-
Notifications
You must be signed in to change notification settings - Fork 5
/
index.html
181 lines (169 loc) · 7.71 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
<!doctype html>
<html lang="en">
<head>
<!-- Google tag (gtag.js) -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-NBPBX93MFR"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'G-NBPBX93MFR');
</script>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="icon" class="js-site-favicon" type="image/png" href="logo.svg">
<title>Petals – Run LLMs at home, BitTorrent-style</title>
<link href="./assets/dist/css/bootstrap.min.css" rel="stylesheet">
<style>
@media (min-width: 1025px) {
.subscribe-form {
padding: 0 100px;
}
}
</style>
<link href="product.css?v=3" rel="stylesheet">
</head>
<body>
<main>
<div class="position-relative overflow-hidden px-5 m-md-4 text-center">
<div class="col-xl-6 col-lg-8 p-md-3 mx-auto my-1" style="max-width: 700px;">
<div>
<div class="d-md-inline-block pe-md-4">
<img src="logo.svg" height="100" class="rot-image">
</div>
<h1 class="display-4 fw-bold d-md-inline-block justify-content-center" style="font-size: 45pt; vertical-align: middle;">Petals</h1>
</div>
<p class="lead fw-normal mt-3 mb-4 pt-3">
Run large language models at home, BitTorrent‑style
</p>
<ul class="feature-list subscription-offer pt-2">
<li class="mb-3">
Generate text with <b>Llama 3.1</b> (up to 405B), <b>Mixtral</b> (8x22B), <b>Falcon</b> (40B+) or <b>BLOOM</b> (176B)
and fine‑tune them for your tasks — using a consumer-grade GPU or Google Colab.
</li>
<li class="mb-3">
You load a part of the model, then join a
<a target="_blank" href="https://health.petals.dev">network</a>
of people serving its other parts.
Single‑batch inference runs at up to <b>6 tokens/sec</b> for <b>Llama 2</b> (70B) and
up to <b>4 tokens/sec</b> for <b>Falcon</b> (180B) — enough for
<a target="_blank" href="https://chat.petals.dev">chatbots</a> and interactive apps.
</li>
<li class="mb-3">
Beyond classic LLM APIs —
you can employ any fine-tuning and sampling methods, execute custom paths through the model, or see its hidden states.
You get the comforts of an API with the flexibility of <b>PyTorch</b> and 🤗 <b>Transformers</b>.
</li>
</ul>
<div class="subscription-success" style="display: none;">
<h5 class="fw-bold pt-4">
Thanks for subscribing!
</h5>
<p class="fw-normal py-4">
We will email you only if we have really exciting updates.
</p>
</div>
<div class="mb-1">
<a target="_blank" class="btn btn-dark mt-2 mb-4" href="https://colab.research.google.com/drive/1uCphNY7gfAUkdDrTx21dZZwCOUDCMPw8?usp=sharing">Try now in Colab</a>
<a target="_blank" class="btn btn-outline-dark ms-4 mt-2 mb-4" href="https://github.com/bigscience-workshop/petals">Docs on GitHub</a>
</div>
<p class="fw-normal pt-3" style="font-size: 11pt;">
<b>Top contributors</b> right now:
</p>
<div class="fw-normal top-contributors" style="font-size: 10.5pt;">
<span class="loading gray">Loading...</span>
<span class="error" style="display: none;">Oops, can't load the network status...</span>
<span class="contents"></span>
<span class="etc" style="display: none;">
<span class="bull">•</span>
and <a target="_blank" href="http://health.petals.dev">more</a>
</span>
</div>
<div class="mt-3">
<a target="_blank" class="btn btn-outline-dark mt-2 mb-4" href="https://github.com/bigscience-workshop/petals#connect-your-gpu-and-increase-petals-capacity">
Contribute my GPU
</a>
</div>
<div class="subscription-offer mb-4">
<p class="fw-normal pt-3" style="font-size: 10.5pt;">
Follow development in <a target="_blank" href="https://discord.gg/D9MwApKgWa">Discord</a> or via email:
</p>
<form class="d-flex col-xl-8 mx-auto">
<input class="form-control me-2" type="email" name="email" id="email" placeholder="Leave your email" style="font-size: 11pt;">
<button class="btn btn-outline-dark my-0" type="submit">Subscribe</button>
</form>
<div class="email-promise mt-2" style="color: #999; font-size: 9.5pt;">
We send updates once a few months. No spam.
</div>
<div class="loading-feedback mt-2" style="display: none; font-size: 9.5pt;">
Submitting...
</div>
<div class="invalid-feedback mt-2" style="font-size: 9.5pt;"></div>
<div class="valid-feedback mt-2" style="font-size: 9.5pt;">
We sent you an email to confirm your address. Click it and you're in!
</div>
</div>
<p class="acknowledgements fw-normal pt-3">
Featured on:
</p>
<div class="mb-3">
<a target="_blank" href="https://techcrunch.com/2022/12/20/petals-is-creating-a-free-distributed-network-for-running-text-generating-ai/">
<img src="techcrunch.png" width="150">
</a>
</div>
<p class="acknowledgements fw-normal pt-4">
This project is a part of the <a target="_blank" href="https://bigscience.huggingface.co/">BigScience</a> research workshop.
</p>
<div>
<a target="_blank" href="https://bigscience.huggingface.co/"><img src="bigscience.png" width="125"></a>
</div>
</div>
</div>
</main>
<script src="https://code.jquery.com/jquery-3.3.1.min.js" crossorigin="anonymous"></script>
<script src="./assets/dist/js/bootstrap.bundle.min.js"></script>
<script>
$(() => {
$(window).on("hashchange", () => {
if (window.location.hash == "#success") {
$(".subscription-offer").hide();
$(".subscription-success").show();
}
});
$(window).trigger("hashchange");
$(".subscription-offer form").submit(event => {
event.preventDefault();
$(".subscription-offer .form-control").removeClass("is-valid is-invalid");
$(".email-promise, .invalid-feedback, .valid-feedback").hide();
$(".loading-feedback").show();
$.post('https://buttondown.email/api/emails/embed-subscribe/petals', $('.subscription-offer form').serialize())
.done(() => {
$(".loading-feedback").hide();
$(".subscription-offer .form-control").addClass("is-valid");
$(".valid-feedback").show();
})
.fail((request, status, message) => {
$(".loading-feedback").hide();
$(".subscription-offer .form-control").addClass("is-invalid");
if (message === "") {
$(".invalid-feedback").text("Oops, something has failed. Please check your Internet connection");
} else {
$(".invalid-feedback").text("Oops, something has failed (" + message + ")");
}
$(".invalid-feedback").show();
});
});
$.get("https://health.petals.dev/")
.done(data => {
$(".top-contributors .loading").hide();
$(".top-contributors .contents").html($(data).find(".top-contributors").html());
$(".top-contributors .etc").show();
})
.fail((request, status, message) => {
$(".top-contributors .loading").hide();
$(".top-contributors .error").show();
});
});
</script>
</body>
</html>