{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 285,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 1.875,
      "learning_rate": 1.7241379310344828e-07,
      "logits/chosen": -2.735238790512085,
      "logits/rejected": -3.2783570289611816,
      "logps/chosen": -165.14926147460938,
      "logps/rejected": -228.84988403320312,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.7421875,
      "learning_rate": 1.724137931034483e-06,
      "logits/chosen": -2.7123961448669434,
      "logits/rejected": -3.239219903945923,
      "logps/chosen": -164.08265686035156,
      "logps/rejected": -227.88357543945312,
      "loss": 0.6921,
      "rewards/accuracies": 0.6111111044883728,
      "rewards/chosen": 0.0008442008402198553,
      "rewards/margins": 0.002071016002446413,
      "rewards/rejected": -0.0012268151622265577,
      "step": 10
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.828125,
      "learning_rate": 3.448275862068966e-06,
      "logits/chosen": -2.7431998252868652,
      "logits/rejected": -3.2384581565856934,
      "logps/chosen": -163.16790771484375,
      "logps/rejected": -229.4684600830078,
      "loss": 0.6817,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.007237003184854984,
      "rewards/margins": 0.023129161447286606,
      "rewards/rejected": -0.015892159193754196,
      "step": 20
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.6640625,
      "learning_rate": 4.999811754597862e-06,
      "logits/chosen": -2.704376220703125,
      "logits/rejected": -3.244763135910034,
      "logps/chosen": -158.67416381835938,
      "logps/rejected": -237.81668090820312,
      "loss": 0.6284,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.03751087933778763,
      "rewards/margins": 0.13478627800941467,
      "rewards/rejected": -0.09727539122104645,
      "step": 30
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.578125,
      "learning_rate": 4.97725658856945e-06,
      "logits/chosen": -2.718313217163086,
      "logits/rejected": -3.2308642864227295,
      "logps/chosen": -150.95199584960938,
      "logps/rejected": -256.7232666015625,
      "loss": 0.5081,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.11078281700611115,
      "rewards/margins": 0.415455162525177,
      "rewards/rejected": -0.3046723008155823,
      "step": 40
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3828125,
      "learning_rate": 4.917441177612131e-06,
      "logits/chosen": -2.699450731277466,
      "logits/rejected": -3.2096571922302246,
      "logps/chosen": -140.04672241210938,
      "logps/rejected": -286.2795104980469,
      "loss": 0.3683,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2172900140285492,
      "rewards/margins": 0.8159033060073853,
      "rewards/rejected": -0.5986132025718689,
      "step": 50
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0390625,
      "learning_rate": 4.82126520118304e-06,
      "logits/chosen": -2.7382638454437256,
      "logits/rejected": -3.209529399871826,
      "logps/chosen": -127.2552261352539,
      "logps/rejected": -322.1763610839844,
      "loss": 0.2407,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.36736661195755005,
      "rewards/margins": 1.3110934495925903,
      "rewards/rejected": -0.9437268972396851,
      "step": 60
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.6875,
      "learning_rate": 4.6901752354885166e-06,
      "logits/chosen": -2.696700096130371,
      "logits/rejected": -3.2043445110321045,
      "logps/chosen": -115.544677734375,
      "logps/rejected": -366.6884765625,
      "loss": 0.1455,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.4782423973083496,
      "rewards/margins": 1.8693335056304932,
      "rewards/rejected": -1.391090750694275,
      "step": 70
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.48828125,
      "learning_rate": 4.526142995631488e-06,
      "logits/chosen": -2.7269909381866455,
      "logits/rejected": -3.2105610370635986,
      "logps/chosen": -109.22001647949219,
      "logps/rejected": -415.4657287597656,
      "loss": 0.0858,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.5552297234535217,
      "rewards/margins": 2.424295663833618,
      "rewards/rejected": -1.8690656423568726,
      "step": 80
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.392578125,
      "learning_rate": 4.331635679181032e-06,
      "logits/chosen": -2.685180425643921,
      "logits/rejected": -3.2150089740753174,
      "logps/chosen": -104.05949401855469,
      "logps/rejected": -453.306884765625,
      "loss": 0.0559,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.5974525213241577,
      "rewards/margins": 2.865081548690796,
      "rewards/rejected": -2.2676291465759277,
      "step": 90
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.302734375,
      "learning_rate": 4.109578857224478e-06,
      "logits/chosen": -2.702922821044922,
      "logits/rejected": -3.216787338256836,
      "logps/chosen": -100.30770111083984,
      "logps/rejected": -488.27081298828125,
      "loss": 0.039,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6299440264701843,
      "rewards/margins": 3.2341396808624268,
      "rewards/rejected": -2.6041955947875977,
      "step": 100
    },
    {
      "epoch": 0.35,
      "eval_logits/chosen": -2.7171199321746826,
      "eval_logits/rejected": -3.2064568996429443,
      "eval_logps/chosen": -79.25934600830078,
      "eval_logps/rejected": -452.1828918457031,
      "eval_loss": 0.052521564066410065,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 0.5119398236274719,
      "eval_rewards/margins": 2.9491662979125977,
      "eval_rewards/rejected": -2.4372265338897705,
      "eval_runtime": 1.2939,
      "eval_samples_per_second": 3.864,
      "eval_steps_per_second": 2.319,
      "step": 100
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.259765625,
      "learning_rate": 3.863312471055116e-06,
      "logits/chosen": -2.7052559852600098,
      "logits/rejected": -3.191067934036255,
      "logps/chosen": -98.0762710571289,
      "logps/rejected": -517.7929077148438,
      "loss": 0.0281,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.654212236404419,
      "rewards/margins": 3.561893939971924,
      "rewards/rejected": -2.907681465148926,
      "step": 110
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.212890625,
      "learning_rate": 3.5965405963463197e-06,
      "logits/chosen": -2.7348670959472656,
      "logits/rejected": -3.205155849456787,
      "logps/chosen": -95.99646759033203,
      "logps/rejected": -544.2891845703125,
      "loss": 0.0213,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6714237332344055,
      "rewards/margins": 3.842583417892456,
      "rewards/rejected": -3.1711599826812744,
      "step": 120
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.1845703125,
      "learning_rate": 3.313275730405658e-06,
      "logits/chosen": -2.698134183883667,
      "logits/rejected": -3.215333938598633,
      "logps/chosen": -94.20452880859375,
      "logps/rejected": -564.6442260742188,
      "loss": 0.0173,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6862392425537109,
      "rewards/margins": 4.055664539337158,
      "rewards/rejected": -3.369424819946289,
      "step": 130
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.1591796875,
      "learning_rate": 3.0177784404805466e-06,
      "logits/chosen": -2.728977680206299,
      "logits/rejected": -3.2040677070617676,
      "logps/chosen": -95.00137329101562,
      "logps/rejected": -582.2269287109375,
      "loss": 0.0145,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6958636045455933,
      "rewards/margins": 4.233551502227783,
      "rewards/rejected": -3.5376884937286377,
      "step": 140
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.1640625,
      "learning_rate": 2.7144932808611002e-06,
      "logits/chosen": -2.705024003982544,
      "logits/rejected": -3.1964850425720215,
      "logps/chosen": -94.14689636230469,
      "logps/rejected": -587.1410522460938,
      "loss": 0.0136,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6878983974456787,
      "rewards/margins": 4.294985294342041,
      "rewards/rejected": -3.6070873737335205,
      "step": 150
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.1572265625,
      "learning_rate": 2.407981942646603e-06,
      "logits/chosen": -2.7173984050750732,
      "logits/rejected": -3.229872226715088,
      "logps/chosen": -93.74874877929688,
      "logps/rejected": -597.3945922851562,
      "loss": 0.0121,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6988341212272644,
      "rewards/margins": 4.410033226013184,
      "rewards/rejected": -3.7111988067626953,
      "step": 160
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.1435546875,
      "learning_rate": 2.102854641665347e-06,
      "logits/chosen": -2.711818218231201,
      "logits/rejected": -3.219580888748169,
      "logps/chosen": -92.11972045898438,
      "logps/rejected": -603.1061401367188,
      "loss": 0.0115,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.7005870342254639,
      "rewards/margins": 4.465435981750488,
      "rewards/rejected": -3.7648491859436035,
      "step": 170
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.154296875,
      "learning_rate": 1.8037007765373677e-06,
      "logits/chosen": -2.736745834350586,
      "logits/rejected": -3.206763505935669,
      "logps/chosen": -93.66650390625,
      "logps/rejected": -605.5599365234375,
      "loss": 0.0113,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.700027346611023,
      "rewards/margins": 4.484536170959473,
      "rewards/rejected": -3.7845091819763184,
      "step": 180
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.1484375,
      "learning_rate": 1.5150198998473802e-06,
      "logits/chosen": -2.6961426734924316,
      "logits/rejected": -3.201364517211914,
      "logps/chosen": -92.75911712646484,
      "logps/rejected": -609.6072998046875,
      "loss": 0.0107,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.7066330313682556,
      "rewards/margins": 4.538401126861572,
      "rewards/rejected": -3.8317675590515137,
      "step": 190
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.142578125,
      "learning_rate": 1.2411540406857064e-06,
      "logits/chosen": -2.697049140930176,
      "logits/rejected": -3.2053134441375732,
      "logps/chosen": -92.19438171386719,
      "logps/rejected": -611.137939453125,
      "loss": 0.0106,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.7058606147766113,
      "rewards/margins": 4.546854496002197,
      "rewards/rejected": -3.840993881225586,
      "step": 200
    },
    {
      "epoch": 0.7,
      "eval_logits/chosen": -2.7109670639038086,
      "eval_logits/rejected": -3.2030317783355713,
      "eval_logps/chosen": -74.4535140991211,
      "eval_logps/rejected": -542.8434448242188,
      "eval_loss": 0.023632029071450233,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 0.5599982738494873,
      "eval_rewards/margins": 3.903829574584961,
      "eval_rewards/rejected": -3.343831777572632,
      "eval_runtime": 1.2832,
      "eval_samples_per_second": 3.896,
      "eval_steps_per_second": 2.338,
      "step": 200
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.1318359375,
      "learning_rate": 9.862223964891864e-07,
      "logits/chosen": -2.728084087371826,
      "logits/rejected": -3.2059364318847656,
      "logps/chosen": -90.77999114990234,
      "logps/rejected": -605.8444213867188,
      "loss": 0.0124,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.694692850112915,
      "rewards/margins": 4.4922895431518555,
      "rewards/rejected": -3.797595977783203,
      "step": 210
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.1279296875,
      "learning_rate": 7.54059376477568e-07,
      "logits/chosen": -2.712827682495117,
      "logits/rejected": -3.2164974212646484,
      "logps/chosen": -93.00579071044922,
      "logps/rejected": -616.4993286132812,
      "loss": 0.0102,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.7107797861099243,
      "rewards/margins": 4.593611717224121,
      "rewards/rejected": -3.8828322887420654,
      "step": 220
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.130859375,
      "learning_rate": 5.481569285697638e-07,
      "logits/chosen": -2.7499520778656006,
      "logits/rejected": -3.2049834728240967,
      "logps/chosen": -93.71913146972656,
      "logps/rejected": -612.4299926757812,
      "loss": 0.0106,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6972833871841431,
      "rewards/margins": 4.548653602600098,
      "rewards/rejected": -3.851370334625244,
      "step": 230
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1328125,
      "learning_rate": 3.71612017236837e-07,
      "logits/chosen": -2.7227025032043457,
      "logits/rejected": -3.198974132537842,
      "logps/chosen": -93.90065002441406,
      "logps/rejected": -616.3263549804688,
      "loss": 0.0101,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.7073711156845093,
      "rewards/margins": 4.595206260681152,
      "rewards/rejected": -3.8878350257873535,
      "step": 240
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1298828125,
      "learning_rate": 2.2708004227369407e-07,
      "logits/chosen": -2.7253494262695312,
      "logits/rejected": -3.200887680053711,
      "logps/chosen": -92.90316009521484,
      "logps/rejected": -616.0049438476562,
      "loss": 0.0102,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6983229517936707,
      "rewards/margins": 4.588388919830322,
      "rewards/rejected": -3.890066146850586,
      "step": 250
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.12890625,
      "learning_rate": 1.1673489911451536e-07,
      "logits/chosen": -2.6972110271453857,
      "logits/rejected": -3.2123961448669434,
      "logps/chosen": -92.834716796875,
      "logps/rejected": -614.0274658203125,
      "loss": 0.0103,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.7053926587104797,
      "rewards/margins": 4.57442569732666,
      "rewards/rejected": -3.869033098220825,
      "step": 260
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.134765625,
      "learning_rate": 4.223628142195929e-08,
      "logits/chosen": -2.706958055496216,
      "logits/rejected": -3.205293655395508,
      "logps/chosen": -93.16792297363281,
      "logps/rejected": -615.3536376953125,
      "loss": 0.0101,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.7087845802307129,
      "rewards/margins": 4.595980167388916,
      "rewards/rejected": -3.887195110321045,
      "step": 270
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1318359375,
      "learning_rate": 4.704717749627052e-09,
      "logits/chosen": -2.6964120864868164,
      "logits/rejected": -3.2064449787139893,
      "logps/chosen": -93.33256530761719,
      "logps/rejected": -614.2545166015625,
      "loss": 0.0104,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.6982468962669373,
      "rewards/margins": 4.565683841705322,
      "rewards/rejected": -3.8674368858337402,
      "step": 280
    },
    {
      "epoch": 1.0,
      "step": 285,
      "total_flos": 0.0,
      "train_loss": 0.12933478820742222,
      "train_runtime": 748.758,
      "train_samples_per_second": 1.521,
      "train_steps_per_second": 0.381
    }
  ],
  "logging_steps": 10,
  "max_steps": 285,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 200,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}