handlebars

assets/data/clustering/info.csv

@@ -1,106 +0,0 @@
-,cluster_id,cluster_summaries,cluster_position_x,cluster_position_y
-0,-1,None,9.926462,4.7121987
-1,0,Philosophical/Spiritual Introspection,10.312462,1.2666532
-2,1,"Scholarships,",8.167274,4.8995786
-3,2,Politics,8.81142,2.4859838
-4,3,Theology,9.615214,0.3783942
-5,4,Dating,4.985182,1.8439052
-6,5,Accommodation,11.457769,5.080919
-7,6,Football,6.6154537,-1.6859366
-8,7,Film Festival,6.9734483,1.4548192
-9,8,Culinary,13.426296,4.5412893
-10,9,Music,6.0653744,0.7536916
-11,10,Gambling,3.124241,3.2533677
-12,11,Baseball,7.133596,-2.4256644
-13,12,Technology,6.4929094,6.768577
-14,13,Website Policies,4.873843,5.771508
-15,14,Weddings,11.815845,3.7894728
-16,15,Gaming,5.529167,2.9530518
-17,16,Commodities/Services Provision,10.453564,5.8489122
-18,17,Crafts,13.287651,6.4237967
-19,18,Automobiles,9.9531145,8.840178
-20,19,Watches,13.893139,9.859185
-21,20,Dogs,12.595798,3.5351615
-22,21,Photography,10.7942295,3.5504062
-23,22,Legalities,8.942016,4.72733
-24,23,Consumer Electronics,7.078649,8.338984
-25,24,Insulation,10.520957,7.914946
-26,25,Cannabis,14.317424,3.2114828
-27,26,Footwear,15.052116,7.6956415
-28,27,Real Estate,9.536316,6.103533
-29,28,Relocation,10.205071,7.1883316
-30,29,Sports betting,3.2779586,2.443366
-31,30,Narratives,7.613535,1.8300554
-32,31,Dating,4.788838,2.1900373
-33,32,Apparel/Clothing,14.394226,7.3073387
-34,33,User Authentication,5.265638,6.4014487
-35,34,Academicwriting,6.9187264,3.4357684
-36,35,Sports,7.4969172,-2.086585
-37,36,Fashion/Lifestyle Products,13.821669,7.7150764
-38,37,Diverse events,9.437052,2.2438836
-39,38,Blockchain/Cryptocurrency,7.7586045,6.9439344
-40,39,Online Businesses/Marketing,6.522259,5.219268
-41,40,Healthcare,11.425277,2.3801014
-42,41,Home Decor,12.878046,7.2632184
-43,42,Biomedicine,12.789575,2.3376262
-44,43,Jewelry,14.259997,8.653363
-45,44,Addiction,11.561383,1.3774762
-46,45,Products,11.711758,8.423251
-47,46,Multi-purposefulness,11.080702,7.4574013
-48,47,"Mass transit,",9.910158,5.4402313
-49,48,Ethernet,6.9763823,7.7909245
-50,49,Legal,9.516912,4.636553
-51,50,E-commerce,13.263438,8.6548195
-52,51,Audio,7.717162,8.903019
-53,52,Infrastructure,10.52904,5.369669
-54,53,Firearms,11.062812,9.268473
-55,54,Freight/Logistics,9.551044,7.0336204
-56,55,Products,12.073747,7.645973
-57,56,Vaccinations,11.9387045,2.7824683
-58,57,Artwork,11.019163,4.1677165
-59,58,Viticulture,14.223523,5.0761614
-60,59,WordPress,5.9597983,5.824579
-61,60,Cosmetics/Dermatology,15.093273,3.4669027
-62,61,Software,6.375921,6.4298844
-63,62,Dentistry,14.76626,1.1620314
-64,63,Pest Control,13.201735,3.6806118
-65,64,SEO,5.720493,5.238112
-66,65,Lottery,1.7142816,2.9782674
-67,66,Narratives,8.460977,1.0804662
-68,67,Waste Reduction & Recycling,10.634534,6.959523
-69,68,Communication,6.438943,5.9467845
-70,69,Orthopedics,13.005415,1.1908791
-71,70,Home Decor & Furniture,12.732457,7.876862
-72,71,Education,7.6568975,3.4944353
-73,72,Sports,7.295141,-0.7343214
-74,73,Social Media Advertising,6.133886,4.8547883
-75,74,Privacy,4.756733,6.3598356
-76,75,Website design,6.1168823,5.465095
-77,76,Roofing,11.389448,8.080609
-78,77,Nutrition/Supplements,13.631578,2.5334294
-79,78,Haircare/Hairstyling,15.544645,4.54254
-80,79,Cookies,4.341592,6.819268
-81,80,International Trade,8.993828,6.4757586
-82,81,Entrepreneurial Resources,9.435777,5.3340797
-83,82,Cricket,6.5171986,-1.245905
-84,83,Crafts,13.852216,7.049825
-85,84,Floristry,13.407425,5.8741536
-86,85,Genealogy,9.530803,1.6548243
-87,86,Mental Health,11.074349,1.6069281
-88,87,Volunteerism,10.145443,3.6734574
-89,88,Lighting,11.385381,8.93693
-90,89,Artificial Intelligence,6.5306387,6.2178063
-91,90,Business,7.471462,6.4142885
-92,91,E-commerce,13.638669,6.5098934
-93,92,Urbanization/Over-tourism,10.221115,6.100654
-94,93,Events,10.8449,3.9822264
-95,94,Pharmaceuticals/Biotechnology,12.318266,2.4331784
-96,95,Professional Wrestling,6.856304,-0.65598303
-97,96,Various,9.3211975,3.4894605
-98,97,Medicine,13.17882,2.1281319
-99,98,Community Engagement,9.848856,3.5187004
-100,99,Fitness,12.504849,0.9134393
-101,100,Bathroom Design & Toilet Engineering,11.779076,7.2920136
-102,101,Business Development,7.328447,5.659843
-103,102,Sports,7.6370654,-1.0701839
-104,103,Sexuality,13.817207,1.6510898

assets/data/plots/all_dumps_bad/agg_score.json

@@ -1 +0,0 @@
-{"data": {"big-run-refinedweb": {"x": [0.0, 2.0971520000000003, 4.194304000000001, 6.291456, 8.388608000000001, 10.48576, 12.582912, 14.680064000000002, 16.777216000000003, 18.874368, 20.97152, 23.068672000000003, 25.165824, 27.262976000000002, 28.311552000000002, 29.360128000000003, 31.45728, 33.554432000000006, 35.651584, 37.748736, 39.845888, 41.94304, 44.040192000000005, 46.137344000000006, 48.234496, 50.331648, 52.4288, 54.525952000000004, 56.623104000000005, 58.720256000000006, 60.817408, 62.91456, 65.011712, 67.10886400000001, 69.206016, 71.303168, 73.40032000000001, 75.497472, 77.59462400000001, 79.691776, 81.788928, 83.88608, 85.983232, 88.08038400000001, 90.177536, 92.27468800000001, 94.37184, 96.468992, 98.56614400000001, 100.663296, 102.76044800000001, 104.8576, 106.95475200000001, 109.05190400000001, 111.149056, 113.24620800000001, 115.34336, 117.44051200000001, 119.537664, 121.634816, 123.73196800000001, 125.82912, 127.92627200000001, 130.023424, 132.120576, 134.21772800000002, 136.31488000000002, 138.412032, 140.509184, 142.606336, 144.70348800000002, 146.80064000000002, 148.897792, 150.994944, 153.092096, 155.18924800000002, 157.28640000000001, 159.383552, 161.480704, 163.577856, 165.67500800000002, 167.77216, 169.869312, 171.966464, 174.06361600000002, 176.16076800000002, 178.25792, 180.355072, 182.452224, 184.54937600000002, 186.64652800000002, 188.74368, 190.840832, 192.937984, 195.03513600000002, 197.13228800000002, 199.22944, 201.326592, 203.423744, 205.52089600000002, 207.61804800000002, 209.7152, 211.812352, 213.90950400000003, 216.00665600000002, 218.10380800000001, 220.20096, 222.298112, 224.39526400000003, 226.49241600000002, 228.589568, 230.68672, 232.783872, 234.88102400000002, 236.97817600000002, 239.075328, 241.17248, 243.269632, 245.36678400000002, 247.46393600000002, 249.561088, 251.65824, 253.75539200000003, 255.85254400000002, 257.949696, 260.046848, 262.144, 264.241152, 266.338304, 268.43545600000004, 270.53260800000004, 272.62976000000003, 274.726912, 276.824064, 278.921216, 281.018368, 283.11552, 285.212672, 287.309824, 289.40697600000004, 291.50412800000004, 293.60128000000003, 295.698432, 297.795584, 299.892736, 301.989888, 304.08704, 306.184192, 308.28134400000005, 310.37849600000004, 312.47564800000004, 314.57280000000003, 316.669952, 318.767104, 320.864256, 322.961408, 325.05856, 327.155712, 329.25286400000005, 331.35001600000004, 333.44716800000003, 335.54432, 337.641472, 339.738624, 341.835776, 343.932928, 346.03008, 348.12723200000005, 350.22438400000004], "y": [0.3308933284133672, 0.3534814938902855, 0.3764607086777687, 0.38782499730587, 0.3981050960719585, 0.4028486795723438, 0.4125883243978023, 0.4117814563214779, 0.414029736071825, 0.4197172522544861, 0.4211113378405571, 0.4279881417751312, 0.4280137903988361, 0.4280424378812313, 0.4291964024305343, 0.4326301179826259, 0.4371833503246307, 0.4346669465303421, 0.4336562640964985, 0.4432648755609989, 0.4401291646063328, 0.4394684173166752, 0.4476612061262131, 0.4465444348752498, 0.4472153298556804, 0.4433343075215816, 0.4510187618434429, 0.4459567815065384, 0.4460812956094742, 0.4498684890568256, 0.4529943652451038, 0.4528274349868297, 0.4551213420927524, 0.4549156539142132, 0.4564928151667118, 0.4576693661510944, 0.4557182416319847, 0.4536240361630916, 0.457439012825489, 0.4570476822555065, 0.4589823484420776, 0.462024375796318, 0.4540738053619861, 0.4550252184271812, 0.4576593860983848, 0.4573238864541054, 0.4575810581445694, 0.4622134491801262, 0.4592566937208175, 0.4614734016358852, 0.4637473002076149, 0.4625372551381588, 0.4613912180066108, 0.4597448222339153, 0.4594792164862156, 0.4662549719214439, 0.4634026065468788, 0.4633508697152138, 0.4635734222829342, 0.4628961533308029, 0.4670135043561458, 0.4639505892992019, 0.4631133340299129, 0.4665167145431041, 0.4672448337078094, 0.4693268723785877, 0.4630668573081493, 0.4676454700529575, 0.4646359197795391, 0.4621579721570015, 0.4692446552217006, 0.4704835228621959, 0.4663223996758461, 0.4680556617677212, 0.466339822858572, 0.4682099223136902, 0.4711195565760135, 0.4722655527293682, 0.4727961830794811, 0.4676857478916645, 0.4719390422105789, 0.4713102728128433, 0.4712141714990139, 0.4721613004803657, 0.4713456854224205, 0.4682970903813839, 0.4679934531450271, 0.4685162976384163, 0.4679946713149547, 0.4681242071092129, 0.4702276065945625, 0.472664151340723, 0.4730790853500366, 0.4731674715876579, 0.4718914777040481, 0.4719801284372806, 0.4761029370129108, 0.4735167175531387, 0.4730370938777923, 0.4730173237621784, 0.4735377207398414, 0.4777223989367485, 0.4796326830983162, 0.4734170883893966, 0.4739485755562782, 0.4748299159109592, 0.4765299335122108, 0.4745025858283043, 0.4754423759877682, 0.4784592799842357, 0.4761341325938701, 0.4760282784700393, 0.4769757278263569, 0.47154351323843, 0.4786738082766533, 0.4804279990494251, 0.4777076803147793, 0.4798569902777672, 0.4759011939167976, 0.4784621745347976, 0.479673832654953, 0.4780617095530033, 0.48076206818223, 0.47995800152421, 0.4790860973298549, 0.4817167408764362, 0.4811586998403072, 0.482547752559185, 0.4816697351634502, 0.4809327870607376, 0.4816545359790325, 0.4804601892828941, 0.4776877984404564, 0.4813711903989315, 0.4844604581594467, 0.4819537848234176, 0.4820829331874847, 0.4778126627206802, 0.482935007661581, 0.48230691999197, 0.4826001971960068, 0.4823969900608063, 0.4811219945549965, 0.4789146520197391, 0.484035175293684, 0.4848698377609253, 0.4855728335678577, 0.4825376532971859, 0.485215101391077, 0.4824351668357849, 0.4835342466831207, 0.4822137206792831, 0.4838785007596016, 0.4837255179882049, 0.4853012599050998, 0.4857851006090641, 0.4863366298377514, 0.4856646582484245, 0.4842503517866134, 0.4838776960968971, 0.4846346862614155, 0.4837041422724724, 0.4813097268342972, 0.4873070046305656, 0.4841253720223903, 0.4837464913725853, 0.483069509267807, 0.4851242564618587, 0.4861010462045669], "label": "RefinedWeb"}, "big-run-sampled_full_filtered_no_dedup": {"x": [0.0, 2.0971520000000003, 4.194304000000001, 6.291456, 8.388608000000001, 10.48576, 12.582912, 14.680064000000002, 16.777216000000003, 18.874368, 20.97152, 23.068672000000003, 25.165824, 27.262976000000002, 28.311552000000002, 29.360128000000003, 31.45728, 33.554432000000006, 35.651584, 37.748736, 39.845888, 41.94304, 44.040192000000005, 46.137344000000006, 48.234496, 50.331648, 52.4288, 54.525952000000004, 56.623104000000005, 58.720256000000006, 60.817408, 62.91456, 65.011712, 67.10886400000001, 69.206016, 71.303168, 73.40032000000001, 75.497472, 77.59462400000001, 79.691776, 81.788928, 83.88608, 85.983232, 88.08038400000001, 90.177536, 92.27468800000001, 94.37184, 96.468992, 98.56614400000001, 100.663296, 102.76044800000001, 104.8576, 106.95475200000001, 109.05190400000001, 111.149056, 113.24620800000001, 115.34336, 117.44051200000001, 119.537664, 121.634816, 123.73196800000001, 125.82912, 127.92627200000001, 130.023424, 132.120576, 134.21772800000002, 136.31488000000002, 138.412032, 140.509184, 142.606336, 144.70348800000002, 146.80064000000002, 148.897792, 150.994944, 153.092096, 155.18924800000002, 157.28640000000001, 159.383552, 161.480704, 163.577856, 165.67500800000002, 167.77216, 169.869312, 171.966464, 174.06361600000002, 176.16076800000002, 178.25792, 180.355072, 182.452224, 184.54937600000002, 186.64652800000002, 188.74368, 190.840832, 192.937984, 195.03513600000002, 197.13228800000002, 199.22944, 201.326592, 203.423744, 205.52089600000002, 207.61804800000002, 209.7152, 211.812352, 213.90950400000003, 216.00665600000002, 218.10380800000001, 220.20096, 222.298112, 224.39526400000003, 226.49241600000002, 228.589568, 230.68672, 232.783872, 234.88102400000002, 236.97817600000002, 239.075328, 241.17248, 243.269632, 245.36678400000002, 247.46393600000002, 249.561088, 251.65824, 253.75539200000003, 255.85254400000002, 257.949696, 260.046848, 262.144, 264.241152, 266.338304, 268.43545600000004, 270.53260800000004, 272.62976000000003, 274.726912, 276.824064, 278.921216, 281.018368, 283.11552, 285.212672, 287.309824, 289.40697600000004, 291.50412800000004, 293.60128000000003, 295.698432, 297.795584, 299.892736, 301.989888, 304.08704, 306.184192, 308.28134400000005, 310.37849600000004, 312.47564800000004, 314.57280000000003, 316.669952, 318.767104, 320.864256, 322.961408, 325.05856, 327.155712, 329.25286400000005, 331.35001600000004, 333.44716800000003, 335.54432, 337.641472, 339.738624, 341.835776, 343.932928, 346.03008, 348.12723200000005, 350.22438400000004], "y": [0.3308933284133672, 0.3605199865996837, 0.3733148723840713, 0.3882005847990513, 0.3934122696518898, 0.3947227671742439, 0.4042885974049568, 0.3974800482392311, 0.4055779427289963, 0.4133470430970192, 0.4117913842201233, 0.4113653488457203, 0.4149517640471458, 0.4187851920723915, 0.4252083078026771, 0.4206527359783649, 0.4240428246557712, 0.422003373503685, 0.4280910938978195, 0.4244147576391697, 0.4316282644867897, 0.4295645765960216, 0.4310102686285972, 0.4360743537545204, 0.4313482865691185, 0.4350991360843181, 0.4378576353192329, 0.4335876516997814, 0.4347924515604973, 0.4348904751241207, 0.436600212007761, 0.430036511272192, 0.4350974671542644, 0.4399556629359722, 0.4371416717767715, 0.4363861419260502, 0.4376698136329651, 0.4405004419386387, 0.4373639523983001, 0.4379038028419018, 0.4371281825006008, 0.4393439553678036, 0.440426729619503, 0.4401675276458263, 0.4429537951946258, 0.4449137263000011, 0.4434786736965179, 0.4450470842421055, 0.4454202279448509, 0.4394537284970283, 0.442185215651989, 0.4461225643754005, 0.4427758157253265, 0.4430646039545536, 0.4476901069283485, 0.4478763341903686, 0.4493869319558143, 0.4448477327823639, 0.450044184923172, 0.4498609118163585, 0.4457665979862213, 0.4506924152374267, 0.449855338782072, 0.448790930211544, 0.4474099352955818, 0.4546772800385952, 0.4529431238770485, 0.452015146613121, 0.4502020999789238, 0.4493804536759853, 0.4523266032338142, 0.4551868587732315, 0.4501944817602634, 0.4493303671479225, 0.4526805207133293, 0.4533850513398647, 0.4518048763275146, 0.4518973492085933, 0.4531301632523536, 0.4518006071448326, 0.4553494565188885, 0.4528752230107784, 0.4536322727799415, 0.4561733976006508, 0.4549491256475448, 0.4574789106845855, 0.4577847123146057, 0.4563642293214798, 0.4578686729073524, 0.4561499990522861, 0.4537816494703293, 0.4542164430022239, 0.4559455662965774, 0.4554723873734474, 0.4575514122843742, 0.4575202167034149, 0.4592722058296203, 0.4585275091230869, 0.4580587856471538, 0.456934317946434, 0.4577495418488979, 0.4540119916200638, 0.4570806957781315, 0.4608120545744896, 0.4588425755500793, 0.4578334167599678, 0.4610816091299057, 0.4598177038133144, 0.461849745362997, 0.4631866924464702, 0.4601576402783394, 0.4646804705262184, 0.4632389545440674, 0.4604574106633663, 0.4602976888418197, 0.4581312239170074, 0.4654182009398937, 0.4655338563024997, 0.4616620391607284, 0.461054053157568, 0.4613021649420261, 0.4658613465726375, 0.4633531905710697, 0.4613638147711754, 0.4643996246159076, 0.462500050663948, 0.4650798961520195, 0.4648764543235302, 0.4639869071543216, 0.4634246975183487, 0.46585888043046, 0.4639799632132053, 0.4630857892334461, 0.4644265696406364, 0.4642998576164245, 0.4686848931014538, 0.4687492996454239, 0.4650243632495403, 0.4627032242715359, 0.4665953740477562, 0.4660026729106903, 0.4664581045508384, 0.4676475040614605, 0.4657339677214622, 0.4664678275585174, 0.4673498086631298, 0.4676674827933311, 0.4680955372750759, 0.4681585058569908, 0.4659864418208599, 0.4686457589268684, 0.4661462865769863, 0.4658931568264961, 0.4674226939678192, 0.46805215254426, 0.4682257212698459, 0.4689070098102093, 0.4699570722877979, 0.4655096270143986, 0.4688013233244419, 0.4707522802054882, 0.4661469310522079, 0.4688841328024864, 0.4671329781413078, 0.4662554152309894, 0.4697433896362781, 0.4698473587632179, 0.4676505327224731, 0.4696521013975143], "label": "FineWeb filtered only"}, "big-run-fineweb-cross-dedup-fixed": {"x": [0.0, 2.0971520000000003, 4.194304000000001, 6.291456, 8.388608000000001, 10.48576, 12.582912, 14.680064000000002, 16.777216000000003, 18.874368, 20.97152, 23.068672000000003, 25.165824, 27.262976000000002, 29.360128000000003, 31.45728, 33.554432000000006, 35.651584, 37.748736, 39.845888, 41.94304, 44.040192000000005, 46.137344000000006, 48.234496, 50.331648, 52.4288, 54.525952000000004, 56.623104000000005, 58.720256000000006, 60.817408, 62.91456, 65.011712, 67.10886400000001, 69.206016, 71.303168, 73.40032000000001, 75.497472, 77.59462400000001, 79.691776, 81.788928, 83.88608, 85.983232, 88.08038400000001, 90.177536, 92.27468800000001, 94.37184, 96.468992, 98.56614400000001, 100.663296, 102.76044800000001, 104.8576, 106.95475200000001, 109.05190400000001, 111.149056, 113.24620800000001, 115.34336, 117.44051200000001, 119.537664, 121.634816, 123.73196800000001, 125.82912, 127.92627200000001, 130.023424, 132.120576, 134.21772800000002, 136.31488000000002, 138.412032, 140.509184, 142.606336, 144.70348800000002, 146.80064000000002, 148.897792, 150.994944, 153.092096, 155.18924800000002, 157.28640000000001, 159.383552, 161.480704, 163.577856, 165.67500800000002, 167.77216, 169.869312, 171.966464, 174.06361600000002, 176.16076800000002, 178.25792, 180.355072, 182.452224, 184.54937600000002, 186.64652800000002, 188.74368, 190.840832, 192.937984, 195.03513600000002, 197.13228800000002, 199.22944, 201.326592, 203.423744, 205.52089600000002, 207.61804800000002, 209.7152, 211.812352, 213.90950400000003, 216.00665600000002, 218.10380800000001, 220.20096, 222.298112, 224.39526400000003, 226.49241600000002, 228.589568, 230.68672, 232.783872, 234.88102400000002, 236.97817600000002, 239.075328, 241.17248, 243.269632, 245.36678400000002, 247.46393600000002, 249.561088, 251.65824, 253.75539200000003, 255.85254400000002, 257.949696, 260.046848, 262.144, 264.241152, 266.338304, 268.43545600000004, 270.53260800000004, 272.62976000000003, 274.726912, 276.824064, 278.921216, 281.018368, 283.11552, 285.212672, 287.309824, 289.40697600000004, 291.50412800000004, 293.60128000000003, 295.698432, 297.795584, 299.892736, 301.989888, 304.08704, 306.184192, 308.28134400000005, 310.37849600000004, 312.47564800000004, 314.57280000000003, 316.669952, 318.767104, 320.864256, 322.961408, 325.05856, 327.155712, 329.25286400000005, 331.35001600000004, 333.44716800000003, 335.54432, 337.641472, 339.738624, 341.835776, 343.932928, 346.03008, 348.12723200000005, 350.22438400000004], "y": [0.3308933284133672, 0.3551952373236418, 0.3736435137689113, 0.3814037963747978, 0.3948809280991554, 0.3996850810945034, 0.4089604057371616, 0.4100853353738785, 0.4119834117591381, 0.4168377220630646, 0.4186493046581745, 0.4169826358556747, 0.4234288297593593, 0.4229162000119686, 0.4273439794778824, 0.4290364980697632, 0.4291782416403293, 0.4296907968819141, 0.4311576783657074, 0.4326641112565994, 0.430318683385849, 0.430436260998249, 0.4339037239551544, 0.4363459683954716, 0.4357402548193931, 0.4342963136732578, 0.4366712383925915, 0.4363959729671478, 0.436981026083231, 0.4447868093848228, 0.4411709941923618, 0.4406092017889023, 0.4424176625907421, 0.4423875361680984, 0.4422253370285034, 0.4410557933151722, 0.4447037056088447, 0.4454837813973427, 0.4435960277915001, 0.4468514993786812, 0.4479999616742134, 0.4428562931716442, 0.445764634758234, 0.4456562362611294, 0.4488007053732872, 0.4475954286754131, 0.4468922987580299, 0.4548408314585686, 0.4511027485132217, 0.4530330970883369, 0.4483681954443455, 0.4531726539134979, 0.45334542542696, 0.4544384703040123, 0.4530758671462536, 0.4540613554418087, 0.4510113634169101, 0.4538320265710354, 0.4518541917204857, 0.4536847211420536, 0.4532708041369915, 0.4552236869931221, 0.455034039914608, 0.4562875479459762, 0.4532428197562694, 0.4574853852391243, 0.4517738744616508, 0.4579889141023159, 0.4538268558681011, 0.456730306148529, 0.4526018649339676, 0.4562746733427048, 0.4560015797615051, 0.4555426277220249, 0.4561501257121563, 0.4524396173655987, 0.4557023830711841, 0.4589769169688225, 0.4581078588962555, 0.4620813727378845, 0.4586601965129375, 0.4568093195557594, 0.4569808952510357, 0.4567535072565079, 0.4575250148773193, 0.4606908001005649, 0.4603964723646641, 0.4622848592698574, 0.4594669193029403, 0.4640629850327968, 0.4604269936680794, 0.4634841009974479, 0.4644578285515308, 0.4642514958977699, 0.4666304066777229, 0.4616626128554344, 0.4588956907391548, 0.4620226770639419, 0.4628621749579906, 0.4595407098531723, 0.4635516740381717, 0.46005355194211, 0.4601523540914058, 0.4644204638898372, 0.4620639197528362, 0.46614545956254, 0.4636696502566337, 0.4610077403485775, 0.4640897810459137, 0.4636163525283336, 0.4630545899271965, 0.466012816876173, 0.4650349207222461, 0.4613720141351223, 0.4644323363900184, 0.4647249802947044, 0.4656480401754379, 0.4651664271950722, 0.4622530452907085, 0.4655019529163837, 0.4650313258171081, 0.466718140989542, 0.4661559611558914, 0.4661237150430679, 0.4664223715662956, 0.4640601389110088, 0.4642657749354839, 0.4633881188929081, 0.4629989042878151, 0.4685831367969513, 0.4675870984792709, 0.467183344066143, 0.4678030684590339, 0.4660939238965511, 0.4691914953291416, 0.4670972637832165, 0.468262892216444, 0.4672016054391861, 0.4676182121038437, 0.4698677137494087, 0.4658828042447567, 0.4701816700398922, 0.4684622809290886, 0.466015312820673, 0.4675401039421558, 0.4693200923502445, 0.4702670983970165, 0.4679145030677318, 0.4676233418285846, 0.4674933589994907, 0.4678357951343059, 0.4669915996491909, 0.4657857678830623, 0.4666901864111423, 0.4669371582567692, 0.4672787226736545, 0.4684535376727581, 0.4685697965323925, 0.4694835692644119, 0.4683254994451999, 0.4712230190634727, 0.4683987610042095, 0.4707653746008873, 0.4663059376180172, 0.4683133698999882, 0.4686385430395603, 0.4657671600580215, 0.4692615270614624], "label": "FineWeb full MinHash"}}, "layout": {"title": {"text": "Dedup across all dumps does not improve performance"}}}

dist/assets/data/clustering/info.csv

@@ -1,106 +0,0 @@
-,cluster_id,cluster_summaries,cluster_position_x,cluster_position_y
-0,-1,None,9.926462,4.7121987
-1,0,Philosophical/Spiritual Introspection,10.312462,1.2666532
-2,1,"Scholarships,",8.167274,4.8995786
-3,2,Politics,8.81142,2.4859838
-4,3,Theology,9.615214,0.3783942
-5,4,Dating,4.985182,1.8439052
-6,5,Accommodation,11.457769,5.080919
-7,6,Football,6.6154537,-1.6859366
-8,7,Film Festival,6.9734483,1.4548192
-9,8,Culinary,13.426296,4.5412893
-10,9,Music,6.0653744,0.7536916
-11,10,Gambling,3.124241,3.2533677
-12,11,Baseball,7.133596,-2.4256644
-13,12,Technology,6.4929094,6.768577
-14,13,Website Policies,4.873843,5.771508
-15,14,Weddings,11.815845,3.7894728
-16,15,Gaming,5.529167,2.9530518
-17,16,Commodities/Services Provision,10.453564,5.8489122
-18,17,Crafts,13.287651,6.4237967
-19,18,Automobiles,9.9531145,8.840178
-20,19,Watches,13.893139,9.859185
-21,20,Dogs,12.595798,3.5351615
-22,21,Photography,10.7942295,3.5504062
-23,22,Legalities,8.942016,4.72733
-24,23,Consumer Electronics,7.078649,8.338984
-25,24,Insulation,10.520957,7.914946
-26,25,Cannabis,14.317424,3.2114828
-27,26,Footwear,15.052116,7.6956415
-28,27,Real Estate,9.536316,6.103533
-29,28,Relocation,10.205071,7.1883316
-30,29,Sports betting,3.2779586,2.443366
-31,30,Narratives,7.613535,1.8300554
-32,31,Dating,4.788838,2.1900373
-33,32,Apparel/Clothing,14.394226,7.3073387
-34,33,User Authentication,5.265638,6.4014487
-35,34,Academicwriting,6.9187264,3.4357684
-36,35,Sports,7.4969172,-2.086585
-37,36,Fashion/Lifestyle Products,13.821669,7.7150764
-38,37,Diverse events,9.437052,2.2438836
-39,38,Blockchain/Cryptocurrency,7.7586045,6.9439344
-40,39,Online Businesses/Marketing,6.522259,5.219268
-41,40,Healthcare,11.425277,2.3801014
-42,41,Home Decor,12.878046,7.2632184
-43,42,Biomedicine,12.789575,2.3376262
-44,43,Jewelry,14.259997,8.653363
-45,44,Addiction,11.561383,1.3774762
-46,45,Products,11.711758,8.423251
-47,46,Multi-purposefulness,11.080702,7.4574013
-48,47,"Mass transit,",9.910158,5.4402313
-49,48,Ethernet,6.9763823,7.7909245
-50,49,Legal,9.516912,4.636553
-51,50,E-commerce,13.263438,8.6548195
-52,51,Audio,7.717162,8.903019
-53,52,Infrastructure,10.52904,5.369669
-54,53,Firearms,11.062812,9.268473
-55,54,Freight/Logistics,9.551044,7.0336204
-56,55,Products,12.073747,7.645973
-57,56,Vaccinations,11.9387045,2.7824683
-58,57,Artwork,11.019163,4.1677165
-59,58,Viticulture,14.223523,5.0761614
-60,59,WordPress,5.9597983,5.824579
-61,60,Cosmetics/Dermatology,15.093273,3.4669027
-62,61,Software,6.375921,6.4298844
-63,62,Dentistry,14.76626,1.1620314
-64,63,Pest Control,13.201735,3.6806118
-65,64,SEO,5.720493,5.238112
-66,65,Lottery,1.7142816,2.9782674
-67,66,Narratives,8.460977,1.0804662
-68,67,Waste Reduction & Recycling,10.634534,6.959523
-69,68,Communication,6.438943,5.9467845
-70,69,Orthopedics,13.005415,1.1908791
-71,70,Home Decor & Furniture,12.732457,7.876862
-72,71,Education,7.6568975,3.4944353
-73,72,Sports,7.295141,-0.7343214
-74,73,Social Media Advertising,6.133886,4.8547883
-75,74,Privacy,4.756733,6.3598356
-76,75,Website design,6.1168823,5.465095
-77,76,Roofing,11.389448,8.080609
-78,77,Nutrition/Supplements,13.631578,2.5334294
-79,78,Haircare/Hairstyling,15.544645,4.54254
-80,79,Cookies,4.341592,6.819268
-81,80,International Trade,8.993828,6.4757586
-82,81,Entrepreneurial Resources,9.435777,5.3340797
-83,82,Cricket,6.5171986,-1.245905
-84,83,Crafts,13.852216,7.049825
-85,84,Floristry,13.407425,5.8741536
-86,85,Genealogy,9.530803,1.6548243
-87,86,Mental Health,11.074349,1.6069281
-88,87,Volunteerism,10.145443,3.6734574
-89,88,Lighting,11.385381,8.93693
-90,89,Artificial Intelligence,6.5306387,6.2178063
-91,90,Business,7.471462,6.4142885
-92,91,E-commerce,13.638669,6.5098934
-93,92,Urbanization/Over-tourism,10.221115,6.100654
-94,93,Events,10.8449,3.9822264
-95,94,Pharmaceuticals/Biotechnology,12.318266,2.4331784
-96,95,Professional Wrestling,6.856304,-0.65598303
-97,96,Various,9.3211975,3.4894605
-98,97,Medicine,13.17882,2.1281319
-99,98,Community Engagement,9.848856,3.5187004
-100,99,Fitness,12.504849,0.9134393
-101,100,Bathroom Design & Toilet Engineering,11.779076,7.2920136
-102,101,Business Development,7.328447,5.659843
-103,102,Sports,7.6370654,-1.0701839
-104,103,Sexuality,13.817207,1.6510898

dist/assets/data/fragments/memory-recomputation.html

@@ -1 +0,0 @@
-<div>                            <div id="3a29a7b6-5a6e-4e51-aabe-515dafefbd63" class="plotly-graph-div" style="height:500px; width:850px;"></div>            <script type="text/javascript">                window.PLOTLYENV=window.PLOTLYENV || {};                                if (document.getElementById("3a29a7b6-5a6e-4e51-aabe-515dafefbd63")) {                    Plotly.newPlot(                        "3a29a7b6-5a6e-4e51-aabe-515dafefbd63",                        [{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(78, 165, 183)"},"name":"parameters","showlegend":true,"visible":true,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAEjnD0AAAAAASPcPQAAAAACkCxBAAAAAAKQrEEA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(227, 138, 66)"},"name":"gradients","showlegend":true,"visible":true,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAEjnD0AAAAAASPcPQAAAAACkCxBAAAAAAKQrEEA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(232, 137, 171)"},"name":"optimizer","showlegend":true,"visible":true,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAEjnH0AAAAAASPcfQAAAAACkCyBAAAAAAKQrIEA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(206, 192, 250)"},"name":"activations","showlegend":true,"visible":true,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAACEDEAAAAAAAEIoQAAAAAAAIUZAAAAAAIAQZUA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(78, 165, 183)"},"name":"parameters","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAKWXKkAAAAAApZ0qQAAAAAClqSpAAAAAAKXBKkA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(227, 138, 66)"},"name":"gradients","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAKWXKkAAAAAApZ0qQAAAAAClqSpAAAAAAKXBKkA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(232, 137, 171)"},"name":"optimizer","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAKWXOkAAAAAApZ06QAAAAAClqTpAAAAAAKXBOkA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(206, 192, 250)"},"name":"activations","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAABLGEAAAAAAgLUyQAAAAACA1U9AAAAAAMAKbUA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(78, 165, 183)"},"name":"parameters","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAKL6OUAAAAAAov45QAAAAACiBjpAAAAAAKIWOkA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(227, 138, 66)"},"name":"gradients","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAKL6OUAAAAAAov45QAAAAACiBjpAAAAAAKIWOkA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(232, 137, 171)"},"name":"optimizer","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAKL6SUAAAAAAov5JQAAAAACiBkpAAAAAAKIWSkA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(206, 192, 250)"},"name":"activations","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAACCIkAAAAAAAII8QAAAAAAAQVhAAAAAAIAgdkA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(78, 165, 183)"},"name":"parameters","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAgER\u002fbkAAAACARIBuQAAAAIBEgm5AAAAAgESGbkA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(227, 138, 66)"},"name":"gradients","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAgER\u002fbkAAAACARIBuQAAAAIBEgm5AAAAAgESGbkA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(232, 137, 171)"},"name":"optimizer","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAgER\u002ffkAAAACARIB+QAAAAIBEgn5AAAAAgESGfkA="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(206, 192, 250)"},"name":"activations","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAAAhR0AAAAAAgNBhQAAAAACAUH5AAAAAAECom0A="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(78, 165, 183)"},"name":"parameters","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAgPa\u002fl0AAAACANsCXQAAAAIC2wJdAAAAAgLbBl0A="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(227, 138, 66)"},"name":"gradients","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAgPa\u002fl0AAAACANsCXQAAAAIC2wJdAAAAAgLbBl0A="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(232, 137, 171)"},"name":"optimizer","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAgPa\u002fp0AAAACANsCnQAAAAIC2wKdAAAAAgLbBp0A="},"type":"bar"},{"hovertemplate":"Seq len=%{x}\u003cbr\u003eMem=%{y:.1f}GB\u003cbr\u003e%{data.name}\u003cextra\u003e\u003c\u002fextra\u003e","marker":{"color":"rgb(206, 192, 250)"},"name":"activations","showlegend":true,"visible":false,"x":["1024","2048","4096","8192"],"y":{"dtype":"f8","bdata":"AAAAAIA2YkAAAAAAgA58QAAAAABA35dAAAAAAKDHtUA="},"type":"bar"}],                        {"template":{"data":{"barpolar":[{"marker":{"line":{"color":"white","width":0.5},"pattern":{"fillmode":"overlay","size":10,"solidity":0.2}},"type":"barpolar"}],"bar":[{"error_x":{"color":"#2a3f5f"},"error_y":{"color":"#2a3f5f"},"marker":{"line":{"color":"white","width":0.5},"pattern":{"fillmode":"overlay","size":10,"solidity":0.2}},"type":"bar"}],"carpet":[{"aaxis":{"endlinecolor":"#2a3f5f","gridcolor":"#C8D4E3","linecolor":"#C8D4E3","minorgridcolor":"#C8D4E3","startlinecolor":"#2a3f5f"},"baxis":{"endlinecolor":"#2a3f5f","gridcolor":"#C8D4E3","linecolor":"#C8D4E3","minorgridcolor":"#C8D4E3","startlinecolor":"#2a3f5f"},"type":"carpet"}],"choropleth":[{"colorbar":{"outlinewidth":0,"ticks":""},"type":"choropleth"}],"contourcarpet":[{"colorbar":{"outlinewidth":0,"ticks":""},"type":"contourcarpet"}],"contour":[{"colorbar":{"outlinewidth":0,"ticks":""},"colorscale":[[0.0,"#0d0887"],[0.1111111111111111,"#46039f"],[0.2222222222222222,"#7201a8"],[0.3333333333333333,"#9c179e"],[0.4444444444444444,"#bd3786"],[0.5555555555555556,"#d8576b"],[0.6666666666666666,"#ed7953"],[0.7777777777777778,"#fb9f3a"],[0.8888888888888888,"#fdca26"],[1.0,"#f0f921"]],"type":"contour"}],"heatmap":[{"colorbar":{"outlinewidth":0,"ticks":""},"colorscale":[[0.0,"#0d0887"],[0.1111111111111111,"#46039f"],[0.2222222222222222,"#7201a8"],[0.3333333333333333,"#9c179e"],[0.4444444444444444,"#bd3786"],[0.5555555555555556,"#d8576b"],[0.6666666666666666,"#ed7953"],[0.7777777777777778,"#fb9f3a"],[0.8888888888888888,"#fdca26"],[1.0,"#f0f921"]],"type":"heatmap"}],"histogram2dcontour":[{"colorbar":{"outlinewidth":0,"ticks":""},"colorscale":[[0.0,"#0d0887"],[0.1111111111111111,"#46039f"],[0.2222222222222222,"#7201a8"],[0.3333333333333333,"#9c179e"],[0.4444444444444444,"#bd3786"],[0.5555555555555556,"#d8576b"],[0.6666666666666666,"#ed7953"],[0.7777777777777778,"#fb9f3a"],[0.8888888888888888,"#fdca26"],[1.0,"#f0f921"]],"type":"histogram2dcontour"}],"histogram2d":[{"colorbar":{"outlinewidth":0,"ticks":""},"colorscale":[[0.0,"#0d0887"],[0.1111111111111111,"#46039f"],[0.2222222222222222,"#7201a8"],[0.3333333333333333,"#9c179e"],[0.4444444444444444,"#bd3786"],[0.5555555555555556,"#d8576b"],[0.6666666666666666,"#ed7953"],[0.7777777777777778,"#fb9f3a"],[0.8888888888888888,"#fdca26"],[1.0,"#f0f921"]],"type":"histogram2d"}],"histogram":[{"marker":{"pattern":{"fillmode":"overlay","size":10,"solidity":0.2}},"type":"histogram"}],"mesh3d":[{"colorbar":{"outlinewidth":0,"ticks":""},"type":"mesh3d"}],"parcoords":[{"line":{"colorbar":{"outlinewidth":0,"ticks":""}},"type":"parcoords"}],"pie":[{"automargin":true,"type":"pie"}],"scatter3d":[{"line":{"colorbar":{"outlinewidth":0,"ticks":""}},"marker":{"colorbar":{"outlinewidth":0,"ticks":""}},"type":"scatter3d"}],"scattercarpet":[{"marker":{"colorbar":{"outlinewidth":0,"ticks":""}},"type":"scattercarpet"}],"scattergeo":[{"marker":{"colorbar":{"outlinewidth":0,"ticks":""}},"type":"scattergeo"}],"scattergl":[{"marker":{"colorbar":{"outlinewidth":0,"ticks":""}},"type":"scattergl"}],"scattermapbox":[{"marker":{"colorbar":{"outlinewidth":0,"ticks":""}},"type":"scattermapbox"}],"scattermap":[{"marker":{"colorbar":{"outlinewidth":0,"ticks":""}},"type":"scattermap"}],"scatterpolargl":[{"marker":{"colorbar":{"outlinewidth":0,"ticks":""}},"type":"scatterpolargl"}],"scatterpolar":[{"marker":{"colorbar":{"outlinewidth":0,"ticks":""}},"type":"scatterpolar"}],"scatter":[{"fillpattern":{"fillmode":"overlay","size":10,"solidity":0.2},"type":"scatter"}],"scatterternary":[{"marker":{"colorbar":{"outlinewidth":0,"ticks":""}},"type":"scatterternary"}],"surface":[{"colorbar":{"outlinewidth":0,"ticks":""},"colorscale":[[0.0,"#0d0887"],[0.1111111111111111,"#46039f"],[0.2222222222222222,"#7201a8"],[0.3333333333333333,"#9c179e"],[0.4444444444444444,"#bd3786"],[0.5555555555555556,"#d8576b"],[0.6666666666666666,"#ed7953"],[0.7777777777777778,"#fb9f3a"],[0.8888888888888888,"#fdca26"],[1.0,"#f0f921"]],"type":"surface"}],"table":[{"cells":{"fill":{"color":"#EBF0F8"},"line":{"color":"white"}},"header":{"fill":{"color":"#C8D4E3"},"line":{"color":"white"}},"type":"table"}]},"layout":{"annotationdefaults":{"arrowcolor":"#2a3f5f","arrowhead":0,"arrowwidth":1},"autotypenumbers":"strict","coloraxis":{"colorbar":{"outlinewidth":0,"ticks":""}},"colorscale":{"diverging":[[0,"#8e0152"],[0.1,"#c51b7d"],[0.2,"#de77ae"],[0.3,"#f1b6da"],[0.4,"#fde0ef"],[0.5,"#f7f7f7"],[0.6,"#e6f5d0"],[0.7,"#b8e186"],[0.8,"#7fbc41"],[0.9,"#4d9221"],[1,"#276419"]],"sequential":[[0.0,"#0d0887"],[0.1111111111111111,"#46039f"],[0.2222222222222222,"#7201a8"],[0.3333333333333333,"#9c179e"],[0.4444444444444444,"#bd3786"],[0.5555555555555556,"#d8576b"],[0.6666666666666666,"#ed7953"],[0.7777777777777778,"#fb9f3a"],[0.8888888888888888,"#fdca26"],[1.0,"#f0f921"]],"sequentialminus":[[0.0,"#0d0887"],[0.1111111111111111,"#46039f"],[0.2222222222222222,"#7201a8"],[0.3333333333333333,"#9c179e"],[0.4444444444444444,"#bd3786"],[0.5555555555555556,"#d8576b"],[0.6666666666666666,"#ed7953"],[0.7777777777777778,"#fb9f3a"],[0.8888888888888888,"#fdca26"],[1.0,"#f0f921"]]},"colorway":["#636efa","#EF553B","#00cc96","#ab63fa","#FFA15A","#19d3f3","#FF6692","#B6E880","#FF97FF","#FECB52"],"font":{"color":"#2a3f5f"},"geo":{"bgcolor":"white","lakecolor":"white","landcolor":"white","showlakes":true,"showland":true,"subunitcolor":"#C8D4E3"},"hoverlabel":{"align":"left"},"hovermode":"closest","mapbox":{"style":"light"},"paper_bgcolor":"white","plot_bgcolor":"white","polar":{"angularaxis":{"gridcolor":"#EBF0F8","linecolor":"#EBF0F8","ticks":""},"bgcolor":"white","radialaxis":{"gridcolor":"#EBF0F8","linecolor":"#EBF0F8","ticks":""}},"scene":{"xaxis":{"backgroundcolor":"white","gridcolor":"#DFE8F3","gridwidth":2,"linecolor":"#EBF0F8","showbackground":true,"ticks":"","zerolinecolor":"#EBF0F8"},"yaxis":{"backgroundcolor":"white","gridcolor":"#DFE8F3","gridwidth":2,"linecolor":"#EBF0F8","showbackground":true,"ticks":"","zerolinecolor":"#EBF0F8"},"zaxis":{"backgroundcolor":"white","gridcolor":"#DFE8F3","gridwidth":2,"linecolor":"#EBF0F8","showbackground":true,"ticks":"","zerolinecolor":"#EBF0F8"}},"shapedefaults":{"line":{"color":"#2a3f5f"}},"ternary":{"aaxis":{"gridcolor":"#DFE8F3","linecolor":"#A2B1C6","ticks":""},"baxis":{"gridcolor":"#DFE8F3","linecolor":"#A2B1C6","ticks":""},"bgcolor":"white","caxis":{"gridcolor":"#DFE8F3","linecolor":"#A2B1C6","ticks":""}},"title":{"x":0.05},"xaxis":{"automargin":true,"gridcolor":"#EBF0F8","linecolor":"#EBF0F8","ticks":"","title":{"standoff":15},"zerolinecolor":"#EBF0F8","zerolinewidth":2},"yaxis":{"automargin":true,"gridcolor":"#EBF0F8","linecolor":"#EBF0F8","ticks":"","title":{"standoff":15},"zerolinecolor":"#EBF0F8","zerolinewidth":2}}},"title":{"text":"Memory Usage with Recomputation","x":0.5,"y":0.95},"margin":{"r":80},"legend":{"y":0.95,"x":1.02,"xanchor":"left","yanchor":"top"},"updatemenus":[{"active":0,"buttons":[{"args":[{"visible":[true,true,true,true,false,false,false,false,false,false,false,false,false,false,false,false,false,false,false,false]},{"yaxis.range":[0,203.1547058105469]}],"label":"1B","method":"update"},{"args":[{"visible":[false,false,false,false,true,true,true,true,false,false,false,false,false,false,false,false,false,false,false,false]},{"yaxis.range":[0,314.4336639404297]}],"label":"3B","method":"update"},{"args":[{"visible":[false,false,false,false,false,false,false,false,true,true,true,true,false,false,false,false,false,false,false,false]},{"yaxis.range":[0,504.2233764648438]}],"label":"8B","method":"update"},{"args":[{"visible":[false,false,false,false,false,false,false,false,false,false,false,false,true,true,true,true,false,false,false,false]},{"yaxis.range":[0,3021.530541992188]}],"label":"70B","method":"update"},{"args":[{"visible":[false,false,false,false,false,false,false,false,false,false,false,false,false,false,false,false,true,true,true,true]},{"yaxis.range":[0,12823.0716796875]}],"label":"405B","method":"update"}],"direction":"down","showactive":true,"x":1.035,"xanchor":"left","y":0.6,"yanchor":"top"},{"active":0,"buttons":[{"args":[{"y":[[3.9879302978515625,3.9957427978515625,4.0113677978515625,4.0426177978515625],[3.9879302978515625,3.9957427978515625,4.0113677978515625,4.0426177978515625],[7.975860595703125,7.991485595703125,8.022735595703125,8.085235595703125],[3.564453125,12.12890625,44.2578125,168.515625],[13.296180725097656,13.307899475097656,13.331336975097656,13.378211975097656],[13.296180725097656,13.307899475097656,13.331336975097656,13.378211975097656],[26.592361450195312,26.615798950195312,26.662673950195312,26.756423950195312],[6.0732421875,18.708984375,63.66796875,232.3359375],[25.979034423828125,25.994659423828125,26.025909423828125,26.088409423828125],[25.979034423828125,25.994659423828125,26.025909423828125,26.088409423828125],[51.95806884765625,51.98931884765625,52.05181884765625,52.17681884765625],[9.25390625,28.5078125,97.015625,354.03125],[243.97711181640625,244.00836181640625,244.07086181640625,244.19586181640625],[243.97711181640625,244.00836181640625,244.07086181640625,244.19586181640625],[487.9542236328125,488.0167236328125,488.1417236328125,488.3917236328125],[46.2578125,142.515625,485.03125,1770.0625],[1519.99072265625,1520.05322265625,1520.17822265625,1520.42822265625],[1519.99072265625,1520.05322265625,1520.17822265625,1520.42822265625],[3039.9814453125,3040.1064453125,3040.3564453125,3040.8564453125],[145.703125,448.90625,1527.8125,5575.625]]}],"label":"None","method":"restyle"},{"args":[{"y":[[3.9879302978515625,3.9957427978515625,4.0113677978515625,4.0426177978515625],[3.9879302978515625,3.9957427978515625,4.0113677978515625,4.0426177978515625],[7.975860595703125,7.991485595703125,8.022735595703125,8.085235595703125],[1.064453125,2.12890625,4.2578125,8.515625],[13.296180725097656,13.307899475097656,13.331336975097656,13.378211975097656],[13.296180725097656,13.307899475097656,13.331336975097656,13.378211975097656],[26.592361450195312,26.615798950195312,26.662673950195312,26.756423950195312],[2.7919921875,5.583984375,11.16796875,22.3359375],[25.979034423828125,25.994659423828125,26.025909423828125,26.088409423828125],[25.979034423828125,25.994659423828125,26.025909423828125,26.088409423828125],[51.95806884765625,51.98931884765625,52.05181884765625,52.17681884765625],[4.25390625,8.5078125,17.015625,34.03125],[243.97711181640625,244.00836181640625,244.07086181640625,244.19586181640625],[243.97711181640625,244.00836181640625,244.07086181640625,244.19586181640625],[487.9542236328125,488.0167236328125,488.1417236328125,488.3917236328125],[21.2578125,42.515625,85.03125,170.0625],[1519.99072265625,1520.05322265625,1520.17822265625,1520.42822265625],[1519.99072265625,1520.05322265625,1520.17822265625,1520.42822265625],[3039.9814453125,3040.1064453125,3040.3564453125,3040.8564453125],[66.953125,133.90625,267.8125,535.625]]}],"label":"selective","method":"restyle"},{"args":[{"y":[[3.9879302978515625,3.9957427978515625,4.0113677978515625,4.0426177978515625],[3.9879302978515625,3.9957427978515625,4.0113677978515625,4.0426177978515625],[7.975860595703125,7.991485595703125,8.022735595703125,8.085235595703125],[0.064453125,0.12890625,0.2578125,0.515625],[13.296180725097656,13.307899475097656,13.331336975097656,13.378211975097656],[13.296180725097656,13.307899475097656,13.331336975097656,13.378211975097656],[26.592361450195312,26.615798950195312,26.662673950195312,26.756423950195312],[0.1669921875,0.333984375,0.66796875,1.3359375],[25.979034423828125,25.994659423828125,26.025909423828125,26.088409423828125],[25.979034423828125,25.994659423828125,26.025909423828125,26.088409423828125],[51.95806884765625,51.98931884765625,52.05181884765625,52.17681884765625],[0.25390625,0.5078125,1.015625,2.03125],[243.97711181640625,244.00836181640625,244.07086181640625,244.19586181640625],[243.97711181640625,244.00836181640625,244.07086181640625,244.19586181640625],[487.9542236328125,488.0167236328125,488.1417236328125,488.3917236328125],[1.2578125,2.515625,5.03125,10.0625],[1519.99072265625,1520.05322265625,1520.17822265625,1520.42822265625],[1519.99072265625,1520.05322265625,1520.17822265625,1520.42822265625],[3039.9814453125,3040.1064453125,3040.3564453125,3040.8564453125],[3.953125,7.90625,15.8125,31.625]]}],"label":"full","method":"restyle"}],"direction":"down","showactive":true,"x":1.035,"xanchor":"left","y":0.4,"yanchor":"top"}],"barmode":"stack","xaxis":{"title":{"text":"Sequence Length"}},"yaxis":{"title":{"text":"Memory (GB)"},"range":[0,203.1547058105469]},"width":850,"height":500,"annotations":[{"showarrow":false,"text":"Model Size:","x":1.035,"xanchor":"left","xref":"paper","y":0.6,"yanchor":"bottom","yref":"paper"},{"showarrow":false,"text":"Recomputation:","x":1.035,"xanchor":"left","xref":"paper","y":0.4,"yanchor":"bottom","yref":"paper"}]},                        {"responsive": true, "scrollZoom": false}                    )                };            </script>        </div>

dist/assets/data/plots/all_dumps_bad/agg_score.json

@@ -1 +0,0 @@
-{"data":{"big-run-refinedweb":{"x":[0,2.0971520000000003,4.194304000000001,6.291456,8.388608000000001,10.48576,12.582912,14.680064000000002,16.777216000000003,18.874368,20.97152,23.068672000000003,25.165824,27.262976000000002,28.311552000000002,29.360128000000003,31.45728,33.554432000000006,35.651584,37.748736,39.845888,41.94304,44.040192000000005,46.137344000000006,48.234496,50.331648,52.4288,54.525952000000004,56.623104000000005,58.720256000000006,60.817408,62.91456,65.011712,67.10886400000001,69.206016,71.303168,73.40032000000001,75.497472,77.59462400000001,79.691776,81.788928,83.88608,85.983232,88.08038400000001,90.177536,92.27468800000001,94.37184,96.468992,98.56614400000001,100.663296,102.76044800000001,104.8576,106.95475200000001,109.05190400000001,111.149056,113.24620800000001,115.34336,117.44051200000001,119.537664,121.634816,123.73196800000001,125.82912,127.92627200000001,130.023424,132.120576,134.21772800000002,136.31488000000002,138.412032,140.509184,142.606336,144.70348800000002,146.80064000000002,148.897792,150.994944,153.092096,155.18924800000002,157.28640000000001,159.383552,161.480704,163.577856,165.67500800000002,167.77216,169.869312,171.966464,174.06361600000002,176.16076800000002,178.25792,180.355072,182.452224,184.54937600000002,186.64652800000002,188.74368,190.840832,192.937984,195.03513600000002,197.13228800000002,199.22944,201.326592,203.423744,205.52089600000002,207.61804800000002,209.7152,211.812352,213.90950400000003,216.00665600000002,218.10380800000001,220.20096,222.298112,224.39526400000003,226.49241600000002,228.589568,230.68672,232.783872,234.88102400000002,236.97817600000002,239.075328,241.17248,243.269632,245.36678400000002,247.46393600000002,249.561088,251.65824,253.75539200000003,255.85254400000002,257.949696,260.046848,262.144,264.241152,266.338304,268.43545600000004,270.53260800000004,272.62976000000003,274.726912,276.824064,278.921216,281.018368,283.11552,285.212672,287.309824,289.40697600000004,291.50412800000004,293.60128000000003,295.698432,297.795584,299.892736,301.989888,304.08704,306.184192,308.28134400000005,310.37849600000004,312.47564800000004,314.57280000000003,316.669952,318.767104,320.864256,322.961408,325.05856,327.155712,329.25286400000005,331.35001600000004,333.44716800000003,335.54432,337.641472,339.738624,341.835776,343.932928,346.03008,348.12723200000005,350.22438400000004],"y":[0.3308933284133672,0.3534814938902855,0.3764607086777687,0.38782499730587,0.3981050960719585,0.4028486795723438,0.4125883243978023,0.4117814563214779,0.414029736071825,0.4197172522544861,0.4211113378405571,0.4279881417751312,0.4280137903988361,0.4280424378812313,0.4291964024305343,0.4326301179826259,0.4371833503246307,0.4346669465303421,0.4336562640964985,0.4432648755609989,0.4401291646063328,0.4394684173166752,0.4476612061262131,0.4465444348752498,0.4472153298556804,0.4433343075215816,0.4510187618434429,0.4459567815065384,0.4460812956094742,0.4498684890568256,0.4529943652451038,0.4528274349868297,0.4551213420927524,0.4549156539142132,0.4564928151667118,0.4576693661510944,0.4557182416319847,0.4536240361630916,0.457439012825489,0.4570476822555065,0.4589823484420776,0.462024375796318,0.4540738053619861,0.4550252184271812,0.4576593860983848,0.4573238864541054,0.4575810581445694,0.4622134491801262,0.4592566937208175,0.4614734016358852,0.4637473002076149,0.4625372551381588,0.4613912180066108,0.4597448222339153,0.4594792164862156,0.4662549719214439,0.4634026065468788,0.4633508697152138,0.4635734222829342,0.4628961533308029,0.4670135043561458,0.4639505892992019,0.4631133340299129,0.4665167145431041,0.4672448337078094,0.4693268723785877,0.4630668573081493,0.4676454700529575,0.4646359197795391,0.4621579721570015,0.4692446552217006,0.4704835228621959,0.4663223996758461,0.4680556617677212,0.466339822858572,0.4682099223136902,0.4711195565760135,0.4722655527293682,0.4727961830794811,0.4676857478916645,0.4719390422105789,0.4713102728128433,0.4712141714990139,0.4721613004803657,0.4713456854224205,0.4682970903813839,0.4679934531450271,0.4685162976384163,0.4679946713149547,0.4681242071092129,0.4702276065945625,0.472664151340723,0.4730790853500366,0.4731674715876579,0.4718914777040481,0.4719801284372806,0.4761029370129108,0.4735167175531387,0.4730370938777923,0.4730173237621784,0.4735377207398414,0.4777223989367485,0.4796326830983162,0.4734170883893966,0.4739485755562782,0.4748299159109592,0.4765299335122108,0.4745025858283043,0.4754423759877682,0.4784592799842357,0.4761341325938701,0.4760282784700393,0.4769757278263569,0.47154351323843,0.4786738082766533,0.4804279990494251,0.4777076803147793,0.4798569902777672,0.4759011939167976,0.4784621745347976,0.479673832654953,0.4780617095530033,0.48076206818223,0.47995800152421,0.4790860973298549,0.4817167408764362,0.4811586998403072,0.482547752559185,0.4816697351634502,0.4809327870607376,0.4816545359790325,0.4804601892828941,0.4776877984404564,0.4813711903989315,0.4844604581594467,0.4819537848234176,0.4820829331874847,0.4778126627206802,0.482935007661581,0.48230691999197,0.4826001971960068,0.4823969900608063,0.4811219945549965,0.4789146520197391,0.484035175293684,0.4848698377609253,0.4855728335678577,0.4825376532971859,0.485215101391077,0.4824351668357849,0.4835342466831207,0.4822137206792831,0.4838785007596016,0.4837255179882049,0.4853012599050998,0.4857851006090641,0.4863366298377514,0.4856646582484245,0.4842503517866134,0.4838776960968971,0.4846346862614155,0.4837041422724724,0.4813097268342972,0.4873070046305656,0.4841253720223903,0.4837464913725853,0.483069509267807,0.4851242564618587,0.4861010462045669],"label":"RefinedWeb"},"big-run-sampled_full_filtered_no_dedup":{"x":[0,2.0971520000000003,4.194304000000001,6.291456,8.388608000000001,10.48576,12.582912,14.680064000000002,16.777216000000003,18.874368,20.97152,23.068672000000003,25.165824,27.262976000000002,28.311552000000002,29.360128000000003,31.45728,33.554432000000006,35.651584,37.748736,39.845888,41.94304,44.040192000000005,46.137344000000006,48.234496,50.331648,52.4288,54.525952000000004,56.623104000000005,58.720256000000006,60.817408,62.91456,65.011712,67.10886400000001,69.206016,71.303168,73.40032000000001,75.497472,77.59462400000001,79.691776,81.788928,83.88608,85.983232,88.08038400000001,90.177536,92.27468800000001,94.37184,96.468992,98.56614400000001,100.663296,102.76044800000001,104.8576,106.95475200000001,109.05190400000001,111.149056,113.24620800000001,115.34336,117.44051200000001,119.537664,121.634816,123.73196800000001,125.82912,127.92627200000001,130.023424,132.120576,134.21772800000002,136.31488000000002,138.412032,140.509184,142.606336,144.70348800000002,146.80064000000002,148.897792,150.994944,153.092096,155.18924800000002,157.28640000000001,159.383552,161.480704,163.577856,165.67500800000002,167.77216,169.869312,171.966464,174.06361600000002,176.16076800000002,178.25792,180.355072,182.452224,184.54937600000002,186.64652800000002,188.74368,190.840832,192.937984,195.03513600000002,197.13228800000002,199.22944,201.326592,203.423744,205.52089600000002,207.61804800000002,209.7152,211.812352,213.90950400000003,216.00665600000002,218.10380800000001,220.20096,222.298112,224.39526400000003,226.49241600000002,228.589568,230.68672,232.783872,234.88102400000002,236.97817600000002,239.075328,241.17248,243.269632,245.36678400000002,247.46393600000002,249.561088,251.65824,253.75539200000003,255.85254400000002,257.949696,260.046848,262.144,264.241152,266.338304,268.43545600000004,270.53260800000004,272.62976000000003,274.726912,276.824064,278.921216,281.018368,283.11552,285.212672,287.309824,289.40697600000004,291.50412800000004,293.60128000000003,295.698432,297.795584,299.892736,301.989888,304.08704,306.184192,308.28134400000005,310.37849600000004,312.47564800000004,314.57280000000003,316.669952,318.767104,320.864256,322.961408,325.05856,327.155712,329.25286400000005,331.35001600000004,333.44716800000003,335.54432,337.641472,339.738624,341.835776,343.932928,346.03008,348.12723200000005,350.22438400000004],"y":[0.3308933284133672,0.3605199865996837,0.3733148723840713,0.3882005847990513,0.3934122696518898,0.3947227671742439,0.4042885974049568,0.3974800482392311,0.4055779427289963,0.4133470430970192,0.4117913842201233,0.4113653488457203,0.4149517640471458,0.4187851920723915,0.4252083078026771,0.4206527359783649,0.4240428246557712,0.422003373503685,0.4280910938978195,0.4244147576391697,0.4316282644867897,0.4295645765960216,0.4310102686285972,0.4360743537545204,0.4313482865691185,0.4350991360843181,0.4378576353192329,0.4335876516997814,0.4347924515604973,0.4348904751241207,0.436600212007761,0.430036511272192,0.4350974671542644,0.4399556629359722,0.4371416717767715,0.4363861419260502,0.4376698136329651,0.4405004419386387,0.4373639523983001,0.4379038028419018,0.4371281825006008,0.4393439553678036,0.440426729619503,0.4401675276458263,0.4429537951946258,0.4449137263000011,0.4434786736965179,0.4450470842421055,0.4454202279448509,0.4394537284970283,0.442185215651989,0.4461225643754005,0.4427758157253265,0.4430646039545536,0.4476901069283485,0.4478763341903686,0.4493869319558143,0.4448477327823639,0.450044184923172,0.4498609118163585,0.4457665979862213,0.4506924152374267,0.449855338782072,0.448790930211544,0.4474099352955818,0.4546772800385952,0.4529431238770485,0.452015146613121,0.4502020999789238,0.4493804536759853,0.4523266032338142,0.4551868587732315,0.4501944817602634,0.4493303671479225,0.4526805207133293,0.4533850513398647,0.4518048763275146,0.4518973492085933,0.4531301632523536,0.4518006071448326,0.4553494565188885,0.4528752230107784,0.4536322727799415,0.4561733976006508,0.4549491256475448,0.4574789106845855,0.4577847123146057,0.4563642293214798,0.4578686729073524,0.4561499990522861,0.4537816494703293,0.4542164430022239,0.4559455662965774,0.4554723873734474,0.4575514122843742,0.4575202167034149,0.4592722058296203,0.4585275091230869,0.4580587856471538,0.456934317946434,0.4577495418488979,0.4540119916200638,0.4570806957781315,0.4608120545744896,0.4588425755500793,0.4578334167599678,0.4610816091299057,0.4598177038133144,0.461849745362997,0.4631866924464702,0.4601576402783394,0.4646804705262184,0.4632389545440674,0.4604574106633663,0.4602976888418197,0.4581312239170074,0.4654182009398937,0.4655338563024997,0.4616620391607284,0.461054053157568,0.4613021649420261,0.4658613465726375,0.4633531905710697,0.4613638147711754,0.4643996246159076,0.462500050663948,0.4650798961520195,0.4648764543235302,0.4639869071543216,0.4634246975183487,0.46585888043046,0.4639799632132053,0.4630857892334461,0.4644265696406364,0.4642998576164245,0.4686848931014538,0.4687492996454239,0.4650243632495403,0.4627032242715359,0.4665953740477562,0.4660026729106903,0.4664581045508384,0.4676475040614605,0.4657339677214622,0.4664678275585174,0.4673498086631298,0.4676674827933311,0.4680955372750759,0.4681585058569908,0.4659864418208599,0.4686457589268684,0.4661462865769863,0.4658931568264961,0.4674226939678192,0.46805215254426,0.4682257212698459,0.4689070098102093,0.4699570722877979,0.4655096270143986,0.4688013233244419,0.4707522802054882,0.4661469310522079,0.4688841328024864,0.4671329781413078,0.4662554152309894,0.4697433896362781,0.4698473587632179,0.4676505327224731,0.4696521013975143],"label":"FineWeb filtered only"},"big-run-fineweb-cross-dedup-fixed":{"x":[0,2.0971520000000003,4.194304000000001,6.291456,8.388608000000001,10.48576,12.582912,14.680064000000002,16.777216000000003,18.874368,20.97152,23.068672000000003,25.165824,27.262976000000002,29.360128000000003,31.45728,33.554432000000006,35.651584,37.748736,39.845888,41.94304,44.040192000000005,46.137344000000006,48.234496,50.331648,52.4288,54.525952000000004,56.623104000000005,58.720256000000006,60.817408,62.91456,65.011712,67.10886400000001,69.206016,71.303168,73.40032000000001,75.497472,77.59462400000001,79.691776,81.788928,83.88608,85.983232,88.08038400000001,90.177536,92.27468800000001,94.37184,96.468992,98.56614400000001,100.663296,102.76044800000001,104.8576,106.95475200000001,109.05190400000001,111.149056,113.24620800000001,115.34336,117.44051200000001,119.537664,121.634816,123.73196800000001,125.82912,127.92627200000001,130.023424,132.120576,134.21772800000002,136.31488000000002,138.412032,140.509184,142.606336,144.70348800000002,146.80064000000002,148.897792,150.994944,153.092096,155.18924800000002,157.28640000000001,159.383552,161.480704,163.577856,165.67500800000002,167.77216,169.869312,171.966464,174.06361600000002,176.16076800000002,178.25792,180.355072,182.452224,184.54937600000002,186.64652800000002,188.74368,190.840832,192.937984,195.03513600000002,197.13228800000002,199.22944,201.326592,203.423744,205.52089600000002,207.61804800000002,209.7152,211.812352,213.90950400000003,216.00665600000002,218.10380800000001,220.20096,222.298112,224.39526400000003,226.49241600000002,228.589568,230.68672,232.783872,234.88102400000002,236.97817600000002,239.075328,241.17248,243.269632,245.36678400000002,247.46393600000002,249.561088,251.65824,253.75539200000003,255.85254400000002,257.949696,260.046848,262.144,264.241152,266.338304,268.43545600000004,270.53260800000004,272.62976000000003,274.726912,276.824064,278.921216,281.018368,283.11552,285.212672,287.309824,289.40697600000004,291.50412800000004,293.60128000000003,295.698432,297.795584,299.892736,301.989888,304.08704,306.184192,308.28134400000005,310.37849600000004,312.47564800000004,314.57280000000003,316.669952,318.767104,320.864256,322.961408,325.05856,327.155712,329.25286400000005,331.35001600000004,333.44716800000003,335.54432,337.641472,339.738624,341.835776,343.932928,346.03008,348.12723200000005,350.22438400000004],"y":[0.3308933284133672,0.3551952373236418,0.3736435137689113,0.3814037963747978,0.3948809280991554,0.3996850810945034,0.4089604057371616,0.4100853353738785,0.4119834117591381,0.4168377220630646,0.4186493046581745,0.4169826358556747,0.4234288297593593,0.4229162000119686,0.4273439794778824,0.4290364980697632,0.4291782416403293,0.4296907968819141,0.4311576783657074,0.4326641112565994,0.430318683385849,0.430436260998249,0.4339037239551544,0.4363459683954716,0.4357402548193931,0.4342963136732578,0.4366712383925915,0.4363959729671478,0.436981026083231,0.4447868093848228,0.4411709941923618,0.4406092017889023,0.4424176625907421,0.4423875361680984,0.4422253370285034,0.4410557933151722,0.4447037056088447,0.4454837813973427,0.4435960277915001,0.4468514993786812,0.4479999616742134,0.4428562931716442,0.445764634758234,0.4456562362611294,0.4488007053732872,0.4475954286754131,0.4468922987580299,0.4548408314585686,0.4511027485132217,0.4530330970883369,0.4483681954443455,0.4531726539134979,0.45334542542696,0.4544384703040123,0.4530758671462536,0.4540613554418087,0.4510113634169101,0.4538320265710354,0.4518541917204857,0.4536847211420536,0.4532708041369915,0.4552236869931221,0.455034039914608,0.4562875479459762,0.4532428197562694,0.4574853852391243,0.4517738744616508,0.4579889141023159,0.4538268558681011,0.456730306148529,0.4526018649339676,0.4562746733427048,0.4560015797615051,0.4555426277220249,0.4561501257121563,0.4524396173655987,0.4557023830711841,0.4589769169688225,0.4581078588962555,0.4620813727378845,0.4586601965129375,0.4568093195557594,0.4569808952510357,0.4567535072565079,0.4575250148773193,0.4606908001005649,0.4603964723646641,0.4622848592698574,0.4594669193029403,0.4640629850327968,0.4604269936680794,0.4634841009974479,0.4644578285515308,0.4642514958977699,0.4666304066777229,0.4616626128554344,0.4588956907391548,0.4620226770639419,0.4628621749579906,0.4595407098531723,0.4635516740381717,0.46005355194211,0.4601523540914058,0.4644204638898372,0.4620639197528362,0.46614545956254,0.4636696502566337,0.4610077403485775,0.4640897810459137,0.4636163525283336,0.4630545899271965,0.466012816876173,0.4650349207222461,0.4613720141351223,0.4644323363900184,0.4647249802947044,0.4656480401754379,0.4651664271950722,0.4622530452907085,0.4655019529163837,0.4650313258171081,0.466718140989542,0.4661559611558914,0.4661237150430679,0.4664223715662956,0.4640601389110088,0.4642657749354839,0.4633881188929081,0.4629989042878151,0.4685831367969513,0.4675870984792709,0.467183344066143,0.4678030684590339,0.4660939238965511,0.4691914953291416,0.4670972637832165,0.468262892216444,0.4672016054391861,0.4676182121038437,0.4698677137494087,0.4658828042447567,0.4701816700398922,0.4684622809290886,0.466015312820673,0.4675401039421558,0.4693200923502445,0.4702670983970165,0.4679145030677318,0.4676233418285846,0.4674933589994907,0.4678357951343059,0.4669915996491909,0.4657857678830623,0.4666901864111423,0.4669371582567692,0.4672787226736545,0.4684535376727581,0.4685697965323925,0.4694835692644119,0.4683254994451999,0.4712230190634727,0.4683987610042095,0.4707653746008873,0.4663059376180172,0.4683133698999882,0.4686385430395603,0.4657671600580215,0.4692615270614624],"label":"FineWeb full MinHash"}},"layout":{"title":{"text":"Dedup across all dumps does not improve performance"}}}

dist/assets/images/activation_recomputation.js

@@ -1 +1,149 @@
-function enhanceSVGContent(t){const e=(new DOMParser).parseFromString(t,"image/svg+xml"),n=e.createElementNS("http://www.w3.org/2000/svg","style");return n.textContent='\n      path[data-element-type="layer"] {\n      transition: all 0.3s;\n      cursor: pointer;\n      }\n      path[data-element-type="layer"]:hover {\n      fill: #b197fc !important;\n      transform: translate(0, -2px);\n      }\n\n      path[data-element-type="layer-updated"] {\n      transition: all 0.3s;\n      cursor: pointer;\n      }\n      \n      path[data-element-type="layer-updated"]:hover {\n      fill:rgb(103, 56, 244) !important;\n      transform: scale(1.02);\n      transform: translate(0, -2px);\n      }\n\n      path[data-element-type="gradient"] {\n      transition: all 0.3s;\n      cursor: pointer;\n      }\n      path[data-element-type="gradient"]:hover {\n      fill: #f06595 !important;\n      transform: translate(0, -2px);\n      }\n\n      path[data-element-type="forward"] {\n      transition: all 0.3s;\n      cursor: pointer;\n      }\n      path[data-element-type="forward"]:hover {\n      stroke: #0c8599 !important;\n      stroke-width: 4 !important;\n      }\n\n      path[data-element-type="backward"] {\n      transition: all 0.3s;\n      cursor: pointer;\n      }\n      path[data-element-type="backward"]:hover {\n      stroke: #e8590c !important;\n      stroke-width: 4 !important;\n      }\n\n      path[data-element-type="optimization"] {\n      transition: all 0.3s;\n      cursor: pointer;\n      }\n      path[data-element-type="optimization"]:hover {\n      stroke: #087f5b !important;\n      stroke-width: 4 !important;\n      }\n',e.documentElement.insertBefore(n,e.documentElement.firstChild),e.querySelectorAll('path[fill="#d0bfff"]').forEach(((t,e)=>{t.setAttribute("data-element-id",`layer-${e}`),t.setAttribute("data-element-type","layer")})),e.querySelectorAll('path[fill="#9775fa"]').forEach(((t,e)=>{t.setAttribute("data-element-id",`layer-updated-${e}`),t.setAttribute("data-element-type","layer-updated")})),e.querySelectorAll('path[fill="#f783ac"]').forEach(((t,e)=>{t.setAttribute("data-element-id",`gradient-${e}`),t.setAttribute("data-element-type","gradient")})),Object.entries({"#15aabf":"forward","#fd7e14":"backward","#099268":"optimization"}).forEach((([t,n])=>{e.querySelectorAll(`path[stroke="${t}"]`).forEach(((t,e)=>{t.setAttribute("data-element-id",`${n}-${e}`),t.setAttribute("data-element-type",n)}))})),e.documentElement.setAttribute("width","100%"),e.documentElement.setAttribute("height","100%"),e.documentElement.setAttribute("preserveAspectRatio","xMidYMid meet"),(new XMLSerializer).serializeToString(e)}async function loadSVG(t,e){try{const n=await fetch(t);if(!n.ok)throw new Error(`HTTP error! Status: ${n.status}`);const a=enhanceSVGContent(await n.text());document.getElementById(e).innerHTML=a}catch(t){console.error("Error loading SVG:",t),document.getElementById(e).innerHTML="<p>Error loading SVG.</p>"}}loadSVG("../assets/images/activation_recomputation.svg","svg-activation_recomputation");const svgContainer3=document.getElementById("svg-activation_recomputation");svgContainer3.addEventListener("mouseover",(function(t){const e=t.target;if("path"===e.tagName.toLowerCase()&&e.hasAttribute("data-element-id")){const t=e.getAttribute("data-element-id"),n=e.getAttribute("data-element-type"),a={layer:"Neural Network Layer","layer-updated":"Neural Network Layer (updated)",gradient:"Gradient Update Layer",forward:"Forward Pass Connection",backward:"Backward Pass Connection",optimization:"Optimization Step"}[n]||n;document.getElementById("svg-activation_recomputation-info").textContent=`Hovering over: ${a} (${t})`}})),svgContainer3.addEventListener("mouseout",(function(){document.getElementById("svg-activation_recomputation-info").textContent="Hover over the network elements to see their details"}));
+
+// Function to enhance the SVG content by adding styles and data attributes
+function enhanceSVGContent(originalContent) {
+      const parser = new DOMParser();
+      const doc = parser.parseFromString(originalContent, 'image/svg+xml');
+
+      // Create a style element with hover effects and insert it as the first child of the SVG
+      const styleElement = doc.createElementNS('http://www.w3.org/2000/svg', 'style');
+      styleElement.textContent = `
+      path[data-element-type="layer"] {
+      transition: all 0.3s;
+      cursor: pointer;
+      }
+      path[data-element-type="layer"]:hover {
+      fill: #b197fc !important;
+      transform: translate(0, -2px);
+      }
+
+      path[data-element-type="layer-updated"] {
+      transition: all 0.3s;
+      cursor: pointer;
+      }
+      
+      path[data-element-type="layer-updated"]:hover {
+      fill:rgb(103, 56, 244) !important;
+      transform: scale(1.02);
+      transform: translate(0, -2px);
+      }
+
+      path[data-element-type="gradient"] {
+      transition: all 0.3s;
+      cursor: pointer;
+      }
+      path[data-element-type="gradient"]:hover {
+      fill: #f06595 !important;
+      transform: translate(0, -2px);
+      }
+
+      path[data-element-type="forward"] {
+      transition: all 0.3s;
+      cursor: pointer;
+      }
+      path[data-element-type="forward"]:hover {
+      stroke: #0c8599 !important;
+      stroke-width: 4 !important;
+      }
+
+      path[data-element-type="backward"] {
+      transition: all 0.3s;
+      cursor: pointer;
+      }
+      path[data-element-type="backward"]:hover {
+      stroke: #e8590c !important;
+      stroke-width: 4 !important;
+      }
+
+      path[data-element-type="optimization"] {
+      transition: all 0.3s;
+      cursor: pointer;
+      }
+      path[data-element-type="optimization"]:hover {
+      stroke: #087f5b !important;
+      stroke-width: 4 !important;
+      }
+`;
+      doc.documentElement.insertBefore(styleElement, doc.documentElement.firstChild);
+
+      // Process neural network layers (purple nodes)
+      doc.querySelectorAll('path[fill="#d0bfff"]').forEach((node, index) => {
+            node.setAttribute('data-element-id', `layer-${index}`);
+            node.setAttribute('data-element-type', 'layer');
+      });
+
+      doc.querySelectorAll('path[fill="#9775fa"]').forEach((node, index) => {
+            node.setAttribute('data-element-id', `layer-updated-${index}`);
+            node.setAttribute('data-element-type', 'layer-updated');
+      });
+
+      // Process gradient nodes (pink nodes)
+      doc.querySelectorAll('path[fill="#f783ac"]').forEach((node, index) => {
+            node.setAttribute('data-element-id', `gradient-${index}`);
+            node.setAttribute('data-element-type', 'gradient');
+      });
+
+      // Process arrows by matching stroke colors
+      const arrowTypes = {
+            '#15aabf': 'forward',
+            '#fd7e14': 'backward',
+            '#099268': 'optimization'
+      };
+
+      Object.entries(arrowTypes).forEach(([color, type]) => {
+            doc.querySelectorAll(`path[stroke="${color}"]`).forEach((arrow, index) => {
+                  arrow.setAttribute('data-element-id', `${type}-${index}`);
+                  arrow.setAttribute('data-element-type', type);
+            });
+      });
+
+      // Make the SVG responsive
+      doc.documentElement.setAttribute('width', '100%');
+      doc.documentElement.setAttribute('height', '100%');
+      doc.documentElement.setAttribute('preserveAspectRatio', 'xMidYMid meet');
+
+      return new XMLSerializer().serializeToString(doc);
+}
+
+// Function to load an SVG file via fetch
+async function loadSVG(url, containerId) {
+      try {
+            const response = await fetch(url);
+            if (!response.ok) {
+                  throw new Error(`HTTP error! Status: ${response.status}`);
+            }
+            const svgText = await response.text();
+            const enhancedSVG = enhanceSVGContent(svgText);
+            document.getElementById(containerId).innerHTML = enhancedSVG;
+      } catch (error) {
+            console.error('Error loading SVG:', error);
+            document.getElementById(containerId).innerHTML = '<p>Error loading SVG.</p>';
+      }
+}
+
+// Load the SVG file (adjust the path if needed)
+loadSVG('../assets/images/activation_recomputation.svg', 'svg-activation_recomputation');
+
+// Set up event listeners to display a description of the hovered element
+const svgContainer3 = document.getElementById('svg-activation_recomputation');
+
+svgContainer3.addEventListener('mouseover', function (event) {
+      const target = event.target;
+      if (target.tagName.toLowerCase() === 'path' && target.hasAttribute('data-element-id')) {
+            const elementId = target.getAttribute('data-element-id');
+            const elementType = target.getAttribute('data-element-type');
+            const descriptions = {
+                  layer: 'Neural Network Layer',
+                  'layer-updated': 'Neural Network Layer (updated)',
+                  gradient: 'Gradient Update Layer',
+                  forward: 'Forward Pass Connection',
+                  backward: 'Backward Pass Connection',
+                  optimization: 'Optimization Step'
+            };
+            const description = descriptions[elementType] || elementType;
+            document.getElementById('svg-activation_recomputation-info').textContent = `Hovering over: ${description} (${elementId})`;
+      }
+});
+
+svgContainer3.addEventListener('mouseout', function () {
+      document.getElementById('svg-activation_recomputation-info').textContent = 'Hover over the network elements to see their details';
+});