-
Notifications
You must be signed in to change notification settings - Fork 2
/
index.html
369 lines (346 loc) · 22.3 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
<!DOCTYPE html>
<html lang="en">
<head>
<!-- Google tag (gtag.js) -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-0FFN6N7318"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'G-0FFN6N7318');
</script>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1, user-scalable=no" />
<!-- SEO -->
<title>LLM360 | Community-Driven AGI via Open-Source LLMs 🚀</title>
<meta name="description" content="Join LLM360 in advancing community-owned AGi with open-source resources for transparent and ethical LLM development. Explore pretraining suites and more."/>
<link rel="canonical" href="https://www.llm360.ai/" />
<meta name="keywords" content="LLM360, Open Source AI, Community AGI, Large Language Models, AI Research, Petuum, CMU, USC, MBZUAI, Amber LLM, CrystalCoder, NLP, Language Processing"/>
<meta name="author" content="Petuum, Carnegie Mellon University, University of Southern California, Mohamed bin Zayed University of Artificial Intelligence"/>
<meta name="robots" content="index, follow"/>
<!-- Open Graph Protocol -->
<meta property="og:title" content="LLM360 | Community-Driven AGI via Open-Source LLMs"/>
<meta property="og:description" content="Empower your AI journey with LLM360’s open-source resources. Drive forward the ethical development and accessibility of LLMs for everyone in the community."/>
<meta property="og:type" content="website"/>
<meta onboarding="og:url" content="https://www.llm360.ai/"/>
<meta property="og:image" content="https://www.llm360.ai/images/open-graph-image.jpg"/>
<!-- Twitter Card -->
<meta name="twitter:card" content="summary_large_image">
<meta name="twitter:site" content="@llm360">
<meta name="twitter:title" content="LLM360 | Community-Driven AGI via Open-Source LLMs">
<meta name="twitter:description" content="Explore how LLM360 is democratizing AGI with open-source large language models, fostering an inclusive and ethical AI development environment.">
<meta name="twitter:image" content="https://www.llm360.ai/images/open-graph-image.jpg">
<!-- Schema Markup -->
<script type="application/ld+json">
{
"@context": "http://schema.org",
"@type": "WebSite",
"name": "LLM360",
"url": "https://www.llm360.ai/",
"description": "LLM360 empowers a community-driven approach to artificial general intelligence (AGI) through open-source large language models (LLMs). We provide comprehensive resources for ethical AGI development, ensuring accessibility and transparency in AI research.",
"publisher": {
"@type": "Organization",
"name": "LLM360",
"logo": {
"@type": "ImageObject",
"url": "https://www.llm360.ai/images/logo_webheader.png",
"width": 160,
"height": 160
},
"sameAs": [
"https://x.com/llm360",
"https://github.com/LLM360"
]
},
"image": {
"@type": "ImageObject",
"url": "https://www.llm360.ai/images/open-graph-image.jpg",
"width": 1920,
"height": 1080
}
}
</script>
<!-- /SEO -->
<!-- Favicon -->
<link rel="icon" type="image/x-icon" href="assets/favicon/favicon.ico" />
<link rel="icon" type="image/png" sizes="192x192" href="assets/favicon/android-chrome-192x192.png">
<link rel="icon" type="image/png" sizes="512x512" href="assets/favicon/android-chrome-512x512.png">
<link rel="icon" type="image/png" sizes="32x32" href="assets/favicon/favicon-32x32.png">
<link rel="icon" type="image/png" sizes="16x16" href="assets/favicon/favicon-16x16.png">
<link rel="apple-touch-icon" sizes="180x180" href="assets/favicon/apple-touch-icon.png">
<!-- <link rel="manifest" href="/site.webmanifest"> -->
<meta name="msapplication-TileColor" content="#da532c">
<meta name="theme-color" content="#ffffff">
<!-- Style -->
<link rel="stylesheet" href="assets/css/main.css" />
</head>
<body class="is-preload">
<!-- Sidebar -->
<button class="toggle-btn" id="toggleBtn">☰</button>
<section id="sidebar" class="sidebar">
<div class="inner">
<nav>
<a class="alt" href="index.html">
<figure class="hover-rotate">
<img src="images/logo-highres.png" alt="logo" />
</figure>
</a>
<h2>LLM360</h2>
<ul>
<li><a href="#one">Models</a></li>
<li><a href="evaluation.html">Performance and Evaluation</a></li>
<li><a href="#two">LLM360 Suites</a></li>
<li><a href="#three">Papers</a></li>
<li><a href="#four">Blogs</a></li>
<li><a href="#six">Get in touch</a></li>
<li><a href="community.html">Open-source Communities</a></li>
<li><a href="about.html">About</a></li>
</ul>
</nav>
</div>
</section>
<!-- Wrapper -->
<div id="wrapper">
<!-- Intro -->
<section id="intro" class="wrapper style1 fullscreen fade-up">
<div class="alert-banner" id="alertBanner">
<a href="blog/several-new-releases-to-further-our-mission.html" target="_blank">Announcing K2-65B: Learn More Here.</a>
</div>
<div class="inner">
<h1><strong>LLM360</strong> enables <strong>community-owned AGI</strong> through <strong>open-source large model</strong> research and development.</h1>
<ul class="actions">
<li><a href="https://twitter.com/llm360" target="_blank" class="icon brands circle fa-twitter"><span class="label">Twitter</span></a></li>
<li><a href="https://github.com/LLM360" target="_blank" class="icon brands circle fa-github"><span class="label">Github</span></a></li>
<li><a href="mailto:[email protected]" target="_blank" class="icon circle fa-envelope"><span class="label">Email</span></a></li>
</ul>
</div>
</section>
<!-- One -->
<section id="one" class="wrapper style2 spotlights">
<div class="inner">
<h1>Our models</h1>
</div>
<section class="pictures-section">
<a href="#k2" class="picture-link">
<img src="images/k2_logo.png" alt="" class="picture">
<h2>K2-65B</h2>
</a>
<a href="#crystal" class="picture-link">
<img src="images/cc_logo.png" alt="Image 2" class="picture">
<h2>CrystalCoder-7B</h2>
</a>
<a href="#amber" class="picture-link">
<img src="images/amber_logo.png" alt="Image 3" class="picture">
<h2>Amber-7B</h2>
</a>
</section>
<section id="k2">
<span class="image">
<img src="images/k2_logo.png" alt=""/>
</span>
<div class="content">
<div class="inner">
<h2>K2-65B</h2>
<p>A <strong>65B parameter</strong> language model trained on <strong>1.4T tokens</strong>. It outperforms <strong>Llama 2 70B</strong>, but uses approximately <strong>35% less</strong> compute to train.</p>
<section>
<ul class="actions fit" >
<li><a href="https://huggingface.co/collections/LLM360/k2-6622ae6911e3eb6219690039" target="_blank" class="button style1 fit">Model</a></li>
<li><a href="https://wandb.ai/llm360/K2?nw=29mu6l0zzqq" target="_blank" class="button style1 fit">W&B</a></li>
<li><a href="https://huggingface.co/datasets/LLM360/K2Datasets" target="_blank" class="button style1 fit">Data</a></li>
<li><a href="https://github.com/LLM360/k2-train" target="_blank" class="button style1 fit">Code</a></li>
</ul>
</section>
</div>
</div>
</section>
<section id="crystal">
<span class="image">
<img src="images/cc_logo.png" alt="" />
</span>
<div class="content">
<div class="inner">
<h2>CrystalCoder-7B</h2>
<p> A 7B parameter language model, distinctively trained on the SlimPajama and StarCoder datasets,
eclipsing the <strong>Llama 2</strong> frontier, skillfully <strong>balances</strong> language and coding.
Its instruction-following variant, <a href="https://huggingface.co/LLM360/CrystalChat" target="_blank">CrystalChat</a>, stands out as a <strong>top-scoring</strong> 7B chat model, trained on a carefully selected mix publicly available language and code datasets.</p>
<ul class="actions fit">
<li><a href="https://huggingface.co/collections/LLM360/crystal-65e733d14e6a0786c4f5a606" target="_blank" class="button style2 fit">Model</a></li>
<li><a href="https://wandb.ai/llm360/CrystalCoder?nw=hdze3lfpuer" target="_blank" class="button style2 fit">W&B</a></li>
<li><a href="https://huggingface.co/datasets/LLM360/CrystalCoderDatasets" target="_blank" class="button style2 fit">Data</a></li>
<li><a href="https://github.com/LLM360" target="_blank" class="button style2 fit">Code</a></li>
</ul>
</div>
</div>
</section>
<section id="amber">
<span class="image">
<img src="images/amber_logo.png" alt="" />
</span>
<div class="content">
<div class="inner">
<h2>Amber-7B</h2>
<p>A 7B parameter English language model based on the <strong>LLaMA</strong> architecture has two fine-tuned instruction-following models named <a href="https://huggingface.co/LLM360/AmberChat" target="_blank">AmberChat</a> and <a href="https://huggingface.co/LLM360/AmberSafe" target="_blank">AmberSafe</a>.</p>
<ul class="actions fit">
<li><a href="https://huggingface.co/collections/LLM360/amber-65e7333ff73c7bbb014f2f2f" target="_blank" class="button style3 fit">Model</a></li>
<li><a href="https://wandb.ai/llm360/Amber/workspace?nw=lnzi8o2g4z" target="_blank" class="button style3 fit">W&B</a></li>
<li><a href="https://huggingface.co/datasets/LLM360/AmberDatasets" target="_blank" class="button style3 fit">Data</a></li>
<li><a href="https://github.com/LLM360" target="_blank" class="button style3 fit">Code</a></li>
</ul>
</div>
</div>
</section>
</section>
<!-- Two -->
<section id="two" class="wrapper style3 fade-up">
<div class="inner">
<h2>LLM360 Suites</h2>
<div class="features three">
<section>
<h3>LLM360 Research Suite</h3>
<p>The Research Suite is a comprehensive set of large language model (LLM) artifacts from each of our models, for academic and industry researchers to explore LLM training dynamics.</p>
<ul class="actions">
<li><a href="research.html" class="button bottom">Learn more</a></li>
</ul>
</section>
<section>
<h3>LLM360 Pretraining Suite</h3>
<p>The Pretraining Suite is a series of step-by-step guides to reproduce each of our models, for tech enthusiasts, AI practitioners, and academic or industry researchers, to transfer knowledge on LLM pretraining techniques.</p>
<ul class="actions">
<li><a href="pretraining.html" class="button bottom">Learn more</a></li>
</ul>
</section>
<section>
<h3>LLM360 Developer Suite</h3>
<p>The Developer Suite is a series of fine-tuning and inference tutorials for tech enthusiasts, AI practitioners, and academic or industry researchers, who are interested in general model usage or downstream task evaluation and research.</p>
<ul class="actions">
<li><a href="developer.html" class="button bottom">Learn more</a></li>
</ul>
</section>
</div>
</div>
</section>
<!-- Three -->
<section id="three" class="wrapper style4 fade-up">
<div class="inner">
<h2>Papers</h2>
<div class="features alt">
<section>
<h3>LLM360 K2-65B: Scaling Up Fully Transparent Open-Source LLMs</h3>
<span class="image right">
<img src="images/paper2.png">
</span>
<p>In this paper, we present LLM360 K2-65B, the most powerful fully transparent open-source large language model (LLM) released to date. K2 is a 65 billion parameter LLM, which follows best practices for reproducibility from the LLM360 project. Despite numerous efforts to develop and release open-source LLMs, full transparency around the training process still remains limited... </p>
<ul class="actions">
<li><a href="paper2.pdf" target="_blank" class="button">Learn more</a></li>
</ul>
</section>
<section>
<span class="image right">
<img src="images/llm360_paper.png">
</span>
<h3>LLM360: Towards Fully Transparent Open-Source LLMs</h3>
<p>The recent surge in open-source Large Language Models (LLMs), such as LLaMA,
Falcon, and Mistral, provides diverse options for AI practitioners and researchers.
However, most LLMs have only released partial artifacts, such as the final model
weights or inference code, and technical reports increasingly limit their scope to
high-level design choices and surface statistics... </p>
<ul class="actions">
<li><a href="paper.pdf" target="_blank" class="button">Learn more</a></li>
</ul>
</section>
</div>
<section>
<h3>Inspired Research:</h3>
<ul>
<li>
<a href="https://arxiv.org/pdf/2402.19465.pdf" target="_blank">Towards Tracing Trustworthiness Dynamics: Revisiting Pre-training Period of Large Language Models</a>
</li>
<li>
<a href="https://arxiv.org/pdf/2401.12255.pdf" target="_blank">Instructional Fingerprinting of Large Language Models</a>
</li>
</ul>
</section>
</div>
</section>
<!-- Four -->
<section id="four" class="wrapper style3 fade-up">
<div class="inner">
<h2>Blogs</h2>
<div class="features alt">
<section>
<span class="image right">
<img src="images/blog2.png">
</span>
<h3>Introducing K2-65B: Charting the Blueprint Towards Open-Source Artificial General Intelligence</h3>
<p>LLM360 is excited to announce several new releases to further our mission enabling community-owned AGI by creating standards and tools to advance the bleeding edge of LLM capability and empower knowledge transfer, research, and development.</p>
<ul class="actions">
<li><a href="blog/several-new-releases-to-further-our-mission.html" class="button">Learn more</a></li>
</ul>
</section>
<section>
<span class="image right">
<img src="images/blog.png">
</span>
<h3>Introducing LLM360: Fully Transparent Open-Source LLMs</h3>
<p>In recent months, the open-source large language model (LLM) community has seen tremendous model contributions. However, model weight releases and overview technical reports do not contain enough information to cover the complexity of LLM training, which hinders openness and transparency, the mechanisms behind trustworthy and innovative research and science for decades.</p>
<ul class="actions">
<li><a href="blog/introducing-llm360-fully-transparent-open-source-llms.html" class="button">Learn more</a></li>
</ul>
</section>
</div>
</div>
</section>
<!-- Four -->
<section id="six" class="wrapper style4 fade-up">
<div class="inner">
<h2>Connect with Us</h2>
<p>The LLM360 team is here to solve the most challenging AI problems. Reach out if you'd like to discuss.</p>
<section>
<form method="post" action="https://formspree.io/f/xqkvrzyy" name="Contact Form">
<div class="fields">
<div class="field half">
<input type="text" name="name" id="name" placeholder="Name" required/>
</div>
<div class="field half">
<input type="text" name="email" id="email" placeholder="Email" required/>
</div>
<div class="field">
<textarea name="message" id="message" rows="5" placeholder="Message" required></textarea>
</div>
</div>
<ul class="actions">
<li><a href="" class="button submit">Send Message</a></li>
<li><input type="reset" value="Reset" /></li>
</ul>
</form>
</section>
</div>
</section>
</div>
<!-- Footer -->
<footer id="footer" class="wrapper style1-alt">
<div class="inner">
<ul class="menu">
<p>
LLM360, proudly sponsored by Petuum and MBZUAI, is dedicated to advancing the field of AI by providing comprehensive access to large language models.<br>
LLM360 enables community-owned AGI by creating standards and tools to advance the bleeding edge of LLM capability and empower knowledge transfer, research, and development.
</p>
<ul class="actions">
<li><a href="https://twitter.com/llm360" target="_blank" class="icon brands circle fa-twitter"><span class="label">Twitter</span></a></li>
<li><a href="https://github.com/LLM360" target="_blank" class="icon brands circle fa-github"><span class="label">Github</span></a></li>
<li><a href="mailto:[email protected]" target="_blank" class="icon circle fa-envelope"><span class="label">Email</span></a></li>
</ul>
<p class="copyright">© LLM360 2023-2024. All rights reserved.</p>
</ul>
</div>
</footer>
<!-- Scripts -->
<script src="assets/js/jquery.min.js"></script>
<script src="assets/js/jquery.scrollex.min.js"></script>
<script src="assets/js/jquery.scrolly.min.js"></script>
<script src="assets/js/browser.min.js"></script>
<script src="assets/js/breakpoints.min.js"></script>
<script src="assets/js/util.js"></script>
<script src="assets/js/main.js"></script>
</body>
</html>