{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 363,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002758620689655172,
      "grad_norm": 2.2012167823201163,
      "learning_rate": 0.0,
      "loss": 0.0852,
      "step": 1
    },
    {
      "epoch": 0.005517241379310344,
      "grad_norm": 2.684327612345962,
      "learning_rate": 5.405405405405406e-07,
      "loss": 0.1171,
      "step": 2
    },
    {
      "epoch": 0.008275862068965517,
      "grad_norm": 2.922848494791373,
      "learning_rate": 1.0810810810810812e-06,
      "loss": 0.1201,
      "step": 3
    },
    {
      "epoch": 0.011034482758620689,
      "grad_norm": 2.923985908786216,
      "learning_rate": 1.6216216216216219e-06,
      "loss": 0.0976,
      "step": 4
    },
    {
      "epoch": 0.013793103448275862,
      "grad_norm": 1.2724240051553533,
      "learning_rate": 2.1621621621621623e-06,
      "loss": 0.1046,
      "step": 5
    },
    {
      "epoch": 0.016551724137931035,
      "grad_norm": 1.673019205535178,
      "learning_rate": 2.702702702702703e-06,
      "loss": 0.0609,
      "step": 6
    },
    {
      "epoch": 0.019310344827586208,
      "grad_norm": 2.504855019369174,
      "learning_rate": 3.2432432432432437e-06,
      "loss": 0.081,
      "step": 7
    },
    {
      "epoch": 0.022068965517241378,
      "grad_norm": 1.878885411504625,
      "learning_rate": 3.7837837837837844e-06,
      "loss": 0.1121,
      "step": 8
    },
    {
      "epoch": 0.02482758620689655,
      "grad_norm": 2.653287813889321,
      "learning_rate": 4.324324324324325e-06,
      "loss": 0.1125,
      "step": 9
    },
    {
      "epoch": 0.027586206896551724,
      "grad_norm": 1.7453919996030804,
      "learning_rate": 4.864864864864866e-06,
      "loss": 0.0724,
      "step": 10
    },
    {
      "epoch": 0.030344827586206897,
      "grad_norm": 3.333638766021421,
      "learning_rate": 5.405405405405406e-06,
      "loss": 0.1475,
      "step": 11
    },
    {
      "epoch": 0.03310344827586207,
      "grad_norm": 2.292734722563975,
      "learning_rate": 5.945945945945947e-06,
      "loss": 0.0974,
      "step": 12
    },
    {
      "epoch": 0.03586206896551724,
      "grad_norm": 1.099049620799708,
      "learning_rate": 6.486486486486487e-06,
      "loss": 0.0597,
      "step": 13
    },
    {
      "epoch": 0.038620689655172416,
      "grad_norm": 1.727408246250002,
      "learning_rate": 7.027027027027028e-06,
      "loss": 0.0778,
      "step": 14
    },
    {
      "epoch": 0.041379310344827586,
      "grad_norm": 1.761342077131181,
      "learning_rate": 7.567567567567569e-06,
      "loss": 0.0955,
      "step": 15
    },
    {
      "epoch": 0.044137931034482755,
      "grad_norm": 1.9277700023635902,
      "learning_rate": 8.108108108108109e-06,
      "loss": 0.0869,
      "step": 16
    },
    {
      "epoch": 0.04689655172413793,
      "grad_norm": 1.568808525211501,
      "learning_rate": 8.64864864864865e-06,
      "loss": 0.0807,
      "step": 17
    },
    {
      "epoch": 0.0496551724137931,
      "grad_norm": 1.3523306531543753,
      "learning_rate": 9.189189189189191e-06,
      "loss": 0.0665,
      "step": 18
    },
    {
      "epoch": 0.05241379310344828,
      "grad_norm": 1.7180837689557336,
      "learning_rate": 9.729729729729732e-06,
      "loss": 0.0868,
      "step": 19
    },
    {
      "epoch": 0.05517241379310345,
      "grad_norm": 1.5131863973051378,
      "learning_rate": 1.027027027027027e-05,
      "loss": 0.0736,
      "step": 20
    },
    {
      "epoch": 0.057931034482758624,
      "grad_norm": 1.6815268723910783,
      "learning_rate": 1.0810810810810812e-05,
      "loss": 0.079,
      "step": 21
    },
    {
      "epoch": 0.060689655172413794,
      "grad_norm": 1.4834028913829107,
      "learning_rate": 1.1351351351351352e-05,
      "loss": 0.0437,
      "step": 22
    },
    {
      "epoch": 0.06344827586206897,
      "grad_norm": 2.1417510445007792,
      "learning_rate": 1.1891891891891894e-05,
      "loss": 0.0905,
      "step": 23
    },
    {
      "epoch": 0.06620689655172414,
      "grad_norm": 1.7871991104284883,
      "learning_rate": 1.2432432432432433e-05,
      "loss": 0.1024,
      "step": 24
    },
    {
      "epoch": 0.06896551724137931,
      "grad_norm": 2.0156063129423485,
      "learning_rate": 1.2972972972972975e-05,
      "loss": 0.0701,
      "step": 25
    },
    {
      "epoch": 0.07172413793103448,
      "grad_norm": 1.9160557707810737,
      "learning_rate": 1.3513513513513515e-05,
      "loss": 0.0748,
      "step": 26
    },
    {
      "epoch": 0.07448275862068965,
      "grad_norm": 2.000911025201981,
      "learning_rate": 1.4054054054054055e-05,
      "loss": 0.101,
      "step": 27
    },
    {
      "epoch": 0.07724137931034483,
      "grad_norm": 2.115047269982351,
      "learning_rate": 1.4594594594594596e-05,
      "loss": 0.0918,
      "step": 28
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.8098326272573262,
      "learning_rate": 1.5135135135135138e-05,
      "loss": 0.0915,
      "step": 29
    },
    {
      "epoch": 0.08275862068965517,
      "grad_norm": 2.1207994498886764,
      "learning_rate": 1.5675675675675676e-05,
      "loss": 0.1025,
      "step": 30
    },
    {
      "epoch": 0.08551724137931034,
      "grad_norm": 2.171868532095742,
      "learning_rate": 1.6216216216216218e-05,
      "loss": 0.1059,
      "step": 31
    },
    {
      "epoch": 0.08827586206896551,
      "grad_norm": 1.6943229112776936,
      "learning_rate": 1.6756756756756757e-05,
      "loss": 0.0677,
      "step": 32
    },
    {
      "epoch": 0.0910344827586207,
      "grad_norm": 1.994899794317826,
      "learning_rate": 1.72972972972973e-05,
      "loss": 0.0894,
      "step": 33
    },
    {
      "epoch": 0.09379310344827586,
      "grad_norm": 2.3711691862715907,
      "learning_rate": 1.783783783783784e-05,
      "loss": 0.1158,
      "step": 34
    },
    {
      "epoch": 0.09655172413793103,
      "grad_norm": 1.9877327316155244,
      "learning_rate": 1.8378378378378383e-05,
      "loss": 0.0924,
      "step": 35
    },
    {
      "epoch": 0.0993103448275862,
      "grad_norm": 2.0641113315593898,
      "learning_rate": 1.891891891891892e-05,
      "loss": 0.0978,
      "step": 36
    },
    {
      "epoch": 0.10206896551724139,
      "grad_norm": 2.5636633454788593,
      "learning_rate": 1.9459459459459463e-05,
      "loss": 0.1327,
      "step": 37
    },
    {
      "epoch": 0.10482758620689656,
      "grad_norm": 2.3262721461116396,
      "learning_rate": 2e-05,
      "loss": 0.095,
      "step": 38
    },
    {
      "epoch": 0.10758620689655173,
      "grad_norm": 2.0685116596847832,
      "learning_rate": 1.9999535665248e-05,
      "loss": 0.0706,
      "step": 39
    },
    {
      "epoch": 0.1103448275862069,
      "grad_norm": 2.5954824427417007,
      "learning_rate": 1.999814270411335e-05,
      "loss": 0.1205,
      "step": 40
    },
    {
      "epoch": 0.11310344827586206,
      "grad_norm": 2.144122161715886,
      "learning_rate": 1.99958212459561e-05,
      "loss": 0.0896,
      "step": 41
    },
    {
      "epoch": 0.11586206896551725,
      "grad_norm": 1.9322201299634434,
      "learning_rate": 1.9992571506362997e-05,
      "loss": 0.1012,
      "step": 42
    },
    {
      "epoch": 0.11862068965517242,
      "grad_norm": 2.075048476162322,
      "learning_rate": 1.9988393787127444e-05,
      "loss": 0.1087,
      "step": 43
    },
    {
      "epoch": 0.12137931034482759,
      "grad_norm": 2.259726318820954,
      "learning_rate": 1.9983288476221482e-05,
      "loss": 0.1382,
      "step": 44
    },
    {
      "epoch": 0.12413793103448276,
      "grad_norm": 1.360664005910794,
      "learning_rate": 1.9977256047759765e-05,
      "loss": 0.0538,
      "step": 45
    },
    {
      "epoch": 0.12689655172413794,
      "grad_norm": 1.6435792074598727,
      "learning_rate": 1.9970297061955533e-05,
      "loss": 0.0855,
      "step": 46
    },
    {
      "epoch": 0.1296551724137931,
      "grad_norm": 2.3601345815092483,
      "learning_rate": 1.9962412165068575e-05,
      "loss": 0.1386,
      "step": 47
    },
    {
      "epoch": 0.13241379310344828,
      "grad_norm": 1.7105484837560168,
      "learning_rate": 1.9953602089345215e-05,
      "loss": 0.0961,
      "step": 48
    },
    {
      "epoch": 0.13517241379310344,
      "grad_norm": 1.6496292161975452,
      "learning_rate": 1.9943867652950323e-05,
      "loss": 0.0857,
      "step": 49
    },
    {
      "epoch": 0.13793103448275862,
      "grad_norm": 1.2468755808496865,
      "learning_rate": 1.9933209759891318e-05,
      "loss": 0.0563,
      "step": 50
    },
    {
      "epoch": 0.1406896551724138,
      "grad_norm": 2.0793413384987183,
      "learning_rate": 1.9921629399934224e-05,
      "loss": 0.108,
      "step": 51
    },
    {
      "epoch": 0.14344827586206896,
      "grad_norm": 2.0093677538398524,
      "learning_rate": 1.9909127648511758e-05,
      "loss": 0.1289,
      "step": 52
    },
    {
      "epoch": 0.14620689655172414,
      "grad_norm": 1.2736371995174656,
      "learning_rate": 1.989570566662345e-05,
      "loss": 0.0615,
      "step": 53
    },
    {
      "epoch": 0.1489655172413793,
      "grad_norm": 1.5368399292599437,
      "learning_rate": 1.9881364700727827e-05,
      "loss": 0.0559,
      "step": 54
    },
    {
      "epoch": 0.15172413793103448,
      "grad_norm": 1.6143124587473046,
      "learning_rate": 1.986610608262665e-05,
      "loss": 0.0512,
      "step": 55
    },
    {
      "epoch": 0.15448275862068966,
      "grad_norm": 2.0640396297545323,
      "learning_rate": 1.9849931229341258e-05,
      "loss": 0.1492,
      "step": 56
    },
    {
      "epoch": 0.15724137931034482,
      "grad_norm": 1.9321341515451917,
      "learning_rate": 1.9832841642980948e-05,
      "loss": 0.1277,
      "step": 57
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.5565020190489753,
      "learning_rate": 1.981483891060348e-05,
      "loss": 0.1266,
      "step": 58
    },
    {
      "epoch": 0.16275862068965516,
      "grad_norm": 1.6731150384360727,
      "learning_rate": 1.979592470406772e-05,
      "loss": 0.095,
      "step": 59
    },
    {
      "epoch": 0.16551724137931034,
      "grad_norm": 1.3348447458614452,
      "learning_rate": 1.9776100779878344e-05,
      "loss": 0.0602,
      "step": 60
    },
    {
      "epoch": 0.16827586206896553,
      "grad_norm": 1.211440894700164,
      "learning_rate": 1.9755368979022734e-05,
      "loss": 0.0597,
      "step": 61
    },
    {
      "epoch": 0.17103448275862068,
      "grad_norm": 2.045395762464389,
      "learning_rate": 1.9733731226800016e-05,
      "loss": 0.0902,
      "step": 62
    },
    {
      "epoch": 0.17379310344827587,
      "grad_norm": 1.681889406229311,
      "learning_rate": 1.9711189532642244e-05,
      "loss": 0.0851,
      "step": 63
    },
    {
      "epoch": 0.17655172413793102,
      "grad_norm": 2.156485405646681,
      "learning_rate": 1.9687745989927823e-05,
      "loss": 0.1022,
      "step": 64
    },
    {
      "epoch": 0.1793103448275862,
      "grad_norm": 2.2694346385487765,
      "learning_rate": 1.9663402775787066e-05,
      "loss": 0.1402,
      "step": 65
    },
    {
      "epoch": 0.1820689655172414,
      "grad_norm": 2.2102729735649502,
      "learning_rate": 1.9638162150900028e-05,
      "loss": 0.0595,
      "step": 66
    },
    {
      "epoch": 0.18482758620689654,
      "grad_norm": 2.4808367225986023,
      "learning_rate": 1.961202645928658e-05,
      "loss": 0.1526,
      "step": 67
    },
    {
      "epoch": 0.18758620689655173,
      "grad_norm": 2.125991623159161,
      "learning_rate": 1.9584998128088686e-05,
      "loss": 0.1149,
      "step": 68
    },
    {
      "epoch": 0.19034482758620688,
      "grad_norm": 3.5959738930248553,
      "learning_rate": 1.955707966734505e-05,
      "loss": 0.1724,
      "step": 69
    },
    {
      "epoch": 0.19310344827586207,
      "grad_norm": 2.217349082976429,
      "learning_rate": 1.9528273669757974e-05,
      "loss": 0.1357,
      "step": 70
    },
    {
      "epoch": 0.19586206896551725,
      "grad_norm": 2.9019793725624288,
      "learning_rate": 1.9498582810452607e-05,
      "loss": 0.1289,
      "step": 71
    },
    {
      "epoch": 0.1986206896551724,
      "grad_norm": 2.677822777383495,
      "learning_rate": 1.9468009846728515e-05,
      "loss": 0.1201,
      "step": 72
    },
    {
      "epoch": 0.2013793103448276,
      "grad_norm": 2.1904068713477693,
      "learning_rate": 1.9436557617803594e-05,
      "loss": 0.1343,
      "step": 73
    },
    {
      "epoch": 0.20413793103448277,
      "grad_norm": 1.9333752870288448,
      "learning_rate": 1.9404229044550432e-05,
      "loss": 0.1014,
      "step": 74
    },
    {
      "epoch": 0.20689655172413793,
      "grad_norm": 1.2594973301567354,
      "learning_rate": 1.9371027129225042e-05,
      "loss": 0.08,
      "step": 75
    },
    {
      "epoch": 0.2096551724137931,
      "grad_norm": 1.1588871349219974,
      "learning_rate": 1.9336954955188042e-05,
      "loss": 0.0664,
      "step": 76
    },
    {
      "epoch": 0.21241379310344827,
      "grad_norm": 2.424390700582275,
      "learning_rate": 1.9302015686618328e-05,
      "loss": 0.1228,
      "step": 77
    },
    {
      "epoch": 0.21517241379310345,
      "grad_norm": 1.839851381331024,
      "learning_rate": 1.9266212568219223e-05,
      "loss": 0.1181,
      "step": 78
    },
    {
      "epoch": 0.21793103448275863,
      "grad_norm": 2.9112052255264627,
      "learning_rate": 1.9229548924917146e-05,
      "loss": 0.1593,
      "step": 79
    },
    {
      "epoch": 0.2206896551724138,
      "grad_norm": 2.6573066473445124,
      "learning_rate": 1.9192028161552848e-05,
      "loss": 0.1068,
      "step": 80
    },
    {
      "epoch": 0.22344827586206897,
      "grad_norm": 2.228711734634549,
      "learning_rate": 1.915365376256519e-05,
      "loss": 0.1395,
      "step": 81
    },
    {
      "epoch": 0.22620689655172413,
      "grad_norm": 2.1487793287842463,
      "learning_rate": 1.9114429291667583e-05,
      "loss": 0.1322,
      "step": 82
    },
    {
      "epoch": 0.2289655172413793,
      "grad_norm": 1.418589529134045,
      "learning_rate": 1.9074358391517026e-05,
      "loss": 0.1298,
      "step": 83
    },
    {
      "epoch": 0.2317241379310345,
      "grad_norm": 1.6184969490675134,
      "learning_rate": 1.9033444783375806e-05,
      "loss": 0.1375,
      "step": 84
    },
    {
      "epoch": 0.23448275862068965,
      "grad_norm": 2.636414425888468,
      "learning_rate": 1.8991692266765947e-05,
      "loss": 0.1454,
      "step": 85
    },
    {
      "epoch": 0.23724137931034484,
      "grad_norm": 1.5999297559125214,
      "learning_rate": 1.8949104719116334e-05,
      "loss": 0.1004,
      "step": 86
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.7047937938863038,
      "learning_rate": 1.8905686095402648e-05,
      "loss": 0.0729,
      "step": 87
    },
    {
      "epoch": 0.24275862068965517,
      "grad_norm": 2.213860086801788,
      "learning_rate": 1.886144042778006e-05,
      "loss": 0.1219,
      "step": 88
    },
    {
      "epoch": 0.24551724137931036,
      "grad_norm": 1.9830213856101813,
      "learning_rate": 1.881637182520879e-05,
      "loss": 0.1277,
      "step": 89
    },
    {
      "epoch": 0.2482758620689655,
      "grad_norm": 2.1526995896994543,
      "learning_rate": 1.8770484473072518e-05,
      "loss": 0.133,
      "step": 90
    },
    {
      "epoch": 0.25103448275862067,
      "grad_norm": 5.615536299423343,
      "learning_rate": 1.87237826327897e-05,
      "loss": 0.1271,
      "step": 91
    },
    {
      "epoch": 0.2537931034482759,
      "grad_norm": 2.9892624763616213,
      "learning_rate": 1.8676270641417824e-05,
      "loss": 0.1337,
      "step": 92
    },
    {
      "epoch": 0.25655172413793104,
      "grad_norm": 1.7670570813048203,
      "learning_rate": 1.8627952911250632e-05,
      "loss": 0.1188,
      "step": 93
    },
    {
      "epoch": 0.2593103448275862,
      "grad_norm": 1.8832817735619782,
      "learning_rate": 1.857883392940837e-05,
      "loss": 0.09,
      "step": 94
    },
    {
      "epoch": 0.2620689655172414,
      "grad_norm": 2.2995757188920964,
      "learning_rate": 1.85289182574211e-05,
      "loss": 0.1084,
      "step": 95
    },
    {
      "epoch": 0.26482758620689656,
      "grad_norm": 2.268436925129853,
      "learning_rate": 1.847821053080505e-05,
      "loss": 0.1003,
      "step": 96
    },
    {
      "epoch": 0.2675862068965517,
      "grad_norm": 1.8417005302085403,
      "learning_rate": 1.8426715458632154e-05,
      "loss": 0.0747,
      "step": 97
    },
    {
      "epoch": 0.27034482758620687,
      "grad_norm": 2.497572736496612,
      "learning_rate": 1.8374437823092726e-05,
      "loss": 0.1513,
      "step": 98
    },
    {
      "epoch": 0.2731034482758621,
      "grad_norm": 2.0212629509138766,
      "learning_rate": 1.832138247905135e-05,
      "loss": 0.088,
      "step": 99
    },
    {
      "epoch": 0.27586206896551724,
      "grad_norm": 1.806243660764203,
      "learning_rate": 1.8267554353596027e-05,
      "loss": 0.0617,
      "step": 100
    },
    {
      "epoch": 0.2786206896551724,
      "grad_norm": 1.43840041921586,
      "learning_rate": 1.8212958445580623e-05,
      "loss": 0.0809,
      "step": 101
    },
    {
      "epoch": 0.2813793103448276,
      "grad_norm": 1.5624416473825227,
      "learning_rate": 1.815759982516061e-05,
      "loss": 0.0987,
      "step": 102
    },
    {
      "epoch": 0.28413793103448276,
      "grad_norm": 1.3722413173440258,
      "learning_rate": 1.8101483633322255e-05,
      "loss": 0.1055,
      "step": 103
    },
    {
      "epoch": 0.2868965517241379,
      "grad_norm": 1.5055322976041905,
      "learning_rate": 1.8044615081405153e-05,
      "loss": 0.12,
      "step": 104
    },
    {
      "epoch": 0.2896551724137931,
      "grad_norm": 1.9238100509571232,
      "learning_rate": 1.7986999450618295e-05,
      "loss": 0.1293,
      "step": 105
    },
    {
      "epoch": 0.2924137931034483,
      "grad_norm": 1.4169309751550185,
      "learning_rate": 1.7928642091549616e-05,
      "loss": 0.0945,
      "step": 106
    },
    {
      "epoch": 0.29517241379310344,
      "grad_norm": 1.7127369268478376,
      "learning_rate": 1.7869548423669075e-05,
      "loss": 0.1124,
      "step": 107
    },
    {
      "epoch": 0.2979310344827586,
      "grad_norm": 1.859903898855105,
      "learning_rate": 1.7809723934825405e-05,
      "loss": 0.1182,
      "step": 108
    },
    {
      "epoch": 0.3006896551724138,
      "grad_norm": 1.4775939025780847,
      "learning_rate": 1.7749174180736443e-05,
      "loss": 0.0857,
      "step": 109
    },
    {
      "epoch": 0.30344827586206896,
      "grad_norm": 1.4354799536836982,
      "learning_rate": 1.768790478447319e-05,
      "loss": 0.0781,
      "step": 110
    },
    {
      "epoch": 0.3062068965517241,
      "grad_norm": 1.0273549786544405,
      "learning_rate": 1.762592143593764e-05,
      "loss": 0.0713,
      "step": 111
    },
    {
      "epoch": 0.30896551724137933,
      "grad_norm": 1.9640486386494376,
      "learning_rate": 1.756322989133434e-05,
      "loss": 0.1301,
      "step": 112
    },
    {
      "epoch": 0.3117241379310345,
      "grad_norm": 1.7089277562540073,
      "learning_rate": 1.749983597263586e-05,
      "loss": 0.0681,
      "step": 113
    },
    {
      "epoch": 0.31448275862068964,
      "grad_norm": 1.9429869321530793,
      "learning_rate": 1.7435745567042096e-05,
      "loss": 0.1331,
      "step": 114
    },
    {
      "epoch": 0.31724137931034485,
      "grad_norm": 1.26536650090277,
      "learning_rate": 1.737096462643357e-05,
      "loss": 0.0604,
      "step": 115
    },
    {
      "epoch": 0.32,
      "grad_norm": 2.087216494160791,
      "learning_rate": 1.730549916681868e-05,
      "loss": 0.1494,
      "step": 116
    },
    {
      "epoch": 0.32275862068965516,
      "grad_norm": 1.9975034441736954,
      "learning_rate": 1.723935526777502e-05,
      "loss": 0.1092,
      "step": 117
    },
    {
      "epoch": 0.3255172413793103,
      "grad_norm": 2.038235893575754,
      "learning_rate": 1.717253907188477e-05,
      "loss": 0.1082,
      "step": 118
    },
    {
      "epoch": 0.32827586206896553,
      "grad_norm": 3.3184113391653396,
      "learning_rate": 1.7105056784164295e-05,
      "loss": 0.1703,
      "step": 119
    },
    {
      "epoch": 0.3310344827586207,
      "grad_norm": 2.0270058507038966,
      "learning_rate": 1.7036914671487854e-05,
      "loss": 0.1303,
      "step": 120
    },
    {
      "epoch": 0.33379310344827584,
      "grad_norm": 1.8154593381383104,
      "learning_rate": 1.6968119062005644e-05,
      "loss": 0.1077,
      "step": 121
    },
    {
      "epoch": 0.33655172413793105,
      "grad_norm": 1.2350092975243878,
      "learning_rate": 1.689867634455612e-05,
      "loss": 0.0924,
      "step": 122
    },
    {
      "epoch": 0.3393103448275862,
      "grad_norm": 1.792900391384705,
      "learning_rate": 1.682859296807268e-05,
      "loss": 0.1221,
      "step": 123
    },
    {
      "epoch": 0.34206896551724136,
      "grad_norm": 1.9404357508554042,
      "learning_rate": 1.675787544098477e-05,
      "loss": 0.1106,
      "step": 124
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 1.7814508590384097,
      "learning_rate": 1.6686530330613472e-05,
      "loss": 0.1122,
      "step": 125
    },
    {
      "epoch": 0.34758620689655173,
      "grad_norm": 1.53815702992395,
      "learning_rate": 1.661456426256161e-05,
      "loss": 0.087,
      "step": 126
    },
    {
      "epoch": 0.3503448275862069,
      "grad_norm": 1.9077349161199095,
      "learning_rate": 1.6541983920098462e-05,
      "loss": 0.087,
      "step": 127
    },
    {
      "epoch": 0.35310344827586204,
      "grad_norm": 2.2996250559408704,
      "learning_rate": 1.6468796043539082e-05,
      "loss": 0.1254,
      "step": 128
    },
    {
      "epoch": 0.35586206896551725,
      "grad_norm": 1.2031295382564096,
      "learning_rate": 1.639500742961838e-05,
      "loss": 0.0498,
      "step": 129
    },
    {
      "epoch": 0.3586206896551724,
      "grad_norm": 1.3436281870681093,
      "learning_rate": 1.6320624930859905e-05,
      "loss": 0.077,
      "step": 130
    },
    {
      "epoch": 0.36137931034482756,
      "grad_norm": 0.6314939772663396,
      "learning_rate": 1.6245655454939474e-05,
      "loss": 0.0352,
      "step": 131
    },
    {
      "epoch": 0.3641379310344828,
      "grad_norm": 2.420910473317855,
      "learning_rate": 1.6170105964043698e-05,
      "loss": 0.1228,
      "step": 132
    },
    {
      "epoch": 0.36689655172413793,
      "grad_norm": 2.80050435155021,
      "learning_rate": 1.6093983474223392e-05,
      "loss": 0.1799,
      "step": 133
    },
    {
      "epoch": 0.3696551724137931,
      "grad_norm": 1.9733179937634817,
      "learning_rate": 1.6017295054742045e-05,
      "loss": 0.1088,
      "step": 134
    },
    {
      "epoch": 0.3724137931034483,
      "grad_norm": 0.8866446082715613,
      "learning_rate": 1.5940047827419305e-05,
      "loss": 0.0722,
      "step": 135
    },
    {
      "epoch": 0.37517241379310345,
      "grad_norm": 1.4747083716633578,
      "learning_rate": 1.5862248965969604e-05,
      "loss": 0.0902,
      "step": 136
    },
    {
      "epoch": 0.3779310344827586,
      "grad_norm": 1.203872871772734,
      "learning_rate": 1.5783905695335947e-05,
      "loss": 0.0763,
      "step": 137
    },
    {
      "epoch": 0.38068965517241377,
      "grad_norm": 1.295759494638433,
      "learning_rate": 1.570502529101896e-05,
      "loss": 0.0582,
      "step": 138
    },
    {
      "epoch": 0.383448275862069,
      "grad_norm": 2.082719193654199,
      "learning_rate": 1.5625615078401244e-05,
      "loss": 0.1131,
      "step": 139
    },
    {
      "epoch": 0.38620689655172413,
      "grad_norm": 2.4303248947365046,
      "learning_rate": 1.5545682432067068e-05,
      "loss": 0.1319,
      "step": 140
    },
    {
      "epoch": 0.3889655172413793,
      "grad_norm": 2.308746027421277,
      "learning_rate": 1.5465234775117538e-05,
      "loss": 0.1232,
      "step": 141
    },
    {
      "epoch": 0.3917241379310345,
      "grad_norm": 2.9364629760972907,
      "learning_rate": 1.5384279578481223e-05,
      "loss": 0.1537,
      "step": 142
    },
    {
      "epoch": 0.39448275862068966,
      "grad_norm": 1.8350345374035169,
      "learning_rate": 1.5302824360220352e-05,
      "loss": 0.1161,
      "step": 143
    },
    {
      "epoch": 0.3972413793103448,
      "grad_norm": 1.8562459888809821,
      "learning_rate": 1.522087668483264e-05,
      "loss": 0.1054,
      "step": 144
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.2050577332052317,
      "learning_rate": 1.5138444162548791e-05,
      "loss": 0.072,
      "step": 145
    },
    {
      "epoch": 0.4027586206896552,
      "grad_norm": 2.576862971551694,
      "learning_rate": 1.5055534448625766e-05,
      "loss": 0.1956,
      "step": 146
    },
    {
      "epoch": 0.40551724137931033,
      "grad_norm": 2.526412596716041,
      "learning_rate": 1.4972155242635853e-05,
      "loss": 0.1163,
      "step": 147
    },
    {
      "epoch": 0.40827586206896554,
      "grad_norm": 2.4122326865926325,
      "learning_rate": 1.488831428775164e-05,
      "loss": 0.1588,
      "step": 148
    },
    {
      "epoch": 0.4110344827586207,
      "grad_norm": 1.5900334279122388,
      "learning_rate": 1.4804019370026927e-05,
      "loss": 0.0874,
      "step": 149
    },
    {
      "epoch": 0.41379310344827586,
      "grad_norm": 2.2583849148863284,
      "learning_rate": 1.4719278317673655e-05,
      "loss": 0.1225,
      "step": 150
    },
    {
      "epoch": 0.416551724137931,
      "grad_norm": 1.5043786808378643,
      "learning_rate": 1.4634099000334932e-05,
      "loss": 0.0628,
      "step": 151
    },
    {
      "epoch": 0.4193103448275862,
      "grad_norm": 2.6867593671378707,
      "learning_rate": 1.4548489328354197e-05,
      "loss": 0.1119,
      "step": 152
    },
    {
      "epoch": 0.4220689655172414,
      "grad_norm": 1.892183158003583,
      "learning_rate": 1.4462457252040606e-05,
      "loss": 0.1371,
      "step": 153
    },
    {
      "epoch": 0.42482758620689653,
      "grad_norm": 0.7376897959603039,
      "learning_rate": 1.437601076093073e-05,
      "loss": 0.0715,
      "step": 154
    },
    {
      "epoch": 0.42758620689655175,
      "grad_norm": 1.2313754414547655,
      "learning_rate": 1.4289157883046567e-05,
      "loss": 0.0671,
      "step": 155
    },
    {
      "epoch": 0.4303448275862069,
      "grad_norm": 2.443959323743539,
      "learning_rate": 1.420190668415002e-05,
      "loss": 0.0937,
      "step": 156
    },
    {
      "epoch": 0.43310344827586206,
      "grad_norm": 1.8279711072151712,
      "learning_rate": 1.4114265266993847e-05,
      "loss": 0.1114,
      "step": 157
    },
    {
      "epoch": 0.43586206896551727,
      "grad_norm": 2.722814716999782,
      "learning_rate": 1.4026241770569198e-05,
      "loss": 0.1388,
      "step": 158
    },
    {
      "epoch": 0.4386206896551724,
      "grad_norm": 2.522323402694141,
      "learning_rate": 1.3937844369349736e-05,
      "loss": 0.0862,
      "step": 159
    },
    {
      "epoch": 0.4413793103448276,
      "grad_norm": 1.0652613518383625,
      "learning_rate": 1.3849081272532545e-05,
      "loss": 0.0752,
      "step": 160
    },
    {
      "epoch": 0.44413793103448274,
      "grad_norm": 2.6542750874407335,
      "learning_rate": 1.375996072327573e-05,
      "loss": 0.1456,
      "step": 161
    },
    {
      "epoch": 0.44689655172413795,
      "grad_norm": 1.8544625800539327,
      "learning_rate": 1.3670490997932922e-05,
      "loss": 0.109,
      "step": 162
    },
    {
      "epoch": 0.4496551724137931,
      "grad_norm": 1.9735546120483556,
      "learning_rate": 1.3580680405284666e-05,
      "loss": 0.1144,
      "step": 163
    },
    {
      "epoch": 0.45241379310344826,
      "grad_norm": 1.2177687301571833,
      "learning_rate": 1.3490537285766809e-05,
      "loss": 0.0838,
      "step": 164
    },
    {
      "epoch": 0.45517241379310347,
      "grad_norm": 2.866049013585789,
      "learning_rate": 1.3400070010695966e-05,
      "loss": 0.1309,
      "step": 165
    },
    {
      "epoch": 0.4579310344827586,
      "grad_norm": 1.7714947880406966,
      "learning_rate": 1.3309286981492084e-05,
      "loss": 0.1336,
      "step": 166
    },
    {
      "epoch": 0.4606896551724138,
      "grad_norm": 2.3528937198669966,
      "learning_rate": 1.3218196628898232e-05,
      "loss": 0.0878,
      "step": 167
    },
    {
      "epoch": 0.463448275862069,
      "grad_norm": 1.32763797639713,
      "learning_rate": 1.3126807412197666e-05,
      "loss": 0.0866,
      "step": 168
    },
    {
      "epoch": 0.46620689655172415,
      "grad_norm": 1.0489060017302254,
      "learning_rate": 1.3035127818428239e-05,
      "loss": 0.0863,
      "step": 169
    },
    {
      "epoch": 0.4689655172413793,
      "grad_norm": 1.4769228384530726,
      "learning_rate": 1.2943166361594242e-05,
      "loss": 0.1393,
      "step": 170
    },
    {
      "epoch": 0.47172413793103446,
      "grad_norm": 1.9367121483246261,
      "learning_rate": 1.2850931581875723e-05,
      "loss": 0.1762,
      "step": 171
    },
    {
      "epoch": 0.47448275862068967,
      "grad_norm": 1.5440366202640854,
      "learning_rate": 1.275843204483539e-05,
      "loss": 0.0599,
      "step": 172
    },
    {
      "epoch": 0.4772413793103448,
      "grad_norm": 1.691410284752133,
      "learning_rate": 1.2665676340623172e-05,
      "loss": 0.0938,
      "step": 173
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.2010801724859,
      "learning_rate": 1.2572673083178448e-05,
      "loss": 0.098,
      "step": 174
    },
    {
      "epoch": 0.4827586206896552,
      "grad_norm": 2.223070626201405,
      "learning_rate": 1.2479430909430109e-05,
      "loss": 0.0936,
      "step": 175
    },
    {
      "epoch": 0.48551724137931035,
      "grad_norm": 2.1176352077764107,
      "learning_rate": 1.2385958478494487e-05,
      "loss": 0.1279,
      "step": 176
    },
    {
      "epoch": 0.4882758620689655,
      "grad_norm": 1.3628256498935367,
      "learning_rate": 1.2292264470871183e-05,
      "loss": 0.0783,
      "step": 177
    },
    {
      "epoch": 0.4910344827586207,
      "grad_norm": 1.064266823164612,
      "learning_rate": 1.2198357587636958e-05,
      "loss": 0.0525,
      "step": 178
    },
    {
      "epoch": 0.49379310344827587,
      "grad_norm": 2.00945904980985,
      "learning_rate": 1.2104246549637683e-05,
      "loss": 0.0978,
      "step": 179
    },
    {
      "epoch": 0.496551724137931,
      "grad_norm": 1.3141391526330848,
      "learning_rate": 1.2009940096678451e-05,
      "loss": 0.0963,
      "step": 180
    },
    {
      "epoch": 0.4993103448275862,
      "grad_norm": 1.6531005740449383,
      "learning_rate": 1.1915446986711953e-05,
      "loss": 0.1092,
      "step": 181
    },
    {
      "epoch": 0.5020689655172413,
      "grad_norm": 1.4945974838001637,
      "learning_rate": 1.1820775995025147e-05,
      "loss": 0.0667,
      "step": 182
    },
    {
      "epoch": 0.5048275862068966,
      "grad_norm": 1.3762608443334923,
      "learning_rate": 1.172593591342432e-05,
      "loss": 0.0882,
      "step": 183
    },
    {
      "epoch": 0.5075862068965518,
      "grad_norm": 1.5415877422060749,
      "learning_rate": 1.1630935549418627e-05,
      "loss": 0.0865,
      "step": 184
    },
    {
      "epoch": 0.5103448275862069,
      "grad_norm": 2.0042018778537622,
      "learning_rate": 1.1535783725402163e-05,
      "loss": 0.1353,
      "step": 185
    },
    {
      "epoch": 0.5131034482758621,
      "grad_norm": 2.6818407651740266,
      "learning_rate": 1.1440489277834645e-05,
      "loss": 0.1393,
      "step": 186
    },
    {
      "epoch": 0.5158620689655172,
      "grad_norm": 2.3807831512162374,
      "learning_rate": 1.134506105642081e-05,
      "loss": 0.1169,
      "step": 187
    },
    {
      "epoch": 0.5186206896551724,
      "grad_norm": 2.202610904326787,
      "learning_rate": 1.1249507923288563e-05,
      "loss": 0.1241,
      "step": 188
    },
    {
      "epoch": 0.5213793103448275,
      "grad_norm": 1.8110167173727552,
      "learning_rate": 1.115383875216598e-05,
      "loss": 0.0945,
      "step": 189
    },
    {
      "epoch": 0.5241379310344828,
      "grad_norm": 1.2395129394737805,
      "learning_rate": 1.105806242755723e-05,
      "loss": 0.0791,
      "step": 190
    },
    {
      "epoch": 0.526896551724138,
      "grad_norm": 2.269164321603063,
      "learning_rate": 1.0962187843917498e-05,
      "loss": 0.0995,
      "step": 191
    },
    {
      "epoch": 0.5296551724137931,
      "grad_norm": 2.1091615955126217,
      "learning_rate": 1.0866223904826992e-05,
      "loss": 0.0978,
      "step": 192
    },
    {
      "epoch": 0.5324137931034483,
      "grad_norm": 1.1661181979597077,
      "learning_rate": 1.0770179522164079e-05,
      "loss": 0.0654,
      "step": 193
    },
    {
      "epoch": 0.5351724137931034,
      "grad_norm": 1.529220319087916,
      "learning_rate": 1.0674063615277681e-05,
      "loss": 0.0795,
      "step": 194
    },
    {
      "epoch": 0.5379310344827586,
      "grad_norm": 1.61462843303553,
      "learning_rate": 1.0577885110158959e-05,
      "loss": 0.0722,
      "step": 195
    },
    {
      "epoch": 0.5406896551724137,
      "grad_norm": 1.0229094996985268,
      "learning_rate": 1.0481652938612374e-05,
      "loss": 0.1059,
      "step": 196
    },
    {
      "epoch": 0.543448275862069,
      "grad_norm": 1.8905308810375994,
      "learning_rate": 1.0385376037426227e-05,
      "loss": 0.1007,
      "step": 197
    },
    {
      "epoch": 0.5462068965517242,
      "grad_norm": 1.4659160591839386,
      "learning_rate": 1.0289063347542727e-05,
      "loss": 0.0997,
      "step": 198
    },
    {
      "epoch": 0.5489655172413793,
      "grad_norm": 3.3227958982974637,
      "learning_rate": 1.0192723813227672e-05,
      "loss": 0.1803,
      "step": 199
    },
    {
      "epoch": 0.5517241379310345,
      "grad_norm": 1.963767479554977,
      "learning_rate": 1.0096366381239808e-05,
      "loss": 0.1275,
      "step": 200
    },
    {
      "epoch": 0.5544827586206896,
      "grad_norm": 1.6295769776645304,
      "learning_rate": 1e-05,
      "loss": 0.0784,
      "step": 201
    },
    {
      "epoch": 0.5572413793103448,
      "grad_norm": 2.440021613726331,
      "learning_rate": 9.903633618760195e-06,
      "loss": 0.0884,
      "step": 202
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4242019780978539,
      "learning_rate": 9.807276186772335e-06,
      "loss": 0.0852,
      "step": 203
    },
    {
      "epoch": 0.5627586206896552,
      "grad_norm": 0.9963914089293635,
      "learning_rate": 9.710936652457276e-06,
      "loss": 0.0528,
      "step": 204
    },
    {
      "epoch": 0.5655172413793104,
      "grad_norm": 2.0171601668683863,
      "learning_rate": 9.614623962573776e-06,
      "loss": 0.1213,
      "step": 205
    },
    {
      "epoch": 0.5682758620689655,
      "grad_norm": 1.8302600165169711,
      "learning_rate": 9.518347061387629e-06,
      "loss": 0.0813,
      "step": 206
    },
    {
      "epoch": 0.5710344827586207,
      "grad_norm": 2.135574797850369,
      "learning_rate": 9.422114889841045e-06,
      "loss": 0.1114,
      "step": 207
    },
    {
      "epoch": 0.5737931034482758,
      "grad_norm": 1.856147430619363,
      "learning_rate": 9.325936384722322e-06,
      "loss": 0.1072,
      "step": 208
    },
    {
      "epoch": 0.576551724137931,
      "grad_norm": 1.3734175489711284,
      "learning_rate": 9.229820477835926e-06,
      "loss": 0.0629,
      "step": 209
    },
    {
      "epoch": 0.5793103448275863,
      "grad_norm": 1.4955582653472665,
      "learning_rate": 9.133776095173015e-06,
      "loss": 0.0798,
      "step": 210
    },
    {
      "epoch": 0.5820689655172414,
      "grad_norm": 2.2098641671152075,
      "learning_rate": 9.037812156082503e-06,
      "loss": 0.0832,
      "step": 211
    },
    {
      "epoch": 0.5848275862068966,
      "grad_norm": 2.1234204791453446,
      "learning_rate": 8.941937572442773e-06,
      "loss": 0.0912,
      "step": 212
    },
    {
      "epoch": 0.5875862068965517,
      "grad_norm": 1.7048969824409477,
      "learning_rate": 8.846161247834024e-06,
      "loss": 0.0757,
      "step": 213
    },
    {
      "epoch": 0.5903448275862069,
      "grad_norm": 1.2386625705533991,
      "learning_rate": 8.750492076711439e-06,
      "loss": 0.0607,
      "step": 214
    },
    {
      "epoch": 0.593103448275862,
      "grad_norm": 2.4577783979073518,
      "learning_rate": 8.654938943579194e-06,
      "loss": 0.1315,
      "step": 215
    },
    {
      "epoch": 0.5958620689655172,
      "grad_norm": 1.7822551401830198,
      "learning_rate": 8.55951072216536e-06,
      "loss": 0.1015,
      "step": 216
    },
    {
      "epoch": 0.5986206896551725,
      "grad_norm": 1.627034418887004,
      "learning_rate": 8.464216274597839e-06,
      "loss": 0.083,
      "step": 217
    },
    {
      "epoch": 0.6013793103448276,
      "grad_norm": 2.4236805021780907,
      "learning_rate": 8.369064450581374e-06,
      "loss": 0.144,
      "step": 218
    },
    {
      "epoch": 0.6041379310344828,
      "grad_norm": 2.0852606023844547,
      "learning_rate": 8.274064086575682e-06,
      "loss": 0.1187,
      "step": 219
    },
    {
      "epoch": 0.6068965517241379,
      "grad_norm": 1.554124619394613,
      "learning_rate": 8.179224004974857e-06,
      "loss": 0.101,
      "step": 220
    },
    {
      "epoch": 0.6096551724137931,
      "grad_norm": 1.0086858963161975,
      "learning_rate": 8.084553013288048e-06,
      "loss": 0.0562,
      "step": 221
    },
    {
      "epoch": 0.6124137931034482,
      "grad_norm": 1.744957411692858,
      "learning_rate": 7.990059903321554e-06,
      "loss": 0.1144,
      "step": 222
    },
    {
      "epoch": 0.6151724137931035,
      "grad_norm": 1.115741854823727,
      "learning_rate": 7.89575345036232e-06,
      "loss": 0.0431,
      "step": 223
    },
    {
      "epoch": 0.6179310344827587,
      "grad_norm": 2.4048467960954523,
      "learning_rate": 7.801642412363042e-06,
      "loss": 0.1239,
      "step": 224
    },
    {
      "epoch": 0.6206896551724138,
      "grad_norm": 1.7045994682708523,
      "learning_rate": 7.707735529128819e-06,
      "loss": 0.0861,
      "step": 225
    },
    {
      "epoch": 0.623448275862069,
      "grad_norm": 1.5844433019812807,
      "learning_rate": 7.614041521505517e-06,
      "loss": 0.0939,
      "step": 226
    },
    {
      "epoch": 0.6262068965517241,
      "grad_norm": 1.817274462365455,
      "learning_rate": 7.520569090569894e-06,
      "loss": 0.0961,
      "step": 227
    },
    {
      "epoch": 0.6289655172413793,
      "grad_norm": 2.169489761841419,
      "learning_rate": 7.427326916821557e-06,
      "loss": 0.1141,
      "step": 228
    },
    {
      "epoch": 0.6317241379310344,
      "grad_norm": 1.2611061121212517,
      "learning_rate": 7.3343236593768295e-06,
      "loss": 0.0631,
      "step": 229
    },
    {
      "epoch": 0.6344827586206897,
      "grad_norm": 1.4981316942094398,
      "learning_rate": 7.24156795516461e-06,
      "loss": 0.0607,
      "step": 230
    },
    {
      "epoch": 0.6372413793103449,
      "grad_norm": 1.889688470210782,
      "learning_rate": 7.149068418124281e-06,
      "loss": 0.1243,
      "step": 231
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6091621479461797,
      "learning_rate": 7.056833638405762e-06,
      "loss": 0.1154,
      "step": 232
    },
    {
      "epoch": 0.6427586206896552,
      "grad_norm": 1.7584988460897566,
      "learning_rate": 6.964872181571765e-06,
      "loss": 0.0782,
      "step": 233
    },
    {
      "epoch": 0.6455172413793103,
      "grad_norm": 1.9608264553670172,
      "learning_rate": 6.87319258780234e-06,
      "loss": 0.1088,
      "step": 234
    },
    {
      "epoch": 0.6482758620689655,
      "grad_norm": 1.1583598512074385,
      "learning_rate": 6.781803371101774e-06,
      "loss": 0.0635,
      "step": 235
    },
    {
      "epoch": 0.6510344827586206,
      "grad_norm": 1.5744270190667782,
      "learning_rate": 6.690713018507917e-06,
      "loss": 0.065,
      "step": 236
    },
    {
      "epoch": 0.6537931034482759,
      "grad_norm": 2.0489496115781147,
      "learning_rate": 6.599929989304034e-06,
      "loss": 0.0941,
      "step": 237
    },
    {
      "epoch": 0.6565517241379311,
      "grad_norm": 1.3833240601648478,
      "learning_rate": 6.509462714233194e-06,
      "loss": 0.1049,
      "step": 238
    },
    {
      "epoch": 0.6593103448275862,
      "grad_norm": 1.105761243006651,
      "learning_rate": 6.419319594715338e-06,
      "loss": 0.0795,
      "step": 239
    },
    {
      "epoch": 0.6620689655172414,
      "grad_norm": 2.2243011538195323,
      "learning_rate": 6.32950900206708e-06,
      "loss": 0.1389,
      "step": 240
    },
    {
      "epoch": 0.6648275862068965,
      "grad_norm": 1.6204675243320001,
      "learning_rate": 6.240039276724273e-06,
      "loss": 0.0904,
      "step": 241
    },
    {
      "epoch": 0.6675862068965517,
      "grad_norm": 1.7416445698806022,
      "learning_rate": 6.150918727467455e-06,
      "loss": 0.0988,
      "step": 242
    },
    {
      "epoch": 0.670344827586207,
      "grad_norm": 1.5865476897173794,
      "learning_rate": 6.062155630650265e-06,
      "loss": 0.0887,
      "step": 243
    },
    {
      "epoch": 0.6731034482758621,
      "grad_norm": 1.0908181878235528,
      "learning_rate": 5.973758229430806e-06,
      "loss": 0.0406,
      "step": 244
    },
    {
      "epoch": 0.6758620689655173,
      "grad_norm": 1.2409267373149415,
      "learning_rate": 5.8857347330061545e-06,
      "loss": 0.0613,
      "step": 245
    },
    {
      "epoch": 0.6786206896551724,
      "grad_norm": 1.5366160191792417,
      "learning_rate": 5.798093315849984e-06,
      "loss": 0.0621,
      "step": 246
    },
    {
      "epoch": 0.6813793103448276,
      "grad_norm": 2.401971583677582,
      "learning_rate": 5.7108421169534376e-06,
      "loss": 0.135,
      "step": 247
    },
    {
      "epoch": 0.6841379310344827,
      "grad_norm": 1.7382959732076737,
      "learning_rate": 5.623989239069275e-06,
      "loss": 0.0912,
      "step": 248
    },
    {
      "epoch": 0.6868965517241379,
      "grad_norm": 1.1164138593496515,
      "learning_rate": 5.5375427479593945e-06,
      "loss": 0.0481,
      "step": 249
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 1.9207044725285578,
      "learning_rate": 5.451510671645806e-06,
      "loss": 0.0915,
      "step": 250
    },
    {
      "epoch": 0.6924137931034483,
      "grad_norm": 1.4554330468757273,
      "learning_rate": 5.3659009996650704e-06,
      "loss": 0.0932,
      "step": 251
    },
    {
      "epoch": 0.6951724137931035,
      "grad_norm": 1.5637751353358034,
      "learning_rate": 5.280721682326349e-06,
      "loss": 0.1258,
      "step": 252
    },
    {
      "epoch": 0.6979310344827586,
      "grad_norm": 1.6864022667103549,
      "learning_rate": 5.195980629973077e-06,
      "loss": 0.136,
      "step": 253
    },
    {
      "epoch": 0.7006896551724138,
      "grad_norm": 1.5424417657042435,
      "learning_rate": 5.111685712248364e-06,
      "loss": 0.0744,
      "step": 254
    },
    {
      "epoch": 0.7034482758620689,
      "grad_norm": 2.4152541945255965,
      "learning_rate": 5.02784475736415e-06,
      "loss": 0.1436,
      "step": 255
    },
    {
      "epoch": 0.7062068965517241,
      "grad_norm": 1.8800510082119968,
      "learning_rate": 4.944465551374238e-06,
      "loss": 0.0864,
      "step": 256
    },
    {
      "epoch": 0.7089655172413794,
      "grad_norm": 1.6391965299287088,
      "learning_rate": 4.861555837451213e-06,
      "loss": 0.0931,
      "step": 257
    },
    {
      "epoch": 0.7117241379310345,
      "grad_norm": 2.1164230178199688,
      "learning_rate": 4.779123315167362e-06,
      "loss": 0.1589,
      "step": 258
    },
    {
      "epoch": 0.7144827586206897,
      "grad_norm": 2.1736287568680615,
      "learning_rate": 4.6971756397796506e-06,
      "loss": 0.075,
      "step": 259
    },
    {
      "epoch": 0.7172413793103448,
      "grad_norm": 0.981201558647046,
      "learning_rate": 4.61572042151878e-06,
      "loss": 0.0423,
      "step": 260
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.2068505093305162,
      "learning_rate": 4.534765224882463e-06,
      "loss": 0.0538,
      "step": 261
    },
    {
      "epoch": 0.7227586206896551,
      "grad_norm": 0.9541230040360105,
      "learning_rate": 4.4543175679329345e-06,
      "loss": 0.0591,
      "step": 262
    },
    {
      "epoch": 0.7255172413793104,
      "grad_norm": 1.7162320823485342,
      "learning_rate": 4.37438492159876e-06,
      "loss": 0.0965,
      "step": 263
    },
    {
      "epoch": 0.7282758620689656,
      "grad_norm": 1.184768855122373,
      "learning_rate": 4.294974708981041e-06,
      "loss": 0.0383,
      "step": 264
    },
    {
      "epoch": 0.7310344827586207,
      "grad_norm": 1.5098290827125584,
      "learning_rate": 4.216094304664056e-06,
      "loss": 0.0952,
      "step": 265
    },
    {
      "epoch": 0.7337931034482759,
      "grad_norm": 1.3641276614652302,
      "learning_rate": 4.1377510340304e-06,
      "loss": 0.0687,
      "step": 266
    },
    {
      "epoch": 0.736551724137931,
      "grad_norm": 1.9886328182742048,
      "learning_rate": 4.059952172580694e-06,
      "loss": 0.1397,
      "step": 267
    },
    {
      "epoch": 0.7393103448275862,
      "grad_norm": 2.0623334755221494,
      "learning_rate": 3.982704945257957e-06,
      "loss": 0.1271,
      "step": 268
    },
    {
      "epoch": 0.7420689655172413,
      "grad_norm": 2.027142606530138,
      "learning_rate": 3.9060165257766116e-06,
      "loss": 0.0736,
      "step": 269
    },
    {
      "epoch": 0.7448275862068966,
      "grad_norm": 1.8884432574197467,
      "learning_rate": 3.829894035956306e-06,
      "loss": 0.1243,
      "step": 270
    },
    {
      "epoch": 0.7475862068965518,
      "grad_norm": 1.5259208901463874,
      "learning_rate": 3.754344545060529e-06,
      "loss": 0.0971,
      "step": 271
    },
    {
      "epoch": 0.7503448275862069,
      "grad_norm": 1.454648528529707,
      "learning_rate": 3.6793750691400996e-06,
      "loss": 0.1226,
      "step": 272
    },
    {
      "epoch": 0.7531034482758621,
      "grad_norm": 1.1852288533881499,
      "learning_rate": 3.604992570381621e-06,
      "loss": 0.0573,
      "step": 273
    },
    {
      "epoch": 0.7558620689655172,
      "grad_norm": 0.9888790128432269,
      "learning_rate": 3.5312039564609203e-06,
      "loss": 0.0357,
      "step": 274
    },
    {
      "epoch": 0.7586206896551724,
      "grad_norm": 1.9227683916123242,
      "learning_rate": 3.458016079901544e-06,
      "loss": 0.0841,
      "step": 275
    },
    {
      "epoch": 0.7613793103448275,
      "grad_norm": 1.6925865088399465,
      "learning_rate": 3.3854357374383905e-06,
      "loss": 0.0749,
      "step": 276
    },
    {
      "epoch": 0.7641379310344828,
      "grad_norm": 2.482457504595889,
      "learning_rate": 3.313469669386532e-06,
      "loss": 0.1234,
      "step": 277
    },
    {
      "epoch": 0.766896551724138,
      "grad_norm": 1.2830007481095762,
      "learning_rate": 3.242124559015234e-06,
      "loss": 0.0752,
      "step": 278
    },
    {
      "epoch": 0.7696551724137931,
      "grad_norm": 2.0768576257082114,
      "learning_rate": 3.171407031927325e-06,
      "loss": 0.0991,
      "step": 279
    },
    {
      "epoch": 0.7724137931034483,
      "grad_norm": 1.8987466668093202,
      "learning_rate": 3.101323655443882e-06,
      "loss": 0.109,
      "step": 280
    },
    {
      "epoch": 0.7751724137931034,
      "grad_norm": 1.6001709673447158,
      "learning_rate": 3.0318809379943594e-06,
      "loss": 0.0599,
      "step": 281
    },
    {
      "epoch": 0.7779310344827586,
      "grad_norm": 1.309542383665605,
      "learning_rate": 2.9630853285121506e-06,
      "loss": 0.0641,
      "step": 282
    },
    {
      "epoch": 0.7806896551724138,
      "grad_norm": 1.5565630583605414,
      "learning_rate": 2.8949432158357083e-06,
      "loss": 0.0909,
      "step": 283
    },
    {
      "epoch": 0.783448275862069,
      "grad_norm": 1.2776706094735155,
      "learning_rate": 2.8274609281152322e-06,
      "loss": 0.1223,
      "step": 284
    },
    {
      "epoch": 0.7862068965517242,
      "grad_norm": 1.6300368148462467,
      "learning_rate": 2.7606447322249876e-06,
      "loss": 0.087,
      "step": 285
    },
    {
      "epoch": 0.7889655172413793,
      "grad_norm": 1.5096693670735182,
      "learning_rate": 2.694500833181323e-06,
      "loss": 0.1002,
      "step": 286
    },
    {
      "epoch": 0.7917241379310345,
      "grad_norm": 1.719399247233759,
      "learning_rate": 2.629035373566433e-06,
      "loss": 0.0915,
      "step": 287
    },
    {
      "epoch": 0.7944827586206896,
      "grad_norm": 1.1425139227274304,
      "learning_rate": 2.5642544329579088e-06,
      "loss": 0.0555,
      "step": 288
    },
    {
      "epoch": 0.7972413793103448,
      "grad_norm": 1.2082856945935607,
      "learning_rate": 2.500164027364147e-06,
      "loss": 0.0555,
      "step": 289
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.028172728126609,
      "learning_rate": 2.4367701086656625e-06,
      "loss": 0.104,
      "step": 290
    },
    {
      "epoch": 0.8027586206896552,
      "grad_norm": 1.498164820422529,
      "learning_rate": 2.374078564062364e-06,
      "loss": 0.0936,
      "step": 291
    },
    {
      "epoch": 0.8055172413793104,
      "grad_norm": 1.3254283577777912,
      "learning_rate": 2.312095215526814e-06,
      "loss": 0.0898,
      "step": 292
    },
    {
      "epoch": 0.8082758620689655,
      "grad_norm": 1.3758428200048072,
      "learning_rate": 2.2508258192635614e-06,
      "loss": 0.073,
      "step": 293
    },
    {
      "epoch": 0.8110344827586207,
      "grad_norm": 1.2825112587820704,
      "learning_rate": 2.190276065174596e-06,
      "loss": 0.0675,
      "step": 294
    },
    {
      "epoch": 0.8137931034482758,
      "grad_norm": 1.704844614821693,
      "learning_rate": 2.130451576330925e-06,
      "loss": 0.0871,
      "step": 295
    },
    {
      "epoch": 0.8165517241379311,
      "grad_norm": 1.8236093781558738,
      "learning_rate": 2.0713579084503877e-06,
      "loss": 0.075,
      "step": 296
    },
    {
      "epoch": 0.8193103448275862,
      "grad_norm": 1.7159210186184939,
      "learning_rate": 2.0130005493817063e-06,
      "loss": 0.0726,
      "step": 297
    },
    {
      "epoch": 0.8220689655172414,
      "grad_norm": 1.5402879375212146,
      "learning_rate": 1.9553849185948514e-06,
      "loss": 0.0585,
      "step": 298
    },
    {
      "epoch": 0.8248275862068966,
      "grad_norm": 2.088587364963122,
      "learning_rate": 1.8985163666777473e-06,
      "loss": 0.1192,
      "step": 299
    },
    {
      "epoch": 0.8275862068965517,
      "grad_norm": 1.385102288521542,
      "learning_rate": 1.8424001748393905e-06,
      "loss": 0.0639,
      "step": 300
    },
    {
      "epoch": 0.8303448275862069,
      "grad_norm": 1.199103692035072,
      "learning_rate": 1.7870415544193808e-06,
      "loss": 0.0644,
      "step": 301
    },
    {
      "epoch": 0.833103448275862,
      "grad_norm": 1.576550399440244,
      "learning_rate": 1.7324456464039751e-06,
      "loss": 0.0806,
      "step": 302
    },
    {
      "epoch": 0.8358620689655173,
      "grad_norm": 1.9033469567963237,
      "learning_rate": 1.6786175209486565e-06,
      "loss": 0.1166,
      "step": 303
    },
    {
      "epoch": 0.8386206896551724,
      "grad_norm": 1.5481679752062283,
      "learning_rate": 1.6255621769072805e-06,
      "loss": 0.0883,
      "step": 304
    },
    {
      "epoch": 0.8413793103448276,
      "grad_norm": 1.7374538978001977,
      "learning_rate": 1.5732845413678477e-06,
      "loss": 0.105,
      "step": 305
    },
    {
      "epoch": 0.8441379310344828,
      "grad_norm": 1.3465892642582866,
      "learning_rate": 1.521789469194952e-06,
      "loss": 0.0618,
      "step": 306
    },
    {
      "epoch": 0.8468965517241379,
      "grad_norm": 2.312489005340463,
      "learning_rate": 1.4710817425789015e-06,
      "loss": 0.0992,
      "step": 307
    },
    {
      "epoch": 0.8496551724137931,
      "grad_norm": 1.1318530850342379,
      "learning_rate": 1.4211660705916286e-06,
      "loss": 0.0458,
      "step": 308
    },
    {
      "epoch": 0.8524137931034482,
      "grad_norm": 1.5063935617388766,
      "learning_rate": 1.372047088749372e-06,
      "loss": 0.0516,
      "step": 309
    },
    {
      "epoch": 0.8551724137931035,
      "grad_norm": 1.4001415457936668,
      "learning_rate": 1.3237293585821786e-06,
      "loss": 0.1156,
      "step": 310
    },
    {
      "epoch": 0.8579310344827586,
      "grad_norm": 2.487040806365276,
      "learning_rate": 1.2762173672102996e-06,
      "loss": 0.1332,
      "step": 311
    },
    {
      "epoch": 0.8606896551724138,
      "grad_norm": 1.514447174356807,
      "learning_rate": 1.2295155269274827e-06,
      "loss": 0.0735,
      "step": 312
    },
    {
      "epoch": 0.863448275862069,
      "grad_norm": 1.9664878278885487,
      "learning_rate": 1.1836281747912125e-06,
      "loss": 0.0825,
      "step": 313
    },
    {
      "epoch": 0.8662068965517241,
      "grad_norm": 1.7248118984472842,
      "learning_rate": 1.1385595722199438e-06,
      "loss": 0.0945,
      "step": 314
    },
    {
      "epoch": 0.8689655172413793,
      "grad_norm": 1.232176840002336,
      "learning_rate": 1.094313904597355e-06,
      "loss": 0.0761,
      "step": 315
    },
    {
      "epoch": 0.8717241379310345,
      "grad_norm": 2.3846939660082636,
      "learning_rate": 1.0508952808836682e-06,
      "loss": 0.1104,
      "step": 316
    },
    {
      "epoch": 0.8744827586206897,
      "grad_norm": 1.5107849502345858,
      "learning_rate": 1.0083077332340563e-06,
      "loss": 0.137,
      "step": 317
    },
    {
      "epoch": 0.8772413793103448,
      "grad_norm": 2.5218758909483077,
      "learning_rate": 9.665552166241965e-07,
      "loss": 0.1395,
      "step": 318
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.913603235284768,
      "learning_rate": 9.256416084829778e-07,
      "loss": 0.1331,
      "step": 319
    },
    {
      "epoch": 0.8827586206896552,
      "grad_norm": 2.4049905043471806,
      "learning_rate": 8.855707083324183e-07,
      "loss": 0.1077,
      "step": 320
    },
    {
      "epoch": 0.8855172413793103,
      "grad_norm": 1.108589495662786,
      "learning_rate": 8.46346237434813e-07,
      "loss": 0.0638,
      "step": 321
    },
    {
      "epoch": 0.8882758620689655,
      "grad_norm": 1.222025656507595,
      "learning_rate": 8.079718384471557e-07,
      "loss": 0.0515,
      "step": 322
    },
    {
      "epoch": 0.8910344827586207,
      "grad_norm": 1.6319377780473996,
      "learning_rate": 7.704510750828542e-07,
      "loss": 0.0623,
      "step": 323
    },
    {
      "epoch": 0.8937931034482759,
      "grad_norm": 1.402983153361783,
      "learning_rate": 7.337874317807803e-07,
      "loss": 0.0516,
      "step": 324
    },
    {
      "epoch": 0.896551724137931,
      "grad_norm": 1.4500270750731776,
      "learning_rate": 6.979843133816744e-07,
      "loss": 0.0612,
      "step": 325
    },
    {
      "epoch": 0.8993103448275862,
      "grad_norm": 1.095211249638046,
      "learning_rate": 6.630450448119618e-07,
      "loss": 0.0333,
      "step": 326
    },
    {
      "epoch": 0.9020689655172414,
      "grad_norm": 1.8925065532997027,
      "learning_rate": 6.289728707749609e-07,
      "loss": 0.0953,
      "step": 327
    },
    {
      "epoch": 0.9048275862068965,
      "grad_norm": 1.747624078137272,
      "learning_rate": 5.957709554495683e-07,
      "loss": 0.0893,
      "step": 328
    },
    {
      "epoch": 0.9075862068965517,
      "grad_norm": 0.9113610050884592,
      "learning_rate": 5.634423821964074e-07,
      "loss": 0.0423,
      "step": 329
    },
    {
      "epoch": 0.9103448275862069,
      "grad_norm": 1.6463257868897792,
      "learning_rate": 5.319901532714877e-07,
      "loss": 0.1055,
      "step": 330
    },
    {
      "epoch": 0.9131034482758621,
      "grad_norm": 1.84950657136513,
      "learning_rate": 5.014171895473929e-07,
      "loss": 0.1232,
      "step": 331
    },
    {
      "epoch": 0.9158620689655173,
      "grad_norm": 1.5561236996647523,
      "learning_rate": 4.717263302420283e-07,
      "loss": 0.076,
      "step": 332
    },
    {
      "epoch": 0.9186206896551724,
      "grad_norm": 1.2285714349711996,
      "learning_rate": 4.429203326549525e-07,
      "loss": 0.0835,
      "step": 333
    },
    {
      "epoch": 0.9213793103448276,
      "grad_norm": 2.8092718918728288,
      "learning_rate": 4.150018719113147e-07,
      "loss": 0.0734,
      "step": 334
    },
    {
      "epoch": 0.9241379310344827,
      "grad_norm": 1.766273016183224,
      "learning_rate": 3.8797354071342443e-07,
      "loss": 0.0768,
      "step": 335
    },
    {
      "epoch": 0.926896551724138,
      "grad_norm": 2.1729531887535285,
      "learning_rate": 3.618378490999719e-07,
      "loss": 0.1027,
      "step": 336
    },
    {
      "epoch": 0.9296551724137931,
      "grad_norm": 1.7482937409483954,
      "learning_rate": 3.365972242129378e-07,
      "loss": 0.1176,
      "step": 337
    },
    {
      "epoch": 0.9324137931034483,
      "grad_norm": 1.2860014321281648,
      "learning_rate": 3.122540100721794e-07,
      "loss": 0.068,
      "step": 338
    },
    {
      "epoch": 0.9351724137931035,
      "grad_norm": 1.8682982161376196,
      "learning_rate": 2.888104673577574e-07,
      "loss": 0.0826,
      "step": 339
    },
    {
      "epoch": 0.9379310344827586,
      "grad_norm": 2.1192579972688548,
      "learning_rate": 2.66268773199988e-07,
      "loss": 0.1094,
      "step": 340
    },
    {
      "epoch": 0.9406896551724138,
      "grad_norm": 1.812620322889224,
      "learning_rate": 2.4463102097726843e-07,
      "loss": 0.1397,
      "step": 341
    },
    {
      "epoch": 0.9434482758620689,
      "grad_norm": 1.55323368762906,
      "learning_rate": 2.2389922012165944e-07,
      "loss": 0.0883,
      "step": 342
    },
    {
      "epoch": 0.9462068965517242,
      "grad_norm": 1.3537994599803411,
      "learning_rate": 2.0407529593228114e-07,
      "loss": 0.0537,
      "step": 343
    },
    {
      "epoch": 0.9489655172413793,
      "grad_norm": 1.9963802514280435,
      "learning_rate": 1.8516108939651945e-07,
      "loss": 0.0841,
      "step": 344
    },
    {
      "epoch": 0.9517241379310345,
      "grad_norm": 1.9489250790151131,
      "learning_rate": 1.6715835701905604e-07,
      "loss": 0.1278,
      "step": 345
    },
    {
      "epoch": 0.9544827586206897,
      "grad_norm": 1.208403698921063,
      "learning_rate": 1.5006877065874338e-07,
      "loss": 0.062,
      "step": 346
    },
    {
      "epoch": 0.9572413793103448,
      "grad_norm": 1.9902454269295837,
      "learning_rate": 1.3389391737335112e-07,
      "loss": 0.1123,
      "step": 347
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.5549649154456184,
      "learning_rate": 1.1863529927217731e-07,
      "loss": 0.0859,
      "step": 348
    },
    {
      "epoch": 0.9627586206896551,
      "grad_norm": 1.5672800792400794,
      "learning_rate": 1.0429433337655115e-07,
      "loss": 0.0664,
      "step": 349
    },
    {
      "epoch": 0.9655172413793104,
      "grad_norm": 1.980266389039632,
      "learning_rate": 9.08723514882437e-08,
      "loss": 0.0978,
      "step": 350
    },
    {
      "epoch": 0.9682758620689655,
      "grad_norm": 1.924180909114094,
      "learning_rate": 7.837060006577801e-08,
      "loss": 0.082,
      "step": 351
    },
    {
      "epoch": 0.9710344827586207,
      "grad_norm": 1.360244814652006,
      "learning_rate": 6.679024010868617e-08,
      "loss": 0.0937,
      "step": 352
    },
    {
      "epoch": 0.9737931034482759,
      "grad_norm": 1.3638474049460683,
      "learning_rate": 5.6132347049679955e-08,
      "loss": 0.0944,
      "step": 353
    },
    {
      "epoch": 0.976551724137931,
      "grad_norm": 1.0043906816455541,
      "learning_rate": 4.639791065478738e-08,
      "loss": 0.0373,
      "step": 354
    },
    {
      "epoch": 0.9793103448275862,
      "grad_norm": 1.3136551040176567,
      "learning_rate": 3.758783493142737e-08,
      "loss": 0.0938,
      "step": 355
    },
    {
      "epoch": 0.9820689655172414,
      "grad_norm": 1.2330770172633143,
      "learning_rate": 2.9702938044468e-08,
      "loss": 0.1083,
      "step": 356
    },
    {
      "epoch": 0.9848275862068966,
      "grad_norm": 1.4318266541987408,
      "learning_rate": 2.274395224023618e-08,
      "loss": 0.0667,
      "step": 357
    },
    {
      "epoch": 0.9875862068965517,
      "grad_norm": 1.4889881534099365,
      "learning_rate": 1.671152377852092e-08,
      "loss": 0.0907,
      "step": 358
    },
    {
      "epoch": 0.9903448275862069,
      "grad_norm": 1.9950572073445185,
      "learning_rate": 1.1606212872559142e-08,
      "loss": 0.0937,
      "step": 359
    },
    {
      "epoch": 0.993103448275862,
      "grad_norm": 1.2536334792292063,
      "learning_rate": 7.42849363700282e-09,
      "loss": 0.0863,
      "step": 360
    },
    {
      "epoch": 0.9958620689655172,
      "grad_norm": 1.2816200143172458,
      "learning_rate": 4.178754043898669e-09,
      "loss": 0.0593,
      "step": 361
    },
    {
      "epoch": 0.9986206896551724,
      "grad_norm": 0.7637431538197891,
      "learning_rate": 1.8572958866514e-09,
      "loss": 0.0859,
      "step": 362
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.3907025925168472,
      "learning_rate": 4.643347520005836e-10,
      "loss": 0.0658,
      "step": 363
    },
    {
      "epoch": 1.0,
      "step": 363,
      "total_flos": 50754799140864.0,
      "train_loss": 0.09636665488652289,
      "train_runtime": 845.3624,
      "train_samples_per_second": 6.855,
      "train_steps_per_second": 0.429
    }
  ],
  "logging_steps": 1,
  "max_steps": 363,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 50754799140864.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}