{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 162,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 4.59239378879475,
      "learning_rate": 2.9411764705882356e-07,
      "logits/chosen": 0.35779184103012085,
      "logits/rejected": 0.16919144988059998,
      "logps/chosen": -358.5450439453125,
      "logps/rejected": -376.22247314453125,
      "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": 4.3884364512467755,
      "learning_rate": 2.9411764705882355e-06,
      "logits/chosen": 0.32741284370422363,
      "logits/rejected": 0.21217964589595795,
      "logps/chosen": -261.8883361816406,
      "logps/rejected": -429.5544738769531,
      "loss": 0.6921,
      "rewards/accuracies": 0.4305555522441864,
      "rewards/chosen": 0.0022237710654735565,
      "rewards/margins": 0.0010415142169222236,
      "rewards/rejected": 0.0011822569649666548,
      "step": 10
    },
    {
      "epoch": 0.12,
      "grad_norm": 4.31235489914423,
      "learning_rate": 4.994720857837211e-06,
      "logits/chosen": 0.2313280999660492,
      "logits/rejected": 0.20647689700126648,
      "logps/chosen": -234.76797485351562,
      "logps/rejected": -427.8453674316406,
      "loss": 0.6608,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 0.06342905759811401,
      "rewards/margins": 0.07273004949092865,
      "rewards/rejected": -0.009300985373556614,
      "step": 20
    },
    {
      "epoch": 0.19,
      "grad_norm": 5.67107114042349,
      "learning_rate": 4.901488388458247e-06,
      "logits/chosen": -0.09879900515079498,
      "logits/rejected": -0.14984218776226044,
      "logps/chosen": -252.62423706054688,
      "logps/rejected": -445.3297424316406,
      "loss": 0.5576,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 0.11789284646511078,
      "rewards/margins": 0.37430933117866516,
      "rewards/rejected": -0.2564164102077484,
      "step": 30
    },
    {
      "epoch": 0.25,
      "grad_norm": 26.138827229575735,
      "learning_rate": 4.6959649910976165e-06,
      "logits/chosen": -0.6740697622299194,
      "logits/rejected": -0.7371362447738647,
      "logps/chosen": -254.4224090576172,
      "logps/rejected": -513.7642822265625,
      "loss": 0.4304,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": -0.1447555273771286,
      "rewards/margins": 0.9142681360244751,
      "rewards/rejected": -1.0590236186981201,
      "step": 40
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.9039717916152203,
      "learning_rate": 4.387760711393052e-06,
      "logits/chosen": -0.8242947459220886,
      "logits/rejected": -0.9119745492935181,
      "logps/chosen": -258.3045959472656,
      "logps/rejected": -554.4381103515625,
      "loss": 0.3277,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": -0.036211688071489334,
      "rewards/margins": 1.6457420587539673,
      "rewards/rejected": -1.6819536685943604,
      "step": 50
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.8055402029988485,
      "learning_rate": 3.991286838919086e-06,
      "logits/chosen": -0.7727451920509338,
      "logits/rejected": -1.0549150705337524,
      "logps/chosen": -215.2255859375,
      "logps/rejected": -712.447998046875,
      "loss": 0.2598,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 0.17725825309753418,
      "rewards/margins": 2.7165703773498535,
      "rewards/rejected": -2.5393126010894775,
      "step": 60
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.719156626451469,
      "learning_rate": 3.5250820513035403e-06,
      "logits/chosen": -0.8838419914245605,
      "logits/rejected": -1.0871695280075073,
      "logps/chosen": -227.4645538330078,
      "logps/rejected": -708.2120361328125,
      "loss": 0.221,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 0.17518171668052673,
      "rewards/margins": 3.272481918334961,
      "rewards/rejected": -3.0973002910614014,
      "step": 70
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.227683945021863,
      "learning_rate": 3.0109455662659126e-06,
      "logits/chosen": -0.6757279634475708,
      "logits/rejected": -1.0420820713043213,
      "logps/chosen": -197.2065887451172,
      "logps/rejected": -840.8824462890625,
      "loss": 0.1685,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 0.3136652112007141,
      "rewards/margins": 4.1078901290893555,
      "rewards/rejected": -3.794224977493286,
      "step": 80
    },
    {
      "epoch": 0.56,
      "grad_norm": 2.4140551247764463,
      "learning_rate": 2.4729178344249007e-06,
      "logits/chosen": -0.8329472541809082,
      "logits/rejected": -0.9393612146377563,
      "logps/chosen": -242.90579223632812,
      "logps/rejected": -966.45849609375,
      "loss": 0.1708,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.4426870346069336,
      "rewards/margins": 6.061380863189697,
      "rewards/rejected": -5.6186933517456055,
      "step": 90
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.6781689743422916,
      "learning_rate": 1.936156434546515e-06,
      "logits/chosen": -0.7473787069320679,
      "logits/rejected": -0.9163832664489746,
      "logps/chosen": -218.4031524658203,
      "logps/rejected": -1023.0726318359375,
      "loss": 0.139,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.4111669063568115,
      "rewards/margins": 6.120899677276611,
      "rewards/rejected": -5.709733009338379,
      "step": 100
    },
    {
      "epoch": 0.62,
      "eval_logits/chosen": -1.1380366086959839,
      "eval_logits/rejected": -0.8936022520065308,
      "eval_logps/chosen": -345.177001953125,
      "eval_logps/rejected": -785.2988891601562,
      "eval_loss": 0.3923305869102478,
      "eval_rewards/accuracies": 0.7421875,
      "eval_rewards/chosen": -1.1034249067306519,
      "eval_rewards/margins": 2.3324625492095947,
      "eval_rewards/rejected": -3.435887336730957,
      "eval_runtime": 58.9627,
      "eval_samples_per_second": 8.565,
      "eval_steps_per_second": 0.271,
      "step": 100
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.6174426297852658,
      "learning_rate": 1.4257597331216211e-06,
      "logits/chosen": -0.6151062250137329,
      "logits/rejected": -0.7789457440376282,
      "logps/chosen": -197.7133026123047,
      "logps/rejected": -1012.9052734375,
      "loss": 0.1543,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 0.32270702719688416,
      "rewards/margins": 6.660803318023682,
      "rewards/rejected": -6.3380961418151855,
      "step": 110
    },
    {
      "epoch": 0.74,
      "grad_norm": 3.0637276007560272,
      "learning_rate": 9.655933126436565e-07,
      "logits/chosen": -0.5019649267196655,
      "logits/rejected": -0.7735512852668762,
      "logps/chosen": -196.74630737304688,
      "logps/rejected": -1048.3336181640625,
      "loss": 0.1251,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": 0.29014357924461365,
      "rewards/margins": 6.463759422302246,
      "rewards/rejected": -6.1736159324646,
      "step": 120
    },
    {
      "epoch": 0.8,
      "grad_norm": 3.0128582000127535,
      "learning_rate": 5.771740434959278e-07,
      "logits/chosen": -0.514670729637146,
      "logits/rejected": -0.7349029183387756,
      "logps/chosen": -192.01687622070312,
      "logps/rejected": -1127.93798828125,
      "loss": 0.1113,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/chosen": 0.35582536458969116,
      "rewards/margins": 7.327975273132324,
      "rewards/rejected": -6.972149848937988,
      "step": 130
    },
    {
      "epoch": 0.86,
      "grad_norm": 3.721314851990241,
      "learning_rate": 2.786639790067719e-07,
      "logits/chosen": -0.5744115114212036,
      "logits/rejected": -0.6790918111801147,
      "logps/chosen": -209.8837127685547,
      "logps/rejected": -1139.7821044921875,
      "loss": 0.1039,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": 0.28541240096092224,
      "rewards/margins": 7.819287300109863,
      "rewards/rejected": -7.53387451171875,
      "step": 140
    },
    {
      "epoch": 0.93,
      "grad_norm": 6.57762550514718,
      "learning_rate": 8.402111802159413e-08,
      "logits/chosen": -0.6430222988128662,
      "logits/rejected": -0.657291054725647,
      "logps/chosen": -197.78465270996094,
      "logps/rejected": -932.9392700195312,
      "loss": 0.111,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.311057984828949,
      "rewards/margins": 5.728517055511475,
      "rewards/rejected": -5.417458534240723,
      "step": 150
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.554107935708671,
      "learning_rate": 2.34674439005822e-09,
      "logits/chosen": -0.689643919467926,
      "logits/rejected": -0.7232618927955627,
      "logps/chosen": -186.9496307373047,
      "logps/rejected": -1079.4962158203125,
      "loss": 0.1113,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 0.31346824765205383,
      "rewards/margins": 7.14910364151001,
      "rewards/rejected": -6.835635185241699,
      "step": 160
    },
    {
      "epoch": 1.0,
      "step": 162,
      "total_flos": 0.0,
      "train_loss": 0.26888106432225967,
      "train_runtime": 2670.405,
      "train_samples_per_second": 3.881,
      "train_steps_per_second": 0.061
    }
  ],
  "logging_steps": 10,
  "max_steps": 162,
  "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
}