Spaces:
Running
Running
testing
Browse files- dist/index.html +1 -146
dist/index.html
CHANGED
@@ -1,148 +1,3 @@
|
|
1 |
-
<!
|
2 |
<html>
|
3 |
-
<head>
|
4 |
-
<script src="distill.bundle.js" type="module" fetchpriority="high" blocking></script>
|
5 |
-
<script src="main.bundle.js" type="module" fetchpriority="low" defer></script>
|
6 |
-
<meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
|
7 |
-
<meta charset="utf8">
|
8 |
-
<base target="_blank">
|
9 |
-
<title>Scaling FineWeb to 1000+ languages: Step 1: finding signal in 100s of evaluation tasks</title>
|
10 |
-
<link rel="stylesheet" href="style.css">
|
11 |
-
</head>
|
12 |
-
|
13 |
-
<body>
|
14 |
-
<d-front-matter>
|
15 |
-
<script id='distill-front-matter' type="text/json">{
|
16 |
-
"title": "📝 Scaling FineWeb to 1000+ languages: Step 1: finding signal in 100s of evaluation tasks",
|
17 |
-
"description": "This blog covers a discussion on multilingual evaluation and task signal, the processes for selecting existing evaluation tasks based on signal resulting in FineTasks, and comparisson of open and closed sourced on the FineTasks.",
|
18 |
-
"published": "Oct 23, 2024",
|
19 |
-
"affiliation": {"name": "HuggingFace"},
|
20 |
-
"authors": [
|
21 |
-
{
|
22 |
-
"author":"Hynek Kydlíček",
|
23 |
-
"authorURL":"https://huggingface.co/hynky"
|
24 |
-
},
|
25 |
-
{
|
26 |
-
"author":"Guilherme Penedo",
|
27 |
-
"authorURL":"https://huggingface.co/guipenedo"
|
28 |
-
},
|
29 |
-
{
|
30 |
-
"author":"Clémentine Fourier",
|
31 |
-
"authorURL":"https://huggingface.co/clefourrier"
|
32 |
-
},
|
33 |
-
{
|
34 |
-
"author":"Nathan Habib",
|
35 |
-
"authorURL":"https://huggingface.co/SaylorTwift"
|
36 |
-
},
|
37 |
-
{
|
38 |
-
"author":"Thomas Wolf",
|
39 |
-
"authorURL":"https://huggingface.co/thomwolf"
|
40 |
-
}
|
41 |
-
]
|
42 |
-
}</script>
|
43 |
-
</d-front-matter>
|
44 |
-
|
45 |
-
<d-byline></d-byline>
|
46 |
-
<d-article>
|
47 |
-
<d-contents>
|
48 |
-
</d-contents>
|
49 |
-
<p>We're looking forward to revisiting this analysis in the future, not with just 9 languages, but at least 50—thanks to community contributions! Let's level the playing field between English and other languages together! 🤗</p>
|
50 |
-
<d-math> 1+1=2 </d-math>
|
51 |
-
</d-article>
|
52 |
-
|
53 |
-
<d-appendix>
|
54 |
-
<d-bibliography src="bibliography.bib"></d-bibliography>
|
55 |
-
<style>
|
56 |
-
d-appendix .citation {
|
57 |
-
font-size: 11px;
|
58 |
-
line-height: 15px;
|
59 |
-
border-left: 1px solid rgba(0, 0, 0, 0.1);
|
60 |
-
padding-left: 18px;
|
61 |
-
border: 1px solid rgba(0,0,0,0.1);
|
62 |
-
background: rgba(0, 0, 0, 0.02);
|
63 |
-
padding: 10px 18px;
|
64 |
-
border-radius: 3px;
|
65 |
-
color: rgba(150, 150, 150, 1);
|
66 |
-
overflow: hidden;
|
67 |
-
margin-top: -12px;
|
68 |
-
white-space: pre-wrap;
|
69 |
-
word-wrap: break-word;
|
70 |
-
}
|
71 |
-
</style>
|
72 |
-
|
73 |
-
<h3 id="citation">Citation</h3>
|
74 |
-
<p>For attribution in academic contexts, please cite this work as</p>
|
75 |
-
<pre class="citation short">Kydlicek, et al., "FineTasks: Finding signal in a haystack of 200+ multilingual tasks", 2024.</pre>
|
76 |
-
<p>BibTeX citation</p>
|
77 |
-
<pre class="citation long">@misc{kydlicek2024finetasksmultilingualtasks,
|
78 |
-
title={FineTasks: Finding signal in a haystack of 200+ multilingual tasks},
|
79 |
-
author={Hynek Kydlíček and Guilherme Penedo and Clémentine Fourier and Nathan Habib and Thomas Wolf},
|
80 |
-
url={https://huggingface.co/spaces/HuggingFaceFW/blogpost-fine-tasks},
|
81 |
-
}</pre>
|
82 |
-
</d-appendix>
|
83 |
-
|
84 |
-
<script>
|
85 |
-
const article = document.querySelector('d-article');
|
86 |
-
const toc = document.querySelector('d-contents');
|
87 |
-
if (toc) {
|
88 |
-
const headings = article.querySelectorAll('h2, h3, h4');
|
89 |
-
let ToC = `<nav role="navigation" class="l-text figcaption"><h3>Table of contents</h3>`;
|
90 |
-
let prevLevel = 0;
|
91 |
-
|
92 |
-
for (const el of headings) {
|
93 |
-
// should element be included in TOC?
|
94 |
-
const isInTitle = el.parentElement.tagName == 'D-TITLE';
|
95 |
-
const isException = el.getAttribute('no-toc');
|
96 |
-
if (isInTitle || isException) continue;
|
97 |
-
el.setAttribute('id', el.textContent.toLowerCase().replaceAll(" ", "_"))
|
98 |
-
const link = '<a target="_self" href="' + '#' + el.getAttribute('id') + '">' + el.textContent + '</a>';
|
99 |
-
|
100 |
-
const level = el.tagName === 'H2' ? 0 : (el.tagName === 'H3' ? 1 : 2);
|
101 |
-
while (prevLevel < level) {
|
102 |
-
ToC += '<ul>'
|
103 |
-
prevLevel++;
|
104 |
-
}
|
105 |
-
while (prevLevel > level) {
|
106 |
-
ToC += '</ul>'
|
107 |
-
prevLevel--;
|
108 |
-
}
|
109 |
-
if (level === 0)
|
110 |
-
ToC += '<div>' + link + '</div>';
|
111 |
-
else
|
112 |
-
ToC += '<li>' + link + '</li>';
|
113 |
-
}
|
114 |
-
|
115 |
-
while (prevLevel > 0) {
|
116 |
-
ToC += '</ul>'
|
117 |
-
prevLevel--;
|
118 |
-
}
|
119 |
-
ToC += '</nav>';
|
120 |
-
toc.innerHTML = ToC;
|
121 |
-
toc.setAttribute('prerendered', 'true');
|
122 |
-
const toc_links = document.querySelectorAll('d-contents > nav a');
|
123 |
-
|
124 |
-
window.addEventListener('scroll', (_event) => {
|
125 |
-
if (typeof (headings) != 'undefined' && headings != null && typeof (toc_links) != 'undefined' && toc_links != null) {
|
126 |
-
// Then iterate forwards, on the first match highlight it and break
|
127 |
-
find_active: {
|
128 |
-
for (let i = headings.length - 1; i >= 0; i--) {
|
129 |
-
if (headings[i].getBoundingClientRect().top - 50 <= 0) {
|
130 |
-
if (!toc_links[i].classList.contains("active")) {
|
131 |
-
toc_links.forEach((link, _index) => {
|
132 |
-
link.classList.remove("active");
|
133 |
-
});
|
134 |
-
toc_links[i].classList.add('active');
|
135 |
-
}
|
136 |
-
break find_active;
|
137 |
-
}
|
138 |
-
}
|
139 |
-
toc_links.forEach((link, _index) => {
|
140 |
-
link.classList.remove("active");
|
141 |
-
});
|
142 |
-
}
|
143 |
-
}
|
144 |
-
});
|
145 |
-
}
|
146 |
-
</script>
|
147 |
-
</body>
|
148 |
</html>
|
|
|
1 |
+
<!doctype html>
|
2 |
<html>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
</html>
|