Files
1.7b-MixtureVitae-300BT-v1-…/trainer_state.json
ModelHub XC 4488999492 初始化项目,由ModelHub XC社区提供模型
Model: ali-elganzory/1.7b-MixtureVitae-300BT-v1-16k-WithChatTemplate
Source: Original Platform
2026-09-10 07:40:17 +08:00

2928 lines
71 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2062,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0024248302618816685,
"grad_norm": 0.38918766379356384,
"learning_rate": 7.692307692307694e-06,
"loss": 0.9711,
"step": 5
},
{
"epoch": 0.004849660523763337,
"grad_norm": 0.39544427394866943,
"learning_rate": 1.730769230769231e-05,
"loss": 0.943,
"step": 10
},
{
"epoch": 0.0072744907856450046,
"grad_norm": 0.2192670702934265,
"learning_rate": 2.6923076923076923e-05,
"loss": 0.9373,
"step": 15
},
{
"epoch": 0.009699321047526674,
"grad_norm": 0.17753367125988007,
"learning_rate": 3.653846153846154e-05,
"loss": 0.8946,
"step": 20
},
{
"epoch": 0.01212415130940834,
"grad_norm": 0.14115607738494873,
"learning_rate": 4.615384615384616e-05,
"loss": 0.8974,
"step": 25
},
{
"epoch": 0.014548981571290009,
"grad_norm": 0.11341860145330429,
"learning_rate": 5.576923076923077e-05,
"loss": 0.8735,
"step": 30
},
{
"epoch": 0.01697381183317168,
"grad_norm": 0.10785568505525589,
"learning_rate": 6.538461538461539e-05,
"loss": 0.8503,
"step": 35
},
{
"epoch": 0.019398642095053348,
"grad_norm": 0.09892906248569489,
"learning_rate": 7.500000000000001e-05,
"loss": 0.8251,
"step": 40
},
{
"epoch": 0.021823472356935016,
"grad_norm": 0.10826981067657471,
"learning_rate": 8.461538461538461e-05,
"loss": 0.8149,
"step": 45
},
{
"epoch": 0.02424830261881668,
"grad_norm": 0.12986481189727783,
"learning_rate": 9.423076923076924e-05,
"loss": 0.8151,
"step": 50
},
{
"epoch": 0.02667313288069835,
"grad_norm": 0.11989903450012207,
"learning_rate": 0.00010384615384615386,
"loss": 0.8155,
"step": 55
},
{
"epoch": 0.029097963142580018,
"grad_norm": 0.12929768860340118,
"learning_rate": 0.00011346153846153846,
"loss": 0.7889,
"step": 60
},
{
"epoch": 0.03152279340446169,
"grad_norm": 0.09861160814762115,
"learning_rate": 0.0001230769230769231,
"loss": 0.7821,
"step": 65
},
{
"epoch": 0.03394762366634336,
"grad_norm": 0.09228499233722687,
"learning_rate": 0.0001326923076923077,
"loss": 0.8075,
"step": 70
},
{
"epoch": 0.036372453928225024,
"grad_norm": 0.09446661174297333,
"learning_rate": 0.0001423076923076923,
"loss": 0.7725,
"step": 75
},
{
"epoch": 0.038797284190106696,
"grad_norm": 0.07715810090303421,
"learning_rate": 0.00015192307692307692,
"loss": 0.7687,
"step": 80
},
{
"epoch": 0.04122211445198836,
"grad_norm": 0.09901192784309387,
"learning_rate": 0.00016153846153846155,
"loss": 0.7555,
"step": 85
},
{
"epoch": 0.04364694471387003,
"grad_norm": 0.07883667945861816,
"learning_rate": 0.00017115384615384616,
"loss": 0.7628,
"step": 90
},
{
"epoch": 0.0460717749757517,
"grad_norm": 0.08181186020374298,
"learning_rate": 0.00018076923076923077,
"loss": 0.7684,
"step": 95
},
{
"epoch": 0.04849660523763336,
"grad_norm": 0.08731997013092041,
"learning_rate": 0.00019038461538461538,
"loss": 0.763,
"step": 100
},
{
"epoch": 0.050921435499515035,
"grad_norm": 0.08396711945533752,
"learning_rate": 0.0002,
"loss": 0.7533,
"step": 105
},
{
"epoch": 0.0533462657613967,
"grad_norm": 0.07599729299545288,
"learning_rate": 0.00019999678202953174,
"loss": 0.7458,
"step": 110
},
{
"epoch": 0.05577109602327837,
"grad_norm": 0.07871965318918228,
"learning_rate": 0.00019998712832523365,
"loss": 0.7679,
"step": 115
},
{
"epoch": 0.058195926285160036,
"grad_norm": 0.08363021910190582,
"learning_rate": 0.0001999710395084124,
"loss": 0.7635,
"step": 120
},
{
"epoch": 0.06062075654704171,
"grad_norm": 0.08509054780006409,
"learning_rate": 0.00019994851661453477,
"loss": 0.7217,
"step": 125
},
{
"epoch": 0.06304558680892337,
"grad_norm": 0.08128708600997925,
"learning_rate": 0.0001999195610931609,
"loss": 0.7518,
"step": 130
},
{
"epoch": 0.06547041707080505,
"grad_norm": 0.08791787177324295,
"learning_rate": 0.00019988417480785105,
"loss": 0.7568,
"step": 135
},
{
"epoch": 0.06789524733268672,
"grad_norm": 0.0770721286535263,
"learning_rate": 0.0001998423600360456,
"loss": 0.7411,
"step": 140
},
{
"epoch": 0.07032007759456838,
"grad_norm": 0.07810170948505402,
"learning_rate": 0.0001997941194689186,
"loss": 0.7461,
"step": 145
},
{
"epoch": 0.07274490785645005,
"grad_norm": 0.08361192792654037,
"learning_rate": 0.00019973945621120452,
"loss": 0.718,
"step": 150
},
{
"epoch": 0.07516973811833172,
"grad_norm": 0.07803226262331009,
"learning_rate": 0.00019967837378099824,
"loss": 0.7473,
"step": 155
},
{
"epoch": 0.07759456838021339,
"grad_norm": 0.08025413751602173,
"learning_rate": 0.0001996108761095289,
"loss": 0.7301,
"step": 160
},
{
"epoch": 0.08001939864209505,
"grad_norm": 0.0751657709479332,
"learning_rate": 0.00019953696754090684,
"loss": 0.7356,
"step": 165
},
{
"epoch": 0.08244422890397672,
"grad_norm": 0.07403068989515305,
"learning_rate": 0.00019945665283184382,
"loss": 0.7414,
"step": 170
},
{
"epoch": 0.0848690591658584,
"grad_norm": 0.08277441561222076,
"learning_rate": 0.0001993699371513471,
"loss": 0.7297,
"step": 175
},
{
"epoch": 0.08729388942774007,
"grad_norm": 0.07375401258468628,
"learning_rate": 0.00019927682608038668,
"loss": 0.732,
"step": 180
},
{
"epoch": 0.08971871968962172,
"grad_norm": 0.07576849311590195,
"learning_rate": 0.00019917732561153606,
"loss": 0.7218,
"step": 185
},
{
"epoch": 0.0921435499515034,
"grad_norm": 0.08011229336261749,
"learning_rate": 0.00019907144214858667,
"loss": 0.7299,
"step": 190
},
{
"epoch": 0.09456838021338507,
"grad_norm": 0.07243379205465317,
"learning_rate": 0.0001989591825061356,
"loss": 0.7361,
"step": 195
},
{
"epoch": 0.09699321047526673,
"grad_norm": 0.08236246556043625,
"learning_rate": 0.00019884055390914721,
"loss": 0.7335,
"step": 200
},
{
"epoch": 0.0994180407371484,
"grad_norm": 0.07557398080825806,
"learning_rate": 0.00019871556399248788,
"loss": 0.7436,
"step": 205
},
{
"epoch": 0.10184287099903007,
"grad_norm": 0.07167453318834305,
"learning_rate": 0.00019858422080043484,
"loss": 0.727,
"step": 210
},
{
"epoch": 0.10426770126091174,
"grad_norm": 0.07908965647220612,
"learning_rate": 0.00019844653278615833,
"loss": 0.7417,
"step": 215
},
{
"epoch": 0.1066925315227934,
"grad_norm": 0.07614396512508392,
"learning_rate": 0.00019830250881117768,
"loss": 0.7244,
"step": 220
},
{
"epoch": 0.10911736178467507,
"grad_norm": 0.0825204849243164,
"learning_rate": 0.0001981521581447908,
"loss": 0.7298,
"step": 225
},
{
"epoch": 0.11154219204655674,
"grad_norm": 0.07835296541452408,
"learning_rate": 0.00019799549046347782,
"loss": 0.7205,
"step": 230
},
{
"epoch": 0.11396702230843841,
"grad_norm": 0.07611602544784546,
"learning_rate": 0.00019783251585027816,
"loss": 0.7171,
"step": 235
},
{
"epoch": 0.11639185257032007,
"grad_norm": 0.07712840288877487,
"learning_rate": 0.00019766324479414166,
"loss": 0.7034,
"step": 240
},
{
"epoch": 0.11881668283220174,
"grad_norm": 0.07362441718578339,
"learning_rate": 0.00019748768818925348,
"loss": 0.7192,
"step": 245
},
{
"epoch": 0.12124151309408342,
"grad_norm": 0.08477923274040222,
"learning_rate": 0.00019730585733433308,
"loss": 0.7205,
"step": 250
},
{
"epoch": 0.12366634335596509,
"grad_norm": 0.07782363146543503,
"learning_rate": 0.00019711776393190687,
"loss": 0.7236,
"step": 255
},
{
"epoch": 0.12609117361784675,
"grad_norm": 0.07701355218887329,
"learning_rate": 0.00019692342008755513,
"loss": 0.7128,
"step": 260
},
{
"epoch": 0.12851600387972842,
"grad_norm": 0.0748991072177887,
"learning_rate": 0.0001967228383091329,
"loss": 0.729,
"step": 265
},
{
"epoch": 0.1309408341416101,
"grad_norm": 0.06640012562274933,
"learning_rate": 0.00019651603150596495,
"loss": 0.708,
"step": 270
},
{
"epoch": 0.13336566440349176,
"grad_norm": 0.0845034196972847,
"learning_rate": 0.00019630301298801504,
"loss": 0.7313,
"step": 275
},
{
"epoch": 0.13579049466537343,
"grad_norm": 0.07433103024959564,
"learning_rate": 0.0001960837964650291,
"loss": 0.7088,
"step": 280
},
{
"epoch": 0.13821532492725508,
"grad_norm": 0.08018206804990768,
"learning_rate": 0.0001958583960456531,
"loss": 0.7071,
"step": 285
},
{
"epoch": 0.14064015518913675,
"grad_norm": 0.07544240355491638,
"learning_rate": 0.00019562682623652493,
"loss": 0.699,
"step": 290
},
{
"epoch": 0.14306498545101842,
"grad_norm": 0.07852726429700851,
"learning_rate": 0.00019538910194134065,
"loss": 0.7252,
"step": 295
},
{
"epoch": 0.1454898157129001,
"grad_norm": 0.08389027416706085,
"learning_rate": 0.0001951452384598956,
"loss": 0.7151,
"step": 300
},
{
"epoch": 0.14791464597478177,
"grad_norm": 0.07259048521518707,
"learning_rate": 0.00019489525148709932,
"loss": 0.718,
"step": 305
},
{
"epoch": 0.15033947623666344,
"grad_norm": 0.07333388179540634,
"learning_rate": 0.0001946391571119658,
"loss": 0.7066,
"step": 310
},
{
"epoch": 0.1527643064985451,
"grad_norm": 0.07507184147834778,
"learning_rate": 0.00019437697181657771,
"loss": 0.7026,
"step": 315
},
{
"epoch": 0.15518913676042678,
"grad_norm": 0.07852832973003387,
"learning_rate": 0.00019410871247502585,
"loss": 0.7062,
"step": 320
},
{
"epoch": 0.15761396702230843,
"grad_norm": 0.07935725152492523,
"learning_rate": 0.00019383439635232294,
"loss": 0.7102,
"step": 325
},
{
"epoch": 0.1600387972841901,
"grad_norm": 0.07529488205909729,
"learning_rate": 0.0001935540411032927,
"loss": 0.7164,
"step": 330
},
{
"epoch": 0.16246362754607177,
"grad_norm": 0.0734439566731453,
"learning_rate": 0.00019326766477143327,
"loss": 0.7038,
"step": 335
},
{
"epoch": 0.16488845780795344,
"grad_norm": 0.07135467231273651,
"learning_rate": 0.0001929752857877563,
"loss": 0.6957,
"step": 340
},
{
"epoch": 0.16731328806983511,
"grad_norm": 0.07417963445186615,
"learning_rate": 0.00019267692296960048,
"loss": 0.7294,
"step": 345
},
{
"epoch": 0.1697381183317168,
"grad_norm": 0.07934129983186722,
"learning_rate": 0.00019237259551942054,
"loss": 0.7138,
"step": 350
},
{
"epoch": 0.17216294859359846,
"grad_norm": 0.07620280236005783,
"learning_rate": 0.00019206232302355143,
"loss": 0.7063,
"step": 355
},
{
"epoch": 0.17458777885548013,
"grad_norm": 0.072400763630867,
"learning_rate": 0.00019174612545094774,
"loss": 0.6784,
"step": 360
},
{
"epoch": 0.17701260911736177,
"grad_norm": 0.07668977975845337,
"learning_rate": 0.0001914240231518985,
"loss": 0.7023,
"step": 365
},
{
"epoch": 0.17943743937924345,
"grad_norm": 0.07105539739131927,
"learning_rate": 0.00019109603685671736,
"loss": 0.6923,
"step": 370
},
{
"epoch": 0.18186226964112512,
"grad_norm": 0.07577413320541382,
"learning_rate": 0.00019076218767440865,
"loss": 0.7077,
"step": 375
},
{
"epoch": 0.1842870999030068,
"grad_norm": 0.07451111823320389,
"learning_rate": 0.0001904224970913085,
"loss": 0.7055,
"step": 380
},
{
"epoch": 0.18671193016488846,
"grad_norm": 0.07368946820497513,
"learning_rate": 0.00019007698696970224,
"loss": 0.6928,
"step": 385
},
{
"epoch": 0.18913676042677013,
"grad_norm": 0.07631614059209824,
"learning_rate": 0.0001897256795464172,
"loss": 0.6912,
"step": 390
},
{
"epoch": 0.1915615906886518,
"grad_norm": 0.07660483568906784,
"learning_rate": 0.00018936859743139166,
"loss": 0.7022,
"step": 395
},
{
"epoch": 0.19398642095053345,
"grad_norm": 0.08595852553844452,
"learning_rate": 0.00018900576360621963,
"loss": 0.693,
"step": 400
},
{
"epoch": 0.19641125121241512,
"grad_norm": 0.07508262246847153,
"learning_rate": 0.00018863720142267183,
"loss": 0.7176,
"step": 405
},
{
"epoch": 0.1988360814742968,
"grad_norm": 0.07464610785245895,
"learning_rate": 0.00018826293460119266,
"loss": 0.7107,
"step": 410
},
{
"epoch": 0.20126091173617847,
"grad_norm": 0.0771806538105011,
"learning_rate": 0.0001878829872293737,
"loss": 0.6831,
"step": 415
},
{
"epoch": 0.20368574199806014,
"grad_norm": 0.07285494357347488,
"learning_rate": 0.00018749738376040346,
"loss": 0.7027,
"step": 420
},
{
"epoch": 0.2061105722599418,
"grad_norm": 0.08320580422878265,
"learning_rate": 0.00018710614901149334,
"loss": 0.7183,
"step": 425
},
{
"epoch": 0.20853540252182348,
"grad_norm": 0.07620342820882797,
"learning_rate": 0.00018670930816228079,
"loss": 0.723,
"step": 430
},
{
"epoch": 0.21096023278370515,
"grad_norm": 0.0763145163655281,
"learning_rate": 0.00018630688675320842,
"loss": 0.689,
"step": 435
},
{
"epoch": 0.2133850630455868,
"grad_norm": 0.07359050959348679,
"learning_rate": 0.00018589891068388048,
"loss": 0.6977,
"step": 440
},
{
"epoch": 0.21580989330746847,
"grad_norm": 0.07169776409864426,
"learning_rate": 0.0001854854062113958,
"loss": 0.7003,
"step": 445
},
{
"epoch": 0.21823472356935014,
"grad_norm": 0.08285032957792282,
"learning_rate": 0.00018506639994865805,
"loss": 0.709,
"step": 450
},
{
"epoch": 0.22065955383123181,
"grad_norm": 0.07976315170526505,
"learning_rate": 0.00018464191886266273,
"loss": 0.7031,
"step": 455
},
{
"epoch": 0.22308438409311349,
"grad_norm": 0.09100283682346344,
"learning_rate": 0.0001842119902727619,
"loss": 0.6808,
"step": 460
},
{
"epoch": 0.22550921435499516,
"grad_norm": 0.07634368538856506,
"learning_rate": 0.00018377664184890562,
"loss": 0.7097,
"step": 465
},
{
"epoch": 0.22793404461687683,
"grad_norm": 0.07575997710227966,
"learning_rate": 0.00018333590160986136,
"loss": 0.707,
"step": 470
},
{
"epoch": 0.23035887487875847,
"grad_norm": 0.08256373554468155,
"learning_rate": 0.00018288979792141058,
"loss": 0.7061,
"step": 475
},
{
"epoch": 0.23278370514064015,
"grad_norm": 0.07805568724870682,
"learning_rate": 0.00018243835949452314,
"loss": 0.6964,
"step": 480
},
{
"epoch": 0.23520853540252182,
"grad_norm": 0.07713499665260315,
"learning_rate": 0.00018198161538350962,
"loss": 0.6883,
"step": 485
},
{
"epoch": 0.2376333656644035,
"grad_norm": 0.07370981574058533,
"learning_rate": 0.00018151959498415122,
"loss": 0.7031,
"step": 490
},
{
"epoch": 0.24005819592628516,
"grad_norm": 0.07156600058078766,
"learning_rate": 0.00018105232803180795,
"loss": 0.7013,
"step": 495
},
{
"epoch": 0.24248302618816683,
"grad_norm": 0.07860364019870758,
"learning_rate": 0.00018057984459950483,
"loss": 0.697,
"step": 500
},
{
"epoch": 0.2449078564500485,
"grad_norm": 0.07430411130189896,
"learning_rate": 0.00018010217509599659,
"loss": 0.7019,
"step": 505
},
{
"epoch": 0.24733268671193018,
"grad_norm": 0.07481130212545395,
"learning_rate": 0.00017961935026381028,
"loss": 0.677,
"step": 510
},
{
"epoch": 0.24975751697381182,
"grad_norm": 0.07289189845323563,
"learning_rate": 0.00017913140117726693,
"loss": 0.6906,
"step": 515
},
{
"epoch": 0.2521823472356935,
"grad_norm": 0.07331232726573944,
"learning_rate": 0.00017863835924048163,
"loss": 0.6741,
"step": 520
},
{
"epoch": 0.25460717749757517,
"grad_norm": 0.07814762741327286,
"learning_rate": 0.00017814025618534212,
"loss": 0.6685,
"step": 525
},
{
"epoch": 0.25703200775945684,
"grad_norm": 0.07258285582065582,
"learning_rate": 0.00017763712406946685,
"loss": 0.7023,
"step": 530
},
{
"epoch": 0.2594568380213385,
"grad_norm": 0.07580193877220154,
"learning_rate": 0.0001771289952741417,
"loss": 0.6749,
"step": 535
},
{
"epoch": 0.2618816682832202,
"grad_norm": 0.07483787834644318,
"learning_rate": 0.00017661590250223576,
"loss": 0.6884,
"step": 540
},
{
"epoch": 0.26430649854510185,
"grad_norm": 0.07170502096414566,
"learning_rate": 0.0001760978787760968,
"loss": 0.6884,
"step": 545
},
{
"epoch": 0.2667313288069835,
"grad_norm": 0.0817745104432106,
"learning_rate": 0.00017557495743542585,
"loss": 0.6912,
"step": 550
},
{
"epoch": 0.2691561590688652,
"grad_norm": 0.07086025178432465,
"learning_rate": 0.00017504717213513157,
"loss": 0.7088,
"step": 555
},
{
"epoch": 0.27158098933074687,
"grad_norm": 0.0802539512515068,
"learning_rate": 0.00017451455684316413,
"loss": 0.6924,
"step": 560
},
{
"epoch": 0.27400581959262854,
"grad_norm": 0.0739421471953392,
"learning_rate": 0.00017397714583832912,
"loss": 0.7049,
"step": 565
},
{
"epoch": 0.27643064985451016,
"grad_norm": 0.0740748941898346,
"learning_rate": 0.00017343497370808146,
"loss": 0.6945,
"step": 570
},
{
"epoch": 0.27885548011639183,
"grad_norm": 0.07814696431159973,
"learning_rate": 0.00017288807534629922,
"loss": 0.6677,
"step": 575
},
{
"epoch": 0.2812803103782735,
"grad_norm": 0.07814390957355499,
"learning_rate": 0.00017233648595103788,
"loss": 0.6944,
"step": 580
},
{
"epoch": 0.2837051406401552,
"grad_norm": 0.07190729677677155,
"learning_rate": 0.0001717802410222652,
"loss": 0.6967,
"step": 585
},
{
"epoch": 0.28612997090203685,
"grad_norm": 0.06977193057537079,
"learning_rate": 0.00017121937635957622,
"loss": 0.6786,
"step": 590
},
{
"epoch": 0.2885548011639185,
"grad_norm": 0.07131995260715485,
"learning_rate": 0.00017065392805988943,
"loss": 0.6949,
"step": 595
},
{
"epoch": 0.2909796314258002,
"grad_norm": 0.07661376148462296,
"learning_rate": 0.00017008393251512332,
"loss": 0.674,
"step": 600
},
{
"epoch": 0.29340446168768186,
"grad_norm": 0.07369566708803177,
"learning_rate": 0.00016950942640985454,
"loss": 0.6743,
"step": 605
},
{
"epoch": 0.29582929194956353,
"grad_norm": 0.07492439448833466,
"learning_rate": 0.00016893044671895672,
"loss": 0.6879,
"step": 610
},
{
"epoch": 0.2982541222114452,
"grad_norm": 0.07045532017946243,
"learning_rate": 0.0001683470307052208,
"loss": 0.671,
"step": 615
},
{
"epoch": 0.3006789524733269,
"grad_norm": 0.07425723969936371,
"learning_rate": 0.0001677592159169568,
"loss": 0.6931,
"step": 620
},
{
"epoch": 0.30310378273520855,
"grad_norm": 0.07944834232330322,
"learning_rate": 0.00016716704018557738,
"loss": 0.6786,
"step": 625
},
{
"epoch": 0.3055286129970902,
"grad_norm": 0.06985245645046234,
"learning_rate": 0.00016657054162316274,
"loss": 0.674,
"step": 630
},
{
"epoch": 0.3079534432589719,
"grad_norm": 0.07267067581415176,
"learning_rate": 0.00016596975862000814,
"loss": 0.6678,
"step": 635
},
{
"epoch": 0.31037827352085356,
"grad_norm": 0.07110784947872162,
"learning_rate": 0.00016536472984215284,
"loss": 0.6914,
"step": 640
},
{
"epoch": 0.3128031037827352,
"grad_norm": 0.07703442126512527,
"learning_rate": 0.0001647554942288915,
"loss": 0.6849,
"step": 645
},
{
"epoch": 0.31522793404461685,
"grad_norm": 0.07916276901960373,
"learning_rate": 0.0001641420909902685,
"loss": 0.6834,
"step": 650
},
{
"epoch": 0.3176527643064985,
"grad_norm": 0.07032134383916855,
"learning_rate": 0.00016352455960455387,
"loss": 0.6966,
"step": 655
},
{
"epoch": 0.3200775945683802,
"grad_norm": 0.07388387620449066,
"learning_rate": 0.00016290293981570291,
"loss": 0.681,
"step": 660
},
{
"epoch": 0.32250242483026187,
"grad_norm": 0.07308702170848846,
"learning_rate": 0.0001622772716307981,
"loss": 0.6698,
"step": 665
},
{
"epoch": 0.32492725509214354,
"grad_norm": 0.06939036399126053,
"learning_rate": 0.00016164759531747425,
"loss": 0.6901,
"step": 670
},
{
"epoch": 0.3273520853540252,
"grad_norm": 0.07716705650091171,
"learning_rate": 0.00016101395140132697,
"loss": 0.6825,
"step": 675
},
{
"epoch": 0.3297769156159069,
"grad_norm": 0.07418633997440338,
"learning_rate": 0.00016037638066330448,
"loss": 0.6744,
"step": 680
},
{
"epoch": 0.33220174587778856,
"grad_norm": 0.07232547551393509,
"learning_rate": 0.00015973492413708283,
"loss": 0.6523,
"step": 685
},
{
"epoch": 0.33462657613967023,
"grad_norm": 0.0700221061706543,
"learning_rate": 0.0001590896231064253,
"loss": 0.6805,
"step": 690
},
{
"epoch": 0.3370514064015519,
"grad_norm": 0.06720143556594849,
"learning_rate": 0.00015844051910252502,
"loss": 0.7044,
"step": 695
},
{
"epoch": 0.3394762366634336,
"grad_norm": 0.07611308246850967,
"learning_rate": 0.00015778765390133228,
"loss": 0.6926,
"step": 700
},
{
"epoch": 0.34190106692531524,
"grad_norm": 0.0685822069644928,
"learning_rate": 0.00015713106952086587,
"loss": 0.6848,
"step": 705
},
{
"epoch": 0.3443258971871969,
"grad_norm": 0.07210035622119904,
"learning_rate": 0.0001564708082185087,
"loss": 0.6607,
"step": 710
},
{
"epoch": 0.3467507274490786,
"grad_norm": 0.07212255895137787,
"learning_rate": 0.00015580691248828817,
"loss": 0.6865,
"step": 715
},
{
"epoch": 0.34917555771096026,
"grad_norm": 0.07362619042396545,
"learning_rate": 0.00015513942505814146,
"loss": 0.6743,
"step": 720
},
{
"epoch": 0.3516003879728419,
"grad_norm": 0.08068300783634186,
"learning_rate": 0.00015446838888716523,
"loss": 0.6712,
"step": 725
},
{
"epoch": 0.35402521823472355,
"grad_norm": 0.06782415509223938,
"learning_rate": 0.00015379384716285112,
"loss": 0.6743,
"step": 730
},
{
"epoch": 0.3564500484966052,
"grad_norm": 0.07503397017717361,
"learning_rate": 0.00015311584329830619,
"loss": 0.6889,
"step": 735
},
{
"epoch": 0.3588748787584869,
"grad_norm": 0.06833958625793457,
"learning_rate": 0.0001524344209294586,
"loss": 0.7017,
"step": 740
},
{
"epoch": 0.36129970902036856,
"grad_norm": 0.07117005437612534,
"learning_rate": 0.0001517496239122496,
"loss": 0.6818,
"step": 745
},
{
"epoch": 0.36372453928225024,
"grad_norm": 0.0768495723605156,
"learning_rate": 0.0001510614963198107,
"loss": 0.6634,
"step": 750
},
{
"epoch": 0.3661493695441319,
"grad_norm": 0.0725046917796135,
"learning_rate": 0.0001503700824396274,
"loss": 0.6835,
"step": 755
},
{
"epoch": 0.3685741998060136,
"grad_norm": 0.07461989670991898,
"learning_rate": 0.00014967542677068858,
"loss": 0.6829,
"step": 760
},
{
"epoch": 0.37099903006789525,
"grad_norm": 0.07148183882236481,
"learning_rate": 0.00014897757402062284,
"loss": 0.6674,
"step": 765
},
{
"epoch": 0.3734238603297769,
"grad_norm": 0.07259765267372131,
"learning_rate": 0.000148276569102821,
"loss": 0.6793,
"step": 770
},
{
"epoch": 0.3758486905916586,
"grad_norm": 0.07724656164646149,
"learning_rate": 0.0001475724571335455,
"loss": 0.6783,
"step": 775
},
{
"epoch": 0.37827352085354027,
"grad_norm": 0.07636789977550507,
"learning_rate": 0.00014686528342902687,
"loss": 0.6737,
"step": 780
},
{
"epoch": 0.38069835111542194,
"grad_norm": 0.07437674701213837,
"learning_rate": 0.00014615509350254699,
"loss": 0.6775,
"step": 785
},
{
"epoch": 0.3831231813773036,
"grad_norm": 0.07093089818954468,
"learning_rate": 0.0001454419330615101,
"loss": 0.6665,
"step": 790
},
{
"epoch": 0.3855480116391853,
"grad_norm": 0.07145288586616516,
"learning_rate": 0.00014472584800450096,
"loss": 0.6634,
"step": 795
},
{
"epoch": 0.3879728419010669,
"grad_norm": 0.07825754582881927,
"learning_rate": 0.00014400688441833088,
"loss": 0.6619,
"step": 800
},
{
"epoch": 0.3903976721629486,
"grad_norm": 0.06800788640975952,
"learning_rate": 0.00014328508857507162,
"loss": 0.6718,
"step": 805
},
{
"epoch": 0.39282250242483024,
"grad_norm": 0.07289911061525345,
"learning_rate": 0.00014256050692907737,
"loss": 0.66,
"step": 810
},
{
"epoch": 0.3952473326867119,
"grad_norm": 0.07747261226177216,
"learning_rate": 0.00014183318611399486,
"loss": 0.6834,
"step": 815
},
{
"epoch": 0.3976721629485936,
"grad_norm": 0.07100475579500198,
"learning_rate": 0.00014110317293976218,
"loss": 0.6675,
"step": 820
},
{
"epoch": 0.40009699321047526,
"grad_norm": 0.07786409556865692,
"learning_rate": 0.0001403705143895961,
"loss": 0.6658,
"step": 825
},
{
"epoch": 0.40252182347235693,
"grad_norm": 0.07454486191272736,
"learning_rate": 0.0001396352576169681,
"loss": 0.6892,
"step": 830
},
{
"epoch": 0.4049466537342386,
"grad_norm": 0.0711129754781723,
"learning_rate": 0.00013889744994256982,
"loss": 0.6652,
"step": 835
},
{
"epoch": 0.4073714839961203,
"grad_norm": 0.0716179683804512,
"learning_rate": 0.00013815713885126748,
"loss": 0.6688,
"step": 840
},
{
"epoch": 0.40979631425800195,
"grad_norm": 0.06912382692098618,
"learning_rate": 0.00013741437198904556,
"loss": 0.6808,
"step": 845
},
{
"epoch": 0.4122211445198836,
"grad_norm": 0.07759840041399002,
"learning_rate": 0.00013666919715994068,
"loss": 0.6613,
"step": 850
},
{
"epoch": 0.4146459747817653,
"grad_norm": 0.07312402129173279,
"learning_rate": 0.00013592166232296465,
"loss": 0.6819,
"step": 855
},
{
"epoch": 0.41707080504364696,
"grad_norm": 0.07156159728765488,
"learning_rate": 0.00013517181558901818,
"loss": 0.6706,
"step": 860
},
{
"epoch": 0.41949563530552864,
"grad_norm": 0.06726700067520142,
"learning_rate": 0.000134419705217794,
"loss": 0.6538,
"step": 865
},
{
"epoch": 0.4219204655674103,
"grad_norm": 0.07086332887411118,
"learning_rate": 0.00013366537961467153,
"loss": 0.647,
"step": 870
},
{
"epoch": 0.4243452958292919,
"grad_norm": 0.07842452824115753,
"learning_rate": 0.000132908887327601,
"loss": 0.677,
"step": 875
},
{
"epoch": 0.4267701260911736,
"grad_norm": 0.07341575622558594,
"learning_rate": 0.0001321502770439792,
"loss": 0.6815,
"step": 880
},
{
"epoch": 0.42919495635305527,
"grad_norm": 0.07531925290822983,
"learning_rate": 0.0001313895975875159,
"loss": 0.6804,
"step": 885
},
{
"epoch": 0.43161978661493694,
"grad_norm": 0.06923644989728928,
"learning_rate": 0.0001306268979150917,
"loss": 0.6739,
"step": 890
},
{
"epoch": 0.4340446168768186,
"grad_norm": 0.06904316693544388,
"learning_rate": 0.00012986222711360693,
"loss": 0.6853,
"step": 895
},
{
"epoch": 0.4364694471387003,
"grad_norm": 0.08328109979629517,
"learning_rate": 0.00012909563439682286,
"loss": 0.6674,
"step": 900
},
{
"epoch": 0.43889427740058196,
"grad_norm": 0.06594524532556534,
"learning_rate": 0.00012832716910219388,
"loss": 0.6808,
"step": 905
},
{
"epoch": 0.44131910766246363,
"grad_norm": 0.07734475284814835,
"learning_rate": 0.0001275568806876925,
"loss": 0.672,
"step": 910
},
{
"epoch": 0.4437439379243453,
"grad_norm": 0.07080471515655518,
"learning_rate": 0.00012678481872862608,
"loss": 0.671,
"step": 915
},
{
"epoch": 0.44616876818622697,
"grad_norm": 0.06919893622398376,
"learning_rate": 0.00012601103291444635,
"loss": 0.6628,
"step": 920
},
{
"epoch": 0.44859359844810864,
"grad_norm": 0.075429767370224,
"learning_rate": 0.0001252355730455513,
"loss": 0.6646,
"step": 925
},
{
"epoch": 0.4510184287099903,
"grad_norm": 0.06842102855443954,
"learning_rate": 0.00012445848903008003,
"loss": 0.6701,
"step": 930
},
{
"epoch": 0.453443258971872,
"grad_norm": 0.07292219996452332,
"learning_rate": 0.00012367983088070077,
"loss": 0.6587,
"step": 935
},
{
"epoch": 0.45586808923375366,
"grad_norm": 0.07115452736616135,
"learning_rate": 0.00012289964871139218,
"loss": 0.6661,
"step": 940
},
{
"epoch": 0.45829291949563533,
"grad_norm": 0.06807117909193039,
"learning_rate": 0.0001221179927342179,
"loss": 0.6747,
"step": 945
},
{
"epoch": 0.46071774975751695,
"grad_norm": 0.07635519653558731,
"learning_rate": 0.00012133491325609485,
"loss": 0.635,
"step": 950
},
{
"epoch": 0.4631425800193986,
"grad_norm": 0.07146444171667099,
"learning_rate": 0.00012055046067555582,
"loss": 0.6772,
"step": 955
},
{
"epoch": 0.4655674102812803,
"grad_norm": 0.07073502987623215,
"learning_rate": 0.00011976468547950547,
"loss": 0.6643,
"step": 960
},
{
"epoch": 0.46799224054316196,
"grad_norm": 0.0681312307715416,
"learning_rate": 0.00011897763823997144,
"loss": 0.6716,
"step": 965
},
{
"epoch": 0.47041707080504364,
"grad_norm": 0.0741071030497551,
"learning_rate": 0.0001181893696108491,
"loss": 0.6792,
"step": 970
},
{
"epoch": 0.4728419010669253,
"grad_norm": 0.07416609674692154,
"learning_rate": 0.0001173999303246419,
"loss": 0.6417,
"step": 975
},
{
"epoch": 0.475266731328807,
"grad_norm": 0.07191631942987442,
"learning_rate": 0.00011660937118919597,
"loss": 0.6592,
"step": 980
},
{
"epoch": 0.47769156159068865,
"grad_norm": 0.07099225372076035,
"learning_rate": 0.0001158177430844304,
"loss": 0.6619,
"step": 985
},
{
"epoch": 0.4801163918525703,
"grad_norm": 0.07223440706729889,
"learning_rate": 0.00011502509695906243,
"loss": 0.6654,
"step": 990
},
{
"epoch": 0.482541222114452,
"grad_norm": 0.06803770363330841,
"learning_rate": 0.00011423148382732853,
"loss": 0.6819,
"step": 995
},
{
"epoch": 0.48496605237633367,
"grad_norm": 0.07934951782226562,
"learning_rate": 0.00011343695476570109,
"loss": 0.6797,
"step": 1000
},
{
"epoch": 0.48739088263821534,
"grad_norm": 0.0722769945859909,
"learning_rate": 0.0001126415609096013,
"loss": 0.6688,
"step": 1005
},
{
"epoch": 0.489815712900097,
"grad_norm": 0.06851601600646973,
"learning_rate": 0.00011184535345010787,
"loss": 0.6548,
"step": 1010
},
{
"epoch": 0.4922405431619787,
"grad_norm": 0.07180795073509216,
"learning_rate": 0.00011104838363066268,
"loss": 0.6674,
"step": 1015
},
{
"epoch": 0.49466537342386036,
"grad_norm": 0.07068037241697311,
"learning_rate": 0.00011025070274377258,
"loss": 0.6521,
"step": 1020
},
{
"epoch": 0.49709020368574197,
"grad_norm": 0.07635895162820816,
"learning_rate": 0.00010945236212770831,
"loss": 0.666,
"step": 1025
},
{
"epoch": 0.49951503394762364,
"grad_norm": 0.06773378700017929,
"learning_rate": 0.0001086534131632004,
"loss": 0.6614,
"step": 1030
},
{
"epoch": 0.5019398642095053,
"grad_norm": 0.06768732517957687,
"learning_rate": 0.00010785390727013231,
"loss": 0.6788,
"step": 1035
},
{
"epoch": 0.504364694471387,
"grad_norm": 0.06725457310676575,
"learning_rate": 0.0001070538959042311,
"loss": 0.6593,
"step": 1040
},
{
"epoch": 0.5067895247332687,
"grad_norm": 0.06976445764303207,
"learning_rate": 0.00010625343055375579,
"loss": 0.6812,
"step": 1045
},
{
"epoch": 0.5092143549951503,
"grad_norm": 0.07576105743646622,
"learning_rate": 0.00010545256273618352,
"loss": 0.6773,
"step": 1050
},
{
"epoch": 0.511639185257032,
"grad_norm": 0.07427694648504257,
"learning_rate": 0.000104651343994894,
"loss": 0.6701,
"step": 1055
},
{
"epoch": 0.5140640155189137,
"grad_norm": 0.06958897411823273,
"learning_rate": 0.00010384982589585227,
"loss": 0.6478,
"step": 1060
},
{
"epoch": 0.5164888457807953,
"grad_norm": 0.06551092863082886,
"learning_rate": 0.00010304806002428966,
"loss": 0.6516,
"step": 1065
},
{
"epoch": 0.518913676042677,
"grad_norm": 0.07074626535177231,
"learning_rate": 0.00010224609798138422,
"loss": 0.6721,
"step": 1070
},
{
"epoch": 0.5213385063045587,
"grad_norm": 0.07591406255960464,
"learning_rate": 0.00010144399138093932,
"loss": 0.6722,
"step": 1075
},
{
"epoch": 0.5237633365664404,
"grad_norm": 0.06842809170484543,
"learning_rate": 0.00010064179184606202,
"loss": 0.6684,
"step": 1080
},
{
"epoch": 0.526188166828322,
"grad_norm": 0.07387987524271011,
"learning_rate": 9.983955100584055e-05,
"loss": 0.6707,
"step": 1085
},
{
"epoch": 0.5286129970902037,
"grad_norm": 0.07045400142669678,
"learning_rate": 9.903732049202157e-05,
"loss": 0.6661,
"step": 1090
},
{
"epoch": 0.5310378273520854,
"grad_norm": 0.07052426785230637,
"learning_rate": 9.823515193568715e-05,
"loss": 0.6528,
"step": 1095
},
{
"epoch": 0.533462657613967,
"grad_norm": 0.0797022134065628,
"learning_rate": 9.743309696393176e-05,
"loss": 0.6731,
"step": 1100
},
{
"epoch": 0.5358874878758487,
"grad_norm": 0.06646578758955002,
"learning_rate": 9.66312071965397e-05,
"loss": 0.671,
"step": 1105
},
{
"epoch": 0.5383123181377304,
"grad_norm": 0.07197515666484833,
"learning_rate": 9.582953424266272e-05,
"loss": 0.6615,
"step": 1110
},
{
"epoch": 0.5407371483996121,
"grad_norm": 0.07122502475976944,
"learning_rate": 9.502812969749863e-05,
"loss": 0.666,
"step": 1115
},
{
"epoch": 0.5431619786614937,
"grad_norm": 0.06962478905916214,
"learning_rate": 9.422704513897068e-05,
"loss": 0.6546,
"step": 1120
},
{
"epoch": 0.5455868089233754,
"grad_norm": 0.0786578431725502,
"learning_rate": 9.342633212440787e-05,
"loss": 0.6602,
"step": 1125
},
{
"epoch": 0.5480116391852571,
"grad_norm": 0.06958270817995071,
"learning_rate": 9.262604218722691e-05,
"loss": 0.667,
"step": 1130
},
{
"epoch": 0.5504364694471388,
"grad_norm": 0.07455138862133026,
"learning_rate": 9.182622683361546e-05,
"loss": 0.644,
"step": 1135
},
{
"epoch": 0.5528612997090203,
"grad_norm": 0.07016881555318832,
"learning_rate": 9.102693753921727e-05,
"loss": 0.6629,
"step": 1140
},
{
"epoch": 0.555286129970902,
"grad_norm": 0.07230067998170853,
"learning_rate": 9.022822574581929e-05,
"loss": 0.6669,
"step": 1145
},
{
"epoch": 0.5577109602327837,
"grad_norm": 0.07794859260320663,
"learning_rate": 8.943014285804072e-05,
"loss": 0.6719,
"step": 1150
},
{
"epoch": 0.5601357904946653,
"grad_norm": 0.0690775066614151,
"learning_rate": 8.863274024002488e-05,
"loss": 0.6645,
"step": 1155
},
{
"epoch": 0.562560620756547,
"grad_norm": 0.07458430528640747,
"learning_rate": 8.78360692121333e-05,
"loss": 0.673,
"step": 1160
},
{
"epoch": 0.5649854510184287,
"grad_norm": 0.07065006345510483,
"learning_rate": 8.70401810476428e-05,
"loss": 0.6679,
"step": 1165
},
{
"epoch": 0.5674102812803103,
"grad_norm": 0.06775312125682831,
"learning_rate": 8.624512696944555e-05,
"loss": 0.6713,
"step": 1170
},
{
"epoch": 0.569835111542192,
"grad_norm": 0.08035821467638016,
"learning_rate": 8.545095814675245e-05,
"loss": 0.6582,
"step": 1175
},
{
"epoch": 0.5722599418040737,
"grad_norm": 0.07017914205789566,
"learning_rate": 8.465772569179981e-05,
"loss": 0.6577,
"step": 1180
},
{
"epoch": 0.5746847720659554,
"grad_norm": 0.06815838068723679,
"learning_rate": 8.386548065656e-05,
"loss": 0.6475,
"step": 1185
},
{
"epoch": 0.577109602327837,
"grad_norm": 0.07126031816005707,
"learning_rate": 8.307427402945556e-05,
"loss": 0.6627,
"step": 1190
},
{
"epoch": 0.5795344325897187,
"grad_norm": 0.07584662735462189,
"learning_rate": 8.228415673207761e-05,
"loss": 0.6717,
"step": 1195
},
{
"epoch": 0.5819592628516004,
"grad_norm": 0.08144812285900116,
"learning_rate": 8.149517961590881e-05,
"loss": 0.6562,
"step": 1200
},
{
"epoch": 0.584384093113482,
"grad_norm": 0.06777734309434891,
"learning_rate": 8.070739345905032e-05,
"loss": 0.6638,
"step": 1205
},
{
"epoch": 0.5868089233753637,
"grad_norm": 0.07002221792936325,
"learning_rate": 7.992084896295395e-05,
"loss": 0.6562,
"step": 1210
},
{
"epoch": 0.5892337536372454,
"grad_norm": 0.07009301334619522,
"learning_rate": 7.913559674915887e-05,
"loss": 0.6683,
"step": 1215
},
{
"epoch": 0.5916585838991271,
"grad_norm": 0.06898258626461029,
"learning_rate": 7.835168735603374e-05,
"loss": 0.6691,
"step": 1220
},
{
"epoch": 0.5940834141610087,
"grad_norm": 0.07859721779823303,
"learning_rate": 7.756917123552416e-05,
"loss": 0.6481,
"step": 1225
},
{
"epoch": 0.5965082444228904,
"grad_norm": 0.07015025615692139,
"learning_rate": 7.678809874990544e-05,
"loss": 0.6832,
"step": 1230
},
{
"epoch": 0.5989330746847721,
"grad_norm": 0.07461626827716827,
"learning_rate": 7.600852016854141e-05,
"loss": 0.652,
"step": 1235
},
{
"epoch": 0.6013579049466538,
"grad_norm": 0.06960247457027435,
"learning_rate": 7.523048566464915e-05,
"loss": 0.6572,
"step": 1240
},
{
"epoch": 0.6037827352085354,
"grad_norm": 0.07379946112632751,
"learning_rate": 7.445404531206975e-05,
"loss": 0.6806,
"step": 1245
},
{
"epoch": 0.6062075654704171,
"grad_norm": 0.0801565870642662,
"learning_rate": 7.367924908204576e-05,
"loss": 0.6592,
"step": 1250
},
{
"epoch": 0.6086323957322988,
"grad_norm": 0.07287748157978058,
"learning_rate": 7.290614684000489e-05,
"loss": 0.6668,
"step": 1255
},
{
"epoch": 0.6110572259941804,
"grad_norm": 0.0685887485742569,
"learning_rate": 7.213478834235079e-05,
"loss": 0.6606,
"step": 1260
},
{
"epoch": 0.6134820562560621,
"grad_norm": 0.07128997147083282,
"learning_rate": 7.136522323326083e-05,
"loss": 0.6804,
"step": 1265
},
{
"epoch": 0.6159068865179438,
"grad_norm": 0.07110767811536789,
"learning_rate": 7.05975010414909e-05,
"loss": 0.6488,
"step": 1270
},
{
"epoch": 0.6183317167798255,
"grad_norm": 0.07725566625595093,
"learning_rate": 6.98316711771878e-05,
"loss": 0.6667,
"step": 1275
},
{
"epoch": 0.6207565470417071,
"grad_norm": 0.07086648046970367,
"learning_rate": 6.906778292870929e-05,
"loss": 0.6565,
"step": 1280
},
{
"epoch": 0.6231813773035888,
"grad_norm": 0.07215660810470581,
"learning_rate": 6.830588545945179e-05,
"loss": 0.6443,
"step": 1285
},
{
"epoch": 0.6256062075654704,
"grad_norm": 0.07043175399303436,
"learning_rate": 6.754602780468651e-05,
"loss": 0.6587,
"step": 1290
},
{
"epoch": 0.628031037827352,
"grad_norm": 0.0720083937048912,
"learning_rate": 6.678825886840329e-05,
"loss": 0.6675,
"step": 1295
},
{
"epoch": 0.6304558680892337,
"grad_norm": 0.07498017698526382,
"learning_rate": 6.603262742016328e-05,
"loss": 0.6477,
"step": 1300
},
{
"epoch": 0.6328806983511154,
"grad_norm": 0.06804075092077255,
"learning_rate": 6.527918209196017e-05,
"loss": 0.6513,
"step": 1305
},
{
"epoch": 0.635305528612997,
"grad_norm": 0.07545261085033417,
"learning_rate": 6.452797137509032e-05,
"loss": 0.6397,
"step": 1310
},
{
"epoch": 0.6377303588748787,
"grad_norm": 0.06824527680873871,
"learning_rate": 6.377904361703178e-05,
"loss": 0.6593,
"step": 1315
},
{
"epoch": 0.6401551891367604,
"grad_norm": 0.07044185698032379,
"learning_rate": 6.303244701833266e-05,
"loss": 0.6543,
"step": 1320
},
{
"epoch": 0.6425800193986421,
"grad_norm": 0.07908555120229721,
"learning_rate": 6.22882296295091e-05,
"loss": 0.6604,
"step": 1325
},
{
"epoch": 0.6450048496605237,
"grad_norm": 0.06805066764354706,
"learning_rate": 6.154643934795272e-05,
"loss": 0.6499,
"step": 1330
},
{
"epoch": 0.6474296799224054,
"grad_norm": 0.06950155645608902,
"learning_rate": 6.080712391484792e-05,
"loss": 0.657,
"step": 1335
},
{
"epoch": 0.6498545101842871,
"grad_norm": 0.06826996058225632,
"learning_rate": 6.007033091209926e-05,
"loss": 0.6584,
"step": 1340
},
{
"epoch": 0.6522793404461688,
"grad_norm": 0.06806496530771255,
"learning_rate": 5.9336107759269257e-05,
"loss": 0.6685,
"step": 1345
},
{
"epoch": 0.6547041707080504,
"grad_norm": 0.07646046578884125,
"learning_rate": 5.8604501710526324e-05,
"loss": 0.6537,
"step": 1350
},
{
"epoch": 0.6571290009699321,
"grad_norm": 0.0673266276717186,
"learning_rate": 5.787555985160371e-05,
"loss": 0.6636,
"step": 1355
},
{
"epoch": 0.6595538312318138,
"grad_norm": 0.06736177206039429,
"learning_rate": 5.7149329096768924e-05,
"loss": 0.6411,
"step": 1360
},
{
"epoch": 0.6619786614936954,
"grad_norm": 0.0655079334974289,
"learning_rate": 5.642585618580434e-05,
"loss": 0.6609,
"step": 1365
},
{
"epoch": 0.6644034917555771,
"grad_norm": 0.06992436945438385,
"learning_rate": 5.570518768099918e-05,
"loss": 0.6638,
"step": 1370
},
{
"epoch": 0.6668283220174588,
"grad_norm": 0.07555795460939407,
"learning_rate": 5.498736996415287e-05,
"loss": 0.6602,
"step": 1375
},
{
"epoch": 0.6692531522793405,
"grad_norm": 0.07073895633220673,
"learning_rate": 5.427244923358967e-05,
"loss": 0.6598,
"step": 1380
},
{
"epoch": 0.6716779825412221,
"grad_norm": 0.07193751633167267,
"learning_rate": 5.356047150118548e-05,
"loss": 0.6537,
"step": 1385
},
{
"epoch": 0.6741028128031038,
"grad_norm": 0.06939923763275146,
"learning_rate": 5.2851482589406675e-05,
"loss": 0.6479,
"step": 1390
},
{
"epoch": 0.6765276430649855,
"grad_norm": 0.06879837065935135,
"learning_rate": 5.2145528128360864e-05,
"loss": 0.6642,
"step": 1395
},
{
"epoch": 0.6789524733268671,
"grad_norm": 0.07876288890838623,
"learning_rate": 5.1442653552860186e-05,
"loss": 0.6491,
"step": 1400
},
{
"epoch": 0.6813773035887488,
"grad_norm": 0.06859347224235535,
"learning_rate": 5.07429040994972e-05,
"loss": 0.6613,
"step": 1405
},
{
"epoch": 0.6838021338506305,
"grad_norm": 0.06807195395231247,
"learning_rate": 5.004632480373341e-05,
"loss": 0.6455,
"step": 1410
},
{
"epoch": 0.6862269641125122,
"grad_norm": 0.070477195084095,
"learning_rate": 4.935296049700087e-05,
"loss": 0.6624,
"step": 1415
},
{
"epoch": 0.6886517943743938,
"grad_norm": 0.06642670184373856,
"learning_rate": 4.866285580381684e-05,
"loss": 0.645,
"step": 1420
},
{
"epoch": 0.6910766246362755,
"grad_norm": 0.07304893434047699,
"learning_rate": 4.797605513891179e-05,
"loss": 0.6697,
"step": 1425
},
{
"epoch": 0.6935014548981572,
"grad_norm": 0.0770198404788971,
"learning_rate": 4.7292602704370834e-05,
"loss": 0.663,
"step": 1430
},
{
"epoch": 0.6959262851600388,
"grad_norm": 0.07229270786046982,
"learning_rate": 4.6612542486789004e-05,
"loss": 0.6481,
"step": 1435
},
{
"epoch": 0.6983511154219205,
"grad_norm": 0.0701506957411766,
"learning_rate": 4.593591825444028e-05,
"loss": 0.6587,
"step": 1440
},
{
"epoch": 0.7007759456838021,
"grad_norm": 0.07267860323190689,
"learning_rate": 4.526277355446049e-05,
"loss": 0.6413,
"step": 1445
},
{
"epoch": 0.7032007759456838,
"grad_norm": 0.0736895501613617,
"learning_rate": 4.4593151710045064e-05,
"loss": 0.6451,
"step": 1450
},
{
"epoch": 0.7056256062075654,
"grad_norm": 0.06733877211809158,
"learning_rate": 4.392709581766038e-05,
"loss": 0.6468,
"step": 1455
},
{
"epoch": 0.7080504364694471,
"grad_norm": 0.07400280982255936,
"learning_rate": 4.326464874427025e-05,
"loss": 0.6492,
"step": 1460
},
{
"epoch": 0.7104752667313288,
"grad_norm": 0.07255461066961288,
"learning_rate": 4.260585312457711e-05,
"loss": 0.6584,
"step": 1465
},
{
"epoch": 0.7129000969932104,
"grad_norm": 0.06916648894548416,
"learning_rate": 4.1950751358277865e-05,
"loss": 0.6614,
"step": 1470
},
{
"epoch": 0.7153249272550921,
"grad_norm": 0.07707880437374115,
"learning_rate": 4.129938560733525e-05,
"loss": 0.6471,
"step": 1475
},
{
"epoch": 0.7177497575169738,
"grad_norm": 0.06957708299160004,
"learning_rate": 4.0651797793264354e-05,
"loss": 0.6496,
"step": 1480
},
{
"epoch": 0.7201745877788555,
"grad_norm": 0.07020954042673111,
"learning_rate": 4.0008029594434414e-05,
"loss": 0.6318,
"step": 1485
},
{
"epoch": 0.7225994180407371,
"grad_norm": 0.06942954659461975,
"learning_rate": 3.93681224433864e-05,
"loss": 0.6631,
"step": 1490
},
{
"epoch": 0.7250242483026188,
"grad_norm": 0.07160889357328415,
"learning_rate": 3.873211752416662e-05,
"loss": 0.6547,
"step": 1495
},
{
"epoch": 0.7274490785645005,
"grad_norm": 0.07599648088216782,
"learning_rate": 3.810005576967606e-05,
"loss": 0.6473,
"step": 1500
},
{
"epoch": 0.7298739088263821,
"grad_norm": 0.07107492536306381,
"learning_rate": 3.74719778590359e-05,
"loss": 0.6559,
"step": 1505
},
{
"epoch": 0.7322987390882638,
"grad_norm": 0.07364842295646667,
"learning_rate": 3.68479242149695e-05,
"loss": 0.6357,
"step": 1510
},
{
"epoch": 0.7347235693501455,
"grad_norm": 0.07083046436309814,
"learning_rate": 3.6227935001200817e-05,
"loss": 0.6518,
"step": 1515
},
{
"epoch": 0.7371483996120272,
"grad_norm": 0.0683782696723938,
"learning_rate": 3.561205011986947e-05,
"loss": 0.6727,
"step": 1520
},
{
"epoch": 0.7395732298739088,
"grad_norm": 0.07474040240049362,
"learning_rate": 3.5000309208962624e-05,
"loss": 0.6344,
"step": 1525
},
{
"epoch": 0.7419980601357905,
"grad_norm": 0.06675226241350174,
"learning_rate": 3.439275163976402e-05,
"loss": 0.6552,
"step": 1530
},
{
"epoch": 0.7444228903976722,
"grad_norm": 0.07602989673614502,
"learning_rate": 3.378941651431996e-05,
"loss": 0.6732,
"step": 1535
},
{
"epoch": 0.7468477206595538,
"grad_norm": 0.06826721876859665,
"learning_rate": 3.319034266292277e-05,
"loss": 0.6722,
"step": 1540
},
{
"epoch": 0.7492725509214355,
"grad_norm": 0.0693931132555008,
"learning_rate": 3.259556864161172e-05,
"loss": 0.663,
"step": 1545
},
{
"epoch": 0.7516973811833172,
"grad_norm": 0.07731709629297256,
"learning_rate": 3.2005132729691424e-05,
"loss": 0.6496,
"step": 1550
},
{
"epoch": 0.7541222114451989,
"grad_norm": 0.0687938928604126,
"learning_rate": 3.1419072927268545e-05,
"loss": 0.6563,
"step": 1555
},
{
"epoch": 0.7565470417070805,
"grad_norm": 0.07114581018686295,
"learning_rate": 3.083742695280578e-05,
"loss": 0.6375,
"step": 1560
},
{
"epoch": 0.7589718719689622,
"grad_norm": 0.06666100770235062,
"learning_rate": 3.026023224069452e-05,
"loss": 0.6444,
"step": 1565
},
{
"epoch": 0.7613967022308439,
"grad_norm": 0.06928212195634842,
"learning_rate": 2.968752593884555e-05,
"loss": 0.6677,
"step": 1570
},
{
"epoch": 0.7638215324927256,
"grad_norm": 0.07700739055871964,
"learning_rate": 2.911934490629813e-05,
"loss": 0.641,
"step": 1575
},
{
"epoch": 0.7662463627546072,
"grad_norm": 0.07009076327085495,
"learning_rate": 2.8555725710847925e-05,
"loss": 0.6586,
"step": 1580
},
{
"epoch": 0.7686711930164889,
"grad_norm": 0.0692385733127594,
"learning_rate": 2.799670462669354e-05,
"loss": 0.6459,
"step": 1585
},
{
"epoch": 0.7710960232783706,
"grad_norm": 0.07023077458143234,
"learning_rate": 2.7442317632101745e-05,
"loss": 0.6478,
"step": 1590
},
{
"epoch": 0.7735208535402521,
"grad_norm": 0.0697379857301712,
"learning_rate": 2.6892600407092038e-05,
"loss": 0.6629,
"step": 1595
},
{
"epoch": 0.7759456838021338,
"grad_norm": 0.07526423782110214,
"learning_rate": 2.634758833114036e-05,
"loss": 0.6589,
"step": 1600
},
{
"epoch": 0.7783705140640155,
"grad_norm": 0.06746955215930939,
"learning_rate": 2.5807316480901976e-05,
"loss": 0.6594,
"step": 1605
},
{
"epoch": 0.7807953443258971,
"grad_norm": 0.07108338922262192,
"learning_rate": 2.5271819627954163e-05,
"loss": 0.648,
"step": 1610
},
{
"epoch": 0.7832201745877788,
"grad_norm": 0.07071094959974289,
"learning_rate": 2.4741132236558006e-05,
"loss": 0.6494,
"step": 1615
},
{
"epoch": 0.7856450048496605,
"grad_norm": 0.07233584672212601,
"learning_rate": 2.42152884614406e-05,
"loss": 0.6582,
"step": 1620
},
{
"epoch": 0.7880698351115422,
"grad_norm": 0.07301636785268784,
"learning_rate": 2.3694322145596735e-05,
"loss": 0.6466,
"step": 1625
},
{
"epoch": 0.7904946653734238,
"grad_norm": 0.06874296814203262,
"learning_rate": 2.3178266818110782e-05,
"loss": 0.6554,
"step": 1630
},
{
"epoch": 0.7929194956353055,
"grad_norm": 0.06953322142362595,
"learning_rate": 2.2667155691998832e-05,
"loss": 0.6356,
"step": 1635
},
{
"epoch": 0.7953443258971872,
"grad_norm": 0.06957842409610748,
"learning_rate": 2.216102166207108e-05,
"loss": 0.6496,
"step": 1640
},
{
"epoch": 0.7977691561590688,
"grad_norm": 0.06773428618907928,
"learning_rate": 2.1659897302814747e-05,
"loss": 0.645,
"step": 1645
},
{
"epoch": 0.8001939864209505,
"grad_norm": 0.07253550738096237,
"learning_rate": 2.1163814866297648e-05,
"loss": 0.6586,
"step": 1650
},
{
"epoch": 0.8026188166828322,
"grad_norm": 0.0699772909283638,
"learning_rate": 2.0672806280092304e-05,
"loss": 0.6567,
"step": 1655
},
{
"epoch": 0.8050436469447139,
"grad_norm": 0.06678451597690582,
"learning_rate": 2.0186903145221403e-05,
"loss": 0.6446,
"step": 1660
},
{
"epoch": 0.8074684772065955,
"grad_norm": 0.07169430702924728,
"learning_rate": 1.9706136734123682e-05,
"loss": 0.6657,
"step": 1665
},
{
"epoch": 0.8098933074684772,
"grad_norm": 0.06794555485248566,
"learning_rate": 1.9230537988641406e-05,
"loss": 0.6489,
"step": 1670
},
{
"epoch": 0.8123181377303589,
"grad_norm": 0.07556099444627762,
"learning_rate": 1.8760137518028953e-05,
"loss": 0.6486,
"step": 1675
},
{
"epoch": 0.8147429679922406,
"grad_norm": 0.06554684042930603,
"learning_rate": 1.8294965596982727e-05,
"loss": 0.6524,
"step": 1680
},
{
"epoch": 0.8171677982541222,
"grad_norm": 0.07229667156934738,
"learning_rate": 1.7835052163692824e-05,
"loss": 0.6424,
"step": 1685
},
{
"epoch": 0.8195926285160039,
"grad_norm": 0.06855263561010361,
"learning_rate": 1.7380426817916208e-05,
"loss": 0.6492,
"step": 1690
},
{
"epoch": 0.8220174587778856,
"grad_norm": 0.06754108518362045,
"learning_rate": 1.6931118819071635e-05,
"loss": 0.6589,
"step": 1695
},
{
"epoch": 0.8244422890397672,
"grad_norm": 0.07303808629512787,
"learning_rate": 1.648715708435645e-05,
"loss": 0.6378,
"step": 1700
},
{
"epoch": 0.8268671193016489,
"grad_norm": 0.07010772824287415,
"learning_rate": 1.6048570186885747e-05,
"loss": 0.6709,
"step": 1705
},
{
"epoch": 0.8292919495635306,
"grad_norm": 0.06951456516981125,
"learning_rate": 1.5615386353853145e-05,
"loss": 0.6557,
"step": 1710
},
{
"epoch": 0.8317167798254123,
"grad_norm": 0.06731481850147247,
"learning_rate": 1.518763346471438e-05,
"loss": 0.6628,
"step": 1715
},
{
"epoch": 0.8341416100872939,
"grad_norm": 0.07028862088918686,
"learning_rate": 1.4765339049392669e-05,
"loss": 0.6543,
"step": 1720
},
{
"epoch": 0.8365664403491756,
"grad_norm": 0.07783721387386322,
"learning_rate": 1.4348530286507179e-05,
"loss": 0.6633,
"step": 1725
},
{
"epoch": 0.8389912706110573,
"grad_norm": 0.06992127746343613,
"learning_rate": 1.3937234001623712e-05,
"loss": 0.6588,
"step": 1730
},
{
"epoch": 0.8414161008729389,
"grad_norm": 0.07088334858417511,
"learning_rate": 1.3531476665528242e-05,
"loss": 0.6437,
"step": 1735
},
{
"epoch": 0.8438409311348206,
"grad_norm": 0.06964202970266342,
"learning_rate": 1.3131284392523257e-05,
"loss": 0.6273,
"step": 1740
},
{
"epoch": 0.8462657613967022,
"grad_norm": 0.07019855082035065,
"learning_rate": 1.273668293874708e-05,
"loss": 0.645,
"step": 1745
},
{
"epoch": 0.8486905916585838,
"grad_norm": 0.0719023197889328,
"learning_rate": 1.2347697700516226e-05,
"loss": 0.6488,
"step": 1750
},
{
"epoch": 0.8511154219204655,
"grad_norm": 0.06849932670593262,
"learning_rate": 1.196435371269089e-05,
"loss": 0.6609,
"step": 1755
},
{
"epoch": 0.8535402521823472,
"grad_norm": 0.07256150245666504,
"learning_rate": 1.158667564706365e-05,
"loss": 0.6526,
"step": 1760
},
{
"epoch": 0.8559650824442289,
"grad_norm": 0.06930581480264664,
"learning_rate": 1.1214687810771796e-05,
"loss": 0.638,
"step": 1765
},
{
"epoch": 0.8583899127061105,
"grad_norm": 0.07124356180429459,
"learning_rate": 1.0848414144732744e-05,
"loss": 0.6503,
"step": 1770
},
{
"epoch": 0.8608147429679922,
"grad_norm": 0.07542183250188828,
"learning_rate": 1.048787822210332e-05,
"loss": 0.6567,
"step": 1775
},
{
"epoch": 0.8632395732298739,
"grad_norm": 0.06727565824985504,
"learning_rate": 1.0133103246762566e-05,
"loss": 0.6544,
"step": 1780
},
{
"epoch": 0.8656644034917556,
"grad_norm": 0.0727323591709137,
"learning_rate": 9.784112051818295e-06,
"loss": 0.649,
"step": 1785
},
{
"epoch": 0.8680892337536372,
"grad_norm": 0.0711803138256073,
"learning_rate": 9.440927098137752e-06,
"loss": 0.6606,
"step": 1790
},
{
"epoch": 0.8705140640155189,
"grad_norm": 0.06881871819496155,
"learning_rate": 9.103570472901835e-06,
"loss": 0.647,
"step": 1795
},
{
"epoch": 0.8729388942774006,
"grad_norm": 0.07360757142305374,
"learning_rate": 8.7720638881837e-06,
"loss": 0.6386,
"step": 1800
},
{
"epoch": 0.8753637245392822,
"grad_norm": 0.06956224888563156,
"learning_rate": 8.44642867955132e-06,
"loss": 0.6556,
"step": 1805
},
{
"epoch": 0.8777885548011639,
"grad_norm": 0.07224294543266296,
"learning_rate": 8.1266858046944e-06,
"loss": 0.6402,
"step": 1810
},
{
"epoch": 0.8802133850630456,
"grad_norm": 0.06778159737586975,
"learning_rate": 7.812855842075495e-06,
"loss": 0.6454,
"step": 1815
},
{
"epoch": 0.8826382153249273,
"grad_norm": 0.068122997879982,
"learning_rate": 7.5049589896056924e-06,
"loss": 0.6435,
"step": 1820
},
{
"epoch": 0.8850630455868089,
"grad_norm": 0.0704505518078804,
"learning_rate": 7.20301506334452e-06,
"loss": 0.6421,
"step": 1825
},
{
"epoch": 0.8874878758486906,
"grad_norm": 0.06601767987012863,
"learning_rate": 6.907043496224741e-06,
"loss": 0.6413,
"step": 1830
},
{
"epoch": 0.8899127061105723,
"grad_norm": 0.07295170426368713,
"learning_rate": 6.61706333680161e-06,
"loss": 0.6527,
"step": 1835
},
{
"epoch": 0.8923375363724539,
"grad_norm": 0.06985466927289963,
"learning_rate": 6.3330932480269116e-06,
"loss": 0.6397,
"step": 1840
},
{
"epoch": 0.8947623666343356,
"grad_norm": 0.07073473930358887,
"learning_rate": 6.0551515060478405e-06,
"loss": 0.665,
"step": 1845
},
{
"epoch": 0.8971871968962173,
"grad_norm": 0.0762571170926094,
"learning_rate": 5.783255999030746e-06,
"loss": 0.6496,
"step": 1850
},
{
"epoch": 0.899612027158099,
"grad_norm": 0.06765305250883102,
"learning_rate": 5.517424226009871e-06,
"loss": 0.64,
"step": 1855
},
{
"epoch": 0.9020368574199806,
"grad_norm": 0.07410787045955658,
"learning_rate": 5.2576732957611005e-06,
"loss": 0.6429,
"step": 1860
},
{
"epoch": 0.9044616876818623,
"grad_norm": 0.0670701190829277,
"learning_rate": 5.00401992570092e-06,
"loss": 0.6498,
"step": 1865
},
{
"epoch": 0.906886517943744,
"grad_norm": 0.06863141059875488,
"learning_rate": 4.756480440810385e-06,
"loss": 0.6457,
"step": 1870
},
{
"epoch": 0.9093113482056256,
"grad_norm": 0.08062388747930527,
"learning_rate": 4.51507077258454e-06,
"loss": 0.6332,
"step": 1875
},
{
"epoch": 0.9117361784675073,
"grad_norm": 0.06748832762241364,
"learning_rate": 4.279806458007052e-06,
"loss": 0.6716,
"step": 1880
},
{
"epoch": 0.914161008729389,
"grad_norm": 0.06559710204601288,
"learning_rate": 4.050702638550275e-06,
"loss": 0.6372,
"step": 1885
},
{
"epoch": 0.9165858389912707,
"grad_norm": 0.06698034703731537,
"learning_rate": 3.82777405920065e-06,
"loss": 0.6529,
"step": 1890
},
{
"epoch": 0.9190106692531523,
"grad_norm": 0.07044559717178345,
"learning_rate": 3.611035067509938e-06,
"loss": 0.6689,
"step": 1895
},
{
"epoch": 0.9214354995150339,
"grad_norm": 0.07434722036123276,
"learning_rate": 3.4004996126716037e-06,
"loss": 0.6371,
"step": 1900
},
{
"epoch": 0.9238603297769156,
"grad_norm": 0.06569571793079376,
"learning_rate": 3.1961812446231864e-06,
"loss": 0.6679,
"step": 1905
},
{
"epoch": 0.9262851600387972,
"grad_norm": 0.07368065416812897,
"learning_rate": 2.9980931131741496e-06,
"loss": 0.6538,
"step": 1910
},
{
"epoch": 0.9287099903006789,
"grad_norm": 0.06916055828332901,
"learning_rate": 2.8062479671596475e-06,
"loss": 0.6608,
"step": 1915
},
{
"epoch": 0.9311348205625606,
"grad_norm": 0.07183555513620377,
"learning_rate": 2.6206581536199594e-06,
"loss": 0.6458,
"step": 1920
},
{
"epoch": 0.9335596508244423,
"grad_norm": 0.07422731071710587,
"learning_rate": 2.441335617005902e-06,
"loss": 0.6437,
"step": 1925
},
{
"epoch": 0.9359844810863239,
"grad_norm": 0.07022187858819962,
"learning_rate": 2.26829189840998e-06,
"loss": 0.6512,
"step": 1930
},
{
"epoch": 0.9384093113482056,
"grad_norm": 0.0708911269903183,
"learning_rate": 2.101538134823733e-06,
"loss": 0.6388,
"step": 1935
},
{
"epoch": 0.9408341416100873,
"grad_norm": 0.07022671401500702,
"learning_rate": 1.9410850584208883e-06,
"loss": 0.6343,
"step": 1940
},
{
"epoch": 0.9432589718719689,
"grad_norm": 0.07077636569738388,
"learning_rate": 1.7869429958666805e-06,
"loss": 0.6465,
"step": 1945
},
{
"epoch": 0.9456838021338506,
"grad_norm": 0.07644498348236084,
"learning_rate": 1.639121867653215e-06,
"loss": 0.6513,
"step": 1950
},
{
"epoch": 0.9481086323957323,
"grad_norm": 0.06702401489019394,
"learning_rate": 1.4976311874610017e-06,
"loss": 0.6332,
"step": 1955
},
{
"epoch": 0.950533462657614,
"grad_norm": 0.07274569571018219,
"learning_rate": 1.3624800615466226e-06,
"loss": 0.6479,
"step": 1960
},
{
"epoch": 0.9529582929194956,
"grad_norm": 0.07144317030906677,
"learning_rate": 1.2336771881567455e-06,
"loss": 0.6491,
"step": 1965
},
{
"epoch": 0.9553831231813773,
"grad_norm": 0.06880391389131546,
"learning_rate": 1.111230856968204e-06,
"loss": 0.6346,
"step": 1970
},
{
"epoch": 0.957807953443259,
"grad_norm": 0.07087037712335587,
"learning_rate": 9.951489485545695e-07,
"loss": 0.6443,
"step": 1975
},
{
"epoch": 0.9602327837051406,
"grad_norm": 0.07291446626186371,
"learning_rate": 8.854389338788904e-07,
"loss": 0.6452,
"step": 1980
},
{
"epoch": 0.9626576139670223,
"grad_norm": 0.0726611539721489,
"learning_rate": 7.821078738129206e-07,
"loss": 0.638,
"step": 1985
},
{
"epoch": 0.965082444228904,
"grad_norm": 0.0732244923710823,
"learning_rate": 6.851624186826722e-07,
"loss": 0.6507,
"step": 1990
},
{
"epoch": 0.9675072744907857,
"grad_norm": 0.06873750686645508,
"learning_rate": 5.946088078403355e-07,
"loss": 0.6555,
"step": 1995
},
{
"epoch": 0.9699321047526673,
"grad_norm": 0.07295786589384079,
"learning_rate": 5.104528692628563e-07,
"loss": 0.638,
"step": 2000
},
{
"epoch": 0.972356935014549,
"grad_norm": 0.06500031799077988,
"learning_rate": 4.32700019176735e-07,
"loss": 0.6375,
"step": 2005
},
{
"epoch": 0.9747817652764307,
"grad_norm": 0.07015930116176605,
"learning_rate": 3.613552617094618e-07,
"loss": 0.6441,
"step": 2010
},
{
"epoch": 0.9772065955383124,
"grad_norm": 0.06755400449037552,
"learning_rate": 2.9642318856749617e-07,
"loss": 0.6413,
"step": 2015
},
{
"epoch": 0.979631425800194,
"grad_norm": 0.07017125934362411,
"learning_rate": 2.3790797874070349e-07,
"loss": 0.6476,
"step": 2020
},
{
"epoch": 0.9820562560620757,
"grad_norm": 0.07523668557405472,
"learning_rate": 1.8581339823345868e-07,
"loss": 0.634,
"step": 2025
},
{
"epoch": 0.9844810863239574,
"grad_norm": 0.06807917356491089,
"learning_rate": 1.401427998221627e-07,
"loss": 0.6699,
"step": 2030
},
{
"epoch": 0.986905916585839,
"grad_norm": 0.0702148899435997,
"learning_rate": 1.0089912283957059e-07,
"loss": 0.6542,
"step": 2035
},
{
"epoch": 0.9893307468477207,
"grad_norm": 0.06573879718780518,
"learning_rate": 6.808489298554266e-08,
"loss": 0.6359,
"step": 2040
},
{
"epoch": 0.9917555771096024,
"grad_norm": 0.0696154534816742,
"learning_rate": 4.170222216454134e-08,
"loss": 0.6616,
"step": 2045
},
{
"epoch": 0.9941804073714839,
"grad_norm": 0.0815952718257904,
"learning_rate": 2.175280834967319e-08,
"loss": 0.6422,
"step": 2050
},
{
"epoch": 0.9966052376333656,
"grad_norm": 0.06667590886354446,
"learning_rate": 8.237935473431879e-09,
"loss": 0.6444,
"step": 2055
},
{
"epoch": 0.9990300678952473,
"grad_norm": 0.07248586416244507,
"learning_rate": 1.1584733450531459e-09,
"loss": 0.641,
"step": 2060
},
{
"epoch": 1.0,
"step": 2062,
"total_flos": 5340644728897536.0,
"train_loss": 0.6803441236137072,
"train_runtime": 63408.1642,
"train_samples_per_second": 16.648,
"train_steps_per_second": 0.033
}
],
"logging_steps": 5,
"max_steps": 2062,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5340644728897536.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}