AIGC, the new driving force for architecture and business | Registration for TF offline event on September 16

The 117th issue of CCF TF "AIGC, the new driving force for architecture and business" event was hosted by CCF TF Architecture SIG and co-organized by Sina. It invited many Internet technology experts from iFlytek, Alibaba Cloud, Qunar.com, Sina Weibo, etc. Experts share and communicate how to drive the efficient development of architecture and business through the application of AIGC technology.

Provide top communication platform for engineers
CCF TF Issue 117

f5abb3d3915eab1bf5d4460b6dc934eb.jpeg

Time: September 16, 2023 14:00-17:30
Location: Roundtable Conference Hall on the second floor of Sina Headquarters Building, Phase II, Zhongguancun Software Park, Haidian District, Beijing
Theme: AIGC, the new driving force for architecture and business

Since this year, AIGC technology has become popular all over the world, and its widespread application is also profoundly changing all walks of life. How to better apply AIGC technology to promote faster business development is also an important issue facing the field of technical architecture.

The 117th issue of CCF TF "AIGC, the new driving force for architecture and business" event was hosted by CCF TF Architecture SIG and co-organized by Sina. It invited many Internet technology experts from iFlytek, Alibaba Cloud, Qunar.com, Sina Weibo, etc. Experts share and communicate how to drive the efficient development of architecture and business through the application of AIGC technology.

The rise of AIGC and large models has led to the emergence of many excellent large model application frameworks. These frameworks make the business implementation of large models more efficient and convenient. Yang Yanbo, an AI technology expert from iFlytek, will bring "Large Model Application Development Framework and Usage" Case", sharing the principles of Langchain/AutoGPT, as well as Autospark core architecture and business implementation cases.

Large model training capabilities are the foundation for the development of AIGC technology. The network of high-performance intelligent computing clusters is also facing unprecedented challenges and opportunities. Cong Peiyan, an architect from Alibaba Cloud, will bring "Lingjun Predictable Network Practical Experience Sharing" to share how Utilize high-performance network architecture based on Ethernet and cloud services to ensure large model training efficiency.

With the rapid development and application of AIGC technology, new production methods have emerged. Zhang Yu, a technical expert from Qunar.com, will bring "AIGC Basic Capability Platform Construction and Implementation Practice" to share how to build a unified platform to reduce costs. The usage threshold allows more business personnel to apply AIGC technology more conveniently and explore more valuable implementation scenarios.

In the field of AIGC code generation, many large vertical models and related classic application cases have emerged. Hu Yunpeng, a technical expert from Sina Weibo, will share "Weibo code assistant WeCode helps improve R&D efficiency" and introduce how Weibo uses WeCode Build a Weibo code generation platform and use AIGC technology to improve the R&D efficiency of the technical team.

This event covers AIGC large model training infrastructure optimization practice sharing, AIGC basic capability platform construction practice introduction, and also code generation platform construction practice sharing in the AIGC vertical field.

Welcome to sign up to participate in the interaction and learn and communicate with technical experts from iFlytek, Alibaba Cloud, Qunar.com, and Sina Weibo.

Meeting schedule

TF117 : AIGC , the new driving force for architecture and business

Moderator: Li Qingfeng,  Chairman of CCF TF Architecture SIG , Senior Director of Weibo R&D Center

time

theme

Speaker

14:00-14:05

activity description

Li Qingfeng

CCF TF Structure SIG Chairman

Senior Director of Weibo R&D Center

14:05-14:15

Event speech

Wang Wei Sina Weibo COO

Wang Xinxia Deputy Secretary-General  of CCF

14:15-14:55

"Large Model Application Development Framework and Use Cases"

Yang Yanbo

Senior R&D Engineer at IFLYTEK AI Engineering Institute

14:55-15:35

 "Lingjun Predictable Network Practical Experience Sharing"

Cong Peiyan

Alibaba Cloud Lingjun Product Architect

15:35-16:15

" AIGC Basic Capability Platform Construction and Implementation Practice"

Zhang Yu

Qunar.com content understanding team leader

16:15-16:55

"Weibo Code Assistant WeCode Helps R&D Efficiency Improvement Practice"

Hu Yunpeng

Director of Weibo’s distributed middleware direction

16:55-17:15

Panel

17:15-17:30

Activity summary

Yang Weihua

CCF TF Chairman & Architecture SIG Chairman Founder and CEO of Westar Labs

Affiliated SIG

CCF TF Architecture SIG

guest speaker

474c9be14b8413a0656895eef00a295e.png


Yang Yanbo, Senior R&D Engineer, iFlytek AI Engineering Institute, HKUST

Topic: "Large Model Application Development Framework and Use Cases"

Topic introduction:
1. Background of large model projects: current development status at home and abroad, problems existing in the implementation of large models;
2. Development of open source agent application framework: introduction to the development status of large model application framework, Langchain/AutoGPT principle, open source AI agent application framework The inspiration brought by it;
3. AutoSpark introduction & principle: autospark panorama, autospark principle;
4. Application cases: implementation cases;
5. Summary.

Personal profile: Core contributor of the open source project sealos, core contributor of the open source project AIGES; one of the first batch of Kubernetes users in China. Loves open source and is often active in open source communities such as K8s, Sealos, Volcano, and Argo. The person in charge of iFlytek's open source is mainly responsible for incubating iFlytek's excellent internal projects into open source projects.

f4f77915de0681dbebc4d97e7afd6cdb.jpeg

Cong Peiyan,
Product Architect, Alibaba Cloud Lingjun

Topic: "Sharing Practical Experience of Lingjun Predictable Network"

Topic introduction: With large model training in full swing, the network of high-performance intelligent computing clusters is also facing unprecedented new challenges and development opportunities. Ethernet's diverse and open ecosystem, rapidly evolving forwarding performance and network scale, and best practices in cloud services are the key cornerstones for ensuring the long-term development and success of intelligent computing clusters. This report will focus on three aspects: changes and challenges brought by large models to AI infrastructure, technology selection: Infiniband or Ethernet, and Lingjun predictable network: high-performance network infrastructure built for AI.

Personal profile: Cong Peiyan, product architect of Alibaba Cloud Lingjun, focuses on the planning and design of high-performance infrastructure products for AI intelligent computing scenarios. He has provided ultra-large-scale intelligent computing cluster full-stack solutions for many large internal and external customers and successfully implemented them.

6387f5efafc9d1aaeaf87718810ddc99.jpeg

Zhang Yu,
head of content understanding team at Qunar.com

Topic: "AIGC Basic Capability Platform Construction and Implementation Practice"

Topic introduction: With the rapid development of AIGC, new production methods have emerged. By building a unified platform and lowering the threshold for use, more business personnel can explore valuable implementation scenarios based on platform capabilities.

Personal profile: Zhang Yu, head of the content understanding team, joined Qunar in 2017. His main research directions are in the fields of NLP and CV. He led the team to build a general content understanding platform and made significant contributions in scenarios such as intelligent review and intelligent customer service. Currently, Mainly responsible for AIGC related work.

148be555239b9bc55f5f0b34008d2424.jpeg

Hu Yunpeng
Weibo Distributed Middleware Director

Topic: "Weibo Code Assistant WeCode Helps R&D Efficiency Improvement Practice"

Topic introduction: In the vertical field of large model code generation, various large models continue to emerge. Weibo has also launched its own code assistant wecode, allowing more developers to use AIGC to improve research and development efficiency. In practice, it has also stepped on I have accumulated some pitfalls and accumulated some experiences to share with you.

Personal profile: Hu Yunpeng is the person in charge of the distributed middleware direction of Weibo. He joined Weibo in 2017. His main research direction is basic component performance optimization and other fields, including hybrid cloud platform, monitoring platform, distributed registration center, DataMesh and other system platforms. Construction, currently mainly responsible for the work related to the AIGC code generation platform project.

SIG Chairman

511626305a4a085aa603e2adffe514b1.jpeg

Yang Weihua,
Chairman of CCF TF & Chairman of Architecture SIG, Founder and CEO of Westar Labs

Personal profile: Yang Weihua, founder and CEO of Westar Labs. He has been involved in the technical architecture work of Sina Weibo since 2009. He has served as the chief architect and technical vice president of Weibo and has long led the research and development of Weibo infrastructure and platform. He is currently the Chairman of CCF TF & Chairman of Architecture SIG, and an executive member of CCF Blockchain Committee.

94eaf757f2c6b489cf183dbb88cd04a0.jpeg

Li Qingfeng,
Chairman of CCF TF Architecture SIG, Senior Director of Weibo R&D Center

Personal profile: Senior Director of Weibo R&D Center, responsible for Weibo infrastructure, streaming media and other research and development directions. He has extensive R&D and management experience in high-availability architecture, microservices, hybrid cloud architecture, and streaming media architecture technology. A supporter of the technology community, he has served as a lecturer and producer at many cutting-edge technology conferences in the industry.

Instructions for attending the meeting

1. The number of places in the event is limited. If you are unable to participate after registration, the fee will not be refunded.

2. Please read the article in detail for the event location and time. Event notification will be sent via email or text message one day before the event . Please follow the email information to sign in and participate on time.

3. The fee for this event is 400 yuan/person for members and 800 yuan/person for non-members . You can enjoy discounts by becoming a member. If you have any questions, please contact us by email: [email protected]

member benefits

Members can participate in CCF TF's 20 online events throughout the year for free, and participate in 14 offline events at discounted prices. It is a good investment for their own technology growth and an excellent way to obtain professional knowledge with high cost performance!

  • Professional member/senior member/distinguished member/fellowship: 360 yuan/year

  • Student membership: 50 yuan/year. Please click to view specific rights and interests: CCF individual member rights

  • Apply for company membership to enjoy more free quotas, brand promotion and other rights. Click to view details: CCF company member rights or consultation hotline: 0512-65900856 ext. 27

Registration contact information

September 16, 2023 (Saturday) 14:00-17:30
Venue: Roundtable Conference Hall on the second floor of Sina Headquarters Building, Phase II, Zhongguancun Software Park, Haidian District, Beijing

To register please contact:

Email: [email protected]
Tel: 0512-6590 0856 extension 27
Mobile: 18852405453

co-organiser

22510c856259cbd22e3cb60eb253cc43.png

Co

c85e2d289a2be166daa1866d49f690f2.jpeg

Guess you like

Origin blog.csdn.net/QbitAI/article/details/132797797