss": 0.1334,
      "rewards/accuracies": 0.9852941036224365,
      "rewards/chosen": 1.394426703453064,
      "rewards/margins": 2.494863271713257,
      "rewards/rejected": -1.1004364490509033,
      "step": 612
    },
    {
      "epoch": 0.9563286454478165,
      "grad_norm": 1.6856123208999634,
      "learning_rate": 4.777594728171334e-08,
      "logits/chosen": -0.5274606347084045,
      "logits/rejected": -0.48015105724334717,
      "logps/chosen": -56.23059844970703,
      "logps/rejected": -23.505773544311523,
      "loss": 0.1363,
      "rewards/accuracies": 0.9852941036224365,
      "rewards/chosen": 1.4247695207595825,
      "rewards/margins": 2.5551772117614746,
      "rewards/rejected": -1.130407691001892,
      "step": 646
    },
    {
      "epoch": 0.999259807549963,
      "step": 675,
      "total_flos": 3.8465011975952794e+17,
      "train_loss": 0.33879759046766494,
      "train_runtime": 9841.7629,
      "train_samples_per_second": 0.549,
      "train_steps_per_second": 0.069
    }
  ],
  "logging_steps": 34,
  "max_steps": 675,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.8465011975952794e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}