{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.984,
  "eval_steps": 100,
  "global_step": 124,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 3.846153846153847e-07,
      "logits/chosen": 0.33759647607803345,
      "logits/rejected": 0.28577902913093567,
      "logps/chosen": -217.51107788085938,
      "logps/rejected": -154.6298065185547,
      "loss": 0.0102,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.16,
      "learning_rate": 3.846153846153847e-06,
      "logits/chosen": 0.23491337895393372,
      "logits/rejected": 0.185601606965065,
      "logps/chosen": -189.5899200439453,
      "logps/rejected": -162.4447784423828,
      "loss": 0.0106,
      "rewards/accuracies": 0.3333333432674408,
      "rewards/chosen": -0.0008325728122144938,
      "rewards/margins": -0.0011730334954336286,
      "rewards/rejected": 0.0003404606832191348,
      "step": 10
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.951096619903317e-06,
      "logits/chosen": 0.11001434177160263,
      "logits/rejected": 0.06706250458955765,
      "logps/chosen": -170.3766632080078,
      "logps/rejected": -145.70187377929688,
      "loss": 0.0102,
      "rewards/accuracies": 0.42500001192092896,
      "rewards/chosen": 0.0003622249641921371,
      "rewards/margins": 0.0007274674135260284,
      "rewards/rejected": -0.000365242303814739,
      "step": 20
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.716164218065246e-06,
      "logits/chosen": 0.15306313335895538,
      "logits/rejected": 0.14907678961753845,
      "logps/chosen": -184.689453125,
      "logps/rejected": -146.67767333984375,
      "loss": 0.01,
      "rewards/accuracies": 0.4937500059604645,
      "rewards/chosen": 0.0019506055396050215,
      "rewards/margins": 0.001728715025819838,
      "rewards/rejected": 0.0002218906593043357,
      "step": 30
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.3048902348863116e-06,
      "logits/chosen": 0.11253954470157623,
      "logits/rejected": 0.2042957991361618,
      "logps/chosen": -182.73019409179688,
      "logps/rejected": -147.92068481445312,
      "loss": 0.0104,
      "rewards/accuracies": 0.3375000059604645,
      "rewards/chosen": 0.0024931146763265133,
      "rewards/margins": -0.0002410466258879751,
      "rewards/rejected": 0.002734161214902997,
      "step": 40
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.7500000000000005e-06,
      "logits/chosen": 0.26959556341171265,
      "logits/rejected": 0.17199485003948212,
      "logps/chosen": -196.38694763183594,
      "logps/rejected": -161.71762084960938,
      "loss": 0.0102,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": 0.00301692564971745,
      "rewards/margins": 0.0006210061255842447,
      "rewards/rejected": 0.0023959192913025618,
      "step": 50
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.0956464785579125e-06,
      "logits/chosen": 0.19740340113639832,
      "logits/rejected": 0.1762491911649704,
      "logps/chosen": -182.07997131347656,
      "logps/rejected": -150.88436889648438,
      "loss": 0.0105,
      "rewards/accuracies": 0.35624998807907104,
      "rewards/chosen": -0.00115151796489954,
      "rewards/margins": -0.0006008323980495334,
      "rewards/rejected": -0.0005506856250576675,
      "step": 60
    },
    {
      "epoch": 1.12,
      "learning_rate": 2.39389699200963e-06,
      "logits/chosen": 0.11103002727031708,
      "logits/rejected": 0.17598359286785126,
      "logps/chosen": -203.237548828125,
      "logps/rejected": -173.33811950683594,
      "loss": 0.0104,
      "rewards/accuracies": 0.35624998807907104,
      "rewards/chosen": 0.001459635910578072,
      "rewards/margins": -0.0005474772187881172,
      "rewards/rejected": 0.0020071130711585283,
      "step": 70
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.700590188571887e-06,
      "logits/chosen": 0.13080066442489624,
      "logits/rejected": 0.154659703373909,
      "logps/chosen": -193.30853271484375,
      "logps/rejected": -161.02102661132812,
      "loss": 0.0102,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": 0.0017246768111363053,
      "rewards/margins": 0.0007789047667756677,
      "rewards/rejected": 0.0009457715786993504,
      "step": 80
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.0708929268538034e-06,
      "logits/chosen": 0.10300946235656738,
      "logits/rejected": 0.11627539247274399,
      "logps/chosen": -184.31507873535156,
      "logps/rejected": -151.22093200683594,
      "loss": 0.0101,
      "rewards/accuracies": 0.41874998807907104,
      "rewards/chosen": 0.0029378323815762997,
      "rewards/margins": 0.0010877925669774413,
      "rewards/rejected": 0.0018500399310141802,
      "step": 90
    },
    {
      "epoch": 1.6,
      "learning_rate": 5.549106142039018e-07,
      "logits/chosen": 0.17798453569412231,
      "logits/rejected": 0.1302061527967453,
      "logps/chosen": -176.29644775390625,
      "logps/rejected": -149.06594848632812,
      "loss": 0.0105,
      "rewards/accuracies": 0.3499999940395355,
      "rewards/chosen": 0.0017444228287786245,
      "rewards/margins": -0.0010295301908627152,
      "rewards/rejected": 0.0027739531360566616,
      "step": 100
    },
    {
      "epoch": 1.6,
      "eval_logits/chosen": -0.002386125735938549,
      "eval_logits/rejected": 0.09546130150556564,
      "eval_logps/chosen": -306.2984924316406,
      "eval_logps/rejected": -278.5430908203125,
      "eval_loss": 0.010826661251485348,
      "eval_rewards/accuracies": 0.4945000112056732,
      "eval_rewards/chosen": 0.0006116966251283884,
      "eval_rewards/margins": -0.00013227059389464557,
      "eval_rewards/rejected": 0.0007439671317115426,
      "eval_runtime": 433.1922,
      "eval_samples_per_second": 4.617,
      "eval_steps_per_second": 1.154,
      "step": 100
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.937002879188285e-07,
      "logits/chosen": 0.13854524493217468,
      "logits/rejected": 0.09755973517894745,
      "logps/chosen": -194.58871459960938,
      "logps/rejected": -159.81277465820312,
      "loss": 0.0105,
      "rewards/accuracies": 0.33125001192092896,
      "rewards/chosen": 0.0010774679249152541,
      "rewards/margins": -0.0008498074603267014,
      "rewards/rejected": 0.0019272754434496164,
      "step": 110
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.6003680950742728e-08,
      "logits/chosen": 0.18497225642204285,
      "logits/rejected": 0.25526371598243713,
      "logps/chosen": -182.73492431640625,
      "logps/rejected": -146.8502197265625,
      "loss": 0.0103,
      "rewards/accuracies": 0.3812499940395355,
      "rewards/chosen": 0.0014311736449599266,
      "rewards/margins": 0.00013411189138423651,
      "rewards/rejected": 0.001297061680816114,
      "step": 120
    },
    {
      "epoch": 1.98,
      "step": 124,
      "total_flos": 0.0,
      "train_loss": 0.010321829197627882,
      "train_runtime": 1406.1718,
      "train_samples_per_second": 1.422,
      "train_steps_per_second": 0.088
    }
  ],
  "logging_steps": 10,
  "max_steps": 124,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}