29:
712:
695:
148:
248:, while WuDao has 1.75 trillion parameters; making it the largest pre-trained model in the world at the time. WuDao was trained on 4.9 terabytes of images and texts (which included 1.2 terabytes of Chinese text and 1.2 terabytes of English text). The chairman of BAAI said that WuDao was an attempt to "create the biggest, most powerful AI model possible"; although direct comparisons between models based on parameter count (i.e. between Wu Dao and GPT-3) do not directly correlate to quality.
192:(AI) research laboratory. BAAI conducts AI research and is dedicated to promoting collaboration among academia and industry, as well as fostering top talent and a focus on long-term research on the fundamentals of AI technology. As a collaborative hub, BAAI's founding members include leading AI companies, universities, and research institutes. BAAI is one of pre-eminent AI research institutes in China. To help it reach its goals, BAAI frequently releases new models and
724:
255:
and image recognition, in addition to generation of text and images. The model can not only write essays, poems and couplets in traditional
Chinese, it can both generate text based on static images and generate nearly photorealistic images based on natural language descriptions. It has also showed
287:
is an AI-computing platform which focuses on supporting AI innovation. As of
September 2022 it provides 1000P computation capacity with 400 Gbit/s high-speed interconnection per server, and support AI chipsets of different architectures. BAAI's platform also includes code compilers for the
271:
is an open-source extensible toolkit for large-scale model training and inference. Its goal is to support training, fine-tuning, and deployment of large-scale models on various downstream tasks with multi-modality. Moreover, its open repository includes not only all source-code, but several
636:
303:(C. elegans) nematode simulating the worm's nervous system along with a "digital body" simulation in real-time. MetaWorm 1.0 exhibits behaviours that parallel C. elegans in the real world.
457:
196:
code. Moreover, BAAI organizes an annual international conference bringing together AI experts, industry leaders, and international talent to discuss challenges and future of AI.
379:
737:
776:
781:
540:
312:
580:
245:
599:
361:
219:
164:
28:
516:
252:
702:
425:
665:
189:
39:
300:
204:
As of 2023, BAAI's research focuses on large pre-trained models (LLMs) and open-source AI infrastructure.
186:
562:
407:
618:
498:
517:"Can A.I. help investors find the next hot technology? A number of companies are betting on it"
272:
pre-trained large models. FlagAI is an approved incubation project at the sandbox level of the
728:
273:
694:
147:
70:
458:"BAAI 2022: 'AI Life', Multimodal Models, AI for Science, Autonomous Driving and More!"
237:
770:
430:
484:
256:
ability to power virtual chat agents and predict the 3D structures of proteins like
716:
563:"China Says WuDao 2.0 AI Is an Even Better Conversationalist than OpenAI, Google"
193:
752:
739:
332:
257:
637:"June 30, 2022 Meeting of the LF AI & Data Technical Advisory Council"
83:
541:"Beijing-funded AI language model tops Google and OpenAI in raw numbers"
227:
172:
97:
240:. WuDao 2.0, was announced on 31 May 2022 and has been compared to
241:
213:
619:"China's 'Wu Dao' AI is 10X bigger than GPT-3, and it can sing"
581:"Cambrian-AI 2022 Predictions: Expect More Than Just New Chips"
666:"Wriggling abiotic worm manifests "intelligent life" rudiment"
600:"Meet Wu Dao 2.0, the Chinese AI model making the West sweat"
688:
136:
380:"Nature Index - Institution Collaborations/Relationships"
499:"Record-breaking AI model WuDao 2.0 unveiled in Beijing"
244:
at the time. But, in comparison, GPT-3 has 175 billion
362:"China's gigantic multi-modal AI is no one-trick pony"
700:
408:"This Chinese Lab Is Aiming for Big AI Breakthroughs"
131:
115:
90:
79:
63:
45:
35:
22:Beijing Academy of Artificial Intelligence (BAAI)
8:
333:"Beijing Academy of Artificial Intelligence"
21:
693:
251:WuDao has demonstrated ability to perform
178:Běijīng Zhìyuán réngōng zhìnéng yánjiùyuàn
157:Beijing Academy of Artificial Intelligence
146:
20:
313:Artificial intelligence industry in China
452:
450:
448:
707:
556:
554:
534:
532:
530:
402:
400:
324:
485:"面向认知,智源研究院联合多家单位发布超大规模新型预训练模型"悟道·文汇""
777:Chinese companies established in 2018
355:
353:
351:
349:
7:
782:Artificial intelligence laboratories
236:) is a large multimodal pre-trained
426:"The race of the AI labs heats up"
14:
598:Heikkilä, Melissa (9 June 2021).
722:
710:
299:is a computational model of the
27:
617:Greene, Tristan (3 June 2021).
561:Mott, Nathaniel (3 June 2021).
232:
223:
177:
168:
16:Chinese AI research laboratory
1:
360:Tarantola, A. (2 June 2021).
121:
288:different AI architectures.
253:natural language processing
798:
539:Feng, Coco (2 June 2021).
211:
51:; 5 years ago
483:智源研究院 (11 January 2021).
200:Products and applications
26:
545:South China Morning Post
190:artificial intelligence
40:Artificial intelligence
301:Caenorhabditis elegans
644:LF AI Data Foundation
749: /
117:Number of employees
23:
670:China.org / Xinhua
587:. 20 January 2022.
414:. 21 January 2021.
49:November 2018
623:TNW | Neural
183:Zhiyuan Institute
181:), also known as
154:
153:
789:
764:
763:
761:
760:
759:
754:
753:39.99°N 116.32°E
750:
747:
746:
745:
742:
727:
726:
725:
715:
714:
713:
706:
697:
692:
691:
689:Official website
674:
673:
662:
656:
655:
653:
651:
641:
633:
627:
626:
614:
608:
607:
595:
589:
588:
577:
571:
570:
558:
549:
548:
536:
525:
524:
513:
507:
506:
495:
489:
488:
480:
474:
473:
471:
469:
454:
443:
442:
440:
438:
422:
416:
415:
404:
395:
394:
392:
390:
376:
370:
369:
357:
344:
343:
341:
339:
329:
274:Linux Foundation
234:
225:
179:
170:
150:
145:
142:
140:
138:
126:
123:
86:, Beijing, China
59:
57:
52:
31:
24:
797:
796:
792:
791:
790:
788:
787:
786:
767:
766:
757:
755:
751:
748:
743:
740:
738:
736:
735:
733:
723:
721:
711:
709:
701:
687:
686:
683:
678:
677:
664:
663:
659:
649:
647:
639:
635:
634:
630:
616:
615:
611:
597:
596:
592:
579:
578:
574:
560:
559:
552:
538:
537:
528:
515:
514:
510:
497:
496:
492:
482:
481:
477:
467:
465:
456:
455:
446:
436:
434:
424:
423:
419:
406:
405:
398:
388:
386:
378:
377:
373:
359:
358:
347:
337:
335:
331:
330:
326:
321:
309:
294:
282:
266:
216:
210:
202:
185:, is a Chinese
135:
124:
118:
111:
75:
71:Zhang Hongjiang
55:
53:
50:
17:
12:
11:
5:
795:
793:
785:
784:
779:
769:
768:
732:
731:
719:
699:
698:
682:
681:External links
679:
676:
675:
672:. 2 June 2022.
657:
646:. 30 June 2022
628:
609:
590:
572:
567:Tom's Hardware
550:
526:
508:
505:. 3 June 2021.
490:
475:
444:
417:
396:
371:
345:
323:
322:
320:
317:
316:
315:
308:
305:
293:
290:
281:
278:
265:
262:
238:language model
212:Main article:
209:
206:
201:
198:
152:
151:
133:
129:
128:
119:
116:
113:
112:
110:
109:
106:
103:
100:
94:
92:
88:
87:
81:
77:
76:
74:
73:
67:
65:
61:
60:
47:
43:
42:
37:
33:
32:
15:
13:
10:
9:
6:
4:
3:
2:
794:
783:
780:
778:
775:
774:
772:
765:
762:
758:39.99; 116.32
730:
720:
718:
708:
704:
696:
690:
685:
684:
680:
671:
667:
661:
658:
645:
638:
632:
629:
624:
620:
613:
610:
605:
601:
594:
591:
586:
582:
576:
573:
568:
564:
557:
555:
551:
546:
542:
535:
533:
531:
527:
522:
518:
512:
509:
504:
500:
494:
491:
486:
479:
476:
464:. 7 June 2022
463:
459:
453:
451:
449:
445:
433:
432:
431:The Economist
427:
421:
418:
413:
409:
403:
401:
397:
385:
381:
375:
372:
367:
363:
356:
354:
352:
350:
346:
334:
328:
325:
318:
314:
311:
310:
306:
304:
302:
298:
291:
289:
286:
279:
277:
275:
270:
263:
261:
259:
254:
249:
247:
243:
239:
235:
229:
221:
215:
207:
205:
199:
197:
195:
191:
188:
184:
180:
174:
166:
162:
158:
149:
144:
134:
130:
120:
114:
107:
104:
101:
99:
96:
95:
93:
89:
85:
82:
78:
72:
69:
68:
66:
62:
48:
44:
41:
38:
34:
30:
25:
19:
734:
669:
660:
648:. Retrieved
643:
631:
622:
612:
603:
593:
584:
575:
566:
544:
520:
511:
503:China.org.cn
502:
493:
478:
466:. Retrieved
461:
435:. Retrieved
429:
420:
411:
387:. Retrieved
383:
374:
365:
336:. Retrieved
327:
296:
295:
284:
283:
268:
267:
250:
231:
217:
203:
182:
176:
160:
156:
155:
80:Headquarters
18:
756: /
194:open source
169:北京智源人工智能研究院
771:Categories
319:References
246:parameters
187:non-profit
125: 100
729:Companies
258:AlphaFold
744:116°19′E
604:Politico
437:28 March
389:28 March
366:Engadget
338:28 March
307:See also
297:MetaWorm
292:MetaWorm
108:MetaWorm
91:Products
84:Wudaokou
64:Founders
36:Industry
741:39°59′N
703:Portals
650:6 April
521:Fortune
468:6 April
285:Jiuding
280:Jiuding
220:Chinese
218:WuDao (
165:Chinese
132:Website
105:Jiuding
56:2018-11
54: (
46:Founded
585:Forbes
462:Synced
384:Nature
269:FlagAI
264:FlagAI
230::
228:pinyin
222::
175::
173:pinyin
167::
127:(2023)
102:FlagAI
717:China
640:(PDF)
412:WIRED
242:GPT-3
233:wùdào
214:WuDao
208:WuDao
139:.baai
98:WuDao
652:2023
470:2023
439:2023
391:2023
340:2023
161:BAAI
163:) (
143:.cn
141:.ac
137:www
773::
668:.
642:.
621:.
602:.
583:.
565:.
553:^
543:.
529:^
519:.
501:.
460:.
447:^
428:.
410:.
399:^
382:.
364:.
348:^
276:.
260:.
226:;
224:悟道
171:;
122:c.
705::
654:.
625:.
606:.
569:.
547:.
523:.
487:.
472:.
441:.
393:.
368:.
342:.
159:(
58:)
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.