food-recognition / trainer_log_history.json
anonymous-eval's picture
Upload dinov3_custom food classifier from run 3: eps_8
2e2eb08 verified
Raw History Blame Contribute Delete
424 kB
Invalid JSON:Unexpected token 'I', ..."ad_norm": Infinity, "... is not valid JSON
[
{
"loss": 5.9751,
"grad_norm": 5.27304744720459,
"learning_rate": 1.492764661081493e-07,
"epoch": 0.0030467369447321918,
"step": 50
},
{
"loss": 5.9993,
"grad_norm": 5.152888298034668,
"learning_rate": 3.015993907083016e-07,
"epoch": 0.0060934738894643835,
"step": 100
},
{
"loss": 6.0075,
"grad_norm": 5.225321292877197,
"learning_rate": 4.53922315308454e-07,
"epoch": 0.009140210834196576,
"step": 150
},
{
"loss": 5.9946,
"grad_norm": 5.201633453369141,
"learning_rate": 6.062452399086063e-07,
"epoch": 0.012186947778928767,
"step": 200
},
{
"loss": 5.9842,
"grad_norm": 5.254497051239014,
"learning_rate": 7.585681645087586e-07,
"epoch": 0.015233684723660958,
"step": 250
},
{
"loss": 5.9924,
"grad_norm": 5.069004535675049,
"learning_rate": 9.10891089108911e-07,
"epoch": 0.01828042166839315,
"step": 300
},
{
"loss": 5.9708,
"grad_norm": 5.476426124572754,
"learning_rate": 1.0632140137090633e-06,
"epoch": 0.021327158613125344,
"step": 350
},
{
"loss": 5.9874,
"grad_norm": 5.278822898864746,
"learning_rate": 1.2155369383092157e-06,
"epoch": 0.024373895557857534,
"step": 400
},
{
"loss": 5.9305,
"grad_norm": 5.147547721862793,
"learning_rate": 1.3678598629093677e-06,
"epoch": 0.027420632502589727,
"step": 450
},
{
"loss": 5.9428,
"grad_norm": 5.292663097381592,
"learning_rate": 1.5201827875095204e-06,
"epoch": 0.030467369447321917,
"step": 500
},
{
"loss": 5.9342,
"grad_norm": 5.335788249969482,
"learning_rate": 1.6725057121096728e-06,
"epoch": 0.03351410639205411,
"step": 550
},
{
"loss": 5.9357,
"grad_norm": 5.418006420135498,
"learning_rate": 1.824828636709825e-06,
"epoch": 0.0365608433367863,
"step": 600
},
{
"loss": 5.9177,
"grad_norm": 5.346350193023682,
"learning_rate": 1.9771515613099775e-06,
"epoch": 0.03960758028151849,
"step": 650
},
{
"loss": 5.9065,
"grad_norm": 5.298623085021973,
"learning_rate": 2.1294744859101295e-06,
"epoch": 0.04265431722625069,
"step": 700
},
{
"loss": 5.8938,
"grad_norm": 5.096645832061768,
"learning_rate": 2.281797410510282e-06,
"epoch": 0.04570105417098288,
"step": 750
},
{
"loss": 5.8845,
"grad_norm": 5.291286468505859,
"learning_rate": 2.4341203351104343e-06,
"epoch": 0.04874779111571507,
"step": 800
},
{
"loss": 5.8681,
"grad_norm": 5.280112266540527,
"learning_rate": 2.5864432597105863e-06,
"epoch": 0.05179452806044726,
"step": 850
},
{
"loss": 5.8101,
"grad_norm": 5.692427158355713,
"learning_rate": 2.7387661843107388e-06,
"epoch": 0.054841265005179454,
"step": 900
},
{
"loss": 5.7986,
"grad_norm": 5.478259086608887,
"learning_rate": 2.8910891089108912e-06,
"epoch": 0.057888001949911644,
"step": 950
},
{
"loss": 5.7799,
"grad_norm": 5.487733364105225,
"learning_rate": 3.0434120335110432e-06,
"epoch": 0.06093473889464383,
"step": 1000
},
{
"loss": 5.7674,
"grad_norm": 5.516610622406006,
"learning_rate": 3.195734958111196e-06,
"epoch": 0.06398147583937602,
"step": 1050
},
{
"loss": 5.742,
"grad_norm": 5.322267532348633,
"learning_rate": 3.3480578827113485e-06,
"epoch": 0.06702821278410823,
"step": 1100
},
{
"loss": 5.7037,
"grad_norm": 5.947967529296875,
"learning_rate": 3.500380807311501e-06,
"epoch": 0.07007494972884042,
"step": 1150
},
{
"loss": 5.6833,
"grad_norm": 5.84609317779541,
"learning_rate": 3.6527037319116534e-06,
"epoch": 0.0731216866735726,
"step": 1200
},
{
"loss": 5.6596,
"grad_norm": 5.530036926269531,
"learning_rate": 3.8050266565118054e-06,
"epoch": 0.0761684236183048,
"step": 1250
},
{
"loss": 5.5964,
"grad_norm": 5.843346118927002,
"learning_rate": 3.957349581111958e-06,
"epoch": 0.07921516056303698,
"step": 1300
},
{
"loss": 5.5659,
"grad_norm": 5.905080318450928,
"learning_rate": 4.10967250571211e-06,
"epoch": 0.08226189750776917,
"step": 1350
},
{
"loss": 5.5292,
"grad_norm": 5.716760158538818,
"learning_rate": 4.261995430312263e-06,
"epoch": 0.08530863445250138,
"step": 1400
},
{
"loss": 5.4888,
"grad_norm": 6.588194847106934,
"learning_rate": 4.414318354912414e-06,
"epoch": 0.08835537139723357,
"step": 1450
},
{
"loss": 5.4326,
"grad_norm": 6.4131340980529785,
"learning_rate": 4.566641279512567e-06,
"epoch": 0.09140210834196576,
"step": 1500
},
{
"loss": 5.4188,
"grad_norm": 5.992576599121094,
"learning_rate": 4.718964204112719e-06,
"epoch": 0.09444884528669795,
"step": 1550
},
{
"loss": 5.39,
"grad_norm": 6.338135719299316,
"learning_rate": 4.8712871287128716e-06,
"epoch": 0.09749558223143014,
"step": 1600
},
{
"loss": 5.3291,
"grad_norm": 7.01084566116333,
"learning_rate": 5.023610053313024e-06,
"epoch": 0.10054231917616233,
"step": 1650
},
{
"loss": 5.2486,
"grad_norm": 7.46585750579834,
"learning_rate": 5.1759329779131764e-06,
"epoch": 0.10358905612089452,
"step": 1700
},
{
"loss": 5.2243,
"grad_norm": 6.56162691116333,
"learning_rate": 5.328255902513329e-06,
"epoch": 0.10663579306562672,
"step": 1750
},
{
"loss": 5.1495,
"grad_norm": 6.631102085113525,
"learning_rate": 5.4805788271134804e-06,
"epoch": 0.10968253001035891,
"step": 1800
},
{
"loss": 5.0484,
"grad_norm": 7.503003120422363,
"learning_rate": 5.632901751713633e-06,
"epoch": 0.1127292669550911,
"step": 1850
},
{
"loss": 5.0551,
"grad_norm": 7.3272528648376465,
"learning_rate": 5.7821782178217824e-06,
"epoch": 0.11577600389982329,
"step": 1900
},
{
"loss": 4.9751,
"grad_norm": 6.4639105796813965,
"learning_rate": 5.934501142421935e-06,
"epoch": 0.11882274084455548,
"step": 1950
},
{
"loss": 4.9326,
"grad_norm": 6.953883647918701,
"learning_rate": 6.0868240670220864e-06,
"epoch": 0.12186947778928767,
"step": 2000
},
{
"loss": 4.8976,
"grad_norm": 7.850661754608154,
"learning_rate": 6.239146991622239e-06,
"epoch": 0.12491621473401987,
"step": 2050
},
{
"loss": 4.8298,
"grad_norm": 7.28013277053833,
"learning_rate": 6.391469916222392e-06,
"epoch": 0.12796295167875205,
"step": 2100
},
{
"loss": 4.7531,
"grad_norm": 7.121276378631592,
"learning_rate": 6.540746382330542e-06,
"epoch": 0.13100968862348425,
"step": 2150
},
{
"loss": 4.6724,
"grad_norm": 8.31320571899414,
"learning_rate": 6.693069306930694e-06,
"epoch": 0.13405642556821645,
"step": 2200
},
{
"loss": 4.6356,
"grad_norm": 7.8836894035339355,
"learning_rate": 6.8453922315308466e-06,
"epoch": 0.13710316251294863,
"step": 2250
},
{
"loss": 4.5831,
"grad_norm": 8.688776969909668,
"learning_rate": 6.997715156130999e-06,
"epoch": 0.14014989945768083,
"step": 2300
},
{
"loss": 4.5336,
"grad_norm": 7.627803325653076,
"learning_rate": 7.150038080731151e-06,
"epoch": 0.143196636402413,
"step": 2350
},
{
"loss": 4.4077,
"grad_norm": 8.144109725952148,
"learning_rate": 7.302361005331303e-06,
"epoch": 0.1462433733471452,
"step": 2400
},
{
"loss": 4.3987,
"grad_norm": 8.045475959777832,
"learning_rate": 7.4546839299314555e-06,
"epoch": 0.1492901102918774,
"step": 2450
},
{
"loss": 4.2843,
"grad_norm": 7.78476095199585,
"learning_rate": 7.607006854531608e-06,
"epoch": 0.1523368472366096,
"step": 2500
},
{
"loss": 4.2641,
"grad_norm": 7.722534656524658,
"learning_rate": 7.75932977913176e-06,
"epoch": 0.1553835841813418,
"step": 2550
},
{
"loss": 4.1534,
"grad_norm": 8.186152458190918,
"learning_rate": 7.911652703731913e-06,
"epoch": 0.15843032112607397,
"step": 2600
},
{
"loss": 4.1537,
"grad_norm": 9.185806274414062,
"learning_rate": 8.063975628332065e-06,
"epoch": 0.16147705807080617,
"step": 2650
},
{
"loss": 4.1343,
"grad_norm": 7.938891410827637,
"learning_rate": 8.216298552932218e-06,
"epoch": 0.16452379501553835,
"step": 2700
},
{
"loss": 4.014,
"grad_norm": 8.34693431854248,
"learning_rate": 8.36862147753237e-06,
"epoch": 0.16757053196027055,
"step": 2750
},
{
"loss": 3.8889,
"grad_norm": 8.644932746887207,
"learning_rate": 8.52094440213252e-06,
"epoch": 0.17061726890500276,
"step": 2800
},
{
"loss": 3.8913,
"grad_norm": 10.140563011169434,
"learning_rate": 8.673267326732673e-06,
"epoch": 0.17366400584973493,
"step": 2850
},
{
"loss": 3.7699,
"grad_norm": 8.651476860046387,
"learning_rate": 8.825590251332826e-06,
"epoch": 0.17671074279446713,
"step": 2900
},
{
"loss": 3.8222,
"grad_norm": 8.98681640625,
"learning_rate": 8.977913175932978e-06,
"epoch": 0.1797574797391993,
"step": 2950
},
{
"loss": 3.7194,
"grad_norm": 8.303566932678223,
"learning_rate": 9.13023610053313e-06,
"epoch": 0.18280421668393151,
"step": 3000
},
{
"loss": 3.7134,
"grad_norm": 9.879234313964844,
"learning_rate": 9.282559025133283e-06,
"epoch": 0.1858509536286637,
"step": 3050
},
{
"loss": 3.629,
"grad_norm": 8.481127738952637,
"learning_rate": 9.434881949733435e-06,
"epoch": 0.1888976905733959,
"step": 3100
},
{
"loss": 3.5589,
"grad_norm": 7.538110256195068,
"learning_rate": 9.587204874333588e-06,
"epoch": 0.1919444275181281,
"step": 3150
},
{
"loss": 3.5331,
"grad_norm": 9.237178802490234,
"learning_rate": 9.73952779893374e-06,
"epoch": 0.19499116446286027,
"step": 3200
},
{
"loss": 3.4027,
"grad_norm": 8.37112045288086,
"learning_rate": 9.891850723533893e-06,
"epoch": 0.19803790140759248,
"step": 3250
},
{
"loss": 3.3643,
"grad_norm": 7.3935465812683105,
"learning_rate": 1.0044173648134045e-05,
"epoch": 0.20108463835232465,
"step": 3300
},
{
"loss": 3.3308,
"grad_norm": 10.61789321899414,
"learning_rate": 1.0196496572734198e-05,
"epoch": 0.20413137529705686,
"step": 3350
},
{
"loss": 3.2593,
"grad_norm": 9.292555809020996,
"learning_rate": 1.034881949733435e-05,
"epoch": 0.20717811224178903,
"step": 3400
},
{
"loss": 3.1899,
"grad_norm": 7.877501010894775,
"learning_rate": 1.0501142421934502e-05,
"epoch": 0.21022484918652123,
"step": 3450
},
{
"loss": 3.1948,
"grad_norm": 9.454145431518555,
"learning_rate": 1.0653465346534655e-05,
"epoch": 0.21327158613125344,
"step": 3500
},
{
"loss": 3.104,
"grad_norm": 10.389566421508789,
"learning_rate": 1.0805788271134807e-05,
"epoch": 0.2163183230759856,
"step": 3550
},
{
"loss": 3.0393,
"grad_norm": 9.737595558166504,
"learning_rate": 1.095811119573496e-05,
"epoch": 0.21936506002071782,
"step": 3600
},
{
"loss": 3.0585,
"grad_norm": 11.218122482299805,
"learning_rate": 1.1110434120335112e-05,
"epoch": 0.22241179696545,
"step": 3650
},
{
"loss": 2.943,
"grad_norm": 11.813579559326172,
"learning_rate": 1.1262757044935265e-05,
"epoch": 0.2254585339101822,
"step": 3700
},
{
"loss": 2.9038,
"grad_norm": 11.747238159179688,
"learning_rate": 1.1415079969535417e-05,
"epoch": 0.2285052708549144,
"step": 3750
},
{
"loss": 2.8321,
"grad_norm": 9.895445823669434,
"learning_rate": 1.156740289413557e-05,
"epoch": 0.23155200779964658,
"step": 3800
},
{
"loss": 2.716,
"grad_norm": 10.228238105773926,
"learning_rate": 1.1719725818735722e-05,
"epoch": 0.23459874474437878,
"step": 3850
},
{
"loss": 2.6983,
"grad_norm": 11.801738739013672,
"learning_rate": 1.1872048743335873e-05,
"epoch": 0.23764548168911095,
"step": 3900
},
{
"loss": 2.6738,
"grad_norm": 9.524603843688965,
"learning_rate": 1.2024371667936025e-05,
"epoch": 0.24069221863384316,
"step": 3950
},
{
"loss": 2.6033,
"grad_norm": 10.915817260742188,
"learning_rate": 1.2176694592536177e-05,
"epoch": 0.24373895557857533,
"step": 4000
},
{
"loss": 2.672,
"grad_norm": 11.490041732788086,
"learning_rate": 1.232901751713633e-05,
"epoch": 0.24678569252330754,
"step": 4050
},
{
"loss": 2.5096,
"grad_norm": 10.562030792236328,
"learning_rate": 1.2481340441736482e-05,
"epoch": 0.24983242946803974,
"step": 4100
},
{
"loss": 2.4444,
"grad_norm": 9.54339599609375,
"learning_rate": 1.2633663366336635e-05,
"epoch": 0.2528791664127719,
"step": 4150
},
{
"loss": 2.4053,
"grad_norm": 9.020502090454102,
"learning_rate": 1.2785986290936787e-05,
"epoch": 0.2559259033575041,
"step": 4200
},
{
"loss": 2.4172,
"grad_norm": 10.14346694946289,
"learning_rate": 1.293830921553694e-05,
"epoch": 0.2589726403022363,
"step": 4250
},
{
"loss": 2.2723,
"grad_norm": 8.143074035644531,
"learning_rate": 1.3090632140137092e-05,
"epoch": 0.2620193772469685,
"step": 4300
},
{
"loss": 2.336,
"grad_norm": 12.532441139221191,
"learning_rate": 1.3242955064737244e-05,
"epoch": 0.2650661141917007,
"step": 4350
},
{
"loss": 2.2545,
"grad_norm": 9.84597110748291,
"learning_rate": 1.3395277989337397e-05,
"epoch": 0.2681128511364329,
"step": 4400
},
{
"loss": 2.2428,
"grad_norm": 14.032034873962402,
"learning_rate": 1.354760091393755e-05,
"epoch": 0.2711595880811651,
"step": 4450
},
{
"loss": 2.0751,
"grad_norm": 9.511150360107422,
"learning_rate": 1.36999238385377e-05,
"epoch": 0.27420632502589726,
"step": 4500
},
{
"loss": 2.0873,
"grad_norm": 9.423986434936523,
"learning_rate": 1.3852246763137853e-05,
"epoch": 0.27725306197062943,
"step": 4550
},
{
"loss": 2.1413,
"grad_norm": 9.668368339538574,
"learning_rate": 1.4004569687738005e-05,
"epoch": 0.28029979891536166,
"step": 4600
},
{
"loss": 2.0886,
"grad_norm": 9.700275421142578,
"learning_rate": 1.4156892612338157e-05,
"epoch": 0.28334653586009384,
"step": 4650
},
{
"loss": 2.005,
"grad_norm": 10.784494400024414,
"learning_rate": 1.430921553693831e-05,
"epoch": 0.286393272804826,
"step": 4700
},
{
"loss": 2.0066,
"grad_norm": 15.585013389587402,
"learning_rate": 1.4461538461538462e-05,
"epoch": 0.28944000974955825,
"step": 4750
},
{
"loss": 1.9403,
"grad_norm": 15.135519981384277,
"learning_rate": 1.4613861386138615e-05,
"epoch": 0.2924867466942904,
"step": 4800
},
{
"loss": 1.8715,
"grad_norm": 12.332038879394531,
"learning_rate": 1.4766184310738767e-05,
"epoch": 0.2955334836390226,
"step": 4850
},
{
"loss": 1.8645,
"grad_norm": 10.842533111572266,
"learning_rate": 1.491850723533892e-05,
"epoch": 0.2985802205837548,
"step": 4900
},
{
"loss": 1.778,
"grad_norm": 12.01357364654541,
"learning_rate": 1.5070830159939072e-05,
"epoch": 0.301626957528487,
"step": 4950
},
{
"loss": 1.7216,
"grad_norm": 11.972970008850098,
"learning_rate": 1.5223153084539224e-05,
"epoch": 0.3046736944732192,
"step": 5000
},
{
"loss": 1.8036,
"grad_norm": 12.13890266418457,
"learning_rate": 1.5375476009139377e-05,
"epoch": 0.30772043141795136,
"step": 5050
},
{
"loss": 1.7432,
"grad_norm": 10.001839637756348,
"learning_rate": 1.5527798933739528e-05,
"epoch": 0.3107671683626836,
"step": 5100
},
{
"loss": 1.6953,
"grad_norm": 14.449841499328613,
"learning_rate": 1.5680121858339682e-05,
"epoch": 0.31381390530741576,
"step": 5150
},
{
"loss": 1.5908,
"grad_norm": 9.0841064453125,
"learning_rate": 1.5832444782939832e-05,
"epoch": 0.31686064225214794,
"step": 5200
},
{
"loss": 1.6883,
"grad_norm": 10.635897636413574,
"learning_rate": 1.5984767707539987e-05,
"epoch": 0.3199073791968801,
"step": 5250
},
{
"loss": 1.6378,
"grad_norm": 8.82083511352539,
"learning_rate": 1.6137090632140137e-05,
"epoch": 0.32295411614161235,
"step": 5300
},
{
"loss": 1.5407,
"grad_norm": 8.60465145111084,
"learning_rate": 1.628941355674029e-05,
"epoch": 0.3260008530863445,
"step": 5350
},
{
"loss": 1.6559,
"grad_norm": 8.506077766418457,
"learning_rate": 1.6441736481340442e-05,
"epoch": 0.3290475900310767,
"step": 5400
},
{
"loss": 1.5501,
"grad_norm": 10.785340309143066,
"learning_rate": 1.6594059405940596e-05,
"epoch": 0.33209432697580893,
"step": 5450
},
{
"loss": 1.5191,
"grad_norm": 11.281582832336426,
"learning_rate": 1.6746382330540747e-05,
"epoch": 0.3351410639205411,
"step": 5500
},
{
"loss": 1.4853,
"grad_norm": 14.768205642700195,
"learning_rate": 1.6898705255140898e-05,
"epoch": 0.3381878008652733,
"step": 5550
},
{
"loss": 1.4734,
"grad_norm": 14.144624710083008,
"learning_rate": 1.7051028179741052e-05,
"epoch": 0.3412345378100055,
"step": 5600
},
{
"loss": 1.4715,
"grad_norm": 9.230982780456543,
"learning_rate": 1.7203351104341203e-05,
"epoch": 0.3442812747547377,
"step": 5650
},
{
"loss": 1.2806,
"grad_norm": 8.654589653015137,
"learning_rate": 1.7355674028941357e-05,
"epoch": 0.34732801169946986,
"step": 5700
},
{
"loss": 1.4039,
"grad_norm": 9.218396186828613,
"learning_rate": 1.7507996953541508e-05,
"epoch": 0.35037474864420204,
"step": 5750
},
{
"loss": 1.3854,
"grad_norm": 11.814713478088379,
"learning_rate": 1.766031987814166e-05,
"epoch": 0.35342148558893427,
"step": 5800
},
{
"loss": 1.4134,
"grad_norm": 27.65602684020996,
"learning_rate": 1.7812642802741812e-05,
"epoch": 0.35646822253366645,
"step": 5850
},
{
"loss": 1.3116,
"grad_norm": 10.332797050476074,
"learning_rate": 1.7964965727341966e-05,
"epoch": 0.3595149594783986,
"step": 5900
},
{
"loss": 1.2907,
"grad_norm": 11.859781265258789,
"learning_rate": 1.811728865194212e-05,
"epoch": 0.36256169642313085,
"step": 5950
},
{
"loss": 1.2568,
"grad_norm": 15.61054801940918,
"learning_rate": 1.826961157654227e-05,
"epoch": 0.36560843336786303,
"step": 6000
},
{
"loss": 1.2427,
"grad_norm": 9.809789657592773,
"learning_rate": 1.8421934501142425e-05,
"epoch": 0.3686551703125952,
"step": 6050
},
{
"loss": 1.2283,
"grad_norm": 7.107547760009766,
"learning_rate": 1.8574257425742576e-05,
"epoch": 0.3717019072573274,
"step": 6100
},
{
"loss": 1.2587,
"grad_norm": 10.349649429321289,
"learning_rate": 1.872658035034273e-05,
"epoch": 0.3747486442020596,
"step": 6150
},
{
"loss": 1.1766,
"grad_norm": 11.582871437072754,
"learning_rate": 1.887890327494288e-05,
"epoch": 0.3777953811467918,
"step": 6200
},
{
"loss": 1.1631,
"grad_norm": 6.336233139038086,
"learning_rate": 1.9031226199543035e-05,
"epoch": 0.38084211809152396,
"step": 6250
},
{
"loss": 1.1911,
"grad_norm": 6.148489475250244,
"learning_rate": 1.9183549124143186e-05,
"epoch": 0.3838888550362562,
"step": 6300
},
{
"loss": 1.1797,
"grad_norm": 11.499217987060547,
"learning_rate": 1.9332825590251336e-05,
"epoch": 0.38693559198098837,
"step": 6350
},
{
"loss": 1.1729,
"grad_norm": 11.640890121459961,
"learning_rate": 1.9485148514851486e-05,
"epoch": 0.38998232892572054,
"step": 6400
},
{
"loss": 1.095,
"grad_norm": 8.160388946533203,
"learning_rate": 1.9637471439451637e-05,
"epoch": 0.3930290658704527,
"step": 6450
},
{
"loss": 1.133,
"grad_norm": 9.691280364990234,
"learning_rate": 1.978979436405179e-05,
"epoch": 0.39607580281518495,
"step": 6500
},
{
"loss": 1.1122,
"grad_norm": 10.121310234069824,
"learning_rate": 1.9942117288651942e-05,
"epoch": 0.3991225397599171,
"step": 6550
},
{
"loss": 1.032,
"grad_norm": 7.770227432250977,
"learning_rate": 1.9995028984229053e-05,
"epoch": 0.4021692767046493,
"step": 6600
},
{
"loss": 1.0221,
"grad_norm": 6.209733963012695,
"learning_rate": 1.9987011216856556e-05,
"epoch": 0.40521601364938153,
"step": 6650
},
{
"loss": 0.9478,
"grad_norm": 15.20450496673584,
"learning_rate": 1.9978993449484058e-05,
"epoch": 0.4082627505941137,
"step": 6700
},
{
"loss": 1.0296,
"grad_norm": 13.067551612854004,
"learning_rate": 1.997097568211156e-05,
"epoch": 0.4113094875388459,
"step": 6750
},
{
"loss": 1.077,
"grad_norm": 8.253534317016602,
"learning_rate": 1.9962957914739063e-05,
"epoch": 0.41435622448357806,
"step": 6800
},
{
"loss": 0.9824,
"grad_norm": 9.505753517150879,
"learning_rate": 1.9954940147366566e-05,
"epoch": 0.4174029614283103,
"step": 6850
},
{
"loss": 0.9678,
"grad_norm": 7.164095401763916,
"learning_rate": 1.9946922379994068e-05,
"epoch": 0.42044969837304247,
"step": 6900
},
{
"loss": 0.9949,
"grad_norm": 10.617815971374512,
"learning_rate": 1.993890461262157e-05,
"epoch": 0.42349643531777464,
"step": 6950
},
{
"loss": 0.9452,
"grad_norm": 7.374587059020996,
"learning_rate": 1.9930886845249073e-05,
"epoch": 0.4265431722625069,
"step": 7000
},
{
"loss": 0.9886,
"grad_norm": 7.502238750457764,
"learning_rate": 1.9922869077876576e-05,
"epoch": 0.42958990920723905,
"step": 7050
},
{
"loss": 0.9514,
"grad_norm": 10.878878593444824,
"learning_rate": 1.9914851310504078e-05,
"epoch": 0.4326366461519712,
"step": 7100
},
{
"loss": 0.993,
"grad_norm": 11.223919868469238,
"learning_rate": 1.990683354313158e-05,
"epoch": 0.43568338309670346,
"step": 7150
},
{
"loss": 0.8666,
"grad_norm": 16.220924377441406,
"learning_rate": 1.9898815775759083e-05,
"epoch": 0.43873012004143563,
"step": 7200
},
{
"loss": 0.9327,
"grad_norm": 6.751389980316162,
"learning_rate": 1.9890798008386586e-05,
"epoch": 0.4417768569861678,
"step": 7250
},
{
"loss": 0.9716,
"grad_norm": 8.510977745056152,
"learning_rate": 1.9882780241014088e-05,
"epoch": 0.4448235939309,
"step": 7300
},
{
"loss": 0.9302,
"grad_norm": 10.246211051940918,
"learning_rate": 1.987476247364159e-05,
"epoch": 0.4478703308756322,
"step": 7350
},
{
"loss": 0.9257,
"grad_norm": 9.319009780883789,
"learning_rate": 1.9866744706269093e-05,
"epoch": 0.4509170678203644,
"step": 7400
},
{
"loss": 0.9944,
"grad_norm": 9.475403785705566,
"learning_rate": 1.98587269388966e-05,
"epoch": 0.45396380476509657,
"step": 7450
},
{
"loss": 0.8847,
"grad_norm": 9.286197662353516,
"learning_rate": 1.9850709171524098e-05,
"epoch": 0.4570105417098288,
"step": 7500
},
{
"loss": 0.8321,
"grad_norm": 7.356568336486816,
"learning_rate": 1.98426914041516e-05,
"epoch": 0.460057278654561,
"step": 7550
},
{
"loss": 0.7723,
"grad_norm": 8.750321388244629,
"learning_rate": 1.9834673636779107e-05,
"epoch": 0.46310401559929315,
"step": 7600
},
{
"loss": 0.8386,
"grad_norm": 9.985953330993652,
"learning_rate": 1.9826655869406606e-05,
"epoch": 0.4661507525440253,
"step": 7650
},
{
"loss": 0.7676,
"grad_norm": 11.913117408752441,
"learning_rate": 1.9818638102034108e-05,
"epoch": 0.46919748948875756,
"step": 7700
},
{
"loss": 0.8457,
"grad_norm": 12.286808967590332,
"learning_rate": 1.981062033466161e-05,
"epoch": 0.47224422643348973,
"step": 7750
},
{
"loss": 0.8349,
"grad_norm": 6.163942337036133,
"learning_rate": 1.9802602567289117e-05,
"epoch": 0.4752909633782219,
"step": 7800
},
{
"loss": 0.7985,
"grad_norm": 8.767670631408691,
"learning_rate": 1.9794584799916616e-05,
"epoch": 0.47833770032295414,
"step": 7850
},
{
"loss": 0.8379,
"grad_norm": 9.287251472473145,
"learning_rate": 1.9786567032544118e-05,
"epoch": 0.4813844372676863,
"step": 7900
},
{
"loss": 0.9086,
"grad_norm": 6.459810733795166,
"learning_rate": 1.9778549265171624e-05,
"epoch": 0.4844311742124185,
"step": 7950
},
{
"loss": 0.8463,
"grad_norm": 21.043193817138672,
"learning_rate": 1.9770531497799123e-05,
"epoch": 0.48747791115715067,
"step": 8000
},
{
"loss": 0.8329,
"grad_norm": 17.127490997314453,
"learning_rate": 1.9762513730426626e-05,
"epoch": 0.4905246481018829,
"step": 8050
},
{
"loss": 0.7992,
"grad_norm": 8.910788536071777,
"learning_rate": 1.975449596305413e-05,
"epoch": 0.4935713850466151,
"step": 8100
},
{
"loss": 0.7982,
"grad_norm": 13.819639205932617,
"learning_rate": 1.9746478195681634e-05,
"epoch": 0.49661812199134725,
"step": 8150
},
{
"loss": 0.7829,
"grad_norm": 15.758248329162598,
"learning_rate": 1.9738460428309133e-05,
"epoch": 0.4996648589360795,
"step": 8200
},
{
"loss": 0.8015,
"grad_norm": 4.369847774505615,
"learning_rate": 1.9730442660936636e-05,
"epoch": 0.5027115958808116,
"step": 8250
},
{
"loss": 0.8182,
"grad_norm": 9.60881519317627,
"learning_rate": 1.972242489356414e-05,
"epoch": 0.5057583328255438,
"step": 8300
},
{
"loss": 0.7888,
"grad_norm": 6.297778606414795,
"learning_rate": 1.9714567481539093e-05,
"epoch": 0.5088050697702761,
"step": 8350
},
{
"loss": 0.7118,
"grad_norm": 11.24323844909668,
"learning_rate": 1.9706549714166593e-05,
"epoch": 0.5118518067150082,
"step": 8400
},
{
"loss": 0.7485,
"grad_norm": 14.08312702178955,
"learning_rate": 1.96985319467941e-05,
"epoch": 0.5148985436597404,
"step": 8450
},
{
"loss": 0.7714,
"grad_norm": 6.063088417053223,
"learning_rate": 1.96905141794216e-05,
"epoch": 0.5179452806044726,
"step": 8500
},
{
"loss": 0.7856,
"grad_norm": 11.223748207092285,
"learning_rate": 1.9682656767396553e-05,
"epoch": 0.5209920175492048,
"step": 8550
},
{
"loss": 0.725,
"grad_norm": 9.955530166625977,
"learning_rate": 1.9674639000024055e-05,
"epoch": 0.524038754493937,
"step": 8600
},
{
"loss": 0.6885,
"grad_norm": 12.986407279968262,
"learning_rate": 1.9666621232651558e-05,
"epoch": 0.5270854914386692,
"step": 8650
},
{
"loss": 0.7431,
"grad_norm": 13.77481460571289,
"learning_rate": 1.965860346527906e-05,
"epoch": 0.5301322283834013,
"step": 8700
},
{
"loss": 0.6785,
"grad_norm": 9.711028099060059,
"learning_rate": 1.9650585697906563e-05,
"epoch": 0.5331789653281336,
"step": 8750
},
{
"loss": 0.6916,
"grad_norm": 13.084221839904785,
"learning_rate": 1.9642567930534065e-05,
"epoch": 0.5362257022728658,
"step": 8800
},
{
"loss": 0.7424,
"grad_norm": 12.57776927947998,
"learning_rate": 1.9634550163161568e-05,
"epoch": 0.5392724392175979,
"step": 8850
},
{
"loss": 0.7015,
"grad_norm": 15.594993591308594,
"learning_rate": 1.962653239578907e-05,
"epoch": 0.5423191761623302,
"step": 8900
},
{
"loss": 0.6371,
"grad_norm": 13.37274169921875,
"learning_rate": 1.9618514628416573e-05,
"epoch": 0.5453659131070623,
"step": 8950
},
{
"loss": 0.6286,
"grad_norm": 6.84214973449707,
"learning_rate": 1.9610496861044075e-05,
"epoch": 0.5484126500517945,
"step": 9000
},
{
"loss": 0.7112,
"grad_norm": 13.600686073303223,
"learning_rate": 1.9602479093671578e-05,
"epoch": 0.5514593869965267,
"step": 9050
},
{
"loss": 0.7127,
"grad_norm": 4.71443510055542,
"learning_rate": 1.959446132629908e-05,
"epoch": 0.5545061239412589,
"step": 9100
},
{
"loss": 0.8043,
"grad_norm": 14.699666023254395,
"learning_rate": 1.9586443558926583e-05,
"epoch": 0.5575528608859911,
"step": 9150
},
{
"loss": 0.718,
"grad_norm": 7.46713924407959,
"learning_rate": 1.9578425791554085e-05,
"epoch": 0.5605995978307233,
"step": 9200
},
{
"loss": 0.6648,
"grad_norm": 9.12059497833252,
"learning_rate": 1.9570408024181588e-05,
"epoch": 0.5636463347754554,
"step": 9250
},
{
"loss": 0.7008,
"grad_norm": 13.465869903564453,
"learning_rate": 1.956239025680909e-05,
"epoch": 0.5666930717201877,
"step": 9300
},
{
"loss": 0.606,
"grad_norm": 7.796365261077881,
"learning_rate": 1.9554372489436593e-05,
"epoch": 0.5697398086649199,
"step": 9350
},
{
"loss": 0.6987,
"grad_norm": 12.26785945892334,
"learning_rate": 1.9546354722064095e-05,
"epoch": 0.572786545609652,
"step": 9400
},
{
"loss": 0.7194,
"grad_norm": 11.7417573928833,
"learning_rate": 1.9538336954691598e-05,
"epoch": 0.5758332825543843,
"step": 9450
},
{
"loss": 0.6301,
"grad_norm": 9.008028984069824,
"learning_rate": 1.95303191873191e-05,
"epoch": 0.5788800194991165,
"step": 9500
},
{
"loss": 0.662,
"grad_norm": 5.153985500335693,
"learning_rate": 1.9522301419946603e-05,
"epoch": 0.5819267564438486,
"step": 9550
},
{
"loss": 0.7229,
"grad_norm": 10.312999725341797,
"learning_rate": 1.9514283652574105e-05,
"epoch": 0.5849734933885808,
"step": 9600
},
{
"loss": 0.608,
"grad_norm": 8.001009941101074,
"learning_rate": 1.9506265885201608e-05,
"epoch": 0.5880202303333131,
"step": 9650
},
{
"loss": 0.6985,
"grad_norm": 4.030726432800293,
"learning_rate": 1.949824811782911e-05,
"epoch": 0.5910669672780452,
"step": 9700
},
{
"loss": 0.64,
"grad_norm": 8.056360244750977,
"learning_rate": 1.9490230350456613e-05,
"epoch": 0.5941137042227774,
"step": 9750
},
{
"loss": 0.6468,
"grad_norm": 8.479045867919922,
"learning_rate": 1.9482212583084115e-05,
"epoch": 0.5971604411675095,
"step": 9800
},
{
"loss": 0.6517,
"grad_norm": 7.774533271789551,
"learning_rate": 1.9474194815711618e-05,
"epoch": 0.6002071781122418,
"step": 9850
},
{
"loss": 0.7232,
"grad_norm": 19.166425704956055,
"learning_rate": 1.946617704833912e-05,
"epoch": 0.603253915056974,
"step": 9900
},
{
"loss": 0.7264,
"grad_norm": 7.891358375549316,
"learning_rate": 1.9458159280966623e-05,
"epoch": 0.6063006520017061,
"step": 9950
},
{
"loss": 0.6401,
"grad_norm": 11.431697845458984,
"learning_rate": 1.9450141513594125e-05,
"epoch": 0.6093473889464384,
"step": 10000
},
{
"loss": 0.6115,
"grad_norm": 8.17441177368164,
"learning_rate": 1.9442123746221628e-05,
"epoch": 0.6123941258911706,
"step": 10050
},
{
"loss": 0.7269,
"grad_norm": 5.700364589691162,
"learning_rate": 1.943426633419658e-05,
"epoch": 0.6154408628359027,
"step": 10100
},
{
"loss": 0.6171,
"grad_norm": 7.469681262969971,
"learning_rate": 1.9426248566824086e-05,
"epoch": 0.618487599780635,
"step": 10150
},
{
"loss": 0.6165,
"grad_norm": 16.24020004272461,
"learning_rate": 1.9418230799451588e-05,
"epoch": 0.6215343367253672,
"step": 10200
},
{
"loss": 0.6107,
"grad_norm": 10.644442558288574,
"learning_rate": 1.9410213032079087e-05,
"epoch": 0.6245810736700993,
"step": 10250
},
{
"loss": 0.6098,
"grad_norm": 12.096433639526367,
"learning_rate": 1.940219526470659e-05,
"epoch": 0.6276278106148315,
"step": 10300
},
{
"loss": 0.6302,
"grad_norm": 3.781506299972534,
"learning_rate": 1.9394177497334096e-05,
"epoch": 0.6306745475595638,
"step": 10350
},
{
"loss": 0.613,
"grad_norm": 6.667350769042969,
"learning_rate": 1.9386159729961595e-05,
"epoch": 0.6337212845042959,
"step": 10400
},
{
"loss": 0.5741,
"grad_norm": 11.330256462097168,
"learning_rate": 1.9378141962589097e-05,
"epoch": 0.6367680214490281,
"step": 10450
},
{
"loss": 0.6214,
"grad_norm": 7.193487644195557,
"learning_rate": 1.9370124195216603e-05,
"epoch": 0.6398147583937602,
"step": 10500
},
{
"loss": 0.6421,
"grad_norm": 6.456230640411377,
"learning_rate": 1.9362106427844106e-05,
"epoch": 0.6428614953384925,
"step": 10550
},
{
"loss": 0.6122,
"grad_norm": 14.182371139526367,
"learning_rate": 1.9354088660471605e-05,
"epoch": 0.6459082322832247,
"step": 10600
},
{
"loss": 0.5761,
"grad_norm": 11.199743270874023,
"learning_rate": 1.934607089309911e-05,
"epoch": 0.6489549692279568,
"step": 10650
},
{
"loss": 0.6089,
"grad_norm": 11.449539184570312,
"learning_rate": 1.9338053125726613e-05,
"epoch": 0.652001706172689,
"step": 10700
},
{
"loss": 0.6511,
"grad_norm": 9.607653617858887,
"learning_rate": 1.9330035358354112e-05,
"epoch": 0.6550484431174213,
"step": 10750
},
{
"loss": 0.5871,
"grad_norm": 17.837121963500977,
"learning_rate": 1.9322017590981615e-05,
"epoch": 0.6580951800621534,
"step": 10800
},
{
"loss": 0.552,
"grad_norm": 8.2896728515625,
"learning_rate": 1.931399982360912e-05,
"epoch": 0.6611419170068856,
"step": 10850
},
{
"loss": 0.6012,
"grad_norm": 5.957419395446777,
"learning_rate": 1.9305982056236623e-05,
"epoch": 0.6641886539516179,
"step": 10900
},
{
"loss": 0.5957,
"grad_norm": 14.288383483886719,
"learning_rate": 1.9297964288864122e-05,
"epoch": 0.66723539089635,
"step": 10950
},
{
"loss": 0.5723,
"grad_norm": 6.8989787101745605,
"learning_rate": 1.928994652149163e-05,
"epoch": 0.6702821278410822,
"step": 11000
},
{
"loss": 0.6045,
"grad_norm": 10.084901809692383,
"learning_rate": 1.928208910946658e-05,
"epoch": 0.6733288647858144,
"step": 11050
},
{
"loss": 0.5012,
"grad_norm": 8.228680610656738,
"learning_rate": 1.9274071342094083e-05,
"epoch": 0.6763756017305466,
"step": 11100
},
{
"loss": 0.5626,
"grad_norm": 13.005243301391602,
"learning_rate": 1.9266053574721585e-05,
"epoch": 0.6794223386752788,
"step": 11150
},
{
"loss": 0.5698,
"grad_norm": 12.052207946777344,
"learning_rate": 1.9258035807349088e-05,
"epoch": 0.682469075620011,
"step": 11200
},
{
"loss": 0.5187,
"grad_norm": 6.421981334686279,
"learning_rate": 1.925001803997659e-05,
"epoch": 0.6855158125647431,
"step": 11250
},
{
"loss": 0.567,
"grad_norm": 13.395944595336914,
"learning_rate": 1.9242000272604093e-05,
"epoch": 0.6885625495094754,
"step": 11300
},
{
"loss": 0.5531,
"grad_norm": 10.468594551086426,
"learning_rate": 1.9233982505231595e-05,
"epoch": 0.6916092864542075,
"step": 11350
},
{
"loss": 0.5889,
"grad_norm": 3.9035274982452393,
"learning_rate": 1.9225964737859098e-05,
"epoch": 0.6946560233989397,
"step": 11400
},
{
"loss": 0.5482,
"grad_norm": 11.460153579711914,
"learning_rate": 1.92179469704866e-05,
"epoch": 0.697702760343672,
"step": 11450
},
{
"loss": 0.5561,
"grad_norm": 4.9930338859558105,
"learning_rate": 1.9209929203114103e-05,
"epoch": 0.7007494972884041,
"step": 11500
},
{
"loss": 0.4342,
"grad_norm": 6.652237415313721,
"learning_rate": 1.9201911435741605e-05,
"epoch": 0.7037962342331363,
"step": 11550
},
{
"loss": 0.5199,
"grad_norm": 10.926090240478516,
"learning_rate": 1.9193893668369108e-05,
"epoch": 0.7068429711778685,
"step": 11600
},
{
"loss": 0.5735,
"grad_norm": 18.5546932220459,
"learning_rate": 1.918587590099661e-05,
"epoch": 0.7098897081226007,
"step": 11650
},
{
"loss": 0.5977,
"grad_norm": 15.148147583007812,
"learning_rate": 1.9177858133624113e-05,
"epoch": 0.7129364450673329,
"step": 11700
},
{
"loss": 0.6166,
"grad_norm": 4.6596221923828125,
"learning_rate": 1.9169840366251615e-05,
"epoch": 0.7159831820120651,
"step": 11750
},
{
"loss": 0.5409,
"grad_norm": 5.004631996154785,
"learning_rate": 1.9161822598879118e-05,
"epoch": 0.7190299189567972,
"step": 11800
},
{
"loss": 0.5211,
"grad_norm": 9.570045471191406,
"learning_rate": 1.915380483150662e-05,
"epoch": 0.7220766559015295,
"step": 11850
},
{
"loss": 0.4908,
"grad_norm": 12.19480037689209,
"learning_rate": 1.9145787064134123e-05,
"epoch": 0.7251233928462617,
"step": 11900
},
{
"loss": 0.5463,
"grad_norm": 13.55962085723877,
"learning_rate": 1.9137769296761625e-05,
"epoch": 0.7281701297909938,
"step": 11950
},
{
"loss": 0.5638,
"grad_norm": 11.181888580322266,
"learning_rate": 1.9129751529389128e-05,
"epoch": 0.7312168667357261,
"step": 12000
},
{
"loss": 0.5757,
"grad_norm": 12.415191650390625,
"learning_rate": 1.912173376201663e-05,
"epoch": 0.7342636036804582,
"step": 12050
},
{
"loss": 0.5435,
"grad_norm": 7.7507524490356445,
"learning_rate": 1.9113715994644133e-05,
"epoch": 0.7373103406251904,
"step": 12100
},
{
"loss": 0.5744,
"grad_norm": 14.383264541625977,
"learning_rate": 1.9105698227271635e-05,
"epoch": 0.7403570775699226,
"step": 12150
},
{
"loss": 0.5429,
"grad_norm": 11.143241882324219,
"learning_rate": 1.9097680459899138e-05,
"epoch": 0.7434038145146548,
"step": 12200
},
{
"loss": 0.5141,
"grad_norm": 7.639148712158203,
"learning_rate": 1.908966269252664e-05,
"epoch": 0.746450551459387,
"step": 12250
},
{
"loss": 0.5921,
"grad_norm": 9.034982681274414,
"learning_rate": 1.9081644925154143e-05,
"epoch": 0.7494972884041192,
"step": 12300
},
{
"loss": 0.5657,
"grad_norm": 17.773897171020508,
"learning_rate": 1.9073627157781645e-05,
"epoch": 0.7525440253488513,
"step": 12350
},
{
"loss": 0.467,
"grad_norm": 3.3905394077301025,
"learning_rate": 1.9065609390409148e-05,
"epoch": 0.7555907622935836,
"step": 12400
},
{
"loss": 0.5315,
"grad_norm": 10.708720207214355,
"learning_rate": 1.905759162303665e-05,
"epoch": 0.7586374992383158,
"step": 12450
},
{
"loss": 0.4799,
"grad_norm": 5.747498989105225,
"learning_rate": 1.9049573855664153e-05,
"epoch": 0.7616842361830479,
"step": 12500
},
{
"loss": 0.4707,
"grad_norm": 13.182746887207031,
"learning_rate": 1.9041556088291655e-05,
"epoch": 0.7647309731277802,
"step": 12550
},
{
"loss": 0.4664,
"grad_norm": 6.879668235778809,
"learning_rate": 1.9033538320919158e-05,
"epoch": 0.7677777100725124,
"step": 12600
},
{
"loss": 0.4323,
"grad_norm": 7.6186933517456055,
"learning_rate": 1.902552055354666e-05,
"epoch": 0.7708244470172445,
"step": 12650
},
{
"loss": 0.4908,
"grad_norm": 10.26600170135498,
"learning_rate": 1.9017502786174166e-05,
"epoch": 0.7738711839619767,
"step": 12700
},
{
"loss": 0.5259,
"grad_norm": 7.036660671234131,
"learning_rate": 1.9009485018801665e-05,
"epoch": 0.776917920906709,
"step": 12750
},
{
"loss": 0.4806,
"grad_norm": 9.241826057434082,
"learning_rate": 1.9001467251429168e-05,
"epoch": 0.7799646578514411,
"step": 12800
},
{
"loss": 0.451,
"grad_norm": 5.064478874206543,
"learning_rate": 1.899344948405667e-05,
"epoch": 0.7830113947961733,
"step": 12850
},
{
"loss": 0.5564,
"grad_norm": 10.820523262023926,
"learning_rate": 1.8985431716684173e-05,
"epoch": 0.7860581317409054,
"step": 12900
},
{
"loss": 0.5595,
"grad_norm": 11.908926010131836,
"learning_rate": 1.8977413949311675e-05,
"epoch": 0.7891048686856377,
"step": 12950
},
{
"loss": 0.5021,
"grad_norm": 8.049078941345215,
"learning_rate": 1.8969396181939178e-05,
"epoch": 0.7921516056303699,
"step": 13000
},
{
"loss": 0.4869,
"grad_norm": 7.749687671661377,
"learning_rate": 1.8961378414566684e-05,
"epoch": 0.795198342575102,
"step": 13050
},
{
"loss": 0.5034,
"grad_norm": 12.783843040466309,
"learning_rate": 1.8953360647194183e-05,
"epoch": 0.7982450795198343,
"step": 13100
},
{
"loss": 0.5797,
"grad_norm": 6.763021469116211,
"learning_rate": 1.8945342879821685e-05,
"epoch": 0.8012918164645665,
"step": 13150
},
{
"loss": 0.5427,
"grad_norm": 10.710930824279785,
"learning_rate": 1.893732511244919e-05,
"epoch": 0.8043385534092986,
"step": 13200
},
{
"loss": 0.4173,
"grad_norm": 9.847426414489746,
"learning_rate": 1.892930734507669e-05,
"epoch": 0.8073852903540308,
"step": 13250
},
{
"loss": 0.5047,
"grad_norm": 9.54844856262207,
"learning_rate": 1.8921289577704193e-05,
"epoch": 0.8104320272987631,
"step": 13300
},
{
"loss": 0.5448,
"grad_norm": 11.02835750579834,
"learning_rate": 1.8913271810331695e-05,
"epoch": 0.8134787642434952,
"step": 13350
},
{
"loss": 0.5698,
"grad_norm": 14.037369728088379,
"learning_rate": 1.89052540429592e-05,
"epoch": 0.8165255011882274,
"step": 13400
},
{
"loss": 0.5461,
"grad_norm": 8.469470024108887,
"learning_rate": 1.88972362755867e-05,
"epoch": 0.8195722381329597,
"step": 13450
},
{
"loss": 0.4949,
"grad_norm": 10.204724311828613,
"learning_rate": 1.8889218508214203e-05,
"epoch": 0.8226189750776918,
"step": 13500
},
{
"loss": 0.5325,
"grad_norm": 14.964487075805664,
"learning_rate": 1.888120074084171e-05,
"epoch": 0.825665712022424,
"step": 13550
},
{
"loss": 0.4322,
"grad_norm": 9.92317008972168,
"learning_rate": 1.8873182973469208e-05,
"epoch": 0.8287124489671561,
"step": 13600
},
{
"loss": 0.5095,
"grad_norm": 3.5198421478271484,
"learning_rate": 1.886516520609671e-05,
"epoch": 0.8317591859118884,
"step": 13650
},
{
"loss": 0.5948,
"grad_norm": 10.616870880126953,
"learning_rate": 1.8857147438724216e-05,
"epoch": 0.8348059228566206,
"step": 13700
},
{
"loss": 0.3746,
"grad_norm": 10.42710018157959,
"learning_rate": 1.884912967135172e-05,
"epoch": 0.8378526598013527,
"step": 13750
},
{
"loss": 0.5017,
"grad_norm": 14.937359809875488,
"learning_rate": 1.8841111903979218e-05,
"epoch": 0.8408993967460849,
"step": 13800
},
{
"loss": 0.4297,
"grad_norm": 10.7739896774292,
"learning_rate": 1.8833094136606724e-05,
"epoch": 0.8439461336908172,
"step": 13850
},
{
"loss": 0.5154,
"grad_norm": 12.550127983093262,
"learning_rate": 1.8825076369234226e-05,
"epoch": 0.8469928706355493,
"step": 13900
},
{
"loss": 0.5466,
"grad_norm": 13.98280143737793,
"learning_rate": 1.8817058601861725e-05,
"epoch": 0.8500396075802815,
"step": 13950
},
{
"loss": 0.456,
"grad_norm": 11.7831392288208,
"learning_rate": 1.8809040834489228e-05,
"epoch": 0.8530863445250138,
"step": 14000
},
{
"loss": 0.4546,
"grad_norm": 4.859154224395752,
"learning_rate": 1.8801023067116734e-05,
"epoch": 0.8561330814697459,
"step": 14050
},
{
"loss": 0.4359,
"grad_norm": 11.597343444824219,
"learning_rate": 1.8793005299744236e-05,
"epoch": 0.8591798184144781,
"step": 14100
},
{
"loss": 0.4884,
"grad_norm": 7.369439125061035,
"learning_rate": 1.8784987532371735e-05,
"epoch": 0.8622265553592103,
"step": 14150
},
{
"loss": 0.5861,
"grad_norm": 8.557465553283691,
"learning_rate": 1.877696976499924e-05,
"epoch": 0.8652732923039425,
"step": 14200
},
{
"loss": 0.421,
"grad_norm": 12.983498573303223,
"learning_rate": 1.8768951997626744e-05,
"epoch": 0.8683200292486747,
"step": 14250
},
{
"loss": 0.3883,
"grad_norm": 2.3741507530212402,
"learning_rate": 1.8760934230254243e-05,
"epoch": 0.8713667661934069,
"step": 14300
},
{
"loss": 0.5239,
"grad_norm": 12.974597930908203,
"learning_rate": 1.875291646288175e-05,
"epoch": 0.874413503138139,
"step": 14350
},
{
"loss": 0.474,
"grad_norm": 6.931816577911377,
"learning_rate": 1.874489869550925e-05,
"epoch": 0.8774602400828713,
"step": 14400
},
{
"loss": 0.5097,
"grad_norm": 9.687259674072266,
"learning_rate": 1.8736880928136754e-05,
"epoch": 0.8805069770276034,
"step": 14450
},
{
"loss": 0.4891,
"grad_norm": 5.1861724853515625,
"learning_rate": 1.8728863160764253e-05,
"epoch": 0.8835537139723356,
"step": 14500
},
{
"loss": 0.4742,
"grad_norm": 4.1977081298828125,
"learning_rate": 1.872084539339176e-05,
"epoch": 0.8866004509170679,
"step": 14550
},
{
"loss": 0.503,
"grad_norm": 13.97655200958252,
"learning_rate": 1.871282762601926e-05,
"epoch": 0.8896471878618,
"step": 14600
},
{
"loss": 0.4944,
"grad_norm": 10.400720596313477,
"learning_rate": 1.870480985864676e-05,
"epoch": 0.8926939248065322,
"step": 14650
},
{
"loss": 0.4731,
"grad_norm": 14.06461238861084,
"learning_rate": 1.8696792091274266e-05,
"epoch": 0.8957406617512644,
"step": 14700
},
{
"loss": 0.4325,
"grad_norm": 12.592815399169922,
"learning_rate": 1.868877432390177e-05,
"epoch": 0.8987873986959966,
"step": 14750
},
{
"loss": 0.4473,
"grad_norm": 8.40234375,
"learning_rate": 1.868075655652927e-05,
"epoch": 0.9018341356407288,
"step": 14800
},
{
"loss": 0.5028,
"grad_norm": 8.714736938476562,
"learning_rate": 1.8672738789156774e-05,
"epoch": 0.904880872585461,
"step": 14850
},
{
"loss": 0.4343,
"grad_norm": 14.819173812866211,
"learning_rate": 1.8664721021784276e-05,
"epoch": 0.9079276095301931,
"step": 14900
},
{
"loss": 0.4869,
"grad_norm": 9.786388397216797,
"learning_rate": 1.865670325441178e-05,
"epoch": 0.9109743464749254,
"step": 14950
},
{
"loss": 0.4581,
"grad_norm": 12.583029747009277,
"learning_rate": 1.864868548703928e-05,
"epoch": 0.9140210834196576,
"step": 15000
},
{
"loss": 0.4507,
"grad_norm": 6.247103691101074,
"learning_rate": 1.8640667719666784e-05,
"epoch": 0.9170678203643897,
"step": 15050
},
{
"loss": 0.4805,
"grad_norm": 8.440277099609375,
"learning_rate": 1.8632649952294286e-05,
"epoch": 0.920114557309122,
"step": 15100
},
{
"loss": 0.4615,
"grad_norm": 31.09827995300293,
"learning_rate": 1.862463218492179e-05,
"epoch": 0.9231612942538541,
"step": 15150
},
{
"loss": 0.4565,
"grad_norm": 5.867329120635986,
"learning_rate": 1.861661441754929e-05,
"epoch": 0.9262080311985863,
"step": 15200
},
{
"loss": 0.4019,
"grad_norm": 11.316119194030762,
"learning_rate": 1.8608596650176794e-05,
"epoch": 0.9292547681433185,
"step": 15250
},
{
"loss": 0.4647,
"grad_norm": 28.179372787475586,
"learning_rate": 1.8600578882804296e-05,
"epoch": 0.9323015050880507,
"step": 15300
},
{
"loss": 0.4663,
"grad_norm": 7.8867106437683105,
"learning_rate": 1.85925611154318e-05,
"epoch": 0.9353482420327829,
"step": 15350
},
{
"loss": 0.4422,
"grad_norm": 13.38543701171875,
"learning_rate": 1.85845433480593e-05,
"epoch": 0.9383949789775151,
"step": 15400
},
{
"loss": 0.4426,
"grad_norm": 6.577428817749023,
"learning_rate": 1.8576525580686804e-05,
"epoch": 0.9414417159222472,
"step": 15450
},
{
"loss": 0.4622,
"grad_norm": 11.758441925048828,
"learning_rate": 1.8568507813314306e-05,
"epoch": 0.9444884528669795,
"step": 15500
},
{
"loss": 0.4605,
"grad_norm": 8.182458877563477,
"learning_rate": 1.856049004594181e-05,
"epoch": 0.9475351898117117,
"step": 15550
},
{
"loss": 0.3987,
"grad_norm": 14.633559226989746,
"learning_rate": 1.855247227856931e-05,
"epoch": 0.9505819267564438,
"step": 15600
},
{
"loss": 0.4348,
"grad_norm": 9.851497650146484,
"learning_rate": 1.8544454511196814e-05,
"epoch": 0.953628663701176,
"step": 15650
},
{
"loss": 0.4501,
"grad_norm": 8.412302017211914,
"learning_rate": 1.8536436743824316e-05,
"epoch": 0.9566754006459083,
"step": 15700
},
{
"loss": 0.4757,
"grad_norm": 2.4255619049072266,
"learning_rate": 1.852841897645182e-05,
"epoch": 0.9597221375906404,
"step": 15750
},
{
"loss": 0.4036,
"grad_norm": 18.82175636291504,
"learning_rate": 1.852040120907932e-05,
"epoch": 0.9627688745353726,
"step": 15800
},
{
"loss": 0.3922,
"grad_norm": 17.99393081665039,
"learning_rate": 1.8512383441706824e-05,
"epoch": 0.9658156114801049,
"step": 15850
},
{
"loss": 0.4478,
"grad_norm": 9.306086540222168,
"learning_rate": 1.8504365674334326e-05,
"epoch": 0.968862348424837,
"step": 15900
},
{
"loss": 0.4588,
"grad_norm": 2.2692463397979736,
"learning_rate": 1.849634790696183e-05,
"epoch": 0.9719090853695692,
"step": 15950
},
{
"loss": 0.4518,
"grad_norm": 11.183533668518066,
"learning_rate": 1.848833013958933e-05,
"epoch": 0.9749558223143013,
"step": 16000
},
{
"loss": 0.4361,
"grad_norm": 20.365339279174805,
"learning_rate": 1.8480312372216834e-05,
"epoch": 0.9780025592590336,
"step": 16050
},
{
"loss": 0.4783,
"grad_norm": 9.099688529968262,
"learning_rate": 1.8472294604844336e-05,
"epoch": 0.9810492962037658,
"step": 16100
},
{
"loss": 0.4841,
"grad_norm": 7.3363871574401855,
"learning_rate": 1.846427683747184e-05,
"epoch": 0.9840960331484979,
"step": 16150
},
{
"loss": 0.4246,
"grad_norm": 18.804698944091797,
"learning_rate": 1.845625907009934e-05,
"epoch": 0.9871427700932301,
"step": 16200
},
{
"loss": 0.3873,
"grad_norm": 7.1848578453063965,
"learning_rate": 1.8448241302726844e-05,
"epoch": 0.9901895070379624,
"step": 16250
},
{
"loss": 0.4886,
"grad_norm": 13.017084121704102,
"learning_rate": 1.8440223535354346e-05,
"epoch": 0.9932362439826945,
"step": 16300
},
{
"loss": 0.4189,
"grad_norm": 11.633439064025879,
"learning_rate": 1.843220576798185e-05,
"epoch": 0.9962829809274267,
"step": 16350
},
{
"loss": 0.3753,
"grad_norm": 11.07437801361084,
"learning_rate": 1.842418800060935e-05,
"epoch": 0.999329717872159,
"step": 16400
},
{
"eval_loss": 0.2600756585597992,
"eval_accuracy": 0.9319280205655527,
"eval_f1_macro": 0.9312117795748152,
"eval_top5_accuracy": 0.99101970865467,
"eval_runtime": 220.5655,
"eval_samples_per_second": 132.274,
"eval_steps_per_second": 8.27,
"epoch": 1.0,
"step": 16411
},
{
"loss": 0.4527,
"grad_norm": 14.677661895751953,
"learning_rate": 1.8416170233236854e-05,
"epoch": 1.0023764548168912,
"step": 16450
},
{
"loss": 0.3592,
"grad_norm": 15.316341400146484,
"learning_rate": 1.8408152465864357e-05,
"epoch": 1.0054231917616232,
"step": 16500
},
{
"loss": 0.4592,
"grad_norm": 15.413875579833984,
"learning_rate": 1.840013469849186e-05,
"epoch": 1.0084699287063554,
"step": 16550
},
{
"loss": 0.3445,
"grad_norm": 7.590103626251221,
"learning_rate": 1.839211693111936e-05,
"epoch": 1.0115166656510877,
"step": 16600
},
{
"loss": 0.3675,
"grad_norm": 4.770218849182129,
"learning_rate": 1.8384099163746864e-05,
"epoch": 1.01456340259582,
"step": 16650
},
{
"loss": 0.3907,
"grad_norm": 12.107449531555176,
"learning_rate": 1.8376081396374367e-05,
"epoch": 1.0176101395405521,
"step": 16700
},
{
"loss": 0.4257,
"grad_norm": 7.213541507720947,
"learning_rate": 1.836806362900187e-05,
"epoch": 1.0206568764852844,
"step": 16750
},
{
"loss": 0.3736,
"grad_norm": 10.05327033996582,
"learning_rate": 1.836004586162937e-05,
"epoch": 1.0237036134300164,
"step": 16800
},
{
"loss": 0.4315,
"grad_norm": 18.48215103149414,
"learning_rate": 1.8352028094256874e-05,
"epoch": 1.0267503503747486,
"step": 16850
},
{
"loss": 0.3958,
"grad_norm": 12.158247947692871,
"learning_rate": 1.834401032688438e-05,
"epoch": 1.0297970873194808,
"step": 16900
},
{
"loss": 0.3613,
"grad_norm": 6.227318286895752,
"learning_rate": 1.833599255951188e-05,
"epoch": 1.032843824264213,
"step": 16950
},
{
"loss": 0.3805,
"grad_norm": 5.82094144821167,
"learning_rate": 1.832797479213938e-05,
"epoch": 1.0358905612089453,
"step": 17000
},
{
"loss": 0.3928,
"grad_norm": 9.13658332824707,
"learning_rate": 1.8319957024766884e-05,
"epoch": 1.0389372981536775,
"step": 17050
},
{
"loss": 0.3804,
"grad_norm": 6.659351348876953,
"learning_rate": 1.8311939257394387e-05,
"epoch": 1.0419840350984095,
"step": 17100
},
{
"loss": 0.4112,
"grad_norm": 16.467992782592773,
"learning_rate": 1.830392149002189e-05,
"epoch": 1.0450307720431418,
"step": 17150
},
{
"loss": 0.3925,
"grad_norm": 3.2631635665893555,
"learning_rate": 1.829590372264939e-05,
"epoch": 1.048077508987874,
"step": 17200
},
{
"loss": 0.4082,
"grad_norm": 18.036739349365234,
"learning_rate": 1.8287885955276897e-05,
"epoch": 1.0511242459326062,
"step": 17250
},
{
"loss": 0.3427,
"grad_norm": 9.820805549621582,
"learning_rate": 1.8279868187904397e-05,
"epoch": 1.0541709828773385,
"step": 17300
},
{
"loss": 0.3804,
"grad_norm": 5.950634002685547,
"learning_rate": 1.82718504205319e-05,
"epoch": 1.0572177198220705,
"step": 17350
},
{
"loss": 0.3658,
"grad_norm": 8.124308586120605,
"learning_rate": 1.8263832653159405e-05,
"epoch": 1.0602644567668027,
"step": 17400
},
{
"loss": 0.3908,
"grad_norm": 8.845989227294922,
"learning_rate": 1.8255814885786904e-05,
"epoch": 1.063311193711535,
"step": 17450
},
{
"loss": 0.2919,
"grad_norm": 5.308241367340088,
"learning_rate": 1.8247797118414407e-05,
"epoch": 1.0663579306562672,
"step": 17500
},
{
"loss": 0.4086,
"grad_norm": 10.882274627685547,
"learning_rate": 1.8239779351041912e-05,
"epoch": 1.0694046676009994,
"step": 17550
},
{
"loss": 0.4138,
"grad_norm": 9.368940353393555,
"learning_rate": 1.8231761583669415e-05,
"epoch": 1.0724514045457316,
"step": 17600
},
{
"loss": 0.3701,
"grad_norm": 12.684194564819336,
"learning_rate": 1.8223743816296914e-05,
"epoch": 1.0754981414904636,
"step": 17650
},
{
"loss": 0.3606,
"grad_norm": 9.783271789550781,
"learning_rate": 1.8215726048924417e-05,
"epoch": 1.0785448784351959,
"step": 17700
},
{
"loss": 0.4143,
"grad_norm": 8.15735149383545,
"learning_rate": 1.8207708281551922e-05,
"epoch": 1.081591615379928,
"step": 17750
},
{
"loss": 0.4024,
"grad_norm": 9.71022891998291,
"learning_rate": 1.819969051417942e-05,
"epoch": 1.0846383523246603,
"step": 17800
},
{
"loss": 0.4063,
"grad_norm": 10.642499923706055,
"learning_rate": 1.8191672746806924e-05,
"epoch": 1.0876850892693926,
"step": 17850
},
{
"loss": 0.4058,
"grad_norm": 14.741069793701172,
"learning_rate": 1.818365497943443e-05,
"epoch": 1.0907318262141246,
"step": 17900
},
{
"loss": 0.3989,
"grad_norm": 24.629985809326172,
"learning_rate": 1.8175637212061932e-05,
"epoch": 1.0937785631588568,
"step": 17950
},
{
"loss": 0.4209,
"grad_norm": 14.127384185791016,
"learning_rate": 1.816761944468943e-05,
"epoch": 1.096825300103589,
"step": 18000
},
{
"loss": 0.3152,
"grad_norm": 10.986865043640137,
"learning_rate": 1.8159601677316937e-05,
"epoch": 1.0998720370483213,
"step": 18050
},
{
"loss": 0.3516,
"grad_norm": 17.347911834716797,
"learning_rate": 1.815158390994444e-05,
"epoch": 1.1029187739930535,
"step": 18100
},
{
"loss": 0.4015,
"grad_norm": 6.95743989944458,
"learning_rate": 1.814356614257194e-05,
"epoch": 1.1059655109377857,
"step": 18150
},
{
"loss": 0.4086,
"grad_norm": 7.002768039703369,
"learning_rate": 1.813554837519944e-05,
"epoch": 1.1090122478825177,
"step": 18200
},
{
"loss": 0.2964,
"grad_norm": 10.233739852905273,
"learning_rate": 1.8127530607826948e-05,
"epoch": 1.11205898482725,
"step": 18250
},
{
"loss": 0.4285,
"grad_norm": 9.420923233032227,
"learning_rate": 1.811951284045445e-05,
"epoch": 1.1151057217719822,
"step": 18300
},
{
"loss": 0.3287,
"grad_norm": 7.177124500274658,
"learning_rate": 1.811149507308195e-05,
"epoch": 1.1181524587167144,
"step": 18350
},
{
"loss": 0.3273,
"grad_norm": 5.678183555603027,
"learning_rate": 1.8103477305709455e-05,
"epoch": 1.1211991956614467,
"step": 18400
},
{
"loss": 0.2935,
"grad_norm": 7.959808826446533,
"learning_rate": 1.8095459538336958e-05,
"epoch": 1.1242459326061787,
"step": 18450
},
{
"loss": 0.3315,
"grad_norm": 9.412012100219727,
"learning_rate": 1.8087441770964457e-05,
"epoch": 1.127292669550911,
"step": 18500
},
{
"loss": 0.3592,
"grad_norm": 6.331026077270508,
"learning_rate": 1.8079424003591963e-05,
"epoch": 1.1303394064956431,
"step": 18550
},
{
"loss": 0.3696,
"grad_norm": 13.279031753540039,
"learning_rate": 1.8071406236219465e-05,
"epoch": 1.1333861434403754,
"step": 18600
},
{
"loss": 0.3363,
"grad_norm": 15.626523971557617,
"learning_rate": 1.8063388468846968e-05,
"epoch": 1.1364328803851076,
"step": 18650
},
{
"loss": 0.3229,
"grad_norm": 5.168631076812744,
"learning_rate": 1.805537070147447e-05,
"epoch": 1.1394796173298398,
"step": 18700
},
{
"loss": 0.3812,
"grad_norm": 7.538236618041992,
"learning_rate": 1.8047352934101973e-05,
"epoch": 1.142526354274572,
"step": 18750
},
{
"loss": 0.384,
"grad_norm": 8.34493350982666,
"learning_rate": 1.8039335166729475e-05,
"epoch": 1.145573091219304,
"step": 18800
},
{
"loss": 0.3882,
"grad_norm": 5.055959701538086,
"learning_rate": 1.8031317399356974e-05,
"epoch": 1.1486198281640363,
"step": 18850
},
{
"loss": 0.3675,
"grad_norm": 25.83802032470703,
"learning_rate": 1.802329963198448e-05,
"epoch": 1.1516665651087685,
"step": 18900
},
{
"loss": 0.2961,
"grad_norm": 3.963958740234375,
"learning_rate": 1.8015281864611983e-05,
"epoch": 1.1547133020535008,
"step": 18950
},
{
"loss": 0.3687,
"grad_norm": 8.482305526733398,
"learning_rate": 1.8007264097239485e-05,
"epoch": 1.157760038998233,
"step": 19000
},
{
"loss": 0.3861,
"grad_norm": 11.421470642089844,
"learning_rate": 1.7999246329866988e-05,
"epoch": 1.160806775942965,
"step": 19050
},
{
"loss": 0.3875,
"grad_norm": 12.134862899780273,
"learning_rate": 1.799122856249449e-05,
"epoch": 1.1638535128876972,
"step": 19100
},
{
"loss": 0.3114,
"grad_norm": 9.052107810974121,
"learning_rate": 1.7983371150469442e-05,
"epoch": 1.1669002498324295,
"step": 19150
},
{
"loss": 0.3293,
"grad_norm": 14.61961841583252,
"learning_rate": 1.7975353383096944e-05,
"epoch": 1.1699469867771617,
"step": 19200
},
{
"loss": 0.3253,
"grad_norm": 10.686649322509766,
"learning_rate": 1.7967335615724447e-05,
"epoch": 1.172993723721894,
"step": 19250
},
{
"loss": 0.3816,
"grad_norm": 8.711565971374512,
"learning_rate": 1.795931784835195e-05,
"epoch": 1.1760404606666262,
"step": 19300
},
{
"loss": 0.3867,
"grad_norm": 9.840596199035645,
"learning_rate": 1.7951300080979452e-05,
"epoch": 1.1790871976113582,
"step": 19350
},
{
"loss": 0.3528,
"grad_norm": 11.06153678894043,
"learning_rate": 1.7943282313606954e-05,
"epoch": 1.1821339345560904,
"step": 19400
},
{
"loss": 0.3408,
"grad_norm": 19.472579956054688,
"learning_rate": 1.7935264546234457e-05,
"epoch": 1.1851806715008226,
"step": 19450
},
{
"loss": 0.3816,
"grad_norm": 3.2399511337280273,
"learning_rate": 1.792724677886196e-05,
"epoch": 1.1882274084455549,
"step": 19500
},
{
"loss": 0.3621,
"grad_norm": 10.839259147644043,
"learning_rate": 1.7919229011489462e-05,
"epoch": 1.191274145390287,
"step": 19550
},
{
"loss": 0.3574,
"grad_norm": 0.16259242594242096,
"learning_rate": 1.7911211244116965e-05,
"epoch": 1.194320882335019,
"step": 19600
},
{
"loss": 0.4216,
"grad_norm": 5.190318584442139,
"learning_rate": 1.7903193476744467e-05,
"epoch": 1.1973676192797513,
"step": 19650
},
{
"loss": 0.3798,
"grad_norm": 11.18252944946289,
"learning_rate": 1.789517570937197e-05,
"epoch": 1.2004143562244836,
"step": 19700
},
{
"loss": 0.3371,
"grad_norm": 8.665257453918457,
"learning_rate": 1.7887157941999472e-05,
"epoch": 1.2034610931692158,
"step": 19750
},
{
"loss": 0.3502,
"grad_norm": 13.515440940856934,
"learning_rate": 1.7879140174626975e-05,
"epoch": 1.206507830113948,
"step": 19800
},
{
"loss": 0.405,
"grad_norm": 7.559476375579834,
"learning_rate": 1.7871122407254477e-05,
"epoch": 1.2095545670586803,
"step": 19850
},
{
"loss": 0.3479,
"grad_norm": 11.12271785736084,
"learning_rate": 1.786310463988198e-05,
"epoch": 1.2126013040034123,
"step": 19900
},
{
"loss": 0.3843,
"grad_norm": 9.343064308166504,
"learning_rate": 1.7855086872509482e-05,
"epoch": 1.2156480409481445,
"step": 19950
},
{
"loss": 0.3712,
"grad_norm": 6.712246894836426,
"learning_rate": 1.7847069105136985e-05,
"epoch": 1.2186947778928767,
"step": 20000
},
{
"loss": 0.3285,
"grad_norm": 14.704537391662598,
"learning_rate": 1.7839051337764487e-05,
"epoch": 1.221741514837609,
"step": 20050
},
{
"loss": 0.3656,
"grad_norm": 7.124002456665039,
"learning_rate": 1.783103357039199e-05,
"epoch": 1.2247882517823412,
"step": 20100
},
{
"loss": 0.3555,
"grad_norm": 7.009684085845947,
"learning_rate": 1.7823015803019492e-05,
"epoch": 1.2278349887270732,
"step": 20150
},
{
"loss": 0.321,
"grad_norm": 7.843273162841797,
"learning_rate": 1.7814998035646995e-05,
"epoch": 1.2308817256718054,
"step": 20200
},
{
"loss": 0.3229,
"grad_norm": 16.953866958618164,
"learning_rate": 1.7806980268274497e-05,
"epoch": 1.2339284626165377,
"step": 20250
},
{
"loss": 0.3605,
"grad_norm": 16.021997451782227,
"learning_rate": 1.7798962500902e-05,
"epoch": 1.23697519956127,
"step": 20300
},
{
"loss": 0.3939,
"grad_norm": 3.6342976093292236,
"learning_rate": 1.7790944733529502e-05,
"epoch": 1.2400219365060021,
"step": 20350
},
{
"loss": 0.3732,
"grad_norm": 15.961114883422852,
"learning_rate": 1.7782926966157005e-05,
"epoch": 1.2430686734507344,
"step": 20400
},
{
"loss": 0.3274,
"grad_norm": 19.7041015625,
"learning_rate": 1.7774909198784507e-05,
"epoch": 1.2461154103954664,
"step": 20450
},
{
"loss": 0.3822,
"grad_norm": 6.556900501251221,
"learning_rate": 1.776689143141201e-05,
"epoch": 1.2491621473401986,
"step": 20500
},
{
"loss": 0.3879,
"grad_norm": 4.181976795196533,
"learning_rate": 1.7758873664039512e-05,
"epoch": 1.2522088842849308,
"step": 20550
},
{
"loss": 0.436,
"grad_norm": 11.57612133026123,
"learning_rate": 1.7750855896667018e-05,
"epoch": 1.255255621229663,
"step": 20600
},
{
"loss": 0.3629,
"grad_norm": 1.8387733697891235,
"learning_rate": 1.7742838129294517e-05,
"epoch": 1.2583023581743953,
"step": 20650
},
{
"loss": 0.3182,
"grad_norm": 3.4361047744750977,
"learning_rate": 1.773482036192202e-05,
"epoch": 1.2613490951191273,
"step": 20700
},
{
"loss": 0.344,
"grad_norm": 12.807571411132812,
"learning_rate": 1.7726802594549522e-05,
"epoch": 1.2643958320638595,
"step": 20750
},
{
"loss": 0.3946,
"grad_norm": 11.238690376281738,
"learning_rate": 1.7718784827177025e-05,
"epoch": 1.2674425690085918,
"step": 20800
},
{
"loss": 0.2983,
"grad_norm": 12.470406532287598,
"learning_rate": 1.7710767059804527e-05,
"epoch": 1.270489305953324,
"step": 20850
},
{
"loss": 0.3855,
"grad_norm": 20.394935607910156,
"learning_rate": 1.770274929243203e-05,
"epoch": 1.2735360428980562,
"step": 20900
},
{
"loss": 0.4636,
"grad_norm": 1.2745529413223267,
"learning_rate": 1.7694731525059535e-05,
"epoch": 1.2765827798427885,
"step": 20950
},
{
"loss": 0.3489,
"grad_norm": 12.434640884399414,
"learning_rate": 1.7686713757687035e-05,
"epoch": 1.2796295167875207,
"step": 21000
},
{
"loss": 0.3862,
"grad_norm": 3.870356798171997,
"learning_rate": 1.7678695990314537e-05,
"epoch": 1.2826762537322527,
"step": 21050
},
{
"loss": 0.3665,
"grad_norm": 25.46006202697754,
"learning_rate": 1.7670678222942043e-05,
"epoch": 1.285722990676985,
"step": 21100
},
{
"loss": 0.3779,
"grad_norm": 17.15304946899414,
"learning_rate": 1.7662660455569542e-05,
"epoch": 1.2887697276217172,
"step": 21150
},
{
"loss": 0.3177,
"grad_norm": 3.7388744354248047,
"learning_rate": 1.7654642688197045e-05,
"epoch": 1.2918164645664494,
"step": 21200
},
{
"loss": 0.41,
"grad_norm": 4.080338478088379,
"learning_rate": 1.764662492082455e-05,
"epoch": 1.2948632015111814,
"step": 21250
},
{
"loss": 0.3996,
"grad_norm": 10.559809684753418,
"learning_rate": 1.7638767508799502e-05,
"epoch": 1.2979099384559136,
"step": 21300
},
{
"loss": 0.3844,
"grad_norm": 1.0747661590576172,
"learning_rate": 1.7630749741427005e-05,
"epoch": 1.3009566754006459,
"step": 21350
},
{
"loss": 0.355,
"grad_norm": 5.847904682159424,
"learning_rate": 1.7622731974054504e-05,
"epoch": 1.304003412345378,
"step": 21400
},
{
"loss": 0.3667,
"grad_norm": 13.829729080200195,
"learning_rate": 1.761471420668201e-05,
"epoch": 1.3070501492901103,
"step": 21450
},
{
"loss": 0.3222,
"grad_norm": 10.61640453338623,
"learning_rate": 1.7606696439309512e-05,
"epoch": 1.3100968862348426,
"step": 21500
},
{
"loss": 0.3842,
"grad_norm": 12.241172790527344,
"learning_rate": 1.7598839027284464e-05,
"epoch": 1.3131436231795748,
"step": 21550
},
{
"loss": 0.3534,
"grad_norm": 6.718263626098633,
"learning_rate": 1.7590821259911967e-05,
"epoch": 1.3161903601243068,
"step": 21600
},
{
"loss": 0.3592,
"grad_norm": 7.976694583892822,
"learning_rate": 1.758280349253947e-05,
"epoch": 1.319237097069039,
"step": 21650
},
{
"loss": 0.3964,
"grad_norm": 7.9949235916137695,
"learning_rate": 1.7574785725166972e-05,
"epoch": 1.3222838340137713,
"step": 21700
},
{
"loss": 0.3722,
"grad_norm": 7.173233985900879,
"learning_rate": 1.7566767957794474e-05,
"epoch": 1.3253305709585035,
"step": 21750
},
{
"loss": 0.3013,
"grad_norm": 13.530651092529297,
"learning_rate": 1.7558750190421977e-05,
"epoch": 1.3283773079032357,
"step": 21800
},
{
"loss": 0.316,
"grad_norm": 4.235117435455322,
"learning_rate": 1.755073242304948e-05,
"epoch": 1.3314240448479677,
"step": 21850
},
{
"loss": 0.2804,
"grad_norm": 13.993218421936035,
"learning_rate": 1.7542714655676982e-05,
"epoch": 1.3344707817927,
"step": 21900
},
{
"loss": 0.3656,
"grad_norm": 8.257148742675781,
"learning_rate": 1.7534696888304484e-05,
"epoch": 1.3375175187374322,
"step": 21950
},
{
"loss": 0.3222,
"grad_norm": 15.376029968261719,
"learning_rate": 1.7526679120931987e-05,
"epoch": 1.3405642556821644,
"step": 22000
},
{
"loss": 0.2988,
"grad_norm": 16.053144454956055,
"learning_rate": 1.751866135355949e-05,
"epoch": 1.3436109926268966,
"step": 22050
},
{
"loss": 0.3607,
"grad_norm": 14.33431625366211,
"learning_rate": 1.7510643586186992e-05,
"epoch": 1.3466577295716289,
"step": 22100
},
{
"loss": 0.3607,
"grad_norm": 9.38208293914795,
"learning_rate": 1.7502625818814494e-05,
"epoch": 1.3497044665163611,
"step": 22150
},
{
"loss": 0.3492,
"grad_norm": 11.134103775024414,
"learning_rate": 1.7494608051441997e-05,
"epoch": 1.3527512034610931,
"step": 22200
},
{
"loss": 0.2781,
"grad_norm": 3.381770133972168,
"learning_rate": 1.74865902840695e-05,
"epoch": 1.3557979404058254,
"step": 22250
},
{
"loss": 0.2867,
"grad_norm": 2.6657536029815674,
"learning_rate": 1.7478572516697002e-05,
"epoch": 1.3588446773505576,
"step": 22300
},
{
"loss": 0.3134,
"grad_norm": 2.461747646331787,
"learning_rate": 1.7470554749324504e-05,
"epoch": 1.3618914142952898,
"step": 22350
},
{
"loss": 0.3945,
"grad_norm": 12.210373878479004,
"learning_rate": 1.7462536981952007e-05,
"epoch": 1.3649381512400218,
"step": 22400
},
{
"loss": 0.4029,
"grad_norm": 11.86012077331543,
"learning_rate": 1.745451921457951e-05,
"epoch": 1.367984888184754,
"step": 22450
},
{
"loss": 0.3762,
"grad_norm": 7.590579509735107,
"learning_rate": 1.7446501447207012e-05,
"epoch": 1.3710316251294863,
"step": 22500
},
{
"loss": 0.3096,
"grad_norm": 8.387678146362305,
"learning_rate": 1.7438483679834514e-05,
"epoch": 1.3740783620742185,
"step": 22550
},
{
"loss": 0.3485,
"grad_norm": 14.225768089294434,
"learning_rate": 1.7430465912462017e-05,
"epoch": 1.3771250990189507,
"step": 22600
},
{
"loss": 0.2867,
"grad_norm": 1.6792166233062744,
"learning_rate": 1.742244814508952e-05,
"epoch": 1.380171835963683,
"step": 22650
},
{
"loss": 0.3608,
"grad_norm": 25.117210388183594,
"learning_rate": 1.7414430377717022e-05,
"epoch": 1.3832185729084152,
"step": 22700
},
{
"loss": 0.3372,
"grad_norm": 12.765902519226074,
"learning_rate": 1.7406412610344524e-05,
"epoch": 1.3862653098531472,
"step": 22750
},
{
"loss": 0.3211,
"grad_norm": 2.188514232635498,
"learning_rate": 1.7398394842972027e-05,
"epoch": 1.3893120467978795,
"step": 22800
},
{
"loss": 0.2653,
"grad_norm": 8.03336238861084,
"learning_rate": 1.739037707559953e-05,
"epoch": 1.3923587837426117,
"step": 22850
},
{
"loss": 0.3673,
"grad_norm": 13.825742721557617,
"learning_rate": 1.7382359308227032e-05,
"epoch": 1.395405520687344,
"step": 22900
},
{
"loss": 0.3183,
"grad_norm": 13.774291038513184,
"learning_rate": 1.7374341540854534e-05,
"epoch": 1.398452257632076,
"step": 22950
},
{
"loss": 0.3199,
"grad_norm": 8.273527145385742,
"learning_rate": 1.7366323773482037e-05,
"epoch": 1.4014989945768082,
"step": 23000
},
{
"loss": 0.3413,
"grad_norm": 20.586673736572266,
"learning_rate": 1.735830600610954e-05,
"epoch": 1.4045457315215404,
"step": 23050
},
{
"loss": 0.2961,
"grad_norm": 3.915707588195801,
"learning_rate": 1.7350288238737042e-05,
"epoch": 1.4075924684662726,
"step": 23100
},
{
"loss": 0.3383,
"grad_norm": 11.608159065246582,
"learning_rate": 1.7342270471364548e-05,
"epoch": 1.4106392054110048,
"step": 23150
},
{
"loss": 0.2955,
"grad_norm": 7.316932678222656,
"learning_rate": 1.7334252703992047e-05,
"epoch": 1.413685942355737,
"step": 23200
},
{
"loss": 0.2842,
"grad_norm": 6.487236499786377,
"learning_rate": 1.732623493661955e-05,
"epoch": 1.4167326793004693,
"step": 23250
},
{
"loss": 0.3348,
"grad_norm": 3.874905586242676,
"learning_rate": 1.7318217169247052e-05,
"epoch": 1.4197794162452013,
"step": 23300
},
{
"loss": 0.3144,
"grad_norm": 7.151147365570068,
"learning_rate": 1.7310199401874554e-05,
"epoch": 1.4228261531899336,
"step": 23350
},
{
"loss": 0.3225,
"grad_norm": 3.7718214988708496,
"learning_rate": 1.7302181634502057e-05,
"epoch": 1.4258728901346658,
"step": 23400
},
{
"loss": 0.3397,
"grad_norm": 10.638415336608887,
"learning_rate": 1.729416386712956e-05,
"epoch": 1.428919627079398,
"step": 23450
},
{
"loss": 0.316,
"grad_norm": 25.834095001220703,
"learning_rate": 1.7286146099757065e-05,
"epoch": 1.43196636402413,
"step": 23500
},
{
"loss": 0.3541,
"grad_norm": 13.270949363708496,
"learning_rate": 1.7278128332384564e-05,
"epoch": 1.4350131009688623,
"step": 23550
},
{
"loss": 0.3225,
"grad_norm": 13.803901672363281,
"learning_rate": 1.7270110565012067e-05,
"epoch": 1.4380598379135945,
"step": 23600
},
{
"loss": 0.3703,
"grad_norm": 9.725604057312012,
"learning_rate": 1.7262092797639573e-05,
"epoch": 1.4411065748583267,
"step": 23650
},
{
"loss": 0.2706,
"grad_norm": 3.4581453800201416,
"learning_rate": 1.7254075030267072e-05,
"epoch": 1.444153311803059,
"step": 23700
},
{
"loss": 0.2523,
"grad_norm": 6.495072364807129,
"learning_rate": 1.7246057262894574e-05,
"epoch": 1.4472000487477912,
"step": 23750
},
{
"loss": 0.3289,
"grad_norm": 9.765362739562988,
"learning_rate": 1.7238039495522077e-05,
"epoch": 1.4502467856925234,
"step": 23800
},
{
"loss": 0.3021,
"grad_norm": 3.8246703147888184,
"learning_rate": 1.7230021728149583e-05,
"epoch": 1.4532935226372554,
"step": 23850
},
{
"loss": 0.3375,
"grad_norm": 7.573472499847412,
"learning_rate": 1.7222003960777082e-05,
"epoch": 1.4563402595819877,
"step": 23900
},
{
"loss": 0.3462,
"grad_norm": 6.002778053283691,
"learning_rate": 1.7213986193404584e-05,
"epoch": 1.4593869965267199,
"step": 23950
},
{
"loss": 0.2978,
"grad_norm": 9.227519035339355,
"learning_rate": 1.720596842603209e-05,
"epoch": 1.4624337334714521,
"step": 24000
},
{
"loss": 0.3243,
"grad_norm": 13.693887710571289,
"learning_rate": 1.719795065865959e-05,
"epoch": 1.4654804704161843,
"step": 24050
},
{
"loss": 0.3189,
"grad_norm": 15.704029083251953,
"learning_rate": 1.7189932891287092e-05,
"epoch": 1.4685272073609164,
"step": 24100
},
{
"loss": 0.3052,
"grad_norm": 13.327622413635254,
"learning_rate": 1.7181915123914598e-05,
"epoch": 1.4715739443056486,
"step": 24150
},
{
"loss": 0.3981,
"grad_norm": 14.926920890808105,
"learning_rate": 1.71738973565421e-05,
"epoch": 1.4746206812503808,
"step": 24200
},
{
"loss": 0.3107,
"grad_norm": 9.864826202392578,
"learning_rate": 1.71658795891696e-05,
"epoch": 1.477667418195113,
"step": 24250
},
{
"loss": 0.3279,
"grad_norm": 1.3453401327133179,
"learning_rate": 1.7157861821797105e-05,
"epoch": 1.4807141551398453,
"step": 24300
},
{
"loss": 0.3289,
"grad_norm": 16.734621047973633,
"learning_rate": 1.7149844054424608e-05,
"epoch": 1.4837608920845775,
"step": 24350
},
{
"loss": 0.3003,
"grad_norm": 4.578878879547119,
"learning_rate": 1.7141826287052107e-05,
"epoch": 1.4868076290293097,
"step": 24400
},
{
"loss": 0.3204,
"grad_norm": 2.33801531791687,
"learning_rate": 1.713380851967961e-05,
"epoch": 1.4898543659740417,
"step": 24450
},
{
"loss": 0.3234,
"grad_norm": 7.014683246612549,
"learning_rate": 1.7125790752307115e-05,
"epoch": 1.492901102918774,
"step": 24500
},
{
"loss": 0.3301,
"grad_norm": 16.192855834960938,
"learning_rate": 1.7117772984934618e-05,
"epoch": 1.4959478398635062,
"step": 24550
},
{
"loss": 0.401,
"grad_norm": 5.642870903015137,
"learning_rate": 1.7109755217562117e-05,
"epoch": 1.4989945768082384,
"step": 24600
},
{
"loss": 0.3604,
"grad_norm": 13.10983943939209,
"learning_rate": 1.7101737450189623e-05,
"epoch": 1.5020413137529705,
"step": 24650
},
{
"loss": 0.3002,
"grad_norm": 6.089293956756592,
"learning_rate": 1.7093719682817125e-05,
"epoch": 1.5050880506977027,
"step": 24700
},
{
"loss": 0.3416,
"grad_norm": 6.627552032470703,
"learning_rate": 1.7085701915444625e-05,
"epoch": 1.508134787642435,
"step": 24750
},
{
"loss": 0.2969,
"grad_norm": 9.110481262207031,
"learning_rate": 1.707768414807213e-05,
"epoch": 1.5111815245871671,
"step": 24800
},
{
"loss": 0.3194,
"grad_norm": 3.8969027996063232,
"learning_rate": 1.7069666380699633e-05,
"epoch": 1.5142282615318994,
"step": 24850
},
{
"loss": 0.3171,
"grad_norm": 7.680835247039795,
"learning_rate": 1.7061648613327135e-05,
"epoch": 1.5172749984766316,
"step": 24900
},
{
"loss": 0.3091,
"grad_norm": 10.085229873657227,
"learning_rate": 1.7053630845954635e-05,
"epoch": 1.5203217354213638,
"step": 24950
},
{
"loss": 0.3228,
"grad_norm": 19.494258880615234,
"learning_rate": 1.704561307858214e-05,
"epoch": 1.523368472366096,
"step": 25000
},
{
"loss": 0.3625,
"grad_norm": 0.5019375085830688,
"learning_rate": 1.7037595311209643e-05,
"epoch": 1.526415209310828,
"step": 25050
},
{
"loss": 0.3307,
"grad_norm": 2.7318079471588135,
"learning_rate": 1.7029577543837142e-05,
"epoch": 1.5294619462555603,
"step": 25100
},
{
"loss": 0.3387,
"grad_norm": 6.749081611633301,
"learning_rate": 1.7021559776464648e-05,
"epoch": 1.5325086832002923,
"step": 25150
},
{
"loss": 0.339,
"grad_norm": 8.64217758178711,
"learning_rate": 1.701354200909215e-05,
"epoch": 1.5355554201450246,
"step": 25200
},
{
"loss": 0.2811,
"grad_norm": 12.908346176147461,
"learning_rate": 1.7005524241719653e-05,
"epoch": 1.5386021570897568,
"step": 25250
},
{
"loss": 0.3244,
"grad_norm": 16.78288459777832,
"learning_rate": 1.6997506474347155e-05,
"epoch": 1.541648894034489,
"step": 25300
},
{
"loss": 0.3364,
"grad_norm": 3.407428741455078,
"learning_rate": 1.6989488706974658e-05,
"epoch": 1.5446956309792212,
"step": 25350
},
{
"loss": 0.2948,
"grad_norm": 11.489734649658203,
"learning_rate": 1.698147093960216e-05,
"epoch": 1.5477423679239535,
"step": 25400
},
{
"loss": 0.2723,
"grad_norm": 2.989764451980591,
"learning_rate": 1.6973453172229663e-05,
"epoch": 1.5507891048686857,
"step": 25450
},
{
"loss": 0.3403,
"grad_norm": 5.868636608123779,
"learning_rate": 1.6965435404857165e-05,
"epoch": 1.553835841813418,
"step": 25500
},
{
"loss": 0.312,
"grad_norm": 10.837996482849121,
"learning_rate": 1.6957417637484668e-05,
"epoch": 1.5568825787581502,
"step": 25550
},
{
"loss": 0.315,
"grad_norm": 15.693818092346191,
"learning_rate": 1.694939987011217e-05,
"epoch": 1.5599293157028822,
"step": 25600
},
{
"loss": 0.3456,
"grad_norm": 6.592184543609619,
"learning_rate": 1.6941382102739673e-05,
"epoch": 1.5629760526476144,
"step": 25650
},
{
"loss": 0.2885,
"grad_norm": 22.89328384399414,
"learning_rate": 1.6933364335367176e-05,
"epoch": 1.5660227895923466,
"step": 25700
},
{
"loss": 0.3457,
"grad_norm": 8.998388290405273,
"learning_rate": 1.6925346567994678e-05,
"epoch": 1.5690695265370787,
"step": 25750
},
{
"loss": 0.2781,
"grad_norm": 1.9641547203063965,
"learning_rate": 1.691732880062218e-05,
"epoch": 1.5721162634818109,
"step": 25800
},
{
"loss": 0.264,
"grad_norm": 7.985995769500732,
"learning_rate": 1.6909311033249683e-05,
"epoch": 1.5751630004265431,
"step": 25850
},
{
"loss": 0.3076,
"grad_norm": 8.128706932067871,
"learning_rate": 1.6901293265877186e-05,
"epoch": 1.5782097373712753,
"step": 25900
},
{
"loss": 0.3225,
"grad_norm": 12.334968566894531,
"learning_rate": 1.6893275498504688e-05,
"epoch": 1.5812564743160076,
"step": 25950
},
{
"loss": 0.3224,
"grad_norm": 11.410700798034668,
"learning_rate": 1.688541808647964e-05,
"epoch": 1.5843032112607398,
"step": 26000
},
{
"loss": 0.352,
"grad_norm": 7.473352909088135,
"learning_rate": 1.6877400319107142e-05,
"epoch": 1.587349948205472,
"step": 26050
},
{
"loss": 0.3278,
"grad_norm": 11.380707740783691,
"learning_rate": 1.6869382551734645e-05,
"epoch": 1.5903966851502043,
"step": 26100
},
{
"loss": 0.3296,
"grad_norm": 2.4631125926971436,
"learning_rate": 1.6861364784362147e-05,
"epoch": 1.5934434220949363,
"step": 26150
},
{
"loss": 0.28,
"grad_norm": 8.952128410339355,
"learning_rate": 1.685334701698965e-05,
"epoch": 1.5964901590396685,
"step": 26200
},
{
"loss": 0.3234,
"grad_norm": 8.922794342041016,
"learning_rate": 1.6845329249617152e-05,
"epoch": 1.5995368959844007,
"step": 26250
},
{
"loss": 0.3435,
"grad_norm": 0.5245643258094788,
"learning_rate": 1.6837311482244655e-05,
"epoch": 1.6025836329291328,
"step": 26300
},
{
"loss": 0.3083,
"grad_norm": 4.835029125213623,
"learning_rate": 1.6829293714872157e-05,
"epoch": 1.605630369873865,
"step": 26350
},
{
"loss": 0.3074,
"grad_norm": 8.97582721710205,
"learning_rate": 1.682127594749966e-05,
"epoch": 1.6086771068185972,
"step": 26400
},
{
"loss": 0.3143,
"grad_norm": 20.144168853759766,
"learning_rate": 1.6813258180127162e-05,
"epoch": 1.6117238437633294,
"step": 26450
},
{
"loss": 0.2952,
"grad_norm": 7.0517730712890625,
"learning_rate": 1.6805240412754665e-05,
"epoch": 1.6147705807080617,
"step": 26500
},
{
"loss": 0.2636,
"grad_norm": 8.716961860656738,
"learning_rate": 1.6797222645382167e-05,
"epoch": 1.617817317652794,
"step": 26550
},
{
"loss": 0.3352,
"grad_norm": 4.337059497833252,
"learning_rate": 1.678920487800967e-05,
"epoch": 1.6208640545975261,
"step": 26600
},
{
"loss": 0.2624,
"grad_norm": 16.230514526367188,
"learning_rate": 1.6781187110637172e-05,
"epoch": 1.6239107915422584,
"step": 26650
},
{
"loss": 0.3333,
"grad_norm": 1.0672177076339722,
"learning_rate": 1.677316934326468e-05,
"epoch": 1.6269575284869904,
"step": 26700
},
{
"loss": 0.3113,
"grad_norm": 8.474405288696289,
"learning_rate": 1.6765151575892177e-05,
"epoch": 1.6300042654317226,
"step": 26750
},
{
"loss": 0.3251,
"grad_norm": 7.372742176055908,
"learning_rate": 1.675713380851968e-05,
"epoch": 1.6330510023764548,
"step": 26800
},
{
"loss": 0.2759,
"grad_norm": 1.649053692817688,
"learning_rate": 1.6749116041147186e-05,
"epoch": 1.6360977393211869,
"step": 26850
},
{
"loss": 0.2941,
"grad_norm": 7.650877952575684,
"learning_rate": 1.6741098273774685e-05,
"epoch": 1.639144476265919,
"step": 26900
},
{
"loss": 0.2976,
"grad_norm": 20.676498413085938,
"learning_rate": 1.6733080506402187e-05,
"epoch": 1.6421912132106513,
"step": 26950
},
{
"loss": 0.3392,
"grad_norm": 21.625612258911133,
"learning_rate": 1.672506273902969e-05,
"epoch": 1.6452379501553835,
"step": 27000
},
{
"loss": 0.3116,
"grad_norm": 13.993587493896484,
"learning_rate": 1.6717044971657196e-05,
"epoch": 1.6482846871001158,
"step": 27050
},
{
"loss": 0.3203,
"grad_norm": 7.262061595916748,
"learning_rate": 1.6709027204284695e-05,
"epoch": 1.651331424044848,
"step": 27100
},
{
"loss": 0.2875,
"grad_norm": 8.886082649230957,
"learning_rate": 1.6701009436912198e-05,
"epoch": 1.6543781609895802,
"step": 27150
},
{
"loss": 0.2901,
"grad_norm": 11.764669418334961,
"learning_rate": 1.6692991669539703e-05,
"epoch": 1.6574248979343125,
"step": 27200
},
{
"loss": 0.275,
"grad_norm": 13.272078514099121,
"learning_rate": 1.6684973902167203e-05,
"epoch": 1.6604716348790447,
"step": 27250
},
{
"loss": 0.2425,
"grad_norm": 10.089193344116211,
"learning_rate": 1.6676956134794705e-05,
"epoch": 1.6635183718237767,
"step": 27300
},
{
"loss": 0.2999,
"grad_norm": 12.321874618530273,
"learning_rate": 1.666893836742221e-05,
"epoch": 1.666565108768509,
"step": 27350
},
{
"loss": 0.3374,
"grad_norm": 42.274723052978516,
"learning_rate": 1.6661080955397163e-05,
"epoch": 1.6696118457132412,
"step": 27400
},
{
"loss": 0.3548,
"grad_norm": 16.946142196655273,
"learning_rate": 1.6653063188024662e-05,
"epoch": 1.6726585826579732,
"step": 27450
},
{
"loss": 0.2994,
"grad_norm": 2.136284589767456,
"learning_rate": 1.6645045420652164e-05,
"epoch": 1.6757053196027054,
"step": 27500
},
{
"loss": 0.3096,
"grad_norm": 17.080486297607422,
"learning_rate": 1.663702765327967e-05,
"epoch": 1.6787520565474376,
"step": 27550
},
{
"loss": 0.3217,
"grad_norm": 14.589789390563965,
"learning_rate": 1.6629009885907173e-05,
"epoch": 1.6817987934921699,
"step": 27600
},
{
"loss": 0.2544,
"grad_norm": 13.077963829040527,
"learning_rate": 1.6620992118534672e-05,
"epoch": 1.684845530436902,
"step": 27650
},
{
"loss": 0.3596,
"grad_norm": 25.065895080566406,
"learning_rate": 1.6612974351162178e-05,
"epoch": 1.6878922673816343,
"step": 27700
},
{
"loss": 0.294,
"grad_norm": 9.362024307250977,
"learning_rate": 1.660495658378968e-05,
"epoch": 1.6909390043263666,
"step": 27750
},
{
"loss": 0.256,
"grad_norm": 10.644511222839355,
"learning_rate": 1.659693881641718e-05,
"epoch": 1.6939857412710988,
"step": 27800
},
{
"loss": 0.3066,
"grad_norm": 7.913748264312744,
"learning_rate": 1.6588921049044685e-05,
"epoch": 1.6970324782158308,
"step": 27850
},
{
"loss": 0.2875,
"grad_norm": 5.977893352508545,
"learning_rate": 1.6580903281672188e-05,
"epoch": 1.700079215160563,
"step": 27900
},
{
"loss": 0.3341,
"grad_norm": 6.6945414543151855,
"learning_rate": 1.657288551429969e-05,
"epoch": 1.7031259521052953,
"step": 27950
},
{
"loss": 0.2992,
"grad_norm": 14.339247703552246,
"learning_rate": 1.6564867746927193e-05,
"epoch": 1.7061726890500273,
"step": 28000
},
{
"loss": 0.3364,
"grad_norm": 0.6013835668563843,
"learning_rate": 1.6556849979554695e-05,
"epoch": 1.7092194259947595,
"step": 28050
},
{
"loss": 0.2342,
"grad_norm": 16.41084098815918,
"learning_rate": 1.6548832212182198e-05,
"epoch": 1.7122661629394917,
"step": 28100
},
{
"loss": 0.2967,
"grad_norm": 1.8200198411941528,
"learning_rate": 1.6540814444809697e-05,
"epoch": 1.715312899884224,
"step": 28150
},
{
"loss": 0.3457,
"grad_norm": 11.769265174865723,
"learning_rate": 1.6532796677437203e-05,
"epoch": 1.7183596368289562,
"step": 28200
},
{
"loss": 0.3509,
"grad_norm": 16.372844696044922,
"learning_rate": 1.6524778910064705e-05,
"epoch": 1.7214063737736884,
"step": 28250
},
{
"loss": 0.3782,
"grad_norm": 9.19438648223877,
"learning_rate": 1.6516761142692208e-05,
"epoch": 1.7244531107184207,
"step": 28300
},
{
"loss": 0.2836,
"grad_norm": 7.3909430503845215,
"learning_rate": 1.650874337531971e-05,
"epoch": 1.727499847663153,
"step": 28350
},
{
"loss": 0.3021,
"grad_norm": 7.053426742553711,
"learning_rate": 1.6500725607947213e-05,
"epoch": 1.730546584607885,
"step": 28400
},
{
"loss": 0.3092,
"grad_norm": 11.572251319885254,
"learning_rate": 1.6492707840574715e-05,
"epoch": 1.7335933215526171,
"step": 28450
},
{
"loss": 0.2824,
"grad_norm": 15.419539451599121,
"learning_rate": 1.6484690073202218e-05,
"epoch": 1.7366400584973494,
"step": 28500
},
{
"loss": 0.3067,
"grad_norm": 5.662206649780273,
"learning_rate": 1.647667230582972e-05,
"epoch": 1.7396867954420814,
"step": 28550
},
{
"loss": 0.3116,
"grad_norm": 3.3203351497650146,
"learning_rate": 1.6468654538457223e-05,
"epoch": 1.7427335323868136,
"step": 28600
},
{
"loss": 0.3655,
"grad_norm": 5.361013889312744,
"learning_rate": 1.6460636771084725e-05,
"epoch": 1.7457802693315458,
"step": 28650
},
{
"loss": 0.3476,
"grad_norm": 22.084739685058594,
"learning_rate": 1.6452619003712228e-05,
"epoch": 1.748827006276278,
"step": 28700
},
{
"loss": 0.3292,
"grad_norm": 4.947688579559326,
"learning_rate": 1.644460123633973e-05,
"epoch": 1.7518737432210103,
"step": 28750
},
{
"loss": 0.3418,
"grad_norm": 12.871330261230469,
"learning_rate": 1.6436583468967233e-05,
"epoch": 1.7549204801657425,
"step": 28800
},
{
"loss": 0.2806,
"grad_norm": 7.678299427032471,
"learning_rate": 1.6428565701594735e-05,
"epoch": 1.7579672171104748,
"step": 28850
},
{
"loss": 0.3075,
"grad_norm": 8.97616195678711,
"learning_rate": 1.6420547934222238e-05,
"epoch": 1.761013954055207,
"step": 28900
},
{
"loss": 0.3357,
"grad_norm": 3.064373016357422,
"learning_rate": 1.641253016684974e-05,
"epoch": 1.7640606909999392,
"step": 28950
},
{
"loss": 0.3204,
"grad_norm": 13.351383209228516,
"learning_rate": 1.6404512399477243e-05,
"epoch": 1.7671074279446712,
"step": 29000
},
{
"loss": 0.3183,
"grad_norm": 15.027705192565918,
"learning_rate": 1.6396494632104745e-05,
"epoch": 1.7701541648894035,
"step": 29050
},
{
"loss": 0.3204,
"grad_norm": 11.368523597717285,
"learning_rate": 1.6388476864732248e-05,
"epoch": 1.7732009018341355,
"step": 29100
},
{
"loss": 0.3568,
"grad_norm": 8.829998970031738,
"learning_rate": 1.638045909735975e-05,
"epoch": 1.7762476387788677,
"step": 29150
},
{
"loss": 0.3042,
"grad_norm": 2.9043288230895996,
"learning_rate": 1.6372441329987253e-05,
"epoch": 1.7792943757236,
"step": 29200
},
{
"loss": 0.2678,
"grad_norm": 6.21755313873291,
"learning_rate": 1.6364423562614755e-05,
"epoch": 1.7823411126683322,
"step": 29250
},
{
"loss": 0.2722,
"grad_norm": 5.3131279945373535,
"learning_rate": 1.6356405795242258e-05,
"epoch": 1.7853878496130644,
"step": 29300
},
{
"loss": 0.2766,
"grad_norm": 17.139389038085938,
"learning_rate": 1.634838802786976e-05,
"epoch": 1.7884345865577966,
"step": 29350
},
{
"loss": 0.2922,
"grad_norm": 7.138078212738037,
"learning_rate": 1.6340370260497263e-05,
"epoch": 1.7914813235025289,
"step": 29400
},
{
"loss": 0.2974,
"grad_norm": 6.688146114349365,
"learning_rate": 1.6332352493124765e-05,
"epoch": 1.794528060447261,
"step": 29450
},
{
"loss": 0.2414,
"grad_norm": 5.749677658081055,
"learning_rate": 1.6324334725752268e-05,
"epoch": 1.7975747973919933,
"step": 29500
},
{
"loss": 0.3287,
"grad_norm": 10.845946311950684,
"learning_rate": 1.631631695837977e-05,
"epoch": 1.8006215343367253,
"step": 29550
},
{
"loss": 0.2326,
"grad_norm": 15.988638877868652,
"learning_rate": 1.6308299191007273e-05,
"epoch": 1.8036682712814576,
"step": 29600
},
{
"loss": 0.3314,
"grad_norm": 5.304831504821777,
"learning_rate": 1.6300281423634775e-05,
"epoch": 1.8067150082261898,
"step": 29650
},
{
"loss": 0.325,
"grad_norm": 0.4449011981487274,
"learning_rate": 1.6292263656262278e-05,
"epoch": 1.8097617451709218,
"step": 29700
},
{
"loss": 0.2902,
"grad_norm": 2.990488290786743,
"learning_rate": 1.628424588888978e-05,
"epoch": 1.812808482115654,
"step": 29750
},
{
"loss": 0.2551,
"grad_norm": 11.540789604187012,
"learning_rate": 1.6276228121517283e-05,
"epoch": 1.8158552190603863,
"step": 29800
},
{
"loss": 0.2732,
"grad_norm": 8.273702621459961,
"learning_rate": 1.6268210354144785e-05,
"epoch": 1.8189019560051185,
"step": 29850
},
{
"loss": 0.3335,
"grad_norm": 14.36491870880127,
"learning_rate": 1.6260192586772288e-05,
"epoch": 1.8219486929498507,
"step": 29900
},
{
"loss": 0.349,
"grad_norm": 1.8825091123580933,
"learning_rate": 1.625217481939979e-05,
"epoch": 1.824995429894583,
"step": 29950
},
{
"loss": 0.2804,
"grad_norm": 2.584400177001953,
"learning_rate": 1.6244157052027293e-05,
"epoch": 1.8280421668393152,
"step": 30000
},
{
"loss": 0.3296,
"grad_norm": 6.209488391876221,
"learning_rate": 1.6236139284654795e-05,
"epoch": 1.8310889037840474,
"step": 30050
},
{
"loss": 0.2436,
"grad_norm": 4.271270275115967,
"learning_rate": 1.6228121517282298e-05,
"epoch": 1.8341356407287794,
"step": 30100
},
{
"loss": 0.297,
"grad_norm": 4.773468494415283,
"learning_rate": 1.62201037499098e-05,
"epoch": 1.8371823776735117,
"step": 30150
},
{
"loss": 0.2392,
"grad_norm": 10.907275199890137,
"learning_rate": 1.6212085982537303e-05,
"epoch": 1.840229114618244,
"step": 30200
},
{
"loss": 0.2326,
"grad_norm": 5.878034591674805,
"learning_rate": 1.6204068215164806e-05,
"epoch": 1.843275851562976,
"step": 30250
},
{
"loss": 0.2756,
"grad_norm": 6.607521057128906,
"learning_rate": 1.6196050447792308e-05,
"epoch": 1.8463225885077081,
"step": 30300
},
{
"loss": 0.2745,
"grad_norm": 38.46754837036133,
"learning_rate": 1.618803268041981e-05,
"epoch": 1.8493693254524404,
"step": 30350
},
{
"loss": 0.2566,
"grad_norm": 2.579451084136963,
"learning_rate": 1.6180014913047316e-05,
"epoch": 1.8524160623971726,
"step": 30400
},
{
"loss": 0.3211,
"grad_norm": 13.066880226135254,
"learning_rate": 1.6171997145674816e-05,
"epoch": 1.8554627993419048,
"step": 30450
},
{
"loss": 0.3101,
"grad_norm": 13.854639053344727,
"learning_rate": 1.6163979378302318e-05,
"epoch": 1.858509536286637,
"step": 30500
},
{
"loss": 0.2991,
"grad_norm": 9.205604553222656,
"learning_rate": 1.6155961610929824e-05,
"epoch": 1.8615562732313693,
"step": 30550
},
{
"loss": 0.2739,
"grad_norm": 19.181583404541016,
"learning_rate": 1.6147943843557323e-05,
"epoch": 1.8646030101761015,
"step": 30600
},
{
"loss": 0.2346,
"grad_norm": 4.946005344390869,
"learning_rate": 1.6139926076184826e-05,
"epoch": 1.8676497471208335,
"step": 30650
},
{
"loss": 0.3341,
"grad_norm": 6.155637264251709,
"learning_rate": 1.6131908308812328e-05,
"epoch": 1.8706964840655658,
"step": 30700
},
{
"loss": 0.295,
"grad_norm": 7.367659091949463,
"learning_rate": 1.6123890541439834e-05,
"epoch": 1.873743221010298,
"step": 30750
},
{
"loss": 0.3736,
"grad_norm": 8.987998962402344,
"learning_rate": 1.6115872774067333e-05,
"epoch": 1.87678995795503,
"step": 30800
},
{
"loss": 0.2669,
"grad_norm": 15.289146423339844,
"learning_rate": 1.6107855006694836e-05,
"epoch": 1.8798366948997622,
"step": 30850
},
{
"loss": 0.287,
"grad_norm": 7.135644435882568,
"learning_rate": 1.609983723932234e-05,
"epoch": 1.8828834318444945,
"step": 30900
},
{
"loss": 0.2929,
"grad_norm": 10.530205726623535,
"learning_rate": 1.609181947194984e-05,
"epoch": 1.8859301687892267,
"step": 30950
},
{
"loss": 0.2887,
"grad_norm": 23.027942657470703,
"learning_rate": 1.6083801704577343e-05,
"epoch": 1.888976905733959,
"step": 31000
},
{
"loss": 0.24,
"grad_norm": 18.42806625366211,
"learning_rate": 1.607578393720485e-05,
"epoch": 1.8920236426786912,
"step": 31050
},
{
"loss": 0.3236,
"grad_norm": 34.70264434814453,
"learning_rate": 1.60679265251798e-05,
"epoch": 1.8950703796234234,
"step": 31100
},
{
"loss": 0.3193,
"grad_norm": 12.506950378417969,
"learning_rate": 1.6059908757807303e-05,
"epoch": 1.8981171165681556,
"step": 31150
},
{
"loss": 0.2401,
"grad_norm": 3.841200113296509,
"learning_rate": 1.6051890990434802e-05,
"epoch": 1.9011638535128879,
"step": 31200
},
{
"loss": 0.3137,
"grad_norm": 3.273566484451294,
"learning_rate": 1.604387322306231e-05,
"epoch": 1.9042105904576199,
"step": 31250
},
{
"loss": 0.291,
"grad_norm": 9.005722999572754,
"learning_rate": 1.603585545568981e-05,
"epoch": 1.907257327402352,
"step": 31300
},
{
"loss": 0.23,
"grad_norm": 3.39867901802063,
"learning_rate": 1.602783768831731e-05,
"epoch": 1.910304064347084,
"step": 31350
},
{
"loss": 0.278,
"grad_norm": 5.881619930267334,
"learning_rate": 1.6019819920944816e-05,
"epoch": 1.9133508012918163,
"step": 31400
},
{
"loss": 0.2842,
"grad_norm": 6.515830039978027,
"learning_rate": 1.601180215357232e-05,
"epoch": 1.9163975382365486,
"step": 31450
},
{
"loss": 0.2614,
"grad_norm": 6.584390163421631,
"learning_rate": 1.600378438619982e-05,
"epoch": 1.9194442751812808,
"step": 31500
},
{
"loss": 0.2657,
"grad_norm": 10.07042407989502,
"learning_rate": 1.5995766618827323e-05,
"epoch": 1.922491012126013,
"step": 31550
},
{
"loss": 0.2967,
"grad_norm": 15.12956428527832,
"learning_rate": 1.5987748851454826e-05,
"epoch": 1.9255377490707453,
"step": 31600
},
{
"loss": 0.2424,
"grad_norm": 10.840766906738281,
"learning_rate": 1.597973108408233e-05,
"epoch": 1.9285844860154775,
"step": 31650
},
{
"loss": 0.236,
"grad_norm": 8.643882751464844,
"learning_rate": 1.597171331670983e-05,
"epoch": 1.9316312229602097,
"step": 31700
},
{
"loss": 0.2682,
"grad_norm": 14.98704719543457,
"learning_rate": 1.5963695549337333e-05,
"epoch": 1.934677959904942,
"step": 31750
},
{
"loss": 0.3028,
"grad_norm": 1.3127882480621338,
"learning_rate": 1.5955677781964836e-05,
"epoch": 1.937724696849674,
"step": 31800
},
{
"loss": 0.2355,
"grad_norm": 11.41299819946289,
"learning_rate": 1.594766001459234e-05,
"epoch": 1.9407714337944062,
"step": 31850
},
{
"loss": 0.3015,
"grad_norm": 4.80661678314209,
"learning_rate": 1.593964224721984e-05,
"epoch": 1.9438181707391384,
"step": 31900
},
{
"loss": 0.2825,
"grad_norm": 5.070887565612793,
"learning_rate": 1.5931624479847343e-05,
"epoch": 1.9468649076838704,
"step": 31950
},
{
"loss": 0.287,
"grad_norm": 2.7772228717803955,
"learning_rate": 1.5923606712474846e-05,
"epoch": 1.9499116446286027,
"step": 32000
},
{
"loss": 0.2558,
"grad_norm": 5.3242573738098145,
"learning_rate": 1.591558894510235e-05,
"epoch": 1.952958381573335,
"step": 32050
},
{
"loss": 0.3142,
"grad_norm": 12.350196838378906,
"learning_rate": 1.590757117772985e-05,
"epoch": 1.9560051185180671,
"step": 32100
},
{
"loss": 0.2799,
"grad_norm": 16.592130661010742,
"learning_rate": 1.5899553410357353e-05,
"epoch": 1.9590518554627994,
"step": 32150
},
{
"loss": 0.2941,
"grad_norm": 11.53604793548584,
"learning_rate": 1.5891535642984856e-05,
"epoch": 1.9620985924075316,
"step": 32200
},
{
"loss": 0.2914,
"grad_norm": 13.886282920837402,
"learning_rate": 1.588351787561236e-05,
"epoch": 1.9651453293522638,
"step": 32250
},
{
"loss": 0.272,
"grad_norm": 8.287816047668457,
"learning_rate": 1.587550010823986e-05,
"epoch": 1.968192066296996,
"step": 32300
},
{
"loss": 0.2374,
"grad_norm": 2.468762159347534,
"learning_rate": 1.5867482340867363e-05,
"epoch": 1.971238803241728,
"step": 32350
},
{
"loss": 0.3393,
"grad_norm": 8.733967781066895,
"learning_rate": 1.5859464573494866e-05,
"epoch": 1.9742855401864603,
"step": 32400
},
{
"loss": 0.2419,
"grad_norm": 2.1739065647125244,
"learning_rate": 1.585144680612237e-05,
"epoch": 1.9773322771311925,
"step": 32450
},
{
"loss": 0.2995,
"grad_norm": 4.084203720092773,
"learning_rate": 1.584342903874987e-05,
"epoch": 1.9803790140759245,
"step": 32500
},
{
"loss": 0.2737,
"grad_norm": 18.032894134521484,
"learning_rate": 1.5835411271377373e-05,
"epoch": 1.9834257510206568,
"step": 32550
},
{
"loss": 0.2844,
"grad_norm": 1.8042428493499756,
"learning_rate": 1.5827393504004876e-05,
"epoch": 1.986472487965389,
"step": 32600
},
{
"loss": 0.2399,
"grad_norm": 7.9712982177734375,
"learning_rate": 1.581937573663238e-05,
"epoch": 1.9895192249101212,
"step": 32650
},
{
"loss": 0.2561,
"grad_norm": 15.293970108032227,
"learning_rate": 1.581135796925988e-05,
"epoch": 1.9925659618548535,
"step": 32700
},
{
"loss": 0.2181,
"grad_norm": 1.9992924928665161,
"learning_rate": 1.5803340201887383e-05,
"epoch": 1.9956126987995857,
"step": 32750
},
{
"loss": 0.2275,
"grad_norm": 9.129408836364746,
"learning_rate": 1.5795322434514886e-05,
"epoch": 1.998659435744318,
"step": 32800
},
{
"eval_loss": 0.16226136684417725,
"eval_accuracy": 0.9560925449871466,
"eval_f1_macro": 0.955396121012068,
"eval_top5_accuracy": 0.9951328191945159,
"eval_runtime": 218.9983,
"eval_samples_per_second": 133.22,
"eval_steps_per_second": 8.329,
"epoch": 2.0,
"step": 32822
},
{
"loss": 0.305,
"grad_norm": 12.002199172973633,
"learning_rate": 1.578730466714239e-05,
"epoch": 2.00170617268905,
"step": 32850
},
{
"loss": 0.2736,
"grad_norm": 17.91629981994629,
"learning_rate": 1.577928689976989e-05,
"epoch": 2.0047529096337824,
"step": 32900
},
{
"loss": 0.228,
"grad_norm": 0.548559308052063,
"learning_rate": 1.5771269132397393e-05,
"epoch": 2.0077996465785146,
"step": 32950
},
{
"loss": 0.2731,
"grad_norm": 16.43472671508789,
"learning_rate": 1.5763251365024896e-05,
"epoch": 2.0108463835232464,
"step": 33000
},
{
"loss": 0.2442,
"grad_norm": 12.24728775024414,
"learning_rate": 1.57552335976524e-05,
"epoch": 2.0138931204679786,
"step": 33050
},
{
"loss": 0.2342,
"grad_norm": 15.659651756286621,
"learning_rate": 1.57472158302799e-05,
"epoch": 2.016939857412711,
"step": 33100
},
{
"loss": 0.2177,
"grad_norm": 14.447957038879395,
"learning_rate": 1.5739198062907403e-05,
"epoch": 2.019986594357443,
"step": 33150
},
{
"loss": 0.2363,
"grad_norm": 4.597593784332275,
"learning_rate": 1.5731180295534906e-05,
"epoch": 2.0230333313021753,
"step": 33200
},
{
"loss": 0.2748,
"grad_norm": 10.542444229125977,
"learning_rate": 1.572316252816241e-05,
"epoch": 2.0260800682469076,
"step": 33250
},
{
"loss": 0.2198,
"grad_norm": 10.905604362487793,
"learning_rate": 1.571514476078991e-05,
"epoch": 2.02912680519164,
"step": 33300
},
{
"loss": 0.2087,
"grad_norm": 0.9764786958694458,
"learning_rate": 1.5707126993417414e-05,
"epoch": 2.032173542136372,
"step": 33350
},
{
"loss": 0.2254,
"grad_norm": 5.941742897033691,
"learning_rate": 1.5699109226044916e-05,
"epoch": 2.0352202790811043,
"step": 33400
},
{
"loss": 0.1647,
"grad_norm": 4.492490291595459,
"learning_rate": 1.569109145867242e-05,
"epoch": 2.0382670160258365,
"step": 33450
},
{
"loss": 0.265,
"grad_norm": 4.393040657043457,
"learning_rate": 1.568307369129992e-05,
"epoch": 2.0413137529705687,
"step": 33500
},
{
"loss": 0.2222,
"grad_norm": 1.1467838287353516,
"learning_rate": 1.5675055923927424e-05,
"epoch": 2.0443604899153005,
"step": 33550
},
{
"loss": 0.2591,
"grad_norm": 7.400262355804443,
"learning_rate": 1.566703815655493e-05,
"epoch": 2.0474072268600327,
"step": 33600
},
{
"loss": 0.2321,
"grad_norm": 11.20752239227295,
"learning_rate": 1.565902038918243e-05,
"epoch": 2.050453963804765,
"step": 33650
},
{
"loss": 0.2041,
"grad_norm": 3.0586113929748535,
"learning_rate": 1.565100262180993e-05,
"epoch": 2.053500700749497,
"step": 33700
},
{
"loss": 0.2192,
"grad_norm": 8.322057723999023,
"learning_rate": 1.5642984854437434e-05,
"epoch": 2.0565474376942294,
"step": 33750
},
{
"loss": 0.2532,
"grad_norm": 16.07746696472168,
"learning_rate": 1.5634967087064936e-05,
"epoch": 2.0595941746389617,
"step": 33800
},
{
"loss": 0.2638,
"grad_norm": 1.6392818689346313,
"learning_rate": 1.562694931969244e-05,
"epoch": 2.062640911583694,
"step": 33850
},
{
"loss": 0.2703,
"grad_norm": 2.3903887271881104,
"learning_rate": 1.561893155231994e-05,
"epoch": 2.065687648528426,
"step": 33900
},
{
"loss": 0.2887,
"grad_norm": 8.792168617248535,
"learning_rate": 1.5610913784947447e-05,
"epoch": 2.0687343854731584,
"step": 33950
},
{
"loss": 0.2203,
"grad_norm": 3.2719595432281494,
"learning_rate": 1.5602896017574946e-05,
"epoch": 2.0717811224178906,
"step": 34000
},
{
"loss": 0.233,
"grad_norm": 4.4483842849731445,
"learning_rate": 1.559487825020245e-05,
"epoch": 2.074827859362623,
"step": 34050
},
{
"loss": 0.2316,
"grad_norm": 5.09189510345459,
"learning_rate": 1.5586860482829954e-05,
"epoch": 2.077874596307355,
"step": 34100
},
{
"loss": 0.266,
"grad_norm": 2.4297544956207275,
"learning_rate": 1.5578842715457454e-05,
"epoch": 2.080921333252087,
"step": 34150
},
{
"loss": 0.2364,
"grad_norm": 21.28923797607422,
"learning_rate": 1.5570824948084956e-05,
"epoch": 2.083968070196819,
"step": 34200
},
{
"loss": 0.2273,
"grad_norm": 23.931238174438477,
"learning_rate": 1.5562807180712462e-05,
"epoch": 2.0870148071415513,
"step": 34250
},
{
"loss": 0.2309,
"grad_norm": 16.536245346069336,
"learning_rate": 1.5554789413339964e-05,
"epoch": 2.0900615440862835,
"step": 34300
},
{
"loss": 0.2384,
"grad_norm": 14.541397094726562,
"learning_rate": 1.5546771645967464e-05,
"epoch": 2.0931082810310158,
"step": 34350
},
{
"loss": 0.2187,
"grad_norm": 0.32076966762542725,
"learning_rate": 1.5538753878594966e-05,
"epoch": 2.096155017975748,
"step": 34400
},
{
"loss": 0.1984,
"grad_norm": 13.719730377197266,
"learning_rate": 1.5530736111222472e-05,
"epoch": 2.0992017549204802,
"step": 34450
},
{
"loss": 0.2252,
"grad_norm": 9.205267906188965,
"learning_rate": 1.552271834384997e-05,
"epoch": 2.1022484918652125,
"step": 34500
},
{
"loss": 0.3275,
"grad_norm": 8.831197738647461,
"learning_rate": 1.5514700576477474e-05,
"epoch": 2.1052952288099447,
"step": 34550
},
{
"loss": 0.266,
"grad_norm": 6.3315300941467285,
"learning_rate": 1.550668280910498e-05,
"epoch": 2.108341965754677,
"step": 34600
},
{
"loss": 0.2866,
"grad_norm": 6.115374565124512,
"learning_rate": 1.5498665041732482e-05,
"epoch": 2.111388702699409,
"step": 34650
},
{
"loss": 0.235,
"grad_norm": 10.484439849853516,
"learning_rate": 1.549064727435998e-05,
"epoch": 2.114435439644141,
"step": 34700
},
{
"loss": 0.2099,
"grad_norm": 7.176803112030029,
"learning_rate": 1.5482629506987487e-05,
"epoch": 2.117482176588873,
"step": 34750
},
{
"loss": 0.2483,
"grad_norm": 15.840865135192871,
"learning_rate": 1.547461173961499e-05,
"epoch": 2.1205289135336054,
"step": 34800
},
{
"loss": 0.2701,
"grad_norm": 14.903778076171875,
"learning_rate": 1.546659397224249e-05,
"epoch": 2.1235756504783376,
"step": 34850
},
{
"loss": 0.2273,
"grad_norm": 9.21725082397461,
"learning_rate": 1.545857620486999e-05,
"epoch": 2.12662238742307,
"step": 34900
},
{
"loss": 0.265,
"grad_norm": 1.673561692237854,
"learning_rate": 1.5450558437497497e-05,
"epoch": 2.129669124367802,
"step": 34950
},
{
"loss": 0.3446,
"grad_norm": 5.9844651222229,
"learning_rate": 1.5442540670125e-05,
"epoch": 2.1327158613125343,
"step": 35000
},
{
"loss": 0.2485,
"grad_norm": 10.696561813354492,
"learning_rate": 1.54345229027525e-05,
"epoch": 2.1357625982572666,
"step": 35050
},
{
"loss": 0.182,
"grad_norm": 4.449814796447754,
"learning_rate": 1.5426505135380005e-05,
"epoch": 2.138809335201999,
"step": 35100
},
{
"loss": 0.2678,
"grad_norm": 18.055715560913086,
"learning_rate": 1.5418487368007507e-05,
"epoch": 2.141856072146731,
"step": 35150
},
{
"loss": 0.2773,
"grad_norm": 7.977348804473877,
"learning_rate": 1.5410469600635006e-05,
"epoch": 2.1449028090914632,
"step": 35200
},
{
"loss": 0.2164,
"grad_norm": 10.132824897766113,
"learning_rate": 1.5402451833262512e-05,
"epoch": 2.147949546036195,
"step": 35250
},
{
"loss": 0.2381,
"grad_norm": 0.2754250764846802,
"learning_rate": 1.5394434065890015e-05,
"epoch": 2.1509962829809273,
"step": 35300
},
{
"loss": 0.2405,
"grad_norm": 0.3685288429260254,
"learning_rate": 1.5386416298517517e-05,
"epoch": 2.1540430199256595,
"step": 35350
},
{
"loss": 0.2924,
"grad_norm": 1.8642265796661377,
"learning_rate": 1.537855888649247e-05,
"epoch": 2.1570897568703917,
"step": 35400
},
{
"loss": 0.2044,
"grad_norm": 3.759467601776123,
"learning_rate": 1.537054111911997e-05,
"epoch": 2.160136493815124,
"step": 35450
},
{
"loss": 0.2565,
"grad_norm": 14.311750411987305,
"learning_rate": 1.5362523351747474e-05,
"epoch": 2.163183230759856,
"step": 35500
},
{
"loss": 0.2479,
"grad_norm": 0.6184712052345276,
"learning_rate": 1.5354505584374976e-05,
"epoch": 2.1662299677045884,
"step": 35550
},
{
"loss": 0.2386,
"grad_norm": 8.407564163208008,
"learning_rate": 1.534648781700248e-05,
"epoch": 2.1692767046493207,
"step": 35600
},
{
"loss": 0.2802,
"grad_norm": 4.442209243774414,
"learning_rate": 1.533847004962998e-05,
"epoch": 2.172323441594053,
"step": 35650
},
{
"loss": 0.1952,
"grad_norm": 10.408587455749512,
"learning_rate": 1.5330452282257484e-05,
"epoch": 2.175370178538785,
"step": 35700
},
{
"loss": 0.2405,
"grad_norm": 5.826844692230225,
"learning_rate": 1.5322434514884986e-05,
"epoch": 2.1784169154835173,
"step": 35750
},
{
"loss": 0.2955,
"grad_norm": 3.78114652633667,
"learning_rate": 1.531441674751249e-05,
"epoch": 2.181463652428249,
"step": 35800
},
{
"loss": 0.2799,
"grad_norm": 15.427207946777344,
"learning_rate": 1.530639898013999e-05,
"epoch": 2.1845103893729814,
"step": 35850
},
{
"loss": 0.2719,
"grad_norm": 9.67601203918457,
"learning_rate": 1.5298381212767494e-05,
"epoch": 2.1875571263177136,
"step": 35900
},
{
"loss": 0.2222,
"grad_norm": 4.227503299713135,
"learning_rate": 1.5290363445394996e-05,
"epoch": 2.190603863262446,
"step": 35950
},
{
"loss": 0.2129,
"grad_norm": 10.22238540649414,
"learning_rate": 1.52823456780225e-05,
"epoch": 2.193650600207178,
"step": 36000
},
{
"loss": 0.2014,
"grad_norm": 10.769732475280762,
"learning_rate": 1.527432791065e-05,
"epoch": 2.1966973371519103,
"step": 36050
},
{
"loss": 0.2307,
"grad_norm": 5.7657599449157715,
"learning_rate": 1.5266310143277504e-05,
"epoch": 2.1997440740966425,
"step": 36100
},
{
"loss": 0.2273,
"grad_norm": 8.071337699890137,
"learning_rate": 1.5258292375905006e-05,
"epoch": 2.2027908110413748,
"step": 36150
},
{
"loss": 0.2398,
"grad_norm": 4.238720893859863,
"learning_rate": 1.5250274608532509e-05,
"epoch": 2.205837547986107,
"step": 36200
},
{
"loss": 0.2166,
"grad_norm": 20.695180892944336,
"learning_rate": 1.5242256841160013e-05,
"epoch": 2.208884284930839,
"step": 36250
},
{
"loss": 0.2181,
"grad_norm": 7.90350866317749,
"learning_rate": 1.5234239073787514e-05,
"epoch": 2.2119310218755714,
"step": 36300
},
{
"loss": 0.2435,
"grad_norm": 12.91663646697998,
"learning_rate": 1.5226221306415017e-05,
"epoch": 2.2149777588203037,
"step": 36350
},
{
"loss": 0.2604,
"grad_norm": 7.939602851867676,
"learning_rate": 1.5218203539042519e-05,
"epoch": 2.2180244957650355,
"step": 36400
},
{
"loss": 0.2515,
"grad_norm": 0.2701917588710785,
"learning_rate": 1.5210185771670023e-05,
"epoch": 2.2210712327097677,
"step": 36450
},
{
"loss": 0.2099,
"grad_norm": 15.202261924743652,
"learning_rate": 1.5202168004297524e-05,
"epoch": 2.2241179696545,
"step": 36500
},
{
"loss": 0.207,
"grad_norm": 10.874486923217773,
"learning_rate": 1.5194150236925027e-05,
"epoch": 2.227164706599232,
"step": 36550
},
{
"loss": 0.2383,
"grad_norm": 14.08298110961914,
"learning_rate": 1.518613246955253e-05,
"epoch": 2.2302114435439644,
"step": 36600
},
{
"loss": 0.2533,
"grad_norm": 8.34710693359375,
"learning_rate": 1.5178114702180032e-05,
"epoch": 2.2332581804886966,
"step": 36650
},
{
"loss": 0.2377,
"grad_norm": 14.197402954101562,
"learning_rate": 1.5170096934807534e-05,
"epoch": 2.236304917433429,
"step": 36700
},
{
"loss": 0.2527,
"grad_norm": 24.362600326538086,
"learning_rate": 1.5162079167435038e-05,
"epoch": 2.239351654378161,
"step": 36750
},
{
"loss": 0.252,
"grad_norm": 8.233315467834473,
"learning_rate": 1.515406140006254e-05,
"epoch": 2.2423983913228933,
"step": 36800
},
{
"loss": 0.2576,
"grad_norm": 3.1376678943634033,
"learning_rate": 1.5146043632690042e-05,
"epoch": 2.2454451282676255,
"step": 36850
},
{
"loss": 0.2034,
"grad_norm": 7.313117027282715,
"learning_rate": 1.5138025865317546e-05,
"epoch": 2.2484918652123573,
"step": 36900
},
{
"loss": 0.3184,
"grad_norm": 10.775823593139648,
"learning_rate": 1.5130008097945048e-05,
"epoch": 2.2515386021570896,
"step": 36950
},
{
"loss": 0.2403,
"grad_norm": 5.567470073699951,
"learning_rate": 1.5121990330572549e-05,
"epoch": 2.254585339101822,
"step": 37000
},
{
"loss": 0.1377,
"grad_norm": 8.842394828796387,
"learning_rate": 1.5113972563200052e-05,
"epoch": 2.257632076046554,
"step": 37050
},
{
"loss": 0.2504,
"grad_norm": 14.804726600646973,
"learning_rate": 1.5105954795827556e-05,
"epoch": 2.2606788129912863,
"step": 37100
},
{
"loss": 0.2565,
"grad_norm": 7.487795352935791,
"learning_rate": 1.5097937028455058e-05,
"epoch": 2.2637255499360185,
"step": 37150
},
{
"loss": 0.2506,
"grad_norm": 3.1857199668884277,
"learning_rate": 1.5089919261082559e-05,
"epoch": 2.2667722868807507,
"step": 37200
},
{
"loss": 0.2151,
"grad_norm": 6.043691158294678,
"learning_rate": 1.5081901493710063e-05,
"epoch": 2.269819023825483,
"step": 37250
},
{
"loss": 0.2206,
"grad_norm": 9.865518569946289,
"learning_rate": 1.5073883726337566e-05,
"epoch": 2.272865760770215,
"step": 37300
},
{
"loss": 0.2363,
"grad_norm": 10.448070526123047,
"learning_rate": 1.5065865958965067e-05,
"epoch": 2.2759124977149474,
"step": 37350
},
{
"loss": 0.2323,
"grad_norm": 9.892102241516113,
"learning_rate": 1.505784819159257e-05,
"epoch": 2.2789592346596796,
"step": 37400
},
{
"loss": 0.2373,
"grad_norm": 11.951998710632324,
"learning_rate": 1.5049990779567523e-05,
"epoch": 2.282005971604412,
"step": 37450
},
{
"loss": 0.1847,
"grad_norm": 6.93466329574585,
"learning_rate": 1.5041973012195025e-05,
"epoch": 2.285052708549144,
"step": 37500
},
{
"loss": 0.254,
"grad_norm": 3.111926794052124,
"learning_rate": 1.503395524482253e-05,
"epoch": 2.288099445493876,
"step": 37550
},
{
"loss": 0.2256,
"grad_norm": 12.360371589660645,
"learning_rate": 1.502593747745003e-05,
"epoch": 2.291146182438608,
"step": 37600
},
{
"loss": 0.225,
"grad_norm": 4.327396869659424,
"learning_rate": 1.5017919710077533e-05,
"epoch": 2.2941929193833404,
"step": 37650
},
{
"loss": 0.2181,
"grad_norm": 4.972620010375977,
"learning_rate": 1.5009901942705037e-05,
"epoch": 2.2972396563280726,
"step": 37700
},
{
"loss": 0.2587,
"grad_norm": 9.710999488830566,
"learning_rate": 1.5001884175332538e-05,
"epoch": 2.300286393272805,
"step": 37750
},
{
"loss": 0.2628,
"grad_norm": 13.191581726074219,
"learning_rate": 1.499386640796004e-05,
"epoch": 2.303333130217537,
"step": 37800
},
{
"loss": 0.2067,
"grad_norm": 9.240971565246582,
"learning_rate": 1.4985848640587544e-05,
"epoch": 2.3063798671622693,
"step": 37850
},
{
"loss": 0.3047,
"grad_norm": 4.321995735168457,
"learning_rate": 1.4977830873215047e-05,
"epoch": 2.3094266041070015,
"step": 37900
},
{
"loss": 0.2328,
"grad_norm": 9.038594245910645,
"learning_rate": 1.4969813105842548e-05,
"epoch": 2.3124733410517337,
"step": 37950
},
{
"loss": 0.2224,
"grad_norm": 13.457297325134277,
"learning_rate": 1.496179533847005e-05,
"epoch": 2.315520077996466,
"step": 38000
},
{
"loss": 0.2392,
"grad_norm": 4.995737075805664,
"learning_rate": 1.4953777571097554e-05,
"epoch": 2.3185668149411978,
"step": 38050
},
{
"loss": 0.2414,
"grad_norm": 0.7766318321228027,
"learning_rate": 1.4945759803725055e-05,
"epoch": 2.32161355188593,
"step": 38100
},
{
"loss": 0.2569,
"grad_norm": 1.4535309076309204,
"learning_rate": 1.4937742036352558e-05,
"epoch": 2.3246602888306622,
"step": 38150
},
{
"loss": 0.2207,
"grad_norm": 11.94556713104248,
"learning_rate": 1.4929724268980062e-05,
"epoch": 2.3277070257753945,
"step": 38200
},
{
"loss": 0.3229,
"grad_norm": 0.12980392575263977,
"learning_rate": 1.4921706501607564e-05,
"epoch": 2.3307537627201267,
"step": 38250
},
{
"loss": 0.2405,
"grad_norm": 3.995821952819824,
"learning_rate": 1.4913688734235065e-05,
"epoch": 2.333800499664859,
"step": 38300
},
{
"loss": 0.2017,
"grad_norm": 19.065113067626953,
"learning_rate": 1.490567096686257e-05,
"epoch": 2.336847236609591,
"step": 38350
},
{
"loss": 0.2514,
"grad_norm": 15.96821403503418,
"learning_rate": 1.4897653199490072e-05,
"epoch": 2.3398939735543234,
"step": 38400
},
{
"loss": 0.2575,
"grad_norm": 7.38139533996582,
"learning_rate": 1.4889635432117573e-05,
"epoch": 2.3429407104990556,
"step": 38450
},
{
"loss": 0.3041,
"grad_norm": 12.211630821228027,
"learning_rate": 1.4881617664745075e-05,
"epoch": 2.345987447443788,
"step": 38500
},
{
"loss": 0.1947,
"grad_norm": 2.677257776260376,
"learning_rate": 1.487359989737258e-05,
"epoch": 2.34903418438852,
"step": 38550
},
{
"loss": 0.1983,
"grad_norm": 0.721404492855072,
"learning_rate": 1.4865582130000082e-05,
"epoch": 2.3520809213332523,
"step": 38600
},
{
"loss": 0.2458,
"grad_norm": 9.244544982910156,
"learning_rate": 1.4857564362627583e-05,
"epoch": 2.355127658277984,
"step": 38650
},
{
"loss": 0.2126,
"grad_norm": 6.6306891441345215,
"learning_rate": 1.4849546595255087e-05,
"epoch": 2.3581743952227163,
"step": 38700
},
{
"loss": 0.227,
"grad_norm": 14.417181968688965,
"learning_rate": 1.484152882788259e-05,
"epoch": 2.3612211321674486,
"step": 38750
},
{
"loss": 0.2575,
"grad_norm": 12.214167594909668,
"learning_rate": 1.483351106051009e-05,
"epoch": 2.364267869112181,
"step": 38800
},
{
"loss": 0.2015,
"grad_norm": 8.66244125366211,
"learning_rate": 1.4825493293137594e-05,
"epoch": 2.367314606056913,
"step": 38850
},
{
"loss": 0.2114,
"grad_norm": 0.5778210759162903,
"learning_rate": 1.4817475525765097e-05,
"epoch": 2.3703613430016452,
"step": 38900
},
{
"loss": 0.2556,
"grad_norm": 4.011947154998779,
"learning_rate": 1.48094577583926e-05,
"epoch": 2.3734080799463775,
"step": 38950
},
{
"loss": 0.2484,
"grad_norm": 0.15622863173484802,
"learning_rate": 1.4801439991020102e-05,
"epoch": 2.3764548168911097,
"step": 39000
},
{
"loss": 0.1823,
"grad_norm": 4.739380836486816,
"learning_rate": 1.4793422223647604e-05,
"epoch": 2.379501553835842,
"step": 39050
},
{
"loss": 0.2146,
"grad_norm": 8.938774108886719,
"learning_rate": 1.4785404456275107e-05,
"epoch": 2.382548290780574,
"step": 39100
},
{
"loss": 0.1862,
"grad_norm": 10.488040924072266,
"learning_rate": 1.4777386688902608e-05,
"epoch": 2.385595027725306,
"step": 39150
},
{
"loss": 0.2314,
"grad_norm": 14.069398880004883,
"learning_rate": 1.4769368921530112e-05,
"epoch": 2.388641764670038,
"step": 39200
},
{
"loss": 0.2255,
"grad_norm": 6.238475799560547,
"learning_rate": 1.4761351154157614e-05,
"epoch": 2.3916885016147704,
"step": 39250
},
{
"loss": 0.2172,
"grad_norm": 4.873676300048828,
"learning_rate": 1.4753333386785119e-05,
"epoch": 2.3947352385595027,
"step": 39300
},
{
"loss": 0.2177,
"grad_norm": 4.463593482971191,
"learning_rate": 1.474531561941262e-05,
"epoch": 2.397781975504235,
"step": 39350
},
{
"loss": 0.2096,
"grad_norm": 14.323698997497559,
"learning_rate": 1.4737297852040122e-05,
"epoch": 2.400828712448967,
"step": 39400
},
{
"loss": 0.2643,
"grad_norm": 19.178359985351562,
"learning_rate": 1.4729280084667626e-05,
"epoch": 2.4038754493936993,
"step": 39450
},
{
"loss": 0.2416,
"grad_norm": 0.37063902616500854,
"learning_rate": 1.4721262317295127e-05,
"epoch": 2.4069221863384316,
"step": 39500
},
{
"loss": 0.2082,
"grad_norm": 4.054755687713623,
"learning_rate": 1.471324454992263e-05,
"epoch": 2.409968923283164,
"step": 39550
},
{
"loss": 0.2746,
"grad_norm": 18.93335723876953,
"learning_rate": 1.470554749324503e-05,
"epoch": 2.413015660227896,
"step": 39600
},
{
"loss": 0.2985,
"grad_norm": 6.553796291351318,
"learning_rate": 1.4697529725872535e-05,
"epoch": 2.4160623971726283,
"step": 39650
},
{
"loss": 0.226,
"grad_norm": 13.31601619720459,
"learning_rate": 1.4689511958500038e-05,
"epoch": 2.4191091341173605,
"step": 39700
},
{
"loss": 0.2529,
"grad_norm": 8.049433708190918,
"learning_rate": 1.4681494191127538e-05,
"epoch": 2.4221558710620927,
"step": 39750
},
{
"loss": 0.2457,
"grad_norm": 3.4679155349731445,
"learning_rate": 1.4673476423755043e-05,
"epoch": 2.4252026080068245,
"step": 39800
},
{
"loss": 0.1778,
"grad_norm": 10.506498336791992,
"learning_rate": 1.4665458656382545e-05,
"epoch": 2.4282493449515568,
"step": 39850
},
{
"loss": 0.2272,
"grad_norm": 2.551778554916382,
"learning_rate": 1.4657440889010048e-05,
"epoch": 2.431296081896289,
"step": 39900
},
{
"loss": 0.28,
"grad_norm": 15.0549955368042,
"learning_rate": 1.464942312163755e-05,
"epoch": 2.434342818841021,
"step": 39950
},
{
"loss": 0.1637,
"grad_norm": 3.878887891769409,
"learning_rate": 1.4641405354265053e-05,
"epoch": 2.4373895557857534,
"step": 40000
},
{
"loss": 0.1905,
"grad_norm": 10.837960243225098,
"learning_rate": 1.4633387586892555e-05,
"epoch": 2.4404362927304857,
"step": 40050
},
{
"loss": 0.2583,
"grad_norm": 1.017728567123413,
"learning_rate": 1.4625369819520056e-05,
"epoch": 2.443483029675218,
"step": 40100
},
{
"loss": 0.2114,
"grad_norm": 0.35611045360565186,
"learning_rate": 1.461735205214756e-05,
"epoch": 2.44652976661995,
"step": 40150
},
{
"loss": 0.2026,
"grad_norm": 7.648980617523193,
"learning_rate": 1.4609334284775063e-05,
"epoch": 2.4495765035646824,
"step": 40200
},
{
"loss": 0.1855,
"grad_norm": 20.588184356689453,
"learning_rate": 1.4601316517402567e-05,
"epoch": 2.4526232405094146,
"step": 40250
},
{
"loss": 0.2567,
"grad_norm": 6.03258752822876,
"learning_rate": 1.4593298750030068e-05,
"epoch": 2.4556699774541464,
"step": 40300
},
{
"loss": 0.2353,
"grad_norm": 7.953198432922363,
"learning_rate": 1.458528098265757e-05,
"epoch": 2.4587167143988786,
"step": 40350
},
{
"loss": 0.2108,
"grad_norm": 5.491347789764404,
"learning_rate": 1.4577263215285074e-05,
"epoch": 2.461763451343611,
"step": 40400
},
{
"loss": 0.2123,
"grad_norm": 2.999462127685547,
"learning_rate": 1.4569245447912575e-05,
"epoch": 2.464810188288343,
"step": 40450
},
{
"loss": 0.2812,
"grad_norm": 1.3468390703201294,
"learning_rate": 1.4561227680540078e-05,
"epoch": 2.4678569252330753,
"step": 40500
},
{
"loss": 0.2831,
"grad_norm": 1.91505765914917,
"learning_rate": 1.455320991316758e-05,
"epoch": 2.4709036621778075,
"step": 40550
},
{
"loss": 0.2008,
"grad_norm": 10.220423698425293,
"learning_rate": 1.4545192145795084e-05,
"epoch": 2.47395039912254,
"step": 40600
},
{
"loss": 0.1838,
"grad_norm": 6.893011569976807,
"learning_rate": 1.4537174378422585e-05,
"epoch": 2.476997136067272,
"step": 40650
},
{
"loss": 0.1758,
"grad_norm": 3.967205047607422,
"learning_rate": 1.4529156611050088e-05,
"epoch": 2.4800438730120042,
"step": 40700
},
{
"loss": 0.1505,
"grad_norm": 0.38301968574523926,
"learning_rate": 1.4521138843677592e-05,
"epoch": 2.4830906099567365,
"step": 40750
},
{
"loss": 0.206,
"grad_norm": 10.306224822998047,
"learning_rate": 1.4513121076305093e-05,
"epoch": 2.4861373469014687,
"step": 40800
},
{
"loss": 0.2383,
"grad_norm": 0.4564184546470642,
"learning_rate": 1.4505263664280045e-05,
"epoch": 2.489184083846201,
"step": 40850
},
{
"loss": 0.1968,
"grad_norm": 8.829743385314941,
"learning_rate": 1.4497245896907549e-05,
"epoch": 2.4922308207909327,
"step": 40900
},
{
"loss": 0.2515,
"grad_norm": 12.544764518737793,
"learning_rate": 1.4489228129535051e-05,
"epoch": 2.495277557735665,
"step": 40950
},
{
"loss": 0.2359,
"grad_norm": 1.0026172399520874,
"learning_rate": 1.4481210362162554e-05,
"epoch": 2.498324294680397,
"step": 41000
},
{
"loss": 0.2486,
"grad_norm": 3.4744980335235596,
"learning_rate": 1.4473192594790055e-05,
"epoch": 2.5013710316251294,
"step": 41050
},
{
"loss": 0.2258,
"grad_norm": 10.205848693847656,
"learning_rate": 1.4465174827417559e-05,
"epoch": 2.5044177685698616,
"step": 41100
},
{
"loss": 0.2746,
"grad_norm": 1.469072699546814,
"learning_rate": 1.4457157060045061e-05,
"epoch": 2.507464505514594,
"step": 41150
},
{
"loss": 0.2574,
"grad_norm": 0.7870273590087891,
"learning_rate": 1.4449139292672562e-05,
"epoch": 2.510511242459326,
"step": 41200
},
{
"loss": 0.2651,
"grad_norm": 10.614931106567383,
"learning_rate": 1.4441121525300066e-05,
"epoch": 2.5135579794040583,
"step": 41250
},
{
"loss": 0.2808,
"grad_norm": 17.5373592376709,
"learning_rate": 1.4433103757927569e-05,
"epoch": 2.5166047163487906,
"step": 41300
},
{
"loss": 0.2519,
"grad_norm": 7.552835941314697,
"learning_rate": 1.4425085990555073e-05,
"epoch": 2.519651453293523,
"step": 41350
},
{
"loss": 0.2081,
"grad_norm": 1.4148788452148438,
"learning_rate": 1.4417068223182574e-05,
"epoch": 2.5226981902382546,
"step": 41400
},
{
"loss": 0.2505,
"grad_norm": 2.076524257659912,
"learning_rate": 1.4409050455810076e-05,
"epoch": 2.525744927182987,
"step": 41450
},
{
"loss": 0.1786,
"grad_norm": 0.7545868158340454,
"learning_rate": 1.4401032688437579e-05,
"epoch": 2.528791664127719,
"step": 41500
},
{
"loss": 0.2878,
"grad_norm": 2.534148931503296,
"learning_rate": 1.4393014921065081e-05,
"epoch": 2.5318384010724513,
"step": 41550
},
{
"loss": 0.2286,
"grad_norm": 11.197412490844727,
"learning_rate": 1.4384997153692584e-05,
"epoch": 2.5348851380171835,
"step": 41600
},
{
"loss": 0.2234,
"grad_norm": 0.10376529395580292,
"learning_rate": 1.4376979386320086e-05,
"epoch": 2.5379318749619157,
"step": 41650
},
{
"loss": 0.2467,
"grad_norm": 9.725293159484863,
"learning_rate": 1.436896161894759e-05,
"epoch": 2.540978611906648,
"step": 41700
},
{
"loss": 0.2286,
"grad_norm": 2.4273295402526855,
"learning_rate": 1.4360943851575091e-05,
"epoch": 2.54402534885138,
"step": 41750
},
{
"loss": 0.2426,
"grad_norm": 5.440350532531738,
"learning_rate": 1.4352926084202594e-05,
"epoch": 2.5470720857961124,
"step": 41800
},
{
"loss": 0.2434,
"grad_norm": 7.610898017883301,
"learning_rate": 1.4344908316830098e-05,
"epoch": 2.5501188227408447,
"step": 41850
},
{
"loss": 0.2174,
"grad_norm": 5.901045322418213,
"learning_rate": 1.4336890549457599e-05,
"epoch": 2.553165559685577,
"step": 41900
},
{
"loss": 0.242,
"grad_norm": 24.036008834838867,
"learning_rate": 1.4328872782085101e-05,
"epoch": 2.556212296630309,
"step": 41950
},
{
"loss": 0.2723,
"grad_norm": 20.48126792907715,
"learning_rate": 1.4320855014712605e-05,
"epoch": 2.5592590335750414,
"step": 42000
},
{
"loss": 0.2294,
"grad_norm": 0.5027254819869995,
"learning_rate": 1.4312837247340108e-05,
"epoch": 2.5623057705197736,
"step": 42050
},
{
"loss": 0.2361,
"grad_norm": 9.130940437316895,
"learning_rate": 1.4304819479967609e-05,
"epoch": 2.5653525074645054,
"step": 42100
},
{
"loss": 0.2392,
"grad_norm": 33.22863006591797,
"learning_rate": 1.4296801712595111e-05,
"epoch": 2.5683992444092376,
"step": 42150
},
{
"loss": 0.2504,
"grad_norm": 15.486459732055664,
"learning_rate": 1.4288783945222615e-05,
"epoch": 2.57144598135397,
"step": 42200
},
{
"loss": 0.1908,
"grad_norm": 11.937169075012207,
"learning_rate": 1.4280766177850116e-05,
"epoch": 2.574492718298702,
"step": 42250
},
{
"loss": 0.1658,
"grad_norm": 1.3168563842773438,
"learning_rate": 1.4272748410477619e-05,
"epoch": 2.5775394552434343,
"step": 42300
},
{
"loss": 0.3314,
"grad_norm": 10.791857719421387,
"learning_rate": 1.4264730643105123e-05,
"epoch": 2.5805861921881665,
"step": 42350
},
{
"loss": 0.2634,
"grad_norm": 11.966769218444824,
"learning_rate": 1.4256712875732626e-05,
"epoch": 2.5836329291328988,
"step": 42400
},
{
"loss": 0.2187,
"grad_norm": 10.748601913452148,
"learning_rate": 1.4248695108360126e-05,
"epoch": 2.586679666077631,
"step": 42450
},
{
"loss": 0.2326,
"grad_norm": 0.5902921557426453,
"learning_rate": 1.424067734098763e-05,
"epoch": 2.589726403022363,
"step": 42500
},
{
"loss": 0.2058,
"grad_norm": 9.101327896118164,
"learning_rate": 1.4232659573615133e-05,
"epoch": 2.592773139967095,
"step": 42550
},
{
"loss": 0.208,
"grad_norm": 1.8679925203323364,
"learning_rate": 1.4224641806242634e-05,
"epoch": 2.5958198769118273,
"step": 42600
},
{
"loss": 0.225,
"grad_norm": 6.802280426025391,
"learning_rate": 1.4216624038870136e-05,
"epoch": 2.5988666138565595,
"step": 42650
},
{
"loss": 0.2006,
"grad_norm": 20.325990676879883,
"learning_rate": 1.420860627149764e-05,
"epoch": 2.6019133508012917,
"step": 42700
},
{
"loss": 0.2167,
"grad_norm": 16.06829261779785,
"learning_rate": 1.4200588504125143e-05,
"epoch": 2.604960087746024,
"step": 42750
},
{
"loss": 0.2274,
"grad_norm": 7.85139799118042,
"learning_rate": 1.4192570736752644e-05,
"epoch": 2.608006824690756,
"step": 42800
},
{
"loss": 0.2443,
"grad_norm": 7.194572448730469,
"learning_rate": 1.4184552969380148e-05,
"epoch": 2.6110535616354884,
"step": 42850
},
{
"loss": 0.2192,
"grad_norm": 12.887028694152832,
"learning_rate": 1.417653520200765e-05,
"epoch": 2.6141002985802206,
"step": 42900
},
{
"loss": 0.2621,
"grad_norm": 14.314068794250488,
"learning_rate": 1.4168838145330054e-05,
"epoch": 2.617147035524953,
"step": 42950
},
{
"loss": 0.1855,
"grad_norm": 14.639540672302246,
"learning_rate": 1.4160820377957556e-05,
"epoch": 2.620193772469685,
"step": 43000
},
{
"loss": 0.2027,
"grad_norm": 1.6619808673858643,
"learning_rate": 1.4152802610585057e-05,
"epoch": 2.6232405094144173,
"step": 43050
},
{
"loss": 0.2798,
"grad_norm": 6.387289524078369,
"learning_rate": 1.414478484321256e-05,
"epoch": 2.6262872463591496,
"step": 43100
},
{
"loss": 0.2468,
"grad_norm": 13.067091941833496,
"learning_rate": 1.4136767075840064e-05,
"epoch": 2.629333983303882,
"step": 43150
},
{
"loss": 0.2246,
"grad_norm": 19.330541610717773,
"learning_rate": 1.4128749308467564e-05,
"epoch": 2.6323807202486136,
"step": 43200
},
{
"loss": 0.1766,
"grad_norm": 0.6762033104896545,
"learning_rate": 1.4120731541095067e-05,
"epoch": 2.635427457193346,
"step": 43250
},
{
"loss": 0.2381,
"grad_norm": 6.1050872802734375,
"learning_rate": 1.4112713773722571e-05,
"epoch": 2.638474194138078,
"step": 43300
},
{
"loss": 0.1681,
"grad_norm": 40.20041275024414,
"learning_rate": 1.4104696006350074e-05,
"epoch": 2.6415209310828103,
"step": 43350
},
{
"loss": 0.2143,
"grad_norm": 13.794843673706055,
"learning_rate": 1.4096678238977574e-05,
"epoch": 2.6445676680275425,
"step": 43400
},
{
"loss": 0.2838,
"grad_norm": 11.035282135009766,
"learning_rate": 1.4088660471605079e-05,
"epoch": 2.6476144049722747,
"step": 43450
},
{
"loss": 0.2251,
"grad_norm": 24.257375717163086,
"learning_rate": 1.4080642704232581e-05,
"epoch": 2.650661141917007,
"step": 43500
},
{
"loss": 0.1904,
"grad_norm": 2.735748291015625,
"learning_rate": 1.4072624936860082e-05,
"epoch": 2.653707878861739,
"step": 43550
},
{
"loss": 0.2234,
"grad_norm": 19.57889175415039,
"learning_rate": 1.4064607169487584e-05,
"epoch": 2.6567546158064714,
"step": 43600
},
{
"loss": 0.2636,
"grad_norm": 3.166445255279541,
"learning_rate": 1.4056589402115089e-05,
"epoch": 2.659801352751203,
"step": 43650
},
{
"loss": 0.2065,
"grad_norm": 4.286762714385986,
"learning_rate": 1.4048571634742591e-05,
"epoch": 2.6628480896959354,
"step": 43700
},
{
"loss": 0.2343,
"grad_norm": 14.490877151489258,
"learning_rate": 1.4040553867370092e-05,
"epoch": 2.6658948266406677,
"step": 43750
},
{
"loss": 0.1894,
"grad_norm": 9.28808307647705,
"learning_rate": 1.4032536099997596e-05,
"epoch": 2.6689415635854,
"step": 43800
},
{
"loss": 0.2441,
"grad_norm": 9.927886962890625,
"learning_rate": 1.4024518332625099e-05,
"epoch": 2.671988300530132,
"step": 43850
},
{
"loss": 0.2248,
"grad_norm": 9.022445678710938,
"learning_rate": 1.40165005652526e-05,
"epoch": 2.6750350374748644,
"step": 43900
},
{
"loss": 0.2295,
"grad_norm": 0.3976965844631195,
"learning_rate": 1.4008482797880104e-05,
"epoch": 2.6780817744195966,
"step": 43950
},
{
"loss": 0.199,
"grad_norm": 5.68612003326416,
"learning_rate": 1.4000465030507606e-05,
"epoch": 2.681128511364329,
"step": 44000
},
{
"loss": 0.2374,
"grad_norm": 8.631611824035645,
"learning_rate": 1.3992447263135109e-05,
"epoch": 2.684175248309061,
"step": 44050
},
{
"loss": 0.1961,
"grad_norm": 3.5786478519439697,
"learning_rate": 1.3984429495762611e-05,
"epoch": 2.6872219852537933,
"step": 44100
},
{
"loss": 0.1708,
"grad_norm": 0.286063551902771,
"learning_rate": 1.3976411728390114e-05,
"epoch": 2.6902687221985255,
"step": 44150
},
{
"loss": 0.2633,
"grad_norm": 5.382233619689941,
"learning_rate": 1.3968393961017616e-05,
"epoch": 2.6933154591432578,
"step": 44200
},
{
"loss": 0.2629,
"grad_norm": 11.39046573638916,
"learning_rate": 1.3960376193645117e-05,
"epoch": 2.69636219608799,
"step": 44250
},
{
"loss": 0.1871,
"grad_norm": 2.758930206298828,
"learning_rate": 1.3952358426272621e-05,
"epoch": 2.6994089330327222,
"step": 44300
},
{
"loss": 0.2328,
"grad_norm": 11.423262596130371,
"learning_rate": 1.3944340658900124e-05,
"epoch": 2.702455669977454,
"step": 44350
},
{
"loss": 0.2014,
"grad_norm": 8.583510398864746,
"learning_rate": 1.3936322891527628e-05,
"epoch": 2.7055024069221862,
"step": 44400
},
{
"loss": 0.2547,
"grad_norm": 1.3459137678146362,
"learning_rate": 1.3928305124155129e-05,
"epoch": 2.7085491438669185,
"step": 44450
},
{
"loss": 0.1992,
"grad_norm": 7.409167289733887,
"learning_rate": 1.3920287356782631e-05,
"epoch": 2.7115958808116507,
"step": 44500
},
{
"loss": 0.2291,
"grad_norm": 2.709199905395508,
"learning_rate": 1.3912269589410134e-05,
"epoch": 2.714642617756383,
"step": 44550
},
{
"loss": 0.2163,
"grad_norm": 10.715925216674805,
"learning_rate": 1.3904251822037636e-05,
"epoch": 2.717689354701115,
"step": 44600
},
{
"loss": 0.193,
"grad_norm": 1.0097488164901733,
"learning_rate": 1.3896234054665139e-05,
"epoch": 2.7207360916458474,
"step": 44650
},
{
"loss": 0.2444,
"grad_norm": 5.5834879875183105,
"learning_rate": 1.3888216287292641e-05,
"epoch": 2.7237828285905796,
"step": 44700
},
{
"loss": 0.2572,
"grad_norm": 5.8961405754089355,
"learning_rate": 1.3880198519920145e-05,
"epoch": 2.7268295655353114,
"step": 44750
},
{
"loss": 0.2439,
"grad_norm": 2.6805806159973145,
"learning_rate": 1.3872180752547646e-05,
"epoch": 2.7298763024800436,
"step": 44800
},
{
"loss": 0.2293,
"grad_norm": 9.95940113067627,
"learning_rate": 1.3864162985175149e-05,
"epoch": 2.732923039424776,
"step": 44850
},
{
"loss": 0.2258,
"grad_norm": 6.230493545532227,
"learning_rate": 1.3856145217802653e-05,
"epoch": 2.735969776369508,
"step": 44900
},
{
"loss": 0.1824,
"grad_norm": 9.444258689880371,
"learning_rate": 1.3848127450430154e-05,
"epoch": 2.7390165133142403,
"step": 44950
},
{
"loss": 0.1832,
"grad_norm": 6.2765350341796875,
"learning_rate": 1.3840109683057656e-05,
"epoch": 2.7420632502589726,
"step": 45000
},
{
"loss": 0.2143,
"grad_norm": 5.38384485244751,
"learning_rate": 1.383209191568516e-05,
"epoch": 2.745109987203705,
"step": 45050
},
{
"loss": 0.1992,
"grad_norm": 15.249722480773926,
"learning_rate": 1.3824074148312663e-05,
"epoch": 2.748156724148437,
"step": 45100
},
{
"loss": 0.2824,
"grad_norm": 3.452012300491333,
"learning_rate": 1.3816056380940164e-05,
"epoch": 2.7512034610931693,
"step": 45150
},
{
"loss": 0.2879,
"grad_norm": 7.320451736450195,
"learning_rate": 1.3808038613567666e-05,
"epoch": 2.7542501980379015,
"step": 45200
},
{
"loss": 0.1857,
"grad_norm": 5.787985801696777,
"learning_rate": 1.380002084619517e-05,
"epoch": 2.7572969349826337,
"step": 45250
},
{
"loss": 0.2611,
"grad_norm": 14.351680755615234,
"learning_rate": 1.3792003078822671e-05,
"epoch": 2.760343671927366,
"step": 45300
},
{
"loss": 0.2231,
"grad_norm": 6.398861408233643,
"learning_rate": 1.3783985311450174e-05,
"epoch": 2.763390408872098,
"step": 45350
},
{
"loss": 0.2291,
"grad_norm": 24.750450134277344,
"learning_rate": 1.3775967544077678e-05,
"epoch": 2.7664371458168304,
"step": 45400
},
{
"loss": 0.2066,
"grad_norm": 7.509747505187988,
"learning_rate": 1.376794977670518e-05,
"epoch": 2.769483882761562,
"step": 45450
},
{
"loss": 0.1675,
"grad_norm": 8.170595169067383,
"learning_rate": 1.3759932009332681e-05,
"epoch": 2.7725306197062944,
"step": 45500
},
{
"loss": 0.2474,
"grad_norm": 8.114269256591797,
"learning_rate": 1.3751914241960185e-05,
"epoch": 2.7755773566510267,
"step": 45550
},
{
"loss": 0.2179,
"grad_norm": 12.927825927734375,
"learning_rate": 1.3743896474587688e-05,
"epoch": 2.778624093595759,
"step": 45600
},
{
"loss": 0.1867,
"grad_norm": 13.283806800842285,
"learning_rate": 1.3735878707215189e-05,
"epoch": 2.781670830540491,
"step": 45650
},
{
"loss": 0.3185,
"grad_norm": 22.413345336914062,
"learning_rate": 1.3727860939842691e-05,
"epoch": 2.7847175674852234,
"step": 45700
},
{
"loss": 0.1511,
"grad_norm": 5.857728481292725,
"learning_rate": 1.3719843172470195e-05,
"epoch": 2.7877643044299556,
"step": 45750
},
{
"loss": 0.2385,
"grad_norm": 17.186080932617188,
"learning_rate": 1.3711825405097698e-05,
"epoch": 2.790811041374688,
"step": 45800
},
{
"loss": 0.1742,
"grad_norm": 14.363241195678711,
"learning_rate": 1.3703807637725199e-05,
"epoch": 2.79385777831942,
"step": 45850
},
{
"loss": 0.2052,
"grad_norm": 6.93228816986084,
"learning_rate": 1.3695789870352703e-05,
"epoch": 2.796904515264152,
"step": 45900
},
{
"loss": 0.201,
"grad_norm": 6.34561014175415,
"learning_rate": 1.3687772102980205e-05,
"epoch": 2.799951252208884,
"step": 45950
},
{
"loss": 0.2583,
"grad_norm": 0.06626395136117935,
"learning_rate": 1.3679754335607706e-05,
"epoch": 2.8029979891536163,
"step": 46000
},
{
"loss": 0.1963,
"grad_norm": 0.5262823104858398,
"learning_rate": 1.367173656823521e-05,
"epoch": 2.8060447260983485,
"step": 46050
},
{
"loss": 0.2383,
"grad_norm": 8.523850440979004,
"learning_rate": 1.3663718800862713e-05,
"epoch": 2.8090914630430808,
"step": 46100
},
{
"loss": 0.1979,
"grad_norm": 7.935704708099365,
"learning_rate": 1.3655701033490215e-05,
"epoch": 2.812138199987813,
"step": 46150
},
{
"loss": 0.1915,
"grad_norm": 9.407266616821289,
"learning_rate": 1.3647683266117718e-05,
"epoch": 2.8151849369325452,
"step": 46200
},
{
"loss": 0.2166,
"grad_norm": 1.5281898975372314,
"learning_rate": 1.363966549874522e-05,
"epoch": 2.8182316738772775,
"step": 46250
},
{
"loss": 0.2592,
"grad_norm": 2.3613524436950684,
"learning_rate": 1.3631647731372723e-05,
"epoch": 2.8212784108220097,
"step": 46300
},
{
"loss": 0.2239,
"grad_norm": 0.8500705361366272,
"learning_rate": 1.3623629964000224e-05,
"epoch": 2.824325147766742,
"step": 46350
},
{
"loss": 0.2146,
"grad_norm": 0.070566825568676,
"learning_rate": 1.3615612196627728e-05,
"epoch": 2.827371884711474,
"step": 46400
},
{
"loss": 0.1954,
"grad_norm": 23.506328582763672,
"learning_rate": 1.360759442925523e-05,
"epoch": 2.8304186216562064,
"step": 46450
},
{
"loss": 0.2264,
"grad_norm": 20.97519302368164,
"learning_rate": 1.3599576661882735e-05,
"epoch": 2.8334653586009386,
"step": 46500
},
{
"loss": 0.1805,
"grad_norm": 0.30535706877708435,
"learning_rate": 1.3591558894510235e-05,
"epoch": 2.836512095545671,
"step": 46550
},
{
"loss": 0.2625,
"grad_norm": 3.5542941093444824,
"learning_rate": 1.3583541127137738e-05,
"epoch": 2.8395588324904026,
"step": 46600
},
{
"loss": 0.2002,
"grad_norm": 1.3966041803359985,
"learning_rate": 1.3575523359765242e-05,
"epoch": 2.842605569435135,
"step": 46650
},
{
"loss": 0.1988,
"grad_norm": 10.330328941345215,
"learning_rate": 1.3567505592392743e-05,
"epoch": 2.845652306379867,
"step": 46700
},
{
"loss": 0.2488,
"grad_norm": 5.144864559173584,
"learning_rate": 1.3559487825020245e-05,
"epoch": 2.8486990433245993,
"step": 46750
},
{
"loss": 0.2356,
"grad_norm": 7.551101207733154,
"learning_rate": 1.3551470057647748e-05,
"epoch": 2.8517457802693316,
"step": 46800
},
{
"loss": 0.2347,
"grad_norm": 5.450807571411133,
"learning_rate": 1.3543452290275252e-05,
"epoch": 2.854792517214064,
"step": 46850
},
{
"loss": 0.1866,
"grad_norm": 5.249316215515137,
"learning_rate": 1.3535434522902753e-05,
"epoch": 2.857839254158796,
"step": 46900
},
{
"loss": 0.2169,
"grad_norm": 4.130272388458252,
"learning_rate": 1.3527416755530256e-05,
"epoch": 2.8608859911035283,
"step": 46950
},
{
"loss": 0.2283,
"grad_norm": 16.014440536499023,
"learning_rate": 1.351939898815776e-05,
"epoch": 2.86393272804826,
"step": 47000
},
{
"loss": 0.173,
"grad_norm": 13.79580020904541,
"learning_rate": 1.351138122078526e-05,
"epoch": 2.8669794649929923,
"step": 47050
},
{
"loss": 0.2649,
"grad_norm": 3.7649166584014893,
"learning_rate": 1.3503363453412763e-05,
"epoch": 2.8700262019377245,
"step": 47100
},
{
"loss": 0.2226,
"grad_norm": 15.53331470489502,
"learning_rate": 1.3495345686040267e-05,
"epoch": 2.8730729388824567,
"step": 47150
},
{
"loss": 0.2888,
"grad_norm": 11.423954010009766,
"learning_rate": 1.348732791866777e-05,
"epoch": 2.876119675827189,
"step": 47200
},
{
"loss": 0.2101,
"grad_norm": 2.3906466960906982,
"learning_rate": 1.347931015129527e-05,
"epoch": 2.879166412771921,
"step": 47250
},
{
"loss": 0.2156,
"grad_norm": 3.551248788833618,
"learning_rate": 1.3471292383922773e-05,
"epoch": 2.8822131497166534,
"step": 47300
},
{
"loss": 0.1508,
"grad_norm": 17.19338607788086,
"learning_rate": 1.3463274616550277e-05,
"epoch": 2.8852598866613857,
"step": 47350
},
{
"loss": 0.2702,
"grad_norm": 3.8021812438964844,
"learning_rate": 1.3455256849177778e-05,
"epoch": 2.888306623606118,
"step": 47400
},
{
"loss": 0.2511,
"grad_norm": 15.451226234436035,
"learning_rate": 1.344723908180528e-05,
"epoch": 2.89135336055085,
"step": 47450
},
{
"loss": 0.1929,
"grad_norm": 7.5608296394348145,
"learning_rate": 1.3439221314432785e-05,
"epoch": 2.8944000974955824,
"step": 47500
},
{
"loss": 0.2385,
"grad_norm": 11.222097396850586,
"learning_rate": 1.3431203547060287e-05,
"epoch": 2.8974468344403146,
"step": 47550
},
{
"loss": 0.1939,
"grad_norm": 0.9156554937362671,
"learning_rate": 1.3423185779687788e-05,
"epoch": 2.900493571385047,
"step": 47600
},
{
"loss": 0.2226,
"grad_norm": 5.364518165588379,
"learning_rate": 1.3415168012315292e-05,
"epoch": 2.903540308329779,
"step": 47650
},
{
"loss": 0.194,
"grad_norm": 7.337564468383789,
"learning_rate": 1.3407150244942795e-05,
"epoch": 2.906587045274511,
"step": 47700
},
{
"loss": 0.2412,
"grad_norm": 1.8656667470932007,
"learning_rate": 1.3399132477570296e-05,
"epoch": 2.909633782219243,
"step": 47750
},
{
"loss": 0.2178,
"grad_norm": 11.3707914352417,
"learning_rate": 1.33911147101978e-05,
"epoch": 2.9126805191639753,
"step": 47800
},
{
"loss": 0.2436,
"grad_norm": 11.585225105285645,
"learning_rate": 1.3383096942825302e-05,
"epoch": 2.9157272561087075,
"step": 47850
},
{
"loss": 0.1734,
"grad_norm": 3.069232225418091,
"learning_rate": 1.3375079175452805e-05,
"epoch": 2.9187739930534398,
"step": 47900
},
{
"loss": 0.2054,
"grad_norm": 8.468469619750977,
"learning_rate": 1.3367061408080306e-05,
"epoch": 2.921820729998172,
"step": 47950
},
{
"loss": 0.1659,
"grad_norm": 5.904839038848877,
"learning_rate": 1.335904364070781e-05,
"epoch": 2.9248674669429042,
"step": 48000
},
{
"loss": 0.1675,
"grad_norm": 10.071244239807129,
"learning_rate": 1.3351025873335312e-05,
"epoch": 2.9279142038876365,
"step": 48050
},
{
"loss": 0.192,
"grad_norm": 4.885236740112305,
"learning_rate": 1.3343008105962813e-05,
"epoch": 2.9309609408323687,
"step": 48100
},
{
"loss": 0.226,
"grad_norm": 6.2569193840026855,
"learning_rate": 1.3334990338590317e-05,
"epoch": 2.9340076777771005,
"step": 48150
},
{
"loss": 0.2317,
"grad_norm": 2.048661708831787,
"learning_rate": 1.332697257121782e-05,
"epoch": 2.9370544147218327,
"step": 48200
},
{
"loss": 0.2323,
"grad_norm": 0.5361374616622925,
"learning_rate": 1.3318954803845322e-05,
"epoch": 2.940101151666565,
"step": 48250
},
{
"loss": 0.2325,
"grad_norm": 12.801497459411621,
"learning_rate": 1.3310937036472825e-05,
"epoch": 2.943147888611297,
"step": 48300
},
{
"loss": 0.259,
"grad_norm": 1.470711350440979,
"learning_rate": 1.3302919269100327e-05,
"epoch": 2.9461946255560294,
"step": 48350
},
{
"loss": 0.1774,
"grad_norm": 23.030967712402344,
"learning_rate": 1.329490150172783e-05,
"epoch": 2.9492413625007616,
"step": 48400
},
{
"loss": 0.2073,
"grad_norm": 3.9478507041931152,
"learning_rate": 1.328688373435533e-05,
"epoch": 2.952288099445494,
"step": 48450
},
{
"loss": 0.137,
"grad_norm": 9.52114486694336,
"learning_rate": 1.3278865966982835e-05,
"epoch": 2.955334836390226,
"step": 48500
},
{
"loss": 0.1911,
"grad_norm": 6.488351345062256,
"learning_rate": 1.3270848199610337e-05,
"epoch": 2.9583815733349583,
"step": 48550
},
{
"loss": 0.2292,
"grad_norm": 12.817031860351562,
"learning_rate": 1.3262830432237842e-05,
"epoch": 2.9614283102796906,
"step": 48600
},
{
"loss": 0.2401,
"grad_norm": 4.986586093902588,
"learning_rate": 1.3254812664865342e-05,
"epoch": 2.964475047224423,
"step": 48650
},
{
"loss": 0.2008,
"grad_norm": 6.5273003578186035,
"learning_rate": 1.3246794897492845e-05,
"epoch": 2.967521784169155,
"step": 48700
},
{
"loss": 0.25,
"grad_norm": 6.1861724853515625,
"learning_rate": 1.3238777130120349e-05,
"epoch": 2.9705685211138873,
"step": 48750
},
{
"loss": 0.2067,
"grad_norm": 1.1591109037399292,
"learning_rate": 1.323075936274785e-05,
"epoch": 2.9736152580586195,
"step": 48800
},
{
"loss": 0.2725,
"grad_norm": 6.211068630218506,
"learning_rate": 1.3222741595375352e-05,
"epoch": 2.9766619950033513,
"step": 48850
},
{
"loss": 0.1994,
"grad_norm": 0.2114630788564682,
"learning_rate": 1.3214723828002855e-05,
"epoch": 2.9797087319480835,
"step": 48900
},
{
"loss": 0.1765,
"grad_norm": 8.591974258422852,
"learning_rate": 1.3206706060630359e-05,
"epoch": 2.9827554688928157,
"step": 48950
},
{
"loss": 0.2106,
"grad_norm": 0.5509645342826843,
"learning_rate": 1.319868829325786e-05,
"epoch": 2.985802205837548,
"step": 49000
},
{
"loss": 0.208,
"grad_norm": 16.50217628479004,
"learning_rate": 1.3190670525885362e-05,
"epoch": 2.98884894278228,
"step": 49050
},
{
"loss": 0.1803,
"grad_norm": 9.82548713684082,
"learning_rate": 1.3182652758512867e-05,
"epoch": 2.9918956797270124,
"step": 49100
},
{
"loss": 0.2008,
"grad_norm": 0.29143351316452026,
"learning_rate": 1.3174634991140367e-05,
"epoch": 2.9949424166717447,
"step": 49150
},
{
"loss": 0.2887,
"grad_norm": 3.37011981010437,
"learning_rate": 1.316661722376787e-05,
"epoch": 2.997989153616477,
"step": 49200
},
{
"eval_loss": 0.1332029551267624,
"eval_accuracy": 0.964764353041988,
"eval_f1_macro": 0.9641926151483103,
"eval_top5_accuracy": 0.9957497857754927,
"eval_runtime": 218.7715,
"eval_samples_per_second": 133.358,
"eval_steps_per_second": 8.337,
"epoch": 3.0,
"step": 49233
},
{
"loss": 0.181,
"grad_norm": 0.959205687046051,
"learning_rate": 1.3158599456395374e-05,
"epoch": 3.001035890561209,
"step": 49250
},
{
"loss": 0.1954,
"grad_norm": 24.304420471191406,
"learning_rate": 1.3150581689022877e-05,
"epoch": 3.0040826275059413,
"step": 49300
},
{
"loss": 0.1729,
"grad_norm": 3.890622854232788,
"learning_rate": 1.3142563921650377e-05,
"epoch": 3.007129364450673,
"step": 49350
},
{
"loss": 0.1867,
"grad_norm": 4.622470855712891,
"learning_rate": 1.313454615427788e-05,
"epoch": 3.0101761013954054,
"step": 49400
},
{
"loss": 0.2533,
"grad_norm": 18.893041610717773,
"learning_rate": 1.3126528386905384e-05,
"epoch": 3.0132228383401376,
"step": 49450
},
{
"loss": 0.2118,
"grad_norm": 3.105787754058838,
"learning_rate": 1.3118510619532885e-05,
"epoch": 3.01626957528487,
"step": 49500
},
{
"loss": 0.2156,
"grad_norm": 5.282907485961914,
"learning_rate": 1.3110492852160387e-05,
"epoch": 3.019316312229602,
"step": 49550
},
{
"loss": 0.2899,
"grad_norm": 4.943941116333008,
"learning_rate": 1.3102475084787892e-05,
"epoch": 3.0223630491743343,
"step": 49600
},
{
"loss": 0.1989,
"grad_norm": 11.13240909576416,
"learning_rate": 1.3094457317415394e-05,
"epoch": 3.0254097861190665,
"step": 49650
},
{
"loss": 0.1873,
"grad_norm": 3.8496713638305664,
"learning_rate": 1.3086439550042895e-05,
"epoch": 3.0284565230637988,
"step": 49700
},
{
"loss": 0.2016,
"grad_norm": 5.319008827209473,
"learning_rate": 1.3078421782670399e-05,
"epoch": 3.031503260008531,
"step": 49750
},
{
"loss": 0.2214,
"grad_norm": 6.1485185623168945,
"learning_rate": 1.3070404015297902e-05,
"epoch": 3.034549996953263,
"step": 49800
},
{
"loss": 0.213,
"grad_norm": 2.6957945823669434,
"learning_rate": 1.3062386247925402e-05,
"epoch": 3.0375967338979954,
"step": 49850
},
{
"loss": 0.1909,
"grad_norm": 4.191739559173584,
"learning_rate": 1.3054368480552907e-05,
"epoch": 3.0406434708427272,
"step": 49900
},
{
"loss": 0.1264,
"grad_norm": 6.86317777633667,
"learning_rate": 1.3046350713180409e-05,
"epoch": 3.0436902077874595,
"step": 49950
},
{
"loss": 0.2413,
"grad_norm": 14.807289123535156,
"learning_rate": 1.3038332945807912e-05,
"epoch": 3.0467369447321917,
"step": 50000
},
{
"loss": 0.1958,
"grad_norm": 0.09081412106752396,
"learning_rate": 1.3030315178435412e-05,
"epoch": 3.049783681676924,
"step": 50050
},
{
"loss": 0.1887,
"grad_norm": 2.240165948867798,
"learning_rate": 1.3022297411062917e-05,
"epoch": 3.052830418621656,
"step": 50100
},
{
"loss": 0.1732,
"grad_norm": 7.532269477844238,
"learning_rate": 1.3014279643690419e-05,
"epoch": 3.0558771555663884,
"step": 50150
},
{
"loss": 0.1398,
"grad_norm": 9.122222900390625,
"learning_rate": 1.300626187631792e-05,
"epoch": 3.0589238925111206,
"step": 50200
},
{
"loss": 0.1473,
"grad_norm": 7.000551223754883,
"learning_rate": 1.2998244108945424e-05,
"epoch": 3.061970629455853,
"step": 50250
},
{
"loss": 0.1988,
"grad_norm": 11.770593643188477,
"learning_rate": 1.2990226341572927e-05,
"epoch": 3.065017366400585,
"step": 50300
},
{
"loss": 0.1469,
"grad_norm": 14.542364120483398,
"learning_rate": 1.298220857420043e-05,
"epoch": 3.0680641033453173,
"step": 50350
},
{
"loss": 0.2883,
"grad_norm": 7.668535232543945,
"learning_rate": 1.2974190806827932e-05,
"epoch": 3.0711108402900495,
"step": 50400
},
{
"loss": 0.2199,
"grad_norm": 9.27001667022705,
"learning_rate": 1.2966173039455434e-05,
"epoch": 3.0741575772347813,
"step": 50450
},
{
"loss": 0.1847,
"grad_norm": 0.16005860269069672,
"learning_rate": 1.2958155272082937e-05,
"epoch": 3.0772043141795136,
"step": 50500
},
{
"loss": 0.1465,
"grad_norm": 6.84523344039917,
"learning_rate": 1.2950137504710437e-05,
"epoch": 3.080251051124246,
"step": 50550
},
{
"loss": 0.1664,
"grad_norm": 4.585577487945557,
"learning_rate": 1.2942119737337942e-05,
"epoch": 3.083297788068978,
"step": 50600
},
{
"loss": 0.2193,
"grad_norm": 5.118870735168457,
"learning_rate": 1.2934101969965444e-05,
"epoch": 3.0863445250137103,
"step": 50650
},
{
"loss": 0.2204,
"grad_norm": 0.22834427654743195,
"learning_rate": 1.2926084202592948e-05,
"epoch": 3.0893912619584425,
"step": 50700
},
{
"loss": 0.2143,
"grad_norm": 4.687326431274414,
"learning_rate": 1.2918066435220449e-05,
"epoch": 3.0924379989031747,
"step": 50750
},
{
"loss": 0.198,
"grad_norm": 11.785704612731934,
"learning_rate": 1.2910048667847952e-05,
"epoch": 3.095484735847907,
"step": 50800
},
{
"loss": 0.1415,
"grad_norm": 15.277908325195312,
"learning_rate": 1.2902030900475456e-05,
"epoch": 3.098531472792639,
"step": 50850
},
{
"loss": 0.2388,
"grad_norm": 17.546405792236328,
"learning_rate": 1.2894013133102957e-05,
"epoch": 3.1015782097373714,
"step": 50900
},
{
"loss": 0.1404,
"grad_norm": 2.95064115524292,
"learning_rate": 1.288599536573046e-05,
"epoch": 3.1046249466821036,
"step": 50950
},
{
"loss": 0.1782,
"grad_norm": 4.244141578674316,
"learning_rate": 1.2877977598357962e-05,
"epoch": 3.107671683626836,
"step": 51000
},
{
"loss": 0.1709,
"grad_norm": 6.86581563949585,
"learning_rate": 1.2869959830985466e-05,
"epoch": 3.1107184205715677,
"step": 51050
},
{
"loss": 0.1785,
"grad_norm": 0.33003708720207214,
"learning_rate": 1.2861942063612967e-05,
"epoch": 3.1137651575163,
"step": 51100
},
{
"loss": 0.2014,
"grad_norm": 1.2162518501281738,
"learning_rate": 1.285392429624047e-05,
"epoch": 3.116811894461032,
"step": 51150
},
{
"loss": 0.1838,
"grad_norm": 7.137429714202881,
"learning_rate": 1.2845906528867973e-05,
"epoch": 3.1198586314057644,
"step": 51200
},
{
"loss": 0.1994,
"grad_norm": 13.012948036193848,
"learning_rate": 1.2837888761495474e-05,
"epoch": 3.1229053683504966,
"step": 51250
},
{
"loss": 0.2352,
"grad_norm": 9.901873588562012,
"learning_rate": 1.2829870994122977e-05,
"epoch": 3.125952105295229,
"step": 51300
},
{
"loss": 0.1957,
"grad_norm": 2.5005834102630615,
"learning_rate": 1.2821853226750481e-05,
"epoch": 3.128998842239961,
"step": 51350
},
{
"loss": 0.2196,
"grad_norm": 2.2109310626983643,
"learning_rate": 1.2813835459377983e-05,
"epoch": 3.1320455791846933,
"step": 51400
},
{
"loss": 0.1763,
"grad_norm": 2.2979929447174072,
"learning_rate": 1.2805817692005484e-05,
"epoch": 3.1350923161294255,
"step": 51450
},
{
"loss": 0.1832,
"grad_norm": 22.76611328125,
"learning_rate": 1.2797799924632988e-05,
"epoch": 3.1381390530741577,
"step": 51500
},
{
"loss": 0.1857,
"grad_norm": 5.223147869110107,
"learning_rate": 1.2789782157260491e-05,
"epoch": 3.1411857900188895,
"step": 51550
},
{
"loss": 0.205,
"grad_norm": 16.374116897583008,
"learning_rate": 1.2781764389887992e-05,
"epoch": 3.1442325269636218,
"step": 51600
},
{
"loss": 0.2135,
"grad_norm": 6.229572296142578,
"learning_rate": 1.2773906977862944e-05,
"epoch": 3.147279263908354,
"step": 51650
},
{
"loss": 0.1985,
"grad_norm": 14.311114311218262,
"learning_rate": 1.2765889210490448e-05,
"epoch": 3.1503260008530862,
"step": 51700
},
{
"loss": 0.1729,
"grad_norm": 3.1758477687835693,
"learning_rate": 1.275787144311795e-05,
"epoch": 3.1533727377978185,
"step": 51750
},
{
"loss": 0.1453,
"grad_norm": 10.041760444641113,
"learning_rate": 1.2749853675745455e-05,
"epoch": 3.1564194747425507,
"step": 51800
},
{
"loss": 0.1215,
"grad_norm": 2.8081419467926025,
"learning_rate": 1.2741835908372955e-05,
"epoch": 3.159466211687283,
"step": 51850
},
{
"loss": 0.1694,
"grad_norm": 0.2606128454208374,
"learning_rate": 1.2733818141000458e-05,
"epoch": 3.162512948632015,
"step": 51900
},
{
"loss": 0.1327,
"grad_norm": 1.5389821529388428,
"learning_rate": 1.272580037362796e-05,
"epoch": 3.1655596855767474,
"step": 51950
},
{
"loss": 0.1934,
"grad_norm": 11.088955879211426,
"learning_rate": 1.2717782606255463e-05,
"epoch": 3.1686064225214796,
"step": 52000
},
{
"loss": 0.1915,
"grad_norm": 1.0898667573928833,
"learning_rate": 1.2709925194230415e-05,
"epoch": 3.171653159466212,
"step": 52050
},
{
"loss": 0.1654,
"grad_norm": 14.428396224975586,
"learning_rate": 1.2701907426857917e-05,
"epoch": 3.174699896410944,
"step": 52100
},
{
"loss": 0.1926,
"grad_norm": 3.730410099029541,
"learning_rate": 1.2693889659485421e-05,
"epoch": 3.177746633355676,
"step": 52150
},
{
"loss": 0.1699,
"grad_norm": 9.546635627746582,
"learning_rate": 1.2685871892112924e-05,
"epoch": 3.180793370300408,
"step": 52200
},
{
"loss": 0.181,
"grad_norm": 19.445940017700195,
"learning_rate": 1.2677854124740425e-05,
"epoch": 3.1838401072451403,
"step": 52250
},
{
"loss": 0.2072,
"grad_norm": 10.81466007232666,
"learning_rate": 1.2669836357367929e-05,
"epoch": 3.1868868441898726,
"step": 52300
},
{
"loss": 0.192,
"grad_norm": 16.28464698791504,
"learning_rate": 1.2661818589995431e-05,
"epoch": 3.189933581134605,
"step": 52350
},
{
"loss": 0.1862,
"grad_norm": 11.830705642700195,
"learning_rate": 1.2653800822622932e-05,
"epoch": 3.192980318079337,
"step": 52400
},
{
"loss": 0.1547,
"grad_norm": 13.580194473266602,
"learning_rate": 1.2645783055250436e-05,
"epoch": 3.1960270550240693,
"step": 52450
},
{
"loss": 0.1955,
"grad_norm": 4.817779064178467,
"learning_rate": 1.2637765287877939e-05,
"epoch": 3.1990737919688015,
"step": 52500
},
{
"loss": 0.214,
"grad_norm": 6.895339488983154,
"learning_rate": 1.2629747520505441e-05,
"epoch": 3.2021205289135337,
"step": 52550
},
{
"loss": 0.1567,
"grad_norm": 20.58013343811035,
"learning_rate": 1.2621729753132942e-05,
"epoch": 3.205167265858266,
"step": 52600
},
{
"loss": 0.1809,
"grad_norm": 14.04248332977295,
"learning_rate": 1.2613711985760446e-05,
"epoch": 3.208214002802998,
"step": 52650
},
{
"loss": 0.1985,
"grad_norm": 2.040515661239624,
"learning_rate": 1.2605694218387949e-05,
"epoch": 3.21126073974773,
"step": 52700
},
{
"loss": 0.1767,
"grad_norm": 12.801034927368164,
"learning_rate": 1.259767645101545e-05,
"epoch": 3.214307476692462,
"step": 52750
},
{
"loss": 0.1854,
"grad_norm": 0.05807257443666458,
"learning_rate": 1.2589658683642954e-05,
"epoch": 3.2173542136371944,
"step": 52800
},
{
"loss": 0.2015,
"grad_norm": 0.42708441615104675,
"learning_rate": 1.2581640916270456e-05,
"epoch": 3.2204009505819267,
"step": 52850
},
{
"loss": 0.1707,
"grad_norm": 11.37332534790039,
"learning_rate": 1.257362314889796e-05,
"epoch": 3.223447687526659,
"step": 52900
},
{
"loss": 0.2015,
"grad_norm": 0.9386255145072937,
"learning_rate": 1.2565605381525462e-05,
"epoch": 3.226494424471391,
"step": 52950
},
{
"loss": 0.1584,
"grad_norm": 5.9182000160217285,
"learning_rate": 1.2557587614152964e-05,
"epoch": 3.2295411614161234,
"step": 53000
},
{
"loss": 0.2074,
"grad_norm": 13.469658851623535,
"learning_rate": 1.2549569846780467e-05,
"epoch": 3.2325878983608556,
"step": 53050
},
{
"loss": 0.1601,
"grad_norm": 10.236445426940918,
"learning_rate": 1.2541552079407967e-05,
"epoch": 3.235634635305588,
"step": 53100
},
{
"loss": 0.126,
"grad_norm": 9.004005432128906,
"learning_rate": 1.2533534312035472e-05,
"epoch": 3.23868137225032,
"step": 53150
},
{
"loss": 0.1491,
"grad_norm": 9.37993049621582,
"learning_rate": 1.2525516544662974e-05,
"epoch": 3.2417281091950523,
"step": 53200
},
{
"loss": 0.1799,
"grad_norm": 12.797863006591797,
"learning_rate": 1.2517498777290478e-05,
"epoch": 3.2447748461397845,
"step": 53250
},
{
"loss": 0.1625,
"grad_norm": 3.5594723224639893,
"learning_rate": 1.2509481009917979e-05,
"epoch": 3.2478215830845163,
"step": 53300
},
{
"loss": 0.1313,
"grad_norm": 2.4726338386535645,
"learning_rate": 1.2501463242545482e-05,
"epoch": 3.2508683200292485,
"step": 53350
},
{
"loss": 0.1666,
"grad_norm": 0.34635454416275024,
"learning_rate": 1.2493445475172986e-05,
"epoch": 3.2539150569739808,
"step": 53400
},
{
"loss": 0.2419,
"grad_norm": 17.786266326904297,
"learning_rate": 1.2485427707800487e-05,
"epoch": 3.256961793918713,
"step": 53450
},
{
"loss": 0.1611,
"grad_norm": 8.61577320098877,
"learning_rate": 1.2477409940427989e-05,
"epoch": 3.260008530863445,
"step": 53500
},
{
"loss": 0.2027,
"grad_norm": 9.45375919342041,
"learning_rate": 1.2469392173055492e-05,
"epoch": 3.2630552678081775,
"step": 53550
},
{
"loss": 0.2283,
"grad_norm": 1.2622352838516235,
"learning_rate": 1.2461374405682996e-05,
"epoch": 3.2661020047529097,
"step": 53600
},
{
"loss": 0.2217,
"grad_norm": 5.125481128692627,
"learning_rate": 1.2453356638310497e-05,
"epoch": 3.269148741697642,
"step": 53650
},
{
"loss": 0.1737,
"grad_norm": 9.393604278564453,
"learning_rate": 1.2445338870937999e-05,
"epoch": 3.272195478642374,
"step": 53700
},
{
"loss": 0.1352,
"grad_norm": 1.1392011642456055,
"learning_rate": 1.2437321103565503e-05,
"epoch": 3.2752422155871064,
"step": 53750
},
{
"loss": 0.1678,
"grad_norm": 11.508397102355957,
"learning_rate": 1.2429303336193004e-05,
"epoch": 3.278288952531838,
"step": 53800
},
{
"loss": 0.2093,
"grad_norm": 8.095763206481934,
"learning_rate": 1.2421285568820507e-05,
"epoch": 3.2813356894765704,
"step": 53850
},
{
"loss": 0.1661,
"grad_norm": 7.244358539581299,
"learning_rate": 1.241326780144801e-05,
"epoch": 3.2843824264213026,
"step": 53900
},
{
"loss": 0.1964,
"grad_norm": 0.8861814141273499,
"learning_rate": 1.2405250034075513e-05,
"epoch": 3.287429163366035,
"step": 53950
},
{
"loss": 0.1967,
"grad_norm": 3.8056399822235107,
"learning_rate": 1.2397232266703014e-05,
"epoch": 3.290475900310767,
"step": 54000
},
{
"loss": 0.1974,
"grad_norm": 1.07497239112854,
"learning_rate": 1.2389214499330518e-05,
"epoch": 3.2935226372554993,
"step": 54050
},
{
"loss": 0.2076,
"grad_norm": 2.8235199451446533,
"learning_rate": 1.238119673195802e-05,
"epoch": 3.2965693742002316,
"step": 54100
},
{
"loss": 0.1736,
"grad_norm": 1.1532771587371826,
"learning_rate": 1.2373178964585522e-05,
"epoch": 3.299616111144964,
"step": 54150
},
{
"loss": 0.1871,
"grad_norm": 3.3390119075775146,
"learning_rate": 1.2365161197213024e-05,
"epoch": 3.302662848089696,
"step": 54200
},
{
"loss": 0.2183,
"grad_norm": 7.56759786605835,
"learning_rate": 1.2357143429840528e-05,
"epoch": 3.3057095850344282,
"step": 54250
},
{
"loss": 0.2119,
"grad_norm": 3.8744874000549316,
"learning_rate": 1.234912566246803e-05,
"epoch": 3.3087563219791605,
"step": 54300
},
{
"loss": 0.1844,
"grad_norm": 7.738255500793457,
"learning_rate": 1.2341107895095532e-05,
"epoch": 3.3118030589238927,
"step": 54350
},
{
"loss": 0.2228,
"grad_norm": 4.160497665405273,
"learning_rate": 1.2333090127723036e-05,
"epoch": 3.314849795868625,
"step": 54400
},
{
"loss": 0.1854,
"grad_norm": 11.288641929626465,
"learning_rate": 1.2325072360350538e-05,
"epoch": 3.3178965328133567,
"step": 54450
},
{
"loss": 0.2318,
"grad_norm": 8.864755630493164,
"learning_rate": 1.2317054592978039e-05,
"epoch": 3.320943269758089,
"step": 54500
},
{
"loss": 0.1825,
"grad_norm": 10.670138359069824,
"learning_rate": 1.2309036825605543e-05,
"epoch": 3.323990006702821,
"step": 54550
},
{
"loss": 0.1949,
"grad_norm": 8.041802406311035,
"learning_rate": 1.2301019058233046e-05,
"epoch": 3.3270367436475534,
"step": 54600
},
{
"loss": 0.1578,
"grad_norm": 5.436008453369141,
"learning_rate": 1.2293001290860548e-05,
"epoch": 3.3300834805922856,
"step": 54650
},
{
"loss": 0.1353,
"grad_norm": 13.034976959228516,
"learning_rate": 1.2284983523488049e-05,
"epoch": 3.333130217537018,
"step": 54700
},
{
"loss": 0.1725,
"grad_norm": 5.594070911407471,
"learning_rate": 1.2276965756115553e-05,
"epoch": 3.33617695448175,
"step": 54750
},
{
"loss": 0.2183,
"grad_norm": 17.82539939880371,
"learning_rate": 1.2268947988743056e-05,
"epoch": 3.3392236914264823,
"step": 54800
},
{
"loss": 0.2131,
"grad_norm": 12.242247581481934,
"learning_rate": 1.2260930221370557e-05,
"epoch": 3.3422704283712146,
"step": 54850
},
{
"loss": 0.1658,
"grad_norm": 7.050077438354492,
"learning_rate": 1.225291245399806e-05,
"epoch": 3.345317165315947,
"step": 54900
},
{
"loss": 0.1985,
"grad_norm": 11.319928169250488,
"learning_rate": 1.2244894686625563e-05,
"epoch": 3.3483639022606786,
"step": 54950
},
{
"loss": 0.2326,
"grad_norm": 0.996658205986023,
"learning_rate": 1.2236876919253068e-05,
"epoch": 3.351410639205411,
"step": 55000
},
{
"loss": 0.227,
"grad_norm": 3.1996378898620605,
"learning_rate": 1.2228859151880568e-05,
"epoch": 3.354457376150143,
"step": 55050
},
{
"loss": 0.1763,
"grad_norm": 0.045249588787555695,
"learning_rate": 1.222084138450807e-05,
"epoch": 3.3575041130948753,
"step": 55100
},
{
"loss": 0.1623,
"grad_norm": 8.66640853881836,
"learning_rate": 1.2212823617135573e-05,
"epoch": 3.3605508500396075,
"step": 55150
},
{
"loss": 0.1508,
"grad_norm": 0.4346877336502075,
"learning_rate": 1.2204805849763076e-05,
"epoch": 3.3635975869843397,
"step": 55200
},
{
"loss": 0.1725,
"grad_norm": 7.728482723236084,
"learning_rate": 1.2196788082390578e-05,
"epoch": 3.366644323929072,
"step": 55250
},
{
"loss": 0.2151,
"grad_norm": 7.137537479400635,
"learning_rate": 1.2188770315018081e-05,
"epoch": 3.369691060873804,
"step": 55300
},
{
"loss": 0.1851,
"grad_norm": 11.769349098205566,
"learning_rate": 1.2180752547645585e-05,
"epoch": 3.3727377978185364,
"step": 55350
},
{
"loss": 0.1975,
"grad_norm": 4.830447196960449,
"learning_rate": 1.2172734780273086e-05,
"epoch": 3.3757845347632687,
"step": 55400
},
{
"loss": 0.1934,
"grad_norm": 8.886012077331543,
"learning_rate": 1.2164717012900588e-05,
"epoch": 3.378831271708001,
"step": 55450
},
{
"loss": 0.1638,
"grad_norm": 0.762124240398407,
"learning_rate": 1.2156699245528093e-05,
"epoch": 3.381878008652733,
"step": 55500
},
{
"loss": 0.1904,
"grad_norm": 13.875347137451172,
"learning_rate": 1.2148681478155593e-05,
"epoch": 3.384924745597465,
"step": 55550
},
{
"loss": 0.1762,
"grad_norm": 10.185726165771484,
"learning_rate": 1.2140663710783096e-05,
"epoch": 3.387971482542197,
"step": 55600
},
{
"loss": 0.1761,
"grad_norm": 18.489660263061523,
"learning_rate": 1.2132645943410598e-05,
"epoch": 3.3910182194869294,
"step": 55650
},
{
"loss": 0.2008,
"grad_norm": 16.125534057617188,
"learning_rate": 1.2124628176038103e-05,
"epoch": 3.3940649564316616,
"step": 55700
},
{
"loss": 0.162,
"grad_norm": 6.575192928314209,
"learning_rate": 1.2116610408665603e-05,
"epoch": 3.397111693376394,
"step": 55750
},
{
"loss": 0.2131,
"grad_norm": 10.228886604309082,
"learning_rate": 1.2108592641293106e-05,
"epoch": 3.400158430321126,
"step": 55800
},
{
"loss": 0.1546,
"grad_norm": 23.764366149902344,
"learning_rate": 1.210057487392061e-05,
"epoch": 3.4032051672658583,
"step": 55850
},
{
"loss": 0.1513,
"grad_norm": 7.559711933135986,
"learning_rate": 1.2092557106548111e-05,
"epoch": 3.4062519042105905,
"step": 55900
},
{
"loss": 0.2306,
"grad_norm": 13.250411987304688,
"learning_rate": 1.2084539339175613e-05,
"epoch": 3.4092986411553228,
"step": 55950
},
{
"loss": 0.179,
"grad_norm": 9.833016395568848,
"learning_rate": 1.2076521571803118e-05,
"epoch": 3.412345378100055,
"step": 56000
},
{
"loss": 0.2192,
"grad_norm": 18.047407150268555,
"learning_rate": 1.206850380443062e-05,
"epoch": 3.415392115044787,
"step": 56050
},
{
"loss": 0.1977,
"grad_norm": 13.977007865905762,
"learning_rate": 1.2060486037058121e-05,
"epoch": 3.418438851989519,
"step": 56100
},
{
"loss": 0.1776,
"grad_norm": 7.3973588943481445,
"learning_rate": 1.2052468269685625e-05,
"epoch": 3.4214855889342513,
"step": 56150
},
{
"loss": 0.1724,
"grad_norm": 6.011740207672119,
"learning_rate": 1.2044450502313128e-05,
"epoch": 3.4245323258789835,
"step": 56200
},
{
"loss": 0.2076,
"grad_norm": 19.67186737060547,
"learning_rate": 1.2036432734940628e-05,
"epoch": 3.4275790628237157,
"step": 56250
},
{
"loss": 0.1682,
"grad_norm": 8.676301002502441,
"learning_rate": 1.2028414967568131e-05,
"epoch": 3.430625799768448,
"step": 56300
},
{
"loss": 0.1785,
"grad_norm": 3.2824344635009766,
"learning_rate": 1.2020397200195635e-05,
"epoch": 3.43367253671318,
"step": 56350
},
{
"loss": 0.2149,
"grad_norm": 11.70221996307373,
"learning_rate": 1.2012379432823138e-05,
"epoch": 3.4367192736579124,
"step": 56400
},
{
"loss": 0.1535,
"grad_norm": 6.6499199867248535,
"learning_rate": 1.2004361665450638e-05,
"epoch": 3.4397660106026446,
"step": 56450
},
{
"loss": 0.2231,
"grad_norm": 0.5234062075614929,
"learning_rate": 1.1996343898078143e-05,
"epoch": 3.442812747547377,
"step": 56500
},
{
"loss": 0.2077,
"grad_norm": 3.492489814758301,
"learning_rate": 1.1988326130705645e-05,
"epoch": 3.445859484492109,
"step": 56550
},
{
"loss": 0.2124,
"grad_norm": 7.748647212982178,
"learning_rate": 1.1980308363333146e-05,
"epoch": 3.4489062214368413,
"step": 56600
},
{
"loss": 0.1725,
"grad_norm": 11.232156753540039,
"learning_rate": 1.197229059596065e-05,
"epoch": 3.4519529583815736,
"step": 56650
},
{
"loss": 0.1668,
"grad_norm": 9.940185546875,
"learning_rate": 1.1964272828588153e-05,
"epoch": 3.4549996953263054,
"step": 56700
},
{
"loss": 0.2249,
"grad_norm": 0.8312782049179077,
"learning_rate": 1.1956255061215655e-05,
"epoch": 3.4580464322710376,
"step": 56750
},
{
"loss": 0.191,
"grad_norm": 0.33266955614089966,
"learning_rate": 1.1948237293843156e-05,
"epoch": 3.46109316921577,
"step": 56800
},
{
"loss": 0.1507,
"grad_norm": 9.008598327636719,
"learning_rate": 1.194021952647066e-05,
"epoch": 3.464139906160502,
"step": 56850
},
{
"loss": 0.2249,
"grad_norm": 12.515632629394531,
"learning_rate": 1.1932201759098163e-05,
"epoch": 3.4671866431052343,
"step": 56900
},
{
"loss": 0.2098,
"grad_norm": 1.4661927223205566,
"learning_rate": 1.1924183991725663e-05,
"epoch": 3.4702333800499665,
"step": 56950
},
{
"loss": 0.1674,
"grad_norm": 13.8302640914917,
"learning_rate": 1.1916166224353168e-05,
"epoch": 3.4732801169946987,
"step": 57000
},
{
"loss": 0.1979,
"grad_norm": 13.61473560333252,
"learning_rate": 1.190814845698067e-05,
"epoch": 3.476326853939431,
"step": 57050
},
{
"loss": 0.1999,
"grad_norm": 0.7185053825378418,
"learning_rate": 1.1900130689608174e-05,
"epoch": 3.479373590884163,
"step": 57100
},
{
"loss": 0.2133,
"grad_norm": 21.016761779785156,
"learning_rate": 1.1892112922235675e-05,
"epoch": 3.4824203278288954,
"step": 57150
},
{
"loss": 0.2262,
"grad_norm": 1.4234304428100586,
"learning_rate": 1.1884255510210627e-05,
"epoch": 3.485467064773627,
"step": 57200
},
{
"loss": 0.1477,
"grad_norm": 1.3733958005905151,
"learning_rate": 1.187623774283813e-05,
"epoch": 3.4885138017183595,
"step": 57250
},
{
"loss": 0.1985,
"grad_norm": 19.46760368347168,
"learning_rate": 1.1868219975465634e-05,
"epoch": 3.4915605386630917,
"step": 57300
},
{
"loss": 0.2104,
"grad_norm": 11.500370025634766,
"learning_rate": 1.1860202208093135e-05,
"epoch": 3.494607275607824,
"step": 57350
},
{
"loss": 0.2437,
"grad_norm": 0.8515303134918213,
"learning_rate": 1.1852184440720637e-05,
"epoch": 3.497654012552556,
"step": 57400
},
{
"loss": 0.1521,
"grad_norm": 7.738093852996826,
"learning_rate": 1.1844166673348141e-05,
"epoch": 3.5007007494972884,
"step": 57450
},
{
"loss": 0.1998,
"grad_norm": 12.039609909057617,
"learning_rate": 1.1836148905975644e-05,
"epoch": 3.5037474864420206,
"step": 57500
},
{
"loss": 0.232,
"grad_norm": 0.31199946999549866,
"learning_rate": 1.1828131138603145e-05,
"epoch": 3.506794223386753,
"step": 57550
},
{
"loss": 0.1592,
"grad_norm": 9.63027572631836,
"learning_rate": 1.1820113371230649e-05,
"epoch": 3.509840960331485,
"step": 57600
},
{
"loss": 0.1655,
"grad_norm": 8.63875961303711,
"learning_rate": 1.1812095603858151e-05,
"epoch": 3.5128876972762173,
"step": 57650
},
{
"loss": 0.2282,
"grad_norm": 0.422468364238739,
"learning_rate": 1.1804077836485652e-05,
"epoch": 3.5159344342209495,
"step": 57700
},
{
"loss": 0.1548,
"grad_norm": 4.973412036895752,
"learning_rate": 1.1796060069113156e-05,
"epoch": 3.5189811711656818,
"step": 57750
},
{
"loss": 0.1939,
"grad_norm": 18.819499969482422,
"learning_rate": 1.1788042301740659e-05,
"epoch": 3.522027908110414,
"step": 57800
},
{
"loss": 0.1749,
"grad_norm": 13.338088989257812,
"learning_rate": 1.178018488971561e-05,
"epoch": 3.525074645055146,
"step": 57850
},
{
"loss": 0.1277,
"grad_norm": 9.301538467407227,
"learning_rate": 1.1772167122343112e-05,
"epoch": 3.528121381999878,
"step": 57900
},
{
"loss": 0.1526,
"grad_norm": 4.308801174163818,
"learning_rate": 1.1764149354970616e-05,
"epoch": 3.5311681189446102,
"step": 57950
},
{
"loss": 0.2195,
"grad_norm": 13.23925495147705,
"learning_rate": 1.1756131587598118e-05,
"epoch": 3.5342148558893425,
"step": 58000
},
{
"loss": 0.2002,
"grad_norm": 6.084113121032715,
"learning_rate": 1.1748113820225622e-05,
"epoch": 3.5372615928340747,
"step": 58050
},
{
"loss": 0.1742,
"grad_norm": 11.012857437133789,
"learning_rate": 1.1740096052853123e-05,
"epoch": 3.540308329778807,
"step": 58100
},
{
"loss": 0.1618,
"grad_norm": 4.274723052978516,
"learning_rate": 1.1732078285480626e-05,
"epoch": 3.543355066723539,
"step": 58150
},
{
"loss": 0.247,
"grad_norm": 21.325281143188477,
"learning_rate": 1.1724060518108128e-05,
"epoch": 3.5464018036682714,
"step": 58200
},
{
"loss": 0.1299,
"grad_norm": 15.806543350219727,
"learning_rate": 1.171604275073563e-05,
"epoch": 3.5494485406130036,
"step": 58250
},
{
"loss": 0.1708,
"grad_norm": 7.6335129737854,
"learning_rate": 1.1708024983363133e-05,
"epoch": 3.5524952775577354,
"step": 58300
},
{
"loss": 0.1565,
"grad_norm": 2.2938215732574463,
"learning_rate": 1.1700007215990636e-05,
"epoch": 3.5555420145024677,
"step": 58350
},
{
"loss": 0.1856,
"grad_norm": 7.947546482086182,
"learning_rate": 1.169198944861814e-05,
"epoch": 3.5585887514472,
"step": 58400
},
{
"loss": 0.2251,
"grad_norm": 10.249144554138184,
"learning_rate": 1.168397168124564e-05,
"epoch": 3.561635488391932,
"step": 58450
},
{
"loss": 0.2069,
"grad_norm": 0.7409939169883728,
"learning_rate": 1.1675953913873143e-05,
"epoch": 3.5646822253366643,
"step": 58500
},
{
"loss": 0.1635,
"grad_norm": 0.8652212619781494,
"learning_rate": 1.1667936146500647e-05,
"epoch": 3.5677289622813966,
"step": 58550
},
{
"loss": 0.2233,
"grad_norm": 13.422883987426758,
"learning_rate": 1.1659918379128148e-05,
"epoch": 3.570775699226129,
"step": 58600
},
{
"loss": 0.1626,
"grad_norm": 18.717967987060547,
"learning_rate": 1.165190061175565e-05,
"epoch": 3.573822436170861,
"step": 58650
},
{
"loss": 0.1397,
"grad_norm": 0.861539900302887,
"learning_rate": 1.1643882844383155e-05,
"epoch": 3.5768691731155933,
"step": 58700
},
{
"loss": 0.1394,
"grad_norm": 14.722729682922363,
"learning_rate": 1.1635865077010657e-05,
"epoch": 3.5799159100603255,
"step": 58750
},
{
"loss": 0.1898,
"grad_norm": 1.8954408168792725,
"learning_rate": 1.1627847309638158e-05,
"epoch": 3.5829626470050577,
"step": 58800
},
{
"loss": 0.1848,
"grad_norm": 0.5409238934516907,
"learning_rate": 1.161982954226566e-05,
"epoch": 3.58600938394979,
"step": 58850
},
{
"loss": 0.1656,
"grad_norm": 1.8600118160247803,
"learning_rate": 1.1611811774893165e-05,
"epoch": 3.589056120894522,
"step": 58900
},
{
"loss": 0.1886,
"grad_norm": 5.577185153961182,
"learning_rate": 1.1603794007520666e-05,
"epoch": 3.5921028578392544,
"step": 58950
},
{
"loss": 0.1147,
"grad_norm": 0.07934561371803284,
"learning_rate": 1.1595776240148168e-05,
"epoch": 3.595149594783986,
"step": 59000
},
{
"loss": 0.1456,
"grad_norm": 9.101874351501465,
"learning_rate": 1.1587758472775672e-05,
"epoch": 3.5981963317287184,
"step": 59050
},
{
"loss": 0.215,
"grad_norm": 10.025199890136719,
"learning_rate": 1.1579740705403175e-05,
"epoch": 3.6012430686734507,
"step": 59100
},
{
"loss": 0.1779,
"grad_norm": 9.788360595703125,
"learning_rate": 1.1571722938030676e-05,
"epoch": 3.604289805618183,
"step": 59150
},
{
"loss": 0.1253,
"grad_norm": 1.0531175136566162,
"learning_rate": 1.156370517065818e-05,
"epoch": 3.607336542562915,
"step": 59200
},
{
"loss": 0.2061,
"grad_norm": 8.481135368347168,
"learning_rate": 1.1555687403285683e-05,
"epoch": 3.6103832795076474,
"step": 59250
},
{
"loss": 0.1827,
"grad_norm": 2.8320040702819824,
"learning_rate": 1.1547669635913183e-05,
"epoch": 3.6134300164523796,
"step": 59300
},
{
"loss": 0.1448,
"grad_norm": 17.753530502319336,
"learning_rate": 1.1539651868540686e-05,
"epoch": 3.616476753397112,
"step": 59350
},
{
"loss": 0.169,
"grad_norm": 6.95145320892334,
"learning_rate": 1.153163410116819e-05,
"epoch": 3.6195234903418436,
"step": 59400
},
{
"loss": 0.2109,
"grad_norm": 2.2689051628112793,
"learning_rate": 1.1523616333795693e-05,
"epoch": 3.622570227286576,
"step": 59450
},
{
"loss": 0.1857,
"grad_norm": 17.519046783447266,
"learning_rate": 1.1515598566423193e-05,
"epoch": 3.625616964231308,
"step": 59500
},
{
"loss": 0.2449,
"grad_norm": 23.458993911743164,
"learning_rate": 1.1507580799050698e-05,
"epoch": 3.6286637011760403,
"step": 59550
},
{
"loss": 0.2053,
"grad_norm": 2.0309958457946777,
"learning_rate": 1.14995630316782e-05,
"epoch": 3.6317104381207725,
"step": 59600
},
{
"loss": 0.1317,
"grad_norm": 7.458314418792725,
"learning_rate": 1.14915452643057e-05,
"epoch": 3.6347571750655048,
"step": 59650
},
{
"loss": 0.1973,
"grad_norm": 5.4756083488464355,
"learning_rate": 1.1483527496933205e-05,
"epoch": 3.637803912010237,
"step": 59700
},
{
"loss": 0.2213,
"grad_norm": 9.060553550720215,
"learning_rate": 1.1475509729560708e-05,
"epoch": 3.6408506489549692,
"step": 59750
},
{
"loss": 0.1504,
"grad_norm": 0.10801022499799728,
"learning_rate": 1.146749196218821e-05,
"epoch": 3.6438973858997015,
"step": 59800
},
{
"loss": 0.2358,
"grad_norm": 6.714724063873291,
"learning_rate": 1.1459474194815713e-05,
"epoch": 3.6469441228444337,
"step": 59850
},
{
"loss": 0.1905,
"grad_norm": 8.820976257324219,
"learning_rate": 1.1451456427443215e-05,
"epoch": 3.649990859789166,
"step": 59900
},
{
"loss": 0.2116,
"grad_norm": 10.387930870056152,
"learning_rate": 1.1443438660070718e-05,
"epoch": 3.653037596733898,
"step": 59950
},
{
"loss": 0.2167,
"grad_norm": 9.958050727844238,
"learning_rate": 1.1435420892698218e-05,
"epoch": 3.6560843336786304,
"step": 60000
},
{
"loss": 0.1947,
"grad_norm": 4.923942565917969,
"learning_rate": 1.1427403125325723e-05,
"epoch": 3.6591310706233626,
"step": 60050
},
{
"loss": 0.201,
"grad_norm": 12.928447723388672,
"learning_rate": 1.1419385357953225e-05,
"epoch": 3.6621778075680944,
"step": 60100
},
{
"loss": 0.1521,
"grad_norm": 9.677907943725586,
"learning_rate": 1.141136759058073e-05,
"epoch": 3.6652245445128266,
"step": 60150
},
{
"loss": 0.1993,
"grad_norm": 0.9910337924957275,
"learning_rate": 1.140334982320823e-05,
"epoch": 3.668271281457559,
"step": 60200
},
{
"loss": 0.1447,
"grad_norm": 0.12287753075361252,
"learning_rate": 1.1395332055835733e-05,
"epoch": 3.671318018402291,
"step": 60250
},
{
"loss": 0.211,
"grad_norm": 1.388326644897461,
"learning_rate": 1.1387314288463237e-05,
"epoch": 3.6743647553470233,
"step": 60300
},
{
"loss": 0.14,
"grad_norm": 12.151650428771973,
"learning_rate": 1.1379296521090738e-05,
"epoch": 3.6774114922917556,
"step": 60350
},
{
"loss": 0.1954,
"grad_norm": 7.885488986968994,
"learning_rate": 1.137127875371824e-05,
"epoch": 3.680458229236488,
"step": 60400
},
{
"loss": 0.1646,
"grad_norm": 11.779284477233887,
"learning_rate": 1.1363260986345743e-05,
"epoch": 3.68350496618122,
"step": 60450
},
{
"loss": 0.226,
"grad_norm": 0.31127092242240906,
"learning_rate": 1.1355243218973247e-05,
"epoch": 3.6865517031259523,
"step": 60500
},
{
"loss": 0.1563,
"grad_norm": 12.790468215942383,
"learning_rate": 1.1347385806948199e-05,
"epoch": 3.689598440070684,
"step": 60550
},
{
"loss": 0.1974,
"grad_norm": 6.824357986450195,
"learning_rate": 1.13393680395757e-05,
"epoch": 3.6926451770154163,
"step": 60600
},
{
"loss": 0.1666,
"grad_norm": 2.2943830490112305,
"learning_rate": 1.1331350272203204e-05,
"epoch": 3.6956919139601485,
"step": 60650
},
{
"loss": 0.2321,
"grad_norm": 12.092047691345215,
"learning_rate": 1.1323332504830706e-05,
"epoch": 3.6987386509048807,
"step": 60700
},
{
"loss": 0.1741,
"grad_norm": 3.462599277496338,
"learning_rate": 1.1315314737458207e-05,
"epoch": 3.701785387849613,
"step": 60750
},
{
"loss": 0.2608,
"grad_norm": 5.341583728790283,
"learning_rate": 1.1307296970085711e-05,
"epoch": 3.704832124794345,
"step": 60800
},
{
"loss": 0.2007,
"grad_norm": 9.245195388793945,
"learning_rate": 1.1299279202713214e-05,
"epoch": 3.7078788617390774,
"step": 60850
},
{
"loss": 0.1724,
"grad_norm": 0.16547225415706635,
"learning_rate": 1.1291261435340716e-05,
"epoch": 3.7109255986838097,
"step": 60900
},
{
"loss": 0.1595,
"grad_norm": 9.750439643859863,
"learning_rate": 1.1283243667968217e-05,
"epoch": 3.713972335628542,
"step": 60950
},
{
"loss": 0.2118,
"grad_norm": 9.630824089050293,
"learning_rate": 1.1275225900595721e-05,
"epoch": 3.717019072573274,
"step": 61000
},
{
"loss": 0.2044,
"grad_norm": 14.20899772644043,
"learning_rate": 1.1267208133223224e-05,
"epoch": 3.7200658095180064,
"step": 61050
},
{
"loss": 0.2689,
"grad_norm": 18.336101531982422,
"learning_rate": 1.1259190365850725e-05,
"epoch": 3.7231125464627386,
"step": 61100
},
{
"loss": 0.1427,
"grad_norm": 0.27455052733421326,
"learning_rate": 1.1251172598478229e-05,
"epoch": 3.726159283407471,
"step": 61150
},
{
"loss": 0.1715,
"grad_norm": 10.76746940612793,
"learning_rate": 1.1243154831105731e-05,
"epoch": 3.729206020352203,
"step": 61200
},
{
"loss": 0.1675,
"grad_norm": 6.859617233276367,
"learning_rate": 1.1235137063733235e-05,
"epoch": 3.732252757296935,
"step": 61250
},
{
"loss": 0.2032,
"grad_norm": 0.7284711599349976,
"learning_rate": 1.1227119296360736e-05,
"epoch": 3.735299494241667,
"step": 61300
},
{
"loss": 0.1974,
"grad_norm": 6.392565727233887,
"learning_rate": 1.1219101528988239e-05,
"epoch": 3.7383462311863993,
"step": 61350
},
{
"loss": 0.1664,
"grad_norm": 11.874924659729004,
"learning_rate": 1.1211083761615741e-05,
"epoch": 3.7413929681311315,
"step": 61400
},
{
"loss": 0.2069,
"grad_norm": 5.095954895019531,
"learning_rate": 1.1203065994243244e-05,
"epoch": 3.7444397050758638,
"step": 61450
},
{
"loss": 0.1128,
"grad_norm": 3.0957977771759033,
"learning_rate": 1.1195048226870746e-05,
"epoch": 3.747486442020596,
"step": 61500
},
{
"loss": 0.1857,
"grad_norm": 5.716493606567383,
"learning_rate": 1.1187030459498249e-05,
"epoch": 3.7505331789653282,
"step": 61550
},
{
"loss": 0.1925,
"grad_norm": 0.9781308174133301,
"learning_rate": 1.1179012692125753e-05,
"epoch": 3.7535799159100605,
"step": 61600
},
{
"loss": 0.2073,
"grad_norm": 0.1202312633395195,
"learning_rate": 1.1170994924753254e-05,
"epoch": 3.7566266528547922,
"step": 61650
},
{
"loss": 0.1798,
"grad_norm": 6.724310874938965,
"learning_rate": 1.1162977157380756e-05,
"epoch": 3.7596733897995245,
"step": 61700
},
{
"loss": 0.1429,
"grad_norm": 6.371334552764893,
"learning_rate": 1.115495939000826e-05,
"epoch": 3.7627201267442567,
"step": 61750
},
{
"loss": 0.1836,
"grad_norm": 17.067386627197266,
"learning_rate": 1.1146941622635761e-05,
"epoch": 3.765766863688989,
"step": 61800
},
{
"loss": 0.1824,
"grad_norm": 2.2249600887298584,
"learning_rate": 1.1138923855263264e-05,
"epoch": 3.768813600633721,
"step": 61850
},
{
"loss": 0.1603,
"grad_norm": 14.501346588134766,
"learning_rate": 1.1130906087890766e-05,
"epoch": 3.7718603375784534,
"step": 61900
},
{
"loss": 0.1543,
"grad_norm": 4.4326090812683105,
"learning_rate": 1.112288832051827e-05,
"epoch": 3.7749070745231856,
"step": 61950
},
{
"loss": 0.1763,
"grad_norm": 10.252082824707031,
"learning_rate": 1.1114870553145771e-05,
"epoch": 3.777953811467918,
"step": 62000
},
{
"loss": 0.175,
"grad_norm": 4.703859329223633,
"learning_rate": 1.1106852785773274e-05,
"epoch": 3.78100054841265,
"step": 62050
},
{
"loss": 0.1507,
"grad_norm": 4.746605396270752,
"learning_rate": 1.1098835018400778e-05,
"epoch": 3.7840472853573823,
"step": 62100
},
{
"loss": 0.1674,
"grad_norm": 6.545734405517578,
"learning_rate": 1.1090817251028279e-05,
"epoch": 3.7870940223021146,
"step": 62150
},
{
"loss": 0.1672,
"grad_norm": 11.161604881286621,
"learning_rate": 1.1082799483655781e-05,
"epoch": 3.790140759246847,
"step": 62200
},
{
"loss": 0.1338,
"grad_norm": 5.916060447692871,
"learning_rate": 1.1074781716283286e-05,
"epoch": 3.793187496191579,
"step": 62250
},
{
"loss": 0.1586,
"grad_norm": 4.1231584548950195,
"learning_rate": 1.1066763948910788e-05,
"epoch": 3.7962342331363113,
"step": 62300
},
{
"loss": 0.2075,
"grad_norm": 36.042293548583984,
"learning_rate": 1.1058746181538289e-05,
"epoch": 3.799280970081043,
"step": 62350
},
{
"loss": 0.1855,
"grad_norm": 1.803945779800415,
"learning_rate": 1.1050728414165793e-05,
"epoch": 3.8023277070257753,
"step": 62400
},
{
"loss": 0.19,
"grad_norm": 6.917568206787109,
"learning_rate": 1.1042710646793296e-05,
"epoch": 3.8053744439705075,
"step": 62450
},
{
"loss": 0.1605,
"grad_norm": 4.814383506774902,
"learning_rate": 1.1034692879420796e-05,
"epoch": 3.8084211809152397,
"step": 62500
},
{
"loss": 0.1406,
"grad_norm": 11.177038192749023,
"learning_rate": 1.1026675112048299e-05,
"epoch": 3.811467917859972,
"step": 62550
},
{
"loss": 0.1768,
"grad_norm": 7.640921115875244,
"learning_rate": 1.1018657344675803e-05,
"epoch": 3.814514654804704,
"step": 62600
},
{
"loss": 0.1141,
"grad_norm": 1.0799803733825684,
"learning_rate": 1.1010639577303306e-05,
"epoch": 3.8175613917494364,
"step": 62650
},
{
"loss": 0.2086,
"grad_norm": 11.758040428161621,
"learning_rate": 1.1002621809930806e-05,
"epoch": 3.8206081286941687,
"step": 62700
},
{
"loss": 0.1595,
"grad_norm": 3.7408828735351562,
"learning_rate": 1.099460404255831e-05,
"epoch": 3.823654865638901,
"step": 62750
},
{
"loss": 0.1705,
"grad_norm": 0.04572775587439537,
"learning_rate": 1.0986586275185813e-05,
"epoch": 3.8267016025836327,
"step": 62800
},
{
"loss": 0.1664,
"grad_norm": 0.3511752784252167,
"learning_rate": 1.0978568507813314e-05,
"epoch": 3.829748339528365,
"step": 62850
},
{
"loss": 0.1437,
"grad_norm": 1.3389674425125122,
"learning_rate": 1.0970550740440818e-05,
"epoch": 3.832795076473097,
"step": 62900
},
{
"loss": 0.15,
"grad_norm": 0.28370460867881775,
"learning_rate": 1.096253297306832e-05,
"epoch": 3.8358418134178294,
"step": 62950
},
{
"loss": 0.1371,
"grad_norm": 6.801644325256348,
"learning_rate": 1.0954515205695823e-05,
"epoch": 3.8388885503625616,
"step": 63000
},
{
"loss": 0.1906,
"grad_norm": 12.617003440856934,
"learning_rate": 1.0946497438323324e-05,
"epoch": 3.841935287307294,
"step": 63050
},
{
"loss": 0.1757,
"grad_norm": 14.75645637512207,
"learning_rate": 1.0938479670950828e-05,
"epoch": 3.844982024252026,
"step": 63100
},
{
"loss": 0.1864,
"grad_norm": 6.594162940979004,
"learning_rate": 1.093046190357833e-05,
"epoch": 3.8480287611967583,
"step": 63150
},
{
"loss": 0.1343,
"grad_norm": 0.11699630320072174,
"learning_rate": 1.0922444136205831e-05,
"epoch": 3.8510754981414905,
"step": 63200
},
{
"loss": 0.2349,
"grad_norm": 10.706260681152344,
"learning_rate": 1.0914426368833336e-05,
"epoch": 3.8541222350862228,
"step": 63250
},
{
"loss": 0.2289,
"grad_norm": 0.9591517448425293,
"learning_rate": 1.0906408601460838e-05,
"epoch": 3.857168972030955,
"step": 63300
},
{
"loss": 0.1506,
"grad_norm": 1.933519721031189,
"learning_rate": 1.0898551189435792e-05,
"epoch": 3.860215708975687,
"step": 63350
},
{
"loss": 0.2145,
"grad_norm": 0.8973000049591064,
"learning_rate": 1.0890533422063294e-05,
"epoch": 3.8632624459204195,
"step": 63400
},
{
"loss": 0.1741,
"grad_norm": 2.4364027976989746,
"learning_rate": 1.0882515654690795e-05,
"epoch": 3.8663091828651517,
"step": 63450
},
{
"loss": 0.2039,
"grad_norm": 4.442420482635498,
"learning_rate": 1.0874497887318297e-05,
"epoch": 3.8693559198098835,
"step": 63500
},
{
"loss": 0.1771,
"grad_norm": 10.859204292297363,
"learning_rate": 1.0866480119945802e-05,
"epoch": 3.8724026567546157,
"step": 63550
},
{
"loss": 0.1835,
"grad_norm": 13.885732650756836,
"learning_rate": 1.0858462352573303e-05,
"epoch": 3.875449393699348,
"step": 63600
},
{
"loss": 0.2079,
"grad_norm": 5.558764934539795,
"learning_rate": 1.0850444585200805e-05,
"epoch": 3.87849613064408,
"step": 63650
},
{
"loss": 0.1298,
"grad_norm": 3.9788265228271484,
"learning_rate": 1.084242681782831e-05,
"epoch": 3.8815428675888124,
"step": 63700
},
{
"loss": 0.1894,
"grad_norm": 9.609243392944336,
"learning_rate": 1.0834409050455812e-05,
"epoch": 3.8845896045335446,
"step": 63750
},
{
"loss": 0.2069,
"grad_norm": 2.2505722045898438,
"learning_rate": 1.0826391283083313e-05,
"epoch": 3.887636341478277,
"step": 63800
},
{
"loss": 0.2211,
"grad_norm": 7.376501560211182,
"learning_rate": 1.0818373515710817e-05,
"epoch": 3.890683078423009,
"step": 63850
},
{
"loss": 0.1551,
"grad_norm": 4.188418865203857,
"learning_rate": 1.081035574833832e-05,
"epoch": 3.893729815367741,
"step": 63900
},
{
"loss": 0.1496,
"grad_norm": 5.205519199371338,
"learning_rate": 1.080233798096582e-05,
"epoch": 3.896776552312473,
"step": 63950
},
{
"loss": 0.2026,
"grad_norm": 3.283127546310425,
"learning_rate": 1.0794320213593324e-05,
"epoch": 3.8998232892572053,
"step": 64000
},
{
"loss": 0.2029,
"grad_norm": 15.604662895202637,
"learning_rate": 1.0786302446220827e-05,
"epoch": 3.9028700262019376,
"step": 64050
},
{
"loss": 0.2061,
"grad_norm": 0.2768491506576538,
"learning_rate": 1.077828467884833e-05,
"epoch": 3.90591676314667,
"step": 64100
},
{
"loss": 0.1694,
"grad_norm": 0.6869677901268005,
"learning_rate": 1.077026691147583e-05,
"epoch": 3.908963500091402,
"step": 64150
},
{
"loss": 0.1459,
"grad_norm": 5.717177867889404,
"learning_rate": 1.0762249144103334e-05,
"epoch": 3.9120102370361343,
"step": 64200
},
{
"loss": 0.1933,
"grad_norm": 0.28584009408950806,
"learning_rate": 1.0754231376730837e-05,
"epoch": 3.9150569739808665,
"step": 64250
},
{
"loss": 0.1692,
"grad_norm": 13.051074981689453,
"learning_rate": 1.0746213609358338e-05,
"epoch": 3.9181037109255987,
"step": 64300
},
{
"loss": 0.136,
"grad_norm": 4.051156520843506,
"learning_rate": 1.0738195841985842e-05,
"epoch": 3.921150447870331,
"step": 64350
},
{
"loss": 0.13,
"grad_norm": 0.6026102304458618,
"learning_rate": 1.0730178074613344e-05,
"epoch": 3.924197184815063,
"step": 64400
},
{
"loss": 0.2427,
"grad_norm": 0.255001425743103,
"learning_rate": 1.0722160307240847e-05,
"epoch": 3.9272439217597954,
"step": 64450
},
{
"loss": 0.2064,
"grad_norm": 14.284687042236328,
"learning_rate": 1.071414253986835e-05,
"epoch": 3.9302906587045277,
"step": 64500
},
{
"loss": 0.2007,
"grad_norm": 6.333935737609863,
"learning_rate": 1.0706124772495852e-05,
"epoch": 3.93333739564926,
"step": 64550
},
{
"loss": 0.2032,
"grad_norm": 9.61050796508789,
"learning_rate": 1.0698107005123354e-05,
"epoch": 3.9363841325939917,
"step": 64600
},
{
"loss": 0.1541,
"grad_norm": 7.274916648864746,
"learning_rate": 1.0690089237750855e-05,
"epoch": 3.939430869538724,
"step": 64650
},
{
"loss": 0.1689,
"grad_norm": 12.28336238861084,
"learning_rate": 1.068207147037836e-05,
"epoch": 3.942477606483456,
"step": 64700
},
{
"loss": 0.1781,
"grad_norm": 8.59406852722168,
"learning_rate": 1.0674053703005862e-05,
"epoch": 3.9455243434281884,
"step": 64750
},
{
"loss": 0.1982,
"grad_norm": 27.270652770996094,
"learning_rate": 1.0666035935633366e-05,
"epoch": 3.9485710803729206,
"step": 64800
},
{
"loss": 0.2046,
"grad_norm": 1.8270009756088257,
"learning_rate": 1.0658018168260867e-05,
"epoch": 3.951617817317653,
"step": 64850
},
{
"loss": 0.1653,
"grad_norm": 4.219080924987793,
"learning_rate": 1.065000040088837e-05,
"epoch": 3.954664554262385,
"step": 64900
},
{
"loss": 0.2235,
"grad_norm": 2.264132499694824,
"learning_rate": 1.0641982633515873e-05,
"epoch": 3.9577112912071173,
"step": 64950
},
{
"loss": 0.1397,
"grad_norm": 2.6758358478546143,
"learning_rate": 1.0633964866143374e-05,
"epoch": 3.9607580281518495,
"step": 65000
},
{
"loss": 0.1442,
"grad_norm": 8.614093780517578,
"learning_rate": 1.0625947098770877e-05,
"epoch": 3.9638047650965813,
"step": 65050
},
{
"loss": 0.1719,
"grad_norm": 4.567839622497559,
"learning_rate": 1.061792933139838e-05,
"epoch": 3.9668515020413135,
"step": 65100
},
{
"loss": 0.1453,
"grad_norm": 2.8265278339385986,
"learning_rate": 1.0609911564025883e-05,
"epoch": 3.9698982389860458,
"step": 65150
},
{
"loss": 0.1886,
"grad_norm": 3.3077237606048584,
"learning_rate": 1.0601893796653384e-05,
"epoch": 3.972944975930778,
"step": 65200
},
{
"loss": 0.2197,
"grad_norm": 28.035741806030273,
"learning_rate": 1.0593876029280887e-05,
"epoch": 3.9759917128755102,
"step": 65250
},
{
"loss": 0.2178,
"grad_norm": 17.373247146606445,
"learning_rate": 1.0585858261908391e-05,
"epoch": 3.9790384498202425,
"step": 65300
},
{
"loss": 0.1216,
"grad_norm": 0.09714843332767487,
"learning_rate": 1.0577840494535892e-05,
"epoch": 3.9820851867649747,
"step": 65350
},
{
"loss": 0.1312,
"grad_norm": 0.07783179730176926,
"learning_rate": 1.0569822727163394e-05,
"epoch": 3.985131923709707,
"step": 65400
},
{
"loss": 0.1851,
"grad_norm": 1.8154667615890503,
"learning_rate": 1.0561804959790899e-05,
"epoch": 3.988178660654439,
"step": 65450
},
{
"loss": 0.196,
"grad_norm": 0.6917868256568909,
"learning_rate": 1.0553787192418401e-05,
"epoch": 3.9912253975991714,
"step": 65500
},
{
"loss": 0.1421,
"grad_norm": 21.771730422973633,
"learning_rate": 1.0545769425045902e-05,
"epoch": 3.9942721345439036,
"step": 65550
},
{
"loss": 0.1551,
"grad_norm": 0.3130871057510376,
"learning_rate": 1.0537751657673404e-05,
"epoch": 3.997318871488636,
"step": 65600
},
{
"eval_loss": 0.123590387403965,
"eval_accuracy": 0.9671979434447301,
"eval_f1_macro": 0.9664556214620569,
"eval_top5_accuracy": 0.9962296486718081,
"eval_runtime": 219.5495,
"eval_samples_per_second": 132.886,
"eval_steps_per_second": 8.308,
"epoch": 4.0,
"step": 65644
},
{
"loss": 0.1555,
"grad_norm": 5.269689559936523,
"learning_rate": 1.0529733890300909e-05,
"epoch": 4.000365608433368,
"step": 65650
},
{
"loss": 0.1578,
"grad_norm": 6.1545586585998535,
"learning_rate": 1.052171612292841e-05,
"epoch": 4.0034123453781,
"step": 65700
},
{
"loss": 0.1438,
"grad_norm": 5.481527328491211,
"learning_rate": 1.0513698355555912e-05,
"epoch": 4.0064590823228325,
"step": 65750
},
{
"loss": 0.1748,
"grad_norm": 4.139859199523926,
"learning_rate": 1.0505680588183416e-05,
"epoch": 4.009505819267565,
"step": 65800
},
{
"loss": 0.1722,
"grad_norm": 7.761731147766113,
"learning_rate": 1.0497662820810919e-05,
"epoch": 4.012552556212297,
"step": 65850
},
{
"loss": 0.1111,
"grad_norm": 8.50834846496582,
"learning_rate": 1.048964505343842e-05,
"epoch": 4.015599293157029,
"step": 65900
},
{
"loss": 0.1245,
"grad_norm": 3.106919050216675,
"learning_rate": 1.0481627286065924e-05,
"epoch": 4.018646030101761,
"step": 65950
},
{
"loss": 0.1728,
"grad_norm": 6.111913204193115,
"learning_rate": 1.0473609518693426e-05,
"epoch": 4.021692767046493,
"step": 66000
},
{
"loss": 0.2039,
"grad_norm": 18.582612991333008,
"learning_rate": 1.0465591751320927e-05,
"epoch": 4.024739503991225,
"step": 66050
},
{
"loss": 0.1712,
"grad_norm": 10.870434761047363,
"learning_rate": 1.0457573983948431e-05,
"epoch": 4.027786240935957,
"step": 66100
},
{
"loss": 0.1351,
"grad_norm": 0.12244825065135956,
"learning_rate": 1.0449556216575934e-05,
"epoch": 4.0308329778806895,
"step": 66150
},
{
"loss": 0.1635,
"grad_norm": 13.035940170288086,
"learning_rate": 1.0441538449203436e-05,
"epoch": 4.033879714825422,
"step": 66200
},
{
"loss": 0.1635,
"grad_norm": 12.398369789123535,
"learning_rate": 1.0433520681830937e-05,
"epoch": 4.036926451770154,
"step": 66250
},
{
"loss": 0.1256,
"grad_norm": 4.048152923583984,
"learning_rate": 1.0425502914458441e-05,
"epoch": 4.039973188714886,
"step": 66300
},
{
"loss": 0.111,
"grad_norm": 0.4533614218235016,
"learning_rate": 1.0417485147085944e-05,
"epoch": 4.043019925659618,
"step": 66350
},
{
"loss": 0.1577,
"grad_norm": 7.1517510414123535,
"learning_rate": 1.0409788090408347e-05,
"epoch": 4.046066662604351,
"step": 66400
},
{
"loss": 0.1532,
"grad_norm": 5.417303562164307,
"learning_rate": 1.0401770323035849e-05,
"epoch": 4.049113399549083,
"step": 66450
},
{
"loss": 0.126,
"grad_norm": 4.442929744720459,
"learning_rate": 1.039375255566335e-05,
"epoch": 4.052160136493815,
"step": 66500
},
{
"loss": 0.1118,
"grad_norm": 1.18214750289917,
"learning_rate": 1.0385734788290852e-05,
"epoch": 4.055206873438547,
"step": 66550
},
{
"loss": 0.1424,
"grad_norm": 6.458086013793945,
"learning_rate": 1.0377717020918357e-05,
"epoch": 4.05825361038328,
"step": 66600
},
{
"loss": 0.1808,
"grad_norm": 5.7554731369018555,
"learning_rate": 1.0369699253545857e-05,
"epoch": 4.061300347328012,
"step": 66650
},
{
"loss": 0.1545,
"grad_norm": 7.057258129119873,
"learning_rate": 1.036168148617336e-05,
"epoch": 4.064347084272744,
"step": 66700
},
{
"loss": 0.1604,
"grad_norm": 1.529481291770935,
"learning_rate": 1.0353663718800864e-05,
"epoch": 4.067393821217476,
"step": 66750
},
{
"loss": 0.1253,
"grad_norm": 0.5276650786399841,
"learning_rate": 1.0345645951428367e-05,
"epoch": 4.0704405581622085,
"step": 66800
},
{
"loss": 0.1359,
"grad_norm": 4.7711591720581055,
"learning_rate": 1.0337628184055867e-05,
"epoch": 4.073487295106941,
"step": 66850
},
{
"loss": 0.1781,
"grad_norm": 0.22231434285640717,
"learning_rate": 1.0329610416683372e-05,
"epoch": 4.076534032051673,
"step": 66900
},
{
"loss": 0.1706,
"grad_norm": 13.07154369354248,
"learning_rate": 1.0321592649310874e-05,
"epoch": 4.079580768996405,
"step": 66950
},
{
"loss": 0.1568,
"grad_norm": 0.5031225681304932,
"learning_rate": 1.0313574881938375e-05,
"epoch": 4.082627505941137,
"step": 67000
},
{
"loss": 0.1632,
"grad_norm": 0.29345402121543884,
"learning_rate": 1.0305557114565879e-05,
"epoch": 4.08567424288587,
"step": 67050
},
{
"loss": 0.1482,
"grad_norm": 14.907662391662598,
"learning_rate": 1.0297539347193382e-05,
"epoch": 4.088720979830601,
"step": 67100
},
{
"loss": 0.1707,
"grad_norm": 2.365347385406494,
"learning_rate": 1.0289521579820884e-05,
"epoch": 4.091767716775333,
"step": 67150
},
{
"loss": 0.1825,
"grad_norm": 6.343519687652588,
"learning_rate": 1.0281503812448385e-05,
"epoch": 4.0948144537200655,
"step": 67200
},
{
"loss": 0.1468,
"grad_norm": 14.620111465454102,
"learning_rate": 1.0273486045075889e-05,
"epoch": 4.097861190664798,
"step": 67250
},
{
"loss": 0.1537,
"grad_norm": 7.050709247589111,
"learning_rate": 1.0265468277703392e-05,
"epoch": 4.10090792760953,
"step": 67300
},
{
"loss": 0.1383,
"grad_norm": 9.39615535736084,
"learning_rate": 1.0257450510330892e-05,
"epoch": 4.103954664554262,
"step": 67350
},
{
"loss": 0.151,
"grad_norm": 0.7197621464729309,
"learning_rate": 1.0249432742958397e-05,
"epoch": 4.107001401498994,
"step": 67400
},
{
"loss": 0.1522,
"grad_norm": 4.250251293182373,
"learning_rate": 1.02414149755859e-05,
"epoch": 4.110048138443727,
"step": 67450
},
{
"loss": 0.1477,
"grad_norm": 1.7124803066253662,
"learning_rate": 1.0233397208213403e-05,
"epoch": 4.113094875388459,
"step": 67500
},
{
"loss": 0.1746,
"grad_norm": 0.5257401466369629,
"learning_rate": 1.0225379440840904e-05,
"epoch": 4.116141612333191,
"step": 67550
},
{
"loss": 0.1763,
"grad_norm": 10.094488143920898,
"learning_rate": 1.0217361673468407e-05,
"epoch": 4.119188349277923,
"step": 67600
},
{
"loss": 0.1593,
"grad_norm": 14.876291275024414,
"learning_rate": 1.020934390609591e-05,
"epoch": 4.1222350862226556,
"step": 67650
},
{
"loss": 0.1855,
"grad_norm": 8.927576065063477,
"learning_rate": 1.020132613872341e-05,
"epoch": 4.125281823167388,
"step": 67700
},
{
"loss": 0.1494,
"grad_norm": Infinity,
"learning_rate": 1.0193308371350914e-05,
"epoch": 4.12832856011212,
"step": 67750
},
{
"loss": 0.1819,
"grad_norm": 7.071376323699951,
"learning_rate": 1.0185450959325866e-05,
"epoch": 4.131375297056852,
"step": 67800
},
{
"loss": 0.1939,
"grad_norm": 0.5017605423927307,
"learning_rate": 1.017743319195337e-05,
"epoch": 4.1344220340015845,
"step": 67850
},
{
"loss": 0.1663,
"grad_norm": 10.274580001831055,
"learning_rate": 1.0169415424580873e-05,
"epoch": 4.137468770946317,
"step": 67900
},
{
"loss": 0.107,
"grad_norm": 1.7769050598144531,
"learning_rate": 1.0161397657208374e-05,
"epoch": 4.140515507891049,
"step": 67950
},
{
"loss": 0.1499,
"grad_norm": 1.9443511962890625,
"learning_rate": 1.0153379889835878e-05,
"epoch": 4.143562244835781,
"step": 68000
},
{
"loss": 0.1698,
"grad_norm": 0.03015199862420559,
"learning_rate": 1.014536212246338e-05,
"epoch": 4.146608981780513,
"step": 68050
},
{
"loss": 0.204,
"grad_norm": 63.50517272949219,
"learning_rate": 1.0137344355090881e-05,
"epoch": 4.149655718725246,
"step": 68100
},
{
"loss": 0.1376,
"grad_norm": 16.454803466796875,
"learning_rate": 1.0129326587718384e-05,
"epoch": 4.152702455669978,
"step": 68150
},
{
"loss": 0.1874,
"grad_norm": 20.75562286376953,
"learning_rate": 1.0121308820345888e-05,
"epoch": 4.15574919261471,
"step": 68200
},
{
"loss": 0.1719,
"grad_norm": 25.64288902282715,
"learning_rate": 1.011329105297339e-05,
"epoch": 4.158795929559441,
"step": 68250
},
{
"loss": 0.1689,
"grad_norm": 0.19461429119110107,
"learning_rate": 1.0105273285600891e-05,
"epoch": 4.161842666504174,
"step": 68300
},
{
"loss": 0.1294,
"grad_norm": 21.52450942993164,
"learning_rate": 1.0097255518228395e-05,
"epoch": 4.164889403448906,
"step": 68350
},
{
"loss": 0.1162,
"grad_norm": 16.224693298339844,
"learning_rate": 1.0089237750855898e-05,
"epoch": 4.167936140393638,
"step": 68400
},
{
"loss": 0.1728,
"grad_norm": 3.495938301086426,
"learning_rate": 1.0081219983483399e-05,
"epoch": 4.17098287733837,
"step": 68450
},
{
"loss": 0.1287,
"grad_norm": 0.36357325315475464,
"learning_rate": 1.0073202216110903e-05,
"epoch": 4.174029614283103,
"step": 68500
},
{
"loss": 0.1902,
"grad_norm": 10.867305755615234,
"learning_rate": 1.0065184448738405e-05,
"epoch": 4.177076351227835,
"step": 68550
},
{
"loss": 0.0995,
"grad_norm": 1.2715249061584473,
"learning_rate": 1.0057166681365908e-05,
"epoch": 4.180123088172567,
"step": 68600
},
{
"loss": 0.1306,
"grad_norm": 0.2404586523771286,
"learning_rate": 1.004914891399341e-05,
"epoch": 4.183169825117299,
"step": 68650
},
{
"loss": 0.1332,
"grad_norm": 5.670309543609619,
"learning_rate": 1.0041131146620913e-05,
"epoch": 4.1862165620620315,
"step": 68700
},
{
"loss": 0.1519,
"grad_norm": 5.621829986572266,
"learning_rate": 1.0033113379248415e-05,
"epoch": 4.189263299006764,
"step": 68750
},
{
"loss": 0.1456,
"grad_norm": 4.919177055358887,
"learning_rate": 1.0025095611875916e-05,
"epoch": 4.192310035951496,
"step": 68800
},
{
"loss": 0.2049,
"grad_norm": 10.926495552062988,
"learning_rate": 1.001707784450342e-05,
"epoch": 4.195356772896228,
"step": 68850
},
{
"loss": 0.1267,
"grad_norm": 16.843975067138672,
"learning_rate": 1.0009060077130923e-05,
"epoch": 4.1984035098409604,
"step": 68900
},
{
"loss": 0.1595,
"grad_norm": 0.2576894462108612,
"learning_rate": 1.0001042309758427e-05,
"epoch": 4.201450246785693,
"step": 68950
},
{
"loss": 0.1449,
"grad_norm": 0.1604314148426056,
"learning_rate": 9.993024542385928e-06,
"epoch": 4.204496983730425,
"step": 69000
},
{
"loss": 0.1368,
"grad_norm": 8.193546295166016,
"learning_rate": 9.98500677501343e-06,
"epoch": 4.207543720675157,
"step": 69050
},
{
"loss": 0.1512,
"grad_norm": 2.784099578857422,
"learning_rate": 9.976989007640933e-06,
"epoch": 4.210590457619889,
"step": 69100
},
{
"loss": 0.1673,
"grad_norm": 27.554716110229492,
"learning_rate": 9.968971240268435e-06,
"epoch": 4.213637194564622,
"step": 69150
},
{
"loss": 0.1678,
"grad_norm": 0.35391202569007874,
"learning_rate": 9.960953472895938e-06,
"epoch": 4.216683931509354,
"step": 69200
},
{
"loss": 0.1272,
"grad_norm": 11.006820678710938,
"learning_rate": 9.95293570552344e-06,
"epoch": 4.219730668454086,
"step": 69250
},
{
"loss": 0.1644,
"grad_norm": 1.2185169458389282,
"learning_rate": 9.944917938150943e-06,
"epoch": 4.222777405398818,
"step": 69300
},
{
"loss": 0.154,
"grad_norm": 6.8227033615112305,
"learning_rate": 9.936900170778447e-06,
"epoch": 4.22582414234355,
"step": 69350
},
{
"loss": 0.1173,
"grad_norm": 16.846935272216797,
"learning_rate": 9.928882403405948e-06,
"epoch": 4.228870879288282,
"step": 69400
},
{
"loss": 0.2016,
"grad_norm": 2.5737788677215576,
"learning_rate": 9.92086463603345e-06,
"epoch": 4.231917616233014,
"step": 69450
},
{
"loss": 0.1448,
"grad_norm": 6.216710567474365,
"learning_rate": 9.912846868660953e-06,
"epoch": 4.234964353177746,
"step": 69500
},
{
"loss": 0.1938,
"grad_norm": 2.4834225177764893,
"learning_rate": 9.904829101288455e-06,
"epoch": 4.238011090122479,
"step": 69550
},
{
"loss": 0.1837,
"grad_norm": 2.0336954593658447,
"learning_rate": 9.89681133391596e-06,
"epoch": 4.241057827067211,
"step": 69600
},
{
"loss": 0.1488,
"grad_norm": 0.253782719373703,
"learning_rate": 9.88879356654346e-06,
"epoch": 4.244104564011943,
"step": 69650
},
{
"loss": 0.1843,
"grad_norm": 18.409990310668945,
"learning_rate": 9.880775799170965e-06,
"epoch": 4.247151300956675,
"step": 69700
},
{
"loss": 0.1713,
"grad_norm": 7.903343200683594,
"learning_rate": 9.872758031798465e-06,
"epoch": 4.2501980379014075,
"step": 69750
},
{
"loss": 0.1265,
"grad_norm": 12.906176567077637,
"learning_rate": 9.864740264425968e-06,
"epoch": 4.25324477484614,
"step": 69800
},
{
"loss": 0.1505,
"grad_norm": 1.4325015544891357,
"learning_rate": 9.856722497053472e-06,
"epoch": 4.256291511790872,
"step": 69850
},
{
"loss": 0.1587,
"grad_norm": 0.3123769462108612,
"learning_rate": 9.848704729680973e-06,
"epoch": 4.259338248735604,
"step": 69900
},
{
"loss": 0.1722,
"grad_norm": 11.920157432556152,
"learning_rate": 9.840686962308477e-06,
"epoch": 4.262384985680336,
"step": 69950
},
{
"loss": 0.142,
"grad_norm": 4.2379150390625,
"learning_rate": 9.832669194935978e-06,
"epoch": 4.265431722625069,
"step": 70000
},
{
"loss": 0.1507,
"grad_norm": 0.45674148201942444,
"learning_rate": 9.824651427563482e-06,
"epoch": 4.268478459569801,
"step": 70050
},
{
"loss": 0.1737,
"grad_norm": 1.3664522171020508,
"learning_rate": 9.816633660190985e-06,
"epoch": 4.271525196514533,
"step": 70100
},
{
"loss": 0.1722,
"grad_norm": 2.290088653564453,
"learning_rate": 9.808615892818485e-06,
"epoch": 4.274571933459265,
"step": 70150
},
{
"loss": 0.1814,
"grad_norm": 0.04270734637975693,
"learning_rate": 9.80059812544599e-06,
"epoch": 4.277618670403998,
"step": 70200
},
{
"loss": 0.1703,
"grad_norm": 0.09741075336933136,
"learning_rate": 9.79258035807349e-06,
"epoch": 4.28066540734873,
"step": 70250
},
{
"loss": 0.1539,
"grad_norm": 10.526651382446289,
"learning_rate": 9.784562590700995e-06,
"epoch": 4.283712144293462,
"step": 70300
},
{
"loss": 0.1804,
"grad_norm": 1.8439863920211792,
"learning_rate": 9.776544823328497e-06,
"epoch": 4.286758881238194,
"step": 70350
},
{
"loss": 0.1837,
"grad_norm": 5.162580966949463,
"learning_rate": 9.768527055956e-06,
"epoch": 4.2898056181829265,
"step": 70400
},
{
"loss": 0.1449,
"grad_norm": 11.055685043334961,
"learning_rate": 9.760509288583502e-06,
"epoch": 4.292852355127659,
"step": 70450
},
{
"loss": 0.1838,
"grad_norm": 12.48595905303955,
"learning_rate": 9.752491521211005e-06,
"epoch": 4.29589909207239,
"step": 70500
},
{
"loss": 0.1423,
"grad_norm": 10.9050931930542,
"learning_rate": 9.744473753838507e-06,
"epoch": 4.298945829017122,
"step": 70550
},
{
"loss": 0.2177,
"grad_norm": 1.2927939891815186,
"learning_rate": 9.73645598646601e-06,
"epoch": 4.3019925659618545,
"step": 70600
},
{
"loss": 0.1709,
"grad_norm": 10.216830253601074,
"learning_rate": 9.728438219093512e-06,
"epoch": 4.305039302906587,
"step": 70650
},
{
"loss": 0.1776,
"grad_norm": 6.0457024574279785,
"learning_rate": 9.720420451721015e-06,
"epoch": 4.308086039851319,
"step": 70700
},
{
"loss": 0.1392,
"grad_norm": 15.577397346496582,
"learning_rate": 9.712402684348517e-06,
"epoch": 4.311132776796051,
"step": 70750
},
{
"loss": 0.1374,
"grad_norm": 9.01161003112793,
"learning_rate": 9.70438491697602e-06,
"epoch": 4.3141795137407835,
"step": 70800
},
{
"loss": 0.143,
"grad_norm": 0.1723499894142151,
"learning_rate": 9.696367149603522e-06,
"epoch": 4.317226250685516,
"step": 70850
},
{
"loss": 0.1424,
"grad_norm": 8.494935989379883,
"learning_rate": 9.688349382231025e-06,
"epoch": 4.320272987630248,
"step": 70900
},
{
"loss": 0.109,
"grad_norm": 13.075262069702148,
"learning_rate": 9.680331614858527e-06,
"epoch": 4.32331972457498,
"step": 70950
},
{
"loss": 0.1636,
"grad_norm": 10.742445945739746,
"learning_rate": 9.67231384748603e-06,
"epoch": 4.326366461519712,
"step": 71000
},
{
"loss": 0.1581,
"grad_norm": 13.296379089355469,
"learning_rate": 9.664296080113532e-06,
"epoch": 4.329413198464445,
"step": 71050
},
{
"loss": 0.19,
"grad_norm": 0.2980782687664032,
"learning_rate": 9.656278312741035e-06,
"epoch": 4.332459935409177,
"step": 71100
},
{
"loss": 0.1471,
"grad_norm": 5.905920028686523,
"learning_rate": 9.648260545368537e-06,
"epoch": 4.335506672353909,
"step": 71150
},
{
"loss": 0.1404,
"grad_norm": 1.3744810819625854,
"learning_rate": 9.64024277799604e-06,
"epoch": 4.338553409298641,
"step": 71200
},
{
"loss": 0.1215,
"grad_norm": 0.4547983705997467,
"learning_rate": 9.632225010623542e-06,
"epoch": 4.3416001462433735,
"step": 71250
},
{
"loss": 0.1408,
"grad_norm": Infinity,
"learning_rate": 9.624207243251045e-06,
"epoch": 4.344646883188106,
"step": 71300
},
{
"loss": 0.174,
"grad_norm": 15.68720531463623,
"learning_rate": 9.616349831225997e-06,
"epoch": 4.347693620132838,
"step": 71350
},
{
"loss": 0.1437,
"grad_norm": 1.7201650142669678,
"learning_rate": 9.6083320638535e-06,
"epoch": 4.35074035707757,
"step": 71400
},
{
"loss": 0.1782,
"grad_norm": 0.1112494170665741,
"learning_rate": 9.600314296481003e-06,
"epoch": 4.3537870940223025,
"step": 71450
},
{
"loss": 0.1512,
"grad_norm": 11.564971923828125,
"learning_rate": 9.592296529108506e-06,
"epoch": 4.356833830967035,
"step": 71500
},
{
"loss": 0.2188,
"grad_norm": 12.624560356140137,
"learning_rate": 9.584278761736008e-06,
"epoch": 4.359880567911766,
"step": 71550
},
{
"loss": 0.167,
"grad_norm": 0.8543098568916321,
"learning_rate": 9.576260994363509e-06,
"epoch": 4.362927304856498,
"step": 71600
},
{
"loss": 0.1596,
"grad_norm": 9.048429489135742,
"learning_rate": 9.568243226991013e-06,
"epoch": 4.3659740418012305,
"step": 71650
},
{
"loss": 0.1718,
"grad_norm": 20.322839736938477,
"learning_rate": 9.560225459618516e-06,
"epoch": 4.369020778745963,
"step": 71700
},
{
"loss": 0.175,
"grad_norm": 5.2872138023376465,
"learning_rate": 9.552207692246018e-06,
"epoch": 4.372067515690695,
"step": 71750
},
{
"loss": 0.1611,
"grad_norm": 16.552467346191406,
"learning_rate": 9.54418992487352e-06,
"epoch": 4.375114252635427,
"step": 71800
},
{
"loss": 0.1943,
"grad_norm": 13.915517807006836,
"learning_rate": 9.536172157501023e-06,
"epoch": 4.378160989580159,
"step": 71850
},
{
"loss": 0.1704,
"grad_norm": 7.444904804229736,
"learning_rate": 9.528154390128526e-06,
"epoch": 4.381207726524892,
"step": 71900
},
{
"loss": 0.1351,
"grad_norm": 0.19871357083320618,
"learning_rate": 9.520136622756028e-06,
"epoch": 4.384254463469624,
"step": 71950
},
{
"loss": 0.1685,
"grad_norm": 0.042566727846860886,
"learning_rate": 9.512118855383531e-06,
"epoch": 4.387301200414356,
"step": 72000
},
{
"loss": 0.1445,
"grad_norm": 11.618751525878906,
"learning_rate": 9.504101088011033e-06,
"epoch": 4.390347937359088,
"step": 72050
},
{
"loss": 0.1237,
"grad_norm": 3.2215969562530518,
"learning_rate": 9.496083320638536e-06,
"epoch": 4.393394674303821,
"step": 72100
},
{
"loss": 0.146,
"grad_norm": 8.024391174316406,
"learning_rate": 9.488065553266038e-06,
"epoch": 4.396441411248553,
"step": 72150
},
{
"loss": 0.1988,
"grad_norm": 5.5678391456604,
"learning_rate": 9.480047785893541e-06,
"epoch": 4.399488148193285,
"step": 72200
},
{
"loss": 0.1544,
"grad_norm": 0.45636528730392456,
"learning_rate": 9.472030018521043e-06,
"epoch": 4.402534885138017,
"step": 72250
},
{
"loss": 0.1577,
"grad_norm": 8.10058879852295,
"learning_rate": 9.464012251148546e-06,
"epoch": 4.4055816220827495,
"step": 72300
},
{
"loss": 0.1709,
"grad_norm": 2.1804676055908203,
"learning_rate": 9.455994483776048e-06,
"epoch": 4.408628359027482,
"step": 72350
},
{
"loss": 0.1097,
"grad_norm": 0.993675708770752,
"learning_rate": 9.447976716403551e-06,
"epoch": 4.411675095972214,
"step": 72400
},
{
"loss": 0.1394,
"grad_norm": 14.475709915161133,
"learning_rate": 9.439958949031053e-06,
"epoch": 4.414721832916946,
"step": 72450
},
{
"loss": 0.1372,
"grad_norm": 7.89212703704834,
"learning_rate": 9.431941181658556e-06,
"epoch": 4.417768569861678,
"step": 72500
},
{
"loss": 0.1539,
"grad_norm": 5.518908977508545,
"learning_rate": 9.423923414286058e-06,
"epoch": 4.420815306806411,
"step": 72550
},
{
"loss": 0.196,
"grad_norm": 8.901643753051758,
"learning_rate": 9.415905646913561e-06,
"epoch": 4.423862043751143,
"step": 72600
},
{
"loss": 0.1617,
"grad_norm": 2.133546829223633,
"learning_rate": 9.407887879541063e-06,
"epoch": 4.426908780695875,
"step": 72650
},
{
"loss": 0.1254,
"grad_norm": 0.07907664775848389,
"learning_rate": 9.399870112168566e-06,
"epoch": 4.429955517640607,
"step": 72700
},
{
"loss": 0.1509,
"grad_norm": 7.087516784667969,
"learning_rate": 9.391852344796068e-06,
"epoch": 4.433002254585339,
"step": 72750
},
{
"loss": 0.1531,
"grad_norm": 18.76205062866211,
"learning_rate": 9.383834577423571e-06,
"epoch": 4.436048991530071,
"step": 72800
},
{
"loss": 0.1476,
"grad_norm": 17.18722915649414,
"learning_rate": 9.375816810051073e-06,
"epoch": 4.439095728474803,
"step": 72850
},
{
"loss": 0.1321,
"grad_norm": 15.510908126831055,
"learning_rate": 9.367799042678578e-06,
"epoch": 4.442142465419535,
"step": 72900
},
{
"loss": 0.1409,
"grad_norm": 9.173558235168457,
"learning_rate": 9.359781275306078e-06,
"epoch": 4.445189202364268,
"step": 72950
},
{
"loss": 0.1504,
"grad_norm": 21.726608276367188,
"learning_rate": 9.351763507933581e-06,
"epoch": 4.448235939309,
"step": 73000
},
{
"loss": 0.1263,
"grad_norm": 12.76060962677002,
"learning_rate": 9.343745740561085e-06,
"epoch": 4.451282676253732,
"step": 73050
},
{
"loss": 0.1627,
"grad_norm": 19.55306625366211,
"learning_rate": 9.335727973188586e-06,
"epoch": 4.454329413198464,
"step": 73100
},
{
"loss": 0.1386,
"grad_norm": 0.40780606865882874,
"learning_rate": 9.32771020581609e-06,
"epoch": 4.4573761501431965,
"step": 73150
},
{
"loss": 0.155,
"grad_norm": 8.28597354888916,
"learning_rate": 9.319692438443591e-06,
"epoch": 4.460422887087929,
"step": 73200
},
{
"loss": 0.1741,
"grad_norm": 10.950077056884766,
"learning_rate": 9.311674671071095e-06,
"epoch": 4.463469624032661,
"step": 73250
},
{
"loss": 0.1933,
"grad_norm": 3.488633155822754,
"learning_rate": 9.303656903698598e-06,
"epoch": 4.466516360977393,
"step": 73300
},
{
"loss": 0.1053,
"grad_norm": 3.3598673343658447,
"learning_rate": 9.295639136326098e-06,
"epoch": 4.4695630979221255,
"step": 73350
},
{
"loss": 0.157,
"grad_norm": 7.0977702140808105,
"learning_rate": 9.287621368953603e-06,
"epoch": 4.472609834866858,
"step": 73400
},
{
"loss": 0.1615,
"grad_norm": 4.10410213470459,
"learning_rate": 9.279603601581103e-06,
"epoch": 4.47565657181159,
"step": 73450
},
{
"loss": 0.1668,
"grad_norm": 23.22998046875,
"learning_rate": 9.271585834208608e-06,
"epoch": 4.478703308756322,
"step": 73500
},
{
"loss": 0.1593,
"grad_norm": 10.562776565551758,
"learning_rate": 9.26356806683611e-06,
"epoch": 4.481750045701054,
"step": 73550
},
{
"loss": 0.1356,
"grad_norm": 6.197351455688477,
"learning_rate": 9.255550299463613e-06,
"epoch": 4.484796782645787,
"step": 73600
},
{
"loss": 0.1509,
"grad_norm": 2.568244218826294,
"learning_rate": 9.247532532091115e-06,
"epoch": 4.487843519590519,
"step": 73650
},
{
"loss": 0.1157,
"grad_norm": 0.352704256772995,
"learning_rate": 9.239514764718616e-06,
"epoch": 4.490890256535251,
"step": 73700
},
{
"loss": 0.1315,
"grad_norm": 11.744978904724121,
"learning_rate": 9.23149699734612e-06,
"epoch": 4.493936993479983,
"step": 73750
},
{
"loss": 0.1717,
"grad_norm": 0.6165456771850586,
"learning_rate": 9.223479229973623e-06,
"epoch": 4.496983730424715,
"step": 73800
},
{
"loss": 0.139,
"grad_norm": 0.6451898217201233,
"learning_rate": 9.215461462601125e-06,
"epoch": 4.500030467369447,
"step": 73850
},
{
"loss": 0.1414,
"grad_norm": 3.256575345993042,
"learning_rate": 9.207443695228628e-06,
"epoch": 4.503077204314179,
"step": 73900
},
{
"loss": 0.1963,
"grad_norm": 5.525998115539551,
"learning_rate": 9.19942592785613e-06,
"epoch": 4.506123941258911,
"step": 73950
},
{
"loss": 0.1678,
"grad_norm": 4.229716777801514,
"learning_rate": 9.191408160483633e-06,
"epoch": 4.509170678203644,
"step": 74000
},
{
"loss": 0.1587,
"grad_norm": 5.144000053405762,
"learning_rate": 9.183390393111135e-06,
"epoch": 4.512217415148376,
"step": 74050
},
{
"loss": 0.1989,
"grad_norm": 14.187276840209961,
"learning_rate": 9.175372625738638e-06,
"epoch": 4.515264152093108,
"step": 74100
},
{
"loss": 0.1569,
"grad_norm": 12.363560676574707,
"learning_rate": 9.16735485836614e-06,
"epoch": 4.51831088903784,
"step": 74150
},
{
"loss": 0.1556,
"grad_norm": 13.659140586853027,
"learning_rate": 9.159337090993643e-06,
"epoch": 4.5213576259825725,
"step": 74200
},
{
"loss": 0.1695,
"grad_norm": 0.2539128065109253,
"learning_rate": 9.151319323621145e-06,
"epoch": 4.524404362927305,
"step": 74250
},
{
"loss": 0.1786,
"grad_norm": 0.0732613131403923,
"learning_rate": 9.143301556248648e-06,
"epoch": 4.527451099872037,
"step": 74300
},
{
"loss": 0.1103,
"grad_norm": 0.16322888433933258,
"learning_rate": 9.13528378887615e-06,
"epoch": 4.530497836816769,
"step": 74350
},
{
"loss": 0.1323,
"grad_norm": 6.227564334869385,
"learning_rate": 9.127266021503653e-06,
"epoch": 4.533544573761501,
"step": 74400
},
{
"loss": 0.1604,
"grad_norm": 7.325923919677734,
"learning_rate": 9.119248254131155e-06,
"epoch": 4.536591310706234,
"step": 74450
},
{
"loss": 0.1221,
"grad_norm": 0.4232766032218933,
"learning_rate": 9.111230486758658e-06,
"epoch": 4.539638047650966,
"step": 74500
},
{
"loss": 0.1627,
"grad_norm": 11.035075187683105,
"learning_rate": 9.10321271938616e-06,
"epoch": 4.542684784595698,
"step": 74550
},
{
"loss": 0.1731,
"grad_norm": 7.738249778747559,
"learning_rate": 9.095194952013663e-06,
"epoch": 4.54573152154043,
"step": 74600
},
{
"loss": 0.1735,
"grad_norm": 0.08011593669652939,
"learning_rate": 9.087177184641165e-06,
"epoch": 4.548778258485163,
"step": 74650
},
{
"loss": 0.1302,
"grad_norm": 0.15644200146198273,
"learning_rate": 9.079159417268668e-06,
"epoch": 4.551824995429895,
"step": 74700
},
{
"loss": 0.1354,
"grad_norm": 2.537287473678589,
"learning_rate": 9.07114164989617e-06,
"epoch": 4.554871732374627,
"step": 74750
},
{
"loss": 0.124,
"grad_norm": 8.051676750183105,
"learning_rate": 9.063123882523673e-06,
"epoch": 4.557918469319359,
"step": 74800
},
{
"loss": 0.1192,
"grad_norm": 3.2524735927581787,
"learning_rate": 9.055106115151175e-06,
"epoch": 4.5609652062640915,
"step": 74850
},
{
"loss": 0.1123,
"grad_norm": 1.4216221570968628,
"learning_rate": 9.04708834777868e-06,
"epoch": 4.564011943208824,
"step": 74900
},
{
"loss": 0.1197,
"grad_norm": 0.07429537922143936,
"learning_rate": 9.03907058040618e-06,
"epoch": 4.567058680153556,
"step": 74950
},
{
"loss": 0.176,
"grad_norm": 15.98138427734375,
"learning_rate": 9.031052813033684e-06,
"epoch": 4.570105417098288,
"step": 75000
},
{
"loss": 0.1288,
"grad_norm": 4.708049774169922,
"learning_rate": 9.023035045661185e-06,
"epoch": 4.57315215404302,
"step": 75050
},
{
"loss": 0.1195,
"grad_norm": 5.169354438781738,
"learning_rate": 9.015017278288688e-06,
"epoch": 4.576198890987752,
"step": 75100
},
{
"loss": 0.1486,
"grad_norm": 2.6833157539367676,
"learning_rate": 9.006999510916192e-06,
"epoch": 4.579245627932484,
"step": 75150
},
{
"loss": 0.2008,
"grad_norm": 0.19553257524967194,
"learning_rate": 8.998981743543693e-06,
"epoch": 4.582292364877216,
"step": 75200
},
{
"loss": 0.1014,
"grad_norm": 4.201247692108154,
"learning_rate": 8.990963976171197e-06,
"epoch": 4.5853391018219485,
"step": 75250
},
{
"loss": 0.1658,
"grad_norm": 2.9434971809387207,
"learning_rate": 8.982946208798698e-06,
"epoch": 4.588385838766681,
"step": 75300
},
{
"loss": 0.1692,
"grad_norm": 13.800713539123535,
"learning_rate": 8.974928441426202e-06,
"epoch": 4.591432575711413,
"step": 75350
},
{
"loss": 0.1513,
"grad_norm": 0.6052752137184143,
"learning_rate": 8.966910674053704e-06,
"epoch": 4.594479312656145,
"step": 75400
},
{
"loss": 0.1396,
"grad_norm": 0.022066060453653336,
"learning_rate": 8.958892906681205e-06,
"epoch": 4.597526049600877,
"step": 75450
},
{
"loss": 0.1506,
"grad_norm": 14.697855949401855,
"learning_rate": 8.95087513930871e-06,
"epoch": 4.60057278654561,
"step": 75500
},
{
"loss": 0.1535,
"grad_norm": 7.615461349487305,
"learning_rate": 8.94285737193621e-06,
"epoch": 4.603619523490342,
"step": 75550
},
{
"loss": 0.1379,
"grad_norm": 6.258397579193115,
"learning_rate": 8.934839604563714e-06,
"epoch": 4.606666260435074,
"step": 75600
},
{
"loss": 0.1453,
"grad_norm": 6.638627529144287,
"learning_rate": 8.926821837191217e-06,
"epoch": 4.609712997379806,
"step": 75650
},
{
"loss": 0.0987,
"grad_norm": 2.085364818572998,
"learning_rate": 8.91880406981872e-06,
"epoch": 4.612759734324539,
"step": 75700
},
{
"loss": 0.1355,
"grad_norm": 0.05569165199995041,
"learning_rate": 8.910786302446222e-06,
"epoch": 4.615806471269271,
"step": 75750
},
{
"loss": 0.1681,
"grad_norm": 12.844086647033691,
"learning_rate": 8.902768535073723e-06,
"epoch": 4.618853208214003,
"step": 75800
},
{
"loss": 0.2137,
"grad_norm": 0.629816472530365,
"learning_rate": 8.894750767701227e-06,
"epoch": 4.621899945158735,
"step": 75850
},
{
"loss": 0.1806,
"grad_norm": 0.4013004004955292,
"learning_rate": 8.88673300032873e-06,
"epoch": 4.6249466821034675,
"step": 75900
},
{
"loss": 0.219,
"grad_norm": 6.622217178344727,
"learning_rate": 8.878715232956232e-06,
"epoch": 4.6279934190482,
"step": 75950
},
{
"loss": 0.1208,
"grad_norm": 8.107784271240234,
"learning_rate": 8.870697465583735e-06,
"epoch": 4.631040155992932,
"step": 76000
},
{
"loss": 0.1572,
"grad_norm": 3.667630434036255,
"learning_rate": 8.862679698211237e-06,
"epoch": 4.634086892937663,
"step": 76050
},
{
"loss": 0.1506,
"grad_norm": 0.0294392928481102,
"learning_rate": 8.85466193083874e-06,
"epoch": 4.6371336298823955,
"step": 76100
},
{
"loss": 0.1361,
"grad_norm": 13.532487869262695,
"learning_rate": 8.846644163466242e-06,
"epoch": 4.640180366827128,
"step": 76150
},
{
"loss": 0.1256,
"grad_norm": 6.133322238922119,
"learning_rate": 8.838626396093745e-06,
"epoch": 4.64322710377186,
"step": 76200
},
{
"loss": 0.1525,
"grad_norm": 14.161111831665039,
"learning_rate": 8.830608628721247e-06,
"epoch": 4.646273840716592,
"step": 76250
},
{
"loss": 0.1567,
"grad_norm": 9.583621978759766,
"learning_rate": 8.82259086134875e-06,
"epoch": 4.6493205776613244,
"step": 76300
},
{
"loss": 0.1285,
"grad_norm": 1.128334879875183,
"learning_rate": 8.814573093976252e-06,
"epoch": 4.652367314606057,
"step": 76350
},
{
"loss": 0.1483,
"grad_norm": 20.707807540893555,
"learning_rate": 8.806555326603755e-06,
"epoch": 4.655414051550789,
"step": 76400
},
{
"loss": 0.1515,
"grad_norm": 8.500483512878418,
"learning_rate": 8.798537559231257e-06,
"epoch": 4.658460788495521,
"step": 76450
},
{
"loss": 0.1138,
"grad_norm": 6.448965072631836,
"learning_rate": 8.79051979185876e-06,
"epoch": 4.661507525440253,
"step": 76500
},
{
"loss": 0.1798,
"grad_norm": 6.490592002868652,
"learning_rate": 8.782502024486262e-06,
"epoch": 4.664554262384986,
"step": 76550
},
{
"loss": 0.1481,
"grad_norm": 20.311704635620117,
"learning_rate": 8.774484257113765e-06,
"epoch": 4.667600999329718,
"step": 76600
},
{
"loss": 0.1318,
"grad_norm": 6.119600772857666,
"learning_rate": 8.766466489741267e-06,
"epoch": 4.67064773627445,
"step": 76650
},
{
"loss": 0.1326,
"grad_norm": 0.02839754708111286,
"learning_rate": 8.75844872236877e-06,
"epoch": 4.673694473219182,
"step": 76700
},
{
"loss": 0.1445,
"grad_norm": 4.742705821990967,
"learning_rate": 8.750430954996274e-06,
"epoch": 4.6767412101639145,
"step": 76750
},
{
"loss": 0.1846,
"grad_norm": 1.4913629293441772,
"learning_rate": 8.742413187623775e-06,
"epoch": 4.679787947108647,
"step": 76800
},
{
"loss": 0.2092,
"grad_norm": 11.706443786621094,
"learning_rate": 8.734395420251277e-06,
"epoch": 4.682834684053379,
"step": 76850
},
{
"loss": 0.146,
"grad_norm": 0.21165534853935242,
"learning_rate": 8.72637765287878e-06,
"epoch": 4.685881420998111,
"step": 76900
},
{
"loss": 0.1722,
"grad_norm": 13.199803352355957,
"learning_rate": 8.718359885506282e-06,
"epoch": 4.6889281579428435,
"step": 76950
},
{
"loss": 0.1197,
"grad_norm": 0.5262232422828674,
"learning_rate": 8.710342118133786e-06,
"epoch": 4.691974894887576,
"step": 77000
},
{
"loss": 0.1098,
"grad_norm": 2.6329002380371094,
"learning_rate": 8.702324350761287e-06,
"epoch": 4.695021631832308,
"step": 77050
},
{
"loss": 0.1228,
"grad_norm": 0.11829888075590134,
"learning_rate": 8.694306583388791e-06,
"epoch": 4.69806836877704,
"step": 77100
},
{
"loss": 0.1614,
"grad_norm": 7.082816123962402,
"learning_rate": 8.686288816016292e-06,
"epoch": 4.701115105721772,
"step": 77150
},
{
"loss": 0.1232,
"grad_norm": 5.528819561004639,
"learning_rate": 8.678271048643795e-06,
"epoch": 4.704161842666505,
"step": 77200
},
{
"loss": 0.1405,
"grad_norm": 7.7945451736450195,
"learning_rate": 8.670253281271299e-06,
"epoch": 4.707208579611237,
"step": 77250
},
{
"loss": 0.1701,
"grad_norm": 1.0242927074432373,
"learning_rate": 8.6622355138988e-06,
"epoch": 4.710255316555968,
"step": 77300
},
{
"loss": 0.1167,
"grad_norm": 6.695906639099121,
"learning_rate": 8.654217746526304e-06,
"epoch": 4.7133020535007,
"step": 77350
},
{
"loss": 0.1835,
"grad_norm": 4.610082626342773,
"learning_rate": 8.646199979153805e-06,
"epoch": 4.716348790445433,
"step": 77400
},
{
"loss": 0.1312,
"grad_norm": 2.486097574234009,
"learning_rate": 8.638182211781309e-06,
"epoch": 4.719395527390165,
"step": 77450
},
{
"loss": 0.1231,
"grad_norm": 13.389360427856445,
"learning_rate": 8.630164444408811e-06,
"epoch": 4.722442264334897,
"step": 77500
},
{
"loss": 0.158,
"grad_norm": 0.1600705236196518,
"learning_rate": 8.622146677036312e-06,
"epoch": 4.725489001279629,
"step": 77550
},
{
"loss": 0.1426,
"grad_norm": 0.45124295353889465,
"learning_rate": 8.614128909663816e-06,
"epoch": 4.728535738224362,
"step": 77600
},
{
"loss": 0.1516,
"grad_norm": 1.6097251176834106,
"learning_rate": 8.606111142291317e-06,
"epoch": 4.731582475169094,
"step": 77650
},
{
"loss": 0.1702,
"grad_norm": 2.6479573249816895,
"learning_rate": 8.598093374918821e-06,
"epoch": 4.734629212113826,
"step": 77700
},
{
"loss": 0.1315,
"grad_norm": 0.5810887217521667,
"learning_rate": 8.590075607546324e-06,
"epoch": 4.737675949058558,
"step": 77750
},
{
"loss": 0.1247,
"grad_norm": 7.257587432861328,
"learning_rate": 8.582057840173826e-06,
"epoch": 4.7407226860032905,
"step": 77800
},
{
"loss": 0.1335,
"grad_norm": 8.074542045593262,
"learning_rate": 8.574040072801329e-06,
"epoch": 4.743769422948023,
"step": 77850
},
{
"loss": 0.1828,
"grad_norm": 9.140806198120117,
"learning_rate": 8.566022305428831e-06,
"epoch": 4.746816159892755,
"step": 77900
},
{
"loss": 0.1816,
"grad_norm": 7.248111724853516,
"learning_rate": 8.558004538056334e-06,
"epoch": 4.749862896837487,
"step": 77950
},
{
"loss": 0.1425,
"grad_norm": 0.023415960371494293,
"learning_rate": 8.549986770683836e-06,
"epoch": 4.752909633782219,
"step": 78000
},
{
"loss": 0.1732,
"grad_norm": 0.06485073268413544,
"learning_rate": 8.541969003311339e-06,
"epoch": 4.755956370726952,
"step": 78050
},
{
"loss": 0.161,
"grad_norm": 14.111425399780273,
"learning_rate": 8.533951235938841e-06,
"epoch": 4.759003107671684,
"step": 78100
},
{
"loss": 0.1747,
"grad_norm": 0.127353698015213,
"learning_rate": 8.525933468566344e-06,
"epoch": 4.762049844616416,
"step": 78150
},
{
"loss": 0.1915,
"grad_norm": 20.735435485839844,
"learning_rate": 8.517915701193846e-06,
"epoch": 4.765096581561148,
"step": 78200
},
{
"loss": 0.1707,
"grad_norm": 14.198448181152344,
"learning_rate": 8.509897933821349e-06,
"epoch": 4.768143318505881,
"step": 78250
},
{
"loss": 0.1426,
"grad_norm": 5.907705783843994,
"learning_rate": 8.501880166448851e-06,
"epoch": 4.771190055450612,
"step": 78300
},
{
"loss": 0.1201,
"grad_norm": 1.1161439418792725,
"learning_rate": 8.493862399076354e-06,
"epoch": 4.774236792395344,
"step": 78350
},
{
"loss": 0.1524,
"grad_norm": 13.004183769226074,
"learning_rate": 8.485844631703856e-06,
"epoch": 4.777283529340076,
"step": 78400
},
{
"loss": 0.1377,
"grad_norm": 2.3549396991729736,
"learning_rate": 8.477826864331359e-06,
"epoch": 4.780330266284809,
"step": 78450
},
{
"loss": 0.1524,
"grad_norm": 0.16806331276893616,
"learning_rate": 8.469809096958861e-06,
"epoch": 4.783377003229541,
"step": 78500
},
{
"loss": 0.1246,
"grad_norm": 0.19073890149593353,
"learning_rate": 8.461791329586364e-06,
"epoch": 4.786423740174273,
"step": 78550
},
{
"loss": 0.1512,
"grad_norm": 12.767096519470215,
"learning_rate": 8.453773562213866e-06,
"epoch": 4.789470477119005,
"step": 78600
},
{
"loss": 0.143,
"grad_norm": 1.1118272542953491,
"learning_rate": 8.445755794841369e-06,
"epoch": 4.7925172140637375,
"step": 78650
},
{
"loss": 0.1248,
"grad_norm": 7.8037800788879395,
"learning_rate": 8.437738027468871e-06,
"epoch": 4.79556395100847,
"step": 78700
},
{
"loss": 0.1468,
"grad_norm": 9.641472816467285,
"learning_rate": 8.429720260096374e-06,
"epoch": 4.798610687953202,
"step": 78750
},
{
"loss": 0.1758,
"grad_norm": 2.647005081176758,
"learning_rate": 8.421702492723876e-06,
"epoch": 4.801657424897934,
"step": 78800
},
{
"loss": 0.1565,
"grad_norm": 0.049469854682683945,
"learning_rate": 8.41368472535138e-06,
"epoch": 4.8047041618426665,
"step": 78850
},
{
"loss": 0.1635,
"grad_norm": 1.1479343175888062,
"learning_rate": 8.405666957978881e-06,
"epoch": 4.807750898787399,
"step": 78900
},
{
"loss": 0.1366,
"grad_norm": 12.654661178588867,
"learning_rate": 8.397649190606384e-06,
"epoch": 4.810797635732131,
"step": 78950
},
{
"loss": 0.124,
"grad_norm": 26.57990837097168,
"learning_rate": 8.389631423233886e-06,
"epoch": 4.813844372676863,
"step": 79000
},
{
"loss": 0.1381,
"grad_norm": 3.692706346511841,
"learning_rate": 8.381613655861389e-06,
"epoch": 4.816891109621595,
"step": 79050
},
{
"loss": 0.1182,
"grad_norm": 6.618845462799072,
"learning_rate": 8.373595888488893e-06,
"epoch": 4.819937846566328,
"step": 79100
},
{
"loss": 0.1131,
"grad_norm": 0.7274699211120605,
"learning_rate": 8.365578121116394e-06,
"epoch": 4.82298458351106,
"step": 79150
},
{
"loss": 0.1713,
"grad_norm": 16.446414947509766,
"learning_rate": 8.357560353743898e-06,
"epoch": 4.826031320455792,
"step": 79200
},
{
"loss": 0.1794,
"grad_norm": 8.824070930480957,
"learning_rate": 8.349542586371399e-06,
"epoch": 4.829078057400524,
"step": 79250
},
{
"loss": 0.1502,
"grad_norm": 1.0957262516021729,
"learning_rate": 8.341524818998901e-06,
"epoch": 4.8321247943452565,
"step": 79300
},
{
"loss": 0.1316,
"grad_norm": 0.32618576288223267,
"learning_rate": 8.333507051626406e-06,
"epoch": 4.835171531289989,
"step": 79350
},
{
"loss": 0.1499,
"grad_norm": 10.803485870361328,
"learning_rate": 8.325489284253906e-06,
"epoch": 4.838218268234721,
"step": 79400
},
{
"loss": 0.0889,
"grad_norm": 0.028994951397180557,
"learning_rate": 8.31747151688141e-06,
"epoch": 4.841265005179453,
"step": 79450
},
{
"loss": 0.1347,
"grad_norm": 7.012624740600586,
"learning_rate": 8.309453749508911e-06,
"epoch": 4.8443117421241855,
"step": 79500
},
{
"loss": 0.1406,
"grad_norm": 9.095114707946777,
"learning_rate": 8.301435982136416e-06,
"epoch": 4.847358479068917,
"step": 79550
},
{
"loss": 0.1888,
"grad_norm": 8.197093963623047,
"learning_rate": 8.293418214763918e-06,
"epoch": 4.850405216013649,
"step": 79600
},
{
"loss": 0.1932,
"grad_norm": 0.11493591964244843,
"learning_rate": 8.285400447391419e-06,
"epoch": 4.853451952958381,
"step": 79650
},
{
"loss": 0.183,
"grad_norm": 1.1331080198287964,
"learning_rate": 8.277382680018923e-06,
"epoch": 4.8564986899031135,
"step": 79700
},
{
"loss": 0.1852,
"grad_norm": 0.15018662810325623,
"learning_rate": 8.269364912646426e-06,
"epoch": 4.859545426847846,
"step": 79750
},
{
"loss": 0.1527,
"grad_norm": 11.478126525878906,
"learning_rate": 8.261347145273928e-06,
"epoch": 4.862592163792578,
"step": 79800
},
{
"loss": 0.1359,
"grad_norm": 0.5378670692443848,
"learning_rate": 8.25332937790143e-06,
"epoch": 4.86563890073731,
"step": 79850
},
{
"loss": 0.1735,
"grad_norm": 7.552121639251709,
"learning_rate": 8.245311610528933e-06,
"epoch": 4.868685637682042,
"step": 79900
},
{
"loss": 0.1008,
"grad_norm": 3.6824395656585693,
"learning_rate": 8.237293843156436e-06,
"epoch": 4.871732374626775,
"step": 79950
},
{
"loss": 0.1494,
"grad_norm": 6.86634635925293,
"learning_rate": 8.229276075783938e-06,
"epoch": 4.874779111571507,
"step": 80000
},
{
"loss": 0.1816,
"grad_norm": 8.14411735534668,
"learning_rate": 8.22141866375889e-06,
"epoch": 4.877825848516239,
"step": 80050
},
{
"loss": 0.1377,
"grad_norm": 10.066082954406738,
"learning_rate": 8.213400896386393e-06,
"epoch": 4.880872585460971,
"step": 80100
},
{
"loss": 0.1843,
"grad_norm": 5.275130271911621,
"learning_rate": 8.205383129013895e-06,
"epoch": 4.883919322405704,
"step": 80150
},
{
"loss": 0.1781,
"grad_norm": 13.97887134552002,
"learning_rate": 8.197365361641398e-06,
"epoch": 4.886966059350436,
"step": 80200
},
{
"loss": 0.1449,
"grad_norm": 8.672468185424805,
"learning_rate": 8.1893475942689e-06,
"epoch": 4.890012796295168,
"step": 80250
},
{
"loss": 0.1418,
"grad_norm": 1.338734745979309,
"learning_rate": 8.181329826896404e-06,
"epoch": 4.8930595332399,
"step": 80300
},
{
"loss": 0.1425,
"grad_norm": 11.119701385498047,
"learning_rate": 8.173312059523905e-06,
"epoch": 4.8961062701846325,
"step": 80350
},
{
"loss": 0.1926,
"grad_norm": 16.262802124023438,
"learning_rate": 8.165294292151408e-06,
"epoch": 4.899153007129365,
"step": 80400
},
{
"loss": 0.1493,
"grad_norm": 11.03242301940918,
"learning_rate": 8.157276524778912e-06,
"epoch": 4.902199744074097,
"step": 80450
},
{
"loss": 0.1419,
"grad_norm": 0.7811722755432129,
"learning_rate": 8.149258757406413e-06,
"epoch": 4.905246481018829,
"step": 80500
},
{
"loss": 0.1712,
"grad_norm": 5.967840671539307,
"learning_rate": 8.141240990033917e-06,
"epoch": 4.9082932179635606,
"step": 80550
},
{
"loss": 0.167,
"grad_norm": 3.886807918548584,
"learning_rate": 8.133223222661418e-06,
"epoch": 4.911339954908293,
"step": 80600
},
{
"loss": 0.1284,
"grad_norm": 6.520562171936035,
"learning_rate": 8.125205455288922e-06,
"epoch": 4.914386691853025,
"step": 80650
},
{
"loss": 0.2092,
"grad_norm": 11.603572845458984,
"learning_rate": 8.117187687916424e-06,
"epoch": 4.917433428797757,
"step": 80700
},
{
"loss": 0.1136,
"grad_norm": 0.2180764526128769,
"learning_rate": 8.109169920543925e-06,
"epoch": 4.9204801657424895,
"step": 80750
},
{
"loss": 0.2012,
"grad_norm": 0.2767029106616974,
"learning_rate": 8.10115215317143e-06,
"epoch": 4.923526902687222,
"step": 80800
},
{
"loss": 0.155,
"grad_norm": 0.22837011516094208,
"learning_rate": 8.09313438579893e-06,
"epoch": 4.926573639631954,
"step": 80850
},
{
"loss": 0.1991,
"grad_norm": 18.10485076904297,
"learning_rate": 8.085116618426434e-06,
"epoch": 4.929620376576686,
"step": 80900
},
{
"loss": 0.1366,
"grad_norm": 13.988286972045898,
"learning_rate": 8.077098851053937e-06,
"epoch": 4.932667113521418,
"step": 80950
},
{
"loss": 0.1526,
"grad_norm": 2.762017250061035,
"learning_rate": 8.06908108368144e-06,
"epoch": 4.935713850466151,
"step": 81000
},
{
"loss": 0.1735,
"grad_norm": 0.5543436408042908,
"learning_rate": 8.061063316308942e-06,
"epoch": 4.938760587410883,
"step": 81050
},
{
"loss": 0.1296,
"grad_norm": 6.978926658630371,
"learning_rate": 8.053045548936443e-06,
"epoch": 4.941807324355615,
"step": 81100
},
{
"loss": 0.147,
"grad_norm": 1.9104408025741577,
"learning_rate": 8.045027781563947e-06,
"epoch": 4.944854061300347,
"step": 81150
},
{
"loss": 0.1381,
"grad_norm": 3.241981267929077,
"learning_rate": 8.03701001419145e-06,
"epoch": 4.94790079824508,
"step": 81200
},
{
"loss": 0.1422,
"grad_norm": 14.468631744384766,
"learning_rate": 8.028992246818952e-06,
"epoch": 4.950947535189812,
"step": 81250
},
{
"loss": 0.1441,
"grad_norm": 0.04659886658191681,
"learning_rate": 8.020974479446454e-06,
"epoch": 4.953994272134544,
"step": 81300
},
{
"loss": 0.1185,
"grad_norm": 6.247297763824463,
"learning_rate": 8.012956712073957e-06,
"epoch": 4.957041009079276,
"step": 81350
},
{
"loss": 0.1274,
"grad_norm": 0.36010149121284485,
"learning_rate": 8.00493894470146e-06,
"epoch": 4.9600877460240085,
"step": 81400
},
{
"loss": 0.1381,
"grad_norm": 0.056266747415065765,
"learning_rate": 7.996921177328962e-06,
"epoch": 4.963134482968741,
"step": 81450
},
{
"loss": 0.119,
"grad_norm": 8.87271785736084,
"learning_rate": 7.988903409956464e-06,
"epoch": 4.966181219913473,
"step": 81500
},
{
"loss": 0.177,
"grad_norm": 2.6044483184814453,
"learning_rate": 7.980885642583967e-06,
"epoch": 4.969227956858205,
"step": 81550
},
{
"loss": 0.1284,
"grad_norm": 0.21755336225032806,
"learning_rate": 7.97286787521147e-06,
"epoch": 4.972274693802937,
"step": 81600
},
{
"loss": 0.1633,
"grad_norm": 6.487043380737305,
"learning_rate": 7.964850107838972e-06,
"epoch": 4.97532143074767,
"step": 81650
},
{
"loss": 0.14,
"grad_norm": 3.60038423538208,
"learning_rate": 7.956832340466474e-06,
"epoch": 4.978368167692402,
"step": 81700
},
{
"loss": 0.1623,
"grad_norm": 13.607037544250488,
"learning_rate": 7.948814573093977e-06,
"epoch": 4.981414904637134,
"step": 81750
},
{
"loss": 0.2025,
"grad_norm": 13.661416053771973,
"learning_rate": 7.94079680572148e-06,
"epoch": 4.984461641581865,
"step": 81800
},
{
"loss": 0.1325,
"grad_norm": 5.776841163635254,
"learning_rate": 7.932779038348982e-06,
"epoch": 4.987508378526598,
"step": 81850
},
{
"loss": 0.1282,
"grad_norm": 0.277293860912323,
"learning_rate": 7.924921626323935e-06,
"epoch": 4.99055511547133,
"step": 81900
},
{
"loss": 0.1409,
"grad_norm": 0.07868258655071259,
"learning_rate": 7.916903858951436e-06,
"epoch": 4.993601852416062,
"step": 81950
},
{
"loss": 0.142,
"grad_norm": 4.016332149505615,
"learning_rate": 7.90888609157894e-06,
"epoch": 4.996648589360794,
"step": 82000
},
{
"loss": 0.1596,
"grad_norm": 28.454986572265625,
"learning_rate": 7.900868324206441e-06,
"epoch": 4.999695326305527,
"step": 82050
},
{
"eval_loss": 0.1185305118560791,
"eval_accuracy": 0.9708997429305912,
"eval_f1_macro": 0.9704648827410302,
"eval_top5_accuracy": 0.996126820908312,
"eval_runtime": 216.758,
"eval_samples_per_second": 134.597,
"eval_steps_per_second": 8.415,
"epoch": 5.0,
"step": 82055
},
{
"loss": 0.1907,
"grad_norm": 5.267197132110596,
"learning_rate": 7.892850556833944e-06,
"epoch": 5.002742063250259,
"step": 82100
},
{
"loss": 0.1327,
"grad_norm": 12.287569999694824,
"learning_rate": 7.884832789461448e-06,
"epoch": 5.005788800194991,
"step": 82150
},
{
"loss": 0.104,
"grad_norm": 1.5745559930801392,
"learning_rate": 7.876815022088949e-06,
"epoch": 5.008835537139723,
"step": 82200
},
{
"loss": 0.0861,
"grad_norm": 0.06769659370183945,
"learning_rate": 7.868797254716453e-06,
"epoch": 5.0118822740844555,
"step": 82250
},
{
"loss": 0.1677,
"grad_norm": 12.604750633239746,
"learning_rate": 7.860779487343956e-06,
"epoch": 5.014929011029188,
"step": 82300
},
{
"loss": 0.1538,
"grad_norm": 3.9857802391052246,
"learning_rate": 7.852761719971458e-06,
"epoch": 5.01797574797392,
"step": 82350
},
{
"loss": 0.1451,
"grad_norm": 2.3828041553497314,
"learning_rate": 7.84474395259896e-06,
"epoch": 5.021022484918652,
"step": 82400
},
{
"loss": 0.1301,
"grad_norm": 1.0668222904205322,
"learning_rate": 7.836726185226461e-06,
"epoch": 5.0240692218633844,
"step": 82450
},
{
"loss": 0.0764,
"grad_norm": 0.03395290672779083,
"learning_rate": 7.828708417853966e-06,
"epoch": 5.027115958808117,
"step": 82500
},
{
"loss": 0.1336,
"grad_norm": 7.711270809173584,
"learning_rate": 7.820690650481468e-06,
"epoch": 5.030162695752849,
"step": 82550
},
{
"loss": 0.1686,
"grad_norm": 26.17399787902832,
"learning_rate": 7.81267288310897e-06,
"epoch": 5.033209432697581,
"step": 82600
},
{
"loss": 0.1584,
"grad_norm": 8.123069763183594,
"learning_rate": 7.804655115736473e-06,
"epoch": 5.036256169642313,
"step": 82650
},
{
"loss": 0.1166,
"grad_norm": 9.66085433959961,
"learning_rate": 7.796637348363976e-06,
"epoch": 5.039302906587046,
"step": 82700
},
{
"loss": 0.1728,
"grad_norm": 21.732318878173828,
"learning_rate": 7.788619580991478e-06,
"epoch": 5.042349643531778,
"step": 82750
},
{
"loss": 0.1575,
"grad_norm": 4.353252410888672,
"learning_rate": 7.78060181361898e-06,
"epoch": 5.04539638047651,
"step": 82800
},
{
"loss": 0.1241,
"grad_norm": 1.0105178356170654,
"learning_rate": 7.772584046246483e-06,
"epoch": 5.048443117421241,
"step": 82850
},
{
"loss": 0.1215,
"grad_norm": 11.825268745422363,
"learning_rate": 7.764566278873986e-06,
"epoch": 5.051489854365974,
"step": 82900
},
{
"loss": 0.1132,
"grad_norm": 2.622793197631836,
"learning_rate": 7.756548511501488e-06,
"epoch": 5.054536591310706,
"step": 82950
},
{
"loss": 0.1465,
"grad_norm": 7.189518451690674,
"learning_rate": 7.74853074412899e-06,
"epoch": 5.057583328255438,
"step": 83000
},
{
"loss": 0.1084,
"grad_norm": 0.45741766691207886,
"learning_rate": 7.740512976756493e-06,
"epoch": 5.06063006520017,
"step": 83050
},
{
"loss": 0.1238,
"grad_norm": 18.018428802490234,
"learning_rate": 7.732495209383996e-06,
"epoch": 5.063676802144903,
"step": 83100
},
{
"loss": 0.1237,
"grad_norm": 23.148338317871094,
"learning_rate": 7.724477442011498e-06,
"epoch": 5.066723539089635,
"step": 83150
},
{
"loss": 0.1792,
"grad_norm": 2.369418144226074,
"learning_rate": 7.716459674639e-06,
"epoch": 5.069770276034367,
"step": 83200
},
{
"loss": 0.1511,
"grad_norm": 0.14156174659729004,
"learning_rate": 7.708441907266503e-06,
"epoch": 5.072817012979099,
"step": 83250
},
{
"loss": 0.1956,
"grad_norm": 6.050409317016602,
"learning_rate": 7.700424139894006e-06,
"epoch": 5.0758637499238315,
"step": 83300
},
{
"loss": 0.185,
"grad_norm": 0.018449582159519196,
"learning_rate": 7.692406372521508e-06,
"epoch": 5.078910486868564,
"step": 83350
},
{
"loss": 0.1428,
"grad_norm": 0.42764395475387573,
"learning_rate": 7.68438860514901e-06,
"epoch": 5.081957223813296,
"step": 83400
},
{
"loss": 0.1327,
"grad_norm": 7.779719829559326,
"learning_rate": 7.676370837776513e-06,
"epoch": 5.085003960758028,
"step": 83450
},
{
"loss": 0.132,
"grad_norm": 9.030720710754395,
"learning_rate": 7.668353070404016e-06,
"epoch": 5.08805069770276,
"step": 83500
},
{
"loss": 0.1316,
"grad_norm": 6.9142937660217285,
"learning_rate": 7.660335303031518e-06,
"epoch": 5.091097434647493,
"step": 83550
},
{
"loss": 0.153,
"grad_norm": 0.07183430343866348,
"learning_rate": 7.65231753565902e-06,
"epoch": 5.094144171592225,
"step": 83600
},
{
"loss": 0.1131,
"grad_norm": 2.218149423599243,
"learning_rate": 7.644299768286523e-06,
"epoch": 5.097190908536957,
"step": 83650
},
{
"loss": 0.1284,
"grad_norm": 0.18849365413188934,
"learning_rate": 7.636282000914026e-06,
"epoch": 5.100237645481689,
"step": 83700
},
{
"loss": 0.1497,
"grad_norm": 9.936875343322754,
"learning_rate": 7.628264233541529e-06,
"epoch": 5.103284382426422,
"step": 83750
},
{
"loss": 0.129,
"grad_norm": 5.664190292358398,
"learning_rate": 7.6202464661690315e-06,
"epoch": 5.106331119371154,
"step": 83800
},
{
"loss": 0.1039,
"grad_norm": 5.729142665863037,
"learning_rate": 7.612228698796533e-06,
"epoch": 5.109377856315886,
"step": 83850
},
{
"loss": 0.1456,
"grad_norm": 0.25635120272636414,
"learning_rate": 7.6042109314240365e-06,
"epoch": 5.112424593260618,
"step": 83900
},
{
"loss": 0.1367,
"grad_norm": 12.674747467041016,
"learning_rate": 7.596193164051538e-06,
"epoch": 5.1154713302053505,
"step": 83950
},
{
"loss": 0.1451,
"grad_norm": 4.0857834815979,
"learning_rate": 7.5881753966790415e-06,
"epoch": 5.118518067150082,
"step": 84000
},
{
"loss": 0.1078,
"grad_norm": 0.581510066986084,
"learning_rate": 7.580157629306544e-06,
"epoch": 5.121564804094814,
"step": 84050
},
{
"loss": 0.1091,
"grad_norm": 0.03809339553117752,
"learning_rate": 7.572139861934047e-06,
"epoch": 5.124611541039546,
"step": 84100
},
{
"loss": 0.1443,
"grad_norm": 0.2127697914838791,
"learning_rate": 7.564122094561549e-06,
"epoch": 5.1276582779842785,
"step": 84150
},
{
"loss": 0.1828,
"grad_norm": 3.460315227508545,
"learning_rate": 7.5561043271890515e-06,
"epoch": 5.130705014929011,
"step": 84200
},
{
"loss": 0.1741,
"grad_norm": 0.09233107417821884,
"learning_rate": 7.548086559816554e-06,
"epoch": 5.133751751873743,
"step": 84250
},
{
"loss": 0.1112,
"grad_norm": 5.463161945343018,
"learning_rate": 7.5400687924440565e-06,
"epoch": 5.136798488818475,
"step": 84300
},
{
"loss": 0.1403,
"grad_norm": 3.8298556804656982,
"learning_rate": 7.53205102507156e-06,
"epoch": 5.1398452257632075,
"step": 84350
},
{
"loss": 0.1446,
"grad_norm": 8.038125991821289,
"learning_rate": 7.5240332576990615e-06,
"epoch": 5.14289196270794,
"step": 84400
},
{
"loss": 0.097,
"grad_norm": 2.0901780128479004,
"learning_rate": 7.516015490326565e-06,
"epoch": 5.145938699652672,
"step": 84450
},
{
"loss": 0.1092,
"grad_norm": 5.212148189544678,
"learning_rate": 7.5079977229540665e-06,
"epoch": 5.148985436597404,
"step": 84500
},
{
"loss": 0.0942,
"grad_norm": 0.11278870701789856,
"learning_rate": 7.499979955581569e-06,
"epoch": 5.152032173542136,
"step": 84550
},
{
"loss": 0.1474,
"grad_norm": 3.698460578918457,
"learning_rate": 7.491962188209072e-06,
"epoch": 5.155078910486869,
"step": 84600
},
{
"loss": 0.1572,
"grad_norm": 8.365859985351562,
"learning_rate": 7.483944420836574e-06,
"epoch": 5.158125647431601,
"step": 84650
},
{
"loss": 0.1213,
"grad_norm": 0.528534471988678,
"learning_rate": 7.475926653464077e-06,
"epoch": 5.161172384376333,
"step": 84700
},
{
"loss": 0.0945,
"grad_norm": 2.412107467651367,
"learning_rate": 7.467908886091579e-06,
"epoch": 5.164219121321065,
"step": 84750
},
{
"loss": 0.1901,
"grad_norm": 6.923273086547852,
"learning_rate": 7.459891118719082e-06,
"epoch": 5.1672658582657975,
"step": 84800
},
{
"loss": 0.1045,
"grad_norm": 1.4270687103271484,
"learning_rate": 7.451873351346585e-06,
"epoch": 5.17031259521053,
"step": 84850
},
{
"loss": 0.1182,
"grad_norm": 16.82732582092285,
"learning_rate": 7.4438555839740865e-06,
"epoch": 5.173359332155262,
"step": 84900
},
{
"loss": 0.1233,
"grad_norm": 6.319582939147949,
"learning_rate": 7.43583781660159e-06,
"epoch": 5.176406069099994,
"step": 84950
},
{
"loss": 0.1151,
"grad_norm": 4.105896472930908,
"learning_rate": 7.4278200492290915e-06,
"epoch": 5.1794528060447265,
"step": 85000
},
{
"loss": 0.1393,
"grad_norm": 15.345377922058105,
"learning_rate": 7.419802281856595e-06,
"epoch": 5.182499542989459,
"step": 85050
},
{
"loss": 0.1659,
"grad_norm": 19.979801177978516,
"learning_rate": 7.411784514484097e-06,
"epoch": 5.18554627993419,
"step": 85100
},
{
"loss": 0.1726,
"grad_norm": 16.18895721435547,
"learning_rate": 7.403766747111601e-06,
"epoch": 5.188593016878922,
"step": 85150
},
{
"loss": 0.1451,
"grad_norm": 2.8700103759765625,
"learning_rate": 7.395748979739102e-06,
"epoch": 5.1916397538236545,
"step": 85200
},
{
"loss": 0.1603,
"grad_norm": 16.855562210083008,
"learning_rate": 7.387731212366605e-06,
"epoch": 5.194686490768387,
"step": 85250
},
{
"loss": 0.1524,
"grad_norm": 8.458178520202637,
"learning_rate": 7.379713444994107e-06,
"epoch": 5.197733227713119,
"step": 85300
},
{
"loss": 0.1595,
"grad_norm": 8.99972152709961,
"learning_rate": 7.37169567762161e-06,
"epoch": 5.200779964657851,
"step": 85350
},
{
"loss": 0.1437,
"grad_norm": 14.312650680541992,
"learning_rate": 7.363677910249113e-06,
"epoch": 5.203826701602583,
"step": 85400
},
{
"loss": 0.1572,
"grad_norm": 3.8768229484558105,
"learning_rate": 7.355660142876615e-06,
"epoch": 5.206873438547316,
"step": 85450
},
{
"loss": 0.1402,
"grad_norm": 4.911172389984131,
"learning_rate": 7.347642375504118e-06,
"epoch": 5.209920175492048,
"step": 85500
},
{
"loss": 0.1164,
"grad_norm": 6.441468238830566,
"learning_rate": 7.33962460813162e-06,
"epoch": 5.21296691243678,
"step": 85550
},
{
"loss": 0.1294,
"grad_norm": 2.7416164875030518,
"learning_rate": 7.3316068407591224e-06,
"epoch": 5.216013649381512,
"step": 85600
},
{
"loss": 0.128,
"grad_norm": 11.977770805358887,
"learning_rate": 7.323589073386626e-06,
"epoch": 5.219060386326245,
"step": 85650
},
{
"loss": 0.138,
"grad_norm": 8.160606384277344,
"learning_rate": 7.3155713060141274e-06,
"epoch": 5.222107123270977,
"step": 85700
},
{
"loss": 0.1346,
"grad_norm": 0.9926206469535828,
"learning_rate": 7.307553538641631e-06,
"epoch": 5.225153860215709,
"step": 85750
},
{
"loss": 0.0941,
"grad_norm": 0.47258642315864563,
"learning_rate": 7.2995357712691324e-06,
"epoch": 5.228200597160441,
"step": 85800
},
{
"loss": 0.1038,
"grad_norm": 0.029850119724869728,
"learning_rate": 7.291518003896636e-06,
"epoch": 5.2312473341051735,
"step": 85850
},
{
"loss": 0.1011,
"grad_norm": 0.019423088058829308,
"learning_rate": 7.283500236524138e-06,
"epoch": 5.234294071049906,
"step": 85900
},
{
"loss": 0.0898,
"grad_norm": 9.811861991882324,
"learning_rate": 7.27548246915164e-06,
"epoch": 5.237340807994638,
"step": 85950
},
{
"loss": 0.1001,
"grad_norm": 6.140851020812988,
"learning_rate": 7.267464701779143e-06,
"epoch": 5.24038754493937,
"step": 86000
},
{
"loss": 0.1323,
"grad_norm": 0.17067290842533112,
"learning_rate": 7.259607289754096e-06,
"epoch": 5.243434281884102,
"step": 86050
},
{
"loss": 0.1399,
"grad_norm": 15.70168685913086,
"learning_rate": 7.251589522381598e-06,
"epoch": 5.246481018828835,
"step": 86100
},
{
"loss": 0.1208,
"grad_norm": 12.917496681213379,
"learning_rate": 7.243571755009101e-06,
"epoch": 5.249527755773567,
"step": 86150
},
{
"loss": 0.1123,
"grad_norm": 6.384840488433838,
"learning_rate": 7.2355539876366036e-06,
"epoch": 5.252574492718299,
"step": 86200
},
{
"loss": 0.1191,
"grad_norm": 0.06860390305519104,
"learning_rate": 7.227536220264106e-06,
"epoch": 5.255621229663031,
"step": 86250
},
{
"loss": 0.1026,
"grad_norm": 0.05497782304883003,
"learning_rate": 7.219518452891609e-06,
"epoch": 5.258667966607763,
"step": 86300
},
{
"loss": 0.1523,
"grad_norm": 29.53835105895996,
"learning_rate": 7.21150068551911e-06,
"epoch": 5.261714703552495,
"step": 86350
},
{
"loss": 0.1184,
"grad_norm": 2.988090991973877,
"learning_rate": 7.203482918146614e-06,
"epoch": 5.264761440497227,
"step": 86400
},
{
"loss": 0.1159,
"grad_norm": 5.291945934295654,
"learning_rate": 7.195465150774116e-06,
"epoch": 5.267808177441959,
"step": 86450
},
{
"loss": 0.1659,
"grad_norm": 7.277029991149902,
"learning_rate": 7.187447383401619e-06,
"epoch": 5.270854914386692,
"step": 86500
},
{
"loss": 0.1211,
"grad_norm": 4.357074737548828,
"learning_rate": 7.179429616029121e-06,
"epoch": 5.273901651331424,
"step": 86550
},
{
"loss": 0.1539,
"grad_norm": 11.240453720092773,
"learning_rate": 7.1714118486566244e-06,
"epoch": 5.276948388276156,
"step": 86600
},
{
"loss": 0.1432,
"grad_norm": 7.438399791717529,
"learning_rate": 7.163394081284126e-06,
"epoch": 5.279995125220888,
"step": 86650
},
{
"loss": 0.1129,
"grad_norm": 26.198026657104492,
"learning_rate": 7.155376313911629e-06,
"epoch": 5.2830418621656206,
"step": 86700
},
{
"loss": 0.1357,
"grad_norm": 6.511169910430908,
"learning_rate": 7.147358546539131e-06,
"epoch": 5.286088599110353,
"step": 86750
},
{
"loss": 0.111,
"grad_norm": 7.824971675872803,
"learning_rate": 7.139340779166634e-06,
"epoch": 5.289135336055085,
"step": 86800
},
{
"loss": 0.0899,
"grad_norm": 4.143392562866211,
"learning_rate": 7.131323011794137e-06,
"epoch": 5.292182072999817,
"step": 86850
},
{
"loss": 0.1288,
"grad_norm": 0.08140899240970612,
"learning_rate": 7.123305244421639e-06,
"epoch": 5.2952288099445495,
"step": 86900
},
{
"loss": 0.1431,
"grad_norm": 2.134497880935669,
"learning_rate": 7.115287477049142e-06,
"epoch": 5.298275546889282,
"step": 86950
},
{
"loss": 0.1201,
"grad_norm": 13.384051322937012,
"learning_rate": 7.1072697096766445e-06,
"epoch": 5.301322283834014,
"step": 87000
},
{
"loss": 0.143,
"grad_norm": 11.723172187805176,
"learning_rate": 7.099251942304146e-06,
"epoch": 5.304369020778746,
"step": 87050
},
{
"loss": 0.1519,
"grad_norm": 15.434045791625977,
"learning_rate": 7.0912341749316495e-06,
"epoch": 5.307415757723478,
"step": 87100
},
{
"loss": 0.1275,
"grad_norm": 9.52042007446289,
"learning_rate": 7.083216407559151e-06,
"epoch": 5.310462494668211,
"step": 87150
},
{
"loss": 0.1134,
"grad_norm": 11.312582015991211,
"learning_rate": 7.0751986401866545e-06,
"epoch": 5.313509231612943,
"step": 87200
},
{
"loss": 0.2056,
"grad_norm": 5.5997796058654785,
"learning_rate": 7.067180872814157e-06,
"epoch": 5.316555968557675,
"step": 87250
},
{
"loss": 0.1404,
"grad_norm": 7.152952671051025,
"learning_rate": 7.0591631054416595e-06,
"epoch": 5.319602705502407,
"step": 87300
},
{
"loss": 0.1276,
"grad_norm": 0.022854411974549294,
"learning_rate": 7.051145338069162e-06,
"epoch": 5.322649442447139,
"step": 87350
},
{
"loss": 0.1126,
"grad_norm": 0.38973402976989746,
"learning_rate": 7.043127570696664e-06,
"epoch": 5.325696179391871,
"step": 87400
},
{
"loss": 0.17,
"grad_norm": 8.142101287841797,
"learning_rate": 7.035109803324167e-06,
"epoch": 5.328742916336603,
"step": 87450
},
{
"loss": 0.1581,
"grad_norm": 9.770666122436523,
"learning_rate": 7.0270920359516695e-06,
"epoch": 5.331789653281335,
"step": 87500
},
{
"loss": 0.1134,
"grad_norm": 6.709127902984619,
"learning_rate": 7.019074268579172e-06,
"epoch": 5.334836390226068,
"step": 87550
},
{
"loss": 0.0752,
"grad_norm": 1.5978096723556519,
"learning_rate": 7.0110565012066745e-06,
"epoch": 5.3378831271708,
"step": 87600
},
{
"loss": 0.1103,
"grad_norm": 0.05920303240418434,
"learning_rate": 7.003038733834178e-06,
"epoch": 5.340929864115532,
"step": 87650
},
{
"loss": 0.1582,
"grad_norm": 18.255706787109375,
"learning_rate": 6.9950209664616795e-06,
"epoch": 5.343976601060264,
"step": 87700
},
{
"loss": 0.109,
"grad_norm": 1.0572799444198608,
"learning_rate": 6.987003199089182e-06,
"epoch": 5.3470233380049965,
"step": 87750
},
{
"loss": 0.151,
"grad_norm": 4.922333717346191,
"learning_rate": 6.978985431716685e-06,
"epoch": 5.350070074949729,
"step": 87800
},
{
"loss": 0.1096,
"grad_norm": 27.421701431274414,
"learning_rate": 6.970967664344187e-06,
"epoch": 5.353116811894461,
"step": 87850
},
{
"loss": 0.1906,
"grad_norm": 12.80792236328125,
"learning_rate": 6.96294989697169e-06,
"epoch": 5.356163548839193,
"step": 87900
},
{
"loss": 0.1327,
"grad_norm": 12.249297142028809,
"learning_rate": 6.954932129599192e-06,
"epoch": 5.359210285783925,
"step": 87950
},
{
"loss": 0.1172,
"grad_norm": 0.6164830923080444,
"learning_rate": 6.946914362226695e-06,
"epoch": 5.362257022728658,
"step": 88000
},
{
"loss": 0.1219,
"grad_norm": 9.190110206604004,
"learning_rate": 6.938896594854198e-06,
"epoch": 5.36530375967339,
"step": 88050
},
{
"loss": 0.1393,
"grad_norm": 9.392322540283203,
"learning_rate": 6.9308788274816995e-06,
"epoch": 5.368350496618122,
"step": 88100
},
{
"loss": 0.105,
"grad_norm": 0.012951495125889778,
"learning_rate": 6.922861060109203e-06,
"epoch": 5.371397233562854,
"step": 88150
},
{
"loss": 0.1546,
"grad_norm": 0.2759097218513489,
"learning_rate": 6.915164003431605e-06,
"epoch": 5.374443970507587,
"step": 88200
},
{
"loss": 0.1319,
"grad_norm": 0.06699617952108383,
"learning_rate": 6.9071462360591076e-06,
"epoch": 5.377490707452319,
"step": 88250
},
{
"loss": 0.0946,
"grad_norm": 5.9987945556640625,
"learning_rate": 6.89912846868661e-06,
"epoch": 5.380537444397051,
"step": 88300
},
{
"loss": 0.182,
"grad_norm": 6.0630879402160645,
"learning_rate": 6.891110701314112e-06,
"epoch": 5.383584181341783,
"step": 88350
},
{
"loss": 0.1517,
"grad_norm": 17.898292541503906,
"learning_rate": 6.883092933941615e-06,
"epoch": 5.3866309182865155,
"step": 88400
},
{
"loss": 0.1194,
"grad_norm": 3.1102070808410645,
"learning_rate": 6.875075166569118e-06,
"epoch": 5.389677655231248,
"step": 88450
},
{
"loss": 0.1109,
"grad_norm": 0.104013592004776,
"learning_rate": 6.867057399196621e-06,
"epoch": 5.39272439217598,
"step": 88500
},
{
"loss": 0.1523,
"grad_norm": 7.351625919342041,
"learning_rate": 6.859039631824123e-06,
"epoch": 5.395771129120711,
"step": 88550
},
{
"loss": 0.1405,
"grad_norm": 0.1634470820426941,
"learning_rate": 6.851021864451626e-06,
"epoch": 5.398817866065444,
"step": 88600
},
{
"loss": 0.1203,
"grad_norm": 5.7577805519104,
"learning_rate": 6.843004097079128e-06,
"epoch": 5.401864603010176,
"step": 88650
},
{
"loss": 0.1463,
"grad_norm": 1.4251629114151,
"learning_rate": 6.83498632970663e-06,
"epoch": 5.404911339954908,
"step": 88700
},
{
"loss": 0.1199,
"grad_norm": 8.578570365905762,
"learning_rate": 6.8269685623341334e-06,
"epoch": 5.40795807689964,
"step": 88750
},
{
"loss": 0.1071,
"grad_norm": 8.347601890563965,
"learning_rate": 6.818950794961635e-06,
"epoch": 5.4110048138443725,
"step": 88800
},
{
"loss": 0.1018,
"grad_norm": 0.0327901653945446,
"learning_rate": 6.8109330275891385e-06,
"epoch": 5.414051550789105,
"step": 88850
},
{
"loss": 0.1054,
"grad_norm": 5.676746845245361,
"learning_rate": 6.80291526021664e-06,
"epoch": 5.417098287733837,
"step": 88900
},
{
"loss": 0.1431,
"grad_norm": 0.18259505927562714,
"learning_rate": 6.7948974928441435e-06,
"epoch": 5.420145024678569,
"step": 88950
},
{
"loss": 0.1545,
"grad_norm": 2.8490512371063232,
"learning_rate": 6.786879725471646e-06,
"epoch": 5.423191761623301,
"step": 89000
},
{
"loss": 0.1512,
"grad_norm": 19.49395751953125,
"learning_rate": 6.7788619580991485e-06,
"epoch": 5.426238498568034,
"step": 89050
},
{
"loss": 0.096,
"grad_norm": 16.408903121948242,
"learning_rate": 6.770844190726651e-06,
"epoch": 5.429285235512766,
"step": 89100
},
{
"loss": 0.1562,
"grad_norm": 6.154325485229492,
"learning_rate": 6.762826423354153e-06,
"epoch": 5.432331972457498,
"step": 89150
},
{
"loss": 0.1396,
"grad_norm": 6.902485370635986,
"learning_rate": 6.754808655981656e-06,
"epoch": 5.43537870940223,
"step": 89200
},
{
"loss": 0.1466,
"grad_norm": 18.579957962036133,
"learning_rate": 6.7467908886091585e-06,
"epoch": 5.438425446346963,
"step": 89250
},
{
"loss": 0.1172,
"grad_norm": 0.044341783970594406,
"learning_rate": 6.738773121236661e-06,
"epoch": 5.441472183291695,
"step": 89300
},
{
"loss": 0.091,
"grad_norm": 5.638580322265625,
"learning_rate": 6.7307553538641635e-06,
"epoch": 5.444518920236427,
"step": 89350
},
{
"loss": 0.131,
"grad_norm": 3.1367039680480957,
"learning_rate": 6.722737586491667e-06,
"epoch": 5.447565657181159,
"step": 89400
},
{
"loss": 0.1608,
"grad_norm": 0.05878010392189026,
"learning_rate": 6.7147198191191685e-06,
"epoch": 5.4506123941258915,
"step": 89450
},
{
"loss": 0.1308,
"grad_norm": 0.7750746011734009,
"learning_rate": 6.706702051746671e-06,
"epoch": 5.453659131070624,
"step": 89500
},
{
"loss": 0.1493,
"grad_norm": 0.8640768527984619,
"learning_rate": 6.698684284374174e-06,
"epoch": 5.456705868015356,
"step": 89550
},
{
"loss": 0.1655,
"grad_norm": 0.3267796039581299,
"learning_rate": 6.690666517001676e-06,
"epoch": 5.459752604960087,
"step": 89600
},
{
"loss": 0.1023,
"grad_norm": 5.872455596923828,
"learning_rate": 6.682648749629179e-06,
"epoch": 5.4627993419048195,
"step": 89650
},
{
"loss": 0.1215,
"grad_norm": 9.562470436096191,
"learning_rate": 6.674630982256681e-06,
"epoch": 5.465846078849552,
"step": 89700
},
{
"loss": 0.1117,
"grad_norm": 12.58138656616211,
"learning_rate": 6.666613214884184e-06,
"epoch": 5.468892815794284,
"step": 89750
},
{
"loss": 0.2105,
"grad_norm": 15.17033863067627,
"learning_rate": 6.658595447511687e-06,
"epoch": 5.471939552739016,
"step": 89800
},
{
"loss": 0.1565,
"grad_norm": 0.32060492038726807,
"learning_rate": 6.6505776801391885e-06,
"epoch": 5.4749862896837485,
"step": 89850
},
{
"loss": 0.1324,
"grad_norm": 0.8286267518997192,
"learning_rate": 6.642559912766692e-06,
"epoch": 5.478033026628481,
"step": 89900
},
{
"loss": 0.1112,
"grad_norm": 15.35359001159668,
"learning_rate": 6.6345421453941935e-06,
"epoch": 5.481079763573213,
"step": 89950
},
{
"loss": 0.1426,
"grad_norm": 0.16423390805721283,
"learning_rate": 6.626524378021697e-06,
"epoch": 5.484126500517945,
"step": 90000
},
{
"loss": 0.1307,
"grad_norm": 28.77798080444336,
"learning_rate": 6.618506610649199e-06,
"epoch": 5.487173237462677,
"step": 90050
},
{
"loss": 0.141,
"grad_norm": 4.807619571685791,
"learning_rate": 6.610488843276702e-06,
"epoch": 5.49021997440741,
"step": 90100
},
{
"loss": 0.1521,
"grad_norm": 12.792724609375,
"learning_rate": 6.602471075904204e-06,
"epoch": 5.493266711352142,
"step": 90150
},
{
"loss": 0.1987,
"grad_norm": 11.406927108764648,
"learning_rate": 6.594453308531706e-06,
"epoch": 5.496313448296874,
"step": 90200
},
{
"loss": 0.1369,
"grad_norm": 5.793826103210449,
"learning_rate": 6.586435541159209e-06,
"epoch": 5.499360185241606,
"step": 90250
},
{
"loss": 0.1098,
"grad_norm": 0.14347496628761292,
"learning_rate": 6.578417773786712e-06,
"epoch": 5.5024069221863385,
"step": 90300
},
{
"loss": 0.1401,
"grad_norm": 4.080855846405029,
"learning_rate": 6.570400006414215e-06,
"epoch": 5.505453659131071,
"step": 90350
},
{
"loss": 0.1553,
"grad_norm": 0.07775057107210159,
"learning_rate": 6.562382239041717e-06,
"epoch": 5.508500396075803,
"step": 90400
},
{
"loss": 0.1272,
"grad_norm": 7.054189682006836,
"learning_rate": 6.55436447166922e-06,
"epoch": 5.511547133020535,
"step": 90450
},
{
"loss": 0.1179,
"grad_norm": 2.0467073917388916,
"learning_rate": 6.546346704296722e-06,
"epoch": 5.5145938699652675,
"step": 90500
},
{
"loss": 0.1324,
"grad_norm": 2.549372673034668,
"learning_rate": 6.538328936924224e-06,
"epoch": 5.51764060691,
"step": 90550
},
{
"loss": 0.1095,
"grad_norm": 16.164710998535156,
"learning_rate": 6.530311169551728e-06,
"epoch": 5.520687343854732,
"step": 90600
},
{
"loss": 0.1527,
"grad_norm": 17.95628547668457,
"learning_rate": 6.522293402179229e-06,
"epoch": 5.523734080799464,
"step": 90650
},
{
"loss": 0.1162,
"grad_norm": 6.167896270751953,
"learning_rate": 6.514275634806733e-06,
"epoch": 5.526780817744196,
"step": 90700
},
{
"loss": 0.1346,
"grad_norm": 2.9473111629486084,
"learning_rate": 6.5062578674342344e-06,
"epoch": 5.529827554688929,
"step": 90750
},
{
"loss": 0.1221,
"grad_norm": 0.8679468631744385,
"learning_rate": 6.498240100061738e-06,
"epoch": 5.532874291633661,
"step": 90800
},
{
"loss": 0.1191,
"grad_norm": 0.5534880757331848,
"learning_rate": 6.49022233268924e-06,
"epoch": 5.535921028578392,
"step": 90850
},
{
"loss": 0.1192,
"grad_norm": 0.1283053159713745,
"learning_rate": 6.482204565316742e-06,
"epoch": 5.538967765523124,
"step": 90900
},
{
"loss": 0.1104,
"grad_norm": 19.87505340576172,
"learning_rate": 6.474186797944245e-06,
"epoch": 5.542014502467857,
"step": 90950
},
{
"loss": 0.1293,
"grad_norm": 1.3316746950149536,
"learning_rate": 6.466169030571747e-06,
"epoch": 5.545061239412589,
"step": 91000
},
{
"loss": 0.1012,
"grad_norm": 0.6940693855285645,
"learning_rate": 6.45815126319925e-06,
"epoch": 5.548107976357321,
"step": 91050
},
{
"loss": 0.1444,
"grad_norm": 11.314190864562988,
"learning_rate": 6.450133495826753e-06,
"epoch": 5.551154713302053,
"step": 91100
},
{
"loss": 0.1015,
"grad_norm": 0.04372856765985489,
"learning_rate": 6.442115728454255e-06,
"epoch": 5.554201450246786,
"step": 91150
},
{
"loss": 0.1594,
"grad_norm": 5.118297100067139,
"learning_rate": 6.434097961081758e-06,
"epoch": 5.557248187191518,
"step": 91200
},
{
"loss": 0.1747,
"grad_norm": 0.33005309104919434,
"learning_rate": 6.4260801937092595e-06,
"epoch": 5.56029492413625,
"step": 91250
},
{
"loss": 0.1182,
"grad_norm": 2.6475133895874023,
"learning_rate": 6.418062426336763e-06,
"epoch": 5.563341661080982,
"step": 91300
},
{
"loss": 0.1369,
"grad_norm": 7.255090713500977,
"learning_rate": 6.410044658964265e-06,
"epoch": 5.5663883980257145,
"step": 91350
},
{
"loss": 0.1338,
"grad_norm": 0.4361167848110199,
"learning_rate": 6.402026891591769e-06,
"epoch": 5.569435134970447,
"step": 91400
},
{
"loss": 0.1034,
"grad_norm": 0.4741666316986084,
"learning_rate": 6.39400912421927e-06,
"epoch": 5.572481871915179,
"step": 91450
},
{
"loss": 0.1017,
"grad_norm": 0.48459458351135254,
"learning_rate": 6.385991356846774e-06,
"epoch": 5.575528608859911,
"step": 91500
},
{
"loss": 0.107,
"grad_norm": 1.0443086624145508,
"learning_rate": 6.377973589474275e-06,
"epoch": 5.578575345804643,
"step": 91550
},
{
"loss": 0.1405,
"grad_norm": 0.012131299823522568,
"learning_rate": 6.369955822101778e-06,
"epoch": 5.581622082749376,
"step": 91600
},
{
"loss": 0.1206,
"grad_norm": 3.075885772705078,
"learning_rate": 6.361938054729281e-06,
"epoch": 5.584668819694108,
"step": 91650
},
{
"loss": 0.1497,
"grad_norm": 4.662344455718994,
"learning_rate": 6.353920287356783e-06,
"epoch": 5.58771555663884,
"step": 91700
},
{
"loss": 0.1445,
"grad_norm": 18.73686981201172,
"learning_rate": 6.345902519984286e-06,
"epoch": 5.590762293583572,
"step": 91750
},
{
"loss": 0.1172,
"grad_norm": 7.2178730964660645,
"learning_rate": 6.337884752611788e-06,
"epoch": 5.593809030528304,
"step": 91800
},
{
"loss": 0.1461,
"grad_norm": 0.6771558523178101,
"learning_rate": 6.329866985239291e-06,
"epoch": 5.596855767473036,
"step": 91850
},
{
"loss": 0.1605,
"grad_norm": 10.380353927612305,
"learning_rate": 6.321849217866794e-06,
"epoch": 5.599902504417768,
"step": 91900
},
{
"loss": 0.1462,
"grad_norm": 0.13167616724967957,
"learning_rate": 6.313831450494295e-06,
"epoch": 5.6029492413625,
"step": 91950
},
{
"loss": 0.1454,
"grad_norm": 0.07599218189716339,
"learning_rate": 6.305813683121799e-06,
"epoch": 5.605995978307233,
"step": 92000
},
{
"loss": 0.1642,
"grad_norm": 7.853109359741211,
"learning_rate": 6.2977959157493e-06,
"epoch": 5.609042715251965,
"step": 92050
},
{
"loss": 0.1252,
"grad_norm": 13.73952579498291,
"learning_rate": 6.289778148376804e-06,
"epoch": 5.612089452196697,
"step": 92100
},
{
"loss": 0.1113,
"grad_norm": 7.313144683837891,
"learning_rate": 6.281760381004306e-06,
"epoch": 5.615136189141429,
"step": 92150
},
{
"loss": 0.1221,
"grad_norm": 3.6213297843933105,
"learning_rate": 6.2737426136318096e-06,
"epoch": 5.6181829260861615,
"step": 92200
},
{
"loss": 0.1335,
"grad_norm": 2.689807415008545,
"learning_rate": 6.265724846259311e-06,
"epoch": 5.621229663030894,
"step": 92250
},
{
"loss": 0.15,
"grad_norm": 0.009239738807082176,
"learning_rate": 6.257707078886813e-06,
"epoch": 5.624276399975626,
"step": 92300
},
{
"loss": 0.1212,
"grad_norm": 2.348738431930542,
"learning_rate": 6.249689311514316e-06,
"epoch": 5.627323136920358,
"step": 92350
},
{
"loss": 0.1536,
"grad_norm": 3.118744134902954,
"learning_rate": 6.241671544141819e-06,
"epoch": 5.6303698738650905,
"step": 92400
},
{
"loss": 0.1202,
"grad_norm": 0.48548606038093567,
"learning_rate": 6.233653776769322e-06,
"epoch": 5.633416610809823,
"step": 92450
},
{
"loss": 0.1233,
"grad_norm": 0.22487905621528625,
"learning_rate": 6.225636009396824e-06,
"epoch": 5.636463347754555,
"step": 92500
},
{
"loss": 0.1853,
"grad_norm": 5.947726249694824,
"learning_rate": 6.217618242024327e-06,
"epoch": 5.639510084699287,
"step": 92550
},
{
"loss": 0.0826,
"grad_norm": 5.513899326324463,
"learning_rate": 6.209600474651829e-06,
"epoch": 5.642556821644019,
"step": 92600
},
{
"loss": 0.17,
"grad_norm": 0.07429134100675583,
"learning_rate": 6.201582707279331e-06,
"epoch": 5.645603558588752,
"step": 92650
},
{
"loss": 0.1019,
"grad_norm": 5.624057769775391,
"learning_rate": 6.193564939906835e-06,
"epoch": 5.648650295533484,
"step": 92700
},
{
"loss": 0.1202,
"grad_norm": 0.23538900911808014,
"learning_rate": 6.185547172534336e-06,
"epoch": 5.651697032478216,
"step": 92750
},
{
"loss": 0.0874,
"grad_norm": 0.024944907054305077,
"learning_rate": 6.17752940516184e-06,
"epoch": 5.654743769422948,
"step": 92800
},
{
"loss": 0.153,
"grad_norm": 0.4419300854206085,
"learning_rate": 6.169511637789341e-06,
"epoch": 5.6577905063676805,
"step": 92850
},
{
"loss": 0.1131,
"grad_norm": 4.591179370880127,
"learning_rate": 6.161493870416845e-06,
"epoch": 5.660837243312413,
"step": 92900
},
{
"loss": 0.1246,
"grad_norm": 5.594015598297119,
"learning_rate": 6.153476103044347e-06,
"epoch": 5.663883980257145,
"step": 92950
},
{
"loss": 0.1222,
"grad_norm": 9.630775451660156,
"learning_rate": 6.145458335671849e-06,
"epoch": 5.666930717201877,
"step": 93000
},
{
"loss": 0.1691,
"grad_norm": 0.3479483723640442,
"learning_rate": 6.137440568299352e-06,
"epoch": 5.6699774541466095,
"step": 93050
},
{
"loss": 0.1301,
"grad_norm": 8.829850196838379,
"learning_rate": 6.129583156274305e-06,
"epoch": 5.673024191091341,
"step": 93100
},
{
"loss": 0.1078,
"grad_norm": 10.85731315612793,
"learning_rate": 6.1215653889018065e-06,
"epoch": 5.676070928036073,
"step": 93150
},
{
"loss": 0.1411,
"grad_norm": 4.869848728179932,
"learning_rate": 6.11354762152931e-06,
"epoch": 5.679117664980805,
"step": 93200
},
{
"loss": 0.1061,
"grad_norm": 0.6373572945594788,
"learning_rate": 6.105529854156812e-06,
"epoch": 5.6821644019255375,
"step": 93250
},
{
"loss": 0.1284,
"grad_norm": 7.1703667640686035,
"learning_rate": 6.097512086784314e-06,
"epoch": 5.68521113887027,
"step": 93300
},
{
"loss": 0.1077,
"grad_norm": 18.066795349121094,
"learning_rate": 6.089494319411817e-06,
"epoch": 5.688257875815002,
"step": 93350
},
{
"loss": 0.1167,
"grad_norm": 1.0252094268798828,
"learning_rate": 6.081476552039319e-06,
"epoch": 5.691304612759734,
"step": 93400
},
{
"loss": 0.0862,
"grad_norm": 0.13383492827415466,
"learning_rate": 6.073458784666822e-06,
"epoch": 5.694351349704466,
"step": 93450
},
{
"loss": 0.1587,
"grad_norm": 13.546426773071289,
"learning_rate": 6.065441017294325e-06,
"epoch": 5.697398086649199,
"step": 93500
},
{
"loss": 0.119,
"grad_norm": 8.705435752868652,
"learning_rate": 6.057423249921827e-06,
"epoch": 5.700444823593931,
"step": 93550
},
{
"loss": 0.135,
"grad_norm": 11.238054275512695,
"learning_rate": 6.04940548254933e-06,
"epoch": 5.703491560538663,
"step": 93600
},
{
"loss": 0.104,
"grad_norm": 6.624605655670166,
"learning_rate": 6.0413877151768316e-06,
"epoch": 5.706538297483395,
"step": 93650
},
{
"loss": 0.1516,
"grad_norm": 0.3517632782459259,
"learning_rate": 6.033369947804335e-06,
"epoch": 5.709585034428128,
"step": 93700
},
{
"loss": 0.1185,
"grad_norm": 1.7899101972579956,
"learning_rate": 6.025352180431837e-06,
"epoch": 5.71263177137286,
"step": 93750
},
{
"loss": 0.1414,
"grad_norm": 6.524298191070557,
"learning_rate": 6.01733441305934e-06,
"epoch": 5.715678508317592,
"step": 93800
},
{
"loss": 0.1593,
"grad_norm": 0.8891724348068237,
"learning_rate": 6.0093166456868424e-06,
"epoch": 5.718725245262324,
"step": 93850
},
{
"loss": 0.1335,
"grad_norm": 4.1735663414001465,
"learning_rate": 6.001298878314346e-06,
"epoch": 5.7217719822070565,
"step": 93900
},
{
"loss": 0.144,
"grad_norm": 3.3270063400268555,
"learning_rate": 5.9932811109418474e-06,
"epoch": 5.724818719151789,
"step": 93950
},
{
"loss": 0.1198,
"grad_norm": 0.7630071640014648,
"learning_rate": 5.98526334356935e-06,
"epoch": 5.727865456096521,
"step": 94000
},
{
"loss": 0.0895,
"grad_norm": 12.740693092346191,
"learning_rate": 5.9772455761968524e-06,
"epoch": 5.730912193041252,
"step": 94050
},
{
"loss": 0.1099,
"grad_norm": 0.6477550864219666,
"learning_rate": 5.969227808824355e-06,
"epoch": 5.7339589299859846,
"step": 94100
},
{
"loss": 0.0878,
"grad_norm": 3.970034599304199,
"learning_rate": 5.961210041451858e-06,
"epoch": 5.737005666930717,
"step": 94150
},
{
"loss": 0.1178,
"grad_norm": 7.248581886291504,
"learning_rate": 5.95319227407936e-06,
"epoch": 5.740052403875449,
"step": 94200
},
{
"loss": 0.0983,
"grad_norm": 2.352372169494629,
"learning_rate": 5.945174506706863e-06,
"epoch": 5.743099140820181,
"step": 94250
},
{
"loss": 0.1184,
"grad_norm": 3.5468578338623047,
"learning_rate": 5.937156739334366e-06,
"epoch": 5.7461458777649135,
"step": 94300
},
{
"loss": 0.0854,
"grad_norm": 3.5943763256073,
"learning_rate": 5.9291389719618675e-06,
"epoch": 5.749192614709646,
"step": 94350
},
{
"loss": 0.1607,
"grad_norm": 16.40985107421875,
"learning_rate": 5.921121204589371e-06,
"epoch": 5.752239351654378,
"step": 94400
},
{
"loss": 0.1393,
"grad_norm": 8.629164695739746,
"learning_rate": 5.9131034372168725e-06,
"epoch": 5.75528608859911,
"step": 94450
},
{
"loss": 0.1696,
"grad_norm": 2.760810136795044,
"learning_rate": 5.905085669844376e-06,
"epoch": 5.758332825543842,
"step": 94500
},
{
"loss": 0.1704,
"grad_norm": 17.170562744140625,
"learning_rate": 5.897067902471878e-06,
"epoch": 5.761379562488575,
"step": 94550
},
{
"loss": 0.1042,
"grad_norm": 8.285941123962402,
"learning_rate": 5.889050135099381e-06,
"epoch": 5.764426299433307,
"step": 94600
},
{
"loss": 0.1111,
"grad_norm": 5.593353271484375,
"learning_rate": 5.881032367726883e-06,
"epoch": 5.767473036378039,
"step": 94650
},
{
"loss": 0.1417,
"grad_norm": 7.027277946472168,
"learning_rate": 5.873014600354385e-06,
"epoch": 5.770519773322771,
"step": 94700
},
{
"loss": 0.1593,
"grad_norm": 0.14826470613479614,
"learning_rate": 5.864996832981888e-06,
"epoch": 5.773566510267504,
"step": 94750
},
{
"loss": 0.1157,
"grad_norm": 2.5026888847351074,
"learning_rate": 5.856979065609391e-06,
"epoch": 5.776613247212236,
"step": 94800
},
{
"loss": 0.1086,
"grad_norm": 0.23201937973499298,
"learning_rate": 5.848961298236893e-06,
"epoch": 5.779659984156968,
"step": 94850
},
{
"loss": 0.1054,
"grad_norm": 7.080366134643555,
"learning_rate": 5.840943530864396e-06,
"epoch": 5.7827067211017,
"step": 94900
},
{
"loss": 0.0911,
"grad_norm": 3.875303030014038,
"learning_rate": 5.832925763491899e-06,
"epoch": 5.7857534580464325,
"step": 94950
},
{
"loss": 0.1216,
"grad_norm": 12.733480453491211,
"learning_rate": 5.824907996119401e-06,
"epoch": 5.788800194991165,
"step": 95000
},
{
"loss": 0.1041,
"grad_norm": 6.427713394165039,
"learning_rate": 5.816890228746903e-06,
"epoch": 5.791846931935897,
"step": 95050
},
{
"loss": 0.1655,
"grad_norm": 15.189671516418457,
"learning_rate": 5.808872461374407e-06,
"epoch": 5.794893668880629,
"step": 95100
},
{
"loss": 0.1517,
"grad_norm": 3.626155138015747,
"learning_rate": 5.801015049349359e-06,
"epoch": 5.797940405825361,
"step": 95150
},
{
"loss": 0.1259,
"grad_norm": 9.089632034301758,
"learning_rate": 5.792997281976861e-06,
"epoch": 5.800987142770094,
"step": 95200
},
{
"loss": 0.129,
"grad_norm": 1.9781885147094727,
"learning_rate": 5.7849795146043645e-06,
"epoch": 5.804033879714826,
"step": 95250
},
{
"loss": 0.1348,
"grad_norm": 6.448685646057129,
"learning_rate": 5.776961747231866e-06,
"epoch": 5.807080616659558,
"step": 95300
},
{
"loss": 0.1558,
"grad_norm": 0.48980405926704407,
"learning_rate": 5.7689439798593695e-06,
"epoch": 5.8101273536042894,
"step": 95350
},
{
"loss": 0.0813,
"grad_norm": 2.787339448928833,
"learning_rate": 5.760926212486871e-06,
"epoch": 5.813174090549022,
"step": 95400
},
{
"loss": 0.1311,
"grad_norm": 2.834056854248047,
"learning_rate": 5.752908445114374e-06,
"epoch": 5.816220827493754,
"step": 95450
},
{
"loss": 0.1695,
"grad_norm": 2.2491960525512695,
"learning_rate": 5.744890677741877e-06,
"epoch": 5.819267564438486,
"step": 95500
},
{
"loss": 0.1045,
"grad_norm": 2.075270652770996,
"learning_rate": 5.736872910369379e-06,
"epoch": 5.822314301383218,
"step": 95550
},
{
"loss": 0.0839,
"grad_norm": 12.828266143798828,
"learning_rate": 5.728855142996882e-06,
"epoch": 5.825361038327951,
"step": 95600
},
{
"loss": 0.135,
"grad_norm": 3.6608312129974365,
"learning_rate": 5.720837375624384e-06,
"epoch": 5.828407775272683,
"step": 95650
},
{
"loss": 0.2489,
"grad_norm": 0.19221070408821106,
"learning_rate": 5.712819608251887e-06,
"epoch": 5.831454512217415,
"step": 95700
},
{
"loss": 0.1127,
"grad_norm": 9.532238960266113,
"learning_rate": 5.7048018408793895e-06,
"epoch": 5.834501249162147,
"step": 95750
},
{
"loss": 0.1086,
"grad_norm": 10.035441398620605,
"learning_rate": 5.696784073506891e-06,
"epoch": 5.8375479861068795,
"step": 95800
},
{
"loss": 0.1021,
"grad_norm": 1.0827982425689697,
"learning_rate": 5.6887663061343945e-06,
"epoch": 5.840594723051612,
"step": 95850
},
{
"loss": 0.155,
"grad_norm": 4.287724494934082,
"learning_rate": 5.680748538761896e-06,
"epoch": 5.843641459996344,
"step": 95900
},
{
"loss": 0.0995,
"grad_norm": 3.3412530422210693,
"learning_rate": 5.6727307713893995e-06,
"epoch": 5.846688196941076,
"step": 95950
},
{
"loss": 0.1262,
"grad_norm": 13.091911315917969,
"learning_rate": 5.664713004016902e-06,
"epoch": 5.8497349338858085,
"step": 96000
},
{
"loss": 0.0964,
"grad_norm": 5.130334854125977,
"learning_rate": 5.656695236644405e-06,
"epoch": 5.852781670830541,
"step": 96050
},
{
"loss": 0.0873,
"grad_norm": 1.551395297050476,
"learning_rate": 5.648677469271907e-06,
"epoch": 5.855828407775273,
"step": 96100
},
{
"loss": 0.1013,
"grad_norm": 6.219453811645508,
"learning_rate": 5.6406597018994095e-06,
"epoch": 5.858875144720005,
"step": 96150
},
{
"loss": 0.1321,
"grad_norm": 7.026537895202637,
"learning_rate": 5.632641934526912e-06,
"epoch": 5.861921881664737,
"step": 96200
},
{
"loss": 0.1455,
"grad_norm": 43.979000091552734,
"learning_rate": 5.6246241671544145e-06,
"epoch": 5.86496861860947,
"step": 96250
},
{
"loss": 0.1258,
"grad_norm": 20.421499252319336,
"learning_rate": 5.616606399781918e-06,
"epoch": 5.868015355554201,
"step": 96300
},
{
"loss": 0.1304,
"grad_norm": 13.100213050842285,
"learning_rate": 5.6085886324094195e-06,
"epoch": 5.871062092498933,
"step": 96350
},
{
"loss": 0.1265,
"grad_norm": 8.620340347290039,
"learning_rate": 5.600570865036923e-06,
"epoch": 5.874108829443665,
"step": 96400
},
{
"loss": 0.1312,
"grad_norm": 3.285090684890747,
"learning_rate": 5.5925530976644245e-06,
"epoch": 5.877155566388398,
"step": 96450
},
{
"loss": 0.105,
"grad_norm": 0.7528755068778992,
"learning_rate": 5.584535330291927e-06,
"epoch": 5.88020230333313,
"step": 96500
},
{
"loss": 0.1183,
"grad_norm": 0.3727531433105469,
"learning_rate": 5.57651756291943e-06,
"epoch": 5.883249040277862,
"step": 96550
},
{
"loss": 0.1336,
"grad_norm": 0.23867398500442505,
"learning_rate": 5.568499795546932e-06,
"epoch": 5.886295777222594,
"step": 96600
},
{
"loss": 0.1347,
"grad_norm": 10.955134391784668,
"learning_rate": 5.560482028174435e-06,
"epoch": 5.889342514167327,
"step": 96650
},
{
"loss": 0.1348,
"grad_norm": 2.6669681072235107,
"learning_rate": 5.552464260801937e-06,
"epoch": 5.892389251112059,
"step": 96700
},
{
"loss": 0.1002,
"grad_norm": 14.541197776794434,
"learning_rate": 5.54444649342944e-06,
"epoch": 5.895435988056791,
"step": 96750
},
{
"loss": 0.1235,
"grad_norm": 0.1584271490573883,
"learning_rate": 5.536428726056943e-06,
"epoch": 5.898482725001523,
"step": 96800
},
{
"loss": 0.1345,
"grad_norm": 5.223379611968994,
"learning_rate": 5.5284109586844446e-06,
"epoch": 5.9015294619462555,
"step": 96850
},
{
"loss": 0.105,
"grad_norm": 0.028294894844293594,
"learning_rate": 5.520393191311948e-06,
"epoch": 5.904576198890988,
"step": 96900
},
{
"loss": 0.1063,
"grad_norm": 0.9608124494552612,
"learning_rate": 5.5123754239394504e-06,
"epoch": 5.90762293583572,
"step": 96950
},
{
"loss": 0.1454,
"grad_norm": 1.07574462890625,
"learning_rate": 5.504357656566953e-06,
"epoch": 5.910669672780452,
"step": 97000
},
{
"loss": 0.1058,
"grad_norm": 8.07391357421875,
"learning_rate": 5.4963398891944554e-06,
"epoch": 5.913716409725184,
"step": 97050
},
{
"loss": 0.1254,
"grad_norm": 31.002660751342773,
"learning_rate": 5.488322121821959e-06,
"epoch": 5.916763146669917,
"step": 97100
},
{
"loss": 0.1101,
"grad_norm": 0.16809025406837463,
"learning_rate": 5.4803043544494604e-06,
"epoch": 5.919809883614649,
"step": 97150
},
{
"loss": 0.1198,
"grad_norm": 6.141071796417236,
"learning_rate": 5.472286587076963e-06,
"epoch": 5.922856620559381,
"step": 97200
},
{
"loss": 0.1407,
"grad_norm": 4.683712005615234,
"learning_rate": 5.4642688197044654e-06,
"epoch": 5.925903357504113,
"step": 97250
},
{
"loss": 0.1271,
"grad_norm": 0.04413143917918205,
"learning_rate": 5.456251052331968e-06,
"epoch": 5.928950094448846,
"step": 97300
},
{
"loss": 0.1306,
"grad_norm": 10.23945140838623,
"learning_rate": 5.448233284959471e-06,
"epoch": 5.931996831393578,
"step": 97350
},
{
"loss": 0.1323,
"grad_norm": 0.680746853351593,
"learning_rate": 5.440215517586973e-06,
"epoch": 5.93504356833831,
"step": 97400
},
{
"loss": 0.1771,
"grad_norm": 25.099395751953125,
"learning_rate": 5.432197750214476e-06,
"epoch": 5.938090305283042,
"step": 97450
},
{
"loss": 0.1109,
"grad_norm": 0.00788741186261177,
"learning_rate": 5.424179982841978e-06,
"epoch": 5.9411370422277745,
"step": 97500
},
{
"loss": 0.1167,
"grad_norm": 0.22878937423229218,
"learning_rate": 5.4161622154694805e-06,
"epoch": 5.944183779172507,
"step": 97550
},
{
"loss": 0.1257,
"grad_norm": 18.001893997192383,
"learning_rate": 5.408144448096984e-06,
"epoch": 5.947230516117238,
"step": 97600
},
{
"loss": 0.1191,
"grad_norm": 0.07348685711622238,
"learning_rate": 5.4001266807244855e-06,
"epoch": 5.95027725306197,
"step": 97650
},
{
"loss": 0.1324,
"grad_norm": 1.1221145391464233,
"learning_rate": 5.392108913351989e-06,
"epoch": 5.9533239900067025,
"step": 97700
},
{
"loss": 0.1525,
"grad_norm": 2.7785961627960205,
"learning_rate": 5.3840911459794905e-06,
"epoch": 5.956370726951435,
"step": 97750
},
{
"loss": 0.1279,
"grad_norm": 4.602593898773193,
"learning_rate": 5.376073378606994e-06,
"epoch": 5.959417463896167,
"step": 97800
},
{
"loss": 0.1488,
"grad_norm": 8.339412689208984,
"learning_rate": 5.368055611234496e-06,
"epoch": 5.962464200840899,
"step": 97850
},
{
"loss": 0.1159,
"grad_norm": 6.0765509605407715,
"learning_rate": 5.360037843861998e-06,
"epoch": 5.9655109377856315,
"step": 97900
},
{
"loss": 0.1458,
"grad_norm": 0.45893725752830505,
"learning_rate": 5.352020076489501e-06,
"epoch": 5.968557674730364,
"step": 97950
},
{
"loss": 0.134,
"grad_norm": 11.298267364501953,
"learning_rate": 5.344002309117004e-06,
"epoch": 5.971604411675096,
"step": 98000
},
{
"loss": 0.1181,
"grad_norm": 7.205318450927734,
"learning_rate": 5.335984541744506e-06,
"epoch": 5.974651148619828,
"step": 98050
},
{
"loss": 0.1482,
"grad_norm": 9.350260734558105,
"learning_rate": 5.327966774372009e-06,
"epoch": 5.97769788556456,
"step": 98100
},
{
"loss": 0.1255,
"grad_norm": 4.684756278991699,
"learning_rate": 5.319949006999512e-06,
"epoch": 5.980744622509293,
"step": 98150
},
{
"loss": 0.0922,
"grad_norm": 9.500842094421387,
"learning_rate": 5.311931239627014e-06,
"epoch": 5.983791359454025,
"step": 98200
},
{
"loss": 0.1216,
"grad_norm": 9.2393217086792,
"learning_rate": 5.303913472254516e-06,
"epoch": 5.986838096398757,
"step": 98250
},
{
"loss": 0.1453,
"grad_norm": 11.546034812927246,
"learning_rate": 5.295895704882019e-06,
"epoch": 5.989884833343489,
"step": 98300
},
{
"loss": 0.12,
"grad_norm": 5.663444519042969,
"learning_rate": 5.288038292856972e-06,
"epoch": 5.9929315702882215,
"step": 98350
},
{
"loss": 0.1101,
"grad_norm": 1.1453205347061157,
"learning_rate": 5.280020525484474e-06,
"epoch": 5.995978307232954,
"step": 98400
},
{
"loss": 0.1004,
"grad_norm": 10.675846099853516,
"learning_rate": 5.272002758111977e-06,
"epoch": 5.999025044177686,
"step": 98450
},
{
"eval_loss": 0.11651920527219772,
"eval_accuracy": 0.9717566409597258,
"eval_f1_macro": 0.9713303556587457,
"eval_top5_accuracy": 0.9965381319622965,
"eval_runtime": 217.8039,
"eval_samples_per_second": 133.951,
"eval_steps_per_second": 8.375,
"epoch": 6.0,
"step": 98466
},
{
"loss": 0.0922,
"grad_norm": 3.5013139247894287,
"learning_rate": 5.263984990739479e-06,
"epoch": 6.002071781122418,
"step": 98500
},
{
"loss": 0.1144,
"grad_norm": 0.06343896687030792,
"learning_rate": 5.255967223366981e-06,
"epoch": 6.0051185180671505,
"step": 98550
},
{
"loss": 0.1005,
"grad_norm": 2.021289587020874,
"learning_rate": 5.247949455994484e-06,
"epoch": 6.008165255011883,
"step": 98600
},
{
"loss": 0.1106,
"grad_norm": 1.1214898824691772,
"learning_rate": 5.239931688621987e-06,
"epoch": 6.011211991956614,
"step": 98650
},
{
"loss": 0.1461,
"grad_norm": 11.966567039489746,
"learning_rate": 5.23191392124949e-06,
"epoch": 6.014258728901346,
"step": 98700
},
{
"loss": 0.0749,
"grad_norm": 7.507488250732422,
"learning_rate": 5.223896153876992e-06,
"epoch": 6.0173054658460785,
"step": 98750
},
{
"loss": 0.1603,
"grad_norm": 16.45906639099121,
"learning_rate": 5.215878386504495e-06,
"epoch": 6.020352202790811,
"step": 98800
},
{
"loss": 0.1184,
"grad_norm": 6.209747791290283,
"learning_rate": 5.207860619131997e-06,
"epoch": 6.023398939735543,
"step": 98850
},
{
"loss": 0.1179,
"grad_norm": 32.41434097290039,
"learning_rate": 5.199842851759499e-06,
"epoch": 6.026445676680275,
"step": 98900
},
{
"loss": 0.1118,
"grad_norm": 0.4788309931755066,
"learning_rate": 5.1918250843870025e-06,
"epoch": 6.029492413625007,
"step": 98950
},
{
"loss": 0.1049,
"grad_norm": 0.13484643399715424,
"learning_rate": 5.183807317014504e-06,
"epoch": 6.03253915056974,
"step": 99000
},
{
"loss": 0.0995,
"grad_norm": 5.8085246086120605,
"learning_rate": 5.1757895496420075e-06,
"epoch": 6.035585887514472,
"step": 99050
},
{
"loss": 0.0984,
"grad_norm": 0.02807340770959854,
"learning_rate": 5.167771782269509e-06,
"epoch": 6.038632624459204,
"step": 99100
},
{
"loss": 0.1296,
"grad_norm": 12.43993091583252,
"learning_rate": 5.1597540148970125e-06,
"epoch": 6.041679361403936,
"step": 99150
},
{
"loss": 0.1034,
"grad_norm": 7.797327518463135,
"learning_rate": 5.151736247524515e-06,
"epoch": 6.044726098348669,
"step": 99200
},
{
"loss": 0.0729,
"grad_norm": 2.413972854614258,
"learning_rate": 5.143718480152017e-06,
"epoch": 6.047772835293401,
"step": 99250
},
{
"loss": 0.1131,
"grad_norm": 0.17976707220077515,
"learning_rate": 5.13570071277952e-06,
"epoch": 6.050819572238133,
"step": 99300
},
{
"loss": 0.0926,
"grad_norm": 0.9632896780967712,
"learning_rate": 5.127682945407022e-06,
"epoch": 6.053866309182865,
"step": 99350
},
{
"loss": 0.1181,
"grad_norm": 0.9285658001899719,
"learning_rate": 5.119665178034525e-06,
"epoch": 6.0569130461275975,
"step": 99400
},
{
"loss": 0.135,
"grad_norm": 5.392992973327637,
"learning_rate": 5.1116474106620275e-06,
"epoch": 6.05995978307233,
"step": 99450
},
{
"loss": 0.0896,
"grad_norm": 0.11282576620578766,
"learning_rate": 5.103629643289531e-06,
"epoch": 6.063006520017062,
"step": 99500
},
{
"loss": 0.1452,
"grad_norm": 6.388367176055908,
"learning_rate": 5.0956118759170325e-06,
"epoch": 6.066053256961794,
"step": 99550
},
{
"loss": 0.1447,
"grad_norm": 7.847933769226074,
"learning_rate": 5.087594108544534e-06,
"epoch": 6.069099993906526,
"step": 99600
},
{
"loss": 0.1279,
"grad_norm": 15.5547513961792,
"learning_rate": 5.0795763411720376e-06,
"epoch": 6.072146730851259,
"step": 99650
},
{
"loss": 0.1151,
"grad_norm": 6.834897518157959,
"learning_rate": 5.07155857379954e-06,
"epoch": 6.075193467795991,
"step": 99700
},
{
"loss": 0.1408,
"grad_norm": 1.0162161588668823,
"learning_rate": 5.063540806427043e-06,
"epoch": 6.078240204740723,
"step": 99750
},
{
"loss": 0.1232,
"grad_norm": 17.747535705566406,
"learning_rate": 5.055523039054545e-06,
"epoch": 6.0812869416854545,
"step": 99800
},
{
"loss": 0.1325,
"grad_norm": 0.015111980959773064,
"learning_rate": 5.047505271682048e-06,
"epoch": 6.084333678630187,
"step": 99850
},
{
"loss": 0.1015,
"grad_norm": 0.005835792515426874,
"learning_rate": 5.03948750430955e-06,
"epoch": 6.087380415574919,
"step": 99900
},
{
"loss": 0.1034,
"grad_norm": 4.3124895095825195,
"learning_rate": 5.0314697369370526e-06,
"epoch": 6.090427152519651,
"step": 99950
},
{
"loss": 0.1009,
"grad_norm": 1.4615000486373901,
"learning_rate": 5.023451969564556e-06,
"epoch": 6.093473889464383,
"step": 100000
},
{
"loss": 0.177,
"grad_norm": 0.12826162576675415,
"learning_rate": 5.015434202192058e-06,
"epoch": 6.096520626409116,
"step": 100050
},
{
"loss": 0.0668,
"grad_norm": 0.9117108583450317,
"learning_rate": 5.007416434819561e-06,
"epoch": 6.099567363353848,
"step": 100100
},
{
"loss": 0.1121,
"grad_norm": 0.0758165791630745,
"learning_rate": 4.999398667447063e-06,
"epoch": 6.10261410029858,
"step": 100150
},
{
"loss": 0.1432,
"grad_norm": 0.17027536034584045,
"learning_rate": 4.991380900074566e-06,
"epoch": 6.105660837243312,
"step": 100200
},
{
"loss": 0.1302,
"grad_norm": 0.5880535244941711,
"learning_rate": 4.9833631327020684e-06,
"epoch": 6.1087075741880446,
"step": 100250
},
{
"loss": 0.1043,
"grad_norm": 12.212766647338867,
"learning_rate": 4.975345365329571e-06,
"epoch": 6.111754311132777,
"step": 100300
},
{
"loss": 0.1219,
"grad_norm": 5.4334282875061035,
"learning_rate": 4.9673275979570734e-06,
"epoch": 6.114801048077509,
"step": 100350
},
{
"loss": 0.0941,
"grad_norm": 5.8105082511901855,
"learning_rate": 4.959309830584576e-06,
"epoch": 6.117847785022241,
"step": 100400
},
{
"loss": 0.1066,
"grad_norm": 6.381535053253174,
"learning_rate": 4.9512920632120785e-06,
"epoch": 6.1208945219669735,
"step": 100450
},
{
"loss": 0.0907,
"grad_norm": 1.9252469539642334,
"learning_rate": 4.943274295839581e-06,
"epoch": 6.123941258911706,
"step": 100500
},
{
"loss": 0.1133,
"grad_norm": 8.460347175598145,
"learning_rate": 4.9352565284670835e-06,
"epoch": 6.126987995856438,
"step": 100550
},
{
"loss": 0.0932,
"grad_norm": 1.4753057956695557,
"learning_rate": 4.927238761094586e-06,
"epoch": 6.13003473280117,
"step": 100600
},
{
"loss": 0.089,
"grad_norm": 15.988399505615234,
"learning_rate": 4.9192209937220885e-06,
"epoch": 6.133081469745902,
"step": 100650
},
{
"loss": 0.1462,
"grad_norm": 1.9219744205474854,
"learning_rate": 4.911203226349591e-06,
"epoch": 6.136128206690635,
"step": 100700
},
{
"loss": 0.1262,
"grad_norm": 0.02226112224161625,
"learning_rate": 4.9031854589770935e-06,
"epoch": 6.139174943635367,
"step": 100750
},
{
"loss": 0.1119,
"grad_norm": Infinity,
"learning_rate": 4.895167691604596e-06,
"epoch": 6.142221680580099,
"step": 100800
},
{
"loss": 0.0805,
"grad_norm": 4.788501739501953,
"learning_rate": 4.887310279579549e-06,
"epoch": 6.145268417524831,
"step": 100850
},
{
"loss": 0.1139,
"grad_norm": 3.6760923862457275,
"learning_rate": 4.879292512207051e-06,
"epoch": 6.148315154469563,
"step": 100900
},
{
"loss": 0.0734,
"grad_norm": 16.159727096557617,
"learning_rate": 4.871274744834554e-06,
"epoch": 6.151361891414295,
"step": 100950
},
{
"loss": 0.107,
"grad_norm": 9.336654663085938,
"learning_rate": 4.863256977462056e-06,
"epoch": 6.154408628359027,
"step": 101000
},
{
"loss": 0.1159,
"grad_norm": 6.604520797729492,
"learning_rate": 4.855239210089559e-06,
"epoch": 6.157455365303759,
"step": 101050
},
{
"loss": 0.1,
"grad_norm": 0.45561620593070984,
"learning_rate": 4.847221442717061e-06,
"epoch": 6.160502102248492,
"step": 101100
},
{
"loss": 0.1178,
"grad_norm": 7.558012962341309,
"learning_rate": 4.839203675344564e-06,
"epoch": 6.163548839193224,
"step": 101150
},
{
"loss": 0.1222,
"grad_norm": 2.358024835586548,
"learning_rate": 4.831185907972066e-06,
"epoch": 6.166595576137956,
"step": 101200
},
{
"loss": 0.1367,
"grad_norm": 0.06460610777139664,
"learning_rate": 4.823168140599569e-06,
"epoch": 6.169642313082688,
"step": 101250
},
{
"loss": 0.0843,
"grad_norm": 8.904021263122559,
"learning_rate": 4.815150373227071e-06,
"epoch": 6.1726890500274205,
"step": 101300
},
{
"loss": 0.1063,
"grad_norm": 0.04964457079768181,
"learning_rate": 4.807132605854575e-06,
"epoch": 6.175735786972153,
"step": 101350
},
{
"loss": 0.0924,
"grad_norm": 19.365276336669922,
"learning_rate": 4.799114838482077e-06,
"epoch": 6.178782523916885,
"step": 101400
},
{
"loss": 0.0997,
"grad_norm": 4.565613746643066,
"learning_rate": 4.79109707110958e-06,
"epoch": 6.181829260861617,
"step": 101450
},
{
"loss": 0.1644,
"grad_norm": 15.264656066894531,
"learning_rate": 4.783079303737081e-06,
"epoch": 6.1848759978063494,
"step": 101500
},
{
"loss": 0.0979,
"grad_norm": 4.733834743499756,
"learning_rate": 4.775061536364584e-06,
"epoch": 6.187922734751082,
"step": 101550
},
{
"loss": 0.107,
"grad_norm": 5.986477851867676,
"learning_rate": 4.767043768992087e-06,
"epoch": 6.190969471695814,
"step": 101600
},
{
"loss": 0.122,
"grad_norm": 7.619190692901611,
"learning_rate": 4.75902600161959e-06,
"epoch": 6.194016208640546,
"step": 101650
},
{
"loss": 0.095,
"grad_norm": 1.2970490455627441,
"learning_rate": 4.751008234247092e-06,
"epoch": 6.197062945585278,
"step": 101700
},
{
"loss": 0.1119,
"grad_norm": 2.0682148933410645,
"learning_rate": 4.742990466874595e-06,
"epoch": 6.200109682530011,
"step": 101750
},
{
"loss": 0.1088,
"grad_norm": 0.8900957107543945,
"learning_rate": 4.734972699502097e-06,
"epoch": 6.203156419474743,
"step": 101800
},
{
"loss": 0.0894,
"grad_norm": 8.241122245788574,
"learning_rate": 4.7269549321296e-06,
"epoch": 6.206203156419475,
"step": 101850
},
{
"loss": 0.1175,
"grad_norm": 0.08001118153333664,
"learning_rate": 4.718937164757102e-06,
"epoch": 6.209249893364207,
"step": 101900
},
{
"loss": 0.1137,
"grad_norm": 13.826456069946289,
"learning_rate": 4.710919397384605e-06,
"epoch": 6.2122966303089395,
"step": 101950
},
{
"loss": 0.1304,
"grad_norm": 24.286483764648438,
"learning_rate": 4.702901630012107e-06,
"epoch": 6.215343367253672,
"step": 102000
},
{
"loss": 0.0737,
"grad_norm": 0.04391051083803177,
"learning_rate": 4.69488386263961e-06,
"epoch": 6.218390104198403,
"step": 102050
},
{
"loss": 0.097,
"grad_norm": 7.025595664978027,
"learning_rate": 4.686866095267112e-06,
"epoch": 6.221436841143135,
"step": 102100
},
{
"loss": 0.1261,
"grad_norm": 5.073550701141357,
"learning_rate": 4.678848327894615e-06,
"epoch": 6.224483578087868,
"step": 102150
},
{
"loss": 0.1233,
"grad_norm": 0.24035510420799255,
"learning_rate": 4.670830560522117e-06,
"epoch": 6.2275303150326,
"step": 102200
},
{
"loss": 0.1226,
"grad_norm": 5.63332462310791,
"learning_rate": 4.66281279314962e-06,
"epoch": 6.230577051977332,
"step": 102250
},
{
"loss": 0.1211,
"grad_norm": 0.3232482969760895,
"learning_rate": 4.654795025777122e-06,
"epoch": 6.233623788922064,
"step": 102300
},
{
"loss": 0.0975,
"grad_norm": 7.074029922485352,
"learning_rate": 4.646777258404625e-06,
"epoch": 6.2366705258667965,
"step": 102350
},
{
"loss": 0.0849,
"grad_norm": 7.9857072830200195,
"learning_rate": 4.638759491032128e-06,
"epoch": 6.239717262811529,
"step": 102400
},
{
"loss": 0.0617,
"grad_norm": 3.251744031906128,
"learning_rate": 4.6307417236596305e-06,
"epoch": 6.242763999756261,
"step": 102450
},
{
"loss": 0.1451,
"grad_norm": 6.38430118560791,
"learning_rate": 4.622723956287133e-06,
"epoch": 6.245810736700993,
"step": 102500
},
{
"loss": 0.0678,
"grad_norm": 4.253998756408691,
"learning_rate": 4.614706188914635e-06,
"epoch": 6.248857473645725,
"step": 102550
},
{
"loss": 0.0764,
"grad_norm": 2.4254322052001953,
"learning_rate": 4.606688421542137e-06,
"epoch": 6.251904210590458,
"step": 102600
},
{
"loss": 0.1484,
"grad_norm": 3.310840606689453,
"learning_rate": 4.5986706541696406e-06,
"epoch": 6.25495094753519,
"step": 102650
},
{
"loss": 0.1402,
"grad_norm": 3.2929577827453613,
"learning_rate": 4.590652886797143e-06,
"epoch": 6.257997684479922,
"step": 102700
},
{
"loss": 0.1111,
"grad_norm": 13.968470573425293,
"learning_rate": 4.5826351194246456e-06,
"epoch": 6.261044421424654,
"step": 102750
},
{
"loss": 0.0902,
"grad_norm": 5.690467357635498,
"learning_rate": 4.574617352052148e-06,
"epoch": 6.264091158369387,
"step": 102800
},
{
"loss": 0.1126,
"grad_norm": 0.029029669240117073,
"learning_rate": 4.5665995846796506e-06,
"epoch": 6.267137895314119,
"step": 102850
},
{
"loss": 0.1129,
"grad_norm": 10.870767593383789,
"learning_rate": 4.558581817307153e-06,
"epoch": 6.270184632258851,
"step": 102900
},
{
"loss": 0.1199,
"grad_norm": 0.05678568407893181,
"learning_rate": 4.5505640499346556e-06,
"epoch": 6.273231369203583,
"step": 102950
},
{
"loss": 0.0998,
"grad_norm": 4.92418098449707,
"learning_rate": 4.542546282562158e-06,
"epoch": 6.2762781061483155,
"step": 103000
},
{
"loss": 0.1157,
"grad_norm": 7.050637722015381,
"learning_rate": 4.534528515189661e-06,
"epoch": 6.279324843093048,
"step": 103050
},
{
"loss": 0.1082,
"grad_norm": 0.30183130502700806,
"learning_rate": 4.526510747817163e-06,
"epoch": 6.282371580037779,
"step": 103100
},
{
"loss": 0.0965,
"grad_norm": 2.2576076984405518,
"learning_rate": 4.518492980444666e-06,
"epoch": 6.285418316982511,
"step": 103150
},
{
"loss": 0.1445,
"grad_norm": 0.015528878197073936,
"learning_rate": 4.510635568419618e-06,
"epoch": 6.2884650539272435,
"step": 103200
},
{
"loss": 0.1245,
"grad_norm": 7.966018199920654,
"learning_rate": 4.502617801047121e-06,
"epoch": 6.291511790871976,
"step": 103250
},
{
"loss": 0.1006,
"grad_norm": 0.0386066734790802,
"learning_rate": 4.494600033674623e-06,
"epoch": 6.294558527816708,
"step": 103300
},
{
"loss": 0.0969,
"grad_norm": 0.012124570086598396,
"learning_rate": 4.486582266302126e-06,
"epoch": 6.29760526476144,
"step": 103350
},
{
"loss": 0.102,
"grad_norm": 18.668869018554688,
"learning_rate": 4.478564498929628e-06,
"epoch": 6.3006520017061725,
"step": 103400
},
{
"loss": 0.0957,
"grad_norm": 7.437379837036133,
"learning_rate": 4.470546731557131e-06,
"epoch": 6.303698738650905,
"step": 103450
},
{
"loss": 0.1569,
"grad_norm": 21.520143508911133,
"learning_rate": 4.462528964184633e-06,
"epoch": 6.306745475595637,
"step": 103500
},
{
"loss": 0.0941,
"grad_norm": 13.22476577758789,
"learning_rate": 4.454511196812136e-06,
"epoch": 6.309792212540369,
"step": 103550
},
{
"loss": 0.1335,
"grad_norm": 0.6633058786392212,
"learning_rate": 4.446493429439638e-06,
"epoch": 6.312838949485101,
"step": 103600
},
{
"loss": 0.072,
"grad_norm": 1.7746915817260742,
"learning_rate": 4.438475662067141e-06,
"epoch": 6.315885686429834,
"step": 103650
},
{
"loss": 0.1094,
"grad_norm": 6.02217435836792,
"learning_rate": 4.430457894694643e-06,
"epoch": 6.318932423374566,
"step": 103700
},
{
"loss": 0.0989,
"grad_norm": 10.789291381835938,
"learning_rate": 4.422440127322146e-06,
"epoch": 6.321979160319298,
"step": 103750
},
{
"loss": 0.0918,
"grad_norm": 1.7923396825790405,
"learning_rate": 4.414422359949649e-06,
"epoch": 6.32502589726403,
"step": 103800
},
{
"loss": 0.0808,
"grad_norm": 0.9456502199172974,
"learning_rate": 4.406404592577152e-06,
"epoch": 6.3280726342087625,
"step": 103850
},
{
"loss": 0.1321,
"grad_norm": 0.048269350081682205,
"learning_rate": 4.398386825204654e-06,
"epoch": 6.331119371153495,
"step": 103900
},
{
"loss": 0.0924,
"grad_norm": 1.43343985080719,
"learning_rate": 4.390369057832156e-06,
"epoch": 6.334166108098227,
"step": 103950
},
{
"loss": 0.0761,
"grad_norm": 2.768300771713257,
"learning_rate": 4.382351290459658e-06,
"epoch": 6.337212845042959,
"step": 104000
},
{
"loss": 0.0753,
"grad_norm": 0.025744518265128136,
"learning_rate": 4.374333523087162e-06,
"epoch": 6.3402595819876915,
"step": 104050
},
{
"loss": 0.0846,
"grad_norm": 0.02048567309975624,
"learning_rate": 4.366315755714664e-06,
"epoch": 6.343306318932424,
"step": 104100
},
{
"loss": 0.107,
"grad_norm": 17.53726577758789,
"learning_rate": 4.358297988342167e-06,
"epoch": 6.346353055877156,
"step": 104150
},
{
"loss": 0.1224,
"grad_norm": 12.585123062133789,
"learning_rate": 4.350280220969669e-06,
"epoch": 6.349399792821888,
"step": 104200
},
{
"loss": 0.1133,
"grad_norm": 0.062179047614336014,
"learning_rate": 4.342262453597172e-06,
"epoch": 6.35244652976662,
"step": 104250
},
{
"loss": 0.0866,
"grad_norm": 0.5241016745567322,
"learning_rate": 4.334244686224674e-06,
"epoch": 6.355493266711352,
"step": 104300
},
{
"loss": 0.1009,
"grad_norm": 4.605782508850098,
"learning_rate": 4.326226918852177e-06,
"epoch": 6.358540003656084,
"step": 104350
},
{
"loss": 0.0955,
"grad_norm": 11.981565475463867,
"learning_rate": 4.318209151479679e-06,
"epoch": 6.361586740600816,
"step": 104400
},
{
"loss": 0.1181,
"grad_norm": 13.790233612060547,
"learning_rate": 4.310191384107182e-06,
"epoch": 6.364633477545548,
"step": 104450
},
{
"loss": 0.1315,
"grad_norm": 1.233720302581787,
"learning_rate": 4.302173616734684e-06,
"epoch": 6.367680214490281,
"step": 104500
},
{
"loss": 0.1016,
"grad_norm": 10.467735290527344,
"learning_rate": 4.294155849362187e-06,
"epoch": 6.370726951435013,
"step": 104550
},
{
"loss": 0.1344,
"grad_norm": 7.432904243469238,
"learning_rate": 4.28613808198969e-06,
"epoch": 6.373773688379745,
"step": 104600
},
{
"loss": 0.0778,
"grad_norm": 1.0015393495559692,
"learning_rate": 4.278120314617192e-06,
"epoch": 6.376820425324477,
"step": 104650
},
{
"loss": 0.1176,
"grad_norm": 0.594997227191925,
"learning_rate": 4.270102547244694e-06,
"epoch": 6.37986716226921,
"step": 104700
},
{
"loss": 0.0939,
"grad_norm": 10.162424087524414,
"learning_rate": 4.262084779872197e-06,
"epoch": 6.382913899213942,
"step": 104750
},
{
"loss": 0.1505,
"grad_norm": 5.383203029632568,
"learning_rate": 4.254067012499699e-06,
"epoch": 6.385960636158674,
"step": 104800
},
{
"loss": 0.1041,
"grad_norm": 3.4818716049194336,
"learning_rate": 4.246049245127203e-06,
"epoch": 6.389007373103406,
"step": 104850
},
{
"loss": 0.0866,
"grad_norm": 7.628481864929199,
"learning_rate": 4.238031477754705e-06,
"epoch": 6.3920541100481385,
"step": 104900
},
{
"loss": 0.1151,
"grad_norm": 10.642709732055664,
"learning_rate": 4.230013710382208e-06,
"epoch": 6.395100846992871,
"step": 104950
},
{
"loss": 0.083,
"grad_norm": 0.22132349014282227,
"learning_rate": 4.221995943009709e-06,
"epoch": 6.398147583937603,
"step": 105000
},
{
"loss": 0.092,
"grad_norm": 3.733302116394043,
"learning_rate": 4.213978175637212e-06,
"epoch": 6.401194320882335,
"step": 105050
},
{
"loss": 0.1102,
"grad_norm": 24.036836624145508,
"learning_rate": 4.205960408264715e-06,
"epoch": 6.404241057827067,
"step": 105100
},
{
"loss": 0.1229,
"grad_norm": 0.004999217577278614,
"learning_rate": 4.197942640892218e-06,
"epoch": 6.4072877947718,
"step": 105150
},
{
"loss": 0.1064,
"grad_norm": 10.873881340026855,
"learning_rate": 4.18992487351972e-06,
"epoch": 6.410334531716532,
"step": 105200
},
{
"loss": 0.1174,
"grad_norm": 0.04077599570155144,
"learning_rate": 4.181907106147223e-06,
"epoch": 6.413381268661264,
"step": 105250
},
{
"loss": 0.1003,
"grad_norm": 0.22582179307937622,
"learning_rate": 4.173889338774725e-06,
"epoch": 6.416428005605996,
"step": 105300
},
{
"loss": 0.1105,
"grad_norm": 6.880581378936768,
"learning_rate": 4.165871571402228e-06,
"epoch": 6.419474742550729,
"step": 105350
},
{
"loss": 0.0868,
"grad_norm": 0.23555010557174683,
"learning_rate": 4.15785380402973e-06,
"epoch": 6.42252147949546,
"step": 105400
},
{
"loss": 0.1433,
"grad_norm": 0.16392719745635986,
"learning_rate": 4.149836036657233e-06,
"epoch": 6.425568216440192,
"step": 105450
},
{
"loss": 0.1106,
"grad_norm": 2.4062070846557617,
"learning_rate": 4.141818269284735e-06,
"epoch": 6.428614953384924,
"step": 105500
},
{
"loss": 0.1382,
"grad_norm": 15.559743881225586,
"learning_rate": 4.133800501912238e-06,
"epoch": 6.431661690329657,
"step": 105550
},
{
"loss": 0.1154,
"grad_norm": 0.07285085320472717,
"learning_rate": 4.12578273453974e-06,
"epoch": 6.434708427274389,
"step": 105600
},
{
"loss": 0.0987,
"grad_norm": 5.930854320526123,
"learning_rate": 4.117925322514693e-06,
"epoch": 6.437755164219121,
"step": 105650
},
{
"loss": 0.1189,
"grad_norm": 3.026083469390869,
"learning_rate": 4.1099075551421955e-06,
"epoch": 6.440801901163853,
"step": 105700
},
{
"loss": 0.1049,
"grad_norm": 0.005459704902023077,
"learning_rate": 4.101889787769698e-06,
"epoch": 6.4438486381085855,
"step": 105750
},
{
"loss": 0.12,
"grad_norm": 0.03457879647612572,
"learning_rate": 4.0938720203972005e-06,
"epoch": 6.446895375053318,
"step": 105800
},
{
"loss": 0.1272,
"grad_norm": 0.03050217032432556,
"learning_rate": 4.085854253024703e-06,
"epoch": 6.44994211199805,
"step": 105850
},
{
"loss": 0.116,
"grad_norm": 0.045939210802316666,
"learning_rate": 4.0778364856522055e-06,
"epoch": 6.452988848942782,
"step": 105900
},
{
"loss": 0.0956,
"grad_norm": 7.571576118469238,
"learning_rate": 4.069818718279708e-06,
"epoch": 6.4560355858875145,
"step": 105950
},
{
"loss": 0.1569,
"grad_norm": 6.989561557769775,
"learning_rate": 4.061800950907211e-06,
"epoch": 6.459082322832247,
"step": 106000
},
{
"loss": 0.0632,
"grad_norm": 8.070648193359375,
"learning_rate": 4.053783183534713e-06,
"epoch": 6.462129059776979,
"step": 106050
},
{
"loss": 0.0763,
"grad_norm": 6.90822696685791,
"learning_rate": 4.0457654161622155e-06,
"epoch": 6.465175796721711,
"step": 106100
},
{
"loss": 0.1331,
"grad_norm": 0.9259748458862305,
"learning_rate": 4.037747648789718e-06,
"epoch": 6.468222533666443,
"step": 106150
},
{
"loss": 0.1013,
"grad_norm": 5.9916157722473145,
"learning_rate": 4.0297298814172205e-06,
"epoch": 6.471269270611176,
"step": 106200
},
{
"loss": 0.1397,
"grad_norm": 23.459564208984375,
"learning_rate": 4.021712114044724e-06,
"epoch": 6.474316007555908,
"step": 106250
},
{
"loss": 0.1071,
"grad_norm": 6.711719989776611,
"learning_rate": 4.013694346672226e-06,
"epoch": 6.47736274450064,
"step": 106300
},
{
"loss": 0.0908,
"grad_norm": 7.34595251083374,
"learning_rate": 4.005676579299729e-06,
"epoch": 6.480409481445372,
"step": 106350
},
{
"loss": 0.1171,
"grad_norm": 0.6432413458824158,
"learning_rate": 3.9976588119272305e-06,
"epoch": 6.4834562183901046,
"step": 106400
},
{
"loss": 0.1454,
"grad_norm": 0.2450372874736786,
"learning_rate": 3.989641044554734e-06,
"epoch": 6.486502955334837,
"step": 106450
},
{
"loss": 0.1239,
"grad_norm": 0.14580930769443512,
"learning_rate": 3.981623277182236e-06,
"epoch": 6.489549692279569,
"step": 106500
},
{
"loss": 0.1514,
"grad_norm": 0.045400507748126984,
"learning_rate": 3.973605509809739e-06,
"epoch": 6.4925964292243,
"step": 106550
},
{
"loss": 0.137,
"grad_norm": 7.579133033752441,
"learning_rate": 3.965587742437241e-06,
"epoch": 6.495643166169033,
"step": 106600
},
{
"loss": 0.0852,
"grad_norm": 0.07898136973381042,
"learning_rate": 3.957569975064744e-06,
"epoch": 6.498689903113765,
"step": 106650
},
{
"loss": 0.1142,
"grad_norm": 0.10030017048120499,
"learning_rate": 3.949552207692246e-06,
"epoch": 6.501736640058497,
"step": 106700
},
{
"loss": 0.1146,
"grad_norm": 9.135345458984375,
"learning_rate": 3.941534440319749e-06,
"epoch": 6.504783377003229,
"step": 106750
},
{
"loss": 0.1218,
"grad_norm": 4.714465618133545,
"learning_rate": 3.933516672947251e-06,
"epoch": 6.5078301139479615,
"step": 106800
},
{
"loss": 0.1196,
"grad_norm": 0.9811190962791443,
"learning_rate": 3.925498905574754e-06,
"epoch": 6.510876850892694,
"step": 106850
},
{
"loss": 0.1107,
"grad_norm": 4.175441741943359,
"learning_rate": 3.917481138202256e-06,
"epoch": 6.513923587837426,
"step": 106900
},
{
"loss": 0.1343,
"grad_norm": 0.02306065708398819,
"learning_rate": 3.909463370829759e-06,
"epoch": 6.516970324782158,
"step": 106950
},
{
"loss": 0.1118,
"grad_norm": 0.09706451743841171,
"learning_rate": 3.901445603457261e-06,
"epoch": 6.52001706172689,
"step": 107000
},
{
"loss": 0.149,
"grad_norm": 1.689697504043579,
"learning_rate": 3.893427836084765e-06,
"epoch": 6.523063798671623,
"step": 107050
},
{
"loss": 0.092,
"grad_norm": 1.3913768529891968,
"learning_rate": 3.885410068712266e-06,
"epoch": 6.526110535616355,
"step": 107100
},
{
"loss": 0.1276,
"grad_norm": 10.348816871643066,
"learning_rate": 3.877392301339769e-06,
"epoch": 6.529157272561087,
"step": 107150
},
{
"loss": 0.1293,
"grad_norm": 16.106571197509766,
"learning_rate": 3.869374533967271e-06,
"epoch": 6.532204009505819,
"step": 107200
},
{
"loss": 0.1134,
"grad_norm": 0.611929714679718,
"learning_rate": 3.861356766594774e-06,
"epoch": 6.535250746450552,
"step": 107250
},
{
"loss": 0.116,
"grad_norm": 16.070056915283203,
"learning_rate": 3.853338999222277e-06,
"epoch": 6.538297483395284,
"step": 107300
},
{
"loss": 0.0934,
"grad_norm": 4.119978904724121,
"learning_rate": 3.84532123184978e-06,
"epoch": 6.541344220340016,
"step": 107350
},
{
"loss": 0.0849,
"grad_norm": 0.26587289571762085,
"learning_rate": 3.837303464477282e-06,
"epoch": 6.544390957284748,
"step": 107400
},
{
"loss": 0.0975,
"grad_norm": 0.1269865185022354,
"learning_rate": 3.829285697104784e-06,
"epoch": 6.5474376942294805,
"step": 107450
},
{
"loss": 0.0919,
"grad_norm": 3.1790404319763184,
"learning_rate": 3.821267929732287e-06,
"epoch": 6.550484431174213,
"step": 107500
},
{
"loss": 0.1357,
"grad_norm": 12.214778900146484,
"learning_rate": 3.8132501623597893e-06,
"epoch": 6.553531168118945,
"step": 107550
},
{
"loss": 0.107,
"grad_norm": 1.6918095350265503,
"learning_rate": 3.8052323949872923e-06,
"epoch": 6.556577905063676,
"step": 107600
},
{
"loss": 0.1703,
"grad_norm": 14.809549331665039,
"learning_rate": 3.7972146276147948e-06,
"epoch": 6.559624642008409,
"step": 107650
},
{
"loss": 0.0993,
"grad_norm": 11.804024696350098,
"learning_rate": 3.7891968602422973e-06,
"epoch": 6.562671378953141,
"step": 107700
},
{
"loss": 0.1044,
"grad_norm": 4.101850986480713,
"learning_rate": 3.7811790928698e-06,
"epoch": 6.565718115897873,
"step": 107750
},
{
"loss": 0.144,
"grad_norm": 7.341966152191162,
"learning_rate": 3.773161325497302e-06,
"epoch": 6.568764852842605,
"step": 107800
},
{
"loss": 0.0848,
"grad_norm": 17.462810516357422,
"learning_rate": 3.765143558124805e-06,
"epoch": 6.5718115897873375,
"step": 107850
},
{
"loss": 0.1114,
"grad_norm": 3.7244746685028076,
"learning_rate": 3.7571257907523073e-06,
"epoch": 6.57485832673207,
"step": 107900
},
{
"loss": 0.0899,
"grad_norm": 0.2401200532913208,
"learning_rate": 3.74910802337981e-06,
"epoch": 6.577905063676802,
"step": 107950
},
{
"loss": 0.0985,
"grad_norm": 16.658544540405273,
"learning_rate": 3.7410902560073127e-06,
"epoch": 6.580951800621534,
"step": 108000
},
{
"loss": 0.1154,
"grad_norm": 0.05522345378994942,
"learning_rate": 3.7330724886348152e-06,
"epoch": 6.583998537566266,
"step": 108050
},
{
"loss": 0.0852,
"grad_norm": 0.020934052765369415,
"learning_rate": 3.7250547212623177e-06,
"epoch": 6.587045274510999,
"step": 108100
},
{
"loss": 0.1004,
"grad_norm": 0.48788294196128845,
"learning_rate": 3.71703695388982e-06,
"epoch": 6.590092011455731,
"step": 108150
},
{
"loss": 0.1054,
"grad_norm": 0.8684332370758057,
"learning_rate": 3.7090191865173223e-06,
"epoch": 6.593138748400463,
"step": 108200
},
{
"loss": 0.1093,
"grad_norm": 8.906766891479492,
"learning_rate": 3.7010014191448252e-06,
"epoch": 6.596185485345195,
"step": 108250
},
{
"loss": 0.1012,
"grad_norm": 4.543704032897949,
"learning_rate": 3.6929836517723277e-06,
"epoch": 6.599232222289928,
"step": 108300
},
{
"loss": 0.1518,
"grad_norm": 17.82200813293457,
"learning_rate": 3.6849658843998302e-06,
"epoch": 6.60227895923466,
"step": 108350
},
{
"loss": 0.1162,
"grad_norm": 9.57278823852539,
"learning_rate": 3.676948117027333e-06,
"epoch": 6.605325696179392,
"step": 108400
},
{
"loss": 0.0979,
"grad_norm": 0.0030305020045489073,
"learning_rate": 3.6689303496548357e-06,
"epoch": 6.608372433124124,
"step": 108450
},
{
"loss": 0.11,
"grad_norm": 13.671561241149902,
"learning_rate": 3.6609125822823378e-06,
"epoch": 6.6114191700688565,
"step": 108500
},
{
"loss": 0.1133,
"grad_norm": 4.691519260406494,
"learning_rate": 3.6528948149098403e-06,
"epoch": 6.614465907013589,
"step": 108550
},
{
"loss": 0.1171,
"grad_norm": 3.799377918243408,
"learning_rate": 3.6448770475373428e-06,
"epoch": 6.617512643958321,
"step": 108600
},
{
"loss": 0.1181,
"grad_norm": 6.710395336151123,
"learning_rate": 3.6368592801648457e-06,
"epoch": 6.620559380903053,
"step": 108650
},
{
"loss": 0.1339,
"grad_norm": 0.18311314284801483,
"learning_rate": 3.628841512792348e-06,
"epoch": 6.623606117847785,
"step": 108700
},
{
"loss": 0.1256,
"grad_norm": 0.17366501688957214,
"learning_rate": 3.6208237454198507e-06,
"epoch": 6.626652854792518,
"step": 108750
},
{
"loss": 0.1225,
"grad_norm": 0.13407079875469208,
"learning_rate": 3.6128059780473536e-06,
"epoch": 6.62969959173725,
"step": 108800
},
{
"loss": 0.1427,
"grad_norm": 0.1014561727643013,
"learning_rate": 3.6047882106748557e-06,
"epoch": 6.632746328681981,
"step": 108850
},
{
"loss": 0.1242,
"grad_norm": 0.6458027362823486,
"learning_rate": 3.596770443302358e-06,
"epoch": 6.6357930656267134,
"step": 108900
},
{
"loss": 0.1219,
"grad_norm": 14.363363265991211,
"learning_rate": 3.5887526759298607e-06,
"epoch": 6.638839802571446,
"step": 108950
},
{
"loss": 0.1487,
"grad_norm": 0.0431794635951519,
"learning_rate": 3.5807349085573632e-06,
"epoch": 6.641886539516178,
"step": 109000
},
{
"loss": 0.0608,
"grad_norm": 0.0029093355406075716,
"learning_rate": 3.572877496532316e-06,
"epoch": 6.64493327646091,
"step": 109050
},
{
"loss": 0.1177,
"grad_norm": 3.1443655490875244,
"learning_rate": 3.5648597291598185e-06,
"epoch": 6.647980013405642,
"step": 109100
},
{
"loss": 0.0859,
"grad_norm": 0.26113003492355347,
"learning_rate": 3.5568419617873214e-06,
"epoch": 6.651026750350375,
"step": 109150
},
{
"loss": 0.1249,
"grad_norm": 0.17804184556007385,
"learning_rate": 3.5488241944148235e-06,
"epoch": 6.654073487295107,
"step": 109200
},
{
"loss": 0.0862,
"grad_norm": 11.204771995544434,
"learning_rate": 3.540806427042326e-06,
"epoch": 6.657120224239839,
"step": 109250
},
{
"loss": 0.1105,
"grad_norm": 8.531720161437988,
"learning_rate": 3.5327886596698285e-06,
"epoch": 6.660166961184571,
"step": 109300
},
{
"loss": 0.1009,
"grad_norm": 1.794236660003662,
"learning_rate": 3.524770892297331e-06,
"epoch": 6.6632136981293035,
"step": 109350
},
{
"loss": 0.0652,
"grad_norm": 0.025617562234401703,
"learning_rate": 3.516753124924834e-06,
"epoch": 6.666260435074036,
"step": 109400
},
{
"loss": 0.1156,
"grad_norm": 0.4970704913139343,
"learning_rate": 3.5087353575523364e-06,
"epoch": 6.669307172018768,
"step": 109450
},
{
"loss": 0.09,
"grad_norm": 5.861164093017578,
"learning_rate": 3.500717590179839e-06,
"epoch": 6.6723539089635,
"step": 109500
},
{
"loss": 0.114,
"grad_norm": 4.488530158996582,
"learning_rate": 3.492699822807341e-06,
"epoch": 6.6754006459082325,
"step": 109550
},
{
"loss": 0.102,
"grad_norm": 20.01913833618164,
"learning_rate": 3.484682055434844e-06,
"epoch": 6.678447382852965,
"step": 109600
},
{
"loss": 0.103,
"grad_norm": 0.7331236004829407,
"learning_rate": 3.4766642880623464e-06,
"epoch": 6.681494119797697,
"step": 109650
},
{
"loss": 0.1398,
"grad_norm": 0.034562163054943085,
"learning_rate": 3.468646520689849e-06,
"epoch": 6.684540856742429,
"step": 109700
},
{
"loss": 0.0888,
"grad_norm": 4.30727481842041,
"learning_rate": 3.4606287533173514e-06,
"epoch": 6.687587593687161,
"step": 109750
},
{
"loss": 0.1302,
"grad_norm": 0.015355073846876621,
"learning_rate": 3.4526109859448544e-06,
"epoch": 6.690634330631894,
"step": 109800
},
{
"loss": 0.1134,
"grad_norm": 4.289348125457764,
"learning_rate": 3.444593218572357e-06,
"epoch": 6.693681067576625,
"step": 109850
},
{
"loss": 0.1393,
"grad_norm": 28.9506893157959,
"learning_rate": 3.436575451199859e-06,
"epoch": 6.696727804521357,
"step": 109900
},
{
"loss": 0.1335,
"grad_norm": 7.488936901092529,
"learning_rate": 3.4285576838273615e-06,
"epoch": 6.699774541466089,
"step": 109950
},
{
"loss": 0.1066,
"grad_norm": 0.4364040791988373,
"learning_rate": 3.420539916454864e-06,
"epoch": 6.702821278410822,
"step": 110000
},
{
"loss": 0.1086,
"grad_norm": 11.219388961791992,
"learning_rate": 3.412522149082367e-06,
"epoch": 6.705868015355554,
"step": 110050
},
{
"loss": 0.0765,
"grad_norm": 1.8903021812438965,
"learning_rate": 3.4045043817098694e-06,
"epoch": 6.708914752300286,
"step": 110100
},
{
"loss": 0.1261,
"grad_norm": 6.067765712738037,
"learning_rate": 3.396486614337372e-06,
"epoch": 6.711961489245018,
"step": 110150
},
{
"loss": 0.1001,
"grad_norm": 0.2357344776391983,
"learning_rate": 3.388468846964875e-06,
"epoch": 6.715008226189751,
"step": 110200
},
{
"loss": 0.0893,
"grad_norm": 3.183743715286255,
"learning_rate": 3.380451079592377e-06,
"epoch": 6.718054963134483,
"step": 110250
},
{
"loss": 0.1639,
"grad_norm": 18.015254974365234,
"learning_rate": 3.3724333122198794e-06,
"epoch": 6.721101700079215,
"step": 110300
},
{
"loss": 0.1059,
"grad_norm": 29.258014678955078,
"learning_rate": 3.364415544847382e-06,
"epoch": 6.724148437023947,
"step": 110350
},
{
"loss": 0.1041,
"grad_norm": 16.73503875732422,
"learning_rate": 3.3563977774748844e-06,
"epoch": 6.7271951739686795,
"step": 110400
},
{
"loss": 0.1175,
"grad_norm": 5.931246280670166,
"learning_rate": 3.3483800101023873e-06,
"epoch": 6.730241910913412,
"step": 110450
},
{
"loss": 0.0955,
"grad_norm": 6.298430919647217,
"learning_rate": 3.34036224272989e-06,
"epoch": 6.733288647858144,
"step": 110500
},
{
"loss": 0.103,
"grad_norm": 0.09814344346523285,
"learning_rate": 3.3323444753573923e-06,
"epoch": 6.736335384802876,
"step": 110550
},
{
"loss": 0.0782,
"grad_norm": 0.22854644060134888,
"learning_rate": 3.3243267079848944e-06,
"epoch": 6.739382121747608,
"step": 110600
},
{
"loss": 0.1117,
"grad_norm": 0.010507500730454922,
"learning_rate": 3.3163089406123974e-06,
"epoch": 6.742428858692341,
"step": 110650
},
{
"loss": 0.1012,
"grad_norm": 0.0423794649541378,
"learning_rate": 3.3082911732399e-06,
"epoch": 6.745475595637073,
"step": 110700
},
{
"loss": 0.0879,
"grad_norm": 5.643665790557861,
"learning_rate": 3.3002734058674024e-06,
"epoch": 6.748522332581805,
"step": 110750
},
{
"loss": 0.0976,
"grad_norm": 0.515412449836731,
"learning_rate": 3.292255638494905e-06,
"epoch": 6.751569069526537,
"step": 110800
},
{
"loss": 0.126,
"grad_norm": 12.693973541259766,
"learning_rate": 3.2842378711224078e-06,
"epoch": 6.75461580647127,
"step": 110850
},
{
"loss": 0.0967,
"grad_norm": 1.1584398746490479,
"learning_rate": 3.2762201037499103e-06,
"epoch": 6.757662543416002,
"step": 110900
},
{
"loss": 0.1043,
"grad_norm": 4.0164103507995605,
"learning_rate": 3.2682023363774124e-06,
"epoch": 6.760709280360734,
"step": 110950
},
{
"loss": 0.0779,
"grad_norm": 5.281560897827148,
"learning_rate": 3.260184569004915e-06,
"epoch": 6.763756017305466,
"step": 111000
},
{
"loss": 0.1353,
"grad_norm": 0.09412240236997604,
"learning_rate": 3.252166801632418e-06,
"epoch": 6.7668027542501985,
"step": 111050
},
{
"loss": 0.1414,
"grad_norm": 5.122351169586182,
"learning_rate": 3.2441490342599203e-06,
"epoch": 6.76984949119493,
"step": 111100
},
{
"loss": 0.1268,
"grad_norm": 9.530380249023438,
"learning_rate": 3.2362916222348726e-06,
"epoch": 6.772896228139662,
"step": 111150
},
{
"loss": 0.1376,
"grad_norm": 1.5909652709960938,
"learning_rate": 3.2282738548623756e-06,
"epoch": 6.775942965084394,
"step": 111200
},
{
"loss": 0.081,
"grad_norm": 7.929418087005615,
"learning_rate": 3.220256087489878e-06,
"epoch": 6.7789897020291265,
"step": 111250
},
{
"loss": 0.1282,
"grad_norm": 0.7203691601753235,
"learning_rate": 3.21223832011738e-06,
"epoch": 6.782036438973859,
"step": 111300
},
{
"loss": 0.0965,
"grad_norm": 5.791773319244385,
"learning_rate": 3.2042205527448827e-06,
"epoch": 6.785083175918591,
"step": 111350
},
{
"loss": 0.122,
"grad_norm": 1.6321055889129639,
"learning_rate": 3.1962027853723856e-06,
"epoch": 6.788129912863323,
"step": 111400
},
{
"loss": 0.1641,
"grad_norm": 1.0251235961914062,
"learning_rate": 3.188185017999888e-06,
"epoch": 6.7911766498080555,
"step": 111450
},
{
"loss": 0.102,
"grad_norm": 3.3390939235687256,
"learning_rate": 3.1801672506273906e-06,
"epoch": 6.794223386752788,
"step": 111500
},
{
"loss": 0.123,
"grad_norm": 9.066089630126953,
"learning_rate": 3.172149483254893e-06,
"epoch": 6.79727012369752,
"step": 111550
},
{
"loss": 0.1528,
"grad_norm": 5.868232727050781,
"learning_rate": 3.164131715882396e-06,
"epoch": 6.800316860642252,
"step": 111600
},
{
"loss": 0.1094,
"grad_norm": 2.686887502670288,
"learning_rate": 3.156113948509898e-06,
"epoch": 6.803363597586984,
"step": 111650
},
{
"loss": 0.1194,
"grad_norm": 8.624323844909668,
"learning_rate": 3.1480961811374006e-06,
"epoch": 6.806410334531717,
"step": 111700
},
{
"loss": 0.1147,
"grad_norm": 3.5322718620300293,
"learning_rate": 3.140078413764903e-06,
"epoch": 6.809457071476449,
"step": 111750
},
{
"loss": 0.121,
"grad_norm": 6.578117370605469,
"learning_rate": 3.1320606463924056e-06,
"epoch": 6.812503808421181,
"step": 111800
},
{
"loss": 0.1158,
"grad_norm": 0.3103256821632385,
"learning_rate": 3.1240428790199085e-06,
"epoch": 6.815550545365913,
"step": 111850
},
{
"loss": 0.0887,
"grad_norm": 1.9541118144989014,
"learning_rate": 3.116025111647411e-06,
"epoch": 6.8185972823106455,
"step": 111900
},
{
"loss": 0.1036,
"grad_norm": 0.16623610258102417,
"learning_rate": 3.1080073442749135e-06,
"epoch": 6.821644019255378,
"step": 111950
},
{
"loss": 0.1013,
"grad_norm": 0.26442739367485046,
"learning_rate": 3.0999895769024156e-06,
"epoch": 6.82469075620011,
"step": 112000
},
{
"loss": 0.085,
"grad_norm": 0.28916820883750916,
"learning_rate": 3.0919718095299185e-06,
"epoch": 6.827737493144842,
"step": 112050
},
{
"loss": 0.1643,
"grad_norm": 4.109094619750977,
"learning_rate": 3.083954042157421e-06,
"epoch": 6.830784230089574,
"step": 112100
},
{
"loss": 0.1341,
"grad_norm": 0.1970689445734024,
"learning_rate": 3.0759362747849236e-06,
"epoch": 6.833830967034306,
"step": 112150
},
{
"loss": 0.1117,
"grad_norm": 5.4792022705078125,
"learning_rate": 3.067918507412426e-06,
"epoch": 6.836877703979038,
"step": 112200
},
{
"loss": 0.1271,
"grad_norm": 0.012952005490660667,
"learning_rate": 3.059900740039929e-06,
"epoch": 6.83992444092377,
"step": 112250
},
{
"loss": 0.0975,
"grad_norm": 19.012514114379883,
"learning_rate": 3.0518829726674315e-06,
"epoch": 6.8429711778685025,
"step": 112300
},
{
"loss": 0.1146,
"grad_norm": 6.967922210693359,
"learning_rate": 3.0438652052949336e-06,
"epoch": 6.846017914813235,
"step": 112350
},
{
"loss": 0.138,
"grad_norm": 2.224149465560913,
"learning_rate": 3.035847437922436e-06,
"epoch": 6.849064651757967,
"step": 112400
},
{
"loss": 0.0817,
"grad_norm": 1.1596438884735107,
"learning_rate": 3.027829670549939e-06,
"epoch": 6.852111388702699,
"step": 112450
},
{
"loss": 0.0959,
"grad_norm": 5.981630325317383,
"learning_rate": 3.0198119031774415e-06,
"epoch": 6.855158125647431,
"step": 112500
},
{
"loss": 0.1131,
"grad_norm": 12.198943138122559,
"learning_rate": 3.011794135804944e-06,
"epoch": 6.858204862592164,
"step": 112550
},
{
"loss": 0.1037,
"grad_norm": 10.507756233215332,
"learning_rate": 3.0037763684324465e-06,
"epoch": 6.861251599536896,
"step": 112600
},
{
"loss": 0.0686,
"grad_norm": 18.165424346923828,
"learning_rate": 2.9957586010599494e-06,
"epoch": 6.864298336481628,
"step": 112650
},
{
"loss": 0.1278,
"grad_norm": 0.5055847764015198,
"learning_rate": 2.9877408336874515e-06,
"epoch": 6.86734507342636,
"step": 112700
},
{
"loss": 0.0677,
"grad_norm": 13.600053787231445,
"learning_rate": 2.979723066314954e-06,
"epoch": 6.870391810371093,
"step": 112750
},
{
"loss": 0.1277,
"grad_norm": 10.777318000793457,
"learning_rate": 2.9717052989424565e-06,
"epoch": 6.873438547315825,
"step": 112800
},
{
"loss": 0.1393,
"grad_norm": 5.368490695953369,
"learning_rate": 2.9636875315699594e-06,
"epoch": 6.876485284260557,
"step": 112850
},
{
"loss": 0.113,
"grad_norm": 0.0565558522939682,
"learning_rate": 2.955669764197462e-06,
"epoch": 6.879532021205289,
"step": 112900
},
{
"loss": 0.0995,
"grad_norm": 0.6518847346305847,
"learning_rate": 2.9476519968249645e-06,
"epoch": 6.8825787581500215,
"step": 112950
},
{
"loss": 0.1249,
"grad_norm": 12.90188217163086,
"learning_rate": 2.939634229452467e-06,
"epoch": 6.885625495094754,
"step": 113000
},
{
"loss": 0.1197,
"grad_norm": 7.218382835388184,
"learning_rate": 2.931616462079969e-06,
"epoch": 6.888672232039486,
"step": 113050
},
{
"loss": 0.0823,
"grad_norm": 2.199795722961426,
"learning_rate": 2.923598694707472e-06,
"epoch": 6.891718968984218,
"step": 113100
},
{
"loss": 0.1036,
"grad_norm": 16.90618324279785,
"learning_rate": 2.9155809273349745e-06,
"epoch": 6.89476570592895,
"step": 113150
},
{
"loss": 0.0977,
"grad_norm": 1.5636283159255981,
"learning_rate": 2.907563159962477e-06,
"epoch": 6.897812442873683,
"step": 113200
},
{
"loss": 0.0994,
"grad_norm": 0.15139245986938477,
"learning_rate": 2.89954539258998e-06,
"epoch": 6.900859179818415,
"step": 113250
},
{
"loss": 0.1482,
"grad_norm": 13.415343284606934,
"learning_rate": 2.8915276252174824e-06,
"epoch": 6.903905916763147,
"step": 113300
},
{
"loss": 0.1089,
"grad_norm": 6.342071056365967,
"learning_rate": 2.883509857844985e-06,
"epoch": 6.9069526537078785,
"step": 113350
},
{
"loss": 0.1077,
"grad_norm": 0.2819023132324219,
"learning_rate": 2.875492090472487e-06,
"epoch": 6.909999390652611,
"step": 113400
},
{
"loss": 0.1433,
"grad_norm": 5.393600940704346,
"learning_rate": 2.8674743230999895e-06,
"epoch": 6.913046127597343,
"step": 113450
},
{
"loss": 0.0898,
"grad_norm": 1.8563032150268555,
"learning_rate": 2.8594565557274924e-06,
"epoch": 6.916092864542075,
"step": 113500
},
{
"loss": 0.0955,
"grad_norm": 6.525058269500732,
"learning_rate": 2.851438788354995e-06,
"epoch": 6.919139601486807,
"step": 113550
},
{
"loss": 0.1096,
"grad_norm": 0.01581917703151703,
"learning_rate": 2.8434210209824974e-06,
"epoch": 6.92218633843154,
"step": 113600
},
{
"loss": 0.1095,
"grad_norm": 5.255688190460205,
"learning_rate": 2.83540325361e-06,
"epoch": 6.925233075376272,
"step": 113650
},
{
"loss": 0.126,
"grad_norm": 18.47319984436035,
"learning_rate": 2.827385486237503e-06,
"epoch": 6.928279812321004,
"step": 113700
},
{
"loss": 0.1017,
"grad_norm": 7.02748966217041,
"learning_rate": 2.819367718865005e-06,
"epoch": 6.931326549265736,
"step": 113750
},
{
"loss": 0.0848,
"grad_norm": 17.112136840820312,
"learning_rate": 2.8113499514925074e-06,
"epoch": 6.934373286210469,
"step": 113800
},
{
"loss": 0.1088,
"grad_norm": 5.9006853103637695,
"learning_rate": 2.80333218412001e-06,
"epoch": 6.937420023155201,
"step": 113850
},
{
"loss": 0.0955,
"grad_norm": 1.1214410066604614,
"learning_rate": 2.795314416747513e-06,
"epoch": 6.940466760099933,
"step": 113900
},
{
"loss": 0.121,
"grad_norm": 0.015278889797627926,
"learning_rate": 2.7872966493750154e-06,
"epoch": 6.943513497044665,
"step": 113950
},
{
"loss": 0.1224,
"grad_norm": 13.605744361877441,
"learning_rate": 2.779278882002518e-06,
"epoch": 6.9465602339893975,
"step": 114000
},
{
"loss": 0.1254,
"grad_norm": 5.2504963874816895,
"learning_rate": 2.7712611146300204e-06,
"epoch": 6.94960697093413,
"step": 114050
},
{
"loss": 0.0955,
"grad_norm": 5.697160243988037,
"learning_rate": 2.7632433472575225e-06,
"epoch": 6.952653707878862,
"step": 114100
},
{
"loss": 0.1064,
"grad_norm": 6.439358234405518,
"learning_rate": 2.7552255798850254e-06,
"epoch": 6.955700444823594,
"step": 114150
},
{
"loss": 0.1247,
"grad_norm": 5.867162704467773,
"learning_rate": 2.747207812512528e-06,
"epoch": 6.958747181768326,
"step": 114200
},
{
"loss": 0.0973,
"grad_norm": 1.2504898309707642,
"learning_rate": 2.7391900451400304e-06,
"epoch": 6.961793918713059,
"step": 114250
},
{
"loss": 0.0859,
"grad_norm": 10.872735023498535,
"learning_rate": 2.7311722777675333e-06,
"epoch": 6.964840655657791,
"step": 114300
},
{
"loss": 0.0865,
"grad_norm": 20.190185546875,
"learning_rate": 2.723154510395036e-06,
"epoch": 6.967887392602522,
"step": 114350
},
{
"loss": 0.1066,
"grad_norm": 7.0147294998168945,
"learning_rate": 2.7151367430225383e-06,
"epoch": 6.970934129547254,
"step": 114400
},
{
"loss": 0.1018,
"grad_norm": 4.0780029296875,
"learning_rate": 2.7071189756500404e-06,
"epoch": 6.973980866491987,
"step": 114450
},
{
"loss": 0.1483,
"grad_norm": 3.171633243560791,
"learning_rate": 2.699261563624993e-06,
"epoch": 6.977027603436719,
"step": 114500
},
{
"loss": 0.1165,
"grad_norm": 6.388818740844727,
"learning_rate": 2.6912437962524957e-06,
"epoch": 6.980074340381451,
"step": 114550
},
{
"loss": 0.0699,
"grad_norm": 8.529497146606445,
"learning_rate": 2.683226028879998e-06,
"epoch": 6.983121077326183,
"step": 114600
},
{
"loss": 0.0656,
"grad_norm": 6.983097553253174,
"learning_rate": 2.675208261507501e-06,
"epoch": 6.986167814270916,
"step": 114650
},
{
"loss": 0.1138,
"grad_norm": 19.808326721191406,
"learning_rate": 2.6671904941350036e-06,
"epoch": 6.989214551215648,
"step": 114700
},
{
"loss": 0.0801,
"grad_norm": 1.2024364471435547,
"learning_rate": 2.659172726762506e-06,
"epoch": 6.99226128816038,
"step": 114750
},
{
"loss": 0.1256,
"grad_norm": 6.075553894042969,
"learning_rate": 2.651154959390008e-06,
"epoch": 6.995308025105112,
"step": 114800
},
{
"loss": 0.0723,
"grad_norm": 5.352075576782227,
"learning_rate": 2.6431371920175107e-06,
"epoch": 6.9983547620498445,
"step": 114850
},
{
"eval_loss": 0.11268351227045059,
"eval_accuracy": 0.9721679520137104,
"eval_f1_macro": 0.9718053472659474,
"eval_top5_accuracy": 0.9967437874892888,
"eval_runtime": 219.6867,
"eval_samples_per_second": 132.803,
"eval_steps_per_second": 8.303,
"epoch": 7.0,
"step": 114877
},
{
"loss": 0.072,
"grad_norm": 4.461543083190918,
"learning_rate": 2.6351194246450136e-06,
"epoch": 7.001401498994577,
"step": 114900
},
{
"loss": 0.1095,
"grad_norm": 10.202938079833984,
"learning_rate": 2.627101657272516e-06,
"epoch": 7.004448235939309,
"step": 114950
},
{
"loss": 0.1286,
"grad_norm": 15.761112213134766,
"learning_rate": 2.6190838899000186e-06,
"epoch": 7.007494972884041,
"step": 115000
},
{
"loss": 0.1014,
"grad_norm": 0.43326061964035034,
"learning_rate": 2.6110661225275215e-06,
"epoch": 7.0105417098287734,
"step": 115050
},
{
"loss": 0.0825,
"grad_norm": 0.09228600561618805,
"learning_rate": 2.603048355155024e-06,
"epoch": 7.013588446773506,
"step": 115100
},
{
"loss": 0.0914,
"grad_norm": 0.012318127788603306,
"learning_rate": 2.595030587782526e-06,
"epoch": 7.016635183718238,
"step": 115150
},
{
"loss": 0.0809,
"grad_norm": 5.912267208099365,
"learning_rate": 2.5870128204100286e-06,
"epoch": 7.01968192066297,
"step": 115200
},
{
"loss": 0.085,
"grad_norm": 0.059856683015823364,
"learning_rate": 2.578995053037531e-06,
"epoch": 7.022728657607702,
"step": 115250
},
{
"loss": 0.107,
"grad_norm": 0.1699170172214508,
"learning_rate": 2.570977285665034e-06,
"epoch": 7.025775394552435,
"step": 115300
},
{
"loss": 0.0769,
"grad_norm": 0.06988510489463806,
"learning_rate": 2.5629595182925366e-06,
"epoch": 7.028822131497167,
"step": 115350
},
{
"loss": 0.0918,
"grad_norm": 11.297406196594238,
"learning_rate": 2.554941750920039e-06,
"epoch": 7.031868868441899,
"step": 115400
},
{
"loss": 0.0739,
"grad_norm": 0.9245764017105103,
"learning_rate": 2.5469239835475416e-06,
"epoch": 7.034915605386631,
"step": 115450
},
{
"loss": 0.1439,
"grad_norm": 0.21616198122501373,
"learning_rate": 2.5389062161750436e-06,
"epoch": 7.0379623423313635,
"step": 115500
},
{
"loss": 0.0854,
"grad_norm": 11.348570823669434,
"learning_rate": 2.5308884488025466e-06,
"epoch": 7.041009079276095,
"step": 115550
},
{
"loss": 0.0721,
"grad_norm": 1.2119930982589722,
"learning_rate": 2.522870681430049e-06,
"epoch": 7.044055816220827,
"step": 115600
},
{
"loss": 0.1162,
"grad_norm": 6.2853803634643555,
"learning_rate": 2.5148529140575516e-06,
"epoch": 7.047102553165559,
"step": 115650
},
{
"loss": 0.1242,
"grad_norm": 5.428760528564453,
"learning_rate": 2.5068351466850545e-06,
"epoch": 7.050149290110292,
"step": 115700
},
{
"loss": 0.0728,
"grad_norm": 0.5153530836105347,
"learning_rate": 2.498817379312557e-06,
"epoch": 7.053196027055024,
"step": 115750
},
{
"loss": 0.0914,
"grad_norm": 5.024087905883789,
"learning_rate": 2.4907996119400595e-06,
"epoch": 7.056242763999756,
"step": 115800
},
{
"loss": 0.1036,
"grad_norm": 0.4718085825443268,
"learning_rate": 2.482781844567562e-06,
"epoch": 7.059289500944488,
"step": 115850
},
{
"loss": 0.0996,
"grad_norm": 4.929214000701904,
"learning_rate": 2.4747640771950645e-06,
"epoch": 7.0623362378892205,
"step": 115900
},
{
"loss": 0.0847,
"grad_norm": 10.664554595947266,
"learning_rate": 2.466746309822567e-06,
"epoch": 7.065382974833953,
"step": 115950
},
{
"loss": 0.1028,
"grad_norm": 0.3477446436882019,
"learning_rate": 2.4587285424500695e-06,
"epoch": 7.068429711778685,
"step": 116000
},
{
"loss": 0.0744,
"grad_norm": 0.11017444729804993,
"learning_rate": 2.450710775077572e-06,
"epoch": 7.071476448723417,
"step": 116050
},
{
"loss": 0.1042,
"grad_norm": 0.03889969736337662,
"learning_rate": 2.442693007705075e-06,
"epoch": 7.074523185668149,
"step": 116100
},
{
"loss": 0.0948,
"grad_norm": 2.102764368057251,
"learning_rate": 2.434675240332577e-06,
"epoch": 7.077569922612882,
"step": 116150
},
{
"loss": 0.0897,
"grad_norm": 0.13599608838558197,
"learning_rate": 2.42665747296008e-06,
"epoch": 7.080616659557614,
"step": 116200
},
{
"loss": 0.0637,
"grad_norm": 1.999712586402893,
"learning_rate": 2.4186397055875825e-06,
"epoch": 7.083663396502346,
"step": 116250
},
{
"loss": 0.0836,
"grad_norm": 1.401777982711792,
"learning_rate": 2.4106219382150845e-06,
"epoch": 7.086710133447078,
"step": 116300
},
{
"loss": 0.1056,
"grad_norm": 2.5041518211364746,
"learning_rate": 2.4026041708425875e-06,
"epoch": 7.089756870391811,
"step": 116350
},
{
"loss": 0.1158,
"grad_norm": 0.009966792538762093,
"learning_rate": 2.39458640347009e-06,
"epoch": 7.092803607336543,
"step": 116400
},
{
"loss": 0.0936,
"grad_norm": 0.013414375483989716,
"learning_rate": 2.3865686360975925e-06,
"epoch": 7.095850344281275,
"step": 116450
},
{
"loss": 0.084,
"grad_norm": 4.469899654388428,
"learning_rate": 2.378550868725095e-06,
"epoch": 7.098897081226007,
"step": 116500
},
{
"loss": 0.0539,
"grad_norm": 0.03635074570775032,
"learning_rate": 2.3705331013525975e-06,
"epoch": 7.1019438181707395,
"step": 116550
},
{
"loss": 0.1125,
"grad_norm": 7.3072638511657715,
"learning_rate": 2.3625153339801e-06,
"epoch": 7.104990555115472,
"step": 116600
},
{
"loss": 0.0676,
"grad_norm": 0.010777551680803299,
"learning_rate": 2.3544975666076025e-06,
"epoch": 7.108037292060203,
"step": 116650
},
{
"loss": 0.1478,
"grad_norm": 0.08819326013326645,
"learning_rate": 2.346479799235105e-06,
"epoch": 7.111084029004935,
"step": 116700
},
{
"loss": 0.0997,
"grad_norm": 0.08724324405193329,
"learning_rate": 2.338462031862608e-06,
"epoch": 7.1141307659496675,
"step": 116750
},
{
"loss": 0.0873,
"grad_norm": 4.976078987121582,
"learning_rate": 2.3304442644901104e-06,
"epoch": 7.1171775028944,
"step": 116800
},
{
"loss": 0.1002,
"grad_norm": 8.376337051391602,
"learning_rate": 2.322426497117613e-06,
"epoch": 7.120224239839132,
"step": 116850
},
{
"loss": 0.0778,
"grad_norm": 6.594933032989502,
"learning_rate": 2.3144087297451154e-06,
"epoch": 7.123270976783864,
"step": 116900
},
{
"loss": 0.1076,
"grad_norm": 2.2587592601776123,
"learning_rate": 2.306390962372618e-06,
"epoch": 7.1263177137285965,
"step": 116950
},
{
"loss": 0.1168,
"grad_norm": 1.7942737340927124,
"learning_rate": 2.2983731950001204e-06,
"epoch": 7.129364450673329,
"step": 117000
},
{
"loss": 0.0893,
"grad_norm": 0.02802767977118492,
"learning_rate": 2.290355427627623e-06,
"epoch": 7.132411187618061,
"step": 117050
},
{
"loss": 0.1139,
"grad_norm": 12.19888687133789,
"learning_rate": 2.2824980156025757e-06,
"epoch": 7.135457924562793,
"step": 117100
},
{
"loss": 0.1271,
"grad_norm": 0.07522892206907272,
"learning_rate": 2.2744802482300778e-06,
"epoch": 7.138504661507525,
"step": 117150
},
{
"loss": 0.1439,
"grad_norm": 0.92941814661026,
"learning_rate": 2.2664624808575807e-06,
"epoch": 7.141551398452258,
"step": 117200
},
{
"loss": 0.1126,
"grad_norm": 0.2575014531612396,
"learning_rate": 2.258444713485083e-06,
"epoch": 7.14459813539699,
"step": 117250
},
{
"loss": 0.0884,
"grad_norm": 7.107641696929932,
"learning_rate": 2.2504269461125857e-06,
"epoch": 7.147644872341722,
"step": 117300
},
{
"loss": 0.1034,
"grad_norm": 0.12390744686126709,
"learning_rate": 2.2424091787400882e-06,
"epoch": 7.150691609286454,
"step": 117350
},
{
"loss": 0.0957,
"grad_norm": 1.9339579343795776,
"learning_rate": 2.2343914113675907e-06,
"epoch": 7.1537383462311865,
"step": 117400
},
{
"loss": 0.0963,
"grad_norm": 5.472762107849121,
"learning_rate": 2.2263736439950932e-06,
"epoch": 7.156785083175919,
"step": 117450
},
{
"loss": 0.0729,
"grad_norm": 0.09180988371372223,
"learning_rate": 2.2183558766225957e-06,
"epoch": 7.159831820120651,
"step": 117500
},
{
"loss": 0.0983,
"grad_norm": 0.176802396774292,
"learning_rate": 2.2103381092500982e-06,
"epoch": 7.162878557065383,
"step": 117550
},
{
"loss": 0.0728,
"grad_norm": 8.294918060302734,
"learning_rate": 2.202320341877601e-06,
"epoch": 7.1659252940101155,
"step": 117600
},
{
"loss": 0.1521,
"grad_norm": 10.783720016479492,
"learning_rate": 2.1943025745051037e-06,
"epoch": 7.168972030954848,
"step": 117650
},
{
"loss": 0.117,
"grad_norm": 21.37734603881836,
"learning_rate": 2.1862848071326057e-06,
"epoch": 7.17201876789958,
"step": 117700
},
{
"loss": 0.1485,
"grad_norm": 10.080018043518066,
"learning_rate": 2.1782670397601087e-06,
"epoch": 7.175065504844312,
"step": 117750
},
{
"loss": 0.1121,
"grad_norm": 0.4653596580028534,
"learning_rate": 2.170249272387611e-06,
"epoch": 7.178112241789044,
"step": 117800
},
{
"loss": 0.1276,
"grad_norm": 0.07687785476446152,
"learning_rate": 2.1622315050151137e-06,
"epoch": 7.181158978733776,
"step": 117850
},
{
"loss": 0.0985,
"grad_norm": 0.036559537053108215,
"learning_rate": 2.154213737642616e-06,
"epoch": 7.184205715678508,
"step": 117900
},
{
"loss": 0.083,
"grad_norm": 0.14159269630908966,
"learning_rate": 2.1461959702701187e-06,
"epoch": 7.18725245262324,
"step": 117950
},
{
"loss": 0.095,
"grad_norm": 0.09861453622579575,
"learning_rate": 2.1381782028976216e-06,
"epoch": 7.190299189567972,
"step": 118000
},
{
"loss": 0.1247,
"grad_norm": 10.685966491699219,
"learning_rate": 2.1301604355251237e-06,
"epoch": 7.193345926512705,
"step": 118050
},
{
"loss": 0.1177,
"grad_norm": 16.73500633239746,
"learning_rate": 2.122142668152626e-06,
"epoch": 7.196392663457437,
"step": 118100
},
{
"loss": 0.0774,
"grad_norm": 9.669848442077637,
"learning_rate": 2.114124900780129e-06,
"epoch": 7.199439400402169,
"step": 118150
},
{
"loss": 0.0923,
"grad_norm": 3.340395450592041,
"learning_rate": 2.1061071334076316e-06,
"epoch": 7.202486137346901,
"step": 118200
},
{
"loss": 0.1125,
"grad_norm": 11.828262329101562,
"learning_rate": 2.098089366035134e-06,
"epoch": 7.205532874291634,
"step": 118250
},
{
"loss": 0.1004,
"grad_norm": 0.07790781557559967,
"learning_rate": 2.0900715986626366e-06,
"epoch": 7.208579611236366,
"step": 118300
},
{
"loss": 0.0788,
"grad_norm": 3.4064042568206787,
"learning_rate": 2.082053831290139e-06,
"epoch": 7.211626348181098,
"step": 118350
},
{
"loss": 0.1277,
"grad_norm": 0.16170386970043182,
"learning_rate": 2.0740360639176416e-06,
"epoch": 7.21467308512583,
"step": 118400
},
{
"loss": 0.0915,
"grad_norm": 9.507930755615234,
"learning_rate": 2.066018296545144e-06,
"epoch": 7.2177198220705625,
"step": 118450
},
{
"loss": 0.1239,
"grad_norm": 13.89639663696289,
"learning_rate": 2.0580005291726466e-06,
"epoch": 7.220766559015295,
"step": 118500
},
{
"loss": 0.1449,
"grad_norm": 25.259172439575195,
"learning_rate": 2.0499827618001496e-06,
"epoch": 7.223813295960027,
"step": 118550
},
{
"loss": 0.102,
"grad_norm": 0.5137035250663757,
"learning_rate": 2.0419649944276516e-06,
"epoch": 7.226860032904759,
"step": 118600
},
{
"loss": 0.0706,
"grad_norm": 1.6865817308425903,
"learning_rate": 2.0339472270551546e-06,
"epoch": 7.229906769849491,
"step": 118650
},
{
"loss": 0.1146,
"grad_norm": 0.1356986165046692,
"learning_rate": 2.025929459682657e-06,
"epoch": 7.232953506794224,
"step": 118700
},
{
"loss": 0.1152,
"grad_norm": 2.088418483734131,
"learning_rate": 2.0179116923101596e-06,
"epoch": 7.236000243738956,
"step": 118750
},
{
"loss": 0.0829,
"grad_norm": 0.2632194757461548,
"learning_rate": 2.009893924937662e-06,
"epoch": 7.239046980683688,
"step": 118800
},
{
"loss": 0.087,
"grad_norm": 9.009870529174805,
"learning_rate": 2.0018761575651646e-06,
"epoch": 7.24209371762842,
"step": 118850
},
{
"loss": 0.0907,
"grad_norm": 0.17808888852596283,
"learning_rate": 1.993858390192667e-06,
"epoch": 7.245140454573152,
"step": 118900
},
{
"loss": 0.1208,
"grad_norm": 13.083930969238281,
"learning_rate": 1.9858406228201696e-06,
"epoch": 7.248187191517884,
"step": 118950
},
{
"loss": 0.0646,
"grad_norm": 0.03696560859680176,
"learning_rate": 1.977822855447672e-06,
"epoch": 7.251233928462616,
"step": 119000
},
{
"loss": 0.0765,
"grad_norm": 0.16806836426258087,
"learning_rate": 1.969805088075175e-06,
"epoch": 7.254280665407348,
"step": 119050
},
{
"loss": 0.0925,
"grad_norm": 8.887118339538574,
"learning_rate": 1.961787320702677e-06,
"epoch": 7.257327402352081,
"step": 119100
},
{
"loss": 0.072,
"grad_norm": 1.7048462629318237,
"learning_rate": 1.9537695533301796e-06,
"epoch": 7.260374139296813,
"step": 119150
},
{
"loss": 0.1089,
"grad_norm": 5.171725273132324,
"learning_rate": 1.9459121413051324e-06,
"epoch": 7.263420876241545,
"step": 119200
},
{
"loss": 0.0815,
"grad_norm": 8.433769226074219,
"learning_rate": 1.937894373932635e-06,
"epoch": 7.266467613186277,
"step": 119250
},
{
"loss": 0.0962,
"grad_norm": 0.466909259557724,
"learning_rate": 1.9298766065601374e-06,
"epoch": 7.2695143501310096,
"step": 119300
},
{
"loss": 0.1276,
"grad_norm": 13.068318367004395,
"learning_rate": 1.92185883918764e-06,
"epoch": 7.272561087075742,
"step": 119350
},
{
"loss": 0.0943,
"grad_norm": 0.24151694774627686,
"learning_rate": 1.913841071815143e-06,
"epoch": 7.275607824020474,
"step": 119400
},
{
"loss": 0.1379,
"grad_norm": 0.14944763481616974,
"learning_rate": 1.9058233044426449e-06,
"epoch": 7.278654560965206,
"step": 119450
},
{
"loss": 0.0995,
"grad_norm": 2.1592702865600586,
"learning_rate": 1.8978055370701476e-06,
"epoch": 7.2817012979099385,
"step": 119500
},
{
"loss": 0.0809,
"grad_norm": 0.9306973218917847,
"learning_rate": 1.8897877696976503e-06,
"epoch": 7.284748034854671,
"step": 119550
},
{
"loss": 0.0759,
"grad_norm": 10.602608680725098,
"learning_rate": 1.8817700023251526e-06,
"epoch": 7.287794771799403,
"step": 119600
},
{
"loss": 0.0794,
"grad_norm": 2.2153685092926025,
"learning_rate": 1.8737522349526551e-06,
"epoch": 7.290841508744135,
"step": 119650
},
{
"loss": 0.1043,
"grad_norm": 0.05096817761659622,
"learning_rate": 1.8657344675801578e-06,
"epoch": 7.293888245688867,
"step": 119700
},
{
"loss": 0.1033,
"grad_norm": 0.02350297011435032,
"learning_rate": 1.8577167002076603e-06,
"epoch": 7.2969349826336,
"step": 119750
},
{
"loss": 0.1325,
"grad_norm": 0.18315041065216064,
"learning_rate": 1.8496989328351628e-06,
"epoch": 7.299981719578332,
"step": 119800
},
{
"loss": 0.1069,
"grad_norm": 0.012904784642159939,
"learning_rate": 1.8416811654626653e-06,
"epoch": 7.303028456523064,
"step": 119850
},
{
"loss": 0.093,
"grad_norm": 0.680059552192688,
"learning_rate": 1.833663398090168e-06,
"epoch": 7.306075193467796,
"step": 119900
},
{
"loss": 0.1128,
"grad_norm": 1.8455207347869873,
"learning_rate": 1.8256456307176703e-06,
"epoch": 7.309121930412529,
"step": 119950
},
{
"loss": 0.1107,
"grad_norm": 0.03370470553636551,
"learning_rate": 1.817627863345173e-06,
"epoch": 7.312168667357261,
"step": 120000
},
{
"loss": 0.0842,
"grad_norm": 0.1649124026298523,
"learning_rate": 1.8096100959726756e-06,
"epoch": 7.315215404301993,
"step": 120050
},
{
"loss": 0.0821,
"grad_norm": 1.4715006351470947,
"learning_rate": 1.8015923286001783e-06,
"epoch": 7.318262141246724,
"step": 120100
},
{
"loss": 0.1016,
"grad_norm": 0.02068573236465454,
"learning_rate": 1.7935745612276806e-06,
"epoch": 7.321308878191457,
"step": 120150
},
{
"loss": 0.0931,
"grad_norm": 2.7271456718444824,
"learning_rate": 1.7855567938551833e-06,
"epoch": 7.324355615136189,
"step": 120200
},
{
"loss": 0.1273,
"grad_norm": 6.8524250984191895,
"learning_rate": 1.7775390264826858e-06,
"epoch": 7.327402352080921,
"step": 120250
},
{
"loss": 0.1224,
"grad_norm": 4.887910842895508,
"learning_rate": 1.7695212591101883e-06,
"epoch": 7.330449089025653,
"step": 120300
},
{
"loss": 0.0927,
"grad_norm": 1.217824101448059,
"learning_rate": 1.7615034917376908e-06,
"epoch": 7.3334958259703855,
"step": 120350
},
{
"loss": 0.0773,
"grad_norm": 2.066261053085327,
"learning_rate": 1.7534857243651935e-06,
"epoch": 7.336542562915118,
"step": 120400
},
{
"loss": 0.0933,
"grad_norm": 3.932084321975708,
"learning_rate": 1.745467956992696e-06,
"epoch": 7.33958929985985,
"step": 120450
},
{
"loss": 0.0897,
"grad_norm": 1.3706263303756714,
"learning_rate": 1.7374501896201983e-06,
"epoch": 7.342636036804582,
"step": 120500
},
{
"loss": 0.0965,
"grad_norm": 0.33276820182800293,
"learning_rate": 1.729432422247701e-06,
"epoch": 7.345682773749314,
"step": 120550
},
{
"loss": 0.0922,
"grad_norm": 0.6193519234657288,
"learning_rate": 1.7214146548752037e-06,
"epoch": 7.348729510694047,
"step": 120600
},
{
"loss": 0.0718,
"grad_norm": 11.159337997436523,
"learning_rate": 1.7133968875027062e-06,
"epoch": 7.351776247638779,
"step": 120650
},
{
"loss": 0.1094,
"grad_norm": 11.781612396240234,
"learning_rate": 1.7053791201302085e-06,
"epoch": 7.354822984583511,
"step": 120700
},
{
"loss": 0.0975,
"grad_norm": 9.19640827178955,
"learning_rate": 1.6973613527577112e-06,
"epoch": 7.357869721528243,
"step": 120750
},
{
"loss": 0.1028,
"grad_norm": 15.428553581237793,
"learning_rate": 1.689343585385214e-06,
"epoch": 7.360916458472976,
"step": 120800
},
{
"loss": 0.1275,
"grad_norm": 2.8069446086883545,
"learning_rate": 1.6813258180127162e-06,
"epoch": 7.363963195417708,
"step": 120850
},
{
"loss": 0.0596,
"grad_norm": 0.9202632308006287,
"learning_rate": 1.6733080506402187e-06,
"epoch": 7.36700993236244,
"step": 120900
},
{
"loss": 0.1136,
"grad_norm": 0.6722289323806763,
"learning_rate": 1.6652902832677215e-06,
"epoch": 7.370056669307172,
"step": 120950
},
{
"loss": 0.0981,
"grad_norm": 0.007202757988125086,
"learning_rate": 1.6572725158952242e-06,
"epoch": 7.3731034062519045,
"step": 121000
},
{
"loss": 0.0922,
"grad_norm": 0.07914326339960098,
"learning_rate": 1.6492547485227265e-06,
"epoch": 7.376150143196637,
"step": 121050
},
{
"loss": 0.0721,
"grad_norm": 0.7331340909004211,
"learning_rate": 1.641236981150229e-06,
"epoch": 7.379196880141369,
"step": 121100
},
{
"loss": 0.0988,
"grad_norm": 0.012063548900187016,
"learning_rate": 1.6332192137777317e-06,
"epoch": 7.3822436170861,
"step": 121150
},
{
"loss": 0.0645,
"grad_norm": 0.2977806627750397,
"learning_rate": 1.625201446405234e-06,
"epoch": 7.385290354030833,
"step": 121200
},
{
"loss": 0.1043,
"grad_norm": 0.627149760723114,
"learning_rate": 1.6171836790327367e-06,
"epoch": 7.388337090975565,
"step": 121250
},
{
"loss": 0.1239,
"grad_norm": 0.2987671494483948,
"learning_rate": 1.6091659116602392e-06,
"epoch": 7.391383827920297,
"step": 121300
},
{
"loss": 0.1207,
"grad_norm": 7.091290473937988,
"learning_rate": 1.6013084996351915e-06,
"epoch": 7.394430564865029,
"step": 121350
},
{
"loss": 0.0737,
"grad_norm": 0.7925600409507751,
"learning_rate": 1.5932907322626942e-06,
"epoch": 7.3974773018097615,
"step": 121400
},
{
"loss": 0.0619,
"grad_norm": 9.005824089050293,
"learning_rate": 1.5852729648901968e-06,
"epoch": 7.400524038754494,
"step": 121450
},
{
"loss": 0.1056,
"grad_norm": 1.3129178285598755,
"learning_rate": 1.5772551975176995e-06,
"epoch": 7.403570775699226,
"step": 121500
},
{
"loss": 0.106,
"grad_norm": 0.047418951988220215,
"learning_rate": 1.5692374301452018e-06,
"epoch": 7.406617512643958,
"step": 121550
},
{
"loss": 0.1053,
"grad_norm": 7.941510200500488,
"learning_rate": 1.5612196627727045e-06,
"epoch": 7.40966424958869,
"step": 121600
},
{
"loss": 0.1093,
"grad_norm": 22.316518783569336,
"learning_rate": 1.553201895400207e-06,
"epoch": 7.412710986533423,
"step": 121650
},
{
"loss": 0.1362,
"grad_norm": 7.48456335067749,
"learning_rate": 1.5451841280277095e-06,
"epoch": 7.415757723478155,
"step": 121700
},
{
"loss": 0.0843,
"grad_norm": 3.5251853466033936,
"learning_rate": 1.537166360655212e-06,
"epoch": 7.418804460422887,
"step": 121750
},
{
"loss": 0.1051,
"grad_norm": 0.03309309110045433,
"learning_rate": 1.5291485932827147e-06,
"epoch": 7.421851197367619,
"step": 121800
},
{
"loss": 0.1275,
"grad_norm": 0.026798337697982788,
"learning_rate": 1.5211308259102172e-06,
"epoch": 7.424897934312352,
"step": 121850
},
{
"loss": 0.0991,
"grad_norm": 7.9568562507629395,
"learning_rate": 1.5131130585377197e-06,
"epoch": 7.427944671257084,
"step": 121900
},
{
"loss": 0.1255,
"grad_norm": 0.7069330215454102,
"learning_rate": 1.5050952911652222e-06,
"epoch": 7.430991408201816,
"step": 121950
},
{
"loss": 0.0892,
"grad_norm": 0.050835639238357544,
"learning_rate": 1.497077523792725e-06,
"epoch": 7.434038145146548,
"step": 122000
},
{
"loss": 0.0906,
"grad_norm": 0.02664211578667164,
"learning_rate": 1.4890597564202272e-06,
"epoch": 7.4370848820912805,
"step": 122050
},
{
"loss": 0.0755,
"grad_norm": 0.024564215913414955,
"learning_rate": 1.48104198904773e-06,
"epoch": 7.440131619036013,
"step": 122100
},
{
"loss": 0.1169,
"grad_norm": 0.027415499091148376,
"learning_rate": 1.4730242216752324e-06,
"epoch": 7.443178355980745,
"step": 122150
},
{
"loss": 0.081,
"grad_norm": 1.9475950002670288,
"learning_rate": 1.4650064543027351e-06,
"epoch": 7.446225092925477,
"step": 122200
},
{
"loss": 0.1083,
"grad_norm": 0.20478984713554382,
"learning_rate": 1.4569886869302374e-06,
"epoch": 7.449271829870209,
"step": 122250
},
{
"loss": 0.0978,
"grad_norm": 13.411532402038574,
"learning_rate": 1.4489709195577402e-06,
"epoch": 7.452318566814942,
"step": 122300
},
{
"loss": 0.0973,
"grad_norm": 0.6007531881332397,
"learning_rate": 1.4409531521852427e-06,
"epoch": 7.455365303759673,
"step": 122350
},
{
"loss": 0.0871,
"grad_norm": 11.390092849731445,
"learning_rate": 1.432935384812745e-06,
"epoch": 7.458412040704405,
"step": 122400
},
{
"loss": 0.0764,
"grad_norm": 10.630624771118164,
"learning_rate": 1.4249176174402477e-06,
"epoch": 7.4614587776491375,
"step": 122450
},
{
"loss": 0.1221,
"grad_norm": 3.3518712520599365,
"learning_rate": 1.4168998500677502e-06,
"epoch": 7.46450551459387,
"step": 122500
},
{
"loss": 0.1072,
"grad_norm": 6.126272201538086,
"learning_rate": 1.4088820826952529e-06,
"epoch": 7.467552251538602,
"step": 122550
},
{
"loss": 0.0951,
"grad_norm": 5.60094690322876,
"learning_rate": 1.4008643153227552e-06,
"epoch": 7.470598988483334,
"step": 122600
},
{
"loss": 0.1146,
"grad_norm": 2.659025192260742,
"learning_rate": 1.3928465479502579e-06,
"epoch": 7.473645725428066,
"step": 122650
},
{
"loss": 0.1121,
"grad_norm": 0.009364171884953976,
"learning_rate": 1.3848287805777604e-06,
"epoch": 7.476692462372799,
"step": 122700
},
{
"loss": 0.1156,
"grad_norm": 1.5403653383255005,
"learning_rate": 1.3768110132052631e-06,
"epoch": 7.479739199317531,
"step": 122750
},
{
"loss": 0.0915,
"grad_norm": 9.259322166442871,
"learning_rate": 1.3687932458327654e-06,
"epoch": 7.482785936262263,
"step": 122800
},
{
"loss": 0.1212,
"grad_norm": 7.430643558502197,
"learning_rate": 1.3607754784602681e-06,
"epoch": 7.485832673206995,
"step": 122850
},
{
"loss": 0.1158,
"grad_norm": 0.5428184866905212,
"learning_rate": 1.3527577110877706e-06,
"epoch": 7.4888794101517275,
"step": 122900
},
{
"loss": 0.0924,
"grad_norm": 8.716285705566406,
"learning_rate": 1.3447399437152731e-06,
"epoch": 7.49192614709646,
"step": 122950
},
{
"loss": 0.1033,
"grad_norm": 2.9800915718078613,
"learning_rate": 1.3367221763427756e-06,
"epoch": 7.494972884041192,
"step": 123000
},
{
"loss": 0.087,
"grad_norm": 12.963000297546387,
"learning_rate": 1.3287044089702783e-06,
"epoch": 7.498019620985924,
"step": 123050
},
{
"loss": 0.0784,
"grad_norm": 0.12310778349637985,
"learning_rate": 1.3206866415977808e-06,
"epoch": 7.5010663579306565,
"step": 123100
},
{
"loss": 0.1358,
"grad_norm": 7.4934234619140625,
"learning_rate": 1.3126688742252833e-06,
"epoch": 7.504113094875389,
"step": 123150
},
{
"loss": 0.1004,
"grad_norm": 1.9572930335998535,
"learning_rate": 1.3046511068527859e-06,
"epoch": 7.507159831820121,
"step": 123200
},
{
"loss": 0.0935,
"grad_norm": 0.5527026057243347,
"learning_rate": 1.2966333394802886e-06,
"epoch": 7.510206568764853,
"step": 123250
},
{
"loss": 0.0958,
"grad_norm": 0.7861798405647278,
"learning_rate": 1.2886155721077909e-06,
"epoch": 7.513253305709585,
"step": 123300
},
{
"loss": 0.0817,
"grad_norm": 0.7424635291099548,
"learning_rate": 1.2805978047352936e-06,
"epoch": 7.516300042654318,
"step": 123350
},
{
"loss": 0.0919,
"grad_norm": 0.007623809855431318,
"learning_rate": 1.272580037362796e-06,
"epoch": 7.519346779599049,
"step": 123400
},
{
"loss": 0.0829,
"grad_norm": 2.8712427616119385,
"learning_rate": 1.2645622699902988e-06,
"epoch": 7.522393516543781,
"step": 123450
},
{
"loss": 0.0761,
"grad_norm": 0.0221791360527277,
"learning_rate": 1.256544502617801e-06,
"epoch": 7.525440253488513,
"step": 123500
},
{
"loss": 0.0814,
"grad_norm": 6.351297855377197,
"learning_rate": 1.2485267352453038e-06,
"epoch": 7.528486990433246,
"step": 123550
},
{
"loss": 0.0954,
"grad_norm": 5.22698974609375,
"learning_rate": 1.2405089678728063e-06,
"epoch": 7.531533727377978,
"step": 123600
},
{
"loss": 0.1179,
"grad_norm": 3.7632925510406494,
"learning_rate": 1.2324912005003088e-06,
"epoch": 7.53458046432271,
"step": 123650
},
{
"loss": 0.0985,
"grad_norm": 3.2731010913848877,
"learning_rate": 1.2244734331278113e-06,
"epoch": 7.537627201267442,
"step": 123700
},
{
"loss": 0.0877,
"grad_norm": 9.08704662322998,
"learning_rate": 1.216455665755314e-06,
"epoch": 7.540673938212175,
"step": 123750
},
{
"loss": 0.0943,
"grad_norm": 15.546286582946777,
"learning_rate": 1.2084378983828163e-06,
"epoch": 7.543720675156907,
"step": 123800
},
{
"loss": 0.0852,
"grad_norm": 7.101307392120361,
"learning_rate": 1.2004201310103188e-06,
"epoch": 7.546767412101639,
"step": 123850
},
{
"loss": 0.0776,
"grad_norm": 0.7207491397857666,
"learning_rate": 1.1924023636378215e-06,
"epoch": 7.549814149046371,
"step": 123900
},
{
"loss": 0.1276,
"grad_norm": 6.661449432373047,
"learning_rate": 1.1845449516127739e-06,
"epoch": 7.5528608859911035,
"step": 123950
},
{
"loss": 0.0956,
"grad_norm": 13.67987060546875,
"learning_rate": 1.1765271842402766e-06,
"epoch": 7.555907622935836,
"step": 124000
},
{
"loss": 0.0768,
"grad_norm": 0.9964921474456787,
"learning_rate": 1.168509416867779e-06,
"epoch": 7.558954359880568,
"step": 124050
},
{
"loss": 0.1033,
"grad_norm": 13.308755874633789,
"learning_rate": 1.1604916494952818e-06,
"epoch": 7.5620010968253,
"step": 124100
},
{
"loss": 0.0767,
"grad_norm": 3.406313896179199,
"learning_rate": 1.152473882122784e-06,
"epoch": 7.565047833770032,
"step": 124150
},
{
"loss": 0.1095,
"grad_norm": 3.1022257804870605,
"learning_rate": 1.1444561147502866e-06,
"epoch": 7.568094570714765,
"step": 124200
},
{
"loss": 0.0878,
"grad_norm": 0.05570690706372261,
"learning_rate": 1.1364383473777893e-06,
"epoch": 7.571141307659497,
"step": 124250
},
{
"loss": 0.0884,
"grad_norm": 2.5508646965026855,
"learning_rate": 1.1284205800052918e-06,
"epoch": 7.574188044604229,
"step": 124300
},
{
"loss": 0.107,
"grad_norm": 29.31032371520996,
"learning_rate": 1.1204028126327943e-06,
"epoch": 7.577234781548961,
"step": 124350
},
{
"loss": 0.1078,
"grad_norm": 0.358432412147522,
"learning_rate": 1.1123850452602968e-06,
"epoch": 7.580281518493694,
"step": 124400
},
{
"loss": 0.1124,
"grad_norm": 5.05091667175293,
"learning_rate": 1.1043672778877995e-06,
"epoch": 7.583328255438426,
"step": 124450
},
{
"loss": 0.1073,
"grad_norm": 0.015557678416371346,
"learning_rate": 1.096349510515302e-06,
"epoch": 7.586374992383158,
"step": 124500
},
{
"loss": 0.0877,
"grad_norm": 1.8221118450164795,
"learning_rate": 1.0883317431428045e-06,
"epoch": 7.58942172932789,
"step": 124550
},
{
"loss": 0.0608,
"grad_norm": 1.4890902042388916,
"learning_rate": 1.080313975770307e-06,
"epoch": 7.5924684662726225,
"step": 124600
},
{
"loss": 0.065,
"grad_norm": 1.4698193073272705,
"learning_rate": 1.0722962083978095e-06,
"epoch": 7.595515203217354,
"step": 124650
},
{
"loss": 0.1072,
"grad_norm": 5.349312782287598,
"learning_rate": 1.0642784410253123e-06,
"epoch": 7.598561940162086,
"step": 124700
},
{
"loss": 0.1092,
"grad_norm": 4.170076847076416,
"learning_rate": 1.0562606736528148e-06,
"epoch": 7.601608677106818,
"step": 124750
},
{
"loss": 0.0971,
"grad_norm": 0.35171884298324585,
"learning_rate": 1.0482429062803173e-06,
"epoch": 7.6046554140515505,
"step": 124800
},
{
"loss": 0.0904,
"grad_norm": 6.645513534545898,
"learning_rate": 1.0402251389078198e-06,
"epoch": 7.607702150996283,
"step": 124850
},
{
"loss": 0.0621,
"grad_norm": 0.08470941334962845,
"learning_rate": 1.0322073715353223e-06,
"epoch": 7.610748887941015,
"step": 124900
},
{
"loss": 0.0955,
"grad_norm": 0.3136177957057953,
"learning_rate": 1.024189604162825e-06,
"epoch": 7.613795624885747,
"step": 124950
},
{
"loss": 0.0814,
"grad_norm": 0.039937473833560944,
"learning_rate": 1.0161718367903275e-06,
"epoch": 7.6168423618304795,
"step": 125000
},
{
"loss": 0.0791,
"grad_norm": 1.1606723070144653,
"learning_rate": 1.00815406941783e-06,
"epoch": 7.619889098775212,
"step": 125050
},
{
"loss": 0.0915,
"grad_norm": 7.000576019287109,
"learning_rate": 1.0001363020453325e-06,
"epoch": 7.622935835719944,
"step": 125100
},
{
"loss": 0.0678,
"grad_norm": 7.432790279388428,
"learning_rate": 9.921185346728352e-07,
"epoch": 7.625982572664676,
"step": 125150
},
{
"loss": 0.0707,
"grad_norm": 0.010940204374492168,
"learning_rate": 9.841007673003377e-07,
"epoch": 7.629029309609408,
"step": 125200
},
{
"loss": 0.1135,
"grad_norm": 0.5122422575950623,
"learning_rate": 9.7608299992784e-07,
"epoch": 7.632076046554141,
"step": 125250
},
{
"loss": 0.0953,
"grad_norm": 3.93365216255188,
"learning_rate": 9.680652325553427e-07,
"epoch": 7.635122783498873,
"step": 125300
},
{
"loss": 0.0683,
"grad_norm": 0.23809945583343506,
"learning_rate": 9.600474651828452e-07,
"epoch": 7.638169520443605,
"step": 125350
},
{
"loss": 0.0855,
"grad_norm": 12.335641860961914,
"learning_rate": 9.520296978103478e-07,
"epoch": 7.641216257388337,
"step": 125400
},
{
"loss": 0.1206,
"grad_norm": 0.2151731699705124,
"learning_rate": 9.440119304378503e-07,
"epoch": 7.6442629943330695,
"step": 125450
},
{
"loss": 0.0818,
"grad_norm": 0.06000524014234543,
"learning_rate": 9.35994163065353e-07,
"epoch": 7.647309731277802,
"step": 125500
},
{
"loss": 0.1176,
"grad_norm": 21.840667724609375,
"learning_rate": 9.279763956928555e-07,
"epoch": 7.650356468222534,
"step": 125550
},
{
"loss": 0.1112,
"grad_norm": 0.020430877804756165,
"learning_rate": 9.19958628320358e-07,
"epoch": 7.653403205167266,
"step": 125600
},
{
"loss": 0.1117,
"grad_norm": 5.955739974975586,
"learning_rate": 9.119408609478606e-07,
"epoch": 7.656449942111998,
"step": 125650
},
{
"loss": 0.1035,
"grad_norm": 0.04285252094268799,
"learning_rate": 9.039230935753631e-07,
"epoch": 7.65949667905673,
"step": 125700
},
{
"loss": 0.1362,
"grad_norm": 0.005909433122724295,
"learning_rate": 8.959053262028657e-07,
"epoch": 7.662543416001462,
"step": 125750
},
{
"loss": 0.0768,
"grad_norm": 6.28010368347168,
"learning_rate": 8.878875588303682e-07,
"epoch": 7.665590152946194,
"step": 125800
},
{
"loss": 0.1177,
"grad_norm": 0.008507492020726204,
"learning_rate": 8.798697914578708e-07,
"epoch": 7.6686368898909265,
"step": 125850
},
{
"loss": 0.0975,
"grad_norm": 0.025365469977259636,
"learning_rate": 8.718520240853733e-07,
"epoch": 7.671683626835659,
"step": 125900
},
{
"loss": 0.1005,
"grad_norm": 0.05381585657596588,
"learning_rate": 8.638342567128757e-07,
"epoch": 7.674730363780391,
"step": 125950
},
{
"loss": 0.0964,
"grad_norm": 0.4972221255302429,
"learning_rate": 8.558164893403784e-07,
"epoch": 7.677777100725123,
"step": 126000
},
{
"loss": 0.066,
"grad_norm": 0.028611665591597557,
"learning_rate": 8.477987219678808e-07,
"epoch": 7.680823837669855,
"step": 126050
},
{
"loss": 0.0931,
"grad_norm": 13.591317176818848,
"learning_rate": 8.397809545953835e-07,
"epoch": 7.683870574614588,
"step": 126100
},
{
"loss": 0.096,
"grad_norm": 0.037239667028188705,
"learning_rate": 8.317631872228859e-07,
"epoch": 7.68691731155932,
"step": 126150
},
{
"loss": 0.1235,
"grad_norm": 0.07474205642938614,
"learning_rate": 8.237454198503886e-07,
"epoch": 7.689964048504052,
"step": 126200
},
{
"loss": 0.0919,
"grad_norm": 2.5956828594207764,
"learning_rate": 8.158880078253411e-07,
"epoch": 7.693010785448784,
"step": 126250
},
{
"loss": 0.0934,
"grad_norm": 8.687529563903809,
"learning_rate": 8.078702404528435e-07,
"epoch": 7.696057522393517,
"step": 126300
},
{
"loss": 0.0925,
"grad_norm": 11.321565628051758,
"learning_rate": 7.998524730803462e-07,
"epoch": 7.699104259338249,
"step": 126350
},
{
"loss": 0.0716,
"grad_norm": 0.0783090591430664,
"learning_rate": 7.918347057078486e-07,
"epoch": 7.702150996282981,
"step": 126400
},
{
"loss": 0.0936,
"grad_norm": 10.89949893951416,
"learning_rate": 7.838169383353512e-07,
"epoch": 7.705197733227713,
"step": 126450
},
{
"loss": 0.1084,
"grad_norm": 4.405998706817627,
"learning_rate": 7.757991709628537e-07,
"epoch": 7.7082444701724455,
"step": 126500
},
{
"loss": 0.135,
"grad_norm": 8.283391952514648,
"learning_rate": 7.677814035903563e-07,
"epoch": 7.711291207117178,
"step": 126550
},
{
"loss": 0.0856,
"grad_norm": 3.188230037689209,
"learning_rate": 7.597636362178588e-07,
"epoch": 7.71433794406191,
"step": 126600
},
{
"loss": 0.0913,
"grad_norm": 19.47659683227539,
"learning_rate": 7.517458688453613e-07,
"epoch": 7.717384681006642,
"step": 126650
},
{
"loss": 0.1139,
"grad_norm": 0.42966464161872864,
"learning_rate": 7.437281014728639e-07,
"epoch": 7.720431417951374,
"step": 126700
},
{
"loss": 0.1094,
"grad_norm": 0.3684709370136261,
"learning_rate": 7.357103341003664e-07,
"epoch": 7.723478154896107,
"step": 126750
},
{
"loss": 0.139,
"grad_norm": 8.219767570495605,
"learning_rate": 7.27692566727869e-07,
"epoch": 7.726524891840839,
"step": 126800
},
{
"loss": 0.1037,
"grad_norm": 5.1842241287231445,
"learning_rate": 7.196747993553715e-07,
"epoch": 7.729571628785571,
"step": 126850
},
{
"loss": 0.1175,
"grad_norm": 2.888819694519043,
"learning_rate": 7.116570319828741e-07,
"epoch": 7.7326183657303025,
"step": 126900
},
{
"loss": 0.1189,
"grad_norm": 7.3930134773254395,
"learning_rate": 7.036392646103767e-07,
"epoch": 7.735665102675035,
"step": 126950
},
{
"loss": 0.1343,
"grad_norm": 0.40427955985069275,
"learning_rate": 6.956214972378792e-07,
"epoch": 7.738711839619767,
"step": 127000
},
{
"loss": 0.0569,
"grad_norm": 7.049976825714111,
"learning_rate": 6.876037298653818e-07,
"epoch": 7.741758576564499,
"step": 127050
},
{
"loss": 0.1393,
"grad_norm": 0.6430282592773438,
"learning_rate": 6.795859624928843e-07,
"epoch": 7.744805313509231,
"step": 127100
},
{
"loss": 0.1024,
"grad_norm": 8.649250984191895,
"learning_rate": 6.715681951203869e-07,
"epoch": 7.747852050453964,
"step": 127150
},
{
"loss": 0.1125,
"grad_norm": 5.9833526611328125,
"learning_rate": 6.635504277478894e-07,
"epoch": 7.750898787398696,
"step": 127200
},
{
"loss": 0.0595,
"grad_norm": 2.9306867122650146,
"learning_rate": 6.55532660375392e-07,
"epoch": 7.753945524343428,
"step": 127250
},
{
"loss": 0.1265,
"grad_norm": 1.2613314390182495,
"learning_rate": 6.475148930028945e-07,
"epoch": 7.75699226128816,
"step": 127300
},
{
"loss": 0.0878,
"grad_norm": 0.02675698883831501,
"learning_rate": 6.39497125630397e-07,
"epoch": 7.760038998232893,
"step": 127350
},
{
"loss": 0.0693,
"grad_norm": 0.3744986057281494,
"learning_rate": 6.314793582578996e-07,
"epoch": 7.763085735177625,
"step": 127400
},
{
"loss": 0.0947,
"grad_norm": 0.035280197858810425,
"learning_rate": 6.234615908854021e-07,
"epoch": 7.766132472122357,
"step": 127450
},
{
"loss": 0.0877,
"grad_norm": 9.804384231567383,
"learning_rate": 6.154438235129046e-07,
"epoch": 7.769179209067089,
"step": 127500
},
{
"loss": 0.0895,
"grad_norm": 4.930053234100342,
"learning_rate": 6.074260561404072e-07,
"epoch": 7.7722259460118215,
"step": 127550
},
{
"loss": 0.0998,
"grad_norm": 17.574127197265625,
"learning_rate": 5.994082887679097e-07,
"epoch": 7.775272682956554,
"step": 127600
},
{
"loss": 0.0637,
"grad_norm": 6.293902397155762,
"learning_rate": 5.913905213954123e-07,
"epoch": 7.778319419901286,
"step": 127650
},
{
"loss": 0.073,
"grad_norm": 25.46702003479004,
"learning_rate": 5.833727540229148e-07,
"epoch": 7.781366156846018,
"step": 127700
},
{
"loss": 0.054,
"grad_norm": 3.1991546154022217,
"learning_rate": 5.753549866504173e-07,
"epoch": 7.78441289379075,
"step": 127750
},
{
"loss": 0.1397,
"grad_norm": 0.015462794341146946,
"learning_rate": 5.6733721927792e-07,
"epoch": 7.787459630735483,
"step": 127800
},
{
"loss": 0.0903,
"grad_norm": 15.515666007995605,
"learning_rate": 5.593194519054225e-07,
"epoch": 7.790506367680215,
"step": 127850
},
{
"loss": 0.0779,
"grad_norm": 0.22044596076011658,
"learning_rate": 5.51301684532925e-07,
"epoch": 7.793553104624946,
"step": 127900
},
{
"loss": 0.0963,
"grad_norm": 7.253251075744629,
"learning_rate": 5.432839171604276e-07,
"epoch": 7.7965998415696784,
"step": 127950
},
{
"loss": 0.1059,
"grad_norm": 13.26782512664795,
"learning_rate": 5.352661497879301e-07,
"epoch": 7.799646578514411,
"step": 128000
},
{
"loss": 0.1096,
"grad_norm": 0.26830458641052246,
"learning_rate": 5.272483824154327e-07,
"epoch": 7.802693315459143,
"step": 128050
},
{
"loss": 0.1461,
"grad_norm": 0.08780431002378464,
"learning_rate": 5.192306150429352e-07,
"epoch": 7.805740052403875,
"step": 128100
},
{
"loss": 0.1507,
"grad_norm": 0.09142392128705978,
"learning_rate": 5.112128476704378e-07,
"epoch": 7.808786789348607,
"step": 128150
},
{
"loss": 0.1322,
"grad_norm": 0.010981561616063118,
"learning_rate": 5.031950802979403e-07,
"epoch": 7.81183352629334,
"step": 128200
},
{
"loss": 0.084,
"grad_norm": 0.03634430095553398,
"learning_rate": 4.951773129254428e-07,
"epoch": 7.814880263238072,
"step": 128250
},
{
"loss": 0.1026,
"grad_norm": 12.6382417678833,
"learning_rate": 4.871595455529454e-07,
"epoch": 7.817927000182804,
"step": 128300
},
{
"loss": 0.1072,
"grad_norm": 0.01119630504399538,
"learning_rate": 4.791417781804479e-07,
"epoch": 7.820973737127536,
"step": 128350
},
{
"loss": 0.1078,
"grad_norm": 0.830100953578949,
"learning_rate": 4.7112401080795046e-07,
"epoch": 7.8240204740722685,
"step": 128400
},
{
"loss": 0.0731,
"grad_norm": 0.11190879344940186,
"learning_rate": 4.63106243435453e-07,
"epoch": 7.827067211017001,
"step": 128450
},
{
"loss": 0.153,
"grad_norm": 19.6695556640625,
"learning_rate": 4.550884760629556e-07,
"epoch": 7.830113947961733,
"step": 128500
},
{
"loss": 0.0826,
"grad_norm": 0.14691300690174103,
"learning_rate": 4.470707086904581e-07,
"epoch": 7.833160684906465,
"step": 128550
},
{
"loss": 0.1004,
"grad_norm": 10.164366722106934,
"learning_rate": 4.3905294131796064e-07,
"epoch": 7.8362074218511975,
"step": 128600
},
{
"loss": 0.086,
"grad_norm": 8.557938575744629,
"learning_rate": 4.3119552929291313e-07,
"epoch": 7.83925415879593,
"step": 128650
},
{
"loss": 0.0925,
"grad_norm": 3.8639073371887207,
"learning_rate": 4.231777619204157e-07,
"epoch": 7.842300895740662,
"step": 128700
},
{
"loss": 0.1125,
"grad_norm": 0.10647812485694885,
"learning_rate": 4.1515999454791824e-07,
"epoch": 7.845347632685394,
"step": 128750
},
{
"loss": 0.1368,
"grad_norm": 0.9484306573867798,
"learning_rate": 4.071422271754208e-07,
"epoch": 7.848394369630126,
"step": 128800
},
{
"loss": 0.1016,
"grad_norm": 0.03759833797812462,
"learning_rate": 3.9912445980292335e-07,
"epoch": 7.851441106574859,
"step": 128850
},
{
"loss": 0.0801,
"grad_norm": 1.4730113744735718,
"learning_rate": 3.911066924304258e-07,
"epoch": 7.854487843519591,
"step": 128900
},
{
"loss": 0.1041,
"grad_norm": 1.2740504741668701,
"learning_rate": 3.8308892505792836e-07,
"epoch": 7.857534580464323,
"step": 128950
},
{
"loss": 0.1325,
"grad_norm": 4.201242446899414,
"learning_rate": 3.750711576854309e-07,
"epoch": 7.860581317409055,
"step": 129000
},
{
"loss": 0.0941,
"grad_norm": 5.97859001159668,
"learning_rate": 3.6705339031293347e-07,
"epoch": 7.8636280543537875,
"step": 129050
},
{
"loss": 0.115,
"grad_norm": 23.125551223754883,
"learning_rate": 3.5903562294043603e-07,
"epoch": 7.86667479129852,
"step": 129100
},
{
"loss": 0.0532,
"grad_norm": 0.47117552161216736,
"learning_rate": 3.510178555679386e-07,
"epoch": 7.869721528243251,
"step": 129150
},
{
"loss": 0.0832,
"grad_norm": 1.7530359029769897,
"learning_rate": 3.4300008819544114e-07,
"epoch": 7.872768265187983,
"step": 129200
},
{
"loss": 0.0798,
"grad_norm": 0.549109160900116,
"learning_rate": 3.3498232082294365e-07,
"epoch": 7.875815002132716,
"step": 129250
},
{
"loss": 0.1029,
"grad_norm": 0.02954717166721821,
"learning_rate": 3.269645534504462e-07,
"epoch": 7.878861739077448,
"step": 129300
},
{
"loss": 0.1079,
"grad_norm": 0.5776802897453308,
"learning_rate": 3.1894678607794876e-07,
"epoch": 7.88190847602218,
"step": 129350
},
{
"loss": 0.0925,
"grad_norm": 10.430231094360352,
"learning_rate": 3.109290187054513e-07,
"epoch": 7.884955212966912,
"step": 129400
},
{
"loss": 0.0995,
"grad_norm": 23.0455322265625,
"learning_rate": 3.0291125133295387e-07,
"epoch": 7.8880019499116445,
"step": 129450
},
{
"loss": 0.1081,
"grad_norm": 0.272248774766922,
"learning_rate": 2.948934839604564e-07,
"epoch": 7.891048686856377,
"step": 129500
},
{
"loss": 0.0601,
"grad_norm": 4.684593677520752,
"learning_rate": 2.8687571658795893e-07,
"epoch": 7.894095423801109,
"step": 129550
},
{
"loss": 0.1114,
"grad_norm": 5.806532382965088,
"learning_rate": 2.788579492154615e-07,
"epoch": 7.897142160745841,
"step": 129600
},
{
"loss": 0.0406,
"grad_norm": 1.7743923664093018,
"learning_rate": 2.7084018184296404e-07,
"epoch": 7.900188897690573,
"step": 129650
},
{
"loss": 0.0949,
"grad_norm": 8.718840599060059,
"learning_rate": 2.6282241447046655e-07,
"epoch": 7.903235634635306,
"step": 129700
},
{
"loss": 0.0918,
"grad_norm": 9.977173805236816,
"learning_rate": 2.548046470979691e-07,
"epoch": 7.906282371580038,
"step": 129750
},
{
"loss": 0.0917,
"grad_norm": 3.153195858001709,
"learning_rate": 2.4678687972547166e-07,
"epoch": 7.90932910852477,
"step": 129800
},
{
"loss": 0.0971,
"grad_norm": 14.393452644348145,
"learning_rate": 2.387691123529742e-07,
"epoch": 7.912375845469502,
"step": 129850
},
{
"loss": 0.0729,
"grad_norm": 12.566503524780273,
"learning_rate": 2.3075134498047674e-07,
"epoch": 7.915422582414235,
"step": 129900
},
{
"loss": 0.1047,
"grad_norm": 4.9103617668151855,
"learning_rate": 2.227335776079793e-07,
"epoch": 7.918469319358967,
"step": 129950
},
{
"loss": 0.1275,
"grad_norm": 12.317655563354492,
"learning_rate": 2.1471581023548186e-07,
"epoch": 7.921516056303699,
"step": 130000
},
{
"loss": 0.191,
"grad_norm": 14.616451263427734,
"learning_rate": 2.0669804286298439e-07,
"epoch": 7.924562793248431,
"step": 130050
},
{
"loss": 0.1177,
"grad_norm": 3.0636723041534424,
"learning_rate": 1.9868027549048694e-07,
"epoch": 7.9276095301931635,
"step": 130100
},
{
"loss": 0.076,
"grad_norm": 1.2362186908721924,
"learning_rate": 1.906625081179895e-07,
"epoch": 7.930656267137895,
"step": 130150
},
{
"loss": 0.1089,
"grad_norm": 0.04036503657698631,
"learning_rate": 1.82644740745492e-07,
"epoch": 7.933703004082627,
"step": 130200
},
{
"loss": 0.1062,
"grad_norm": 0.9977279305458069,
"learning_rate": 1.7462697337299456e-07,
"epoch": 7.936749741027359,
"step": 130250
},
{
"loss": 0.082,
"grad_norm": 0.0652448982000351,
"learning_rate": 1.6676956134794708e-07,
"epoch": 7.9397964779720915,
"step": 130300
},
{
"loss": 0.1372,
"grad_norm": 4.645798206329346,
"learning_rate": 1.587517939754496e-07,
"epoch": 7.942843214916824,
"step": 130350
},
{
"loss": 0.1056,
"grad_norm": 3.977147340774536,
"learning_rate": 1.5073402660295216e-07,
"epoch": 7.945889951861556,
"step": 130400
},
{
"loss": 0.0988,
"grad_norm": 0.9203997850418091,
"learning_rate": 1.427162592304547e-07,
"epoch": 7.948936688806288,
"step": 130450
},
{
"loss": 0.1158,
"grad_norm": 5.688488006591797,
"learning_rate": 1.3469849185795725e-07,
"epoch": 7.9519834257510205,
"step": 130500
},
{
"loss": 0.1193,
"grad_norm": 7.420154094696045,
"learning_rate": 1.2668072448545978e-07,
"epoch": 7.955030162695753,
"step": 130550
},
{
"loss": 0.0888,
"grad_norm": 0.1533631831407547,
"learning_rate": 1.1866295711296234e-07,
"epoch": 7.958076899640485,
"step": 130600
},
{
"loss": 0.1162,
"grad_norm": 0.011089220643043518,
"learning_rate": 1.1064518974046488e-07,
"epoch": 7.961123636585217,
"step": 130650
},
{
"loss": 0.0787,
"grad_norm": 0.19594143331050873,
"learning_rate": 1.0262742236796742e-07,
"epoch": 7.964170373529949,
"step": 130700
},
{
"loss": 0.0819,
"grad_norm": 0.13441698253154755,
"learning_rate": 9.460965499546997e-08,
"epoch": 7.967217110474682,
"step": 130750
},
{
"loss": 0.082,
"grad_norm": 0.0023586221504956484,
"learning_rate": 8.659188762297252e-08,
"epoch": 7.970263847419414,
"step": 130800
},
{
"loss": 0.1086,
"grad_norm": 13.072683334350586,
"learning_rate": 7.857412025047505e-08,
"epoch": 7.973310584364146,
"step": 130850
},
{
"loss": 0.0894,
"grad_norm": 14.6112642288208,
"learning_rate": 7.055635287797761e-08,
"epoch": 7.976357321308878,
"step": 130900
},
{
"loss": 0.1224,
"grad_norm": 17.01149559020996,
"learning_rate": 6.253858550548015e-08,
"epoch": 7.9794040582536105,
"step": 130950
},
{
"loss": 0.0861,
"grad_norm": 17.6811466217041,
"learning_rate": 5.4520818132982694e-08,
"epoch": 7.982450795198343,
"step": 131000
},
{
"loss": 0.097,
"grad_norm": 3.3666467666625977,
"learning_rate": 4.6503050760485244e-08,
"epoch": 7.985497532143075,
"step": 131050
},
{
"loss": 0.1184,
"grad_norm": 10.550926208496094,
"learning_rate": 3.848528338798778e-08,
"epoch": 7.988544269087807,
"step": 131100
},
{
"loss": 0.1183,
"grad_norm": 0.11769083142280579,
"learning_rate": 3.046751601549033e-08,
"epoch": 7.9915910060325395,
"step": 131150
},
{
"loss": 0.0706,
"grad_norm": 11.574141502380371,
"learning_rate": 2.2449748642992876e-08,
"epoch": 7.994637742977272,
"step": 131200
},
{
"loss": 0.0745,
"grad_norm": 0.03502688184380531,
"learning_rate": 1.443198127049542e-08,
"epoch": 7.997684479922004,
"step": 131250
},
{
"eval_loss": 0.11002838611602783,
"eval_accuracy": 0.9730591259640102,
"eval_f1_macro": 0.9727023139698624,
"eval_top5_accuracy": 0.996846615252785,
"eval_runtime": 216.7058,
"eval_samples_per_second": 134.63,
"eval_steps_per_second": 8.417,
"epoch": 8.0,
"step": 131288
},
{
"train_runtime": 37253.6996,
"train_samples_per_second": 56.386,
"train_steps_per_second": 3.524,
"total_flos": 0.0,
"train_loss": 0.37596295774437,
"epoch": 8.0,
"step": 131288
},
{
"eval_loss": 0.11002838611602783,
"eval_accuracy": 0.9730591259640102,
"eval_f1_macro": 0.9727023139698624,
"eval_top5_accuracy": 0.996846615252785,
"eval_runtime": 216.4171,
"eval_samples_per_second": 134.809,
"eval_steps_per_second": 8.428,
"epoch": 8.0,
"step": 131288
}
]