{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.971563981042654,
  "eval_steps": 100,
  "global_step": 104,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.018957345971563982,
      "grad_norm": 133.64062565621384,
      "learning_rate": 4.545454545454545e-08,
      "logits/chosen": 119.0696792602539,
      "logits/rejected": 120.28123474121094,
      "logps/chosen": -394.1268310546875,
      "logps/rejected": -419.3145446777344,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.1895734597156398,
      "grad_norm": 130.60842697521545,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": 133.6595001220703,
      "logits/rejected": 136.7303466796875,
      "logps/chosen": -410.0771484375,
      "logps/rejected": -445.1907653808594,
      "loss": 0.7019,
      "rewards/accuracies": 0.46875,
      "rewards/chosen": 0.020121444016695023,
      "rewards/margins": 0.041466910392045975,
      "rewards/rejected": -0.021345460787415504,
      "step": 10
    },
    {
      "epoch": 0.3791469194312796,
      "grad_norm": 127.29787487076526,
      "learning_rate": 4.885348141000122e-07,
      "logits/chosen": 122.2022476196289,
      "logits/rejected": 128.57586669921875,
      "logps/chosen": -357.1582336425781,
      "logps/rejected": -416.08087158203125,
      "loss": 0.6346,
      "rewards/accuracies": 0.653124988079071,
      "rewards/chosen": 0.25420495867729187,
      "rewards/margins": 0.4108888506889343,
      "rewards/rejected": -0.15668384730815887,
      "step": 20
    },
    {
      "epoch": 0.5687203791469194,
      "grad_norm": 110.05011163607695,
      "learning_rate": 4.5025027361734613e-07,
      "logits/chosen": 121.9586181640625,
      "logits/rejected": 125.2878646850586,
      "logps/chosen": -387.713134765625,
      "logps/rejected": -442.55206298828125,
      "loss": 0.5698,
      "rewards/accuracies": 0.703125,
      "rewards/chosen": -1.2848999500274658,
      "rewards/margins": 0.9555079340934753,
      "rewards/rejected": -2.240407943725586,
      "step": 30
    },
    {
      "epoch": 0.7582938388625592,
      "grad_norm": 111.08969508053838,
      "learning_rate": 3.893311157806091e-07,
      "logits/chosen": 121.52265930175781,
      "logits/rejected": 119.2688980102539,
      "logps/chosen": -402.15716552734375,
      "logps/rejected": -444.649169921875,
      "loss": 0.5496,
      "rewards/accuracies": 0.7593749761581421,
      "rewards/chosen": -2.0494799613952637,
      "rewards/margins": 1.070623517036438,
      "rewards/rejected": -3.120103359222412,
      "step": 40
    },
    {
      "epoch": 0.9478672985781991,
      "grad_norm": 122.82358054602282,
      "learning_rate": 3.126631330646801e-07,
      "logits/chosen": 128.3933868408203,
      "logits/rejected": 133.44308471679688,
      "logps/chosen": -431.0421447753906,
      "logps/rejected": -497.99420166015625,
      "loss": 0.4959,
      "rewards/accuracies": 0.746874988079071,
      "rewards/chosen": -2.2058186531066895,
      "rewards/margins": 1.2984471321105957,
      "rewards/rejected": -3.504265546798706,
      "step": 50
    },
    {
      "epoch": 1.1374407582938388,
      "grad_norm": 65.56687198861316,
      "learning_rate": 2.2891223348923882e-07,
      "logits/chosen": 124.70857238769531,
      "logits/rejected": 126.91219329833984,
      "logps/chosen": -420.981201171875,
      "logps/rejected": -505.5345153808594,
      "loss": 0.307,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -2.4674336910247803,
      "rewards/margins": 2.305318832397461,
      "rewards/rejected": -4.772752285003662,
      "step": 60
    },
    {
      "epoch": 1.3270142180094786,
      "grad_norm": 56.106028687537446,
      "learning_rate": 1.4754491880085317e-07,
      "logits/chosen": 121.775146484375,
      "logits/rejected": 125.95316314697266,
      "logps/chosen": -425.7054138183594,
      "logps/rejected": -518.8656005859375,
      "loss": 0.1907,
      "rewards/accuracies": 0.940625011920929,
      "rewards/chosen": -2.623661518096924,
      "rewards/margins": 2.869920253753662,
      "rewards/rejected": -5.493582248687744,
      "step": 70
    },
    {
      "epoch": 1.5165876777251186,
      "grad_norm": 50.43661058282089,
      "learning_rate": 7.775827023107834e-08,
      "logits/chosen": 114.5962142944336,
      "logits/rejected": 126.1790771484375,
      "logps/chosen": -426.8082580566406,
      "logps/rejected": -527.3065185546875,
      "loss": 0.1761,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": -2.979158401489258,
      "rewards/margins": 3.0644469261169434,
      "rewards/rejected": -6.043605804443359,
      "step": 80
    },
    {
      "epoch": 1.7061611374407581,
      "grad_norm": 45.81843583580765,
      "learning_rate": 2.7440387297912122e-08,
      "logits/chosen": 117.46388244628906,
      "logits/rejected": 123.80489349365234,
      "logps/chosen": -449.65399169921875,
      "logps/rejected": -544.6094970703125,
      "loss": 0.1515,
      "rewards/accuracies": 0.984375,
      "rewards/chosen": -2.8478360176086426,
      "rewards/margins": 3.247156858444214,
      "rewards/rejected": -6.0949931144714355,
      "step": 90
    },
    {
      "epoch": 1.8957345971563981,
      "grad_norm": 42.75820426735574,
      "learning_rate": 2.27878296044029e-09,
      "logits/chosen": 114.7729721069336,
      "logits/rejected": 119.34477233886719,
      "logps/chosen": -437.2296447753906,
      "logps/rejected": -523.9191284179688,
      "loss": 0.1578,
      "rewards/accuracies": 0.9593750238418579,
      "rewards/chosen": -2.8138155937194824,
      "rewards/margins": 3.170293092727661,
      "rewards/rejected": -5.984108924865723,
      "step": 100
    },
    {
      "epoch": 1.8957345971563981,
      "eval_logits/chosen": 91.35408782958984,
      "eval_logits/rejected": 94.07221221923828,
      "eval_logps/chosen": -428.1683349609375,
      "eval_logps/rejected": -515.7637939453125,
      "eval_loss": 0.4643263816833496,
      "eval_rewards/accuracies": 0.75,
      "eval_rewards/chosen": -3.5909416675567627,
      "eval_rewards/margins": 1.7481167316436768,
      "eval_rewards/rejected": -5.339057922363281,
      "eval_runtime": 88.3612,
      "eval_samples_per_second": 8.488,
      "eval_steps_per_second": 0.532,
      "step": 100
    },
    {
      "epoch": 1.971563981042654,
      "step": 104,
      "total_flos": 0.0,
      "train_loss": 0.3883641087091886,
      "train_runtime": 2802.2739,
      "train_samples_per_second": 4.818,
      "train_steps_per_second": 0.037
    }
  ],
  "logging_steps": 10,
  "max_steps": 104,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}