rqadri's picture
Add files using upload-large-folder tool
d9e781c verified
Raw History Blame Contribute Delete
180 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9996427933559564,
"eval_steps": 500,
"global_step": 11194,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.000893016610108948,
"grad_norm": 3.7484283447265625,
"learning_rate": 1.1061797752808989e-06,
"loss": 1.4058,
"step": 10
},
{
"epoch": 0.001786033220217896,
"grad_norm": 1.9000362157821655,
"learning_rate": 2.2241573033707866e-06,
"loss": 1.3045,
"step": 20
},
{
"epoch": 0.002679049830326844,
"grad_norm": 2.4037389755249023,
"learning_rate": 3.342134831460675e-06,
"loss": 1.1338,
"step": 30
},
{
"epoch": 0.003572066440435792,
"grad_norm": 1.0070688724517822,
"learning_rate": 4.4601123595505624e-06,
"loss": 1.0082,
"step": 40
},
{
"epoch": 0.00446508305054474,
"grad_norm": 0.6030862927436829,
"learning_rate": 5.5780898876404495e-06,
"loss": 0.8595,
"step": 50
},
{
"epoch": 0.005358099660653688,
"grad_norm": 0.6932407021522522,
"learning_rate": 6.6960674157303374e-06,
"loss": 0.7768,
"step": 60
},
{
"epoch": 0.006251116270762636,
"grad_norm": 0.5126384496688843,
"learning_rate": 7.814044943820226e-06,
"loss": 0.7961,
"step": 70
},
{
"epoch": 0.007144132880871584,
"grad_norm": 0.5326073169708252,
"learning_rate": 8.932022471910112e-06,
"loss": 0.777,
"step": 80
},
{
"epoch": 0.008037149490980533,
"grad_norm": 0.4840937554836273,
"learning_rate": 1.005e-05,
"loss": 0.8092,
"step": 90
},
{
"epoch": 0.00893016610108948,
"grad_norm": 0.4524643123149872,
"learning_rate": 1.1167977528089888e-05,
"loss": 0.7689,
"step": 100
},
{
"epoch": 0.009823182711198428,
"grad_norm": 0.4209403097629547,
"learning_rate": 1.2285955056179776e-05,
"loss": 0.7862,
"step": 110
},
{
"epoch": 0.010716199321307376,
"grad_norm": 0.47385457158088684,
"learning_rate": 1.3403932584269664e-05,
"loss": 0.7922,
"step": 120
},
{
"epoch": 0.011609215931416325,
"grad_norm": 0.4067447781562805,
"learning_rate": 1.4521910112359552e-05,
"loss": 0.7441,
"step": 130
},
{
"epoch": 0.012502232541525273,
"grad_norm": 0.467846542596817,
"learning_rate": 1.563988764044944e-05,
"loss": 0.8082,
"step": 140
},
{
"epoch": 0.01339524915163422,
"grad_norm": 0.5435770153999329,
"learning_rate": 1.6757865168539328e-05,
"loss": 0.8008,
"step": 150
},
{
"epoch": 0.014288265761743168,
"grad_norm": 0.4837871789932251,
"learning_rate": 1.7875842696629214e-05,
"loss": 0.7348,
"step": 160
},
{
"epoch": 0.015181282371852116,
"grad_norm": 0.44955676794052124,
"learning_rate": 1.89938202247191e-05,
"loss": 0.7796,
"step": 170
},
{
"epoch": 0.016074298981961065,
"grad_norm": 0.4254562258720398,
"learning_rate": 2e-05,
"loss": 0.7696,
"step": 180
},
{
"epoch": 0.01696731559207001,
"grad_norm": 0.4308949410915375,
"learning_rate": 2e-05,
"loss": 0.7623,
"step": 190
},
{
"epoch": 0.01786033220217896,
"grad_norm": 0.4657016098499298,
"learning_rate": 2e-05,
"loss": 0.741,
"step": 200
},
{
"epoch": 0.01875334881228791,
"grad_norm": 0.4654051661491394,
"learning_rate": 2e-05,
"loss": 0.7624,
"step": 210
},
{
"epoch": 0.019646365422396856,
"grad_norm": 0.4155106544494629,
"learning_rate": 2e-05,
"loss": 0.7521,
"step": 220
},
{
"epoch": 0.020539382032505805,
"grad_norm": 0.47818872332572937,
"learning_rate": 2e-05,
"loss": 0.7431,
"step": 230
},
{
"epoch": 0.02143239864261475,
"grad_norm": 0.43436357378959656,
"learning_rate": 2e-05,
"loss": 0.7559,
"step": 240
},
{
"epoch": 0.0223254152527237,
"grad_norm": 0.4438256621360779,
"learning_rate": 2e-05,
"loss": 0.7874,
"step": 250
},
{
"epoch": 0.02321843186283265,
"grad_norm": 0.42037931084632874,
"learning_rate": 2e-05,
"loss": 0.7776,
"step": 260
},
{
"epoch": 0.024111448472941596,
"grad_norm": 0.4763539433479309,
"learning_rate": 2e-05,
"loss": 0.7769,
"step": 270
},
{
"epoch": 0.025004465083050546,
"grad_norm": 0.4746975898742676,
"learning_rate": 2e-05,
"loss": 0.7761,
"step": 280
},
{
"epoch": 0.02589748169315949,
"grad_norm": 0.4424160122871399,
"learning_rate": 2e-05,
"loss": 0.7603,
"step": 290
},
{
"epoch": 0.02679049830326844,
"grad_norm": 0.4469321668148041,
"learning_rate": 2e-05,
"loss": 0.756,
"step": 300
},
{
"epoch": 0.02768351491337739,
"grad_norm": 0.42572203278541565,
"learning_rate": 2e-05,
"loss": 0.7527,
"step": 310
},
{
"epoch": 0.028576531523486336,
"grad_norm": 0.37948864698410034,
"learning_rate": 2e-05,
"loss": 0.6629,
"step": 320
},
{
"epoch": 0.029469548133595286,
"grad_norm": 0.4720393717288971,
"learning_rate": 2e-05,
"loss": 0.7598,
"step": 330
},
{
"epoch": 0.030362564743704232,
"grad_norm": 0.4645335078239441,
"learning_rate": 2e-05,
"loss": 0.7635,
"step": 340
},
{
"epoch": 0.03125558135381318,
"grad_norm": 0.4660659432411194,
"learning_rate": 2e-05,
"loss": 0.77,
"step": 350
},
{
"epoch": 0.03214859796392213,
"grad_norm": 0.47986748814582825,
"learning_rate": 2e-05,
"loss": 0.7442,
"step": 360
},
{
"epoch": 0.03304161457403108,
"grad_norm": 0.4416871964931488,
"learning_rate": 2e-05,
"loss": 0.7316,
"step": 370
},
{
"epoch": 0.03393463118414002,
"grad_norm": 0.3982086181640625,
"learning_rate": 2e-05,
"loss": 0.7805,
"step": 380
},
{
"epoch": 0.034827647794248975,
"grad_norm": 0.4491710662841797,
"learning_rate": 2e-05,
"loss": 0.743,
"step": 390
},
{
"epoch": 0.03572066440435792,
"grad_norm": 0.4009242653846741,
"learning_rate": 2e-05,
"loss": 0.7401,
"step": 400
},
{
"epoch": 0.03661368101446687,
"grad_norm": 0.3790673017501831,
"learning_rate": 2e-05,
"loss": 0.7317,
"step": 410
},
{
"epoch": 0.03750669762457582,
"grad_norm": 0.38693296909332275,
"learning_rate": 2e-05,
"loss": 0.7373,
"step": 420
},
{
"epoch": 0.038399714234684766,
"grad_norm": 0.4592219293117523,
"learning_rate": 2e-05,
"loss": 0.7421,
"step": 430
},
{
"epoch": 0.03929273084479371,
"grad_norm": 0.48258858919143677,
"learning_rate": 2e-05,
"loss": 0.7904,
"step": 440
},
{
"epoch": 0.04018574745490266,
"grad_norm": 0.8017710447311401,
"learning_rate": 2e-05,
"loss": 0.7501,
"step": 450
},
{
"epoch": 0.04107876406501161,
"grad_norm": 0.41467684507369995,
"learning_rate": 2e-05,
"loss": 0.7129,
"step": 460
},
{
"epoch": 0.04197178067512056,
"grad_norm": 0.490904301404953,
"learning_rate": 2e-05,
"loss": 0.7671,
"step": 470
},
{
"epoch": 0.0428647972852295,
"grad_norm": 0.4521087110042572,
"learning_rate": 2e-05,
"loss": 0.7397,
"step": 480
},
{
"epoch": 0.043757813895338456,
"grad_norm": 0.4126375913619995,
"learning_rate": 2e-05,
"loss": 0.7391,
"step": 490
},
{
"epoch": 0.0446508305054474,
"grad_norm": 0.4476797878742218,
"learning_rate": 2e-05,
"loss": 0.7796,
"step": 500
},
{
"epoch": 0.04554384711555635,
"grad_norm": 0.43812263011932373,
"learning_rate": 2e-05,
"loss": 0.7352,
"step": 510
},
{
"epoch": 0.0464368637256653,
"grad_norm": 0.48256680369377136,
"learning_rate": 2e-05,
"loss": 0.755,
"step": 520
},
{
"epoch": 0.047329880335774246,
"grad_norm": 0.39438191056251526,
"learning_rate": 2e-05,
"loss": 0.7438,
"step": 530
},
{
"epoch": 0.04822289694588319,
"grad_norm": 0.4146062731742859,
"learning_rate": 2e-05,
"loss": 0.7275,
"step": 540
},
{
"epoch": 0.04911591355599214,
"grad_norm": 0.41129109263420105,
"learning_rate": 2e-05,
"loss": 0.7226,
"step": 550
},
{
"epoch": 0.05000893016610109,
"grad_norm": 0.38930240273475647,
"learning_rate": 2e-05,
"loss": 0.7625,
"step": 560
},
{
"epoch": 0.05090194677621004,
"grad_norm": 0.48423701524734497,
"learning_rate": 2e-05,
"loss": 0.7631,
"step": 570
},
{
"epoch": 0.05179496338631898,
"grad_norm": 0.40557992458343506,
"learning_rate": 2e-05,
"loss": 0.7481,
"step": 580
},
{
"epoch": 0.052687979996427936,
"grad_norm": 0.4137013554573059,
"learning_rate": 2e-05,
"loss": 0.7481,
"step": 590
},
{
"epoch": 0.05358099660653688,
"grad_norm": 0.45856934785842896,
"learning_rate": 2e-05,
"loss": 0.7721,
"step": 600
},
{
"epoch": 0.05447401321664583,
"grad_norm": 0.4169819951057434,
"learning_rate": 2e-05,
"loss": 0.7525,
"step": 610
},
{
"epoch": 0.05536702982675478,
"grad_norm": 0.38590312004089355,
"learning_rate": 2e-05,
"loss": 0.7374,
"step": 620
},
{
"epoch": 0.05626004643686373,
"grad_norm": 0.40914642810821533,
"learning_rate": 2e-05,
"loss": 0.7495,
"step": 630
},
{
"epoch": 0.05715306304697267,
"grad_norm": 0.41702547669410706,
"learning_rate": 2e-05,
"loss": 0.7398,
"step": 640
},
{
"epoch": 0.05804607965708162,
"grad_norm": 0.39613404870033264,
"learning_rate": 2e-05,
"loss": 0.74,
"step": 650
},
{
"epoch": 0.05893909626719057,
"grad_norm": 0.39417654275894165,
"learning_rate": 2e-05,
"loss": 0.7308,
"step": 660
},
{
"epoch": 0.05983211287729952,
"grad_norm": 0.3835000693798065,
"learning_rate": 2e-05,
"loss": 0.7184,
"step": 670
},
{
"epoch": 0.060725129487408464,
"grad_norm": 0.40290987491607666,
"learning_rate": 2e-05,
"loss": 0.7252,
"step": 680
},
{
"epoch": 0.061618146097517416,
"grad_norm": 0.5707727074623108,
"learning_rate": 2e-05,
"loss": 0.7349,
"step": 690
},
{
"epoch": 0.06251116270762636,
"grad_norm": 0.44035205245018005,
"learning_rate": 2e-05,
"loss": 0.7322,
"step": 700
},
{
"epoch": 0.06340417931773532,
"grad_norm": 0.4347422420978546,
"learning_rate": 2e-05,
"loss": 0.7574,
"step": 710
},
{
"epoch": 0.06429719592784426,
"grad_norm": 0.4078957140445709,
"learning_rate": 2e-05,
"loss": 0.7621,
"step": 720
},
{
"epoch": 0.06519021253795321,
"grad_norm": 0.3920077979564667,
"learning_rate": 2e-05,
"loss": 0.7195,
"step": 730
},
{
"epoch": 0.06608322914806215,
"grad_norm": 0.38933590054512024,
"learning_rate": 2e-05,
"loss": 0.7471,
"step": 740
},
{
"epoch": 0.0669762457581711,
"grad_norm": 0.4065713584423065,
"learning_rate": 2e-05,
"loss": 0.7767,
"step": 750
},
{
"epoch": 0.06786926236828004,
"grad_norm": 0.41228511929512024,
"learning_rate": 2e-05,
"loss": 0.7547,
"step": 760
},
{
"epoch": 0.068762278978389,
"grad_norm": 0.3922688364982605,
"learning_rate": 2e-05,
"loss": 0.7148,
"step": 770
},
{
"epoch": 0.06965529558849795,
"grad_norm": 0.4028988778591156,
"learning_rate": 2e-05,
"loss": 0.7262,
"step": 780
},
{
"epoch": 0.0705483121986069,
"grad_norm": 0.44637545943260193,
"learning_rate": 2e-05,
"loss": 0.7614,
"step": 790
},
{
"epoch": 0.07144132880871584,
"grad_norm": 0.40410199761390686,
"learning_rate": 2e-05,
"loss": 0.7524,
"step": 800
},
{
"epoch": 0.07233434541882479,
"grad_norm": 0.41571539640426636,
"learning_rate": 2e-05,
"loss": 0.7423,
"step": 810
},
{
"epoch": 0.07322736202893373,
"grad_norm": 0.48081284761428833,
"learning_rate": 2e-05,
"loss": 0.7067,
"step": 820
},
{
"epoch": 0.07412037863904268,
"grad_norm": 0.40906158089637756,
"learning_rate": 2e-05,
"loss": 0.7576,
"step": 830
},
{
"epoch": 0.07501339524915164,
"grad_norm": 0.37821027636528015,
"learning_rate": 2e-05,
"loss": 0.7579,
"step": 840
},
{
"epoch": 0.07590641185926059,
"grad_norm": 0.3791477382183075,
"learning_rate": 2e-05,
"loss": 0.7153,
"step": 850
},
{
"epoch": 0.07679942846936953,
"grad_norm": 0.38766372203826904,
"learning_rate": 2e-05,
"loss": 0.7393,
"step": 860
},
{
"epoch": 0.07769244507947848,
"grad_norm": 0.3937142491340637,
"learning_rate": 2e-05,
"loss": 0.7371,
"step": 870
},
{
"epoch": 0.07858546168958742,
"grad_norm": 0.3776763677597046,
"learning_rate": 2e-05,
"loss": 0.7221,
"step": 880
},
{
"epoch": 0.07947847829969637,
"grad_norm": 0.4015044867992401,
"learning_rate": 2e-05,
"loss": 0.723,
"step": 890
},
{
"epoch": 0.08037149490980532,
"grad_norm": 0.43104034662246704,
"learning_rate": 2e-05,
"loss": 0.7621,
"step": 900
},
{
"epoch": 0.08126451151991428,
"grad_norm": 0.3779524564743042,
"learning_rate": 2e-05,
"loss": 0.7332,
"step": 910
},
{
"epoch": 0.08215752813002322,
"grad_norm": 0.3575296998023987,
"learning_rate": 2e-05,
"loss": 0.7575,
"step": 920
},
{
"epoch": 0.08305054474013217,
"grad_norm": 0.43516188859939575,
"learning_rate": 2e-05,
"loss": 0.7601,
"step": 930
},
{
"epoch": 0.08394356135024111,
"grad_norm": 0.3753051161766052,
"learning_rate": 2e-05,
"loss": 0.7269,
"step": 940
},
{
"epoch": 0.08483657796035006,
"grad_norm": 0.3971746861934662,
"learning_rate": 2e-05,
"loss": 0.7396,
"step": 950
},
{
"epoch": 0.085729594570459,
"grad_norm": 0.41950562596321106,
"learning_rate": 2e-05,
"loss": 0.7804,
"step": 960
},
{
"epoch": 0.08662261118056797,
"grad_norm": 0.3685847222805023,
"learning_rate": 2e-05,
"loss": 0.722,
"step": 970
},
{
"epoch": 0.08751562779067691,
"grad_norm": 0.3798861801624298,
"learning_rate": 2e-05,
"loss": 0.752,
"step": 980
},
{
"epoch": 0.08840864440078586,
"grad_norm": 0.40628236532211304,
"learning_rate": 2e-05,
"loss": 0.7563,
"step": 990
},
{
"epoch": 0.0893016610108948,
"grad_norm": 0.37584495544433594,
"learning_rate": 2e-05,
"loss": 0.719,
"step": 1000
},
{
"epoch": 0.09019467762100375,
"grad_norm": 0.3755060136318207,
"learning_rate": 2e-05,
"loss": 0.7321,
"step": 1010
},
{
"epoch": 0.0910876942311127,
"grad_norm": 0.3935806155204773,
"learning_rate": 2e-05,
"loss": 0.7296,
"step": 1020
},
{
"epoch": 0.09198071084122164,
"grad_norm": 0.4500720798969269,
"learning_rate": 2e-05,
"loss": 0.7681,
"step": 1030
},
{
"epoch": 0.0928737274513306,
"grad_norm": 0.4131600856781006,
"learning_rate": 2e-05,
"loss": 0.7691,
"step": 1040
},
{
"epoch": 0.09376674406143955,
"grad_norm": 0.38126325607299805,
"learning_rate": 2e-05,
"loss": 0.7203,
"step": 1050
},
{
"epoch": 0.09465976067154849,
"grad_norm": 0.438177227973938,
"learning_rate": 2e-05,
"loss": 0.7332,
"step": 1060
},
{
"epoch": 0.09555277728165744,
"grad_norm": 0.3714941143989563,
"learning_rate": 2e-05,
"loss": 0.7107,
"step": 1070
},
{
"epoch": 0.09644579389176638,
"grad_norm": 0.37563997507095337,
"learning_rate": 2e-05,
"loss": 0.7064,
"step": 1080
},
{
"epoch": 0.09733881050187533,
"grad_norm": 0.38376548886299133,
"learning_rate": 2e-05,
"loss": 0.7471,
"step": 1090
},
{
"epoch": 0.09823182711198428,
"grad_norm": 0.4249208867549896,
"learning_rate": 2e-05,
"loss": 0.7632,
"step": 1100
},
{
"epoch": 0.09912484372209324,
"grad_norm": 0.39996418356895447,
"learning_rate": 2e-05,
"loss": 0.7302,
"step": 1110
},
{
"epoch": 0.10001786033220218,
"grad_norm": 0.4173353612422943,
"learning_rate": 2e-05,
"loss": 0.7412,
"step": 1120
},
{
"epoch": 0.10091087694231113,
"grad_norm": 0.3551906943321228,
"learning_rate": 2e-05,
"loss": 0.7452,
"step": 1130
},
{
"epoch": 0.10180389355242007,
"grad_norm": 0.4338817894458771,
"learning_rate": 2e-05,
"loss": 0.7549,
"step": 1140
},
{
"epoch": 0.10269691016252902,
"grad_norm": 0.49487829208374023,
"learning_rate": 2e-05,
"loss": 0.7578,
"step": 1150
},
{
"epoch": 0.10358992677263797,
"grad_norm": 0.37487703561782837,
"learning_rate": 2e-05,
"loss": 0.7532,
"step": 1160
},
{
"epoch": 0.10448294338274691,
"grad_norm": 0.36092835664749146,
"learning_rate": 2e-05,
"loss": 0.7386,
"step": 1170
},
{
"epoch": 0.10537595999285587,
"grad_norm": 0.3902466297149658,
"learning_rate": 2e-05,
"loss": 0.7287,
"step": 1180
},
{
"epoch": 0.10626897660296482,
"grad_norm": 0.36106592416763306,
"learning_rate": 2e-05,
"loss": 0.7497,
"step": 1190
},
{
"epoch": 0.10716199321307376,
"grad_norm": 0.42885327339172363,
"learning_rate": 2e-05,
"loss": 0.7801,
"step": 1200
},
{
"epoch": 0.10805500982318271,
"grad_norm": 0.36564674973487854,
"learning_rate": 2e-05,
"loss": 0.7152,
"step": 1210
},
{
"epoch": 0.10894802643329166,
"grad_norm": 0.34475770592689514,
"learning_rate": 2e-05,
"loss": 0.7162,
"step": 1220
},
{
"epoch": 0.1098410430434006,
"grad_norm": 0.379625529050827,
"learning_rate": 2e-05,
"loss": 0.7211,
"step": 1230
},
{
"epoch": 0.11073405965350956,
"grad_norm": 0.3452484905719757,
"learning_rate": 2e-05,
"loss": 0.7218,
"step": 1240
},
{
"epoch": 0.11162707626361851,
"grad_norm": 0.4212513566017151,
"learning_rate": 2e-05,
"loss": 0.7304,
"step": 1250
},
{
"epoch": 0.11252009287372745,
"grad_norm": 0.40886190533638,
"learning_rate": 2e-05,
"loss": 0.7373,
"step": 1260
},
{
"epoch": 0.1134131094838364,
"grad_norm": 0.3536968529224396,
"learning_rate": 2e-05,
"loss": 0.7323,
"step": 1270
},
{
"epoch": 0.11430612609394535,
"grad_norm": 0.38670650124549866,
"learning_rate": 2e-05,
"loss": 0.741,
"step": 1280
},
{
"epoch": 0.11519914270405429,
"grad_norm": 0.42771264910697937,
"learning_rate": 2e-05,
"loss": 0.7424,
"step": 1290
},
{
"epoch": 0.11609215931416324,
"grad_norm": 0.39867374300956726,
"learning_rate": 2e-05,
"loss": 0.7251,
"step": 1300
},
{
"epoch": 0.1169851759242722,
"grad_norm": 0.38744813203811646,
"learning_rate": 2e-05,
"loss": 0.715,
"step": 1310
},
{
"epoch": 0.11787819253438114,
"grad_norm": 0.4313223958015442,
"learning_rate": 2e-05,
"loss": 0.6988,
"step": 1320
},
{
"epoch": 0.11877120914449009,
"grad_norm": 0.3686710298061371,
"learning_rate": 2e-05,
"loss": 0.7162,
"step": 1330
},
{
"epoch": 0.11966422575459904,
"grad_norm": 0.8750284314155579,
"learning_rate": 2e-05,
"loss": 0.7319,
"step": 1340
},
{
"epoch": 0.12055724236470798,
"grad_norm": 0.3907322287559509,
"learning_rate": 2e-05,
"loss": 0.7475,
"step": 1350
},
{
"epoch": 0.12145025897481693,
"grad_norm": 0.355712354183197,
"learning_rate": 2e-05,
"loss": 0.7475,
"step": 1360
},
{
"epoch": 0.12234327558492587,
"grad_norm": 0.3635517954826355,
"learning_rate": 2e-05,
"loss": 0.7576,
"step": 1370
},
{
"epoch": 0.12323629219503483,
"grad_norm": 0.39474037289619446,
"learning_rate": 2e-05,
"loss": 0.6988,
"step": 1380
},
{
"epoch": 0.12412930880514378,
"grad_norm": 0.38797375559806824,
"learning_rate": 2e-05,
"loss": 0.7089,
"step": 1390
},
{
"epoch": 0.1250223254152527,
"grad_norm": 0.35904696583747864,
"learning_rate": 2e-05,
"loss": 0.7187,
"step": 1400
},
{
"epoch": 0.12591534202536167,
"grad_norm": 0.4828942120075226,
"learning_rate": 2e-05,
"loss": 0.7326,
"step": 1410
},
{
"epoch": 0.12680835863547063,
"grad_norm": 0.3671641945838928,
"learning_rate": 2e-05,
"loss": 0.7149,
"step": 1420
},
{
"epoch": 0.12770137524557956,
"grad_norm": 0.40876835584640503,
"learning_rate": 2e-05,
"loss": 0.7398,
"step": 1430
},
{
"epoch": 0.12859439185568852,
"grad_norm": 0.3652034401893616,
"learning_rate": 2e-05,
"loss": 0.724,
"step": 1440
},
{
"epoch": 0.12948740846579745,
"grad_norm": 0.3712545931339264,
"learning_rate": 2e-05,
"loss": 0.7169,
"step": 1450
},
{
"epoch": 0.13038042507590641,
"grad_norm": 0.39639317989349365,
"learning_rate": 2e-05,
"loss": 0.7623,
"step": 1460
},
{
"epoch": 0.13127344168601535,
"grad_norm": 0.38050130009651184,
"learning_rate": 2e-05,
"loss": 0.7259,
"step": 1470
},
{
"epoch": 0.1321664582961243,
"grad_norm": 0.39309820532798767,
"learning_rate": 2e-05,
"loss": 0.7601,
"step": 1480
},
{
"epoch": 0.13305947490623327,
"grad_norm": 0.3582014739513397,
"learning_rate": 2e-05,
"loss": 0.7489,
"step": 1490
},
{
"epoch": 0.1339524915163422,
"grad_norm": 0.37961331009864807,
"learning_rate": 2e-05,
"loss": 0.7129,
"step": 1500
},
{
"epoch": 0.13484550812645116,
"grad_norm": 0.5690667033195496,
"learning_rate": 2e-05,
"loss": 0.7345,
"step": 1510
},
{
"epoch": 0.1357385247365601,
"grad_norm": 0.3793013095855713,
"learning_rate": 2e-05,
"loss": 0.718,
"step": 1520
},
{
"epoch": 0.13663154134666905,
"grad_norm": 0.38825497031211853,
"learning_rate": 2e-05,
"loss": 0.7198,
"step": 1530
},
{
"epoch": 0.137524557956778,
"grad_norm": 0.38483718037605286,
"learning_rate": 2e-05,
"loss": 0.716,
"step": 1540
},
{
"epoch": 0.13841757456688694,
"grad_norm": 0.3974905014038086,
"learning_rate": 2e-05,
"loss": 0.7388,
"step": 1550
},
{
"epoch": 0.1393105911769959,
"grad_norm": 0.390345960855484,
"learning_rate": 2e-05,
"loss": 0.729,
"step": 1560
},
{
"epoch": 0.14020360778710483,
"grad_norm": 0.3711761236190796,
"learning_rate": 2e-05,
"loss": 0.7439,
"step": 1570
},
{
"epoch": 0.1410966243972138,
"grad_norm": 0.39335206151008606,
"learning_rate": 2e-05,
"loss": 0.6857,
"step": 1580
},
{
"epoch": 0.14198964100732273,
"grad_norm": 0.431006520986557,
"learning_rate": 2e-05,
"loss": 0.6739,
"step": 1590
},
{
"epoch": 0.14288265761743169,
"grad_norm": 0.37578749656677246,
"learning_rate": 2e-05,
"loss": 0.7253,
"step": 1600
},
{
"epoch": 0.14377567422754065,
"grad_norm": 0.36983808875083923,
"learning_rate": 2e-05,
"loss": 0.7087,
"step": 1610
},
{
"epoch": 0.14466869083764958,
"grad_norm": 0.36534303426742554,
"learning_rate": 2e-05,
"loss": 0.7048,
"step": 1620
},
{
"epoch": 0.14556170744775854,
"grad_norm": 0.3659409284591675,
"learning_rate": 2e-05,
"loss": 0.7145,
"step": 1630
},
{
"epoch": 0.14645472405786747,
"grad_norm": 0.4026823341846466,
"learning_rate": 2e-05,
"loss": 0.7084,
"step": 1640
},
{
"epoch": 0.14734774066797643,
"grad_norm": 0.40227675437927246,
"learning_rate": 2e-05,
"loss": 0.7521,
"step": 1650
},
{
"epoch": 0.14824075727808536,
"grad_norm": 0.3589552640914917,
"learning_rate": 2e-05,
"loss": 0.6962,
"step": 1660
},
{
"epoch": 0.14913377388819432,
"grad_norm": 0.40876489877700806,
"learning_rate": 2e-05,
"loss": 0.7581,
"step": 1670
},
{
"epoch": 0.15002679049830328,
"grad_norm": 0.3787183463573456,
"learning_rate": 2e-05,
"loss": 0.712,
"step": 1680
},
{
"epoch": 0.1509198071084122,
"grad_norm": 0.3726568818092346,
"learning_rate": 2e-05,
"loss": 0.7329,
"step": 1690
},
{
"epoch": 0.15181282371852117,
"grad_norm": 0.36480188369750977,
"learning_rate": 2e-05,
"loss": 0.7192,
"step": 1700
},
{
"epoch": 0.1527058403286301,
"grad_norm": 0.4598802328109741,
"learning_rate": 2e-05,
"loss": 0.742,
"step": 1710
},
{
"epoch": 0.15359885693873906,
"grad_norm": 0.34519508481025696,
"learning_rate": 2e-05,
"loss": 0.7418,
"step": 1720
},
{
"epoch": 0.154491873548848,
"grad_norm": 0.39450502395629883,
"learning_rate": 2e-05,
"loss": 0.7245,
"step": 1730
},
{
"epoch": 0.15538489015895696,
"grad_norm": 0.44403451681137085,
"learning_rate": 2e-05,
"loss": 0.7168,
"step": 1740
},
{
"epoch": 0.15627790676906592,
"grad_norm": 0.3881482183933258,
"learning_rate": 2e-05,
"loss": 0.7322,
"step": 1750
},
{
"epoch": 0.15717092337917485,
"grad_norm": 0.3961481750011444,
"learning_rate": 2e-05,
"loss": 0.7593,
"step": 1760
},
{
"epoch": 0.1580639399892838,
"grad_norm": 0.38921841979026794,
"learning_rate": 2e-05,
"loss": 0.7356,
"step": 1770
},
{
"epoch": 0.15895695659939274,
"grad_norm": 0.3746980130672455,
"learning_rate": 2e-05,
"loss": 0.7109,
"step": 1780
},
{
"epoch": 0.1598499732095017,
"grad_norm": 0.36172837018966675,
"learning_rate": 2e-05,
"loss": 0.7193,
"step": 1790
},
{
"epoch": 0.16074298981961063,
"grad_norm": 0.34081071615219116,
"learning_rate": 2e-05,
"loss": 0.734,
"step": 1800
},
{
"epoch": 0.1616360064297196,
"grad_norm": 0.3822826147079468,
"learning_rate": 2e-05,
"loss": 0.7213,
"step": 1810
},
{
"epoch": 0.16252902303982855,
"grad_norm": 0.39521777629852295,
"learning_rate": 2e-05,
"loss": 0.7071,
"step": 1820
},
{
"epoch": 0.16342203964993748,
"grad_norm": 0.3572978973388672,
"learning_rate": 2e-05,
"loss": 0.7122,
"step": 1830
},
{
"epoch": 0.16431505626004644,
"grad_norm": 0.37090566754341125,
"learning_rate": 2e-05,
"loss": 0.7656,
"step": 1840
},
{
"epoch": 0.16520807287015538,
"grad_norm": 1.1004981994628906,
"learning_rate": 2e-05,
"loss": 0.7045,
"step": 1850
},
{
"epoch": 0.16610108948026434,
"grad_norm": 0.406376451253891,
"learning_rate": 2e-05,
"loss": 0.7121,
"step": 1860
},
{
"epoch": 0.16699410609037327,
"grad_norm": 0.410037100315094,
"learning_rate": 2e-05,
"loss": 0.6982,
"step": 1870
},
{
"epoch": 0.16788712270048223,
"grad_norm": 0.4311705231666565,
"learning_rate": 2e-05,
"loss": 0.6981,
"step": 1880
},
{
"epoch": 0.1687801393105912,
"grad_norm": 0.4563307464122772,
"learning_rate": 2e-05,
"loss": 0.7065,
"step": 1890
},
{
"epoch": 0.16967315592070012,
"grad_norm": 0.4294278621673584,
"learning_rate": 2e-05,
"loss": 0.7177,
"step": 1900
},
{
"epoch": 0.17056617253080908,
"grad_norm": 0.41810211539268494,
"learning_rate": 2e-05,
"loss": 0.7447,
"step": 1910
},
{
"epoch": 0.171459189140918,
"grad_norm": 0.47829675674438477,
"learning_rate": 2e-05,
"loss": 0.7345,
"step": 1920
},
{
"epoch": 0.17235220575102697,
"grad_norm": 0.40736883878707886,
"learning_rate": 2e-05,
"loss": 0.7146,
"step": 1930
},
{
"epoch": 0.17324522236113593,
"grad_norm": 0.43190640211105347,
"learning_rate": 2e-05,
"loss": 0.7641,
"step": 1940
},
{
"epoch": 0.17413823897124486,
"grad_norm": 0.4052496552467346,
"learning_rate": 2e-05,
"loss": 0.6961,
"step": 1950
},
{
"epoch": 0.17503125558135382,
"grad_norm": 0.3576786518096924,
"learning_rate": 2e-05,
"loss": 0.7191,
"step": 1960
},
{
"epoch": 0.17592427219146275,
"grad_norm": 0.38320717215538025,
"learning_rate": 2e-05,
"loss": 0.7182,
"step": 1970
},
{
"epoch": 0.17681728880157171,
"grad_norm": 0.33529797196388245,
"learning_rate": 2e-05,
"loss": 0.6998,
"step": 1980
},
{
"epoch": 0.17771030541168065,
"grad_norm": 0.4263617992401123,
"learning_rate": 2e-05,
"loss": 0.7237,
"step": 1990
},
{
"epoch": 0.1786033220217896,
"grad_norm": 0.3683549761772156,
"learning_rate": 2e-05,
"loss": 0.7177,
"step": 2000
},
{
"epoch": 0.17949633863189857,
"grad_norm": 0.37825068831443787,
"learning_rate": 2e-05,
"loss": 0.6936,
"step": 2010
},
{
"epoch": 0.1803893552420075,
"grad_norm": 0.38845622539520264,
"learning_rate": 2e-05,
"loss": 0.6886,
"step": 2020
},
{
"epoch": 0.18128237185211646,
"grad_norm": 0.41564154624938965,
"learning_rate": 2e-05,
"loss": 0.6744,
"step": 2030
},
{
"epoch": 0.1821753884622254,
"grad_norm": 0.375347763299942,
"learning_rate": 2e-05,
"loss": 0.6926,
"step": 2040
},
{
"epoch": 0.18306840507233435,
"grad_norm": 0.39925751090049744,
"learning_rate": 2e-05,
"loss": 0.7214,
"step": 2050
},
{
"epoch": 0.18396142168244328,
"grad_norm": 0.3923713266849518,
"learning_rate": 2e-05,
"loss": 0.7478,
"step": 2060
},
{
"epoch": 0.18485443829255224,
"grad_norm": 0.33927154541015625,
"learning_rate": 2e-05,
"loss": 0.7259,
"step": 2070
},
{
"epoch": 0.1857474549026612,
"grad_norm": 0.3772149384021759,
"learning_rate": 2e-05,
"loss": 0.7127,
"step": 2080
},
{
"epoch": 0.18664047151277013,
"grad_norm": 0.4021603763103485,
"learning_rate": 2e-05,
"loss": 0.6927,
"step": 2090
},
{
"epoch": 0.1875334881228791,
"grad_norm": 0.400551974773407,
"learning_rate": 2e-05,
"loss": 0.7256,
"step": 2100
},
{
"epoch": 0.18842650473298803,
"grad_norm": 0.41825857758522034,
"learning_rate": 2e-05,
"loss": 0.6991,
"step": 2110
},
{
"epoch": 0.18931952134309699,
"grad_norm": 0.3584500551223755,
"learning_rate": 2e-05,
"loss": 0.7141,
"step": 2120
},
{
"epoch": 0.19021253795320592,
"grad_norm": 0.8151934146881104,
"learning_rate": 2e-05,
"loss": 0.6834,
"step": 2130
},
{
"epoch": 0.19110555456331488,
"grad_norm": 0.36020809412002563,
"learning_rate": 2e-05,
"loss": 0.7175,
"step": 2140
},
{
"epoch": 0.19199857117342384,
"grad_norm": 0.3733537495136261,
"learning_rate": 2e-05,
"loss": 0.7156,
"step": 2150
},
{
"epoch": 0.19289158778353277,
"grad_norm": 0.4166247844696045,
"learning_rate": 2e-05,
"loss": 0.7636,
"step": 2160
},
{
"epoch": 0.19378460439364173,
"grad_norm": 0.3874996602535248,
"learning_rate": 2e-05,
"loss": 0.699,
"step": 2170
},
{
"epoch": 0.19467762100375066,
"grad_norm": 0.3566342294216156,
"learning_rate": 2e-05,
"loss": 0.7453,
"step": 2180
},
{
"epoch": 0.19557063761385962,
"grad_norm": 0.35978227853775024,
"learning_rate": 2e-05,
"loss": 0.7152,
"step": 2190
},
{
"epoch": 0.19646365422396855,
"grad_norm": 0.34371668100357056,
"learning_rate": 2e-05,
"loss": 0.7258,
"step": 2200
},
{
"epoch": 0.1973566708340775,
"grad_norm": 0.36636587977409363,
"learning_rate": 2e-05,
"loss": 0.7109,
"step": 2210
},
{
"epoch": 0.19824968744418647,
"grad_norm": 0.3694961667060852,
"learning_rate": 2e-05,
"loss": 0.7266,
"step": 2220
},
{
"epoch": 0.1991427040542954,
"grad_norm": 0.3831530809402466,
"learning_rate": 2e-05,
"loss": 0.7165,
"step": 2230
},
{
"epoch": 0.20003572066440437,
"grad_norm": 0.4752442538738251,
"learning_rate": 2e-05,
"loss": 0.7054,
"step": 2240
},
{
"epoch": 0.2009287372745133,
"grad_norm": 0.3605307638645172,
"learning_rate": 2e-05,
"loss": 0.7232,
"step": 2250
},
{
"epoch": 0.20182175388462226,
"grad_norm": 0.3741367757320404,
"learning_rate": 2e-05,
"loss": 0.7164,
"step": 2260
},
{
"epoch": 0.2027147704947312,
"grad_norm": 0.44306308031082153,
"learning_rate": 2e-05,
"loss": 0.7054,
"step": 2270
},
{
"epoch": 0.20360778710484015,
"grad_norm": 0.35885170102119446,
"learning_rate": 2e-05,
"loss": 0.7193,
"step": 2280
},
{
"epoch": 0.2045008037149491,
"grad_norm": 0.42225345969200134,
"learning_rate": 2e-05,
"loss": 0.7167,
"step": 2290
},
{
"epoch": 0.20539382032505804,
"grad_norm": 0.36834093928337097,
"learning_rate": 2e-05,
"loss": 0.699,
"step": 2300
},
{
"epoch": 0.206286836935167,
"grad_norm": 0.37331831455230713,
"learning_rate": 2e-05,
"loss": 0.6864,
"step": 2310
},
{
"epoch": 0.20717985354527593,
"grad_norm": 0.33582958579063416,
"learning_rate": 2e-05,
"loss": 0.7447,
"step": 2320
},
{
"epoch": 0.2080728701553849,
"grad_norm": 0.3406867980957031,
"learning_rate": 2e-05,
"loss": 0.6888,
"step": 2330
},
{
"epoch": 0.20896588676549382,
"grad_norm": 0.7537785768508911,
"learning_rate": 2e-05,
"loss": 0.7387,
"step": 2340
},
{
"epoch": 0.20985890337560278,
"grad_norm": 0.3987394869327545,
"learning_rate": 2e-05,
"loss": 0.6995,
"step": 2350
},
{
"epoch": 0.21075191998571174,
"grad_norm": 0.3712831437587738,
"learning_rate": 2e-05,
"loss": 0.6763,
"step": 2360
},
{
"epoch": 0.21164493659582068,
"grad_norm": 0.3862318992614746,
"learning_rate": 2e-05,
"loss": 0.7021,
"step": 2370
},
{
"epoch": 0.21253795320592964,
"grad_norm": 0.35739755630493164,
"learning_rate": 2e-05,
"loss": 0.7143,
"step": 2380
},
{
"epoch": 0.21343096981603857,
"grad_norm": 0.37976863980293274,
"learning_rate": 2e-05,
"loss": 0.721,
"step": 2390
},
{
"epoch": 0.21432398642614753,
"grad_norm": 0.3507783114910126,
"learning_rate": 2e-05,
"loss": 0.7008,
"step": 2400
},
{
"epoch": 0.2152170030362565,
"grad_norm": 0.37717995047569275,
"learning_rate": 2e-05,
"loss": 0.7123,
"step": 2410
},
{
"epoch": 0.21611001964636542,
"grad_norm": 0.5691996216773987,
"learning_rate": 2e-05,
"loss": 0.7418,
"step": 2420
},
{
"epoch": 0.21700303625647438,
"grad_norm": 0.38756728172302246,
"learning_rate": 2e-05,
"loss": 0.7104,
"step": 2430
},
{
"epoch": 0.2178960528665833,
"grad_norm": 0.554707407951355,
"learning_rate": 2e-05,
"loss": 0.68,
"step": 2440
},
{
"epoch": 0.21878906947669227,
"grad_norm": 0.3568492829799652,
"learning_rate": 2e-05,
"loss": 0.6845,
"step": 2450
},
{
"epoch": 0.2196820860868012,
"grad_norm": 0.3751198947429657,
"learning_rate": 2e-05,
"loss": 0.6868,
"step": 2460
},
{
"epoch": 0.22057510269691016,
"grad_norm": 0.39954832196235657,
"learning_rate": 2e-05,
"loss": 0.7364,
"step": 2470
},
{
"epoch": 0.22146811930701912,
"grad_norm": 0.37244030833244324,
"learning_rate": 2e-05,
"loss": 0.7365,
"step": 2480
},
{
"epoch": 0.22236113591712806,
"grad_norm": 0.38988977670669556,
"learning_rate": 2e-05,
"loss": 0.7324,
"step": 2490
},
{
"epoch": 0.22325415252723702,
"grad_norm": 0.3926638662815094,
"learning_rate": 2e-05,
"loss": 0.7181,
"step": 2500
},
{
"epoch": 0.22414716913734595,
"grad_norm": 0.361802339553833,
"learning_rate": 2e-05,
"loss": 0.6984,
"step": 2510
},
{
"epoch": 0.2250401857474549,
"grad_norm": 0.38118746876716614,
"learning_rate": 2e-05,
"loss": 0.7376,
"step": 2520
},
{
"epoch": 0.22593320235756384,
"grad_norm": 0.39917808771133423,
"learning_rate": 2e-05,
"loss": 0.7166,
"step": 2530
},
{
"epoch": 0.2268262189676728,
"grad_norm": 0.39203181862831116,
"learning_rate": 2e-05,
"loss": 0.7142,
"step": 2540
},
{
"epoch": 0.22771923557778176,
"grad_norm": 0.82660973072052,
"learning_rate": 2e-05,
"loss": 0.6914,
"step": 2550
},
{
"epoch": 0.2286122521878907,
"grad_norm": 0.3741285502910614,
"learning_rate": 2e-05,
"loss": 0.7014,
"step": 2560
},
{
"epoch": 0.22950526879799965,
"grad_norm": 0.3964354395866394,
"learning_rate": 2e-05,
"loss": 0.6982,
"step": 2570
},
{
"epoch": 0.23039828540810858,
"grad_norm": 0.44241467118263245,
"learning_rate": 2e-05,
"loss": 0.7137,
"step": 2580
},
{
"epoch": 0.23129130201821754,
"grad_norm": 0.36581575870513916,
"learning_rate": 2e-05,
"loss": 0.6891,
"step": 2590
},
{
"epoch": 0.23218431862832647,
"grad_norm": 0.3831479549407959,
"learning_rate": 2e-05,
"loss": 0.7266,
"step": 2600
},
{
"epoch": 0.23307733523843543,
"grad_norm": 0.36546680331230164,
"learning_rate": 2e-05,
"loss": 0.7137,
"step": 2610
},
{
"epoch": 0.2339703518485444,
"grad_norm": 0.3919263780117035,
"learning_rate": 2e-05,
"loss": 0.6868,
"step": 2620
},
{
"epoch": 0.23486336845865333,
"grad_norm": 0.37016761302948,
"learning_rate": 2e-05,
"loss": 0.7009,
"step": 2630
},
{
"epoch": 0.2357563850687623,
"grad_norm": 0.37380245327949524,
"learning_rate": 2e-05,
"loss": 0.7025,
"step": 2640
},
{
"epoch": 0.23664940167887122,
"grad_norm": 0.647921085357666,
"learning_rate": 2e-05,
"loss": 0.709,
"step": 2650
},
{
"epoch": 0.23754241828898018,
"grad_norm": 0.3720865845680237,
"learning_rate": 2e-05,
"loss": 0.6891,
"step": 2660
},
{
"epoch": 0.2384354348990891,
"grad_norm": 0.38328051567077637,
"learning_rate": 2e-05,
"loss": 0.7039,
"step": 2670
},
{
"epoch": 0.23932845150919807,
"grad_norm": 0.46826261281967163,
"learning_rate": 2e-05,
"loss": 0.7047,
"step": 2680
},
{
"epoch": 0.24022146811930703,
"grad_norm": 0.37531983852386475,
"learning_rate": 2e-05,
"loss": 0.7185,
"step": 2690
},
{
"epoch": 0.24111448472941596,
"grad_norm": 0.3817111551761627,
"learning_rate": 2e-05,
"loss": 0.7202,
"step": 2700
},
{
"epoch": 0.24200750133952492,
"grad_norm": 0.4027620851993561,
"learning_rate": 2e-05,
"loss": 0.7064,
"step": 2710
},
{
"epoch": 0.24290051794963385,
"grad_norm": 0.3872319459915161,
"learning_rate": 2e-05,
"loss": 0.7163,
"step": 2720
},
{
"epoch": 0.2437935345597428,
"grad_norm": 0.37678542733192444,
"learning_rate": 2e-05,
"loss": 0.6815,
"step": 2730
},
{
"epoch": 0.24468655116985175,
"grad_norm": 0.35330837965011597,
"learning_rate": 2e-05,
"loss": 0.685,
"step": 2740
},
{
"epoch": 0.2455795677799607,
"grad_norm": 0.3747386634349823,
"learning_rate": 2e-05,
"loss": 0.6878,
"step": 2750
},
{
"epoch": 0.24647258439006967,
"grad_norm": 0.41168853640556335,
"learning_rate": 2e-05,
"loss": 0.6705,
"step": 2760
},
{
"epoch": 0.2473656010001786,
"grad_norm": 0.32120653986930847,
"learning_rate": 2e-05,
"loss": 0.6855,
"step": 2770
},
{
"epoch": 0.24825861761028756,
"grad_norm": 0.3390073776245117,
"learning_rate": 2e-05,
"loss": 0.7101,
"step": 2780
},
{
"epoch": 0.2491516342203965,
"grad_norm": 0.37074482440948486,
"learning_rate": 2e-05,
"loss": 0.6986,
"step": 2790
},
{
"epoch": 0.2500446508305054,
"grad_norm": 0.3548227548599243,
"learning_rate": 2e-05,
"loss": 0.6841,
"step": 2800
},
{
"epoch": 0.2509376674406144,
"grad_norm": 0.3685583174228668,
"learning_rate": 2e-05,
"loss": 0.7048,
"step": 2810
},
{
"epoch": 0.25183068405072334,
"grad_norm": 0.3529677391052246,
"learning_rate": 2e-05,
"loss": 0.7216,
"step": 2820
},
{
"epoch": 0.2527237006608323,
"grad_norm": 0.3592994809150696,
"learning_rate": 2e-05,
"loss": 0.6864,
"step": 2830
},
{
"epoch": 0.25361671727094126,
"grad_norm": 0.36516258120536804,
"learning_rate": 2e-05,
"loss": 0.701,
"step": 2840
},
{
"epoch": 0.2545097338810502,
"grad_norm": 0.3898088037967682,
"learning_rate": 2e-05,
"loss": 0.7021,
"step": 2850
},
{
"epoch": 0.2554027504911591,
"grad_norm": 0.34210461378097534,
"learning_rate": 2e-05,
"loss": 0.6915,
"step": 2860
},
{
"epoch": 0.25629576710126806,
"grad_norm": 0.34576666355133057,
"learning_rate": 2e-05,
"loss": 0.704,
"step": 2870
},
{
"epoch": 0.25718878371137704,
"grad_norm": 0.3502766788005829,
"learning_rate": 2e-05,
"loss": 0.7146,
"step": 2880
},
{
"epoch": 0.258081800321486,
"grad_norm": 0.37955331802368164,
"learning_rate": 2e-05,
"loss": 0.7165,
"step": 2890
},
{
"epoch": 0.2589748169315949,
"grad_norm": 0.3612077832221985,
"learning_rate": 2e-05,
"loss": 0.7198,
"step": 2900
},
{
"epoch": 0.2598678335417039,
"grad_norm": 0.3505285084247589,
"learning_rate": 2e-05,
"loss": 0.7248,
"step": 2910
},
{
"epoch": 0.26076085015181283,
"grad_norm": 0.3673901855945587,
"learning_rate": 2e-05,
"loss": 0.7286,
"step": 2920
},
{
"epoch": 0.26165386676192176,
"grad_norm": 0.37945684790611267,
"learning_rate": 2e-05,
"loss": 0.6932,
"step": 2930
},
{
"epoch": 0.2625468833720307,
"grad_norm": 0.3634501099586487,
"learning_rate": 2e-05,
"loss": 0.6931,
"step": 2940
},
{
"epoch": 0.2634398999821397,
"grad_norm": 0.3780888020992279,
"learning_rate": 2e-05,
"loss": 0.7335,
"step": 2950
},
{
"epoch": 0.2643329165922486,
"grad_norm": 0.3801252245903015,
"learning_rate": 2e-05,
"loss": 0.6956,
"step": 2960
},
{
"epoch": 0.26522593320235754,
"grad_norm": 0.36999383568763733,
"learning_rate": 2e-05,
"loss": 0.7009,
"step": 2970
},
{
"epoch": 0.26611894981246653,
"grad_norm": 0.34486889839172363,
"learning_rate": 2e-05,
"loss": 0.6622,
"step": 2980
},
{
"epoch": 0.26701196642257546,
"grad_norm": 0.39070191979408264,
"learning_rate": 2e-05,
"loss": 0.6658,
"step": 2990
},
{
"epoch": 0.2679049830326844,
"grad_norm": 0.37735554575920105,
"learning_rate": 2e-05,
"loss": 0.7351,
"step": 3000
},
{
"epoch": 0.26879799964279333,
"grad_norm": 0.45589423179626465,
"learning_rate": 2e-05,
"loss": 0.7251,
"step": 3010
},
{
"epoch": 0.2696910162529023,
"grad_norm": 0.36697840690612793,
"learning_rate": 2e-05,
"loss": 0.7083,
"step": 3020
},
{
"epoch": 0.27058403286301125,
"grad_norm": 0.6844993233680725,
"learning_rate": 2e-05,
"loss": 0.7378,
"step": 3030
},
{
"epoch": 0.2714770494731202,
"grad_norm": 0.36845651268959045,
"learning_rate": 2e-05,
"loss": 0.7181,
"step": 3040
},
{
"epoch": 0.27237006608322917,
"grad_norm": 0.3911682069301605,
"learning_rate": 2e-05,
"loss": 0.7009,
"step": 3050
},
{
"epoch": 0.2732630826933381,
"grad_norm": 0.3813776671886444,
"learning_rate": 2e-05,
"loss": 0.6905,
"step": 3060
},
{
"epoch": 0.27415609930344703,
"grad_norm": 0.39326590299606323,
"learning_rate": 2e-05,
"loss": 0.7043,
"step": 3070
},
{
"epoch": 0.275049115913556,
"grad_norm": 0.3827492594718933,
"learning_rate": 2e-05,
"loss": 0.7136,
"step": 3080
},
{
"epoch": 0.27594213252366495,
"grad_norm": 0.3630823493003845,
"learning_rate": 2e-05,
"loss": 0.6918,
"step": 3090
},
{
"epoch": 0.2768351491337739,
"grad_norm": 0.3365762233734131,
"learning_rate": 2e-05,
"loss": 0.675,
"step": 3100
},
{
"epoch": 0.2777281657438828,
"grad_norm": 0.39159443974494934,
"learning_rate": 2e-05,
"loss": 0.6669,
"step": 3110
},
{
"epoch": 0.2786211823539918,
"grad_norm": 0.3930347263813019,
"learning_rate": 2e-05,
"loss": 0.7083,
"step": 3120
},
{
"epoch": 0.27951419896410074,
"grad_norm": 0.35313093662261963,
"learning_rate": 2e-05,
"loss": 0.6891,
"step": 3130
},
{
"epoch": 0.28040721557420967,
"grad_norm": 0.33984461426734924,
"learning_rate": 2e-05,
"loss": 0.6683,
"step": 3140
},
{
"epoch": 0.28130023218431865,
"grad_norm": 0.40427762269973755,
"learning_rate": 2e-05,
"loss": 0.7342,
"step": 3150
},
{
"epoch": 0.2821932487944276,
"grad_norm": 0.3635815978050232,
"learning_rate": 2e-05,
"loss": 0.6897,
"step": 3160
},
{
"epoch": 0.2830862654045365,
"grad_norm": 0.41092535853385925,
"learning_rate": 2e-05,
"loss": 0.6868,
"step": 3170
},
{
"epoch": 0.28397928201464545,
"grad_norm": 0.3885297179222107,
"learning_rate": 2e-05,
"loss": 0.6929,
"step": 3180
},
{
"epoch": 0.28487229862475444,
"grad_norm": 0.3909347653388977,
"learning_rate": 2e-05,
"loss": 0.6872,
"step": 3190
},
{
"epoch": 0.28576531523486337,
"grad_norm": 0.34606102108955383,
"learning_rate": 2e-05,
"loss": 0.673,
"step": 3200
},
{
"epoch": 0.2866583318449723,
"grad_norm": 0.3511039614677429,
"learning_rate": 2e-05,
"loss": 0.7187,
"step": 3210
},
{
"epoch": 0.2875513484550813,
"grad_norm": 0.48329827189445496,
"learning_rate": 2e-05,
"loss": 0.7078,
"step": 3220
},
{
"epoch": 0.2884443650651902,
"grad_norm": 0.36455851793289185,
"learning_rate": 2e-05,
"loss": 0.6885,
"step": 3230
},
{
"epoch": 0.28933738167529915,
"grad_norm": 0.36680030822753906,
"learning_rate": 2e-05,
"loss": 0.6744,
"step": 3240
},
{
"epoch": 0.2902303982854081,
"grad_norm": 0.39824068546295166,
"learning_rate": 2e-05,
"loss": 0.7061,
"step": 3250
},
{
"epoch": 0.2911234148955171,
"grad_norm": 0.3565004765987396,
"learning_rate": 2e-05,
"loss": 0.6894,
"step": 3260
},
{
"epoch": 0.292016431505626,
"grad_norm": 0.34405532479286194,
"learning_rate": 2e-05,
"loss": 0.6807,
"step": 3270
},
{
"epoch": 0.29290944811573494,
"grad_norm": 0.3882642090320587,
"learning_rate": 2e-05,
"loss": 0.696,
"step": 3280
},
{
"epoch": 0.2938024647258439,
"grad_norm": 0.3908495008945465,
"learning_rate": 2e-05,
"loss": 0.6932,
"step": 3290
},
{
"epoch": 0.29469548133595286,
"grad_norm": 0.39699220657348633,
"learning_rate": 2e-05,
"loss": 0.7252,
"step": 3300
},
{
"epoch": 0.2955884979460618,
"grad_norm": 0.37945181131362915,
"learning_rate": 2e-05,
"loss": 0.7179,
"step": 3310
},
{
"epoch": 0.2964815145561707,
"grad_norm": 0.3482162058353424,
"learning_rate": 2e-05,
"loss": 0.6994,
"step": 3320
},
{
"epoch": 0.2973745311662797,
"grad_norm": 0.3984408378601074,
"learning_rate": 2e-05,
"loss": 0.7088,
"step": 3330
},
{
"epoch": 0.29826754777638864,
"grad_norm": 0.3412688076496124,
"learning_rate": 2e-05,
"loss": 0.6745,
"step": 3340
},
{
"epoch": 0.2991605643864976,
"grad_norm": 0.34985795617103577,
"learning_rate": 2e-05,
"loss": 0.7055,
"step": 3350
},
{
"epoch": 0.30005358099660656,
"grad_norm": 0.3310694694519043,
"learning_rate": 2e-05,
"loss": 0.6952,
"step": 3360
},
{
"epoch": 0.3009465976067155,
"grad_norm": 0.34946319460868835,
"learning_rate": 2e-05,
"loss": 0.6766,
"step": 3370
},
{
"epoch": 0.3018396142168244,
"grad_norm": 0.36742743849754333,
"learning_rate": 2e-05,
"loss": 0.6916,
"step": 3380
},
{
"epoch": 0.30273263082693336,
"grad_norm": 0.3899418115615845,
"learning_rate": 2e-05,
"loss": 0.7216,
"step": 3390
},
{
"epoch": 0.30362564743704235,
"grad_norm": 0.3610823154449463,
"learning_rate": 2e-05,
"loss": 0.7164,
"step": 3400
},
{
"epoch": 0.3045186640471513,
"grad_norm": 0.36853116750717163,
"learning_rate": 2e-05,
"loss": 0.6832,
"step": 3410
},
{
"epoch": 0.3054116806572602,
"grad_norm": 0.3777369260787964,
"learning_rate": 2e-05,
"loss": 0.7184,
"step": 3420
},
{
"epoch": 0.3063046972673692,
"grad_norm": 0.37218400835990906,
"learning_rate": 2e-05,
"loss": 0.6983,
"step": 3430
},
{
"epoch": 0.30719771387747813,
"grad_norm": 0.39339882135391235,
"learning_rate": 2e-05,
"loss": 0.7022,
"step": 3440
},
{
"epoch": 0.30809073048758706,
"grad_norm": 0.36650851368904114,
"learning_rate": 2e-05,
"loss": 0.6761,
"step": 3450
},
{
"epoch": 0.308983747097696,
"grad_norm": 0.3942379951477051,
"learning_rate": 2e-05,
"loss": 0.6723,
"step": 3460
},
{
"epoch": 0.309876763707805,
"grad_norm": 0.3716135621070862,
"learning_rate": 2e-05,
"loss": 0.7155,
"step": 3470
},
{
"epoch": 0.3107697803179139,
"grad_norm": 0.3720799684524536,
"learning_rate": 2e-05,
"loss": 0.6939,
"step": 3480
},
{
"epoch": 0.31166279692802284,
"grad_norm": 0.35542234778404236,
"learning_rate": 2e-05,
"loss": 0.6962,
"step": 3490
},
{
"epoch": 0.31255581353813183,
"grad_norm": 0.41491395235061646,
"learning_rate": 2e-05,
"loss": 0.7012,
"step": 3500
},
{
"epoch": 0.31344883014824076,
"grad_norm": 0.35415154695510864,
"learning_rate": 2e-05,
"loss": 0.6697,
"step": 3510
},
{
"epoch": 0.3143418467583497,
"grad_norm": 0.4248461425304413,
"learning_rate": 2e-05,
"loss": 0.7122,
"step": 3520
},
{
"epoch": 0.31523486336845863,
"grad_norm": 0.42884528636932373,
"learning_rate": 2e-05,
"loss": 0.7165,
"step": 3530
},
{
"epoch": 0.3161278799785676,
"grad_norm": 0.4538537263870239,
"learning_rate": 2e-05,
"loss": 0.7215,
"step": 3540
},
{
"epoch": 0.31702089658867655,
"grad_norm": 0.39300262928009033,
"learning_rate": 2e-05,
"loss": 0.6708,
"step": 3550
},
{
"epoch": 0.3179139131987855,
"grad_norm": 0.3674524426460266,
"learning_rate": 2e-05,
"loss": 0.6796,
"step": 3560
},
{
"epoch": 0.31880692980889447,
"grad_norm": 0.3366934359073639,
"learning_rate": 2e-05,
"loss": 0.7064,
"step": 3570
},
{
"epoch": 0.3196999464190034,
"grad_norm": 0.46228548884391785,
"learning_rate": 2e-05,
"loss": 0.6933,
"step": 3580
},
{
"epoch": 0.32059296302911233,
"grad_norm": 0.3779584765434265,
"learning_rate": 2e-05,
"loss": 0.668,
"step": 3590
},
{
"epoch": 0.32148597963922126,
"grad_norm": 0.3647182583808899,
"learning_rate": 2e-05,
"loss": 0.6798,
"step": 3600
},
{
"epoch": 0.32237899624933025,
"grad_norm": 0.35239359736442566,
"learning_rate": 2e-05,
"loss": 0.6978,
"step": 3610
},
{
"epoch": 0.3232720128594392,
"grad_norm": 0.34447768330574036,
"learning_rate": 2e-05,
"loss": 0.6887,
"step": 3620
},
{
"epoch": 0.3241650294695481,
"grad_norm": 0.5083383321762085,
"learning_rate": 2e-05,
"loss": 0.7006,
"step": 3630
},
{
"epoch": 0.3250580460796571,
"grad_norm": 0.37443387508392334,
"learning_rate": 2e-05,
"loss": 0.683,
"step": 3640
},
{
"epoch": 0.32595106268976604,
"grad_norm": 0.36775442957878113,
"learning_rate": 2e-05,
"loss": 0.6841,
"step": 3650
},
{
"epoch": 0.32684407929987497,
"grad_norm": 0.43107321858406067,
"learning_rate": 2e-05,
"loss": 0.6962,
"step": 3660
},
{
"epoch": 0.3277370959099839,
"grad_norm": 0.3437502682209015,
"learning_rate": 2e-05,
"loss": 0.6788,
"step": 3670
},
{
"epoch": 0.3286301125200929,
"grad_norm": 0.3835112452507019,
"learning_rate": 2e-05,
"loss": 0.679,
"step": 3680
},
{
"epoch": 0.3295231291302018,
"grad_norm": 0.3627762794494629,
"learning_rate": 2e-05,
"loss": 0.6925,
"step": 3690
},
{
"epoch": 0.33041614574031075,
"grad_norm": 0.36223524808883667,
"learning_rate": 2e-05,
"loss": 0.707,
"step": 3700
},
{
"epoch": 0.33130916235041974,
"grad_norm": 0.380140095949173,
"learning_rate": 2e-05,
"loss": 0.6735,
"step": 3710
},
{
"epoch": 0.33220217896052867,
"grad_norm": 0.3285742998123169,
"learning_rate": 2e-05,
"loss": 0.6674,
"step": 3720
},
{
"epoch": 0.3330951955706376,
"grad_norm": 0.39423131942749023,
"learning_rate": 2e-05,
"loss": 0.7027,
"step": 3730
},
{
"epoch": 0.33398821218074654,
"grad_norm": 0.3929256498813629,
"learning_rate": 2e-05,
"loss": 0.7024,
"step": 3740
},
{
"epoch": 0.3348812287908555,
"grad_norm": 0.3465358018875122,
"learning_rate": 2e-05,
"loss": 0.6796,
"step": 3750
},
{
"epoch": 0.33577424540096446,
"grad_norm": 0.4118633568286896,
"learning_rate": 2e-05,
"loss": 0.7036,
"step": 3760
},
{
"epoch": 0.3366672620110734,
"grad_norm": 0.3681686818599701,
"learning_rate": 2e-05,
"loss": 0.7157,
"step": 3770
},
{
"epoch": 0.3375602786211824,
"grad_norm": 0.4659908413887024,
"learning_rate": 2e-05,
"loss": 0.6902,
"step": 3780
},
{
"epoch": 0.3384532952312913,
"grad_norm": 0.35910534858703613,
"learning_rate": 2e-05,
"loss": 0.7047,
"step": 3790
},
{
"epoch": 0.33934631184140024,
"grad_norm": 0.38420364260673523,
"learning_rate": 2e-05,
"loss": 0.6894,
"step": 3800
},
{
"epoch": 0.34023932845150917,
"grad_norm": 0.4698358178138733,
"learning_rate": 2e-05,
"loss": 0.6686,
"step": 3810
},
{
"epoch": 0.34113234506161816,
"grad_norm": 0.33641931414604187,
"learning_rate": 2e-05,
"loss": 0.7071,
"step": 3820
},
{
"epoch": 0.3420253616717271,
"grad_norm": 0.39272478222846985,
"learning_rate": 2e-05,
"loss": 0.6958,
"step": 3830
},
{
"epoch": 0.342918378281836,
"grad_norm": 0.36490029096603394,
"learning_rate": 2e-05,
"loss": 0.6921,
"step": 3840
},
{
"epoch": 0.343811394891945,
"grad_norm": 0.34400805830955505,
"learning_rate": 2e-05,
"loss": 0.6763,
"step": 3850
},
{
"epoch": 0.34470441150205394,
"grad_norm": 0.3541272282600403,
"learning_rate": 2e-05,
"loss": 0.7135,
"step": 3860
},
{
"epoch": 0.3455974281121629,
"grad_norm": 0.34691932797431946,
"learning_rate": 2e-05,
"loss": 0.6809,
"step": 3870
},
{
"epoch": 0.34649044472227186,
"grad_norm": 0.3615210950374603,
"learning_rate": 2e-05,
"loss": 0.6811,
"step": 3880
},
{
"epoch": 0.3473834613323808,
"grad_norm": 0.37400543689727783,
"learning_rate": 2e-05,
"loss": 0.6974,
"step": 3890
},
{
"epoch": 0.3482764779424897,
"grad_norm": 0.41673043370246887,
"learning_rate": 2e-05,
"loss": 0.7509,
"step": 3900
},
{
"epoch": 0.34916949455259866,
"grad_norm": 0.4203324615955353,
"learning_rate": 2e-05,
"loss": 0.7048,
"step": 3910
},
{
"epoch": 0.35006251116270765,
"grad_norm": 0.3847067058086395,
"learning_rate": 2e-05,
"loss": 0.6853,
"step": 3920
},
{
"epoch": 0.3509555277728166,
"grad_norm": 0.3331018388271332,
"learning_rate": 2e-05,
"loss": 0.6173,
"step": 3930
},
{
"epoch": 0.3518485443829255,
"grad_norm": 0.38789236545562744,
"learning_rate": 2e-05,
"loss": 0.6686,
"step": 3940
},
{
"epoch": 0.3527415609930345,
"grad_norm": 0.34841862320899963,
"learning_rate": 2e-05,
"loss": 0.6713,
"step": 3950
},
{
"epoch": 0.35363457760314343,
"grad_norm": 0.36600327491760254,
"learning_rate": 2e-05,
"loss": 0.6757,
"step": 3960
},
{
"epoch": 0.35452759421325236,
"grad_norm": 0.35073408484458923,
"learning_rate": 2e-05,
"loss": 0.7055,
"step": 3970
},
{
"epoch": 0.3554206108233613,
"grad_norm": 0.33693766593933105,
"learning_rate": 2e-05,
"loss": 0.6924,
"step": 3980
},
{
"epoch": 0.3563136274334703,
"grad_norm": 0.3566618263721466,
"learning_rate": 2e-05,
"loss": 0.6803,
"step": 3990
},
{
"epoch": 0.3572066440435792,
"grad_norm": 0.3572796881198883,
"learning_rate": 2e-05,
"loss": 0.6805,
"step": 4000
},
{
"epoch": 0.35809966065368815,
"grad_norm": 0.40160486102104187,
"learning_rate": 2e-05,
"loss": 0.6705,
"step": 4010
},
{
"epoch": 0.35899267726379713,
"grad_norm": 0.36571747064590454,
"learning_rate": 2e-05,
"loss": 0.6738,
"step": 4020
},
{
"epoch": 0.35988569387390607,
"grad_norm": 0.34542420506477356,
"learning_rate": 2e-05,
"loss": 0.6895,
"step": 4030
},
{
"epoch": 0.360778710484015,
"grad_norm": 0.3404637575149536,
"learning_rate": 2e-05,
"loss": 0.6949,
"step": 4040
},
{
"epoch": 0.36167172709412393,
"grad_norm": 0.4743908643722534,
"learning_rate": 2e-05,
"loss": 0.6849,
"step": 4050
},
{
"epoch": 0.3625647437042329,
"grad_norm": 0.3539956510066986,
"learning_rate": 2e-05,
"loss": 0.6772,
"step": 4060
},
{
"epoch": 0.36345776031434185,
"grad_norm": 0.35375577211380005,
"learning_rate": 2e-05,
"loss": 0.6803,
"step": 4070
},
{
"epoch": 0.3643507769244508,
"grad_norm": 0.37531569600105286,
"learning_rate": 2e-05,
"loss": 0.6856,
"step": 4080
},
{
"epoch": 0.36524379353455977,
"grad_norm": 0.3987765312194824,
"learning_rate": 2e-05,
"loss": 0.6841,
"step": 4090
},
{
"epoch": 0.3661368101446687,
"grad_norm": 0.3849702775478363,
"learning_rate": 2e-05,
"loss": 0.7117,
"step": 4100
},
{
"epoch": 0.36702982675477763,
"grad_norm": 0.364816278219223,
"learning_rate": 2e-05,
"loss": 0.6823,
"step": 4110
},
{
"epoch": 0.36792284336488656,
"grad_norm": 0.372311532497406,
"learning_rate": 2e-05,
"loss": 0.6535,
"step": 4120
},
{
"epoch": 0.36881585997499555,
"grad_norm": 0.33558011054992676,
"learning_rate": 2e-05,
"loss": 0.6663,
"step": 4130
},
{
"epoch": 0.3697088765851045,
"grad_norm": 0.36152487993240356,
"learning_rate": 2e-05,
"loss": 0.7104,
"step": 4140
},
{
"epoch": 0.3706018931952134,
"grad_norm": 0.3753606379032135,
"learning_rate": 2e-05,
"loss": 0.702,
"step": 4150
},
{
"epoch": 0.3714949098053224,
"grad_norm": 0.36609184741973877,
"learning_rate": 2e-05,
"loss": 0.6831,
"step": 4160
},
{
"epoch": 0.37238792641543134,
"grad_norm": 0.34632691740989685,
"learning_rate": 2e-05,
"loss": 0.6892,
"step": 4170
},
{
"epoch": 0.37328094302554027,
"grad_norm": 0.3701765239238739,
"learning_rate": 2e-05,
"loss": 0.6784,
"step": 4180
},
{
"epoch": 0.3741739596356492,
"grad_norm": 0.40695920586586,
"learning_rate": 2e-05,
"loss": 0.668,
"step": 4190
},
{
"epoch": 0.3750669762457582,
"grad_norm": 0.3674159348011017,
"learning_rate": 2e-05,
"loss": 0.6903,
"step": 4200
},
{
"epoch": 0.3759599928558671,
"grad_norm": 0.32752901315689087,
"learning_rate": 2e-05,
"loss": 0.6563,
"step": 4210
},
{
"epoch": 0.37685300946597605,
"grad_norm": 0.3936874568462372,
"learning_rate": 2e-05,
"loss": 0.6835,
"step": 4220
},
{
"epoch": 0.37774602607608504,
"grad_norm": 0.3815043568611145,
"learning_rate": 2e-05,
"loss": 0.7096,
"step": 4230
},
{
"epoch": 0.37863904268619397,
"grad_norm": 0.3459341824054718,
"learning_rate": 2e-05,
"loss": 0.6846,
"step": 4240
},
{
"epoch": 0.3795320592963029,
"grad_norm": 0.39455604553222656,
"learning_rate": 2e-05,
"loss": 0.6611,
"step": 4250
},
{
"epoch": 0.38042507590641184,
"grad_norm": 0.43168190121650696,
"learning_rate": 2e-05,
"loss": 0.6748,
"step": 4260
},
{
"epoch": 0.3813180925165208,
"grad_norm": 0.35696497559547424,
"learning_rate": 2e-05,
"loss": 0.6423,
"step": 4270
},
{
"epoch": 0.38221110912662976,
"grad_norm": 0.36734700202941895,
"learning_rate": 2e-05,
"loss": 0.673,
"step": 4280
},
{
"epoch": 0.3831041257367387,
"grad_norm": 0.3984774053096771,
"learning_rate": 2e-05,
"loss": 0.6811,
"step": 4290
},
{
"epoch": 0.3839971423468477,
"grad_norm": 0.33982765674591064,
"learning_rate": 2e-05,
"loss": 0.6909,
"step": 4300
},
{
"epoch": 0.3848901589569566,
"grad_norm": 0.37864384055137634,
"learning_rate": 2e-05,
"loss": 0.6893,
"step": 4310
},
{
"epoch": 0.38578317556706554,
"grad_norm": 0.36823466420173645,
"learning_rate": 2e-05,
"loss": 0.6905,
"step": 4320
},
{
"epoch": 0.38667619217717447,
"grad_norm": 0.40874573588371277,
"learning_rate": 2e-05,
"loss": 0.6901,
"step": 4330
},
{
"epoch": 0.38756920878728346,
"grad_norm": 0.43925756216049194,
"learning_rate": 2e-05,
"loss": 0.6699,
"step": 4340
},
{
"epoch": 0.3884622253973924,
"grad_norm": 0.4438115656375885,
"learning_rate": 2e-05,
"loss": 0.7074,
"step": 4350
},
{
"epoch": 0.3893552420075013,
"grad_norm": 0.36280399560928345,
"learning_rate": 2e-05,
"loss": 0.6643,
"step": 4360
},
{
"epoch": 0.3902482586176103,
"grad_norm": 0.3471735715866089,
"learning_rate": 2e-05,
"loss": 0.6921,
"step": 4370
},
{
"epoch": 0.39114127522771924,
"grad_norm": 0.3418101370334625,
"learning_rate": 2e-05,
"loss": 0.6394,
"step": 4380
},
{
"epoch": 0.3920342918378282,
"grad_norm": 0.3582875430583954,
"learning_rate": 2e-05,
"loss": 0.7202,
"step": 4390
},
{
"epoch": 0.3929273084479371,
"grad_norm": 0.37227901816368103,
"learning_rate": 2e-05,
"loss": 0.6704,
"step": 4400
},
{
"epoch": 0.3938203250580461,
"grad_norm": 0.3621814548969269,
"learning_rate": 2e-05,
"loss": 0.7006,
"step": 4410
},
{
"epoch": 0.394713341668155,
"grad_norm": 0.35347166657447815,
"learning_rate": 2e-05,
"loss": 0.6776,
"step": 4420
},
{
"epoch": 0.39560635827826396,
"grad_norm": 0.42193514108657837,
"learning_rate": 2e-05,
"loss": 0.6937,
"step": 4430
},
{
"epoch": 0.39649937488837295,
"grad_norm": 0.3808785676956177,
"learning_rate": 2e-05,
"loss": 0.6377,
"step": 4440
},
{
"epoch": 0.3973923914984819,
"grad_norm": 0.37089216709136963,
"learning_rate": 2e-05,
"loss": 0.67,
"step": 4450
},
{
"epoch": 0.3982854081085908,
"grad_norm": 0.3868139982223511,
"learning_rate": 2e-05,
"loss": 0.6919,
"step": 4460
},
{
"epoch": 0.39917842471869974,
"grad_norm": 0.37381669878959656,
"learning_rate": 2e-05,
"loss": 0.7009,
"step": 4470
},
{
"epoch": 0.40007144132880873,
"grad_norm": 0.33514586091041565,
"learning_rate": 2e-05,
"loss": 0.6514,
"step": 4480
},
{
"epoch": 0.40096445793891766,
"grad_norm": 0.34450769424438477,
"learning_rate": 2e-05,
"loss": 0.6847,
"step": 4490
},
{
"epoch": 0.4018574745490266,
"grad_norm": 0.3657180368900299,
"learning_rate": 2e-05,
"loss": 0.6901,
"step": 4500
},
{
"epoch": 0.4027504911591356,
"grad_norm": 0.38325607776641846,
"learning_rate": 2e-05,
"loss": 0.6751,
"step": 4510
},
{
"epoch": 0.4036435077692445,
"grad_norm": 0.3803079128265381,
"learning_rate": 2e-05,
"loss": 0.6819,
"step": 4520
},
{
"epoch": 0.40453652437935345,
"grad_norm": 0.4088026285171509,
"learning_rate": 2e-05,
"loss": 0.6617,
"step": 4530
},
{
"epoch": 0.4054295409894624,
"grad_norm": 0.3512919843196869,
"learning_rate": 2e-05,
"loss": 0.6587,
"step": 4540
},
{
"epoch": 0.40632255759957137,
"grad_norm": 0.34363865852355957,
"learning_rate": 2e-05,
"loss": 0.6863,
"step": 4550
},
{
"epoch": 0.4072155742096803,
"grad_norm": 0.34950166940689087,
"learning_rate": 2e-05,
"loss": 0.6724,
"step": 4560
},
{
"epoch": 0.40810859081978923,
"grad_norm": 0.394996702671051,
"learning_rate": 2e-05,
"loss": 0.6752,
"step": 4570
},
{
"epoch": 0.4090016074298982,
"grad_norm": 0.4020107090473175,
"learning_rate": 2e-05,
"loss": 0.6739,
"step": 4580
},
{
"epoch": 0.40989462404000715,
"grad_norm": 0.5412101745605469,
"learning_rate": 2e-05,
"loss": 0.6879,
"step": 4590
},
{
"epoch": 0.4107876406501161,
"grad_norm": 0.33643534779548645,
"learning_rate": 2e-05,
"loss": 0.6516,
"step": 4600
},
{
"epoch": 0.411680657260225,
"grad_norm": 0.3697690963745117,
"learning_rate": 2e-05,
"loss": 0.6901,
"step": 4610
},
{
"epoch": 0.412573673870334,
"grad_norm": 0.33381566405296326,
"learning_rate": 2e-05,
"loss": 0.6622,
"step": 4620
},
{
"epoch": 0.41346669048044293,
"grad_norm": 0.3048006594181061,
"learning_rate": 2e-05,
"loss": 0.6491,
"step": 4630
},
{
"epoch": 0.41435970709055187,
"grad_norm": 0.35339999198913574,
"learning_rate": 2e-05,
"loss": 0.6826,
"step": 4640
},
{
"epoch": 0.41525272370066085,
"grad_norm": 0.40563100576400757,
"learning_rate": 2e-05,
"loss": 0.6453,
"step": 4650
},
{
"epoch": 0.4161457403107698,
"grad_norm": 0.380941241979599,
"learning_rate": 2e-05,
"loss": 0.6991,
"step": 4660
},
{
"epoch": 0.4170387569208787,
"grad_norm": 0.3585035800933838,
"learning_rate": 2e-05,
"loss": 0.6627,
"step": 4670
},
{
"epoch": 0.41793177353098765,
"grad_norm": 0.3435126841068268,
"learning_rate": 2e-05,
"loss": 0.6869,
"step": 4680
},
{
"epoch": 0.41882479014109664,
"grad_norm": 0.3623499572277069,
"learning_rate": 2e-05,
"loss": 0.6498,
"step": 4690
},
{
"epoch": 0.41971780675120557,
"grad_norm": 0.3764245808124542,
"learning_rate": 2e-05,
"loss": 0.718,
"step": 4700
},
{
"epoch": 0.4206108233613145,
"grad_norm": 0.3425867557525635,
"learning_rate": 2e-05,
"loss": 0.6717,
"step": 4710
},
{
"epoch": 0.4215038399714235,
"grad_norm": 0.3577650189399719,
"learning_rate": 2e-05,
"loss": 0.6627,
"step": 4720
},
{
"epoch": 0.4223968565815324,
"grad_norm": 0.36620646715164185,
"learning_rate": 2e-05,
"loss": 0.6828,
"step": 4730
},
{
"epoch": 0.42328987319164135,
"grad_norm": 0.36670416593551636,
"learning_rate": 2e-05,
"loss": 0.6876,
"step": 4740
},
{
"epoch": 0.42418288980175034,
"grad_norm": 0.3564507067203522,
"learning_rate": 2e-05,
"loss": 0.7026,
"step": 4750
},
{
"epoch": 0.4250759064118593,
"grad_norm": 0.4080749452114105,
"learning_rate": 2e-05,
"loss": 0.6925,
"step": 4760
},
{
"epoch": 0.4259689230219682,
"grad_norm": 0.3509368598461151,
"learning_rate": 2e-05,
"loss": 0.6667,
"step": 4770
},
{
"epoch": 0.42686193963207714,
"grad_norm": 0.3736545443534851,
"learning_rate": 2e-05,
"loss": 0.6771,
"step": 4780
},
{
"epoch": 0.4277549562421861,
"grad_norm": 0.3350389301776886,
"learning_rate": 2e-05,
"loss": 0.6885,
"step": 4790
},
{
"epoch": 0.42864797285229506,
"grad_norm": 0.3636820614337921,
"learning_rate": 2e-05,
"loss": 0.6824,
"step": 4800
},
{
"epoch": 0.429540989462404,
"grad_norm": 0.3685387969017029,
"learning_rate": 2e-05,
"loss": 0.7062,
"step": 4810
},
{
"epoch": 0.430434006072513,
"grad_norm": 0.39025336503982544,
"learning_rate": 2e-05,
"loss": 0.6726,
"step": 4820
},
{
"epoch": 0.4313270226826219,
"grad_norm": 0.3786081075668335,
"learning_rate": 2e-05,
"loss": 0.6905,
"step": 4830
},
{
"epoch": 0.43222003929273084,
"grad_norm": 0.36511704325675964,
"learning_rate": 2e-05,
"loss": 0.669,
"step": 4840
},
{
"epoch": 0.43311305590283977,
"grad_norm": 0.32465630769729614,
"learning_rate": 2e-05,
"loss": 0.6582,
"step": 4850
},
{
"epoch": 0.43400607251294876,
"grad_norm": 0.36687803268432617,
"learning_rate": 2e-05,
"loss": 0.6927,
"step": 4860
},
{
"epoch": 0.4348990891230577,
"grad_norm": 0.35513487458229065,
"learning_rate": 2e-05,
"loss": 0.6798,
"step": 4870
},
{
"epoch": 0.4357921057331666,
"grad_norm": 0.38180556893348694,
"learning_rate": 2e-05,
"loss": 0.66,
"step": 4880
},
{
"epoch": 0.4366851223432756,
"grad_norm": 0.39429476857185364,
"learning_rate": 2e-05,
"loss": 0.6777,
"step": 4890
},
{
"epoch": 0.43757813895338454,
"grad_norm": 0.38248512148857117,
"learning_rate": 2e-05,
"loss": 0.6685,
"step": 4900
},
{
"epoch": 0.4384711555634935,
"grad_norm": 0.3432442545890808,
"learning_rate": 2e-05,
"loss": 0.6747,
"step": 4910
},
{
"epoch": 0.4393641721736024,
"grad_norm": 0.35280707478523254,
"learning_rate": 2e-05,
"loss": 0.6851,
"step": 4920
},
{
"epoch": 0.4402571887837114,
"grad_norm": 0.3475905954837799,
"learning_rate": 2e-05,
"loss": 0.6685,
"step": 4930
},
{
"epoch": 0.4411502053938203,
"grad_norm": 0.3824846148490906,
"learning_rate": 2e-05,
"loss": 0.6698,
"step": 4940
},
{
"epoch": 0.44204322200392926,
"grad_norm": 0.3793005049228668,
"learning_rate": 2e-05,
"loss": 0.6754,
"step": 4950
},
{
"epoch": 0.44293623861403825,
"grad_norm": 0.36466237902641296,
"learning_rate": 2e-05,
"loss": 0.6867,
"step": 4960
},
{
"epoch": 0.4438292552241472,
"grad_norm": 0.3465329110622406,
"learning_rate": 2e-05,
"loss": 0.664,
"step": 4970
},
{
"epoch": 0.4447222718342561,
"grad_norm": 0.3506736159324646,
"learning_rate": 2e-05,
"loss": 0.6785,
"step": 4980
},
{
"epoch": 0.44561528844436504,
"grad_norm": 0.34552523493766785,
"learning_rate": 2e-05,
"loss": 0.6867,
"step": 4990
},
{
"epoch": 0.44650830505447403,
"grad_norm": 0.38686060905456543,
"learning_rate": 2e-05,
"loss": 0.6207,
"step": 5000
},
{
"epoch": 0.44740132166458296,
"grad_norm": 0.3621189296245575,
"learning_rate": 2e-05,
"loss": 0.6745,
"step": 5010
},
{
"epoch": 0.4482943382746919,
"grad_norm": 0.36412402987480164,
"learning_rate": 2e-05,
"loss": 0.6785,
"step": 5020
},
{
"epoch": 0.4491873548848009,
"grad_norm": 0.3481839895248413,
"learning_rate": 2e-05,
"loss": 0.6755,
"step": 5030
},
{
"epoch": 0.4500803714949098,
"grad_norm": 0.3802565336227417,
"learning_rate": 2e-05,
"loss": 0.6705,
"step": 5040
},
{
"epoch": 0.45097338810501875,
"grad_norm": 0.3811277449131012,
"learning_rate": 2e-05,
"loss": 0.6751,
"step": 5050
},
{
"epoch": 0.4518664047151277,
"grad_norm": 0.36518847942352295,
"learning_rate": 2e-05,
"loss": 0.6939,
"step": 5060
},
{
"epoch": 0.45275942132523667,
"grad_norm": 0.3774931728839874,
"learning_rate": 2e-05,
"loss": 0.6897,
"step": 5070
},
{
"epoch": 0.4536524379353456,
"grad_norm": 0.355311781167984,
"learning_rate": 2e-05,
"loss": 0.6793,
"step": 5080
},
{
"epoch": 0.45454545454545453,
"grad_norm": 0.3612700402736664,
"learning_rate": 2e-05,
"loss": 0.6198,
"step": 5090
},
{
"epoch": 0.4554384711555635,
"grad_norm": 0.3959670066833496,
"learning_rate": 2e-05,
"loss": 0.6884,
"step": 5100
},
{
"epoch": 0.45633148776567245,
"grad_norm": 0.7453445792198181,
"learning_rate": 2e-05,
"loss": 0.6794,
"step": 5110
},
{
"epoch": 0.4572245043757814,
"grad_norm": 0.3486466705799103,
"learning_rate": 2e-05,
"loss": 0.6519,
"step": 5120
},
{
"epoch": 0.4581175209858903,
"grad_norm": 0.3771803677082062,
"learning_rate": 2e-05,
"loss": 0.6735,
"step": 5130
},
{
"epoch": 0.4590105375959993,
"grad_norm": 0.33258822560310364,
"learning_rate": 2e-05,
"loss": 0.6714,
"step": 5140
},
{
"epoch": 0.45990355420610823,
"grad_norm": 0.34192121028900146,
"learning_rate": 2e-05,
"loss": 0.6609,
"step": 5150
},
{
"epoch": 0.46079657081621717,
"grad_norm": 0.37189429998397827,
"learning_rate": 2e-05,
"loss": 0.6534,
"step": 5160
},
{
"epoch": 0.46168958742632615,
"grad_norm": 0.3807178735733032,
"learning_rate": 2e-05,
"loss": 0.6621,
"step": 5170
},
{
"epoch": 0.4625826040364351,
"grad_norm": 0.3872874975204468,
"learning_rate": 2e-05,
"loss": 0.6375,
"step": 5180
},
{
"epoch": 0.463475620646544,
"grad_norm": 0.3766045868396759,
"learning_rate": 2e-05,
"loss": 0.6724,
"step": 5190
},
{
"epoch": 0.46436863725665295,
"grad_norm": 0.41265395283699036,
"learning_rate": 2e-05,
"loss": 0.6824,
"step": 5200
},
{
"epoch": 0.46526165386676194,
"grad_norm": 0.38400959968566895,
"learning_rate": 2e-05,
"loss": 0.6743,
"step": 5210
},
{
"epoch": 0.46615467047687087,
"grad_norm": 0.34264642000198364,
"learning_rate": 2e-05,
"loss": 0.6653,
"step": 5220
},
{
"epoch": 0.4670476870869798,
"grad_norm": 0.38337865471839905,
"learning_rate": 2e-05,
"loss": 0.6813,
"step": 5230
},
{
"epoch": 0.4679407036970888,
"grad_norm": 0.3864377737045288,
"learning_rate": 2e-05,
"loss": 0.6665,
"step": 5240
},
{
"epoch": 0.4688337203071977,
"grad_norm": 0.3777919411659241,
"learning_rate": 2e-05,
"loss": 0.6758,
"step": 5250
},
{
"epoch": 0.46972673691730665,
"grad_norm": 0.33093130588531494,
"learning_rate": 2e-05,
"loss": 0.6497,
"step": 5260
},
{
"epoch": 0.4706197535274156,
"grad_norm": 0.33650872111320496,
"learning_rate": 2e-05,
"loss": 0.6789,
"step": 5270
},
{
"epoch": 0.4715127701375246,
"grad_norm": 0.32984471321105957,
"learning_rate": 2e-05,
"loss": 0.6561,
"step": 5280
},
{
"epoch": 0.4724057867476335,
"grad_norm": 0.3533443212509155,
"learning_rate": 2e-05,
"loss": 0.6979,
"step": 5290
},
{
"epoch": 0.47329880335774244,
"grad_norm": 0.3933804929256439,
"learning_rate": 2e-05,
"loss": 0.6506,
"step": 5300
},
{
"epoch": 0.4741918199678514,
"grad_norm": 0.37119993567466736,
"learning_rate": 2e-05,
"loss": 0.6629,
"step": 5310
},
{
"epoch": 0.47508483657796036,
"grad_norm": 0.37831130623817444,
"learning_rate": 2e-05,
"loss": 0.6675,
"step": 5320
},
{
"epoch": 0.4759778531880693,
"grad_norm": 0.3645733594894409,
"learning_rate": 2e-05,
"loss": 0.6593,
"step": 5330
},
{
"epoch": 0.4768708697981782,
"grad_norm": 0.3728773593902588,
"learning_rate": 2e-05,
"loss": 0.6594,
"step": 5340
},
{
"epoch": 0.4777638864082872,
"grad_norm": 0.3517187833786011,
"learning_rate": 2e-05,
"loss": 0.6652,
"step": 5350
},
{
"epoch": 0.47865690301839614,
"grad_norm": 0.3658277094364166,
"learning_rate": 2e-05,
"loss": 0.6902,
"step": 5360
},
{
"epoch": 0.4795499196285051,
"grad_norm": 0.3474698066711426,
"learning_rate": 2e-05,
"loss": 0.6655,
"step": 5370
},
{
"epoch": 0.48044293623861406,
"grad_norm": 0.3642910420894623,
"learning_rate": 2e-05,
"loss": 0.6506,
"step": 5380
},
{
"epoch": 0.481335952848723,
"grad_norm": 0.3575136661529541,
"learning_rate": 2e-05,
"loss": 0.6848,
"step": 5390
},
{
"epoch": 0.4822289694588319,
"grad_norm": 0.3715827167034149,
"learning_rate": 2e-05,
"loss": 0.6527,
"step": 5400
},
{
"epoch": 0.48312198606894086,
"grad_norm": 0.3397136330604553,
"learning_rate": 2e-05,
"loss": 0.6363,
"step": 5410
},
{
"epoch": 0.48401500267904984,
"grad_norm": 0.3929939270019531,
"learning_rate": 2e-05,
"loss": 0.663,
"step": 5420
},
{
"epoch": 0.4849080192891588,
"grad_norm": 0.4231276214122772,
"learning_rate": 2e-05,
"loss": 0.6717,
"step": 5430
},
{
"epoch": 0.4858010358992677,
"grad_norm": 0.35676369071006775,
"learning_rate": 2e-05,
"loss": 0.6851,
"step": 5440
},
{
"epoch": 0.4866940525093767,
"grad_norm": 0.3371965289115906,
"learning_rate": 2e-05,
"loss": 0.6662,
"step": 5450
},
{
"epoch": 0.4875870691194856,
"grad_norm": 0.35079243779182434,
"learning_rate": 2e-05,
"loss": 0.6634,
"step": 5460
},
{
"epoch": 0.48848008572959456,
"grad_norm": 0.3426143229007721,
"learning_rate": 2e-05,
"loss": 0.6855,
"step": 5470
},
{
"epoch": 0.4893731023397035,
"grad_norm": 0.4098520278930664,
"learning_rate": 2e-05,
"loss": 0.6607,
"step": 5480
},
{
"epoch": 0.4902661189498125,
"grad_norm": 0.3547561466693878,
"learning_rate": 2e-05,
"loss": 0.658,
"step": 5490
},
{
"epoch": 0.4911591355599214,
"grad_norm": 0.33377352356910706,
"learning_rate": 2e-05,
"loss": 0.6571,
"step": 5500
},
{
"epoch": 0.49205215217003034,
"grad_norm": 0.33643361926078796,
"learning_rate": 2e-05,
"loss": 0.6766,
"step": 5510
},
{
"epoch": 0.49294516878013933,
"grad_norm": 0.3627377450466156,
"learning_rate": 2e-05,
"loss": 0.6646,
"step": 5520
},
{
"epoch": 0.49383818539024826,
"grad_norm": 0.3409472405910492,
"learning_rate": 2e-05,
"loss": 0.6542,
"step": 5530
},
{
"epoch": 0.4947312020003572,
"grad_norm": 0.3511272668838501,
"learning_rate": 2e-05,
"loss": 0.6734,
"step": 5540
},
{
"epoch": 0.4956242186104661,
"grad_norm": 0.3915940523147583,
"learning_rate": 2e-05,
"loss": 0.6674,
"step": 5550
},
{
"epoch": 0.4965172352205751,
"grad_norm": 0.43016305565834045,
"learning_rate": 2e-05,
"loss": 0.6701,
"step": 5560
},
{
"epoch": 0.49741025183068405,
"grad_norm": 0.3571012616157532,
"learning_rate": 2e-05,
"loss": 0.6907,
"step": 5570
},
{
"epoch": 0.498303268440793,
"grad_norm": 0.43640032410621643,
"learning_rate": 2e-05,
"loss": 0.6803,
"step": 5580
},
{
"epoch": 0.49919628505090197,
"grad_norm": 0.40161386132240295,
"learning_rate": 2e-05,
"loss": 0.6651,
"step": 5590
},
{
"epoch": 0.5000893016610108,
"grad_norm": 0.36353665590286255,
"learning_rate": 2e-05,
"loss": 0.6549,
"step": 5600
},
{
"epoch": 0.5009823182711198,
"grad_norm": 0.3861052691936493,
"learning_rate": 2e-05,
"loss": 0.6679,
"step": 5610
},
{
"epoch": 0.5018753348812288,
"grad_norm": 0.43473106622695923,
"learning_rate": 2e-05,
"loss": 0.6834,
"step": 5620
},
{
"epoch": 0.5027683514913377,
"grad_norm": 0.35996997356414795,
"learning_rate": 2e-05,
"loss": 0.682,
"step": 5630
},
{
"epoch": 0.5036613681014467,
"grad_norm": 0.35466209053993225,
"learning_rate": 2e-05,
"loss": 0.6972,
"step": 5640
},
{
"epoch": 0.5045543847115557,
"grad_norm": 0.4709925055503845,
"learning_rate": 2e-05,
"loss": 0.6961,
"step": 5650
},
{
"epoch": 0.5054474013216645,
"grad_norm": 0.3815552890300751,
"learning_rate": 2e-05,
"loss": 0.6278,
"step": 5660
},
{
"epoch": 0.5063404179317735,
"grad_norm": 0.42588570713996887,
"learning_rate": 2e-05,
"loss": 0.6636,
"step": 5670
},
{
"epoch": 0.5072334345418825,
"grad_norm": 0.3429091274738312,
"learning_rate": 2e-05,
"loss": 0.6406,
"step": 5680
},
{
"epoch": 0.5081264511519914,
"grad_norm": 0.3587746024131775,
"learning_rate": 2e-05,
"loss": 0.6685,
"step": 5690
},
{
"epoch": 0.5090194677621004,
"grad_norm": 0.3580346703529358,
"learning_rate": 2e-05,
"loss": 0.6532,
"step": 5700
},
{
"epoch": 0.5099124843722094,
"grad_norm": 0.36586418747901917,
"learning_rate": 2e-05,
"loss": 0.6657,
"step": 5710
},
{
"epoch": 0.5108055009823183,
"grad_norm": 0.3661688268184662,
"learning_rate": 2e-05,
"loss": 0.6584,
"step": 5720
},
{
"epoch": 0.5116985175924272,
"grad_norm": 0.34977465867996216,
"learning_rate": 2e-05,
"loss": 0.688,
"step": 5730
},
{
"epoch": 0.5125915342025361,
"grad_norm": 0.3609250485897064,
"learning_rate": 2e-05,
"loss": 0.687,
"step": 5740
},
{
"epoch": 0.5134845508126451,
"grad_norm": 0.39893829822540283,
"learning_rate": 2e-05,
"loss": 0.6849,
"step": 5750
},
{
"epoch": 0.5143775674227541,
"grad_norm": 0.3651576340198517,
"learning_rate": 2e-05,
"loss": 0.6697,
"step": 5760
},
{
"epoch": 0.515270584032863,
"grad_norm": 0.3337726891040802,
"learning_rate": 2e-05,
"loss": 0.6753,
"step": 5770
},
{
"epoch": 0.516163600642972,
"grad_norm": 0.3550962805747986,
"learning_rate": 2e-05,
"loss": 0.6536,
"step": 5780
},
{
"epoch": 0.5170566172530809,
"grad_norm": 0.3426112234592438,
"learning_rate": 2e-05,
"loss": 0.6724,
"step": 5790
},
{
"epoch": 0.5179496338631898,
"grad_norm": 0.3464663028717041,
"learning_rate": 2e-05,
"loss": 0.6487,
"step": 5800
},
{
"epoch": 0.5188426504732988,
"grad_norm": 0.36796367168426514,
"learning_rate": 2e-05,
"loss": 0.66,
"step": 5810
},
{
"epoch": 0.5197356670834078,
"grad_norm": 0.3438207507133484,
"learning_rate": 2e-05,
"loss": 0.6729,
"step": 5820
},
{
"epoch": 0.5206286836935167,
"grad_norm": 0.37432757019996643,
"learning_rate": 2e-05,
"loss": 0.6496,
"step": 5830
},
{
"epoch": 0.5215217003036257,
"grad_norm": 0.35678356885910034,
"learning_rate": 2e-05,
"loss": 0.6666,
"step": 5840
},
{
"epoch": 0.5224147169137346,
"grad_norm": 0.37033793330192566,
"learning_rate": 2e-05,
"loss": 0.658,
"step": 5850
},
{
"epoch": 0.5233077335238435,
"grad_norm": 0.36231255531311035,
"learning_rate": 2e-05,
"loss": 0.6687,
"step": 5860
},
{
"epoch": 0.5242007501339525,
"grad_norm": 0.3425047993659973,
"learning_rate": 2e-05,
"loss": 0.6402,
"step": 5870
},
{
"epoch": 0.5250937667440614,
"grad_norm": 0.33811652660369873,
"learning_rate": 2e-05,
"loss": 0.667,
"step": 5880
},
{
"epoch": 0.5259867833541704,
"grad_norm": 0.3571730852127075,
"learning_rate": 2e-05,
"loss": 0.6771,
"step": 5890
},
{
"epoch": 0.5268797999642794,
"grad_norm": 0.37353089451789856,
"learning_rate": 2e-05,
"loss": 0.6471,
"step": 5900
},
{
"epoch": 0.5277728165743882,
"grad_norm": 0.4042361378669739,
"learning_rate": 2e-05,
"loss": 0.6515,
"step": 5910
},
{
"epoch": 0.5286658331844972,
"grad_norm": 0.35085824131965637,
"learning_rate": 2e-05,
"loss": 0.6547,
"step": 5920
},
{
"epoch": 0.5295588497946062,
"grad_norm": 0.3579235374927521,
"learning_rate": 2e-05,
"loss": 0.6557,
"step": 5930
},
{
"epoch": 0.5304518664047151,
"grad_norm": 0.3455997705459595,
"learning_rate": 2e-05,
"loss": 0.6471,
"step": 5940
},
{
"epoch": 0.5313448830148241,
"grad_norm": 0.36393439769744873,
"learning_rate": 2e-05,
"loss": 0.6698,
"step": 5950
},
{
"epoch": 0.5322378996249331,
"grad_norm": 0.3506855368614197,
"learning_rate": 2e-05,
"loss": 0.6517,
"step": 5960
},
{
"epoch": 0.5331309162350419,
"grad_norm": 0.3504025340080261,
"learning_rate": 2e-05,
"loss": 0.6728,
"step": 5970
},
{
"epoch": 0.5340239328451509,
"grad_norm": 0.348411500453949,
"learning_rate": 2e-05,
"loss": 0.6544,
"step": 5980
},
{
"epoch": 0.5349169494552599,
"grad_norm": 0.5087844133377075,
"learning_rate": 2e-05,
"loss": 0.6555,
"step": 5990
},
{
"epoch": 0.5358099660653688,
"grad_norm": 0.3430831730365753,
"learning_rate": 2e-05,
"loss": 0.6972,
"step": 6000
},
{
"epoch": 0.5367029826754778,
"grad_norm": 0.35714656114578247,
"learning_rate": 2e-05,
"loss": 0.6673,
"step": 6010
},
{
"epoch": 0.5375959992855867,
"grad_norm": 0.3810299336910248,
"learning_rate": 2e-05,
"loss": 0.6385,
"step": 6020
},
{
"epoch": 0.5384890158956956,
"grad_norm": 0.3410446047782898,
"learning_rate": 2e-05,
"loss": 0.6812,
"step": 6030
},
{
"epoch": 0.5393820325058046,
"grad_norm": 0.3585295081138611,
"learning_rate": 2e-05,
"loss": 0.6961,
"step": 6040
},
{
"epoch": 0.5402750491159135,
"grad_norm": 0.36647966504096985,
"learning_rate": 2e-05,
"loss": 0.6339,
"step": 6050
},
{
"epoch": 0.5411680657260225,
"grad_norm": 0.36925992369651794,
"learning_rate": 2e-05,
"loss": 0.6375,
"step": 6060
},
{
"epoch": 0.5420610823361315,
"grad_norm": 0.358527272939682,
"learning_rate": 2e-05,
"loss": 0.6527,
"step": 6070
},
{
"epoch": 0.5429540989462404,
"grad_norm": 0.3600867688655853,
"learning_rate": 2e-05,
"loss": 0.6537,
"step": 6080
},
{
"epoch": 0.5438471155563493,
"grad_norm": 0.4136565327644348,
"learning_rate": 2e-05,
"loss": 0.6967,
"step": 6090
},
{
"epoch": 0.5447401321664583,
"grad_norm": 0.36484768986701965,
"learning_rate": 2e-05,
"loss": 0.6527,
"step": 6100
},
{
"epoch": 0.5456331487765672,
"grad_norm": 0.36393529176712036,
"learning_rate": 2e-05,
"loss": 0.6521,
"step": 6110
},
{
"epoch": 0.5465261653866762,
"grad_norm": 0.42837387323379517,
"learning_rate": 2e-05,
"loss": 0.671,
"step": 6120
},
{
"epoch": 0.5474191819967852,
"grad_norm": 0.36187878251075745,
"learning_rate": 2e-05,
"loss": 0.6583,
"step": 6130
},
{
"epoch": 0.5483121986068941,
"grad_norm": 0.5120518207550049,
"learning_rate": 2e-05,
"loss": 0.6511,
"step": 6140
},
{
"epoch": 0.549205215217003,
"grad_norm": 0.39847511053085327,
"learning_rate": 2e-05,
"loss": 0.6486,
"step": 6150
},
{
"epoch": 0.550098231827112,
"grad_norm": 0.3621254861354828,
"learning_rate": 2e-05,
"loss": 0.676,
"step": 6160
},
{
"epoch": 0.5509912484372209,
"grad_norm": 0.39098846912384033,
"learning_rate": 2e-05,
"loss": 0.6609,
"step": 6170
},
{
"epoch": 0.5518842650473299,
"grad_norm": 0.40693438053131104,
"learning_rate": 2e-05,
"loss": 0.6423,
"step": 6180
},
{
"epoch": 0.5527772816574388,
"grad_norm": 0.3641411364078522,
"learning_rate": 2e-05,
"loss": 0.6714,
"step": 6190
},
{
"epoch": 0.5536702982675478,
"grad_norm": 0.420174241065979,
"learning_rate": 2e-05,
"loss": 0.6586,
"step": 6200
},
{
"epoch": 0.5545633148776568,
"grad_norm": 0.3791487216949463,
"learning_rate": 2e-05,
"loss": 0.6986,
"step": 6210
},
{
"epoch": 0.5554563314877656,
"grad_norm": 0.3721368610858917,
"learning_rate": 2e-05,
"loss": 0.6438,
"step": 6220
},
{
"epoch": 0.5563493480978746,
"grad_norm": 0.3461293578147888,
"learning_rate": 2e-05,
"loss": 0.658,
"step": 6230
},
{
"epoch": 0.5572423647079836,
"grad_norm": 0.338593453168869,
"learning_rate": 2e-05,
"loss": 0.6467,
"step": 6240
},
{
"epoch": 0.5581353813180925,
"grad_norm": 0.33129143714904785,
"learning_rate": 2e-05,
"loss": 0.6551,
"step": 6250
},
{
"epoch": 0.5590283979282015,
"grad_norm": 0.3500993251800537,
"learning_rate": 2e-05,
"loss": 0.6878,
"step": 6260
},
{
"epoch": 0.5599214145383105,
"grad_norm": 0.3759903013706207,
"learning_rate": 2e-05,
"loss": 0.6594,
"step": 6270
},
{
"epoch": 0.5608144311484193,
"grad_norm": 0.38059449195861816,
"learning_rate": 2e-05,
"loss": 0.6607,
"step": 6280
},
{
"epoch": 0.5617074477585283,
"grad_norm": 0.3667088747024536,
"learning_rate": 2e-05,
"loss": 0.6491,
"step": 6290
},
{
"epoch": 0.5626004643686373,
"grad_norm": 0.3926546275615692,
"learning_rate": 2e-05,
"loss": 0.6626,
"step": 6300
},
{
"epoch": 0.5634934809787462,
"grad_norm": 0.36864620447158813,
"learning_rate": 2e-05,
"loss": 0.6418,
"step": 6310
},
{
"epoch": 0.5643864975888552,
"grad_norm": 0.36084607243537903,
"learning_rate": 2e-05,
"loss": 0.6695,
"step": 6320
},
{
"epoch": 0.565279514198964,
"grad_norm": 0.34905868768692017,
"learning_rate": 2e-05,
"loss": 0.6487,
"step": 6330
},
{
"epoch": 0.566172530809073,
"grad_norm": 0.3538414239883423,
"learning_rate": 2e-05,
"loss": 0.6722,
"step": 6340
},
{
"epoch": 0.567065547419182,
"grad_norm": 0.3844115734100342,
"learning_rate": 2e-05,
"loss": 0.6523,
"step": 6350
},
{
"epoch": 0.5679585640292909,
"grad_norm": 0.34614577889442444,
"learning_rate": 2e-05,
"loss": 0.6193,
"step": 6360
},
{
"epoch": 0.5688515806393999,
"grad_norm": 0.4562835395336151,
"learning_rate": 2e-05,
"loss": 0.6839,
"step": 6370
},
{
"epoch": 0.5697445972495089,
"grad_norm": 0.37612417340278625,
"learning_rate": 2e-05,
"loss": 0.6779,
"step": 6380
},
{
"epoch": 0.5706376138596178,
"grad_norm": 0.42284253239631653,
"learning_rate": 2e-05,
"loss": 0.6494,
"step": 6390
},
{
"epoch": 0.5715306304697267,
"grad_norm": 0.3244997560977936,
"learning_rate": 2e-05,
"loss": 0.6057,
"step": 6400
},
{
"epoch": 0.5724236470798357,
"grad_norm": 0.34958600997924805,
"learning_rate": 2e-05,
"loss": 0.6495,
"step": 6410
},
{
"epoch": 0.5733166636899446,
"grad_norm": 0.36616605520248413,
"learning_rate": 2e-05,
"loss": 0.6535,
"step": 6420
},
{
"epoch": 0.5742096803000536,
"grad_norm": 0.3466321527957916,
"learning_rate": 2e-05,
"loss": 0.6461,
"step": 6430
},
{
"epoch": 0.5751026969101626,
"grad_norm": 0.35527801513671875,
"learning_rate": 2e-05,
"loss": 0.6684,
"step": 6440
},
{
"epoch": 0.5759957135202715,
"grad_norm": 0.39283958077430725,
"learning_rate": 2e-05,
"loss": 0.6643,
"step": 6450
},
{
"epoch": 0.5768887301303804,
"grad_norm": 0.40512293577194214,
"learning_rate": 2e-05,
"loss": 0.6408,
"step": 6460
},
{
"epoch": 0.5777817467404893,
"grad_norm": 0.39400383830070496,
"learning_rate": 2e-05,
"loss": 0.6629,
"step": 6470
},
{
"epoch": 0.5786747633505983,
"grad_norm": 0.3738294839859009,
"learning_rate": 2e-05,
"loss": 0.6472,
"step": 6480
},
{
"epoch": 0.5795677799607073,
"grad_norm": 0.36141011118888855,
"learning_rate": 2e-05,
"loss": 0.6394,
"step": 6490
},
{
"epoch": 0.5804607965708162,
"grad_norm": 0.3352036774158478,
"learning_rate": 2e-05,
"loss": 0.6332,
"step": 6500
},
{
"epoch": 0.5813538131809252,
"grad_norm": 0.3833242654800415,
"learning_rate": 2e-05,
"loss": 0.6534,
"step": 6510
},
{
"epoch": 0.5822468297910341,
"grad_norm": 0.33181020617485046,
"learning_rate": 2e-05,
"loss": 0.6519,
"step": 6520
},
{
"epoch": 0.583139846401143,
"grad_norm": 0.33593302965164185,
"learning_rate": 2e-05,
"loss": 0.6506,
"step": 6530
},
{
"epoch": 0.584032863011252,
"grad_norm": 0.37441152334213257,
"learning_rate": 2e-05,
"loss": 0.6723,
"step": 6540
},
{
"epoch": 0.584925879621361,
"grad_norm": 0.39254680275917053,
"learning_rate": 2e-05,
"loss": 0.6394,
"step": 6550
},
{
"epoch": 0.5858188962314699,
"grad_norm": 0.37452268600463867,
"learning_rate": 2e-05,
"loss": 0.6821,
"step": 6560
},
{
"epoch": 0.5867119128415789,
"grad_norm": 0.39086517691612244,
"learning_rate": 2e-05,
"loss": 0.661,
"step": 6570
},
{
"epoch": 0.5876049294516879,
"grad_norm": 0.34630292654037476,
"learning_rate": 2e-05,
"loss": 0.6549,
"step": 6580
},
{
"epoch": 0.5884979460617967,
"grad_norm": 0.36574527621269226,
"learning_rate": 2e-05,
"loss": 0.6517,
"step": 6590
},
{
"epoch": 0.5893909626719057,
"grad_norm": 0.3634946942329407,
"learning_rate": 2e-05,
"loss": 0.6525,
"step": 6600
},
{
"epoch": 0.5902839792820146,
"grad_norm": 0.3633476793766022,
"learning_rate": 2e-05,
"loss": 0.6584,
"step": 6610
},
{
"epoch": 0.5911769958921236,
"grad_norm": 0.3540787696838379,
"learning_rate": 2e-05,
"loss": 0.648,
"step": 6620
},
{
"epoch": 0.5920700125022326,
"grad_norm": 0.38122430443763733,
"learning_rate": 2e-05,
"loss": 0.6485,
"step": 6630
},
{
"epoch": 0.5929630291123414,
"grad_norm": 0.3683073818683624,
"learning_rate": 2e-05,
"loss": 0.6718,
"step": 6640
},
{
"epoch": 0.5938560457224504,
"grad_norm": 0.40289875864982605,
"learning_rate": 2e-05,
"loss": 0.68,
"step": 6650
},
{
"epoch": 0.5947490623325594,
"grad_norm": 0.35463404655456543,
"learning_rate": 2e-05,
"loss": 0.6434,
"step": 6660
},
{
"epoch": 0.5956420789426683,
"grad_norm": 0.3831791877746582,
"learning_rate": 2e-05,
"loss": 0.6442,
"step": 6670
},
{
"epoch": 0.5965350955527773,
"grad_norm": 0.3458722233772278,
"learning_rate": 2e-05,
"loss": 0.6619,
"step": 6680
},
{
"epoch": 0.5974281121628863,
"grad_norm": 0.3601379692554474,
"learning_rate": 2e-05,
"loss": 0.6586,
"step": 6690
},
{
"epoch": 0.5983211287729951,
"grad_norm": 0.42295390367507935,
"learning_rate": 2e-05,
"loss": 0.6629,
"step": 6700
},
{
"epoch": 0.5992141453831041,
"grad_norm": 0.3600584864616394,
"learning_rate": 2e-05,
"loss": 0.6604,
"step": 6710
},
{
"epoch": 0.6001071619932131,
"grad_norm": 0.3739805817604065,
"learning_rate": 2e-05,
"loss": 0.6873,
"step": 6720
},
{
"epoch": 0.601000178603322,
"grad_norm": 0.39393553137779236,
"learning_rate": 2e-05,
"loss": 0.6666,
"step": 6730
},
{
"epoch": 0.601893195213431,
"grad_norm": 0.4042629897594452,
"learning_rate": 2e-05,
"loss": 0.6508,
"step": 6740
},
{
"epoch": 0.6027862118235399,
"grad_norm": 0.3513331711292267,
"learning_rate": 2e-05,
"loss": 0.6811,
"step": 6750
},
{
"epoch": 0.6036792284336489,
"grad_norm": 0.345931738615036,
"learning_rate": 2e-05,
"loss": 0.6578,
"step": 6760
},
{
"epoch": 0.6045722450437578,
"grad_norm": 0.3750261068344116,
"learning_rate": 2e-05,
"loss": 0.6143,
"step": 6770
},
{
"epoch": 0.6054652616538667,
"grad_norm": 0.38943830132484436,
"learning_rate": 2e-05,
"loss": 0.6623,
"step": 6780
},
{
"epoch": 0.6063582782639757,
"grad_norm": 0.3595246970653534,
"learning_rate": 2e-05,
"loss": 0.6665,
"step": 6790
},
{
"epoch": 0.6072512948740847,
"grad_norm": 0.3731057643890381,
"learning_rate": 2e-05,
"loss": 0.6332,
"step": 6800
},
{
"epoch": 0.6081443114841936,
"grad_norm": 0.4783044457435608,
"learning_rate": 2e-05,
"loss": 0.6621,
"step": 6810
},
{
"epoch": 0.6090373280943026,
"grad_norm": 0.4021615982055664,
"learning_rate": 2e-05,
"loss": 0.658,
"step": 6820
},
{
"epoch": 0.6099303447044115,
"grad_norm": 0.3629353940486908,
"learning_rate": 2e-05,
"loss": 0.6446,
"step": 6830
},
{
"epoch": 0.6108233613145204,
"grad_norm": 0.41887161135673523,
"learning_rate": 2e-05,
"loss": 0.6712,
"step": 6840
},
{
"epoch": 0.6117163779246294,
"grad_norm": 0.32228341698646545,
"learning_rate": 2e-05,
"loss": 0.6607,
"step": 6850
},
{
"epoch": 0.6126093945347384,
"grad_norm": 0.35648855566978455,
"learning_rate": 2e-05,
"loss": 0.6632,
"step": 6860
},
{
"epoch": 0.6135024111448473,
"grad_norm": 0.36995384097099304,
"learning_rate": 2e-05,
"loss": 0.6475,
"step": 6870
},
{
"epoch": 0.6143954277549563,
"grad_norm": 0.3997064232826233,
"learning_rate": 2e-05,
"loss": 0.6115,
"step": 6880
},
{
"epoch": 0.6152884443650651,
"grad_norm": 0.38138753175735474,
"learning_rate": 2e-05,
"loss": 0.6547,
"step": 6890
},
{
"epoch": 0.6161814609751741,
"grad_norm": 0.3442397713661194,
"learning_rate": 2e-05,
"loss": 0.6439,
"step": 6900
},
{
"epoch": 0.6170744775852831,
"grad_norm": 0.35359567403793335,
"learning_rate": 2e-05,
"loss": 0.6469,
"step": 6910
},
{
"epoch": 0.617967494195392,
"grad_norm": 0.36267122626304626,
"learning_rate": 2e-05,
"loss": 0.6548,
"step": 6920
},
{
"epoch": 0.618860510805501,
"grad_norm": 0.3338748812675476,
"learning_rate": 2e-05,
"loss": 0.7242,
"step": 6930
},
{
"epoch": 0.61975352741561,
"grad_norm": 0.39297038316726685,
"learning_rate": 2e-05,
"loss": 0.6504,
"step": 6940
},
{
"epoch": 0.6206465440257188,
"grad_norm": 0.40164870023727417,
"learning_rate": 2e-05,
"loss": 0.6802,
"step": 6950
},
{
"epoch": 0.6215395606358278,
"grad_norm": 0.33879491686820984,
"learning_rate": 2e-05,
"loss": 0.6506,
"step": 6960
},
{
"epoch": 0.6224325772459368,
"grad_norm": 0.35674700140953064,
"learning_rate": 2e-05,
"loss": 0.6544,
"step": 6970
},
{
"epoch": 0.6233255938560457,
"grad_norm": 0.32872486114501953,
"learning_rate": 2e-05,
"loss": 0.6229,
"step": 6980
},
{
"epoch": 0.6242186104661547,
"grad_norm": 0.3782084882259369,
"learning_rate": 2e-05,
"loss": 0.6871,
"step": 6990
},
{
"epoch": 0.6251116270762637,
"grad_norm": 0.34868884086608887,
"learning_rate": 2e-05,
"loss": 0.6429,
"step": 7000
},
{
"epoch": 0.6260046436863725,
"grad_norm": 0.3346202075481415,
"learning_rate": 2e-05,
"loss": 0.6876,
"step": 7010
},
{
"epoch": 0.6268976602964815,
"grad_norm": 0.36817246675491333,
"learning_rate": 2e-05,
"loss": 0.6421,
"step": 7020
},
{
"epoch": 0.6277906769065905,
"grad_norm": 0.34746721386909485,
"learning_rate": 2e-05,
"loss": 0.6211,
"step": 7030
},
{
"epoch": 0.6286836935166994,
"grad_norm": 0.5210012197494507,
"learning_rate": 2e-05,
"loss": 0.6458,
"step": 7040
},
{
"epoch": 0.6295767101268084,
"grad_norm": 0.3762451112270355,
"learning_rate": 2e-05,
"loss": 0.6492,
"step": 7050
},
{
"epoch": 0.6304697267369173,
"grad_norm": 0.37917590141296387,
"learning_rate": 2e-05,
"loss": 0.6605,
"step": 7060
},
{
"epoch": 0.6313627433470262,
"grad_norm": 0.37249645590782166,
"learning_rate": 2e-05,
"loss": 0.6759,
"step": 7070
},
{
"epoch": 0.6322557599571352,
"grad_norm": 0.36519092321395874,
"learning_rate": 2e-05,
"loss": 0.6406,
"step": 7080
},
{
"epoch": 0.6331487765672441,
"grad_norm": 0.3679816722869873,
"learning_rate": 2e-05,
"loss": 0.6691,
"step": 7090
},
{
"epoch": 0.6340417931773531,
"grad_norm": 0.5378724336624146,
"learning_rate": 2e-05,
"loss": 0.6418,
"step": 7100
},
{
"epoch": 0.6349348097874621,
"grad_norm": 0.3780600130558014,
"learning_rate": 2e-05,
"loss": 0.6594,
"step": 7110
},
{
"epoch": 0.635827826397571,
"grad_norm": 0.378917932510376,
"learning_rate": 2e-05,
"loss": 0.6292,
"step": 7120
},
{
"epoch": 0.63672084300768,
"grad_norm": 0.39086970686912537,
"learning_rate": 2e-05,
"loss": 0.6253,
"step": 7130
},
{
"epoch": 0.6376138596177889,
"grad_norm": 0.39849790930747986,
"learning_rate": 2e-05,
"loss": 0.6717,
"step": 7140
},
{
"epoch": 0.6385068762278978,
"grad_norm": 0.3468756675720215,
"learning_rate": 2e-05,
"loss": 0.6247,
"step": 7150
},
{
"epoch": 0.6393998928380068,
"grad_norm": 0.37662196159362793,
"learning_rate": 2e-05,
"loss": 0.6489,
"step": 7160
},
{
"epoch": 0.6402929094481158,
"grad_norm": 0.3498980700969696,
"learning_rate": 2e-05,
"loss": 0.6676,
"step": 7170
},
{
"epoch": 0.6411859260582247,
"grad_norm": 0.3483699560165405,
"learning_rate": 2e-05,
"loss": 0.6469,
"step": 7180
},
{
"epoch": 0.6420789426683337,
"grad_norm": 0.35714176297187805,
"learning_rate": 2e-05,
"loss": 0.6224,
"step": 7190
},
{
"epoch": 0.6429719592784425,
"grad_norm": 0.3551608920097351,
"learning_rate": 2e-05,
"loss": 0.6724,
"step": 7200
},
{
"epoch": 0.6438649758885515,
"grad_norm": 0.3582582473754883,
"learning_rate": 2e-05,
"loss": 0.6526,
"step": 7210
},
{
"epoch": 0.6447579924986605,
"grad_norm": 0.32978641986846924,
"learning_rate": 2e-05,
"loss": 0.6478,
"step": 7220
},
{
"epoch": 0.6456510091087694,
"grad_norm": 0.3717423677444458,
"learning_rate": 2e-05,
"loss": 0.6646,
"step": 7230
},
{
"epoch": 0.6465440257188784,
"grad_norm": 0.38326701521873474,
"learning_rate": 2e-05,
"loss": 0.6677,
"step": 7240
},
{
"epoch": 0.6474370423289874,
"grad_norm": 0.3557390570640564,
"learning_rate": 2e-05,
"loss": 0.6425,
"step": 7250
},
{
"epoch": 0.6483300589390962,
"grad_norm": 0.3725491464138031,
"learning_rate": 2e-05,
"loss": 0.6271,
"step": 7260
},
{
"epoch": 0.6492230755492052,
"grad_norm": 0.33208444714546204,
"learning_rate": 2e-05,
"loss": 0.648,
"step": 7270
},
{
"epoch": 0.6501160921593142,
"grad_norm": 0.35812655091285706,
"learning_rate": 2e-05,
"loss": 0.6625,
"step": 7280
},
{
"epoch": 0.6510091087694231,
"grad_norm": 0.37675929069519043,
"learning_rate": 2e-05,
"loss": 0.6483,
"step": 7290
},
{
"epoch": 0.6519021253795321,
"grad_norm": 0.37389448285102844,
"learning_rate": 2e-05,
"loss": 0.6877,
"step": 7300
},
{
"epoch": 0.6527951419896411,
"grad_norm": 0.3832888901233673,
"learning_rate": 2e-05,
"loss": 0.6783,
"step": 7310
},
{
"epoch": 0.6536881585997499,
"grad_norm": 0.3738185167312622,
"learning_rate": 2e-05,
"loss": 0.6541,
"step": 7320
},
{
"epoch": 0.6545811752098589,
"grad_norm": 0.34761908650398254,
"learning_rate": 2e-05,
"loss": 0.6367,
"step": 7330
},
{
"epoch": 0.6554741918199678,
"grad_norm": 0.34760138392448425,
"learning_rate": 2e-05,
"loss": 0.6563,
"step": 7340
},
{
"epoch": 0.6563672084300768,
"grad_norm": 0.3394586443901062,
"learning_rate": 2e-05,
"loss": 0.6416,
"step": 7350
},
{
"epoch": 0.6572602250401858,
"grad_norm": 0.3827344477176666,
"learning_rate": 2e-05,
"loss": 0.6559,
"step": 7360
},
{
"epoch": 0.6581532416502947,
"grad_norm": 0.3585679531097412,
"learning_rate": 2e-05,
"loss": 0.6452,
"step": 7370
},
{
"epoch": 0.6590462582604036,
"grad_norm": 0.35019639134407043,
"learning_rate": 2e-05,
"loss": 0.6484,
"step": 7380
},
{
"epoch": 0.6599392748705126,
"grad_norm": 0.38184788823127747,
"learning_rate": 2e-05,
"loss": 0.6593,
"step": 7390
},
{
"epoch": 0.6608322914806215,
"grad_norm": 0.35426050424575806,
"learning_rate": 2e-05,
"loss": 0.6348,
"step": 7400
},
{
"epoch": 0.6617253080907305,
"grad_norm": 0.39316391944885254,
"learning_rate": 2e-05,
"loss": 0.6471,
"step": 7410
},
{
"epoch": 0.6626183247008395,
"grad_norm": 0.3381967842578888,
"learning_rate": 2e-05,
"loss": 0.6581,
"step": 7420
},
{
"epoch": 0.6635113413109484,
"grad_norm": 0.34994620084762573,
"learning_rate": 2e-05,
"loss": 0.6971,
"step": 7430
},
{
"epoch": 0.6644043579210573,
"grad_norm": 0.34434041380882263,
"learning_rate": 2e-05,
"loss": 0.6502,
"step": 7440
},
{
"epoch": 0.6652973745311663,
"grad_norm": 0.39835429191589355,
"learning_rate": 2e-05,
"loss": 0.6682,
"step": 7450
},
{
"epoch": 0.6661903911412752,
"grad_norm": 0.3319467306137085,
"learning_rate": 2e-05,
"loss": 0.6207,
"step": 7460
},
{
"epoch": 0.6670834077513842,
"grad_norm": 0.36640501022338867,
"learning_rate": 2e-05,
"loss": 0.6574,
"step": 7470
},
{
"epoch": 0.6679764243614931,
"grad_norm": 0.3805921971797943,
"learning_rate": 2e-05,
"loss": 0.6326,
"step": 7480
},
{
"epoch": 0.6688694409716021,
"grad_norm": 0.35379254817962646,
"learning_rate": 2e-05,
"loss": 0.6412,
"step": 7490
},
{
"epoch": 0.669762457581711,
"grad_norm": 0.35636794567108154,
"learning_rate": 2e-05,
"loss": 0.644,
"step": 7500
},
{
"epoch": 0.6706554741918199,
"grad_norm": 0.36402857303619385,
"learning_rate": 2e-05,
"loss": 0.6233,
"step": 7510
},
{
"epoch": 0.6715484908019289,
"grad_norm": 0.3685232996940613,
"learning_rate": 2e-05,
"loss": 0.6592,
"step": 7520
},
{
"epoch": 0.6724415074120379,
"grad_norm": 0.38051891326904297,
"learning_rate": 2e-05,
"loss": 0.6678,
"step": 7530
},
{
"epoch": 0.6733345240221468,
"grad_norm": 0.3412979245185852,
"learning_rate": 2e-05,
"loss": 0.6499,
"step": 7540
},
{
"epoch": 0.6742275406322558,
"grad_norm": 0.3998122811317444,
"learning_rate": 2e-05,
"loss": 0.6379,
"step": 7550
},
{
"epoch": 0.6751205572423647,
"grad_norm": 0.3297918736934662,
"learning_rate": 2e-05,
"loss": 0.6474,
"step": 7560
},
{
"epoch": 0.6760135738524736,
"grad_norm": 0.3857010006904602,
"learning_rate": 2e-05,
"loss": 0.6323,
"step": 7570
},
{
"epoch": 0.6769065904625826,
"grad_norm": 0.37469497323036194,
"learning_rate": 2e-05,
"loss": 0.6421,
"step": 7580
},
{
"epoch": 0.6777996070726916,
"grad_norm": 0.3465240001678467,
"learning_rate": 2e-05,
"loss": 0.646,
"step": 7590
},
{
"epoch": 0.6786926236828005,
"grad_norm": 0.364776611328125,
"learning_rate": 2e-05,
"loss": 0.6654,
"step": 7600
},
{
"epoch": 0.6795856402929095,
"grad_norm": 0.3513811528682709,
"learning_rate": 2e-05,
"loss": 0.6279,
"step": 7610
},
{
"epoch": 0.6804786569030183,
"grad_norm": 0.3765827417373657,
"learning_rate": 2e-05,
"loss": 0.6385,
"step": 7620
},
{
"epoch": 0.6813716735131273,
"grad_norm": 0.395460307598114,
"learning_rate": 2e-05,
"loss": 0.6696,
"step": 7630
},
{
"epoch": 0.6822646901232363,
"grad_norm": 0.3784414231777191,
"learning_rate": 2e-05,
"loss": 0.7024,
"step": 7640
},
{
"epoch": 0.6831577067333452,
"grad_norm": 0.3830471932888031,
"learning_rate": 2e-05,
"loss": 0.6501,
"step": 7650
},
{
"epoch": 0.6840507233434542,
"grad_norm": 0.38837364315986633,
"learning_rate": 2e-05,
"loss": 0.6441,
"step": 7660
},
{
"epoch": 0.6849437399535632,
"grad_norm": 0.380912184715271,
"learning_rate": 2e-05,
"loss": 0.6238,
"step": 7670
},
{
"epoch": 0.685836756563672,
"grad_norm": 0.40643543004989624,
"learning_rate": 2e-05,
"loss": 0.6184,
"step": 7680
},
{
"epoch": 0.686729773173781,
"grad_norm": 0.35792428255081177,
"learning_rate": 2e-05,
"loss": 0.6855,
"step": 7690
},
{
"epoch": 0.68762278978389,
"grad_norm": 0.35817697644233704,
"learning_rate": 2e-05,
"loss": 0.6384,
"step": 7700
},
{
"epoch": 0.6885158063939989,
"grad_norm": 0.377506285905838,
"learning_rate": 2e-05,
"loss": 0.6679,
"step": 7710
},
{
"epoch": 0.6894088230041079,
"grad_norm": 0.40117689967155457,
"learning_rate": 2e-05,
"loss": 0.6361,
"step": 7720
},
{
"epoch": 0.6903018396142169,
"grad_norm": 0.3447059392929077,
"learning_rate": 2e-05,
"loss": 0.6367,
"step": 7730
},
{
"epoch": 0.6911948562243257,
"grad_norm": 0.3850513696670532,
"learning_rate": 2e-05,
"loss": 0.6521,
"step": 7740
},
{
"epoch": 0.6920878728344347,
"grad_norm": 0.36426204442977905,
"learning_rate": 2e-05,
"loss": 0.6567,
"step": 7750
},
{
"epoch": 0.6929808894445437,
"grad_norm": 0.37759077548980713,
"learning_rate": 2e-05,
"loss": 0.6569,
"step": 7760
},
{
"epoch": 0.6938739060546526,
"grad_norm": 0.41851291060447693,
"learning_rate": 2e-05,
"loss": 0.6472,
"step": 7770
},
{
"epoch": 0.6947669226647616,
"grad_norm": 0.3735768795013428,
"learning_rate": 2e-05,
"loss": 0.6283,
"step": 7780
},
{
"epoch": 0.6956599392748705,
"grad_norm": 0.38447844982147217,
"learning_rate": 2e-05,
"loss": 0.6457,
"step": 7790
},
{
"epoch": 0.6965529558849795,
"grad_norm": 0.3695277273654938,
"learning_rate": 2e-05,
"loss": 0.6386,
"step": 7800
},
{
"epoch": 0.6974459724950884,
"grad_norm": 0.3485957980155945,
"learning_rate": 2e-05,
"loss": 0.6502,
"step": 7810
},
{
"epoch": 0.6983389891051973,
"grad_norm": 0.45563650131225586,
"learning_rate": 2e-05,
"loss": 0.6724,
"step": 7820
},
{
"epoch": 0.6992320057153063,
"grad_norm": 0.38755181431770325,
"learning_rate": 2e-05,
"loss": 0.6589,
"step": 7830
},
{
"epoch": 0.7001250223254153,
"grad_norm": 0.33155113458633423,
"learning_rate": 2e-05,
"loss": 0.6731,
"step": 7840
},
{
"epoch": 0.7010180389355242,
"grad_norm": 0.367754727602005,
"learning_rate": 2e-05,
"loss": 0.6423,
"step": 7850
},
{
"epoch": 0.7019110555456332,
"grad_norm": 0.32001858949661255,
"learning_rate": 2e-05,
"loss": 0.6656,
"step": 7860
},
{
"epoch": 0.7028040721557421,
"grad_norm": 0.38766756653785706,
"learning_rate": 2e-05,
"loss": 0.628,
"step": 7870
},
{
"epoch": 0.703697088765851,
"grad_norm": 0.4228326082229614,
"learning_rate": 2e-05,
"loss": 0.6394,
"step": 7880
},
{
"epoch": 0.70459010537596,
"grad_norm": 0.370313435792923,
"learning_rate": 2e-05,
"loss": 0.6322,
"step": 7890
},
{
"epoch": 0.705483121986069,
"grad_norm": 0.42004263401031494,
"learning_rate": 2e-05,
"loss": 0.6553,
"step": 7900
},
{
"epoch": 0.7063761385961779,
"grad_norm": 0.3541529178619385,
"learning_rate": 2e-05,
"loss": 0.6511,
"step": 7910
},
{
"epoch": 0.7072691552062869,
"grad_norm": 0.37485307455062866,
"learning_rate": 2e-05,
"loss": 0.6636,
"step": 7920
},
{
"epoch": 0.7081621718163957,
"grad_norm": 0.33004194498062134,
"learning_rate": 2e-05,
"loss": 0.616,
"step": 7930
},
{
"epoch": 0.7090551884265047,
"grad_norm": 0.3753902316093445,
"learning_rate": 2e-05,
"loss": 0.6354,
"step": 7940
},
{
"epoch": 0.7099482050366137,
"grad_norm": 0.37645214796066284,
"learning_rate": 2e-05,
"loss": 0.6556,
"step": 7950
},
{
"epoch": 0.7108412216467226,
"grad_norm": 0.3821773827075958,
"learning_rate": 2e-05,
"loss": 0.628,
"step": 7960
},
{
"epoch": 0.7117342382568316,
"grad_norm": 0.3525018095970154,
"learning_rate": 2e-05,
"loss": 0.6541,
"step": 7970
},
{
"epoch": 0.7126272548669406,
"grad_norm": 0.37279772758483887,
"learning_rate": 2e-05,
"loss": 0.6746,
"step": 7980
},
{
"epoch": 0.7135202714770494,
"grad_norm": 0.35871294140815735,
"learning_rate": 2e-05,
"loss": 0.6306,
"step": 7990
},
{
"epoch": 0.7144132880871584,
"grad_norm": 0.3641284704208374,
"learning_rate": 2e-05,
"loss": 0.672,
"step": 8000
},
{
"epoch": 0.7153063046972674,
"grad_norm": 0.3854856491088867,
"learning_rate": 2e-05,
"loss": 0.6441,
"step": 8010
},
{
"epoch": 0.7161993213073763,
"grad_norm": 0.35918375849723816,
"learning_rate": 2e-05,
"loss": 0.6152,
"step": 8020
},
{
"epoch": 0.7170923379174853,
"grad_norm": 0.3735085725784302,
"learning_rate": 2e-05,
"loss": 0.6425,
"step": 8030
},
{
"epoch": 0.7179853545275943,
"grad_norm": 0.3780219256877899,
"learning_rate": 2e-05,
"loss": 0.6114,
"step": 8040
},
{
"epoch": 0.7188783711377031,
"grad_norm": 0.3764614164829254,
"learning_rate": 2e-05,
"loss": 0.6748,
"step": 8050
},
{
"epoch": 0.7197713877478121,
"grad_norm": 0.3699946403503418,
"learning_rate": 2e-05,
"loss": 0.63,
"step": 8060
},
{
"epoch": 0.720664404357921,
"grad_norm": 0.33643773198127747,
"learning_rate": 2e-05,
"loss": 0.6333,
"step": 8070
},
{
"epoch": 0.72155742096803,
"grad_norm": 0.36439046263694763,
"learning_rate": 2e-05,
"loss": 0.6499,
"step": 8080
},
{
"epoch": 0.722450437578139,
"grad_norm": 0.38973066210746765,
"learning_rate": 2e-05,
"loss": 0.6357,
"step": 8090
},
{
"epoch": 0.7233434541882479,
"grad_norm": 0.37671512365341187,
"learning_rate": 2e-05,
"loss": 0.6345,
"step": 8100
},
{
"epoch": 0.7242364707983568,
"grad_norm": 0.3631742298603058,
"learning_rate": 2e-05,
"loss": 0.6463,
"step": 8110
},
{
"epoch": 0.7251294874084658,
"grad_norm": 0.3698227107524872,
"learning_rate": 2e-05,
"loss": 0.6134,
"step": 8120
},
{
"epoch": 0.7260225040185747,
"grad_norm": 0.36946287751197815,
"learning_rate": 2e-05,
"loss": 0.6196,
"step": 8130
},
{
"epoch": 0.7269155206286837,
"grad_norm": 0.34809666872024536,
"learning_rate": 2e-05,
"loss": 0.6454,
"step": 8140
},
{
"epoch": 0.7278085372387927,
"grad_norm": 0.3508375883102417,
"learning_rate": 2e-05,
"loss": 0.6525,
"step": 8150
},
{
"epoch": 0.7287015538489016,
"grad_norm": 0.36184972524642944,
"learning_rate": 2e-05,
"loss": 0.6527,
"step": 8160
},
{
"epoch": 0.7295945704590105,
"grad_norm": 0.3500848710536957,
"learning_rate": 2e-05,
"loss": 0.6506,
"step": 8170
},
{
"epoch": 0.7304875870691195,
"grad_norm": 0.3669527769088745,
"learning_rate": 2e-05,
"loss": 0.627,
"step": 8180
},
{
"epoch": 0.7313806036792284,
"grad_norm": 0.36657291650772095,
"learning_rate": 2e-05,
"loss": 0.6357,
"step": 8190
},
{
"epoch": 0.7322736202893374,
"grad_norm": 0.3328060507774353,
"learning_rate": 2e-05,
"loss": 0.6764,
"step": 8200
},
{
"epoch": 0.7331666368994463,
"grad_norm": 0.39230093359947205,
"learning_rate": 2e-05,
"loss": 0.6573,
"step": 8210
},
{
"epoch": 0.7340596535095553,
"grad_norm": 0.3294333219528198,
"learning_rate": 2e-05,
"loss": 0.6281,
"step": 8220
},
{
"epoch": 0.7349526701196643,
"grad_norm": 0.39646658301353455,
"learning_rate": 2e-05,
"loss": 0.6313,
"step": 8230
},
{
"epoch": 0.7358456867297731,
"grad_norm": 0.40107738971710205,
"learning_rate": 2e-05,
"loss": 0.675,
"step": 8240
},
{
"epoch": 0.7367387033398821,
"grad_norm": 0.37273871898651123,
"learning_rate": 2e-05,
"loss": 0.6123,
"step": 8250
},
{
"epoch": 0.7376317199499911,
"grad_norm": 0.3266337513923645,
"learning_rate": 2e-05,
"loss": 0.6383,
"step": 8260
},
{
"epoch": 0.7385247365601,
"grad_norm": 0.3725298345088959,
"learning_rate": 2e-05,
"loss": 0.6613,
"step": 8270
},
{
"epoch": 0.739417753170209,
"grad_norm": 0.4053983688354492,
"learning_rate": 2e-05,
"loss": 0.6407,
"step": 8280
},
{
"epoch": 0.740310769780318,
"grad_norm": 0.33429980278015137,
"learning_rate": 2e-05,
"loss": 0.6373,
"step": 8290
},
{
"epoch": 0.7412037863904268,
"grad_norm": 0.3500809669494629,
"learning_rate": 2e-05,
"loss": 0.6401,
"step": 8300
},
{
"epoch": 0.7420968030005358,
"grad_norm": 0.3745899498462677,
"learning_rate": 2e-05,
"loss": 0.6432,
"step": 8310
},
{
"epoch": 0.7429898196106448,
"grad_norm": 0.3830415904521942,
"learning_rate": 2e-05,
"loss": 0.6819,
"step": 8320
},
{
"epoch": 0.7438828362207537,
"grad_norm": 0.3654642105102539,
"learning_rate": 2e-05,
"loss": 0.6343,
"step": 8330
},
{
"epoch": 0.7447758528308627,
"grad_norm": 0.3727662265300751,
"learning_rate": 2e-05,
"loss": 0.6428,
"step": 8340
},
{
"epoch": 0.7456688694409715,
"grad_norm": 0.3408336341381073,
"learning_rate": 2e-05,
"loss": 0.6382,
"step": 8350
},
{
"epoch": 0.7465618860510805,
"grad_norm": 0.34580788016319275,
"learning_rate": 2e-05,
"loss": 0.6513,
"step": 8360
},
{
"epoch": 0.7474549026611895,
"grad_norm": 0.38336580991744995,
"learning_rate": 2e-05,
"loss": 0.6437,
"step": 8370
},
{
"epoch": 0.7483479192712984,
"grad_norm": 0.35544657707214355,
"learning_rate": 2e-05,
"loss": 0.6323,
"step": 8380
},
{
"epoch": 0.7492409358814074,
"grad_norm": 0.3235292434692383,
"learning_rate": 2e-05,
"loss": 0.6494,
"step": 8390
},
{
"epoch": 0.7501339524915164,
"grad_norm": 0.3829690217971802,
"learning_rate": 2e-05,
"loss": 0.6415,
"step": 8400
},
{
"epoch": 0.7510269691016253,
"grad_norm": 0.38451266288757324,
"learning_rate": 2e-05,
"loss": 0.6544,
"step": 8410
},
{
"epoch": 0.7519199857117342,
"grad_norm": 0.34611839056015015,
"learning_rate": 2e-05,
"loss": 0.6162,
"step": 8420
},
{
"epoch": 0.7528130023218432,
"grad_norm": 0.3335798978805542,
"learning_rate": 2e-05,
"loss": 0.5984,
"step": 8430
},
{
"epoch": 0.7537060189319521,
"grad_norm": 0.37450137734413147,
"learning_rate": 2e-05,
"loss": 0.6467,
"step": 8440
},
{
"epoch": 0.7545990355420611,
"grad_norm": 0.36449572443962097,
"learning_rate": 2e-05,
"loss": 0.622,
"step": 8450
},
{
"epoch": 0.7554920521521701,
"grad_norm": 0.3585810363292694,
"learning_rate": 2e-05,
"loss": 0.6334,
"step": 8460
},
{
"epoch": 0.756385068762279,
"grad_norm": 0.3503415286540985,
"learning_rate": 2e-05,
"loss": 0.6503,
"step": 8470
},
{
"epoch": 0.7572780853723879,
"grad_norm": 0.36462873220443726,
"learning_rate": 2e-05,
"loss": 0.6442,
"step": 8480
},
{
"epoch": 0.7581711019824968,
"grad_norm": 0.3520074784755707,
"learning_rate": 2e-05,
"loss": 0.6269,
"step": 8490
},
{
"epoch": 0.7590641185926058,
"grad_norm": 0.3810393512248993,
"learning_rate": 2e-05,
"loss": 0.6084,
"step": 8500
},
{
"epoch": 0.7599571352027148,
"grad_norm": 0.3559550344944,
"learning_rate": 2e-05,
"loss": 0.6584,
"step": 8510
},
{
"epoch": 0.7608501518128237,
"grad_norm": 0.3632529079914093,
"learning_rate": 2e-05,
"loss": 0.6427,
"step": 8520
},
{
"epoch": 0.7617431684229327,
"grad_norm": 0.3899143934249878,
"learning_rate": 2e-05,
"loss": 0.6233,
"step": 8530
},
{
"epoch": 0.7626361850330416,
"grad_norm": 0.35882511734962463,
"learning_rate": 2e-05,
"loss": 0.6282,
"step": 8540
},
{
"epoch": 0.7635292016431505,
"grad_norm": 0.35905560851097107,
"learning_rate": 2e-05,
"loss": 0.6175,
"step": 8550
},
{
"epoch": 0.7644222182532595,
"grad_norm": 0.40037915110588074,
"learning_rate": 2e-05,
"loss": 0.6288,
"step": 8560
},
{
"epoch": 0.7653152348633685,
"grad_norm": 0.3584715723991394,
"learning_rate": 2e-05,
"loss": 0.6622,
"step": 8570
},
{
"epoch": 0.7662082514734774,
"grad_norm": 0.36507561802864075,
"learning_rate": 2e-05,
"loss": 0.622,
"step": 8580
},
{
"epoch": 0.7671012680835864,
"grad_norm": 0.3779599666595459,
"learning_rate": 2e-05,
"loss": 0.6648,
"step": 8590
},
{
"epoch": 0.7679942846936954,
"grad_norm": 0.357250452041626,
"learning_rate": 2e-05,
"loss": 0.6288,
"step": 8600
},
{
"epoch": 0.7688873013038042,
"grad_norm": 0.37146446108818054,
"learning_rate": 2e-05,
"loss": 0.6314,
"step": 8610
},
{
"epoch": 0.7697803179139132,
"grad_norm": 0.3315790891647339,
"learning_rate": 2e-05,
"loss": 0.6313,
"step": 8620
},
{
"epoch": 0.7706733345240222,
"grad_norm": 0.3690609633922577,
"learning_rate": 2e-05,
"loss": 0.6391,
"step": 8630
},
{
"epoch": 0.7715663511341311,
"grad_norm": 0.36267754435539246,
"learning_rate": 2e-05,
"loss": 0.6426,
"step": 8640
},
{
"epoch": 0.7724593677442401,
"grad_norm": 0.35805806517601013,
"learning_rate": 2e-05,
"loss": 0.6545,
"step": 8650
},
{
"epoch": 0.7733523843543489,
"grad_norm": 0.7660143375396729,
"learning_rate": 2e-05,
"loss": 0.6623,
"step": 8660
},
{
"epoch": 0.7742454009644579,
"grad_norm": 0.38856565952301025,
"learning_rate": 2e-05,
"loss": 0.6398,
"step": 8670
},
{
"epoch": 0.7751384175745669,
"grad_norm": 0.3529595732688904,
"learning_rate": 2e-05,
"loss": 0.6375,
"step": 8680
},
{
"epoch": 0.7760314341846758,
"grad_norm": 0.34518077969551086,
"learning_rate": 2e-05,
"loss": 0.6261,
"step": 8690
},
{
"epoch": 0.7769244507947848,
"grad_norm": 0.3565385341644287,
"learning_rate": 2e-05,
"loss": 0.6336,
"step": 8700
},
{
"epoch": 0.7778174674048938,
"grad_norm": 0.3693068325519562,
"learning_rate": 2e-05,
"loss": 0.6312,
"step": 8710
},
{
"epoch": 0.7787104840150026,
"grad_norm": 0.409981906414032,
"learning_rate": 2e-05,
"loss": 0.6316,
"step": 8720
},
{
"epoch": 0.7796035006251116,
"grad_norm": 0.40905171632766724,
"learning_rate": 2e-05,
"loss": 0.6345,
"step": 8730
},
{
"epoch": 0.7804965172352206,
"grad_norm": 0.4453680217266083,
"learning_rate": 2e-05,
"loss": 0.6436,
"step": 8740
},
{
"epoch": 0.7813895338453295,
"grad_norm": 0.37390002608299255,
"learning_rate": 2e-05,
"loss": 0.6147,
"step": 8750
},
{
"epoch": 0.7822825504554385,
"grad_norm": 0.40038150548934937,
"learning_rate": 2e-05,
"loss": 0.6401,
"step": 8760
},
{
"epoch": 0.7831755670655475,
"grad_norm": 0.37419605255126953,
"learning_rate": 2e-05,
"loss": 0.6332,
"step": 8770
},
{
"epoch": 0.7840685836756563,
"grad_norm": 0.3553370237350464,
"learning_rate": 2e-05,
"loss": 0.6436,
"step": 8780
},
{
"epoch": 0.7849616002857653,
"grad_norm": 0.3607825040817261,
"learning_rate": 2e-05,
"loss": 0.6262,
"step": 8790
},
{
"epoch": 0.7858546168958742,
"grad_norm": 0.3781571388244629,
"learning_rate": 2e-05,
"loss": 0.6416,
"step": 8800
},
{
"epoch": 0.7867476335059832,
"grad_norm": 0.39754539728164673,
"learning_rate": 2e-05,
"loss": 0.6405,
"step": 8810
},
{
"epoch": 0.7876406501160922,
"grad_norm": 0.3528376519680023,
"learning_rate": 2e-05,
"loss": 0.6534,
"step": 8820
},
{
"epoch": 0.7885336667262011,
"grad_norm": 0.387594610452652,
"learning_rate": 2e-05,
"loss": 0.6239,
"step": 8830
},
{
"epoch": 0.78942668333631,
"grad_norm": 0.34396862983703613,
"learning_rate": 2e-05,
"loss": 0.6584,
"step": 8840
},
{
"epoch": 0.790319699946419,
"grad_norm": 0.3587445914745331,
"learning_rate": 2e-05,
"loss": 0.6267,
"step": 8850
},
{
"epoch": 0.7912127165565279,
"grad_norm": 0.34561896324157715,
"learning_rate": 2e-05,
"loss": 0.6581,
"step": 8860
},
{
"epoch": 0.7921057331666369,
"grad_norm": 0.353670209646225,
"learning_rate": 2e-05,
"loss": 0.6156,
"step": 8870
},
{
"epoch": 0.7929987497767459,
"grad_norm": 0.3640616536140442,
"learning_rate": 2e-05,
"loss": 0.6532,
"step": 8880
},
{
"epoch": 0.7938917663868548,
"grad_norm": 0.3393080532550812,
"learning_rate": 2e-05,
"loss": 0.6173,
"step": 8890
},
{
"epoch": 0.7947847829969638,
"grad_norm": 0.38873931765556335,
"learning_rate": 2e-05,
"loss": 0.6413,
"step": 8900
},
{
"epoch": 0.7956777996070727,
"grad_norm": 0.3523586094379425,
"learning_rate": 2e-05,
"loss": 0.5893,
"step": 8910
},
{
"epoch": 0.7965708162171816,
"grad_norm": 0.3958098292350769,
"learning_rate": 2e-05,
"loss": 0.6347,
"step": 8920
},
{
"epoch": 0.7974638328272906,
"grad_norm": 0.3697707951068878,
"learning_rate": 2e-05,
"loss": 0.6297,
"step": 8930
},
{
"epoch": 0.7983568494373995,
"grad_norm": 0.32838818430900574,
"learning_rate": 2e-05,
"loss": 0.6459,
"step": 8940
},
{
"epoch": 0.7992498660475085,
"grad_norm": 0.40657302737236023,
"learning_rate": 2e-05,
"loss": 0.6194,
"step": 8950
},
{
"epoch": 0.8001428826576175,
"grad_norm": 0.38677093386650085,
"learning_rate": 2e-05,
"loss": 0.6446,
"step": 8960
},
{
"epoch": 0.8010358992677263,
"grad_norm": 0.36992794275283813,
"learning_rate": 2e-05,
"loss": 0.6386,
"step": 8970
},
{
"epoch": 0.8019289158778353,
"grad_norm": 0.3429109454154968,
"learning_rate": 2e-05,
"loss": 0.6407,
"step": 8980
},
{
"epoch": 0.8028219324879443,
"grad_norm": 0.3723035156726837,
"learning_rate": 2e-05,
"loss": 0.6255,
"step": 8990
},
{
"epoch": 0.8037149490980532,
"grad_norm": 0.32330772280693054,
"learning_rate": 2e-05,
"loss": 0.6384,
"step": 9000
},
{
"epoch": 0.8046079657081622,
"grad_norm": 0.3547590374946594,
"learning_rate": 2e-05,
"loss": 0.6502,
"step": 9010
},
{
"epoch": 0.8055009823182712,
"grad_norm": 0.39275795221328735,
"learning_rate": 2e-05,
"loss": 0.6169,
"step": 9020
},
{
"epoch": 0.80639399892838,
"grad_norm": 0.4927731454372406,
"learning_rate": 2e-05,
"loss": 0.6464,
"step": 9030
},
{
"epoch": 0.807287015538489,
"grad_norm": 0.3671863377094269,
"learning_rate": 2e-05,
"loss": 0.6682,
"step": 9040
},
{
"epoch": 0.808180032148598,
"grad_norm": 0.39080289006233215,
"learning_rate": 2e-05,
"loss": 0.66,
"step": 9050
},
{
"epoch": 0.8090730487587069,
"grad_norm": 0.3432101607322693,
"learning_rate": 2e-05,
"loss": 0.6475,
"step": 9060
},
{
"epoch": 0.8099660653688159,
"grad_norm": 0.42142486572265625,
"learning_rate": 2e-05,
"loss": 0.6273,
"step": 9070
},
{
"epoch": 0.8108590819789248,
"grad_norm": 0.34850892424583435,
"learning_rate": 2e-05,
"loss": 0.6346,
"step": 9080
},
{
"epoch": 0.8117520985890337,
"grad_norm": 0.6122007369995117,
"learning_rate": 2e-05,
"loss": 0.6118,
"step": 9090
},
{
"epoch": 0.8126451151991427,
"grad_norm": 0.4129777252674103,
"learning_rate": 2e-05,
"loss": 0.6168,
"step": 9100
},
{
"epoch": 0.8135381318092516,
"grad_norm": 0.3217753469944,
"learning_rate": 2e-05,
"loss": 0.6338,
"step": 9110
},
{
"epoch": 0.8144311484193606,
"grad_norm": 0.4058667719364166,
"learning_rate": 2e-05,
"loss": 0.6597,
"step": 9120
},
{
"epoch": 0.8153241650294696,
"grad_norm": 0.37627971172332764,
"learning_rate": 2e-05,
"loss": 0.6183,
"step": 9130
},
{
"epoch": 0.8162171816395785,
"grad_norm": 0.382609486579895,
"learning_rate": 2e-05,
"loss": 0.6324,
"step": 9140
},
{
"epoch": 0.8171101982496874,
"grad_norm": 0.35089224576950073,
"learning_rate": 2e-05,
"loss": 0.6507,
"step": 9150
},
{
"epoch": 0.8180032148597964,
"grad_norm": 0.3670852482318878,
"learning_rate": 2e-05,
"loss": 0.6365,
"step": 9160
},
{
"epoch": 0.8188962314699053,
"grad_norm": 0.33786916732788086,
"learning_rate": 2e-05,
"loss": 0.6017,
"step": 9170
},
{
"epoch": 0.8197892480800143,
"grad_norm": 0.3557758629322052,
"learning_rate": 2e-05,
"loss": 0.6289,
"step": 9180
},
{
"epoch": 0.8206822646901233,
"grad_norm": 0.3517480492591858,
"learning_rate": 2e-05,
"loss": 0.6416,
"step": 9190
},
{
"epoch": 0.8215752813002322,
"grad_norm": 0.3295474648475647,
"learning_rate": 2e-05,
"loss": 0.6065,
"step": 9200
},
{
"epoch": 0.8224682979103412,
"grad_norm": 0.3888343572616577,
"learning_rate": 2e-05,
"loss": 0.6401,
"step": 9210
},
{
"epoch": 0.82336131452045,
"grad_norm": 0.36942827701568604,
"learning_rate": 2e-05,
"loss": 0.6403,
"step": 9220
},
{
"epoch": 0.824254331130559,
"grad_norm": 0.3594754636287689,
"learning_rate": 2e-05,
"loss": 0.6086,
"step": 9230
},
{
"epoch": 0.825147347740668,
"grad_norm": 0.4078496992588043,
"learning_rate": 2e-05,
"loss": 0.6093,
"step": 9240
},
{
"epoch": 0.8260403643507769,
"grad_norm": 0.372263640165329,
"learning_rate": 2e-05,
"loss": 0.6446,
"step": 9250
},
{
"epoch": 0.8269333809608859,
"grad_norm": 0.35985904932022095,
"learning_rate": 2e-05,
"loss": 0.6378,
"step": 9260
},
{
"epoch": 0.8278263975709949,
"grad_norm": 0.36655333638191223,
"learning_rate": 2e-05,
"loss": 0.6682,
"step": 9270
},
{
"epoch": 0.8287194141811037,
"grad_norm": 0.369455486536026,
"learning_rate": 2e-05,
"loss": 0.6461,
"step": 9280
},
{
"epoch": 0.8296124307912127,
"grad_norm": 0.3896237909793854,
"learning_rate": 2e-05,
"loss": 0.5976,
"step": 9290
},
{
"epoch": 0.8305054474013217,
"grad_norm": 0.3790488839149475,
"learning_rate": 2e-05,
"loss": 0.6233,
"step": 9300
},
{
"epoch": 0.8313984640114306,
"grad_norm": 0.36344683170318604,
"learning_rate": 2e-05,
"loss": 0.6466,
"step": 9310
},
{
"epoch": 0.8322914806215396,
"grad_norm": 0.36092326045036316,
"learning_rate": 2e-05,
"loss": 0.6327,
"step": 9320
},
{
"epoch": 0.8331844972316486,
"grad_norm": 0.3741254508495331,
"learning_rate": 2e-05,
"loss": 0.6214,
"step": 9330
},
{
"epoch": 0.8340775138417574,
"grad_norm": 0.39041784405708313,
"learning_rate": 2e-05,
"loss": 0.6329,
"step": 9340
},
{
"epoch": 0.8349705304518664,
"grad_norm": 0.3576848804950714,
"learning_rate": 2e-05,
"loss": 0.6408,
"step": 9350
},
{
"epoch": 0.8358635470619753,
"grad_norm": 0.3822234869003296,
"learning_rate": 2e-05,
"loss": 0.6237,
"step": 9360
},
{
"epoch": 0.8367565636720843,
"grad_norm": 0.3773120045661926,
"learning_rate": 2e-05,
"loss": 0.662,
"step": 9370
},
{
"epoch": 0.8376495802821933,
"grad_norm": 0.41020745038986206,
"learning_rate": 2e-05,
"loss": 0.6225,
"step": 9380
},
{
"epoch": 0.8385425968923022,
"grad_norm": 0.35917946696281433,
"learning_rate": 2e-05,
"loss": 0.6467,
"step": 9390
},
{
"epoch": 0.8394356135024111,
"grad_norm": 0.36384376883506775,
"learning_rate": 2e-05,
"loss": 0.6229,
"step": 9400
},
{
"epoch": 0.8403286301125201,
"grad_norm": 0.3806622624397278,
"learning_rate": 2e-05,
"loss": 0.6509,
"step": 9410
},
{
"epoch": 0.841221646722629,
"grad_norm": 0.38903409242630005,
"learning_rate": 2e-05,
"loss": 0.6164,
"step": 9420
},
{
"epoch": 0.842114663332738,
"grad_norm": 0.36390170454978943,
"learning_rate": 2e-05,
"loss": 0.6072,
"step": 9430
},
{
"epoch": 0.843007679942847,
"grad_norm": 0.37405890226364136,
"learning_rate": 2e-05,
"loss": 0.6365,
"step": 9440
},
{
"epoch": 0.8439006965529559,
"grad_norm": 0.3474353551864624,
"learning_rate": 2e-05,
"loss": 0.6334,
"step": 9450
},
{
"epoch": 0.8447937131630648,
"grad_norm": 0.4031187891960144,
"learning_rate": 2e-05,
"loss": 0.6235,
"step": 9460
},
{
"epoch": 0.8456867297731738,
"grad_norm": 0.34947916865348816,
"learning_rate": 2e-05,
"loss": 0.6555,
"step": 9470
},
{
"epoch": 0.8465797463832827,
"grad_norm": 0.3819867968559265,
"learning_rate": 2e-05,
"loss": 0.6521,
"step": 9480
},
{
"epoch": 0.8474727629933917,
"grad_norm": 0.3800252079963684,
"learning_rate": 2e-05,
"loss": 0.5997,
"step": 9490
},
{
"epoch": 0.8483657796035007,
"grad_norm": 0.34814706444740295,
"learning_rate": 2e-05,
"loss": 0.6091,
"step": 9500
},
{
"epoch": 0.8492587962136096,
"grad_norm": 0.3752995729446411,
"learning_rate": 2e-05,
"loss": 0.5996,
"step": 9510
},
{
"epoch": 0.8501518128237185,
"grad_norm": 0.5471505522727966,
"learning_rate": 2e-05,
"loss": 0.6326,
"step": 9520
},
{
"epoch": 0.8510448294338274,
"grad_norm": 0.4234468936920166,
"learning_rate": 2e-05,
"loss": 0.6528,
"step": 9530
},
{
"epoch": 0.8519378460439364,
"grad_norm": 0.38045734167099,
"learning_rate": 2e-05,
"loss": 0.6121,
"step": 9540
},
{
"epoch": 0.8528308626540454,
"grad_norm": 0.3527081310749054,
"learning_rate": 2e-05,
"loss": 0.6269,
"step": 9550
},
{
"epoch": 0.8537238792641543,
"grad_norm": 0.32407793402671814,
"learning_rate": 2e-05,
"loss": 0.6193,
"step": 9560
},
{
"epoch": 0.8546168958742633,
"grad_norm": 0.4124983847141266,
"learning_rate": 2e-05,
"loss": 0.6539,
"step": 9570
},
{
"epoch": 0.8555099124843722,
"grad_norm": 0.3486443758010864,
"learning_rate": 2e-05,
"loss": 0.6255,
"step": 9580
},
{
"epoch": 0.8564029290944811,
"grad_norm": 0.3669477105140686,
"learning_rate": 2e-05,
"loss": 0.6317,
"step": 9590
},
{
"epoch": 0.8572959457045901,
"grad_norm": 0.38984769582748413,
"learning_rate": 2e-05,
"loss": 0.6436,
"step": 9600
},
{
"epoch": 0.8581889623146991,
"grad_norm": 0.3532237410545349,
"learning_rate": 2e-05,
"loss": 0.6419,
"step": 9610
},
{
"epoch": 0.859081978924808,
"grad_norm": 0.3785175085067749,
"learning_rate": 2e-05,
"loss": 0.6455,
"step": 9620
},
{
"epoch": 0.859974995534917,
"grad_norm": 0.3686520755290985,
"learning_rate": 2e-05,
"loss": 0.6652,
"step": 9630
},
{
"epoch": 0.860868012145026,
"grad_norm": 0.36306267976760864,
"learning_rate": 2e-05,
"loss": 0.6096,
"step": 9640
},
{
"epoch": 0.8617610287551348,
"grad_norm": 0.3526270091533661,
"learning_rate": 2e-05,
"loss": 0.6416,
"step": 9650
},
{
"epoch": 0.8626540453652438,
"grad_norm": 0.364076167345047,
"learning_rate": 2e-05,
"loss": 0.6075,
"step": 9660
},
{
"epoch": 0.8635470619753527,
"grad_norm": 0.382335901260376,
"learning_rate": 2e-05,
"loss": 0.6455,
"step": 9670
},
{
"epoch": 0.8644400785854617,
"grad_norm": 0.39734891057014465,
"learning_rate": 2e-05,
"loss": 0.6182,
"step": 9680
},
{
"epoch": 0.8653330951955707,
"grad_norm": 0.3347090482711792,
"learning_rate": 2e-05,
"loss": 0.6132,
"step": 9690
},
{
"epoch": 0.8662261118056795,
"grad_norm": 0.3821641802787781,
"learning_rate": 2e-05,
"loss": 0.6421,
"step": 9700
},
{
"epoch": 0.8671191284157885,
"grad_norm": 0.37654832005500793,
"learning_rate": 2e-05,
"loss": 0.6854,
"step": 9710
},
{
"epoch": 0.8680121450258975,
"grad_norm": 0.45461076498031616,
"learning_rate": 2e-05,
"loss": 0.6116,
"step": 9720
},
{
"epoch": 0.8689051616360064,
"grad_norm": 0.35106271505355835,
"learning_rate": 2e-05,
"loss": 0.6496,
"step": 9730
},
{
"epoch": 0.8697981782461154,
"grad_norm": 0.3606022894382477,
"learning_rate": 2e-05,
"loss": 0.6335,
"step": 9740
},
{
"epoch": 0.8706911948562244,
"grad_norm": 0.3571542799472809,
"learning_rate": 2e-05,
"loss": 0.6369,
"step": 9750
},
{
"epoch": 0.8715842114663332,
"grad_norm": 0.3824685215950012,
"learning_rate": 2e-05,
"loss": 0.6296,
"step": 9760
},
{
"epoch": 0.8724772280764422,
"grad_norm": 0.35880377888679504,
"learning_rate": 2e-05,
"loss": 0.6717,
"step": 9770
},
{
"epoch": 0.8733702446865512,
"grad_norm": 0.3360402584075928,
"learning_rate": 2e-05,
"loss": 0.6306,
"step": 9780
},
{
"epoch": 0.8742632612966601,
"grad_norm": 0.379660964012146,
"learning_rate": 2e-05,
"loss": 0.6304,
"step": 9790
},
{
"epoch": 0.8751562779067691,
"grad_norm": 0.4060397744178772,
"learning_rate": 2e-05,
"loss": 0.6139,
"step": 9800
},
{
"epoch": 0.876049294516878,
"grad_norm": 0.37415921688079834,
"learning_rate": 2e-05,
"loss": 0.6295,
"step": 9810
},
{
"epoch": 0.876942311126987,
"grad_norm": 0.34107649326324463,
"learning_rate": 2e-05,
"loss": 0.6013,
"step": 9820
},
{
"epoch": 0.8778353277370959,
"grad_norm": 0.3327794075012207,
"learning_rate": 2e-05,
"loss": 0.6228,
"step": 9830
},
{
"epoch": 0.8787283443472048,
"grad_norm": 0.42529264092445374,
"learning_rate": 2e-05,
"loss": 0.6079,
"step": 9840
},
{
"epoch": 0.8796213609573138,
"grad_norm": 0.369464635848999,
"learning_rate": 2e-05,
"loss": 0.6256,
"step": 9850
},
{
"epoch": 0.8805143775674228,
"grad_norm": 0.3954029679298401,
"learning_rate": 2e-05,
"loss": 0.5891,
"step": 9860
},
{
"epoch": 0.8814073941775317,
"grad_norm": 0.3886130154132843,
"learning_rate": 2e-05,
"loss": 0.6407,
"step": 9870
},
{
"epoch": 0.8823004107876407,
"grad_norm": 0.3608984649181366,
"learning_rate": 2e-05,
"loss": 0.6195,
"step": 9880
},
{
"epoch": 0.8831934273977496,
"grad_norm": 0.3250412940979004,
"learning_rate": 2e-05,
"loss": 0.6646,
"step": 9890
},
{
"epoch": 0.8840864440078585,
"grad_norm": 0.398472398519516,
"learning_rate": 2e-05,
"loss": 0.6059,
"step": 9900
},
{
"epoch": 0.8849794606179675,
"grad_norm": 0.39279335737228394,
"learning_rate": 2e-05,
"loss": 0.6093,
"step": 9910
},
{
"epoch": 0.8858724772280765,
"grad_norm": 0.38058897852897644,
"learning_rate": 2e-05,
"loss": 0.6388,
"step": 9920
},
{
"epoch": 0.8867654938381854,
"grad_norm": 0.35318806767463684,
"learning_rate": 2e-05,
"loss": 0.6035,
"step": 9930
},
{
"epoch": 0.8876585104482944,
"grad_norm": 0.4062493145465851,
"learning_rate": 2e-05,
"loss": 0.6292,
"step": 9940
},
{
"epoch": 0.8885515270584032,
"grad_norm": 0.3581417202949524,
"learning_rate": 2e-05,
"loss": 0.6399,
"step": 9950
},
{
"epoch": 0.8894445436685122,
"grad_norm": 0.39318400621414185,
"learning_rate": 2e-05,
"loss": 0.6428,
"step": 9960
},
{
"epoch": 0.8903375602786212,
"grad_norm": 0.4005088210105896,
"learning_rate": 2e-05,
"loss": 0.6631,
"step": 9970
},
{
"epoch": 0.8912305768887301,
"grad_norm": 0.3549855947494507,
"learning_rate": 2e-05,
"loss": 0.6514,
"step": 9980
},
{
"epoch": 0.8921235934988391,
"grad_norm": 0.37502971291542053,
"learning_rate": 2e-05,
"loss": 0.6409,
"step": 9990
},
{
"epoch": 0.8930166101089481,
"grad_norm": 0.4062226712703705,
"learning_rate": 2e-05,
"loss": 0.6486,
"step": 10000
},
{
"epoch": 0.8939096267190569,
"grad_norm": 0.3586188554763794,
"learning_rate": 2e-05,
"loss": 0.6401,
"step": 10010
},
{
"epoch": 0.8948026433291659,
"grad_norm": 0.3621748685836792,
"learning_rate": 2e-05,
"loss": 0.6217,
"step": 10020
},
{
"epoch": 0.8956956599392749,
"grad_norm": 0.3818952143192291,
"learning_rate": 2e-05,
"loss": 0.6349,
"step": 10030
},
{
"epoch": 0.8965886765493838,
"grad_norm": 0.38034072518348694,
"learning_rate": 2e-05,
"loss": 0.6104,
"step": 10040
},
{
"epoch": 0.8974816931594928,
"grad_norm": 0.3757418096065521,
"learning_rate": 2e-05,
"loss": 0.6388,
"step": 10050
},
{
"epoch": 0.8983747097696018,
"grad_norm": 0.3484813868999481,
"learning_rate": 2e-05,
"loss": 0.6174,
"step": 10060
},
{
"epoch": 0.8992677263797106,
"grad_norm": 0.46023502945899963,
"learning_rate": 2e-05,
"loss": 0.6346,
"step": 10070
},
{
"epoch": 0.9001607429898196,
"grad_norm": 0.343302845954895,
"learning_rate": 2e-05,
"loss": 0.6409,
"step": 10080
},
{
"epoch": 0.9010537595999285,
"grad_norm": 0.43893858790397644,
"learning_rate": 2e-05,
"loss": 0.6346,
"step": 10090
},
{
"epoch": 0.9019467762100375,
"grad_norm": 0.360633909702301,
"learning_rate": 2e-05,
"loss": 0.5662,
"step": 10100
},
{
"epoch": 0.9028397928201465,
"grad_norm": 0.41140517592430115,
"learning_rate": 2e-05,
"loss": 0.6208,
"step": 10110
},
{
"epoch": 0.9037328094302554,
"grad_norm": 0.41604629158973694,
"learning_rate": 2e-05,
"loss": 0.6441,
"step": 10120
},
{
"epoch": 0.9046258260403643,
"grad_norm": 0.38773632049560547,
"learning_rate": 2e-05,
"loss": 0.6221,
"step": 10130
},
{
"epoch": 0.9055188426504733,
"grad_norm": 0.37810635566711426,
"learning_rate": 2e-05,
"loss": 0.628,
"step": 10140
},
{
"epoch": 0.9064118592605822,
"grad_norm": 0.3822040855884552,
"learning_rate": 2e-05,
"loss": 0.6211,
"step": 10150
},
{
"epoch": 0.9073048758706912,
"grad_norm": 0.3787420690059662,
"learning_rate": 2e-05,
"loss": 0.6514,
"step": 10160
},
{
"epoch": 0.9081978924808002,
"grad_norm": 0.5229046940803528,
"learning_rate": 2e-05,
"loss": 0.6014,
"step": 10170
},
{
"epoch": 0.9090909090909091,
"grad_norm": 0.3930092751979828,
"learning_rate": 1.999957358086795e-05,
"loss": 0.6463,
"step": 10180
},
{
"epoch": 0.909983925701018,
"grad_norm": 0.38098448514938354,
"learning_rate": 1.9991993813020483e-05,
"loss": 0.6465,
"step": 10190
},
{
"epoch": 0.910876942311127,
"grad_norm": 0.33956673741340637,
"learning_rate": 1.9974946373028628e-05,
"loss": 0.6126,
"step": 10200
},
{
"epoch": 0.9117699589212359,
"grad_norm": 0.3713693618774414,
"learning_rate": 1.994844749501764e-05,
"loss": 0.5829,
"step": 10210
},
{
"epoch": 0.9126629755313449,
"grad_norm": 0.3689557909965515,
"learning_rate": 1.9912522413632988e-05,
"loss": 0.6438,
"step": 10220
},
{
"epoch": 0.9135559921414538,
"grad_norm": 0.3251681625843048,
"learning_rate": 1.9867205340009656e-05,
"loss": 0.6152,
"step": 10230
},
{
"epoch": 0.9144490087515628,
"grad_norm": 0.36217454075813293,
"learning_rate": 1.9812539429193164e-05,
"loss": 0.6124,
"step": 10240
},
{
"epoch": 0.9153420253616718,
"grad_norm": 0.36874791979789734,
"learning_rate": 1.9748576739043423e-05,
"loss": 0.6247,
"step": 10250
},
{
"epoch": 0.9162350419717806,
"grad_norm": 0.37588948011398315,
"learning_rate": 1.967537818066047e-05,
"loss": 0.6162,
"step": 10260
},
{
"epoch": 0.9171280585818896,
"grad_norm": 0.38607147336006165,
"learning_rate": 1.959301346037941e-05,
"loss": 0.6533,
"step": 10270
},
{
"epoch": 0.9180210751919986,
"grad_norm": 0.3462531566619873,
"learning_rate": 1.9501561013389646e-05,
"loss": 0.6117,
"step": 10280
},
{
"epoch": 0.9189140918021075,
"grad_norm": 0.35778751969337463,
"learning_rate": 1.9401107929041795e-05,
"loss": 0.6223,
"step": 10290
},
{
"epoch": 0.9198071084122165,
"grad_norm": 0.39281153678894043,
"learning_rate": 1.9291749867913225e-05,
"loss": 0.6281,
"step": 10300
},
{
"epoch": 0.9207001250223255,
"grad_norm": 0.358835369348526,
"learning_rate": 1.91735909707114e-05,
"loss": 0.6237,
"step": 10310
},
{
"epoch": 0.9215931416324343,
"grad_norm": 0.35125985741615295,
"learning_rate": 1.9046743759101545e-05,
"loss": 0.6402,
"step": 10320
},
{
"epoch": 0.9224861582425433,
"grad_norm": 0.3486235737800598,
"learning_rate": 1.8911329028553324e-05,
"loss": 0.6294,
"step": 10330
},
{
"epoch": 0.9233791748526523,
"grad_norm": 0.3921213448047638,
"learning_rate": 1.876747573330835e-05,
"loss": 0.603,
"step": 10340
},
{
"epoch": 0.9242721914627612,
"grad_norm": 0.39618730545043945,
"learning_rate": 1.8615320863578244e-05,
"loss": 0.6136,
"step": 10350
},
{
"epoch": 0.9251652080728702,
"grad_norm": 0.3554159104824066,
"learning_rate": 1.8455009315090036e-05,
"loss": 0.6505,
"step": 10360
},
{
"epoch": 0.9260582246829792,
"grad_norm": 0.3428751230239868,
"learning_rate": 1.8286693751103294e-05,
"loss": 0.6217,
"step": 10370
},
{
"epoch": 0.926951241293088,
"grad_norm": 0.3795991539955139,
"learning_rate": 1.811053445703024e-05,
"loss": 0.5984,
"step": 10380
},
{
"epoch": 0.927844257903197,
"grad_norm": 0.404371440410614,
"learning_rate": 1.7926699187797375e-05,
"loss": 0.6431,
"step": 10390
},
{
"epoch": 0.9287372745133059,
"grad_norm": 0.38412851095199585,
"learning_rate": 1.7735363008093995e-05,
"loss": 0.6275,
"step": 10400
},
{
"epoch": 0.9296302911234149,
"grad_norm": 0.3620508015155792,
"learning_rate": 1.7536708125659577e-05,
"loss": 0.6119,
"step": 10410
},
{
"epoch": 0.9305233077335239,
"grad_norm": 0.36473074555397034,
"learning_rate": 1.733092371776905e-05,
"loss": 0.6076,
"step": 10420
},
{
"epoch": 0.9314163243436328,
"grad_norm": 0.36890533566474915,
"learning_rate": 1.7118205751080915e-05,
"loss": 0.6323,
"step": 10430
},
{
"epoch": 0.9323093409537417,
"grad_norm": 0.4714130759239197,
"learning_rate": 1.6898756795019935e-05,
"loss": 0.607,
"step": 10440
},
{
"epoch": 0.9332023575638507,
"grad_norm": 0.36841025948524475,
"learning_rate": 1.667278582887211e-05,
"loss": 0.6461,
"step": 10450
},
{
"epoch": 0.9340953741739596,
"grad_norm": 0.35519856214523315,
"learning_rate": 1.6440508042775545e-05,
"loss": 0.6141,
"step": 10460
},
{
"epoch": 0.9349883907840686,
"grad_norm": 0.436336487531662,
"learning_rate": 1.6202144632796796e-05,
"loss": 0.6098,
"step": 10470
},
{
"epoch": 0.9358814073941776,
"grad_norm": 0.3691329061985016,
"learning_rate": 1.595792259028783e-05,
"loss": 0.6287,
"step": 10480
},
{
"epoch": 0.9367744240042865,
"grad_norm": 0.3849637508392334,
"learning_rate": 1.570807448572419e-05,
"loss": 0.6638,
"step": 10490
},
{
"epoch": 0.9376674406143954,
"grad_norm": 0.35689228773117065,
"learning_rate": 1.5452838247230197e-05,
"loss": 0.6267,
"step": 10500
},
{
"epoch": 0.9385604572245044,
"grad_norm": 0.5085926651954651,
"learning_rate": 1.5192456934002106e-05,
"loss": 0.6335,
"step": 10510
},
{
"epoch": 0.9394534738346133,
"grad_norm": 0.36765962839126587,
"learning_rate": 1.492717850484502e-05,
"loss": 0.6047,
"step": 10520
},
{
"epoch": 0.9403464904447223,
"grad_norm": 0.35301855206489563,
"learning_rate": 1.4657255582043904e-05,
"loss": 0.6254,
"step": 10530
},
{
"epoch": 0.9412395070548312,
"grad_norm": 0.3616034686565399,
"learning_rate": 1.4382945210793663e-05,
"loss": 0.6169,
"step": 10540
},
{
"epoch": 0.9421325236649402,
"grad_norm": 0.35207030177116394,
"learning_rate": 1.4104508614417267e-05,
"loss": 0.6379,
"step": 10550
},
{
"epoch": 0.9430255402750491,
"grad_norm": 0.3955402076244354,
"learning_rate": 1.382221094560512e-05,
"loss": 0.6207,
"step": 10560
},
{
"epoch": 0.943918556885158,
"grad_norm": 0.39667633175849915,
"learning_rate": 1.3536321033912506e-05,
"loss": 0.6364,
"step": 10570
},
{
"epoch": 0.944811573495267,
"grad_norm": 0.33852139115333557,
"learning_rate": 1.3247111129755609e-05,
"loss": 0.6207,
"step": 10580
},
{
"epoch": 0.945704590105376,
"grad_norm": 0.35518282651901245,
"learning_rate": 1.2954856645149868e-05,
"loss": 0.6121,
"step": 10590
},
{
"epoch": 0.9465976067154849,
"grad_norm": 0.341487854719162,
"learning_rate": 1.2659835891437553e-05,
"loss": 0.6389,
"step": 10600
},
{
"epoch": 0.9474906233255939,
"grad_norm": 0.376911461353302,
"learning_rate": 1.236232981425436e-05,
"loss": 0.597,
"step": 10610
},
{
"epoch": 0.9483836399357028,
"grad_norm": 0.40024319291114807,
"learning_rate": 1.2062621725987402e-05,
"loss": 0.6053,
"step": 10620
},
{
"epoch": 0.9492766565458117,
"grad_norm": 0.3672933876514435,
"learning_rate": 1.1760997035979318e-05,
"loss": 0.6155,
"step": 10630
},
{
"epoch": 0.9501696731559207,
"grad_norm": 0.3369959592819214,
"learning_rate": 1.1457742978735514e-05,
"loss": 0.6034,
"step": 10640
},
{
"epoch": 0.9510626897660297,
"grad_norm": 0.36264780163764954,
"learning_rate": 1.1153148340393298e-05,
"loss": 0.5998,
"step": 10650
},
{
"epoch": 0.9519557063761386,
"grad_norm": 0.37920981645584106,
"learning_rate": 1.0847503183713424e-05,
"loss": 0.5908,
"step": 10660
},
{
"epoch": 0.9528487229862476,
"grad_norm": 0.42272019386291504,
"learning_rate": 1.0541098571855896e-05,
"loss": 0.6021,
"step": 10670
},
{
"epoch": 0.9537417395963564,
"grad_norm": 0.3429796099662781,
"learning_rate": 1.0234226291203123e-05,
"loss": 0.5707,
"step": 10680
},
{
"epoch": 0.9546347562064654,
"grad_norm": 0.3957381546497345,
"learning_rate": 9.927178573494333e-06,
"loss": 0.615,
"step": 10690
},
{
"epoch": 0.9555277728165744,
"grad_norm": 0.3588882088661194,
"learning_rate": 9.620247817535902e-06,
"loss": 0.6268,
"step": 10700
},
{
"epoch": 0.9564207894266833,
"grad_norm": 0.3256109654903412,
"learning_rate": 9.313726310752581e-06,
"loss": 0.6108,
"step": 10710
},
{
"epoch": 0.9573138060367923,
"grad_norm": 0.3188807964324951,
"learning_rate": 9.00790595084477e-06,
"loss": 0.6077,
"step": 10720
},
{
"epoch": 0.9582068226469013,
"grad_norm": 0.36516305804252625,
"learning_rate": 8.703077967816944e-06,
"loss": 0.6144,
"step": 10730
},
{
"epoch": 0.9590998392570101,
"grad_norm": 0.36421048641204834,
"learning_rate": 8.39953264664192e-06,
"loss": 0.6089,
"step": 10740
},
{
"epoch": 0.9599928558671191,
"grad_norm": 0.3440021276473999,
"learning_rate": 8.097559050825058e-06,
"loss": 0.5947,
"step": 10750
},
{
"epoch": 0.9608858724772281,
"grad_norm": 0.37342238426208496,
"learning_rate": 7.79744474713166e-06,
"loss": 0.5703,
"step": 10760
},
{
"epoch": 0.961778889087337,
"grad_norm": 0.33532771468162537,
"learning_rate": 7.499475531739696e-06,
"loss": 0.5893,
"step": 10770
},
{
"epoch": 0.962671905697446,
"grad_norm": 0.36765265464782715,
"learning_rate": 7.203935158078662e-06,
"loss": 0.6186,
"step": 10780
},
{
"epoch": 0.963564922307555,
"grad_norm": 0.38002893328666687,
"learning_rate": 6.911105066613709e-06,
"loss": 0.6039,
"step": 10790
},
{
"epoch": 0.9644579389176638,
"grad_norm": 0.5010676383972168,
"learning_rate": 6.621264116832394e-06,
"loss": 0.6082,
"step": 10800
},
{
"epoch": 0.9653509555277728,
"grad_norm": 0.3839252293109894,
"learning_rate": 6.334688321689274e-06,
"loss": 0.6194,
"step": 10810
},
{
"epoch": 0.9662439721378817,
"grad_norm": 0.35580697655677795,
"learning_rate": 6.051650584761237e-06,
"loss": 0.6032,
"step": 10820
},
{
"epoch": 0.9671369887479907,
"grad_norm": 0.4455433487892151,
"learning_rate": 5.772420440363841e-06,
"loss": 0.6256,
"step": 10830
},
{
"epoch": 0.9680300053580997,
"grad_norm": 0.48495927453041077,
"learning_rate": 5.497263796876207e-06,
"loss": 0.5905,
"step": 10840
},
{
"epoch": 0.9689230219682086,
"grad_norm": 0.3414197564125061,
"learning_rate": 5.226442683518813e-06,
"loss": 0.5985,
"step": 10850
},
{
"epoch": 0.9698160385783176,
"grad_norm": 0.45709747076034546,
"learning_rate": 4.960215000825432e-06,
"loss": 0.596,
"step": 10860
},
{
"epoch": 0.9707090551884265,
"grad_norm": 0.3490951955318451,
"learning_rate": 4.698834275046711e-06,
"loss": 0.6297,
"step": 10870
},
{
"epoch": 0.9716020717985354,
"grad_norm": 0.3531166613101959,
"learning_rate": 4.442549416719423e-06,
"loss": 0.5763,
"step": 10880
},
{
"epoch": 0.9724950884086444,
"grad_norm": 0.3562028110027313,
"learning_rate": 4.1916044836311565e-06,
"loss": 0.6064,
"step": 10890
},
{
"epoch": 0.9733881050187534,
"grad_norm": 0.3873286843299866,
"learning_rate": 3.94623844840626e-06,
"loss": 0.5846,
"step": 10900
},
{
"epoch": 0.9742811216288623,
"grad_norm": 0.3266199827194214,
"learning_rate": 3.7066849709343467e-06,
"loss": 0.5851,
"step": 10910
},
{
"epoch": 0.9751741382389713,
"grad_norm": 0.36563774943351746,
"learning_rate": 3.473172175858059e-06,
"loss": 0.6047,
"step": 10920
},
{
"epoch": 0.9760671548490802,
"grad_norm": 0.3520563542842865,
"learning_rate": 3.2459224353320046e-06,
"loss": 0.5874,
"step": 10930
},
{
"epoch": 0.9769601714591891,
"grad_norm": 0.3633574843406677,
"learning_rate": 3.025152157259735e-06,
"loss": 0.5894,
"step": 10940
},
{
"epoch": 0.9778531880692981,
"grad_norm": 0.40328413248062134,
"learning_rate": 2.811071579210425e-06,
"loss": 0.6169,
"step": 10950
},
{
"epoch": 0.978746204679407,
"grad_norm": 0.35317492485046387,
"learning_rate": 2.603884568211513e-06,
"loss": 0.5771,
"step": 10960
},
{
"epoch": 0.979639221289516,
"grad_norm": 0.38915857672691345,
"learning_rate": 2.4037884266079434e-06,
"loss": 0.6025,
"step": 10970
},
{
"epoch": 0.980532237899625,
"grad_norm": 0.36653295159339905,
"learning_rate": 2.2109737041728955e-06,
"loss": 0.6235,
"step": 10980
},
{
"epoch": 0.9814252545097338,
"grad_norm": 0.380196750164032,
"learning_rate": 2.0256240166489454e-06,
"loss": 0.6031,
"step": 10990
},
{
"epoch": 0.9823182711198428,
"grad_norm": 0.3558424413204193,
"learning_rate": 1.847915870892413e-06,
"loss": 0.6319,
"step": 11000
},
{
"epoch": 0.9832112877299518,
"grad_norm": 0.40899258852005005,
"learning_rate": 1.6780184967874715e-06,
"loss": 0.6088,
"step": 11010
},
{
"epoch": 0.9841043043400607,
"grad_norm": 0.333664208650589,
"learning_rate": 1.5160936860900294e-06,
"loss": 0.599,
"step": 11020
},
{
"epoch": 0.9849973209501697,
"grad_norm": 0.36042433977127075,
"learning_rate": 1.3622956383548762e-06,
"loss": 0.6244,
"step": 11030
},
{
"epoch": 0.9858903375602787,
"grad_norm": 0.3818886876106262,
"learning_rate": 1.216770814092834e-06,
"loss": 0.6089,
"step": 11040
},
{
"epoch": 0.9867833541703875,
"grad_norm": 0.5575549006462097,
"learning_rate": 1.0796577952977013e-06,
"loss": 0.6211,
"step": 11050
},
{
"epoch": 0.9876763707804965,
"grad_norm": 0.383680522441864,
"learning_rate": 9.510871534758705e-07,
"loss": 0.6057,
"step": 11060
},
{
"epoch": 0.9885693873906055,
"grad_norm": 0.3265126645565033,
"learning_rate": 8.311813253042334e-07,
"loss": 0.5875,
"step": 11070
},
{
"epoch": 0.9894624040007144,
"grad_norm": 0.3824686110019684,
"learning_rate": 7.200544960348308e-07,
"loss": 0.6101,
"step": 11080
},
{
"epoch": 0.9903554206108234,
"grad_norm": 0.34627270698547363,
"learning_rate": 6.178124907572545e-07,
"loss": 0.6042,
"step": 11090
},
{
"epoch": 0.9912484372209323,
"grad_norm": 0.3606855869293213,
"learning_rate": 5.245526736223528e-07,
"loss": 0.6044,
"step": 11100
},
{
"epoch": 0.9921414538310412,
"grad_norm": 0.36044779419898987,
"learning_rate": 4.403638551232205e-07,
"loss": 0.6059,
"step": 11110
},
{
"epoch": 0.9930344704411502,
"grad_norm": 0.3836638927459717,
"learning_rate": 3.6532620752175616e-07,
"loss": 0.6216,
"step": 11120
},
{
"epoch": 0.9939274870512591,
"grad_norm": 0.40283387899398804,
"learning_rate": 2.9951118850133247e-07,
"loss": 0.6126,
"step": 11130
},
{
"epoch": 0.9948205036613681,
"grad_norm": 0.38970088958740234,
"learning_rate": 2.429814731182774e-07,
"loss": 0.6104,
"step": 11140
},
{
"epoch": 0.9957135202714771,
"grad_norm": 0.3077683746814728,
"learning_rate": 1.957908941169813e-07,
"loss": 0.5801,
"step": 11150
},
{
"epoch": 0.996606536881586,
"grad_norm": 0.38186129927635193,
"learning_rate": 1.5798439066545254e-07,
"loss": 0.6235,
"step": 11160
},
{
"epoch": 0.997499553491695,
"grad_norm": 0.3379780948162079,
"learning_rate": 1.2959796556014471e-07,
"loss": 0.601,
"step": 11170
},
{
"epoch": 0.9983925701018039,
"grad_norm": 0.3538696765899658,
"learning_rate": 1.1065865094081884e-07,
"loss": 0.6549,
"step": 11180
},
{
"epoch": 0.9992855867119128,
"grad_norm": 0.31598612666130066,
"learning_rate": 1.011844825480734e-07,
"loss": 0.6135,
"step": 11190
},
{
"epoch": 0.9996427933559564,
"step": 11194,
"total_flos": 4.547514943786923e+19,
"train_loss": 0.0035392726683842407,
"train_runtime": 601.378,
"train_samples_per_second": 223.367,
"train_steps_per_second": 18.614
}
],
"logging_steps": 10,
"max_steps": 11194,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 159,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 4.547514943786923e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}