9b-82 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
e5198e0 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.0,
"eval_steps": 500,
"global_step": 1804,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004434589800443459,
"grad_norm": 3.3554506301879883,
"learning_rate": 1.098901098901099e-07,
"loss": 1.8640056848526,
"step": 2
},
{
"epoch": 0.008869179600886918,
"grad_norm": 6.643158435821533,
"learning_rate": 3.296703296703297e-07,
"loss": 2.1273629665374756,
"step": 4
},
{
"epoch": 0.013303769401330377,
"grad_norm": 3.7049906253814697,
"learning_rate": 5.494505494505495e-07,
"loss": 1.900314450263977,
"step": 6
},
{
"epoch": 0.017738359201773836,
"grad_norm": 1.3328758478164673,
"learning_rate": 7.692307692307694e-07,
"loss": 1.8142547607421875,
"step": 8
},
{
"epoch": 0.022172949002217297,
"grad_norm": 3.059246778488159,
"learning_rate": 9.890109890109891e-07,
"loss": 1.6169909238815308,
"step": 10
},
{
"epoch": 0.026607538802660754,
"grad_norm": 2.269116163253784,
"learning_rate": 1.2087912087912089e-06,
"loss": 2.002399444580078,
"step": 12
},
{
"epoch": 0.031042128603104215,
"grad_norm": 41.05389404296875,
"learning_rate": 1.4285714285714286e-06,
"loss": 1.5568478107452393,
"step": 14
},
{
"epoch": 0.03547671840354767,
"grad_norm": 5.92053747177124,
"learning_rate": 1.6483516483516484e-06,
"loss": 1.1748180389404297,
"step": 16
},
{
"epoch": 0.03991130820399113,
"grad_norm": 2.922490119934082,
"learning_rate": 1.8681318681318684e-06,
"loss": 1.1770696640014648,
"step": 18
},
{
"epoch": 0.04434589800443459,
"grad_norm": 3.8268938064575195,
"learning_rate": 2.0879120879120883e-06,
"loss": 1.273401141166687,
"step": 20
},
{
"epoch": 0.04878048780487805,
"grad_norm": 2.3410561084747314,
"learning_rate": 2.307692307692308e-06,
"loss": 1.3016290664672852,
"step": 22
},
{
"epoch": 0.05321507760532151,
"grad_norm": 1.2151767015457153,
"learning_rate": 2.5274725274725274e-06,
"loss": 1.5213745832443237,
"step": 24
},
{
"epoch": 0.057649667405764965,
"grad_norm": 1.567148208618164,
"learning_rate": 2.7472527472527476e-06,
"loss": 1.2907153367996216,
"step": 26
},
{
"epoch": 0.06208425720620843,
"grad_norm": 1.8262158632278442,
"learning_rate": 2.9670329670329673e-06,
"loss": 1.0924012660980225,
"step": 28
},
{
"epoch": 0.06651884700665188,
"grad_norm": 2.100053071975708,
"learning_rate": 3.1868131868131867e-06,
"loss": 1.216428518295288,
"step": 30
},
{
"epoch": 0.07095343680709534,
"grad_norm": 2.2148892879486084,
"learning_rate": 3.406593406593407e-06,
"loss": 1.120039463043213,
"step": 32
},
{
"epoch": 0.07538802660753881,
"grad_norm": 1.2118117809295654,
"learning_rate": 3.6263736263736266e-06,
"loss": 1.4959304332733154,
"step": 34
},
{
"epoch": 0.07982261640798226,
"grad_norm": 23.823530197143555,
"learning_rate": 3.846153846153847e-06,
"loss": 1.2191643714904785,
"step": 36
},
{
"epoch": 0.08425720620842572,
"grad_norm": 2.678205728530884,
"learning_rate": 4.065934065934066e-06,
"loss": 1.5424106121063232,
"step": 38
},
{
"epoch": 0.08869179600886919,
"grad_norm": 1.8756370544433594,
"learning_rate": 4.2857142857142855e-06,
"loss": 0.7748823165893555,
"step": 40
},
{
"epoch": 0.09312638580931264,
"grad_norm": 1.0674728155136108,
"learning_rate": 4.505494505494506e-06,
"loss": 1.4076530933380127,
"step": 42
},
{
"epoch": 0.0975609756097561,
"grad_norm": 0.9798895716667175,
"learning_rate": 4.725274725274726e-06,
"loss": 1.3454558849334717,
"step": 44
},
{
"epoch": 0.10199556541019955,
"grad_norm": 3.0464468002319336,
"learning_rate": 4.945054945054946e-06,
"loss": 1.3695911169052124,
"step": 46
},
{
"epoch": 0.10643015521064302,
"grad_norm": 5.054316520690918,
"learning_rate": 5.164835164835166e-06,
"loss": 1.5866491794586182,
"step": 48
},
{
"epoch": 0.11086474501108648,
"grad_norm": 2.4682066440582275,
"learning_rate": 5.384615384615385e-06,
"loss": 1.4569354057312012,
"step": 50
},
{
"epoch": 0.11529933481152993,
"grad_norm": 1.1039724349975586,
"learning_rate": 5.604395604395605e-06,
"loss": 1.375510334968567,
"step": 52
},
{
"epoch": 0.1197339246119734,
"grad_norm": 0.911780595779419,
"learning_rate": 5.824175824175825e-06,
"loss": 1.3685683012008667,
"step": 54
},
{
"epoch": 0.12416851441241686,
"grad_norm": 6.081076145172119,
"learning_rate": 6.043956043956044e-06,
"loss": 1.1247367858886719,
"step": 56
},
{
"epoch": 0.1286031042128603,
"grad_norm": 1.635683536529541,
"learning_rate": 6.2637362637362645e-06,
"loss": 1.3490444421768188,
"step": 58
},
{
"epoch": 0.13303769401330376,
"grad_norm": 1.733303427696228,
"learning_rate": 6.483516483516485e-06,
"loss": 1.4053246974945068,
"step": 60
},
{
"epoch": 0.13747228381374724,
"grad_norm": 1.308205485343933,
"learning_rate": 6.703296703296703e-06,
"loss": 1.3603203296661377,
"step": 62
},
{
"epoch": 0.1419068736141907,
"grad_norm": 2.4009580612182617,
"learning_rate": 6.923076923076923e-06,
"loss": 1.403708577156067,
"step": 64
},
{
"epoch": 0.14634146341463414,
"grad_norm": 1.5370208024978638,
"learning_rate": 7.1428571428571436e-06,
"loss": 1.4732989072799683,
"step": 66
},
{
"epoch": 0.15077605321507762,
"grad_norm": 0.9826664328575134,
"learning_rate": 7.362637362637364e-06,
"loss": 1.347627878189087,
"step": 68
},
{
"epoch": 0.15521064301552107,
"grad_norm": 0.939538300037384,
"learning_rate": 7.582417582417583e-06,
"loss": 1.3008030652999878,
"step": 70
},
{
"epoch": 0.15964523281596452,
"grad_norm": 1.8689913749694824,
"learning_rate": 7.802197802197802e-06,
"loss": 1.3441245555877686,
"step": 72
},
{
"epoch": 0.164079822616408,
"grad_norm": 1.0277022123336792,
"learning_rate": 8.021978021978023e-06,
"loss": 1.249913215637207,
"step": 74
},
{
"epoch": 0.16851441241685144,
"grad_norm": 2.182894706726074,
"learning_rate": 8.241758241758243e-06,
"loss": 1.3225178718566895,
"step": 76
},
{
"epoch": 0.1729490022172949,
"grad_norm": 1.0984985828399658,
"learning_rate": 8.461538461538462e-06,
"loss": 1.3755261898040771,
"step": 78
},
{
"epoch": 0.17738359201773837,
"grad_norm": 1.6256319284439087,
"learning_rate": 8.681318681318681e-06,
"loss": 1.3338983058929443,
"step": 80
},
{
"epoch": 0.18181818181818182,
"grad_norm": 1.7835297584533691,
"learning_rate": 8.9010989010989e-06,
"loss": 0.840368926525116,
"step": 82
},
{
"epoch": 0.18625277161862527,
"grad_norm": 1.2216118574142456,
"learning_rate": 9.120879120879122e-06,
"loss": 1.0529584884643555,
"step": 84
},
{
"epoch": 0.19068736141906872,
"grad_norm": 8.783498764038086,
"learning_rate": 9.340659340659341e-06,
"loss": 0.9759446978569031,
"step": 86
},
{
"epoch": 0.1951219512195122,
"grad_norm": 1.1936975717544556,
"learning_rate": 9.560439560439562e-06,
"loss": 1.2668921947479248,
"step": 88
},
{
"epoch": 0.19955654101995565,
"grad_norm": 2.084754705429077,
"learning_rate": 9.780219780219781e-06,
"loss": 1.4002450704574585,
"step": 90
},
{
"epoch": 0.2039911308203991,
"grad_norm": 1.1517976522445679,
"learning_rate": 1e-05,
"loss": 1.2500206232070923,
"step": 92
},
{
"epoch": 0.20842572062084258,
"grad_norm": 0.9133433699607849,
"learning_rate": 9.99996972898091e-06,
"loss": 1.260838270187378,
"step": 94
},
{
"epoch": 0.21286031042128603,
"grad_norm": 1.9567794799804688,
"learning_rate": 9.999878916330893e-06,
"loss": 1.7782005071640015,
"step": 96
},
{
"epoch": 0.21729490022172948,
"grad_norm": 1.7563012838363647,
"learning_rate": 9.999727563271727e-06,
"loss": 0.9444801807403564,
"step": 98
},
{
"epoch": 0.22172949002217296,
"grad_norm": 0.9970119595527649,
"learning_rate": 9.999515671839682e-06,
"loss": 1.1939334869384766,
"step": 100
},
{
"epoch": 0.2261640798226164,
"grad_norm": 0.8175113201141357,
"learning_rate": 9.999243244885499e-06,
"loss": 1.3335109949111938,
"step": 102
},
{
"epoch": 0.23059866962305986,
"grad_norm": 0.8858981728553772,
"learning_rate": 9.998910286074355e-06,
"loss": 1.0767525434494019,
"step": 104
},
{
"epoch": 0.23503325942350334,
"grad_norm": 1.0978095531463623,
"learning_rate": 9.998516799885806e-06,
"loss": 1.301979899406433,
"step": 106
},
{
"epoch": 0.2394678492239468,
"grad_norm": 1.8789381980895996,
"learning_rate": 9.998062791613729e-06,
"loss": 1.5071189403533936,
"step": 108
},
{
"epoch": 0.24390243902439024,
"grad_norm": 0.8671095371246338,
"learning_rate": 9.997548267366255e-06,
"loss": 1.0696842670440674,
"step": 110
},
{
"epoch": 0.24833702882483372,
"grad_norm": 2.358786106109619,
"learning_rate": 9.996973234065685e-06,
"loss": 1.3439608812332153,
"step": 112
},
{
"epoch": 0.25277161862527714,
"grad_norm": 2.1430983543395996,
"learning_rate": 9.996337699448392e-06,
"loss": 0.8265036940574646,
"step": 114
},
{
"epoch": 0.2572062084257206,
"grad_norm": 0.876884400844574,
"learning_rate": 9.995641672064726e-06,
"loss": 1.3629298210144043,
"step": 116
},
{
"epoch": 0.2616407982261641,
"grad_norm": 0.48178285360336304,
"learning_rate": 9.994885161278885e-06,
"loss": 1.1036193370819092,
"step": 118
},
{
"epoch": 0.2660753880266075,
"grad_norm": 0.9723570942878723,
"learning_rate": 9.994068177268807e-06,
"loss": 1.2973068952560425,
"step": 120
},
{
"epoch": 0.270509977827051,
"grad_norm": 2.7924370765686035,
"learning_rate": 9.993190731026024e-06,
"loss": 1.2149885892868042,
"step": 122
},
{
"epoch": 0.2749445676274945,
"grad_norm": 2.260781764984131,
"learning_rate": 9.992252834355503e-06,
"loss": 1.5927814245224,
"step": 124
},
{
"epoch": 0.2793791574279379,
"grad_norm": 1.4566712379455566,
"learning_rate": 9.99125449987551e-06,
"loss": 0.9197668433189392,
"step": 126
},
{
"epoch": 0.2838137472283814,
"grad_norm": 1.0135411024093628,
"learning_rate": 9.990195741017422e-06,
"loss": 1.2308801412582397,
"step": 128
},
{
"epoch": 0.28824833702882485,
"grad_norm": 1.2406107187271118,
"learning_rate": 9.989076572025554e-06,
"loss": 0.9607541561126709,
"step": 130
},
{
"epoch": 0.2926829268292683,
"grad_norm": 9.23150634765625,
"learning_rate": 9.987897007956968e-06,
"loss": 1.0329269170761108,
"step": 132
},
{
"epoch": 0.29711751662971175,
"grad_norm": 7.180088520050049,
"learning_rate": 9.986657064681267e-06,
"loss": 1.0436151027679443,
"step": 134
},
{
"epoch": 0.30155210643015523,
"grad_norm": 5.097459316253662,
"learning_rate": 9.98535675888038e-06,
"loss": 1.2045552730560303,
"step": 136
},
{
"epoch": 0.30598669623059865,
"grad_norm": 4.074262619018555,
"learning_rate": 9.983996108048345e-06,
"loss": 0.8822503685951233,
"step": 138
},
{
"epoch": 0.31042128603104213,
"grad_norm": 2.92777156829834,
"learning_rate": 9.982575130491068e-06,
"loss": 1.1243302822113037,
"step": 140
},
{
"epoch": 0.3148558758314856,
"grad_norm": 1.5820231437683105,
"learning_rate": 9.981093845326079e-06,
"loss": 1.0163542032241821,
"step": 142
},
{
"epoch": 0.31929046563192903,
"grad_norm": 7.281858921051025,
"learning_rate": 9.979552272482268e-06,
"loss": 0.9830036163330078,
"step": 144
},
{
"epoch": 0.3237250554323725,
"grad_norm": 2.2801249027252197,
"learning_rate": 9.977950432699629e-06,
"loss": 0.952194094657898,
"step": 146
},
{
"epoch": 0.328159645232816,
"grad_norm": 1.4058592319488525,
"learning_rate": 9.976288347528972e-06,
"loss": 1.4162338972091675,
"step": 148
},
{
"epoch": 0.3325942350332594,
"grad_norm": 6.441311359405518,
"learning_rate": 9.974566039331634e-06,
"loss": 1.170148253440857,
"step": 150
},
{
"epoch": 0.3370288248337029,
"grad_norm": 2.101698160171509,
"learning_rate": 9.972783531279184e-06,
"loss": 1.0924369096755981,
"step": 152
},
{
"epoch": 0.34146341463414637,
"grad_norm": 1.7578390836715698,
"learning_rate": 9.970940847353103e-06,
"loss": 1.272911548614502,
"step": 154
},
{
"epoch": 0.3458980044345898,
"grad_norm": 1.6119791269302368,
"learning_rate": 9.969038012344465e-06,
"loss": 1.215100884437561,
"step": 156
},
{
"epoch": 0.35033259423503327,
"grad_norm": 3.253216028213501,
"learning_rate": 9.967075051853609e-06,
"loss": 1.2995957136154175,
"step": 158
},
{
"epoch": 0.35476718403547675,
"grad_norm": 0.5216699242591858,
"learning_rate": 9.965051992289782e-06,
"loss": 1.125442624092102,
"step": 160
},
{
"epoch": 0.35920177383592017,
"grad_norm": 1.3521455526351929,
"learning_rate": 9.962968860870798e-06,
"loss": 0.8663116097450256,
"step": 162
},
{
"epoch": 0.36363636363636365,
"grad_norm": 1.3028908967971802,
"learning_rate": 9.96082568562266e-06,
"loss": 0.9397580027580261,
"step": 164
},
{
"epoch": 0.36807095343680707,
"grad_norm": 1.2469340562820435,
"learning_rate": 9.958622495379193e-06,
"loss": 1.2785670757293701,
"step": 166
},
{
"epoch": 0.37250554323725055,
"grad_norm": 2.306950330734253,
"learning_rate": 9.956359319781642e-06,
"loss": 1.2098650932312012,
"step": 168
},
{
"epoch": 0.376940133037694,
"grad_norm": 2.1364083290100098,
"learning_rate": 9.954036189278292e-06,
"loss": 1.1903345584869385,
"step": 170
},
{
"epoch": 0.38137472283813745,
"grad_norm": 3.892174482345581,
"learning_rate": 9.951653135124045e-06,
"loss": 0.795183539390564,
"step": 172
},
{
"epoch": 0.3858093126385809,
"grad_norm": 2.057490825653076,
"learning_rate": 9.94921018938e-06,
"loss": 1.6345411539077759,
"step": 174
},
{
"epoch": 0.3902439024390244,
"grad_norm": 2.5715115070343018,
"learning_rate": 9.946707384913027e-06,
"loss": 1.2782708406448364,
"step": 176
},
{
"epoch": 0.3946784922394678,
"grad_norm": 1.30599045753479,
"learning_rate": 9.944144755395321e-06,
"loss": 1.3271397352218628,
"step": 178
},
{
"epoch": 0.3991130820399113,
"grad_norm": 3.162742853164673,
"learning_rate": 9.941522335303955e-06,
"loss": 1.079615592956543,
"step": 180
},
{
"epoch": 0.4035476718403548,
"grad_norm": 1.9985105991363525,
"learning_rate": 9.938840159920406e-06,
"loss": 1.2768008708953857,
"step": 182
},
{
"epoch": 0.4079822616407982,
"grad_norm": 1.0763951539993286,
"learning_rate": 9.93609826533009e-06,
"loss": 0.8742649555206299,
"step": 184
},
{
"epoch": 0.4124168514412417,
"grad_norm": 1.6700925827026367,
"learning_rate": 9.933296688421872e-06,
"loss": 0.9816460013389587,
"step": 186
},
{
"epoch": 0.41685144124168516,
"grad_norm": 1.3022819757461548,
"learning_rate": 9.930435466887564e-06,
"loss": 0.9694766998291016,
"step": 188
},
{
"epoch": 0.4212860310421286,
"grad_norm": 4.169641017913818,
"learning_rate": 9.927514639221433e-06,
"loss": 1.0260610580444336,
"step": 190
},
{
"epoch": 0.42572062084257206,
"grad_norm": 2.331894636154175,
"learning_rate": 9.92453424471967e-06,
"loss": 0.8971779346466064,
"step": 192
},
{
"epoch": 0.43015521064301554,
"grad_norm": 1.6991792917251587,
"learning_rate": 9.921494323479862e-06,
"loss": 1.3024404048919678,
"step": 194
},
{
"epoch": 0.43458980044345896,
"grad_norm": 1.732059121131897,
"learning_rate": 9.918394916400465e-06,
"loss": 1.5722870826721191,
"step": 196
},
{
"epoch": 0.43902439024390244,
"grad_norm": 1.7187920808792114,
"learning_rate": 9.915236065180235e-06,
"loss": 1.2906628847122192,
"step": 198
},
{
"epoch": 0.4434589800443459,
"grad_norm": 1.5556683540344238,
"learning_rate": 9.912017812317684e-06,
"loss": 1.1513667106628418,
"step": 200
},
{
"epoch": 0.44789356984478934,
"grad_norm": 1.5006672143936157,
"learning_rate": 9.908740201110497e-06,
"loss": 1.2683374881744385,
"step": 202
},
{
"epoch": 0.4523281596452328,
"grad_norm": 0.682464063167572,
"learning_rate": 9.905403275654951e-06,
"loss": 1.2315711975097656,
"step": 204
},
{
"epoch": 0.4567627494456763,
"grad_norm": 1.3839211463928223,
"learning_rate": 9.902007080845336e-06,
"loss": 1.0794446468353271,
"step": 206
},
{
"epoch": 0.4611973392461197,
"grad_norm": 2.0738892555236816,
"learning_rate": 9.898551662373325e-06,
"loss": 1.0966854095458984,
"step": 208
},
{
"epoch": 0.4656319290465632,
"grad_norm": 3.4559130668640137,
"learning_rate": 9.895037066727382e-06,
"loss": 0.6951467394828796,
"step": 210
},
{
"epoch": 0.4700665188470067,
"grad_norm": 4.855388641357422,
"learning_rate": 9.891463341192124e-06,
"loss": 0.8868162035942078,
"step": 212
},
{
"epoch": 0.4745011086474501,
"grad_norm": 0.7046754360198975,
"learning_rate": 9.88783053384769e-06,
"loss": 1.0986382961273193,
"step": 214
},
{
"epoch": 0.4789356984478936,
"grad_norm": 1.4185415506362915,
"learning_rate": 9.884138693569095e-06,
"loss": 1.1152902841567993,
"step": 216
},
{
"epoch": 0.48337028824833705,
"grad_norm": 0.9065267443656921,
"learning_rate": 9.88038787002557e-06,
"loss": 1.130263090133667,
"step": 218
},
{
"epoch": 0.4878048780487805,
"grad_norm": 1.9596226215362549,
"learning_rate": 9.876578113679891e-06,
"loss": 1.2346633672714233,
"step": 220
},
{
"epoch": 0.49223946784922396,
"grad_norm": 1.4536895751953125,
"learning_rate": 9.872709475787708e-06,
"loss": 1.2405025959014893,
"step": 222
},
{
"epoch": 0.49667405764966743,
"grad_norm": 0.745212972164154,
"learning_rate": 9.868782008396848e-06,
"loss": 1.2145315408706665,
"step": 224
},
{
"epoch": 0.5011086474501109,
"grad_norm": 0.7731147408485413,
"learning_rate": 9.864795764346615e-06,
"loss": 1.2132314443588257,
"step": 226
},
{
"epoch": 0.5055432372505543,
"grad_norm": 0.931675374507904,
"learning_rate": 9.860750797267085e-06,
"loss": 1.2666246891021729,
"step": 228
},
{
"epoch": 0.5099778270509978,
"grad_norm": 1.4641870260238647,
"learning_rate": 9.856647161578384e-06,
"loss": 1.855782151222229,
"step": 230
},
{
"epoch": 0.5144124168514412,
"grad_norm": 1.5169061422348022,
"learning_rate": 9.852484912489946e-06,
"loss": 0.8463820219039917,
"step": 232
},
{
"epoch": 0.5188470066518847,
"grad_norm": 0.7717849016189575,
"learning_rate": 9.848264105999783e-06,
"loss": 1.2967286109924316,
"step": 234
},
{
"epoch": 0.5232815964523282,
"grad_norm": 0.6193521022796631,
"learning_rate": 9.843984798893722e-06,
"loss": 0.9528148174285889,
"step": 236
},
{
"epoch": 0.5277161862527716,
"grad_norm": 0.6214136481285095,
"learning_rate": 9.839647048744645e-06,
"loss": 1.0307953357696533,
"step": 238
},
{
"epoch": 0.532150776053215,
"grad_norm": 0.6043981909751892,
"learning_rate": 9.83525091391172e-06,
"loss": 1.2823783159255981,
"step": 240
},
{
"epoch": 0.5365853658536586,
"grad_norm": 0.8168203234672546,
"learning_rate": 9.8307964535396e-06,
"loss": 1.3375117778778076,
"step": 242
},
{
"epoch": 0.541019955654102,
"grad_norm": 10.725733757019043,
"learning_rate": 9.826283727557644e-06,
"loss": 1.0199899673461914,
"step": 244
},
{
"epoch": 0.5454545454545454,
"grad_norm": 1.418264389038086,
"learning_rate": 9.821712796679106e-06,
"loss": 1.2732073068618774,
"step": 246
},
{
"epoch": 0.549889135254989,
"grad_norm": 0.9490866661071777,
"learning_rate": 9.817083722400309e-06,
"loss": 1.4992234706878662,
"step": 248
},
{
"epoch": 0.5543237250554324,
"grad_norm": 1.1178134679794312,
"learning_rate": 9.812396566999832e-06,
"loss": 1.2512468099594116,
"step": 250
},
{
"epoch": 0.5587583148558758,
"grad_norm": 5.028557777404785,
"learning_rate": 9.807651393537659e-06,
"loss": 0.7450082302093506,
"step": 252
},
{
"epoch": 0.5631929046563193,
"grad_norm": 1.3556923866271973,
"learning_rate": 9.802848265854343e-06,
"loss": 1.4887409210205078,
"step": 254
},
{
"epoch": 0.5676274944567627,
"grad_norm": 1.1928884983062744,
"learning_rate": 9.797987248570137e-06,
"loss": 1.3401446342468262,
"step": 256
},
{
"epoch": 0.5720620842572062,
"grad_norm": 0.7480881214141846,
"learning_rate": 9.793068407084125e-06,
"loss": 1.306605577468872,
"step": 258
},
{
"epoch": 0.5764966740576497,
"grad_norm": 0.641482412815094,
"learning_rate": 9.78809180757335e-06,
"loss": 1.3040629625320435,
"step": 260
},
{
"epoch": 0.5809312638580931,
"grad_norm": 3.0096657276153564,
"learning_rate": 9.783057516991921e-06,
"loss": 0.7431791424751282,
"step": 262
},
{
"epoch": 0.5853658536585366,
"grad_norm": 0.928672194480896,
"learning_rate": 9.777965603070106e-06,
"loss": 1.3897572755813599,
"step": 264
},
{
"epoch": 0.5898004434589801,
"grad_norm": 1.2365378141403198,
"learning_rate": 9.772816134313424e-06,
"loss": 1.0928809642791748,
"step": 266
},
{
"epoch": 0.5942350332594235,
"grad_norm": 0.8595125079154968,
"learning_rate": 9.76760918000173e-06,
"loss": 1.5679038763046265,
"step": 268
},
{
"epoch": 0.5986696230598669,
"grad_norm": 3.3064000606536865,
"learning_rate": 9.762344810188276e-06,
"loss": 1.21007239818573,
"step": 270
},
{
"epoch": 0.6031042128603105,
"grad_norm": 0.6491636037826538,
"learning_rate": 9.757023095698766e-06,
"loss": 1.2509498596191406,
"step": 272
},
{
"epoch": 0.6075388026607539,
"grad_norm": 2.3098983764648438,
"learning_rate": 9.751644108130405e-06,
"loss": 1.2461967468261719,
"step": 274
},
{
"epoch": 0.6119733924611973,
"grad_norm": 0.7247368097305298,
"learning_rate": 9.746207919850951e-06,
"loss": 1.2410979270935059,
"step": 276
},
{
"epoch": 0.6164079822616408,
"grad_norm": 3.6421871185302734,
"learning_rate": 9.740714603997712e-06,
"loss": 1.2530133724212646,
"step": 278
},
{
"epoch": 0.6208425720620843,
"grad_norm": 0.7265106439590454,
"learning_rate": 9.735164234476588e-06,
"loss": 1.3131635189056396,
"step": 280
},
{
"epoch": 0.6252771618625277,
"grad_norm": 1.2309120893478394,
"learning_rate": 9.729556885961064e-06,
"loss": 0.9114726781845093,
"step": 282
},
{
"epoch": 0.6297117516629712,
"grad_norm": 0.7240830063819885,
"learning_rate": 9.72389263389121e-06,
"loss": 1.2823668718338013,
"step": 284
},
{
"epoch": 0.6341463414634146,
"grad_norm": 1.1647893190383911,
"learning_rate": 9.718171554472662e-06,
"loss": 1.3458237648010254,
"step": 286
},
{
"epoch": 0.6385809312638581,
"grad_norm": 0.8723113536834717,
"learning_rate": 9.712393724675597e-06,
"loss": 1.2615196704864502,
"step": 288
},
{
"epoch": 0.6430155210643016,
"grad_norm": 0.7617486119270325,
"learning_rate": 9.706559222233704e-06,
"loss": 1.2500044107437134,
"step": 290
},
{
"epoch": 0.647450110864745,
"grad_norm": 1.2043062448501587,
"learning_rate": 9.700668125643132e-06,
"loss": 1.3805018663406372,
"step": 292
},
{
"epoch": 0.6518847006651884,
"grad_norm": 0.8987176418304443,
"learning_rate": 9.694720514161437e-06,
"loss": 0.9510623216629028,
"step": 294
},
{
"epoch": 0.656319290465632,
"grad_norm": 1.2220165729522705,
"learning_rate": 9.688716467806508e-06,
"loss": 1.0256685018539429,
"step": 296
},
{
"epoch": 0.6607538802660754,
"grad_norm": 1.0878409147262573,
"learning_rate": 9.682656067355505e-06,
"loss": 1.230331301689148,
"step": 298
},
{
"epoch": 0.6651884700665188,
"grad_norm": 0.7459701299667358,
"learning_rate": 9.67653939434376e-06,
"loss": 1.272396206855774,
"step": 300
},
{
"epoch": 0.6696230598669624,
"grad_norm": 0.6311793327331543,
"learning_rate": 9.670366531063686e-06,
"loss": 1.216810703277588,
"step": 302
},
{
"epoch": 0.6740576496674058,
"grad_norm": 0.9430868625640869,
"learning_rate": 9.664137560563663e-06,
"loss": 1.2837876081466675,
"step": 304
},
{
"epoch": 0.6784922394678492,
"grad_norm": 0.6982176303863525,
"learning_rate": 9.657852566646929e-06,
"loss": 1.2521125078201294,
"step": 306
},
{
"epoch": 0.6829268292682927,
"grad_norm": 0.9578081369400024,
"learning_rate": 9.651511633870451e-06,
"loss": 0.9001311659812927,
"step": 308
},
{
"epoch": 0.6873614190687362,
"grad_norm": 0.7129071354866028,
"learning_rate": 9.645114847543781e-06,
"loss": 1.2543275356292725,
"step": 310
},
{
"epoch": 0.6917960088691796,
"grad_norm": 0.7872418761253357,
"learning_rate": 9.638662293727916e-06,
"loss": 1.225979208946228,
"step": 312
},
{
"epoch": 0.6962305986696231,
"grad_norm": 0.8527798652648926,
"learning_rate": 9.632154059234137e-06,
"loss": 1.2440277338027954,
"step": 314
},
{
"epoch": 0.7006651884700665,
"grad_norm": 0.8361867666244507,
"learning_rate": 9.625590231622837e-06,
"loss": 1.3528887033462524,
"step": 316
},
{
"epoch": 0.70509977827051,
"grad_norm": 1.285349726676941,
"learning_rate": 9.618970899202354e-06,
"loss": 1.0397502183914185,
"step": 318
},
{
"epoch": 0.7095343680709535,
"grad_norm": 1.2274253368377686,
"learning_rate": 9.612296151027765e-06,
"loss": 1.0072234869003296,
"step": 320
},
{
"epoch": 0.7139689578713969,
"grad_norm": 1.74605393409729,
"learning_rate": 9.605566076899714e-06,
"loss": 1.0066006183624268,
"step": 322
},
{
"epoch": 0.7184035476718403,
"grad_norm": 1.2762601375579834,
"learning_rate": 9.598780767363174e-06,
"loss": 1.1060127019882202,
"step": 324
},
{
"epoch": 0.7228381374722838,
"grad_norm": 2.128355026245117,
"learning_rate": 9.591940313706248e-06,
"loss": 1.1243398189544678,
"step": 326
},
{
"epoch": 0.7272727272727273,
"grad_norm": 0.4223320186138153,
"learning_rate": 9.585044807958942e-06,
"loss": 0.8515225648880005,
"step": 328
},
{
"epoch": 0.7317073170731707,
"grad_norm": 1.9096373319625854,
"learning_rate": 9.578094342891915e-06,
"loss": 0.8710092902183533,
"step": 330
},
{
"epoch": 0.7361419068736141,
"grad_norm": 1.7320899963378906,
"learning_rate": 9.571089012015237e-06,
"loss": 1.403598666191101,
"step": 332
},
{
"epoch": 0.7405764966740577,
"grad_norm": 2.438868999481201,
"learning_rate": 9.564028909577132e-06,
"loss": 1.208489179611206,
"step": 334
},
{
"epoch": 0.7450110864745011,
"grad_norm": 1.0341501235961914,
"learning_rate": 9.55691413056271e-06,
"loss": 1.2458648681640625,
"step": 336
},
{
"epoch": 0.7494456762749445,
"grad_norm": 2.2480390071868896,
"learning_rate": 9.54974477069269e-06,
"loss": 0.7539517283439636,
"step": 338
},
{
"epoch": 0.753880266075388,
"grad_norm": 1.8287423849105835,
"learning_rate": 9.542520926422105e-06,
"loss": 0.3899018466472626,
"step": 340
},
{
"epoch": 0.7583148558758315,
"grad_norm": 1.3207381963729858,
"learning_rate": 9.535242694939011e-06,
"loss": 0.9129378199577332,
"step": 342
},
{
"epoch": 0.7627494456762749,
"grad_norm": 0.7297150492668152,
"learning_rate": 9.527910174163179e-06,
"loss": 1.3148428201675415,
"step": 344
},
{
"epoch": 0.7671840354767184,
"grad_norm": 0.6148672103881836,
"learning_rate": 9.520523462744776e-06,
"loss": 1.2478430271148682,
"step": 346
},
{
"epoch": 0.7716186252771619,
"grad_norm": 2.0105557441711426,
"learning_rate": 9.51308266006304e-06,
"loss": 1.108123540878296,
"step": 348
},
{
"epoch": 0.7760532150776053,
"grad_norm": 0.5692620873451233,
"learning_rate": 9.505587866224939e-06,
"loss": 1.2364296913146973,
"step": 350
},
{
"epoch": 0.7804878048780488,
"grad_norm": 1.4871420860290527,
"learning_rate": 9.498039182063828e-06,
"loss": 1.178447961807251,
"step": 352
},
{
"epoch": 0.7849223946784922,
"grad_norm": 2.9917361736297607,
"learning_rate": 9.49043670913809e-06,
"loss": 1.0383565425872803,
"step": 354
},
{
"epoch": 0.7893569844789357,
"grad_norm": 1.1324853897094727,
"learning_rate": 9.48278054972977e-06,
"loss": 1.2439326047897339,
"step": 356
},
{
"epoch": 0.7937915742793792,
"grad_norm": 1.664652943611145,
"learning_rate": 9.475070806843202e-06,
"loss": 1.293602466583252,
"step": 358
},
{
"epoch": 0.7982261640798226,
"grad_norm": 0.6531599164009094,
"learning_rate": 9.467307584203619e-06,
"loss": 1.2421692609786987,
"step": 360
},
{
"epoch": 0.802660753880266,
"grad_norm": 0.9146925806999207,
"learning_rate": 9.459490986255756e-06,
"loss": 0.6786575317382812,
"step": 362
},
{
"epoch": 0.8070953436807096,
"grad_norm": 1.0817216634750366,
"learning_rate": 9.451621118162453e-06,
"loss": 1.3574942350387573,
"step": 364
},
{
"epoch": 0.811529933481153,
"grad_norm": 0.8012416958808899,
"learning_rate": 9.443698085803235e-06,
"loss": 1.2184292078018188,
"step": 366
},
{
"epoch": 0.8159645232815964,
"grad_norm": 0.6288676261901855,
"learning_rate": 9.435721995772884e-06,
"loss": 1.0145061016082764,
"step": 368
},
{
"epoch": 0.8203991130820399,
"grad_norm": 2.079406499862671,
"learning_rate": 9.42769295538001e-06,
"loss": 0.8563152551651001,
"step": 370
},
{
"epoch": 0.8248337028824834,
"grad_norm": 0.6172999143600464,
"learning_rate": 9.419611072645608e-06,
"loss": 1.246902585029602,
"step": 372
},
{
"epoch": 0.8292682926829268,
"grad_norm": 0.6710400581359863,
"learning_rate": 9.4114764563016e-06,
"loss": 1.2490512132644653,
"step": 374
},
{
"epoch": 0.8337028824833703,
"grad_norm": 1.4091862440109253,
"learning_rate": 9.403289215789373e-06,
"loss": 1.2173607349395752,
"step": 376
},
{
"epoch": 0.8381374722838137,
"grad_norm": 1.2435530424118042,
"learning_rate": 9.395049461258318e-06,
"loss": 1.2395451068878174,
"step": 378
},
{
"epoch": 0.8425720620842572,
"grad_norm": 1.2862529754638672,
"learning_rate": 9.386757303564323e-06,
"loss": 0.807468593120575,
"step": 380
},
{
"epoch": 0.8470066518847007,
"grad_norm": 0.6037282943725586,
"learning_rate": 9.37841285426831e-06,
"loss": 1.3310133218765259,
"step": 382
},
{
"epoch": 0.8514412416851441,
"grad_norm": 0.6190339922904968,
"learning_rate": 9.370016225634719e-06,
"loss": 1.168778896331787,
"step": 384
},
{
"epoch": 0.8558758314855875,
"grad_norm": 0.557336688041687,
"learning_rate": 9.361567530629988e-06,
"loss": 1.2637746334075928,
"step": 386
},
{
"epoch": 0.8603104212860311,
"grad_norm": 0.4134604334831238,
"learning_rate": 9.353066882921063e-06,
"loss": 1.192622423171997,
"step": 388
},
{
"epoch": 0.8647450110864745,
"grad_norm": 0.5234290957450867,
"learning_rate": 9.344514396873837e-06,
"loss": 1.2653976678848267,
"step": 390
},
{
"epoch": 0.8691796008869179,
"grad_norm": 1.246576189994812,
"learning_rate": 9.335910187551628e-06,
"loss": 0.9260438084602356,
"step": 392
},
{
"epoch": 0.8736141906873615,
"grad_norm": 0.5154649615287781,
"learning_rate": 9.327254370713636e-06,
"loss": 1.2277519702911377,
"step": 394
},
{
"epoch": 0.8780487804878049,
"grad_norm": 1.1650840044021606,
"learning_rate": 9.31854706281336e-06,
"loss": 1.2231098413467407,
"step": 396
},
{
"epoch": 0.8824833702882483,
"grad_norm": 0.8881715536117554,
"learning_rate": 9.309788380997069e-06,
"loss": 1.219472050666809,
"step": 398
},
{
"epoch": 0.8869179600886918,
"grad_norm": 1.3225759267807007,
"learning_rate": 9.30097844310219e-06,
"loss": 0.5170519351959229,
"step": 400
},
{
"epoch": 0.8913525498891353,
"grad_norm": 0.6402235627174377,
"learning_rate": 9.292117367655749e-06,
"loss": 1.1000133752822876,
"step": 402
},
{
"epoch": 0.8957871396895787,
"grad_norm": 0.25668174028396606,
"learning_rate": 9.283205273872757e-06,
"loss": 0.9859495162963867,
"step": 404
},
{
"epoch": 0.9002217294900222,
"grad_norm": 0.659263014793396,
"learning_rate": 9.274242281654621e-06,
"loss": 1.2722219228744507,
"step": 406
},
{
"epoch": 0.9046563192904656,
"grad_norm": 1.187609314918518,
"learning_rate": 9.265228511587525e-06,
"loss": 1.1655584573745728,
"step": 408
},
{
"epoch": 0.9090909090909091,
"grad_norm": 0.9958807229995728,
"learning_rate": 9.2561640849408e-06,
"loss": 1.1174354553222656,
"step": 410
},
{
"epoch": 0.9135254988913526,
"grad_norm": 0.497241348028183,
"learning_rate": 9.247049123665306e-06,
"loss": 1.240470051765442,
"step": 412
},
{
"epoch": 0.917960088691796,
"grad_norm": 0.5129937529563904,
"learning_rate": 9.237883750391786e-06,
"loss": 1.2640633583068848,
"step": 414
},
{
"epoch": 0.9223946784922394,
"grad_norm": 1.200805902481079,
"learning_rate": 9.228668088429212e-06,
"loss": 1.023282766342163,
"step": 416
},
{
"epoch": 0.926829268292683,
"grad_norm": 1.1263169050216675,
"learning_rate": 9.219402261763129e-06,
"loss": 1.2812139987945557,
"step": 418
},
{
"epoch": 0.9312638580931264,
"grad_norm": 8.3773832321167,
"learning_rate": 9.210086395053992e-06,
"loss": 0.8014230728149414,
"step": 420
},
{
"epoch": 0.9356984478935698,
"grad_norm": 1.9554097652435303,
"learning_rate": 9.200720613635476e-06,
"loss": 1.4770504236221313,
"step": 422
},
{
"epoch": 0.9401330376940134,
"grad_norm": 9.581804275512695,
"learning_rate": 9.191305043512806e-06,
"loss": 1.068493366241455,
"step": 424
},
{
"epoch": 0.9445676274944568,
"grad_norm": 2.6033480167388916,
"learning_rate": 9.181839811361048e-06,
"loss": 1.225175380706787,
"step": 426
},
{
"epoch": 0.9490022172949002,
"grad_norm": 0.506930947303772,
"learning_rate": 9.172325044523413e-06,
"loss": 1.1149027347564697,
"step": 428
},
{
"epoch": 0.9534368070953437,
"grad_norm": 2.6233949661254883,
"learning_rate": 9.16276087100954e-06,
"loss": 0.7080057859420776,
"step": 430
},
{
"epoch": 0.9578713968957872,
"grad_norm": 0.4864356815814972,
"learning_rate": 9.153147419493774e-06,
"loss": 1.2395509481430054,
"step": 432
},
{
"epoch": 0.9623059866962306,
"grad_norm": 10.547813415527344,
"learning_rate": 9.143484819313441e-06,
"loss": 1.1166844367980957,
"step": 434
},
{
"epoch": 0.9667405764966741,
"grad_norm": 0.5617235898971558,
"learning_rate": 9.133773200467095e-06,
"loss": 1.2907624244689941,
"step": 436
},
{
"epoch": 0.9711751662971175,
"grad_norm": 0.6111188530921936,
"learning_rate": 9.12401269361278e-06,
"loss": 1.2545419931411743,
"step": 438
},
{
"epoch": 0.975609756097561,
"grad_norm": 2.3575284481048584,
"learning_rate": 9.114203430066273e-06,
"loss": 0.825833797454834,
"step": 440
},
{
"epoch": 0.9800443458980045,
"grad_norm": 0.5967570543289185,
"learning_rate": 9.104345541799304e-06,
"loss": 1.0687720775604248,
"step": 442
},
{
"epoch": 0.9844789356984479,
"grad_norm": 0.7675039172172546,
"learning_rate": 9.094439161437797e-06,
"loss": 1.3078503608703613,
"step": 444
},
{
"epoch": 0.9889135254988913,
"grad_norm": 0.5262892842292786,
"learning_rate": 9.084484422260079e-06,
"loss": 1.2215782403945923,
"step": 446
},
{
"epoch": 0.9933481152993349,
"grad_norm": 0.5987316370010376,
"learning_rate": 9.074481458195077e-06,
"loss": 1.2558292150497437,
"step": 448
},
{
"epoch": 0.9977827050997783,
"grad_norm": 0.6052929162979126,
"learning_rate": 9.064430403820538e-06,
"loss": 1.0702413320541382,
"step": 450
},
{
"epoch": 1.0022172949002217,
"grad_norm": 2.1425399780273438,
"learning_rate": 9.054331394361195e-06,
"loss": 1.0677464008331299,
"step": 452
},
{
"epoch": 1.0066518847006651,
"grad_norm": 1.4416669607162476,
"learning_rate": 9.044184565686963e-06,
"loss": 1.2889671325683594,
"step": 454
},
{
"epoch": 1.0110864745011086,
"grad_norm": 0.6405385136604309,
"learning_rate": 9.033990054311108e-06,
"loss": 1.169727087020874,
"step": 456
},
{
"epoch": 1.0155210643015522,
"grad_norm": 0.6425080895423889,
"learning_rate": 9.023747997388409e-06,
"loss": 0.8333257436752319,
"step": 458
},
{
"epoch": 1.0199556541019956,
"grad_norm": 0.2827882170677185,
"learning_rate": 9.013458532713303e-06,
"loss": 0.76810622215271,
"step": 460
},
{
"epoch": 1.024390243902439,
"grad_norm": 0.7950817942619324,
"learning_rate": 9.003121798718055e-06,
"loss": 0.5744680166244507,
"step": 462
},
{
"epoch": 1.0288248337028825,
"grad_norm": 2.603362560272217,
"learning_rate": 8.992737934470875e-06,
"loss": 0.8523436188697815,
"step": 464
},
{
"epoch": 1.033259423503326,
"grad_norm": 0.9304072856903076,
"learning_rate": 8.982307079674051e-06,
"loss": 0.723950207233429,
"step": 466
},
{
"epoch": 1.0376940133037693,
"grad_norm": 1.9990200996398926,
"learning_rate": 8.971829374662075e-06,
"loss": 1.0109994411468506,
"step": 468
},
{
"epoch": 1.042128603104213,
"grad_norm": 0.8482561111450195,
"learning_rate": 8.961304960399746e-06,
"loss": 0.6301277875900269,
"step": 470
},
{
"epoch": 1.0465631929046564,
"grad_norm": 1.0088205337524414,
"learning_rate": 8.950733978480295e-06,
"loss": 0.9854256510734558,
"step": 472
},
{
"epoch": 1.0509977827050998,
"grad_norm": 1.1430635452270508,
"learning_rate": 8.940116571123442e-06,
"loss": 1.022680401802063,
"step": 474
},
{
"epoch": 1.0554323725055432,
"grad_norm": 0.5003657341003418,
"learning_rate": 8.929452881173522e-06,
"loss": 1.0202995538711548,
"step": 476
},
{
"epoch": 1.0598669623059866,
"grad_norm": 0.5999464988708496,
"learning_rate": 8.91874305209754e-06,
"loss": 1.0750170946121216,
"step": 478
},
{
"epoch": 1.06430155210643,
"grad_norm": 1.1878728866577148,
"learning_rate": 8.907987227983244e-06,
"loss": 0.6575189828872681,
"step": 480
},
{
"epoch": 1.0687361419068737,
"grad_norm": 0.7090225219726562,
"learning_rate": 8.897185553537199e-06,
"loss": 1.1403692960739136,
"step": 482
},
{
"epoch": 1.0731707317073171,
"grad_norm": 0.3266684412956238,
"learning_rate": 8.886338174082818e-06,
"loss": 0.8844245076179504,
"step": 484
},
{
"epoch": 1.0776053215077606,
"grad_norm": 0.8314698934555054,
"learning_rate": 8.875445235558429e-06,
"loss": 1.0011812448501587,
"step": 486
},
{
"epoch": 1.082039911308204,
"grad_norm": 0.7279475927352905,
"learning_rate": 8.864506884515298e-06,
"loss": 0.7617220282554626,
"step": 488
},
{
"epoch": 1.0864745011086474,
"grad_norm": 0.42657408118247986,
"learning_rate": 8.853523268115662e-06,
"loss": 0.922903835773468,
"step": 490
},
{
"epoch": 1.0909090909090908,
"grad_norm": 0.6559341549873352,
"learning_rate": 8.84249453413075e-06,
"loss": 1.0911064147949219,
"step": 492
},
{
"epoch": 1.0953436807095343,
"grad_norm": 1.2124531269073486,
"learning_rate": 8.831420830938787e-06,
"loss": 1.0689539909362793,
"step": 494
},
{
"epoch": 1.099778270509978,
"grad_norm": 1.9062159061431885,
"learning_rate": 8.820302307523012e-06,
"loss": 0.9026176929473877,
"step": 496
},
{
"epoch": 1.1042128603104213,
"grad_norm": 0.4806932806968689,
"learning_rate": 8.809139113469664e-06,
"loss": 0.9866113662719727,
"step": 498
},
{
"epoch": 1.1086474501108647,
"grad_norm": 0.7793009281158447,
"learning_rate": 8.797931398965968e-06,
"loss": 0.6823930144309998,
"step": 500
},
{
"epoch": 1.1130820399113082,
"grad_norm": 0.4156184494495392,
"learning_rate": 8.78667931479812e-06,
"loss": 0.9841465353965759,
"step": 502
},
{
"epoch": 1.1175166297117516,
"grad_norm": 0.5767391920089722,
"learning_rate": 8.775383012349255e-06,
"loss": 0.8618515729904175,
"step": 504
},
{
"epoch": 1.1219512195121952,
"grad_norm": 1.2682608366012573,
"learning_rate": 8.764042643597413e-06,
"loss": 0.9228280782699585,
"step": 506
},
{
"epoch": 1.1263858093126387,
"grad_norm": 1.26870596408844,
"learning_rate": 8.75265836111349e-06,
"loss": 0.6945858001708984,
"step": 508
},
{
"epoch": 1.130820399113082,
"grad_norm": 0.9481813311576843,
"learning_rate": 8.741230318059188e-06,
"loss": 1.1894787549972534,
"step": 510
},
{
"epoch": 1.1352549889135255,
"grad_norm": 1.9494104385375977,
"learning_rate": 8.72975866818496e-06,
"loss": 0.5956374406814575,
"step": 512
},
{
"epoch": 1.139689578713969,
"grad_norm": 0.6274094581604004,
"learning_rate": 8.718243565827927e-06,
"loss": 0.4184134900569916,
"step": 514
},
{
"epoch": 1.1441241685144123,
"grad_norm": 0.6972094178199768,
"learning_rate": 8.706685165909817e-06,
"loss": 0.8942562937736511,
"step": 516
},
{
"epoch": 1.1485587583148558,
"grad_norm": 0.9098567366600037,
"learning_rate": 8.695083623934872e-06,
"loss": 1.0386896133422852,
"step": 518
},
{
"epoch": 1.1529933481152994,
"grad_norm": 0.43444758653640747,
"learning_rate": 8.683439095987758e-06,
"loss": 1.1053950786590576,
"step": 520
},
{
"epoch": 1.1574279379157428,
"grad_norm": 0.6920401453971863,
"learning_rate": 8.671751738731464e-06,
"loss": 0.5241594314575195,
"step": 522
},
{
"epoch": 1.1618625277161863,
"grad_norm": 0.7615489363670349,
"learning_rate": 8.660021709405197e-06,
"loss": 0.9574247002601624,
"step": 524
},
{
"epoch": 1.1662971175166297,
"grad_norm": 0.9507637023925781,
"learning_rate": 8.648249165822265e-06,
"loss": 1.1572948694229126,
"step": 526
},
{
"epoch": 1.170731707317073,
"grad_norm": 1.4296756982803345,
"learning_rate": 8.636434266367956e-06,
"loss": 0.6602965593338013,
"step": 528
},
{
"epoch": 1.1751662971175167,
"grad_norm": 1.0683335065841675,
"learning_rate": 8.624577169997394e-06,
"loss": 0.8447040319442749,
"step": 530
},
{
"epoch": 1.1796008869179602,
"grad_norm": 0.4889877438545227,
"learning_rate": 8.612678036233428e-06,
"loss": 0.8834771513938904,
"step": 532
},
{
"epoch": 1.1840354767184036,
"grad_norm": 0.42208215594291687,
"learning_rate": 8.600737025164454e-06,
"loss": 0.9915164709091187,
"step": 534
},
{
"epoch": 1.188470066518847,
"grad_norm": 0.32866641879081726,
"learning_rate": 8.588754297442288e-06,
"loss": 0.7454093098640442,
"step": 536
},
{
"epoch": 1.1929046563192904,
"grad_norm": 1.7769194841384888,
"learning_rate": 8.576730014279982e-06,
"loss": 1.0999675989151,
"step": 538
},
{
"epoch": 1.1973392461197339,
"grad_norm": 0.7567623257637024,
"learning_rate": 8.564664337449677e-06,
"loss": 0.7065927982330322,
"step": 540
},
{
"epoch": 1.2017738359201773,
"grad_norm": 1.5624749660491943,
"learning_rate": 8.552557429280407e-06,
"loss": 0.6832781434059143,
"step": 542
},
{
"epoch": 1.206208425720621,
"grad_norm": 1.4154813289642334,
"learning_rate": 8.540409452655927e-06,
"loss": 1.0527125597000122,
"step": 544
},
{
"epoch": 1.2106430155210643,
"grad_norm": 0.3779228627681732,
"learning_rate": 8.528220571012518e-06,
"loss": 0.8481003642082214,
"step": 546
},
{
"epoch": 1.2150776053215078,
"grad_norm": 0.7003396153450012,
"learning_rate": 8.51599094833679e-06,
"loss": 1.2240773439407349,
"step": 548
},
{
"epoch": 1.2195121951219512,
"grad_norm": 0.9287251234054565,
"learning_rate": 8.503720749163472e-06,
"loss": 0.629622220993042,
"step": 550
},
{
"epoch": 1.2239467849223946,
"grad_norm": 1.5095727443695068,
"learning_rate": 8.491410138573201e-06,
"loss": 1.0369082689285278,
"step": 552
},
{
"epoch": 1.2283813747228383,
"grad_norm": 1.6644355058670044,
"learning_rate": 8.479059282190298e-06,
"loss": 0.7387758493423462,
"step": 554
},
{
"epoch": 1.2328159645232817,
"grad_norm": 0.320863276720047,
"learning_rate": 8.466668346180548e-06,
"loss": 0.810981035232544,
"step": 556
},
{
"epoch": 1.237250554323725,
"grad_norm": 1.620963215827942,
"learning_rate": 8.454237497248956e-06,
"loss": 0.6092830300331116,
"step": 558
},
{
"epoch": 1.2416851441241685,
"grad_norm": 1.0662033557891846,
"learning_rate": 8.441766902637506e-06,
"loss": 1.2609363794326782,
"step": 560
},
{
"epoch": 1.246119733924612,
"grad_norm": 1.3801884651184082,
"learning_rate": 8.429256730122909e-06,
"loss": 0.859289824962616,
"step": 562
},
{
"epoch": 1.2505543237250554,
"grad_norm": 0.6064059138298035,
"learning_rate": 8.416707148014358e-06,
"loss": 0.6251385807991028,
"step": 564
},
{
"epoch": 1.2549889135254988,
"grad_norm": 0.5259643197059631,
"learning_rate": 8.404118325151245e-06,
"loss": 0.9294739961624146,
"step": 566
},
{
"epoch": 1.2594235033259422,
"grad_norm": 0.9450163841247559,
"learning_rate": 8.391490430900902e-06,
"loss": 0.9038935303688049,
"step": 568
},
{
"epoch": 1.2638580931263859,
"grad_norm": 0.5424582362174988,
"learning_rate": 8.378823635156319e-06,
"loss": 1.1193132400512695,
"step": 570
},
{
"epoch": 1.2682926829268293,
"grad_norm": 0.3963155448436737,
"learning_rate": 8.366118108333861e-06,
"loss": 0.7388672232627869,
"step": 572
},
{
"epoch": 1.2727272727272727,
"grad_norm": 0.823639452457428,
"learning_rate": 8.353374021370967e-06,
"loss": 1.1983304023742676,
"step": 574
},
{
"epoch": 1.2771618625277161,
"grad_norm": 0.5981648564338684,
"learning_rate": 8.340591545723861e-06,
"loss": 1.0332672595977783,
"step": 576
},
{
"epoch": 1.2815964523281598,
"grad_norm": 0.6439179182052612,
"learning_rate": 8.327770853365238e-06,
"loss": 1.1301418542861938,
"step": 578
},
{
"epoch": 1.2860310421286032,
"grad_norm": 0.5708024501800537,
"learning_rate": 8.314912116781954e-06,
"loss": 1.0946240425109863,
"step": 580
},
{
"epoch": 1.2904656319290466,
"grad_norm": 1.2981244325637817,
"learning_rate": 8.302015508972702e-06,
"loss": 1.3642780780792236,
"step": 582
},
{
"epoch": 1.29490022172949,
"grad_norm": 0.48043203353881836,
"learning_rate": 8.289081203445686e-06,
"loss": 0.7540053129196167,
"step": 584
},
{
"epoch": 1.2993348115299335,
"grad_norm": 1.8940509557724,
"learning_rate": 8.276109374216286e-06,
"loss": 0.764034628868103,
"step": 586
},
{
"epoch": 1.3037694013303769,
"grad_norm": 1.8195059299468994,
"learning_rate": 8.263100195804722e-06,
"loss": 0.5129228234291077,
"step": 588
},
{
"epoch": 1.3082039911308203,
"grad_norm": 1.5529930591583252,
"learning_rate": 8.250053843233704e-06,
"loss": 1.0290241241455078,
"step": 590
},
{
"epoch": 1.3126385809312637,
"grad_norm": 0.5189142823219299,
"learning_rate": 8.236970492026063e-06,
"loss": 1.0039256811141968,
"step": 592
},
{
"epoch": 1.3170731707317074,
"grad_norm": 1.0931061506271362,
"learning_rate": 8.223850318202415e-06,
"loss": 1.0637818574905396,
"step": 594
},
{
"epoch": 1.3215077605321508,
"grad_norm": 0.4218553602695465,
"learning_rate": 8.210693498278773e-06,
"loss": 1.108015537261963,
"step": 596
},
{
"epoch": 1.3259423503325942,
"grad_norm": 0.7798098921775818,
"learning_rate": 8.197500209264181e-06,
"loss": 1.029466152191162,
"step": 598
},
{
"epoch": 1.3303769401330376,
"grad_norm": 0.5583693981170654,
"learning_rate": 8.18427062865833e-06,
"loss": 0.49875786900520325,
"step": 600
},
{
"epoch": 1.3348115299334813,
"grad_norm": 1.674991488456726,
"learning_rate": 8.171004934449166e-06,
"loss": 0.9001069664955139,
"step": 602
},
{
"epoch": 1.3392461197339247,
"grad_norm": 1.0978165864944458,
"learning_rate": 8.157703305110508e-06,
"loss": 0.860775887966156,
"step": 604
},
{
"epoch": 1.3436807095343681,
"grad_norm": 0.3894875645637512,
"learning_rate": 8.144365919599632e-06,
"loss": 0.7187979221343994,
"step": 606
},
{
"epoch": 1.3481152993348116,
"grad_norm": 0.8448476791381836,
"learning_rate": 8.130992957354872e-06,
"loss": 1.1956108808517456,
"step": 608
},
{
"epoch": 1.352549889135255,
"grad_norm": 2.5347466468811035,
"learning_rate": 8.117584598293204e-06,
"loss": 1.1394492387771606,
"step": 610
},
{
"epoch": 1.3569844789356984,
"grad_norm": 0.7789635062217712,
"learning_rate": 8.104141022807824e-06,
"loss": 0.7438350319862366,
"step": 612
},
{
"epoch": 1.3614190687361418,
"grad_norm": 0.6115071773529053,
"learning_rate": 8.090662411765726e-06,
"loss": 1.1242707967758179,
"step": 614
},
{
"epoch": 1.3658536585365852,
"grad_norm": 0.8269447088241577,
"learning_rate": 8.077148946505258e-06,
"loss": 0.8168088793754578,
"step": 616
},
{
"epoch": 1.370288248337029,
"grad_norm": 0.5584709048271179,
"learning_rate": 8.063600808833698e-06,
"loss": 0.9382954239845276,
"step": 618
},
{
"epoch": 1.3747228381374723,
"grad_norm": 0.3414313495159149,
"learning_rate": 8.050018181024788e-06,
"loss": 0.1388121247291565,
"step": 620
},
{
"epoch": 1.3791574279379157,
"grad_norm": 0.46918442845344543,
"learning_rate": 8.036401245816306e-06,
"loss": 0.9869102835655212,
"step": 622
},
{
"epoch": 1.3835920177383592,
"grad_norm": 0.9420056939125061,
"learning_rate": 8.022750186407586e-06,
"loss": 0.9787265062332153,
"step": 624
},
{
"epoch": 1.3880266075388026,
"grad_norm": 1.6520426273345947,
"learning_rate": 8.009065186457061e-06,
"loss": 0.9564573764801025,
"step": 626
},
{
"epoch": 1.3924611973392462,
"grad_norm": 0.7199759483337402,
"learning_rate": 7.995346430079799e-06,
"loss": 0.949640691280365,
"step": 628
},
{
"epoch": 1.3968957871396896,
"grad_norm": 0.4535790979862213,
"learning_rate": 7.981594101845012e-06,
"loss": 1.0623891353607178,
"step": 630
},
{
"epoch": 1.401330376940133,
"grad_norm": 0.4923252463340759,
"learning_rate": 7.967808386773591e-06,
"loss": 0.736829400062561,
"step": 632
},
{
"epoch": 1.4057649667405765,
"grad_norm": 1.7496610879898071,
"learning_rate": 7.953989470335592e-06,
"loss": 1.0442922115325928,
"step": 634
},
{
"epoch": 1.41019955654102,
"grad_norm": 2.604696750640869,
"learning_rate": 7.940137538447769e-06,
"loss": 0.9943015575408936,
"step": 636
},
{
"epoch": 1.4146341463414633,
"grad_norm": 0.8832758069038391,
"learning_rate": 7.92625277747105e-06,
"loss": 0.6925525069236755,
"step": 638
},
{
"epoch": 1.4190687361419068,
"grad_norm": 0.976290225982666,
"learning_rate": 7.912335374208043e-06,
"loss": 0.8791317939758301,
"step": 640
},
{
"epoch": 1.4235033259423504,
"grad_norm": 1.112613558769226,
"learning_rate": 7.898385515900517e-06,
"loss": 0.8423402309417725,
"step": 642
},
{
"epoch": 1.4279379157427938,
"grad_norm": 4.072666645050049,
"learning_rate": 7.884403390226883e-06,
"loss": 0.5500181317329407,
"step": 644
},
{
"epoch": 1.4323725055432373,
"grad_norm": 1.1002585887908936,
"learning_rate": 7.870389185299672e-06,
"loss": 1.159914255142212,
"step": 646
},
{
"epoch": 1.4368070953436807,
"grad_norm": 1.006394386291504,
"learning_rate": 7.856343089663002e-06,
"loss": 1.0645030736923218,
"step": 648
},
{
"epoch": 1.441241685144124,
"grad_norm": 0.7567451000213623,
"learning_rate": 7.842265292290039e-06,
"loss": 1.143546223640442,
"step": 650
},
{
"epoch": 1.4456762749445677,
"grad_norm": 1.4760500192642212,
"learning_rate": 7.828155982580465e-06,
"loss": 0.9194218516349792,
"step": 652
},
{
"epoch": 1.4501108647450112,
"grad_norm": 1.3499608039855957,
"learning_rate": 7.814015350357912e-06,
"loss": 1.0461726188659668,
"step": 654
},
{
"epoch": 1.4545454545454546,
"grad_norm": 0.5506664514541626,
"learning_rate": 7.799843585867426e-06,
"loss": 0.7790355086326599,
"step": 656
},
{
"epoch": 1.458980044345898,
"grad_norm": 0.4337323307991028,
"learning_rate": 7.785640879772897e-06,
"loss": 0.7051459550857544,
"step": 658
},
{
"epoch": 1.4634146341463414,
"grad_norm": 0.4734741449356079,
"learning_rate": 7.771407423154498e-06,
"loss": 1.015939712524414,
"step": 660
},
{
"epoch": 1.4678492239467849,
"grad_norm": 0.440094918012619,
"learning_rate": 7.757143407506111e-06,
"loss": 0.9792048931121826,
"step": 662
},
{
"epoch": 1.4722838137472283,
"grad_norm": 0.7068251967430115,
"learning_rate": 7.742849024732754e-06,
"loss": 0.7712748050689697,
"step": 664
},
{
"epoch": 1.476718403547672,
"grad_norm": 0.85067218542099,
"learning_rate": 7.728524467148e-06,
"loss": 0.6657487750053406,
"step": 666
},
{
"epoch": 1.4811529933481153,
"grad_norm": 1.339653730392456,
"learning_rate": 7.714169927471379e-06,
"loss": 0.5245797634124756,
"step": 668
},
{
"epoch": 1.4855875831485588,
"grad_norm": 1.0584568977355957,
"learning_rate": 7.699785598825805e-06,
"loss": 0.9373866319656372,
"step": 670
},
{
"epoch": 1.4900221729490022,
"grad_norm": 0.9347496032714844,
"learning_rate": 7.68537167473496e-06,
"loss": 1.0927127599716187,
"step": 672
},
{
"epoch": 1.4944567627494456,
"grad_norm": 0.5240381956100464,
"learning_rate": 7.670928349120699e-06,
"loss": 0.6736969351768494,
"step": 674
},
{
"epoch": 1.4988913525498893,
"grad_norm": 0.4406070411205292,
"learning_rate": 7.656455816300434e-06,
"loss": 0.7100923657417297,
"step": 676
},
{
"epoch": 1.5033259423503327,
"grad_norm": 2.0313918590545654,
"learning_rate": 7.641954270984532e-06,
"loss": 1.05730402469635,
"step": 678
},
{
"epoch": 1.507760532150776,
"grad_norm": 0.420768678188324,
"learning_rate": 7.627423908273683e-06,
"loss": 0.570051908493042,
"step": 680
},
{
"epoch": 1.5121951219512195,
"grad_norm": 0.6675831079483032,
"learning_rate": 7.61286492365628e-06,
"loss": 0.5550310015678406,
"step": 682
},
{
"epoch": 1.516629711751663,
"grad_norm": 0.7205184102058411,
"learning_rate": 7.598277513005793e-06,
"loss": 1.0391545295715332,
"step": 684
},
{
"epoch": 1.5210643015521064,
"grad_norm": 0.4342437982559204,
"learning_rate": 7.583661872578124e-06,
"loss": 0.9450809955596924,
"step": 686
},
{
"epoch": 1.5254988913525498,
"grad_norm": 1.2098145484924316,
"learning_rate": 7.569018199008976e-06,
"loss": 0.5710777640342712,
"step": 688
},
{
"epoch": 1.5299334811529932,
"grad_norm": 0.5424330234527588,
"learning_rate": 7.554346689311205e-06,
"loss": 0.29212188720703125,
"step": 690
},
{
"epoch": 1.5343680709534369,
"grad_norm": 0.7510618567466736,
"learning_rate": 7.539647540872165e-06,
"loss": 0.7841662764549255,
"step": 692
},
{
"epoch": 1.5388026607538803,
"grad_norm": 1.6880568265914917,
"learning_rate": 7.5249209514510595e-06,
"loss": 0.9736351370811462,
"step": 694
},
{
"epoch": 1.5432372505543237,
"grad_norm": 1.3462926149368286,
"learning_rate": 7.510167119176273e-06,
"loss": 0.6195929646492004,
"step": 696
},
{
"epoch": 1.5476718403547673,
"grad_norm": 0.47400352358818054,
"learning_rate": 7.49538624254271e-06,
"loss": 1.0942884683609009,
"step": 698
},
{
"epoch": 1.5521064301552108,
"grad_norm": 0.4019659161567688,
"learning_rate": 7.48057852040913e-06,
"loss": 1.01500403881073,
"step": 700
},
{
"epoch": 1.5565410199556542,
"grad_norm": 0.7360739707946777,
"learning_rate": 7.465744151995458e-06,
"loss": 1.0028947591781616,
"step": 702
},
{
"epoch": 1.5609756097560976,
"grad_norm": 1.1567124128341675,
"learning_rate": 7.450883336880116e-06,
"loss": 0.9144392609596252,
"step": 704
},
{
"epoch": 1.565410199556541,
"grad_norm": 0.9256898760795593,
"learning_rate": 7.435996274997337e-06,
"loss": 0.45123380422592163,
"step": 706
},
{
"epoch": 1.5698447893569845,
"grad_norm": 0.4010877311229706,
"learning_rate": 7.421083166634466e-06,
"loss": 1.0565845966339111,
"step": 708
},
{
"epoch": 1.5742793791574279,
"grad_norm": 0.986020565032959,
"learning_rate": 7.40614421242928e-06,
"loss": 0.8794451951980591,
"step": 710
},
{
"epoch": 1.5787139689578713,
"grad_norm": 1.3509045839309692,
"learning_rate": 7.391179613367272e-06,
"loss": 0.6458362936973572,
"step": 712
},
{
"epoch": 1.5831485587583147,
"grad_norm": 0.43885716795921326,
"learning_rate": 7.37618957077896e-06,
"loss": 1.005838394165039,
"step": 714
},
{
"epoch": 1.5875831485587582,
"grad_norm": 0.5049751400947571,
"learning_rate": 7.361174286337175e-06,
"loss": 1.0728709697723389,
"step": 716
},
{
"epoch": 1.5920177383592018,
"grad_norm": 1.5042154788970947,
"learning_rate": 7.346133962054341e-06,
"loss": 1.202171802520752,
"step": 718
},
{
"epoch": 1.5964523281596452,
"grad_norm": 2.809225559234619,
"learning_rate": 7.33106880027977e-06,
"loss": 0.5905061364173889,
"step": 720
},
{
"epoch": 1.6008869179600886,
"grad_norm": 0.27043482661247253,
"learning_rate": 7.315979003696927e-06,
"loss": 0.521863579750061,
"step": 722
},
{
"epoch": 1.6053215077605323,
"grad_norm": 0.41005781292915344,
"learning_rate": 7.300864775320708e-06,
"loss": 1.0308693647384644,
"step": 724
},
{
"epoch": 1.6097560975609757,
"grad_norm": 1.3939813375473022,
"learning_rate": 7.285726318494717e-06,
"loss": 1.0042451620101929,
"step": 726
},
{
"epoch": 1.6141906873614191,
"grad_norm": 0.4165089428424835,
"learning_rate": 7.2705638368885105e-06,
"loss": 0.7745837569236755,
"step": 728
},
{
"epoch": 1.6186252771618626,
"grad_norm": 0.5574784874916077,
"learning_rate": 7.255377534494875e-06,
"loss": 1.0095455646514893,
"step": 730
},
{
"epoch": 1.623059866962306,
"grad_norm": 0.5945760607719421,
"learning_rate": 7.240167615627082e-06,
"loss": 1.0015901327133179,
"step": 732
},
{
"epoch": 1.6274944567627494,
"grad_norm": 0.5041150450706482,
"learning_rate": 7.224934284916127e-06,
"loss": 0.9732311964035034,
"step": 734
},
{
"epoch": 1.6319290465631928,
"grad_norm": 0.3137003481388092,
"learning_rate": 7.209677747307982e-06,
"loss": 0.5148752331733704,
"step": 736
},
{
"epoch": 1.6363636363636362,
"grad_norm": 0.47104862332344055,
"learning_rate": 7.194398208060848e-06,
"loss": 1.2126970291137695,
"step": 738
},
{
"epoch": 1.6407982261640797,
"grad_norm": 0.3913891613483429,
"learning_rate": 7.179095872742378e-06,
"loss": 0.985205888748169,
"step": 740
},
{
"epoch": 1.6452328159645233,
"grad_norm": 1.6343456506729126,
"learning_rate": 7.16377094722692e-06,
"loss": 1.1372039318084717,
"step": 742
},
{
"epoch": 1.6496674057649667,
"grad_norm": 5.484721660614014,
"learning_rate": 7.148423637692748e-06,
"loss": 0.8813603520393372,
"step": 744
},
{
"epoch": 1.6541019955654102,
"grad_norm": 1.7607204914093018,
"learning_rate": 7.133054150619282e-06,
"loss": 0.8659079074859619,
"step": 746
},
{
"epoch": 1.6585365853658538,
"grad_norm": 0.5767048597335815,
"learning_rate": 7.117662692784318e-06,
"loss": 0.6968775987625122,
"step": 748
},
{
"epoch": 1.6629711751662972,
"grad_norm": 0.4937343895435333,
"learning_rate": 7.102249471261241e-06,
"loss": 0.9998278617858887,
"step": 750
},
{
"epoch": 1.6674057649667406,
"grad_norm": 0.5310612320899963,
"learning_rate": 7.0868146934162365e-06,
"loss": 0.7577317357063293,
"step": 752
},
{
"epoch": 1.671840354767184,
"grad_norm": 0.5454942584037781,
"learning_rate": 7.071358566905507e-06,
"loss": 0.6115183234214783,
"step": 754
},
{
"epoch": 1.6762749445676275,
"grad_norm": 0.6721274256706238,
"learning_rate": 7.055881299672476e-06,
"loss": 0.9478708505630493,
"step": 756
},
{
"epoch": 1.680709534368071,
"grad_norm": 0.5245542526245117,
"learning_rate": 7.040383099944988e-06,
"loss": 0.9483290314674377,
"step": 758
},
{
"epoch": 1.6851441241685143,
"grad_norm": 1.0505019426345825,
"learning_rate": 7.02486417623251e-06,
"loss": 0.9622657895088196,
"step": 760
},
{
"epoch": 1.6895787139689578,
"grad_norm": 1.2052677869796753,
"learning_rate": 7.009324737323325e-06,
"loss": 0.9271701574325562,
"step": 762
},
{
"epoch": 1.6940133037694012,
"grad_norm": 0.8925173282623291,
"learning_rate": 6.993764992281722e-06,
"loss": 0.962370753288269,
"step": 764
},
{
"epoch": 1.6984478935698448,
"grad_norm": 1.8376715183258057,
"learning_rate": 6.978185150445187e-06,
"loss": 1.0133888721466064,
"step": 766
},
{
"epoch": 1.7028824833702882,
"grad_norm": 0.8429685235023499,
"learning_rate": 6.96258542142158e-06,
"loss": 0.7847933769226074,
"step": 768
},
{
"epoch": 1.7073170731707317,
"grad_norm": 0.451034277677536,
"learning_rate": 6.946966015086321e-06,
"loss": 0.9764997363090515,
"step": 770
},
{
"epoch": 1.7117516629711753,
"grad_norm": 2.5368645191192627,
"learning_rate": 6.931327141579565e-06,
"loss": 1.0000364780426025,
"step": 772
},
{
"epoch": 1.7161862527716187,
"grad_norm": 0.48420414328575134,
"learning_rate": 6.915669011303374e-06,
"loss": 0.8809909820556641,
"step": 774
},
{
"epoch": 1.7206208425720622,
"grad_norm": 0.5247390866279602,
"learning_rate": 6.899991834918884e-06,
"loss": 1.0189868211746216,
"step": 776
},
{
"epoch": 1.7250554323725056,
"grad_norm": 0.690968930721283,
"learning_rate": 6.884295823343479e-06,
"loss": 1.035364031791687,
"step": 778
},
{
"epoch": 1.729490022172949,
"grad_norm": 0.9502907395362854,
"learning_rate": 6.868581187747941e-06,
"loss": 0.6463431715965271,
"step": 780
},
{
"epoch": 1.7339246119733924,
"grad_norm": 0.5312066674232483,
"learning_rate": 6.852848139553619e-06,
"loss": 0.9688764810562134,
"step": 782
},
{
"epoch": 1.7383592017738358,
"grad_norm": 0.5402536392211914,
"learning_rate": 6.837096890429582e-06,
"loss": 0.6454051733016968,
"step": 784
},
{
"epoch": 1.7427937915742793,
"grad_norm": 0.48874494433403015,
"learning_rate": 6.821327652289768e-06,
"loss": 0.7339925169944763,
"step": 786
},
{
"epoch": 1.7472283813747227,
"grad_norm": 0.6470894813537598,
"learning_rate": 6.8055406372901344e-06,
"loss": 0.6527358293533325,
"step": 788
},
{
"epoch": 1.7516629711751663,
"grad_norm": 0.9488041400909424,
"learning_rate": 6.789736057825812e-06,
"loss": 0.650081992149353,
"step": 790
},
{
"epoch": 1.7560975609756098,
"grad_norm": 0.4228132665157318,
"learning_rate": 6.77391412652823e-06,
"loss": 0.650942862033844,
"step": 792
},
{
"epoch": 1.7605321507760532,
"grad_norm": 0.5584551692008972,
"learning_rate": 6.758075056262271e-06,
"loss": 0.8236819505691528,
"step": 794
},
{
"epoch": 1.7649667405764968,
"grad_norm": 1.0599656105041504,
"learning_rate": 6.742219060123403e-06,
"loss": 0.7987365126609802,
"step": 796
},
{
"epoch": 1.7694013303769403,
"grad_norm": 0.6103171110153198,
"learning_rate": 6.7263463514348095e-06,
"loss": 0.9921965003013611,
"step": 798
},
{
"epoch": 1.7738359201773837,
"grad_norm": 0.6360469460487366,
"learning_rate": 6.710457143744519e-06,
"loss": 1.0037264823913574,
"step": 800
},
{
"epoch": 1.778270509977827,
"grad_norm": 0.5509381890296936,
"learning_rate": 6.6945516508225325e-06,
"loss": 0.992444634437561,
"step": 802
},
{
"epoch": 1.7827050997782705,
"grad_norm": 0.6882392168045044,
"learning_rate": 6.678630086657959e-06,
"loss": 1.0927345752716064,
"step": 804
},
{
"epoch": 1.787139689578714,
"grad_norm": 0.6332179307937622,
"learning_rate": 6.662692665456115e-06,
"loss": 0.7333803772926331,
"step": 806
},
{
"epoch": 1.7915742793791574,
"grad_norm": 0.5961802005767822,
"learning_rate": 6.646739601635661e-06,
"loss": 1.0272966623306274,
"step": 808
},
{
"epoch": 1.7960088691796008,
"grad_norm": 1.2394970655441284,
"learning_rate": 6.6307711098257074e-06,
"loss": 0.6059154272079468,
"step": 810
},
{
"epoch": 1.8004434589800442,
"grad_norm": 1.1179450750350952,
"learning_rate": 6.6147874048629294e-06,
"loss": 0.6130510568618774,
"step": 812
},
{
"epoch": 1.8048780487804879,
"grad_norm": 1.7669285535812378,
"learning_rate": 6.598788701788677e-06,
"loss": 1.039893388748169,
"step": 814
},
{
"epoch": 1.8093126385809313,
"grad_norm": 4.2482709884643555,
"learning_rate": 6.582775215846082e-06,
"loss": 0.9353387355804443,
"step": 816
},
{
"epoch": 1.8137472283813747,
"grad_norm": 0.8275613188743591,
"learning_rate": 6.566747162477164e-06,
"loss": 0.4844290316104889,
"step": 818
},
{
"epoch": 1.8181818181818183,
"grad_norm": 0.8788929581642151,
"learning_rate": 6.5507047573199235e-06,
"loss": 1.1249672174453735,
"step": 820
},
{
"epoch": 1.8226164079822618,
"grad_norm": 0.8800422549247742,
"learning_rate": 6.5346482162054526e-06,
"loss": 1.1043124198913574,
"step": 822
},
{
"epoch": 1.8270509977827052,
"grad_norm": 0.4852801263332367,
"learning_rate": 6.518577755155024e-06,
"loss": 1.1537041664123535,
"step": 824
},
{
"epoch": 1.8314855875831486,
"grad_norm": 1.2431832551956177,
"learning_rate": 6.502493590377184e-06,
"loss": 0.8821326494216919,
"step": 826
},
{
"epoch": 1.835920177383592,
"grad_norm": 0.9608327746391296,
"learning_rate": 6.48639593826485e-06,
"loss": 0.8124348521232605,
"step": 828
},
{
"epoch": 1.8403547671840355,
"grad_norm": 0.7294570803642273,
"learning_rate": 6.4702850153923915e-06,
"loss": 0.9093976616859436,
"step": 830
},
{
"epoch": 1.8447893569844789,
"grad_norm": 0.575531542301178,
"learning_rate": 6.45416103851272e-06,
"loss": 1.0434691905975342,
"step": 832
},
{
"epoch": 1.8492239467849223,
"grad_norm": 0.4895908236503601,
"learning_rate": 6.438024224554378e-06,
"loss": 0.9609898328781128,
"step": 834
},
{
"epoch": 1.8536585365853657,
"grad_norm": 0.33139824867248535,
"learning_rate": 6.421874790618608e-06,
"loss": 1.0044034719467163,
"step": 836
},
{
"epoch": 1.8580931263858091,
"grad_norm": 0.6011033654212952,
"learning_rate": 6.405712953976444e-06,
"loss": 0.9612147808074951,
"step": 838
},
{
"epoch": 1.8625277161862528,
"grad_norm": 0.7546729445457458,
"learning_rate": 6.389538932065783e-06,
"loss": 1.141190767288208,
"step": 840
},
{
"epoch": 1.8669623059866962,
"grad_norm": 0.6122239828109741,
"learning_rate": 6.373352942488455e-06,
"loss": 0.6330398321151733,
"step": 842
},
{
"epoch": 1.8713968957871396,
"grad_norm": 0.5186591148376465,
"learning_rate": 6.357155203007307e-06,
"loss": 0.7644712924957275,
"step": 844
},
{
"epoch": 1.8758314855875833,
"grad_norm": 0.8722115159034729,
"learning_rate": 6.340945931543263e-06,
"loss": 0.7418773770332336,
"step": 846
},
{
"epoch": 1.8802660753880267,
"grad_norm": 1.3620861768722534,
"learning_rate": 6.324725346172399e-06,
"loss": 0.9680192470550537,
"step": 848
},
{
"epoch": 1.8847006651884701,
"grad_norm": 0.49945542216300964,
"learning_rate": 6.308493665123e-06,
"loss": 1.0896743535995483,
"step": 850
},
{
"epoch": 1.8891352549889135,
"grad_norm": 0.8075128793716431,
"learning_rate": 6.2922511067726365e-06,
"loss": 0.667526364326477,
"step": 852
},
{
"epoch": 1.893569844789357,
"grad_norm": 0.7008575201034546,
"learning_rate": 6.2759978896452155e-06,
"loss": 1.0155971050262451,
"step": 854
},
{
"epoch": 1.8980044345898004,
"grad_norm": 0.6262828707695007,
"learning_rate": 6.259734232408047e-06,
"loss": 0.9141192436218262,
"step": 856
},
{
"epoch": 1.9024390243902438,
"grad_norm": 0.5417189002037048,
"learning_rate": 6.2434603538688975e-06,
"loss": 0.7273485660552979,
"step": 858
},
{
"epoch": 1.9068736141906872,
"grad_norm": 1.6091915369033813,
"learning_rate": 6.2271764729730525e-06,
"loss": 0.7897534370422363,
"step": 860
},
{
"epoch": 1.9113082039911307,
"grad_norm": 1.3260005712509155,
"learning_rate": 6.210882808800366e-06,
"loss": 1.0450389385223389,
"step": 862
},
{
"epoch": 1.9157427937915743,
"grad_norm": 0.47777101397514343,
"learning_rate": 6.19457958056231e-06,
"loss": 0.7668794393539429,
"step": 864
},
{
"epoch": 1.9201773835920177,
"grad_norm": 0.9053187966346741,
"learning_rate": 6.178267007599034e-06,
"loss": 0.6413880586624146,
"step": 866
},
{
"epoch": 1.9246119733924612,
"grad_norm": 0.6551345586776733,
"learning_rate": 6.161945309376409e-06,
"loss": 1.095296859741211,
"step": 868
},
{
"epoch": 1.9290465631929048,
"grad_norm": 0.878730297088623,
"learning_rate": 6.145614705483075e-06,
"loss": 0.9538235068321228,
"step": 870
},
{
"epoch": 1.9334811529933482,
"grad_norm": 0.5680310726165771,
"learning_rate": 6.129275415627485e-06,
"loss": 0.6358171701431274,
"step": 872
},
{
"epoch": 1.9379157427937916,
"grad_norm": 0.6420407295227051,
"learning_rate": 6.11292765963495e-06,
"loss": 0.6467251181602478,
"step": 874
},
{
"epoch": 1.942350332594235,
"grad_norm": 0.5206342935562134,
"learning_rate": 6.09657165744469e-06,
"loss": 0.6561703681945801,
"step": 876
},
{
"epoch": 1.9467849223946785,
"grad_norm": 0.6010205745697021,
"learning_rate": 6.080207629106859e-06,
"loss": 0.7634451389312744,
"step": 878
},
{
"epoch": 1.951219512195122,
"grad_norm": 0.5291740894317627,
"learning_rate": 6.063835794779598e-06,
"loss": 0.85489821434021,
"step": 880
},
{
"epoch": 1.9556541019955653,
"grad_norm": 4.159445762634277,
"learning_rate": 6.047456374726067e-06,
"loss": 0.6732465028762817,
"step": 882
},
{
"epoch": 1.9600886917960088,
"grad_norm": 0.47579193115234375,
"learning_rate": 6.031069589311481e-06,
"loss": 0.7860453128814697,
"step": 884
},
{
"epoch": 1.9645232815964522,
"grad_norm": 0.4395957887172699,
"learning_rate": 6.01467565900015e-06,
"loss": 0.9734126925468445,
"step": 886
},
{
"epoch": 1.9689578713968958,
"grad_norm": 2.2899482250213623,
"learning_rate": 5.99827480435251e-06,
"loss": 0.9537763595581055,
"step": 888
},
{
"epoch": 1.9733924611973392,
"grad_norm": 0.5516371130943298,
"learning_rate": 5.981867246022149e-06,
"loss": 0.99347984790802,
"step": 890
},
{
"epoch": 1.9778270509977827,
"grad_norm": 0.5941886901855469,
"learning_rate": 5.965453204752855e-06,
"loss": 1.0517255067825317,
"step": 892
},
{
"epoch": 1.9822616407982263,
"grad_norm": 0.631015419960022,
"learning_rate": 5.949032901375627e-06,
"loss": 1.0343537330627441,
"step": 894
},
{
"epoch": 1.9866962305986697,
"grad_norm": 1.527363657951355,
"learning_rate": 5.932606556805719e-06,
"loss": 0.8417173624038696,
"step": 896
},
{
"epoch": 1.9911308203991132,
"grad_norm": 0.7326000332832336,
"learning_rate": 5.916174392039659e-06,
"loss": 0.9509387016296387,
"step": 898
},
{
"epoch": 1.9955654101995566,
"grad_norm": 0.5246658325195312,
"learning_rate": 5.899736628152284e-06,
"loss": 0.798784613609314,
"step": 900
},
{
"epoch": 2.0,
"grad_norm": 0.41008785367012024,
"learning_rate": 5.88329348629375e-06,
"loss": 0.985916018486023,
"step": 902
},
{
"epoch": 2.0044345898004434,
"grad_norm": 0.4405089318752289,
"learning_rate": 5.8668451876865736e-06,
"loss": 0.7073619365692139,
"step": 904
},
{
"epoch": 2.008869179600887,
"grad_norm": 1.7787858247756958,
"learning_rate": 5.850391953622652e-06,
"loss": 0.8173685669898987,
"step": 906
},
{
"epoch": 2.0133037694013303,
"grad_norm": 0.5108776092529297,
"learning_rate": 5.8339340054602775e-06,
"loss": 0.6087190508842468,
"step": 908
},
{
"epoch": 2.0177383592017737,
"grad_norm": 0.5111180543899536,
"learning_rate": 5.817471564621169e-06,
"loss": 0.7211825847625732,
"step": 910
},
{
"epoch": 2.022172949002217,
"grad_norm": 0.7296714186668396,
"learning_rate": 5.801004852587485e-06,
"loss": 0.507999837398529,
"step": 912
},
{
"epoch": 2.0266075388026605,
"grad_norm": 1.6111247539520264,
"learning_rate": 5.784534090898849e-06,
"loss": 0.5558914542198181,
"step": 914
},
{
"epoch": 2.0310421286031044,
"grad_norm": 0.9995886087417603,
"learning_rate": 5.768059501149369e-06,
"loss": 0.5170501470565796,
"step": 916
},
{
"epoch": 2.035476718403548,
"grad_norm": 0.5618699193000793,
"learning_rate": 5.751581304984657e-06,
"loss": 0.7065697908401489,
"step": 918
},
{
"epoch": 2.0399113082039912,
"grad_norm": 0.554235577583313,
"learning_rate": 5.735099724098838e-06,
"loss": 0.5513534545898438,
"step": 920
},
{
"epoch": 2.0443458980044347,
"grad_norm": 0.6221592426300049,
"learning_rate": 5.718614980231582e-06,
"loss": 0.580018162727356,
"step": 922
},
{
"epoch": 2.048780487804878,
"grad_norm": 0.7114652395248413,
"learning_rate": 5.702127295165107e-06,
"loss": 0.26266422867774963,
"step": 924
},
{
"epoch": 2.0532150776053215,
"grad_norm": 0.7338118553161621,
"learning_rate": 5.685636890721205e-06,
"loss": 0.7764136791229248,
"step": 926
},
{
"epoch": 2.057649667405765,
"grad_norm": 1.6588897705078125,
"learning_rate": 5.669143988758253e-06,
"loss": 0.3647955656051636,
"step": 928
},
{
"epoch": 2.0620842572062084,
"grad_norm": 0.25482437014579773,
"learning_rate": 5.652648811168228e-06,
"loss": 0.5260996222496033,
"step": 930
},
{
"epoch": 2.066518847006652,
"grad_norm": 1.088817834854126,
"learning_rate": 5.636151579873726e-06,
"loss": 0.5555666089057922,
"step": 932
},
{
"epoch": 2.070953436807095,
"grad_norm": 0.6878843307495117,
"learning_rate": 5.619652516824967e-06,
"loss": 0.6538913249969482,
"step": 934
},
{
"epoch": 2.0753880266075386,
"grad_norm": 0.5582793354988098,
"learning_rate": 5.603151843996822e-06,
"loss": 0.7564462423324585,
"step": 936
},
{
"epoch": 2.079822616407982,
"grad_norm": 0.5043266415596008,
"learning_rate": 5.586649783385813e-06,
"loss": 0.43487629294395447,
"step": 938
},
{
"epoch": 2.084257206208426,
"grad_norm": 0.964514434337616,
"learning_rate": 5.570146557007141e-06,
"loss": 0.4970490336418152,
"step": 940
},
{
"epoch": 2.0886917960088693,
"grad_norm": 0.8735406398773193,
"learning_rate": 5.553642386891683e-06,
"loss": 0.6485542058944702,
"step": 942
},
{
"epoch": 2.0931263858093128,
"grad_norm": 0.5761396288871765,
"learning_rate": 5.537137495083018e-06,
"loss": 0.49649330973625183,
"step": 944
},
{
"epoch": 2.097560975609756,
"grad_norm": 0.5437949895858765,
"learning_rate": 5.5206321036344304e-06,
"loss": 0.7156221270561218,
"step": 946
},
{
"epoch": 2.1019955654101996,
"grad_norm": 0.9228906035423279,
"learning_rate": 5.504126434605932e-06,
"loss": 0.7050195932388306,
"step": 948
},
{
"epoch": 2.106430155210643,
"grad_norm": 3.377256393432617,
"learning_rate": 5.487620710061262e-06,
"loss": 0.5274829268455505,
"step": 950
},
{
"epoch": 2.1108647450110865,
"grad_norm": 0.6513805985450745,
"learning_rate": 5.471115152064916e-06,
"loss": 0.5962385535240173,
"step": 952
},
{
"epoch": 2.11529933481153,
"grad_norm": 2.8682925701141357,
"learning_rate": 5.454609982679138e-06,
"loss": 0.6596555709838867,
"step": 954
},
{
"epoch": 2.1197339246119733,
"grad_norm": 0.6173861026763916,
"learning_rate": 5.4381054239609525e-06,
"loss": 0.7276114821434021,
"step": 956
},
{
"epoch": 2.1241685144124167,
"grad_norm": 0.5581717491149902,
"learning_rate": 5.421601697959164e-06,
"loss": 0.5600399374961853,
"step": 958
},
{
"epoch": 2.12860310421286,
"grad_norm": 1.5211955308914185,
"learning_rate": 5.405099026711374e-06,
"loss": 0.5050497055053711,
"step": 960
},
{
"epoch": 2.1330376940133036,
"grad_norm": 1.1427018642425537,
"learning_rate": 5.388597632240994e-06,
"loss": 0.36328691244125366,
"step": 962
},
{
"epoch": 2.1374722838137474,
"grad_norm": 1.5880637168884277,
"learning_rate": 5.372097736554261e-06,
"loss": 0.6904159784317017,
"step": 964
},
{
"epoch": 2.141906873614191,
"grad_norm": 0.6093219518661499,
"learning_rate": 5.35559956163724e-06,
"loss": 0.5827292799949646,
"step": 966
},
{
"epoch": 2.1463414634146343,
"grad_norm": 1.1137672662734985,
"learning_rate": 5.339103329452856e-06,
"loss": 0.47683635354042053,
"step": 968
},
{
"epoch": 2.1507760532150777,
"grad_norm": 0.8449851870536804,
"learning_rate": 5.322609261937887e-06,
"loss": 0.5655550360679626,
"step": 970
},
{
"epoch": 2.155210643015521,
"grad_norm": 0.5638793110847473,
"learning_rate": 5.306117580999993e-06,
"loss": 0.5006866455078125,
"step": 972
},
{
"epoch": 2.1596452328159645,
"grad_norm": 0.6014860272407532,
"learning_rate": 5.289628508514725e-06,
"loss": 0.7802212834358215,
"step": 974
},
{
"epoch": 2.164079822616408,
"grad_norm": 0.6840709447860718,
"learning_rate": 5.2731422663225385e-06,
"loss": 0.6838681101799011,
"step": 976
},
{
"epoch": 2.1685144124168514,
"grad_norm": 1.3036504983901978,
"learning_rate": 5.256659076225813e-06,
"loss": 0.41791778802871704,
"step": 978
},
{
"epoch": 2.172949002217295,
"grad_norm": 2.5424842834472656,
"learning_rate": 5.240179159985866e-06,
"loss": 0.7272178530693054,
"step": 980
},
{
"epoch": 2.1773835920177382,
"grad_norm": 0.5120069980621338,
"learning_rate": 5.2237027393199645e-06,
"loss": 0.3696378171443939,
"step": 982
},
{
"epoch": 2.1818181818181817,
"grad_norm": 0.654542088508606,
"learning_rate": 5.207230035898356e-06,
"loss": 0.21627305448055267,
"step": 984
},
{
"epoch": 2.186252771618625,
"grad_norm": 3.1496362686157227,
"learning_rate": 5.190761271341268e-06,
"loss": 0.5596635937690735,
"step": 986
},
{
"epoch": 2.1906873614190685,
"grad_norm": 1.2095357179641724,
"learning_rate": 5.174296667215939e-06,
"loss": 0.3137320280075073,
"step": 988
},
{
"epoch": 2.1951219512195124,
"grad_norm": 0.5719432234764099,
"learning_rate": 5.157836445033636e-06,
"loss": 0.7822216749191284,
"step": 990
},
{
"epoch": 2.199556541019956,
"grad_norm": 2.0039100646972656,
"learning_rate": 5.141380826246667e-06,
"loss": 0.7932789921760559,
"step": 992
},
{
"epoch": 2.203991130820399,
"grad_norm": 0.8511418700218201,
"learning_rate": 5.124930032245415e-06,
"loss": 0.5880918502807617,
"step": 994
},
{
"epoch": 2.2084257206208426,
"grad_norm": 0.6295179724693298,
"learning_rate": 5.108484284355339e-06,
"loss": 0.795894980430603,
"step": 996
},
{
"epoch": 2.212860310421286,
"grad_norm": 0.783069372177124,
"learning_rate": 5.0920438038340194e-06,
"loss": 0.6926091313362122,
"step": 998
},
{
"epoch": 2.2172949002217295,
"grad_norm": 0.6107291579246521,
"learning_rate": 5.075608811868169e-06,
"loss": 0.5917972326278687,
"step": 1000
},
{
"epoch": 2.221729490022173,
"grad_norm": 0.68993079662323,
"learning_rate": 5.059179529570657e-06,
"loss": 0.3001978397369385,
"step": 1002
},
{
"epoch": 2.2261640798226163,
"grad_norm": 0.775002121925354,
"learning_rate": 5.042756177977534e-06,
"loss": 0.7339842319488525,
"step": 1004
},
{
"epoch": 2.2305986696230597,
"grad_norm": 0.5174007415771484,
"learning_rate": 5.026338978045062e-06,
"loss": 0.600027322769165,
"step": 1006
},
{
"epoch": 2.235033259423503,
"grad_norm": 0.4449722468852997,
"learning_rate": 5.009928150646741e-06,
"loss": 0.7295263409614563,
"step": 1008
},
{
"epoch": 2.2394678492239466,
"grad_norm": 0.5435464382171631,
"learning_rate": 4.993523916570334e-06,
"loss": 0.45167410373687744,
"step": 1010
},
{
"epoch": 2.2439024390243905,
"grad_norm": 0.6709751486778259,
"learning_rate": 4.977126496514902e-06,
"loss": 0.7127041816711426,
"step": 1012
},
{
"epoch": 2.248337028824834,
"grad_norm": 1.1186648607254028,
"learning_rate": 4.960736111087827e-06,
"loss": 0.6675184369087219,
"step": 1014
},
{
"epoch": 2.2527716186252773,
"grad_norm": 0.6485116481781006,
"learning_rate": 4.9443529808018545e-06,
"loss": 0.7672496438026428,
"step": 1016
},
{
"epoch": 2.2572062084257207,
"grad_norm": 0.9844960570335388,
"learning_rate": 4.927977326072115e-06,
"loss": 0.35495874285697937,
"step": 1018
},
{
"epoch": 2.261640798226164,
"grad_norm": 0.7903894186019897,
"learning_rate": 4.911609367213168e-06,
"loss": 0.7469250559806824,
"step": 1020
},
{
"epoch": 2.2660753880266076,
"grad_norm": 0.3099033832550049,
"learning_rate": 4.895249324436035e-06,
"loss": 0.40709808468818665,
"step": 1022
},
{
"epoch": 2.270509977827051,
"grad_norm": 1.1226407289505005,
"learning_rate": 4.8788974178452316e-06,
"loss": 0.9115332365036011,
"step": 1024
},
{
"epoch": 2.2749445676274944,
"grad_norm": 0.6671955585479736,
"learning_rate": 4.86255386743582e-06,
"loss": 0.5178396105766296,
"step": 1026
},
{
"epoch": 2.279379157427938,
"grad_norm": 0.5912867188453674,
"learning_rate": 4.846218893090426e-06,
"loss": 0.7622277736663818,
"step": 1028
},
{
"epoch": 2.2838137472283813,
"grad_norm": 0.6438608765602112,
"learning_rate": 4.829892714576307e-06,
"loss": 0.7653957009315491,
"step": 1030
},
{
"epoch": 2.2882483370288247,
"grad_norm": 0.4936668574810028,
"learning_rate": 4.813575551542381e-06,
"loss": 0.5053679347038269,
"step": 1032
},
{
"epoch": 2.292682926829268,
"grad_norm": 0.8145703673362732,
"learning_rate": 4.7972676235162714e-06,
"loss": 0.4616087079048157,
"step": 1034
},
{
"epoch": 2.2971175166297115,
"grad_norm": 0.9270339608192444,
"learning_rate": 4.780969149901354e-06,
"loss": 0.6849997043609619,
"step": 1036
},
{
"epoch": 2.3015521064301554,
"grad_norm": 0.9029300808906555,
"learning_rate": 4.764680349973812e-06,
"loss": 0.6518898010253906,
"step": 1038
},
{
"epoch": 2.305986696230599,
"grad_norm": 1.1026612520217896,
"learning_rate": 4.748401442879674e-06,
"loss": 0.575007438659668,
"step": 1040
},
{
"epoch": 2.3104212860310422,
"grad_norm": 1.147935152053833,
"learning_rate": 4.732132647631881e-06,
"loss": 0.7933302521705627,
"step": 1042
},
{
"epoch": 2.3148558758314857,
"grad_norm": 0.4572734236717224,
"learning_rate": 4.715874183107324e-06,
"loss": 0.5096555352210999,
"step": 1044
},
{
"epoch": 2.319290465631929,
"grad_norm": 0.623441219329834,
"learning_rate": 4.699626268043911e-06,
"loss": 0.7747717499732971,
"step": 1046
},
{
"epoch": 2.3237250554323725,
"grad_norm": 0.5140348076820374,
"learning_rate": 4.683389121037618e-06,
"loss": 0.661880612373352,
"step": 1048
},
{
"epoch": 2.328159645232816,
"grad_norm": 0.5512512922286987,
"learning_rate": 4.667162960539552e-06,
"loss": 0.7839210033416748,
"step": 1050
},
{
"epoch": 2.3325942350332594,
"grad_norm": 1.5569580793380737,
"learning_rate": 4.650948004853006e-06,
"loss": 0.5787954330444336,
"step": 1052
},
{
"epoch": 2.337028824833703,
"grad_norm": 0.5759721398353577,
"learning_rate": 4.634744472130529e-06,
"loss": 0.5592602491378784,
"step": 1054
},
{
"epoch": 2.341463414634146,
"grad_norm": 0.5359609127044678,
"learning_rate": 4.618552580370988e-06,
"loss": 0.5897871851921082,
"step": 1056
},
{
"epoch": 2.3458980044345896,
"grad_norm": 0.3575122356414795,
"learning_rate": 4.6023725474166324e-06,
"loss": 0.47162723541259766,
"step": 1058
},
{
"epoch": 2.3503325942350335,
"grad_norm": 0.6168625950813293,
"learning_rate": 4.586204590950169e-06,
"loss": 0.8093945384025574,
"step": 1060
},
{
"epoch": 2.354767184035477,
"grad_norm": 0.6786574721336365,
"learning_rate": 4.570048928491824e-06,
"loss": 0.3322570323944092,
"step": 1062
},
{
"epoch": 2.3592017738359203,
"grad_norm": 0.3918606638908386,
"learning_rate": 4.5539057773964316e-06,
"loss": 0.4114290177822113,
"step": 1064
},
{
"epoch": 2.3636363636363638,
"grad_norm": 1.4159204959869385,
"learning_rate": 4.537775354850496e-06,
"loss": 0.6567169427871704,
"step": 1066
},
{
"epoch": 2.368070953436807,
"grad_norm": 0.5967050194740295,
"learning_rate": 4.5216578778692725e-06,
"loss": 0.734316349029541,
"step": 1068
},
{
"epoch": 2.3725055432372506,
"grad_norm": 0.5874074697494507,
"learning_rate": 4.5055535632938526e-06,
"loss": 0.7882179021835327,
"step": 1070
},
{
"epoch": 2.376940133037694,
"grad_norm": 1.1057568788528442,
"learning_rate": 4.489462627788242e-06,
"loss": 0.6857914924621582,
"step": 1072
},
{
"epoch": 2.3813747228381374,
"grad_norm": 0.6758074164390564,
"learning_rate": 4.473385287836448e-06,
"loss": 0.04006706923246384,
"step": 1074
},
{
"epoch": 2.385809312638581,
"grad_norm": 0.6099052429199219,
"learning_rate": 4.457321759739567e-06,
"loss": 0.43440449237823486,
"step": 1076
},
{
"epoch": 2.3902439024390243,
"grad_norm": 0.838287353515625,
"learning_rate": 4.4412722596128686e-06,
"loss": 0.19644105434417725,
"step": 1078
},
{
"epoch": 2.3946784922394677,
"grad_norm": 1.0746098756790161,
"learning_rate": 4.425237003382903e-06,
"loss": 0.6213241815567017,
"step": 1080
},
{
"epoch": 2.399113082039911,
"grad_norm": 0.2223556786775589,
"learning_rate": 4.409216206784577e-06,
"loss": 0.43604913353919983,
"step": 1082
},
{
"epoch": 2.4035476718403546,
"grad_norm": 1.25760018825531,
"learning_rate": 4.393210085358265e-06,
"loss": 0.5678226947784424,
"step": 1084
},
{
"epoch": 2.4079822616407984,
"grad_norm": 0.4775606393814087,
"learning_rate": 4.3772188544469016e-06,
"loss": 0.6794226765632629,
"step": 1086
},
{
"epoch": 2.412416851441242,
"grad_norm": 0.33771422505378723,
"learning_rate": 4.3612427291930915e-06,
"loss": 0.1676941215991974,
"step": 1088
},
{
"epoch": 2.4168514412416853,
"grad_norm": 0.485148549079895,
"learning_rate": 4.345281924536208e-06,
"loss": 0.7475972175598145,
"step": 1090
},
{
"epoch": 2.4212860310421287,
"grad_norm": 1.9281665086746216,
"learning_rate": 4.329336655209505e-06,
"loss": 0.5811200141906738,
"step": 1092
},
{
"epoch": 2.425720620842572,
"grad_norm": 0.96091628074646,
"learning_rate": 4.31340713573723e-06,
"loss": 0.3428981304168701,
"step": 1094
},
{
"epoch": 2.4301552106430155,
"grad_norm": 1.0888780355453491,
"learning_rate": 4.297493580431732e-06,
"loss": 0.4109230041503906,
"step": 1096
},
{
"epoch": 2.434589800443459,
"grad_norm": 0.2943512201309204,
"learning_rate": 4.281596203390582e-06,
"loss": 0.3823660910129547,
"step": 1098
},
{
"epoch": 2.4390243902439024,
"grad_norm": 1.013668179512024,
"learning_rate": 4.265715218493695e-06,
"loss": 0.6369096040725708,
"step": 1100
},
{
"epoch": 2.443458980044346,
"grad_norm": 0.41791486740112305,
"learning_rate": 4.249850839400446e-06,
"loss": 0.8213289380073547,
"step": 1102
},
{
"epoch": 2.4478935698447892,
"grad_norm": 1.5500551462173462,
"learning_rate": 4.2340032795468e-06,
"loss": 0.38706445693969727,
"step": 1104
},
{
"epoch": 2.4523281596452327,
"grad_norm": 1.2918660640716553,
"learning_rate": 4.218172752142442e-06,
"loss": 0.8401514291763306,
"step": 1106
},
{
"epoch": 2.4567627494456765,
"grad_norm": 1.1685500144958496,
"learning_rate": 4.202359470167903e-06,
"loss": 0.6909348964691162,
"step": 1108
},
{
"epoch": 2.4611973392461195,
"grad_norm": 0.5699374079704285,
"learning_rate": 4.186563646371696e-06,
"loss": 0.7599969506263733,
"step": 1110
},
{
"epoch": 2.4656319290465634,
"grad_norm": 0.4491530954837799,
"learning_rate": 4.170785493267463e-06,
"loss": 0.5214658975601196,
"step": 1112
},
{
"epoch": 2.470066518847007,
"grad_norm": 0.5719757676124573,
"learning_rate": 4.155025223131102e-06,
"loss": 0.7296863794326782,
"step": 1114
},
{
"epoch": 2.47450110864745,
"grad_norm": 1.554071307182312,
"learning_rate": 4.139283047997919e-06,
"loss": 0.2472933977842331,
"step": 1116
},
{
"epoch": 2.4789356984478936,
"grad_norm": 0.4289678931236267,
"learning_rate": 4.123559179659771e-06,
"loss": 0.7022151350975037,
"step": 1118
},
{
"epoch": 2.483370288248337,
"grad_norm": 0.6308154463768005,
"learning_rate": 4.107853829662224e-06,
"loss": 0.6172606945037842,
"step": 1120
},
{
"epoch": 2.4878048780487805,
"grad_norm": 0.5270412564277649,
"learning_rate": 4.0921672093017e-06,
"loss": 0.7171418070793152,
"step": 1122
},
{
"epoch": 2.492239467849224,
"grad_norm": 1.6710797548294067,
"learning_rate": 4.076499529622636e-06,
"loss": 0.7357991933822632,
"step": 1124
},
{
"epoch": 2.4966740576496673,
"grad_norm": 0.7722313404083252,
"learning_rate": 4.0608510014146455e-06,
"loss": 0.7875959277153015,
"step": 1126
},
{
"epoch": 2.5011086474501107,
"grad_norm": 0.580020010471344,
"learning_rate": 4.045221835209684e-06,
"loss": 0.4804987907409668,
"step": 1128
},
{
"epoch": 2.505543237250554,
"grad_norm": 1.0426214933395386,
"learning_rate": 4.02961224127921e-06,
"loss": 0.517763078212738,
"step": 1130
},
{
"epoch": 2.5099778270509976,
"grad_norm": 10.173712730407715,
"learning_rate": 4.014022429631368e-06,
"loss": 0.6159908771514893,
"step": 1132
},
{
"epoch": 2.5144124168514415,
"grad_norm": 0.5241939425468445,
"learning_rate": 3.998452610008147e-06,
"loss": 0.4526905417442322,
"step": 1134
},
{
"epoch": 2.5188470066518844,
"grad_norm": 0.6242166757583618,
"learning_rate": 3.982902991882578e-06,
"loss": 0.8381586670875549,
"step": 1136
},
{
"epoch": 2.5232815964523283,
"grad_norm": 0.5091142058372498,
"learning_rate": 3.967373784455896e-06,
"loss": 0.5896800756454468,
"step": 1138
},
{
"epoch": 2.5277161862527717,
"grad_norm": 0.6498262286186218,
"learning_rate": 3.951865196654738e-06,
"loss": 0.7107540965080261,
"step": 1140
},
{
"epoch": 2.532150776053215,
"grad_norm": 1.1750524044036865,
"learning_rate": 3.936377437128329e-06,
"loss": 0.35893017053604126,
"step": 1142
},
{
"epoch": 2.5365853658536586,
"grad_norm": 0.3123358190059662,
"learning_rate": 3.920910714245679e-06,
"loss": 0.4256970286369324,
"step": 1144
},
{
"epoch": 2.541019955654102,
"grad_norm": 0.533246636390686,
"learning_rate": 3.905465236092771e-06,
"loss": 0.7800043225288391,
"step": 1146
},
{
"epoch": 2.5454545454545454,
"grad_norm": 0.6559436321258545,
"learning_rate": 3.890041210469765e-06,
"loss": 0.7252997756004333,
"step": 1148
},
{
"epoch": 2.549889135254989,
"grad_norm": 0.8817735314369202,
"learning_rate": 3.8746388448882055e-06,
"loss": 0.6513781547546387,
"step": 1150
},
{
"epoch": 2.5543237250554323,
"grad_norm": 0.6646784543991089,
"learning_rate": 3.859258346568228e-06,
"loss": 0.3836185336112976,
"step": 1152
},
{
"epoch": 2.5587583148558757,
"grad_norm": 0.2349381297826767,
"learning_rate": 3.843899922435767e-06,
"loss": 0.24700769782066345,
"step": 1154
},
{
"epoch": 2.5631929046563195,
"grad_norm": 0.49869391322135925,
"learning_rate": 3.8285637791197815e-06,
"loss": 0.38181808590888977,
"step": 1156
},
{
"epoch": 2.5676274944567625,
"grad_norm": 4.80774450302124,
"learning_rate": 3.8132501229494635e-06,
"loss": 0.33294159173965454,
"step": 1158
},
{
"epoch": 2.5720620842572064,
"grad_norm": 1.0527015924453735,
"learning_rate": 3.7979591599514696e-06,
"loss": 0.44462141394615173,
"step": 1160
},
{
"epoch": 2.57649667405765,
"grad_norm": 0.4929417669773102,
"learning_rate": 3.782691095847151e-06,
"loss": 0.7220413088798523,
"step": 1162
},
{
"epoch": 2.5809312638580932,
"grad_norm": 0.9527971744537354,
"learning_rate": 3.767446136049775e-06,
"loss": 0.8302287459373474,
"step": 1164
},
{
"epoch": 2.5853658536585367,
"grad_norm": 0.634364664554596,
"learning_rate": 3.752224485661775e-06,
"loss": 0.5409159660339355,
"step": 1166
},
{
"epoch": 2.58980044345898,
"grad_norm": 0.5789323449134827,
"learning_rate": 3.7370263494719805e-06,
"loss": 0.6534980535507202,
"step": 1168
},
{
"epoch": 2.5942350332594235,
"grad_norm": 0.8696730732917786,
"learning_rate": 3.721851931952869e-06,
"loss": 0.7098603248596191,
"step": 1170
},
{
"epoch": 2.598669623059867,
"grad_norm": 0.7621782422065735,
"learning_rate": 3.706701437257808e-06,
"loss": 0.25166431069374084,
"step": 1172
},
{
"epoch": 2.6031042128603104,
"grad_norm": 0.5634275078773499,
"learning_rate": 3.691575069218314e-06,
"loss": 0.5319882035255432,
"step": 1174
},
{
"epoch": 2.6075388026607538,
"grad_norm": 0.6240749955177307,
"learning_rate": 3.676473031341313e-06,
"loss": 0.5341241359710693,
"step": 1176
},
{
"epoch": 2.611973392461197,
"grad_norm": 0.5434238314628601,
"learning_rate": 3.661395526806395e-06,
"loss": 0.43045657873153687,
"step": 1178
},
{
"epoch": 2.6164079822616406,
"grad_norm": 1.4436023235321045,
"learning_rate": 3.6463427584630806e-06,
"loss": 0.5974833965301514,
"step": 1180
},
{
"epoch": 2.6208425720620845,
"grad_norm": 0.4753042757511139,
"learning_rate": 3.631314928828099e-06,
"loss": 0.6559595465660095,
"step": 1182
},
{
"epoch": 2.6252771618625275,
"grad_norm": 0.5069239139556885,
"learning_rate": 3.616312240082659e-06,
"loss": 0.7961637377738953,
"step": 1184
},
{
"epoch": 2.6297117516629713,
"grad_norm": 0.5121076703071594,
"learning_rate": 3.601334894069728e-06,
"loss": 0.7357903718948364,
"step": 1186
},
{
"epoch": 2.6341463414634148,
"grad_norm": 1.606050968170166,
"learning_rate": 3.5863830922913147e-06,
"loss": 0.7505167722702026,
"step": 1188
},
{
"epoch": 2.638580931263858,
"grad_norm": 0.6388944387435913,
"learning_rate": 3.5714570359057676e-06,
"loss": 0.43692290782928467,
"step": 1190
},
{
"epoch": 2.6430155210643016,
"grad_norm": 1.7327947616577148,
"learning_rate": 3.556556925725061e-06,
"loss": 0.688886821269989,
"step": 1192
},
{
"epoch": 2.647450110864745,
"grad_norm": 1.5744296312332153,
"learning_rate": 3.5416829622120875e-06,
"loss": 0.4643045663833618,
"step": 1194
},
{
"epoch": 2.6518847006651884,
"grad_norm": 0.5231549739837646,
"learning_rate": 3.526835345477978e-06,
"loss": 0.6326463222503662,
"step": 1196
},
{
"epoch": 2.656319290465632,
"grad_norm": 0.3871590495109558,
"learning_rate": 3.5120142752793907e-06,
"loss": 0.2630988359451294,
"step": 1198
},
{
"epoch": 2.6607538802660753,
"grad_norm": 0.6133447885513306,
"learning_rate": 3.4972199510158393e-06,
"loss": 0.8605660796165466,
"step": 1200
},
{
"epoch": 2.6651884700665187,
"grad_norm": 0.49589312076568604,
"learning_rate": 3.4824525717269975e-06,
"loss": 0.7755805850028992,
"step": 1202
},
{
"epoch": 2.6696230598669626,
"grad_norm": 2.0768580436706543,
"learning_rate": 3.4677123360900342e-06,
"loss": 0.3658536672592163,
"step": 1204
},
{
"epoch": 2.6740576496674056,
"grad_norm": 0.6321128606796265,
"learning_rate": 3.4529994424169233e-06,
"loss": 0.5949617624282837,
"step": 1206
},
{
"epoch": 2.6784922394678494,
"grad_norm": 0.720273494720459,
"learning_rate": 3.4383140886517953e-06,
"loss": 0.6172499060630798,
"step": 1208
},
{
"epoch": 2.682926829268293,
"grad_norm": 3.1463682651519775,
"learning_rate": 3.423656472368262e-06,
"loss": 0.38858482241630554,
"step": 1210
},
{
"epoch": 2.6873614190687363,
"grad_norm": 1.7371691465377808,
"learning_rate": 3.409026790766756e-06,
"loss": 0.25850269198417664,
"step": 1212
},
{
"epoch": 2.6917960088691797,
"grad_norm": 0.6183602213859558,
"learning_rate": 3.394425240671891e-06,
"loss": 0.4422842860221863,
"step": 1214
},
{
"epoch": 2.696230598669623,
"grad_norm": 1.6302183866500854,
"learning_rate": 3.379852018529799e-06,
"loss": 0.5197972059249878,
"step": 1216
},
{
"epoch": 2.7006651884700665,
"grad_norm": 1.3484753370285034,
"learning_rate": 3.3653073204054942e-06,
"loss": 0.4698919355869293,
"step": 1218
},
{
"epoch": 2.70509977827051,
"grad_norm": 0.6926991939544678,
"learning_rate": 3.3507913419802403e-06,
"loss": 0.7520615458488464,
"step": 1220
},
{
"epoch": 2.7095343680709534,
"grad_norm": 0.7605311870574951,
"learning_rate": 3.336304278548903e-06,
"loss": 0.5865758061408997,
"step": 1222
},
{
"epoch": 2.713968957871397,
"grad_norm": 1.110818862915039,
"learning_rate": 3.321846325017342e-06,
"loss": 0.759763240814209,
"step": 1224
},
{
"epoch": 2.7184035476718402,
"grad_norm": 1.391050934791565,
"learning_rate": 3.3074176758997744e-06,
"loss": 0.34599849581718445,
"step": 1226
},
{
"epoch": 2.7228381374722836,
"grad_norm": 0.7273187041282654,
"learning_rate": 3.2930185253161574e-06,
"loss": 0.8604745864868164,
"step": 1228
},
{
"epoch": 2.7272727272727275,
"grad_norm": 0.6875346899032593,
"learning_rate": 3.2786490669895883e-06,
"loss": 0.575916588306427,
"step": 1230
},
{
"epoch": 2.7317073170731705,
"grad_norm": 0.5254970788955688,
"learning_rate": 3.2643094942436865e-06,
"loss": 0.654707670211792,
"step": 1232
},
{
"epoch": 2.7361419068736144,
"grad_norm": 1.1695361137390137,
"learning_rate": 3.2500000000000015e-06,
"loss": 0.33159804344177246,
"step": 1234
},
{
"epoch": 2.740576496674058,
"grad_norm": 0.5698898434638977,
"learning_rate": 3.2357207767754063e-06,
"loss": 0.7334117293357849,
"step": 1236
},
{
"epoch": 2.745011086474501,
"grad_norm": 0.5489004254341125,
"learning_rate": 3.221472016679521e-06,
"loss": 0.4428709149360657,
"step": 1238
},
{
"epoch": 2.7494456762749446,
"grad_norm": 0.5650718212127686,
"learning_rate": 3.2072539114121188e-06,
"loss": 0.44653040170669556,
"step": 1240
},
{
"epoch": 2.753880266075388,
"grad_norm": 0.5928937792778015,
"learning_rate": 3.193066652260547e-06,
"loss": 0.782917320728302,
"step": 1242
},
{
"epoch": 2.7583148558758315,
"grad_norm": 0.5975604057312012,
"learning_rate": 3.1789104300971603e-06,
"loss": 0.7677651047706604,
"step": 1244
},
{
"epoch": 2.762749445676275,
"grad_norm": 1.0852833986282349,
"learning_rate": 3.164785435376745e-06,
"loss": 0.29107940196990967,
"step": 1246
},
{
"epoch": 2.7671840354767183,
"grad_norm": 1.17136549949646,
"learning_rate": 3.1506918581339583e-06,
"loss": 0.4287865459918976,
"step": 1248
},
{
"epoch": 2.7716186252771617,
"grad_norm": 0.9015536308288574,
"learning_rate": 3.136629887980781e-06,
"loss": 0.49741506576538086,
"step": 1250
},
{
"epoch": 2.776053215077605,
"grad_norm": 0.573412299156189,
"learning_rate": 3.122599714103949e-06,
"loss": 0.8777692317962646,
"step": 1252
},
{
"epoch": 2.7804878048780486,
"grad_norm": 0.5751115679740906,
"learning_rate": 3.1086015252624257e-06,
"loss": 0.8302069306373596,
"step": 1254
},
{
"epoch": 2.7849223946784925,
"grad_norm": 0.4423518180847168,
"learning_rate": 3.0946355097848535e-06,
"loss": 0.7646183371543884,
"step": 1256
},
{
"epoch": 2.7893569844789354,
"grad_norm": 0.403327614068985,
"learning_rate": 3.0807018555670153e-06,
"loss": 0.15144073963165283,
"step": 1258
},
{
"epoch": 2.7937915742793793,
"grad_norm": 0.5304398536682129,
"learning_rate": 3.0668007500693216e-06,
"loss": 0.711767852306366,
"step": 1260
},
{
"epoch": 2.7982261640798227,
"grad_norm": 0.8088931441307068,
"learning_rate": 3.0529323803142697e-06,
"loss": 0.41084200143814087,
"step": 1262
},
{
"epoch": 2.802660753880266,
"grad_norm": 0.5133860111236572,
"learning_rate": 3.0390969328839464e-06,
"loss": 0.42946991324424744,
"step": 1264
},
{
"epoch": 2.8070953436807096,
"grad_norm": 0.5490220189094543,
"learning_rate": 3.0252945939175004e-06,
"loss": 0.6761541366577148,
"step": 1266
},
{
"epoch": 2.811529933481153,
"grad_norm": 1.2554481029510498,
"learning_rate": 3.0115255491086537e-06,
"loss": 0.6815743446350098,
"step": 1268
},
{
"epoch": 2.8159645232815964,
"grad_norm": 0.7049453258514404,
"learning_rate": 2.9977899837031895e-06,
"loss": 0.7387552261352539,
"step": 1270
},
{
"epoch": 2.82039911308204,
"grad_norm": 1.1351791620254517,
"learning_rate": 2.984088082496469e-06,
"loss": 0.21328306198120117,
"step": 1272
},
{
"epoch": 2.8248337028824833,
"grad_norm": 0.5017495155334473,
"learning_rate": 2.970420029830946e-06,
"loss": 0.4749630391597748,
"step": 1274
},
{
"epoch": 2.8292682926829267,
"grad_norm": 0.6224768161773682,
"learning_rate": 2.9567860095936775e-06,
"loss": 0.7620760798454285,
"step": 1276
},
{
"epoch": 2.8337028824833705,
"grad_norm": 0.6925188899040222,
"learning_rate": 2.9431862052138545e-06,
"loss": 0.7254369258880615,
"step": 1278
},
{
"epoch": 2.8381374722838135,
"grad_norm": 0.7251638174057007,
"learning_rate": 2.929620799660343e-06,
"loss": 0.3205258846282959,
"step": 1280
},
{
"epoch": 2.8425720620842574,
"grad_norm": 6.005808353424072,
"learning_rate": 2.916089975439207e-06,
"loss": 0.44966956973075867,
"step": 1282
},
{
"epoch": 2.847006651884701,
"grad_norm": 0.6776400208473206,
"learning_rate": 2.9025939145912655e-06,
"loss": 0.5670607089996338,
"step": 1284
},
{
"epoch": 2.8514412416851442,
"grad_norm": 0.5101284980773926,
"learning_rate": 2.8891327986896345e-06,
"loss": 0.7107180953025818,
"step": 1286
},
{
"epoch": 2.8558758314855877,
"grad_norm": 3.805804967880249,
"learning_rate": 2.875706808837292e-06,
"loss": 0.1649579256772995,
"step": 1288
},
{
"epoch": 2.860310421286031,
"grad_norm": 0.5193606615066528,
"learning_rate": 2.862316125664636e-06,
"loss": 0.7318248152732849,
"step": 1290
},
{
"epoch": 2.8647450110864745,
"grad_norm": 1.6448044776916504,
"learning_rate": 2.848960929327053e-06,
"loss": 0.5531004071235657,
"step": 1292
},
{
"epoch": 2.869179600886918,
"grad_norm": 0.5975578427314758,
"learning_rate": 2.8356413995025044e-06,
"loss": 0.7080908417701721,
"step": 1294
},
{
"epoch": 2.8736141906873613,
"grad_norm": 0.5933077335357666,
"learning_rate": 2.8223577153890934e-06,
"loss": 0.7111974358558655,
"step": 1296
},
{
"epoch": 2.8780487804878048,
"grad_norm": 1.380179762840271,
"learning_rate": 2.8091100557026702e-06,
"loss": 0.568051815032959,
"step": 1298
},
{
"epoch": 2.882483370288248,
"grad_norm": 0.5458784699440002,
"learning_rate": 2.795898598674415e-06,
"loss": 0.7207503318786621,
"step": 1300
},
{
"epoch": 2.8869179600886916,
"grad_norm": 1.1317119598388672,
"learning_rate": 2.782723522048444e-06,
"loss": 0.2552308738231659,
"step": 1302
},
{
"epoch": 2.8913525498891355,
"grad_norm": 0.7093331217765808,
"learning_rate": 2.7695850030794293e-06,
"loss": 0.6807090044021606,
"step": 1304
},
{
"epoch": 2.8957871396895785,
"grad_norm": 0.9031757116317749,
"learning_rate": 2.7564832185301915e-06,
"loss": 0.6350713968276978,
"step": 1306
},
{
"epoch": 2.9002217294900223,
"grad_norm": 0.3347035348415375,
"learning_rate": 2.7434183446693397e-06,
"loss": 0.2564702033996582,
"step": 1308
},
{
"epoch": 2.9046563192904657,
"grad_norm": 0.7028330564498901,
"learning_rate": 2.730390557268897e-06,
"loss": 0.37558212876319885,
"step": 1310
},
{
"epoch": 2.909090909090909,
"grad_norm": 1.2789974212646484,
"learning_rate": 2.7174000316019277e-06,
"loss": 0.7542843222618103,
"step": 1312
},
{
"epoch": 2.9135254988913526,
"grad_norm": 0.5418480634689331,
"learning_rate": 2.704446942440191e-06,
"loss": 0.7207885980606079,
"step": 1314
},
{
"epoch": 2.917960088691796,
"grad_norm": 1.9766089916229248,
"learning_rate": 2.6915314640517755e-06,
"loss": 0.5357068181037903,
"step": 1316
},
{
"epoch": 2.9223946784922394,
"grad_norm": 0.6477085947990417,
"learning_rate": 2.6786537701987703e-06,
"loss": 0.813840925693512,
"step": 1318
},
{
"epoch": 2.926829268292683,
"grad_norm": 0.5277987122535706,
"learning_rate": 2.665814034134916e-06,
"loss": 0.4807235598564148,
"step": 1320
},
{
"epoch": 2.9312638580931263,
"grad_norm": 0.4396364390850067,
"learning_rate": 2.6530124286032755e-06,
"loss": 0.7007963061332703,
"step": 1322
},
{
"epoch": 2.9356984478935697,
"grad_norm": 0.8923091292381287,
"learning_rate": 2.640249125833915e-06,
"loss": 0.7382723689079285,
"step": 1324
},
{
"epoch": 2.9401330376940136,
"grad_norm": 0.4707968533039093,
"learning_rate": 2.6275242975415804e-06,
"loss": 0.638217568397522,
"step": 1326
},
{
"epoch": 2.9445676274944566,
"grad_norm": 0.5562184453010559,
"learning_rate": 2.614838114923394e-06,
"loss": 0.8006188869476318,
"step": 1328
},
{
"epoch": 2.9490022172949004,
"grad_norm": 0.6460016369819641,
"learning_rate": 2.6021907486565447e-06,
"loss": 0.7337307333946228,
"step": 1330
},
{
"epoch": 2.953436807095344,
"grad_norm": 0.600333034992218,
"learning_rate": 2.589582368895992e-06,
"loss": 0.7522150278091431,
"step": 1332
},
{
"epoch": 2.9578713968957873,
"grad_norm": 0.47927215695381165,
"learning_rate": 2.577013145272185e-06,
"loss": 0.38160401582717896,
"step": 1334
},
{
"epoch": 2.9623059866962307,
"grad_norm": 0.6014701128005981,
"learning_rate": 2.564483246888772e-06,
"loss": 0.7021808624267578,
"step": 1336
},
{
"epoch": 2.966740576496674,
"grad_norm": 1.0804996490478516,
"learning_rate": 2.5519928423203266e-06,
"loss": 0.7971820831298828,
"step": 1338
},
{
"epoch": 2.9711751662971175,
"grad_norm": 0.5336750745773315,
"learning_rate": 2.539542099610084e-06,
"loss": 0.5081973075866699,
"step": 1340
},
{
"epoch": 2.975609756097561,
"grad_norm": 2.955822467803955,
"learning_rate": 2.5271311862676727e-06,
"loss": 0.4675310254096985,
"step": 1342
},
{
"epoch": 2.9800443458980044,
"grad_norm": 0.7282045483589172,
"learning_rate": 2.514760269266871e-06,
"loss": 0.16825823485851288,
"step": 1344
},
{
"epoch": 2.984478935698448,
"grad_norm": 0.5478266477584839,
"learning_rate": 2.50242951504335e-06,
"loss": 0.4867478311061859,
"step": 1346
},
{
"epoch": 2.988913525498891,
"grad_norm": 0.22204731404781342,
"learning_rate": 2.490139089492443e-06,
"loss": 0.4467255175113678,
"step": 1348
},
{
"epoch": 2.9933481152993346,
"grad_norm": 4.795533657073975,
"learning_rate": 2.4778891579669067e-06,
"loss": 0.3938838839530945,
"step": 1350
},
{
"epoch": 2.9977827050997785,
"grad_norm": 0.496768981218338,
"learning_rate": 2.4656798852747023e-06,
"loss": 0.5702564120292664,
"step": 1352
},
{
"epoch": 3.002217294900222,
"grad_norm": 0.7324950098991394,
"learning_rate": 2.453511435676777e-06,
"loss": 0.5976329445838928,
"step": 1354
},
{
"epoch": 3.0066518847006654,
"grad_norm": 0.5204404592514038,
"learning_rate": 2.441383972884848e-06,
"loss": 0.2519686818122864,
"step": 1356
},
{
"epoch": 3.011086474501109,
"grad_norm": 0.6069619655609131,
"learning_rate": 2.4292976600592095e-06,
"loss": 0.3836074471473694,
"step": 1358
},
{
"epoch": 3.015521064301552,
"grad_norm": 0.6137452721595764,
"learning_rate": 2.4172526598065304e-06,
"loss": 0.5434846878051758,
"step": 1360
},
{
"epoch": 3.0199556541019956,
"grad_norm": 1.0837262868881226,
"learning_rate": 2.4052491341776686e-06,
"loss": 0.4407395124435425,
"step": 1362
},
{
"epoch": 3.024390243902439,
"grad_norm": 0.8856160640716553,
"learning_rate": 2.393287244665494e-06,
"loss": 0.5295175313949585,
"step": 1364
},
{
"epoch": 3.0288248337028825,
"grad_norm": 0.12850333750247955,
"learning_rate": 2.3813671522027094e-06,
"loss": 0.13662414252758026,
"step": 1366
},
{
"epoch": 3.033259423503326,
"grad_norm": 0.762299120426178,
"learning_rate": 2.369489017159692e-06,
"loss": 0.2511744797229767,
"step": 1368
},
{
"epoch": 3.0376940133037693,
"grad_norm": 0.7555387616157532,
"learning_rate": 2.357652999342334e-06,
"loss": 0.28912025690078735,
"step": 1370
},
{
"epoch": 3.0421286031042127,
"grad_norm": 0.7977571487426758,
"learning_rate": 2.345859257989886e-06,
"loss": 0.4399268925189972,
"step": 1372
},
{
"epoch": 3.046563192904656,
"grad_norm": 0.47056496143341064,
"learning_rate": 2.334107951772826e-06,
"loss": 0.13275624811649323,
"step": 1374
},
{
"epoch": 3.0509977827050996,
"grad_norm": 1.6538853645324707,
"learning_rate": 2.3223992387907137e-06,
"loss": 0.2295996993780136,
"step": 1376
},
{
"epoch": 3.0554323725055434,
"grad_norm": 0.35899239778518677,
"learning_rate": 2.3107332765700733e-06,
"loss": 0.1384759098291397,
"step": 1378
},
{
"epoch": 3.059866962305987,
"grad_norm": 3.042902708053589,
"learning_rate": 2.2991102220622647e-06,
"loss": 0.21765412390232086,
"step": 1380
},
{
"epoch": 3.0643015521064303,
"grad_norm": 1.761853575706482,
"learning_rate": 2.2875302316413807e-06,
"loss": 0.10996652394533157,
"step": 1382
},
{
"epoch": 3.0687361419068737,
"grad_norm": 1.1698344945907593,
"learning_rate": 2.275993461102138e-06,
"loss": 0.19330096244812012,
"step": 1384
},
{
"epoch": 3.073170731707317,
"grad_norm": 0.32080015540122986,
"learning_rate": 2.2645000656577793e-06,
"loss": 0.14769285917282104,
"step": 1386
},
{
"epoch": 3.0776053215077606,
"grad_norm": 0.19042012095451355,
"learning_rate": 2.2530501999379932e-06,
"loss": 0.1551397293806076,
"step": 1388
},
{
"epoch": 3.082039911308204,
"grad_norm": 0.5313494801521301,
"learning_rate": 2.2416440179868236e-06,
"loss": 0.3368770480155945,
"step": 1390
},
{
"epoch": 3.0864745011086474,
"grad_norm": 1.5543434619903564,
"learning_rate": 2.230281673260605e-06,
"loss": 0.09978589415550232,
"step": 1392
},
{
"epoch": 3.090909090909091,
"grad_norm": 2.5066184997558594,
"learning_rate": 2.218963318625895e-06,
"loss": 0.2049475759267807,
"step": 1394
},
{
"epoch": 3.0953436807095343,
"grad_norm": 0.8068859577178955,
"learning_rate": 2.2076891063574167e-06,
"loss": 0.4105300307273865,
"step": 1396
},
{
"epoch": 3.0997782705099777,
"grad_norm": 0.9558060765266418,
"learning_rate": 2.196459188136014e-06,
"loss": 0.40823429822921753,
"step": 1398
},
{
"epoch": 3.104212860310421,
"grad_norm": 0.2851542830467224,
"learning_rate": 2.1852737150466064e-06,
"loss": 0.31296348571777344,
"step": 1400
},
{
"epoch": 3.1086474501108645,
"grad_norm": 0.3126837909221649,
"learning_rate": 2.174132837576156e-06,
"loss": 0.0858599990606308,
"step": 1402
},
{
"epoch": 3.1130820399113084,
"grad_norm": 1.1437623500823975,
"learning_rate": 2.1630367056116496e-06,
"loss": 0.16902470588684082,
"step": 1404
},
{
"epoch": 3.117516629711752,
"grad_norm": 1.4211009740829468,
"learning_rate": 2.1519854684380724e-06,
"loss": 0.44321006536483765,
"step": 1406
},
{
"epoch": 3.1219512195121952,
"grad_norm": 0.5866262912750244,
"learning_rate": 2.1409792747364103e-06,
"loss": 0.4670795202255249,
"step": 1408
},
{
"epoch": 3.1263858093126387,
"grad_norm": 0.5906457901000977,
"learning_rate": 2.1300182725816378e-06,
"loss": 0.3769230246543884,
"step": 1410
},
{
"epoch": 3.130820399113082,
"grad_norm": 0.6624063849449158,
"learning_rate": 2.1191026094407386e-06,
"loss": 0.3948790431022644,
"step": 1412
},
{
"epoch": 3.1352549889135255,
"grad_norm": 0.45142582058906555,
"learning_rate": 2.1082324321707075e-06,
"loss": 0.26148101687431335,
"step": 1414
},
{
"epoch": 3.139689578713969,
"grad_norm": 0.08900812268257141,
"learning_rate": 2.0974078870165882e-06,
"loss": 0.0190610121935606,
"step": 1416
},
{
"epoch": 3.1441241685144123,
"grad_norm": 0.7073218822479248,
"learning_rate": 2.086629119609499e-06,
"loss": 0.3549107611179352,
"step": 1418
},
{
"epoch": 3.1485587583148558,
"grad_norm": 0.5838420391082764,
"learning_rate": 2.0758962749646716e-06,
"loss": 0.39513227343559265,
"step": 1420
},
{
"epoch": 3.152993348115299,
"grad_norm": 0.4758341312408447,
"learning_rate": 2.065209497479502e-06,
"loss": 0.3184177279472351,
"step": 1422
},
{
"epoch": 3.1574279379157426,
"grad_norm": 0.33181366324424744,
"learning_rate": 2.0545689309316138e-06,
"loss": 0.05200305953621864,
"step": 1424
},
{
"epoch": 3.1618625277161865,
"grad_norm": 1.9294317960739136,
"learning_rate": 2.043974718476911e-06,
"loss": 0.34890684485435486,
"step": 1426
},
{
"epoch": 3.16629711751663,
"grad_norm": 1.2482831478118896,
"learning_rate": 2.033427002647668e-06,
"loss": 0.2931206524372101,
"step": 1428
},
{
"epoch": 3.1707317073170733,
"grad_norm": 0.8384535312652588,
"learning_rate": 2.0229259253505946e-06,
"loss": 0.15022891759872437,
"step": 1430
},
{
"epoch": 3.1751662971175167,
"grad_norm": 0.5024890303611755,
"learning_rate": 2.012471627864943e-06,
"loss": 0.24945132434368134,
"step": 1432
},
{
"epoch": 3.17960088691796,
"grad_norm": 0.7453292608261108,
"learning_rate": 2.0020642508405984e-06,
"loss": 0.27870461344718933,
"step": 1434
},
{
"epoch": 3.1840354767184036,
"grad_norm": 0.5649028420448303,
"learning_rate": 1.9917039342961837e-06,
"loss": 0.04178987070918083,
"step": 1436
},
{
"epoch": 3.188470066518847,
"grad_norm": 0.5637784600257874,
"learning_rate": 1.9813908176171857e-06,
"loss": 0.2853415310382843,
"step": 1438
},
{
"epoch": 3.1929046563192904,
"grad_norm": 1.2113947868347168,
"learning_rate": 1.97112503955407e-06,
"loss": 0.3286460340023041,
"step": 1440
},
{
"epoch": 3.197339246119734,
"grad_norm": 0.7789623737335205,
"learning_rate": 1.9609067382204224e-06,
"loss": 0.41224563121795654,
"step": 1442
},
{
"epoch": 3.2017738359201773,
"grad_norm": 1.0445942878723145,
"learning_rate": 1.950736051091084e-06,
"loss": 0.5211740136146545,
"step": 1444
},
{
"epoch": 3.2062084257206207,
"grad_norm": 0.4848485589027405,
"learning_rate": 1.9406131150003036e-06,
"loss": 0.37894371151924133,
"step": 1446
},
{
"epoch": 3.210643015521064,
"grad_norm": 0.6881194710731506,
"learning_rate": 1.930538066139904e-06,
"loss": 0.1102185845375061,
"step": 1448
},
{
"epoch": 3.2150776053215075,
"grad_norm": 0.7072343230247498,
"learning_rate": 1.9205110400574368e-06,
"loss": 0.46290311217308044,
"step": 1450
},
{
"epoch": 3.2195121951219514,
"grad_norm": 0.7401267290115356,
"learning_rate": 1.910532171654367e-06,
"loss": 0.1061285138130188,
"step": 1452
},
{
"epoch": 3.223946784922395,
"grad_norm": 0.6795309782028198,
"learning_rate": 1.9006015951842587e-06,
"loss": 0.30657243728637695,
"step": 1454
},
{
"epoch": 3.2283813747228383,
"grad_norm": 1.8200849294662476,
"learning_rate": 1.8907194442509642e-06,
"loss": 0.29492244124412537,
"step": 1456
},
{
"epoch": 3.2328159645232817,
"grad_norm": 0.4630158841609955,
"learning_rate": 1.8808858518068312e-06,
"loss": 0.2674718499183655,
"step": 1458
},
{
"epoch": 3.237250554323725,
"grad_norm": 1.3433810472488403,
"learning_rate": 1.8711009501509087e-06,
"loss": 0.287025511264801,
"step": 1460
},
{
"epoch": 3.2416851441241685,
"grad_norm": 0.596284031867981,
"learning_rate": 1.8613648709271732e-06,
"loss": 0.30234360694885254,
"step": 1462
},
{
"epoch": 3.246119733924612,
"grad_norm": 1.2054084539413452,
"learning_rate": 1.8516777451227552e-06,
"loss": 0.26377224922180176,
"step": 1464
},
{
"epoch": 3.2505543237250554,
"grad_norm": 1.0872267484664917,
"learning_rate": 1.842039703066172e-06,
"loss": 0.44247758388519287,
"step": 1466
},
{
"epoch": 3.254988913525499,
"grad_norm": 0.12414921820163727,
"learning_rate": 1.8324508744255842e-06,
"loss": 0.028142806142568588,
"step": 1468
},
{
"epoch": 3.259423503325942,
"grad_norm": 1.1836111545562744,
"learning_rate": 1.8229113882070398e-06,
"loss": 0.258672297000885,
"step": 1470
},
{
"epoch": 3.2638580931263856,
"grad_norm": 0.5712287425994873,
"learning_rate": 1.8134213727527504e-06,
"loss": 0.3421834409236908,
"step": 1472
},
{
"epoch": 3.2682926829268295,
"grad_norm": 0.5858505964279175,
"learning_rate": 1.803980955739354e-06,
"loss": 0.08477696031332016,
"step": 1474
},
{
"epoch": 3.2727272727272725,
"grad_norm": 0.5068755149841309,
"learning_rate": 1.7945902641762027e-06,
"loss": 0.2908667027950287,
"step": 1476
},
{
"epoch": 3.2771618625277164,
"grad_norm": 0.4555286169052124,
"learning_rate": 1.785249424403654e-06,
"loss": 0.23708750307559967,
"step": 1478
},
{
"epoch": 3.2815964523281598,
"grad_norm": 0.4396187365055084,
"learning_rate": 1.7759585620913723e-06,
"loss": 0.2890844941139221,
"step": 1480
},
{
"epoch": 3.286031042128603,
"grad_norm": 0.7039048671722412,
"learning_rate": 1.7667178022366294e-06,
"loss": 0.42803099751472473,
"step": 1482
},
{
"epoch": 3.2904656319290466,
"grad_norm": 0.6853517293930054,
"learning_rate": 1.757527269162636e-06,
"loss": 0.31256699562072754,
"step": 1484
},
{
"epoch": 3.29490022172949,
"grad_norm": 0.5980814695358276,
"learning_rate": 1.7483870865168585e-06,
"loss": 0.23612843453884125,
"step": 1486
},
{
"epoch": 3.2993348115299335,
"grad_norm": 0.6018959879875183,
"learning_rate": 1.739297377269361e-06,
"loss": 0.1575717329978943,
"step": 1488
},
{
"epoch": 3.303769401330377,
"grad_norm": 0.9202043414115906,
"learning_rate": 1.730258263711149e-06,
"loss": 0.0800296813249588,
"step": 1490
},
{
"epoch": 3.3082039911308203,
"grad_norm": 0.804567277431488,
"learning_rate": 1.7212698674525246e-06,
"loss": 0.384365975856781,
"step": 1492
},
{
"epoch": 3.3126385809312637,
"grad_norm": 0.6774597764015198,
"learning_rate": 1.7123323094214485e-06,
"loss": 0.35183945298194885,
"step": 1494
},
{
"epoch": 3.317073170731707,
"grad_norm": 0.5413676500320435,
"learning_rate": 1.7034457098619176e-06,
"loss": 0.3031509518623352,
"step": 1496
},
{
"epoch": 3.3215077605321506,
"grad_norm": 0.6860851645469666,
"learning_rate": 1.6946101883323435e-06,
"loss": 0.4678665101528168,
"step": 1498
},
{
"epoch": 3.3259423503325944,
"grad_norm": 0.8729199767112732,
"learning_rate": 1.6858258637039421e-06,
"loss": 0.2541211247444153,
"step": 1500
},
{
"epoch": 3.330376940133038,
"grad_norm": 0.5793163776397705,
"learning_rate": 1.677092854159142e-06,
"loss": 0.409402996301651,
"step": 1502
},
{
"epoch": 3.3348115299334813,
"grad_norm": 0.6284571886062622,
"learning_rate": 1.6684112771899858e-06,
"loss": 0.44005826115608215,
"step": 1504
},
{
"epoch": 3.3392461197339247,
"grad_norm": 0.23389166593551636,
"learning_rate": 1.6597812495965537e-06,
"loss": 0.18956291675567627,
"step": 1506
},
{
"epoch": 3.343680709534368,
"grad_norm": 0.6742655038833618,
"learning_rate": 1.651202887485394e-06,
"loss": 0.19848978519439697,
"step": 1508
},
{
"epoch": 3.3481152993348116,
"grad_norm": 1.077938199043274,
"learning_rate": 1.6426763062679553e-06,
"loss": 0.2809881567955017,
"step": 1510
},
{
"epoch": 3.352549889135255,
"grad_norm": 0.15749704837799072,
"learning_rate": 1.63420162065904e-06,
"loss": 0.2086062878370285,
"step": 1512
},
{
"epoch": 3.3569844789356984,
"grad_norm": 0.796462893486023,
"learning_rate": 1.625778944675257e-06,
"loss": 0.4808502495288849,
"step": 1514
},
{
"epoch": 3.361419068736142,
"grad_norm": 0.3194371163845062,
"learning_rate": 1.6174083916334877e-06,
"loss": 0.2608613967895508,
"step": 1516
},
{
"epoch": 3.3658536585365852,
"grad_norm": 0.6312937140464783,
"learning_rate": 1.609090074149366e-06,
"loss": 0.30664098262786865,
"step": 1518
},
{
"epoch": 3.3702882483370287,
"grad_norm": 0.6045692563056946,
"learning_rate": 1.6008241041357535e-06,
"loss": 0.4134957492351532,
"step": 1520
},
{
"epoch": 3.374722838137472,
"grad_norm": 0.35350367426872253,
"learning_rate": 1.5926105928012486e-06,
"loss": 0.25765693187713623,
"step": 1522
},
{
"epoch": 3.3791574279379155,
"grad_norm": 0.6642202138900757,
"learning_rate": 1.5844496506486734e-06,
"loss": 0.45118772983551025,
"step": 1524
},
{
"epoch": 3.3835920177383594,
"grad_norm": 1.6221054792404175,
"learning_rate": 1.576341387473601e-06,
"loss": 0.26244989037513733,
"step": 1526
},
{
"epoch": 3.388026607538803,
"grad_norm": 0.6710484623908997,
"learning_rate": 1.568285912362872e-06,
"loss": 0.18830682337284088,
"step": 1528
},
{
"epoch": 3.3924611973392462,
"grad_norm": 0.8072285652160645,
"learning_rate": 1.5602833336931242e-06,
"loss": 0.20579344034194946,
"step": 1530
},
{
"epoch": 3.3968957871396896,
"grad_norm": 1.560577392578125,
"learning_rate": 1.552333759129344e-06,
"loss": 0.07648878544569016,
"step": 1532
},
{
"epoch": 3.401330376940133,
"grad_norm": 0.8869048357009888,
"learning_rate": 1.5444372956234062e-06,
"loss": 0.3228890001773834,
"step": 1534
},
{
"epoch": 3.4057649667405765,
"grad_norm": 0.6825631260871887,
"learning_rate": 1.5365940494126424e-06,
"loss": 0.412392795085907,
"step": 1536
},
{
"epoch": 3.41019955654102,
"grad_norm": 0.7702367901802063,
"learning_rate": 1.5288041260184132e-06,
"loss": 0.33175259828567505,
"step": 1538
},
{
"epoch": 3.4146341463414633,
"grad_norm": 0.9447566270828247,
"learning_rate": 1.5210676302446801e-06,
"loss": 0.437308669090271,
"step": 1540
},
{
"epoch": 3.4190687361419068,
"grad_norm": 1.0833908319473267,
"learning_rate": 1.5133846661766058e-06,
"loss": 0.29791808128356934,
"step": 1542
},
{
"epoch": 3.42350332594235,
"grad_norm": 0.7799423933029175,
"learning_rate": 1.5057553371791461e-06,
"loss": 0.40058910846710205,
"step": 1544
},
{
"epoch": 3.4279379157427936,
"grad_norm": 1.4001895189285278,
"learning_rate": 1.4981797458956624e-06,
"loss": 0.031288325786590576,
"step": 1546
},
{
"epoch": 3.4323725055432375,
"grad_norm": 0.24610136449337006,
"learning_rate": 1.490657994246542e-06,
"loss": 0.09260788559913635,
"step": 1548
},
{
"epoch": 3.436807095343681,
"grad_norm": 0.6165451407432556,
"learning_rate": 1.4831901834278212e-06,
"loss": 0.5242594480514526,
"step": 1550
},
{
"epoch": 3.4412416851441243,
"grad_norm": 0.738479495048523,
"learning_rate": 1.4757764139098332e-06,
"loss": 0.37381070852279663,
"step": 1552
},
{
"epoch": 3.4456762749445677,
"grad_norm": 0.11110386252403259,
"learning_rate": 1.468416785435847e-06,
"loss": 0.22732020914554596,
"step": 1554
},
{
"epoch": 3.450110864745011,
"grad_norm": 0.8456872701644897,
"learning_rate": 1.461111397020732e-06,
"loss": 0.47603723406791687,
"step": 1556
},
{
"epoch": 3.4545454545454546,
"grad_norm": 1.9469002485275269,
"learning_rate": 1.4538603469496215e-06,
"loss": 0.6115579009056091,
"step": 1558
},
{
"epoch": 3.458980044345898,
"grad_norm": 0.6602691411972046,
"learning_rate": 1.4466637327765937e-06,
"loss": 0.4858705699443817,
"step": 1560
},
{
"epoch": 3.4634146341463414,
"grad_norm": 0.5199502110481262,
"learning_rate": 1.4395216513233584e-06,
"loss": 0.06964029371738434,
"step": 1562
},
{
"epoch": 3.467849223946785,
"grad_norm": 0.6388575434684753,
"learning_rate": 1.4324341986779527e-06,
"loss": 0.4515414237976074,
"step": 1564
},
{
"epoch": 3.4722838137472283,
"grad_norm": 0.5326871871948242,
"learning_rate": 1.4254014701934481e-06,
"loss": 0.20355406403541565,
"step": 1566
},
{
"epoch": 3.4767184035476717,
"grad_norm": 0.6119282841682434,
"learning_rate": 1.4184235604866725e-06,
"loss": 0.3266424834728241,
"step": 1568
},
{
"epoch": 3.481152993348115,
"grad_norm": 0.621752917766571,
"learning_rate": 1.4115005634369296e-06,
"loss": 0.19875553250312805,
"step": 1570
},
{
"epoch": 3.4855875831485585,
"grad_norm": 0.22762054204940796,
"learning_rate": 1.4046325721847443e-06,
"loss": 0.3008459210395813,
"step": 1572
},
{
"epoch": 3.4900221729490024,
"grad_norm": 0.6187798380851746,
"learning_rate": 1.397819679130601e-06,
"loss": 0.6530707478523254,
"step": 1574
},
{
"epoch": 3.494456762749446,
"grad_norm": 0.4245932996273041,
"learning_rate": 1.3910619759337074e-06,
"loss": 0.21562115848064423,
"step": 1576
},
{
"epoch": 3.4988913525498893,
"grad_norm": 0.7150433659553528,
"learning_rate": 1.3843595535107587e-06,
"loss": 0.21964870393276215,
"step": 1578
},
{
"epoch": 3.5033259423503327,
"grad_norm": 1.0825732946395874,
"learning_rate": 1.377712502034712e-06,
"loss": 0.4543288052082062,
"step": 1580
},
{
"epoch": 3.507760532150776,
"grad_norm": 0.8185349702835083,
"learning_rate": 1.3711209109335793e-06,
"loss": 0.5709398984909058,
"step": 1582
},
{
"epoch": 3.5121951219512195,
"grad_norm": 0.7394730448722839,
"learning_rate": 1.3645848688892162e-06,
"loss": 0.31390172243118286,
"step": 1584
},
{
"epoch": 3.516629711751663,
"grad_norm": 1.0575904846191406,
"learning_rate": 1.3581044638361373e-06,
"loss": 0.2191072404384613,
"step": 1586
},
{
"epoch": 3.5210643015521064,
"grad_norm": 1.3772166967391968,
"learning_rate": 1.3516797829603256e-06,
"loss": 0.40813693404197693,
"step": 1588
},
{
"epoch": 3.52549889135255,
"grad_norm": 1.6636643409729004,
"learning_rate": 1.3453109126980643e-06,
"loss": 0.13340415060520172,
"step": 1590
},
{
"epoch": 3.529933481152993,
"grad_norm": 1.4304341077804565,
"learning_rate": 1.3389979387347743e-06,
"loss": 0.24174071848392487,
"step": 1592
},
{
"epoch": 3.5343680709534366,
"grad_norm": 0.7794627547264099,
"learning_rate": 1.332740946003857e-06,
"loss": 0.485975980758667,
"step": 1594
},
{
"epoch": 3.5388026607538805,
"grad_norm": 1.178125023841858,
"learning_rate": 1.3265400186855548e-06,
"loss": 0.07965230196714401,
"step": 1596
},
{
"epoch": 3.5432372505543235,
"grad_norm": 1.4382803440093994,
"learning_rate": 1.320395240205819e-06,
"loss": 0.16565276682376862,
"step": 1598
},
{
"epoch": 3.5476718403547673,
"grad_norm": 0.6275985836982727,
"learning_rate": 1.3143066932351856e-06,
"loss": 0.25433921813964844,
"step": 1600
},
{
"epoch": 3.5521064301552108,
"grad_norm": 0.6336963772773743,
"learning_rate": 1.308274459687665e-06,
"loss": 0.3951624035835266,
"step": 1602
},
{
"epoch": 3.556541019955654,
"grad_norm": 0.611705482006073,
"learning_rate": 1.3022986207196367e-06,
"loss": 0.623420000076294,
"step": 1604
},
{
"epoch": 3.5609756097560976,
"grad_norm": 0.7854931354522705,
"learning_rate": 1.2963792567287617e-06,
"loss": 0.5146090984344482,
"step": 1606
},
{
"epoch": 3.565410199556541,
"grad_norm": 0.1012968197464943,
"learning_rate": 1.290516447352899e-06,
"loss": 0.023020246997475624,
"step": 1608
},
{
"epoch": 3.5698447893569845,
"grad_norm": 1.4004569053649902,
"learning_rate": 1.2847102714690308e-06,
"loss": 0.2902672588825226,
"step": 1610
},
{
"epoch": 3.574279379157428,
"grad_norm": 1.025610089302063,
"learning_rate": 1.2789608071922076e-06,
"loss": 0.03987615182995796,
"step": 1612
},
{
"epoch": 3.5787139689578713,
"grad_norm": 0.9752769470214844,
"learning_rate": 1.2732681318744923e-06,
"loss": 0.30136820673942566,
"step": 1614
},
{
"epoch": 3.5831485587583147,
"grad_norm": 0.9108896851539612,
"learning_rate": 1.2676323221039236e-06,
"loss": 0.5326212644577026,
"step": 1616
},
{
"epoch": 3.587583148558758,
"grad_norm": 0.5897319316864014,
"learning_rate": 1.2620534537034795e-06,
"loss": 0.2587262690067291,
"step": 1618
},
{
"epoch": 3.5920177383592016,
"grad_norm": 2.681466817855835,
"learning_rate": 1.2565316017300635e-06,
"loss": 0.4006616473197937,
"step": 1620
},
{
"epoch": 3.5964523281596454,
"grad_norm": 1.1970038414001465,
"learning_rate": 1.2510668404734924e-06,
"loss": 0.4482450783252716,
"step": 1622
},
{
"epoch": 3.6008869179600884,
"grad_norm": 0.7043033838272095,
"learning_rate": 1.2456592434554963e-06,
"loss": 0.3349531888961792,
"step": 1624
},
{
"epoch": 3.6053215077605323,
"grad_norm": 0.3154018223285675,
"learning_rate": 1.2403088834287282e-06,
"loss": 0.08102969080209732,
"step": 1626
},
{
"epoch": 3.6097560975609757,
"grad_norm": 0.5838890671730042,
"learning_rate": 1.2350158323757903e-06,
"loss": 0.41221365332603455,
"step": 1628
},
{
"epoch": 3.614190687361419,
"grad_norm": 1.219826102256775,
"learning_rate": 1.229780161508259e-06,
"loss": 0.25824180245399475,
"step": 1630
},
{
"epoch": 3.6186252771618626,
"grad_norm": 0.6512527465820312,
"learning_rate": 1.2246019412657319e-06,
"loss": 0.298809289932251,
"step": 1632
},
{
"epoch": 3.623059866962306,
"grad_norm": 0.8826231360435486,
"learning_rate": 1.2194812413148756e-06,
"loss": 0.4713267683982849,
"step": 1634
},
{
"epoch": 3.6274944567627494,
"grad_norm": 1.1182727813720703,
"learning_rate": 1.214418130548495e-06,
"loss": 0.23571792244911194,
"step": 1636
},
{
"epoch": 3.631929046563193,
"grad_norm": 3.134111166000366,
"learning_rate": 1.2094126770845986e-06,
"loss": 0.2846839427947998,
"step": 1638
},
{
"epoch": 3.6363636363636362,
"grad_norm": 0.7134626507759094,
"learning_rate": 1.2044649482654876e-06,
"loss": 0.41180163621902466,
"step": 1640
},
{
"epoch": 3.6407982261640797,
"grad_norm": 0.2941051423549652,
"learning_rate": 1.1995750106568496e-06,
"loss": 0.18607045710086823,
"step": 1642
},
{
"epoch": 3.6452328159645235,
"grad_norm": 0.8682871460914612,
"learning_rate": 1.1947429300468575e-06,
"loss": 0.2725341320037842,
"step": 1644
},
{
"epoch": 3.6496674057649665,
"grad_norm": 0.22906602919101715,
"learning_rate": 1.1899687714452932e-06,
"loss": 0.1510142982006073,
"step": 1646
},
{
"epoch": 3.6541019955654104,
"grad_norm": 0.5846609473228455,
"learning_rate": 1.1852525990826658e-06,
"loss": 0.17548693716526031,
"step": 1648
},
{
"epoch": 3.658536585365854,
"grad_norm": 0.8649150729179382,
"learning_rate": 1.1805944764093484e-06,
"loss": 0.3178820312023163,
"step": 1650
},
{
"epoch": 3.662971175166297,
"grad_norm": 0.7545923590660095,
"learning_rate": 1.1759944660947301e-06,
"loss": 0.4674905240535736,
"step": 1652
},
{
"epoch": 3.6674057649667406,
"grad_norm": 0.5282281041145325,
"learning_rate": 1.171452630026365e-06,
"loss": 0.22073353826999664,
"step": 1654
},
{
"epoch": 3.671840354767184,
"grad_norm": 0.6379572749137878,
"learning_rate": 1.1669690293091452e-06,
"loss": 0.44962742924690247,
"step": 1656
},
{
"epoch": 3.6762749445676275,
"grad_norm": 0.8793405890464783,
"learning_rate": 1.1625437242644772e-06,
"loss": 0.21119025349617004,
"step": 1658
},
{
"epoch": 3.680709534368071,
"grad_norm": 0.5134512186050415,
"learning_rate": 1.1581767744294682e-06,
"loss": 0.27141135931015015,
"step": 1660
},
{
"epoch": 3.6851441241685143,
"grad_norm": 0.17565104365348816,
"learning_rate": 1.1538682385561286e-06,
"loss": 0.23845592141151428,
"step": 1662
},
{
"epoch": 3.6895787139689578,
"grad_norm": 0.32826969027519226,
"learning_rate": 1.1496181746105784e-06,
"loss": 0.14338834583759308,
"step": 1664
},
{
"epoch": 3.694013303769401,
"grad_norm": 1.2581610679626465,
"learning_rate": 1.1454266397722707e-06,
"loss": 0.4462907016277313,
"step": 1666
},
{
"epoch": 3.6984478935698446,
"grad_norm": 0.7050212025642395,
"learning_rate": 1.1412936904332181e-06,
"loss": 0.2973634600639343,
"step": 1668
},
{
"epoch": 3.7028824833702885,
"grad_norm": 0.8381198644638062,
"learning_rate": 1.1372193821972379e-06,
"loss": 0.3616812229156494,
"step": 1670
},
{
"epoch": 3.7073170731707314,
"grad_norm": 0.5880244374275208,
"learning_rate": 1.1332037698792033e-06,
"loss": 0.317374587059021,
"step": 1672
},
{
"epoch": 3.7117516629711753,
"grad_norm": 0.5628880262374878,
"learning_rate": 1.1292469075043026e-06,
"loss": 0.5282438397407532,
"step": 1674
},
{
"epoch": 3.7161862527716187,
"grad_norm": 1.1323150396347046,
"learning_rate": 1.1253488483073177e-06,
"loss": 0.5078736543655396,
"step": 1676
},
{
"epoch": 3.720620842572062,
"grad_norm": 0.5901404023170471,
"learning_rate": 1.1215096447319038e-06,
"loss": 0.3895023763179779,
"step": 1678
},
{
"epoch": 3.7250554323725056,
"grad_norm": 1.2219599485397339,
"learning_rate": 1.117729348429884e-06,
"loss": 0.08083083480596542,
"step": 1680
},
{
"epoch": 3.729490022172949,
"grad_norm": 0.5432906150817871,
"learning_rate": 1.114008010260558e-06,
"loss": 0.27891483902931213,
"step": 1682
},
{
"epoch": 3.7339246119733924,
"grad_norm": 0.590613603591919,
"learning_rate": 1.1103456802900134e-06,
"loss": 0.25149670243263245,
"step": 1684
},
{
"epoch": 3.738359201773836,
"grad_norm": 1.1293264627456665,
"learning_rate": 1.1067424077904555e-06,
"loss": 0.2085466980934143,
"step": 1686
},
{
"epoch": 3.7427937915742793,
"grad_norm": 0.21245625615119934,
"learning_rate": 1.103198241239542e-06,
"loss": 0.08489034324884415,
"step": 1688
},
{
"epoch": 3.7472283813747227,
"grad_norm": 1.0891178846359253,
"learning_rate": 1.0997132283197324e-06,
"loss": 0.4237329363822937,
"step": 1690
},
{
"epoch": 3.7516629711751666,
"grad_norm": 0.7265070080757141,
"learning_rate": 1.0962874159176454e-06,
"loss": 0.3137209713459015,
"step": 1692
},
{
"epoch": 3.7560975609756095,
"grad_norm": 0.6292901635169983,
"learning_rate": 1.0929208501234286e-06,
"loss": 0.5190775394439697,
"step": 1694
},
{
"epoch": 3.7605321507760534,
"grad_norm": 0.8824191093444824,
"learning_rate": 1.0896135762301393e-06,
"loss": 0.4801103472709656,
"step": 1696
},
{
"epoch": 3.764966740576497,
"grad_norm": 1.3474018573760986,
"learning_rate": 1.0863656387331328e-06,
"loss": 0.2235155552625656,
"step": 1698
},
{
"epoch": 3.7694013303769403,
"grad_norm": 0.7029469013214111,
"learning_rate": 1.0831770813294668e-06,
"loss": 0.226326584815979,
"step": 1700
},
{
"epoch": 3.7738359201773837,
"grad_norm": 0.7810203433036804,
"learning_rate": 1.0800479469173101e-06,
"loss": 0.7419723272323608,
"step": 1702
},
{
"epoch": 3.778270509977827,
"grad_norm": 0.42261672019958496,
"learning_rate": 1.076978277595369e-06,
"loss": 0.07026417553424835,
"step": 1704
},
{
"epoch": 3.7827050997782705,
"grad_norm": 0.1440313309431076,
"learning_rate": 1.0739681146623185e-06,
"loss": 0.24512067437171936,
"step": 1706
},
{
"epoch": 3.787139689578714,
"grad_norm": 0.6102575659751892,
"learning_rate": 1.0710174986162471e-06,
"loss": 0.22913426160812378,
"step": 1708
},
{
"epoch": 3.7915742793791574,
"grad_norm": 0.6097363233566284,
"learning_rate": 1.0681264691541127e-06,
"loss": 0.5384188294410706,
"step": 1710
},
{
"epoch": 3.796008869179601,
"grad_norm": 0.7359468936920166,
"learning_rate": 1.0652950651712072e-06,
"loss": 0.24644701182842255,
"step": 1712
},
{
"epoch": 3.800443458980044,
"grad_norm": 1.484494924545288,
"learning_rate": 1.0625233247606348e-06,
"loss": 0.2666923403739929,
"step": 1714
},
{
"epoch": 3.8048780487804876,
"grad_norm": 0.5770846605300903,
"learning_rate": 1.059811285212799e-06,
"loss": 0.39251670241355896,
"step": 1716
},
{
"epoch": 3.8093126385809315,
"grad_norm": 0.5093443393707275,
"learning_rate": 1.0571589830149e-06,
"loss": 0.2737593352794647,
"step": 1718
},
{
"epoch": 3.8137472283813745,
"grad_norm": 0.7546675205230713,
"learning_rate": 1.054566453850444e-06,
"loss": 0.30364954471588135,
"step": 1720
},
{
"epoch": 3.8181818181818183,
"grad_norm": 0.9405109286308289,
"learning_rate": 1.0520337325987649e-06,
"loss": 0.464808851480484,
"step": 1722
},
{
"epoch": 3.8226164079822618,
"grad_norm": 0.9105086922645569,
"learning_rate": 1.049560853334553e-06,
"loss": 0.4284362494945526,
"step": 1724
},
{
"epoch": 3.827050997782705,
"grad_norm": 0.5973249673843384,
"learning_rate": 1.0471478493273976e-06,
"loss": 0.45257946848869324,
"step": 1726
},
{
"epoch": 3.8314855875831486,
"grad_norm": 0.5348697900772095,
"learning_rate": 1.0447947530413389e-06,
"loss": 0.19484858214855194,
"step": 1728
},
{
"epoch": 3.835920177383592,
"grad_norm": 1.245671033859253,
"learning_rate": 1.042501596134431e-06,
"loss": 0.24068011343479156,
"step": 1730
},
{
"epoch": 3.8403547671840355,
"grad_norm": 0.6368365287780762,
"learning_rate": 1.0402684094583173e-06,
"loss": 0.5048990249633789,
"step": 1732
},
{
"epoch": 3.844789356984479,
"grad_norm": 2.171820640563965,
"learning_rate": 1.0380952230578125e-06,
"loss": 0.29761821031570435,
"step": 1734
},
{
"epoch": 3.8492239467849223,
"grad_norm": 2.3513731956481934,
"learning_rate": 1.0359820661705042e-06,
"loss": 0.3343811333179474,
"step": 1736
},
{
"epoch": 3.8536585365853657,
"grad_norm": 0.7197679877281189,
"learning_rate": 1.0339289672263519e-06,
"loss": 0.4054519534111023,
"step": 1738
},
{
"epoch": 3.858093126385809,
"grad_norm": 0.6081451773643494,
"learning_rate": 1.0319359538473107e-06,
"loss": 0.219305619597435,
"step": 1740
},
{
"epoch": 3.8625277161862526,
"grad_norm": 0.8351064324378967,
"learning_rate": 1.0300030528469564e-06,
"loss": 0.20432864129543304,
"step": 1742
},
{
"epoch": 3.8669623059866964,
"grad_norm": 0.7205188274383545,
"learning_rate": 1.0281302902301254e-06,
"loss": 0.31986573338508606,
"step": 1744
},
{
"epoch": 3.8713968957871394,
"grad_norm": 0.6454185843467712,
"learning_rate": 1.026317691192567e-06,
"loss": 0.35778456926345825,
"step": 1746
},
{
"epoch": 3.8758314855875833,
"grad_norm": 0.6062719225883484,
"learning_rate": 1.0245652801205999e-06,
"loss": 0.21619272232055664,
"step": 1748
},
{
"epoch": 3.8802660753880267,
"grad_norm": 0.6591951251029968,
"learning_rate": 1.0228730805907891e-06,
"loss": 0.29133281111717224,
"step": 1750
},
{
"epoch": 3.88470066518847,
"grad_norm": 2.8523550033569336,
"learning_rate": 1.0212411153696247e-06,
"loss": 0.36294376850128174,
"step": 1752
},
{
"epoch": 3.8891352549889135,
"grad_norm": 0.8326625227928162,
"learning_rate": 1.019669406413218e-06,
"loss": 0.3876746892929077,
"step": 1754
},
{
"epoch": 3.893569844789357,
"grad_norm": 2.4112775325775146,
"learning_rate": 1.0181579748670054e-06,
"loss": 0.3648989200592041,
"step": 1756
},
{
"epoch": 3.8980044345898004,
"grad_norm": 0.5865671038627625,
"learning_rate": 1.0167068410654643e-06,
"loss": 0.4565258026123047,
"step": 1758
},
{
"epoch": 3.902439024390244,
"grad_norm": 0.07741643488407135,
"learning_rate": 1.0153160245318384e-06,
"loss": 0.0015238127671182156,
"step": 1760
},
{
"epoch": 3.9068736141906872,
"grad_norm": 1.0577608346939087,
"learning_rate": 1.0139855439778766e-06,
"loss": 0.08189034461975098,
"step": 1762
},
{
"epoch": 3.9113082039911307,
"grad_norm": 0.07507964968681335,
"learning_rate": 1.0127154173035787e-06,
"loss": 0.17938612401485443,
"step": 1764
},
{
"epoch": 3.9157427937915745,
"grad_norm": 0.647083044052124,
"learning_rate": 1.0115056615969584e-06,
"loss": 0.3286856710910797,
"step": 1766
},
{
"epoch": 3.9201773835920175,
"grad_norm": 0.8342782855033875,
"learning_rate": 1.0103562931338105e-06,
"loss": 0.47932499647140503,
"step": 1768
},
{
"epoch": 3.9246119733924614,
"grad_norm": 0.5373421907424927,
"learning_rate": 1.009267327377492e-06,
"loss": 0.465029776096344,
"step": 1770
},
{
"epoch": 3.929046563192905,
"grad_norm": 0.5635690689086914,
"learning_rate": 1.008238778978716e-06,
"loss": 0.05487694591283798,
"step": 1772
},
{
"epoch": 3.933481152993348,
"grad_norm": 0.8110321164131165,
"learning_rate": 1.0072706617753528e-06,
"loss": 0.3653331696987152,
"step": 1774
},
{
"epoch": 3.9379157427937916,
"grad_norm": 1.7579855918884277,
"learning_rate": 1.0063629887922441e-06,
"loss": 0.44433099031448364,
"step": 1776
},
{
"epoch": 3.942350332594235,
"grad_norm": 0.13488821685314178,
"learning_rate": 1.0055157722410279e-06,
"loss": 0.02499130368232727,
"step": 1778
},
{
"epoch": 3.9467849223946785,
"grad_norm": 1.413166880607605,
"learning_rate": 1.0047290235199753e-06,
"loss": 0.21658755838871002,
"step": 1780
},
{
"epoch": 3.951219512195122,
"grad_norm": 0.14947852492332458,
"learning_rate": 1.0040027532138351e-06,
"loss": 0.2654849886894226,
"step": 1782
},
{
"epoch": 3.9556541019955653,
"grad_norm": 0.6576520204544067,
"learning_rate": 1.0033369710936928e-06,
"loss": 0.27802592515945435,
"step": 1784
},
{
"epoch": 3.9600886917960088,
"grad_norm": 0.6993730068206787,
"learning_rate": 1.0027316861168388e-06,
"loss": 0.35919517278671265,
"step": 1786
},
{
"epoch": 3.964523281596452,
"grad_norm": 0.5883303284645081,
"learning_rate": 1.0021869064266472e-06,
"loss": 0.2748097777366638,
"step": 1788
},
{
"epoch": 3.9689578713968956,
"grad_norm": 0.6255540251731873,
"learning_rate": 1.0017026393524684e-06,
"loss": 0.281380295753479,
"step": 1790
},
{
"epoch": 3.9733924611973395,
"grad_norm": 0.6950498223304749,
"learning_rate": 1.0012788914095275e-06,
"loss": 0.3902501165866852,
"step": 1792
},
{
"epoch": 3.9778270509977824,
"grad_norm": 0.7842665314674377,
"learning_rate": 1.0009156682988395e-06,
"loss": 0.19079414010047913,
"step": 1794
},
{
"epoch": 3.9822616407982263,
"grad_norm": 0.9547889232635498,
"learning_rate": 1.0006129749071298e-06,
"loss": 0.2992710769176483,
"step": 1796
},
{
"epoch": 3.9866962305986697,
"grad_norm": 0.7517938017845154,
"learning_rate": 1.00037081530677e-06,
"loss": 0.3780803680419922,
"step": 1798
},
{
"epoch": 3.991130820399113,
"grad_norm": 1.7430669069290161,
"learning_rate": 1.0001891927557255e-06,
"loss": 0.32227784395217896,
"step": 1800
},
{
"epoch": 3.9955654101995566,
"grad_norm": 0.7806469202041626,
"learning_rate": 1.0000681096975056e-06,
"loss": 0.3458644151687622,
"step": 1802
},
{
"epoch": 4.0,
"grad_norm": 0.5159145593643188,
"learning_rate": 1.0000075677611364e-06,
"loss": 0.12477380037307739,
"step": 1804
},
{
"epoch": 4.0,
"step": 1804,
"total_flos": 3.4175049861232067e+18,
"train_loss": 0.7412061936664963,
"train_runtime": 8152.0644,
"train_samples_per_second": 6.639,
"train_steps_per_second": 0.221
}
],
"logging_steps": 2,
"max_steps": 1804,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 99999,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3.4175049861232067e+18,
"train_batch_size": 3,
"trial_name": null,
"trial_params": null
}