OpenELM-1_1B-SFT-3 / trainer_state.json
CharlesLi's picture
Model save
b72d8f3 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 6849,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0004380201489268506,
"grad_norm": 244.05783349993243,
"learning_rate": 2.9197080291970803e-08,
"loss": 8.7812,
"step": 1
},
{
"epoch": 0.002190100744634253,
"grad_norm": 242.80510323947792,
"learning_rate": 1.4598540145985402e-07,
"loss": 8.8047,
"step": 5
},
{
"epoch": 0.004380201489268506,
"grad_norm": 225.0335670574099,
"learning_rate": 2.9197080291970804e-07,
"loss": 8.7531,
"step": 10
},
{
"epoch": 0.006570302233902759,
"grad_norm": 228.72582215453716,
"learning_rate": 4.379562043795621e-07,
"loss": 8.7125,
"step": 15
},
{
"epoch": 0.008760402978537012,
"grad_norm": 202.96584235540402,
"learning_rate": 5.839416058394161e-07,
"loss": 8.5219,
"step": 20
},
{
"epoch": 0.010950503723171266,
"grad_norm": 125.01523628364068,
"learning_rate": 7.299270072992701e-07,
"loss": 8.2953,
"step": 25
},
{
"epoch": 0.013140604467805518,
"grad_norm": 110.40505502297347,
"learning_rate": 8.759124087591242e-07,
"loss": 7.7859,
"step": 30
},
{
"epoch": 0.015330705212439772,
"grad_norm": 68.45912276297945,
"learning_rate": 1.0218978102189781e-06,
"loss": 7.3016,
"step": 35
},
{
"epoch": 0.017520805957074025,
"grad_norm": 55.590663092336015,
"learning_rate": 1.1678832116788322e-06,
"loss": 6.7047,
"step": 40
},
{
"epoch": 0.01971090670170828,
"grad_norm": 40.0844953425019,
"learning_rate": 1.3138686131386864e-06,
"loss": 6.3906,
"step": 45
},
{
"epoch": 0.021901007446342532,
"grad_norm": 34.48948950173287,
"learning_rate": 1.4598540145985402e-06,
"loss": 5.575,
"step": 50
},
{
"epoch": 0.024091108190976786,
"grad_norm": 31.68892363070654,
"learning_rate": 1.6058394160583942e-06,
"loss": 5.1031,
"step": 55
},
{
"epoch": 0.026281208935611037,
"grad_norm": 24.216002249213382,
"learning_rate": 1.7518248175182485e-06,
"loss": 4.5187,
"step": 60
},
{
"epoch": 0.02847130968024529,
"grad_norm": 16.153964519214966,
"learning_rate": 1.8978102189781023e-06,
"loss": 3.6203,
"step": 65
},
{
"epoch": 0.030661410424879545,
"grad_norm": 13.836195759408959,
"learning_rate": 2.0437956204379563e-06,
"loss": 3.1273,
"step": 70
},
{
"epoch": 0.0328515111695138,
"grad_norm": 11.652340272071394,
"learning_rate": 2.1897810218978103e-06,
"loss": 2.6828,
"step": 75
},
{
"epoch": 0.03504161191414805,
"grad_norm": 6.333091358966209,
"learning_rate": 2.3357664233576643e-06,
"loss": 2.2922,
"step": 80
},
{
"epoch": 0.03723171265878231,
"grad_norm": 2.9939785076399814,
"learning_rate": 2.4817518248175183e-06,
"loss": 2.0016,
"step": 85
},
{
"epoch": 0.03942181340341656,
"grad_norm": 2.6606355077846895,
"learning_rate": 2.627737226277373e-06,
"loss": 1.8633,
"step": 90
},
{
"epoch": 0.04161191414805081,
"grad_norm": 1.2346022297538288,
"learning_rate": 2.7737226277372264e-06,
"loss": 1.7285,
"step": 95
},
{
"epoch": 0.043802014892685065,
"grad_norm": 0.9865905253164456,
"learning_rate": 2.9197080291970804e-06,
"loss": 1.5969,
"step": 100
},
{
"epoch": 0.045992115637319315,
"grad_norm": 0.8076275483624458,
"learning_rate": 3.065693430656935e-06,
"loss": 1.5781,
"step": 105
},
{
"epoch": 0.04818221638195357,
"grad_norm": 0.6036339872996322,
"learning_rate": 3.2116788321167884e-06,
"loss": 1.5227,
"step": 110
},
{
"epoch": 0.05037231712658782,
"grad_norm": 0.5227962690437455,
"learning_rate": 3.3576642335766425e-06,
"loss": 1.466,
"step": 115
},
{
"epoch": 0.052562417871222074,
"grad_norm": 0.5546021102398396,
"learning_rate": 3.503649635036497e-06,
"loss": 1.4375,
"step": 120
},
{
"epoch": 0.05475251861585633,
"grad_norm": 0.5037444771795104,
"learning_rate": 3.6496350364963505e-06,
"loss": 1.4559,
"step": 125
},
{
"epoch": 0.05694261936049058,
"grad_norm": 0.4581953982365608,
"learning_rate": 3.7956204379562045e-06,
"loss": 1.3863,
"step": 130
},
{
"epoch": 0.05913272010512484,
"grad_norm": 0.42553885665065,
"learning_rate": 3.9416058394160585e-06,
"loss": 1.3832,
"step": 135
},
{
"epoch": 0.06132282084975909,
"grad_norm": 0.4179093460042933,
"learning_rate": 4.0875912408759126e-06,
"loss": 1.3824,
"step": 140
},
{
"epoch": 0.06351292159439334,
"grad_norm": 0.40273353738694206,
"learning_rate": 4.233576642335767e-06,
"loss": 1.3465,
"step": 145
},
{
"epoch": 0.0657030223390276,
"grad_norm": 0.41160581205347213,
"learning_rate": 4.379562043795621e-06,
"loss": 1.3945,
"step": 150
},
{
"epoch": 0.06789312308366185,
"grad_norm": 0.3949551182499691,
"learning_rate": 4.525547445255475e-06,
"loss": 1.3434,
"step": 155
},
{
"epoch": 0.0700832238282961,
"grad_norm": 0.387740320945072,
"learning_rate": 4.671532846715329e-06,
"loss": 1.3551,
"step": 160
},
{
"epoch": 0.07227332457293036,
"grad_norm": 0.37700600156165553,
"learning_rate": 4.8175182481751835e-06,
"loss": 1.3637,
"step": 165
},
{
"epoch": 0.07446342531756461,
"grad_norm": 0.36760005726354067,
"learning_rate": 4.963503649635037e-06,
"loss": 1.3211,
"step": 170
},
{
"epoch": 0.07665352606219886,
"grad_norm": 0.36407987579057216,
"learning_rate": 5.1094890510948916e-06,
"loss": 1.3168,
"step": 175
},
{
"epoch": 0.07884362680683311,
"grad_norm": 0.3639367357256408,
"learning_rate": 5.255474452554746e-06,
"loss": 1.3477,
"step": 180
},
{
"epoch": 0.08103372755146737,
"grad_norm": 0.3695352411819165,
"learning_rate": 5.401459854014599e-06,
"loss": 1.3062,
"step": 185
},
{
"epoch": 0.08322382829610162,
"grad_norm": 0.33882020026658927,
"learning_rate": 5.547445255474453e-06,
"loss": 1.3195,
"step": 190
},
{
"epoch": 0.08541392904073587,
"grad_norm": 0.3519909386154269,
"learning_rate": 5.693430656934307e-06,
"loss": 1.298,
"step": 195
},
{
"epoch": 0.08760402978537013,
"grad_norm": 0.3677025946352492,
"learning_rate": 5.839416058394161e-06,
"loss": 1.3055,
"step": 200
},
{
"epoch": 0.08979413053000437,
"grad_norm": 0.3553999324591866,
"learning_rate": 5.985401459854016e-06,
"loss": 1.3102,
"step": 205
},
{
"epoch": 0.09198423127463863,
"grad_norm": 0.339876890506739,
"learning_rate": 6.13138686131387e-06,
"loss": 1.2914,
"step": 210
},
{
"epoch": 0.09417433201927289,
"grad_norm": 0.36515602431235566,
"learning_rate": 6.277372262773723e-06,
"loss": 1.2844,
"step": 215
},
{
"epoch": 0.09636443276390715,
"grad_norm": 0.3937869670816605,
"learning_rate": 6.423357664233577e-06,
"loss": 1.3137,
"step": 220
},
{
"epoch": 0.09855453350854139,
"grad_norm": 0.34727218148619815,
"learning_rate": 6.569343065693431e-06,
"loss": 1.2637,
"step": 225
},
{
"epoch": 0.10074463425317565,
"grad_norm": 0.3582494230319763,
"learning_rate": 6.715328467153285e-06,
"loss": 1.2852,
"step": 230
},
{
"epoch": 0.1029347349978099,
"grad_norm": 0.3570600153368335,
"learning_rate": 6.86131386861314e-06,
"loss": 1.2887,
"step": 235
},
{
"epoch": 0.10512483574244415,
"grad_norm": 0.3346226350011156,
"learning_rate": 7.007299270072994e-06,
"loss": 1.2633,
"step": 240
},
{
"epoch": 0.1073149364870784,
"grad_norm": 0.33400332391330406,
"learning_rate": 7.153284671532848e-06,
"loss": 1.2336,
"step": 245
},
{
"epoch": 0.10950503723171266,
"grad_norm": 0.3315126078930196,
"learning_rate": 7.299270072992701e-06,
"loss": 1.2664,
"step": 250
},
{
"epoch": 0.1116951379763469,
"grad_norm": 0.3709768940217129,
"learning_rate": 7.445255474452555e-06,
"loss": 1.2621,
"step": 255
},
{
"epoch": 0.11388523872098116,
"grad_norm": 0.32948286354246736,
"learning_rate": 7.591240875912409e-06,
"loss": 1.2867,
"step": 260
},
{
"epoch": 0.11607533946561542,
"grad_norm": 0.33294579346715825,
"learning_rate": 7.737226277372264e-06,
"loss": 1.2348,
"step": 265
},
{
"epoch": 0.11826544021024968,
"grad_norm": 0.323499147467889,
"learning_rate": 7.883211678832117e-06,
"loss": 1.2281,
"step": 270
},
{
"epoch": 0.12045554095488392,
"grad_norm": 0.3615814763070892,
"learning_rate": 8.029197080291972e-06,
"loss": 1.2637,
"step": 275
},
{
"epoch": 0.12264564169951818,
"grad_norm": 0.3538080065403599,
"learning_rate": 8.175182481751825e-06,
"loss": 1.2617,
"step": 280
},
{
"epoch": 0.12483574244415244,
"grad_norm": 0.336790971357565,
"learning_rate": 8.32116788321168e-06,
"loss": 1.2527,
"step": 285
},
{
"epoch": 0.12702584318878668,
"grad_norm": 0.37615337138920746,
"learning_rate": 8.467153284671533e-06,
"loss": 1.2527,
"step": 290
},
{
"epoch": 0.12921594393342092,
"grad_norm": 0.33532555778207584,
"learning_rate": 8.613138686131386e-06,
"loss": 1.223,
"step": 295
},
{
"epoch": 0.1314060446780552,
"grad_norm": 0.33649845965070907,
"learning_rate": 8.759124087591241e-06,
"loss": 1.2332,
"step": 300
},
{
"epoch": 0.13359614542268944,
"grad_norm": 0.3202250374399337,
"learning_rate": 8.905109489051096e-06,
"loss": 1.2219,
"step": 305
},
{
"epoch": 0.1357862461673237,
"grad_norm": 0.34514025665060105,
"learning_rate": 9.05109489051095e-06,
"loss": 1.2477,
"step": 310
},
{
"epoch": 0.13797634691195795,
"grad_norm": 0.33020924127428763,
"learning_rate": 9.197080291970804e-06,
"loss": 1.2105,
"step": 315
},
{
"epoch": 0.1401664476565922,
"grad_norm": 0.35152454516434467,
"learning_rate": 9.343065693430657e-06,
"loss": 1.1941,
"step": 320
},
{
"epoch": 0.14235654840122647,
"grad_norm": 0.316946041818773,
"learning_rate": 9.48905109489051e-06,
"loss": 1.2348,
"step": 325
},
{
"epoch": 0.1445466491458607,
"grad_norm": 0.3171098592158982,
"learning_rate": 9.635036496350367e-06,
"loss": 1.2172,
"step": 330
},
{
"epoch": 0.14673674989049496,
"grad_norm": 0.32347075273927617,
"learning_rate": 9.78102189781022e-06,
"loss": 1.2082,
"step": 335
},
{
"epoch": 0.14892685063512923,
"grad_norm": 0.31992621701805507,
"learning_rate": 9.927007299270073e-06,
"loss": 1.2039,
"step": 340
},
{
"epoch": 0.15111695137976347,
"grad_norm": 0.33669777465454753,
"learning_rate": 1.0072992700729928e-05,
"loss": 1.2199,
"step": 345
},
{
"epoch": 0.1533070521243977,
"grad_norm": 0.3242034424279669,
"learning_rate": 1.0218978102189783e-05,
"loss": 1.2219,
"step": 350
},
{
"epoch": 0.15549715286903198,
"grad_norm": 0.3383473976099628,
"learning_rate": 1.0364963503649636e-05,
"loss": 1.2082,
"step": 355
},
{
"epoch": 0.15768725361366623,
"grad_norm": 0.3375586020448848,
"learning_rate": 1.0510948905109491e-05,
"loss": 1.2156,
"step": 360
},
{
"epoch": 0.15987735435830047,
"grad_norm": 0.3277741096524666,
"learning_rate": 1.0656934306569344e-05,
"loss": 1.207,
"step": 365
},
{
"epoch": 0.16206745510293474,
"grad_norm": 0.3337717738586561,
"learning_rate": 1.0802919708029198e-05,
"loss": 1.2176,
"step": 370
},
{
"epoch": 0.164257555847569,
"grad_norm": 0.34816058383267184,
"learning_rate": 1.0948905109489052e-05,
"loss": 1.2023,
"step": 375
},
{
"epoch": 0.16644765659220323,
"grad_norm": 0.31393085773851587,
"learning_rate": 1.1094890510948906e-05,
"loss": 1.2066,
"step": 380
},
{
"epoch": 0.1686377573368375,
"grad_norm": 0.34170466792783555,
"learning_rate": 1.124087591240876e-05,
"loss": 1.2457,
"step": 385
},
{
"epoch": 0.17082785808147175,
"grad_norm": 0.32289309385714443,
"learning_rate": 1.1386861313868614e-05,
"loss": 1.2234,
"step": 390
},
{
"epoch": 0.173017958826106,
"grad_norm": 0.31607508887925223,
"learning_rate": 1.1532846715328467e-05,
"loss": 1.2184,
"step": 395
},
{
"epoch": 0.17520805957074026,
"grad_norm": 0.3324717224401286,
"learning_rate": 1.1678832116788322e-05,
"loss": 1.2223,
"step": 400
},
{
"epoch": 0.1773981603153745,
"grad_norm": 0.29789000161051715,
"learning_rate": 1.1824817518248176e-05,
"loss": 1.1777,
"step": 405
},
{
"epoch": 0.17958826106000875,
"grad_norm": 0.31662586926381675,
"learning_rate": 1.1970802919708031e-05,
"loss": 1.198,
"step": 410
},
{
"epoch": 0.18177836180464302,
"grad_norm": 0.31124419804470477,
"learning_rate": 1.2116788321167885e-05,
"loss": 1.1855,
"step": 415
},
{
"epoch": 0.18396846254927726,
"grad_norm": 0.31302760748427677,
"learning_rate": 1.226277372262774e-05,
"loss": 1.1934,
"step": 420
},
{
"epoch": 0.18615856329391153,
"grad_norm": 0.34309852924466416,
"learning_rate": 1.2408759124087593e-05,
"loss": 1.2152,
"step": 425
},
{
"epoch": 0.18834866403854578,
"grad_norm": 0.319117428903726,
"learning_rate": 1.2554744525547446e-05,
"loss": 1.2125,
"step": 430
},
{
"epoch": 0.19053876478318002,
"grad_norm": 0.33718547337261756,
"learning_rate": 1.27007299270073e-05,
"loss": 1.207,
"step": 435
},
{
"epoch": 0.1927288655278143,
"grad_norm": 0.32668740457102435,
"learning_rate": 1.2846715328467154e-05,
"loss": 1.1836,
"step": 440
},
{
"epoch": 0.19491896627244854,
"grad_norm": 0.32088829919202544,
"learning_rate": 1.2992700729927009e-05,
"loss": 1.1941,
"step": 445
},
{
"epoch": 0.19710906701708278,
"grad_norm": 0.3431063641531018,
"learning_rate": 1.3138686131386862e-05,
"loss": 1.2031,
"step": 450
},
{
"epoch": 0.19929916776171705,
"grad_norm": 0.3268567982657968,
"learning_rate": 1.3284671532846715e-05,
"loss": 1.1941,
"step": 455
},
{
"epoch": 0.2014892685063513,
"grad_norm": 0.31848928369863305,
"learning_rate": 1.343065693430657e-05,
"loss": 1.1812,
"step": 460
},
{
"epoch": 0.20367936925098554,
"grad_norm": 0.31588925727105266,
"learning_rate": 1.3576642335766423e-05,
"loss": 1.1785,
"step": 465
},
{
"epoch": 0.2058694699956198,
"grad_norm": 0.32886857710303746,
"learning_rate": 1.372262773722628e-05,
"loss": 1.1816,
"step": 470
},
{
"epoch": 0.20805957074025405,
"grad_norm": 0.30015782909456273,
"learning_rate": 1.3868613138686133e-05,
"loss": 1.1641,
"step": 475
},
{
"epoch": 0.2102496714848883,
"grad_norm": 0.30042222734891383,
"learning_rate": 1.4014598540145988e-05,
"loss": 1.1879,
"step": 480
},
{
"epoch": 0.21243977222952257,
"grad_norm": 0.3082053551352324,
"learning_rate": 1.416058394160584e-05,
"loss": 1.1762,
"step": 485
},
{
"epoch": 0.2146298729741568,
"grad_norm": 0.3069511061152226,
"learning_rate": 1.4306569343065696e-05,
"loss": 1.1852,
"step": 490
},
{
"epoch": 0.21681997371879105,
"grad_norm": 0.30587098461412204,
"learning_rate": 1.4452554744525549e-05,
"loss": 1.1887,
"step": 495
},
{
"epoch": 0.21901007446342532,
"grad_norm": 0.3472745496182846,
"learning_rate": 1.4598540145985402e-05,
"loss": 1.1621,
"step": 500
},
{
"epoch": 0.22120017520805957,
"grad_norm": 0.3383694408481445,
"learning_rate": 1.4744525547445257e-05,
"loss": 1.1691,
"step": 505
},
{
"epoch": 0.2233902759526938,
"grad_norm": 0.30917693397844287,
"learning_rate": 1.489051094890511e-05,
"loss": 1.2016,
"step": 510
},
{
"epoch": 0.22558037669732808,
"grad_norm": 0.32724498784049777,
"learning_rate": 1.5036496350364965e-05,
"loss": 1.1824,
"step": 515
},
{
"epoch": 0.22777047744196233,
"grad_norm": 0.30742723190833204,
"learning_rate": 1.5182481751824818e-05,
"loss": 1.1797,
"step": 520
},
{
"epoch": 0.22996057818659657,
"grad_norm": 0.34475225249504454,
"learning_rate": 1.5328467153284673e-05,
"loss": 1.184,
"step": 525
},
{
"epoch": 0.23215067893123084,
"grad_norm": 0.3129106984746414,
"learning_rate": 1.5474452554744528e-05,
"loss": 1.2012,
"step": 530
},
{
"epoch": 0.23434077967586509,
"grad_norm": 0.31049881643452515,
"learning_rate": 1.5620437956204383e-05,
"loss": 1.1604,
"step": 535
},
{
"epoch": 0.23653088042049936,
"grad_norm": 0.3086241541949012,
"learning_rate": 1.5766423357664234e-05,
"loss": 1.1648,
"step": 540
},
{
"epoch": 0.2387209811651336,
"grad_norm": 0.3189864576022323,
"learning_rate": 1.591240875912409e-05,
"loss": 1.1602,
"step": 545
},
{
"epoch": 0.24091108190976784,
"grad_norm": 0.32841841575252745,
"learning_rate": 1.6058394160583944e-05,
"loss": 1.166,
"step": 550
},
{
"epoch": 0.24310118265440211,
"grad_norm": 0.3238032151207836,
"learning_rate": 1.62043795620438e-05,
"loss": 1.1871,
"step": 555
},
{
"epoch": 0.24529128339903636,
"grad_norm": 0.3105247743890357,
"learning_rate": 1.635036496350365e-05,
"loss": 1.193,
"step": 560
},
{
"epoch": 0.2474813841436706,
"grad_norm": 0.326200553936659,
"learning_rate": 1.6496350364963505e-05,
"loss": 1.1703,
"step": 565
},
{
"epoch": 0.24967148488830487,
"grad_norm": 0.3137884080668668,
"learning_rate": 1.664233576642336e-05,
"loss": 1.1707,
"step": 570
},
{
"epoch": 0.2518615856329391,
"grad_norm": 0.327163922779202,
"learning_rate": 1.678832116788321e-05,
"loss": 1.148,
"step": 575
},
{
"epoch": 0.25405168637757336,
"grad_norm": 0.3063941514339094,
"learning_rate": 1.6934306569343066e-05,
"loss": 1.1613,
"step": 580
},
{
"epoch": 0.25624178712220763,
"grad_norm": 0.3108041781813456,
"learning_rate": 1.708029197080292e-05,
"loss": 1.1535,
"step": 585
},
{
"epoch": 0.25843188786684185,
"grad_norm": 0.3133301402975079,
"learning_rate": 1.7226277372262773e-05,
"loss": 1.1914,
"step": 590
},
{
"epoch": 0.2606219886114761,
"grad_norm": 0.3154737069033297,
"learning_rate": 1.737226277372263e-05,
"loss": 1.1836,
"step": 595
},
{
"epoch": 0.2628120893561104,
"grad_norm": 0.3011426797487246,
"learning_rate": 1.7518248175182482e-05,
"loss": 1.1559,
"step": 600
},
{
"epoch": 0.26500219010074466,
"grad_norm": 0.3416641336195657,
"learning_rate": 1.7664233576642337e-05,
"loss": 1.1572,
"step": 605
},
{
"epoch": 0.2671922908453789,
"grad_norm": 0.30613307226505193,
"learning_rate": 1.7810218978102192e-05,
"loss": 1.1762,
"step": 610
},
{
"epoch": 0.26938239159001315,
"grad_norm": 0.29981894574558626,
"learning_rate": 1.7956204379562047e-05,
"loss": 1.1562,
"step": 615
},
{
"epoch": 0.2715724923346474,
"grad_norm": 0.29887010220063054,
"learning_rate": 1.81021897810219e-05,
"loss": 1.1645,
"step": 620
},
{
"epoch": 0.27376259307928164,
"grad_norm": 0.2966732076878118,
"learning_rate": 1.8248175182481753e-05,
"loss": 1.1824,
"step": 625
},
{
"epoch": 0.2759526938239159,
"grad_norm": 0.29935159320241506,
"learning_rate": 1.8394160583941608e-05,
"loss": 1.1684,
"step": 630
},
{
"epoch": 0.2781427945685502,
"grad_norm": 0.2884901367267061,
"learning_rate": 1.854014598540146e-05,
"loss": 1.1625,
"step": 635
},
{
"epoch": 0.2803328953131844,
"grad_norm": 0.29675051316205636,
"learning_rate": 1.8686131386861315e-05,
"loss": 1.1711,
"step": 640
},
{
"epoch": 0.28252299605781866,
"grad_norm": 0.2994167858357806,
"learning_rate": 1.883211678832117e-05,
"loss": 1.1691,
"step": 645
},
{
"epoch": 0.28471309680245294,
"grad_norm": 0.3058796858332876,
"learning_rate": 1.897810218978102e-05,
"loss": 1.1621,
"step": 650
},
{
"epoch": 0.28690319754708715,
"grad_norm": 0.2972115649292112,
"learning_rate": 1.912408759124088e-05,
"loss": 1.1687,
"step": 655
},
{
"epoch": 0.2890932982917214,
"grad_norm": 0.3246749282257886,
"learning_rate": 1.9270072992700734e-05,
"loss": 1.176,
"step": 660
},
{
"epoch": 0.2912833990363557,
"grad_norm": 0.2949778092608126,
"learning_rate": 1.9416058394160586e-05,
"loss": 1.1602,
"step": 665
},
{
"epoch": 0.2934734997809899,
"grad_norm": 0.29601206389636064,
"learning_rate": 1.956204379562044e-05,
"loss": 1.1543,
"step": 670
},
{
"epoch": 0.2956636005256242,
"grad_norm": 0.31730205467856953,
"learning_rate": 1.9708029197080295e-05,
"loss": 1.1484,
"step": 675
},
{
"epoch": 0.29785370127025845,
"grad_norm": 0.29248038800784815,
"learning_rate": 1.9854014598540147e-05,
"loss": 1.1629,
"step": 680
},
{
"epoch": 0.30004380201489267,
"grad_norm": 0.324442556976409,
"learning_rate": 2e-05,
"loss": 1.1504,
"step": 685
},
{
"epoch": 0.30223390275952694,
"grad_norm": 0.29652947714633526,
"learning_rate": 1.9999967529852368e-05,
"loss": 1.134,
"step": 690
},
{
"epoch": 0.3044240035041612,
"grad_norm": 0.28765128738836787,
"learning_rate": 1.9999870119620323e-05,
"loss": 1.1391,
"step": 695
},
{
"epoch": 0.3066141042487954,
"grad_norm": 0.2977547290672057,
"learning_rate": 1.999970776993646e-05,
"loss": 1.1738,
"step": 700
},
{
"epoch": 0.3088042049934297,
"grad_norm": 0.3092291985228049,
"learning_rate": 1.9999480481855074e-05,
"loss": 1.166,
"step": 705
},
{
"epoch": 0.31099430573806397,
"grad_norm": 0.3056767165612742,
"learning_rate": 1.9999188256852184e-05,
"loss": 1.1516,
"step": 710
},
{
"epoch": 0.3131844064826982,
"grad_norm": 0.2963548040389902,
"learning_rate": 1.999883109682551e-05,
"loss": 1.1465,
"step": 715
},
{
"epoch": 0.31537450722733246,
"grad_norm": 0.30234744446219075,
"learning_rate": 1.9998409004094455e-05,
"loss": 1.1414,
"step": 720
},
{
"epoch": 0.31756460797196673,
"grad_norm": 0.30066223568227457,
"learning_rate": 1.999792198140011e-05,
"loss": 1.1504,
"step": 725
},
{
"epoch": 0.31975470871660094,
"grad_norm": 0.2889539765396387,
"learning_rate": 1.9997370031905204e-05,
"loss": 1.1687,
"step": 730
},
{
"epoch": 0.3219448094612352,
"grad_norm": 0.3119167096280213,
"learning_rate": 1.999675315919412e-05,
"loss": 1.1258,
"step": 735
},
{
"epoch": 0.3241349102058695,
"grad_norm": 0.3082566009232655,
"learning_rate": 1.9996071367272846e-05,
"loss": 1.1398,
"step": 740
},
{
"epoch": 0.3263250109505037,
"grad_norm": 0.3190515342314494,
"learning_rate": 1.999532466056896e-05,
"loss": 1.1422,
"step": 745
},
{
"epoch": 0.328515111695138,
"grad_norm": 0.3137645553915184,
"learning_rate": 1.9994513043931596e-05,
"loss": 1.1367,
"step": 750
},
{
"epoch": 0.33070521243977224,
"grad_norm": 0.309872007079329,
"learning_rate": 1.9993636522631417e-05,
"loss": 1.1664,
"step": 755
},
{
"epoch": 0.33289531318440646,
"grad_norm": 0.27597463143785256,
"learning_rate": 1.9992695102360576e-05,
"loss": 1.1438,
"step": 760
},
{
"epoch": 0.33508541392904073,
"grad_norm": 0.29218301221271276,
"learning_rate": 1.9991688789232687e-05,
"loss": 1.1551,
"step": 765
},
{
"epoch": 0.337275514673675,
"grad_norm": 0.30069339984117943,
"learning_rate": 1.9990617589782778e-05,
"loss": 1.1582,
"step": 770
},
{
"epoch": 0.3394656154183092,
"grad_norm": 0.3092042415654092,
"learning_rate": 1.9989481510967245e-05,
"loss": 1.1572,
"step": 775
},
{
"epoch": 0.3416557161629435,
"grad_norm": 0.286252883483907,
"learning_rate": 1.9988280560163825e-05,
"loss": 1.1359,
"step": 780
},
{
"epoch": 0.34384581690757776,
"grad_norm": 0.2965378350747857,
"learning_rate": 1.998701474517152e-05,
"loss": 1.1404,
"step": 785
},
{
"epoch": 0.346035917652212,
"grad_norm": 0.32399058891091354,
"learning_rate": 1.9985684074210573e-05,
"loss": 1.141,
"step": 790
},
{
"epoch": 0.34822601839684625,
"grad_norm": 0.29631463717134965,
"learning_rate": 1.9984288555922402e-05,
"loss": 1.141,
"step": 795
},
{
"epoch": 0.3504161191414805,
"grad_norm": 0.3066454043832438,
"learning_rate": 1.998282819936954e-05,
"loss": 1.1346,
"step": 800
},
{
"epoch": 0.35260621988611474,
"grad_norm": 0.3002290039251923,
"learning_rate": 1.9981303014035593e-05,
"loss": 1.1344,
"step": 805
},
{
"epoch": 0.354796320630749,
"grad_norm": 0.31658840783678355,
"learning_rate": 1.9979713009825152e-05,
"loss": 1.1422,
"step": 810
},
{
"epoch": 0.3569864213753833,
"grad_norm": 0.28116056868919015,
"learning_rate": 1.9978058197063756e-05,
"loss": 1.1645,
"step": 815
},
{
"epoch": 0.3591765221200175,
"grad_norm": 0.2988466651254511,
"learning_rate": 1.9976338586497804e-05,
"loss": 1.1383,
"step": 820
},
{
"epoch": 0.36136662286465177,
"grad_norm": 0.29802628898129574,
"learning_rate": 1.9974554189294504e-05,
"loss": 1.1344,
"step": 825
},
{
"epoch": 0.36355672360928604,
"grad_norm": 0.2847009818824244,
"learning_rate": 1.9972705017041777e-05,
"loss": 1.1383,
"step": 830
},
{
"epoch": 0.3657468243539203,
"grad_norm": 0.3028159290266467,
"learning_rate": 1.9970791081748208e-05,
"loss": 1.1316,
"step": 835
},
{
"epoch": 0.3679369250985545,
"grad_norm": 0.287479339437057,
"learning_rate": 1.9968812395842942e-05,
"loss": 1.1543,
"step": 840
},
{
"epoch": 0.3701270258431888,
"grad_norm": 0.28447649711277756,
"learning_rate": 1.996676897217563e-05,
"loss": 1.0986,
"step": 845
},
{
"epoch": 0.37231712658782307,
"grad_norm": 0.28951032395742343,
"learning_rate": 1.9964660824016327e-05,
"loss": 1.1508,
"step": 850
},
{
"epoch": 0.3745072273324573,
"grad_norm": 0.2977760588633454,
"learning_rate": 1.9962487965055403e-05,
"loss": 1.1322,
"step": 855
},
{
"epoch": 0.37669732807709155,
"grad_norm": 0.2955483265627134,
"learning_rate": 1.996025040940347e-05,
"loss": 1.1285,
"step": 860
},
{
"epoch": 0.3788874288217258,
"grad_norm": 0.287939559842973,
"learning_rate": 1.9957948171591286e-05,
"loss": 1.15,
"step": 865
},
{
"epoch": 0.38107752956636004,
"grad_norm": 0.30141508811651624,
"learning_rate": 1.9955581266569648e-05,
"loss": 1.1465,
"step": 870
},
{
"epoch": 0.3832676303109943,
"grad_norm": 0.28984961674810866,
"learning_rate": 1.9953149709709304e-05,
"loss": 1.1328,
"step": 875
},
{
"epoch": 0.3854577310556286,
"grad_norm": 0.2693443491254494,
"learning_rate": 1.995065351680086e-05,
"loss": 1.1348,
"step": 880
},
{
"epoch": 0.3876478318002628,
"grad_norm": 0.28451110876798263,
"learning_rate": 1.9948092704054664e-05,
"loss": 1.1684,
"step": 885
},
{
"epoch": 0.38983793254489707,
"grad_norm": 0.27443501432665496,
"learning_rate": 1.9945467288100714e-05,
"loss": 1.1422,
"step": 890
},
{
"epoch": 0.39202803328953134,
"grad_norm": 0.285419754074599,
"learning_rate": 1.9942777285988527e-05,
"loss": 1.1064,
"step": 895
},
{
"epoch": 0.39421813403416556,
"grad_norm": 0.283204897016455,
"learning_rate": 1.9940022715187074e-05,
"loss": 1.1402,
"step": 900
},
{
"epoch": 0.39640823477879983,
"grad_norm": 0.2975243255381765,
"learning_rate": 1.9937203593584604e-05,
"loss": 1.1422,
"step": 905
},
{
"epoch": 0.3985983355234341,
"grad_norm": 0.286375682431649,
"learning_rate": 1.993431993948858e-05,
"loss": 1.1395,
"step": 910
},
{
"epoch": 0.4007884362680683,
"grad_norm": 0.2935563141727228,
"learning_rate": 1.9931371771625545e-05,
"loss": 1.15,
"step": 915
},
{
"epoch": 0.4029785370127026,
"grad_norm": 0.2667947999658235,
"learning_rate": 1.9928359109140976e-05,
"loss": 1.1398,
"step": 920
},
{
"epoch": 0.40516863775733686,
"grad_norm": 0.2743643642469239,
"learning_rate": 1.99252819715992e-05,
"loss": 1.125,
"step": 925
},
{
"epoch": 0.4073587385019711,
"grad_norm": 0.2791187567058821,
"learning_rate": 1.9922140378983235e-05,
"loss": 1.116,
"step": 930
},
{
"epoch": 0.40954883924660535,
"grad_norm": 0.2798610384611816,
"learning_rate": 1.991893435169468e-05,
"loss": 1.1199,
"step": 935
},
{
"epoch": 0.4117389399912396,
"grad_norm": 0.28612583032215144,
"learning_rate": 1.991566391055357e-05,
"loss": 1.1344,
"step": 940
},
{
"epoch": 0.41392904073587383,
"grad_norm": 0.30128821800137673,
"learning_rate": 1.991232907679824e-05,
"loss": 1.1574,
"step": 945
},
{
"epoch": 0.4161191414805081,
"grad_norm": 0.2819831413759113,
"learning_rate": 1.9908929872085205e-05,
"loss": 1.1438,
"step": 950
},
{
"epoch": 0.4183092422251424,
"grad_norm": 0.2758868673995733,
"learning_rate": 1.9905466318489e-05,
"loss": 1.1332,
"step": 955
},
{
"epoch": 0.4204993429697766,
"grad_norm": 0.2720381391855681,
"learning_rate": 1.990193843850205e-05,
"loss": 1.1363,
"step": 960
},
{
"epoch": 0.42268944371441086,
"grad_norm": 0.2844557339068452,
"learning_rate": 1.9898346255034503e-05,
"loss": 1.1555,
"step": 965
},
{
"epoch": 0.42487954445904513,
"grad_norm": 0.2892067533735414,
"learning_rate": 1.9894689791414102e-05,
"loss": 1.1207,
"step": 970
},
{
"epoch": 0.42706964520367935,
"grad_norm": 0.28260412206082436,
"learning_rate": 1.9890969071386038e-05,
"loss": 1.1281,
"step": 975
},
{
"epoch": 0.4292597459483136,
"grad_norm": 0.27462057678308865,
"learning_rate": 1.9887184119112776e-05,
"loss": 1.1555,
"step": 980
},
{
"epoch": 0.4314498466929479,
"grad_norm": 0.2777174448301392,
"learning_rate": 1.9883334959173905e-05,
"loss": 1.1336,
"step": 985
},
{
"epoch": 0.4336399474375821,
"grad_norm": 0.29745925368348247,
"learning_rate": 1.9879421616565984e-05,
"loss": 1.1445,
"step": 990
},
{
"epoch": 0.4358300481822164,
"grad_norm": 0.2977067104078218,
"learning_rate": 1.9875444116702377e-05,
"loss": 1.1336,
"step": 995
},
{
"epoch": 0.43802014892685065,
"grad_norm": 0.29064501481596583,
"learning_rate": 1.987140248541308e-05,
"loss": 1.1187,
"step": 1000
},
{
"epoch": 0.44021024967148487,
"grad_norm": 0.28202894919724386,
"learning_rate": 1.9867296748944575e-05,
"loss": 1.1426,
"step": 1005
},
{
"epoch": 0.44240035041611914,
"grad_norm": 0.2671051370902169,
"learning_rate": 1.9863126933959623e-05,
"loss": 1.1398,
"step": 1010
},
{
"epoch": 0.4445904511607534,
"grad_norm": 0.2829190919447246,
"learning_rate": 1.985889306753714e-05,
"loss": 1.1262,
"step": 1015
},
{
"epoch": 0.4467805519053876,
"grad_norm": 0.27938603807782114,
"learning_rate": 1.9854595177171968e-05,
"loss": 1.1396,
"step": 1020
},
{
"epoch": 0.4489706526500219,
"grad_norm": 0.2726811187740653,
"learning_rate": 1.9850233290774744e-05,
"loss": 1.1283,
"step": 1025
},
{
"epoch": 0.45116075339465617,
"grad_norm": 0.2811174352400791,
"learning_rate": 1.984580743667168e-05,
"loss": 1.1625,
"step": 1030
},
{
"epoch": 0.4533508541392904,
"grad_norm": 0.28726266639080117,
"learning_rate": 1.9841317643604404e-05,
"loss": 1.1379,
"step": 1035
},
{
"epoch": 0.45554095488392465,
"grad_norm": 0.2790715625514297,
"learning_rate": 1.9836763940729765e-05,
"loss": 1.1465,
"step": 1040
},
{
"epoch": 0.4577310556285589,
"grad_norm": 0.2756123869189833,
"learning_rate": 1.9832146357619647e-05,
"loss": 1.132,
"step": 1045
},
{
"epoch": 0.45992115637319314,
"grad_norm": 0.28910724603592386,
"learning_rate": 1.982746492426077e-05,
"loss": 1.1613,
"step": 1050
},
{
"epoch": 0.4621112571178274,
"grad_norm": 0.28038299192799726,
"learning_rate": 1.9822719671054498e-05,
"loss": 1.1141,
"step": 1055
},
{
"epoch": 0.4643013578624617,
"grad_norm": 0.29312368250159127,
"learning_rate": 1.981791062881665e-05,
"loss": 1.1715,
"step": 1060
},
{
"epoch": 0.4664914586070959,
"grad_norm": 0.28833172213041786,
"learning_rate": 1.9813037828777276e-05,
"loss": 1.1328,
"step": 1065
},
{
"epoch": 0.46868155935173017,
"grad_norm": 0.2951339006961912,
"learning_rate": 1.98081013025805e-05,
"loss": 1.1234,
"step": 1070
},
{
"epoch": 0.47087166009636444,
"grad_norm": 0.2797023938943954,
"learning_rate": 1.980310108228426e-05,
"loss": 1.1385,
"step": 1075
},
{
"epoch": 0.4730617608409987,
"grad_norm": 0.2759952236533748,
"learning_rate": 1.9798037200360135e-05,
"loss": 1.1221,
"step": 1080
},
{
"epoch": 0.47525186158563293,
"grad_norm": 0.28412146048511483,
"learning_rate": 1.9792909689693124e-05,
"loss": 1.1246,
"step": 1085
},
{
"epoch": 0.4774419623302672,
"grad_norm": 0.29446603696048546,
"learning_rate": 1.978771858358143e-05,
"loss": 1.148,
"step": 1090
},
{
"epoch": 0.47963206307490147,
"grad_norm": 0.27023047414464874,
"learning_rate": 1.9782463915736255e-05,
"loss": 1.108,
"step": 1095
},
{
"epoch": 0.4818221638195357,
"grad_norm": 0.2834629828500795,
"learning_rate": 1.977714572028157e-05,
"loss": 1.116,
"step": 1100
},
{
"epoch": 0.48401226456416996,
"grad_norm": 0.2928628093709197,
"learning_rate": 1.977176403175388e-05,
"loss": 1.1217,
"step": 1105
},
{
"epoch": 0.48620236530880423,
"grad_norm": 0.2798052929515212,
"learning_rate": 1.976631888510204e-05,
"loss": 1.125,
"step": 1110
},
{
"epoch": 0.48839246605343845,
"grad_norm": 0.2815012858261469,
"learning_rate": 1.976081031568699e-05,
"loss": 1.132,
"step": 1115
},
{
"epoch": 0.4905825667980727,
"grad_norm": 0.28767140618274256,
"learning_rate": 1.9755238359281545e-05,
"loss": 1.117,
"step": 1120
},
{
"epoch": 0.492772667542707,
"grad_norm": 0.2818985849019624,
"learning_rate": 1.9749603052070147e-05,
"loss": 1.1289,
"step": 1125
},
{
"epoch": 0.4949627682873412,
"grad_norm": 0.2786359761353917,
"learning_rate": 1.9743904430648654e-05,
"loss": 1.1086,
"step": 1130
},
{
"epoch": 0.4971528690319755,
"grad_norm": 0.30082948310927515,
"learning_rate": 1.973814253202408e-05,
"loss": 1.1326,
"step": 1135
},
{
"epoch": 0.49934296977660975,
"grad_norm": 0.28823463423411355,
"learning_rate": 1.9732317393614365e-05,
"loss": 1.1418,
"step": 1140
},
{
"epoch": 0.501533070521244,
"grad_norm": 0.2835666012503442,
"learning_rate": 1.972642905324813e-05,
"loss": 1.1066,
"step": 1145
},
{
"epoch": 0.5037231712658782,
"grad_norm": 0.28208696915543563,
"learning_rate": 1.9720477549164435e-05,
"loss": 1.0969,
"step": 1150
},
{
"epoch": 0.5059132720105125,
"grad_norm": 0.27273009984309005,
"learning_rate": 1.9714462920012515e-05,
"loss": 1.1406,
"step": 1155
},
{
"epoch": 0.5081033727551467,
"grad_norm": 0.2833154120822429,
"learning_rate": 1.9708385204851552e-05,
"loss": 1.1293,
"step": 1160
},
{
"epoch": 0.5102934734997809,
"grad_norm": 0.2734595441879901,
"learning_rate": 1.9702244443150412e-05,
"loss": 1.1037,
"step": 1165
},
{
"epoch": 0.5124835742444153,
"grad_norm": 0.28386641782695426,
"learning_rate": 1.9696040674787383e-05,
"loss": 1.1309,
"step": 1170
},
{
"epoch": 0.5146736749890495,
"grad_norm": 0.2770001252770981,
"learning_rate": 1.9689773940049913e-05,
"loss": 1.116,
"step": 1175
},
{
"epoch": 0.5168637757336837,
"grad_norm": 0.2738518763278271,
"learning_rate": 1.9683444279634364e-05,
"loss": 1.1316,
"step": 1180
},
{
"epoch": 0.519053876478318,
"grad_norm": 0.2758683563717482,
"learning_rate": 1.9677051734645744e-05,
"loss": 1.134,
"step": 1185
},
{
"epoch": 0.5212439772229522,
"grad_norm": 0.2709237152430827,
"learning_rate": 1.9670596346597425e-05,
"loss": 1.1512,
"step": 1190
},
{
"epoch": 0.5234340779675866,
"grad_norm": 0.2970767298116832,
"learning_rate": 1.966407815741089e-05,
"loss": 1.1371,
"step": 1195
},
{
"epoch": 0.5256241787122208,
"grad_norm": 0.28308210048437704,
"learning_rate": 1.9657497209415443e-05,
"loss": 1.098,
"step": 1200
},
{
"epoch": 0.527814279456855,
"grad_norm": 0.28960642709808776,
"learning_rate": 1.9650853545347967e-05,
"loss": 1.1119,
"step": 1205
},
{
"epoch": 0.5300043802014893,
"grad_norm": 0.275714098894066,
"learning_rate": 1.96441472083526e-05,
"loss": 1.1066,
"step": 1210
},
{
"epoch": 0.5321944809461235,
"grad_norm": 0.27558043103307944,
"learning_rate": 1.9637378241980504e-05,
"loss": 1.1391,
"step": 1215
},
{
"epoch": 0.5343845816907578,
"grad_norm": 0.27922191954409015,
"learning_rate": 1.9630546690189544e-05,
"loss": 1.1238,
"step": 1220
},
{
"epoch": 0.5365746824353921,
"grad_norm": 0.27077948716119843,
"learning_rate": 1.9623652597344015e-05,
"loss": 1.091,
"step": 1225
},
{
"epoch": 0.5387647831800263,
"grad_norm": 0.2722865027554513,
"learning_rate": 1.961669600821436e-05,
"loss": 1.1236,
"step": 1230
},
{
"epoch": 0.5409548839246605,
"grad_norm": 0.28922569582983776,
"learning_rate": 1.9609676967976873e-05,
"loss": 1.1262,
"step": 1235
},
{
"epoch": 0.5431449846692948,
"grad_norm": 0.2965889559255691,
"learning_rate": 1.9602595522213414e-05,
"loss": 1.1086,
"step": 1240
},
{
"epoch": 0.545335085413929,
"grad_norm": 0.2762851160005322,
"learning_rate": 1.9595451716911097e-05,
"loss": 1.1309,
"step": 1245
},
{
"epoch": 0.5475251861585633,
"grad_norm": 0.2669273952815999,
"learning_rate": 1.958824559846201e-05,
"loss": 1.1297,
"step": 1250
},
{
"epoch": 0.5497152869031976,
"grad_norm": 0.2730652771013225,
"learning_rate": 1.958097721366289e-05,
"loss": 1.11,
"step": 1255
},
{
"epoch": 0.5519053876478318,
"grad_norm": 0.2842957029219278,
"learning_rate": 1.957364660971485e-05,
"loss": 1.1273,
"step": 1260
},
{
"epoch": 0.554095488392466,
"grad_norm": 0.2840418180943608,
"learning_rate": 1.9566253834223042e-05,
"loss": 1.1006,
"step": 1265
},
{
"epoch": 0.5562855891371004,
"grad_norm": 0.27712160762674326,
"learning_rate": 1.9558798935196376e-05,
"loss": 1.1135,
"step": 1270
},
{
"epoch": 0.5584756898817346,
"grad_norm": 0.2857322595691661,
"learning_rate": 1.955128196104718e-05,
"loss": 1.107,
"step": 1275
},
{
"epoch": 0.5606657906263688,
"grad_norm": 0.28868118619459854,
"learning_rate": 1.9543702960590903e-05,
"loss": 1.1141,
"step": 1280
},
{
"epoch": 0.5628558913710031,
"grad_norm": 0.26786400432490776,
"learning_rate": 1.9536061983045812e-05,
"loss": 1.1277,
"step": 1285
},
{
"epoch": 0.5650459921156373,
"grad_norm": 0.27964141513626756,
"learning_rate": 1.9528359078032624e-05,
"loss": 1.1027,
"step": 1290
},
{
"epoch": 0.5672360928602715,
"grad_norm": 0.27097789156538216,
"learning_rate": 1.9520594295574245e-05,
"loss": 1.1102,
"step": 1295
},
{
"epoch": 0.5694261936049059,
"grad_norm": 0.2881923750285588,
"learning_rate": 1.9512767686095397e-05,
"loss": 1.1137,
"step": 1300
},
{
"epoch": 0.5716162943495401,
"grad_norm": 0.28928505042046504,
"learning_rate": 1.9504879300422308e-05,
"loss": 1.1414,
"step": 1305
},
{
"epoch": 0.5738063950941743,
"grad_norm": 0.2870532574710609,
"learning_rate": 1.9496929189782393e-05,
"loss": 1.1092,
"step": 1310
},
{
"epoch": 0.5759964958388086,
"grad_norm": 0.28512356312967324,
"learning_rate": 1.9488917405803902e-05,
"loss": 1.1244,
"step": 1315
},
{
"epoch": 0.5781865965834428,
"grad_norm": 0.31392762060438606,
"learning_rate": 1.94808440005156e-05,
"loss": 1.1406,
"step": 1320
},
{
"epoch": 0.5803766973280771,
"grad_norm": 0.2767171970254651,
"learning_rate": 1.947270902634642e-05,
"loss": 1.1107,
"step": 1325
},
{
"epoch": 0.5825667980727114,
"grad_norm": 0.27800092456236136,
"learning_rate": 1.946451253612512e-05,
"loss": 1.1324,
"step": 1330
},
{
"epoch": 0.5847568988173456,
"grad_norm": 0.2885057649952374,
"learning_rate": 1.9456254583079954e-05,
"loss": 1.085,
"step": 1335
},
{
"epoch": 0.5869469995619798,
"grad_norm": 0.2761292875453245,
"learning_rate": 1.9447935220838308e-05,
"loss": 1.1422,
"step": 1340
},
{
"epoch": 0.5891371003066141,
"grad_norm": 0.28880466519196507,
"learning_rate": 1.943955450342637e-05,
"loss": 1.1199,
"step": 1345
},
{
"epoch": 0.5913272010512484,
"grad_norm": 0.2700534828962362,
"learning_rate": 1.9431112485268764e-05,
"loss": 1.1172,
"step": 1350
},
{
"epoch": 0.5935173017958826,
"grad_norm": 0.2747262063569435,
"learning_rate": 1.9422609221188208e-05,
"loss": 1.1137,
"step": 1355
},
{
"epoch": 0.5957074025405169,
"grad_norm": 0.26972085330648815,
"learning_rate": 1.9414044766405146e-05,
"loss": 1.1125,
"step": 1360
},
{
"epoch": 0.5978975032851511,
"grad_norm": 0.2855247406870803,
"learning_rate": 1.9405419176537405e-05,
"loss": 1.1207,
"step": 1365
},
{
"epoch": 0.6000876040297853,
"grad_norm": 0.2601474614595161,
"learning_rate": 1.9396732507599816e-05,
"loss": 1.1211,
"step": 1370
},
{
"epoch": 0.6022777047744197,
"grad_norm": 0.2854935378780939,
"learning_rate": 1.9387984816003868e-05,
"loss": 1.1094,
"step": 1375
},
{
"epoch": 0.6044678055190539,
"grad_norm": 0.2679659367257591,
"learning_rate": 1.937917615855732e-05,
"loss": 1.1199,
"step": 1380
},
{
"epoch": 0.6066579062636881,
"grad_norm": 0.2817414130754905,
"learning_rate": 1.9370306592463866e-05,
"loss": 1.125,
"step": 1385
},
{
"epoch": 0.6088480070083224,
"grad_norm": 0.2598674707004089,
"learning_rate": 1.936137617532272e-05,
"loss": 1.1064,
"step": 1390
},
{
"epoch": 0.6110381077529566,
"grad_norm": 0.2691317233389999,
"learning_rate": 1.935238496512828e-05,
"loss": 1.0988,
"step": 1395
},
{
"epoch": 0.6132282084975909,
"grad_norm": 0.2859835765343508,
"learning_rate": 1.9343333020269724e-05,
"loss": 1.1227,
"step": 1400
},
{
"epoch": 0.6154183092422252,
"grad_norm": 0.2694914982156586,
"learning_rate": 1.9334220399530662e-05,
"loss": 1.1449,
"step": 1405
},
{
"epoch": 0.6176084099868594,
"grad_norm": 0.30824208779553697,
"learning_rate": 1.9325047162088707e-05,
"loss": 1.1332,
"step": 1410
},
{
"epoch": 0.6197985107314936,
"grad_norm": 0.26679438460192956,
"learning_rate": 1.9315813367515142e-05,
"loss": 1.1146,
"step": 1415
},
{
"epoch": 0.6219886114761279,
"grad_norm": 0.2797877766039786,
"learning_rate": 1.9306519075774504e-05,
"loss": 1.107,
"step": 1420
},
{
"epoch": 0.6241787122207622,
"grad_norm": 0.2760924748552255,
"learning_rate": 1.929716434722419e-05,
"loss": 1.1313,
"step": 1425
},
{
"epoch": 0.6263688129653964,
"grad_norm": 0.26866416743615296,
"learning_rate": 1.928774924261409e-05,
"loss": 1.0984,
"step": 1430
},
{
"epoch": 0.6285589137100307,
"grad_norm": 0.2771243619587273,
"learning_rate": 1.927827382308617e-05,
"loss": 1.1062,
"step": 1435
},
{
"epoch": 0.6307490144546649,
"grad_norm": 0.26961265825183367,
"learning_rate": 1.9268738150174084e-05,
"loss": 1.0975,
"step": 1440
},
{
"epoch": 0.6329391151992991,
"grad_norm": 0.28071640780132767,
"learning_rate": 1.925914228580277e-05,
"loss": 1.1156,
"step": 1445
},
{
"epoch": 0.6351292159439335,
"grad_norm": 0.266937201343873,
"learning_rate": 1.924948629228806e-05,
"loss": 1.0809,
"step": 1450
},
{
"epoch": 0.6373193166885677,
"grad_norm": 0.27461403399812495,
"learning_rate": 1.9239770232336258e-05,
"loss": 1.1105,
"step": 1455
},
{
"epoch": 0.6395094174332019,
"grad_norm": 0.2986924180856064,
"learning_rate": 1.922999416904374e-05,
"loss": 1.134,
"step": 1460
},
{
"epoch": 0.6416995181778362,
"grad_norm": 0.2714601828449212,
"learning_rate": 1.9220158165896557e-05,
"loss": 1.1152,
"step": 1465
},
{
"epoch": 0.6438896189224704,
"grad_norm": 0.2793036577569871,
"learning_rate": 1.9210262286770005e-05,
"loss": 1.1055,
"step": 1470
},
{
"epoch": 0.6460797196671046,
"grad_norm": 0.2909420705461525,
"learning_rate": 1.9200306595928207e-05,
"loss": 1.1107,
"step": 1475
},
{
"epoch": 0.648269820411739,
"grad_norm": 0.28963291188379814,
"learning_rate": 1.919029115802372e-05,
"loss": 1.0994,
"step": 1480
},
{
"epoch": 0.6504599211563732,
"grad_norm": 0.2870590741002279,
"learning_rate": 1.918021603809709e-05,
"loss": 1.1002,
"step": 1485
},
{
"epoch": 0.6526500219010074,
"grad_norm": 0.2871696912380895,
"learning_rate": 1.9170081301576444e-05,
"loss": 1.1187,
"step": 1490
},
{
"epoch": 0.6548401226456417,
"grad_norm": 0.27381019486120817,
"learning_rate": 1.915988701427706e-05,
"loss": 1.1137,
"step": 1495
},
{
"epoch": 0.657030223390276,
"grad_norm": 0.27657948021093287,
"learning_rate": 1.9149633242400942e-05,
"loss": 1.1352,
"step": 1500
},
{
"epoch": 0.6592203241349102,
"grad_norm": 0.2751875181246786,
"learning_rate": 1.9139320052536388e-05,
"loss": 1.1193,
"step": 1505
},
{
"epoch": 0.6614104248795445,
"grad_norm": 0.27766900983637177,
"learning_rate": 1.9128947511657558e-05,
"loss": 1.1221,
"step": 1510
},
{
"epoch": 0.6636005256241787,
"grad_norm": 0.2749718770075062,
"learning_rate": 1.911851568712403e-05,
"loss": 1.1176,
"step": 1515
},
{
"epoch": 0.6657906263688129,
"grad_norm": 0.2772174321273592,
"learning_rate": 1.910802464668039e-05,
"loss": 1.0959,
"step": 1520
},
{
"epoch": 0.6679807271134472,
"grad_norm": 0.2620155432938436,
"learning_rate": 1.9097474458455766e-05,
"loss": 1.1064,
"step": 1525
},
{
"epoch": 0.6701708278580815,
"grad_norm": 0.2979089630943025,
"learning_rate": 1.9086865190963384e-05,
"loss": 1.1059,
"step": 1530
},
{
"epoch": 0.6723609286027157,
"grad_norm": 0.27611457258125455,
"learning_rate": 1.9076196913100146e-05,
"loss": 1.116,
"step": 1535
},
{
"epoch": 0.67455102934735,
"grad_norm": 0.27818118501142747,
"learning_rate": 1.906546969414616e-05,
"loss": 1.1115,
"step": 1540
},
{
"epoch": 0.6767411300919842,
"grad_norm": 0.27699675699597703,
"learning_rate": 1.9054683603764305e-05,
"loss": 1.0965,
"step": 1545
},
{
"epoch": 0.6789312308366184,
"grad_norm": 0.27740082677306727,
"learning_rate": 1.904383871199977e-05,
"loss": 1.0965,
"step": 1550
},
{
"epoch": 0.6811213315812528,
"grad_norm": 0.2809043424196064,
"learning_rate": 1.90329350892796e-05,
"loss": 1.098,
"step": 1555
},
{
"epoch": 0.683311432325887,
"grad_norm": 0.2639066667854838,
"learning_rate": 1.9021972806412248e-05,
"loss": 1.1078,
"step": 1560
},
{
"epoch": 0.6855015330705212,
"grad_norm": 0.276203210313125,
"learning_rate": 1.9010951934587098e-05,
"loss": 1.1457,
"step": 1565
},
{
"epoch": 0.6876916338151555,
"grad_norm": 0.27076390154989444,
"learning_rate": 1.899987254537402e-05,
"loss": 1.0775,
"step": 1570
},
{
"epoch": 0.6898817345597897,
"grad_norm": 0.2774038504373021,
"learning_rate": 1.898873471072289e-05,
"loss": 1.1191,
"step": 1575
},
{
"epoch": 0.692071835304424,
"grad_norm": 0.26935741887190473,
"learning_rate": 1.897753850296314e-05,
"loss": 1.1195,
"step": 1580
},
{
"epoch": 0.6942619360490583,
"grad_norm": 0.27030811654802295,
"learning_rate": 1.896628399480327e-05,
"loss": 1.0945,
"step": 1585
},
{
"epoch": 0.6964520367936925,
"grad_norm": 0.27721250821361343,
"learning_rate": 1.8954971259330393e-05,
"loss": 1.1582,
"step": 1590
},
{
"epoch": 0.6986421375383267,
"grad_norm": 0.26784617062540544,
"learning_rate": 1.8943600370009744e-05,
"loss": 1.0934,
"step": 1595
},
{
"epoch": 0.700832238282961,
"grad_norm": 0.2715953252890686,
"learning_rate": 1.8932171400684215e-05,
"loss": 1.1187,
"step": 1600
},
{
"epoch": 0.7030223390275953,
"grad_norm": 0.2827134309634548,
"learning_rate": 1.8920684425573865e-05,
"loss": 1.1193,
"step": 1605
},
{
"epoch": 0.7052124397722295,
"grad_norm": 0.26957691881610674,
"learning_rate": 1.890913951927546e-05,
"loss": 1.1219,
"step": 1610
},
{
"epoch": 0.7074025405168638,
"grad_norm": 0.27433959110638084,
"learning_rate": 1.8897536756761954e-05,
"loss": 1.0973,
"step": 1615
},
{
"epoch": 0.709592641261498,
"grad_norm": 0.2818454037953233,
"learning_rate": 1.8885876213382032e-05,
"loss": 1.1164,
"step": 1620
},
{
"epoch": 0.7117827420061322,
"grad_norm": 0.27685333665328266,
"learning_rate": 1.887415796485961e-05,
"loss": 1.1262,
"step": 1625
},
{
"epoch": 0.7139728427507666,
"grad_norm": 0.2744600694708599,
"learning_rate": 1.8862382087293333e-05,
"loss": 1.1148,
"step": 1630
},
{
"epoch": 0.7161629434954008,
"grad_norm": 0.26756099510595055,
"learning_rate": 1.8850548657156106e-05,
"loss": 1.1148,
"step": 1635
},
{
"epoch": 0.718353044240035,
"grad_norm": 0.26844914903776257,
"learning_rate": 1.8838657751294564e-05,
"loss": 1.085,
"step": 1640
},
{
"epoch": 0.7205431449846693,
"grad_norm": 0.2769618210155935,
"learning_rate": 1.8826709446928612e-05,
"loss": 1.1254,
"step": 1645
},
{
"epoch": 0.7227332457293035,
"grad_norm": 0.2640705844978224,
"learning_rate": 1.8814703821650883e-05,
"loss": 1.1105,
"step": 1650
},
{
"epoch": 0.7249233464739377,
"grad_norm": 0.25859071008307577,
"learning_rate": 1.8802640953426263e-05,
"loss": 1.1096,
"step": 1655
},
{
"epoch": 0.7271134472185721,
"grad_norm": 0.262574438661076,
"learning_rate": 1.8790520920591376e-05,
"loss": 1.0914,
"step": 1660
},
{
"epoch": 0.7293035479632063,
"grad_norm": 0.27713581623207095,
"learning_rate": 1.8778343801854074e-05,
"loss": 1.1324,
"step": 1665
},
{
"epoch": 0.7314936487078406,
"grad_norm": 0.2616321001095485,
"learning_rate": 1.8766109676292922e-05,
"loss": 1.0879,
"step": 1670
},
{
"epoch": 0.7336837494524748,
"grad_norm": 0.26898963160284506,
"learning_rate": 1.8753818623356702e-05,
"loss": 1.0885,
"step": 1675
},
{
"epoch": 0.735873850197109,
"grad_norm": 0.2691621602639776,
"learning_rate": 1.874147072286386e-05,
"loss": 1.1187,
"step": 1680
},
{
"epoch": 0.7380639509417434,
"grad_norm": 0.26783674649445033,
"learning_rate": 1.8729066055002037e-05,
"loss": 1.1117,
"step": 1685
},
{
"epoch": 0.7402540516863776,
"grad_norm": 0.2666441183588182,
"learning_rate": 1.8716604700327516e-05,
"loss": 1.1078,
"step": 1690
},
{
"epoch": 0.7424441524310118,
"grad_norm": 0.2738375436810693,
"learning_rate": 1.870408673976469e-05,
"loss": 1.1277,
"step": 1695
},
{
"epoch": 0.7446342531756461,
"grad_norm": 0.2738186764180608,
"learning_rate": 1.869151225460557e-05,
"loss": 1.1234,
"step": 1700
},
{
"epoch": 0.7468243539202803,
"grad_norm": 0.27101152403848194,
"learning_rate": 1.8678881326509236e-05,
"loss": 1.1023,
"step": 1705
},
{
"epoch": 0.7490144546649146,
"grad_norm": 0.2727444445299331,
"learning_rate": 1.8666194037501308e-05,
"loss": 1.1055,
"step": 1710
},
{
"epoch": 0.7512045554095489,
"grad_norm": 0.2975855374946256,
"learning_rate": 1.865345046997341e-05,
"loss": 1.0961,
"step": 1715
},
{
"epoch": 0.7533946561541831,
"grad_norm": 0.26858349969011003,
"learning_rate": 1.864065070668265e-05,
"loss": 1.0971,
"step": 1720
},
{
"epoch": 0.7555847568988173,
"grad_norm": 0.265988934542856,
"learning_rate": 1.8627794830751068e-05,
"loss": 1.1025,
"step": 1725
},
{
"epoch": 0.7577748576434516,
"grad_norm": 0.27148968054194217,
"learning_rate": 1.8614882925665107e-05,
"loss": 1.1094,
"step": 1730
},
{
"epoch": 0.7599649583880859,
"grad_norm": 0.27790159663787395,
"learning_rate": 1.8601915075275052e-05,
"loss": 1.1211,
"step": 1735
},
{
"epoch": 0.7621550591327201,
"grad_norm": 0.2758132531875579,
"learning_rate": 1.858889136379451e-05,
"loss": 1.1076,
"step": 1740
},
{
"epoch": 0.7643451598773544,
"grad_norm": 0.2623757976766024,
"learning_rate": 1.8575811875799846e-05,
"loss": 1.0887,
"step": 1745
},
{
"epoch": 0.7665352606219886,
"grad_norm": 0.26151492176840657,
"learning_rate": 1.8562676696229646e-05,
"loss": 1.124,
"step": 1750
},
{
"epoch": 0.7687253613666228,
"grad_norm": 0.271833755533329,
"learning_rate": 1.8549485910384145e-05,
"loss": 1.1453,
"step": 1755
},
{
"epoch": 0.7709154621112572,
"grad_norm": 0.2716257068895584,
"learning_rate": 1.8536239603924707e-05,
"loss": 1.1084,
"step": 1760
},
{
"epoch": 0.7731055628558914,
"grad_norm": 0.27019868315503554,
"learning_rate": 1.852293786287323e-05,
"loss": 1.1328,
"step": 1765
},
{
"epoch": 0.7752956636005256,
"grad_norm": 0.27402054743522397,
"learning_rate": 1.850958077361161e-05,
"loss": 1.1211,
"step": 1770
},
{
"epoch": 0.7774857643451599,
"grad_norm": 0.26295913539462323,
"learning_rate": 1.8496168422881192e-05,
"loss": 1.1033,
"step": 1775
},
{
"epoch": 0.7796758650897941,
"grad_norm": 0.26609455658638365,
"learning_rate": 1.8482700897782166e-05,
"loss": 1.0914,
"step": 1780
},
{
"epoch": 0.7818659658344284,
"grad_norm": 0.2611758289091626,
"learning_rate": 1.846917828577304e-05,
"loss": 1.107,
"step": 1785
},
{
"epoch": 0.7840560665790627,
"grad_norm": 0.2733500279507379,
"learning_rate": 1.8455600674670054e-05,
"loss": 1.1066,
"step": 1790
},
{
"epoch": 0.7862461673236969,
"grad_norm": 0.27632403418777396,
"learning_rate": 1.8441968152646623e-05,
"loss": 1.0865,
"step": 1795
},
{
"epoch": 0.7884362680683311,
"grad_norm": 0.2613214944333076,
"learning_rate": 1.842828080823274e-05,
"loss": 1.0889,
"step": 1800
},
{
"epoch": 0.7906263688129654,
"grad_norm": 0.26071387464509554,
"learning_rate": 1.8414538730314428e-05,
"loss": 1.0898,
"step": 1805
},
{
"epoch": 0.7928164695575997,
"grad_norm": 0.2697069485769201,
"learning_rate": 1.8400742008133146e-05,
"loss": 1.1227,
"step": 1810
},
{
"epoch": 0.7950065703022339,
"grad_norm": 0.2686347723972552,
"learning_rate": 1.838689073128521e-05,
"loss": 1.1051,
"step": 1815
},
{
"epoch": 0.7971966710468682,
"grad_norm": 0.2749670020061178,
"learning_rate": 1.8372984989721232e-05,
"loss": 1.0992,
"step": 1820
},
{
"epoch": 0.7993867717915024,
"grad_norm": 0.26279534392981274,
"learning_rate": 1.83590248737455e-05,
"loss": 1.1102,
"step": 1825
},
{
"epoch": 0.8015768725361366,
"grad_norm": 0.27530772454657076,
"learning_rate": 1.8345010474015418e-05,
"loss": 1.0838,
"step": 1830
},
{
"epoch": 0.803766973280771,
"grad_norm": 0.273234030161823,
"learning_rate": 1.8330941881540917e-05,
"loss": 1.1107,
"step": 1835
},
{
"epoch": 0.8059570740254052,
"grad_norm": 0.2710268252498512,
"learning_rate": 1.831681918768385e-05,
"loss": 1.1148,
"step": 1840
},
{
"epoch": 0.8081471747700394,
"grad_norm": 0.2991459154978479,
"learning_rate": 1.8302642484157407e-05,
"loss": 1.1016,
"step": 1845
},
{
"epoch": 0.8103372755146737,
"grad_norm": 0.2644212945605006,
"learning_rate": 1.8288411863025518e-05,
"loss": 1.0885,
"step": 1850
},
{
"epoch": 0.8125273762593079,
"grad_norm": 0.26633412635914927,
"learning_rate": 1.8274127416702262e-05,
"loss": 1.0869,
"step": 1855
},
{
"epoch": 0.8147174770039421,
"grad_norm": 0.28059854237511905,
"learning_rate": 1.8259789237951248e-05,
"loss": 1.0859,
"step": 1860
},
{
"epoch": 0.8169075777485765,
"grad_norm": 0.2618027813328137,
"learning_rate": 1.824539741988504e-05,
"loss": 1.123,
"step": 1865
},
{
"epoch": 0.8190976784932107,
"grad_norm": 0.27214989117780813,
"learning_rate": 1.8230952055964518e-05,
"loss": 1.1027,
"step": 1870
},
{
"epoch": 0.8212877792378449,
"grad_norm": 0.2666332713432548,
"learning_rate": 1.8216453239998318e-05,
"loss": 1.1191,
"step": 1875
},
{
"epoch": 0.8234778799824792,
"grad_norm": 0.28508298692561346,
"learning_rate": 1.8201901066142166e-05,
"loss": 1.0912,
"step": 1880
},
{
"epoch": 0.8256679807271134,
"grad_norm": 0.26214833981514946,
"learning_rate": 1.8187295628898315e-05,
"loss": 1.1059,
"step": 1885
},
{
"epoch": 0.8278580814717477,
"grad_norm": 0.2582420612465903,
"learning_rate": 1.81726370231149e-05,
"loss": 1.1027,
"step": 1890
},
{
"epoch": 0.830048182216382,
"grad_norm": 0.2648228026122393,
"learning_rate": 1.8157925343985344e-05,
"loss": 1.0934,
"step": 1895
},
{
"epoch": 0.8322382829610162,
"grad_norm": 0.26577633349371715,
"learning_rate": 1.814316068704773e-05,
"loss": 1.0934,
"step": 1900
},
{
"epoch": 0.8344283837056504,
"grad_norm": 0.2767314302797419,
"learning_rate": 1.812834314818417e-05,
"loss": 1.0918,
"step": 1905
},
{
"epoch": 0.8366184844502847,
"grad_norm": 0.2654350963998228,
"learning_rate": 1.8113472823620196e-05,
"loss": 1.0906,
"step": 1910
},
{
"epoch": 0.838808585194919,
"grad_norm": 0.2788726099269675,
"learning_rate": 1.8098549809924143e-05,
"loss": 1.0926,
"step": 1915
},
{
"epoch": 0.8409986859395532,
"grad_norm": 0.27306335537213267,
"learning_rate": 1.8083574204006496e-05,
"loss": 1.0895,
"step": 1920
},
{
"epoch": 0.8431887866841875,
"grad_norm": 0.2680339031442287,
"learning_rate": 1.8068546103119286e-05,
"loss": 1.1086,
"step": 1925
},
{
"epoch": 0.8453788874288217,
"grad_norm": 0.26391750461525426,
"learning_rate": 1.805346560485544e-05,
"loss": 1.1084,
"step": 1930
},
{
"epoch": 0.8475689881734559,
"grad_norm": 0.27946114587100535,
"learning_rate": 1.8038332807148165e-05,
"loss": 1.0947,
"step": 1935
},
{
"epoch": 0.8497590889180903,
"grad_norm": 0.25762923925047615,
"learning_rate": 1.8023147808270293e-05,
"loss": 1.1301,
"step": 1940
},
{
"epoch": 0.8519491896627245,
"grad_norm": 0.2704818192172617,
"learning_rate": 1.8007910706833648e-05,
"loss": 1.1143,
"step": 1945
},
{
"epoch": 0.8541392904073587,
"grad_norm": 0.27049489080275974,
"learning_rate": 1.799262160178843e-05,
"loss": 1.1086,
"step": 1950
},
{
"epoch": 0.856329391151993,
"grad_norm": 0.2659644550309275,
"learning_rate": 1.797728059242253e-05,
"loss": 1.1098,
"step": 1955
},
{
"epoch": 0.8585194918966272,
"grad_norm": 0.26336208498756253,
"learning_rate": 1.7961887778360916e-05,
"loss": 1.1074,
"step": 1960
},
{
"epoch": 0.8607095926412615,
"grad_norm": 0.2801383860699238,
"learning_rate": 1.7946443259564982e-05,
"loss": 1.1187,
"step": 1965
},
{
"epoch": 0.8628996933858958,
"grad_norm": 0.27770606630491224,
"learning_rate": 1.7930947136331884e-05,
"loss": 1.0789,
"step": 1970
},
{
"epoch": 0.86508979413053,
"grad_norm": 0.2998967999026723,
"learning_rate": 1.7915399509293906e-05,
"loss": 1.1016,
"step": 1975
},
{
"epoch": 0.8672798948751642,
"grad_norm": 0.2648191447375537,
"learning_rate": 1.7899800479417796e-05,
"loss": 1.1051,
"step": 1980
},
{
"epoch": 0.8694699956197985,
"grad_norm": 0.2749847601598037,
"learning_rate": 1.7884150148004117e-05,
"loss": 1.1215,
"step": 1985
},
{
"epoch": 0.8716600963644328,
"grad_norm": 0.26906571047619404,
"learning_rate": 1.7868448616686577e-05,
"loss": 1.0973,
"step": 1990
},
{
"epoch": 0.873850197109067,
"grad_norm": 0.2641533486933865,
"learning_rate": 1.785269598743139e-05,
"loss": 1.1125,
"step": 1995
},
{
"epoch": 0.8760402978537013,
"grad_norm": 0.2745178776293995,
"learning_rate": 1.78368923625366e-05,
"loss": 1.0943,
"step": 2000
},
{
"epoch": 0.8782303985983355,
"grad_norm": 0.26767546603196474,
"learning_rate": 1.7821037844631404e-05,
"loss": 1.0961,
"step": 2005
},
{
"epoch": 0.8804204993429697,
"grad_norm": 0.2674524087347779,
"learning_rate": 1.7805132536675512e-05,
"loss": 1.0988,
"step": 2010
},
{
"epoch": 0.8826106000876041,
"grad_norm": 0.276219025861902,
"learning_rate": 1.7789176541958466e-05,
"loss": 1.1166,
"step": 2015
},
{
"epoch": 0.8848007008322383,
"grad_norm": 0.2726202235697706,
"learning_rate": 1.7773169964098965e-05,
"loss": 1.1184,
"step": 2020
},
{
"epoch": 0.8869908015768725,
"grad_norm": 0.2634542597431643,
"learning_rate": 1.77571129070442e-05,
"loss": 1.1344,
"step": 2025
},
{
"epoch": 0.8891809023215068,
"grad_norm": 0.2848518603393168,
"learning_rate": 1.774100547506917e-05,
"loss": 1.125,
"step": 2030
},
{
"epoch": 0.891371003066141,
"grad_norm": 0.2786252562346639,
"learning_rate": 1.772484777277602e-05,
"loss": 1.1127,
"step": 2035
},
{
"epoch": 0.8935611038107752,
"grad_norm": 0.26740734293412094,
"learning_rate": 1.7708639905093337e-05,
"loss": 1.1203,
"step": 2040
},
{
"epoch": 0.8957512045554096,
"grad_norm": 0.2747291608481027,
"learning_rate": 1.76923819772755e-05,
"loss": 1.0998,
"step": 2045
},
{
"epoch": 0.8979413053000438,
"grad_norm": 0.2869604556453588,
"learning_rate": 1.7676074094901975e-05,
"loss": 1.1035,
"step": 2050
},
{
"epoch": 0.900131406044678,
"grad_norm": 0.2651779430966185,
"learning_rate": 1.7659716363876616e-05,
"loss": 1.0871,
"step": 2055
},
{
"epoch": 0.9023215067893123,
"grad_norm": 0.2729012408536021,
"learning_rate": 1.7643308890427027e-05,
"loss": 1.0877,
"step": 2060
},
{
"epoch": 0.9045116075339465,
"grad_norm": 0.27230010821588146,
"learning_rate": 1.762685178110382e-05,
"loss": 1.0955,
"step": 2065
},
{
"epoch": 0.9067017082785808,
"grad_norm": 0.2859811730936888,
"learning_rate": 1.761034514277995e-05,
"loss": 1.1016,
"step": 2070
},
{
"epoch": 0.9088918090232151,
"grad_norm": 0.2729495630885254,
"learning_rate": 1.759378908265001e-05,
"loss": 1.1006,
"step": 2075
},
{
"epoch": 0.9110819097678493,
"grad_norm": 0.29579631337299594,
"learning_rate": 1.7577183708229545e-05,
"loss": 1.1277,
"step": 2080
},
{
"epoch": 0.9132720105124835,
"grad_norm": 0.2678446315371196,
"learning_rate": 1.756052912735435e-05,
"loss": 1.0807,
"step": 2085
},
{
"epoch": 0.9154621112571178,
"grad_norm": 0.27329266459572743,
"learning_rate": 1.754382544817976e-05,
"loss": 1.0998,
"step": 2090
},
{
"epoch": 0.9176522120017521,
"grad_norm": 0.2744618273856731,
"learning_rate": 1.752707277917996e-05,
"loss": 1.0861,
"step": 2095
},
{
"epoch": 0.9198423127463863,
"grad_norm": 0.26947221466303073,
"learning_rate": 1.7510271229147286e-05,
"loss": 1.1004,
"step": 2100
},
{
"epoch": 0.9220324134910206,
"grad_norm": 0.2862528549331771,
"learning_rate": 1.7493420907191493e-05,
"loss": 1.109,
"step": 2105
},
{
"epoch": 0.9242225142356548,
"grad_norm": 0.2620725898469788,
"learning_rate": 1.7476521922739067e-05,
"loss": 1.1082,
"step": 2110
},
{
"epoch": 0.926412614980289,
"grad_norm": 0.2785905306845852,
"learning_rate": 1.7459574385532518e-05,
"loss": 1.1191,
"step": 2115
},
{
"epoch": 0.9286027157249234,
"grad_norm": 0.2637140594813735,
"learning_rate": 1.744257840562965e-05,
"loss": 1.0779,
"step": 2120
},
{
"epoch": 0.9307928164695576,
"grad_norm": 0.2714092567203416,
"learning_rate": 1.742553409340286e-05,
"loss": 1.1043,
"step": 2125
},
{
"epoch": 0.9329829172141918,
"grad_norm": 0.2700689511416106,
"learning_rate": 1.740844155953841e-05,
"loss": 1.1018,
"step": 2130
},
{
"epoch": 0.9351730179588261,
"grad_norm": 0.278450005569163,
"learning_rate": 1.7391300915035728e-05,
"loss": 1.1096,
"step": 2135
},
{
"epoch": 0.9373631187034603,
"grad_norm": 0.2817205107300403,
"learning_rate": 1.7374112271206658e-05,
"loss": 1.0885,
"step": 2140
},
{
"epoch": 0.9395532194480947,
"grad_norm": 0.2747437897419016,
"learning_rate": 1.735687573967476e-05,
"loss": 1.1262,
"step": 2145
},
{
"epoch": 0.9417433201927289,
"grad_norm": 0.2723091223016179,
"learning_rate": 1.733959143237459e-05,
"loss": 1.093,
"step": 2150
},
{
"epoch": 0.9439334209373631,
"grad_norm": 0.2549767505673877,
"learning_rate": 1.732225946155094e-05,
"loss": 1.0664,
"step": 2155
},
{
"epoch": 0.9461235216819974,
"grad_norm": 0.26305174791961355,
"learning_rate": 1.730487993975814e-05,
"loss": 1.0848,
"step": 2160
},
{
"epoch": 0.9483136224266316,
"grad_norm": 0.2643783576244969,
"learning_rate": 1.728745297985932e-05,
"loss": 1.1121,
"step": 2165
},
{
"epoch": 0.9505037231712659,
"grad_norm": 0.2743918837868961,
"learning_rate": 1.7269978695025674e-05,
"loss": 1.1375,
"step": 2170
},
{
"epoch": 0.9526938239159002,
"grad_norm": 0.26269555514318277,
"learning_rate": 1.7252457198735722e-05,
"loss": 1.1109,
"step": 2175
},
{
"epoch": 0.9548839246605344,
"grad_norm": 0.26810701345053684,
"learning_rate": 1.7234888604774576e-05,
"loss": 1.1102,
"step": 2180
},
{
"epoch": 0.9570740254051686,
"grad_norm": 0.2637883296717608,
"learning_rate": 1.721727302723321e-05,
"loss": 1.1318,
"step": 2185
},
{
"epoch": 0.9592641261498029,
"grad_norm": 0.2719515531587497,
"learning_rate": 1.7199610580507696e-05,
"loss": 1.1156,
"step": 2190
},
{
"epoch": 0.9614542268944372,
"grad_norm": 0.2594336301706438,
"learning_rate": 1.7181901379298493e-05,
"loss": 1.0939,
"step": 2195
},
{
"epoch": 0.9636443276390714,
"grad_norm": 0.2610060749161414,
"learning_rate": 1.7164145538609668e-05,
"loss": 1.0693,
"step": 2200
},
{
"epoch": 0.9658344283837057,
"grad_norm": 0.26884834201604224,
"learning_rate": 1.7146343173748185e-05,
"loss": 1.102,
"step": 2205
},
{
"epoch": 0.9680245291283399,
"grad_norm": 0.2678668170662298,
"learning_rate": 1.712849440032312e-05,
"loss": 1.1195,
"step": 2210
},
{
"epoch": 0.9702146298729741,
"grad_norm": 0.2621622732714593,
"learning_rate": 1.7110599334244936e-05,
"loss": 1.0922,
"step": 2215
},
{
"epoch": 0.9724047306176085,
"grad_norm": 0.25732866295844903,
"learning_rate": 1.7092658091724722e-05,
"loss": 1.1021,
"step": 2220
},
{
"epoch": 0.9745948313622427,
"grad_norm": 0.2596998662779607,
"learning_rate": 1.7074670789273434e-05,
"loss": 1.0922,
"step": 2225
},
{
"epoch": 0.9767849321068769,
"grad_norm": 0.2663765137611305,
"learning_rate": 1.7056637543701145e-05,
"loss": 1.0643,
"step": 2230
},
{
"epoch": 0.9789750328515112,
"grad_norm": 0.26236680268310164,
"learning_rate": 1.7038558472116288e-05,
"loss": 1.1059,
"step": 2235
},
{
"epoch": 0.9811651335961454,
"grad_norm": 0.2694955545673036,
"learning_rate": 1.7020433691924885e-05,
"loss": 1.0875,
"step": 2240
},
{
"epoch": 0.9833552343407796,
"grad_norm": 0.2684464368662257,
"learning_rate": 1.7002263320829796e-05,
"loss": 1.1076,
"step": 2245
},
{
"epoch": 0.985545335085414,
"grad_norm": 0.2765402521694828,
"learning_rate": 1.6984047476829944e-05,
"loss": 1.152,
"step": 2250
},
{
"epoch": 0.9877354358300482,
"grad_norm": 0.2750352215522162,
"learning_rate": 1.696578627821956e-05,
"loss": 1.0938,
"step": 2255
},
{
"epoch": 0.9899255365746824,
"grad_norm": 0.2652853027178437,
"learning_rate": 1.6947479843587406e-05,
"loss": 1.0824,
"step": 2260
},
{
"epoch": 0.9921156373193167,
"grad_norm": 0.27290730679118463,
"learning_rate": 1.692912829181601e-05,
"loss": 1.098,
"step": 2265
},
{
"epoch": 0.994305738063951,
"grad_norm": 0.25125441787653247,
"learning_rate": 1.691073174208089e-05,
"loss": 1.0924,
"step": 2270
},
{
"epoch": 0.9964958388085852,
"grad_norm": 0.25849286326507437,
"learning_rate": 1.6892290313849783e-05,
"loss": 1.101,
"step": 2275
},
{
"epoch": 0.9986859395532195,
"grad_norm": 0.2748639842655102,
"learning_rate": 1.687380412688187e-05,
"loss": 1.0953,
"step": 2280
},
{
"epoch": 1.0008760402978536,
"grad_norm": 0.26892282605547124,
"learning_rate": 1.6855273301226997e-05,
"loss": 1.0813,
"step": 2285
},
{
"epoch": 1.003066141042488,
"grad_norm": 0.2798609549581797,
"learning_rate": 1.683669795722489e-05,
"loss": 1.0699,
"step": 2290
},
{
"epoch": 1.0052562417871223,
"grad_norm": 0.2922624161897083,
"learning_rate": 1.681807821550438e-05,
"loss": 1.0848,
"step": 2295
},
{
"epoch": 1.0074463425317564,
"grad_norm": 0.259860907163824,
"learning_rate": 1.679941419698262e-05,
"loss": 1.0492,
"step": 2300
},
{
"epoch": 1.0096364432763907,
"grad_norm": 0.2712856437968212,
"learning_rate": 1.67807060228643e-05,
"loss": 1.0705,
"step": 2305
},
{
"epoch": 1.011826544021025,
"grad_norm": 0.26676093929792954,
"learning_rate": 1.6761953814640853e-05,
"loss": 1.0822,
"step": 2310
},
{
"epoch": 1.0140166447656591,
"grad_norm": 0.2645720914080042,
"learning_rate": 1.6743157694089675e-05,
"loss": 1.0393,
"step": 2315
},
{
"epoch": 1.0162067455102934,
"grad_norm": 0.273166262887767,
"learning_rate": 1.6724317783273323e-05,
"loss": 1.0701,
"step": 2320
},
{
"epoch": 1.0183968462549278,
"grad_norm": 0.2732861335077073,
"learning_rate": 1.670543420453874e-05,
"loss": 1.0686,
"step": 2325
},
{
"epoch": 1.0205869469995619,
"grad_norm": 0.27513602666197295,
"learning_rate": 1.668650708051644e-05,
"loss": 1.0537,
"step": 2330
},
{
"epoch": 1.0227770477441962,
"grad_norm": 0.2924048948993731,
"learning_rate": 1.666753653411973e-05,
"loss": 1.0666,
"step": 2335
},
{
"epoch": 1.0249671484888305,
"grad_norm": 0.28407032693449563,
"learning_rate": 1.6648522688543893e-05,
"loss": 1.0896,
"step": 2340
},
{
"epoch": 1.0271572492334646,
"grad_norm": 0.27607464243536883,
"learning_rate": 1.66294656672654e-05,
"loss": 1.0766,
"step": 2345
},
{
"epoch": 1.029347349978099,
"grad_norm": 0.2775777140300583,
"learning_rate": 1.6610365594041127e-05,
"loss": 1.0645,
"step": 2350
},
{
"epoch": 1.0315374507227333,
"grad_norm": 0.28212695538172056,
"learning_rate": 1.6591222592907493e-05,
"loss": 1.0746,
"step": 2355
},
{
"epoch": 1.0337275514673676,
"grad_norm": 0.26708386148841196,
"learning_rate": 1.6572036788179728e-05,
"loss": 1.0709,
"step": 2360
},
{
"epoch": 1.0359176522120017,
"grad_norm": 0.2647326361883933,
"learning_rate": 1.6552808304451e-05,
"loss": 1.0715,
"step": 2365
},
{
"epoch": 1.038107752956636,
"grad_norm": 0.27682819921543267,
"learning_rate": 1.6533537266591665e-05,
"loss": 1.0797,
"step": 2370
},
{
"epoch": 1.0402978537012704,
"grad_norm": 0.2736226995262712,
"learning_rate": 1.6514223799748402e-05,
"loss": 1.0703,
"step": 2375
},
{
"epoch": 1.0424879544459045,
"grad_norm": 0.27393594039449576,
"learning_rate": 1.649486802934344e-05,
"loss": 1.042,
"step": 2380
},
{
"epoch": 1.0446780551905388,
"grad_norm": 0.26691728521745745,
"learning_rate": 1.6475470081073724e-05,
"loss": 1.0688,
"step": 2385
},
{
"epoch": 1.0468681559351731,
"grad_norm": 0.2687411433553384,
"learning_rate": 1.6456030080910093e-05,
"loss": 1.0721,
"step": 2390
},
{
"epoch": 1.0490582566798072,
"grad_norm": 0.26858936807043243,
"learning_rate": 1.6436548155096495e-05,
"loss": 1.0723,
"step": 2395
},
{
"epoch": 1.0512483574244416,
"grad_norm": 0.2631218458201007,
"learning_rate": 1.6417024430149126e-05,
"loss": 1.0354,
"step": 2400
},
{
"epoch": 1.0534384581690759,
"grad_norm": 0.2715486126833225,
"learning_rate": 1.639745903285563e-05,
"loss": 1.0779,
"step": 2405
},
{
"epoch": 1.05562855891371,
"grad_norm": 0.2770109959912853,
"learning_rate": 1.6377852090274278e-05,
"loss": 1.0754,
"step": 2410
},
{
"epoch": 1.0578186596583443,
"grad_norm": 0.27924640149556096,
"learning_rate": 1.635820372973313e-05,
"loss": 1.0688,
"step": 2415
},
{
"epoch": 1.0600087604029786,
"grad_norm": 0.28178668312812843,
"learning_rate": 1.6338514078829227e-05,
"loss": 1.0996,
"step": 2420
},
{
"epoch": 1.0621988611476127,
"grad_norm": 0.27443250432543503,
"learning_rate": 1.6318783265427736e-05,
"loss": 1.0734,
"step": 2425
},
{
"epoch": 1.064388961892247,
"grad_norm": 0.27067404276566015,
"learning_rate": 1.6299011417661146e-05,
"loss": 1.0912,
"step": 2430
},
{
"epoch": 1.0665790626368814,
"grad_norm": 0.26818614853532363,
"learning_rate": 1.6279198663928415e-05,
"loss": 1.0727,
"step": 2435
},
{
"epoch": 1.0687691633815155,
"grad_norm": 0.27692110633499695,
"learning_rate": 1.625934513289416e-05,
"loss": 1.0918,
"step": 2440
},
{
"epoch": 1.0709592641261498,
"grad_norm": 0.2683984093970962,
"learning_rate": 1.6239450953487786e-05,
"loss": 1.0533,
"step": 2445
},
{
"epoch": 1.0731493648707842,
"grad_norm": 0.2680808660452308,
"learning_rate": 1.6219516254902697e-05,
"loss": 1.0691,
"step": 2450
},
{
"epoch": 1.0753394656154183,
"grad_norm": 0.276528567930805,
"learning_rate": 1.6199541166595397e-05,
"loss": 1.067,
"step": 2455
},
{
"epoch": 1.0775295663600526,
"grad_norm": 0.27014467052751223,
"learning_rate": 1.617952581828471e-05,
"loss": 1.0555,
"step": 2460
},
{
"epoch": 1.079719667104687,
"grad_norm": 0.2670920984515106,
"learning_rate": 1.6159470339950898e-05,
"loss": 1.0645,
"step": 2465
},
{
"epoch": 1.081909767849321,
"grad_norm": 0.26650906620699233,
"learning_rate": 1.6139374861834827e-05,
"loss": 1.0488,
"step": 2470
},
{
"epoch": 1.0840998685939554,
"grad_norm": 0.28613285356865026,
"learning_rate": 1.6119239514437125e-05,
"loss": 1.0527,
"step": 2475
},
{
"epoch": 1.0862899693385897,
"grad_norm": 0.2687551351307193,
"learning_rate": 1.6099064428517334e-05,
"loss": 1.0445,
"step": 2480
},
{
"epoch": 1.0884800700832238,
"grad_norm": 0.26480671253991495,
"learning_rate": 1.6078849735093057e-05,
"loss": 1.0588,
"step": 2485
},
{
"epoch": 1.090670170827858,
"grad_norm": 0.25864231413225064,
"learning_rate": 1.605859556543911e-05,
"loss": 1.0623,
"step": 2490
},
{
"epoch": 1.0928602715724924,
"grad_norm": 0.2714842359217486,
"learning_rate": 1.603830205108667e-05,
"loss": 1.0584,
"step": 2495
},
{
"epoch": 1.0950503723171265,
"grad_norm": 0.2754034686856655,
"learning_rate": 1.6017969323822417e-05,
"loss": 1.0939,
"step": 2500
},
{
"epoch": 1.0972404730617609,
"grad_norm": 0.26974950416433874,
"learning_rate": 1.5997597515687678e-05,
"loss": 1.058,
"step": 2505
},
{
"epoch": 1.0994305738063952,
"grad_norm": 0.2846310921186313,
"learning_rate": 1.5977186758977586e-05,
"loss": 1.0715,
"step": 2510
},
{
"epoch": 1.1016206745510293,
"grad_norm": 0.2648151245986546,
"learning_rate": 1.595673718624019e-05,
"loss": 1.0555,
"step": 2515
},
{
"epoch": 1.1038107752956636,
"grad_norm": 0.26291167200610804,
"learning_rate": 1.5936248930275625e-05,
"loss": 1.0656,
"step": 2520
},
{
"epoch": 1.106000876040298,
"grad_norm": 0.26794335368410055,
"learning_rate": 1.5915722124135227e-05,
"loss": 1.0785,
"step": 2525
},
{
"epoch": 1.108190976784932,
"grad_norm": 0.2700870903979981,
"learning_rate": 1.589515690112068e-05,
"loss": 1.0752,
"step": 2530
},
{
"epoch": 1.1103810775295664,
"grad_norm": 0.2761573536656991,
"learning_rate": 1.587455339478315e-05,
"loss": 1.0414,
"step": 2535
},
{
"epoch": 1.1125711782742007,
"grad_norm": 0.26100878660436766,
"learning_rate": 1.5853911738922416e-05,
"loss": 1.0742,
"step": 2540
},
{
"epoch": 1.1147612790188348,
"grad_norm": 0.2699651621314423,
"learning_rate": 1.5833232067586003e-05,
"loss": 1.0598,
"step": 2545
},
{
"epoch": 1.1169513797634691,
"grad_norm": 0.2643500086322407,
"learning_rate": 1.5812514515068304e-05,
"loss": 1.0758,
"step": 2550
},
{
"epoch": 1.1191414805081035,
"grad_norm": 0.26158426366408605,
"learning_rate": 1.579175921590972e-05,
"loss": 1.0414,
"step": 2555
},
{
"epoch": 1.1213315812527376,
"grad_norm": 0.2747194691521627,
"learning_rate": 1.5770966304895773e-05,
"loss": 1.0461,
"step": 2560
},
{
"epoch": 1.123521681997372,
"grad_norm": 0.2816065347391309,
"learning_rate": 1.5750135917056245e-05,
"loss": 1.04,
"step": 2565
},
{
"epoch": 1.1257117827420062,
"grad_norm": 0.27210023210552514,
"learning_rate": 1.5729268187664284e-05,
"loss": 1.0736,
"step": 2570
},
{
"epoch": 1.1279018834866403,
"grad_norm": 0.26597948498457047,
"learning_rate": 1.570836325223555e-05,
"loss": 1.0779,
"step": 2575
},
{
"epoch": 1.1300919842312747,
"grad_norm": 0.27937206024905653,
"learning_rate": 1.56874212465273e-05,
"loss": 1.066,
"step": 2580
},
{
"epoch": 1.132282084975909,
"grad_norm": 0.2710428960854059,
"learning_rate": 1.5666442306537542e-05,
"loss": 1.0652,
"step": 2585
},
{
"epoch": 1.134472185720543,
"grad_norm": 0.26833840792123953,
"learning_rate": 1.5645426568504135e-05,
"loss": 1.0826,
"step": 2590
},
{
"epoch": 1.1366622864651774,
"grad_norm": 0.27620664297921765,
"learning_rate": 1.5624374168903894e-05,
"loss": 1.0475,
"step": 2595
},
{
"epoch": 1.1388523872098117,
"grad_norm": 0.26832807339683407,
"learning_rate": 1.560328524445174e-05,
"loss": 1.0529,
"step": 2600
},
{
"epoch": 1.1410424879544458,
"grad_norm": 0.2771183218880352,
"learning_rate": 1.558215993209975e-05,
"loss": 1.0668,
"step": 2605
},
{
"epoch": 1.1432325886990802,
"grad_norm": 0.27093032012891505,
"learning_rate": 1.556099836903634e-05,
"loss": 1.0719,
"step": 2610
},
{
"epoch": 1.1454226894437145,
"grad_norm": 0.26498102520925576,
"learning_rate": 1.5539800692685326e-05,
"loss": 1.0437,
"step": 2615
},
{
"epoch": 1.1476127901883486,
"grad_norm": 0.2784489095370376,
"learning_rate": 1.5518567040705036e-05,
"loss": 1.0525,
"step": 2620
},
{
"epoch": 1.149802890932983,
"grad_norm": 0.25991368601083464,
"learning_rate": 1.5497297550987438e-05,
"loss": 1.0521,
"step": 2625
},
{
"epoch": 1.1519929916776173,
"grad_norm": 0.27972691777909914,
"learning_rate": 1.5475992361657228e-05,
"loss": 1.0402,
"step": 2630
},
{
"epoch": 1.1541830924222514,
"grad_norm": 0.2662833992907027,
"learning_rate": 1.545465161107093e-05,
"loss": 1.059,
"step": 2635
},
{
"epoch": 1.1563731931668857,
"grad_norm": 0.2736333701331792,
"learning_rate": 1.5433275437816004e-05,
"loss": 1.0498,
"step": 2640
},
{
"epoch": 1.15856329391152,
"grad_norm": 0.27897424796493925,
"learning_rate": 1.5411863980709962e-05,
"loss": 1.0717,
"step": 2645
},
{
"epoch": 1.1607533946561541,
"grad_norm": 0.2840334326774718,
"learning_rate": 1.539041737879943e-05,
"loss": 1.0652,
"step": 2650
},
{
"epoch": 1.1629434954007885,
"grad_norm": 0.26536289863258733,
"learning_rate": 1.536893577135928e-05,
"loss": 1.0662,
"step": 2655
},
{
"epoch": 1.1651335961454228,
"grad_norm": 0.26830852227928204,
"learning_rate": 1.5347419297891706e-05,
"loss": 1.0736,
"step": 2660
},
{
"epoch": 1.1673236968900569,
"grad_norm": 0.2648868718385574,
"learning_rate": 1.5325868098125312e-05,
"loss": 1.06,
"step": 2665
},
{
"epoch": 1.1695137976346912,
"grad_norm": 0.2807483907173534,
"learning_rate": 1.5304282312014235e-05,
"loss": 1.0676,
"step": 2670
},
{
"epoch": 1.1717038983793255,
"grad_norm": 0.2861338277467685,
"learning_rate": 1.5282662079737206e-05,
"loss": 1.0738,
"step": 2675
},
{
"epoch": 1.1738939991239596,
"grad_norm": 0.28438493797626496,
"learning_rate": 1.526100754169665e-05,
"loss": 1.0527,
"step": 2680
},
{
"epoch": 1.176084099868594,
"grad_norm": 0.2680358924154674,
"learning_rate": 1.5239318838517772e-05,
"loss": 1.0805,
"step": 2685
},
{
"epoch": 1.1782742006132283,
"grad_norm": 0.28144376585272174,
"learning_rate": 1.5217596111047656e-05,
"loss": 1.0715,
"step": 2690
},
{
"epoch": 1.1804643013578624,
"grad_norm": 0.27730236843145956,
"learning_rate": 1.5195839500354337e-05,
"loss": 1.0402,
"step": 2695
},
{
"epoch": 1.1826544021024967,
"grad_norm": 0.2739919694755342,
"learning_rate": 1.5174049147725885e-05,
"loss": 1.0498,
"step": 2700
},
{
"epoch": 1.184844502847131,
"grad_norm": 0.27303973140567345,
"learning_rate": 1.5152225194669495e-05,
"loss": 1.0508,
"step": 2705
},
{
"epoch": 1.1870346035917652,
"grad_norm": 0.27671454231651477,
"learning_rate": 1.5130367782910557e-05,
"loss": 1.0604,
"step": 2710
},
{
"epoch": 1.1892247043363995,
"grad_norm": 0.2641264661635561,
"learning_rate": 1.5108477054391756e-05,
"loss": 1.0721,
"step": 2715
},
{
"epoch": 1.1914148050810338,
"grad_norm": 0.26276026147789583,
"learning_rate": 1.5086553151272126e-05,
"loss": 1.0445,
"step": 2720
},
{
"epoch": 1.193604905825668,
"grad_norm": 0.27530344528775347,
"learning_rate": 1.5064596215926139e-05,
"loss": 1.0324,
"step": 2725
},
{
"epoch": 1.1957950065703022,
"grad_norm": 0.26851666471510427,
"learning_rate": 1.5042606390942785e-05,
"loss": 1.0758,
"step": 2730
},
{
"epoch": 1.1979851073149366,
"grad_norm": 0.27786184283527277,
"learning_rate": 1.5020583819124633e-05,
"loss": 1.0736,
"step": 2735
},
{
"epoch": 1.2001752080595707,
"grad_norm": 0.2748103368049444,
"learning_rate": 1.4998528643486915e-05,
"loss": 1.073,
"step": 2740
},
{
"epoch": 1.202365308804205,
"grad_norm": 0.2641215184110416,
"learning_rate": 1.4976441007256599e-05,
"loss": 1.0604,
"step": 2745
},
{
"epoch": 1.2045554095488393,
"grad_norm": 0.2717405510915443,
"learning_rate": 1.4954321053871444e-05,
"loss": 1.0283,
"step": 2750
},
{
"epoch": 1.2067455102934734,
"grad_norm": 0.26585144660715426,
"learning_rate": 1.4932168926979074e-05,
"loss": 1.0799,
"step": 2755
},
{
"epoch": 1.2089356110381078,
"grad_norm": 0.26852091548064677,
"learning_rate": 1.4909984770436064e-05,
"loss": 1.0727,
"step": 2760
},
{
"epoch": 1.211125711782742,
"grad_norm": 0.2832438116695105,
"learning_rate": 1.4887768728306974e-05,
"loss": 1.0629,
"step": 2765
},
{
"epoch": 1.2133158125273762,
"grad_norm": 0.2584259442397719,
"learning_rate": 1.4865520944863445e-05,
"loss": 1.0502,
"step": 2770
},
{
"epoch": 1.2155059132720105,
"grad_norm": 0.27567941486908737,
"learning_rate": 1.4843241564583236e-05,
"loss": 1.0723,
"step": 2775
},
{
"epoch": 1.2176960140166448,
"grad_norm": 0.27009311931492574,
"learning_rate": 1.4820930732149297e-05,
"loss": 1.0848,
"step": 2780
},
{
"epoch": 1.219886114761279,
"grad_norm": 0.2813111088617309,
"learning_rate": 1.4798588592448834e-05,
"loss": 1.0867,
"step": 2785
},
{
"epoch": 1.2220762155059133,
"grad_norm": 0.26607157411575105,
"learning_rate": 1.477621529057237e-05,
"loss": 1.082,
"step": 2790
},
{
"epoch": 1.2242663162505476,
"grad_norm": 0.2803710769666951,
"learning_rate": 1.4753810971812779e-05,
"loss": 1.0607,
"step": 2795
},
{
"epoch": 1.2264564169951817,
"grad_norm": 0.27648257549873206,
"learning_rate": 1.473137578166437e-05,
"loss": 1.0436,
"step": 2800
},
{
"epoch": 1.228646517739816,
"grad_norm": 0.2771472647218359,
"learning_rate": 1.4708909865821932e-05,
"loss": 1.0676,
"step": 2805
},
{
"epoch": 1.2308366184844504,
"grad_norm": 0.2587044578585095,
"learning_rate": 1.4686413370179786e-05,
"loss": 1.0531,
"step": 2810
},
{
"epoch": 1.2330267192290845,
"grad_norm": 0.26482046648511565,
"learning_rate": 1.466388644083084e-05,
"loss": 1.0615,
"step": 2815
},
{
"epoch": 1.2352168199737188,
"grad_norm": 0.26764213866360803,
"learning_rate": 1.4641329224065635e-05,
"loss": 1.0891,
"step": 2820
},
{
"epoch": 1.2374069207183531,
"grad_norm": 0.25538458354646704,
"learning_rate": 1.4618741866371404e-05,
"loss": 1.0562,
"step": 2825
},
{
"epoch": 1.2395970214629872,
"grad_norm": 0.2620736158974201,
"learning_rate": 1.4596124514431117e-05,
"loss": 1.0555,
"step": 2830
},
{
"epoch": 1.2417871222076216,
"grad_norm": 0.2568051923995238,
"learning_rate": 1.457347731512252e-05,
"loss": 1.0471,
"step": 2835
},
{
"epoch": 1.2439772229522559,
"grad_norm": 0.2582881510400402,
"learning_rate": 1.4550800415517201e-05,
"loss": 1.0586,
"step": 2840
},
{
"epoch": 1.24616732369689,
"grad_norm": 0.2590493852659133,
"learning_rate": 1.452809396287961e-05,
"loss": 1.0453,
"step": 2845
},
{
"epoch": 1.2483574244415243,
"grad_norm": 0.26593849032587,
"learning_rate": 1.4505358104666125e-05,
"loss": 1.0633,
"step": 2850
},
{
"epoch": 1.2505475251861586,
"grad_norm": 0.260260223527052,
"learning_rate": 1.4482592988524076e-05,
"loss": 1.0553,
"step": 2855
},
{
"epoch": 1.2527376259307927,
"grad_norm": 0.27392055036341906,
"learning_rate": 1.4459798762290807e-05,
"loss": 1.0533,
"step": 2860
},
{
"epoch": 1.254927726675427,
"grad_norm": 0.2854949267142112,
"learning_rate": 1.4436975573992687e-05,
"loss": 1.0551,
"step": 2865
},
{
"epoch": 1.2571178274200614,
"grad_norm": 0.26738529406237527,
"learning_rate": 1.4414123571844178e-05,
"loss": 1.0508,
"step": 2870
},
{
"epoch": 1.2593079281646955,
"grad_norm": 0.2714894374228558,
"learning_rate": 1.439124290424686e-05,
"loss": 1.0609,
"step": 2875
},
{
"epoch": 1.2614980289093298,
"grad_norm": 0.26386369834766954,
"learning_rate": 1.436833371978846e-05,
"loss": 1.0516,
"step": 2880
},
{
"epoch": 1.2636881296539642,
"grad_norm": 0.26154902396246443,
"learning_rate": 1.4345396167241903e-05,
"loss": 1.082,
"step": 2885
},
{
"epoch": 1.2658782303985983,
"grad_norm": 0.27009603608193167,
"learning_rate": 1.4322430395564326e-05,
"loss": 1.073,
"step": 2890
},
{
"epoch": 1.2680683311432326,
"grad_norm": 0.2607395450893893,
"learning_rate": 1.4299436553896133e-05,
"loss": 1.0578,
"step": 2895
},
{
"epoch": 1.270258431887867,
"grad_norm": 0.26182126573431935,
"learning_rate": 1.4276414791560012e-05,
"loss": 1.0455,
"step": 2900
},
{
"epoch": 1.272448532632501,
"grad_norm": 0.26055004587884284,
"learning_rate": 1.4253365258059964e-05,
"loss": 1.041,
"step": 2905
},
{
"epoch": 1.2746386333771353,
"grad_norm": 0.26542997354867837,
"learning_rate": 1.4230288103080339e-05,
"loss": 1.0973,
"step": 2910
},
{
"epoch": 1.2768287341217697,
"grad_norm": 0.28297483635713,
"learning_rate": 1.4207183476484864e-05,
"loss": 1.0398,
"step": 2915
},
{
"epoch": 1.2790188348664038,
"grad_norm": 0.2720216676413036,
"learning_rate": 1.418405152831567e-05,
"loss": 1.0619,
"step": 2920
},
{
"epoch": 1.281208935611038,
"grad_norm": 0.25854803685185235,
"learning_rate": 1.4160892408792307e-05,
"loss": 1.0756,
"step": 2925
},
{
"epoch": 1.2833990363556724,
"grad_norm": 0.27269182259029,
"learning_rate": 1.4137706268310784e-05,
"loss": 1.0521,
"step": 2930
},
{
"epoch": 1.2855891371003065,
"grad_norm": 0.2656775626074289,
"learning_rate": 1.411449325744258e-05,
"loss": 1.0713,
"step": 2935
},
{
"epoch": 1.2877792378449409,
"grad_norm": 0.2578007150112824,
"learning_rate": 1.4091253526933672e-05,
"loss": 1.0256,
"step": 2940
},
{
"epoch": 1.2899693385895752,
"grad_norm": 0.2578283345450258,
"learning_rate": 1.4067987227703558e-05,
"loss": 1.0398,
"step": 2945
},
{
"epoch": 1.2921594393342093,
"grad_norm": 0.25629426766106156,
"learning_rate": 1.4044694510844274e-05,
"loss": 1.0334,
"step": 2950
},
{
"epoch": 1.2943495400788436,
"grad_norm": 0.2608317134978303,
"learning_rate": 1.4021375527619409e-05,
"loss": 1.0443,
"step": 2955
},
{
"epoch": 1.296539640823478,
"grad_norm": 0.2633979929205152,
"learning_rate": 1.3998030429463127e-05,
"loss": 1.0436,
"step": 2960
},
{
"epoch": 1.298729741568112,
"grad_norm": 0.27013153256191164,
"learning_rate": 1.3974659367979186e-05,
"loss": 1.0682,
"step": 2965
},
{
"epoch": 1.3009198423127464,
"grad_norm": 0.26004507355183054,
"learning_rate": 1.395126249493995e-05,
"loss": 1.0529,
"step": 2970
},
{
"epoch": 1.3031099430573807,
"grad_norm": 0.26291757732552634,
"learning_rate": 1.3927839962285406e-05,
"loss": 1.0496,
"step": 2975
},
{
"epoch": 1.3053000438020148,
"grad_norm": 0.26738381407622364,
"learning_rate": 1.3904391922122165e-05,
"loss": 1.0385,
"step": 2980
},
{
"epoch": 1.3074901445466491,
"grad_norm": 0.2695960064844618,
"learning_rate": 1.3880918526722497e-05,
"loss": 1.0418,
"step": 2985
},
{
"epoch": 1.3096802452912835,
"grad_norm": 0.2786190745701212,
"learning_rate": 1.3857419928523328e-05,
"loss": 1.0584,
"step": 2990
},
{
"epoch": 1.3118703460359176,
"grad_norm": 0.25865708239628743,
"learning_rate": 1.3833896280125245e-05,
"loss": 1.0664,
"step": 2995
},
{
"epoch": 1.314060446780552,
"grad_norm": 0.27043550106818665,
"learning_rate": 1.3810347734291511e-05,
"loss": 1.0713,
"step": 3000
},
{
"epoch": 1.3162505475251862,
"grad_norm": 0.2786466259752348,
"learning_rate": 1.3786774443947087e-05,
"loss": 1.0654,
"step": 3005
},
{
"epoch": 1.3184406482698203,
"grad_norm": 0.2643200048634759,
"learning_rate": 1.3763176562177609e-05,
"loss": 1.0482,
"step": 3010
},
{
"epoch": 1.3206307490144547,
"grad_norm": 0.2713055037479327,
"learning_rate": 1.3739554242228421e-05,
"loss": 1.0584,
"step": 3015
},
{
"epoch": 1.322820849759089,
"grad_norm": 0.271441501246053,
"learning_rate": 1.3715907637503569e-05,
"loss": 1.0602,
"step": 3020
},
{
"epoch": 1.325010950503723,
"grad_norm": 0.2829209579812269,
"learning_rate": 1.3692236901564797e-05,
"loss": 1.0604,
"step": 3025
},
{
"epoch": 1.3272010512483574,
"grad_norm": 0.2767208834441249,
"learning_rate": 1.3668542188130567e-05,
"loss": 1.0646,
"step": 3030
},
{
"epoch": 1.3293911519929917,
"grad_norm": 0.2598545734511999,
"learning_rate": 1.3644823651075046e-05,
"loss": 1.0635,
"step": 3035
},
{
"epoch": 1.3315812527376258,
"grad_norm": 0.2651920796665307,
"learning_rate": 1.3621081444427114e-05,
"loss": 1.0689,
"step": 3040
},
{
"epoch": 1.3337713534822602,
"grad_norm": 0.26723030516899665,
"learning_rate": 1.3597315722369363e-05,
"loss": 1.0531,
"step": 3045
},
{
"epoch": 1.3359614542268945,
"grad_norm": 0.26169000253454394,
"learning_rate": 1.3573526639237089e-05,
"loss": 1.0404,
"step": 3050
},
{
"epoch": 1.3381515549715286,
"grad_norm": 0.2796414283834146,
"learning_rate": 1.3549714349517307e-05,
"loss": 1.058,
"step": 3055
},
{
"epoch": 1.340341655716163,
"grad_norm": 0.28015821037583777,
"learning_rate": 1.352587900784773e-05,
"loss": 1.042,
"step": 3060
},
{
"epoch": 1.3425317564607973,
"grad_norm": 0.26587778740964085,
"learning_rate": 1.3502020769015762e-05,
"loss": 1.049,
"step": 3065
},
{
"epoch": 1.3447218572054314,
"grad_norm": 0.27151273634562134,
"learning_rate": 1.3478139787957521e-05,
"loss": 1.0533,
"step": 3070
},
{
"epoch": 1.3469119579500657,
"grad_norm": 0.27774671237927817,
"learning_rate": 1.3454236219756794e-05,
"loss": 1.074,
"step": 3075
},
{
"epoch": 1.3491020586947,
"grad_norm": 0.30548698754204073,
"learning_rate": 1.3430310219644063e-05,
"loss": 1.0541,
"step": 3080
},
{
"epoch": 1.3512921594393341,
"grad_norm": 0.2769958664505269,
"learning_rate": 1.3406361942995481e-05,
"loss": 1.0709,
"step": 3085
},
{
"epoch": 1.3534822601839684,
"grad_norm": 0.26621017281065856,
"learning_rate": 1.3382391545331861e-05,
"loss": 1.0289,
"step": 3090
},
{
"epoch": 1.3556723609286028,
"grad_norm": 0.2654596239842093,
"learning_rate": 1.3358399182317672e-05,
"loss": 1.0676,
"step": 3095
},
{
"epoch": 1.3578624616732369,
"grad_norm": 0.2605361900805774,
"learning_rate": 1.3334385009760032e-05,
"loss": 1.0711,
"step": 3100
},
{
"epoch": 1.3600525624178712,
"grad_norm": 0.2734264004539264,
"learning_rate": 1.3310349183607682e-05,
"loss": 1.0678,
"step": 3105
},
{
"epoch": 1.3622426631625055,
"grad_norm": 0.2607045359371018,
"learning_rate": 1.3286291859949992e-05,
"loss": 1.0584,
"step": 3110
},
{
"epoch": 1.3644327639071396,
"grad_norm": 0.26946117404138387,
"learning_rate": 1.3262213195015928e-05,
"loss": 1.0473,
"step": 3115
},
{
"epoch": 1.366622864651774,
"grad_norm": 0.2649730727308054,
"learning_rate": 1.323811334517305e-05,
"loss": 1.0521,
"step": 3120
},
{
"epoch": 1.3688129653964083,
"grad_norm": 0.2714465080707081,
"learning_rate": 1.3213992466926498e-05,
"loss": 1.0771,
"step": 3125
},
{
"epoch": 1.3710030661410424,
"grad_norm": 0.2703058507590705,
"learning_rate": 1.3189850716917967e-05,
"loss": 1.0582,
"step": 3130
},
{
"epoch": 1.3731931668856767,
"grad_norm": 0.2628441678919807,
"learning_rate": 1.3165688251924694e-05,
"loss": 1.0656,
"step": 3135
},
{
"epoch": 1.375383267630311,
"grad_norm": 0.2683866578537538,
"learning_rate": 1.3141505228858438e-05,
"loss": 1.0287,
"step": 3140
},
{
"epoch": 1.3775733683749452,
"grad_norm": 0.27295127357891913,
"learning_rate": 1.3117301804764467e-05,
"loss": 1.0592,
"step": 3145
},
{
"epoch": 1.3797634691195795,
"grad_norm": 0.2662649769317535,
"learning_rate": 1.3093078136820534e-05,
"loss": 1.0418,
"step": 3150
},
{
"epoch": 1.3819535698642138,
"grad_norm": 0.2721595648282966,
"learning_rate": 1.3068834382335847e-05,
"loss": 1.0617,
"step": 3155
},
{
"epoch": 1.384143670608848,
"grad_norm": 0.27547589811798506,
"learning_rate": 1.304457069875007e-05,
"loss": 1.0607,
"step": 3160
},
{
"epoch": 1.3863337713534822,
"grad_norm": 0.2773048299021727,
"learning_rate": 1.3020287243632279e-05,
"loss": 1.0328,
"step": 3165
},
{
"epoch": 1.3885238720981166,
"grad_norm": 0.26593053511889214,
"learning_rate": 1.2995984174679946e-05,
"loss": 1.0564,
"step": 3170
},
{
"epoch": 1.3907139728427507,
"grad_norm": 0.26522350000493905,
"learning_rate": 1.2971661649717921e-05,
"loss": 1.0553,
"step": 3175
},
{
"epoch": 1.392904073587385,
"grad_norm": 0.2739759291020745,
"learning_rate": 1.2947319826697398e-05,
"loss": 1.0828,
"step": 3180
},
{
"epoch": 1.3950941743320193,
"grad_norm": 0.2653056984623836,
"learning_rate": 1.2922958863694897e-05,
"loss": 1.058,
"step": 3185
},
{
"epoch": 1.3972842750766534,
"grad_norm": 0.2703400930314117,
"learning_rate": 1.2898578918911225e-05,
"loss": 1.0578,
"step": 3190
},
{
"epoch": 1.3994743758212878,
"grad_norm": 0.26118537050181506,
"learning_rate": 1.287418015067047e-05,
"loss": 1.0775,
"step": 3195
},
{
"epoch": 1.401664476565922,
"grad_norm": 0.2674813901144202,
"learning_rate": 1.2849762717418952e-05,
"loss": 1.0314,
"step": 3200
},
{
"epoch": 1.4038545773105562,
"grad_norm": 0.2696947890894864,
"learning_rate": 1.2825326777724199e-05,
"loss": 1.0627,
"step": 3205
},
{
"epoch": 1.4060446780551905,
"grad_norm": 0.26602192805217667,
"learning_rate": 1.280087249027393e-05,
"loss": 1.0561,
"step": 3210
},
{
"epoch": 1.4082347787998248,
"grad_norm": 0.26896422544093695,
"learning_rate": 1.2776400013875006e-05,
"loss": 1.073,
"step": 3215
},
{
"epoch": 1.410424879544459,
"grad_norm": 0.27033565405608107,
"learning_rate": 1.2751909507452416e-05,
"loss": 1.059,
"step": 3220
},
{
"epoch": 1.4126149802890933,
"grad_norm": 0.2815290747437615,
"learning_rate": 1.2727401130048227e-05,
"loss": 1.0789,
"step": 3225
},
{
"epoch": 1.4148050810337276,
"grad_norm": 0.2667387754809302,
"learning_rate": 1.270287504082057e-05,
"loss": 1.0639,
"step": 3230
},
{
"epoch": 1.4169951817783617,
"grad_norm": 0.27178389185852914,
"learning_rate": 1.2678331399042585e-05,
"loss": 1.0469,
"step": 3235
},
{
"epoch": 1.419185282522996,
"grad_norm": 0.2760250342430515,
"learning_rate": 1.2653770364101416e-05,
"loss": 1.0783,
"step": 3240
},
{
"epoch": 1.4213753832676304,
"grad_norm": 0.2789228780835345,
"learning_rate": 1.2629192095497143e-05,
"loss": 1.0654,
"step": 3245
},
{
"epoch": 1.4235654840122645,
"grad_norm": 0.26547832460110765,
"learning_rate": 1.260459675284177e-05,
"loss": 1.0646,
"step": 3250
},
{
"epoch": 1.4257555847568988,
"grad_norm": 0.26556213425120645,
"learning_rate": 1.2579984495858179e-05,
"loss": 1.0348,
"step": 3255
},
{
"epoch": 1.4279456855015331,
"grad_norm": 0.2661116149656591,
"learning_rate": 1.255535548437909e-05,
"loss": 1.0637,
"step": 3260
},
{
"epoch": 1.4301357862461672,
"grad_norm": 0.2737166623631643,
"learning_rate": 1.2530709878346037e-05,
"loss": 1.0309,
"step": 3265
},
{
"epoch": 1.4323258869908015,
"grad_norm": 0.27818732407942237,
"learning_rate": 1.2506047837808308e-05,
"loss": 1.049,
"step": 3270
},
{
"epoch": 1.4345159877354359,
"grad_norm": 0.2559208288121395,
"learning_rate": 1.2481369522921925e-05,
"loss": 1.0439,
"step": 3275
},
{
"epoch": 1.43670608848007,
"grad_norm": 0.2652327115101503,
"learning_rate": 1.2456675093948592e-05,
"loss": 1.0732,
"step": 3280
},
{
"epoch": 1.4388961892247043,
"grad_norm": 0.2672613734124192,
"learning_rate": 1.243196471125466e-05,
"loss": 1.0754,
"step": 3285
},
{
"epoch": 1.4410862899693386,
"grad_norm": 0.26771969463623996,
"learning_rate": 1.2407238535310084e-05,
"loss": 1.0664,
"step": 3290
},
{
"epoch": 1.4432763907139727,
"grad_norm": 0.26928472122583363,
"learning_rate": 1.2382496726687383e-05,
"loss": 1.0736,
"step": 3295
},
{
"epoch": 1.445466491458607,
"grad_norm": 0.26495634423319697,
"learning_rate": 1.2357739446060593e-05,
"loss": 1.0602,
"step": 3300
},
{
"epoch": 1.4476565922032414,
"grad_norm": 0.26514448226579496,
"learning_rate": 1.2332966854204219e-05,
"loss": 1.0582,
"step": 3305
},
{
"epoch": 1.4498466929478755,
"grad_norm": 0.26861128400347345,
"learning_rate": 1.230817911199221e-05,
"loss": 1.0766,
"step": 3310
},
{
"epoch": 1.4520367936925098,
"grad_norm": 0.26551344406201943,
"learning_rate": 1.2283376380396894e-05,
"loss": 1.0525,
"step": 3315
},
{
"epoch": 1.4542268944371441,
"grad_norm": 0.2710500634387481,
"learning_rate": 1.2258558820487944e-05,
"loss": 1.0854,
"step": 3320
},
{
"epoch": 1.4564169951817783,
"grad_norm": 0.26286300066867274,
"learning_rate": 1.2233726593431322e-05,
"loss": 1.0578,
"step": 3325
},
{
"epoch": 1.4586070959264126,
"grad_norm": 0.26988115045815153,
"learning_rate": 1.220887986048825e-05,
"loss": 1.0605,
"step": 3330
},
{
"epoch": 1.460797196671047,
"grad_norm": 0.26499152320248637,
"learning_rate": 1.218401878301414e-05,
"loss": 1.0506,
"step": 3335
},
{
"epoch": 1.462987297415681,
"grad_norm": 0.2894193501769946,
"learning_rate": 1.2159143522457568e-05,
"loss": 1.0463,
"step": 3340
},
{
"epoch": 1.4651773981603153,
"grad_norm": 0.3054608297159258,
"learning_rate": 1.2134254240359206e-05,
"loss": 1.0648,
"step": 3345
},
{
"epoch": 1.4673674989049497,
"grad_norm": 0.2726029462886979,
"learning_rate": 1.210935109835079e-05,
"loss": 1.0445,
"step": 3350
},
{
"epoch": 1.4695575996495838,
"grad_norm": 0.2701355868570653,
"learning_rate": 1.2084434258154059e-05,
"loss": 1.0764,
"step": 3355
},
{
"epoch": 1.471747700394218,
"grad_norm": 0.2689474409120184,
"learning_rate": 1.2059503881579708e-05,
"loss": 1.0551,
"step": 3360
},
{
"epoch": 1.4739378011388524,
"grad_norm": 0.2657448254752709,
"learning_rate": 1.2034560130526341e-05,
"loss": 1.0586,
"step": 3365
},
{
"epoch": 1.4761279018834865,
"grad_norm": 0.2831462519055798,
"learning_rate": 1.2009603166979409e-05,
"loss": 1.0494,
"step": 3370
},
{
"epoch": 1.4783180026281209,
"grad_norm": 0.2643934359070407,
"learning_rate": 1.1984633153010175e-05,
"loss": 1.0506,
"step": 3375
},
{
"epoch": 1.4805081033727552,
"grad_norm": 0.26997363450497397,
"learning_rate": 1.1959650250774644e-05,
"loss": 1.0344,
"step": 3380
},
{
"epoch": 1.4826982041173893,
"grad_norm": 0.26573567629495287,
"learning_rate": 1.1934654622512521e-05,
"loss": 1.0496,
"step": 3385
},
{
"epoch": 1.4848883048620236,
"grad_norm": 0.27525372160484834,
"learning_rate": 1.1909646430546155e-05,
"loss": 1.0766,
"step": 3390
},
{
"epoch": 1.487078405606658,
"grad_norm": 0.27889170137924035,
"learning_rate": 1.1884625837279483e-05,
"loss": 1.0816,
"step": 3395
},
{
"epoch": 1.489268506351292,
"grad_norm": 0.2661192505700371,
"learning_rate": 1.1859593005196979e-05,
"loss": 1.0592,
"step": 3400
},
{
"epoch": 1.4914586070959264,
"grad_norm": 0.2735408402306572,
"learning_rate": 1.1834548096862588e-05,
"loss": 1.0807,
"step": 3405
},
{
"epoch": 1.4936487078405607,
"grad_norm": 0.2673420736962167,
"learning_rate": 1.1809491274918692e-05,
"loss": 1.0715,
"step": 3410
},
{
"epoch": 1.4958388085851948,
"grad_norm": 0.26821892445211293,
"learning_rate": 1.1784422702085024e-05,
"loss": 1.0467,
"step": 3415
},
{
"epoch": 1.4980289093298291,
"grad_norm": 0.2575092291317386,
"learning_rate": 1.1759342541157638e-05,
"loss": 1.0402,
"step": 3420
},
{
"epoch": 1.5002190100744635,
"grad_norm": 0.28334435595013674,
"learning_rate": 1.1734250955007844e-05,
"loss": 1.06,
"step": 3425
},
{
"epoch": 1.5024091108190976,
"grad_norm": 0.2606484435934838,
"learning_rate": 1.170914810658114e-05,
"loss": 1.0244,
"step": 3430
},
{
"epoch": 1.5045992115637319,
"grad_norm": 0.274800688330061,
"learning_rate": 1.168403415889617e-05,
"loss": 1.0402,
"step": 3435
},
{
"epoch": 1.5067893123083662,
"grad_norm": 0.27404489640156304,
"learning_rate": 1.1658909275043644e-05,
"loss": 1.0605,
"step": 3440
},
{
"epoch": 1.5089794130530003,
"grad_norm": 0.2901107001823462,
"learning_rate": 1.1633773618185302e-05,
"loss": 1.0562,
"step": 3445
},
{
"epoch": 1.5111695137976346,
"grad_norm": 0.2727747313422447,
"learning_rate": 1.1608627351552846e-05,
"loss": 1.0807,
"step": 3450
},
{
"epoch": 1.513359614542269,
"grad_norm": 0.2717095695436325,
"learning_rate": 1.1583470638446872e-05,
"loss": 1.0604,
"step": 3455
},
{
"epoch": 1.515549715286903,
"grad_norm": 0.2673851569182466,
"learning_rate": 1.1558303642235813e-05,
"loss": 1.0363,
"step": 3460
},
{
"epoch": 1.5177398160315374,
"grad_norm": 0.2613355528607941,
"learning_rate": 1.1533126526354892e-05,
"loss": 1.0441,
"step": 3465
},
{
"epoch": 1.5199299167761717,
"grad_norm": 0.2732685633304171,
"learning_rate": 1.1507939454305037e-05,
"loss": 1.0379,
"step": 3470
},
{
"epoch": 1.5221200175208058,
"grad_norm": 0.26638353891069905,
"learning_rate": 1.1482742589651843e-05,
"loss": 1.041,
"step": 3475
},
{
"epoch": 1.5243101182654402,
"grad_norm": 0.2647954737777941,
"learning_rate": 1.1457536096024493e-05,
"loss": 1.0908,
"step": 3480
},
{
"epoch": 1.5265002190100745,
"grad_norm": 0.2745155747431527,
"learning_rate": 1.1432320137114697e-05,
"loss": 1.0682,
"step": 3485
},
{
"epoch": 1.5286903197547086,
"grad_norm": 0.26032815898238637,
"learning_rate": 1.1407094876675638e-05,
"loss": 1.0568,
"step": 3490
},
{
"epoch": 1.530880420499343,
"grad_norm": 0.2678966018798207,
"learning_rate": 1.13818604785209e-05,
"loss": 1.0572,
"step": 3495
},
{
"epoch": 1.5330705212439772,
"grad_norm": 0.2625845890583784,
"learning_rate": 1.1356617106523418e-05,
"loss": 1.0645,
"step": 3500
},
{
"epoch": 1.5352606219886114,
"grad_norm": 0.2702043111010277,
"learning_rate": 1.1331364924614387e-05,
"loss": 1.0742,
"step": 3505
},
{
"epoch": 1.5374507227332457,
"grad_norm": 0.26793561456953935,
"learning_rate": 1.1306104096782223e-05,
"loss": 1.0475,
"step": 3510
},
{
"epoch": 1.53964082347788,
"grad_norm": 0.25713647329994405,
"learning_rate": 1.1280834787071488e-05,
"loss": 1.06,
"step": 3515
},
{
"epoch": 1.541830924222514,
"grad_norm": 0.27664155878420915,
"learning_rate": 1.1255557159581829e-05,
"loss": 1.0729,
"step": 3520
},
{
"epoch": 1.5440210249671484,
"grad_norm": 0.28295251997882537,
"learning_rate": 1.1230271378466907e-05,
"loss": 1.0494,
"step": 3525
},
{
"epoch": 1.5462111257117828,
"grad_norm": 0.2702429683089981,
"learning_rate": 1.1204977607933321e-05,
"loss": 1.0693,
"step": 3530
},
{
"epoch": 1.5484012264564169,
"grad_norm": 0.26693259673082964,
"learning_rate": 1.117967601223957e-05,
"loss": 1.0648,
"step": 3535
},
{
"epoch": 1.5505913272010512,
"grad_norm": 0.26707808657095433,
"learning_rate": 1.1154366755694964e-05,
"loss": 1.0539,
"step": 3540
},
{
"epoch": 1.5527814279456855,
"grad_norm": 0.26813358409375265,
"learning_rate": 1.1129050002658563e-05,
"loss": 1.073,
"step": 3545
},
{
"epoch": 1.5549715286903196,
"grad_norm": 0.2808358737118327,
"learning_rate": 1.1103725917538106e-05,
"loss": 1.0449,
"step": 3550
},
{
"epoch": 1.557161629434954,
"grad_norm": 0.2653292385507869,
"learning_rate": 1.107839466478895e-05,
"loss": 1.0426,
"step": 3555
},
{
"epoch": 1.5593517301795883,
"grad_norm": 0.2606618878063292,
"learning_rate": 1.1053056408913001e-05,
"loss": 1.0307,
"step": 3560
},
{
"epoch": 1.5615418309242224,
"grad_norm": 0.27074122465063255,
"learning_rate": 1.1027711314457637e-05,
"loss": 1.0561,
"step": 3565
},
{
"epoch": 1.563731931668857,
"grad_norm": 0.26005870439969836,
"learning_rate": 1.1002359546014654e-05,
"loss": 1.0613,
"step": 3570
},
{
"epoch": 1.565922032413491,
"grad_norm": 0.2639594548692859,
"learning_rate": 1.097700126821918e-05,
"loss": 1.0494,
"step": 3575
},
{
"epoch": 1.5681121331581251,
"grad_norm": 0.2766008683933031,
"learning_rate": 1.0951636645748624e-05,
"loss": 1.073,
"step": 3580
},
{
"epoch": 1.5703022339027597,
"grad_norm": 0.28265496721239436,
"learning_rate": 1.0926265843321591e-05,
"loss": 1.066,
"step": 3585
},
{
"epoch": 1.5724923346473938,
"grad_norm": 0.2693455630957891,
"learning_rate": 1.0900889025696824e-05,
"loss": 1.0572,
"step": 3590
},
{
"epoch": 1.574682435392028,
"grad_norm": 0.27906985295838976,
"learning_rate": 1.0875506357672121e-05,
"loss": 1.0457,
"step": 3595
},
{
"epoch": 1.5768725361366625,
"grad_norm": 0.2624004808687805,
"learning_rate": 1.0850118004083279e-05,
"loss": 1.0697,
"step": 3600
},
{
"epoch": 1.5790626368812966,
"grad_norm": 0.26122640361086863,
"learning_rate": 1.0824724129803019e-05,
"loss": 1.0334,
"step": 3605
},
{
"epoch": 1.5812527376259307,
"grad_norm": 0.2619259175583039,
"learning_rate": 1.0799324899739907e-05,
"loss": 1.057,
"step": 3610
},
{
"epoch": 1.5834428383705652,
"grad_norm": 0.261437829445439,
"learning_rate": 1.0773920478837297e-05,
"loss": 1.0529,
"step": 3615
},
{
"epoch": 1.5856329391151993,
"grad_norm": 0.26033725853769885,
"learning_rate": 1.0748511032072244e-05,
"loss": 1.0355,
"step": 3620
},
{
"epoch": 1.5878230398598334,
"grad_norm": 0.2566394274506097,
"learning_rate": 1.0723096724454447e-05,
"loss": 1.0559,
"step": 3625
},
{
"epoch": 1.590013140604468,
"grad_norm": 0.26276427225977134,
"learning_rate": 1.069767772102517e-05,
"loss": 1.0566,
"step": 3630
},
{
"epoch": 1.592203241349102,
"grad_norm": 0.2632371371848792,
"learning_rate": 1.0672254186856177e-05,
"loss": 1.0482,
"step": 3635
},
{
"epoch": 1.5943933420937362,
"grad_norm": 0.2641026930868911,
"learning_rate": 1.064682628704864e-05,
"loss": 1.04,
"step": 3640
},
{
"epoch": 1.5965834428383707,
"grad_norm": 0.26424518556825116,
"learning_rate": 1.0621394186732098e-05,
"loss": 1.0549,
"step": 3645
},
{
"epoch": 1.5987735435830048,
"grad_norm": 0.27923099117512623,
"learning_rate": 1.0595958051063357e-05,
"loss": 1.0773,
"step": 3650
},
{
"epoch": 1.600963644327639,
"grad_norm": 0.26998119469928633,
"learning_rate": 1.0570518045225438e-05,
"loss": 1.0725,
"step": 3655
},
{
"epoch": 1.6031537450722735,
"grad_norm": 0.2685343508566471,
"learning_rate": 1.0545074334426489e-05,
"loss": 1.027,
"step": 3660
},
{
"epoch": 1.6053438458169076,
"grad_norm": 0.2595911690857489,
"learning_rate": 1.0519627083898716e-05,
"loss": 1.043,
"step": 3665
},
{
"epoch": 1.6075339465615417,
"grad_norm": 0.27641158129272464,
"learning_rate": 1.0494176458897316e-05,
"loss": 1.0598,
"step": 3670
},
{
"epoch": 1.6097240473061762,
"grad_norm": 0.2648548237861228,
"learning_rate": 1.0468722624699401e-05,
"loss": 1.0371,
"step": 3675
},
{
"epoch": 1.6119141480508103,
"grad_norm": 0.2702767226474014,
"learning_rate": 1.0443265746602924e-05,
"loss": 1.0434,
"step": 3680
},
{
"epoch": 1.6141042487954445,
"grad_norm": 0.2621468550352411,
"learning_rate": 1.0417805989925599e-05,
"loss": 1.0529,
"step": 3685
},
{
"epoch": 1.616294349540079,
"grad_norm": 0.2637165057836258,
"learning_rate": 1.0392343520003836e-05,
"loss": 1.0516,
"step": 3690
},
{
"epoch": 1.618484450284713,
"grad_norm": 0.2667005261052336,
"learning_rate": 1.036687850219167e-05,
"loss": 1.0539,
"step": 3695
},
{
"epoch": 1.6206745510293472,
"grad_norm": 0.26569655972150674,
"learning_rate": 1.034141110185968e-05,
"loss": 1.0701,
"step": 3700
},
{
"epoch": 1.6228646517739818,
"grad_norm": 0.26574857790395595,
"learning_rate": 1.0315941484393915e-05,
"loss": 1.0473,
"step": 3705
},
{
"epoch": 1.6250547525186159,
"grad_norm": 0.2768540484061404,
"learning_rate": 1.0290469815194816e-05,
"loss": 1.0416,
"step": 3710
},
{
"epoch": 1.62724485326325,
"grad_norm": 0.270209664019903,
"learning_rate": 1.0264996259676166e-05,
"loss": 1.0434,
"step": 3715
},
{
"epoch": 1.6294349540078845,
"grad_norm": 0.27949726855416546,
"learning_rate": 1.0239520983263977e-05,
"loss": 1.0547,
"step": 3720
},
{
"epoch": 1.6316250547525186,
"grad_norm": 0.2638818494131406,
"learning_rate": 1.0214044151395455e-05,
"loss": 1.0682,
"step": 3725
},
{
"epoch": 1.6338151554971527,
"grad_norm": 0.2675675222607541,
"learning_rate": 1.018856592951789e-05,
"loss": 1.0639,
"step": 3730
},
{
"epoch": 1.6360052562417873,
"grad_norm": 0.2672082249953611,
"learning_rate": 1.0163086483087612e-05,
"loss": 1.0562,
"step": 3735
},
{
"epoch": 1.6381953569864214,
"grad_norm": 0.26705014594504956,
"learning_rate": 1.0137605977568896e-05,
"loss": 1.0473,
"step": 3740
},
{
"epoch": 1.6403854577310555,
"grad_norm": 0.2660961478360659,
"learning_rate": 1.0112124578432901e-05,
"loss": 1.0559,
"step": 3745
},
{
"epoch": 1.64257555847569,
"grad_norm": 0.26645965796828935,
"learning_rate": 1.0086642451156583e-05,
"loss": 1.0648,
"step": 3750
},
{
"epoch": 1.6447656592203241,
"grad_norm": 0.2738983150978546,
"learning_rate": 1.0061159761221629e-05,
"loss": 1.043,
"step": 3755
},
{
"epoch": 1.6469557599649582,
"grad_norm": 0.2798777975444317,
"learning_rate": 1.0035676674113379e-05,
"loss": 1.0551,
"step": 3760
},
{
"epoch": 1.6491458607095928,
"grad_norm": 0.265220883116891,
"learning_rate": 1.001019335531975e-05,
"loss": 1.0543,
"step": 3765
},
{
"epoch": 1.651335961454227,
"grad_norm": 0.2682284038098586,
"learning_rate": 9.984709970330176e-06,
"loss": 1.0549,
"step": 3770
},
{
"epoch": 1.653526062198861,
"grad_norm": 0.25832893494460807,
"learning_rate": 9.959226684634502e-06,
"loss": 1.032,
"step": 3775
},
{
"epoch": 1.6557161629434956,
"grad_norm": 0.257904307751333,
"learning_rate": 9.933743663721945e-06,
"loss": 1.0686,
"step": 3780
},
{
"epoch": 1.6579062636881297,
"grad_norm": 0.2689638363753139,
"learning_rate": 9.908261073079991e-06,
"loss": 1.0602,
"step": 3785
},
{
"epoch": 1.6600963644327638,
"grad_norm": 0.2686614118320103,
"learning_rate": 9.88277907819334e-06,
"loss": 1.0617,
"step": 3790
},
{
"epoch": 1.6622864651773983,
"grad_norm": 0.2627709715046705,
"learning_rate": 9.857297844542812e-06,
"loss": 1.0422,
"step": 3795
},
{
"epoch": 1.6644765659220324,
"grad_norm": 0.26431236568218625,
"learning_rate": 9.831817537604296e-06,
"loss": 1.0258,
"step": 3800
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.28223545312633347,
"learning_rate": 9.80633832284766e-06,
"loss": 1.0355,
"step": 3805
},
{
"epoch": 1.668856767411301,
"grad_norm": 0.2598098177349332,
"learning_rate": 9.78086036573567e-06,
"loss": 1.0344,
"step": 3810
},
{
"epoch": 1.6710468681559352,
"grad_norm": 0.26435044314602746,
"learning_rate": 9.755383831722939e-06,
"loss": 1.0223,
"step": 3815
},
{
"epoch": 1.6732369689005693,
"grad_norm": 0.26109212937735365,
"learning_rate": 9.729908886254825e-06,
"loss": 1.0592,
"step": 3820
},
{
"epoch": 1.6754270696452038,
"grad_norm": 0.2654720571684063,
"learning_rate": 9.704435694766377e-06,
"loss": 1.0408,
"step": 3825
},
{
"epoch": 1.677617170389838,
"grad_norm": 0.258522042520765,
"learning_rate": 9.678964422681258e-06,
"loss": 1.0346,
"step": 3830
},
{
"epoch": 1.679807271134472,
"grad_norm": 0.26289147613332475,
"learning_rate": 9.653495235410654e-06,
"loss": 1.0488,
"step": 3835
},
{
"epoch": 1.6819973718791066,
"grad_norm": 0.25642033853461843,
"learning_rate": 9.628028298352226e-06,
"loss": 1.0359,
"step": 3840
},
{
"epoch": 1.6841874726237407,
"grad_norm": 0.2808061166004908,
"learning_rate": 9.602563776889008e-06,
"loss": 1.0516,
"step": 3845
},
{
"epoch": 1.6863775733683748,
"grad_norm": 0.26470050862147015,
"learning_rate": 9.57710183638836e-06,
"loss": 1.0697,
"step": 3850
},
{
"epoch": 1.6885676741130093,
"grad_norm": 0.26606691954269995,
"learning_rate": 9.551642642200875e-06,
"loss": 1.0758,
"step": 3855
},
{
"epoch": 1.6907577748576434,
"grad_norm": 0.2736697572447457,
"learning_rate": 9.526186359659307e-06,
"loss": 1.0545,
"step": 3860
},
{
"epoch": 1.6929478756022776,
"grad_norm": 0.2603250399961473,
"learning_rate": 9.500733154077508e-06,
"loss": 1.0471,
"step": 3865
},
{
"epoch": 1.695137976346912,
"grad_norm": 0.274926658101149,
"learning_rate": 9.475283190749356e-06,
"loss": 1.0699,
"step": 3870
},
{
"epoch": 1.6973280770915462,
"grad_norm": 0.26279297784971317,
"learning_rate": 9.44983663494765e-06,
"loss": 1.0512,
"step": 3875
},
{
"epoch": 1.6995181778361803,
"grad_norm": 0.2641150816670752,
"learning_rate": 9.424393651923084e-06,
"loss": 1.0564,
"step": 3880
},
{
"epoch": 1.7017082785808149,
"grad_norm": 0.25856068896220813,
"learning_rate": 9.398954406903137e-06,
"loss": 1.0494,
"step": 3885
},
{
"epoch": 1.703898379325449,
"grad_norm": 0.27030707323741804,
"learning_rate": 9.373519065091019e-06,
"loss": 1.0475,
"step": 3890
},
{
"epoch": 1.706088480070083,
"grad_norm": 0.2666495942417461,
"learning_rate": 9.348087791664591e-06,
"loss": 1.0553,
"step": 3895
},
{
"epoch": 1.7082785808147176,
"grad_norm": 0.2612477606748569,
"learning_rate": 9.32266075177529e-06,
"loss": 1.0531,
"step": 3900
},
{
"epoch": 1.7104686815593517,
"grad_norm": 0.25589631959691345,
"learning_rate": 9.297238110547075e-06,
"loss": 1.0408,
"step": 3905
},
{
"epoch": 1.7126587823039858,
"grad_norm": 0.2616645026641856,
"learning_rate": 9.271820033075311e-06,
"loss": 1.0406,
"step": 3910
},
{
"epoch": 1.7148488830486204,
"grad_norm": 0.2586391146779032,
"learning_rate": 9.246406684425757e-06,
"loss": 1.0682,
"step": 3915
},
{
"epoch": 1.7170389837932545,
"grad_norm": 0.2627435639301313,
"learning_rate": 9.220998229633447e-06,
"loss": 1.0512,
"step": 3920
},
{
"epoch": 1.7192290845378886,
"grad_norm": 0.27131216390216845,
"learning_rate": 9.195594833701634e-06,
"loss": 1.0637,
"step": 3925
},
{
"epoch": 1.7214191852825231,
"grad_norm": 0.2669482190799787,
"learning_rate": 9.170196661600725e-06,
"loss": 1.0574,
"step": 3930
},
{
"epoch": 1.7236092860271572,
"grad_norm": 0.26420187232497483,
"learning_rate": 9.144803878267191e-06,
"loss": 1.0557,
"step": 3935
},
{
"epoch": 1.7257993867717913,
"grad_norm": 0.2659324687775046,
"learning_rate": 9.119416648602526e-06,
"loss": 1.0875,
"step": 3940
},
{
"epoch": 1.727989487516426,
"grad_norm": 0.2704024968469015,
"learning_rate": 9.094035137472148e-06,
"loss": 1.0291,
"step": 3945
},
{
"epoch": 1.73017958826106,
"grad_norm": 0.2704315331590373,
"learning_rate": 9.068659509704337e-06,
"loss": 1.0303,
"step": 3950
},
{
"epoch": 1.732369689005694,
"grad_norm": 0.26075951030660144,
"learning_rate": 9.043289930089166e-06,
"loss": 1.0416,
"step": 3955
},
{
"epoch": 1.7345597897503287,
"grad_norm": 0.2798473730769966,
"learning_rate": 9.017926563377444e-06,
"loss": 1.0432,
"step": 3960
},
{
"epoch": 1.7367498904949628,
"grad_norm": 0.287189918659511,
"learning_rate": 8.992569574279613e-06,
"loss": 1.0646,
"step": 3965
},
{
"epoch": 1.7389399912395969,
"grad_norm": 0.271002464852271,
"learning_rate": 8.967219127464714e-06,
"loss": 1.0738,
"step": 3970
},
{
"epoch": 1.7411300919842314,
"grad_norm": 0.28869898620552076,
"learning_rate": 8.941875387559296e-06,
"loss": 1.0512,
"step": 3975
},
{
"epoch": 1.7433201927288655,
"grad_norm": 0.2580747875213114,
"learning_rate": 8.916538519146355e-06,
"loss": 1.0697,
"step": 3980
},
{
"epoch": 1.7455102934734996,
"grad_norm": 0.26720671681904395,
"learning_rate": 8.891208686764263e-06,
"loss": 1.0516,
"step": 3985
},
{
"epoch": 1.7477003942181342,
"grad_norm": 0.27472866838079185,
"learning_rate": 8.865886054905695e-06,
"loss": 1.0564,
"step": 3990
},
{
"epoch": 1.7498904949627683,
"grad_norm": 0.26848146534574147,
"learning_rate": 8.840570788016578e-06,
"loss": 1.0631,
"step": 3995
},
{
"epoch": 1.7520805957074024,
"grad_norm": 0.26914337171369807,
"learning_rate": 8.815263050494994e-06,
"loss": 1.0875,
"step": 4000
},
{
"epoch": 1.754270696452037,
"grad_norm": 0.2636833138291496,
"learning_rate": 8.789963006690142e-06,
"loss": 1.0604,
"step": 4005
},
{
"epoch": 1.756460797196671,
"grad_norm": 0.2583287936852473,
"learning_rate": 8.764670820901256e-06,
"loss": 1.0395,
"step": 4010
},
{
"epoch": 1.7586508979413051,
"grad_norm": 0.2785046304924881,
"learning_rate": 8.739386657376532e-06,
"loss": 1.091,
"step": 4015
},
{
"epoch": 1.7608409986859397,
"grad_norm": 0.27037923704909567,
"learning_rate": 8.714110680312082e-06,
"loss": 1.0301,
"step": 4020
},
{
"epoch": 1.7630310994305738,
"grad_norm": 0.2649755210179008,
"learning_rate": 8.688843053850834e-06,
"loss": 1.0574,
"step": 4025
},
{
"epoch": 1.765221200175208,
"grad_norm": 0.2671017408683602,
"learning_rate": 8.663583942081514e-06,
"loss": 1.0371,
"step": 4030
},
{
"epoch": 1.7674113009198424,
"grad_norm": 0.26430762243636857,
"learning_rate": 8.638333509037537e-06,
"loss": 1.0664,
"step": 4035
},
{
"epoch": 1.7696014016644765,
"grad_norm": 0.281966657857878,
"learning_rate": 8.613091918695954e-06,
"loss": 1.0617,
"step": 4040
},
{
"epoch": 1.7717915024091109,
"grad_norm": 0.2641603417837672,
"learning_rate": 8.587859334976405e-06,
"loss": 1.0414,
"step": 4045
},
{
"epoch": 1.7739816031537452,
"grad_norm": 0.2550399068901032,
"learning_rate": 8.562635921740029e-06,
"loss": 1.0186,
"step": 4050
},
{
"epoch": 1.7761717038983793,
"grad_norm": 0.258436517607505,
"learning_rate": 8.537421842788416e-06,
"loss": 1.0506,
"step": 4055
},
{
"epoch": 1.7783618046430136,
"grad_norm": 0.27070416646617784,
"learning_rate": 8.512217261862546e-06,
"loss": 1.0707,
"step": 4060
},
{
"epoch": 1.780551905387648,
"grad_norm": 0.2623515052424796,
"learning_rate": 8.487022342641706e-06,
"loss": 1.0268,
"step": 4065
},
{
"epoch": 1.782742006132282,
"grad_norm": 0.2580098583224272,
"learning_rate": 8.461837248742445e-06,
"loss": 1.0547,
"step": 4070
},
{
"epoch": 1.7849321068769164,
"grad_norm": 0.271853498810519,
"learning_rate": 8.43666214371751e-06,
"loss": 1.0709,
"step": 4075
},
{
"epoch": 1.7871222076215507,
"grad_norm": 0.2659287637845158,
"learning_rate": 8.411497191054772e-06,
"loss": 1.0518,
"step": 4080
},
{
"epoch": 1.7893123083661848,
"grad_norm": 0.27283290836176466,
"learning_rate": 8.386342554176184e-06,
"loss": 1.0584,
"step": 4085
},
{
"epoch": 1.7915024091108191,
"grad_norm": 0.2651207636177441,
"learning_rate": 8.361198396436688e-06,
"loss": 1.0291,
"step": 4090
},
{
"epoch": 1.7936925098554535,
"grad_norm": 0.2703741744888326,
"learning_rate": 8.336064881123199e-06,
"loss": 1.0535,
"step": 4095
},
{
"epoch": 1.7958826106000876,
"grad_norm": 0.277174424727332,
"learning_rate": 8.310942171453502e-06,
"loss": 1.0611,
"step": 4100
},
{
"epoch": 1.798072711344722,
"grad_norm": 0.2605646974649158,
"learning_rate": 8.285830430575218e-06,
"loss": 1.0518,
"step": 4105
},
{
"epoch": 1.8002628120893562,
"grad_norm": 0.27113214575980377,
"learning_rate": 8.260729821564733e-06,
"loss": 1.066,
"step": 4110
},
{
"epoch": 1.8024529128339903,
"grad_norm": 0.26188100923955243,
"learning_rate": 8.235640507426136e-06,
"loss": 1.0637,
"step": 4115
},
{
"epoch": 1.8046430135786247,
"grad_norm": 0.2647927642271762,
"learning_rate": 8.210562651090184e-06,
"loss": 1.0709,
"step": 4120
},
{
"epoch": 1.806833114323259,
"grad_norm": 0.2659054715179537,
"learning_rate": 8.185496415413215e-06,
"loss": 1.0664,
"step": 4125
},
{
"epoch": 1.809023215067893,
"grad_norm": 0.26152364121681076,
"learning_rate": 8.160441963176097e-06,
"loss": 1.0395,
"step": 4130
},
{
"epoch": 1.8112133158125274,
"grad_norm": 0.2598225560908046,
"learning_rate": 8.13539945708319e-06,
"loss": 1.0629,
"step": 4135
},
{
"epoch": 1.8134034165571618,
"grad_norm": 0.26952205543318625,
"learning_rate": 8.110369059761264e-06,
"loss": 1.0576,
"step": 4140
},
{
"epoch": 1.8155935173017959,
"grad_norm": 0.27030120261610047,
"learning_rate": 8.085350933758456e-06,
"loss": 1.0762,
"step": 4145
},
{
"epoch": 1.8177836180464302,
"grad_norm": 0.25627881011325426,
"learning_rate": 8.060345241543222e-06,
"loss": 1.051,
"step": 4150
},
{
"epoch": 1.8199737187910645,
"grad_norm": 0.2655782481715189,
"learning_rate": 8.035352145503258e-06,
"loss": 1.0602,
"step": 4155
},
{
"epoch": 1.8221638195356986,
"grad_norm": 0.26293288923024916,
"learning_rate": 8.010371807944475e-06,
"loss": 1.0561,
"step": 4160
},
{
"epoch": 1.824353920280333,
"grad_norm": 0.2670317031054168,
"learning_rate": 7.985404391089917e-06,
"loss": 1.0318,
"step": 4165
},
{
"epoch": 1.8265440210249673,
"grad_norm": 0.2599529233299643,
"learning_rate": 7.960450057078726e-06,
"loss": 1.0555,
"step": 4170
},
{
"epoch": 1.8287341217696014,
"grad_norm": 0.2640429458071465,
"learning_rate": 7.935508967965092e-06,
"loss": 1.0582,
"step": 4175
},
{
"epoch": 1.8309242225142357,
"grad_norm": 0.2611223291897593,
"learning_rate": 7.910581285717173e-06,
"loss": 1.058,
"step": 4180
},
{
"epoch": 1.83311432325887,
"grad_norm": 0.27332241824491227,
"learning_rate": 7.885667172216079e-06,
"loss": 1.0605,
"step": 4185
},
{
"epoch": 1.8353044240035041,
"grad_norm": 0.26068149812972996,
"learning_rate": 7.860766789254802e-06,
"loss": 1.0338,
"step": 4190
},
{
"epoch": 1.8374945247481385,
"grad_norm": 0.27123772619112196,
"learning_rate": 7.835880298537158e-06,
"loss": 1.0496,
"step": 4195
},
{
"epoch": 1.8396846254927728,
"grad_norm": 0.2595476192894215,
"learning_rate": 7.811007861676759e-06,
"loss": 1.0666,
"step": 4200
},
{
"epoch": 1.841874726237407,
"grad_norm": 0.26422252826709297,
"learning_rate": 7.786149640195933e-06,
"loss": 1.0271,
"step": 4205
},
{
"epoch": 1.8440648269820412,
"grad_norm": 0.26134111456960313,
"learning_rate": 7.761305795524716e-06,
"loss": 1.0363,
"step": 4210
},
{
"epoch": 1.8462549277266755,
"grad_norm": 0.2756776421734931,
"learning_rate": 7.736476488999768e-06,
"loss": 1.0475,
"step": 4215
},
{
"epoch": 1.8484450284713096,
"grad_norm": 0.2702364297034732,
"learning_rate": 7.711661881863331e-06,
"loss": 1.0268,
"step": 4220
},
{
"epoch": 1.850635129215944,
"grad_norm": 0.26975541398998254,
"learning_rate": 7.686862135262205e-06,
"loss": 1.0371,
"step": 4225
},
{
"epoch": 1.8528252299605783,
"grad_norm": 0.2644048495956391,
"learning_rate": 7.66207741024667e-06,
"loss": 1.0611,
"step": 4230
},
{
"epoch": 1.8550153307052124,
"grad_norm": 0.26966735694678484,
"learning_rate": 7.637307867769462e-06,
"loss": 1.0771,
"step": 4235
},
{
"epoch": 1.8572054314498467,
"grad_norm": 0.26866968091520127,
"learning_rate": 7.612553668684729e-06,
"loss": 1.06,
"step": 4240
},
{
"epoch": 1.859395532194481,
"grad_norm": 0.25872064254050153,
"learning_rate": 7.587814973746963e-06,
"loss": 1.051,
"step": 4245
},
{
"epoch": 1.8615856329391152,
"grad_norm": 0.27201780645953577,
"learning_rate": 7.563091943609984e-06,
"loss": 1.0908,
"step": 4250
},
{
"epoch": 1.8637757336837495,
"grad_norm": 0.27623982941637193,
"learning_rate": 7.538384738825876e-06,
"loss": 1.0432,
"step": 4255
},
{
"epoch": 1.8659658344283838,
"grad_norm": 0.2658326988358835,
"learning_rate": 7.513693519843956e-06,
"loss": 1.0855,
"step": 4260
},
{
"epoch": 1.868155935173018,
"grad_norm": 0.2603354692417316,
"learning_rate": 7.489018447009737e-06,
"loss": 1.0502,
"step": 4265
},
{
"epoch": 1.8703460359176522,
"grad_norm": 0.2634605332883379,
"learning_rate": 7.46435968056386e-06,
"loss": 1.0662,
"step": 4270
},
{
"epoch": 1.8725361366622866,
"grad_norm": 0.26436138724333197,
"learning_rate": 7.439717380641087e-06,
"loss": 1.0361,
"step": 4275
},
{
"epoch": 1.8747262374069207,
"grad_norm": 0.2695505088765098,
"learning_rate": 7.415091707269246e-06,
"loss": 1.0701,
"step": 4280
},
{
"epoch": 1.876916338151555,
"grad_norm": 0.2711600184846312,
"learning_rate": 7.390482820368179e-06,
"loss": 1.06,
"step": 4285
},
{
"epoch": 1.8791064388961893,
"grad_norm": 0.27114060939285184,
"learning_rate": 7.365890879748732e-06,
"loss": 1.0307,
"step": 4290
},
{
"epoch": 1.8812965396408234,
"grad_norm": 0.27852182132375797,
"learning_rate": 7.341316045111683e-06,
"loss": 1.0596,
"step": 4295
},
{
"epoch": 1.8834866403854578,
"grad_norm": 0.2657978196594915,
"learning_rate": 7.3167584760467415e-06,
"loss": 1.0471,
"step": 4300
},
{
"epoch": 1.885676741130092,
"grad_norm": 0.2674058529573235,
"learning_rate": 7.292218332031489e-06,
"loss": 1.0408,
"step": 4305
},
{
"epoch": 1.8878668418747262,
"grad_norm": 0.26909849042401646,
"learning_rate": 7.267695772430336e-06,
"loss": 1.05,
"step": 4310
},
{
"epoch": 1.8900569426193605,
"grad_norm": 0.2717698485268081,
"learning_rate": 7.243190956493518e-06,
"loss": 1.0623,
"step": 4315
},
{
"epoch": 1.8922470433639949,
"grad_norm": 0.2841986212548669,
"learning_rate": 7.218704043356026e-06,
"loss": 1.0555,
"step": 4320
},
{
"epoch": 1.894437144108629,
"grad_norm": 0.27274737898653567,
"learning_rate": 7.194235192036598e-06,
"loss": 1.0453,
"step": 4325
},
{
"epoch": 1.8966272448532633,
"grad_norm": 0.26224406560969943,
"learning_rate": 7.1697845614366825e-06,
"loss": 1.0391,
"step": 4330
},
{
"epoch": 1.8988173455978976,
"grad_norm": 0.26110104192389827,
"learning_rate": 7.145352310339391e-06,
"loss": 1.0475,
"step": 4335
},
{
"epoch": 1.9010074463425317,
"grad_norm": 0.2624783893516754,
"learning_rate": 7.120938597408487e-06,
"loss": 1.0494,
"step": 4340
},
{
"epoch": 1.903197547087166,
"grad_norm": 0.2642985663653997,
"learning_rate": 7.0965435811873384e-06,
"loss": 1.075,
"step": 4345
},
{
"epoch": 1.9053876478318004,
"grad_norm": 0.26634192070298696,
"learning_rate": 7.072167420097904e-06,
"loss": 1.0734,
"step": 4350
},
{
"epoch": 1.9075777485764345,
"grad_norm": 0.2672053957248454,
"learning_rate": 7.047810272439699e-06,
"loss": 1.0496,
"step": 4355
},
{
"epoch": 1.9097678493210688,
"grad_norm": 0.25758269759882013,
"learning_rate": 7.023472296388747e-06,
"loss": 1.05,
"step": 4360
},
{
"epoch": 1.9119579500657031,
"grad_norm": 0.26583912661168085,
"learning_rate": 6.999153649996595e-06,
"loss": 1.0494,
"step": 4365
},
{
"epoch": 1.9141480508103372,
"grad_norm": 0.2614518285010944,
"learning_rate": 6.974854491189243e-06,
"loss": 1.0479,
"step": 4370
},
{
"epoch": 1.9163381515549716,
"grad_norm": 0.25956078456974213,
"learning_rate": 6.95057497776615e-06,
"loss": 1.0346,
"step": 4375
},
{
"epoch": 1.9185282522996059,
"grad_norm": 0.26309155812958857,
"learning_rate": 6.926315267399193e-06,
"loss": 1.0738,
"step": 4380
},
{
"epoch": 1.92071835304424,
"grad_norm": 0.26694616315083763,
"learning_rate": 6.902075517631642e-06,
"loss": 1.057,
"step": 4385
},
{
"epoch": 1.9229084537888743,
"grad_norm": 0.2715223161888118,
"learning_rate": 6.877855885877154e-06,
"loss": 1.0488,
"step": 4390
},
{
"epoch": 1.9250985545335086,
"grad_norm": 0.2692563426788601,
"learning_rate": 6.853656529418733e-06,
"loss": 1.0666,
"step": 4395
},
{
"epoch": 1.9272886552781427,
"grad_norm": 0.2637247892608681,
"learning_rate": 6.82947760540771e-06,
"loss": 1.0316,
"step": 4400
},
{
"epoch": 1.929478756022777,
"grad_norm": 0.2631628957902492,
"learning_rate": 6.805319270862736e-06,
"loss": 1.0568,
"step": 4405
},
{
"epoch": 1.9316688567674114,
"grad_norm": 0.2639983365666047,
"learning_rate": 6.7811816826687425e-06,
"loss": 1.0564,
"step": 4410
},
{
"epoch": 1.9338589575120455,
"grad_norm": 0.26678744265203513,
"learning_rate": 6.757064997575944e-06,
"loss": 1.0738,
"step": 4415
},
{
"epoch": 1.9360490582566798,
"grad_norm": 0.276712279089104,
"learning_rate": 6.732969372198808e-06,
"loss": 1.0814,
"step": 4420
},
{
"epoch": 1.9382391590013142,
"grad_norm": 0.25482014051465834,
"learning_rate": 6.708894963015034e-06,
"loss": 1.0412,
"step": 4425
},
{
"epoch": 1.9404292597459483,
"grad_norm": 0.2632248605954728,
"learning_rate": 6.684841926364547e-06,
"loss": 1.0496,
"step": 4430
},
{
"epoch": 1.9426193604905826,
"grad_norm": 0.2612596534133882,
"learning_rate": 6.660810418448475e-06,
"loss": 1.0391,
"step": 4435
},
{
"epoch": 1.944809461235217,
"grad_norm": 0.28939889023140064,
"learning_rate": 6.63680059532814e-06,
"loss": 1.0693,
"step": 4440
},
{
"epoch": 1.946999561979851,
"grad_norm": 0.2586881368868953,
"learning_rate": 6.612812612924051e-06,
"loss": 1.016,
"step": 4445
},
{
"epoch": 1.9491896627244854,
"grad_norm": 0.27522451220029226,
"learning_rate": 6.58884662701486e-06,
"loss": 1.0527,
"step": 4450
},
{
"epoch": 1.9513797634691197,
"grad_norm": 0.2598623624479845,
"learning_rate": 6.564902793236398e-06,
"loss": 1.0336,
"step": 4455
},
{
"epoch": 1.9535698642137538,
"grad_norm": 0.27748246484682554,
"learning_rate": 6.540981267080624e-06,
"loss": 1.0297,
"step": 4460
},
{
"epoch": 1.955759964958388,
"grad_norm": 0.2685656534630219,
"learning_rate": 6.517082203894636e-06,
"loss": 1.0537,
"step": 4465
},
{
"epoch": 1.9579500657030224,
"grad_norm": 0.26891368891515,
"learning_rate": 6.493205758879657e-06,
"loss": 1.0734,
"step": 4470
},
{
"epoch": 1.9601401664476565,
"grad_norm": 0.2674766199421492,
"learning_rate": 6.469352087090022e-06,
"loss": 1.0562,
"step": 4475
},
{
"epoch": 1.9623302671922909,
"grad_norm": 0.25788874477297874,
"learning_rate": 6.445521343432189e-06,
"loss": 1.0404,
"step": 4480
},
{
"epoch": 1.9645203679369252,
"grad_norm": 0.2819872242228695,
"learning_rate": 6.421713682663697e-06,
"loss": 1.0604,
"step": 4485
},
{
"epoch": 1.9667104686815593,
"grad_norm": 0.27010244363463976,
"learning_rate": 6.397929259392208e-06,
"loss": 1.0295,
"step": 4490
},
{
"epoch": 1.9689005694261936,
"grad_norm": 0.25928825181925214,
"learning_rate": 6.374168228074471e-06,
"loss": 1.0434,
"step": 4495
},
{
"epoch": 1.971090670170828,
"grad_norm": 0.2670727572943277,
"learning_rate": 6.3504307430153165e-06,
"loss": 1.0598,
"step": 4500
},
{
"epoch": 1.973280770915462,
"grad_norm": 0.27662824636944733,
"learning_rate": 6.326716958366676e-06,
"loss": 1.052,
"step": 4505
},
{
"epoch": 1.9754708716600964,
"grad_norm": 0.2645319859255144,
"learning_rate": 6.3030270281265745e-06,
"loss": 1.0461,
"step": 4510
},
{
"epoch": 1.9776609724047307,
"grad_norm": 0.27311415864779,
"learning_rate": 6.27936110613811e-06,
"loss": 1.0367,
"step": 4515
},
{
"epoch": 1.9798510731493648,
"grad_norm": 0.2640788171682172,
"learning_rate": 6.255719346088483e-06,
"loss": 1.068,
"step": 4520
},
{
"epoch": 1.9820411738939991,
"grad_norm": 0.26269876183491886,
"learning_rate": 6.232101901507978e-06,
"loss": 1.0424,
"step": 4525
},
{
"epoch": 1.9842312746386335,
"grad_norm": 0.2666587948643593,
"learning_rate": 6.20850892576898e-06,
"loss": 1.0307,
"step": 4530
},
{
"epoch": 1.9864213753832676,
"grad_norm": 0.25956508969478437,
"learning_rate": 6.184940572084972e-06,
"loss": 1.048,
"step": 4535
},
{
"epoch": 1.988611476127902,
"grad_norm": 0.2716220580126288,
"learning_rate": 6.16139699350953e-06,
"loss": 1.0408,
"step": 4540
},
{
"epoch": 1.9908015768725362,
"grad_norm": 0.27436626928192,
"learning_rate": 6.1378783429353596e-06,
"loss": 1.0406,
"step": 4545
},
{
"epoch": 1.9929916776171703,
"grad_norm": 0.267050414240659,
"learning_rate": 6.114384773093265e-06,
"loss": 1.0527,
"step": 4550
},
{
"epoch": 1.9951817783618047,
"grad_norm": 0.2791092291038011,
"learning_rate": 6.090916436551185e-06,
"loss": 1.0594,
"step": 4555
},
{
"epoch": 1.997371879106439,
"grad_norm": 0.2536778001853949,
"learning_rate": 6.067473485713191e-06,
"loss": 1.0494,
"step": 4560
},
{
"epoch": 1.999561979851073,
"grad_norm": 0.26436103598489785,
"learning_rate": 6.044056072818495e-06,
"loss": 1.05,
"step": 4565
},
{
"epoch": 2.001752080595707,
"grad_norm": 0.26093422619007045,
"learning_rate": 6.020664349940471e-06,
"loss": 1.0168,
"step": 4570
},
{
"epoch": 2.0039421813403417,
"grad_norm": 0.2689011008218334,
"learning_rate": 5.997298468985652e-06,
"loss": 1.0332,
"step": 4575
},
{
"epoch": 2.006132282084976,
"grad_norm": 0.27970037489217886,
"learning_rate": 5.973958581692765e-06,
"loss": 1.0086,
"step": 4580
},
{
"epoch": 2.00832238282961,
"grad_norm": 0.257258807469164,
"learning_rate": 5.950644839631727e-06,
"loss": 1.0227,
"step": 4585
},
{
"epoch": 2.0105124835742445,
"grad_norm": 0.2623857732646142,
"learning_rate": 5.927357394202665e-06,
"loss": 1.0256,
"step": 4590
},
{
"epoch": 2.0127025843188786,
"grad_norm": 0.26667045216828666,
"learning_rate": 5.904096396634935e-06,
"loss": 0.9852,
"step": 4595
},
{
"epoch": 2.0148926850635127,
"grad_norm": 0.2654589837312285,
"learning_rate": 5.880861997986151e-06,
"loss": 1.0217,
"step": 4600
},
{
"epoch": 2.0170827858081473,
"grad_norm": 0.2622730969078804,
"learning_rate": 5.8576543491411734e-06,
"loss": 1.0072,
"step": 4605
},
{
"epoch": 2.0192728865527814,
"grad_norm": 0.25609804027452565,
"learning_rate": 5.834473600811168e-06,
"loss": 0.9986,
"step": 4610
},
{
"epoch": 2.0214629872974155,
"grad_norm": 0.26444734638461304,
"learning_rate": 5.8113199035325885e-06,
"loss": 0.9914,
"step": 4615
},
{
"epoch": 2.02365308804205,
"grad_norm": 0.2759313276801203,
"learning_rate": 5.788193407666234e-06,
"loss": 1.0354,
"step": 4620
},
{
"epoch": 2.025843188786684,
"grad_norm": 0.26725617714670247,
"learning_rate": 5.765094263396257e-06,
"loss": 1.0139,
"step": 4625
},
{
"epoch": 2.0280332895313182,
"grad_norm": 0.2664407521165841,
"learning_rate": 5.742022620729177e-06,
"loss": 1.0203,
"step": 4630
},
{
"epoch": 2.0302233902759528,
"grad_norm": 0.2610201181123097,
"learning_rate": 5.718978629492924e-06,
"loss": 1.0201,
"step": 4635
},
{
"epoch": 2.032413491020587,
"grad_norm": 0.2709130624912754,
"learning_rate": 5.695962439335853e-06,
"loss": 1.0299,
"step": 4640
},
{
"epoch": 2.034603591765221,
"grad_norm": 0.2685235958201942,
"learning_rate": 5.672974199725785e-06,
"loss": 1.0191,
"step": 4645
},
{
"epoch": 2.0367936925098555,
"grad_norm": 0.26748654369266783,
"learning_rate": 5.650014059949033e-06,
"loss": 0.982,
"step": 4650
},
{
"epoch": 2.0389837932544896,
"grad_norm": 0.2605321089033867,
"learning_rate": 5.627082169109417e-06,
"loss": 1.0314,
"step": 4655
},
{
"epoch": 2.0411738939991237,
"grad_norm": 0.2587831303240616,
"learning_rate": 5.604178676127314e-06,
"loss": 1.0309,
"step": 4660
},
{
"epoch": 2.0433639947437583,
"grad_norm": 0.2756781260939094,
"learning_rate": 5.581303729738681e-06,
"loss": 1.0369,
"step": 4665
},
{
"epoch": 2.0455540954883924,
"grad_norm": 0.26243765749849607,
"learning_rate": 5.558457478494096e-06,
"loss": 1.0271,
"step": 4670
},
{
"epoch": 2.0477441962330265,
"grad_norm": 0.2674529202379436,
"learning_rate": 5.5356400707578e-06,
"loss": 1.0256,
"step": 4675
},
{
"epoch": 2.049934296977661,
"grad_norm": 0.26296862723387604,
"learning_rate": 5.512851654706694e-06,
"loss": 1.0414,
"step": 4680
},
{
"epoch": 2.052124397722295,
"grad_norm": 0.27045918514005246,
"learning_rate": 5.490092378329441e-06,
"loss": 1.0229,
"step": 4685
},
{
"epoch": 2.0543144984669293,
"grad_norm": 0.27477672949598386,
"learning_rate": 5.467362389425441e-06,
"loss": 1.0129,
"step": 4690
},
{
"epoch": 2.056504599211564,
"grad_norm": 0.2805167534895326,
"learning_rate": 5.444661835603925e-06,
"loss": 1.0311,
"step": 4695
},
{
"epoch": 2.058694699956198,
"grad_norm": 0.26414962391577573,
"learning_rate": 5.421990864282953e-06,
"loss": 1.0109,
"step": 4700
},
{
"epoch": 2.060884800700832,
"grad_norm": 0.27441404134914826,
"learning_rate": 5.399349622688479e-06,
"loss": 1.0135,
"step": 4705
},
{
"epoch": 2.0630749014454666,
"grad_norm": 0.2579083757623819,
"learning_rate": 5.3767382578534e-06,
"loss": 0.9945,
"step": 4710
},
{
"epoch": 2.0652650021901007,
"grad_norm": 0.26992308351772354,
"learning_rate": 5.354156916616589e-06,
"loss": 1.0139,
"step": 4715
},
{
"epoch": 2.0674551029347352,
"grad_norm": 0.27138173828265577,
"learning_rate": 5.331605745621938e-06,
"loss": 1.0236,
"step": 4720
},
{
"epoch": 2.0696452036793693,
"grad_norm": 0.2643761326181544,
"learning_rate": 5.30908489131742e-06,
"loss": 1.0051,
"step": 4725
},
{
"epoch": 2.0718353044240034,
"grad_norm": 0.2701535434252902,
"learning_rate": 5.286594499954121e-06,
"loss": 1.0361,
"step": 4730
},
{
"epoch": 2.074025405168638,
"grad_norm": 0.2792019275770248,
"learning_rate": 5.264134717585313e-06,
"loss": 0.9674,
"step": 4735
},
{
"epoch": 2.076215505913272,
"grad_norm": 0.2731806421275931,
"learning_rate": 5.241705690065489e-06,
"loss": 1.0164,
"step": 4740
},
{
"epoch": 2.078405606657906,
"grad_norm": 0.28547898058692295,
"learning_rate": 5.219307563049411e-06,
"loss": 1.0262,
"step": 4745
},
{
"epoch": 2.0805957074025407,
"grad_norm": 0.26085931171959603,
"learning_rate": 5.196940481991179e-06,
"loss": 1.0084,
"step": 4750
},
{
"epoch": 2.082785808147175,
"grad_norm": 0.2819568795934252,
"learning_rate": 5.174604592143273e-06,
"loss": 1.0195,
"step": 4755
},
{
"epoch": 2.084975908891809,
"grad_norm": 0.25860822102705483,
"learning_rate": 5.152300038555624e-06,
"loss": 1.0324,
"step": 4760
},
{
"epoch": 2.0871660096364435,
"grad_norm": 0.2672824104490307,
"learning_rate": 5.130026966074671e-06,
"loss": 1.0164,
"step": 4765
},
{
"epoch": 2.0893561103810776,
"grad_norm": 0.2659892744135754,
"learning_rate": 5.107785519342388e-06,
"loss": 1.0213,
"step": 4770
},
{
"epoch": 2.0915462111257117,
"grad_norm": 0.26571378582465105,
"learning_rate": 5.085575842795398e-06,
"loss": 1.0381,
"step": 4775
},
{
"epoch": 2.0937363118703463,
"grad_norm": 0.2688782038959789,
"learning_rate": 5.06339808066399e-06,
"loss": 1.0475,
"step": 4780
},
{
"epoch": 2.0959264126149804,
"grad_norm": 0.26362420646124834,
"learning_rate": 5.041252376971213e-06,
"loss": 1.0102,
"step": 4785
},
{
"epoch": 2.0981165133596145,
"grad_norm": 0.27101641391517345,
"learning_rate": 5.019138875531917e-06,
"loss": 1.0389,
"step": 4790
},
{
"epoch": 2.100306614104249,
"grad_norm": 0.2576814050017412,
"learning_rate": 4.99705771995183e-06,
"loss": 1.0223,
"step": 4795
},
{
"epoch": 2.102496714848883,
"grad_norm": 0.27045735827617223,
"learning_rate": 4.975009053626634e-06,
"loss": 1.0145,
"step": 4800
},
{
"epoch": 2.1046868155935172,
"grad_norm": 0.26113258147724255,
"learning_rate": 4.952993019741013e-06,
"loss": 1.0297,
"step": 4805
},
{
"epoch": 2.1068769163381518,
"grad_norm": 0.27448549506036557,
"learning_rate": 4.93100976126775e-06,
"loss": 1.0096,
"step": 4810
},
{
"epoch": 2.109067017082786,
"grad_norm": 0.27280933957719383,
"learning_rate": 4.909059420966768e-06,
"loss": 1.0092,
"step": 4815
},
{
"epoch": 2.11125711782742,
"grad_norm": 0.264506277380279,
"learning_rate": 4.887142141384222e-06,
"loss": 1.0426,
"step": 4820
},
{
"epoch": 2.1134472185720545,
"grad_norm": 0.2770787791773093,
"learning_rate": 4.865258064851579e-06,
"loss": 1.0316,
"step": 4825
},
{
"epoch": 2.1156373193166886,
"grad_norm": 0.2557758077437797,
"learning_rate": 4.8434073334846785e-06,
"loss": 1.0371,
"step": 4830
},
{
"epoch": 2.1178274200613227,
"grad_norm": 0.2610392778424005,
"learning_rate": 4.821590089182814e-06,
"loss": 1.0043,
"step": 4835
},
{
"epoch": 2.1200175208059573,
"grad_norm": 0.26958938148423095,
"learning_rate": 4.799806473627813e-06,
"loss": 1.0297,
"step": 4840
},
{
"epoch": 2.1222076215505914,
"grad_norm": 0.2660024950094293,
"learning_rate": 4.778056628283114e-06,
"loss": 0.9949,
"step": 4845
},
{
"epoch": 2.1243977222952255,
"grad_norm": 0.2703860686810657,
"learning_rate": 4.756340694392863e-06,
"loss": 1.0152,
"step": 4850
},
{
"epoch": 2.12658782303986,
"grad_norm": 0.2726552697268506,
"learning_rate": 4.734658812980973e-06,
"loss": 1.0508,
"step": 4855
},
{
"epoch": 2.128777923784494,
"grad_norm": 0.2606367742943384,
"learning_rate": 4.7130111248502175e-06,
"loss": 1.008,
"step": 4860
},
{
"epoch": 2.1309680245291283,
"grad_norm": 0.26232587434114535,
"learning_rate": 4.6913977705813296e-06,
"loss": 1.0148,
"step": 4865
},
{
"epoch": 2.133158125273763,
"grad_norm": 0.27500445217893804,
"learning_rate": 4.6698188905320615e-06,
"loss": 1.0271,
"step": 4870
},
{
"epoch": 2.135348226018397,
"grad_norm": 0.26713378585912395,
"learning_rate": 4.648274624836309e-06,
"loss": 1.0221,
"step": 4875
},
{
"epoch": 2.137538326763031,
"grad_norm": 0.2762560729124893,
"learning_rate": 4.626765113403161e-06,
"loss": 1.0234,
"step": 4880
},
{
"epoch": 2.1397284275076656,
"grad_norm": 0.2689363110437547,
"learning_rate": 4.605290495916018e-06,
"loss": 1.0357,
"step": 4885
},
{
"epoch": 2.1419185282522997,
"grad_norm": 0.26265385480996006,
"learning_rate": 4.5838509118316896e-06,
"loss": 1.0123,
"step": 4890
},
{
"epoch": 2.1441086289969338,
"grad_norm": 0.26738349884810464,
"learning_rate": 4.5624465003794575e-06,
"loss": 1.0416,
"step": 4895
},
{
"epoch": 2.1462987297415683,
"grad_norm": 0.2641315930523429,
"learning_rate": 4.54107740056021e-06,
"loss": 1.0109,
"step": 4900
},
{
"epoch": 2.1484888304862024,
"grad_norm": 0.2759532152378268,
"learning_rate": 4.51974375114551e-06,
"loss": 1.035,
"step": 4905
},
{
"epoch": 2.1506789312308365,
"grad_norm": 0.2711223796119766,
"learning_rate": 4.4984456906767e-06,
"loss": 1.0186,
"step": 4910
},
{
"epoch": 2.152869031975471,
"grad_norm": 0.2622897744480132,
"learning_rate": 4.477183357464021e-06,
"loss": 1.0242,
"step": 4915
},
{
"epoch": 2.155059132720105,
"grad_norm": 0.2750413280265702,
"learning_rate": 4.455956889585698e-06,
"loss": 1.0242,
"step": 4920
},
{
"epoch": 2.1572492334647393,
"grad_norm": 0.2756527614489596,
"learning_rate": 4.434766424887024e-06,
"loss": 1.0461,
"step": 4925
},
{
"epoch": 2.159439334209374,
"grad_norm": 0.2701065410839812,
"learning_rate": 4.413612100979516e-06,
"loss": 1.0447,
"step": 4930
},
{
"epoch": 2.161629434954008,
"grad_norm": 0.26414864633307716,
"learning_rate": 4.392494055239971e-06,
"loss": 1.0115,
"step": 4935
},
{
"epoch": 2.163819535698642,
"grad_norm": 0.2597662178542703,
"learning_rate": 4.371412424809607e-06,
"loss": 0.9982,
"step": 4940
},
{
"epoch": 2.1660096364432766,
"grad_norm": 0.265049318136912,
"learning_rate": 4.350367346593151e-06,
"loss": 1.0076,
"step": 4945
},
{
"epoch": 2.1681997371879107,
"grad_norm": 0.26480844249657876,
"learning_rate": 4.3293589572579585e-06,
"loss": 1.0127,
"step": 4950
},
{
"epoch": 2.170389837932545,
"grad_norm": 0.26650586635339474,
"learning_rate": 4.30838739323314e-06,
"loss": 1.0211,
"step": 4955
},
{
"epoch": 2.1725799386771794,
"grad_norm": 0.2615690992438174,
"learning_rate": 4.287452790708641e-06,
"loss": 1.0082,
"step": 4960
},
{
"epoch": 2.1747700394218135,
"grad_norm": 0.26000121930930187,
"learning_rate": 4.266555285634398e-06,
"loss": 1.0309,
"step": 4965
},
{
"epoch": 2.1769601401664476,
"grad_norm": 0.2651977563284172,
"learning_rate": 4.245695013719421e-06,
"loss": 1.01,
"step": 4970
},
{
"epoch": 2.179150240911082,
"grad_norm": 0.26742093385668575,
"learning_rate": 4.224872110430929e-06,
"loss": 1.0119,
"step": 4975
},
{
"epoch": 2.181340341655716,
"grad_norm": 0.2685843051356311,
"learning_rate": 4.204086710993477e-06,
"loss": 0.9998,
"step": 4980
},
{
"epoch": 2.1835304424003503,
"grad_norm": 0.2741238172517032,
"learning_rate": 4.1833389503880574e-06,
"loss": 1.0191,
"step": 4985
},
{
"epoch": 2.185720543144985,
"grad_norm": 0.26609648886123183,
"learning_rate": 4.162628963351245e-06,
"loss": 1.0188,
"step": 4990
},
{
"epoch": 2.187910643889619,
"grad_norm": 0.26640975647603027,
"learning_rate": 4.141956884374304e-06,
"loss": 1.034,
"step": 4995
},
{
"epoch": 2.190100744634253,
"grad_norm": 0.2616553108981328,
"learning_rate": 4.1213228477023235e-06,
"loss": 1.007,
"step": 5000
},
{
"epoch": 2.1922908453788876,
"grad_norm": 0.25930380462776687,
"learning_rate": 4.100726987333353e-06,
"loss": 1.0047,
"step": 5005
},
{
"epoch": 2.1944809461235217,
"grad_norm": 0.26376308398121895,
"learning_rate": 4.080169437017516e-06,
"loss": 1.017,
"step": 5010
},
{
"epoch": 2.196671046868156,
"grad_norm": 0.2708631621391155,
"learning_rate": 4.059650330256145e-06,
"loss": 1.0291,
"step": 5015
},
{
"epoch": 2.1988611476127904,
"grad_norm": 0.26235140363314685,
"learning_rate": 4.039169800300933e-06,
"loss": 1.0307,
"step": 5020
},
{
"epoch": 2.2010512483574245,
"grad_norm": 0.27081362352421917,
"learning_rate": 4.018727980153041e-06,
"loss": 1.0273,
"step": 5025
},
{
"epoch": 2.2032413491020586,
"grad_norm": 0.2680987698793304,
"learning_rate": 3.998325002562258e-06,
"loss": 1.0236,
"step": 5030
},
{
"epoch": 2.205431449846693,
"grad_norm": 0.2723369033529761,
"learning_rate": 3.977961000026119e-06,
"loss": 1.033,
"step": 5035
},
{
"epoch": 2.2076215505913273,
"grad_norm": 0.27329071897001517,
"learning_rate": 3.957636104789056e-06,
"loss": 1.0227,
"step": 5040
},
{
"epoch": 2.2098116513359614,
"grad_norm": 0.266253622886688,
"learning_rate": 3.937350448841544e-06,
"loss": 1.0,
"step": 5045
},
{
"epoch": 2.212001752080596,
"grad_norm": 0.2845347733345193,
"learning_rate": 3.917104163919226e-06,
"loss": 1.0328,
"step": 5050
},
{
"epoch": 2.21419185282523,
"grad_norm": 0.2671087468266786,
"learning_rate": 3.896897381502081e-06,
"loss": 1.0188,
"step": 5055
},
{
"epoch": 2.216381953569864,
"grad_norm": 0.2647712357087782,
"learning_rate": 3.8767302328135456e-06,
"loss": 1.015,
"step": 5060
},
{
"epoch": 2.2185720543144987,
"grad_norm": 0.2647110909009258,
"learning_rate": 3.856602848819677e-06,
"loss": 1.0416,
"step": 5065
},
{
"epoch": 2.2207621550591328,
"grad_norm": 0.26237178804400635,
"learning_rate": 3.836515360228308e-06,
"loss": 1.0424,
"step": 5070
},
{
"epoch": 2.222952255803767,
"grad_norm": 0.2658640348878226,
"learning_rate": 3.816467897488175e-06,
"loss": 1.0449,
"step": 5075
},
{
"epoch": 2.2251423565484014,
"grad_norm": 0.2601983313467295,
"learning_rate": 3.7964605907881014e-06,
"loss": 1.0,
"step": 5080
},
{
"epoch": 2.2273324572930355,
"grad_norm": 0.27724490452236145,
"learning_rate": 3.7764935700561222e-06,
"loss": 1.0033,
"step": 5085
},
{
"epoch": 2.2295225580376696,
"grad_norm": 0.28674009424588764,
"learning_rate": 3.7565669649586577e-06,
"loss": 1.0377,
"step": 5090
},
{
"epoch": 2.231712658782304,
"grad_norm": 0.26041573255312417,
"learning_rate": 3.7366809048996755e-06,
"loss": 1.0219,
"step": 5095
},
{
"epoch": 2.2339027595269383,
"grad_norm": 0.26439782352388286,
"learning_rate": 3.7168355190198336e-06,
"loss": 1.0268,
"step": 5100
},
{
"epoch": 2.2360928602715724,
"grad_norm": 0.2666477044206433,
"learning_rate": 3.6970309361956514e-06,
"loss": 1.0215,
"step": 5105
},
{
"epoch": 2.238282961016207,
"grad_norm": 0.27347497537887744,
"learning_rate": 3.6772672850386803e-06,
"loss": 0.9965,
"step": 5110
},
{
"epoch": 2.240473061760841,
"grad_norm": 0.2733236300572501,
"learning_rate": 3.657544693894648e-06,
"loss": 1.0197,
"step": 5115
},
{
"epoch": 2.242663162505475,
"grad_norm": 0.27287749992945903,
"learning_rate": 3.6378632908426504e-06,
"loss": 1.034,
"step": 5120
},
{
"epoch": 2.2448532632501097,
"grad_norm": 0.2585480868756628,
"learning_rate": 3.6182232036942967e-06,
"loss": 1.0184,
"step": 5125
},
{
"epoch": 2.247043363994744,
"grad_norm": 0.26361194041457353,
"learning_rate": 3.5986245599928903e-06,
"loss": 1.0135,
"step": 5130
},
{
"epoch": 2.249233464739378,
"grad_norm": 0.2662253344158049,
"learning_rate": 3.5790674870126073e-06,
"loss": 1.0063,
"step": 5135
},
{
"epoch": 2.2514235654840125,
"grad_norm": 0.26616630879092884,
"learning_rate": 3.5595521117576513e-06,
"loss": 1.0043,
"step": 5140
},
{
"epoch": 2.2536136662286466,
"grad_norm": 0.2736827752131189,
"learning_rate": 3.540078560961452e-06,
"loss": 1.0125,
"step": 5145
},
{
"epoch": 2.2558037669732807,
"grad_norm": 0.2710314880260435,
"learning_rate": 3.5206469610858207e-06,
"loss": 1.0432,
"step": 5150
},
{
"epoch": 2.257993867717915,
"grad_norm": 0.268667093843008,
"learning_rate": 3.501257438320136e-06,
"loss": 1.0404,
"step": 5155
},
{
"epoch": 2.2601839684625493,
"grad_norm": 0.2723797726229181,
"learning_rate": 3.4819101185805383e-06,
"loss": 1.0123,
"step": 5160
},
{
"epoch": 2.2623740692071834,
"grad_norm": 0.2603114146629379,
"learning_rate": 3.4626051275090865e-06,
"loss": 0.9975,
"step": 5165
},
{
"epoch": 2.264564169951818,
"grad_norm": 0.27790374707917465,
"learning_rate": 3.44334259047297e-06,
"loss": 1.0217,
"step": 5170
},
{
"epoch": 2.266754270696452,
"grad_norm": 0.25899914233411553,
"learning_rate": 3.424122632563669e-06,
"loss": 1.0504,
"step": 5175
},
{
"epoch": 2.268944371441086,
"grad_norm": 0.26446806135188367,
"learning_rate": 3.404945378596155e-06,
"loss": 1.0127,
"step": 5180
},
{
"epoch": 2.2711344721857207,
"grad_norm": 0.2612602747990249,
"learning_rate": 3.3858109531080887e-06,
"loss": 1.0156,
"step": 5185
},
{
"epoch": 2.273324572930355,
"grad_norm": 0.26445200917852263,
"learning_rate": 3.3667194803589885e-06,
"loss": 1.0119,
"step": 5190
},
{
"epoch": 2.275514673674989,
"grad_norm": 0.28094575273309763,
"learning_rate": 3.3476710843294415e-06,
"loss": 1.0281,
"step": 5195
},
{
"epoch": 2.2777047744196235,
"grad_norm": 0.26736737190018384,
"learning_rate": 3.3286658887202994e-06,
"loss": 1.0037,
"step": 5200
},
{
"epoch": 2.2798948751642576,
"grad_norm": 0.2640795352734308,
"learning_rate": 3.3097040169518567e-06,
"loss": 1.0406,
"step": 5205
},
{
"epoch": 2.2820849759088917,
"grad_norm": 0.26426495072100187,
"learning_rate": 3.2907855921630747e-06,
"loss": 1.033,
"step": 5210
},
{
"epoch": 2.2842750766535262,
"grad_norm": 0.26566279121236197,
"learning_rate": 3.2719107372107615e-06,
"loss": 1.0197,
"step": 5215
},
{
"epoch": 2.2864651773981604,
"grad_norm": 0.2653552548457873,
"learning_rate": 3.253079574668777e-06,
"loss": 1.0006,
"step": 5220
},
{
"epoch": 2.2886552781427945,
"grad_norm": 0.26398157254205407,
"learning_rate": 3.234292226827255e-06,
"loss": 0.9961,
"step": 5225
},
{
"epoch": 2.290845378887429,
"grad_norm": 0.2768441293910038,
"learning_rate": 3.215548815691779e-06,
"loss": 1.0348,
"step": 5230
},
{
"epoch": 2.293035479632063,
"grad_norm": 0.26131639882439045,
"learning_rate": 3.196849462982622e-06,
"loss": 1.0018,
"step": 5235
},
{
"epoch": 2.295225580376697,
"grad_norm": 0.2722287901413304,
"learning_rate": 3.178194290133929e-06,
"loss": 1.0232,
"step": 5240
},
{
"epoch": 2.2974156811213318,
"grad_norm": 0.2761279729955395,
"learning_rate": 3.1595834182929396e-06,
"loss": 1.0309,
"step": 5245
},
{
"epoch": 2.299605781865966,
"grad_norm": 0.264451768689309,
"learning_rate": 3.1410169683192114e-06,
"loss": 1.0291,
"step": 5250
},
{
"epoch": 2.3017958826106,
"grad_norm": 0.2720396323147097,
"learning_rate": 3.1224950607838147e-06,
"loss": 1.0318,
"step": 5255
},
{
"epoch": 2.3039859833552345,
"grad_norm": 0.2659690024816963,
"learning_rate": 3.1040178159685675e-06,
"loss": 1.0279,
"step": 5260
},
{
"epoch": 2.3061760840998686,
"grad_norm": 0.26752050191400945,
"learning_rate": 3.08558535386524e-06,
"loss": 1.0295,
"step": 5265
},
{
"epoch": 2.3083661848445027,
"grad_norm": 0.26719252797672,
"learning_rate": 3.0671977941747843e-06,
"loss": 1.0074,
"step": 5270
},
{
"epoch": 2.3105562855891373,
"grad_norm": 0.27119586183987965,
"learning_rate": 3.048855256306562e-06,
"loss": 0.9959,
"step": 5275
},
{
"epoch": 2.3127463863337714,
"grad_norm": 0.2599492206066672,
"learning_rate": 3.030557859377551e-06,
"loss": 1.027,
"step": 5280
},
{
"epoch": 2.3149364870784055,
"grad_norm": 0.2622696675757602,
"learning_rate": 3.0123057222115835e-06,
"loss": 1.015,
"step": 5285
},
{
"epoch": 2.31712658782304,
"grad_norm": 0.26724171900385185,
"learning_rate": 2.9940989633385866e-06,
"loss": 1.0105,
"step": 5290
},
{
"epoch": 2.319316688567674,
"grad_norm": 0.26121450618999226,
"learning_rate": 2.9759377009937807e-06,
"loss": 1.017,
"step": 5295
},
{
"epoch": 2.3215067893123082,
"grad_norm": 0.2705484002824616,
"learning_rate": 2.957822053116948e-06,
"loss": 1.0334,
"step": 5300
},
{
"epoch": 2.323696890056943,
"grad_norm": 0.27808085682505257,
"learning_rate": 2.9397521373516357e-06,
"loss": 1.0045,
"step": 5305
},
{
"epoch": 2.325886990801577,
"grad_norm": 0.26449757221896497,
"learning_rate": 2.9217280710444084e-06,
"loss": 1.0076,
"step": 5310
},
{
"epoch": 2.328077091546211,
"grad_norm": 0.2681810130131999,
"learning_rate": 2.9037499712440907e-06,
"loss": 1.0102,
"step": 5315
},
{
"epoch": 2.3302671922908456,
"grad_norm": 0.267694962057186,
"learning_rate": 2.885817954700988e-06,
"loss": 1.0125,
"step": 5320
},
{
"epoch": 2.3324572930354797,
"grad_norm": 0.27046123405151434,
"learning_rate": 2.8679321378661506e-06,
"loss": 1.0197,
"step": 5325
},
{
"epoch": 2.3346473937801138,
"grad_norm": 0.2602187883267958,
"learning_rate": 2.850092636890599e-06,
"loss": 1.0006,
"step": 5330
},
{
"epoch": 2.3368374945247483,
"grad_norm": 0.2697969479776123,
"learning_rate": 2.8322995676245756e-06,
"loss": 1.0207,
"step": 5335
},
{
"epoch": 2.3390275952693824,
"grad_norm": 0.26593532687029453,
"learning_rate": 2.8145530456168034e-06,
"loss": 1.0064,
"step": 5340
},
{
"epoch": 2.3412176960140165,
"grad_norm": 0.260443619701446,
"learning_rate": 2.796853186113715e-06,
"loss": 1.0381,
"step": 5345
},
{
"epoch": 2.343407796758651,
"grad_norm": 0.2656577780398318,
"learning_rate": 2.7792001040587256e-06,
"loss": 1.0186,
"step": 5350
},
{
"epoch": 2.345597897503285,
"grad_norm": 0.26695059912132757,
"learning_rate": 2.7615939140914704e-06,
"loss": 0.9973,
"step": 5355
},
{
"epoch": 2.3477879982479193,
"grad_norm": 0.2698545962387733,
"learning_rate": 2.7440347305470604e-06,
"loss": 1.0262,
"step": 5360
},
{
"epoch": 2.349978098992554,
"grad_norm": 0.2621284105960352,
"learning_rate": 2.7265226674553613e-06,
"loss": 1.0186,
"step": 5365
},
{
"epoch": 2.352168199737188,
"grad_norm": 0.2629369792774262,
"learning_rate": 2.7090578385402243e-06,
"loss": 0.9812,
"step": 5370
},
{
"epoch": 2.354358300481822,
"grad_norm": 0.26223979769748595,
"learning_rate": 2.691640357218759e-06,
"loss": 1.0314,
"step": 5375
},
{
"epoch": 2.3565484012264566,
"grad_norm": 0.2638015005792946,
"learning_rate": 2.674270336600612e-06,
"loss": 1.024,
"step": 5380
},
{
"epoch": 2.3587385019710907,
"grad_norm": 0.2643561856353618,
"learning_rate": 2.6569478894872014e-06,
"loss": 1.0102,
"step": 5385
},
{
"epoch": 2.360928602715725,
"grad_norm": 0.27732040588272344,
"learning_rate": 2.6396731283710185e-06,
"loss": 1.0422,
"step": 5390
},
{
"epoch": 2.3631187034603593,
"grad_norm": 0.2644682180760317,
"learning_rate": 2.6224461654348686e-06,
"loss": 1.0139,
"step": 5395
},
{
"epoch": 2.3653088042049935,
"grad_norm": 0.26210054764945073,
"learning_rate": 2.605267112551154e-06,
"loss": 1.0309,
"step": 5400
},
{
"epoch": 2.3674989049496276,
"grad_norm": 0.2709828298883479,
"learning_rate": 2.5881360812811584e-06,
"loss": 1.0332,
"step": 5405
},
{
"epoch": 2.369689005694262,
"grad_norm": 0.2730880205333291,
"learning_rate": 2.571053182874298e-06,
"loss": 1.0281,
"step": 5410
},
{
"epoch": 2.371879106438896,
"grad_norm": 0.2596551205971999,
"learning_rate": 2.5540185282674247e-06,
"loss": 1.0342,
"step": 5415
},
{
"epoch": 2.3740692071835303,
"grad_norm": 0.2695558996090938,
"learning_rate": 2.5370322280840863e-06,
"loss": 0.9879,
"step": 5420
},
{
"epoch": 2.376259307928165,
"grad_norm": 0.26055214066353855,
"learning_rate": 2.520094392633815e-06,
"loss": 1.0318,
"step": 5425
},
{
"epoch": 2.378449408672799,
"grad_norm": 0.26376924748943736,
"learning_rate": 2.5032051319114203e-06,
"loss": 0.984,
"step": 5430
},
{
"epoch": 2.380639509417433,
"grad_norm": 0.2659522914893529,
"learning_rate": 2.4863645555962535e-06,
"loss": 1.0074,
"step": 5435
},
{
"epoch": 2.3828296101620676,
"grad_norm": 0.26652092795558513,
"learning_rate": 2.469572773051524e-06,
"loss": 1.0002,
"step": 5440
},
{
"epoch": 2.3850197109067017,
"grad_norm": 0.2721077024859027,
"learning_rate": 2.4528298933235504e-06,
"loss": 1.0012,
"step": 5445
},
{
"epoch": 2.387209811651336,
"grad_norm": 0.2714038096859095,
"learning_rate": 2.4361360251410958e-06,
"loss": 1.0395,
"step": 5450
},
{
"epoch": 2.3893999123959704,
"grad_norm": 0.259937301986236,
"learning_rate": 2.419491276914635e-06,
"loss": 0.9941,
"step": 5455
},
{
"epoch": 2.3915900131406045,
"grad_norm": 0.2693725837181157,
"learning_rate": 2.4028957567356504e-06,
"loss": 1.0162,
"step": 5460
},
{
"epoch": 2.3937801138852386,
"grad_norm": 0.2614519634595169,
"learning_rate": 2.386349572375939e-06,
"loss": 1.0273,
"step": 5465
},
{
"epoch": 2.395970214629873,
"grad_norm": 0.26516778440783295,
"learning_rate": 2.3698528312869164e-06,
"loss": 1.0143,
"step": 5470
},
{
"epoch": 2.3981603153745072,
"grad_norm": 0.2617725830901367,
"learning_rate": 2.3534056405988992e-06,
"loss": 1.016,
"step": 5475
},
{
"epoch": 2.4003504161191414,
"grad_norm": 0.26739935073726256,
"learning_rate": 2.3370081071204355e-06,
"loss": 1.0293,
"step": 5480
},
{
"epoch": 2.402540516863776,
"grad_norm": 0.26968306076232107,
"learning_rate": 2.3206603373375893e-06,
"loss": 1.002,
"step": 5485
},
{
"epoch": 2.40473061760841,
"grad_norm": 0.27295936217638855,
"learning_rate": 2.304362437413258e-06,
"loss": 1.0021,
"step": 5490
},
{
"epoch": 2.406920718353044,
"grad_norm": 0.2708387281457088,
"learning_rate": 2.28811451318649e-06,
"loss": 1.0225,
"step": 5495
},
{
"epoch": 2.4091108190976787,
"grad_norm": 0.26354121947186615,
"learning_rate": 2.2719166701717786e-06,
"loss": 1.0031,
"step": 5500
},
{
"epoch": 2.4113009198423128,
"grad_norm": 0.2637002110660547,
"learning_rate": 2.2557690135584008e-06,
"loss": 1.0184,
"step": 5505
},
{
"epoch": 2.413491020586947,
"grad_norm": 0.2698479127999328,
"learning_rate": 2.2396716482097135e-06,
"loss": 1.0189,
"step": 5510
},
{
"epoch": 2.4156811213315814,
"grad_norm": 0.26409864234783814,
"learning_rate": 2.2236246786624794e-06,
"loss": 1.0211,
"step": 5515
},
{
"epoch": 2.4178712220762155,
"grad_norm": 0.27268499867925555,
"learning_rate": 2.207628209126198e-06,
"loss": 1.0412,
"step": 5520
},
{
"epoch": 2.4200613228208496,
"grad_norm": 0.2622606967406724,
"learning_rate": 2.1916823434824087e-06,
"loss": 1.0209,
"step": 5525
},
{
"epoch": 2.422251423565484,
"grad_norm": 0.27635573904338784,
"learning_rate": 2.175787185284042e-06,
"loss": 1.0305,
"step": 5530
},
{
"epoch": 2.4244415243101183,
"grad_norm": 0.25710396782274697,
"learning_rate": 2.1599428377547137e-06,
"loss": 1.0164,
"step": 5535
},
{
"epoch": 2.4266316250547524,
"grad_norm": 0.2633283094129851,
"learning_rate": 2.144149403788092e-06,
"loss": 1.0199,
"step": 5540
},
{
"epoch": 2.428821725799387,
"grad_norm": 0.26415517310658565,
"learning_rate": 2.1284069859472035e-06,
"loss": 1.0223,
"step": 5545
},
{
"epoch": 2.431011826544021,
"grad_norm": 0.265930449211507,
"learning_rate": 2.112715686463772e-06,
"loss": 1.0098,
"step": 5550
},
{
"epoch": 2.433201927288655,
"grad_norm": 0.25723984456483956,
"learning_rate": 2.097075607237559e-06,
"loss": 1.0248,
"step": 5555
},
{
"epoch": 2.4353920280332897,
"grad_norm": 0.2598804894344425,
"learning_rate": 2.0814868498357042e-06,
"loss": 1.0082,
"step": 5560
},
{
"epoch": 2.437582128777924,
"grad_norm": 0.2697688309874498,
"learning_rate": 2.065949515492054e-06,
"loss": 1.0238,
"step": 5565
},
{
"epoch": 2.439772229522558,
"grad_norm": 0.27620146863759415,
"learning_rate": 2.0504637051065212e-06,
"loss": 1.0223,
"step": 5570
},
{
"epoch": 2.4419623302671924,
"grad_norm": 0.26231043488916006,
"learning_rate": 2.0350295192444135e-06,
"loss": 1.0211,
"step": 5575
},
{
"epoch": 2.4441524310118266,
"grad_norm": 0.2634847594383747,
"learning_rate": 2.0196470581357865e-06,
"loss": 1.0063,
"step": 5580
},
{
"epoch": 2.4463425317564607,
"grad_norm": 0.2585906098421852,
"learning_rate": 2.0043164216748034e-06,
"loss": 1.0312,
"step": 5585
},
{
"epoch": 2.448532632501095,
"grad_norm": 0.2599866945351205,
"learning_rate": 1.9890377094190626e-06,
"loss": 1.0141,
"step": 5590
},
{
"epoch": 2.4507227332457293,
"grad_norm": 0.26235044645446814,
"learning_rate": 1.9738110205889803e-06,
"loss": 1.0279,
"step": 5595
},
{
"epoch": 2.4529128339903634,
"grad_norm": 0.26350829783533997,
"learning_rate": 1.958636454067119e-06,
"loss": 1.0377,
"step": 5600
},
{
"epoch": 2.455102934734998,
"grad_norm": 0.2697500929246074,
"learning_rate": 1.9435141083975596e-06,
"loss": 0.9988,
"step": 5605
},
{
"epoch": 2.457293035479632,
"grad_norm": 0.25991940084532544,
"learning_rate": 1.9284440817852666e-06,
"loss": 0.9961,
"step": 5610
},
{
"epoch": 2.459483136224266,
"grad_norm": 0.2609678334326761,
"learning_rate": 1.9134264720954353e-06,
"loss": 1.0229,
"step": 5615
},
{
"epoch": 2.4616732369689007,
"grad_norm": 0.2563823694294751,
"learning_rate": 1.8984613768528658e-06,
"loss": 1.0031,
"step": 5620
},
{
"epoch": 2.463863337713535,
"grad_norm": 0.26389657889097734,
"learning_rate": 1.8835488932413272e-06,
"loss": 0.9926,
"step": 5625
},
{
"epoch": 2.466053438458169,
"grad_norm": 0.2679236979583576,
"learning_rate": 1.868689118102931e-06,
"loss": 1.0238,
"step": 5630
},
{
"epoch": 2.4682435392028035,
"grad_norm": 0.2715746821595108,
"learning_rate": 1.8538821479374968e-06,
"loss": 1.024,
"step": 5635
},
{
"epoch": 2.4704336399474376,
"grad_norm": 0.26957317386333224,
"learning_rate": 1.839128078901924e-06,
"loss": 1.0096,
"step": 5640
},
{
"epoch": 2.4726237406920717,
"grad_norm": 0.2622405945557025,
"learning_rate": 1.8244270068095727e-06,
"loss": 1.0381,
"step": 5645
},
{
"epoch": 2.4748138414367062,
"grad_norm": 0.2740756090414041,
"learning_rate": 1.8097790271296378e-06,
"loss": 1.0223,
"step": 5650
},
{
"epoch": 2.4770039421813403,
"grad_norm": 0.2661221374455693,
"learning_rate": 1.7951842349865323e-06,
"loss": 1.025,
"step": 5655
},
{
"epoch": 2.4791940429259745,
"grad_norm": 0.25667749629367814,
"learning_rate": 1.7806427251592717e-06,
"loss": 0.9939,
"step": 5660
},
{
"epoch": 2.481384143670609,
"grad_norm": 0.2647393667585526,
"learning_rate": 1.7661545920808465e-06,
"loss": 1.0166,
"step": 5665
},
{
"epoch": 2.483574244415243,
"grad_norm": 0.2571248265493812,
"learning_rate": 1.7517199298376187e-06,
"loss": 1.0135,
"step": 5670
},
{
"epoch": 2.485764345159877,
"grad_norm": 0.26602653787423514,
"learning_rate": 1.7373388321687145e-06,
"loss": 1.0326,
"step": 5675
},
{
"epoch": 2.4879544459045118,
"grad_norm": 0.25813672302306245,
"learning_rate": 1.7230113924654047e-06,
"loss": 1.0357,
"step": 5680
},
{
"epoch": 2.490144546649146,
"grad_norm": 0.2579671571163789,
"learning_rate": 1.7087377037705121e-06,
"loss": 1.0311,
"step": 5685
},
{
"epoch": 2.49233464739378,
"grad_norm": 0.2563802679936426,
"learning_rate": 1.694517858777781e-06,
"loss": 1.0252,
"step": 5690
},
{
"epoch": 2.4945247481384145,
"grad_norm": 0.2662164827240698,
"learning_rate": 1.6803519498313126e-06,
"loss": 1.0135,
"step": 5695
},
{
"epoch": 2.4967148488830486,
"grad_norm": 0.26122565812958937,
"learning_rate": 1.666240068924937e-06,
"loss": 1.0414,
"step": 5700
},
{
"epoch": 2.4989049496276827,
"grad_norm": 0.2605998958832977,
"learning_rate": 1.6521823077016253e-06,
"loss": 1.0152,
"step": 5705
},
{
"epoch": 2.5010950503723173,
"grad_norm": 0.26609122798906276,
"learning_rate": 1.638178757452894e-06,
"loss": 1.0201,
"step": 5710
},
{
"epoch": 2.5032851511169514,
"grad_norm": 0.26241620338438515,
"learning_rate": 1.6242295091182082e-06,
"loss": 0.9906,
"step": 5715
},
{
"epoch": 2.5054752518615855,
"grad_norm": 0.26393016094462074,
"learning_rate": 1.6103346532844011e-06,
"loss": 1.0312,
"step": 5720
},
{
"epoch": 2.50766535260622,
"grad_norm": 0.256500142243546,
"learning_rate": 1.5964942801850802e-06,
"loss": 1.0223,
"step": 5725
},
{
"epoch": 2.509855453350854,
"grad_norm": 0.26878064802677665,
"learning_rate": 1.5827084797000336e-06,
"loss": 1.0113,
"step": 5730
},
{
"epoch": 2.5120455540954882,
"grad_norm": 0.2707677013243957,
"learning_rate": 1.5689773413546571e-06,
"loss": 0.9873,
"step": 5735
},
{
"epoch": 2.514235654840123,
"grad_norm": 0.26025385637699205,
"learning_rate": 1.5553009543193677e-06,
"loss": 1.002,
"step": 5740
},
{
"epoch": 2.516425755584757,
"grad_norm": 0.26424240891929196,
"learning_rate": 1.5416794074090258e-06,
"loss": 1.0338,
"step": 5745
},
{
"epoch": 2.518615856329391,
"grad_norm": 0.2715208586256904,
"learning_rate": 1.5281127890823633e-06,
"loss": 1.0029,
"step": 5750
},
{
"epoch": 2.5208059570740255,
"grad_norm": 0.26863340452995355,
"learning_rate": 1.514601187441399e-06,
"loss": 1.0311,
"step": 5755
},
{
"epoch": 2.5229960578186597,
"grad_norm": 0.26339075175746157,
"learning_rate": 1.5011446902308714e-06,
"loss": 1.0016,
"step": 5760
},
{
"epoch": 2.5251861585632938,
"grad_norm": 0.2780659159629088,
"learning_rate": 1.4877433848376688e-06,
"loss": 1.0357,
"step": 5765
},
{
"epoch": 2.5273762593079283,
"grad_norm": 0.2824820354932007,
"learning_rate": 1.4743973582902659e-06,
"loss": 1.0451,
"step": 5770
},
{
"epoch": 2.5295663600525624,
"grad_norm": 0.26541836104789196,
"learning_rate": 1.4611066972581588e-06,
"loss": 1.0393,
"step": 5775
},
{
"epoch": 2.5317564607971965,
"grad_norm": 0.26392038520386774,
"learning_rate": 1.4478714880512824e-06,
"loss": 1.0098,
"step": 5780
},
{
"epoch": 2.533946561541831,
"grad_norm": 0.2670274065680639,
"learning_rate": 1.4346918166194811e-06,
"loss": 1.0318,
"step": 5785
},
{
"epoch": 2.536136662286465,
"grad_norm": 0.26637327926132287,
"learning_rate": 1.4215677685519335e-06,
"loss": 1.0457,
"step": 5790
},
{
"epoch": 2.5383267630310993,
"grad_norm": 0.2718315157534932,
"learning_rate": 1.4084994290765918e-06,
"loss": 1.0229,
"step": 5795
},
{
"epoch": 2.540516863775734,
"grad_norm": 0.265091011136329,
"learning_rate": 1.3954868830596391e-06,
"loss": 1.0135,
"step": 5800
},
{
"epoch": 2.542706964520368,
"grad_norm": 0.27179242757473226,
"learning_rate": 1.3825302150049313e-06,
"loss": 1.0445,
"step": 5805
},
{
"epoch": 2.544897065265002,
"grad_norm": 0.2634719242658282,
"learning_rate": 1.369629509053455e-06,
"loss": 1.0252,
"step": 5810
},
{
"epoch": 2.5470871660096366,
"grad_norm": 0.26428182587169274,
"learning_rate": 1.3567848489827773e-06,
"loss": 1.0268,
"step": 5815
},
{
"epoch": 2.5492772667542707,
"grad_norm": 0.27171679493352563,
"learning_rate": 1.3439963182064996e-06,
"loss": 1.0414,
"step": 5820
},
{
"epoch": 2.551467367498905,
"grad_norm": 0.26868016597068656,
"learning_rate": 1.3312639997737166e-06,
"loss": 1.0059,
"step": 5825
},
{
"epoch": 2.5536574682435393,
"grad_norm": 0.26796885929306313,
"learning_rate": 1.3185879763684784e-06,
"loss": 1.0229,
"step": 5830
},
{
"epoch": 2.5558475689881734,
"grad_norm": 0.2747441448985468,
"learning_rate": 1.3059683303092574e-06,
"loss": 1.0383,
"step": 5835
},
{
"epoch": 2.5580376697328076,
"grad_norm": 0.2656630564773151,
"learning_rate": 1.2934051435484107e-06,
"loss": 1.017,
"step": 5840
},
{
"epoch": 2.560227770477442,
"grad_norm": 0.2714058989519959,
"learning_rate": 1.2808984976716421e-06,
"loss": 1.0297,
"step": 5845
},
{
"epoch": 2.562417871222076,
"grad_norm": 0.2672456887749402,
"learning_rate": 1.2684484738974778e-06,
"loss": 1.0324,
"step": 5850
},
{
"epoch": 2.5646079719667103,
"grad_norm": 0.26136420409749284,
"learning_rate": 1.2560551530767384e-06,
"loss": 1.0004,
"step": 5855
},
{
"epoch": 2.566798072711345,
"grad_norm": 0.2605185305648504,
"learning_rate": 1.2437186156920167e-06,
"loss": 1.0158,
"step": 5860
},
{
"epoch": 2.568988173455979,
"grad_norm": 0.2617789541796642,
"learning_rate": 1.231438941857156e-06,
"loss": 1.0293,
"step": 5865
},
{
"epoch": 2.571178274200613,
"grad_norm": 0.26468651594530146,
"learning_rate": 1.2192162113167117e-06,
"loss": 1.0162,
"step": 5870
},
{
"epoch": 2.5733683749452476,
"grad_norm": 0.26905249931056374,
"learning_rate": 1.207050503445464e-06,
"loss": 1.007,
"step": 5875
},
{
"epoch": 2.5755584756898817,
"grad_norm": 0.2560987056326573,
"learning_rate": 1.1949418972478743e-06,
"loss": 0.9949,
"step": 5880
},
{
"epoch": 2.577748576434516,
"grad_norm": 0.2659226997423034,
"learning_rate": 1.1828904713575917e-06,
"loss": 1.0291,
"step": 5885
},
{
"epoch": 2.5799386771791504,
"grad_norm": 0.252859900042723,
"learning_rate": 1.1708963040369315e-06,
"loss": 0.9893,
"step": 5890
},
{
"epoch": 2.5821287779237845,
"grad_norm": 0.2626662868814629,
"learning_rate": 1.1589594731763676e-06,
"loss": 1.0025,
"step": 5895
},
{
"epoch": 2.5843188786684186,
"grad_norm": 0.26008729946148035,
"learning_rate": 1.1470800562940355e-06,
"loss": 1.0303,
"step": 5900
},
{
"epoch": 2.586508979413053,
"grad_norm": 0.26338286175111314,
"learning_rate": 1.1352581305352184e-06,
"loss": 1.0236,
"step": 5905
},
{
"epoch": 2.5886990801576872,
"grad_norm": 0.2730522605992461,
"learning_rate": 1.1234937726718508e-06,
"loss": 1.0174,
"step": 5910
},
{
"epoch": 2.5908891809023213,
"grad_norm": 0.26606670762402423,
"learning_rate": 1.1117870591020207e-06,
"loss": 1.0275,
"step": 5915
},
{
"epoch": 2.593079281646956,
"grad_norm": 0.2635576121376948,
"learning_rate": 1.1001380658494688e-06,
"loss": 1.0195,
"step": 5920
},
{
"epoch": 2.59526938239159,
"grad_norm": 0.2632812590069056,
"learning_rate": 1.088546868563104e-06,
"loss": 1.0059,
"step": 5925
},
{
"epoch": 2.597459483136224,
"grad_norm": 0.26590583110590243,
"learning_rate": 1.077013542516504e-06,
"loss": 1.0133,
"step": 5930
},
{
"epoch": 2.5996495838808586,
"grad_norm": 0.2592095997153406,
"learning_rate": 1.065538162607429e-06,
"loss": 1.0439,
"step": 5935
},
{
"epoch": 2.6018396846254928,
"grad_norm": 0.26304009255367145,
"learning_rate": 1.0541208033573325e-06,
"loss": 1.0258,
"step": 5940
},
{
"epoch": 2.604029785370127,
"grad_norm": 0.26783388307534434,
"learning_rate": 1.0427615389108836e-06,
"loss": 1.0203,
"step": 5945
},
{
"epoch": 2.6062198861147614,
"grad_norm": 0.25445386428299266,
"learning_rate": 1.0314604430354803e-06,
"loss": 1.0096,
"step": 5950
},
{
"epoch": 2.6084099868593955,
"grad_norm": 0.25432480950933944,
"learning_rate": 1.0202175891207777e-06,
"loss": 1.0211,
"step": 5955
},
{
"epoch": 2.6106000876040296,
"grad_norm": 0.2599533384422445,
"learning_rate": 1.0090330501781942e-06,
"loss": 1.0381,
"step": 5960
},
{
"epoch": 2.612790188348664,
"grad_norm": 0.2687854621922093,
"learning_rate": 9.979068988404607e-07,
"loss": 1.0324,
"step": 5965
},
{
"epoch": 2.6149802890932983,
"grad_norm": 0.2747054491371782,
"learning_rate": 9.868392073611277e-07,
"loss": 1.016,
"step": 5970
},
{
"epoch": 2.6171703898379324,
"grad_norm": 0.2605323280615755,
"learning_rate": 9.75830047614117e-07,
"loss": 1.0221,
"step": 5975
},
{
"epoch": 2.619360490582567,
"grad_norm": 0.26132755345912295,
"learning_rate": 9.648794910932334e-07,
"loss": 1.0146,
"step": 5980
},
{
"epoch": 2.621550591327201,
"grad_norm": 0.26679877508377275,
"learning_rate": 9.539876089117128e-07,
"loss": 1.0129,
"step": 5985
},
{
"epoch": 2.623740692071835,
"grad_norm": 0.25873816940118294,
"learning_rate": 9.431544718017615e-07,
"loss": 1.0215,
"step": 5990
},
{
"epoch": 2.6259307928164697,
"grad_norm": 0.26420753117085755,
"learning_rate": 9.323801501140927e-07,
"loss": 1.0291,
"step": 5995
},
{
"epoch": 2.628120893561104,
"grad_norm": 0.2673456929005932,
"learning_rate": 9.216647138174695e-07,
"loss": 1.0328,
"step": 6000
},
{
"epoch": 2.630310994305738,
"grad_norm": 0.27110840415327153,
"learning_rate": 9.110082324982506e-07,
"loss": 1.0119,
"step": 6005
},
{
"epoch": 2.6325010950503724,
"grad_norm": 0.25958577599037735,
"learning_rate": 9.004107753599389e-07,
"loss": 0.9982,
"step": 6010
},
{
"epoch": 2.6346911957950065,
"grad_norm": 0.26952299942882346,
"learning_rate": 8.898724112227353e-07,
"loss": 0.9979,
"step": 6015
},
{
"epoch": 2.6368812965396407,
"grad_norm": 0.25890135882874327,
"learning_rate": 8.793932085230894e-07,
"loss": 1.001,
"step": 6020
},
{
"epoch": 2.639071397284275,
"grad_norm": 0.26093797224431725,
"learning_rate": 8.689732353132518e-07,
"loss": 1.0008,
"step": 6025
},
{
"epoch": 2.6412614980289093,
"grad_norm": 0.2630437536693983,
"learning_rate": 8.586125592608363e-07,
"loss": 0.9936,
"step": 6030
},
{
"epoch": 2.6434515987735434,
"grad_norm": 0.2781661808204497,
"learning_rate": 8.483112476483757e-07,
"loss": 1.0029,
"step": 6035
},
{
"epoch": 2.645641699518178,
"grad_norm": 0.266907247409542,
"learning_rate": 8.380693673728957e-07,
"loss": 1.0256,
"step": 6040
},
{
"epoch": 2.647831800262812,
"grad_norm": 0.27026510023326716,
"learning_rate": 8.278869849454718e-07,
"loss": 1.0441,
"step": 6045
},
{
"epoch": 2.650021901007446,
"grad_norm": 0.2715664818445047,
"learning_rate": 8.177641664907887e-07,
"loss": 1.0318,
"step": 6050
},
{
"epoch": 2.6522120017520807,
"grad_norm": 0.262452995725068,
"learning_rate": 8.077009777467337e-07,
"loss": 1.0148,
"step": 6055
},
{
"epoch": 2.654402102496715,
"grad_norm": 0.26944180313288235,
"learning_rate": 7.976974840639484e-07,
"loss": 1.0268,
"step": 6060
},
{
"epoch": 2.656592203241349,
"grad_norm": 0.26970515074858986,
"learning_rate": 7.8775375040542e-07,
"loss": 1.035,
"step": 6065
},
{
"epoch": 2.6587823039859835,
"grad_norm": 0.26603146543022527,
"learning_rate": 7.778698413460462e-07,
"loss": 1.0385,
"step": 6070
},
{
"epoch": 2.6609724047306176,
"grad_norm": 0.26268076522726097,
"learning_rate": 7.68045821072223e-07,
"loss": 1.0205,
"step": 6075
},
{
"epoch": 2.6631625054752517,
"grad_norm": 0.26151474090954424,
"learning_rate": 7.582817533814313e-07,
"loss": 1.0076,
"step": 6080
},
{
"epoch": 2.6653526062198862,
"grad_norm": 0.26380289604041596,
"learning_rate": 7.485777016818119e-07,
"loss": 1.0047,
"step": 6085
},
{
"epoch": 2.6675427069645203,
"grad_norm": 0.2640739120564622,
"learning_rate": 7.389337289917653e-07,
"loss": 1.0238,
"step": 6090
},
{
"epoch": 2.6697328077091544,
"grad_norm": 0.25824476005629765,
"learning_rate": 7.293498979395353e-07,
"loss": 1.033,
"step": 6095
},
{
"epoch": 2.671922908453789,
"grad_norm": 0.2761915579791404,
"learning_rate": 7.198262707628e-07,
"loss": 1.0457,
"step": 6100
},
{
"epoch": 2.674113009198423,
"grad_norm": 0.2591538186339242,
"learning_rate": 7.103629093082776e-07,
"loss": 1.0104,
"step": 6105
},
{
"epoch": 2.676303109943057,
"grad_norm": 0.2667664421930206,
"learning_rate": 7.009598750313195e-07,
"loss": 1.0127,
"step": 6110
},
{
"epoch": 2.6784932106876918,
"grad_norm": 0.2718642700436106,
"learning_rate": 6.916172289955059e-07,
"loss": 1.0279,
"step": 6115
},
{
"epoch": 2.680683311432326,
"grad_norm": 0.268922489593917,
"learning_rate": 6.823350318722555e-07,
"loss": 1.0236,
"step": 6120
},
{
"epoch": 2.68287341217696,
"grad_norm": 0.26563093628712836,
"learning_rate": 6.731133439404291e-07,
"loss": 1.0229,
"step": 6125
},
{
"epoch": 2.6850635129215945,
"grad_norm": 0.27212083901058975,
"learning_rate": 6.63952225085942e-07,
"loss": 1.0072,
"step": 6130
},
{
"epoch": 2.6872536136662286,
"grad_norm": 0.27909050468631363,
"learning_rate": 6.548517348013739e-07,
"loss": 1.0186,
"step": 6135
},
{
"epoch": 2.6894437144108627,
"grad_norm": 0.2634021147116165,
"learning_rate": 6.458119321855705e-07,
"loss": 1.0383,
"step": 6140
},
{
"epoch": 2.6916338151554973,
"grad_norm": 0.2604570689525165,
"learning_rate": 6.368328759432829e-07,
"loss": 1.0018,
"step": 6145
},
{
"epoch": 2.6938239159001314,
"grad_norm": 0.2651829616366712,
"learning_rate": 6.279146243847622e-07,
"loss": 1.0262,
"step": 6150
},
{
"epoch": 2.6960140166447655,
"grad_norm": 0.2626234864167782,
"learning_rate": 6.190572354254031e-07,
"loss": 1.0094,
"step": 6155
},
{
"epoch": 2.6982041173894,
"grad_norm": 0.2676356622187676,
"learning_rate": 6.102607665853477e-07,
"loss": 1.023,
"step": 6160
},
{
"epoch": 2.700394218134034,
"grad_norm": 0.2573872899177252,
"learning_rate": 6.015252749891231e-07,
"loss": 1.0111,
"step": 6165
},
{
"epoch": 2.7025843188786682,
"grad_norm": 0.26768422170634076,
"learning_rate": 5.928508173652736e-07,
"loss": 1.0414,
"step": 6170
},
{
"epoch": 2.704774419623303,
"grad_norm": 0.2702917236559311,
"learning_rate": 5.842374500459791e-07,
"loss": 1.0123,
"step": 6175
},
{
"epoch": 2.706964520367937,
"grad_norm": 0.26063052725857316,
"learning_rate": 5.756852289667048e-07,
"loss": 1.0143,
"step": 6180
},
{
"epoch": 2.709154621112571,
"grad_norm": 0.2736945051716679,
"learning_rate": 5.671942096658256e-07,
"loss": 1.0219,
"step": 6185
},
{
"epoch": 2.7113447218572055,
"grad_norm": 0.26033814003152445,
"learning_rate": 5.587644472842691e-07,
"loss": 1.027,
"step": 6190
},
{
"epoch": 2.7135348226018396,
"grad_norm": 0.2747695664781286,
"learning_rate": 5.503959965651651e-07,
"loss": 1.0328,
"step": 6195
},
{
"epoch": 2.7157249233464738,
"grad_norm": 0.2639436223194078,
"learning_rate": 5.420889118534767e-07,
"loss": 1.0184,
"step": 6200
},
{
"epoch": 2.7179150240911083,
"grad_norm": 0.26625978732822686,
"learning_rate": 5.33843247095659e-07,
"loss": 1.0137,
"step": 6205
},
{
"epoch": 2.7201051248357424,
"grad_norm": 0.2649792774062934,
"learning_rate": 5.256590558393016e-07,
"loss": 1.0195,
"step": 6210
},
{
"epoch": 2.7222952255803765,
"grad_norm": 0.2617036738785984,
"learning_rate": 5.175363912327835e-07,
"loss": 1.017,
"step": 6215
},
{
"epoch": 2.724485326325011,
"grad_norm": 0.2699060497399551,
"learning_rate": 5.094753060249302e-07,
"loss": 1.0291,
"step": 6220
},
{
"epoch": 2.726675427069645,
"grad_norm": 0.2649872703564233,
"learning_rate": 5.014758525646679e-07,
"loss": 0.9908,
"step": 6225
},
{
"epoch": 2.7288655278142793,
"grad_norm": 0.2651642492888453,
"learning_rate": 4.935380828006797e-07,
"loss": 1.0137,
"step": 6230
},
{
"epoch": 2.731055628558914,
"grad_norm": 0.26425590598727583,
"learning_rate": 4.856620482810803e-07,
"loss": 1.0154,
"step": 6235
},
{
"epoch": 2.733245729303548,
"grad_norm": 0.2761774875687218,
"learning_rate": 4.778478001530673e-07,
"loss": 1.0252,
"step": 6240
},
{
"epoch": 2.735435830048182,
"grad_norm": 0.2690718325322844,
"learning_rate": 4.7009538916260254e-07,
"loss": 1.0211,
"step": 6245
},
{
"epoch": 2.7376259307928166,
"grad_norm": 0.26094127741106216,
"learning_rate": 4.624048656540703e-07,
"loss": 1.0018,
"step": 6250
},
{
"epoch": 2.7398160315374507,
"grad_norm": 0.26073608470470544,
"learning_rate": 4.547762795699551e-07,
"loss": 1.0293,
"step": 6255
},
{
"epoch": 2.742006132282085,
"grad_norm": 0.26176116543356426,
"learning_rate": 4.4720968045052327e-07,
"loss": 1.0281,
"step": 6260
},
{
"epoch": 2.7441962330267193,
"grad_norm": 0.2587633523361001,
"learning_rate": 4.3970511743349095e-07,
"loss": 1.0119,
"step": 6265
},
{
"epoch": 2.7463863337713534,
"grad_norm": 0.26467827125109494,
"learning_rate": 4.3226263925371323e-07,
"loss": 1.0189,
"step": 6270
},
{
"epoch": 2.7485764345159875,
"grad_norm": 0.26838751461155635,
"learning_rate": 4.248822942428643e-07,
"loss": 1.0141,
"step": 6275
},
{
"epoch": 2.750766535260622,
"grad_norm": 0.27455488470046974,
"learning_rate": 4.175641303291178e-07,
"loss": 1.0303,
"step": 6280
},
{
"epoch": 2.752956636005256,
"grad_norm": 0.2608592038076227,
"learning_rate": 4.103081950368526e-07,
"loss": 1.0201,
"step": 6285
},
{
"epoch": 2.7551467367498903,
"grad_norm": 0.26905266300546427,
"learning_rate": 4.0311453548632084e-07,
"loss": 1.0504,
"step": 6290
},
{
"epoch": 2.757336837494525,
"grad_norm": 0.26225270534619144,
"learning_rate": 3.9598319839336484e-07,
"loss": 1.0301,
"step": 6295
},
{
"epoch": 2.759526938239159,
"grad_norm": 0.275637498653482,
"learning_rate": 3.8891423006909623e-07,
"loss": 1.0344,
"step": 6300
},
{
"epoch": 2.761717038983793,
"grad_norm": 0.25859464573598573,
"learning_rate": 3.819076764196028e-07,
"loss": 1.023,
"step": 6305
},
{
"epoch": 2.7639071397284276,
"grad_norm": 0.2623767091254652,
"learning_rate": 3.7496358294565325e-07,
"loss": 1.0355,
"step": 6310
},
{
"epoch": 2.7660972404730617,
"grad_norm": 0.2614628167949124,
"learning_rate": 3.6808199474239413e-07,
"loss": 1.0074,
"step": 6315
},
{
"epoch": 2.768287341217696,
"grad_norm": 0.2589581748365727,
"learning_rate": 3.612629564990622e-07,
"loss": 1.0012,
"step": 6320
},
{
"epoch": 2.7704774419623304,
"grad_norm": 0.26135089658558563,
"learning_rate": 3.545065124986935e-07,
"loss": 1.0516,
"step": 6325
},
{
"epoch": 2.7726675427069645,
"grad_norm": 0.2550260966747239,
"learning_rate": 3.4781270661783495e-07,
"loss": 1.0309,
"step": 6330
},
{
"epoch": 2.7748576434515986,
"grad_norm": 0.26241337371995144,
"learning_rate": 3.4118158232626095e-07,
"loss": 1.0102,
"step": 6335
},
{
"epoch": 2.777047744196233,
"grad_norm": 0.2717236091770032,
"learning_rate": 3.346131826866872e-07,
"loss": 1.0031,
"step": 6340
},
{
"epoch": 2.7792378449408672,
"grad_norm": 0.2627537948727433,
"learning_rate": 3.281075503544928e-07,
"loss": 1.0246,
"step": 6345
},
{
"epoch": 2.7814279456855013,
"grad_norm": 0.2733462694900371,
"learning_rate": 3.216647275774509e-07,
"loss": 1.0326,
"step": 6350
},
{
"epoch": 2.783618046430136,
"grad_norm": 0.2653361262256602,
"learning_rate": 3.1528475619543755e-07,
"loss": 1.0336,
"step": 6355
},
{
"epoch": 2.78580814717477,
"grad_norm": 0.2637059071556029,
"learning_rate": 3.0896767764018066e-07,
"loss": 1.0447,
"step": 6360
},
{
"epoch": 2.787998247919404,
"grad_norm": 0.2643417975733627,
"learning_rate": 3.0271353293497174e-07,
"loss": 1.0256,
"step": 6365
},
{
"epoch": 2.7901883486640386,
"grad_norm": 0.2664848830571204,
"learning_rate": 2.9652236269441027e-07,
"loss": 1.0205,
"step": 6370
},
{
"epoch": 2.7923784494086727,
"grad_norm": 0.25803930539316744,
"learning_rate": 2.903942071241406e-07,
"loss": 1.0258,
"step": 6375
},
{
"epoch": 2.794568550153307,
"grad_norm": 0.2595092497202521,
"learning_rate": 2.843291060205855e-07,
"loss": 1.0129,
"step": 6380
},
{
"epoch": 2.7967586508979414,
"grad_norm": 0.26891951593800467,
"learning_rate": 2.7832709877068985e-07,
"loss": 1.0174,
"step": 6385
},
{
"epoch": 2.7989487516425755,
"grad_norm": 0.2618144209855571,
"learning_rate": 2.723882243516685e-07,
"loss": 1.0289,
"step": 6390
},
{
"epoch": 2.8011388523872096,
"grad_norm": 0.26115034114899394,
"learning_rate": 2.665125213307429e-07,
"loss": 1.0096,
"step": 6395
},
{
"epoch": 2.803328953131844,
"grad_norm": 0.2635263953836736,
"learning_rate": 2.6070002786490523e-07,
"loss": 1.034,
"step": 6400
},
{
"epoch": 2.8055190538764783,
"grad_norm": 0.26418026561714664,
"learning_rate": 2.549507817006602e-07,
"loss": 0.9963,
"step": 6405
},
{
"epoch": 2.8077091546211124,
"grad_norm": 0.2603026964757617,
"learning_rate": 2.492648201737791e-07,
"loss": 1.0178,
"step": 6410
},
{
"epoch": 2.809899255365747,
"grad_norm": 0.2627929976284597,
"learning_rate": 2.4364218020906605e-07,
"loss": 1.0445,
"step": 6415
},
{
"epoch": 2.812089356110381,
"grad_norm": 0.2669549562791691,
"learning_rate": 2.3808289832011112e-07,
"loss": 1.0066,
"step": 6420
},
{
"epoch": 2.814279456855015,
"grad_norm": 0.2721770667997164,
"learning_rate": 2.325870106090544e-07,
"loss": 1.0373,
"step": 6425
},
{
"epoch": 2.8164695575996497,
"grad_norm": 0.26354214498011763,
"learning_rate": 2.2715455276635323e-07,
"loss": 1.0289,
"step": 6430
},
{
"epoch": 2.818659658344284,
"grad_norm": 0.27146122428424413,
"learning_rate": 2.2178556007054876e-07,
"loss": 1.0225,
"step": 6435
},
{
"epoch": 2.820849759088918,
"grad_norm": 0.2680864657143132,
"learning_rate": 2.164800673880385e-07,
"loss": 0.9902,
"step": 6440
},
{
"epoch": 2.8230398598335524,
"grad_norm": 0.256788359012176,
"learning_rate": 2.1123810917284883e-07,
"loss": 1.0104,
"step": 6445
},
{
"epoch": 2.8252299605781865,
"grad_norm": 0.25921069217588105,
"learning_rate": 2.0605971946641045e-07,
"loss": 1.018,
"step": 6450
},
{
"epoch": 2.8274200613228206,
"grad_norm": 0.262881975020387,
"learning_rate": 2.0094493189734e-07,
"loss": 1.0277,
"step": 6455
},
{
"epoch": 2.829610162067455,
"grad_norm": 0.26033018837371813,
"learning_rate": 1.958937796812177e-07,
"loss": 1.0266,
"step": 6460
},
{
"epoch": 2.8318002628120893,
"grad_norm": 0.2725677982906727,
"learning_rate": 1.9090629562037777e-07,
"loss": 1.0311,
"step": 6465
},
{
"epoch": 2.8339903635567234,
"grad_norm": 0.2578506507690712,
"learning_rate": 1.8598251210368624e-07,
"loss": 1.0258,
"step": 6470
},
{
"epoch": 2.836180464301358,
"grad_norm": 0.26040068151802814,
"learning_rate": 1.811224611063378e-07,
"loss": 1.0254,
"step": 6475
},
{
"epoch": 2.838370565045992,
"grad_norm": 0.2654370924009657,
"learning_rate": 1.7632617418965047e-07,
"loss": 1.0264,
"step": 6480
},
{
"epoch": 2.840560665790626,
"grad_norm": 0.2624447866987324,
"learning_rate": 1.7159368250085128e-07,
"loss": 1.008,
"step": 6485
},
{
"epoch": 2.8427507665352607,
"grad_norm": 0.2693327061312894,
"learning_rate": 1.66925016772882e-07,
"loss": 1.0328,
"step": 6490
},
{
"epoch": 2.844940867279895,
"grad_norm": 0.2623948762195147,
"learning_rate": 1.6232020732419363e-07,
"loss": 1.025,
"step": 6495
},
{
"epoch": 2.847130968024529,
"grad_norm": 0.2682313581066649,
"learning_rate": 1.5777928405855681e-07,
"loss": 1.0533,
"step": 6500
},
{
"epoch": 2.8493210687691635,
"grad_norm": 0.2634435740939065,
"learning_rate": 1.5330227646486063e-07,
"loss": 1.0012,
"step": 6505
},
{
"epoch": 2.8515111695137976,
"grad_norm": 0.2648570311945631,
"learning_rate": 1.4888921361692288e-07,
"loss": 1.0111,
"step": 6510
},
{
"epoch": 2.8537012702584317,
"grad_norm": 0.2523702071185083,
"learning_rate": 1.4454012417330688e-07,
"loss": 1.0203,
"step": 6515
},
{
"epoch": 2.8558913710030662,
"grad_norm": 0.25829191001092155,
"learning_rate": 1.402550363771271e-07,
"loss": 1.0205,
"step": 6520
},
{
"epoch": 2.8580814717477003,
"grad_norm": 0.2711962263676663,
"learning_rate": 1.3603397805586837e-07,
"loss": 1.0012,
"step": 6525
},
{
"epoch": 2.8602715724923344,
"grad_norm": 0.26761518076252255,
"learning_rate": 1.3187697662121135e-07,
"loss": 1.0025,
"step": 6530
},
{
"epoch": 2.862461673236969,
"grad_norm": 0.25953013799411634,
"learning_rate": 1.2778405906884173e-07,
"loss": 1.0326,
"step": 6535
},
{
"epoch": 2.864651773981603,
"grad_norm": 0.26481196168192916,
"learning_rate": 1.2375525197829142e-07,
"loss": 1.0117,
"step": 6540
},
{
"epoch": 2.866841874726237,
"grad_norm": 0.26480303588564624,
"learning_rate": 1.197905815127498e-07,
"loss": 1.0092,
"step": 6545
},
{
"epoch": 2.8690319754708717,
"grad_norm": 0.26673811546819176,
"learning_rate": 1.158900734189039e-07,
"loss": 1.0354,
"step": 6550
},
{
"epoch": 2.871222076215506,
"grad_norm": 0.25622503188012385,
"learning_rate": 1.1205375302677068e-07,
"loss": 0.9949,
"step": 6555
},
{
"epoch": 2.87341217696014,
"grad_norm": 0.26585114917527647,
"learning_rate": 1.0828164524952611e-07,
"loss": 1.0129,
"step": 6560
},
{
"epoch": 2.8756022777047745,
"grad_norm": 0.2658125353437615,
"learning_rate": 1.0457377458334972e-07,
"loss": 1.0309,
"step": 6565
},
{
"epoch": 2.8777923784494086,
"grad_norm": 0.2644621372001252,
"learning_rate": 1.0093016510726361e-07,
"loss": 1.0238,
"step": 6570
},
{
"epoch": 2.8799824791940427,
"grad_norm": 0.26285181246145756,
"learning_rate": 9.735084048297482e-08,
"loss": 1.057,
"step": 6575
},
{
"epoch": 2.8821725799386773,
"grad_norm": 0.26347424345721254,
"learning_rate": 9.383582395472435e-08,
"loss": 1.0172,
"step": 6580
},
{
"epoch": 2.8843626806833114,
"grad_norm": 0.25880950335332537,
"learning_rate": 9.038513834913165e-08,
"loss": 1.0102,
"step": 6585
},
{
"epoch": 2.8865527814279455,
"grad_norm": 0.2626233208204072,
"learning_rate": 8.699880607505151e-08,
"loss": 1.034,
"step": 6590
},
{
"epoch": 2.88874288217258,
"grad_norm": 0.26481783365515854,
"learning_rate": 8.367684912342522e-08,
"loss": 1.0252,
"step": 6595
},
{
"epoch": 2.890932982917214,
"grad_norm": 0.25531476024902067,
"learning_rate": 8.041928906714069e-08,
"loss": 1.0205,
"step": 6600
},
{
"epoch": 2.8931230836618482,
"grad_norm": 0.2622947023545746,
"learning_rate": 7.722614706088815e-08,
"loss": 0.9975,
"step": 6605
},
{
"epoch": 2.8953131844064828,
"grad_norm": 0.26062229732752173,
"learning_rate": 7.409744384102691e-08,
"loss": 1.0113,
"step": 6610
},
{
"epoch": 2.897503285151117,
"grad_norm": 0.2723792115437162,
"learning_rate": 7.103319972544653e-08,
"loss": 1.0068,
"step": 6615
},
{
"epoch": 2.899693385895751,
"grad_norm": 0.2616779503781998,
"learning_rate": 6.803343461344037e-08,
"loss": 1.0088,
"step": 6620
},
{
"epoch": 2.9018834866403855,
"grad_norm": 0.2525101695328338,
"learning_rate": 6.509816798557112e-08,
"loss": 1.0082,
"step": 6625
},
{
"epoch": 2.9040735873850196,
"grad_norm": 0.25967364820532146,
"learning_rate": 6.222741890354655e-08,
"loss": 0.9977,
"step": 6630
},
{
"epoch": 2.9062636881296537,
"grad_norm": 0.26366528034026343,
"learning_rate": 5.942120601009627e-08,
"loss": 1.0125,
"step": 6635
},
{
"epoch": 2.9084537888742883,
"grad_norm": 0.25629635035286313,
"learning_rate": 5.667954752884841e-08,
"loss": 1.0059,
"step": 6640
},
{
"epoch": 2.9106438896189224,
"grad_norm": 0.2596464176195038,
"learning_rate": 5.400246126421649e-08,
"loss": 1.008,
"step": 6645
},
{
"epoch": 2.9128339903635565,
"grad_norm": 0.2679323332344789,
"learning_rate": 5.13899646012761e-08,
"loss": 1.0508,
"step": 6650
},
{
"epoch": 2.915024091108191,
"grad_norm": 0.26006077791672805,
"learning_rate": 4.8842074505658366e-08,
"loss": 1.0084,
"step": 6655
},
{
"epoch": 2.917214191852825,
"grad_norm": 0.2636907745428655,
"learning_rate": 4.635880752343669e-08,
"loss": 1.0357,
"step": 6660
},
{
"epoch": 2.9194042925974593,
"grad_norm": 0.26795706149424414,
"learning_rate": 4.394017978101905e-08,
"loss": 1.0348,
"step": 6665
},
{
"epoch": 2.921594393342094,
"grad_norm": 0.2608332255135537,
"learning_rate": 4.1586206985047006e-08,
"loss": 1.0221,
"step": 6670
},
{
"epoch": 2.923784494086728,
"grad_norm": 0.26982387956347165,
"learning_rate": 3.9296904422289064e-08,
"loss": 1.0326,
"step": 6675
},
{
"epoch": 2.925974594831362,
"grad_norm": 0.26654827231521183,
"learning_rate": 3.707228695954301e-08,
"loss": 1.0227,
"step": 6680
},
{
"epoch": 2.9281646955759966,
"grad_norm": 0.26353562404229947,
"learning_rate": 3.491236904354156e-08,
"loss": 1.0156,
"step": 6685
},
{
"epoch": 2.9303547963206307,
"grad_norm": 0.26222190486801167,
"learning_rate": 3.28171647008535e-08,
"loss": 0.9871,
"step": 6690
},
{
"epoch": 2.9325448970652648,
"grad_norm": 0.26937698718561326,
"learning_rate": 3.0786687537800454e-08,
"loss": 1.0158,
"step": 6695
},
{
"epoch": 2.9347349978098993,
"grad_norm": 0.2667931921324664,
"learning_rate": 2.8820950740358066e-08,
"loss": 1.0244,
"step": 6700
},
{
"epoch": 2.9369250985545334,
"grad_norm": 0.25675719984449435,
"learning_rate": 2.691996707408051e-08,
"loss": 1.0252,
"step": 6705
},
{
"epoch": 2.9391151992991675,
"grad_norm": 0.26641359009583415,
"learning_rate": 2.5083748884012772e-08,
"loss": 1.0232,
"step": 6710
},
{
"epoch": 2.941305300043802,
"grad_norm": 0.2643669870540917,
"learning_rate": 2.3312308094607382e-08,
"loss": 1.0402,
"step": 6715
},
{
"epoch": 2.943495400788436,
"grad_norm": 0.2640482226409207,
"learning_rate": 2.1605656209656713e-08,
"loss": 1.0123,
"step": 6720
},
{
"epoch": 2.9456855015330703,
"grad_norm": 0.257416143001543,
"learning_rate": 1.996380431220635e-08,
"loss": 1.0018,
"step": 6725
},
{
"epoch": 2.947875602277705,
"grad_norm": 0.25925869910092936,
"learning_rate": 1.838676306448961e-08,
"loss": 1.0105,
"step": 6730
},
{
"epoch": 2.950065703022339,
"grad_norm": 0.2682720189373006,
"learning_rate": 1.687454270786204e-08,
"loss": 1.0258,
"step": 6735
},
{
"epoch": 2.952255803766973,
"grad_norm": 0.2638121872050763,
"learning_rate": 1.542715306272591e-08,
"loss": 1.0398,
"step": 6740
},
{
"epoch": 2.9544459045116076,
"grad_norm": 0.2552510538750082,
"learning_rate": 1.4044603528471368e-08,
"loss": 1.0221,
"step": 6745
},
{
"epoch": 2.9566360052562417,
"grad_norm": 0.2660512986505126,
"learning_rate": 1.2726903083417619e-08,
"loss": 1.0199,
"step": 6750
},
{
"epoch": 2.958826106000876,
"grad_norm": 0.26097986341434426,
"learning_rate": 1.1474060284748512e-08,
"loss": 0.9941,
"step": 6755
},
{
"epoch": 2.9610162067455104,
"grad_norm": 0.2696098253592492,
"learning_rate": 1.0286083268463698e-08,
"loss": 1.0227,
"step": 6760
},
{
"epoch": 2.9632063074901445,
"grad_norm": 0.2651201330068607,
"learning_rate": 9.16297974931979e-09,
"loss": 1.0469,
"step": 6765
},
{
"epoch": 2.9653964082347786,
"grad_norm": 0.261068060678048,
"learning_rate": 8.104757020784838e-09,
"loss": 1.0203,
"step": 6770
},
{
"epoch": 2.967586508979413,
"grad_norm": 0.26764709627692834,
"learning_rate": 7.111421954989484e-09,
"loss": 1.0379,
"step": 6775
},
{
"epoch": 2.9697766097240472,
"grad_norm": 0.26991775591350575,
"learning_rate": 6.182981002679223e-09,
"loss": 1.027,
"step": 6780
},
{
"epoch": 2.9719667104686813,
"grad_norm": 0.2662528611485649,
"learning_rate": 5.31944019317665e-09,
"loss": 1.0146,
"step": 6785
},
{
"epoch": 2.974156811213316,
"grad_norm": 0.265221109043864,
"learning_rate": 4.52080513434372e-09,
"loss": 1.0186,
"step": 6790
},
{
"epoch": 2.97634691195795,
"grad_norm": 0.26743085714777837,
"learning_rate": 3.7870810125384406e-09,
"loss": 1.0064,
"step": 6795
},
{
"epoch": 2.978537012702584,
"grad_norm": 0.26402623363889577,
"learning_rate": 3.118272592587124e-09,
"loss": 1.0344,
"step": 6800
},
{
"epoch": 2.9807271134472186,
"grad_norm": 0.25773362294157315,
"learning_rate": 2.5143842177510758e-09,
"loss": 1.0084,
"step": 6805
},
{
"epoch": 2.9829172141918527,
"grad_norm": 0.26980590288864464,
"learning_rate": 1.975419809699952e-09,
"loss": 1.0227,
"step": 6810
},
{
"epoch": 2.985107314936487,
"grad_norm": 0.2530142845867315,
"learning_rate": 1.5013828684840025e-09,
"loss": 0.9992,
"step": 6815
},
{
"epoch": 2.9872974156811214,
"grad_norm": 0.26702416469760076,
"learning_rate": 1.0922764725140867e-09,
"loss": 1.0303,
"step": 6820
},
{
"epoch": 2.9894875164257555,
"grad_norm": 0.26248596298655585,
"learning_rate": 7.481032785372488e-10,
"loss": 1.0215,
"step": 6825
},
{
"epoch": 2.9916776171703896,
"grad_norm": 0.26614527758419954,
"learning_rate": 4.688655216256167e-10,
"loss": 1.0164,
"step": 6830
},
{
"epoch": 2.993867717915024,
"grad_norm": 0.2620390464236342,
"learning_rate": 2.545650151586365e-10,
"loss": 1.0002,
"step": 6835
},
{
"epoch": 2.9960578186596583,
"grad_norm": 0.25671053477383604,
"learning_rate": 1.0520315080753129e-10,
"loss": 1.0336,
"step": 6840
},
{
"epoch": 2.9982479194042924,
"grad_norm": 0.26588732851444397,
"learning_rate": 2.0780898535299742e-11,
"loss": 1.0246,
"step": 6845
},
{
"epoch": 3.0,
"step": 6849,
"total_flos": 648076339445760.0,
"train_loss": 1.1352002231165133,
"train_runtime": 23571.371,
"train_samples_per_second": 18.589,
"train_steps_per_second": 0.291
}
],
"logging_steps": 5,
"max_steps": 6849,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"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": 648076339445760.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}