{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9962825278810409,
  "eval_steps": 100,
  "global_step": 134,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 7.393402431483719,
      "learning_rate": 3.571428571428571e-08,
      "logits/chosen": -0.5970903635025024,
      "logits/rejected": -0.02967279776930809,
      "logps/chosen": -254.73361206054688,
      "logps/rejected": -449.335693359375,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.07,
      "grad_norm": 6.97808175581924,
      "learning_rate": 3.5714285714285716e-07,
      "logits/chosen": -0.502315104007721,
      "logits/rejected": -0.24948199093341827,
      "logps/chosen": -339.6366271972656,
      "logps/rejected": -657.9154663085938,
      "loss": 0.6929,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": -0.0001597129157744348,
      "rewards/margins": 0.00033807966974563897,
      "rewards/rejected": -0.0004977926146239042,
      "step": 10
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.671337358423795,
      "learning_rate": 4.969220851487844e-07,
      "logits/chosen": -0.5676344037055969,
      "logits/rejected": -0.3287120759487152,
      "logps/chosen": -378.62664794921875,
      "logps/rejected": -670.591552734375,
      "loss": 0.6827,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": -0.0052469945512712,
      "rewards/margins": 0.02313617616891861,
      "rewards/rejected": -0.028383171185851097,
      "step": 20
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.338334841496394,
      "learning_rate": 4.783863644106502e-07,
      "logits/chosen": -0.5812798738479614,
      "logits/rejected": -0.34431666135787964,
      "logps/chosen": -353.8559875488281,
      "logps/rejected": -717.6322021484375,
      "loss": 0.6376,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": -0.018838122487068176,
      "rewards/margins": 0.11194054782390594,
      "rewards/rejected": -0.13077868521213531,
      "step": 30
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.540070129895064,
      "learning_rate": 4.442864903642427e-07,
      "logits/chosen": -0.4757254719734192,
      "logits/rejected": -0.3771602213382721,
      "logps/chosen": -361.31365966796875,
      "logps/rejected": -770.7361450195312,
      "loss": 0.5666,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -0.04972491413354874,
      "rewards/margins": 0.309120774269104,
      "rewards/rejected": -0.35884565114974976,
      "step": 40
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.028650135342188,
      "learning_rate": 3.9694631307311825e-07,
      "logits/chosen": -0.5247567892074585,
      "logits/rejected": -0.45507222414016724,
      "logps/chosen": -293.3636169433594,
      "logps/rejected": -845.0416870117188,
      "loss": 0.4561,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -0.1027413159608841,
      "rewards/margins": 0.9158787727355957,
      "rewards/rejected": -1.0186201333999634,
      "step": 50
    },
    {
      "epoch": 0.45,
      "grad_norm": 5.965290480598111,
      "learning_rate": 3.39591987386325e-07,
      "logits/chosen": -0.5522093772888184,
      "logits/rejected": -0.4290170669555664,
      "logps/chosen": -363.2244873046875,
      "logps/rejected": -951.5051879882812,
      "loss": 0.3684,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": -0.2073672115802765,
      "rewards/margins": 2.0674309730529785,
      "rewards/rejected": -2.2747981548309326,
      "step": 60
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.3960424532927345,
      "learning_rate": 2.761321158169134e-07,
      "logits/chosen": -0.47125476598739624,
      "logits/rejected": -0.4486091136932373,
      "logps/chosen": -341.56646728515625,
      "logps/rejected": -1014.1658935546875,
      "loss": 0.338,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.219242125749588,
      "rewards/margins": 3.1757972240448,
      "rewards/rejected": -3.3950393199920654,
      "step": 70
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.338829151255127,
      "learning_rate": 2.1089138373994222e-07,
      "logits/chosen": -0.5217522382736206,
      "logits/rejected": -0.5397945642471313,
      "logps/chosen": -321.6473083496094,
      "logps/rejected": -1235.2142333984375,
      "loss": 0.2868,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": -0.31155842542648315,
      "rewards/margins": 4.753512382507324,
      "rewards/rejected": -5.065071105957031,
      "step": 80
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.274164409019951,
      "learning_rate": 1.4831583923104998e-07,
      "logits/chosen": -0.44177961349487305,
      "logits/rejected": -0.528927743434906,
      "logps/chosen": -327.2131042480469,
      "logps/rejected": -1242.676513671875,
      "loss": 0.2479,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": -0.2681874930858612,
      "rewards/margins": 5.103245735168457,
      "rewards/rejected": -5.371432781219482,
      "step": 90
    },
    {
      "epoch": 0.74,
      "grad_norm": 3.847681923589658,
      "learning_rate": 9.266990223754067e-08,
      "logits/chosen": -0.3910934329032898,
      "logits/rejected": -0.5766850709915161,
      "logps/chosen": -417.525390625,
      "logps/rejected": -1386.257080078125,
      "loss": 0.2497,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.5247961282730103,
      "rewards/margins": 5.741795539855957,
      "rewards/rejected": -6.266592979431152,
      "step": 100
    },
    {
      "epoch": 0.74,
      "eval_logits/chosen": -0.7826768159866333,
      "eval_logits/rejected": -0.5636682510375977,
      "eval_logps/chosen": -311.038330078125,
      "eval_logps/rejected": -748.6944580078125,
      "eval_loss": 0.3023545444011688,
      "eval_rewards/accuracies": 0.949999988079071,
      "eval_rewards/chosen": -0.08790449053049088,
      "eval_rewards/margins": 1.8343137502670288,
      "eval_rewards/rejected": -1.9222180843353271,
      "eval_runtime": 15.884,
      "eval_samples_per_second": 9.569,
      "eval_steps_per_second": 0.315,
      "step": 100
    },
    {
      "epoch": 0.82,
      "grad_norm": 4.194490591135143,
      "learning_rate": 4.774575140626316e-08,
      "logits/chosen": -0.42816129326820374,
      "logits/rejected": -0.4562205374240875,
      "logps/chosen": -361.47930908203125,
      "logps/rejected": -1313.421630859375,
      "loss": 0.2458,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.5410887002944946,
      "rewards/margins": 5.888722896575928,
      "rewards/rejected": -6.429811954498291,
      "step": 110
    },
    {
      "epoch": 0.89,
      "grad_norm": 3.9693871459743573,
      "learning_rate": 1.6604893375699592e-08,
      "logits/chosen": -0.4562758803367615,
      "logits/rejected": -0.5703433156013489,
      "logps/chosen": -393.4559631347656,
      "logps/rejected": -1514.4947509765625,
      "loss": 0.2112,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.518619954586029,
      "rewards/margins": 7.754377841949463,
      "rewards/rejected": -8.272997856140137,
      "step": 120
    },
    {
      "epoch": 0.97,
      "grad_norm": 4.192543496278868,
      "learning_rate": 1.3695261579316775e-09,
      "logits/chosen": -0.39360299706459045,
      "logits/rejected": -0.4867175221443176,
      "logps/chosen": -386.438232421875,
      "logps/rejected": -1469.9609375,
      "loss": 0.2181,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.6078141927719116,
      "rewards/margins": 7.127106666564941,
      "rewards/rejected": -7.734920501708984,
      "step": 130
    },
    {
      "epoch": 1.0,
      "step": 134,
      "total_flos": 0.0,
      "train_loss": 0.3946254751575527,
      "train_runtime": 1910.8543,
      "train_samples_per_second": 4.489,
      "train_steps_per_second": 0.07
    }
  ],
  "logging_steps": 10,
  "max_steps": 134,
  "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
}