MobileWorld-html-8B / trainer_state.json
xwk123's picture
Upload model checkpoint 4965
61240ba verified
Raw
History Blame Contribute Delete
94.4 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 4965,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0060435132957292505,
"grad_norm": 2.6791505813598633,
"learning_rate": 1.8108651911468814e-07,
"loss": 0.43424110412597655,
"step": 10
},
{
"epoch": 0.012087026591458501,
"grad_norm": 2.683911085128784,
"learning_rate": 3.8229376257545275e-07,
"loss": 0.42812118530273435,
"step": 20
},
{
"epoch": 0.01813053988718775,
"grad_norm": 1.6422226428985596,
"learning_rate": 5.835010060362174e-07,
"loss": 0.419526481628418,
"step": 30
},
{
"epoch": 0.024174053182917002,
"grad_norm": 0.8506667017936707,
"learning_rate": 7.847082494969819e-07,
"loss": 0.3855069637298584,
"step": 40
},
{
"epoch": 0.030217566478646252,
"grad_norm": 0.5741743445396423,
"learning_rate": 9.859154929577465e-07,
"loss": 0.3555116653442383,
"step": 50
},
{
"epoch": 0.0362610797743755,
"grad_norm": 0.4924837350845337,
"learning_rate": 1.187122736418511e-06,
"loss": 0.3286411762237549,
"step": 60
},
{
"epoch": 0.04230459307010476,
"grad_norm": 0.33213233947753906,
"learning_rate": 1.3883299798792758e-06,
"loss": 0.31856331825256345,
"step": 70
},
{
"epoch": 0.048348106365834004,
"grad_norm": 0.2811642587184906,
"learning_rate": 1.5895372233400403e-06,
"loss": 0.29794213771820066,
"step": 80
},
{
"epoch": 0.05439161966156326,
"grad_norm": 0.27522650361061096,
"learning_rate": 1.790744466800805e-06,
"loss": 0.2954577445983887,
"step": 90
},
{
"epoch": 0.060435132957292505,
"grad_norm": 0.30479246377944946,
"learning_rate": 1.9919517102615694e-06,
"loss": 0.275892972946167,
"step": 100
},
{
"epoch": 0.06647864625302176,
"grad_norm": 0.2663208842277527,
"learning_rate": 2.193158953722334e-06,
"loss": 0.2677940845489502,
"step": 110
},
{
"epoch": 0.072522159548751,
"grad_norm": 0.2940618097782135,
"learning_rate": 2.3943661971830984e-06,
"loss": 0.27342398166656495,
"step": 120
},
{
"epoch": 0.07856567284448025,
"grad_norm": 0.2885012626647949,
"learning_rate": 2.595573440643863e-06,
"loss": 0.25932297706604,
"step": 130
},
{
"epoch": 0.08460918614020951,
"grad_norm": 0.26729241013526917,
"learning_rate": 2.796780684104628e-06,
"loss": 0.2515596389770508,
"step": 140
},
{
"epoch": 0.09065269943593876,
"grad_norm": 0.25183233618736267,
"learning_rate": 2.9979879275653927e-06,
"loss": 0.25308003425598147,
"step": 150
},
{
"epoch": 0.09669621273166801,
"grad_norm": 0.27593663334846497,
"learning_rate": 3.199195171026157e-06,
"loss": 0.24183225631713867,
"step": 160
},
{
"epoch": 0.10273972602739725,
"grad_norm": 0.26931726932525635,
"learning_rate": 3.4004024144869217e-06,
"loss": 0.24087212085723878,
"step": 170
},
{
"epoch": 0.10878323932312652,
"grad_norm": 0.28699785470962524,
"learning_rate": 3.601609657947687e-06,
"loss": 0.23806214332580566,
"step": 180
},
{
"epoch": 0.11482675261885576,
"grad_norm": 0.2808695435523987,
"learning_rate": 3.8028169014084508e-06,
"loss": 0.23471593856811523,
"step": 190
},
{
"epoch": 0.12087026591458501,
"grad_norm": 0.26960834860801697,
"learning_rate": 4.0040241448692155e-06,
"loss": 0.22571153640747071,
"step": 200
},
{
"epoch": 0.12691377921031427,
"grad_norm": 0.2849039137363434,
"learning_rate": 4.20523138832998e-06,
"loss": 0.2278608798980713,
"step": 210
},
{
"epoch": 0.13295729250604352,
"grad_norm": 0.2857084572315216,
"learning_rate": 4.406438631790745e-06,
"loss": 0.2277141809463501,
"step": 220
},
{
"epoch": 0.13900080580177276,
"grad_norm": 0.28886616230010986,
"learning_rate": 4.60764587525151e-06,
"loss": 0.2221583366394043,
"step": 230
},
{
"epoch": 0.145044319097502,
"grad_norm": 0.28347155451774597,
"learning_rate": 4.808853118712274e-06,
"loss": 0.2190323829650879,
"step": 240
},
{
"epoch": 0.15108783239323126,
"grad_norm": 0.27866682410240173,
"learning_rate": 5.010060362173038e-06,
"loss": 0.21880054473876953,
"step": 250
},
{
"epoch": 0.1571313456889605,
"grad_norm": 0.28653502464294434,
"learning_rate": 5.211267605633803e-06,
"loss": 0.22141761779785157,
"step": 260
},
{
"epoch": 0.16317485898468978,
"grad_norm": 0.2667871117591858,
"learning_rate": 5.412474849094568e-06,
"loss": 0.2136612892150879,
"step": 270
},
{
"epoch": 0.16921837228041903,
"grad_norm": 0.26795583963394165,
"learning_rate": 5.6136820925553326e-06,
"loss": 0.2144620656967163,
"step": 280
},
{
"epoch": 0.17526188557614827,
"grad_norm": 0.2649329602718353,
"learning_rate": 5.814889336016097e-06,
"loss": 0.20957472324371337,
"step": 290
},
{
"epoch": 0.18130539887187752,
"grad_norm": 0.29122769832611084,
"learning_rate": 6.016096579476862e-06,
"loss": 0.21188030242919922,
"step": 300
},
{
"epoch": 0.18734891216760677,
"grad_norm": 0.2643492817878723,
"learning_rate": 6.217303822937626e-06,
"loss": 0.20421910285949707,
"step": 310
},
{
"epoch": 0.19339242546333602,
"grad_norm": 0.2602762281894684,
"learning_rate": 6.418511066398391e-06,
"loss": 0.20553312301635743,
"step": 320
},
{
"epoch": 0.19943593875906526,
"grad_norm": 0.2506936490535736,
"learning_rate": 6.619718309859155e-06,
"loss": 0.20259194374084472,
"step": 330
},
{
"epoch": 0.2054794520547945,
"grad_norm": 0.26634275913238525,
"learning_rate": 6.82092555331992e-06,
"loss": 0.20387697219848633,
"step": 340
},
{
"epoch": 0.21152296535052378,
"grad_norm": 0.293220192193985,
"learning_rate": 7.022132796780685e-06,
"loss": 0.1974625825881958,
"step": 350
},
{
"epoch": 0.21756647864625303,
"grad_norm": 0.26546773314476013,
"learning_rate": 7.223340040241449e-06,
"loss": 0.1918584704399109,
"step": 360
},
{
"epoch": 0.22360999194198228,
"grad_norm": 0.2801249027252197,
"learning_rate": 7.4245472837022135e-06,
"loss": 0.2003864526748657,
"step": 370
},
{
"epoch": 0.22965350523771152,
"grad_norm": 0.26137885451316833,
"learning_rate": 7.625754527162978e-06,
"loss": 0.19211164712905884,
"step": 380
},
{
"epoch": 0.23569701853344077,
"grad_norm": 0.23950669169425964,
"learning_rate": 7.826961770623743e-06,
"loss": 0.1997000217437744,
"step": 390
},
{
"epoch": 0.24174053182917002,
"grad_norm": 0.2763938307762146,
"learning_rate": 8.028169014084509e-06,
"loss": 0.19561755657196045,
"step": 400
},
{
"epoch": 0.24778404512489927,
"grad_norm": 0.2883632183074951,
"learning_rate": 8.229376257545272e-06,
"loss": 0.1947632312774658,
"step": 410
},
{
"epoch": 0.25382755842062854,
"grad_norm": 0.2761209011077881,
"learning_rate": 8.430583501006036e-06,
"loss": 0.19417184591293335,
"step": 420
},
{
"epoch": 0.25987107171635776,
"grad_norm": 0.29035407304763794,
"learning_rate": 8.631790744466802e-06,
"loss": 0.19058849811553955,
"step": 430
},
{
"epoch": 0.26591458501208703,
"grad_norm": 0.24615156650543213,
"learning_rate": 8.832997987927566e-06,
"loss": 0.18921122550964356,
"step": 440
},
{
"epoch": 0.27195809830781625,
"grad_norm": 0.2587921619415283,
"learning_rate": 9.034205231388331e-06,
"loss": 0.18803141117095948,
"step": 450
},
{
"epoch": 0.27800161160354553,
"grad_norm": 0.27339935302734375,
"learning_rate": 9.235412474849095e-06,
"loss": 0.1917414426803589,
"step": 460
},
{
"epoch": 0.2840451248992748,
"grad_norm": 0.26091912388801575,
"learning_rate": 9.43661971830986e-06,
"loss": 0.18812184333801268,
"step": 470
},
{
"epoch": 0.290088638195004,
"grad_norm": 0.26199963688850403,
"learning_rate": 9.637826961770625e-06,
"loss": 0.18310174942016602,
"step": 480
},
{
"epoch": 0.2961321514907333,
"grad_norm": 0.26349949836730957,
"learning_rate": 9.839034205231389e-06,
"loss": 0.1794055223464966,
"step": 490
},
{
"epoch": 0.3021756647864625,
"grad_norm": 0.2564206123352051,
"learning_rate": 9.999995056058254e-06,
"loss": 0.181296706199646,
"step": 500
},
{
"epoch": 0.3082191780821918,
"grad_norm": 0.2618381679058075,
"learning_rate": 9.999822019123702e-06,
"loss": 0.18494994640350343,
"step": 510
},
{
"epoch": 0.314262691377921,
"grad_norm": 0.2703542709350586,
"learning_rate": 9.999401794878746e-06,
"loss": 0.17985227108001708,
"step": 520
},
{
"epoch": 0.3203062046736503,
"grad_norm": 0.29166391491889954,
"learning_rate": 9.998734404098946e-06,
"loss": 0.18138515949249268,
"step": 530
},
{
"epoch": 0.32634971796937956,
"grad_norm": 0.24447283148765564,
"learning_rate": 9.997819879779586e-06,
"loss": 0.18624844551086425,
"step": 540
},
{
"epoch": 0.3323932312651088,
"grad_norm": 0.24413910508155823,
"learning_rate": 9.996658267134033e-06,
"loss": 0.18135397434234618,
"step": 550
},
{
"epoch": 0.33843674456083805,
"grad_norm": 0.26679185032844543,
"learning_rate": 9.995249623591515e-06,
"loss": 0.17672932147979736,
"step": 560
},
{
"epoch": 0.3444802578565673,
"grad_norm": 0.25367289781570435,
"learning_rate": 9.993594018794271e-06,
"loss": 0.17478251457214355,
"step": 570
},
{
"epoch": 0.35052377115229655,
"grad_norm": 0.24849548935890198,
"learning_rate": 9.991691534594113e-06,
"loss": 0.1787455201148987,
"step": 580
},
{
"epoch": 0.35656728444802577,
"grad_norm": 0.263493150472641,
"learning_rate": 9.989542265048383e-06,
"loss": 0.17695608139038085,
"step": 590
},
{
"epoch": 0.36261079774375504,
"grad_norm": 0.23856207728385925,
"learning_rate": 9.987146316415292e-06,
"loss": 0.17286056280136108,
"step": 600
},
{
"epoch": 0.36865431103948426,
"grad_norm": 0.3688794672489166,
"learning_rate": 9.984503807148677e-06,
"loss": 0.17635351419448853,
"step": 610
},
{
"epoch": 0.37469782433521354,
"grad_norm": 0.25517919659614563,
"learning_rate": 9.981614867892143e-06,
"loss": 0.17394735813140869,
"step": 620
},
{
"epoch": 0.3807413376309428,
"grad_norm": 0.24865099787712097,
"learning_rate": 9.978479641472593e-06,
"loss": 0.1734173536300659,
"step": 630
},
{
"epoch": 0.38678485092667203,
"grad_norm": 0.2586166262626648,
"learning_rate": 9.975098282893187e-06,
"loss": 0.17834866046905518,
"step": 640
},
{
"epoch": 0.3928283642224013,
"grad_norm": 0.2314433753490448,
"learning_rate": 9.97147095932566e-06,
"loss": 0.1714148998260498,
"step": 650
},
{
"epoch": 0.3988718775181305,
"grad_norm": 0.25793835520744324,
"learning_rate": 9.967597850102069e-06,
"loss": 0.17357563972473145,
"step": 660
},
{
"epoch": 0.4049153908138598,
"grad_norm": 0.2363017499446869,
"learning_rate": 9.963479146705916e-06,
"loss": 0.17228338718414307,
"step": 670
},
{
"epoch": 0.410958904109589,
"grad_norm": 0.24678575992584229,
"learning_rate": 9.959115052762697e-06,
"loss": 0.17418212890625,
"step": 680
},
{
"epoch": 0.4170024174053183,
"grad_norm": 0.23901484906673431,
"learning_rate": 9.95450578402982e-06,
"loss": 0.16737285852432252,
"step": 690
},
{
"epoch": 0.42304593070104757,
"grad_norm": 0.2519242465496063,
"learning_rate": 9.949651568385942e-06,
"loss": 0.1672084093093872,
"step": 700
},
{
"epoch": 0.4290894439967768,
"grad_norm": 0.24263595044612885,
"learning_rate": 9.94455264581971e-06,
"loss": 0.16824760437011718,
"step": 710
},
{
"epoch": 0.43513295729250606,
"grad_norm": 0.23398065567016602,
"learning_rate": 9.939209268417886e-06,
"loss": 0.1633542537689209,
"step": 720
},
{
"epoch": 0.4411764705882353,
"grad_norm": 0.28119024634361267,
"learning_rate": 9.933621700352896e-06,
"loss": 0.1700071930885315,
"step": 730
},
{
"epoch": 0.44721998388396456,
"grad_norm": 0.24456487596035004,
"learning_rate": 9.927790217869753e-06,
"loss": 0.16781260967254638,
"step": 740
},
{
"epoch": 0.4532634971796938,
"grad_norm": 0.2353801131248474,
"learning_rate": 9.921715109272413e-06,
"loss": 0.16569297313690184,
"step": 750
},
{
"epoch": 0.45930701047542305,
"grad_norm": 0.24682150781154633,
"learning_rate": 9.915396674909522e-06,
"loss": 0.16531245708465575,
"step": 760
},
{
"epoch": 0.46535052377115227,
"grad_norm": 0.22643272578716278,
"learning_rate": 9.908835227159559e-06,
"loss": 0.16031930446624756,
"step": 770
},
{
"epoch": 0.47139403706688154,
"grad_norm": 0.21679629385471344,
"learning_rate": 9.902031090415391e-06,
"loss": 0.1657174348831177,
"step": 780
},
{
"epoch": 0.4774375503626108,
"grad_norm": 0.31907621026039124,
"learning_rate": 9.894984601068248e-06,
"loss": 0.16643948554992677,
"step": 790
},
{
"epoch": 0.48348106365834004,
"grad_norm": 0.22850657999515533,
"learning_rate": 9.887696107491079e-06,
"loss": 0.16328579187393188,
"step": 800
},
{
"epoch": 0.4895245769540693,
"grad_norm": 0.2363438904285431,
"learning_rate": 9.880165970021337e-06,
"loss": 0.16295944452285765,
"step": 810
},
{
"epoch": 0.49556809024979853,
"grad_norm": 0.21235784888267517,
"learning_rate": 9.872394560943157e-06,
"loss": 0.1601036310195923,
"step": 820
},
{
"epoch": 0.5016116035455278,
"grad_norm": 0.23377592861652374,
"learning_rate": 9.86438226446896e-06,
"loss": 0.16438961029052734,
"step": 830
},
{
"epoch": 0.5076551168412571,
"grad_norm": 0.2163173109292984,
"learning_rate": 9.856129476720449e-06,
"loss": 0.1559450387954712,
"step": 840
},
{
"epoch": 0.5136986301369864,
"grad_norm": 0.23434801399707794,
"learning_rate": 9.847636605709028e-06,
"loss": 0.1650776147842407,
"step": 850
},
{
"epoch": 0.5197421434327155,
"grad_norm": 0.2319152057170868,
"learning_rate": 9.838904071315632e-06,
"loss": 0.1596601963043213,
"step": 860
},
{
"epoch": 0.5257856567284448,
"grad_norm": 0.21985158324241638,
"learning_rate": 9.829932305269964e-06,
"loss": 0.16151692867279052,
"step": 870
},
{
"epoch": 0.5318291700241741,
"grad_norm": 0.2180061638355255,
"learning_rate": 9.820721751129163e-06,
"loss": 0.16482841968536377,
"step": 880
},
{
"epoch": 0.5378726833199033,
"grad_norm": 0.22528943419456482,
"learning_rate": 9.811272864255853e-06,
"loss": 0.158658766746521,
"step": 890
},
{
"epoch": 0.5439161966156325,
"grad_norm": 0.21785296499729156,
"learning_rate": 9.801586111795655e-06,
"loss": 0.16361989974975585,
"step": 900
},
{
"epoch": 0.5499597099113618,
"grad_norm": 0.2318844050168991,
"learning_rate": 9.791661972654069e-06,
"loss": 0.15711183547973634,
"step": 910
},
{
"epoch": 0.5560032232070911,
"grad_norm": 0.21686816215515137,
"learning_rate": 9.781500937472814e-06,
"loss": 0.15947939157485963,
"step": 920
},
{
"epoch": 0.5620467365028203,
"grad_norm": 0.2278321385383606,
"learning_rate": 9.771103508605563e-06,
"loss": 0.1608155846595764,
"step": 930
},
{
"epoch": 0.5680902497985496,
"grad_norm": 0.22438998520374298,
"learning_rate": 9.760470200093111e-06,
"loss": 0.15645452737808227,
"step": 940
},
{
"epoch": 0.5741337630942788,
"grad_norm": 0.2459161877632141,
"learning_rate": 9.749601537637956e-06,
"loss": 0.15974905490875244,
"step": 950
},
{
"epoch": 0.580177276390008,
"grad_norm": 0.22639086842536926,
"learning_rate": 9.738498058578311e-06,
"loss": 0.16175615787506104,
"step": 960
},
{
"epoch": 0.5862207896857373,
"grad_norm": 0.21432101726531982,
"learning_rate": 9.727160311861546e-06,
"loss": 0.1513558268547058,
"step": 970
},
{
"epoch": 0.5922643029814666,
"grad_norm": 0.2377699315547943,
"learning_rate": 9.715588858017033e-06,
"loss": 0.15713939666748047,
"step": 980
},
{
"epoch": 0.5983078162771958,
"grad_norm": 0.23188424110412598,
"learning_rate": 9.703784269128449e-06,
"loss": 0.15651999711990355,
"step": 990
},
{
"epoch": 0.604351329572925,
"grad_norm": 0.22484862804412842,
"learning_rate": 9.691747128805482e-06,
"loss": 0.15296961069107057,
"step": 1000
},
{
"epoch": 0.6103948428686543,
"grad_norm": 0.21575000882148743,
"learning_rate": 9.679478032154986e-06,
"loss": 0.15741455554962158,
"step": 1010
},
{
"epoch": 0.6164383561643836,
"grad_norm": 0.2111063003540039,
"learning_rate": 9.666977585751551e-06,
"loss": 0.1585395574569702,
"step": 1020
},
{
"epoch": 0.6224818694601129,
"grad_norm": 0.2278982400894165,
"learning_rate": 9.654246407607522e-06,
"loss": 0.15823476314544677,
"step": 1030
},
{
"epoch": 0.628525382755842,
"grad_norm": 0.21201825141906738,
"learning_rate": 9.641285127142442e-06,
"loss": 0.15284976959228516,
"step": 1040
},
{
"epoch": 0.6345688960515713,
"grad_norm": 0.22854183614253998,
"learning_rate": 9.628094385151931e-06,
"loss": 0.14952181577682494,
"step": 1050
},
{
"epoch": 0.6406124093473006,
"grad_norm": 0.21677425503730774,
"learning_rate": 9.61467483377601e-06,
"loss": 0.14994876384735106,
"step": 1060
},
{
"epoch": 0.6466559226430298,
"grad_norm": 0.24580124020576477,
"learning_rate": 9.601027136466854e-06,
"loss": 0.1575385332107544,
"step": 1070
},
{
"epoch": 0.6526994359387591,
"grad_norm": 0.22464416921138763,
"learning_rate": 9.587151967956006e-06,
"loss": 0.15378172397613527,
"step": 1080
},
{
"epoch": 0.6587429492344883,
"grad_norm": 0.22706280648708344,
"learning_rate": 9.573050014220995e-06,
"loss": 0.15251691341400148,
"step": 1090
},
{
"epoch": 0.6647864625302176,
"grad_norm": 0.2249971181154251,
"learning_rate": 9.558721972451447e-06,
"loss": 0.1516406536102295,
"step": 1100
},
{
"epoch": 0.6708299758259468,
"grad_norm": 0.21184460818767548,
"learning_rate": 9.544168551014594e-06,
"loss": 0.1552306056022644,
"step": 1110
},
{
"epoch": 0.6768734891216761,
"grad_norm": 0.21974043548107147,
"learning_rate": 9.52939046942027e-06,
"loss": 0.14897232055664061,
"step": 1120
},
{
"epoch": 0.6829170024174053,
"grad_norm": 0.22956401109695435,
"learning_rate": 9.514388458285332e-06,
"loss": 0.15314501523971558,
"step": 1130
},
{
"epoch": 0.6889605157131345,
"grad_norm": 0.20469701290130615,
"learning_rate": 9.499163259297535e-06,
"loss": 0.15435714721679689,
"step": 1140
},
{
"epoch": 0.6950040290088638,
"grad_norm": 0.2130810171365738,
"learning_rate": 9.483715625178874e-06,
"loss": 0.14373431205749512,
"step": 1150
},
{
"epoch": 0.7010475423045931,
"grad_norm": 0.21660926938056946,
"learning_rate": 9.468046319648358e-06,
"loss": 0.15549267530441285,
"step": 1160
},
{
"epoch": 0.7070910556003224,
"grad_norm": 0.22545422613620758,
"learning_rate": 9.45215611738426e-06,
"loss": 0.1464073419570923,
"step": 1170
},
{
"epoch": 0.7131345688960515,
"grad_norm": 0.2041836678981781,
"learning_rate": 9.436045803985819e-06,
"loss": 0.14939506053924562,
"step": 1180
},
{
"epoch": 0.7191780821917808,
"grad_norm": 0.21615348756313324,
"learning_rate": 9.41971617593439e-06,
"loss": 0.1525928258895874,
"step": 1190
},
{
"epoch": 0.7252215954875101,
"grad_norm": 0.21271154284477234,
"learning_rate": 9.403168040554085e-06,
"loss": 0.14430539608001708,
"step": 1200
},
{
"epoch": 0.7312651087832394,
"grad_norm": 0.22406251728534698,
"learning_rate": 9.386402215971835e-06,
"loss": 0.14692449569702148,
"step": 1210
},
{
"epoch": 0.7373086220789685,
"grad_norm": 0.22445231676101685,
"learning_rate": 9.369419531076967e-06,
"loss": 0.15025017261505128,
"step": 1220
},
{
"epoch": 0.7433521353746978,
"grad_norm": 0.21477998793125153,
"learning_rate": 9.352220825480205e-06,
"loss": 0.15201830863952637,
"step": 1230
},
{
"epoch": 0.7493956486704271,
"grad_norm": 0.20397023856639862,
"learning_rate": 9.334806949472176e-06,
"loss": 0.14703288078308105,
"step": 1240
},
{
"epoch": 0.7554391619661563,
"grad_norm": 0.20194675028324127,
"learning_rate": 9.317178763981354e-06,
"loss": 0.14660085439682008,
"step": 1250
},
{
"epoch": 0.7614826752618856,
"grad_norm": 0.2115529626607895,
"learning_rate": 9.29933714053152e-06,
"loss": 0.1503981828689575,
"step": 1260
},
{
"epoch": 0.7675261885576148,
"grad_norm": 0.20875698328018188,
"learning_rate": 9.281282961198651e-06,
"loss": 0.14698731899261475,
"step": 1270
},
{
"epoch": 0.7735697018533441,
"grad_norm": 0.21880143880844116,
"learning_rate": 9.263017118567329e-06,
"loss": 0.1496628999710083,
"step": 1280
},
{
"epoch": 0.7796132151490733,
"grad_norm": 0.20595993101596832,
"learning_rate": 9.2445405156866e-06,
"loss": 0.14186606407165528,
"step": 1290
},
{
"epoch": 0.7856567284448026,
"grad_norm": 0.2238467037677765,
"learning_rate": 9.225854066025336e-06,
"loss": 0.14565565586090087,
"step": 1300
},
{
"epoch": 0.7917002417405318,
"grad_norm": 0.20252187550067902,
"learning_rate": 9.206958693427063e-06,
"loss": 0.14660226106643676,
"step": 1310
},
{
"epoch": 0.797743755036261,
"grad_norm": 0.2454824447631836,
"learning_rate": 9.187855332064307e-06,
"loss": 0.14617925882339478,
"step": 1320
},
{
"epoch": 0.8037872683319903,
"grad_norm": 0.20953455567359924,
"learning_rate": 9.168544926392388e-06,
"loss": 0.14425084590911866,
"step": 1330
},
{
"epoch": 0.8098307816277196,
"grad_norm": 0.22163881361484528,
"learning_rate": 9.149028431102738e-06,
"loss": 0.1437234401702881,
"step": 1340
},
{
"epoch": 0.8158742949234489,
"grad_norm": 0.21825546026229858,
"learning_rate": 9.129306811075698e-06,
"loss": 0.14237486124038695,
"step": 1350
},
{
"epoch": 0.821917808219178,
"grad_norm": 0.21504727005958557,
"learning_rate": 9.109381041332815e-06,
"loss": 0.1479846477508545,
"step": 1360
},
{
"epoch": 0.8279613215149073,
"grad_norm": 0.3299539387226105,
"learning_rate": 9.089252106988645e-06,
"loss": 0.1492680311203003,
"step": 1370
},
{
"epoch": 0.8340048348106366,
"grad_norm": 0.2820020318031311,
"learning_rate": 9.06892100320204e-06,
"loss": 0.14104161262512208,
"step": 1380
},
{
"epoch": 0.8400483481063659,
"grad_norm": 0.24350281059741974,
"learning_rate": 9.048388735126949e-06,
"loss": 0.14567277431488038,
"step": 1390
},
{
"epoch": 0.8460918614020951,
"grad_norm": 0.19603702425956726,
"learning_rate": 9.027656317862733e-06,
"loss": 0.14261494874954223,
"step": 1400
},
{
"epoch": 0.8521353746978243,
"grad_norm": 0.2087412029504776,
"learning_rate": 9.00672477640397e-06,
"loss": 0.14455668926239013,
"step": 1410
},
{
"epoch": 0.8581788879935536,
"grad_norm": 0.2269335836172104,
"learning_rate": 8.98559514558978e-06,
"loss": 0.14213941097259522,
"step": 1420
},
{
"epoch": 0.8642224012892828,
"grad_norm": 0.20767658948898315,
"learning_rate": 8.964268470052676e-06,
"loss": 0.14149200916290283,
"step": 1430
},
{
"epoch": 0.8702659145850121,
"grad_norm": 0.20842665433883667,
"learning_rate": 8.942745804166898e-06,
"loss": 0.14532248973846434,
"step": 1440
},
{
"epoch": 0.8763094278807413,
"grad_norm": 0.2045581191778183,
"learning_rate": 8.921028211996304e-06,
"loss": 0.14435100555419922,
"step": 1450
},
{
"epoch": 0.8823529411764706,
"grad_norm": 0.19654729962348938,
"learning_rate": 8.89911676724175e-06,
"loss": 0.14154709577560426,
"step": 1460
},
{
"epoch": 0.8883964544721998,
"grad_norm": 0.21717678010463715,
"learning_rate": 8.87701255318802e-06,
"loss": 0.14401607513427733,
"step": 1470
},
{
"epoch": 0.8944399677679291,
"grad_norm": 0.1900615245103836,
"learning_rate": 8.854716662650255e-06,
"loss": 0.14473350048065187,
"step": 1480
},
{
"epoch": 0.9004834810636584,
"grad_norm": 0.2035045623779297,
"learning_rate": 8.832230197919938e-06,
"loss": 0.14227144718170165,
"step": 1490
},
{
"epoch": 0.9065269943593876,
"grad_norm": 0.2273212969303131,
"learning_rate": 8.809554270710387e-06,
"loss": 0.13896039724349976,
"step": 1500
},
{
"epoch": 0.9125705076551168,
"grad_norm": 0.22352223098278046,
"learning_rate": 8.7866900021018e-06,
"loss": 0.14246771335601807,
"step": 1510
},
{
"epoch": 0.9186140209508461,
"grad_norm": 0.2079835683107376,
"learning_rate": 8.763638522485827e-06,
"loss": 0.1394651174545288,
"step": 1520
},
{
"epoch": 0.9246575342465754,
"grad_norm": 0.20067137479782104,
"learning_rate": 8.740400971509686e-06,
"loss": 0.13732972145080566,
"step": 1530
},
{
"epoch": 0.9307010475423045,
"grad_norm": 0.28111594915390015,
"learning_rate": 8.716978498019814e-06,
"loss": 0.134084153175354,
"step": 1540
},
{
"epoch": 0.9367445608380338,
"grad_norm": 0.2192552238702774,
"learning_rate": 8.693372260005078e-06,
"loss": 0.13721671104431152,
"step": 1550
},
{
"epoch": 0.9427880741337631,
"grad_norm": 0.21726588904857635,
"learning_rate": 8.669583424539518e-06,
"loss": 0.13847979307174682,
"step": 1560
},
{
"epoch": 0.9488315874294924,
"grad_norm": 0.20951680839061737,
"learning_rate": 8.64561316772465e-06,
"loss": 0.13922054767608644,
"step": 1570
},
{
"epoch": 0.9548751007252216,
"grad_norm": 0.2030678689479828,
"learning_rate": 8.62146267463132e-06,
"loss": 0.14072515964508056,
"step": 1580
},
{
"epoch": 0.9609186140209508,
"grad_norm": 0.21917769312858582,
"learning_rate": 8.597133139241113e-06,
"loss": 0.1449857234954834,
"step": 1590
},
{
"epoch": 0.9669621273166801,
"grad_norm": 0.2006089687347412,
"learning_rate": 8.572625764387334e-06,
"loss": 0.14181168079376222,
"step": 1600
},
{
"epoch": 0.9730056406124094,
"grad_norm": 0.20863375067710876,
"learning_rate": 8.547941761695525e-06,
"loss": 0.13618249893188478,
"step": 1610
},
{
"epoch": 0.9790491539081386,
"grad_norm": 0.20326712727546692,
"learning_rate": 8.523082351523573e-06,
"loss": 0.13995929956436157,
"step": 1620
},
{
"epoch": 0.9850926672038679,
"grad_norm": 0.20966173708438873,
"learning_rate": 8.498048762901371e-06,
"loss": 0.13684163093566895,
"step": 1630
},
{
"epoch": 0.9911361804995971,
"grad_norm": 0.2103189378976822,
"learning_rate": 8.47284223347007e-06,
"loss": 0.13694381713867188,
"step": 1640
},
{
"epoch": 0.9971796937953263,
"grad_norm": 0.21439479291439056,
"learning_rate": 8.447464009420863e-06,
"loss": 0.1397465229034424,
"step": 1650
},
{
"epoch": 1.0,
"eval_loss": 0.13645927608013153,
"eval_runtime": 73.0138,
"eval_samples_per_second": 10.998,
"eval_steps_per_second": 0.698,
"step": 1655
},
{
"epoch": 1.0030217566478645,
"grad_norm": 0.25065481662750244,
"learning_rate": 8.42191534543341e-06,
"loss": 0.13378231525421141,
"step": 1660
},
{
"epoch": 1.0090652699435938,
"grad_norm": 0.23352180421352386,
"learning_rate": 8.396197504613774e-06,
"loss": 0.12866504192352296,
"step": 1670
},
{
"epoch": 1.015108783239323,
"grad_norm": 0.23111934959888458,
"learning_rate": 8.370311758431998e-06,
"loss": 0.1263338088989258,
"step": 1680
},
{
"epoch": 1.0211522965350524,
"grad_norm": 0.21575987339019775,
"learning_rate": 8.344259386659231e-06,
"loss": 0.12696999311447144,
"step": 1690
},
{
"epoch": 1.0271958098307816,
"grad_norm": 0.208969384431839,
"learning_rate": 8.318041677304463e-06,
"loss": 0.12824112176895142,
"step": 1700
},
{
"epoch": 1.033239323126511,
"grad_norm": 0.21562984585762024,
"learning_rate": 8.291659926550845e-06,
"loss": 0.126296329498291,
"step": 1710
},
{
"epoch": 1.0392828364222402,
"grad_norm": 0.2121679186820984,
"learning_rate": 8.265115438691607e-06,
"loss": 0.12731149196624755,
"step": 1720
},
{
"epoch": 1.0453263497179694,
"grad_norm": 0.20733845233917236,
"learning_rate": 8.238409526065575e-06,
"loss": 0.12744966745376587,
"step": 1730
},
{
"epoch": 1.0513698630136987,
"grad_norm": 0.20615026354789734,
"learning_rate": 8.211543508992293e-06,
"loss": 0.13107160329818726,
"step": 1740
},
{
"epoch": 1.0574133763094278,
"grad_norm": 0.20750097930431366,
"learning_rate": 8.184518715706737e-06,
"loss": 0.12762153148651123,
"step": 1750
},
{
"epoch": 1.063456889605157,
"grad_norm": 0.21405155956745148,
"learning_rate": 8.157336482293661e-06,
"loss": 0.12602468729019164,
"step": 1760
},
{
"epoch": 1.0695004029008863,
"grad_norm": 0.20859122276306152,
"learning_rate": 8.129998152621539e-06,
"loss": 0.12637439966201783,
"step": 1770
},
{
"epoch": 1.0755439161966156,
"grad_norm": 0.2062380611896515,
"learning_rate": 8.102505078276115e-06,
"loss": 0.12189408540725707,
"step": 1780
},
{
"epoch": 1.0815874294923449,
"grad_norm": 0.23061694204807281,
"learning_rate": 8.074858618493593e-06,
"loss": 0.1286104679107666,
"step": 1790
},
{
"epoch": 1.0876309427880742,
"grad_norm": 0.20338337123394012,
"learning_rate": 8.047060140093436e-06,
"loss": 0.12919163703918457,
"step": 1800
},
{
"epoch": 1.0936744560838034,
"grad_norm": 0.2183983027935028,
"learning_rate": 8.019111017410785e-06,
"loss": 0.1271709680557251,
"step": 1810
},
{
"epoch": 1.0997179693795327,
"grad_norm": 0.22303783893585205,
"learning_rate": 7.991012632228519e-06,
"loss": 0.12489502429962158,
"step": 1820
},
{
"epoch": 1.105761482675262,
"grad_norm": 0.24084198474884033,
"learning_rate": 7.96276637370894e-06,
"loss": 0.12184426784515381,
"step": 1830
},
{
"epoch": 1.1118049959709912,
"grad_norm": 0.21119949221611023,
"learning_rate": 7.93437363832509e-06,
"loss": 0.12558618783950806,
"step": 1840
},
{
"epoch": 1.1178485092667203,
"grad_norm": 0.21568512916564941,
"learning_rate": 7.905835829791717e-06,
"loss": 0.12440104484558105,
"step": 1850
},
{
"epoch": 1.1238920225624496,
"grad_norm": 0.22111637890338898,
"learning_rate": 7.87715435899587e-06,
"loss": 0.12415385246276855,
"step": 1860
},
{
"epoch": 1.1299355358581789,
"grad_norm": 0.20567971467971802,
"learning_rate": 7.848330643927146e-06,
"loss": 0.1168677568435669,
"step": 1870
},
{
"epoch": 1.1359790491539081,
"grad_norm": 0.2187441736459732,
"learning_rate": 7.819366109607595e-06,
"loss": 0.1290172219276428,
"step": 1880
},
{
"epoch": 1.1420225624496374,
"grad_norm": 0.22424107789993286,
"learning_rate": 7.790262188021257e-06,
"loss": 0.12380149364471435,
"step": 1890
},
{
"epoch": 1.1480660757453667,
"grad_norm": 0.22608165442943573,
"learning_rate": 7.761020318043366e-06,
"loss": 0.12190085649490356,
"step": 1900
},
{
"epoch": 1.154109589041096,
"grad_norm": 0.22297121584415436,
"learning_rate": 7.731641945369223e-06,
"loss": 0.12813954353332518,
"step": 1910
},
{
"epoch": 1.1601531023368252,
"grad_norm": 0.21130020916461945,
"learning_rate": 7.702128522442717e-06,
"loss": 0.12408671379089356,
"step": 1920
},
{
"epoch": 1.1661966156325545,
"grad_norm": 0.21228879690170288,
"learning_rate": 7.672481508384512e-06,
"loss": 0.11850497722625733,
"step": 1930
},
{
"epoch": 1.1722401289282836,
"grad_norm": 0.24546727538108826,
"learning_rate": 7.642702368919919e-06,
"loss": 0.12462244033813477,
"step": 1940
},
{
"epoch": 1.1782836422240128,
"grad_norm": 0.2104247659444809,
"learning_rate": 7.612792576306416e-06,
"loss": 0.12534818649291993,
"step": 1950
},
{
"epoch": 1.184327155519742,
"grad_norm": 0.22319723665714264,
"learning_rate": 7.582753609260885e-06,
"loss": 0.12887107133865355,
"step": 1960
},
{
"epoch": 1.1903706688154714,
"grad_norm": 0.2151956409215927,
"learning_rate": 7.552586952886478e-06,
"loss": 0.12561689615249633,
"step": 1970
},
{
"epoch": 1.1964141821112007,
"grad_norm": 0.24011431634426117,
"learning_rate": 7.5222940985992186e-06,
"loss": 0.124991774559021,
"step": 1980
},
{
"epoch": 1.20245769540693,
"grad_norm": 0.224945068359375,
"learning_rate": 7.491876544054251e-06,
"loss": 0.12340732812881469,
"step": 1990
},
{
"epoch": 1.2085012087026592,
"grad_norm": 0.21621409058570862,
"learning_rate": 7.4613357930718044e-06,
"loss": 0.12357243299484252,
"step": 2000
},
{
"epoch": 1.2145447219983885,
"grad_norm": 0.2096043974161148,
"learning_rate": 7.430673355562847e-06,
"loss": 0.1263098359107971,
"step": 2010
},
{
"epoch": 1.2205882352941178,
"grad_norm": 0.23898456990718842,
"learning_rate": 7.3998907474544294e-06,
"loss": 0.1227832555770874,
"step": 2020
},
{
"epoch": 1.2266317485898468,
"grad_norm": 0.2461117058992386,
"learning_rate": 7.368989490614744e-06,
"loss": 0.12333790063858033,
"step": 2030
},
{
"epoch": 1.232675261885576,
"grad_norm": 0.22805091738700867,
"learning_rate": 7.337971112777891e-06,
"loss": 0.12160630226135254,
"step": 2040
},
{
"epoch": 1.2387187751813054,
"grad_norm": 0.21315939724445343,
"learning_rate": 7.306837147468331e-06,
"loss": 0.1259782910346985,
"step": 2050
},
{
"epoch": 1.2447622884770346,
"grad_norm": 0.2278059870004654,
"learning_rate": 7.275589133925085e-06,
"loss": 0.1212125301361084,
"step": 2060
},
{
"epoch": 1.250805801772764,
"grad_norm": 0.2272907793521881,
"learning_rate": 7.244228617025633e-06,
"loss": 0.12110664844512939,
"step": 2070
},
{
"epoch": 1.2568493150684932,
"grad_norm": 0.22618329524993896,
"learning_rate": 7.212757147209525e-06,
"loss": 0.1240962266921997,
"step": 2080
},
{
"epoch": 1.2628928283642225,
"grad_norm": 0.2161119282245636,
"learning_rate": 7.181176280401745e-06,
"loss": 0.12139427661895752,
"step": 2090
},
{
"epoch": 1.2689363416599517,
"grad_norm": 0.20182672142982483,
"learning_rate": 7.149487577935773e-06,
"loss": 0.12184669971466064,
"step": 2100
},
{
"epoch": 1.274979854955681,
"grad_norm": 0.22437644004821777,
"learning_rate": 7.117692606476405e-06,
"loss": 0.12244476079940796,
"step": 2110
},
{
"epoch": 1.28102336825141,
"grad_norm": 0.22313347458839417,
"learning_rate": 7.08579293794229e-06,
"loss": 0.12389321327209472,
"step": 2120
},
{
"epoch": 1.2870668815471393,
"grad_norm": 0.22929154336452484,
"learning_rate": 7.053790149428223e-06,
"loss": 0.1189534306526184,
"step": 2130
},
{
"epoch": 1.2931103948428686,
"grad_norm": 0.22624798119068146,
"learning_rate": 7.021685823127163e-06,
"loss": 0.12425057888031006,
"step": 2140
},
{
"epoch": 1.2991539081385979,
"grad_norm": 0.2135779857635498,
"learning_rate": 6.989481546252027e-06,
"loss": 0.12051645517349244,
"step": 2150
},
{
"epoch": 1.3051974214343272,
"grad_norm": 0.20884054899215698,
"learning_rate": 6.957178910957203e-06,
"loss": 0.12027026414871216,
"step": 2160
},
{
"epoch": 1.3112409347300564,
"grad_norm": 0.20178496837615967,
"learning_rate": 6.924779514259849e-06,
"loss": 0.11570076942443848,
"step": 2170
},
{
"epoch": 1.3172844480257857,
"grad_norm": 0.22677868604660034,
"learning_rate": 6.892284957960926e-06,
"loss": 0.12076947689056397,
"step": 2180
},
{
"epoch": 1.323327961321515,
"grad_norm": 0.27102935314178467,
"learning_rate": 6.859696848566017e-06,
"loss": 0.12137305736541748,
"step": 2190
},
{
"epoch": 1.3293714746172443,
"grad_norm": 0.2308349609375,
"learning_rate": 6.827016797205892e-06,
"loss": 0.12035479545593261,
"step": 2200
},
{
"epoch": 1.3354149879129733,
"grad_norm": 0.2503156363964081,
"learning_rate": 6.794246419556866e-06,
"loss": 0.1176835060119629,
"step": 2210
},
{
"epoch": 1.3414585012087026,
"grad_norm": 0.9529595375061035,
"learning_rate": 6.76138733576091e-06,
"loss": 0.12373212575912476,
"step": 2220
},
{
"epoch": 1.3475020145044319,
"grad_norm": 0.23353809118270874,
"learning_rate": 6.728441170345559e-06,
"loss": 0.120170259475708,
"step": 2230
},
{
"epoch": 1.3535455278001611,
"grad_norm": 0.21173524856567383,
"learning_rate": 6.695409552143599e-06,
"loss": 0.12025713920593262,
"step": 2240
},
{
"epoch": 1.3595890410958904,
"grad_norm": 0.2226538062095642,
"learning_rate": 6.662294114212533e-06,
"loss": 0.11859447956085205,
"step": 2250
},
{
"epoch": 1.3656325543916197,
"grad_norm": 0.23403026163578033,
"learning_rate": 6.629096493753844e-06,
"loss": 0.1168973207473755,
"step": 2260
},
{
"epoch": 1.371676067687349,
"grad_norm": 0.2118721455335617,
"learning_rate": 6.595818332032058e-06,
"loss": 0.12062954902648926,
"step": 2270
},
{
"epoch": 1.3777195809830782,
"grad_norm": 0.22576157748699188,
"learning_rate": 6.562461274293599e-06,
"loss": 0.1222001314163208,
"step": 2280
},
{
"epoch": 1.3837630942788075,
"grad_norm": 0.22125409543514252,
"learning_rate": 6.529026969685446e-06,
"loss": 0.11852601766586304,
"step": 2290
},
{
"epoch": 1.3898066075745366,
"grad_norm": 0.22153203189373016,
"learning_rate": 6.495517071173607e-06,
"loss": 0.11841906309127807,
"step": 2300
},
{
"epoch": 1.3958501208702658,
"grad_norm": 0.21368172764778137,
"learning_rate": 6.461933235461388e-06,
"loss": 0.11996223926544189,
"step": 2310
},
{
"epoch": 1.401893634165995,
"grad_norm": 0.23732133209705353,
"learning_rate": 6.428277122907498e-06,
"loss": 0.12129522562026977,
"step": 2320
},
{
"epoch": 1.4079371474617244,
"grad_norm": 0.20500251650810242,
"learning_rate": 6.394550397443953e-06,
"loss": 0.1194156527519226,
"step": 2330
},
{
"epoch": 1.4139806607574537,
"grad_norm": 0.22899705171585083,
"learning_rate": 6.360754726493821e-06,
"loss": 0.11767797470092774,
"step": 2340
},
{
"epoch": 1.420024174053183,
"grad_norm": 0.23209087550640106,
"learning_rate": 6.326891780888775e-06,
"loss": 0.1170041799545288,
"step": 2350
},
{
"epoch": 1.4260676873489122,
"grad_norm": 0.24385273456573486,
"learning_rate": 6.2929632347865e-06,
"loss": 0.11813467741012573,
"step": 2360
},
{
"epoch": 1.4321112006446415,
"grad_norm": 0.20373444259166718,
"learning_rate": 6.2589707655879165e-06,
"loss": 0.11844906806945801,
"step": 2370
},
{
"epoch": 1.4381547139403708,
"grad_norm": 0.22466544806957245,
"learning_rate": 6.224916053854252e-06,
"loss": 0.11670351028442383,
"step": 2380
},
{
"epoch": 1.4441982272360998,
"grad_norm": 0.20104901492595673,
"learning_rate": 6.190800783223959e-06,
"loss": 0.11748621463775635,
"step": 2390
},
{
"epoch": 1.4502417405318293,
"grad_norm": 0.21001137793064117,
"learning_rate": 6.156626640329471e-06,
"loss": 0.11688747406005859,
"step": 2400
},
{
"epoch": 1.4562852538275584,
"grad_norm": 0.23640118539333344,
"learning_rate": 6.122395314713825e-06,
"loss": 0.1174784779548645,
"step": 2410
},
{
"epoch": 1.4623287671232876,
"grad_norm": 0.20448064804077148,
"learning_rate": 6.088108498747123e-06,
"loss": 0.11605008840560913,
"step": 2420
},
{
"epoch": 1.468372280419017,
"grad_norm": 0.21227361261844635,
"learning_rate": 6.0537678875428715e-06,
"loss": 0.1141992449760437,
"step": 2430
},
{
"epoch": 1.4744157937147462,
"grad_norm": 0.21395207941532135,
"learning_rate": 6.019375178874166e-06,
"loss": 0.11088359355926514,
"step": 2440
},
{
"epoch": 1.4804593070104755,
"grad_norm": 0.20509681105613708,
"learning_rate": 5.984932073089767e-06,
"loss": 0.11901907920837403,
"step": 2450
},
{
"epoch": 1.4865028203062047,
"grad_norm": 0.23024727404117584,
"learning_rate": 5.950440273030019e-06,
"loss": 0.11589326858520507,
"step": 2460
},
{
"epoch": 1.492546333601934,
"grad_norm": 0.22049815952777863,
"learning_rate": 5.915901483942687e-06,
"loss": 0.11147205829620362,
"step": 2470
},
{
"epoch": 1.498589846897663,
"grad_norm": 0.24279910326004028,
"learning_rate": 5.881317413398624e-06,
"loss": 0.1184835433959961,
"step": 2480
},
{
"epoch": 1.5046333601933926,
"grad_norm": 0.2349153608083725,
"learning_rate": 5.846689771207372e-06,
"loss": 0.11466156244277954,
"step": 2490
},
{
"epoch": 1.5106768734891216,
"grad_norm": 0.22500401735305786,
"learning_rate": 5.8120202693326164e-06,
"loss": 0.11465930938720703,
"step": 2500
},
{
"epoch": 1.5167203867848509,
"grad_norm": 0.2286197394132614,
"learning_rate": 5.777310621807556e-06,
"loss": 0.11652610301971436,
"step": 2510
},
{
"epoch": 1.5227639000805802,
"grad_norm": 0.23102812469005585,
"learning_rate": 5.7425625446501575e-06,
"loss": 0.11454565525054931,
"step": 2520
},
{
"epoch": 1.5288074133763094,
"grad_norm": 0.2146225869655609,
"learning_rate": 5.707777755778319e-06,
"loss": 0.11421661376953125,
"step": 2530
},
{
"epoch": 1.5348509266720387,
"grad_norm": 0.21997742354869843,
"learning_rate": 5.672957974924935e-06,
"loss": 0.11838579177856445,
"step": 2540
},
{
"epoch": 1.540894439967768,
"grad_norm": 0.21023927628993988,
"learning_rate": 5.638104923552884e-06,
"loss": 0.111042058467865,
"step": 2550
},
{
"epoch": 1.5469379532634973,
"grad_norm": 0.21510760486125946,
"learning_rate": 5.603220324769901e-06,
"loss": 0.11168649196624755,
"step": 2560
},
{
"epoch": 1.5529814665592263,
"grad_norm": 0.218896746635437,
"learning_rate": 5.568305903243413e-06,
"loss": 0.11467195749282837,
"step": 2570
},
{
"epoch": 1.5590249798549558,
"grad_norm": 0.20666085183620453,
"learning_rate": 5.533363385115251e-06,
"loss": 0.11230006217956542,
"step": 2580
},
{
"epoch": 1.5650684931506849,
"grad_norm": 0.21789900958538055,
"learning_rate": 5.498394497916324e-06,
"loss": 0.11302851438522339,
"step": 2590
},
{
"epoch": 1.5711120064464141,
"grad_norm": 0.22432194650173187,
"learning_rate": 5.463400970481206e-06,
"loss": 0.1176527500152588,
"step": 2600
},
{
"epoch": 1.5771555197421434,
"grad_norm": 0.21380209922790527,
"learning_rate": 5.428384532862667e-06,
"loss": 0.11611467599868774,
"step": 2610
},
{
"epoch": 1.5831990330378727,
"grad_norm": 0.2121361345052719,
"learning_rate": 5.393346916246139e-06,
"loss": 0.1127274751663208,
"step": 2620
},
{
"epoch": 1.589242546333602,
"grad_norm": 0.23556004464626312,
"learning_rate": 5.358289852864123e-06,
"loss": 0.11342973709106445,
"step": 2630
},
{
"epoch": 1.5952860596293312,
"grad_norm": 0.22848543524742126,
"learning_rate": 5.323215075910556e-06,
"loss": 0.11354776620864868,
"step": 2640
},
{
"epoch": 1.6013295729250605,
"grad_norm": 0.22549968957901,
"learning_rate": 5.288124319455121e-06,
"loss": 0.11867868900299072,
"step": 2650
},
{
"epoch": 1.6073730862207896,
"grad_norm": 0.20736844837665558,
"learning_rate": 5.253019318357515e-06,
"loss": 0.11953473091125488,
"step": 2660
},
{
"epoch": 1.613416599516519,
"grad_norm": 0.21321389079093933,
"learning_rate": 5.217901808181675e-06,
"loss": 0.11560553312301636,
"step": 2670
},
{
"epoch": 1.619460112812248,
"grad_norm": 0.20952194929122925,
"learning_rate": 5.182773525109983e-06,
"loss": 0.11379563808441162,
"step": 2680
},
{
"epoch": 1.6255036261079774,
"grad_norm": 0.2412755936384201,
"learning_rate": 5.1476362058574205e-06,
"loss": 0.11462094783782958,
"step": 2690
},
{
"epoch": 1.6315471394037067,
"grad_norm": 0.21880988776683807,
"learning_rate": 5.112491587585712e-06,
"loss": 0.11484885215759277,
"step": 2700
},
{
"epoch": 1.637590652699436,
"grad_norm": 0.24393679201602936,
"learning_rate": 5.077341407817439e-06,
"loss": 0.11311535835266114,
"step": 2710
},
{
"epoch": 1.6436341659951652,
"grad_norm": 0.21137407422065735,
"learning_rate": 5.042187404350142e-06,
"loss": 0.1140148401260376,
"step": 2720
},
{
"epoch": 1.6496776792908945,
"grad_norm": 0.26122796535491943,
"learning_rate": 5.007031315170397e-06,
"loss": 0.10838596820831299,
"step": 2730
},
{
"epoch": 1.6557211925866238,
"grad_norm": 0.21811725199222565,
"learning_rate": 4.971874878367902e-06,
"loss": 0.11355844736099244,
"step": 2740
},
{
"epoch": 1.6617647058823528,
"grad_norm": 0.25574809312820435,
"learning_rate": 4.936719832049537e-06,
"loss": 0.11144473552703857,
"step": 2750
},
{
"epoch": 1.6678082191780823,
"grad_norm": 0.23158828914165497,
"learning_rate": 4.901567914253439e-06,
"loss": 0.11145066022872925,
"step": 2760
},
{
"epoch": 1.6738517324738114,
"grad_norm": 0.2101062536239624,
"learning_rate": 4.8664208628630726e-06,
"loss": 0.11440601348876953,
"step": 2770
},
{
"epoch": 1.6798952457695409,
"grad_norm": 0.23863664269447327,
"learning_rate": 4.831280415521312e-06,
"loss": 0.11194591522216797,
"step": 2780
},
{
"epoch": 1.68593875906527,
"grad_norm": 0.24111558496952057,
"learning_rate": 4.796148309544532e-06,
"loss": 0.11374534368515014,
"step": 2790
},
{
"epoch": 1.6919822723609992,
"grad_norm": 0.2214035540819168,
"learning_rate": 4.761026281836718e-06,
"loss": 0.1109438419342041,
"step": 2800
},
{
"epoch": 1.6980257856567285,
"grad_norm": 0.23886463046073914,
"learning_rate": 4.725916068803589e-06,
"loss": 0.10815801620483398,
"step": 2810
},
{
"epoch": 1.7040692989524577,
"grad_norm": 0.24336574971675873,
"learning_rate": 4.690819406266763e-06,
"loss": 0.111271333694458,
"step": 2820
},
{
"epoch": 1.710112812248187,
"grad_norm": 0.25888592004776,
"learning_rate": 4.6557380293779266e-06,
"loss": 0.11122057437896729,
"step": 2830
},
{
"epoch": 1.716156325543916,
"grad_norm": 0.21733474731445312,
"learning_rate": 4.620673672533057e-06,
"loss": 0.1081722617149353,
"step": 2840
},
{
"epoch": 1.7221998388396456,
"grad_norm": 0.2514082193374634,
"learning_rate": 4.585628069286671e-06,
"loss": 0.11077737808227539,
"step": 2850
},
{
"epoch": 1.7282433521353746,
"grad_norm": 0.22327256202697754,
"learning_rate": 4.5506029522661325e-06,
"loss": 0.1123847484588623,
"step": 2860
},
{
"epoch": 1.734286865431104,
"grad_norm": 0.23829680681228638,
"learning_rate": 4.515600053085969e-06,
"loss": 0.10773321390151977,
"step": 2870
},
{
"epoch": 1.7403303787268332,
"grad_norm": 0.25309762358665466,
"learning_rate": 4.480621102262283e-06,
"loss": 0.11210131645202637,
"step": 2880
},
{
"epoch": 1.7463738920225624,
"grad_norm": 0.22266383469104767,
"learning_rate": 4.445667829127185e-06,
"loss": 0.10769500732421874,
"step": 2890
},
{
"epoch": 1.7524174053182917,
"grad_norm": 0.23095478117465973,
"learning_rate": 4.410741961743306e-06,
"loss": 0.11072452068328857,
"step": 2900
},
{
"epoch": 1.758460918614021,
"grad_norm": 0.2459106594324112,
"learning_rate": 4.375845226818354e-06,
"loss": 0.1095454216003418,
"step": 2910
},
{
"epoch": 1.7645044319097503,
"grad_norm": 0.23950867354869843,
"learning_rate": 4.34097934961975e-06,
"loss": 0.10627870559692383,
"step": 2920
},
{
"epoch": 1.7705479452054793,
"grad_norm": 0.24852554500102997,
"learning_rate": 4.306146053889336e-06,
"loss": 0.10963752269744872,
"step": 2930
},
{
"epoch": 1.7765914585012088,
"grad_norm": 0.20720911026000977,
"learning_rate": 4.271347061758146e-06,
"loss": 0.10962324142456055,
"step": 2940
},
{
"epoch": 1.7826349717969379,
"grad_norm": 0.2429235428571701,
"learning_rate": 4.23658409366128e-06,
"loss": 0.10711297988891602,
"step": 2950
},
{
"epoch": 1.7886784850926674,
"grad_norm": 0.25143516063690186,
"learning_rate": 4.201858868252832e-06,
"loss": 0.11143739223480224,
"step": 2960
},
{
"epoch": 1.7947219983883964,
"grad_norm": 0.28281307220458984,
"learning_rate": 4.167173102320925e-06,
"loss": 0.10768698453903199,
"step": 2970
},
{
"epoch": 1.8007655116841257,
"grad_norm": 0.262405127286911,
"learning_rate": 4.132528510702839e-06,
"loss": 0.11116191148757934,
"step": 2980
},
{
"epoch": 1.806809024979855,
"grad_norm": 0.24316293001174927,
"learning_rate": 4.0979268062002295e-06,
"loss": 0.11136236190795898,
"step": 2990
},
{
"epoch": 1.8128525382755842,
"grad_norm": 0.2162204533815384,
"learning_rate": 4.063369699494443e-06,
"loss": 0.10728955268859863,
"step": 3000
},
{
"epoch": 1.8188960515713135,
"grad_norm": 0.22905303537845612,
"learning_rate": 4.0288588990619475e-06,
"loss": 0.11171340942382812,
"step": 3010
},
{
"epoch": 1.8249395648670426,
"grad_norm": 0.25043120980262756,
"learning_rate": 3.9943961110898645e-06,
"loss": 0.10901223421096802,
"step": 3020
},
{
"epoch": 1.830983078162772,
"grad_norm": 0.23680433630943298,
"learning_rate": 3.959983039391617e-06,
"loss": 0.1083879828453064,
"step": 3030
},
{
"epoch": 1.8370265914585011,
"grad_norm": 0.2268252670764923,
"learning_rate": 3.925621385322693e-06,
"loss": 0.10923205614089966,
"step": 3040
},
{
"epoch": 1.8430701047542306,
"grad_norm": 0.23718592524528503,
"learning_rate": 3.8913128476965356e-06,
"loss": 0.10960838794708253,
"step": 3050
},
{
"epoch": 1.8491136180499597,
"grad_norm": 0.2261449545621872,
"learning_rate": 3.857059122700548e-06,
"loss": 0.10592516660690307,
"step": 3060
},
{
"epoch": 1.855157131345689,
"grad_norm": 0.2755502462387085,
"learning_rate": 3.822861903812244e-06,
"loss": 0.11440458297729492,
"step": 3070
},
{
"epoch": 1.8612006446414182,
"grad_norm": 0.23894339799880981,
"learning_rate": 3.788722881715518e-06,
"loss": 0.105678129196167,
"step": 3080
},
{
"epoch": 1.8672441579371475,
"grad_norm": 0.24882644414901733,
"learning_rate": 3.7546437442170593e-06,
"loss": 0.10813140869140625,
"step": 3090
},
{
"epoch": 1.8732876712328768,
"grad_norm": 0.2674005925655365,
"learning_rate": 3.720626176162909e-06,
"loss": 0.10488420724868774,
"step": 3100
},
{
"epoch": 1.8793311845286058,
"grad_norm": 0.2316111922264099,
"learning_rate": 3.686671859355162e-06,
"loss": 0.10722472667694091,
"step": 3110
},
{
"epoch": 1.8853746978243353,
"grad_norm": 0.2484072744846344,
"learning_rate": 3.6527824724688267e-06,
"loss": 0.10736773014068604,
"step": 3120
},
{
"epoch": 1.8914182111200644,
"grad_norm": 0.20056864619255066,
"learning_rate": 3.6189596909688194e-06,
"loss": 0.10886975526809692,
"step": 3130
},
{
"epoch": 1.8974617244157939,
"grad_norm": 0.23614878952503204,
"learning_rate": 3.585205187027142e-06,
"loss": 0.10807681083679199,
"step": 3140
},
{
"epoch": 1.903505237711523,
"grad_norm": 0.24152705073356628,
"learning_rate": 3.5515206294402026e-06,
"loss": 0.10454709529876709,
"step": 3150
},
{
"epoch": 1.9095487510072522,
"grad_norm": 0.23608173429965973,
"learning_rate": 3.517907683546325e-06,
"loss": 0.10878932476043701,
"step": 3160
},
{
"epoch": 1.9155922643029815,
"grad_norm": 0.23896652460098267,
"learning_rate": 3.484368011143398e-06,
"loss": 0.10882582664489746,
"step": 3170
},
{
"epoch": 1.9216357775987107,
"grad_norm": 0.22567640244960785,
"learning_rate": 3.450903270406729e-06,
"loss": 0.10648589134216309,
"step": 3180
},
{
"epoch": 1.92767929089444,
"grad_norm": 0.21789751946926117,
"learning_rate": 3.4175151158070595e-06,
"loss": 0.10857722759246827,
"step": 3190
},
{
"epoch": 1.933722804190169,
"grad_norm": 0.2438836544752121,
"learning_rate": 3.3842051980287794e-06,
"loss": 0.10621727705001831,
"step": 3200
},
{
"epoch": 1.9397663174858986,
"grad_norm": 0.23827660083770752,
"learning_rate": 3.3509751638883038e-06,
"loss": 0.10356340408325196,
"step": 3210
},
{
"epoch": 1.9458098307816276,
"grad_norm": 0.23436710238456726,
"learning_rate": 3.3178266562526644e-06,
"loss": 0.10494379997253418,
"step": 3220
},
{
"epoch": 1.951853344077357,
"grad_norm": 0.2269088178873062,
"learning_rate": 3.284761313958286e-06,
"loss": 0.10738768577575683,
"step": 3230
},
{
"epoch": 1.9578968573730862,
"grad_norm": 0.21747054159641266,
"learning_rate": 3.251780771729968e-06,
"loss": 0.10422486066818237,
"step": 3240
},
{
"epoch": 1.9639403706688154,
"grad_norm": 0.22467920184135437,
"learning_rate": 3.218886660100055e-06,
"loss": 0.10968996286392212,
"step": 3250
},
{
"epoch": 1.9699838839645447,
"grad_norm": 0.24414922297000885,
"learning_rate": 3.186080605327831e-06,
"loss": 0.10581851005554199,
"step": 3260
},
{
"epoch": 1.976027397260274,
"grad_norm": 0.23211994767189026,
"learning_rate": 3.1533642293191204e-06,
"loss": 0.10720672607421874,
"step": 3270
},
{
"epoch": 1.9820709105560033,
"grad_norm": 0.24336817860603333,
"learning_rate": 3.120739149546095e-06,
"loss": 0.10917339324951172,
"step": 3280
},
{
"epoch": 1.9881144238517323,
"grad_norm": 0.2525531053543091,
"learning_rate": 3.088206978967315e-06,
"loss": 0.10552664995193481,
"step": 3290
},
{
"epoch": 1.9941579371474618,
"grad_norm": 0.23478975892066956,
"learning_rate": 3.0557693259479836e-06,
"loss": 0.1081237554550171,
"step": 3300
},
{
"epoch": 2.0,
"grad_norm": 0.28233903646469116,
"learning_rate": 3.0234277941804247e-06,
"loss": 0.10643197298049926,
"step": 3310
},
{
"epoch": 2.0,
"eval_loss": 0.11082334816455841,
"eval_runtime": 73.8328,
"eval_samples_per_second": 10.876,
"eval_steps_per_second": 0.691,
"step": 3310
},
{
"epoch": 2.006043513295729,
"grad_norm": 0.25113484263420105,
"learning_rate": 2.991183982604804e-06,
"loss": 0.0850183367729187,
"step": 3320
},
{
"epoch": 2.0120870265914585,
"grad_norm": 0.23909488320350647,
"learning_rate": 2.959039485330082e-06,
"loss": 0.08990187644958496,
"step": 3330
},
{
"epoch": 2.0181305398871876,
"grad_norm": 0.24318291246891022,
"learning_rate": 2.92699589155519e-06,
"loss": 0.08521405458450318,
"step": 3340
},
{
"epoch": 2.024174053182917,
"grad_norm": 0.2429560273885727,
"learning_rate": 2.895054785490474e-06,
"loss": 0.0877398431301117,
"step": 3350
},
{
"epoch": 2.030217566478646,
"grad_norm": 0.2468767911195755,
"learning_rate": 2.8632177462793605e-06,
"loss": 0.08729259371757507,
"step": 3360
},
{
"epoch": 2.0362610797743756,
"grad_norm": 0.2928377687931061,
"learning_rate": 2.8314863479203026e-06,
"loss": 0.08801011443138122,
"step": 3370
},
{
"epoch": 2.0423045930701047,
"grad_norm": 0.2536792755126953,
"learning_rate": 2.7998621591889423e-06,
"loss": 0.0870097041130066,
"step": 3380
},
{
"epoch": 2.048348106365834,
"grad_norm": 0.3121541440486908,
"learning_rate": 2.768346743560565e-06,
"loss": 0.08584122657775879,
"step": 3390
},
{
"epoch": 2.0543916196615633,
"grad_norm": 0.2636508643627167,
"learning_rate": 2.736941659132799e-06,
"loss": 0.08660758137702942,
"step": 3400
},
{
"epoch": 2.0604351329572923,
"grad_norm": 0.24208393692970276,
"learning_rate": 2.705648458548584e-06,
"loss": 0.08983132839202881,
"step": 3410
},
{
"epoch": 2.066478646253022,
"grad_norm": 0.2726794183254242,
"learning_rate": 2.674468688919405e-06,
"loss": 0.08874686360359192,
"step": 3420
},
{
"epoch": 2.072522159548751,
"grad_norm": 0.29161110520362854,
"learning_rate": 2.6434038917488163e-06,
"loss": 0.08659718036651612,
"step": 3430
},
{
"epoch": 2.0785656728444803,
"grad_norm": 0.2472435086965561,
"learning_rate": 2.6124556028562174e-06,
"loss": 0.09049414396286011,
"step": 3440
},
{
"epoch": 2.0846091861402094,
"grad_norm": 0.25984734296798706,
"learning_rate": 2.5816253523009305e-06,
"loss": 0.0886120080947876,
"step": 3450
},
{
"epoch": 2.090652699435939,
"grad_norm": 0.24759183824062347,
"learning_rate": 2.5509146643065563e-06,
"loss": 0.09088815450668335,
"step": 3460
},
{
"epoch": 2.096696212731668,
"grad_norm": 0.22352154552936554,
"learning_rate": 2.520325057185614e-06,
"loss": 0.0859030842781067,
"step": 3470
},
{
"epoch": 2.1027397260273974,
"grad_norm": 0.24618713557720184,
"learning_rate": 2.489858043264479e-06,
"loss": 0.08588383793830871,
"step": 3480
},
{
"epoch": 2.1087832393231265,
"grad_norm": 0.25341397523880005,
"learning_rate": 2.4595151288086104e-06,
"loss": 0.08556797504425048,
"step": 3490
},
{
"epoch": 2.1148267526188556,
"grad_norm": 0.2574682831764221,
"learning_rate": 2.4292978139480956e-06,
"loss": 0.08930388689041138,
"step": 3500
},
{
"epoch": 2.120870265914585,
"grad_norm": 0.2640034556388855,
"learning_rate": 2.3992075926034663e-06,
"loss": 0.08774279952049255,
"step": 3510
},
{
"epoch": 2.126913779210314,
"grad_norm": 0.24985980987548828,
"learning_rate": 2.369245952411855e-06,
"loss": 0.08940563797950744,
"step": 3520
},
{
"epoch": 2.1329572925060436,
"grad_norm": 0.26639437675476074,
"learning_rate": 2.3394143746534383e-06,
"loss": 0.08724412322044373,
"step": 3530
},
{
"epoch": 2.1390008058017727,
"grad_norm": 0.2554488480091095,
"learning_rate": 2.3097143341782126e-06,
"loss": 0.09025406837463379,
"step": 3540
},
{
"epoch": 2.145044319097502,
"grad_norm": 0.2665922939777374,
"learning_rate": 2.280147299333068e-06,
"loss": 0.08426840901374817,
"step": 3550
},
{
"epoch": 2.151087832393231,
"grad_norm": 0.2911239564418793,
"learning_rate": 2.250714731889203e-06,
"loss": 0.08560506105422974,
"step": 3560
},
{
"epoch": 2.1571313456889607,
"grad_norm": 0.26575127243995667,
"learning_rate": 2.221418086969849e-06,
"loss": 0.08637800216674804,
"step": 3570
},
{
"epoch": 2.1631748589846898,
"grad_norm": 0.24241304397583008,
"learning_rate": 2.1922588129783403e-06,
"loss": 0.08939040899276733,
"step": 3580
},
{
"epoch": 2.169218372280419,
"grad_norm": 0.25169071555137634,
"learning_rate": 2.163238351526492e-06,
"loss": 0.08800272941589356,
"step": 3590
},
{
"epoch": 2.1752618855761483,
"grad_norm": 0.2451711893081665,
"learning_rate": 2.1343581373633394e-06,
"loss": 0.08782594799995422,
"step": 3600
},
{
"epoch": 2.1813053988718774,
"grad_norm": 0.24838466942310333,
"learning_rate": 2.1056195983041987e-06,
"loss": 0.08553072214126586,
"step": 3610
},
{
"epoch": 2.187348912167607,
"grad_norm": 0.2609288692474365,
"learning_rate": 2.077024155160079e-06,
"loss": 0.08314642310142517,
"step": 3620
},
{
"epoch": 2.193392425463336,
"grad_norm": 0.2765029966831207,
"learning_rate": 2.048573221667444e-06,
"loss": 0.08574669361114502,
"step": 3630
},
{
"epoch": 2.1994359387590654,
"grad_norm": 0.2791488468647003,
"learning_rate": 2.0202682044183066e-06,
"loss": 0.08509529232978821,
"step": 3640
},
{
"epoch": 2.2054794520547945,
"grad_norm": 0.2832431495189667,
"learning_rate": 1.992110502790696e-06,
"loss": 0.08729273676872254,
"step": 3650
},
{
"epoch": 2.211522965350524,
"grad_norm": 0.2647612690925598,
"learning_rate": 1.964101508879471e-06,
"loss": 0.08810110092163086,
"step": 3660
},
{
"epoch": 2.217566478646253,
"grad_norm": 0.2782638669013977,
"learning_rate": 1.9362426074275003e-06,
"loss": 0.08275524973869323,
"step": 3670
},
{
"epoch": 2.2236099919419825,
"grad_norm": 0.251492977142334,
"learning_rate": 1.9085351757571935e-06,
"loss": 0.08382174968719483,
"step": 3680
},
{
"epoch": 2.2296535052377116,
"grad_norm": 0.258375346660614,
"learning_rate": 1.8809805837024148e-06,
"loss": 0.0837257981300354,
"step": 3690
},
{
"epoch": 2.2356970185334406,
"grad_norm": 0.26384153962135315,
"learning_rate": 1.8535801935407538e-06,
"loss": 0.08655069470405578,
"step": 3700
},
{
"epoch": 2.24174053182917,
"grad_norm": 0.2561194896697998,
"learning_rate": 1.826335359926184e-06,
"loss": 0.0858405888080597,
"step": 3710
},
{
"epoch": 2.247784045124899,
"grad_norm": 0.29886385798454285,
"learning_rate": 1.7992474298220792e-06,
"loss": 0.0856437087059021,
"step": 3720
},
{
"epoch": 2.2538275584206287,
"grad_norm": 0.2630106806755066,
"learning_rate": 1.7723177424346272e-06,
"loss": 0.08493099212646485,
"step": 3730
},
{
"epoch": 2.2598710717163577,
"grad_norm": 0.24556013941764832,
"learning_rate": 1.7455476291466177e-06,
"loss": 0.0812730073928833,
"step": 3740
},
{
"epoch": 2.265914585012087,
"grad_norm": 0.2617291808128357,
"learning_rate": 1.7189384134516262e-06,
"loss": 0.08479050397872925,
"step": 3750
},
{
"epoch": 2.2719580983078163,
"grad_norm": 0.26498010754585266,
"learning_rate": 1.6924914108885709e-06,
"loss": 0.08670079112052917,
"step": 3760
},
{
"epoch": 2.2780016116035453,
"grad_norm": 0.2495623528957367,
"learning_rate": 1.6662079289766802e-06,
"loss": 0.08470574617385865,
"step": 3770
},
{
"epoch": 2.284045124899275,
"grad_norm": 0.27399682998657227,
"learning_rate": 1.6400892671508505e-06,
"loss": 0.08199700117111205,
"step": 3780
},
{
"epoch": 2.290088638195004,
"grad_norm": 0.26449716091156006,
"learning_rate": 1.6141367166973975e-06,
"loss": 0.08422369956970215,
"step": 3790
},
{
"epoch": 2.2961321514907334,
"grad_norm": 0.2562250792980194,
"learning_rate": 1.5883515606902272e-06,
"loss": 0.08601263761520386,
"step": 3800
},
{
"epoch": 2.3021756647864624,
"grad_norm": 0.2867155075073242,
"learning_rate": 1.5627350739273873e-06,
"loss": 0.08274335861206054,
"step": 3810
},
{
"epoch": 2.308219178082192,
"grad_norm": 0.25865477323532104,
"learning_rate": 1.5372885228680518e-06,
"loss": 0.08601617813110352,
"step": 3820
},
{
"epoch": 2.314262691377921,
"grad_norm": 0.2661864757537842,
"learning_rate": 1.512013165569905e-06,
"loss": 0.08428186178207397,
"step": 3830
},
{
"epoch": 2.3203062046736505,
"grad_norm": 0.2708439230918884,
"learning_rate": 1.4869102516269479e-06,
"loss": 0.08554589748382568,
"step": 3840
},
{
"epoch": 2.3263497179693795,
"grad_norm": 0.2581521272659302,
"learning_rate": 1.4619810221077135e-06,
"loss": 0.08260358572006225,
"step": 3850
},
{
"epoch": 2.332393231265109,
"grad_norm": 0.2732486128807068,
"learning_rate": 1.4372267094939113e-06,
"loss": 0.08420640230178833,
"step": 3860
},
{
"epoch": 2.338436744560838,
"grad_norm": 0.26530545949935913,
"learning_rate": 1.4126485376194954e-06,
"loss": 0.08369795680046081,
"step": 3870
},
{
"epoch": 2.344480257856567,
"grad_norm": 0.2732124924659729,
"learning_rate": 1.3882477216101626e-06,
"loss": 0.08537374138832092,
"step": 3880
},
{
"epoch": 2.3505237711522966,
"grad_norm": 0.2569216787815094,
"learning_rate": 1.3640254678232696e-06,
"loss": 0.08633752465248108,
"step": 3890
},
{
"epoch": 2.3565672844480257,
"grad_norm": 0.25759920477867126,
"learning_rate": 1.3399829737881953e-06,
"loss": 0.08359290361404419,
"step": 3900
},
{
"epoch": 2.362610797743755,
"grad_norm": 0.26242437958717346,
"learning_rate": 1.3161214281471406e-06,
"loss": 0.08388763666152954,
"step": 3910
},
{
"epoch": 2.368654311039484,
"grad_norm": 0.2549450099468231,
"learning_rate": 1.2924420105963549e-06,
"loss": 0.08403220176696777,
"step": 3920
},
{
"epoch": 2.3746978243352137,
"grad_norm": 0.2518737316131592,
"learning_rate": 1.2689458918278158e-06,
"loss": 0.07982347011566163,
"step": 3930
},
{
"epoch": 2.3807413376309428,
"grad_norm": 0.2545059025287628,
"learning_rate": 1.2456342334713566e-06,
"loss": 0.08241246342658996,
"step": 3940
},
{
"epoch": 2.386784850926672,
"grad_norm": 0.24779869616031647,
"learning_rate": 1.2225081880372275e-06,
"loss": 0.08269934058189392,
"step": 3950
},
{
"epoch": 2.3928283642224013,
"grad_norm": 0.29499658942222595,
"learning_rate": 1.1995688988591186e-06,
"loss": 0.08604642152786254,
"step": 3960
},
{
"epoch": 2.3988718775181304,
"grad_norm": 0.25372618436813354,
"learning_rate": 1.176817500037642e-06,
"loss": 0.08472075462341308,
"step": 3970
},
{
"epoch": 2.40491539081386,
"grad_norm": 0.27962806820869446,
"learning_rate": 1.1542551163842507e-06,
"loss": 0.08208303451538086,
"step": 3980
},
{
"epoch": 2.410958904109589,
"grad_norm": 0.27031031250953674,
"learning_rate": 1.1318828633656392e-06,
"loss": 0.08391703367233276,
"step": 3990
},
{
"epoch": 2.4170024174053184,
"grad_norm": 0.27157852053642273,
"learning_rate": 1.1097018470485866e-06,
"loss": 0.08238473534584045,
"step": 4000
},
{
"epoch": 2.4230459307010475,
"grad_norm": 0.28021591901779175,
"learning_rate": 1.0877131640452842e-06,
"loss": 0.08168457150459289,
"step": 4010
},
{
"epoch": 2.429089443996777,
"grad_norm": 0.2608916461467743,
"learning_rate": 1.0659179014591103e-06,
"loss": 0.08118799924850464,
"step": 4020
},
{
"epoch": 2.435132957292506,
"grad_norm": 0.3028624355792999,
"learning_rate": 1.0443171368308879e-06,
"loss": 0.08596821427345276,
"step": 4030
},
{
"epoch": 2.4411764705882355,
"grad_norm": 0.2674548923969269,
"learning_rate": 1.0229119380856122e-06,
"loss": 0.08125120997428895,
"step": 4040
},
{
"epoch": 2.4472199838839646,
"grad_norm": 0.25821155309677124,
"learning_rate": 1.0017033634796575e-06,
"loss": 0.07936325073242187,
"step": 4050
},
{
"epoch": 2.4532634971796936,
"grad_norm": 0.294775128364563,
"learning_rate": 9.806924615484492e-07,
"loss": 0.07957918643951416,
"step": 4060
},
{
"epoch": 2.459307010475423,
"grad_norm": 0.2851545512676239,
"learning_rate": 9.5988027105463e-07,
"loss": 0.08183679580688477,
"step": 4070
},
{
"epoch": 2.465350523771152,
"grad_norm": 0.27319052815437317,
"learning_rate": 9.392678209367035e-07,
"loss": 0.08182142376899719,
"step": 4080
},
{
"epoch": 2.4713940370668817,
"grad_norm": 0.26378685235977173,
"learning_rate": 9.188561302581689e-07,
"loss": 0.08690798282623291,
"step": 4090
},
{
"epoch": 2.4774375503626107,
"grad_norm": 0.2635554075241089,
"learning_rate": 8.986462081571296e-07,
"loss": 0.08294034600257874,
"step": 4100
},
{
"epoch": 2.48348106365834,
"grad_norm": 0.27604904770851135,
"learning_rate": 8.78639053796409e-07,
"loss": 0.08500493764877319,
"step": 4110
},
{
"epoch": 2.4895245769540693,
"grad_norm": 0.26117950677871704,
"learning_rate": 8.588356563141531e-07,
"loss": 0.08091460466384888,
"step": 4120
},
{
"epoch": 2.4955680902497983,
"grad_norm": 0.2721405327320099,
"learning_rate": 8.392369947749224e-07,
"loss": 0.08403721451759338,
"step": 4130
},
{
"epoch": 2.501611603545528,
"grad_norm": 0.2595398426055908,
"learning_rate": 8.198440381212974e-07,
"loss": 0.0813600778579712,
"step": 4140
},
{
"epoch": 2.5076551168412573,
"grad_norm": 0.28272923827171326,
"learning_rate": 8.006577451259645e-07,
"loss": 0.07888777852058411,
"step": 4150
},
{
"epoch": 2.5136986301369864,
"grad_norm": 0.31115999817848206,
"learning_rate": 7.816790643443206e-07,
"loss": 0.08272597789764405,
"step": 4160
},
{
"epoch": 2.5197421434327154,
"grad_norm": 0.2709655463695526,
"learning_rate": 7.629089340675767e-07,
"loss": 0.08107831478118896,
"step": 4170
},
{
"epoch": 2.525785656728445,
"grad_norm": 0.27046215534210205,
"learning_rate": 7.443482822763709e-07,
"loss": 0.0840245544910431,
"step": 4180
},
{
"epoch": 2.531829170024174,
"grad_norm": 0.26231053471565247,
"learning_rate": 7.259980265948846e-07,
"loss": 0.08337774872779846,
"step": 4190
},
{
"epoch": 2.5378726833199035,
"grad_norm": 0.2597627639770508,
"learning_rate": 7.078590742454816e-07,
"loss": 0.08387154340744019,
"step": 4200
},
{
"epoch": 2.5439161966156325,
"grad_norm": 0.2701342701911926,
"learning_rate": 6.899323220038512e-07,
"loss": 0.08139055967330933,
"step": 4210
},
{
"epoch": 2.549959709911362,
"grad_norm": 0.2797119617462158,
"learning_rate": 6.722186561546779e-07,
"loss": 0.08418188095092774,
"step": 4220
},
{
"epoch": 2.556003223207091,
"grad_norm": 0.2397763431072235,
"learning_rate": 6.547189524478182e-07,
"loss": 0.08225395679473876,
"step": 4230
},
{
"epoch": 2.56204673650282,
"grad_norm": 0.2523185610771179,
"learning_rate": 6.374340760550068e-07,
"loss": 0.08198237419128418,
"step": 4240
},
{
"epoch": 2.5680902497985496,
"grad_norm": 0.24639897048473358,
"learning_rate": 6.20364881527083e-07,
"loss": 0.08555344939231872,
"step": 4250
},
{
"epoch": 2.5741337630942787,
"grad_norm": 0.2564961016178131,
"learning_rate": 6.035122127517456e-07,
"loss": 0.07994904518127441,
"step": 4260
},
{
"epoch": 2.580177276390008,
"grad_norm": 0.2582924962043762,
"learning_rate": 5.868769029118243e-07,
"loss": 0.08319709300994874,
"step": 4270
},
{
"epoch": 2.586220789685737,
"grad_norm": 0.2715415358543396,
"learning_rate": 5.704597744440949e-07,
"loss": 0.08228959441184998,
"step": 4280
},
{
"epoch": 2.5922643029814667,
"grad_norm": 0.2580704987049103,
"learning_rate": 5.542616389986144e-07,
"loss": 0.08236298561096192,
"step": 4290
},
{
"epoch": 2.5983078162771958,
"grad_norm": 0.24521584808826447,
"learning_rate": 5.382832973985958e-07,
"loss": 0.08108619451522828,
"step": 4300
},
{
"epoch": 2.604351329572925,
"grad_norm": 0.24205389618873596,
"learning_rate": 5.225255396008172e-07,
"loss": 0.08536337018013,
"step": 4310
},
{
"epoch": 2.6103948428686543,
"grad_norm": 0.40597712993621826,
"learning_rate": 5.069891446565617e-07,
"loss": 0.08480771780014038,
"step": 4320
},
{
"epoch": 2.616438356164384,
"grad_norm": 0.2616465985774994,
"learning_rate": 4.916748806731081e-07,
"loss": 0.0817090928554535,
"step": 4330
},
{
"epoch": 2.622481869460113,
"grad_norm": 0.269398033618927,
"learning_rate": 4.7658350477575e-07,
"loss": 0.08234010934829712,
"step": 4340
},
{
"epoch": 2.628525382755842,
"grad_norm": 0.32039719820022583,
"learning_rate": 4.6171576307037226e-07,
"loss": 0.08119097948074341,
"step": 4350
},
{
"epoch": 2.6345688960515714,
"grad_norm": 0.296687513589859,
"learning_rate": 4.4707239060655604e-07,
"loss": 0.07519071102142334,
"step": 4360
},
{
"epoch": 2.6406124093473005,
"grad_norm": 0.2812546193599701,
"learning_rate": 4.3265411134124023e-07,
"loss": 0.08451371788978576,
"step": 4370
},
{
"epoch": 2.64665592264303,
"grad_norm": 0.2701050341129303,
"learning_rate": 4.1846163810293607e-07,
"loss": 0.07716899514198303,
"step": 4380
},
{
"epoch": 2.652699435938759,
"grad_norm": 0.26015597581863403,
"learning_rate": 4.0449567255647637e-07,
"loss": 0.08344717025756836,
"step": 4390
},
{
"epoch": 2.6587429492344885,
"grad_norm": 0.2669179141521454,
"learning_rate": 3.907569051683302e-07,
"loss": 0.08082753419876099,
"step": 4400
},
{
"epoch": 2.6647864625302176,
"grad_norm": 0.2712559700012207,
"learning_rate": 3.7724601517246937e-07,
"loss": 0.07941591143608093,
"step": 4410
},
{
"epoch": 2.6708299758259466,
"grad_norm": 0.2751019597053528,
"learning_rate": 3.6396367053678274e-07,
"loss": 0.08138488531112671,
"step": 4420
},
{
"epoch": 2.676873489121676,
"grad_norm": 0.2488052248954773,
"learning_rate": 3.509105279300529e-07,
"loss": 0.0773822546005249,
"step": 4430
},
{
"epoch": 2.682917002417405,
"grad_norm": 0.25115492939949036,
"learning_rate": 3.380872326894946e-07,
"loss": 0.08337308168411255,
"step": 4440
},
{
"epoch": 2.6889605157131347,
"grad_norm": 0.3224640190601349,
"learning_rate": 3.254944187888481e-07,
"loss": 0.08367630243301391,
"step": 4450
},
{
"epoch": 2.6950040290088637,
"grad_norm": 0.2521418333053589,
"learning_rate": 3.131327088070335e-07,
"loss": 0.08279047012329102,
"step": 4460
},
{
"epoch": 2.701047542304593,
"grad_norm": 0.27216672897338867,
"learning_rate": 3.0100271389737365e-07,
"loss": 0.07915748357772827,
"step": 4470
},
{
"epoch": 2.7070910556003223,
"grad_norm": 0.2813398838043213,
"learning_rate": 2.891050337573803e-07,
"loss": 0.08114453554153442,
"step": 4480
},
{
"epoch": 2.7131345688960513,
"grad_norm": 0.292023241519928,
"learning_rate": 2.774402565991019e-07,
"loss": 0.07939202785491943,
"step": 4490
},
{
"epoch": 2.719178082191781,
"grad_norm": 0.2795642018318176,
"learning_rate": 2.6600895912004433e-07,
"loss": 0.07853795289993286,
"step": 4500
},
{
"epoch": 2.7252215954875103,
"grad_norm": 0.25801974534988403,
"learning_rate": 2.548117064746608e-07,
"loss": 0.08058047294616699,
"step": 4510
},
{
"epoch": 2.7312651087832394,
"grad_norm": 0.2468692809343338,
"learning_rate": 2.438490522464104e-07,
"loss": 0.08399761915206909,
"step": 4520
},
{
"epoch": 2.7373086220789684,
"grad_norm": 0.2499784380197525,
"learning_rate": 2.3312153842038775e-07,
"loss": 0.08040977716445923,
"step": 4530
},
{
"epoch": 2.743352135374698,
"grad_norm": 0.2497340440750122,
"learning_rate": 2.2262969535653044e-07,
"loss": 0.08070325255393981,
"step": 4540
},
{
"epoch": 2.749395648670427,
"grad_norm": 0.2518327832221985,
"learning_rate": 2.123740417633946e-07,
"loss": 0.0766767144203186,
"step": 4550
},
{
"epoch": 2.7554391619661565,
"grad_norm": 0.28237244486808777,
"learning_rate": 2.0235508467251642e-07,
"loss": 0.0834607720375061,
"step": 4560
},
{
"epoch": 2.7614826752618855,
"grad_norm": 0.24840769171714783,
"learning_rate": 1.9257331941333723e-07,
"loss": 0.08156927824020385,
"step": 4570
},
{
"epoch": 2.767526188557615,
"grad_norm": 0.26519617438316345,
"learning_rate": 1.830292295887215e-07,
"loss": 0.08160626292228698,
"step": 4580
},
{
"epoch": 2.773569701853344,
"grad_norm": 0.26683536171913147,
"learning_rate": 1.7372328705104356e-07,
"loss": 0.083110111951828,
"step": 4590
},
{
"epoch": 2.779613215149073,
"grad_norm": 0.2819823920726776,
"learning_rate": 1.6465595187886209e-07,
"loss": 0.07948071956634521,
"step": 4600
},
{
"epoch": 2.7856567284448026,
"grad_norm": 0.25477421283721924,
"learning_rate": 1.558276723541735e-07,
"loss": 0.08223173618316651,
"step": 4610
},
{
"epoch": 2.7917002417405317,
"grad_norm": 0.25511234998703003,
"learning_rate": 1.472388849402473e-07,
"loss": 0.07758338451385498,
"step": 4620
},
{
"epoch": 2.797743755036261,
"grad_norm": 0.2947482466697693,
"learning_rate": 1.3889001426005134e-07,
"loss": 0.07938053011894226,
"step": 4630
},
{
"epoch": 2.80378726833199,
"grad_norm": 0.27020299434661865,
"learning_rate": 1.3078147307525491e-07,
"loss": 0.08611318469047546,
"step": 4640
},
{
"epoch": 2.8098307816277197,
"grad_norm": 0.26552340388298035,
"learning_rate": 1.2291366226582724e-07,
"loss": 0.08001590371131898,
"step": 4650
},
{
"epoch": 2.8158742949234488,
"grad_norm": 0.2595081031322479,
"learning_rate": 1.1528697081021112e-07,
"loss": 0.07811046242713929,
"step": 4660
},
{
"epoch": 2.821917808219178,
"grad_norm": 0.2572011947631836,
"learning_rate": 1.0790177576609939e-07,
"loss": 0.0812339425086975,
"step": 4670
},
{
"epoch": 2.8279613215149073,
"grad_norm": 0.2580914795398712,
"learning_rate": 1.007584422517871e-07,
"loss": 0.08156332969665528,
"step": 4680
},
{
"epoch": 2.834004834810637,
"grad_norm": 0.26249319314956665,
"learning_rate": 9.385732342812703e-08,
"loss": 0.07958078384399414,
"step": 4690
},
{
"epoch": 2.840048348106366,
"grad_norm": 0.26565834879875183,
"learning_rate": 8.719876048106313e-08,
"loss": 0.07611703276634216,
"step": 4700
},
{
"epoch": 2.846091861402095,
"grad_norm": 0.2693769037723541,
"learning_rate": 8.078308260476675e-08,
"loss": 0.0787214994430542,
"step": 4710
},
{
"epoch": 2.8521353746978244,
"grad_norm": 0.28061386942863464,
"learning_rate": 7.461060698536026e-08,
"loss": 0.07973547577857971,
"step": 4720
},
{
"epoch": 2.8581788879935535,
"grad_norm": 0.2562766969203949,
"learning_rate": 6.868163878523626e-08,
"loss": 0.07998113632202149,
"step": 4730
},
{
"epoch": 2.864222401289283,
"grad_norm": 0.2675401270389557,
"learning_rate": 6.299647112796903e-08,
"loss": 0.08009319305419922,
"step": 4740
},
{
"epoch": 2.870265914585012,
"grad_norm": 0.3061717450618744,
"learning_rate": 5.7555385083823964e-08,
"loss": 0.08148778080940247,
"step": 4750
},
{
"epoch": 2.8763094278807415,
"grad_norm": 0.2746511697769165,
"learning_rate": 5.235864965586146e-08,
"loss": 0.07808900475502015,
"step": 4760
},
{
"epoch": 2.8823529411764706,
"grad_norm": 0.24901708960533142,
"learning_rate": 4.7406521766637515e-08,
"loss": 0.08119735717773438,
"step": 4770
},
{
"epoch": 2.8883964544721996,
"grad_norm": 0.27437493205070496,
"learning_rate": 4.269924624550226e-08,
"loss": 0.0810629665851593,
"step": 4780
},
{
"epoch": 2.894439967767929,
"grad_norm": 0.2583921253681183,
"learning_rate": 3.8237055816495174e-08,
"loss": 0.08207312226295471,
"step": 4790
},
{
"epoch": 2.9004834810636586,
"grad_norm": 0.24652065336704254,
"learning_rate": 3.4020171086838196e-08,
"loss": 0.07873359322547913,
"step": 4800
},
{
"epoch": 2.9065269943593877,
"grad_norm": 0.25214189291000366,
"learning_rate": 3.0048800536031654e-08,
"loss": 0.07919525504112243,
"step": 4810
},
{
"epoch": 2.9125705076551167,
"grad_norm": 0.2710227072238922,
"learning_rate": 2.6323140505545853e-08,
"loss": 0.082087242603302,
"step": 4820
},
{
"epoch": 2.918614020950846,
"grad_norm": 0.25783440470695496,
"learning_rate": 2.2843375189113835e-08,
"loss": 0.07796441316604615,
"step": 4830
},
{
"epoch": 2.9246575342465753,
"grad_norm": 0.284587025642395,
"learning_rate": 1.960967662362534e-08,
"loss": 0.08019020557403564,
"step": 4840
},
{
"epoch": 2.9307010475423043,
"grad_norm": 0.27005454897880554,
"learning_rate": 1.662220468062081e-08,
"loss": 0.08582592010498047,
"step": 4850
},
{
"epoch": 2.936744560838034,
"grad_norm": 0.27235037088394165,
"learning_rate": 1.3881107058389409e-08,
"loss": 0.0842123031616211,
"step": 4860
},
{
"epoch": 2.9427880741337633,
"grad_norm": 0.2919752299785614,
"learning_rate": 1.1386519274664831e-08,
"loss": 0.08081229925155639,
"step": 4870
},
{
"epoch": 2.9488315874294924,
"grad_norm": 0.256102055311203,
"learning_rate": 9.138564659925131e-09,
"loss": 0.07858948707580567,
"step": 4880
},
{
"epoch": 2.9548751007252214,
"grad_norm": 0.2619668245315552,
"learning_rate": 7.137354351298698e-09,
"loss": 0.08291544318199158,
"step": 4890
},
{
"epoch": 2.960918614020951,
"grad_norm": 0.2574392259120941,
"learning_rate": 5.382987287065322e-09,
"loss": 0.08099491596221924,
"step": 4900
},
{
"epoch": 2.96696212731668,
"grad_norm": 0.2536278963088989,
"learning_rate": 3.875550201767331e-09,
"loss": 0.07988271713256836,
"step": 4910
},
{
"epoch": 2.9730056406124095,
"grad_norm": 0.2615348696708679,
"learning_rate": 2.6151176219224584e-09,
"loss": 0.08206771612167359,
"step": 4920
},
{
"epoch": 2.9790491539081385,
"grad_norm": 0.25114062428474426,
"learning_rate": 1.6017518623356876e-09,
"loss": 0.08417447209358216,
"step": 4930
},
{
"epoch": 2.985092667203868,
"grad_norm": 0.2586587071418762,
"learning_rate": 8.35503023022266e-10,
"loss": 0.07650007009506225,
"step": 4940
},
{
"epoch": 2.991136180499597,
"grad_norm": 0.2562219798564911,
"learning_rate": 3.1640898672802287e-10,
"loss": 0.08685986995697022,
"step": 4950
},
{
"epoch": 2.997179693795326,
"grad_norm": 0.24136510491371155,
"learning_rate": 4.4495417059198246e-11,
"loss": 0.07802990674972535,
"step": 4960
},
{
"epoch": 3.0,
"eval_loss": 0.10261032730340958,
"eval_runtime": 74.3074,
"eval_samples_per_second": 10.806,
"eval_steps_per_second": 0.686,
"step": 4965
},
{
"epoch": 3.0,
"step": 4965,
"total_flos": 6.832000780639352e+19,
"train_loss": 0.12692538877388262,
"train_runtime": 145498.6824,
"train_samples_per_second": 1.637,
"train_steps_per_second": 0.034
}
],
"logging_steps": 10,
"max_steps": 4965,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6.832000780639352e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}