u0 / trainer_state.json
Kaipengm2's picture
Upload folder using huggingface_hub
3218fd2 verified
Raw
History Blame Contribute Delete
87.2 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2328,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00214799699280421,
"grad_norm": 1.6932775974273682,
"learning_rate": 5.7142857142857145e-06,
"loss": 2.9927764892578126,
"step": 5
},
{
"epoch": 0.00429599398560842,
"grad_norm": 1.2265691757202148,
"learning_rate": 1.2857142857142857e-05,
"loss": 2.8719606399536133,
"step": 10
},
{
"epoch": 0.0064439909784126306,
"grad_norm": 0.6039227247238159,
"learning_rate": 2e-05,
"loss": 2.7400981903076174,
"step": 15
},
{
"epoch": 0.00859198797121684,
"grad_norm": 0.29856061935424805,
"learning_rate": 2.714285714285714e-05,
"loss": 2.593766784667969,
"step": 20
},
{
"epoch": 0.01073998496402105,
"grad_norm": 0.3089836835861206,
"learning_rate": 3.428571428571429e-05,
"loss": 2.472783660888672,
"step": 25
},
{
"epoch": 0.012887981956825261,
"grad_norm": 0.30488327145576477,
"learning_rate": 4.1428571428571437e-05,
"loss": 2.3986921310424805,
"step": 30
},
{
"epoch": 0.01503597894962947,
"grad_norm": 0.243967205286026,
"learning_rate": 4.8571428571428576e-05,
"loss": 2.3093551635742187,
"step": 35
},
{
"epoch": 0.01718397594243368,
"grad_norm": 0.23277634382247925,
"learning_rate": 5.571428571428572e-05,
"loss": 2.2308311462402344,
"step": 40
},
{
"epoch": 0.01933197293523789,
"grad_norm": 0.2143218070268631,
"learning_rate": 6.285714285714286e-05,
"loss": 2.2247371673583984,
"step": 45
},
{
"epoch": 0.0214799699280421,
"grad_norm": 0.2504544258117676,
"learning_rate": 7e-05,
"loss": 2.113149070739746,
"step": 50
},
{
"epoch": 0.02362796692084631,
"grad_norm": 0.23724156618118286,
"learning_rate": 7.714285714285715e-05,
"loss": 2.064219856262207,
"step": 55
},
{
"epoch": 0.025775963913650522,
"grad_norm": 0.2178376168012619,
"learning_rate": 8.428571428571429e-05,
"loss": 1.9700836181640624,
"step": 60
},
{
"epoch": 0.027923960906454732,
"grad_norm": 0.2316979169845581,
"learning_rate": 9.142857142857143e-05,
"loss": 1.9064104080200195,
"step": 65
},
{
"epoch": 0.03007195789925894,
"grad_norm": 0.22800028324127197,
"learning_rate": 9.857142857142858e-05,
"loss": 1.8772602081298828,
"step": 70
},
{
"epoch": 0.032219954892063155,
"grad_norm": 0.21327374875545502,
"learning_rate": 9.999922569735892e-05,
"loss": 1.898672103881836,
"step": 75
},
{
"epoch": 0.03436795188486736,
"grad_norm": 0.21225009858608246,
"learning_rate": 9.999608013398099e-05,
"loss": 1.7929576873779296,
"step": 80
},
{
"epoch": 0.036515948877671574,
"grad_norm": 0.202202707529068,
"learning_rate": 9.999051506806e-05,
"loss": 1.8280349731445313,
"step": 85
},
{
"epoch": 0.03866394587047578,
"grad_norm": 0.21529272198677063,
"learning_rate": 9.998253076891087e-05,
"loss": 1.7857349395751954,
"step": 90
},
{
"epoch": 0.04081194286327999,
"grad_norm": 0.21258528530597687,
"learning_rate": 9.997212762292454e-05,
"loss": 1.7510189056396483,
"step": 95
},
{
"epoch": 0.0429599398560842,
"grad_norm": 0.23243068158626556,
"learning_rate": 9.995930613354923e-05,
"loss": 1.7037622451782226,
"step": 100
},
{
"epoch": 0.04510793684888841,
"grad_norm": 0.23923775553703308,
"learning_rate": 9.99440669212661e-05,
"loss": 1.7596521377563477,
"step": 105
},
{
"epoch": 0.04725593384169262,
"grad_norm": 0.20659016072750092,
"learning_rate": 9.992641072355927e-05,
"loss": 1.7518970489501953,
"step": 110
},
{
"epoch": 0.04940393083449683,
"grad_norm": 0.24057933688163757,
"learning_rate": 9.990633839487996e-05,
"loss": 1.64355411529541,
"step": 115
},
{
"epoch": 0.051551927827301045,
"grad_norm": 0.22050833702087402,
"learning_rate": 9.98838509066054e-05,
"loss": 1.7364654541015625,
"step": 120
},
{
"epoch": 0.05369992482010525,
"grad_norm": 0.23694919049739838,
"learning_rate": 9.985894934699154e-05,
"loss": 1.646773338317871,
"step": 125
},
{
"epoch": 0.055847921812909464,
"grad_norm": 0.22422945499420166,
"learning_rate": 9.983163492112066e-05,
"loss": 1.685190200805664,
"step": 130
},
{
"epoch": 0.05799591880571367,
"grad_norm": 0.23080891370773315,
"learning_rate": 9.980190895084279e-05,
"loss": 1.6647254943847656,
"step": 135
},
{
"epoch": 0.06014391579851788,
"grad_norm": 0.26722222566604614,
"learning_rate": 9.976977287471197e-05,
"loss": 1.702375602722168,
"step": 140
},
{
"epoch": 0.06229191279132209,
"grad_norm": 0.3150525689125061,
"learning_rate": 9.973522824791644e-05,
"loss": 1.6489757537841796,
"step": 145
},
{
"epoch": 0.06443990978412631,
"grad_norm": 0.2626664340496063,
"learning_rate": 9.969827674220353e-05,
"loss": 1.649013137817383,
"step": 150
},
{
"epoch": 0.06658790677693051,
"grad_norm": 0.22747838497161865,
"learning_rate": 9.965892014579867e-05,
"loss": 1.6707025527954102,
"step": 155
},
{
"epoch": 0.06873590376973472,
"grad_norm": 0.2435477077960968,
"learning_rate": 9.961716036331886e-05,
"loss": 1.6180774688720703,
"step": 160
},
{
"epoch": 0.07088390076253893,
"grad_norm": 0.28238314390182495,
"learning_rate": 9.957299941568057e-05,
"loss": 1.5666514396667481,
"step": 165
},
{
"epoch": 0.07303189775534315,
"grad_norm": 0.2607802748680115,
"learning_rate": 9.952643944000181e-05,
"loss": 1.6158615112304688,
"step": 170
},
{
"epoch": 0.07517989474814735,
"grad_norm": 0.2308652698993683,
"learning_rate": 9.947748268949885e-05,
"loss": 1.6345035552978515,
"step": 175
},
{
"epoch": 0.07732789174095156,
"grad_norm": 0.28599420189857483,
"learning_rate": 9.942613153337705e-05,
"loss": 1.6306491851806642,
"step": 180
},
{
"epoch": 0.07947588873375577,
"grad_norm": 0.22085429728031158,
"learning_rate": 9.937238845671631e-05,
"loss": 1.591522789001465,
"step": 185
},
{
"epoch": 0.08162388572655999,
"grad_norm": 0.2728830575942993,
"learning_rate": 9.931625606035072e-05,
"loss": 1.6046226501464844,
"step": 190
},
{
"epoch": 0.0837718827193642,
"grad_norm": 0.2554752826690674,
"learning_rate": 9.925773706074278e-05,
"loss": 1.6327739715576173,
"step": 195
},
{
"epoch": 0.0859198797121684,
"grad_norm": 0.25561556220054626,
"learning_rate": 9.919683428985185e-05,
"loss": 1.5859023094177247,
"step": 200
},
{
"epoch": 0.08806787670497261,
"grad_norm": 0.2333848774433136,
"learning_rate": 9.91335506949972e-05,
"loss": 1.5448897361755372,
"step": 205
},
{
"epoch": 0.09021587369777682,
"grad_norm": 0.28334739804267883,
"learning_rate": 9.906788933871529e-05,
"loss": 1.6283615112304688,
"step": 210
},
{
"epoch": 0.09236387069058104,
"grad_norm": 0.22497156262397766,
"learning_rate": 9.89998533986116e-05,
"loss": 1.6024372100830078,
"step": 215
},
{
"epoch": 0.09451186768338524,
"grad_norm": 0.24012039601802826,
"learning_rate": 9.892944616720685e-05,
"loss": 1.5988900184631347,
"step": 220
},
{
"epoch": 0.09665986467618945,
"grad_norm": 0.23542068898677826,
"learning_rate": 9.885667105177768e-05,
"loss": 1.5697736740112305,
"step": 225
},
{
"epoch": 0.09880786166899366,
"grad_norm": 0.2410110980272293,
"learning_rate": 9.878153157419174e-05,
"loss": 1.619461441040039,
"step": 230
},
{
"epoch": 0.10095585866179788,
"grad_norm": 0.24803216755390167,
"learning_rate": 9.870403137073722e-05,
"loss": 1.5621639251708985,
"step": 235
},
{
"epoch": 0.10310385565460209,
"grad_norm": 0.2517596185207367,
"learning_rate": 9.862417419194698e-05,
"loss": 1.5740959167480468,
"step": 240
},
{
"epoch": 0.10525185264740629,
"grad_norm": 0.24555464088916779,
"learning_rate": 9.854196390241691e-05,
"loss": 1.5601798057556153,
"step": 245
},
{
"epoch": 0.1073998496402105,
"grad_norm": 0.29419493675231934,
"learning_rate": 9.8457404480619e-05,
"loss": 1.5268316268920898,
"step": 250
},
{
"epoch": 0.10954784663301471,
"grad_norm": 0.29217928647994995,
"learning_rate": 9.83705000187088e-05,
"loss": 1.5419147491455079,
"step": 255
},
{
"epoch": 0.11169584362581893,
"grad_norm": 0.29212093353271484,
"learning_rate": 9.828125472232734e-05,
"loss": 1.5635286331176759,
"step": 260
},
{
"epoch": 0.11384384061862314,
"grad_norm": 0.26869240403175354,
"learning_rate": 9.818967291039767e-05,
"loss": 1.557928466796875,
"step": 265
},
{
"epoch": 0.11599183761142734,
"grad_norm": 0.24330586194992065,
"learning_rate": 9.809575901491574e-05,
"loss": 1.510034942626953,
"step": 270
},
{
"epoch": 0.11813983460423155,
"grad_norm": 0.2569581866264343,
"learning_rate": 9.799951758073606e-05,
"loss": 1.5425460815429688,
"step": 275
},
{
"epoch": 0.12028783159703577,
"grad_norm": 0.2696671187877655,
"learning_rate": 9.790095326535167e-05,
"loss": 1.5411754608154298,
"step": 280
},
{
"epoch": 0.12243582858983998,
"grad_norm": 0.251804918050766,
"learning_rate": 9.780007083866872e-05,
"loss": 1.47529296875,
"step": 285
},
{
"epoch": 0.12458382558264418,
"grad_norm": 0.25417736172676086,
"learning_rate": 9.769687518277573e-05,
"loss": 1.497340774536133,
"step": 290
},
{
"epoch": 0.1267318225754484,
"grad_norm": 0.24474380910396576,
"learning_rate": 9.759137129170728e-05,
"loss": 1.5088220596313477,
"step": 295
},
{
"epoch": 0.12887981956825262,
"grad_norm": 0.2617799639701843,
"learning_rate": 9.748356427120226e-05,
"loss": 1.4764927864074706,
"step": 300
},
{
"epoch": 0.1310278165610568,
"grad_norm": 0.27879655361175537,
"learning_rate": 9.737345933845691e-05,
"loss": 1.5067102432250976,
"step": 305
},
{
"epoch": 0.13317581355386102,
"grad_norm": 0.2816133201122284,
"learning_rate": 9.726106182187226e-05,
"loss": 1.4711156845092774,
"step": 310
},
{
"epoch": 0.13532381054666523,
"grad_norm": 0.2718169689178467,
"learning_rate": 9.714637716079627e-05,
"loss": 1.5501375198364258,
"step": 315
},
{
"epoch": 0.13747180753946944,
"grad_norm": 0.2559225559234619,
"learning_rate": 9.702941090526065e-05,
"loss": 1.504294490814209,
"step": 320
},
{
"epoch": 0.13961980453227366,
"grad_norm": 0.24129146337509155,
"learning_rate": 9.69101687157122e-05,
"loss": 1.5170598030090332,
"step": 325
},
{
"epoch": 0.14176780152507787,
"grad_norm": 0.2434818297624588,
"learning_rate": 9.678865636273895e-05,
"loss": 1.4940474510192872,
"step": 330
},
{
"epoch": 0.14391579851788208,
"grad_norm": 0.24998946487903595,
"learning_rate": 9.666487972679084e-05,
"loss": 1.489492416381836,
"step": 335
},
{
"epoch": 0.1460637955106863,
"grad_norm": 0.2829580008983612,
"learning_rate": 9.653884479789519e-05,
"loss": 1.454762077331543,
"step": 340
},
{
"epoch": 0.1482117925034905,
"grad_norm": 0.2580244839191437,
"learning_rate": 9.641055767536679e-05,
"loss": 1.4957307815551757,
"step": 345
},
{
"epoch": 0.1503597894962947,
"grad_norm": 0.27733874320983887,
"learning_rate": 9.628002456751273e-05,
"loss": 1.4771961212158202,
"step": 350
},
{
"epoch": 0.1525077864890989,
"grad_norm": 0.281941294670105,
"learning_rate": 9.614725179133197e-05,
"loss": 1.4683259963989257,
"step": 355
},
{
"epoch": 0.15465578348190312,
"grad_norm": 0.24962927401065826,
"learning_rate": 9.601224577220966e-05,
"loss": 1.5095586776733398,
"step": 360
},
{
"epoch": 0.15680378047470733,
"grad_norm": 0.26393774151802063,
"learning_rate": 9.587501304360613e-05,
"loss": 1.549782657623291,
"step": 365
},
{
"epoch": 0.15895177746751155,
"grad_norm": 0.24991562962532043,
"learning_rate": 9.573556024674073e-05,
"loss": 1.4670140266418457,
"step": 370
},
{
"epoch": 0.16109977446031576,
"grad_norm": 0.2363777905702591,
"learning_rate": 9.559389413027048e-05,
"loss": 1.4712093353271485,
"step": 375
},
{
"epoch": 0.16324777145311997,
"grad_norm": 0.2319713532924652,
"learning_rate": 9.545002154996344e-05,
"loss": 1.5290300369262695,
"step": 380
},
{
"epoch": 0.16539576844592419,
"grad_norm": 0.25370731949806213,
"learning_rate": 9.530394946836694e-05,
"loss": 1.4795835494995118,
"step": 385
},
{
"epoch": 0.1675437654387284,
"grad_norm": 0.30085310339927673,
"learning_rate": 9.51556849544706e-05,
"loss": 1.4818733215332032,
"step": 390
},
{
"epoch": 0.16969176243153258,
"grad_norm": 0.2645235061645508,
"learning_rate": 9.500523518336434e-05,
"loss": 1.5029242515563965,
"step": 395
},
{
"epoch": 0.1718397594243368,
"grad_norm": 0.2732005715370178,
"learning_rate": 9.4852607435891e-05,
"loss": 1.4794482231140136,
"step": 400
},
{
"epoch": 0.173987756417141,
"grad_norm": 0.2590753138065338,
"learning_rate": 9.46978090982941e-05,
"loss": 1.4937355995178223,
"step": 405
},
{
"epoch": 0.17613575340994522,
"grad_norm": 0.24807235598564148,
"learning_rate": 9.45408476618604e-05,
"loss": 1.455815887451172,
"step": 410
},
{
"epoch": 0.17828375040274944,
"grad_norm": 0.27246901392936707,
"learning_rate": 9.438173072255727e-05,
"loss": 1.4796748161315918,
"step": 415
},
{
"epoch": 0.18043174739555365,
"grad_norm": 0.28435900807380676,
"learning_rate": 9.422046598066519e-05,
"loss": 1.4520401000976562,
"step": 420
},
{
"epoch": 0.18257974438835786,
"grad_norm": 0.23311451077461243,
"learning_rate": 9.405706124040505e-05,
"loss": 1.522452449798584,
"step": 425
},
{
"epoch": 0.18472774138116207,
"grad_norm": 0.260882169008255,
"learning_rate": 9.389152440956048e-05,
"loss": 1.4505027770996093,
"step": 430
},
{
"epoch": 0.1868757383739663,
"grad_norm": 0.2527851462364197,
"learning_rate": 9.37238634990952e-05,
"loss": 1.4480948448181152,
"step": 435
},
{
"epoch": 0.18902373536677047,
"grad_norm": 0.2491643875837326,
"learning_rate": 9.355408662276528e-05,
"loss": 1.4588994979858398,
"step": 440
},
{
"epoch": 0.1911717323595747,
"grad_norm": 0.267521470785141,
"learning_rate": 9.338220199672651e-05,
"loss": 1.3949003219604492,
"step": 445
},
{
"epoch": 0.1933197293523789,
"grad_norm": 0.243647962808609,
"learning_rate": 9.320821793913685e-05,
"loss": 1.4993959426879884,
"step": 450
},
{
"epoch": 0.1954677263451831,
"grad_norm": 0.2857939898967743,
"learning_rate": 9.303214286975373e-05,
"loss": 1.4854547500610351,
"step": 455
},
{
"epoch": 0.19761572333798733,
"grad_norm": 0.2699887752532959,
"learning_rate": 9.285398530952679e-05,
"loss": 1.4479030609130858,
"step": 460
},
{
"epoch": 0.19976372033079154,
"grad_norm": 0.2546408772468567,
"learning_rate": 9.267375388018529e-05,
"loss": 1.448690700531006,
"step": 465
},
{
"epoch": 0.20191171732359575,
"grad_norm": 0.2458326518535614,
"learning_rate": 9.249145730382109e-05,
"loss": 1.4833657264709472,
"step": 470
},
{
"epoch": 0.20405971431639996,
"grad_norm": 0.23938611149787903,
"learning_rate": 9.230710440246642e-05,
"loss": 1.4777969360351562,
"step": 475
},
{
"epoch": 0.20620771130920418,
"grad_norm": 0.24163755774497986,
"learning_rate": 9.212070409766696e-05,
"loss": 1.4708952903747559,
"step": 480
},
{
"epoch": 0.20835570830200836,
"grad_norm": 0.25088000297546387,
"learning_rate": 9.193226541005014e-05,
"loss": 1.457043170928955,
"step": 485
},
{
"epoch": 0.21050370529481258,
"grad_norm": 0.2366296648979187,
"learning_rate": 9.174179745888857e-05,
"loss": 1.4701215744018554,
"step": 490
},
{
"epoch": 0.2126517022876168,
"grad_norm": 0.25736457109451294,
"learning_rate": 9.15493094616587e-05,
"loss": 1.4858928680419923,
"step": 495
},
{
"epoch": 0.214799699280421,
"grad_norm": 0.2697490453720093,
"learning_rate": 9.135481073359482e-05,
"loss": 1.4081520080566405,
"step": 500
},
{
"epoch": 0.21694769627322522,
"grad_norm": 0.31537505984306335,
"learning_rate": 9.115831068723817e-05,
"loss": 1.3567978858947753,
"step": 505
},
{
"epoch": 0.21909569326602943,
"grad_norm": 0.27851346135139465,
"learning_rate": 9.095981883198153e-05,
"loss": 1.4485319137573243,
"step": 510
},
{
"epoch": 0.22124369025883364,
"grad_norm": 0.2801581919193268,
"learning_rate": 9.07593447736089e-05,
"loss": 1.3925585746765137,
"step": 515
},
{
"epoch": 0.22339168725163785,
"grad_norm": 0.28237661719322205,
"learning_rate": 9.055689821383077e-05,
"loss": 1.460464096069336,
"step": 520
},
{
"epoch": 0.22553968424444207,
"grad_norm": 0.24527870118618011,
"learning_rate": 9.035248894981453e-05,
"loss": 1.4803878784179687,
"step": 525
},
{
"epoch": 0.22768768123724628,
"grad_norm": 0.27649515867233276,
"learning_rate": 9.014612687371039e-05,
"loss": 1.3901119232177734,
"step": 530
},
{
"epoch": 0.22983567823005047,
"grad_norm": 0.2523747384548187,
"learning_rate": 8.993782197217262e-05,
"loss": 1.4215583801269531,
"step": 535
},
{
"epoch": 0.23198367522285468,
"grad_norm": 0.25827792286872864,
"learning_rate": 8.972758432587632e-05,
"loss": 1.4548846244812013,
"step": 540
},
{
"epoch": 0.2341316722156589,
"grad_norm": 0.26774686574935913,
"learning_rate": 8.951542410902948e-05,
"loss": 1.3848633766174316,
"step": 545
},
{
"epoch": 0.2362796692084631,
"grad_norm": 0.2456275224685669,
"learning_rate": 8.930135158888078e-05,
"loss": 1.404214572906494,
"step": 550
},
{
"epoch": 0.23842766620126732,
"grad_norm": 0.31159016489982605,
"learning_rate": 8.908537712522244e-05,
"loss": 1.3782641410827636,
"step": 555
},
{
"epoch": 0.24057566319407153,
"grad_norm": 0.34996986389160156,
"learning_rate": 8.886751116988923e-05,
"loss": 1.3972426414489747,
"step": 560
},
{
"epoch": 0.24272366018687574,
"grad_norm": 0.2576439678668976,
"learning_rate": 8.864776426625231e-05,
"loss": 1.422494602203369,
"step": 565
},
{
"epoch": 0.24487165717967996,
"grad_norm": 0.2605086863040924,
"learning_rate": 8.842614704870929e-05,
"loss": 1.4705518722534179,
"step": 570
},
{
"epoch": 0.24701965417248417,
"grad_norm": 0.2480541169643402,
"learning_rate": 8.820267024216938e-05,
"loss": 1.4339286804199218,
"step": 575
},
{
"epoch": 0.24916765116528836,
"grad_norm": 0.2904776632785797,
"learning_rate": 8.797734466153445e-05,
"loss": 1.4545066833496094,
"step": 580
},
{
"epoch": 0.2513156481580926,
"grad_norm": 0.2635369896888733,
"learning_rate": 8.775018121117568e-05,
"loss": 1.4260221481323243,
"step": 585
},
{
"epoch": 0.2534636451508968,
"grad_norm": 0.2347613275051117,
"learning_rate": 8.752119088440589e-05,
"loss": 1.4399876594543457,
"step": 590
},
{
"epoch": 0.255611642143701,
"grad_norm": 0.2483147829771042,
"learning_rate": 8.729038476294737e-05,
"loss": 1.4108962059020995,
"step": 595
},
{
"epoch": 0.25775963913650524,
"grad_norm": 0.23239263892173767,
"learning_rate": 8.705777401639574e-05,
"loss": 1.4686798095703124,
"step": 600
},
{
"epoch": 0.2599076361293094,
"grad_norm": 0.281330943107605,
"learning_rate": 8.68233699016794e-05,
"loss": 1.4563122749328614,
"step": 605
},
{
"epoch": 0.2620556331221136,
"grad_norm": 0.26226606965065,
"learning_rate": 8.658718376251466e-05,
"loss": 1.415459442138672,
"step": 610
},
{
"epoch": 0.2642036301149178,
"grad_norm": 0.24768346548080444,
"learning_rate": 8.634922702885693e-05,
"loss": 1.419222354888916,
"step": 615
},
{
"epoch": 0.26635162710772203,
"grad_norm": 0.25416624546051025,
"learning_rate": 8.610951121634742e-05,
"loss": 1.4490538597106934,
"step": 620
},
{
"epoch": 0.26849962410052625,
"grad_norm": 0.28159448504447937,
"learning_rate": 8.586804792575597e-05,
"loss": 1.4109480857849122,
"step": 625
},
{
"epoch": 0.27064762109333046,
"grad_norm": 0.21726375818252563,
"learning_rate": 8.562484884241959e-05,
"loss": 1.4314626693725585,
"step": 630
},
{
"epoch": 0.2727956180861347,
"grad_norm": 0.2320200353860855,
"learning_rate": 8.537992573567697e-05,
"loss": 1.374193000793457,
"step": 635
},
{
"epoch": 0.2749436150789389,
"grad_norm": 0.25649043917655945,
"learning_rate": 8.513329045829893e-05,
"loss": 1.3780905723571777,
"step": 640
},
{
"epoch": 0.2770916120717431,
"grad_norm": 0.2817293107509613,
"learning_rate": 8.488495494591482e-05,
"loss": 1.3980112075805664,
"step": 645
},
{
"epoch": 0.2792396090645473,
"grad_norm": 0.264091819524765,
"learning_rate": 8.463493121643487e-05,
"loss": 1.4563077926635741,
"step": 650
},
{
"epoch": 0.2813876060573515,
"grad_norm": 0.3143632113933563,
"learning_rate": 8.438323136946865e-05,
"loss": 1.3955091476440429,
"step": 655
},
{
"epoch": 0.28353560305015574,
"grad_norm": 0.2942863702774048,
"learning_rate": 8.412986758573949e-05,
"loss": 1.4141192436218262,
"step": 660
},
{
"epoch": 0.28568360004295995,
"grad_norm": 0.2809731662273407,
"learning_rate": 8.387485212649503e-05,
"loss": 1.3962980270385743,
"step": 665
},
{
"epoch": 0.28783159703576416,
"grad_norm": 0.24791520833969116,
"learning_rate": 8.361819733291383e-05,
"loss": 1.3947994232177734,
"step": 670
},
{
"epoch": 0.2899795940285684,
"grad_norm": 0.27428653836250305,
"learning_rate": 8.335991562550812e-05,
"loss": 1.3811721801757812,
"step": 675
},
{
"epoch": 0.2921275910213726,
"grad_norm": 0.30165061354637146,
"learning_rate": 8.31000195035228e-05,
"loss": 1.3478551864624024,
"step": 680
},
{
"epoch": 0.2942755880141768,
"grad_norm": 0.2590092420578003,
"learning_rate": 8.283852154433041e-05,
"loss": 1.4236961364746095,
"step": 685
},
{
"epoch": 0.296423585006981,
"grad_norm": 0.24623268842697144,
"learning_rate": 8.257543440282266e-05,
"loss": 1.3732666015625,
"step": 690
},
{
"epoch": 0.2985715819997852,
"grad_norm": 0.25951796770095825,
"learning_rate": 8.231077081079782e-05,
"loss": 1.3912155151367187,
"step": 695
},
{
"epoch": 0.3007195789925894,
"grad_norm": 0.266838014125824,
"learning_rate": 8.204454357634469e-05,
"loss": 1.3673189163208008,
"step": 700
},
{
"epoch": 0.3028675759853936,
"grad_norm": 0.21887816488742828,
"learning_rate": 8.177676558322274e-05,
"loss": 1.346531867980957,
"step": 705
},
{
"epoch": 0.3050155729781978,
"grad_norm": 0.25916555523872375,
"learning_rate": 8.150744979023861e-05,
"loss": 1.4149486541748046,
"step": 710
},
{
"epoch": 0.307163569971002,
"grad_norm": 0.2560618221759796,
"learning_rate": 8.1236609230619e-05,
"loss": 1.4099782943725585,
"step": 715
},
{
"epoch": 0.30931156696380624,
"grad_norm": 0.2647530734539032,
"learning_rate": 8.096425701137992e-05,
"loss": 1.3830817222595215,
"step": 720
},
{
"epoch": 0.31145956395661045,
"grad_norm": 0.2625667154788971,
"learning_rate": 8.06904063126924e-05,
"loss": 1.416067886352539,
"step": 725
},
{
"epoch": 0.31360756094941467,
"grad_norm": 0.23775933682918549,
"learning_rate": 8.041507038724463e-05,
"loss": 1.4043415069580079,
"step": 730
},
{
"epoch": 0.3157555579422189,
"grad_norm": 0.24000824987888336,
"learning_rate": 8.013826255960069e-05,
"loss": 1.4054168701171874,
"step": 735
},
{
"epoch": 0.3179035549350231,
"grad_norm": 0.31010645627975464,
"learning_rate": 7.985999622555562e-05,
"loss": 1.413990592956543,
"step": 740
},
{
"epoch": 0.3200515519278273,
"grad_norm": 0.26660218834877014,
"learning_rate": 7.95802848514872e-05,
"loss": 1.3784582138061523,
"step": 745
},
{
"epoch": 0.3221995489206315,
"grad_norm": 0.2455047070980072,
"learning_rate": 7.929914197370425e-05,
"loss": 1.411691951751709,
"step": 750
},
{
"epoch": 0.32434754591343573,
"grad_norm": 0.26444971561431885,
"learning_rate": 7.90165811977916e-05,
"loss": 1.3853992462158202,
"step": 755
},
{
"epoch": 0.32649554290623994,
"grad_norm": 0.2267448753118515,
"learning_rate": 7.873261619795156e-05,
"loss": 1.363304042816162,
"step": 760
},
{
"epoch": 0.32864353989904416,
"grad_norm": 0.2788371443748474,
"learning_rate": 7.844726071634228e-05,
"loss": 1.3655694007873536,
"step": 765
},
{
"epoch": 0.33079153689184837,
"grad_norm": 0.24006222188472748,
"learning_rate": 7.816052856241261e-05,
"loss": 1.3904157638549806,
"step": 770
},
{
"epoch": 0.3329395338846526,
"grad_norm": 0.22869496047496796,
"learning_rate": 7.787243361223397e-05,
"loss": 1.3929327011108399,
"step": 775
},
{
"epoch": 0.3350875308774568,
"grad_norm": 0.25210389494895935,
"learning_rate": 7.758298980782863e-05,
"loss": 1.4066057205200195,
"step": 780
},
{
"epoch": 0.337235527870261,
"grad_norm": 0.2500297725200653,
"learning_rate": 7.729221115649517e-05,
"loss": 1.3727193832397462,
"step": 785
},
{
"epoch": 0.33938352486306517,
"grad_norm": 0.258533775806427,
"learning_rate": 7.700011173013049e-05,
"loss": 1.3815672874450684,
"step": 790
},
{
"epoch": 0.3415315218558694,
"grad_norm": 0.25695502758026123,
"learning_rate": 7.670670566454891e-05,
"loss": 1.4345569610595703,
"step": 795
},
{
"epoch": 0.3436795188486736,
"grad_norm": 0.2510787546634674,
"learning_rate": 7.641200715879804e-05,
"loss": 1.33048677444458,
"step": 800
},
{
"epoch": 0.3458275158414778,
"grad_norm": 0.24779732525348663,
"learning_rate": 7.611603047447161e-05,
"loss": 1.3819402694702148,
"step": 805
},
{
"epoch": 0.347975512834282,
"grad_norm": 0.35685786604881287,
"learning_rate": 7.581878993501937e-05,
"loss": 1.492457389831543,
"step": 810
},
{
"epoch": 0.35012350982708623,
"grad_norm": 0.26960375905036926,
"learning_rate": 7.552029992505385e-05,
"loss": 1.4139192581176758,
"step": 815
},
{
"epoch": 0.35227150681989045,
"grad_norm": 0.2798788845539093,
"learning_rate": 7.522057488965427e-05,
"loss": 1.373062515258789,
"step": 820
},
{
"epoch": 0.35441950381269466,
"grad_norm": 0.2582121193408966,
"learning_rate": 7.491962933366747e-05,
"loss": 1.3300559997558594,
"step": 825
},
{
"epoch": 0.35656750080549887,
"grad_norm": 0.2511405646800995,
"learning_rate": 7.461747782100602e-05,
"loss": 1.351390552520752,
"step": 830
},
{
"epoch": 0.3587154977983031,
"grad_norm": 0.2690654397010803,
"learning_rate": 7.431413497394328e-05,
"loss": 1.3780029296875,
"step": 835
},
{
"epoch": 0.3608634947911073,
"grad_norm": 0.28491783142089844,
"learning_rate": 7.400961547240595e-05,
"loss": 1.379025650024414,
"step": 840
},
{
"epoch": 0.3630114917839115,
"grad_norm": 0.28490254282951355,
"learning_rate": 7.370393405326351e-05,
"loss": 1.3628406524658203,
"step": 845
},
{
"epoch": 0.3651594887767157,
"grad_norm": 0.29991158843040466,
"learning_rate": 7.339710550961513e-05,
"loss": 1.3672489166259765,
"step": 850
},
{
"epoch": 0.36730748576951994,
"grad_norm": 0.2680410146713257,
"learning_rate": 7.308914469007372e-05,
"loss": 1.3925317764282226,
"step": 855
},
{
"epoch": 0.36945548276232415,
"grad_norm": 0.2888924181461334,
"learning_rate": 7.27800664980474e-05,
"loss": 1.3208433151245118,
"step": 860
},
{
"epoch": 0.37160347975512836,
"grad_norm": 0.2586013674736023,
"learning_rate": 7.246988589101824e-05,
"loss": 1.3503068923950194,
"step": 865
},
{
"epoch": 0.3737514767479326,
"grad_norm": 0.2911142110824585,
"learning_rate": 7.215861787981839e-05,
"loss": 1.3667444229125976,
"step": 870
},
{
"epoch": 0.3758994737407368,
"grad_norm": 0.2524161636829376,
"learning_rate": 7.184627752790367e-05,
"loss": 1.3751562118530274,
"step": 875
},
{
"epoch": 0.37804747073354095,
"grad_norm": 0.2730039656162262,
"learning_rate": 7.153287995062459e-05,
"loss": 1.3180736541748046,
"step": 880
},
{
"epoch": 0.38019546772634516,
"grad_norm": 0.20326928794384003,
"learning_rate": 7.12184403144949e-05,
"loss": 1.367253017425537,
"step": 885
},
{
"epoch": 0.3823434647191494,
"grad_norm": 0.2840406000614166,
"learning_rate": 7.090297383645753e-05,
"loss": 1.3786996841430663,
"step": 890
},
{
"epoch": 0.3844914617119536,
"grad_norm": 0.29089492559432983,
"learning_rate": 7.058649578314828e-05,
"loss": 1.3182512283325196,
"step": 895
},
{
"epoch": 0.3866394587047578,
"grad_norm": 0.3280753195285797,
"learning_rate": 7.02690214701569e-05,
"loss": 1.337693691253662,
"step": 900
},
{
"epoch": 0.388787455697562,
"grad_norm": 0.26715826988220215,
"learning_rate": 6.995056626128609e-05,
"loss": 1.380331802368164,
"step": 905
},
{
"epoch": 0.3909354526903662,
"grad_norm": 0.276676744222641,
"learning_rate": 6.963114556780775e-05,
"loss": 1.3760421752929688,
"step": 910
},
{
"epoch": 0.39308344968317044,
"grad_norm": 0.3139406740665436,
"learning_rate": 6.931077484771739e-05,
"loss": 1.3976767539978028,
"step": 915
},
{
"epoch": 0.39523144667597465,
"grad_norm": 0.32448112964630127,
"learning_rate": 6.898946960498589e-05,
"loss": 1.3961191177368164,
"step": 920
},
{
"epoch": 0.39737944366877886,
"grad_norm": 0.2545507252216339,
"learning_rate": 6.866724538880931e-05,
"loss": 1.3693606376647949,
"step": 925
},
{
"epoch": 0.3995274406615831,
"grad_norm": 0.2889951467514038,
"learning_rate": 6.834411779285638e-05,
"loss": 1.3285128593444824,
"step": 930
},
{
"epoch": 0.4016754376543873,
"grad_norm": 0.24984769523143768,
"learning_rate": 6.802010245451383e-05,
"loss": 1.3431509017944336,
"step": 935
},
{
"epoch": 0.4038234346471915,
"grad_norm": 0.2542552947998047,
"learning_rate": 6.769521505412964e-05,
"loss": 1.3805252075195313,
"step": 940
},
{
"epoch": 0.4059714316399957,
"grad_norm": 0.2676757276058197,
"learning_rate": 6.736947131425422e-05,
"loss": 1.3348833084106446,
"step": 945
},
{
"epoch": 0.40811942863279993,
"grad_norm": 0.27421504259109497,
"learning_rate": 6.704288699887965e-05,
"loss": 1.3977334022521972,
"step": 950
},
{
"epoch": 0.41026742562560414,
"grad_norm": 0.2610110938549042,
"learning_rate": 6.671547791267652e-05,
"loss": 1.321683692932129,
"step": 955
},
{
"epoch": 0.41241542261840836,
"grad_norm": 0.279585063457489,
"learning_rate": 6.638725990022938e-05,
"loss": 1.3916077613830566,
"step": 960
},
{
"epoch": 0.41456341961121257,
"grad_norm": 0.2708588242530823,
"learning_rate": 6.605824884526977e-05,
"loss": 1.3744972229003907,
"step": 965
},
{
"epoch": 0.4167114166040167,
"grad_norm": 0.2506023943424225,
"learning_rate": 6.572846066990765e-05,
"loss": 1.4075494766235352,
"step": 970
},
{
"epoch": 0.41885941359682094,
"grad_norm": 0.3166026771068573,
"learning_rate": 6.539791133386078e-05,
"loss": 1.3285415649414063,
"step": 975
},
{
"epoch": 0.42100741058962515,
"grad_norm": 0.26006966829299927,
"learning_rate": 6.506661683368242e-05,
"loss": 1.3573419570922851,
"step": 980
},
{
"epoch": 0.42315540758242937,
"grad_norm": 0.21595826745033264,
"learning_rate": 6.47345932019872e-05,
"loss": 1.2939266204833983,
"step": 985
},
{
"epoch": 0.4253034045752336,
"grad_norm": 0.26269346475601196,
"learning_rate": 6.440185650667526e-05,
"loss": 1.3140311241149902,
"step": 990
},
{
"epoch": 0.4274514015680378,
"grad_norm": 0.2812257707118988,
"learning_rate": 6.406842285015455e-05,
"loss": 1.3527883529663085,
"step": 995
},
{
"epoch": 0.429599398560842,
"grad_norm": 0.2505556344985962,
"learning_rate": 6.373430836856173e-05,
"loss": 1.3523382186889648,
"step": 1000
},
{
"epoch": 0.4317473955536462,
"grad_norm": 0.28471699357032776,
"learning_rate": 6.339952923098116e-05,
"loss": 1.32545166015625,
"step": 1005
},
{
"epoch": 0.43389539254645043,
"grad_norm": 0.2407894879579544,
"learning_rate": 6.306410163866247e-05,
"loss": 1.3143939971923828,
"step": 1010
},
{
"epoch": 0.43604338953925464,
"grad_norm": 0.2561395764350891,
"learning_rate": 6.27280418242365e-05,
"loss": 1.3232194900512695,
"step": 1015
},
{
"epoch": 0.43819138653205886,
"grad_norm": 0.23729200661182404,
"learning_rate": 6.239136605092975e-05,
"loss": 1.2960392951965332,
"step": 1020
},
{
"epoch": 0.44033938352486307,
"grad_norm": 0.3065226674079895,
"learning_rate": 6.205409061177738e-05,
"loss": 1.3518625259399415,
"step": 1025
},
{
"epoch": 0.4424873805176673,
"grad_norm": 0.2600172460079193,
"learning_rate": 6.171623182883462e-05,
"loss": 1.3439982414245606,
"step": 1030
},
{
"epoch": 0.4446353775104715,
"grad_norm": 0.2614782452583313,
"learning_rate": 6.137780605238698e-05,
"loss": 1.3319015502929688,
"step": 1035
},
{
"epoch": 0.4467833745032757,
"grad_norm": 0.25418734550476074,
"learning_rate": 6.103882966015897e-05,
"loss": 1.3518550872802735,
"step": 1040
},
{
"epoch": 0.4489313714960799,
"grad_norm": 0.3305002450942993,
"learning_rate": 6.069931905652151e-05,
"loss": 1.2984891891479493,
"step": 1045
},
{
"epoch": 0.45107936848888414,
"grad_norm": 0.29671943187713623,
"learning_rate": 6.035929067169805e-05,
"loss": 1.3011366844177246,
"step": 1050
},
{
"epoch": 0.45322736548168835,
"grad_norm": 0.30703550577163696,
"learning_rate": 6.001876096096951e-05,
"loss": 1.3626813888549805,
"step": 1055
},
{
"epoch": 0.45537536247449256,
"grad_norm": 0.2879122495651245,
"learning_rate": 5.967774640387783e-05,
"loss": 1.3248126983642579,
"step": 1060
},
{
"epoch": 0.4575233594672967,
"grad_norm": 0.27624985575675964,
"learning_rate": 5.933626350342858e-05,
"loss": 1.3089393615722655,
"step": 1065
},
{
"epoch": 0.45967135646010093,
"grad_norm": 0.2764015793800354,
"learning_rate": 5.8994328785292264e-05,
"loss": 1.379379653930664,
"step": 1070
},
{
"epoch": 0.46181935345290515,
"grad_norm": 0.26699161529541016,
"learning_rate": 5.865195879700454e-05,
"loss": 1.355512809753418,
"step": 1075
},
{
"epoch": 0.46396735044570936,
"grad_norm": 0.23895306885242462,
"learning_rate": 5.830917010716549e-05,
"loss": 1.335743522644043,
"step": 1080
},
{
"epoch": 0.46611534743851357,
"grad_norm": 0.2280833125114441,
"learning_rate": 5.796597930463776e-05,
"loss": 1.3431385993957519,
"step": 1085
},
{
"epoch": 0.4682633444313178,
"grad_norm": 0.28493252396583557,
"learning_rate": 5.7622402997743844e-05,
"loss": 1.326830005645752,
"step": 1090
},
{
"epoch": 0.470411341424122,
"grad_norm": 0.2926257848739624,
"learning_rate": 5.727845781346216e-05,
"loss": 1.3231845855712892,
"step": 1095
},
{
"epoch": 0.4725593384169262,
"grad_norm": 0.2646852433681488,
"learning_rate": 5.6934160396622613e-05,
"loss": 1.3631766319274903,
"step": 1100
},
{
"epoch": 0.4747073354097304,
"grad_norm": 0.23445414006710052,
"learning_rate": 5.658952740910094e-05,
"loss": 1.3835596084594726,
"step": 1105
},
{
"epoch": 0.47685533240253464,
"grad_norm": 0.3288275897502899,
"learning_rate": 5.624457552901248e-05,
"loss": 1.3219709396362305,
"step": 1110
},
{
"epoch": 0.47900332939533885,
"grad_norm": 0.2447737753391266,
"learning_rate": 5.589932144990495e-05,
"loss": 1.2648405075073241,
"step": 1115
},
{
"epoch": 0.48115132638814306,
"grad_norm": 0.269762247800827,
"learning_rate": 5.555378187995069e-05,
"loss": 1.3840081214904785,
"step": 1120
},
{
"epoch": 0.4832993233809473,
"grad_norm": 0.2805151343345642,
"learning_rate": 5.5207973541138034e-05,
"loss": 1.4001265525817872,
"step": 1125
},
{
"epoch": 0.4854473203737515,
"grad_norm": 0.29017549753189087,
"learning_rate": 5.486191316846207e-05,
"loss": 1.308724594116211,
"step": 1130
},
{
"epoch": 0.4875953173665557,
"grad_norm": 0.25178658962249756,
"learning_rate": 5.451561750911475e-05,
"loss": 1.3248611450195313,
"step": 1135
},
{
"epoch": 0.4897433143593599,
"grad_norm": 0.26356133818626404,
"learning_rate": 5.416910332167446e-05,
"loss": 1.2571813583374023,
"step": 1140
},
{
"epoch": 0.49189131135216413,
"grad_norm": 0.26561856269836426,
"learning_rate": 5.382238737529505e-05,
"loss": 1.3139129638671876,
"step": 1145
},
{
"epoch": 0.49403930834496834,
"grad_norm": 0.3105677366256714,
"learning_rate": 5.3475486448894185e-05,
"loss": 1.3246249198913573,
"step": 1150
},
{
"epoch": 0.4961873053377725,
"grad_norm": 0.27671149373054504,
"learning_rate": 5.312841733034147e-05,
"loss": 1.364235782623291,
"step": 1155
},
{
"epoch": 0.4983353023305767,
"grad_norm": 0.26342299580574036,
"learning_rate": 5.278119681564597e-05,
"loss": 1.2932592391967774,
"step": 1160
},
{
"epoch": 0.500483299323381,
"grad_norm": 0.25300517678260803,
"learning_rate": 5.243384170814341e-05,
"loss": 1.317176628112793,
"step": 1165
},
{
"epoch": 0.5026312963161852,
"grad_norm": 0.3036137521266937,
"learning_rate": 5.208636881768292e-05,
"loss": 1.3103975296020507,
"step": 1170
},
{
"epoch": 0.5047792933089894,
"grad_norm": 0.27075737714767456,
"learning_rate": 5.173879495981367e-05,
"loss": 1.2820294380187989,
"step": 1175
},
{
"epoch": 0.5069272903017936,
"grad_norm": 0.2537965178489685,
"learning_rate": 5.1391136954970954e-05,
"loss": 1.3398639678955078,
"step": 1180
},
{
"epoch": 0.5090752872945978,
"grad_norm": 0.31024762988090515,
"learning_rate": 5.104341162766234e-05,
"loss": 1.3471439361572266,
"step": 1185
},
{
"epoch": 0.511223284287402,
"grad_norm": 0.31161966919898987,
"learning_rate": 5.069563580565334e-05,
"loss": 1.3478769302368163,
"step": 1190
},
{
"epoch": 0.5133712812802063,
"grad_norm": 0.24639959633350372,
"learning_rate": 5.0347826319153135e-05,
"loss": 1.299079704284668,
"step": 1195
},
{
"epoch": 0.5155192782730105,
"grad_norm": 0.29902616143226624,
"learning_rate": 5e-05,
"loss": 1.3132919311523437,
"step": 1200
},
{
"epoch": 0.5176672752658146,
"grad_norm": 0.27302250266075134,
"learning_rate": 4.965217368084688e-05,
"loss": 1.3421066284179688,
"step": 1205
},
{
"epoch": 0.5198152722586188,
"grad_norm": 0.27624455094337463,
"learning_rate": 4.930436419434666e-05,
"loss": 1.3211750030517577,
"step": 1210
},
{
"epoch": 0.521963269251423,
"grad_norm": 0.3103443384170532,
"learning_rate": 4.895658837233767e-05,
"loss": 1.3382830619812012,
"step": 1215
},
{
"epoch": 0.5241112662442272,
"grad_norm": 0.22491198778152466,
"learning_rate": 4.860886304502906e-05,
"loss": 1.3661996841430664,
"step": 1220
},
{
"epoch": 0.5262592632370314,
"grad_norm": 0.27289167046546936,
"learning_rate": 4.8261205040186345e-05,
"loss": 1.3404071807861329,
"step": 1225
},
{
"epoch": 0.5284072602298356,
"grad_norm": 0.27364280819892883,
"learning_rate": 4.791363118231708e-05,
"loss": 1.350501251220703,
"step": 1230
},
{
"epoch": 0.5305552572226399,
"grad_norm": 0.2737840414047241,
"learning_rate": 4.756615829185661e-05,
"loss": 1.289029598236084,
"step": 1235
},
{
"epoch": 0.5327032542154441,
"grad_norm": 0.303817480802536,
"learning_rate": 4.7218803184354035e-05,
"loss": 1.2715542793273926,
"step": 1240
},
{
"epoch": 0.5348512512082483,
"grad_norm": 0.3596574664115906,
"learning_rate": 4.687158266965854e-05,
"loss": 1.3193989753723145,
"step": 1245
},
{
"epoch": 0.5369992482010525,
"grad_norm": 0.24622918665409088,
"learning_rate": 4.652451355110583e-05,
"loss": 1.3230259895324707,
"step": 1250
},
{
"epoch": 0.5391472451938567,
"grad_norm": 0.2619636058807373,
"learning_rate": 4.6177612624704975e-05,
"loss": 1.3224468231201172,
"step": 1255
},
{
"epoch": 0.5412952421866609,
"grad_norm": 0.289656400680542,
"learning_rate": 4.5830896678325544e-05,
"loss": 1.2713048934936524,
"step": 1260
},
{
"epoch": 0.5434432391794651,
"grad_norm": 0.24629399180412292,
"learning_rate": 4.5484382490885265e-05,
"loss": 1.3336850166320802,
"step": 1265
},
{
"epoch": 0.5455912361722693,
"grad_norm": 0.2697427570819855,
"learning_rate": 4.513808683153793e-05,
"loss": 1.2737080574035644,
"step": 1270
},
{
"epoch": 0.5477392331650736,
"grad_norm": 0.3097023069858551,
"learning_rate": 4.479202645886196e-05,
"loss": 1.3050345420837401,
"step": 1275
},
{
"epoch": 0.5498872301578778,
"grad_norm": 0.27639704942703247,
"learning_rate": 4.444621812004931e-05,
"loss": 1.2819522857666015,
"step": 1280
},
{
"epoch": 0.552035227150682,
"grad_norm": 0.25524887442588806,
"learning_rate": 4.410067855009506e-05,
"loss": 1.3360523223876952,
"step": 1285
},
{
"epoch": 0.5541832241434862,
"grad_norm": 0.25183290243148804,
"learning_rate": 4.375542447098753e-05,
"loss": 1.316969108581543,
"step": 1290
},
{
"epoch": 0.5563312211362904,
"grad_norm": 0.22324296832084656,
"learning_rate": 4.341047259089905e-05,
"loss": 1.3311479568481446,
"step": 1295
},
{
"epoch": 0.5584792181290946,
"grad_norm": 0.2849888801574707,
"learning_rate": 4.30658396033774e-05,
"loss": 1.3290605545043945,
"step": 1300
},
{
"epoch": 0.5606272151218988,
"grad_norm": 0.2956770062446594,
"learning_rate": 4.272154218653785e-05,
"loss": 1.2997495651245117,
"step": 1305
},
{
"epoch": 0.562775212114703,
"grad_norm": 0.3384418189525604,
"learning_rate": 4.237759700225616e-05,
"loss": 1.3090848922729492,
"step": 1310
},
{
"epoch": 0.5649232091075073,
"grad_norm": 0.2551332116127014,
"learning_rate": 4.203402069536223e-05,
"loss": 1.328516960144043,
"step": 1315
},
{
"epoch": 0.5670712061003115,
"grad_norm": 0.32826900482177734,
"learning_rate": 4.169082989283453e-05,
"loss": 1.3074729919433594,
"step": 1320
},
{
"epoch": 0.5692192030931157,
"grad_norm": 0.3241807520389557,
"learning_rate": 4.134804120299548e-05,
"loss": 1.34509220123291,
"step": 1325
},
{
"epoch": 0.5713672000859199,
"grad_norm": 0.25115394592285156,
"learning_rate": 4.100567121470775e-05,
"loss": 1.3247136116027831,
"step": 1330
},
{
"epoch": 0.5735151970787241,
"grad_norm": 0.2897872030735016,
"learning_rate": 4.066373649657141e-05,
"loss": 1.310714340209961,
"step": 1335
},
{
"epoch": 0.5756631940715283,
"grad_norm": 0.28196126222610474,
"learning_rate": 4.032225359612216e-05,
"loss": 1.311944580078125,
"step": 1340
},
{
"epoch": 0.5778111910643325,
"grad_norm": 0.26805579662323,
"learning_rate": 3.998123903903051e-05,
"loss": 1.3038825035095214,
"step": 1345
},
{
"epoch": 0.5799591880571368,
"grad_norm": 0.29111260175704956,
"learning_rate": 3.964070932830196e-05,
"loss": 1.2750447273254395,
"step": 1350
},
{
"epoch": 0.582107185049941,
"grad_norm": 0.2827833294868469,
"learning_rate": 3.93006809434785e-05,
"loss": 1.3284080505371094,
"step": 1355
},
{
"epoch": 0.5842551820427452,
"grad_norm": 0.2655515968799591,
"learning_rate": 3.896117033984104e-05,
"loss": 1.3065573692321777,
"step": 1360
},
{
"epoch": 0.5864031790355494,
"grad_norm": 0.24827344715595245,
"learning_rate": 3.8622193947613046e-05,
"loss": 1.3143893241882325,
"step": 1365
},
{
"epoch": 0.5885511760283536,
"grad_norm": 0.2720562815666199,
"learning_rate": 3.82837681711654e-05,
"loss": 1.2964927673339843,
"step": 1370
},
{
"epoch": 0.5906991730211578,
"grad_norm": 0.35056138038635254,
"learning_rate": 3.7945909388222633e-05,
"loss": 1.3251249313354492,
"step": 1375
},
{
"epoch": 0.592847170013962,
"grad_norm": 0.2703423798084259,
"learning_rate": 3.760863394907025e-05,
"loss": 1.2811177253723145,
"step": 1380
},
{
"epoch": 0.5949951670067662,
"grad_norm": 0.2835351228713989,
"learning_rate": 3.7271958175763514e-05,
"loss": 1.2531731605529786,
"step": 1385
},
{
"epoch": 0.5971431639995703,
"grad_norm": 0.3012356460094452,
"learning_rate": 3.693589836133754e-05,
"loss": 1.2826812744140625,
"step": 1390
},
{
"epoch": 0.5992911609923746,
"grad_norm": 0.2844890356063843,
"learning_rate": 3.660047076901885e-05,
"loss": 1.3259981155395508,
"step": 1395
},
{
"epoch": 0.6014391579851788,
"grad_norm": 0.24462057650089264,
"learning_rate": 3.626569163143827e-05,
"loss": 1.304189109802246,
"step": 1400
},
{
"epoch": 0.603587154977983,
"grad_norm": 0.2675747275352478,
"learning_rate": 3.5931577149845465e-05,
"loss": 1.330357551574707,
"step": 1405
},
{
"epoch": 0.6057351519707872,
"grad_norm": 0.28438204526901245,
"learning_rate": 3.559814349332476e-05,
"loss": 1.221456241607666,
"step": 1410
},
{
"epoch": 0.6078831489635914,
"grad_norm": 0.3155391216278076,
"learning_rate": 3.52654067980128e-05,
"loss": 1.3078535079956055,
"step": 1415
},
{
"epoch": 0.6100311459563956,
"grad_norm": 0.3022688329219818,
"learning_rate": 3.4933383166317596e-05,
"loss": 1.236587905883789,
"step": 1420
},
{
"epoch": 0.6121791429491998,
"grad_norm": 0.3147773742675781,
"learning_rate": 3.460208866613923e-05,
"loss": 1.268033504486084,
"step": 1425
},
{
"epoch": 0.614327139942004,
"grad_norm": 0.25222596526145935,
"learning_rate": 3.427153933009237e-05,
"loss": 1.3179887771606444,
"step": 1430
},
{
"epoch": 0.6164751369348083,
"grad_norm": 0.2633180320262909,
"learning_rate": 3.394175115473024e-05,
"loss": 1.2902313232421876,
"step": 1435
},
{
"epoch": 0.6186231339276125,
"grad_norm": 0.2732374370098114,
"learning_rate": 3.3612740099770636e-05,
"loss": 1.3039766311645509,
"step": 1440
},
{
"epoch": 0.6207711309204167,
"grad_norm": 0.229779452085495,
"learning_rate": 3.328452208732349e-05,
"loss": 1.3249471664428711,
"step": 1445
},
{
"epoch": 0.6229191279132209,
"grad_norm": 0.32926133275032043,
"learning_rate": 3.295711300112038e-05,
"loss": 1.2507265090942383,
"step": 1450
},
{
"epoch": 0.6250671249060251,
"grad_norm": 0.2564751207828522,
"learning_rate": 3.2630528685745774e-05,
"loss": 1.2961599349975585,
"step": 1455
},
{
"epoch": 0.6272151218988293,
"grad_norm": 0.26443812251091003,
"learning_rate": 3.230478494587038e-05,
"loss": 1.333937931060791,
"step": 1460
},
{
"epoch": 0.6293631188916335,
"grad_norm": 0.2585863769054413,
"learning_rate": 3.197989754548618e-05,
"loss": 1.3062692642211915,
"step": 1465
},
{
"epoch": 0.6315111158844378,
"grad_norm": 0.2903755307197571,
"learning_rate": 3.1655882207143636e-05,
"loss": 1.3401302337646483,
"step": 1470
},
{
"epoch": 0.633659112877242,
"grad_norm": 0.26031243801116943,
"learning_rate": 3.13327546111907e-05,
"loss": 1.3259515762329102,
"step": 1475
},
{
"epoch": 0.6358071098700462,
"grad_norm": 0.2582283914089203,
"learning_rate": 3.1010530395014126e-05,
"loss": 1.2527307510375976,
"step": 1480
},
{
"epoch": 0.6379551068628504,
"grad_norm": 0.27913805842399597,
"learning_rate": 3.0689225152282625e-05,
"loss": 1.352574348449707,
"step": 1485
},
{
"epoch": 0.6401031038556546,
"grad_norm": 0.29610851407051086,
"learning_rate": 3.036885443219225e-05,
"loss": 1.312398338317871,
"step": 1490
},
{
"epoch": 0.6422511008484588,
"grad_norm": 0.24300889670848846,
"learning_rate": 3.0049433738713928e-05,
"loss": 1.3338730812072754,
"step": 1495
},
{
"epoch": 0.644399097841263,
"grad_norm": 0.27786949276924133,
"learning_rate": 2.97309785298431e-05,
"loss": 1.3038150787353515,
"step": 1500
},
{
"epoch": 0.6465470948340672,
"grad_norm": 0.29308438301086426,
"learning_rate": 2.941350421685174e-05,
"loss": 1.3312264442443849,
"step": 1505
},
{
"epoch": 0.6486950918268715,
"grad_norm": 0.2860047519207001,
"learning_rate": 2.9097026163542462e-05,
"loss": 1.2409782409667969,
"step": 1510
},
{
"epoch": 0.6508430888196757,
"grad_norm": 0.2621942162513733,
"learning_rate": 2.8781559685505106e-05,
"loss": 1.3369476318359375,
"step": 1515
},
{
"epoch": 0.6529910858124799,
"grad_norm": 0.2630119025707245,
"learning_rate": 2.8467120049375406e-05,
"loss": 1.2919260025024415,
"step": 1520
},
{
"epoch": 0.6551390828052841,
"grad_norm": 0.3010934591293335,
"learning_rate": 2.8153722472096332e-05,
"loss": 1.2760833740234374,
"step": 1525
},
{
"epoch": 0.6572870797980883,
"grad_norm": 0.25142380595207214,
"learning_rate": 2.7841382120181613e-05,
"loss": 1.3006477355957031,
"step": 1530
},
{
"epoch": 0.6594350767908925,
"grad_norm": 0.3040289580821991,
"learning_rate": 2.7530114108981774e-05,
"loss": 1.262589168548584,
"step": 1535
},
{
"epoch": 0.6615830737836967,
"grad_norm": 0.303268164396286,
"learning_rate": 2.7219933501952606e-05,
"loss": 1.2814247131347656,
"step": 1540
},
{
"epoch": 0.663731070776501,
"grad_norm": 0.29907235503196716,
"learning_rate": 2.6910855309926293e-05,
"loss": 1.2651835441589356,
"step": 1545
},
{
"epoch": 0.6658790677693052,
"grad_norm": 0.24670934677124023,
"learning_rate": 2.660289449038489e-05,
"loss": 1.2371265411376953,
"step": 1550
},
{
"epoch": 0.6680270647621094,
"grad_norm": 0.3393242061138153,
"learning_rate": 2.6296065946736502e-05,
"loss": 1.2900367736816407,
"step": 1555
},
{
"epoch": 0.6701750617549136,
"grad_norm": 0.33548596501350403,
"learning_rate": 2.599038452759408e-05,
"loss": 1.3156991958618165,
"step": 1560
},
{
"epoch": 0.6723230587477178,
"grad_norm": 0.24356567859649658,
"learning_rate": 2.5685865026056744e-05,
"loss": 1.27972412109375,
"step": 1565
},
{
"epoch": 0.674471055740522,
"grad_norm": 0.25120630860328674,
"learning_rate": 2.5382522178994e-05,
"loss": 1.3083066940307617,
"step": 1570
},
{
"epoch": 0.6766190527333261,
"grad_norm": 0.27809250354766846,
"learning_rate": 2.508037066633253e-05,
"loss": 1.3442459106445312,
"step": 1575
},
{
"epoch": 0.6787670497261303,
"grad_norm": 0.2727198004722595,
"learning_rate": 2.477942511034574e-05,
"loss": 1.326646327972412,
"step": 1580
},
{
"epoch": 0.6809150467189345,
"grad_norm": 0.2595716118812561,
"learning_rate": 2.4479700074946155e-05,
"loss": 1.3192209243774413,
"step": 1585
},
{
"epoch": 0.6830630437117388,
"grad_norm": 0.27739694714546204,
"learning_rate": 2.4181210064980636e-05,
"loss": 1.258584213256836,
"step": 1590
},
{
"epoch": 0.685211040704543,
"grad_norm": 0.2736615836620331,
"learning_rate": 2.3883969525528394e-05,
"loss": 1.2861428260803223,
"step": 1595
},
{
"epoch": 0.6873590376973472,
"grad_norm": 0.28235530853271484,
"learning_rate": 2.3587992841201983e-05,
"loss": 1.3235095977783202,
"step": 1600
},
{
"epoch": 0.6895070346901514,
"grad_norm": 0.2958912253379822,
"learning_rate": 2.3293294335451105e-05,
"loss": 1.266238307952881,
"step": 1605
},
{
"epoch": 0.6916550316829556,
"grad_norm": 0.28194791078567505,
"learning_rate": 2.299988826986952e-05,
"loss": 1.2568707466125488,
"step": 1610
},
{
"epoch": 0.6938030286757598,
"grad_norm": 0.2527979910373688,
"learning_rate": 2.2707788843504835e-05,
"loss": 1.2381067276000977,
"step": 1615
},
{
"epoch": 0.695951025668564,
"grad_norm": 0.2859898507595062,
"learning_rate": 2.241701019217137e-05,
"loss": 1.3089589118957519,
"step": 1620
},
{
"epoch": 0.6980990226613683,
"grad_norm": 0.23987330496311188,
"learning_rate": 2.2127566387766047e-05,
"loss": 1.3086300849914552,
"step": 1625
},
{
"epoch": 0.7002470196541725,
"grad_norm": 0.22601911425590515,
"learning_rate": 2.18394714375874e-05,
"loss": 1.2787461280822754,
"step": 1630
},
{
"epoch": 0.7023950166469767,
"grad_norm": 0.27868494391441345,
"learning_rate": 2.1552739283657752e-05,
"loss": 1.2973819732666017,
"step": 1635
},
{
"epoch": 0.7045430136397809,
"grad_norm": 0.2973822057247162,
"learning_rate": 2.1267383802048456e-05,
"loss": 1.2784087181091308,
"step": 1640
},
{
"epoch": 0.7066910106325851,
"grad_norm": 0.29257938265800476,
"learning_rate": 2.0983418802208417e-05,
"loss": 1.2020733833312989,
"step": 1645
},
{
"epoch": 0.7088390076253893,
"grad_norm": 0.274156391620636,
"learning_rate": 2.070085802629575e-05,
"loss": 1.2446213722229005,
"step": 1650
},
{
"epoch": 0.7109870046181935,
"grad_norm": 0.2804328203201294,
"learning_rate": 2.041971514851281e-05,
"loss": 1.2708898544311524,
"step": 1655
},
{
"epoch": 0.7131350016109977,
"grad_norm": 0.3227699398994446,
"learning_rate": 2.0140003774444387e-05,
"loss": 1.2751914978027343,
"step": 1660
},
{
"epoch": 0.715282998603802,
"grad_norm": 0.3017464876174927,
"learning_rate": 1.9861737440399325e-05,
"loss": 1.2952705383300782,
"step": 1665
},
{
"epoch": 0.7174309955966062,
"grad_norm": 0.28101837635040283,
"learning_rate": 1.9584929612755383e-05,
"loss": 1.2265684127807617,
"step": 1670
},
{
"epoch": 0.7195789925894104,
"grad_norm": 0.28632572293281555,
"learning_rate": 1.930959368730762e-05,
"loss": 1.2267627716064453,
"step": 1675
},
{
"epoch": 0.7217269895822146,
"grad_norm": 0.2715645730495453,
"learning_rate": 1.9035742988620088e-05,
"loss": 1.3036357879638671,
"step": 1680
},
{
"epoch": 0.7238749865750188,
"grad_norm": 0.3325082063674927,
"learning_rate": 1.876339076938102e-05,
"loss": 1.2487954139709472,
"step": 1685
},
{
"epoch": 0.726022983567823,
"grad_norm": 0.2875954508781433,
"learning_rate": 1.8492550209761407e-05,
"loss": 1.2789426803588868,
"step": 1690
},
{
"epoch": 0.7281709805606272,
"grad_norm": 0.3244427740573883,
"learning_rate": 1.8223234416777274e-05,
"loss": 1.257153034210205,
"step": 1695
},
{
"epoch": 0.7303189775534314,
"grad_norm": 0.2947673797607422,
"learning_rate": 1.7955456423655327e-05,
"loss": 1.3045918464660644,
"step": 1700
},
{
"epoch": 0.7324669745462357,
"grad_norm": 0.3650137782096863,
"learning_rate": 1.7689229189202196e-05,
"loss": 1.273738193511963,
"step": 1705
},
{
"epoch": 0.7346149715390399,
"grad_norm": 0.29794642329216003,
"learning_rate": 1.7424565597177344e-05,
"loss": 1.2713969230651856,
"step": 1710
},
{
"epoch": 0.7367629685318441,
"grad_norm": 0.3232552707195282,
"learning_rate": 1.7161478455669588e-05,
"loss": 1.2479372024536133,
"step": 1715
},
{
"epoch": 0.7389109655246483,
"grad_norm": 0.2830597460269928,
"learning_rate": 1.6899980496477208e-05,
"loss": 1.2470852851867675,
"step": 1720
},
{
"epoch": 0.7410589625174525,
"grad_norm": 0.30554208159446716,
"learning_rate": 1.664008437449187e-05,
"loss": 1.2826133728027345,
"step": 1725
},
{
"epoch": 0.7432069595102567,
"grad_norm": 0.3143727779388428,
"learning_rate": 1.6381802667086183e-05,
"loss": 1.3243678092956543,
"step": 1730
},
{
"epoch": 0.7453549565030609,
"grad_norm": 0.2651905119419098,
"learning_rate": 1.612514787350497e-05,
"loss": 1.2746322631835938,
"step": 1735
},
{
"epoch": 0.7475029534958652,
"grad_norm": 0.2581336200237274,
"learning_rate": 1.5870132414260514e-05,
"loss": 1.3463025093078613,
"step": 1740
},
{
"epoch": 0.7496509504886694,
"grad_norm": 0.2893677353858948,
"learning_rate": 1.5616768630531355e-05,
"loss": 1.2779773712158202,
"step": 1745
},
{
"epoch": 0.7517989474814736,
"grad_norm": 0.2981060743331909,
"learning_rate": 1.536506878356515e-05,
"loss": 1.260219192504883,
"step": 1750
},
{
"epoch": 0.7539469444742778,
"grad_norm": 0.29954954981803894,
"learning_rate": 1.5115045054085203e-05,
"loss": 1.284311866760254,
"step": 1755
},
{
"epoch": 0.7560949414670819,
"grad_norm": 0.2798292636871338,
"learning_rate": 1.4866709541701084e-05,
"loss": 1.2953471183776855,
"step": 1760
},
{
"epoch": 0.7582429384598861,
"grad_norm": 0.2985500693321228,
"learning_rate": 1.4620074264323046e-05,
"loss": 1.276895523071289,
"step": 1765
},
{
"epoch": 0.7603909354526903,
"grad_norm": 0.30349722504615784,
"learning_rate": 1.4375151157580424e-05,
"loss": 1.2596096038818358,
"step": 1770
},
{
"epoch": 0.7625389324454945,
"grad_norm": 0.304608017206192,
"learning_rate": 1.4131952074244037e-05,
"loss": 1.317898654937744,
"step": 1775
},
{
"epoch": 0.7646869294382987,
"grad_norm": 0.3377379775047302,
"learning_rate": 1.3890488783652578e-05,
"loss": 1.2119129180908204,
"step": 1780
},
{
"epoch": 0.766834926431103,
"grad_norm": 0.2883898913860321,
"learning_rate": 1.3650772971143066e-05,
"loss": 1.2787168502807618,
"step": 1785
},
{
"epoch": 0.7689829234239072,
"grad_norm": 0.2671854496002197,
"learning_rate": 1.3412816237485326e-05,
"loss": 1.2940439224243163,
"step": 1790
},
{
"epoch": 0.7711309204167114,
"grad_norm": 0.2621491849422455,
"learning_rate": 1.3176630098320614e-05,
"loss": 1.269842529296875,
"step": 1795
},
{
"epoch": 0.7732789174095156,
"grad_norm": 0.28728634119033813,
"learning_rate": 1.2942225983604267e-05,
"loss": 1.2921714782714844,
"step": 1800
},
{
"epoch": 0.7754269144023198,
"grad_norm": 0.3033546209335327,
"learning_rate": 1.2709615237052642e-05,
"loss": 1.2844670295715332,
"step": 1805
},
{
"epoch": 0.777574911395124,
"grad_norm": 0.2372342348098755,
"learning_rate": 1.2478809115594115e-05,
"loss": 1.2834739685058594,
"step": 1810
},
{
"epoch": 0.7797229083879282,
"grad_norm": 0.29628515243530273,
"learning_rate": 1.2249818788824325e-05,
"loss": 1.2726101875305176,
"step": 1815
},
{
"epoch": 0.7818709053807325,
"grad_norm": 0.2835840880870819,
"learning_rate": 1.2022655338465576e-05,
"loss": 1.301106834411621,
"step": 1820
},
{
"epoch": 0.7840189023735367,
"grad_norm": 0.2993585765361786,
"learning_rate": 1.179732975783065e-05,
"loss": 1.2663209915161133,
"step": 1825
},
{
"epoch": 0.7861668993663409,
"grad_norm": 0.24655503034591675,
"learning_rate": 1.157385295129072e-05,
"loss": 1.3068438529968263,
"step": 1830
},
{
"epoch": 0.7883148963591451,
"grad_norm": 0.3119746446609497,
"learning_rate": 1.1352235733747685e-05,
"loss": 1.3065114974975587,
"step": 1835
},
{
"epoch": 0.7904628933519493,
"grad_norm": 0.2524607479572296,
"learning_rate": 1.1132488830110777e-05,
"loss": 1.3178667068481444,
"step": 1840
},
{
"epoch": 0.7926108903447535,
"grad_norm": 0.28341639041900635,
"learning_rate": 1.0914622874777548e-05,
"loss": 1.2802416801452636,
"step": 1845
},
{
"epoch": 0.7947588873375577,
"grad_norm": 0.30353984236717224,
"learning_rate": 1.0698648411119232e-05,
"loss": 1.2557078361511231,
"step": 1850
},
{
"epoch": 0.7969068843303619,
"grad_norm": 0.25872737169265747,
"learning_rate": 1.0484575890970506e-05,
"loss": 1.2983148574829102,
"step": 1855
},
{
"epoch": 0.7990548813231662,
"grad_norm": 0.30864548683166504,
"learning_rate": 1.0272415674123691e-05,
"loss": 1.3070337295532226,
"step": 1860
},
{
"epoch": 0.8012028783159704,
"grad_norm": 0.2922477424144745,
"learning_rate": 1.0062178027827385e-05,
"loss": 1.321023178100586,
"step": 1865
},
{
"epoch": 0.8033508753087746,
"grad_norm": 0.30001819133758545,
"learning_rate": 9.85387312628962e-06,
"loss": 1.2790605545043945,
"step": 1870
},
{
"epoch": 0.8054988723015788,
"grad_norm": 0.3499928414821625,
"learning_rate": 9.647511050185476e-06,
"loss": 1.2983773231506348,
"step": 1875
},
{
"epoch": 0.807646869294383,
"grad_norm": 0.2724253535270691,
"learning_rate": 9.443101786169251e-06,
"loss": 1.2426782608032227,
"step": 1880
},
{
"epoch": 0.8097948662871872,
"grad_norm": 0.28038036823272705,
"learning_rate": 9.24065522639112e-06,
"loss": 1.3003353118896483,
"step": 1885
},
{
"epoch": 0.8119428632799914,
"grad_norm": 0.30941468477249146,
"learning_rate": 9.04018116801849e-06,
"loss": 1.2399185180664063,
"step": 1890
},
{
"epoch": 0.8140908602727956,
"grad_norm": 0.2753143608570099,
"learning_rate": 8.841689312761836e-06,
"loss": 1.2843408584594727,
"step": 1895
},
{
"epoch": 0.8162388572655999,
"grad_norm": 0.26509711146354675,
"learning_rate": 8.64518926640519e-06,
"loss": 1.281872272491455,
"step": 1900
},
{
"epoch": 0.8183868542584041,
"grad_norm": 0.28912749886512756,
"learning_rate": 8.4506905383413e-06,
"loss": 1.2745040893554687,
"step": 1905
},
{
"epoch": 0.8205348512512083,
"grad_norm": 0.2923169732093811,
"learning_rate": 8.258202541111437e-06,
"loss": 1.2869806289672852,
"step": 1910
},
{
"epoch": 0.8226828482440125,
"grad_norm": 0.3156343698501587,
"learning_rate": 8.067734589949861e-06,
"loss": 1.2547836303710938,
"step": 1915
},
{
"epoch": 0.8248308452368167,
"grad_norm": 0.28888848423957825,
"learning_rate": 7.879295902333034e-06,
"loss": 1.3018421173095702,
"step": 1920
},
{
"epoch": 0.8269788422296209,
"grad_norm": 0.24648155272006989,
"learning_rate": 7.692895597533584e-06,
"loss": 1.2865129470825196,
"step": 1925
},
{
"epoch": 0.8291268392224251,
"grad_norm": 0.2558462619781494,
"learning_rate": 7.50854269617891e-06,
"loss": 1.2889558792114257,
"step": 1930
},
{
"epoch": 0.8312748362152294,
"grad_norm": 0.2943022847175598,
"learning_rate": 7.326246119814711e-06,
"loss": 1.240505313873291,
"step": 1935
},
{
"epoch": 0.8334228332080335,
"grad_norm": 0.2633776366710663,
"learning_rate": 7.146014690473224e-06,
"loss": 1.3057293891906738,
"step": 1940
},
{
"epoch": 0.8355708302008377,
"grad_norm": 0.2737610936164856,
"learning_rate": 6.967857130246281e-06,
"loss": 1.2296178817749024,
"step": 1945
},
{
"epoch": 0.8377188271936419,
"grad_norm": 0.31707239151000977,
"learning_rate": 6.791782060863172e-06,
"loss": 1.2079605102539062,
"step": 1950
},
{
"epoch": 0.8398668241864461,
"grad_norm": 0.32487425208091736,
"learning_rate": 6.617798003273495e-06,
"loss": 1.2449745178222655,
"step": 1955
},
{
"epoch": 0.8420148211792503,
"grad_norm": 0.26417678594589233,
"learning_rate": 6.445913377234736e-06,
"loss": 1.2077396392822266,
"step": 1960
},
{
"epoch": 0.8441628181720545,
"grad_norm": 0.31679192185401917,
"learning_rate": 6.276136500904822e-06,
"loss": 1.2758880615234376,
"step": 1965
},
{
"epoch": 0.8463108151648587,
"grad_norm": 0.33341100811958313,
"learning_rate": 6.108475590439522e-06,
"loss": 1.2420418739318848,
"step": 1970
},
{
"epoch": 0.848458812157663,
"grad_norm": 0.31125956773757935,
"learning_rate": 5.942938759594952e-06,
"loss": 1.28662109375,
"step": 1975
},
{
"epoch": 0.8506068091504672,
"grad_norm": 0.31284815073013306,
"learning_rate": 5.7795340193348035e-06,
"loss": 1.265916633605957,
"step": 1980
},
{
"epoch": 0.8527548061432714,
"grad_norm": 0.3097616732120514,
"learning_rate": 5.618269277442722e-06,
"loss": 1.2471879959106444,
"step": 1985
},
{
"epoch": 0.8549028031360756,
"grad_norm": 0.29485225677490234,
"learning_rate": 5.459152338139606e-06,
"loss": 1.2311647415161133,
"step": 1990
},
{
"epoch": 0.8570508001288798,
"grad_norm": 0.27610111236572266,
"learning_rate": 5.3021909017059e-06,
"loss": 1.2552482604980468,
"step": 1995
},
{
"epoch": 0.859198797121684,
"grad_norm": 0.43170663714408875,
"learning_rate": 5.14739256410901e-06,
"loss": 1.2518775939941407,
"step": 2000
},
{
"epoch": 0.8613467941144882,
"grad_norm": 0.2928427755832672,
"learning_rate": 4.994764816635666e-06,
"loss": 1.2499638557434083,
"step": 2005
},
{
"epoch": 0.8634947911072924,
"grad_norm": 0.33848118782043457,
"learning_rate": 4.844315045529407e-06,
"loss": 1.2733563423156737,
"step": 2010
},
{
"epoch": 0.8656427881000966,
"grad_norm": 0.2882305681705475,
"learning_rate": 4.696050531633078e-06,
"loss": 1.2659109115600586,
"step": 2015
},
{
"epoch": 0.8677907850929009,
"grad_norm": 0.3241490125656128,
"learning_rate": 4.549978450036568e-06,
"loss": 1.2454627990722655,
"step": 2020
},
{
"epoch": 0.8699387820857051,
"grad_norm": 0.29955822229385376,
"learning_rate": 4.406105869729532e-06,
"loss": 1.2802690505981444,
"step": 2025
},
{
"epoch": 0.8720867790785093,
"grad_norm": 0.2901234030723572,
"learning_rate": 4.264439753259292e-06,
"loss": 1.2602261543273925,
"step": 2030
},
{
"epoch": 0.8742347760713135,
"grad_norm": 0.2517850995063782,
"learning_rate": 4.124986956393895e-06,
"loss": 1.2761344909667969,
"step": 2035
},
{
"epoch": 0.8763827730641177,
"grad_norm": 0.3021356761455536,
"learning_rate": 3.987754227790341e-06,
"loss": 1.2739974975585937,
"step": 2040
},
{
"epoch": 0.8785307700569219,
"grad_norm": 0.27927812933921814,
"learning_rate": 3.852748208668028e-06,
"loss": 1.2833540916442872,
"step": 2045
},
{
"epoch": 0.8806787670497261,
"grad_norm": 0.22617845237255096,
"learning_rate": 3.719975432487277e-06,
"loss": 1.2788030624389648,
"step": 2050
},
{
"epoch": 0.8828267640425304,
"grad_norm": 0.2822844386100769,
"learning_rate": 3.589442324633224e-06,
"loss": 1.2651789665222168,
"step": 2055
},
{
"epoch": 0.8849747610353346,
"grad_norm": 0.24618762731552124,
"learning_rate": 3.461155202104821e-06,
"loss": 1.2958842277526856,
"step": 2060
},
{
"epoch": 0.8871227580281388,
"grad_norm": 0.330362468957901,
"learning_rate": 3.335120273209169e-06,
"loss": 1.2687481880187987,
"step": 2065
},
{
"epoch": 0.889270755020943,
"grad_norm": 0.267100065946579,
"learning_rate": 3.2113436372610605e-06,
"loss": 1.2643348693847656,
"step": 2070
},
{
"epoch": 0.8914187520137472,
"grad_norm": 0.2676951289176941,
"learning_rate": 3.089831284287814e-06,
"loss": 1.2248193740844726,
"step": 2075
},
{
"epoch": 0.8935667490065514,
"grad_norm": 0.2669046223163605,
"learning_rate": 2.9705890947393624e-06,
"loss": 1.3051311492919921,
"step": 2080
},
{
"epoch": 0.8957147459993556,
"grad_norm": 0.29845312237739563,
"learning_rate": 2.8536228392037403e-06,
"loss": 1.2953030586242675,
"step": 2085
},
{
"epoch": 0.8978627429921598,
"grad_norm": 0.26989129185676575,
"learning_rate": 2.7389381781277533e-06,
"loss": 1.2524210929870605,
"step": 2090
},
{
"epoch": 0.9000107399849641,
"grad_norm": 0.31527775526046753,
"learning_rate": 2.6265406615431033e-06,
"loss": 1.299118423461914,
"step": 2095
},
{
"epoch": 0.9021587369777683,
"grad_norm": 0.2844144105911255,
"learning_rate": 2.51643572879775e-06,
"loss": 1.2464170455932617,
"step": 2100
},
{
"epoch": 0.9043067339705725,
"grad_norm": 0.303445965051651,
"learning_rate": 2.4086287082927317e-06,
"loss": 1.306422233581543,
"step": 2105
},
{
"epoch": 0.9064547309633767,
"grad_norm": 0.2937731146812439,
"learning_rate": 2.303124817224267e-06,
"loss": 1.280664825439453,
"step": 2110
},
{
"epoch": 0.9086027279561809,
"grad_norm": 0.27328890562057495,
"learning_rate": 2.1999291613312823e-06,
"loss": 1.296454620361328,
"step": 2115
},
{
"epoch": 0.9107507249489851,
"grad_norm": 0.2823365032672882,
"learning_rate": 2.0990467346483466e-06,
"loss": 1.3039875030517578,
"step": 2120
},
{
"epoch": 0.9128987219417892,
"grad_norm": 0.27393898367881775,
"learning_rate": 2.000482419263944e-06,
"loss": 1.2848936080932618,
"step": 2125
},
{
"epoch": 0.9150467189345934,
"grad_norm": 0.2775443494319916,
"learning_rate": 1.9042409850842658e-06,
"loss": 1.3306063652038573,
"step": 2130
},
{
"epoch": 0.9171947159273977,
"grad_norm": 0.2975456416606903,
"learning_rate": 1.8103270896023427e-06,
"loss": 1.3341773986816405,
"step": 2135
},
{
"epoch": 0.9193427129202019,
"grad_norm": 0.27994534373283386,
"learning_rate": 1.718745277672662e-06,
"loss": 1.2672662734985352,
"step": 2140
},
{
"epoch": 0.9214907099130061,
"grad_norm": 0.2889311611652374,
"learning_rate": 1.629499981291205e-06,
"loss": 1.2739192962646484,
"step": 2145
},
{
"epoch": 0.9236387069058103,
"grad_norm": 0.26127493381500244,
"learning_rate": 1.5425955193810016e-06,
"loss": 1.2637762069702148,
"step": 2150
},
{
"epoch": 0.9257867038986145,
"grad_norm": 0.2669077217578888,
"learning_rate": 1.4580360975830986e-06,
"loss": 1.287917709350586,
"step": 2155
},
{
"epoch": 0.9279347008914187,
"grad_norm": 0.33365505933761597,
"learning_rate": 1.3758258080530273e-06,
"loss": 1.2490084648132325,
"step": 2160
},
{
"epoch": 0.9300826978842229,
"grad_norm": 0.2899354100227356,
"learning_rate": 1.2959686292627782e-06,
"loss": 1.28909912109375,
"step": 2165
},
{
"epoch": 0.9322306948770271,
"grad_norm": 0.30653882026672363,
"learning_rate": 1.2184684258082723e-06,
"loss": 1.2363777160644531,
"step": 2170
},
{
"epoch": 0.9343786918698314,
"grad_norm": 0.3222835659980774,
"learning_rate": 1.1433289482223274e-06,
"loss": 1.2729862213134766,
"step": 2175
},
{
"epoch": 0.9365266888626356,
"grad_norm": 0.3175857961177826,
"learning_rate": 1.0705538327931587e-06,
"loss": 1.2359086990356445,
"step": 2180
},
{
"epoch": 0.9386746858554398,
"grad_norm": 0.23822984099388123,
"learning_rate": 1.000146601388413e-06,
"loss": 1.3249876022338867,
"step": 2185
},
{
"epoch": 0.940822682848244,
"grad_norm": 0.3522796928882599,
"learning_rate": 9.321106612847185e-07,
"loss": 1.3001674652099608,
"step": 2190
},
{
"epoch": 0.9429706798410482,
"grad_norm": 0.30886110663414,
"learning_rate": 8.664493050028033e-07,
"loss": 1.251561737060547,
"step": 2195
},
{
"epoch": 0.9451186768338524,
"grad_norm": 0.296652227640152,
"learning_rate": 8.031657101481516e-07,
"loss": 1.2635051727294921,
"step": 2200
},
{
"epoch": 0.9472666738266566,
"grad_norm": 0.29425349831581116,
"learning_rate": 7.422629392572323e-07,
"loss": 1.2742862701416016,
"step": 2205
},
{
"epoch": 0.9494146708194608,
"grad_norm": 0.28834399580955505,
"learning_rate": 6.83743939649284e-07,
"loss": 1.2603504180908203,
"step": 2210
},
{
"epoch": 0.9515626678122651,
"grad_norm": 0.34383395314216614,
"learning_rate": 6.27611543283696e-07,
"loss": 1.2798163414001464,
"step": 2215
},
{
"epoch": 0.9537106648050693,
"grad_norm": 0.3192986845970154,
"learning_rate": 5.73868466622951e-07,
"loss": 1.222487735748291,
"step": 2220
},
{
"epoch": 0.9558586617978735,
"grad_norm": 0.2906084358692169,
"learning_rate": 5.225173105011582e-07,
"loss": 1.322528076171875,
"step": 2225
},
{
"epoch": 0.9580066587906777,
"grad_norm": 0.30103063583374023,
"learning_rate": 4.735605599981874e-07,
"loss": 1.2273588180541992,
"step": 2230
},
{
"epoch": 0.9601546557834819,
"grad_norm": 0.2715284824371338,
"learning_rate": 4.270005843194369e-07,
"loss": 1.31506290435791,
"step": 2235
},
{
"epoch": 0.9623026527762861,
"grad_norm": 0.3000585734844208,
"learning_rate": 3.8283963668114266e-07,
"loss": 1.2237860679626464,
"step": 2240
},
{
"epoch": 0.9644506497690903,
"grad_norm": 0.3379741311073303,
"learning_rate": 3.4107985420134824e-07,
"loss": 1.264302635192871,
"step": 2245
},
{
"epoch": 0.9665986467618946,
"grad_norm": 0.37124356627464294,
"learning_rate": 3.017232577964824e-07,
"loss": 1.2457963943481445,
"step": 2250
},
{
"epoch": 0.9687466437546988,
"grad_norm": 0.26245129108428955,
"learning_rate": 2.647717520835702e-07,
"loss": 1.2051671028137207,
"step": 2255
},
{
"epoch": 0.970894640747503,
"grad_norm": 0.342318058013916,
"learning_rate": 2.3022712528804612e-07,
"loss": 1.2445919036865234,
"step": 2260
},
{
"epoch": 0.9730426377403072,
"grad_norm": 0.27479904890060425,
"learning_rate": 1.9809104915721187e-07,
"loss": 1.2522891998291015,
"step": 2265
},
{
"epoch": 0.9751906347331114,
"grad_norm": 0.2656843364238739,
"learning_rate": 1.683650788793567e-07,
"loss": 1.289116096496582,
"step": 2270
},
{
"epoch": 0.9773386317259156,
"grad_norm": 0.2872951924800873,
"learning_rate": 1.410506530084621e-07,
"loss": 1.2644329071044922,
"step": 2275
},
{
"epoch": 0.9794866287187198,
"grad_norm": 0.3204033374786377,
"learning_rate": 1.161490933946241e-07,
"loss": 1.2509076118469238,
"step": 2280
},
{
"epoch": 0.981634625711524,
"grad_norm": 0.308625727891922,
"learning_rate": 9.366160512004341e-08,
"loss": 1.265985870361328,
"step": 2285
},
{
"epoch": 0.9837826227043283,
"grad_norm": 0.27000510692596436,
"learning_rate": 7.358927644075531e-08,
"loss": 1.2676359176635743,
"step": 2290
},
{
"epoch": 0.9859306196971325,
"grad_norm": 0.2328324317932129,
"learning_rate": 5.5933078733899634e-08,
"loss": 1.2606651306152343,
"step": 2295
},
{
"epoch": 0.9880786166899367,
"grad_norm": 0.3303236663341522,
"learning_rate": 4.069386645078055e-08,
"loss": 1.2653938293457032,
"step": 2300
},
{
"epoch": 0.9902266136827409,
"grad_norm": 0.2890678942203522,
"learning_rate": 2.7872377075471857e-08,
"loss": 1.2857831954956054,
"step": 2305
},
{
"epoch": 0.992374610675545,
"grad_norm": 0.26837414503097534,
"learning_rate": 1.746923108913445e-08,
"loss": 1.3168321609497071,
"step": 2310
},
{
"epoch": 0.9945226076683492,
"grad_norm": 0.30601441860198975,
"learning_rate": 9.48493194000144e-09,
"loss": 1.2843009948730468,
"step": 2315
},
{
"epoch": 0.9966706046611534,
"grad_norm": 0.3158735930919647,
"learning_rate": 3.919866019008733e-09,
"loss": 1.2655706405639648,
"step": 2320
},
{
"epoch": 0.9988186016539576,
"grad_norm": 0.30448752641677856,
"learning_rate": 7.743026410989007e-10,
"loss": 1.3010024070739745,
"step": 2325
}
],
"logging_steps": 5,
"max_steps": 2328,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.7382193779533414e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}