{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 292,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 7.594140625,
      "epoch": 0.03432003432003432,
      "grad_norm": 86.33226300367427,
      "kl": 6.39765625,
      "learning_rate": 1.5e-07,
      "logits/chosen": -107184128.0,
      "logits/rejected": -105178726.4,
      "logps/chosen": -9729.8,
      "logps/rejected": -10091.2,
      "loss": 0.4985076904296875,
      "num_tokens": 1000098.0,
      "rewards/chosen": -0.04769744873046875,
      "rewards/margins": 0.019524288177490235,
      "rewards/rejected": -0.06722173690795899,
      "step": 10
    },
    {
      "entropy": 7.608984375,
      "epoch": 0.06864006864006864,
      "grad_norm": 127.51161758392736,
      "kl": 9.834375,
      "learning_rate": 3.166666666666666e-07,
      "logits/chosen": -111175270.4,
      "logits/rejected": -110572339.2,
      "logps/chosen": -10525.6,
      "logps/rejected": -10866.0,
      "loss": 0.5047119140625,
      "num_tokens": 2059302.0,
      "rewards/chosen": -0.09103813171386718,
      "rewards/margins": -0.06356315612792969,
      "rewards/rejected": -0.0274749755859375,
      "step": 20
    },
    {
      "entropy": 7.582421875,
      "epoch": 0.10296010296010295,
      "grad_norm": 92.88485309688424,
      "kl": 0.2046875,
      "learning_rate": 4.833333333333333e-07,
      "logits/chosen": -106351820.8,
      "logits/rejected": -105293414.4,
      "logps/chosen": -10041.6,
      "logps/rejected": -10261.0,
      "loss": 0.49964599609375,
      "num_tokens": 3087752.0,
      "rewards/chosen": -0.5374847412109375,
      "rewards/margins": 0.03810577392578125,
      "rewards/rejected": -0.5755905151367188,
      "step": 30
    },
    {
      "entropy": 7.533984375,
      "epoch": 0.13728013728013727,
      "grad_norm": 61.54426686277499,
      "kl": 0.0,
      "learning_rate": 4.985456442051682e-07,
      "logits/chosen": -99627827.2,
      "logits/rejected": -95980748.8,
      "logps/chosen": -9881.0,
      "logps/rejected": -10339.0,
      "loss": 0.5043426513671875,
      "num_tokens": 4100561.0,
      "rewards/chosen": -1.321875,
      "rewards/margins": 0.0261962890625,
      "rewards/rejected": -1.3480712890625,
      "step": 40
    },
    {
      "entropy": 7.523828125,
      "epoch": 0.1716001716001716,
      "grad_norm": 122.84522486033784,
      "kl": 0.09375,
      "learning_rate": 4.935399618791793e-07,
      "logits/chosen": -98726707.2,
      "logits/rejected": -96527974.4,
      "logps/chosen": -9919.9,
      "logps/rejected": -10279.2,
      "loss": 0.4984130859375,
      "num_tokens": 5111368.0,
      "rewards/chosen": -1.801904296875,
      "rewards/margins": 0.1624755859375,
      "rewards/rejected": -1.9643798828125,
      "step": 50
    },
    {
      "entropy": 7.54296875,
      "epoch": 0.2059202059202059,
      "grad_norm": 75.96355510905495,
      "kl": 3.9984375,
      "learning_rate": 4.850368660819092e-07,
      "logits/chosen": -106171596.8,
      "logits/rejected": -105942220.8,
      "logps/chosen": -10453.6,
      "logps/rejected": -10825.2,
      "loss": 0.495751953125,
      "num_tokens": 6159442.0,
      "rewards/chosen": -0.5387272834777832,
      "rewards/margins": 0.12462263107299805,
      "rewards/rejected": -0.6633499145507813,
      "step": 60
    },
    {
      "entropy": 7.62265625,
      "epoch": 0.24024024024024024,
      "grad_norm": 109.08719788934083,
      "kl": 37.352734375,
      "learning_rate": 4.731584675403184e-07,
      "logits/chosen": -113573888.0,
      "logits/rejected": -110988492.8,
      "logps/chosen": -9306.4,
      "logps/rejected": -9616.2,
      "loss": 0.5021636962890625,
      "num_tokens": 7148746.0,
      "rewards/chosen": 0.21960601806640626,
      "rewards/margins": 0.003122711181640625,
      "rewards/rejected": 0.21648330688476564,
      "step": 70
    },
    {
      "entropy": 7.659765625,
      "epoch": 0.27456027456027454,
      "grad_norm": 58.939359796249065,
      "kl": 171.0875,
      "learning_rate": 4.5807534881817183e-07,
      "logits/chosen": -114412748.8,
      "logits/rejected": -113252761.6,
      "logps/chosen": -9437.6,
      "logps/rejected": -9837.6,
      "loss": 0.4959259033203125,
      "num_tokens": 8139737.0,
      "rewards/chosen": 1.7338134765625,
      "rewards/margins": 0.0449676513671875,
      "rewards/rejected": 1.6888458251953125,
      "step": 80
    },
    {
      "entropy": 7.712109375,
      "epoch": 0.3088803088803089,
      "grad_norm": 83.72781850857909,
      "kl": 330.3125,
      "learning_rate": 4.400041146246136e-07,
      "logits/chosen": -129309081.6,
      "logits/rejected": -125432627.2,
      "logps/chosen": -9531.0,
      "logps/rejected": -10017.2,
      "loss": 0.50328369140625,
      "num_tokens": 9171482.0,
      "rewards/chosen": 3.218115234375,
      "rewards/margins": -0.171630859375,
      "rewards/rejected": 3.38974609375,
      "step": 90
    },
    {
      "entropy": 7.815625,
      "epoch": 0.3432003432003432,
      "grad_norm": 61.619761951920744,
      "kl": 569.35,
      "learning_rate": 4.1920428121079e-07,
      "logits/chosen": -142609612.8,
      "logits/rejected": -140650086.4,
      "logps/chosen": -9252.6,
      "logps/rejected": -9623.2,
      "loss": 0.5056427001953125,
      "num_tokens": 10204959.0,
      "rewards/chosen": 5.598828125,
      "rewards/margins": -0.18271484375,
      "rewards/rejected": 5.78154296875,
      "step": 100
    },
    {
      "entropy": 7.884765625,
      "epoch": 0.37752037752037754,
      "grad_norm": 45.03461918952044,
      "kl": 831.45,
      "learning_rate": 3.959745495250818e-07,
      "logits/chosen": -151001497.6,
      "logits/rejected": -147940966.4,
      "logps/chosen": -8605.2,
      "logps/rejected": -9020.0,
      "loss": 0.50517578125,
      "num_tokens": 11188498.0,
      "rewards/chosen": 8.1416015625,
      "rewards/margins": -0.340234375,
      "rewards/rejected": 8.4818359375,
      "step": 110
    },
    {
      "entropy": 7.975390625,
      "epoch": 0.4118404118404118,
      "grad_norm": 36.350000413332154,
      "kl": 1201.625,
      "learning_rate": 3.7064851564718353e-07,
      "logits/chosen": -181908275.2,
      "logits/rejected": -178755993.6,
      "logps/chosen": -8662.0,
      "logps/rejected": -8825.2,
      "loss": 0.496246337890625,
      "num_tokens": 12220511.0,
      "rewards/chosen": 12.016015625,
      "rewards/margins": -0.000390625,
      "rewards/rejected": 12.01640625,
      "step": 120
    },
    {
      "entropy": 7.977734375,
      "epoch": 0.44616044616044614,
      "grad_norm": 31.597257913649155,
      "kl": 1435.55,
      "learning_rate": 3.4358988010236103e-07,
      "logits/chosen": -203227136.0,
      "logits/rejected": -201883648.0,
      "logps/chosen": -8452.6,
      "logps/rejected": -8621.8,
      "loss": 0.4972412109375,
      "num_tokens": 13250325.0,
      "rewards/chosen": 14.191796875,
      "rewards/margins": -0.3341796875,
      "rewards/rejected": 14.5259765625,
      "step": 130
    },
    {
      "entropy": 7.975390625,
      "epoch": 0.4804804804804805,
      "grad_norm": 21.621622249090635,
      "kl": 1731.0,
      "learning_rate": 3.1518722485366754e-07,
      "logits/chosen": -222298112.0,
      "logits/rejected": -219827404.8,
      "logps/chosen": -8690.0,
      "logps/rejected": -8949.2,
      "loss": 0.50311279296875,
      "num_tokens": 14326309.0,
      "rewards/chosen": 16.971484375,
      "rewards/margins": -0.696484375,
      "rewards/rejected": 17.66796875,
      "step": 140
    },
    {
      "entropy": 7.954296875,
      "epoch": 0.5148005148005148,
      "grad_norm": 21.811975809486956,
      "kl": 1867.15,
      "learning_rate": 2.8584843297836277e-07,
      "logits/chosen": -228825497.6,
      "logits/rejected": -228071833.6,
      "logps/chosen": -7983.8,
      "logps/rejected": -8123.0,
      "loss": 0.4952362060546875,
      "num_tokens": 15346427.0,
      "rewards/chosen": 18.575390625,
      "rewards/margins": -0.193359375,
      "rewards/rejected": 18.76875,
      "step": 150
    },
    {
      "entropy": 7.976953125,
      "epoch": 0.5491205491205491,
      "grad_norm": 33.183727711509015,
      "kl": 1883.8,
      "learning_rate": 2.5599483116609544e-07,
      "logits/chosen": -228923801.6,
      "logits/rejected": -224670515.2,
      "logps/chosen": -8062.4,
      "logps/rejected": -8387.2,
      "loss": 0.5054046630859375,
      "num_tokens": 16366642.0,
      "rewards/chosen": 18.32421875,
      "rewards/margins": -1.030078125,
      "rewards/rejected": 19.354296875,
      "step": 160
    },
    {
      "entropy": 8.000390625,
      "epoch": 0.5834405834405835,
      "grad_norm": 12.850165739096068,
      "kl": 1952.45,
      "learning_rate": 2.2605513915689874e-07,
      "logits/chosen": -230529433.6,
      "logits/rejected": -228432281.6,
      "logps/chosen": -7903.6,
      "logps/rejected": -8130.6,
      "loss": 0.50506591796875,
      "num_tokens": 17380939.0,
      "rewards/chosen": 19.238671875,
      "rewards/margins": -0.577734375,
      "rewards/rejected": 19.81640625,
      "step": 170
    },
    {
      "entropy": 7.98984375,
      "epoch": 0.6177606177606177,
      "grad_norm": 9.466663342172486,
      "kl": 2113.5,
      "learning_rate": 1.9645931300952795e-07,
      "logits/chosen": -245137408.0,
      "logits/rejected": -241355980.8,
      "logps/chosen": -7811.6,
      "logps/rejected": -8251.0,
      "loss": 0.5172149658203125,
      "num_tokens": 18421034.0,
      "rewards/chosen": 20.554296875,
      "rewards/margins": -1.14609375,
      "rewards/rejected": 21.700390625,
      "step": 180
    },
    {
      "entropy": 7.917578125,
      "epoch": 0.6520806520806521,
      "grad_norm": 8.76687486946907,
      "kl": 2130.25,
      "learning_rate": 1.6763237061535008e-07,
      "logits/chosen": -243918438.4,
      "logits/rejected": -240589209.6,
      "logps/chosen": -7463.2,
      "logps/rejected": -7685.8,
      "loss": 0.500433349609375,
      "num_tokens": 19430090.0,
      "rewards/chosen": 20.763671875,
      "rewards/margins": -1.073828125,
      "rewards/rejected": 21.8375,
      "step": 190
    },
    {
      "entropy": 7.755078125,
      "epoch": 0.6864006864006864,
      "grad_norm": 11.01172644517593,
      "kl": 2340.1,
      "learning_rate": 1.3998828812795932e-07,
      "logits/chosen": -260092723.2,
      "logits/rejected": -258008678.4,
      "logps/chosen": -7685.4,
      "logps/rejected": -7937.2,
      "loss": 0.5029571533203125,
      "num_tokens": 20459910.0,
      "rewards/chosen": 22.9265625,
      "rewards/margins": -0.95234375,
      "rewards/rejected": 23.87890625,
      "step": 200
    },
    {
      "entropy": 7.675390625,
      "epoch": 0.7207207207207207,
      "grad_norm": 30.85091198701311,
      "kl": 2452.7,
      "learning_rate": 1.1392405496029076e-07,
      "logits/chosen": -270807859.2,
      "logits/rejected": -269208780.8,
      "logps/chosen": -7861.9,
      "logps/rejected": -8111.2,
      "loss": 0.5022857666015625,
      "num_tokens": 21493036.0,
      "rewards/chosen": 23.9083984375,
      "rewards/margins": -1.2330078125,
      "rewards/rejected": 25.14140625,
      "step": 210
    },
    {
      "entropy": 7.585546875,
      "epoch": 0.7550407550407551,
      "grad_norm": 15.36265057914587,
      "kl": 2365.5,
      "learning_rate": 8.981397272385738e-08,
      "logits/chosen": -266141696.0,
      "logits/rejected": -263979008.0,
      "logps/chosen": -7420.0,
      "logps/rejected": -7809.0,
      "loss": 0.510125732421875,
      "num_tokens": 22497542.0,
      "rewards/chosen": 23.06484375,
      "rewards/margins": -1.197265625,
      "rewards/rejected": 24.262109375,
      "step": 220
    },
    {
      "entropy": 7.594140625,
      "epoch": 0.7893607893607893,
      "grad_norm": 6.546387154998616,
      "kl": 2397.8,
      "learning_rate": 6.800427998154968e-08,
      "logits/chosen": -275847577.6,
      "logits/rejected": -274078105.6,
      "logps/chosen": -7580.4,
      "logps/rejected": -7842.4,
      "loss": 0.4974273681640625,
      "num_tokens": 23534012.0,
      "rewards/chosen": 23.476953125,
      "rewards/margins": -0.99921875,
      "rewards/rejected": 24.476171875,
      "step": 230
    },
    {
      "entropy": 7.50078125,
      "epoch": 0.8236808236808236,
      "grad_norm": 24.903868525956547,
      "kl": 2610.4,
      "learning_rate": 4.8808180006509356e-08,
      "logits/chosen": -296511078.4,
      "logits/rejected": -293443993.6,
      "logps/chosen": -8122.0,
      "logps/rejected": -8423.6,
      "loss": 0.50238037109375,
      "num_tokens": 24628737.0,
      "rewards/chosen": 25.5359375,
      "rewards/margins": -1.14375,
      "rewards/rejected": 26.6796875,
      "step": 240
    },
    {
      "entropy": 7.55234375,
      "epoch": 0.858000858000858,
      "grad_norm": 15.268579913892744,
      "kl": 2629.2,
      "learning_rate": 3.250134295213053e-08,
      "logits/chosen": -288122470.4,
      "logits/rejected": -284799795.2,
      "logps/chosen": -7932.2,
      "logps/rejected": -8312.4,
      "loss": 0.5093109130859375,
      "num_tokens": 25690351.0,
      "rewards/chosen": 25.858984375,
      "rewards/margins": -0.870703125,
      "rewards/rejected": 26.7296875,
      "step": 250
    },
    {
      "entropy": 7.44453125,
      "epoch": 0.8923208923208923,
      "grad_norm": 15.110634441696666,
      "kl": 2613.15,
      "learning_rate": 1.9317947025340232e-08,
      "logits/chosen": -292644454.4,
      "logits/rejected": -290822553.6,
      "logps/chosen": -8003.2,
      "logps/rejected": -8455.0,
      "loss": 0.5151123046875,
      "num_tokens": 26763232.0,
      "rewards/chosen": 25.44453125,
      "rewards/margins": -1.375,
      "rewards/rejected": 26.81953125,
      "step": 260
    },
    {
      "entropy": 7.478515625,
      "epoch": 0.9266409266409267,
      "grad_norm": 13.312675942972383,
      "kl": 2433.15,
      "learning_rate": 9.447315514833353e-09,
      "logits/chosen": -276883046.4,
      "logits/rejected": -270277017.6,
      "logps/chosen": -7410.8,
      "logps/rejected": -7717.0,
      "loss": 0.5004669189453125,
      "num_tokens": 27765238.0,
      "rewards/chosen": 23.782421875,
      "rewards/margins": -1.104296875,
      "rewards/rejected": 24.88671875,
      "step": 270
    },
    {
      "entropy": 7.510546875,
      "epoch": 0.960960960960961,
      "grad_norm": 10.274017437444142,
      "kl": 2346.1,
      "learning_rate": 3.0311979690147426e-09,
      "logits/chosen": -259797811.2,
      "logits/rejected": -256979763.2,
      "logps/chosen": -7306.8,
      "logps/rejected": -7390.2,
      "loss": 0.500830078125,
      "num_tokens": 28731039.0,
      "rewards/chosen": 23.248828125,
      "rewards/margins": -0.418359375,
      "rewards/rejected": 23.6671875,
      "step": 280
    },
    {
      "entropy": 7.46171875,
      "epoch": 0.9952809952809952,
      "grad_norm": 6.810163426768989,
      "kl": 2498.55,
      "learning_rate": 1.6173456793908135e-10,
      "logits/chosen": -279746969.6,
      "logits/rejected": -278174105.6,
      "logps/chosen": -7759.2,
      "logps/rejected": -7908.0,
      "loss": 0.5092376708984375,
      "num_tokens": 29763064.0,
      "rewards/chosen": 24.6875,
      "rewards/margins": -0.61796875,
      "rewards/rejected": 25.30546875,
      "step": 290
    },
    {
      "epoch": 1.0,
      "eval_entropy": 7.5152138157894735,
      "eval_kl": 2432.842105263158,
      "eval_logits/chosen": -280314718.31578946,
      "eval_logits/rejected": -277313859.3684211,
      "eval_logps/chosen": -7510.631578947368,
      "eval_logps/rejected": -7763.578947368421,
      "eval_loss": 0.5068327188491821,
      "eval_num_tokens": 29885865.0,
      "eval_rewards/chosen": 24.02014802631579,
      "eval_rewards/margins": -0.6217105263157895,
      "eval_rewards/rejected": 24.64185855263158,
      "eval_runtime": 43.731,
      "eval_samples_per_second": 27.623,
      "eval_steps_per_second": 1.738,
      "step": 292
    }
  ],
  "logging_steps": 10,
  "max_steps": 292,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7245703888896.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}