-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
538 lines (474 loc) · 23.7 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
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
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta name="description" content="Chimera">
<meta name="keywords" content="Large Multi-modal Models, multi-modal reasoning, visual structural extraction">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Chimera</title>
<link rel="stylesheet" href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/bulma@0.9.1/css/bulma.min.css">
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/4.5.2/css/bootstrap.min.css">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/5.15.1/css/all.min.css">
<link rel="stylesheet" href="./static/css/index.css">
<link rel="icon" href="./static/pure_logo.png">
<link href="https://fonts.googleapis.com/icon?family=Material+Icons" rel="stylesheet">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script defer src="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/5.15.1/js/all.min.js"></script>
<script type="module" src="https://gradio.s3-us-west-2.amazonaws.com/4.16.0/gradio.js"></script>
<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="stylesheet" href="./static/css/index.css">
<link rel="stylesheet" href="./static/css/leaderboard.css">
<!-- <link href="https://unpkg.com/tabulator-tables@5.5.2/dist/css/tabulator_bulma.min.css" rel="stylesheet">
<script type="text/javascript" src="https://unpkg.com/tabulator-tables@5.5.2/dist/js/tabulator.min.js"></script> -->
<script type="text/javascript" src="static/js/sort-table.js" defer></script>
<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/explorer-index.js"></script>
<script src="./static/js/question_card.js"></script>
<script src="./static/js/leaderboard_testmini.js"></script>
<script src="./data/results/output_folders.js" defer></script>
<script src="./data/results/model_scores.js" defer></script>
<script src="./visualizer/data/data_public.js" defer></script>
<style>
.center-container {
display: flex;
justify-content: center;
align-items: center;
height: 100%;
margin-top: -20px;
}
.node {
fill: #f8f1e4;
stroke: #000;
stroke-width: 1;
rx: 10;
ry: 10;
}
.node text {
font-size: 14px;
text-anchor: middle;
}
.link {
fill: none;
stroke: #000;
stroke-width: 2;
}
.badge {
font-size: 12px;
}
</style>
<style>
/* Basic styling for the bar */
.navbar {
background-color: #333;
overflow: hidden;
}
.navbar a {
float: left;
display: block;
color: #f2f2f2;
text-align: center;
padding: 14px 16px;
text-decoration: none;
}
.navbar a:hover {
background-color: #ddd;
color: black;
}
</style>
</head>
<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">Chimera: <span class="is-size-2">Improving Generalist Model with <br>Domain-Specific Experts</span></h1>
<h5 class="subtitle is-5 publication-awards has-text-centered">Arxiv 2024</h5> -->
<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">Chimera: <span class="is-size-2">Improving Generalist Model with <br>Domain-Specific Experts</span></h1>
<div class="is-size-5 publication-authors">
<span class="author-block">
<a href="https://pengts.github.io/" style="color:#f68946;font-weight:normal;">Tianshuo Peng<sup>1,2,*</sup></a>,
</span>
<span class="author-block">
<a href="https://msheng-lee.github.io/" style="color:#008AD7;font-weight:normal;">Mingsheng Li<sup>3,*</sup></a>,
</span>
<span class="author-block">
<a href="https://scholar.google.com/citations?user=fb_WgAEAAAAJ&hl=en&oi=sra" style="color:#F2A900;font-weight:normal;">Hongbin Zhou<sup>1</sup></a>,
</span>
<span class="author-block">
<a href="https://scholar.google.com/citations?user=E520fqQAAAAJ&hl=zh-CN" style="color:#F2A900;font-weight:normal;">Renqiu Xia<sup>1,4</sup></a>,
</span>
<span class="author-block">
<a href="https://zrrskywalker.github.io/" style="color:#F2A900;font-weight:normal;">Renrui Zhang<sup>2</sup></a>,
</span>
<span class="author-block">
<a href="http://leibai.site/" style="color:#F2A900;font-weight:normal;">Lei Bai<sup>1</sup></a>,
</span>
<span class="author-block">
<a href="https://scholar.google.com/citations?user=BaqGkQQAAAAJ&hl=zh-CN" style="color:#F2A900;font-weight:normal;">Song Mao<sup>1</sup></a>,
</span>
<span class="author-block">
<a href="https://wangbindl.github.io/" style="color:#F2A900;font-weight:normal;">Bin Wang<sup>1</sup></a>,
</span>
<span class="author-block">
<a href="https://conghui.github.io/" style="color:#F2A900;font-weight:normal;">Conghui He<sup>1</sup></a>,
</span>
<span class="author-block">
<a href="https://scholar.google.com.hk/citations?user=cC8lXi8AAAAJ&hl=en" style="color:#F2A900;font-weight:normal;">Aojun Zhou<sup>2</sup></a>,
</span>
<span class="author-block">
<a href="https://scholar.google.com/citations?user=K0PpvLkAAAAJ&hl=en" style="color:#F2A900;font-weight:normal;">Botian Shi<sup>1</sup></a>,
</span>
<span class="author-block">
<a href="https://eetchen.github.io/" style="color:#F2A900;font-weight:normal;">Tao Chen<sup>3</sup></a>,
</span>
<span class="author-block">
<a href="https://bobrown.github.io/boZhang.github.io/" style="color:#F2A900;font-weight:normal;">Bo Zhang<sup>1,✉,‡</sup></a>,
</span>
<span class="author-block">
<a href="https://www.ie.cuhk.edu.hk/faculty/yue-xiangyu/" style="color:#F2A900;font-weight:normal;">Xiangyu Yue<sup>2,✉</sup></a>,
</span>
</div>
<div class="is-size-5 publication-authors">
<span class="author-block"><sup>1</sup>Shanghai Artificial Intelligence Laboratory,</span>
<span class="author-block"><sup>2</sup>MMLab, The Chinese University of Hong Kong,</span><br>
<span class="author-block"><sup>3</sup>Fudan University,</span>
<span class="author-block"><sup>4</sup>Shanghai Jiao Tong University</span><br>
</div>
<div class="is-size-5 publication-authors">
<sup>*</sup> Equal contribution, <sup>✉</sup> Corresponding author, <sup>‡</sup> Project Leader
</div>
<!-- <div class="is-size-5 publication-authors">
<span class="author-block"><b style="color:#f68946; font-weight:normal">▶ </b> University of
Wisconsin-Madison</b></span>
<span class="author-block"><b style="color:#008AD7; font-weight:normal">▶ </b> Microsoft Research</span>
<span class="author-block"><b style="color:#F2A900; font-weight:normal">▶ </b>Columbia
University</span>
</div>
-->
<!-- <div class="column has-text-centered">
<h3 class="title is-3 publication-title">Improved Baselines with Visual Instruction Fine-tuning</h3>
<div class="is-size-5 publication-authors">
<span class="author-block">
<a href="https://hliu.cc/" style="color:#f68946;font-weight:normal;">Haotian Liu<sup>*</sup></a>,
</span>
<span class="author-block">
<a href="https://chunyuan.li/" style="color:#008AD7;font-weight:normal;">Chunyuan Li<sup>*</sup></a>,
</span>
<span class="author-block">
<a href="https://yuheng-li.github.io" style="color:#008AD7;font-weight:normal;">Yuheng Li</a>,
</span>
<span class="author-block">
<a href="https://pages.cs.wisc.edu/~yongjaelee/" style="color:#f68946;font-weight:normal;">Yong Jae
Lee</a>
</span>
</div>
<div class="is-size-5 publication-authors">
<span class="author-block"><b style="color:#f68946; font-weight:normal">▶ </b> University of
Wisconsin-Madison</b></span>
<span class="author-block"><b style="color:#008AD7; font-weight:normal">▶ </b> Microsoft Research</span>
</div> -->
<div class="column has-text-centered">
<div class="publication-links">
<span class="link-block">
<a href="https://arxiv.org/abs/2412.05983" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="ai ai-arxiv"></i>
</span>
<span>ArXiv (Chimera-1.0)</span>
</a>
</span>
<span class="link-block">
<a href="https://github.com/UniModal4Reasoning/Chimera" 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://llava.hliu.cc" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="far fa-images"></i>
</span>
<span>Demo</span>
</a>
</span> -->
<!-- <span class="link-block">
<a href="https://huggingface.co/datasets/liuhaotian/LLaVA-Instruct-150K" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fas fa-database"></i>
</span>
<span>Dataset</span>
</a>
</span> -->
<span class="link-block">
<a href="https://huggingface.co/collections/U4R/chimera-10-6749542e2f0dfa09414232c0" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fas fa-share-square"></i>
</span>
<span>Model</span>
</a>
</span>
<!-- <span class="link-block">
<a href="#"
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> -->
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<section class="hero teaser">
<div class="container is-max-desktop">
<div class="hero-body">
<h4 class="subtitle">
🔥<span style="color: #ff3860">[NEW!]</span>
Chimera-Reasoner-8B gets 64.9 on MathVista, achieves new SoTA under 10B scale models! <br>
🔥<span style="color: #ff3860">[NEW!]</span>
Chimera-Extractor demonstrates powerful extraction performance on various types of documents! <br>
🔥<span style="color: #ff3860">[NEW!]</span>
<a href="https://huggingface.co/collections/U4R/chimera-10-6749542e2f0dfa09414232c0">Weights</a> & <a href="https://github.com/UniModal4Reasoning/Chimera">Inference code</a> have been released！
</h4>
</div>
</div>
</section>
<section class="section">
<div class="container is-max-desktop">
<!-- Abstract. -->
<div class="columns is-centered has-text-centered">
<div class="column is-full">
<h2 class="title is-3">Abstract</h2>
<div class="content has-text-justified">
<p>
Recent advancements in Large Multi-modal Models (LMMs) underscore the importance of scaling by increasing image-text paired data, achieving impressive performance on general tasks.
Despite their effectiveness in broad applications, generalist models are primarily trained on web-scale datasets dominated by natural images, resulting in the sacrifice of specialized capabilities for domain-specific tasks that require extensive domain prior knowledge.
Moreover, directly integrating expert models tailored for specific domains is challenging due to the representational gap and imbalanced optimization between the generalist model and experts.
To address these challenges, we introduce <b>Chimera</b>, a scalable and low-cost multi-modal pipeline designed to boost the ability of existing LMMs with domain-specific experts.
Specifically, we design a progressive training strategy to integrate features from expert models into the input of a generalist LMM.
To address the imbalanced optimization caused by the well-aligned general visual encoder, we introduce a novel Generalist-Specialist Collaboration Masking (GSCM) mechanism.
This results in a versatile model that excels across the chart, table, math, and document domains, achieving state-of-the-art performance on multi-modal reasoning and visual content extraction tasks, both of which are challenging tasks for assessing existing LMMs.
We will release Chimera's weights, along with the data used for training and evaluation, to facilitate future research on LMMs.
</p>
</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-full">
<h2 class="title is-3">Method</h2>
</div>
</div>
</div>
<!--/ Results. -->
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column is-full-width">
<div class="content has-text-justified">
<p>
We introduce Chimera, a <b>scalable</b> pipeline that integrates specialist models into generalist LMMs, facilitating their adaptation to diverse specialized tasks.
Chimera comprises a general visual encoder, a general projector paired with a language model pre-trained from an LMM, alongside a router and an expert model suite, which includes specialized expert models and their corresponding expert projectors.
Chimera uses a <b>Generalist-Specialist Collaboration Masking</b> (GSCM) mechanism to facilitate the alignment with expert models.
<br>
We consider a progressive two-stage training procedure:
<ul type="1">
<li><b>Stage 1: Domain-General Knowledge Alignment</b>. <span style="font-size: 95%;">To initially align domain-specific knowledge with the semantic space of the generalist LMM, we train the model using tasks that directly perceive diverse image content. We only train the general projector and expert projectors during this stage.</span></li>
<li><b>Stage 2: Visual Instruction Tuning</b>. <span style="font-size: 95%;">To further enhance the performance of Chimera on specialized tasks, we perform visual instruction tuning on different domain-specific tasks with the proposed GSCM. All projectors and LLM are updated during this stage.</span></li>
</ul>
Please check out our
<a href="https://huggingface.co/collections/U4R/chimera-10-6749542e2f0dfa09414232c0">[Model Zoo]</a>.
</p>
</div>
<centering>
<div style="text-align: center;">
<img id="teaser" width="70%" src="static/overview.png">
</div>
</centering>
</div>
</div>
</section>
<section class="section">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-full">
<h2 class="title is-3">Performance</h2>
</div>
</div>
</div>
<!--/ Results. -->
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column is-full-width">
<div class="content has-text-justified">
<p>
We conduct quantitative experiments to evaluate Chimera's capabilities in multi-modal reasoning and visual content extraction.
Chimera achieves a new SOTA for LMMs of comparable scale on two multi-modal reasoning benchmarks.
It also surpasses or matches the performance of representative expert models in visual content extraction tasks across chart, table, and document domains.
</p>
</div>
<centering>
<div style="text-align: center;">
<img id="teaser" width="70%" src="static/reasoning_mathvista.png">
</div>
<br>
<div style="text-align: center;">
<img id="teaser" width="70%" src="static/reasoning_mathverse.png">
</div>
<br>
<div style="text-align: center;">
<img id="teaser" width="70%" src="static/extraction_all.png">
</div>
</centering>
</div>
</div>
</section>
<section class="section">
<div class="container" style="margin-bottom: 2vh;">
<!-- Abstract. -->
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Examples</h2>
<div class="content has-text-justified">
<p>
In addition to quantitatively reporting Chimera's performance across various benchmarks, we also provide several demos below to showcase Chimera's capabilities on challenging domain-specific tasks, such as table format transformation, chart structural extraction, and document context extraction.
</p>
<div id="results-carousel" class="carousel results-carousel">
<div class="box m-5">
<div class="content has-text-centered">
<img src="static/demo_1.png" alt="algebraic reasoning" width="100%"/>
<p> Output of Chimera-Reasoner-8B on Table Format Transformation.
</div>
</div>
<div class="box m-5">
<div class="content has-text-centered">
<img src="static/demo_2.png" alt="arithmetic reasoning" width="80%"/>
<p> Output of Chimera-Reasoner-8B on Table Format Transformation.
</div>
</div>
<div class="box m-5">
<div class="content has-text-centered">
<img src="static/demo_3.png" alt="arithmetic reasoning" width="100%"/>
<p> Output of Chimera-Reasoner-8B on Chart Structural Extraction.
</div>
</div>
<div class="box m-5">
<div class="content has-text-centered">
<img src="static/demo_4.png" alt="arithmetic reasoning" width="100%"/>
<p>
Output of Chimera-Reasoner-8B on Chart Structural Extraction.
</div>
</div>
<div class="box m-5">
<div class="content has-text-centered">
<img src="static/demo_5.png" alt="arithmetic reasoning" width="100%"/>
<p>
Output of Chimera-Extractor-1B on Document Context Extraction.</div>
</div>
<div class="box m-5">
<div class="content has-text-centered">
<img src="static/demo_6.png" alt="arithmetic reasoning" width="80%"/>
<p>
Output of Chimera-Extractor-1B on Document Context Extraction.</div>
</div>
</div>
</div>
<!--/ Abstract. -->
</div>
</section>
<section class="section" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title">BibTeX</h2>
<pre><code>
@misc{peng2024chimeraimprovinggeneralistmodel,
title={Chimera: Improving Generalist Model with Domain-Specific Experts},
author={Tianshuo Peng and Mingsheng Li and Hongbin Zhou and Renqiu Xia and Renrui Zhang and Lei Bai and Song Mao and Bin Wang and Conghui He and Aojun Zhou and Botian Shi and Tao Chen and Bo Zhang and Xiangyu Yue},
year={2024},
eprint={2412.05983},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2412.05983},
}
</code></pre>
</div>
</section>
<footer class="footer">
<div class="container">
<div class="content has-text-centered">
<a class="icon-link"
href="https://arxiv.org/abs/2412.05983">
<i class="fas fa-file-pdf"></i>
</a>
<a class="icon-link" href="https://github.com/UniModal4Reasoning/Chimera" class="external-link" disabled>
<i class="fab fa-github"></i>
</a>
</div>
<div class="columns is-centered">
<div class="column is-8">
<div class="content">
<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>
</div>
</div>
</div>
</footer>
<script src="https://code.jquery.com/jquery-3.5.1.js"></script>
<script src="https://cdn.datatables.net/1.10.24/js/jquery.dataTables.min.js"></script>
<script src="https://cdn.datatables.net/1.10.24/js/dataTables.bootstrap4.min.js"></script>
<script>
$(document).ready(function () {
$('#nr3d_table').DataTable({
columnDefs: [ // disable sorting of the first column
{orderable: false, targets: 0}
],
"order": [[1, "desc"]], // Default Ordering on the overall column
searching: false, // Disable search bar
paging: false, // Stop pagination for now
info: false // hide "Showing 1 to M of N entries" line
});
$('#sr3d_table').DataTable({
columnDefs: [ // disable sorting of the first column
{orderable: false, targets: 0}
],
"order": [[1, "desc"]], // Default Ordering on the overall column
searching: false, // Disable search bar
paging: false, // Stop pagination for now
info: false // hide "Showing 1 to M of N entries" line
});
})
</script>
</body>
</html>