mistral-v0.1-7b-sft-ultrachat / trainer_state.json
AmberYifan's picture
Model save
2ccf9bc verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1624,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006157635467980296,
"grad_norm": 12.449194669272266,
"learning_rate": 1.226993865030675e-07,
"loss": 1.1446,
"step": 1
},
{
"epoch": 0.003078817733990148,
"grad_norm": 9.980892973363595,
"learning_rate": 6.134969325153375e-07,
"loss": 1.1078,
"step": 5
},
{
"epoch": 0.006157635467980296,
"grad_norm": 6.2466283379921395,
"learning_rate": 1.226993865030675e-06,
"loss": 1.0505,
"step": 10
},
{
"epoch": 0.009236453201970444,
"grad_norm": 6.415224868024984,
"learning_rate": 1.8404907975460124e-06,
"loss": 0.9422,
"step": 15
},
{
"epoch": 0.012315270935960592,
"grad_norm": 4.21159705861286,
"learning_rate": 2.45398773006135e-06,
"loss": 0.9679,
"step": 20
},
{
"epoch": 0.01539408866995074,
"grad_norm": 4.31110729184732,
"learning_rate": 3.0674846625766875e-06,
"loss": 1.0075,
"step": 25
},
{
"epoch": 0.01847290640394089,
"grad_norm": 4.666213753845418,
"learning_rate": 3.680981595092025e-06,
"loss": 0.9588,
"step": 30
},
{
"epoch": 0.021551724137931036,
"grad_norm": 3.935337955011237,
"learning_rate": 4.294478527607362e-06,
"loss": 0.9337,
"step": 35
},
{
"epoch": 0.024630541871921183,
"grad_norm": 4.178703761956684,
"learning_rate": 4.9079754601227e-06,
"loss": 1.0124,
"step": 40
},
{
"epoch": 0.02770935960591133,
"grad_norm": 4.594637443489114,
"learning_rate": 5.521472392638038e-06,
"loss": 0.9178,
"step": 45
},
{
"epoch": 0.03078817733990148,
"grad_norm": 4.119966867370525,
"learning_rate": 6.134969325153375e-06,
"loss": 0.9392,
"step": 50
},
{
"epoch": 0.033866995073891626,
"grad_norm": 4.008194128332786,
"learning_rate": 6.748466257668712e-06,
"loss": 0.899,
"step": 55
},
{
"epoch": 0.03694581280788178,
"grad_norm": 3.9381003827245156,
"learning_rate": 7.36196319018405e-06,
"loss": 1.0394,
"step": 60
},
{
"epoch": 0.04002463054187192,
"grad_norm": 6.38641506050558,
"learning_rate": 7.975460122699386e-06,
"loss": 0.9469,
"step": 65
},
{
"epoch": 0.04310344827586207,
"grad_norm": 5.021312099859512,
"learning_rate": 8.588957055214725e-06,
"loss": 0.9827,
"step": 70
},
{
"epoch": 0.046182266009852216,
"grad_norm": 4.729632295314948,
"learning_rate": 9.202453987730062e-06,
"loss": 1.0044,
"step": 75
},
{
"epoch": 0.04926108374384237,
"grad_norm": 4.802753082733405,
"learning_rate": 9.8159509202454e-06,
"loss": 1.0064,
"step": 80
},
{
"epoch": 0.05233990147783251,
"grad_norm": 4.163608188743648,
"learning_rate": 1.0429447852760737e-05,
"loss": 0.9655,
"step": 85
},
{
"epoch": 0.05541871921182266,
"grad_norm": 3.8919936616069624,
"learning_rate": 1.1042944785276076e-05,
"loss": 0.9869,
"step": 90
},
{
"epoch": 0.058497536945812806,
"grad_norm": 3.7931965993281636,
"learning_rate": 1.1656441717791411e-05,
"loss": 0.9936,
"step": 95
},
{
"epoch": 0.06157635467980296,
"grad_norm": 3.629752539724635,
"learning_rate": 1.226993865030675e-05,
"loss": 0.9779,
"step": 100
},
{
"epoch": 0.06465517241379311,
"grad_norm": 3.8409355694066156,
"learning_rate": 1.2883435582822085e-05,
"loss": 1.0271,
"step": 105
},
{
"epoch": 0.06773399014778325,
"grad_norm": 3.8907579324108337,
"learning_rate": 1.3496932515337424e-05,
"loss": 1.0038,
"step": 110
},
{
"epoch": 0.0708128078817734,
"grad_norm": 4.049981630301336,
"learning_rate": 1.4110429447852763e-05,
"loss": 1.0149,
"step": 115
},
{
"epoch": 0.07389162561576355,
"grad_norm": 3.9174505355398965,
"learning_rate": 1.47239263803681e-05,
"loss": 0.9734,
"step": 120
},
{
"epoch": 0.0769704433497537,
"grad_norm": 3.9635299751402244,
"learning_rate": 1.5337423312883436e-05,
"loss": 1.0132,
"step": 125
},
{
"epoch": 0.08004926108374384,
"grad_norm": 3.8132424368148903,
"learning_rate": 1.5950920245398772e-05,
"loss": 1.0053,
"step": 130
},
{
"epoch": 0.08312807881773399,
"grad_norm": 3.9830819677934115,
"learning_rate": 1.656441717791411e-05,
"loss": 1.0691,
"step": 135
},
{
"epoch": 0.08620689655172414,
"grad_norm": 3.992267174130198,
"learning_rate": 1.717791411042945e-05,
"loss": 1.077,
"step": 140
},
{
"epoch": 0.08928571428571429,
"grad_norm": 3.6504275295174224,
"learning_rate": 1.7791411042944788e-05,
"loss": 1.0921,
"step": 145
},
{
"epoch": 0.09236453201970443,
"grad_norm": 4.071204318793922,
"learning_rate": 1.8404907975460123e-05,
"loss": 1.0861,
"step": 150
},
{
"epoch": 0.09544334975369458,
"grad_norm": 4.040027623798463,
"learning_rate": 1.9018404907975462e-05,
"loss": 1.0303,
"step": 155
},
{
"epoch": 0.09852216748768473,
"grad_norm": 4.1635142079857745,
"learning_rate": 1.96319018404908e-05,
"loss": 1.0646,
"step": 160
},
{
"epoch": 0.10160098522167488,
"grad_norm": 3.85597486126407,
"learning_rate": 1.999990752408443e-05,
"loss": 1.1117,
"step": 165
},
{
"epoch": 0.10467980295566502,
"grad_norm": 4.0881998871528396,
"learning_rate": 1.9998867189676517e-05,
"loss": 1.102,
"step": 170
},
{
"epoch": 0.10775862068965517,
"grad_norm": 3.4866367002740715,
"learning_rate": 1.999667104662347e-05,
"loss": 1.0925,
"step": 175
},
{
"epoch": 0.11083743842364532,
"grad_norm": 3.893706039967695,
"learning_rate": 1.9993319348786157e-05,
"loss": 1.1304,
"step": 180
},
{
"epoch": 0.11391625615763547,
"grad_norm": 4.0387550915074275,
"learning_rate": 1.9988812483600597e-05,
"loss": 1.0482,
"step": 185
},
{
"epoch": 0.11699507389162561,
"grad_norm": 3.8886429488865635,
"learning_rate": 1.9983150972033186e-05,
"loss": 1.1137,
"step": 190
},
{
"epoch": 0.12007389162561577,
"grad_norm": 4.385070480027836,
"learning_rate": 1.9976335468520452e-05,
"loss": 1.1425,
"step": 195
},
{
"epoch": 0.12315270935960591,
"grad_norm": 3.8660261941026697,
"learning_rate": 1.9968366760893437e-05,
"loss": 1.1345,
"step": 200
},
{
"epoch": 0.12623152709359606,
"grad_norm": 4.078216623217991,
"learning_rate": 1.9959245770286602e-05,
"loss": 1.1244,
"step": 205
},
{
"epoch": 0.12931034482758622,
"grad_norm": 4.194625427771513,
"learning_rate": 1.994897355103136e-05,
"loss": 1.0923,
"step": 210
},
{
"epoch": 0.13238916256157635,
"grad_norm": 3.768805639067731,
"learning_rate": 1.9937551290534208e-05,
"loss": 1.0704,
"step": 215
},
{
"epoch": 0.1354679802955665,
"grad_norm": 4.01482102680058,
"learning_rate": 1.9924980309139455e-05,
"loss": 1.1115,
"step": 220
},
{
"epoch": 0.13854679802955666,
"grad_norm": 3.851672790811978,
"learning_rate": 1.9911262059976614e-05,
"loss": 1.1198,
"step": 225
},
{
"epoch": 0.1416256157635468,
"grad_norm": 4.078454606346042,
"learning_rate": 1.9896398128792413e-05,
"loss": 1.1155,
"step": 230
},
{
"epoch": 0.14470443349753695,
"grad_norm": 10.140724702954135,
"learning_rate": 1.988039023376751e-05,
"loss": 1.138,
"step": 235
},
{
"epoch": 0.1477832512315271,
"grad_norm": 4.258959621761418,
"learning_rate": 1.9863240225317868e-05,
"loss": 1.1956,
"step": 240
},
{
"epoch": 0.15086206896551724,
"grad_norm": 3.2929723550833048,
"learning_rate": 1.984495008588086e-05,
"loss": 1.094,
"step": 245
},
{
"epoch": 0.1539408866995074,
"grad_norm": 3.9641040307483832,
"learning_rate": 1.982552192968612e-05,
"loss": 1.1461,
"step": 250
},
{
"epoch": 0.15701970443349753,
"grad_norm": 3.86458571871174,
"learning_rate": 1.9804958002511137e-05,
"loss": 1.1136,
"step": 255
},
{
"epoch": 0.16009852216748768,
"grad_norm": 7.792239297246812,
"learning_rate": 1.9783260681421667e-05,
"loss": 1.1341,
"step": 260
},
{
"epoch": 0.16317733990147784,
"grad_norm": 3.5904936166470685,
"learning_rate": 1.9760432474496963e-05,
"loss": 1.1218,
"step": 265
},
{
"epoch": 0.16625615763546797,
"grad_norm": 3.651660444895842,
"learning_rate": 1.973647602053984e-05,
"loss": 1.1235,
"step": 270
},
{
"epoch": 0.16933497536945813,
"grad_norm": 4.105229601726613,
"learning_rate": 1.9711394088771658e-05,
"loss": 1.1527,
"step": 275
},
{
"epoch": 0.1724137931034483,
"grad_norm": 4.895983896468738,
"learning_rate": 1.9685189578512206e-05,
"loss": 1.1452,
"step": 280
},
{
"epoch": 0.17549261083743842,
"grad_norm": 3.7372875576159243,
"learning_rate": 1.9657865518844578e-05,
"loss": 1.1101,
"step": 285
},
{
"epoch": 0.17857142857142858,
"grad_norm": 4.306528584992038,
"learning_rate": 1.962942506826501e-05,
"loss": 1.1254,
"step": 290
},
{
"epoch": 0.1816502463054187,
"grad_norm": 3.753362540475309,
"learning_rate": 1.9599871514317785e-05,
"loss": 1.1301,
"step": 295
},
{
"epoch": 0.18472906403940886,
"grad_norm": 4.326975851282091,
"learning_rate": 1.9569208273215204e-05,
"loss": 1.0988,
"step": 300
},
{
"epoch": 0.18780788177339902,
"grad_norm": 15.699939417537761,
"learning_rate": 1.953743888944271e-05,
"loss": 1.0966,
"step": 305
},
{
"epoch": 0.19088669950738915,
"grad_norm": 4.073364244471582,
"learning_rate": 1.950456703534915e-05,
"loss": 1.1672,
"step": 310
},
{
"epoch": 0.1939655172413793,
"grad_norm": 5.186383331729901,
"learning_rate": 1.9470596510722285e-05,
"loss": 1.0839,
"step": 315
},
{
"epoch": 0.19704433497536947,
"grad_norm": 4.102384568352007,
"learning_rate": 1.9435531242349545e-05,
"loss": 1.2208,
"step": 320
},
{
"epoch": 0.2001231527093596,
"grad_norm": 4.6573438540845515,
"learning_rate": 1.9399375283564134e-05,
"loss": 1.1336,
"step": 325
},
{
"epoch": 0.20320197044334976,
"grad_norm": 4.158210329894318,
"learning_rate": 1.9362132813776472e-05,
"loss": 1.1211,
"step": 330
},
{
"epoch": 0.2062807881773399,
"grad_norm": 3.6782546597796055,
"learning_rate": 1.9323808137991084e-05,
"loss": 1.1362,
"step": 335
},
{
"epoch": 0.20935960591133004,
"grad_norm": 3.797488442245091,
"learning_rate": 1.9284405686308982e-05,
"loss": 1.1101,
"step": 340
},
{
"epoch": 0.2124384236453202,
"grad_norm": 3.5436391605527056,
"learning_rate": 1.924393001341555e-05,
"loss": 1.1014,
"step": 345
},
{
"epoch": 0.21551724137931033,
"grad_norm": 3.9778714009049567,
"learning_rate": 1.9202385798054073e-05,
"loss": 1.0881,
"step": 350
},
{
"epoch": 0.2185960591133005,
"grad_norm": 3.9821081347340734,
"learning_rate": 1.9159777842484878e-05,
"loss": 1.1333,
"step": 355
},
{
"epoch": 0.22167487684729065,
"grad_norm": 3.68983403491795,
"learning_rate": 1.911611107193024e-05,
"loss": 1.1419,
"step": 360
},
{
"epoch": 0.22475369458128078,
"grad_norm": 3.4569634083709158,
"learning_rate": 1.9071390534005045e-05,
"loss": 1.1569,
"step": 365
},
{
"epoch": 0.22783251231527094,
"grad_norm": 6.612046067872354,
"learning_rate": 1.9025621398133333e-05,
"loss": 1.1434,
"step": 370
},
{
"epoch": 0.2309113300492611,
"grad_norm": 4.196018514891641,
"learning_rate": 1.8978808954950722e-05,
"loss": 1.1327,
"step": 375
},
{
"epoch": 0.23399014778325122,
"grad_norm": 3.528290894860675,
"learning_rate": 1.8930958615692854e-05,
"loss": 1.1616,
"step": 380
},
{
"epoch": 0.23706896551724138,
"grad_norm": 3.3678211092258263,
"learning_rate": 1.8882075911569887e-05,
"loss": 1.0569,
"step": 385
},
{
"epoch": 0.24014778325123154,
"grad_norm": 3.7755412664234904,
"learning_rate": 1.8832166493127128e-05,
"loss": 1.1112,
"step": 390
},
{
"epoch": 0.24322660098522167,
"grad_norm": 3.512696869964461,
"learning_rate": 1.8781236129591847e-05,
"loss": 1.0967,
"step": 395
},
{
"epoch": 0.24630541871921183,
"grad_norm": 3.7091686615796253,
"learning_rate": 1.8729290708206412e-05,
"loss": 1.1516,
"step": 400
},
{
"epoch": 0.24938423645320196,
"grad_norm": 3.556116508453499,
"learning_rate": 1.867633623354773e-05,
"loss": 1.1399,
"step": 405
},
{
"epoch": 0.2524630541871921,
"grad_norm": 3.588557880072793,
"learning_rate": 1.8622378826833186e-05,
"loss": 1.1707,
"step": 410
},
{
"epoch": 0.2555418719211823,
"grad_norm": 3.7765823975111403,
"learning_rate": 1.856742472521304e-05,
"loss": 1.154,
"step": 415
},
{
"epoch": 0.25862068965517243,
"grad_norm": 3.3290448833305772,
"learning_rate": 1.8511480281049475e-05,
"loss": 1.1221,
"step": 420
},
{
"epoch": 0.2616995073891626,
"grad_norm": 3.2508526123377472,
"learning_rate": 1.8454551961182276e-05,
"loss": 1.0971,
"step": 425
},
{
"epoch": 0.2647783251231527,
"grad_norm": 3.4156091471191226,
"learning_rate": 1.8396646346181327e-05,
"loss": 1.1352,
"step": 430
},
{
"epoch": 0.26785714285714285,
"grad_norm": 3.524814697100625,
"learning_rate": 1.8337770129585918e-05,
"loss": 1.1722,
"step": 435
},
{
"epoch": 0.270935960591133,
"grad_norm": 3.187493349372238,
"learning_rate": 1.8277930117131025e-05,
"loss": 1.1673,
"step": 440
},
{
"epoch": 0.27401477832512317,
"grad_norm": 3.2884074522724305,
"learning_rate": 1.8217133225960597e-05,
"loss": 1.0869,
"step": 445
},
{
"epoch": 0.2770935960591133,
"grad_norm": 4.149826683807174,
"learning_rate": 1.8155386483827995e-05,
"loss": 1.1292,
"step": 450
},
{
"epoch": 0.2801724137931034,
"grad_norm": 17.052593860696543,
"learning_rate": 1.8092697028283598e-05,
"loss": 1.1306,
"step": 455
},
{
"epoch": 0.2832512315270936,
"grad_norm": 3.3062225089999084,
"learning_rate": 1.8029072105849767e-05,
"loss": 1.1015,
"step": 460
},
{
"epoch": 0.28633004926108374,
"grad_norm": 3.945656027594156,
"learning_rate": 1.7964519071183188e-05,
"loss": 1.153,
"step": 465
},
{
"epoch": 0.2894088669950739,
"grad_norm": 3.7891484410248406,
"learning_rate": 1.789904538622471e-05,
"loss": 1.1286,
"step": 470
},
{
"epoch": 0.29248768472906406,
"grad_norm": 3.3275138905716575,
"learning_rate": 1.7832658619336794e-05,
"loss": 1.1379,
"step": 475
},
{
"epoch": 0.2955665024630542,
"grad_norm": 3.2403040922808746,
"learning_rate": 1.7765366444428655e-05,
"loss": 1.1404,
"step": 480
},
{
"epoch": 0.2986453201970443,
"grad_norm": 3.4281568102023177,
"learning_rate": 1.7697176640069217e-05,
"loss": 1.0973,
"step": 485
},
{
"epoch": 0.3017241379310345,
"grad_norm": 3.640944599610315,
"learning_rate": 1.762809708858793e-05,
"loss": 1.1312,
"step": 490
},
{
"epoch": 0.30480295566502463,
"grad_norm": 3.285189628705901,
"learning_rate": 1.7558135775163645e-05,
"loss": 1.1359,
"step": 495
},
{
"epoch": 0.3078817733990148,
"grad_norm": 3.9243990171686303,
"learning_rate": 1.7487300786901568e-05,
"loss": 1.0783,
"step": 500
},
{
"epoch": 0.31096059113300495,
"grad_norm": 4.195861576140969,
"learning_rate": 1.7415600311898436e-05,
"loss": 1.1746,
"step": 505
},
{
"epoch": 0.31403940886699505,
"grad_norm": 3.854423327657789,
"learning_rate": 1.734304263829602e-05,
"loss": 1.162,
"step": 510
},
{
"epoch": 0.3171182266009852,
"grad_norm": 3.882121350852715,
"learning_rate": 1.726963615332308e-05,
"loss": 1.1592,
"step": 515
},
{
"epoch": 0.32019704433497537,
"grad_norm": 3.197191626464174,
"learning_rate": 1.7195389342325843e-05,
"loss": 1.147,
"step": 520
},
{
"epoch": 0.3232758620689655,
"grad_norm": 3.272219729154457,
"learning_rate": 1.7120310787787136e-05,
"loss": 1.1042,
"step": 525
},
{
"epoch": 0.3263546798029557,
"grad_norm": 3.3755245131238514,
"learning_rate": 1.7044409168334327e-05,
"loss": 1.1234,
"step": 530
},
{
"epoch": 0.3294334975369458,
"grad_norm": 3.160894964086755,
"learning_rate": 1.696769325773611e-05,
"loss": 1.098,
"step": 535
},
{
"epoch": 0.33251231527093594,
"grad_norm": 3.287291928978594,
"learning_rate": 1.6890171923888323e-05,
"loss": 1.1156,
"step": 540
},
{
"epoch": 0.3355911330049261,
"grad_norm": 3.314816838985308,
"learning_rate": 1.6811854127788857e-05,
"loss": 1.1726,
"step": 545
},
{
"epoch": 0.33866995073891626,
"grad_norm": 3.1449775230026367,
"learning_rate": 1.6732748922501832e-05,
"loss": 1.195,
"step": 550
},
{
"epoch": 0.3417487684729064,
"grad_norm": 3.0892909073528783,
"learning_rate": 1.6652865452111115e-05,
"loss": 1.0833,
"step": 555
},
{
"epoch": 0.3448275862068966,
"grad_norm": 3.5148241235863877,
"learning_rate": 1.657221295066332e-05,
"loss": 1.1503,
"step": 560
},
{
"epoch": 0.3479064039408867,
"grad_norm": 3.0988462012904536,
"learning_rate": 1.6490800741100396e-05,
"loss": 1.1277,
"step": 565
},
{
"epoch": 0.35098522167487683,
"grad_norm": 3.57661343292914,
"learning_rate": 1.6408638234181975e-05,
"loss": 1.1287,
"step": 570
},
{
"epoch": 0.354064039408867,
"grad_norm": 3.0472132018873315,
"learning_rate": 1.6325734927397514e-05,
"loss": 1.1228,
"step": 575
},
{
"epoch": 0.35714285714285715,
"grad_norm": 2.982568519902514,
"learning_rate": 1.6242100403868472e-05,
"loss": 1.1062,
"step": 580
},
{
"epoch": 0.3602216748768473,
"grad_norm": 3.613410135680105,
"learning_rate": 1.615774433124054e-05,
"loss": 1.1194,
"step": 585
},
{
"epoch": 0.3633004926108374,
"grad_norm": 3.7090605636080807,
"learning_rate": 1.6072676460566136e-05,
"loss": 1.1434,
"step": 590
},
{
"epoch": 0.36637931034482757,
"grad_norm": 3.588372664901903,
"learning_rate": 1.5986906625177215e-05,
"loss": 1.1328,
"step": 595
},
{
"epoch": 0.3694581280788177,
"grad_norm": 3.2769423951164263,
"learning_rate": 1.590044473954863e-05,
"loss": 1.1814,
"step": 600
},
{
"epoch": 0.3725369458128079,
"grad_norm": 3.242767583172485,
"learning_rate": 1.5813300798152048e-05,
"loss": 1.1142,
"step": 605
},
{
"epoch": 0.37561576354679804,
"grad_norm": 3.061738502102936,
"learning_rate": 1.5725484874300673e-05,
"loss": 1.1541,
"step": 610
},
{
"epoch": 0.3786945812807882,
"grad_norm": 3.2368216535383825,
"learning_rate": 1.5637007118984814e-05,
"loss": 1.1176,
"step": 615
},
{
"epoch": 0.3817733990147783,
"grad_norm": 2.949253864191439,
"learning_rate": 1.5547877759698498e-05,
"loss": 1.1163,
"step": 620
},
{
"epoch": 0.38485221674876846,
"grad_norm": 3.7534336954232885,
"learning_rate": 1.5458107099257245e-05,
"loss": 1.1225,
"step": 625
},
{
"epoch": 0.3879310344827586,
"grad_norm": 3.4548119559883643,
"learning_rate": 1.5367705514607107e-05,
"loss": 1.0937,
"step": 630
},
{
"epoch": 0.3910098522167488,
"grad_norm": 3.320619549175968,
"learning_rate": 1.527668345562516e-05,
"loss": 1.1053,
"step": 635
},
{
"epoch": 0.39408866995073893,
"grad_norm": 3.687965187609941,
"learning_rate": 1.518505144391157e-05,
"loss": 1.1132,
"step": 640
},
{
"epoch": 0.39716748768472904,
"grad_norm": 3.2865367885031684,
"learning_rate": 1.5092820071573358e-05,
"loss": 1.1472,
"step": 645
},
{
"epoch": 0.4002463054187192,
"grad_norm": 3.34191572938925,
"learning_rate": 1.5000000000000002e-05,
"loss": 1.1196,
"step": 650
},
{
"epoch": 0.40332512315270935,
"grad_norm": 3.139464852124613,
"learning_rate": 1.490660195863106e-05,
"loss": 1.0889,
"step": 655
},
{
"epoch": 0.4064039408866995,
"grad_norm": 3.4649339044328338,
"learning_rate": 1.4812636743715912e-05,
"loss": 1.1097,
"step": 660
},
{
"epoch": 0.40948275862068967,
"grad_norm": 3.441398082194117,
"learning_rate": 1.4718115217065766e-05,
"loss": 1.1264,
"step": 665
},
{
"epoch": 0.4125615763546798,
"grad_norm": 8.680790262828156,
"learning_rate": 1.462304830479811e-05,
"loss": 1.0911,
"step": 670
},
{
"epoch": 0.41564039408866993,
"grad_norm": 3.2720209752403515,
"learning_rate": 1.4527446996073714e-05,
"loss": 1.1086,
"step": 675
},
{
"epoch": 0.4187192118226601,
"grad_norm": 3.3808864910732312,
"learning_rate": 1.4431322341826348e-05,
"loss": 1.0823,
"step": 680
},
{
"epoch": 0.42179802955665024,
"grad_norm": 5.895874470875083,
"learning_rate": 1.433468545348537e-05,
"loss": 1.1478,
"step": 685
},
{
"epoch": 0.4248768472906404,
"grad_norm": 3.298327972592576,
"learning_rate": 1.4237547501691298e-05,
"loss": 1.1225,
"step": 690
},
{
"epoch": 0.42795566502463056,
"grad_norm": 3.656397824200903,
"learning_rate": 1.4139919715004558e-05,
"loss": 1.1113,
"step": 695
},
{
"epoch": 0.43103448275862066,
"grad_norm": 3.275379275650544,
"learning_rate": 1.4041813378607534e-05,
"loss": 1.1237,
"step": 700
},
{
"epoch": 0.4341133004926108,
"grad_norm": 3.4602636465262337,
"learning_rate": 1.3943239833000068e-05,
"loss": 1.0995,
"step": 705
},
{
"epoch": 0.437192118226601,
"grad_norm": 3.724366842289642,
"learning_rate": 1.3844210472688557e-05,
"loss": 1.0788,
"step": 710
},
{
"epoch": 0.44027093596059114,
"grad_norm": 2.942039515863392,
"learning_rate": 1.3744736744868832e-05,
"loss": 1.0871,
"step": 715
},
{
"epoch": 0.4433497536945813,
"grad_norm": 3.138084437622199,
"learning_rate": 1.3644830148102915e-05,
"loss": 1.1348,
"step": 720
},
{
"epoch": 0.44642857142857145,
"grad_norm": 3.408982293188815,
"learning_rate": 1.3544502230989868e-05,
"loss": 1.0979,
"step": 725
},
{
"epoch": 0.44950738916256155,
"grad_norm": 3.1002717980756893,
"learning_rate": 1.344376459083085e-05,
"loss": 1.1005,
"step": 730
},
{
"epoch": 0.4525862068965517,
"grad_norm": 17.278045877818894,
"learning_rate": 1.3342628872288505e-05,
"loss": 1.1218,
"step": 735
},
{
"epoch": 0.45566502463054187,
"grad_norm": 3.272049638934581,
"learning_rate": 1.3241106766040957e-05,
"loss": 1.0785,
"step": 740
},
{
"epoch": 0.45874384236453203,
"grad_norm": 17.295405551641185,
"learning_rate": 1.3139210007430404e-05,
"loss": 1.1061,
"step": 745
},
{
"epoch": 0.4618226600985222,
"grad_norm": 3.2990052865548223,
"learning_rate": 1.3036950375106588e-05,
"loss": 1.1092,
"step": 750
},
{
"epoch": 0.4649014778325123,
"grad_norm": 4.499883081117569,
"learning_rate": 1.2934339689665274e-05,
"loss": 1.0674,
"step": 755
},
{
"epoch": 0.46798029556650245,
"grad_norm": 4.597401478352803,
"learning_rate": 1.283138981228184e-05,
"loss": 1.0939,
"step": 760
},
{
"epoch": 0.4710591133004926,
"grad_norm": 3.112132890870631,
"learning_rate": 1.2728112643340201e-05,
"loss": 1.1078,
"step": 765
},
{
"epoch": 0.47413793103448276,
"grad_norm": 5.123554337974174,
"learning_rate": 1.26245201210572e-05,
"loss": 1.0935,
"step": 770
},
{
"epoch": 0.4772167487684729,
"grad_norm": 3.8895887481118425,
"learning_rate": 1.2520624220102623e-05,
"loss": 1.093,
"step": 775
},
{
"epoch": 0.4802955665024631,
"grad_norm": 3.4882210948428756,
"learning_rate": 1.2416436950215001e-05,
"loss": 1.1131,
"step": 780
},
{
"epoch": 0.4833743842364532,
"grad_norm": 3.1639683319312994,
"learning_rate": 1.2311970354813345e-05,
"loss": 1.1663,
"step": 785
},
{
"epoch": 0.48645320197044334,
"grad_norm": 2.867109832403359,
"learning_rate": 1.220723650960502e-05,
"loss": 1.0963,
"step": 790
},
{
"epoch": 0.4895320197044335,
"grad_norm": 3.224277807128918,
"learning_rate": 1.2102247521189838e-05,
"loss": 1.067,
"step": 795
},
{
"epoch": 0.49261083743842365,
"grad_norm": 3.075861959330623,
"learning_rate": 1.199701552566064e-05,
"loss": 1.0475,
"step": 800
},
{
"epoch": 0.4956896551724138,
"grad_norm": 2.997668053101941,
"learning_rate": 1.1891552687200414e-05,
"loss": 1.1301,
"step": 805
},
{
"epoch": 0.4987684729064039,
"grad_norm": 3.132754877029608,
"learning_rate": 1.1785871196676196e-05,
"loss": 1.1388,
"step": 810
},
{
"epoch": 0.5018472906403941,
"grad_norm": 2.958248113770891,
"learning_rate": 1.167998327022988e-05,
"loss": 1.1606,
"step": 815
},
{
"epoch": 0.5049261083743842,
"grad_norm": 2.8737357739639653,
"learning_rate": 1.1573901147866108e-05,
"loss": 1.0883,
"step": 820
},
{
"epoch": 0.5080049261083743,
"grad_norm": 3.4058120438396524,
"learning_rate": 1.1467637092037399e-05,
"loss": 1.0785,
"step": 825
},
{
"epoch": 0.5110837438423645,
"grad_norm": 2.770234210397766,
"learning_rate": 1.1361203386226672e-05,
"loss": 1.1077,
"step": 830
},
{
"epoch": 0.5141625615763546,
"grad_norm": 2.994536644985356,
"learning_rate": 1.1254612333527368e-05,
"loss": 1.0519,
"step": 835
},
{
"epoch": 0.5172413793103449,
"grad_norm": 3.179306210634924,
"learning_rate": 1.1147876255221274e-05,
"loss": 1.1212,
"step": 840
},
{
"epoch": 0.520320197044335,
"grad_norm": 3.3351728688701,
"learning_rate": 1.1041007489354263e-05,
"loss": 1.0775,
"step": 845
},
{
"epoch": 0.5233990147783252,
"grad_norm": 3.07280819000468,
"learning_rate": 1.093401838931009e-05,
"loss": 1.0952,
"step": 850
},
{
"epoch": 0.5264778325123153,
"grad_norm": 2.6954433427584807,
"learning_rate": 1.0826921322382407e-05,
"loss": 1.0792,
"step": 855
},
{
"epoch": 0.5295566502463054,
"grad_norm": 2.9633204153778654,
"learning_rate": 1.071972866834519e-05,
"loss": 1.031,
"step": 860
},
{
"epoch": 0.5326354679802956,
"grad_norm": 3.30721069792099,
"learning_rate": 1.061245281802171e-05,
"loss": 1.108,
"step": 865
},
{
"epoch": 0.5357142857142857,
"grad_norm": 3.2641700714045987,
"learning_rate": 1.0505106171852226e-05,
"loss": 1.0095,
"step": 870
},
{
"epoch": 0.5387931034482759,
"grad_norm": 10.327984728125847,
"learning_rate": 1.039770113846056e-05,
"loss": 1.0472,
"step": 875
},
{
"epoch": 0.541871921182266,
"grad_norm": 3.0160222520909845,
"learning_rate": 1.0290250133219755e-05,
"loss": 1.0745,
"step": 880
},
{
"epoch": 0.5449507389162561,
"grad_norm": 2.9721114234303,
"learning_rate": 1.0182765576816916e-05,
"loss": 1.1172,
"step": 885
},
{
"epoch": 0.5480295566502463,
"grad_norm": 3.1033969859284083,
"learning_rate": 1.0075259893817465e-05,
"loss": 1.0773,
"step": 890
},
{
"epoch": 0.5511083743842364,
"grad_norm": 2.7649824268967094,
"learning_rate": 9.967745511228922e-06,
"loss": 1.0311,
"step": 895
},
{
"epoch": 0.5541871921182266,
"grad_norm": 3.4409095300232666,
"learning_rate": 9.860234857064439e-06,
"loss": 1.083,
"step": 900
},
{
"epoch": 0.5572660098522167,
"grad_norm": 3.021260984140304,
"learning_rate": 9.752740358906168e-06,
"loss": 1.092,
"step": 905
},
{
"epoch": 0.5603448275862069,
"grad_norm": 2.9279135292288307,
"learning_rate": 9.645274442468746e-06,
"loss": 1.0758,
"step": 910
},
{
"epoch": 0.5634236453201971,
"grad_norm": 3.030875547738876,
"learning_rate": 9.537849530162915e-06,
"loss": 1.0711,
"step": 915
},
{
"epoch": 0.5665024630541872,
"grad_norm": 3.1459804862679546,
"learning_rate": 9.430478039659604e-06,
"loss": 1.0596,
"step": 920
},
{
"epoch": 0.5695812807881774,
"grad_norm": 2.8645753245525913,
"learning_rate": 9.323172382454494e-06,
"loss": 1.0684,
"step": 925
},
{
"epoch": 0.5726600985221675,
"grad_norm": 2.8133898242396036,
"learning_rate": 9.215944962433329e-06,
"loss": 0.9945,
"step": 930
},
{
"epoch": 0.5757389162561576,
"grad_norm": 3.2494439673022035,
"learning_rate": 9.108808174438128e-06,
"loss": 1.0888,
"step": 935
},
{
"epoch": 0.5788177339901478,
"grad_norm": 3.4879876220279686,
"learning_rate": 9.001774402834375e-06,
"loss": 1.0256,
"step": 940
},
{
"epoch": 0.5818965517241379,
"grad_norm": 3.0362561734597957,
"learning_rate": 8.89485602007949e-06,
"loss": 1.0825,
"step": 945
},
{
"epoch": 0.5849753694581281,
"grad_norm": 3.250810792553236,
"learning_rate": 8.788065385292637e-06,
"loss": 1.0507,
"step": 950
},
{
"epoch": 0.5880541871921182,
"grad_norm": 2.997776398210721,
"learning_rate": 8.68141484282609e-06,
"loss": 1.0853,
"step": 955
},
{
"epoch": 0.5911330049261084,
"grad_norm": 3.5628103235180726,
"learning_rate": 8.574916720838293e-06,
"loss": 1.0984,
"step": 960
},
{
"epoch": 0.5942118226600985,
"grad_norm": 3.3227883638914677,
"learning_rate": 8.468583329868805e-06,
"loss": 1.1062,
"step": 965
},
{
"epoch": 0.5972906403940886,
"grad_norm": 3.3145071637154477,
"learning_rate": 8.362426961415274e-06,
"loss": 1.0777,
"step": 970
},
{
"epoch": 0.6003694581280788,
"grad_norm": 3.0550411344589654,
"learning_rate": 8.256459886512618e-06,
"loss": 1.0519,
"step": 975
},
{
"epoch": 0.603448275862069,
"grad_norm": 3.2015935398471655,
"learning_rate": 8.150694354314556e-06,
"loss": 1.1033,
"step": 980
},
{
"epoch": 0.6065270935960592,
"grad_norm": 3.485028556188331,
"learning_rate": 8.0451425906777e-06,
"loss": 1.0681,
"step": 985
},
{
"epoch": 0.6096059113300493,
"grad_norm": 3.0476445800134075,
"learning_rate": 7.939816796748296e-06,
"loss": 1.0281,
"step": 990
},
{
"epoch": 0.6126847290640394,
"grad_norm": 3.144923746254212,
"learning_rate": 7.834729147551858e-06,
"loss": 1.0612,
"step": 995
},
{
"epoch": 0.6157635467980296,
"grad_norm": 3.0052901270401824,
"learning_rate": 7.729891790585817e-06,
"loss": 1.0095,
"step": 1000
},
{
"epoch": 0.6188423645320197,
"grad_norm": 2.934306426182672,
"learning_rate": 7.625316844415327e-06,
"loss": 1.0074,
"step": 1005
},
{
"epoch": 0.6219211822660099,
"grad_norm": 3.6722713412111476,
"learning_rate": 7.521016397272436e-06,
"loss": 1.0577,
"step": 1010
},
{
"epoch": 0.625,
"grad_norm": 3.13588694551087,
"learning_rate": 7.417002505658773e-06,
"loss": 1.0982,
"step": 1015
},
{
"epoch": 0.6280788177339901,
"grad_norm": 3.0404241698093104,
"learning_rate": 7.313287192951866e-06,
"loss": 1.0631,
"step": 1020
},
{
"epoch": 0.6311576354679803,
"grad_norm": 3.1970819124393675,
"learning_rate": 7.209882448015338e-06,
"loss": 1.045,
"step": 1025
},
{
"epoch": 0.6342364532019704,
"grad_norm": 3.2738547129785966,
"learning_rate": 7.1068002238130465e-06,
"loss": 1.0533,
"step": 1030
},
{
"epoch": 0.6373152709359606,
"grad_norm": 3.1062744336330663,
"learning_rate": 7.004052436027397e-06,
"loss": 1.0432,
"step": 1035
},
{
"epoch": 0.6403940886699507,
"grad_norm": 3.231866509753256,
"learning_rate": 6.901650961681976e-06,
"loss": 1.0375,
"step": 1040
},
{
"epoch": 0.6434729064039408,
"grad_norm": 3.007322111337476,
"learning_rate": 6.799607637768621e-06,
"loss": 1.0498,
"step": 1045
},
{
"epoch": 0.646551724137931,
"grad_norm": 2.79973094934948,
"learning_rate": 6.6979342598791395e-06,
"loss": 1.0807,
"step": 1050
},
{
"epoch": 0.6496305418719212,
"grad_norm": 3.1659372475891017,
"learning_rate": 6.596642580841827e-06,
"loss": 1.0424,
"step": 1055
},
{
"epoch": 0.6527093596059114,
"grad_norm": 3.513601567378622,
"learning_rate": 6.495744309362879e-06,
"loss": 1.0168,
"step": 1060
},
{
"epoch": 0.6557881773399015,
"grad_norm": 3.1210869518303306,
"learning_rate": 6.395251108672975e-06,
"loss": 1.0178,
"step": 1065
},
{
"epoch": 0.6588669950738916,
"grad_norm": 2.90314830604983,
"learning_rate": 6.295174595179041e-06,
"loss": 1.036,
"step": 1070
},
{
"epoch": 0.6619458128078818,
"grad_norm": 3.2971277127054788,
"learning_rate": 6.195526337121483e-06,
"loss": 1.0754,
"step": 1075
},
{
"epoch": 0.6650246305418719,
"grad_norm": 2.9315751247607595,
"learning_rate": 6.096317853236975e-06,
"loss": 0.9825,
"step": 1080
},
{
"epoch": 0.6681034482758621,
"grad_norm": 3.0304652796065334,
"learning_rate": 5.997560611426947e-06,
"loss": 1.0364,
"step": 1085
},
{
"epoch": 0.6711822660098522,
"grad_norm": 3.790881779725883,
"learning_rate": 5.899266027431965e-06,
"loss": 1.0662,
"step": 1090
},
{
"epoch": 0.6742610837438424,
"grad_norm": 2.8525366943170707,
"learning_rate": 5.80144546351216e-06,
"loss": 0.9966,
"step": 1095
},
{
"epoch": 0.6773399014778325,
"grad_norm": 2.954888560101942,
"learning_rate": 5.704110227133792e-06,
"loss": 1.0385,
"step": 1100
},
{
"epoch": 0.6804187192118226,
"grad_norm": 2.84954197112955,
"learning_rate": 5.607271569662203e-06,
"loss": 1.0201,
"step": 1105
},
{
"epoch": 0.6834975369458128,
"grad_norm": 3.258827328403599,
"learning_rate": 5.510940685061202e-06,
"loss": 1.0265,
"step": 1110
},
{
"epoch": 0.6865763546798029,
"grad_norm": 3.246354458124418,
"learning_rate": 5.41512870859912e-06,
"loss": 1.0551,
"step": 1115
},
{
"epoch": 0.6896551724137931,
"grad_norm": 3.5591457944202016,
"learning_rate": 5.319846715561656e-06,
"loss": 1.0579,
"step": 1120
},
{
"epoch": 0.6927339901477833,
"grad_norm": 2.9491452747854106,
"learning_rate": 5.225105719971615e-06,
"loss": 1.0111,
"step": 1125
},
{
"epoch": 0.6958128078817734,
"grad_norm": 3.194740090743344,
"learning_rate": 5.130916673315762e-06,
"loss": 1.0657,
"step": 1130
},
{
"epoch": 0.6988916256157636,
"grad_norm": 2.904300602561608,
"learning_rate": 5.037290463278914e-06,
"loss": 1.021,
"step": 1135
},
{
"epoch": 0.7019704433497537,
"grad_norm": 3.082833149576533,
"learning_rate": 4.94423791248536e-06,
"loss": 1.0149,
"step": 1140
},
{
"epoch": 0.7050492610837439,
"grad_norm": 2.856518798210586,
"learning_rate": 4.851769777247857e-06,
"loss": 1.0762,
"step": 1145
},
{
"epoch": 0.708128078817734,
"grad_norm": 3.0965298694963868,
"learning_rate": 4.759896746324247e-06,
"loss": 1.0273,
"step": 1150
},
{
"epoch": 0.7112068965517241,
"grad_norm": 2.8259971969585624,
"learning_rate": 4.668629439681907e-06,
"loss": 1.0449,
"step": 1155
},
{
"epoch": 0.7142857142857143,
"grad_norm": 2.926544221652753,
"learning_rate": 4.577978407270156e-06,
"loss": 1.0513,
"step": 1160
},
{
"epoch": 0.7173645320197044,
"grad_norm": 4.440971093938638,
"learning_rate": 4.487954127800726e-06,
"loss": 1.0264,
"step": 1165
},
{
"epoch": 0.7204433497536946,
"grad_norm": 2.9730831964061832,
"learning_rate": 4.398567007536493e-06,
"loss": 1.087,
"step": 1170
},
{
"epoch": 0.7235221674876847,
"grad_norm": 2.764796339014182,
"learning_rate": 4.309827379088589e-06,
"loss": 1.0262,
"step": 1175
},
{
"epoch": 0.7266009852216748,
"grad_norm": 3.1698729349677275,
"learning_rate": 4.221745500221992e-06,
"loss": 0.9967,
"step": 1180
},
{
"epoch": 0.729679802955665,
"grad_norm": 3.2707913443905587,
"learning_rate": 4.134331552669812e-06,
"loss": 1.0684,
"step": 1185
},
{
"epoch": 0.7327586206896551,
"grad_norm": 2.9190647733844135,
"learning_rate": 4.047595640956326e-06,
"loss": 1.0383,
"step": 1190
},
{
"epoch": 0.7358374384236454,
"grad_norm": 2.920602079542603,
"learning_rate": 3.961547791228963e-06,
"loss": 0.9809,
"step": 1195
},
{
"epoch": 0.7389162561576355,
"grad_norm": 2.9297071969995883,
"learning_rate": 3.876197950099351e-06,
"loss": 0.9973,
"step": 1200
},
{
"epoch": 0.7419950738916257,
"grad_norm": 3.0268815588920086,
"learning_rate": 3.7915559834935355e-06,
"loss": 1.0361,
"step": 1205
},
{
"epoch": 0.7450738916256158,
"grad_norm": 3.0226493271136055,
"learning_rate": 3.7076316755115407e-06,
"loss": 1.0132,
"step": 1210
},
{
"epoch": 0.7481527093596059,
"grad_norm": 2.975777490352993,
"learning_rate": 3.6244347272963974e-06,
"loss": 1.0227,
"step": 1215
},
{
"epoch": 0.7512315270935961,
"grad_norm": 3.2531462819381924,
"learning_rate": 3.5419747559127294e-06,
"loss": 1.029,
"step": 1220
},
{
"epoch": 0.7543103448275862,
"grad_norm": 2.933887832357094,
"learning_rate": 3.4602612932351065e-06,
"loss": 1.0397,
"step": 1225
},
{
"epoch": 0.7573891625615764,
"grad_norm": 3.225771278101853,
"learning_rate": 3.3793037848461873e-06,
"loss": 1.0031,
"step": 1230
},
{
"epoch": 0.7604679802955665,
"grad_norm": 3.0822132779428313,
"learning_rate": 3.2991115889448877e-06,
"loss": 1.0152,
"step": 1235
},
{
"epoch": 0.7635467980295566,
"grad_norm": 2.8782654123233566,
"learning_rate": 3.2196939752646183e-06,
"loss": 1.0161,
"step": 1240
},
{
"epoch": 0.7666256157635468,
"grad_norm": 3.207017045075226,
"learning_rate": 3.141060124001776e-06,
"loss": 1.0331,
"step": 1245
},
{
"epoch": 0.7697044334975369,
"grad_norm": 3.006885129221621,
"learning_rate": 3.063219124754543e-06,
"loss": 1.0185,
"step": 1250
},
{
"epoch": 0.7727832512315271,
"grad_norm": 2.9936192104802073,
"learning_rate": 2.9861799754722033e-06,
"loss": 0.9919,
"step": 1255
},
{
"epoch": 0.7758620689655172,
"grad_norm": 3.1127606400696077,
"learning_rate": 2.9099515814150335e-06,
"loss": 0.9851,
"step": 1260
},
{
"epoch": 0.7789408866995073,
"grad_norm": 3.105938780311132,
"learning_rate": 2.8345427541248993e-06,
"loss": 0.9745,
"step": 1265
},
{
"epoch": 0.7820197044334976,
"grad_norm": 2.692901045482411,
"learning_rate": 2.7599622104066937e-06,
"loss": 1.0484,
"step": 1270
},
{
"epoch": 0.7850985221674877,
"grad_norm": 2.8591413978889966,
"learning_rate": 2.6862185713207467e-06,
"loss": 1.0181,
"step": 1275
},
{
"epoch": 0.7881773399014779,
"grad_norm": 3.141024770129843,
"learning_rate": 2.6133203611862554e-06,
"loss": 0.9986,
"step": 1280
},
{
"epoch": 0.791256157635468,
"grad_norm": 2.7093877839779283,
"learning_rate": 2.5412760065959386e-06,
"loss": 1.0032,
"step": 1285
},
{
"epoch": 0.7943349753694581,
"grad_norm": 2.6389276156236985,
"learning_rate": 2.4700938354419823e-06,
"loss": 0.9894,
"step": 1290
},
{
"epoch": 0.7974137931034483,
"grad_norm": 3.031879834724528,
"learning_rate": 2.3997820759533654e-06,
"loss": 1.02,
"step": 1295
},
{
"epoch": 0.8004926108374384,
"grad_norm": 2.9357036501776514,
"learning_rate": 2.3303488557447374e-06,
"loss": 0.9764,
"step": 1300
},
{
"epoch": 0.8035714285714286,
"grad_norm": 3.0411835757416994,
"learning_rate": 2.26180220087692e-06,
"loss": 0.9981,
"step": 1305
},
{
"epoch": 0.8066502463054187,
"grad_norm": 2.9846054359780316,
"learning_rate": 2.194150034929133e-06,
"loss": 1.0144,
"step": 1310
},
{
"epoch": 0.8097290640394089,
"grad_norm": 3.0082859977783047,
"learning_rate": 2.1274001780830776e-06,
"loss": 1.0371,
"step": 1315
},
{
"epoch": 0.812807881773399,
"grad_norm": 2.9061280452250537,
"learning_rate": 2.0615603462189824e-06,
"loss": 1.0334,
"step": 1320
},
{
"epoch": 0.8158866995073891,
"grad_norm": 3.156684287503527,
"learning_rate": 1.9966381500236786e-06,
"loss": 0.9631,
"step": 1325
},
{
"epoch": 0.8189655172413793,
"grad_norm": 2.8124160392042428,
"learning_rate": 1.932641094110855e-06,
"loss": 1.0026,
"step": 1330
},
{
"epoch": 0.8220443349753694,
"grad_norm": 2.9788470369935323,
"learning_rate": 1.869576576153581e-06,
"loss": 0.9504,
"step": 1335
},
{
"epoch": 0.8251231527093597,
"grad_norm": 3.1432151528869228,
"learning_rate": 1.8074518860291646e-06,
"loss": 1.0159,
"step": 1340
},
{
"epoch": 0.8282019704433498,
"grad_norm": 4.199947622169966,
"learning_rate": 1.746274204976498e-06,
"loss": 1.0151,
"step": 1345
},
{
"epoch": 0.8312807881773399,
"grad_norm": 3.0295798860885723,
"learning_rate": 1.6860506047659442e-06,
"loss": 1.0179,
"step": 1350
},
{
"epoch": 0.8343596059113301,
"grad_norm": 3.048559496196436,
"learning_rate": 1.6267880468818787e-06,
"loss": 0.9803,
"step": 1355
},
{
"epoch": 0.8374384236453202,
"grad_norm": 3.0170419912915243,
"learning_rate": 1.5684933817180014e-06,
"loss": 1.0048,
"step": 1360
},
{
"epoch": 0.8405172413793104,
"grad_norm": 3.295111747998583,
"learning_rate": 1.5111733477854507e-06,
"loss": 1.0255,
"step": 1365
},
{
"epoch": 0.8435960591133005,
"grad_norm": 2.7573975682591576,
"learning_rate": 1.454834570933884e-06,
"loss": 0.9872,
"step": 1370
},
{
"epoch": 0.8466748768472906,
"grad_norm": 2.9681857274891805,
"learning_rate": 1.399483563585573e-06,
"loss": 0.9924,
"step": 1375
},
{
"epoch": 0.8497536945812808,
"grad_norm": 3.3019052202961725,
"learning_rate": 1.345126723982594e-06,
"loss": 1.0681,
"step": 1380
},
{
"epoch": 0.8528325123152709,
"grad_norm": 3.061278743842955,
"learning_rate": 1.2917703354472467e-06,
"loss": 0.9942,
"step": 1385
},
{
"epoch": 0.8559113300492611,
"grad_norm": 3.0707996456504993,
"learning_rate": 1.2394205656557224e-06,
"loss": 0.9719,
"step": 1390
},
{
"epoch": 0.8589901477832512,
"grad_norm": 2.8229314400382104,
"learning_rate": 1.1880834659251706e-06,
"loss": 1.0067,
"step": 1395
},
{
"epoch": 0.8620689655172413,
"grad_norm": 3.1104952330602504,
"learning_rate": 1.1377649705142012e-06,
"loss": 1.0015,
"step": 1400
},
{
"epoch": 0.8651477832512315,
"grad_norm": 3.177017032481223,
"learning_rate": 1.0884708959369116e-06,
"loss": 1.0355,
"step": 1405
},
{
"epoch": 0.8682266009852216,
"grad_norm": 2.8878442026210247,
"learning_rate": 1.040206940290547e-06,
"loss": 0.9765,
"step": 1410
},
{
"epoch": 0.8713054187192119,
"grad_norm": 2.7376219446157597,
"learning_rate": 9.929786825968213e-07,
"loss": 0.992,
"step": 1415
},
{
"epoch": 0.874384236453202,
"grad_norm": 3.3638865553088557,
"learning_rate": 9.467915821570228e-07,
"loss": 1.0484,
"step": 1420
},
{
"epoch": 0.8774630541871922,
"grad_norm": 3.045132136096421,
"learning_rate": 9.016509779209536e-07,
"loss": 1.0107,
"step": 1425
},
{
"epoch": 0.8805418719211823,
"grad_norm": 3.4420152557153,
"learning_rate": 8.575620878697744e-07,
"loss": 0.9689,
"step": 1430
},
{
"epoch": 0.8836206896551724,
"grad_norm": 3.0414602766772663,
"learning_rate": 8.145300084128349e-07,
"loss": 1.0062,
"step": 1435
},
{
"epoch": 0.8866995073891626,
"grad_norm": 2.668099678425544,
"learning_rate": 7.725597137985741e-07,
"loss": 0.9514,
"step": 1440
},
{
"epoch": 0.8897783251231527,
"grad_norm": 2.937074165368976,
"learning_rate": 7.316560555395069e-07,
"loss": 0.9664,
"step": 1445
},
{
"epoch": 0.8928571428571429,
"grad_norm": 2.9267866902532393,
"learning_rate": 6.918237618514378e-07,
"loss": 0.9464,
"step": 1450
},
{
"epoch": 0.895935960591133,
"grad_norm": 2.899665156091937,
"learning_rate": 6.530674371068946e-07,
"loss": 0.9955,
"step": 1455
},
{
"epoch": 0.8990147783251231,
"grad_norm": 3.49062205552801,
"learning_rate": 6.153915613028915e-07,
"loss": 0.9914,
"step": 1460
},
{
"epoch": 0.9020935960591133,
"grad_norm": 2.7915828222108954,
"learning_rate": 5.788004895430799e-07,
"loss": 0.9601,
"step": 1465
},
{
"epoch": 0.9051724137931034,
"grad_norm": 3.011491247234021,
"learning_rate": 5.43298451534312e-07,
"loss": 1.0462,
"step": 1470
},
{
"epoch": 0.9082512315270936,
"grad_norm": 2.9498211031946533,
"learning_rate": 5.088895510977154e-07,
"loss": 0.9752,
"step": 1475
},
{
"epoch": 0.9113300492610837,
"grad_norm": 2.7665957742386613,
"learning_rate": 4.755777656943239e-07,
"loss": 0.9998,
"step": 1480
},
{
"epoch": 0.9144088669950738,
"grad_norm": 3.2520133328690295,
"learning_rate": 4.433669459652945e-07,
"loss": 0.9742,
"step": 1485
},
{
"epoch": 0.9174876847290641,
"grad_norm": 3.098137802269807,
"learning_rate": 4.1226081528680907e-07,
"loss": 1.0136,
"step": 1490
},
{
"epoch": 0.9205665024630542,
"grad_norm": 2.830576924784849,
"learning_rate": 3.822629693396651e-07,
"loss": 0.9708,
"step": 1495
},
{
"epoch": 0.9236453201970444,
"grad_norm": 3.3994115488707237,
"learning_rate": 3.5337687569363734e-07,
"loss": 1.0167,
"step": 1500
},
{
"epoch": 0.9267241379310345,
"grad_norm": 3.0103914691893014,
"learning_rate": 3.2560587340665694e-07,
"loss": 1.0011,
"step": 1505
},
{
"epoch": 0.9298029556650246,
"grad_norm": 3.3519220072935565,
"learning_rate": 2.989531726388262e-07,
"loss": 1.0026,
"step": 1510
},
{
"epoch": 0.9328817733990148,
"grad_norm": 3.150785383803172,
"learning_rate": 2.7342185428134604e-07,
"loss": 1.007,
"step": 1515
},
{
"epoch": 0.9359605911330049,
"grad_norm": 2.5884532661862356,
"learning_rate": 2.4901486960039025e-07,
"loss": 1.0052,
"step": 1520
},
{
"epoch": 0.9390394088669951,
"grad_norm": 3.297888826907389,
"learning_rate": 2.257350398959457e-07,
"loss": 0.9834,
"step": 1525
},
{
"epoch": 0.9421182266009852,
"grad_norm": 3.3590473746519356,
"learning_rate": 2.035850561756969e-07,
"loss": 1.0001,
"step": 1530
},
{
"epoch": 0.9451970443349754,
"grad_norm": 3.1703193136587347,
"learning_rate": 1.8256747884395577e-07,
"loss": 0.9944,
"step": 1535
},
{
"epoch": 0.9482758620689655,
"grad_norm": 2.890061326174278,
"learning_rate": 1.6268473740569723e-07,
"loss": 1.0598,
"step": 1540
},
{
"epoch": 0.9513546798029556,
"grad_norm": 3.410088768269518,
"learning_rate": 1.4393913018572182e-07,
"loss": 0.9718,
"step": 1545
},
{
"epoch": 0.9544334975369458,
"grad_norm": 2.961401630871241,
"learning_rate": 1.2633282406298576e-07,
"loss": 0.9845,
"step": 1550
},
{
"epoch": 0.9575123152709359,
"grad_norm": 2.900907749720405,
"learning_rate": 1.0986785422011593e-07,
"loss": 0.9852,
"step": 1555
},
{
"epoch": 0.9605911330049262,
"grad_norm": 2.7677132240531934,
"learning_rate": 9.454612390816686e-08,
"loss": 0.9936,
"step": 1560
},
{
"epoch": 0.9636699507389163,
"grad_norm": 2.8750875541860035,
"learning_rate": 8.036940422660345e-08,
"loss": 1.0016,
"step": 1565
},
{
"epoch": 0.9667487684729064,
"grad_norm": 3.0158244079592484,
"learning_rate": 6.733933391858238e-08,
"loss": 1.0278,
"step": 1570
},
{
"epoch": 0.9698275862068966,
"grad_norm": 2.986649180793057,
"learning_rate": 5.5457419181517145e-08,
"loss": 0.9841,
"step": 1575
},
{
"epoch": 0.9729064039408867,
"grad_norm": 2.7712739447981996,
"learning_rate": 4.472503349297497e-08,
"loss": 0.9599,
"step": 1580
},
{
"epoch": 0.9759852216748769,
"grad_norm": 3.034332488316358,
"learning_rate": 3.5143417451907195e-08,
"loss": 0.94,
"step": 1585
},
{
"epoch": 0.979064039408867,
"grad_norm": 3.1089847502543173,
"learning_rate": 2.671367863524732e-08,
"loss": 1.0172,
"step": 1590
},
{
"epoch": 0.9821428571428571,
"grad_norm": 3.0903160037664534,
"learning_rate": 1.943679146988009e-08,
"loss": 1.0085,
"step": 1595
},
{
"epoch": 0.9852216748768473,
"grad_norm": 3.218218364976497,
"learning_rate": 1.3313597120002686e-08,
"loss": 1.0207,
"step": 1600
},
{
"epoch": 0.9883004926108374,
"grad_norm": 2.995299761218313,
"learning_rate": 8.34480338989141e-09,
"loss": 1.017,
"step": 1605
},
{
"epoch": 0.9913793103448276,
"grad_norm": 2.7187289642325423,
"learning_rate": 4.530984642087122e-09,
"loss": 0.9361,
"step": 1610
},
{
"epoch": 0.9944581280788177,
"grad_norm": 2.689389109504255,
"learning_rate": 1.872581730997247e-09,
"loss": 1.0238,
"step": 1615
},
{
"epoch": 0.9975369458128078,
"grad_norm": 3.0072927851931466,
"learning_rate": 3.699019519387559e-10,
"loss": 0.9547,
"step": 1620
},
{
"epoch": 1.0,
"eval_loss": 0.9928293228149414,
"eval_runtime": 1399.3933,
"eval_samples_per_second": 16.514,
"eval_steps_per_second": 0.517,
"step": 1624
},
{
"epoch": 1.0,
"step": 1624,
"total_flos": 163322853457920.0,
"train_loss": 1.062036202605722,
"train_runtime": 12545.798,
"train_samples_per_second": 4.142,
"train_steps_per_second": 0.129
}
],
"logging_steps": 5,
"max_steps": 1624,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 163322853457920.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}