-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathindex.html
More file actions
667 lines (575 loc) · 25 KB
/
index.html
File metadata and controls
667 lines (575 loc) · 25 KB
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
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta name="description"
content="VPGTrans">
<meta name="keywords" content="VPGTrans, open-source, vision-language, LLM">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>VPGTrans</title>
<!-- Global site tag (gtag.js) - Google Analytics
<script async src="https://www.googletagmanager.com/gtag/js?id=G-PYVRSFMDRL"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag() {
dataLayer.push(arguments);
}
gtag('js', new Date());
gtag('config', 'G-PYVRSFMDRL');
</script> -->
<meta name="google-site-verification" content="6lbYN1vX7A4sD8SrVniq84UEKyEUSBgxeP7d3FjuuK0" />
<link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro"
rel="stylesheet">
<link rel="stylesheet" href="./static/css/bulma.min.css">
<link rel="stylesheet" href="./static/css/bulma-carousel.min.css">
<link rel="stylesheet" href="./static/css/bulma-slider.min.css">
<link rel="stylesheet" href="./static/css/fontawesome.all.min.css">
<link rel="stylesheet"
href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<link rel="icon" href="images/icon.png">
<link rel="stylesheet" href="./static/css/index.css">
<link rel="shortcut icon" href="images/icon.png" type="image/x-icon">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script defer src="./static/js/fontawesome.all.min.js"></script>
<script src="./static/js/bulma-carousel.min.js"></script>
<script src="./static/js/bulma-slider.min.js"></script>
<script src="./static/js/index.js"></script>
</head>
<style>
#main{
position: relative;;
width: 1200px;
}
.box{
float: left;
padding: 15px 0 0 15px;
/* background-color: red;*/
}
.pic{
width: 500px;
padding: 10px;
border: 1px solid #ccc;
border-radius: 5px;
background-color: #fff;
}
.pic img{
width: 500px;
}
</style>
<body>
<section class="hero">
<div class="hero-body">
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column has-text-centered">
<h1 class="title is-1 publication-title">VPGTrans:</h1>
<h2 class="title is-2 publication-title">Transfer Visual Prompt Generator across LLMs</h2>
<div class="is-size-5">
<span class="author-block">
<a href="https://waxnkw.github.io/" style="color:#008AD7;font-weight:normal;">Ao Zhang<sup>1</sup>
</a>,
</span>
<span class="author-block">
<a href="https://haofei.vip" style="color:#008AD7;font-weight:normal;">Hao Fei<sup>1*</sup></a>,</span>
<span class="author-block">
<a href="https://yaoyuanthu.github.io/" style="color:#008AD7;font-weight:normal;">Yuan Yao<sup>2*</sup></a>,
</span>
<span class="author-block">
<a href="https://jiwei0523.github.io/" style="color:#008AD7;font-weight:normal;">Wei Ji<sup>1</sup></a>,
</span>
<span class="author-block">
<a href="#" style="color:#008AD7;font-weight:normal;">Li Li<sup>1</sup></a>,
</span>
<span class="author-block">
<a href="http://nlp.csai.tsinghua.edu.cn/~lzy/" style="color:#008AD7;font-weight:normal;">Zhiyuan Liu<sup>2</sup></a>,
</span>
<span class="author-block">
<a href="https://www.chuatatseng.com/" style="color:#008AD7;font-weight:normal;">Tat-Seng Chua<sup>1</sup></a>
</span>
</div>
<br>
<div class="is-size-5 publication-authors">
<span class="author-block"><b style="color:#F2A900; font-weight:normal">▶ </b>1. Sea-NExT Joint Lab, National University of Singapore </span>
<br>
<span class="author-block"><b style="color:#00A4EF; font-weight:normal">▶ </b>2. Department of Computer Science and Technology, Tsinghua University </span>
<!-- <span class="author-block"><b style="color:#00A4EF; font-weight:normal">▶ </b>Microsoft Research, Redmond; </span> -->
<!-- <span class="author-block"><b style="color:#008AD7; font-weight:normal">▶ </b>Microsoft Cloud & AI </span> -->
</div>
<div class="is-size-5 publication-authors">
<span class="author-block"><sup>*</sup>Correspondence</span>
<!-- <span class="author-block"><sup>☨</sup>Equal Advisory Contribution, </span> -->
<!-- <span class="author-block"><sup>⚑</sup>Project Lead </span> -->
</div>
<br>
<!-- <div class="is-size-5 publication-authors">
<span class="author-block"><b style="color:#e08ba0; font-weight:normal"> <b>In CVPR2023</b> </b></span>
</div> -->
<div class="column has-text-centered">
<div class="publication-links">
<span class="link-block">
<a href="https://arxiv.org/abs/2305.01278" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="ai ai-arxiv"></i>
</span>
<span>Paper</span>
</a>
</span>
<span class="link-block">
<a href="https://github.com/VPGTrans/VPGTrans" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fab fa-github"></i>
</span>
<span>Code</span>
</a>
</span>
<!-- <span class="link-block">-->
<!-- <a href="https://huggingface.co/spaces/##" target="_blank"-->
<!-- class="external-link button is-normal is-rounded is-dark">-->
<!-- <span class="icon">-->
<!-- 🤗-->
<!-- </span>-->
<!-- <span>Space</span>-->
<!-- </a>-->
<!-- </span>-->
<script>
window.addEventListener('load', function() {
const urls = [
// 'https://8a2f5fddb53d0d87fb.gradio.live', // dgx02-07
// 'https://adabddb02eb6cc5112.gradio.live', // dgx01-00
// 'https://1a42399a0aab566720.gradio.live', // dgx02-06
'https://7116ee36952d2ef342.gradio.live', // dgx02-07
// 'https://73aa8417f578aa1573.gradio.live', // dgx01-05
// 'https://83a25cef578d89c67d.gradio.live', // dgx01-06
// 'https://60ac677ed79d2da927.gradio.live',
// 'https://c9cd51f7cae3c9fec1.gradio.live',
// 'https://f2283aee1e9ab52ea9.gradio.live',
// 'https://ecf463e1131cfb7099.gradio.live',
// 'https://d6a81e24b072749a4d.gradio.live',
// 'https://4da94f3035b0462a03.gradio.live',
];
const randomIndex = Math.floor(Math.random() * urls.length);
const randomURL = urls[randomIndex];
document.getElementById('randomLink').href = randomURL;
});
</script>
<span class="link-block">
<a id="randomLink" href="https://7116ee36952d2ef342.gradio.live" target="_blank" class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fa fa-play"></i>
</span>
<span>Demo</span>
</a>
</span>
<!-- <span class="link-block">-->
<!-- <a href="https://youtu.be/###" target="_blank"-->
<!-- class="external-link button is-normal is-rounded is-dark">-->
<!-- <span class="icon">-->
<!-- <i class="fab fa-youtube"></i>-->
<!-- </span>-->
<!-- <span>Video</span>-->
<!-- </a>-->
<!-- </span>-->
<!-- <span class="link-block">
<a href="https://huggingface.co/datasets/Vision-CAIR/cc_sbu_align" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fa fa-database"></i>
</span>
<span>Dataset</span>
</a>
</span> -->
<!-- <span class="link-block">-->
<!-- <a href="https://huggingface.co/###" target="_blank"-->
<!-- class="external-link button is-normal is-rounded is-dark">-->
<!-- <span class="icon">-->
<!-- <i class="fa fa-laugh"></i>-->
<!-- </span>-->
<!-- <span>Model</span>-->
<!-- </a>-->
<!-- </span>-->
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<!-- <section class="section">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<div class="content has-text-justified">
<p>
Thanks for your interest in our work. Currently, the number of users has exceeded our expectations. We provide <strong><font color="#008AD7">alternative demo links</font></strong> here:
<a href="https://b2517615b965687635.gradio.live" target="_blank">Demo1</a>
<a href="https://c8de8ff74b6a6c6a9b.gradio.live" target="_blank">Demo2</a>
<a href="https://90bc0bac96e6457e8f.gradio.live" target="_blank">Demo3</a>
<a href="https://cd772059965a71f9e6.gradio.live" target="_blank">Demo4</a>
<a href="https://48da7e23bcadec7551.gradio.live" target="_blank">Demo5</a>
<a href="https://687d119023cd37e5fb.gradio.live" target="_blank">Demo6</a>
<a href="https://0810e8582bcad31944.gradio.live" target="_blank">Demo7</a>
<a href="https://31c7cdb7e3594e851e.gradio.live" target="_blank">Demo8</a>
<strong><font>News</font></strong>: We now provide a pretrained MiniGPT-4 aligned with <strong><font color="#008AD7">Vicuna-7B</font></strong>! The demo GPU memory consumption now can be <strong><font color="#008AD7">as low as 12GB</font></strong>.
<br>
</p>
</div>
</div>
</div>
</section>
-->
<script>
window.addEventListener('load', function() {
const urls = [
// 'https://8a2f5fddb53d0d87fb.gradio.live', // dgx02-07
// 'https://adabddb02eb6cc5112.gradio.live', // dgx01-00
// 'https://1a42399a0aab566720.gradio.live', // dgx02-06
'https://7116ee36952d2ef342.gradio.live', // dgx02-07
// 'https://73aa8417f578aa1573.gradio.live', // dgx01-05
// 'https://83a25cef578d89c67d.gradio.live', // dgx01-06
// 'https://60ac677ed79d2da927.gradio.live',
// 'https://c9cd51f7cae3c9fec1.gradio.live',
// 'https://f2283aee1e9ab52ea9.gradio.live',
// 'https://ecf463e1131cfb7099.gradio.live',
// 'https://d6a81e24b072749a4d.gradio.live',
// 'https://4da94f3035b0462a03.gradio.live',
];
const randomIndex = Math.floor(Math.random() * urls.length);
const randomURL = urls[randomIndex];
const iframe = document.getElementById('gradio');
iframe.setAttribute('src', randomURL);
});
</script>
<iframe id="gradio" width="100%" height="950">
<p>Gradio.</p>
</iframe>
<!-- <link rel="stylesheet" href="js/ft-carousel.css" />
<script src="js/jquery.min.js"></script>
<script src="js/ft-carousel.min.js"></script>
<script type="text/javascript">
$("#carousel_1").FtCarousel();
$("#carousel_2").FtCarousel({
index: 1,
auto: false
});
$("#carousel_3").FtCarousel({
index: 0,
auto: true,
time: 3000,
indicators: false,
buttons: true
});
</script> -->
<!-- <section class="hero teaser">
<div class="container is-max-desktop">
<div class="example">
<div class="ft-carousel" id="carousel_1">
<ul class="carousel-inner"> -->
<!-- <li class="carousel-item"><img src="demos/wop_2.png" /></li>
<li class="carousel-item"><img src="demos/cook_1.png" /></li>
<li class="carousel-item"><img src="demos/fix_1.png" /></li>
<li class="carousel-item"><img src="demos/rhyme_1.png" /></li> -->
<!-- <li class="carousel-item"><img src="img/a1.jpg" /></li>
<li class="carousel-item"><img src="img/a2.jpg" /></li>
<li class="carousel-item"><img src="img/a3.jpg" /></li>
<li class="carousel-item"><img src="img/a4.jpg" /></li>
<li class="carousel-item"><img src="img/a5.jpg" /></li>
<li class="carousel-item"><img src="img/a6.jpg" /></li>
</ul>
</div>
</div>
</div>
</section> -->
<!--
<section class="hero is-small">
<div class="hero-body">
<div class="container">
<div id="results-carousel" class="carousel results-carousel">
<div class="item">
<img src="demos/wop_2.png" alt="MY ALT TEXT"/>
<h2 class="subtitle has-text-centered">
</h2>
</div>
<div class="item">
<img src="demos/cook_1.png" alt="MY ALT TEXT"/>
<h2 class="subtitle has-text-centered">
</h2>
</div>
<div class="item">
<img src="demos/fix_1.png" alt="MY ALT TEXT"/>
<h2 class="subtitle has-text-centered">
</h2>
</div>
<div class="item">
<img src="demos/rhyme_1.png" alt="MY ALT TEXT"/>
<h2 class="subtitle has-text-centered">
</h2>
</div>
</div>
</div>
</div>
</section>
-->
<link rel="stylesheet" type="text/css" href="js/simple_style.css" />
<script type="text/javascript" src="js/simple_swiper.js"></script>
<!-- <div class="app">
<div id="swiper-demo" class="simple-swiper-container">
<a id="prev" class="btn btn-prev"></a>
<a id="next" class="btn btn-next"></a>
<div class="pagination"></div>
</div>
</div>
<p id="index"></p>
<script type="text/javascript">
new SimSwiper("#swiper-demo", {
autoplay: 4000,
duration: 300,
easing: 'ease',
button: {
prev: "#prev", // 前进后退按钮
next: "#next"
},
pagination: {
el: '.pagination',
click: true// 是否可以点击
},
// 轮播图数据
data: [{
index: 0,
href: '#',
src: 'demos/wop_2.png'
}, {
index: 1,
href: '#',
src: 'demos/cook_1.png'
}, {
index: 2,
href: '#',
src: 'demos/fix_1.png'
}, {
index: 3,
href: '#',
src: 'demos/rhyme_1.png'
}]
});
</script> -->
<section class="section">
<div class="container is-max-desktop">
<!-- Abstract. -->
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Abstract</h2>
<div class="content has-text-justified">
<p>
While developing a new vision-language LLM (VL-LLM) by pre-training on tremendous image-text pairs from scratch can be exceedingly resource-consuming, connecting an existing LLM with a comparatively lightweight visual prompt generator (VPG) becomes a feasible paradigm.
However, further tuning the VPG part of the VL-LLM still suffers from indispensable computational costs, i.e., requiring thousands of GPU hours and millions of training data.
One alternative solution is to transfer an existing VPG from any existing VL-LLMs for the target VL-LLM.
<br>
<br>
In this work, we for the first time investigate the VPG transferability across LLMs, and explore a solution to reduce the cost of VPG transfer.
We first study the VPG transfer across different LLM sizes (e.g., small-to-large), and across different LLM types, through which we diagnose the key factors to maximize the transfer efficiency.
Based on our observation, we design a two-stage transfer framework named <b>VPGTrans</b>, which is simple yet highly effective.
Through extensive experiments, we demonstrate that VPGTrans helps significantly speed up the transfer learning process without compromising performance.
Remarkably, it helps achieve the VPG transfer from BLIP-2 OPT<sub>2.7B</sub> to BLIP-2 OPT<sub>6.7B</sub> with over 10 times speed-up and 10.7% training data compared with connecting a VPG to OPT<sub>6.7B</sub> from scratch.
Further, a series of intriguing findings and potential rationales behind them are provided and discussed.
Finally, we showcase the application value of our VPGTrans approach, by newly customizing two novel VL-LLMs, including <b>VL-LLaMA</b> and <b>VL-Vicuna</b>, with recently released LLaMA and Vicuna models.
<br>
<br>
With our VPGTrans, one can build a novel high-performance VL-LLM based on any existing text-based LLMs and an existing VPG with considerably lower cost. For example, the BLIP-2 traing cost can be reduced strikingly:
<!-- Our work will show great potential in the multimodal LLM community. -->
<br>
<!-- <br>
The
</b> -->
</p>
</div>
</div>
</div>
<!--/ Abstract. -->
<img id="model" width="100%" src="images/cost.png">
<!-- <h3 class="subtitle has-text-centered">
<p style="font-family:Times New Roman"></p>
</h3> -->
<br>
<br>
<!-- Paper video. -->
<!-- <div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Video</h2>
<div class="publication-video">
<iframe width=“560” height=“315" src=“https://www.youtube.com/embed/__tftoxpBAw” title=“YouTube video player” frameborder=“0” allow=“accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share” allowfullscreen></iframe>
</div>
</div>
</div> -->
<!-- <section class="hero is-small is-light">
<div class="hero-body"> -->
<!--
<div class="container">
<h2 class="title has-text-centered">Video Presentation</h2>
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<div class="publication-video">
<iframe width="560" height="315" src="https://www.youtube.com/embed/__tftoxpBAw" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen></iframe>
</div>
</div>
</div>
</div>
-->
<!-- </div>
</section> -->
<!--/ Demo. -->
<!-- <br>
<br>
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Demo</h2>
</div>
</div>
<div class="column is-full-width">
<div class="columns is-centered">
<img id="teaser" width="90%" src="images/demo6_AdobeExpress.gif">
</div>
<div class="columns is-centered">
<h1>
<p style="font-family:Times New Roman"><b>X-GPT: Connecting generalist X-Decoder with GPT-3</b>
</h1>
</div>
</div>
<br>
<div class="column is-full-width">
<div class="columns is-centered">
<img id="teaser" width="90%" src="images/inpaint.gif">
</div>
<div class="columns is-centered">
<h1>
<p style="font-family:Times New Roman"><b>Instruct-X-Decoder: Object-centric instructional image editing</b>
</h1>
</div>
</div> -->
<!--/ Paper video. -->
<br>
<br>
<!-- Paper Model. -->
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths">
<h2 class="title is-3">Method</h2>
<div class="content has-text-justified">
<p>
<b>The general architecture of VL-LLMs and the VPG transfer</b>:
</p>
<ul>
<!-- <li>It has two types of queries (latent queries and text queries) and outputs (semantic outputs and pixel-level outputs).</li>
<li>It uses a single text encoder for all text corpus, ranging from class concepts, referring phrases to image captions.</li>
<li>It decouples image and text encoder to accomadate cross-image tasks (e.g., image-text retrieval) and within-image tasks (e.g., segmentation and captioning).</li> -->
</ul>
</div>
<img id="model" width="100%" src="images/Intro.png">
<!-- <h3 class="subtitle has-text-centered">
<p style="font-family:Times New Roman"></p>
</h3> -->
<br>
<br>
<div class="content has-text-justified">
<p>
<b>Our proposed two-stage VPGTrans Framework</b>:
</p>
</div>
<img id="model" width="100%" src="images/VPGTrans.png">
<br>
<br>
</div>
</div>
<br>
<br>
<!--/ Paper video. -->
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths">
<h2 class="title is-3">Experiments</h2>
<div class="content has-text-justified">
<p>
<b>Our built BLIP-2 models</b>:
</p>
<ul>
<!-- <li>It has two types of queries (latent queries and text queries) and outputs (semantic outputs and pixel-level outputs).</li>
<li>It uses a single text encoder for all text corpus, ranging from class concepts, referring phrases to image captions.</li>
<li>It decouples image and text encoder to accomadate cross-image tasks (e.g., image-text retrieval) and within-image tasks (e.g., segmentation and captioning).</li> -->
</ul>
</div>
<img id="model" width="80%" src="images/blip2.png">
<!-- <h3 class="subtitle has-text-centered">
<p style="font-family:Times New Roman"></p>
</h3> -->
<br>
<br>
<div class="content has-text-justified">
<p>
<b>Our built VL-LLaMA</b>:
</p>
</div>
<img id="model" width="60%" src="images/vl-llama.png">
<br>
<br>
</div>
</div>
<br>
<br>
</div>
</section>
<section class="section" id="BibTeX">
<div class="container is-max-desktop content">
<div class="columns is-centered has-text-centered">
<h2 class="title is-3">BibTeX</h2>
</div>
<pre><code>
@article{2023vpgtrans,
author = {Ao Zhang and Hao Fei and Yuan Yao and Wei Ji and Li Li and Zhiyuan Liu and Tat-Seng Chua},
title = {Transfer Visual Prompt Generator across LLMs},
journal = {CoRR},
volume = {abs/23045.01278},
year = {2023},
url = {https://doi.org/10.48550/arXiv.2305.01278},
}
</code></pre>
</div>
</section>
<section class="section">
<!-- Results. -->
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths">
<h2 class="title is-3">VL-Vicuna Demo</h2>
</div>
</div>
<!--/ Results. -->
<!-- <div class="container is-max-desktop">
</section> -->
<!-- <section class="section"> -->
<div id="main">
<div class="box"><div class="pic"><img src="demos/p1.png" alt=""></div></div>
<div class="box"><div class="pic"><img src="demos/p2.png" alt=""></div></div>
<div class="box"><div class="pic"><img src="demos/p3.png" alt=""></div></div>
<div class="box"><div class="pic"><img src="demos/p4.png" alt=""></div></div>
<div class="box"><div class="pic"><img src="demos/p5.png" alt=""></div></div>
<div class="box"><div class="pic"><img src="demos/p6.png" alt=""></div></div>
<div class="box"><div class="pic"><img src="demos/p7.png" alt=""></div></div>
<div class="box"><div class="pic"><img src="demos/p8.png" alt=""></div></div>
<div class="box"><div class="pic"><img src="demos/p9.png" alt=""></div></div>
</div>
</section>
<!--
<section class="section" id="Acknowledgement">
<div class="container is-max-desktop content">
<h2 class="title is-3">Acknowledgement</h2>
<p>
This website is adapted from <a
href="https://github.com/nerfies/nerfies.github.io">Nerfies</a>, licensed under a <a rel="license"
href="http://creativecommons.org/licenses/by-sa/4.0/">Creative
Commons Attribution-ShareAlike 4.0 International License</a>.
</p>
</div>
</section>
-->
<script src="js/Underscore-min.js"></script>
<script src="js/index.js"></script>
</body>
</html>