{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 193,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 18.047405023374292,
      "learning_rate": 2.5e-08,
      "logits/chosen": -2.3762412071228027,
      "logits/rejected": -2.8935458660125732,
      "logps/chosen": -633.0433349609375,
      "logps/rejected": -1050.6610107421875,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.05,
      "grad_norm": 19.360233123290687,
      "learning_rate": 2.5e-07,
      "logits/chosen": -2.7744946479797363,
      "logits/rejected": -2.76593017578125,
      "logps/chosen": -479.1276550292969,
      "logps/rejected": -986.8477783203125,
      "loss": 0.6924,
      "rewards/accuracies": 0.4305555522441864,
      "rewards/chosen": 0.0005723661743104458,
      "rewards/margins": 0.0007126606651581824,
      "rewards/rejected": -0.00014029450539965183,
      "step": 10
    },
    {
      "epoch": 0.1,
      "grad_norm": 20.190218055543376,
      "learning_rate": 5e-07,
      "logits/chosen": -2.8035528659820557,
      "logits/rejected": -2.70589280128479,
      "logps/chosen": -569.3781127929688,
      "logps/rejected": -963.7088012695312,
      "loss": 0.6772,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 0.028690308332443237,
      "rewards/margins": 0.031130677089095116,
      "rewards/rejected": -0.002440371550619602,
      "step": 20
    },
    {
      "epoch": 0.16,
      "grad_norm": 23.740010487903497,
      "learning_rate": 4.958892245285593e-07,
      "logits/chosen": -2.8474576473236084,
      "logits/rejected": -2.897707462310791,
      "logps/chosen": -587.6588134765625,
      "logps/rejected": -1056.246337890625,
      "loss": 0.6377,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.03852864354848862,
      "rewards/margins": 0.11339948326349258,
      "rewards/rejected": -0.07487084716558456,
      "step": 30
    },
    {
      "epoch": 0.21,
      "grad_norm": 41.01050783833045,
      "learning_rate": 4.836920859140499e-07,
      "logits/chosen": -3.0017783641815186,
      "logits/rejected": -3.0174295902252197,
      "logps/chosen": -484.53204345703125,
      "logps/rejected": -1140.9752197265625,
      "loss": 0.4858,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": -0.15590691566467285,
      "rewards/margins": 0.6636868119239807,
      "rewards/rejected": -0.8195937275886536,
      "step": 40
    },
    {
      "epoch": 0.26,
      "grad_norm": 29.37990905047941,
      "learning_rate": 4.638097017423782e-07,
      "logits/chosen": -3.24768328666687,
      "logits/rejected": -3.380809783935547,
      "logps/chosen": -613.0138549804688,
      "logps/rejected": -1256.5755615234375,
      "loss": 0.3608,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": -0.6601036190986633,
      "rewards/margins": 2.2774975299835205,
      "rewards/rejected": -2.937601327896118,
      "step": 50
    },
    {
      "epoch": 0.31,
      "grad_norm": 76.05443101336851,
      "learning_rate": 4.3689592815087764e-07,
      "logits/chosen": -3.0450172424316406,
      "logits/rejected": -3.2747421264648438,
      "logps/chosen": -602.8151245117188,
      "logps/rejected": -1274.7425537109375,
      "loss": 0.3022,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": -0.6452581882476807,
      "rewards/margins": 2.6640372276306152,
      "rewards/rejected": -3.309295654296875,
      "step": 60
    },
    {
      "epoch": 0.36,
      "grad_norm": 33.12857974455004,
      "learning_rate": 4.038358569821387e-07,
      "logits/chosen": -3.0854501724243164,
      "logits/rejected": -3.26892352104187,
      "logps/chosen": -661.424072265625,
      "logps/rejected": -1598.431884765625,
      "loss": 0.2321,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": -1.154323697090149,
      "rewards/margins": 4.336134910583496,
      "rewards/rejected": -5.4904584884643555,
      "step": 70
    },
    {
      "epoch": 0.41,
      "grad_norm": 40.2672306106755,
      "learning_rate": 3.65716708473318e-07,
      "logits/chosen": -3.1047134399414062,
      "logits/rejected": -3.264630079269409,
      "logps/chosen": -605.7648315429688,
      "logps/rejected": -1600.621826171875,
      "loss": 0.1992,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": -0.9309484362602234,
      "rewards/margins": 4.815203666687012,
      "rewards/rejected": -5.746151924133301,
      "step": 80
    },
    {
      "epoch": 0.47,
      "grad_norm": 31.009610555577336,
      "learning_rate": 3.2379207670987346e-07,
      "logits/chosen": -3.1034348011016846,
      "logits/rejected": -3.04689359664917,
      "logps/chosen": -630.8348388671875,
      "logps/rejected": -1542.02685546875,
      "loss": 0.1838,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.8493152856826782,
      "rewards/margins": 5.170994758605957,
      "rewards/rejected": -6.020310401916504,
      "step": 90
    },
    {
      "epoch": 0.52,
      "grad_norm": 49.251429406444444,
      "learning_rate": 2.79440703675024e-07,
      "logits/chosen": -3.0847506523132324,
      "logits/rejected": -3.1429543495178223,
      "logps/chosen": -606.3587036132812,
      "logps/rejected": -1756.322265625,
      "loss": 0.1625,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": -0.8517366647720337,
      "rewards/margins": 6.2657976150512695,
      "rewards/rejected": -7.1175336837768555,
      "step": 100
    },
    {
      "epoch": 0.52,
      "eval_logits/chosen": -3.3846025466918945,
      "eval_logits/rejected": -3.3073556423187256,
      "eval_logps/chosen": -972.7769165039062,
      "eval_logps/rejected": -1444.57763671875,
      "eval_loss": 0.33459049463272095,
      "eval_rewards/accuracies": 0.7916666865348816,
      "eval_rewards/chosen": -1.212856411933899,
      "eval_rewards/margins": 4.115774631500244,
      "eval_rewards/rejected": -5.328631401062012,
      "eval_runtime": 34.9981,
      "eval_samples_per_second": 8.115,
      "eval_steps_per_second": 0.257,
      "step": 100
    },
    {
      "epoch": 0.57,
      "grad_norm": 17.436897320860403,
      "learning_rate": 2.341211376599406e-07,
      "logits/chosen": -3.0867555141448975,
      "logits/rejected": -3.298828125,
      "logps/chosen": -614.1629028320312,
      "logps/rejected": -1688.4345703125,
      "loss": 0.1466,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -0.7798967361450195,
      "rewards/margins": 5.776389122009277,
      "rewards/rejected": -6.556285858154297,
      "step": 110
    },
    {
      "epoch": 0.62,
      "grad_norm": 31.209651266501922,
      "learning_rate": 1.8932376714743234e-07,
      "logits/chosen": -3.143540620803833,
      "logits/rejected": -3.1130645275115967,
      "logps/chosen": -650.592041015625,
      "logps/rejected": -1678.760498046875,
      "loss": 0.1488,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": -0.9372560381889343,
      "rewards/margins": 5.595232963562012,
      "rewards/rejected": -6.532488822937012,
      "step": 120
    },
    {
      "epoch": 0.67,
      "grad_norm": 27.932007137908535,
      "learning_rate": 1.465218075926022e-07,
      "logits/chosen": -3.0368003845214844,
      "logits/rejected": -3.2008068561553955,
      "logps/chosen": -692.443359375,
      "logps/rejected": -1812.3238525390625,
      "loss": 0.1393,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.2292582988739014,
      "rewards/margins": 6.150660514831543,
      "rewards/rejected": -7.379919528961182,
      "step": 130
    },
    {
      "epoch": 0.73,
      "grad_norm": 38.49048498668927,
      "learning_rate": 1.071228529591909e-07,
      "logits/chosen": -3.0865681171417236,
      "logits/rejected": -3.1455368995666504,
      "logps/chosen": -677.2738037109375,
      "logps/rejected": -1936.8187255859375,
      "loss": 0.1349,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -1.3584263324737549,
      "rewards/margins": 7.760031700134277,
      "rewards/rejected": -9.118457794189453,
      "step": 140
    },
    {
      "epoch": 0.78,
      "grad_norm": 131.57411160189204,
      "learning_rate": 7.242258529765793e-08,
      "logits/chosen": -3.0663182735443115,
      "logits/rejected": -3.130246877670288,
      "logps/chosen": -719.6699829101562,
      "logps/rejected": -2027.244873046875,
      "loss": 0.1059,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": -1.5801483392715454,
      "rewards/margins": 8.059651374816895,
      "rewards/rejected": -9.639799118041992,
      "step": 150
    },
    {
      "epoch": 0.83,
      "grad_norm": 20.032754023483275,
      "learning_rate": 4.3562164681246824e-08,
      "logits/chosen": -3.028076410293579,
      "logits/rejected": -3.1169402599334717,
      "logps/chosen": -710.513427734375,
      "logps/rejected": -2098.9404296875,
      "loss": 0.1314,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": -1.6005052328109741,
      "rewards/margins": 8.92524528503418,
      "rewards/rejected": -10.525751113891602,
      "step": 160
    },
    {
      "epoch": 0.88,
      "grad_norm": 43.60438550060245,
      "learning_rate": 2.1490700783280884e-08,
      "logits/chosen": -2.9330756664276123,
      "logits/rejected": -3.0294029712677,
      "logps/chosen": -678.80615234375,
      "logps/rejected": -1974.254638671875,
      "loss": 0.1197,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": -1.2993422746658325,
      "rewards/margins": 9.22239875793457,
      "rewards/rejected": -10.52174186706543,
      "step": 170
    },
    {
      "epoch": 0.93,
      "grad_norm": 26.084829370342806,
      "learning_rate": 6.9340402630445885e-09,
      "logits/chosen": -3.012688159942627,
      "logits/rejected": -3.114205837249756,
      "logps/chosen": -689.8292236328125,
      "logps/rejected": -2119.27197265625,
      "loss": 0.1299,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": -1.4270641803741455,
      "rewards/margins": 9.246889114379883,
      "rewards/rejected": -10.673952102661133,
      "step": 180
    },
    {
      "epoch": 0.98,
      "grad_norm": 56.458996538652926,
      "learning_rate": 3.708964246392221e-10,
      "logits/chosen": -3.041151523590088,
      "logits/rejected": -2.961495876312256,
      "logps/chosen": -715.9195556640625,
      "logps/rejected": -2031.1800537109375,
      "loss": 0.1285,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": -1.2981992959976196,
      "rewards/margins": 9.372197151184082,
      "rewards/rejected": -10.67039680480957,
      "step": 190
    },
    {
      "epoch": 1.0,
      "step": 193,
      "total_flos": 0.0,
      "train_loss": 0.2672084366101675,
      "train_runtime": 3083.1249,
      "train_samples_per_second": 3.999,
      "train_steps_per_second": 0.063
    }
  ],
  "logging_steps": 10,
  "max_steps": 193,
  "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
}