{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 108,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 7.125032920473296,
      "learning_rate": 4.545454545454545e-08,
      "logits/chosen": -0.3348100781440735,
      "logits/rejected": -0.12572002410888672,
      "logps/chosen": -508.3168640136719,
      "logps/rejected": -567.79248046875,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.12412398575055,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": -0.27360814809799194,
      "logits/rejected": -0.16230103373527527,
      "logps/chosen": -385.29803466796875,
      "logps/rejected": -683.7905883789062,
      "loss": 0.6926,
      "rewards/accuracies": 0.5555555820465088,
      "rewards/chosen": -0.00015664617239963263,
      "rewards/margins": 0.002182987052947283,
      "rewards/rejected": -0.0023396331816911697,
      "step": 10
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.588825243356651,
      "learning_rate": 4.894543310469967e-07,
      "logits/chosen": -0.3925573229789734,
      "logits/rejected": -0.16251607239246368,
      "logps/chosen": -445.0148010253906,
      "logps/rejected": -654.8276977539062,
      "loss": 0.6702,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -0.006402991712093353,
      "rewards/margins": 0.051536716520786285,
      "rewards/rejected": -0.05793970823287964,
      "step": 20
    },
    {
      "epoch": 0.28,
      "grad_norm": 6.258154713919184,
      "learning_rate": 4.541409157643027e-07,
      "logits/chosen": -0.3203420341014862,
      "logits/rejected": -0.23488692939281464,
      "logps/chosen": -424.0899963378906,
      "logps/rejected": -731.0706787109375,
      "loss": 0.5838,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.03505988419055939,
      "rewards/margins": 0.31399136781692505,
      "rewards/rejected": -0.34905126690864563,
      "step": 30
    },
    {
      "epoch": 0.37,
      "grad_norm": 6.572704084485822,
      "learning_rate": 3.9760059325148063e-07,
      "logits/chosen": -0.4178282618522644,
      "logits/rejected": -0.29288938641548157,
      "logps/chosen": -426.16741943359375,
      "logps/rejected": -792.8697509765625,
      "loss": 0.5087,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": -0.0784226730465889,
      "rewards/margins": 0.7482975721359253,
      "rewards/rejected": -0.8267202377319336,
      "step": 40
    },
    {
      "epoch": 0.46,
      "grad_norm": 4.8730878124658625,
      "learning_rate": 3.2571251897448763e-07,
      "logits/chosen": -0.3789535164833069,
      "logits/rejected": -0.36692285537719727,
      "logps/chosen": -423.10565185546875,
      "logps/rejected": -962.5338134765625,
      "loss": 0.3946,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": -0.20811887085437775,
      "rewards/margins": 1.7078611850738525,
      "rewards/rejected": -1.9159801006317139,
      "step": 50
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.313374029373249,
      "learning_rate": 2.459517327993746e-07,
      "logits/chosen": -0.41928666830062866,
      "logits/rejected": -0.36349207162857056,
      "logps/chosen": -402.4577331542969,
      "logps/rejected": -1017.8125,
      "loss": 0.3308,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.25981393456459045,
      "rewards/margins": 3.3628947734832764,
      "rewards/rejected": -3.622708797454834,
      "step": 60
    },
    {
      "epoch": 0.65,
      "grad_norm": 4.014655004416041,
      "learning_rate": 1.6661189208729489e-07,
      "logits/chosen": -0.37841862440109253,
      "logits/rejected": -0.39744123816490173,
      "logps/chosen": -384.15191650390625,
      "logps/rejected": -1127.722412109375,
      "loss": 0.2913,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -0.3124866485595703,
      "rewards/margins": 3.733828067779541,
      "rewards/rejected": -4.0463151931762695,
      "step": 70
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.145768949783251,
      "learning_rate": 9.594288359976815e-08,
      "logits/chosen": -0.35634320974349976,
      "logits/rejected": -0.42021504044532776,
      "logps/chosen": -447.4537048339844,
      "logps/rejected": -1247.1588134765625,
      "loss": 0.285,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": -0.3937515318393707,
      "rewards/margins": 5.310124397277832,
      "rewards/rejected": -5.703875541687012,
      "step": 80
    },
    {
      "epoch": 0.83,
      "grad_norm": 5.591694328641131,
      "learning_rate": 4.1292986742682254e-08,
      "logits/chosen": -0.38446053862571716,
      "logits/rejected": -0.381165087223053,
      "logps/chosen": -381.900634765625,
      "logps/rejected": -1111.2266845703125,
      "loss": 0.2676,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.32613903284072876,
      "rewards/margins": 4.088562488555908,
      "rewards/rejected": -4.414700984954834,
      "step": 90
    },
    {
      "epoch": 0.93,
      "grad_norm": 5.087642389120506,
      "learning_rate": 8.344787421847216e-09,
      "logits/chosen": -0.3562104105949402,
      "logits/rejected": -0.39805513620376587,
      "logps/chosen": -437.3785705566406,
      "logps/rejected": -1228.094970703125,
      "loss": 0.2705,
      "rewards/accuracies": 0.875,
      "rewards/chosen": -0.33504000306129456,
      "rewards/margins": 4.651824474334717,
      "rewards/rejected": -4.986865043640137,
      "step": 100
    },
    {
      "epoch": 0.93,
      "eval_logits/chosen": -0.9202428460121155,
      "eval_logits/rejected": -0.40449342131614685,
      "eval_logps/chosen": -108.84640502929688,
      "eval_logps/rejected": -714.0911254882812,
      "eval_loss": 0.44560545682907104,
      "eval_rewards/accuracies": 0.9418103694915771,
      "eval_rewards/chosen": -0.030723365023732185,
      "eval_rewards/margins": 1.1136174201965332,
      "eval_rewards/rejected": -1.1443408727645874,
      "eval_runtime": 191.5153,
      "eval_samples_per_second": 9.634,
      "eval_steps_per_second": 0.303,
      "step": 100
    },
    {
      "epoch": 1.0,
      "step": 108,
      "total_flos": 0.0,
      "train_loss": 0.41410978966289097,
      "train_runtime": 1778.1435,
      "train_samples_per_second": 3.872,
      "train_steps_per_second": 0.061
    }
  ],
  "logging_steps": 10,
  "max_steps": 108,
  "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
}