9b-24 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
3b3d2ed verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 2055,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00291970802919708,
"grad_norm": 0.3534850478172302,
"learning_rate": 9.70873786407767e-13,
"loss": 1.7768380641937256,
"step": 2
},
{
"epoch": 0.00583941605839416,
"grad_norm": 0.2643459439277649,
"learning_rate": 2.912621359223301e-12,
"loss": 1.800352931022644,
"step": 4
},
{
"epoch": 0.008759124087591242,
"grad_norm": 0.25368738174438477,
"learning_rate": 4.854368932038835e-12,
"loss": 1.815911889076233,
"step": 6
},
{
"epoch": 0.01167883211678832,
"grad_norm": 0.2690427601337433,
"learning_rate": 6.796116504854369e-12,
"loss": 1.8008553981781006,
"step": 8
},
{
"epoch": 0.014598540145985401,
"grad_norm": 0.27801981568336487,
"learning_rate": 8.737864077669904e-12,
"loss": 1.7111756801605225,
"step": 10
},
{
"epoch": 0.017518248175182483,
"grad_norm": 0.30863869190216064,
"learning_rate": 1.0679611650485436e-11,
"loss": 1.7506548166275024,
"step": 12
},
{
"epoch": 0.020437956204379562,
"grad_norm": 0.25328800082206726,
"learning_rate": 1.2621359223300972e-11,
"loss": 1.6444569826126099,
"step": 14
},
{
"epoch": 0.02335766423357664,
"grad_norm": 0.5986829996109009,
"learning_rate": 1.4563106796116506e-11,
"loss": 1.7497678995132446,
"step": 16
},
{
"epoch": 0.026277372262773723,
"grad_norm": 0.20500004291534424,
"learning_rate": 1.6504854368932042e-11,
"loss": 1.5234403610229492,
"step": 18
},
{
"epoch": 0.029197080291970802,
"grad_norm": 0.2636180520057678,
"learning_rate": 1.8446601941747574e-11,
"loss": 1.8194037675857544,
"step": 20
},
{
"epoch": 0.032116788321167884,
"grad_norm": 0.6845226883888245,
"learning_rate": 2.0388349514563107e-11,
"loss": 1.7494784593582153,
"step": 22
},
{
"epoch": 0.035036496350364967,
"grad_norm": 0.29200759530067444,
"learning_rate": 2.2330097087378642e-11,
"loss": 1.57957124710083,
"step": 24
},
{
"epoch": 0.03795620437956204,
"grad_norm": 0.25369206070899963,
"learning_rate": 2.4271844660194175e-11,
"loss": 1.7397639751434326,
"step": 26
},
{
"epoch": 0.040875912408759124,
"grad_norm": 0.6018857359886169,
"learning_rate": 2.621359223300971e-11,
"loss": 1.9077666997909546,
"step": 28
},
{
"epoch": 0.043795620437956206,
"grad_norm": 0.35382694005966187,
"learning_rate": 2.8155339805825243e-11,
"loss": 1.7949832677841187,
"step": 30
},
{
"epoch": 0.04671532846715328,
"grad_norm": 1.6109180450439453,
"learning_rate": 3.009708737864078e-11,
"loss": 1.751054048538208,
"step": 32
},
{
"epoch": 0.049635036496350364,
"grad_norm": 0.19936665892601013,
"learning_rate": 3.203883495145632e-11,
"loss": 1.4549548625946045,
"step": 34
},
{
"epoch": 0.052554744525547446,
"grad_norm": 0.31096768379211426,
"learning_rate": 3.398058252427185e-11,
"loss": 1.3109430074691772,
"step": 36
},
{
"epoch": 0.05547445255474453,
"grad_norm": 0.24408775568008423,
"learning_rate": 3.592233009708738e-11,
"loss": 1.644413948059082,
"step": 38
},
{
"epoch": 0.058394160583941604,
"grad_norm": 0.3106284439563751,
"learning_rate": 3.7864077669902915e-11,
"loss": 1.7961643934249878,
"step": 40
},
{
"epoch": 0.061313868613138686,
"grad_norm": 0.32802024483680725,
"learning_rate": 3.980582524271845e-11,
"loss": 1.8175064325332642,
"step": 42
},
{
"epoch": 0.06423357664233577,
"grad_norm": 0.3910239040851593,
"learning_rate": 4.174757281553398e-11,
"loss": 1.8361930847167969,
"step": 44
},
{
"epoch": 0.06715328467153285,
"grad_norm": 0.5940719246864319,
"learning_rate": 4.368932038834951e-11,
"loss": 2.0858569145202637,
"step": 46
},
{
"epoch": 0.07007299270072993,
"grad_norm": 0.2699325382709503,
"learning_rate": 4.563106796116505e-11,
"loss": 1.9244903326034546,
"step": 48
},
{
"epoch": 0.072992700729927,
"grad_norm": 0.3519454002380371,
"learning_rate": 4.757281553398058e-11,
"loss": 1.6795828342437744,
"step": 50
},
{
"epoch": 0.07591240875912408,
"grad_norm": 1.271136999130249,
"learning_rate": 4.9514563106796115e-11,
"loss": 1.8167542219161987,
"step": 52
},
{
"epoch": 0.07883211678832117,
"grad_norm": 0.32721835374832153,
"learning_rate": 5.1456310679611654e-11,
"loss": 1.8619266748428345,
"step": 54
},
{
"epoch": 0.08175182481751825,
"grad_norm": 0.35416823625564575,
"learning_rate": 5.339805825242719e-11,
"loss": 1.8907665014266968,
"step": 56
},
{
"epoch": 0.08467153284671533,
"grad_norm": 0.1937120109796524,
"learning_rate": 5.533980582524272e-11,
"loss": 1.4969103336334229,
"step": 58
},
{
"epoch": 0.08759124087591241,
"grad_norm": 0.2578057646751404,
"learning_rate": 5.728155339805825e-11,
"loss": 1.4486453533172607,
"step": 60
},
{
"epoch": 0.0905109489051095,
"grad_norm": 0.23345084488391876,
"learning_rate": 5.922330097087378e-11,
"loss": 1.5223218202590942,
"step": 62
},
{
"epoch": 0.09343065693430656,
"grad_norm": 0.23968848586082458,
"learning_rate": 6.116504854368932e-11,
"loss": 1.6385910511016846,
"step": 64
},
{
"epoch": 0.09635036496350365,
"grad_norm": 0.23735839128494263,
"learning_rate": 6.310679611650486e-11,
"loss": 1.6215620040893555,
"step": 66
},
{
"epoch": 0.09927007299270073,
"grad_norm": 0.274890661239624,
"learning_rate": 6.50485436893204e-11,
"loss": 1.5964994430541992,
"step": 68
},
{
"epoch": 0.10218978102189781,
"grad_norm": 0.26082801818847656,
"learning_rate": 6.699029126213593e-11,
"loss": 1.681908369064331,
"step": 70
},
{
"epoch": 0.10510948905109489,
"grad_norm": 0.4017055928707123,
"learning_rate": 6.893203883495146e-11,
"loss": 1.853885531425476,
"step": 72
},
{
"epoch": 0.10802919708029197,
"grad_norm": 0.26503318548202515,
"learning_rate": 7.087378640776699e-11,
"loss": 1.830824375152588,
"step": 74
},
{
"epoch": 0.11094890510948906,
"grad_norm": 0.4232666492462158,
"learning_rate": 7.281553398058252e-11,
"loss": 1.809700608253479,
"step": 76
},
{
"epoch": 0.11386861313868613,
"grad_norm": 0.5037844181060791,
"learning_rate": 7.475728155339806e-11,
"loss": 1.9286326169967651,
"step": 78
},
{
"epoch": 0.11678832116788321,
"grad_norm": 0.39150989055633545,
"learning_rate": 7.669902912621359e-11,
"loss": 1.816016435623169,
"step": 80
},
{
"epoch": 0.11970802919708029,
"grad_norm": 0.9030010104179382,
"learning_rate": 7.864077669902912e-11,
"loss": 1.778789758682251,
"step": 82
},
{
"epoch": 0.12262773722627737,
"grad_norm": 1.9095746278762817,
"learning_rate": 8.058252427184467e-11,
"loss": 1.9802308082580566,
"step": 84
},
{
"epoch": 0.12554744525547445,
"grad_norm": 0.30885621905326843,
"learning_rate": 8.25242718446602e-11,
"loss": 1.7350918054580688,
"step": 86
},
{
"epoch": 0.12846715328467154,
"grad_norm": 0.3458031415939331,
"learning_rate": 8.446601941747573e-11,
"loss": 1.5395787954330444,
"step": 88
},
{
"epoch": 0.13138686131386862,
"grad_norm": 0.5105660557746887,
"learning_rate": 8.640776699029126e-11,
"loss": 1.821498990058899,
"step": 90
},
{
"epoch": 0.1343065693430657,
"grad_norm": 0.23492082953453064,
"learning_rate": 8.834951456310681e-11,
"loss": 1.8252274990081787,
"step": 92
},
{
"epoch": 0.13722627737226278,
"grad_norm": 0.36012178659439087,
"learning_rate": 9.029126213592234e-11,
"loss": 1.8968510627746582,
"step": 94
},
{
"epoch": 0.14014598540145987,
"grad_norm": 0.4315379858016968,
"learning_rate": 9.223300970873787e-11,
"loss": 1.7809394598007202,
"step": 96
},
{
"epoch": 0.14306569343065692,
"grad_norm": 0.2982156276702881,
"learning_rate": 9.41747572815534e-11,
"loss": 1.7060034275054932,
"step": 98
},
{
"epoch": 0.145985401459854,
"grad_norm": 0.3450309634208679,
"learning_rate": 9.611650485436894e-11,
"loss": 1.9719310998916626,
"step": 100
},
{
"epoch": 0.14890510948905109,
"grad_norm": 0.32693490386009216,
"learning_rate": 9.805825242718447e-11,
"loss": 1.9070788621902466,
"step": 102
},
{
"epoch": 0.15182481751824817,
"grad_norm": 1.445264220237732,
"learning_rate": 1e-10,
"loss": 1.7498117685317993,
"step": 104
},
{
"epoch": 0.15474452554744525,
"grad_norm": 0.3428294062614441,
"learning_rate": 9.999976687854062e-11,
"loss": 1.8611483573913574,
"step": 106
},
{
"epoch": 0.15766423357664233,
"grad_norm": 0.3272436857223511,
"learning_rate": 9.999906751657785e-11,
"loss": 1.9175266027450562,
"step": 108
},
{
"epoch": 0.16058394160583941,
"grad_norm": 0.909197986125946,
"learning_rate": 9.999790192135772e-11,
"loss": 1.8308496475219727,
"step": 110
},
{
"epoch": 0.1635036496350365,
"grad_norm": 0.38203534483909607,
"learning_rate": 9.999627010495695e-11,
"loss": 1.8768818378448486,
"step": 112
},
{
"epoch": 0.16642335766423358,
"grad_norm": 0.2889254689216614,
"learning_rate": 9.999417208428267e-11,
"loss": 1.636149287223816,
"step": 114
},
{
"epoch": 0.16934306569343066,
"grad_norm": 0.4831934869289398,
"learning_rate": 9.999160788107241e-11,
"loss": 1.942168116569519,
"step": 116
},
{
"epoch": 0.17226277372262774,
"grad_norm": 0.2783659100532532,
"learning_rate": 9.998857752189376e-11,
"loss": 1.8512721061706543,
"step": 118
},
{
"epoch": 0.17518248175182483,
"grad_norm": 0.31743311882019043,
"learning_rate": 9.99850810381441e-11,
"loss": 1.9434564113616943,
"step": 120
},
{
"epoch": 0.1781021897810219,
"grad_norm": 0.26472657918930054,
"learning_rate": 9.998111846605035e-11,
"loss": 1.8089823722839355,
"step": 122
},
{
"epoch": 0.181021897810219,
"grad_norm": 0.47778958082199097,
"learning_rate": 9.997668984666856e-11,
"loss": 1.7395007610321045,
"step": 124
},
{
"epoch": 0.18394160583941604,
"grad_norm": 0.3084363639354706,
"learning_rate": 9.99717952258834e-11,
"loss": 1.9149442911148071,
"step": 126
},
{
"epoch": 0.18686131386861313,
"grad_norm": 0.381334513425827,
"learning_rate": 9.996643465440788e-11,
"loss": 1.7895768880844116,
"step": 128
},
{
"epoch": 0.1897810218978102,
"grad_norm": 0.3909507393836975,
"learning_rate": 9.996060818778257e-11,
"loss": 1.8174935579299927,
"step": 130
},
{
"epoch": 0.1927007299270073,
"grad_norm": 0.40029406547546387,
"learning_rate": 9.995431588637526e-11,
"loss": 1.8789606094360352,
"step": 132
},
{
"epoch": 0.19562043795620437,
"grad_norm": 0.24144038558006287,
"learning_rate": 9.994755781538018e-11,
"loss": 1.704660177230835,
"step": 134
},
{
"epoch": 0.19854014598540146,
"grad_norm": 0.15783090889453888,
"learning_rate": 9.994033404481737e-11,
"loss": 1.5398731231689453,
"step": 136
},
{
"epoch": 0.20145985401459854,
"grad_norm": 0.1733558028936386,
"learning_rate": 9.993264464953202e-11,
"loss": 1.304384469985962,
"step": 138
},
{
"epoch": 0.20437956204379562,
"grad_norm": 0.3875671923160553,
"learning_rate": 9.992448970919358e-11,
"loss": 1.786711573600769,
"step": 140
},
{
"epoch": 0.2072992700729927,
"grad_norm": 0.294428288936615,
"learning_rate": 9.991586930829501e-11,
"loss": 1.7124841213226318,
"step": 142
},
{
"epoch": 0.21021897810218979,
"grad_norm": 0.19950203597545624,
"learning_rate": 9.990678353615189e-11,
"loss": 1.663377285003662,
"step": 144
},
{
"epoch": 0.21313868613138687,
"grad_norm": 0.26574021577835083,
"learning_rate": 9.989723248690149e-11,
"loss": 1.7981501817703247,
"step": 146
},
{
"epoch": 0.21605839416058395,
"grad_norm": 0.3551180958747864,
"learning_rate": 9.988721625950177e-11,
"loss": 1.694273591041565,
"step": 148
},
{
"epoch": 0.21897810218978103,
"grad_norm": 0.22780469059944153,
"learning_rate": 9.987673495773042e-11,
"loss": 1.8389781713485718,
"step": 150
},
{
"epoch": 0.22189781021897811,
"grad_norm": 0.29115864634513855,
"learning_rate": 9.986578869018371e-11,
"loss": 1.8860565423965454,
"step": 152
},
{
"epoch": 0.22481751824817517,
"grad_norm": 0.44859740138053894,
"learning_rate": 9.985437757027541e-11,
"loss": 1.9679546356201172,
"step": 154
},
{
"epoch": 0.22773722627737225,
"grad_norm": 0.8019328117370605,
"learning_rate": 9.98425017162356e-11,
"loss": 1.8494609594345093,
"step": 156
},
{
"epoch": 0.23065693430656933,
"grad_norm": 0.4708123505115509,
"learning_rate": 9.98301612511095e-11,
"loss": 1.9518005847930908,
"step": 158
},
{
"epoch": 0.23357664233576642,
"grad_norm": 0.39831846952438354,
"learning_rate": 9.981735630275602e-11,
"loss": 2.0240159034729004,
"step": 160
},
{
"epoch": 0.2364963503649635,
"grad_norm": 0.5212871432304382,
"learning_rate": 9.980408700384671e-11,
"loss": 2.01702618598938,
"step": 162
},
{
"epoch": 0.23941605839416058,
"grad_norm": 0.9411134123802185,
"learning_rate": 9.979035349186414e-11,
"loss": 2.1166324615478516,
"step": 164
},
{
"epoch": 0.24233576642335766,
"grad_norm": 0.5210184454917908,
"learning_rate": 9.977615590910056e-11,
"loss": 1.9885681867599487,
"step": 166
},
{
"epoch": 0.24525547445255474,
"grad_norm": 0.8235350251197815,
"learning_rate": 9.976149440265651e-11,
"loss": 2.311216354370117,
"step": 168
},
{
"epoch": 0.24817518248175183,
"grad_norm": 0.9275002479553223,
"learning_rate": 9.974636912443916e-11,
"loss": 1.8653641939163208,
"step": 170
},
{
"epoch": 0.2510948905109489,
"grad_norm": 0.6768398880958557,
"learning_rate": 9.973078023116083e-11,
"loss": 1.848363995552063,
"step": 172
},
{
"epoch": 0.25401459854014596,
"grad_norm": 0.7485382556915283,
"learning_rate": 9.971472788433731e-11,
"loss": 1.9982264041900635,
"step": 174
},
{
"epoch": 0.2569343065693431,
"grad_norm": 4.182097911834717,
"learning_rate": 9.969821225028622e-11,
"loss": 2.407541513442993,
"step": 176
},
{
"epoch": 0.25985401459854013,
"grad_norm": 0.42571887373924255,
"learning_rate": 9.968123350012531e-11,
"loss": 1.7301071882247925,
"step": 178
},
{
"epoch": 0.26277372262773724,
"grad_norm": 0.5800350308418274,
"learning_rate": 9.966379180977059e-11,
"loss": 1.6739298105239868,
"step": 180
},
{
"epoch": 0.2656934306569343,
"grad_norm": 0.34787094593048096,
"learning_rate": 9.964588735993461e-11,
"loss": 1.5040230751037598,
"step": 182
},
{
"epoch": 0.2686131386861314,
"grad_norm": 0.7273563146591187,
"learning_rate": 9.962752033612457e-11,
"loss": 1.762050747871399,
"step": 184
},
{
"epoch": 0.27153284671532846,
"grad_norm": 1.0910371541976929,
"learning_rate": 9.960869092864034e-11,
"loss": 1.5401787757873535,
"step": 186
},
{
"epoch": 0.27445255474452557,
"grad_norm": 0.3126949667930603,
"learning_rate": 9.958939933257254e-11,
"loss": 1.420863389968872,
"step": 188
},
{
"epoch": 0.2773722627737226,
"grad_norm": 0.37012311816215515,
"learning_rate": 9.95696457478005e-11,
"loss": 1.5388368368148804,
"step": 190
},
{
"epoch": 0.28029197080291973,
"grad_norm": 0.20049013197422028,
"learning_rate": 9.954943037899022e-11,
"loss": 1.1660946607589722,
"step": 192
},
{
"epoch": 0.2832116788321168,
"grad_norm": 0.6460405588150024,
"learning_rate": 9.952875343559217e-11,
"loss": 1.5849355459213257,
"step": 194
},
{
"epoch": 0.28613138686131384,
"grad_norm": 0.7101665139198303,
"learning_rate": 9.950761513183924e-11,
"loss": 1.9086475372314453,
"step": 196
},
{
"epoch": 0.28905109489051095,
"grad_norm": 0.8782652020454407,
"learning_rate": 9.948601568674439e-11,
"loss": 1.9283380508422852,
"step": 198
},
{
"epoch": 0.291970802919708,
"grad_norm": 0.8213207125663757,
"learning_rate": 9.946395532409846e-11,
"loss": 1.6717543601989746,
"step": 200
},
{
"epoch": 0.2948905109489051,
"grad_norm": 0.8597667217254639,
"learning_rate": 9.94414342724679e-11,
"loss": 2.2654058933258057,
"step": 202
},
{
"epoch": 0.29781021897810217,
"grad_norm": 0.31900808215141296,
"learning_rate": 9.941845276519223e-11,
"loss": 1.2488539218902588,
"step": 204
},
{
"epoch": 0.3007299270072993,
"grad_norm": 0.2743098735809326,
"learning_rate": 9.93950110403818e-11,
"loss": 1.5524955987930298,
"step": 206
},
{
"epoch": 0.30364963503649633,
"grad_norm": 0.44529157876968384,
"learning_rate": 9.937110934091525e-11,
"loss": 1.9126129150390625,
"step": 208
},
{
"epoch": 0.30656934306569344,
"grad_norm": 0.2431587129831314,
"learning_rate": 9.934674791443698e-11,
"loss": 1.6700682640075684,
"step": 210
},
{
"epoch": 0.3094890510948905,
"grad_norm": 0.25222745537757874,
"learning_rate": 9.932192701335459e-11,
"loss": 1.792035460472107,
"step": 212
},
{
"epoch": 0.3124087591240876,
"grad_norm": 0.29361167550086975,
"learning_rate": 9.929664689483631e-11,
"loss": 1.7590653896331787,
"step": 214
},
{
"epoch": 0.31532846715328466,
"grad_norm": 0.8557894229888916,
"learning_rate": 9.927090782080824e-11,
"loss": 1.598631739616394,
"step": 216
},
{
"epoch": 0.3182481751824818,
"grad_norm": 0.19789443910121918,
"learning_rate": 9.924471005795176e-11,
"loss": 1.61155104637146,
"step": 218
},
{
"epoch": 0.32116788321167883,
"grad_norm": 1.04644775390625,
"learning_rate": 9.92180538777007e-11,
"loss": 1.526228904724121,
"step": 220
},
{
"epoch": 0.32408759124087594,
"grad_norm": 0.22464589774608612,
"learning_rate": 9.919093955623848e-11,
"loss": 1.368972659111023,
"step": 222
},
{
"epoch": 0.327007299270073,
"grad_norm": 0.8330907225608826,
"learning_rate": 9.916336737449534e-11,
"loss": 2.178337335586548,
"step": 224
},
{
"epoch": 0.32992700729927005,
"grad_norm": 0.24324865639209747,
"learning_rate": 9.913533761814537e-11,
"loss": 1.7530229091644287,
"step": 226
},
{
"epoch": 0.33284671532846716,
"grad_norm": 0.6675015091896057,
"learning_rate": 9.91068505776036e-11,
"loss": 1.697060227394104,
"step": 228
},
{
"epoch": 0.3357664233576642,
"grad_norm": 0.5403636693954468,
"learning_rate": 9.907790654802293e-11,
"loss": 1.4822953939437866,
"step": 230
},
{
"epoch": 0.3386861313868613,
"grad_norm": 0.7010171413421631,
"learning_rate": 9.90485058292911e-11,
"loss": 1.9493117332458496,
"step": 232
},
{
"epoch": 0.3416058394160584,
"grad_norm": 0.38018572330474854,
"learning_rate": 9.901864872602762e-11,
"loss": 1.9160114526748657,
"step": 234
},
{
"epoch": 0.3445255474452555,
"grad_norm": 1.252315640449524,
"learning_rate": 9.898833554758054e-11,
"loss": 1.9603207111358643,
"step": 236
},
{
"epoch": 0.34744525547445254,
"grad_norm": 0.41888001561164856,
"learning_rate": 9.895756660802331e-11,
"loss": 1.890695571899414,
"step": 238
},
{
"epoch": 0.35036496350364965,
"grad_norm": 0.4101668894290924,
"learning_rate": 9.892634222615147e-11,
"loss": 1.7084157466888428,
"step": 240
},
{
"epoch": 0.3532846715328467,
"grad_norm": 1.959101915359497,
"learning_rate": 9.889466272547941e-11,
"loss": 1.9957774877548218,
"step": 242
},
{
"epoch": 0.3562043795620438,
"grad_norm": 1.0332179069519043,
"learning_rate": 9.886252843423699e-11,
"loss": 1.825124740600586,
"step": 244
},
{
"epoch": 0.35912408759124087,
"grad_norm": 0.2979462146759033,
"learning_rate": 9.882993968536607e-11,
"loss": 1.7160868644714355,
"step": 246
},
{
"epoch": 0.362043795620438,
"grad_norm": 0.712623655796051,
"learning_rate": 9.879689681651721e-11,
"loss": 2.088395595550537,
"step": 248
},
{
"epoch": 0.36496350364963503,
"grad_norm": 0.4699522852897644,
"learning_rate": 9.876340017004604e-11,
"loss": 2.0031797885894775,
"step": 250
},
{
"epoch": 0.3678832116788321,
"grad_norm": 0.4952646493911743,
"learning_rate": 9.872945009300975e-11,
"loss": 2.03470516204834,
"step": 252
},
{
"epoch": 0.3708029197080292,
"grad_norm": 0.8986260294914246,
"learning_rate": 9.869504693716353e-11,
"loss": 2.13348650932312,
"step": 254
},
{
"epoch": 0.37372262773722625,
"grad_norm": 0.7052081823348999,
"learning_rate": 9.866019105895686e-11,
"loss": 1.8968867063522339,
"step": 256
},
{
"epoch": 0.37664233576642336,
"grad_norm": 0.47716858983039856,
"learning_rate": 9.862488281952992e-11,
"loss": 1.7656636238098145,
"step": 258
},
{
"epoch": 0.3795620437956204,
"grad_norm": 0.21338705718517303,
"learning_rate": 9.858912258470973e-11,
"loss": 1.604569435119629,
"step": 260
},
{
"epoch": 0.38248175182481753,
"grad_norm": 0.441703200340271,
"learning_rate": 9.855291072500643e-11,
"loss": 1.8019328117370605,
"step": 262
},
{
"epoch": 0.3854014598540146,
"grad_norm": 0.4217037558555603,
"learning_rate": 9.851624761560942e-11,
"loss": 2.2849860191345215,
"step": 264
},
{
"epoch": 0.3883211678832117,
"grad_norm": 0.358407586812973,
"learning_rate": 9.847913363638348e-11,
"loss": 1.927869200706482,
"step": 266
},
{
"epoch": 0.39124087591240875,
"grad_norm": 0.44805291295051575,
"learning_rate": 9.844156917186485e-11,
"loss": 1.5878663063049316,
"step": 268
},
{
"epoch": 0.39416058394160586,
"grad_norm": 0.5878186821937561,
"learning_rate": 9.840355461125717e-11,
"loss": 1.739422082901001,
"step": 270
},
{
"epoch": 0.3970802919708029,
"grad_norm": 0.3898945748806,
"learning_rate": 9.836509034842758e-11,
"loss": 1.6820441484451294,
"step": 272
},
{
"epoch": 0.4,
"grad_norm": 0.3018088638782501,
"learning_rate": 9.832617678190252e-11,
"loss": 1.6972731351852417,
"step": 274
},
{
"epoch": 0.4029197080291971,
"grad_norm": 0.3607556223869324,
"learning_rate": 9.828681431486365e-11,
"loss": 1.8021835088729858,
"step": 276
},
{
"epoch": 0.4058394160583942,
"grad_norm": 0.7184948921203613,
"learning_rate": 9.824700335514367e-11,
"loss": 1.8424593210220337,
"step": 278
},
{
"epoch": 0.40875912408759124,
"grad_norm": 0.9060298204421997,
"learning_rate": 9.820674431522208e-11,
"loss": 1.7280572652816772,
"step": 280
},
{
"epoch": 0.4116788321167883,
"grad_norm": 0.5940150022506714,
"learning_rate": 9.816603761222096e-11,
"loss": 1.8406977653503418,
"step": 282
},
{
"epoch": 0.4145985401459854,
"grad_norm": 0.5251798629760742,
"learning_rate": 9.812488366790056e-11,
"loss": 1.8819663524627686,
"step": 284
},
{
"epoch": 0.41751824817518246,
"grad_norm": 0.6706494688987732,
"learning_rate": 9.808328290865499e-11,
"loss": 1.6763763427734375,
"step": 286
},
{
"epoch": 0.42043795620437957,
"grad_norm": 0.4086778461933136,
"learning_rate": 9.80412357655078e-11,
"loss": 1.5377238988876343,
"step": 288
},
{
"epoch": 0.4233576642335766,
"grad_norm": 0.3526064455509186,
"learning_rate": 9.799874267410747e-11,
"loss": 1.6129189729690552,
"step": 290
},
{
"epoch": 0.42627737226277373,
"grad_norm": 0.32802364230155945,
"learning_rate": 9.7955804074723e-11,
"loss": 1.7900128364562988,
"step": 292
},
{
"epoch": 0.4291970802919708,
"grad_norm": 0.5461502075195312,
"learning_rate": 9.791242041223921e-11,
"loss": 1.7179615497589111,
"step": 294
},
{
"epoch": 0.4321167883211679,
"grad_norm": 0.3147372603416443,
"learning_rate": 9.786859213615222e-11,
"loss": 1.3112092018127441,
"step": 296
},
{
"epoch": 0.43503649635036495,
"grad_norm": 0.5890180468559265,
"learning_rate": 9.782431970056477e-11,
"loss": 1.637431025505066,
"step": 298
},
{
"epoch": 0.43795620437956206,
"grad_norm": 0.5663186311721802,
"learning_rate": 9.777960356418152e-11,
"loss": 1.6922191381454468,
"step": 300
},
{
"epoch": 0.4408759124087591,
"grad_norm": 0.27407780289649963,
"learning_rate": 9.773444419030429e-11,
"loss": 1.6063798666000366,
"step": 302
},
{
"epoch": 0.44379562043795623,
"grad_norm": 0.3785094618797302,
"learning_rate": 9.768884204682725e-11,
"loss": 1.7761635780334473,
"step": 304
},
{
"epoch": 0.4467153284671533,
"grad_norm": 0.29776790738105774,
"learning_rate": 9.764279760623213e-11,
"loss": 1.8873343467712402,
"step": 306
},
{
"epoch": 0.44963503649635034,
"grad_norm": 0.27831506729125977,
"learning_rate": 9.759631134558321e-11,
"loss": 1.5000873804092407,
"step": 308
},
{
"epoch": 0.45255474452554745,
"grad_norm": 0.3883137106895447,
"learning_rate": 9.754938374652254e-11,
"loss": 1.4682884216308594,
"step": 310
},
{
"epoch": 0.4554744525547445,
"grad_norm": 0.3289071321487427,
"learning_rate": 9.750201529526476e-11,
"loss": 1.616572380065918,
"step": 312
},
{
"epoch": 0.4583941605839416,
"grad_norm": 0.39953532814979553,
"learning_rate": 9.745420648259221e-11,
"loss": 1.5080758333206177,
"step": 314
},
{
"epoch": 0.46131386861313867,
"grad_norm": 0.4054333567619324,
"learning_rate": 9.74059578038498e-11,
"loss": 1.9431893825531006,
"step": 316
},
{
"epoch": 0.4642335766423358,
"grad_norm": 0.2948903441429138,
"learning_rate": 9.735726975893986e-11,
"loss": 1.6747829914093018,
"step": 318
},
{
"epoch": 0.46715328467153283,
"grad_norm": 0.2299513816833496,
"learning_rate": 9.730814285231695e-11,
"loss": 1.6036838293075562,
"step": 320
},
{
"epoch": 0.47007299270072994,
"grad_norm": 0.5000912547111511,
"learning_rate": 9.725857759298269e-11,
"loss": 1.6960393190383911,
"step": 322
},
{
"epoch": 0.472992700729927,
"grad_norm": 0.44810453057289124,
"learning_rate": 9.720857449448044e-11,
"loss": 1.5583330392837524,
"step": 324
},
{
"epoch": 0.4759124087591241,
"grad_norm": 0.3251849412918091,
"learning_rate": 9.715813407489001e-11,
"loss": 1.4970282316207886,
"step": 326
},
{
"epoch": 0.47883211678832116,
"grad_norm": 0.23649266362190247,
"learning_rate": 9.710725685682221e-11,
"loss": 1.4132741689682007,
"step": 328
},
{
"epoch": 0.48175182481751827,
"grad_norm": 0.34182292222976685,
"learning_rate": 9.70559433674136e-11,
"loss": 1.5327880382537842,
"step": 330
},
{
"epoch": 0.4846715328467153,
"grad_norm": 0.7462207078933716,
"learning_rate": 9.700419413832081e-11,
"loss": 1.5539097785949707,
"step": 332
},
{
"epoch": 0.48759124087591244,
"grad_norm": 0.23400592803955078,
"learning_rate": 9.695200970571525e-11,
"loss": 1.5931789875030518,
"step": 334
},
{
"epoch": 0.4905109489051095,
"grad_norm": 0.4315284192562103,
"learning_rate": 9.68993906102774e-11,
"loss": 1.7818516492843628,
"step": 336
},
{
"epoch": 0.49343065693430654,
"grad_norm": 0.5085203647613525,
"learning_rate": 9.684633739719126e-11,
"loss": 1.6964106559753418,
"step": 338
},
{
"epoch": 0.49635036496350365,
"grad_norm": 0.32666438817977905,
"learning_rate": 9.679285061613877e-11,
"loss": 1.8903117179870605,
"step": 340
},
{
"epoch": 0.4992700729927007,
"grad_norm": 0.696006715297699,
"learning_rate": 9.673893082129394e-11,
"loss": 1.6851402521133423,
"step": 342
},
{
"epoch": 0.5021897810218978,
"grad_norm": 0.3322364091873169,
"learning_rate": 9.668457857131728e-11,
"loss": 1.4909449815750122,
"step": 344
},
{
"epoch": 0.5051094890510949,
"grad_norm": 0.32292819023132324,
"learning_rate": 9.662979442934995e-11,
"loss": 1.3091050386428833,
"step": 346
},
{
"epoch": 0.5080291970802919,
"grad_norm": 0.7607306838035583,
"learning_rate": 9.657457896300791e-11,
"loss": 1.4538644552230835,
"step": 348
},
{
"epoch": 0.5109489051094891,
"grad_norm": 0.8117560744285583,
"learning_rate": 9.651893274437606e-11,
"loss": 1.715744137763977,
"step": 350
},
{
"epoch": 0.5138686131386861,
"grad_norm": 0.42647185921669006,
"learning_rate": 9.646285635000226e-11,
"loss": 1.8959312438964844,
"step": 352
},
{
"epoch": 0.5167883211678832,
"grad_norm": 0.36822986602783203,
"learning_rate": 9.640635036089149e-11,
"loss": 1.9616965055465698,
"step": 354
},
{
"epoch": 0.5197080291970803,
"grad_norm": 0.49794426560401917,
"learning_rate": 9.634941536249966e-11,
"loss": 1.6823753118515015,
"step": 356
},
{
"epoch": 0.5226277372262774,
"grad_norm": 0.5548354983329773,
"learning_rate": 9.629205194472766e-11,
"loss": 1.640671730041504,
"step": 358
},
{
"epoch": 0.5255474452554745,
"grad_norm": 0.5667994618415833,
"learning_rate": 9.623426070191521e-11,
"loss": 1.4344934225082397,
"step": 360
},
{
"epoch": 0.5284671532846715,
"grad_norm": 0.6978429555892944,
"learning_rate": 9.617604223283472e-11,
"loss": 1.8851178884506226,
"step": 362
},
{
"epoch": 0.5313868613138686,
"grad_norm": 3.7346668243408203,
"learning_rate": 9.611739714068503e-11,
"loss": 2.039613723754883,
"step": 364
},
{
"epoch": 0.5343065693430656,
"grad_norm": 4.44684362411499,
"learning_rate": 9.605832603308523e-11,
"loss": 1.6009081602096558,
"step": 366
},
{
"epoch": 0.5372262773722628,
"grad_norm": 0.8611459136009216,
"learning_rate": 9.599882952206834e-11,
"loss": 1.4058667421340942,
"step": 368
},
{
"epoch": 0.5401459854014599,
"grad_norm": 0.3514321744441986,
"learning_rate": 9.593890822407497e-11,
"loss": 1.6223829984664917,
"step": 370
},
{
"epoch": 0.5430656934306569,
"grad_norm": 0.7854035496711731,
"learning_rate": 9.58785627599469e-11,
"loss": 1.837151050567627,
"step": 372
},
{
"epoch": 0.545985401459854,
"grad_norm": 0.7137376666069031,
"learning_rate": 9.58177937549207e-11,
"loss": 1.902309536933899,
"step": 374
},
{
"epoch": 0.5489051094890511,
"grad_norm": 0.217190682888031,
"learning_rate": 9.575660183862122e-11,
"loss": 1.6437760591506958,
"step": 376
},
{
"epoch": 0.5518248175182482,
"grad_norm": 0.5008941292762756,
"learning_rate": 9.569498764505507e-11,
"loss": 1.7193173170089722,
"step": 378
},
{
"epoch": 0.5547445255474452,
"grad_norm": 0.3317372500896454,
"learning_rate": 9.563295181260408e-11,
"loss": 1.4700264930725098,
"step": 380
},
{
"epoch": 0.5576642335766423,
"grad_norm": 0.24804271757602692,
"learning_rate": 9.55704949840186e-11,
"loss": 1.3685392141342163,
"step": 382
},
{
"epoch": 0.5605839416058395,
"grad_norm": 0.5515373349189758,
"learning_rate": 9.550761780641101e-11,
"loss": 1.272115707397461,
"step": 384
},
{
"epoch": 0.5635036496350365,
"grad_norm": 0.8238644003868103,
"learning_rate": 9.544432093124877e-11,
"loss": 1.5264108180999756,
"step": 386
},
{
"epoch": 0.5664233576642336,
"grad_norm": 0.6252793073654175,
"learning_rate": 9.53806050143479e-11,
"loss": 1.3205806016921997,
"step": 388
},
{
"epoch": 0.5693430656934306,
"grad_norm": 0.5369510054588318,
"learning_rate": 9.531647071586608e-11,
"loss": 1.6523046493530273,
"step": 390
},
{
"epoch": 0.5722627737226277,
"grad_norm": 0.5264310836791992,
"learning_rate": 9.52519187002958e-11,
"loss": 1.37351393699646,
"step": 392
},
{
"epoch": 0.5751824817518248,
"grad_norm": 0.8609051704406738,
"learning_rate": 9.518694963645751e-11,
"loss": 1.3692678213119507,
"step": 394
},
{
"epoch": 0.5781021897810219,
"grad_norm": 0.28163954615592957,
"learning_rate": 9.512156419749268e-11,
"loss": 1.3704659938812256,
"step": 396
},
{
"epoch": 0.581021897810219,
"grad_norm": 0.4875539541244507,
"learning_rate": 9.505576306085681e-11,
"loss": 1.4559335708618164,
"step": 398
},
{
"epoch": 0.583941605839416,
"grad_norm": 0.3505980372428894,
"learning_rate": 9.498954690831245e-11,
"loss": 1.6419453620910645,
"step": 400
},
{
"epoch": 0.5868613138686132,
"grad_norm": 0.5778207182884216,
"learning_rate": 9.492291642592205e-11,
"loss": 1.2554142475128174,
"step": 402
},
{
"epoch": 0.5897810218978102,
"grad_norm": 0.24265649914741516,
"learning_rate": 9.485587230404102e-11,
"loss": 1.2532234191894531,
"step": 404
},
{
"epoch": 0.5927007299270073,
"grad_norm": 0.6797131896018982,
"learning_rate": 9.478841523731036e-11,
"loss": 1.2978441715240479,
"step": 406
},
{
"epoch": 0.5956204379562043,
"grad_norm": 0.41169965267181396,
"learning_rate": 9.472054592464963e-11,
"loss": 1.3374297618865967,
"step": 408
},
{
"epoch": 0.5985401459854015,
"grad_norm": 0.5518661737442017,
"learning_rate": 9.465226506924965e-11,
"loss": 1.652148723602295,
"step": 410
},
{
"epoch": 0.6014598540145986,
"grad_norm": 0.1583075225353241,
"learning_rate": 9.458357337856519e-11,
"loss": 1.443751335144043,
"step": 412
},
{
"epoch": 0.6043795620437956,
"grad_norm": 0.34667131304740906,
"learning_rate": 9.451447156430769e-11,
"loss": 1.2133233547210693,
"step": 414
},
{
"epoch": 0.6072992700729927,
"grad_norm": 0.503765344619751,
"learning_rate": 9.444496034243786e-11,
"loss": 1.3155429363250732,
"step": 416
},
{
"epoch": 0.6102189781021898,
"grad_norm": 0.43029820919036865,
"learning_rate": 9.437504043315824e-11,
"loss": 1.536086082458496,
"step": 418
},
{
"epoch": 0.6131386861313869,
"grad_norm": 1.2249265909194946,
"learning_rate": 9.430471256090578e-11,
"loss": 1.691704511642456,
"step": 420
},
{
"epoch": 0.6160583941605839,
"grad_norm": 0.25307124853134155,
"learning_rate": 9.42339774543443e-11,
"loss": 1.1645077466964722,
"step": 422
},
{
"epoch": 0.618978102189781,
"grad_norm": 0.36436426639556885,
"learning_rate": 9.4162835846357e-11,
"loss": 1.5609655380249023,
"step": 424
},
{
"epoch": 0.621897810218978,
"grad_norm": 0.26339077949523926,
"learning_rate": 9.409128847403874e-11,
"loss": 1.4671318531036377,
"step": 426
},
{
"epoch": 0.6248175182481752,
"grad_norm": 0.4291604459285736,
"learning_rate": 9.401933607868859e-11,
"loss": 1.2365825176239014,
"step": 428
},
{
"epoch": 0.6277372262773723,
"grad_norm": 0.24072806537151337,
"learning_rate": 9.394697940580195e-11,
"loss": 1.2259232997894287,
"step": 430
},
{
"epoch": 0.6306569343065693,
"grad_norm": 0.38607171177864075,
"learning_rate": 9.387421920506298e-11,
"loss": 1.1846305131912231,
"step": 432
},
{
"epoch": 0.6335766423357664,
"grad_norm": 0.27039477229118347,
"learning_rate": 9.380105623033677e-11,
"loss": 1.6850125789642334,
"step": 434
},
{
"epoch": 0.6364963503649635,
"grad_norm": 0.27692389488220215,
"learning_rate": 9.372749123966148e-11,
"loss": 1.5691750049591064,
"step": 436
},
{
"epoch": 0.6394160583941606,
"grad_norm": 0.506200909614563,
"learning_rate": 9.365352499524059e-11,
"loss": 1.6711288690567017,
"step": 438
},
{
"epoch": 0.6423357664233577,
"grad_norm": 0.2726089656352997,
"learning_rate": 9.357915826343496e-11,
"loss": 1.3896371126174927,
"step": 440
},
{
"epoch": 0.6452554744525547,
"grad_norm": 0.5812091827392578,
"learning_rate": 9.350439181475483e-11,
"loss": 1.6031279563903809,
"step": 442
},
{
"epoch": 0.6481751824817519,
"grad_norm": 0.27857470512390137,
"learning_rate": 9.342922642385193e-11,
"loss": 1.507469892501831,
"step": 444
},
{
"epoch": 0.6510948905109489,
"grad_norm": 0.7666060924530029,
"learning_rate": 9.33536628695114e-11,
"loss": 1.5827025175094604,
"step": 446
},
{
"epoch": 0.654014598540146,
"grad_norm": 0.716264545917511,
"learning_rate": 9.327770193464374e-11,
"loss": 1.3550753593444824,
"step": 448
},
{
"epoch": 0.656934306569343,
"grad_norm": 0.3254503309726715,
"learning_rate": 9.320134440627665e-11,
"loss": 1.4587416648864746,
"step": 450
},
{
"epoch": 0.6598540145985401,
"grad_norm": 0.4242144823074341,
"learning_rate": 9.312459107554698e-11,
"loss": 1.5037614107131958,
"step": 452
},
{
"epoch": 0.6627737226277373,
"grad_norm": 0.2750391364097595,
"learning_rate": 9.304744273769243e-11,
"loss": 1.6231236457824707,
"step": 454
},
{
"epoch": 0.6656934306569343,
"grad_norm": 0.4245796501636505,
"learning_rate": 9.296990019204335e-11,
"loss": 1.3407589197158813,
"step": 456
},
{
"epoch": 0.6686131386861314,
"grad_norm": 0.32759228348731995,
"learning_rate": 9.289196424201451e-11,
"loss": 1.3652153015136719,
"step": 458
},
{
"epoch": 0.6715328467153284,
"grad_norm": 0.5771019458770752,
"learning_rate": 9.281363569509662e-11,
"loss": 1.6398756504058838,
"step": 460
},
{
"epoch": 0.6744525547445256,
"grad_norm": 0.7310810089111328,
"learning_rate": 9.273491536284819e-11,
"loss": 1.7764878273010254,
"step": 462
},
{
"epoch": 0.6773722627737226,
"grad_norm": 1.0177778005599976,
"learning_rate": 9.265580406088692e-11,
"loss": 1.4794597625732422,
"step": 464
},
{
"epoch": 0.6802919708029197,
"grad_norm": 0.23670396208763123,
"learning_rate": 9.257630260888133e-11,
"loss": 1.2997967004776,
"step": 466
},
{
"epoch": 0.6832116788321168,
"grad_norm": 0.4635494649410248,
"learning_rate": 9.249641183054232e-11,
"loss": 1.4545753002166748,
"step": 468
},
{
"epoch": 0.6861313868613139,
"grad_norm": 0.5670157670974731,
"learning_rate": 9.241613255361454e-11,
"loss": 1.6660820245742798,
"step": 470
},
{
"epoch": 0.689051094890511,
"grad_norm": 0.27268993854522705,
"learning_rate": 9.233546560986786e-11,
"loss": 1.6526291370391846,
"step": 472
},
{
"epoch": 0.691970802919708,
"grad_norm": 1.366753339767456,
"learning_rate": 9.225441183508875e-11,
"loss": 1.6750726699829102,
"step": 474
},
{
"epoch": 0.6948905109489051,
"grad_norm": 0.32852044701576233,
"learning_rate": 9.217297206907161e-11,
"loss": 1.4638712406158447,
"step": 476
},
{
"epoch": 0.6978102189781021,
"grad_norm": 0.373755544424057,
"learning_rate": 9.209114715561012e-11,
"loss": 1.6217398643493652,
"step": 478
},
{
"epoch": 0.7007299270072993,
"grad_norm": 0.287253737449646,
"learning_rate": 9.200893794248843e-11,
"loss": 1.5670347213745117,
"step": 480
},
{
"epoch": 0.7036496350364964,
"grad_norm": 0.5298172831535339,
"learning_rate": 9.192634528147234e-11,
"loss": 1.6435707807540894,
"step": 482
},
{
"epoch": 0.7065693430656934,
"grad_norm": 0.23429054021835327,
"learning_rate": 9.184337002830065e-11,
"loss": 1.3780916929244995,
"step": 484
},
{
"epoch": 0.7094890510948905,
"grad_norm": 0.4721639156341553,
"learning_rate": 9.176001304267607e-11,
"loss": 1.7100255489349365,
"step": 486
},
{
"epoch": 0.7124087591240876,
"grad_norm": 0.32687824964523315,
"learning_rate": 9.167627518825651e-11,
"loss": 1.5640754699707031,
"step": 488
},
{
"epoch": 0.7153284671532847,
"grad_norm": 0.4479893147945404,
"learning_rate": 9.1592157332646e-11,
"loss": 1.6100982427597046,
"step": 490
},
{
"epoch": 0.7182481751824817,
"grad_norm": 0.3041948080062866,
"learning_rate": 9.150766034738575e-11,
"loss": 1.625577449798584,
"step": 492
},
{
"epoch": 0.7211678832116788,
"grad_norm": 0.523112952709198,
"learning_rate": 9.14227851079451e-11,
"loss": 1.5553128719329834,
"step": 494
},
{
"epoch": 0.724087591240876,
"grad_norm": 0.37522390484809875,
"learning_rate": 9.133753249371251e-11,
"loss": 1.6051119565963745,
"step": 496
},
{
"epoch": 0.727007299270073,
"grad_norm": 0.5746657252311707,
"learning_rate": 9.125190338798634e-11,
"loss": 1.5265920162200928,
"step": 498
},
{
"epoch": 0.7299270072992701,
"grad_norm": 0.48165905475616455,
"learning_rate": 9.116589867796582e-11,
"loss": 1.7097959518432617,
"step": 500
},
{
"epoch": 0.7328467153284671,
"grad_norm": 1.95722234249115,
"learning_rate": 9.107951925474176e-11,
"loss": 1.5973304510116577,
"step": 502
},
{
"epoch": 0.7357664233576642,
"grad_norm": 0.4828181266784668,
"learning_rate": 9.099276601328738e-11,
"loss": 1.4544634819030762,
"step": 504
},
{
"epoch": 0.7386861313868613,
"grad_norm": 0.35941943526268005,
"learning_rate": 9.090563985244899e-11,
"loss": 1.5493344068527222,
"step": 506
},
{
"epoch": 0.7416058394160584,
"grad_norm": 0.4957440197467804,
"learning_rate": 9.081814167493671e-11,
"loss": 1.6227343082427979,
"step": 508
},
{
"epoch": 0.7445255474452555,
"grad_norm": 0.8167078495025635,
"learning_rate": 9.073027238731513e-11,
"loss": 1.6121728420257568,
"step": 510
},
{
"epoch": 0.7474452554744525,
"grad_norm": 0.2773600220680237,
"learning_rate": 9.064203289999385e-11,
"loss": 1.6567003726959229,
"step": 512
},
{
"epoch": 0.7503649635036497,
"grad_norm": 0.38320213556289673,
"learning_rate": 9.055342412721815e-11,
"loss": 1.559574842453003,
"step": 514
},
{
"epoch": 0.7532846715328467,
"grad_norm": 0.2601078450679779,
"learning_rate": 9.04644469870594e-11,
"loss": 1.634887456893921,
"step": 516
},
{
"epoch": 0.7562043795620438,
"grad_norm": 0.3223051428794861,
"learning_rate": 9.037510240140563e-11,
"loss": 1.5338331460952759,
"step": 518
},
{
"epoch": 0.7591240875912408,
"grad_norm": 0.6172982454299927,
"learning_rate": 9.028539129595198e-11,
"loss": 1.6613255739212036,
"step": 520
},
{
"epoch": 0.762043795620438,
"grad_norm": 0.2970709204673767,
"learning_rate": 9.019531460019105e-11,
"loss": 1.66960608959198,
"step": 522
},
{
"epoch": 0.7649635036496351,
"grad_norm": 0.2917795479297638,
"learning_rate": 9.010487324740333e-11,
"loss": 1.5973072052001953,
"step": 524
},
{
"epoch": 0.7678832116788321,
"grad_norm": 0.40561556816101074,
"learning_rate": 9.001406817464748e-11,
"loss": 1.8639005422592163,
"step": 526
},
{
"epoch": 0.7708029197080292,
"grad_norm": 0.4241611659526825,
"learning_rate": 8.992290032275067e-11,
"loss": 1.574570894241333,
"step": 528
},
{
"epoch": 0.7737226277372263,
"grad_norm": 0.3406296670436859,
"learning_rate": 8.983137063629879e-11,
"loss": 1.5100054740905762,
"step": 530
},
{
"epoch": 0.7766423357664234,
"grad_norm": 0.38461536169052124,
"learning_rate": 8.97394800636267e-11,
"loss": 1.744242787361145,
"step": 532
},
{
"epoch": 0.7795620437956204,
"grad_norm": 0.37936046719551086,
"learning_rate": 8.964722955680835e-11,
"loss": 1.5281338691711426,
"step": 534
},
{
"epoch": 0.7824817518248175,
"grad_norm": 0.42358264327049255,
"learning_rate": 8.955462007164699e-11,
"loss": 1.5278962850570679,
"step": 536
},
{
"epoch": 0.7854014598540145,
"grad_norm": 0.5687382817268372,
"learning_rate": 8.946165256766524e-11,
"loss": 1.286256194114685,
"step": 538
},
{
"epoch": 0.7883211678832117,
"grad_norm": 0.34203967452049255,
"learning_rate": 8.936832800809506e-11,
"loss": 1.2884188890457153,
"step": 540
},
{
"epoch": 0.7912408759124088,
"grad_norm": 0.33001580834388733,
"learning_rate": 8.927464735986792e-11,
"loss": 1.4524775743484497,
"step": 542
},
{
"epoch": 0.7941605839416058,
"grad_norm": 0.24470959603786469,
"learning_rate": 8.918061159360468e-11,
"loss": 1.5741451978683472,
"step": 544
},
{
"epoch": 0.7970802919708029,
"grad_norm": 0.22205418348312378,
"learning_rate": 8.908622168360558e-11,
"loss": 1.415770411491394,
"step": 546
},
{
"epoch": 0.8,
"grad_norm": 0.12828628718852997,
"learning_rate": 8.899147860784006e-11,
"loss": 1.2092006206512451,
"step": 548
},
{
"epoch": 0.8029197080291971,
"grad_norm": 0.19067053496837616,
"learning_rate": 8.889638334793682e-11,
"loss": 1.2418055534362793,
"step": 550
},
{
"epoch": 0.8058394160583942,
"grad_norm": 0.2096225470304489,
"learning_rate": 8.88009368891734e-11,
"loss": 1.150226354598999,
"step": 552
},
{
"epoch": 0.8087591240875912,
"grad_norm": 0.23198847472667694,
"learning_rate": 8.870514022046617e-11,
"loss": 1.1433364152908325,
"step": 554
},
{
"epoch": 0.8116788321167884,
"grad_norm": 0.23092319071292877,
"learning_rate": 8.860899433435994e-11,
"loss": 1.0709772109985352,
"step": 556
},
{
"epoch": 0.8145985401459854,
"grad_norm": 1.7060472965240479,
"learning_rate": 8.851250022701784e-11,
"loss": 1.0478689670562744,
"step": 558
},
{
"epoch": 0.8175182481751825,
"grad_norm": 0.20169571042060852,
"learning_rate": 8.841565889821084e-11,
"loss": 1.0691083669662476,
"step": 560
},
{
"epoch": 0.8204379562043795,
"grad_norm": 0.2585836946964264,
"learning_rate": 8.831847135130746e-11,
"loss": 1.0405468940734863,
"step": 562
},
{
"epoch": 0.8233576642335766,
"grad_norm": 0.15813350677490234,
"learning_rate": 8.822093859326339e-11,
"loss": 1.119150996208191,
"step": 564
},
{
"epoch": 0.8262773722627738,
"grad_norm": 0.2605457007884979,
"learning_rate": 8.8123061634611e-11,
"loss": 1.451040506362915,
"step": 566
},
{
"epoch": 0.8291970802919708,
"grad_norm": 0.29428720474243164,
"learning_rate": 8.802484148944897e-11,
"loss": 1.498022198677063,
"step": 568
},
{
"epoch": 0.8321167883211679,
"grad_norm": 0.32284432649612427,
"learning_rate": 8.792627917543165e-11,
"loss": 1.5406850576400757,
"step": 570
},
{
"epoch": 0.8350364963503649,
"grad_norm": 0.2935868799686432,
"learning_rate": 8.782737571375861e-11,
"loss": 1.4694265127182007,
"step": 572
},
{
"epoch": 0.8379562043795621,
"grad_norm": 0.1954089254140854,
"learning_rate": 8.772813212916408e-11,
"loss": 1.497194528579712,
"step": 574
},
{
"epoch": 0.8408759124087591,
"grad_norm": 0.29668334126472473,
"learning_rate": 8.762854944990623e-11,
"loss": 1.4351569414138794,
"step": 576
},
{
"epoch": 0.8437956204379562,
"grad_norm": 0.3149370849132538,
"learning_rate": 8.752862870775658e-11,
"loss": 1.4598360061645508,
"step": 578
},
{
"epoch": 0.8467153284671532,
"grad_norm": 0.3693007528781891,
"learning_rate": 8.742837093798934e-11,
"loss": 1.5218864679336548,
"step": 580
},
{
"epoch": 0.8496350364963504,
"grad_norm": 0.27900761365890503,
"learning_rate": 8.732777717937063e-11,
"loss": 1.4627201557159424,
"step": 582
},
{
"epoch": 0.8525547445255475,
"grad_norm": 0.303118497133255,
"learning_rate": 8.72268484741477e-11,
"loss": 1.4716007709503174,
"step": 584
},
{
"epoch": 0.8554744525547445,
"grad_norm": 0.3150079846382141,
"learning_rate": 8.712558586803822e-11,
"loss": 1.4263559579849243,
"step": 586
},
{
"epoch": 0.8583941605839416,
"grad_norm": 0.3285709619522095,
"learning_rate": 8.702399041021937e-11,
"loss": 1.4390225410461426,
"step": 588
},
{
"epoch": 0.8613138686131386,
"grad_norm": 0.786037266254425,
"learning_rate": 8.692206315331697e-11,
"loss": 1.4343092441558838,
"step": 590
},
{
"epoch": 0.8642335766423358,
"grad_norm": 0.2554452419281006,
"learning_rate": 8.681980515339464e-11,
"loss": 1.3966456651687622,
"step": 592
},
{
"epoch": 0.8671532846715329,
"grad_norm": 0.3106688857078552,
"learning_rate": 8.671721746994278e-11,
"loss": 1.47050142288208,
"step": 594
},
{
"epoch": 0.8700729927007299,
"grad_norm": 0.36648714542388916,
"learning_rate": 8.661430116586763e-11,
"loss": 1.4842816591262817,
"step": 596
},
{
"epoch": 0.872992700729927,
"grad_norm": 0.2672443985939026,
"learning_rate": 8.651105730748023e-11,
"loss": 1.5009945631027222,
"step": 598
},
{
"epoch": 0.8759124087591241,
"grad_norm": 0.3231627345085144,
"learning_rate": 8.640748696448546e-11,
"loss": 1.4662688970565796,
"step": 600
},
{
"epoch": 0.8788321167883212,
"grad_norm": 0.447077214717865,
"learning_rate": 8.630359120997084e-11,
"loss": 1.4902161359786987,
"step": 602
},
{
"epoch": 0.8817518248175182,
"grad_norm": 1.3244692087173462,
"learning_rate": 8.61993711203955e-11,
"loss": 1.4570201635360718,
"step": 604
},
{
"epoch": 0.8846715328467153,
"grad_norm": 0.323764443397522,
"learning_rate": 8.609482777557891e-11,
"loss": 1.4848111867904663,
"step": 606
},
{
"epoch": 0.8875912408759125,
"grad_norm": 0.21813516318798065,
"learning_rate": 8.59899622586899e-11,
"loss": 1.410193920135498,
"step": 608
},
{
"epoch": 0.8905109489051095,
"grad_norm": 0.4606623947620392,
"learning_rate": 8.58847756562352e-11,
"loss": 1.4347537755966187,
"step": 610
},
{
"epoch": 0.8934306569343066,
"grad_norm": 0.2534776031970978,
"learning_rate": 8.577926905804833e-11,
"loss": 1.4541630744934082,
"step": 612
},
{
"epoch": 0.8963503649635036,
"grad_norm": 0.24322006106376648,
"learning_rate": 8.567344355727829e-11,
"loss": 1.432068943977356,
"step": 614
},
{
"epoch": 0.8992700729927007,
"grad_norm": 0.21942275762557983,
"learning_rate": 8.556730025037819e-11,
"loss": 1.4721245765686035,
"step": 616
},
{
"epoch": 0.9021897810218978,
"grad_norm": 0.32879260182380676,
"learning_rate": 8.546084023709393e-11,
"loss": 1.4534900188446045,
"step": 618
},
{
"epoch": 0.9051094890510949,
"grad_norm": 0.33828699588775635,
"learning_rate": 8.535406462045279e-11,
"loss": 1.4511345624923706,
"step": 620
},
{
"epoch": 0.908029197080292,
"grad_norm": 0.25151824951171875,
"learning_rate": 8.524697450675199e-11,
"loss": 1.4202616214752197,
"step": 622
},
{
"epoch": 0.910948905109489,
"grad_norm": 0.4613194465637207,
"learning_rate": 8.513957100554721e-11,
"loss": 1.4088406562805176,
"step": 624
},
{
"epoch": 0.9138686131386862,
"grad_norm": 0.2900555431842804,
"learning_rate": 8.503185522964122e-11,
"loss": 1.4401381015777588,
"step": 626
},
{
"epoch": 0.9167883211678832,
"grad_norm": 0.2588656544685364,
"learning_rate": 8.492382829507215e-11,
"loss": 1.497382402420044,
"step": 628
},
{
"epoch": 0.9197080291970803,
"grad_norm": 0.36505982279777527,
"learning_rate": 8.48154913211021e-11,
"loss": 1.4394062757492065,
"step": 630
},
{
"epoch": 0.9226277372262773,
"grad_norm": 0.23443567752838135,
"learning_rate": 8.470684543020542e-11,
"loss": 1.4333170652389526,
"step": 632
},
{
"epoch": 0.9255474452554745,
"grad_norm": 0.2820832133293152,
"learning_rate": 8.45978917480572e-11,
"loss": 1.3851203918457031,
"step": 634
},
{
"epoch": 0.9284671532846716,
"grad_norm": 0.30184608697891235,
"learning_rate": 8.448863140352144e-11,
"loss": 1.4456335306167603,
"step": 636
},
{
"epoch": 0.9313868613138686,
"grad_norm": 0.3142789304256439,
"learning_rate": 8.437906552863958e-11,
"loss": 1.450710654258728,
"step": 638
},
{
"epoch": 0.9343065693430657,
"grad_norm": 0.4998232424259186,
"learning_rate": 8.426919525861854e-11,
"loss": 1.4164352416992188,
"step": 640
},
{
"epoch": 0.9372262773722628,
"grad_norm": 0.2807002365589142,
"learning_rate": 8.41590217318191e-11,
"loss": 1.4481756687164307,
"step": 642
},
{
"epoch": 0.9401459854014599,
"grad_norm": 0.3465396463871002,
"learning_rate": 8.404854608974414e-11,
"loss": 1.4438577890396118,
"step": 644
},
{
"epoch": 0.9430656934306569,
"grad_norm": 0.2985764145851135,
"learning_rate": 8.39377694770266e-11,
"loss": 1.5020420551300049,
"step": 646
},
{
"epoch": 0.945985401459854,
"grad_norm": 0.3235088884830475,
"learning_rate": 8.38266930414179e-11,
"loss": 1.4755083322525024,
"step": 648
},
{
"epoch": 0.948905109489051,
"grad_norm": 0.23725037276744843,
"learning_rate": 8.37153179337758e-11,
"loss": 1.4602628946304321,
"step": 650
},
{
"epoch": 0.9518248175182482,
"grad_norm": 0.2622622847557068,
"learning_rate": 8.360364530805266e-11,
"loss": 1.4472800493240356,
"step": 652
},
{
"epoch": 0.9547445255474453,
"grad_norm": 0.662328839302063,
"learning_rate": 8.349167632128342e-11,
"loss": 1.4257826805114746,
"step": 654
},
{
"epoch": 0.9576642335766423,
"grad_norm": 0.3589712083339691,
"learning_rate": 8.337941213357355e-11,
"loss": 1.384074330329895,
"step": 656
},
{
"epoch": 0.9605839416058394,
"grad_norm": 0.2395324409008026,
"learning_rate": 8.326685390808711e-11,
"loss": 1.3845038414001465,
"step": 658
},
{
"epoch": 0.9635036496350365,
"grad_norm": 0.29251864552497864,
"learning_rate": 8.315400281103466e-11,
"loss": 1.3974683284759521,
"step": 660
},
{
"epoch": 0.9664233576642336,
"grad_norm": 0.3395976126194,
"learning_rate": 8.304086001166119e-11,
"loss": 1.4027167558670044,
"step": 662
},
{
"epoch": 0.9693430656934306,
"grad_norm": 0.29036200046539307,
"learning_rate": 8.292742668223403e-11,
"loss": 1.4590882062911987,
"step": 664
},
{
"epoch": 0.9722627737226277,
"grad_norm": 0.3950836658477783,
"learning_rate": 8.281370399803065e-11,
"loss": 1.4444341659545898,
"step": 666
},
{
"epoch": 0.9751824817518249,
"grad_norm": 0.2706378400325775,
"learning_rate": 8.269969313732654e-11,
"loss": 1.4351338148117065,
"step": 668
},
{
"epoch": 0.9781021897810219,
"grad_norm": 0.45173388719558716,
"learning_rate": 8.258539528138293e-11,
"loss": 1.4339206218719482,
"step": 670
},
{
"epoch": 0.981021897810219,
"grad_norm": 0.25517186522483826,
"learning_rate": 8.247081161443463e-11,
"loss": 1.4406836032867432,
"step": 672
},
{
"epoch": 0.983941605839416,
"grad_norm": 0.23826317489147186,
"learning_rate": 8.23559433236777e-11,
"loss": 1.3740425109863281,
"step": 674
},
{
"epoch": 0.9868613138686131,
"grad_norm": 0.37786585092544556,
"learning_rate": 8.224079159925721e-11,
"loss": 1.4657270908355713,
"step": 676
},
{
"epoch": 0.9897810218978103,
"grad_norm": 0.34262749552726746,
"learning_rate": 8.212535763425484e-11,
"loss": 1.50901460647583,
"step": 678
},
{
"epoch": 0.9927007299270073,
"grad_norm": 0.505266547203064,
"learning_rate": 8.200964262467657e-11,
"loss": 1.7625021934509277,
"step": 680
},
{
"epoch": 0.9956204379562044,
"grad_norm": 0.4917811453342438,
"learning_rate": 8.189364776944026e-11,
"loss": 2.1940176486968994,
"step": 682
},
{
"epoch": 0.9985401459854014,
"grad_norm": 0.4079400300979614,
"learning_rate": 8.177737427036323e-11,
"loss": 2.142901659011841,
"step": 684
},
{
"epoch": 1.0014598540145985,
"grad_norm": 0.32731911540031433,
"learning_rate": 8.166082333214983e-11,
"loss": 1.792697548866272,
"step": 686
},
{
"epoch": 1.0043795620437956,
"grad_norm": 0.3955923020839691,
"learning_rate": 8.154399616237894e-11,
"loss": 1.509110450744629,
"step": 688
},
{
"epoch": 1.0072992700729928,
"grad_norm": 0.23040631413459778,
"learning_rate": 8.142689397149148e-11,
"loss": 1.4758418798446655,
"step": 690
},
{
"epoch": 1.0102189781021897,
"grad_norm": 0.23065227270126343,
"learning_rate": 8.130951797277783e-11,
"loss": 1.4547432661056519,
"step": 692
},
{
"epoch": 1.013138686131387,
"grad_norm": 0.27488207817077637,
"learning_rate": 8.119186938236524e-11,
"loss": 1.5417169332504272,
"step": 694
},
{
"epoch": 1.0160583941605839,
"grad_norm": 0.3516688644886017,
"learning_rate": 8.107394941920536e-11,
"loss": 1.4036191701889038,
"step": 696
},
{
"epoch": 1.018978102189781,
"grad_norm": 0.641706645488739,
"learning_rate": 8.095575930506148e-11,
"loss": 1.498883605003357,
"step": 698
},
{
"epoch": 1.0218978102189782,
"grad_norm": 0.2665751874446869,
"learning_rate": 8.083730026449588e-11,
"loss": 1.324239730834961,
"step": 700
},
{
"epoch": 1.0248175182481751,
"grad_norm": 0.18407660722732544,
"learning_rate": 8.07185735248572e-11,
"loss": 1.3473355770111084,
"step": 702
},
{
"epoch": 1.0277372262773723,
"grad_norm": 0.32454606890678406,
"learning_rate": 8.059958031626771e-11,
"loss": 1.3403379917144775,
"step": 704
},
{
"epoch": 1.0306569343065692,
"grad_norm": 0.3869016468524933,
"learning_rate": 8.048032187161055e-11,
"loss": 1.4723657369613647,
"step": 706
},
{
"epoch": 1.0335766423357664,
"grad_norm": 0.2613857090473175,
"learning_rate": 8.036079942651694e-11,
"loss": 1.4147893190383911,
"step": 708
},
{
"epoch": 1.0364963503649636,
"grad_norm": 0.19323043525218964,
"learning_rate": 8.02410142193534e-11,
"loss": 1.2198904752731323,
"step": 710
},
{
"epoch": 1.0394160583941605,
"grad_norm": 0.28235873579978943,
"learning_rate": 8.01209674912089e-11,
"loss": 1.4131911993026733,
"step": 712
},
{
"epoch": 1.0423357664233577,
"grad_norm": 0.2318020910024643,
"learning_rate": 8.00006604858821e-11,
"loss": 1.516754388809204,
"step": 714
},
{
"epoch": 1.0452554744525548,
"grad_norm": 0.4528334438800812,
"learning_rate": 7.988009444986827e-11,
"loss": 1.4079957008361816,
"step": 716
},
{
"epoch": 1.0481751824817518,
"grad_norm": 0.2093593031167984,
"learning_rate": 7.975927063234655e-11,
"loss": 1.4088261127471924,
"step": 718
},
{
"epoch": 1.051094890510949,
"grad_norm": 0.16769030690193176,
"learning_rate": 7.963819028516691e-11,
"loss": 1.084653377532959,
"step": 720
},
{
"epoch": 1.054014598540146,
"grad_norm": 0.3169795870780945,
"learning_rate": 7.951685466283724e-11,
"loss": 1.2415783405303955,
"step": 722
},
{
"epoch": 1.056934306569343,
"grad_norm": 0.3813137114048004,
"learning_rate": 7.93952650225103e-11,
"loss": 1.4431004524230957,
"step": 724
},
{
"epoch": 1.0598540145985402,
"grad_norm": 0.3302343189716339,
"learning_rate": 7.927342262397074e-11,
"loss": 1.3890019655227661,
"step": 726
},
{
"epoch": 1.0627737226277372,
"grad_norm": 0.2482709437608719,
"learning_rate": 7.915132872962204e-11,
"loss": 1.4420756101608276,
"step": 728
},
{
"epoch": 1.0656934306569343,
"grad_norm": 0.3268437683582306,
"learning_rate": 7.902898460447333e-11,
"loss": 1.5538722276687622,
"step": 730
},
{
"epoch": 1.0686131386861315,
"grad_norm": 0.45657190680503845,
"learning_rate": 7.890639151612647e-11,
"loss": 1.4880216121673584,
"step": 732
},
{
"epoch": 1.0715328467153284,
"grad_norm": 0.24117252230644226,
"learning_rate": 7.878355073476282e-11,
"loss": 1.3895787000656128,
"step": 734
},
{
"epoch": 1.0744525547445256,
"grad_norm": 0.25678950548171997,
"learning_rate": 7.866046353312994e-11,
"loss": 1.428212285041809,
"step": 736
},
{
"epoch": 1.0773722627737226,
"grad_norm": 0.2189822942018509,
"learning_rate": 7.853713118652872e-11,
"loss": 1.4189248085021973,
"step": 738
},
{
"epoch": 1.0802919708029197,
"grad_norm": 0.30409157276153564,
"learning_rate": 7.841355497279987e-11,
"loss": 1.4963862895965576,
"step": 740
},
{
"epoch": 1.0832116788321169,
"grad_norm": 0.18158665299415588,
"learning_rate": 7.828973617231079e-11,
"loss": 1.384456992149353,
"step": 742
},
{
"epoch": 1.0861313868613138,
"grad_norm": 0.28591710329055786,
"learning_rate": 7.816567606794239e-11,
"loss": 1.1959834098815918,
"step": 744
},
{
"epoch": 1.089051094890511,
"grad_norm": 0.18725569546222687,
"learning_rate": 7.804137594507565e-11,
"loss": 1.2433326244354248,
"step": 746
},
{
"epoch": 1.091970802919708,
"grad_norm": 0.18714189529418945,
"learning_rate": 7.791683709157841e-11,
"loss": 1.3405526876449585,
"step": 748
},
{
"epoch": 1.094890510948905,
"grad_norm": 0.3646625876426697,
"learning_rate": 7.779206079779196e-11,
"loss": 1.3495800495147705,
"step": 750
},
{
"epoch": 1.0978102189781023,
"grad_norm": 0.4056638777256012,
"learning_rate": 7.766704835651772e-11,
"loss": 1.2475500106811523,
"step": 752
},
{
"epoch": 1.1007299270072992,
"grad_norm": 0.5528532862663269,
"learning_rate": 7.75418010630038e-11,
"loss": 1.3547694683074951,
"step": 754
},
{
"epoch": 1.1036496350364964,
"grad_norm": 0.45539093017578125,
"learning_rate": 7.741632021493161e-11,
"loss": 1.3677958250045776,
"step": 756
},
{
"epoch": 1.1065693430656935,
"grad_norm": 0.26975327730178833,
"learning_rate": 7.729060711240244e-11,
"loss": 1.6038460731506348,
"step": 758
},
{
"epoch": 1.1094890510948905,
"grad_norm": 0.31061655282974243,
"learning_rate": 7.71646630579239e-11,
"loss": 1.3528101444244385,
"step": 760
},
{
"epoch": 1.1124087591240877,
"grad_norm": 0.5260862708091736,
"learning_rate": 7.703848935639653e-11,
"loss": 1.6168701648712158,
"step": 762
},
{
"epoch": 1.1153284671532846,
"grad_norm": 0.29489070177078247,
"learning_rate": 7.691208731510022e-11,
"loss": 1.455543875694275,
"step": 764
},
{
"epoch": 1.1182481751824818,
"grad_norm": 0.42220962047576904,
"learning_rate": 7.678545824368068e-11,
"loss": 1.2831261157989502,
"step": 766
},
{
"epoch": 1.121167883211679,
"grad_norm": 0.8123577833175659,
"learning_rate": 7.665860345413583e-11,
"loss": 1.337751865386963,
"step": 768
},
{
"epoch": 1.1240875912408759,
"grad_norm": 0.29676365852355957,
"learning_rate": 7.653152426080231e-11,
"loss": 1.1918377876281738,
"step": 770
},
{
"epoch": 1.127007299270073,
"grad_norm": 0.2986065447330475,
"learning_rate": 7.640422198034175e-11,
"loss": 1.2475813627243042,
"step": 772
},
{
"epoch": 1.12992700729927,
"grad_norm": 0.5116667747497559,
"learning_rate": 7.627669793172719e-11,
"loss": 1.365373969078064,
"step": 774
},
{
"epoch": 1.1328467153284671,
"grad_norm": 0.2317640781402588,
"learning_rate": 7.614895343622941e-11,
"loss": 1.4166637659072876,
"step": 776
},
{
"epoch": 1.1357664233576643,
"grad_norm": 0.39576828479766846,
"learning_rate": 7.60209898174032e-11,
"loss": 1.5067970752716064,
"step": 778
},
{
"epoch": 1.1386861313868613,
"grad_norm": 0.2449396550655365,
"learning_rate": 7.589280840107372e-11,
"loss": 1.4031476974487305,
"step": 780
},
{
"epoch": 1.1416058394160584,
"grad_norm": 0.14816564321517944,
"learning_rate": 7.576441051532268e-11,
"loss": 1.3129007816314697,
"step": 782
},
{
"epoch": 1.1445255474452556,
"grad_norm": 0.3485988676548004,
"learning_rate": 7.563579749047463e-11,
"loss": 1.4514999389648438,
"step": 784
},
{
"epoch": 1.1474452554744525,
"grad_norm": 0.5801905393600464,
"learning_rate": 7.55069706590832e-11,
"loss": 1.5295640230178833,
"step": 786
},
{
"epoch": 1.1503649635036497,
"grad_norm": 0.29618364572525024,
"learning_rate": 7.537793135591721e-11,
"loss": 1.3779162168502808,
"step": 788
},
{
"epoch": 1.1532846715328466,
"grad_norm": 0.342839777469635,
"learning_rate": 7.524868091794691e-11,
"loss": 1.4380520582199097,
"step": 790
},
{
"epoch": 1.1562043795620438,
"grad_norm": 0.34696051478385925,
"learning_rate": 7.511922068433012e-11,
"loss": 1.4846165180206299,
"step": 792
},
{
"epoch": 1.159124087591241,
"grad_norm": 0.7464672327041626,
"learning_rate": 7.498955199639834e-11,
"loss": 1.4885586500167847,
"step": 794
},
{
"epoch": 1.162043795620438,
"grad_norm": 0.3263891637325287,
"learning_rate": 7.485967619764282e-11,
"loss": 1.4655848741531372,
"step": 796
},
{
"epoch": 1.164963503649635,
"grad_norm": 0.4067479372024536,
"learning_rate": 7.472959463370074e-11,
"loss": 1.2892862558364868,
"step": 798
},
{
"epoch": 1.167883211678832,
"grad_norm": 0.3189847469329834,
"learning_rate": 7.459930865234115e-11,
"loss": 1.4570207595825195,
"step": 800
},
{
"epoch": 1.1708029197080292,
"grad_norm": 0.5341388583183289,
"learning_rate": 7.446881960345107e-11,
"loss": 1.4906622171401978,
"step": 802
},
{
"epoch": 1.1737226277372264,
"grad_norm": 0.30527299642562866,
"learning_rate": 7.433812883902151e-11,
"loss": 1.4360853433609009,
"step": 804
},
{
"epoch": 1.1766423357664233,
"grad_norm": 0.26061245799064636,
"learning_rate": 7.420723771313342e-11,
"loss": 1.499984860420227,
"step": 806
},
{
"epoch": 1.1795620437956205,
"grad_norm": 0.3794720470905304,
"learning_rate": 7.407614758194375e-11,
"loss": 1.4522984027862549,
"step": 808
},
{
"epoch": 1.1824817518248176,
"grad_norm": 0.32486245036125183,
"learning_rate": 7.394485980367121e-11,
"loss": 1.3547807931900024,
"step": 810
},
{
"epoch": 1.1854014598540146,
"grad_norm": 0.3289315104484558,
"learning_rate": 7.381337573858245e-11,
"loss": 1.3755017518997192,
"step": 812
},
{
"epoch": 1.1883211678832117,
"grad_norm": 0.6499447226524353,
"learning_rate": 7.36816967489778e-11,
"loss": 1.4222335815429688,
"step": 814
},
{
"epoch": 1.1912408759124087,
"grad_norm": 0.26047471165657043,
"learning_rate": 7.354982419917714e-11,
"loss": 1.3651729822158813,
"step": 816
},
{
"epoch": 1.1941605839416058,
"grad_norm": 0.3220467269420624,
"learning_rate": 7.34177594555059e-11,
"loss": 1.4682464599609375,
"step": 818
},
{
"epoch": 1.197080291970803,
"grad_norm": 0.26152145862579346,
"learning_rate": 7.328550388628072e-11,
"loss": 1.3593947887420654,
"step": 820
},
{
"epoch": 1.2,
"grad_norm": 0.3044966459274292,
"learning_rate": 7.31530588617955e-11,
"loss": 1.1967570781707764,
"step": 822
},
{
"epoch": 1.2029197080291971,
"grad_norm": 0.27290457487106323,
"learning_rate": 7.3020425754307e-11,
"loss": 1.276015281677246,
"step": 824
},
{
"epoch": 1.205839416058394,
"grad_norm": 0.2363789975643158,
"learning_rate": 7.288760593802067e-11,
"loss": 1.3282839059829712,
"step": 826
},
{
"epoch": 1.2087591240875912,
"grad_norm": 0.25797799229621887,
"learning_rate": 7.275460078907654e-11,
"loss": 1.3120553493499756,
"step": 828
},
{
"epoch": 1.2116788321167884,
"grad_norm": 0.2650147080421448,
"learning_rate": 7.262141168553478e-11,
"loss": 1.4232453107833862,
"step": 830
},
{
"epoch": 1.2145985401459853,
"grad_norm": 0.38055944442749023,
"learning_rate": 7.248804000736153e-11,
"loss": 1.3195230960845947,
"step": 832
},
{
"epoch": 1.2175182481751825,
"grad_norm": 0.8058264255523682,
"learning_rate": 7.235448713641457e-11,
"loss": 1.3820394277572632,
"step": 834
},
{
"epoch": 1.2204379562043797,
"grad_norm": 0.27823376655578613,
"learning_rate": 7.222075445642904e-11,
"loss": 1.477295994758606,
"step": 836
},
{
"epoch": 1.2233576642335766,
"grad_norm": 0.30671426653862,
"learning_rate": 7.208684335300305e-11,
"loss": 1.6407607793807983,
"step": 838
},
{
"epoch": 1.2262773722627738,
"grad_norm": 0.2697727382183075,
"learning_rate": 7.195275521358333e-11,
"loss": 1.5532571077346802,
"step": 840
},
{
"epoch": 1.2291970802919707,
"grad_norm": 0.34386980533599854,
"learning_rate": 7.181849142745091e-11,
"loss": 1.5054192543029785,
"step": 842
},
{
"epoch": 1.2321167883211679,
"grad_norm": 0.81573086977005,
"learning_rate": 7.168405338570667e-11,
"loss": 1.5115678310394287,
"step": 844
},
{
"epoch": 1.235036496350365,
"grad_norm": 0.5397254228591919,
"learning_rate": 7.154944248125693e-11,
"loss": 1.638497233390808,
"step": 846
},
{
"epoch": 1.237956204379562,
"grad_norm": 0.5890994668006897,
"learning_rate": 7.141466010879904e-11,
"loss": 1.4858747720718384,
"step": 848
},
{
"epoch": 1.2408759124087592,
"grad_norm": 0.44891178607940674,
"learning_rate": 7.127970766480695e-11,
"loss": 1.6084527969360352,
"step": 850
},
{
"epoch": 1.243795620437956,
"grad_norm": 0.48259827494621277,
"learning_rate": 7.114458654751666e-11,
"loss": 1.5594782829284668,
"step": 852
},
{
"epoch": 1.2467153284671533,
"grad_norm": 0.7319297790527344,
"learning_rate": 7.100929815691185e-11,
"loss": 1.5059024095535278,
"step": 854
},
{
"epoch": 1.2496350364963504,
"grad_norm": 0.5670822262763977,
"learning_rate": 7.087384389470928e-11,
"loss": 1.0539096593856812,
"step": 856
},
{
"epoch": 1.2525547445255474,
"grad_norm": 0.7238962054252625,
"learning_rate": 7.073822516434425e-11,
"loss": 1.2365281581878662,
"step": 858
},
{
"epoch": 1.2554744525547445,
"grad_norm": 0.5037800073623657,
"learning_rate": 7.060244337095619e-11,
"loss": 1.411769986152649,
"step": 860
},
{
"epoch": 1.2583941605839417,
"grad_norm": 0.4767591655254364,
"learning_rate": 7.046649992137399e-11,
"loss": 1.281144380569458,
"step": 862
},
{
"epoch": 1.2613138686131387,
"grad_norm": 0.3418749272823334,
"learning_rate": 7.03303962241014e-11,
"loss": 1.2626243829727173,
"step": 864
},
{
"epoch": 1.2642335766423358,
"grad_norm": 0.29436102509498596,
"learning_rate": 7.019413368930259e-11,
"loss": 1.3997855186462402,
"step": 866
},
{
"epoch": 1.2671532846715328,
"grad_norm": 0.5215297341346741,
"learning_rate": 7.00577137287873e-11,
"loss": 1.2505439519882202,
"step": 868
},
{
"epoch": 1.27007299270073,
"grad_norm": 0.8407970666885376,
"learning_rate": 6.992113775599654e-11,
"loss": 0.9684967994689941,
"step": 870
},
{
"epoch": 1.2729927007299269,
"grad_norm": 1.0256235599517822,
"learning_rate": 6.978440718598756e-11,
"loss": 1.1375266313552856,
"step": 872
},
{
"epoch": 1.275912408759124,
"grad_norm": 0.5392014384269714,
"learning_rate": 6.964752343541952e-11,
"loss": 1.3095382452011108,
"step": 874
},
{
"epoch": 1.2788321167883212,
"grad_norm": 0.3686670958995819,
"learning_rate": 6.95104879225386e-11,
"loss": 1.0908489227294922,
"step": 876
},
{
"epoch": 1.2817518248175181,
"grad_norm": 0.5632182359695435,
"learning_rate": 6.937330206716343e-11,
"loss": 1.1164169311523438,
"step": 878
},
{
"epoch": 1.2846715328467153,
"grad_norm": 0.32079240679740906,
"learning_rate": 6.923596729067029e-11,
"loss": 1.232709527015686,
"step": 880
},
{
"epoch": 1.2875912408759125,
"grad_norm": 0.3154524564743042,
"learning_rate": 6.909848501597848e-11,
"loss": 1.255340576171875,
"step": 882
},
{
"epoch": 1.2905109489051094,
"grad_norm": 0.28902700543403625,
"learning_rate": 6.896085666753544e-11,
"loss": 1.377984881401062,
"step": 884
},
{
"epoch": 1.2934306569343066,
"grad_norm": 0.4770423173904419,
"learning_rate": 6.882308367130217e-11,
"loss": 1.4664411544799805,
"step": 886
},
{
"epoch": 1.2963503649635038,
"grad_norm": 1.9368946552276611,
"learning_rate": 6.868516745473831e-11,
"loss": 1.1397716999053955,
"step": 888
},
{
"epoch": 1.2992700729927007,
"grad_norm": 0.6642079949378967,
"learning_rate": 6.854710944678737e-11,
"loss": 1.1195887327194214,
"step": 890
},
{
"epoch": 1.3021897810218979,
"grad_norm": 0.4591664671897888,
"learning_rate": 6.840891107786203e-11,
"loss": 1.4305362701416016,
"step": 892
},
{
"epoch": 1.305109489051095,
"grad_norm": 0.4840669631958008,
"learning_rate": 6.827057377982915e-11,
"loss": 1.4628136157989502,
"step": 894
},
{
"epoch": 1.308029197080292,
"grad_norm": 0.3429550230503082,
"learning_rate": 6.81320989859951e-11,
"loss": 1.4985345602035522,
"step": 896
},
{
"epoch": 1.310948905109489,
"grad_norm": 0.3146049380302429,
"learning_rate": 6.799348813109082e-11,
"loss": 1.6390399932861328,
"step": 898
},
{
"epoch": 1.313868613138686,
"grad_norm": 0.23190069198608398,
"learning_rate": 6.785474265125695e-11,
"loss": 1.2628161907196045,
"step": 900
},
{
"epoch": 1.3167883211678832,
"grad_norm": 0.4373994767665863,
"learning_rate": 6.771586398402901e-11,
"loss": 1.3325560092926025,
"step": 902
},
{
"epoch": 1.3197080291970802,
"grad_norm": 0.3247528672218323,
"learning_rate": 6.757685356832243e-11,
"loss": 1.2263948917388916,
"step": 904
},
{
"epoch": 1.3226277372262774,
"grad_norm": 0.2387562245130539,
"learning_rate": 6.743771284441771e-11,
"loss": 1.206580638885498,
"step": 906
},
{
"epoch": 1.3255474452554745,
"grad_norm": 0.5596776008605957,
"learning_rate": 6.729844325394546e-11,
"loss": 1.2958136796951294,
"step": 908
},
{
"epoch": 1.3284671532846715,
"grad_norm": 0.20952443778514862,
"learning_rate": 6.715904623987145e-11,
"loss": 1.6216472387313843,
"step": 910
},
{
"epoch": 1.3313868613138686,
"grad_norm": 0.5061504244804382,
"learning_rate": 6.701952324648167e-11,
"loss": 1.4907279014587402,
"step": 912
},
{
"epoch": 1.3343065693430658,
"grad_norm": 0.3604317903518677,
"learning_rate": 6.687987571936747e-11,
"loss": 1.1069793701171875,
"step": 914
},
{
"epoch": 1.3372262773722627,
"grad_norm": 0.2837623953819275,
"learning_rate": 6.674010510541039e-11,
"loss": 1.3102599382400513,
"step": 916
},
{
"epoch": 1.34014598540146,
"grad_norm": 0.4640481472015381,
"learning_rate": 6.660021285276727e-11,
"loss": 1.6540470123291016,
"step": 918
},
{
"epoch": 1.343065693430657,
"grad_norm": 0.3107167184352875,
"learning_rate": 6.646020041085529e-11,
"loss": 1.4386550188064575,
"step": 920
},
{
"epoch": 1.345985401459854,
"grad_norm": 0.3694753646850586,
"learning_rate": 6.632006923033689e-11,
"loss": 1.5420114994049072,
"step": 922
},
{
"epoch": 1.348905109489051,
"grad_norm": 0.26372024416923523,
"learning_rate": 6.617982076310476e-11,
"loss": 1.5645748376846313,
"step": 924
},
{
"epoch": 1.3518248175182481,
"grad_norm": 0.9855953454971313,
"learning_rate": 6.603945646226675e-11,
"loss": 1.351561188697815,
"step": 926
},
{
"epoch": 1.3547445255474453,
"grad_norm": 0.425919771194458,
"learning_rate": 6.58989777821309e-11,
"loss": 1.2176544666290283,
"step": 928
},
{
"epoch": 1.3576642335766422,
"grad_norm": 0.26680389046669006,
"learning_rate": 6.575838617819036e-11,
"loss": 1.3588489294052124,
"step": 930
},
{
"epoch": 1.3605839416058394,
"grad_norm": 0.49709874391555786,
"learning_rate": 6.561768310710817e-11,
"loss": 1.4943910837173462,
"step": 932
},
{
"epoch": 1.3635036496350366,
"grad_norm": 0.5887729525566101,
"learning_rate": 6.547687002670238e-11,
"loss": 1.7052679061889648,
"step": 934
},
{
"epoch": 1.3664233576642335,
"grad_norm": 0.4863436222076416,
"learning_rate": 6.53359483959308e-11,
"loss": 1.5693230628967285,
"step": 936
},
{
"epoch": 1.3693430656934307,
"grad_norm": 0.57687908411026,
"learning_rate": 6.519491967487592e-11,
"loss": 1.7266438007354736,
"step": 938
},
{
"epoch": 1.3722627737226278,
"grad_norm": 0.27145904302597046,
"learning_rate": 6.505378532472983e-11,
"loss": 1.4036791324615479,
"step": 940
},
{
"epoch": 1.3751824817518248,
"grad_norm": 0.4891437292098999,
"learning_rate": 6.491254680777894e-11,
"loss": 1.4227482080459595,
"step": 942
},
{
"epoch": 1.378102189781022,
"grad_norm": 0.2843855619430542,
"learning_rate": 6.477120558738903e-11,
"loss": 1.0680041313171387,
"step": 944
},
{
"epoch": 1.3810218978102191,
"grad_norm": 0.5112000703811646,
"learning_rate": 6.462976312798997e-11,
"loss": 1.4661967754364014,
"step": 946
},
{
"epoch": 1.383941605839416,
"grad_norm": 1.1939078569412231,
"learning_rate": 6.44882208950605e-11,
"loss": 1.5780178308486938,
"step": 948
},
{
"epoch": 1.3868613138686132,
"grad_norm": 0.5088428258895874,
"learning_rate": 6.434658035511315e-11,
"loss": 1.398094892501831,
"step": 950
},
{
"epoch": 1.3897810218978102,
"grad_norm": 0.2530817985534668,
"learning_rate": 6.420484297567905e-11,
"loss": 1.348752737045288,
"step": 952
},
{
"epoch": 1.3927007299270073,
"grad_norm": 0.8184604048728943,
"learning_rate": 6.406301022529257e-11,
"loss": 1.2859934568405151,
"step": 954
},
{
"epoch": 1.3956204379562043,
"grad_norm": 0.2624606490135193,
"learning_rate": 6.392108357347634e-11,
"loss": 1.185379981994629,
"step": 956
},
{
"epoch": 1.3985401459854014,
"grad_norm": 0.3192046582698822,
"learning_rate": 6.377906449072578e-11,
"loss": 1.074135422706604,
"step": 958
},
{
"epoch": 1.4014598540145986,
"grad_norm": 1.6579471826553345,
"learning_rate": 6.363695444849407e-11,
"loss": 1.3976491689682007,
"step": 960
},
{
"epoch": 1.4043795620437955,
"grad_norm": 1.3667290210723877,
"learning_rate": 6.349475491917677e-11,
"loss": 1.4886680841445923,
"step": 962
},
{
"epoch": 1.4072992700729927,
"grad_norm": 0.25920569896698,
"learning_rate": 6.335246737609664e-11,
"loss": 1.5233007669448853,
"step": 964
},
{
"epoch": 1.4102189781021899,
"grad_norm": 0.2395065426826477,
"learning_rate": 6.321009329348832e-11,
"loss": 1.3459280729293823,
"step": 966
},
{
"epoch": 1.4131386861313868,
"grad_norm": 0.6280210614204407,
"learning_rate": 6.306763414648311e-11,
"loss": 1.6271347999572754,
"step": 968
},
{
"epoch": 1.416058394160584,
"grad_norm": 0.2879810631275177,
"learning_rate": 6.292509141109364e-11,
"loss": 1.3719046115875244,
"step": 970
},
{
"epoch": 1.4189781021897812,
"grad_norm": 0.1828605681657791,
"learning_rate": 6.278246656419859e-11,
"loss": 1.2339953184127808,
"step": 972
},
{
"epoch": 1.421897810218978,
"grad_norm": 0.4212108552455902,
"learning_rate": 6.263976108352739e-11,
"loss": 1.4191358089447021,
"step": 974
},
{
"epoch": 1.4248175182481753,
"grad_norm": 0.576755166053772,
"learning_rate": 6.249697644764493e-11,
"loss": 1.3736066818237305,
"step": 976
},
{
"epoch": 1.4277372262773722,
"grad_norm": 0.39886170625686646,
"learning_rate": 6.235411413593627e-11,
"loss": 1.3101552724838257,
"step": 978
},
{
"epoch": 1.4306569343065694,
"grad_norm": 3.171804666519165,
"learning_rate": 6.221117562859115e-11,
"loss": 1.160914421081543,
"step": 980
},
{
"epoch": 1.4335766423357663,
"grad_norm": 0.28551721572875977,
"learning_rate": 6.206816240658887e-11,
"loss": 1.2121726274490356,
"step": 982
},
{
"epoch": 1.4364963503649635,
"grad_norm": 0.41442498564720154,
"learning_rate": 6.192507595168279e-11,
"loss": 1.4567291736602783,
"step": 984
},
{
"epoch": 1.4394160583941606,
"grad_norm": 0.2783399224281311,
"learning_rate": 6.178191774638506e-11,
"loss": 1.2854819297790527,
"step": 986
},
{
"epoch": 1.4423357664233576,
"grad_norm": 0.3941711187362671,
"learning_rate": 6.163868927395123e-11,
"loss": 1.473130226135254,
"step": 988
},
{
"epoch": 1.4452554744525548,
"grad_norm": 0.34142744541168213,
"learning_rate": 6.149539201836484e-11,
"loss": 1.5927597284317017,
"step": 990
},
{
"epoch": 1.448175182481752,
"grad_norm": 0.2587209641933441,
"learning_rate": 6.135202746432217e-11,
"loss": 1.50771963596344,
"step": 992
},
{
"epoch": 1.4510948905109489,
"grad_norm": 0.23753675818443298,
"learning_rate": 6.120859709721674e-11,
"loss": 1.2842893600463867,
"step": 994
},
{
"epoch": 1.454014598540146,
"grad_norm": 0.3801853060722351,
"learning_rate": 6.106510240312393e-11,
"loss": 1.3453171253204346,
"step": 996
},
{
"epoch": 1.4569343065693432,
"grad_norm": 0.3056936264038086,
"learning_rate": 6.092154486878562e-11,
"loss": 1.2746431827545166,
"step": 998
},
{
"epoch": 1.4598540145985401,
"grad_norm": 0.7941753268241882,
"learning_rate": 6.077792598159479e-11,
"loss": 1.3984593152999878,
"step": 1000
},
{
"epoch": 1.4627737226277373,
"grad_norm": 0.7185998558998108,
"learning_rate": 6.06342472295801e-11,
"loss": 1.588934302330017,
"step": 1002
},
{
"epoch": 1.4656934306569342,
"grad_norm": 0.36543312668800354,
"learning_rate": 6.049051010139046e-11,
"loss": 1.4541923999786377,
"step": 1004
},
{
"epoch": 1.4686131386861314,
"grad_norm": 0.39861613512039185,
"learning_rate": 6.034671608627957e-11,
"loss": 1.330125331878662,
"step": 1006
},
{
"epoch": 1.4715328467153284,
"grad_norm": 0.39358770847320557,
"learning_rate": 6.020286667409061e-11,
"loss": 1.4012455940246582,
"step": 1008
},
{
"epoch": 1.4744525547445255,
"grad_norm": 0.6596587896347046,
"learning_rate": 6.005896335524069e-11,
"loss": 1.3372608423233032,
"step": 1010
},
{
"epoch": 1.4773722627737227,
"grad_norm": 0.19527080655097961,
"learning_rate": 5.99150076207054e-11,
"loss": 1.2573559284210205,
"step": 1012
},
{
"epoch": 1.4802919708029196,
"grad_norm": 0.439053475856781,
"learning_rate": 5.977100096200347e-11,
"loss": 1.3484768867492676,
"step": 1014
},
{
"epoch": 1.4832116788321168,
"grad_norm": 1.0919126272201538,
"learning_rate": 5.962694487118122e-11,
"loss": 1.2934248447418213,
"step": 1016
},
{
"epoch": 1.486131386861314,
"grad_norm": 0.4834900498390198,
"learning_rate": 5.948284084079716e-11,
"loss": 1.359639286994934,
"step": 1018
},
{
"epoch": 1.489051094890511,
"grad_norm": 0.32894521951675415,
"learning_rate": 5.933869036390651e-11,
"loss": 1.3357242345809937,
"step": 1020
},
{
"epoch": 1.491970802919708,
"grad_norm": 0.25190240144729614,
"learning_rate": 5.919449493404563e-11,
"loss": 1.5525610446929932,
"step": 1022
},
{
"epoch": 1.4948905109489052,
"grad_norm": 0.36443647742271423,
"learning_rate": 5.905025604521674e-11,
"loss": 1.7397009134292603,
"step": 1024
},
{
"epoch": 1.4978102189781022,
"grad_norm": 0.28807729482650757,
"learning_rate": 5.890597519187229e-11,
"loss": 1.5908589363098145,
"step": 1026
},
{
"epoch": 1.5007299270072991,
"grad_norm": 1.0596458911895752,
"learning_rate": 5.876165386889952e-11,
"loss": 1.2815005779266357,
"step": 1028
},
{
"epoch": 1.5036496350364965,
"grad_norm": 0.6740530133247375,
"learning_rate": 5.861729357160501e-11,
"loss": 1.121368408203125,
"step": 1030
},
{
"epoch": 1.5065693430656935,
"grad_norm": 0.25546136498451233,
"learning_rate": 5.84728957956991e-11,
"loss": 1.2286404371261597,
"step": 1032
},
{
"epoch": 1.5094890510948904,
"grad_norm": 0.6252645254135132,
"learning_rate": 5.832846203728045e-11,
"loss": 1.3033808469772339,
"step": 1034
},
{
"epoch": 1.5124087591240876,
"grad_norm": 0.5182809829711914,
"learning_rate": 5.81839937928206e-11,
"loss": 1.7270663976669312,
"step": 1036
},
{
"epoch": 1.5153284671532847,
"grad_norm": 0.458566814661026,
"learning_rate": 5.8039492559148314e-11,
"loss": 1.8393914699554443,
"step": 1038
},
{
"epoch": 1.5182481751824817,
"grad_norm": 0.522005021572113,
"learning_rate": 5.789495983343418e-11,
"loss": 1.5257136821746826,
"step": 1040
},
{
"epoch": 1.5211678832116788,
"grad_norm": 0.3969437777996063,
"learning_rate": 5.7750397113175114e-11,
"loss": 1.3572144508361816,
"step": 1042
},
{
"epoch": 1.524087591240876,
"grad_norm": 0.22145526111125946,
"learning_rate": 5.760580589617876e-11,
"loss": 1.3691705465316772,
"step": 1044
},
{
"epoch": 1.527007299270073,
"grad_norm": 0.8607689142227173,
"learning_rate": 5.746118768054806e-11,
"loss": 1.4887243509292603,
"step": 1046
},
{
"epoch": 1.5299270072992701,
"grad_norm": 0.5395622253417969,
"learning_rate": 5.73165439646656e-11,
"loss": 1.6568856239318848,
"step": 1048
},
{
"epoch": 1.5328467153284673,
"grad_norm": 0.587317168712616,
"learning_rate": 5.717187624717827e-11,
"loss": 1.622255563735962,
"step": 1050
},
{
"epoch": 1.5357664233576642,
"grad_norm": 0.40985527634620667,
"learning_rate": 5.7027186026981604e-11,
"loss": 1.221044659614563,
"step": 1052
},
{
"epoch": 1.5386861313868612,
"grad_norm": 0.7627142071723938,
"learning_rate": 5.6882474803204254e-11,
"loss": 1.3784704208374023,
"step": 1054
},
{
"epoch": 1.5416058394160586,
"grad_norm": 0.7809776067733765,
"learning_rate": 5.673774407519253e-11,
"loss": 1.5165852308273315,
"step": 1056
},
{
"epoch": 1.5445255474452555,
"grad_norm": 0.7891870737075806,
"learning_rate": 5.6592995342494836e-11,
"loss": 1.712707757949829,
"step": 1058
},
{
"epoch": 1.5474452554744524,
"grad_norm": 0.7284282445907593,
"learning_rate": 5.644823010484604e-11,
"loss": 1.7109613418579102,
"step": 1060
},
{
"epoch": 1.5503649635036496,
"grad_norm": 0.4619555175304413,
"learning_rate": 5.6303449862152144e-11,
"loss": 1.3397283554077148,
"step": 1062
},
{
"epoch": 1.5532846715328468,
"grad_norm": 0.7122825980186462,
"learning_rate": 5.61586561144745e-11,
"loss": 1.565988540649414,
"step": 1064
},
{
"epoch": 1.5562043795620437,
"grad_norm": 0.3608238101005554,
"learning_rate": 5.601385036201443e-11,
"loss": 1.2163538932800293,
"step": 1066
},
{
"epoch": 1.5591240875912409,
"grad_norm": 1.3949124813079834,
"learning_rate": 5.586903410509765e-11,
"loss": 1.230873942375183,
"step": 1068
},
{
"epoch": 1.562043795620438,
"grad_norm": 0.38245004415512085,
"learning_rate": 5.572420884415871e-11,
"loss": 1.021897315979004,
"step": 1070
},
{
"epoch": 1.564963503649635,
"grad_norm": 0.34170427918434143,
"learning_rate": 5.557937607972542e-11,
"loss": 1.0241258144378662,
"step": 1072
},
{
"epoch": 1.5678832116788322,
"grad_norm": 0.315836638212204,
"learning_rate": 5.543453731240338e-11,
"loss": 1.202805519104004,
"step": 1074
},
{
"epoch": 1.5708029197080293,
"grad_norm": 0.3503977656364441,
"learning_rate": 5.528969404286032e-11,
"loss": 1.213181734085083,
"step": 1076
},
{
"epoch": 1.5737226277372263,
"grad_norm": 0.46344345808029175,
"learning_rate": 5.51448477718107e-11,
"loss": 1.1099185943603516,
"step": 1078
},
{
"epoch": 1.5766423357664232,
"grad_norm": 0.37887194752693176,
"learning_rate": 5.5000000000000004e-11,
"loss": 1.25699782371521,
"step": 1080
},
{
"epoch": 1.5795620437956206,
"grad_norm": 0.26375895738601685,
"learning_rate": 5.485515222818931e-11,
"loss": 1.2985727787017822,
"step": 1082
},
{
"epoch": 1.5824817518248175,
"grad_norm": 0.587094783782959,
"learning_rate": 5.4710305957139695e-11,
"loss": 1.3389722108840942,
"step": 1084
},
{
"epoch": 1.5854014598540145,
"grad_norm": 0.26151588559150696,
"learning_rate": 5.4565462687596646e-11,
"loss": 1.3401594161987305,
"step": 1086
},
{
"epoch": 1.5883211678832116,
"grad_norm": 0.33108699321746826,
"learning_rate": 5.442062392027459e-11,
"loss": 1.1463489532470703,
"step": 1088
},
{
"epoch": 1.5912408759124088,
"grad_norm": 0.12337132543325424,
"learning_rate": 5.4275791155841305e-11,
"loss": 1.1203198432922363,
"step": 1090
},
{
"epoch": 1.5941605839416058,
"grad_norm": 0.4239310622215271,
"learning_rate": 5.4130965894902354e-11,
"loss": 1.1963176727294922,
"step": 1092
},
{
"epoch": 1.597080291970803,
"grad_norm": 0.3204934000968933,
"learning_rate": 5.398614963798558e-11,
"loss": 1.313279151916504,
"step": 1094
},
{
"epoch": 1.6,
"grad_norm": 0.28497034311294556,
"learning_rate": 5.384134388552552e-11,
"loss": 1.5193721055984497,
"step": 1096
},
{
"epoch": 1.602919708029197,
"grad_norm": 0.1569192260503769,
"learning_rate": 5.3696550137847864e-11,
"loss": 1.2503010034561157,
"step": 1098
},
{
"epoch": 1.6058394160583942,
"grad_norm": 0.7396665811538696,
"learning_rate": 5.355176989515396e-11,
"loss": 1.1561777591705322,
"step": 1100
},
{
"epoch": 1.6087591240875914,
"grad_norm": 0.27860796451568604,
"learning_rate": 5.3407004657505185e-11,
"loss": 1.3068069219589233,
"step": 1102
},
{
"epoch": 1.6116788321167883,
"grad_norm": 0.653042733669281,
"learning_rate": 5.326225592480748e-11,
"loss": 1.439526081085205,
"step": 1104
},
{
"epoch": 1.6145985401459853,
"grad_norm": 0.335859090089798,
"learning_rate": 5.3117525196795767e-11,
"loss": 1.2161372900009155,
"step": 1106
},
{
"epoch": 1.6175182481751826,
"grad_norm": 0.3679628074169159,
"learning_rate": 5.297281397301841e-11,
"loss": 1.167518973350525,
"step": 1108
},
{
"epoch": 1.6204379562043796,
"grad_norm": 0.17621150612831116,
"learning_rate": 5.282812375282173e-11,
"loss": 1.581429362297058,
"step": 1110
},
{
"epoch": 1.6233576642335765,
"grad_norm": 0.3105796277523041,
"learning_rate": 5.268345603533441e-11,
"loss": 1.1999684572219849,
"step": 1112
},
{
"epoch": 1.6262773722627737,
"grad_norm": 0.5713252425193787,
"learning_rate": 5.2538812319451956e-11,
"loss": 1.129384994506836,
"step": 1114
},
{
"epoch": 1.6291970802919709,
"grad_norm": 0.245794415473938,
"learning_rate": 5.2394194103821245e-11,
"loss": 1.1704528331756592,
"step": 1116
},
{
"epoch": 1.6321167883211678,
"grad_norm": 0.8822201490402222,
"learning_rate": 5.22496028868249e-11,
"loss": 1.220594048500061,
"step": 1118
},
{
"epoch": 1.635036496350365,
"grad_norm": 0.394856721162796,
"learning_rate": 5.210504016656583e-11,
"loss": 1.506517767906189,
"step": 1120
},
{
"epoch": 1.6379562043795621,
"grad_norm": 0.3811044991016388,
"learning_rate": 5.1960507440851726e-11,
"loss": 1.4898362159729004,
"step": 1122
},
{
"epoch": 1.640875912408759,
"grad_norm": 0.7341606020927429,
"learning_rate": 5.1816006207179414e-11,
"loss": 1.4160077571868896,
"step": 1124
},
{
"epoch": 1.6437956204379562,
"grad_norm": 0.27225232124328613,
"learning_rate": 5.167153796271955e-11,
"loss": 1.1867581605911255,
"step": 1126
},
{
"epoch": 1.6467153284671534,
"grad_norm": 0.250491738319397,
"learning_rate": 5.152710420430091e-11,
"loss": 1.469104290008545,
"step": 1128
},
{
"epoch": 1.6496350364963503,
"grad_norm": 0.3392203748226166,
"learning_rate": 5.1382706428395e-11,
"loss": 1.2819688320159912,
"step": 1130
},
{
"epoch": 1.6525547445255473,
"grad_norm": 0.2684314548969269,
"learning_rate": 5.1238346131100486e-11,
"loss": 1.3832699060440063,
"step": 1132
},
{
"epoch": 1.6554744525547447,
"grad_norm": 0.27301549911499023,
"learning_rate": 5.1094024808127713e-11,
"loss": 1.3297004699707031,
"step": 1134
},
{
"epoch": 1.6583941605839416,
"grad_norm": 0.8703414797782898,
"learning_rate": 5.094974395478327e-11,
"loss": 1.3337628841400146,
"step": 1136
},
{
"epoch": 1.6613138686131386,
"grad_norm": 0.48248016834259033,
"learning_rate": 5.080550506595439e-11,
"loss": 1.4017032384872437,
"step": 1138
},
{
"epoch": 1.6642335766423357,
"grad_norm": 0.6301165223121643,
"learning_rate": 5.0661309636093525e-11,
"loss": 1.3380167484283447,
"step": 1140
},
{
"epoch": 1.667153284671533,
"grad_norm": 0.5514047741889954,
"learning_rate": 5.0517159159202843e-11,
"loss": 1.1581861972808838,
"step": 1142
},
{
"epoch": 1.6700729927007298,
"grad_norm": 0.3049074709415436,
"learning_rate": 5.037305512881878e-11,
"loss": 1.3807573318481445,
"step": 1144
},
{
"epoch": 1.672992700729927,
"grad_norm": 0.42181819677352905,
"learning_rate": 5.022899903799655e-11,
"loss": 1.4547597169876099,
"step": 1146
},
{
"epoch": 1.6759124087591242,
"grad_norm": 3.0331032276153564,
"learning_rate": 5.0084992379294625e-11,
"loss": 1.6627700328826904,
"step": 1148
},
{
"epoch": 1.6788321167883211,
"grad_norm": 0.5782711505889893,
"learning_rate": 4.994103664475933e-11,
"loss": 1.1384409666061401,
"step": 1150
},
{
"epoch": 1.6817518248175183,
"grad_norm": 0.2746776342391968,
"learning_rate": 4.979713332590939e-11,
"loss": 1.2250109910964966,
"step": 1152
},
{
"epoch": 1.6846715328467154,
"grad_norm": 0.3908049464225769,
"learning_rate": 4.9653283913720435e-11,
"loss": 1.4306840896606445,
"step": 1154
},
{
"epoch": 1.6875912408759124,
"grad_norm": 0.4787120521068573,
"learning_rate": 4.950948989860955e-11,
"loss": 1.584275722503662,
"step": 1156
},
{
"epoch": 1.6905109489051093,
"grad_norm": 0.3665141463279724,
"learning_rate": 4.9365752770419915e-11,
"loss": 1.496099591255188,
"step": 1158
},
{
"epoch": 1.6934306569343067,
"grad_norm": 0.36038002371788025,
"learning_rate": 4.922207401840521e-11,
"loss": 1.3737331628799438,
"step": 1160
},
{
"epoch": 1.6963503649635037,
"grad_norm": 0.38061758875846863,
"learning_rate": 4.9078455131214394e-11,
"loss": 1.3694974184036255,
"step": 1162
},
{
"epoch": 1.6992700729927006,
"grad_norm": 0.28912433981895447,
"learning_rate": 4.8934897596876084e-11,
"loss": 1.5625628232955933,
"step": 1164
},
{
"epoch": 1.7021897810218978,
"grad_norm": 0.3381723463535309,
"learning_rate": 4.879140290278327e-11,
"loss": 1.3396795988082886,
"step": 1166
},
{
"epoch": 1.705109489051095,
"grad_norm": 0.21722273528575897,
"learning_rate": 4.8647972535677835e-11,
"loss": 1.4543912410736084,
"step": 1168
},
{
"epoch": 1.7080291970802919,
"grad_norm": 0.7460881471633911,
"learning_rate": 4.8504607981635173e-11,
"loss": 1.4622814655303955,
"step": 1170
},
{
"epoch": 1.710948905109489,
"grad_norm": 1.0124627351760864,
"learning_rate": 4.83613107260488e-11,
"loss": 1.5359344482421875,
"step": 1172
},
{
"epoch": 1.7138686131386862,
"grad_norm": 0.3741784393787384,
"learning_rate": 4.821808225361497e-11,
"loss": 1.4243507385253906,
"step": 1174
},
{
"epoch": 1.7167883211678832,
"grad_norm": 0.3962346911430359,
"learning_rate": 4.8074924048317217e-11,
"loss": 1.4934279918670654,
"step": 1176
},
{
"epoch": 1.7197080291970803,
"grad_norm": 0.4626739025115967,
"learning_rate": 4.793183759341114e-11,
"loss": 1.478497862815857,
"step": 1178
},
{
"epoch": 1.7226277372262775,
"grad_norm": 0.3711540102958679,
"learning_rate": 4.7788824371408855e-11,
"loss": 1.4607126712799072,
"step": 1180
},
{
"epoch": 1.7255474452554744,
"grad_norm": 0.6520898938179016,
"learning_rate": 4.764588586406373e-11,
"loss": 1.3575026988983154,
"step": 1182
},
{
"epoch": 1.7284671532846714,
"grad_norm": 0.8203355669975281,
"learning_rate": 4.750302355235506e-11,
"loss": 1.6894546747207642,
"step": 1184
},
{
"epoch": 1.7313868613138688,
"grad_norm": 0.37086644768714905,
"learning_rate": 4.7360238916472634e-11,
"loss": 1.42867910861969,
"step": 1186
},
{
"epoch": 1.7343065693430657,
"grad_norm": 0.9349713325500488,
"learning_rate": 4.721753343580143e-11,
"loss": 1.4776962995529175,
"step": 1188
},
{
"epoch": 1.7372262773722627,
"grad_norm": 0.6913307905197144,
"learning_rate": 4.7074908588906385e-11,
"loss": 1.3140593767166138,
"step": 1190
},
{
"epoch": 1.7401459854014598,
"grad_norm": 0.38876020908355713,
"learning_rate": 4.6932365853516914e-11,
"loss": 1.4444148540496826,
"step": 1192
},
{
"epoch": 1.743065693430657,
"grad_norm": 0.26223206520080566,
"learning_rate": 4.678990670651169e-11,
"loss": 1.5006850957870483,
"step": 1194
},
{
"epoch": 1.745985401459854,
"grad_norm": 0.4009760916233063,
"learning_rate": 4.664753262390337e-11,
"loss": 1.4603517055511475,
"step": 1196
},
{
"epoch": 1.748905109489051,
"grad_norm": 0.2159179002046585,
"learning_rate": 4.6505245080823234e-11,
"loss": 1.4046112298965454,
"step": 1198
},
{
"epoch": 1.7518248175182483,
"grad_norm": 0.7279269695281982,
"learning_rate": 4.6363045551505944e-11,
"loss": 1.41069757938385,
"step": 1200
},
{
"epoch": 1.7547445255474452,
"grad_norm": 0.34113809466362,
"learning_rate": 4.622093550927423e-11,
"loss": 1.4951008558273315,
"step": 1202
},
{
"epoch": 1.7576642335766424,
"grad_norm": 0.2637474834918976,
"learning_rate": 4.6078916426523674e-11,
"loss": 1.3514400720596313,
"step": 1204
},
{
"epoch": 1.7605839416058395,
"grad_norm": 0.316500723361969,
"learning_rate": 4.593698977470744e-11,
"loss": 1.4736438989639282,
"step": 1206
},
{
"epoch": 1.7635036496350365,
"grad_norm": 0.3700624406337738,
"learning_rate": 4.579515702432098e-11,
"loss": 1.4961127042770386,
"step": 1208
},
{
"epoch": 1.7664233576642334,
"grad_norm": 1.0039310455322266,
"learning_rate": 4.565341964488686e-11,
"loss": 1.5141915082931519,
"step": 1210
},
{
"epoch": 1.7693430656934308,
"grad_norm": 0.4779518246650696,
"learning_rate": 4.551177910493951e-11,
"loss": 1.5229568481445312,
"step": 1212
},
{
"epoch": 1.7722627737226277,
"grad_norm": 0.3397350013256073,
"learning_rate": 4.537023687201004e-11,
"loss": 1.368076205253601,
"step": 1214
},
{
"epoch": 1.7751824817518247,
"grad_norm": 2.7952754497528076,
"learning_rate": 4.522879441261097e-11,
"loss": 1.538074016571045,
"step": 1216
},
{
"epoch": 1.7781021897810219,
"grad_norm": 1.0508959293365479,
"learning_rate": 4.5087453192221077e-11,
"loss": 1.4159822463989258,
"step": 1218
},
{
"epoch": 1.781021897810219,
"grad_norm": 0.3718510866165161,
"learning_rate": 4.4946214675270194e-11,
"loss": 1.3541284799575806,
"step": 1220
},
{
"epoch": 1.783941605839416,
"grad_norm": 0.3654969036579132,
"learning_rate": 4.4805080325124085e-11,
"loss": 1.2151225805282593,
"step": 1222
},
{
"epoch": 1.7868613138686131,
"grad_norm": 0.40370652079582214,
"learning_rate": 4.4664051604069215e-11,
"loss": 1.1107691526412964,
"step": 1224
},
{
"epoch": 1.7897810218978103,
"grad_norm": 0.2435276061296463,
"learning_rate": 4.452312997329763e-11,
"loss": 1.1705870628356934,
"step": 1226
},
{
"epoch": 1.7927007299270072,
"grad_norm": 0.22067563235759735,
"learning_rate": 4.438231689289185e-11,
"loss": 1.4070066213607788,
"step": 1228
},
{
"epoch": 1.7956204379562044,
"grad_norm": 0.22086317837238312,
"learning_rate": 4.4241613821809646e-11,
"loss": 1.4132890701293945,
"step": 1230
},
{
"epoch": 1.7985401459854016,
"grad_norm": 0.17591090500354767,
"learning_rate": 4.410102221786909e-11,
"loss": 1.191218614578247,
"step": 1232
},
{
"epoch": 1.8014598540145985,
"grad_norm": 0.14714421331882477,
"learning_rate": 4.3960543537733255e-11,
"loss": 1.2550400495529175,
"step": 1234
},
{
"epoch": 1.8043795620437955,
"grad_norm": 0.23958607017993927,
"learning_rate": 4.382017923689525e-11,
"loss": 1.1083970069885254,
"step": 1236
},
{
"epoch": 1.8072992700729928,
"grad_norm": 0.16653022170066833,
"learning_rate": 4.3679930769663116e-11,
"loss": 1.0630451440811157,
"step": 1238
},
{
"epoch": 1.8102189781021898,
"grad_norm": 0.48908543586730957,
"learning_rate": 4.3539799589144715e-11,
"loss": 1.1045050621032715,
"step": 1240
},
{
"epoch": 1.8131386861313867,
"grad_norm": 0.2572908103466034,
"learning_rate": 4.339978714723274e-11,
"loss": 1.0034579038619995,
"step": 1242
},
{
"epoch": 1.816058394160584,
"grad_norm": 0.19536933302879333,
"learning_rate": 4.325989489458963e-11,
"loss": 1.0157605409622192,
"step": 1244
},
{
"epoch": 1.818978102189781,
"grad_norm": 0.3267905116081238,
"learning_rate": 4.312012428063252e-11,
"loss": 1.004263997077942,
"step": 1246
},
{
"epoch": 1.821897810218978,
"grad_norm": 0.2991647720336914,
"learning_rate": 4.298047675351832e-11,
"loss": 1.0326032638549805,
"step": 1248
},
{
"epoch": 1.8248175182481752,
"grad_norm": 0.68552166223526,
"learning_rate": 4.2840953760128575e-11,
"loss": 1.171319603919983,
"step": 1250
},
{
"epoch": 1.8277372262773723,
"grad_norm": 0.30608347058296204,
"learning_rate": 4.270155674605455e-11,
"loss": 1.382737636566162,
"step": 1252
},
{
"epoch": 1.8306569343065693,
"grad_norm": 0.38271135091781616,
"learning_rate": 4.25622871555823e-11,
"loss": 1.37453293800354,
"step": 1254
},
{
"epoch": 1.8335766423357664,
"grad_norm": 0.20676696300506592,
"learning_rate": 4.242314643167759e-11,
"loss": 1.3673899173736572,
"step": 1256
},
{
"epoch": 1.8364963503649636,
"grad_norm": 0.40299898386001587,
"learning_rate": 4.2284136015971e-11,
"loss": 1.342198133468628,
"step": 1258
},
{
"epoch": 1.8394160583941606,
"grad_norm": 0.19543984532356262,
"learning_rate": 4.214525734874306e-11,
"loss": 1.3685353994369507,
"step": 1260
},
{
"epoch": 1.8423357664233575,
"grad_norm": 0.18539777398109436,
"learning_rate": 4.2006511868909207e-11,
"loss": 1.296134114265442,
"step": 1262
},
{
"epoch": 1.845255474452555,
"grad_norm": 0.21791622042655945,
"learning_rate": 4.186790101400491e-11,
"loss": 1.3704713582992554,
"step": 1264
},
{
"epoch": 1.8481751824817518,
"grad_norm": 0.5269507169723511,
"learning_rate": 4.172942622017086e-11,
"loss": 1.366510272026062,
"step": 1266
},
{
"epoch": 1.8510948905109488,
"grad_norm": 0.34738093614578247,
"learning_rate": 4.159108892213799e-11,
"loss": 1.3568106889724731,
"step": 1268
},
{
"epoch": 1.854014598540146,
"grad_norm": 0.6385413408279419,
"learning_rate": 4.145289055321263e-11,
"loss": 1.2829129695892334,
"step": 1270
},
{
"epoch": 1.856934306569343,
"grad_norm": 0.17486143112182617,
"learning_rate": 4.1314832545261706e-11,
"loss": 1.3451555967330933,
"step": 1272
},
{
"epoch": 1.85985401459854,
"grad_norm": 0.20117926597595215,
"learning_rate": 4.1176916328697836e-11,
"loss": 1.31380033493042,
"step": 1274
},
{
"epoch": 1.8627737226277372,
"grad_norm": 0.3269534111022949,
"learning_rate": 4.103914333246458e-11,
"loss": 1.295755386352539,
"step": 1276
},
{
"epoch": 1.8656934306569344,
"grad_norm": 0.2715974450111389,
"learning_rate": 4.090151498402155e-11,
"loss": 1.322004795074463,
"step": 1278
},
{
"epoch": 1.8686131386861313,
"grad_norm": 0.29098689556121826,
"learning_rate": 4.076403270932973e-11,
"loss": 1.3752413988113403,
"step": 1280
},
{
"epoch": 1.8715328467153285,
"grad_norm": 0.3070428967475891,
"learning_rate": 4.0626697932836585e-11,
"loss": 1.3564178943634033,
"step": 1282
},
{
"epoch": 1.8744525547445257,
"grad_norm": 0.24386434257030487,
"learning_rate": 4.048951207746142e-11,
"loss": 1.3851372003555298,
"step": 1284
},
{
"epoch": 1.8773722627737226,
"grad_norm": 0.2833121120929718,
"learning_rate": 4.0352476564580485e-11,
"loss": 1.3745468854904175,
"step": 1286
},
{
"epoch": 1.8802919708029195,
"grad_norm": 0.22633326053619385,
"learning_rate": 4.021559281401244e-11,
"loss": 1.3696306943893433,
"step": 1288
},
{
"epoch": 1.883211678832117,
"grad_norm": 0.2947603464126587,
"learning_rate": 4.0078862244003477e-11,
"loss": 1.4001069068908691,
"step": 1290
},
{
"epoch": 1.8861313868613139,
"grad_norm": 0.24118058383464813,
"learning_rate": 3.994228627121269e-11,
"loss": 1.3127013444900513,
"step": 1292
},
{
"epoch": 1.8890510948905108,
"grad_norm": 0.32797881960868835,
"learning_rate": 3.980586631069744e-11,
"loss": 1.3768930435180664,
"step": 1294
},
{
"epoch": 1.891970802919708,
"grad_norm": 0.38555362820625305,
"learning_rate": 3.9669603775898614e-11,
"loss": 1.3195134401321411,
"step": 1296
},
{
"epoch": 1.8948905109489051,
"grad_norm": 0.25967398285865784,
"learning_rate": 3.953350007862603e-11,
"loss": 1.3479055166244507,
"step": 1298
},
{
"epoch": 1.897810218978102,
"grad_norm": 0.21513788402080536,
"learning_rate": 3.9397556629043816e-11,
"loss": 1.358872890472412,
"step": 1300
},
{
"epoch": 1.9007299270072993,
"grad_norm": 0.1812690943479538,
"learning_rate": 3.926177483565576e-11,
"loss": 1.3878893852233887,
"step": 1302
},
{
"epoch": 1.9036496350364964,
"grad_norm": 0.2857212722301483,
"learning_rate": 3.912615610529074e-11,
"loss": 1.3680543899536133,
"step": 1304
},
{
"epoch": 1.9065693430656934,
"grad_norm": 0.17340198159217834,
"learning_rate": 3.899070184308815e-11,
"loss": 1.299302339553833,
"step": 1306
},
{
"epoch": 1.9094890510948905,
"grad_norm": 0.22302907705307007,
"learning_rate": 3.885541345248335e-11,
"loss": 1.333414912223816,
"step": 1308
},
{
"epoch": 1.9124087591240877,
"grad_norm": 0.2631547749042511,
"learning_rate": 3.8720292335193065e-11,
"loss": 1.3633759021759033,
"step": 1310
},
{
"epoch": 1.9153284671532846,
"grad_norm": 0.20353633165359497,
"learning_rate": 3.858533989120098e-11,
"loss": 1.3834773302078247,
"step": 1312
},
{
"epoch": 1.9182481751824818,
"grad_norm": 0.38067495822906494,
"learning_rate": 3.845055751874309e-11,
"loss": 1.369248390197754,
"step": 1314
},
{
"epoch": 1.921167883211679,
"grad_norm": 0.2600199282169342,
"learning_rate": 3.831594661429334e-11,
"loss": 1.3464183807373047,
"step": 1316
},
{
"epoch": 1.924087591240876,
"grad_norm": 0.2344445139169693,
"learning_rate": 3.8181508572549096e-11,
"loss": 1.3000050783157349,
"step": 1318
},
{
"epoch": 1.9270072992700729,
"grad_norm": 0.2518412470817566,
"learning_rate": 3.8047244786416667e-11,
"loss": 1.3449358940124512,
"step": 1320
},
{
"epoch": 1.92992700729927,
"grad_norm": 0.3552137017250061,
"learning_rate": 3.791315664699697e-11,
"loss": 1.3735779523849487,
"step": 1322
},
{
"epoch": 1.9328467153284672,
"grad_norm": 0.8765981793403625,
"learning_rate": 3.777924554357096e-11,
"loss": 1.3276495933532715,
"step": 1324
},
{
"epoch": 1.9357664233576641,
"grad_norm": 0.2266838103532791,
"learning_rate": 3.764551286358544e-11,
"loss": 1.355583906173706,
"step": 1326
},
{
"epoch": 1.9386861313868613,
"grad_norm": 0.21473616361618042,
"learning_rate": 3.751195999263849e-11,
"loss": 1.3490177392959595,
"step": 1328
},
{
"epoch": 1.9416058394160585,
"grad_norm": 0.3912813365459442,
"learning_rate": 3.737858831446524e-11,
"loss": 1.3904234170913696,
"step": 1330
},
{
"epoch": 1.9445255474452554,
"grad_norm": 0.2954106628894806,
"learning_rate": 3.724539921092347e-11,
"loss": 1.4133667945861816,
"step": 1332
},
{
"epoch": 1.9474452554744526,
"grad_norm": 0.2548486590385437,
"learning_rate": 3.711239406197932e-11,
"loss": 1.3827425241470337,
"step": 1334
},
{
"epoch": 1.9503649635036497,
"grad_norm": 0.4357367753982544,
"learning_rate": 3.6979574245693e-11,
"loss": 1.3763043880462646,
"step": 1336
},
{
"epoch": 1.9532846715328467,
"grad_norm": 0.26733970642089844,
"learning_rate": 3.68469411382045e-11,
"loss": 1.3509302139282227,
"step": 1338
},
{
"epoch": 1.9562043795620438,
"grad_norm": 0.20610250532627106,
"learning_rate": 3.671449611371928e-11,
"loss": 1.2913023233413696,
"step": 1340
},
{
"epoch": 1.959124087591241,
"grad_norm": 0.2972968518733978,
"learning_rate": 3.658224054449412e-11,
"loss": 1.3234026432037354,
"step": 1342
},
{
"epoch": 1.962043795620438,
"grad_norm": 0.2919084131717682,
"learning_rate": 3.645017580082287e-11,
"loss": 1.3362833261489868,
"step": 1344
},
{
"epoch": 1.964963503649635,
"grad_norm": 0.27695927023887634,
"learning_rate": 3.631830325102221e-11,
"loss": 1.3145451545715332,
"step": 1346
},
{
"epoch": 1.967883211678832,
"grad_norm": 0.42268478870391846,
"learning_rate": 3.618662426141754e-11,
"loss": 1.337138295173645,
"step": 1348
},
{
"epoch": 1.9708029197080292,
"grad_norm": 0.2148294746875763,
"learning_rate": 3.6055140196328793e-11,
"loss": 1.380530834197998,
"step": 1350
},
{
"epoch": 1.9737226277372262,
"grad_norm": 0.2545860707759857,
"learning_rate": 3.592385241805628e-11,
"loss": 1.3746337890625,
"step": 1352
},
{
"epoch": 1.9766423357664233,
"grad_norm": 1.5863921642303467,
"learning_rate": 3.579276228686658e-11,
"loss": 1.3830804824829102,
"step": 1354
},
{
"epoch": 1.9795620437956205,
"grad_norm": 0.35024622082710266,
"learning_rate": 3.56618711609785e-11,
"loss": 1.33449125289917,
"step": 1356
},
{
"epoch": 1.9824817518248175,
"grad_norm": 0.2447447031736374,
"learning_rate": 3.5531180396548933e-11,
"loss": 1.3325806856155396,
"step": 1358
},
{
"epoch": 1.9854014598540146,
"grad_norm": 0.2641764283180237,
"learning_rate": 3.5400691347658855e-11,
"loss": 1.3506903648376465,
"step": 1360
},
{
"epoch": 1.9883211678832118,
"grad_norm": 0.4542999267578125,
"learning_rate": 3.527040536629927e-11,
"loss": 1.3841180801391602,
"step": 1362
},
{
"epoch": 1.9912408759124087,
"grad_norm": 0.29342707991600037,
"learning_rate": 3.514032380235719e-11,
"loss": 1.4284846782684326,
"step": 1364
},
{
"epoch": 1.994160583941606,
"grad_norm": 0.5308235287666321,
"learning_rate": 3.5010448003601684e-11,
"loss": 1.8175984621047974,
"step": 1366
},
{
"epoch": 1.997080291970803,
"grad_norm": 0.4636997580528259,
"learning_rate": 3.4880779315669865e-11,
"loss": 1.7896119356155396,
"step": 1368
},
{
"epoch": 2.0,
"grad_norm": 0.8472878932952881,
"learning_rate": 3.475131908205308e-11,
"loss": 1.7334907054901123,
"step": 1370
},
{
"epoch": 2.002919708029197,
"grad_norm": 0.27248427271842957,
"learning_rate": 3.4622068644082805e-11,
"loss": 1.4240479469299316,
"step": 1372
},
{
"epoch": 2.0058394160583943,
"grad_norm": 0.21887673437595367,
"learning_rate": 3.44930293409168e-11,
"loss": 1.4103740453720093,
"step": 1374
},
{
"epoch": 2.0087591240875913,
"grad_norm": 0.22277064621448517,
"learning_rate": 3.436420250952537e-11,
"loss": 1.4105690717697144,
"step": 1376
},
{
"epoch": 2.011678832116788,
"grad_norm": 0.24514085054397583,
"learning_rate": 3.423558948467733e-11,
"loss": 1.4243745803833008,
"step": 1378
},
{
"epoch": 2.0145985401459856,
"grad_norm": 0.2835189402103424,
"learning_rate": 3.4107191598926294e-11,
"loss": 1.3502079248428345,
"step": 1380
},
{
"epoch": 2.0175182481751825,
"grad_norm": 0.29972848296165466,
"learning_rate": 3.3979010182596804e-11,
"loss": 1.3972976207733154,
"step": 1382
},
{
"epoch": 2.0204379562043795,
"grad_norm": 0.23961658775806427,
"learning_rate": 3.3851046563770615e-11,
"loss": 1.3070876598358154,
"step": 1384
},
{
"epoch": 2.0233576642335764,
"grad_norm": 0.7350632548332214,
"learning_rate": 3.372330206827281e-11,
"loss": 1.2981512546539307,
"step": 1386
},
{
"epoch": 2.026277372262774,
"grad_norm": 0.3386096656322479,
"learning_rate": 3.3595778019658256e-11,
"loss": 1.2211928367614746,
"step": 1388
},
{
"epoch": 2.0291970802919708,
"grad_norm": 0.20256347954273224,
"learning_rate": 3.346847573919769e-11,
"loss": 1.4054863452911377,
"step": 1390
},
{
"epoch": 2.0321167883211677,
"grad_norm": 0.48250511288642883,
"learning_rate": 3.3341396545864166e-11,
"loss": 1.3012176752090454,
"step": 1392
},
{
"epoch": 2.035036496350365,
"grad_norm": 0.16802747547626495,
"learning_rate": 3.321454175631934e-11,
"loss": 1.249161720275879,
"step": 1394
},
{
"epoch": 2.037956204379562,
"grad_norm": 0.29464009404182434,
"learning_rate": 3.308791268489979e-11,
"loss": 1.303439974784851,
"step": 1396
},
{
"epoch": 2.040875912408759,
"grad_norm": 0.6583344340324402,
"learning_rate": 3.2961510643603477e-11,
"loss": 1.3693758249282837,
"step": 1398
},
{
"epoch": 2.0437956204379564,
"grad_norm": 0.4720025658607483,
"learning_rate": 3.283533694207611e-11,
"loss": 1.371114730834961,
"step": 1400
},
{
"epoch": 2.0467153284671533,
"grad_norm": 0.4682556688785553,
"learning_rate": 3.270939288759758e-11,
"loss": 1.3761752843856812,
"step": 1402
},
{
"epoch": 2.0496350364963503,
"grad_norm": 0.23124952614307404,
"learning_rate": 3.25836797850684e-11,
"loss": 1.1644366979599,
"step": 1404
},
{
"epoch": 2.0525547445255476,
"grad_norm": 0.21318967640399933,
"learning_rate": 3.2458198936996216e-11,
"loss": 1.1261866092681885,
"step": 1406
},
{
"epoch": 2.0554744525547446,
"grad_norm": 0.2597343921661377,
"learning_rate": 3.233295164348229e-11,
"loss": 1.2748457193374634,
"step": 1408
},
{
"epoch": 2.0583941605839415,
"grad_norm": 0.36715397238731384,
"learning_rate": 3.2207939202208046e-11,
"loss": 1.3361799716949463,
"step": 1410
},
{
"epoch": 2.0613138686131385,
"grad_norm": 0.24065475165843964,
"learning_rate": 3.208316290842159e-11,
"loss": 1.361956238746643,
"step": 1412
},
{
"epoch": 2.064233576642336,
"grad_norm": 0.33580994606018066,
"learning_rate": 3.1958624054924346e-11,
"loss": 1.3905457258224487,
"step": 1414
},
{
"epoch": 2.067153284671533,
"grad_norm": 0.4103996753692627,
"learning_rate": 3.183432393205763e-11,
"loss": 1.6151682138442993,
"step": 1416
},
{
"epoch": 2.0700729927007298,
"grad_norm": 0.2529504597187042,
"learning_rate": 3.171026382768923e-11,
"loss": 1.2452044486999512,
"step": 1418
},
{
"epoch": 2.072992700729927,
"grad_norm": 0.352078378200531,
"learning_rate": 3.158644502720015e-11,
"loss": 1.301306962966919,
"step": 1420
},
{
"epoch": 2.075912408759124,
"grad_norm": 0.23457178473472595,
"learning_rate": 3.146286881347128e-11,
"loss": 1.3487284183502197,
"step": 1422
},
{
"epoch": 2.078832116788321,
"grad_norm": 0.30150166153907776,
"learning_rate": 3.1339536466870046e-11,
"loss": 1.3885747194290161,
"step": 1424
},
{
"epoch": 2.0817518248175184,
"grad_norm": 0.377258837223053,
"learning_rate": 3.121644926523721e-11,
"loss": 1.405432939529419,
"step": 1426
},
{
"epoch": 2.0846715328467154,
"grad_norm": 0.13830381631851196,
"learning_rate": 3.1093608483873524e-11,
"loss": 1.209243893623352,
"step": 1428
},
{
"epoch": 2.0875912408759123,
"grad_norm": 0.18046823143959045,
"learning_rate": 3.097101539552668e-11,
"loss": 1.2217992544174194,
"step": 1430
},
{
"epoch": 2.0905109489051097,
"grad_norm": 0.22244705259799957,
"learning_rate": 3.0848671270377985e-11,
"loss": 1.2246205806732178,
"step": 1432
},
{
"epoch": 2.0934306569343066,
"grad_norm": 0.21235236525535583,
"learning_rate": 3.0726577376029264e-11,
"loss": 1.2530802488327026,
"step": 1434
},
{
"epoch": 2.0963503649635036,
"grad_norm": 0.2744048237800598,
"learning_rate": 3.0604734977489704e-11,
"loss": 1.2528806924819946,
"step": 1436
},
{
"epoch": 2.0992700729927005,
"grad_norm": 0.16244126856327057,
"learning_rate": 3.048314533716277e-11,
"loss": 1.2494616508483887,
"step": 1438
},
{
"epoch": 2.102189781021898,
"grad_norm": 0.2635323107242584,
"learning_rate": 3.03618097148331e-11,
"loss": 1.31374192237854,
"step": 1440
},
{
"epoch": 2.105109489051095,
"grad_norm": 0.343732625246048,
"learning_rate": 3.0240729367653456e-11,
"loss": 1.4676549434661865,
"step": 1442
},
{
"epoch": 2.108029197080292,
"grad_norm": 0.6998947262763977,
"learning_rate": 3.0119905550131735e-11,
"loss": 1.4312711954116821,
"step": 1444
},
{
"epoch": 2.110948905109489,
"grad_norm": 0.3323758840560913,
"learning_rate": 2.9999339514117915e-11,
"loss": 1.3852202892303467,
"step": 1446
},
{
"epoch": 2.113868613138686,
"grad_norm": 0.3845193386077881,
"learning_rate": 2.987903250879109e-11,
"loss": 1.4576423168182373,
"step": 1448
},
{
"epoch": 2.116788321167883,
"grad_norm": 0.3180224597454071,
"learning_rate": 2.975898578064662e-11,
"loss": 1.3024555444717407,
"step": 1450
},
{
"epoch": 2.1197080291970805,
"grad_norm": 0.4924963116645813,
"learning_rate": 2.9639200573483075e-11,
"loss": 1.2051639556884766,
"step": 1452
},
{
"epoch": 2.1226277372262774,
"grad_norm": 0.2880760729312897,
"learning_rate": 2.951967812838946e-11,
"loss": 1.1336140632629395,
"step": 1454
},
{
"epoch": 2.1255474452554743,
"grad_norm": 0.42956921458244324,
"learning_rate": 2.940041968373229e-11,
"loss": 1.2710269689559937,
"step": 1456
},
{
"epoch": 2.1284671532846717,
"grad_norm": 1.3568767309188843,
"learning_rate": 2.928142647514281e-11,
"loss": 1.1412277221679688,
"step": 1458
},
{
"epoch": 2.1313868613138687,
"grad_norm": 0.1728566586971283,
"learning_rate": 2.916269973550413e-11,
"loss": 1.3375706672668457,
"step": 1460
},
{
"epoch": 2.1343065693430656,
"grad_norm": 0.2527841627597809,
"learning_rate": 2.904424069493853e-11,
"loss": 1.3697013854980469,
"step": 1462
},
{
"epoch": 2.137226277372263,
"grad_norm": 0.18707573413848877,
"learning_rate": 2.892605058079464e-11,
"loss": 1.4107820987701416,
"step": 1464
},
{
"epoch": 2.14014598540146,
"grad_norm": 0.7027086019515991,
"learning_rate": 2.8808130617634766e-11,
"loss": 1.3240883350372314,
"step": 1466
},
{
"epoch": 2.143065693430657,
"grad_norm": 0.2566525340080261,
"learning_rate": 2.8690482027222204e-11,
"loss": 1.3276679515838623,
"step": 1468
},
{
"epoch": 2.145985401459854,
"grad_norm": 0.33008280396461487,
"learning_rate": 2.857310602850854e-11,
"loss": 1.4535688161849976,
"step": 1470
},
{
"epoch": 2.1489051094890512,
"grad_norm": 0.2940777540206909,
"learning_rate": 2.845600383762107e-11,
"loss": 1.3314485549926758,
"step": 1472
},
{
"epoch": 2.151824817518248,
"grad_norm": 0.263931542634964,
"learning_rate": 2.833917666785017e-11,
"loss": 1.3149549961090088,
"step": 1474
},
{
"epoch": 2.154744525547445,
"grad_norm": 0.23515236377716064,
"learning_rate": 2.8222625729636774e-11,
"loss": 1.3796401023864746,
"step": 1476
},
{
"epoch": 2.1576642335766425,
"grad_norm": 3.0907609462738037,
"learning_rate": 2.8106352230559756e-11,
"loss": 1.441535472869873,
"step": 1478
},
{
"epoch": 2.1605839416058394,
"grad_norm": 0.28067415952682495,
"learning_rate": 2.799035737532344e-11,
"loss": 1.3874502182006836,
"step": 1480
},
{
"epoch": 2.1635036496350364,
"grad_norm": 0.2478429526090622,
"learning_rate": 2.7874642365745163e-11,
"loss": 1.433276891708374,
"step": 1482
},
{
"epoch": 2.1664233576642338,
"grad_norm": 0.19423043727874756,
"learning_rate": 2.7759208400742797e-11,
"loss": 1.2716385126113892,
"step": 1484
},
{
"epoch": 2.1693430656934307,
"grad_norm": 0.4264819622039795,
"learning_rate": 2.764405667632231e-11,
"loss": 1.450852394104004,
"step": 1486
},
{
"epoch": 2.1722627737226277,
"grad_norm": 0.2181134819984436,
"learning_rate": 2.7529188385565386e-11,
"loss": 1.2895259857177734,
"step": 1488
},
{
"epoch": 2.1751824817518246,
"grad_norm": 0.2849787175655365,
"learning_rate": 2.741460471861708e-11,
"loss": 1.4493159055709839,
"step": 1490
},
{
"epoch": 2.178102189781022,
"grad_norm": 0.20413418114185333,
"learning_rate": 2.730030686267347e-11,
"loss": 1.3337498903274536,
"step": 1492
},
{
"epoch": 2.181021897810219,
"grad_norm": 0.2518576383590698,
"learning_rate": 2.7186296001969352e-11,
"loss": 1.3332432508468628,
"step": 1494
},
{
"epoch": 2.183941605839416,
"grad_norm": 0.4149719774723053,
"learning_rate": 2.7072573317765974e-11,
"loss": 1.3538711071014404,
"step": 1496
},
{
"epoch": 2.1868613138686133,
"grad_norm": 0.3612917363643646,
"learning_rate": 2.695913998833881e-11,
"loss": 1.3591725826263428,
"step": 1498
},
{
"epoch": 2.18978102189781,
"grad_norm": 0.26884135603904724,
"learning_rate": 2.6845997188965365e-11,
"loss": 1.304358720779419,
"step": 1500
},
{
"epoch": 2.192700729927007,
"grad_norm": 0.34493714570999146,
"learning_rate": 2.673314609191291e-11,
"loss": 1.3458648920059204,
"step": 1502
},
{
"epoch": 2.1956204379562045,
"grad_norm": 0.15932005643844604,
"learning_rate": 2.662058786642646e-11,
"loss": 1.371425986289978,
"step": 1504
},
{
"epoch": 2.1985401459854015,
"grad_norm": 0.2096133679151535,
"learning_rate": 2.6508323678716584e-11,
"loss": 1.2557461261749268,
"step": 1506
},
{
"epoch": 2.2014598540145984,
"grad_norm": 0.253741592168808,
"learning_rate": 2.6396354691947322e-11,
"loss": 1.1103127002716064,
"step": 1508
},
{
"epoch": 2.204379562043796,
"grad_norm": 0.23460233211517334,
"learning_rate": 2.628468206622421e-11,
"loss": 1.3412158489227295,
"step": 1510
},
{
"epoch": 2.2072992700729928,
"grad_norm": 0.19601471722126007,
"learning_rate": 2.617330695858212e-11,
"loss": 1.2751479148864746,
"step": 1512
},
{
"epoch": 2.2102189781021897,
"grad_norm": 0.26254287362098694,
"learning_rate": 2.6062230522973407e-11,
"loss": 1.2706090211868286,
"step": 1514
},
{
"epoch": 2.213138686131387,
"grad_norm": 0.2905943691730499,
"learning_rate": 2.5951453910255874e-11,
"loss": 1.3488496541976929,
"step": 1516
},
{
"epoch": 2.216058394160584,
"grad_norm": 0.24975769221782684,
"learning_rate": 2.5840978268180892e-11,
"loss": 1.267167329788208,
"step": 1518
},
{
"epoch": 2.218978102189781,
"grad_norm": 0.34441080689430237,
"learning_rate": 2.5730804741381475e-11,
"loss": 1.3788913488388062,
"step": 1520
},
{
"epoch": 2.221897810218978,
"grad_norm": 0.5073427557945251,
"learning_rate": 2.5620934471360446e-11,
"loss": 1.5342570543289185,
"step": 1522
},
{
"epoch": 2.2248175182481753,
"grad_norm": 0.3326243758201599,
"learning_rate": 2.5511368596478575e-11,
"loss": 1.589991807937622,
"step": 1524
},
{
"epoch": 2.2277372262773723,
"grad_norm": 0.7213902473449707,
"learning_rate": 2.5402108251942813e-11,
"loss": 1.393171787261963,
"step": 1526
},
{
"epoch": 2.230656934306569,
"grad_norm": 0.3098500072956085,
"learning_rate": 2.529315456979457e-11,
"loss": 1.482409119606018,
"step": 1528
},
{
"epoch": 2.2335766423357666,
"grad_norm": 1.4665156602859497,
"learning_rate": 2.5184508678897894e-11,
"loss": 1.541229248046875,
"step": 1530
},
{
"epoch": 2.2364963503649635,
"grad_norm": 0.3888106942176819,
"learning_rate": 2.5076171704927847e-11,
"loss": 1.5156506299972534,
"step": 1532
},
{
"epoch": 2.2394160583941605,
"grad_norm": 0.4767511487007141,
"learning_rate": 2.4968144770358785e-11,
"loss": 1.474044919013977,
"step": 1534
},
{
"epoch": 2.242335766423358,
"grad_norm": 0.671013593673706,
"learning_rate": 2.4860428994452785e-11,
"loss": 1.3635804653167725,
"step": 1536
},
{
"epoch": 2.245255474452555,
"grad_norm": 0.5932507514953613,
"learning_rate": 2.4753025493248032e-11,
"loss": 1.5467637777328491,
"step": 1538
},
{
"epoch": 2.2481751824817517,
"grad_norm": 0.4003777503967285,
"learning_rate": 2.464593537954722e-11,
"loss": 1.075831651687622,
"step": 1540
},
{
"epoch": 2.2510948905109487,
"grad_norm": 2.531550168991089,
"learning_rate": 2.453915976290607e-11,
"loss": 1.1271113157272339,
"step": 1542
},
{
"epoch": 2.254014598540146,
"grad_norm": 0.5367615222930908,
"learning_rate": 2.4432699749621813e-11,
"loss": 1.1544005870819092,
"step": 1544
},
{
"epoch": 2.256934306569343,
"grad_norm": 0.5404489636421204,
"learning_rate": 2.4326556442721715e-11,
"loss": 1.0723111629486084,
"step": 1546
},
{
"epoch": 2.25985401459854,
"grad_norm": 0.4654960632324219,
"learning_rate": 2.4220730941951673e-11,
"loss": 1.3666132688522339,
"step": 1548
},
{
"epoch": 2.2627737226277373,
"grad_norm": 0.4631255567073822,
"learning_rate": 2.411522434376481e-11,
"loss": 1.3214212656021118,
"step": 1550
},
{
"epoch": 2.2656934306569343,
"grad_norm": 3.1865177154541016,
"learning_rate": 2.4010037741310103e-11,
"loss": 1.2218841314315796,
"step": 1552
},
{
"epoch": 2.2686131386861312,
"grad_norm": 0.3000134527683258,
"learning_rate": 2.3905172224421092e-11,
"loss": 1.1274064779281616,
"step": 1554
},
{
"epoch": 2.2715328467153286,
"grad_norm": 0.2830018699169159,
"learning_rate": 2.3800628879604523e-11,
"loss": 0.938373327255249,
"step": 1556
},
{
"epoch": 2.2744525547445256,
"grad_norm": 0.2360849827528,
"learning_rate": 2.3696408790029166e-11,
"loss": 1.1650950908660889,
"step": 1558
},
{
"epoch": 2.2773722627737225,
"grad_norm": 0.34656640887260437,
"learning_rate": 2.3592513035514534e-11,
"loss": 1.2450108528137207,
"step": 1560
},
{
"epoch": 2.28029197080292,
"grad_norm": 0.31296175718307495,
"learning_rate": 2.348894269251978e-11,
"loss": 0.9854579567909241,
"step": 1562
},
{
"epoch": 2.283211678832117,
"grad_norm": 0.4444359540939331,
"learning_rate": 2.3385698834132398e-11,
"loss": 1.1185884475708008,
"step": 1564
},
{
"epoch": 2.286131386861314,
"grad_norm": 0.5716680288314819,
"learning_rate": 2.3282782530057236e-11,
"loss": 1.3133447170257568,
"step": 1566
},
{
"epoch": 2.289051094890511,
"grad_norm": 0.5097543597221375,
"learning_rate": 2.3180194846605366e-11,
"loss": 1.203965187072754,
"step": 1568
},
{
"epoch": 2.291970802919708,
"grad_norm": 0.3655783534049988,
"learning_rate": 2.307793684668303e-11,
"loss": 1.262579083442688,
"step": 1570
},
{
"epoch": 2.294890510948905,
"grad_norm": 0.4481443166732788,
"learning_rate": 2.297600958978064e-11,
"loss": 1.3191816806793213,
"step": 1572
},
{
"epoch": 2.297810218978102,
"grad_norm": 0.2837509512901306,
"learning_rate": 2.2874414131961783e-11,
"loss": 1.0644941329956055,
"step": 1574
},
{
"epoch": 2.3007299270072994,
"grad_norm": 0.44186243414878845,
"learning_rate": 2.277315152585231e-11,
"loss": 1.1767611503601074,
"step": 1576
},
{
"epoch": 2.3036496350364963,
"grad_norm": 0.476301908493042,
"learning_rate": 2.2672222820629375e-11,
"loss": 1.5055090188980103,
"step": 1578
},
{
"epoch": 2.3065693430656933,
"grad_norm": 0.5160381197929382,
"learning_rate": 2.2571629062010654e-11,
"loss": 1.3974980115890503,
"step": 1580
},
{
"epoch": 2.3094890510948907,
"grad_norm": 0.22894681990146637,
"learning_rate": 2.2471371292243415e-11,
"loss": 1.4789299964904785,
"step": 1582
},
{
"epoch": 2.3124087591240876,
"grad_norm": 0.24650397896766663,
"learning_rate": 2.2371450550093786e-11,
"loss": 1.438536524772644,
"step": 1584
},
{
"epoch": 2.3153284671532846,
"grad_norm": 0.26311492919921875,
"learning_rate": 2.227186787083593e-11,
"loss": 1.152450680732727,
"step": 1586
},
{
"epoch": 2.318248175182482,
"grad_norm": 0.3128260374069214,
"learning_rate": 2.2172624286241394e-11,
"loss": 1.2790557146072388,
"step": 1588
},
{
"epoch": 2.321167883211679,
"grad_norm": 0.5634860396385193,
"learning_rate": 2.2073720824568366e-11,
"loss": 1.2126829624176025,
"step": 1590
},
{
"epoch": 2.324087591240876,
"grad_norm": 0.3091767430305481,
"learning_rate": 2.1975158510551046e-11,
"loss": 1.0758181810379028,
"step": 1592
},
{
"epoch": 2.3270072992700728,
"grad_norm": 0.5447919964790344,
"learning_rate": 2.1876938365389005e-11,
"loss": 1.5189749002456665,
"step": 1594
},
{
"epoch": 2.32992700729927,
"grad_norm": 0.25837060809135437,
"learning_rate": 2.1779061406736623e-11,
"loss": 1.4296108484268188,
"step": 1596
},
{
"epoch": 2.332846715328467,
"grad_norm": 0.29859817028045654,
"learning_rate": 2.1681528648692546e-11,
"loss": 1.3460187911987305,
"step": 1598
},
{
"epoch": 2.335766423357664,
"grad_norm": 0.3790282607078552,
"learning_rate": 2.1584341101789163e-11,
"loss": 1.0567677021026611,
"step": 1600
},
{
"epoch": 2.3386861313868614,
"grad_norm": 0.37978944182395935,
"learning_rate": 2.1487499772982154e-11,
"loss": 1.4715806245803833,
"step": 1602
},
{
"epoch": 2.3416058394160584,
"grad_norm": 0.33095240592956543,
"learning_rate": 2.1391005665640057e-11,
"loss": 1.4167277812957764,
"step": 1604
},
{
"epoch": 2.3445255474452553,
"grad_norm": 0.7214593291282654,
"learning_rate": 2.129485977953386e-11,
"loss": 1.5120526552200317,
"step": 1606
},
{
"epoch": 2.3474452554744527,
"grad_norm": 0.575445830821991,
"learning_rate": 2.1199063110826615e-11,
"loss": 1.4879270792007446,
"step": 1608
},
{
"epoch": 2.3503649635036497,
"grad_norm": 0.30642205476760864,
"learning_rate": 2.1103616652063197e-11,
"loss": 1.3315616846084595,
"step": 1610
},
{
"epoch": 2.3532846715328466,
"grad_norm": 1.8541874885559082,
"learning_rate": 2.1008521392159927e-11,
"loss": 1.2089598178863525,
"step": 1612
},
{
"epoch": 2.356204379562044,
"grad_norm": 0.6219841241836548,
"learning_rate": 2.0913778316394434e-11,
"loss": 1.2917388677597046,
"step": 1614
},
{
"epoch": 2.359124087591241,
"grad_norm": 0.7050977945327759,
"learning_rate": 2.081938840639533e-11,
"loss": 1.3587377071380615,
"step": 1616
},
{
"epoch": 2.362043795620438,
"grad_norm": 0.42739561200141907,
"learning_rate": 2.072535264013209e-11,
"loss": 1.522105097770691,
"step": 1618
},
{
"epoch": 2.3649635036496353,
"grad_norm": 1.166688323020935,
"learning_rate": 2.063167199190495e-11,
"loss": 1.4914265871047974,
"step": 1620
},
{
"epoch": 2.367883211678832,
"grad_norm": 0.29809263348579407,
"learning_rate": 2.053834743233477e-11,
"loss": 1.6098909378051758,
"step": 1622
},
{
"epoch": 2.370802919708029,
"grad_norm": 0.33959904313087463,
"learning_rate": 2.0445379928353005e-11,
"loss": 1.443519949913025,
"step": 1624
},
{
"epoch": 2.373722627737226,
"grad_norm": 0.7890218496322632,
"learning_rate": 2.035277044319165e-11,
"loss": 1.422911524772644,
"step": 1626
},
{
"epoch": 2.3766423357664235,
"grad_norm": 0.3700925409793854,
"learning_rate": 2.026051993637332e-11,
"loss": 1.1868336200714111,
"step": 1628
},
{
"epoch": 2.3795620437956204,
"grad_norm": 0.25010111927986145,
"learning_rate": 2.0168629363701215e-11,
"loss": 1.1865566968917847,
"step": 1630
},
{
"epoch": 2.3824817518248174,
"grad_norm": 0.4608599841594696,
"learning_rate": 2.0077099677249334e-11,
"loss": 1.4894611835479736,
"step": 1632
},
{
"epoch": 2.3854014598540147,
"grad_norm": 0.3453768193721771,
"learning_rate": 1.9985931825352528e-11,
"loss": 1.3610191345214844,
"step": 1634
},
{
"epoch": 2.3883211678832117,
"grad_norm": 0.323233038187027,
"learning_rate": 1.989512675259668e-11,
"loss": 1.4355524778366089,
"step": 1636
},
{
"epoch": 2.3912408759124086,
"grad_norm": 0.3967125415802002,
"learning_rate": 1.980468539980897e-11,
"loss": 1.1551700830459595,
"step": 1638
},
{
"epoch": 2.394160583941606,
"grad_norm": 0.2280547320842743,
"learning_rate": 1.9714608704048038e-11,
"loss": 1.136304497718811,
"step": 1640
},
{
"epoch": 2.397080291970803,
"grad_norm": 0.2169790118932724,
"learning_rate": 1.9624897598594387e-11,
"loss": 1.1340363025665283,
"step": 1642
},
{
"epoch": 2.4,
"grad_norm": 0.2436823695898056,
"learning_rate": 1.953555301294062e-11,
"loss": 1.1512538194656372,
"step": 1644
},
{
"epoch": 2.402919708029197,
"grad_norm": 0.2881166636943817,
"learning_rate": 1.9446575872781863e-11,
"loss": 1.4590939283370972,
"step": 1646
},
{
"epoch": 2.4058394160583942,
"grad_norm": 0.7346351742744446,
"learning_rate": 1.9357967100006153e-11,
"loss": 1.4373379945755005,
"step": 1648
},
{
"epoch": 2.408759124087591,
"grad_norm": 0.2763451039791107,
"learning_rate": 1.9269727612684882e-11,
"loss": 1.4031740427017212,
"step": 1650
},
{
"epoch": 2.411678832116788,
"grad_norm": 0.3692055642604828,
"learning_rate": 1.9181858325063297e-11,
"loss": 1.413076400756836,
"step": 1652
},
{
"epoch": 2.4145985401459855,
"grad_norm": 0.4550726115703583,
"learning_rate": 1.9094360147551022e-11,
"loss": 1.4948763847351074,
"step": 1654
},
{
"epoch": 2.4175182481751825,
"grad_norm": 0.3476276695728302,
"learning_rate": 1.900723398671263e-11,
"loss": 1.2074034214019775,
"step": 1656
},
{
"epoch": 2.4204379562043794,
"grad_norm": 0.2987731993198395,
"learning_rate": 1.892048074525824e-11,
"loss": 1.208572506904602,
"step": 1658
},
{
"epoch": 2.423357664233577,
"grad_norm": 0.30724695324897766,
"learning_rate": 1.8834101322034192e-11,
"loss": 1.3055994510650635,
"step": 1660
},
{
"epoch": 2.4262773722627737,
"grad_norm": 0.3010522723197937,
"learning_rate": 1.874809661201367e-11,
"loss": 1.384286880493164,
"step": 1662
},
{
"epoch": 2.4291970802919707,
"grad_norm": 0.48422113060951233,
"learning_rate": 1.8662467506287496e-11,
"loss": 1.2188966274261475,
"step": 1664
},
{
"epoch": 2.432116788321168,
"grad_norm": 0.22298695147037506,
"learning_rate": 1.8577214892054895e-11,
"loss": 1.1136993169784546,
"step": 1666
},
{
"epoch": 2.435036496350365,
"grad_norm": 0.3978652358055115,
"learning_rate": 1.8492339652614253e-11,
"loss": 1.3406078815460205,
"step": 1668
},
{
"epoch": 2.437956204379562,
"grad_norm": 0.35969844460487366,
"learning_rate": 1.8407842667354008e-11,
"loss": 1.2762571573257446,
"step": 1670
},
{
"epoch": 2.4408759124087593,
"grad_norm": 0.21341219544410706,
"learning_rate": 1.8323724811743495e-11,
"loss": 1.310207724571228,
"step": 1672
},
{
"epoch": 2.4437956204379563,
"grad_norm": 0.27594175934791565,
"learning_rate": 1.8239986957323938e-11,
"loss": 1.4516338109970093,
"step": 1674
},
{
"epoch": 2.4467153284671532,
"grad_norm": 0.5084519386291504,
"learning_rate": 1.8156629971699374e-11,
"loss": 1.572962999343872,
"step": 1676
},
{
"epoch": 2.44963503649635,
"grad_norm": 0.24259944260120392,
"learning_rate": 1.807365471852767e-11,
"loss": 1.2814817428588867,
"step": 1678
},
{
"epoch": 2.4525547445255476,
"grad_norm": 0.32336655259132385,
"learning_rate": 1.7991062057511587e-11,
"loss": 1.2371183633804321,
"step": 1680
},
{
"epoch": 2.4554744525547445,
"grad_norm": 0.3355330526828766,
"learning_rate": 1.7908852844389878e-11,
"loss": 1.329166054725647,
"step": 1682
},
{
"epoch": 2.4583941605839414,
"grad_norm": 0.23916582763195038,
"learning_rate": 1.7827027930928388e-11,
"loss": 1.2250287532806396,
"step": 1684
},
{
"epoch": 2.461313868613139,
"grad_norm": 0.509010910987854,
"learning_rate": 1.7745588164911263e-11,
"loss": 1.5886147022247314,
"step": 1686
},
{
"epoch": 2.4642335766423358,
"grad_norm": 0.24120256304740906,
"learning_rate": 1.766453439013215e-11,
"loss": 1.3899035453796387,
"step": 1688
},
{
"epoch": 2.4671532846715327,
"grad_norm": 0.34605830907821655,
"learning_rate": 1.758386744638546e-11,
"loss": 1.3214712142944336,
"step": 1690
},
{
"epoch": 2.47007299270073,
"grad_norm": 0.47554486989974976,
"learning_rate": 1.7503588169457688e-11,
"loss": 1.3142483234405518,
"step": 1692
},
{
"epoch": 2.472992700729927,
"grad_norm": 0.42094510793685913,
"learning_rate": 1.7423697391118673e-11,
"loss": 1.274762511253357,
"step": 1694
},
{
"epoch": 2.475912408759124,
"grad_norm": 0.9319076538085938,
"learning_rate": 1.7344195939113094e-11,
"loss": 1.3079981803894043,
"step": 1696
},
{
"epoch": 2.478832116788321,
"grad_norm": 0.25632455945014954,
"learning_rate": 1.7265084637151817e-11,
"loss": 1.2449252605438232,
"step": 1698
},
{
"epoch": 2.4817518248175183,
"grad_norm": 0.4942947030067444,
"learning_rate": 1.718636430490338e-11,
"loss": 1.2960014343261719,
"step": 1700
},
{
"epoch": 2.4846715328467153,
"grad_norm": 0.20054902136325836,
"learning_rate": 1.7108035757985506e-11,
"loss": 1.3219631910324097,
"step": 1702
},
{
"epoch": 2.487591240875912,
"grad_norm": 0.34794673323631287,
"learning_rate": 1.7030099807956648e-11,
"loss": 1.2419843673706055,
"step": 1704
},
{
"epoch": 2.4905109489051096,
"grad_norm": 1.3033686876296997,
"learning_rate": 1.695255726230758e-11,
"loss": 1.5983386039733887,
"step": 1706
},
{
"epoch": 2.4934306569343065,
"grad_norm": 0.44935575127601624,
"learning_rate": 1.6875408924453033e-11,
"loss": 1.4845508337020874,
"step": 1708
},
{
"epoch": 2.4963503649635035,
"grad_norm": 0.3333362936973572,
"learning_rate": 1.6798655593723362e-11,
"loss": 1.6482346057891846,
"step": 1710
},
{
"epoch": 2.499270072992701,
"grad_norm": 0.568465530872345,
"learning_rate": 1.6722298065356283e-11,
"loss": 1.282273292541504,
"step": 1712
},
{
"epoch": 2.502189781021898,
"grad_norm": 0.285226434469223,
"learning_rate": 1.6646337130488608e-11,
"loss": 1.2390754222869873,
"step": 1714
},
{
"epoch": 2.5051094890510948,
"grad_norm": 0.35810387134552,
"learning_rate": 1.657077357614808e-11,
"loss": 1.1541006565093994,
"step": 1716
},
{
"epoch": 2.508029197080292,
"grad_norm": 0.22877416014671326,
"learning_rate": 1.649560818524517e-11,
"loss": 1.2132242918014526,
"step": 1718
},
{
"epoch": 2.510948905109489,
"grad_norm": 0.650364875793457,
"learning_rate": 1.6420841736565042e-11,
"loss": 1.4655767679214478,
"step": 1720
},
{
"epoch": 2.513868613138686,
"grad_norm": 2.0470120906829834,
"learning_rate": 1.6346475004759414e-11,
"loss": 1.6555049419403076,
"step": 1722
},
{
"epoch": 2.5167883211678834,
"grad_norm": 0.3082306385040283,
"learning_rate": 1.627250876033853e-11,
"loss": 1.7491514682769775,
"step": 1724
},
{
"epoch": 2.5197080291970804,
"grad_norm": 0.8077488541603088,
"learning_rate": 1.619894376966325e-11,
"loss": 1.3254021406173706,
"step": 1726
},
{
"epoch": 2.5226277372262773,
"grad_norm": 0.48932626843452454,
"learning_rate": 1.612578079493702e-11,
"loss": 1.440732717514038,
"step": 1728
},
{
"epoch": 2.5255474452554747,
"grad_norm": 0.4108842611312866,
"learning_rate": 1.6053020594198053e-11,
"loss": 1.2445178031921387,
"step": 1730
},
{
"epoch": 2.5284671532846716,
"grad_norm": 0.746144711971283,
"learning_rate": 1.598066392131142e-11,
"loss": 1.5634169578552246,
"step": 1732
},
{
"epoch": 2.5313868613138686,
"grad_norm": 0.809232771396637,
"learning_rate": 1.5908711525961263e-11,
"loss": 1.751116156578064,
"step": 1734
},
{
"epoch": 2.5343065693430655,
"grad_norm": 0.31983619928359985,
"learning_rate": 1.5837164153643013e-11,
"loss": 1.1314088106155396,
"step": 1736
},
{
"epoch": 2.537226277372263,
"grad_norm": 0.35743609070777893,
"learning_rate": 1.5766022545655703e-11,
"loss": 1.249573826789856,
"step": 1738
},
{
"epoch": 2.54014598540146,
"grad_norm": 0.25417211651802063,
"learning_rate": 1.569528743909423e-11,
"loss": 1.3246897459030151,
"step": 1740
},
{
"epoch": 2.543065693430657,
"grad_norm": 0.521004319190979,
"learning_rate": 1.5624959566841764e-11,
"loss": 1.5900986194610596,
"step": 1742
},
{
"epoch": 2.5459854014598537,
"grad_norm": 0.5410935878753662,
"learning_rate": 1.5555039657562147e-11,
"loss": 1.7000155448913574,
"step": 1744
},
{
"epoch": 2.548905109489051,
"grad_norm": 0.1969938427209854,
"learning_rate": 1.548552843569231e-11,
"loss": 1.4420926570892334,
"step": 1746
},
{
"epoch": 2.551824817518248,
"grad_norm": 0.9155759215354919,
"learning_rate": 1.541642662143481e-11,
"loss": 1.553064227104187,
"step": 1748
},
{
"epoch": 2.554744525547445,
"grad_norm": 0.33890122175216675,
"learning_rate": 1.5347734930750357e-11,
"loss": 1.3361493349075317,
"step": 1750
},
{
"epoch": 2.5576642335766424,
"grad_norm": 0.2208409309387207,
"learning_rate": 1.5279454075350363e-11,
"loss": 1.1686358451843262,
"step": 1752
},
{
"epoch": 2.5605839416058394,
"grad_norm": 0.2458084523677826,
"learning_rate": 1.521158476268965e-11,
"loss": 0.9916486740112305,
"step": 1754
},
{
"epoch": 2.5635036496350363,
"grad_norm": 0.5456938743591309,
"learning_rate": 1.514412769595899e-11,
"loss": 1.1382449865341187,
"step": 1756
},
{
"epoch": 2.5664233576642337,
"grad_norm": 0.32823365926742554,
"learning_rate": 1.5077083574077948e-11,
"loss": 0.9516903162002563,
"step": 1758
},
{
"epoch": 2.5693430656934306,
"grad_norm": 0.5023689270019531,
"learning_rate": 1.5010453091687567e-11,
"loss": 1.284109115600586,
"step": 1760
},
{
"epoch": 2.5722627737226276,
"grad_norm": 0.30572813749313354,
"learning_rate": 1.494423693914319e-11,
"loss": 1.1433796882629395,
"step": 1762
},
{
"epoch": 2.575182481751825,
"grad_norm": 0.8459102511405945,
"learning_rate": 1.4878435802507326e-11,
"loss": 1.1362444162368774,
"step": 1764
},
{
"epoch": 2.578102189781022,
"grad_norm": 0.44648513197898865,
"learning_rate": 1.48130503635425e-11,
"loss": 1.214508295059204,
"step": 1766
},
{
"epoch": 2.581021897810219,
"grad_norm": 0.1434440016746521,
"learning_rate": 1.474808129970421e-11,
"loss": 1.1404409408569336,
"step": 1768
},
{
"epoch": 2.5839416058394162,
"grad_norm": 0.4274137020111084,
"learning_rate": 1.468352928413392e-11,
"loss": 1.478256106376648,
"step": 1770
},
{
"epoch": 2.586861313868613,
"grad_norm": 0.20402590930461884,
"learning_rate": 1.4619394985652097e-11,
"loss": 1.0844522714614868,
"step": 1772
},
{
"epoch": 2.58978102189781,
"grad_norm": 0.6682690978050232,
"learning_rate": 1.4555679068751232e-11,
"loss": 1.0875853300094604,
"step": 1774
},
{
"epoch": 2.5927007299270075,
"grad_norm": 0.1818685382604599,
"learning_rate": 1.4492382193589005e-11,
"loss": 1.1487854719161987,
"step": 1776
},
{
"epoch": 2.5956204379562045,
"grad_norm": 0.2316112518310547,
"learning_rate": 1.4429505015981392e-11,
"loss": 1.1742403507232666,
"step": 1778
},
{
"epoch": 2.5985401459854014,
"grad_norm": 0.5319095849990845,
"learning_rate": 1.4367048187395926e-11,
"loss": 1.4584426879882812,
"step": 1780
},
{
"epoch": 2.601459854014599,
"grad_norm": 0.20086680352687836,
"learning_rate": 1.430501235494493e-11,
"loss": 1.3289785385131836,
"step": 1782
},
{
"epoch": 2.6043795620437957,
"grad_norm": 0.12394100427627563,
"learning_rate": 1.4243398161378788e-11,
"loss": 1.0840532779693604,
"step": 1784
},
{
"epoch": 2.6072992700729927,
"grad_norm": 0.44066891074180603,
"learning_rate": 1.418220624507931e-11,
"loss": 1.1645475625991821,
"step": 1786
},
{
"epoch": 2.61021897810219,
"grad_norm": 0.5535151362419128,
"learning_rate": 1.412143724005311e-11,
"loss": 1.4112097024917603,
"step": 1788
},
{
"epoch": 2.613138686131387,
"grad_norm": 0.5862986445426941,
"learning_rate": 1.4061091775925042e-11,
"loss": 1.57581627368927,
"step": 1790
},
{
"epoch": 2.616058394160584,
"grad_norm": 0.7794012427330017,
"learning_rate": 1.4001170477931665e-11,
"loss": 0.9675093293190002,
"step": 1792
},
{
"epoch": 2.618978102189781,
"grad_norm": 0.39823436737060547,
"learning_rate": 1.3941673966914778e-11,
"loss": 1.2986432313919067,
"step": 1794
},
{
"epoch": 2.621897810218978,
"grad_norm": 0.34392768144607544,
"learning_rate": 1.3882602859314983e-11,
"loss": 1.3523763418197632,
"step": 1796
},
{
"epoch": 2.624817518248175,
"grad_norm": 0.7295415997505188,
"learning_rate": 1.3823957767165299e-11,
"loss": 1.1150239706039429,
"step": 1798
},
{
"epoch": 2.627737226277372,
"grad_norm": 0.8080652952194214,
"learning_rate": 1.3765739298084793e-11,
"loss": 1.1515074968338013,
"step": 1800
},
{
"epoch": 2.630656934306569,
"grad_norm": 0.3648310601711273,
"learning_rate": 1.3707948055272349e-11,
"loss": 1.0616378784179688,
"step": 1802
},
{
"epoch": 2.6335766423357665,
"grad_norm": 0.281863272190094,
"learning_rate": 1.3650584637500337e-11,
"loss": 1.4396766424179077,
"step": 1804
},
{
"epoch": 2.6364963503649634,
"grad_norm": 0.33628129959106445,
"learning_rate": 1.3593649639108521e-11,
"loss": 1.3579171895980835,
"step": 1806
},
{
"epoch": 2.6394160583941604,
"grad_norm": 0.43391627073287964,
"learning_rate": 1.3537143649997743e-11,
"loss": 1.4490206241607666,
"step": 1808
},
{
"epoch": 2.6423357664233578,
"grad_norm": 0.38222044706344604,
"learning_rate": 1.3481067255623958e-11,
"loss": 1.1668449640274048,
"step": 1810
},
{
"epoch": 2.6452554744525547,
"grad_norm": 0.3526589870452881,
"learning_rate": 1.3425421036992098e-11,
"loss": 1.387364387512207,
"step": 1812
},
{
"epoch": 2.6481751824817517,
"grad_norm": 0.2548363506793976,
"learning_rate": 1.337020557065006e-11,
"loss": 1.2788268327713013,
"step": 1814
},
{
"epoch": 2.651094890510949,
"grad_norm": 0.41914018988609314,
"learning_rate": 1.3315421428682727e-11,
"loss": 1.3902523517608643,
"step": 1816
},
{
"epoch": 2.654014598540146,
"grad_norm": 0.45936498045921326,
"learning_rate": 1.326106917870607e-11,
"loss": 1.1867153644561768,
"step": 1818
},
{
"epoch": 2.656934306569343,
"grad_norm": 0.3785250782966614,
"learning_rate": 1.320714938386125e-11,
"loss": 1.2837830781936646,
"step": 1820
},
{
"epoch": 2.6598540145985403,
"grad_norm": 0.5063408017158508,
"learning_rate": 1.3153662602808731e-11,
"loss": 1.2772409915924072,
"step": 1822
},
{
"epoch": 2.6627737226277373,
"grad_norm": 0.42612773180007935,
"learning_rate": 1.3100609389722604e-11,
"loss": 1.4097553491592407,
"step": 1824
},
{
"epoch": 2.665693430656934,
"grad_norm": 1.5975621938705444,
"learning_rate": 1.3047990294284753e-11,
"loss": 1.1698918342590332,
"step": 1826
},
{
"epoch": 2.6686131386861316,
"grad_norm": 0.5364381074905396,
"learning_rate": 1.29958058616792e-11,
"loss": 1.182719349861145,
"step": 1828
},
{
"epoch": 2.6715328467153285,
"grad_norm": 0.26085424423217773,
"learning_rate": 1.2944056632586418e-11,
"loss": 1.4151949882507324,
"step": 1830
},
{
"epoch": 2.6744525547445255,
"grad_norm": 0.4390961825847626,
"learning_rate": 1.2892743143177793e-11,
"loss": 1.6285353899002075,
"step": 1832
},
{
"epoch": 2.677372262773723,
"grad_norm": 1.3621900081634521,
"learning_rate": 1.284186592511e-11,
"loss": 1.251800775527954,
"step": 1834
},
{
"epoch": 2.68029197080292,
"grad_norm": 0.5485129356384277,
"learning_rate": 1.2791425505519557e-11,
"loss": 1.1241846084594727,
"step": 1836
},
{
"epoch": 2.6832116788321168,
"grad_norm": 0.24132966995239258,
"learning_rate": 1.2741422407017312e-11,
"loss": 1.27989661693573,
"step": 1838
},
{
"epoch": 2.686131386861314,
"grad_norm": 2.0611886978149414,
"learning_rate": 1.2691857147683055e-11,
"loss": 1.456813097000122,
"step": 1840
},
{
"epoch": 2.689051094890511,
"grad_norm": 0.25726816058158875,
"learning_rate": 1.2642730241060149e-11,
"loss": 1.4273651838302612,
"step": 1842
},
{
"epoch": 2.691970802919708,
"grad_norm": 0.6684473752975464,
"learning_rate": 1.2594042196150196e-11,
"loss": 1.4390050172805786,
"step": 1844
},
{
"epoch": 2.694890510948905,
"grad_norm": 0.31177812814712524,
"learning_rate": 1.254579351740779e-11,
"loss": 1.2763824462890625,
"step": 1846
},
{
"epoch": 2.697810218978102,
"grad_norm": 0.4132930636405945,
"learning_rate": 1.2497984704735244e-11,
"loss": 1.4589380025863647,
"step": 1848
},
{
"epoch": 2.7007299270072993,
"grad_norm": 0.5510165691375732,
"learning_rate": 1.2450616253477472e-11,
"loss": 1.4122458696365356,
"step": 1850
},
{
"epoch": 2.7036496350364962,
"grad_norm": 0.7220962643623352,
"learning_rate": 1.2403688654416788e-11,
"loss": 1.4138315916061401,
"step": 1852
},
{
"epoch": 2.706569343065693,
"grad_norm": 0.33070141077041626,
"learning_rate": 1.2357202393767884e-11,
"loss": 1.2186477184295654,
"step": 1854
},
{
"epoch": 2.7094890510948906,
"grad_norm": 0.5884044170379639,
"learning_rate": 1.2311157953172753e-11,
"loss": 1.546258568763733,
"step": 1856
},
{
"epoch": 2.7124087591240875,
"grad_norm": 0.2696082592010498,
"learning_rate": 1.2265555809695725e-11,
"loss": 1.3984894752502441,
"step": 1858
},
{
"epoch": 2.7153284671532845,
"grad_norm": 0.38939806818962097,
"learning_rate": 1.2220396435818494e-11,
"loss": 1.4466756582260132,
"step": 1860
},
{
"epoch": 2.718248175182482,
"grad_norm": 0.5408584475517273,
"learning_rate": 1.217568029943524e-11,
"loss": 1.4394067525863647,
"step": 1862
},
{
"epoch": 2.721167883211679,
"grad_norm": 0.3784240782260895,
"learning_rate": 1.2131407863847788e-11,
"loss": 1.4045777320861816,
"step": 1864
},
{
"epoch": 2.7240875912408757,
"grad_norm": 0.4126115143299103,
"learning_rate": 1.2087579587760794e-11,
"loss": 1.4569664001464844,
"step": 1866
},
{
"epoch": 2.727007299270073,
"grad_norm": 0.40902724862098694,
"learning_rate": 1.2044195925277e-11,
"loss": 1.3677961826324463,
"step": 1868
},
{
"epoch": 2.72992700729927,
"grad_norm": 1.7693758010864258,
"learning_rate": 1.2001257325892525e-11,
"loss": 1.5177785158157349,
"step": 1870
},
{
"epoch": 2.732846715328467,
"grad_norm": 0.592065691947937,
"learning_rate": 1.1958764234492219e-11,
"loss": 1.395766258239746,
"step": 1872
},
{
"epoch": 2.7357664233576644,
"grad_norm": 0.5635214447975159,
"learning_rate": 1.1916717091345023e-11,
"loss": 1.3134100437164307,
"step": 1874
},
{
"epoch": 2.7386861313868613,
"grad_norm": 0.41125619411468506,
"learning_rate": 1.1875116332099453e-11,
"loss": 1.3772425651550293,
"step": 1876
},
{
"epoch": 2.7416058394160583,
"grad_norm": 0.35679513216018677,
"learning_rate": 1.1833962387779047e-11,
"loss": 1.4052398204803467,
"step": 1878
},
{
"epoch": 2.7445255474452557,
"grad_norm": 0.26886072754859924,
"learning_rate": 1.1793255684777923e-11,
"loss": 1.4129647016525269,
"step": 1880
},
{
"epoch": 2.7474452554744526,
"grad_norm": 0.2874450087547302,
"learning_rate": 1.1752996644856346e-11,
"loss": 1.4309979677200317,
"step": 1882
},
{
"epoch": 2.7503649635036496,
"grad_norm": 0.5160027146339417,
"learning_rate": 1.1713185685136364e-11,
"loss": 1.3455803394317627,
"step": 1884
},
{
"epoch": 2.753284671532847,
"grad_norm": 0.2807236313819885,
"learning_rate": 1.167382321809749e-11,
"loss": 1.4191386699676514,
"step": 1886
},
{
"epoch": 2.756204379562044,
"grad_norm": 0.3917008638381958,
"learning_rate": 1.1634909651572425e-11,
"loss": 1.3121984004974365,
"step": 1888
},
{
"epoch": 2.759124087591241,
"grad_norm": 0.3935042917728424,
"learning_rate": 1.1596445388742837e-11,
"loss": 1.4232765436172485,
"step": 1890
},
{
"epoch": 2.7620437956204382,
"grad_norm": 0.2913266718387604,
"learning_rate": 1.1558430828135168e-11,
"loss": 1.4180057048797607,
"step": 1892
},
{
"epoch": 2.764963503649635,
"grad_norm": 0.24990983307361603,
"learning_rate": 1.1520866363616524e-11,
"loss": 1.3848999738693237,
"step": 1894
},
{
"epoch": 2.767883211678832,
"grad_norm": 2.3751542568206787,
"learning_rate": 1.1483752384390583e-11,
"loss": 1.5882635116577148,
"step": 1896
},
{
"epoch": 2.770802919708029,
"grad_norm": 0.4262540340423584,
"learning_rate": 1.1447089274993575e-11,
"loss": 1.3331663608551025,
"step": 1898
},
{
"epoch": 2.7737226277372264,
"grad_norm": 0.44628793001174927,
"learning_rate": 1.1410877415290269e-11,
"loss": 1.2856853008270264,
"step": 1900
},
{
"epoch": 2.7766423357664234,
"grad_norm": 0.39279812574386597,
"learning_rate": 1.1375117180470078e-11,
"loss": 1.4837859869003296,
"step": 1902
},
{
"epoch": 2.7795620437956203,
"grad_norm": 0.3202086389064789,
"learning_rate": 1.133980894104314e-11,
"loss": 1.2985246181488037,
"step": 1904
},
{
"epoch": 2.7824817518248173,
"grad_norm": 0.5055567622184753,
"learning_rate": 1.1304953062836486e-11,
"loss": 1.2733244895935059,
"step": 1906
},
{
"epoch": 2.7854014598540147,
"grad_norm": 0.42744991183280945,
"learning_rate": 1.1270549906990256e-11,
"loss": 1.0509121417999268,
"step": 1908
},
{
"epoch": 2.7883211678832116,
"grad_norm": 0.3167889416217804,
"learning_rate": 1.1236599829953968e-11,
"loss": 1.095268964767456,
"step": 1910
},
{
"epoch": 2.7912408759124085,
"grad_norm": 0.2934759557247162,
"learning_rate": 1.1203103183482787e-11,
"loss": 1.2570604085922241,
"step": 1912
},
{
"epoch": 2.794160583941606,
"grad_norm": 0.3890041410923004,
"learning_rate": 1.1170060314633928e-11,
"loss": 1.4037665128707886,
"step": 1914
},
{
"epoch": 2.797080291970803,
"grad_norm": 0.14640799164772034,
"learning_rate": 1.1137471565763024e-11,
"loss": 1.2871328592300415,
"step": 1916
},
{
"epoch": 2.8,
"grad_norm": 0.1849203109741211,
"learning_rate": 1.1105337274520589e-11,
"loss": 1.1548789739608765,
"step": 1918
},
{
"epoch": 2.802919708029197,
"grad_norm": 0.18686549365520477,
"learning_rate": 1.1073657773848535e-11,
"loss": 1.1780543327331543,
"step": 1920
},
{
"epoch": 2.805839416058394,
"grad_norm": 0.3322215676307678,
"learning_rate": 1.10424333919767e-11,
"loss": 1.088761329650879,
"step": 1922
},
{
"epoch": 2.808759124087591,
"grad_norm": 0.18809480965137482,
"learning_rate": 1.1011664452419465e-11,
"loss": 1.0823053121566772,
"step": 1924
},
{
"epoch": 2.8116788321167885,
"grad_norm": 0.17418228089809418,
"learning_rate": 1.0981351273972383e-11,
"loss": 1.0169756412506104,
"step": 1926
},
{
"epoch": 2.8145985401459854,
"grad_norm": 0.5126947164535522,
"learning_rate": 1.09514941707089e-11,
"loss": 0.994587242603302,
"step": 1928
},
{
"epoch": 2.8175182481751824,
"grad_norm": 0.9765654802322388,
"learning_rate": 1.0922093451977073e-11,
"loss": 1.0193537473678589,
"step": 1930
},
{
"epoch": 2.8204379562043798,
"grad_norm": 0.20624187588691711,
"learning_rate": 1.0893149422396403e-11,
"loss": 0.989415168762207,
"step": 1932
},
{
"epoch": 2.8233576642335767,
"grad_norm": 0.24683533608913422,
"learning_rate": 1.0864662381854632e-11,
"loss": 1.0691274404525757,
"step": 1934
},
{
"epoch": 2.8262773722627736,
"grad_norm": 0.24659498035907745,
"learning_rate": 1.0836632625504674e-11,
"loss": 1.2702159881591797,
"step": 1936
},
{
"epoch": 2.829197080291971,
"grad_norm": 0.5301987528800964,
"learning_rate": 1.0809060443761531e-11,
"loss": 1.3106719255447388,
"step": 1938
},
{
"epoch": 2.832116788321168,
"grad_norm": 0.22561290860176086,
"learning_rate": 1.0781946122299307e-11,
"loss": 1.3655246496200562,
"step": 1940
},
{
"epoch": 2.835036496350365,
"grad_norm": 0.2964118719100952,
"learning_rate": 1.0755289942048237e-11,
"loss": 1.298119306564331,
"step": 1942
},
{
"epoch": 2.8379562043795623,
"grad_norm": 0.2465992569923401,
"learning_rate": 1.0729092179191765e-11,
"loss": 1.3384490013122559,
"step": 1944
},
{
"epoch": 2.8408759124087593,
"grad_norm": 0.2166963517665863,
"learning_rate": 1.0703353105163708e-11,
"loss": 1.2705466747283936,
"step": 1946
},
{
"epoch": 2.843795620437956,
"grad_norm": 0.2356664091348648,
"learning_rate": 1.0678072986645414e-11,
"loss": 1.2964879274368286,
"step": 1948
},
{
"epoch": 2.846715328467153,
"grad_norm": 0.1898549348115921,
"learning_rate": 1.0653252085563021e-11,
"loss": 1.3481731414794922,
"step": 1950
},
{
"epoch": 2.8496350364963505,
"grad_norm": 0.22473156452178955,
"learning_rate": 1.0628890659084748e-11,
"loss": 1.2900036573410034,
"step": 1952
},
{
"epoch": 2.8525547445255475,
"grad_norm": 0.37998268008232117,
"learning_rate": 1.0604988959618193e-11,
"loss": 1.3140424489974976,
"step": 1954
},
{
"epoch": 2.8554744525547444,
"grad_norm": 0.4351639449596405,
"learning_rate": 1.0581547234807778e-11,
"loss": 1.2694008350372314,
"step": 1956
},
{
"epoch": 2.8583941605839414,
"grad_norm": 0.2931780219078064,
"learning_rate": 1.055856572753211e-11,
"loss": 1.2881304025650024,
"step": 1958
},
{
"epoch": 2.8613138686131387,
"grad_norm": 0.3236960172653198,
"learning_rate": 1.0536044675901533e-11,
"loss": 1.2829375267028809,
"step": 1960
},
{
"epoch": 2.8642335766423357,
"grad_norm": 0.2794724702835083,
"learning_rate": 1.0513984313255612e-11,
"loss": 1.2485144138336182,
"step": 1962
},
{
"epoch": 2.8671532846715326,
"grad_norm": 0.3454572856426239,
"learning_rate": 1.0492384868160763e-11,
"loss": 1.3162847757339478,
"step": 1964
},
{
"epoch": 2.87007299270073,
"grad_norm": 0.40962308645248413,
"learning_rate": 1.0471246564407825e-11,
"loss": 1.327282428741455,
"step": 1966
},
{
"epoch": 2.872992700729927,
"grad_norm": 0.6799861788749695,
"learning_rate": 1.0450569621009781e-11,
"loss": 1.3605397939682007,
"step": 1968
},
{
"epoch": 2.875912408759124,
"grad_norm": 0.5539891123771667,
"learning_rate": 1.0430354252199495e-11,
"loss": 1.3370953798294067,
"step": 1970
},
{
"epoch": 2.8788321167883213,
"grad_norm": 0.25739309191703796,
"learning_rate": 1.0410600667427462e-11,
"loss": 1.354380488395691,
"step": 1972
},
{
"epoch": 2.8817518248175182,
"grad_norm": 0.3870704770088196,
"learning_rate": 1.0391309071359665e-11,
"loss": 1.3251416683197021,
"step": 1974
},
{
"epoch": 2.884671532846715,
"grad_norm": 0.20314446091651917,
"learning_rate": 1.0372479663875433e-11,
"loss": 1.3485803604125977,
"step": 1976
},
{
"epoch": 2.8875912408759126,
"grad_norm": 0.31354570388793945,
"learning_rate": 1.0354112640065392e-11,
"loss": 1.2970892190933228,
"step": 1978
},
{
"epoch": 2.8905109489051095,
"grad_norm": 0.4762999713420868,
"learning_rate": 1.0336208190229425e-11,
"loss": 1.302712321281433,
"step": 1980
},
{
"epoch": 2.8934306569343065,
"grad_norm": 0.24994267523288727,
"learning_rate": 1.0318766499874702e-11,
"loss": 1.3298907279968262,
"step": 1982
},
{
"epoch": 2.896350364963504,
"grad_norm": 0.256294846534729,
"learning_rate": 1.0301787749713778e-11,
"loss": 1.3118003606796265,
"step": 1984
},
{
"epoch": 2.899270072992701,
"grad_norm": 0.7621220946311951,
"learning_rate": 1.0285272115662697e-11,
"loss": 1.3401591777801514,
"step": 1986
},
{
"epoch": 2.9021897810218977,
"grad_norm": 0.30471134185791016,
"learning_rate": 1.0269219768839177e-11,
"loss": 1.3362921476364136,
"step": 1988
},
{
"epoch": 2.905109489051095,
"grad_norm": 0.41349539160728455,
"learning_rate": 1.0253630875560841e-11,
"loss": 1.330914855003357,
"step": 1990
},
{
"epoch": 2.908029197080292,
"grad_norm": 0.2509981691837311,
"learning_rate": 1.0238505597343493e-11,
"loss": 1.288212537765503,
"step": 1992
},
{
"epoch": 2.910948905109489,
"grad_norm": 0.2082396149635315,
"learning_rate": 1.0223844090899445e-11,
"loss": 1.2984920740127563,
"step": 1994
},
{
"epoch": 2.9138686131386864,
"grad_norm": 0.32519200444221497,
"learning_rate": 1.0209646508135873e-11,
"loss": 1.3151469230651855,
"step": 1996
},
{
"epoch": 2.9167883211678833,
"grad_norm": 0.26818934082984924,
"learning_rate": 1.0195912996153294e-11,
"loss": 1.3731813430786133,
"step": 1998
},
{
"epoch": 2.9197080291970803,
"grad_norm": 0.43063884973526,
"learning_rate": 1.0182643697243976e-11,
"loss": 1.3158377408981323,
"step": 2000
},
{
"epoch": 2.9226277372262772,
"grad_norm": 0.280540406703949,
"learning_rate": 1.0169838748890516e-11,
"loss": 1.3163559436798096,
"step": 2002
},
{
"epoch": 2.9255474452554746,
"grad_norm": 0.22299645841121674,
"learning_rate": 1.0157498283764395e-11,
"loss": 1.2669562101364136,
"step": 2004
},
{
"epoch": 2.9284671532846716,
"grad_norm": 0.4416807293891907,
"learning_rate": 1.0145622429724598e-11,
"loss": 1.320312261581421,
"step": 2006
},
{
"epoch": 2.9313868613138685,
"grad_norm": 0.311170756816864,
"learning_rate": 1.0134211309816299e-11,
"loss": 1.3365989923477173,
"step": 2008
},
{
"epoch": 2.9343065693430654,
"grad_norm": 0.26800814270973206,
"learning_rate": 1.0123265042269589e-11,
"loss": 1.2997403144836426,
"step": 2010
},
{
"epoch": 2.937226277372263,
"grad_norm": 0.334674596786499,
"learning_rate": 1.0112783740498235e-11,
"loss": 1.3328243494033813,
"step": 2012
},
{
"epoch": 2.9401459854014598,
"grad_norm": 0.3755459189414978,
"learning_rate": 1.0102767513098516e-11,
"loss": 1.3310189247131348,
"step": 2014
},
{
"epoch": 2.9430656934306567,
"grad_norm": 0.2287389189004898,
"learning_rate": 1.0093216463848107e-11,
"loss": 1.383652687072754,
"step": 2016
},
{
"epoch": 2.945985401459854,
"grad_norm": 0.2728535830974579,
"learning_rate": 1.0084130691704985e-11,
"loss": 1.357974648475647,
"step": 2018
},
{
"epoch": 2.948905109489051,
"grad_norm": 0.31655973196029663,
"learning_rate": 1.0075510290806418e-11,
"loss": 1.3427679538726807,
"step": 2020
},
{
"epoch": 2.951824817518248,
"grad_norm": 0.24858510494232178,
"learning_rate": 1.0067355350467981e-11,
"loss": 1.3368231058120728,
"step": 2022
},
{
"epoch": 2.9547445255474454,
"grad_norm": 0.262162446975708,
"learning_rate": 1.0059665955182627e-11,
"loss": 1.3229783773422241,
"step": 2024
},
{
"epoch": 2.9576642335766423,
"grad_norm": 0.2759620249271393,
"learning_rate": 1.0052442184619831e-11,
"loss": 1.2838077545166016,
"step": 2026
},
{
"epoch": 2.9605839416058393,
"grad_norm": 0.2278604358434677,
"learning_rate": 1.0045684113624746e-11,
"loss": 1.286117434501648,
"step": 2028
},
{
"epoch": 2.9635036496350367,
"grad_norm": 0.2492004632949829,
"learning_rate": 1.0039391812217433e-11,
"loss": 1.2981648445129395,
"step": 2030
},
{
"epoch": 2.9664233576642336,
"grad_norm": 0.37556126713752747,
"learning_rate": 1.0033565345592127e-11,
"loss": 1.2995569705963135,
"step": 2032
},
{
"epoch": 2.9693430656934305,
"grad_norm": 0.2894899249076843,
"learning_rate": 1.0028204774116592e-11,
"loss": 1.3662912845611572,
"step": 2034
},
{
"epoch": 2.972262773722628,
"grad_norm": 0.310973197221756,
"learning_rate": 1.0023310153331455e-11,
"loss": 1.3318140506744385,
"step": 2036
},
{
"epoch": 2.975182481751825,
"grad_norm": 0.2315581887960434,
"learning_rate": 1.0018881533949651e-11,
"loss": 1.334710955619812,
"step": 2038
},
{
"epoch": 2.978102189781022,
"grad_norm": 0.3643578290939331,
"learning_rate": 1.0014918961855914e-11,
"loss": 1.330309510231018,
"step": 2040
},
{
"epoch": 2.981021897810219,
"grad_norm": 0.30834120512008667,
"learning_rate": 1.0011422478106256e-11,
"loss": 1.3401741981506348,
"step": 2042
},
{
"epoch": 2.983941605839416,
"grad_norm": 0.6918445229530334,
"learning_rate": 1.000839211892759e-11,
"loss": 1.2814037799835205,
"step": 2044
},
{
"epoch": 2.986861313868613,
"grad_norm": 0.26494693756103516,
"learning_rate": 1.0005827915717327e-11,
"loss": 1.3775584697723389,
"step": 2046
},
{
"epoch": 2.9897810218978105,
"grad_norm": 0.2850579619407654,
"learning_rate": 1.0003729895043056e-11,
"loss": 1.4125664234161377,
"step": 2048
},
{
"epoch": 2.9927007299270074,
"grad_norm": 0.544228732585907,
"learning_rate": 1.0002098078642278e-11,
"loss": 1.465717077255249,
"step": 2050
},
{
"epoch": 2.9956204379562044,
"grad_norm": 0.5321989059448242,
"learning_rate": 1.000093248342216e-11,
"loss": 1.5938138961791992,
"step": 2052
},
{
"epoch": 2.9985401459854013,
"grad_norm": 0.5018138885498047,
"learning_rate": 1.0000233121459382e-11,
"loss": 1.6167956590652466,
"step": 2054
},
{
"epoch": 3.0,
"step": 2055,
"total_flos": 3.6341125364349665e+18,
"train_loss": 1.4411861958004843,
"train_runtime": 21029.937,
"train_samples_per_second": 1.563,
"train_steps_per_second": 0.098
}
],
"logging_steps": 2,
"max_steps": 2055,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 9999999,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.6341125364349665e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}