Jeronymous commited on
Commit
1e8a35a
·
1 Parent(s): 969b0e4

Add figures

Browse files
figures/convergence-curve-pretraining.png ADDED
figures/learning-curve-evaluation-benchmarks-in-english.png ADDED
figures/learning-curve-evaluation-french-bench.png ADDED
figures/learning-curve-evaluation-multilingual-arc-benchmark.png ADDED
figures/needle-in-a-haystack/Lucie-7B-annealing.png ADDED
figures/needle-in-a-haystack/Lucie-7B-extension.png ADDED
figures/needle-in-a-haystack/Lucie-7B-main.png ADDED
metadata/training_logs/1_pretraining.csv ADDED
@@ -0,0 +1,387 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ training_steps,training_tokens,training_loss,walltime,gputime,learning_rate
2
+ 1865,1955594240,5.540690021514893,0.7975208023282855,408.33065079208217,7.161599933169782e-05
3
+ 3790,4114087936,3.00810284614563,1.6655786928923269,852.7762907608713,0.00015066239575389773
4
+ 5313,6110314496,2.6936181354522706,2.419871313692319,1238.9741126104673,0.0002237664011772722
5
+ 7402,9252634624,2.525974779129028,3.525297917627946,1804.9525338255085,0.000299999926937744
6
+ 9616,13150715904,2.3369024181365967,4.839082616515136,2477.6102996557497,0.00029999829712323844
7
+ 11655,17390895104,2.276480207443237,6.211945974481024,3180.5163389342842,0.00029999419348314404
8
+ 13968,23110877184,2.2075621032714845,7.9962738511337434,4094.0922117804766,0.00029998470563441515
9
+ 15910,28752740352,2.1356527423858642,9.694892240850036,4963.7848273152185,0.0002999709395226091
10
+ 17655,34566045696,2.187712240219116,11.403844424466975,5838.768345327091,0.000299952196655795
11
+ 19561,41906601984,2.1478218460083007,13.512215933090056,6918.254557742109,0.0002999218995682895
12
+ 21599,50454593536,2.108550395965576,15.963069546280945,8173.091607695844,0.00029987728339619935
13
+ 22929,56033017856,2.08190354347229,17.555688533989233,8988.512529402487,0.0002998427371494472
14
+ 24879,64211910656,2.063254547119141,20.013122328626352,10246.718632256692,0.00029978438396938145
15
+ 26435,70738247680,2.0415384721755983,22.095803890509558,11313.051591940894,0.00029973124037496746
16
+ 28408,79013609472,2.0200176572799684,24.742381052788318,12668.099099027619,0.0002996554540004581
17
+ 30712,88677285888,2.009900689125061,27.834729681593487,14251.381596975865,0.00029955507488921285
18
+ 32719,97095254016,1.997526035308838,30.529593802309225,15631.152026782323,0.00029945719870738685
19
+ 34620,105068625920,1.9822303581237792,33.088940144262565,16941.537353862434,0.00029935556813143194
20
+ 37346,116502298624,1.9698213863372802,36.75907196609288,18820.644846639556,0.0002991946239490062
21
+ 39233,124416950272,1.9548299026489258,39.309065435078146,20126.24150276001,0.00029907276621088386
22
+ 40846,131182362624,1.951894497871399,41.485595499098174,21240.624895538265,0.0002989618224091828
23
+ 42956,140032344064,1.9350375270843505,44.33258508443755,22698.283563232024,0.0002988072519656271
24
+ 44939,148349648896,1.9308405590057374,47.01033564593735,24069.291850719925,0.00029865227406844497
25
+ 46768,156021030912,1.9222344255447388,49.47616130587861,25331.79458860985,0.00029850099235773087
26
+ 48882,164887789568,1.9197535276412965,52.35843906614072,26807.520801864048,0.00029831615393050015
27
+ 50706,172538200064,1.912511978149414,54.81750936548799,28066.564795129852,0.00029814810841344297
28
+ 52358,179467190272,1.907254514694214,57.04858303080124,29208.874511770235,0.0002979890559799969
29
+ 53857,185754451968,1.9000687837600707,59.087224691490626,30252.6590420432,0.00029783911304548383
30
+ 56177,195485237248,1.897755184173584,62.22864813516615,31861.067845205067,0.0002975965035147965
31
+ 58449,205014695936,1.8944290637969972,65.2858384233869,33426.349272774096,0.0002973465307150036
32
+ 60122,212031766528,1.885751051902771,67.54159698311844,34581.29765535664,0.00029715464916080236
33
+ 62532,222140039168,1.8851725578308105,70.83456301174779,36267.29626201487,0.0002968665794469416
34
+ 64820,231736606720,1.875647120475769,73.90514303951494,37839.43323623165,0.0002965803723782301
35
+ 67145,241488363520,1.8717713451385498,77.07916604916753,39464.53301717377,0.0002962769358418882
36
+ 69418,251022016512,1.8673745584487915,80.13863182021647,41030.979491950835,0.00029596799868158996
37
+ 70813,256873070592,1.8599135208129882,82.01500265421159,41991.68135895633,0.0002957723627332598
38
+ 73257,267123949568,1.8653322410583497,85.29531497185086,43671.20126558764,0.00029541869298554957
39
+ 74982,274359123968,1.8608362674713135,87.62793199229114,44865.50118005306,0.0002951606293208897
40
+ 76878,282311524352,1.8511997079849243,90.18867519614679,46176.601700427156,0.00029486900893971324
41
+ 79546,293501927424,1.8491973686218262,93.81005986021884,48030.75064843205,0.0002944445004686713
42
+ 81690,302494515200,1.8464100503921508,96.70326836685862,49512.073403831615,0.00029409138369373977
43
+ 83549,310291726336,1.847438826560974,99.1904628869521,50785.51699811948,0.0002937766257673502
44
+ 86305,321851228160,1.8410191345214844,102.87963783826878,52674.374573193614,0.00029329530661925673
45
+ 87821,328209793024,1.8423179483413696,104.95056398590805,53734.68876078492,0.0002930230984929949
46
+ 89624,335772123136,1.8375876760482788,107.36518641720015,54970.97544560648,0.00029269250808283687
47
+ 91624,344160731136,1.8310838747024536,110.08795571576191,56365.0333264701,0.00029231709777377546
48
+ 94363,355648929792,1.8548078060150146,113.81864126870086,58275.14432957484,0.0002917881647590548
49
+ 96400,364192727040,1.8301151275634766,116.56662833875006,59682.11370944003,0.000291383737931028
50
+ 98285,372098990080,1.825421872138977,119.117614549062,60988.21864911974,0.0002910011389758438
51
+ 99761,378289782784,1.828107204437256,121.11521686213217,62010.99103341167,0.0002906959562096745
52
+ 101660,386254766080,1.8232884979248047,123.67062551266181,63319.36026248285,0.0002902960986830294
53
+ 102797,391023689728,1.8284022760391236,125.20398345403945,64104.439528468196,0.0002900528197642416
54
+ 104466,398023983104,1.8237161207199097,127.48315444020703,65271.375073386,0.0002896904479712248
55
+ 106544,406739746816,1.8209293365478516,130.37456855220438,66751.77909872864,0.0002892306074500084
56
+ 108654,415589728256,1.813926682472229,133.23329634167231,68215.44772693623,0.0002887538284994662
57
+ 110464,423181418496,1.8184159755706788,135.6778559462546,69467.06224448235,0.00028833700343966484
58
+ 113489,435869188096,1.8118912790502821,139.763474430842,71558.8989085911,0.0002876241924241185
59
+ 115252,443263746048,1.8117296981811524,142.15046140346098,72781.03623857202,0.0002871994802262634
60
+ 116687,449282572288,1.8083889770507813,144.07612540268525,73766.97620617485,0.00028684877906925976
61
+ 119106,459428593664,1.8078338527679443,147.34195422902133,75439.08056525892,0.0002862474066205323
62
+ 121073,467678789632,1.808600254058838,149.9825934684281,76791.08785583518,0.00028574903262779117
63
+ 122958,475585052672,1.7992388534545898,152.52741438835127,78094.03616683585,0.0002852636098396033
64
+ 124630,482597928960,1.8028169393539428,154.7699610268964,79242.22004577096,0.000284826586721465
65
+ 126358,489845686272,1.8049570083618165,157.08673302915514,80428.40731092743,0.00028436866705305874
66
+ 128236,497722589184,1.8046081829071046,159.62801352390318,81729.54292423843,0.00028386374469846487
67
+ 130926,509005266944,1.7990764093399048,163.24389280245367,83580.87311485628,0.0002831274177879095
68
+ 132497,515594518528,1.794449429512024,165.37276175601704,84670.85401908073,0.0002826903073582798
69
+ 135020,526176747520,1.799844126701355,168.77996288452678,86415.34099687771,0.0002819774381350726
70
+ 136777,533546139648,1.7945206785202026,171.14425289637555,87625.85748294428,0.0002814731269609183
71
+ 139067,543151095808,1.7959566926956176,174.2254043554347,89203.40702998257,0.00028080615447834134
72
+ 140573,549467717632,1.8007401327292125,176.28781320277358,90259.36035982007,0.00028036159346811473
73
+ 142990,559605350400,1.790226879119873,179.52959129813544,91919.15074464535,0.00027963833417743444
74
+ 144717,566848913408,1.7950389575958252,181.85608978862726,93110.31797177716,0.00027911417419090867
75
+ 147235,577410170880,1.787166004180908,185.26090506606388,94853.58339382471,0.0002783390518743545
76
+ 149646,587522637824,1.7874946737289428,188.52470080996864,96524.64681470394,0.00027758482610806823
77
+ 152254,598461382656,1.7876266622543335,192.05541756899714,98332.37379532654,0.00027675574528984725
78
+ 153497,603674902528,1.7870644330978394,193.73033408004696,99189.93104898404,0.0002763558004517108
79
+ 155091,610360623104,1.7821037721633912,195.90790076383007,100304.845191081,0.00027583842165768147
80
+ 157080,618703093760,1.7816897821426392,198.57830149684983,101672.09036638711,0.00027518573915585876
81
+ 159083,627104284672,1.7820564126968383,201.2724561886386,103051.49756858297,0.00027452060021460056
82
+ 160774,634196852736,1.7861653804779052,203.56257374424416,104224.03775705301,0.00027395293000154197
83
+ 162912,643164274688,1.7875264167785645,206.45703154702142,105706.00015207496,0.00027322719688527286
84
+ 164504,649841606656,1.7890281875928242,208.59348014120815,106799.86183229857,0.00027268106350675225
85
+ 166091,656497967104,1.7810548543930054,210.73933679447862,107898.54043877305,0.0002721317869145423
86
+ 168208,665377308672,1.7790726804733277,213.599654382709,109363.023043947,0.0002713915309868753
87
+ 170798,676240556032,1.7792628765106202,217.10682849737182,111158.69619065437,0.0002704742655623704
88
+ 173178,686222999552,1.7723922634124756,220.3199128641763,112803.79538645827,0.00026962021365761757
89
+ 175477,695865704448,1.7773733282089232,223.42362898933956,114392.89804254186,0.0002687851374503225
90
+ 177319,703591612416,1.7751788663864136,225.91816708412628,115670.10154707266,0.0002681089681573212
91
+ 179285,711837614080,1.773937292098999,228.55287790967017,117019.07348975113,0.00026738038286566734
92
+ 181269,720159113216,1.7757549047470094,231.22714584410738,118388.29867218298,0.00026663794415071607
93
+ 182472,725204860928,1.7724720859527587,232.86696931923933,119227.88829145054,0.00026618424453772604
94
+ 183903,731206909952,1.7717243003845216,234.78111301701074,120207.9298647095,0.0002656411670614034
95
+ 186140,740589568000,1.7713909292221068,237.79312360069395,121750.0792835553,0.00026478481595404446
96
+ 188681,751247294464,1.7675371074676514,241.22784125286,123508.65472146432,0.00026380125200375915
97
+ 190733,759854006272,1.770745587348938,243.9752635411917,124915.33493309015,0.00026299862656742334
98
+ 192470,767139512320,1.7692016410827636,246.32182628229677,126116.77505653595,0.00026231343508698046
99
+ 194277,774718619648,1.7701459550857543,248.78614687347053,127378.50719921691,0.00026159503613598645
100
+ 195792,781072990208,1.7644649791717528,250.8326970803534,128426.34090514094,0.00026098836679011583
101
+ 198006,790359179264,1.7638875579833984,253.84096263601015,129966.5728696372,0.0002600947336759418
102
+ 199191,795329429504,1.7611687517166137,255.44903973567233,130789.90834466423,0.0002596129779703915
103
+ 201402,804603035648,1.7673775005340575,258.4134064139623,132307.6640839487,0.0002587077615316957
104
+ 203281,812484132864,1.7651812601089478,260.93973858808965,133601.1461571019,0.00025793202803470194
105
+ 204903,819287293952,1.7612399768829345,263.1386813809225,134727.00486703232,0.00025725766317918897
106
+ 206833,827382300672,1.7639558458328246,265.7458362247564,136061.86814707526,0.00025644959532655776
107
+ 209473,838455263232,1.7610173416137695,269.30228242161377,137882.76859986625,0.00025533439475111663
108
+ 211312,846168588288,1.759269905090332,271.79195232891027,139157.47959240206,0.00025455086142756045
109
+ 213227,854200680448,1.7630469226837158,274.3732711282936,140479.11481768632,0.00025372920208610594
110
+ 215847,865189756928,1.7609153509140014,277.9089087475696,142289.36127875565,0.00025259560788981616
111
+ 217356,871518961664,1.7579796981811524,279.9560044781967,143337.4742928367,0.0002519378322176635
112
+ 219172,879135817728,1.7634607887268066,282.2283169674041,144500.8982873109,0.0002511414932087064
113
+ 222150,891626455040,1.7575871229171753,285.76951404988944,146313.9911935434,0.000249824661295861
114
+ 223853,898769354752,1.758929591178894,287.8028389246148,147355.05352940276,0.000249065546086058
115
+ 225145,904188395520,1.7552893447875977,289.3465113940193,148145.41383373787,0.00024848670000210404
116
+ 226928,911666839552,1.751872878074646,291.46764839194356,149231.4359766751,0.00024768381263129413
117
+ 227290,913185177600,1.756528417269389,291.96401525144137,149485.57580873798,0.00024752022000029683
118
+ 229251,921410207744,1.7513698720932007,294.6090092848358,150839.81275383593,0.0002466306905262172
119
+ 231418,930499264512,1.7536880302429199,297.51891305445406,152329.68348388048,0.00024564118939451873
120
+ 232756,936111243264,1.7567100238800049,299.3165955686522,153250.09693114992,0.0002450268075335771
121
+ 234747,944462102528,1.7647992753982544,301.9957619307597,154621.83010854898,0.00024410788319073617
122
+ 236656,952469028864,1.759487557411194,304.57088663315164,155940.29395617364,0.00024322151148226112
123
+ 238174,958835982336,1.760659966468811,306.609707181298,156984.17007682458,0.00024251305148936808
124
+ 240473,968478687232,1.754089126586914,309.6952503732213,158563.9681910893,0.00024143399787135422
125
+ 242768,978104614912,1.7523606967926026,312.7955168953312,160151.30465040958,0.00024034960370045155
126
+ 244243,984291213312,1.7513489294052125,314.7860534490736,161170.4593659257,0.00023964889987837523
127
+ 245978,991568330752,1.7525734424591064,317.14563100169306,162378.56307286685,0.0002388209686614573
128
+ 248018,1000124710912,1.7499562788009644,319.9244360278179,163801.31124624275,0.00023784241057001054
129
+ 249886,1007959670784,1.7509011316299439,322.45674101591925,165097.85140015066,0.00023694158880971372
130
+ 251357,1014129491968,1.7487824440002442,324.4791230457826,166133.3109994407,0.00023622905428055674
131
+ 252908,1020634857472,1.7466645431518555,326.58318169927685,167210.58903002975,0.00023547477030660957
132
+ 255231,1030378225664,1.7461054420471191,329.7483051645596,168831.13224425452,0.00023433937167283148
133
+ 256784,1036891979776,1.7455737209320068,331.87937631953497,169922.2406756019,0.00023357658938039094
134
+ 258875,1045662269440,1.7457578039169313,334.731770063507,171382.66627251558,0.00023254487314261496
135
+ 260952,1054373838848,1.7445444059371948,337.5729135723391,172837.3317490376,0.00023151483037509024
136
+ 262527,1060979867648,1.741570553779602,339.7600884215228,173957.1652718197,0.00023073032207321376
137
+ 264811,1070559657984,1.740293960571289,342.88246244716737,175555.8207729497,0.00022958747285883874
138
+ 266420,1077308293120,1.7429306316375732,345.0796831538715,176680.7977747822,0.00022877875017002225
139
+ 268308,1085227139072,1.744356451034546,347.65395848590146,177998.82674478155,0.00022782600717619061
140
+ 269559,1090474213376,1.7428239250183106,349.3460391367528,178865.17203801742,0.00022719251865055412
141
+ 271809,1099911397376,1.7441408920288086,352.4030745393055,180430.37416412443,0.00022604875266551971
142
+ 274026,1109210169344,1.7409614515304566,355.4137426253769,181971.836224193,0.00022491635172627866
143
+ 276200,1118328586240,1.7383298921585082,358.35239492776185,183476.42620301407,0.00022380080190487206
144
+ 279273,1131217682432,1.7371133943883383,362.5356638252912,185618.25987854908,0.00022221545805223286
145
+ 281879,1142148038656,1.7415697383880615,366.08764050357314,187436.87193782945,0.00022086345416028053
146
+ 283803,1150217879552,1.733839235305786,368.711490408947,188780.28308938086,0.0002198608999606222
147
+ 286646,1162142285824,1.7431989669799806,372.57302363409656,190757.38810065744,0.00021837285021319985
148
+ 288431,1169629118464,1.739035325050354,375.010250589488,192005.24830181786,0.00021743458637502044
149
+ 289869,1175660527616,1.733289074897766,376.974500554174,193010.9442837371,0.00021667654800694436
150
+ 292495,1186674769920,1.7315478420257568,380.5901660706151,194862.16502815494,0.00021528734941966832
151
+ 295211,1198066499584,1.733106060028076,384.2868979963938,196754.89177415363,0.00021384400315582752
152
+ 297195,1206387998720,1.7410508108139038,386.9830853080197,198135.3396777061,0.00021278555504977703
153
+ 298925,1213644144640,1.734706358909607,389.3362156377362,199340.14240652093,0.00021185987861827016
154
+ 300889,1221881757696,1.73500732421875,392.0144997620718,200711.42387818077,0.00021080594160594046
155
+ 302662,1229318258688,1.7418826770782472,394.42366878365414,201944.91841723092,0.00020985178707633168
156
+ 304473,1236914143232,1.7310642719268798,396.8901623202001,203207.76310794245,0.00020887458231300116
157
+ 306476,1245315334144,1.7335577774047852,399.62016656875767,204605.52528320393,0.0002077907556667924
158
+ 308518,1253880102912,1.7285248136520386,402.42136555782304,206039.7391656054,0.00020668267097789794
159
+ 310574,1262503591936,1.733399453163147,405.2239282874955,207474.6512831977,0.00020556384697556496
160
+ 312781,1271760420864,1.730285539627075,408.3160193839066,209057.8019245602,0.00020435944315977395
161
+ 314612,1279440191488,1.733029899597168,410.81628584431604,210337.9383522898,0.0002033576020039618
162
+ 316647,1287975600128,1.731144299507141,413.5669219080583,211746.26401692585,0.00020224145555403084
163
+ 319268,1298968870912,1.729205231666565,416.9754043734229,213491.40703919253,0.00020079984096810222
164
+ 320942,1305990135808,1.7270024967193605,418.99536520049196,214525.62698265188,0.00019987679843325168
165
+ 323057,1314861088768,1.7282018089294433,421.5389839125796,215827.95976324077,0.0001987080613616854
166
+ 324688,1321701998592,1.727188115119934,423.4974892227614,216830.71448205382,0.0001978049403987825
167
+ 325782,1326290567168,1.7273711681365966,424.81380970657455,217504.67056976617,0.00019719830015674233
168
+ 328381,1337191563264,1.7251214504241943,427.95426755280107,219112.58498703415,0.00019575434271246195
169
+ 330172,1344703561728,1.7238556051254272,430.13040363881544,220226.7666630735,0.0001947571145137772
170
+ 332386,1353989750784,1.7249505424499512,432.82640173550084,221607.11768857643,0.00019352202070876956
171
+ 334457,1362676154368,1.7219355773925782,435.3433918957347,222895.81665061618,0.00019236441585235298
172
+ 336042,1369324126208,1.72216224193573,437.23262586380747,223863.10444226942,0.0001914770546136424
173
+ 337560,1375691079680,1.7201724815368653,439.0764706419921,224807.15296869996,0.00019062607316300273
174
+ 338748,1380673912832,1.7237060013271512,440.51795167100903,225545.19125555662,0.00018995934806298465
175
+ 340933,1389838467072,1.7192433309555053,443.16397790562996,226899.95668768254,0.00018873147200793028
176
+ 342332,1395706298368,1.7205400276184082,444.8788870963792,227777.99019334614,0.00018794421339407563
177
+ 344302,1403969077248,1.7203184127807618,447.26131719529656,228997.79440399184,0.00018683428061194718
178
+ 345968,1410956787712,1.720102686882019,449.29559398623377,230039.3441209517,0.0001858944451669231
179
+ 347502,1417390850048,1.720639853477478,451.16690631718046,230997.4560343964,0.0001850281551014632
180
+ 349991,1427830472704,1.7197772884368896,454.2264308151358,232563.93257734954,0.00018362075206823647
181
+ 352052,1436474933248,1.7142712354660035,456.81043543511663,233886.94294277971,0.00018245380488224328
182
+ 354054,1444871929856,1.7160462188720702,459.453295596552,235240.08734543461,0.00018131898832507432
183
+ 355452,1450735566848,1.7196543836593627,461.30180067648325,236186.52194635943,0.0001805258507374674
184
+ 357345,1458675384320,1.7139066517353059,463.9481888459466,237541.47268912467,0.00017945100262295455
185
+ 359270,1466749419520,1.7155957555770873,466.5414323894858,238869.21338341673,0.00017835704784374684
186
+ 361118,1474500493312,1.7160774993896484,469.038111011795,240147.51283803905,0.0001773060066625476
187
+ 363330,1483778293760,1.716701912879944,472.0381865977867,241683.5515380668,0.0001760469749569893
188
+ 365934,1494700261376,1.7132270240783691,475.5838090708114,243498.91024425544,0.0001745635672705248
189
+ 367689,1502061264896,1.7156493997573852,477.96177319155544,244716.42787407638,0.00017356315220240504
190
+ 369392,1509204164608,1.7158074569702149,480.26144646743097,245893.86059132466,0.00017259192827623338
191
+ 371001,1515952799744,1.715286192893982,482.4363151349443,247007.3933490915,0.00017167393525596708
192
+ 373275,1525490647040,1.712519268989563,485.5007621892336,248576.39024088762,0.0001703760353848338
193
+ 375882,1536425197568,1.7095552492141723,489.0597850280192,250398.60993434582,0.00016888746176846325
194
+ 378308,1546600579072,1.7065624713897705,492.38097163360044,252099.05747640342,0.0001675017992965877
195
+ 380320,1555039518720,1.7018836069107055,495.12656034960173,253504.7988989961,0.00016635240172035992
196
+ 381423,1559665836032,1.712470350265503,496.61899996729517,254268.92798325513,0.00016572224558331072
197
+ 384180,1571229532160,1.70383526802063,500.35769686916666,256183.14079701333,0.0001641471026232466
198
+ 386393,1580511526912,1.7026185607910156,503.3778410258571,257729.45460523883,0.0001628828322282061
199
+ 388342,1588686225408,1.703223738670349,506.02086480091833,259082.6827780702,0.00016176952340174466
200
+ 389711,1594428227584,1.7003657007217408,507.8724601782208,260030.69961124906,0.00016098766354843974
201
+ 391539,1602095415296,1.6955804538726806,510.35535180901985,261301.94012621816,0.000159943854669109
202
+ 393541,1610492411904,1.6995060205459596,513.0894965861288,262701.82225209795,0.00015880104911047965
203
+ 395741,1619719880704,1.7011328125,516.0827098217362,264234.34742872894,0.00015754574269521981
204
+ 398543,1631472320512,1.7008357238769531,519.8783201021156,266177.6998922832,0.000155947869643569
205
+ 400565,1639953203200,1.695354881286621,522.6083451644168,267575.4727241814,0.000154795590788126
206
+ 402527,1648182427648,1.6946661138534547,525.2668895711776,268936.6474604429,0.00015367820742540061
207
+ 403945,1654129950720,1.6960719108581543,527.1914807180993,269922.03812766686,0.00015287112910300493
208
+ 405313,1659867758592,1.697601842880249,529.0370912977813,270866.99074446404,0.0001520929072285071
209
+ 406788,1666054356992,1.6927887296676636,531.0349452814536,271889.89198410424,0.000151254324009642
210
+ 409387,1676955353088,1.69403892993927,534.5653393185967,273697.4537311215,0.00014977800310589373
211
+ 411155,1684370882560,1.691395869255066,536.9709613285441,274929.1322002146,0.0001487747795181349
212
+ 413089,1692482666496,1.6922231245040893,539.5964738321247,276273.39460204786,0.00014767838001716882
213
+ 414732,1699373907968,1.6947362184524537,541.7996070183738,277401.39879340737,0.00014674787234980613
214
+ 416436,1706521001984,1.6899096250534058,544.0972045027258,278577.76870539563,0.0001457837497582659
215
+ 417863,1712506273792,1.6881012392044068,545.9991046208406,279551.5415658704,0.0001449771225452423
216
+ 419237,1718269247488,1.691229190826416,547.8411327792202,280494.65998296073,0.00014420114166568965
217
+ 421285,1726859182080,1.6884182071685792,550.5797526560408,281896.8333598929,0.00014304582145996392
218
+ 424030,1738372546560,1.6905427885055542,554.2562450989998,283779.1974906879,0.0001414999132975936
219
+ 426068,1746920538112,1.6902626466751098,557.0390545870044,285203.9959485463,0.00014035419735591859
220
+ 427364,1752356356096,1.6865596675872803,558.7964797184435,286103.79761584307,0.0001396265724906698
221
+ 429281,1760396836864,1.6867563104629517,561.412200978866,287443.0469011794,0.00013855169527232647
222
+ 432073,1772107333632,1.6869472122192384,565.1983350192647,289381.54752986354,0.0001369893434457481
223
+ 433564,1778361040896,1.6836950588226318,567.304830058465,290460.07298993407,0.00013615658099297434
224
+ 435451,1786275692544,1.682388458251953,569.8640171882282,291770.37680037285,0.0001351043174508959
225
+ 436975,1792667811840,1.6846920680999755,571.9226006626227,292824.3715392628,0.00013425585348159075
226
+ 439052,1801379381248,1.6854845952987672,574.7291045055265,294261.30150682956,0.00013310158101376146
227
+ 440966,1809407279104,1.6832574605941772,577.334748714053,295595.39134159515,0.000132040077005513
228
+ 443227,1818890600448,1.6824970388412475,580.0256552258438,296973.13547563204,0.00013078891788609326
229
+ 445179,1827077881856,1.6792938709259033,582.2320406351785,298102.8048052114,0.00012971126125194132
230
+ 447934,1838633189376,1.6817043399810792,585.2915628118471,299669.28015966574,0.00012819441326428205
231
+ 449710,1846082273280,1.6782426357269287,587.2799727474821,300687.34604671085,0.00012721921666525304
232
+ 452524,1857885044736,1.6820737409591675,590.4152128734677,302292.58899121545,0.00012567844532895833
233
+ 454426,1865862610944,1.6787804031372071,592.544702035982,303382.8874424228,0.00012464018072932959
234
+ 456289,1873676599296,1.681654314994812,594.6135522814485,304442.13876810163,0.00012362573761492968
235
+ 458583,1883298332672,1.6730261087417602,597.1887328853452,305760.63123729674,0.00012238013732712716
236
+ 460981,1893356273664,1.672509970664978,599.8846338122594,307140.9325118768,0.00012108237569918856
237
+ 462659,1900394315776,1.6751681089401245,601.761152752987,308101.7102095293,0.00012017694825772196
238
+ 464445,1907885342720,1.671654839515686,603.7396514160039,309114.701524994,0.0001192157287732698
239
+ 465759,1913396658176,1.6746321821212768,605.2131596599303,309869.1377458843,0.00011851020099129528
240
+ 467953,1922598961152,1.6700562286376952,607.6498691665898,311116.733013294,0.00011733539577107877
241
+ 470480,1933197967360,1.6706900882720948,610.4716868327131,312561.5036583491,0.000115987379103899
242
+ 472362,1941091647488,1.670096197128296,612.5532063371096,313627.2416446001,0.00011498706589918584
243
+ 474479,1949970989056,1.672633442878723,614.9096566533404,314833.7442065103,0.00011386564437998459
244
+ 475718,1955167731712,1.670849027633667,616.2887570926199,315539.8436314214,0.00011321121564833447
245
+ 477678,1963388567552,1.6683212041854858,618.4896738461576,316666.7130092327,0.00011217887367820367
246
+ 479405,1970632130560,1.6645479679107666,620.5010973832233,317696.5618602103,0.00011127226753160357
247
+ 481490,1979377254400,1.6670279264450074,622.9267749258964,318938.508762059,0.00011018155055353418
248
+ 483395,1987367403520,1.6700633478164673,625.1531309454637,320078.4030440774,0.00010918873158516362
249
+ 484660,1992673198080,1.665715732574463,626.6322892833138,320835.73211305664,0.00010853145067812875
250
+ 486388,1999920955392,1.6684691619873047,628.6401761626622,321863.77019528305,0.00010763623140519485
251
+ 487760,2005675540480,1.6899535751342774,630.2428063296861,322684.3168407993,0.00010692761861719191
252
+ 489679,2013724409856,1.6870677042007447,632.4661689389804,323822.678496758,0.00010593978367978707
253
+ 491707,2022230458368,1.683250117301941,634.8101256682668,325022.7843421526,0.00010490007844055071
254
+ 493795,2030988165120,1.6834172248840331,637.2431921455864,326268.5143785402,0.00010383423796156421
255
+ 495562,2038399500288,1.6794629859924317,639.3029817152653,327323.12663821585,0.00010293597733834758
256
+ 497249,2045475291136,1.6757073497772217,641.2957346101667,328343.41612040537,0.00010208162711933255
257
+ 499270,2053951979520,1.6750809478759765,643.6868357490317,329567.65990350425,0.00010106235276907682
258
+ 501917,2065054302208,1.6795844745635986,646.8346841992168,331179.358309999,9.973444684874266e-05
259
+ 504501,2075892383744,1.6791791486740113,649.9504268523947,332774.6185484261,9.844604937825352e-05
260
+ 505920,2081844101120,1.6794115686416626,651.6345685349747,333636.89908990706,9.774190402822569e-05
261
+ 507735,2089456762880,1.6736989212036133,653.8060941808081,334748.72022057377,9.684478573035449e-05
262
+ 509781,2098038308864,1.6739554023742675,656.2464802015245,335998.19786318054,9.583831706549972e-05
263
+ 511352,2104627560448,1.6752780342102052,658.1126194182707,336953.6611421546,9.50690227909945e-05
264
+ 512792,2110667358208,1.6734393405914307,659.8307014095916,337833.3191217109,9.436659456696361e-05
265
+ 514516,2117898338304,1.6712262153625488,661.8950038467839,338890.24196955335,9.352908818982542e-05
266
+ 516978,2128224714752,1.670324263572693,664.8669427431114,340411.87468447303,9.233966557076201e-05
267
+ 519037,2136860786688,1.6687762260437011,667.3400292872751,341678.09499508486,9.135099389823154e-05
268
+ 520968,2144959987712,1.6679233884811402,669.6824876479714,342877.43367576133,9.042886085808277e-05
269
+ 523402,2155168923648,1.6661997795104981,672.6088669252829,344375.73986574484,8.927362796384841e-05
270
+ 525489,2163922436096,1.667925977706909,675.1255832034733,345664.2986001783,8.828948193695396e-05
271
+ 527861,2173871325184,1.6650946950912475,677.9682576375237,347119.74791041214,8.71782103786245e-05
272
+ 529924,2182524174336,1.6656306838989259,680.4565219490987,348393.73923793854,8.621808228781447e-05
273
+ 531732,2190107475968,1.6626214647293092,682.6502407960503,349516.92328757775,8.538155816495419e-05
274
+ 533512,2197573337088,1.6671656608581542,684.8457100742603,350641.0035580213,8.456254727207124e-05
275
+ 535979,2207920685056,1.6623902940750122,687.8088072102388,352158.10929164226,8.343499212060124e-05
276
+ 538441,2218247061504,1.6625070142745972,690.8060061686901,353692.67515836935,8.231859101215377e-05
277
+ 539611,2223154397184,1.6638607454299927,692.2126377288533,354412.8705171729,8.179118594853207e-05
278
+ 541835,2232482529280,1.6632291555404664,694.9052978233159,355791.51248553774,8.07943069958128e-05
279
+ 543513,2239520571392,1.6591477632522582,696.9350562256079,356830.74878751126,8.004709525266662e-05
280
+ 544834,2245061246976,1.6582461643218993,698.5149932145702,357639.6765258599,7.946186815388501e-05
281
+ 546541,2252220923904,1.6569991302490235,700.5529090210858,358683.08941879595,7.870959962019697e-05
282
+ 548386,2259959414784,1.6534462308883666,702.7537166691092,359809.9029345839,7.790157542331144e-05
283
+ 550107,2267177811968,1.6563579320907593,704.8354743227627,360875.7628532545,7.715264655416831e-05
284
+ 551929,2274819833856,1.652633171081543,707.0277662998835,361998.21634554036,7.636484951945022e-05
285
+ 553695,2282226974720,1.6558599662780762,709.1717543189402,363095.9382112974,7.560628728242591e-05
286
+ 555203,2288551985152,1.6531555843353272,710.9832715057181,364023.4350109277,7.496249600080773e-05
287
+ 557414,2297825591296,1.658251905441284,713.6318656351355,365379.51520518935,7.402522169286385e-05
288
+ 559184,2305249509376,1.6528305721282959,715.7625381680682,366470.4195420509,7.328063657041639e-05
289
+ 561554,2315190009856,1.6525710678100587,718.5529555138435,367899.11322308786,7.229171023936942e-05
290
+ 562859,2320663576576,1.651809163093567,720.0850926347636,368683.56742899894,7.175114296842366e-05
291
+ 565169,2330352418816,1.6541441106796264,722.8299401458767,370088.92935468885,7.080127397784963e-05
292
+ 567134,2338594226176,1.6530265474319459,725.1825567879315,371293.46907542093,7.000035111559555e-05
293
+ 569714,2349415530496,1.6502766132354736,728.2318701144516,372854.7174985992,6.895873957546428e-05
294
+ 571324,2356168359936,1.651259379386902,730.1611943364807,373842.5315002781,6.831453356426209e-05
295
+ 574248,2368432504832,1.6482793807983398,733.6468108414203,375627.1671508072,6.715607014484704e-05
296
+ 576334,2377181822976,1.644599723815918,736.1349917772365,376901.11578994506,6.63387545500882e-05
297
+ 578782,2387449479168,1.6505870151519775,739.0739024216011,378405.8380398598,6.538941670442e-05
298
+ 580516,2394722402304,1.6470952558517455,741.1653197225369,379476.6436979389,6.472343375207856e-05
299
+ 582668,2403748544512,1.6487390184402466,743.7736260008398,380812.09651243,6.390442285919562e-05
300
+ 584669,2412141346816,1.6448141527175903,746.1684989620086,382038.2714685484,6.315039354376495e-05
301
+ 586818,2421154906112,1.646544461250305,748.775306891994,383372.9571287009,6.234873580979183e-05
302
+ 589001,2430311071744,1.6475654029846192,751.4020625109125,384717.8560055872,6.154309085104614e-05
303
+ 591113,2439169441792,1.6450910377502441,753.9527754451661,386023.82102792506,6.077205034671351e-05
304
+ 593228,2448040394752,1.6460276412963868,756.5354509950179,387346.15090944915,6.0008260334143415e-05
305
+ 594605,2453815951360,1.6390350008010863,758.1939404388438,388195.29750468803,5.951549974270165e-05
306
+ 597153,2464503037952,1.637695927619934,761.2579777614424,389764.0846138585,5.8613157307263464e-05
307
+ 598759,2471239090176,1.6295817375183106,763.2197744939834,390768.5245409195,5.8050762163475156e-05
308
+ 600863,2480063905792,1.6312317228317261,765.7391971130864,392058.46892190026,5.732145291403867e-05
309
+ 602611,2487395549184,1.6288472127914428,767.8321921499497,393130.08238077426,5.67220376979094e-05
310
+ 604453,2495121457152,1.6286871242523193,770.052846124808,394267.0572159017,5.609680010820739e-05
311
+ 606186,2502390185984,1.626492328643799,772.1559839896788,395343.86380271555,5.5514599807793275e-05
312
+ 606868,2505250701312,1.6294204493363698,772.9715416562736,395761.4293280121,5.528709516511299e-05
313
+ 608527,2512209051648,1.6318567752838136,774.9540150309879,396776.4556958658,5.4737502068746835e-05
314
+ 609912,2518018162688,1.6299163818359375,776.6102461643201,397624.4460361319,5.4282838391372934e-05
315
+ 612696,2529695105024,1.6283214426040649,779.9288053733754,399323.5483511682,5.338044138625264e-05
316
+ 614703,2538113073152,1.624042534828186,782.3509216785588,400563.67189942213,5.2739502280019224e-05
317
+ 617059,2547994853376,1.64087806224823,785.185724072868,402015.0907253084,5.19974491908215e-05
318
+ 618445,2553808158720,1.642133264541626,786.8536925417527,402869.0905813774,5.156615225132555e-05
319
+ 620429,2562129657856,1.6360367107391358,789.3145963788388,404129.0733459655,5.0955564802279696e-05
320
+ 622331,2570107224064,1.6323289918899535,791.575088748329,405286.44543914445,5.0377762818243355e-05
321
+ 624509,2579242418176,1.6336705684661865,794.1814413488138,406620.89797059266,4.9725240387488157e-05
322
+ 625869,2584946671616,1.63454927444458,795.8170756094358,407458.34271203115,4.932274896418676e-05
323
+ 628359,2595390488576,1.6327994871139526,798.8134971637937,408992.5105478624,4.8595778935123235e-05
324
+ 629947,2602051043328,1.634876675605774,800.7342204672283,409975.9208792209,4.8138899728655815e-05
325
+ 631685,2609340743680,1.6314793634414673,802.8435091534811,411055.87668658234,4.764491313835606e-05
326
+ 633406,2616559140864,1.630139570236206,804.9092638572018,412113.54309488734,4.716201510746032e-05
327
+ 635185,2624020807680,1.6301208162307739,807.085845822091,413227.9530609106,4.666941094910726e-05
328
+ 636953,2631436337152,1.6280076551437377,809.2236022954875,414322.4843752896,4.61865020042751e-05
329
+ 638743,2638944141312,1.6293576526641846,811.3632367052068,415417.97719306586,4.570435703499243e-05
330
+ 641116,2648897224704,1.6281150197982788,814.1943204899674,416867.4920908633,4.507573976297863e-05
331
+ 643518,2658971942912,1.626808156967163,817.0790731657314,418344.4854608545,4.4451757275965065e-05
332
+ 645200,2666026762240,1.6276088953018188,819.1059377729686,419382.2401397599,4.402222475619055e-05
333
+ 646903,2673169661952,1.6250265979766845,821.1531966983101,420430.4367095348,4.3593576265266165e-05
334
+ 649246,2682996916224,1.6259414958953857,823.9828252710962,421879.20653880126,4.3014148104703054e-05
335
+ 652174,2695277838336,1.6242961645126344,827.4949100539828,423677.3939476392,4.2306914110668004e-05
336
+ 654123,2703452536832,1.623875789642334,829.8569429315521,424886.7547809547,4.184658610029146e-05
337
+ 656975,2715414691840,1.621957221031189,833.2863375114017,426642.6048058377,4.1188093746313825e-05
338
+ 658694,2722624700416,1.626935167312622,835.3677633732899,427708.29484712443,4.07998995797243e-05
339
+ 660347,2729557884928,1.6193980121612548,837.4232649987488,428760.7116793594,4.0432812966173515e-05
340
+ 662381,2738089099264,1.6226060914993286,839.8879330915463,430022.6217428717,3.9989481592783704e-05
341
+ 664665,2747668889600,1.6181461000442505,842.6497982943592,431436.69672671193,3.9502705476479605e-05
342
+ 666543,2755545792512,1.6224317026138306,844.9222846863987,432600.20975943614,3.91112407669425e-05
343
+ 668035,2761803694080,1.6199790287017821,846.7305949538775,433526.06461638527,3.8805901567684487e-05
344
+ 669716,2768854319104,1.6205505228042603,848.7580324526521,434564.11261575785,3.8467911508632824e-05
345
+ 671548,2776538284032,1.6194300413131715,850.955616959009,435689.2758830126,3.81068566639442e-05
346
+ 673792,2785950302208,1.618673267364502,853.719618993503,437104.44492467353,3.767499583773315e-05
347
+ 675542,2793290334208,1.6200080347061157,855.8460515671908,438193.1784024017,3.7346177123254165e-05
348
+ 677331,2800793944064,1.6195425510406494,858.0345433676414,439313.6862042324,3.7017267459305e-05
349
+ 679068,2808079450112,1.6178715658187866,860.1212525989619,440382.0813306685,3.670493606477976e-05
350
+ 681041,2816354811904,1.619278564453125,862.4906129164967,441595.1938132463,3.6358578654471785e-05
351
+ 682885,2824089108480,1.6154922294616698,864.7152048434652,442734.1848798542,3.6042976716998965e-05
352
+ 684471,2830741274624,1.6204235363006592,866.6373480686542,443718.32221115095,3.577781535568647e-05
353
+ 686548,2839452844032,1.6162836599349975,869.1610365124169,445010.45069435745,3.5439366911305115e-05
354
+ 688512,2847690457088,1.6139143562316896,871.5226151134167,446219.57893806935,3.51285380020272e-05
355
+ 690770,2857161195520,1.6140108585357666,874.2491468152524,447615.56316940923,3.47822715411894e-05
356
+ 692653,2865059069952,1.612514958381653,876.5247202096293,448780.6567473302,3.450260192039423e-05
357
+ 695067,2875184119808,1.6133524417877196,879.4364041243566,450271.4389116706,3.4156193578382954e-05
358
+ 697014,2883350429696,1.613981342315674,881.7887764691106,451475.85355218465,3.3886746678035706e-05
359
+ 698746,2890614964224,1.613626651763916,883.8767033670731,452544.87212394143,3.3654530852800235e-05
360
+ 701452,2901964750848,1.6106493425369264,887.1577838441956,454224.78532822814,3.330586332594976e-05
361
+ 703123,2908973432832,1.6123323392868043,889.1781213319924,455259.1981219801,3.309917883598246e-05
362
+ 704756,2915822731264,1.6120843172073365,891.1229818776349,456254.96672134905,3.2903564715525135e-05
363
+ 706984,2925167640576,1.6084727716445923,893.7894090776833,457620.17744777387,3.264685801696032e-05
364
+ 708614,2932004356096,1.6117757987976074,895.7409758135146,458619.3796165195,3.2466501579619944e-05
365
+ 710902,2941600923648,1.6067533922195434,898.5324710232566,460048.62516390736,3.222398299840279e-05
366
+ 712655,2948953538560,1.6069003009796143,900.6844297494815,461150.42803173454,3.204659151379019e-05
367
+ 714446,2956465537024,1.6114630460739137,902.828141202652,462248.0082957578,3.187291440553963e-05
368
+ 716271,2964120141824,1.6145262241363525,905.0301348181574,463375.4290268966,3.170380659867078e-05
369
+ 718439,2973213392896,1.6167470741271972,907.64100600771,464712.1950759475,3.1513252906734124e-05
370
+ 720429,2981560057856,1.615721011161804,910.0234193265205,465931.9906951785,3.1348234188044444e-05
371
+ 722346,2989600538624,1.6144047689437866,912.3105630227034,467103.0082676241,3.119822940789163e-05
372
+ 724097,2996944764928,1.6113934993743897,914.3946314318039,468170.0512930836,3.106891381321475e-05
373
+ 726062,3005186572288,1.6150265264511108,916.7757618370391,469389.190060564,3.093255145358853e-05
374
+ 728431,3015122878464,1.6120684432983399,919.6071913458889,470838.8819690951,3.078047666349448e-05
375
+ 730800,3025059184640,1.6116199398040771,922.4232348246237,472280.69623020734,3.064189513679594e-05
376
+ 732477,3032093032448,1.60649507522583,924.4540804354708,473320.48918296106,3.0551960662705824e-05
377
+ 734574,3040888487936,1.6101762390136718,927.0469926404139,474648.0602318919,3.0449025871348567e-05
378
+ 736132,3047423213568,1.6150142908096314,929.0061162371555,475651.13151342364,3.0379411327885464e-05
379
+ 737978,3055165898752,1.6060742330551148,931.2157941709505,476782.48661552666,3.0304503525258042e-05
380
+ 740325,3065009930240,1.5952490282058716,934.0287519650691,478222.7210061154,3.0221137421904132e-05
381
+ 741809,3071234277376,1.5957392265922146,935.8210044055174,479140.35425562493,3.0175287974998355e-05
382
+ 744212,3081313189888,1.5954618740081787,938.745490208727,480637.6909868682,3.0112320018815808e-05
383
+ 746818,3092243546112,1.5991949892044068,941.9034764933106,482254.579964575,3.0059802156756632e-05
384
+ 748203,3098052657152,1.5962308692932128,943.6019803061257,483124.21391673636,3.0038570912438445e-05
385
+ 749620,3103995985920,1.597407283782959,945.3358183610009,484011.93900083244,3.0021647035027854e-05
386
+ 751627,3112413954048,1.5941282081604005,947.7966657645534,485271.89287145133,3.0005981898284517e-05
387
+ 753851,3121742086144,1.5939428043365478,950.4965920289866,486654.25511884113,2.9999999242136255e-05
metadata/training_logs/2_extension.csv ADDED
@@ -0,0 +1,50 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ training_steps,training_tokens,training_loss,walltime,gputime,learning_rate
2
+ 25,102400000,4.309772815704346,0.26922979407836667,34.461413642030934,1.9999999494757503e-05
3
+ 50,204800000,2.6095064067840577,0.5344728346386659,68.41252283374924,1.9999999494757503e-05
4
+ 75,307200000,2.1840998935699463,0.7994251677144519,102.32642146744985,1.9999999494757503e-05
5
+ 100,409600000,1.9669664239883422,1.066323601762864,136.48942102564658,1.9999999494757503e-05
6
+ 125,512000000,1.8014992809295653,1.3312230805499679,170.39655431039589,1.9999999494757503e-05
7
+ 150,614400000,1.7220078945159911,1.5958814102547383,204.2728205126065,1.9999999494757503e-05
8
+ 175,716800000,1.6801378870010375,1.8606056269711062,238.1575202523016,1.9999999494757503e-05
9
+ 200,819200000,1.6575293684005736,2.126106606112505,272.14164558240066,1.9999999494757503e-05
10
+ 225,921600000,1.639773211479187,2.391510878373186,306.1133924317678,1.9999999494757503e-05
11
+ 250,1024000000,1.6256405591964722,2.6564186745483864,340.02159034219346,1.9999999494757503e-05
12
+ 275,1126400000,1.614159688949585,2.922717172069702,374.10779802492186,1.9999999494757503e-05
13
+ 300,1228800000,1.6140153646469115,3.187462724955874,407.9952287943519,1.9999999494757503e-05
14
+ 325,1331200000,1.6082664394378663,3.453434134056764,442.0395691592658,1.9999999494757503e-05
15
+ 350,1433600000,1.5916787338256837,3.718426102696622,475.9585411451676,1.9999999494757503e-05
16
+ 375,1536000000,1.5903722620010377,3.9834007768839617,509.8752994411471,1.9999999494757503e-05
17
+ 400,1638400000,1.5937609910964965,4.248280933908213,543.7799595402513,1.9999999494757503e-05
18
+ 425,1740800000,1.5833971071243287,4.514559538923331,577.8636209821864,1.9999999494757503e-05
19
+ 450,1843200000,1.588483600616455,4.7794486881652904,611.7694320851572,1.9999999494757503e-05
20
+ 475,1945600000,1.5811502838134766,5.044273908992154,645.6670603509957,1.9999999494757503e-05
21
+ 500,2048000000,1.5776061773300172,5.308712950952829,679.5152577219621,1.9999999494757503e-05
22
+ 525,2150400000,1.5762306451797485,5.575429370141868,713.6549593781591,1.9999999494757503e-05
23
+ 550,2252800000,1.577330994606018,5.8403744682127785,747.5679319312356,1.9999999494757503e-05
24
+ 575,2355200000,1.5774771738052369,6.105136559348899,781.457479596659,1.9999999494757503e-05
25
+ 600,2457600000,1.571889362335205,6.370161146941192,815.3806268084726,1.9999999494757503e-05
26
+ 625,2560000000,1.5669999837875366,6.636842231150017,849.5158055872022,1.9999999494757503e-05
27
+ 650,2662400000,1.5683012199401856,6.901824726711475,883.4335650190689,1.9999999494757503e-05
28
+ 675,2764800000,1.5606089782714845,7.166626843748192,917.3282359997686,1.9999999494757503e-05
29
+ 700,2867200000,1.569625825881958,7.431464089033584,951.2274033962988,1.9999999494757503e-05
30
+ 725,2969600000,1.5637955999374389,7.697967706009089,985.3398663691634,1.9999999494757503e-05
31
+ 750,3072000000,1.5669568061828614,7.962892068477961,1019.250184765179,1.9999999494757503e-05
32
+ 775,3174400000,1.578919801712036,8.229167067304108,1053.3333846149258,1.9999999494757503e-05
33
+ 800,3276800000,1.5597226810455322,8.493772274765757,1087.2028511700169,1.9999999494757503e-05
34
+ 825,3379200000,1.5684496641159058,8.760366790807014,1121.3269492232978,1.9999999494757503e-05
35
+ 850,3481600000,1.555274577140808,9.025527741439927,1155.2675509043106,1.9999999494757503e-05
36
+ 875,3584000000,1.5589488697052003,9.290825104582503,1189.2256133865603,1.9999999494757503e-05
37
+ 900,3686400000,1.56228581905365,9.555587218982518,1223.1151640297624,1.9999999494757503e-05
38
+ 925,3788800000,1.5693172216415405,9.821702597393193,1257.1779324663287,1.9999999494757503e-05
39
+ 950,3891200000,1.547282567024231,10.086582040948787,1291.0825012414448,1.9999999494757503e-05
40
+ 975,3993600000,1.552180905342102,10.35216691473301,1325.0773650858252,1.9999999494757503e-05
41
+ 1000,4096000000,1.5544623231887817,10.617501541588314,1359.0401973233043,1.9999999494757503e-05
42
+ 1025,4198400000,1.5621129417419433,10.884025346612145,1393.1552443663545,1.9999999494757503e-05
43
+ 1050,4300800000,1.5600895547866822,11.148848565988725,1427.0526164465568,1.9999999494757503e-05
44
+ 1075,4403200000,1.5528885984420777,11.413467440123783,1460.9238323358443,1.9999999494757503e-05
45
+ 1100,4505600000,1.5599483346939087,11.678333667439569,1494.8267094322648,1.9999999494757503e-05
46
+ 1125,4608000000,1.5648639726638793,11.943580217023005,1528.7782677789446,1.9999999494757503e-05
47
+ 1150,4710400000,1.549267168045044,12.20822147437584,1562.6523487201075,1.9999999494757503e-05
48
+ 1175,4812800000,1.5537393379211426,12.473067252154218,1596.5526082757399,1.9999999494757503e-05
49
+ 1200,4915200000,1.5565906286239624,12.737777670184224,1630.4355417835807,1.9999999494757503e-05
50
+ 1220,4997120000,1.549389386177063,12.94956159459549,1657.5438841082228,1.9999999494757503e-05
metadata/training_logs/3_annealing.csv ADDED
@@ -0,0 +1,50 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ training_steps,training_tokens,training_loss,walltime,gputime,learning_rate
2
+ 25,102400000,1.4206320714950562,0.26831979325136307,34.34493353617447,2.938559919130057e-05
3
+ 50,204800000,1.3607354307174682,0.532874067107453,68.20788058975398,2.8771199140464887e-05
4
+ 75,307200000,1.344892144203186,0.797669098578563,102.10164461805607,2.8156800908618607e-05
5
+ 100,409600000,1.3339987087249756,1.0622922554949321,135.9734087033513,2.7542400857782923e-05
6
+ 125,512000000,1.3272197246551514,1.3275567747589183,169.92726716914154,2.692800080694724e-05
7
+ 150,614400000,1.3201901483535767,1.5918107887810053,203.75178096396868,2.6313600756111555e-05
8
+ 175,716800000,1.3260539388656616,1.856745663885765,237.6634449773779,2.569920070527587e-05
9
+ 200,819200000,1.3183754396438598,2.1223753731163395,271.66404775889146,2.5084800654440187e-05
10
+ 225,921600000,1.3132305669784545,2.3882311399600185,305.69358591488236,2.4470400603604503e-05
11
+ 250,1024000000,1.3129970407485962,2.65354331628931,339.65354448503166,2.385600055276882e-05
12
+ 275,1126400000,1.3064435338973999,2.9190579786122703,373.6394212623706,2.3241600501933135e-05
13
+ 300,1228800000,1.3097565698623657,3.1846308088092967,407.63274352759,2.262720045109745e-05
14
+ 325,1331200000,1.2990926265716554,3.450655307807062,441.6838793993039,2.2012800400261767e-05
15
+ 350,1433600000,1.3015284061431884,3.7156531358282527,475.60360138601635,2.1398400349426083e-05
16
+ 375,1536000000,1.300846767425537,3.9803074113379786,509.47934865126126,2.07840002985904e-05
17
+ 400,1638400000,1.2993078660964965,4.245298594229775,543.3982200614112,2.0169600247754715e-05
18
+ 425,1740800000,1.2959114503860474,4.510613332994379,577.3585066232805,1.955520019691903e-05
19
+ 450,1843200000,1.2931818628311158,4.775614008341895,611.2785930677626,1.8940800146083347e-05
20
+ 475,1945600000,1.2935280084609986,5.040500104048112,645.1840133181583,1.8326400095247664e-05
21
+ 500,2048000000,1.297581434249878,5.305721228775273,679.1323172832349,1.771200004441198e-05
22
+ 525,2150400000,1.2973516607284545,5.571934178461597,713.2075748430844,1.7097599993576296e-05
23
+ 550,2252800000,1.2920558738708496,5.836718603762894,747.0999812816505,1.6483199942740612e-05
24
+ 575,2355200000,1.2921710443496703,6.101621121729093,781.0075035813239,1.5868799891904928e-05
25
+ 600,2457600000,1.2922473382949828,6.3664986443806235,814.9118264807198,1.5254399841069244e-05
26
+ 625,2560000000,1.286066074371338,6.632426781060178,848.9506279757028,1.463999979023356e-05
27
+ 650,2662400000,1.2801355123519897,6.897544031576831,882.8856360418343,1.4025599739397876e-05
28
+ 675,2764800000,1.2844274616241456,7.16281749378081,916.8406392039436,1.3411199688562192e-05
29
+ 700,2867200000,1.2837993860244752,7.427888308341158,950.7697034676683,1.2796799637726508e-05
30
+ 725,2969600000,1.277207851409912,7.694053462864512,984.8388432466576,1.2182399586890824e-05
31
+ 750,3072000000,1.2725739479064941,7.9594902915458245,1018.8147573178655,1.1568000445549842e-05
32
+ 775,3174400000,1.279445676803589,8.224737335718995,1052.7663789720314,1.0953600394714158e-05
33
+ 800,3276800000,1.2785338878631591,8.4897940390995,1086.693637004736,1.0339200343878474e-05
34
+ 825,3379200000,1.2763902473449706,8.754925038348853,1120.6304049086532,9.72480029304279e-06
35
+ 850,3481600000,1.273976821899414,9.020435715932496,1154.6157716393595,9.110400242207106e-06
36
+ 875,3584000000,1.2713893747329712,9.285925992178,1188.598526998784,8.496000191371422e-06
37
+ 900,3686400000,1.2784578609466553,9.551624662507189,1222.6079568009202,7.881600140535738e-06
38
+ 925,3788800000,1.2703066444396973,9.81712425603072,1256.5919047719321,7.267200089700054e-06
39
+ 950,3891200000,1.271108751296997,10.082206830408738,1290.5224742923185,6.6528000388643704e-06
40
+ 975,3993600000,1.2680458974838258,10.347317189690123,1324.4566002803358,6.0383999880286865e-06
41
+ 1000,4096000000,1.2702019023895263,10.612797046162356,1358.4380219087816,5.4239999371930026e-06
42
+ 1025,4198400000,1.2751475191116333,10.87931673718002,1392.5525423590425,4.809599886357319e-06
43
+ 1050,4300800000,1.266355185508728,11.143856210205902,1426.4135949063555,4.195199835521635e-06
44
+ 1075,4403200000,1.2681057167053222,11.408356219250548,1460.2695960640701,3.580800012059626e-06
45
+ 1100,4505600000,1.271327452659607,11.672805533763803,1494.1191083217668,2.9663999612239422e-06
46
+ 1125,4608000000,1.2650820541381835,11.938276334099923,1528.0993707647901,2.3519999103882583e-06
47
+ 1150,4710400000,1.270536971092224,12.203680372205294,1562.0710876422777,1.737599973239412e-06
48
+ 1175,4812800000,1.2631586408615112,12.468635068036095,1595.9852887086201,1.1232000360905658e-06
49
+ 1200,4915200000,1.2647824430465697,12.734026105942206,1629.9553415606024,5.087999852548819e-07
50
+ 1220,4997120000,1.2646446466445922,12.946103368596539,1657.101231180357,1.727999965339677e-08