tttw1004m / trainer_state.json
Kaipengm2's picture
Upload folder using huggingface_hub
ac93783 verified
Raw
History Blame Contribute Delete
75.1 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.7587973062695628,
"eval_steps": 500,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.001896993265673907,
"grad_norm": 1.204211711883545,
"learning_rate": 5e-06,
"loss": 2.802743911743164,
"step": 5
},
{
"epoch": 0.003793986531347814,
"grad_norm": 1.3055319786071777,
"learning_rate": 1.125e-05,
"loss": 2.922507476806641,
"step": 10
},
{
"epoch": 0.005690979797021721,
"grad_norm": 0.5632550716400146,
"learning_rate": 1.75e-05,
"loss": 2.7133102416992188,
"step": 15
},
{
"epoch": 0.007587973062695628,
"grad_norm": 0.2928524911403656,
"learning_rate": 2.375e-05,
"loss": 2.5979026794433593,
"step": 20
},
{
"epoch": 0.009484966328369534,
"grad_norm": 0.3215028941631317,
"learning_rate": 3e-05,
"loss": 2.5264644622802734,
"step": 25
},
{
"epoch": 0.011381959594043442,
"grad_norm": 0.30555373430252075,
"learning_rate": 3.625e-05,
"loss": 2.5048519134521485,
"step": 30
},
{
"epoch": 0.013278952859717348,
"grad_norm": 0.2635839283466339,
"learning_rate": 4.25e-05,
"loss": 2.3574724197387695,
"step": 35
},
{
"epoch": 0.015175946125391256,
"grad_norm": 0.23266227543354034,
"learning_rate": 4.875e-05,
"loss": 2.2846229553222654,
"step": 40
},
{
"epoch": 0.01707293939106516,
"grad_norm": 0.21502430737018585,
"learning_rate": 5.500000000000001e-05,
"loss": 2.2320762634277345,
"step": 45
},
{
"epoch": 0.018969932656739068,
"grad_norm": 0.23130150139331818,
"learning_rate": 6.125000000000001e-05,
"loss": 2.1551288604736327,
"step": 50
},
{
"epoch": 0.020866925922412974,
"grad_norm": 0.2014365941286087,
"learning_rate": 6.750000000000001e-05,
"loss": 2.078548622131348,
"step": 55
},
{
"epoch": 0.022763919188086883,
"grad_norm": 0.2039063721895218,
"learning_rate": 7.375e-05,
"loss": 2.03682861328125,
"step": 60
},
{
"epoch": 0.02466091245376079,
"grad_norm": 0.22200852632522583,
"learning_rate": 8e-05,
"loss": 1.9837913513183594,
"step": 65
},
{
"epoch": 0.026557905719434696,
"grad_norm": 0.19454964995384216,
"learning_rate": 8.625000000000001e-05,
"loss": 1.997808837890625,
"step": 70
},
{
"epoch": 0.0284548989851086,
"grad_norm": 0.22495104372501373,
"learning_rate": 9.250000000000001e-05,
"loss": 1.8923458099365233,
"step": 75
},
{
"epoch": 0.03035189225078251,
"grad_norm": 0.23704491555690765,
"learning_rate": 9.875000000000002e-05,
"loss": 1.876457405090332,
"step": 80
},
{
"epoch": 0.03224888551645642,
"grad_norm": 0.23261992633342743,
"learning_rate": 9.999939572147152e-05,
"loss": 1.840583038330078,
"step": 85
},
{
"epoch": 0.03414587878213032,
"grad_norm": 0.2339024692773819,
"learning_rate": 9.999694086498248e-05,
"loss": 1.855813980102539,
"step": 90
},
{
"epoch": 0.03604287204780423,
"grad_norm": 0.21744900941848755,
"learning_rate": 9.999259775576696e-05,
"loss": 1.8163593292236329,
"step": 95
},
{
"epoch": 0.037939865313478136,
"grad_norm": 0.2505708634853363,
"learning_rate": 9.998636655785278e-05,
"loss": 1.7104930877685547,
"step": 100
},
{
"epoch": 0.03983685857915204,
"grad_norm": 0.2241698056459427,
"learning_rate": 9.997824750657585e-05,
"loss": 1.7659015655517578,
"step": 105
},
{
"epoch": 0.04173385184482595,
"grad_norm": 0.22831183671951294,
"learning_rate": 9.996824090857136e-05,
"loss": 1.7662044525146485,
"step": 110
},
{
"epoch": 0.04363084511049986,
"grad_norm": 0.24719521403312683,
"learning_rate": 9.995634714176216e-05,
"loss": 1.7465341567993165,
"step": 115
},
{
"epoch": 0.04552783837617377,
"grad_norm": 0.21913813054561615,
"learning_rate": 9.994256665534445e-05,
"loss": 1.6910831451416015,
"step": 120
},
{
"epoch": 0.04742483164184767,
"grad_norm": 0.24229924380779266,
"learning_rate": 9.992689996977091e-05,
"loss": 1.6647504806518554,
"step": 125
},
{
"epoch": 0.04932182490752158,
"grad_norm": 0.25045862793922424,
"learning_rate": 9.9909347676731e-05,
"loss": 1.7728565216064454,
"step": 130
},
{
"epoch": 0.051218818173195485,
"grad_norm": 0.23526810109615326,
"learning_rate": 9.988991043912857e-05,
"loss": 1.703898048400879,
"step": 135
},
{
"epoch": 0.05311581143886939,
"grad_norm": 0.25613462924957275,
"learning_rate": 9.986858899105684e-05,
"loss": 1.6876985549926757,
"step": 140
},
{
"epoch": 0.0550128047045433,
"grad_norm": 0.26118171215057373,
"learning_rate": 9.984538413777076e-05,
"loss": 1.6936487197875976,
"step": 145
},
{
"epoch": 0.0569097979702172,
"grad_norm": 0.21433641016483307,
"learning_rate": 9.98202967556565e-05,
"loss": 1.6974334716796875,
"step": 150
},
{
"epoch": 0.05880679123589111,
"grad_norm": 0.26981523633003235,
"learning_rate": 9.979332779219836e-05,
"loss": 1.6554887771606446,
"step": 155
},
{
"epoch": 0.06070378450156502,
"grad_norm": 0.23291254043579102,
"learning_rate": 9.976447826594305e-05,
"loss": 1.6636981964111328,
"step": 160
},
{
"epoch": 0.06260077776723892,
"grad_norm": 0.24142742156982422,
"learning_rate": 9.973374926646116e-05,
"loss": 1.6125133514404297,
"step": 165
},
{
"epoch": 0.06449777103291283,
"grad_norm": 0.2866875231266022,
"learning_rate": 9.970114195430604e-05,
"loss": 1.5909235954284668,
"step": 170
},
{
"epoch": 0.06639476429858673,
"grad_norm": 0.2811475098133087,
"learning_rate": 9.966665756096995e-05,
"loss": 1.5830004692077637,
"step": 175
},
{
"epoch": 0.06829175756426065,
"grad_norm": 0.24986404180526733,
"learning_rate": 9.96302973888376e-05,
"loss": 1.5703643798828124,
"step": 180
},
{
"epoch": 0.07018875082993456,
"grad_norm": 0.21419142186641693,
"learning_rate": 9.959206281113684e-05,
"loss": 1.607168197631836,
"step": 185
},
{
"epoch": 0.07208574409560846,
"grad_norm": 0.2700653374195099,
"learning_rate": 9.955195527188697e-05,
"loss": 1.591099166870117,
"step": 190
},
{
"epoch": 0.07398273736128237,
"grad_norm": 0.24841217696666718,
"learning_rate": 9.950997628584404e-05,
"loss": 1.555056381225586,
"step": 195
},
{
"epoch": 0.07587973062695627,
"grad_norm": 0.20358560979366302,
"learning_rate": 9.946612743844379e-05,
"loss": 1.5962520599365235,
"step": 200
},
{
"epoch": 0.07777672389263018,
"grad_norm": 0.26881343126296997,
"learning_rate": 9.942041038574162e-05,
"loss": 1.6076745986938477,
"step": 205
},
{
"epoch": 0.07967371715830408,
"grad_norm": 0.2529791593551636,
"learning_rate": 9.937282685435019e-05,
"loss": 1.6203548431396484,
"step": 210
},
{
"epoch": 0.081570710423978,
"grad_norm": 0.24552154541015625,
"learning_rate": 9.932337864137407e-05,
"loss": 1.5111330986022948,
"step": 215
},
{
"epoch": 0.0834677036896519,
"grad_norm": 0.2761000096797943,
"learning_rate": 9.9272067614342e-05,
"loss": 1.586027240753174,
"step": 220
},
{
"epoch": 0.08536469695532581,
"grad_norm": 0.24447834491729736,
"learning_rate": 9.921889571113628e-05,
"loss": 1.5407236099243165,
"step": 225
},
{
"epoch": 0.08726169022099972,
"grad_norm": 0.31283313035964966,
"learning_rate": 9.91638649399196e-05,
"loss": 1.6450325012207032,
"step": 230
},
{
"epoch": 0.08915868348667362,
"grad_norm": 0.30715709924697876,
"learning_rate": 9.910697737905919e-05,
"loss": 1.5557141304016113,
"step": 235
},
{
"epoch": 0.09105567675234753,
"grad_norm": 0.22031593322753906,
"learning_rate": 9.904823517704836e-05,
"loss": 1.5843168258666993,
"step": 240
},
{
"epoch": 0.09295267001802143,
"grad_norm": 0.25055739283561707,
"learning_rate": 9.898764055242526e-05,
"loss": 1.5050912857055665,
"step": 245
},
{
"epoch": 0.09484966328369535,
"grad_norm": 0.2518811821937561,
"learning_rate": 9.892519579368926e-05,
"loss": 1.6115983963012694,
"step": 250
},
{
"epoch": 0.09674665654936924,
"grad_norm": 0.23363788425922394,
"learning_rate": 9.886090325921434e-05,
"loss": 1.5730743408203125,
"step": 255
},
{
"epoch": 0.09864364981504316,
"grad_norm": 0.2886348068714142,
"learning_rate": 9.879476537716015e-05,
"loss": 1.5849692344665527,
"step": 260
},
{
"epoch": 0.10054064308071706,
"grad_norm": 0.24128347635269165,
"learning_rate": 9.872678464538021e-05,
"loss": 1.5591470718383789,
"step": 265
},
{
"epoch": 0.10243763634639097,
"grad_norm": 0.2330770194530487,
"learning_rate": 9.865696363132769e-05,
"loss": 1.5730993270874023,
"step": 270
},
{
"epoch": 0.10433462961206488,
"grad_norm": 0.2483072429895401,
"learning_rate": 9.858530497195828e-05,
"loss": 1.5576128959655762,
"step": 275
},
{
"epoch": 0.10623162287773878,
"grad_norm": 0.3136220872402191,
"learning_rate": 9.851181137363079e-05,
"loss": 1.5509288787841797,
"step": 280
},
{
"epoch": 0.1081286161434127,
"grad_norm": 0.27949973940849304,
"learning_rate": 9.843648561200475e-05,
"loss": 1.5363675117492677,
"step": 285
},
{
"epoch": 0.1100256094090866,
"grad_norm": 0.28136932849884033,
"learning_rate": 9.835933053193574e-05,
"loss": 1.563088321685791,
"step": 290
},
{
"epoch": 0.11192260267476051,
"grad_norm": 0.2632514238357544,
"learning_rate": 9.828034904736782e-05,
"loss": 1.5300712585449219,
"step": 295
},
{
"epoch": 0.1138195959404344,
"grad_norm": 0.2528294324874878,
"learning_rate": 9.819954414122359e-05,
"loss": 1.4897494316101074,
"step": 300
},
{
"epoch": 0.11571658920610832,
"grad_norm": 0.24739930033683777,
"learning_rate": 9.811691886529144e-05,
"loss": 1.5607781410217285,
"step": 305
},
{
"epoch": 0.11761358247178222,
"grad_norm": 0.2602822482585907,
"learning_rate": 9.803247634011034e-05,
"loss": 1.5098703384399415,
"step": 310
},
{
"epoch": 0.11951057573745613,
"grad_norm": 0.28818434476852417,
"learning_rate": 9.794621975485201e-05,
"loss": 1.51871337890625,
"step": 315
},
{
"epoch": 0.12140756900313004,
"grad_norm": 0.2329605221748352,
"learning_rate": 9.785815236720037e-05,
"loss": 1.5149699211120606,
"step": 320
},
{
"epoch": 0.12330456226880394,
"grad_norm": 0.2775144577026367,
"learning_rate": 9.776827750322865e-05,
"loss": 1.5283567428588867,
"step": 325
},
{
"epoch": 0.12520155553447784,
"grad_norm": 0.22660040855407715,
"learning_rate": 9.767659855727364e-05,
"loss": 1.5318140029907226,
"step": 330
},
{
"epoch": 0.12709854880015176,
"grad_norm": 0.28303974866867065,
"learning_rate": 9.758311899180758e-05,
"loss": 1.5514656066894532,
"step": 335
},
{
"epoch": 0.12899554206582567,
"grad_norm": 0.2359406054019928,
"learning_rate": 9.748784233730736e-05,
"loss": 1.515054988861084,
"step": 340
},
{
"epoch": 0.13089253533149958,
"grad_norm": 0.23715272545814514,
"learning_rate": 9.73907721921212e-05,
"loss": 1.50885009765625,
"step": 345
},
{
"epoch": 0.13278952859717347,
"grad_norm": 0.2521718740463257,
"learning_rate": 9.729191222233272e-05,
"loss": 1.5373132705688477,
"step": 350
},
{
"epoch": 0.13468652186284738,
"grad_norm": 0.25664857029914856,
"learning_rate": 9.719126616162251e-05,
"loss": 1.506770420074463,
"step": 355
},
{
"epoch": 0.1365835151285213,
"grad_norm": 0.24742132425308228,
"learning_rate": 9.70888378111271e-05,
"loss": 1.5162310600280762,
"step": 360
},
{
"epoch": 0.1384805083941952,
"grad_norm": 0.3116005063056946,
"learning_rate": 9.698463103929542e-05,
"loss": 1.515502643585205,
"step": 365
},
{
"epoch": 0.14037750165986912,
"grad_norm": 0.2673533260822296,
"learning_rate": 9.687864978174268e-05,
"loss": 1.5125722885131836,
"step": 370
},
{
"epoch": 0.142274494925543,
"grad_norm": 0.2768329977989197,
"learning_rate": 9.677089804110174e-05,
"loss": 1.515056324005127,
"step": 375
},
{
"epoch": 0.14417148819121692,
"grad_norm": 0.2856305241584778,
"learning_rate": 9.666137988687193e-05,
"loss": 1.533064365386963,
"step": 380
},
{
"epoch": 0.14606848145689083,
"grad_norm": 0.2695675790309906,
"learning_rate": 9.655009945526541e-05,
"loss": 1.5210994720458983,
"step": 385
},
{
"epoch": 0.14796547472256474,
"grad_norm": 0.2579309046268463,
"learning_rate": 9.643706094905086e-05,
"loss": 1.5141459465026856,
"step": 390
},
{
"epoch": 0.14986246798823863,
"grad_norm": 0.2593337595462799,
"learning_rate": 9.632226863739484e-05,
"loss": 1.509230613708496,
"step": 395
},
{
"epoch": 0.15175946125391254,
"grad_norm": 0.22641733288764954,
"learning_rate": 9.620572685570052e-05,
"loss": 1.4663333892822266,
"step": 400
},
{
"epoch": 0.15365645451958646,
"grad_norm": 0.2912636399269104,
"learning_rate": 9.608744000544392e-05,
"loss": 1.4927945137023926,
"step": 405
},
{
"epoch": 0.15555344778526037,
"grad_norm": 0.2990904152393341,
"learning_rate": 9.596741255400773e-05,
"loss": 1.4595409393310548,
"step": 410
},
{
"epoch": 0.15745044105093428,
"grad_norm": 0.2548057436943054,
"learning_rate": 9.584564903451252e-05,
"loss": 1.5293696403503418,
"step": 415
},
{
"epoch": 0.15934743431660817,
"grad_norm": 0.2348567694425583,
"learning_rate": 9.572215404564564e-05,
"loss": 1.5123428344726562,
"step": 420
},
{
"epoch": 0.16124442758228208,
"grad_norm": 0.29777541756629944,
"learning_rate": 9.559693225148743e-05,
"loss": 1.4895515441894531,
"step": 425
},
{
"epoch": 0.163141420847956,
"grad_norm": 0.25626078248023987,
"learning_rate": 9.54699883813351e-05,
"loss": 1.5255406379699707,
"step": 430
},
{
"epoch": 0.1650384141136299,
"grad_norm": 0.27839428186416626,
"learning_rate": 9.534132722952418e-05,
"loss": 1.4806964874267579,
"step": 435
},
{
"epoch": 0.1669354073793038,
"grad_norm": 0.26998913288116455,
"learning_rate": 9.521095365524739e-05,
"loss": 1.487065315246582,
"step": 440
},
{
"epoch": 0.1688324006449777,
"grad_norm": 0.25939276814460754,
"learning_rate": 9.507887258237112e-05,
"loss": 1.4892295837402343,
"step": 445
},
{
"epoch": 0.17072939391065162,
"grad_norm": 0.26628801226615906,
"learning_rate": 9.494508899924948e-05,
"loss": 1.4731863975524901,
"step": 450
},
{
"epoch": 0.17262638717632553,
"grad_norm": 0.2473168522119522,
"learning_rate": 9.480960795853591e-05,
"loss": 1.454709815979004,
"step": 455
},
{
"epoch": 0.17452338044199944,
"grad_norm": 0.2799444794654846,
"learning_rate": 9.467243457699238e-05,
"loss": 1.4769775390625,
"step": 460
},
{
"epoch": 0.17642037370767333,
"grad_norm": 0.2699210047721863,
"learning_rate": 9.453357403529609e-05,
"loss": 1.459995937347412,
"step": 465
},
{
"epoch": 0.17831736697334724,
"grad_norm": 0.2729581594467163,
"learning_rate": 9.439303157784383e-05,
"loss": 1.4660751342773437,
"step": 470
},
{
"epoch": 0.18021436023902115,
"grad_norm": 0.2628643214702606,
"learning_rate": 9.425081251255392e-05,
"loss": 1.4800500869750977,
"step": 475
},
{
"epoch": 0.18211135350469507,
"grad_norm": 0.24878107011318207,
"learning_rate": 9.410692221066573e-05,
"loss": 1.466275691986084,
"step": 480
},
{
"epoch": 0.18400834677036895,
"grad_norm": 0.26707327365875244,
"learning_rate": 9.396136610653687e-05,
"loss": 1.4521776199340821,
"step": 485
},
{
"epoch": 0.18590534003604287,
"grad_norm": 0.28566935658454895,
"learning_rate": 9.381414969743788e-05,
"loss": 1.4933480262756347,
"step": 490
},
{
"epoch": 0.18780233330171678,
"grad_norm": 0.2836907207965851,
"learning_rate": 9.366527854334463e-05,
"loss": 1.4608956336975099,
"step": 495
},
{
"epoch": 0.1896993265673907,
"grad_norm": 0.35665667057037354,
"learning_rate": 9.351475826672842e-05,
"loss": 1.4607025146484376,
"step": 500
},
{
"epoch": 0.1915963198330646,
"grad_norm": 0.2649923264980316,
"learning_rate": 9.336259455234347e-05,
"loss": 1.4632261276245118,
"step": 505
},
{
"epoch": 0.1934933130987385,
"grad_norm": 0.25335875153541565,
"learning_rate": 9.320879314701233e-05,
"loss": 1.4849108695983886,
"step": 510
},
{
"epoch": 0.1953903063644124,
"grad_norm": 0.250990629196167,
"learning_rate": 9.305335985940895e-05,
"loss": 1.4326010704040528,
"step": 515
},
{
"epoch": 0.19728729963008632,
"grad_norm": 0.27390721440315247,
"learning_rate": 9.289630055983897e-05,
"loss": 1.502902889251709,
"step": 520
},
{
"epoch": 0.19918429289576023,
"grad_norm": 0.27056121826171875,
"learning_rate": 9.273762118001837e-05,
"loss": 1.483546829223633,
"step": 525
},
{
"epoch": 0.20108128616143411,
"grad_norm": 0.26807019114494324,
"learning_rate": 9.257732771284924e-05,
"loss": 1.445467758178711,
"step": 530
},
{
"epoch": 0.20297827942710803,
"grad_norm": 0.3302270770072937,
"learning_rate": 9.241542621219348e-05,
"loss": 1.4597110748291016,
"step": 535
},
{
"epoch": 0.20487527269278194,
"grad_norm": 0.27566173672676086,
"learning_rate": 9.225192279264423e-05,
"loss": 1.4078072547912597,
"step": 540
},
{
"epoch": 0.20677226595845585,
"grad_norm": 0.2839542329311371,
"learning_rate": 9.208682362929484e-05,
"loss": 1.452101707458496,
"step": 545
},
{
"epoch": 0.20866925922412977,
"grad_norm": 0.2720314562320709,
"learning_rate": 9.19201349575057e-05,
"loss": 1.4331625938415526,
"step": 550
},
{
"epoch": 0.21056625248980365,
"grad_norm": 0.26887810230255127,
"learning_rate": 9.175186307266876e-05,
"loss": 1.4134525299072265,
"step": 555
},
{
"epoch": 0.21246324575547756,
"grad_norm": 0.25607389211654663,
"learning_rate": 9.158201432996978e-05,
"loss": 1.4619136810302735,
"step": 560
},
{
"epoch": 0.21436023902115148,
"grad_norm": 0.31376007199287415,
"learning_rate": 9.141059514414824e-05,
"loss": 1.4151582717895508,
"step": 565
},
{
"epoch": 0.2162572322868254,
"grad_norm": 0.2648751735687256,
"learning_rate": 9.123761198925514e-05,
"loss": 1.536808967590332,
"step": 570
},
{
"epoch": 0.21815422555249928,
"grad_norm": 0.24469561874866486,
"learning_rate": 9.106307139840849e-05,
"loss": 1.494936752319336,
"step": 575
},
{
"epoch": 0.2200512188181732,
"grad_norm": 0.24914619326591492,
"learning_rate": 9.088697996354649e-05,
"loss": 1.4340242385864257,
"step": 580
},
{
"epoch": 0.2219482120838471,
"grad_norm": 0.24026300013065338,
"learning_rate": 9.070934433517873e-05,
"loss": 1.4562675476074218,
"step": 585
},
{
"epoch": 0.22384520534952101,
"grad_norm": 0.2660679519176483,
"learning_rate": 9.053017122213481e-05,
"loss": 1.4040220260620118,
"step": 590
},
{
"epoch": 0.22574219861519493,
"grad_norm": 0.23900838196277618,
"learning_rate": 9.034946739131119e-05,
"loss": 1.4403876304626464,
"step": 595
},
{
"epoch": 0.2276391918808688,
"grad_norm": 0.24892283976078033,
"learning_rate": 9.016723966741538e-05,
"loss": 1.4478761672973632,
"step": 600
},
{
"epoch": 0.22953618514654273,
"grad_norm": 0.2921767234802246,
"learning_rate": 8.998349493270846e-05,
"loss": 1.4169246673583984,
"step": 605
},
{
"epoch": 0.23143317841221664,
"grad_norm": 0.24210797250270844,
"learning_rate": 8.979824012674489e-05,
"loss": 1.4091812133789063,
"step": 610
},
{
"epoch": 0.23333017167789055,
"grad_norm": 0.2604314684867859,
"learning_rate": 8.96114822461106e-05,
"loss": 1.4544056892395019,
"step": 615
},
{
"epoch": 0.23522716494356444,
"grad_norm": 0.2753211259841919,
"learning_rate": 8.942322834415869e-05,
"loss": 1.4028273582458497,
"step": 620
},
{
"epoch": 0.23712415820923835,
"grad_norm": 0.25446656346321106,
"learning_rate": 8.9233485530743e-05,
"loss": 1.4523637771606446,
"step": 625
},
{
"epoch": 0.23902115147491226,
"grad_norm": 0.2570347189903259,
"learning_rate": 8.90422609719497e-05,
"loss": 1.4611879348754884,
"step": 630
},
{
"epoch": 0.24091814474058618,
"grad_norm": 0.28898608684539795,
"learning_rate": 8.884956188982651e-05,
"loss": 1.3687424659729004,
"step": 635
},
{
"epoch": 0.2428151380062601,
"grad_norm": 0.25607869029045105,
"learning_rate": 8.865539556211008e-05,
"loss": 1.4238749504089356,
"step": 640
},
{
"epoch": 0.24471213127193397,
"grad_norm": 0.24087631702423096,
"learning_rate": 8.845976932195103e-05,
"loss": 1.4357139587402343,
"step": 645
},
{
"epoch": 0.2466091245376079,
"grad_norm": 0.2392406463623047,
"learning_rate": 8.826269055763701e-05,
"loss": 1.4272048950195313,
"step": 650
},
{
"epoch": 0.2485061178032818,
"grad_norm": 0.24851970374584198,
"learning_rate": 8.806416671231366e-05,
"loss": 1.4464078903198243,
"step": 655
},
{
"epoch": 0.2504031110689557,
"grad_norm": 0.25366175174713135,
"learning_rate": 8.786420528370361e-05,
"loss": 1.4115880012512207,
"step": 660
},
{
"epoch": 0.2523001043346296,
"grad_norm": 0.2184738665819168,
"learning_rate": 8.766281382382311e-05,
"loss": 1.4369379997253418,
"step": 665
},
{
"epoch": 0.2541970976003035,
"grad_norm": 0.31689387559890747,
"learning_rate": 8.745999993869703e-05,
"loss": 1.3705198287963867,
"step": 670
},
{
"epoch": 0.25609409086597745,
"grad_norm": 0.28559598326683044,
"learning_rate": 8.725577128807143e-05,
"loss": 1.4334306716918945,
"step": 675
},
{
"epoch": 0.25799108413165134,
"grad_norm": 0.24641846120357513,
"learning_rate": 8.705013558512435e-05,
"loss": 1.4052096366882325,
"step": 680
},
{
"epoch": 0.2598880773973252,
"grad_norm": 0.2755413055419922,
"learning_rate": 8.684310059617448e-05,
"loss": 1.4064033508300782,
"step": 685
},
{
"epoch": 0.26178507066299916,
"grad_norm": 0.2912381887435913,
"learning_rate": 8.663467414038787e-05,
"loss": 1.4062003135681151,
"step": 690
},
{
"epoch": 0.26368206392867305,
"grad_norm": 0.2753216624259949,
"learning_rate": 8.642486408948261e-05,
"loss": 1.3890793800354004,
"step": 695
},
{
"epoch": 0.26557905719434693,
"grad_norm": 0.2836456298828125,
"learning_rate": 8.62136783674315e-05,
"loss": 1.4482671737670898,
"step": 700
},
{
"epoch": 0.2674760504600209,
"grad_norm": 0.28797534108161926,
"learning_rate": 8.600112495016288e-05,
"loss": 1.4231111526489257,
"step": 705
},
{
"epoch": 0.26937304372569476,
"grad_norm": 0.21396268904209137,
"learning_rate": 8.578721186525925e-05,
"loss": 1.4082660675048828,
"step": 710
},
{
"epoch": 0.2712700369913687,
"grad_norm": 0.29503434896469116,
"learning_rate": 8.557194719165424e-05,
"loss": 1.4204629898071288,
"step": 715
},
{
"epoch": 0.2731670302570426,
"grad_norm": 0.2756027281284332,
"learning_rate": 8.535533905932738e-05,
"loss": 1.4463971138000489,
"step": 720
},
{
"epoch": 0.27506402352271647,
"grad_norm": 0.24828635156154633,
"learning_rate": 8.513739564899715e-05,
"loss": 1.4123985290527343,
"step": 725
},
{
"epoch": 0.2769610167883904,
"grad_norm": 0.30589962005615234,
"learning_rate": 8.49181251918119e-05,
"loss": 1.4363461494445802,
"step": 730
},
{
"epoch": 0.2788580100540643,
"grad_norm": 0.22183258831501007,
"learning_rate": 8.46975359690391e-05,
"loss": 1.4241199493408203,
"step": 735
},
{
"epoch": 0.28075500331973824,
"grad_norm": 0.2492808699607849,
"learning_rate": 8.447563631175245e-05,
"loss": 1.422759437561035,
"step": 740
},
{
"epoch": 0.2826519965854121,
"grad_norm": 0.27179956436157227,
"learning_rate": 8.425243460051739e-05,
"loss": 1.4010013580322265,
"step": 745
},
{
"epoch": 0.284548989851086,
"grad_norm": 0.28707119822502136,
"learning_rate": 8.402793926507448e-05,
"loss": 1.4091285705566405,
"step": 750
},
{
"epoch": 0.28644598311675995,
"grad_norm": 0.2424633949995041,
"learning_rate": 8.380215878402101e-05,
"loss": 1.4144980430603027,
"step": 755
},
{
"epoch": 0.28834297638243384,
"grad_norm": 0.29242298007011414,
"learning_rate": 8.35751016844909e-05,
"loss": 1.4693922042846679,
"step": 760
},
{
"epoch": 0.2902399696481078,
"grad_norm": 0.24107855558395386,
"learning_rate": 8.334677654183254e-05,
"loss": 1.3732561111450194,
"step": 765
},
{
"epoch": 0.29213696291378166,
"grad_norm": 0.25916820764541626,
"learning_rate": 8.3117191979285e-05,
"loss": 1.439093780517578,
"step": 770
},
{
"epoch": 0.29403395617945555,
"grad_norm": 0.28402164578437805,
"learning_rate": 8.288635666765231e-05,
"loss": 1.3702613830566406,
"step": 775
},
{
"epoch": 0.2959309494451295,
"grad_norm": 0.24597717821598053,
"learning_rate": 8.265427932497598e-05,
"loss": 1.4019732475280762,
"step": 780
},
{
"epoch": 0.2978279427108034,
"grad_norm": 0.2891501784324646,
"learning_rate": 8.242096871620578e-05,
"loss": 1.3701047897338867,
"step": 785
},
{
"epoch": 0.29972493597647726,
"grad_norm": 0.28862521052360535,
"learning_rate": 8.218643365286866e-05,
"loss": 1.438255214691162,
"step": 790
},
{
"epoch": 0.3016219292421512,
"grad_norm": 0.22568482160568237,
"learning_rate": 8.195068299273608e-05,
"loss": 1.3949029922485352,
"step": 795
},
{
"epoch": 0.3035189225078251,
"grad_norm": 0.23400996625423431,
"learning_rate": 8.171372563948927e-05,
"loss": 1.4001420974731444,
"step": 800
},
{
"epoch": 0.305415915773499,
"grad_norm": 0.2480504959821701,
"learning_rate": 8.147557054238317e-05,
"loss": 1.3827430725097656,
"step": 805
},
{
"epoch": 0.3073129090391729,
"grad_norm": 0.24780896306037903,
"learning_rate": 8.123622669590831e-05,
"loss": 1.3203132629394532,
"step": 810
},
{
"epoch": 0.3092099023048468,
"grad_norm": 0.23441602289676666,
"learning_rate": 8.099570313945113e-05,
"loss": 1.4034107208251954,
"step": 815
},
{
"epoch": 0.31110689557052074,
"grad_norm": 0.24393552541732788,
"learning_rate": 8.075400895695266e-05,
"loss": 1.4224702835083007,
"step": 820
},
{
"epoch": 0.3130038888361946,
"grad_norm": 0.24471324682235718,
"learning_rate": 8.051115327656538e-05,
"loss": 1.3615886688232421,
"step": 825
},
{
"epoch": 0.31490088210186856,
"grad_norm": 0.2637152671813965,
"learning_rate": 8.026714527030843e-05,
"loss": 1.4766820907592773,
"step": 830
},
{
"epoch": 0.31679787536754245,
"grad_norm": 0.21350067853927612,
"learning_rate": 8.002199415372134e-05,
"loss": 1.3876203536987304,
"step": 835
},
{
"epoch": 0.31869486863321633,
"grad_norm": 0.27970579266548157,
"learning_rate": 7.977570918551584e-05,
"loss": 1.3938987731933594,
"step": 840
},
{
"epoch": 0.3205918618988903,
"grad_norm": 0.23193910717964172,
"learning_rate": 7.95282996672263e-05,
"loss": 1.409254264831543,
"step": 845
},
{
"epoch": 0.32248885516456416,
"grad_norm": 0.23503375053405762,
"learning_rate": 7.927977494285835e-05,
"loss": 1.3622031211853027,
"step": 850
},
{
"epoch": 0.3243858484302381,
"grad_norm": 0.26904213428497314,
"learning_rate": 7.903014439853605e-05,
"loss": 1.3633729934692382,
"step": 855
},
{
"epoch": 0.326282841695912,
"grad_norm": 0.23989248275756836,
"learning_rate": 7.87794174621473e-05,
"loss": 1.3692960739135742,
"step": 860
},
{
"epoch": 0.32817983496158587,
"grad_norm": 0.2922050356864929,
"learning_rate": 7.85276036029879e-05,
"loss": 1.4186046600341797,
"step": 865
},
{
"epoch": 0.3300768282272598,
"grad_norm": 0.2780815660953522,
"learning_rate": 7.827471233140385e-05,
"loss": 1.3955347061157226,
"step": 870
},
{
"epoch": 0.3319738214929337,
"grad_norm": 0.22726669907569885,
"learning_rate": 7.802075319843217e-05,
"loss": 1.346540641784668,
"step": 875
},
{
"epoch": 0.3338708147586076,
"grad_norm": 0.22620946168899536,
"learning_rate": 7.776573579544017e-05,
"loss": 1.3656360626220703,
"step": 880
},
{
"epoch": 0.3357678080242815,
"grad_norm": 0.27938517928123474,
"learning_rate": 7.750966975376328e-05,
"loss": 1.3472801208496095,
"step": 885
},
{
"epoch": 0.3376648012899554,
"grad_norm": 0.26527056097984314,
"learning_rate": 7.72525647443412e-05,
"loss": 1.3760688781738282,
"step": 890
},
{
"epoch": 0.33956179455562935,
"grad_norm": 0.27444592118263245,
"learning_rate": 7.699443047735274e-05,
"loss": 1.4339282035827636,
"step": 895
},
{
"epoch": 0.34145878782130323,
"grad_norm": 0.26562467217445374,
"learning_rate": 7.673527670184902e-05,
"loss": 1.3343040466308593,
"step": 900
},
{
"epoch": 0.3433557810869771,
"grad_norm": 0.3343700170516968,
"learning_rate": 7.647511320538536e-05,
"loss": 1.349220085144043,
"step": 905
},
{
"epoch": 0.34525277435265106,
"grad_norm": 0.2355739027261734,
"learning_rate": 7.621394981365153e-05,
"loss": 1.387210750579834,
"step": 910
},
{
"epoch": 0.34714976761832494,
"grad_norm": 0.26889568567276,
"learning_rate": 7.595179639010075e-05,
"loss": 1.4053765296936036,
"step": 915
},
{
"epoch": 0.3490467608839989,
"grad_norm": 0.2567550539970398,
"learning_rate": 7.56886628355771e-05,
"loss": 1.362953281402588,
"step": 920
},
{
"epoch": 0.35094375414967277,
"grad_norm": 0.2534516453742981,
"learning_rate": 7.542455908794163e-05,
"loss": 1.3884872436523437,
"step": 925
},
{
"epoch": 0.35284074741534666,
"grad_norm": 0.22958898544311523,
"learning_rate": 7.515949512169708e-05,
"loss": 1.3981565475463866,
"step": 930
},
{
"epoch": 0.3547377406810206,
"grad_norm": 0.25677797198295593,
"learning_rate": 7.489348094761104e-05,
"loss": 1.4092668533325194,
"step": 935
},
{
"epoch": 0.3566347339466945,
"grad_norm": 0.2739006280899048,
"learning_rate": 7.4626526612338e-05,
"loss": 1.3539085388183594,
"step": 940
},
{
"epoch": 0.3585317272123684,
"grad_norm": 0.2807665169239044,
"learning_rate": 7.435864219803983e-05,
"loss": 1.3749590873718263,
"step": 945
},
{
"epoch": 0.3604287204780423,
"grad_norm": 0.24743717908859253,
"learning_rate": 7.408983782200505e-05,
"loss": 1.3704416275024414,
"step": 950
},
{
"epoch": 0.3623257137437162,
"grad_norm": 0.2931808531284332,
"learning_rate": 7.382012363626672e-05,
"loss": 1.372137451171875,
"step": 955
},
{
"epoch": 0.36422270700939013,
"grad_norm": 0.2241211086511612,
"learning_rate": 7.354950982721897e-05,
"loss": 1.361600685119629,
"step": 960
},
{
"epoch": 0.366119700275064,
"grad_norm": 0.23293639719486237,
"learning_rate": 7.327800661523238e-05,
"loss": 1.3370338439941407,
"step": 965
},
{
"epoch": 0.3680166935407379,
"grad_norm": 0.272787868976593,
"learning_rate": 7.300562425426793e-05,
"loss": 1.326862907409668,
"step": 970
},
{
"epoch": 0.36991368680641185,
"grad_norm": 0.2674017548561096,
"learning_rate": 7.273237303148975e-05,
"loss": 1.3624146461486817,
"step": 975
},
{
"epoch": 0.37181068007208573,
"grad_norm": 0.25618863105773926,
"learning_rate": 7.245826326687652e-05,
"loss": 1.3387451171875,
"step": 980
},
{
"epoch": 0.37370767333775967,
"grad_norm": 0.2501315772533417,
"learning_rate": 7.21833053128319e-05,
"loss": 1.3274173736572266,
"step": 985
},
{
"epoch": 0.37560466660343356,
"grad_norm": 0.2657756507396698,
"learning_rate": 7.190750955379331e-05,
"loss": 1.3742895126342773,
"step": 990
},
{
"epoch": 0.37750165986910744,
"grad_norm": 0.35259178280830383,
"learning_rate": 7.163088640583993e-05,
"loss": 1.3802461624145508,
"step": 995
},
{
"epoch": 0.3793986531347814,
"grad_norm": 0.25012266635894775,
"learning_rate": 7.135344631629923e-05,
"loss": 1.3813657760620117,
"step": 1000
},
{
"epoch": 0.38129564640045527,
"grad_norm": 0.27602267265319824,
"learning_rate": 7.10751997633524e-05,
"loss": 1.344485378265381,
"step": 1005
},
{
"epoch": 0.3831926396661292,
"grad_norm": 0.2319122701883316,
"learning_rate": 7.079615725563864e-05,
"loss": 1.3822796821594239,
"step": 1010
},
{
"epoch": 0.3850896329318031,
"grad_norm": 0.2415914237499237,
"learning_rate": 7.051632933185824e-05,
"loss": 1.3857491493225098,
"step": 1015
},
{
"epoch": 0.386986626197477,
"grad_norm": 0.23108910024166107,
"learning_rate": 7.02357265603746e-05,
"loss": 1.3276800155639648,
"step": 1020
},
{
"epoch": 0.3888836194631509,
"grad_norm": 0.2377890944480896,
"learning_rate": 6.995435953881509e-05,
"loss": 1.3621143341064452,
"step": 1025
},
{
"epoch": 0.3907806127288248,
"grad_norm": 0.2662498652935028,
"learning_rate": 6.967223889367079e-05,
"loss": 1.3481254577636719,
"step": 1030
},
{
"epoch": 0.39267760599449875,
"grad_norm": 0.2958657145500183,
"learning_rate": 6.938937527989511e-05,
"loss": 1.3370389938354492,
"step": 1035
},
{
"epoch": 0.39457459926017263,
"grad_norm": 0.28385475277900696,
"learning_rate": 6.91057793805015e-05,
"loss": 1.3931269645690918,
"step": 1040
},
{
"epoch": 0.3964715925258465,
"grad_norm": 0.22935713827610016,
"learning_rate": 6.882146190615984e-05,
"loss": 1.329813575744629,
"step": 1045
},
{
"epoch": 0.39836858579152046,
"grad_norm": 0.26869210600852966,
"learning_rate": 6.853643359479207e-05,
"loss": 1.3835307121276856,
"step": 1050
},
{
"epoch": 0.40026557905719434,
"grad_norm": 0.2622896432876587,
"learning_rate": 6.825070521116649e-05,
"loss": 1.3417768478393555,
"step": 1055
},
{
"epoch": 0.40216257232286823,
"grad_norm": 0.2867312431335449,
"learning_rate": 6.796428754649134e-05,
"loss": 1.416060447692871,
"step": 1060
},
{
"epoch": 0.40405956558854217,
"grad_norm": 0.24947425723075867,
"learning_rate": 6.767719141800717e-05,
"loss": 1.302861499786377,
"step": 1065
},
{
"epoch": 0.40595655885421605,
"grad_norm": 0.2498169094324112,
"learning_rate": 6.738942766857834e-05,
"loss": 1.360936450958252,
"step": 1070
},
{
"epoch": 0.40785355211989,
"grad_norm": 0.26447436213493347,
"learning_rate": 6.710100716628344e-05,
"loss": 1.3067538261413574,
"step": 1075
},
{
"epoch": 0.4097505453855639,
"grad_norm": 0.2777959704399109,
"learning_rate": 6.681194080400495e-05,
"loss": 1.385844612121582,
"step": 1080
},
{
"epoch": 0.41164753865123777,
"grad_norm": 0.2564963698387146,
"learning_rate": 6.652223949901774e-05,
"loss": 1.3263046264648437,
"step": 1085
},
{
"epoch": 0.4135445319169117,
"grad_norm": 0.24474751949310303,
"learning_rate": 6.623191419257683e-05,
"loss": 1.334895133972168,
"step": 1090
},
{
"epoch": 0.4154415251825856,
"grad_norm": 0.23888425529003143,
"learning_rate": 6.594097584950406e-05,
"loss": 1.384188175201416,
"step": 1095
},
{
"epoch": 0.41733851844825953,
"grad_norm": 0.286937952041626,
"learning_rate": 6.564943545777415e-05,
"loss": 1.359019660949707,
"step": 1100
},
{
"epoch": 0.4192355117139334,
"grad_norm": 0.29415133595466614,
"learning_rate": 6.53573040280995e-05,
"loss": 1.3365095138549805,
"step": 1105
},
{
"epoch": 0.4211325049796073,
"grad_norm": 0.2529754936695099,
"learning_rate": 6.506459259351457e-05,
"loss": 1.3641948699951172,
"step": 1110
},
{
"epoch": 0.42302949824528124,
"grad_norm": 0.2588936388492584,
"learning_rate": 6.477131220895896e-05,
"loss": 1.3539860725402832,
"step": 1115
},
{
"epoch": 0.42492649151095513,
"grad_norm": 0.26479557156562805,
"learning_rate": 6.447747395086013e-05,
"loss": 1.3693425178527832,
"step": 1120
},
{
"epoch": 0.42682348477662907,
"grad_norm": 0.2439931184053421,
"learning_rate": 6.418308891671484e-05,
"loss": 1.331023597717285,
"step": 1125
},
{
"epoch": 0.42872047804230295,
"grad_norm": 0.29339954257011414,
"learning_rate": 6.388816822467023e-05,
"loss": 1.327955722808838,
"step": 1130
},
{
"epoch": 0.43061747130797684,
"grad_norm": 0.2536592483520508,
"learning_rate": 6.35927230131038e-05,
"loss": 1.3192357063293456,
"step": 1135
},
{
"epoch": 0.4325144645736508,
"grad_norm": 0.23217377066612244,
"learning_rate": 6.329676444020273e-05,
"loss": 1.3994966506958009,
"step": 1140
},
{
"epoch": 0.43441145783932467,
"grad_norm": 0.2740131914615631,
"learning_rate": 6.300030368354255e-05,
"loss": 1.3187359809875487,
"step": 1145
},
{
"epoch": 0.43630845110499855,
"grad_norm": 0.25305646657943726,
"learning_rate": 6.270335193966495e-05,
"loss": 1.3131250381469726,
"step": 1150
},
{
"epoch": 0.4382054443706725,
"grad_norm": 0.23598378896713257,
"learning_rate": 6.240592042365488e-05,
"loss": 1.3426215171813964,
"step": 1155
},
{
"epoch": 0.4401024376363464,
"grad_norm": 0.32366031408309937,
"learning_rate": 6.210802036871707e-05,
"loss": 1.377659797668457,
"step": 1160
},
{
"epoch": 0.4419994309020203,
"grad_norm": 0.26842397451400757,
"learning_rate": 6.180966302575166e-05,
"loss": 1.3077903747558595,
"step": 1165
},
{
"epoch": 0.4438964241676942,
"grad_norm": 0.26482513546943665,
"learning_rate": 6.151085966292942e-05,
"loss": 1.2868882179260255,
"step": 1170
},
{
"epoch": 0.4457934174333681,
"grad_norm": 0.23800063133239746,
"learning_rate": 6.121162156526606e-05,
"loss": 1.3256553649902343,
"step": 1175
},
{
"epoch": 0.44769041069904203,
"grad_norm": 0.2715536057949066,
"learning_rate": 6.0911960034196156e-05,
"loss": 1.343141746520996,
"step": 1180
},
{
"epoch": 0.4495874039647159,
"grad_norm": 0.31748896837234497,
"learning_rate": 6.0611886387146156e-05,
"loss": 1.3046430587768554,
"step": 1185
},
{
"epoch": 0.45148439723038986,
"grad_norm": 0.2605152130126953,
"learning_rate": 6.031141195710711e-05,
"loss": 1.349820327758789,
"step": 1190
},
{
"epoch": 0.45338139049606374,
"grad_norm": 0.3222675919532776,
"learning_rate": 6.001054809220655e-05,
"loss": 1.3373382568359375,
"step": 1195
},
{
"epoch": 0.4552783837617376,
"grad_norm": 0.3359321355819702,
"learning_rate": 5.970930615527997e-05,
"loss": 1.2910158157348632,
"step": 1200
},
{
"epoch": 0.45717537702741157,
"grad_norm": 0.2673036754131317,
"learning_rate": 5.9407697523441576e-05,
"loss": 1.3502120971679688,
"step": 1205
},
{
"epoch": 0.45907237029308545,
"grad_norm": 0.3085758090019226,
"learning_rate": 5.9105733587654785e-05,
"loss": 1.3679951667785644,
"step": 1210
},
{
"epoch": 0.4609693635587594,
"grad_norm": 0.2747040390968323,
"learning_rate": 5.880342575230181e-05,
"loss": 1.365695095062256,
"step": 1215
},
{
"epoch": 0.4628663568244333,
"grad_norm": 0.2672959268093109,
"learning_rate": 5.850078543475311e-05,
"loss": 1.3438910484313964,
"step": 1220
},
{
"epoch": 0.46476335009010716,
"grad_norm": 0.27118828892707825,
"learning_rate": 5.8197824064936066e-05,
"loss": 1.300242805480957,
"step": 1225
},
{
"epoch": 0.4666603433557811,
"grad_norm": 0.25283756852149963,
"learning_rate": 5.7894553084903416e-05,
"loss": 1.2881752967834472,
"step": 1230
},
{
"epoch": 0.468557336621455,
"grad_norm": 0.27304574847221375,
"learning_rate": 5.7590983948401004e-05,
"loss": 1.4113713264465333,
"step": 1235
},
{
"epoch": 0.4704543298871289,
"grad_norm": 0.2680307626724243,
"learning_rate": 5.7287128120435284e-05,
"loss": 1.3545337677001954,
"step": 1240
},
{
"epoch": 0.4723513231528028,
"grad_norm": 0.2750462591648102,
"learning_rate": 5.6982997076840306e-05,
"loss": 1.334061050415039,
"step": 1245
},
{
"epoch": 0.4742483164184767,
"grad_norm": 0.24703142046928406,
"learning_rate": 5.667860230384427e-05,
"loss": 1.36375732421875,
"step": 1250
},
{
"epoch": 0.47614530968415064,
"grad_norm": 0.2511509358882904,
"learning_rate": 5.637395529763575e-05,
"loss": 1.3220529556274414,
"step": 1255
},
{
"epoch": 0.4780423029498245,
"grad_norm": 0.30084121227264404,
"learning_rate": 5.606906756392949e-05,
"loss": 1.270376205444336,
"step": 1260
},
{
"epoch": 0.4799392962154984,
"grad_norm": 0.2661868631839752,
"learning_rate": 5.576395061753187e-05,
"loss": 1.3405677795410156,
"step": 1265
},
{
"epoch": 0.48183628948117235,
"grad_norm": 0.31216517090797424,
"learning_rate": 5.545861598190607e-05,
"loss": 1.366221046447754,
"step": 1270
},
{
"epoch": 0.48373328274684624,
"grad_norm": 0.32247743010520935,
"learning_rate": 5.5153075188736766e-05,
"loss": 1.404726791381836,
"step": 1275
},
{
"epoch": 0.4856302760125202,
"grad_norm": 0.23004227876663208,
"learning_rate": 5.484733977749473e-05,
"loss": 1.3186360359191895,
"step": 1280
},
{
"epoch": 0.48752726927819406,
"grad_norm": 0.25774309039115906,
"learning_rate": 5.4541421295000864e-05,
"loss": 1.408599281311035,
"step": 1285
},
{
"epoch": 0.48942426254386795,
"grad_norm": 0.2686976194381714,
"learning_rate": 5.4235331294990266e-05,
"loss": 1.3324304580688477,
"step": 1290
},
{
"epoch": 0.4913212558095419,
"grad_norm": 0.2881031036376953,
"learning_rate": 5.392908133767572e-05,
"loss": 1.3579701423645019,
"step": 1295
},
{
"epoch": 0.4932182490752158,
"grad_norm": 0.30309098958969116,
"learning_rate": 5.362268298931127e-05,
"loss": 1.3555720329284668,
"step": 1300
},
{
"epoch": 0.4951152423408897,
"grad_norm": 0.24863609671592712,
"learning_rate": 5.3316147821755205e-05,
"loss": 1.3139801979064942,
"step": 1305
},
{
"epoch": 0.4970122356065636,
"grad_norm": 0.26706060767173767,
"learning_rate": 5.3009487412033166e-05,
"loss": 1.338627243041992,
"step": 1310
},
{
"epoch": 0.4989092288722375,
"grad_norm": 0.24213336408138275,
"learning_rate": 5.2702713341900866e-05,
"loss": 1.3194496154785156,
"step": 1315
},
{
"epoch": 0.5008062221379114,
"grad_norm": 0.27462077140808105,
"learning_rate": 5.239583719740664e-05,
"loss": 1.3304800987243652,
"step": 1320
},
{
"epoch": 0.5027032154035853,
"grad_norm": 0.32555609941482544,
"learning_rate": 5.208887056845394e-05,
"loss": 1.3162017822265626,
"step": 1325
},
{
"epoch": 0.5046002086692593,
"grad_norm": 0.2596307694911957,
"learning_rate": 5.178182504836352e-05,
"loss": 1.2875455856323241,
"step": 1330
},
{
"epoch": 0.5064972019349331,
"grad_norm": 0.2573661804199219,
"learning_rate": 5.147471223343573e-05,
"loss": 1.240321731567383,
"step": 1335
},
{
"epoch": 0.508394195200607,
"grad_norm": 0.2391667664051056,
"learning_rate": 5.116754372251238e-05,
"loss": 1.3177503585815429,
"step": 1340
},
{
"epoch": 0.510291188466281,
"grad_norm": 0.34576547145843506,
"learning_rate": 5.086033111653884e-05,
"loss": 1.3279438972473145,
"step": 1345
},
{
"epoch": 0.5121881817319549,
"grad_norm": 0.28111353516578674,
"learning_rate": 5.05530860181258e-05,
"loss": 1.3314611434936523,
"step": 1350
},
{
"epoch": 0.5140851749976287,
"grad_norm": 0.27309855818748474,
"learning_rate": 5.024582003111109e-05,
"loss": 1.3239142417907714,
"step": 1355
},
{
"epoch": 0.5159821682633027,
"grad_norm": 0.34345176815986633,
"learning_rate": 4.99385447601215e-05,
"loss": 1.3664510726928711,
"step": 1360
},
{
"epoch": 0.5178791615289766,
"grad_norm": 0.23667146265506744,
"learning_rate": 4.963127181013438e-05,
"loss": 1.3228439331054687,
"step": 1365
},
{
"epoch": 0.5197761547946504,
"grad_norm": 0.23325034976005554,
"learning_rate": 4.932401278603949e-05,
"loss": 1.3221150398254395,
"step": 1370
},
{
"epoch": 0.5216731480603244,
"grad_norm": 0.2858182489871979,
"learning_rate": 4.9016779292200614e-05,
"loss": 1.3203374862670898,
"step": 1375
},
{
"epoch": 0.5235701413259983,
"grad_norm": 0.24023060500621796,
"learning_rate": 4.8709582932017324e-05,
"loss": 1.285901165008545,
"step": 1380
},
{
"epoch": 0.5254671345916722,
"grad_norm": 0.34592846035957336,
"learning_rate": 4.8402435307486734e-05,
"loss": 1.3382051467895508,
"step": 1385
},
{
"epoch": 0.5273641278573461,
"grad_norm": 0.26865020394325256,
"learning_rate": 4.809534801876538e-05,
"loss": 1.3635382652282715,
"step": 1390
},
{
"epoch": 0.52926112112302,
"grad_norm": 0.2539322078227997,
"learning_rate": 4.7788332663731066e-05,
"loss": 1.3386174201965333,
"step": 1395
},
{
"epoch": 0.5311581143886939,
"grad_norm": 0.2540978193283081,
"learning_rate": 4.7481400837544825e-05,
"loss": 1.371791934967041,
"step": 1400
},
{
"epoch": 0.5330551076543678,
"grad_norm": 0.29954543709754944,
"learning_rate": 4.717456413221302e-05,
"loss": 1.4154946327209472,
"step": 1405
},
{
"epoch": 0.5349521009200418,
"grad_norm": 0.23236536979675293,
"learning_rate": 4.6867834136149615e-05,
"loss": 1.3220270156860352,
"step": 1410
},
{
"epoch": 0.5368490941857157,
"grad_norm": 0.24130581319332123,
"learning_rate": 4.656122243373838e-05,
"loss": 1.3156471252441406,
"step": 1415
},
{
"epoch": 0.5387460874513895,
"grad_norm": 0.23898352682590485,
"learning_rate": 4.625474060489546e-05,
"loss": 1.3023578643798828,
"step": 1420
},
{
"epoch": 0.5406430807170635,
"grad_norm": 0.2676070034503937,
"learning_rate": 4.5948400224632e-05,
"loss": 1.3223109245300293,
"step": 1425
},
{
"epoch": 0.5425400739827374,
"grad_norm": 0.295886367559433,
"learning_rate": 4.564221286261709e-05,
"loss": 1.3164411544799806,
"step": 1430
},
{
"epoch": 0.5444370672484112,
"grad_norm": 0.2820796072483063,
"learning_rate": 4.5336190082740625e-05,
"loss": 1.2911755561828613,
"step": 1435
},
{
"epoch": 0.5463340605140852,
"grad_norm": 0.24667897820472717,
"learning_rate": 4.503034344267671e-05,
"loss": 1.2966703414916991,
"step": 1440
},
{
"epoch": 0.5482310537797591,
"grad_norm": 0.2428257316350937,
"learning_rate": 4.472468449344709e-05,
"loss": 1.2589365005493165,
"step": 1445
},
{
"epoch": 0.5501280470454329,
"grad_norm": 0.2763369679450989,
"learning_rate": 4.441922477898497e-05,
"loss": 1.3005006790161133,
"step": 1450
},
{
"epoch": 0.5520250403111069,
"grad_norm": 0.2695941925048828,
"learning_rate": 4.411397583569894e-05,
"loss": 1.3524924278259278,
"step": 1455
},
{
"epoch": 0.5539220335767808,
"grad_norm": 0.29227209091186523,
"learning_rate": 4.3808949192037325e-05,
"loss": 1.3824165344238282,
"step": 1460
},
{
"epoch": 0.5558190268424547,
"grad_norm": 0.24675939977169037,
"learning_rate": 4.3504156368052765e-05,
"loss": 1.3211604118347169,
"step": 1465
},
{
"epoch": 0.5577160201081286,
"grad_norm": 0.2530978322029114,
"learning_rate": 4.3199608874967193e-05,
"loss": 1.3258498191833497,
"step": 1470
},
{
"epoch": 0.5596130133738025,
"grad_norm": 0.2500854730606079,
"learning_rate": 4.289531821473702e-05,
"loss": 1.318402099609375,
"step": 1475
},
{
"epoch": 0.5615100066394765,
"grad_norm": 0.26825377345085144,
"learning_rate": 4.259129587961873e-05,
"loss": 1.297877025604248,
"step": 1480
},
{
"epoch": 0.5634069999051503,
"grad_norm": 0.3033313751220703,
"learning_rate": 4.228755335173488e-05,
"loss": 1.3168060302734375,
"step": 1485
},
{
"epoch": 0.5653039931708242,
"grad_norm": 0.341702401638031,
"learning_rate": 4.19841021026405e-05,
"loss": 1.2362494468688965,
"step": 1490
},
{
"epoch": 0.5672009864364982,
"grad_norm": 0.2761666774749756,
"learning_rate": 4.168095359288972e-05,
"loss": 1.3344615936279296,
"step": 1495
},
{
"epoch": 0.569097979702172,
"grad_norm": 0.25564298033714294,
"learning_rate": 4.137811927160304e-05,
"loss": 1.2510947227478026,
"step": 1500
},
{
"epoch": 0.570994972967846,
"grad_norm": 0.29818210005760193,
"learning_rate": 4.107561057603486e-05,
"loss": 1.3625399589538574,
"step": 1505
},
{
"epoch": 0.5728919662335199,
"grad_norm": 0.30387359857559204,
"learning_rate": 4.077343893114159e-05,
"loss": 1.3149986267089844,
"step": 1510
},
{
"epoch": 0.5747889594991937,
"grad_norm": 0.27656859159469604,
"learning_rate": 4.047161574915011e-05,
"loss": 1.2504589080810546,
"step": 1515
},
{
"epoch": 0.5766859527648677,
"grad_norm": 0.2697388827800751,
"learning_rate": 4.017015242912677e-05,
"loss": 1.317511749267578,
"step": 1520
},
{
"epoch": 0.5785829460305416,
"grad_norm": 0.2973303496837616,
"learning_rate": 3.9869060356546855e-05,
"loss": 1.332118034362793,
"step": 1525
},
{
"epoch": 0.5804799392962156,
"grad_norm": 0.275661438703537,
"learning_rate": 3.956835090286468e-05,
"loss": 1.3418188095092773,
"step": 1530
},
{
"epoch": 0.5823769325618894,
"grad_norm": 0.23285940289497375,
"learning_rate": 3.9268035425084006e-05,
"loss": 1.3276552200317382,
"step": 1535
},
{
"epoch": 0.5842739258275633,
"grad_norm": 0.2696329355239868,
"learning_rate": 3.896812526532916e-05,
"loss": 1.3003293991088867,
"step": 1540
},
{
"epoch": 0.5861709190932373,
"grad_norm": 0.34393227100372314,
"learning_rate": 3.866863175041666e-05,
"loss": 1.3268285751342774,
"step": 1545
},
{
"epoch": 0.5880679123589111,
"grad_norm": 0.29866087436676025,
"learning_rate": 3.836956619142751e-05,
"loss": 1.3523910522460938,
"step": 1550
},
{
"epoch": 0.589964905624585,
"grad_norm": 0.2631630003452301,
"learning_rate": 3.80709398832799e-05,
"loss": 1.3111137390136718,
"step": 1555
},
{
"epoch": 0.591861898890259,
"grad_norm": 0.2786312401294708,
"learning_rate": 3.777276410430267e-05,
"loss": 1.311577606201172,
"step": 1560
},
{
"epoch": 0.5937588921559328,
"grad_norm": 0.2818828523159027,
"learning_rate": 3.747505011580936e-05,
"loss": 1.3115427017211914,
"step": 1565
},
{
"epoch": 0.5956558854216067,
"grad_norm": 0.2687928378582001,
"learning_rate": 3.7177809161672934e-05,
"loss": 1.3268730163574218,
"step": 1570
},
{
"epoch": 0.5975528786872807,
"grad_norm": 0.2781391441822052,
"learning_rate": 3.688105246790106e-05,
"loss": 1.3441936492919921,
"step": 1575
},
{
"epoch": 0.5994498719529545,
"grad_norm": 0.29178586602211,
"learning_rate": 3.6584791242212166e-05,
"loss": 1.3358587265014648,
"step": 1580
},
{
"epoch": 0.6013468652186285,
"grad_norm": 0.28743571043014526,
"learning_rate": 3.628903667361216e-05,
"loss": 1.3186549186706542,
"step": 1585
},
{
"epoch": 0.6032438584843024,
"grad_norm": 0.2775789499282837,
"learning_rate": 3.5993799931971905e-05,
"loss": 1.3131633758544923,
"step": 1590
},
{
"epoch": 0.6051408517499763,
"grad_norm": 0.2770810127258301,
"learning_rate": 3.569909216760522e-05,
"loss": 1.297097873687744,
"step": 1595
},
{
"epoch": 0.6070378450156502,
"grad_norm": 0.25473111867904663,
"learning_rate": 3.5404924510847884e-05,
"loss": 1.335860824584961,
"step": 1600
},
{
"epoch": 0.6089348382813241,
"grad_norm": 0.2893996834754944,
"learning_rate": 3.511130807163724e-05,
"loss": 1.2592733383178711,
"step": 1605
},
{
"epoch": 0.610831831546998,
"grad_norm": 0.2767355442047119,
"learning_rate": 3.481825393909255e-05,
"loss": 1.3512200355529784,
"step": 1610
},
{
"epoch": 0.6127288248126719,
"grad_norm": 0.2869298756122589,
"learning_rate": 3.452577318109632e-05,
"loss": 1.3195438385009766,
"step": 1615
},
{
"epoch": 0.6146258180783458,
"grad_norm": 0.2852262854576111,
"learning_rate": 3.423387684387615e-05,
"loss": 1.3630094528198242,
"step": 1620
},
{
"epoch": 0.6165228113440198,
"grad_norm": 0.3178340494632721,
"learning_rate": 3.394257595158761e-05,
"loss": 1.305915641784668,
"step": 1625
},
{
"epoch": 0.6184198046096936,
"grad_norm": 0.28296470642089844,
"learning_rate": 3.365188150589789e-05,
"loss": 1.3004621505737304,
"step": 1630
},
{
"epoch": 0.6203167978753675,
"grad_norm": 0.27672746777534485,
"learning_rate": 3.336180448557031e-05,
"loss": 1.2676490783691405,
"step": 1635
},
{
"epoch": 0.6222137911410415,
"grad_norm": 0.2730206549167633,
"learning_rate": 3.307235584604964e-05,
"loss": 1.2916270256042481,
"step": 1640
},
{
"epoch": 0.6241107844067153,
"grad_norm": 0.24284686148166656,
"learning_rate": 3.278354651904836e-05,
"loss": 1.3340941429138184,
"step": 1645
},
{
"epoch": 0.6260077776723892,
"grad_norm": 0.2962943911552429,
"learning_rate": 3.249538741213378e-05,
"loss": 1.2806172370910645,
"step": 1650
},
{
"epoch": 0.6279047709380632,
"grad_norm": 0.2714846134185791,
"learning_rate": 3.220788940831619e-05,
"loss": 1.245294952392578,
"step": 1655
},
{
"epoch": 0.6298017642037371,
"grad_norm": 0.3047868311405182,
"learning_rate": 3.192106336563768e-05,
"loss": 1.3315817832946777,
"step": 1660
},
{
"epoch": 0.631698757469411,
"grad_norm": 0.2749776840209961,
"learning_rate": 3.1634920116762176e-05,
"loss": 1.283216667175293,
"step": 1665
},
{
"epoch": 0.6335957507350849,
"grad_norm": 0.27595117688179016,
"learning_rate": 3.1349470468566226e-05,
"loss": 1.3455049514770507,
"step": 1670
},
{
"epoch": 0.6354927440007588,
"grad_norm": 0.31003913283348083,
"learning_rate": 3.106472520173101e-05,
"loss": 1.3287286758422852,
"step": 1675
},
{
"epoch": 0.6373897372664327,
"grad_norm": 0.30897730588912964,
"learning_rate": 3.0780695070334995e-05,
"loss": 1.276911163330078,
"step": 1680
},
{
"epoch": 0.6392867305321066,
"grad_norm": 0.29919230937957764,
"learning_rate": 3.0497390801447883e-05,
"loss": 1.2657719612121583,
"step": 1685
},
{
"epoch": 0.6411837237977805,
"grad_norm": 0.3092974126338959,
"learning_rate": 3.0214823094725474e-05,
"loss": 1.2836290359497071,
"step": 1690
},
{
"epoch": 0.6430807170634544,
"grad_norm": 0.38299307227134705,
"learning_rate": 2.993300262200556e-05,
"loss": 1.2660253524780274,
"step": 1695
},
{
"epoch": 0.6449777103291283,
"grad_norm": 0.26396632194519043,
"learning_rate": 2.965194002690487e-05,
"loss": 1.2721212387084961,
"step": 1700
},
{
"epoch": 0.6468747035948023,
"grad_norm": 0.2829623222351074,
"learning_rate": 2.937164592441708e-05,
"loss": 1.3260103225708009,
"step": 1705
},
{
"epoch": 0.6487716968604762,
"grad_norm": 0.25421395897865295,
"learning_rate": 2.909213090051191e-05,
"loss": 1.2883237838745116,
"step": 1710
},
{
"epoch": 0.65066869012615,
"grad_norm": 0.2579733729362488,
"learning_rate": 2.8813405511735393e-05,
"loss": 1.3096033096313477,
"step": 1715
},
{
"epoch": 0.652565683391824,
"grad_norm": 0.24724499881267548,
"learning_rate": 2.85354802848111e-05,
"loss": 1.2852197647094727,
"step": 1720
},
{
"epoch": 0.6544626766574979,
"grad_norm": 0.4333013594150543,
"learning_rate": 2.825836571624254e-05,
"loss": 1.2738615036010743,
"step": 1725
},
{
"epoch": 0.6563596699231717,
"grad_norm": 0.25200602412223816,
"learning_rate": 2.7982072271916852e-05,
"loss": 1.2166935920715332,
"step": 1730
},
{
"epoch": 0.6582566631888457,
"grad_norm": 0.3310626745223999,
"learning_rate": 2.7706610386709498e-05,
"loss": 1.3035533905029297,
"step": 1735
},
{
"epoch": 0.6601536564545196,
"grad_norm": 0.2914412319660187,
"learning_rate": 2.743199046409005e-05,
"loss": 1.3005502700805665,
"step": 1740
},
{
"epoch": 0.6620506497201935,
"grad_norm": 0.28406983613967896,
"learning_rate": 2.7158222875729472e-05,
"loss": 1.2131826400756835,
"step": 1745
},
{
"epoch": 0.6639476429858674,
"grad_norm": 0.3953897953033447,
"learning_rate": 2.6885317961108202e-05,
"loss": 1.3039031982421876,
"step": 1750
},
{
"epoch": 0.6658446362515413,
"grad_norm": 0.28072869777679443,
"learning_rate": 2.6613286027125915e-05,
"loss": 1.2851948738098145,
"step": 1755
},
{
"epoch": 0.6677416295172152,
"grad_norm": 0.24564129114151,
"learning_rate": 2.6342137347711952e-05,
"loss": 1.30118989944458,
"step": 1760
},
{
"epoch": 0.6696386227828891,
"grad_norm": 0.24667160212993622,
"learning_rate": 2.607188216343756e-05,
"loss": 1.265779972076416,
"step": 1765
},
{
"epoch": 0.671535616048563,
"grad_norm": 0.27209749817848206,
"learning_rate": 2.5802530681128935e-05,
"loss": 1.318216037750244,
"step": 1770
},
{
"epoch": 0.673432609314237,
"grad_norm": 0.3034723699092865,
"learning_rate": 2.553409307348197e-05,
"loss": 1.2217689514160157,
"step": 1775
},
{
"epoch": 0.6753296025799108,
"grad_norm": 0.3007160425186157,
"learning_rate": 2.5266579478677765e-05,
"loss": 1.2883822441101074,
"step": 1780
},
{
"epoch": 0.6772265958455848,
"grad_norm": 0.27932560443878174,
"learning_rate": 2.500000000000001e-05,
"loss": 1.3245506286621094,
"step": 1785
},
{
"epoch": 0.6791235891112587,
"grad_norm": 0.2890223264694214,
"learning_rate": 2.47343647054532e-05,
"loss": 1.3349345207214356,
"step": 1790
},
{
"epoch": 0.6810205823769325,
"grad_norm": 0.2815885543823242,
"learning_rate": 2.446968362738255e-05,
"loss": 1.2415434837341308,
"step": 1795
},
{
"epoch": 0.6829175756426065,
"grad_norm": 0.28444790840148926,
"learning_rate": 2.4205966762095016e-05,
"loss": 1.272564697265625,
"step": 1800
},
{
"epoch": 0.6848145689082804,
"grad_norm": 0.36384108662605286,
"learning_rate": 2.3943224069481734e-05,
"loss": 1.295553970336914,
"step": 1805
},
{
"epoch": 0.6867115621739542,
"grad_norm": 0.29714980721473694,
"learning_rate": 2.3681465472641967e-05,
"loss": 1.2582395553588868,
"step": 1810
},
{
"epoch": 0.6886085554396282,
"grad_norm": 0.263151079416275,
"learning_rate": 2.3420700857508232e-05,
"loss": 1.283635711669922,
"step": 1815
},
{
"epoch": 0.6905055487053021,
"grad_norm": 0.30710259079933167,
"learning_rate": 2.316094007247302e-05,
"loss": 1.3082473754882813,
"step": 1820
},
{
"epoch": 0.692402541970976,
"grad_norm": 0.24722664058208466,
"learning_rate": 2.2902192928016734e-05,
"loss": 1.273237133026123,
"step": 1825
},
{
"epoch": 0.6942995352366499,
"grad_norm": 0.2690366506576538,
"learning_rate": 2.26444691963373e-05,
"loss": 1.3232413291931153,
"step": 1830
},
{
"epoch": 0.6961965285023238,
"grad_norm": 0.2856862545013428,
"learning_rate": 2.238777861098102e-05,
"loss": 1.2787722587585448,
"step": 1835
},
{
"epoch": 0.6980935217679978,
"grad_norm": 0.27475956082344055,
"learning_rate": 2.213213086647501e-05,
"loss": 1.2913030624389648,
"step": 1840
},
{
"epoch": 0.6999905150336716,
"grad_norm": 0.25229784846305847,
"learning_rate": 2.187753561796097e-05,
"loss": 1.2863998413085938,
"step": 1845
},
{
"epoch": 0.7018875082993455,
"grad_norm": 0.23411835730075836,
"learning_rate": 2.162400248083065e-05,
"loss": 1.3396069526672363,
"step": 1850
},
{
"epoch": 0.7037845015650195,
"grad_norm": 0.34195366501808167,
"learning_rate": 2.1371541030362674e-05,
"loss": 1.3385623931884765,
"step": 1855
},
{
"epoch": 0.7056814948306933,
"grad_norm": 0.2691027522087097,
"learning_rate": 2.1120160801360794e-05,
"loss": 1.275947380065918,
"step": 1860
},
{
"epoch": 0.7075784880963673,
"grad_norm": 0.3500896990299225,
"learning_rate": 2.0869871287793985e-05,
"loss": 1.3039764404296874,
"step": 1865
},
{
"epoch": 0.7094754813620412,
"grad_norm": 0.30469706654548645,
"learning_rate": 2.0620681942437665e-05,
"loss": 1.277679443359375,
"step": 1870
},
{
"epoch": 0.711372474627715,
"grad_norm": 0.3132733106613159,
"learning_rate": 2.0372602176516914e-05,
"loss": 1.2633010864257812,
"step": 1875
},
{
"epoch": 0.713269467893389,
"grad_norm": 0.26674601435661316,
"learning_rate": 2.0125641359350815e-05,
"loss": 1.2836494445800781,
"step": 1880
},
{
"epoch": 0.7151664611590629,
"grad_norm": 0.2870766222476959,
"learning_rate": 1.987980881799879e-05,
"loss": 1.2626848220825195,
"step": 1885
},
{
"epoch": 0.7170634544247368,
"grad_norm": 0.25851741433143616,
"learning_rate": 1.963511383690817e-05,
"loss": 1.3008936882019042,
"step": 1890
},
{
"epoch": 0.7189604476904107,
"grad_norm": 0.2628515660762787,
"learning_rate": 1.9391565657563698e-05,
"loss": 1.2477121353149414,
"step": 1895
},
{
"epoch": 0.7208574409560846,
"grad_norm": 0.23573534190654755,
"learning_rate": 1.9149173478138422e-05,
"loss": 1.2801203727722168,
"step": 1900
},
{
"epoch": 0.7227544342217586,
"grad_norm": 0.28856074810028076,
"learning_rate": 1.890794645314633e-05,
"loss": 1.294834804534912,
"step": 1905
},
{
"epoch": 0.7246514274874324,
"grad_norm": 0.3120816946029663,
"learning_rate": 1.866789369309654e-05,
"loss": 1.2395054817199707,
"step": 1910
},
{
"epoch": 0.7265484207531063,
"grad_norm": 0.2720513641834259,
"learning_rate": 1.8429024264149337e-05,
"loss": 1.2873690605163575,
"step": 1915
},
{
"epoch": 0.7284454140187803,
"grad_norm": 0.2862343490123749,
"learning_rate": 1.8191347187773704e-05,
"loss": 1.2763073921203614,
"step": 1920
},
{
"epoch": 0.7303424072844541,
"grad_norm": 0.29964083433151245,
"learning_rate": 1.7954871440406562e-05,
"loss": 1.2217901229858399,
"step": 1925
},
{
"epoch": 0.732239400550128,
"grad_norm": 0.26421642303466797,
"learning_rate": 1.7719605953113865e-05,
"loss": 1.2820805549621581,
"step": 1930
},
{
"epoch": 0.734136393815802,
"grad_norm": 0.30602434277534485,
"learning_rate": 1.748555961125315e-05,
"loss": 1.3283105850219727,
"step": 1935
},
{
"epoch": 0.7360333870814758,
"grad_norm": 0.29646191000938416,
"learning_rate": 1.7252741254138162e-05,
"loss": 1.3078016281127929,
"step": 1940
},
{
"epoch": 0.7379303803471498,
"grad_norm": 0.2863771915435791,
"learning_rate": 1.702115967470477e-05,
"loss": 1.2773143768310546,
"step": 1945
},
{
"epoch": 0.7398273736128237,
"grad_norm": 0.3262138366699219,
"learning_rate": 1.6790823619179107e-05,
"loss": 1.310184860229492,
"step": 1950
},
{
"epoch": 0.7417243668784976,
"grad_norm": 0.2814273536205292,
"learning_rate": 1.6561741786747058e-05,
"loss": 1.2708742141723632,
"step": 1955
},
{
"epoch": 0.7436213601441715,
"grad_norm": 0.32096412777900696,
"learning_rate": 1.633392282922593e-05,
"loss": 1.2818357467651367,
"step": 1960
},
{
"epoch": 0.7455183534098454,
"grad_norm": 0.38933348655700684,
"learning_rate": 1.6107375350737437e-05,
"loss": 1.3111585617065429,
"step": 1965
},
{
"epoch": 0.7474153466755193,
"grad_norm": 0.24294926226139069,
"learning_rate": 1.588210790738299e-05,
"loss": 1.3092460632324219,
"step": 1970
},
{
"epoch": 0.7493123399411932,
"grad_norm": 0.25524646043777466,
"learning_rate": 1.5658129006920352e-05,
"loss": 1.2489660263061524,
"step": 1975
},
{
"epoch": 0.7512093332068671,
"grad_norm": 0.2851572632789612,
"learning_rate": 1.5435447108442497e-05,
"loss": 1.3061258316040039,
"step": 1980
},
{
"epoch": 0.753106326472541,
"grad_norm": 0.3529151380062103,
"learning_rate": 1.5214070622058013e-05,
"loss": 1.276413345336914,
"step": 1985
},
{
"epoch": 0.7550033197382149,
"grad_norm": 0.27792128920555115,
"learning_rate": 1.4994007908573499e-05,
"loss": 1.2463918685913087,
"step": 1990
},
{
"epoch": 0.7569003130038888,
"grad_norm": 0.25739148259162903,
"learning_rate": 1.477526727917784e-05,
"loss": 1.2394469261169434,
"step": 1995
},
{
"epoch": 0.7587973062695628,
"grad_norm": 0.27975818514823914,
"learning_rate": 1.4557856995128282e-05,
"loss": 1.32928409576416,
"step": 2000
}
],
"logging_steps": 5,
"max_steps": 2636,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.2927736456609792e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}