{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9968652037617555,
  "eval_steps": 500,
  "global_step": 159,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 3.125e-08,
      "logits/chosen": -2.5757241249084473,
      "logits/rejected": -2.6334033012390137,
      "logps/chosen": -158.64126586914062,
      "logps/rejected": -129.17214965820312,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.06,
      "learning_rate": 3.1249999999999997e-07,
      "logits/chosen": -2.689120292663574,
      "logits/rejected": -2.7100415229797363,
      "logps/chosen": -231.9307403564453,
      "logps/rejected": -230.61669921875,
      "loss": 0.6931,
      "rewards/accuracies": 0.3541666567325592,
      "rewards/chosen": -0.005180968437343836,
      "rewards/margins": -0.0007737001869827509,
      "rewards/rejected": -0.004407268483191729,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.990353313429303e-07,
      "logits/chosen": -2.7198684215545654,
      "logits/rejected": -2.677248477935791,
      "logps/chosen": -246.5954132080078,
      "logps/rejected": -250.72412109375,
      "loss": 0.6907,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": -0.14306680858135223,
      "rewards/margins": 0.009777521714568138,
      "rewards/rejected": -0.15284433960914612,
      "step": 20
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.882681251368548e-07,
      "logits/chosen": -2.576653242111206,
      "logits/rejected": -2.5379045009613037,
      "logps/chosen": -247.7720489501953,
      "logps/rejected": -245.41921997070312,
      "loss": 0.6804,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.21431183815002441,
      "rewards/margins": 0.036602433770895004,
      "rewards/rejected": -0.2509142756462097,
      "step": 30
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.6604720940421207e-07,
      "logits/chosen": -2.4344098567962646,
      "logits/rejected": -2.4352052211761475,
      "logps/chosen": -253.05126953125,
      "logps/rejected": -293.6170654296875,
      "loss": 0.666,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": -0.521602988243103,
      "rewards/margins": 0.07865401357412338,
      "rewards/rejected": -0.6002570390701294,
      "step": 40
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.3344075855595097e-07,
      "logits/chosen": -2.3450260162353516,
      "logits/rejected": -2.3353271484375,
      "logps/chosen": -307.8914794921875,
      "logps/rejected": -316.11505126953125,
      "loss": 0.6678,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.8211005926132202,
      "rewards/margins": 0.08438173681497574,
      "rewards/rejected": -0.9054821729660034,
      "step": 50
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.920161866827889e-07,
      "logits/chosen": -2.2511789798736572,
      "logits/rejected": -2.2335622310638428,
      "logps/chosen": -282.9967956542969,
      "logps/rejected": -303.55426025390625,
      "loss": 0.6582,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": -0.6935502886772156,
      "rewards/margins": 0.16655965149402618,
      "rewards/rejected": -0.8601099848747253,
      "step": 60
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4376480090239047e-07,
      "logits/chosen": -2.2453956604003906,
      "logits/rejected": -2.226074457168579,
      "logps/chosen": -286.8125915527344,
      "logps/rejected": -300.02789306640625,
      "loss": 0.665,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.6816620826721191,
      "rewards/margins": 0.20877805352210999,
      "rewards/rejected": -0.8904401659965515,
      "step": 70
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.910060778827554e-07,
      "logits/chosen": -2.2295961380004883,
      "logits/rejected": -2.209394931793213,
      "logps/chosen": -282.4059753417969,
      "logps/rejected": -335.4136657714844,
      "loss": 0.6747,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.8680456876754761,
      "rewards/margins": 0.1932743787765503,
      "rewards/rejected": -1.0613200664520264,
      "step": 80
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.3627616503391812e-07,
      "logits/chosen": -2.2832415103912354,
      "logits/rejected": -2.2923521995544434,
      "logps/chosen": -280.1626892089844,
      "logps/rejected": -269.5970153808594,
      "loss": 0.6566,
      "rewards/accuracies": 0.574999988079071,
      "rewards/chosen": -0.7220357656478882,
      "rewards/margins": 0.12294892966747284,
      "rewards/rejected": -0.8449847102165222,
      "step": 90
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8220596619089573e-07,
      "logits/chosen": -2.2097229957580566,
      "logits/rejected": -2.204942464828491,
      "logps/chosen": -277.9078674316406,
      "logps/rejected": -311.76177978515625,
      "loss": 0.6557,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.695646345615387,
      "rewards/margins": 0.10921863466501236,
      "rewards/rejected": -0.8048648834228516,
      "step": 100
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.3139467229135998e-07,
      "logits/chosen": -2.2849056720733643,
      "logits/rejected": -2.301518201828003,
      "logps/chosen": -298.0511474609375,
      "logps/rejected": -290.65155029296875,
      "loss": 0.6557,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.8137847781181335,
      "rewards/margins": 0.09719870239496231,
      "rewards/rejected": -0.9109834432601929,
      "step": 110
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.628481651367875e-08,
      "logits/chosen": -2.277188777923584,
      "logits/rejected": -2.2617883682250977,
      "logps/chosen": -308.5967102050781,
      "logps/rejected": -313.9825134277344,
      "loss": 0.6536,
      "rewards/accuracies": 0.5687500238418579,
      "rewards/chosen": -0.7052074670791626,
      "rewards/margins": 0.12203893810510635,
      "rewards/rejected": -0.827246367931366,
      "step": 120
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.904486005914027e-08,
      "logits/chosen": -2.3243308067321777,
      "logits/rejected": -2.3333945274353027,
      "logps/chosen": -274.1197204589844,
      "logps/rejected": -313.65740966796875,
      "loss": 0.6559,
      "rewards/accuracies": 0.5687500238418579,
      "rewards/chosen": -0.615770161151886,
      "rewards/margins": 0.12384297698736191,
      "rewards/rejected": -0.7396131753921509,
      "step": 130
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.1464952759020856e-08,
      "logits/chosen": -2.321314573287964,
      "logits/rejected": -2.318671703338623,
      "logps/chosen": -298.8154602050781,
      "logps/rejected": -320.1452331542969,
      "loss": 0.6433,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": -0.7064296007156372,
      "rewards/margins": 0.19559960067272186,
      "rewards/rejected": -0.9020291566848755,
      "step": 140
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.8708793644441086e-09,
      "logits/chosen": -2.2854580879211426,
      "logits/rejected": -2.2632009983062744,
      "logps/chosen": -282.5638732910156,
      "logps/rejected": -272.4760437011719,
      "loss": 0.6327,
      "rewards/accuracies": 0.5562499761581421,
      "rewards/chosen": -0.6732932925224304,
      "rewards/margins": 0.1675054430961609,
      "rewards/rejected": -0.8407986760139465,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 159,
      "total_flos": 0.0,
      "train_loss": 0.6621600247029239,
      "train_runtime": 2660.7412,
      "train_samples_per_second": 7.659,
      "train_steps_per_second": 0.06
    }
  ],
  "logging_steps": 10,
  "max_steps": 159,
  "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
}