gyung's picture
Upload Gemma 4 31B Terminal SFT epoch 1 checkpoint
b9265c5 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1467,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006816632583503749,
"grad_norm": 4383.8388671875,
"learning_rate": 0.0,
"loss": 6.527473449707031,
"step": 1
},
{
"epoch": 0.0013633265167007499,
"grad_norm": 1823.0880126953125,
"learning_rate": 2.247191011235955e-07,
"loss": 7.508894443511963,
"step": 2
},
{
"epoch": 0.002044989775051125,
"grad_norm": 40114.21875,
"learning_rate": 4.49438202247191e-07,
"loss": 7.247808456420898,
"step": 3
},
{
"epoch": 0.0027266530334014998,
"grad_norm": 575.8884887695312,
"learning_rate": 6.741573033707865e-07,
"loss": 6.052981376647949,
"step": 4
},
{
"epoch": 0.0034083162917518746,
"grad_norm": 76116.6796875,
"learning_rate": 8.98876404494382e-07,
"loss": 5.867002964019775,
"step": 5
},
{
"epoch": 0.00408997955010225,
"grad_norm": 4967.98291015625,
"learning_rate": 1.1235955056179777e-06,
"loss": 6.827726364135742,
"step": 6
},
{
"epoch": 0.004771642808452625,
"grad_norm": 22014.96484375,
"learning_rate": 1.348314606741573e-06,
"loss": 7.764345169067383,
"step": 7
},
{
"epoch": 0.0054533060668029995,
"grad_norm": 2535.458251953125,
"learning_rate": 1.5730337078651686e-06,
"loss": 6.417271137237549,
"step": 8
},
{
"epoch": 0.006134969325153374,
"grad_norm": 4581.03662109375,
"learning_rate": 1.797752808988764e-06,
"loss": 7.255514144897461,
"step": 9
},
{
"epoch": 0.006816632583503749,
"grad_norm": 4421.4501953125,
"learning_rate": 2.02247191011236e-06,
"loss": 4.907093048095703,
"step": 10
},
{
"epoch": 0.007498295841854124,
"grad_norm": 9041.5,
"learning_rate": 2.2471910112359554e-06,
"loss": 4.617020606994629,
"step": 11
},
{
"epoch": 0.0081799591002045,
"grad_norm": 2413.45556640625,
"learning_rate": 2.4719101123595505e-06,
"loss": 5.7079010009765625,
"step": 12
},
{
"epoch": 0.008861622358554875,
"grad_norm": 1334.906982421875,
"learning_rate": 2.696629213483146e-06,
"loss": 3.1134910583496094,
"step": 13
},
{
"epoch": 0.00954328561690525,
"grad_norm": 2017.1741943359375,
"learning_rate": 2.9213483146067416e-06,
"loss": 4.945159912109375,
"step": 14
},
{
"epoch": 0.010224948875255624,
"grad_norm": 1304.0582275390625,
"learning_rate": 3.146067415730337e-06,
"loss": 4.624004364013672,
"step": 15
},
{
"epoch": 0.010906612133605999,
"grad_norm": 3951.13134765625,
"learning_rate": 3.3707865168539327e-06,
"loss": 3.8251986503601074,
"step": 16
},
{
"epoch": 0.011588275391956374,
"grad_norm": 906.8524780273438,
"learning_rate": 3.595505617977528e-06,
"loss": 3.313896656036377,
"step": 17
},
{
"epoch": 0.012269938650306749,
"grad_norm": 1372.37353515625,
"learning_rate": 3.820224719101124e-06,
"loss": 3.498051404953003,
"step": 18
},
{
"epoch": 0.012951601908657124,
"grad_norm": 494.13616943359375,
"learning_rate": 4.04494382022472e-06,
"loss": 3.268960952758789,
"step": 19
},
{
"epoch": 0.013633265167007498,
"grad_norm": 282.08331298828125,
"learning_rate": 4.269662921348315e-06,
"loss": 2.357489585876465,
"step": 20
},
{
"epoch": 0.014314928425357873,
"grad_norm": 3040.841064453125,
"learning_rate": 4.494382022471911e-06,
"loss": 3.6005187034606934,
"step": 21
},
{
"epoch": 0.014996591683708248,
"grad_norm": 1250.9658203125,
"learning_rate": 4.719101123595506e-06,
"loss": 3.457188129425049,
"step": 22
},
{
"epoch": 0.015678254942058625,
"grad_norm": 5198.23974609375,
"learning_rate": 4.943820224719101e-06,
"loss": 3.725776433944702,
"step": 23
},
{
"epoch": 0.016359918200409,
"grad_norm": 454.5555114746094,
"learning_rate": 5.168539325842698e-06,
"loss": 2.6974992752075195,
"step": 24
},
{
"epoch": 0.017041581458759374,
"grad_norm": 1461.07763671875,
"learning_rate": 5.393258426966292e-06,
"loss": 2.363149642944336,
"step": 25
},
{
"epoch": 0.01772324471710975,
"grad_norm": 223.5943145751953,
"learning_rate": 5.617977528089889e-06,
"loss": 1.918904185295105,
"step": 26
},
{
"epoch": 0.018404907975460124,
"grad_norm": 819.2132568359375,
"learning_rate": 5.842696629213483e-06,
"loss": 2.747713565826416,
"step": 27
},
{
"epoch": 0.0190865712338105,
"grad_norm": 329.3885803222656,
"learning_rate": 6.06741573033708e-06,
"loss": 2.5517754554748535,
"step": 28
},
{
"epoch": 0.019768234492160874,
"grad_norm": 578.25927734375,
"learning_rate": 6.292134831460674e-06,
"loss": 3.0694191455841064,
"step": 29
},
{
"epoch": 0.02044989775051125,
"grad_norm": 521.046142578125,
"learning_rate": 6.51685393258427e-06,
"loss": 2.9562296867370605,
"step": 30
},
{
"epoch": 0.021131561008861623,
"grad_norm": 479.71722412109375,
"learning_rate": 6.741573033707865e-06,
"loss": 3.020545721054077,
"step": 31
},
{
"epoch": 0.021813224267211998,
"grad_norm": 734.2029418945312,
"learning_rate": 6.966292134831461e-06,
"loss": 3.466350555419922,
"step": 32
},
{
"epoch": 0.022494887525562373,
"grad_norm": 254.68246459960938,
"learning_rate": 7.191011235955056e-06,
"loss": 3.2190747261047363,
"step": 33
},
{
"epoch": 0.023176550783912748,
"grad_norm": 130.5963592529297,
"learning_rate": 7.415730337078652e-06,
"loss": 2.0983405113220215,
"step": 34
},
{
"epoch": 0.023858214042263123,
"grad_norm": 73.80171966552734,
"learning_rate": 7.640449438202247e-06,
"loss": 1.889816164970398,
"step": 35
},
{
"epoch": 0.024539877300613498,
"grad_norm": 56.68690490722656,
"learning_rate": 7.865168539325843e-06,
"loss": 3.1251683235168457,
"step": 36
},
{
"epoch": 0.025221540558963872,
"grad_norm": 97.23187255859375,
"learning_rate": 8.08988764044944e-06,
"loss": 3.2829296588897705,
"step": 37
},
{
"epoch": 0.025903203817314247,
"grad_norm": 135.62997436523438,
"learning_rate": 8.314606741573035e-06,
"loss": 2.9910526275634766,
"step": 38
},
{
"epoch": 0.026584867075664622,
"grad_norm": 77.11801147460938,
"learning_rate": 8.53932584269663e-06,
"loss": 3.5667967796325684,
"step": 39
},
{
"epoch": 0.027266530334014997,
"grad_norm": 141.92796325683594,
"learning_rate": 8.764044943820226e-06,
"loss": 1.9355766773223877,
"step": 40
},
{
"epoch": 0.02794819359236537,
"grad_norm": 49.39057159423828,
"learning_rate": 8.988764044943822e-06,
"loss": 1.2001816034317017,
"step": 41
},
{
"epoch": 0.028629856850715747,
"grad_norm": 78.03803253173828,
"learning_rate": 9.213483146067417e-06,
"loss": 2.352545976638794,
"step": 42
},
{
"epoch": 0.02931152010906612,
"grad_norm": 26.599884033203125,
"learning_rate": 9.438202247191012e-06,
"loss": 2.0077102184295654,
"step": 43
},
{
"epoch": 0.029993183367416496,
"grad_norm": 49.84416961669922,
"learning_rate": 9.662921348314608e-06,
"loss": 2.672440528869629,
"step": 44
},
{
"epoch": 0.03067484662576687,
"grad_norm": 35.74471664428711,
"learning_rate": 9.887640449438202e-06,
"loss": 1.4557160139083862,
"step": 45
},
{
"epoch": 0.03135650988411725,
"grad_norm": 29.79218864440918,
"learning_rate": 1.01123595505618e-05,
"loss": 2.77484393119812,
"step": 46
},
{
"epoch": 0.032038173142467624,
"grad_norm": 242.94088745117188,
"learning_rate": 1.0337078651685396e-05,
"loss": 1.2804923057556152,
"step": 47
},
{
"epoch": 0.032719836400818,
"grad_norm": 33.16373825073242,
"learning_rate": 1.0561797752808988e-05,
"loss": 1.6744909286499023,
"step": 48
},
{
"epoch": 0.033401499659168374,
"grad_norm": 47.615360260009766,
"learning_rate": 1.0786516853932584e-05,
"loss": 2.003718376159668,
"step": 49
},
{
"epoch": 0.03408316291751875,
"grad_norm": 37.541908264160156,
"learning_rate": 1.101123595505618e-05,
"loss": 2.387068271636963,
"step": 50
},
{
"epoch": 0.034764826175869123,
"grad_norm": 29.1586971282959,
"learning_rate": 1.1235955056179778e-05,
"loss": 2.4086151123046875,
"step": 51
},
{
"epoch": 0.0354464894342195,
"grad_norm": 16.95186996459961,
"learning_rate": 1.146067415730337e-05,
"loss": 1.283890962600708,
"step": 52
},
{
"epoch": 0.03612815269256987,
"grad_norm": 16.21533203125,
"learning_rate": 1.1685393258426966e-05,
"loss": 1.3941848278045654,
"step": 53
},
{
"epoch": 0.03680981595092025,
"grad_norm": 15.245625495910645,
"learning_rate": 1.1910112359550562e-05,
"loss": 1.799220323562622,
"step": 54
},
{
"epoch": 0.03749147920927062,
"grad_norm": 1592.2939453125,
"learning_rate": 1.213483146067416e-05,
"loss": 2.0396103858947754,
"step": 55
},
{
"epoch": 0.038173142467621,
"grad_norm": 19.32965087890625,
"learning_rate": 1.2359550561797752e-05,
"loss": 1.3870739936828613,
"step": 56
},
{
"epoch": 0.03885480572597137,
"grad_norm": 55.4034423828125,
"learning_rate": 1.2584269662921348e-05,
"loss": 2.1296072006225586,
"step": 57
},
{
"epoch": 0.03953646898432175,
"grad_norm": 82.31266784667969,
"learning_rate": 1.2808988764044944e-05,
"loss": 0.8768513798713684,
"step": 58
},
{
"epoch": 0.04021813224267212,
"grad_norm": 9.533697128295898,
"learning_rate": 1.303370786516854e-05,
"loss": 2.0535597801208496,
"step": 59
},
{
"epoch": 0.0408997955010225,
"grad_norm": 15.587124824523926,
"learning_rate": 1.3258426966292135e-05,
"loss": 1.9851542711257935,
"step": 60
},
{
"epoch": 0.04158145875937287,
"grad_norm": 30.805431365966797,
"learning_rate": 1.348314606741573e-05,
"loss": 1.4596028327941895,
"step": 61
},
{
"epoch": 0.04226312201772325,
"grad_norm": 9.697044372558594,
"learning_rate": 1.3707865168539327e-05,
"loss": 1.6669340133666992,
"step": 62
},
{
"epoch": 0.04294478527607362,
"grad_norm": 27.938608169555664,
"learning_rate": 1.3932584269662923e-05,
"loss": 0.9620833396911621,
"step": 63
},
{
"epoch": 0.043626448534423996,
"grad_norm": 7.904890060424805,
"learning_rate": 1.4157303370786517e-05,
"loss": 1.2511191368103027,
"step": 64
},
{
"epoch": 0.04430811179277437,
"grad_norm": 6.319112777709961,
"learning_rate": 1.4382022471910113e-05,
"loss": 0.6878035068511963,
"step": 65
},
{
"epoch": 0.044989775051124746,
"grad_norm": 22.30373191833496,
"learning_rate": 1.4606741573033709e-05,
"loss": 2.279068946838379,
"step": 66
},
{
"epoch": 0.04567143830947512,
"grad_norm": 12.7396879196167,
"learning_rate": 1.4831460674157305e-05,
"loss": 1.6731181144714355,
"step": 67
},
{
"epoch": 0.046353101567825496,
"grad_norm": 20.310932159423828,
"learning_rate": 1.5056179775280899e-05,
"loss": 0.8521436452865601,
"step": 68
},
{
"epoch": 0.04703476482617587,
"grad_norm": 13.107906341552734,
"learning_rate": 1.5280898876404495e-05,
"loss": 1.511030912399292,
"step": 69
},
{
"epoch": 0.047716428084526245,
"grad_norm": 63.82345962524414,
"learning_rate": 1.5505617977528093e-05,
"loss": 1.378786563873291,
"step": 70
},
{
"epoch": 0.04839809134287662,
"grad_norm": 19.864704132080078,
"learning_rate": 1.5730337078651687e-05,
"loss": 1.5483001470565796,
"step": 71
},
{
"epoch": 0.049079754601226995,
"grad_norm": 16.860614776611328,
"learning_rate": 1.595505617977528e-05,
"loss": 1.5748002529144287,
"step": 72
},
{
"epoch": 0.04976141785957737,
"grad_norm": 35.4315185546875,
"learning_rate": 1.617977528089888e-05,
"loss": 1.4997893571853638,
"step": 73
},
{
"epoch": 0.050443081117927745,
"grad_norm": 39.20589065551758,
"learning_rate": 1.6404494382022473e-05,
"loss": 2.2233829498291016,
"step": 74
},
{
"epoch": 0.05112474437627812,
"grad_norm": 1053.0841064453125,
"learning_rate": 1.662921348314607e-05,
"loss": 2.754089117050171,
"step": 75
},
{
"epoch": 0.051806407634628494,
"grad_norm": 13.371782302856445,
"learning_rate": 1.6853932584269665e-05,
"loss": 0.9811263680458069,
"step": 76
},
{
"epoch": 0.05248807089297887,
"grad_norm": 10.675305366516113,
"learning_rate": 1.707865168539326e-05,
"loss": 1.7070579528808594,
"step": 77
},
{
"epoch": 0.053169734151329244,
"grad_norm": 14.619195938110352,
"learning_rate": 1.7303370786516857e-05,
"loss": 1.7202574014663696,
"step": 78
},
{
"epoch": 0.05385139740967962,
"grad_norm": 3042.486083984375,
"learning_rate": 1.752808988764045e-05,
"loss": 1.9486947059631348,
"step": 79
},
{
"epoch": 0.054533060668029994,
"grad_norm": 55.10081481933594,
"learning_rate": 1.7752808988764045e-05,
"loss": 1.7752485275268555,
"step": 80
},
{
"epoch": 0.05521472392638037,
"grad_norm": 19.896421432495117,
"learning_rate": 1.7977528089887643e-05,
"loss": 2.0431885719299316,
"step": 81
},
{
"epoch": 0.05589638718473074,
"grad_norm": 20.129390716552734,
"learning_rate": 1.8202247191011237e-05,
"loss": 1.767225980758667,
"step": 82
},
{
"epoch": 0.05657805044308112,
"grad_norm": 68.15953826904297,
"learning_rate": 1.8426966292134835e-05,
"loss": 1.4359931945800781,
"step": 83
},
{
"epoch": 0.05725971370143149,
"grad_norm": 33.0981330871582,
"learning_rate": 1.8651685393258426e-05,
"loss": 0.9191545248031616,
"step": 84
},
{
"epoch": 0.05794137695978187,
"grad_norm": 17.07698631286621,
"learning_rate": 1.8876404494382024e-05,
"loss": 1.7767210006713867,
"step": 85
},
{
"epoch": 0.05862304021813224,
"grad_norm": 27.136503219604492,
"learning_rate": 1.910112359550562e-05,
"loss": 1.3165686130523682,
"step": 86
},
{
"epoch": 0.05930470347648262,
"grad_norm": 234.57733154296875,
"learning_rate": 1.9325842696629215e-05,
"loss": 1.5632712841033936,
"step": 87
},
{
"epoch": 0.05998636673483299,
"grad_norm": 164.1422882080078,
"learning_rate": 1.955056179775281e-05,
"loss": 1.1874092817306519,
"step": 88
},
{
"epoch": 0.06066802999318337,
"grad_norm": 34.60917663574219,
"learning_rate": 1.9775280898876404e-05,
"loss": 0.8331537842750549,
"step": 89
},
{
"epoch": 0.06134969325153374,
"grad_norm": 10.945159912109375,
"learning_rate": 2e-05,
"loss": 1.2834336757659912,
"step": 90
},
{
"epoch": 0.06203135650988412,
"grad_norm": 15.15926742553711,
"learning_rate": 1.9999993903154863e-05,
"loss": 1.7447407245635986,
"step": 91
},
{
"epoch": 0.0627130197682345,
"grad_norm": 144.3293914794922,
"learning_rate": 1.9999975612626872e-05,
"loss": 1.3708195686340332,
"step": 92
},
{
"epoch": 0.06339468302658487,
"grad_norm": 25.28166961669922,
"learning_rate": 1.9999945128438338e-05,
"loss": 2.8497724533081055,
"step": 93
},
{
"epoch": 0.06407634628493525,
"grad_norm": 19.966693878173828,
"learning_rate": 1.9999902450626434e-05,
"loss": 1.6539599895477295,
"step": 94
},
{
"epoch": 0.06475800954328562,
"grad_norm": 15.393426895141602,
"learning_rate": 1.9999847579243196e-05,
"loss": 2.0296273231506348,
"step": 95
},
{
"epoch": 0.065439672801636,
"grad_norm": 11.24446964263916,
"learning_rate": 1.9999780514355533e-05,
"loss": 1.3424549102783203,
"step": 96
},
{
"epoch": 0.06612133605998637,
"grad_norm": 27.612884521484375,
"learning_rate": 1.9999701256045223e-05,
"loss": 1.194828987121582,
"step": 97
},
{
"epoch": 0.06680299931833675,
"grad_norm": 12.825202941894531,
"learning_rate": 1.999960980440891e-05,
"loss": 0.8488439321517944,
"step": 98
},
{
"epoch": 0.06748466257668712,
"grad_norm": 1464.5302734375,
"learning_rate": 1.9999506159558107e-05,
"loss": 0.9869670867919922,
"step": 99
},
{
"epoch": 0.0681663258350375,
"grad_norm": 111.20191955566406,
"learning_rate": 1.9999390321619196e-05,
"loss": 1.3429677486419678,
"step": 100
},
{
"epoch": 0.06884798909338787,
"grad_norm": 16.313430786132812,
"learning_rate": 1.9999262290733427e-05,
"loss": 1.9419188499450684,
"step": 101
},
{
"epoch": 0.06952965235173825,
"grad_norm": 9.3142728805542,
"learning_rate": 1.9999122067056915e-05,
"loss": 1.5300267934799194,
"step": 102
},
{
"epoch": 0.07021131561008861,
"grad_norm": 22.08498191833496,
"learning_rate": 1.9998969650760646e-05,
"loss": 1.9083926677703857,
"step": 103
},
{
"epoch": 0.070892978868439,
"grad_norm": 15.76110553741455,
"learning_rate": 1.9998805042030472e-05,
"loss": 1.6124863624572754,
"step": 104
},
{
"epoch": 0.07157464212678936,
"grad_norm": 30.93997573852539,
"learning_rate": 1.9998628241067113e-05,
"loss": 0.9653267860412598,
"step": 105
},
{
"epoch": 0.07225630538513975,
"grad_norm": 12.660534858703613,
"learning_rate": 1.999843924808615e-05,
"loss": 1.7511203289031982,
"step": 106
},
{
"epoch": 0.07293796864349011,
"grad_norm": 18.625757217407227,
"learning_rate": 1.999823806331804e-05,
"loss": 1.297489881515503,
"step": 107
},
{
"epoch": 0.0736196319018405,
"grad_norm": 56.3899040222168,
"learning_rate": 1.9998024687008098e-05,
"loss": 1.449361801147461,
"step": 108
},
{
"epoch": 0.07430129516019086,
"grad_norm": 162.703857421875,
"learning_rate": 1.9997799119416508e-05,
"loss": 1.7976202964782715,
"step": 109
},
{
"epoch": 0.07498295841854125,
"grad_norm": 9.930502891540527,
"learning_rate": 1.9997561360818322e-05,
"loss": 1.5131696462631226,
"step": 110
},
{
"epoch": 0.07566462167689161,
"grad_norm": 31.915830612182617,
"learning_rate": 1.999731141150346e-05,
"loss": 1.3789496421813965,
"step": 111
},
{
"epoch": 0.076346284935242,
"grad_norm": 11.755589485168457,
"learning_rate": 1.9997049271776693e-05,
"loss": 1.747565746307373,
"step": 112
},
{
"epoch": 0.07702794819359236,
"grad_norm": 31.262149810791016,
"learning_rate": 1.9996774941957673e-05,
"loss": 1.4239745140075684,
"step": 113
},
{
"epoch": 0.07770961145194274,
"grad_norm": 63.65169143676758,
"learning_rate": 1.999648842238091e-05,
"loss": 1.815131425857544,
"step": 114
},
{
"epoch": 0.07839127471029311,
"grad_norm": 18.11605453491211,
"learning_rate": 1.999618971339577e-05,
"loss": 1.7555679082870483,
"step": 115
},
{
"epoch": 0.0790729379686435,
"grad_norm": 160.328369140625,
"learning_rate": 1.9995878815366498e-05,
"loss": 1.8470728397369385,
"step": 116
},
{
"epoch": 0.07975460122699386,
"grad_norm": 619.9039916992188,
"learning_rate": 1.999555572867218e-05,
"loss": 1.7569350004196167,
"step": 117
},
{
"epoch": 0.08043626448534424,
"grad_norm": 19.946969985961914,
"learning_rate": 1.9995220453706797e-05,
"loss": 1.3383402824401855,
"step": 118
},
{
"epoch": 0.08111792774369461,
"grad_norm": 20.062156677246094,
"learning_rate": 1.9994872990879163e-05,
"loss": 1.3586251735687256,
"step": 119
},
{
"epoch": 0.081799591002045,
"grad_norm": 19.174461364746094,
"learning_rate": 1.9994513340612957e-05,
"loss": 1.8789781332015991,
"step": 120
},
{
"epoch": 0.08248125426039536,
"grad_norm": 14.19649887084961,
"learning_rate": 1.9994141503346735e-05,
"loss": 1.6669260263442993,
"step": 121
},
{
"epoch": 0.08316291751874574,
"grad_norm": 182.03321838378906,
"learning_rate": 1.99937574795339e-05,
"loss": 2.1706442832946777,
"step": 122
},
{
"epoch": 0.08384458077709611,
"grad_norm": 25.119983673095703,
"learning_rate": 1.999336126964272e-05,
"loss": 1.6401561498641968,
"step": 123
},
{
"epoch": 0.0845262440354465,
"grad_norm": 11.802164077758789,
"learning_rate": 1.9992952874156323e-05,
"loss": 2.053577184677124,
"step": 124
},
{
"epoch": 0.08520790729379686,
"grad_norm": 53.97670364379883,
"learning_rate": 1.9992532293572688e-05,
"loss": 0.6542696952819824,
"step": 125
},
{
"epoch": 0.08588957055214724,
"grad_norm": 23.286149978637695,
"learning_rate": 1.9992099528404664e-05,
"loss": 1.7017326354980469,
"step": 126
},
{
"epoch": 0.08657123381049761,
"grad_norm": 17.239919662475586,
"learning_rate": 1.9991654579179948e-05,
"loss": 1.2679834365844727,
"step": 127
},
{
"epoch": 0.08725289706884799,
"grad_norm": 16.885732650756836,
"learning_rate": 1.9991197446441096e-05,
"loss": 1.4786620140075684,
"step": 128
},
{
"epoch": 0.08793456032719836,
"grad_norm": 23.140071868896484,
"learning_rate": 1.9990728130745524e-05,
"loss": 1.5406410694122314,
"step": 129
},
{
"epoch": 0.08861622358554874,
"grad_norm": 14.451329231262207,
"learning_rate": 1.9990246632665503e-05,
"loss": 1.145501732826233,
"step": 130
},
{
"epoch": 0.08929788684389911,
"grad_norm": 10.32580852508545,
"learning_rate": 1.998975295278815e-05,
"loss": 1.1438711881637573,
"step": 131
},
{
"epoch": 0.08997955010224949,
"grad_norm": 10.931632041931152,
"learning_rate": 1.9989247091715454e-05,
"loss": 2.1025326251983643,
"step": 132
},
{
"epoch": 0.09066121336059986,
"grad_norm": 6.50631856918335,
"learning_rate": 1.998872905006423e-05,
"loss": 1.4954837560653687,
"step": 133
},
{
"epoch": 0.09134287661895024,
"grad_norm": 4.921420574188232,
"learning_rate": 1.9988198828466182e-05,
"loss": 1.5103764533996582,
"step": 134
},
{
"epoch": 0.09202453987730061,
"grad_norm": 5.854800224304199,
"learning_rate": 1.998765642756783e-05,
"loss": 1.2965787649154663,
"step": 135
},
{
"epoch": 0.09270620313565099,
"grad_norm": 9.764752388000488,
"learning_rate": 1.9987101848030566e-05,
"loss": 1.9255067110061646,
"step": 136
},
{
"epoch": 0.09338786639400136,
"grad_norm": 7.5960798263549805,
"learning_rate": 1.998653509053063e-05,
"loss": 1.791310429573059,
"step": 137
},
{
"epoch": 0.09406952965235174,
"grad_norm": 6.565906524658203,
"learning_rate": 1.9985956155759104e-05,
"loss": 1.3865487575531006,
"step": 138
},
{
"epoch": 0.09475119291070211,
"grad_norm": 14.060343742370605,
"learning_rate": 1.9985365044421926e-05,
"loss": 2.089996814727783,
"step": 139
},
{
"epoch": 0.09543285616905249,
"grad_norm": 7.5861687660217285,
"learning_rate": 1.9984761757239878e-05,
"loss": 0.8783048391342163,
"step": 140
},
{
"epoch": 0.09611451942740286,
"grad_norm": 7.125285625457764,
"learning_rate": 1.9984146294948585e-05,
"loss": 1.721929669380188,
"step": 141
},
{
"epoch": 0.09679618268575324,
"grad_norm": 7.363461971282959,
"learning_rate": 1.998351865829853e-05,
"loss": 0.9673935770988464,
"step": 142
},
{
"epoch": 0.09747784594410361,
"grad_norm": 4.273559093475342,
"learning_rate": 1.9982878848055036e-05,
"loss": 1.318180799484253,
"step": 143
},
{
"epoch": 0.09815950920245399,
"grad_norm": 6.188460350036621,
"learning_rate": 1.9982226864998256e-05,
"loss": 1.5355404615402222,
"step": 144
},
{
"epoch": 0.09884117246080436,
"grad_norm": 3.5846171379089355,
"learning_rate": 1.998156270992321e-05,
"loss": 1.4460152387619019,
"step": 145
},
{
"epoch": 0.09952283571915474,
"grad_norm": 4.710892200469971,
"learning_rate": 1.998088638363974e-05,
"loss": 1.4085346460342407,
"step": 146
},
{
"epoch": 0.10020449897750511,
"grad_norm": 11.747404098510742,
"learning_rate": 1.998019788697254e-05,
"loss": 1.7807292938232422,
"step": 147
},
{
"epoch": 0.10088616223585549,
"grad_norm": 3.7539148330688477,
"learning_rate": 1.9979497220761142e-05,
"loss": 1.1747300624847412,
"step": 148
},
{
"epoch": 0.10156782549420586,
"grad_norm": 6.504991054534912,
"learning_rate": 1.997878438585992e-05,
"loss": 1.5703426599502563,
"step": 149
},
{
"epoch": 0.10224948875255624,
"grad_norm": 5.009849548339844,
"learning_rate": 1.9978059383138073e-05,
"loss": 0.9697821736335754,
"step": 150
},
{
"epoch": 0.1029311520109066,
"grad_norm": 3.9481139183044434,
"learning_rate": 1.9977322213479655e-05,
"loss": 1.7657496929168701,
"step": 151
},
{
"epoch": 0.10361281526925699,
"grad_norm": 5.673348903656006,
"learning_rate": 1.9976572877783548e-05,
"loss": 0.8409373164176941,
"step": 152
},
{
"epoch": 0.10429447852760736,
"grad_norm": 3.3935084342956543,
"learning_rate": 1.9975811376963464e-05,
"loss": 1.313270926475525,
"step": 153
},
{
"epoch": 0.10497614178595774,
"grad_norm": 4.8457465171813965,
"learning_rate": 1.997503771194796e-05,
"loss": 1.44279146194458,
"step": 154
},
{
"epoch": 0.1056578050443081,
"grad_norm": 5.6174116134643555,
"learning_rate": 1.997425188368041e-05,
"loss": 1.989031434059143,
"step": 155
},
{
"epoch": 0.10633946830265849,
"grad_norm": 13.403592109680176,
"learning_rate": 1.9973453893119033e-05,
"loss": 0.7220730185508728,
"step": 156
},
{
"epoch": 0.10702113156100886,
"grad_norm": 4.496342658996582,
"learning_rate": 1.9972643741236882e-05,
"loss": 1.0831682682037354,
"step": 157
},
{
"epoch": 0.10770279481935924,
"grad_norm": 3.356137752532959,
"learning_rate": 1.9971821429021817e-05,
"loss": 1.2503924369812012,
"step": 158
},
{
"epoch": 0.1083844580777096,
"grad_norm": 3.4993855953216553,
"learning_rate": 1.997098695747655e-05,
"loss": 0.8627510070800781,
"step": 159
},
{
"epoch": 0.10906612133605999,
"grad_norm": 11.131978988647461,
"learning_rate": 1.9970140327618612e-05,
"loss": 2.386359214782715,
"step": 160
},
{
"epoch": 0.10974778459441036,
"grad_norm": 4.573759078979492,
"learning_rate": 1.9969281540480346e-05,
"loss": 1.5444923639297485,
"step": 161
},
{
"epoch": 0.11042944785276074,
"grad_norm": 7.6408538818359375,
"learning_rate": 1.9968410597108935e-05,
"loss": 1.550155520439148,
"step": 162
},
{
"epoch": 0.1111111111111111,
"grad_norm": 9.762516975402832,
"learning_rate": 1.9967527498566387e-05,
"loss": 1.9585456848144531,
"step": 163
},
{
"epoch": 0.11179277436946149,
"grad_norm": 9.570932388305664,
"learning_rate": 1.9966632245929515e-05,
"loss": 1.7024056911468506,
"step": 164
},
{
"epoch": 0.11247443762781185,
"grad_norm": 556.8116455078125,
"learning_rate": 1.9965724840289972e-05,
"loss": 0.9666114449501038,
"step": 165
},
{
"epoch": 0.11315610088616224,
"grad_norm": 11.168703079223633,
"learning_rate": 1.996480528275421e-05,
"loss": 1.1300976276397705,
"step": 166
},
{
"epoch": 0.1138377641445126,
"grad_norm": 17.5162410736084,
"learning_rate": 1.996387357444352e-05,
"loss": 1.6300952434539795,
"step": 167
},
{
"epoch": 0.11451942740286299,
"grad_norm": 9.438365936279297,
"learning_rate": 1.9962929716493987e-05,
"loss": 1.9170794486999512,
"step": 168
},
{
"epoch": 0.11520109066121335,
"grad_norm": 4.525749683380127,
"learning_rate": 1.9961973710056535e-05,
"loss": 0.7288682460784912,
"step": 169
},
{
"epoch": 0.11588275391956374,
"grad_norm": 9.106972694396973,
"learning_rate": 1.9961005556296875e-05,
"loss": 1.1762197017669678,
"step": 170
},
{
"epoch": 0.1165644171779141,
"grad_norm": 5.086156845092773,
"learning_rate": 1.9960025256395556e-05,
"loss": 0.9640843868255615,
"step": 171
},
{
"epoch": 0.11724608043626449,
"grad_norm": 5.4781012535095215,
"learning_rate": 1.9959032811547912e-05,
"loss": 1.398245096206665,
"step": 172
},
{
"epoch": 0.11792774369461487,
"grad_norm": 8.261008262634277,
"learning_rate": 1.9958028222964114e-05,
"loss": 1.9427658319473267,
"step": 173
},
{
"epoch": 0.11860940695296524,
"grad_norm": 10.69827651977539,
"learning_rate": 1.9957011491869118e-05,
"loss": 2.783135414123535,
"step": 174
},
{
"epoch": 0.11929107021131562,
"grad_norm": 14.744508743286133,
"learning_rate": 1.9955982619502693e-05,
"loss": 1.108130931854248,
"step": 175
},
{
"epoch": 0.11997273346966598,
"grad_norm": 12.835195541381836,
"learning_rate": 1.995494160711942e-05,
"loss": 1.1446030139923096,
"step": 176
},
{
"epoch": 0.12065439672801637,
"grad_norm": 19.859210968017578,
"learning_rate": 1.9953888455988674e-05,
"loss": 1.5773301124572754,
"step": 177
},
{
"epoch": 0.12133605998636673,
"grad_norm": 8.095100402832031,
"learning_rate": 1.995282316739463e-05,
"loss": 1.3076493740081787,
"step": 178
},
{
"epoch": 0.12201772324471712,
"grad_norm": 53.70732498168945,
"learning_rate": 1.995174574263628e-05,
"loss": 1.727512240409851,
"step": 179
},
{
"epoch": 0.12269938650306748,
"grad_norm": 8.65125846862793,
"learning_rate": 1.9950656183027392e-05,
"loss": 1.158308506011963,
"step": 180
},
{
"epoch": 0.12338104976141787,
"grad_norm": 10.54161262512207,
"learning_rate": 1.9949554489896542e-05,
"loss": 2.202761173248291,
"step": 181
},
{
"epoch": 0.12406271301976823,
"grad_norm": 10.052018165588379,
"learning_rate": 1.994844066458711e-05,
"loss": 2.5571885108947754,
"step": 182
},
{
"epoch": 0.12474437627811862,
"grad_norm": 18.71660804748535,
"learning_rate": 1.9947314708457245e-05,
"loss": 0.9831675887107849,
"step": 183
},
{
"epoch": 0.125426039536469,
"grad_norm": 25.179962158203125,
"learning_rate": 1.9946176622879915e-05,
"loss": 1.7804163694381714,
"step": 184
},
{
"epoch": 0.12610770279481937,
"grad_norm": 10.924018859863281,
"learning_rate": 1.994502640924286e-05,
"loss": 1.2503122091293335,
"step": 185
},
{
"epoch": 0.12678936605316973,
"grad_norm": 16.85798454284668,
"learning_rate": 1.994386406894862e-05,
"loss": 1.3904640674591064,
"step": 186
},
{
"epoch": 0.1274710293115201,
"grad_norm": 12.353588104248047,
"learning_rate": 1.9942689603414513e-05,
"loss": 1.937730312347412,
"step": 187
},
{
"epoch": 0.1281526925698705,
"grad_norm": 17.090003967285156,
"learning_rate": 1.9941503014072646e-05,
"loss": 2.1551363468170166,
"step": 188
},
{
"epoch": 0.12883435582822086,
"grad_norm": 6.396425724029541,
"learning_rate": 1.9940304302369912e-05,
"loss": 1.046122670173645,
"step": 189
},
{
"epoch": 0.12951601908657123,
"grad_norm": 5.918776988983154,
"learning_rate": 1.993909346976798e-05,
"loss": 0.5687689781188965,
"step": 190
},
{
"epoch": 0.1301976823449216,
"grad_norm": 5.968915939331055,
"learning_rate": 1.9937870517743304e-05,
"loss": 0.9140273332595825,
"step": 191
},
{
"epoch": 0.130879345603272,
"grad_norm": 23.541154861450195,
"learning_rate": 1.9936635447787112e-05,
"loss": 0.6421869397163391,
"step": 192
},
{
"epoch": 0.13156100886162236,
"grad_norm": 20.733173370361328,
"learning_rate": 1.9935388261405412e-05,
"loss": 1.4283093214035034,
"step": 193
},
{
"epoch": 0.13224267211997273,
"grad_norm": 5.4685821533203125,
"learning_rate": 1.993412896011898e-05,
"loss": 0.9551104307174683,
"step": 194
},
{
"epoch": 0.1329243353783231,
"grad_norm": 6.2597737312316895,
"learning_rate": 1.993285754546338e-05,
"loss": 1.2597966194152832,
"step": 195
},
{
"epoch": 0.1336059986366735,
"grad_norm": 470.6517028808594,
"learning_rate": 1.9931574018988916e-05,
"loss": 1.691474437713623,
"step": 196
},
{
"epoch": 0.13428766189502386,
"grad_norm": 5.0026469230651855,
"learning_rate": 1.99302783822607e-05,
"loss": 0.9203601479530334,
"step": 197
},
{
"epoch": 0.13496932515337423,
"grad_norm": 3.4127256870269775,
"learning_rate": 1.9928970636858584e-05,
"loss": 0.9550940990447998,
"step": 198
},
{
"epoch": 0.1356509884117246,
"grad_norm": 5.123717784881592,
"learning_rate": 1.992765078437719e-05,
"loss": 2.055361747741699,
"step": 199
},
{
"epoch": 0.136332651670075,
"grad_norm": 3.371704339981079,
"learning_rate": 1.9926318826425905e-05,
"loss": 1.0328929424285889,
"step": 200
},
{
"epoch": 0.13701431492842536,
"grad_norm": 3.506957769393921,
"learning_rate": 1.992497476462888e-05,
"loss": 0.9359508156776428,
"step": 201
},
{
"epoch": 0.13769597818677573,
"grad_norm": 2.7216708660125732,
"learning_rate": 1.9923618600625025e-05,
"loss": 1.400390386581421,
"step": 202
},
{
"epoch": 0.1383776414451261,
"grad_norm": 4.871631145477295,
"learning_rate": 1.9922250336068e-05,
"loss": 1.239612102508545,
"step": 203
},
{
"epoch": 0.1390593047034765,
"grad_norm": 2.473804473876953,
"learning_rate": 1.9920869972626225e-05,
"loss": 1.3447606563568115,
"step": 204
},
{
"epoch": 0.13974096796182686,
"grad_norm": 6.894538879394531,
"learning_rate": 1.9919477511982873e-05,
"loss": 1.5631332397460938,
"step": 205
},
{
"epoch": 0.14042263122017723,
"grad_norm": 3.913583517074585,
"learning_rate": 1.991807295583587e-05,
"loss": 1.618501901626587,
"step": 206
},
{
"epoch": 0.1411042944785276,
"grad_norm": 3.0639336109161377,
"learning_rate": 1.991665630589788e-05,
"loss": 1.318312644958496,
"step": 207
},
{
"epoch": 0.141785957736878,
"grad_norm": 5.246321678161621,
"learning_rate": 1.9915227563896327e-05,
"loss": 0.876002848148346,
"step": 208
},
{
"epoch": 0.14246762099522836,
"grad_norm": 3.295959949493408,
"learning_rate": 1.9913786731573382e-05,
"loss": 0.4557987451553345,
"step": 209
},
{
"epoch": 0.14314928425357873,
"grad_norm": 3.8275537490844727,
"learning_rate": 1.991233381068594e-05,
"loss": 0.9583624601364136,
"step": 210
},
{
"epoch": 0.1438309475119291,
"grad_norm": 4.7201128005981445,
"learning_rate": 1.991086880300565e-05,
"loss": 1.0539599657058716,
"step": 211
},
{
"epoch": 0.1445126107702795,
"grad_norm": 4.5735039710998535,
"learning_rate": 1.9909391710318907e-05,
"loss": 0.9567979574203491,
"step": 212
},
{
"epoch": 0.14519427402862986,
"grad_norm": 2.669004440307617,
"learning_rate": 1.990790253442682e-05,
"loss": 0.7672552466392517,
"step": 213
},
{
"epoch": 0.14587593728698023,
"grad_norm": 9.868934631347656,
"learning_rate": 1.990640127714525e-05,
"loss": 2.299375057220459,
"step": 214
},
{
"epoch": 0.1465576005453306,
"grad_norm": 4.517803192138672,
"learning_rate": 1.990488794030478e-05,
"loss": 0.5457000732421875,
"step": 215
},
{
"epoch": 0.147239263803681,
"grad_norm": 2.341135263442993,
"learning_rate": 1.990336252575073e-05,
"loss": 0.6983858346939087,
"step": 216
},
{
"epoch": 0.14792092706203136,
"grad_norm": 2.1451797485351562,
"learning_rate": 1.990182503534314e-05,
"loss": 0.9473968148231506,
"step": 217
},
{
"epoch": 0.14860259032038173,
"grad_norm": 2.1667206287384033,
"learning_rate": 1.9900275470956778e-05,
"loss": 0.6302728056907654,
"step": 218
},
{
"epoch": 0.1492842535787321,
"grad_norm": 6.366608619689941,
"learning_rate": 1.9898713834481137e-05,
"loss": 1.6893022060394287,
"step": 219
},
{
"epoch": 0.1499659168370825,
"grad_norm": 2.48405122756958,
"learning_rate": 1.9897140127820432e-05,
"loss": 0.5234397053718567,
"step": 220
},
{
"epoch": 0.15064758009543286,
"grad_norm": 4.939483642578125,
"learning_rate": 1.9895554352893584e-05,
"loss": 1.7856894731521606,
"step": 221
},
{
"epoch": 0.15132924335378323,
"grad_norm": 12.190360069274902,
"learning_rate": 1.9893956511634242e-05,
"loss": 1.9232455492019653,
"step": 222
},
{
"epoch": 0.1520109066121336,
"grad_norm": 8.300911903381348,
"learning_rate": 1.9892346605990764e-05,
"loss": 0.9950571656227112,
"step": 223
},
{
"epoch": 0.152692569870484,
"grad_norm": 6.4139814376831055,
"learning_rate": 1.989072463792622e-05,
"loss": 1.2645041942596436,
"step": 224
},
{
"epoch": 0.15337423312883436,
"grad_norm": 6.562028408050537,
"learning_rate": 1.9889090609418384e-05,
"loss": 2.0295963287353516,
"step": 225
},
{
"epoch": 0.15405589638718473,
"grad_norm": 5.604927062988281,
"learning_rate": 1.9887444522459743e-05,
"loss": 1.2279257774353027,
"step": 226
},
{
"epoch": 0.1547375596455351,
"grad_norm": 5.523647785186768,
"learning_rate": 1.9885786379057487e-05,
"loss": 2.1371541023254395,
"step": 227
},
{
"epoch": 0.1554192229038855,
"grad_norm": 5.242480754852295,
"learning_rate": 1.9884116181233496e-05,
"loss": 1.2968311309814453,
"step": 228
},
{
"epoch": 0.15610088616223586,
"grad_norm": 5.943783283233643,
"learning_rate": 1.9882433931024367e-05,
"loss": 0.5952497124671936,
"step": 229
},
{
"epoch": 0.15678254942058623,
"grad_norm": 210.10638427734375,
"learning_rate": 1.9880739630481376e-05,
"loss": 1.4420039653778076,
"step": 230
},
{
"epoch": 0.1574642126789366,
"grad_norm": 5.810100078582764,
"learning_rate": 1.9879033281670508e-05,
"loss": 1.5601913928985596,
"step": 231
},
{
"epoch": 0.158145875937287,
"grad_norm": 5.746815204620361,
"learning_rate": 1.9877314886672424e-05,
"loss": 1.6530776023864746,
"step": 232
},
{
"epoch": 0.15882753919563736,
"grad_norm": 4.114201545715332,
"learning_rate": 1.987558444758249e-05,
"loss": 1.1385934352874756,
"step": 233
},
{
"epoch": 0.15950920245398773,
"grad_norm": 4.840907096862793,
"learning_rate": 1.9873841966510744e-05,
"loss": 1.0146422386169434,
"step": 234
},
{
"epoch": 0.1601908657123381,
"grad_norm": 3.263094425201416,
"learning_rate": 1.9872087445581912e-05,
"loss": 1.216705560684204,
"step": 235
},
{
"epoch": 0.1608725289706885,
"grad_norm": 5.41212797164917,
"learning_rate": 1.987032088693541e-05,
"loss": 1.4565320014953613,
"step": 236
},
{
"epoch": 0.16155419222903886,
"grad_norm": 5.620246410369873,
"learning_rate": 1.9868542292725316e-05,
"loss": 1.5943286418914795,
"step": 237
},
{
"epoch": 0.16223585548738922,
"grad_norm": 7.258380889892578,
"learning_rate": 1.9866751665120398e-05,
"loss": 0.7804931402206421,
"step": 238
},
{
"epoch": 0.1629175187457396,
"grad_norm": 20.362430572509766,
"learning_rate": 1.9864949006304094e-05,
"loss": 1.0291253328323364,
"step": 239
},
{
"epoch": 0.16359918200409,
"grad_norm": 4.7177300453186035,
"learning_rate": 1.9863134318474504e-05,
"loss": 1.457777738571167,
"step": 240
},
{
"epoch": 0.16428084526244036,
"grad_norm": 3.546522378921509,
"learning_rate": 1.986130760384441e-05,
"loss": 1.5622825622558594,
"step": 241
},
{
"epoch": 0.16496250852079072,
"grad_norm": 3.295022964477539,
"learning_rate": 1.9859468864641242e-05,
"loss": 0.9937570095062256,
"step": 242
},
{
"epoch": 0.1656441717791411,
"grad_norm": 4.15577507019043,
"learning_rate": 1.985761810310711e-05,
"loss": 1.1751668453216553,
"step": 243
},
{
"epoch": 0.1663258350374915,
"grad_norm": 7.615987777709961,
"learning_rate": 1.985575532149877e-05,
"loss": 1.6934175491333008,
"step": 244
},
{
"epoch": 0.16700749829584186,
"grad_norm": 4.522923469543457,
"learning_rate": 1.985388052208764e-05,
"loss": 1.5461843013763428,
"step": 245
},
{
"epoch": 0.16768916155419222,
"grad_norm": 4.890939712524414,
"learning_rate": 1.9851993707159794e-05,
"loss": 0.9180086255073547,
"step": 246
},
{
"epoch": 0.1683708248125426,
"grad_norm": 7.1219096183776855,
"learning_rate": 1.985009487901596e-05,
"loss": 1.4076170921325684,
"step": 247
},
{
"epoch": 0.169052488070893,
"grad_norm": 6.1358184814453125,
"learning_rate": 1.9848184039971502e-05,
"loss": 1.320831060409546,
"step": 248
},
{
"epoch": 0.16973415132924335,
"grad_norm": 5.749042510986328,
"learning_rate": 1.9846261192356442e-05,
"loss": 1.53205406665802,
"step": 249
},
{
"epoch": 0.17041581458759372,
"grad_norm": 2.8806917667388916,
"learning_rate": 1.9844326338515444e-05,
"loss": 1.121159553527832,
"step": 250
},
{
"epoch": 0.1710974778459441,
"grad_norm": 3.486462354660034,
"learning_rate": 1.9842379480807804e-05,
"loss": 0.8663906455039978,
"step": 251
},
{
"epoch": 0.17177914110429449,
"grad_norm": 2.8455584049224854,
"learning_rate": 1.984042062160746e-05,
"loss": 1.0783741474151611,
"step": 252
},
{
"epoch": 0.17246080436264485,
"grad_norm": 8.036139488220215,
"learning_rate": 1.983844976330299e-05,
"loss": 1.7741141319274902,
"step": 253
},
{
"epoch": 0.17314246762099522,
"grad_norm": 6.524624824523926,
"learning_rate": 1.983646690829759e-05,
"loss": 1.7694973945617676,
"step": 254
},
{
"epoch": 0.1738241308793456,
"grad_norm": 5.912774562835693,
"learning_rate": 1.9834472059009097e-05,
"loss": 1.4783644676208496,
"step": 255
},
{
"epoch": 0.17450579413769599,
"grad_norm": 3.9578967094421387,
"learning_rate": 1.9832465217869964e-05,
"loss": 1.1265515089035034,
"step": 256
},
{
"epoch": 0.17518745739604635,
"grad_norm": 9.286506652832031,
"learning_rate": 1.9830446387327277e-05,
"loss": 2.3992319107055664,
"step": 257
},
{
"epoch": 0.17586912065439672,
"grad_norm": 42.55002975463867,
"learning_rate": 1.9828415569842732e-05,
"loss": 2.409393787384033,
"step": 258
},
{
"epoch": 0.17655078391274712,
"grad_norm": 33.78275680541992,
"learning_rate": 1.9826372767892644e-05,
"loss": 0.7688829898834229,
"step": 259
},
{
"epoch": 0.17723244717109748,
"grad_norm": 8.896828651428223,
"learning_rate": 1.982431798396794e-05,
"loss": 1.742114782333374,
"step": 260
},
{
"epoch": 0.17791411042944785,
"grad_norm": 14.53412914276123,
"learning_rate": 1.982225122057417e-05,
"loss": 1.395640254020691,
"step": 261
},
{
"epoch": 0.17859577368779822,
"grad_norm": 8.733621597290039,
"learning_rate": 1.9820172480231476e-05,
"loss": 1.3212776184082031,
"step": 262
},
{
"epoch": 0.17927743694614862,
"grad_norm": 6.948071479797363,
"learning_rate": 1.9818081765474602e-05,
"loss": 1.8596258163452148,
"step": 263
},
{
"epoch": 0.17995910020449898,
"grad_norm": 7.661222457885742,
"learning_rate": 1.981597907885291e-05,
"loss": 1.5962014198303223,
"step": 264
},
{
"epoch": 0.18064076346284935,
"grad_norm": 5.942171096801758,
"learning_rate": 1.9813864422930345e-05,
"loss": 1.7828670740127563,
"step": 265
},
{
"epoch": 0.18132242672119972,
"grad_norm": 8.718239784240723,
"learning_rate": 1.981173780028546e-05,
"loss": 1.8014458417892456,
"step": 266
},
{
"epoch": 0.18200408997955012,
"grad_norm": 10.833904266357422,
"learning_rate": 1.980959921351139e-05,
"loss": 1.397186517715454,
"step": 267
},
{
"epoch": 0.18268575323790048,
"grad_norm": 3.20481014251709,
"learning_rate": 1.980744866521586e-05,
"loss": 0.9240924715995789,
"step": 268
},
{
"epoch": 0.18336741649625085,
"grad_norm": 3.304488182067871,
"learning_rate": 1.980528615802118e-05,
"loss": 1.1787257194519043,
"step": 269
},
{
"epoch": 0.18404907975460122,
"grad_norm": 3.4648053646087646,
"learning_rate": 1.9803111694564246e-05,
"loss": 0.8863788843154907,
"step": 270
},
{
"epoch": 0.18473074301295161,
"grad_norm": 2.342740297317505,
"learning_rate": 1.9800925277496532e-05,
"loss": 0.9875154495239258,
"step": 271
},
{
"epoch": 0.18541240627130198,
"grad_norm": 3.5584747791290283,
"learning_rate": 1.979872690948409e-05,
"loss": 0.8974324464797974,
"step": 272
},
{
"epoch": 0.18609406952965235,
"grad_norm": 4.4199628829956055,
"learning_rate": 1.9796516593207537e-05,
"loss": 1.2789227962493896,
"step": 273
},
{
"epoch": 0.18677573278800272,
"grad_norm": 7.764851093292236,
"learning_rate": 1.979429433136207e-05,
"loss": 2.7128586769104004,
"step": 274
},
{
"epoch": 0.18745739604635311,
"grad_norm": 7.048688888549805,
"learning_rate": 1.9792060126657437e-05,
"loss": 1.1671762466430664,
"step": 275
},
{
"epoch": 0.18813905930470348,
"grad_norm": 19.91363525390625,
"learning_rate": 1.9789813981817963e-05,
"loss": 1.2773725986480713,
"step": 276
},
{
"epoch": 0.18882072256305385,
"grad_norm": 13.8008394241333,
"learning_rate": 1.9787555899582533e-05,
"loss": 1.5615161657333374,
"step": 277
},
{
"epoch": 0.18950238582140422,
"grad_norm": 6.971741199493408,
"learning_rate": 1.978528588270458e-05,
"loss": 1.249703049659729,
"step": 278
},
{
"epoch": 0.1901840490797546,
"grad_norm": 8.02096939086914,
"learning_rate": 1.9783003933952082e-05,
"loss": 1.4458056688308716,
"step": 279
},
{
"epoch": 0.19086571233810498,
"grad_norm": 4.437285423278809,
"learning_rate": 1.9780710056107587e-05,
"loss": 1.5829441547393799,
"step": 280
},
{
"epoch": 0.19154737559645535,
"grad_norm": 7.333526611328125,
"learning_rate": 1.9778404251968176e-05,
"loss": 1.612710952758789,
"step": 281
},
{
"epoch": 0.19222903885480572,
"grad_norm": 5.8669843673706055,
"learning_rate": 1.977608652434548e-05,
"loss": 1.884479284286499,
"step": 282
},
{
"epoch": 0.1929107021131561,
"grad_norm": 17.100505828857422,
"learning_rate": 1.9773756876065655e-05,
"loss": 2.108002185821533,
"step": 283
},
{
"epoch": 0.19359236537150648,
"grad_norm": 5.0063090324401855,
"learning_rate": 1.9771415309969406e-05,
"loss": 1.2777209281921387,
"step": 284
},
{
"epoch": 0.19427402862985685,
"grad_norm": 5.124573707580566,
"learning_rate": 1.976906182891197e-05,
"loss": 1.4534913301467896,
"step": 285
},
{
"epoch": 0.19495569188820722,
"grad_norm": 5.953328609466553,
"learning_rate": 1.97666964357631e-05,
"loss": 1.2062658071517944,
"step": 286
},
{
"epoch": 0.1956373551465576,
"grad_norm": 5.880934715270996,
"learning_rate": 1.9764319133407095e-05,
"loss": 1.77345871925354,
"step": 287
},
{
"epoch": 0.19631901840490798,
"grad_norm": 3.9601995944976807,
"learning_rate": 1.9761929924742756e-05,
"loss": 1.3594872951507568,
"step": 288
},
{
"epoch": 0.19700068166325835,
"grad_norm": 4.685924530029297,
"learning_rate": 1.975952881268341e-05,
"loss": 1.5763859748840332,
"step": 289
},
{
"epoch": 0.19768234492160872,
"grad_norm": 3.826180934906006,
"learning_rate": 1.97571158001569e-05,
"loss": 1.8830716609954834,
"step": 290
},
{
"epoch": 0.1983640081799591,
"grad_norm": 3.6553056240081787,
"learning_rate": 1.975469089010558e-05,
"loss": 1.1618760824203491,
"step": 291
},
{
"epoch": 0.19904567143830948,
"grad_norm": 5.389840126037598,
"learning_rate": 1.975225408548631e-05,
"loss": 1.0185208320617676,
"step": 292
},
{
"epoch": 0.19972733469665985,
"grad_norm": 6.093509674072266,
"learning_rate": 1.9749805389270453e-05,
"loss": 2.012967109680176,
"step": 293
},
{
"epoch": 0.20040899795501022,
"grad_norm": 3.8174962997436523,
"learning_rate": 1.9747344804443873e-05,
"loss": 1.5239055156707764,
"step": 294
},
{
"epoch": 0.2010906612133606,
"grad_norm": 2.4020743370056152,
"learning_rate": 1.9744872334006936e-05,
"loss": 1.4020321369171143,
"step": 295
},
{
"epoch": 0.20177232447171098,
"grad_norm": 2.9826319217681885,
"learning_rate": 1.9742387980974484e-05,
"loss": 1.3844187259674072,
"step": 296
},
{
"epoch": 0.20245398773006135,
"grad_norm": 6.365255355834961,
"learning_rate": 1.973989174837587e-05,
"loss": 1.321211338043213,
"step": 297
},
{
"epoch": 0.20313565098841171,
"grad_norm": 100.61157989501953,
"learning_rate": 1.9737383639254924e-05,
"loss": 1.522669792175293,
"step": 298
},
{
"epoch": 0.2038173142467621,
"grad_norm": 7.6911940574646,
"learning_rate": 1.9734863656669948e-05,
"loss": 1.4330592155456543,
"step": 299
},
{
"epoch": 0.20449897750511248,
"grad_norm": 4.5384721755981445,
"learning_rate": 1.973233180369374e-05,
"loss": 1.3101820945739746,
"step": 300
},
{
"epoch": 0.20518064076346285,
"grad_norm": 4.599635601043701,
"learning_rate": 1.9729788083413558e-05,
"loss": 1.6237077713012695,
"step": 301
},
{
"epoch": 0.2058623040218132,
"grad_norm": 4.026865005493164,
"learning_rate": 1.9727232498931135e-05,
"loss": 1.1968040466308594,
"step": 302
},
{
"epoch": 0.2065439672801636,
"grad_norm": 3.2824039459228516,
"learning_rate": 1.972466505336267e-05,
"loss": 1.1076245307922363,
"step": 303
},
{
"epoch": 0.20722563053851398,
"grad_norm": 5.328464031219482,
"learning_rate": 1.9722085749838835e-05,
"loss": 1.4950063228607178,
"step": 304
},
{
"epoch": 0.20790729379686435,
"grad_norm": 5.068223476409912,
"learning_rate": 1.9719494591504747e-05,
"loss": 1.2632877826690674,
"step": 305
},
{
"epoch": 0.2085889570552147,
"grad_norm": 8.749210357666016,
"learning_rate": 1.9716891581519983e-05,
"loss": 2.029329776763916,
"step": 306
},
{
"epoch": 0.2092706203135651,
"grad_norm": 5.680600643157959,
"learning_rate": 1.9714276723058576e-05,
"loss": 2.031003713607788,
"step": 307
},
{
"epoch": 0.20995228357191548,
"grad_norm": 8.02801513671875,
"learning_rate": 1.9711650019309e-05,
"loss": 2.549004077911377,
"step": 308
},
{
"epoch": 0.21063394683026584,
"grad_norm": 15.788527488708496,
"learning_rate": 1.970901147347418e-05,
"loss": 1.2446638345718384,
"step": 309
},
{
"epoch": 0.2113156100886162,
"grad_norm": 7.427382469177246,
"learning_rate": 1.9706361088771474e-05,
"loss": 1.8798837661743164,
"step": 310
},
{
"epoch": 0.2119972733469666,
"grad_norm": 9.844267845153809,
"learning_rate": 1.9703698868432676e-05,
"loss": 1.9102556705474854,
"step": 311
},
{
"epoch": 0.21267893660531698,
"grad_norm": 3.674903631210327,
"learning_rate": 1.9701024815704023e-05,
"loss": 2.105062484741211,
"step": 312
},
{
"epoch": 0.21336059986366734,
"grad_norm": 7.803150177001953,
"learning_rate": 1.9698338933846172e-05,
"loss": 1.1954119205474854,
"step": 313
},
{
"epoch": 0.2140422631220177,
"grad_norm": 9.175468444824219,
"learning_rate": 1.9695641226134196e-05,
"loss": 1.0092734098434448,
"step": 314
},
{
"epoch": 0.2147239263803681,
"grad_norm": 5.410757541656494,
"learning_rate": 1.96929316958576e-05,
"loss": 1.2513163089752197,
"step": 315
},
{
"epoch": 0.21540558963871848,
"grad_norm": 1.7223652601242065,
"learning_rate": 1.9690210346320304e-05,
"loss": 0.8200163841247559,
"step": 316
},
{
"epoch": 0.21608725289706884,
"grad_norm": 4.928730487823486,
"learning_rate": 1.9687477180840634e-05,
"loss": 1.9416441917419434,
"step": 317
},
{
"epoch": 0.2167689161554192,
"grad_norm": 3.1371078491210938,
"learning_rate": 1.9684732202751328e-05,
"loss": 1.1810110807418823,
"step": 318
},
{
"epoch": 0.2174505794137696,
"grad_norm": 4.844557762145996,
"learning_rate": 1.968197541539953e-05,
"loss": 0.7698321342468262,
"step": 319
},
{
"epoch": 0.21813224267211997,
"grad_norm": 5.738413333892822,
"learning_rate": 1.9679206822146776e-05,
"loss": 1.2975870370864868,
"step": 320
},
{
"epoch": 0.21881390593047034,
"grad_norm": 4.841405868530273,
"learning_rate": 1.967642642636901e-05,
"loss": 1.0371558666229248,
"step": 321
},
{
"epoch": 0.2194955691888207,
"grad_norm": 2.5466487407684326,
"learning_rate": 1.9673634231456554e-05,
"loss": 0.5535352230072021,
"step": 322
},
{
"epoch": 0.2201772324471711,
"grad_norm": 6.791589736938477,
"learning_rate": 1.9670830240814127e-05,
"loss": 2.2420239448547363,
"step": 323
},
{
"epoch": 0.22085889570552147,
"grad_norm": 4.811392307281494,
"learning_rate": 1.9668014457860828e-05,
"loss": 0.7682366371154785,
"step": 324
},
{
"epoch": 0.22154055896387184,
"grad_norm": 5.438621520996094,
"learning_rate": 1.9665186886030135e-05,
"loss": 1.7728581428527832,
"step": 325
},
{
"epoch": 0.2222222222222222,
"grad_norm": 3.9589333534240723,
"learning_rate": 1.96623475287699e-05,
"loss": 1.0751643180847168,
"step": 326
},
{
"epoch": 0.2229038854805726,
"grad_norm": 3.6448564529418945,
"learning_rate": 1.965949638954235e-05,
"loss": 1.2647238969802856,
"step": 327
},
{
"epoch": 0.22358554873892297,
"grad_norm": 4.940537929534912,
"learning_rate": 1.9656633471824075e-05,
"loss": 1.1155469417572021,
"step": 328
},
{
"epoch": 0.22426721199727334,
"grad_norm": 7.523629188537598,
"learning_rate": 1.9653758779106028e-05,
"loss": 1.4137378931045532,
"step": 329
},
{
"epoch": 0.2249488752556237,
"grad_norm": 7.168045997619629,
"learning_rate": 1.9650872314893523e-05,
"loss": 0.5647242069244385,
"step": 330
},
{
"epoch": 0.2256305385139741,
"grad_norm": 2.690207004547119,
"learning_rate": 1.964797408270622e-05,
"loss": 1.0476772785186768,
"step": 331
},
{
"epoch": 0.22631220177232447,
"grad_norm": 2.781921148300171,
"learning_rate": 1.9645064086078135e-05,
"loss": 0.7483969926834106,
"step": 332
},
{
"epoch": 0.22699386503067484,
"grad_norm": 2.627107858657837,
"learning_rate": 1.964214232855763e-05,
"loss": 0.7267999649047852,
"step": 333
},
{
"epoch": 0.2276755282890252,
"grad_norm": 6.753015995025635,
"learning_rate": 1.9639208813707407e-05,
"loss": 1.645385503768921,
"step": 334
},
{
"epoch": 0.2283571915473756,
"grad_norm": 1.9686328172683716,
"learning_rate": 1.9636263545104498e-05,
"loss": 0.6474498510360718,
"step": 335
},
{
"epoch": 0.22903885480572597,
"grad_norm": 5.547754287719727,
"learning_rate": 1.9633306526340273e-05,
"loss": 1.437086582183838,
"step": 336
},
{
"epoch": 0.22972051806407634,
"grad_norm": 3.0991756916046143,
"learning_rate": 1.9630337761020436e-05,
"loss": 1.0682692527770996,
"step": 337
},
{
"epoch": 0.2304021813224267,
"grad_norm": 7.258819103240967,
"learning_rate": 1.9627357252765e-05,
"loss": 2.262204647064209,
"step": 338
},
{
"epoch": 0.2310838445807771,
"grad_norm": 5.0217390060424805,
"learning_rate": 1.9624365005208303e-05,
"loss": 1.6506143808364868,
"step": 339
},
{
"epoch": 0.23176550783912747,
"grad_norm": 11.728917121887207,
"learning_rate": 1.962136102199901e-05,
"loss": 2.1931800842285156,
"step": 340
},
{
"epoch": 0.23244717109747784,
"grad_norm": 9.160211563110352,
"learning_rate": 1.961834530680007e-05,
"loss": 1.0702080726623535,
"step": 341
},
{
"epoch": 0.2331288343558282,
"grad_norm": 9.413346290588379,
"learning_rate": 1.9615317863288765e-05,
"loss": 0.8787815570831299,
"step": 342
},
{
"epoch": 0.2338104976141786,
"grad_norm": 13.654266357421875,
"learning_rate": 1.9612278695156662e-05,
"loss": 1.8223183155059814,
"step": 343
},
{
"epoch": 0.23449216087252897,
"grad_norm": 347.9715881347656,
"learning_rate": 1.9609227806109627e-05,
"loss": 1.0823979377746582,
"step": 344
},
{
"epoch": 0.23517382413087934,
"grad_norm": 6.461592674255371,
"learning_rate": 1.9606165199867822e-05,
"loss": 2.422264337539673,
"step": 345
},
{
"epoch": 0.23585548738922973,
"grad_norm": 5.662838459014893,
"learning_rate": 1.960309088016569e-05,
"loss": 1.9186232089996338,
"step": 346
},
{
"epoch": 0.2365371506475801,
"grad_norm": 6.0235395431518555,
"learning_rate": 1.9600004850751967e-05,
"loss": 1.1358025074005127,
"step": 347
},
{
"epoch": 0.23721881390593047,
"grad_norm": 3.422726631164551,
"learning_rate": 1.9596907115389656e-05,
"loss": 1.333078384399414,
"step": 348
},
{
"epoch": 0.23790047716428084,
"grad_norm": 5.582715034484863,
"learning_rate": 1.9593797677856043e-05,
"loss": 1.3061596155166626,
"step": 349
},
{
"epoch": 0.23858214042263123,
"grad_norm": 3.6086502075195312,
"learning_rate": 1.959067654194268e-05,
"loss": 1.1203460693359375,
"step": 350
},
{
"epoch": 0.2392638036809816,
"grad_norm": 2.288213014602661,
"learning_rate": 1.9587543711455383e-05,
"loss": 0.8664935827255249,
"step": 351
},
{
"epoch": 0.23994546693933197,
"grad_norm": 3.412855625152588,
"learning_rate": 1.958439919021423e-05,
"loss": 1.3929232358932495,
"step": 352
},
{
"epoch": 0.24062713019768234,
"grad_norm": 2.167471170425415,
"learning_rate": 1.9581242982053546e-05,
"loss": 0.7552188038825989,
"step": 353
},
{
"epoch": 0.24130879345603273,
"grad_norm": 4.317893028259277,
"learning_rate": 1.957807509082192e-05,
"loss": 1.6116098165512085,
"step": 354
},
{
"epoch": 0.2419904567143831,
"grad_norm": 5.659145832061768,
"learning_rate": 1.9574895520382183e-05,
"loss": 1.7634755373001099,
"step": 355
},
{
"epoch": 0.24267211997273347,
"grad_norm": 3.4220974445343018,
"learning_rate": 1.9571704274611397e-05,
"loss": 1.3048694133758545,
"step": 356
},
{
"epoch": 0.24335378323108384,
"grad_norm": 3.2889888286590576,
"learning_rate": 1.956850135740087e-05,
"loss": 1.5248088836669922,
"step": 357
},
{
"epoch": 0.24403544648943423,
"grad_norm": 3.1788370609283447,
"learning_rate": 1.9565286772656145e-05,
"loss": 1.4584882259368896,
"step": 358
},
{
"epoch": 0.2447171097477846,
"grad_norm": 4.251647472381592,
"learning_rate": 1.9562060524296983e-05,
"loss": 1.024428367614746,
"step": 359
},
{
"epoch": 0.24539877300613497,
"grad_norm": 4.799529552459717,
"learning_rate": 1.955882261625737e-05,
"loss": 1.2644996643066406,
"step": 360
},
{
"epoch": 0.24608043626448534,
"grad_norm": 4.792227745056152,
"learning_rate": 1.9555573052485518e-05,
"loss": 2.040235996246338,
"step": 361
},
{
"epoch": 0.24676209952283573,
"grad_norm": 4.666950702667236,
"learning_rate": 1.9552311836943832e-05,
"loss": 1.5763626098632812,
"step": 362
},
{
"epoch": 0.2474437627811861,
"grad_norm": 4.873711585998535,
"learning_rate": 1.9549038973608952e-05,
"loss": 1.2719221115112305,
"step": 363
},
{
"epoch": 0.24812542603953647,
"grad_norm": 5.893383502960205,
"learning_rate": 1.9545754466471696e-05,
"loss": 2.0642623901367188,
"step": 364
},
{
"epoch": 0.24880708929788684,
"grad_norm": 6.61902379989624,
"learning_rate": 1.9542458319537094e-05,
"loss": 1.2182157039642334,
"step": 365
},
{
"epoch": 0.24948875255623723,
"grad_norm": 5.884739875793457,
"learning_rate": 1.9539150536824363e-05,
"loss": 1.106882929801941,
"step": 366
},
{
"epoch": 0.2501704158145876,
"grad_norm": 5.0897650718688965,
"learning_rate": 1.953583112236691e-05,
"loss": 0.5587277412414551,
"step": 367
},
{
"epoch": 0.250852079072938,
"grad_norm": 2.1747047901153564,
"learning_rate": 1.9532500080212333e-05,
"loss": 1.0599546432495117,
"step": 368
},
{
"epoch": 0.25153374233128833,
"grad_norm": 1.9114444255828857,
"learning_rate": 1.9529157414422398e-05,
"loss": 1.4041662216186523,
"step": 369
},
{
"epoch": 0.25221540558963873,
"grad_norm": 4.506577014923096,
"learning_rate": 1.9525803129073046e-05,
"loss": 1.44583261013031,
"step": 370
},
{
"epoch": 0.25289706884798907,
"grad_norm": 5.992087364196777,
"learning_rate": 1.9522437228254386e-05,
"loss": 1.9638934135437012,
"step": 371
},
{
"epoch": 0.25357873210633947,
"grad_norm": 2.0379884243011475,
"learning_rate": 1.9519059716070702e-05,
"loss": 0.9091148972511292,
"step": 372
},
{
"epoch": 0.25426039536468986,
"grad_norm": 5.675978660583496,
"learning_rate": 1.951567059664042e-05,
"loss": 1.7077932357788086,
"step": 373
},
{
"epoch": 0.2549420586230402,
"grad_norm": 8.352693557739258,
"learning_rate": 1.9512269874096132e-05,
"loss": 2.6432814598083496,
"step": 374
},
{
"epoch": 0.2556237218813906,
"grad_norm": 16.751062393188477,
"learning_rate": 1.9508857552584574e-05,
"loss": 1.5420377254486084,
"step": 375
},
{
"epoch": 0.256305385139741,
"grad_norm": 7.535068511962891,
"learning_rate": 1.9505433636266618e-05,
"loss": 1.5092883110046387,
"step": 376
},
{
"epoch": 0.25698704839809133,
"grad_norm": 5.2896246910095215,
"learning_rate": 1.9501998129317288e-05,
"loss": 1.8087358474731445,
"step": 377
},
{
"epoch": 0.25766871165644173,
"grad_norm": 3.864203453063965,
"learning_rate": 1.9498551035925736e-05,
"loss": 1.727104663848877,
"step": 378
},
{
"epoch": 0.25835037491479207,
"grad_norm": 4.895890712738037,
"learning_rate": 1.9495092360295236e-05,
"loss": 0.7697995901107788,
"step": 379
},
{
"epoch": 0.25903203817314246,
"grad_norm": 4.769713401794434,
"learning_rate": 1.9491622106643195e-05,
"loss": 1.466042399406433,
"step": 380
},
{
"epoch": 0.25971370143149286,
"grad_norm": 6.704110145568848,
"learning_rate": 1.9488140279201128e-05,
"loss": 1.9534811973571777,
"step": 381
},
{
"epoch": 0.2603953646898432,
"grad_norm": 3.5679831504821777,
"learning_rate": 1.948464688221467e-05,
"loss": 1.6868244409561157,
"step": 382
},
{
"epoch": 0.2610770279481936,
"grad_norm": 4.395662307739258,
"learning_rate": 1.948114191994356e-05,
"loss": 2.0778064727783203,
"step": 383
},
{
"epoch": 0.261758691206544,
"grad_norm": 3.936203718185425,
"learning_rate": 1.9477625396661643e-05,
"loss": 1.0802814960479736,
"step": 384
},
{
"epoch": 0.26244035446489433,
"grad_norm": 3.488983631134033,
"learning_rate": 1.9474097316656856e-05,
"loss": 1.6045266389846802,
"step": 385
},
{
"epoch": 0.2631220177232447,
"grad_norm": 4.06668758392334,
"learning_rate": 1.947055768423123e-05,
"loss": 1.3836003541946411,
"step": 386
},
{
"epoch": 0.26380368098159507,
"grad_norm": 3.609584331512451,
"learning_rate": 1.9467006503700886e-05,
"loss": 1.6604485511779785,
"step": 387
},
{
"epoch": 0.26448534423994546,
"grad_norm": 4.5511064529418945,
"learning_rate": 1.946344377939602e-05,
"loss": 1.2709903717041016,
"step": 388
},
{
"epoch": 0.26516700749829586,
"grad_norm": 3.8128294944763184,
"learning_rate": 1.945986951566091e-05,
"loss": 1.0375221967697144,
"step": 389
},
{
"epoch": 0.2658486707566462,
"grad_norm": 3.75148606300354,
"learning_rate": 1.9456283716853906e-05,
"loss": 0.6834596991539001,
"step": 390
},
{
"epoch": 0.2665303340149966,
"grad_norm": 6.924191474914551,
"learning_rate": 1.9452686387347414e-05,
"loss": 1.6890287399291992,
"step": 391
},
{
"epoch": 0.267211997273347,
"grad_norm": 3.1461596488952637,
"learning_rate": 1.9449077531527906e-05,
"loss": 1.2176769971847534,
"step": 392
},
{
"epoch": 0.26789366053169733,
"grad_norm": 1.6956125497817993,
"learning_rate": 1.9445457153795912e-05,
"loss": 0.3354303240776062,
"step": 393
},
{
"epoch": 0.2685753237900477,
"grad_norm": 16.950252532958984,
"learning_rate": 1.944182525856601e-05,
"loss": 1.2652069330215454,
"step": 394
},
{
"epoch": 0.26925698704839807,
"grad_norm": 4.389862060546875,
"learning_rate": 1.9438181850266815e-05,
"loss": 1.260068416595459,
"step": 395
},
{
"epoch": 0.26993865030674846,
"grad_norm": 4.1714582443237305,
"learning_rate": 1.9434526933340993e-05,
"loss": 1.3269736766815186,
"step": 396
},
{
"epoch": 0.27062031356509886,
"grad_norm": 7.820770740509033,
"learning_rate": 1.9430860512245233e-05,
"loss": 2.191876173019409,
"step": 397
},
{
"epoch": 0.2713019768234492,
"grad_norm": 8.434717178344727,
"learning_rate": 1.9427182591450252e-05,
"loss": 1.7254667282104492,
"step": 398
},
{
"epoch": 0.2719836400817996,
"grad_norm": 10.429789543151855,
"learning_rate": 1.9423493175440797e-05,
"loss": 2.233715534210205,
"step": 399
},
{
"epoch": 0.27266530334015,
"grad_norm": 6.914125442504883,
"learning_rate": 1.941979226871563e-05,
"loss": 1.3225195407867432,
"step": 400
},
{
"epoch": 0.27334696659850033,
"grad_norm": 5.250466823577881,
"learning_rate": 1.9416079875787518e-05,
"loss": 1.5940214395523071,
"step": 401
},
{
"epoch": 0.2740286298568507,
"grad_norm": 3.1187331676483154,
"learning_rate": 1.9412356001183234e-05,
"loss": 1.6403329372406006,
"step": 402
},
{
"epoch": 0.27471029311520107,
"grad_norm": 3.826421022415161,
"learning_rate": 1.9408620649443563e-05,
"loss": 1.121032953262329,
"step": 403
},
{
"epoch": 0.27539195637355146,
"grad_norm": 4.634712219238281,
"learning_rate": 1.940487382512328e-05,
"loss": 2.140951156616211,
"step": 404
},
{
"epoch": 0.27607361963190186,
"grad_norm": 4.077300071716309,
"learning_rate": 1.9401115532791134e-05,
"loss": 1.3883341550827026,
"step": 405
},
{
"epoch": 0.2767552828902522,
"grad_norm": 5.884093761444092,
"learning_rate": 1.9397345777029877e-05,
"loss": 1.732025384902954,
"step": 406
},
{
"epoch": 0.2774369461486026,
"grad_norm": 8.901309967041016,
"learning_rate": 1.9393564562436235e-05,
"loss": 2.0747663974761963,
"step": 407
},
{
"epoch": 0.278118609406953,
"grad_norm": 5.239158630371094,
"learning_rate": 1.93897718936209e-05,
"loss": 2.3918912410736084,
"step": 408
},
{
"epoch": 0.27880027266530333,
"grad_norm": 6.223039150238037,
"learning_rate": 1.9385967775208538e-05,
"loss": 0.9834439754486084,
"step": 409
},
{
"epoch": 0.2794819359236537,
"grad_norm": 70.54039001464844,
"learning_rate": 1.938215221183777e-05,
"loss": 0.9655174016952515,
"step": 410
},
{
"epoch": 0.28016359918200406,
"grad_norm": 8.442602157592773,
"learning_rate": 1.9378325208161178e-05,
"loss": 2.2607247829437256,
"step": 411
},
{
"epoch": 0.28084526244035446,
"grad_norm": 10.892454147338867,
"learning_rate": 1.937448676884529e-05,
"loss": 2.0963754653930664,
"step": 412
},
{
"epoch": 0.28152692569870486,
"grad_norm": 4.259152412414551,
"learning_rate": 1.9370636898570585e-05,
"loss": 0.9232777953147888,
"step": 413
},
{
"epoch": 0.2822085889570552,
"grad_norm": 4.880434513092041,
"learning_rate": 1.9366775602031466e-05,
"loss": 1.79977285861969,
"step": 414
},
{
"epoch": 0.2828902522154056,
"grad_norm": 3.0103135108947754,
"learning_rate": 1.936290288393629e-05,
"loss": 0.7951688766479492,
"step": 415
},
{
"epoch": 0.283571915473756,
"grad_norm": 7.5906171798706055,
"learning_rate": 1.935901874900732e-05,
"loss": 1.6573779582977295,
"step": 416
},
{
"epoch": 0.2842535787321063,
"grad_norm": 3.9085164070129395,
"learning_rate": 1.9355123201980756e-05,
"loss": 1.1650043725967407,
"step": 417
},
{
"epoch": 0.2849352419904567,
"grad_norm": 10.126424789428711,
"learning_rate": 1.93512162476067e-05,
"loss": 1.5728178024291992,
"step": 418
},
{
"epoch": 0.28561690524880706,
"grad_norm": 3.047975540161133,
"learning_rate": 1.934729789064918e-05,
"loss": 1.0197397470474243,
"step": 419
},
{
"epoch": 0.28629856850715746,
"grad_norm": 3.7309322357177734,
"learning_rate": 1.9343368135886112e-05,
"loss": 1.2601120471954346,
"step": 420
},
{
"epoch": 0.28698023176550785,
"grad_norm": 8.821162223815918,
"learning_rate": 1.9339426988109325e-05,
"loss": 0.5328366160392761,
"step": 421
},
{
"epoch": 0.2876618950238582,
"grad_norm": 4.5019097328186035,
"learning_rate": 1.9335474452124524e-05,
"loss": 0.8345069885253906,
"step": 422
},
{
"epoch": 0.2883435582822086,
"grad_norm": 19.59219741821289,
"learning_rate": 1.9331510532751313e-05,
"loss": 1.2992053031921387,
"step": 423
},
{
"epoch": 0.289025221540559,
"grad_norm": 5.920818328857422,
"learning_rate": 1.9327535234823174e-05,
"loss": 1.5236142873764038,
"step": 424
},
{
"epoch": 0.2897068847989093,
"grad_norm": 2.4441089630126953,
"learning_rate": 1.932354856318746e-05,
"loss": 0.8194361925125122,
"step": 425
},
{
"epoch": 0.2903885480572597,
"grad_norm": 3.7855710983276367,
"learning_rate": 1.9319550522705394e-05,
"loss": 1.374048113822937,
"step": 426
},
{
"epoch": 0.29107021131561006,
"grad_norm": 3.0214314460754395,
"learning_rate": 1.9315541118252068e-05,
"loss": 0.8251756429672241,
"step": 427
},
{
"epoch": 0.29175187457396046,
"grad_norm": 7.234747886657715,
"learning_rate": 1.9311520354716417e-05,
"loss": 1.16514253616333,
"step": 428
},
{
"epoch": 0.29243353783231085,
"grad_norm": 1.991133451461792,
"learning_rate": 1.9307488237001247e-05,
"loss": 0.714950442314148,
"step": 429
},
{
"epoch": 0.2931152010906612,
"grad_norm": 22.254261016845703,
"learning_rate": 1.9303444770023184e-05,
"loss": 0.9669209122657776,
"step": 430
},
{
"epoch": 0.2937968643490116,
"grad_norm": 3.0376803874969482,
"learning_rate": 1.9299389958712717e-05,
"loss": 1.1408151388168335,
"step": 431
},
{
"epoch": 0.294478527607362,
"grad_norm": 4.4752726554870605,
"learning_rate": 1.9295323808014152e-05,
"loss": 1.4061379432678223,
"step": 432
},
{
"epoch": 0.2951601908657123,
"grad_norm": 2.8586220741271973,
"learning_rate": 1.9291246322885627e-05,
"loss": 1.4070537090301514,
"step": 433
},
{
"epoch": 0.2958418541240627,
"grad_norm": 2.5731873512268066,
"learning_rate": 1.9287157508299104e-05,
"loss": 1.1559405326843262,
"step": 434
},
{
"epoch": 0.2965235173824131,
"grad_norm": 3.9198739528656006,
"learning_rate": 1.9283057369240358e-05,
"loss": 1.6907612085342407,
"step": 435
},
{
"epoch": 0.29720518064076346,
"grad_norm": 2.753835916519165,
"learning_rate": 1.9278945910708967e-05,
"loss": 1.4664416313171387,
"step": 436
},
{
"epoch": 0.29788684389911385,
"grad_norm": 3.459925889968872,
"learning_rate": 1.927482313771832e-05,
"loss": 1.1719900369644165,
"step": 437
},
{
"epoch": 0.2985685071574642,
"grad_norm": 3.5883936882019043,
"learning_rate": 1.9270689055295596e-05,
"loss": 1.4042953252792358,
"step": 438
},
{
"epoch": 0.2992501704158146,
"grad_norm": 4.524599075317383,
"learning_rate": 1.926654366848177e-05,
"loss": 1.5059282779693604,
"step": 439
},
{
"epoch": 0.299931833674165,
"grad_norm": 3.603348970413208,
"learning_rate": 1.9262386982331596e-05,
"loss": 0.7287573218345642,
"step": 440
},
{
"epoch": 0.3006134969325153,
"grad_norm": 3.01680850982666,
"learning_rate": 1.9258219001913607e-05,
"loss": 0.9661978483200073,
"step": 441
},
{
"epoch": 0.3012951601908657,
"grad_norm": 4.1275482177734375,
"learning_rate": 1.9254039732310113e-05,
"loss": 1.4540948867797852,
"step": 442
},
{
"epoch": 0.3019768234492161,
"grad_norm": 11.864845275878906,
"learning_rate": 1.9249849178617182e-05,
"loss": 2.273348331451416,
"step": 443
},
{
"epoch": 0.30265848670756645,
"grad_norm": 3.5200693607330322,
"learning_rate": 1.9245647345944647e-05,
"loss": 1.7083361148834229,
"step": 444
},
{
"epoch": 0.30334014996591685,
"grad_norm": 2.3711342811584473,
"learning_rate": 1.9241434239416093e-05,
"loss": 1.1357401609420776,
"step": 445
},
{
"epoch": 0.3040218132242672,
"grad_norm": 130.5746612548828,
"learning_rate": 1.9237209864168855e-05,
"loss": 1.9227306842803955,
"step": 446
},
{
"epoch": 0.3047034764826176,
"grad_norm": 3.825364589691162,
"learning_rate": 1.9232974225354e-05,
"loss": 0.8690076470375061,
"step": 447
},
{
"epoch": 0.305385139740968,
"grad_norm": 2.0452988147735596,
"learning_rate": 1.9228727328136337e-05,
"loss": 1.1484711170196533,
"step": 448
},
{
"epoch": 0.3060668029993183,
"grad_norm": 15.158158302307129,
"learning_rate": 1.92244691776944e-05,
"loss": 1.3377666473388672,
"step": 449
},
{
"epoch": 0.3067484662576687,
"grad_norm": 2.2981607913970947,
"learning_rate": 1.922019977922045e-05,
"loss": 1.059500813484192,
"step": 450
},
{
"epoch": 0.3074301295160191,
"grad_norm": 6.6916327476501465,
"learning_rate": 1.9215919137920452e-05,
"loss": 0.831468939781189,
"step": 451
},
{
"epoch": 0.30811179277436945,
"grad_norm": 1.9115036725997925,
"learning_rate": 1.921162725901409e-05,
"loss": 0.36133643984794617,
"step": 452
},
{
"epoch": 0.30879345603271985,
"grad_norm": 4.945026874542236,
"learning_rate": 1.920732414773475e-05,
"loss": 1.5682481527328491,
"step": 453
},
{
"epoch": 0.3094751192910702,
"grad_norm": 3.5810322761535645,
"learning_rate": 1.9203009809329515e-05,
"loss": 1.7324963808059692,
"step": 454
},
{
"epoch": 0.3101567825494206,
"grad_norm": 2.092665433883667,
"learning_rate": 1.919868424905915e-05,
"loss": 0.765301525592804,
"step": 455
},
{
"epoch": 0.310838445807771,
"grad_norm": 4.602624893188477,
"learning_rate": 1.9194347472198112e-05,
"loss": 0.37820541858673096,
"step": 456
},
{
"epoch": 0.3115201090661213,
"grad_norm": 3.6788721084594727,
"learning_rate": 1.9189999484034533e-05,
"loss": 0.5887208580970764,
"step": 457
},
{
"epoch": 0.3122017723244717,
"grad_norm": 190.16064453125,
"learning_rate": 1.9185640289870213e-05,
"loss": 0.9291536211967468,
"step": 458
},
{
"epoch": 0.3128834355828221,
"grad_norm": 7.2953410148620605,
"learning_rate": 1.9181269895020624e-05,
"loss": 1.3983128070831299,
"step": 459
},
{
"epoch": 0.31356509884117245,
"grad_norm": 2.5724921226501465,
"learning_rate": 1.9176888304814882e-05,
"loss": 0.994464635848999,
"step": 460
},
{
"epoch": 0.31424676209952285,
"grad_norm": 5.118542194366455,
"learning_rate": 1.9172495524595764e-05,
"loss": 1.9143290519714355,
"step": 461
},
{
"epoch": 0.3149284253578732,
"grad_norm": 29.36760139465332,
"learning_rate": 1.9168091559719696e-05,
"loss": 2.25297212600708,
"step": 462
},
{
"epoch": 0.3156100886162236,
"grad_norm": 3.8203303813934326,
"learning_rate": 1.9163676415556734e-05,
"loss": 0.7553459405899048,
"step": 463
},
{
"epoch": 0.316291751874574,
"grad_norm": 7.625334739685059,
"learning_rate": 1.9159250097490563e-05,
"loss": 2.2948050498962402,
"step": 464
},
{
"epoch": 0.3169734151329243,
"grad_norm": 10.45826244354248,
"learning_rate": 1.9154812610918503e-05,
"loss": 1.3943508863449097,
"step": 465
},
{
"epoch": 0.3176550783912747,
"grad_norm": 6.181829929351807,
"learning_rate": 1.9150363961251485e-05,
"loss": 0.7830055952072144,
"step": 466
},
{
"epoch": 0.3183367416496251,
"grad_norm": 5.109697341918945,
"learning_rate": 1.914590415391406e-05,
"loss": 1.253631591796875,
"step": 467
},
{
"epoch": 0.31901840490797545,
"grad_norm": 8.534515380859375,
"learning_rate": 1.9141433194344374e-05,
"loss": 1.3872311115264893,
"step": 468
},
{
"epoch": 0.31970006816632585,
"grad_norm": 4.851257801055908,
"learning_rate": 1.9136951087994176e-05,
"loss": 2.131148338317871,
"step": 469
},
{
"epoch": 0.3203817314246762,
"grad_norm": 8.646773338317871,
"learning_rate": 1.913245784032881e-05,
"loss": 1.5957164764404297,
"step": 470
},
{
"epoch": 0.3210633946830266,
"grad_norm": 4.015015125274658,
"learning_rate": 1.9127953456827205e-05,
"loss": 0.8922734260559082,
"step": 471
},
{
"epoch": 0.321745057941377,
"grad_norm": 5.832241535186768,
"learning_rate": 1.912343794298186e-05,
"loss": 1.2595834732055664,
"step": 472
},
{
"epoch": 0.3224267211997273,
"grad_norm": 2.19978404045105,
"learning_rate": 1.911891130429886e-05,
"loss": 0.8392937183380127,
"step": 473
},
{
"epoch": 0.3231083844580777,
"grad_norm": 5.221693515777588,
"learning_rate": 1.9114373546297844e-05,
"loss": 1.1632442474365234,
"step": 474
},
{
"epoch": 0.3237900477164281,
"grad_norm": 5.705185890197754,
"learning_rate": 1.9109824674512014e-05,
"loss": 1.3697636127471924,
"step": 475
},
{
"epoch": 0.32447171097477845,
"grad_norm": 5.748249530792236,
"learning_rate": 1.9105264694488124e-05,
"loss": 1.4544509649276733,
"step": 476
},
{
"epoch": 0.32515337423312884,
"grad_norm": 4.253024101257324,
"learning_rate": 1.9100693611786472e-05,
"loss": 0.7538120746612549,
"step": 477
},
{
"epoch": 0.3258350374914792,
"grad_norm": 2.40335750579834,
"learning_rate": 1.9096111431980896e-05,
"loss": 0.9789775013923645,
"step": 478
},
{
"epoch": 0.3265167007498296,
"grad_norm": 15.633095741271973,
"learning_rate": 1.9091518160658763e-05,
"loss": 1.4342238903045654,
"step": 479
},
{
"epoch": 0.32719836400818,
"grad_norm": 2.747086763381958,
"learning_rate": 1.9086913803420966e-05,
"loss": 1.5781259536743164,
"step": 480
},
{
"epoch": 0.3278800272665303,
"grad_norm": 3.0719778537750244,
"learning_rate": 1.9082298365881916e-05,
"loss": 1.3554413318634033,
"step": 481
},
{
"epoch": 0.3285616905248807,
"grad_norm": 2.799806594848633,
"learning_rate": 1.907767185366953e-05,
"loss": 1.06112539768219,
"step": 482
},
{
"epoch": 0.3292433537832311,
"grad_norm": 1.8001681566238403,
"learning_rate": 1.9073034272425245e-05,
"loss": 0.9077748656272888,
"step": 483
},
{
"epoch": 0.32992501704158145,
"grad_norm": 2.3758649826049805,
"learning_rate": 1.9068385627803972e-05,
"loss": 1.0518077611923218,
"step": 484
},
{
"epoch": 0.33060668029993184,
"grad_norm": 4.855216026306152,
"learning_rate": 1.906372592547413e-05,
"loss": 0.8362563252449036,
"step": 485
},
{
"epoch": 0.3312883435582822,
"grad_norm": 3.3203141689300537,
"learning_rate": 1.905905517111761e-05,
"loss": 1.283669114112854,
"step": 486
},
{
"epoch": 0.3319700068166326,
"grad_norm": 9.288386344909668,
"learning_rate": 1.90543733704298e-05,
"loss": 0.9806363582611084,
"step": 487
},
{
"epoch": 0.332651670074983,
"grad_norm": 2.7394044399261475,
"learning_rate": 1.9049680529119524e-05,
"loss": 1.4370222091674805,
"step": 488
},
{
"epoch": 0.3333333333333333,
"grad_norm": 3.870551586151123,
"learning_rate": 1.9044976652909102e-05,
"loss": 2.071592330932617,
"step": 489
},
{
"epoch": 0.3340149965916837,
"grad_norm": 2.8228816986083984,
"learning_rate": 1.9040261747534282e-05,
"loss": 0.5981768369674683,
"step": 490
},
{
"epoch": 0.3346966598500341,
"grad_norm": 4.471041679382324,
"learning_rate": 1.9035535818744286e-05,
"loss": 1.2177919149398804,
"step": 491
},
{
"epoch": 0.33537832310838445,
"grad_norm": 9.052423477172852,
"learning_rate": 1.9030798872301758e-05,
"loss": 1.2425520420074463,
"step": 492
},
{
"epoch": 0.33605998636673484,
"grad_norm": 3.4728076457977295,
"learning_rate": 1.9026050913982788e-05,
"loss": 0.8533923625946045,
"step": 493
},
{
"epoch": 0.3367416496250852,
"grad_norm": 1.923197865486145,
"learning_rate": 1.9021291949576883e-05,
"loss": 0.8370528221130371,
"step": 494
},
{
"epoch": 0.3374233128834356,
"grad_norm": 3.5292904376983643,
"learning_rate": 1.9016521984886984e-05,
"loss": 1.0731617212295532,
"step": 495
},
{
"epoch": 0.338104976141786,
"grad_norm": 2.2630650997161865,
"learning_rate": 1.901174102572943e-05,
"loss": 0.8468523621559143,
"step": 496
},
{
"epoch": 0.3387866394001363,
"grad_norm": 4.356873989105225,
"learning_rate": 1.9006949077933984e-05,
"loss": 1.2603099346160889,
"step": 497
},
{
"epoch": 0.3394683026584867,
"grad_norm": 2.700840473175049,
"learning_rate": 1.900214614734379e-05,
"loss": 1.1067051887512207,
"step": 498
},
{
"epoch": 0.3401499659168371,
"grad_norm": 5.208966255187988,
"learning_rate": 1.8997332239815403e-05,
"loss": 1.839177131652832,
"step": 499
},
{
"epoch": 0.34083162917518744,
"grad_norm": 5.6692633628845215,
"learning_rate": 1.8992507361218743e-05,
"loss": 1.3088819980621338,
"step": 500
},
{
"epoch": 0.34151329243353784,
"grad_norm": 8.791088104248047,
"learning_rate": 1.8987671517437122e-05,
"loss": 2.0345616340637207,
"step": 501
},
{
"epoch": 0.3421949556918882,
"grad_norm": 6.057740211486816,
"learning_rate": 1.8982824714367214e-05,
"loss": 0.721781313419342,
"step": 502
},
{
"epoch": 0.3428766189502386,
"grad_norm": 4.495492935180664,
"learning_rate": 1.8977966957919068e-05,
"loss": 0.5859732627868652,
"step": 503
},
{
"epoch": 0.34355828220858897,
"grad_norm": 3.266132116317749,
"learning_rate": 1.8973098254016074e-05,
"loss": 1.680146336555481,
"step": 504
},
{
"epoch": 0.3442399454669393,
"grad_norm": 3.1697919368743896,
"learning_rate": 1.8968218608594987e-05,
"loss": 1.515268087387085,
"step": 505
},
{
"epoch": 0.3449216087252897,
"grad_norm": 3.3663229942321777,
"learning_rate": 1.8963328027605886e-05,
"loss": 0.5287567377090454,
"step": 506
},
{
"epoch": 0.3456032719836401,
"grad_norm": 2.6414244174957275,
"learning_rate": 1.8958426517012203e-05,
"loss": 0.8111573457717896,
"step": 507
},
{
"epoch": 0.34628493524199044,
"grad_norm": 6.3523454666137695,
"learning_rate": 1.8953514082790683e-05,
"loss": 1.9260203838348389,
"step": 508
},
{
"epoch": 0.34696659850034084,
"grad_norm": 3.327556610107422,
"learning_rate": 1.8948590730931394e-05,
"loss": 0.9914911985397339,
"step": 509
},
{
"epoch": 0.3476482617586912,
"grad_norm": 2.532057523727417,
"learning_rate": 1.8943656467437726e-05,
"loss": 0.9729138612747192,
"step": 510
},
{
"epoch": 0.3483299250170416,
"grad_norm": 4.6469268798828125,
"learning_rate": 1.893871129832636e-05,
"loss": 2.3366539478302,
"step": 511
},
{
"epoch": 0.34901158827539197,
"grad_norm": 2.643529176712036,
"learning_rate": 1.893375522962729e-05,
"loss": 0.7275136709213257,
"step": 512
},
{
"epoch": 0.3496932515337423,
"grad_norm": 3.031022787094116,
"learning_rate": 1.8928788267383783e-05,
"loss": 1.2368642091751099,
"step": 513
},
{
"epoch": 0.3503749147920927,
"grad_norm": 2.6746222972869873,
"learning_rate": 1.8923810417652404e-05,
"loss": 1.6857903003692627,
"step": 514
},
{
"epoch": 0.3510565780504431,
"grad_norm": 2.2445945739746094,
"learning_rate": 1.8918821686502992e-05,
"loss": 1.0779474973678589,
"step": 515
},
{
"epoch": 0.35173824130879344,
"grad_norm": 3.943788528442383,
"learning_rate": 1.8913822080018645e-05,
"loss": 1.655326008796692,
"step": 516
},
{
"epoch": 0.35241990456714384,
"grad_norm": 4.718496322631836,
"learning_rate": 1.8908811604295728e-05,
"loss": 0.8111345767974854,
"step": 517
},
{
"epoch": 0.35310156782549423,
"grad_norm": 2.1908249855041504,
"learning_rate": 1.8903790265443865e-05,
"loss": 0.7612648010253906,
"step": 518
},
{
"epoch": 0.3537832310838446,
"grad_norm": 3.267055034637451,
"learning_rate": 1.8898758069585923e-05,
"loss": 0.9781118631362915,
"step": 519
},
{
"epoch": 0.35446489434219497,
"grad_norm": 5.063033103942871,
"learning_rate": 1.8893715022858e-05,
"loss": 1.366690754890442,
"step": 520
},
{
"epoch": 0.3551465576005453,
"grad_norm": 5.890979290008545,
"learning_rate": 1.888866113140943e-05,
"loss": 2.1514992713928223,
"step": 521
},
{
"epoch": 0.3558282208588957,
"grad_norm": 4.908336639404297,
"learning_rate": 1.8883596401402777e-05,
"loss": 1.4852519035339355,
"step": 522
},
{
"epoch": 0.3565098841172461,
"grad_norm": 2.1742019653320312,
"learning_rate": 1.8878520839013812e-05,
"loss": 1.223859429359436,
"step": 523
},
{
"epoch": 0.35719154737559644,
"grad_norm": 6.2009124755859375,
"learning_rate": 1.8873434450431522e-05,
"loss": 1.1366631984710693,
"step": 524
},
{
"epoch": 0.35787321063394684,
"grad_norm": 4.069712162017822,
"learning_rate": 1.886833724185809e-05,
"loss": 0.6879901885986328,
"step": 525
},
{
"epoch": 0.35855487389229723,
"grad_norm": 7.754619121551514,
"learning_rate": 1.8863229219508892e-05,
"loss": 2.0778799057006836,
"step": 526
},
{
"epoch": 0.3592365371506476,
"grad_norm": 5.237489223480225,
"learning_rate": 1.8858110389612495e-05,
"loss": 1.3526272773742676,
"step": 527
},
{
"epoch": 0.35991820040899797,
"grad_norm": 2.5144271850585938,
"learning_rate": 1.885298075841064e-05,
"loss": 1.1102807521820068,
"step": 528
},
{
"epoch": 0.3605998636673483,
"grad_norm": 3.9299678802490234,
"learning_rate": 1.8847840332158243e-05,
"loss": 2.0058670043945312,
"step": 529
},
{
"epoch": 0.3612815269256987,
"grad_norm": 3.4149467945098877,
"learning_rate": 1.8842689117123377e-05,
"loss": 0.6895046830177307,
"step": 530
},
{
"epoch": 0.3619631901840491,
"grad_norm": 4.579489707946777,
"learning_rate": 1.8837527119587277e-05,
"loss": 1.8255746364593506,
"step": 531
},
{
"epoch": 0.36264485344239944,
"grad_norm": 2.0138349533081055,
"learning_rate": 1.883235434584432e-05,
"loss": 0.8882697224617004,
"step": 532
},
{
"epoch": 0.36332651670074984,
"grad_norm": 3.6011276245117188,
"learning_rate": 1.8827170802202027e-05,
"loss": 1.4054481983184814,
"step": 533
},
{
"epoch": 0.36400817995910023,
"grad_norm": 129.46046447753906,
"learning_rate": 1.8821976494981055e-05,
"loss": 1.6898300647735596,
"step": 534
},
{
"epoch": 0.36468984321745057,
"grad_norm": 2.155585527420044,
"learning_rate": 1.8816771430515178e-05,
"loss": 0.7753575444221497,
"step": 535
},
{
"epoch": 0.36537150647580097,
"grad_norm": 4.523682117462158,
"learning_rate": 1.8811555615151286e-05,
"loss": 1.6820876598358154,
"step": 536
},
{
"epoch": 0.3660531697341513,
"grad_norm": 4.429269790649414,
"learning_rate": 1.880632905524939e-05,
"loss": 0.6383548378944397,
"step": 537
},
{
"epoch": 0.3667348329925017,
"grad_norm": 3.7983267307281494,
"learning_rate": 1.8801091757182593e-05,
"loss": 1.0961453914642334,
"step": 538
},
{
"epoch": 0.3674164962508521,
"grad_norm": 3.1742172241210938,
"learning_rate": 1.879584372733709e-05,
"loss": 1.4234809875488281,
"step": 539
},
{
"epoch": 0.36809815950920244,
"grad_norm": 2.4020187854766846,
"learning_rate": 1.8790584972112174e-05,
"loss": 1.1758843660354614,
"step": 540
},
{
"epoch": 0.36877982276755283,
"grad_norm": 2.9260194301605225,
"learning_rate": 1.87853154979202e-05,
"loss": 1.0781900882720947,
"step": 541
},
{
"epoch": 0.36946148602590323,
"grad_norm": 2.0199899673461914,
"learning_rate": 1.8780035311186605e-05,
"loss": 0.7699449062347412,
"step": 542
},
{
"epoch": 0.37014314928425357,
"grad_norm": 9.097823143005371,
"learning_rate": 1.8774744418349886e-05,
"loss": 1.7699912786483765,
"step": 543
},
{
"epoch": 0.37082481254260397,
"grad_norm": 3.701662302017212,
"learning_rate": 1.8769442825861594e-05,
"loss": 1.1131584644317627,
"step": 544
},
{
"epoch": 0.3715064758009543,
"grad_norm": 3.3982136249542236,
"learning_rate": 1.876413054018633e-05,
"loss": 1.6480752229690552,
"step": 545
},
{
"epoch": 0.3721881390593047,
"grad_norm": 2.128912925720215,
"learning_rate": 1.875880756780172e-05,
"loss": 1.0749809741973877,
"step": 546
},
{
"epoch": 0.3728698023176551,
"grad_norm": 2.3384532928466797,
"learning_rate": 1.8753473915198437e-05,
"loss": 1.4570391178131104,
"step": 547
},
{
"epoch": 0.37355146557600544,
"grad_norm": 2.2841551303863525,
"learning_rate": 1.874812958888018e-05,
"loss": 1.0104687213897705,
"step": 548
},
{
"epoch": 0.37423312883435583,
"grad_norm": 1.5464543104171753,
"learning_rate": 1.874277459536364e-05,
"loss": 1.2241737842559814,
"step": 549
},
{
"epoch": 0.37491479209270623,
"grad_norm": 5.0516815185546875,
"learning_rate": 1.873740894117854e-05,
"loss": 1.6643073558807373,
"step": 550
},
{
"epoch": 0.37559645535105657,
"grad_norm": 2.9442920684814453,
"learning_rate": 1.8732032632867592e-05,
"loss": 1.5176081657409668,
"step": 551
},
{
"epoch": 0.37627811860940696,
"grad_norm": 2.480600118637085,
"learning_rate": 1.8726645676986503e-05,
"loss": 1.3346457481384277,
"step": 552
},
{
"epoch": 0.3769597818677573,
"grad_norm": 2.3903183937072754,
"learning_rate": 1.872124808010395e-05,
"loss": 1.3713115453720093,
"step": 553
},
{
"epoch": 0.3776414451261077,
"grad_norm": 4.027533531188965,
"learning_rate": 1.8715839848801604e-05,
"loss": 0.8249767422676086,
"step": 554
},
{
"epoch": 0.3783231083844581,
"grad_norm": 2.3533899784088135,
"learning_rate": 1.8710420989674093e-05,
"loss": 1.4343526363372803,
"step": 555
},
{
"epoch": 0.37900477164280844,
"grad_norm": 2.4576408863067627,
"learning_rate": 1.8704991509329002e-05,
"loss": 1.4026904106140137,
"step": 556
},
{
"epoch": 0.37968643490115883,
"grad_norm": 3.576802968978882,
"learning_rate": 1.8699551414386877e-05,
"loss": 1.5657179355621338,
"step": 557
},
{
"epoch": 0.3803680981595092,
"grad_norm": 2.6988046169281006,
"learning_rate": 1.8694100711481195e-05,
"loss": 0.6028250455856323,
"step": 558
},
{
"epoch": 0.38104976141785957,
"grad_norm": 3.8934648036956787,
"learning_rate": 1.868863940725838e-05,
"loss": 1.2468817234039307,
"step": 559
},
{
"epoch": 0.38173142467620996,
"grad_norm": 2.7071168422698975,
"learning_rate": 1.8683167508377775e-05,
"loss": 0.9047738313674927,
"step": 560
},
{
"epoch": 0.3824130879345603,
"grad_norm": 3.0257182121276855,
"learning_rate": 1.8677685021511643e-05,
"loss": 1.2457003593444824,
"step": 561
},
{
"epoch": 0.3830947511929107,
"grad_norm": 4.956380367279053,
"learning_rate": 1.8672191953345156e-05,
"loss": 1.811165452003479,
"step": 562
},
{
"epoch": 0.3837764144512611,
"grad_norm": 6.11633825302124,
"learning_rate": 1.86666883105764e-05,
"loss": 1.6507594585418701,
"step": 563
},
{
"epoch": 0.38445807770961143,
"grad_norm": 4.5495991706848145,
"learning_rate": 1.866117409991634e-05,
"loss": 1.5494463443756104,
"step": 564
},
{
"epoch": 0.38513974096796183,
"grad_norm": 4.0347795486450195,
"learning_rate": 1.8655649328088836e-05,
"loss": 1.0738871097564697,
"step": 565
},
{
"epoch": 0.3858214042263122,
"grad_norm": 3.920830249786377,
"learning_rate": 1.865011400183062e-05,
"loss": 0.739644467830658,
"step": 566
},
{
"epoch": 0.38650306748466257,
"grad_norm": 4.003133296966553,
"learning_rate": 1.8644568127891303e-05,
"loss": 1.6497926712036133,
"step": 567
},
{
"epoch": 0.38718473074301296,
"grad_norm": 3.9445290565490723,
"learning_rate": 1.8639011713033347e-05,
"loss": 0.5134506225585938,
"step": 568
},
{
"epoch": 0.3878663940013633,
"grad_norm": 2.661447048187256,
"learning_rate": 1.8633444764032074e-05,
"loss": 1.0034466981887817,
"step": 569
},
{
"epoch": 0.3885480572597137,
"grad_norm": 2.3295094966888428,
"learning_rate": 1.862786728767565e-05,
"loss": 1.208392858505249,
"step": 570
},
{
"epoch": 0.3892297205180641,
"grad_norm": 2.4365479946136475,
"learning_rate": 1.8622279290765078e-05,
"loss": 0.5979467630386353,
"step": 571
},
{
"epoch": 0.38991138377641443,
"grad_norm": 5.058823108673096,
"learning_rate": 1.8616680780114183e-05,
"loss": 1.4158920049667358,
"step": 572
},
{
"epoch": 0.39059304703476483,
"grad_norm": 2.1014676094055176,
"learning_rate": 1.8611071762549623e-05,
"loss": 1.025521993637085,
"step": 573
},
{
"epoch": 0.3912747102931152,
"grad_norm": 2.405318260192871,
"learning_rate": 1.860545224491085e-05,
"loss": 0.9537474513053894,
"step": 574
},
{
"epoch": 0.39195637355146556,
"grad_norm": 4.968394756317139,
"learning_rate": 1.8599822234050143e-05,
"loss": 1.0375311374664307,
"step": 575
},
{
"epoch": 0.39263803680981596,
"grad_norm": 2.6036109924316406,
"learning_rate": 1.859418173683255e-05,
"loss": 1.193580150604248,
"step": 576
},
{
"epoch": 0.3933197000681663,
"grad_norm": 3.131683826446533,
"learning_rate": 1.858853076013593e-05,
"loss": 1.137315034866333,
"step": 577
},
{
"epoch": 0.3940013633265167,
"grad_norm": 4.054980278015137,
"learning_rate": 1.8582869310850903e-05,
"loss": 0.7231044173240662,
"step": 578
},
{
"epoch": 0.3946830265848671,
"grad_norm": 2.896749496459961,
"learning_rate": 1.8577197395880866e-05,
"loss": 0.8693374395370483,
"step": 579
},
{
"epoch": 0.39536468984321743,
"grad_norm": 3.2366976737976074,
"learning_rate": 1.8571515022141974e-05,
"loss": 0.869022786617279,
"step": 580
},
{
"epoch": 0.3960463531015678,
"grad_norm": 1.8894977569580078,
"learning_rate": 1.856582219656314e-05,
"loss": 0.7851641178131104,
"step": 581
},
{
"epoch": 0.3967280163599182,
"grad_norm": 1.6904579401016235,
"learning_rate": 1.856011892608602e-05,
"loss": 0.6581047773361206,
"step": 582
},
{
"epoch": 0.39740967961826856,
"grad_norm": 3.3967642784118652,
"learning_rate": 1.8554405217665004e-05,
"loss": 0.8824714422225952,
"step": 583
},
{
"epoch": 0.39809134287661896,
"grad_norm": 3.3108932971954346,
"learning_rate": 1.854868107826721e-05,
"loss": 1.6939976215362549,
"step": 584
},
{
"epoch": 0.3987730061349693,
"grad_norm": 5.848761558532715,
"learning_rate": 1.8542946514872478e-05,
"loss": 2.178783416748047,
"step": 585
},
{
"epoch": 0.3994546693933197,
"grad_norm": 3.158979654312134,
"learning_rate": 1.8537201534473353e-05,
"loss": 1.70957612991333,
"step": 586
},
{
"epoch": 0.4001363326516701,
"grad_norm": 8.649386405944824,
"learning_rate": 1.8531446144075093e-05,
"loss": 0.9513790607452393,
"step": 587
},
{
"epoch": 0.40081799591002043,
"grad_norm": 7.115112781524658,
"learning_rate": 1.852568035069564e-05,
"loss": 1.7480049133300781,
"step": 588
},
{
"epoch": 0.4014996591683708,
"grad_norm": 8.075335502624512,
"learning_rate": 1.8519904161365624e-05,
"loss": 1.7717616558074951,
"step": 589
},
{
"epoch": 0.4021813224267212,
"grad_norm": 8.434859275817871,
"learning_rate": 1.851411758312835e-05,
"loss": 1.2135674953460693,
"step": 590
},
{
"epoch": 0.40286298568507156,
"grad_norm": 1.8380115032196045,
"learning_rate": 1.8508320623039792e-05,
"loss": 1.348672866821289,
"step": 591
},
{
"epoch": 0.40354464894342196,
"grad_norm": 2.730818033218384,
"learning_rate": 1.8502513288168584e-05,
"loss": 1.261131763458252,
"step": 592
},
{
"epoch": 0.4042263122017723,
"grad_norm": 16.677165985107422,
"learning_rate": 1.8496695585596013e-05,
"loss": 1.3604838848114014,
"step": 593
},
{
"epoch": 0.4049079754601227,
"grad_norm": 7.098438262939453,
"learning_rate": 1.8490867522416e-05,
"loss": 1.7277930974960327,
"step": 594
},
{
"epoch": 0.4055896387184731,
"grad_norm": 3.663721799850464,
"learning_rate": 1.8485029105735112e-05,
"loss": 1.1387404203414917,
"step": 595
},
{
"epoch": 0.40627130197682343,
"grad_norm": 3.6738693714141846,
"learning_rate": 1.8479180342672525e-05,
"loss": 0.8586311340332031,
"step": 596
},
{
"epoch": 0.4069529652351738,
"grad_norm": 4.40339469909668,
"learning_rate": 1.8473321240360048e-05,
"loss": 1.3938498497009277,
"step": 597
},
{
"epoch": 0.4076346284935242,
"grad_norm": 3.4013774394989014,
"learning_rate": 1.846745180594208e-05,
"loss": 1.429935097694397,
"step": 598
},
{
"epoch": 0.40831629175187456,
"grad_norm": 2.294057607650757,
"learning_rate": 1.8461572046575638e-05,
"loss": 1.4736216068267822,
"step": 599
},
{
"epoch": 0.40899795501022496,
"grad_norm": 5.40059232711792,
"learning_rate": 1.8455681969430307e-05,
"loss": 1.7445282936096191,
"step": 600
},
{
"epoch": 0.4096796182685753,
"grad_norm": 8.884811401367188,
"learning_rate": 1.8449781581688274e-05,
"loss": 3.040208578109741,
"step": 601
},
{
"epoch": 0.4103612815269257,
"grad_norm": 5.935024261474609,
"learning_rate": 1.8443870890544287e-05,
"loss": 0.8014417886734009,
"step": 602
},
{
"epoch": 0.4110429447852761,
"grad_norm": 3.7073605060577393,
"learning_rate": 1.8437949903205657e-05,
"loss": 0.7055726647377014,
"step": 603
},
{
"epoch": 0.4117246080436264,
"grad_norm": 3.1601369380950928,
"learning_rate": 1.843201862689225e-05,
"loss": 0.5688856840133667,
"step": 604
},
{
"epoch": 0.4124062713019768,
"grad_norm": 3.3272056579589844,
"learning_rate": 1.8426077068836487e-05,
"loss": 1.3306844234466553,
"step": 605
},
{
"epoch": 0.4130879345603272,
"grad_norm": 2.028167247772217,
"learning_rate": 1.842012523628332e-05,
"loss": 0.9651132225990295,
"step": 606
},
{
"epoch": 0.41376959781867756,
"grad_norm": 3.8314738273620605,
"learning_rate": 1.8414163136490224e-05,
"loss": 1.5358803272247314,
"step": 607
},
{
"epoch": 0.41445126107702795,
"grad_norm": 6.4982733726501465,
"learning_rate": 1.84081907767272e-05,
"loss": 2.0918755531311035,
"step": 608
},
{
"epoch": 0.41513292433537835,
"grad_norm": 1.8055447340011597,
"learning_rate": 1.840220816427676e-05,
"loss": 0.6877099275588989,
"step": 609
},
{
"epoch": 0.4158145875937287,
"grad_norm": 2.6051347255706787,
"learning_rate": 1.839621530643392e-05,
"loss": 1.0851391553878784,
"step": 610
},
{
"epoch": 0.4164962508520791,
"grad_norm": 4.730994701385498,
"learning_rate": 1.839021221050618e-05,
"loss": 1.643237829208374,
"step": 611
},
{
"epoch": 0.4171779141104294,
"grad_norm": 2.420578956604004,
"learning_rate": 1.838419888381353e-05,
"loss": 0.8266706466674805,
"step": 612
},
{
"epoch": 0.4178595773687798,
"grad_norm": 2.6848230361938477,
"learning_rate": 1.8378175333688438e-05,
"loss": 0.33754855394363403,
"step": 613
},
{
"epoch": 0.4185412406271302,
"grad_norm": 4.421104431152344,
"learning_rate": 1.837214156747583e-05,
"loss": 1.3408927917480469,
"step": 614
},
{
"epoch": 0.41922290388548056,
"grad_norm": 5.390198707580566,
"learning_rate": 1.8366097592533095e-05,
"loss": 0.5011366009712219,
"step": 615
},
{
"epoch": 0.41990456714383095,
"grad_norm": 3.1726467609405518,
"learning_rate": 1.8360043416230067e-05,
"loss": 1.6905428171157837,
"step": 616
},
{
"epoch": 0.42058623040218135,
"grad_norm": 2.5729990005493164,
"learning_rate": 1.8353979045949023e-05,
"loss": 1.076096534729004,
"step": 617
},
{
"epoch": 0.4212678936605317,
"grad_norm": 3.2037148475646973,
"learning_rate": 1.834790448908467e-05,
"loss": 0.5505931377410889,
"step": 618
},
{
"epoch": 0.4219495569188821,
"grad_norm": 3.306154251098633,
"learning_rate": 1.8341819753044135e-05,
"loss": 0.9625562429428101,
"step": 619
},
{
"epoch": 0.4226312201772324,
"grad_norm": 2.3715250492095947,
"learning_rate": 1.8335724845246948e-05,
"loss": 0.53058260679245,
"step": 620
},
{
"epoch": 0.4233128834355828,
"grad_norm": 4.0207839012146,
"learning_rate": 1.8329619773125064e-05,
"loss": 1.3297679424285889,
"step": 621
},
{
"epoch": 0.4239945466939332,
"grad_norm": 6.694503307342529,
"learning_rate": 1.832350454412281e-05,
"loss": 2.504713296890259,
"step": 622
},
{
"epoch": 0.42467620995228356,
"grad_norm": 3.3091583251953125,
"learning_rate": 1.8317379165696908e-05,
"loss": 1.2117489576339722,
"step": 623
},
{
"epoch": 0.42535787321063395,
"grad_norm": 2.701916217803955,
"learning_rate": 1.8311243645316458e-05,
"loss": 0.6958295106887817,
"step": 624
},
{
"epoch": 0.42603953646898435,
"grad_norm": 1.5155378580093384,
"learning_rate": 1.830509799046292e-05,
"loss": 0.8749496936798096,
"step": 625
},
{
"epoch": 0.4267211997273347,
"grad_norm": 1.740338683128357,
"learning_rate": 1.829894220863012e-05,
"loss": 0.7681260108947754,
"step": 626
},
{
"epoch": 0.4274028629856851,
"grad_norm": 3.138975143432617,
"learning_rate": 1.8292776307324217e-05,
"loss": 1.421783685684204,
"step": 627
},
{
"epoch": 0.4280845262440354,
"grad_norm": 2.65615177154541,
"learning_rate": 1.8286600294063732e-05,
"loss": 0.5015825629234314,
"step": 628
},
{
"epoch": 0.4287661895023858,
"grad_norm": 3.291783094406128,
"learning_rate": 1.82804141763795e-05,
"loss": 1.7397900819778442,
"step": 629
},
{
"epoch": 0.4294478527607362,
"grad_norm": 6.939148426055908,
"learning_rate": 1.8274217961814682e-05,
"loss": 1.7783665657043457,
"step": 630
},
{
"epoch": 0.43012951601908656,
"grad_norm": 2.345168352127075,
"learning_rate": 1.8268011657924746e-05,
"loss": 0.9429031610488892,
"step": 631
},
{
"epoch": 0.43081117927743695,
"grad_norm": 2.665942907333374,
"learning_rate": 1.8261795272277472e-05,
"loss": 0.8730664849281311,
"step": 632
},
{
"epoch": 0.43149284253578735,
"grad_norm": 2.2719271183013916,
"learning_rate": 1.8255568812452923e-05,
"loss": 1.1947182416915894,
"step": 633
},
{
"epoch": 0.4321745057941377,
"grad_norm": 3.0505599975585938,
"learning_rate": 1.8249332286043456e-05,
"loss": 1.4211301803588867,
"step": 634
},
{
"epoch": 0.4328561690524881,
"grad_norm": 1.8321754932403564,
"learning_rate": 1.8243085700653698e-05,
"loss": 1.4799766540527344,
"step": 635
},
{
"epoch": 0.4335378323108384,
"grad_norm": 1.7265902757644653,
"learning_rate": 1.8236829063900535e-05,
"loss": 1.4023617506027222,
"step": 636
},
{
"epoch": 0.4342194955691888,
"grad_norm": 2.2862470149993896,
"learning_rate": 1.8230562383413127e-05,
"loss": 1.5263274908065796,
"step": 637
},
{
"epoch": 0.4349011588275392,
"grad_norm": 2.731217384338379,
"learning_rate": 1.822428566683286e-05,
"loss": 1.7954682111740112,
"step": 638
},
{
"epoch": 0.43558282208588955,
"grad_norm": 2.999441385269165,
"learning_rate": 1.8217998921813375e-05,
"loss": 0.7107315063476562,
"step": 639
},
{
"epoch": 0.43626448534423995,
"grad_norm": 2.7480056285858154,
"learning_rate": 1.821170215602053e-05,
"loss": 1.6439075469970703,
"step": 640
},
{
"epoch": 0.43694614860259035,
"grad_norm": 5.756744384765625,
"learning_rate": 1.8205395377132407e-05,
"loss": 0.6778331995010376,
"step": 641
},
{
"epoch": 0.4376278118609407,
"grad_norm": 2.9490387439727783,
"learning_rate": 1.81990785928393e-05,
"loss": 1.2749234437942505,
"step": 642
},
{
"epoch": 0.4383094751192911,
"grad_norm": 2.444660186767578,
"learning_rate": 1.8192751810843697e-05,
"loss": 1.1050373315811157,
"step": 643
},
{
"epoch": 0.4389911383776414,
"grad_norm": 2.1506237983703613,
"learning_rate": 1.8186415038860276e-05,
"loss": 0.3917778432369232,
"step": 644
},
{
"epoch": 0.4396728016359918,
"grad_norm": 2.1962451934814453,
"learning_rate": 1.818006828461591e-05,
"loss": 0.9240995645523071,
"step": 645
},
{
"epoch": 0.4403544648943422,
"grad_norm": 4.333210468292236,
"learning_rate": 1.8173711555849626e-05,
"loss": 1.3042209148406982,
"step": 646
},
{
"epoch": 0.44103612815269255,
"grad_norm": 3.983182191848755,
"learning_rate": 1.8167344860312627e-05,
"loss": 1.383401870727539,
"step": 647
},
{
"epoch": 0.44171779141104295,
"grad_norm": 1.7078006267547607,
"learning_rate": 1.8160968205768264e-05,
"loss": 0.7653818130493164,
"step": 648
},
{
"epoch": 0.44239945466939334,
"grad_norm": 1.8697181940078735,
"learning_rate": 1.815458159999203e-05,
"loss": 0.8399191498756409,
"step": 649
},
{
"epoch": 0.4430811179277437,
"grad_norm": 3.080517053604126,
"learning_rate": 1.8148185050771554e-05,
"loss": 1.2301526069641113,
"step": 650
},
{
"epoch": 0.4437627811860941,
"grad_norm": 4.8597588539123535,
"learning_rate": 1.8141778565906594e-05,
"loss": 1.6627204418182373,
"step": 651
},
{
"epoch": 0.4444444444444444,
"grad_norm": 4.034822940826416,
"learning_rate": 1.8135362153209014e-05,
"loss": 1.291428565979004,
"step": 652
},
{
"epoch": 0.4451261077027948,
"grad_norm": 4.153934955596924,
"learning_rate": 1.8128935820502792e-05,
"loss": 1.1491031646728516,
"step": 653
},
{
"epoch": 0.4458077709611452,
"grad_norm": 3.6002509593963623,
"learning_rate": 1.8122499575624e-05,
"loss": 1.6600425243377686,
"step": 654
},
{
"epoch": 0.44648943421949555,
"grad_norm": 1.543675184249878,
"learning_rate": 1.8116053426420793e-05,
"loss": 0.8339579105377197,
"step": 655
},
{
"epoch": 0.44717109747784595,
"grad_norm": 2.568967819213867,
"learning_rate": 1.8109597380753404e-05,
"loss": 0.8489872813224792,
"step": 656
},
{
"epoch": 0.44785276073619634,
"grad_norm": 3.282658815383911,
"learning_rate": 1.8103131446494144e-05,
"loss": 1.2508018016815186,
"step": 657
},
{
"epoch": 0.4485344239945467,
"grad_norm": 3.4995203018188477,
"learning_rate": 1.8096655631527365e-05,
"loss": 1.5129706859588623,
"step": 658
},
{
"epoch": 0.4492160872528971,
"grad_norm": 2.7602193355560303,
"learning_rate": 1.8090169943749477e-05,
"loss": 1.2756472826004028,
"step": 659
},
{
"epoch": 0.4498977505112474,
"grad_norm": 11.795307159423828,
"learning_rate": 1.8083674391068925e-05,
"loss": 1.1611485481262207,
"step": 660
},
{
"epoch": 0.4505794137695978,
"grad_norm": 1.9726192951202393,
"learning_rate": 1.807716898140619e-05,
"loss": 0.83235764503479,
"step": 661
},
{
"epoch": 0.4512610770279482,
"grad_norm": 2.960864543914795,
"learning_rate": 1.807065372269376e-05,
"loss": 0.7254927158355713,
"step": 662
},
{
"epoch": 0.45194274028629855,
"grad_norm": 2.131723165512085,
"learning_rate": 1.8064128622876146e-05,
"loss": 0.8774275183677673,
"step": 663
},
{
"epoch": 0.45262440354464895,
"grad_norm": 3.856238603591919,
"learning_rate": 1.805759368990985e-05,
"loss": 0.6111989617347717,
"step": 664
},
{
"epoch": 0.45330606680299934,
"grad_norm": 2.0575990676879883,
"learning_rate": 1.8051048931763366e-05,
"loss": 1.1646723747253418,
"step": 665
},
{
"epoch": 0.4539877300613497,
"grad_norm": 2.3782901763916016,
"learning_rate": 1.804449435641717e-05,
"loss": 1.2033634185791016,
"step": 666
},
{
"epoch": 0.4546693933197001,
"grad_norm": 3.0704290866851807,
"learning_rate": 1.803792997186371e-05,
"loss": 1.3781287670135498,
"step": 667
},
{
"epoch": 0.4553510565780504,
"grad_norm": 3.466710090637207,
"learning_rate": 1.803135578610739e-05,
"loss": 0.7485005259513855,
"step": 668
},
{
"epoch": 0.4560327198364008,
"grad_norm": 2.91351580619812,
"learning_rate": 1.802477180716457e-05,
"loss": 1.451478362083435,
"step": 669
},
{
"epoch": 0.4567143830947512,
"grad_norm": 2.3317689895629883,
"learning_rate": 1.8018178043063554e-05,
"loss": 1.2862257957458496,
"step": 670
},
{
"epoch": 0.45739604635310155,
"grad_norm": 2.3442463874816895,
"learning_rate": 1.801157450184457e-05,
"loss": 1.1389061212539673,
"step": 671
},
{
"epoch": 0.45807770961145194,
"grad_norm": 2.2962794303894043,
"learning_rate": 1.8004961191559765e-05,
"loss": 0.920241117477417,
"step": 672
},
{
"epoch": 0.45875937286980234,
"grad_norm": 8.112113952636719,
"learning_rate": 1.7998338120273218e-05,
"loss": 0.3879634439945221,
"step": 673
},
{
"epoch": 0.4594410361281527,
"grad_norm": 2.143040657043457,
"learning_rate": 1.7991705296060888e-05,
"loss": 1.3585686683654785,
"step": 674
},
{
"epoch": 0.4601226993865031,
"grad_norm": 6.662731647491455,
"learning_rate": 1.798506272701064e-05,
"loss": 1.844017505645752,
"step": 675
},
{
"epoch": 0.4608043626448534,
"grad_norm": 2.191220760345459,
"learning_rate": 1.797841042122221e-05,
"loss": 1.460452675819397,
"step": 676
},
{
"epoch": 0.4614860259032038,
"grad_norm": 3.355347156524658,
"learning_rate": 1.797174838680722e-05,
"loss": 1.8264518976211548,
"step": 677
},
{
"epoch": 0.4621676891615542,
"grad_norm": 1.9022427797317505,
"learning_rate": 1.7965076631889146e-05,
"loss": 1.0955109596252441,
"step": 678
},
{
"epoch": 0.46284935241990455,
"grad_norm": 2.4177889823913574,
"learning_rate": 1.7958395164603323e-05,
"loss": 1.0090279579162598,
"step": 679
},
{
"epoch": 0.46353101567825494,
"grad_norm": 3.373490571975708,
"learning_rate": 1.795170399309692e-05,
"loss": 1.6577601432800293,
"step": 680
},
{
"epoch": 0.46421267893660534,
"grad_norm": 2.119124174118042,
"learning_rate": 1.794500312552895e-05,
"loss": 1.2227559089660645,
"step": 681
},
{
"epoch": 0.4648943421949557,
"grad_norm": 4.674558639526367,
"learning_rate": 1.7938292570070238e-05,
"loss": 1.5070548057556152,
"step": 682
},
{
"epoch": 0.4655760054533061,
"grad_norm": 4.290884494781494,
"learning_rate": 1.7931572334903427e-05,
"loss": 0.7182972431182861,
"step": 683
},
{
"epoch": 0.4662576687116564,
"grad_norm": 2.9920401573181152,
"learning_rate": 1.792484242822297e-05,
"loss": 1.6099597215652466,
"step": 684
},
{
"epoch": 0.4669393319700068,
"grad_norm": 2.643561840057373,
"learning_rate": 1.7918102858235103e-05,
"loss": 0.6941257119178772,
"step": 685
},
{
"epoch": 0.4676209952283572,
"grad_norm": 5.84933614730835,
"learning_rate": 1.7911353633157844e-05,
"loss": 0.5451070070266724,
"step": 686
},
{
"epoch": 0.46830265848670755,
"grad_norm": 2.0181429386138916,
"learning_rate": 1.7904594761221e-05,
"loss": 1.117553949356079,
"step": 687
},
{
"epoch": 0.46898432174505794,
"grad_norm": 2.363997459411621,
"learning_rate": 1.789782625066612e-05,
"loss": 1.1420124769210815,
"step": 688
},
{
"epoch": 0.46966598500340834,
"grad_norm": 8.58547592163086,
"learning_rate": 1.7891048109746522e-05,
"loss": 1.1073362827301025,
"step": 689
},
{
"epoch": 0.4703476482617587,
"grad_norm": 6.624891757965088,
"learning_rate": 1.7884260346727257e-05,
"loss": 1.8669713735580444,
"step": 690
},
{
"epoch": 0.4710293115201091,
"grad_norm": 4.729033470153809,
"learning_rate": 1.7877462969885114e-05,
"loss": 1.9904009103775024,
"step": 691
},
{
"epoch": 0.47171097477845947,
"grad_norm": 1.7405580282211304,
"learning_rate": 1.7870655987508613e-05,
"loss": 0.7781103849411011,
"step": 692
},
{
"epoch": 0.4723926380368098,
"grad_norm": 3.082596778869629,
"learning_rate": 1.7863839407897962e-05,
"loss": 1.954369306564331,
"step": 693
},
{
"epoch": 0.4730743012951602,
"grad_norm": 2.463020086288452,
"learning_rate": 1.7857013239365098e-05,
"loss": 0.8323026299476624,
"step": 694
},
{
"epoch": 0.47375596455351054,
"grad_norm": 1.7679860591888428,
"learning_rate": 1.7850177490233635e-05,
"loss": 0.9151184558868408,
"step": 695
},
{
"epoch": 0.47443762781186094,
"grad_norm": 3.689037799835205,
"learning_rate": 1.784333216883887e-05,
"loss": 1.7770330905914307,
"step": 696
},
{
"epoch": 0.47511929107021134,
"grad_norm": 3.824721097946167,
"learning_rate": 1.7836477283527787e-05,
"loss": 0.8589752912521362,
"step": 697
},
{
"epoch": 0.4758009543285617,
"grad_norm": 3.058713674545288,
"learning_rate": 1.782961284265901e-05,
"loss": 1.1384007930755615,
"step": 698
},
{
"epoch": 0.47648261758691207,
"grad_norm": 2.086995840072632,
"learning_rate": 1.7822738854602835e-05,
"loss": 0.9110372066497803,
"step": 699
},
{
"epoch": 0.47716428084526247,
"grad_norm": 2.521082639694214,
"learning_rate": 1.7815855327741185e-05,
"loss": 1.4321343898773193,
"step": 700
},
{
"epoch": 0.4778459441036128,
"grad_norm": 3.35274338722229,
"learning_rate": 1.780896227046762e-05,
"loss": 1.0381226539611816,
"step": 701
},
{
"epoch": 0.4785276073619632,
"grad_norm": 2.294340133666992,
"learning_rate": 1.7802059691187316e-05,
"loss": 1.3280988931655884,
"step": 702
},
{
"epoch": 0.47920927062031354,
"grad_norm": 4.459254741668701,
"learning_rate": 1.7795147598317067e-05,
"loss": 1.4365578889846802,
"step": 703
},
{
"epoch": 0.47989093387866394,
"grad_norm": 2.573671579360962,
"learning_rate": 1.7788226000285272e-05,
"loss": 0.9661579132080078,
"step": 704
},
{
"epoch": 0.48057259713701433,
"grad_norm": 3.37911319732666,
"learning_rate": 1.7781294905531908e-05,
"loss": 1.8097542524337769,
"step": 705
},
{
"epoch": 0.4812542603953647,
"grad_norm": 2.7699577808380127,
"learning_rate": 1.7774354322508535e-05,
"loss": 1.1084448099136353,
"step": 706
},
{
"epoch": 0.48193592365371507,
"grad_norm": 4.020786762237549,
"learning_rate": 1.776740425967829e-05,
"loss": 0.7846464514732361,
"step": 707
},
{
"epoch": 0.48261758691206547,
"grad_norm": 1.7012214660644531,
"learning_rate": 1.7760444725515856e-05,
"loss": 1.0971169471740723,
"step": 708
},
{
"epoch": 0.4832992501704158,
"grad_norm": 2.0524704456329346,
"learning_rate": 1.775347572850748e-05,
"loss": 0.6552506685256958,
"step": 709
},
{
"epoch": 0.4839809134287662,
"grad_norm": 2.792614698410034,
"learning_rate": 1.774649727715094e-05,
"loss": 0.76331627368927,
"step": 710
},
{
"epoch": 0.48466257668711654,
"grad_norm": 2.1224751472473145,
"learning_rate": 1.7739509379955548e-05,
"loss": 1.1910077333450317,
"step": 711
},
{
"epoch": 0.48534423994546694,
"grad_norm": 6.945074558258057,
"learning_rate": 1.7732512045442125e-05,
"loss": 1.9449994564056396,
"step": 712
},
{
"epoch": 0.48602590320381733,
"grad_norm": 7.90489387512207,
"learning_rate": 1.7725505282142997e-05,
"loss": 2.0441508293151855,
"step": 713
},
{
"epoch": 0.4867075664621677,
"grad_norm": 3.1438591480255127,
"learning_rate": 1.771848909860201e-05,
"loss": 1.3289283514022827,
"step": 714
},
{
"epoch": 0.48738922972051807,
"grad_norm": 4.14089822769165,
"learning_rate": 1.7711463503374466e-05,
"loss": 1.4352498054504395,
"step": 715
},
{
"epoch": 0.48807089297886846,
"grad_norm": 5.588857650756836,
"learning_rate": 1.7704428505027165e-05,
"loss": 1.6422855854034424,
"step": 716
},
{
"epoch": 0.4887525562372188,
"grad_norm": 5.233851432800293,
"learning_rate": 1.7697384112138367e-05,
"loss": 1.6709481477737427,
"step": 717
},
{
"epoch": 0.4894342194955692,
"grad_norm": 6.2413649559021,
"learning_rate": 1.769033033329778e-05,
"loss": 1.1305233240127563,
"step": 718
},
{
"epoch": 0.49011588275391954,
"grad_norm": 5.878918647766113,
"learning_rate": 1.7683267177106573e-05,
"loss": 0.5765758156776428,
"step": 719
},
{
"epoch": 0.49079754601226994,
"grad_norm": 3.9863340854644775,
"learning_rate": 1.7676194652177333e-05,
"loss": 1.8041818141937256,
"step": 720
},
{
"epoch": 0.49147920927062033,
"grad_norm": 2.9146947860717773,
"learning_rate": 1.7669112767134084e-05,
"loss": 1.0354045629501343,
"step": 721
},
{
"epoch": 0.49216087252897067,
"grad_norm": 4.547295570373535,
"learning_rate": 1.766202153061225e-05,
"loss": 1.101534128189087,
"step": 722
},
{
"epoch": 0.49284253578732107,
"grad_norm": 5.544919490814209,
"learning_rate": 1.7654920951258668e-05,
"loss": 0.9572349786758423,
"step": 723
},
{
"epoch": 0.49352419904567146,
"grad_norm": 3.9510440826416016,
"learning_rate": 1.7647811037731565e-05,
"loss": 0.8672992587089539,
"step": 724
},
{
"epoch": 0.4942058623040218,
"grad_norm": 4.034142017364502,
"learning_rate": 1.764069179870055e-05,
"loss": 0.8420158624649048,
"step": 725
},
{
"epoch": 0.4948875255623722,
"grad_norm": 4.926695346832275,
"learning_rate": 1.76335632428466e-05,
"loss": 1.3785490989685059,
"step": 726
},
{
"epoch": 0.49556918882072254,
"grad_norm": 6.231317043304443,
"learning_rate": 1.762642537886206e-05,
"loss": 1.6120049953460693,
"step": 727
},
{
"epoch": 0.49625085207907293,
"grad_norm": 2.347989320755005,
"learning_rate": 1.7619278215450615e-05,
"loss": 0.36018455028533936,
"step": 728
},
{
"epoch": 0.49693251533742333,
"grad_norm": 2.300482988357544,
"learning_rate": 1.76121217613273e-05,
"loss": 0.7264208793640137,
"step": 729
},
{
"epoch": 0.49761417859577367,
"grad_norm": 3.5808284282684326,
"learning_rate": 1.760495602521847e-05,
"loss": 1.2240182161331177,
"step": 730
},
{
"epoch": 0.49829584185412407,
"grad_norm": 2.3418526649475098,
"learning_rate": 1.7597781015861797e-05,
"loss": 0.4540298581123352,
"step": 731
},
{
"epoch": 0.49897750511247446,
"grad_norm": 5.1293864250183105,
"learning_rate": 1.7590596742006276e-05,
"loss": 0.9630584716796875,
"step": 732
},
{
"epoch": 0.4996591683708248,
"grad_norm": 4.265676975250244,
"learning_rate": 1.7583403212412183e-05,
"loss": 2.012369394302368,
"step": 733
},
{
"epoch": 0.5003408316291752,
"grad_norm": 14.98397445678711,
"learning_rate": 1.7576200435851082e-05,
"loss": 1.671754002571106,
"step": 734
},
{
"epoch": 0.5010224948875256,
"grad_norm": 6.400186538696289,
"learning_rate": 1.756898842110582e-05,
"loss": 2.1458077430725098,
"step": 735
},
{
"epoch": 0.501704158145876,
"grad_norm": 5.821322917938232,
"learning_rate": 1.75617671769705e-05,
"loss": 0.9948315620422363,
"step": 736
},
{
"epoch": 0.5023858214042263,
"grad_norm": 4.912761211395264,
"learning_rate": 1.7554536712250488e-05,
"loss": 0.9431995153427124,
"step": 737
},
{
"epoch": 0.5030674846625767,
"grad_norm": 8.251131057739258,
"learning_rate": 1.7547297035762387e-05,
"loss": 1.3028041124343872,
"step": 738
},
{
"epoch": 0.5037491479209271,
"grad_norm": 4.36862325668335,
"learning_rate": 1.7540048156334035e-05,
"loss": 0.7506901621818542,
"step": 739
},
{
"epoch": 0.5044308111792775,
"grad_norm": 3.1437900066375732,
"learning_rate": 1.753279008280449e-05,
"loss": 1.8270992040634155,
"step": 740
},
{
"epoch": 0.5051124744376279,
"grad_norm": 1.6939406394958496,
"learning_rate": 1.7525522824024023e-05,
"loss": 0.8369800448417664,
"step": 741
},
{
"epoch": 0.5057941376959781,
"grad_norm": 5.759995460510254,
"learning_rate": 1.75182463888541e-05,
"loss": 1.7431058883666992,
"step": 742
},
{
"epoch": 0.5064758009543285,
"grad_norm": 3.341987133026123,
"learning_rate": 1.751096078616739e-05,
"loss": 1.688689947128296,
"step": 743
},
{
"epoch": 0.5071574642126789,
"grad_norm": 2.5472590923309326,
"learning_rate": 1.7503666024847722e-05,
"loss": 0.9975855350494385,
"step": 744
},
{
"epoch": 0.5078391274710293,
"grad_norm": 3.2179603576660156,
"learning_rate": 1.74963621137901e-05,
"loss": 0.6496344804763794,
"step": 745
},
{
"epoch": 0.5085207907293797,
"grad_norm": 5.027393817901611,
"learning_rate": 1.7489049061900702e-05,
"loss": 1.0913028717041016,
"step": 746
},
{
"epoch": 0.50920245398773,
"grad_norm": 3.2983977794647217,
"learning_rate": 1.7481726878096824e-05,
"loss": 0.8632484674453735,
"step": 747
},
{
"epoch": 0.5098841172460804,
"grad_norm": 2.075984239578247,
"learning_rate": 1.7474395571306914e-05,
"loss": 0.8893939256668091,
"step": 748
},
{
"epoch": 0.5105657805044308,
"grad_norm": 3.399078130722046,
"learning_rate": 1.746705515047054e-05,
"loss": 0.7946146130561829,
"step": 749
},
{
"epoch": 0.5112474437627812,
"grad_norm": 3.5610105991363525,
"learning_rate": 1.7459705624538383e-05,
"loss": 1.0715913772583008,
"step": 750
},
{
"epoch": 0.5119291070211316,
"grad_norm": 2.419004440307617,
"learning_rate": 1.745234700247223e-05,
"loss": 1.079657793045044,
"step": 751
},
{
"epoch": 0.512610770279482,
"grad_norm": 4.799737930297852,
"learning_rate": 1.7444979293244953e-05,
"loss": 1.1214956045150757,
"step": 752
},
{
"epoch": 0.5132924335378323,
"grad_norm": 2.3297369480133057,
"learning_rate": 1.7437602505840513e-05,
"loss": 1.031355857849121,
"step": 753
},
{
"epoch": 0.5139740967961827,
"grad_norm": 1.1967624425888062,
"learning_rate": 1.7430216649253934e-05,
"loss": 0.6558365821838379,
"step": 754
},
{
"epoch": 0.5146557600545331,
"grad_norm": 2.59171462059021,
"learning_rate": 1.7422821732491297e-05,
"loss": 0.8954111337661743,
"step": 755
},
{
"epoch": 0.5153374233128835,
"grad_norm": 3.414041519165039,
"learning_rate": 1.741541776456974e-05,
"loss": 1.0040534734725952,
"step": 756
},
{
"epoch": 0.5160190865712339,
"grad_norm": 3.0361270904541016,
"learning_rate": 1.7408004754517428e-05,
"loss": 0.7500423789024353,
"step": 757
},
{
"epoch": 0.5167007498295841,
"grad_norm": 3.6555752754211426,
"learning_rate": 1.7400582711373558e-05,
"loss": 0.7112274169921875,
"step": 758
},
{
"epoch": 0.5173824130879345,
"grad_norm": 5.004818916320801,
"learning_rate": 1.739315164418834e-05,
"loss": 1.391385555267334,
"step": 759
},
{
"epoch": 0.5180640763462849,
"grad_norm": 3.7263758182525635,
"learning_rate": 1.7385711562022988e-05,
"loss": 1.5679736137390137,
"step": 760
},
{
"epoch": 0.5187457396046353,
"grad_norm": 4.171016693115234,
"learning_rate": 1.7378262473949705e-05,
"loss": 1.539245843887329,
"step": 761
},
{
"epoch": 0.5194274028629857,
"grad_norm": 2.8865065574645996,
"learning_rate": 1.737080438905168e-05,
"loss": 0.8579041957855225,
"step": 762
},
{
"epoch": 0.520109066121336,
"grad_norm": 2.8619775772094727,
"learning_rate": 1.736333731642307e-05,
"loss": 1.6193718910217285,
"step": 763
},
{
"epoch": 0.5207907293796864,
"grad_norm": 3.0032424926757812,
"learning_rate": 1.735586126516899e-05,
"loss": 1.8853137493133545,
"step": 764
},
{
"epoch": 0.5214723926380368,
"grad_norm": 3.6113500595092773,
"learning_rate": 1.734837624440551e-05,
"loss": 1.374289870262146,
"step": 765
},
{
"epoch": 0.5221540558963872,
"grad_norm": 4.458436489105225,
"learning_rate": 1.734088226325963e-05,
"loss": 0.8771038055419922,
"step": 766
},
{
"epoch": 0.5228357191547376,
"grad_norm": 3.736870050430298,
"learning_rate": 1.733337933086928e-05,
"loss": 1.342976450920105,
"step": 767
},
{
"epoch": 0.523517382413088,
"grad_norm": 3.629173755645752,
"learning_rate": 1.7325867456383303e-05,
"loss": 1.500548243522644,
"step": 768
},
{
"epoch": 0.5241990456714383,
"grad_norm": 2.3277621269226074,
"learning_rate": 1.7318346648961444e-05,
"loss": 0.8789968490600586,
"step": 769
},
{
"epoch": 0.5248807089297887,
"grad_norm": 3.080993175506592,
"learning_rate": 1.731081691777434e-05,
"loss": 1.8359971046447754,
"step": 770
},
{
"epoch": 0.5255623721881391,
"grad_norm": 2.8816323280334473,
"learning_rate": 1.7303278272003524e-05,
"loss": 1.0419747829437256,
"step": 771
},
{
"epoch": 0.5262440354464895,
"grad_norm": 2.5966172218322754,
"learning_rate": 1.7295730720841372e-05,
"loss": 0.6071522235870361,
"step": 772
},
{
"epoch": 0.5269256987048399,
"grad_norm": 2.166282892227173,
"learning_rate": 1.7288174273491144e-05,
"loss": 1.092464566230774,
"step": 773
},
{
"epoch": 0.5276073619631901,
"grad_norm": 2.050143241882324,
"learning_rate": 1.7280608939166937e-05,
"loss": 0.3647843599319458,
"step": 774
},
{
"epoch": 0.5282890252215405,
"grad_norm": 4.108394622802734,
"learning_rate": 1.7273034727093677e-05,
"loss": 1.2300772666931152,
"step": 775
},
{
"epoch": 0.5289706884798909,
"grad_norm": 3.4278016090393066,
"learning_rate": 1.726545164650714e-05,
"loss": 1.1582565307617188,
"step": 776
},
{
"epoch": 0.5296523517382413,
"grad_norm": 3.606982707977295,
"learning_rate": 1.725785970665388e-05,
"loss": 1.1168954372406006,
"step": 777
},
{
"epoch": 0.5303340149965917,
"grad_norm": 1.5078668594360352,
"learning_rate": 1.7250258916791286e-05,
"loss": 0.8669070601463318,
"step": 778
},
{
"epoch": 0.5310156782549421,
"grad_norm": 2.591376543045044,
"learning_rate": 1.7242649286187524e-05,
"loss": 0.9776431322097778,
"step": 779
},
{
"epoch": 0.5316973415132924,
"grad_norm": 3.464742422103882,
"learning_rate": 1.7235030824121542e-05,
"loss": 1.2853455543518066,
"step": 780
},
{
"epoch": 0.5323790047716428,
"grad_norm": 5.493305683135986,
"learning_rate": 1.722740353988305e-05,
"loss": 1.8691112995147705,
"step": 781
},
{
"epoch": 0.5330606680299932,
"grad_norm": 2.410437822341919,
"learning_rate": 1.721976744277253e-05,
"loss": 0.814149796962738,
"step": 782
},
{
"epoch": 0.5337423312883436,
"grad_norm": 16.686614990234375,
"learning_rate": 1.7212122542101202e-05,
"loss": 0.375972181558609,
"step": 783
},
{
"epoch": 0.534423994546694,
"grad_norm": 2.9362881183624268,
"learning_rate": 1.7204468847191017e-05,
"loss": 0.7596120834350586,
"step": 784
},
{
"epoch": 0.5351056578050443,
"grad_norm": 9.278156280517578,
"learning_rate": 1.7196806367374656e-05,
"loss": 1.4426381587982178,
"step": 785
},
{
"epoch": 0.5357873210633947,
"grad_norm": 1.8680217266082764,
"learning_rate": 1.718913511199551e-05,
"loss": 0.5805763006210327,
"step": 786
},
{
"epoch": 0.5364689843217451,
"grad_norm": 5.596136569976807,
"learning_rate": 1.7181455090407667e-05,
"loss": 1.9274470806121826,
"step": 787
},
{
"epoch": 0.5371506475800955,
"grad_norm": 1.6634202003479004,
"learning_rate": 1.717376631197591e-05,
"loss": 0.9097037315368652,
"step": 788
},
{
"epoch": 0.5378323108384458,
"grad_norm": 2.053905963897705,
"learning_rate": 1.7166068786075697e-05,
"loss": 0.7156496047973633,
"step": 789
},
{
"epoch": 0.5385139740967961,
"grad_norm": 1.8190159797668457,
"learning_rate": 1.7158362522093153e-05,
"loss": 0.9084742665290833,
"step": 790
},
{
"epoch": 0.5391956373551465,
"grad_norm": 6.989439010620117,
"learning_rate": 1.715064752942506e-05,
"loss": 1.7869974374771118,
"step": 791
},
{
"epoch": 0.5398773006134969,
"grad_norm": 2.9668173789978027,
"learning_rate": 1.714292381747883e-05,
"loss": 1.452651023864746,
"step": 792
},
{
"epoch": 0.5405589638718473,
"grad_norm": 2.09556245803833,
"learning_rate": 1.7135191395672532e-05,
"loss": 1.1105618476867676,
"step": 793
},
{
"epoch": 0.5412406271301977,
"grad_norm": 2.4879820346832275,
"learning_rate": 1.7127450273434837e-05,
"loss": 1.349273681640625,
"step": 794
},
{
"epoch": 0.5419222903885481,
"grad_norm": 2.475329875946045,
"learning_rate": 1.7119700460205026e-05,
"loss": 1.1380001306533813,
"step": 795
},
{
"epoch": 0.5426039536468984,
"grad_norm": 2.9453556537628174,
"learning_rate": 1.7111941965432985e-05,
"loss": 1.2032630443572998,
"step": 796
},
{
"epoch": 0.5432856169052488,
"grad_norm": 1.9774304628372192,
"learning_rate": 1.710417479857918e-05,
"loss": 1.3968552350997925,
"step": 797
},
{
"epoch": 0.5439672801635992,
"grad_norm": 1.4213351011276245,
"learning_rate": 1.709639896911466e-05,
"loss": 1.2200727462768555,
"step": 798
},
{
"epoch": 0.5446489434219496,
"grad_norm": 2.5062761306762695,
"learning_rate": 1.708861448652102e-05,
"loss": 1.232042908668518,
"step": 799
},
{
"epoch": 0.5453306066803,
"grad_norm": 2.286937713623047,
"learning_rate": 1.7080821360290426e-05,
"loss": 1.0151350498199463,
"step": 800
},
{
"epoch": 0.5460122699386503,
"grad_norm": 3.6527047157287598,
"learning_rate": 1.707301959992557e-05,
"loss": 0.5274887084960938,
"step": 801
},
{
"epoch": 0.5466939331970007,
"grad_norm": 5.329606056213379,
"learning_rate": 1.7065209214939677e-05,
"loss": 1.9559152126312256,
"step": 802
},
{
"epoch": 0.547375596455351,
"grad_norm": 3.702806234359741,
"learning_rate": 1.7057390214856493e-05,
"loss": 1.6446794271469116,
"step": 803
},
{
"epoch": 0.5480572597137015,
"grad_norm": 3.3207600116729736,
"learning_rate": 1.704956260921026e-05,
"loss": 0.7604854106903076,
"step": 804
},
{
"epoch": 0.5487389229720518,
"grad_norm": 3.5969300270080566,
"learning_rate": 1.7041726407545716e-05,
"loss": 2.1249866485595703,
"step": 805
},
{
"epoch": 0.5494205862304021,
"grad_norm": 3.1857168674468994,
"learning_rate": 1.703388161941809e-05,
"loss": 1.656814455986023,
"step": 806
},
{
"epoch": 0.5501022494887525,
"grad_norm": 6.424698829650879,
"learning_rate": 1.7026028254393067e-05,
"loss": 0.5891388058662415,
"step": 807
},
{
"epoch": 0.5507839127471029,
"grad_norm": 2.86966872215271,
"learning_rate": 1.7018166322046798e-05,
"loss": 1.0807417631149292,
"step": 808
},
{
"epoch": 0.5514655760054533,
"grad_norm": 2.9843268394470215,
"learning_rate": 1.7010295831965886e-05,
"loss": 1.485582947731018,
"step": 809
},
{
"epoch": 0.5521472392638037,
"grad_norm": 2.941580295562744,
"learning_rate": 1.7002416793747354e-05,
"loss": 0.7899935841560364,
"step": 810
},
{
"epoch": 0.5528289025221541,
"grad_norm": 3.1687259674072266,
"learning_rate": 1.6994529216998664e-05,
"loss": 0.9913202524185181,
"step": 811
},
{
"epoch": 0.5535105657805044,
"grad_norm": 3.191603183746338,
"learning_rate": 1.698663311133768e-05,
"loss": 1.5229895114898682,
"step": 812
},
{
"epoch": 0.5541922290388548,
"grad_norm": 3.0717062950134277,
"learning_rate": 1.697872848639267e-05,
"loss": 1.01241135597229,
"step": 813
},
{
"epoch": 0.5548738922972052,
"grad_norm": 3.8809890747070312,
"learning_rate": 1.6970815351802285e-05,
"loss": 1.1171845197677612,
"step": 814
},
{
"epoch": 0.5555555555555556,
"grad_norm": 2.6531922817230225,
"learning_rate": 1.696289371721556e-05,
"loss": 1.0600652694702148,
"step": 815
},
{
"epoch": 0.556237218813906,
"grad_norm": 2.629523754119873,
"learning_rate": 1.695496359229189e-05,
"loss": 1.5618984699249268,
"step": 816
},
{
"epoch": 0.5569188820722563,
"grad_norm": 3.4782345294952393,
"learning_rate": 1.694702498670102e-05,
"loss": 1.4636913537979126,
"step": 817
},
{
"epoch": 0.5576005453306067,
"grad_norm": 5.301943778991699,
"learning_rate": 1.693907791012305e-05,
"loss": 2.062746524810791,
"step": 818
},
{
"epoch": 0.558282208588957,
"grad_norm": 8.220975875854492,
"learning_rate": 1.6931122372248386e-05,
"loss": 1.9080711603164673,
"step": 819
},
{
"epoch": 0.5589638718473074,
"grad_norm": 9.341310501098633,
"learning_rate": 1.692315838277778e-05,
"loss": 2.5527243614196777,
"step": 820
},
{
"epoch": 0.5596455351056578,
"grad_norm": 3.4971513748168945,
"learning_rate": 1.6915185951422256e-05,
"loss": 0.818741500377655,
"step": 821
},
{
"epoch": 0.5603271983640081,
"grad_norm": 2.5906221866607666,
"learning_rate": 1.690720508790316e-05,
"loss": 1.187768578529358,
"step": 822
},
{
"epoch": 0.5610088616223585,
"grad_norm": 2.4157657623291016,
"learning_rate": 1.6899215801952112e-05,
"loss": 0.821315348148346,
"step": 823
},
{
"epoch": 0.5616905248807089,
"grad_norm": 3.0064284801483154,
"learning_rate": 1.6891218103310994e-05,
"loss": 1.7353657484054565,
"step": 824
},
{
"epoch": 0.5623721881390593,
"grad_norm": 2.2080047130584717,
"learning_rate": 1.6883212001731956e-05,
"loss": 0.8247939348220825,
"step": 825
},
{
"epoch": 0.5630538513974097,
"grad_norm": 3.125454902648926,
"learning_rate": 1.6875197506977387e-05,
"loss": 0.867475152015686,
"step": 826
},
{
"epoch": 0.5637355146557601,
"grad_norm": 3.8463499546051025,
"learning_rate": 1.686717462881992e-05,
"loss": 1.0625464916229248,
"step": 827
},
{
"epoch": 0.5644171779141104,
"grad_norm": 5.8361406326293945,
"learning_rate": 1.685914337704239e-05,
"loss": 2.3758678436279297,
"step": 828
},
{
"epoch": 0.5650988411724608,
"grad_norm": 2.165736436843872,
"learning_rate": 1.6851103761437876e-05,
"loss": 1.1737010478973389,
"step": 829
},
{
"epoch": 0.5657805044308112,
"grad_norm": 2.5541231632232666,
"learning_rate": 1.6843055791809623e-05,
"loss": 1.3654487133026123,
"step": 830
},
{
"epoch": 0.5664621676891616,
"grad_norm": 1.6458773612976074,
"learning_rate": 1.6834999477971078e-05,
"loss": 1.1628589630126953,
"step": 831
},
{
"epoch": 0.567143830947512,
"grad_norm": 2.6931536197662354,
"learning_rate": 1.6826934829745868e-05,
"loss": 1.0243465900421143,
"step": 832
},
{
"epoch": 0.5678254942058623,
"grad_norm": 2.4119794368743896,
"learning_rate": 1.6818861856967762e-05,
"loss": 1.3056764602661133,
"step": 833
},
{
"epoch": 0.5685071574642127,
"grad_norm": 2.928074836730957,
"learning_rate": 1.68107805694807e-05,
"loss": 1.3191547393798828,
"step": 834
},
{
"epoch": 0.569188820722563,
"grad_norm": 1.8255895376205444,
"learning_rate": 1.680269097713876e-05,
"loss": 1.5903112888336182,
"step": 835
},
{
"epoch": 0.5698704839809134,
"grad_norm": 2.3143746852874756,
"learning_rate": 1.6794593089806134e-05,
"loss": 1.0325926542282104,
"step": 836
},
{
"epoch": 0.5705521472392638,
"grad_norm": 3.990488052368164,
"learning_rate": 1.678648691735713e-05,
"loss": 1.4269142150878906,
"step": 837
},
{
"epoch": 0.5712338104976141,
"grad_norm": 3.2297160625457764,
"learning_rate": 1.6778372469676173e-05,
"loss": 1.53303861618042,
"step": 838
},
{
"epoch": 0.5719154737559645,
"grad_norm": 2.850156545639038,
"learning_rate": 1.6770249756657762e-05,
"loss": 0.9840932488441467,
"step": 839
},
{
"epoch": 0.5725971370143149,
"grad_norm": 1.9680904150009155,
"learning_rate": 1.6762118788206488e-05,
"loss": 1.3542200326919556,
"step": 840
},
{
"epoch": 0.5732788002726653,
"grad_norm": 2.278607130050659,
"learning_rate": 1.6753979574236996e-05,
"loss": 0.4934600591659546,
"step": 841
},
{
"epoch": 0.5739604635310157,
"grad_norm": 2.5241875648498535,
"learning_rate": 1.6745832124673996e-05,
"loss": 0.564077615737915,
"step": 842
},
{
"epoch": 0.5746421267893661,
"grad_norm": 1.576717495918274,
"learning_rate": 1.6737676449452235e-05,
"loss": 1.2056704759597778,
"step": 843
},
{
"epoch": 0.5753237900477164,
"grad_norm": 3.5793991088867188,
"learning_rate": 1.672951255851649e-05,
"loss": 1.4175639152526855,
"step": 844
},
{
"epoch": 0.5760054533060668,
"grad_norm": 3.2483086585998535,
"learning_rate": 1.6721340461821555e-05,
"loss": 1.3645559549331665,
"step": 845
},
{
"epoch": 0.5766871165644172,
"grad_norm": 1.2694355249404907,
"learning_rate": 1.671316016933223e-05,
"loss": 1.0463674068450928,
"step": 846
},
{
"epoch": 0.5773687798227676,
"grad_norm": 1.5353333950042725,
"learning_rate": 1.6704971691023316e-05,
"loss": 0.9055666923522949,
"step": 847
},
{
"epoch": 0.578050443081118,
"grad_norm": 7.366179943084717,
"learning_rate": 1.6696775036879588e-05,
"loss": 2.724958896636963,
"step": 848
},
{
"epoch": 0.5787321063394683,
"grad_norm": 3.2687127590179443,
"learning_rate": 1.6688570216895793e-05,
"loss": 1.4015331268310547,
"step": 849
},
{
"epoch": 0.5794137695978187,
"grad_norm": 4.6578898429870605,
"learning_rate": 1.6680357241076632e-05,
"loss": 0.6592810153961182,
"step": 850
},
{
"epoch": 0.580095432856169,
"grad_norm": 2.0691535472869873,
"learning_rate": 1.667213611943675e-05,
"loss": 0.654139518737793,
"step": 851
},
{
"epoch": 0.5807770961145194,
"grad_norm": 2.046271324157715,
"learning_rate": 1.6663906862000736e-05,
"loss": 1.0838494300842285,
"step": 852
},
{
"epoch": 0.5814587593728698,
"grad_norm": 2.3908135890960693,
"learning_rate": 1.6655669478803086e-05,
"loss": 0.9503333568572998,
"step": 853
},
{
"epoch": 0.5821404226312201,
"grad_norm": 2.4451968669891357,
"learning_rate": 1.6647423979888214e-05,
"loss": 1.4373940229415894,
"step": 854
},
{
"epoch": 0.5828220858895705,
"grad_norm": 1.9011889696121216,
"learning_rate": 1.6639170375310422e-05,
"loss": 1.1368272304534912,
"step": 855
},
{
"epoch": 0.5835037491479209,
"grad_norm": 5.264576435089111,
"learning_rate": 1.6630908675133905e-05,
"loss": 2.110563278198242,
"step": 856
},
{
"epoch": 0.5841854124062713,
"grad_norm": 2.0514638423919678,
"learning_rate": 1.6622638889432716e-05,
"loss": 1.768547773361206,
"step": 857
},
{
"epoch": 0.5848670756646217,
"grad_norm": 2.4283359050750732,
"learning_rate": 1.6614361028290783e-05,
"loss": 1.0054471492767334,
"step": 858
},
{
"epoch": 0.5855487389229721,
"grad_norm": 2.2833070755004883,
"learning_rate": 1.660607510180187e-05,
"loss": 1.6151342391967773,
"step": 859
},
{
"epoch": 0.5862304021813224,
"grad_norm": 2.7579729557037354,
"learning_rate": 1.6597781120069584e-05,
"loss": 1.5430269241333008,
"step": 860
},
{
"epoch": 0.5869120654396728,
"grad_norm": 2.7292988300323486,
"learning_rate": 1.658947909320734e-05,
"loss": 0.5025774240493774,
"step": 861
},
{
"epoch": 0.5875937286980232,
"grad_norm": 2.3136188983917236,
"learning_rate": 1.658116903133838e-05,
"loss": 1.8564517498016357,
"step": 862
},
{
"epoch": 0.5882753919563736,
"grad_norm": 2.4586784839630127,
"learning_rate": 1.6572850944595735e-05,
"loss": 1.1788643598556519,
"step": 863
},
{
"epoch": 0.588957055214724,
"grad_norm": 6.834049701690674,
"learning_rate": 1.6564524843122223e-05,
"loss": 2.2534217834472656,
"step": 864
},
{
"epoch": 0.5896387184730743,
"grad_norm": 2.284600257873535,
"learning_rate": 1.655619073707043e-05,
"loss": 0.9232459664344788,
"step": 865
},
{
"epoch": 0.5903203817314246,
"grad_norm": 5.458041667938232,
"learning_rate": 1.6547848636602708e-05,
"loss": 1.7762432098388672,
"step": 866
},
{
"epoch": 0.591002044989775,
"grad_norm": 4.977056980133057,
"learning_rate": 1.653949855189116e-05,
"loss": 1.9290387630462646,
"step": 867
},
{
"epoch": 0.5916837082481254,
"grad_norm": 2.7046382427215576,
"learning_rate": 1.653114049311762e-05,
"loss": 1.8991743326187134,
"step": 868
},
{
"epoch": 0.5923653715064758,
"grad_norm": 3.796870708465576,
"learning_rate": 1.6522774470473642e-05,
"loss": 1.9370957612991333,
"step": 869
},
{
"epoch": 0.5930470347648262,
"grad_norm": 2.8540728092193604,
"learning_rate": 1.6514400494160498e-05,
"loss": 1.2413911819458008,
"step": 870
},
{
"epoch": 0.5937286980231765,
"grad_norm": 2.58767032623291,
"learning_rate": 1.6506018574389152e-05,
"loss": 1.1953917741775513,
"step": 871
},
{
"epoch": 0.5944103612815269,
"grad_norm": 4.319695472717285,
"learning_rate": 1.6497628721380257e-05,
"loss": 1.3608784675598145,
"step": 872
},
{
"epoch": 0.5950920245398773,
"grad_norm": 3.1396563053131104,
"learning_rate": 1.6489230945364148e-05,
"loss": 1.0947556495666504,
"step": 873
},
{
"epoch": 0.5957736877982277,
"grad_norm": 1.7502825260162354,
"learning_rate": 1.648082525658081e-05,
"loss": 0.942647397518158,
"step": 874
},
{
"epoch": 0.5964553510565781,
"grad_norm": 3.145920991897583,
"learning_rate": 1.6472411665279872e-05,
"loss": 0.8317646980285645,
"step": 875
},
{
"epoch": 0.5971370143149284,
"grad_norm": 2.4428975582122803,
"learning_rate": 1.646399018172061e-05,
"loss": 1.2358999252319336,
"step": 876
},
{
"epoch": 0.5978186775732788,
"grad_norm": 2.7724978923797607,
"learning_rate": 1.6455560816171928e-05,
"loss": 0.3633544445037842,
"step": 877
},
{
"epoch": 0.5985003408316292,
"grad_norm": 3.3238282203674316,
"learning_rate": 1.644712357891232e-05,
"loss": 1.467452049255371,
"step": 878
},
{
"epoch": 0.5991820040899796,
"grad_norm": 1.9840031862258911,
"learning_rate": 1.643867848022991e-05,
"loss": 1.1709258556365967,
"step": 879
},
{
"epoch": 0.59986366734833,
"grad_norm": 3.40712308883667,
"learning_rate": 1.643022553042237e-05,
"loss": 1.1537139415740967,
"step": 880
},
{
"epoch": 0.6005453306066802,
"grad_norm": 6.150967597961426,
"learning_rate": 1.6421764739796974e-05,
"loss": 1.8220754861831665,
"step": 881
},
{
"epoch": 0.6012269938650306,
"grad_norm": 3.064323663711548,
"learning_rate": 1.6413296118670553e-05,
"loss": 0.8146998882293701,
"step": 882
},
{
"epoch": 0.601908657123381,
"grad_norm": 3.7720255851745605,
"learning_rate": 1.6404819677369474e-05,
"loss": 1.929551124572754,
"step": 883
},
{
"epoch": 0.6025903203817314,
"grad_norm": 4.224287033081055,
"learning_rate": 1.639633542622965e-05,
"loss": 0.8990840911865234,
"step": 884
},
{
"epoch": 0.6032719836400818,
"grad_norm": 7.684881687164307,
"learning_rate": 1.6387843375596513e-05,
"loss": 1.4321951866149902,
"step": 885
},
{
"epoch": 0.6039536468984322,
"grad_norm": 3.8415868282318115,
"learning_rate": 1.6379343535825004e-05,
"loss": 0.9163957834243774,
"step": 886
},
{
"epoch": 0.6046353101567825,
"grad_norm": 3.149662971496582,
"learning_rate": 1.6370835917279573e-05,
"loss": 1.1009467840194702,
"step": 887
},
{
"epoch": 0.6053169734151329,
"grad_norm": 4.081305503845215,
"learning_rate": 1.636232053033414e-05,
"loss": 1.9384831190109253,
"step": 888
},
{
"epoch": 0.6059986366734833,
"grad_norm": 1.7843787670135498,
"learning_rate": 1.63537973853721e-05,
"loss": 0.908522367477417,
"step": 889
},
{
"epoch": 0.6066802999318337,
"grad_norm": 2.2492616176605225,
"learning_rate": 1.634526649278632e-05,
"loss": 1.1304497718811035,
"step": 890
},
{
"epoch": 0.6073619631901841,
"grad_norm": 2.759291648864746,
"learning_rate": 1.6336727862979108e-05,
"loss": 0.9259381294250488,
"step": 891
},
{
"epoch": 0.6080436264485344,
"grad_norm": 3.2604150772094727,
"learning_rate": 1.6328181506362193e-05,
"loss": 1.2496137619018555,
"step": 892
},
{
"epoch": 0.6087252897068848,
"grad_norm": 2.1876111030578613,
"learning_rate": 1.631962743335675e-05,
"loss": 1.1682977676391602,
"step": 893
},
{
"epoch": 0.6094069529652352,
"grad_norm": 2.5494682788848877,
"learning_rate": 1.631106565439334e-05,
"loss": 1.04998779296875,
"step": 894
},
{
"epoch": 0.6100886162235856,
"grad_norm": 1.378842830657959,
"learning_rate": 1.630249617991194e-05,
"loss": 0.8368078470230103,
"step": 895
},
{
"epoch": 0.610770279481936,
"grad_norm": 3.7163796424865723,
"learning_rate": 1.6293919020361895e-05,
"loss": 1.5429718494415283,
"step": 896
},
{
"epoch": 0.6114519427402862,
"grad_norm": 7.239688873291016,
"learning_rate": 1.6285334186201933e-05,
"loss": 2.765354871749878,
"step": 897
},
{
"epoch": 0.6121336059986366,
"grad_norm": 1.3090721368789673,
"learning_rate": 1.6276741687900134e-05,
"loss": 0.7607836127281189,
"step": 898
},
{
"epoch": 0.612815269256987,
"grad_norm": 3.923982858657837,
"learning_rate": 1.626814153593392e-05,
"loss": 1.0957961082458496,
"step": 899
},
{
"epoch": 0.6134969325153374,
"grad_norm": 3.3863680362701416,
"learning_rate": 1.6259533740790055e-05,
"loss": 1.135899305343628,
"step": 900
},
{
"epoch": 0.6141785957736878,
"grad_norm": 2.371218681335449,
"learning_rate": 1.6250918312964613e-05,
"loss": 1.238204002380371,
"step": 901
},
{
"epoch": 0.6148602590320382,
"grad_norm": 1.8255348205566406,
"learning_rate": 1.6242295262962983e-05,
"loss": 0.9338752031326294,
"step": 902
},
{
"epoch": 0.6155419222903885,
"grad_norm": 4.244454383850098,
"learning_rate": 1.6233664601299848e-05,
"loss": 1.579453468322754,
"step": 903
},
{
"epoch": 0.6162235855487389,
"grad_norm": 1.443424940109253,
"learning_rate": 1.6225026338499166e-05,
"loss": 0.8247011303901672,
"step": 904
},
{
"epoch": 0.6169052488070893,
"grad_norm": 1.967061996459961,
"learning_rate": 1.6216380485094164e-05,
"loss": 0.9666604995727539,
"step": 905
},
{
"epoch": 0.6175869120654397,
"grad_norm": 3.75724720954895,
"learning_rate": 1.6207727051627334e-05,
"loss": 1.3404090404510498,
"step": 906
},
{
"epoch": 0.6182685753237901,
"grad_norm": 2.1765472888946533,
"learning_rate": 1.61990660486504e-05,
"loss": 0.6874090433120728,
"step": 907
},
{
"epoch": 0.6189502385821404,
"grad_norm": 1.8902146816253662,
"learning_rate": 1.6190397486724324e-05,
"loss": 1.1136112213134766,
"step": 908
},
{
"epoch": 0.6196319018404908,
"grad_norm": 1.558221459388733,
"learning_rate": 1.6181721376419282e-05,
"loss": 1.1098493337631226,
"step": 909
},
{
"epoch": 0.6203135650988412,
"grad_norm": 5.35651159286499,
"learning_rate": 1.617303772831465e-05,
"loss": 2.0255041122436523,
"step": 910
},
{
"epoch": 0.6209952283571916,
"grad_norm": 5.652133464813232,
"learning_rate": 1.6164346552999e-05,
"loss": 2.2431159019470215,
"step": 911
},
{
"epoch": 0.621676891615542,
"grad_norm": 2.664736747741699,
"learning_rate": 1.615564786107009e-05,
"loss": 1.6007494926452637,
"step": 912
},
{
"epoch": 0.6223585548738922,
"grad_norm": 3.1986606121063232,
"learning_rate": 1.614694166313483e-05,
"loss": 0.9353867769241333,
"step": 913
},
{
"epoch": 0.6230402181322426,
"grad_norm": 3.8483238220214844,
"learning_rate": 1.6138227969809283e-05,
"loss": 1.6212551593780518,
"step": 914
},
{
"epoch": 0.623721881390593,
"grad_norm": 6.442831039428711,
"learning_rate": 1.6129506791718665e-05,
"loss": 1.4637436866760254,
"step": 915
},
{
"epoch": 0.6244035446489434,
"grad_norm": 3.796579122543335,
"learning_rate": 1.6120778139497307e-05,
"loss": 1.3633496761322021,
"step": 916
},
{
"epoch": 0.6250852079072938,
"grad_norm": 4.911407470703125,
"learning_rate": 1.611204202378866e-05,
"loss": 1.731675386428833,
"step": 917
},
{
"epoch": 0.6257668711656442,
"grad_norm": 3.0223350524902344,
"learning_rate": 1.6103298455245267e-05,
"loss": 2.0269124507904053,
"step": 918
},
{
"epoch": 0.6264485344239945,
"grad_norm": 1.8000354766845703,
"learning_rate": 1.6094547444528767e-05,
"loss": 1.2296888828277588,
"step": 919
},
{
"epoch": 0.6271301976823449,
"grad_norm": 2.296272039413452,
"learning_rate": 1.6085789002309873e-05,
"loss": 0.943554699420929,
"step": 920
},
{
"epoch": 0.6278118609406953,
"grad_norm": 4.166633605957031,
"learning_rate": 1.607702313926836e-05,
"loss": 1.6951007843017578,
"step": 921
},
{
"epoch": 0.6284935241990457,
"grad_norm": 2.0115439891815186,
"learning_rate": 1.6068249866093046e-05,
"loss": 1.4381415843963623,
"step": 922
},
{
"epoch": 0.6291751874573961,
"grad_norm": 2.222892999649048,
"learning_rate": 1.605946919348179e-05,
"loss": 0.7805427312850952,
"step": 923
},
{
"epoch": 0.6298568507157464,
"grad_norm": 7.529016494750977,
"learning_rate": 1.6050681132141473e-05,
"loss": 2.7365059852600098,
"step": 924
},
{
"epoch": 0.6305385139740968,
"grad_norm": 2.209188461303711,
"learning_rate": 1.6041885692787985e-05,
"loss": 0.549006998538971,
"step": 925
},
{
"epoch": 0.6312201772324472,
"grad_norm": 1.747800588607788,
"learning_rate": 1.603308288614621e-05,
"loss": 1.1981534957885742,
"step": 926
},
{
"epoch": 0.6319018404907976,
"grad_norm": 1.9363545179367065,
"learning_rate": 1.602427272295002e-05,
"loss": 1.0019506216049194,
"step": 927
},
{
"epoch": 0.632583503749148,
"grad_norm": 2.520519256591797,
"learning_rate": 1.6015455213942253e-05,
"loss": 1.3125648498535156,
"step": 928
},
{
"epoch": 0.6332651670074982,
"grad_norm": 1.7315162420272827,
"learning_rate": 1.600663036987471e-05,
"loss": 1.2615352869033813,
"step": 929
},
{
"epoch": 0.6339468302658486,
"grad_norm": 2.635585308074951,
"learning_rate": 1.599779820150813e-05,
"loss": 1.0167396068572998,
"step": 930
},
{
"epoch": 0.634628493524199,
"grad_norm": 2.0008440017700195,
"learning_rate": 1.5988958719612182e-05,
"loss": 1.4046450853347778,
"step": 931
},
{
"epoch": 0.6353101567825494,
"grad_norm": 1.9721522331237793,
"learning_rate": 1.5980111934965467e-05,
"loss": 1.0870532989501953,
"step": 932
},
{
"epoch": 0.6359918200408998,
"grad_norm": 2.249335289001465,
"learning_rate": 1.5971257858355467e-05,
"loss": 1.325969934463501,
"step": 933
},
{
"epoch": 0.6366734832992502,
"grad_norm": 5.9346747398376465,
"learning_rate": 1.596239650057858e-05,
"loss": 2.0995099544525146,
"step": 934
},
{
"epoch": 0.6373551465576005,
"grad_norm": 4.217530250549316,
"learning_rate": 1.5953527872440063e-05,
"loss": 1.4458832740783691,
"step": 935
},
{
"epoch": 0.6380368098159509,
"grad_norm": 4.074642658233643,
"learning_rate": 1.5944651984754053e-05,
"loss": 0.7346776723861694,
"step": 936
},
{
"epoch": 0.6387184730743013,
"grad_norm": 3.239382743835449,
"learning_rate": 1.5935768848343527e-05,
"loss": 1.4215084314346313,
"step": 937
},
{
"epoch": 0.6394001363326517,
"grad_norm": 2.162938117980957,
"learning_rate": 1.5926878474040313e-05,
"loss": 1.3204901218414307,
"step": 938
},
{
"epoch": 0.6400817995910021,
"grad_norm": 3.4916062355041504,
"learning_rate": 1.591798087268505e-05,
"loss": 0.5219067335128784,
"step": 939
},
{
"epoch": 0.6407634628493524,
"grad_norm": 3.3285751342773438,
"learning_rate": 1.5909076055127202e-05,
"loss": 1.9166496992111206,
"step": 940
},
{
"epoch": 0.6414451261077028,
"grad_norm": 3.0225143432617188,
"learning_rate": 1.590016403222503e-05,
"loss": 1.4402289390563965,
"step": 941
},
{
"epoch": 0.6421267893660532,
"grad_norm": 3.20241117477417,
"learning_rate": 1.5891244814845575e-05,
"loss": 1.6794233322143555,
"step": 942
},
{
"epoch": 0.6428084526244036,
"grad_norm": 2.3573503494262695,
"learning_rate": 1.5882318413864653e-05,
"loss": 1.4372035264968872,
"step": 943
},
{
"epoch": 0.643490115882754,
"grad_norm": 4.339122772216797,
"learning_rate": 1.5873384840166846e-05,
"loss": 2.0290374755859375,
"step": 944
},
{
"epoch": 0.6441717791411042,
"grad_norm": 2.5518579483032227,
"learning_rate": 1.5864444104645473e-05,
"loss": 1.3544771671295166,
"step": 945
},
{
"epoch": 0.6448534423994546,
"grad_norm": 3.1364567279815674,
"learning_rate": 1.5855496218202592e-05,
"loss": 1.642664909362793,
"step": 946
},
{
"epoch": 0.645535105657805,
"grad_norm": 4.1128249168396,
"learning_rate": 1.5846541191748978e-05,
"loss": 1.026753306388855,
"step": 947
},
{
"epoch": 0.6462167689161554,
"grad_norm": 2.6246652603149414,
"learning_rate": 1.5837579036204114e-05,
"loss": 1.317074179649353,
"step": 948
},
{
"epoch": 0.6468984321745058,
"grad_norm": 2.8232638835906982,
"learning_rate": 1.582860976249617e-05,
"loss": 0.8829560279846191,
"step": 949
},
{
"epoch": 0.6475800954328562,
"grad_norm": 2.4629805088043213,
"learning_rate": 1.581963338156201e-05,
"loss": 1.2965718507766724,
"step": 950
},
{
"epoch": 0.6482617586912065,
"grad_norm": 1.6031711101531982,
"learning_rate": 1.5810649904347145e-05,
"loss": 0.9839382171630859,
"step": 951
},
{
"epoch": 0.6489434219495569,
"grad_norm": 2.805589199066162,
"learning_rate": 1.5801659341805752e-05,
"loss": 1.3879642486572266,
"step": 952
},
{
"epoch": 0.6496250852079073,
"grad_norm": 3.0134496688842773,
"learning_rate": 1.5792661704900648e-05,
"loss": 1.470901608467102,
"step": 953
},
{
"epoch": 0.6503067484662577,
"grad_norm": 5.161888122558594,
"learning_rate": 1.5783657004603273e-05,
"loss": 2.198390483856201,
"step": 954
},
{
"epoch": 0.6509884117246081,
"grad_norm": 2.703739881515503,
"learning_rate": 1.5774645251893673e-05,
"loss": 1.5957465171813965,
"step": 955
},
{
"epoch": 0.6516700749829584,
"grad_norm": 2.825924873352051,
"learning_rate": 1.5765626457760506e-05,
"loss": 0.8621901273727417,
"step": 956
},
{
"epoch": 0.6523517382413088,
"grad_norm": 2.693979501724243,
"learning_rate": 1.575660063320101e-05,
"loss": 1.402087688446045,
"step": 957
},
{
"epoch": 0.6530334014996592,
"grad_norm": 2.7669472694396973,
"learning_rate": 1.574756778922099e-05,
"loss": 1.3180046081542969,
"step": 958
},
{
"epoch": 0.6537150647580096,
"grad_norm": 2.372765302658081,
"learning_rate": 1.5738527936834824e-05,
"loss": 0.9966622591018677,
"step": 959
},
{
"epoch": 0.65439672801636,
"grad_norm": 1.8964282274246216,
"learning_rate": 1.5729481087065423e-05,
"loss": 1.2379252910614014,
"step": 960
},
{
"epoch": 0.6550783912747103,
"grad_norm": 4.117354869842529,
"learning_rate": 1.5720427250944237e-05,
"loss": 2.0293562412261963,
"step": 961
},
{
"epoch": 0.6557600545330606,
"grad_norm": 2.0985937118530273,
"learning_rate": 1.5711366439511234e-05,
"loss": 0.9524621963500977,
"step": 962
},
{
"epoch": 0.656441717791411,
"grad_norm": 1.853117823600769,
"learning_rate": 1.5702298663814884e-05,
"loss": 0.8067352771759033,
"step": 963
},
{
"epoch": 0.6571233810497614,
"grad_norm": 2.6025946140289307,
"learning_rate": 1.569322393491216e-05,
"loss": 1.2105220556259155,
"step": 964
},
{
"epoch": 0.6578050443081118,
"grad_norm": 1.4278172254562378,
"learning_rate": 1.5684142263868493e-05,
"loss": 1.3550410270690918,
"step": 965
},
{
"epoch": 0.6584867075664622,
"grad_norm": 1.4979023933410645,
"learning_rate": 1.5675053661757802e-05,
"loss": 0.5304239988327026,
"step": 966
},
{
"epoch": 0.6591683708248125,
"grad_norm": 1.8251065015792847,
"learning_rate": 1.566595813966244e-05,
"loss": 0.6945926547050476,
"step": 967
},
{
"epoch": 0.6598500340831629,
"grad_norm": 1.7620820999145508,
"learning_rate": 1.5656855708673208e-05,
"loss": 1.5049657821655273,
"step": 968
},
{
"epoch": 0.6605316973415133,
"grad_norm": 1.326943039894104,
"learning_rate": 1.564774637988933e-05,
"loss": 0.8265421986579895,
"step": 969
},
{
"epoch": 0.6612133605998637,
"grad_norm": 1.3603960275650024,
"learning_rate": 1.5638630164418435e-05,
"loss": 0.9551488757133484,
"step": 970
},
{
"epoch": 0.6618950238582141,
"grad_norm": 3.848780632019043,
"learning_rate": 1.5629507073376556e-05,
"loss": 1.197197675704956,
"step": 971
},
{
"epoch": 0.6625766871165644,
"grad_norm": 3.2928988933563232,
"learning_rate": 1.562037711788811e-05,
"loss": 1.1889151334762573,
"step": 972
},
{
"epoch": 0.6632583503749148,
"grad_norm": 2.450935125350952,
"learning_rate": 1.5611240309085877e-05,
"loss": 1.1293702125549316,
"step": 973
},
{
"epoch": 0.6639400136332652,
"grad_norm": 1.8610042333602905,
"learning_rate": 1.5602096658111003e-05,
"loss": 0.689786970615387,
"step": 974
},
{
"epoch": 0.6646216768916156,
"grad_norm": 2.4367642402648926,
"learning_rate": 1.5592946176112973e-05,
"loss": 0.8688675761222839,
"step": 975
},
{
"epoch": 0.665303340149966,
"grad_norm": 2.765475034713745,
"learning_rate": 1.55837888742496e-05,
"loss": 1.4046356678009033,
"step": 976
},
{
"epoch": 0.6659850034083163,
"grad_norm": 2.2495126724243164,
"learning_rate": 1.5574624763687006e-05,
"loss": 1.0725502967834473,
"step": 977
},
{
"epoch": 0.6666666666666666,
"grad_norm": 2.21120548248291,
"learning_rate": 1.556545385559964e-05,
"loss": 0.50251704454422,
"step": 978
},
{
"epoch": 0.667348329925017,
"grad_norm": 2.6886188983917236,
"learning_rate": 1.5556276161170214e-05,
"loss": 1.5717726945877075,
"step": 979
},
{
"epoch": 0.6680299931833674,
"grad_norm": 2.472787380218506,
"learning_rate": 1.554709169158972e-05,
"loss": 0.9982832670211792,
"step": 980
},
{
"epoch": 0.6687116564417178,
"grad_norm": 2.3250319957733154,
"learning_rate": 1.5537900458057426e-05,
"loss": 1.3237147331237793,
"step": 981
},
{
"epoch": 0.6693933197000682,
"grad_norm": 3.5646936893463135,
"learning_rate": 1.5528702471780832e-05,
"loss": 1.4620048999786377,
"step": 982
},
{
"epoch": 0.6700749829584185,
"grad_norm": 3.5376639366149902,
"learning_rate": 1.5519497743975676e-05,
"loss": 1.7363507747650146,
"step": 983
},
{
"epoch": 0.6707566462167689,
"grad_norm": 2.8820950984954834,
"learning_rate": 1.551028628586592e-05,
"loss": 0.7051398754119873,
"step": 984
},
{
"epoch": 0.6714383094751193,
"grad_norm": 4.920443058013916,
"learning_rate": 1.550106810868373e-05,
"loss": 2.031773567199707,
"step": 985
},
{
"epoch": 0.6721199727334697,
"grad_norm": 5.214235305786133,
"learning_rate": 1.549184322366947e-05,
"loss": 1.7547272443771362,
"step": 986
},
{
"epoch": 0.6728016359918201,
"grad_norm": 3.160337209701538,
"learning_rate": 1.5482611642071672e-05,
"loss": 0.8350422978401184,
"step": 987
},
{
"epoch": 0.6734832992501704,
"grad_norm": 2.562685966491699,
"learning_rate": 1.5473373375147046e-05,
"loss": 1.5145583152770996,
"step": 988
},
{
"epoch": 0.6741649625085208,
"grad_norm": 2.3527350425720215,
"learning_rate": 1.546412843416045e-05,
"loss": 1.2808892726898193,
"step": 989
},
{
"epoch": 0.6748466257668712,
"grad_norm": 2.424290180206299,
"learning_rate": 1.5454876830384868e-05,
"loss": 1.0751357078552246,
"step": 990
},
{
"epoch": 0.6755282890252216,
"grad_norm": 3.4699552059173584,
"learning_rate": 1.544561857510143e-05,
"loss": 1.7717998027801514,
"step": 991
},
{
"epoch": 0.676209952283572,
"grad_norm": 5.047571182250977,
"learning_rate": 1.5436353679599363e-05,
"loss": 1.4212061166763306,
"step": 992
},
{
"epoch": 0.6768916155419223,
"grad_norm": 1.3153938055038452,
"learning_rate": 1.5427082155175993e-05,
"loss": 0.9446808695793152,
"step": 993
},
{
"epoch": 0.6775732788002726,
"grad_norm": 3.4039230346679688,
"learning_rate": 1.541780401313673e-05,
"loss": 0.8643260598182678,
"step": 994
},
{
"epoch": 0.678254942058623,
"grad_norm": 1.9677404165267944,
"learning_rate": 1.540851926479505e-05,
"loss": 1.230716586112976,
"step": 995
},
{
"epoch": 0.6789366053169734,
"grad_norm": 3.332017183303833,
"learning_rate": 1.5399227921472493e-05,
"loss": 1.6875895261764526,
"step": 996
},
{
"epoch": 0.6796182685753238,
"grad_norm": 2.3876984119415283,
"learning_rate": 1.5389929994498635e-05,
"loss": 0.35820311307907104,
"step": 997
},
{
"epoch": 0.6802999318336742,
"grad_norm": 3.2951459884643555,
"learning_rate": 1.5380625495211072e-05,
"loss": 0.7848578095436096,
"step": 998
},
{
"epoch": 0.6809815950920245,
"grad_norm": 2.0296571254730225,
"learning_rate": 1.537131443495543e-05,
"loss": 1.5027271509170532,
"step": 999
},
{
"epoch": 0.6816632583503749,
"grad_norm": 3.0071208477020264,
"learning_rate": 1.536199682508533e-05,
"loss": 1.7935607433319092,
"step": 1000
},
{
"epoch": 0.6823449216087253,
"grad_norm": 3.048903226852417,
"learning_rate": 1.5352672676962365e-05,
"loss": 1.4052588939666748,
"step": 1001
},
{
"epoch": 0.6830265848670757,
"grad_norm": 2.0453131198883057,
"learning_rate": 1.5343342001956125e-05,
"loss": 0.7496938705444336,
"step": 1002
},
{
"epoch": 0.6837082481254261,
"grad_norm": 2.2888898849487305,
"learning_rate": 1.533400481144414e-05,
"loss": 0.9012516736984253,
"step": 1003
},
{
"epoch": 0.6843899113837764,
"grad_norm": 4.049041271209717,
"learning_rate": 1.5324661116811887e-05,
"loss": 1.8084564208984375,
"step": 1004
},
{
"epoch": 0.6850715746421268,
"grad_norm": 3.5373117923736572,
"learning_rate": 1.531531092945279e-05,
"loss": 1.5952699184417725,
"step": 1005
},
{
"epoch": 0.6857532379004772,
"grad_norm": 1.6792633533477783,
"learning_rate": 1.5305954260768166e-05,
"loss": 0.2857983112335205,
"step": 1006
},
{
"epoch": 0.6864349011588275,
"grad_norm": 1.9056538343429565,
"learning_rate": 1.5296591122167254e-05,
"loss": 0.3411814272403717,
"step": 1007
},
{
"epoch": 0.6871165644171779,
"grad_norm": 1.7760215997695923,
"learning_rate": 1.5287221525067168e-05,
"loss": 0.6943484544754028,
"step": 1008
},
{
"epoch": 0.6877982276755283,
"grad_norm": 2.4310076236724854,
"learning_rate": 1.5277845480892914e-05,
"loss": 1.2629508972167969,
"step": 1009
},
{
"epoch": 0.6884798909338786,
"grad_norm": 1.9507763385772705,
"learning_rate": 1.526846300107734e-05,
"loss": 1.1614940166473389,
"step": 1010
},
{
"epoch": 0.689161554192229,
"grad_norm": 4.232208728790283,
"learning_rate": 1.5259074097061166e-05,
"loss": 1.4028643369674683,
"step": 1011
},
{
"epoch": 0.6898432174505794,
"grad_norm": 3.713007688522339,
"learning_rate": 1.5249678780292913e-05,
"loss": 1.0402387380599976,
"step": 1012
},
{
"epoch": 0.6905248807089298,
"grad_norm": 2.0313596725463867,
"learning_rate": 1.5240277062228952e-05,
"loss": 0.4616890251636505,
"step": 1013
},
{
"epoch": 0.6912065439672802,
"grad_norm": 2.8289709091186523,
"learning_rate": 1.523086895433344e-05,
"loss": 1.3428205251693726,
"step": 1014
},
{
"epoch": 0.6918882072256305,
"grad_norm": 4.949586868286133,
"learning_rate": 1.5221454468078336e-05,
"loss": 1.4365060329437256,
"step": 1015
},
{
"epoch": 0.6925698704839809,
"grad_norm": 9.691402435302734,
"learning_rate": 1.5212033614943371e-05,
"loss": 3.060600519180298,
"step": 1016
},
{
"epoch": 0.6932515337423313,
"grad_norm": 2.583277463912964,
"learning_rate": 1.5202606406416043e-05,
"loss": 1.2374069690704346,
"step": 1017
},
{
"epoch": 0.6939331970006817,
"grad_norm": 3.875859022140503,
"learning_rate": 1.5193172853991596e-05,
"loss": 1.4892609119415283,
"step": 1018
},
{
"epoch": 0.6946148602590321,
"grad_norm": 4.193378448486328,
"learning_rate": 1.518373296917301e-05,
"loss": 1.2775728702545166,
"step": 1019
},
{
"epoch": 0.6952965235173824,
"grad_norm": 4.090426921844482,
"learning_rate": 1.5174286763470995e-05,
"loss": 0.4450388550758362,
"step": 1020
},
{
"epoch": 0.6959781867757328,
"grad_norm": 4.702569007873535,
"learning_rate": 1.5164834248403959e-05,
"loss": 0.5944312214851379,
"step": 1021
},
{
"epoch": 0.6966598500340832,
"grad_norm": 2.688467264175415,
"learning_rate": 1.5155375435498001e-05,
"loss": 1.0980937480926514,
"step": 1022
},
{
"epoch": 0.6973415132924335,
"grad_norm": 1.5542073249816895,
"learning_rate": 1.5145910336286912e-05,
"loss": 0.9820501208305359,
"step": 1023
},
{
"epoch": 0.6980231765507839,
"grad_norm": 2.9898831844329834,
"learning_rate": 1.5136438962312134e-05,
"loss": 1.3177458047866821,
"step": 1024
},
{
"epoch": 0.6987048398091343,
"grad_norm": 3.9392518997192383,
"learning_rate": 1.5126961325122773e-05,
"loss": 1.6820330619812012,
"step": 1025
},
{
"epoch": 0.6993865030674846,
"grad_norm": 1.8033751249313354,
"learning_rate": 1.511747743627556e-05,
"loss": 1.1170791387557983,
"step": 1026
},
{
"epoch": 0.700068166325835,
"grad_norm": 1.4688934087753296,
"learning_rate": 1.5107987307334864e-05,
"loss": 0.9317739605903625,
"step": 1027
},
{
"epoch": 0.7007498295841854,
"grad_norm": 2.8467535972595215,
"learning_rate": 1.5098490949872648e-05,
"loss": 1.3663240671157837,
"step": 1028
},
{
"epoch": 0.7014314928425358,
"grad_norm": 3.6989221572875977,
"learning_rate": 1.5088988375468473e-05,
"loss": 1.3530757427215576,
"step": 1029
},
{
"epoch": 0.7021131561008862,
"grad_norm": 3.1766574382781982,
"learning_rate": 1.5079479595709493e-05,
"loss": 1.1704943180084229,
"step": 1030
},
{
"epoch": 0.7027948193592365,
"grad_norm": 1.4907842874526978,
"learning_rate": 1.5069964622190409e-05,
"loss": 0.32295915484428406,
"step": 1031
},
{
"epoch": 0.7034764826175869,
"grad_norm": 2.6720528602600098,
"learning_rate": 1.5060443466513497e-05,
"loss": 1.995790958404541,
"step": 1032
},
{
"epoch": 0.7041581458759373,
"grad_norm": 2.4431049823760986,
"learning_rate": 1.5050916140288552e-05,
"loss": 1.0614051818847656,
"step": 1033
},
{
"epoch": 0.7048398091342877,
"grad_norm": 2.8213205337524414,
"learning_rate": 1.5041382655132899e-05,
"loss": 1.234609842300415,
"step": 1034
},
{
"epoch": 0.7055214723926381,
"grad_norm": 1.7441692352294922,
"learning_rate": 1.5031843022671377e-05,
"loss": 1.0156347751617432,
"step": 1035
},
{
"epoch": 0.7062031356509885,
"grad_norm": 1.8287917375564575,
"learning_rate": 1.5022297254536321e-05,
"loss": 1.3329904079437256,
"step": 1036
},
{
"epoch": 0.7068847989093388,
"grad_norm": 1.5410305261611938,
"learning_rate": 1.5012745362367543e-05,
"loss": 1.0349948406219482,
"step": 1037
},
{
"epoch": 0.7075664621676891,
"grad_norm": 1.9430551528930664,
"learning_rate": 1.5003187357812323e-05,
"loss": 1.4377586841583252,
"step": 1038
},
{
"epoch": 0.7082481254260395,
"grad_norm": 1.7912523746490479,
"learning_rate": 1.4993623252525398e-05,
"loss": 1.3896902799606323,
"step": 1039
},
{
"epoch": 0.7089297886843899,
"grad_norm": 2.173234701156616,
"learning_rate": 1.4984053058168936e-05,
"loss": 1.353006362915039,
"step": 1040
},
{
"epoch": 0.7096114519427403,
"grad_norm": 3.65002179145813,
"learning_rate": 1.4974476786412542e-05,
"loss": 1.4411238431930542,
"step": 1041
},
{
"epoch": 0.7102931152010906,
"grad_norm": 2.9713757038116455,
"learning_rate": 1.4964894448933227e-05,
"loss": 0.813322901725769,
"step": 1042
},
{
"epoch": 0.710974778459441,
"grad_norm": 1.54813814163208,
"learning_rate": 1.4955306057415388e-05,
"loss": 1.206241488456726,
"step": 1043
},
{
"epoch": 0.7116564417177914,
"grad_norm": 2.3598363399505615,
"learning_rate": 1.4945711623550822e-05,
"loss": 1.1738507747650146,
"step": 1044
},
{
"epoch": 0.7123381049761418,
"grad_norm": 2.5693681240081787,
"learning_rate": 1.4936111159038677e-05,
"loss": 1.855168342590332,
"step": 1045
},
{
"epoch": 0.7130197682344922,
"grad_norm": 1.6126163005828857,
"learning_rate": 1.4926504675585467e-05,
"loss": 1.1066110134124756,
"step": 1046
},
{
"epoch": 0.7137014314928425,
"grad_norm": 1.2783271074295044,
"learning_rate": 1.4916892184905037e-05,
"loss": 0.6282731294631958,
"step": 1047
},
{
"epoch": 0.7143830947511929,
"grad_norm": 2.2483839988708496,
"learning_rate": 1.4907273698718562e-05,
"loss": 0.9469138383865356,
"step": 1048
},
{
"epoch": 0.7150647580095433,
"grad_norm": 4.296797752380371,
"learning_rate": 1.4897649228754527e-05,
"loss": 1.7291085720062256,
"step": 1049
},
{
"epoch": 0.7157464212678937,
"grad_norm": 3.502342462539673,
"learning_rate": 1.4888018786748713e-05,
"loss": 1.0134327411651611,
"step": 1050
},
{
"epoch": 0.7164280845262441,
"grad_norm": 3.2990827560424805,
"learning_rate": 1.487838238444418e-05,
"loss": 1.590086579322815,
"step": 1051
},
{
"epoch": 0.7171097477845945,
"grad_norm": 3.785966157913208,
"learning_rate": 1.4868740033591258e-05,
"loss": 1.6262550354003906,
"step": 1052
},
{
"epoch": 0.7177914110429447,
"grad_norm": 6.545089244842529,
"learning_rate": 1.485909174594753e-05,
"loss": 2.984455108642578,
"step": 1053
},
{
"epoch": 0.7184730743012951,
"grad_norm": 2.6544864177703857,
"learning_rate": 1.484943753327783e-05,
"loss": 1.136319637298584,
"step": 1054
},
{
"epoch": 0.7191547375596455,
"grad_norm": 2.4620413780212402,
"learning_rate": 1.4839777407354194e-05,
"loss": 0.72819983959198,
"step": 1055
},
{
"epoch": 0.7198364008179959,
"grad_norm": 7.4918951988220215,
"learning_rate": 1.4830111379955886e-05,
"loss": 1.7762341499328613,
"step": 1056
},
{
"epoch": 0.7205180640763463,
"grad_norm": 6.07712459564209,
"learning_rate": 1.4820439462869353e-05,
"loss": 1.6564725637435913,
"step": 1057
},
{
"epoch": 0.7211997273346966,
"grad_norm": 2.416376829147339,
"learning_rate": 1.481076166788824e-05,
"loss": 1.1260161399841309,
"step": 1058
},
{
"epoch": 0.721881390593047,
"grad_norm": 3.4239766597747803,
"learning_rate": 1.480107800681335e-05,
"loss": 1.4130642414093018,
"step": 1059
},
{
"epoch": 0.7225630538513974,
"grad_norm": 2.0911810398101807,
"learning_rate": 1.4791388491452637e-05,
"loss": 1.4372596740722656,
"step": 1060
},
{
"epoch": 0.7232447171097478,
"grad_norm": 2.543943405151367,
"learning_rate": 1.4781693133621191e-05,
"loss": 0.859691858291626,
"step": 1061
},
{
"epoch": 0.7239263803680982,
"grad_norm": 2.474747657775879,
"learning_rate": 1.4771991945141237e-05,
"loss": 1.1199817657470703,
"step": 1062
},
{
"epoch": 0.7246080436264485,
"grad_norm": 7.446532249450684,
"learning_rate": 1.4762284937842103e-05,
"loss": 2.407529354095459,
"step": 1063
},
{
"epoch": 0.7252897068847989,
"grad_norm": 3.7536842823028564,
"learning_rate": 1.4752572123560216e-05,
"loss": 1.3106913566589355,
"step": 1064
},
{
"epoch": 0.7259713701431493,
"grad_norm": 3.0671236515045166,
"learning_rate": 1.4742853514139076e-05,
"loss": 0.7006158232688904,
"step": 1065
},
{
"epoch": 0.7266530334014997,
"grad_norm": 5.503408432006836,
"learning_rate": 1.4733129121429253e-05,
"loss": 2.3005611896514893,
"step": 1066
},
{
"epoch": 0.7273346966598501,
"grad_norm": 4.134151458740234,
"learning_rate": 1.4723398957288373e-05,
"loss": 1.7469112873077393,
"step": 1067
},
{
"epoch": 0.7280163599182005,
"grad_norm": 3.927537441253662,
"learning_rate": 1.4713663033581099e-05,
"loss": 1.7568607330322266,
"step": 1068
},
{
"epoch": 0.7286980231765507,
"grad_norm": 6.670135021209717,
"learning_rate": 1.470392136217911e-05,
"loss": 1.0784207582473755,
"step": 1069
},
{
"epoch": 0.7293796864349011,
"grad_norm": 4.407510280609131,
"learning_rate": 1.4694173954961105e-05,
"loss": 1.3891305923461914,
"step": 1070
},
{
"epoch": 0.7300613496932515,
"grad_norm": 3.5949463844299316,
"learning_rate": 1.4684420823812763e-05,
"loss": 0.9051351547241211,
"step": 1071
},
{
"epoch": 0.7307430129516019,
"grad_norm": 3.821967363357544,
"learning_rate": 1.4674661980626754e-05,
"loss": 0.6995428800582886,
"step": 1072
},
{
"epoch": 0.7314246762099523,
"grad_norm": 2.605086326599121,
"learning_rate": 1.466489743730271e-05,
"loss": 1.1066350936889648,
"step": 1073
},
{
"epoch": 0.7321063394683026,
"grad_norm": 2.183931350708008,
"learning_rate": 1.4655127205747208e-05,
"loss": 1.5164170265197754,
"step": 1074
},
{
"epoch": 0.732788002726653,
"grad_norm": 1.955289602279663,
"learning_rate": 1.4645351297873774e-05,
"loss": 1.0034122467041016,
"step": 1075
},
{
"epoch": 0.7334696659850034,
"grad_norm": 1.7779383659362793,
"learning_rate": 1.463556972560284e-05,
"loss": 1.0842978954315186,
"step": 1076
},
{
"epoch": 0.7341513292433538,
"grad_norm": 2.3080132007598877,
"learning_rate": 1.4625782500861756e-05,
"loss": 0.9262508749961853,
"step": 1077
},
{
"epoch": 0.7348329925017042,
"grad_norm": 2.3780229091644287,
"learning_rate": 1.4615989635584757e-05,
"loss": 0.780438244342804,
"step": 1078
},
{
"epoch": 0.7355146557600545,
"grad_norm": 1.959047794342041,
"learning_rate": 1.4606191141712964e-05,
"loss": 1.0655291080474854,
"step": 1079
},
{
"epoch": 0.7361963190184049,
"grad_norm": 2.633192300796509,
"learning_rate": 1.4596387031194354e-05,
"loss": 0.8464280366897583,
"step": 1080
},
{
"epoch": 0.7368779822767553,
"grad_norm": 1.6572291851043701,
"learning_rate": 1.458657731598376e-05,
"loss": 0.3969135880470276,
"step": 1081
},
{
"epoch": 0.7375596455351057,
"grad_norm": 2.454660177230835,
"learning_rate": 1.4576762008042837e-05,
"loss": 1.2453688383102417,
"step": 1082
},
{
"epoch": 0.7382413087934561,
"grad_norm": 4.7392144203186035,
"learning_rate": 1.4566941119340074e-05,
"loss": 1.502977967262268,
"step": 1083
},
{
"epoch": 0.7389229720518065,
"grad_norm": 2.6455323696136475,
"learning_rate": 1.4557114661850755e-05,
"loss": 1.0623095035552979,
"step": 1084
},
{
"epoch": 0.7396046353101567,
"grad_norm": 4.488423824310303,
"learning_rate": 1.4547282647556964e-05,
"loss": 1.8041034936904907,
"step": 1085
},
{
"epoch": 0.7402862985685071,
"grad_norm": 1.9584484100341797,
"learning_rate": 1.4537445088447547e-05,
"loss": 0.7578965425491333,
"step": 1086
},
{
"epoch": 0.7409679618268575,
"grad_norm": 2.6489927768707275,
"learning_rate": 1.4527601996518122e-05,
"loss": 0.7286869287490845,
"step": 1087
},
{
"epoch": 0.7416496250852079,
"grad_norm": 1.6441211700439453,
"learning_rate": 1.4517753383771052e-05,
"loss": 0.9616504311561584,
"step": 1088
},
{
"epoch": 0.7423312883435583,
"grad_norm": 2.347137212753296,
"learning_rate": 1.4507899262215426e-05,
"loss": 1.5748083591461182,
"step": 1089
},
{
"epoch": 0.7430129516019086,
"grad_norm": 2.179248809814453,
"learning_rate": 1.449803964386706e-05,
"loss": 1.1422452926635742,
"step": 1090
},
{
"epoch": 0.743694614860259,
"grad_norm": 2.3553104400634766,
"learning_rate": 1.4488174540748463e-05,
"loss": 1.0269725322723389,
"step": 1091
},
{
"epoch": 0.7443762781186094,
"grad_norm": 0.9572274684906006,
"learning_rate": 1.4478303964888842e-05,
"loss": 0.5475146770477295,
"step": 1092
},
{
"epoch": 0.7450579413769598,
"grad_norm": 2.5568504333496094,
"learning_rate": 1.4468427928324065e-05,
"loss": 1.681059718132019,
"step": 1093
},
{
"epoch": 0.7457396046353102,
"grad_norm": 1.5831090211868286,
"learning_rate": 1.4458546443096663e-05,
"loss": 1.3471312522888184,
"step": 1094
},
{
"epoch": 0.7464212678936605,
"grad_norm": 1.6717140674591064,
"learning_rate": 1.4448659521255823e-05,
"loss": 1.0710022449493408,
"step": 1095
},
{
"epoch": 0.7471029311520109,
"grad_norm": 2.684556007385254,
"learning_rate": 1.4438767174857346e-05,
"loss": 1.708756685256958,
"step": 1096
},
{
"epoch": 0.7477845944103613,
"grad_norm": 2.5712358951568604,
"learning_rate": 1.442886941596365e-05,
"loss": 1.1187701225280762,
"step": 1097
},
{
"epoch": 0.7484662576687117,
"grad_norm": 1.3731135129928589,
"learning_rate": 1.4418966256643762e-05,
"loss": 1.2401134967803955,
"step": 1098
},
{
"epoch": 0.7491479209270621,
"grad_norm": 2.568267345428467,
"learning_rate": 1.4409057708973282e-05,
"loss": 0.7313793897628784,
"step": 1099
},
{
"epoch": 0.7498295841854125,
"grad_norm": 2.1688249111175537,
"learning_rate": 1.4399143785034388e-05,
"loss": 1.3187694549560547,
"step": 1100
},
{
"epoch": 0.7505112474437627,
"grad_norm": 2.1575472354888916,
"learning_rate": 1.4389224496915814e-05,
"loss": 0.8957356214523315,
"step": 1101
},
{
"epoch": 0.7511929107021131,
"grad_norm": 2.9634883403778076,
"learning_rate": 1.4379299856712827e-05,
"loss": 0.5563441514968872,
"step": 1102
},
{
"epoch": 0.7518745739604635,
"grad_norm": 3.4954793453216553,
"learning_rate": 1.4369369876527234e-05,
"loss": 1.5315556526184082,
"step": 1103
},
{
"epoch": 0.7525562372188139,
"grad_norm": 1.7266521453857422,
"learning_rate": 1.4359434568467341e-05,
"loss": 0.7705234289169312,
"step": 1104
},
{
"epoch": 0.7532379004771643,
"grad_norm": 4.004641056060791,
"learning_rate": 1.4349493944647953e-05,
"loss": 1.7995654344558716,
"step": 1105
},
{
"epoch": 0.7539195637355146,
"grad_norm": 1.724542498588562,
"learning_rate": 1.4339548017190356e-05,
"loss": 0.3790285587310791,
"step": 1106
},
{
"epoch": 0.754601226993865,
"grad_norm": 2.3936798572540283,
"learning_rate": 1.4329596798222318e-05,
"loss": 0.6810147762298584,
"step": 1107
},
{
"epoch": 0.7552828902522154,
"grad_norm": 5.291030406951904,
"learning_rate": 1.4319640299878038e-05,
"loss": 2.0598695278167725,
"step": 1108
},
{
"epoch": 0.7559645535105658,
"grad_norm": 3.289830446243286,
"learning_rate": 1.4309678534298164e-05,
"loss": 1.0737022161483765,
"step": 1109
},
{
"epoch": 0.7566462167689162,
"grad_norm": 9.536825180053711,
"learning_rate": 1.4299711513629759e-05,
"loss": 1.3822963237762451,
"step": 1110
},
{
"epoch": 0.7573278800272665,
"grad_norm": 2.259124755859375,
"learning_rate": 1.428973925002631e-05,
"loss": 1.4830787181854248,
"step": 1111
},
{
"epoch": 0.7580095432856169,
"grad_norm": 1.6364065408706665,
"learning_rate": 1.4279761755647679e-05,
"loss": 1.0743396282196045,
"step": 1112
},
{
"epoch": 0.7586912065439673,
"grad_norm": 1.8119486570358276,
"learning_rate": 1.4269779042660112e-05,
"loss": 0.804845929145813,
"step": 1113
},
{
"epoch": 0.7593728698023177,
"grad_norm": 1.8661983013153076,
"learning_rate": 1.4259791123236227e-05,
"loss": 0.7559719085693359,
"step": 1114
},
{
"epoch": 0.7600545330606681,
"grad_norm": 1.2363063097000122,
"learning_rate": 1.4249798009554979e-05,
"loss": 0.8930497169494629,
"step": 1115
},
{
"epoch": 0.7607361963190185,
"grad_norm": 1.8818364143371582,
"learning_rate": 1.4239799713801662e-05,
"loss": 1.1456657648086548,
"step": 1116
},
{
"epoch": 0.7614178595773687,
"grad_norm": 1.87226140499115,
"learning_rate": 1.4229796248167888e-05,
"loss": 0.5097870230674744,
"step": 1117
},
{
"epoch": 0.7620995228357191,
"grad_norm": 3.1565282344818115,
"learning_rate": 1.421978762485157e-05,
"loss": 1.0278431177139282,
"step": 1118
},
{
"epoch": 0.7627811860940695,
"grad_norm": 5.5133957862854,
"learning_rate": 1.4209773856056925e-05,
"loss": 2.1597955226898193,
"step": 1119
},
{
"epoch": 0.7634628493524199,
"grad_norm": 3.499448537826538,
"learning_rate": 1.419975495399442e-05,
"loss": 1.8728885650634766,
"step": 1120
},
{
"epoch": 0.7641445126107703,
"grad_norm": 1.7962192296981812,
"learning_rate": 1.4189730930880799e-05,
"loss": 1.4974309206008911,
"step": 1121
},
{
"epoch": 0.7648261758691206,
"grad_norm": 2.1655869483947754,
"learning_rate": 1.417970179893904e-05,
"loss": 1.0126588344573975,
"step": 1122
},
{
"epoch": 0.765507839127471,
"grad_norm": 3.2068166732788086,
"learning_rate": 1.4169667570398367e-05,
"loss": 1.8077328205108643,
"step": 1123
},
{
"epoch": 0.7661895023858214,
"grad_norm": 2.407762289047241,
"learning_rate": 1.4159628257494195e-05,
"loss": 0.6827515363693237,
"step": 1124
},
{
"epoch": 0.7668711656441718,
"grad_norm": 3.49932599067688,
"learning_rate": 1.4149583872468165e-05,
"loss": 1.3755052089691162,
"step": 1125
},
{
"epoch": 0.7675528289025222,
"grad_norm": 1.8937991857528687,
"learning_rate": 1.4139534427568073e-05,
"loss": 1.2675697803497314,
"step": 1126
},
{
"epoch": 0.7682344921608726,
"grad_norm": 1.550775170326233,
"learning_rate": 1.4129479935047914e-05,
"loss": 0.6430014371871948,
"step": 1127
},
{
"epoch": 0.7689161554192229,
"grad_norm": 2.033308744430542,
"learning_rate": 1.4119420407167817e-05,
"loss": 0.9698971509933472,
"step": 1128
},
{
"epoch": 0.7695978186775733,
"grad_norm": 1.069423794746399,
"learning_rate": 1.4109355856194062e-05,
"loss": 0.7354646325111389,
"step": 1129
},
{
"epoch": 0.7702794819359237,
"grad_norm": 2.814561128616333,
"learning_rate": 1.4099286294399051e-05,
"loss": 1.539209008216858,
"step": 1130
},
{
"epoch": 0.7709611451942741,
"grad_norm": 3.4009978771209717,
"learning_rate": 1.4089211734061294e-05,
"loss": 0.6647955179214478,
"step": 1131
},
{
"epoch": 0.7716428084526245,
"grad_norm": 1.6567187309265137,
"learning_rate": 1.4079132187465403e-05,
"loss": 0.864957332611084,
"step": 1132
},
{
"epoch": 0.7723244717109747,
"grad_norm": 2.7520089149475098,
"learning_rate": 1.4069047666902056e-05,
"loss": 0.6806797385215759,
"step": 1133
},
{
"epoch": 0.7730061349693251,
"grad_norm": 6.400692939758301,
"learning_rate": 1.405895818466801e-05,
"loss": 2.2545909881591797,
"step": 1134
},
{
"epoch": 0.7736877982276755,
"grad_norm": 2.637324333190918,
"learning_rate": 1.404886375306607e-05,
"loss": 1.1330559253692627,
"step": 1135
},
{
"epoch": 0.7743694614860259,
"grad_norm": 4.983392238616943,
"learning_rate": 1.403876438440507e-05,
"loss": 2.1453890800476074,
"step": 1136
},
{
"epoch": 0.7750511247443763,
"grad_norm": 24.753963470458984,
"learning_rate": 1.4028660090999866e-05,
"loss": 0.6390861868858337,
"step": 1137
},
{
"epoch": 0.7757327880027266,
"grad_norm": 2.1686322689056396,
"learning_rate": 1.4018550885171322e-05,
"loss": 1.2475088834762573,
"step": 1138
},
{
"epoch": 0.776414451261077,
"grad_norm": 1.1742987632751465,
"learning_rate": 1.4008436779246288e-05,
"loss": 0.4145359694957733,
"step": 1139
},
{
"epoch": 0.7770961145194274,
"grad_norm": 3.770195722579956,
"learning_rate": 1.3998317785557597e-05,
"loss": 1.0192723274230957,
"step": 1140
},
{
"epoch": 0.7777777777777778,
"grad_norm": 3.6359922885894775,
"learning_rate": 1.3988193916444036e-05,
"loss": 1.6351332664489746,
"step": 1141
},
{
"epoch": 0.7784594410361282,
"grad_norm": 1.383095383644104,
"learning_rate": 1.3978065184250334e-05,
"loss": 0.6680575609207153,
"step": 1142
},
{
"epoch": 0.7791411042944786,
"grad_norm": 3.0580966472625732,
"learning_rate": 1.396793160132715e-05,
"loss": 2.0260071754455566,
"step": 1143
},
{
"epoch": 0.7798227675528289,
"grad_norm": 1.9012495279312134,
"learning_rate": 1.3957793180031067e-05,
"loss": 0.7958223819732666,
"step": 1144
},
{
"epoch": 0.7805044308111793,
"grad_norm": 3.0579073429107666,
"learning_rate": 1.3947649932724563e-05,
"loss": 1.2729113101959229,
"step": 1145
},
{
"epoch": 0.7811860940695297,
"grad_norm": 3.44350266456604,
"learning_rate": 1.3937501871775995e-05,
"loss": 1.5779197216033936,
"step": 1146
},
{
"epoch": 0.78186775732788,
"grad_norm": 1.746261715888977,
"learning_rate": 1.3927349009559597e-05,
"loss": 0.3372318744659424,
"step": 1147
},
{
"epoch": 0.7825494205862304,
"grad_norm": 5.006249904632568,
"learning_rate": 1.3917191358455453e-05,
"loss": 2.2594876289367676,
"step": 1148
},
{
"epoch": 0.7832310838445807,
"grad_norm": 5.940740585327148,
"learning_rate": 1.3907028930849489e-05,
"loss": 0.8546584844589233,
"step": 1149
},
{
"epoch": 0.7839127471029311,
"grad_norm": 3.4024150371551514,
"learning_rate": 1.3896861739133456e-05,
"loss": 1.9201159477233887,
"step": 1150
},
{
"epoch": 0.7845944103612815,
"grad_norm": 1.3676657676696777,
"learning_rate": 1.388668979570491e-05,
"loss": 0.8639696836471558,
"step": 1151
},
{
"epoch": 0.7852760736196319,
"grad_norm": 1.9111442565917969,
"learning_rate": 1.3876513112967208e-05,
"loss": 0.7849295139312744,
"step": 1152
},
{
"epoch": 0.7859577368779823,
"grad_norm": 1.646479606628418,
"learning_rate": 1.3866331703329477e-05,
"loss": 0.9975169897079468,
"step": 1153
},
{
"epoch": 0.7866394001363326,
"grad_norm": 1.8222453594207764,
"learning_rate": 1.3856145579206612e-05,
"loss": 1.2962346076965332,
"step": 1154
},
{
"epoch": 0.787321063394683,
"grad_norm": 3.7591559886932373,
"learning_rate": 1.384595475301926e-05,
"loss": 1.5346903800964355,
"step": 1155
},
{
"epoch": 0.7880027266530334,
"grad_norm": 1.4336442947387695,
"learning_rate": 1.38357592371938e-05,
"loss": 1.160192847251892,
"step": 1156
},
{
"epoch": 0.7886843899113838,
"grad_norm": 2.9270148277282715,
"learning_rate": 1.3825559044162327e-05,
"loss": 1.0325207710266113,
"step": 1157
},
{
"epoch": 0.7893660531697342,
"grad_norm": 3.7184078693389893,
"learning_rate": 1.381535418636264e-05,
"loss": 0.982779860496521,
"step": 1158
},
{
"epoch": 0.7900477164280846,
"grad_norm": 1.879891276359558,
"learning_rate": 1.3805144676238225e-05,
"loss": 2.0673606395721436,
"step": 1159
},
{
"epoch": 0.7907293796864349,
"grad_norm": 1.9370105266571045,
"learning_rate": 1.3794930526238246e-05,
"loss": 0.573197603225708,
"step": 1160
},
{
"epoch": 0.7914110429447853,
"grad_norm": 2.1516077518463135,
"learning_rate": 1.3784711748817519e-05,
"loss": 0.9432398080825806,
"step": 1161
},
{
"epoch": 0.7920927062031357,
"grad_norm": 2.1923482418060303,
"learning_rate": 1.3774488356436505e-05,
"loss": 1.4591634273529053,
"step": 1162
},
{
"epoch": 0.792774369461486,
"grad_norm": 3.7260212898254395,
"learning_rate": 1.376426036156129e-05,
"loss": 1.241162896156311,
"step": 1163
},
{
"epoch": 0.7934560327198364,
"grad_norm": 2.8687682151794434,
"learning_rate": 1.3754027776663579e-05,
"loss": 1.5285322666168213,
"step": 1164
},
{
"epoch": 0.7941376959781867,
"grad_norm": 1.8364909887313843,
"learning_rate": 1.3743790614220664e-05,
"loss": 0.9934275150299072,
"step": 1165
},
{
"epoch": 0.7948193592365371,
"grad_norm": 2.3077330589294434,
"learning_rate": 1.3733548886715427e-05,
"loss": 1.2558547258377075,
"step": 1166
},
{
"epoch": 0.7955010224948875,
"grad_norm": 1.8851512670516968,
"learning_rate": 1.3723302606636311e-05,
"loss": 0.9629714488983154,
"step": 1167
},
{
"epoch": 0.7961826857532379,
"grad_norm": 2.393500804901123,
"learning_rate": 1.3713051786477319e-05,
"loss": 1.1742806434631348,
"step": 1168
},
{
"epoch": 0.7968643490115883,
"grad_norm": 3.2605621814727783,
"learning_rate": 1.3702796438737974e-05,
"loss": 2.027634382247925,
"step": 1169
},
{
"epoch": 0.7975460122699386,
"grad_norm": 2.644625425338745,
"learning_rate": 1.3692536575923334e-05,
"loss": 1.0252785682678223,
"step": 1170
},
{
"epoch": 0.798227675528289,
"grad_norm": 3.954820394515991,
"learning_rate": 1.3682272210543959e-05,
"loss": 2.3715176582336426,
"step": 1171
},
{
"epoch": 0.7989093387866394,
"grad_norm": 2.5120863914489746,
"learning_rate": 1.3672003355115897e-05,
"loss": 2.089763641357422,
"step": 1172
},
{
"epoch": 0.7995910020449898,
"grad_norm": 2.659560441970825,
"learning_rate": 1.3661730022160673e-05,
"loss": 0.6313329339027405,
"step": 1173
},
{
"epoch": 0.8002726653033402,
"grad_norm": 3.3629090785980225,
"learning_rate": 1.365145222420527e-05,
"loss": 0.580939531326294,
"step": 1174
},
{
"epoch": 0.8009543285616906,
"grad_norm": 2.3204710483551025,
"learning_rate": 1.3641169973782117e-05,
"loss": 1.5531153678894043,
"step": 1175
},
{
"epoch": 0.8016359918200409,
"grad_norm": 3.2493536472320557,
"learning_rate": 1.3630883283429071e-05,
"loss": 1.7264301776885986,
"step": 1176
},
{
"epoch": 0.8023176550783913,
"grad_norm": 2.0848870277404785,
"learning_rate": 1.3620592165689405e-05,
"loss": 1.3860576152801514,
"step": 1177
},
{
"epoch": 0.8029993183367417,
"grad_norm": 3.532919406890869,
"learning_rate": 1.3610296633111788e-05,
"loss": 0.6948189735412598,
"step": 1178
},
{
"epoch": 0.803680981595092,
"grad_norm": 1.9613711833953857,
"learning_rate": 1.3599996698250274e-05,
"loss": 1.3784890174865723,
"step": 1179
},
{
"epoch": 0.8043626448534424,
"grad_norm": 2.8494956493377686,
"learning_rate": 1.3589692373664288e-05,
"loss": 0.40545928478240967,
"step": 1180
},
{
"epoch": 0.8050443081117927,
"grad_norm": 1.2043472528457642,
"learning_rate": 1.3579383671918598e-05,
"loss": 0.9535353183746338,
"step": 1181
},
{
"epoch": 0.8057259713701431,
"grad_norm": 2.0384488105773926,
"learning_rate": 1.3569070605583319e-05,
"loss": 1.616403341293335,
"step": 1182
},
{
"epoch": 0.8064076346284935,
"grad_norm": 2.768143892288208,
"learning_rate": 1.3558753187233881e-05,
"loss": 2.2363321781158447,
"step": 1183
},
{
"epoch": 0.8070892978868439,
"grad_norm": 2.2129340171813965,
"learning_rate": 1.3548431429451032e-05,
"loss": 1.5187221765518188,
"step": 1184
},
{
"epoch": 0.8077709611451943,
"grad_norm": 2.341991424560547,
"learning_rate": 1.3538105344820798e-05,
"loss": 1.0091967582702637,
"step": 1185
},
{
"epoch": 0.8084526244035446,
"grad_norm": 2.6940460205078125,
"learning_rate": 1.352777494593449e-05,
"loss": 1.6440155506134033,
"step": 1186
},
{
"epoch": 0.809134287661895,
"grad_norm": 2.276599168777466,
"learning_rate": 1.3517440245388672e-05,
"loss": 1.2129032611846924,
"step": 1187
},
{
"epoch": 0.8098159509202454,
"grad_norm": 3.9329209327697754,
"learning_rate": 1.350710125578516e-05,
"loss": 1.818249225616455,
"step": 1188
},
{
"epoch": 0.8104976141785958,
"grad_norm": 2.928197145462036,
"learning_rate": 1.3496757989730997e-05,
"loss": 1.140410304069519,
"step": 1189
},
{
"epoch": 0.8111792774369462,
"grad_norm": 2.0259079933166504,
"learning_rate": 1.3486410459838448e-05,
"loss": 0.7515283823013306,
"step": 1190
},
{
"epoch": 0.8118609406952966,
"grad_norm": 5.171846389770508,
"learning_rate": 1.347605867872496e-05,
"loss": 1.5359857082366943,
"step": 1191
},
{
"epoch": 0.8125426039536469,
"grad_norm": 2.310758590698242,
"learning_rate": 1.346570265901318e-05,
"loss": 1.3270797729492188,
"step": 1192
},
{
"epoch": 0.8132242672119973,
"grad_norm": 2.1669437885284424,
"learning_rate": 1.3455342413330916e-05,
"loss": 1.3661367893218994,
"step": 1193
},
{
"epoch": 0.8139059304703476,
"grad_norm": 2.030578136444092,
"learning_rate": 1.3444977954311133e-05,
"loss": 1.123007893562317,
"step": 1194
},
{
"epoch": 0.814587593728698,
"grad_norm": 3.647148847579956,
"learning_rate": 1.343460929459193e-05,
"loss": 1.5400652885437012,
"step": 1195
},
{
"epoch": 0.8152692569870484,
"grad_norm": 1.4944745302200317,
"learning_rate": 1.3424236446816528e-05,
"loss": 1.2711783647537231,
"step": 1196
},
{
"epoch": 0.8159509202453987,
"grad_norm": 4.806166172027588,
"learning_rate": 1.3413859423633259e-05,
"loss": 1.698742389678955,
"step": 1197
},
{
"epoch": 0.8166325835037491,
"grad_norm": 1.8429958820343018,
"learning_rate": 1.3403478237695542e-05,
"loss": 1.910172939300537,
"step": 1198
},
{
"epoch": 0.8173142467620995,
"grad_norm": 3.9592504501342773,
"learning_rate": 1.3393092901661873e-05,
"loss": 1.5437757968902588,
"step": 1199
},
{
"epoch": 0.8179959100204499,
"grad_norm": 3.513111114501953,
"learning_rate": 1.3382703428195812e-05,
"loss": 1.042831540107727,
"step": 1200
},
{
"epoch": 0.8186775732788003,
"grad_norm": 4.69813871383667,
"learning_rate": 1.3372309829965957e-05,
"loss": 2.1290884017944336,
"step": 1201
},
{
"epoch": 0.8193592365371506,
"grad_norm": 4.101344585418701,
"learning_rate": 1.3361912119645943e-05,
"loss": 0.5426227450370789,
"step": 1202
},
{
"epoch": 0.820040899795501,
"grad_norm": 3.597198724746704,
"learning_rate": 1.3351510309914415e-05,
"loss": 0.7113486528396606,
"step": 1203
},
{
"epoch": 0.8207225630538514,
"grad_norm": 2.5351171493530273,
"learning_rate": 1.3341104413455014e-05,
"loss": 0.8288606405258179,
"step": 1204
},
{
"epoch": 0.8214042263122018,
"grad_norm": 2.3007466793060303,
"learning_rate": 1.3330694442956376e-05,
"loss": 1.4681382179260254,
"step": 1205
},
{
"epoch": 0.8220858895705522,
"grad_norm": 2.452038049697876,
"learning_rate": 1.3320280411112092e-05,
"loss": 1.3009843826293945,
"step": 1206
},
{
"epoch": 0.8227675528289026,
"grad_norm": 2.4166622161865234,
"learning_rate": 1.3309862330620709e-05,
"loss": 0.875541090965271,
"step": 1207
},
{
"epoch": 0.8234492160872529,
"grad_norm": 3.604154348373413,
"learning_rate": 1.3299440214185707e-05,
"loss": 1.0196988582611084,
"step": 1208
},
{
"epoch": 0.8241308793456033,
"grad_norm": 2.63765025138855,
"learning_rate": 1.3289014074515505e-05,
"loss": 0.8730734586715698,
"step": 1209
},
{
"epoch": 0.8248125426039536,
"grad_norm": 2.459676504135132,
"learning_rate": 1.3278583924323405e-05,
"loss": 2.113161087036133,
"step": 1210
},
{
"epoch": 0.825494205862304,
"grad_norm": 2.1859536170959473,
"learning_rate": 1.326814977632761e-05,
"loss": 1.2790409326553345,
"step": 1211
},
{
"epoch": 0.8261758691206544,
"grad_norm": 2.1957504749298096,
"learning_rate": 1.3257711643251201e-05,
"loss": 0.5538440346717834,
"step": 1212
},
{
"epoch": 0.8268575323790047,
"grad_norm": 1.5440486669540405,
"learning_rate": 1.3247269537822109e-05,
"loss": 0.7184191942214966,
"step": 1213
},
{
"epoch": 0.8275391956373551,
"grad_norm": 1.557157278060913,
"learning_rate": 1.3236823472773116e-05,
"loss": 0.47062918543815613,
"step": 1214
},
{
"epoch": 0.8282208588957055,
"grad_norm": 5.947640419006348,
"learning_rate": 1.3226373460841835e-05,
"loss": 1.872814655303955,
"step": 1215
},
{
"epoch": 0.8289025221540559,
"grad_norm": 1.9622546434402466,
"learning_rate": 1.3215919514770676e-05,
"loss": 1.1466459035873413,
"step": 1216
},
{
"epoch": 0.8295841854124063,
"grad_norm": 1.643409013748169,
"learning_rate": 1.3205461647306872e-05,
"loss": 1.1066665649414062,
"step": 1217
},
{
"epoch": 0.8302658486707567,
"grad_norm": 1.5371270179748535,
"learning_rate": 1.3194999871202408e-05,
"loss": 0.630884051322937,
"step": 1218
},
{
"epoch": 0.830947511929107,
"grad_norm": 1.544770359992981,
"learning_rate": 1.3184534199214059e-05,
"loss": 0.7396583557128906,
"step": 1219
},
{
"epoch": 0.8316291751874574,
"grad_norm": 3.373950481414795,
"learning_rate": 1.3174064644103334e-05,
"loss": 1.3077354431152344,
"step": 1220
},
{
"epoch": 0.8323108384458078,
"grad_norm": 1.815501093864441,
"learning_rate": 1.3163591218636494e-05,
"loss": 1.2874424457550049,
"step": 1221
},
{
"epoch": 0.8329925017041582,
"grad_norm": 5.293972015380859,
"learning_rate": 1.31531139355845e-05,
"loss": 1.6489955186843872,
"step": 1222
},
{
"epoch": 0.8336741649625086,
"grad_norm": 1.9330779314041138,
"learning_rate": 1.3142632807723035e-05,
"loss": 1.4095408916473389,
"step": 1223
},
{
"epoch": 0.8343558282208589,
"grad_norm": 2.491067886352539,
"learning_rate": 1.3132147847832453e-05,
"loss": 1.4048972129821777,
"step": 1224
},
{
"epoch": 0.8350374914792092,
"grad_norm": 2.374852180480957,
"learning_rate": 1.3121659068697797e-05,
"loss": 1.1800034046173096,
"step": 1225
},
{
"epoch": 0.8357191547375596,
"grad_norm": 3.167478561401367,
"learning_rate": 1.3111166483108753e-05,
"loss": 1.1449928283691406,
"step": 1226
},
{
"epoch": 0.83640081799591,
"grad_norm": 1.7197049856185913,
"learning_rate": 1.310067010385966e-05,
"loss": 1.2595001459121704,
"step": 1227
},
{
"epoch": 0.8370824812542604,
"grad_norm": 2.305821657180786,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.5387115478515625,
"step": 1228
},
{
"epoch": 0.8377641445126107,
"grad_norm": 1.9236418008804321,
"learning_rate": 1.307966601558177e-05,
"loss": 1.238250970840454,
"step": 1229
},
{
"epoch": 0.8384458077709611,
"grad_norm": 1.9397454261779785,
"learning_rate": 1.306915833216471e-05,
"loss": 1.2780492305755615,
"step": 1230
},
{
"epoch": 0.8391274710293115,
"grad_norm": 5.5808820724487305,
"learning_rate": 1.3058646906311032e-05,
"loss": 0.4606386423110962,
"step": 1231
},
{
"epoch": 0.8398091342876619,
"grad_norm": 1.8887132406234741,
"learning_rate": 1.304813175083805e-05,
"loss": 0.5161309838294983,
"step": 1232
},
{
"epoch": 0.8404907975460123,
"grad_norm": 2.4710726737976074,
"learning_rate": 1.3037612878567623e-05,
"loss": 0.7300988435745239,
"step": 1233
},
{
"epoch": 0.8411724608043627,
"grad_norm": 3.93088960647583,
"learning_rate": 1.3027090302326127e-05,
"loss": 1.6448276042938232,
"step": 1234
},
{
"epoch": 0.841854124062713,
"grad_norm": 5.784260272979736,
"learning_rate": 1.3016564034944473e-05,
"loss": 2.1232104301452637,
"step": 1235
},
{
"epoch": 0.8425357873210634,
"grad_norm": 9.557524681091309,
"learning_rate": 1.3006034089258059e-05,
"loss": 0.8751978874206543,
"step": 1236
},
{
"epoch": 0.8432174505794138,
"grad_norm": 3.1583986282348633,
"learning_rate": 1.2995500478106781e-05,
"loss": 1.4895182847976685,
"step": 1237
},
{
"epoch": 0.8438991138377642,
"grad_norm": 1.9370651245117188,
"learning_rate": 1.2984963214335e-05,
"loss": 1.2552303075790405,
"step": 1238
},
{
"epoch": 0.8445807770961146,
"grad_norm": 2.8927485942840576,
"learning_rate": 1.2974422310791524e-05,
"loss": 0.8721221089363098,
"step": 1239
},
{
"epoch": 0.8452624403544649,
"grad_norm": 3.3905813694000244,
"learning_rate": 1.29638777803296e-05,
"loss": 2.2706828117370605,
"step": 1240
},
{
"epoch": 0.8459441036128152,
"grad_norm": 2.136692762374878,
"learning_rate": 1.2953329635806914e-05,
"loss": 0.9477517604827881,
"step": 1241
},
{
"epoch": 0.8466257668711656,
"grad_norm": 6.7342963218688965,
"learning_rate": 1.2942777890085538e-05,
"loss": 2.1006269454956055,
"step": 1242
},
{
"epoch": 0.847307430129516,
"grad_norm": 6.7248616218566895,
"learning_rate": 1.2932222556031946e-05,
"loss": 2.246072292327881,
"step": 1243
},
{
"epoch": 0.8479890933878664,
"grad_norm": 5.852639675140381,
"learning_rate": 1.2921663646516985e-05,
"loss": 1.0963563919067383,
"step": 1244
},
{
"epoch": 0.8486707566462167,
"grad_norm": 3.222761631011963,
"learning_rate": 1.2911101174415861e-05,
"loss": 1.497108817100525,
"step": 1245
},
{
"epoch": 0.8493524199045671,
"grad_norm": 2.8850924968719482,
"learning_rate": 1.290053515260813e-05,
"loss": 1.030866026878357,
"step": 1246
},
{
"epoch": 0.8500340831629175,
"grad_norm": 2.141866445541382,
"learning_rate": 1.288996559397767e-05,
"loss": 1.2230726480484009,
"step": 1247
},
{
"epoch": 0.8507157464212679,
"grad_norm": 2.458935499191284,
"learning_rate": 1.2879392511412668e-05,
"loss": 1.5903170108795166,
"step": 1248
},
{
"epoch": 0.8513974096796183,
"grad_norm": 2.7172353267669678,
"learning_rate": 1.2868815917805619e-05,
"loss": 0.705411970615387,
"step": 1249
},
{
"epoch": 0.8520790729379687,
"grad_norm": 1.890684723854065,
"learning_rate": 1.2858235826053294e-05,
"loss": 0.7464626431465149,
"step": 1250
},
{
"epoch": 0.852760736196319,
"grad_norm": 2.983992576599121,
"learning_rate": 1.2847652249056726e-05,
"loss": 1.1335132122039795,
"step": 1251
},
{
"epoch": 0.8534423994546694,
"grad_norm": 2.6148171424865723,
"learning_rate": 1.2837065199721204e-05,
"loss": 0.6404839754104614,
"step": 1252
},
{
"epoch": 0.8541240627130198,
"grad_norm": 2.2352583408355713,
"learning_rate": 1.2826474690956243e-05,
"loss": 0.8552806377410889,
"step": 1253
},
{
"epoch": 0.8548057259713702,
"grad_norm": 1.6480870246887207,
"learning_rate": 1.2815880735675588e-05,
"loss": 0.601885199546814,
"step": 1254
},
{
"epoch": 0.8554873892297206,
"grad_norm": 11.749967575073242,
"learning_rate": 1.2805283346797179e-05,
"loss": 1.6991385221481323,
"step": 1255
},
{
"epoch": 0.8561690524880708,
"grad_norm": 2.079409122467041,
"learning_rate": 1.279468253724314e-05,
"loss": 1.1878442764282227,
"step": 1256
},
{
"epoch": 0.8568507157464212,
"grad_norm": 4.077375888824463,
"learning_rate": 1.2784078319939769e-05,
"loss": 1.33461594581604,
"step": 1257
},
{
"epoch": 0.8575323790047716,
"grad_norm": 3.0802202224731445,
"learning_rate": 1.2773470707817524e-05,
"loss": 1.5822620391845703,
"step": 1258
},
{
"epoch": 0.858214042263122,
"grad_norm": 3.6835556030273438,
"learning_rate": 1.2762859713810998e-05,
"loss": 1.514790415763855,
"step": 1259
},
{
"epoch": 0.8588957055214724,
"grad_norm": 2.215324878692627,
"learning_rate": 1.2752245350858905e-05,
"loss": 1.1391081809997559,
"step": 1260
},
{
"epoch": 0.8595773687798227,
"grad_norm": 4.364893436431885,
"learning_rate": 1.2741627631904077e-05,
"loss": 1.7329888343811035,
"step": 1261
},
{
"epoch": 0.8602590320381731,
"grad_norm": 3.4366931915283203,
"learning_rate": 1.2731006569893427e-05,
"loss": 1.4349662065505981,
"step": 1262
},
{
"epoch": 0.8609406952965235,
"grad_norm": 4.452749729156494,
"learning_rate": 1.272038217777795e-05,
"loss": 1.5987038612365723,
"step": 1263
},
{
"epoch": 0.8616223585548739,
"grad_norm": 232.9865264892578,
"learning_rate": 1.27097544685127e-05,
"loss": 0.9670251607894897,
"step": 1264
},
{
"epoch": 0.8623040218132243,
"grad_norm": 3.7537894248962402,
"learning_rate": 1.2699123455056777e-05,
"loss": 1.1016144752502441,
"step": 1265
},
{
"epoch": 0.8629856850715747,
"grad_norm": 2.8255839347839355,
"learning_rate": 1.268848915037331e-05,
"loss": 0.9010818600654602,
"step": 1266
},
{
"epoch": 0.863667348329925,
"grad_norm": 2.359464406967163,
"learning_rate": 1.2677851567429442e-05,
"loss": 0.7024168968200684,
"step": 1267
},
{
"epoch": 0.8643490115882754,
"grad_norm": 2.3273167610168457,
"learning_rate": 1.2667210719196308e-05,
"loss": 1.6643811464309692,
"step": 1268
},
{
"epoch": 0.8650306748466258,
"grad_norm": 2.9757978916168213,
"learning_rate": 1.2656566618649031e-05,
"loss": 1.4444563388824463,
"step": 1269
},
{
"epoch": 0.8657123381049762,
"grad_norm": 1.7979950904846191,
"learning_rate": 1.26459192787667e-05,
"loss": 0.7493287920951843,
"step": 1270
},
{
"epoch": 0.8663940013633266,
"grad_norm": 2.6119706630706787,
"learning_rate": 1.263526871253235e-05,
"loss": 1.1911826133728027,
"step": 1271
},
{
"epoch": 0.8670756646216768,
"grad_norm": 1.5572715997695923,
"learning_rate": 1.2624614932932953e-05,
"loss": 1.0380706787109375,
"step": 1272
},
{
"epoch": 0.8677573278800272,
"grad_norm": 2.26962947845459,
"learning_rate": 1.261395795295939e-05,
"loss": 1.4516929388046265,
"step": 1273
},
{
"epoch": 0.8684389911383776,
"grad_norm": 2.810452938079834,
"learning_rate": 1.260329778560646e-05,
"loss": 0.6930328607559204,
"step": 1274
},
{
"epoch": 0.869120654396728,
"grad_norm": 2.030446767807007,
"learning_rate": 1.2592634443872842e-05,
"loss": 0.3848379850387573,
"step": 1275
},
{
"epoch": 0.8698023176550784,
"grad_norm": 2.2821154594421387,
"learning_rate": 1.2581967940761079e-05,
"loss": 0.6541157364845276,
"step": 1276
},
{
"epoch": 0.8704839809134287,
"grad_norm": 3.862823247909546,
"learning_rate": 1.257129828927758e-05,
"loss": 1.4444286823272705,
"step": 1277
},
{
"epoch": 0.8711656441717791,
"grad_norm": 1.122416615486145,
"learning_rate": 1.2560625502432581e-05,
"loss": 0.7840420007705688,
"step": 1278
},
{
"epoch": 0.8718473074301295,
"grad_norm": 2.7760159969329834,
"learning_rate": 1.2549949593240156e-05,
"loss": 1.672440528869629,
"step": 1279
},
{
"epoch": 0.8725289706884799,
"grad_norm": 1.3836902379989624,
"learning_rate": 1.2539270574718172e-05,
"loss": 0.9528952836990356,
"step": 1280
},
{
"epoch": 0.8732106339468303,
"grad_norm": 2.4517650604248047,
"learning_rate": 1.2528588459888291e-05,
"loss": 1.0446099042892456,
"step": 1281
},
{
"epoch": 0.8738922972051807,
"grad_norm": 1.6478908061981201,
"learning_rate": 1.2517903261775963e-05,
"loss": 1.1493809223175049,
"step": 1282
},
{
"epoch": 0.874573960463531,
"grad_norm": 1.7826266288757324,
"learning_rate": 1.2507214993410382e-05,
"loss": 0.873786449432373,
"step": 1283
},
{
"epoch": 0.8752556237218814,
"grad_norm": 1.3265115022659302,
"learning_rate": 1.2496523667824487e-05,
"loss": 0.9068435430526733,
"step": 1284
},
{
"epoch": 0.8759372869802318,
"grad_norm": 4.726826190948486,
"learning_rate": 1.2485829298054952e-05,
"loss": 2.038616418838501,
"step": 1285
},
{
"epoch": 0.8766189502385822,
"grad_norm": 2.2327160835266113,
"learning_rate": 1.2475131897142165e-05,
"loss": 1.1699340343475342,
"step": 1286
},
{
"epoch": 0.8773006134969326,
"grad_norm": 2.6090691089630127,
"learning_rate": 1.2464431478130204e-05,
"loss": 1.4517043828964233,
"step": 1287
},
{
"epoch": 0.8779822767552828,
"grad_norm": 3.6594088077545166,
"learning_rate": 1.2453728054066825e-05,
"loss": 1.3603757619857788,
"step": 1288
},
{
"epoch": 0.8786639400136332,
"grad_norm": 1.8763833045959473,
"learning_rate": 1.244302163800345e-05,
"loss": 1.105027198791504,
"step": 1289
},
{
"epoch": 0.8793456032719836,
"grad_norm": 3.21225643157959,
"learning_rate": 1.2432312242995158e-05,
"loss": 1.1615551710128784,
"step": 1290
},
{
"epoch": 0.880027266530334,
"grad_norm": 1.9490127563476562,
"learning_rate": 1.2421599882100647e-05,
"loss": 1.3426179885864258,
"step": 1291
},
{
"epoch": 0.8807089297886844,
"grad_norm": 2.9683022499084473,
"learning_rate": 1.2410884568382245e-05,
"loss": 0.8526750802993774,
"step": 1292
},
{
"epoch": 0.8813905930470347,
"grad_norm": 1.6657367944717407,
"learning_rate": 1.2400166314905868e-05,
"loss": 0.974453330039978,
"step": 1293
},
{
"epoch": 0.8820722563053851,
"grad_norm": 2.1943328380584717,
"learning_rate": 1.2389445134741022e-05,
"loss": 1.0904216766357422,
"step": 1294
},
{
"epoch": 0.8827539195637355,
"grad_norm": 3.002743721008301,
"learning_rate": 1.2378721040960788e-05,
"loss": 1.1856180429458618,
"step": 1295
},
{
"epoch": 0.8834355828220859,
"grad_norm": 3.0109915733337402,
"learning_rate": 1.2367994046641787e-05,
"loss": 1.7809927463531494,
"step": 1296
},
{
"epoch": 0.8841172460804363,
"grad_norm": 1.7773187160491943,
"learning_rate": 1.2357264164864186e-05,
"loss": 0.9765096306800842,
"step": 1297
},
{
"epoch": 0.8847989093387867,
"grad_norm": 1.5350987911224365,
"learning_rate": 1.2346531408711675e-05,
"loss": 1.1852858066558838,
"step": 1298
},
{
"epoch": 0.885480572597137,
"grad_norm": 2.7784996032714844,
"learning_rate": 1.233579579127144e-05,
"loss": 1.5897890329360962,
"step": 1299
},
{
"epoch": 0.8861622358554874,
"grad_norm": 5.824115753173828,
"learning_rate": 1.232505732563416e-05,
"loss": 1.3956981897354126,
"step": 1300
},
{
"epoch": 0.8868438991138378,
"grad_norm": 1.4179086685180664,
"learning_rate": 1.2314316024893987e-05,
"loss": 0.5982742309570312,
"step": 1301
},
{
"epoch": 0.8875255623721882,
"grad_norm": 5.166876792907715,
"learning_rate": 1.2303571902148532e-05,
"loss": 1.2101175785064697,
"step": 1302
},
{
"epoch": 0.8882072256305386,
"grad_norm": 2.7313640117645264,
"learning_rate": 1.2292824970498847e-05,
"loss": 1.0363984107971191,
"step": 1303
},
{
"epoch": 0.8888888888888888,
"grad_norm": 2.3922321796417236,
"learning_rate": 1.2282075243049408e-05,
"loss": 1.292094349861145,
"step": 1304
},
{
"epoch": 0.8895705521472392,
"grad_norm": 2.692531108856201,
"learning_rate": 1.2271322732908091e-05,
"loss": 0.5840368866920471,
"step": 1305
},
{
"epoch": 0.8902522154055896,
"grad_norm": 1.4718867540359497,
"learning_rate": 1.2260567453186185e-05,
"loss": 0.5893124341964722,
"step": 1306
},
{
"epoch": 0.89093387866394,
"grad_norm": 4.281642436981201,
"learning_rate": 1.2249809416998339e-05,
"loss": 1.3083157539367676,
"step": 1307
},
{
"epoch": 0.8916155419222904,
"grad_norm": 1.3782727718353271,
"learning_rate": 1.2239048637462572e-05,
"loss": 0.3187441825866699,
"step": 1308
},
{
"epoch": 0.8922972051806408,
"grad_norm": 6.625371932983398,
"learning_rate": 1.2228285127700244e-05,
"loss": 2.061591386795044,
"step": 1309
},
{
"epoch": 0.8929788684389911,
"grad_norm": 1.15508234500885,
"learning_rate": 1.2217518900836045e-05,
"loss": 0.7010345458984375,
"step": 1310
},
{
"epoch": 0.8936605316973415,
"grad_norm": 2.2813875675201416,
"learning_rate": 1.2206749969997979e-05,
"loss": 1.2704284191131592,
"step": 1311
},
{
"epoch": 0.8943421949556919,
"grad_norm": 1.3211407661437988,
"learning_rate": 1.2195978348317347e-05,
"loss": 0.8025050163269043,
"step": 1312
},
{
"epoch": 0.8950238582140423,
"grad_norm": 3.8039255142211914,
"learning_rate": 1.2185204048928729e-05,
"loss": 1.2957152128219604,
"step": 1313
},
{
"epoch": 0.8957055214723927,
"grad_norm": 2.144146203994751,
"learning_rate": 1.2174427084969973e-05,
"loss": 1.082507610321045,
"step": 1314
},
{
"epoch": 0.896387184730743,
"grad_norm": 2.7374770641326904,
"learning_rate": 1.2163647469582181e-05,
"loss": 1.42399001121521,
"step": 1315
},
{
"epoch": 0.8970688479890934,
"grad_norm": 6.868131637573242,
"learning_rate": 1.2152865215909673e-05,
"loss": 1.5974717140197754,
"step": 1316
},
{
"epoch": 0.8977505112474438,
"grad_norm": 2.1185736656188965,
"learning_rate": 1.2142080337099998e-05,
"loss": 1.3208156824111938,
"step": 1317
},
{
"epoch": 0.8984321745057942,
"grad_norm": 4.58150053024292,
"learning_rate": 1.2131292846303901e-05,
"loss": 1.752542495727539,
"step": 1318
},
{
"epoch": 0.8991138377641446,
"grad_norm": 3.0584685802459717,
"learning_rate": 1.2120502756675324e-05,
"loss": 0.8576693534851074,
"step": 1319
},
{
"epoch": 0.8997955010224948,
"grad_norm": 2.083662748336792,
"learning_rate": 1.210971008137136e-05,
"loss": 0.9496646523475647,
"step": 1320
},
{
"epoch": 0.9004771642808452,
"grad_norm": 2.260331869125366,
"learning_rate": 1.2098914833552262e-05,
"loss": 1.085833191871643,
"step": 1321
},
{
"epoch": 0.9011588275391956,
"grad_norm": 2.5296528339385986,
"learning_rate": 1.2088117026381422e-05,
"loss": 1.541821002960205,
"step": 1322
},
{
"epoch": 0.901840490797546,
"grad_norm": 1.7845332622528076,
"learning_rate": 1.2077316673025354e-05,
"loss": 0.7951586246490479,
"step": 1323
},
{
"epoch": 0.9025221540558964,
"grad_norm": 1.9708597660064697,
"learning_rate": 1.2066513786653675e-05,
"loss": 1.301053762435913,
"step": 1324
},
{
"epoch": 0.9032038173142468,
"grad_norm": 2.5523624420166016,
"learning_rate": 1.2055708380439089e-05,
"loss": 1.4844532012939453,
"step": 1325
},
{
"epoch": 0.9038854805725971,
"grad_norm": 1.4186663627624512,
"learning_rate": 1.204490046755737e-05,
"loss": 0.7520268559455872,
"step": 1326
},
{
"epoch": 0.9045671438309475,
"grad_norm": 3.578901767730713,
"learning_rate": 1.2034090061187358e-05,
"loss": 1.8416643142700195,
"step": 1327
},
{
"epoch": 0.9052488070892979,
"grad_norm": 1.7172307968139648,
"learning_rate": 1.2023277174510923e-05,
"loss": 0.981949508190155,
"step": 1328
},
{
"epoch": 0.9059304703476483,
"grad_norm": 4.107685565948486,
"learning_rate": 1.2012461820712966e-05,
"loss": 2.1946287155151367,
"step": 1329
},
{
"epoch": 0.9066121336059987,
"grad_norm": 2.601182460784912,
"learning_rate": 1.2001644012981392e-05,
"loss": 1.226723313331604,
"step": 1330
},
{
"epoch": 0.907293796864349,
"grad_norm": 1.4821761846542358,
"learning_rate": 1.1990823764507108e-05,
"loss": 1.1278297901153564,
"step": 1331
},
{
"epoch": 0.9079754601226994,
"grad_norm": 3.0466654300689697,
"learning_rate": 1.1980001088483986e-05,
"loss": 1.0089236497879028,
"step": 1332
},
{
"epoch": 0.9086571233810498,
"grad_norm": 1.8594928979873657,
"learning_rate": 1.1969175998108857e-05,
"loss": 1.0941195487976074,
"step": 1333
},
{
"epoch": 0.9093387866394002,
"grad_norm": 6.154777526855469,
"learning_rate": 1.1958348506581503e-05,
"loss": 2.400118112564087,
"step": 1334
},
{
"epoch": 0.9100204498977505,
"grad_norm": 3.1990363597869873,
"learning_rate": 1.1947518627104637e-05,
"loss": 0.6849943995475769,
"step": 1335
},
{
"epoch": 0.9107021131561008,
"grad_norm": 2.930387496948242,
"learning_rate": 1.1936686372883877e-05,
"loss": 0.5548266768455505,
"step": 1336
},
{
"epoch": 0.9113837764144512,
"grad_norm": 2.311358690261841,
"learning_rate": 1.1925851757127735e-05,
"loss": 0.995979905128479,
"step": 1337
},
{
"epoch": 0.9120654396728016,
"grad_norm": 3.1525368690490723,
"learning_rate": 1.1915014793047606e-05,
"loss": 0.44815969467163086,
"step": 1338
},
{
"epoch": 0.912747102931152,
"grad_norm": 3.2352945804595947,
"learning_rate": 1.190417549385775e-05,
"loss": 1.2947889566421509,
"step": 1339
},
{
"epoch": 0.9134287661895024,
"grad_norm": 1.5861382484436035,
"learning_rate": 1.1893333872775275e-05,
"loss": 0.9021660089492798,
"step": 1340
},
{
"epoch": 0.9141104294478528,
"grad_norm": 4.334277629852295,
"learning_rate": 1.1882489943020115e-05,
"loss": 1.8276174068450928,
"step": 1341
},
{
"epoch": 0.9147920927062031,
"grad_norm": 2.043137788772583,
"learning_rate": 1.187164371781502e-05,
"loss": 0.3331798017024994,
"step": 1342
},
{
"epoch": 0.9154737559645535,
"grad_norm": 3.3330094814300537,
"learning_rate": 1.1860795210385547e-05,
"loss": 2.193763256072998,
"step": 1343
},
{
"epoch": 0.9161554192229039,
"grad_norm": 2.059584379196167,
"learning_rate": 1.1849944433960026e-05,
"loss": 0.5906944870948792,
"step": 1344
},
{
"epoch": 0.9168370824812543,
"grad_norm": 2.9722602367401123,
"learning_rate": 1.1839091401769559e-05,
"loss": 1.6175780296325684,
"step": 1345
},
{
"epoch": 0.9175187457396047,
"grad_norm": 3.0332117080688477,
"learning_rate": 1.1828236127047991e-05,
"loss": 1.4893518686294556,
"step": 1346
},
{
"epoch": 0.918200408997955,
"grad_norm": 1.3801240921020508,
"learning_rate": 1.1817378623031921e-05,
"loss": 1.227494716644287,
"step": 1347
},
{
"epoch": 0.9188820722563054,
"grad_norm": 1.5017725229263306,
"learning_rate": 1.1806518902960643e-05,
"loss": 0.8263968229293823,
"step": 1348
},
{
"epoch": 0.9195637355146558,
"grad_norm": 2.4677810668945312,
"learning_rate": 1.1795656980076164e-05,
"loss": 1.325760006904602,
"step": 1349
},
{
"epoch": 0.9202453987730062,
"grad_norm": 6.49123477935791,
"learning_rate": 1.1784792867623179e-05,
"loss": 1.9121781587600708,
"step": 1350
},
{
"epoch": 0.9209270620313565,
"grad_norm": 1.6628427505493164,
"learning_rate": 1.1773926578849049e-05,
"loss": 0.689493715763092,
"step": 1351
},
{
"epoch": 0.9216087252897068,
"grad_norm": 3.224045753479004,
"learning_rate": 1.1763058127003793e-05,
"loss": 1.6399683952331543,
"step": 1352
},
{
"epoch": 0.9222903885480572,
"grad_norm": 2.3120028972625732,
"learning_rate": 1.1752187525340061e-05,
"loss": 0.34457504749298096,
"step": 1353
},
{
"epoch": 0.9229720518064076,
"grad_norm": 2.854567289352417,
"learning_rate": 1.1741314787113129e-05,
"loss": 1.567234992980957,
"step": 1354
},
{
"epoch": 0.923653715064758,
"grad_norm": 2.151095151901245,
"learning_rate": 1.1730439925580876e-05,
"loss": 1.320873737335205,
"step": 1355
},
{
"epoch": 0.9243353783231084,
"grad_norm": 1.4363987445831299,
"learning_rate": 1.1719562954003774e-05,
"loss": 0.8847675919532776,
"step": 1356
},
{
"epoch": 0.9250170415814588,
"grad_norm": 3.134401321411133,
"learning_rate": 1.1708683885644865e-05,
"loss": 1.750314474105835,
"step": 1357
},
{
"epoch": 0.9256987048398091,
"grad_norm": 1.9255026578903198,
"learning_rate": 1.1697802733769745e-05,
"loss": 1.4583141803741455,
"step": 1358
},
{
"epoch": 0.9263803680981595,
"grad_norm": 3.1198015213012695,
"learning_rate": 1.1686919511646557e-05,
"loss": 1.9378410577774048,
"step": 1359
},
{
"epoch": 0.9270620313565099,
"grad_norm": 2.286750555038452,
"learning_rate": 1.1676034232545963e-05,
"loss": 0.9694126844406128,
"step": 1360
},
{
"epoch": 0.9277436946148603,
"grad_norm": 1.7516858577728271,
"learning_rate": 1.1665146909741134e-05,
"loss": 1.0411911010742188,
"step": 1361
},
{
"epoch": 0.9284253578732107,
"grad_norm": 3.267704963684082,
"learning_rate": 1.1654257556507735e-05,
"loss": 1.0651662349700928,
"step": 1362
},
{
"epoch": 0.929107021131561,
"grad_norm": 1.5885710716247559,
"learning_rate": 1.1643366186123913e-05,
"loss": 0.6454929113388062,
"step": 1363
},
{
"epoch": 0.9297886843899114,
"grad_norm": 3.1804027557373047,
"learning_rate": 1.163247281187026e-05,
"loss": 1.0113682746887207,
"step": 1364
},
{
"epoch": 0.9304703476482618,
"grad_norm": 2.0777599811553955,
"learning_rate": 1.1621577447029816e-05,
"loss": 0.6646759510040283,
"step": 1365
},
{
"epoch": 0.9311520109066121,
"grad_norm": 4.906405448913574,
"learning_rate": 1.1610680104888057e-05,
"loss": 2.013087272644043,
"step": 1366
},
{
"epoch": 0.9318336741649625,
"grad_norm": 1.0533791780471802,
"learning_rate": 1.1599780798732868e-05,
"loss": 0.5444119572639465,
"step": 1367
},
{
"epoch": 0.9325153374233128,
"grad_norm": 3.627973794937134,
"learning_rate": 1.158887954185452e-05,
"loss": 1.0098434686660767,
"step": 1368
},
{
"epoch": 0.9331970006816632,
"grad_norm": 2.2816169261932373,
"learning_rate": 1.157797634754567e-05,
"loss": 0.5894101858139038,
"step": 1369
},
{
"epoch": 0.9338786639400136,
"grad_norm": 1.3498238325119019,
"learning_rate": 1.1567071229101332e-05,
"loss": 0.8812987804412842,
"step": 1370
},
{
"epoch": 0.934560327198364,
"grad_norm": 2.0164008140563965,
"learning_rate": 1.1556164199818871e-05,
"loss": 1.0886428356170654,
"step": 1371
},
{
"epoch": 0.9352419904567144,
"grad_norm": 5.006718635559082,
"learning_rate": 1.1545255272997983e-05,
"loss": 2.039119005203247,
"step": 1372
},
{
"epoch": 0.9359236537150648,
"grad_norm": 1.3555481433868408,
"learning_rate": 1.153434446194068e-05,
"loss": 1.6098264455795288,
"step": 1373
},
{
"epoch": 0.9366053169734151,
"grad_norm": 3.5221757888793945,
"learning_rate": 1.1523431779951255e-05,
"loss": 1.293176531791687,
"step": 1374
},
{
"epoch": 0.9372869802317655,
"grad_norm": 2.4867475032806396,
"learning_rate": 1.1512517240336304e-05,
"loss": 1.3837792873382568,
"step": 1375
},
{
"epoch": 0.9379686434901159,
"grad_norm": 5.508114337921143,
"learning_rate": 1.1501600856404676e-05,
"loss": 2.0230002403259277,
"step": 1376
},
{
"epoch": 0.9386503067484663,
"grad_norm": 2.9103028774261475,
"learning_rate": 1.149068264146747e-05,
"loss": 1.313584327697754,
"step": 1377
},
{
"epoch": 0.9393319700068167,
"grad_norm": 4.0926594734191895,
"learning_rate": 1.1479762608838018e-05,
"loss": 1.785624623298645,
"step": 1378
},
{
"epoch": 0.940013633265167,
"grad_norm": 2.0462605953216553,
"learning_rate": 1.1468840771831874e-05,
"loss": 0.5250790119171143,
"step": 1379
},
{
"epoch": 0.9406952965235174,
"grad_norm": 2.402515172958374,
"learning_rate": 1.1457917143766786e-05,
"loss": 0.5881788730621338,
"step": 1380
},
{
"epoch": 0.9413769597818678,
"grad_norm": 1.6995738744735718,
"learning_rate": 1.1446991737962688e-05,
"loss": 0.9168304800987244,
"step": 1381
},
{
"epoch": 0.9420586230402181,
"grad_norm": 1.801107406616211,
"learning_rate": 1.1436064567741679e-05,
"loss": 0.9895972013473511,
"step": 1382
},
{
"epoch": 0.9427402862985685,
"grad_norm": 4.026597023010254,
"learning_rate": 1.1425135646428011e-05,
"loss": 2.1194210052490234,
"step": 1383
},
{
"epoch": 0.9434219495569189,
"grad_norm": 3.322819232940674,
"learning_rate": 1.141420498734808e-05,
"loss": 0.801817774772644,
"step": 1384
},
{
"epoch": 0.9441036128152692,
"grad_norm": 1.8702592849731445,
"learning_rate": 1.1403272603830384e-05,
"loss": 1.154260277748108,
"step": 1385
},
{
"epoch": 0.9447852760736196,
"grad_norm": 4.2028489112854,
"learning_rate": 1.139233850920554e-05,
"loss": 2.3136725425720215,
"step": 1386
},
{
"epoch": 0.94546693933197,
"grad_norm": 2.436260223388672,
"learning_rate": 1.1381402716806237e-05,
"loss": 0.9741934537887573,
"step": 1387
},
{
"epoch": 0.9461486025903204,
"grad_norm": 1.9411702156066895,
"learning_rate": 1.137046523996725e-05,
"loss": 1.2022333145141602,
"step": 1388
},
{
"epoch": 0.9468302658486708,
"grad_norm": 2.452498435974121,
"learning_rate": 1.1359526092025395e-05,
"loss": 0.7073385715484619,
"step": 1389
},
{
"epoch": 0.9475119291070211,
"grad_norm": 1.9482067823410034,
"learning_rate": 1.1348585286319529e-05,
"loss": 0.8009145259857178,
"step": 1390
},
{
"epoch": 0.9481935923653715,
"grad_norm": 1.6864196062088013,
"learning_rate": 1.1337642836190532e-05,
"loss": 0.9256442785263062,
"step": 1391
},
{
"epoch": 0.9488752556237219,
"grad_norm": 1.9843751192092896,
"learning_rate": 1.1326698754981292e-05,
"loss": 0.8110345602035522,
"step": 1392
},
{
"epoch": 0.9495569188820723,
"grad_norm": 2.8555917739868164,
"learning_rate": 1.131575305603668e-05,
"loss": 0.9550151824951172,
"step": 1393
},
{
"epoch": 0.9502385821404227,
"grad_norm": 1.2877635955810547,
"learning_rate": 1.130480575270354e-05,
"loss": 0.7633803486824036,
"step": 1394
},
{
"epoch": 0.950920245398773,
"grad_norm": 1.3084089756011963,
"learning_rate": 1.1293856858330678e-05,
"loss": 1.1857316493988037,
"step": 1395
},
{
"epoch": 0.9516019086571234,
"grad_norm": 2.01818585395813,
"learning_rate": 1.1282906386268842e-05,
"loss": 1.207491159439087,
"step": 1396
},
{
"epoch": 0.9522835719154737,
"grad_norm": 2.232659101486206,
"learning_rate": 1.1271954349870686e-05,
"loss": 1.1294128894805908,
"step": 1397
},
{
"epoch": 0.9529652351738241,
"grad_norm": 2.3538355827331543,
"learning_rate": 1.1261000762490793e-05,
"loss": 1.0520989894866943,
"step": 1398
},
{
"epoch": 0.9536468984321745,
"grad_norm": 2.2052597999572754,
"learning_rate": 1.1250045637485624e-05,
"loss": 0.3434506356716156,
"step": 1399
},
{
"epoch": 0.9543285616905249,
"grad_norm": 2.2298030853271484,
"learning_rate": 1.1239088988213522e-05,
"loss": 1.4217681884765625,
"step": 1400
},
{
"epoch": 0.9550102249488752,
"grad_norm": 3.030216932296753,
"learning_rate": 1.1228130828034685e-05,
"loss": 0.8571851253509521,
"step": 1401
},
{
"epoch": 0.9556918882072256,
"grad_norm": 1.500774621963501,
"learning_rate": 1.1217171170311157e-05,
"loss": 1.0443930625915527,
"step": 1402
},
{
"epoch": 0.956373551465576,
"grad_norm": 2.5589287281036377,
"learning_rate": 1.1206210028406797e-05,
"loss": 1.8444973230361938,
"step": 1403
},
{
"epoch": 0.9570552147239264,
"grad_norm": 1.1014353036880493,
"learning_rate": 1.1195247415687286e-05,
"loss": 0.6612838506698608,
"step": 1404
},
{
"epoch": 0.9577368779822768,
"grad_norm": 1.461776614189148,
"learning_rate": 1.11842833455201e-05,
"loss": 0.9575937986373901,
"step": 1405
},
{
"epoch": 0.9584185412406271,
"grad_norm": 2.6493754386901855,
"learning_rate": 1.1173317831274479e-05,
"loss": 1.0812537670135498,
"step": 1406
},
{
"epoch": 0.9591002044989775,
"grad_norm": 2.1644461154937744,
"learning_rate": 1.1162350886321435e-05,
"loss": 0.9594922065734863,
"step": 1407
},
{
"epoch": 0.9597818677573279,
"grad_norm": 3.7024781703948975,
"learning_rate": 1.115138252403372e-05,
"loss": 1.1270642280578613,
"step": 1408
},
{
"epoch": 0.9604635310156783,
"grad_norm": 2.0624656677246094,
"learning_rate": 1.1140412757785818e-05,
"loss": 0.8593518733978271,
"step": 1409
},
{
"epoch": 0.9611451942740287,
"grad_norm": 1.5966134071350098,
"learning_rate": 1.1129441600953916e-05,
"loss": 1.2847955226898193,
"step": 1410
},
{
"epoch": 0.961826857532379,
"grad_norm": 1.5607731342315674,
"learning_rate": 1.1118469066915907e-05,
"loss": 0.973434329032898,
"step": 1411
},
{
"epoch": 0.9625085207907293,
"grad_norm": 2.171595811843872,
"learning_rate": 1.1107495169051364e-05,
"loss": 1.2536604404449463,
"step": 1412
},
{
"epoch": 0.9631901840490797,
"grad_norm": 2.4465882778167725,
"learning_rate": 1.1096519920741509e-05,
"loss": 1.2964740991592407,
"step": 1413
},
{
"epoch": 0.9638718473074301,
"grad_norm": 1.1267685890197754,
"learning_rate": 1.1085543335369224e-05,
"loss": 0.756505012512207,
"step": 1414
},
{
"epoch": 0.9645535105657805,
"grad_norm": 4.980567455291748,
"learning_rate": 1.1074565426319014e-05,
"loss": 1.916570782661438,
"step": 1415
},
{
"epoch": 0.9652351738241309,
"grad_norm": 2.2414605617523193,
"learning_rate": 1.1063586206977009e-05,
"loss": 1.068878173828125,
"step": 1416
},
{
"epoch": 0.9659168370824812,
"grad_norm": 1.3340094089508057,
"learning_rate": 1.1052605690730922e-05,
"loss": 0.5894954204559326,
"step": 1417
},
{
"epoch": 0.9665985003408316,
"grad_norm": 1.699578881263733,
"learning_rate": 1.1041623890970061e-05,
"loss": 0.9399139881134033,
"step": 1418
},
{
"epoch": 0.967280163599182,
"grad_norm": 1.8976777791976929,
"learning_rate": 1.1030640821085284e-05,
"loss": 1.1349092721939087,
"step": 1419
},
{
"epoch": 0.9679618268575324,
"grad_norm": 1.6042039394378662,
"learning_rate": 1.101965649446901e-05,
"loss": 1.9341599941253662,
"step": 1420
},
{
"epoch": 0.9686434901158828,
"grad_norm": 1.7874032258987427,
"learning_rate": 1.1008670924515191e-05,
"loss": 1.09175443649292,
"step": 1421
},
{
"epoch": 0.9693251533742331,
"grad_norm": 2.457379102706909,
"learning_rate": 1.0997684124619286e-05,
"loss": 1.2264299392700195,
"step": 1422
},
{
"epoch": 0.9700068166325835,
"grad_norm": 2.286973476409912,
"learning_rate": 1.0986696108178259e-05,
"loss": 1.0105416774749756,
"step": 1423
},
{
"epoch": 0.9706884798909339,
"grad_norm": 1.4236822128295898,
"learning_rate": 1.0975706888590556e-05,
"loss": 0.9824390411376953,
"step": 1424
},
{
"epoch": 0.9713701431492843,
"grad_norm": 2.462789297103882,
"learning_rate": 1.0964716479256094e-05,
"loss": 1.3617198467254639,
"step": 1425
},
{
"epoch": 0.9720518064076347,
"grad_norm": 2.2924046516418457,
"learning_rate": 1.0953724893576236e-05,
"loss": 1.625471830368042,
"step": 1426
},
{
"epoch": 0.972733469665985,
"grad_norm": 2.718590021133423,
"learning_rate": 1.0942732144953782e-05,
"loss": 1.3157684803009033,
"step": 1427
},
{
"epoch": 0.9734151329243353,
"grad_norm": 2.6177055835723877,
"learning_rate": 1.0931738246792947e-05,
"loss": 1.2948482036590576,
"step": 1428
},
{
"epoch": 0.9740967961826857,
"grad_norm": 2.5654430389404297,
"learning_rate": 1.0920743212499355e-05,
"loss": 1.319617509841919,
"step": 1429
},
{
"epoch": 0.9747784594410361,
"grad_norm": 1.715665578842163,
"learning_rate": 1.0909747055480004e-05,
"loss": 0.8499696254730225,
"step": 1430
},
{
"epoch": 0.9754601226993865,
"grad_norm": 5.237400531768799,
"learning_rate": 1.089874978914327e-05,
"loss": 1.848403811454773,
"step": 1431
},
{
"epoch": 0.9761417859577369,
"grad_norm": 3.2037734985351562,
"learning_rate": 1.0887751426898878e-05,
"loss": 1.5934137105941772,
"step": 1432
},
{
"epoch": 0.9768234492160872,
"grad_norm": 5.400890350341797,
"learning_rate": 1.0876751982157892e-05,
"loss": 2.2045936584472656,
"step": 1433
},
{
"epoch": 0.9775051124744376,
"grad_norm": 2.7312662601470947,
"learning_rate": 1.0865751468332695e-05,
"loss": 1.0877323150634766,
"step": 1434
},
{
"epoch": 0.978186775732788,
"grad_norm": 3.9003102779388428,
"learning_rate": 1.0854749898836974e-05,
"loss": 1.2857639789581299,
"step": 1435
},
{
"epoch": 0.9788684389911384,
"grad_norm": 2.6747279167175293,
"learning_rate": 1.0843747287085693e-05,
"loss": 0.6983846426010132,
"step": 1436
},
{
"epoch": 0.9795501022494888,
"grad_norm": 2.099173069000244,
"learning_rate": 1.0832743646495105e-05,
"loss": 0.3862391412258148,
"step": 1437
},
{
"epoch": 0.9802317655078391,
"grad_norm": 3.4450464248657227,
"learning_rate": 1.0821738990482709e-05,
"loss": 2.163217782974243,
"step": 1438
},
{
"epoch": 0.9809134287661895,
"grad_norm": 2.7090086936950684,
"learning_rate": 1.0810733332467235e-05,
"loss": 1.9412810802459717,
"step": 1439
},
{
"epoch": 0.9815950920245399,
"grad_norm": 2.2056849002838135,
"learning_rate": 1.0799726685868648e-05,
"loss": 0.9925412535667419,
"step": 1440
},
{
"epoch": 0.9822767552828903,
"grad_norm": 1.637809157371521,
"learning_rate": 1.0788719064108108e-05,
"loss": 0.6028140783309937,
"step": 1441
},
{
"epoch": 0.9829584185412407,
"grad_norm": 22.958332061767578,
"learning_rate": 1.077771048060797e-05,
"loss": 1.2065813541412354,
"step": 1442
},
{
"epoch": 0.983640081799591,
"grad_norm": 4.294493198394775,
"learning_rate": 1.076670094879176e-05,
"loss": 1.4867355823516846,
"step": 1443
},
{
"epoch": 0.9843217450579413,
"grad_norm": 6.976797103881836,
"learning_rate": 1.0755690482084154e-05,
"loss": 2.184006929397583,
"step": 1444
},
{
"epoch": 0.9850034083162917,
"grad_norm": 1.6113888025283813,
"learning_rate": 1.0744679093910987e-05,
"loss": 1.124434232711792,
"step": 1445
},
{
"epoch": 0.9856850715746421,
"grad_norm": 1.7234139442443848,
"learning_rate": 1.0733666797699191e-05,
"loss": 1.2631725072860718,
"step": 1446
},
{
"epoch": 0.9863667348329925,
"grad_norm": 2.597224712371826,
"learning_rate": 1.0722653606876828e-05,
"loss": 1.4020732641220093,
"step": 1447
},
{
"epoch": 0.9870483980913429,
"grad_norm": 1.8286712169647217,
"learning_rate": 1.0711639534873035e-05,
"loss": 0.7092963457107544,
"step": 1448
},
{
"epoch": 0.9877300613496932,
"grad_norm": 2.565237283706665,
"learning_rate": 1.0700624595118037e-05,
"loss": 1.0420867204666138,
"step": 1449
},
{
"epoch": 0.9884117246080436,
"grad_norm": 2.8494246006011963,
"learning_rate": 1.0689608801043107e-05,
"loss": 1.3831346035003662,
"step": 1450
},
{
"epoch": 0.989093387866394,
"grad_norm": 2.3130369186401367,
"learning_rate": 1.0678592166080565e-05,
"loss": 1.2310330867767334,
"step": 1451
},
{
"epoch": 0.9897750511247444,
"grad_norm": 1.8310436010360718,
"learning_rate": 1.066757470366375e-05,
"loss": 1.2459018230438232,
"step": 1452
},
{
"epoch": 0.9904567143830948,
"grad_norm": 2.529782772064209,
"learning_rate": 1.0656556427227019e-05,
"loss": 1.8146663904190063,
"step": 1453
},
{
"epoch": 0.9911383776414451,
"grad_norm": 1.7871593236923218,
"learning_rate": 1.0645537350205714e-05,
"loss": 1.0763839483261108,
"step": 1454
},
{
"epoch": 0.9918200408997955,
"grad_norm": 1.7131450176239014,
"learning_rate": 1.063451748603616e-05,
"loss": 1.2468634843826294,
"step": 1455
},
{
"epoch": 0.9925017041581459,
"grad_norm": 2.5768184661865234,
"learning_rate": 1.0623496848155635e-05,
"loss": 1.4966133832931519,
"step": 1456
},
{
"epoch": 0.9931833674164963,
"grad_norm": 11.853314399719238,
"learning_rate": 1.0612475450002363e-05,
"loss": 1.127371907234192,
"step": 1457
},
{
"epoch": 0.9938650306748467,
"grad_norm": 2.7855846881866455,
"learning_rate": 1.0601453305015497e-05,
"loss": 1.5127134323120117,
"step": 1458
},
{
"epoch": 0.994546693933197,
"grad_norm": 2.1716721057891846,
"learning_rate": 1.0590430426635098e-05,
"loss": 0.47972676157951355,
"step": 1459
},
{
"epoch": 0.9952283571915473,
"grad_norm": 2.356139898300171,
"learning_rate": 1.0579406828302124e-05,
"loss": 1.1699801683425903,
"step": 1460
},
{
"epoch": 0.9959100204498977,
"grad_norm": 2.288517713546753,
"learning_rate": 1.0568382523458412e-05,
"loss": 0.9673724174499512,
"step": 1461
},
{
"epoch": 0.9965916837082481,
"grad_norm": 2.9393985271453857,
"learning_rate": 1.0557357525546651e-05,
"loss": 1.2749524116516113,
"step": 1462
},
{
"epoch": 0.9972733469665985,
"grad_norm": 3.3071682453155518,
"learning_rate": 1.054633184801039e-05,
"loss": 1.4630143642425537,
"step": 1463
},
{
"epoch": 0.9979550102249489,
"grad_norm": 3.298875331878662,
"learning_rate": 1.0535305504293988e-05,
"loss": 2.061471700668335,
"step": 1464
},
{
"epoch": 0.9986366734832992,
"grad_norm": 2.4312033653259277,
"learning_rate": 1.0524278507842637e-05,
"loss": 1.497416377067566,
"step": 1465
},
{
"epoch": 0.9993183367416496,
"grad_norm": 1.8498071432113647,
"learning_rate": 1.0513250872102312e-05,
"loss": 0.9513588547706604,
"step": 1466
},
{
"epoch": 1.0,
"grad_norm": 5.430646896362305,
"learning_rate": 1.0502222610519772e-05,
"loss": 1.1737076044082642,
"step": 1467
}
],
"logging_steps": 1,
"max_steps": 2934,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 5.504286382316585e+17,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}