{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.998272884283247,
  "eval_steps": 100,
  "global_step": 289,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 0.314453125,
      "learning_rate": 1.7241379310344828e-07,
      "logits/chosen": -2.648426055908203,
      "logits/rejected": -2.6488823890686035,
      "logps/chosen": -29.924301147460938,
      "logps/rejected": -43.789119720458984,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.294921875,
      "learning_rate": 1.724137931034483e-06,
      "logits/chosen": -2.791017532348633,
      "logits/rejected": -2.792222023010254,
      "logps/chosen": -30.778282165527344,
      "logps/rejected": -44.59426498413086,
      "loss": 0.6929,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.0004109032452106476,
      "rewards/margins": 0.0004249625199008733,
      "rewards/rejected": -1.4059260138310492e-05,
      "step": 10
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.3203125,
      "learning_rate": 3.448275862068966e-06,
      "logits/chosen": -2.8013782501220703,
      "logits/rejected": -2.80298113822937,
      "logps/chosen": -30.407978057861328,
      "logps/rejected": -46.99640655517578,
      "loss": 0.6919,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 0.0021479069255292416,
      "rewards/margins": 0.0025836334098130465,
      "rewards/rejected": -0.00043572625145316124,
      "step": 20
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.30078125,
      "learning_rate": 4.999817502139027e-06,
      "logits/chosen": -2.8514046669006348,
      "logits/rejected": -2.851513385772705,
      "logps/chosen": -29.340972900390625,
      "logps/rejected": -46.38640594482422,
      "loss": 0.6853,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.016690988093614578,
      "rewards/margins": 0.01585637591779232,
      "rewards/rejected": 0.0008346118847839534,
      "step": 30
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.34765625,
      "learning_rate": 4.977949980164773e-06,
      "logits/chosen": -2.77081036567688,
      "logits/rejected": -2.7683987617492676,
      "logps/chosen": -24.99424934387207,
      "logps/rejected": -43.76649475097656,
      "loss": 0.6703,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.05648866295814514,
      "rewards/margins": 0.04632381349802017,
      "rewards/rejected": 0.010164851322770119,
      "step": 40
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.53125,
      "learning_rate": 4.919948367622307e-06,
      "logits/chosen": -2.826796054840088,
      "logits/rejected": -2.8314404487609863,
      "logps/chosen": -20.64249610900879,
      "logps/rejected": -46.383785247802734,
      "loss": 0.6438,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.09788469970226288,
      "rewards/margins": 0.10135942697525024,
      "rewards/rejected": -0.003474730532616377,
      "step": 50
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.57421875,
      "learning_rate": 4.8266584586307555e-06,
      "logits/chosen": -2.835439682006836,
      "logits/rejected": -2.8375213146209717,
      "logps/chosen": -19.745527267456055,
      "logps/rejected": -48.08259582519531,
      "loss": 0.6272,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.10895518958568573,
      "rewards/margins": 0.13693755865097046,
      "rewards/rejected": -0.02798236906528473,
      "step": 60
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.80859375,
      "learning_rate": 4.699440630133794e-06,
      "logits/chosen": -2.8537039756774902,
      "logits/rejected": -2.8549795150756836,
      "logps/chosen": -16.767593383789062,
      "logps/rejected": -53.961273193359375,
      "loss": 0.5912,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.13634786009788513,
      "rewards/margins": 0.2162630558013916,
      "rewards/rejected": -0.07991518080234528,
      "step": 70
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.7109375,
      "learning_rate": 4.5401500045405126e-06,
      "logits/chosen": -2.7856647968292236,
      "logits/rejected": -2.7866556644439697,
      "logps/chosen": -10.72642993927002,
      "logps/rejected": -63.57000732421875,
      "loss": 0.522,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.20102627575397491,
      "rewards/margins": 0.37968239188194275,
      "rewards/rejected": -0.17865613102912903,
      "step": 80
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.73828125,
      "learning_rate": 4.351109397863309e-06,
      "logits/chosen": -2.784008026123047,
      "logits/rejected": -2.7872986793518066,
      "logps/chosen": -4.303569793701172,
      "logps/rejected": -73.54255676269531,
      "loss": 0.4602,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.26261404156684875,
      "rewards/margins": 0.5405046343803406,
      "rewards/rejected": -0.277890682220459,
      "step": 90
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.671875,
      "learning_rate": 4.135075447829912e-06,
      "logits/chosen": -2.7918694019317627,
      "logits/rejected": -2.797271490097046,
      "logps/chosen": -2.3328211307525635,
      "logps/rejected": -88.87330627441406,
      "loss": 0.3995,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2833644449710846,
      "rewards/margins": 0.7134484052658081,
      "rewards/rejected": -0.4300839304924011,
      "step": 100
    },
    {
      "epoch": 0.35,
      "eval_logits/chosen": -2.780285358428955,
      "eval_logits/rejected": -2.7814674377441406,
      "eval_logps/chosen": -1.9368014335632324,
      "eval_logps/rejected": -97.03541564941406,
      "eval_loss": 0.3704952299594879,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 0.28558260202407837,
      "eval_rewards/margins": 0.8057959675788879,
      "eval_rewards/rejected": -0.5202134251594543,
      "eval_runtime": 59.0118,
      "eval_samples_per_second": 4.897,
      "eval_steps_per_second": 2.457,
      "step": 100
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.87890625,
      "learning_rate": 3.895198415897896e-06,
      "logits/chosen": -2.757932186126709,
      "logits/rejected": -2.75920033454895,
      "logps/chosen": -1.1944591999053955,
      "logps/rejected": -102.75828552246094,
      "loss": 0.3536,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.29440590739250183,
      "rewards/margins": 0.8625706434249878,
      "rewards/rejected": -0.5681648254394531,
      "step": 110
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.5546875,
      "learning_rate": 3.634976249348867e-06,
      "logits/chosen": -2.7818870544433594,
      "logits/rejected": -2.7803432941436768,
      "logps/chosen": -1.4344559907913208,
      "logps/rejected": -116.496337890625,
      "loss": 0.3157,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2936590015888214,
      "rewards/margins": 0.9929488897323608,
      "rewards/rejected": -0.6992899179458618,
      "step": 120
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.52734375,
      "learning_rate": 3.3582035733403963e-06,
      "logits/chosen": -2.784778594970703,
      "logits/rejected": -2.78275465965271,
      "logps/chosen": -1.0116920471191406,
      "logps/rejected": -123.45560455322266,
      "loss": 0.2951,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2951065003871918,
      "rewards/margins": 1.0736355781555176,
      "rewards/rejected": -0.7785290479660034,
      "step": 130
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.47265625,
      "learning_rate": 3.068916356726475e-06,
      "logits/chosen": -2.732774019241333,
      "logits/rejected": -2.730987071990967,
      "logps/chosen": -0.8254967927932739,
      "logps/rejected": -134.45706176757812,
      "loss": 0.2701,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.29469770193099976,
      "rewards/margins": 1.1738824844360352,
      "rewards/rejected": -0.8791847229003906,
      "step": 140
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.451171875,
      "learning_rate": 2.771333058543416e-06,
      "logits/chosen": -2.7359633445739746,
      "logits/rejected": -2.7314648628234863,
      "logps/chosen": -0.7030662894248962,
      "logps/rejected": -138.85061645507812,
      "loss": 0.2582,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.29896482825279236,
      "rewards/margins": 1.2272465229034424,
      "rewards/rejected": -0.9282817840576172,
      "step": 150
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.37109375,
      "learning_rate": 2.4697931133779566e-06,
      "logits/chosen": -2.760554790496826,
      "logits/rejected": -2.754014492034912,
      "logps/chosen": -4.322839736938477,
      "logps/rejected": -146.63290405273438,
      "loss": 0.2537,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.26600345969200134,
      "rewards/margins": 1.265166163444519,
      "rewards/rejected": -0.9991627931594849,
      "step": 160
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.5,
      "learning_rate": 2.168693652639432e-06,
      "logits/chosen": -2.749978542327881,
      "logits/rejected": -2.7451446056365967,
      "logps/chosen": -0.6626571416854858,
      "logps/rejected": -147.86549377441406,
      "loss": 0.2367,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.3002447485923767,
      "rewards/margins": 1.324704647064209,
      "rewards/rejected": -1.0244598388671875,
      "step": 170
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.50390625,
      "learning_rate": 1.872425384482346e-06,
      "logits/chosen": -2.7601447105407715,
      "logits/rejected": -2.7531185150146484,
      "logps/chosen": -0.5638567209243774,
      "logps/rejected": -151.16998291015625,
      "loss": 0.2307,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.30341148376464844,
      "rewards/margins": 1.3561573028564453,
      "rewards/rejected": -1.0527456998825073,
      "step": 180
    },
    {
      "epoch": 0.66,
      "grad_norm": 3.484375,
      "learning_rate": 1.5853085673944695e-06,
      "logits/chosen": -2.7549314498901367,
      "logits/rejected": -2.748093366622925,
      "logps/chosen": -1.0212732553482056,
      "logps/rejected": -151.91207885742188,
      "loss": 0.2298,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2937738597393036,
      "rewards/margins": 1.36028254032135,
      "rewards/rejected": -1.0665085315704346,
      "step": 190
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.37109375,
      "learning_rate": 1.3115300110997097e-06,
      "logits/chosen": -2.73872447013855,
      "logits/rejected": -2.734360933303833,
      "logps/chosen": -0.8876104354858398,
      "logps/rejected": -154.93753051757812,
      "loss": 0.2236,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2938212454319,
      "rewards/margins": 1.3927398920059204,
      "rewards/rejected": -1.0989186763763428,
      "step": 200
    },
    {
      "epoch": 0.69,
      "eval_logits/chosen": -2.7480151653289795,
      "eval_logits/rejected": -2.741860866546631,
      "eval_logps/chosen": -0.6285810470581055,
      "eval_logps/rejected": -154.59251403808594,
      "eval_loss": 0.22360171377658844,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 0.29866480827331543,
      "eval_rewards/margins": 1.3944491147994995,
      "eval_rewards/rejected": -1.095784306526184,
      "eval_runtime": 58.9766,
      "eval_samples_per_second": 4.9,
      "eval_steps_per_second": 2.459,
      "step": 200
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.56640625,
      "learning_rate": 1.0550820234444627e-06,
      "logits/chosen": -2.7622146606445312,
      "logits/rejected": -2.7575910091400146,
      "logps/chosen": -0.2955681085586548,
      "logps/rejected": -156.5863494873047,
      "loss": 0.2204,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.3056582808494568,
      "rewards/margins": 1.4064788818359375,
      "rewards/rejected": -1.1008204221725464,
      "step": 210
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.3125,
      "learning_rate": 8.197041935593181e-07,
      "logits/chosen": -2.7583320140838623,
      "logits/rejected": -2.7535462379455566,
      "logps/chosen": -0.26544955372810364,
      "logps/rejected": -156.95286560058594,
      "loss": 0.2193,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.30188480019569397,
      "rewards/margins": 1.4142131805419922,
      "rewards/rejected": -1.112328290939331,
      "step": 220
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.373046875,
      "learning_rate": 6.088288602287159e-07,
      "logits/chosen": -2.723353385925293,
      "logits/rejected": -2.71772837638855,
      "logps/chosen": -0.17958000302314758,
      "logps/rejected": -157.92813110351562,
      "loss": 0.2166,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.3015274405479431,
      "rewards/margins": 1.4261300563812256,
      "rewards/rejected": -1.1246027946472168,
      "step": 230
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.6328125,
      "learning_rate": 4.255310606625124e-07,
      "logits/chosen": -2.762199640274048,
      "logits/rejected": -2.758517265319824,
      "logps/chosen": -0.2033136636018753,
      "logps/rejected": -158.05380249023438,
      "loss": 0.2164,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.3013545870780945,
      "rewards/margins": 1.4260454177856445,
      "rewards/rejected": -1.1246907711029053,
      "step": 240
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.478515625,
      "learning_rate": 2.7248368952908055e-07,
      "logits/chosen": -2.7318742275238037,
      "logits/rejected": -2.724575996398926,
      "logps/chosen": -0.2196466624736786,
      "logps/rejected": -155.65301513671875,
      "loss": 0.2205,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.3047715723514557,
      "rewards/margins": 1.406855583190918,
      "rewards/rejected": -1.1020839214324951,
      "step": 250
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.4453125,
      "learning_rate": 1.5191852213221198e-07,
      "logits/chosen": -2.7768027782440186,
      "logits/rejected": -2.7685720920562744,
      "logps/chosen": -1.5333685874938965,
      "logps/rejected": -156.28958129882812,
      "loss": 0.2237,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.29162219166755676,
      "rewards/margins": 1.3938542604446411,
      "rewards/rejected": -1.1022322177886963,
      "step": 260
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.345703125,
      "learning_rate": 6.559367010166629e-08,
      "logits/chosen": -2.699760675430298,
      "logits/rejected": -2.696230411529541,
      "logps/chosen": -0.23872259259223938,
      "logps/rejected": -159.244140625,
      "loss": 0.2139,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.3005048334598541,
      "rewards/margins": 1.4419862031936646,
      "rewards/rejected": -1.1414815187454224,
      "step": 270
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.400390625,
      "learning_rate": 1.4767944166687032e-08,
      "logits/chosen": -2.699716091156006,
      "logits/rejected": -2.6911072731018066,
      "logps/chosen": -0.23203042149543762,
      "logps/rejected": -159.73312377929688,
      "loss": 0.2128,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.3060615360736847,
      "rewards/margins": 1.44607675075531,
      "rewards/rejected": -1.1400153636932373,
      "step": 280
    },
    {
      "epoch": 1.0,
      "step": 289,
      "total_flos": 0.0,
      "train_loss": 0.36633619635162884,
      "train_runtime": 628.6241,
      "train_samples_per_second": 1.841,
      "train_steps_per_second": 0.46
    }
  ],
  "logging_steps": 10,
  "max_steps": 289,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}