{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8833271991166728,
  "eval_steps": 500,
  "global_step": 600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001472211998527788,
      "grad_norm": 0.33429867029190063,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.8228,
      "step": 1
    },
    {
      "epoch": 0.002944423997055576,
      "grad_norm": 0.36462727189064026,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.7988,
      "step": 2
    },
    {
      "epoch": 0.004416635995583364,
      "grad_norm": 0.3369121253490448,
      "learning_rate": 6e-06,
      "loss": 1.8082,
      "step": 3
    },
    {
      "epoch": 0.005888847994111152,
      "grad_norm": 0.33166587352752686,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.8245,
      "step": 4
    },
    {
      "epoch": 0.00736105999263894,
      "grad_norm": 0.3309018015861511,
      "learning_rate": 1e-05,
      "loss": 1.8154,
      "step": 5
    },
    {
      "epoch": 0.008833271991166729,
      "grad_norm": 0.3230971097946167,
      "learning_rate": 1.2e-05,
      "loss": 1.8277,
      "step": 6
    },
    {
      "epoch": 0.010305483989694516,
      "grad_norm": 0.3316158950328827,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 1.8233,
      "step": 7
    },
    {
      "epoch": 0.011777695988222304,
      "grad_norm": 0.3074355721473694,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.8251,
      "step": 8
    },
    {
      "epoch": 0.013249907986750091,
      "grad_norm": 0.3094817101955414,
      "learning_rate": 1.8e-05,
      "loss": 1.8244,
      "step": 9
    },
    {
      "epoch": 0.01472211998527788,
      "grad_norm": 0.30950072407722473,
      "learning_rate": 2e-05,
      "loss": 1.8118,
      "step": 10
    },
    {
      "epoch": 0.016194331983805668,
      "grad_norm": 0.28165385127067566,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 1.8374,
      "step": 11
    },
    {
      "epoch": 0.017666543982333457,
      "grad_norm": 0.25515854358673096,
      "learning_rate": 2.4e-05,
      "loss": 1.7765,
      "step": 12
    },
    {
      "epoch": 0.019138755980861243,
      "grad_norm": 0.2388654500246048,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 1.7686,
      "step": 13
    },
    {
      "epoch": 0.020610967979389033,
      "grad_norm": 0.22097326815128326,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 1.7423,
      "step": 14
    },
    {
      "epoch": 0.02208317997791682,
      "grad_norm": 0.19228965044021606,
      "learning_rate": 3e-05,
      "loss": 1.7438,
      "step": 15
    },
    {
      "epoch": 0.023555391976444608,
      "grad_norm": 0.18450209498405457,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 1.8003,
      "step": 16
    },
    {
      "epoch": 0.025027603974972397,
      "grad_norm": 0.17064838111400604,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 1.7162,
      "step": 17
    },
    {
      "epoch": 0.026499815973500183,
      "grad_norm": 0.1780705749988556,
      "learning_rate": 3.6e-05,
      "loss": 1.7139,
      "step": 18
    },
    {
      "epoch": 0.027972027972027972,
      "grad_norm": 0.1894058734178543,
      "learning_rate": 3.8e-05,
      "loss": 1.7546,
      "step": 19
    },
    {
      "epoch": 0.02944423997055576,
      "grad_norm": 0.18374145030975342,
      "learning_rate": 4e-05,
      "loss": 1.7763,
      "step": 20
    },
    {
      "epoch": 0.030916451969083547,
      "grad_norm": 0.17798911035060883,
      "learning_rate": 4.2e-05,
      "loss": 1.7508,
      "step": 21
    },
    {
      "epoch": 0.032388663967611336,
      "grad_norm": 0.1605505496263504,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.7107,
      "step": 22
    },
    {
      "epoch": 0.033860875966139126,
      "grad_norm": 0.1549708992242813,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.7292,
      "step": 23
    },
    {
      "epoch": 0.035333087964666915,
      "grad_norm": 0.14511638879776,
      "learning_rate": 4.8e-05,
      "loss": 1.7095,
      "step": 24
    },
    {
      "epoch": 0.0368052999631947,
      "grad_norm": 0.14705853164196014,
      "learning_rate": 5e-05,
      "loss": 1.7171,
      "step": 25
    },
    {
      "epoch": 0.03827751196172249,
      "grad_norm": 0.14739441871643066,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 1.683,
      "step": 26
    },
    {
      "epoch": 0.039749723960250276,
      "grad_norm": 0.16744551062583923,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 1.6609,
      "step": 27
    },
    {
      "epoch": 0.041221935958778065,
      "grad_norm": 0.17307262122631073,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 1.6956,
      "step": 28
    },
    {
      "epoch": 0.042694147957305854,
      "grad_norm": 0.19107450544834137,
      "learning_rate": 5.8e-05,
      "loss": 1.6701,
      "step": 29
    },
    {
      "epoch": 0.04416635995583364,
      "grad_norm": 0.16990923881530762,
      "learning_rate": 6e-05,
      "loss": 1.6794,
      "step": 30
    },
    {
      "epoch": 0.045638571954361426,
      "grad_norm": 0.18232804536819458,
      "learning_rate": 6.2e-05,
      "loss": 1.6426,
      "step": 31
    },
    {
      "epoch": 0.047110783952889215,
      "grad_norm": 0.18065060675144196,
      "learning_rate": 6.400000000000001e-05,
      "loss": 1.6751,
      "step": 32
    },
    {
      "epoch": 0.048582995951417005,
      "grad_norm": 0.17246262729167938,
      "learning_rate": 6.6e-05,
      "loss": 1.6627,
      "step": 33
    },
    {
      "epoch": 0.050055207949944794,
      "grad_norm": 0.1384831815958023,
      "learning_rate": 6.800000000000001e-05,
      "loss": 1.6444,
      "step": 34
    },
    {
      "epoch": 0.05152741994847258,
      "grad_norm": 0.15911152958869934,
      "learning_rate": 7e-05,
      "loss": 1.6706,
      "step": 35
    },
    {
      "epoch": 0.052999631947000365,
      "grad_norm": 0.16192561388015747,
      "learning_rate": 7.2e-05,
      "loss": 1.6637,
      "step": 36
    },
    {
      "epoch": 0.054471843945528155,
      "grad_norm": 0.18282459676265717,
      "learning_rate": 7.4e-05,
      "loss": 1.6537,
      "step": 37
    },
    {
      "epoch": 0.055944055944055944,
      "grad_norm": 0.16843295097351074,
      "learning_rate": 7.6e-05,
      "loss": 1.6721,
      "step": 38
    },
    {
      "epoch": 0.05741626794258373,
      "grad_norm": 0.15258659422397614,
      "learning_rate": 7.800000000000001e-05,
      "loss": 1.6398,
      "step": 39
    },
    {
      "epoch": 0.05888847994111152,
      "grad_norm": 0.12701424956321716,
      "learning_rate": 8e-05,
      "loss": 1.6641,
      "step": 40
    },
    {
      "epoch": 0.060360691939639305,
      "grad_norm": 0.13082149624824524,
      "learning_rate": 8.2e-05,
      "loss": 1.6672,
      "step": 41
    },
    {
      "epoch": 0.061832903938167094,
      "grad_norm": 0.13053245842456818,
      "learning_rate": 8.4e-05,
      "loss": 1.644,
      "step": 42
    },
    {
      "epoch": 0.06330511593669488,
      "grad_norm": 0.15370412170886993,
      "learning_rate": 8.6e-05,
      "loss": 1.6594,
      "step": 43
    },
    {
      "epoch": 0.06477732793522267,
      "grad_norm": 0.12709803879261017,
      "learning_rate": 8.800000000000001e-05,
      "loss": 1.6529,
      "step": 44
    },
    {
      "epoch": 0.06624953993375046,
      "grad_norm": 0.1236925944685936,
      "learning_rate": 9e-05,
      "loss": 1.6452,
      "step": 45
    },
    {
      "epoch": 0.06772175193227825,
      "grad_norm": 0.12472256273031235,
      "learning_rate": 9.200000000000001e-05,
      "loss": 1.6597,
      "step": 46
    },
    {
      "epoch": 0.06919396393080604,
      "grad_norm": 0.12761995196342468,
      "learning_rate": 9.4e-05,
      "loss": 1.6371,
      "step": 47
    },
    {
      "epoch": 0.07066617592933383,
      "grad_norm": 0.11738269031047821,
      "learning_rate": 9.6e-05,
      "loss": 1.623,
      "step": 48
    },
    {
      "epoch": 0.0721383879278616,
      "grad_norm": 0.11436701565980911,
      "learning_rate": 9.8e-05,
      "loss": 1.6148,
      "step": 49
    },
    {
      "epoch": 0.0736105999263894,
      "grad_norm": 0.10698351263999939,
      "learning_rate": 0.0001,
      "loss": 1.6027,
      "step": 50
    },
    {
      "epoch": 0.07508281192491718,
      "grad_norm": 0.11186650395393372,
      "learning_rate": 0.00010200000000000001,
      "loss": 1.5946,
      "step": 51
    },
    {
      "epoch": 0.07655502392344497,
      "grad_norm": 0.11112627387046814,
      "learning_rate": 0.00010400000000000001,
      "loss": 1.6015,
      "step": 52
    },
    {
      "epoch": 0.07802723592197276,
      "grad_norm": 0.11840241402387619,
      "learning_rate": 0.00010600000000000002,
      "loss": 1.6208,
      "step": 53
    },
    {
      "epoch": 0.07949944792050055,
      "grad_norm": 0.11877040565013885,
      "learning_rate": 0.00010800000000000001,
      "loss": 1.5988,
      "step": 54
    },
    {
      "epoch": 0.08097165991902834,
      "grad_norm": 0.10934694856405258,
      "learning_rate": 0.00011000000000000002,
      "loss": 1.5885,
      "step": 55
    },
    {
      "epoch": 0.08244387191755613,
      "grad_norm": 0.11492514610290527,
      "learning_rate": 0.00011200000000000001,
      "loss": 1.6151,
      "step": 56
    },
    {
      "epoch": 0.08391608391608392,
      "grad_norm": 0.1252240389585495,
      "learning_rate": 0.00011399999999999999,
      "loss": 1.6192,
      "step": 57
    },
    {
      "epoch": 0.08538829591461171,
      "grad_norm": 0.11753290891647339,
      "learning_rate": 0.000116,
      "loss": 1.6421,
      "step": 58
    },
    {
      "epoch": 0.0868605079131395,
      "grad_norm": 0.11350328475236893,
      "learning_rate": 0.000118,
      "loss": 1.6151,
      "step": 59
    },
    {
      "epoch": 0.08833271991166727,
      "grad_norm": 0.1290436089038849,
      "learning_rate": 0.00012,
      "loss": 1.5855,
      "step": 60
    },
    {
      "epoch": 0.08980493191019506,
      "grad_norm": 0.12414577603340149,
      "learning_rate": 0.000122,
      "loss": 1.6219,
      "step": 61
    },
    {
      "epoch": 0.09127714390872285,
      "grad_norm": 0.1235835999250412,
      "learning_rate": 0.000124,
      "loss": 1.5546,
      "step": 62
    },
    {
      "epoch": 0.09274935590725064,
      "grad_norm": 0.11756890267133713,
      "learning_rate": 0.000126,
      "loss": 1.553,
      "step": 63
    },
    {
      "epoch": 0.09422156790577843,
      "grad_norm": 0.11887472122907639,
      "learning_rate": 0.00012800000000000002,
      "loss": 1.6016,
      "step": 64
    },
    {
      "epoch": 0.09569377990430622,
      "grad_norm": 0.12895843386650085,
      "learning_rate": 0.00013000000000000002,
      "loss": 1.6052,
      "step": 65
    },
    {
      "epoch": 0.09716599190283401,
      "grad_norm": 0.12211526930332184,
      "learning_rate": 0.000132,
      "loss": 1.5857,
      "step": 66
    },
    {
      "epoch": 0.0986382039013618,
      "grad_norm": 0.11842519044876099,
      "learning_rate": 0.000134,
      "loss": 1.5881,
      "step": 67
    },
    {
      "epoch": 0.10011041589988959,
      "grad_norm": 0.12883660197257996,
      "learning_rate": 0.00013600000000000003,
      "loss": 1.6071,
      "step": 68
    },
    {
      "epoch": 0.10158262789841738,
      "grad_norm": 0.12024354934692383,
      "learning_rate": 0.000138,
      "loss": 1.6027,
      "step": 69
    },
    {
      "epoch": 0.10305483989694517,
      "grad_norm": 0.1253923773765564,
      "learning_rate": 0.00014,
      "loss": 1.625,
      "step": 70
    },
    {
      "epoch": 0.10452705189547294,
      "grad_norm": 0.12402202934026718,
      "learning_rate": 0.000142,
      "loss": 1.5773,
      "step": 71
    },
    {
      "epoch": 0.10599926389400073,
      "grad_norm": 0.1243094876408577,
      "learning_rate": 0.000144,
      "loss": 1.5523,
      "step": 72
    },
    {
      "epoch": 0.10747147589252852,
      "grad_norm": 0.12207358330488205,
      "learning_rate": 0.000146,
      "loss": 1.5869,
      "step": 73
    },
    {
      "epoch": 0.10894368789105631,
      "grad_norm": 0.13517668843269348,
      "learning_rate": 0.000148,
      "loss": 1.5948,
      "step": 74
    },
    {
      "epoch": 0.1104158998895841,
      "grad_norm": 0.1349286139011383,
      "learning_rate": 0.00015000000000000001,
      "loss": 1.591,
      "step": 75
    },
    {
      "epoch": 0.11188811188811189,
      "grad_norm": 0.12710173428058624,
      "learning_rate": 0.000152,
      "loss": 1.5743,
      "step": 76
    },
    {
      "epoch": 0.11336032388663968,
      "grad_norm": 0.13487382233142853,
      "learning_rate": 0.000154,
      "loss": 1.5994,
      "step": 77
    },
    {
      "epoch": 0.11483253588516747,
      "grad_norm": 0.1304217427968979,
      "learning_rate": 0.00015600000000000002,
      "loss": 1.5785,
      "step": 78
    },
    {
      "epoch": 0.11630474788369526,
      "grad_norm": 0.12587234377861023,
      "learning_rate": 0.00015800000000000002,
      "loss": 1.5467,
      "step": 79
    },
    {
      "epoch": 0.11777695988222305,
      "grad_norm": 0.13064821064472198,
      "learning_rate": 0.00016,
      "loss": 1.5514,
      "step": 80
    },
    {
      "epoch": 0.11924917188075083,
      "grad_norm": 0.1304921656847,
      "learning_rate": 0.000162,
      "loss": 1.5731,
      "step": 81
    },
    {
      "epoch": 0.12072138387927861,
      "grad_norm": 0.13384824991226196,
      "learning_rate": 0.000164,
      "loss": 1.5959,
      "step": 82
    },
    {
      "epoch": 0.1221935958778064,
      "grad_norm": 0.13977950811386108,
      "learning_rate": 0.000166,
      "loss": 1.5706,
      "step": 83
    },
    {
      "epoch": 0.12366580787633419,
      "grad_norm": 0.13425523042678833,
      "learning_rate": 0.000168,
      "loss": 1.5773,
      "step": 84
    },
    {
      "epoch": 0.12513801987486198,
      "grad_norm": 0.1266559362411499,
      "learning_rate": 0.00017,
      "loss": 1.5867,
      "step": 85
    },
    {
      "epoch": 0.12661023187338977,
      "grad_norm": 0.12553836405277252,
      "learning_rate": 0.000172,
      "loss": 1.593,
      "step": 86
    },
    {
      "epoch": 0.12808244387191756,
      "grad_norm": 0.1316322237253189,
      "learning_rate": 0.000174,
      "loss": 1.5671,
      "step": 87
    },
    {
      "epoch": 0.12955465587044535,
      "grad_norm": 0.13156810402870178,
      "learning_rate": 0.00017600000000000002,
      "loss": 1.565,
      "step": 88
    },
    {
      "epoch": 0.13102686786897313,
      "grad_norm": 0.13208283483982086,
      "learning_rate": 0.00017800000000000002,
      "loss": 1.605,
      "step": 89
    },
    {
      "epoch": 0.13249907986750092,
      "grad_norm": 0.13334494829177856,
      "learning_rate": 0.00018,
      "loss": 1.5466,
      "step": 90
    },
    {
      "epoch": 0.1339712918660287,
      "grad_norm": 0.1409706324338913,
      "learning_rate": 0.000182,
      "loss": 1.5891,
      "step": 91
    },
    {
      "epoch": 0.1354435038645565,
      "grad_norm": 0.1378515660762787,
      "learning_rate": 0.00018400000000000003,
      "loss": 1.5779,
      "step": 92
    },
    {
      "epoch": 0.1369157158630843,
      "grad_norm": 0.13051100075244904,
      "learning_rate": 0.00018600000000000002,
      "loss": 1.5436,
      "step": 93
    },
    {
      "epoch": 0.13838792786161208,
      "grad_norm": 0.14391683042049408,
      "learning_rate": 0.000188,
      "loss": 1.55,
      "step": 94
    },
    {
      "epoch": 0.13986013986013987,
      "grad_norm": 0.14169299602508545,
      "learning_rate": 0.00019,
      "loss": 1.5492,
      "step": 95
    },
    {
      "epoch": 0.14133235185866766,
      "grad_norm": 0.13894174993038177,
      "learning_rate": 0.000192,
      "loss": 1.5697,
      "step": 96
    },
    {
      "epoch": 0.14280456385719545,
      "grad_norm": 0.14716342091560364,
      "learning_rate": 0.000194,
      "loss": 1.5542,
      "step": 97
    },
    {
      "epoch": 0.1442767758557232,
      "grad_norm": 0.1475650668144226,
      "learning_rate": 0.000196,
      "loss": 1.5943,
      "step": 98
    },
    {
      "epoch": 0.145748987854251,
      "grad_norm": 0.1504179686307907,
      "learning_rate": 0.00019800000000000002,
      "loss": 1.5918,
      "step": 99
    },
    {
      "epoch": 0.1472211998527788,
      "grad_norm": 0.14042824506759644,
      "learning_rate": 0.0002,
      "loss": 1.5576,
      "step": 100
    },
    {
      "epoch": 0.14869341185130658,
      "grad_norm": 0.1358688473701477,
      "learning_rate": 0.0001999980260856137,
      "loss": 1.5576,
      "step": 101
    },
    {
      "epoch": 0.15016562384983437,
      "grad_norm": 0.1377555876970291,
      "learning_rate": 0.00019999210442038162,
      "loss": 1.5894,
      "step": 102
    },
    {
      "epoch": 0.15163783584836216,
      "grad_norm": 0.13653920590877533,
      "learning_rate": 0.0001999822352380809,
      "loss": 1.562,
      "step": 103
    },
    {
      "epoch": 0.15311004784688995,
      "grad_norm": 0.1352362483739853,
      "learning_rate": 0.00019996841892833,
      "loss": 1.5868,
      "step": 104
    },
    {
      "epoch": 0.15458225984541774,
      "grad_norm": 0.13514937460422516,
      "learning_rate": 0.00019995065603657316,
      "loss": 1.579,
      "step": 105
    },
    {
      "epoch": 0.15605447184394552,
      "grad_norm": 0.13642875850200653,
      "learning_rate": 0.00019992894726405893,
      "loss": 1.5628,
      "step": 106
    },
    {
      "epoch": 0.15752668384247331,
      "grad_norm": 0.13795313239097595,
      "learning_rate": 0.0001999032934678125,
      "loss": 1.5788,
      "step": 107
    },
    {
      "epoch": 0.1589988958410011,
      "grad_norm": 0.13564889132976532,
      "learning_rate": 0.00019987369566060176,
      "loss": 1.5302,
      "step": 108
    },
    {
      "epoch": 0.1604711078395289,
      "grad_norm": 0.13674072921276093,
      "learning_rate": 0.00019984015501089752,
      "loss": 1.556,
      "step": 109
    },
    {
      "epoch": 0.16194331983805668,
      "grad_norm": 0.13632620871067047,
      "learning_rate": 0.00019980267284282717,
      "loss": 1.5641,
      "step": 110
    },
    {
      "epoch": 0.16341553183658447,
      "grad_norm": 0.1406012922525406,
      "learning_rate": 0.00019976125063612252,
      "loss": 1.5464,
      "step": 111
    },
    {
      "epoch": 0.16488774383511226,
      "grad_norm": 0.14117242395877838,
      "learning_rate": 0.0001997158900260614,
      "loss": 1.5638,
      "step": 112
    },
    {
      "epoch": 0.16635995583364005,
      "grad_norm": 0.1342194825410843,
      "learning_rate": 0.00019966659280340297,
      "loss": 1.5847,
      "step": 113
    },
    {
      "epoch": 0.16783216783216784,
      "grad_norm": 0.13204339146614075,
      "learning_rate": 0.00019961336091431727,
      "loss": 1.5544,
      "step": 114
    },
    {
      "epoch": 0.16930437983069563,
      "grad_norm": 0.13027901947498322,
      "learning_rate": 0.00019955619646030802,
      "loss": 1.5825,
      "step": 115
    },
    {
      "epoch": 0.17077659182922342,
      "grad_norm": 0.12915514409542084,
      "learning_rate": 0.00019949510169813003,
      "loss": 1.5547,
      "step": 116
    },
    {
      "epoch": 0.1722488038277512,
      "grad_norm": 0.13460314273834229,
      "learning_rate": 0.0001994300790396999,
      "loss": 1.5668,
      "step": 117
    },
    {
      "epoch": 0.173721015826279,
      "grad_norm": 0.13463035225868225,
      "learning_rate": 0.00019936113105200085,
      "loss": 1.5628,
      "step": 118
    },
    {
      "epoch": 0.17519322782480679,
      "grad_norm": 0.13116863369941711,
      "learning_rate": 0.00019928826045698136,
      "loss": 1.5694,
      "step": 119
    },
    {
      "epoch": 0.17666543982333455,
      "grad_norm": 0.1322312206029892,
      "learning_rate": 0.0001992114701314478,
      "loss": 1.5706,
      "step": 120
    },
    {
      "epoch": 0.17813765182186234,
      "grad_norm": 0.1320774406194687,
      "learning_rate": 0.00019913076310695068,
      "loss": 1.5586,
      "step": 121
    },
    {
      "epoch": 0.17960986382039013,
      "grad_norm": 0.12637224793434143,
      "learning_rate": 0.00019904614256966512,
      "loss": 1.5483,
      "step": 122
    },
    {
      "epoch": 0.18108207581891791,
      "grad_norm": 0.13226664066314697,
      "learning_rate": 0.0001989576118602651,
      "loss": 1.522,
      "step": 123
    },
    {
      "epoch": 0.1825542878174457,
      "grad_norm": 0.1337641328573227,
      "learning_rate": 0.0001988651744737914,
      "loss": 1.5813,
      "step": 124
    },
    {
      "epoch": 0.1840264998159735,
      "grad_norm": 0.1351972222328186,
      "learning_rate": 0.00019876883405951377,
      "loss": 1.5256,
      "step": 125
    },
    {
      "epoch": 0.18549871181450128,
      "grad_norm": 0.14576929807662964,
      "learning_rate": 0.0001986685944207868,
      "loss": 1.5435,
      "step": 126
    },
    {
      "epoch": 0.18697092381302907,
      "grad_norm": 0.13020643591880798,
      "learning_rate": 0.00019856445951489982,
      "loss": 1.5929,
      "step": 127
    },
    {
      "epoch": 0.18844313581155686,
      "grad_norm": 0.12671110033988953,
      "learning_rate": 0.00019845643345292054,
      "loss": 1.5455,
      "step": 128
    },
    {
      "epoch": 0.18991534781008465,
      "grad_norm": 0.12815219163894653,
      "learning_rate": 0.00019834452049953297,
      "loss": 1.5267,
      "step": 129
    },
    {
      "epoch": 0.19138755980861244,
      "grad_norm": 0.13211888074874878,
      "learning_rate": 0.0001982287250728689,
      "loss": 1.579,
      "step": 130
    },
    {
      "epoch": 0.19285977180714023,
      "grad_norm": 0.1327330321073532,
      "learning_rate": 0.0001981090517443334,
      "loss": 1.5789,
      "step": 131
    },
    {
      "epoch": 0.19433198380566802,
      "grad_norm": 0.13214127719402313,
      "learning_rate": 0.0001979855052384247,
      "loss": 1.5321,
      "step": 132
    },
    {
      "epoch": 0.1958041958041958,
      "grad_norm": 0.13302084803581238,
      "learning_rate": 0.00019785809043254722,
      "loss": 1.5513,
      "step": 133
    },
    {
      "epoch": 0.1972764078027236,
      "grad_norm": 0.1224193349480629,
      "learning_rate": 0.00019772681235681936,
      "loss": 1.5603,
      "step": 134
    },
    {
      "epoch": 0.19874861980125139,
      "grad_norm": 0.12664197385311127,
      "learning_rate": 0.00019759167619387476,
      "loss": 1.5906,
      "step": 135
    },
    {
      "epoch": 0.20022083179977918,
      "grad_norm": 0.12762169539928436,
      "learning_rate": 0.00019745268727865774,
      "loss": 1.5591,
      "step": 136
    },
    {
      "epoch": 0.20169304379830696,
      "grad_norm": 0.1285737156867981,
      "learning_rate": 0.00019730985109821266,
      "loss": 1.5365,
      "step": 137
    },
    {
      "epoch": 0.20316525579683475,
      "grad_norm": 0.12642493844032288,
      "learning_rate": 0.0001971631732914674,
      "loss": 1.5529,
      "step": 138
    },
    {
      "epoch": 0.20463746779536254,
      "grad_norm": 0.13150472939014435,
      "learning_rate": 0.0001970126596490106,
      "loss": 1.5483,
      "step": 139
    },
    {
      "epoch": 0.20610967979389033,
      "grad_norm": 0.1243419200181961,
      "learning_rate": 0.0001968583161128631,
      "loss": 1.5493,
      "step": 140
    },
    {
      "epoch": 0.20758189179241812,
      "grad_norm": 0.13332587480545044,
      "learning_rate": 0.00019670014877624353,
      "loss": 1.5708,
      "step": 141
    },
    {
      "epoch": 0.20905410379094588,
      "grad_norm": 0.1288883537054062,
      "learning_rate": 0.0001965381638833274,
      "loss": 1.5447,
      "step": 142
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 0.12580043077468872,
      "learning_rate": 0.000196372367829001,
      "loss": 1.5399,
      "step": 143
    },
    {
      "epoch": 0.21199852778800146,
      "grad_norm": 0.12434295564889908,
      "learning_rate": 0.0001962027671586086,
      "loss": 1.5683,
      "step": 144
    },
    {
      "epoch": 0.21347073978652925,
      "grad_norm": 0.1273431032896042,
      "learning_rate": 0.0001960293685676943,
      "loss": 1.5666,
      "step": 145
    },
    {
      "epoch": 0.21494295178505704,
      "grad_norm": 0.1347806453704834,
      "learning_rate": 0.0001958521789017376,
      "loss": 1.5455,
      "step": 146
    },
    {
      "epoch": 0.21641516378358483,
      "grad_norm": 0.12940432131290436,
      "learning_rate": 0.00019567120515588308,
      "loss": 1.5639,
      "step": 147
    },
    {
      "epoch": 0.21788737578211262,
      "grad_norm": 0.12149060517549515,
      "learning_rate": 0.00019548645447466431,
      "loss": 1.5843,
      "step": 148
    },
    {
      "epoch": 0.2193595877806404,
      "grad_norm": 0.12301753461360931,
      "learning_rate": 0.00019529793415172192,
      "loss": 1.5361,
      "step": 149
    },
    {
      "epoch": 0.2208317997791682,
      "grad_norm": 0.13079455494880676,
      "learning_rate": 0.00019510565162951537,
      "loss": 1.5477,
      "step": 150
    },
    {
      "epoch": 0.222304011777696,
      "grad_norm": 0.12788613140583038,
      "learning_rate": 0.00019490961449902946,
      "loss": 1.5362,
      "step": 151
    },
    {
      "epoch": 0.22377622377622378,
      "grad_norm": 0.12581096589565277,
      "learning_rate": 0.00019470983049947444,
      "loss": 1.5336,
      "step": 152
    },
    {
      "epoch": 0.22524843577475157,
      "grad_norm": 0.12982791662216187,
      "learning_rate": 0.00019450630751798048,
      "loss": 1.5316,
      "step": 153
    },
    {
      "epoch": 0.22672064777327935,
      "grad_norm": 0.13706190884113312,
      "learning_rate": 0.00019429905358928646,
      "loss": 1.5428,
      "step": 154
    },
    {
      "epoch": 0.22819285977180714,
      "grad_norm": 0.1302231401205063,
      "learning_rate": 0.00019408807689542257,
      "loss": 1.5427,
      "step": 155
    },
    {
      "epoch": 0.22966507177033493,
      "grad_norm": 0.12531238794326782,
      "learning_rate": 0.00019387338576538744,
      "loss": 1.5302,
      "step": 156
    },
    {
      "epoch": 0.23113728376886272,
      "grad_norm": 0.13268686830997467,
      "learning_rate": 0.00019365498867481923,
      "loss": 1.5347,
      "step": 157
    },
    {
      "epoch": 0.2326094957673905,
      "grad_norm": 0.13343119621276855,
      "learning_rate": 0.00019343289424566122,
      "loss": 1.5552,
      "step": 158
    },
    {
      "epoch": 0.2340817077659183,
      "grad_norm": 0.1270790994167328,
      "learning_rate": 0.0001932071112458211,
      "loss": 1.5272,
      "step": 159
    },
    {
      "epoch": 0.2355539197644461,
      "grad_norm": 0.12997180223464966,
      "learning_rate": 0.00019297764858882514,
      "loss": 1.5504,
      "step": 160
    },
    {
      "epoch": 0.23702613176297388,
      "grad_norm": 0.12340233474969864,
      "learning_rate": 0.00019274451533346615,
      "loss": 1.5634,
      "step": 161
    },
    {
      "epoch": 0.23849834376150167,
      "grad_norm": 0.1311703324317932,
      "learning_rate": 0.0001925077206834458,
      "loss": 1.5017,
      "step": 162
    },
    {
      "epoch": 0.23997055576002946,
      "grad_norm": 0.12580685317516327,
      "learning_rate": 0.0001922672739870115,
      "loss": 1.5544,
      "step": 163
    },
    {
      "epoch": 0.24144276775855722,
      "grad_norm": 0.1292380392551422,
      "learning_rate": 0.00019202318473658705,
      "loss": 1.5349,
      "step": 164
    },
    {
      "epoch": 0.242914979757085,
      "grad_norm": 0.1275525838136673,
      "learning_rate": 0.00019177546256839812,
      "loss": 1.5421,
      "step": 165
    },
    {
      "epoch": 0.2443871917556128,
      "grad_norm": 0.1311991810798645,
      "learning_rate": 0.00019152411726209176,
      "loss": 1.5448,
      "step": 166
    },
    {
      "epoch": 0.2458594037541406,
      "grad_norm": 0.12919144332408905,
      "learning_rate": 0.0001912691587403503,
      "loss": 1.5526,
      "step": 167
    },
    {
      "epoch": 0.24733161575266838,
      "grad_norm": 0.1315072923898697,
      "learning_rate": 0.00019101059706849957,
      "loss": 1.5595,
      "step": 168
    },
    {
      "epoch": 0.24880382775119617,
      "grad_norm": 0.13200980424880981,
      "learning_rate": 0.0001907484424541117,
      "loss": 1.5158,
      "step": 169
    },
    {
      "epoch": 0.25027603974972396,
      "grad_norm": 0.12845145165920258,
      "learning_rate": 0.00019048270524660196,
      "loss": 1.5328,
      "step": 170
    },
    {
      "epoch": 0.2517482517482518,
      "grad_norm": 0.1307184100151062,
      "learning_rate": 0.00019021339593682028,
      "loss": 1.5358,
      "step": 171
    },
    {
      "epoch": 0.25322046374677953,
      "grad_norm": 0.13154615461826324,
      "learning_rate": 0.0001899405251566371,
      "loss": 1.5048,
      "step": 172
    },
    {
      "epoch": 0.25469267574530735,
      "grad_norm": 0.12833188474178314,
      "learning_rate": 0.00018966410367852362,
      "loss": 1.519,
      "step": 173
    },
    {
      "epoch": 0.2561648877438351,
      "grad_norm": 0.12930724024772644,
      "learning_rate": 0.0001893841424151264,
      "loss": 1.513,
      "step": 174
    },
    {
      "epoch": 0.2576370997423629,
      "grad_norm": 0.13588398694992065,
      "learning_rate": 0.0001891006524188368,
      "loss": 1.5409,
      "step": 175
    },
    {
      "epoch": 0.2591093117408907,
      "grad_norm": 0.1312905102968216,
      "learning_rate": 0.00018881364488135448,
      "loss": 1.583,
      "step": 176
    },
    {
      "epoch": 0.26058152373941845,
      "grad_norm": 0.12696287035942078,
      "learning_rate": 0.00018852313113324552,
      "loss": 1.5449,
      "step": 177
    },
    {
      "epoch": 0.26205373573794627,
      "grad_norm": 0.12995372712612152,
      "learning_rate": 0.00018822912264349534,
      "loss": 1.5357,
      "step": 178
    },
    {
      "epoch": 0.26352594773647403,
      "grad_norm": 0.1284688264131546,
      "learning_rate": 0.00018793163101905563,
      "loss": 1.5273,
      "step": 179
    },
    {
      "epoch": 0.26499815973500185,
      "grad_norm": 0.1321571320295334,
      "learning_rate": 0.00018763066800438636,
      "loss": 1.5436,
      "step": 180
    },
    {
      "epoch": 0.2664703717335296,
      "grad_norm": 0.13060812652111053,
      "learning_rate": 0.00018732624548099204,
      "loss": 1.5164,
      "step": 181
    },
    {
      "epoch": 0.2679425837320574,
      "grad_norm": 0.12950077652931213,
      "learning_rate": 0.0001870183754669526,
      "loss": 1.5413,
      "step": 182
    },
    {
      "epoch": 0.2694147957305852,
      "grad_norm": 0.13167962431907654,
      "learning_rate": 0.000186707070116449,
      "loss": 1.53,
      "step": 183
    },
    {
      "epoch": 0.270887007729113,
      "grad_norm": 0.12349419295787811,
      "learning_rate": 0.00018639234171928353,
      "loss": 1.5489,
      "step": 184
    },
    {
      "epoch": 0.27235921972764077,
      "grad_norm": 0.12604287266731262,
      "learning_rate": 0.0001860742027003944,
      "loss": 1.5491,
      "step": 185
    },
    {
      "epoch": 0.2738314317261686,
      "grad_norm": 0.1273098886013031,
      "learning_rate": 0.00018575266561936523,
      "loss": 1.5516,
      "step": 186
    },
    {
      "epoch": 0.27530364372469635,
      "grad_norm": 0.12621380388736725,
      "learning_rate": 0.0001854277431699295,
      "loss": 1.522,
      "step": 187
    },
    {
      "epoch": 0.27677585572322416,
      "grad_norm": 0.1286550909280777,
      "learning_rate": 0.00018509944817946922,
      "loss": 1.5585,
      "step": 188
    },
    {
      "epoch": 0.2782480677217519,
      "grad_norm": 0.12399259209632874,
      "learning_rate": 0.00018476779360850832,
      "loss": 1.5303,
      "step": 189
    },
    {
      "epoch": 0.27972027972027974,
      "grad_norm": 0.12871603667736053,
      "learning_rate": 0.00018443279255020152,
      "loss": 1.539,
      "step": 190
    },
    {
      "epoch": 0.2811924917188075,
      "grad_norm": 0.13571785390377045,
      "learning_rate": 0.00018409445822981693,
      "loss": 1.5658,
      "step": 191
    },
    {
      "epoch": 0.2826647037173353,
      "grad_norm": 0.12659163773059845,
      "learning_rate": 0.0001837528040042142,
      "loss": 1.544,
      "step": 192
    },
    {
      "epoch": 0.2841369157158631,
      "grad_norm": 0.12647713720798492,
      "learning_rate": 0.00018340784336131713,
      "loss": 1.5378,
      "step": 193
    },
    {
      "epoch": 0.2856091277143909,
      "grad_norm": 0.12926916778087616,
      "learning_rate": 0.00018305958991958127,
      "loss": 1.5408,
      "step": 194
    },
    {
      "epoch": 0.28708133971291866,
      "grad_norm": 0.1257879137992859,
      "learning_rate": 0.00018270805742745617,
      "loss": 1.5332,
      "step": 195
    },
    {
      "epoch": 0.2885535517114464,
      "grad_norm": 0.12876833975315094,
      "learning_rate": 0.00018235325976284275,
      "loss": 1.5116,
      "step": 196
    },
    {
      "epoch": 0.29002576370997424,
      "grad_norm": 0.12502498924732208,
      "learning_rate": 0.00018199521093254523,
      "loss": 1.5126,
      "step": 197
    },
    {
      "epoch": 0.291497975708502,
      "grad_norm": 0.12516739964485168,
      "learning_rate": 0.00018163392507171842,
      "loss": 1.5047,
      "step": 198
    },
    {
      "epoch": 0.2929701877070298,
      "grad_norm": 0.1297457367181778,
      "learning_rate": 0.0001812694164433094,
      "loss": 1.5269,
      "step": 199
    },
    {
      "epoch": 0.2944423997055576,
      "grad_norm": 0.1266913115978241,
      "learning_rate": 0.00018090169943749476,
      "loss": 1.5198,
      "step": 200
    },
    {
      "epoch": 0.2959146117040854,
      "grad_norm": 0.12404821813106537,
      "learning_rate": 0.0001805307885711122,
      "loss": 1.5311,
      "step": 201
    },
    {
      "epoch": 0.29738682370261316,
      "grad_norm": 0.12168893218040466,
      "learning_rate": 0.00018015669848708767,
      "loss": 1.5555,
      "step": 202
    },
    {
      "epoch": 0.298859035701141,
      "grad_norm": 0.1262127012014389,
      "learning_rate": 0.0001797794439538571,
      "loss": 1.5527,
      "step": 203
    },
    {
      "epoch": 0.30033124769966874,
      "grad_norm": 0.12916256487369537,
      "learning_rate": 0.00017939903986478355,
      "loss": 1.5193,
      "step": 204
    },
    {
      "epoch": 0.30180345969819655,
      "grad_norm": 0.12800262868404388,
      "learning_rate": 0.00017901550123756906,
      "loss": 1.5241,
      "step": 205
    },
    {
      "epoch": 0.3032756716967243,
      "grad_norm": 0.12354377657175064,
      "learning_rate": 0.00017862884321366188,
      "loss": 1.5144,
      "step": 206
    },
    {
      "epoch": 0.30474788369525213,
      "grad_norm": 0.13330724835395813,
      "learning_rate": 0.0001782390810576588,
      "loss": 1.4818,
      "step": 207
    },
    {
      "epoch": 0.3062200956937799,
      "grad_norm": 0.12449085712432861,
      "learning_rate": 0.00017784623015670238,
      "loss": 1.5516,
      "step": 208
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 0.1266268938779831,
      "learning_rate": 0.00017745030601987337,
      "loss": 1.5136,
      "step": 209
    },
    {
      "epoch": 0.30916451969083547,
      "grad_norm": 0.1184929758310318,
      "learning_rate": 0.00017705132427757895,
      "loss": 1.5476,
      "step": 210
    },
    {
      "epoch": 0.3106367316893633,
      "grad_norm": 0.12397826462984085,
      "learning_rate": 0.00017664930068093498,
      "loss": 1.5411,
      "step": 211
    },
    {
      "epoch": 0.31210894368789105,
      "grad_norm": 0.12503477931022644,
      "learning_rate": 0.0001762442511011448,
      "loss": 1.5438,
      "step": 212
    },
    {
      "epoch": 0.31358115568641887,
      "grad_norm": 0.12811186909675598,
      "learning_rate": 0.0001758361915288722,
      "loss": 1.534,
      "step": 213
    },
    {
      "epoch": 0.31505336768494663,
      "grad_norm": 0.1272122710943222,
      "learning_rate": 0.00017542513807361037,
      "loss": 1.4977,
      "step": 214
    },
    {
      "epoch": 0.31652557968347445,
      "grad_norm": 0.12660300731658936,
      "learning_rate": 0.00017501110696304596,
      "loss": 1.5256,
      "step": 215
    },
    {
      "epoch": 0.3179977916820022,
      "grad_norm": 0.12921711802482605,
      "learning_rate": 0.00017459411454241822,
      "loss": 1.5161,
      "step": 216
    },
    {
      "epoch": 0.31947000368053,
      "grad_norm": 0.13279977440834045,
      "learning_rate": 0.00017417417727387394,
      "loss": 1.5273,
      "step": 217
    },
    {
      "epoch": 0.3209422156790578,
      "grad_norm": 0.12619398534297943,
      "learning_rate": 0.0001737513117358174,
      "loss": 1.4882,
      "step": 218
    },
    {
      "epoch": 0.32241442767758555,
      "grad_norm": 0.12969842553138733,
      "learning_rate": 0.00017332553462225602,
      "loss": 1.5171,
      "step": 219
    },
    {
      "epoch": 0.32388663967611336,
      "grad_norm": 0.12548337876796722,
      "learning_rate": 0.00017289686274214118,
      "loss": 1.567,
      "step": 220
    },
    {
      "epoch": 0.3253588516746411,
      "grad_norm": 0.12587487697601318,
      "learning_rate": 0.0001724653130187047,
      "loss": 1.5275,
      "step": 221
    },
    {
      "epoch": 0.32683106367316894,
      "grad_norm": 0.12818500399589539,
      "learning_rate": 0.0001720309024887907,
      "loss": 1.5385,
      "step": 222
    },
    {
      "epoch": 0.3283032756716967,
      "grad_norm": 0.12866437435150146,
      "learning_rate": 0.00017159364830218312,
      "loss": 1.5309,
      "step": 223
    },
    {
      "epoch": 0.3297754876702245,
      "grad_norm": 0.12270201742649078,
      "learning_rate": 0.00017115356772092857,
      "loss": 1.5222,
      "step": 224
    },
    {
      "epoch": 0.3312476996687523,
      "grad_norm": 0.1288307011127472,
      "learning_rate": 0.00017071067811865476,
      "loss": 1.5364,
      "step": 225
    },
    {
      "epoch": 0.3327199116672801,
      "grad_norm": 0.12552323937416077,
      "learning_rate": 0.00017026499697988493,
      "loss": 1.5437,
      "step": 226
    },
    {
      "epoch": 0.33419212366580786,
      "grad_norm": 0.12705761194229126,
      "learning_rate": 0.00016981654189934727,
      "loss": 1.5326,
      "step": 227
    },
    {
      "epoch": 0.3356643356643357,
      "grad_norm": 0.12445786595344543,
      "learning_rate": 0.0001693653305812805,
      "loss": 1.5561,
      "step": 228
    },
    {
      "epoch": 0.33713654766286344,
      "grad_norm": 0.1243494376540184,
      "learning_rate": 0.00016891138083873487,
      "loss": 1.505,
      "step": 229
    },
    {
      "epoch": 0.33860875966139126,
      "grad_norm": 0.13060913980007172,
      "learning_rate": 0.00016845471059286887,
      "loss": 1.5399,
      "step": 230
    },
    {
      "epoch": 0.340080971659919,
      "grad_norm": 0.12850089371204376,
      "learning_rate": 0.00016799533787224192,
      "loss": 1.5115,
      "step": 231
    },
    {
      "epoch": 0.34155318365844684,
      "grad_norm": 0.12945328652858734,
      "learning_rate": 0.00016753328081210245,
      "loss": 1.544,
      "step": 232
    },
    {
      "epoch": 0.3430253956569746,
      "grad_norm": 0.127003014087677,
      "learning_rate": 0.000167068557653672,
      "loss": 1.5238,
      "step": 233
    },
    {
      "epoch": 0.3444976076555024,
      "grad_norm": 0.1306627243757248,
      "learning_rate": 0.00016660118674342517,
      "loss": 1.5147,
      "step": 234
    },
    {
      "epoch": 0.3459698196540302,
      "grad_norm": 0.12929843366146088,
      "learning_rate": 0.00016613118653236518,
      "loss": 1.5252,
      "step": 235
    },
    {
      "epoch": 0.347442031652558,
      "grad_norm": 0.1339099258184433,
      "learning_rate": 0.00016565857557529566,
      "loss": 1.5035,
      "step": 236
    },
    {
      "epoch": 0.34891424365108575,
      "grad_norm": 0.12324322015047073,
      "learning_rate": 0.0001651833725300879,
      "loss": 1.5172,
      "step": 237
    },
    {
      "epoch": 0.35038645564961357,
      "grad_norm": 0.12321919947862625,
      "learning_rate": 0.00016470559615694446,
      "loss": 1.5209,
      "step": 238
    },
    {
      "epoch": 0.35185866764814133,
      "grad_norm": 0.12979108095169067,
      "learning_rate": 0.00016422526531765846,
      "loss": 1.5286,
      "step": 239
    },
    {
      "epoch": 0.3533308796466691,
      "grad_norm": 0.12395581603050232,
      "learning_rate": 0.000163742398974869,
      "loss": 1.5364,
      "step": 240
    },
    {
      "epoch": 0.3548030916451969,
      "grad_norm": 0.12904591858386993,
      "learning_rate": 0.00016325701619131246,
      "loss": 1.5488,
      "step": 241
    },
    {
      "epoch": 0.3562753036437247,
      "grad_norm": 0.12751571834087372,
      "learning_rate": 0.00016276913612907007,
      "loss": 1.5481,
      "step": 242
    },
    {
      "epoch": 0.3577475156422525,
      "grad_norm": 0.12640038132667542,
      "learning_rate": 0.00016227877804881127,
      "loss": 1.5488,
      "step": 243
    },
    {
      "epoch": 0.35921972764078025,
      "grad_norm": 0.12787500023841858,
      "learning_rate": 0.00016178596130903344,
      "loss": 1.5773,
      "step": 244
    },
    {
      "epoch": 0.36069193963930807,
      "grad_norm": 0.12736569344997406,
      "learning_rate": 0.00016129070536529766,
      "loss": 1.5244,
      "step": 245
    },
    {
      "epoch": 0.36216415163783583,
      "grad_norm": 0.1278051882982254,
      "learning_rate": 0.00016079302976946055,
      "loss": 1.4955,
      "step": 246
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 0.13505584001541138,
      "learning_rate": 0.00016029295416890248,
      "loss": 1.511,
      "step": 247
    },
    {
      "epoch": 0.3651085756348914,
      "grad_norm": 0.12807372212409973,
      "learning_rate": 0.0001597904983057519,
      "loss": 1.5466,
      "step": 248
    },
    {
      "epoch": 0.3665807876334192,
      "grad_norm": 0.1321410834789276,
      "learning_rate": 0.00015928568201610595,
      "loss": 1.5148,
      "step": 249
    },
    {
      "epoch": 0.368052999631947,
      "grad_norm": 0.124916672706604,
      "learning_rate": 0.00015877852522924732,
      "loss": 1.5196,
      "step": 250
    },
    {
      "epoch": 0.3695252116304748,
      "grad_norm": 0.1248563751578331,
      "learning_rate": 0.00015826904796685762,
      "loss": 1.5315,
      "step": 251
    },
    {
      "epoch": 0.37099742362900257,
      "grad_norm": 0.12782709300518036,
      "learning_rate": 0.00015775727034222675,
      "loss": 1.4711,
      "step": 252
    },
    {
      "epoch": 0.3724696356275304,
      "grad_norm": 0.12841551005840302,
      "learning_rate": 0.0001572432125594591,
      "loss": 1.5044,
      "step": 253
    },
    {
      "epoch": 0.37394184762605814,
      "grad_norm": 0.12746386229991913,
      "learning_rate": 0.00015672689491267567,
      "loss": 1.5193,
      "step": 254
    },
    {
      "epoch": 0.37541405962458596,
      "grad_norm": 0.1278688609600067,
      "learning_rate": 0.00015620833778521307,
      "loss": 1.5091,
      "step": 255
    },
    {
      "epoch": 0.3768862716231137,
      "grad_norm": 0.13400790095329285,
      "learning_rate": 0.00015568756164881882,
      "loss": 1.5451,
      "step": 256
    },
    {
      "epoch": 0.37835848362164154,
      "grad_norm": 0.12973400950431824,
      "learning_rate": 0.00015516458706284303,
      "loss": 1.5009,
      "step": 257
    },
    {
      "epoch": 0.3798306956201693,
      "grad_norm": 0.1290912926197052,
      "learning_rate": 0.00015463943467342693,
      "loss": 1.5416,
      "step": 258
    },
    {
      "epoch": 0.3813029076186971,
      "grad_norm": 0.13089627027511597,
      "learning_rate": 0.00015411212521268758,
      "loss": 1.5258,
      "step": 259
    },
    {
      "epoch": 0.3827751196172249,
      "grad_norm": 0.12819084525108337,
      "learning_rate": 0.00015358267949789966,
      "loss": 1.5409,
      "step": 260
    },
    {
      "epoch": 0.38424733161575264,
      "grad_norm": 0.1323699951171875,
      "learning_rate": 0.0001530511184306734,
      "loss": 1.4914,
      "step": 261
    },
    {
      "epoch": 0.38571954361428046,
      "grad_norm": 0.13444304466247559,
      "learning_rate": 0.0001525174629961296,
      "loss": 1.5082,
      "step": 262
    },
    {
      "epoch": 0.3871917556128082,
      "grad_norm": 0.12922947108745575,
      "learning_rate": 0.00015198173426207094,
      "loss": 1.526,
      "step": 263
    },
    {
      "epoch": 0.38866396761133604,
      "grad_norm": 0.13330569863319397,
      "learning_rate": 0.00015144395337815064,
      "loss": 1.5234,
      "step": 264
    },
    {
      "epoch": 0.3901361796098638,
      "grad_norm": 0.13401193916797638,
      "learning_rate": 0.00015090414157503714,
      "loss": 1.496,
      "step": 265
    },
    {
      "epoch": 0.3916083916083916,
      "grad_norm": 0.1333129107952118,
      "learning_rate": 0.0001503623201635761,
      "loss": 1.5069,
      "step": 266
    },
    {
      "epoch": 0.3930806036069194,
      "grad_norm": 0.1272551715373993,
      "learning_rate": 0.0001498185105339491,
      "loss": 1.5036,
      "step": 267
    },
    {
      "epoch": 0.3945528156054472,
      "grad_norm": 0.12798148393630981,
      "learning_rate": 0.00014927273415482915,
      "loss": 1.4968,
      "step": 268
    },
    {
      "epoch": 0.39602502760397496,
      "grad_norm": 0.1261674463748932,
      "learning_rate": 0.00014872501257253323,
      "loss": 1.5261,
      "step": 269
    },
    {
      "epoch": 0.39749723960250277,
      "grad_norm": 0.13026413321495056,
      "learning_rate": 0.00014817536741017152,
      "loss": 1.5141,
      "step": 270
    },
    {
      "epoch": 0.39896945160103053,
      "grad_norm": 0.12507426738739014,
      "learning_rate": 0.0001476238203667939,
      "loss": 1.5253,
      "step": 271
    },
    {
      "epoch": 0.40044166359955835,
      "grad_norm": 0.12942412495613098,
      "learning_rate": 0.0001470703932165333,
      "loss": 1.5218,
      "step": 272
    },
    {
      "epoch": 0.4019138755980861,
      "grad_norm": 0.13566623628139496,
      "learning_rate": 0.00014651510780774583,
      "loss": 1.5249,
      "step": 273
    },
    {
      "epoch": 0.40338608759661393,
      "grad_norm": 0.13030388951301575,
      "learning_rate": 0.00014595798606214882,
      "loss": 1.4935,
      "step": 274
    },
    {
      "epoch": 0.4048582995951417,
      "grad_norm": 0.1307370364665985,
      "learning_rate": 0.00014539904997395468,
      "loss": 1.4912,
      "step": 275
    },
    {
      "epoch": 0.4063305115936695,
      "grad_norm": 0.13302060961723328,
      "learning_rate": 0.00014483832160900326,
      "loss": 1.5311,
      "step": 276
    },
    {
      "epoch": 0.40780272359219727,
      "grad_norm": 0.13240724802017212,
      "learning_rate": 0.0001442758231038902,
      "loss": 1.5176,
      "step": 277
    },
    {
      "epoch": 0.4092749355907251,
      "grad_norm": 0.13307149708271027,
      "learning_rate": 0.0001437115766650933,
      "loss": 1.5167,
      "step": 278
    },
    {
      "epoch": 0.41074714758925285,
      "grad_norm": 0.13026264309883118,
      "learning_rate": 0.0001431456045680959,
      "loss": 1.519,
      "step": 279
    },
    {
      "epoch": 0.41221935958778066,
      "grad_norm": 0.12934176623821259,
      "learning_rate": 0.00014257792915650728,
      "loss": 1.4979,
      "step": 280
    },
    {
      "epoch": 0.4136915715863084,
      "grad_norm": 0.12649208307266235,
      "learning_rate": 0.00014200857284118066,
      "loss": 1.4716,
      "step": 281
    },
    {
      "epoch": 0.41516378358483624,
      "grad_norm": 0.13500629365444183,
      "learning_rate": 0.00014143755809932845,
      "loss": 1.5176,
      "step": 282
    },
    {
      "epoch": 0.416635995583364,
      "grad_norm": 0.135499507188797,
      "learning_rate": 0.00014086490747363493,
      "loss": 1.5274,
      "step": 283
    },
    {
      "epoch": 0.41810820758189177,
      "grad_norm": 0.12690576910972595,
      "learning_rate": 0.00014029064357136628,
      "loss": 1.5351,
      "step": 284
    },
    {
      "epoch": 0.4195804195804196,
      "grad_norm": 0.13042376935482025,
      "learning_rate": 0.00013971478906347806,
      "loss": 1.5006,
      "step": 285
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 0.13196174800395966,
      "learning_rate": 0.00013913736668372026,
      "loss": 1.5131,
      "step": 286
    },
    {
      "epoch": 0.42252484357747516,
      "grad_norm": 0.1333332359790802,
      "learning_rate": 0.00013855839922773968,
      "loss": 1.4935,
      "step": 287
    },
    {
      "epoch": 0.4239970555760029,
      "grad_norm": 0.1266212910413742,
      "learning_rate": 0.00013797790955218014,
      "loss": 1.4935,
      "step": 288
    },
    {
      "epoch": 0.42546926757453074,
      "grad_norm": 0.13191771507263184,
      "learning_rate": 0.00013739592057378003,
      "loss": 1.4827,
      "step": 289
    },
    {
      "epoch": 0.4269414795730585,
      "grad_norm": 0.13043944537639618,
      "learning_rate": 0.00013681245526846783,
      "loss": 1.5146,
      "step": 290
    },
    {
      "epoch": 0.4284136915715863,
      "grad_norm": 0.1315285563468933,
      "learning_rate": 0.00013622753667045457,
      "loss": 1.5129,
      "step": 291
    },
    {
      "epoch": 0.4298859035701141,
      "grad_norm": 0.12743107974529266,
      "learning_rate": 0.00013564118787132506,
      "loss": 1.4806,
      "step": 292
    },
    {
      "epoch": 0.4313581155686419,
      "grad_norm": 0.13012798130512238,
      "learning_rate": 0.0001350534320191259,
      "loss": 1.5476,
      "step": 293
    },
    {
      "epoch": 0.43283032756716966,
      "grad_norm": 0.130373015999794,
      "learning_rate": 0.0001344642923174517,
      "loss": 1.4977,
      "step": 294
    },
    {
      "epoch": 0.4343025395656975,
      "grad_norm": 0.13684207201004028,
      "learning_rate": 0.00013387379202452917,
      "loss": 1.4642,
      "step": 295
    },
    {
      "epoch": 0.43577475156422524,
      "grad_norm": 0.1254444420337677,
      "learning_rate": 0.00013328195445229868,
      "loss": 1.5317,
      "step": 296
    },
    {
      "epoch": 0.43724696356275305,
      "grad_norm": 0.13003680109977722,
      "learning_rate": 0.00013268880296549425,
      "loss": 1.5281,
      "step": 297
    },
    {
      "epoch": 0.4387191755612808,
      "grad_norm": 0.13377824425697327,
      "learning_rate": 0.00013209436098072095,
      "loss": 1.5127,
      "step": 298
    },
    {
      "epoch": 0.44019138755980863,
      "grad_norm": 0.12856221199035645,
      "learning_rate": 0.0001314986519655305,
      "loss": 1.4857,
      "step": 299
    },
    {
      "epoch": 0.4416635995583364,
      "grad_norm": 0.12562119960784912,
      "learning_rate": 0.00013090169943749476,
      "loss": 1.5121,
      "step": 300
    },
    {
      "epoch": 0.4431358115568642,
      "grad_norm": 0.12989653646945953,
      "learning_rate": 0.00013030352696327742,
      "loss": 1.4812,
      "step": 301
    },
    {
      "epoch": 0.444608023555392,
      "grad_norm": 0.1322024166584015,
      "learning_rate": 0.0001297041581577035,
      "loss": 1.5141,
      "step": 302
    },
    {
      "epoch": 0.4460802355539198,
      "grad_norm": 0.13190506398677826,
      "learning_rate": 0.00012910361668282719,
      "loss": 1.5016,
      "step": 303
    },
    {
      "epoch": 0.44755244755244755,
      "grad_norm": 0.1277342438697815,
      "learning_rate": 0.0001285019262469976,
      "loss": 1.5107,
      "step": 304
    },
    {
      "epoch": 0.4490246595509753,
      "grad_norm": 0.13322456181049347,
      "learning_rate": 0.00012789911060392294,
      "loss": 1.466,
      "step": 305
    },
    {
      "epoch": 0.45049687154950313,
      "grad_norm": 0.12962105870246887,
      "learning_rate": 0.00012729519355173254,
      "loss": 1.5227,
      "step": 306
    },
    {
      "epoch": 0.4519690835480309,
      "grad_norm": 0.13476352393627167,
      "learning_rate": 0.00012669019893203759,
      "loss": 1.477,
      "step": 307
    },
    {
      "epoch": 0.4534412955465587,
      "grad_norm": 0.13467338681221008,
      "learning_rate": 0.00012608415062898972,
      "loss": 1.5132,
      "step": 308
    },
    {
      "epoch": 0.45491350754508647,
      "grad_norm": 0.1326906681060791,
      "learning_rate": 0.00012547707256833823,
      "loss": 1.4596,
      "step": 309
    },
    {
      "epoch": 0.4563857195436143,
      "grad_norm": 0.128783717751503,
      "learning_rate": 0.0001248689887164855,
      "loss": 1.5144,
      "step": 310
    },
    {
      "epoch": 0.45785793154214205,
      "grad_norm": 0.1335008442401886,
      "learning_rate": 0.00012425992307954075,
      "loss": 1.5093,
      "step": 311
    },
    {
      "epoch": 0.45933014354066987,
      "grad_norm": 0.13134045898914337,
      "learning_rate": 0.00012364989970237248,
      "loss": 1.4961,
      "step": 312
    },
    {
      "epoch": 0.46080235553919763,
      "grad_norm": 0.1284698247909546,
      "learning_rate": 0.00012303894266765908,
      "loss": 1.5132,
      "step": 313
    },
    {
      "epoch": 0.46227456753772544,
      "grad_norm": 0.1301211714744568,
      "learning_rate": 0.00012242707609493814,
      "loss": 1.5013,
      "step": 314
    },
    {
      "epoch": 0.4637467795362532,
      "grad_norm": 0.13129912316799164,
      "learning_rate": 0.00012181432413965428,
      "loss": 1.5067,
      "step": 315
    },
    {
      "epoch": 0.465218991534781,
      "grad_norm": 0.13301070034503937,
      "learning_rate": 0.00012120071099220549,
      "loss": 1.4861,
      "step": 316
    },
    {
      "epoch": 0.4666912035333088,
      "grad_norm": 0.1287582516670227,
      "learning_rate": 0.00012058626087698814,
      "loss": 1.4912,
      "step": 317
    },
    {
      "epoch": 0.4681634155318366,
      "grad_norm": 0.13155058026313782,
      "learning_rate": 0.00011997099805144069,
      "loss": 1.4863,
      "step": 318
    },
    {
      "epoch": 0.46963562753036436,
      "grad_norm": 0.13563479483127594,
      "learning_rate": 0.00011935494680508606,
      "loss": 1.5353,
      "step": 319
    },
    {
      "epoch": 0.4711078395288922,
      "grad_norm": 0.1328849047422409,
      "learning_rate": 0.00011873813145857249,
      "loss": 1.514,
      "step": 320
    },
    {
      "epoch": 0.47258005152741994,
      "grad_norm": 0.13374167680740356,
      "learning_rate": 0.00011812057636271374,
      "loss": 1.5126,
      "step": 321
    },
    {
      "epoch": 0.47405226352594776,
      "grad_norm": 0.12980063259601593,
      "learning_rate": 0.00011750230589752762,
      "loss": 1.5269,
      "step": 322
    },
    {
      "epoch": 0.4755244755244755,
      "grad_norm": 0.12554620206356049,
      "learning_rate": 0.00011688334447127338,
      "loss": 1.4823,
      "step": 323
    },
    {
      "epoch": 0.47699668752300334,
      "grad_norm": 0.1304844319820404,
      "learning_rate": 0.00011626371651948838,
      "loss": 1.485,
      "step": 324
    },
    {
      "epoch": 0.4784688995215311,
      "grad_norm": 0.12832032144069672,
      "learning_rate": 0.0001156434465040231,
      "loss": 1.5102,
      "step": 325
    },
    {
      "epoch": 0.4799411115200589,
      "grad_norm": 0.13046902418136597,
      "learning_rate": 0.00011502255891207572,
      "loss": 1.5026,
      "step": 326
    },
    {
      "epoch": 0.4814133235185867,
      "grad_norm": 0.12792472541332245,
      "learning_rate": 0.00011440107825522521,
      "loss": 1.5254,
      "step": 327
    },
    {
      "epoch": 0.48288553551711444,
      "grad_norm": 0.12645626068115234,
      "learning_rate": 0.0001137790290684638,
      "loss": 1.521,
      "step": 328
    },
    {
      "epoch": 0.48435774751564226,
      "grad_norm": 0.1297602504491806,
      "learning_rate": 0.00011315643590922827,
      "loss": 1.5172,
      "step": 329
    },
    {
      "epoch": 0.48582995951417,
      "grad_norm": 0.12985859811306,
      "learning_rate": 0.00011253332335643043,
      "loss": 1.518,
      "step": 330
    },
    {
      "epoch": 0.48730217151269783,
      "grad_norm": 0.1302354633808136,
      "learning_rate": 0.00011190971600948699,
      "loss": 1.5216,
      "step": 331
    },
    {
      "epoch": 0.4887743835112256,
      "grad_norm": 0.13054907321929932,
      "learning_rate": 0.00011128563848734816,
      "loss": 1.5101,
      "step": 332
    },
    {
      "epoch": 0.4902465955097534,
      "grad_norm": 0.13132669031620026,
      "learning_rate": 0.000110661115427526,
      "loss": 1.4761,
      "step": 333
    },
    {
      "epoch": 0.4917188075082812,
      "grad_norm": 0.1325066238641739,
      "learning_rate": 0.00011003617148512149,
      "loss": 1.5254,
      "step": 334
    },
    {
      "epoch": 0.493191019506809,
      "grad_norm": 0.12612849473953247,
      "learning_rate": 0.00010941083133185146,
      "loss": 1.5047,
      "step": 335
    },
    {
      "epoch": 0.49466323150533675,
      "grad_norm": 0.13487622141838074,
      "learning_rate": 0.00010878511965507434,
      "loss": 1.5185,
      "step": 336
    },
    {
      "epoch": 0.49613544350386457,
      "grad_norm": 0.13075707852840424,
      "learning_rate": 0.00010815906115681578,
      "loss": 1.5103,
      "step": 337
    },
    {
      "epoch": 0.49760765550239233,
      "grad_norm": 0.12982569634914398,
      "learning_rate": 0.00010753268055279329,
      "loss": 1.5045,
      "step": 338
    },
    {
      "epoch": 0.49907986750092015,
      "grad_norm": 0.13324594497680664,
      "learning_rate": 0.00010690600257144061,
      "loss": 1.5082,
      "step": 339
    },
    {
      "epoch": 0.5005520794994479,
      "grad_norm": 0.13495436310768127,
      "learning_rate": 0.00010627905195293135,
      "loss": 1.4749,
      "step": 340
    },
    {
      "epoch": 0.5020242914979757,
      "grad_norm": 0.13739405572414398,
      "learning_rate": 0.00010565185344820247,
      "loss": 1.5228,
      "step": 341
    },
    {
      "epoch": 0.5034965034965035,
      "grad_norm": 0.13414184749126434,
      "learning_rate": 0.00010502443181797697,
      "loss": 1.5117,
      "step": 342
    },
    {
      "epoch": 0.5049687154950313,
      "grad_norm": 0.1293892115354538,
      "learning_rate": 0.0001043968118317865,
      "loss": 1.5017,
      "step": 343
    },
    {
      "epoch": 0.5064409274935591,
      "grad_norm": 0.13166604936122894,
      "learning_rate": 0.00010376901826699348,
      "loss": 1.5201,
      "step": 344
    },
    {
      "epoch": 0.5079131394920868,
      "grad_norm": 0.13358330726623535,
      "learning_rate": 0.00010314107590781284,
      "loss": 1.5155,
      "step": 345
    },
    {
      "epoch": 0.5093853514906147,
      "grad_norm": 0.13341715931892395,
      "learning_rate": 0.00010251300954433376,
      "loss": 1.5367,
      "step": 346
    },
    {
      "epoch": 0.5108575634891425,
      "grad_norm": 0.12866951525211334,
      "learning_rate": 0.00010188484397154084,
      "loss": 1.4939,
      "step": 347
    },
    {
      "epoch": 0.5123297754876702,
      "grad_norm": 0.12844346463680267,
      "learning_rate": 0.00010125660398833528,
      "loss": 1.5175,
      "step": 348
    },
    {
      "epoch": 0.513801987486198,
      "grad_norm": 0.13152100145816803,
      "learning_rate": 0.00010062831439655591,
      "loss": 1.4871,
      "step": 349
    },
    {
      "epoch": 0.5152741994847257,
      "grad_norm": 0.1332051008939743,
      "learning_rate": 0.0001,
      "loss": 1.5156,
      "step": 350
    },
    {
      "epoch": 0.5167464114832536,
      "grad_norm": 0.13279229402542114,
      "learning_rate": 9.937168560344412e-05,
      "loss": 1.4889,
      "step": 351
    },
    {
      "epoch": 0.5182186234817814,
      "grad_norm": 0.13278289139270782,
      "learning_rate": 9.874339601166473e-05,
      "loss": 1.5125,
      "step": 352
    },
    {
      "epoch": 0.5196908354803091,
      "grad_norm": 0.13304084539413452,
      "learning_rate": 9.81151560284592e-05,
      "loss": 1.4954,
      "step": 353
    },
    {
      "epoch": 0.5211630474788369,
      "grad_norm": 0.12908920645713806,
      "learning_rate": 9.748699045566626e-05,
      "loss": 1.5185,
      "step": 354
    },
    {
      "epoch": 0.5226352594773648,
      "grad_norm": 0.1331358551979065,
      "learning_rate": 9.685892409218717e-05,
      "loss": 1.5097,
      "step": 355
    },
    {
      "epoch": 0.5241074714758925,
      "grad_norm": 0.13341255486011505,
      "learning_rate": 9.623098173300654e-05,
      "loss": 1.5299,
      "step": 356
    },
    {
      "epoch": 0.5255796834744203,
      "grad_norm": 0.13090355694293976,
      "learning_rate": 9.560318816821353e-05,
      "loss": 1.4925,
      "step": 357
    },
    {
      "epoch": 0.5270518954729481,
      "grad_norm": 0.12971143424510956,
      "learning_rate": 9.497556818202306e-05,
      "loss": 1.5061,
      "step": 358
    },
    {
      "epoch": 0.5285241074714759,
      "grad_norm": 0.13158752024173737,
      "learning_rate": 9.434814655179755e-05,
      "loss": 1.4943,
      "step": 359
    },
    {
      "epoch": 0.5299963194700037,
      "grad_norm": 0.12975053489208221,
      "learning_rate": 9.372094804706867e-05,
      "loss": 1.5021,
      "step": 360
    },
    {
      "epoch": 0.5314685314685315,
      "grad_norm": 0.13350540399551392,
      "learning_rate": 9.309399742855942e-05,
      "loss": 1.5324,
      "step": 361
    },
    {
      "epoch": 0.5329407434670592,
      "grad_norm": 0.12975060939788818,
      "learning_rate": 9.246731944720675e-05,
      "loss": 1.4826,
      "step": 362
    },
    {
      "epoch": 0.5344129554655871,
      "grad_norm": 0.13513784110546112,
      "learning_rate": 9.184093884318425e-05,
      "loss": 1.4827,
      "step": 363
    },
    {
      "epoch": 0.5358851674641149,
      "grad_norm": 0.13155671954154968,
      "learning_rate": 9.121488034492569e-05,
      "loss": 1.4977,
      "step": 364
    },
    {
      "epoch": 0.5373573794626426,
      "grad_norm": 0.13195782899856567,
      "learning_rate": 9.058916866814858e-05,
      "loss": 1.4776,
      "step": 365
    },
    {
      "epoch": 0.5388295914611704,
      "grad_norm": 0.1299077868461609,
      "learning_rate": 8.99638285148785e-05,
      "loss": 1.4987,
      "step": 366
    },
    {
      "epoch": 0.5403018034596982,
      "grad_norm": 0.13166947662830353,
      "learning_rate": 8.933888457247402e-05,
      "loss": 1.4735,
      "step": 367
    },
    {
      "epoch": 0.541774015458226,
      "grad_norm": 0.1331235021352768,
      "learning_rate": 8.871436151265184e-05,
      "loss": 1.4672,
      "step": 368
    },
    {
      "epoch": 0.5432462274567538,
      "grad_norm": 0.13196539878845215,
      "learning_rate": 8.809028399051302e-05,
      "loss": 1.5089,
      "step": 369
    },
    {
      "epoch": 0.5447184394552815,
      "grad_norm": 0.1342688351869583,
      "learning_rate": 8.746667664356956e-05,
      "loss": 1.5031,
      "step": 370
    },
    {
      "epoch": 0.5461906514538093,
      "grad_norm": 0.13164031505584717,
      "learning_rate": 8.684356409077176e-05,
      "loss": 1.4897,
      "step": 371
    },
    {
      "epoch": 0.5476628634523372,
      "grad_norm": 0.13104857504367828,
      "learning_rate": 8.62209709315362e-05,
      "loss": 1.4855,
      "step": 372
    },
    {
      "epoch": 0.5491350754508649,
      "grad_norm": 0.13208110630512238,
      "learning_rate": 8.559892174477479e-05,
      "loss": 1.5078,
      "step": 373
    },
    {
      "epoch": 0.5506072874493927,
      "grad_norm": 0.13312184810638428,
      "learning_rate": 8.497744108792429e-05,
      "loss": 1.5159,
      "step": 374
    },
    {
      "epoch": 0.5520794994479205,
      "grad_norm": 0.13393402099609375,
      "learning_rate": 8.435655349597689e-05,
      "loss": 1.4963,
      "step": 375
    },
    {
      "epoch": 0.5535517114464483,
      "grad_norm": 0.13724307715892792,
      "learning_rate": 8.373628348051165e-05,
      "loss": 1.4967,
      "step": 376
    },
    {
      "epoch": 0.5550239234449761,
      "grad_norm": 0.13683705031871796,
      "learning_rate": 8.311665552872662e-05,
      "loss": 1.5119,
      "step": 377
    },
    {
      "epoch": 0.5564961354435038,
      "grad_norm": 0.13283081352710724,
      "learning_rate": 8.249769410247239e-05,
      "loss": 1.4868,
      "step": 378
    },
    {
      "epoch": 0.5579683474420316,
      "grad_norm": 0.13637295365333557,
      "learning_rate": 8.187942363728625e-05,
      "loss": 1.4909,
      "step": 379
    },
    {
      "epoch": 0.5594405594405595,
      "grad_norm": 0.13765372335910797,
      "learning_rate": 8.126186854142752e-05,
      "loss": 1.4663,
      "step": 380
    },
    {
      "epoch": 0.5609127714390872,
      "grad_norm": 0.1324106901884079,
      "learning_rate": 8.064505319491398e-05,
      "loss": 1.488,
      "step": 381
    },
    {
      "epoch": 0.562384983437615,
      "grad_norm": 0.13049346208572388,
      "learning_rate": 8.002900194855932e-05,
      "loss": 1.5099,
      "step": 382
    },
    {
      "epoch": 0.5638571954361428,
      "grad_norm": 0.1328127533197403,
      "learning_rate": 7.941373912301189e-05,
      "loss": 1.4883,
      "step": 383
    },
    {
      "epoch": 0.5653294074346706,
      "grad_norm": 0.1314123123884201,
      "learning_rate": 7.879928900779456e-05,
      "loss": 1.5202,
      "step": 384
    },
    {
      "epoch": 0.5668016194331984,
      "grad_norm": 0.13271084427833557,
      "learning_rate": 7.818567586034577e-05,
      "loss": 1.4894,
      "step": 385
    },
    {
      "epoch": 0.5682738314317262,
      "grad_norm": 0.13358627259731293,
      "learning_rate": 7.75729239050619e-05,
      "loss": 1.5011,
      "step": 386
    },
    {
      "epoch": 0.5697460434302539,
      "grad_norm": 0.13315322995185852,
      "learning_rate": 7.696105733234098e-05,
      "loss": 1.5229,
      "step": 387
    },
    {
      "epoch": 0.5712182554287818,
      "grad_norm": 0.1348247230052948,
      "learning_rate": 7.635010029762756e-05,
      "loss": 1.5123,
      "step": 388
    },
    {
      "epoch": 0.5726904674273096,
      "grad_norm": 0.13394920527935028,
      "learning_rate": 7.574007692045928e-05,
      "loss": 1.495,
      "step": 389
    },
    {
      "epoch": 0.5741626794258373,
      "grad_norm": 0.13603061437606812,
      "learning_rate": 7.513101128351454e-05,
      "loss": 1.4856,
      "step": 390
    },
    {
      "epoch": 0.5756348914243651,
      "grad_norm": 0.13142289221286774,
      "learning_rate": 7.45229274316618e-05,
      "loss": 1.5608,
      "step": 391
    },
    {
      "epoch": 0.5771071034228928,
      "grad_norm": 0.13422873616218567,
      "learning_rate": 7.391584937101033e-05,
      "loss": 1.5043,
      "step": 392
    },
    {
      "epoch": 0.5785793154214207,
      "grad_norm": 0.13258948922157288,
      "learning_rate": 7.330980106796246e-05,
      "loss": 1.4855,
      "step": 393
    },
    {
      "epoch": 0.5800515274199485,
      "grad_norm": 0.13000285625457764,
      "learning_rate": 7.270480644826749e-05,
      "loss": 1.5094,
      "step": 394
    },
    {
      "epoch": 0.5815237394184762,
      "grad_norm": 0.13080152869224548,
      "learning_rate": 7.210088939607708e-05,
      "loss": 1.4993,
      "step": 395
    },
    {
      "epoch": 0.582995951417004,
      "grad_norm": 0.1318790316581726,
      "learning_rate": 7.149807375300239e-05,
      "loss": 1.5071,
      "step": 396
    },
    {
      "epoch": 0.5844681634155319,
      "grad_norm": 0.1368306428194046,
      "learning_rate": 7.089638331717284e-05,
      "loss": 1.4791,
      "step": 397
    },
    {
      "epoch": 0.5859403754140596,
      "grad_norm": 0.13718125224113464,
      "learning_rate": 7.029584184229653e-05,
      "loss": 1.4936,
      "step": 398
    },
    {
      "epoch": 0.5874125874125874,
      "grad_norm": 0.13892057538032532,
      "learning_rate": 6.969647303672262e-05,
      "loss": 1.515,
      "step": 399
    },
    {
      "epoch": 0.5888847994111152,
      "grad_norm": 0.13341866433620453,
      "learning_rate": 6.909830056250527e-05,
      "loss": 1.5339,
      "step": 400
    },
    {
      "epoch": 0.590357011409643,
      "grad_norm": 0.13300460577011108,
      "learning_rate": 6.850134803446954e-05,
      "loss": 1.4831,
      "step": 401
    },
    {
      "epoch": 0.5918292234081708,
      "grad_norm": 0.13199138641357422,
      "learning_rate": 6.790563901927907e-05,
      "loss": 1.4638,
      "step": 402
    },
    {
      "epoch": 0.5933014354066986,
      "grad_norm": 0.13444076478481293,
      "learning_rate": 6.731119703450577e-05,
      "loss": 1.4739,
      "step": 403
    },
    {
      "epoch": 0.5947736474052263,
      "grad_norm": 0.13487786054611206,
      "learning_rate": 6.671804554770135e-05,
      "loss": 1.5074,
      "step": 404
    },
    {
      "epoch": 0.5962458594037542,
      "grad_norm": 0.13607552647590637,
      "learning_rate": 6.612620797547087e-05,
      "loss": 1.4938,
      "step": 405
    },
    {
      "epoch": 0.597718071402282,
      "grad_norm": 0.13471536338329315,
      "learning_rate": 6.55357076825483e-05,
      "loss": 1.4987,
      "step": 406
    },
    {
      "epoch": 0.5991902834008097,
      "grad_norm": 0.13304749131202698,
      "learning_rate": 6.494656798087412e-05,
      "loss": 1.5281,
      "step": 407
    },
    {
      "epoch": 0.6006624953993375,
      "grad_norm": 0.13069571554660797,
      "learning_rate": 6.435881212867493e-05,
      "loss": 1.4995,
      "step": 408
    },
    {
      "epoch": 0.6021347073978653,
      "grad_norm": 0.13390809297561646,
      "learning_rate": 6.377246332954544e-05,
      "loss": 1.4832,
      "step": 409
    },
    {
      "epoch": 0.6036069193963931,
      "grad_norm": 0.13501663506031036,
      "learning_rate": 6.318754473153221e-05,
      "loss": 1.4796,
      "step": 410
    },
    {
      "epoch": 0.6050791313949209,
      "grad_norm": 0.13101309537887573,
      "learning_rate": 6.260407942621998e-05,
      "loss": 1.5158,
      "step": 411
    },
    {
      "epoch": 0.6065513433934486,
      "grad_norm": 0.13380104303359985,
      "learning_rate": 6.20220904478199e-05,
      "loss": 1.4903,
      "step": 412
    },
    {
      "epoch": 0.6080235553919764,
      "grad_norm": 0.1364106833934784,
      "learning_rate": 6.144160077226036e-05,
      "loss": 1.5233,
      "step": 413
    },
    {
      "epoch": 0.6094957673905043,
      "grad_norm": 0.13187123835086823,
      "learning_rate": 6.086263331627976e-05,
      "loss": 1.4854,
      "step": 414
    },
    {
      "epoch": 0.610967979389032,
      "grad_norm": 0.13597944378852844,
      "learning_rate": 6.0285210936521955e-05,
      "loss": 1.4675,
      "step": 415
    },
    {
      "epoch": 0.6124401913875598,
      "grad_norm": 0.1344550997018814,
      "learning_rate": 5.9709356428633746e-05,
      "loss": 1.4986,
      "step": 416
    },
    {
      "epoch": 0.6139124033860875,
      "grad_norm": 0.13348796963691711,
      "learning_rate": 5.913509252636511e-05,
      "loss": 1.5225,
      "step": 417
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 0.1340065449476242,
      "learning_rate": 5.856244190067159e-05,
      "loss": 1.5027,
      "step": 418
    },
    {
      "epoch": 0.6168568273831432,
      "grad_norm": 0.13286136090755463,
      "learning_rate": 5.799142715881938e-05,
      "loss": 1.4899,
      "step": 419
    },
    {
      "epoch": 0.6183290393816709,
      "grad_norm": 0.13747188448905945,
      "learning_rate": 5.7422070843492734e-05,
      "loss": 1.5028,
      "step": 420
    },
    {
      "epoch": 0.6198012513801987,
      "grad_norm": 0.13422006368637085,
      "learning_rate": 5.6854395431904094e-05,
      "loss": 1.4868,
      "step": 421
    },
    {
      "epoch": 0.6212734633787266,
      "grad_norm": 0.13465476036071777,
      "learning_rate": 5.6288423334906735e-05,
      "loss": 1.4949,
      "step": 422
    },
    {
      "epoch": 0.6227456753772543,
      "grad_norm": 0.13342292606830597,
      "learning_rate": 5.572417689610987e-05,
      "loss": 1.4778,
      "step": 423
    },
    {
      "epoch": 0.6242178873757821,
      "grad_norm": 0.1357763260602951,
      "learning_rate": 5.5161678390996796e-05,
      "loss": 1.5088,
      "step": 424
    },
    {
      "epoch": 0.6256900993743099,
      "grad_norm": 0.13748598098754883,
      "learning_rate": 5.4600950026045326e-05,
      "loss": 1.4999,
      "step": 425
    },
    {
      "epoch": 0.6271623113728377,
      "grad_norm": 0.13297002017498016,
      "learning_rate": 5.404201393785122e-05,
      "loss": 1.496,
      "step": 426
    },
    {
      "epoch": 0.6286345233713655,
      "grad_norm": 0.13217949867248535,
      "learning_rate": 5.348489219225416e-05,
      "loss": 1.4717,
      "step": 427
    },
    {
      "epoch": 0.6301067353698933,
      "grad_norm": 0.1306363046169281,
      "learning_rate": 5.292960678346675e-05,
      "loss": 1.49,
      "step": 428
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 0.13448570668697357,
      "learning_rate": 5.237617963320608e-05,
      "loss": 1.493,
      "step": 429
    },
    {
      "epoch": 0.6330511593669489,
      "grad_norm": 0.13820518553256989,
      "learning_rate": 5.182463258982846e-05,
      "loss": 1.5271,
      "step": 430
    },
    {
      "epoch": 0.6345233713654767,
      "grad_norm": 0.1368243545293808,
      "learning_rate": 5.127498742746675e-05,
      "loss": 1.4695,
      "step": 431
    },
    {
      "epoch": 0.6359955833640044,
      "grad_norm": 0.13261400163173676,
      "learning_rate": 5.072726584517086e-05,
      "loss": 1.5143,
      "step": 432
    },
    {
      "epoch": 0.6374677953625322,
      "grad_norm": 0.13191469013690948,
      "learning_rate": 5.018148946605092e-05,
      "loss": 1.4873,
      "step": 433
    },
    {
      "epoch": 0.63894000736106,
      "grad_norm": 0.13407787680625916,
      "learning_rate": 4.9637679836423924e-05,
      "loss": 1.4909,
      "step": 434
    },
    {
      "epoch": 0.6404122193595878,
      "grad_norm": 0.13669706881046295,
      "learning_rate": 4.909585842496287e-05,
      "loss": 1.509,
      "step": 435
    },
    {
      "epoch": 0.6418844313581156,
      "grad_norm": 0.132285013794899,
      "learning_rate": 4.8556046621849346e-05,
      "loss": 1.532,
      "step": 436
    },
    {
      "epoch": 0.6433566433566433,
      "grad_norm": 0.13264308869838715,
      "learning_rate": 4.8018265737929044e-05,
      "loss": 1.4943,
      "step": 437
    },
    {
      "epoch": 0.6448288553551711,
      "grad_norm": 0.13449332118034363,
      "learning_rate": 4.748253700387042e-05,
      "loss": 1.5179,
      "step": 438
    },
    {
      "epoch": 0.646301067353699,
      "grad_norm": 0.1325521618127823,
      "learning_rate": 4.694888156932658e-05,
      "loss": 1.4932,
      "step": 439
    },
    {
      "epoch": 0.6477732793522267,
      "grad_norm": 0.13812002539634705,
      "learning_rate": 4.6417320502100316e-05,
      "loss": 1.4918,
      "step": 440
    },
    {
      "epoch": 0.6492454913507545,
      "grad_norm": 0.13224385678768158,
      "learning_rate": 4.588787478731242e-05,
      "loss": 1.5026,
      "step": 441
    },
    {
      "epoch": 0.6507177033492823,
      "grad_norm": 0.1375768482685089,
      "learning_rate": 4.5360565326573104e-05,
      "loss": 1.459,
      "step": 442
    },
    {
      "epoch": 0.6521899153478101,
      "grad_norm": 0.1358180046081543,
      "learning_rate": 4.483541293715698e-05,
      "loss": 1.4691,
      "step": 443
    },
    {
      "epoch": 0.6536621273463379,
      "grad_norm": 0.13343274593353271,
      "learning_rate": 4.431243835118124e-05,
      "loss": 1.5098,
      "step": 444
    },
    {
      "epoch": 0.6551343393448656,
      "grad_norm": 0.1360098421573639,
      "learning_rate": 4.379166221478697e-05,
      "loss": 1.5008,
      "step": 445
    },
    {
      "epoch": 0.6566065513433934,
      "grad_norm": 0.12845464050769806,
      "learning_rate": 4.327310508732437e-05,
      "loss": 1.5002,
      "step": 446
    },
    {
      "epoch": 0.6580787633419213,
      "grad_norm": 0.13384786248207092,
      "learning_rate": 4.2756787440540936e-05,
      "loss": 1.4796,
      "step": 447
    },
    {
      "epoch": 0.659550975340449,
      "grad_norm": 0.12972962856292725,
      "learning_rate": 4.224272965777326e-05,
      "loss": 1.5039,
      "step": 448
    },
    {
      "epoch": 0.6610231873389768,
      "grad_norm": 0.1361577957868576,
      "learning_rate": 4.173095203314241e-05,
      "loss": 1.467,
      "step": 449
    },
    {
      "epoch": 0.6624953993375046,
      "grad_norm": 0.13251101970672607,
      "learning_rate": 4.12214747707527e-05,
      "loss": 1.5112,
      "step": 450
    },
    {
      "epoch": 0.6639676113360324,
      "grad_norm": 0.13592743873596191,
      "learning_rate": 4.071431798389408e-05,
      "loss": 1.5007,
      "step": 451
    },
    {
      "epoch": 0.6654398233345602,
      "grad_norm": 0.1373070627450943,
      "learning_rate": 4.020950169424815e-05,
      "loss": 1.5036,
      "step": 452
    },
    {
      "epoch": 0.666912035333088,
      "grad_norm": 0.13762016594409943,
      "learning_rate": 3.9707045831097555e-05,
      "loss": 1.5029,
      "step": 453
    },
    {
      "epoch": 0.6683842473316157,
      "grad_norm": 0.13593138754367828,
      "learning_rate": 3.920697023053949e-05,
      "loss": 1.4634,
      "step": 454
    },
    {
      "epoch": 0.6698564593301436,
      "grad_norm": 0.13403427600860596,
      "learning_rate": 3.8709294634702376e-05,
      "loss": 1.4959,
      "step": 455
    },
    {
      "epoch": 0.6713286713286714,
      "grad_norm": 0.13755904138088226,
      "learning_rate": 3.821403869096658e-05,
      "loss": 1.4872,
      "step": 456
    },
    {
      "epoch": 0.6728008833271991,
      "grad_norm": 0.13614238798618317,
      "learning_rate": 3.7721221951188765e-05,
      "loss": 1.521,
      "step": 457
    },
    {
      "epoch": 0.6742730953257269,
      "grad_norm": 0.13538721203804016,
      "learning_rate": 3.7230863870929964e-05,
      "loss": 1.5415,
      "step": 458
    },
    {
      "epoch": 0.6757453073242546,
      "grad_norm": 0.13466057181358337,
      "learning_rate": 3.674298380868756e-05,
      "loss": 1.4752,
      "step": 459
    },
    {
      "epoch": 0.6772175193227825,
      "grad_norm": 0.13793981075286865,
      "learning_rate": 3.6257601025131026e-05,
      "loss": 1.4996,
      "step": 460
    },
    {
      "epoch": 0.6786897313213103,
      "grad_norm": 0.13543136417865753,
      "learning_rate": 3.577473468234156e-05,
      "loss": 1.4936,
      "step": 461
    },
    {
      "epoch": 0.680161943319838,
      "grad_norm": 0.1373487114906311,
      "learning_rate": 3.52944038430556e-05,
      "loss": 1.4803,
      "step": 462
    },
    {
      "epoch": 0.6816341553183658,
      "grad_norm": 0.13347399234771729,
      "learning_rate": 3.481662746991214e-05,
      "loss": 1.5085,
      "step": 463
    },
    {
      "epoch": 0.6831063673168937,
      "grad_norm": 0.13949111104011536,
      "learning_rate": 3.4341424424704375e-05,
      "loss": 1.5072,
      "step": 464
    },
    {
      "epoch": 0.6845785793154214,
      "grad_norm": 0.13755980134010315,
      "learning_rate": 3.386881346763483e-05,
      "loss": 1.499,
      "step": 465
    },
    {
      "epoch": 0.6860507913139492,
      "grad_norm": 0.1401529610157013,
      "learning_rate": 3.339881325657484e-05,
      "loss": 1.4602,
      "step": 466
    },
    {
      "epoch": 0.687523003312477,
      "grad_norm": 0.13102111220359802,
      "learning_rate": 3.2931442346328004e-05,
      "loss": 1.4891,
      "step": 467
    },
    {
      "epoch": 0.6889952153110048,
      "grad_norm": 0.13376979529857635,
      "learning_rate": 3.246671918789755e-05,
      "loss": 1.4706,
      "step": 468
    },
    {
      "epoch": 0.6904674273095326,
      "grad_norm": 0.1339038610458374,
      "learning_rate": 3.200466212775808e-05,
      "loss": 1.4991,
      "step": 469
    },
    {
      "epoch": 0.6919396393080604,
      "grad_norm": 0.13536672294139862,
      "learning_rate": 3.154528940713113e-05,
      "loss": 1.4909,
      "step": 470
    },
    {
      "epoch": 0.6934118513065881,
      "grad_norm": 0.139925017952919,
      "learning_rate": 3.108861916126518e-05,
      "loss": 1.4971,
      "step": 471
    },
    {
      "epoch": 0.694884063305116,
      "grad_norm": 0.1338726282119751,
      "learning_rate": 3.063466941871952e-05,
      "loss": 1.4559,
      "step": 472
    },
    {
      "epoch": 0.6963562753036437,
      "grad_norm": 0.13365983963012695,
      "learning_rate": 3.018345810065275e-05,
      "loss": 1.509,
      "step": 473
    },
    {
      "epoch": 0.6978284873021715,
      "grad_norm": 0.1354924589395523,
      "learning_rate": 2.9735003020115092e-05,
      "loss": 1.5307,
      "step": 474
    },
    {
      "epoch": 0.6993006993006993,
      "grad_norm": 0.13422416150569916,
      "learning_rate": 2.9289321881345254e-05,
      "loss": 1.508,
      "step": 475
    },
    {
      "epoch": 0.7007729112992271,
      "grad_norm": 0.13424506783485413,
      "learning_rate": 2.8846432279071467e-05,
      "loss": 1.4862,
      "step": 476
    },
    {
      "epoch": 0.7022451232977549,
      "grad_norm": 0.1364148110151291,
      "learning_rate": 2.840635169781688e-05,
      "loss": 1.4953,
      "step": 477
    },
    {
      "epoch": 0.7037173352962827,
      "grad_norm": 0.13503850996494293,
      "learning_rate": 2.7969097511209308e-05,
      "loss": 1.4765,
      "step": 478
    },
    {
      "epoch": 0.7051895472948104,
      "grad_norm": 0.13706064224243164,
      "learning_rate": 2.753468698129533e-05,
      "loss": 1.4836,
      "step": 479
    },
    {
      "epoch": 0.7066617592933382,
      "grad_norm": 0.1361798346042633,
      "learning_rate": 2.7103137257858868e-05,
      "loss": 1.4539,
      "step": 480
    },
    {
      "epoch": 0.7081339712918661,
      "grad_norm": 0.1333787888288498,
      "learning_rate": 2.6674465377744017e-05,
      "loss": 1.4765,
      "step": 481
    },
    {
      "epoch": 0.7096061832903938,
      "grad_norm": 0.1392863541841507,
      "learning_rate": 2.624868826418262e-05,
      "loss": 1.4811,
      "step": 482
    },
    {
      "epoch": 0.7110783952889216,
      "grad_norm": 0.13769815862178802,
      "learning_rate": 2.582582272612609e-05,
      "loss": 1.4952,
      "step": 483
    },
    {
      "epoch": 0.7125506072874493,
      "grad_norm": 0.13681407272815704,
      "learning_rate": 2.540588545758179e-05,
      "loss": 1.4624,
      "step": 484
    },
    {
      "epoch": 0.7140228192859772,
      "grad_norm": 0.13446034491062164,
      "learning_rate": 2.4988893036954043e-05,
      "loss": 1.4792,
      "step": 485
    },
    {
      "epoch": 0.715495031284505,
      "grad_norm": 0.1320485919713974,
      "learning_rate": 2.4574861926389615e-05,
      "loss": 1.5093,
      "step": 486
    },
    {
      "epoch": 0.7169672432830327,
      "grad_norm": 0.1359575241804123,
      "learning_rate": 2.4163808471127812e-05,
      "loss": 1.5053,
      "step": 487
    },
    {
      "epoch": 0.7184394552815605,
      "grad_norm": 0.1347683072090149,
      "learning_rate": 2.37557488988552e-05,
      "loss": 1.4721,
      "step": 488
    },
    {
      "epoch": 0.7199116672800884,
      "grad_norm": 0.13442374765872955,
      "learning_rate": 2.3350699319065026e-05,
      "loss": 1.4853,
      "step": 489
    },
    {
      "epoch": 0.7213838792786161,
      "grad_norm": 0.13414570689201355,
      "learning_rate": 2.2948675722421086e-05,
      "loss": 1.4859,
      "step": 490
    },
    {
      "epoch": 0.7228560912771439,
      "grad_norm": 0.13509652018547058,
      "learning_rate": 2.254969398012663e-05,
      "loss": 1.4945,
      "step": 491
    },
    {
      "epoch": 0.7243283032756717,
      "grad_norm": 0.1353853940963745,
      "learning_rate": 2.2153769843297667e-05,
      "loss": 1.5139,
      "step": 492
    },
    {
      "epoch": 0.7258005152741995,
      "grad_norm": 0.1363237500190735,
      "learning_rate": 2.1760918942341192e-05,
      "loss": 1.4621,
      "step": 493
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.13358739018440247,
      "learning_rate": 2.137115678633811e-05,
      "loss": 1.4904,
      "step": 494
    },
    {
      "epoch": 0.728744939271255,
      "grad_norm": 0.13719165325164795,
      "learning_rate": 2.098449876243096e-05,
      "loss": 1.4819,
      "step": 495
    },
    {
      "epoch": 0.7302171512697828,
      "grad_norm": 0.13620305061340332,
      "learning_rate": 2.0600960135216462e-05,
      "loss": 1.4847,
      "step": 496
    },
    {
      "epoch": 0.7316893632683107,
      "grad_norm": 0.13385285437107086,
      "learning_rate": 2.0220556046142893e-05,
      "loss": 1.4856,
      "step": 497
    },
    {
      "epoch": 0.7331615752668385,
      "grad_norm": 0.13555505871772766,
      "learning_rate": 1.9843301512912327e-05,
      "loss": 1.4985,
      "step": 498
    },
    {
      "epoch": 0.7346337872653662,
      "grad_norm": 0.1352514922618866,
      "learning_rate": 1.946921142888781e-05,
      "loss": 1.4604,
      "step": 499
    },
    {
      "epoch": 0.736105999263894,
      "grad_norm": 0.13447973132133484,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 1.4937,
      "step": 500
    },
    {
      "epoch": 0.7375782112624217,
      "grad_norm": 0.13288331031799316,
      "learning_rate": 1.8730583556690605e-05,
      "loss": 1.4737,
      "step": 501
    },
    {
      "epoch": 0.7390504232609496,
      "grad_norm": 0.13768433034420013,
      "learning_rate": 1.8366074928281607e-05,
      "loss": 1.4945,
      "step": 502
    },
    {
      "epoch": 0.7405226352594774,
      "grad_norm": 0.1377992033958435,
      "learning_rate": 1.8004789067454764e-05,
      "loss": 1.5027,
      "step": 503
    },
    {
      "epoch": 0.7419948472580051,
      "grad_norm": 0.13398252427577972,
      "learning_rate": 1.7646740237157256e-05,
      "loss": 1.4893,
      "step": 504
    },
    {
      "epoch": 0.7434670592565329,
      "grad_norm": 0.12989221513271332,
      "learning_rate": 1.7291942572543807e-05,
      "loss": 1.4819,
      "step": 505
    },
    {
      "epoch": 0.7449392712550608,
      "grad_norm": 0.13395771384239197,
      "learning_rate": 1.6940410080418723e-05,
      "loss": 1.4873,
      "step": 506
    },
    {
      "epoch": 0.7464114832535885,
      "grad_norm": 0.1362403780221939,
      "learning_rate": 1.6592156638682886e-05,
      "loss": 1.4932,
      "step": 507
    },
    {
      "epoch": 0.7478836952521163,
      "grad_norm": 0.13380947709083557,
      "learning_rate": 1.6247195995785837e-05,
      "loss": 1.5001,
      "step": 508
    },
    {
      "epoch": 0.749355907250644,
      "grad_norm": 0.1316158026456833,
      "learning_rate": 1.5905541770183096e-05,
      "loss": 1.4952,
      "step": 509
    },
    {
      "epoch": 0.7508281192491719,
      "grad_norm": 0.13325780630111694,
      "learning_rate": 1.5567207449798515e-05,
      "loss": 1.4768,
      "step": 510
    },
    {
      "epoch": 0.7523003312476997,
      "grad_norm": 0.13417863845825195,
      "learning_rate": 1.5232206391491699e-05,
      "loss": 1.4612,
      "step": 511
    },
    {
      "epoch": 0.7537725432462274,
      "grad_norm": 0.13535621762275696,
      "learning_rate": 1.4900551820530828e-05,
      "loss": 1.4739,
      "step": 512
    },
    {
      "epoch": 0.7552447552447552,
      "grad_norm": 0.13501548767089844,
      "learning_rate": 1.4572256830070497e-05,
      "loss": 1.4371,
      "step": 513
    },
    {
      "epoch": 0.7567169672432831,
      "grad_norm": 0.13107748329639435,
      "learning_rate": 1.4247334380634792e-05,
      "loss": 1.4796,
      "step": 514
    },
    {
      "epoch": 0.7581891792418108,
      "grad_norm": 0.13535360991954803,
      "learning_rate": 1.3925797299605647e-05,
      "loss": 1.499,
      "step": 515
    },
    {
      "epoch": 0.7596613912403386,
      "grad_norm": 0.14287012815475464,
      "learning_rate": 1.3607658280716473e-05,
      "loss": 1.4779,
      "step": 516
    },
    {
      "epoch": 0.7611336032388664,
      "grad_norm": 0.13279961049556732,
      "learning_rate": 1.3292929883550998e-05,
      "loss": 1.4581,
      "step": 517
    },
    {
      "epoch": 0.7626058152373942,
      "grad_norm": 0.13332952558994293,
      "learning_rate": 1.2981624533047432e-05,
      "loss": 1.4959,
      "step": 518
    },
    {
      "epoch": 0.764078027235922,
      "grad_norm": 0.13875915110111237,
      "learning_rate": 1.2673754519008008e-05,
      "loss": 1.4875,
      "step": 519
    },
    {
      "epoch": 0.7655502392344498,
      "grad_norm": 0.13662198185920715,
      "learning_rate": 1.2369331995613665e-05,
      "loss": 1.4989,
      "step": 520
    },
    {
      "epoch": 0.7670224512329775,
      "grad_norm": 0.13764922320842743,
      "learning_rate": 1.206836898094439e-05,
      "loss": 1.4715,
      "step": 521
    },
    {
      "epoch": 0.7684946632315053,
      "grad_norm": 0.14062534272670746,
      "learning_rate": 1.1770877356504683e-05,
      "loss": 1.4439,
      "step": 522
    },
    {
      "epoch": 0.7699668752300332,
      "grad_norm": 0.13690273463726044,
      "learning_rate": 1.1476868866754486e-05,
      "loss": 1.4845,
      "step": 523
    },
    {
      "epoch": 0.7714390872285609,
      "grad_norm": 0.1317390352487564,
      "learning_rate": 1.1186355118645554e-05,
      "loss": 1.5173,
      "step": 524
    },
    {
      "epoch": 0.7729112992270887,
      "grad_norm": 0.13805943727493286,
      "learning_rate": 1.0899347581163221e-05,
      "loss": 1.4615,
      "step": 525
    },
    {
      "epoch": 0.7743835112256164,
      "grad_norm": 0.13799770176410675,
      "learning_rate": 1.0615857584873623e-05,
      "loss": 1.4893,
      "step": 526
    },
    {
      "epoch": 0.7758557232241443,
      "grad_norm": 0.13743866980075836,
      "learning_rate": 1.0335896321476413e-05,
      "loss": 1.4882,
      "step": 527
    },
    {
      "epoch": 0.7773279352226721,
      "grad_norm": 0.135455921292305,
      "learning_rate": 1.0059474843362892e-05,
      "loss": 1.5039,
      "step": 528
    },
    {
      "epoch": 0.7788001472211998,
      "grad_norm": 0.13624714314937592,
      "learning_rate": 9.786604063179728e-06,
      "loss": 1.5252,
      "step": 529
    },
    {
      "epoch": 0.7802723592197276,
      "grad_norm": 0.13567432761192322,
      "learning_rate": 9.517294753398064e-06,
      "loss": 1.4649,
      "step": 530
    },
    {
      "epoch": 0.7817445712182555,
      "grad_norm": 0.1356220543384552,
      "learning_rate": 9.251557545888312e-06,
      "loss": 1.4873,
      "step": 531
    },
    {
      "epoch": 0.7832167832167832,
      "grad_norm": 0.1355055421590805,
      "learning_rate": 8.989402931500434e-06,
      "loss": 1.4994,
      "step": 532
    },
    {
      "epoch": 0.784688995215311,
      "grad_norm": 0.13455788791179657,
      "learning_rate": 8.730841259649725e-06,
      "loss": 1.5005,
      "step": 533
    },
    {
      "epoch": 0.7861612072138388,
      "grad_norm": 0.13485857844352722,
      "learning_rate": 8.475882737908248e-06,
      "loss": 1.482,
      "step": 534
    },
    {
      "epoch": 0.7876334192123666,
      "grad_norm": 0.13655389845371246,
      "learning_rate": 8.224537431601886e-06,
      "loss": 1.4647,
      "step": 535
    },
    {
      "epoch": 0.7891056312108944,
      "grad_norm": 0.13773468136787415,
      "learning_rate": 7.976815263412963e-06,
      "loss": 1.4592,
      "step": 536
    },
    {
      "epoch": 0.7905778432094221,
      "grad_norm": 0.13726471364498138,
      "learning_rate": 7.73272601298851e-06,
      "loss": 1.4813,
      "step": 537
    },
    {
      "epoch": 0.7920500552079499,
      "grad_norm": 0.13571199774742126,
      "learning_rate": 7.492279316554207e-06,
      "loss": 1.494,
      "step": 538
    },
    {
      "epoch": 0.7935222672064778,
      "grad_norm": 0.1306808590888977,
      "learning_rate": 7.255484666533874e-06,
      "loss": 1.505,
      "step": 539
    },
    {
      "epoch": 0.7949944792050055,
      "grad_norm": 0.13337427377700806,
      "learning_rate": 7.022351411174866e-06,
      "loss": 1.4757,
      "step": 540
    },
    {
      "epoch": 0.7964666912035333,
      "grad_norm": 0.12852667272090912,
      "learning_rate": 6.7928887541789055e-06,
      "loss": 1.4721,
      "step": 541
    },
    {
      "epoch": 0.7979389032020611,
      "grad_norm": 0.1298791915178299,
      "learning_rate": 6.5671057543387985e-06,
      "loss": 1.4999,
      "step": 542
    },
    {
      "epoch": 0.7994111152005889,
      "grad_norm": 0.13449601829051971,
      "learning_rate": 6.345011325180772e-06,
      "loss": 1.4624,
      "step": 543
    },
    {
      "epoch": 0.8008833271991167,
      "grad_norm": 0.13329100608825684,
      "learning_rate": 6.126614234612593e-06,
      "loss": 1.4966,
      "step": 544
    },
    {
      "epoch": 0.8023555391976445,
      "grad_norm": 0.1387065052986145,
      "learning_rate": 5.911923104577455e-06,
      "loss": 1.4912,
      "step": 545
    },
    {
      "epoch": 0.8038277511961722,
      "grad_norm": 0.138751819729805,
      "learning_rate": 5.700946410713548e-06,
      "loss": 1.4778,
      "step": 546
    },
    {
      "epoch": 0.8052999631947,
      "grad_norm": 0.1354382187128067,
      "learning_rate": 5.49369248201953e-06,
      "loss": 1.4795,
      "step": 547
    },
    {
      "epoch": 0.8067721751932279,
      "grad_norm": 0.13986830413341522,
      "learning_rate": 5.290169500525577e-06,
      "loss": 1.5086,
      "step": 548
    },
    {
      "epoch": 0.8082443871917556,
      "grad_norm": 0.13272003829479218,
      "learning_rate": 5.0903855009705514e-06,
      "loss": 1.4534,
      "step": 549
    },
    {
      "epoch": 0.8097165991902834,
      "grad_norm": 0.13290782272815704,
      "learning_rate": 4.8943483704846475e-06,
      "loss": 1.4954,
      "step": 550
    },
    {
      "epoch": 0.8111888111888111,
      "grad_norm": 0.13395152986049652,
      "learning_rate": 4.702065848278126e-06,
      "loss": 1.5313,
      "step": 551
    },
    {
      "epoch": 0.812661023187339,
      "grad_norm": 0.1314895749092102,
      "learning_rate": 4.513545525335705e-06,
      "loss": 1.478,
      "step": 552
    },
    {
      "epoch": 0.8141332351858668,
      "grad_norm": 0.13987351953983307,
      "learning_rate": 4.328794844116946e-06,
      "loss": 1.4918,
      "step": 553
    },
    {
      "epoch": 0.8156054471843945,
      "grad_norm": 0.1387564241886139,
      "learning_rate": 4.147821098262405e-06,
      "loss": 1.4904,
      "step": 554
    },
    {
      "epoch": 0.8170776591829223,
      "grad_norm": 0.1383361518383026,
      "learning_rate": 3.970631432305694e-06,
      "loss": 1.5298,
      "step": 555
    },
    {
      "epoch": 0.8185498711814502,
      "grad_norm": 0.13355666399002075,
      "learning_rate": 3.797232841391407e-06,
      "loss": 1.5135,
      "step": 556
    },
    {
      "epoch": 0.8200220831799779,
      "grad_norm": 0.13369423151016235,
      "learning_rate": 3.627632170999029e-06,
      "loss": 1.4622,
      "step": 557
    },
    {
      "epoch": 0.8214942951785057,
      "grad_norm": 0.13357460498809814,
      "learning_rate": 3.461836116672612e-06,
      "loss": 1.4937,
      "step": 558
    },
    {
      "epoch": 0.8229665071770335,
      "grad_norm": 0.13675329089164734,
      "learning_rate": 3.2998512237565005e-06,
      "loss": 1.49,
      "step": 559
    },
    {
      "epoch": 0.8244387191755613,
      "grad_norm": 0.13170136511325836,
      "learning_rate": 3.1416838871368924e-06,
      "loss": 1.4749,
      "step": 560
    },
    {
      "epoch": 0.8259109311740891,
      "grad_norm": 0.13150560855865479,
      "learning_rate": 2.9873403509894203e-06,
      "loss": 1.4605,
      "step": 561
    },
    {
      "epoch": 0.8273831431726169,
      "grad_norm": 0.13412930071353912,
      "learning_rate": 2.836826708532603e-06,
      "loss": 1.4909,
      "step": 562
    },
    {
      "epoch": 0.8288553551711446,
      "grad_norm": 0.1372586339712143,
      "learning_rate": 2.690148901787337e-06,
      "loss": 1.4934,
      "step": 563
    },
    {
      "epoch": 0.8303275671696725,
      "grad_norm": 0.13502952456474304,
      "learning_rate": 2.5473127213422763e-06,
      "loss": 1.4881,
      "step": 564
    },
    {
      "epoch": 0.8317997791682002,
      "grad_norm": 0.13261441886425018,
      "learning_rate": 2.4083238061252567e-06,
      "loss": 1.4736,
      "step": 565
    },
    {
      "epoch": 0.833271991166728,
      "grad_norm": 0.1365126222372055,
      "learning_rate": 2.273187643180652e-06,
      "loss": 1.4694,
      "step": 566
    },
    {
      "epoch": 0.8347442031652558,
      "grad_norm": 0.13771353662014008,
      "learning_rate": 2.141909567452793e-06,
      "loss": 1.4555,
      "step": 567
    },
    {
      "epoch": 0.8362164151637835,
      "grad_norm": 0.13247795403003693,
      "learning_rate": 2.014494761575314e-06,
      "loss": 1.4901,
      "step": 568
    },
    {
      "epoch": 0.8376886271623114,
      "grad_norm": 0.1359633505344391,
      "learning_rate": 1.8909482556666024e-06,
      "loss": 1.4958,
      "step": 569
    },
    {
      "epoch": 0.8391608391608392,
      "grad_norm": 0.13325655460357666,
      "learning_rate": 1.771274927131139e-06,
      "loss": 1.4449,
      "step": 570
    },
    {
      "epoch": 0.8406330511593669,
      "grad_norm": 0.13798603415489197,
      "learning_rate": 1.6554795004670388e-06,
      "loss": 1.4682,
      "step": 571
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.13382163643836975,
      "learning_rate": 1.543566547079467e-06,
      "loss": 1.4885,
      "step": 572
    },
    {
      "epoch": 0.8435774751564226,
      "grad_norm": 0.13109365105628967,
      "learning_rate": 1.4355404851001952e-06,
      "loss": 1.494,
      "step": 573
    },
    {
      "epoch": 0.8450496871549503,
      "grad_norm": 0.13858172297477722,
      "learning_rate": 1.3314055792131964e-06,
      "loss": 1.4757,
      "step": 574
    },
    {
      "epoch": 0.8465218991534781,
      "grad_norm": 0.1343967318534851,
      "learning_rate": 1.231165940486234e-06,
      "loss": 1.4737,
      "step": 575
    },
    {
      "epoch": 0.8479941111520058,
      "grad_norm": 0.13387811183929443,
      "learning_rate": 1.134825526208605e-06,
      "loss": 1.4856,
      "step": 576
    },
    {
      "epoch": 0.8494663231505337,
      "grad_norm": 0.13163875043392181,
      "learning_rate": 1.0423881397349068e-06,
      "loss": 1.447,
      "step": 577
    },
    {
      "epoch": 0.8509385351490615,
      "grad_norm": 0.13378369808197021,
      "learning_rate": 9.538574303348813e-07,
      "loss": 1.507,
      "step": 578
    },
    {
      "epoch": 0.8524107471475892,
      "grad_norm": 0.13452619314193726,
      "learning_rate": 8.692368930493521e-07,
      "loss": 1.482,
      "step": 579
    },
    {
      "epoch": 0.853882959146117,
      "grad_norm": 0.13239046931266785,
      "learning_rate": 7.885298685522235e-07,
      "loss": 1.4941,
      "step": 580
    },
    {
      "epoch": 0.8553551711446449,
      "grad_norm": 0.13270816206932068,
      "learning_rate": 7.117395430186414e-07,
      "loss": 1.4735,
      "step": 581
    },
    {
      "epoch": 0.8568273831431726,
      "grad_norm": 0.13482458889484406,
      "learning_rate": 6.388689479991605e-07,
      "loss": 1.4705,
      "step": 582
    },
    {
      "epoch": 0.8582995951417004,
      "grad_norm": 0.1361265927553177,
      "learning_rate": 5.699209603001076e-07,
      "loss": 1.5051,
      "step": 583
    },
    {
      "epoch": 0.8597718071402282,
      "grad_norm": 0.13906005024909973,
      "learning_rate": 5.048983018699827e-07,
      "loss": 1.4908,
      "step": 584
    },
    {
      "epoch": 0.861244019138756,
      "grad_norm": 0.1362522542476654,
      "learning_rate": 4.438035396920004e-07,
      "loss": 1.5123,
      "step": 585
    },
    {
      "epoch": 0.8627162311372838,
      "grad_norm": 0.13366208970546722,
      "learning_rate": 3.866390856827495e-07,
      "loss": 1.5009,
      "step": 586
    },
    {
      "epoch": 0.8641884431358116,
      "grad_norm": 0.1322133094072342,
      "learning_rate": 3.3340719659701313e-07,
      "loss": 1.5068,
      "step": 587
    },
    {
      "epoch": 0.8656606551343393,
      "grad_norm": 0.13434545695781708,
      "learning_rate": 2.841099739386066e-07,
      "loss": 1.4791,
      "step": 588
    },
    {
      "epoch": 0.8671328671328671,
      "grad_norm": 0.13443467020988464,
      "learning_rate": 2.387493638774774e-07,
      "loss": 1.5091,
      "step": 589
    },
    {
      "epoch": 0.868605079131395,
      "grad_norm": 0.1373066008090973,
      "learning_rate": 1.973271571728441e-07,
      "loss": 1.4863,
      "step": 590
    },
    {
      "epoch": 0.8700772911299227,
      "grad_norm": 0.13269129395484924,
      "learning_rate": 1.598449891024978e-07,
      "loss": 1.4988,
      "step": 591
    },
    {
      "epoch": 0.8715495031284505,
      "grad_norm": 0.13867610692977905,
      "learning_rate": 1.2630433939825327e-07,
      "loss": 1.4899,
      "step": 592
    },
    {
      "epoch": 0.8730217151269782,
      "grad_norm": 0.1365261971950531,
      "learning_rate": 9.670653218752934e-08,
      "loss": 1.4736,
      "step": 593
    },
    {
      "epoch": 0.8744939271255061,
      "grad_norm": 0.1346593052148819,
      "learning_rate": 7.105273594107953e-08,
      "loss": 1.5087,
      "step": 594
    },
    {
      "epoch": 0.8759661391240339,
      "grad_norm": 0.1361488550901413,
      "learning_rate": 4.934396342684e-08,
      "loss": 1.5038,
      "step": 595
    },
    {
      "epoch": 0.8774383511225616,
      "grad_norm": 0.13418179750442505,
      "learning_rate": 3.1581071670006015e-08,
      "loss": 1.5087,
      "step": 596
    },
    {
      "epoch": 0.8789105631210894,
      "grad_norm": 0.1340452879667282,
      "learning_rate": 1.7764761919103477e-08,
      "loss": 1.522,
      "step": 597
    },
    {
      "epoch": 0.8803827751196173,
      "grad_norm": 0.13247017562389374,
      "learning_rate": 7.895579618388827e-09,
      "loss": 1.4889,
      "step": 598
    },
    {
      "epoch": 0.881854987118145,
      "grad_norm": 0.13550199568271637,
      "learning_rate": 1.973914386288467e-09,
      "loss": 1.5147,
      "step": 599
    },
    {
      "epoch": 0.8833271991166728,
      "grad_norm": 0.13783395290374756,
      "learning_rate": 0.0,
      "loss": 1.4738,
      "step": 600
    },
    {
      "epoch": 0.8833271991166728,
      "eval_loss": 1.4882882833480835,
      "eval_runtime": 44.6471,
      "eval_samples_per_second": 102.515,
      "eval_steps_per_second": 3.225,
      "step": 600
    }
  ],
  "logging_steps": 1,
  "max_steps": 600,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3380424723029033e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}