{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.968,
  "eval_steps": 100,
  "global_step": 248,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 2.0000000000000002e-07,
      "logits/chosen": -0.026431415230035782,
      "logits/rejected": -0.16986289620399475,
      "logps/chosen": -151.98907470703125,
      "logps/rejected": -93.52606964111328,
      "loss": 0.0011,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.16,
      "learning_rate": 2.0000000000000003e-06,
      "logits/chosen": -0.147756889462471,
      "logits/rejected": 0.020012276247143745,
      "logps/chosen": -136.29202270507812,
      "logps/rejected": -75.50608825683594,
      "loss": 0.0013,
      "rewards/accuracies": 0.4027777910232544,
      "rewards/chosen": -0.0007271924405358732,
      "rewards/margins": -0.0012592646526172757,
      "rewards/rejected": 0.0005320722702890635,
      "step": 10
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.000000000000001e-06,
      "logits/chosen": -0.055388081818819046,
      "logits/rejected": 0.15466341376304626,
      "logps/chosen": -146.0823974609375,
      "logps/rejected": -90.69671630859375,
      "loss": 0.0013,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": -3.685036062961444e-05,
      "rewards/margins": -0.0004525856929831207,
      "rewards/rejected": 0.00041573523776605725,
      "step": 20
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.993800445762451e-06,
      "logits/chosen": -0.2320336550474167,
      "logits/rejected": 0.11852385103702545,
      "logps/chosen": -137.83750915527344,
      "logps/rejected": -78.5675048828125,
      "loss": 0.0012,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.0010912430007010698,
      "rewards/margins": 0.001036056550219655,
      "rewards/rejected": 5.518653051694855e-05,
      "step": 30
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.944388344834205e-06,
      "logits/chosen": -0.051973842084407806,
      "logits/rejected": 0.04229766130447388,
      "logps/chosen": -142.12911987304688,
      "logps/rejected": -96.51297760009766,
      "loss": 0.0013,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.0018935591215267777,
      "rewards/margins": 0.0001399043685523793,
      "rewards/rejected": 0.001753654913045466,
      "step": 40
    },
    {
      "epoch": 0.8,
      "learning_rate": 4.8465431931347904e-06,
      "logits/chosen": -0.16981413960456848,
      "logits/rejected": 0.09270621091127396,
      "logps/chosen": -133.02362060546875,
      "logps/rejected": -90.3238525390625,
      "loss": 0.0012,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": 0.002572892000898719,
      "rewards/margins": 0.0004979773075319827,
      "rewards/rejected": 0.002074914751574397,
      "step": 50
    },
    {
      "epoch": 0.96,
      "learning_rate": 4.702203692102539e-06,
      "logits/chosen": -0.10357820987701416,
      "logits/rejected": 0.13292734324932098,
      "logps/chosen": -143.26065063476562,
      "logps/rejected": -81.9512710571289,
      "loss": 0.0013,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.0026912898756563663,
      "rewards/margins": 0.0002925347362179309,
      "rewards/rejected": 0.0023987549357116222,
      "step": 60
    },
    {
      "epoch": 1.12,
      "learning_rate": 4.514229781074239e-06,
      "logits/chosen": -0.08756460249423981,
      "logits/rejected": -0.010441536083817482,
      "logps/chosen": -155.8980255126953,
      "logps/rejected": -105.80094146728516,
      "loss": 0.0011,
      "rewards/accuracies": 0.543749988079071,
      "rewards/chosen": 0.0051867421716451645,
      "rewards/margins": 0.003287202911451459,
      "rewards/rejected": 0.00189953891094774,
      "step": 70
    },
    {
      "epoch": 1.28,
      "learning_rate": 4.286345970517195e-06,
      "logits/chosen": -0.11689458042383194,
      "logits/rejected": 0.07064902782440186,
      "logps/chosen": -124.151123046875,
      "logps/rejected": -64.7757339477539,
      "loss": 0.001,
      "rewards/accuracies": 0.5562499761581421,
      "rewards/chosen": 0.005000757984817028,
      "rewards/margins": 0.0033935843966901302,
      "rewards/rejected": 0.0016071733552962542,
      "step": 80
    },
    {
      "epoch": 1.44,
      "learning_rate": 4.023067544670082e-06,
      "logits/chosen": -0.1919364184141159,
      "logits/rejected": 0.19178633391857147,
      "logps/chosen": -134.85496520996094,
      "logps/rejected": -79.26893615722656,
      "loss": 0.0011,
      "rewards/accuracies": 0.543749988079071,
      "rewards/chosen": 0.005402157548815012,
      "rewards/margins": 0.0031424053013324738,
      "rewards/rejected": 0.0022597520146518946,
      "step": 90
    },
    {
      "epoch": 1.6,
      "learning_rate": 3.7296110958116845e-06,
      "logits/chosen": -0.21874800324440002,
      "logits/rejected": 0.08748228847980499,
      "logps/chosen": -143.6033935546875,
      "logps/rejected": -93.09342956542969,
      "loss": 0.0011,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.0063162692822515965,
      "rewards/margins": 0.0023192143999040127,
      "rewards/rejected": 0.003997053951025009,
      "step": 100
    },
    {
      "epoch": 1.6,
      "eval_logits/chosen": -0.0479925237596035,
      "eval_logits/rejected": 0.05092725157737732,
      "eval_logps/chosen": -307.1817321777344,
      "eval_logps/rejected": -279.2488708496094,
      "eval_loss": 0.0018035146640613675,
      "eval_rewards/accuracies": 0.47999998927116394,
      "eval_rewards/chosen": -0.003050154075026512,
      "eval_rewards/margins": -0.0009030703222379088,
      "eval_rewards/rejected": -0.0021470836363732815,
      "eval_runtime": 412.557,
      "eval_samples_per_second": 4.848,
      "eval_steps_per_second": 1.212,
      "step": 100
    },
    {
      "epoch": 1.76,
      "learning_rate": 3.4117911628292944e-06,
      "logits/chosen": -0.05666325241327286,
      "logits/rejected": 0.15881529450416565,
      "logps/chosen": -156.90457153320312,
      "logps/rejected": -89.17283630371094,
      "loss": 0.001,
      "rewards/accuracies": 0.606249988079071,
      "rewards/chosen": 0.009410916827619076,
      "rewards/margins": 0.005939081776887178,
      "rewards/rejected": 0.0034718364477157593,
      "step": 110
    },
    {
      "epoch": 1.92,
      "learning_rate": 3.075905022087675e-06,
      "logits/chosen": -0.07111762464046478,
      "logits/rejected": 0.03149420768022537,
      "logps/chosen": -154.131103515625,
      "logps/rejected": -102.81583404541016,
      "loss": 0.0012,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.008770602755248547,
      "rewards/margins": 0.0032143122516572475,
      "rewards/rejected": 0.005556290503591299,
      "step": 120
    },
    {
      "epoch": 2.08,
      "learning_rate": 2.728607913349464e-06,
      "logits/chosen": -0.25201350450515747,
      "logits/rejected": 0.08731786906719208,
      "logps/chosen": -133.0045928955078,
      "logps/rejected": -75.28916931152344,
      "loss": 0.001,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": 0.010033163242042065,
      "rewards/margins": 0.005545603111386299,
      "rewards/rejected": 0.004487560596317053,
      "step": 130
    },
    {
      "epoch": 2.24,
      "learning_rate": 2.376781173017589e-06,
      "logits/chosen": -0.23511283099651337,
      "logits/rejected": 0.056128501892089844,
      "logps/chosen": -139.34347534179688,
      "logps/rejected": -83.03041076660156,
      "loss": 0.0009,
      "rewards/accuracies": 0.643750011920929,
      "rewards/chosen": 0.011506117880344391,
      "rewards/margins": 0.007246524095535278,
      "rewards/rejected": 0.0042595937848091125,
      "step": 140
    },
    {
      "epoch": 2.4,
      "learning_rate": 2.0273958875043877e-06,
      "logits/chosen": -0.15843239426612854,
      "logits/rejected": -0.025659451261162758,
      "logps/chosen": -128.82235717773438,
      "logps/rejected": -71.82875061035156,
      "loss": 0.001,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": 0.010396704077720642,
      "rewards/margins": 0.006070839706808329,
      "rewards/rejected": 0.004325864836573601,
      "step": 150
    },
    {
      "epoch": 2.56,
      "learning_rate": 1.6873747682962393e-06,
      "logits/chosen": -0.10364966094493866,
      "logits/rejected": 0.15611091256141663,
      "logps/chosen": -146.84365844726562,
      "logps/rejected": -94.91715240478516,
      "loss": 0.0011,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": 0.01151433028280735,
      "rewards/margins": 0.005733857862651348,
      "rewards/rejected": 0.005780472420156002,
      "step": 160
    },
    {
      "epoch": 2.72,
      "learning_rate": 1.363454985517803e-06,
      "logits/chosen": -0.14020611345767975,
      "logits/rejected": 0.0140873147174716,
      "logps/chosen": -130.91256713867188,
      "logps/rejected": -88.08763122558594,
      "loss": 0.001,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.01202808041125536,
      "rewards/margins": 0.005936866160482168,
      "rewards/rejected": 0.006091213319450617,
      "step": 170
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.062054677808238e-06,
      "logits/chosen": -0.07270533591508865,
      "logits/rejected": 0.16701629757881165,
      "logps/chosen": -149.4364471435547,
      "logps/rejected": -83.87244415283203,
      "loss": 0.001,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.011001331731677055,
      "rewards/margins": 0.0055719343945384026,
      "rewards/rejected": 0.00542939780279994,
      "step": 180
    },
    {
      "epoch": 3.04,
      "learning_rate": 7.891457834794711e-07,
      "logits/chosen": -0.13850273191928864,
      "logits/rejected": 0.006671518087387085,
      "logps/chosen": -151.9923553466797,
      "logps/rejected": -94.79319763183594,
      "loss": 0.001,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": 0.014899802394211292,
      "rewards/margins": 0.006995724979788065,
      "rewards/rejected": 0.00790407694876194,
      "step": 190
    },
    {
      "epoch": 3.2,
      "learning_rate": 5.501357126768117e-07,
      "logits/chosen": -0.14668866991996765,
      "logits/rejected": 0.10788450390100479,
      "logps/chosen": -140.62472534179688,
      "logps/rejected": -88.3201904296875,
      "loss": 0.001,
      "rewards/accuracies": 0.6187499761581421,
      "rewards/chosen": 0.011035704985260963,
      "rewards/margins": 0.006984876934438944,
      "rewards/rejected": 0.004050827585160732,
      "step": 200
    },
    {
      "epoch": 3.2,
      "eval_logits/chosen": -0.0663725882768631,
      "eval_logits/rejected": 0.03233730420470238,
      "eval_logps/chosen": -307.4276123046875,
      "eval_logps/rejected": -279.4666748046875,
      "eval_loss": 0.0018860435811802745,
      "eval_rewards/accuracies": 0.476500004529953,
      "eval_rewards/chosen": -0.0055097793228924274,
      "eval_rewards/margins": -0.001184401917271316,
      "eval_rewards/rejected": -0.0043253772892057896,
      "eval_runtime": 412.8262,
      "eval_samples_per_second": 4.845,
      "eval_steps_per_second": 1.211,
      "step": 200
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.4976020508682345e-07,
      "logits/chosen": -0.2199704945087433,
      "logits/rejected": 0.13789795339107513,
      "logps/chosen": -151.80128479003906,
      "logps/rejected": -87.3798599243164,
      "loss": 0.001,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.012573355808854103,
      "rewards/margins": 0.006345916539430618,
      "rewards/rejected": 0.00622743833810091,
      "step": 210
    },
    {
      "epoch": 3.52,
      "learning_rate": 1.9198949610721273e-07,
      "logits/chosen": -0.19111016392707825,
      "logits/rejected": -0.011430763639509678,
      "logps/chosen": -132.64901733398438,
      "logps/rejected": -73.41009521484375,
      "loss": 0.0009,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.010676460340619087,
      "rewards/margins": 0.006789586041122675,
      "rewards/rejected": 0.0038868754636496305,
      "step": 220
    },
    {
      "epoch": 3.68,
      "learning_rate": 7.994965069994143e-08,
      "logits/chosen": -0.1486184000968933,
      "logits/rejected": 0.17951056361198425,
      "logps/chosen": -139.86245727539062,
      "logps/rejected": -91.14982604980469,
      "loss": 0.001,
      "rewards/accuracies": 0.606249988079071,
      "rewards/chosen": 0.012494201771914959,
      "rewards/margins": 0.006681007333099842,
      "rewards/rejected": 0.005813195835798979,
      "step": 230
    },
    {
      "epoch": 3.84,
      "learning_rate": 1.5860623616664183e-08,
      "logits/chosen": -0.13247700035572052,
      "logits/rejected": -0.05245450884103775,
      "logps/chosen": -140.35784912109375,
      "logps/rejected": -86.07032012939453,
      "loss": 0.0009,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.014865470118820667,
      "rewards/margins": 0.008430338464677334,
      "rewards/rejected": 0.006435131188482046,
      "step": 240
    },
    {
      "epoch": 3.97,
      "step": 248,
      "total_flos": 0.0,
      "train_loss": 0.0010713144931228712,
      "train_runtime": 2635.242,
      "train_samples_per_second": 1.518,
      "train_steps_per_second": 0.094
    }
  ],
  "logging_steps": 10,
  "max_steps": 248,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}