openchat-7b-galbaat / checkpoint-30000 /trainer_state.json
codingninja's picture
Training in progress, step 30000, checkpoint
8f6c2aa verified
Raw
History Blame Contribute Delete
485 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.5706785367802317,
"eval_steps": 500,
"global_step": 30000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0001902261789267439,
"grad_norm": 7.031089782714844,
"learning_rate": 1.9022256039566294e-08,
"loss": 2.4444,
"step": 10
},
{
"epoch": 0.0003804523578534878,
"grad_norm": 13.060206413269043,
"learning_rate": 3.804451207913259e-08,
"loss": 2.614,
"step": 20
},
{
"epoch": 0.0005706785367802317,
"grad_norm": 15.187254905700684,
"learning_rate": 5.7066768118698886e-08,
"loss": 3.1213,
"step": 30
},
{
"epoch": 0.0007609047157069756,
"grad_norm": 19.703323364257812,
"learning_rate": 7.608902415826518e-08,
"loss": 4.1664,
"step": 40
},
{
"epoch": 0.0009511308946337195,
"grad_norm": 32.09828186035156,
"learning_rate": 9.511128019783147e-08,
"loss": 5.9187,
"step": 50
},
{
"epoch": 0.0011413570735604635,
"grad_norm": 6.527578830718994,
"learning_rate": 1.1413353623739777e-07,
"loss": 2.5169,
"step": 60
},
{
"epoch": 0.0013315832524872073,
"grad_norm": 8.701460838317871,
"learning_rate": 1.3315579227696404e-07,
"loss": 2.6031,
"step": 70
},
{
"epoch": 0.0015218094314139513,
"grad_norm": 19.97081184387207,
"learning_rate": 1.5217804831653035e-07,
"loss": 3.1127,
"step": 80
},
{
"epoch": 0.001712035610340695,
"grad_norm": 14.284954071044922,
"learning_rate": 1.7120030435609666e-07,
"loss": 3.9951,
"step": 90
},
{
"epoch": 0.001902261789267439,
"grad_norm": 25.341217041015625,
"learning_rate": 1.9022256039566295e-07,
"loss": 5.4591,
"step": 100
},
{
"epoch": 0.002092487968194183,
"grad_norm": 5.924285411834717,
"learning_rate": 2.0924481643522923e-07,
"loss": 2.4404,
"step": 110
},
{
"epoch": 0.002282714147120927,
"grad_norm": 11.63301944732666,
"learning_rate": 2.2826707247479554e-07,
"loss": 2.5059,
"step": 120
},
{
"epoch": 0.0024729403260476705,
"grad_norm": 11.728666305541992,
"learning_rate": 2.472893285143618e-07,
"loss": 3.1191,
"step": 130
},
{
"epoch": 0.0026631665049744145,
"grad_norm": 9.490508079528809,
"learning_rate": 2.663115845539281e-07,
"loss": 3.8237,
"step": 140
},
{
"epoch": 0.0028533926839011585,
"grad_norm": 24.791996002197266,
"learning_rate": 2.853338405934944e-07,
"loss": 5.5867,
"step": 150
},
{
"epoch": 0.0030436188628279025,
"grad_norm": 4.845604419708252,
"learning_rate": 3.043560966330607e-07,
"loss": 2.3633,
"step": 160
},
{
"epoch": 0.003233845041754646,
"grad_norm": 6.938657283782959,
"learning_rate": 3.2337835267262704e-07,
"loss": 2.3907,
"step": 170
},
{
"epoch": 0.00342407122068139,
"grad_norm": 8.972224235534668,
"learning_rate": 3.424006087121933e-07,
"loss": 2.6722,
"step": 180
},
{
"epoch": 0.003614297399608134,
"grad_norm": 9.294358253479004,
"learning_rate": 3.614228647517596e-07,
"loss": 3.4523,
"step": 190
},
{
"epoch": 0.003804523578534878,
"grad_norm": 19.425222396850586,
"learning_rate": 3.804451207913259e-07,
"loss": 5.2828,
"step": 200
},
{
"epoch": 0.003994749757461622,
"grad_norm": 3.1245577335357666,
"learning_rate": 3.994673768308922e-07,
"loss": 2.191,
"step": 210
},
{
"epoch": 0.004184975936388366,
"grad_norm": 4.416078567504883,
"learning_rate": 4.1848963287045846e-07,
"loss": 2.4138,
"step": 220
},
{
"epoch": 0.00437520211531511,
"grad_norm": 8.128155708312988,
"learning_rate": 4.3751188891002475e-07,
"loss": 2.4941,
"step": 230
},
{
"epoch": 0.004565428294241854,
"grad_norm": 5.850177764892578,
"learning_rate": 4.565341449495911e-07,
"loss": 3.153,
"step": 240
},
{
"epoch": 0.004755654473168598,
"grad_norm": 19.345319747924805,
"learning_rate": 4.7555640098915737e-07,
"loss": 5.0348,
"step": 250
},
{
"epoch": 0.004945880652095341,
"grad_norm": 2.327831983566284,
"learning_rate": 4.945786570287236e-07,
"loss": 2.1702,
"step": 260
},
{
"epoch": 0.005136106831022085,
"grad_norm": 2.7916853427886963,
"learning_rate": 5.136009130682899e-07,
"loss": 2.1161,
"step": 270
},
{
"epoch": 0.005326333009948829,
"grad_norm": 3.9358408451080322,
"learning_rate": 5.326231691078562e-07,
"loss": 2.2815,
"step": 280
},
{
"epoch": 0.005516559188875573,
"grad_norm": 4.891003131866455,
"learning_rate": 5.516454251474225e-07,
"loss": 2.8836,
"step": 290
},
{
"epoch": 0.005706785367802317,
"grad_norm": 22.499956130981445,
"learning_rate": 5.706676811869888e-07,
"loss": 4.7026,
"step": 300
},
{
"epoch": 0.005897011546729061,
"grad_norm": 1.9951050281524658,
"learning_rate": 5.896899372265551e-07,
"loss": 2.0446,
"step": 310
},
{
"epoch": 0.006087237725655805,
"grad_norm": 3.147122859954834,
"learning_rate": 6.087121932661214e-07,
"loss": 2.138,
"step": 320
},
{
"epoch": 0.006277463904582548,
"grad_norm": 3.4201011657714844,
"learning_rate": 6.277344493056876e-07,
"loss": 2.1024,
"step": 330
},
{
"epoch": 0.006467690083509292,
"grad_norm": 4.823904991149902,
"learning_rate": 6.467567053452541e-07,
"loss": 2.7573,
"step": 340
},
{
"epoch": 0.006657916262436036,
"grad_norm": 21.016569137573242,
"learning_rate": 6.657789613848204e-07,
"loss": 4.1976,
"step": 350
},
{
"epoch": 0.00684814244136278,
"grad_norm": 1.872706413269043,
"learning_rate": 6.848012174243867e-07,
"loss": 1.9986,
"step": 360
},
{
"epoch": 0.007038368620289524,
"grad_norm": 2.8600945472717285,
"learning_rate": 7.038234734639529e-07,
"loss": 1.9909,
"step": 370
},
{
"epoch": 0.007228594799216268,
"grad_norm": 3.7441070079803467,
"learning_rate": 7.228457295035192e-07,
"loss": 2.0065,
"step": 380
},
{
"epoch": 0.007418820978143012,
"grad_norm": 5.4876627922058105,
"learning_rate": 7.418679855430855e-07,
"loss": 2.7059,
"step": 390
},
{
"epoch": 0.007609047157069756,
"grad_norm": 24.044830322265625,
"learning_rate": 7.608902415826518e-07,
"loss": 4.0807,
"step": 400
},
{
"epoch": 0.0077992733359964995,
"grad_norm": 1.6859079599380493,
"learning_rate": 7.799124976222181e-07,
"loss": 1.974,
"step": 410
},
{
"epoch": 0.007989499514923244,
"grad_norm": 2.8750579357147217,
"learning_rate": 7.989347536617844e-07,
"loss": 1.835,
"step": 420
},
{
"epoch": 0.008179725693849988,
"grad_norm": 4.05927038192749,
"learning_rate": 8.179570097013506e-07,
"loss": 1.9796,
"step": 430
},
{
"epoch": 0.008369951872776732,
"grad_norm": 6.277343273162842,
"learning_rate": 8.369792657409169e-07,
"loss": 2.474,
"step": 440
},
{
"epoch": 0.008560178051703475,
"grad_norm": 32.71267318725586,
"learning_rate": 8.560015217804832e-07,
"loss": 3.7252,
"step": 450
},
{
"epoch": 0.00875040423063022,
"grad_norm": 1.6604670286178589,
"learning_rate": 8.750237778200495e-07,
"loss": 1.851,
"step": 460
},
{
"epoch": 0.008940630409556963,
"grad_norm": 2.756211280822754,
"learning_rate": 8.940460338596159e-07,
"loss": 1.6797,
"step": 470
},
{
"epoch": 0.009130856588483708,
"grad_norm": 4.476970195770264,
"learning_rate": 9.130682898991822e-07,
"loss": 1.8404,
"step": 480
},
{
"epoch": 0.00932108276741045,
"grad_norm": 6.676766395568848,
"learning_rate": 9.320905459387485e-07,
"loss": 2.2254,
"step": 490
},
{
"epoch": 0.009511308946337196,
"grad_norm": 33.66838836669922,
"learning_rate": 9.511128019783147e-07,
"loss": 3.3597,
"step": 500
},
{
"epoch": 0.009701535125263939,
"grad_norm": 1.771323561668396,
"learning_rate": 9.70135058017881e-07,
"loss": 1.7942,
"step": 510
},
{
"epoch": 0.009891761304190682,
"grad_norm": 3.0909101963043213,
"learning_rate": 9.891573140574472e-07,
"loss": 1.7758,
"step": 520
},
{
"epoch": 0.010081987483117427,
"grad_norm": 4.743190765380859,
"learning_rate": 1.0081795700970136e-06,
"loss": 1.7564,
"step": 530
},
{
"epoch": 0.01027221366204417,
"grad_norm": 7.06041145324707,
"learning_rate": 1.0272018261365798e-06,
"loss": 2.1511,
"step": 540
},
{
"epoch": 0.010462439840970915,
"grad_norm": 26.422157287597656,
"learning_rate": 1.0462240821761462e-06,
"loss": 2.8421,
"step": 550
},
{
"epoch": 0.010652666019897658,
"grad_norm": 1.8234245777130127,
"learning_rate": 1.0652463382157123e-06,
"loss": 1.7149,
"step": 560
},
{
"epoch": 0.010842892198824403,
"grad_norm": 3.059401273727417,
"learning_rate": 1.0842685942552787e-06,
"loss": 1.647,
"step": 570
},
{
"epoch": 0.011033118377751146,
"grad_norm": 4.818949222564697,
"learning_rate": 1.103290850294845e-06,
"loss": 1.5044,
"step": 580
},
{
"epoch": 0.01122334455667789,
"grad_norm": 6.994847774505615,
"learning_rate": 1.1223131063344113e-06,
"loss": 1.8919,
"step": 590
},
{
"epoch": 0.011413570735604634,
"grad_norm": 38.19530487060547,
"learning_rate": 1.1413353623739777e-06,
"loss": 2.4963,
"step": 600
},
{
"epoch": 0.011603796914531377,
"grad_norm": 2.0052971839904785,
"learning_rate": 1.160357618413544e-06,
"loss": 1.6709,
"step": 610
},
{
"epoch": 0.011794023093458122,
"grad_norm": 1.6200284957885742,
"learning_rate": 1.1793798744531103e-06,
"loss": 1.5749,
"step": 620
},
{
"epoch": 0.011984249272384865,
"grad_norm": 2.2375597953796387,
"learning_rate": 1.1984021304926766e-06,
"loss": 1.4411,
"step": 630
},
{
"epoch": 0.01217447545131161,
"grad_norm": 5.535067558288574,
"learning_rate": 1.2174243865322428e-06,
"loss": 1.845,
"step": 640
},
{
"epoch": 0.012364701630238353,
"grad_norm": 9.7356595993042,
"learning_rate": 1.2364466425718092e-06,
"loss": 2.524,
"step": 650
},
{
"epoch": 0.012554927809165096,
"grad_norm": 1.4515656232833862,
"learning_rate": 1.2554688986113752e-06,
"loss": 1.6263,
"step": 660
},
{
"epoch": 0.012745153988091841,
"grad_norm": 1.6473021507263184,
"learning_rate": 1.2744911546509418e-06,
"loss": 1.5024,
"step": 670
},
{
"epoch": 0.012935380167018585,
"grad_norm": 2.3517134189605713,
"learning_rate": 1.2935134106905082e-06,
"loss": 1.4407,
"step": 680
},
{
"epoch": 0.01312560634594533,
"grad_norm": 2.8996622562408447,
"learning_rate": 1.3125356667300743e-06,
"loss": 1.6358,
"step": 690
},
{
"epoch": 0.013315832524872073,
"grad_norm": 11.34484577178955,
"learning_rate": 1.3315579227696407e-06,
"loss": 2.2334,
"step": 700
},
{
"epoch": 0.013506058703798817,
"grad_norm": 1.1996574401855469,
"learning_rate": 1.350580178809207e-06,
"loss": 1.4753,
"step": 710
},
{
"epoch": 0.01369628488272556,
"grad_norm": 1.6772912740707397,
"learning_rate": 1.3696024348487733e-06,
"loss": 1.5256,
"step": 720
},
{
"epoch": 0.013886511061652304,
"grad_norm": 2.156625986099243,
"learning_rate": 1.3886246908883395e-06,
"loss": 1.4419,
"step": 730
},
{
"epoch": 0.014076737240579049,
"grad_norm": 3.710660219192505,
"learning_rate": 1.4076469469279059e-06,
"loss": 1.6054,
"step": 740
},
{
"epoch": 0.014266963419505792,
"grad_norm": 7.250610828399658,
"learning_rate": 1.426669202967472e-06,
"loss": 2.1039,
"step": 750
},
{
"epoch": 0.014457189598432537,
"grad_norm": 1.4629435539245605,
"learning_rate": 1.4456914590070384e-06,
"loss": 1.5061,
"step": 760
},
{
"epoch": 0.01464741577735928,
"grad_norm": 1.7025142908096313,
"learning_rate": 1.4647137150466046e-06,
"loss": 1.4471,
"step": 770
},
{
"epoch": 0.014837641956286025,
"grad_norm": 2.418815851211548,
"learning_rate": 1.483735971086171e-06,
"loss": 1.3384,
"step": 780
},
{
"epoch": 0.015027868135212768,
"grad_norm": 4.666192531585693,
"learning_rate": 1.5027582271257372e-06,
"loss": 1.599,
"step": 790
},
{
"epoch": 0.015218094314139513,
"grad_norm": 8.951377868652344,
"learning_rate": 1.5217804831653036e-06,
"loss": 1.9391,
"step": 800
},
{
"epoch": 0.015408320493066256,
"grad_norm": 1.383745551109314,
"learning_rate": 1.54080273920487e-06,
"loss": 1.5076,
"step": 810
},
{
"epoch": 0.015598546671992999,
"grad_norm": 1.4020535945892334,
"learning_rate": 1.5598249952444361e-06,
"loss": 1.3977,
"step": 820
},
{
"epoch": 0.015788772850919744,
"grad_norm": 2.2491354942321777,
"learning_rate": 1.5788472512840025e-06,
"loss": 1.3356,
"step": 830
},
{
"epoch": 0.015978999029846487,
"grad_norm": 3.5490710735321045,
"learning_rate": 1.5978695073235687e-06,
"loss": 1.5511,
"step": 840
},
{
"epoch": 0.01616922520877323,
"grad_norm": 11.946330070495605,
"learning_rate": 1.616891763363135e-06,
"loss": 1.8743,
"step": 850
},
{
"epoch": 0.016359451387699977,
"grad_norm": 1.4720561504364014,
"learning_rate": 1.6359140194027013e-06,
"loss": 1.4478,
"step": 860
},
{
"epoch": 0.01654967756662672,
"grad_norm": 1.6156224012374878,
"learning_rate": 1.6549362754422677e-06,
"loss": 1.4235,
"step": 870
},
{
"epoch": 0.016739903745553463,
"grad_norm": 2.14105224609375,
"learning_rate": 1.6739585314818339e-06,
"loss": 1.3183,
"step": 880
},
{
"epoch": 0.016930129924480206,
"grad_norm": 4.404923915863037,
"learning_rate": 1.6929807875214002e-06,
"loss": 1.3946,
"step": 890
},
{
"epoch": 0.01712035610340695,
"grad_norm": 7.182434558868408,
"learning_rate": 1.7120030435609664e-06,
"loss": 1.8036,
"step": 900
},
{
"epoch": 0.017310582282333696,
"grad_norm": 1.494486689567566,
"learning_rate": 1.7310252996005328e-06,
"loss": 1.4478,
"step": 910
},
{
"epoch": 0.01750080846126044,
"grad_norm": 1.6941497325897217,
"learning_rate": 1.750047555640099e-06,
"loss": 1.3241,
"step": 920
},
{
"epoch": 0.017691034640187182,
"grad_norm": 2.1194875240325928,
"learning_rate": 1.7690698116796654e-06,
"loss": 1.189,
"step": 930
},
{
"epoch": 0.017881260819113926,
"grad_norm": 2.8642826080322266,
"learning_rate": 1.7880920677192318e-06,
"loss": 1.3617,
"step": 940
},
{
"epoch": 0.01807148699804067,
"grad_norm": 9.907108306884766,
"learning_rate": 1.807114323758798e-06,
"loss": 1.7392,
"step": 950
},
{
"epoch": 0.018261713176967415,
"grad_norm": 1.627024531364441,
"learning_rate": 1.8261365797983643e-06,
"loss": 1.4536,
"step": 960
},
{
"epoch": 0.01845193935589416,
"grad_norm": 1.7732610702514648,
"learning_rate": 1.8451588358379305e-06,
"loss": 1.3351,
"step": 970
},
{
"epoch": 0.0186421655348209,
"grad_norm": 2.412036180496216,
"learning_rate": 1.864181091877497e-06,
"loss": 1.2028,
"step": 980
},
{
"epoch": 0.018832391713747645,
"grad_norm": 4.754890441894531,
"learning_rate": 1.883203347917063e-06,
"loss": 1.3517,
"step": 990
},
{
"epoch": 0.01902261789267439,
"grad_norm": 8.927350997924805,
"learning_rate": 1.9022256039566295e-06,
"loss": 1.6531,
"step": 1000
},
{
"epoch": 0.019212844071601134,
"grad_norm": 1.6348631381988525,
"learning_rate": 1.9212478599961954e-06,
"loss": 1.4934,
"step": 1010
},
{
"epoch": 0.019403070250527878,
"grad_norm": 1.9770840406417847,
"learning_rate": 1.940270116035762e-06,
"loss": 1.3571,
"step": 1020
},
{
"epoch": 0.01959329642945462,
"grad_norm": 2.404418468475342,
"learning_rate": 1.9592923720753282e-06,
"loss": 1.1291,
"step": 1030
},
{
"epoch": 0.019783522608381364,
"grad_norm": 3.4404196739196777,
"learning_rate": 1.9783146281148944e-06,
"loss": 1.2498,
"step": 1040
},
{
"epoch": 0.01997374878730811,
"grad_norm": 9.02733325958252,
"learning_rate": 1.9973368841544606e-06,
"loss": 1.5646,
"step": 1050
},
{
"epoch": 0.020163974966234854,
"grad_norm": 1.9403765201568604,
"learning_rate": 2.016359140194027e-06,
"loss": 1.3972,
"step": 1060
},
{
"epoch": 0.020354201145161597,
"grad_norm": 2.009138584136963,
"learning_rate": 2.0353813962335938e-06,
"loss": 1.303,
"step": 1070
},
{
"epoch": 0.02054442732408834,
"grad_norm": 2.1750669479370117,
"learning_rate": 2.0544036522731595e-06,
"loss": 1.1875,
"step": 1080
},
{
"epoch": 0.020734653503015087,
"grad_norm": 3.3346753120422363,
"learning_rate": 2.073425908312726e-06,
"loss": 1.262,
"step": 1090
},
{
"epoch": 0.02092487968194183,
"grad_norm": 8.939334869384766,
"learning_rate": 2.0924481643522923e-06,
"loss": 1.547,
"step": 1100
},
{
"epoch": 0.021115105860868573,
"grad_norm": 1.614304780960083,
"learning_rate": 2.111470420391859e-06,
"loss": 1.3818,
"step": 1110
},
{
"epoch": 0.021305332039795316,
"grad_norm": 1.98068106174469,
"learning_rate": 2.1304926764314247e-06,
"loss": 1.2253,
"step": 1120
},
{
"epoch": 0.02149555821872206,
"grad_norm": 2.3504743576049805,
"learning_rate": 2.1495149324709913e-06,
"loss": 1.1185,
"step": 1130
},
{
"epoch": 0.021685784397648806,
"grad_norm": 3.188448667526245,
"learning_rate": 2.1685371885105575e-06,
"loss": 1.2133,
"step": 1140
},
{
"epoch": 0.02187601057657555,
"grad_norm": 11.752523422241211,
"learning_rate": 2.187559444550124e-06,
"loss": 1.5604,
"step": 1150
},
{
"epoch": 0.022066236755502292,
"grad_norm": 1.4586597681045532,
"learning_rate": 2.20658170058969e-06,
"loss": 1.2922,
"step": 1160
},
{
"epoch": 0.022256462934429035,
"grad_norm": 1.83343505859375,
"learning_rate": 2.2256039566292564e-06,
"loss": 1.2738,
"step": 1170
},
{
"epoch": 0.02244668911335578,
"grad_norm": 2.553067684173584,
"learning_rate": 2.2446262126688226e-06,
"loss": 1.096,
"step": 1180
},
{
"epoch": 0.022636915292282525,
"grad_norm": 4.024831771850586,
"learning_rate": 2.263648468708389e-06,
"loss": 1.1661,
"step": 1190
},
{
"epoch": 0.022827141471209268,
"grad_norm": 11.252798080444336,
"learning_rate": 2.2826707247479554e-06,
"loss": 1.4848,
"step": 1200
},
{
"epoch": 0.02301736765013601,
"grad_norm": 2.0187082290649414,
"learning_rate": 2.3016929807875215e-06,
"loss": 1.2914,
"step": 1210
},
{
"epoch": 0.023207593829062755,
"grad_norm": 2.2791354656219482,
"learning_rate": 2.320715236827088e-06,
"loss": 1.2912,
"step": 1220
},
{
"epoch": 0.0233978200079895,
"grad_norm": 2.755721092224121,
"learning_rate": 2.3397374928666543e-06,
"loss": 1.1446,
"step": 1230
},
{
"epoch": 0.023588046186916244,
"grad_norm": 4.635268688201904,
"learning_rate": 2.3587597489062205e-06,
"loss": 1.156,
"step": 1240
},
{
"epoch": 0.023778272365842987,
"grad_norm": 11.180182456970215,
"learning_rate": 2.3777820049457867e-06,
"loss": 1.5128,
"step": 1250
},
{
"epoch": 0.02396849854476973,
"grad_norm": 2.6687514781951904,
"learning_rate": 2.3968042609853533e-06,
"loss": 1.3078,
"step": 1260
},
{
"epoch": 0.024158724723696474,
"grad_norm": 2.1617326736450195,
"learning_rate": 2.4158265170249195e-06,
"loss": 1.187,
"step": 1270
},
{
"epoch": 0.02434895090262322,
"grad_norm": 2.7917706966400146,
"learning_rate": 2.4348487730644856e-06,
"loss": 1.1008,
"step": 1280
},
{
"epoch": 0.024539177081549964,
"grad_norm": 3.9902455806732178,
"learning_rate": 2.453871029104052e-06,
"loss": 1.1694,
"step": 1290
},
{
"epoch": 0.024729403260476707,
"grad_norm": 10.19775104522705,
"learning_rate": 2.4728932851436184e-06,
"loss": 1.4882,
"step": 1300
},
{
"epoch": 0.02491962943940345,
"grad_norm": 1.815943956375122,
"learning_rate": 2.4919155411831846e-06,
"loss": 1.3296,
"step": 1310
},
{
"epoch": 0.025109855618330193,
"grad_norm": 1.7802958488464355,
"learning_rate": 2.5109377972227504e-06,
"loss": 1.1588,
"step": 1320
},
{
"epoch": 0.02530008179725694,
"grad_norm": 2.8108973503112793,
"learning_rate": 2.529960053262317e-06,
"loss": 1.0912,
"step": 1330
},
{
"epoch": 0.025490307976183683,
"grad_norm": 3.822803497314453,
"learning_rate": 2.5489823093018836e-06,
"loss": 1.1713,
"step": 1340
},
{
"epoch": 0.025680534155110426,
"grad_norm": 10.885619163513184,
"learning_rate": 2.5680045653414497e-06,
"loss": 1.4048,
"step": 1350
},
{
"epoch": 0.02587076033403717,
"grad_norm": 1.5321264266967773,
"learning_rate": 2.5870268213810163e-06,
"loss": 1.2555,
"step": 1360
},
{
"epoch": 0.026060986512963916,
"grad_norm": 2.115370512008667,
"learning_rate": 2.606049077420582e-06,
"loss": 1.1987,
"step": 1370
},
{
"epoch": 0.02625121269189066,
"grad_norm": 2.6233675479888916,
"learning_rate": 2.6250713334601487e-06,
"loss": 1.0758,
"step": 1380
},
{
"epoch": 0.026441438870817402,
"grad_norm": 3.8407323360443115,
"learning_rate": 2.644093589499715e-06,
"loss": 1.1887,
"step": 1390
},
{
"epoch": 0.026631665049744145,
"grad_norm": 12.603764533996582,
"learning_rate": 2.6631158455392815e-06,
"loss": 1.4082,
"step": 1400
},
{
"epoch": 0.026821891228670888,
"grad_norm": 1.6481189727783203,
"learning_rate": 2.6821381015788472e-06,
"loss": 1.2527,
"step": 1410
},
{
"epoch": 0.027012117407597635,
"grad_norm": 2.1282284259796143,
"learning_rate": 2.701160357618414e-06,
"loss": 1.2139,
"step": 1420
},
{
"epoch": 0.027202343586524378,
"grad_norm": 2.9524528980255127,
"learning_rate": 2.72018261365798e-06,
"loss": 1.09,
"step": 1430
},
{
"epoch": 0.02739256976545112,
"grad_norm": 4.275539398193359,
"learning_rate": 2.7392048696975466e-06,
"loss": 1.1149,
"step": 1440
},
{
"epoch": 0.027582795944377864,
"grad_norm": 12.224852561950684,
"learning_rate": 2.7582271257371124e-06,
"loss": 1.3974,
"step": 1450
},
{
"epoch": 0.027773022123304607,
"grad_norm": 2.3314363956451416,
"learning_rate": 2.777249381776679e-06,
"loss": 1.3061,
"step": 1460
},
{
"epoch": 0.027963248302231354,
"grad_norm": 2.170639991760254,
"learning_rate": 2.796271637816245e-06,
"loss": 1.2254,
"step": 1470
},
{
"epoch": 0.028153474481158097,
"grad_norm": 2.8608880043029785,
"learning_rate": 2.8152938938558117e-06,
"loss": 1.0807,
"step": 1480
},
{
"epoch": 0.02834370066008484,
"grad_norm": 3.440014600753784,
"learning_rate": 2.834316149895378e-06,
"loss": 1.1255,
"step": 1490
},
{
"epoch": 0.028533926839011584,
"grad_norm": 10.175577163696289,
"learning_rate": 2.853338405934944e-06,
"loss": 1.3387,
"step": 1500
},
{
"epoch": 0.02872415301793833,
"grad_norm": 2.22709321975708,
"learning_rate": 2.8723606619745103e-06,
"loss": 1.2654,
"step": 1510
},
{
"epoch": 0.028914379196865073,
"grad_norm": 1.8235417604446411,
"learning_rate": 2.891382918014077e-06,
"loss": 1.2287,
"step": 1520
},
{
"epoch": 0.029104605375791816,
"grad_norm": 2.5544168949127197,
"learning_rate": 2.910405174053643e-06,
"loss": 0.9776,
"step": 1530
},
{
"epoch": 0.02929483155471856,
"grad_norm": 4.163671016693115,
"learning_rate": 2.9294274300932092e-06,
"loss": 1.0837,
"step": 1540
},
{
"epoch": 0.029485057733645303,
"grad_norm": 10.698822021484375,
"learning_rate": 2.9484496861327754e-06,
"loss": 1.3228,
"step": 1550
},
{
"epoch": 0.02967528391257205,
"grad_norm": 1.6194937229156494,
"learning_rate": 2.967471942172342e-06,
"loss": 1.2534,
"step": 1560
},
{
"epoch": 0.029865510091498793,
"grad_norm": 1.9834601879119873,
"learning_rate": 2.986494198211908e-06,
"loss": 1.201,
"step": 1570
},
{
"epoch": 0.030055736270425536,
"grad_norm": 2.591292142868042,
"learning_rate": 3.0055164542514744e-06,
"loss": 1.0589,
"step": 1580
},
{
"epoch": 0.03024596244935228,
"grad_norm": 4.2067670822143555,
"learning_rate": 3.0245387102910406e-06,
"loss": 1.0603,
"step": 1590
},
{
"epoch": 0.030436188628279025,
"grad_norm": 9.938916206359863,
"learning_rate": 3.043560966330607e-06,
"loss": 1.2619,
"step": 1600
},
{
"epoch": 0.03062641480720577,
"grad_norm": 2.7832484245300293,
"learning_rate": 3.0625832223701733e-06,
"loss": 1.2533,
"step": 1610
},
{
"epoch": 0.030816640986132512,
"grad_norm": 2.117129325866699,
"learning_rate": 3.08160547840974e-06,
"loss": 1.1825,
"step": 1620
},
{
"epoch": 0.031006867165059255,
"grad_norm": 3.0919687747955322,
"learning_rate": 3.1006277344493057e-06,
"loss": 1.095,
"step": 1630
},
{
"epoch": 0.031197093343985998,
"grad_norm": 4.166520595550537,
"learning_rate": 3.1196499904888723e-06,
"loss": 1.047,
"step": 1640
},
{
"epoch": 0.031387319522912745,
"grad_norm": 12.0921630859375,
"learning_rate": 3.1386722465284385e-06,
"loss": 1.3526,
"step": 1650
},
{
"epoch": 0.03157754570183949,
"grad_norm": 1.9751317501068115,
"learning_rate": 3.157694502568005e-06,
"loss": 1.2864,
"step": 1660
},
{
"epoch": 0.03176777188076623,
"grad_norm": 2.2880234718322754,
"learning_rate": 3.176716758607571e-06,
"loss": 1.1079,
"step": 1670
},
{
"epoch": 0.031957998059692974,
"grad_norm": 2.804194688796997,
"learning_rate": 3.1957390146471374e-06,
"loss": 0.9666,
"step": 1680
},
{
"epoch": 0.03214822423861972,
"grad_norm": 4.810030937194824,
"learning_rate": 3.2147612706867036e-06,
"loss": 1.0979,
"step": 1690
},
{
"epoch": 0.03233845041754646,
"grad_norm": 9.718046188354492,
"learning_rate": 3.23378352672627e-06,
"loss": 1.2383,
"step": 1700
},
{
"epoch": 0.032528676596473204,
"grad_norm": 2.0624125003814697,
"learning_rate": 3.252805782765836e-06,
"loss": 1.2589,
"step": 1710
},
{
"epoch": 0.032718902775399954,
"grad_norm": 2.8751795291900635,
"learning_rate": 3.2718280388054026e-06,
"loss": 1.1319,
"step": 1720
},
{
"epoch": 0.0329091289543267,
"grad_norm": 2.805138349533081,
"learning_rate": 3.2908502948449687e-06,
"loss": 0.9922,
"step": 1730
},
{
"epoch": 0.03309935513325344,
"grad_norm": 4.693927764892578,
"learning_rate": 3.3098725508845353e-06,
"loss": 1.0518,
"step": 1740
},
{
"epoch": 0.03328958131218018,
"grad_norm": 9.922603607177734,
"learning_rate": 3.3288948069241015e-06,
"loss": 1.2417,
"step": 1750
},
{
"epoch": 0.033479807491106926,
"grad_norm": 1.9441560506820679,
"learning_rate": 3.3479170629636677e-06,
"loss": 1.2156,
"step": 1760
},
{
"epoch": 0.03367003367003367,
"grad_norm": 2.1576952934265137,
"learning_rate": 3.366939319003234e-06,
"loss": 1.1009,
"step": 1770
},
{
"epoch": 0.03386025984896041,
"grad_norm": 2.93638014793396,
"learning_rate": 3.3859615750428005e-06,
"loss": 0.9806,
"step": 1780
},
{
"epoch": 0.034050486027887156,
"grad_norm": 4.772079944610596,
"learning_rate": 3.4049838310823667e-06,
"loss": 1.0505,
"step": 1790
},
{
"epoch": 0.0342407122068139,
"grad_norm": 13.75841236114502,
"learning_rate": 3.424006087121933e-06,
"loss": 1.2653,
"step": 1800
},
{
"epoch": 0.03443093838574065,
"grad_norm": 2.011861801147461,
"learning_rate": 3.443028343161499e-06,
"loss": 1.1962,
"step": 1810
},
{
"epoch": 0.03462116456466739,
"grad_norm": 2.2049922943115234,
"learning_rate": 3.4620505992010656e-06,
"loss": 1.1636,
"step": 1820
},
{
"epoch": 0.034811390743594135,
"grad_norm": 3.468641757965088,
"learning_rate": 3.481072855240632e-06,
"loss": 1.006,
"step": 1830
},
{
"epoch": 0.03500161692252088,
"grad_norm": 5.194962024688721,
"learning_rate": 3.500095111280198e-06,
"loss": 1.035,
"step": 1840
},
{
"epoch": 0.03519184310144762,
"grad_norm": 9.455286979675293,
"learning_rate": 3.519117367319764e-06,
"loss": 1.1639,
"step": 1850
},
{
"epoch": 0.035382069280374365,
"grad_norm": 2.2892448902130127,
"learning_rate": 3.5381396233593308e-06,
"loss": 1.2648,
"step": 1860
},
{
"epoch": 0.03557229545930111,
"grad_norm": 2.668325185775757,
"learning_rate": 3.557161879398897e-06,
"loss": 1.1385,
"step": 1870
},
{
"epoch": 0.03576252163822785,
"grad_norm": 2.897562265396118,
"learning_rate": 3.5761841354384635e-06,
"loss": 0.9518,
"step": 1880
},
{
"epoch": 0.035952747817154594,
"grad_norm": 5.355794906616211,
"learning_rate": 3.5952063914780293e-06,
"loss": 1.059,
"step": 1890
},
{
"epoch": 0.03614297399608134,
"grad_norm": 11.731189727783203,
"learning_rate": 3.614228647517596e-06,
"loss": 1.2406,
"step": 1900
},
{
"epoch": 0.03633320017500809,
"grad_norm": 2.4913065433502197,
"learning_rate": 3.633250903557162e-06,
"loss": 1.1554,
"step": 1910
},
{
"epoch": 0.03652342635393483,
"grad_norm": 2.4660778045654297,
"learning_rate": 3.6522731595967287e-06,
"loss": 1.1604,
"step": 1920
},
{
"epoch": 0.036713652532861574,
"grad_norm": 2.8727564811706543,
"learning_rate": 3.6712954156362944e-06,
"loss": 0.986,
"step": 1930
},
{
"epoch": 0.03690387871178832,
"grad_norm": 5.380633354187012,
"learning_rate": 3.690317671675861e-06,
"loss": 0.9519,
"step": 1940
},
{
"epoch": 0.03709410489071506,
"grad_norm": 9.196549415588379,
"learning_rate": 3.709339927715427e-06,
"loss": 1.253,
"step": 1950
},
{
"epoch": 0.0372843310696418,
"grad_norm": 1.7812504768371582,
"learning_rate": 3.728362183754994e-06,
"loss": 1.2104,
"step": 1960
},
{
"epoch": 0.037474557248568546,
"grad_norm": 2.1316113471984863,
"learning_rate": 3.7473844397945596e-06,
"loss": 1.1207,
"step": 1970
},
{
"epoch": 0.03766478342749529,
"grad_norm": 2.813387155532837,
"learning_rate": 3.766406695834126e-06,
"loss": 0.9173,
"step": 1980
},
{
"epoch": 0.03785500960642203,
"grad_norm": 4.0899128913879395,
"learning_rate": 3.7854289518736923e-06,
"loss": 0.9493,
"step": 1990
},
{
"epoch": 0.03804523578534878,
"grad_norm": 10.877523422241211,
"learning_rate": 3.804451207913259e-06,
"loss": 1.2015,
"step": 2000
},
{
"epoch": 0.038235461964275526,
"grad_norm": 2.7726292610168457,
"learning_rate": 3.8234734639528255e-06,
"loss": 1.2061,
"step": 2010
},
{
"epoch": 0.03842568814320227,
"grad_norm": 2.1731584072113037,
"learning_rate": 3.842495719992391e-06,
"loss": 1.106,
"step": 2020
},
{
"epoch": 0.03861591432212901,
"grad_norm": 3.7517712116241455,
"learning_rate": 3.861517976031958e-06,
"loss": 0.9798,
"step": 2030
},
{
"epoch": 0.038806140501055755,
"grad_norm": 5.391069412231445,
"learning_rate": 3.880540232071524e-06,
"loss": 0.9718,
"step": 2040
},
{
"epoch": 0.0389963666799825,
"grad_norm": 9.744902610778809,
"learning_rate": 3.89956248811109e-06,
"loss": 1.2262,
"step": 2050
},
{
"epoch": 0.03918659285890924,
"grad_norm": 2.6106269359588623,
"learning_rate": 3.9185847441506564e-06,
"loss": 1.2071,
"step": 2060
},
{
"epoch": 0.039376819037835985,
"grad_norm": 3.034111738204956,
"learning_rate": 3.937607000190223e-06,
"loss": 1.1291,
"step": 2070
},
{
"epoch": 0.03956704521676273,
"grad_norm": 3.5086514949798584,
"learning_rate": 3.956629256229789e-06,
"loss": 0.9248,
"step": 2080
},
{
"epoch": 0.03975727139568948,
"grad_norm": 4.488050937652588,
"learning_rate": 3.975651512269356e-06,
"loss": 1.0084,
"step": 2090
},
{
"epoch": 0.03994749757461622,
"grad_norm": 10.10869312286377,
"learning_rate": 3.994673768308921e-06,
"loss": 1.188,
"step": 2100
},
{
"epoch": 0.040137723753542964,
"grad_norm": 2.787724256515503,
"learning_rate": 4.013696024348488e-06,
"loss": 1.153,
"step": 2110
},
{
"epoch": 0.04032794993246971,
"grad_norm": 2.309357166290283,
"learning_rate": 4.032718280388054e-06,
"loss": 1.0842,
"step": 2120
},
{
"epoch": 0.04051817611139645,
"grad_norm": 2.9359734058380127,
"learning_rate": 4.0517405364276205e-06,
"loss": 0.9549,
"step": 2130
},
{
"epoch": 0.040708402290323194,
"grad_norm": 4.532121658325195,
"learning_rate": 4.0707627924671876e-06,
"loss": 0.9548,
"step": 2140
},
{
"epoch": 0.04089862846924994,
"grad_norm": 9.86694049835205,
"learning_rate": 4.089785048506753e-06,
"loss": 1.1928,
"step": 2150
},
{
"epoch": 0.04108885464817668,
"grad_norm": 2.4668540954589844,
"learning_rate": 4.108807304546319e-06,
"loss": 1.2318,
"step": 2160
},
{
"epoch": 0.04127908082710342,
"grad_norm": 2.501939296722412,
"learning_rate": 4.127829560585886e-06,
"loss": 1.11,
"step": 2170
},
{
"epoch": 0.04146930700603017,
"grad_norm": 2.8509225845336914,
"learning_rate": 4.146851816625452e-06,
"loss": 0.8994,
"step": 2180
},
{
"epoch": 0.041659533184956916,
"grad_norm": 4.271004676818848,
"learning_rate": 4.1658740726650185e-06,
"loss": 0.9308,
"step": 2190
},
{
"epoch": 0.04184975936388366,
"grad_norm": 8.929543495178223,
"learning_rate": 4.184896328704585e-06,
"loss": 1.1605,
"step": 2200
},
{
"epoch": 0.0420399855428104,
"grad_norm": 2.176539421081543,
"learning_rate": 4.203918584744151e-06,
"loss": 1.1261,
"step": 2210
},
{
"epoch": 0.042230211721737146,
"grad_norm": 3.0560004711151123,
"learning_rate": 4.222940840783718e-06,
"loss": 1.1041,
"step": 2220
},
{
"epoch": 0.04242043790066389,
"grad_norm": 2.956686496734619,
"learning_rate": 4.241963096823283e-06,
"loss": 0.9188,
"step": 2230
},
{
"epoch": 0.04261066407959063,
"grad_norm": 4.251719951629639,
"learning_rate": 4.260985352862849e-06,
"loss": 1.0,
"step": 2240
},
{
"epoch": 0.042800890258517375,
"grad_norm": 11.712019920349121,
"learning_rate": 4.280007608902416e-06,
"loss": 1.1977,
"step": 2250
},
{
"epoch": 0.04299111643744412,
"grad_norm": 3.014216899871826,
"learning_rate": 4.2990298649419825e-06,
"loss": 1.175,
"step": 2260
},
{
"epoch": 0.04318134261637086,
"grad_norm": 2.524491786956787,
"learning_rate": 4.318052120981549e-06,
"loss": 1.0855,
"step": 2270
},
{
"epoch": 0.04337156879529761,
"grad_norm": 3.5122768878936768,
"learning_rate": 4.337074377021115e-06,
"loss": 0.8957,
"step": 2280
},
{
"epoch": 0.043561794974224355,
"grad_norm": 5.294569492340088,
"learning_rate": 4.356096633060681e-06,
"loss": 0.9801,
"step": 2290
},
{
"epoch": 0.0437520211531511,
"grad_norm": 12.318851470947266,
"learning_rate": 4.375118889100248e-06,
"loss": 1.2072,
"step": 2300
},
{
"epoch": 0.04394224733207784,
"grad_norm": 3.4695818424224854,
"learning_rate": 4.394141145139814e-06,
"loss": 1.1855,
"step": 2310
},
{
"epoch": 0.044132473511004584,
"grad_norm": 2.193481683731079,
"learning_rate": 4.41316340117938e-06,
"loss": 1.0368,
"step": 2320
},
{
"epoch": 0.04432269968993133,
"grad_norm": 3.3650193214416504,
"learning_rate": 4.432185657218947e-06,
"loss": 0.9339,
"step": 2330
},
{
"epoch": 0.04451292586885807,
"grad_norm": 4.817994594573975,
"learning_rate": 4.451207913258513e-06,
"loss": 0.9878,
"step": 2340
},
{
"epoch": 0.044703152047784814,
"grad_norm": 11.410785675048828,
"learning_rate": 4.470230169298079e-06,
"loss": 1.1553,
"step": 2350
},
{
"epoch": 0.04489337822671156,
"grad_norm": 2.329460620880127,
"learning_rate": 4.489252425337645e-06,
"loss": 1.1137,
"step": 2360
},
{
"epoch": 0.04508360440563831,
"grad_norm": 4.1374287605285645,
"learning_rate": 4.508274681377211e-06,
"loss": 1.0883,
"step": 2370
},
{
"epoch": 0.04527383058456505,
"grad_norm": 3.188446521759033,
"learning_rate": 4.527296937416778e-06,
"loss": 0.9459,
"step": 2380
},
{
"epoch": 0.04546405676349179,
"grad_norm": 6.629786014556885,
"learning_rate": 4.5463191934563446e-06,
"loss": 0.9519,
"step": 2390
},
{
"epoch": 0.045654282942418536,
"grad_norm": 12.22085189819336,
"learning_rate": 4.565341449495911e-06,
"loss": 1.1074,
"step": 2400
},
{
"epoch": 0.04584450912134528,
"grad_norm": 3.0764682292938232,
"learning_rate": 4.584363705535477e-06,
"loss": 1.1458,
"step": 2410
},
{
"epoch": 0.04603473530027202,
"grad_norm": 2.767094135284424,
"learning_rate": 4.603385961575043e-06,
"loss": 1.0658,
"step": 2420
},
{
"epoch": 0.046224961479198766,
"grad_norm": 3.757636308670044,
"learning_rate": 4.622408217614609e-06,
"loss": 0.9337,
"step": 2430
},
{
"epoch": 0.04641518765812551,
"grad_norm": 5.4339447021484375,
"learning_rate": 4.641430473654176e-06,
"loss": 0.9154,
"step": 2440
},
{
"epoch": 0.04660541383705225,
"grad_norm": 11.860993385314941,
"learning_rate": 4.660452729693742e-06,
"loss": 1.1127,
"step": 2450
},
{
"epoch": 0.046795640015979,
"grad_norm": 2.824115037918091,
"learning_rate": 4.679474985733309e-06,
"loss": 1.1366,
"step": 2460
},
{
"epoch": 0.046985866194905745,
"grad_norm": 2.538586139678955,
"learning_rate": 4.698497241772875e-06,
"loss": 1.0307,
"step": 2470
},
{
"epoch": 0.04717609237383249,
"grad_norm": 3.817042589187622,
"learning_rate": 4.717519497812441e-06,
"loss": 0.9,
"step": 2480
},
{
"epoch": 0.04736631855275923,
"grad_norm": 5.339418888092041,
"learning_rate": 4.736541753852007e-06,
"loss": 0.9951,
"step": 2490
},
{
"epoch": 0.047556544731685975,
"grad_norm": 12.489927291870117,
"learning_rate": 4.755564009891573e-06,
"loss": 1.1541,
"step": 2500
},
{
"epoch": 0.04774677091061272,
"grad_norm": 2.1420531272888184,
"learning_rate": 4.7745862659311395e-06,
"loss": 1.1063,
"step": 2510
},
{
"epoch": 0.04793699708953946,
"grad_norm": 2.2218401432037354,
"learning_rate": 4.7936085219707066e-06,
"loss": 1.0871,
"step": 2520
},
{
"epoch": 0.048127223268466204,
"grad_norm": 3.700441360473633,
"learning_rate": 4.812630778010273e-06,
"loss": 0.873,
"step": 2530
},
{
"epoch": 0.04831744944739295,
"grad_norm": 5.405223369598389,
"learning_rate": 4.831653034049839e-06,
"loss": 0.985,
"step": 2540
},
{
"epoch": 0.04850767562631969,
"grad_norm": 11.029736518859863,
"learning_rate": 4.850675290089405e-06,
"loss": 1.2141,
"step": 2550
},
{
"epoch": 0.04869790180524644,
"grad_norm": 2.7142674922943115,
"learning_rate": 4.869697546128971e-06,
"loss": 1.1559,
"step": 2560
},
{
"epoch": 0.048888127984173184,
"grad_norm": 2.459038734436035,
"learning_rate": 4.8887198021685375e-06,
"loss": 1.1262,
"step": 2570
},
{
"epoch": 0.04907835416309993,
"grad_norm": 3.2301881313323975,
"learning_rate": 4.907742058208104e-06,
"loss": 0.9321,
"step": 2580
},
{
"epoch": 0.04926858034202667,
"grad_norm": 5.523041248321533,
"learning_rate": 4.92676431424767e-06,
"loss": 0.9226,
"step": 2590
},
{
"epoch": 0.04945880652095341,
"grad_norm": 10.131287574768066,
"learning_rate": 4.945786570287237e-06,
"loss": 1.1409,
"step": 2600
},
{
"epoch": 0.049649032699880156,
"grad_norm": 2.8806002140045166,
"learning_rate": 4.964808826326803e-06,
"loss": 1.1364,
"step": 2610
},
{
"epoch": 0.0498392588788069,
"grad_norm": 2.400009870529175,
"learning_rate": 4.983831082366369e-06,
"loss": 1.0847,
"step": 2620
},
{
"epoch": 0.05002948505773364,
"grad_norm": 4.137524604797363,
"learning_rate": 5.002853338405935e-06,
"loss": 0.9026,
"step": 2630
},
{
"epoch": 0.050219711236660386,
"grad_norm": 4.834102153778076,
"learning_rate": 5.021875594445501e-06,
"loss": 0.936,
"step": 2640
},
{
"epoch": 0.050409937415587136,
"grad_norm": 10.480986595153809,
"learning_rate": 5.040897850485068e-06,
"loss": 1.1391,
"step": 2650
},
{
"epoch": 0.05060016359451388,
"grad_norm": 2.876462697982788,
"learning_rate": 5.059920106524634e-06,
"loss": 1.1186,
"step": 2660
},
{
"epoch": 0.05079038977344062,
"grad_norm": 3.254173755645752,
"learning_rate": 5.078942362564201e-06,
"loss": 1.0468,
"step": 2670
},
{
"epoch": 0.050980615952367365,
"grad_norm": 3.5466344356536865,
"learning_rate": 5.097964618603767e-06,
"loss": 0.8944,
"step": 2680
},
{
"epoch": 0.05117084213129411,
"grad_norm": 4.0647382736206055,
"learning_rate": 5.1169868746433324e-06,
"loss": 0.8855,
"step": 2690
},
{
"epoch": 0.05136106831022085,
"grad_norm": 12.590734481811523,
"learning_rate": 5.1360091306828995e-06,
"loss": 1.1354,
"step": 2700
},
{
"epoch": 0.051551294489147595,
"grad_norm": 2.858893632888794,
"learning_rate": 5.155031386722466e-06,
"loss": 1.1345,
"step": 2710
},
{
"epoch": 0.05174152066807434,
"grad_norm": 2.8196043968200684,
"learning_rate": 5.174053642762033e-06,
"loss": 1.0826,
"step": 2720
},
{
"epoch": 0.05193174684700108,
"grad_norm": 3.493431806564331,
"learning_rate": 5.193075898801598e-06,
"loss": 0.8708,
"step": 2730
},
{
"epoch": 0.05212197302592783,
"grad_norm": 5.68046236038208,
"learning_rate": 5.212098154841164e-06,
"loss": 0.937,
"step": 2740
},
{
"epoch": 0.052312199204854574,
"grad_norm": 13.049890518188477,
"learning_rate": 5.231120410880731e-06,
"loss": 1.1392,
"step": 2750
},
{
"epoch": 0.05250242538378132,
"grad_norm": 4.613514423370361,
"learning_rate": 5.250142666920297e-06,
"loss": 1.1284,
"step": 2760
},
{
"epoch": 0.05269265156270806,
"grad_norm": 2.3258795738220215,
"learning_rate": 5.269164922959863e-06,
"loss": 1.0144,
"step": 2770
},
{
"epoch": 0.052882877741634804,
"grad_norm": 3.8380930423736572,
"learning_rate": 5.28818717899943e-06,
"loss": 0.9067,
"step": 2780
},
{
"epoch": 0.05307310392056155,
"grad_norm": 5.028414249420166,
"learning_rate": 5.307209435038996e-06,
"loss": 0.9445,
"step": 2790
},
{
"epoch": 0.05326333009948829,
"grad_norm": 9.695432662963867,
"learning_rate": 5.326231691078563e-06,
"loss": 1.0988,
"step": 2800
},
{
"epoch": 0.05345355627841503,
"grad_norm": 3.135622262954712,
"learning_rate": 5.345253947118128e-06,
"loss": 1.1195,
"step": 2810
},
{
"epoch": 0.053643782457341777,
"grad_norm": 3.463083505630493,
"learning_rate": 5.3642762031576945e-06,
"loss": 1.0298,
"step": 2820
},
{
"epoch": 0.05383400863626853,
"grad_norm": 2.976191997528076,
"learning_rate": 5.3832984591972615e-06,
"loss": 0.8963,
"step": 2830
},
{
"epoch": 0.05402423481519527,
"grad_norm": 5.485313892364502,
"learning_rate": 5.402320715236828e-06,
"loss": 0.9498,
"step": 2840
},
{
"epoch": 0.05421446099412201,
"grad_norm": 12.000988006591797,
"learning_rate": 5.421342971276395e-06,
"loss": 1.1117,
"step": 2850
},
{
"epoch": 0.054404687173048756,
"grad_norm": 3.293025493621826,
"learning_rate": 5.44036522731596e-06,
"loss": 1.1391,
"step": 2860
},
{
"epoch": 0.0545949133519755,
"grad_norm": 3.216097354888916,
"learning_rate": 5.459387483355526e-06,
"loss": 1.0316,
"step": 2870
},
{
"epoch": 0.05478513953090224,
"grad_norm": 3.772177219390869,
"learning_rate": 5.478409739395093e-06,
"loss": 0.8565,
"step": 2880
},
{
"epoch": 0.054975365709828986,
"grad_norm": 4.893718242645264,
"learning_rate": 5.4974319954346586e-06,
"loss": 0.8826,
"step": 2890
},
{
"epoch": 0.05516559188875573,
"grad_norm": 12.316627502441406,
"learning_rate": 5.516454251474225e-06,
"loss": 1.1345,
"step": 2900
},
{
"epoch": 0.05535581806768247,
"grad_norm": 2.428457021713257,
"learning_rate": 5.535476507513792e-06,
"loss": 1.1332,
"step": 2910
},
{
"epoch": 0.055546044246609215,
"grad_norm": 2.9872822761535645,
"learning_rate": 5.554498763553358e-06,
"loss": 1.0649,
"step": 2920
},
{
"epoch": 0.055736270425535965,
"grad_norm": 3.698500156402588,
"learning_rate": 5.573521019592925e-06,
"loss": 0.9102,
"step": 2930
},
{
"epoch": 0.05592649660446271,
"grad_norm": 5.477416515350342,
"learning_rate": 5.59254327563249e-06,
"loss": 0.9307,
"step": 2940
},
{
"epoch": 0.05611672278338945,
"grad_norm": 13.570015907287598,
"learning_rate": 5.6115655316720565e-06,
"loss": 1.1574,
"step": 2950
},
{
"epoch": 0.056306948962316195,
"grad_norm": 4.1514434814453125,
"learning_rate": 5.6305877877116235e-06,
"loss": 1.0952,
"step": 2960
},
{
"epoch": 0.05649717514124294,
"grad_norm": 2.7112059593200684,
"learning_rate": 5.649610043751189e-06,
"loss": 1.1087,
"step": 2970
},
{
"epoch": 0.05668740132016968,
"grad_norm": 3.348348617553711,
"learning_rate": 5.668632299790756e-06,
"loss": 0.893,
"step": 2980
},
{
"epoch": 0.056877627499096424,
"grad_norm": 5.3222126960754395,
"learning_rate": 5.687654555830322e-06,
"loss": 0.9064,
"step": 2990
},
{
"epoch": 0.05706785367802317,
"grad_norm": 11.816397666931152,
"learning_rate": 5.706676811869888e-06,
"loss": 1.0827,
"step": 3000
},
{
"epoch": 0.05725807985694991,
"grad_norm": 2.0702426433563232,
"learning_rate": 5.725699067909455e-06,
"loss": 1.144,
"step": 3010
},
{
"epoch": 0.05744830603587666,
"grad_norm": 3.0293846130371094,
"learning_rate": 5.7447213239490206e-06,
"loss": 1.0186,
"step": 3020
},
{
"epoch": 0.057638532214803403,
"grad_norm": 3.4096548557281494,
"learning_rate": 5.763743579988587e-06,
"loss": 0.8879,
"step": 3030
},
{
"epoch": 0.05782875839373015,
"grad_norm": 6.768439769744873,
"learning_rate": 5.782765836028154e-06,
"loss": 0.8991,
"step": 3040
},
{
"epoch": 0.05801898457265689,
"grad_norm": 14.614982604980469,
"learning_rate": 5.801788092067719e-06,
"loss": 1.0877,
"step": 3050
},
{
"epoch": 0.05820921075158363,
"grad_norm": 2.3279948234558105,
"learning_rate": 5.820810348107286e-06,
"loss": 1.081,
"step": 3060
},
{
"epoch": 0.058399436930510376,
"grad_norm": 3.4593429565429688,
"learning_rate": 5.839832604146852e-06,
"loss": 1.0231,
"step": 3070
},
{
"epoch": 0.05858966310943712,
"grad_norm": 3.165771484375,
"learning_rate": 5.8588548601864185e-06,
"loss": 0.8792,
"step": 3080
},
{
"epoch": 0.05877988928836386,
"grad_norm": 5.759438991546631,
"learning_rate": 5.8778771162259855e-06,
"loss": 0.9131,
"step": 3090
},
{
"epoch": 0.058970115467290606,
"grad_norm": 10.794733047485352,
"learning_rate": 5.896899372265551e-06,
"loss": 1.0556,
"step": 3100
},
{
"epoch": 0.059160341646217356,
"grad_norm": 2.0568151473999023,
"learning_rate": 5.915921628305118e-06,
"loss": 1.0646,
"step": 3110
},
{
"epoch": 0.0593505678251441,
"grad_norm": 3.4895856380462646,
"learning_rate": 5.934943884344684e-06,
"loss": 1.0352,
"step": 3120
},
{
"epoch": 0.05954079400407084,
"grad_norm": 3.709806442260742,
"learning_rate": 5.953966140384249e-06,
"loss": 0.8928,
"step": 3130
},
{
"epoch": 0.059731020182997585,
"grad_norm": 4.89426326751709,
"learning_rate": 5.972988396423816e-06,
"loss": 0.881,
"step": 3140
},
{
"epoch": 0.05992124636192433,
"grad_norm": 13.595807075500488,
"learning_rate": 5.992010652463383e-06,
"loss": 1.0825,
"step": 3150
},
{
"epoch": 0.06011147254085107,
"grad_norm": 3.318986177444458,
"learning_rate": 6.011032908502949e-06,
"loss": 1.1474,
"step": 3160
},
{
"epoch": 0.060301698719777815,
"grad_norm": 3.450166940689087,
"learning_rate": 6.030055164542516e-06,
"loss": 1.0437,
"step": 3170
},
{
"epoch": 0.06049192489870456,
"grad_norm": 3.8872604370117188,
"learning_rate": 6.049077420582081e-06,
"loss": 0.8966,
"step": 3180
},
{
"epoch": 0.0606821510776313,
"grad_norm": 4.919650554656982,
"learning_rate": 6.068099676621648e-06,
"loss": 0.9155,
"step": 3190
},
{
"epoch": 0.06087237725655805,
"grad_norm": 12.513046264648438,
"learning_rate": 6.087121932661214e-06,
"loss": 1.0846,
"step": 3200
},
{
"epoch": 0.061062603435484794,
"grad_norm": 3.5747430324554443,
"learning_rate": 6.10614418870078e-06,
"loss": 1.1239,
"step": 3210
},
{
"epoch": 0.06125282961441154,
"grad_norm": 2.480370044708252,
"learning_rate": 6.125166444740347e-06,
"loss": 1.0142,
"step": 3220
},
{
"epoch": 0.06144305579333828,
"grad_norm": 3.4333581924438477,
"learning_rate": 6.144188700779913e-06,
"loss": 0.9134,
"step": 3230
},
{
"epoch": 0.061633281972265024,
"grad_norm": 5.920953750610352,
"learning_rate": 6.16321095681948e-06,
"loss": 0.9426,
"step": 3240
},
{
"epoch": 0.06182350815119177,
"grad_norm": 11.068296432495117,
"learning_rate": 6.182233212859046e-06,
"loss": 0.9753,
"step": 3250
},
{
"epoch": 0.06201373433011851,
"grad_norm": 2.670910120010376,
"learning_rate": 6.201255468898611e-06,
"loss": 1.0863,
"step": 3260
},
{
"epoch": 0.06220396050904525,
"grad_norm": 2.3734078407287598,
"learning_rate": 6.220277724938178e-06,
"loss": 1.0623,
"step": 3270
},
{
"epoch": 0.062394186687971996,
"grad_norm": 5.874599933624268,
"learning_rate": 6.239299980977745e-06,
"loss": 0.8873,
"step": 3280
},
{
"epoch": 0.06258441286689874,
"grad_norm": 6.023142337799072,
"learning_rate": 6.25832223701731e-06,
"loss": 0.8765,
"step": 3290
},
{
"epoch": 0.06277463904582549,
"grad_norm": 12.106161117553711,
"learning_rate": 6.277344493056877e-06,
"loss": 1.0472,
"step": 3300
},
{
"epoch": 0.06296486522475223,
"grad_norm": 2.7873780727386475,
"learning_rate": 6.296366749096443e-06,
"loss": 1.0871,
"step": 3310
},
{
"epoch": 0.06315509140367898,
"grad_norm": 2.410355806350708,
"learning_rate": 6.31538900513601e-06,
"loss": 1.0264,
"step": 3320
},
{
"epoch": 0.06334531758260571,
"grad_norm": 3.43695068359375,
"learning_rate": 6.334411261175576e-06,
"loss": 0.8591,
"step": 3330
},
{
"epoch": 0.06353554376153246,
"grad_norm": 5.210920810699463,
"learning_rate": 6.353433517215142e-06,
"loss": 0.9006,
"step": 3340
},
{
"epoch": 0.06372576994045921,
"grad_norm": 13.646373748779297,
"learning_rate": 6.372455773254709e-06,
"loss": 1.039,
"step": 3350
},
{
"epoch": 0.06391599611938595,
"grad_norm": 3.9488167762756348,
"learning_rate": 6.391478029294275e-06,
"loss": 1.1264,
"step": 3360
},
{
"epoch": 0.0641062222983127,
"grad_norm": 2.9603095054626465,
"learning_rate": 6.410500285333842e-06,
"loss": 1.0785,
"step": 3370
},
{
"epoch": 0.06429644847723943,
"grad_norm": 6.125018119812012,
"learning_rate": 6.429522541373407e-06,
"loss": 0.8835,
"step": 3380
},
{
"epoch": 0.06448667465616618,
"grad_norm": 6.037325382232666,
"learning_rate": 6.448544797412973e-06,
"loss": 0.9131,
"step": 3390
},
{
"epoch": 0.06467690083509292,
"grad_norm": 16.08112144470215,
"learning_rate": 6.46756705345254e-06,
"loss": 1.111,
"step": 3400
},
{
"epoch": 0.06486712701401967,
"grad_norm": 2.1979708671569824,
"learning_rate": 6.486589309492107e-06,
"loss": 1.0869,
"step": 3410
},
{
"epoch": 0.06505735319294641,
"grad_norm": 3.191075086593628,
"learning_rate": 6.505611565531672e-06,
"loss": 1.0359,
"step": 3420
},
{
"epoch": 0.06524757937187316,
"grad_norm": 4.361050128936768,
"learning_rate": 6.524633821571239e-06,
"loss": 0.8945,
"step": 3430
},
{
"epoch": 0.06543780555079991,
"grad_norm": 5.246515274047852,
"learning_rate": 6.543656077610805e-06,
"loss": 0.898,
"step": 3440
},
{
"epoch": 0.06562803172972664,
"grad_norm": 11.343035697937012,
"learning_rate": 6.562678333650372e-06,
"loss": 1.0329,
"step": 3450
},
{
"epoch": 0.0658182579086534,
"grad_norm": 3.319098711013794,
"learning_rate": 6.5817005896899375e-06,
"loss": 1.1156,
"step": 3460
},
{
"epoch": 0.06600848408758013,
"grad_norm": 2.946639060974121,
"learning_rate": 6.600722845729504e-06,
"loss": 1.018,
"step": 3470
},
{
"epoch": 0.06619871026650688,
"grad_norm": 2.6161160469055176,
"learning_rate": 6.619745101769071e-06,
"loss": 0.8507,
"step": 3480
},
{
"epoch": 0.06638893644543362,
"grad_norm": 5.574872970581055,
"learning_rate": 6.638767357808637e-06,
"loss": 0.9083,
"step": 3490
},
{
"epoch": 0.06657916262436037,
"grad_norm": 13.42422866821289,
"learning_rate": 6.657789613848203e-06,
"loss": 1.0902,
"step": 3500
},
{
"epoch": 0.0667693888032871,
"grad_norm": 2.8455140590667725,
"learning_rate": 6.676811869887769e-06,
"loss": 1.06,
"step": 3510
},
{
"epoch": 0.06695961498221385,
"grad_norm": 2.6613621711730957,
"learning_rate": 6.695834125927335e-06,
"loss": 0.989,
"step": 3520
},
{
"epoch": 0.0671498411611406,
"grad_norm": 3.7422332763671875,
"learning_rate": 6.7148563819669024e-06,
"loss": 0.8448,
"step": 3530
},
{
"epoch": 0.06734006734006734,
"grad_norm": 5.147526264190674,
"learning_rate": 6.733878638006468e-06,
"loss": 0.8717,
"step": 3540
},
{
"epoch": 0.06753029351899409,
"grad_norm": 12.03034782409668,
"learning_rate": 6.752900894046034e-06,
"loss": 1.048,
"step": 3550
},
{
"epoch": 0.06772051969792083,
"grad_norm": 3.214045763015747,
"learning_rate": 6.771923150085601e-06,
"loss": 1.0937,
"step": 3560
},
{
"epoch": 0.06791074587684758,
"grad_norm": 2.7989227771759033,
"learning_rate": 6.790945406125167e-06,
"loss": 1.0337,
"step": 3570
},
{
"epoch": 0.06810097205577431,
"grad_norm": 4.597136497497559,
"learning_rate": 6.809967662164733e-06,
"loss": 0.8806,
"step": 3580
},
{
"epoch": 0.06829119823470106,
"grad_norm": 4.616443634033203,
"learning_rate": 6.8289899182042995e-06,
"loss": 0.8547,
"step": 3590
},
{
"epoch": 0.0684814244136278,
"grad_norm": 11.670774459838867,
"learning_rate": 6.848012174243866e-06,
"loss": 1.0259,
"step": 3600
},
{
"epoch": 0.06867165059255455,
"grad_norm": 2.3249611854553223,
"learning_rate": 6.867034430283433e-06,
"loss": 1.045,
"step": 3610
},
{
"epoch": 0.0688618767714813,
"grad_norm": 4.721532344818115,
"learning_rate": 6.886056686322998e-06,
"loss": 1.0629,
"step": 3620
},
{
"epoch": 0.06905210295040803,
"grad_norm": 3.195577621459961,
"learning_rate": 6.905078942362565e-06,
"loss": 0.8311,
"step": 3630
},
{
"epoch": 0.06924232912933478,
"grad_norm": 5.499499320983887,
"learning_rate": 6.924101198402131e-06,
"loss": 0.8811,
"step": 3640
},
{
"epoch": 0.06943255530826152,
"grad_norm": 12.689728736877441,
"learning_rate": 6.943123454441697e-06,
"loss": 1.0668,
"step": 3650
},
{
"epoch": 0.06962278148718827,
"grad_norm": 2.8369383811950684,
"learning_rate": 6.962145710481264e-06,
"loss": 1.066,
"step": 3660
},
{
"epoch": 0.069813007666115,
"grad_norm": 5.4221415519714355,
"learning_rate": 6.98116796652083e-06,
"loss": 1.0518,
"step": 3670
},
{
"epoch": 0.07000323384504176,
"grad_norm": 2.920156717300415,
"learning_rate": 7.000190222560396e-06,
"loss": 0.8456,
"step": 3680
},
{
"epoch": 0.07019346002396849,
"grad_norm": 6.14293909072876,
"learning_rate": 7.019212478599963e-06,
"loss": 0.8598,
"step": 3690
},
{
"epoch": 0.07038368620289524,
"grad_norm": 14.368626594543457,
"learning_rate": 7.038234734639528e-06,
"loss": 1.0146,
"step": 3700
},
{
"epoch": 0.070573912381822,
"grad_norm": 3.4393813610076904,
"learning_rate": 7.057256990679095e-06,
"loss": 1.1011,
"step": 3710
},
{
"epoch": 0.07076413856074873,
"grad_norm": 2.811824083328247,
"learning_rate": 7.0762792467186615e-06,
"loss": 1.0224,
"step": 3720
},
{
"epoch": 0.07095436473967548,
"grad_norm": 3.577205181121826,
"learning_rate": 7.095301502758227e-06,
"loss": 0.8828,
"step": 3730
},
{
"epoch": 0.07114459091860222,
"grad_norm": 5.109273910522461,
"learning_rate": 7.114323758797794e-06,
"loss": 0.8378,
"step": 3740
},
{
"epoch": 0.07133481709752897,
"grad_norm": 12.113494873046875,
"learning_rate": 7.13334601483736e-06,
"loss": 1.0819,
"step": 3750
},
{
"epoch": 0.0715250432764557,
"grad_norm": 2.632756471633911,
"learning_rate": 7.152368270876927e-06,
"loss": 1.0574,
"step": 3760
},
{
"epoch": 0.07171526945538245,
"grad_norm": 2.616201162338257,
"learning_rate": 7.171390526916493e-06,
"loss": 1.0205,
"step": 3770
},
{
"epoch": 0.07190549563430919,
"grad_norm": 4.030309200286865,
"learning_rate": 7.190412782956059e-06,
"loss": 0.8747,
"step": 3780
},
{
"epoch": 0.07209572181323594,
"grad_norm": 5.87321138381958,
"learning_rate": 7.209435038995626e-06,
"loss": 0.9093,
"step": 3790
},
{
"epoch": 0.07228594799216267,
"grad_norm": 13.199514389038086,
"learning_rate": 7.228457295035192e-06,
"loss": 1.0841,
"step": 3800
},
{
"epoch": 0.07247617417108942,
"grad_norm": 3.2827208042144775,
"learning_rate": 7.247479551074757e-06,
"loss": 1.0602,
"step": 3810
},
{
"epoch": 0.07266640035001617,
"grad_norm": 2.744004726409912,
"learning_rate": 7.266501807114324e-06,
"loss": 0.9903,
"step": 3820
},
{
"epoch": 0.07285662652894291,
"grad_norm": 4.373607635498047,
"learning_rate": 7.28552406315389e-06,
"loss": 0.8325,
"step": 3830
},
{
"epoch": 0.07304685270786966,
"grad_norm": 5.0198140144348145,
"learning_rate": 7.304546319193457e-06,
"loss": 0.9334,
"step": 3840
},
{
"epoch": 0.0732370788867964,
"grad_norm": 13.514430046081543,
"learning_rate": 7.3235685752330235e-06,
"loss": 1.1131,
"step": 3850
},
{
"epoch": 0.07342730506572315,
"grad_norm": 2.1664793491363525,
"learning_rate": 7.342590831272589e-06,
"loss": 1.0527,
"step": 3860
},
{
"epoch": 0.07361753124464988,
"grad_norm": 6.562690258026123,
"learning_rate": 7.361613087312156e-06,
"loss": 0.9892,
"step": 3870
},
{
"epoch": 0.07380775742357663,
"grad_norm": 3.725823163986206,
"learning_rate": 7.380635343351722e-06,
"loss": 0.8403,
"step": 3880
},
{
"epoch": 0.07399798360250337,
"grad_norm": 5.953464031219482,
"learning_rate": 7.399657599391289e-06,
"loss": 0.8569,
"step": 3890
},
{
"epoch": 0.07418820978143012,
"grad_norm": 10.803670883178711,
"learning_rate": 7.418679855430854e-06,
"loss": 1.0628,
"step": 3900
},
{
"epoch": 0.07437843596035687,
"grad_norm": 2.3450770378112793,
"learning_rate": 7.437702111470421e-06,
"loss": 1.0782,
"step": 3910
},
{
"epoch": 0.0745686621392836,
"grad_norm": 2.5900728702545166,
"learning_rate": 7.456724367509988e-06,
"loss": 0.9397,
"step": 3920
},
{
"epoch": 0.07475888831821036,
"grad_norm": 3.811370849609375,
"learning_rate": 7.475746623549554e-06,
"loss": 0.8472,
"step": 3930
},
{
"epoch": 0.07494911449713709,
"grad_norm": 8.507999420166016,
"learning_rate": 7.494768879589119e-06,
"loss": 0.8761,
"step": 3940
},
{
"epoch": 0.07513934067606384,
"grad_norm": 11.520349502563477,
"learning_rate": 7.513791135628686e-06,
"loss": 1.0814,
"step": 3950
},
{
"epoch": 0.07532956685499058,
"grad_norm": 2.7183773517608643,
"learning_rate": 7.532813391668252e-06,
"loss": 1.0746,
"step": 3960
},
{
"epoch": 0.07551979303391733,
"grad_norm": 2.7066807746887207,
"learning_rate": 7.551835647707819e-06,
"loss": 0.9913,
"step": 3970
},
{
"epoch": 0.07571001921284407,
"grad_norm": 2.972797393798828,
"learning_rate": 7.570857903747385e-06,
"loss": 0.8899,
"step": 3980
},
{
"epoch": 0.07590024539177082,
"grad_norm": 5.18887996673584,
"learning_rate": 7.589880159786951e-06,
"loss": 0.8659,
"step": 3990
},
{
"epoch": 0.07609047157069757,
"grad_norm": 13.543824195861816,
"learning_rate": 7.608902415826518e-06,
"loss": 1.003,
"step": 4000
},
{
"epoch": 0.0762806977496243,
"grad_norm": 2.2497000694274902,
"learning_rate": 7.627924671866084e-06,
"loss": 1.0616,
"step": 4010
},
{
"epoch": 0.07647092392855105,
"grad_norm": 2.5572075843811035,
"learning_rate": 7.646946927905651e-06,
"loss": 0.9807,
"step": 4020
},
{
"epoch": 0.07666115010747779,
"grad_norm": 4.120774745941162,
"learning_rate": 7.665969183945216e-06,
"loss": 0.8635,
"step": 4030
},
{
"epoch": 0.07685137628640454,
"grad_norm": 5.530590534210205,
"learning_rate": 7.684991439984782e-06,
"loss": 0.8859,
"step": 4040
},
{
"epoch": 0.07704160246533127,
"grad_norm": 12.142224311828613,
"learning_rate": 7.704013696024349e-06,
"loss": 1.0475,
"step": 4050
},
{
"epoch": 0.07723182864425802,
"grad_norm": 3.392179250717163,
"learning_rate": 7.723035952063916e-06,
"loss": 1.0639,
"step": 4060
},
{
"epoch": 0.07742205482318476,
"grad_norm": 2.641266345977783,
"learning_rate": 7.742058208103481e-06,
"loss": 0.9801,
"step": 4070
},
{
"epoch": 0.07761228100211151,
"grad_norm": 4.2979350090026855,
"learning_rate": 7.761080464143048e-06,
"loss": 0.8867,
"step": 4080
},
{
"epoch": 0.07780250718103826,
"grad_norm": 5.953768730163574,
"learning_rate": 7.780102720182613e-06,
"loss": 0.8286,
"step": 4090
},
{
"epoch": 0.077992733359965,
"grad_norm": 9.22087574005127,
"learning_rate": 7.79912497622218e-06,
"loss": 1.0132,
"step": 4100
},
{
"epoch": 0.07818295953889175,
"grad_norm": 2.587709665298462,
"learning_rate": 7.818147232261748e-06,
"loss": 1.064,
"step": 4110
},
{
"epoch": 0.07837318571781848,
"grad_norm": 2.669003486633301,
"learning_rate": 7.837169488301313e-06,
"loss": 0.9844,
"step": 4120
},
{
"epoch": 0.07856341189674523,
"grad_norm": 3.0199615955352783,
"learning_rate": 7.85619174434088e-06,
"loss": 0.8422,
"step": 4130
},
{
"epoch": 0.07875363807567197,
"grad_norm": 4.246236324310303,
"learning_rate": 7.875214000380445e-06,
"loss": 0.8418,
"step": 4140
},
{
"epoch": 0.07894386425459872,
"grad_norm": 10.317477226257324,
"learning_rate": 7.894236256420012e-06,
"loss": 1.0039,
"step": 4150
},
{
"epoch": 0.07913409043352546,
"grad_norm": 2.923060655593872,
"learning_rate": 7.913258512459578e-06,
"loss": 1.0689,
"step": 4160
},
{
"epoch": 0.0793243166124522,
"grad_norm": 1.7650673389434814,
"learning_rate": 7.932280768499145e-06,
"loss": 0.9898,
"step": 4170
},
{
"epoch": 0.07951454279137896,
"grad_norm": 4.0497822761535645,
"learning_rate": 7.951303024538712e-06,
"loss": 0.8677,
"step": 4180
},
{
"epoch": 0.07970476897030569,
"grad_norm": 4.521123886108398,
"learning_rate": 7.970325280578277e-06,
"loss": 0.8269,
"step": 4190
},
{
"epoch": 0.07989499514923244,
"grad_norm": 12.454859733581543,
"learning_rate": 7.989347536617842e-06,
"loss": 0.9994,
"step": 4200
},
{
"epoch": 0.08008522132815918,
"grad_norm": 2.702120542526245,
"learning_rate": 8.00836979265741e-06,
"loss": 1.0442,
"step": 4210
},
{
"epoch": 0.08027544750708593,
"grad_norm": 2.9055328369140625,
"learning_rate": 8.027392048696976e-06,
"loss": 1.0382,
"step": 4220
},
{
"epoch": 0.08046567368601266,
"grad_norm": 3.210998773574829,
"learning_rate": 8.046414304736543e-06,
"loss": 0.8021,
"step": 4230
},
{
"epoch": 0.08065589986493941,
"grad_norm": 4.988993167877197,
"learning_rate": 8.065436560776109e-06,
"loss": 0.8846,
"step": 4240
},
{
"epoch": 0.08084612604386615,
"grad_norm": 13.53798770904541,
"learning_rate": 8.084458816815674e-06,
"loss": 1.0493,
"step": 4250
},
{
"epoch": 0.0810363522227929,
"grad_norm": 4.448582172393799,
"learning_rate": 8.103481072855241e-06,
"loss": 1.0174,
"step": 4260
},
{
"epoch": 0.08122657840171965,
"grad_norm": 3.8949131965637207,
"learning_rate": 8.122503328894808e-06,
"loss": 1.024,
"step": 4270
},
{
"epoch": 0.08141680458064639,
"grad_norm": 3.751685619354248,
"learning_rate": 8.141525584934375e-06,
"loss": 0.797,
"step": 4280
},
{
"epoch": 0.08160703075957314,
"grad_norm": 4.3078436851501465,
"learning_rate": 8.16054784097394e-06,
"loss": 0.8257,
"step": 4290
},
{
"epoch": 0.08179725693849987,
"grad_norm": 10.666089057922363,
"learning_rate": 8.179570097013506e-06,
"loss": 1.0389,
"step": 4300
},
{
"epoch": 0.08198748311742662,
"grad_norm": 2.153872489929199,
"learning_rate": 8.198592353053073e-06,
"loss": 1.0465,
"step": 4310
},
{
"epoch": 0.08217770929635336,
"grad_norm": 2.911468744277954,
"learning_rate": 8.217614609092638e-06,
"loss": 0.9862,
"step": 4320
},
{
"epoch": 0.08236793547528011,
"grad_norm": 3.601792573928833,
"learning_rate": 8.236636865132205e-06,
"loss": 0.8786,
"step": 4330
},
{
"epoch": 0.08255816165420685,
"grad_norm": 7.536445140838623,
"learning_rate": 8.255659121171772e-06,
"loss": 0.8042,
"step": 4340
},
{
"epoch": 0.0827483878331336,
"grad_norm": 12.88973617553711,
"learning_rate": 8.274681377211338e-06,
"loss": 1.0235,
"step": 4350
},
{
"epoch": 0.08293861401206035,
"grad_norm": 3.1542675495147705,
"learning_rate": 8.293703633250905e-06,
"loss": 1.0693,
"step": 4360
},
{
"epoch": 0.08312884019098708,
"grad_norm": 3.760368824005127,
"learning_rate": 8.31272588929047e-06,
"loss": 1.0306,
"step": 4370
},
{
"epoch": 0.08331906636991383,
"grad_norm": 3.9413628578186035,
"learning_rate": 8.331748145330037e-06,
"loss": 0.8654,
"step": 4380
},
{
"epoch": 0.08350929254884057,
"grad_norm": 4.6039862632751465,
"learning_rate": 8.350770401369604e-06,
"loss": 0.8311,
"step": 4390
},
{
"epoch": 0.08369951872776732,
"grad_norm": 11.079093933105469,
"learning_rate": 8.36979265740917e-06,
"loss": 0.9909,
"step": 4400
},
{
"epoch": 0.08388974490669406,
"grad_norm": 2.3281502723693848,
"learning_rate": 8.388814913448736e-06,
"loss": 1.0569,
"step": 4410
},
{
"epoch": 0.0840799710856208,
"grad_norm": 3.260143518447876,
"learning_rate": 8.407837169488302e-06,
"loss": 1.0086,
"step": 4420
},
{
"epoch": 0.08427019726454754,
"grad_norm": 2.812483310699463,
"learning_rate": 8.426859425527869e-06,
"loss": 0.8608,
"step": 4430
},
{
"epoch": 0.08446042344347429,
"grad_norm": 4.584824085235596,
"learning_rate": 8.445881681567436e-06,
"loss": 0.8322,
"step": 4440
},
{
"epoch": 0.08465064962240103,
"grad_norm": 9.15593147277832,
"learning_rate": 8.464903937607001e-06,
"loss": 1.0237,
"step": 4450
},
{
"epoch": 0.08484087580132778,
"grad_norm": 2.053774118423462,
"learning_rate": 8.483926193646566e-06,
"loss": 1.0547,
"step": 4460
},
{
"epoch": 0.08503110198025453,
"grad_norm": 2.387960910797119,
"learning_rate": 8.502948449686133e-06,
"loss": 0.9456,
"step": 4470
},
{
"epoch": 0.08522132815918126,
"grad_norm": 3.73870849609375,
"learning_rate": 8.521970705725699e-06,
"loss": 0.8484,
"step": 4480
},
{
"epoch": 0.08541155433810801,
"grad_norm": 5.459985256195068,
"learning_rate": 8.540992961765266e-06,
"loss": 0.8449,
"step": 4490
},
{
"epoch": 0.08560178051703475,
"grad_norm": 14.273500442504883,
"learning_rate": 8.560015217804833e-06,
"loss": 1.0177,
"step": 4500
},
{
"epoch": 0.0857920066959615,
"grad_norm": 3.626765012741089,
"learning_rate": 8.579037473844398e-06,
"loss": 1.0674,
"step": 4510
},
{
"epoch": 0.08598223287488824,
"grad_norm": 5.053386688232422,
"learning_rate": 8.598059729883965e-06,
"loss": 1.0287,
"step": 4520
},
{
"epoch": 0.08617245905381499,
"grad_norm": 3.2023839950561523,
"learning_rate": 8.61708198592353e-06,
"loss": 0.7877,
"step": 4530
},
{
"epoch": 0.08636268523274172,
"grad_norm": 9.036783218383789,
"learning_rate": 8.636104241963097e-06,
"loss": 0.8622,
"step": 4540
},
{
"epoch": 0.08655291141166847,
"grad_norm": 12.49517822265625,
"learning_rate": 8.655126498002664e-06,
"loss": 1.0252,
"step": 4550
},
{
"epoch": 0.08674313759059522,
"grad_norm": 2.4512569904327393,
"learning_rate": 8.67414875404223e-06,
"loss": 1.0266,
"step": 4560
},
{
"epoch": 0.08693336376952196,
"grad_norm": 2.0704665184020996,
"learning_rate": 8.693171010081797e-06,
"loss": 0.9304,
"step": 4570
},
{
"epoch": 0.08712358994844871,
"grad_norm": 2.670184850692749,
"learning_rate": 8.712193266121362e-06,
"loss": 0.8504,
"step": 4580
},
{
"epoch": 0.08731381612737545,
"grad_norm": 7.81472635269165,
"learning_rate": 8.73121552216093e-06,
"loss": 0.8383,
"step": 4590
},
{
"epoch": 0.0875040423063022,
"grad_norm": 11.205781936645508,
"learning_rate": 8.750237778200496e-06,
"loss": 1.0539,
"step": 4600
},
{
"epoch": 0.08769426848522893,
"grad_norm": 2.879502773284912,
"learning_rate": 8.769260034240062e-06,
"loss": 1.0363,
"step": 4610
},
{
"epoch": 0.08788449466415568,
"grad_norm": 4.640857696533203,
"learning_rate": 8.788282290279629e-06,
"loss": 1.0272,
"step": 4620
},
{
"epoch": 0.08807472084308242,
"grad_norm": 3.4126992225646973,
"learning_rate": 8.807304546319194e-06,
"loss": 0.8409,
"step": 4630
},
{
"epoch": 0.08826494702200917,
"grad_norm": 3.9957668781280518,
"learning_rate": 8.82632680235876e-06,
"loss": 0.876,
"step": 4640
},
{
"epoch": 0.08845517320093592,
"grad_norm": 11.795957565307617,
"learning_rate": 8.845349058398326e-06,
"loss": 1.0023,
"step": 4650
},
{
"epoch": 0.08864539937986265,
"grad_norm": 2.539607524871826,
"learning_rate": 8.864371314437893e-06,
"loss": 1.0589,
"step": 4660
},
{
"epoch": 0.0888356255587894,
"grad_norm": 2.778062343597412,
"learning_rate": 8.88339357047746e-06,
"loss": 0.9831,
"step": 4670
},
{
"epoch": 0.08902585173771614,
"grad_norm": 2.902723789215088,
"learning_rate": 8.902415826517026e-06,
"loss": 0.8115,
"step": 4680
},
{
"epoch": 0.08921607791664289,
"grad_norm": 6.224371910095215,
"learning_rate": 8.921438082556591e-06,
"loss": 0.8285,
"step": 4690
},
{
"epoch": 0.08940630409556963,
"grad_norm": 13.158405303955078,
"learning_rate": 8.940460338596158e-06,
"loss": 1.056,
"step": 4700
},
{
"epoch": 0.08959653027449638,
"grad_norm": 2.7186522483825684,
"learning_rate": 8.959482594635725e-06,
"loss": 1.0455,
"step": 4710
},
{
"epoch": 0.08978675645342311,
"grad_norm": 2.863405227661133,
"learning_rate": 8.97850485067529e-06,
"loss": 1.0269,
"step": 4720
},
{
"epoch": 0.08997698263234986,
"grad_norm": 2.861708641052246,
"learning_rate": 8.997527106714857e-06,
"loss": 0.7955,
"step": 4730
},
{
"epoch": 0.09016720881127661,
"grad_norm": 4.701228618621826,
"learning_rate": 9.016549362754423e-06,
"loss": 0.8354,
"step": 4740
},
{
"epoch": 0.09035743499020335,
"grad_norm": 12.41553020477295,
"learning_rate": 9.03557161879399e-06,
"loss": 1.131,
"step": 4750
},
{
"epoch": 0.0905476611691301,
"grad_norm": 2.774181842803955,
"learning_rate": 9.054593874833557e-06,
"loss": 1.045,
"step": 4760
},
{
"epoch": 0.09073788734805684,
"grad_norm": 2.170881748199463,
"learning_rate": 9.073616130873122e-06,
"loss": 0.974,
"step": 4770
},
{
"epoch": 0.09092811352698359,
"grad_norm": 3.6522154808044434,
"learning_rate": 9.092638386912689e-06,
"loss": 0.8233,
"step": 4780
},
{
"epoch": 0.09111833970591032,
"grad_norm": 4.515476226806641,
"learning_rate": 9.111660642952254e-06,
"loss": 0.8054,
"step": 4790
},
{
"epoch": 0.09130856588483707,
"grad_norm": 10.058849334716797,
"learning_rate": 9.130682898991821e-06,
"loss": 1.0239,
"step": 4800
},
{
"epoch": 0.09149879206376381,
"grad_norm": 4.930131912231445,
"learning_rate": 9.149705155031387e-06,
"loss": 1.0748,
"step": 4810
},
{
"epoch": 0.09168901824269056,
"grad_norm": 3.753122091293335,
"learning_rate": 9.168727411070954e-06,
"loss": 0.9262,
"step": 4820
},
{
"epoch": 0.09187924442161731,
"grad_norm": 3.604468822479248,
"learning_rate": 9.187749667110521e-06,
"loss": 0.8119,
"step": 4830
},
{
"epoch": 0.09206947060054405,
"grad_norm": 4.331033706665039,
"learning_rate": 9.206771923150086e-06,
"loss": 0.8029,
"step": 4840
},
{
"epoch": 0.0922596967794708,
"grad_norm": 11.139960289001465,
"learning_rate": 9.225794179189652e-06,
"loss": 0.9785,
"step": 4850
},
{
"epoch": 0.09244992295839753,
"grad_norm": 3.0269057750701904,
"learning_rate": 9.244816435229219e-06,
"loss": 1.0565,
"step": 4860
},
{
"epoch": 0.09264014913732428,
"grad_norm": 2.4621336460113525,
"learning_rate": 9.263838691268786e-06,
"loss": 0.9695,
"step": 4870
},
{
"epoch": 0.09283037531625102,
"grad_norm": 3.160784959793091,
"learning_rate": 9.282860947308353e-06,
"loss": 0.7952,
"step": 4880
},
{
"epoch": 0.09302060149517777,
"grad_norm": 6.189260005950928,
"learning_rate": 9.301883203347918e-06,
"loss": 0.8586,
"step": 4890
},
{
"epoch": 0.0932108276741045,
"grad_norm": 13.33193588256836,
"learning_rate": 9.320905459387483e-06,
"loss": 1.0753,
"step": 4900
},
{
"epoch": 0.09340105385303125,
"grad_norm": 2.4601781368255615,
"learning_rate": 9.33992771542705e-06,
"loss": 1.0345,
"step": 4910
},
{
"epoch": 0.093591280031958,
"grad_norm": 3.0945911407470703,
"learning_rate": 9.358949971466617e-06,
"loss": 0.9604,
"step": 4920
},
{
"epoch": 0.09378150621088474,
"grad_norm": 3.2006800174713135,
"learning_rate": 9.377972227506184e-06,
"loss": 0.8122,
"step": 4930
},
{
"epoch": 0.09397173238981149,
"grad_norm": 6.865549564361572,
"learning_rate": 9.39699448354575e-06,
"loss": 0.8086,
"step": 4940
},
{
"epoch": 0.09416195856873823,
"grad_norm": 11.536060333251953,
"learning_rate": 9.416016739585315e-06,
"loss": 0.9879,
"step": 4950
},
{
"epoch": 0.09435218474766498,
"grad_norm": 4.649012088775635,
"learning_rate": 9.435038995624882e-06,
"loss": 1.0615,
"step": 4960
},
{
"epoch": 0.09454241092659171,
"grad_norm": 2.240013599395752,
"learning_rate": 9.454061251664447e-06,
"loss": 0.9851,
"step": 4970
},
{
"epoch": 0.09473263710551846,
"grad_norm": 3.457646131515503,
"learning_rate": 9.473083507704014e-06,
"loss": 0.7976,
"step": 4980
},
{
"epoch": 0.0949228632844452,
"grad_norm": 4.621734142303467,
"learning_rate": 9.492105763743581e-06,
"loss": 0.7787,
"step": 4990
},
{
"epoch": 0.09511308946337195,
"grad_norm": 12.112364768981934,
"learning_rate": 9.511128019783147e-06,
"loss": 1.0011,
"step": 5000
},
{
"epoch": 0.0953033156422987,
"grad_norm": 2.524951696395874,
"learning_rate": 9.530150275822714e-06,
"loss": 1.0024,
"step": 5010
},
{
"epoch": 0.09549354182122544,
"grad_norm": 2.1611406803131104,
"learning_rate": 9.549172531862279e-06,
"loss": 0.9895,
"step": 5020
},
{
"epoch": 0.09568376800015219,
"grad_norm": 3.647883415222168,
"learning_rate": 9.568194787901846e-06,
"loss": 0.8134,
"step": 5030
},
{
"epoch": 0.09587399417907892,
"grad_norm": 5.40878963470459,
"learning_rate": 9.587217043941413e-06,
"loss": 0.7944,
"step": 5040
},
{
"epoch": 0.09606422035800567,
"grad_norm": 11.021536827087402,
"learning_rate": 9.606239299980978e-06,
"loss": 1.0177,
"step": 5050
},
{
"epoch": 0.09625444653693241,
"grad_norm": 2.3612380027770996,
"learning_rate": 9.625261556020545e-06,
"loss": 1.0267,
"step": 5060
},
{
"epoch": 0.09644467271585916,
"grad_norm": 2.3841519355773926,
"learning_rate": 9.64428381206011e-06,
"loss": 0.9979,
"step": 5070
},
{
"epoch": 0.0966348988947859,
"grad_norm": 2.819408416748047,
"learning_rate": 9.663306068099678e-06,
"loss": 0.7921,
"step": 5080
},
{
"epoch": 0.09682512507371265,
"grad_norm": 4.353105068206787,
"learning_rate": 9.682328324139245e-06,
"loss": 0.8087,
"step": 5090
},
{
"epoch": 0.09701535125263938,
"grad_norm": 9.611862182617188,
"learning_rate": 9.70135058017881e-06,
"loss": 0.9903,
"step": 5100
},
{
"epoch": 0.09720557743156613,
"grad_norm": 3.051971197128296,
"learning_rate": 9.720372836218376e-06,
"loss": 1.0251,
"step": 5110
},
{
"epoch": 0.09739580361049288,
"grad_norm": 3.2879152297973633,
"learning_rate": 9.739395092257943e-06,
"loss": 0.9787,
"step": 5120
},
{
"epoch": 0.09758602978941962,
"grad_norm": 3.3315176963806152,
"learning_rate": 9.758417348297508e-06,
"loss": 0.8526,
"step": 5130
},
{
"epoch": 0.09777625596834637,
"grad_norm": 4.358604907989502,
"learning_rate": 9.777439604337075e-06,
"loss": 0.841,
"step": 5140
},
{
"epoch": 0.0979664821472731,
"grad_norm": 10.932659149169922,
"learning_rate": 9.796461860376642e-06,
"loss": 1.0112,
"step": 5150
},
{
"epoch": 0.09815670832619985,
"grad_norm": 3.0365395545959473,
"learning_rate": 9.815484116416207e-06,
"loss": 1.0484,
"step": 5160
},
{
"epoch": 0.09834693450512659,
"grad_norm": 2.4415764808654785,
"learning_rate": 9.834506372455774e-06,
"loss": 0.9546,
"step": 5170
},
{
"epoch": 0.09853716068405334,
"grad_norm": 3.0862984657287598,
"learning_rate": 9.85352862849534e-06,
"loss": 0.8239,
"step": 5180
},
{
"epoch": 0.09872738686298008,
"grad_norm": 4.357607841491699,
"learning_rate": 9.872550884534907e-06,
"loss": 0.8166,
"step": 5190
},
{
"epoch": 0.09891761304190683,
"grad_norm": 10.748388290405273,
"learning_rate": 9.891573140574474e-06,
"loss": 1.0355,
"step": 5200
},
{
"epoch": 0.09910783922083358,
"grad_norm": 2.779499053955078,
"learning_rate": 9.910595396614039e-06,
"loss": 1.0779,
"step": 5210
},
{
"epoch": 0.09929806539976031,
"grad_norm": 3.65140700340271,
"learning_rate": 9.929617652653606e-06,
"loss": 0.9539,
"step": 5220
},
{
"epoch": 0.09948829157868706,
"grad_norm": 3.2343590259552,
"learning_rate": 9.948639908693171e-06,
"loss": 0.7727,
"step": 5230
},
{
"epoch": 0.0996785177576138,
"grad_norm": 5.205893516540527,
"learning_rate": 9.967662164732738e-06,
"loss": 0.8448,
"step": 5240
},
{
"epoch": 0.09986874393654055,
"grad_norm": 11.548497200012207,
"learning_rate": 9.986684420772305e-06,
"loss": 1.0132,
"step": 5250
},
{
"epoch": 0.10005897011546729,
"grad_norm": 2.170140504837036,
"learning_rate": 1e-05,
"loss": 0.9852,
"step": 5260
},
{
"epoch": 0.10024919629439404,
"grad_norm": 2.476912260055542,
"learning_rate": 1e-05,
"loss": 0.9804,
"step": 5270
},
{
"epoch": 0.10043942247332077,
"grad_norm": 3.0934178829193115,
"learning_rate": 1e-05,
"loss": 0.8043,
"step": 5280
},
{
"epoch": 0.10062964865224752,
"grad_norm": 4.3917999267578125,
"learning_rate": 1e-05,
"loss": 0.8387,
"step": 5290
},
{
"epoch": 0.10081987483117427,
"grad_norm": 14.04360580444336,
"learning_rate": 1e-05,
"loss": 1.0638,
"step": 5300
},
{
"epoch": 0.10101010101010101,
"grad_norm": 1.9007540941238403,
"learning_rate": 1e-05,
"loss": 1.0097,
"step": 5310
},
{
"epoch": 0.10120032718902776,
"grad_norm": 2.8841350078582764,
"learning_rate": 1e-05,
"loss": 0.9581,
"step": 5320
},
{
"epoch": 0.1013905533679545,
"grad_norm": 4.027973175048828,
"learning_rate": 1e-05,
"loss": 0.7922,
"step": 5330
},
{
"epoch": 0.10158077954688124,
"grad_norm": 4.386593818664551,
"learning_rate": 1e-05,
"loss": 0.8356,
"step": 5340
},
{
"epoch": 0.10177100572580798,
"grad_norm": 10.778451919555664,
"learning_rate": 1e-05,
"loss": 0.987,
"step": 5350
},
{
"epoch": 0.10196123190473473,
"grad_norm": 2.4819626808166504,
"learning_rate": 1e-05,
"loss": 1.0063,
"step": 5360
},
{
"epoch": 0.10215145808366147,
"grad_norm": 2.8411858081817627,
"learning_rate": 1e-05,
"loss": 0.9477,
"step": 5370
},
{
"epoch": 0.10234168426258822,
"grad_norm": 3.0888965129852295,
"learning_rate": 1e-05,
"loss": 0.7993,
"step": 5380
},
{
"epoch": 0.10253191044151497,
"grad_norm": 4.141193866729736,
"learning_rate": 1e-05,
"loss": 0.8134,
"step": 5390
},
{
"epoch": 0.1027221366204417,
"grad_norm": 9.52237319946289,
"learning_rate": 1e-05,
"loss": 0.9566,
"step": 5400
},
{
"epoch": 0.10291236279936845,
"grad_norm": 3.151825428009033,
"learning_rate": 1e-05,
"loss": 1.0167,
"step": 5410
},
{
"epoch": 0.10310258897829519,
"grad_norm": 2.040436267852783,
"learning_rate": 1e-05,
"loss": 0.9696,
"step": 5420
},
{
"epoch": 0.10329281515722194,
"grad_norm": 3.3553965091705322,
"learning_rate": 1e-05,
"loss": 0.8425,
"step": 5430
},
{
"epoch": 0.10348304133614868,
"grad_norm": 5.130828857421875,
"learning_rate": 1e-05,
"loss": 0.8082,
"step": 5440
},
{
"epoch": 0.10367326751507543,
"grad_norm": 9.019664764404297,
"learning_rate": 1e-05,
"loss": 0.9742,
"step": 5450
},
{
"epoch": 0.10386349369400216,
"grad_norm": 2.3726866245269775,
"learning_rate": 1e-05,
"loss": 0.9999,
"step": 5460
},
{
"epoch": 0.10405371987292891,
"grad_norm": 2.137408494949341,
"learning_rate": 1e-05,
"loss": 0.9862,
"step": 5470
},
{
"epoch": 0.10424394605185566,
"grad_norm": 3.0669353008270264,
"learning_rate": 1e-05,
"loss": 0.7806,
"step": 5480
},
{
"epoch": 0.1044341722307824,
"grad_norm": 3.7706305980682373,
"learning_rate": 1e-05,
"loss": 0.8121,
"step": 5490
},
{
"epoch": 0.10462439840970915,
"grad_norm": 10.3373441696167,
"learning_rate": 1e-05,
"loss": 0.9813,
"step": 5500
},
{
"epoch": 0.10481462458863589,
"grad_norm": 2.218104124069214,
"learning_rate": 1e-05,
"loss": 1.0445,
"step": 5510
},
{
"epoch": 0.10500485076756264,
"grad_norm": 2.3944590091705322,
"learning_rate": 1e-05,
"loss": 0.9092,
"step": 5520
},
{
"epoch": 0.10519507694648937,
"grad_norm": 3.1064295768737793,
"learning_rate": 1e-05,
"loss": 0.7949,
"step": 5530
},
{
"epoch": 0.10538530312541612,
"grad_norm": 5.8249030113220215,
"learning_rate": 1e-05,
"loss": 0.8293,
"step": 5540
},
{
"epoch": 0.10557552930434286,
"grad_norm": 9.393155097961426,
"learning_rate": 1e-05,
"loss": 0.9622,
"step": 5550
},
{
"epoch": 0.10576575548326961,
"grad_norm": 4.191644668579102,
"learning_rate": 1e-05,
"loss": 1.0034,
"step": 5560
},
{
"epoch": 0.10595598166219636,
"grad_norm": 2.126408576965332,
"learning_rate": 1e-05,
"loss": 0.9413,
"step": 5570
},
{
"epoch": 0.1061462078411231,
"grad_norm": 2.6248719692230225,
"learning_rate": 1e-05,
"loss": 0.7711,
"step": 5580
},
{
"epoch": 0.10633643402004984,
"grad_norm": 4.614845275878906,
"learning_rate": 1e-05,
"loss": 0.8127,
"step": 5590
},
{
"epoch": 0.10652666019897658,
"grad_norm": 10.200318336486816,
"learning_rate": 1e-05,
"loss": 0.9967,
"step": 5600
},
{
"epoch": 0.10671688637790333,
"grad_norm": 2.541682481765747,
"learning_rate": 1e-05,
"loss": 1.0298,
"step": 5610
},
{
"epoch": 0.10690711255683007,
"grad_norm": 2.1859960556030273,
"learning_rate": 1e-05,
"loss": 0.9212,
"step": 5620
},
{
"epoch": 0.10709733873575682,
"grad_norm": 2.971053123474121,
"learning_rate": 1e-05,
"loss": 0.7827,
"step": 5630
},
{
"epoch": 0.10728756491468355,
"grad_norm": 5.017271518707275,
"learning_rate": 1e-05,
"loss": 0.7759,
"step": 5640
},
{
"epoch": 0.1074777910936103,
"grad_norm": 9.602482795715332,
"learning_rate": 1e-05,
"loss": 0.9141,
"step": 5650
},
{
"epoch": 0.10766801727253705,
"grad_norm": 2.660590887069702,
"learning_rate": 1e-05,
"loss": 1.0346,
"step": 5660
},
{
"epoch": 0.10785824345146379,
"grad_norm": 2.3705356121063232,
"learning_rate": 1e-05,
"loss": 0.9302,
"step": 5670
},
{
"epoch": 0.10804846963039054,
"grad_norm": 2.7425448894500732,
"learning_rate": 1e-05,
"loss": 0.7813,
"step": 5680
},
{
"epoch": 0.10823869580931728,
"grad_norm": 4.8073601722717285,
"learning_rate": 1e-05,
"loss": 0.7984,
"step": 5690
},
{
"epoch": 0.10842892198824403,
"grad_norm": 11.548068046569824,
"learning_rate": 1e-05,
"loss": 0.9622,
"step": 5700
},
{
"epoch": 0.10861914816717076,
"grad_norm": 2.775019407272339,
"learning_rate": 1e-05,
"loss": 0.984,
"step": 5710
},
{
"epoch": 0.10880937434609751,
"grad_norm": 3.1400108337402344,
"learning_rate": 1e-05,
"loss": 0.9565,
"step": 5720
},
{
"epoch": 0.10899960052502425,
"grad_norm": 2.912726879119873,
"learning_rate": 1e-05,
"loss": 0.7693,
"step": 5730
},
{
"epoch": 0.109189826703951,
"grad_norm": 4.758162021636963,
"learning_rate": 1e-05,
"loss": 0.8397,
"step": 5740
},
{
"epoch": 0.10938005288287775,
"grad_norm": 11.496520042419434,
"learning_rate": 1e-05,
"loss": 0.9572,
"step": 5750
},
{
"epoch": 0.10957027906180448,
"grad_norm": 2.6489996910095215,
"learning_rate": 1e-05,
"loss": 1.0163,
"step": 5760
},
{
"epoch": 0.10976050524073123,
"grad_norm": 2.03450345993042,
"learning_rate": 1e-05,
"loss": 0.9425,
"step": 5770
},
{
"epoch": 0.10995073141965797,
"grad_norm": 2.9618072509765625,
"learning_rate": 1e-05,
"loss": 0.7925,
"step": 5780
},
{
"epoch": 0.11014095759858472,
"grad_norm": 3.8805148601531982,
"learning_rate": 1e-05,
"loss": 0.8022,
"step": 5790
},
{
"epoch": 0.11033118377751146,
"grad_norm": 10.308512687683105,
"learning_rate": 1e-05,
"loss": 0.9776,
"step": 5800
},
{
"epoch": 0.11052140995643821,
"grad_norm": 2.324984312057495,
"learning_rate": 1e-05,
"loss": 1.0327,
"step": 5810
},
{
"epoch": 0.11071163613536494,
"grad_norm": 2.211977005004883,
"learning_rate": 1e-05,
"loss": 0.9078,
"step": 5820
},
{
"epoch": 0.1109018623142917,
"grad_norm": 3.970721483230591,
"learning_rate": 1e-05,
"loss": 0.7686,
"step": 5830
},
{
"epoch": 0.11109208849321843,
"grad_norm": 4.543284893035889,
"learning_rate": 1e-05,
"loss": 0.7989,
"step": 5840
},
{
"epoch": 0.11128231467214518,
"grad_norm": 10.333953857421875,
"learning_rate": 1e-05,
"loss": 0.965,
"step": 5850
},
{
"epoch": 0.11147254085107193,
"grad_norm": 2.5103368759155273,
"learning_rate": 1e-05,
"loss": 1.0208,
"step": 5860
},
{
"epoch": 0.11166276702999867,
"grad_norm": 2.713374614715576,
"learning_rate": 1e-05,
"loss": 0.9197,
"step": 5870
},
{
"epoch": 0.11185299320892542,
"grad_norm": 2.925102710723877,
"learning_rate": 1e-05,
"loss": 0.763,
"step": 5880
},
{
"epoch": 0.11204321938785215,
"grad_norm": 4.616085052490234,
"learning_rate": 1e-05,
"loss": 0.7851,
"step": 5890
},
{
"epoch": 0.1122334455667789,
"grad_norm": 11.197551727294922,
"learning_rate": 1e-05,
"loss": 0.941,
"step": 5900
},
{
"epoch": 0.11242367174570564,
"grad_norm": 5.09884786605835,
"learning_rate": 1e-05,
"loss": 0.9611,
"step": 5910
},
{
"epoch": 0.11261389792463239,
"grad_norm": 2.7460286617279053,
"learning_rate": 1e-05,
"loss": 0.9698,
"step": 5920
},
{
"epoch": 0.11280412410355913,
"grad_norm": 3.23636794090271,
"learning_rate": 1e-05,
"loss": 0.7744,
"step": 5930
},
{
"epoch": 0.11299435028248588,
"grad_norm": 3.779413938522339,
"learning_rate": 1e-05,
"loss": 0.7861,
"step": 5940
},
{
"epoch": 0.11318457646141263,
"grad_norm": 12.377265930175781,
"learning_rate": 1e-05,
"loss": 0.9692,
"step": 5950
},
{
"epoch": 0.11337480264033936,
"grad_norm": 2.6297824382781982,
"learning_rate": 1e-05,
"loss": 0.9445,
"step": 5960
},
{
"epoch": 0.11356502881926611,
"grad_norm": 3.0455784797668457,
"learning_rate": 1e-05,
"loss": 0.9369,
"step": 5970
},
{
"epoch": 0.11375525499819285,
"grad_norm": 3.3172903060913086,
"learning_rate": 1e-05,
"loss": 0.7328,
"step": 5980
},
{
"epoch": 0.1139454811771196,
"grad_norm": 6.561709880828857,
"learning_rate": 1e-05,
"loss": 0.7805,
"step": 5990
},
{
"epoch": 0.11413570735604633,
"grad_norm": 8.712008476257324,
"learning_rate": 1e-05,
"loss": 0.9122,
"step": 6000
},
{
"epoch": 0.11432593353497308,
"grad_norm": 2.2284984588623047,
"learning_rate": 1e-05,
"loss": 1.0115,
"step": 6010
},
{
"epoch": 0.11451615971389982,
"grad_norm": 2.497851610183716,
"learning_rate": 1e-05,
"loss": 0.9679,
"step": 6020
},
{
"epoch": 0.11470638589282657,
"grad_norm": 3.0872201919555664,
"learning_rate": 1e-05,
"loss": 0.8158,
"step": 6030
},
{
"epoch": 0.11489661207175332,
"grad_norm": 4.245258808135986,
"learning_rate": 1e-05,
"loss": 0.7766,
"step": 6040
},
{
"epoch": 0.11508683825068006,
"grad_norm": 12.017084121704102,
"learning_rate": 1e-05,
"loss": 0.9771,
"step": 6050
},
{
"epoch": 0.11527706442960681,
"grad_norm": 2.125140905380249,
"learning_rate": 1e-05,
"loss": 0.9782,
"step": 6060
},
{
"epoch": 0.11546729060853354,
"grad_norm": 2.8148915767669678,
"learning_rate": 1e-05,
"loss": 0.9655,
"step": 6070
},
{
"epoch": 0.1156575167874603,
"grad_norm": 3.286608934402466,
"learning_rate": 1e-05,
"loss": 0.7886,
"step": 6080
},
{
"epoch": 0.11584774296638703,
"grad_norm": 3.6205978393554688,
"learning_rate": 1e-05,
"loss": 0.7683,
"step": 6090
},
{
"epoch": 0.11603796914531378,
"grad_norm": 8.98322582244873,
"learning_rate": 1e-05,
"loss": 0.9018,
"step": 6100
},
{
"epoch": 0.11622819532424052,
"grad_norm": 2.4430668354034424,
"learning_rate": 1e-05,
"loss": 0.9824,
"step": 6110
},
{
"epoch": 0.11641842150316727,
"grad_norm": 1.9802136421203613,
"learning_rate": 1e-05,
"loss": 0.9655,
"step": 6120
},
{
"epoch": 0.11660864768209402,
"grad_norm": 3.089259386062622,
"learning_rate": 1e-05,
"loss": 0.8234,
"step": 6130
},
{
"epoch": 0.11679887386102075,
"grad_norm": 3.642322063446045,
"learning_rate": 1e-05,
"loss": 0.7563,
"step": 6140
},
{
"epoch": 0.1169891000399475,
"grad_norm": 9.399459838867188,
"learning_rate": 1e-05,
"loss": 0.9738,
"step": 6150
},
{
"epoch": 0.11717932621887424,
"grad_norm": 4.041309356689453,
"learning_rate": 1e-05,
"loss": 0.9792,
"step": 6160
},
{
"epoch": 0.11736955239780099,
"grad_norm": 2.730517864227295,
"learning_rate": 1e-05,
"loss": 0.9379,
"step": 6170
},
{
"epoch": 0.11755977857672772,
"grad_norm": 2.9666225910186768,
"learning_rate": 1e-05,
"loss": 0.7589,
"step": 6180
},
{
"epoch": 0.11775000475565447,
"grad_norm": 3.7376034259796143,
"learning_rate": 1e-05,
"loss": 0.7737,
"step": 6190
},
{
"epoch": 0.11794023093458121,
"grad_norm": 9.174288749694824,
"learning_rate": 1e-05,
"loss": 0.9367,
"step": 6200
},
{
"epoch": 0.11813045711350796,
"grad_norm": 2.6808459758758545,
"learning_rate": 1e-05,
"loss": 0.9981,
"step": 6210
},
{
"epoch": 0.11832068329243471,
"grad_norm": 2.970728874206543,
"learning_rate": 1e-05,
"loss": 0.9225,
"step": 6220
},
{
"epoch": 0.11851090947136145,
"grad_norm": 3.5266144275665283,
"learning_rate": 1e-05,
"loss": 0.7492,
"step": 6230
},
{
"epoch": 0.1187011356502882,
"grad_norm": 3.797211170196533,
"learning_rate": 1e-05,
"loss": 0.7845,
"step": 6240
},
{
"epoch": 0.11889136182921493,
"grad_norm": 9.619593620300293,
"learning_rate": 1e-05,
"loss": 0.9271,
"step": 6250
},
{
"epoch": 0.11908158800814168,
"grad_norm": 2.430349588394165,
"learning_rate": 1e-05,
"loss": 0.9484,
"step": 6260
},
{
"epoch": 0.11927181418706842,
"grad_norm": 2.7586164474487305,
"learning_rate": 1e-05,
"loss": 0.9234,
"step": 6270
},
{
"epoch": 0.11946204036599517,
"grad_norm": 2.546567440032959,
"learning_rate": 1e-05,
"loss": 0.7825,
"step": 6280
},
{
"epoch": 0.1196522665449219,
"grad_norm": 4.157527923583984,
"learning_rate": 1e-05,
"loss": 0.7748,
"step": 6290
},
{
"epoch": 0.11984249272384866,
"grad_norm": 8.950289726257324,
"learning_rate": 1e-05,
"loss": 0.9233,
"step": 6300
},
{
"epoch": 0.1200327189027754,
"grad_norm": 3.151874542236328,
"learning_rate": 1e-05,
"loss": 0.9943,
"step": 6310
},
{
"epoch": 0.12022294508170214,
"grad_norm": 2.2693397998809814,
"learning_rate": 1e-05,
"loss": 0.9402,
"step": 6320
},
{
"epoch": 0.12041317126062889,
"grad_norm": 3.1120898723602295,
"learning_rate": 1e-05,
"loss": 0.7721,
"step": 6330
},
{
"epoch": 0.12060339743955563,
"grad_norm": 4.523067951202393,
"learning_rate": 1e-05,
"loss": 0.7868,
"step": 6340
},
{
"epoch": 0.12079362361848238,
"grad_norm": 9.654865264892578,
"learning_rate": 1e-05,
"loss": 0.9464,
"step": 6350
},
{
"epoch": 0.12098384979740912,
"grad_norm": 5.636833190917969,
"learning_rate": 1e-05,
"loss": 0.9955,
"step": 6360
},
{
"epoch": 0.12117407597633587,
"grad_norm": 2.3272182941436768,
"learning_rate": 1e-05,
"loss": 0.8959,
"step": 6370
},
{
"epoch": 0.1213643021552626,
"grad_norm": 2.844841718673706,
"learning_rate": 1e-05,
"loss": 0.7521,
"step": 6380
},
{
"epoch": 0.12155452833418935,
"grad_norm": 3.4328556060791016,
"learning_rate": 1e-05,
"loss": 0.7829,
"step": 6390
},
{
"epoch": 0.1217447545131161,
"grad_norm": 11.463356018066406,
"learning_rate": 1e-05,
"loss": 0.9494,
"step": 6400
},
{
"epoch": 0.12193498069204284,
"grad_norm": 2.4653143882751465,
"learning_rate": 1e-05,
"loss": 0.9929,
"step": 6410
},
{
"epoch": 0.12212520687096959,
"grad_norm": 2.027087926864624,
"learning_rate": 1e-05,
"loss": 0.902,
"step": 6420
},
{
"epoch": 0.12231543304989632,
"grad_norm": 3.557056427001953,
"learning_rate": 1e-05,
"loss": 0.7847,
"step": 6430
},
{
"epoch": 0.12250565922882307,
"grad_norm": 4.602181911468506,
"learning_rate": 1e-05,
"loss": 0.774,
"step": 6440
},
{
"epoch": 0.12269588540774981,
"grad_norm": 8.64316463470459,
"learning_rate": 1e-05,
"loss": 1.0009,
"step": 6450
},
{
"epoch": 0.12288611158667656,
"grad_norm": 1.7479580640792847,
"learning_rate": 1e-05,
"loss": 0.9745,
"step": 6460
},
{
"epoch": 0.1230763377656033,
"grad_norm": 2.0574121475219727,
"learning_rate": 1e-05,
"loss": 0.9184,
"step": 6470
},
{
"epoch": 0.12326656394453005,
"grad_norm": 3.0321197509765625,
"learning_rate": 1e-05,
"loss": 0.8146,
"step": 6480
},
{
"epoch": 0.12345679012345678,
"grad_norm": 4.120352745056152,
"learning_rate": 1e-05,
"loss": 0.7856,
"step": 6490
},
{
"epoch": 0.12364701630238353,
"grad_norm": 9.343111038208008,
"learning_rate": 1e-05,
"loss": 0.866,
"step": 6500
},
{
"epoch": 0.12383724248131028,
"grad_norm": 2.2107670307159424,
"learning_rate": 1e-05,
"loss": 0.9627,
"step": 6510
},
{
"epoch": 0.12402746866023702,
"grad_norm": 1.997926115989685,
"learning_rate": 1e-05,
"loss": 0.9453,
"step": 6520
},
{
"epoch": 0.12421769483916377,
"grad_norm": 3.56750226020813,
"learning_rate": 1e-05,
"loss": 0.7545,
"step": 6530
},
{
"epoch": 0.1244079210180905,
"grad_norm": 5.119503021240234,
"learning_rate": 1e-05,
"loss": 0.7673,
"step": 6540
},
{
"epoch": 0.12459814719701726,
"grad_norm": 10.422179222106934,
"learning_rate": 1e-05,
"loss": 0.9992,
"step": 6550
},
{
"epoch": 0.12478837337594399,
"grad_norm": 2.8418734073638916,
"learning_rate": 1e-05,
"loss": 1.0037,
"step": 6560
},
{
"epoch": 0.12497859955487074,
"grad_norm": 2.3695666790008545,
"learning_rate": 1e-05,
"loss": 0.9351,
"step": 6570
},
{
"epoch": 0.12516882573379748,
"grad_norm": 2.5930049419403076,
"learning_rate": 1e-05,
"loss": 0.7711,
"step": 6580
},
{
"epoch": 0.12535905191272423,
"grad_norm": 4.580657005310059,
"learning_rate": 1e-05,
"loss": 0.7722,
"step": 6590
},
{
"epoch": 0.12554927809165098,
"grad_norm": 9.7059907913208,
"learning_rate": 1e-05,
"loss": 0.937,
"step": 6600
},
{
"epoch": 0.12573950427057773,
"grad_norm": 2.2432050704956055,
"learning_rate": 1e-05,
"loss": 1.0081,
"step": 6610
},
{
"epoch": 0.12592973044950445,
"grad_norm": 2.191602945327759,
"learning_rate": 1e-05,
"loss": 0.8972,
"step": 6620
},
{
"epoch": 0.1261199566284312,
"grad_norm": 3.507627010345459,
"learning_rate": 1e-05,
"loss": 0.7619,
"step": 6630
},
{
"epoch": 0.12631018280735795,
"grad_norm": 3.8176779747009277,
"learning_rate": 1e-05,
"loss": 0.7549,
"step": 6640
},
{
"epoch": 0.1265004089862847,
"grad_norm": 9.30078125,
"learning_rate": 1e-05,
"loss": 0.9071,
"step": 6650
},
{
"epoch": 0.12669063516521142,
"grad_norm": 2.264244556427002,
"learning_rate": 1e-05,
"loss": 0.9968,
"step": 6660
},
{
"epoch": 0.12688086134413817,
"grad_norm": 2.0131940841674805,
"learning_rate": 1e-05,
"loss": 0.8852,
"step": 6670
},
{
"epoch": 0.12707108752306492,
"grad_norm": 3.041452407836914,
"learning_rate": 1e-05,
"loss": 0.7694,
"step": 6680
},
{
"epoch": 0.12726131370199167,
"grad_norm": 4.399232387542725,
"learning_rate": 1e-05,
"loss": 0.7868,
"step": 6690
},
{
"epoch": 0.12745153988091842,
"grad_norm": 10.927412033081055,
"learning_rate": 1e-05,
"loss": 0.9408,
"step": 6700
},
{
"epoch": 0.12764176605984515,
"grad_norm": 2.2591021060943604,
"learning_rate": 1e-05,
"loss": 0.976,
"step": 6710
},
{
"epoch": 0.1278319922387719,
"grad_norm": 2.965693473815918,
"learning_rate": 1e-05,
"loss": 0.9298,
"step": 6720
},
{
"epoch": 0.12802221841769865,
"grad_norm": 3.462883949279785,
"learning_rate": 1e-05,
"loss": 0.8064,
"step": 6730
},
{
"epoch": 0.1282124445966254,
"grad_norm": 3.5158207416534424,
"learning_rate": 1e-05,
"loss": 0.7598,
"step": 6740
},
{
"epoch": 0.12840267077555212,
"grad_norm": 9.568047523498535,
"learning_rate": 1e-05,
"loss": 0.9074,
"step": 6750
},
{
"epoch": 0.12859289695447887,
"grad_norm": 2.028777599334717,
"learning_rate": 1e-05,
"loss": 0.9531,
"step": 6760
},
{
"epoch": 0.12878312313340562,
"grad_norm": 2.2654993534088135,
"learning_rate": 1e-05,
"loss": 0.9049,
"step": 6770
},
{
"epoch": 0.12897334931233237,
"grad_norm": 2.993471145629883,
"learning_rate": 1e-05,
"loss": 0.7565,
"step": 6780
},
{
"epoch": 0.12916357549125912,
"grad_norm": 3.730862617492676,
"learning_rate": 1e-05,
"loss": 0.7777,
"step": 6790
},
{
"epoch": 0.12935380167018584,
"grad_norm": 13.323833465576172,
"learning_rate": 1e-05,
"loss": 0.9326,
"step": 6800
},
{
"epoch": 0.1295440278491126,
"grad_norm": 3.4739651679992676,
"learning_rate": 1e-05,
"loss": 0.9834,
"step": 6810
},
{
"epoch": 0.12973425402803934,
"grad_norm": 2.4134697914123535,
"learning_rate": 1e-05,
"loss": 0.9505,
"step": 6820
},
{
"epoch": 0.1299244802069661,
"grad_norm": 2.6052565574645996,
"learning_rate": 1e-05,
"loss": 0.7357,
"step": 6830
},
{
"epoch": 0.13011470638589281,
"grad_norm": 3.5280425548553467,
"learning_rate": 1e-05,
"loss": 0.7585,
"step": 6840
},
{
"epoch": 0.13030493256481956,
"grad_norm": 9.52997875213623,
"learning_rate": 1e-05,
"loss": 0.9279,
"step": 6850
},
{
"epoch": 0.13049515874374631,
"grad_norm": 1.4204837083816528,
"learning_rate": 1e-05,
"loss": 0.9548,
"step": 6860
},
{
"epoch": 0.13068538492267306,
"grad_norm": 2.220087766647339,
"learning_rate": 1e-05,
"loss": 0.879,
"step": 6870
},
{
"epoch": 0.13087561110159981,
"grad_norm": 3.0759003162384033,
"learning_rate": 1e-05,
"loss": 0.7423,
"step": 6880
},
{
"epoch": 0.13106583728052654,
"grad_norm": 4.504390716552734,
"learning_rate": 1e-05,
"loss": 0.7591,
"step": 6890
},
{
"epoch": 0.1312560634594533,
"grad_norm": 9.588232040405273,
"learning_rate": 1e-05,
"loss": 0.8487,
"step": 6900
},
{
"epoch": 0.13144628963838004,
"grad_norm": 1.95205819606781,
"learning_rate": 1e-05,
"loss": 1.0052,
"step": 6910
},
{
"epoch": 0.1316365158173068,
"grad_norm": 2.5651416778564453,
"learning_rate": 1e-05,
"loss": 0.8973,
"step": 6920
},
{
"epoch": 0.1318267419962335,
"grad_norm": 3.936758041381836,
"learning_rate": 1e-05,
"loss": 0.7852,
"step": 6930
},
{
"epoch": 0.13201696817516026,
"grad_norm": 3.85343337059021,
"learning_rate": 1e-05,
"loss": 0.7637,
"step": 6940
},
{
"epoch": 0.132207194354087,
"grad_norm": 8.253195762634277,
"learning_rate": 1e-05,
"loss": 0.9287,
"step": 6950
},
{
"epoch": 0.13239742053301376,
"grad_norm": 2.9102096557617188,
"learning_rate": 1e-05,
"loss": 0.932,
"step": 6960
},
{
"epoch": 0.1325876467119405,
"grad_norm": 2.1497514247894287,
"learning_rate": 1e-05,
"loss": 0.8979,
"step": 6970
},
{
"epoch": 0.13277787289086723,
"grad_norm": 3.1719298362731934,
"learning_rate": 1e-05,
"loss": 0.7428,
"step": 6980
},
{
"epoch": 0.13296809906979398,
"grad_norm": 4.168227672576904,
"learning_rate": 1e-05,
"loss": 0.7603,
"step": 6990
},
{
"epoch": 0.13315832524872073,
"grad_norm": 10.031818389892578,
"learning_rate": 1e-05,
"loss": 0.9581,
"step": 7000
},
{
"epoch": 0.13334855142764748,
"grad_norm": 2.893181800842285,
"learning_rate": 1e-05,
"loss": 0.966,
"step": 7010
},
{
"epoch": 0.1335387776065742,
"grad_norm": 1.9807931184768677,
"learning_rate": 1e-05,
"loss": 0.904,
"step": 7020
},
{
"epoch": 0.13372900378550096,
"grad_norm": 2.5196480751037598,
"learning_rate": 1e-05,
"loss": 0.7399,
"step": 7030
},
{
"epoch": 0.1339192299644277,
"grad_norm": 4.4074273109436035,
"learning_rate": 1e-05,
"loss": 0.7363,
"step": 7040
},
{
"epoch": 0.13410945614335446,
"grad_norm": 8.36719799041748,
"learning_rate": 1e-05,
"loss": 0.923,
"step": 7050
},
{
"epoch": 0.1342996823222812,
"grad_norm": 2.3951706886291504,
"learning_rate": 1e-05,
"loss": 0.9963,
"step": 7060
},
{
"epoch": 0.13448990850120793,
"grad_norm": 2.871250629425049,
"learning_rate": 1e-05,
"loss": 0.9313,
"step": 7070
},
{
"epoch": 0.13468013468013468,
"grad_norm": 2.757784366607666,
"learning_rate": 1e-05,
"loss": 0.7336,
"step": 7080
},
{
"epoch": 0.13487036085906143,
"grad_norm": 3.7014901638031006,
"learning_rate": 1e-05,
"loss": 0.7444,
"step": 7090
},
{
"epoch": 0.13506058703798818,
"grad_norm": 9.950662612915039,
"learning_rate": 1e-05,
"loss": 0.9171,
"step": 7100
},
{
"epoch": 0.1352508132169149,
"grad_norm": 2.358494758605957,
"learning_rate": 1e-05,
"loss": 0.9731,
"step": 7110
},
{
"epoch": 0.13544103939584165,
"grad_norm": 3.2696361541748047,
"learning_rate": 1e-05,
"loss": 0.9092,
"step": 7120
},
{
"epoch": 0.1356312655747684,
"grad_norm": 3.1684041023254395,
"learning_rate": 1e-05,
"loss": 0.7599,
"step": 7130
},
{
"epoch": 0.13582149175369515,
"grad_norm": 3.9581055641174316,
"learning_rate": 1e-05,
"loss": 0.7249,
"step": 7140
},
{
"epoch": 0.1360117179326219,
"grad_norm": 9.73874568939209,
"learning_rate": 1e-05,
"loss": 0.9604,
"step": 7150
},
{
"epoch": 0.13620194411154862,
"grad_norm": 2.1463685035705566,
"learning_rate": 1e-05,
"loss": 0.9504,
"step": 7160
},
{
"epoch": 0.13639217029047537,
"grad_norm": 2.0909180641174316,
"learning_rate": 1e-05,
"loss": 0.924,
"step": 7170
},
{
"epoch": 0.13658239646940212,
"grad_norm": 2.8187594413757324,
"learning_rate": 1e-05,
"loss": 0.7798,
"step": 7180
},
{
"epoch": 0.13677262264832887,
"grad_norm": 3.9886257648468018,
"learning_rate": 1e-05,
"loss": 0.7516,
"step": 7190
},
{
"epoch": 0.1369628488272556,
"grad_norm": 7.883686542510986,
"learning_rate": 1e-05,
"loss": 0.8497,
"step": 7200
},
{
"epoch": 0.13715307500618235,
"grad_norm": 2.2434940338134766,
"learning_rate": 1e-05,
"loss": 0.9753,
"step": 7210
},
{
"epoch": 0.1373433011851091,
"grad_norm": 1.9893772602081299,
"learning_rate": 1e-05,
"loss": 0.9292,
"step": 7220
},
{
"epoch": 0.13753352736403585,
"grad_norm": 2.490360975265503,
"learning_rate": 1e-05,
"loss": 0.7271,
"step": 7230
},
{
"epoch": 0.1377237535429626,
"grad_norm": 4.713914394378662,
"learning_rate": 1e-05,
"loss": 0.7262,
"step": 7240
},
{
"epoch": 0.13791397972188932,
"grad_norm": 10.599442481994629,
"learning_rate": 1e-05,
"loss": 0.9414,
"step": 7250
},
{
"epoch": 0.13810420590081607,
"grad_norm": 1.8025403022766113,
"learning_rate": 1e-05,
"loss": 0.9935,
"step": 7260
},
{
"epoch": 0.13829443207974282,
"grad_norm": 2.0245046615600586,
"learning_rate": 1e-05,
"loss": 0.9039,
"step": 7270
},
{
"epoch": 0.13848465825866957,
"grad_norm": 2.433586359024048,
"learning_rate": 1e-05,
"loss": 0.7233,
"step": 7280
},
{
"epoch": 0.1386748844375963,
"grad_norm": 3.7603626251220703,
"learning_rate": 1e-05,
"loss": 0.7435,
"step": 7290
},
{
"epoch": 0.13886511061652304,
"grad_norm": 9.949060440063477,
"learning_rate": 1e-05,
"loss": 0.9193,
"step": 7300
},
{
"epoch": 0.1390553367954498,
"grad_norm": 1.9665907621383667,
"learning_rate": 1e-05,
"loss": 0.9694,
"step": 7310
},
{
"epoch": 0.13924556297437654,
"grad_norm": 2.3955485820770264,
"learning_rate": 1e-05,
"loss": 0.8834,
"step": 7320
},
{
"epoch": 0.1394357891533033,
"grad_norm": 3.56719708442688,
"learning_rate": 1e-05,
"loss": 0.7467,
"step": 7330
},
{
"epoch": 0.13962601533223,
"grad_norm": 5.839688777923584,
"learning_rate": 1e-05,
"loss": 0.7826,
"step": 7340
},
{
"epoch": 0.13981624151115676,
"grad_norm": 8.73769760131836,
"learning_rate": 1e-05,
"loss": 0.8996,
"step": 7350
},
{
"epoch": 0.1400064676900835,
"grad_norm": 2.4210596084594727,
"learning_rate": 1e-05,
"loss": 0.9457,
"step": 7360
},
{
"epoch": 0.14019669386901026,
"grad_norm": 2.2819197177886963,
"learning_rate": 1e-05,
"loss": 0.9146,
"step": 7370
},
{
"epoch": 0.14038692004793699,
"grad_norm": 2.937675714492798,
"learning_rate": 1e-05,
"loss": 0.7391,
"step": 7380
},
{
"epoch": 0.14057714622686374,
"grad_norm": 3.7016258239746094,
"learning_rate": 1e-05,
"loss": 0.7802,
"step": 7390
},
{
"epoch": 0.1407673724057905,
"grad_norm": 11.490416526794434,
"learning_rate": 1e-05,
"loss": 0.9383,
"step": 7400
},
{
"epoch": 0.14095759858471724,
"grad_norm": 1.945784091949463,
"learning_rate": 1e-05,
"loss": 0.9442,
"step": 7410
},
{
"epoch": 0.141147824763644,
"grad_norm": 2.2445881366729736,
"learning_rate": 1e-05,
"loss": 0.8851,
"step": 7420
},
{
"epoch": 0.1413380509425707,
"grad_norm": 2.87397837638855,
"learning_rate": 1e-05,
"loss": 0.7344,
"step": 7430
},
{
"epoch": 0.14152827712149746,
"grad_norm": 3.980924129486084,
"learning_rate": 1e-05,
"loss": 0.7429,
"step": 7440
},
{
"epoch": 0.1417185033004242,
"grad_norm": 10.149578094482422,
"learning_rate": 1e-05,
"loss": 0.8999,
"step": 7450
},
{
"epoch": 0.14190872947935096,
"grad_norm": 1.817115306854248,
"learning_rate": 1e-05,
"loss": 0.9626,
"step": 7460
},
{
"epoch": 0.14209895565827768,
"grad_norm": 3.103394031524658,
"learning_rate": 1e-05,
"loss": 0.8718,
"step": 7470
},
{
"epoch": 0.14228918183720443,
"grad_norm": 2.643470287322998,
"learning_rate": 1e-05,
"loss": 0.7589,
"step": 7480
},
{
"epoch": 0.14247940801613118,
"grad_norm": 3.57576060295105,
"learning_rate": 1e-05,
"loss": 0.7347,
"step": 7490
},
{
"epoch": 0.14266963419505793,
"grad_norm": 9.406258583068848,
"learning_rate": 1e-05,
"loss": 0.8546,
"step": 7500
},
{
"epoch": 0.14285986037398465,
"grad_norm": 1.9683119058609009,
"learning_rate": 1e-05,
"loss": 0.9823,
"step": 7510
},
{
"epoch": 0.1430500865529114,
"grad_norm": 2.1903293132781982,
"learning_rate": 1e-05,
"loss": 0.8892,
"step": 7520
},
{
"epoch": 0.14324031273183815,
"grad_norm": 2.3271307945251465,
"learning_rate": 1e-05,
"loss": 0.7399,
"step": 7530
},
{
"epoch": 0.1434305389107649,
"grad_norm": 3.696786880493164,
"learning_rate": 1e-05,
"loss": 0.765,
"step": 7540
},
{
"epoch": 0.14362076508969165,
"grad_norm": 10.260873794555664,
"learning_rate": 1e-05,
"loss": 0.9512,
"step": 7550
},
{
"epoch": 0.14381099126861838,
"grad_norm": 2.8486692905426025,
"learning_rate": 1e-05,
"loss": 0.9529,
"step": 7560
},
{
"epoch": 0.14400121744754513,
"grad_norm": 1.8218411207199097,
"learning_rate": 1e-05,
"loss": 0.8659,
"step": 7570
},
{
"epoch": 0.14419144362647188,
"grad_norm": 2.820119857788086,
"learning_rate": 1e-05,
"loss": 0.739,
"step": 7580
},
{
"epoch": 0.14438166980539863,
"grad_norm": 3.814169406890869,
"learning_rate": 1e-05,
"loss": 0.7099,
"step": 7590
},
{
"epoch": 0.14457189598432535,
"grad_norm": 8.425736427307129,
"learning_rate": 1e-05,
"loss": 0.939,
"step": 7600
},
{
"epoch": 0.1447621221632521,
"grad_norm": 2.2623238563537598,
"learning_rate": 1e-05,
"loss": 0.9889,
"step": 7610
},
{
"epoch": 0.14495234834217885,
"grad_norm": 2.1133148670196533,
"learning_rate": 1e-05,
"loss": 0.8873,
"step": 7620
},
{
"epoch": 0.1451425745211056,
"grad_norm": 2.5928361415863037,
"learning_rate": 1e-05,
"loss": 0.7357,
"step": 7630
},
{
"epoch": 0.14533280070003235,
"grad_norm": 4.55255651473999,
"learning_rate": 1e-05,
"loss": 0.7393,
"step": 7640
},
{
"epoch": 0.14552302687895907,
"grad_norm": 9.313065528869629,
"learning_rate": 1e-05,
"loss": 0.9329,
"step": 7650
},
{
"epoch": 0.14571325305788582,
"grad_norm": 1.8915058374404907,
"learning_rate": 1e-05,
"loss": 0.9375,
"step": 7660
},
{
"epoch": 0.14590347923681257,
"grad_norm": 2.0189549922943115,
"learning_rate": 1e-05,
"loss": 0.8861,
"step": 7670
},
{
"epoch": 0.14609370541573932,
"grad_norm": 3.0433082580566406,
"learning_rate": 1e-05,
"loss": 0.7762,
"step": 7680
},
{
"epoch": 0.14628393159466604,
"grad_norm": 4.068163871765137,
"learning_rate": 1e-05,
"loss": 0.7496,
"step": 7690
},
{
"epoch": 0.1464741577735928,
"grad_norm": 9.165461540222168,
"learning_rate": 1e-05,
"loss": 0.9122,
"step": 7700
},
{
"epoch": 0.14666438395251954,
"grad_norm": 1.5687580108642578,
"learning_rate": 1e-05,
"loss": 0.9181,
"step": 7710
},
{
"epoch": 0.1468546101314463,
"grad_norm": 2.31510066986084,
"learning_rate": 1e-05,
"loss": 0.9334,
"step": 7720
},
{
"epoch": 0.14704483631037304,
"grad_norm": 2.3600876331329346,
"learning_rate": 1e-05,
"loss": 0.714,
"step": 7730
},
{
"epoch": 0.14723506248929977,
"grad_norm": 4.805253028869629,
"learning_rate": 1e-05,
"loss": 0.7549,
"step": 7740
},
{
"epoch": 0.14742528866822652,
"grad_norm": 8.319846153259277,
"learning_rate": 1e-05,
"loss": 0.8928,
"step": 7750
},
{
"epoch": 0.14761551484715327,
"grad_norm": 1.9584218263626099,
"learning_rate": 1e-05,
"loss": 0.9447,
"step": 7760
},
{
"epoch": 0.14780574102608002,
"grad_norm": 2.880420446395874,
"learning_rate": 1e-05,
"loss": 0.9119,
"step": 7770
},
{
"epoch": 0.14799596720500674,
"grad_norm": 3.016834020614624,
"learning_rate": 1e-05,
"loss": 0.7319,
"step": 7780
},
{
"epoch": 0.1481861933839335,
"grad_norm": 3.644460916519165,
"learning_rate": 1e-05,
"loss": 0.7329,
"step": 7790
},
{
"epoch": 0.14837641956286024,
"grad_norm": 10.653403282165527,
"learning_rate": 1e-05,
"loss": 0.8525,
"step": 7800
},
{
"epoch": 0.148566645741787,
"grad_norm": 1.8029781579971313,
"learning_rate": 1e-05,
"loss": 0.9418,
"step": 7810
},
{
"epoch": 0.14875687192071374,
"grad_norm": 2.104005813598633,
"learning_rate": 1e-05,
"loss": 0.8934,
"step": 7820
},
{
"epoch": 0.14894709809964046,
"grad_norm": 2.5508646965026855,
"learning_rate": 1e-05,
"loss": 0.752,
"step": 7830
},
{
"epoch": 0.1491373242785672,
"grad_norm": 3.7671616077423096,
"learning_rate": 1e-05,
"loss": 0.7303,
"step": 7840
},
{
"epoch": 0.14932755045749396,
"grad_norm": 10.065755844116211,
"learning_rate": 1e-05,
"loss": 0.8709,
"step": 7850
},
{
"epoch": 0.1495177766364207,
"grad_norm": 2.4404067993164062,
"learning_rate": 1e-05,
"loss": 0.9554,
"step": 7860
},
{
"epoch": 0.14970800281534744,
"grad_norm": 2.2561497688293457,
"learning_rate": 1e-05,
"loss": 0.8792,
"step": 7870
},
{
"epoch": 0.14989822899427419,
"grad_norm": 2.597222328186035,
"learning_rate": 1e-05,
"loss": 0.7474,
"step": 7880
},
{
"epoch": 0.15008845517320094,
"grad_norm": 3.484921455383301,
"learning_rate": 1e-05,
"loss": 0.7259,
"step": 7890
},
{
"epoch": 0.15027868135212769,
"grad_norm": 8.056356430053711,
"learning_rate": 1e-05,
"loss": 0.8913,
"step": 7900
},
{
"epoch": 0.15046890753105444,
"grad_norm": 1.918716549873352,
"learning_rate": 1e-05,
"loss": 0.9598,
"step": 7910
},
{
"epoch": 0.15065913370998116,
"grad_norm": 2.484280586242676,
"learning_rate": 1e-05,
"loss": 0.9302,
"step": 7920
},
{
"epoch": 0.1508493598889079,
"grad_norm": 2.303973436355591,
"learning_rate": 1e-05,
"loss": 0.7595,
"step": 7930
},
{
"epoch": 0.15103958606783466,
"grad_norm": 3.4837100505828857,
"learning_rate": 1e-05,
"loss": 0.7496,
"step": 7940
},
{
"epoch": 0.1512298122467614,
"grad_norm": 8.684762001037598,
"learning_rate": 1e-05,
"loss": 0.8852,
"step": 7950
},
{
"epoch": 0.15142003842568813,
"grad_norm": 1.741068720817566,
"learning_rate": 1e-05,
"loss": 0.9448,
"step": 7960
},
{
"epoch": 0.15161026460461488,
"grad_norm": 1.7397018671035767,
"learning_rate": 1e-05,
"loss": 0.8763,
"step": 7970
},
{
"epoch": 0.15180049078354163,
"grad_norm": 2.4503140449523926,
"learning_rate": 1e-05,
"loss": 0.7383,
"step": 7980
},
{
"epoch": 0.15199071696246838,
"grad_norm": 3.6492369174957275,
"learning_rate": 1e-05,
"loss": 0.7236,
"step": 7990
},
{
"epoch": 0.15218094314139513,
"grad_norm": 6.908052444458008,
"learning_rate": 1e-05,
"loss": 0.8749,
"step": 8000
},
{
"epoch": 0.15237116932032185,
"grad_norm": 2.341634511947632,
"learning_rate": 1e-05,
"loss": 0.928,
"step": 8010
},
{
"epoch": 0.1525613954992486,
"grad_norm": 1.9240483045578003,
"learning_rate": 1e-05,
"loss": 0.9184,
"step": 8020
},
{
"epoch": 0.15275162167817535,
"grad_norm": 2.744624137878418,
"learning_rate": 1e-05,
"loss": 0.6819,
"step": 8030
},
{
"epoch": 0.1529418478571021,
"grad_norm": 3.6284239292144775,
"learning_rate": 1e-05,
"loss": 0.6876,
"step": 8040
},
{
"epoch": 0.15313207403602883,
"grad_norm": 9.622000694274902,
"learning_rate": 1e-05,
"loss": 0.9168,
"step": 8050
},
{
"epoch": 0.15332230021495558,
"grad_norm": 2.9141807556152344,
"learning_rate": 1e-05,
"loss": 0.9517,
"step": 8060
},
{
"epoch": 0.15351252639388233,
"grad_norm": 2.321643590927124,
"learning_rate": 1e-05,
"loss": 0.9107,
"step": 8070
},
{
"epoch": 0.15370275257280908,
"grad_norm": 2.6926326751708984,
"learning_rate": 1e-05,
"loss": 0.7405,
"step": 8080
},
{
"epoch": 0.15389297875173583,
"grad_norm": 3.6644179821014404,
"learning_rate": 1e-05,
"loss": 0.7286,
"step": 8090
},
{
"epoch": 0.15408320493066255,
"grad_norm": 8.381803512573242,
"learning_rate": 1e-05,
"loss": 0.9141,
"step": 8100
},
{
"epoch": 0.1542734311095893,
"grad_norm": 2.255265712738037,
"learning_rate": 1e-05,
"loss": 0.9584,
"step": 8110
},
{
"epoch": 0.15446365728851605,
"grad_norm": 2.1524147987365723,
"learning_rate": 1e-05,
"loss": 0.9034,
"step": 8120
},
{
"epoch": 0.1546538834674428,
"grad_norm": 2.6393749713897705,
"learning_rate": 1e-05,
"loss": 0.7285,
"step": 8130
},
{
"epoch": 0.15484410964636952,
"grad_norm": 3.686825752258301,
"learning_rate": 1e-05,
"loss": 0.7418,
"step": 8140
},
{
"epoch": 0.15503433582529627,
"grad_norm": 8.094237327575684,
"learning_rate": 1e-05,
"loss": 0.8536,
"step": 8150
},
{
"epoch": 0.15522456200422302,
"grad_norm": 1.788849949836731,
"learning_rate": 1e-05,
"loss": 0.957,
"step": 8160
},
{
"epoch": 0.15541478818314977,
"grad_norm": 2.4223103523254395,
"learning_rate": 1e-05,
"loss": 0.9089,
"step": 8170
},
{
"epoch": 0.15560501436207652,
"grad_norm": 2.830402374267578,
"learning_rate": 1e-05,
"loss": 0.7133,
"step": 8180
},
{
"epoch": 0.15579524054100324,
"grad_norm": 3.6556434631347656,
"learning_rate": 1e-05,
"loss": 0.7285,
"step": 8190
},
{
"epoch": 0.15598546671993,
"grad_norm": 7.380302906036377,
"learning_rate": 1e-05,
"loss": 0.8263,
"step": 8200
},
{
"epoch": 0.15617569289885674,
"grad_norm": 1.9398459196090698,
"learning_rate": 1e-05,
"loss": 0.9365,
"step": 8210
},
{
"epoch": 0.1563659190777835,
"grad_norm": 2.962153434753418,
"learning_rate": 1e-05,
"loss": 0.9047,
"step": 8220
},
{
"epoch": 0.15655614525671022,
"grad_norm": 4.362739562988281,
"learning_rate": 1e-05,
"loss": 0.7256,
"step": 8230
},
{
"epoch": 0.15674637143563697,
"grad_norm": 4.4467010498046875,
"learning_rate": 1e-05,
"loss": 0.7371,
"step": 8240
},
{
"epoch": 0.15693659761456372,
"grad_norm": 8.591287612915039,
"learning_rate": 1e-05,
"loss": 0.9162,
"step": 8250
},
{
"epoch": 0.15712682379349047,
"grad_norm": 3.129237651824951,
"learning_rate": 1e-05,
"loss": 0.9463,
"step": 8260
},
{
"epoch": 0.15731704997241722,
"grad_norm": 2.2711048126220703,
"learning_rate": 1e-05,
"loss": 0.8854,
"step": 8270
},
{
"epoch": 0.15750727615134394,
"grad_norm": 2.182875394821167,
"learning_rate": 1e-05,
"loss": 0.7583,
"step": 8280
},
{
"epoch": 0.1576975023302707,
"grad_norm": 3.1378896236419678,
"learning_rate": 1e-05,
"loss": 0.7369,
"step": 8290
},
{
"epoch": 0.15788772850919744,
"grad_norm": 8.022656440734863,
"learning_rate": 1e-05,
"loss": 0.898,
"step": 8300
},
{
"epoch": 0.1580779546881242,
"grad_norm": 2.3261232376098633,
"learning_rate": 1e-05,
"loss": 0.9574,
"step": 8310
},
{
"epoch": 0.1582681808670509,
"grad_norm": 2.309971809387207,
"learning_rate": 1e-05,
"loss": 0.8824,
"step": 8320
},
{
"epoch": 0.15845840704597766,
"grad_norm": 2.7919437885284424,
"learning_rate": 1e-05,
"loss": 0.7214,
"step": 8330
},
{
"epoch": 0.1586486332249044,
"grad_norm": 3.387028932571411,
"learning_rate": 1e-05,
"loss": 0.707,
"step": 8340
},
{
"epoch": 0.15883885940383116,
"grad_norm": 7.674351215362549,
"learning_rate": 1e-05,
"loss": 0.8624,
"step": 8350
},
{
"epoch": 0.1590290855827579,
"grad_norm": 1.8700567483901978,
"learning_rate": 1e-05,
"loss": 0.9195,
"step": 8360
},
{
"epoch": 0.15921931176168463,
"grad_norm": 2.174476385116577,
"learning_rate": 1e-05,
"loss": 0.8863,
"step": 8370
},
{
"epoch": 0.15940953794061138,
"grad_norm": 2.63166880607605,
"learning_rate": 1e-05,
"loss": 0.7617,
"step": 8380
},
{
"epoch": 0.15959976411953813,
"grad_norm": 3.3343350887298584,
"learning_rate": 1e-05,
"loss": 0.7556,
"step": 8390
},
{
"epoch": 0.15978999029846488,
"grad_norm": 8.57804012298584,
"learning_rate": 1e-05,
"loss": 0.8198,
"step": 8400
},
{
"epoch": 0.1599802164773916,
"grad_norm": 2.156667947769165,
"learning_rate": 1e-05,
"loss": 0.938,
"step": 8410
},
{
"epoch": 0.16017044265631836,
"grad_norm": 2.5154175758361816,
"learning_rate": 1e-05,
"loss": 0.8755,
"step": 8420
},
{
"epoch": 0.1603606688352451,
"grad_norm": 2.3889803886413574,
"learning_rate": 1e-05,
"loss": 0.7032,
"step": 8430
},
{
"epoch": 0.16055089501417186,
"grad_norm": 3.2766594886779785,
"learning_rate": 1e-05,
"loss": 0.7127,
"step": 8440
},
{
"epoch": 0.1607411211930986,
"grad_norm": 9.414206504821777,
"learning_rate": 1e-05,
"loss": 0.9007,
"step": 8450
},
{
"epoch": 0.16093134737202533,
"grad_norm": 1.7306021451950073,
"learning_rate": 1e-05,
"loss": 0.9316,
"step": 8460
},
{
"epoch": 0.16112157355095208,
"grad_norm": 1.7641631364822388,
"learning_rate": 1e-05,
"loss": 0.8551,
"step": 8470
},
{
"epoch": 0.16131179972987883,
"grad_norm": 2.6620235443115234,
"learning_rate": 1e-05,
"loss": 0.7133,
"step": 8480
},
{
"epoch": 0.16150202590880558,
"grad_norm": 3.5259556770324707,
"learning_rate": 1e-05,
"loss": 0.7205,
"step": 8490
},
{
"epoch": 0.1616922520877323,
"grad_norm": 9.95064640045166,
"learning_rate": 1e-05,
"loss": 0.9022,
"step": 8500
},
{
"epoch": 0.16188247826665905,
"grad_norm": 2.9001455307006836,
"learning_rate": 1e-05,
"loss": 0.9695,
"step": 8510
},
{
"epoch": 0.1620727044455858,
"grad_norm": 1.9286770820617676,
"learning_rate": 1e-05,
"loss": 0.9122,
"step": 8520
},
{
"epoch": 0.16226293062451255,
"grad_norm": 2.233778953552246,
"learning_rate": 1e-05,
"loss": 0.7198,
"step": 8530
},
{
"epoch": 0.1624531568034393,
"grad_norm": 3.4565389156341553,
"learning_rate": 1e-05,
"loss": 0.7234,
"step": 8540
},
{
"epoch": 0.16264338298236602,
"grad_norm": 9.445456504821777,
"learning_rate": 1e-05,
"loss": 0.9314,
"step": 8550
},
{
"epoch": 0.16283360916129277,
"grad_norm": 1.7565070390701294,
"learning_rate": 1e-05,
"loss": 0.9528,
"step": 8560
},
{
"epoch": 0.16302383534021953,
"grad_norm": 2.12056303024292,
"learning_rate": 1e-05,
"loss": 0.8396,
"step": 8570
},
{
"epoch": 0.16321406151914628,
"grad_norm": 2.5299785137176514,
"learning_rate": 1e-05,
"loss": 0.7643,
"step": 8580
},
{
"epoch": 0.163404287698073,
"grad_norm": 4.489746570587158,
"learning_rate": 1e-05,
"loss": 0.7293,
"step": 8590
},
{
"epoch": 0.16359451387699975,
"grad_norm": 8.609288215637207,
"learning_rate": 1e-05,
"loss": 0.8638,
"step": 8600
},
{
"epoch": 0.1637847400559265,
"grad_norm": 2.2143959999084473,
"learning_rate": 1e-05,
"loss": 0.9491,
"step": 8610
},
{
"epoch": 0.16397496623485325,
"grad_norm": 2.3459243774414062,
"learning_rate": 1e-05,
"loss": 0.9087,
"step": 8620
},
{
"epoch": 0.16416519241378,
"grad_norm": 2.3523526191711426,
"learning_rate": 1e-05,
"loss": 0.7329,
"step": 8630
},
{
"epoch": 0.16435541859270672,
"grad_norm": 5.09869909286499,
"learning_rate": 1e-05,
"loss": 0.7398,
"step": 8640
},
{
"epoch": 0.16454564477163347,
"grad_norm": 7.432270526885986,
"learning_rate": 1e-05,
"loss": 0.8723,
"step": 8650
},
{
"epoch": 0.16473587095056022,
"grad_norm": 1.9303075075149536,
"learning_rate": 1e-05,
"loss": 0.9571,
"step": 8660
},
{
"epoch": 0.16492609712948697,
"grad_norm": 2.0262880325317383,
"learning_rate": 1e-05,
"loss": 0.8436,
"step": 8670
},
{
"epoch": 0.1651163233084137,
"grad_norm": 3.6900765895843506,
"learning_rate": 1e-05,
"loss": 0.7486,
"step": 8680
},
{
"epoch": 0.16530654948734044,
"grad_norm": 3.4136927127838135,
"learning_rate": 1e-05,
"loss": 0.7499,
"step": 8690
},
{
"epoch": 0.1654967756662672,
"grad_norm": 8.807273864746094,
"learning_rate": 1e-05,
"loss": 0.9128,
"step": 8700
},
{
"epoch": 0.16568700184519394,
"grad_norm": 1.9169867038726807,
"learning_rate": 1e-05,
"loss": 0.943,
"step": 8710
},
{
"epoch": 0.1658772280241207,
"grad_norm": 2.6366212368011475,
"learning_rate": 1e-05,
"loss": 0.8563,
"step": 8720
},
{
"epoch": 0.16606745420304742,
"grad_norm": 2.97066068649292,
"learning_rate": 1e-05,
"loss": 0.7313,
"step": 8730
},
{
"epoch": 0.16625768038197417,
"grad_norm": 3.364522933959961,
"learning_rate": 1e-05,
"loss": 0.7164,
"step": 8740
},
{
"epoch": 0.16644790656090092,
"grad_norm": 8.377469062805176,
"learning_rate": 1e-05,
"loss": 0.8551,
"step": 8750
},
{
"epoch": 0.16663813273982767,
"grad_norm": 1.9407504796981812,
"learning_rate": 1e-05,
"loss": 0.9249,
"step": 8760
},
{
"epoch": 0.1668283589187544,
"grad_norm": 1.4594062566757202,
"learning_rate": 1e-05,
"loss": 0.8756,
"step": 8770
},
{
"epoch": 0.16701858509768114,
"grad_norm": 2.8266334533691406,
"learning_rate": 1e-05,
"loss": 0.7437,
"step": 8780
},
{
"epoch": 0.1672088112766079,
"grad_norm": 4.619261264801025,
"learning_rate": 1e-05,
"loss": 0.7063,
"step": 8790
},
{
"epoch": 0.16739903745553464,
"grad_norm": 9.557561874389648,
"learning_rate": 1e-05,
"loss": 0.8977,
"step": 8800
},
{
"epoch": 0.1675892636344614,
"grad_norm": 3.0954744815826416,
"learning_rate": 1e-05,
"loss": 0.9476,
"step": 8810
},
{
"epoch": 0.1677794898133881,
"grad_norm": 1.9986836910247803,
"learning_rate": 1e-05,
"loss": 0.9016,
"step": 8820
},
{
"epoch": 0.16796971599231486,
"grad_norm": 2.36448073387146,
"learning_rate": 1e-05,
"loss": 0.7133,
"step": 8830
},
{
"epoch": 0.1681599421712416,
"grad_norm": 3.5756449699401855,
"learning_rate": 1e-05,
"loss": 0.7154,
"step": 8840
},
{
"epoch": 0.16835016835016836,
"grad_norm": 7.382334232330322,
"learning_rate": 1e-05,
"loss": 0.8213,
"step": 8850
},
{
"epoch": 0.16854039452909508,
"grad_norm": 2.1712677478790283,
"learning_rate": 1e-05,
"loss": 0.9437,
"step": 8860
},
{
"epoch": 0.16873062070802183,
"grad_norm": 2.593846559524536,
"learning_rate": 1e-05,
"loss": 0.9031,
"step": 8870
},
{
"epoch": 0.16892084688694858,
"grad_norm": 2.551229953765869,
"learning_rate": 1e-05,
"loss": 0.7129,
"step": 8880
},
{
"epoch": 0.16911107306587533,
"grad_norm": 2.7769222259521484,
"learning_rate": 1e-05,
"loss": 0.6873,
"step": 8890
},
{
"epoch": 0.16930129924480206,
"grad_norm": 9.250907897949219,
"learning_rate": 1e-05,
"loss": 0.8555,
"step": 8900
},
{
"epoch": 0.1694915254237288,
"grad_norm": 2.5075912475585938,
"learning_rate": 1e-05,
"loss": 0.9223,
"step": 8910
},
{
"epoch": 0.16968175160265556,
"grad_norm": 2.4540295600891113,
"learning_rate": 1e-05,
"loss": 0.8999,
"step": 8920
},
{
"epoch": 0.1698719777815823,
"grad_norm": 3.2890870571136475,
"learning_rate": 1e-05,
"loss": 0.7297,
"step": 8930
},
{
"epoch": 0.17006220396050906,
"grad_norm": 3.711933135986328,
"learning_rate": 1e-05,
"loss": 0.6689,
"step": 8940
},
{
"epoch": 0.17025243013943578,
"grad_norm": 8.429349899291992,
"learning_rate": 1e-05,
"loss": 0.8658,
"step": 8950
},
{
"epoch": 0.17044265631836253,
"grad_norm": 2.0525553226470947,
"learning_rate": 1e-05,
"loss": 0.9062,
"step": 8960
},
{
"epoch": 0.17063288249728928,
"grad_norm": 2.2169885635375977,
"learning_rate": 1e-05,
"loss": 0.8231,
"step": 8970
},
{
"epoch": 0.17082310867621603,
"grad_norm": 2.522550582885742,
"learning_rate": 1e-05,
"loss": 0.7469,
"step": 8980
},
{
"epoch": 0.17101333485514275,
"grad_norm": 3.1226565837860107,
"learning_rate": 1e-05,
"loss": 0.6757,
"step": 8990
},
{
"epoch": 0.1712035610340695,
"grad_norm": 9.755632400512695,
"learning_rate": 1e-05,
"loss": 0.8663,
"step": 9000
},
{
"epoch": 0.17139378721299625,
"grad_norm": 1.725836157798767,
"learning_rate": 1e-05,
"loss": 0.9309,
"step": 9010
},
{
"epoch": 0.171584013391923,
"grad_norm": 2.495276927947998,
"learning_rate": 1e-05,
"loss": 0.9254,
"step": 9020
},
{
"epoch": 0.17177423957084975,
"grad_norm": 2.578331470489502,
"learning_rate": 1e-05,
"loss": 0.7242,
"step": 9030
},
{
"epoch": 0.17196446574977647,
"grad_norm": 3.5339763164520264,
"learning_rate": 1e-05,
"loss": 0.6829,
"step": 9040
},
{
"epoch": 0.17215469192870322,
"grad_norm": 8.083335876464844,
"learning_rate": 1e-05,
"loss": 0.8474,
"step": 9050
},
{
"epoch": 0.17234491810762997,
"grad_norm": 2.3775265216827393,
"learning_rate": 1e-05,
"loss": 0.9268,
"step": 9060
},
{
"epoch": 0.17253514428655672,
"grad_norm": 1.87567937374115,
"learning_rate": 1e-05,
"loss": 0.853,
"step": 9070
},
{
"epoch": 0.17272537046548345,
"grad_norm": 2.3250389099121094,
"learning_rate": 1e-05,
"loss": 0.7393,
"step": 9080
},
{
"epoch": 0.1729155966444102,
"grad_norm": 3.2631142139434814,
"learning_rate": 1e-05,
"loss": 0.692,
"step": 9090
},
{
"epoch": 0.17310582282333695,
"grad_norm": 7.674604892730713,
"learning_rate": 1e-05,
"loss": 0.8567,
"step": 9100
},
{
"epoch": 0.1732960490022637,
"grad_norm": 2.177569627761841,
"learning_rate": 1e-05,
"loss": 0.9285,
"step": 9110
},
{
"epoch": 0.17348627518119045,
"grad_norm": 1.7919386625289917,
"learning_rate": 1e-05,
"loss": 0.8675,
"step": 9120
},
{
"epoch": 0.17367650136011717,
"grad_norm": 2.678382635116577,
"learning_rate": 1e-05,
"loss": 0.7594,
"step": 9130
},
{
"epoch": 0.17386672753904392,
"grad_norm": 5.0587239265441895,
"learning_rate": 1e-05,
"loss": 0.697,
"step": 9140
},
{
"epoch": 0.17405695371797067,
"grad_norm": 8.538005828857422,
"learning_rate": 1e-05,
"loss": 0.8431,
"step": 9150
},
{
"epoch": 0.17424717989689742,
"grad_norm": 1.460579514503479,
"learning_rate": 1e-05,
"loss": 0.8771,
"step": 9160
},
{
"epoch": 0.17443740607582414,
"grad_norm": 1.7728286981582642,
"learning_rate": 1e-05,
"loss": 0.8712,
"step": 9170
},
{
"epoch": 0.1746276322547509,
"grad_norm": 3.0225491523742676,
"learning_rate": 1e-05,
"loss": 0.7478,
"step": 9180
},
{
"epoch": 0.17481785843367764,
"grad_norm": 3.1023576259613037,
"learning_rate": 1e-05,
"loss": 0.7111,
"step": 9190
},
{
"epoch": 0.1750080846126044,
"grad_norm": 8.56110954284668,
"learning_rate": 1e-05,
"loss": 0.8342,
"step": 9200
},
{
"epoch": 0.17519831079153114,
"grad_norm": 2.0546131134033203,
"learning_rate": 1e-05,
"loss": 0.909,
"step": 9210
},
{
"epoch": 0.17538853697045786,
"grad_norm": 2.559865951538086,
"learning_rate": 1e-05,
"loss": 0.8631,
"step": 9220
},
{
"epoch": 0.17557876314938461,
"grad_norm": 2.4749162197113037,
"learning_rate": 1e-05,
"loss": 0.7308,
"step": 9230
},
{
"epoch": 0.17576898932831136,
"grad_norm": 4.281092643737793,
"learning_rate": 1e-05,
"loss": 0.6866,
"step": 9240
},
{
"epoch": 0.17595921550723811,
"grad_norm": 7.077030181884766,
"learning_rate": 1e-05,
"loss": 0.8617,
"step": 9250
},
{
"epoch": 0.17614944168616484,
"grad_norm": 1.7959973812103271,
"learning_rate": 1e-05,
"loss": 0.9331,
"step": 9260
},
{
"epoch": 0.1763396678650916,
"grad_norm": 2.3159968852996826,
"learning_rate": 1e-05,
"loss": 0.8711,
"step": 9270
},
{
"epoch": 0.17652989404401834,
"grad_norm": 2.575800657272339,
"learning_rate": 1e-05,
"loss": 0.7345,
"step": 9280
},
{
"epoch": 0.1767201202229451,
"grad_norm": 3.4078171253204346,
"learning_rate": 1e-05,
"loss": 0.7206,
"step": 9290
},
{
"epoch": 0.17691034640187184,
"grad_norm": 8.762450218200684,
"learning_rate": 1e-05,
"loss": 0.8692,
"step": 9300
},
{
"epoch": 0.17710057258079856,
"grad_norm": 2.1218979358673096,
"learning_rate": 1e-05,
"loss": 0.9229,
"step": 9310
},
{
"epoch": 0.1772907987597253,
"grad_norm": 2.0216665267944336,
"learning_rate": 1e-05,
"loss": 0.9111,
"step": 9320
},
{
"epoch": 0.17748102493865206,
"grad_norm": 3.9459786415100098,
"learning_rate": 1e-05,
"loss": 0.7021,
"step": 9330
},
{
"epoch": 0.1776712511175788,
"grad_norm": 3.8002572059631348,
"learning_rate": 1e-05,
"loss": 0.7225,
"step": 9340
},
{
"epoch": 0.17786147729650553,
"grad_norm": 9.027454376220703,
"learning_rate": 1e-05,
"loss": 0.8801,
"step": 9350
},
{
"epoch": 0.17805170347543228,
"grad_norm": 2.1301164627075195,
"learning_rate": 1e-05,
"loss": 0.9465,
"step": 9360
},
{
"epoch": 0.17824192965435903,
"grad_norm": 2.0452616214752197,
"learning_rate": 1e-05,
"loss": 0.8692,
"step": 9370
},
{
"epoch": 0.17843215583328578,
"grad_norm": 2.2002103328704834,
"learning_rate": 1e-05,
"loss": 0.7513,
"step": 9380
},
{
"epoch": 0.17862238201221253,
"grad_norm": 2.8925886154174805,
"learning_rate": 1e-05,
"loss": 0.6952,
"step": 9390
},
{
"epoch": 0.17881260819113926,
"grad_norm": 8.817564010620117,
"learning_rate": 1e-05,
"loss": 0.8474,
"step": 9400
},
{
"epoch": 0.179002834370066,
"grad_norm": 1.5694103240966797,
"learning_rate": 1e-05,
"loss": 0.9642,
"step": 9410
},
{
"epoch": 0.17919306054899276,
"grad_norm": 2.286940097808838,
"learning_rate": 1e-05,
"loss": 0.869,
"step": 9420
},
{
"epoch": 0.1793832867279195,
"grad_norm": 2.020352840423584,
"learning_rate": 1e-05,
"loss": 0.6763,
"step": 9430
},
{
"epoch": 0.17957351290684623,
"grad_norm": 3.237088203430176,
"learning_rate": 1e-05,
"loss": 0.7002,
"step": 9440
},
{
"epoch": 0.17976373908577298,
"grad_norm": 8.736642837524414,
"learning_rate": 1e-05,
"loss": 0.8431,
"step": 9450
},
{
"epoch": 0.17995396526469973,
"grad_norm": 1.9268271923065186,
"learning_rate": 1e-05,
"loss": 0.9151,
"step": 9460
},
{
"epoch": 0.18014419144362648,
"grad_norm": 2.028158664703369,
"learning_rate": 1e-05,
"loss": 0.8848,
"step": 9470
},
{
"epoch": 0.18033441762255323,
"grad_norm": 2.3369863033294678,
"learning_rate": 1e-05,
"loss": 0.708,
"step": 9480
},
{
"epoch": 0.18052464380147995,
"grad_norm": 3.3326122760772705,
"learning_rate": 1e-05,
"loss": 0.6853,
"step": 9490
},
{
"epoch": 0.1807148699804067,
"grad_norm": 7.981497287750244,
"learning_rate": 1e-05,
"loss": 0.8363,
"step": 9500
},
{
"epoch": 0.18090509615933345,
"grad_norm": 1.7365386486053467,
"learning_rate": 1e-05,
"loss": 0.927,
"step": 9510
},
{
"epoch": 0.1810953223382602,
"grad_norm": 2.0339159965515137,
"learning_rate": 1e-05,
"loss": 0.913,
"step": 9520
},
{
"epoch": 0.18128554851718692,
"grad_norm": 2.474303960800171,
"learning_rate": 1e-05,
"loss": 0.6947,
"step": 9530
},
{
"epoch": 0.18147577469611367,
"grad_norm": 4.409385681152344,
"learning_rate": 1e-05,
"loss": 0.7381,
"step": 9540
},
{
"epoch": 0.18166600087504042,
"grad_norm": 8.506572723388672,
"learning_rate": 1e-05,
"loss": 0.8328,
"step": 9550
},
{
"epoch": 0.18185622705396717,
"grad_norm": 1.7876067161560059,
"learning_rate": 1e-05,
"loss": 0.9413,
"step": 9560
},
{
"epoch": 0.18204645323289392,
"grad_norm": 2.1218321323394775,
"learning_rate": 1e-05,
"loss": 0.8329,
"step": 9570
},
{
"epoch": 0.18223667941182065,
"grad_norm": 2.97168231010437,
"learning_rate": 1e-05,
"loss": 0.7217,
"step": 9580
},
{
"epoch": 0.1824269055907474,
"grad_norm": 3.7951321601867676,
"learning_rate": 1e-05,
"loss": 0.7221,
"step": 9590
},
{
"epoch": 0.18261713176967415,
"grad_norm": 7.260524749755859,
"learning_rate": 1e-05,
"loss": 0.8523,
"step": 9600
},
{
"epoch": 0.1828073579486009,
"grad_norm": 1.7398523092269897,
"learning_rate": 1e-05,
"loss": 0.9023,
"step": 9610
},
{
"epoch": 0.18299758412752762,
"grad_norm": 1.8647395372390747,
"learning_rate": 1e-05,
"loss": 0.8748,
"step": 9620
},
{
"epoch": 0.18318781030645437,
"grad_norm": 2.751552104949951,
"learning_rate": 1e-05,
"loss": 0.6877,
"step": 9630
},
{
"epoch": 0.18337803648538112,
"grad_norm": 3.6920166015625,
"learning_rate": 1e-05,
"loss": 0.6974,
"step": 9640
},
{
"epoch": 0.18356826266430787,
"grad_norm": 8.56862735748291,
"learning_rate": 1e-05,
"loss": 0.8843,
"step": 9650
},
{
"epoch": 0.18375848884323462,
"grad_norm": 2.4071204662323,
"learning_rate": 1e-05,
"loss": 0.9086,
"step": 9660
},
{
"epoch": 0.18394871502216134,
"grad_norm": 2.1661462783813477,
"learning_rate": 1e-05,
"loss": 0.8484,
"step": 9670
},
{
"epoch": 0.1841389412010881,
"grad_norm": 2.463552474975586,
"learning_rate": 1e-05,
"loss": 0.702,
"step": 9680
},
{
"epoch": 0.18432916738001484,
"grad_norm": 3.848130702972412,
"learning_rate": 1e-05,
"loss": 0.7331,
"step": 9690
},
{
"epoch": 0.1845193935589416,
"grad_norm": 8.107160568237305,
"learning_rate": 1e-05,
"loss": 0.8518,
"step": 9700
},
{
"epoch": 0.1847096197378683,
"grad_norm": 2.2690823078155518,
"learning_rate": 1e-05,
"loss": 0.901,
"step": 9710
},
{
"epoch": 0.18489984591679506,
"grad_norm": 2.0350871086120605,
"learning_rate": 1e-05,
"loss": 0.8776,
"step": 9720
},
{
"epoch": 0.1850900720957218,
"grad_norm": 2.4634194374084473,
"learning_rate": 1e-05,
"loss": 0.6835,
"step": 9730
},
{
"epoch": 0.18528029827464856,
"grad_norm": 3.224806547164917,
"learning_rate": 1e-05,
"loss": 0.7068,
"step": 9740
},
{
"epoch": 0.18547052445357531,
"grad_norm": 9.610673904418945,
"learning_rate": 1e-05,
"loss": 0.8401,
"step": 9750
},
{
"epoch": 0.18566075063250204,
"grad_norm": 2.0625853538513184,
"learning_rate": 1e-05,
"loss": 0.9121,
"step": 9760
},
{
"epoch": 0.1858509768114288,
"grad_norm": 2.440350294113159,
"learning_rate": 1e-05,
"loss": 0.8798,
"step": 9770
},
{
"epoch": 0.18604120299035554,
"grad_norm": 2.7017195224761963,
"learning_rate": 1e-05,
"loss": 0.722,
"step": 9780
},
{
"epoch": 0.1862314291692823,
"grad_norm": 3.6038119792938232,
"learning_rate": 1e-05,
"loss": 0.7228,
"step": 9790
},
{
"epoch": 0.186421655348209,
"grad_norm": 7.01503849029541,
"learning_rate": 1e-05,
"loss": 0.7854,
"step": 9800
},
{
"epoch": 0.18661188152713576,
"grad_norm": 1.946079134941101,
"learning_rate": 1e-05,
"loss": 0.9159,
"step": 9810
},
{
"epoch": 0.1868021077060625,
"grad_norm": 1.9338854551315308,
"learning_rate": 1e-05,
"loss": 0.841,
"step": 9820
},
{
"epoch": 0.18699233388498926,
"grad_norm": 2.2061033248901367,
"learning_rate": 1e-05,
"loss": 0.7121,
"step": 9830
},
{
"epoch": 0.187182560063916,
"grad_norm": 3.604792356491089,
"learning_rate": 1e-05,
"loss": 0.6947,
"step": 9840
},
{
"epoch": 0.18737278624284273,
"grad_norm": 8.015079498291016,
"learning_rate": 1e-05,
"loss": 0.8018,
"step": 9850
},
{
"epoch": 0.18756301242176948,
"grad_norm": 1.7840076684951782,
"learning_rate": 1e-05,
"loss": 0.9221,
"step": 9860
},
{
"epoch": 0.18775323860069623,
"grad_norm": 1.8281898498535156,
"learning_rate": 1e-05,
"loss": 0.8306,
"step": 9870
},
{
"epoch": 0.18794346477962298,
"grad_norm": 2.212019205093384,
"learning_rate": 1e-05,
"loss": 0.7094,
"step": 9880
},
{
"epoch": 0.1881336909585497,
"grad_norm": 3.6437084674835205,
"learning_rate": 1e-05,
"loss": 0.6616,
"step": 9890
},
{
"epoch": 0.18832391713747645,
"grad_norm": 6.506216526031494,
"learning_rate": 1e-05,
"loss": 0.8494,
"step": 9900
},
{
"epoch": 0.1885141433164032,
"grad_norm": 1.7926994562149048,
"learning_rate": 1e-05,
"loss": 0.9581,
"step": 9910
},
{
"epoch": 0.18870436949532995,
"grad_norm": 2.541463851928711,
"learning_rate": 1e-05,
"loss": 0.8566,
"step": 9920
},
{
"epoch": 0.1888945956742567,
"grad_norm": 2.160651922225952,
"learning_rate": 1e-05,
"loss": 0.7317,
"step": 9930
},
{
"epoch": 0.18908482185318343,
"grad_norm": 3.0525214672088623,
"learning_rate": 1e-05,
"loss": 0.689,
"step": 9940
},
{
"epoch": 0.18927504803211018,
"grad_norm": 8.882493019104004,
"learning_rate": 1e-05,
"loss": 0.8568,
"step": 9950
},
{
"epoch": 0.18946527421103693,
"grad_norm": 1.8088182210922241,
"learning_rate": 1e-05,
"loss": 0.939,
"step": 9960
},
{
"epoch": 0.18965550038996368,
"grad_norm": 2.463989496231079,
"learning_rate": 1e-05,
"loss": 0.8348,
"step": 9970
},
{
"epoch": 0.1898457265688904,
"grad_norm": 2.688507556915283,
"learning_rate": 1e-05,
"loss": 0.7359,
"step": 9980
},
{
"epoch": 0.19003595274781715,
"grad_norm": 3.7007699012756348,
"learning_rate": 1e-05,
"loss": 0.7183,
"step": 9990
},
{
"epoch": 0.1902261789267439,
"grad_norm": 7.293190002441406,
"learning_rate": 1e-05,
"loss": 0.8481,
"step": 10000
},
{
"epoch": 0.19041640510567065,
"grad_norm": 1.6456226110458374,
"learning_rate": 1e-05,
"loss": 0.9294,
"step": 10010
},
{
"epoch": 0.1906066312845974,
"grad_norm": 2.9795262813568115,
"learning_rate": 1e-05,
"loss": 0.8775,
"step": 10020
},
{
"epoch": 0.19079685746352412,
"grad_norm": 2.4334216117858887,
"learning_rate": 1e-05,
"loss": 0.6956,
"step": 10030
},
{
"epoch": 0.19098708364245087,
"grad_norm": 3.2898142337799072,
"learning_rate": 1e-05,
"loss": 0.6851,
"step": 10040
},
{
"epoch": 0.19117730982137762,
"grad_norm": 8.827730178833008,
"learning_rate": 1e-05,
"loss": 0.8012,
"step": 10050
},
{
"epoch": 0.19136753600030437,
"grad_norm": 1.9843683242797852,
"learning_rate": 1e-05,
"loss": 0.9156,
"step": 10060
},
{
"epoch": 0.1915577621792311,
"grad_norm": 2.0664703845977783,
"learning_rate": 1e-05,
"loss": 0.8657,
"step": 10070
},
{
"epoch": 0.19174798835815784,
"grad_norm": 2.2624592781066895,
"learning_rate": 1e-05,
"loss": 0.6674,
"step": 10080
},
{
"epoch": 0.1919382145370846,
"grad_norm": 3.4906821250915527,
"learning_rate": 1e-05,
"loss": 0.6607,
"step": 10090
},
{
"epoch": 0.19212844071601134,
"grad_norm": 7.082588195800781,
"learning_rate": 1e-05,
"loss": 0.8551,
"step": 10100
},
{
"epoch": 0.1923186668949381,
"grad_norm": 2.082470178604126,
"learning_rate": 1e-05,
"loss": 0.8989,
"step": 10110
},
{
"epoch": 0.19250889307386482,
"grad_norm": 1.6501373052597046,
"learning_rate": 1e-05,
"loss": 0.8569,
"step": 10120
},
{
"epoch": 0.19269911925279157,
"grad_norm": 2.4363276958465576,
"learning_rate": 1e-05,
"loss": 0.7236,
"step": 10130
},
{
"epoch": 0.19288934543171832,
"grad_norm": 3.374803066253662,
"learning_rate": 1e-05,
"loss": 0.7006,
"step": 10140
},
{
"epoch": 0.19307957161064507,
"grad_norm": 8.738495826721191,
"learning_rate": 1e-05,
"loss": 0.8371,
"step": 10150
},
{
"epoch": 0.1932697977895718,
"grad_norm": 1.710397720336914,
"learning_rate": 1e-05,
"loss": 0.8992,
"step": 10160
},
{
"epoch": 0.19346002396849854,
"grad_norm": 2.1988213062286377,
"learning_rate": 1e-05,
"loss": 0.8717,
"step": 10170
},
{
"epoch": 0.1936502501474253,
"grad_norm": 2.576183319091797,
"learning_rate": 1e-05,
"loss": 0.6986,
"step": 10180
},
{
"epoch": 0.19384047632635204,
"grad_norm": 3.7219669818878174,
"learning_rate": 1e-05,
"loss": 0.6656,
"step": 10190
},
{
"epoch": 0.19403070250527876,
"grad_norm": 8.551814079284668,
"learning_rate": 1e-05,
"loss": 0.8986,
"step": 10200
},
{
"epoch": 0.1942209286842055,
"grad_norm": 2.1695873737335205,
"learning_rate": 1e-05,
"loss": 0.9325,
"step": 10210
},
{
"epoch": 0.19441115486313226,
"grad_norm": 2.070331335067749,
"learning_rate": 1e-05,
"loss": 0.8373,
"step": 10220
},
{
"epoch": 0.194601381042059,
"grad_norm": 2.5308446884155273,
"learning_rate": 1e-05,
"loss": 0.7873,
"step": 10230
},
{
"epoch": 0.19479160722098576,
"grad_norm": 2.967834711074829,
"learning_rate": 1e-05,
"loss": 0.7459,
"step": 10240
},
{
"epoch": 0.19498183339991249,
"grad_norm": 7.878716468811035,
"learning_rate": 1e-05,
"loss": 0.8654,
"step": 10250
},
{
"epoch": 0.19517205957883924,
"grad_norm": 1.727189064025879,
"learning_rate": 1e-05,
"loss": 0.8961,
"step": 10260
},
{
"epoch": 0.19536228575776599,
"grad_norm": 1.8486124277114868,
"learning_rate": 1e-05,
"loss": 0.8474,
"step": 10270
},
{
"epoch": 0.19555251193669274,
"grad_norm": 2.146573305130005,
"learning_rate": 1e-05,
"loss": 0.7126,
"step": 10280
},
{
"epoch": 0.19574273811561946,
"grad_norm": 3.1944596767425537,
"learning_rate": 1e-05,
"loss": 0.6653,
"step": 10290
},
{
"epoch": 0.1959329642945462,
"grad_norm": 6.696405410766602,
"learning_rate": 1e-05,
"loss": 0.815,
"step": 10300
},
{
"epoch": 0.19612319047347296,
"grad_norm": 1.8762474060058594,
"learning_rate": 1e-05,
"loss": 0.936,
"step": 10310
},
{
"epoch": 0.1963134166523997,
"grad_norm": 2.0236756801605225,
"learning_rate": 1e-05,
"loss": 0.8972,
"step": 10320
},
{
"epoch": 0.19650364283132646,
"grad_norm": 2.0975186824798584,
"learning_rate": 1e-05,
"loss": 0.7004,
"step": 10330
},
{
"epoch": 0.19669386901025318,
"grad_norm": 3.289216995239258,
"learning_rate": 1e-05,
"loss": 0.6818,
"step": 10340
},
{
"epoch": 0.19688409518917993,
"grad_norm": 8.432727813720703,
"learning_rate": 1e-05,
"loss": 0.866,
"step": 10350
},
{
"epoch": 0.19707432136810668,
"grad_norm": 1.8026350736618042,
"learning_rate": 1e-05,
"loss": 0.9483,
"step": 10360
},
{
"epoch": 0.19726454754703343,
"grad_norm": 2.2063632011413574,
"learning_rate": 1e-05,
"loss": 0.8123,
"step": 10370
},
{
"epoch": 0.19745477372596015,
"grad_norm": 2.476705551147461,
"learning_rate": 1e-05,
"loss": 0.6864,
"step": 10380
},
{
"epoch": 0.1976449999048869,
"grad_norm": 3.307602882385254,
"learning_rate": 1e-05,
"loss": 0.6489,
"step": 10390
},
{
"epoch": 0.19783522608381365,
"grad_norm": 8.887639045715332,
"learning_rate": 1e-05,
"loss": 0.8333,
"step": 10400
},
{
"epoch": 0.1980254522627404,
"grad_norm": 3.0926196575164795,
"learning_rate": 1e-05,
"loss": 0.9244,
"step": 10410
},
{
"epoch": 0.19821567844166715,
"grad_norm": 1.730660080909729,
"learning_rate": 1e-05,
"loss": 0.8555,
"step": 10420
},
{
"epoch": 0.19840590462059388,
"grad_norm": 2.3197708129882812,
"learning_rate": 1e-05,
"loss": 0.6777,
"step": 10430
},
{
"epoch": 0.19859613079952063,
"grad_norm": 3.3812198638916016,
"learning_rate": 1e-05,
"loss": 0.724,
"step": 10440
},
{
"epoch": 0.19878635697844738,
"grad_norm": 8.763368606567383,
"learning_rate": 1e-05,
"loss": 0.8285,
"step": 10450
},
{
"epoch": 0.19897658315737413,
"grad_norm": 2.35097599029541,
"learning_rate": 1e-05,
"loss": 0.9673,
"step": 10460
},
{
"epoch": 0.19916680933630085,
"grad_norm": 1.920229434967041,
"learning_rate": 1e-05,
"loss": 0.8442,
"step": 10470
},
{
"epoch": 0.1993570355152276,
"grad_norm": 2.342207908630371,
"learning_rate": 1e-05,
"loss": 0.6922,
"step": 10480
},
{
"epoch": 0.19954726169415435,
"grad_norm": 3.731473207473755,
"learning_rate": 1e-05,
"loss": 0.7068,
"step": 10490
},
{
"epoch": 0.1997374878730811,
"grad_norm": 7.806982040405273,
"learning_rate": 1e-05,
"loss": 0.8493,
"step": 10500
},
{
"epoch": 0.19992771405200785,
"grad_norm": 1.723568081855774,
"learning_rate": 1e-05,
"loss": 0.8976,
"step": 10510
},
{
"epoch": 0.20011794023093457,
"grad_norm": 2.171288251876831,
"learning_rate": 1e-05,
"loss": 0.8391,
"step": 10520
},
{
"epoch": 0.20030816640986132,
"grad_norm": 2.5706255435943604,
"learning_rate": 1e-05,
"loss": 0.7383,
"step": 10530
},
{
"epoch": 0.20049839258878807,
"grad_norm": 3.221074104309082,
"learning_rate": 1e-05,
"loss": 0.7178,
"step": 10540
},
{
"epoch": 0.20068861876771482,
"grad_norm": 9.466782569885254,
"learning_rate": 1e-05,
"loss": 0.8624,
"step": 10550
},
{
"epoch": 0.20087884494664154,
"grad_norm": 2.056270122528076,
"learning_rate": 1e-05,
"loss": 0.9207,
"step": 10560
},
{
"epoch": 0.2010690711255683,
"grad_norm": 2.054572820663452,
"learning_rate": 1e-05,
"loss": 0.9097,
"step": 10570
},
{
"epoch": 0.20125929730449504,
"grad_norm": 2.8823249340057373,
"learning_rate": 1e-05,
"loss": 0.7217,
"step": 10580
},
{
"epoch": 0.2014495234834218,
"grad_norm": 3.0235226154327393,
"learning_rate": 1e-05,
"loss": 0.693,
"step": 10590
},
{
"epoch": 0.20163974966234854,
"grad_norm": 13.29204273223877,
"learning_rate": 1e-05,
"loss": 0.8226,
"step": 10600
},
{
"epoch": 0.20182997584127527,
"grad_norm": 1.7713842391967773,
"learning_rate": 1e-05,
"loss": 0.9011,
"step": 10610
},
{
"epoch": 0.20202020202020202,
"grad_norm": 1.7549028396606445,
"learning_rate": 1e-05,
"loss": 0.8941,
"step": 10620
},
{
"epoch": 0.20221042819912877,
"grad_norm": 2.501962184906006,
"learning_rate": 1e-05,
"loss": 0.7105,
"step": 10630
},
{
"epoch": 0.20240065437805552,
"grad_norm": 3.16438364982605,
"learning_rate": 1e-05,
"loss": 0.634,
"step": 10640
},
{
"epoch": 0.20259088055698224,
"grad_norm": 8.223575592041016,
"learning_rate": 1e-05,
"loss": 0.8489,
"step": 10650
},
{
"epoch": 0.202781106735909,
"grad_norm": 1.5903781652450562,
"learning_rate": 1e-05,
"loss": 0.9112,
"step": 10660
},
{
"epoch": 0.20297133291483574,
"grad_norm": 1.9833863973617554,
"learning_rate": 1e-05,
"loss": 0.836,
"step": 10670
},
{
"epoch": 0.2031615590937625,
"grad_norm": 2.7223188877105713,
"learning_rate": 1e-05,
"loss": 0.7173,
"step": 10680
},
{
"epoch": 0.20335178527268924,
"grad_norm": 3.5782735347747803,
"learning_rate": 1e-05,
"loss": 0.6489,
"step": 10690
},
{
"epoch": 0.20354201145161596,
"grad_norm": 9.124385833740234,
"learning_rate": 1e-05,
"loss": 0.8517,
"step": 10700
},
{
"epoch": 0.2037322376305427,
"grad_norm": 1.43975830078125,
"learning_rate": 1e-05,
"loss": 0.9183,
"step": 10710
},
{
"epoch": 0.20392246380946946,
"grad_norm": 1.9790184497833252,
"learning_rate": 1e-05,
"loss": 0.8438,
"step": 10720
},
{
"epoch": 0.2041126899883962,
"grad_norm": 2.6500227451324463,
"learning_rate": 1e-05,
"loss": 0.6688,
"step": 10730
},
{
"epoch": 0.20430291616732293,
"grad_norm": 3.210527181625366,
"learning_rate": 1e-05,
"loss": 0.7114,
"step": 10740
},
{
"epoch": 0.20449314234624968,
"grad_norm": 8.70366096496582,
"learning_rate": 1e-05,
"loss": 0.8203,
"step": 10750
},
{
"epoch": 0.20468336852517643,
"grad_norm": 1.5692269802093506,
"learning_rate": 1e-05,
"loss": 0.9044,
"step": 10760
},
{
"epoch": 0.20487359470410318,
"grad_norm": 1.8724249601364136,
"learning_rate": 1e-05,
"loss": 0.8995,
"step": 10770
},
{
"epoch": 0.20506382088302993,
"grad_norm": 2.4753377437591553,
"learning_rate": 1e-05,
"loss": 0.6898,
"step": 10780
},
{
"epoch": 0.20525404706195666,
"grad_norm": 3.5420470237731934,
"learning_rate": 1e-05,
"loss": 0.6578,
"step": 10790
},
{
"epoch": 0.2054442732408834,
"grad_norm": 8.560863494873047,
"learning_rate": 1e-05,
"loss": 0.791,
"step": 10800
},
{
"epoch": 0.20563449941981016,
"grad_norm": 1.704290509223938,
"learning_rate": 1e-05,
"loss": 0.9089,
"step": 10810
},
{
"epoch": 0.2058247255987369,
"grad_norm": 2.0210866928100586,
"learning_rate": 1e-05,
"loss": 0.8616,
"step": 10820
},
{
"epoch": 0.20601495177766363,
"grad_norm": 2.2871251106262207,
"learning_rate": 1e-05,
"loss": 0.6824,
"step": 10830
},
{
"epoch": 0.20620517795659038,
"grad_norm": 2.8320086002349854,
"learning_rate": 1e-05,
"loss": 0.6457,
"step": 10840
},
{
"epoch": 0.20639540413551713,
"grad_norm": 8.928088188171387,
"learning_rate": 1e-05,
"loss": 0.8491,
"step": 10850
},
{
"epoch": 0.20658563031444388,
"grad_norm": 1.8365607261657715,
"learning_rate": 1e-05,
"loss": 0.8782,
"step": 10860
},
{
"epoch": 0.20677585649337063,
"grad_norm": 1.8146898746490479,
"learning_rate": 1e-05,
"loss": 0.8199,
"step": 10870
},
{
"epoch": 0.20696608267229735,
"grad_norm": 2.558598756790161,
"learning_rate": 1e-05,
"loss": 0.7011,
"step": 10880
},
{
"epoch": 0.2071563088512241,
"grad_norm": 3.4497337341308594,
"learning_rate": 1e-05,
"loss": 0.706,
"step": 10890
},
{
"epoch": 0.20734653503015085,
"grad_norm": 7.7153706550598145,
"learning_rate": 1e-05,
"loss": 0.8093,
"step": 10900
},
{
"epoch": 0.2075367612090776,
"grad_norm": 1.5823546648025513,
"learning_rate": 1e-05,
"loss": 0.9059,
"step": 10910
},
{
"epoch": 0.20772698738800432,
"grad_norm": 1.872671365737915,
"learning_rate": 1e-05,
"loss": 0.8557,
"step": 10920
},
{
"epoch": 0.20791721356693108,
"grad_norm": 2.473158597946167,
"learning_rate": 1e-05,
"loss": 0.6862,
"step": 10930
},
{
"epoch": 0.20810743974585783,
"grad_norm": 3.515918731689453,
"learning_rate": 1e-05,
"loss": 0.7119,
"step": 10940
},
{
"epoch": 0.20829766592478458,
"grad_norm": 7.3194074630737305,
"learning_rate": 1e-05,
"loss": 0.8303,
"step": 10950
},
{
"epoch": 0.20848789210371133,
"grad_norm": 2.0247702598571777,
"learning_rate": 1e-05,
"loss": 0.9072,
"step": 10960
},
{
"epoch": 0.20867811828263805,
"grad_norm": 2.0145211219787598,
"learning_rate": 1e-05,
"loss": 0.8821,
"step": 10970
},
{
"epoch": 0.2088683444615648,
"grad_norm": 2.2125120162963867,
"learning_rate": 1e-05,
"loss": 0.6992,
"step": 10980
},
{
"epoch": 0.20905857064049155,
"grad_norm": 4.4064178466796875,
"learning_rate": 1e-05,
"loss": 0.6786,
"step": 10990
},
{
"epoch": 0.2092487968194183,
"grad_norm": 8.505256652832031,
"learning_rate": 1e-05,
"loss": 0.8531,
"step": 11000
},
{
"epoch": 0.20943902299834502,
"grad_norm": 2.3176887035369873,
"learning_rate": 1e-05,
"loss": 0.8953,
"step": 11010
},
{
"epoch": 0.20962924917727177,
"grad_norm": 1.7218331098556519,
"learning_rate": 1e-05,
"loss": 0.9051,
"step": 11020
},
{
"epoch": 0.20981947535619852,
"grad_norm": 2.257793664932251,
"learning_rate": 1e-05,
"loss": 0.6921,
"step": 11030
},
{
"epoch": 0.21000970153512527,
"grad_norm": 3.0849592685699463,
"learning_rate": 1e-05,
"loss": 0.7115,
"step": 11040
},
{
"epoch": 0.21019992771405202,
"grad_norm": 9.030512809753418,
"learning_rate": 1e-05,
"loss": 0.8001,
"step": 11050
},
{
"epoch": 0.21039015389297874,
"grad_norm": 1.8325556516647339,
"learning_rate": 1e-05,
"loss": 0.9223,
"step": 11060
},
{
"epoch": 0.2105803800719055,
"grad_norm": 1.912457823753357,
"learning_rate": 1e-05,
"loss": 0.8395,
"step": 11070
},
{
"epoch": 0.21077060625083224,
"grad_norm": 2.5100717544555664,
"learning_rate": 1e-05,
"loss": 0.7089,
"step": 11080
},
{
"epoch": 0.210960832429759,
"grad_norm": 3.052577495574951,
"learning_rate": 1e-05,
"loss": 0.677,
"step": 11090
},
{
"epoch": 0.21115105860868572,
"grad_norm": 9.417335510253906,
"learning_rate": 1e-05,
"loss": 0.8838,
"step": 11100
},
{
"epoch": 0.21134128478761247,
"grad_norm": 1.9494776725769043,
"learning_rate": 1e-05,
"loss": 0.8791,
"step": 11110
},
{
"epoch": 0.21153151096653922,
"grad_norm": 1.9042080640792847,
"learning_rate": 1e-05,
"loss": 0.8614,
"step": 11120
},
{
"epoch": 0.21172173714546597,
"grad_norm": 2.9908151626586914,
"learning_rate": 1e-05,
"loss": 0.756,
"step": 11130
},
{
"epoch": 0.21191196332439272,
"grad_norm": 3.3369860649108887,
"learning_rate": 1e-05,
"loss": 0.6793,
"step": 11140
},
{
"epoch": 0.21210218950331944,
"grad_norm": 7.909424781799316,
"learning_rate": 1e-05,
"loss": 0.8698,
"step": 11150
},
{
"epoch": 0.2122924156822462,
"grad_norm": 2.1398510932922363,
"learning_rate": 1e-05,
"loss": 0.9081,
"step": 11160
},
{
"epoch": 0.21248264186117294,
"grad_norm": 1.725948691368103,
"learning_rate": 1e-05,
"loss": 0.8344,
"step": 11170
},
{
"epoch": 0.2126728680400997,
"grad_norm": 2.211113691329956,
"learning_rate": 1e-05,
"loss": 0.6968,
"step": 11180
},
{
"epoch": 0.2128630942190264,
"grad_norm": 3.1788854598999023,
"learning_rate": 1e-05,
"loss": 0.698,
"step": 11190
},
{
"epoch": 0.21305332039795316,
"grad_norm": 6.950159549713135,
"learning_rate": 1e-05,
"loss": 0.7759,
"step": 11200
},
{
"epoch": 0.2132435465768799,
"grad_norm": 1.9211266040802002,
"learning_rate": 1e-05,
"loss": 0.894,
"step": 11210
},
{
"epoch": 0.21343377275580666,
"grad_norm": 2.643813133239746,
"learning_rate": 1e-05,
"loss": 0.8672,
"step": 11220
},
{
"epoch": 0.2136239989347334,
"grad_norm": 3.018766164779663,
"learning_rate": 1e-05,
"loss": 0.7093,
"step": 11230
},
{
"epoch": 0.21381422511366013,
"grad_norm": 3.3418798446655273,
"learning_rate": 1e-05,
"loss": 0.6871,
"step": 11240
},
{
"epoch": 0.21400445129258688,
"grad_norm": 7.862785816192627,
"learning_rate": 1e-05,
"loss": 0.8582,
"step": 11250
},
{
"epoch": 0.21419467747151363,
"grad_norm": 1.8714709281921387,
"learning_rate": 1e-05,
"loss": 0.9146,
"step": 11260
},
{
"epoch": 0.21438490365044038,
"grad_norm": 2.029397487640381,
"learning_rate": 1e-05,
"loss": 0.8621,
"step": 11270
},
{
"epoch": 0.2145751298293671,
"grad_norm": 2.297769069671631,
"learning_rate": 1e-05,
"loss": 0.6778,
"step": 11280
},
{
"epoch": 0.21476535600829386,
"grad_norm": 3.518481731414795,
"learning_rate": 1e-05,
"loss": 0.6532,
"step": 11290
},
{
"epoch": 0.2149555821872206,
"grad_norm": 8.1442289352417,
"learning_rate": 1e-05,
"loss": 0.7952,
"step": 11300
},
{
"epoch": 0.21514580836614736,
"grad_norm": 1.6912665367126465,
"learning_rate": 1e-05,
"loss": 0.8844,
"step": 11310
},
{
"epoch": 0.2153360345450741,
"grad_norm": 2.051668643951416,
"learning_rate": 1e-05,
"loss": 0.8225,
"step": 11320
},
{
"epoch": 0.21552626072400083,
"grad_norm": 2.2250289916992188,
"learning_rate": 1e-05,
"loss": 0.6959,
"step": 11330
},
{
"epoch": 0.21571648690292758,
"grad_norm": 3.1313180923461914,
"learning_rate": 1e-05,
"loss": 0.6914,
"step": 11340
},
{
"epoch": 0.21590671308185433,
"grad_norm": 8.869209289550781,
"learning_rate": 1e-05,
"loss": 0.808,
"step": 11350
},
{
"epoch": 0.21609693926078108,
"grad_norm": 1.7078214883804321,
"learning_rate": 1e-05,
"loss": 0.8764,
"step": 11360
},
{
"epoch": 0.2162871654397078,
"grad_norm": 1.7782410383224487,
"learning_rate": 1e-05,
"loss": 0.8458,
"step": 11370
},
{
"epoch": 0.21647739161863455,
"grad_norm": 2.590075731277466,
"learning_rate": 1e-05,
"loss": 0.732,
"step": 11380
},
{
"epoch": 0.2166676177975613,
"grad_norm": 2.7843222618103027,
"learning_rate": 1e-05,
"loss": 0.6477,
"step": 11390
},
{
"epoch": 0.21685784397648805,
"grad_norm": 7.939810752868652,
"learning_rate": 1e-05,
"loss": 0.8014,
"step": 11400
},
{
"epoch": 0.2170480701554148,
"grad_norm": 1.53331458568573,
"learning_rate": 1e-05,
"loss": 0.9162,
"step": 11410
},
{
"epoch": 0.21723829633434152,
"grad_norm": 1.763517141342163,
"learning_rate": 1e-05,
"loss": 0.839,
"step": 11420
},
{
"epoch": 0.21742852251326827,
"grad_norm": 2.2777090072631836,
"learning_rate": 1e-05,
"loss": 0.7011,
"step": 11430
},
{
"epoch": 0.21761874869219502,
"grad_norm": 3.469245433807373,
"learning_rate": 1e-05,
"loss": 0.6751,
"step": 11440
},
{
"epoch": 0.21780897487112177,
"grad_norm": 9.501341819763184,
"learning_rate": 1e-05,
"loss": 0.8277,
"step": 11450
},
{
"epoch": 0.2179992010500485,
"grad_norm": 1.403679609298706,
"learning_rate": 1e-05,
"loss": 0.8802,
"step": 11460
},
{
"epoch": 0.21818942722897525,
"grad_norm": 1.7505896091461182,
"learning_rate": 1e-05,
"loss": 0.8516,
"step": 11470
},
{
"epoch": 0.218379653407902,
"grad_norm": 2.9275999069213867,
"learning_rate": 1e-05,
"loss": 0.7019,
"step": 11480
},
{
"epoch": 0.21856987958682875,
"grad_norm": 3.2237815856933594,
"learning_rate": 1e-05,
"loss": 0.6852,
"step": 11490
},
{
"epoch": 0.2187601057657555,
"grad_norm": 8.424261093139648,
"learning_rate": 1e-05,
"loss": 0.863,
"step": 11500
},
{
"epoch": 0.21895033194468222,
"grad_norm": 1.6581450700759888,
"learning_rate": 1e-05,
"loss": 0.9228,
"step": 11510
},
{
"epoch": 0.21914055812360897,
"grad_norm": 2.0493242740631104,
"learning_rate": 1e-05,
"loss": 0.8508,
"step": 11520
},
{
"epoch": 0.21933078430253572,
"grad_norm": 2.502948045730591,
"learning_rate": 1e-05,
"loss": 0.6656,
"step": 11530
},
{
"epoch": 0.21952101048146247,
"grad_norm": 3.2436251640319824,
"learning_rate": 1e-05,
"loss": 0.6602,
"step": 11540
},
{
"epoch": 0.2197112366603892,
"grad_norm": 8.516260147094727,
"learning_rate": 1e-05,
"loss": 0.8216,
"step": 11550
},
{
"epoch": 0.21990146283931594,
"grad_norm": 1.7573206424713135,
"learning_rate": 1e-05,
"loss": 0.9076,
"step": 11560
},
{
"epoch": 0.2200916890182427,
"grad_norm": 1.9056260585784912,
"learning_rate": 1e-05,
"loss": 0.869,
"step": 11570
},
{
"epoch": 0.22028191519716944,
"grad_norm": 2.78605055809021,
"learning_rate": 1e-05,
"loss": 0.697,
"step": 11580
},
{
"epoch": 0.22047214137609616,
"grad_norm": 3.34475040435791,
"learning_rate": 1e-05,
"loss": 0.6799,
"step": 11590
},
{
"epoch": 0.22066236755502291,
"grad_norm": 8.150701522827148,
"learning_rate": 1e-05,
"loss": 0.8361,
"step": 11600
},
{
"epoch": 0.22085259373394966,
"grad_norm": 1.6616748571395874,
"learning_rate": 1e-05,
"loss": 0.9,
"step": 11610
},
{
"epoch": 0.22104281991287641,
"grad_norm": 1.9119235277175903,
"learning_rate": 1e-05,
"loss": 0.8355,
"step": 11620
},
{
"epoch": 0.22123304609180316,
"grad_norm": 2.6226744651794434,
"learning_rate": 1e-05,
"loss": 0.6801,
"step": 11630
},
{
"epoch": 0.2214232722707299,
"grad_norm": 3.234544515609741,
"learning_rate": 1e-05,
"loss": 0.6524,
"step": 11640
},
{
"epoch": 0.22161349844965664,
"grad_norm": 8.400768280029297,
"learning_rate": 1e-05,
"loss": 0.8381,
"step": 11650
},
{
"epoch": 0.2218037246285834,
"grad_norm": 1.7022336721420288,
"learning_rate": 1e-05,
"loss": 0.9327,
"step": 11660
},
{
"epoch": 0.22199395080751014,
"grad_norm": 2.2398221492767334,
"learning_rate": 1e-05,
"loss": 0.8161,
"step": 11670
},
{
"epoch": 0.22218417698643686,
"grad_norm": 2.4238078594207764,
"learning_rate": 1e-05,
"loss": 0.6774,
"step": 11680
},
{
"epoch": 0.2223744031653636,
"grad_norm": 3.259014129638672,
"learning_rate": 1e-05,
"loss": 0.6819,
"step": 11690
},
{
"epoch": 0.22256462934429036,
"grad_norm": 8.916090965270996,
"learning_rate": 1e-05,
"loss": 0.8275,
"step": 11700
},
{
"epoch": 0.2227548555232171,
"grad_norm": 1.6667135953903198,
"learning_rate": 1e-05,
"loss": 0.8989,
"step": 11710
},
{
"epoch": 0.22294508170214386,
"grad_norm": 1.8889349699020386,
"learning_rate": 1e-05,
"loss": 0.8343,
"step": 11720
},
{
"epoch": 0.22313530788107058,
"grad_norm": 2.473635673522949,
"learning_rate": 1e-05,
"loss": 0.6453,
"step": 11730
},
{
"epoch": 0.22332553405999733,
"grad_norm": 3.0168142318725586,
"learning_rate": 1e-05,
"loss": 0.6613,
"step": 11740
},
{
"epoch": 0.22351576023892408,
"grad_norm": 7.629496097564697,
"learning_rate": 1e-05,
"loss": 0.8272,
"step": 11750
},
{
"epoch": 0.22370598641785083,
"grad_norm": 1.5201396942138672,
"learning_rate": 1e-05,
"loss": 0.8966,
"step": 11760
},
{
"epoch": 0.22389621259677756,
"grad_norm": 1.6035292148590088,
"learning_rate": 1e-05,
"loss": 0.7771,
"step": 11770
},
{
"epoch": 0.2240864387757043,
"grad_norm": 2.3732476234436035,
"learning_rate": 1e-05,
"loss": 0.6666,
"step": 11780
},
{
"epoch": 0.22427666495463106,
"grad_norm": 3.351346731185913,
"learning_rate": 1e-05,
"loss": 0.6609,
"step": 11790
},
{
"epoch": 0.2244668911335578,
"grad_norm": 8.000328063964844,
"learning_rate": 1e-05,
"loss": 0.8348,
"step": 11800
},
{
"epoch": 0.22465711731248456,
"grad_norm": 2.0371391773223877,
"learning_rate": 1e-05,
"loss": 0.9081,
"step": 11810
},
{
"epoch": 0.22484734349141128,
"grad_norm": 1.8373336791992188,
"learning_rate": 1e-05,
"loss": 0.8421,
"step": 11820
},
{
"epoch": 0.22503756967033803,
"grad_norm": 2.426450490951538,
"learning_rate": 1e-05,
"loss": 0.6459,
"step": 11830
},
{
"epoch": 0.22522779584926478,
"grad_norm": 3.183286428451538,
"learning_rate": 1e-05,
"loss": 0.6885,
"step": 11840
},
{
"epoch": 0.22541802202819153,
"grad_norm": 7.636185169219971,
"learning_rate": 1e-05,
"loss": 0.8584,
"step": 11850
},
{
"epoch": 0.22560824820711825,
"grad_norm": 1.6561346054077148,
"learning_rate": 1e-05,
"loss": 0.895,
"step": 11860
},
{
"epoch": 0.225798474386045,
"grad_norm": 1.8181359767913818,
"learning_rate": 1e-05,
"loss": 0.8327,
"step": 11870
},
{
"epoch": 0.22598870056497175,
"grad_norm": 2.2778751850128174,
"learning_rate": 1e-05,
"loss": 0.6695,
"step": 11880
},
{
"epoch": 0.2261789267438985,
"grad_norm": 3.0406625270843506,
"learning_rate": 1e-05,
"loss": 0.6853,
"step": 11890
},
{
"epoch": 0.22636915292282525,
"grad_norm": 6.587218284606934,
"learning_rate": 1e-05,
"loss": 0.7878,
"step": 11900
},
{
"epoch": 0.22655937910175197,
"grad_norm": 1.6370172500610352,
"learning_rate": 1e-05,
"loss": 0.9343,
"step": 11910
},
{
"epoch": 0.22674960528067872,
"grad_norm": 2.0910651683807373,
"learning_rate": 1e-05,
"loss": 0.8663,
"step": 11920
},
{
"epoch": 0.22693983145960547,
"grad_norm": 2.447936534881592,
"learning_rate": 1e-05,
"loss": 0.7038,
"step": 11930
},
{
"epoch": 0.22713005763853222,
"grad_norm": 3.1258623600006104,
"learning_rate": 1e-05,
"loss": 0.6851,
"step": 11940
},
{
"epoch": 0.22732028381745895,
"grad_norm": 9.921146392822266,
"learning_rate": 1e-05,
"loss": 0.8211,
"step": 11950
},
{
"epoch": 0.2275105099963857,
"grad_norm": 1.6214704513549805,
"learning_rate": 1e-05,
"loss": 0.9084,
"step": 11960
},
{
"epoch": 0.22770073617531245,
"grad_norm": 2.229059934616089,
"learning_rate": 1e-05,
"loss": 0.833,
"step": 11970
},
{
"epoch": 0.2278909623542392,
"grad_norm": 2.1819334030151367,
"learning_rate": 1e-05,
"loss": 0.7154,
"step": 11980
},
{
"epoch": 0.22808118853316595,
"grad_norm": 2.60158371925354,
"learning_rate": 1e-05,
"loss": 0.6689,
"step": 11990
},
{
"epoch": 0.22827141471209267,
"grad_norm": 7.86823844909668,
"learning_rate": 1e-05,
"loss": 0.8265,
"step": 12000
},
{
"epoch": 0.22846164089101942,
"grad_norm": 2.524681568145752,
"learning_rate": 1e-05,
"loss": 0.903,
"step": 12010
},
{
"epoch": 0.22865186706994617,
"grad_norm": 1.828129529953003,
"learning_rate": 1e-05,
"loss": 0.8307,
"step": 12020
},
{
"epoch": 0.22884209324887292,
"grad_norm": 2.0472328662872314,
"learning_rate": 1e-05,
"loss": 0.674,
"step": 12030
},
{
"epoch": 0.22903231942779964,
"grad_norm": 2.722903251647949,
"learning_rate": 1e-05,
"loss": 0.6541,
"step": 12040
},
{
"epoch": 0.2292225456067264,
"grad_norm": 9.293694496154785,
"learning_rate": 1e-05,
"loss": 0.7853,
"step": 12050
},
{
"epoch": 0.22941277178565314,
"grad_norm": 1.4744130373001099,
"learning_rate": 1e-05,
"loss": 0.8812,
"step": 12060
},
{
"epoch": 0.2296029979645799,
"grad_norm": 2.0159759521484375,
"learning_rate": 1e-05,
"loss": 0.8134,
"step": 12070
},
{
"epoch": 0.22979322414350664,
"grad_norm": 2.542578935623169,
"learning_rate": 1e-05,
"loss": 0.7185,
"step": 12080
},
{
"epoch": 0.22998345032243336,
"grad_norm": 3.231062173843384,
"learning_rate": 1e-05,
"loss": 0.7022,
"step": 12090
},
{
"epoch": 0.2301736765013601,
"grad_norm": 8.560314178466797,
"learning_rate": 1e-05,
"loss": 0.8614,
"step": 12100
},
{
"epoch": 0.23036390268028686,
"grad_norm": 1.8245794773101807,
"learning_rate": 1e-05,
"loss": 0.9044,
"step": 12110
},
{
"epoch": 0.23055412885921361,
"grad_norm": 2.1264288425445557,
"learning_rate": 1e-05,
"loss": 0.867,
"step": 12120
},
{
"epoch": 0.23074435503814034,
"grad_norm": 2.0867459774017334,
"learning_rate": 1e-05,
"loss": 0.6969,
"step": 12130
},
{
"epoch": 0.2309345812170671,
"grad_norm": 3.31634783744812,
"learning_rate": 1e-05,
"loss": 0.6555,
"step": 12140
},
{
"epoch": 0.23112480739599384,
"grad_norm": 8.601350784301758,
"learning_rate": 1e-05,
"loss": 0.8114,
"step": 12150
},
{
"epoch": 0.2313150335749206,
"grad_norm": 2.365330934524536,
"learning_rate": 1e-05,
"loss": 0.9047,
"step": 12160
},
{
"epoch": 0.23150525975384734,
"grad_norm": 1.6276649236679077,
"learning_rate": 1e-05,
"loss": 0.8254,
"step": 12170
},
{
"epoch": 0.23169548593277406,
"grad_norm": 2.794624090194702,
"learning_rate": 1e-05,
"loss": 0.6747,
"step": 12180
},
{
"epoch": 0.2318857121117008,
"grad_norm": 2.976040840148926,
"learning_rate": 1e-05,
"loss": 0.654,
"step": 12190
},
{
"epoch": 0.23207593829062756,
"grad_norm": 6.933033466339111,
"learning_rate": 1e-05,
"loss": 0.8057,
"step": 12200
},
{
"epoch": 0.2322661644695543,
"grad_norm": 1.7779545783996582,
"learning_rate": 1e-05,
"loss": 0.9081,
"step": 12210
},
{
"epoch": 0.23245639064848103,
"grad_norm": 1.9526350498199463,
"learning_rate": 1e-05,
"loss": 0.857,
"step": 12220
},
{
"epoch": 0.23264661682740778,
"grad_norm": 2.0491960048675537,
"learning_rate": 1e-05,
"loss": 0.6714,
"step": 12230
},
{
"epoch": 0.23283684300633453,
"grad_norm": 3.1386642456054688,
"learning_rate": 1e-05,
"loss": 0.6317,
"step": 12240
},
{
"epoch": 0.23302706918526128,
"grad_norm": 8.413360595703125,
"learning_rate": 1e-05,
"loss": 0.7791,
"step": 12250
},
{
"epoch": 0.23321729536418803,
"grad_norm": 1.8893053531646729,
"learning_rate": 1e-05,
"loss": 0.8869,
"step": 12260
},
{
"epoch": 0.23340752154311475,
"grad_norm": 1.7962968349456787,
"learning_rate": 1e-05,
"loss": 0.823,
"step": 12270
},
{
"epoch": 0.2335977477220415,
"grad_norm": 2.5221433639526367,
"learning_rate": 1e-05,
"loss": 0.7192,
"step": 12280
},
{
"epoch": 0.23378797390096825,
"grad_norm": 2.8311657905578613,
"learning_rate": 1e-05,
"loss": 0.7063,
"step": 12290
},
{
"epoch": 0.233978200079895,
"grad_norm": 7.7757649421691895,
"learning_rate": 1e-05,
"loss": 0.7973,
"step": 12300
},
{
"epoch": 0.23416842625882173,
"grad_norm": 1.7501600980758667,
"learning_rate": 1e-05,
"loss": 0.8655,
"step": 12310
},
{
"epoch": 0.23435865243774848,
"grad_norm": 2.095082998275757,
"learning_rate": 1e-05,
"loss": 0.8283,
"step": 12320
},
{
"epoch": 0.23454887861667523,
"grad_norm": 2.532461166381836,
"learning_rate": 1e-05,
"loss": 0.6826,
"step": 12330
},
{
"epoch": 0.23473910479560198,
"grad_norm": 2.7468090057373047,
"learning_rate": 1e-05,
"loss": 0.6384,
"step": 12340
},
{
"epoch": 0.23492933097452873,
"grad_norm": 7.482028007507324,
"learning_rate": 1e-05,
"loss": 0.7903,
"step": 12350
},
{
"epoch": 0.23511955715345545,
"grad_norm": 1.808935284614563,
"learning_rate": 1e-05,
"loss": 0.8941,
"step": 12360
},
{
"epoch": 0.2353097833323822,
"grad_norm": 1.7405446767807007,
"learning_rate": 1e-05,
"loss": 0.8,
"step": 12370
},
{
"epoch": 0.23550000951130895,
"grad_norm": 2.3710715770721436,
"learning_rate": 1e-05,
"loss": 0.6524,
"step": 12380
},
{
"epoch": 0.2356902356902357,
"grad_norm": 3.4030723571777344,
"learning_rate": 1e-05,
"loss": 0.6481,
"step": 12390
},
{
"epoch": 0.23588046186916242,
"grad_norm": 8.050148963928223,
"learning_rate": 1e-05,
"loss": 0.7871,
"step": 12400
},
{
"epoch": 0.23607068804808917,
"grad_norm": 1.569789171218872,
"learning_rate": 1e-05,
"loss": 0.8716,
"step": 12410
},
{
"epoch": 0.23626091422701592,
"grad_norm": 1.7262279987335205,
"learning_rate": 1e-05,
"loss": 0.8299,
"step": 12420
},
{
"epoch": 0.23645114040594267,
"grad_norm": 2.2987582683563232,
"learning_rate": 1e-05,
"loss": 0.65,
"step": 12430
},
{
"epoch": 0.23664136658486942,
"grad_norm": 3.7272424697875977,
"learning_rate": 1e-05,
"loss": 0.6652,
"step": 12440
},
{
"epoch": 0.23683159276379614,
"grad_norm": 7.83486270904541,
"learning_rate": 1e-05,
"loss": 0.8159,
"step": 12450
},
{
"epoch": 0.2370218189427229,
"grad_norm": 2.0631587505340576,
"learning_rate": 1e-05,
"loss": 0.8735,
"step": 12460
},
{
"epoch": 0.23721204512164965,
"grad_norm": 2.02669358253479,
"learning_rate": 1e-05,
"loss": 0.8,
"step": 12470
},
{
"epoch": 0.2374022713005764,
"grad_norm": 2.4576754570007324,
"learning_rate": 1e-05,
"loss": 0.6536,
"step": 12480
},
{
"epoch": 0.23759249747950312,
"grad_norm": 3.5122756958007812,
"learning_rate": 1e-05,
"loss": 0.671,
"step": 12490
},
{
"epoch": 0.23778272365842987,
"grad_norm": 9.034518241882324,
"learning_rate": 1e-05,
"loss": 0.8778,
"step": 12500
},
{
"epoch": 0.23797294983735662,
"grad_norm": 2.5387027263641357,
"learning_rate": 1e-05,
"loss": 0.9088,
"step": 12510
},
{
"epoch": 0.23816317601628337,
"grad_norm": 1.7956695556640625,
"learning_rate": 1e-05,
"loss": 0.7954,
"step": 12520
},
{
"epoch": 0.23835340219521012,
"grad_norm": 2.3624777793884277,
"learning_rate": 1e-05,
"loss": 0.6658,
"step": 12530
},
{
"epoch": 0.23854362837413684,
"grad_norm": 2.8864974975585938,
"learning_rate": 1e-05,
"loss": 0.6733,
"step": 12540
},
{
"epoch": 0.2387338545530636,
"grad_norm": 8.089028358459473,
"learning_rate": 1e-05,
"loss": 0.8344,
"step": 12550
},
{
"epoch": 0.23892408073199034,
"grad_norm": 1.7586256265640259,
"learning_rate": 1e-05,
"loss": 0.8788,
"step": 12560
},
{
"epoch": 0.2391143069109171,
"grad_norm": 1.607397198677063,
"learning_rate": 1e-05,
"loss": 0.8078,
"step": 12570
},
{
"epoch": 0.2393045330898438,
"grad_norm": 2.2670743465423584,
"learning_rate": 1e-05,
"loss": 0.7031,
"step": 12580
},
{
"epoch": 0.23949475926877056,
"grad_norm": 3.146979570388794,
"learning_rate": 1e-05,
"loss": 0.6681,
"step": 12590
},
{
"epoch": 0.2396849854476973,
"grad_norm": 7.837850570678711,
"learning_rate": 1e-05,
"loss": 0.8245,
"step": 12600
},
{
"epoch": 0.23987521162662406,
"grad_norm": 2.0519442558288574,
"learning_rate": 1e-05,
"loss": 0.8655,
"step": 12610
},
{
"epoch": 0.2400654378055508,
"grad_norm": 1.9732800722122192,
"learning_rate": 1e-05,
"loss": 0.809,
"step": 12620
},
{
"epoch": 0.24025566398447754,
"grad_norm": 2.4767050743103027,
"learning_rate": 1e-05,
"loss": 0.6899,
"step": 12630
},
{
"epoch": 0.24044589016340429,
"grad_norm": 3.0085601806640625,
"learning_rate": 1e-05,
"loss": 0.6648,
"step": 12640
},
{
"epoch": 0.24063611634233104,
"grad_norm": 7.591806411743164,
"learning_rate": 1e-05,
"loss": 0.8023,
"step": 12650
},
{
"epoch": 0.24082634252125779,
"grad_norm": 1.7809135913848877,
"learning_rate": 1e-05,
"loss": 0.8769,
"step": 12660
},
{
"epoch": 0.2410165687001845,
"grad_norm": 2.0374767780303955,
"learning_rate": 1e-05,
"loss": 0.8862,
"step": 12670
},
{
"epoch": 0.24120679487911126,
"grad_norm": 2.2392523288726807,
"learning_rate": 1e-05,
"loss": 0.6338,
"step": 12680
},
{
"epoch": 0.241397021058038,
"grad_norm": 3.12538480758667,
"learning_rate": 1e-05,
"loss": 0.6466,
"step": 12690
},
{
"epoch": 0.24158724723696476,
"grad_norm": 7.700732707977295,
"learning_rate": 1e-05,
"loss": 0.823,
"step": 12700
},
{
"epoch": 0.2417774734158915,
"grad_norm": 1.6536741256713867,
"learning_rate": 1e-05,
"loss": 0.8899,
"step": 12710
},
{
"epoch": 0.24196769959481823,
"grad_norm": 1.803846001625061,
"learning_rate": 1e-05,
"loss": 0.8104,
"step": 12720
},
{
"epoch": 0.24215792577374498,
"grad_norm": 2.369276523590088,
"learning_rate": 1e-05,
"loss": 0.682,
"step": 12730
},
{
"epoch": 0.24234815195267173,
"grad_norm": 3.3642470836639404,
"learning_rate": 1e-05,
"loss": 0.6414,
"step": 12740
},
{
"epoch": 0.24253837813159848,
"grad_norm": 9.56998062133789,
"learning_rate": 1e-05,
"loss": 0.8104,
"step": 12750
},
{
"epoch": 0.2427286043105252,
"grad_norm": 1.84210205078125,
"learning_rate": 1e-05,
"loss": 0.9058,
"step": 12760
},
{
"epoch": 0.24291883048945195,
"grad_norm": 1.4530168771743774,
"learning_rate": 1e-05,
"loss": 0.8024,
"step": 12770
},
{
"epoch": 0.2431090566683787,
"grad_norm": 2.2654664516448975,
"learning_rate": 1e-05,
"loss": 0.7175,
"step": 12780
},
{
"epoch": 0.24329928284730545,
"grad_norm": 2.610100030899048,
"learning_rate": 1e-05,
"loss": 0.6259,
"step": 12790
},
{
"epoch": 0.2434895090262322,
"grad_norm": 8.209732055664062,
"learning_rate": 1e-05,
"loss": 0.7929,
"step": 12800
},
{
"epoch": 0.24367973520515893,
"grad_norm": 1.6762723922729492,
"learning_rate": 1e-05,
"loss": 0.9031,
"step": 12810
},
{
"epoch": 0.24386996138408568,
"grad_norm": 2.2293217182159424,
"learning_rate": 1e-05,
"loss": 0.8437,
"step": 12820
},
{
"epoch": 0.24406018756301243,
"grad_norm": 2.734182357788086,
"learning_rate": 1e-05,
"loss": 0.6601,
"step": 12830
},
{
"epoch": 0.24425041374193918,
"grad_norm": 3.0669283866882324,
"learning_rate": 1e-05,
"loss": 0.6398,
"step": 12840
},
{
"epoch": 0.2444406399208659,
"grad_norm": 9.472503662109375,
"learning_rate": 1e-05,
"loss": 0.8245,
"step": 12850
},
{
"epoch": 0.24463086609979265,
"grad_norm": 1.7405332326889038,
"learning_rate": 1e-05,
"loss": 0.8789,
"step": 12860
},
{
"epoch": 0.2448210922787194,
"grad_norm": 2.245732545852661,
"learning_rate": 1e-05,
"loss": 0.8795,
"step": 12870
},
{
"epoch": 0.24501131845764615,
"grad_norm": 1.9449738264083862,
"learning_rate": 1e-05,
"loss": 0.6514,
"step": 12880
},
{
"epoch": 0.2452015446365729,
"grad_norm": 3.132315158843994,
"learning_rate": 1e-05,
"loss": 0.7077,
"step": 12890
},
{
"epoch": 0.24539177081549962,
"grad_norm": 8.021961212158203,
"learning_rate": 1e-05,
"loss": 0.8287,
"step": 12900
},
{
"epoch": 0.24558199699442637,
"grad_norm": 1.3967387676239014,
"learning_rate": 1e-05,
"loss": 0.8634,
"step": 12910
},
{
"epoch": 0.24577222317335312,
"grad_norm": 1.8983734846115112,
"learning_rate": 1e-05,
"loss": 0.8078,
"step": 12920
},
{
"epoch": 0.24596244935227987,
"grad_norm": 2.4123551845550537,
"learning_rate": 1e-05,
"loss": 0.6415,
"step": 12930
},
{
"epoch": 0.2461526755312066,
"grad_norm": 3.2492337226867676,
"learning_rate": 1e-05,
"loss": 0.6929,
"step": 12940
},
{
"epoch": 0.24634290171013334,
"grad_norm": 7.201467514038086,
"learning_rate": 1e-05,
"loss": 0.7836,
"step": 12950
},
{
"epoch": 0.2465331278890601,
"grad_norm": 1.4045886993408203,
"learning_rate": 1e-05,
"loss": 0.8673,
"step": 12960
},
{
"epoch": 0.24672335406798684,
"grad_norm": 1.7008304595947266,
"learning_rate": 1e-05,
"loss": 0.7649,
"step": 12970
},
{
"epoch": 0.24691358024691357,
"grad_norm": 2.7933273315429688,
"learning_rate": 1e-05,
"loss": 0.6998,
"step": 12980
},
{
"epoch": 0.24710380642584032,
"grad_norm": 4.158901691436768,
"learning_rate": 1e-05,
"loss": 0.6763,
"step": 12990
},
{
"epoch": 0.24729403260476707,
"grad_norm": 9.045417785644531,
"learning_rate": 1e-05,
"loss": 0.848,
"step": 13000
},
{
"epoch": 0.24748425878369382,
"grad_norm": 1.6765538454055786,
"learning_rate": 1e-05,
"loss": 0.889,
"step": 13010
},
{
"epoch": 0.24767448496262057,
"grad_norm": 2.151571273803711,
"learning_rate": 1e-05,
"loss": 0.8139,
"step": 13020
},
{
"epoch": 0.2478647111415473,
"grad_norm": 2.350522994995117,
"learning_rate": 1e-05,
"loss": 0.6716,
"step": 13030
},
{
"epoch": 0.24805493732047404,
"grad_norm": 3.107022762298584,
"learning_rate": 1e-05,
"loss": 0.6345,
"step": 13040
},
{
"epoch": 0.2482451634994008,
"grad_norm": 8.354117393493652,
"learning_rate": 1e-05,
"loss": 0.8032,
"step": 13050
},
{
"epoch": 0.24843538967832754,
"grad_norm": 1.5396760702133179,
"learning_rate": 1e-05,
"loss": 0.8851,
"step": 13060
},
{
"epoch": 0.24862561585725426,
"grad_norm": 1.6520079374313354,
"learning_rate": 1e-05,
"loss": 0.821,
"step": 13070
},
{
"epoch": 0.248815842036181,
"grad_norm": 2.0819594860076904,
"learning_rate": 1e-05,
"loss": 0.6544,
"step": 13080
},
{
"epoch": 0.24900606821510776,
"grad_norm": 3.4308531284332275,
"learning_rate": 1e-05,
"loss": 0.6439,
"step": 13090
},
{
"epoch": 0.2491962943940345,
"grad_norm": 7.857069492340088,
"learning_rate": 1e-05,
"loss": 0.7988,
"step": 13100
},
{
"epoch": 0.24938652057296126,
"grad_norm": 1.571454644203186,
"learning_rate": 1e-05,
"loss": 0.8618,
"step": 13110
},
{
"epoch": 0.24957674675188798,
"grad_norm": 2.005012035369873,
"learning_rate": 1e-05,
"loss": 0.822,
"step": 13120
},
{
"epoch": 0.24976697293081473,
"grad_norm": 2.6577112674713135,
"learning_rate": 1e-05,
"loss": 0.7083,
"step": 13130
},
{
"epoch": 0.24995719910974148,
"grad_norm": 4.0507988929748535,
"learning_rate": 1e-05,
"loss": 0.6757,
"step": 13140
},
{
"epoch": 0.25014742528866823,
"grad_norm": 6.790675640106201,
"learning_rate": 1e-05,
"loss": 0.7965,
"step": 13150
},
{
"epoch": 0.25033765146759496,
"grad_norm": 1.5484857559204102,
"learning_rate": 1e-05,
"loss": 0.8649,
"step": 13160
},
{
"epoch": 0.25052787764652173,
"grad_norm": 1.6114444732666016,
"learning_rate": 1e-05,
"loss": 0.8328,
"step": 13170
},
{
"epoch": 0.25071810382544846,
"grad_norm": 2.0124263763427734,
"learning_rate": 1e-05,
"loss": 0.6564,
"step": 13180
},
{
"epoch": 0.2509083300043752,
"grad_norm": 3.197026014328003,
"learning_rate": 1e-05,
"loss": 0.6616,
"step": 13190
},
{
"epoch": 0.25109855618330196,
"grad_norm": 8.478826522827148,
"learning_rate": 1e-05,
"loss": 0.8385,
"step": 13200
},
{
"epoch": 0.2512887823622287,
"grad_norm": 1.4479085206985474,
"learning_rate": 1e-05,
"loss": 0.8998,
"step": 13210
},
{
"epoch": 0.25147900854115546,
"grad_norm": 1.858893871307373,
"learning_rate": 1e-05,
"loss": 0.8882,
"step": 13220
},
{
"epoch": 0.2516692347200822,
"grad_norm": 2.4273767471313477,
"learning_rate": 1e-05,
"loss": 0.7108,
"step": 13230
},
{
"epoch": 0.2518594608990089,
"grad_norm": 2.7225241661071777,
"learning_rate": 1e-05,
"loss": 0.6573,
"step": 13240
},
{
"epoch": 0.2520496870779357,
"grad_norm": 8.1727876663208,
"learning_rate": 1e-05,
"loss": 0.8141,
"step": 13250
},
{
"epoch": 0.2522399132568624,
"grad_norm": 1.6315418481826782,
"learning_rate": 1e-05,
"loss": 0.8768,
"step": 13260
},
{
"epoch": 0.2524301394357892,
"grad_norm": 1.9085664749145508,
"learning_rate": 1e-05,
"loss": 0.8455,
"step": 13270
},
{
"epoch": 0.2526203656147159,
"grad_norm": 2.238964796066284,
"learning_rate": 1e-05,
"loss": 0.6881,
"step": 13280
},
{
"epoch": 0.2528105917936426,
"grad_norm": 2.955059289932251,
"learning_rate": 1e-05,
"loss": 0.6555,
"step": 13290
},
{
"epoch": 0.2530008179725694,
"grad_norm": 8.048650741577148,
"learning_rate": 1e-05,
"loss": 0.7763,
"step": 13300
},
{
"epoch": 0.2531910441514961,
"grad_norm": 1.9061994552612305,
"learning_rate": 1e-05,
"loss": 0.9032,
"step": 13310
},
{
"epoch": 0.25338127033042285,
"grad_norm": 1.9454113245010376,
"learning_rate": 1e-05,
"loss": 0.7711,
"step": 13320
},
{
"epoch": 0.2535714965093496,
"grad_norm": 2.8815934658050537,
"learning_rate": 1e-05,
"loss": 0.6653,
"step": 13330
},
{
"epoch": 0.25376172268827635,
"grad_norm": 3.361938714981079,
"learning_rate": 1e-05,
"loss": 0.6566,
"step": 13340
},
{
"epoch": 0.2539519488672031,
"grad_norm": 8.011330604553223,
"learning_rate": 1e-05,
"loss": 0.7957,
"step": 13350
},
{
"epoch": 0.25414217504612985,
"grad_norm": 1.9102609157562256,
"learning_rate": 1e-05,
"loss": 0.8955,
"step": 13360
},
{
"epoch": 0.25433240122505657,
"grad_norm": 1.944886326789856,
"learning_rate": 1e-05,
"loss": 0.8874,
"step": 13370
},
{
"epoch": 0.25452262740398335,
"grad_norm": 2.2420082092285156,
"learning_rate": 1e-05,
"loss": 0.6522,
"step": 13380
},
{
"epoch": 0.25471285358291007,
"grad_norm": 3.000147581100464,
"learning_rate": 1e-05,
"loss": 0.6657,
"step": 13390
},
{
"epoch": 0.25490307976183685,
"grad_norm": 7.219781875610352,
"learning_rate": 1e-05,
"loss": 0.7894,
"step": 13400
},
{
"epoch": 0.25509330594076357,
"grad_norm": 1.84577476978302,
"learning_rate": 1e-05,
"loss": 0.8971,
"step": 13410
},
{
"epoch": 0.2552835321196903,
"grad_norm": 1.638624906539917,
"learning_rate": 1e-05,
"loss": 0.7917,
"step": 13420
},
{
"epoch": 0.25547375829861707,
"grad_norm": 2.2567522525787354,
"learning_rate": 1e-05,
"loss": 0.6842,
"step": 13430
},
{
"epoch": 0.2556639844775438,
"grad_norm": 3.0142083168029785,
"learning_rate": 1e-05,
"loss": 0.6391,
"step": 13440
},
{
"epoch": 0.25585421065647057,
"grad_norm": 7.75400447845459,
"learning_rate": 1e-05,
"loss": 0.7372,
"step": 13450
},
{
"epoch": 0.2560444368353973,
"grad_norm": 1.9061027765274048,
"learning_rate": 1e-05,
"loss": 0.8809,
"step": 13460
},
{
"epoch": 0.256234663014324,
"grad_norm": 1.7697067260742188,
"learning_rate": 1e-05,
"loss": 0.8833,
"step": 13470
},
{
"epoch": 0.2564248891932508,
"grad_norm": 2.4031708240509033,
"learning_rate": 1e-05,
"loss": 0.6646,
"step": 13480
},
{
"epoch": 0.2566151153721775,
"grad_norm": 2.779425859451294,
"learning_rate": 1e-05,
"loss": 0.648,
"step": 13490
},
{
"epoch": 0.25680534155110424,
"grad_norm": 8.156201362609863,
"learning_rate": 1e-05,
"loss": 0.8291,
"step": 13500
},
{
"epoch": 0.256995567730031,
"grad_norm": 1.522715449333191,
"learning_rate": 1e-05,
"loss": 0.8679,
"step": 13510
},
{
"epoch": 0.25718579390895774,
"grad_norm": 1.7634680271148682,
"learning_rate": 1e-05,
"loss": 0.8535,
"step": 13520
},
{
"epoch": 0.2573760200878845,
"grad_norm": 2.1683990955352783,
"learning_rate": 1e-05,
"loss": 0.6921,
"step": 13530
},
{
"epoch": 0.25756624626681124,
"grad_norm": 3.1832690238952637,
"learning_rate": 1e-05,
"loss": 0.6686,
"step": 13540
},
{
"epoch": 0.25775647244573796,
"grad_norm": 7.043651103973389,
"learning_rate": 1e-05,
"loss": 0.7945,
"step": 13550
},
{
"epoch": 0.25794669862466474,
"grad_norm": 1.6289427280426025,
"learning_rate": 1e-05,
"loss": 0.8858,
"step": 13560
},
{
"epoch": 0.25813692480359146,
"grad_norm": 1.9320590496063232,
"learning_rate": 1e-05,
"loss": 0.7927,
"step": 13570
},
{
"epoch": 0.25832715098251824,
"grad_norm": 2.4451706409454346,
"learning_rate": 1e-05,
"loss": 0.6729,
"step": 13580
},
{
"epoch": 0.25851737716144496,
"grad_norm": 3.379577398300171,
"learning_rate": 1e-05,
"loss": 0.6556,
"step": 13590
},
{
"epoch": 0.2587076033403717,
"grad_norm": 7.359285354614258,
"learning_rate": 1e-05,
"loss": 0.7734,
"step": 13600
},
{
"epoch": 0.25889782951929846,
"grad_norm": 1.994218349456787,
"learning_rate": 1e-05,
"loss": 0.8606,
"step": 13610
},
{
"epoch": 0.2590880556982252,
"grad_norm": 1.7727612257003784,
"learning_rate": 1e-05,
"loss": 0.7995,
"step": 13620
},
{
"epoch": 0.2592782818771519,
"grad_norm": 2.2640380859375,
"learning_rate": 1e-05,
"loss": 0.6815,
"step": 13630
},
{
"epoch": 0.2594685080560787,
"grad_norm": 3.029695749282837,
"learning_rate": 1e-05,
"loss": 0.6527,
"step": 13640
},
{
"epoch": 0.2596587342350054,
"grad_norm": 7.429525375366211,
"learning_rate": 1e-05,
"loss": 0.7905,
"step": 13650
},
{
"epoch": 0.2598489604139322,
"grad_norm": 1.6208535432815552,
"learning_rate": 1e-05,
"loss": 0.9017,
"step": 13660
},
{
"epoch": 0.2600391865928589,
"grad_norm": 1.614573359489441,
"learning_rate": 1e-05,
"loss": 0.8432,
"step": 13670
},
{
"epoch": 0.26022941277178563,
"grad_norm": 2.1695950031280518,
"learning_rate": 1e-05,
"loss": 0.7121,
"step": 13680
},
{
"epoch": 0.2604196389507124,
"grad_norm": 2.8136610984802246,
"learning_rate": 1e-05,
"loss": 0.6594,
"step": 13690
},
{
"epoch": 0.26060986512963913,
"grad_norm": 6.667365074157715,
"learning_rate": 1e-05,
"loss": 0.771,
"step": 13700
},
{
"epoch": 0.2608000913085659,
"grad_norm": 1.7707114219665527,
"learning_rate": 1e-05,
"loss": 0.9198,
"step": 13710
},
{
"epoch": 0.26099031748749263,
"grad_norm": 1.7706857919692993,
"learning_rate": 1e-05,
"loss": 0.8036,
"step": 13720
},
{
"epoch": 0.26118054366641935,
"grad_norm": 2.390261650085449,
"learning_rate": 1e-05,
"loss": 0.6905,
"step": 13730
},
{
"epoch": 0.26137076984534613,
"grad_norm": 3.2801735401153564,
"learning_rate": 1e-05,
"loss": 0.6849,
"step": 13740
},
{
"epoch": 0.26156099602427285,
"grad_norm": 6.89001989364624,
"learning_rate": 1e-05,
"loss": 0.801,
"step": 13750
},
{
"epoch": 0.26175122220319963,
"grad_norm": 1.4974653720855713,
"learning_rate": 1e-05,
"loss": 0.8438,
"step": 13760
},
{
"epoch": 0.26194144838212635,
"grad_norm": 2.0863194465637207,
"learning_rate": 1e-05,
"loss": 0.8149,
"step": 13770
},
{
"epoch": 0.2621316745610531,
"grad_norm": 2.336155652999878,
"learning_rate": 1e-05,
"loss": 0.6196,
"step": 13780
},
{
"epoch": 0.26232190073997985,
"grad_norm": 2.8712093830108643,
"learning_rate": 1e-05,
"loss": 0.6647,
"step": 13790
},
{
"epoch": 0.2625121269189066,
"grad_norm": 6.549112796783447,
"learning_rate": 1e-05,
"loss": 0.7513,
"step": 13800
},
{
"epoch": 0.2627023530978333,
"grad_norm": 1.6849238872528076,
"learning_rate": 1e-05,
"loss": 0.8774,
"step": 13810
},
{
"epoch": 0.2628925792767601,
"grad_norm": 2.1048007011413574,
"learning_rate": 1e-05,
"loss": 0.8456,
"step": 13820
},
{
"epoch": 0.2630828054556868,
"grad_norm": 2.24222993850708,
"learning_rate": 1e-05,
"loss": 0.6526,
"step": 13830
},
{
"epoch": 0.2632730316346136,
"grad_norm": 2.8760387897491455,
"learning_rate": 1e-05,
"loss": 0.6531,
"step": 13840
},
{
"epoch": 0.2634632578135403,
"grad_norm": 8.288125991821289,
"learning_rate": 1e-05,
"loss": 0.8068,
"step": 13850
},
{
"epoch": 0.263653483992467,
"grad_norm": 1.4013222455978394,
"learning_rate": 1e-05,
"loss": 0.8563,
"step": 13860
},
{
"epoch": 0.2638437101713938,
"grad_norm": 1.674559235572815,
"learning_rate": 1e-05,
"loss": 0.8418,
"step": 13870
},
{
"epoch": 0.2640339363503205,
"grad_norm": 2.215543508529663,
"learning_rate": 1e-05,
"loss": 0.6867,
"step": 13880
},
{
"epoch": 0.2642241625292473,
"grad_norm": 2.9042210578918457,
"learning_rate": 1e-05,
"loss": 0.6535,
"step": 13890
},
{
"epoch": 0.264414388708174,
"grad_norm": 7.516767978668213,
"learning_rate": 1e-05,
"loss": 0.8198,
"step": 13900
},
{
"epoch": 0.26460461488710074,
"grad_norm": 1.7848894596099854,
"learning_rate": 1e-05,
"loss": 0.8682,
"step": 13910
},
{
"epoch": 0.2647948410660275,
"grad_norm": 2.2028093338012695,
"learning_rate": 1e-05,
"loss": 0.8053,
"step": 13920
},
{
"epoch": 0.26498506724495424,
"grad_norm": 2.1048998832702637,
"learning_rate": 1e-05,
"loss": 0.6609,
"step": 13930
},
{
"epoch": 0.265175293423881,
"grad_norm": 3.248894691467285,
"learning_rate": 1e-05,
"loss": 0.6401,
"step": 13940
},
{
"epoch": 0.26536551960280774,
"grad_norm": 9.998050689697266,
"learning_rate": 1e-05,
"loss": 0.8063,
"step": 13950
},
{
"epoch": 0.26555574578173446,
"grad_norm": 2.4291651248931885,
"learning_rate": 1e-05,
"loss": 0.8986,
"step": 13960
},
{
"epoch": 0.26574597196066124,
"grad_norm": 2.147336483001709,
"learning_rate": 1e-05,
"loss": 0.8006,
"step": 13970
},
{
"epoch": 0.26593619813958796,
"grad_norm": 3.2079710960388184,
"learning_rate": 1e-05,
"loss": 0.7033,
"step": 13980
},
{
"epoch": 0.2661264243185147,
"grad_norm": 2.52164888381958,
"learning_rate": 1e-05,
"loss": 0.6432,
"step": 13990
},
{
"epoch": 0.26631665049744147,
"grad_norm": 7.505655288696289,
"learning_rate": 1e-05,
"loss": 0.7923,
"step": 14000
},
{
"epoch": 0.2665068766763682,
"grad_norm": 2.1677727699279785,
"learning_rate": 1e-05,
"loss": 0.8853,
"step": 14010
},
{
"epoch": 0.26669710285529497,
"grad_norm": 2.2618408203125,
"learning_rate": 1e-05,
"loss": 0.842,
"step": 14020
},
{
"epoch": 0.2668873290342217,
"grad_norm": 2.5630223751068115,
"learning_rate": 1e-05,
"loss": 0.6452,
"step": 14030
},
{
"epoch": 0.2670775552131484,
"grad_norm": 2.7240421772003174,
"learning_rate": 1e-05,
"loss": 0.6585,
"step": 14040
},
{
"epoch": 0.2672677813920752,
"grad_norm": 7.502673149108887,
"learning_rate": 1e-05,
"loss": 0.7881,
"step": 14050
},
{
"epoch": 0.2674580075710019,
"grad_norm": 1.4357889890670776,
"learning_rate": 1e-05,
"loss": 0.8723,
"step": 14060
},
{
"epoch": 0.2676482337499287,
"grad_norm": 1.913122534751892,
"learning_rate": 1e-05,
"loss": 0.8176,
"step": 14070
},
{
"epoch": 0.2678384599288554,
"grad_norm": 2.229461193084717,
"learning_rate": 1e-05,
"loss": 0.6839,
"step": 14080
},
{
"epoch": 0.26802868610778213,
"grad_norm": 3.1840813159942627,
"learning_rate": 1e-05,
"loss": 0.6523,
"step": 14090
},
{
"epoch": 0.2682189122867089,
"grad_norm": 7.3194580078125,
"learning_rate": 1e-05,
"loss": 0.7949,
"step": 14100
},
{
"epoch": 0.26840913846563563,
"grad_norm": 1.605735421180725,
"learning_rate": 1e-05,
"loss": 0.8902,
"step": 14110
},
{
"epoch": 0.2685993646445624,
"grad_norm": 1.9395052194595337,
"learning_rate": 1e-05,
"loss": 0.8162,
"step": 14120
},
{
"epoch": 0.26878959082348913,
"grad_norm": 2.147433042526245,
"learning_rate": 1e-05,
"loss": 0.6348,
"step": 14130
},
{
"epoch": 0.26897981700241586,
"grad_norm": 3.189103603363037,
"learning_rate": 1e-05,
"loss": 0.6634,
"step": 14140
},
{
"epoch": 0.26917004318134263,
"grad_norm": 7.577734470367432,
"learning_rate": 1e-05,
"loss": 0.7908,
"step": 14150
},
{
"epoch": 0.26936026936026936,
"grad_norm": 1.5861204862594604,
"learning_rate": 1e-05,
"loss": 0.8658,
"step": 14160
},
{
"epoch": 0.2695504955391961,
"grad_norm": 1.5528852939605713,
"learning_rate": 1e-05,
"loss": 0.819,
"step": 14170
},
{
"epoch": 0.26974072171812286,
"grad_norm": 2.3312697410583496,
"learning_rate": 1e-05,
"loss": 0.6555,
"step": 14180
},
{
"epoch": 0.2699309478970496,
"grad_norm": 3.033217191696167,
"learning_rate": 1e-05,
"loss": 0.6424,
"step": 14190
},
{
"epoch": 0.27012117407597636,
"grad_norm": 7.0703511238098145,
"learning_rate": 1e-05,
"loss": 0.7949,
"step": 14200
},
{
"epoch": 0.2703114002549031,
"grad_norm": 1.5363426208496094,
"learning_rate": 1e-05,
"loss": 0.8869,
"step": 14210
},
{
"epoch": 0.2705016264338298,
"grad_norm": 2.165447235107422,
"learning_rate": 1e-05,
"loss": 0.82,
"step": 14220
},
{
"epoch": 0.2706918526127566,
"grad_norm": 2.7397921085357666,
"learning_rate": 1e-05,
"loss": 0.6905,
"step": 14230
},
{
"epoch": 0.2708820787916833,
"grad_norm": 3.9881529808044434,
"learning_rate": 1e-05,
"loss": 0.677,
"step": 14240
},
{
"epoch": 0.2710723049706101,
"grad_norm": 7.556854248046875,
"learning_rate": 1e-05,
"loss": 0.7926,
"step": 14250
},
{
"epoch": 0.2712625311495368,
"grad_norm": 2.1996848583221436,
"learning_rate": 1e-05,
"loss": 0.8791,
"step": 14260
},
{
"epoch": 0.2714527573284635,
"grad_norm": 1.9084094762802124,
"learning_rate": 1e-05,
"loss": 0.7795,
"step": 14270
},
{
"epoch": 0.2716429835073903,
"grad_norm": 2.744328498840332,
"learning_rate": 1e-05,
"loss": 0.642,
"step": 14280
},
{
"epoch": 0.271833209686317,
"grad_norm": 3.258359670639038,
"learning_rate": 1e-05,
"loss": 0.6631,
"step": 14290
},
{
"epoch": 0.2720234358652438,
"grad_norm": 7.210597038269043,
"learning_rate": 1e-05,
"loss": 0.7635,
"step": 14300
},
{
"epoch": 0.2722136620441705,
"grad_norm": 1.5209240913391113,
"learning_rate": 1e-05,
"loss": 0.8551,
"step": 14310
},
{
"epoch": 0.27240388822309725,
"grad_norm": 1.8904902935028076,
"learning_rate": 1e-05,
"loss": 0.8118,
"step": 14320
},
{
"epoch": 0.272594114402024,
"grad_norm": 2.1927058696746826,
"learning_rate": 1e-05,
"loss": 0.6817,
"step": 14330
},
{
"epoch": 0.27278434058095075,
"grad_norm": 2.927259922027588,
"learning_rate": 1e-05,
"loss": 0.6501,
"step": 14340
},
{
"epoch": 0.27297456675987747,
"grad_norm": 7.322691917419434,
"learning_rate": 1e-05,
"loss": 0.7506,
"step": 14350
},
{
"epoch": 0.27316479293880425,
"grad_norm": 1.5430567264556885,
"learning_rate": 1e-05,
"loss": 0.8528,
"step": 14360
},
{
"epoch": 0.27335501911773097,
"grad_norm": 2.5950794219970703,
"learning_rate": 1e-05,
"loss": 0.7964,
"step": 14370
},
{
"epoch": 0.27354524529665775,
"grad_norm": 2.008544445037842,
"learning_rate": 1e-05,
"loss": 0.6778,
"step": 14380
},
{
"epoch": 0.27373547147558447,
"grad_norm": 3.445490837097168,
"learning_rate": 1e-05,
"loss": 0.6289,
"step": 14390
},
{
"epoch": 0.2739256976545112,
"grad_norm": 7.733455181121826,
"learning_rate": 1e-05,
"loss": 0.7909,
"step": 14400
},
{
"epoch": 0.27411592383343797,
"grad_norm": 1.3672651052474976,
"learning_rate": 1e-05,
"loss": 0.8642,
"step": 14410
},
{
"epoch": 0.2743061500123647,
"grad_norm": 1.9755691289901733,
"learning_rate": 1e-05,
"loss": 0.8306,
"step": 14420
},
{
"epoch": 0.27449637619129147,
"grad_norm": 2.4375479221343994,
"learning_rate": 1e-05,
"loss": 0.6314,
"step": 14430
},
{
"epoch": 0.2746866023702182,
"grad_norm": 3.6759612560272217,
"learning_rate": 1e-05,
"loss": 0.6647,
"step": 14440
},
{
"epoch": 0.2748768285491449,
"grad_norm": 9.145861625671387,
"learning_rate": 1e-05,
"loss": 0.7885,
"step": 14450
},
{
"epoch": 0.2750670547280717,
"grad_norm": 1.5727124214172363,
"learning_rate": 1e-05,
"loss": 0.8722,
"step": 14460
},
{
"epoch": 0.2752572809069984,
"grad_norm": 2.4320907592773438,
"learning_rate": 1e-05,
"loss": 0.7852,
"step": 14470
},
{
"epoch": 0.2754475070859252,
"grad_norm": 2.339585542678833,
"learning_rate": 1e-05,
"loss": 0.6546,
"step": 14480
},
{
"epoch": 0.2756377332648519,
"grad_norm": 3.0163166522979736,
"learning_rate": 1e-05,
"loss": 0.6915,
"step": 14490
},
{
"epoch": 0.27582795944377864,
"grad_norm": 7.441839694976807,
"learning_rate": 1e-05,
"loss": 0.8146,
"step": 14500
},
{
"epoch": 0.2760181856227054,
"grad_norm": 1.7810478210449219,
"learning_rate": 1e-05,
"loss": 0.8356,
"step": 14510
},
{
"epoch": 0.27620841180163214,
"grad_norm": 1.6769726276397705,
"learning_rate": 1e-05,
"loss": 0.7822,
"step": 14520
},
{
"epoch": 0.27639863798055886,
"grad_norm": 2.1768784523010254,
"learning_rate": 1e-05,
"loss": 0.665,
"step": 14530
},
{
"epoch": 0.27658886415948564,
"grad_norm": 3.1060454845428467,
"learning_rate": 1e-05,
"loss": 0.6232,
"step": 14540
},
{
"epoch": 0.27677909033841236,
"grad_norm": 7.601314544677734,
"learning_rate": 1e-05,
"loss": 0.7755,
"step": 14550
},
{
"epoch": 0.27696931651733914,
"grad_norm": 1.4545385837554932,
"learning_rate": 1e-05,
"loss": 0.8769,
"step": 14560
},
{
"epoch": 0.27715954269626586,
"grad_norm": 1.9256722927093506,
"learning_rate": 1e-05,
"loss": 0.8175,
"step": 14570
},
{
"epoch": 0.2773497688751926,
"grad_norm": 2.372667074203491,
"learning_rate": 1e-05,
"loss": 0.6374,
"step": 14580
},
{
"epoch": 0.27753999505411936,
"grad_norm": 2.6548144817352295,
"learning_rate": 1e-05,
"loss": 0.6095,
"step": 14590
},
{
"epoch": 0.2777302212330461,
"grad_norm": 5.9940290451049805,
"learning_rate": 1e-05,
"loss": 0.7546,
"step": 14600
},
{
"epoch": 0.27792044741197286,
"grad_norm": 1.7117999792099,
"learning_rate": 1e-05,
"loss": 0.8566,
"step": 14610
},
{
"epoch": 0.2781106735908996,
"grad_norm": 1.7675544023513794,
"learning_rate": 1e-05,
"loss": 0.8202,
"step": 14620
},
{
"epoch": 0.2783008997698263,
"grad_norm": 2.8261663913726807,
"learning_rate": 1e-05,
"loss": 0.656,
"step": 14630
},
{
"epoch": 0.2784911259487531,
"grad_norm": 3.2293360233306885,
"learning_rate": 1e-05,
"loss": 0.6668,
"step": 14640
},
{
"epoch": 0.2786813521276798,
"grad_norm": 7.594730854034424,
"learning_rate": 1e-05,
"loss": 0.7645,
"step": 14650
},
{
"epoch": 0.2788715783066066,
"grad_norm": 2.0959665775299072,
"learning_rate": 1e-05,
"loss": 0.8706,
"step": 14660
},
{
"epoch": 0.2790618044855333,
"grad_norm": 1.9151928424835205,
"learning_rate": 1e-05,
"loss": 0.8327,
"step": 14670
},
{
"epoch": 0.27925203066446,
"grad_norm": 2.522365093231201,
"learning_rate": 1e-05,
"loss": 0.686,
"step": 14680
},
{
"epoch": 0.2794422568433868,
"grad_norm": 3.016352653503418,
"learning_rate": 1e-05,
"loss": 0.6403,
"step": 14690
},
{
"epoch": 0.2796324830223135,
"grad_norm": 8.589140892028809,
"learning_rate": 1e-05,
"loss": 0.7975,
"step": 14700
},
{
"epoch": 0.27982270920124025,
"grad_norm": 1.7063238620758057,
"learning_rate": 1e-05,
"loss": 0.8667,
"step": 14710
},
{
"epoch": 0.280012935380167,
"grad_norm": 1.881842017173767,
"learning_rate": 1e-05,
"loss": 0.7922,
"step": 14720
},
{
"epoch": 0.28020316155909375,
"grad_norm": 2.4525134563446045,
"learning_rate": 1e-05,
"loss": 0.6455,
"step": 14730
},
{
"epoch": 0.2803933877380205,
"grad_norm": 2.8054866790771484,
"learning_rate": 1e-05,
"loss": 0.664,
"step": 14740
},
{
"epoch": 0.28058361391694725,
"grad_norm": 7.305141925811768,
"learning_rate": 1e-05,
"loss": 0.8027,
"step": 14750
},
{
"epoch": 0.28077384009587397,
"grad_norm": 1.7342288494110107,
"learning_rate": 1e-05,
"loss": 0.8735,
"step": 14760
},
{
"epoch": 0.28096406627480075,
"grad_norm": 2.190155267715454,
"learning_rate": 1e-05,
"loss": 0.8022,
"step": 14770
},
{
"epoch": 0.2811542924537275,
"grad_norm": 2.4192986488342285,
"learning_rate": 1e-05,
"loss": 0.6382,
"step": 14780
},
{
"epoch": 0.28134451863265425,
"grad_norm": 2.657390832901001,
"learning_rate": 1e-05,
"loss": 0.6559,
"step": 14790
},
{
"epoch": 0.281534744811581,
"grad_norm": 8.331223487854004,
"learning_rate": 1e-05,
"loss": 0.8003,
"step": 14800
},
{
"epoch": 0.2817249709905077,
"grad_norm": 1.5674375295639038,
"learning_rate": 1e-05,
"loss": 0.8816,
"step": 14810
},
{
"epoch": 0.2819151971694345,
"grad_norm": 1.9171380996704102,
"learning_rate": 1e-05,
"loss": 0.8364,
"step": 14820
},
{
"epoch": 0.2821054233483612,
"grad_norm": 2.3145523071289062,
"learning_rate": 1e-05,
"loss": 0.6687,
"step": 14830
},
{
"epoch": 0.282295649527288,
"grad_norm": 3.048172950744629,
"learning_rate": 1e-05,
"loss": 0.6288,
"step": 14840
},
{
"epoch": 0.2824858757062147,
"grad_norm": 9.130074501037598,
"learning_rate": 1e-05,
"loss": 0.7882,
"step": 14850
},
{
"epoch": 0.2826761018851414,
"grad_norm": 1.5021427869796753,
"learning_rate": 1e-05,
"loss": 0.8445,
"step": 14860
},
{
"epoch": 0.2828663280640682,
"grad_norm": 1.6946682929992676,
"learning_rate": 1e-05,
"loss": 0.8165,
"step": 14870
},
{
"epoch": 0.2830565542429949,
"grad_norm": 2.3710122108459473,
"learning_rate": 1e-05,
"loss": 0.6142,
"step": 14880
},
{
"epoch": 0.28324678042192164,
"grad_norm": 3.219972848892212,
"learning_rate": 1e-05,
"loss": 0.6338,
"step": 14890
},
{
"epoch": 0.2834370066008484,
"grad_norm": 8.351262092590332,
"learning_rate": 1e-05,
"loss": 0.7789,
"step": 14900
},
{
"epoch": 0.28362723277977514,
"grad_norm": 1.4915364980697632,
"learning_rate": 1e-05,
"loss": 0.8644,
"step": 14910
},
{
"epoch": 0.2838174589587019,
"grad_norm": 1.9973068237304688,
"learning_rate": 1e-05,
"loss": 0.8181,
"step": 14920
},
{
"epoch": 0.28400768513762864,
"grad_norm": 2.6096749305725098,
"learning_rate": 1e-05,
"loss": 0.6351,
"step": 14930
},
{
"epoch": 0.28419791131655536,
"grad_norm": 3.0387916564941406,
"learning_rate": 1e-05,
"loss": 0.649,
"step": 14940
},
{
"epoch": 0.28438813749548214,
"grad_norm": 9.28211784362793,
"learning_rate": 1e-05,
"loss": 0.8039,
"step": 14950
},
{
"epoch": 0.28457836367440886,
"grad_norm": 2.1682565212249756,
"learning_rate": 1e-05,
"loss": 0.8681,
"step": 14960
},
{
"epoch": 0.28476858985333564,
"grad_norm": 1.9600428342819214,
"learning_rate": 1e-05,
"loss": 0.8488,
"step": 14970
},
{
"epoch": 0.28495881603226236,
"grad_norm": 2.605666160583496,
"learning_rate": 1e-05,
"loss": 0.652,
"step": 14980
},
{
"epoch": 0.2851490422111891,
"grad_norm": 3.303462028503418,
"learning_rate": 1e-05,
"loss": 0.6382,
"step": 14990
},
{
"epoch": 0.28533926839011586,
"grad_norm": 7.363654136657715,
"learning_rate": 1e-05,
"loss": 0.7854,
"step": 15000
},
{
"epoch": 0.2855294945690426,
"grad_norm": 1.9826524257659912,
"learning_rate": 1e-05,
"loss": 0.8481,
"step": 15010
},
{
"epoch": 0.2857197207479693,
"grad_norm": 2.012927532196045,
"learning_rate": 1e-05,
"loss": 0.7991,
"step": 15020
},
{
"epoch": 0.2859099469268961,
"grad_norm": 2.1477842330932617,
"learning_rate": 1e-05,
"loss": 0.6363,
"step": 15030
},
{
"epoch": 0.2861001731058228,
"grad_norm": 3.0967814922332764,
"learning_rate": 1e-05,
"loss": 0.6554,
"step": 15040
},
{
"epoch": 0.2862903992847496,
"grad_norm": 6.655501842498779,
"learning_rate": 1e-05,
"loss": 0.7585,
"step": 15050
},
{
"epoch": 0.2864806254636763,
"grad_norm": 1.7653157711029053,
"learning_rate": 1e-05,
"loss": 0.8757,
"step": 15060
},
{
"epoch": 0.28667085164260303,
"grad_norm": 1.907548189163208,
"learning_rate": 1e-05,
"loss": 0.8465,
"step": 15070
},
{
"epoch": 0.2868610778215298,
"grad_norm": 2.543829917907715,
"learning_rate": 1e-05,
"loss": 0.6446,
"step": 15080
},
{
"epoch": 0.28705130400045653,
"grad_norm": 2.998492479324341,
"learning_rate": 1e-05,
"loss": 0.6334,
"step": 15090
},
{
"epoch": 0.2872415301793833,
"grad_norm": 8.46926498413086,
"learning_rate": 1e-05,
"loss": 0.762,
"step": 15100
},
{
"epoch": 0.28743175635831003,
"grad_norm": 2.2793540954589844,
"learning_rate": 1e-05,
"loss": 0.8644,
"step": 15110
},
{
"epoch": 0.28762198253723675,
"grad_norm": 1.9191033840179443,
"learning_rate": 1e-05,
"loss": 0.8134,
"step": 15120
},
{
"epoch": 0.28781220871616353,
"grad_norm": 1.9256781339645386,
"learning_rate": 1e-05,
"loss": 0.6757,
"step": 15130
},
{
"epoch": 0.28800243489509025,
"grad_norm": 2.737562894821167,
"learning_rate": 1e-05,
"loss": 0.6345,
"step": 15140
},
{
"epoch": 0.28819266107401703,
"grad_norm": 9.471795082092285,
"learning_rate": 1e-05,
"loss": 0.8122,
"step": 15150
},
{
"epoch": 0.28838288725294375,
"grad_norm": 1.6299372911453247,
"learning_rate": 1e-05,
"loss": 0.8483,
"step": 15160
},
{
"epoch": 0.2885731134318705,
"grad_norm": 1.8826267719268799,
"learning_rate": 1e-05,
"loss": 0.7852,
"step": 15170
},
{
"epoch": 0.28876333961079725,
"grad_norm": 2.1511967182159424,
"learning_rate": 1e-05,
"loss": 0.6573,
"step": 15180
},
{
"epoch": 0.288953565789724,
"grad_norm": 3.9024434089660645,
"learning_rate": 1e-05,
"loss": 0.6605,
"step": 15190
},
{
"epoch": 0.2891437919686507,
"grad_norm": 6.50871467590332,
"learning_rate": 1e-05,
"loss": 0.7296,
"step": 15200
},
{
"epoch": 0.2893340181475775,
"grad_norm": 1.4598047733306885,
"learning_rate": 1e-05,
"loss": 0.8574,
"step": 15210
},
{
"epoch": 0.2895242443265042,
"grad_norm": 1.9483811855316162,
"learning_rate": 1e-05,
"loss": 0.8324,
"step": 15220
},
{
"epoch": 0.289714470505431,
"grad_norm": 2.1774086952209473,
"learning_rate": 1e-05,
"loss": 0.6395,
"step": 15230
},
{
"epoch": 0.2899046966843577,
"grad_norm": 3.2058489322662354,
"learning_rate": 1e-05,
"loss": 0.6342,
"step": 15240
},
{
"epoch": 0.2900949228632844,
"grad_norm": 6.3300862312316895,
"learning_rate": 1e-05,
"loss": 0.7943,
"step": 15250
},
{
"epoch": 0.2902851490422112,
"grad_norm": 1.5692402124404907,
"learning_rate": 1e-05,
"loss": 0.8574,
"step": 15260
},
{
"epoch": 0.2904753752211379,
"grad_norm": 1.789862871170044,
"learning_rate": 1e-05,
"loss": 0.8306,
"step": 15270
},
{
"epoch": 0.2906656014000647,
"grad_norm": 2.18183970451355,
"learning_rate": 1e-05,
"loss": 0.6689,
"step": 15280
},
{
"epoch": 0.2908558275789914,
"grad_norm": 2.8390350341796875,
"learning_rate": 1e-05,
"loss": 0.63,
"step": 15290
},
{
"epoch": 0.29104605375791814,
"grad_norm": 6.4474639892578125,
"learning_rate": 1e-05,
"loss": 0.7731,
"step": 15300
},
{
"epoch": 0.2912362799368449,
"grad_norm": 1.421912670135498,
"learning_rate": 1e-05,
"loss": 0.8658,
"step": 15310
},
{
"epoch": 0.29142650611577164,
"grad_norm": 2.0187103748321533,
"learning_rate": 1e-05,
"loss": 0.7831,
"step": 15320
},
{
"epoch": 0.2916167322946984,
"grad_norm": 2.346890926361084,
"learning_rate": 1e-05,
"loss": 0.592,
"step": 15330
},
{
"epoch": 0.29180695847362514,
"grad_norm": 2.880852699279785,
"learning_rate": 1e-05,
"loss": 0.6396,
"step": 15340
},
{
"epoch": 0.29199718465255187,
"grad_norm": 9.754484176635742,
"learning_rate": 1e-05,
"loss": 0.8041,
"step": 15350
},
{
"epoch": 0.29218741083147864,
"grad_norm": 1.4410114288330078,
"learning_rate": 1e-05,
"loss": 0.8438,
"step": 15360
},
{
"epoch": 0.29237763701040537,
"grad_norm": 1.7649128437042236,
"learning_rate": 1e-05,
"loss": 0.842,
"step": 15370
},
{
"epoch": 0.2925678631893321,
"grad_norm": 2.5158565044403076,
"learning_rate": 1e-05,
"loss": 0.6822,
"step": 15380
},
{
"epoch": 0.29275808936825887,
"grad_norm": 2.4811694622039795,
"learning_rate": 1e-05,
"loss": 0.6186,
"step": 15390
},
{
"epoch": 0.2929483155471856,
"grad_norm": 7.777284622192383,
"learning_rate": 1e-05,
"loss": 0.7985,
"step": 15400
},
{
"epoch": 0.29313854172611237,
"grad_norm": 1.5489392280578613,
"learning_rate": 1e-05,
"loss": 0.857,
"step": 15410
},
{
"epoch": 0.2933287679050391,
"grad_norm": 1.4409061670303345,
"learning_rate": 1e-05,
"loss": 0.7677,
"step": 15420
},
{
"epoch": 0.2935189940839658,
"grad_norm": 2.335170030593872,
"learning_rate": 1e-05,
"loss": 0.6796,
"step": 15430
},
{
"epoch": 0.2937092202628926,
"grad_norm": 2.682518243789673,
"learning_rate": 1e-05,
"loss": 0.6388,
"step": 15440
},
{
"epoch": 0.2938994464418193,
"grad_norm": 7.9400553703308105,
"learning_rate": 1e-05,
"loss": 0.7537,
"step": 15450
},
{
"epoch": 0.2940896726207461,
"grad_norm": 1.3789311647415161,
"learning_rate": 1e-05,
"loss": 0.8371,
"step": 15460
},
{
"epoch": 0.2942798987996728,
"grad_norm": 1.7134215831756592,
"learning_rate": 1e-05,
"loss": 0.7804,
"step": 15470
},
{
"epoch": 0.29447012497859953,
"grad_norm": 2.3252980709075928,
"learning_rate": 1e-05,
"loss": 0.6379,
"step": 15480
},
{
"epoch": 0.2946603511575263,
"grad_norm": 2.6960790157318115,
"learning_rate": 1e-05,
"loss": 0.6441,
"step": 15490
},
{
"epoch": 0.29485057733645303,
"grad_norm": 7.48028039932251,
"learning_rate": 1e-05,
"loss": 0.7999,
"step": 15500
},
{
"epoch": 0.2950408035153798,
"grad_norm": 1.6065644025802612,
"learning_rate": 1e-05,
"loss": 0.8588,
"step": 15510
},
{
"epoch": 0.29523102969430653,
"grad_norm": 1.789211392402649,
"learning_rate": 1e-05,
"loss": 0.8266,
"step": 15520
},
{
"epoch": 0.29542125587323326,
"grad_norm": 2.702014684677124,
"learning_rate": 1e-05,
"loss": 0.7141,
"step": 15530
},
{
"epoch": 0.29561148205216004,
"grad_norm": 2.7151732444763184,
"learning_rate": 1e-05,
"loss": 0.6204,
"step": 15540
},
{
"epoch": 0.29580170823108676,
"grad_norm": 6.7979512214660645,
"learning_rate": 1e-05,
"loss": 0.7588,
"step": 15550
},
{
"epoch": 0.2959919344100135,
"grad_norm": 1.5403528213500977,
"learning_rate": 1e-05,
"loss": 0.8793,
"step": 15560
},
{
"epoch": 0.29618216058894026,
"grad_norm": 1.6201974153518677,
"learning_rate": 1e-05,
"loss": 0.8032,
"step": 15570
},
{
"epoch": 0.296372386767867,
"grad_norm": 2.0255391597747803,
"learning_rate": 1e-05,
"loss": 0.6353,
"step": 15580
},
{
"epoch": 0.29656261294679376,
"grad_norm": 2.948060989379883,
"learning_rate": 1e-05,
"loss": 0.6332,
"step": 15590
},
{
"epoch": 0.2967528391257205,
"grad_norm": 7.696873188018799,
"learning_rate": 1e-05,
"loss": 0.8034,
"step": 15600
},
{
"epoch": 0.2969430653046472,
"grad_norm": 1.5631297826766968,
"learning_rate": 1e-05,
"loss": 0.8347,
"step": 15610
},
{
"epoch": 0.297133291483574,
"grad_norm": 1.65414559841156,
"learning_rate": 1e-05,
"loss": 0.8105,
"step": 15620
},
{
"epoch": 0.2973235176625007,
"grad_norm": 2.1988933086395264,
"learning_rate": 1e-05,
"loss": 0.6641,
"step": 15630
},
{
"epoch": 0.2975137438414275,
"grad_norm": 3.074617862701416,
"learning_rate": 1e-05,
"loss": 0.6052,
"step": 15640
},
{
"epoch": 0.2977039700203542,
"grad_norm": 8.723734855651855,
"learning_rate": 1e-05,
"loss": 0.7711,
"step": 15650
},
{
"epoch": 0.2978941961992809,
"grad_norm": 1.7039355039596558,
"learning_rate": 1e-05,
"loss": 0.9062,
"step": 15660
},
{
"epoch": 0.2980844223782077,
"grad_norm": 1.8725625276565552,
"learning_rate": 1e-05,
"loss": 0.8125,
"step": 15670
},
{
"epoch": 0.2982746485571344,
"grad_norm": 1.8811205625534058,
"learning_rate": 1e-05,
"loss": 0.6284,
"step": 15680
},
{
"epoch": 0.2984648747360612,
"grad_norm": 2.8209848403930664,
"learning_rate": 1e-05,
"loss": 0.6293,
"step": 15690
},
{
"epoch": 0.2986551009149879,
"grad_norm": 7.707814693450928,
"learning_rate": 1e-05,
"loss": 0.7995,
"step": 15700
},
{
"epoch": 0.29884532709391465,
"grad_norm": 1.4202220439910889,
"learning_rate": 1e-05,
"loss": 0.8707,
"step": 15710
},
{
"epoch": 0.2990355532728414,
"grad_norm": 2.014759063720703,
"learning_rate": 1e-05,
"loss": 0.8067,
"step": 15720
},
{
"epoch": 0.29922577945176815,
"grad_norm": 2.167865753173828,
"learning_rate": 1e-05,
"loss": 0.6378,
"step": 15730
},
{
"epoch": 0.29941600563069487,
"grad_norm": 2.86718487739563,
"learning_rate": 1e-05,
"loss": 0.6501,
"step": 15740
},
{
"epoch": 0.29960623180962165,
"grad_norm": 7.032610893249512,
"learning_rate": 1e-05,
"loss": 0.7707,
"step": 15750
},
{
"epoch": 0.29979645798854837,
"grad_norm": 1.7989959716796875,
"learning_rate": 1e-05,
"loss": 0.8947,
"step": 15760
},
{
"epoch": 0.29998668416747515,
"grad_norm": 1.6317957639694214,
"learning_rate": 1e-05,
"loss": 0.7795,
"step": 15770
},
{
"epoch": 0.30017691034640187,
"grad_norm": 2.062312364578247,
"learning_rate": 1e-05,
"loss": 0.6479,
"step": 15780
},
{
"epoch": 0.3003671365253286,
"grad_norm": 3.1726346015930176,
"learning_rate": 1e-05,
"loss": 0.6432,
"step": 15790
},
{
"epoch": 0.30055736270425537,
"grad_norm": 6.379366874694824,
"learning_rate": 1e-05,
"loss": 0.7566,
"step": 15800
},
{
"epoch": 0.3007475888831821,
"grad_norm": 1.74537992477417,
"learning_rate": 1e-05,
"loss": 0.8655,
"step": 15810
},
{
"epoch": 0.30093781506210887,
"grad_norm": 2.377631902694702,
"learning_rate": 1e-05,
"loss": 0.8043,
"step": 15820
},
{
"epoch": 0.3011280412410356,
"grad_norm": 2.2452874183654785,
"learning_rate": 1e-05,
"loss": 0.6609,
"step": 15830
},
{
"epoch": 0.3013182674199623,
"grad_norm": 2.7078030109405518,
"learning_rate": 1e-05,
"loss": 0.6536,
"step": 15840
},
{
"epoch": 0.3015084935988891,
"grad_norm": 7.873197555541992,
"learning_rate": 1e-05,
"loss": 0.8253,
"step": 15850
},
{
"epoch": 0.3016987197778158,
"grad_norm": 1.848899483680725,
"learning_rate": 1e-05,
"loss": 0.857,
"step": 15860
},
{
"epoch": 0.3018889459567426,
"grad_norm": 1.7001627683639526,
"learning_rate": 1e-05,
"loss": 0.8171,
"step": 15870
},
{
"epoch": 0.3020791721356693,
"grad_norm": 2.9449894428253174,
"learning_rate": 1e-05,
"loss": 0.6494,
"step": 15880
},
{
"epoch": 0.30226939831459604,
"grad_norm": 3.467072010040283,
"learning_rate": 1e-05,
"loss": 0.6766,
"step": 15890
},
{
"epoch": 0.3024596244935228,
"grad_norm": 7.12153959274292,
"learning_rate": 1e-05,
"loss": 0.7532,
"step": 15900
},
{
"epoch": 0.30264985067244954,
"grad_norm": 1.415823221206665,
"learning_rate": 1e-05,
"loss": 0.847,
"step": 15910
},
{
"epoch": 0.30284007685137626,
"grad_norm": 1.9057809114456177,
"learning_rate": 1e-05,
"loss": 0.7953,
"step": 15920
},
{
"epoch": 0.30303030303030304,
"grad_norm": 3.1442956924438477,
"learning_rate": 1e-05,
"loss": 0.662,
"step": 15930
},
{
"epoch": 0.30322052920922976,
"grad_norm": 3.0042428970336914,
"learning_rate": 1e-05,
"loss": 0.6381,
"step": 15940
},
{
"epoch": 0.30341075538815654,
"grad_norm": 6.438754558563232,
"learning_rate": 1e-05,
"loss": 0.7573,
"step": 15950
},
{
"epoch": 0.30360098156708326,
"grad_norm": 1.6751583814620972,
"learning_rate": 1e-05,
"loss": 0.8344,
"step": 15960
},
{
"epoch": 0.30379120774601,
"grad_norm": 2.1287903785705566,
"learning_rate": 1e-05,
"loss": 0.8072,
"step": 15970
},
{
"epoch": 0.30398143392493676,
"grad_norm": 2.2405827045440674,
"learning_rate": 1e-05,
"loss": 0.6519,
"step": 15980
},
{
"epoch": 0.3041716601038635,
"grad_norm": 3.1466052532196045,
"learning_rate": 1e-05,
"loss": 0.612,
"step": 15990
},
{
"epoch": 0.30436188628279026,
"grad_norm": 7.578444480895996,
"learning_rate": 1e-05,
"loss": 0.774,
"step": 16000
},
{
"epoch": 0.304552112461717,
"grad_norm": 1.6151862144470215,
"learning_rate": 1e-05,
"loss": 0.8601,
"step": 16010
},
{
"epoch": 0.3047423386406437,
"grad_norm": 2.017993450164795,
"learning_rate": 1e-05,
"loss": 0.7716,
"step": 16020
},
{
"epoch": 0.3049325648195705,
"grad_norm": 2.9746615886688232,
"learning_rate": 1e-05,
"loss": 0.6447,
"step": 16030
},
{
"epoch": 0.3051227909984972,
"grad_norm": 3.3719658851623535,
"learning_rate": 1e-05,
"loss": 0.6327,
"step": 16040
},
{
"epoch": 0.305313017177424,
"grad_norm": 6.4755144119262695,
"learning_rate": 1e-05,
"loss": 0.7417,
"step": 16050
},
{
"epoch": 0.3055032433563507,
"grad_norm": 1.7517694234848022,
"learning_rate": 1e-05,
"loss": 0.8549,
"step": 16060
},
{
"epoch": 0.30569346953527743,
"grad_norm": 1.574364185333252,
"learning_rate": 1e-05,
"loss": 0.7834,
"step": 16070
},
{
"epoch": 0.3058836957142042,
"grad_norm": 2.175368070602417,
"learning_rate": 1e-05,
"loss": 0.5863,
"step": 16080
},
{
"epoch": 0.30607392189313093,
"grad_norm": 2.425565242767334,
"learning_rate": 1e-05,
"loss": 0.6499,
"step": 16090
},
{
"epoch": 0.30626414807205765,
"grad_norm": 6.630768299102783,
"learning_rate": 1e-05,
"loss": 0.7639,
"step": 16100
},
{
"epoch": 0.30645437425098443,
"grad_norm": 1.853054404258728,
"learning_rate": 1e-05,
"loss": 0.8764,
"step": 16110
},
{
"epoch": 0.30664460042991115,
"grad_norm": 1.6852649450302124,
"learning_rate": 1e-05,
"loss": 0.785,
"step": 16120
},
{
"epoch": 0.30683482660883793,
"grad_norm": 2.204566240310669,
"learning_rate": 1e-05,
"loss": 0.7049,
"step": 16130
},
{
"epoch": 0.30702505278776465,
"grad_norm": 2.9731392860412598,
"learning_rate": 1e-05,
"loss": 0.6338,
"step": 16140
},
{
"epoch": 0.3072152789666914,
"grad_norm": 7.570809841156006,
"learning_rate": 1e-05,
"loss": 0.7554,
"step": 16150
},
{
"epoch": 0.30740550514561815,
"grad_norm": 1.6046339273452759,
"learning_rate": 1e-05,
"loss": 0.8625,
"step": 16160
},
{
"epoch": 0.3075957313245449,
"grad_norm": 1.660595178604126,
"learning_rate": 1e-05,
"loss": 0.8186,
"step": 16170
},
{
"epoch": 0.30778595750347165,
"grad_norm": 2.438683271408081,
"learning_rate": 1e-05,
"loss": 0.6621,
"step": 16180
},
{
"epoch": 0.3079761836823984,
"grad_norm": 2.4547557830810547,
"learning_rate": 1e-05,
"loss": 0.6153,
"step": 16190
},
{
"epoch": 0.3081664098613251,
"grad_norm": 7.473080635070801,
"learning_rate": 1e-05,
"loss": 0.7992,
"step": 16200
},
{
"epoch": 0.3083566360402519,
"grad_norm": 2.110948085784912,
"learning_rate": 1e-05,
"loss": 0.8484,
"step": 16210
},
{
"epoch": 0.3085468622191786,
"grad_norm": 1.4378507137298584,
"learning_rate": 1e-05,
"loss": 0.7591,
"step": 16220
},
{
"epoch": 0.3087370883981054,
"grad_norm": 2.265608549118042,
"learning_rate": 1e-05,
"loss": 0.6511,
"step": 16230
},
{
"epoch": 0.3089273145770321,
"grad_norm": 2.9768824577331543,
"learning_rate": 1e-05,
"loss": 0.6333,
"step": 16240
},
{
"epoch": 0.3091175407559588,
"grad_norm": 5.973684787750244,
"learning_rate": 1e-05,
"loss": 0.7638,
"step": 16250
},
{
"epoch": 0.3093077669348856,
"grad_norm": 1.8720929622650146,
"learning_rate": 1e-05,
"loss": 0.8657,
"step": 16260
},
{
"epoch": 0.3094979931138123,
"grad_norm": 1.8256558179855347,
"learning_rate": 1e-05,
"loss": 0.8253,
"step": 16270
},
{
"epoch": 0.30968821929273904,
"grad_norm": 2.33085036277771,
"learning_rate": 1e-05,
"loss": 0.6558,
"step": 16280
},
{
"epoch": 0.3098784454716658,
"grad_norm": 2.6149489879608154,
"learning_rate": 1e-05,
"loss": 0.6367,
"step": 16290
},
{
"epoch": 0.31006867165059254,
"grad_norm": 7.9340081214904785,
"learning_rate": 1e-05,
"loss": 0.7471,
"step": 16300
},
{
"epoch": 0.3102588978295193,
"grad_norm": 1.554776668548584,
"learning_rate": 1e-05,
"loss": 0.8505,
"step": 16310
},
{
"epoch": 0.31044912400844604,
"grad_norm": 1.770323395729065,
"learning_rate": 1e-05,
"loss": 0.7853,
"step": 16320
},
{
"epoch": 0.31063935018737276,
"grad_norm": 2.2204089164733887,
"learning_rate": 1e-05,
"loss": 0.6496,
"step": 16330
},
{
"epoch": 0.31082957636629954,
"grad_norm": 3.55918025970459,
"learning_rate": 1e-05,
"loss": 0.6555,
"step": 16340
},
{
"epoch": 0.31101980254522626,
"grad_norm": 8.057416915893555,
"learning_rate": 1e-05,
"loss": 0.7748,
"step": 16350
},
{
"epoch": 0.31121002872415304,
"grad_norm": 2.0971598625183105,
"learning_rate": 1e-05,
"loss": 0.8394,
"step": 16360
},
{
"epoch": 0.31140025490307977,
"grad_norm": 1.75458562374115,
"learning_rate": 1e-05,
"loss": 0.7784,
"step": 16370
},
{
"epoch": 0.3115904810820065,
"grad_norm": 2.14908504486084,
"learning_rate": 1e-05,
"loss": 0.6873,
"step": 16380
},
{
"epoch": 0.31178070726093327,
"grad_norm": 2.843341112136841,
"learning_rate": 1e-05,
"loss": 0.6425,
"step": 16390
},
{
"epoch": 0.31197093343986,
"grad_norm": 6.85654878616333,
"learning_rate": 1e-05,
"loss": 0.7818,
"step": 16400
},
{
"epoch": 0.3121611596187867,
"grad_norm": 1.5873843431472778,
"learning_rate": 1e-05,
"loss": 0.857,
"step": 16410
},
{
"epoch": 0.3123513857977135,
"grad_norm": 1.8644886016845703,
"learning_rate": 1e-05,
"loss": 0.7697,
"step": 16420
},
{
"epoch": 0.3125416119766402,
"grad_norm": 2.434981107711792,
"learning_rate": 1e-05,
"loss": 0.6105,
"step": 16430
},
{
"epoch": 0.312731838155567,
"grad_norm": 2.9216392040252686,
"learning_rate": 1e-05,
"loss": 0.6239,
"step": 16440
},
{
"epoch": 0.3129220643344937,
"grad_norm": 7.214798450469971,
"learning_rate": 1e-05,
"loss": 0.7389,
"step": 16450
},
{
"epoch": 0.31311229051342043,
"grad_norm": 1.6312718391418457,
"learning_rate": 1e-05,
"loss": 0.8728,
"step": 16460
},
{
"epoch": 0.3133025166923472,
"grad_norm": 1.5557498931884766,
"learning_rate": 1e-05,
"loss": 0.8483,
"step": 16470
},
{
"epoch": 0.31349274287127393,
"grad_norm": 2.1639347076416016,
"learning_rate": 1e-05,
"loss": 0.6382,
"step": 16480
},
{
"epoch": 0.3136829690502007,
"grad_norm": 2.782951831817627,
"learning_rate": 1e-05,
"loss": 0.6015,
"step": 16490
},
{
"epoch": 0.31387319522912743,
"grad_norm": 7.034222602844238,
"learning_rate": 1e-05,
"loss": 0.7714,
"step": 16500
},
{
"epoch": 0.31406342140805416,
"grad_norm": 2.4815218448638916,
"learning_rate": 1e-05,
"loss": 0.8301,
"step": 16510
},
{
"epoch": 0.31425364758698093,
"grad_norm": 1.8007875680923462,
"learning_rate": 1e-05,
"loss": 0.8119,
"step": 16520
},
{
"epoch": 0.31444387376590766,
"grad_norm": 2.028233289718628,
"learning_rate": 1e-05,
"loss": 0.6591,
"step": 16530
},
{
"epoch": 0.31463409994483443,
"grad_norm": 2.870704174041748,
"learning_rate": 1e-05,
"loss": 0.6034,
"step": 16540
},
{
"epoch": 0.31482432612376116,
"grad_norm": 6.206780910491943,
"learning_rate": 1e-05,
"loss": 0.7657,
"step": 16550
},
{
"epoch": 0.3150145523026879,
"grad_norm": 1.396316409111023,
"learning_rate": 1e-05,
"loss": 0.8512,
"step": 16560
},
{
"epoch": 0.31520477848161466,
"grad_norm": 2.0313045978546143,
"learning_rate": 1e-05,
"loss": 0.7784,
"step": 16570
},
{
"epoch": 0.3153950046605414,
"grad_norm": 2.118194341659546,
"learning_rate": 1e-05,
"loss": 0.6285,
"step": 16580
},
{
"epoch": 0.3155852308394681,
"grad_norm": 2.7404439449310303,
"learning_rate": 1e-05,
"loss": 0.6092,
"step": 16590
},
{
"epoch": 0.3157754570183949,
"grad_norm": 6.128175258636475,
"learning_rate": 1e-05,
"loss": 0.8047,
"step": 16600
},
{
"epoch": 0.3159656831973216,
"grad_norm": 1.7313777208328247,
"learning_rate": 1e-05,
"loss": 0.836,
"step": 16610
},
{
"epoch": 0.3161559093762484,
"grad_norm": 2.262467861175537,
"learning_rate": 1e-05,
"loss": 0.8053,
"step": 16620
},
{
"epoch": 0.3163461355551751,
"grad_norm": 1.9950371980667114,
"learning_rate": 1e-05,
"loss": 0.6821,
"step": 16630
},
{
"epoch": 0.3165363617341018,
"grad_norm": 2.9390523433685303,
"learning_rate": 1e-05,
"loss": 0.6111,
"step": 16640
},
{
"epoch": 0.3167265879130286,
"grad_norm": 8.057096481323242,
"learning_rate": 1e-05,
"loss": 0.7712,
"step": 16650
},
{
"epoch": 0.3169168140919553,
"grad_norm": 1.3481460809707642,
"learning_rate": 1e-05,
"loss": 0.8274,
"step": 16660
},
{
"epoch": 0.3171070402708821,
"grad_norm": 2.0134177207946777,
"learning_rate": 1e-05,
"loss": 0.8356,
"step": 16670
},
{
"epoch": 0.3172972664498088,
"grad_norm": 2.1858694553375244,
"learning_rate": 1e-05,
"loss": 0.6282,
"step": 16680
},
{
"epoch": 0.31748749262873555,
"grad_norm": 2.8450186252593994,
"learning_rate": 1e-05,
"loss": 0.6342,
"step": 16690
},
{
"epoch": 0.3176777188076623,
"grad_norm": 7.341821670532227,
"learning_rate": 1e-05,
"loss": 0.7793,
"step": 16700
},
{
"epoch": 0.31786794498658905,
"grad_norm": 1.7662394046783447,
"learning_rate": 1e-05,
"loss": 0.8629,
"step": 16710
},
{
"epoch": 0.3180581711655158,
"grad_norm": 1.781969428062439,
"learning_rate": 1e-05,
"loss": 0.7844,
"step": 16720
},
{
"epoch": 0.31824839734444255,
"grad_norm": 2.096881628036499,
"learning_rate": 1e-05,
"loss": 0.6706,
"step": 16730
},
{
"epoch": 0.31843862352336927,
"grad_norm": 2.65963077545166,
"learning_rate": 1e-05,
"loss": 0.6359,
"step": 16740
},
{
"epoch": 0.31862884970229605,
"grad_norm": 6.949047565460205,
"learning_rate": 1e-05,
"loss": 0.7419,
"step": 16750
},
{
"epoch": 0.31881907588122277,
"grad_norm": 1.4740442037582397,
"learning_rate": 1e-05,
"loss": 0.8186,
"step": 16760
},
{
"epoch": 0.3190093020601495,
"grad_norm": 1.5900787115097046,
"learning_rate": 1e-05,
"loss": 0.8301,
"step": 16770
},
{
"epoch": 0.31919952823907627,
"grad_norm": 1.9719778299331665,
"learning_rate": 1e-05,
"loss": 0.6329,
"step": 16780
},
{
"epoch": 0.319389754418003,
"grad_norm": 3.0982918739318848,
"learning_rate": 1e-05,
"loss": 0.6583,
"step": 16790
},
{
"epoch": 0.31957998059692977,
"grad_norm": 7.582902908325195,
"learning_rate": 1e-05,
"loss": 0.8151,
"step": 16800
},
{
"epoch": 0.3197702067758565,
"grad_norm": 1.3722530603408813,
"learning_rate": 1e-05,
"loss": 0.8194,
"step": 16810
},
{
"epoch": 0.3199604329547832,
"grad_norm": 1.6688120365142822,
"learning_rate": 1e-05,
"loss": 0.7833,
"step": 16820
},
{
"epoch": 0.32015065913371,
"grad_norm": 2.0163159370422363,
"learning_rate": 1e-05,
"loss": 0.655,
"step": 16830
},
{
"epoch": 0.3203408853126367,
"grad_norm": 2.846965789794922,
"learning_rate": 1e-05,
"loss": 0.6243,
"step": 16840
},
{
"epoch": 0.3205311114915635,
"grad_norm": 7.88754940032959,
"learning_rate": 1e-05,
"loss": 0.7556,
"step": 16850
},
{
"epoch": 0.3207213376704902,
"grad_norm": 1.5737555027008057,
"learning_rate": 1e-05,
"loss": 0.8413,
"step": 16860
},
{
"epoch": 0.32091156384941694,
"grad_norm": 1.832712173461914,
"learning_rate": 1e-05,
"loss": 0.827,
"step": 16870
},
{
"epoch": 0.3211017900283437,
"grad_norm": 2.1304585933685303,
"learning_rate": 1e-05,
"loss": 0.649,
"step": 16880
},
{
"epoch": 0.32129201620727044,
"grad_norm": 3.117626905441284,
"learning_rate": 1e-05,
"loss": 0.6262,
"step": 16890
},
{
"epoch": 0.3214822423861972,
"grad_norm": 5.951064586639404,
"learning_rate": 1e-05,
"loss": 0.7351,
"step": 16900
},
{
"epoch": 0.32167246856512394,
"grad_norm": 1.5526176691055298,
"learning_rate": 1e-05,
"loss": 0.8585,
"step": 16910
},
{
"epoch": 0.32186269474405066,
"grad_norm": 1.7932369709014893,
"learning_rate": 1e-05,
"loss": 0.8011,
"step": 16920
},
{
"epoch": 0.32205292092297744,
"grad_norm": 2.9471065998077393,
"learning_rate": 1e-05,
"loss": 0.6956,
"step": 16930
},
{
"epoch": 0.32224314710190416,
"grad_norm": 2.784172773361206,
"learning_rate": 1e-05,
"loss": 0.6576,
"step": 16940
},
{
"epoch": 0.3224333732808309,
"grad_norm": 6.327900409698486,
"learning_rate": 1e-05,
"loss": 0.7173,
"step": 16950
},
{
"epoch": 0.32262359945975766,
"grad_norm": 1.5088452100753784,
"learning_rate": 1e-05,
"loss": 0.8187,
"step": 16960
},
{
"epoch": 0.3228138256386844,
"grad_norm": 1.81622314453125,
"learning_rate": 1e-05,
"loss": 0.7759,
"step": 16970
},
{
"epoch": 0.32300405181761116,
"grad_norm": 2.4694132804870605,
"learning_rate": 1e-05,
"loss": 0.6729,
"step": 16980
},
{
"epoch": 0.3231942779965379,
"grad_norm": 3.0028185844421387,
"learning_rate": 1e-05,
"loss": 0.6408,
"step": 16990
},
{
"epoch": 0.3233845041754646,
"grad_norm": 8.298627853393555,
"learning_rate": 1e-05,
"loss": 0.7335,
"step": 17000
},
{
"epoch": 0.3235747303543914,
"grad_norm": 1.465479850769043,
"learning_rate": 1e-05,
"loss": 0.8383,
"step": 17010
},
{
"epoch": 0.3237649565333181,
"grad_norm": 1.8554552793502808,
"learning_rate": 1e-05,
"loss": 0.8489,
"step": 17020
},
{
"epoch": 0.3239551827122449,
"grad_norm": 2.2836194038391113,
"learning_rate": 1e-05,
"loss": 0.6507,
"step": 17030
},
{
"epoch": 0.3241454088911716,
"grad_norm": 3.168001174926758,
"learning_rate": 1e-05,
"loss": 0.6524,
"step": 17040
},
{
"epoch": 0.3243356350700983,
"grad_norm": 7.013669967651367,
"learning_rate": 1e-05,
"loss": 0.7215,
"step": 17050
},
{
"epoch": 0.3245258612490251,
"grad_norm": 1.4595328569412231,
"learning_rate": 1e-05,
"loss": 0.8243,
"step": 17060
},
{
"epoch": 0.3247160874279518,
"grad_norm": 1.6026345491409302,
"learning_rate": 1e-05,
"loss": 0.8059,
"step": 17070
},
{
"epoch": 0.3249063136068786,
"grad_norm": 2.0784807205200195,
"learning_rate": 1e-05,
"loss": 0.6789,
"step": 17080
},
{
"epoch": 0.3250965397858053,
"grad_norm": 2.44980525970459,
"learning_rate": 1e-05,
"loss": 0.5976,
"step": 17090
},
{
"epoch": 0.32528676596473205,
"grad_norm": 7.785951137542725,
"learning_rate": 1e-05,
"loss": 0.7449,
"step": 17100
},
{
"epoch": 0.3254769921436588,
"grad_norm": 1.5732204914093018,
"learning_rate": 1e-05,
"loss": 0.8532,
"step": 17110
},
{
"epoch": 0.32566721832258555,
"grad_norm": 1.5644340515136719,
"learning_rate": 1e-05,
"loss": 0.8052,
"step": 17120
},
{
"epoch": 0.3258574445015123,
"grad_norm": 2.173517942428589,
"learning_rate": 1e-05,
"loss": 0.6481,
"step": 17130
},
{
"epoch": 0.32604767068043905,
"grad_norm": 2.6543405055999756,
"learning_rate": 1e-05,
"loss": 0.61,
"step": 17140
},
{
"epoch": 0.3262378968593658,
"grad_norm": 6.667696475982666,
"learning_rate": 1e-05,
"loss": 0.7429,
"step": 17150
},
{
"epoch": 0.32642812303829255,
"grad_norm": 1.6218584775924683,
"learning_rate": 1e-05,
"loss": 0.867,
"step": 17160
},
{
"epoch": 0.3266183492172193,
"grad_norm": 1.899166226387024,
"learning_rate": 1e-05,
"loss": 0.8144,
"step": 17170
},
{
"epoch": 0.326808575396146,
"grad_norm": 2.4350476264953613,
"learning_rate": 1e-05,
"loss": 0.6223,
"step": 17180
},
{
"epoch": 0.3269988015750728,
"grad_norm": 3.00462007522583,
"learning_rate": 1e-05,
"loss": 0.6133,
"step": 17190
},
{
"epoch": 0.3271890277539995,
"grad_norm": 7.418691158294678,
"learning_rate": 1e-05,
"loss": 0.7944,
"step": 17200
},
{
"epoch": 0.3273792539329263,
"grad_norm": 1.7549774646759033,
"learning_rate": 1e-05,
"loss": 0.8662,
"step": 17210
},
{
"epoch": 0.327569480111853,
"grad_norm": 1.8458225727081299,
"learning_rate": 1e-05,
"loss": 0.7623,
"step": 17220
},
{
"epoch": 0.3277597062907797,
"grad_norm": 2.241943836212158,
"learning_rate": 1e-05,
"loss": 0.678,
"step": 17230
},
{
"epoch": 0.3279499324697065,
"grad_norm": 3.1956396102905273,
"learning_rate": 1e-05,
"loss": 0.6734,
"step": 17240
},
{
"epoch": 0.3281401586486332,
"grad_norm": 9.091144561767578,
"learning_rate": 1e-05,
"loss": 0.7931,
"step": 17250
},
{
"epoch": 0.32833038482756,
"grad_norm": 1.7532674074172974,
"learning_rate": 1e-05,
"loss": 0.8587,
"step": 17260
},
{
"epoch": 0.3285206110064867,
"grad_norm": 1.8452141284942627,
"learning_rate": 1e-05,
"loss": 0.8043,
"step": 17270
},
{
"epoch": 0.32871083718541344,
"grad_norm": 2.2617251873016357,
"learning_rate": 1e-05,
"loss": 0.666,
"step": 17280
},
{
"epoch": 0.3289010633643402,
"grad_norm": 2.936025619506836,
"learning_rate": 1e-05,
"loss": 0.6232,
"step": 17290
},
{
"epoch": 0.32909128954326694,
"grad_norm": 8.573421478271484,
"learning_rate": 1e-05,
"loss": 0.7573,
"step": 17300
},
{
"epoch": 0.32928151572219366,
"grad_norm": 1.4912211894989014,
"learning_rate": 1e-05,
"loss": 0.8354,
"step": 17310
},
{
"epoch": 0.32947174190112044,
"grad_norm": 1.5418314933776855,
"learning_rate": 1e-05,
"loss": 0.7972,
"step": 17320
},
{
"epoch": 0.32966196808004716,
"grad_norm": 2.2443690299987793,
"learning_rate": 1e-05,
"loss": 0.626,
"step": 17330
},
{
"epoch": 0.32985219425897394,
"grad_norm": 2.4687271118164062,
"learning_rate": 1e-05,
"loss": 0.6368,
"step": 17340
},
{
"epoch": 0.33004242043790066,
"grad_norm": 6.7355217933654785,
"learning_rate": 1e-05,
"loss": 0.7238,
"step": 17350
},
{
"epoch": 0.3302326466168274,
"grad_norm": 1.4340784549713135,
"learning_rate": 1e-05,
"loss": 0.8648,
"step": 17360
},
{
"epoch": 0.33042287279575416,
"grad_norm": 2.2490570545196533,
"learning_rate": 1e-05,
"loss": 0.7709,
"step": 17370
},
{
"epoch": 0.3306130989746809,
"grad_norm": 2.3127803802490234,
"learning_rate": 1e-05,
"loss": 0.6745,
"step": 17380
},
{
"epoch": 0.33080332515360766,
"grad_norm": 3.154917001724243,
"learning_rate": 1e-05,
"loss": 0.664,
"step": 17390
},
{
"epoch": 0.3309935513325344,
"grad_norm": 7.320753574371338,
"learning_rate": 1e-05,
"loss": 0.7335,
"step": 17400
},
{
"epoch": 0.3311837775114611,
"grad_norm": 1.8334399461746216,
"learning_rate": 1e-05,
"loss": 0.8441,
"step": 17410
},
{
"epoch": 0.3313740036903879,
"grad_norm": 1.7856636047363281,
"learning_rate": 1e-05,
"loss": 0.7784,
"step": 17420
},
{
"epoch": 0.3315642298693146,
"grad_norm": 2.225019693374634,
"learning_rate": 1e-05,
"loss": 0.6298,
"step": 17430
},
{
"epoch": 0.3317544560482414,
"grad_norm": 2.6612188816070557,
"learning_rate": 1e-05,
"loss": 0.6167,
"step": 17440
},
{
"epoch": 0.3319446822271681,
"grad_norm": 6.736233234405518,
"learning_rate": 1e-05,
"loss": 0.7448,
"step": 17450
},
{
"epoch": 0.33213490840609483,
"grad_norm": 1.3639291524887085,
"learning_rate": 1e-05,
"loss": 0.8397,
"step": 17460
},
{
"epoch": 0.3323251345850216,
"grad_norm": 1.65357506275177,
"learning_rate": 1e-05,
"loss": 0.8053,
"step": 17470
},
{
"epoch": 0.33251536076394833,
"grad_norm": 2.016624689102173,
"learning_rate": 1e-05,
"loss": 0.6259,
"step": 17480
},
{
"epoch": 0.33270558694287505,
"grad_norm": 2.8766391277313232,
"learning_rate": 1e-05,
"loss": 0.5987,
"step": 17490
},
{
"epoch": 0.33289581312180183,
"grad_norm": 6.8968610763549805,
"learning_rate": 1e-05,
"loss": 0.7521,
"step": 17500
},
{
"epoch": 0.33308603930072855,
"grad_norm": 1.5295982360839844,
"learning_rate": 1e-05,
"loss": 0.8496,
"step": 17510
},
{
"epoch": 0.33327626547965533,
"grad_norm": 1.6879425048828125,
"learning_rate": 1e-05,
"loss": 0.7872,
"step": 17520
},
{
"epoch": 0.33346649165858205,
"grad_norm": 2.554042339324951,
"learning_rate": 1e-05,
"loss": 0.6945,
"step": 17530
},
{
"epoch": 0.3336567178375088,
"grad_norm": 2.786318302154541,
"learning_rate": 1e-05,
"loss": 0.6687,
"step": 17540
},
{
"epoch": 0.33384694401643555,
"grad_norm": 7.835622310638428,
"learning_rate": 1e-05,
"loss": 0.7697,
"step": 17550
},
{
"epoch": 0.3340371701953623,
"grad_norm": 1.4356480836868286,
"learning_rate": 1e-05,
"loss": 0.8591,
"step": 17560
},
{
"epoch": 0.33422739637428905,
"grad_norm": 1.7435493469238281,
"learning_rate": 1e-05,
"loss": 0.8116,
"step": 17570
},
{
"epoch": 0.3344176225532158,
"grad_norm": 2.059699296951294,
"learning_rate": 1e-05,
"loss": 0.6366,
"step": 17580
},
{
"epoch": 0.3346078487321425,
"grad_norm": 2.780322551727295,
"learning_rate": 1e-05,
"loss": 0.641,
"step": 17590
},
{
"epoch": 0.3347980749110693,
"grad_norm": 7.039731979370117,
"learning_rate": 1e-05,
"loss": 0.7666,
"step": 17600
},
{
"epoch": 0.334988301089996,
"grad_norm": 1.7043027877807617,
"learning_rate": 1e-05,
"loss": 0.8552,
"step": 17610
},
{
"epoch": 0.3351785272689228,
"grad_norm": 1.8412545919418335,
"learning_rate": 1e-05,
"loss": 0.8014,
"step": 17620
},
{
"epoch": 0.3353687534478495,
"grad_norm": 2.265721082687378,
"learning_rate": 1e-05,
"loss": 0.6351,
"step": 17630
},
{
"epoch": 0.3355589796267762,
"grad_norm": 2.7243497371673584,
"learning_rate": 1e-05,
"loss": 0.5833,
"step": 17640
},
{
"epoch": 0.335749205805703,
"grad_norm": 6.635868072509766,
"learning_rate": 1e-05,
"loss": 0.7676,
"step": 17650
},
{
"epoch": 0.3359394319846297,
"grad_norm": 1.5984292030334473,
"learning_rate": 1e-05,
"loss": 0.8457,
"step": 17660
},
{
"epoch": 0.33612965816355644,
"grad_norm": 1.657456398010254,
"learning_rate": 1e-05,
"loss": 0.774,
"step": 17670
},
{
"epoch": 0.3363198843424832,
"grad_norm": 1.9562004804611206,
"learning_rate": 1e-05,
"loss": 0.6503,
"step": 17680
},
{
"epoch": 0.33651011052140994,
"grad_norm": 2.654813766479492,
"learning_rate": 1e-05,
"loss": 0.6674,
"step": 17690
},
{
"epoch": 0.3367003367003367,
"grad_norm": 7.267960071563721,
"learning_rate": 1e-05,
"loss": 0.7556,
"step": 17700
},
{
"epoch": 0.33689056287926344,
"grad_norm": 1.45785391330719,
"learning_rate": 1e-05,
"loss": 0.8412,
"step": 17710
},
{
"epoch": 0.33708078905819017,
"grad_norm": 1.9008221626281738,
"learning_rate": 1e-05,
"loss": 0.7839,
"step": 17720
},
{
"epoch": 0.33727101523711694,
"grad_norm": 1.9482413530349731,
"learning_rate": 1e-05,
"loss": 0.642,
"step": 17730
},
{
"epoch": 0.33746124141604367,
"grad_norm": 3.1671218872070312,
"learning_rate": 1e-05,
"loss": 0.6119,
"step": 17740
},
{
"epoch": 0.33765146759497044,
"grad_norm": 7.470427989959717,
"learning_rate": 1e-05,
"loss": 0.7979,
"step": 17750
},
{
"epoch": 0.33784169377389717,
"grad_norm": 1.4068573713302612,
"learning_rate": 1e-05,
"loss": 0.8207,
"step": 17760
},
{
"epoch": 0.3380319199528239,
"grad_norm": 1.6802282333374023,
"learning_rate": 1e-05,
"loss": 0.803,
"step": 17770
},
{
"epoch": 0.33822214613175067,
"grad_norm": 1.889751672744751,
"learning_rate": 1e-05,
"loss": 0.6203,
"step": 17780
},
{
"epoch": 0.3384123723106774,
"grad_norm": 3.1426007747650146,
"learning_rate": 1e-05,
"loss": 0.6153,
"step": 17790
},
{
"epoch": 0.3386025984896041,
"grad_norm": 7.829723834991455,
"learning_rate": 1e-05,
"loss": 0.7566,
"step": 17800
},
{
"epoch": 0.3387928246685309,
"grad_norm": 1.7207051515579224,
"learning_rate": 1e-05,
"loss": 0.8457,
"step": 17810
},
{
"epoch": 0.3389830508474576,
"grad_norm": 1.8551545143127441,
"learning_rate": 1e-05,
"loss": 0.7925,
"step": 17820
},
{
"epoch": 0.3391732770263844,
"grad_norm": 2.306788682937622,
"learning_rate": 1e-05,
"loss": 0.624,
"step": 17830
},
{
"epoch": 0.3393635032053111,
"grad_norm": 2.951352596282959,
"learning_rate": 1e-05,
"loss": 0.6119,
"step": 17840
},
{
"epoch": 0.33955372938423783,
"grad_norm": 8.6023588180542,
"learning_rate": 1e-05,
"loss": 0.7726,
"step": 17850
},
{
"epoch": 0.3397439555631646,
"grad_norm": 1.5706382989883423,
"learning_rate": 1e-05,
"loss": 0.824,
"step": 17860
},
{
"epoch": 0.33993418174209133,
"grad_norm": 1.620394229888916,
"learning_rate": 1e-05,
"loss": 0.7768,
"step": 17870
},
{
"epoch": 0.3401244079210181,
"grad_norm": 2.3519740104675293,
"learning_rate": 1e-05,
"loss": 0.6639,
"step": 17880
},
{
"epoch": 0.34031463409994483,
"grad_norm": 2.8097429275512695,
"learning_rate": 1e-05,
"loss": 0.6416,
"step": 17890
},
{
"epoch": 0.34050486027887156,
"grad_norm": 6.9569478034973145,
"learning_rate": 1e-05,
"loss": 0.7666,
"step": 17900
},
{
"epoch": 0.34069508645779834,
"grad_norm": 1.444406270980835,
"learning_rate": 1e-05,
"loss": 0.8435,
"step": 17910
},
{
"epoch": 0.34088531263672506,
"grad_norm": 1.8301011323928833,
"learning_rate": 1e-05,
"loss": 0.7561,
"step": 17920
},
{
"epoch": 0.34107553881565184,
"grad_norm": 1.9897793531417847,
"learning_rate": 1e-05,
"loss": 0.6525,
"step": 17930
},
{
"epoch": 0.34126576499457856,
"grad_norm": 2.551945209503174,
"learning_rate": 1e-05,
"loss": 0.6284,
"step": 17940
},
{
"epoch": 0.3414559911735053,
"grad_norm": 7.724950313568115,
"learning_rate": 1e-05,
"loss": 0.7336,
"step": 17950
},
{
"epoch": 0.34164621735243206,
"grad_norm": 1.6402499675750732,
"learning_rate": 1e-05,
"loss": 0.8258,
"step": 17960
},
{
"epoch": 0.3418364435313588,
"grad_norm": 1.8207011222839355,
"learning_rate": 1e-05,
"loss": 0.7646,
"step": 17970
},
{
"epoch": 0.3420266697102855,
"grad_norm": 2.191542863845825,
"learning_rate": 1e-05,
"loss": 0.6504,
"step": 17980
},
{
"epoch": 0.3422168958892123,
"grad_norm": 2.716728448867798,
"learning_rate": 1e-05,
"loss": 0.5966,
"step": 17990
},
{
"epoch": 0.342407122068139,
"grad_norm": 7.114306449890137,
"learning_rate": 1e-05,
"loss": 0.7278,
"step": 18000
},
{
"epoch": 0.3425973482470658,
"grad_norm": 1.7861384153366089,
"learning_rate": 1e-05,
"loss": 0.8438,
"step": 18010
},
{
"epoch": 0.3427875744259925,
"grad_norm": 1.8029905557632446,
"learning_rate": 1e-05,
"loss": 0.7977,
"step": 18020
},
{
"epoch": 0.3429778006049192,
"grad_norm": 2.335706949234009,
"learning_rate": 1e-05,
"loss": 0.6065,
"step": 18030
},
{
"epoch": 0.343168026783846,
"grad_norm": 3.215322732925415,
"learning_rate": 1e-05,
"loss": 0.634,
"step": 18040
},
{
"epoch": 0.3433582529627727,
"grad_norm": 7.653074741363525,
"learning_rate": 1e-05,
"loss": 0.8175,
"step": 18050
},
{
"epoch": 0.3435484791416995,
"grad_norm": 1.5260872840881348,
"learning_rate": 1e-05,
"loss": 0.817,
"step": 18060
},
{
"epoch": 0.3437387053206262,
"grad_norm": 1.6873350143432617,
"learning_rate": 1e-05,
"loss": 0.7906,
"step": 18070
},
{
"epoch": 0.34392893149955295,
"grad_norm": 1.9461485147476196,
"learning_rate": 1e-05,
"loss": 0.6236,
"step": 18080
},
{
"epoch": 0.3441191576784797,
"grad_norm": 3.1558187007904053,
"learning_rate": 1e-05,
"loss": 0.6344,
"step": 18090
},
{
"epoch": 0.34430938385740645,
"grad_norm": 7.34866189956665,
"learning_rate": 1e-05,
"loss": 0.7396,
"step": 18100
},
{
"epoch": 0.3444996100363332,
"grad_norm": 2.009432077407837,
"learning_rate": 1e-05,
"loss": 0.8504,
"step": 18110
},
{
"epoch": 0.34468983621525995,
"grad_norm": 1.6450684070587158,
"learning_rate": 1e-05,
"loss": 0.8253,
"step": 18120
},
{
"epoch": 0.34488006239418667,
"grad_norm": 2.346203327178955,
"learning_rate": 1e-05,
"loss": 0.6409,
"step": 18130
},
{
"epoch": 0.34507028857311345,
"grad_norm": 2.981313943862915,
"learning_rate": 1e-05,
"loss": 0.6459,
"step": 18140
},
{
"epoch": 0.34526051475204017,
"grad_norm": 7.191849708557129,
"learning_rate": 1e-05,
"loss": 0.752,
"step": 18150
},
{
"epoch": 0.3454507409309669,
"grad_norm": 1.4715839624404907,
"learning_rate": 1e-05,
"loss": 0.8305,
"step": 18160
},
{
"epoch": 0.34564096710989367,
"grad_norm": 1.598312258720398,
"learning_rate": 1e-05,
"loss": 0.8109,
"step": 18170
},
{
"epoch": 0.3458311932888204,
"grad_norm": 2.6912858486175537,
"learning_rate": 1e-05,
"loss": 0.6626,
"step": 18180
},
{
"epoch": 0.34602141946774717,
"grad_norm": 2.5340702533721924,
"learning_rate": 1e-05,
"loss": 0.6351,
"step": 18190
},
{
"epoch": 0.3462116456466739,
"grad_norm": 7.2905707359313965,
"learning_rate": 1e-05,
"loss": 0.7806,
"step": 18200
},
{
"epoch": 0.3464018718256006,
"grad_norm": 1.9511455297470093,
"learning_rate": 1e-05,
"loss": 0.8678,
"step": 18210
},
{
"epoch": 0.3465920980045274,
"grad_norm": 1.734151840209961,
"learning_rate": 1e-05,
"loss": 0.796,
"step": 18220
},
{
"epoch": 0.3467823241834541,
"grad_norm": 2.0906589031219482,
"learning_rate": 1e-05,
"loss": 0.6804,
"step": 18230
},
{
"epoch": 0.3469725503623809,
"grad_norm": 2.8573203086853027,
"learning_rate": 1e-05,
"loss": 0.666,
"step": 18240
},
{
"epoch": 0.3471627765413076,
"grad_norm": 7.564403057098389,
"learning_rate": 1e-05,
"loss": 0.7729,
"step": 18250
},
{
"epoch": 0.34735300272023434,
"grad_norm": 1.6833574771881104,
"learning_rate": 1e-05,
"loss": 0.8385,
"step": 18260
},
{
"epoch": 0.3475432288991611,
"grad_norm": 1.5758486986160278,
"learning_rate": 1e-05,
"loss": 0.7964,
"step": 18270
},
{
"epoch": 0.34773345507808784,
"grad_norm": 2.214982271194458,
"learning_rate": 1e-05,
"loss": 0.6677,
"step": 18280
},
{
"epoch": 0.3479236812570146,
"grad_norm": 2.9456260204315186,
"learning_rate": 1e-05,
"loss": 0.6186,
"step": 18290
},
{
"epoch": 0.34811390743594134,
"grad_norm": 6.65476655960083,
"learning_rate": 1e-05,
"loss": 0.7814,
"step": 18300
},
{
"epoch": 0.34830413361486806,
"grad_norm": 1.4759432077407837,
"learning_rate": 1e-05,
"loss": 0.8366,
"step": 18310
},
{
"epoch": 0.34849435979379484,
"grad_norm": 1.6910377740859985,
"learning_rate": 1e-05,
"loss": 0.8187,
"step": 18320
},
{
"epoch": 0.34868458597272156,
"grad_norm": 2.7149925231933594,
"learning_rate": 1e-05,
"loss": 0.6448,
"step": 18330
},
{
"epoch": 0.3488748121516483,
"grad_norm": 2.670668363571167,
"learning_rate": 1e-05,
"loss": 0.6382,
"step": 18340
},
{
"epoch": 0.34906503833057506,
"grad_norm": 7.046491622924805,
"learning_rate": 1e-05,
"loss": 0.7513,
"step": 18350
},
{
"epoch": 0.3492552645095018,
"grad_norm": 1.6136091947555542,
"learning_rate": 1e-05,
"loss": 0.839,
"step": 18360
},
{
"epoch": 0.34944549068842856,
"grad_norm": 1.6167227029800415,
"learning_rate": 1e-05,
"loss": 0.7683,
"step": 18370
},
{
"epoch": 0.3496357168673553,
"grad_norm": 1.9725382328033447,
"learning_rate": 1e-05,
"loss": 0.6071,
"step": 18380
},
{
"epoch": 0.349825943046282,
"grad_norm": 2.8352761268615723,
"learning_rate": 1e-05,
"loss": 0.6433,
"step": 18390
},
{
"epoch": 0.3500161692252088,
"grad_norm": 6.600358009338379,
"learning_rate": 1e-05,
"loss": 0.7491,
"step": 18400
},
{
"epoch": 0.3502063954041355,
"grad_norm": 1.540473222732544,
"learning_rate": 1e-05,
"loss": 0.8691,
"step": 18410
},
{
"epoch": 0.3503966215830623,
"grad_norm": 1.4649937152862549,
"learning_rate": 1e-05,
"loss": 0.8094,
"step": 18420
},
{
"epoch": 0.350586847761989,
"grad_norm": 1.875489592552185,
"learning_rate": 1e-05,
"loss": 0.656,
"step": 18430
},
{
"epoch": 0.35077707394091573,
"grad_norm": 2.3585963249206543,
"learning_rate": 1e-05,
"loss": 0.6381,
"step": 18440
},
{
"epoch": 0.3509673001198425,
"grad_norm": 6.578237056732178,
"learning_rate": 1e-05,
"loss": 0.7174,
"step": 18450
},
{
"epoch": 0.35115752629876923,
"grad_norm": 1.289381742477417,
"learning_rate": 1e-05,
"loss": 0.8567,
"step": 18460
},
{
"epoch": 0.351347752477696,
"grad_norm": 1.6858856678009033,
"learning_rate": 1e-05,
"loss": 0.7829,
"step": 18470
},
{
"epoch": 0.35153797865662273,
"grad_norm": 2.21474027633667,
"learning_rate": 1e-05,
"loss": 0.6419,
"step": 18480
},
{
"epoch": 0.35172820483554945,
"grad_norm": 2.6466288566589355,
"learning_rate": 1e-05,
"loss": 0.5833,
"step": 18490
},
{
"epoch": 0.35191843101447623,
"grad_norm": 6.731773376464844,
"learning_rate": 1e-05,
"loss": 0.7315,
"step": 18500
},
{
"epoch": 0.35210865719340295,
"grad_norm": 2.4983060359954834,
"learning_rate": 1e-05,
"loss": 0.8232,
"step": 18510
},
{
"epoch": 0.3522988833723297,
"grad_norm": 1.8547838926315308,
"learning_rate": 1e-05,
"loss": 0.8262,
"step": 18520
},
{
"epoch": 0.35248910955125645,
"grad_norm": 1.991874098777771,
"learning_rate": 1e-05,
"loss": 0.6273,
"step": 18530
},
{
"epoch": 0.3526793357301832,
"grad_norm": 2.498522996902466,
"learning_rate": 1e-05,
"loss": 0.6336,
"step": 18540
},
{
"epoch": 0.35286956190910995,
"grad_norm": 6.589123249053955,
"learning_rate": 1e-05,
"loss": 0.7297,
"step": 18550
},
{
"epoch": 0.3530597880880367,
"grad_norm": 2.085104465484619,
"learning_rate": 1e-05,
"loss": 0.8282,
"step": 18560
},
{
"epoch": 0.3532500142669634,
"grad_norm": 1.6305805444717407,
"learning_rate": 1e-05,
"loss": 0.8024,
"step": 18570
},
{
"epoch": 0.3534402404458902,
"grad_norm": 2.2769906520843506,
"learning_rate": 1e-05,
"loss": 0.6237,
"step": 18580
},
{
"epoch": 0.3536304666248169,
"grad_norm": 2.789093017578125,
"learning_rate": 1e-05,
"loss": 0.6338,
"step": 18590
},
{
"epoch": 0.3538206928037437,
"grad_norm": 8.674882888793945,
"learning_rate": 1e-05,
"loss": 0.7689,
"step": 18600
},
{
"epoch": 0.3540109189826704,
"grad_norm": 1.4302926063537598,
"learning_rate": 1e-05,
"loss": 0.8453,
"step": 18610
},
{
"epoch": 0.3542011451615971,
"grad_norm": 1.5999759435653687,
"learning_rate": 1e-05,
"loss": 0.8172,
"step": 18620
},
{
"epoch": 0.3543913713405239,
"grad_norm": 2.4344184398651123,
"learning_rate": 1e-05,
"loss": 0.6441,
"step": 18630
},
{
"epoch": 0.3545815975194506,
"grad_norm": 3.4894847869873047,
"learning_rate": 1e-05,
"loss": 0.5928,
"step": 18640
},
{
"epoch": 0.3547718236983774,
"grad_norm": 7.913034439086914,
"learning_rate": 1e-05,
"loss": 0.8005,
"step": 18650
},
{
"epoch": 0.3549620498773041,
"grad_norm": 1.5340771675109863,
"learning_rate": 1e-05,
"loss": 0.8158,
"step": 18660
},
{
"epoch": 0.35515227605623084,
"grad_norm": 1.508939266204834,
"learning_rate": 1e-05,
"loss": 0.7691,
"step": 18670
},
{
"epoch": 0.3553425022351576,
"grad_norm": 1.8768864870071411,
"learning_rate": 1e-05,
"loss": 0.597,
"step": 18680
},
{
"epoch": 0.35553272841408434,
"grad_norm": 2.606522798538208,
"learning_rate": 1e-05,
"loss": 0.6274,
"step": 18690
},
{
"epoch": 0.35572295459301106,
"grad_norm": 7.808927536010742,
"learning_rate": 1e-05,
"loss": 0.7968,
"step": 18700
},
{
"epoch": 0.35591318077193784,
"grad_norm": 1.436602234840393,
"learning_rate": 1e-05,
"loss": 0.8436,
"step": 18710
},
{
"epoch": 0.35610340695086456,
"grad_norm": 1.7927616834640503,
"learning_rate": 1e-05,
"loss": 0.8241,
"step": 18720
},
{
"epoch": 0.35629363312979134,
"grad_norm": 2.026174783706665,
"learning_rate": 1e-05,
"loss": 0.6322,
"step": 18730
},
{
"epoch": 0.35648385930871807,
"grad_norm": 2.670032024383545,
"learning_rate": 1e-05,
"loss": 0.6488,
"step": 18740
},
{
"epoch": 0.3566740854876448,
"grad_norm": 6.692541599273682,
"learning_rate": 1e-05,
"loss": 0.7468,
"step": 18750
},
{
"epoch": 0.35686431166657157,
"grad_norm": 1.4814190864562988,
"learning_rate": 1e-05,
"loss": 0.8575,
"step": 18760
},
{
"epoch": 0.3570545378454983,
"grad_norm": 1.734423279762268,
"learning_rate": 1e-05,
"loss": 0.7841,
"step": 18770
},
{
"epoch": 0.35724476402442507,
"grad_norm": 2.280099868774414,
"learning_rate": 1e-05,
"loss": 0.6524,
"step": 18780
},
{
"epoch": 0.3574349902033518,
"grad_norm": 2.9057765007019043,
"learning_rate": 1e-05,
"loss": 0.6502,
"step": 18790
},
{
"epoch": 0.3576252163822785,
"grad_norm": 7.063267707824707,
"learning_rate": 1e-05,
"loss": 0.7481,
"step": 18800
},
{
"epoch": 0.3578154425612053,
"grad_norm": 1.5933854579925537,
"learning_rate": 1e-05,
"loss": 0.8507,
"step": 18810
},
{
"epoch": 0.358005668740132,
"grad_norm": 2.190952777862549,
"learning_rate": 1e-05,
"loss": 0.8135,
"step": 18820
},
{
"epoch": 0.3581958949190588,
"grad_norm": 2.135913133621216,
"learning_rate": 1e-05,
"loss": 0.6231,
"step": 18830
},
{
"epoch": 0.3583861210979855,
"grad_norm": 2.984184741973877,
"learning_rate": 1e-05,
"loss": 0.5974,
"step": 18840
},
{
"epoch": 0.35857634727691223,
"grad_norm": 6.511662483215332,
"learning_rate": 1e-05,
"loss": 0.7262,
"step": 18850
},
{
"epoch": 0.358766573455839,
"grad_norm": 1.5873311758041382,
"learning_rate": 1e-05,
"loss": 0.8736,
"step": 18860
},
{
"epoch": 0.35895679963476573,
"grad_norm": 2.1583216190338135,
"learning_rate": 1e-05,
"loss": 0.7468,
"step": 18870
},
{
"epoch": 0.35914702581369246,
"grad_norm": 2.478818893432617,
"learning_rate": 1e-05,
"loss": 0.6978,
"step": 18880
},
{
"epoch": 0.35933725199261923,
"grad_norm": 2.8479950428009033,
"learning_rate": 1e-05,
"loss": 0.5962,
"step": 18890
},
{
"epoch": 0.35952747817154596,
"grad_norm": 7.277161121368408,
"learning_rate": 1e-05,
"loss": 0.7184,
"step": 18900
},
{
"epoch": 0.35971770435047273,
"grad_norm": 1.4554587602615356,
"learning_rate": 1e-05,
"loss": 0.8253,
"step": 18910
},
{
"epoch": 0.35990793052939946,
"grad_norm": 1.7399671077728271,
"learning_rate": 1e-05,
"loss": 0.7872,
"step": 18920
},
{
"epoch": 0.3600981567083262,
"grad_norm": 2.170468330383301,
"learning_rate": 1e-05,
"loss": 0.6992,
"step": 18930
},
{
"epoch": 0.36028838288725296,
"grad_norm": 2.6035072803497314,
"learning_rate": 1e-05,
"loss": 0.5862,
"step": 18940
},
{
"epoch": 0.3604786090661797,
"grad_norm": 6.511167526245117,
"learning_rate": 1e-05,
"loss": 0.7423,
"step": 18950
},
{
"epoch": 0.36066883524510646,
"grad_norm": 1.4190443754196167,
"learning_rate": 1e-05,
"loss": 0.8146,
"step": 18960
},
{
"epoch": 0.3608590614240332,
"grad_norm": 1.7635581493377686,
"learning_rate": 1e-05,
"loss": 0.8065,
"step": 18970
},
{
"epoch": 0.3610492876029599,
"grad_norm": 2.19046950340271,
"learning_rate": 1e-05,
"loss": 0.6506,
"step": 18980
},
{
"epoch": 0.3612395137818867,
"grad_norm": 2.8667054176330566,
"learning_rate": 1e-05,
"loss": 0.6103,
"step": 18990
},
{
"epoch": 0.3614297399608134,
"grad_norm": 6.743791580200195,
"learning_rate": 1e-05,
"loss": 0.7512,
"step": 19000
},
{
"epoch": 0.3616199661397402,
"grad_norm": 1.7846150398254395,
"learning_rate": 1e-05,
"loss": 0.8234,
"step": 19010
},
{
"epoch": 0.3618101923186669,
"grad_norm": 1.683529257774353,
"learning_rate": 1e-05,
"loss": 0.7792,
"step": 19020
},
{
"epoch": 0.3620004184975936,
"grad_norm": 2.1498847007751465,
"learning_rate": 1e-05,
"loss": 0.6622,
"step": 19030
},
{
"epoch": 0.3621906446765204,
"grad_norm": 2.7236878871917725,
"learning_rate": 1e-05,
"loss": 0.5691,
"step": 19040
},
{
"epoch": 0.3623808708554471,
"grad_norm": 13.231972694396973,
"learning_rate": 1e-05,
"loss": 0.7062,
"step": 19050
},
{
"epoch": 0.36257109703437385,
"grad_norm": 1.6943469047546387,
"learning_rate": 1e-05,
"loss": 0.8358,
"step": 19060
},
{
"epoch": 0.3627613232133006,
"grad_norm": 1.6543833017349243,
"learning_rate": 1e-05,
"loss": 0.7797,
"step": 19070
},
{
"epoch": 0.36295154939222735,
"grad_norm": 1.7396007776260376,
"learning_rate": 1e-05,
"loss": 0.6012,
"step": 19080
},
{
"epoch": 0.3631417755711541,
"grad_norm": 2.5125463008880615,
"learning_rate": 1e-05,
"loss": 0.6335,
"step": 19090
},
{
"epoch": 0.36333200175008085,
"grad_norm": 7.000112056732178,
"learning_rate": 1e-05,
"loss": 0.7512,
"step": 19100
},
{
"epoch": 0.36352222792900757,
"grad_norm": 1.5076160430908203,
"learning_rate": 1e-05,
"loss": 0.8435,
"step": 19110
},
{
"epoch": 0.36371245410793435,
"grad_norm": 1.7014366388320923,
"learning_rate": 1e-05,
"loss": 0.7628,
"step": 19120
},
{
"epoch": 0.36390268028686107,
"grad_norm": 1.8652722835540771,
"learning_rate": 1e-05,
"loss": 0.6337,
"step": 19130
},
{
"epoch": 0.36409290646578785,
"grad_norm": 3.1641438007354736,
"learning_rate": 1e-05,
"loss": 0.6054,
"step": 19140
},
{
"epoch": 0.36428313264471457,
"grad_norm": 7.466098308563232,
"learning_rate": 1e-05,
"loss": 0.7673,
"step": 19150
},
{
"epoch": 0.3644733588236413,
"grad_norm": 1.6421300172805786,
"learning_rate": 1e-05,
"loss": 0.8537,
"step": 19160
},
{
"epoch": 0.36466358500256807,
"grad_norm": 1.5945026874542236,
"learning_rate": 1e-05,
"loss": 0.8042,
"step": 19170
},
{
"epoch": 0.3648538111814948,
"grad_norm": 2.0663678646087646,
"learning_rate": 1e-05,
"loss": 0.62,
"step": 19180
},
{
"epoch": 0.3650440373604215,
"grad_norm": 2.6924874782562256,
"learning_rate": 1e-05,
"loss": 0.5826,
"step": 19190
},
{
"epoch": 0.3652342635393483,
"grad_norm": 8.550240516662598,
"learning_rate": 1e-05,
"loss": 0.7877,
"step": 19200
},
{
"epoch": 0.365424489718275,
"grad_norm": 1.4179600477218628,
"learning_rate": 1e-05,
"loss": 0.8498,
"step": 19210
},
{
"epoch": 0.3656147158972018,
"grad_norm": 1.5007336139678955,
"learning_rate": 1e-05,
"loss": 0.7576,
"step": 19220
},
{
"epoch": 0.3658049420761285,
"grad_norm": 2.0936832427978516,
"learning_rate": 1e-05,
"loss": 0.6697,
"step": 19230
},
{
"epoch": 0.36599516825505524,
"grad_norm": 2.728520154953003,
"learning_rate": 1e-05,
"loss": 0.6098,
"step": 19240
},
{
"epoch": 0.366185394433982,
"grad_norm": 8.544076919555664,
"learning_rate": 1e-05,
"loss": 0.7678,
"step": 19250
},
{
"epoch": 0.36637562061290874,
"grad_norm": 1.9753940105438232,
"learning_rate": 1e-05,
"loss": 0.859,
"step": 19260
},
{
"epoch": 0.3665658467918355,
"grad_norm": 1.9429312944412231,
"learning_rate": 1e-05,
"loss": 0.7853,
"step": 19270
},
{
"epoch": 0.36675607297076224,
"grad_norm": 2.0017690658569336,
"learning_rate": 1e-05,
"loss": 0.5965,
"step": 19280
},
{
"epoch": 0.36694629914968896,
"grad_norm": 2.6853249073028564,
"learning_rate": 1e-05,
"loss": 0.5918,
"step": 19290
},
{
"epoch": 0.36713652532861574,
"grad_norm": 7.348158836364746,
"learning_rate": 1e-05,
"loss": 0.731,
"step": 19300
},
{
"epoch": 0.36732675150754246,
"grad_norm": 1.3817493915557861,
"learning_rate": 1e-05,
"loss": 0.8093,
"step": 19310
},
{
"epoch": 0.36751697768646924,
"grad_norm": 1.6991790533065796,
"learning_rate": 1e-05,
"loss": 0.786,
"step": 19320
},
{
"epoch": 0.36770720386539596,
"grad_norm": 1.9250372648239136,
"learning_rate": 1e-05,
"loss": 0.6371,
"step": 19330
},
{
"epoch": 0.3678974300443227,
"grad_norm": 2.869899034500122,
"learning_rate": 1e-05,
"loss": 0.6255,
"step": 19340
},
{
"epoch": 0.36808765622324946,
"grad_norm": 6.925943851470947,
"learning_rate": 1e-05,
"loss": 0.7344,
"step": 19350
},
{
"epoch": 0.3682778824021762,
"grad_norm": 1.4145034551620483,
"learning_rate": 1e-05,
"loss": 0.8236,
"step": 19360
},
{
"epoch": 0.3684681085811029,
"grad_norm": 1.7110106945037842,
"learning_rate": 1e-05,
"loss": 0.7744,
"step": 19370
},
{
"epoch": 0.3686583347600297,
"grad_norm": 2.2476234436035156,
"learning_rate": 1e-05,
"loss": 0.6793,
"step": 19380
},
{
"epoch": 0.3688485609389564,
"grad_norm": 2.9286696910858154,
"learning_rate": 1e-05,
"loss": 0.6182,
"step": 19390
},
{
"epoch": 0.3690387871178832,
"grad_norm": 6.292760372161865,
"learning_rate": 1e-05,
"loss": 0.7302,
"step": 19400
},
{
"epoch": 0.3692290132968099,
"grad_norm": 1.6311912536621094,
"learning_rate": 1e-05,
"loss": 0.8022,
"step": 19410
},
{
"epoch": 0.3694192394757366,
"grad_norm": 1.536869764328003,
"learning_rate": 1e-05,
"loss": 0.8289,
"step": 19420
},
{
"epoch": 0.3696094656546634,
"grad_norm": 1.8690853118896484,
"learning_rate": 1e-05,
"loss": 0.6347,
"step": 19430
},
{
"epoch": 0.3697996918335901,
"grad_norm": 2.8210017681121826,
"learning_rate": 1e-05,
"loss": 0.6142,
"step": 19440
},
{
"epoch": 0.3699899180125169,
"grad_norm": 6.502079010009766,
"learning_rate": 1e-05,
"loss": 0.7448,
"step": 19450
},
{
"epoch": 0.3701801441914436,
"grad_norm": 1.5034239292144775,
"learning_rate": 1e-05,
"loss": 0.8612,
"step": 19460
},
{
"epoch": 0.37037037037037035,
"grad_norm": 1.7813000679016113,
"learning_rate": 1e-05,
"loss": 0.7968,
"step": 19470
},
{
"epoch": 0.3705605965492971,
"grad_norm": 1.8723803758621216,
"learning_rate": 1e-05,
"loss": 0.6275,
"step": 19480
},
{
"epoch": 0.37075082272822385,
"grad_norm": 2.5647337436676025,
"learning_rate": 1e-05,
"loss": 0.6219,
"step": 19490
},
{
"epoch": 0.37094104890715063,
"grad_norm": 7.042048454284668,
"learning_rate": 1e-05,
"loss": 0.7174,
"step": 19500
},
{
"epoch": 0.37113127508607735,
"grad_norm": 1.6857272386550903,
"learning_rate": 1e-05,
"loss": 0.8212,
"step": 19510
},
{
"epoch": 0.3713215012650041,
"grad_norm": 1.604725956916809,
"learning_rate": 1e-05,
"loss": 0.7496,
"step": 19520
},
{
"epoch": 0.37151172744393085,
"grad_norm": 2.1342358589172363,
"learning_rate": 1e-05,
"loss": 0.6451,
"step": 19530
},
{
"epoch": 0.3717019536228576,
"grad_norm": 2.772266149520874,
"learning_rate": 1e-05,
"loss": 0.642,
"step": 19540
},
{
"epoch": 0.3718921798017843,
"grad_norm": 6.602941989898682,
"learning_rate": 1e-05,
"loss": 0.7469,
"step": 19550
},
{
"epoch": 0.3720824059807111,
"grad_norm": 1.7314958572387695,
"learning_rate": 1e-05,
"loss": 0.8211,
"step": 19560
},
{
"epoch": 0.3722726321596378,
"grad_norm": 1.7661234140396118,
"learning_rate": 1e-05,
"loss": 0.7875,
"step": 19570
},
{
"epoch": 0.3724628583385646,
"grad_norm": 2.228301525115967,
"learning_rate": 1e-05,
"loss": 0.6524,
"step": 19580
},
{
"epoch": 0.3726530845174913,
"grad_norm": 2.7345993518829346,
"learning_rate": 1e-05,
"loss": 0.6167,
"step": 19590
},
{
"epoch": 0.372843310696418,
"grad_norm": 8.211006164550781,
"learning_rate": 1e-05,
"loss": 0.7664,
"step": 19600
},
{
"epoch": 0.3730335368753448,
"grad_norm": 1.6909074783325195,
"learning_rate": 1e-05,
"loss": 0.8241,
"step": 19610
},
{
"epoch": 0.3732237630542715,
"grad_norm": 1.8774691820144653,
"learning_rate": 1e-05,
"loss": 0.7869,
"step": 19620
},
{
"epoch": 0.3734139892331983,
"grad_norm": 1.9746757745742798,
"learning_rate": 1e-05,
"loss": 0.6067,
"step": 19630
},
{
"epoch": 0.373604215412125,
"grad_norm": 3.1322102546691895,
"learning_rate": 1e-05,
"loss": 0.6382,
"step": 19640
},
{
"epoch": 0.37379444159105174,
"grad_norm": 6.851424217224121,
"learning_rate": 1e-05,
"loss": 0.7312,
"step": 19650
},
{
"epoch": 0.3739846677699785,
"grad_norm": 1.4527931213378906,
"learning_rate": 1e-05,
"loss": 0.7868,
"step": 19660
},
{
"epoch": 0.37417489394890524,
"grad_norm": 1.655470609664917,
"learning_rate": 1e-05,
"loss": 0.8111,
"step": 19670
},
{
"epoch": 0.374365120127832,
"grad_norm": 1.8805464506149292,
"learning_rate": 1e-05,
"loss": 0.6089,
"step": 19680
},
{
"epoch": 0.37455534630675874,
"grad_norm": 2.7048580646514893,
"learning_rate": 1e-05,
"loss": 0.6472,
"step": 19690
},
{
"epoch": 0.37474557248568546,
"grad_norm": 6.733554840087891,
"learning_rate": 1e-05,
"loss": 0.7364,
"step": 19700
},
{
"epoch": 0.37493579866461224,
"grad_norm": 1.5801007747650146,
"learning_rate": 1e-05,
"loss": 0.85,
"step": 19710
},
{
"epoch": 0.37512602484353896,
"grad_norm": 1.6578195095062256,
"learning_rate": 1e-05,
"loss": 0.8118,
"step": 19720
},
{
"epoch": 0.3753162510224657,
"grad_norm": 2.4759578704833984,
"learning_rate": 1e-05,
"loss": 0.6153,
"step": 19730
},
{
"epoch": 0.37550647720139246,
"grad_norm": 2.9755969047546387,
"learning_rate": 1e-05,
"loss": 0.6615,
"step": 19740
},
{
"epoch": 0.3756967033803192,
"grad_norm": 8.543745040893555,
"learning_rate": 1e-05,
"loss": 0.7685,
"step": 19750
},
{
"epoch": 0.37588692955924596,
"grad_norm": 1.5876930952072144,
"learning_rate": 1e-05,
"loss": 0.8431,
"step": 19760
},
{
"epoch": 0.3760771557381727,
"grad_norm": 1.731096863746643,
"learning_rate": 1e-05,
"loss": 0.8229,
"step": 19770
},
{
"epoch": 0.3762673819170994,
"grad_norm": 2.745640277862549,
"learning_rate": 1e-05,
"loss": 0.6533,
"step": 19780
},
{
"epoch": 0.3764576080960262,
"grad_norm": 2.834723711013794,
"learning_rate": 1e-05,
"loss": 0.6153,
"step": 19790
},
{
"epoch": 0.3766478342749529,
"grad_norm": 7.093329429626465,
"learning_rate": 1e-05,
"loss": 0.7299,
"step": 19800
},
{
"epoch": 0.3768380604538797,
"grad_norm": 1.4555355310440063,
"learning_rate": 1e-05,
"loss": 0.8549,
"step": 19810
},
{
"epoch": 0.3770282866328064,
"grad_norm": 1.7423663139343262,
"learning_rate": 1e-05,
"loss": 0.7857,
"step": 19820
},
{
"epoch": 0.37721851281173313,
"grad_norm": 2.084831953048706,
"learning_rate": 1e-05,
"loss": 0.6123,
"step": 19830
},
{
"epoch": 0.3774087389906599,
"grad_norm": 2.4425032138824463,
"learning_rate": 1e-05,
"loss": 0.5895,
"step": 19840
},
{
"epoch": 0.37759896516958663,
"grad_norm": 8.438631057739258,
"learning_rate": 1e-05,
"loss": 0.7451,
"step": 19850
},
{
"epoch": 0.3777891913485134,
"grad_norm": 1.5773450136184692,
"learning_rate": 1e-05,
"loss": 0.8345,
"step": 19860
},
{
"epoch": 0.37797941752744013,
"grad_norm": 1.9074832201004028,
"learning_rate": 1e-05,
"loss": 0.7903,
"step": 19870
},
{
"epoch": 0.37816964370636685,
"grad_norm": 1.9741617441177368,
"learning_rate": 1e-05,
"loss": 0.6277,
"step": 19880
},
{
"epoch": 0.37835986988529363,
"grad_norm": 2.625777244567871,
"learning_rate": 1e-05,
"loss": 0.641,
"step": 19890
},
{
"epoch": 0.37855009606422035,
"grad_norm": 6.641512393951416,
"learning_rate": 1e-05,
"loss": 0.7567,
"step": 19900
},
{
"epoch": 0.3787403222431471,
"grad_norm": 1.3250898122787476,
"learning_rate": 1e-05,
"loss": 0.7931,
"step": 19910
},
{
"epoch": 0.37893054842207385,
"grad_norm": 1.757511019706726,
"learning_rate": 1e-05,
"loss": 0.8135,
"step": 19920
},
{
"epoch": 0.3791207746010006,
"grad_norm": 2.54929256439209,
"learning_rate": 1e-05,
"loss": 0.6755,
"step": 19930
},
{
"epoch": 0.37931100077992735,
"grad_norm": 2.462949275970459,
"learning_rate": 1e-05,
"loss": 0.6095,
"step": 19940
},
{
"epoch": 0.3795012269588541,
"grad_norm": 8.830132484436035,
"learning_rate": 1e-05,
"loss": 0.7994,
"step": 19950
},
{
"epoch": 0.3796914531377808,
"grad_norm": 1.8836718797683716,
"learning_rate": 1e-05,
"loss": 0.8366,
"step": 19960
},
{
"epoch": 0.3798816793167076,
"grad_norm": 1.6502013206481934,
"learning_rate": 1e-05,
"loss": 0.7839,
"step": 19970
},
{
"epoch": 0.3800719054956343,
"grad_norm": 2.0403785705566406,
"learning_rate": 1e-05,
"loss": 0.6186,
"step": 19980
},
{
"epoch": 0.3802621316745611,
"grad_norm": 2.6063833236694336,
"learning_rate": 1e-05,
"loss": 0.627,
"step": 19990
},
{
"epoch": 0.3804523578534878,
"grad_norm": 7.821043491363525,
"learning_rate": 1e-05,
"loss": 0.7485,
"step": 20000
},
{
"epoch": 0.3806425840324145,
"grad_norm": 1.9438045024871826,
"learning_rate": 1e-05,
"loss": 0.8408,
"step": 20010
},
{
"epoch": 0.3808328102113413,
"grad_norm": 1.8496999740600586,
"learning_rate": 1e-05,
"loss": 0.7863,
"step": 20020
},
{
"epoch": 0.381023036390268,
"grad_norm": 1.834145426750183,
"learning_rate": 1e-05,
"loss": 0.6387,
"step": 20030
},
{
"epoch": 0.3812132625691948,
"grad_norm": 2.7673659324645996,
"learning_rate": 1e-05,
"loss": 0.6072,
"step": 20040
},
{
"epoch": 0.3814034887481215,
"grad_norm": 6.970427513122559,
"learning_rate": 1e-05,
"loss": 0.7161,
"step": 20050
},
{
"epoch": 0.38159371492704824,
"grad_norm": 1.4438550472259521,
"learning_rate": 1e-05,
"loss": 0.8815,
"step": 20060
},
{
"epoch": 0.381783941105975,
"grad_norm": 1.6741695404052734,
"learning_rate": 1e-05,
"loss": 0.7965,
"step": 20070
},
{
"epoch": 0.38197416728490174,
"grad_norm": 2.322519302368164,
"learning_rate": 1e-05,
"loss": 0.6302,
"step": 20080
},
{
"epoch": 0.38216439346382847,
"grad_norm": 2.5832953453063965,
"learning_rate": 1e-05,
"loss": 0.6084,
"step": 20090
},
{
"epoch": 0.38235461964275524,
"grad_norm": 6.573698997497559,
"learning_rate": 1e-05,
"loss": 0.7487,
"step": 20100
},
{
"epoch": 0.38254484582168197,
"grad_norm": 1.4620909690856934,
"learning_rate": 1e-05,
"loss": 0.8293,
"step": 20110
},
{
"epoch": 0.38273507200060874,
"grad_norm": 1.6676819324493408,
"learning_rate": 1e-05,
"loss": 0.7982,
"step": 20120
},
{
"epoch": 0.38292529817953547,
"grad_norm": 2.0680060386657715,
"learning_rate": 1e-05,
"loss": 0.6771,
"step": 20130
},
{
"epoch": 0.3831155243584622,
"grad_norm": 2.7260758876800537,
"learning_rate": 1e-05,
"loss": 0.6238,
"step": 20140
},
{
"epoch": 0.38330575053738897,
"grad_norm": 6.992647171020508,
"learning_rate": 1e-05,
"loss": 0.7053,
"step": 20150
},
{
"epoch": 0.3834959767163157,
"grad_norm": 1.6186981201171875,
"learning_rate": 1e-05,
"loss": 0.8139,
"step": 20160
},
{
"epoch": 0.38368620289524247,
"grad_norm": 1.8367278575897217,
"learning_rate": 1e-05,
"loss": 0.785,
"step": 20170
},
{
"epoch": 0.3838764290741692,
"grad_norm": 2.121630907058716,
"learning_rate": 1e-05,
"loss": 0.6358,
"step": 20180
},
{
"epoch": 0.3840666552530959,
"grad_norm": 2.8505945205688477,
"learning_rate": 1e-05,
"loss": 0.6366,
"step": 20190
},
{
"epoch": 0.3842568814320227,
"grad_norm": 7.4002909660339355,
"learning_rate": 1e-05,
"loss": 0.6942,
"step": 20200
},
{
"epoch": 0.3844471076109494,
"grad_norm": 1.4614757299423218,
"learning_rate": 1e-05,
"loss": 0.8546,
"step": 20210
},
{
"epoch": 0.3846373337898762,
"grad_norm": 1.5963836908340454,
"learning_rate": 1e-05,
"loss": 0.7889,
"step": 20220
},
{
"epoch": 0.3848275599688029,
"grad_norm": 2.2764158248901367,
"learning_rate": 1e-05,
"loss": 0.6195,
"step": 20230
},
{
"epoch": 0.38501778614772963,
"grad_norm": 2.745134115219116,
"learning_rate": 1e-05,
"loss": 0.5979,
"step": 20240
},
{
"epoch": 0.3852080123266564,
"grad_norm": 6.876772403717041,
"learning_rate": 1e-05,
"loss": 0.7448,
"step": 20250
},
{
"epoch": 0.38539823850558314,
"grad_norm": 1.6288174390792847,
"learning_rate": 1e-05,
"loss": 0.8442,
"step": 20260
},
{
"epoch": 0.38558846468450986,
"grad_norm": 1.68385910987854,
"learning_rate": 1e-05,
"loss": 0.7359,
"step": 20270
},
{
"epoch": 0.38577869086343664,
"grad_norm": 2.1139776706695557,
"learning_rate": 1e-05,
"loss": 0.6234,
"step": 20280
},
{
"epoch": 0.38596891704236336,
"grad_norm": 3.015326976776123,
"learning_rate": 1e-05,
"loss": 0.5806,
"step": 20290
},
{
"epoch": 0.38615914322129014,
"grad_norm": 7.759620666503906,
"learning_rate": 1e-05,
"loss": 0.7443,
"step": 20300
},
{
"epoch": 0.38634936940021686,
"grad_norm": 1.4787245988845825,
"learning_rate": 1e-05,
"loss": 0.8317,
"step": 20310
},
{
"epoch": 0.3865395955791436,
"grad_norm": 1.6797029972076416,
"learning_rate": 1e-05,
"loss": 0.7768,
"step": 20320
},
{
"epoch": 0.38672982175807036,
"grad_norm": 2.1213343143463135,
"learning_rate": 1e-05,
"loss": 0.6241,
"step": 20330
},
{
"epoch": 0.3869200479369971,
"grad_norm": 2.2526659965515137,
"learning_rate": 1e-05,
"loss": 0.6125,
"step": 20340
},
{
"epoch": 0.38711027411592386,
"grad_norm": 6.841493129730225,
"learning_rate": 1e-05,
"loss": 0.7394,
"step": 20350
},
{
"epoch": 0.3873005002948506,
"grad_norm": 1.574538230895996,
"learning_rate": 1e-05,
"loss": 0.8488,
"step": 20360
},
{
"epoch": 0.3874907264737773,
"grad_norm": 1.864216685295105,
"learning_rate": 1e-05,
"loss": 0.7657,
"step": 20370
},
{
"epoch": 0.3876809526527041,
"grad_norm": 1.9863324165344238,
"learning_rate": 1e-05,
"loss": 0.6612,
"step": 20380
},
{
"epoch": 0.3878711788316308,
"grad_norm": 2.6953065395355225,
"learning_rate": 1e-05,
"loss": 0.6065,
"step": 20390
},
{
"epoch": 0.3880614050105575,
"grad_norm": 7.9635820388793945,
"learning_rate": 1e-05,
"loss": 0.7655,
"step": 20400
},
{
"epoch": 0.3882516311894843,
"grad_norm": 1.21963632106781,
"learning_rate": 1e-05,
"loss": 0.8263,
"step": 20410
},
{
"epoch": 0.388441857368411,
"grad_norm": 1.5361708402633667,
"learning_rate": 1e-05,
"loss": 0.7769,
"step": 20420
},
{
"epoch": 0.3886320835473378,
"grad_norm": 2.0140774250030518,
"learning_rate": 1e-05,
"loss": 0.6025,
"step": 20430
},
{
"epoch": 0.3888223097262645,
"grad_norm": 2.5272417068481445,
"learning_rate": 1e-05,
"loss": 0.6088,
"step": 20440
},
{
"epoch": 0.38901253590519125,
"grad_norm": 7.379544258117676,
"learning_rate": 1e-05,
"loss": 0.7224,
"step": 20450
},
{
"epoch": 0.389202762084118,
"grad_norm": 1.4019469022750854,
"learning_rate": 1e-05,
"loss": 0.8466,
"step": 20460
},
{
"epoch": 0.38939298826304475,
"grad_norm": 2.4000778198242188,
"learning_rate": 1e-05,
"loss": 0.7969,
"step": 20470
},
{
"epoch": 0.3895832144419715,
"grad_norm": 1.9260022640228271,
"learning_rate": 1e-05,
"loss": 0.6063,
"step": 20480
},
{
"epoch": 0.38977344062089825,
"grad_norm": 2.8490819931030273,
"learning_rate": 1e-05,
"loss": 0.6567,
"step": 20490
},
{
"epoch": 0.38996366679982497,
"grad_norm": 7.016058921813965,
"learning_rate": 1e-05,
"loss": 0.7307,
"step": 20500
},
{
"epoch": 0.39015389297875175,
"grad_norm": 1.4212769269943237,
"learning_rate": 1e-05,
"loss": 0.8433,
"step": 20510
},
{
"epoch": 0.39034411915767847,
"grad_norm": 1.5372602939605713,
"learning_rate": 1e-05,
"loss": 0.7657,
"step": 20520
},
{
"epoch": 0.39053434533660525,
"grad_norm": 2.183230400085449,
"learning_rate": 1e-05,
"loss": 0.6454,
"step": 20530
},
{
"epoch": 0.39072457151553197,
"grad_norm": 2.927985906600952,
"learning_rate": 1e-05,
"loss": 0.6227,
"step": 20540
},
{
"epoch": 0.3909147976944587,
"grad_norm": 7.6321940422058105,
"learning_rate": 1e-05,
"loss": 0.7332,
"step": 20550
},
{
"epoch": 0.39110502387338547,
"grad_norm": 1.5500627756118774,
"learning_rate": 1e-05,
"loss": 0.8212,
"step": 20560
},
{
"epoch": 0.3912952500523122,
"grad_norm": 1.8475933074951172,
"learning_rate": 1e-05,
"loss": 0.7934,
"step": 20570
},
{
"epoch": 0.3914854762312389,
"grad_norm": 2.045667886734009,
"learning_rate": 1e-05,
"loss": 0.6056,
"step": 20580
},
{
"epoch": 0.3916757024101657,
"grad_norm": 3.0658962726593018,
"learning_rate": 1e-05,
"loss": 0.6416,
"step": 20590
},
{
"epoch": 0.3918659285890924,
"grad_norm": 6.326318264007568,
"learning_rate": 1e-05,
"loss": 0.7148,
"step": 20600
},
{
"epoch": 0.3920561547680192,
"grad_norm": 1.3966017961502075,
"learning_rate": 1e-05,
"loss": 0.8129,
"step": 20610
},
{
"epoch": 0.3922463809469459,
"grad_norm": 1.6497981548309326,
"learning_rate": 1e-05,
"loss": 0.7952,
"step": 20620
},
{
"epoch": 0.39243660712587264,
"grad_norm": 2.156029462814331,
"learning_rate": 1e-05,
"loss": 0.6453,
"step": 20630
},
{
"epoch": 0.3926268333047994,
"grad_norm": 2.9816219806671143,
"learning_rate": 1e-05,
"loss": 0.5876,
"step": 20640
},
{
"epoch": 0.39281705948372614,
"grad_norm": 7.04098653793335,
"learning_rate": 1e-05,
"loss": 0.7425,
"step": 20650
},
{
"epoch": 0.3930072856626529,
"grad_norm": 1.7080111503601074,
"learning_rate": 1e-05,
"loss": 0.8459,
"step": 20660
},
{
"epoch": 0.39319751184157964,
"grad_norm": 1.966111183166504,
"learning_rate": 1e-05,
"loss": 0.7818,
"step": 20670
},
{
"epoch": 0.39338773802050636,
"grad_norm": 1.9661368131637573,
"learning_rate": 1e-05,
"loss": 0.6521,
"step": 20680
},
{
"epoch": 0.39357796419943314,
"grad_norm": 3.0567848682403564,
"learning_rate": 1e-05,
"loss": 0.6078,
"step": 20690
},
{
"epoch": 0.39376819037835986,
"grad_norm": 7.279221534729004,
"learning_rate": 1e-05,
"loss": 0.7404,
"step": 20700
},
{
"epoch": 0.39395841655728664,
"grad_norm": 1.4563050270080566,
"learning_rate": 1e-05,
"loss": 0.8182,
"step": 20710
},
{
"epoch": 0.39414864273621336,
"grad_norm": 1.635894536972046,
"learning_rate": 1e-05,
"loss": 0.7875,
"step": 20720
},
{
"epoch": 0.3943388689151401,
"grad_norm": 2.322808265686035,
"learning_rate": 1e-05,
"loss": 0.6445,
"step": 20730
},
{
"epoch": 0.39452909509406686,
"grad_norm": 2.761725902557373,
"learning_rate": 1e-05,
"loss": 0.6101,
"step": 20740
},
{
"epoch": 0.3947193212729936,
"grad_norm": 7.4666314125061035,
"learning_rate": 1e-05,
"loss": 0.7385,
"step": 20750
},
{
"epoch": 0.3949095474519203,
"grad_norm": 1.3985857963562012,
"learning_rate": 1e-05,
"loss": 0.832,
"step": 20760
},
{
"epoch": 0.3950997736308471,
"grad_norm": 1.5213103294372559,
"learning_rate": 1e-05,
"loss": 0.7868,
"step": 20770
},
{
"epoch": 0.3952899998097738,
"grad_norm": 2.022300958633423,
"learning_rate": 1e-05,
"loss": 0.6558,
"step": 20780
},
{
"epoch": 0.3954802259887006,
"grad_norm": 3.0033295154571533,
"learning_rate": 1e-05,
"loss": 0.6352,
"step": 20790
},
{
"epoch": 0.3956704521676273,
"grad_norm": 6.658642768859863,
"learning_rate": 1e-05,
"loss": 0.7475,
"step": 20800
},
{
"epoch": 0.39586067834655403,
"grad_norm": 1.4247969388961792,
"learning_rate": 1e-05,
"loss": 0.8338,
"step": 20810
},
{
"epoch": 0.3960509045254808,
"grad_norm": 1.5818713903427124,
"learning_rate": 1e-05,
"loss": 0.7864,
"step": 20820
},
{
"epoch": 0.39624113070440753,
"grad_norm": 2.0734338760375977,
"learning_rate": 1e-05,
"loss": 0.6405,
"step": 20830
},
{
"epoch": 0.3964313568833343,
"grad_norm": 2.8314590454101562,
"learning_rate": 1e-05,
"loss": 0.5961,
"step": 20840
},
{
"epoch": 0.39662158306226103,
"grad_norm": 6.884099960327148,
"learning_rate": 1e-05,
"loss": 0.7903,
"step": 20850
},
{
"epoch": 0.39681180924118775,
"grad_norm": 1.8323087692260742,
"learning_rate": 1e-05,
"loss": 0.8212,
"step": 20860
},
{
"epoch": 0.39700203542011453,
"grad_norm": 1.6293091773986816,
"learning_rate": 1e-05,
"loss": 0.7497,
"step": 20870
},
{
"epoch": 0.39719226159904125,
"grad_norm": 2.019087791442871,
"learning_rate": 1e-05,
"loss": 0.6258,
"step": 20880
},
{
"epoch": 0.39738248777796803,
"grad_norm": 2.7192296981811523,
"learning_rate": 1e-05,
"loss": 0.6139,
"step": 20890
},
{
"epoch": 0.39757271395689475,
"grad_norm": 7.233087062835693,
"learning_rate": 1e-05,
"loss": 0.71,
"step": 20900
},
{
"epoch": 0.3977629401358215,
"grad_norm": 1.397848129272461,
"learning_rate": 1e-05,
"loss": 0.8358,
"step": 20910
},
{
"epoch": 0.39795316631474825,
"grad_norm": 1.5518430471420288,
"learning_rate": 1e-05,
"loss": 0.7982,
"step": 20920
},
{
"epoch": 0.398143392493675,
"grad_norm": 2.403047561645508,
"learning_rate": 1e-05,
"loss": 0.6358,
"step": 20930
},
{
"epoch": 0.3983336186726017,
"grad_norm": 2.6010377407073975,
"learning_rate": 1e-05,
"loss": 0.6369,
"step": 20940
},
{
"epoch": 0.3985238448515285,
"grad_norm": 6.847127914428711,
"learning_rate": 1e-05,
"loss": 0.7103,
"step": 20950
},
{
"epoch": 0.3987140710304552,
"grad_norm": 1.4623287916183472,
"learning_rate": 1e-05,
"loss": 0.8129,
"step": 20960
},
{
"epoch": 0.398904297209382,
"grad_norm": 1.565334677696228,
"learning_rate": 1e-05,
"loss": 0.7362,
"step": 20970
},
{
"epoch": 0.3990945233883087,
"grad_norm": 2.1012327671051025,
"learning_rate": 1e-05,
"loss": 0.6055,
"step": 20980
},
{
"epoch": 0.3992847495672354,
"grad_norm": 2.788709878921509,
"learning_rate": 1e-05,
"loss": 0.6357,
"step": 20990
},
{
"epoch": 0.3994749757461622,
"grad_norm": 6.261979103088379,
"learning_rate": 1e-05,
"loss": 0.751,
"step": 21000
},
{
"epoch": 0.3996652019250889,
"grad_norm": 1.5507466793060303,
"learning_rate": 1e-05,
"loss": 0.8229,
"step": 21010
},
{
"epoch": 0.3998554281040157,
"grad_norm": 1.5293151140213013,
"learning_rate": 1e-05,
"loss": 0.7642,
"step": 21020
},
{
"epoch": 0.4000456542829424,
"grad_norm": 2.0063936710357666,
"learning_rate": 1e-05,
"loss": 0.6279,
"step": 21030
},
{
"epoch": 0.40023588046186914,
"grad_norm": 3.0445287227630615,
"learning_rate": 1e-05,
"loss": 0.6085,
"step": 21040
},
{
"epoch": 0.4004261066407959,
"grad_norm": 7.4967217445373535,
"learning_rate": 1e-05,
"loss": 0.7403,
"step": 21050
},
{
"epoch": 0.40061633281972264,
"grad_norm": 1.3526582717895508,
"learning_rate": 1e-05,
"loss": 0.8366,
"step": 21060
},
{
"epoch": 0.4008065589986494,
"grad_norm": 1.5526952743530273,
"learning_rate": 1e-05,
"loss": 0.7759,
"step": 21070
},
{
"epoch": 0.40099678517757614,
"grad_norm": 2.1743087768554688,
"learning_rate": 1e-05,
"loss": 0.6099,
"step": 21080
},
{
"epoch": 0.40118701135650287,
"grad_norm": 2.901949405670166,
"learning_rate": 1e-05,
"loss": 0.6304,
"step": 21090
},
{
"epoch": 0.40137723753542964,
"grad_norm": 5.788262367248535,
"learning_rate": 1e-05,
"loss": 0.741,
"step": 21100
},
{
"epoch": 0.40156746371435637,
"grad_norm": 1.5440911054611206,
"learning_rate": 1e-05,
"loss": 0.8252,
"step": 21110
},
{
"epoch": 0.4017576898932831,
"grad_norm": 1.737626552581787,
"learning_rate": 1e-05,
"loss": 0.7454,
"step": 21120
},
{
"epoch": 0.40194791607220987,
"grad_norm": 1.975041151046753,
"learning_rate": 1e-05,
"loss": 0.6195,
"step": 21130
},
{
"epoch": 0.4021381422511366,
"grad_norm": 2.6315276622772217,
"learning_rate": 1e-05,
"loss": 0.6058,
"step": 21140
},
{
"epoch": 0.40232836843006337,
"grad_norm": 6.199795722961426,
"learning_rate": 1e-05,
"loss": 0.6957,
"step": 21150
},
{
"epoch": 0.4025185946089901,
"grad_norm": 1.4145256280899048,
"learning_rate": 1e-05,
"loss": 0.8179,
"step": 21160
},
{
"epoch": 0.4027088207879168,
"grad_norm": 1.6122918128967285,
"learning_rate": 1e-05,
"loss": 0.8137,
"step": 21170
},
{
"epoch": 0.4028990469668436,
"grad_norm": 2.0701076984405518,
"learning_rate": 1e-05,
"loss": 0.6342,
"step": 21180
},
{
"epoch": 0.4030892731457703,
"grad_norm": 2.8882083892822266,
"learning_rate": 1e-05,
"loss": 0.5969,
"step": 21190
},
{
"epoch": 0.4032794993246971,
"grad_norm": 5.7476019859313965,
"learning_rate": 1e-05,
"loss": 0.7289,
"step": 21200
},
{
"epoch": 0.4034697255036238,
"grad_norm": 1.9890129566192627,
"learning_rate": 1e-05,
"loss": 0.8212,
"step": 21210
},
{
"epoch": 0.40365995168255053,
"grad_norm": 1.7185076475143433,
"learning_rate": 1e-05,
"loss": 0.778,
"step": 21220
},
{
"epoch": 0.4038501778614773,
"grad_norm": 2.169844388961792,
"learning_rate": 1e-05,
"loss": 0.61,
"step": 21230
},
{
"epoch": 0.40404040404040403,
"grad_norm": 2.6390838623046875,
"learning_rate": 1e-05,
"loss": 0.6287,
"step": 21240
},
{
"epoch": 0.4042306302193308,
"grad_norm": 8.314827919006348,
"learning_rate": 1e-05,
"loss": 0.7842,
"step": 21250
},
{
"epoch": 0.40442085639825753,
"grad_norm": 1.4161739349365234,
"learning_rate": 1e-05,
"loss": 0.8306,
"step": 21260
},
{
"epoch": 0.40461108257718426,
"grad_norm": 1.5647960901260376,
"learning_rate": 1e-05,
"loss": 0.7668,
"step": 21270
},
{
"epoch": 0.40480130875611103,
"grad_norm": 2.1744017601013184,
"learning_rate": 1e-05,
"loss": 0.6495,
"step": 21280
},
{
"epoch": 0.40499153493503776,
"grad_norm": 3.1150834560394287,
"learning_rate": 1e-05,
"loss": 0.6201,
"step": 21290
},
{
"epoch": 0.4051817611139645,
"grad_norm": 7.626542568206787,
"learning_rate": 1e-05,
"loss": 0.7756,
"step": 21300
},
{
"epoch": 0.40537198729289126,
"grad_norm": 1.438798189163208,
"learning_rate": 1e-05,
"loss": 0.8819,
"step": 21310
},
{
"epoch": 0.405562213471818,
"grad_norm": 1.7915786504745483,
"learning_rate": 1e-05,
"loss": 0.7255,
"step": 21320
},
{
"epoch": 0.40575243965074476,
"grad_norm": 1.9837087392807007,
"learning_rate": 1e-05,
"loss": 0.5977,
"step": 21330
},
{
"epoch": 0.4059426658296715,
"grad_norm": 2.7124886512756348,
"learning_rate": 1e-05,
"loss": 0.6102,
"step": 21340
},
{
"epoch": 0.4061328920085982,
"grad_norm": 6.7429022789001465,
"learning_rate": 1e-05,
"loss": 0.7206,
"step": 21350
},
{
"epoch": 0.406323118187525,
"grad_norm": 1.4629786014556885,
"learning_rate": 1e-05,
"loss": 0.8342,
"step": 21360
},
{
"epoch": 0.4065133443664517,
"grad_norm": 1.5456796884536743,
"learning_rate": 1e-05,
"loss": 0.7672,
"step": 21370
},
{
"epoch": 0.4067035705453785,
"grad_norm": 1.902247428894043,
"learning_rate": 1e-05,
"loss": 0.6286,
"step": 21380
},
{
"epoch": 0.4068937967243052,
"grad_norm": 2.741138219833374,
"learning_rate": 1e-05,
"loss": 0.6276,
"step": 21390
},
{
"epoch": 0.4070840229032319,
"grad_norm": 5.38442325592041,
"learning_rate": 1e-05,
"loss": 0.7012,
"step": 21400
},
{
"epoch": 0.4072742490821587,
"grad_norm": 1.665209412574768,
"learning_rate": 1e-05,
"loss": 0.8151,
"step": 21410
},
{
"epoch": 0.4074644752610854,
"grad_norm": 1.6590564250946045,
"learning_rate": 1e-05,
"loss": 0.7642,
"step": 21420
},
{
"epoch": 0.4076547014400122,
"grad_norm": 1.9104335308074951,
"learning_rate": 1e-05,
"loss": 0.6063,
"step": 21430
},
{
"epoch": 0.4078449276189389,
"grad_norm": 2.5906612873077393,
"learning_rate": 1e-05,
"loss": 0.6054,
"step": 21440
},
{
"epoch": 0.40803515379786565,
"grad_norm": 6.86151647567749,
"learning_rate": 1e-05,
"loss": 0.7324,
"step": 21450
},
{
"epoch": 0.4082253799767924,
"grad_norm": 1.7165383100509644,
"learning_rate": 1e-05,
"loss": 0.8376,
"step": 21460
},
{
"epoch": 0.40841560615571915,
"grad_norm": 1.6326996088027954,
"learning_rate": 1e-05,
"loss": 0.7714,
"step": 21470
},
{
"epoch": 0.40860583233464587,
"grad_norm": 2.2232749462127686,
"learning_rate": 1e-05,
"loss": 0.6149,
"step": 21480
},
{
"epoch": 0.40879605851357265,
"grad_norm": 2.8421921730041504,
"learning_rate": 1e-05,
"loss": 0.6248,
"step": 21490
},
{
"epoch": 0.40898628469249937,
"grad_norm": 6.839955806732178,
"learning_rate": 1e-05,
"loss": 0.7828,
"step": 21500
},
{
"epoch": 0.40917651087142615,
"grad_norm": 1.5385832786560059,
"learning_rate": 1e-05,
"loss": 0.8298,
"step": 21510
},
{
"epoch": 0.40936673705035287,
"grad_norm": 1.499190092086792,
"learning_rate": 1e-05,
"loss": 0.7664,
"step": 21520
},
{
"epoch": 0.4095569632292796,
"grad_norm": 1.9949061870574951,
"learning_rate": 1e-05,
"loss": 0.6562,
"step": 21530
},
{
"epoch": 0.40974718940820637,
"grad_norm": 3.3526968955993652,
"learning_rate": 1e-05,
"loss": 0.5987,
"step": 21540
},
{
"epoch": 0.4099374155871331,
"grad_norm": 6.7234601974487305,
"learning_rate": 1e-05,
"loss": 0.7263,
"step": 21550
},
{
"epoch": 0.41012764176605987,
"grad_norm": 1.8269189596176147,
"learning_rate": 1e-05,
"loss": 0.8321,
"step": 21560
},
{
"epoch": 0.4103178679449866,
"grad_norm": 1.6492648124694824,
"learning_rate": 1e-05,
"loss": 0.7872,
"step": 21570
},
{
"epoch": 0.4105080941239133,
"grad_norm": 1.9041223526000977,
"learning_rate": 1e-05,
"loss": 0.6366,
"step": 21580
},
{
"epoch": 0.4106983203028401,
"grad_norm": 3.013411045074463,
"learning_rate": 1e-05,
"loss": 0.5808,
"step": 21590
},
{
"epoch": 0.4108885464817668,
"grad_norm": 6.961212158203125,
"learning_rate": 1e-05,
"loss": 0.7427,
"step": 21600
},
{
"epoch": 0.4110787726606936,
"grad_norm": 1.3953548669815063,
"learning_rate": 1e-05,
"loss": 0.808,
"step": 21610
},
{
"epoch": 0.4112689988396203,
"grad_norm": 1.548154592514038,
"learning_rate": 1e-05,
"loss": 0.8112,
"step": 21620
},
{
"epoch": 0.41145922501854704,
"grad_norm": 1.9667688608169556,
"learning_rate": 1e-05,
"loss": 0.6133,
"step": 21630
},
{
"epoch": 0.4116494511974738,
"grad_norm": 3.429926872253418,
"learning_rate": 1e-05,
"loss": 0.5926,
"step": 21640
},
{
"epoch": 0.41183967737640054,
"grad_norm": 6.866636753082275,
"learning_rate": 1e-05,
"loss": 0.7448,
"step": 21650
},
{
"epoch": 0.41202990355532726,
"grad_norm": 1.4171299934387207,
"learning_rate": 1e-05,
"loss": 0.8105,
"step": 21660
},
{
"epoch": 0.41222012973425404,
"grad_norm": 1.4290895462036133,
"learning_rate": 1e-05,
"loss": 0.7704,
"step": 21670
},
{
"epoch": 0.41241035591318076,
"grad_norm": 1.8702961206436157,
"learning_rate": 1e-05,
"loss": 0.6069,
"step": 21680
},
{
"epoch": 0.41260058209210754,
"grad_norm": 2.6292471885681152,
"learning_rate": 1e-05,
"loss": 0.6143,
"step": 21690
},
{
"epoch": 0.41279080827103426,
"grad_norm": 6.300690650939941,
"learning_rate": 1e-05,
"loss": 0.7189,
"step": 21700
},
{
"epoch": 0.412981034449961,
"grad_norm": 1.5759576559066772,
"learning_rate": 1e-05,
"loss": 0.8004,
"step": 21710
},
{
"epoch": 0.41317126062888776,
"grad_norm": 1.8417686223983765,
"learning_rate": 1e-05,
"loss": 0.8045,
"step": 21720
},
{
"epoch": 0.4133614868078145,
"grad_norm": 2.0611612796783447,
"learning_rate": 1e-05,
"loss": 0.6583,
"step": 21730
},
{
"epoch": 0.41355171298674126,
"grad_norm": 3.0942375659942627,
"learning_rate": 1e-05,
"loss": 0.5949,
"step": 21740
},
{
"epoch": 0.413741939165668,
"grad_norm": 7.431186676025391,
"learning_rate": 1e-05,
"loss": 0.771,
"step": 21750
},
{
"epoch": 0.4139321653445947,
"grad_norm": 1.4004002809524536,
"learning_rate": 1e-05,
"loss": 0.7881,
"step": 21760
},
{
"epoch": 0.4141223915235215,
"grad_norm": 1.548725962638855,
"learning_rate": 1e-05,
"loss": 0.7844,
"step": 21770
},
{
"epoch": 0.4143126177024482,
"grad_norm": 2.0773072242736816,
"learning_rate": 1e-05,
"loss": 0.6185,
"step": 21780
},
{
"epoch": 0.4145028438813749,
"grad_norm": 2.8360114097595215,
"learning_rate": 1e-05,
"loss": 0.6056,
"step": 21790
},
{
"epoch": 0.4146930700603017,
"grad_norm": 7.396265506744385,
"learning_rate": 1e-05,
"loss": 0.7874,
"step": 21800
},
{
"epoch": 0.4148832962392284,
"grad_norm": 1.4704854488372803,
"learning_rate": 1e-05,
"loss": 0.8027,
"step": 21810
},
{
"epoch": 0.4150735224181552,
"grad_norm": 1.6272562742233276,
"learning_rate": 1e-05,
"loss": 0.7887,
"step": 21820
},
{
"epoch": 0.4152637485970819,
"grad_norm": 2.059877634048462,
"learning_rate": 1e-05,
"loss": 0.5901,
"step": 21830
},
{
"epoch": 0.41545397477600865,
"grad_norm": 2.800795316696167,
"learning_rate": 1e-05,
"loss": 0.5978,
"step": 21840
},
{
"epoch": 0.4156442009549354,
"grad_norm": 7.532629489898682,
"learning_rate": 1e-05,
"loss": 0.773,
"step": 21850
},
{
"epoch": 0.41583442713386215,
"grad_norm": 1.183721661567688,
"learning_rate": 1e-05,
"loss": 0.7837,
"step": 21860
},
{
"epoch": 0.41602465331278893,
"grad_norm": 1.6058815717697144,
"learning_rate": 1e-05,
"loss": 0.7783,
"step": 21870
},
{
"epoch": 0.41621487949171565,
"grad_norm": 1.9045753479003906,
"learning_rate": 1e-05,
"loss": 0.6504,
"step": 21880
},
{
"epoch": 0.4164051056706424,
"grad_norm": 2.571465492248535,
"learning_rate": 1e-05,
"loss": 0.5848,
"step": 21890
},
{
"epoch": 0.41659533184956915,
"grad_norm": 7.120177268981934,
"learning_rate": 1e-05,
"loss": 0.7677,
"step": 21900
},
{
"epoch": 0.4167855580284959,
"grad_norm": 1.5746983289718628,
"learning_rate": 1e-05,
"loss": 0.8226,
"step": 21910
},
{
"epoch": 0.41697578420742265,
"grad_norm": 1.6912455558776855,
"learning_rate": 1e-05,
"loss": 0.7392,
"step": 21920
},
{
"epoch": 0.4171660103863494,
"grad_norm": 2.148684024810791,
"learning_rate": 1e-05,
"loss": 0.6172,
"step": 21930
},
{
"epoch": 0.4173562365652761,
"grad_norm": 2.8162217140197754,
"learning_rate": 1e-05,
"loss": 0.6063,
"step": 21940
},
{
"epoch": 0.4175464627442029,
"grad_norm": 7.236839294433594,
"learning_rate": 1e-05,
"loss": 0.7394,
"step": 21950
},
{
"epoch": 0.4177366889231296,
"grad_norm": 1.4926108121871948,
"learning_rate": 1e-05,
"loss": 0.8317,
"step": 21960
},
{
"epoch": 0.4179269151020563,
"grad_norm": 1.8474189043045044,
"learning_rate": 1e-05,
"loss": 0.7661,
"step": 21970
},
{
"epoch": 0.4181171412809831,
"grad_norm": 2.1197712421417236,
"learning_rate": 1e-05,
"loss": 0.6165,
"step": 21980
},
{
"epoch": 0.4183073674599098,
"grad_norm": 2.797433376312256,
"learning_rate": 1e-05,
"loss": 0.6225,
"step": 21990
},
{
"epoch": 0.4184975936388366,
"grad_norm": 6.228893280029297,
"learning_rate": 1e-05,
"loss": 0.7483,
"step": 22000
},
{
"epoch": 0.4186878198177633,
"grad_norm": 1.4215539693832397,
"learning_rate": 1e-05,
"loss": 0.8321,
"step": 22010
},
{
"epoch": 0.41887804599669004,
"grad_norm": 1.4604730606079102,
"learning_rate": 1e-05,
"loss": 0.715,
"step": 22020
},
{
"epoch": 0.4190682721756168,
"grad_norm": 2.1636555194854736,
"learning_rate": 1e-05,
"loss": 0.6216,
"step": 22030
},
{
"epoch": 0.41925849835454354,
"grad_norm": 3.073380470275879,
"learning_rate": 1e-05,
"loss": 0.5693,
"step": 22040
},
{
"epoch": 0.4194487245334703,
"grad_norm": 7.762996673583984,
"learning_rate": 1e-05,
"loss": 0.7157,
"step": 22050
},
{
"epoch": 0.41963895071239704,
"grad_norm": 1.5549002885818481,
"learning_rate": 1e-05,
"loss": 0.8056,
"step": 22060
},
{
"epoch": 0.41982917689132376,
"grad_norm": 1.6262388229370117,
"learning_rate": 1e-05,
"loss": 0.7594,
"step": 22070
},
{
"epoch": 0.42001940307025054,
"grad_norm": 2.0899667739868164,
"learning_rate": 1e-05,
"loss": 0.6003,
"step": 22080
},
{
"epoch": 0.42020962924917726,
"grad_norm": 2.8294174671173096,
"learning_rate": 1e-05,
"loss": 0.6254,
"step": 22090
},
{
"epoch": 0.42039985542810404,
"grad_norm": 6.044638633728027,
"learning_rate": 1e-05,
"loss": 0.6987,
"step": 22100
},
{
"epoch": 0.42059008160703076,
"grad_norm": 1.456489086151123,
"learning_rate": 1e-05,
"loss": 0.8292,
"step": 22110
},
{
"epoch": 0.4207803077859575,
"grad_norm": 1.6765515804290771,
"learning_rate": 1e-05,
"loss": 0.7341,
"step": 22120
},
{
"epoch": 0.42097053396488426,
"grad_norm": 2.4537079334259033,
"learning_rate": 1e-05,
"loss": 0.6253,
"step": 22130
},
{
"epoch": 0.421160760143811,
"grad_norm": 3.135146379470825,
"learning_rate": 1e-05,
"loss": 0.5861,
"step": 22140
},
{
"epoch": 0.4213509863227377,
"grad_norm": 8.208638191223145,
"learning_rate": 1e-05,
"loss": 0.7258,
"step": 22150
},
{
"epoch": 0.4215412125016645,
"grad_norm": 1.7043077945709229,
"learning_rate": 1e-05,
"loss": 0.8339,
"step": 22160
},
{
"epoch": 0.4217314386805912,
"grad_norm": 1.5617111921310425,
"learning_rate": 1e-05,
"loss": 0.7694,
"step": 22170
},
{
"epoch": 0.421921664859518,
"grad_norm": 2.646005868911743,
"learning_rate": 1e-05,
"loss": 0.6193,
"step": 22180
},
{
"epoch": 0.4221118910384447,
"grad_norm": 2.784905433654785,
"learning_rate": 1e-05,
"loss": 0.6259,
"step": 22190
},
{
"epoch": 0.42230211721737143,
"grad_norm": 6.417455673217773,
"learning_rate": 1e-05,
"loss": 0.7281,
"step": 22200
},
{
"epoch": 0.4224923433962982,
"grad_norm": 1.6433660984039307,
"learning_rate": 1e-05,
"loss": 0.8395,
"step": 22210
},
{
"epoch": 0.42268256957522493,
"grad_norm": 1.7666829824447632,
"learning_rate": 1e-05,
"loss": 0.7857,
"step": 22220
},
{
"epoch": 0.4228727957541517,
"grad_norm": 2.1245760917663574,
"learning_rate": 1e-05,
"loss": 0.6343,
"step": 22230
},
{
"epoch": 0.42306302193307843,
"grad_norm": 2.488694667816162,
"learning_rate": 1e-05,
"loss": 0.617,
"step": 22240
},
{
"epoch": 0.42325324811200515,
"grad_norm": 7.654630184173584,
"learning_rate": 1e-05,
"loss": 0.7663,
"step": 22250
},
{
"epoch": 0.42344347429093193,
"grad_norm": 1.3615024089813232,
"learning_rate": 1e-05,
"loss": 0.8351,
"step": 22260
},
{
"epoch": 0.42363370046985865,
"grad_norm": 1.6781669855117798,
"learning_rate": 1e-05,
"loss": 0.766,
"step": 22270
},
{
"epoch": 0.42382392664878543,
"grad_norm": 2.8877549171447754,
"learning_rate": 1e-05,
"loss": 0.5753,
"step": 22280
},
{
"epoch": 0.42401415282771215,
"grad_norm": 2.656270742416382,
"learning_rate": 1e-05,
"loss": 0.6138,
"step": 22290
},
{
"epoch": 0.4242043790066389,
"grad_norm": 7.828327655792236,
"learning_rate": 1e-05,
"loss": 0.7702,
"step": 22300
},
{
"epoch": 0.42439460518556565,
"grad_norm": 1.6173627376556396,
"learning_rate": 1e-05,
"loss": 0.8362,
"step": 22310
},
{
"epoch": 0.4245848313644924,
"grad_norm": 1.4991214275360107,
"learning_rate": 1e-05,
"loss": 0.7478,
"step": 22320
},
{
"epoch": 0.4247750575434191,
"grad_norm": 2.2232842445373535,
"learning_rate": 1e-05,
"loss": 0.5771,
"step": 22330
},
{
"epoch": 0.4249652837223459,
"grad_norm": 2.848970890045166,
"learning_rate": 1e-05,
"loss": 0.6082,
"step": 22340
},
{
"epoch": 0.4251555099012726,
"grad_norm": 7.872457027435303,
"learning_rate": 1e-05,
"loss": 0.7543,
"step": 22350
},
{
"epoch": 0.4253457360801994,
"grad_norm": 1.367767333984375,
"learning_rate": 1e-05,
"loss": 0.8528,
"step": 22360
},
{
"epoch": 0.4255359622591261,
"grad_norm": 1.8605856895446777,
"learning_rate": 1e-05,
"loss": 0.7733,
"step": 22370
},
{
"epoch": 0.4257261884380528,
"grad_norm": 2.0029051303863525,
"learning_rate": 1e-05,
"loss": 0.6246,
"step": 22380
},
{
"epoch": 0.4259164146169796,
"grad_norm": 2.5023515224456787,
"learning_rate": 1e-05,
"loss": 0.6093,
"step": 22390
},
{
"epoch": 0.4261066407959063,
"grad_norm": 6.068182945251465,
"learning_rate": 1e-05,
"loss": 0.7066,
"step": 22400
},
{
"epoch": 0.4262968669748331,
"grad_norm": 1.3323262929916382,
"learning_rate": 1e-05,
"loss": 0.7923,
"step": 22410
},
{
"epoch": 0.4264870931537598,
"grad_norm": 2.335211753845215,
"learning_rate": 1e-05,
"loss": 0.7631,
"step": 22420
},
{
"epoch": 0.42667731933268654,
"grad_norm": 2.137937068939209,
"learning_rate": 1e-05,
"loss": 0.61,
"step": 22430
},
{
"epoch": 0.4268675455116133,
"grad_norm": 2.510817289352417,
"learning_rate": 1e-05,
"loss": 0.5958,
"step": 22440
},
{
"epoch": 0.42705777169054004,
"grad_norm": 7.117891311645508,
"learning_rate": 1e-05,
"loss": 0.7055,
"step": 22450
},
{
"epoch": 0.4272479978694668,
"grad_norm": 2.3258843421936035,
"learning_rate": 1e-05,
"loss": 0.8345,
"step": 22460
},
{
"epoch": 0.42743822404839354,
"grad_norm": 1.9138954877853394,
"learning_rate": 1e-05,
"loss": 0.7544,
"step": 22470
},
{
"epoch": 0.42762845022732027,
"grad_norm": 2.1563355922698975,
"learning_rate": 1e-05,
"loss": 0.626,
"step": 22480
},
{
"epoch": 0.42781867640624704,
"grad_norm": 2.685288429260254,
"learning_rate": 1e-05,
"loss": 0.6049,
"step": 22490
},
{
"epoch": 0.42800890258517377,
"grad_norm": 6.471707820892334,
"learning_rate": 1e-05,
"loss": 0.7089,
"step": 22500
},
{
"epoch": 0.4281991287641005,
"grad_norm": 1.3516937494277954,
"learning_rate": 1e-05,
"loss": 0.7849,
"step": 22510
},
{
"epoch": 0.42838935494302727,
"grad_norm": 1.5422897338867188,
"learning_rate": 1e-05,
"loss": 0.794,
"step": 22520
},
{
"epoch": 0.428579581121954,
"grad_norm": 2.0554656982421875,
"learning_rate": 1e-05,
"loss": 0.6386,
"step": 22530
},
{
"epoch": 0.42876980730088077,
"grad_norm": 3.0782418251037598,
"learning_rate": 1e-05,
"loss": 0.6065,
"step": 22540
},
{
"epoch": 0.4289600334798075,
"grad_norm": 6.9862470626831055,
"learning_rate": 1e-05,
"loss": 0.706,
"step": 22550
},
{
"epoch": 0.4291502596587342,
"grad_norm": 1.397106647491455,
"learning_rate": 1e-05,
"loss": 0.8038,
"step": 22560
},
{
"epoch": 0.429340485837661,
"grad_norm": 1.9687145948410034,
"learning_rate": 1e-05,
"loss": 0.7645,
"step": 22570
},
{
"epoch": 0.4295307120165877,
"grad_norm": 1.645969033241272,
"learning_rate": 1e-05,
"loss": 0.6125,
"step": 22580
},
{
"epoch": 0.4297209381955145,
"grad_norm": 2.6927411556243896,
"learning_rate": 1e-05,
"loss": 0.5706,
"step": 22590
},
{
"epoch": 0.4299111643744412,
"grad_norm": 6.746013164520264,
"learning_rate": 1e-05,
"loss": 0.738,
"step": 22600
},
{
"epoch": 0.43010139055336793,
"grad_norm": 1.368550181388855,
"learning_rate": 1e-05,
"loss": 0.8104,
"step": 22610
},
{
"epoch": 0.4302916167322947,
"grad_norm": 1.8987951278686523,
"learning_rate": 1e-05,
"loss": 0.7362,
"step": 22620
},
{
"epoch": 0.43048184291122144,
"grad_norm": 1.9571783542633057,
"learning_rate": 1e-05,
"loss": 0.6013,
"step": 22630
},
{
"epoch": 0.4306720690901482,
"grad_norm": 2.759227991104126,
"learning_rate": 1e-05,
"loss": 0.5899,
"step": 22640
},
{
"epoch": 0.43086229526907494,
"grad_norm": 6.114544868469238,
"learning_rate": 1e-05,
"loss": 0.7282,
"step": 22650
},
{
"epoch": 0.43105252144800166,
"grad_norm": 1.5622162818908691,
"learning_rate": 1e-05,
"loss": 0.822,
"step": 22660
},
{
"epoch": 0.43124274762692844,
"grad_norm": 1.8977807760238647,
"learning_rate": 1e-05,
"loss": 0.7646,
"step": 22670
},
{
"epoch": 0.43143297380585516,
"grad_norm": 2.2254631519317627,
"learning_rate": 1e-05,
"loss": 0.579,
"step": 22680
},
{
"epoch": 0.4316231999847819,
"grad_norm": 2.9056873321533203,
"learning_rate": 1e-05,
"loss": 0.605,
"step": 22690
},
{
"epoch": 0.43181342616370866,
"grad_norm": 7.09505558013916,
"learning_rate": 1e-05,
"loss": 0.7203,
"step": 22700
},
{
"epoch": 0.4320036523426354,
"grad_norm": 1.3744453191757202,
"learning_rate": 1e-05,
"loss": 0.7824,
"step": 22710
},
{
"epoch": 0.43219387852156216,
"grad_norm": 1.7175064086914062,
"learning_rate": 1e-05,
"loss": 0.7373,
"step": 22720
},
{
"epoch": 0.4323841047004889,
"grad_norm": 1.9812932014465332,
"learning_rate": 1e-05,
"loss": 0.6307,
"step": 22730
},
{
"epoch": 0.4325743308794156,
"grad_norm": 2.5877597332000732,
"learning_rate": 1e-05,
"loss": 0.5944,
"step": 22740
},
{
"epoch": 0.4327645570583424,
"grad_norm": 6.665163040161133,
"learning_rate": 1e-05,
"loss": 0.7252,
"step": 22750
},
{
"epoch": 0.4329547832372691,
"grad_norm": 1.4962937831878662,
"learning_rate": 1e-05,
"loss": 0.8133,
"step": 22760
},
{
"epoch": 0.4331450094161959,
"grad_norm": 1.5577149391174316,
"learning_rate": 1e-05,
"loss": 0.7439,
"step": 22770
},
{
"epoch": 0.4333352355951226,
"grad_norm": 2.3257436752319336,
"learning_rate": 1e-05,
"loss": 0.6175,
"step": 22780
},
{
"epoch": 0.4335254617740493,
"grad_norm": 3.2422592639923096,
"learning_rate": 1e-05,
"loss": 0.5829,
"step": 22790
},
{
"epoch": 0.4337156879529761,
"grad_norm": 6.280292987823486,
"learning_rate": 1e-05,
"loss": 0.7542,
"step": 22800
},
{
"epoch": 0.4339059141319028,
"grad_norm": 1.4401211738586426,
"learning_rate": 1e-05,
"loss": 0.8324,
"step": 22810
},
{
"epoch": 0.4340961403108296,
"grad_norm": 1.5887025594711304,
"learning_rate": 1e-05,
"loss": 0.7628,
"step": 22820
},
{
"epoch": 0.4342863664897563,
"grad_norm": 2.085007429122925,
"learning_rate": 1e-05,
"loss": 0.6241,
"step": 22830
},
{
"epoch": 0.43447659266868305,
"grad_norm": 2.6930272579193115,
"learning_rate": 1e-05,
"loss": 0.5983,
"step": 22840
},
{
"epoch": 0.4346668188476098,
"grad_norm": 6.500943183898926,
"learning_rate": 1e-05,
"loss": 0.7077,
"step": 22850
},
{
"epoch": 0.43485704502653655,
"grad_norm": 1.7516988515853882,
"learning_rate": 1e-05,
"loss": 0.8278,
"step": 22860
},
{
"epoch": 0.43504727120546327,
"grad_norm": 1.5914795398712158,
"learning_rate": 1e-05,
"loss": 0.7357,
"step": 22870
},
{
"epoch": 0.43523749738439005,
"grad_norm": 2.1554529666900635,
"learning_rate": 1e-05,
"loss": 0.6002,
"step": 22880
},
{
"epoch": 0.43542772356331677,
"grad_norm": 2.6188783645629883,
"learning_rate": 1e-05,
"loss": 0.5992,
"step": 22890
},
{
"epoch": 0.43561794974224355,
"grad_norm": 7.268838882446289,
"learning_rate": 1e-05,
"loss": 0.7139,
"step": 22900
},
{
"epoch": 0.43580817592117027,
"grad_norm": 1.4615405797958374,
"learning_rate": 1e-05,
"loss": 0.8125,
"step": 22910
},
{
"epoch": 0.435998402100097,
"grad_norm": 1.8655891418457031,
"learning_rate": 1e-05,
"loss": 0.7636,
"step": 22920
},
{
"epoch": 0.43618862827902377,
"grad_norm": 2.240283966064453,
"learning_rate": 1e-05,
"loss": 0.5995,
"step": 22930
},
{
"epoch": 0.4363788544579505,
"grad_norm": 2.645955801010132,
"learning_rate": 1e-05,
"loss": 0.6215,
"step": 22940
},
{
"epoch": 0.43656908063687727,
"grad_norm": 6.524368762969971,
"learning_rate": 1e-05,
"loss": 0.6731,
"step": 22950
},
{
"epoch": 0.436759306815804,
"grad_norm": 1.6127771139144897,
"learning_rate": 1e-05,
"loss": 0.8325,
"step": 22960
},
{
"epoch": 0.4369495329947307,
"grad_norm": 1.5488996505737305,
"learning_rate": 1e-05,
"loss": 0.7742,
"step": 22970
},
{
"epoch": 0.4371397591736575,
"grad_norm": 2.2502658367156982,
"learning_rate": 1e-05,
"loss": 0.6599,
"step": 22980
},
{
"epoch": 0.4373299853525842,
"grad_norm": 2.424576997756958,
"learning_rate": 1e-05,
"loss": 0.6171,
"step": 22990
},
{
"epoch": 0.437520211531511,
"grad_norm": 7.69747257232666,
"learning_rate": 1e-05,
"loss": 0.7553,
"step": 23000
},
{
"epoch": 0.4377104377104377,
"grad_norm": 1.5736782550811768,
"learning_rate": 1e-05,
"loss": 0.8189,
"step": 23010
},
{
"epoch": 0.43790066388936444,
"grad_norm": 1.785718321800232,
"learning_rate": 1e-05,
"loss": 0.8043,
"step": 23020
},
{
"epoch": 0.4380908900682912,
"grad_norm": 1.899800181388855,
"learning_rate": 1e-05,
"loss": 0.6123,
"step": 23030
},
{
"epoch": 0.43828111624721794,
"grad_norm": 2.558415174484253,
"learning_rate": 1e-05,
"loss": 0.5976,
"step": 23040
},
{
"epoch": 0.43847134242614466,
"grad_norm": 6.158372402191162,
"learning_rate": 1e-05,
"loss": 0.6982,
"step": 23050
},
{
"epoch": 0.43866156860507144,
"grad_norm": 1.514840006828308,
"learning_rate": 1e-05,
"loss": 0.8181,
"step": 23060
},
{
"epoch": 0.43885179478399816,
"grad_norm": 2.2898526191711426,
"learning_rate": 1e-05,
"loss": 0.7609,
"step": 23070
},
{
"epoch": 0.43904202096292494,
"grad_norm": 1.978365421295166,
"learning_rate": 1e-05,
"loss": 0.5872,
"step": 23080
},
{
"epoch": 0.43923224714185166,
"grad_norm": 2.61547589302063,
"learning_rate": 1e-05,
"loss": 0.5949,
"step": 23090
},
{
"epoch": 0.4394224733207784,
"grad_norm": 5.580928802490234,
"learning_rate": 1e-05,
"loss": 0.7139,
"step": 23100
},
{
"epoch": 0.43961269949970516,
"grad_norm": 1.5627914667129517,
"learning_rate": 1e-05,
"loss": 0.816,
"step": 23110
},
{
"epoch": 0.4398029256786319,
"grad_norm": 1.7263432741165161,
"learning_rate": 1e-05,
"loss": 0.761,
"step": 23120
},
{
"epoch": 0.43999315185755866,
"grad_norm": 1.9462865591049194,
"learning_rate": 1e-05,
"loss": 0.5814,
"step": 23130
},
{
"epoch": 0.4401833780364854,
"grad_norm": 2.481372356414795,
"learning_rate": 1e-05,
"loss": 0.6155,
"step": 23140
},
{
"epoch": 0.4403736042154121,
"grad_norm": 7.702070236206055,
"learning_rate": 1e-05,
"loss": 0.7482,
"step": 23150
},
{
"epoch": 0.4405638303943389,
"grad_norm": 1.4825353622436523,
"learning_rate": 1e-05,
"loss": 0.8167,
"step": 23160
},
{
"epoch": 0.4407540565732656,
"grad_norm": 1.7170199155807495,
"learning_rate": 1e-05,
"loss": 0.7624,
"step": 23170
},
{
"epoch": 0.44094428275219233,
"grad_norm": 1.8658539056777954,
"learning_rate": 1e-05,
"loss": 0.6151,
"step": 23180
},
{
"epoch": 0.4411345089311191,
"grad_norm": 2.8086531162261963,
"learning_rate": 1e-05,
"loss": 0.6043,
"step": 23190
},
{
"epoch": 0.44132473511004583,
"grad_norm": 7.021834850311279,
"learning_rate": 1e-05,
"loss": 0.7315,
"step": 23200
},
{
"epoch": 0.4415149612889726,
"grad_norm": 1.3526475429534912,
"learning_rate": 1e-05,
"loss": 0.8351,
"step": 23210
},
{
"epoch": 0.44170518746789933,
"grad_norm": 1.532271385192871,
"learning_rate": 1e-05,
"loss": 0.7427,
"step": 23220
},
{
"epoch": 0.44189541364682605,
"grad_norm": 2.2991366386413574,
"learning_rate": 1e-05,
"loss": 0.6382,
"step": 23230
},
{
"epoch": 0.44208563982575283,
"grad_norm": 2.8424527645111084,
"learning_rate": 1e-05,
"loss": 0.5845,
"step": 23240
},
{
"epoch": 0.44227586600467955,
"grad_norm": 6.907649040222168,
"learning_rate": 1e-05,
"loss": 0.7537,
"step": 23250
},
{
"epoch": 0.44246609218360633,
"grad_norm": 1.3680170774459839,
"learning_rate": 1e-05,
"loss": 0.8278,
"step": 23260
},
{
"epoch": 0.44265631836253305,
"grad_norm": 1.736412763595581,
"learning_rate": 1e-05,
"loss": 0.78,
"step": 23270
},
{
"epoch": 0.4428465445414598,
"grad_norm": 2.0505850315093994,
"learning_rate": 1e-05,
"loss": 0.6556,
"step": 23280
},
{
"epoch": 0.44303677072038655,
"grad_norm": 2.707890272140503,
"learning_rate": 1e-05,
"loss": 0.6193,
"step": 23290
},
{
"epoch": 0.4432269968993133,
"grad_norm": 6.543889999389648,
"learning_rate": 1e-05,
"loss": 0.7132,
"step": 23300
},
{
"epoch": 0.44341722307824005,
"grad_norm": 1.4513194561004639,
"learning_rate": 1e-05,
"loss": 0.8179,
"step": 23310
},
{
"epoch": 0.4436074492571668,
"grad_norm": 1.7616751194000244,
"learning_rate": 1e-05,
"loss": 0.7679,
"step": 23320
},
{
"epoch": 0.4437976754360935,
"grad_norm": 1.9056065082550049,
"learning_rate": 1e-05,
"loss": 0.607,
"step": 23330
},
{
"epoch": 0.4439879016150203,
"grad_norm": 2.4347009658813477,
"learning_rate": 1e-05,
"loss": 0.5824,
"step": 23340
},
{
"epoch": 0.444178127793947,
"grad_norm": 6.513676643371582,
"learning_rate": 1e-05,
"loss": 0.7336,
"step": 23350
},
{
"epoch": 0.4443683539728737,
"grad_norm": 1.38242506980896,
"learning_rate": 1e-05,
"loss": 0.8242,
"step": 23360
},
{
"epoch": 0.4445585801518005,
"grad_norm": 1.6413469314575195,
"learning_rate": 1e-05,
"loss": 0.815,
"step": 23370
},
{
"epoch": 0.4447488063307272,
"grad_norm": 1.9574592113494873,
"learning_rate": 1e-05,
"loss": 0.624,
"step": 23380
},
{
"epoch": 0.444939032509654,
"grad_norm": 2.696187973022461,
"learning_rate": 1e-05,
"loss": 0.6153,
"step": 23390
},
{
"epoch": 0.4451292586885807,
"grad_norm": 6.090272903442383,
"learning_rate": 1e-05,
"loss": 0.6806,
"step": 23400
},
{
"epoch": 0.44531948486750744,
"grad_norm": 1.6440588235855103,
"learning_rate": 1e-05,
"loss": 0.808,
"step": 23410
},
{
"epoch": 0.4455097110464342,
"grad_norm": 1.650538682937622,
"learning_rate": 1e-05,
"loss": 0.7556,
"step": 23420
},
{
"epoch": 0.44569993722536094,
"grad_norm": 1.969576358795166,
"learning_rate": 1e-05,
"loss": 0.6265,
"step": 23430
},
{
"epoch": 0.4458901634042877,
"grad_norm": 2.900421380996704,
"learning_rate": 1e-05,
"loss": 0.5996,
"step": 23440
},
{
"epoch": 0.44608038958321444,
"grad_norm": 8.251358032226562,
"learning_rate": 1e-05,
"loss": 0.7232,
"step": 23450
},
{
"epoch": 0.44627061576214117,
"grad_norm": 1.752803087234497,
"learning_rate": 1e-05,
"loss": 0.7837,
"step": 23460
},
{
"epoch": 0.44646084194106794,
"grad_norm": 1.6273314952850342,
"learning_rate": 1e-05,
"loss": 0.7396,
"step": 23470
},
{
"epoch": 0.44665106811999467,
"grad_norm": 2.0402679443359375,
"learning_rate": 1e-05,
"loss": 0.6017,
"step": 23480
},
{
"epoch": 0.44684129429892144,
"grad_norm": 2.5864932537078857,
"learning_rate": 1e-05,
"loss": 0.6143,
"step": 23490
},
{
"epoch": 0.44703152047784817,
"grad_norm": 6.227820873260498,
"learning_rate": 1e-05,
"loss": 0.7225,
"step": 23500
},
{
"epoch": 0.4472217466567749,
"grad_norm": 1.5930750370025635,
"learning_rate": 1e-05,
"loss": 0.8203,
"step": 23510
},
{
"epoch": 0.44741197283570167,
"grad_norm": 1.580527424812317,
"learning_rate": 1e-05,
"loss": 0.7683,
"step": 23520
},
{
"epoch": 0.4476021990146284,
"grad_norm": 1.8535393476486206,
"learning_rate": 1e-05,
"loss": 0.5765,
"step": 23530
},
{
"epoch": 0.4477924251935551,
"grad_norm": 2.8945422172546387,
"learning_rate": 1e-05,
"loss": 0.6289,
"step": 23540
},
{
"epoch": 0.4479826513724819,
"grad_norm": 6.095447540283203,
"learning_rate": 1e-05,
"loss": 0.7309,
"step": 23550
},
{
"epoch": 0.4481728775514086,
"grad_norm": 1.2353800535202026,
"learning_rate": 1e-05,
"loss": 0.7993,
"step": 23560
},
{
"epoch": 0.4483631037303354,
"grad_norm": 1.5261456966400146,
"learning_rate": 1e-05,
"loss": 0.7575,
"step": 23570
},
{
"epoch": 0.4485533299092621,
"grad_norm": 2.1640236377716064,
"learning_rate": 1e-05,
"loss": 0.6019,
"step": 23580
},
{
"epoch": 0.44874355608818883,
"grad_norm": 2.27223539352417,
"learning_rate": 1e-05,
"loss": 0.5656,
"step": 23590
},
{
"epoch": 0.4489337822671156,
"grad_norm": 7.814069747924805,
"learning_rate": 1e-05,
"loss": 0.6964,
"step": 23600
},
{
"epoch": 0.44912400844604233,
"grad_norm": 1.4168845415115356,
"learning_rate": 1e-05,
"loss": 0.813,
"step": 23610
},
{
"epoch": 0.4493142346249691,
"grad_norm": 1.67057204246521,
"learning_rate": 1e-05,
"loss": 0.7512,
"step": 23620
},
{
"epoch": 0.44950446080389583,
"grad_norm": 2.0266611576080322,
"learning_rate": 1e-05,
"loss": 0.6165,
"step": 23630
},
{
"epoch": 0.44969468698282256,
"grad_norm": 2.473586082458496,
"learning_rate": 1e-05,
"loss": 0.5918,
"step": 23640
},
{
"epoch": 0.44988491316174933,
"grad_norm": 7.469400882720947,
"learning_rate": 1e-05,
"loss": 0.752,
"step": 23650
},
{
"epoch": 0.45007513934067606,
"grad_norm": 1.4267483949661255,
"learning_rate": 1e-05,
"loss": 0.8281,
"step": 23660
},
{
"epoch": 0.45026536551960283,
"grad_norm": 1.8579682111740112,
"learning_rate": 1e-05,
"loss": 0.7508,
"step": 23670
},
{
"epoch": 0.45045559169852956,
"grad_norm": 1.9884873628616333,
"learning_rate": 1e-05,
"loss": 0.6538,
"step": 23680
},
{
"epoch": 0.4506458178774563,
"grad_norm": 2.5367233753204346,
"learning_rate": 1e-05,
"loss": 0.5947,
"step": 23690
},
{
"epoch": 0.45083604405638306,
"grad_norm": 6.5690226554870605,
"learning_rate": 1e-05,
"loss": 0.6997,
"step": 23700
},
{
"epoch": 0.4510262702353098,
"grad_norm": 1.4327682256698608,
"learning_rate": 1e-05,
"loss": 0.8114,
"step": 23710
},
{
"epoch": 0.4512164964142365,
"grad_norm": 1.6522828340530396,
"learning_rate": 1e-05,
"loss": 0.8045,
"step": 23720
},
{
"epoch": 0.4514067225931633,
"grad_norm": 2.082068920135498,
"learning_rate": 1e-05,
"loss": 0.6196,
"step": 23730
},
{
"epoch": 0.45159694877209,
"grad_norm": 2.446380138397217,
"learning_rate": 1e-05,
"loss": 0.5864,
"step": 23740
},
{
"epoch": 0.4517871749510168,
"grad_norm": 6.408164024353027,
"learning_rate": 1e-05,
"loss": 0.7009,
"step": 23750
},
{
"epoch": 0.4519774011299435,
"grad_norm": 1.531535029411316,
"learning_rate": 1e-05,
"loss": 0.7992,
"step": 23760
},
{
"epoch": 0.4521676273088702,
"grad_norm": 2.251401901245117,
"learning_rate": 1e-05,
"loss": 0.756,
"step": 23770
},
{
"epoch": 0.452357853487797,
"grad_norm": 2.0824103355407715,
"learning_rate": 1e-05,
"loss": 0.6296,
"step": 23780
},
{
"epoch": 0.4525480796667237,
"grad_norm": 2.495861530303955,
"learning_rate": 1e-05,
"loss": 0.5596,
"step": 23790
},
{
"epoch": 0.4527383058456505,
"grad_norm": 8.60665512084961,
"learning_rate": 1e-05,
"loss": 0.7058,
"step": 23800
},
{
"epoch": 0.4529285320245772,
"grad_norm": 1.3445286750793457,
"learning_rate": 1e-05,
"loss": 0.796,
"step": 23810
},
{
"epoch": 0.45311875820350395,
"grad_norm": 1.6318038702011108,
"learning_rate": 1e-05,
"loss": 0.7264,
"step": 23820
},
{
"epoch": 0.4533089843824307,
"grad_norm": 1.8862273693084717,
"learning_rate": 1e-05,
"loss": 0.5866,
"step": 23830
},
{
"epoch": 0.45349921056135745,
"grad_norm": 2.9340245723724365,
"learning_rate": 1e-05,
"loss": 0.5795,
"step": 23840
},
{
"epoch": 0.4536894367402842,
"grad_norm": 6.770589351654053,
"learning_rate": 1e-05,
"loss": 0.7518,
"step": 23850
},
{
"epoch": 0.45387966291921095,
"grad_norm": 1.2478692531585693,
"learning_rate": 1e-05,
"loss": 0.8181,
"step": 23860
},
{
"epoch": 0.45406988909813767,
"grad_norm": 1.9352004528045654,
"learning_rate": 1e-05,
"loss": 0.7746,
"step": 23870
},
{
"epoch": 0.45426011527706445,
"grad_norm": 1.8664720058441162,
"learning_rate": 1e-05,
"loss": 0.5927,
"step": 23880
},
{
"epoch": 0.45445034145599117,
"grad_norm": 2.474200487136841,
"learning_rate": 1e-05,
"loss": 0.631,
"step": 23890
},
{
"epoch": 0.4546405676349179,
"grad_norm": 6.4189934730529785,
"learning_rate": 1e-05,
"loss": 0.696,
"step": 23900
},
{
"epoch": 0.45483079381384467,
"grad_norm": 1.2441041469573975,
"learning_rate": 1e-05,
"loss": 0.7993,
"step": 23910
},
{
"epoch": 0.4550210199927714,
"grad_norm": 1.379363775253296,
"learning_rate": 1e-05,
"loss": 0.7654,
"step": 23920
},
{
"epoch": 0.45521124617169817,
"grad_norm": 1.9366620779037476,
"learning_rate": 1e-05,
"loss": 0.6052,
"step": 23930
},
{
"epoch": 0.4554014723506249,
"grad_norm": 2.2613441944122314,
"learning_rate": 1e-05,
"loss": 0.5943,
"step": 23940
},
{
"epoch": 0.4555916985295516,
"grad_norm": 7.525345802307129,
"learning_rate": 1e-05,
"loss": 0.7072,
"step": 23950
},
{
"epoch": 0.4557819247084784,
"grad_norm": 1.5408786535263062,
"learning_rate": 1e-05,
"loss": 0.8031,
"step": 23960
},
{
"epoch": 0.4559721508874051,
"grad_norm": 2.0759196281433105,
"learning_rate": 1e-05,
"loss": 0.7403,
"step": 23970
},
{
"epoch": 0.4561623770663319,
"grad_norm": 2.095139503479004,
"learning_rate": 1e-05,
"loss": 0.6011,
"step": 23980
},
{
"epoch": 0.4563526032452586,
"grad_norm": 2.4406020641326904,
"learning_rate": 1e-05,
"loss": 0.6019,
"step": 23990
},
{
"epoch": 0.45654282942418534,
"grad_norm": 7.5866475105285645,
"learning_rate": 1e-05,
"loss": 0.7595,
"step": 24000
},
{
"epoch": 0.4567330556031121,
"grad_norm": 1.2872250080108643,
"learning_rate": 1e-05,
"loss": 0.7997,
"step": 24010
},
{
"epoch": 0.45692328178203884,
"grad_norm": 1.8803125619888306,
"learning_rate": 1e-05,
"loss": 0.7596,
"step": 24020
},
{
"epoch": 0.4571135079609656,
"grad_norm": 1.9355621337890625,
"learning_rate": 1e-05,
"loss": 0.6326,
"step": 24030
},
{
"epoch": 0.45730373413989234,
"grad_norm": 2.4504873752593994,
"learning_rate": 1e-05,
"loss": 0.604,
"step": 24040
},
{
"epoch": 0.45749396031881906,
"grad_norm": 6.2335638999938965,
"learning_rate": 1e-05,
"loss": 0.7257,
"step": 24050
},
{
"epoch": 0.45768418649774584,
"grad_norm": 1.6137391328811646,
"learning_rate": 1e-05,
"loss": 0.8022,
"step": 24060
},
{
"epoch": 0.45787441267667256,
"grad_norm": 1.7394353151321411,
"learning_rate": 1e-05,
"loss": 0.7518,
"step": 24070
},
{
"epoch": 0.4580646388555993,
"grad_norm": 2.1530978679656982,
"learning_rate": 1e-05,
"loss": 0.6173,
"step": 24080
},
{
"epoch": 0.45825486503452606,
"grad_norm": 2.3911030292510986,
"learning_rate": 1e-05,
"loss": 0.581,
"step": 24090
},
{
"epoch": 0.4584450912134528,
"grad_norm": 7.079417705535889,
"learning_rate": 1e-05,
"loss": 0.7471,
"step": 24100
},
{
"epoch": 0.45863531739237956,
"grad_norm": 1.3778908252716064,
"learning_rate": 1e-05,
"loss": 0.8156,
"step": 24110
},
{
"epoch": 0.4588255435713063,
"grad_norm": 1.5596013069152832,
"learning_rate": 1e-05,
"loss": 0.7428,
"step": 24120
},
{
"epoch": 0.459015769750233,
"grad_norm": 2.008509874343872,
"learning_rate": 1e-05,
"loss": 0.6269,
"step": 24130
},
{
"epoch": 0.4592059959291598,
"grad_norm": 2.210686445236206,
"learning_rate": 1e-05,
"loss": 0.5757,
"step": 24140
},
{
"epoch": 0.4593962221080865,
"grad_norm": 7.073795318603516,
"learning_rate": 1e-05,
"loss": 0.7382,
"step": 24150
},
{
"epoch": 0.4595864482870133,
"grad_norm": 1.1860955953598022,
"learning_rate": 1e-05,
"loss": 0.78,
"step": 24160
},
{
"epoch": 0.45977667446594,
"grad_norm": 1.7176934480667114,
"learning_rate": 1e-05,
"loss": 0.7608,
"step": 24170
},
{
"epoch": 0.4599669006448667,
"grad_norm": 1.7659779787063599,
"learning_rate": 1e-05,
"loss": 0.6197,
"step": 24180
},
{
"epoch": 0.4601571268237935,
"grad_norm": 2.610393524169922,
"learning_rate": 1e-05,
"loss": 0.6015,
"step": 24190
},
{
"epoch": 0.4603473530027202,
"grad_norm": 6.386167526245117,
"learning_rate": 1e-05,
"loss": 0.6888,
"step": 24200
},
{
"epoch": 0.460537579181647,
"grad_norm": 1.4856669902801514,
"learning_rate": 1e-05,
"loss": 0.8259,
"step": 24210
},
{
"epoch": 0.46072780536057373,
"grad_norm": 1.7215503454208374,
"learning_rate": 1e-05,
"loss": 0.7627,
"step": 24220
},
{
"epoch": 0.46091803153950045,
"grad_norm": 1.8146615028381348,
"learning_rate": 1e-05,
"loss": 0.6253,
"step": 24230
},
{
"epoch": 0.46110825771842723,
"grad_norm": 2.6651651859283447,
"learning_rate": 1e-05,
"loss": 0.6135,
"step": 24240
},
{
"epoch": 0.46129848389735395,
"grad_norm": 5.836438179016113,
"learning_rate": 1e-05,
"loss": 0.7351,
"step": 24250
},
{
"epoch": 0.4614887100762807,
"grad_norm": 1.5189788341522217,
"learning_rate": 1e-05,
"loss": 0.824,
"step": 24260
},
{
"epoch": 0.46167893625520745,
"grad_norm": 1.7803634405136108,
"learning_rate": 1e-05,
"loss": 0.7529,
"step": 24270
},
{
"epoch": 0.4618691624341342,
"grad_norm": 1.8436952829360962,
"learning_rate": 1e-05,
"loss": 0.6346,
"step": 24280
},
{
"epoch": 0.46205938861306095,
"grad_norm": 2.8803257942199707,
"learning_rate": 1e-05,
"loss": 0.5955,
"step": 24290
},
{
"epoch": 0.4622496147919877,
"grad_norm": 7.429708957672119,
"learning_rate": 1e-05,
"loss": 0.7754,
"step": 24300
},
{
"epoch": 0.4624398409709144,
"grad_norm": 1.3199936151504517,
"learning_rate": 1e-05,
"loss": 0.8046,
"step": 24310
},
{
"epoch": 0.4626300671498412,
"grad_norm": 1.5775935649871826,
"learning_rate": 1e-05,
"loss": 0.7562,
"step": 24320
},
{
"epoch": 0.4628202933287679,
"grad_norm": 2.075223922729492,
"learning_rate": 1e-05,
"loss": 0.6112,
"step": 24330
},
{
"epoch": 0.4630105195076947,
"grad_norm": 2.5630037784576416,
"learning_rate": 1e-05,
"loss": 0.5967,
"step": 24340
},
{
"epoch": 0.4632007456866214,
"grad_norm": 6.622848987579346,
"learning_rate": 1e-05,
"loss": 0.7225,
"step": 24350
},
{
"epoch": 0.4633909718655481,
"grad_norm": 1.4858283996582031,
"learning_rate": 1e-05,
"loss": 0.8085,
"step": 24360
},
{
"epoch": 0.4635811980444749,
"grad_norm": 1.6341520547866821,
"learning_rate": 1e-05,
"loss": 0.742,
"step": 24370
},
{
"epoch": 0.4637714242234016,
"grad_norm": 1.9799001216888428,
"learning_rate": 1e-05,
"loss": 0.5981,
"step": 24380
},
{
"epoch": 0.4639616504023284,
"grad_norm": 2.5379247665405273,
"learning_rate": 1e-05,
"loss": 0.5884,
"step": 24390
},
{
"epoch": 0.4641518765812551,
"grad_norm": 7.345506191253662,
"learning_rate": 1e-05,
"loss": 0.7484,
"step": 24400
},
{
"epoch": 0.46434210276018184,
"grad_norm": 1.257645606994629,
"learning_rate": 1e-05,
"loss": 0.7793,
"step": 24410
},
{
"epoch": 0.4645323289391086,
"grad_norm": 1.4955521821975708,
"learning_rate": 1e-05,
"loss": 0.7958,
"step": 24420
},
{
"epoch": 0.46472255511803534,
"grad_norm": 1.7558790445327759,
"learning_rate": 1e-05,
"loss": 0.6161,
"step": 24430
},
{
"epoch": 0.46491278129696206,
"grad_norm": 2.6672909259796143,
"learning_rate": 1e-05,
"loss": 0.6261,
"step": 24440
},
{
"epoch": 0.46510300747588884,
"grad_norm": 6.626492500305176,
"learning_rate": 1e-05,
"loss": 0.7259,
"step": 24450
},
{
"epoch": 0.46529323365481556,
"grad_norm": 1.4956393241882324,
"learning_rate": 1e-05,
"loss": 0.7895,
"step": 24460
},
{
"epoch": 0.46548345983374234,
"grad_norm": 1.7034528255462646,
"learning_rate": 1e-05,
"loss": 0.7915,
"step": 24470
},
{
"epoch": 0.46567368601266906,
"grad_norm": 1.8485451936721802,
"learning_rate": 1e-05,
"loss": 0.6038,
"step": 24480
},
{
"epoch": 0.4658639121915958,
"grad_norm": 2.574549913406372,
"learning_rate": 1e-05,
"loss": 0.6173,
"step": 24490
},
{
"epoch": 0.46605413837052256,
"grad_norm": 7.465858459472656,
"learning_rate": 1e-05,
"loss": 0.731,
"step": 24500
},
{
"epoch": 0.4662443645494493,
"grad_norm": 1.8520907163619995,
"learning_rate": 1e-05,
"loss": 0.7946,
"step": 24510
},
{
"epoch": 0.46643459072837606,
"grad_norm": 2.0016281604766846,
"learning_rate": 1e-05,
"loss": 0.7524,
"step": 24520
},
{
"epoch": 0.4666248169073028,
"grad_norm": 2.0822525024414062,
"learning_rate": 1e-05,
"loss": 0.6286,
"step": 24530
},
{
"epoch": 0.4668150430862295,
"grad_norm": 2.6511826515197754,
"learning_rate": 1e-05,
"loss": 0.5918,
"step": 24540
},
{
"epoch": 0.4670052692651563,
"grad_norm": 5.872143745422363,
"learning_rate": 1e-05,
"loss": 0.6921,
"step": 24550
},
{
"epoch": 0.467195495444083,
"grad_norm": 1.3494436740875244,
"learning_rate": 1e-05,
"loss": 0.7953,
"step": 24560
},
{
"epoch": 0.46738572162300973,
"grad_norm": 1.481616497039795,
"learning_rate": 1e-05,
"loss": 0.7586,
"step": 24570
},
{
"epoch": 0.4675759478019365,
"grad_norm": 1.921842098236084,
"learning_rate": 1e-05,
"loss": 0.6024,
"step": 24580
},
{
"epoch": 0.46776617398086323,
"grad_norm": 2.53204345703125,
"learning_rate": 1e-05,
"loss": 0.5881,
"step": 24590
},
{
"epoch": 0.46795640015979,
"grad_norm": 6.100655555725098,
"learning_rate": 1e-05,
"loss": 0.7231,
"step": 24600
},
{
"epoch": 0.46814662633871673,
"grad_norm": 1.4967926740646362,
"learning_rate": 1e-05,
"loss": 0.8082,
"step": 24610
},
{
"epoch": 0.46833685251764345,
"grad_norm": 1.6108191013336182,
"learning_rate": 1e-05,
"loss": 0.7761,
"step": 24620
},
{
"epoch": 0.46852707869657023,
"grad_norm": 1.97259521484375,
"learning_rate": 1e-05,
"loss": 0.6384,
"step": 24630
},
{
"epoch": 0.46871730487549695,
"grad_norm": 2.352874517440796,
"learning_rate": 1e-05,
"loss": 0.5775,
"step": 24640
},
{
"epoch": 0.46890753105442373,
"grad_norm": 7.416808605194092,
"learning_rate": 1e-05,
"loss": 0.7198,
"step": 24650
},
{
"epoch": 0.46909775723335045,
"grad_norm": 1.6956251859664917,
"learning_rate": 1e-05,
"loss": 0.8133,
"step": 24660
},
{
"epoch": 0.4692879834122772,
"grad_norm": 1.7639371156692505,
"learning_rate": 1e-05,
"loss": 0.7622,
"step": 24670
},
{
"epoch": 0.46947820959120395,
"grad_norm": 2.1144707202911377,
"learning_rate": 1e-05,
"loss": 0.6044,
"step": 24680
},
{
"epoch": 0.4696684357701307,
"grad_norm": 2.719249725341797,
"learning_rate": 1e-05,
"loss": 0.6185,
"step": 24690
},
{
"epoch": 0.46985866194905745,
"grad_norm": 8.062496185302734,
"learning_rate": 1e-05,
"loss": 0.7156,
"step": 24700
},
{
"epoch": 0.4700488881279842,
"grad_norm": 1.5634870529174805,
"learning_rate": 1e-05,
"loss": 0.8168,
"step": 24710
},
{
"epoch": 0.4702391143069109,
"grad_norm": 1.6957346200942993,
"learning_rate": 1e-05,
"loss": 0.7849,
"step": 24720
},
{
"epoch": 0.4704293404858377,
"grad_norm": 2.1655750274658203,
"learning_rate": 1e-05,
"loss": 0.6231,
"step": 24730
},
{
"epoch": 0.4706195666647644,
"grad_norm": 2.843968152999878,
"learning_rate": 1e-05,
"loss": 0.6155,
"step": 24740
},
{
"epoch": 0.4708097928436911,
"grad_norm": 5.857408046722412,
"learning_rate": 1e-05,
"loss": 0.707,
"step": 24750
},
{
"epoch": 0.4710000190226179,
"grad_norm": 1.1980286836624146,
"learning_rate": 1e-05,
"loss": 0.8204,
"step": 24760
},
{
"epoch": 0.4711902452015446,
"grad_norm": 1.6137747764587402,
"learning_rate": 1e-05,
"loss": 0.7626,
"step": 24770
},
{
"epoch": 0.4713804713804714,
"grad_norm": 1.9890812635421753,
"learning_rate": 1e-05,
"loss": 0.6365,
"step": 24780
},
{
"epoch": 0.4715706975593981,
"grad_norm": 2.363523244857788,
"learning_rate": 1e-05,
"loss": 0.5737,
"step": 24790
},
{
"epoch": 0.47176092373832484,
"grad_norm": 5.6321516036987305,
"learning_rate": 1e-05,
"loss": 0.6848,
"step": 24800
},
{
"epoch": 0.4719511499172516,
"grad_norm": 1.341628074645996,
"learning_rate": 1e-05,
"loss": 0.7823,
"step": 24810
},
{
"epoch": 0.47214137609617834,
"grad_norm": 1.5668390989303589,
"learning_rate": 1e-05,
"loss": 0.7511,
"step": 24820
},
{
"epoch": 0.4723316022751051,
"grad_norm": 2.7752296924591064,
"learning_rate": 1e-05,
"loss": 0.6157,
"step": 24830
},
{
"epoch": 0.47252182845403184,
"grad_norm": 2.6439831256866455,
"learning_rate": 1e-05,
"loss": 0.5784,
"step": 24840
},
{
"epoch": 0.47271205463295857,
"grad_norm": 7.426585674285889,
"learning_rate": 1e-05,
"loss": 0.7152,
"step": 24850
},
{
"epoch": 0.47290228081188534,
"grad_norm": 1.7817015647888184,
"learning_rate": 1e-05,
"loss": 0.8172,
"step": 24860
},
{
"epoch": 0.47309250699081207,
"grad_norm": 1.811673879623413,
"learning_rate": 1e-05,
"loss": 0.7643,
"step": 24870
},
{
"epoch": 0.47328273316973885,
"grad_norm": 2.219876766204834,
"learning_rate": 1e-05,
"loss": 0.6595,
"step": 24880
},
{
"epoch": 0.47347295934866557,
"grad_norm": 2.5547797679901123,
"learning_rate": 1e-05,
"loss": 0.5618,
"step": 24890
},
{
"epoch": 0.4736631855275923,
"grad_norm": 5.76616907119751,
"learning_rate": 1e-05,
"loss": 0.681,
"step": 24900
},
{
"epoch": 0.47385341170651907,
"grad_norm": 1.3932499885559082,
"learning_rate": 1e-05,
"loss": 0.798,
"step": 24910
},
{
"epoch": 0.4740436378854458,
"grad_norm": 1.677416443824768,
"learning_rate": 1e-05,
"loss": 0.7876,
"step": 24920
},
{
"epoch": 0.4742338640643725,
"grad_norm": 1.9274797439575195,
"learning_rate": 1e-05,
"loss": 0.6653,
"step": 24930
},
{
"epoch": 0.4744240902432993,
"grad_norm": 2.853856086730957,
"learning_rate": 1e-05,
"loss": 0.6098,
"step": 24940
},
{
"epoch": 0.474614316422226,
"grad_norm": 6.967202186584473,
"learning_rate": 1e-05,
"loss": 0.7139,
"step": 24950
},
{
"epoch": 0.4748045426011528,
"grad_norm": 1.4107186794281006,
"learning_rate": 1e-05,
"loss": 0.8241,
"step": 24960
},
{
"epoch": 0.4749947687800795,
"grad_norm": 1.5797606706619263,
"learning_rate": 1e-05,
"loss": 0.7461,
"step": 24970
},
{
"epoch": 0.47518499495900623,
"grad_norm": 1.8848415613174438,
"learning_rate": 1e-05,
"loss": 0.6207,
"step": 24980
},
{
"epoch": 0.475375221137933,
"grad_norm": 2.5875403881073,
"learning_rate": 1e-05,
"loss": 0.6017,
"step": 24990
},
{
"epoch": 0.47556544731685974,
"grad_norm": 6.65854024887085,
"learning_rate": 1e-05,
"loss": 0.7184,
"step": 25000
},
{
"epoch": 0.4757556734957865,
"grad_norm": 1.3997161388397217,
"learning_rate": 1e-05,
"loss": 0.8136,
"step": 25010
},
{
"epoch": 0.47594589967471324,
"grad_norm": 1.486290693283081,
"learning_rate": 1e-05,
"loss": 0.7167,
"step": 25020
},
{
"epoch": 0.47613612585363996,
"grad_norm": 2.206482172012329,
"learning_rate": 1e-05,
"loss": 0.58,
"step": 25030
},
{
"epoch": 0.47632635203256674,
"grad_norm": 2.526432752609253,
"learning_rate": 1e-05,
"loss": 0.5839,
"step": 25040
},
{
"epoch": 0.47651657821149346,
"grad_norm": 7.106001377105713,
"learning_rate": 1e-05,
"loss": 0.7298,
"step": 25050
},
{
"epoch": 0.47670680439042024,
"grad_norm": 1.397973895072937,
"learning_rate": 1e-05,
"loss": 0.7853,
"step": 25060
},
{
"epoch": 0.47689703056934696,
"grad_norm": 1.4500606060028076,
"learning_rate": 1e-05,
"loss": 0.7151,
"step": 25070
},
{
"epoch": 0.4770872567482737,
"grad_norm": 1.8139972686767578,
"learning_rate": 1e-05,
"loss": 0.5968,
"step": 25080
},
{
"epoch": 0.47727748292720046,
"grad_norm": 2.8111655712127686,
"learning_rate": 1e-05,
"loss": 0.5966,
"step": 25090
},
{
"epoch": 0.4774677091061272,
"grad_norm": 6.381850242614746,
"learning_rate": 1e-05,
"loss": 0.6928,
"step": 25100
},
{
"epoch": 0.4776579352850539,
"grad_norm": 1.2442795038223267,
"learning_rate": 1e-05,
"loss": 0.8155,
"step": 25110
},
{
"epoch": 0.4778481614639807,
"grad_norm": 1.4345344305038452,
"learning_rate": 1e-05,
"loss": 0.7557,
"step": 25120
},
{
"epoch": 0.4780383876429074,
"grad_norm": 1.7890595197677612,
"learning_rate": 1e-05,
"loss": 0.6216,
"step": 25130
},
{
"epoch": 0.4782286138218342,
"grad_norm": 2.9044179916381836,
"learning_rate": 1e-05,
"loss": 0.5896,
"step": 25140
},
{
"epoch": 0.4784188400007609,
"grad_norm": 6.361065864562988,
"learning_rate": 1e-05,
"loss": 0.6764,
"step": 25150
},
{
"epoch": 0.4786090661796876,
"grad_norm": 1.385246992111206,
"learning_rate": 1e-05,
"loss": 0.797,
"step": 25160
},
{
"epoch": 0.4787992923586144,
"grad_norm": 1.622786283493042,
"learning_rate": 1e-05,
"loss": 0.7581,
"step": 25170
},
{
"epoch": 0.4789895185375411,
"grad_norm": 1.8420246839523315,
"learning_rate": 1e-05,
"loss": 0.6032,
"step": 25180
},
{
"epoch": 0.4791797447164679,
"grad_norm": 3.0507521629333496,
"learning_rate": 1e-05,
"loss": 0.5636,
"step": 25190
},
{
"epoch": 0.4793699708953946,
"grad_norm": 6.347072601318359,
"learning_rate": 1e-05,
"loss": 0.73,
"step": 25200
},
{
"epoch": 0.47956019707432135,
"grad_norm": 1.2836600542068481,
"learning_rate": 1e-05,
"loss": 0.7922,
"step": 25210
},
{
"epoch": 0.4797504232532481,
"grad_norm": 1.6343671083450317,
"learning_rate": 1e-05,
"loss": 0.7452,
"step": 25220
},
{
"epoch": 0.47994064943217485,
"grad_norm": 2.178748846054077,
"learning_rate": 1e-05,
"loss": 0.6283,
"step": 25230
},
{
"epoch": 0.4801308756111016,
"grad_norm": 3.1259818077087402,
"learning_rate": 1e-05,
"loss": 0.6148,
"step": 25240
},
{
"epoch": 0.48032110179002835,
"grad_norm": 6.1668877601623535,
"learning_rate": 1e-05,
"loss": 0.7429,
"step": 25250
},
{
"epoch": 0.48051132796895507,
"grad_norm": 1.6452568769454956,
"learning_rate": 1e-05,
"loss": 0.8344,
"step": 25260
},
{
"epoch": 0.48070155414788185,
"grad_norm": 1.728680968284607,
"learning_rate": 1e-05,
"loss": 0.764,
"step": 25270
},
{
"epoch": 0.48089178032680857,
"grad_norm": 1.8995915651321411,
"learning_rate": 1e-05,
"loss": 0.61,
"step": 25280
},
{
"epoch": 0.4810820065057353,
"grad_norm": 2.302121162414551,
"learning_rate": 1e-05,
"loss": 0.5881,
"step": 25290
},
{
"epoch": 0.48127223268466207,
"grad_norm": 5.718452453613281,
"learning_rate": 1e-05,
"loss": 0.6793,
"step": 25300
},
{
"epoch": 0.4814624588635888,
"grad_norm": 1.3672212362289429,
"learning_rate": 1e-05,
"loss": 0.8101,
"step": 25310
},
{
"epoch": 0.48165268504251557,
"grad_norm": 1.5332621335983276,
"learning_rate": 1e-05,
"loss": 0.7735,
"step": 25320
},
{
"epoch": 0.4818429112214423,
"grad_norm": 2.7715306282043457,
"learning_rate": 1e-05,
"loss": 0.6226,
"step": 25330
},
{
"epoch": 0.482033137400369,
"grad_norm": 2.5493557453155518,
"learning_rate": 1e-05,
"loss": 0.6004,
"step": 25340
},
{
"epoch": 0.4822233635792958,
"grad_norm": 6.952243804931641,
"learning_rate": 1e-05,
"loss": 0.7326,
"step": 25350
},
{
"epoch": 0.4824135897582225,
"grad_norm": 1.6663097143173218,
"learning_rate": 1e-05,
"loss": 0.8034,
"step": 25360
},
{
"epoch": 0.4826038159371493,
"grad_norm": 1.616231918334961,
"learning_rate": 1e-05,
"loss": 0.7415,
"step": 25370
},
{
"epoch": 0.482794042116076,
"grad_norm": 1.9222723245620728,
"learning_rate": 1e-05,
"loss": 0.5971,
"step": 25380
},
{
"epoch": 0.48298426829500274,
"grad_norm": 2.5769152641296387,
"learning_rate": 1e-05,
"loss": 0.6171,
"step": 25390
},
{
"epoch": 0.4831744944739295,
"grad_norm": 7.6587958335876465,
"learning_rate": 1e-05,
"loss": 0.7103,
"step": 25400
},
{
"epoch": 0.48336472065285624,
"grad_norm": 1.6322726011276245,
"learning_rate": 1e-05,
"loss": 0.8092,
"step": 25410
},
{
"epoch": 0.483554946831783,
"grad_norm": 1.6653474569320679,
"learning_rate": 1e-05,
"loss": 0.7956,
"step": 25420
},
{
"epoch": 0.48374517301070974,
"grad_norm": 1.968806266784668,
"learning_rate": 1e-05,
"loss": 0.6469,
"step": 25430
},
{
"epoch": 0.48393539918963646,
"grad_norm": 2.7097020149230957,
"learning_rate": 1e-05,
"loss": 0.6219,
"step": 25440
},
{
"epoch": 0.48412562536856324,
"grad_norm": 7.540827751159668,
"learning_rate": 1e-05,
"loss": 0.7216,
"step": 25450
},
{
"epoch": 0.48431585154748996,
"grad_norm": 1.3733558654785156,
"learning_rate": 1e-05,
"loss": 0.8021,
"step": 25460
},
{
"epoch": 0.4845060777264167,
"grad_norm": 2.147263765335083,
"learning_rate": 1e-05,
"loss": 0.7461,
"step": 25470
},
{
"epoch": 0.48469630390534346,
"grad_norm": 2.248168468475342,
"learning_rate": 1e-05,
"loss": 0.5995,
"step": 25480
},
{
"epoch": 0.4848865300842702,
"grad_norm": 2.6114723682403564,
"learning_rate": 1e-05,
"loss": 0.5679,
"step": 25490
},
{
"epoch": 0.48507675626319696,
"grad_norm": 6.517694473266602,
"learning_rate": 1e-05,
"loss": 0.7132,
"step": 25500
},
{
"epoch": 0.4852669824421237,
"grad_norm": 1.6420058012008667,
"learning_rate": 1e-05,
"loss": 0.8246,
"step": 25510
},
{
"epoch": 0.4854572086210504,
"grad_norm": 1.7278486490249634,
"learning_rate": 1e-05,
"loss": 0.7459,
"step": 25520
},
{
"epoch": 0.4856474347999772,
"grad_norm": 1.928666353225708,
"learning_rate": 1e-05,
"loss": 0.6095,
"step": 25530
},
{
"epoch": 0.4858376609789039,
"grad_norm": 2.6522419452667236,
"learning_rate": 1e-05,
"loss": 0.6203,
"step": 25540
},
{
"epoch": 0.4860278871578307,
"grad_norm": 7.48252010345459,
"learning_rate": 1e-05,
"loss": 0.6998,
"step": 25550
},
{
"epoch": 0.4862181133367574,
"grad_norm": 1.3695486783981323,
"learning_rate": 1e-05,
"loss": 0.7952,
"step": 25560
},
{
"epoch": 0.48640833951568413,
"grad_norm": 1.5086663961410522,
"learning_rate": 1e-05,
"loss": 0.7583,
"step": 25570
},
{
"epoch": 0.4865985656946109,
"grad_norm": 1.9904680252075195,
"learning_rate": 1e-05,
"loss": 0.5771,
"step": 25580
},
{
"epoch": 0.48678879187353763,
"grad_norm": 2.377136468887329,
"learning_rate": 1e-05,
"loss": 0.5724,
"step": 25590
},
{
"epoch": 0.4869790180524644,
"grad_norm": 6.3017473220825195,
"learning_rate": 1e-05,
"loss": 0.7235,
"step": 25600
},
{
"epoch": 0.48716924423139113,
"grad_norm": 1.3907005786895752,
"learning_rate": 1e-05,
"loss": 0.7998,
"step": 25610
},
{
"epoch": 0.48735947041031785,
"grad_norm": 1.5196592807769775,
"learning_rate": 1e-05,
"loss": 0.7551,
"step": 25620
},
{
"epoch": 0.48754969658924463,
"grad_norm": 2.0708792209625244,
"learning_rate": 1e-05,
"loss": 0.5688,
"step": 25630
},
{
"epoch": 0.48773992276817135,
"grad_norm": 2.8068413734436035,
"learning_rate": 1e-05,
"loss": 0.5794,
"step": 25640
},
{
"epoch": 0.4879301489470981,
"grad_norm": 7.299570560455322,
"learning_rate": 1e-05,
"loss": 0.7296,
"step": 25650
},
{
"epoch": 0.48812037512602485,
"grad_norm": 1.3350303173065186,
"learning_rate": 1e-05,
"loss": 0.7957,
"step": 25660
},
{
"epoch": 0.4883106013049516,
"grad_norm": 1.5439625978469849,
"learning_rate": 1e-05,
"loss": 0.731,
"step": 25670
},
{
"epoch": 0.48850082748387835,
"grad_norm": 2.0365967750549316,
"learning_rate": 1e-05,
"loss": 0.5883,
"step": 25680
},
{
"epoch": 0.4886910536628051,
"grad_norm": 2.5288565158843994,
"learning_rate": 1e-05,
"loss": 0.5814,
"step": 25690
},
{
"epoch": 0.4888812798417318,
"grad_norm": 7.27092170715332,
"learning_rate": 1e-05,
"loss": 0.7028,
"step": 25700
},
{
"epoch": 0.4890715060206586,
"grad_norm": 1.5296874046325684,
"learning_rate": 1e-05,
"loss": 0.7922,
"step": 25710
},
{
"epoch": 0.4892617321995853,
"grad_norm": 1.5780446529388428,
"learning_rate": 1e-05,
"loss": 0.7573,
"step": 25720
},
{
"epoch": 0.4894519583785121,
"grad_norm": 2.3514628410339355,
"learning_rate": 1e-05,
"loss": 0.6341,
"step": 25730
},
{
"epoch": 0.4896421845574388,
"grad_norm": 2.4253222942352295,
"learning_rate": 1e-05,
"loss": 0.5814,
"step": 25740
},
{
"epoch": 0.4898324107363655,
"grad_norm": 6.163158893585205,
"learning_rate": 1e-05,
"loss": 0.7399,
"step": 25750
},
{
"epoch": 0.4900226369152923,
"grad_norm": 1.1654995679855347,
"learning_rate": 1e-05,
"loss": 0.8041,
"step": 25760
},
{
"epoch": 0.490212863094219,
"grad_norm": 1.6642873287200928,
"learning_rate": 1e-05,
"loss": 0.7849,
"step": 25770
},
{
"epoch": 0.4904030892731458,
"grad_norm": 1.7302231788635254,
"learning_rate": 1e-05,
"loss": 0.6281,
"step": 25780
},
{
"epoch": 0.4905933154520725,
"grad_norm": 2.697882652282715,
"learning_rate": 1e-05,
"loss": 0.5876,
"step": 25790
},
{
"epoch": 0.49078354163099924,
"grad_norm": 6.980758190155029,
"learning_rate": 1e-05,
"loss": 0.7138,
"step": 25800
},
{
"epoch": 0.490973767809926,
"grad_norm": 1.4509748220443726,
"learning_rate": 1e-05,
"loss": 0.8098,
"step": 25810
},
{
"epoch": 0.49116399398885274,
"grad_norm": 1.5892177820205688,
"learning_rate": 1e-05,
"loss": 0.7664,
"step": 25820
},
{
"epoch": 0.49135422016777947,
"grad_norm": 1.837883472442627,
"learning_rate": 1e-05,
"loss": 0.6103,
"step": 25830
},
{
"epoch": 0.49154444634670624,
"grad_norm": 2.920602798461914,
"learning_rate": 1e-05,
"loss": 0.6131,
"step": 25840
},
{
"epoch": 0.49173467252563297,
"grad_norm": 7.1568379402160645,
"learning_rate": 1e-05,
"loss": 0.7428,
"step": 25850
},
{
"epoch": 0.49192489870455974,
"grad_norm": 1.4557758569717407,
"learning_rate": 1e-05,
"loss": 0.8294,
"step": 25860
},
{
"epoch": 0.49211512488348647,
"grad_norm": 1.7474267482757568,
"learning_rate": 1e-05,
"loss": 0.7394,
"step": 25870
},
{
"epoch": 0.4923053510624132,
"grad_norm": 2.0510482788085938,
"learning_rate": 1e-05,
"loss": 0.6034,
"step": 25880
},
{
"epoch": 0.49249557724133997,
"grad_norm": 2.5900800228118896,
"learning_rate": 1e-05,
"loss": 0.5847,
"step": 25890
},
{
"epoch": 0.4926858034202667,
"grad_norm": 6.460900783538818,
"learning_rate": 1e-05,
"loss": 0.6996,
"step": 25900
},
{
"epoch": 0.49287602959919347,
"grad_norm": 1.2711211442947388,
"learning_rate": 1e-05,
"loss": 0.7865,
"step": 25910
},
{
"epoch": 0.4930662557781202,
"grad_norm": 1.7260485887527466,
"learning_rate": 1e-05,
"loss": 0.7696,
"step": 25920
},
{
"epoch": 0.4932564819570469,
"grad_norm": 2.494281053543091,
"learning_rate": 1e-05,
"loss": 0.5973,
"step": 25930
},
{
"epoch": 0.4934467081359737,
"grad_norm": 2.5816214084625244,
"learning_rate": 1e-05,
"loss": 0.6255,
"step": 25940
},
{
"epoch": 0.4936369343149004,
"grad_norm": 6.672122001647949,
"learning_rate": 1e-05,
"loss": 0.7298,
"step": 25950
},
{
"epoch": 0.49382716049382713,
"grad_norm": 1.522822380065918,
"learning_rate": 1e-05,
"loss": 0.7936,
"step": 25960
},
{
"epoch": 0.4940173866727539,
"grad_norm": 1.690618872642517,
"learning_rate": 1e-05,
"loss": 0.7752,
"step": 25970
},
{
"epoch": 0.49420761285168063,
"grad_norm": 1.8049094676971436,
"learning_rate": 1e-05,
"loss": 0.596,
"step": 25980
},
{
"epoch": 0.4943978390306074,
"grad_norm": 2.081824541091919,
"learning_rate": 1e-05,
"loss": 0.5504,
"step": 25990
},
{
"epoch": 0.49458806520953413,
"grad_norm": 6.052360534667969,
"learning_rate": 1e-05,
"loss": 0.7111,
"step": 26000
},
{
"epoch": 0.49477829138846086,
"grad_norm": 1.48108971118927,
"learning_rate": 1e-05,
"loss": 0.8027,
"step": 26010
},
{
"epoch": 0.49496851756738763,
"grad_norm": 1.8303556442260742,
"learning_rate": 1e-05,
"loss": 0.766,
"step": 26020
},
{
"epoch": 0.49515874374631436,
"grad_norm": 2.2966580390930176,
"learning_rate": 1e-05,
"loss": 0.6017,
"step": 26030
},
{
"epoch": 0.49534896992524113,
"grad_norm": 2.715561866760254,
"learning_rate": 1e-05,
"loss": 0.6268,
"step": 26040
},
{
"epoch": 0.49553919610416786,
"grad_norm": 6.248926162719727,
"learning_rate": 1e-05,
"loss": 0.757,
"step": 26050
},
{
"epoch": 0.4957294222830946,
"grad_norm": 1.357933759689331,
"learning_rate": 1e-05,
"loss": 0.8051,
"step": 26060
},
{
"epoch": 0.49591964846202136,
"grad_norm": 1.6126124858856201,
"learning_rate": 1e-05,
"loss": 0.7506,
"step": 26070
},
{
"epoch": 0.4961098746409481,
"grad_norm": 2.367105484008789,
"learning_rate": 1e-05,
"loss": 0.6157,
"step": 26080
},
{
"epoch": 0.49630010081987486,
"grad_norm": 2.717240571975708,
"learning_rate": 1e-05,
"loss": 0.5922,
"step": 26090
},
{
"epoch": 0.4964903269988016,
"grad_norm": 6.814499855041504,
"learning_rate": 1e-05,
"loss": 0.7035,
"step": 26100
},
{
"epoch": 0.4966805531777283,
"grad_norm": 1.5942243337631226,
"learning_rate": 1e-05,
"loss": 0.8056,
"step": 26110
},
{
"epoch": 0.4968707793566551,
"grad_norm": 1.5243202447891235,
"learning_rate": 1e-05,
"loss": 0.7321,
"step": 26120
},
{
"epoch": 0.4970610055355818,
"grad_norm": 1.956007957458496,
"learning_rate": 1e-05,
"loss": 0.5837,
"step": 26130
},
{
"epoch": 0.4972512317145085,
"grad_norm": 2.8430559635162354,
"learning_rate": 1e-05,
"loss": 0.6095,
"step": 26140
},
{
"epoch": 0.4974414578934353,
"grad_norm": 6.209444046020508,
"learning_rate": 1e-05,
"loss": 0.6888,
"step": 26150
},
{
"epoch": 0.497631684072362,
"grad_norm": 1.3413457870483398,
"learning_rate": 1e-05,
"loss": 0.7843,
"step": 26160
},
{
"epoch": 0.4978219102512888,
"grad_norm": 1.4749476909637451,
"learning_rate": 1e-05,
"loss": 0.7513,
"step": 26170
},
{
"epoch": 0.4980121364302155,
"grad_norm": 2.261862277984619,
"learning_rate": 1e-05,
"loss": 0.5725,
"step": 26180
},
{
"epoch": 0.49820236260914225,
"grad_norm": 2.2241382598876953,
"learning_rate": 1e-05,
"loss": 0.6044,
"step": 26190
},
{
"epoch": 0.498392588788069,
"grad_norm": 8.47707748413086,
"learning_rate": 1e-05,
"loss": 0.7779,
"step": 26200
},
{
"epoch": 0.49858281496699575,
"grad_norm": 1.810894250869751,
"learning_rate": 1e-05,
"loss": 0.7907,
"step": 26210
},
{
"epoch": 0.4987730411459225,
"grad_norm": 1.6950432062149048,
"learning_rate": 1e-05,
"loss": 0.7589,
"step": 26220
},
{
"epoch": 0.49896326732484925,
"grad_norm": 2.165290117263794,
"learning_rate": 1e-05,
"loss": 0.6089,
"step": 26230
},
{
"epoch": 0.49915349350377597,
"grad_norm": 2.5153884887695312,
"learning_rate": 1e-05,
"loss": 0.5537,
"step": 26240
},
{
"epoch": 0.49934371968270275,
"grad_norm": 6.540546417236328,
"learning_rate": 1e-05,
"loss": 0.7066,
"step": 26250
},
{
"epoch": 0.49953394586162947,
"grad_norm": 1.8814890384674072,
"learning_rate": 1e-05,
"loss": 0.7931,
"step": 26260
},
{
"epoch": 0.49972417204055625,
"grad_norm": 1.8376352787017822,
"learning_rate": 1e-05,
"loss": 0.7613,
"step": 26270
},
{
"epoch": 0.49991439821948297,
"grad_norm": 2.261735200881958,
"learning_rate": 1e-05,
"loss": 0.6134,
"step": 26280
},
{
"epoch": 0.5001046243984097,
"grad_norm": 2.868830919265747,
"learning_rate": 1e-05,
"loss": 0.5735,
"step": 26290
},
{
"epoch": 0.5002948505773365,
"grad_norm": 7.201417922973633,
"learning_rate": 1e-05,
"loss": 0.7119,
"step": 26300
},
{
"epoch": 0.5004850767562632,
"grad_norm": 2.9679315090179443,
"learning_rate": 1e-05,
"loss": 0.7942,
"step": 26310
},
{
"epoch": 0.5006753029351899,
"grad_norm": 1.9978197813034058,
"learning_rate": 1e-05,
"loss": 0.7696,
"step": 26320
},
{
"epoch": 0.5008655291141166,
"grad_norm": 2.186264991760254,
"learning_rate": 1e-05,
"loss": 0.6325,
"step": 26330
},
{
"epoch": 0.5010557552930435,
"grad_norm": 2.630091905593872,
"learning_rate": 1e-05,
"loss": 0.5993,
"step": 26340
},
{
"epoch": 0.5012459814719702,
"grad_norm": 6.750666618347168,
"learning_rate": 1e-05,
"loss": 0.6829,
"step": 26350
},
{
"epoch": 0.5014362076508969,
"grad_norm": 1.3985573053359985,
"learning_rate": 1e-05,
"loss": 0.8016,
"step": 26360
},
{
"epoch": 0.5016264338298236,
"grad_norm": 1.584041714668274,
"learning_rate": 1e-05,
"loss": 0.7455,
"step": 26370
},
{
"epoch": 0.5018166600087504,
"grad_norm": 1.7916698455810547,
"learning_rate": 1e-05,
"loss": 0.5937,
"step": 26380
},
{
"epoch": 0.5020068861876772,
"grad_norm": 2.706805944442749,
"learning_rate": 1e-05,
"loss": 0.593,
"step": 26390
},
{
"epoch": 0.5021971123666039,
"grad_norm": 6.90195894241333,
"learning_rate": 1e-05,
"loss": 0.6856,
"step": 26400
},
{
"epoch": 0.5023873385455306,
"grad_norm": 1.4350513219833374,
"learning_rate": 1e-05,
"loss": 0.8175,
"step": 26410
},
{
"epoch": 0.5025775647244574,
"grad_norm": 1.7008193731307983,
"learning_rate": 1e-05,
"loss": 0.706,
"step": 26420
},
{
"epoch": 0.5027677909033841,
"grad_norm": 1.8652887344360352,
"learning_rate": 1e-05,
"loss": 0.5729,
"step": 26430
},
{
"epoch": 0.5029580170823109,
"grad_norm": 3.1122312545776367,
"learning_rate": 1e-05,
"loss": 0.5739,
"step": 26440
},
{
"epoch": 0.5031482432612376,
"grad_norm": 6.325319290161133,
"learning_rate": 1e-05,
"loss": 0.7145,
"step": 26450
},
{
"epoch": 0.5033384694401644,
"grad_norm": 1.5242714881896973,
"learning_rate": 1e-05,
"loss": 0.8009,
"step": 26460
},
{
"epoch": 0.5035286956190911,
"grad_norm": 1.502831220626831,
"learning_rate": 1e-05,
"loss": 0.7324,
"step": 26470
},
{
"epoch": 0.5037189217980178,
"grad_norm": 1.8909034729003906,
"learning_rate": 1e-05,
"loss": 0.5917,
"step": 26480
},
{
"epoch": 0.5039091479769446,
"grad_norm": 2.689692497253418,
"learning_rate": 1e-05,
"loss": 0.6235,
"step": 26490
},
{
"epoch": 0.5040993741558714,
"grad_norm": 6.4771528244018555,
"learning_rate": 1e-05,
"loss": 0.7342,
"step": 26500
},
{
"epoch": 0.5042896003347981,
"grad_norm": 1.348188877105713,
"learning_rate": 1e-05,
"loss": 0.8088,
"step": 26510
},
{
"epoch": 0.5044798265137248,
"grad_norm": 1.8132826089859009,
"learning_rate": 1e-05,
"loss": 0.7372,
"step": 26520
},
{
"epoch": 0.5046700526926515,
"grad_norm": 2.127190589904785,
"learning_rate": 1e-05,
"loss": 0.6067,
"step": 26530
},
{
"epoch": 0.5048602788715784,
"grad_norm": 2.6336286067962646,
"learning_rate": 1e-05,
"loss": 0.5653,
"step": 26540
},
{
"epoch": 0.5050505050505051,
"grad_norm": 6.656616687774658,
"learning_rate": 1e-05,
"loss": 0.7127,
"step": 26550
},
{
"epoch": 0.5052407312294318,
"grad_norm": 1.653061032295227,
"learning_rate": 1e-05,
"loss": 0.7943,
"step": 26560
},
{
"epoch": 0.5054309574083585,
"grad_norm": 1.4513689279556274,
"learning_rate": 1e-05,
"loss": 0.7295,
"step": 26570
},
{
"epoch": 0.5056211835872853,
"grad_norm": 2.0724902153015137,
"learning_rate": 1e-05,
"loss": 0.6216,
"step": 26580
},
{
"epoch": 0.505811409766212,
"grad_norm": 2.418262004852295,
"learning_rate": 1e-05,
"loss": 0.5855,
"step": 26590
},
{
"epoch": 0.5060016359451388,
"grad_norm": 6.561469078063965,
"learning_rate": 1e-05,
"loss": 0.7375,
"step": 26600
},
{
"epoch": 0.5061918621240655,
"grad_norm": 1.6499059200286865,
"learning_rate": 1e-05,
"loss": 0.8205,
"step": 26610
},
{
"epoch": 0.5063820883029923,
"grad_norm": 1.6943861246109009,
"learning_rate": 1e-05,
"loss": 0.7614,
"step": 26620
},
{
"epoch": 0.506572314481919,
"grad_norm": 2.216977596282959,
"learning_rate": 1e-05,
"loss": 0.5906,
"step": 26630
},
{
"epoch": 0.5067625406608457,
"grad_norm": 2.4687340259552,
"learning_rate": 1e-05,
"loss": 0.5722,
"step": 26640
},
{
"epoch": 0.5069527668397725,
"grad_norm": 6.7068095207214355,
"learning_rate": 1e-05,
"loss": 0.7114,
"step": 26650
},
{
"epoch": 0.5071429930186993,
"grad_norm": 1.319226861000061,
"learning_rate": 1e-05,
"loss": 0.8044,
"step": 26660
},
{
"epoch": 0.507333219197626,
"grad_norm": 1.6044725179672241,
"learning_rate": 1e-05,
"loss": 0.7156,
"step": 26670
},
{
"epoch": 0.5075234453765527,
"grad_norm": 1.9069814682006836,
"learning_rate": 1e-05,
"loss": 0.6308,
"step": 26680
},
{
"epoch": 0.5077136715554794,
"grad_norm": 2.634343147277832,
"learning_rate": 1e-05,
"loss": 0.6058,
"step": 26690
},
{
"epoch": 0.5079038977344063,
"grad_norm": 7.282718658447266,
"learning_rate": 1e-05,
"loss": 0.738,
"step": 26700
},
{
"epoch": 0.508094123913333,
"grad_norm": 1.279189944267273,
"learning_rate": 1e-05,
"loss": 0.7926,
"step": 26710
},
{
"epoch": 0.5082843500922597,
"grad_norm": 1.5796113014221191,
"learning_rate": 1e-05,
"loss": 0.7845,
"step": 26720
},
{
"epoch": 0.5084745762711864,
"grad_norm": 1.8787798881530762,
"learning_rate": 1e-05,
"loss": 0.6195,
"step": 26730
},
{
"epoch": 0.5086648024501131,
"grad_norm": 2.8469300270080566,
"learning_rate": 1e-05,
"loss": 0.5805,
"step": 26740
},
{
"epoch": 0.50885502862904,
"grad_norm": 5.278876304626465,
"learning_rate": 1e-05,
"loss": 0.6918,
"step": 26750
},
{
"epoch": 0.5090452548079667,
"grad_norm": 1.2685539722442627,
"learning_rate": 1e-05,
"loss": 0.8069,
"step": 26760
},
{
"epoch": 0.5092354809868934,
"grad_norm": 1.4827028512954712,
"learning_rate": 1e-05,
"loss": 0.745,
"step": 26770
},
{
"epoch": 0.5094257071658201,
"grad_norm": 1.82663094997406,
"learning_rate": 1e-05,
"loss": 0.591,
"step": 26780
},
{
"epoch": 0.5096159333447469,
"grad_norm": 2.636378765106201,
"learning_rate": 1e-05,
"loss": 0.5805,
"step": 26790
},
{
"epoch": 0.5098061595236737,
"grad_norm": 6.358584880828857,
"learning_rate": 1e-05,
"loss": 0.7202,
"step": 26800
},
{
"epoch": 0.5099963857026004,
"grad_norm": 1.4085490703582764,
"learning_rate": 1e-05,
"loss": 0.8086,
"step": 26810
},
{
"epoch": 0.5101866118815271,
"grad_norm": 1.5895555019378662,
"learning_rate": 1e-05,
"loss": 0.7179,
"step": 26820
},
{
"epoch": 0.5103768380604539,
"grad_norm": 2.108654737472534,
"learning_rate": 1e-05,
"loss": 0.6136,
"step": 26830
},
{
"epoch": 0.5105670642393806,
"grad_norm": 2.4762609004974365,
"learning_rate": 1e-05,
"loss": 0.5918,
"step": 26840
},
{
"epoch": 0.5107572904183074,
"grad_norm": 8.02391529083252,
"learning_rate": 1e-05,
"loss": 0.7235,
"step": 26850
},
{
"epoch": 0.5109475165972341,
"grad_norm": 1.5284092426300049,
"learning_rate": 1e-05,
"loss": 0.8001,
"step": 26860
},
{
"epoch": 0.5111377427761609,
"grad_norm": 1.51941978931427,
"learning_rate": 1e-05,
"loss": 0.7877,
"step": 26870
},
{
"epoch": 0.5113279689550876,
"grad_norm": 2.0881569385528564,
"learning_rate": 1e-05,
"loss": 0.6336,
"step": 26880
},
{
"epoch": 0.5115181951340143,
"grad_norm": 2.3765969276428223,
"learning_rate": 1e-05,
"loss": 0.5844,
"step": 26890
},
{
"epoch": 0.5117084213129411,
"grad_norm": 6.924647331237793,
"learning_rate": 1e-05,
"loss": 0.6925,
"step": 26900
},
{
"epoch": 0.5118986474918679,
"grad_norm": 1.3793188333511353,
"learning_rate": 1e-05,
"loss": 0.7947,
"step": 26910
},
{
"epoch": 0.5120888736707946,
"grad_norm": 1.4719302654266357,
"learning_rate": 1e-05,
"loss": 0.7731,
"step": 26920
},
{
"epoch": 0.5122790998497213,
"grad_norm": 1.8981635570526123,
"learning_rate": 1e-05,
"loss": 0.6076,
"step": 26930
},
{
"epoch": 0.512469326028648,
"grad_norm": 2.526689291000366,
"learning_rate": 1e-05,
"loss": 0.5831,
"step": 26940
},
{
"epoch": 0.5126595522075748,
"grad_norm": 5.960160255432129,
"learning_rate": 1e-05,
"loss": 0.7166,
"step": 26950
},
{
"epoch": 0.5128497783865016,
"grad_norm": 1.2402483224868774,
"learning_rate": 1e-05,
"loss": 0.795,
"step": 26960
},
{
"epoch": 0.5130400045654283,
"grad_norm": 1.534856915473938,
"learning_rate": 1e-05,
"loss": 0.7512,
"step": 26970
},
{
"epoch": 0.513230230744355,
"grad_norm": 2.110203981399536,
"learning_rate": 1e-05,
"loss": 0.639,
"step": 26980
},
{
"epoch": 0.5134204569232818,
"grad_norm": 2.555514335632324,
"learning_rate": 1e-05,
"loss": 0.5796,
"step": 26990
},
{
"epoch": 0.5136106831022085,
"grad_norm": 6.482339859008789,
"learning_rate": 1e-05,
"loss": 0.7479,
"step": 27000
},
{
"epoch": 0.5138009092811353,
"grad_norm": 1.2397505044937134,
"learning_rate": 1e-05,
"loss": 0.8022,
"step": 27010
},
{
"epoch": 0.513991135460062,
"grad_norm": 1.5713386535644531,
"learning_rate": 1e-05,
"loss": 0.7579,
"step": 27020
},
{
"epoch": 0.5141813616389888,
"grad_norm": 1.8647968769073486,
"learning_rate": 1e-05,
"loss": 0.5756,
"step": 27030
},
{
"epoch": 0.5143715878179155,
"grad_norm": 2.618239402770996,
"learning_rate": 1e-05,
"loss": 0.586,
"step": 27040
},
{
"epoch": 0.5145618139968422,
"grad_norm": 6.593390941619873,
"learning_rate": 1e-05,
"loss": 0.7243,
"step": 27050
},
{
"epoch": 0.514752040175769,
"grad_norm": 1.3481703996658325,
"learning_rate": 1e-05,
"loss": 0.7883,
"step": 27060
},
{
"epoch": 0.5149422663546958,
"grad_norm": 1.5756845474243164,
"learning_rate": 1e-05,
"loss": 0.7464,
"step": 27070
},
{
"epoch": 0.5151324925336225,
"grad_norm": 1.934828281402588,
"learning_rate": 1e-05,
"loss": 0.6301,
"step": 27080
},
{
"epoch": 0.5153227187125492,
"grad_norm": 3.1743288040161133,
"learning_rate": 1e-05,
"loss": 0.604,
"step": 27090
},
{
"epoch": 0.5155129448914759,
"grad_norm": 7.107223987579346,
"learning_rate": 1e-05,
"loss": 0.6882,
"step": 27100
},
{
"epoch": 0.5157031710704028,
"grad_norm": 1.3347837924957275,
"learning_rate": 1e-05,
"loss": 0.7995,
"step": 27110
},
{
"epoch": 0.5158933972493295,
"grad_norm": 1.679954171180725,
"learning_rate": 1e-05,
"loss": 0.7452,
"step": 27120
},
{
"epoch": 0.5160836234282562,
"grad_norm": 2.0433168411254883,
"learning_rate": 1e-05,
"loss": 0.5876,
"step": 27130
},
{
"epoch": 0.5162738496071829,
"grad_norm": 2.7278661727905273,
"learning_rate": 1e-05,
"loss": 0.5765,
"step": 27140
},
{
"epoch": 0.5164640757861096,
"grad_norm": 6.587141036987305,
"learning_rate": 1e-05,
"loss": 0.6922,
"step": 27150
},
{
"epoch": 0.5166543019650365,
"grad_norm": 1.6837540864944458,
"learning_rate": 1e-05,
"loss": 0.8,
"step": 27160
},
{
"epoch": 0.5168445281439632,
"grad_norm": 1.9510124921798706,
"learning_rate": 1e-05,
"loss": 0.7427,
"step": 27170
},
{
"epoch": 0.5170347543228899,
"grad_norm": 2.3558545112609863,
"learning_rate": 1e-05,
"loss": 0.6454,
"step": 27180
},
{
"epoch": 0.5172249805018166,
"grad_norm": 2.728874444961548,
"learning_rate": 1e-05,
"loss": 0.5709,
"step": 27190
},
{
"epoch": 0.5174152066807434,
"grad_norm": 7.075480937957764,
"learning_rate": 1e-05,
"loss": 0.7133,
"step": 27200
},
{
"epoch": 0.5176054328596702,
"grad_norm": 1.4894939661026,
"learning_rate": 1e-05,
"loss": 0.8027,
"step": 27210
},
{
"epoch": 0.5177956590385969,
"grad_norm": 1.4669134616851807,
"learning_rate": 1e-05,
"loss": 0.7053,
"step": 27220
},
{
"epoch": 0.5179858852175236,
"grad_norm": 2.2579185962677,
"learning_rate": 1e-05,
"loss": 0.5656,
"step": 27230
},
{
"epoch": 0.5181761113964504,
"grad_norm": 2.7445151805877686,
"learning_rate": 1e-05,
"loss": 0.5868,
"step": 27240
},
{
"epoch": 0.5183663375753771,
"grad_norm": 6.485048770904541,
"learning_rate": 1e-05,
"loss": 0.6835,
"step": 27250
},
{
"epoch": 0.5185565637543038,
"grad_norm": 1.2838879823684692,
"learning_rate": 1e-05,
"loss": 0.8109,
"step": 27260
},
{
"epoch": 0.5187467899332306,
"grad_norm": 1.4693782329559326,
"learning_rate": 1e-05,
"loss": 0.7378,
"step": 27270
},
{
"epoch": 0.5189370161121574,
"grad_norm": 1.8807647228240967,
"learning_rate": 1e-05,
"loss": 0.5801,
"step": 27280
},
{
"epoch": 0.5191272422910841,
"grad_norm": 2.7925591468811035,
"learning_rate": 1e-05,
"loss": 0.5779,
"step": 27290
},
{
"epoch": 0.5193174684700108,
"grad_norm": 7.334691524505615,
"learning_rate": 1e-05,
"loss": 0.7173,
"step": 27300
},
{
"epoch": 0.5195076946489375,
"grad_norm": 1.4690724611282349,
"learning_rate": 1e-05,
"loss": 0.7852,
"step": 27310
},
{
"epoch": 0.5196979208278644,
"grad_norm": 1.6412367820739746,
"learning_rate": 1e-05,
"loss": 0.7446,
"step": 27320
},
{
"epoch": 0.5198881470067911,
"grad_norm": 2.0946903228759766,
"learning_rate": 1e-05,
"loss": 0.5961,
"step": 27330
},
{
"epoch": 0.5200783731857178,
"grad_norm": 2.699122905731201,
"learning_rate": 1e-05,
"loss": 0.5874,
"step": 27340
},
{
"epoch": 0.5202685993646445,
"grad_norm": 6.867517471313477,
"learning_rate": 1e-05,
"loss": 0.7326,
"step": 27350
},
{
"epoch": 0.5204588255435713,
"grad_norm": 1.4884153604507446,
"learning_rate": 1e-05,
"loss": 0.8042,
"step": 27360
},
{
"epoch": 0.5206490517224981,
"grad_norm": 1.608587384223938,
"learning_rate": 1e-05,
"loss": 0.7506,
"step": 27370
},
{
"epoch": 0.5208392779014248,
"grad_norm": 2.0665712356567383,
"learning_rate": 1e-05,
"loss": 0.5875,
"step": 27380
},
{
"epoch": 0.5210295040803515,
"grad_norm": 2.7834298610687256,
"learning_rate": 1e-05,
"loss": 0.5666,
"step": 27390
},
{
"epoch": 0.5212197302592783,
"grad_norm": 7.0573835372924805,
"learning_rate": 1e-05,
"loss": 0.6988,
"step": 27400
},
{
"epoch": 0.521409956438205,
"grad_norm": 1.5806615352630615,
"learning_rate": 1e-05,
"loss": 0.7958,
"step": 27410
},
{
"epoch": 0.5216001826171318,
"grad_norm": 2.0366570949554443,
"learning_rate": 1e-05,
"loss": 0.773,
"step": 27420
},
{
"epoch": 0.5217904087960585,
"grad_norm": 1.837448000907898,
"learning_rate": 1e-05,
"loss": 0.6216,
"step": 27430
},
{
"epoch": 0.5219806349749853,
"grad_norm": 2.48103666305542,
"learning_rate": 1e-05,
"loss": 0.6063,
"step": 27440
},
{
"epoch": 0.522170861153912,
"grad_norm": 5.834935188293457,
"learning_rate": 1e-05,
"loss": 0.7025,
"step": 27450
},
{
"epoch": 0.5223610873328387,
"grad_norm": 1.5611969232559204,
"learning_rate": 1e-05,
"loss": 0.8203,
"step": 27460
},
{
"epoch": 0.5225513135117655,
"grad_norm": 1.716864824295044,
"learning_rate": 1e-05,
"loss": 0.7305,
"step": 27470
},
{
"epoch": 0.5227415396906923,
"grad_norm": 1.816564917564392,
"learning_rate": 1e-05,
"loss": 0.5886,
"step": 27480
},
{
"epoch": 0.522931765869619,
"grad_norm": 2.760875701904297,
"learning_rate": 1e-05,
"loss": 0.5545,
"step": 27490
},
{
"epoch": 0.5231219920485457,
"grad_norm": 6.841777801513672,
"learning_rate": 1e-05,
"loss": 0.7145,
"step": 27500
},
{
"epoch": 0.5233122182274724,
"grad_norm": 1.434421181678772,
"learning_rate": 1e-05,
"loss": 0.801,
"step": 27510
},
{
"epoch": 0.5235024444063993,
"grad_norm": 1.951026439666748,
"learning_rate": 1e-05,
"loss": 0.7595,
"step": 27520
},
{
"epoch": 0.523692670585326,
"grad_norm": 2.1492252349853516,
"learning_rate": 1e-05,
"loss": 0.5507,
"step": 27530
},
{
"epoch": 0.5238828967642527,
"grad_norm": 2.586805582046509,
"learning_rate": 1e-05,
"loss": 0.5568,
"step": 27540
},
{
"epoch": 0.5240731229431794,
"grad_norm": 5.794380187988281,
"learning_rate": 1e-05,
"loss": 0.7159,
"step": 27550
},
{
"epoch": 0.5242633491221061,
"grad_norm": 1.3293472528457642,
"learning_rate": 1e-05,
"loss": 0.778,
"step": 27560
},
{
"epoch": 0.524453575301033,
"grad_norm": 1.5989394187927246,
"learning_rate": 1e-05,
"loss": 0.758,
"step": 27570
},
{
"epoch": 0.5246438014799597,
"grad_norm": 1.8876078128814697,
"learning_rate": 1e-05,
"loss": 0.5894,
"step": 27580
},
{
"epoch": 0.5248340276588864,
"grad_norm": 2.6294922828674316,
"learning_rate": 1e-05,
"loss": 0.581,
"step": 27590
},
{
"epoch": 0.5250242538378131,
"grad_norm": 6.965528964996338,
"learning_rate": 1e-05,
"loss": 0.6715,
"step": 27600
},
{
"epoch": 0.5252144800167399,
"grad_norm": 1.3446075916290283,
"learning_rate": 1e-05,
"loss": 0.7774,
"step": 27610
},
{
"epoch": 0.5254047061956666,
"grad_norm": 1.8073546886444092,
"learning_rate": 1e-05,
"loss": 0.73,
"step": 27620
},
{
"epoch": 0.5255949323745934,
"grad_norm": 2.443486213684082,
"learning_rate": 1e-05,
"loss": 0.5971,
"step": 27630
},
{
"epoch": 0.5257851585535201,
"grad_norm": 2.417875051498413,
"learning_rate": 1e-05,
"loss": 0.578,
"step": 27640
},
{
"epoch": 0.5259753847324469,
"grad_norm": 6.705755710601807,
"learning_rate": 1e-05,
"loss": 0.6901,
"step": 27650
},
{
"epoch": 0.5261656109113736,
"grad_norm": 1.3031730651855469,
"learning_rate": 1e-05,
"loss": 0.8128,
"step": 27660
},
{
"epoch": 0.5263558370903003,
"grad_norm": 1.4837942123413086,
"learning_rate": 1e-05,
"loss": 0.7585,
"step": 27670
},
{
"epoch": 0.5265460632692271,
"grad_norm": 1.951445460319519,
"learning_rate": 1e-05,
"loss": 0.565,
"step": 27680
},
{
"epoch": 0.5267362894481539,
"grad_norm": 2.5272397994995117,
"learning_rate": 1e-05,
"loss": 0.5918,
"step": 27690
},
{
"epoch": 0.5269265156270806,
"grad_norm": 6.136901378631592,
"learning_rate": 1e-05,
"loss": 0.7087,
"step": 27700
},
{
"epoch": 0.5271167418060073,
"grad_norm": 1.2595221996307373,
"learning_rate": 1e-05,
"loss": 0.8017,
"step": 27710
},
{
"epoch": 0.527306967984934,
"grad_norm": 2.2021961212158203,
"learning_rate": 1e-05,
"loss": 0.789,
"step": 27720
},
{
"epoch": 0.5274971941638609,
"grad_norm": 2.069411277770996,
"learning_rate": 1e-05,
"loss": 0.6014,
"step": 27730
},
{
"epoch": 0.5276874203427876,
"grad_norm": 2.618126392364502,
"learning_rate": 1e-05,
"loss": 0.5771,
"step": 27740
},
{
"epoch": 0.5278776465217143,
"grad_norm": 5.805337905883789,
"learning_rate": 1e-05,
"loss": 0.6807,
"step": 27750
},
{
"epoch": 0.528067872700641,
"grad_norm": 1.3324886560440063,
"learning_rate": 1e-05,
"loss": 0.7998,
"step": 27760
},
{
"epoch": 0.5282580988795678,
"grad_norm": 1.811241626739502,
"learning_rate": 1e-05,
"loss": 0.7366,
"step": 27770
},
{
"epoch": 0.5284483250584946,
"grad_norm": 1.9403773546218872,
"learning_rate": 1e-05,
"loss": 0.6137,
"step": 27780
},
{
"epoch": 0.5286385512374213,
"grad_norm": 2.476445436477661,
"learning_rate": 1e-05,
"loss": 0.5828,
"step": 27790
},
{
"epoch": 0.528828777416348,
"grad_norm": 5.76503849029541,
"learning_rate": 1e-05,
"loss": 0.6884,
"step": 27800
},
{
"epoch": 0.5290190035952748,
"grad_norm": 1.2381008863449097,
"learning_rate": 1e-05,
"loss": 0.8055,
"step": 27810
},
{
"epoch": 0.5292092297742015,
"grad_norm": 1.7735977172851562,
"learning_rate": 1e-05,
"loss": 0.7436,
"step": 27820
},
{
"epoch": 0.5293994559531283,
"grad_norm": 2.0617969036102295,
"learning_rate": 1e-05,
"loss": 0.6169,
"step": 27830
},
{
"epoch": 0.529589682132055,
"grad_norm": 2.406144857406616,
"learning_rate": 1e-05,
"loss": 0.5567,
"step": 27840
},
{
"epoch": 0.5297799083109818,
"grad_norm": 6.201722621917725,
"learning_rate": 1e-05,
"loss": 0.7078,
"step": 27850
},
{
"epoch": 0.5299701344899085,
"grad_norm": 1.6492078304290771,
"learning_rate": 1e-05,
"loss": 0.8179,
"step": 27860
},
{
"epoch": 0.5301603606688352,
"grad_norm": 1.7765238285064697,
"learning_rate": 1e-05,
"loss": 0.7297,
"step": 27870
},
{
"epoch": 0.530350586847762,
"grad_norm": 2.128385543823242,
"learning_rate": 1e-05,
"loss": 0.5951,
"step": 27880
},
{
"epoch": 0.5305408130266888,
"grad_norm": 2.5266244411468506,
"learning_rate": 1e-05,
"loss": 0.5837,
"step": 27890
},
{
"epoch": 0.5307310392056155,
"grad_norm": 5.7893829345703125,
"learning_rate": 1e-05,
"loss": 0.6587,
"step": 27900
},
{
"epoch": 0.5309212653845422,
"grad_norm": 1.5904130935668945,
"learning_rate": 1e-05,
"loss": 0.8236,
"step": 27910
},
{
"epoch": 0.5311114915634689,
"grad_norm": 1.6347166299819946,
"learning_rate": 1e-05,
"loss": 0.7315,
"step": 27920
},
{
"epoch": 0.5313017177423958,
"grad_norm": 1.873299479484558,
"learning_rate": 1e-05,
"loss": 0.6016,
"step": 27930
},
{
"epoch": 0.5314919439213225,
"grad_norm": 2.6605935096740723,
"learning_rate": 1e-05,
"loss": 0.5482,
"step": 27940
},
{
"epoch": 0.5316821701002492,
"grad_norm": 8.03520679473877,
"learning_rate": 1e-05,
"loss": 0.7169,
"step": 27950
},
{
"epoch": 0.5318723962791759,
"grad_norm": 2.1033520698547363,
"learning_rate": 1e-05,
"loss": 0.7935,
"step": 27960
},
{
"epoch": 0.5320626224581027,
"grad_norm": 1.467359185218811,
"learning_rate": 1e-05,
"loss": 0.7689,
"step": 27970
},
{
"epoch": 0.5322528486370294,
"grad_norm": 1.8925598859786987,
"learning_rate": 1e-05,
"loss": 0.6117,
"step": 27980
},
{
"epoch": 0.5324430748159562,
"grad_norm": 2.5618064403533936,
"learning_rate": 1e-05,
"loss": 0.5669,
"step": 27990
},
{
"epoch": 0.5326333009948829,
"grad_norm": 5.857771873474121,
"learning_rate": 1e-05,
"loss": 0.7239,
"step": 28000
},
{
"epoch": 0.5328235271738097,
"grad_norm": 1.3273088932037354,
"learning_rate": 1e-05,
"loss": 0.7783,
"step": 28010
},
{
"epoch": 0.5330137533527364,
"grad_norm": 1.5500215291976929,
"learning_rate": 1e-05,
"loss": 0.7402,
"step": 28020
},
{
"epoch": 0.5332039795316631,
"grad_norm": 1.9283325672149658,
"learning_rate": 1e-05,
"loss": 0.6126,
"step": 28030
},
{
"epoch": 0.5333942057105899,
"grad_norm": 2.4763705730438232,
"learning_rate": 1e-05,
"loss": 0.6044,
"step": 28040
},
{
"epoch": 0.5335844318895167,
"grad_norm": 6.787921905517578,
"learning_rate": 1e-05,
"loss": 0.6806,
"step": 28050
},
{
"epoch": 0.5337746580684434,
"grad_norm": 1.9250437021255493,
"learning_rate": 1e-05,
"loss": 0.8101,
"step": 28060
},
{
"epoch": 0.5339648842473701,
"grad_norm": 1.572171926498413,
"learning_rate": 1e-05,
"loss": 0.7626,
"step": 28070
},
{
"epoch": 0.5341551104262968,
"grad_norm": 1.7359269857406616,
"learning_rate": 1e-05,
"loss": 0.6203,
"step": 28080
},
{
"epoch": 0.5343453366052237,
"grad_norm": 2.3607969284057617,
"learning_rate": 1e-05,
"loss": 0.5532,
"step": 28090
},
{
"epoch": 0.5345355627841504,
"grad_norm": 7.541353225708008,
"learning_rate": 1e-05,
"loss": 0.7453,
"step": 28100
},
{
"epoch": 0.5347257889630771,
"grad_norm": 1.2965593338012695,
"learning_rate": 1e-05,
"loss": 0.8077,
"step": 28110
},
{
"epoch": 0.5349160151420038,
"grad_norm": 1.5686266422271729,
"learning_rate": 1e-05,
"loss": 0.7523,
"step": 28120
},
{
"epoch": 0.5351062413209305,
"grad_norm": 1.9476683139801025,
"learning_rate": 1e-05,
"loss": 0.6086,
"step": 28130
},
{
"epoch": 0.5352964674998574,
"grad_norm": 2.743095874786377,
"learning_rate": 1e-05,
"loss": 0.5786,
"step": 28140
},
{
"epoch": 0.5354866936787841,
"grad_norm": 6.631075859069824,
"learning_rate": 1e-05,
"loss": 0.7093,
"step": 28150
},
{
"epoch": 0.5356769198577108,
"grad_norm": 1.331789493560791,
"learning_rate": 1e-05,
"loss": 0.7631,
"step": 28160
},
{
"epoch": 0.5358671460366375,
"grad_norm": 1.439956784248352,
"learning_rate": 1e-05,
"loss": 0.7659,
"step": 28170
},
{
"epoch": 0.5360573722155643,
"grad_norm": 2.074004888534546,
"learning_rate": 1e-05,
"loss": 0.6159,
"step": 28180
},
{
"epoch": 0.5362475983944911,
"grad_norm": 2.34525990486145,
"learning_rate": 1e-05,
"loss": 0.5601,
"step": 28190
},
{
"epoch": 0.5364378245734178,
"grad_norm": 6.267715930938721,
"learning_rate": 1e-05,
"loss": 0.6614,
"step": 28200
},
{
"epoch": 0.5366280507523445,
"grad_norm": 1.861405372619629,
"learning_rate": 1e-05,
"loss": 0.7887,
"step": 28210
},
{
"epoch": 0.5368182769312713,
"grad_norm": 1.4569647312164307,
"learning_rate": 1e-05,
"loss": 0.7356,
"step": 28220
},
{
"epoch": 0.537008503110198,
"grad_norm": 2.006636619567871,
"learning_rate": 1e-05,
"loss": 0.5684,
"step": 28230
},
{
"epoch": 0.5371987292891248,
"grad_norm": 3.0889835357666016,
"learning_rate": 1e-05,
"loss": 0.5686,
"step": 28240
},
{
"epoch": 0.5373889554680515,
"grad_norm": 7.1055684089660645,
"learning_rate": 1e-05,
"loss": 0.7217,
"step": 28250
},
{
"epoch": 0.5375791816469783,
"grad_norm": 1.3795398473739624,
"learning_rate": 1e-05,
"loss": 0.7849,
"step": 28260
},
{
"epoch": 0.537769407825905,
"grad_norm": 1.5904273986816406,
"learning_rate": 1e-05,
"loss": 0.737,
"step": 28270
},
{
"epoch": 0.5379596340048317,
"grad_norm": 1.823179841041565,
"learning_rate": 1e-05,
"loss": 0.632,
"step": 28280
},
{
"epoch": 0.5381498601837585,
"grad_norm": 2.9460597038269043,
"learning_rate": 1e-05,
"loss": 0.5892,
"step": 28290
},
{
"epoch": 0.5383400863626853,
"grad_norm": 6.5957865715026855,
"learning_rate": 1e-05,
"loss": 0.7289,
"step": 28300
},
{
"epoch": 0.538530312541612,
"grad_norm": 1.48269522190094,
"learning_rate": 1e-05,
"loss": 0.7837,
"step": 28310
},
{
"epoch": 0.5387205387205387,
"grad_norm": 1.7960617542266846,
"learning_rate": 1e-05,
"loss": 0.7619,
"step": 28320
},
{
"epoch": 0.5389107648994654,
"grad_norm": 1.9189598560333252,
"learning_rate": 1e-05,
"loss": 0.6173,
"step": 28330
},
{
"epoch": 0.5391009910783922,
"grad_norm": 2.705803155899048,
"learning_rate": 1e-05,
"loss": 0.5721,
"step": 28340
},
{
"epoch": 0.539291217257319,
"grad_norm": 5.9613752365112305,
"learning_rate": 1e-05,
"loss": 0.7004,
"step": 28350
},
{
"epoch": 0.5394814434362457,
"grad_norm": 1.3911089897155762,
"learning_rate": 1e-05,
"loss": 0.7784,
"step": 28360
},
{
"epoch": 0.5396716696151724,
"grad_norm": 1.8370181322097778,
"learning_rate": 1e-05,
"loss": 0.7541,
"step": 28370
},
{
"epoch": 0.5398618957940992,
"grad_norm": 1.9959443807601929,
"learning_rate": 1e-05,
"loss": 0.6267,
"step": 28380
},
{
"epoch": 0.5400521219730259,
"grad_norm": 2.6368703842163086,
"learning_rate": 1e-05,
"loss": 0.5865,
"step": 28390
},
{
"epoch": 0.5402423481519527,
"grad_norm": 6.700469970703125,
"learning_rate": 1e-05,
"loss": 0.6965,
"step": 28400
},
{
"epoch": 0.5404325743308794,
"grad_norm": 1.4312291145324707,
"learning_rate": 1e-05,
"loss": 0.7941,
"step": 28410
},
{
"epoch": 0.5406228005098062,
"grad_norm": 1.8878716230392456,
"learning_rate": 1e-05,
"loss": 0.7944,
"step": 28420
},
{
"epoch": 0.5408130266887329,
"grad_norm": 1.9028757810592651,
"learning_rate": 1e-05,
"loss": 0.6026,
"step": 28430
},
{
"epoch": 0.5410032528676596,
"grad_norm": 2.4574577808380127,
"learning_rate": 1e-05,
"loss": 0.6096,
"step": 28440
},
{
"epoch": 0.5411934790465864,
"grad_norm": 7.736937046051025,
"learning_rate": 1e-05,
"loss": 0.7299,
"step": 28450
},
{
"epoch": 0.5413837052255132,
"grad_norm": 1.3780299425125122,
"learning_rate": 1e-05,
"loss": 0.7854,
"step": 28460
},
{
"epoch": 0.5415739314044399,
"grad_norm": 1.4352824687957764,
"learning_rate": 1e-05,
"loss": 0.7319,
"step": 28470
},
{
"epoch": 0.5417641575833666,
"grad_norm": 1.86263108253479,
"learning_rate": 1e-05,
"loss": 0.602,
"step": 28480
},
{
"epoch": 0.5419543837622933,
"grad_norm": 2.620088577270508,
"learning_rate": 1e-05,
"loss": 0.5917,
"step": 28490
},
{
"epoch": 0.5421446099412202,
"grad_norm": 7.4325761795043945,
"learning_rate": 1e-05,
"loss": 0.7105,
"step": 28500
},
{
"epoch": 0.5423348361201469,
"grad_norm": 1.291056752204895,
"learning_rate": 1e-05,
"loss": 0.7583,
"step": 28510
},
{
"epoch": 0.5425250622990736,
"grad_norm": 1.6262853145599365,
"learning_rate": 1e-05,
"loss": 0.7513,
"step": 28520
},
{
"epoch": 0.5427152884780003,
"grad_norm": 1.86658775806427,
"learning_rate": 1e-05,
"loss": 0.571,
"step": 28530
},
{
"epoch": 0.542905514656927,
"grad_norm": 2.3650619983673096,
"learning_rate": 1e-05,
"loss": 0.5814,
"step": 28540
},
{
"epoch": 0.5430957408358539,
"grad_norm": 6.558077335357666,
"learning_rate": 1e-05,
"loss": 0.7064,
"step": 28550
},
{
"epoch": 0.5432859670147806,
"grad_norm": 1.5061731338500977,
"learning_rate": 1e-05,
"loss": 0.8089,
"step": 28560
},
{
"epoch": 0.5434761931937073,
"grad_norm": 1.644956350326538,
"learning_rate": 1e-05,
"loss": 0.7583,
"step": 28570
},
{
"epoch": 0.543666419372634,
"grad_norm": 1.7633553743362427,
"learning_rate": 1e-05,
"loss": 0.5866,
"step": 28580
},
{
"epoch": 0.5438566455515608,
"grad_norm": 2.7508761882781982,
"learning_rate": 1e-05,
"loss": 0.6144,
"step": 28590
},
{
"epoch": 0.5440468717304876,
"grad_norm": 6.460179328918457,
"learning_rate": 1e-05,
"loss": 0.7035,
"step": 28600
},
{
"epoch": 0.5442370979094143,
"grad_norm": 1.2889879941940308,
"learning_rate": 1e-05,
"loss": 0.7571,
"step": 28610
},
{
"epoch": 0.544427324088341,
"grad_norm": 1.448184609413147,
"learning_rate": 1e-05,
"loss": 0.7322,
"step": 28620
},
{
"epoch": 0.5446175502672678,
"grad_norm": 1.8823974132537842,
"learning_rate": 1e-05,
"loss": 0.6042,
"step": 28630
},
{
"epoch": 0.5448077764461945,
"grad_norm": 2.188398838043213,
"learning_rate": 1e-05,
"loss": 0.5659,
"step": 28640
},
{
"epoch": 0.5449980026251212,
"grad_norm": 5.888668537139893,
"learning_rate": 1e-05,
"loss": 0.6582,
"step": 28650
},
{
"epoch": 0.545188228804048,
"grad_norm": 1.266444206237793,
"learning_rate": 1e-05,
"loss": 0.8042,
"step": 28660
},
{
"epoch": 0.5453784549829748,
"grad_norm": 1.710341453552246,
"learning_rate": 1e-05,
"loss": 0.7606,
"step": 28670
},
{
"epoch": 0.5455686811619015,
"grad_norm": 1.9488623142242432,
"learning_rate": 1e-05,
"loss": 0.6225,
"step": 28680
},
{
"epoch": 0.5457589073408282,
"grad_norm": 2.4624128341674805,
"learning_rate": 1e-05,
"loss": 0.5729,
"step": 28690
},
{
"epoch": 0.5459491335197549,
"grad_norm": 6.178300380706787,
"learning_rate": 1e-05,
"loss": 0.7127,
"step": 28700
},
{
"epoch": 0.5461393596986818,
"grad_norm": 1.4813989400863647,
"learning_rate": 1e-05,
"loss": 0.8329,
"step": 28710
},
{
"epoch": 0.5463295858776085,
"grad_norm": 1.5241448879241943,
"learning_rate": 1e-05,
"loss": 0.718,
"step": 28720
},
{
"epoch": 0.5465198120565352,
"grad_norm": 1.9688202142715454,
"learning_rate": 1e-05,
"loss": 0.6262,
"step": 28730
},
{
"epoch": 0.5467100382354619,
"grad_norm": 2.5945465564727783,
"learning_rate": 1e-05,
"loss": 0.5942,
"step": 28740
},
{
"epoch": 0.5469002644143887,
"grad_norm": 6.7968645095825195,
"learning_rate": 1e-05,
"loss": 0.7103,
"step": 28750
},
{
"epoch": 0.5470904905933155,
"grad_norm": 1.1676732301712036,
"learning_rate": 1e-05,
"loss": 0.7726,
"step": 28760
},
{
"epoch": 0.5472807167722422,
"grad_norm": 1.6644660234451294,
"learning_rate": 1e-05,
"loss": 0.7388,
"step": 28770
},
{
"epoch": 0.5474709429511689,
"grad_norm": 2.0268590450286865,
"learning_rate": 1e-05,
"loss": 0.614,
"step": 28780
},
{
"epoch": 0.5476611691300957,
"grad_norm": 2.428539991378784,
"learning_rate": 1e-05,
"loss": 0.5849,
"step": 28790
},
{
"epoch": 0.5478513953090224,
"grad_norm": 6.891856670379639,
"learning_rate": 1e-05,
"loss": 0.7027,
"step": 28800
},
{
"epoch": 0.5480416214879492,
"grad_norm": 1.2530388832092285,
"learning_rate": 1e-05,
"loss": 0.7782,
"step": 28810
},
{
"epoch": 0.5482318476668759,
"grad_norm": 2.073314666748047,
"learning_rate": 1e-05,
"loss": 0.7399,
"step": 28820
},
{
"epoch": 0.5484220738458027,
"grad_norm": 1.8712562322616577,
"learning_rate": 1e-05,
"loss": 0.5919,
"step": 28830
},
{
"epoch": 0.5486123000247294,
"grad_norm": 2.826935052871704,
"learning_rate": 1e-05,
"loss": 0.5787,
"step": 28840
},
{
"epoch": 0.5488025262036561,
"grad_norm": 6.0296735763549805,
"learning_rate": 1e-05,
"loss": 0.6842,
"step": 28850
},
{
"epoch": 0.5489927523825829,
"grad_norm": 1.3791965246200562,
"learning_rate": 1e-05,
"loss": 0.798,
"step": 28860
},
{
"epoch": 0.5491829785615097,
"grad_norm": 1.4636938571929932,
"learning_rate": 1e-05,
"loss": 0.7471,
"step": 28870
},
{
"epoch": 0.5493732047404364,
"grad_norm": 2.024298906326294,
"learning_rate": 1e-05,
"loss": 0.6018,
"step": 28880
},
{
"epoch": 0.5495634309193631,
"grad_norm": 2.59021258354187,
"learning_rate": 1e-05,
"loss": 0.5867,
"step": 28890
},
{
"epoch": 0.5497536570982898,
"grad_norm": 5.938200950622559,
"learning_rate": 1e-05,
"loss": 0.6914,
"step": 28900
},
{
"epoch": 0.5499438832772167,
"grad_norm": 1.4610862731933594,
"learning_rate": 1e-05,
"loss": 0.7969,
"step": 28910
},
{
"epoch": 0.5501341094561434,
"grad_norm": 1.8848310708999634,
"learning_rate": 1e-05,
"loss": 0.7427,
"step": 28920
},
{
"epoch": 0.5503243356350701,
"grad_norm": 1.7692400217056274,
"learning_rate": 1e-05,
"loss": 0.5859,
"step": 28930
},
{
"epoch": 0.5505145618139968,
"grad_norm": 2.7332236766815186,
"learning_rate": 1e-05,
"loss": 0.5774,
"step": 28940
},
{
"epoch": 0.5507047879929236,
"grad_norm": 6.071177959442139,
"learning_rate": 1e-05,
"loss": 0.7158,
"step": 28950
},
{
"epoch": 0.5508950141718504,
"grad_norm": 2.0782177448272705,
"learning_rate": 1e-05,
"loss": 0.7985,
"step": 28960
},
{
"epoch": 0.5510852403507771,
"grad_norm": 1.8426365852355957,
"learning_rate": 1e-05,
"loss": 0.7949,
"step": 28970
},
{
"epoch": 0.5512754665297038,
"grad_norm": 2.006798267364502,
"learning_rate": 1e-05,
"loss": 0.5925,
"step": 28980
},
{
"epoch": 0.5514656927086306,
"grad_norm": 2.813978672027588,
"learning_rate": 1e-05,
"loss": 0.5934,
"step": 28990
},
{
"epoch": 0.5516559188875573,
"grad_norm": 5.671996116638184,
"learning_rate": 1e-05,
"loss": 0.6623,
"step": 29000
},
{
"epoch": 0.551846145066484,
"grad_norm": 1.1891865730285645,
"learning_rate": 1e-05,
"loss": 0.7861,
"step": 29010
},
{
"epoch": 0.5520363712454108,
"grad_norm": 1.7034006118774414,
"learning_rate": 1e-05,
"loss": 0.7486,
"step": 29020
},
{
"epoch": 0.5522265974243376,
"grad_norm": 2.028172254562378,
"learning_rate": 1e-05,
"loss": 0.6221,
"step": 29030
},
{
"epoch": 0.5524168236032643,
"grad_norm": 2.6445059776306152,
"learning_rate": 1e-05,
"loss": 0.5592,
"step": 29040
},
{
"epoch": 0.552607049782191,
"grad_norm": 5.840847015380859,
"learning_rate": 1e-05,
"loss": 0.6607,
"step": 29050
},
{
"epoch": 0.5527972759611177,
"grad_norm": 1.402751088142395,
"learning_rate": 1e-05,
"loss": 0.7967,
"step": 29060
},
{
"epoch": 0.5529875021400446,
"grad_norm": 1.5487130880355835,
"learning_rate": 1e-05,
"loss": 0.7103,
"step": 29070
},
{
"epoch": 0.5531777283189713,
"grad_norm": 2.0249884128570557,
"learning_rate": 1e-05,
"loss": 0.6109,
"step": 29080
},
{
"epoch": 0.553367954497898,
"grad_norm": 2.302168607711792,
"learning_rate": 1e-05,
"loss": 0.5663,
"step": 29090
},
{
"epoch": 0.5535581806768247,
"grad_norm": 7.091302871704102,
"learning_rate": 1e-05,
"loss": 0.6886,
"step": 29100
},
{
"epoch": 0.5537484068557514,
"grad_norm": 1.418855905532837,
"learning_rate": 1e-05,
"loss": 0.8091,
"step": 29110
},
{
"epoch": 0.5539386330346783,
"grad_norm": 1.8696565628051758,
"learning_rate": 1e-05,
"loss": 0.7307,
"step": 29120
},
{
"epoch": 0.554128859213605,
"grad_norm": 2.113225221633911,
"learning_rate": 1e-05,
"loss": 0.6343,
"step": 29130
},
{
"epoch": 0.5543190853925317,
"grad_norm": 2.322115659713745,
"learning_rate": 1e-05,
"loss": 0.5764,
"step": 29140
},
{
"epoch": 0.5545093115714584,
"grad_norm": 7.4178242683410645,
"learning_rate": 1e-05,
"loss": 0.7379,
"step": 29150
},
{
"epoch": 0.5546995377503852,
"grad_norm": 1.2225861549377441,
"learning_rate": 1e-05,
"loss": 0.7992,
"step": 29160
},
{
"epoch": 0.554889763929312,
"grad_norm": 1.5799157619476318,
"learning_rate": 1e-05,
"loss": 0.7607,
"step": 29170
},
{
"epoch": 0.5550799901082387,
"grad_norm": 1.8895138502120972,
"learning_rate": 1e-05,
"loss": 0.5903,
"step": 29180
},
{
"epoch": 0.5552702162871654,
"grad_norm": 2.50899338722229,
"learning_rate": 1e-05,
"loss": 0.5569,
"step": 29190
},
{
"epoch": 0.5554604424660922,
"grad_norm": 7.346343994140625,
"learning_rate": 1e-05,
"loss": 0.668,
"step": 29200
},
{
"epoch": 0.5556506686450189,
"grad_norm": 1.3231711387634277,
"learning_rate": 1e-05,
"loss": 0.8064,
"step": 29210
},
{
"epoch": 0.5558408948239457,
"grad_norm": 1.5447133779525757,
"learning_rate": 1e-05,
"loss": 0.7292,
"step": 29220
},
{
"epoch": 0.5560311210028724,
"grad_norm": 2.454815626144409,
"learning_rate": 1e-05,
"loss": 0.616,
"step": 29230
},
{
"epoch": 0.5562213471817992,
"grad_norm": 2.5089452266693115,
"learning_rate": 1e-05,
"loss": 0.5772,
"step": 29240
},
{
"epoch": 0.5564115733607259,
"grad_norm": 7.431902885437012,
"learning_rate": 1e-05,
"loss": 0.6926,
"step": 29250
},
{
"epoch": 0.5566017995396526,
"grad_norm": 2.4867982864379883,
"learning_rate": 1e-05,
"loss": 0.7928,
"step": 29260
},
{
"epoch": 0.5567920257185794,
"grad_norm": 1.7437180280685425,
"learning_rate": 1e-05,
"loss": 0.7606,
"step": 29270
},
{
"epoch": 0.5569822518975062,
"grad_norm": 1.9877288341522217,
"learning_rate": 1e-05,
"loss": 0.6407,
"step": 29280
},
{
"epoch": 0.5571724780764329,
"grad_norm": 2.6790881156921387,
"learning_rate": 1e-05,
"loss": 0.6013,
"step": 29290
},
{
"epoch": 0.5573627042553596,
"grad_norm": 7.395862579345703,
"learning_rate": 1e-05,
"loss": 0.7164,
"step": 29300
},
{
"epoch": 0.5575529304342863,
"grad_norm": 1.2900116443634033,
"learning_rate": 1e-05,
"loss": 0.789,
"step": 29310
},
{
"epoch": 0.5577431566132132,
"grad_norm": 1.5753931999206543,
"learning_rate": 1e-05,
"loss": 0.748,
"step": 29320
},
{
"epoch": 0.5579333827921399,
"grad_norm": 1.9359933137893677,
"learning_rate": 1e-05,
"loss": 0.6034,
"step": 29330
},
{
"epoch": 0.5581236089710666,
"grad_norm": 2.5064215660095215,
"learning_rate": 1e-05,
"loss": 0.6155,
"step": 29340
},
{
"epoch": 0.5583138351499933,
"grad_norm": 5.94927978515625,
"learning_rate": 1e-05,
"loss": 0.6878,
"step": 29350
},
{
"epoch": 0.55850406132892,
"grad_norm": 1.2428947687149048,
"learning_rate": 1e-05,
"loss": 0.7878,
"step": 29360
},
{
"epoch": 0.5586942875078468,
"grad_norm": 1.607889175415039,
"learning_rate": 1e-05,
"loss": 0.7015,
"step": 29370
},
{
"epoch": 0.5588845136867736,
"grad_norm": 2.1764302253723145,
"learning_rate": 1e-05,
"loss": 0.6229,
"step": 29380
},
{
"epoch": 0.5590747398657003,
"grad_norm": 2.385596752166748,
"learning_rate": 1e-05,
"loss": 0.5844,
"step": 29390
},
{
"epoch": 0.559264966044627,
"grad_norm": 6.2451324462890625,
"learning_rate": 1e-05,
"loss": 0.6719,
"step": 29400
},
{
"epoch": 0.5594551922235538,
"grad_norm": 1.982694149017334,
"learning_rate": 1e-05,
"loss": 0.7665,
"step": 29410
},
{
"epoch": 0.5596454184024805,
"grad_norm": 1.3985553979873657,
"learning_rate": 1e-05,
"loss": 0.7426,
"step": 29420
},
{
"epoch": 0.5598356445814073,
"grad_norm": 1.9726155996322632,
"learning_rate": 1e-05,
"loss": 0.6028,
"step": 29430
},
{
"epoch": 0.560025870760334,
"grad_norm": 2.49698543548584,
"learning_rate": 1e-05,
"loss": 0.5642,
"step": 29440
},
{
"epoch": 0.5602160969392608,
"grad_norm": 6.6375508308410645,
"learning_rate": 1e-05,
"loss": 0.7128,
"step": 29450
},
{
"epoch": 0.5604063231181875,
"grad_norm": 1.296148657798767,
"learning_rate": 1e-05,
"loss": 0.7893,
"step": 29460
},
{
"epoch": 0.5605965492971142,
"grad_norm": 1.5590506792068481,
"learning_rate": 1e-05,
"loss": 0.7235,
"step": 29470
},
{
"epoch": 0.560786775476041,
"grad_norm": 1.9726529121398926,
"learning_rate": 1e-05,
"loss": 0.5702,
"step": 29480
},
{
"epoch": 0.5609770016549678,
"grad_norm": 2.3355813026428223,
"learning_rate": 1e-05,
"loss": 0.5986,
"step": 29490
},
{
"epoch": 0.5611672278338945,
"grad_norm": 6.404568672180176,
"learning_rate": 1e-05,
"loss": 0.7461,
"step": 29500
},
{
"epoch": 0.5613574540128212,
"grad_norm": 1.2093815803527832,
"learning_rate": 1e-05,
"loss": 0.7704,
"step": 29510
},
{
"epoch": 0.5615476801917479,
"grad_norm": 1.661629319190979,
"learning_rate": 1e-05,
"loss": 0.753,
"step": 29520
},
{
"epoch": 0.5617379063706748,
"grad_norm": 2.0196964740753174,
"learning_rate": 1e-05,
"loss": 0.5848,
"step": 29530
},
{
"epoch": 0.5619281325496015,
"grad_norm": 2.4349427223205566,
"learning_rate": 1e-05,
"loss": 0.5452,
"step": 29540
},
{
"epoch": 0.5621183587285282,
"grad_norm": 5.383221626281738,
"learning_rate": 1e-05,
"loss": 0.6494,
"step": 29550
},
{
"epoch": 0.562308584907455,
"grad_norm": 1.4114487171173096,
"learning_rate": 1e-05,
"loss": 0.7888,
"step": 29560
},
{
"epoch": 0.5624988110863817,
"grad_norm": 1.720810055732727,
"learning_rate": 1e-05,
"loss": 0.7237,
"step": 29570
},
{
"epoch": 0.5626890372653085,
"grad_norm": 1.9097247123718262,
"learning_rate": 1e-05,
"loss": 0.5776,
"step": 29580
},
{
"epoch": 0.5628792634442352,
"grad_norm": 2.5241849422454834,
"learning_rate": 1e-05,
"loss": 0.5926,
"step": 29590
},
{
"epoch": 0.563069489623162,
"grad_norm": 6.2696146965026855,
"learning_rate": 1e-05,
"loss": 0.7242,
"step": 29600
},
{
"epoch": 0.5632597158020887,
"grad_norm": 1.378063440322876,
"learning_rate": 1e-05,
"loss": 0.7861,
"step": 29610
},
{
"epoch": 0.5634499419810154,
"grad_norm": 1.3331849575042725,
"learning_rate": 1e-05,
"loss": 0.764,
"step": 29620
},
{
"epoch": 0.5636401681599422,
"grad_norm": 1.8952465057373047,
"learning_rate": 1e-05,
"loss": 0.5873,
"step": 29630
},
{
"epoch": 0.563830394338869,
"grad_norm": 2.4615895748138428,
"learning_rate": 1e-05,
"loss": 0.5998,
"step": 29640
},
{
"epoch": 0.5640206205177957,
"grad_norm": 7.339219570159912,
"learning_rate": 1e-05,
"loss": 0.7147,
"step": 29650
},
{
"epoch": 0.5642108466967224,
"grad_norm": 1.293250560760498,
"learning_rate": 1e-05,
"loss": 0.7952,
"step": 29660
},
{
"epoch": 0.5644010728756491,
"grad_norm": 1.456028699874878,
"learning_rate": 1e-05,
"loss": 0.754,
"step": 29670
},
{
"epoch": 0.564591299054576,
"grad_norm": 2.108973264694214,
"learning_rate": 1e-05,
"loss": 0.6056,
"step": 29680
},
{
"epoch": 0.5647815252335027,
"grad_norm": 2.4971702098846436,
"learning_rate": 1e-05,
"loss": 0.5801,
"step": 29690
},
{
"epoch": 0.5649717514124294,
"grad_norm": 6.77570915222168,
"learning_rate": 1e-05,
"loss": 0.68,
"step": 29700
},
{
"epoch": 0.5651619775913561,
"grad_norm": 1.587699294090271,
"learning_rate": 1e-05,
"loss": 0.7847,
"step": 29710
},
{
"epoch": 0.5653522037702828,
"grad_norm": 1.5664024353027344,
"learning_rate": 1e-05,
"loss": 0.7658,
"step": 29720
},
{
"epoch": 0.5655424299492096,
"grad_norm": 2.1801064014434814,
"learning_rate": 1e-05,
"loss": 0.6238,
"step": 29730
},
{
"epoch": 0.5657326561281364,
"grad_norm": 2.177845001220703,
"learning_rate": 1e-05,
"loss": 0.5513,
"step": 29740
},
{
"epoch": 0.5659228823070631,
"grad_norm": 6.509383678436279,
"learning_rate": 1e-05,
"loss": 0.721,
"step": 29750
},
{
"epoch": 0.5661131084859898,
"grad_norm": 1.2346396446228027,
"learning_rate": 1e-05,
"loss": 0.7853,
"step": 29760
},
{
"epoch": 0.5663033346649166,
"grad_norm": 1.6915191411972046,
"learning_rate": 1e-05,
"loss": 0.7531,
"step": 29770
},
{
"epoch": 0.5664935608438433,
"grad_norm": 2.1893465518951416,
"learning_rate": 1e-05,
"loss": 0.6338,
"step": 29780
},
{
"epoch": 0.5666837870227701,
"grad_norm": 2.505868434906006,
"learning_rate": 1e-05,
"loss": 0.5842,
"step": 29790
},
{
"epoch": 0.5668740132016968,
"grad_norm": 5.915696620941162,
"learning_rate": 1e-05,
"loss": 0.7308,
"step": 29800
},
{
"epoch": 0.5670642393806236,
"grad_norm": 1.4407780170440674,
"learning_rate": 1e-05,
"loss": 0.8031,
"step": 29810
},
{
"epoch": 0.5672544655595503,
"grad_norm": 1.5830825567245483,
"learning_rate": 1e-05,
"loss": 0.7637,
"step": 29820
},
{
"epoch": 0.567444691738477,
"grad_norm": 2.0608913898468018,
"learning_rate": 1e-05,
"loss": 0.6216,
"step": 29830
},
{
"epoch": 0.5676349179174038,
"grad_norm": 2.5357120037078857,
"learning_rate": 1e-05,
"loss": 0.5686,
"step": 29840
},
{
"epoch": 0.5678251440963306,
"grad_norm": 7.026393413543701,
"learning_rate": 1e-05,
"loss": 0.6914,
"step": 29850
},
{
"epoch": 0.5680153702752573,
"grad_norm": 1.4412094354629517,
"learning_rate": 1e-05,
"loss": 0.8015,
"step": 29860
},
{
"epoch": 0.568205596454184,
"grad_norm": 1.4673103094100952,
"learning_rate": 1e-05,
"loss": 0.6977,
"step": 29870
},
{
"epoch": 0.5683958226331107,
"grad_norm": 1.9802727699279785,
"learning_rate": 1e-05,
"loss": 0.6062,
"step": 29880
},
{
"epoch": 0.5685860488120376,
"grad_norm": 2.661513090133667,
"learning_rate": 1e-05,
"loss": 0.5649,
"step": 29890
},
{
"epoch": 0.5687762749909643,
"grad_norm": 5.5551066398620605,
"learning_rate": 1e-05,
"loss": 0.682,
"step": 29900
},
{
"epoch": 0.568966501169891,
"grad_norm": 1.4657974243164062,
"learning_rate": 1e-05,
"loss": 0.7817,
"step": 29910
},
{
"epoch": 0.5691567273488177,
"grad_norm": 1.6034162044525146,
"learning_rate": 1e-05,
"loss": 0.7396,
"step": 29920
},
{
"epoch": 0.5693469535277444,
"grad_norm": 1.9244917631149292,
"learning_rate": 1e-05,
"loss": 0.6039,
"step": 29930
},
{
"epoch": 0.5695371797066713,
"grad_norm": 2.7198832035064697,
"learning_rate": 1e-05,
"loss": 0.5765,
"step": 29940
},
{
"epoch": 0.569727405885598,
"grad_norm": 5.565457344055176,
"learning_rate": 1e-05,
"loss": 0.7202,
"step": 29950
},
{
"epoch": 0.5699176320645247,
"grad_norm": 2.0305943489074707,
"learning_rate": 1e-05,
"loss": 0.7947,
"step": 29960
},
{
"epoch": 0.5701078582434514,
"grad_norm": 1.5212080478668213,
"learning_rate": 1e-05,
"loss": 0.733,
"step": 29970
},
{
"epoch": 0.5702980844223782,
"grad_norm": 1.8028953075408936,
"learning_rate": 1e-05,
"loss": 0.5961,
"step": 29980
},
{
"epoch": 0.570488310601305,
"grad_norm": 2.40333890914917,
"learning_rate": 1e-05,
"loss": 0.5743,
"step": 29990
},
{
"epoch": 0.5706785367802317,
"grad_norm": 6.989930629730225,
"learning_rate": 1e-05,
"loss": 0.7271,
"step": 30000
}
],
"logging_steps": 10,
"max_steps": 105138,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 6000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4.040907885477388e+19,
"train_batch_size": 7,
"trial_name": null,
"trial_params": null
}