{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 60.0,
  "eval_steps": 500,
  "global_step": 197880,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 5.0535678188801295e-08,
      "loss": 12.7149,
      "step": 1
    },
    {
      "epoch": 0.3,
      "learning_rate": 5.0030321406913285e-05,
      "loss": 3.3584,
      "step": 990
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00010006064281382657,
      "loss": 0.3384,
      "step": 1980
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00015009096422073984,
      "loss": 0.2236,
      "step": 2970
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.11203870922327042,
      "eval_max_distance": 133,
      "eval_mean_distance": 5,
      "eval_runtime": 0.5965,
      "eval_samples_per_second": 83.828,
      "eval_steps_per_second": 3.353,
      "step": 3298
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00020012128562765314,
      "loss": 0.1679,
      "step": 3960
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002501516070345664,
      "loss": 0.1395,
      "step": 4950
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0003001819284414797,
      "loss": 0.1179,
      "step": 5940
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.05475025996565819,
      "eval_max_distance": 82,
      "eval_mean_distance": 3,
      "eval_runtime": 0.5422,
      "eval_samples_per_second": 92.223,
      "eval_steps_per_second": 3.689,
      "step": 6596
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0003502122498483929,
      "loss": 0.1022,
      "step": 6930
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.0004002425712553063,
      "loss": 0.0917,
      "step": 7920
    },
    {
      "epoch": 2.7,
      "learning_rate": 0.0004502728926622195,
      "loss": 0.0829,
      "step": 8910
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.042510777711868286,
      "eval_max_distance": 46,
      "eval_mean_distance": 1,
      "eval_runtime": 0.5158,
      "eval_samples_per_second": 96.928,
      "eval_steps_per_second": 3.877,
      "step": 9894
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0005003032140691328,
      "loss": 0.0769,
      "step": 9900
    },
    {
      "epoch": 3.3,
      "learning_rate": 0.0005503335354760462,
      "loss": 0.0667,
      "step": 10890
    },
    {
      "epoch": 3.6,
      "learning_rate": 0.0006003638568829594,
      "loss": 0.0653,
      "step": 11880
    },
    {
      "epoch": 3.9,
      "learning_rate": 0.0006503941782898727,
      "loss": 0.0643,
      "step": 12870
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.03110930137336254,
      "eval_max_distance": 64,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4848,
      "eval_samples_per_second": 103.129,
      "eval_steps_per_second": 4.125,
      "step": 13192
    },
    {
      "epoch": 4.2,
      "learning_rate": 0.0007004244996967858,
      "loss": 0.0589,
      "step": 13860
    },
    {
      "epoch": 4.5,
      "learning_rate": 0.0007504548211036993,
      "loss": 0.0549,
      "step": 14850
    },
    {
      "epoch": 4.8,
      "learning_rate": 0.0008004851425106126,
      "loss": 0.0538,
      "step": 15840
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.026651622727513313,
      "eval_max_distance": 48,
      "eval_mean_distance": 1,
      "eval_runtime": 0.5057,
      "eval_samples_per_second": 98.878,
      "eval_steps_per_second": 3.955,
      "step": 16490
    },
    {
      "epoch": 5.1,
      "learning_rate": 0.0008505154639175257,
      "loss": 0.048,
      "step": 16830
    },
    {
      "epoch": 5.4,
      "learning_rate": 0.000900545785324439,
      "loss": 0.0461,
      "step": 17820
    },
    {
      "epoch": 5.7,
      "learning_rate": 0.0009505761067313523,
      "loss": 0.0469,
      "step": 18810
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.039574604481458664,
      "eval_max_distance": 80,
      "eval_mean_distance": 2,
      "eval_runtime": 0.5179,
      "eval_samples_per_second": 96.548,
      "eval_steps_per_second": 3.862,
      "step": 19788
    },
    {
      "epoch": 6.0,
      "learning_rate": 0.0009999326190957482,
      "loss": 0.0464,
      "step": 19800
    },
    {
      "epoch": 6.3,
      "learning_rate": 0.0009943736944949802,
      "loss": 0.0393,
      "step": 20790
    },
    {
      "epoch": 6.6,
      "learning_rate": 0.000988814769894212,
      "loss": 0.0426,
      "step": 21780
    },
    {
      "epoch": 6.9,
      "learning_rate": 0.000983255845293444,
      "loss": 0.0385,
      "step": 22770
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.026188833639025688,
      "eval_max_distance": 73,
      "eval_mean_distance": 2,
      "eval_runtime": 0.4896,
      "eval_samples_per_second": 102.115,
      "eval_steps_per_second": 4.085,
      "step": 23086
    },
    {
      "epoch": 7.2,
      "learning_rate": 0.0009776969206926756,
      "loss": 0.034,
      "step": 23760
    },
    {
      "epoch": 7.5,
      "learning_rate": 0.0009721379960919076,
      "loss": 0.0315,
      "step": 24750
    },
    {
      "epoch": 7.8,
      "learning_rate": 0.0009665790714911395,
      "loss": 0.0316,
      "step": 25740
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.02234221063554287,
      "eval_max_distance": 40,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4837,
      "eval_samples_per_second": 103.365,
      "eval_steps_per_second": 4.135,
      "step": 26384
    },
    {
      "epoch": 8.1,
      "learning_rate": 0.0009610201468903713,
      "loss": 0.0305,
      "step": 26730
    },
    {
      "epoch": 8.41,
      "learning_rate": 0.0009554612222896032,
      "loss": 0.0271,
      "step": 27720
    },
    {
      "epoch": 8.71,
      "learning_rate": 0.0009499022976888349,
      "loss": 0.0263,
      "step": 28710
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.023996921256184578,
      "eval_max_distance": 69,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4894,
      "eval_samples_per_second": 102.167,
      "eval_steps_per_second": 4.087,
      "step": 29682
    },
    {
      "epoch": 9.01,
      "learning_rate": 0.0009443433730880669,
      "loss": 0.0282,
      "step": 29700
    },
    {
      "epoch": 9.31,
      "learning_rate": 0.0009387844484872987,
      "loss": 0.0229,
      "step": 30690
    },
    {
      "epoch": 9.61,
      "learning_rate": 0.0009332255238865306,
      "loss": 0.0226,
      "step": 31680
    },
    {
      "epoch": 9.91,
      "learning_rate": 0.0009276665992857625,
      "loss": 0.0226,
      "step": 32670
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.02030733972787857,
      "eval_max_distance": 60,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4797,
      "eval_samples_per_second": 104.236,
      "eval_steps_per_second": 4.169,
      "step": 32980
    },
    {
      "epoch": 10.21,
      "learning_rate": 0.0009221076746849943,
      "loss": 0.0209,
      "step": 33660
    },
    {
      "epoch": 10.51,
      "learning_rate": 0.0009165487500842261,
      "loss": 0.02,
      "step": 34650
    },
    {
      "epoch": 10.81,
      "learning_rate": 0.000910989825483458,
      "loss": 0.0203,
      "step": 35640
    },
    {
      "epoch": 11.0,
      "eval_loss": 0.017732510343194008,
      "eval_max_distance": 54,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4814,
      "eval_samples_per_second": 103.858,
      "eval_steps_per_second": 4.154,
      "step": 36278
    },
    {
      "epoch": 11.11,
      "learning_rate": 0.0009054309008826899,
      "loss": 0.0183,
      "step": 36630
    },
    {
      "epoch": 11.41,
      "learning_rate": 0.0008998719762819217,
      "loss": 0.0174,
      "step": 37620
    },
    {
      "epoch": 11.71,
      "learning_rate": 0.0008943130516811536,
      "loss": 0.0178,
      "step": 38610
    },
    {
      "epoch": 12.0,
      "eval_loss": 0.018777821213006973,
      "eval_max_distance": 61,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4893,
      "eval_samples_per_second": 102.185,
      "eval_steps_per_second": 4.087,
      "step": 39576
    },
    {
      "epoch": 12.01,
      "learning_rate": 0.0008887541270803853,
      "loss": 0.0174,
      "step": 39600
    },
    {
      "epoch": 12.31,
      "learning_rate": 0.0008831952024796173,
      "loss": 0.0153,
      "step": 40590
    },
    {
      "epoch": 12.61,
      "learning_rate": 0.0008776362778788492,
      "loss": 0.015,
      "step": 41580
    },
    {
      "epoch": 12.91,
      "learning_rate": 0.000872077353278081,
      "loss": 0.0154,
      "step": 42570
    },
    {
      "epoch": 13.0,
      "eval_loss": 0.029613599181175232,
      "eval_max_distance": 65,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4669,
      "eval_samples_per_second": 107.079,
      "eval_steps_per_second": 4.283,
      "step": 42874
    },
    {
      "epoch": 13.21,
      "learning_rate": 0.0008665184286773129,
      "loss": 0.014,
      "step": 43560
    },
    {
      "epoch": 13.51,
      "learning_rate": 0.0008609595040765447,
      "loss": 0.0135,
      "step": 44550
    },
    {
      "epoch": 13.81,
      "learning_rate": 0.0008554005794757766,
      "loss": 0.0138,
      "step": 45540
    },
    {
      "epoch": 14.0,
      "eval_loss": 0.02011469565331936,
      "eval_max_distance": 55,
      "eval_mean_distance": 1,
      "eval_runtime": 0.5034,
      "eval_samples_per_second": 99.332,
      "eval_steps_per_second": 3.973,
      "step": 46172
    },
    {
      "epoch": 14.11,
      "learning_rate": 0.0008498416548750084,
      "loss": 0.0128,
      "step": 46530
    },
    {
      "epoch": 14.41,
      "learning_rate": 0.0008442827302742403,
      "loss": 0.0121,
      "step": 47520
    },
    {
      "epoch": 14.71,
      "learning_rate": 0.0008387238056734722,
      "loss": 0.012,
      "step": 48510
    },
    {
      "epoch": 15.0,
      "eval_loss": 0.026753582060337067,
      "eval_max_distance": 67,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4716,
      "eval_samples_per_second": 106.031,
      "eval_steps_per_second": 4.241,
      "step": 49470
    },
    {
      "epoch": 15.01,
      "learning_rate": 0.000833164881072704,
      "loss": 0.0123,
      "step": 49500
    },
    {
      "epoch": 15.31,
      "learning_rate": 0.0008276059564719359,
      "loss": 0.0104,
      "step": 50490
    },
    {
      "epoch": 15.61,
      "learning_rate": 0.0008220470318711677,
      "loss": 0.0109,
      "step": 51480
    },
    {
      "epoch": 15.91,
      "learning_rate": 0.0008164881072703996,
      "loss": 0.0109,
      "step": 52470
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.01633359119296074,
      "eval_max_distance": 35,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4971,
      "eval_samples_per_second": 100.579,
      "eval_steps_per_second": 4.023,
      "step": 52768
    },
    {
      "epoch": 16.21,
      "learning_rate": 0.0008109291826696314,
      "loss": 0.0098,
      "step": 53460
    },
    {
      "epoch": 16.51,
      "learning_rate": 0.0008053702580688633,
      "loss": 0.0094,
      "step": 54450
    },
    {
      "epoch": 16.81,
      "learning_rate": 0.0007998113334680952,
      "loss": 0.0105,
      "step": 55440
    },
    {
      "epoch": 17.0,
      "eval_loss": 0.013592842034995556,
      "eval_max_distance": 26,
      "eval_mean_distance": 1,
      "eval_runtime": 0.48,
      "eval_samples_per_second": 104.157,
      "eval_steps_per_second": 4.166,
      "step": 56066
    },
    {
      "epoch": 17.11,
      "learning_rate": 0.000794252408867327,
      "loss": 0.0097,
      "step": 56430
    },
    {
      "epoch": 17.41,
      "learning_rate": 0.0007886934842665589,
      "loss": 0.0083,
      "step": 57420
    },
    {
      "epoch": 17.71,
      "learning_rate": 0.0007831345596657907,
      "loss": 0.0092,
      "step": 58410
    },
    {
      "epoch": 18.0,
      "eval_loss": 0.020196767523884773,
      "eval_max_distance": 65,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4567,
      "eval_samples_per_second": 109.487,
      "eval_steps_per_second": 4.379,
      "step": 59364
    },
    {
      "epoch": 18.01,
      "learning_rate": 0.0007775756350650226,
      "loss": 0.009,
      "step": 59400
    },
    {
      "epoch": 18.31,
      "learning_rate": 0.0007720167104642545,
      "loss": 0.0075,
      "step": 60390
    },
    {
      "epoch": 18.61,
      "learning_rate": 0.0007664577858634864,
      "loss": 0.0078,
      "step": 61380
    },
    {
      "epoch": 18.91,
      "learning_rate": 0.0007608988612627181,
      "loss": 0.0087,
      "step": 62370
    },
    {
      "epoch": 19.0,
      "eval_loss": 0.02213277295231819,
      "eval_max_distance": 65,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4707,
      "eval_samples_per_second": 106.233,
      "eval_steps_per_second": 4.249,
      "step": 62662
    },
    {
      "epoch": 19.21,
      "learning_rate": 0.00075533993666195,
      "loss": 0.0077,
      "step": 63360
    },
    {
      "epoch": 19.51,
      "learning_rate": 0.0007497810120611818,
      "loss": 0.0071,
      "step": 64350
    },
    {
      "epoch": 19.81,
      "learning_rate": 0.0007442220874604138,
      "loss": 0.0075,
      "step": 65340
    },
    {
      "epoch": 20.0,
      "eval_loss": 0.020336275920271873,
      "eval_max_distance": 33,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4773,
      "eval_samples_per_second": 104.749,
      "eval_steps_per_second": 4.19,
      "step": 65960
    },
    {
      "epoch": 20.11,
      "learning_rate": 0.0007386631628596457,
      "loss": 0.0073,
      "step": 66330
    },
    {
      "epoch": 20.41,
      "learning_rate": 0.0007331042382588774,
      "loss": 0.0063,
      "step": 67320
    },
    {
      "epoch": 20.71,
      "learning_rate": 0.0007275453136581093,
      "loss": 0.0067,
      "step": 68310
    },
    {
      "epoch": 21.0,
      "eval_loss": 0.022562623023986816,
      "eval_max_distance": 26,
      "eval_mean_distance": 1,
      "eval_runtime": 0.5033,
      "eval_samples_per_second": 99.35,
      "eval_steps_per_second": 3.974,
      "step": 69258
    },
    {
      "epoch": 21.01,
      "learning_rate": 0.0007219863890573411,
      "loss": 0.007,
      "step": 69300
    },
    {
      "epoch": 21.31,
      "learning_rate": 0.000716427464456573,
      "loss": 0.0061,
      "step": 70290
    },
    {
      "epoch": 21.61,
      "learning_rate": 0.0007108685398558049,
      "loss": 0.006,
      "step": 71280
    },
    {
      "epoch": 21.91,
      "learning_rate": 0.0007053096152550368,
      "loss": 0.0062,
      "step": 72270
    },
    {
      "epoch": 22.0,
      "eval_loss": 0.01839238964021206,
      "eval_max_distance": 24,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4856,
      "eval_samples_per_second": 102.959,
      "eval_steps_per_second": 4.118,
      "step": 72556
    },
    {
      "epoch": 22.21,
      "learning_rate": 0.0006997506906542685,
      "loss": 0.0057,
      "step": 73260
    },
    {
      "epoch": 22.51,
      "learning_rate": 0.0006941917660535004,
      "loss": 0.0058,
      "step": 74250
    },
    {
      "epoch": 22.81,
      "learning_rate": 0.0006886328414527323,
      "loss": 0.0059,
      "step": 75240
    },
    {
      "epoch": 23.0,
      "eval_loss": 0.013111269101500511,
      "eval_max_distance": 18,
      "eval_mean_distance": 0,
      "eval_runtime": 0.5001,
      "eval_samples_per_second": 99.983,
      "eval_steps_per_second": 3.999,
      "step": 75854
    },
    {
      "epoch": 23.11,
      "learning_rate": 0.0006830739168519642,
      "loss": 0.0055,
      "step": 76230
    },
    {
      "epoch": 23.41,
      "learning_rate": 0.0006775149922511961,
      "loss": 0.0051,
      "step": 77220
    },
    {
      "epoch": 23.71,
      "learning_rate": 0.0006719560676504279,
      "loss": 0.0054,
      "step": 78210
    },
    {
      "epoch": 24.0,
      "eval_loss": 0.026959825307130814,
      "eval_max_distance": 58,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4725,
      "eval_samples_per_second": 105.825,
      "eval_steps_per_second": 4.233,
      "step": 79152
    },
    {
      "epoch": 24.01,
      "learning_rate": 0.0006663971430496597,
      "loss": 0.0055,
      "step": 79200
    },
    {
      "epoch": 24.31,
      "learning_rate": 0.0006608382184488915,
      "loss": 0.0046,
      "step": 80190
    },
    {
      "epoch": 24.61,
      "learning_rate": 0.0006552792938481235,
      "loss": 0.005,
      "step": 81180
    },
    {
      "epoch": 24.92,
      "learning_rate": 0.0006497203692473554,
      "loss": 0.0052,
      "step": 82170
    },
    {
      "epoch": 25.0,
      "eval_loss": 0.024379713460803032,
      "eval_max_distance": 45,
      "eval_mean_distance": 1,
      "eval_runtime": 0.47,
      "eval_samples_per_second": 106.387,
      "eval_steps_per_second": 4.255,
      "step": 82450
    },
    {
      "epoch": 25.22,
      "learning_rate": 0.0006441614446465872,
      "loss": 0.0048,
      "step": 83160
    },
    {
      "epoch": 25.52,
      "learning_rate": 0.000638602520045819,
      "loss": 0.0045,
      "step": 84150
    },
    {
      "epoch": 25.82,
      "learning_rate": 0.0006330435954450508,
      "loss": 0.0044,
      "step": 85140
    },
    {
      "epoch": 26.0,
      "eval_loss": 0.014908027835190296,
      "eval_max_distance": 23,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4819,
      "eval_samples_per_second": 103.748,
      "eval_steps_per_second": 4.15,
      "step": 85748
    },
    {
      "epoch": 26.12,
      "learning_rate": 0.0006274846708442828,
      "loss": 0.0044,
      "step": 86130
    },
    {
      "epoch": 26.42,
      "learning_rate": 0.0006219257462435146,
      "loss": 0.0042,
      "step": 87120
    },
    {
      "epoch": 26.72,
      "learning_rate": 0.0006163668216427465,
      "loss": 0.0043,
      "step": 88110
    },
    {
      "epoch": 27.0,
      "eval_loss": 0.0256387647241354,
      "eval_max_distance": 63,
      "eval_mean_distance": 1,
      "eval_runtime": 0.5104,
      "eval_samples_per_second": 97.954,
      "eval_steps_per_second": 3.918,
      "step": 89046
    },
    {
      "epoch": 27.02,
      "learning_rate": 0.0006108078970419783,
      "loss": 0.0043,
      "step": 89100
    },
    {
      "epoch": 27.32,
      "learning_rate": 0.0006052489724412101,
      "loss": 0.004,
      "step": 90090
    },
    {
      "epoch": 27.62,
      "learning_rate": 0.0005996900478404421,
      "loss": 0.0037,
      "step": 91080
    },
    {
      "epoch": 27.92,
      "learning_rate": 0.0005941311232396739,
      "loss": 0.0038,
      "step": 92070
    },
    {
      "epoch": 28.0,
      "eval_loss": 0.017227506265044212,
      "eval_max_distance": 30,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4632,
      "eval_samples_per_second": 107.934,
      "eval_steps_per_second": 4.317,
      "step": 92344
    },
    {
      "epoch": 28.22,
      "learning_rate": 0.0005885721986389058,
      "loss": 0.0037,
      "step": 93060
    },
    {
      "epoch": 28.52,
      "learning_rate": 0.0005830132740381376,
      "loss": 0.0038,
      "step": 94050
    },
    {
      "epoch": 28.82,
      "learning_rate": 0.0005774543494373694,
      "loss": 0.0036,
      "step": 95040
    },
    {
      "epoch": 29.0,
      "eval_loss": 0.022354494780302048,
      "eval_max_distance": 37,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4846,
      "eval_samples_per_second": 103.187,
      "eval_steps_per_second": 4.127,
      "step": 95642
    },
    {
      "epoch": 29.12,
      "learning_rate": 0.0005718954248366013,
      "loss": 0.0037,
      "step": 96030
    },
    {
      "epoch": 29.42,
      "learning_rate": 0.0005663365002358332,
      "loss": 0.0033,
      "step": 97020
    },
    {
      "epoch": 29.72,
      "learning_rate": 0.000560777575635065,
      "loss": 0.0033,
      "step": 98010
    },
    {
      "epoch": 30.0,
      "eval_loss": 0.01936698891222477,
      "eval_max_distance": 30,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4829,
      "eval_samples_per_second": 103.544,
      "eval_steps_per_second": 4.142,
      "step": 98940
    },
    {
      "epoch": 30.02,
      "learning_rate": 0.0005552186510342969,
      "loss": 0.0035,
      "step": 99000
    },
    {
      "epoch": 30.32,
      "learning_rate": 0.0005496597264335288,
      "loss": 0.003,
      "step": 99990
    },
    {
      "epoch": 30.62,
      "learning_rate": 0.0005441008018327606,
      "loss": 0.0033,
      "step": 100980
    },
    {
      "epoch": 30.92,
      "learning_rate": 0.0005385418772319925,
      "loss": 0.0031,
      "step": 101970
    },
    {
      "epoch": 31.0,
      "eval_loss": 0.023793019354343414,
      "eval_max_distance": 59,
      "eval_mean_distance": 1,
      "eval_runtime": 0.5012,
      "eval_samples_per_second": 99.754,
      "eval_steps_per_second": 3.99,
      "step": 102238
    },
    {
      "epoch": 31.22,
      "learning_rate": 0.0005329829526312243,
      "loss": 0.0029,
      "step": 102960
    },
    {
      "epoch": 31.52,
      "learning_rate": 0.0005274240280304562,
      "loss": 0.003,
      "step": 103950
    },
    {
      "epoch": 31.82,
      "learning_rate": 0.000521865103429688,
      "loss": 0.003,
      "step": 104940
    },
    {
      "epoch": 32.0,
      "eval_loss": 0.02003033086657524,
      "eval_max_distance": 28,
      "eval_mean_distance": 1,
      "eval_runtime": 0.475,
      "eval_samples_per_second": 105.268,
      "eval_steps_per_second": 4.211,
      "step": 105536
    },
    {
      "epoch": 32.12,
      "learning_rate": 0.00051630617882892,
      "loss": 0.0028,
      "step": 105930
    },
    {
      "epoch": 32.42,
      "learning_rate": 0.0005107472542281517,
      "loss": 0.0027,
      "step": 106920
    },
    {
      "epoch": 32.72,
      "learning_rate": 0.0005051883296273836,
      "loss": 0.0028,
      "step": 107910
    },
    {
      "epoch": 33.0,
      "eval_loss": 0.01606147363781929,
      "eval_max_distance": 18,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4673,
      "eval_samples_per_second": 107.008,
      "eval_steps_per_second": 4.28,
      "step": 108834
    },
    {
      "epoch": 33.02,
      "learning_rate": 0.0004996294050266155,
      "loss": 0.0028,
      "step": 108900
    },
    {
      "epoch": 33.32,
      "learning_rate": 0.0004940704804258473,
      "loss": 0.0026,
      "step": 109890
    },
    {
      "epoch": 33.62,
      "learning_rate": 0.0004885115558250792,
      "loss": 0.0026,
      "step": 110880
    },
    {
      "epoch": 33.92,
      "learning_rate": 0.00048295263122431103,
      "loss": 0.0027,
      "step": 111870
    },
    {
      "epoch": 34.0,
      "eval_loss": 0.021506933495402336,
      "eval_max_distance": 26,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4763,
      "eval_samples_per_second": 104.968,
      "eval_steps_per_second": 4.199,
      "step": 112132
    },
    {
      "epoch": 34.22,
      "learning_rate": 0.00047739370662354294,
      "loss": 0.0024,
      "step": 112860
    },
    {
      "epoch": 34.52,
      "learning_rate": 0.00047183478202277474,
      "loss": 0.0023,
      "step": 113850
    },
    {
      "epoch": 34.82,
      "learning_rate": 0.0004662758574220066,
      "loss": 0.0025,
      "step": 114840
    },
    {
      "epoch": 35.0,
      "eval_loss": 0.019841769710183144,
      "eval_max_distance": 19,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4767,
      "eval_samples_per_second": 104.884,
      "eval_steps_per_second": 4.195,
      "step": 115430
    },
    {
      "epoch": 35.12,
      "learning_rate": 0.00046071693282123845,
      "loss": 0.0023,
      "step": 115830
    },
    {
      "epoch": 35.42,
      "learning_rate": 0.0004551580082204703,
      "loss": 0.0021,
      "step": 116820
    },
    {
      "epoch": 35.72,
      "learning_rate": 0.0004495990836197022,
      "loss": 0.0023,
      "step": 117810
    },
    {
      "epoch": 36.0,
      "eval_loss": 0.01675160974264145,
      "eval_max_distance": 24,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4591,
      "eval_samples_per_second": 108.901,
      "eval_steps_per_second": 4.356,
      "step": 118728
    },
    {
      "epoch": 36.02,
      "learning_rate": 0.000444040159018934,
      "loss": 0.0023,
      "step": 118800
    },
    {
      "epoch": 36.32,
      "learning_rate": 0.0004384812344181659,
      "loss": 0.0021,
      "step": 119790
    },
    {
      "epoch": 36.62,
      "learning_rate": 0.0004329223098173978,
      "loss": 0.0021,
      "step": 120780
    },
    {
      "epoch": 36.92,
      "learning_rate": 0.0004273633852166296,
      "loss": 0.002,
      "step": 121770
    },
    {
      "epoch": 37.0,
      "eval_loss": 0.022139811888337135,
      "eval_max_distance": 32,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4713,
      "eval_samples_per_second": 106.08,
      "eval_steps_per_second": 4.243,
      "step": 122026
    },
    {
      "epoch": 37.22,
      "learning_rate": 0.0004218044606158615,
      "loss": 0.002,
      "step": 122760
    },
    {
      "epoch": 37.52,
      "learning_rate": 0.00041624553601509335,
      "loss": 0.0019,
      "step": 123750
    },
    {
      "epoch": 37.82,
      "learning_rate": 0.00041068661141432515,
      "loss": 0.0019,
      "step": 124740
    },
    {
      "epoch": 38.0,
      "eval_loss": 0.02140805311501026,
      "eval_max_distance": 32,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4808,
      "eval_samples_per_second": 104.001,
      "eval_steps_per_second": 4.16,
      "step": 125324
    },
    {
      "epoch": 38.12,
      "learning_rate": 0.00040512768681355706,
      "loss": 0.0019,
      "step": 125730
    },
    {
      "epoch": 38.42,
      "learning_rate": 0.0003995687622127889,
      "loss": 0.0018,
      "step": 126720
    },
    {
      "epoch": 38.72,
      "learning_rate": 0.0003940098376120208,
      "loss": 0.0017,
      "step": 127710
    },
    {
      "epoch": 39.0,
      "eval_loss": 0.018618840724229813,
      "eval_max_distance": 19,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4752,
      "eval_samples_per_second": 105.222,
      "eval_steps_per_second": 4.209,
      "step": 128622
    },
    {
      "epoch": 39.02,
      "learning_rate": 0.00038845091301125263,
      "loss": 0.002,
      "step": 128700
    },
    {
      "epoch": 39.32,
      "learning_rate": 0.0003828919884104845,
      "loss": 0.0016,
      "step": 129690
    },
    {
      "epoch": 39.62,
      "learning_rate": 0.00037733306380971634,
      "loss": 0.0017,
      "step": 130680
    },
    {
      "epoch": 39.92,
      "learning_rate": 0.0003717741392089482,
      "loss": 0.0017,
      "step": 131670
    },
    {
      "epoch": 40.0,
      "eval_loss": 0.017086679115891457,
      "eval_max_distance": 23,
      "eval_mean_distance": 0,
      "eval_runtime": 0.458,
      "eval_samples_per_second": 109.178,
      "eval_steps_per_second": 4.367,
      "step": 131920
    },
    {
      "epoch": 40.22,
      "learning_rate": 0.00036621521460818,
      "loss": 0.0015,
      "step": 132660
    },
    {
      "epoch": 40.52,
      "learning_rate": 0.0003606562900074119,
      "loss": 0.0016,
      "step": 133650
    },
    {
      "epoch": 40.82,
      "learning_rate": 0.00035509736540664376,
      "loss": 0.0016,
      "step": 134640
    },
    {
      "epoch": 41.0,
      "eval_loss": 0.01638130471110344,
      "eval_max_distance": 17,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4581,
      "eval_samples_per_second": 109.147,
      "eval_steps_per_second": 4.366,
      "step": 135218
    },
    {
      "epoch": 41.12,
      "learning_rate": 0.0003495384408058756,
      "loss": 0.0015,
      "step": 135630
    },
    {
      "epoch": 41.43,
      "learning_rate": 0.0003439795162051075,
      "loss": 0.0014,
      "step": 136620
    },
    {
      "epoch": 41.73,
      "learning_rate": 0.00033842059160433933,
      "loss": 0.0015,
      "step": 137610
    },
    {
      "epoch": 42.0,
      "eval_loss": 0.016585057601332664,
      "eval_max_distance": 21,
      "eval_mean_distance": 1,
      "eval_runtime": 0.479,
      "eval_samples_per_second": 104.393,
      "eval_steps_per_second": 4.176,
      "step": 138516
    },
    {
      "epoch": 42.03,
      "learning_rate": 0.0003328616670035712,
      "loss": 0.0014,
      "step": 138600
    },
    {
      "epoch": 42.33,
      "learning_rate": 0.00032730274240280304,
      "loss": 0.0015,
      "step": 139590
    },
    {
      "epoch": 42.63,
      "learning_rate": 0.00032174381780203495,
      "loss": 0.0015,
      "step": 140580
    },
    {
      "epoch": 42.93,
      "learning_rate": 0.00031618489320126675,
      "loss": 0.0014,
      "step": 141570
    },
    {
      "epoch": 43.0,
      "eval_loss": 0.016704820096492767,
      "eval_max_distance": 21,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4809,
      "eval_samples_per_second": 103.976,
      "eval_steps_per_second": 4.159,
      "step": 141814
    },
    {
      "epoch": 43.23,
      "learning_rate": 0.0003106259686004986,
      "loss": 0.0011,
      "step": 142560
    },
    {
      "epoch": 43.53,
      "learning_rate": 0.0003050670439997305,
      "loss": 0.0013,
      "step": 143550
    },
    {
      "epoch": 43.83,
      "learning_rate": 0.0002995081193989623,
      "loss": 0.0019,
      "step": 144540
    },
    {
      "epoch": 44.0,
      "eval_loss": 0.019240867346525192,
      "eval_max_distance": 32,
      "eval_mean_distance": 1,
      "eval_runtime": 0.6494,
      "eval_samples_per_second": 76.999,
      "eval_steps_per_second": 3.08,
      "step": 145112
    },
    {
      "epoch": 44.13,
      "learning_rate": 0.00029394919479819423,
      "loss": 0.0012,
      "step": 145530
    },
    {
      "epoch": 44.43,
      "learning_rate": 0.00028839027019742603,
      "loss": 0.0011,
      "step": 146520
    },
    {
      "epoch": 44.73,
      "learning_rate": 0.0002828313455966579,
      "loss": 0.0011,
      "step": 147510
    },
    {
      "epoch": 45.0,
      "eval_loss": 0.02091757208108902,
      "eval_max_distance": 27,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4646,
      "eval_samples_per_second": 107.608,
      "eval_steps_per_second": 4.304,
      "step": 148410
    },
    {
      "epoch": 45.03,
      "learning_rate": 0.0002772724209958898,
      "loss": 0.0011,
      "step": 148500
    },
    {
      "epoch": 45.33,
      "learning_rate": 0.0002717134963951216,
      "loss": 0.0011,
      "step": 149490
    },
    {
      "epoch": 45.63,
      "learning_rate": 0.0002661545717943535,
      "loss": 0.001,
      "step": 150480
    },
    {
      "epoch": 45.93,
      "learning_rate": 0.00026059564719358537,
      "loss": 0.0011,
      "step": 151470
    },
    {
      "epoch": 46.0,
      "eval_loss": 0.02175173908472061,
      "eval_max_distance": 23,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4863,
      "eval_samples_per_second": 102.827,
      "eval_steps_per_second": 4.113,
      "step": 151708
    },
    {
      "epoch": 46.23,
      "learning_rate": 0.00025503672259281717,
      "loss": 0.001,
      "step": 152460
    },
    {
      "epoch": 46.53,
      "learning_rate": 0.0002494777979920491,
      "loss": 0.001,
      "step": 153450
    },
    {
      "epoch": 46.83,
      "learning_rate": 0.0002439188733912809,
      "loss": 0.001,
      "step": 154440
    },
    {
      "epoch": 47.0,
      "eval_loss": 0.01951581984758377,
      "eval_max_distance": 25,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4608,
      "eval_samples_per_second": 108.512,
      "eval_steps_per_second": 4.34,
      "step": 155006
    },
    {
      "epoch": 47.13,
      "learning_rate": 0.0002383599487905128,
      "loss": 0.001,
      "step": 155430
    },
    {
      "epoch": 47.43,
      "learning_rate": 0.00023280102418974464,
      "loss": 0.0009,
      "step": 156420
    },
    {
      "epoch": 47.73,
      "learning_rate": 0.00022724209958897647,
      "loss": 0.0009,
      "step": 157410
    },
    {
      "epoch": 48.0,
      "eval_loss": 0.01657327450811863,
      "eval_max_distance": 15,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4688,
      "eval_samples_per_second": 106.651,
      "eval_steps_per_second": 4.266,
      "step": 158304
    },
    {
      "epoch": 48.03,
      "learning_rate": 0.00022168317498820833,
      "loss": 0.0009,
      "step": 158400
    },
    {
      "epoch": 48.33,
      "learning_rate": 0.0002161242503874402,
      "loss": 0.0008,
      "step": 159390
    },
    {
      "epoch": 48.63,
      "learning_rate": 0.00021056532578667207,
      "loss": 0.0008,
      "step": 160380
    },
    {
      "epoch": 48.93,
      "learning_rate": 0.00020500640118590392,
      "loss": 0.0008,
      "step": 161370
    },
    {
      "epoch": 49.0,
      "eval_loss": 0.020961837843060493,
      "eval_max_distance": 31,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4893,
      "eval_samples_per_second": 102.188,
      "eval_steps_per_second": 4.088,
      "step": 161602
    },
    {
      "epoch": 49.23,
      "learning_rate": 0.00019944747658513578,
      "loss": 0.0008,
      "step": 162360
    },
    {
      "epoch": 49.53,
      "learning_rate": 0.00019388855198436764,
      "loss": 0.0008,
      "step": 163350
    },
    {
      "epoch": 49.83,
      "learning_rate": 0.0001883296273835995,
      "loss": 0.0008,
      "step": 164340
    },
    {
      "epoch": 50.0,
      "eval_loss": 0.022983456030488014,
      "eval_max_distance": 22,
      "eval_mean_distance": 0,
      "eval_runtime": 0.479,
      "eval_samples_per_second": 104.39,
      "eval_steps_per_second": 4.176,
      "step": 164900
    },
    {
      "epoch": 50.13,
      "learning_rate": 0.00018277070278283135,
      "loss": 0.0008,
      "step": 165330
    },
    {
      "epoch": 50.43,
      "learning_rate": 0.0001772117781820632,
      "loss": 0.0007,
      "step": 166320
    },
    {
      "epoch": 50.73,
      "learning_rate": 0.00017165285358129506,
      "loss": 0.0008,
      "step": 167310
    },
    {
      "epoch": 51.0,
      "eval_loss": 0.018444916233420372,
      "eval_max_distance": 15,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4866,
      "eval_samples_per_second": 102.75,
      "eval_steps_per_second": 4.11,
      "step": 168198
    },
    {
      "epoch": 51.03,
      "learning_rate": 0.00016609392898052691,
      "loss": 0.0007,
      "step": 168300
    },
    {
      "epoch": 51.33,
      "learning_rate": 0.0001605350043797588,
      "loss": 0.0007,
      "step": 169290
    },
    {
      "epoch": 51.63,
      "learning_rate": 0.00015497607977899065,
      "loss": 0.0007,
      "step": 170280
    },
    {
      "epoch": 51.93,
      "learning_rate": 0.00014941715517822248,
      "loss": 0.0007,
      "step": 171270
    },
    {
      "epoch": 52.0,
      "eval_loss": 0.01832015998661518,
      "eval_max_distance": 15,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4672,
      "eval_samples_per_second": 107.025,
      "eval_steps_per_second": 4.281,
      "step": 171496
    },
    {
      "epoch": 52.23,
      "learning_rate": 0.00014385823057745434,
      "loss": 0.0006,
      "step": 172260
    },
    {
      "epoch": 52.53,
      "learning_rate": 0.00013829930597668622,
      "loss": 0.0006,
      "step": 173250
    },
    {
      "epoch": 52.83,
      "learning_rate": 0.00013274038137591808,
      "loss": 0.0006,
      "step": 174240
    },
    {
      "epoch": 53.0,
      "eval_loss": 0.023398304358124733,
      "eval_max_distance": 32,
      "eval_mean_distance": 1,
      "eval_runtime": 0.4822,
      "eval_samples_per_second": 103.698,
      "eval_steps_per_second": 4.148,
      "step": 174794
    },
    {
      "epoch": 53.13,
      "learning_rate": 0.0001271814567751499,
      "loss": 0.0006,
      "step": 175230
    },
    {
      "epoch": 53.43,
      "learning_rate": 0.00012162253217438179,
      "loss": 0.0006,
      "step": 176220
    },
    {
      "epoch": 53.73,
      "learning_rate": 0.00011606360757361364,
      "loss": 0.0005,
      "step": 177210
    },
    {
      "epoch": 54.0,
      "eval_loss": 0.022733934223651886,
      "eval_max_distance": 24,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4789,
      "eval_samples_per_second": 104.41,
      "eval_steps_per_second": 4.176,
      "step": 178092
    },
    {
      "epoch": 54.03,
      "learning_rate": 0.0001105046829728455,
      "loss": 0.0005,
      "step": 178200
    },
    {
      "epoch": 54.33,
      "learning_rate": 0.00010494575837207735,
      "loss": 0.0005,
      "step": 179190
    },
    {
      "epoch": 54.63,
      "learning_rate": 9.938683377130921e-05,
      "loss": 0.0005,
      "step": 180180
    },
    {
      "epoch": 54.93,
      "learning_rate": 9.382790917054107e-05,
      "loss": 0.0004,
      "step": 181170
    },
    {
      "epoch": 55.0,
      "eval_loss": 0.018815917894244194,
      "eval_max_distance": 15,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4798,
      "eval_samples_per_second": 104.21,
      "eval_steps_per_second": 4.168,
      "step": 181390
    },
    {
      "epoch": 55.23,
      "learning_rate": 8.826898456977294e-05,
      "loss": 0.0005,
      "step": 182160
    },
    {
      "epoch": 55.53,
      "learning_rate": 8.271005996900478e-05,
      "loss": 0.0004,
      "step": 183150
    },
    {
      "epoch": 55.83,
      "learning_rate": 7.715113536823665e-05,
      "loss": 0.0005,
      "step": 184140
    },
    {
      "epoch": 56.0,
      "eval_loss": 0.01906018890440464,
      "eval_max_distance": 15,
      "eval_mean_distance": 0,
      "eval_runtime": 0.48,
      "eval_samples_per_second": 104.168,
      "eval_steps_per_second": 4.167,
      "step": 184688
    },
    {
      "epoch": 56.13,
      "learning_rate": 7.15922107674685e-05,
      "loss": 0.0004,
      "step": 185130
    },
    {
      "epoch": 56.43,
      "learning_rate": 6.603328616670036e-05,
      "loss": 0.0004,
      "step": 186120
    },
    {
      "epoch": 56.73,
      "learning_rate": 6.0474361565932214e-05,
      "loss": 0.0004,
      "step": 187110
    },
    {
      "epoch": 57.0,
      "eval_loss": 0.018282707780599594,
      "eval_max_distance": 15,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4797,
      "eval_samples_per_second": 104.233,
      "eval_steps_per_second": 4.169,
      "step": 187986
    },
    {
      "epoch": 57.03,
      "learning_rate": 5.491543696516407e-05,
      "loss": 0.0004,
      "step": 188100
    },
    {
      "epoch": 57.33,
      "learning_rate": 4.935651236439593e-05,
      "loss": 0.0004,
      "step": 189090
    },
    {
      "epoch": 57.63,
      "learning_rate": 4.379758776362779e-05,
      "loss": 0.0004,
      "step": 190080
    },
    {
      "epoch": 57.94,
      "learning_rate": 3.823866316285965e-05,
      "loss": 0.0003,
      "step": 191070
    },
    {
      "epoch": 58.0,
      "eval_loss": 0.018019111827015877,
      "eval_max_distance": 15,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4619,
      "eval_samples_per_second": 108.242,
      "eval_steps_per_second": 4.33,
      "step": 191284
    },
    {
      "epoch": 58.24,
      "learning_rate": 3.2679738562091506e-05,
      "loss": 0.0004,
      "step": 192060
    },
    {
      "epoch": 58.54,
      "learning_rate": 2.7120813961323362e-05,
      "loss": 0.0004,
      "step": 193050
    },
    {
      "epoch": 58.84,
      "learning_rate": 2.1561889360555218e-05,
      "loss": 0.0003,
      "step": 194040
    },
    {
      "epoch": 59.0,
      "eval_loss": 0.01795811764895916,
      "eval_max_distance": 15,
      "eval_mean_distance": 0,
      "eval_runtime": 0.475,
      "eval_samples_per_second": 105.265,
      "eval_steps_per_second": 4.211,
      "step": 194582
    },
    {
      "epoch": 59.14,
      "learning_rate": 1.6002964759787074e-05,
      "loss": 0.0004,
      "step": 195030
    },
    {
      "epoch": 59.44,
      "learning_rate": 1.0444040159018933e-05,
      "loss": 0.0004,
      "step": 196020
    },
    {
      "epoch": 59.74,
      "learning_rate": 4.885115558250792e-06,
      "loss": 0.0004,
      "step": 197010
    },
    {
      "epoch": 60.0,
      "eval_loss": 0.017678335309028625,
      "eval_max_distance": 15,
      "eval_mean_distance": 0,
      "eval_runtime": 0.4798,
      "eval_samples_per_second": 104.214,
      "eval_steps_per_second": 4.169,
      "step": 197880
    },
    {
      "epoch": 60.0,
      "step": 197880,
      "total_flos": 1.1400109636858675e+17,
      "train_loss": 0.031872519274052644,
      "train_runtime": 16366.2485,
      "train_samples_per_second": 362.656,
      "train_steps_per_second": 12.091
    }
  ],
  "logging_steps": 990,
  "max_steps": 197880,
  "num_train_epochs": 60,
  "save_steps": 1979,
  "total_flos": 1.1400109636858675e+17,
  "trial_name": null,
  "trial_params": null
}