混元文本/多模态预训练框架研发工程师（深圳/北京/上海/杭州）
Tencent
混元文本/多模态预训练框架研发工程师（深圳/北京/上海/杭州）
Job Description
混元文本/多模态预训练框架研发工程师（深圳/北京/上海/杭州）
📋 Job Overview
The role involves developing and optimizing large-scale pre-training frameworks for text and multimodal models at Tencent. Engineers will design model structures, verify training efficiency, and accelerate performance for generative AI applications like text-to-image, text-to-video, and text-to-3D. The position is available in Shenzhen, Beijing, Shanghai, or Hangzhou, focusing on high-efficiency training at scales exceeding 10,000 GPUs.
📍 Location: Shenzhen, China
🏢 Business Unit: TEG
📄 Full Description
2.参与NLP、多模态大模型结构设计，并联合业务进行模型训练效率和效果验证；
3.参与文生图、文生视频、文生3D等业务的训练性能加速；
4.参与低精度训练性能优化和业务推广、参与大窗口训练性能优化。
🎯 Key Responsibilities
- Develop and optimize large model training frameworks to support efficient and stable training at scales over 10,000 GPUs for single tasks
- Participate in NLP and multimodal large model structure design, and collaborate with business teams to verify model training efficiency and effectiveness
- Contribute to training performance acceleration for text-to-image, text-to-video, text-to-3D, and similar business applications
- Engage in low-precision training performance optimization and business promotion, as well as large window training performance optimization
Locations
- Shenzhen, China
Salary
300,000 - 800,000 CNY / yearly
Source: ai estimated
* This is an estimated range based on market data and may vary based on experience and qualifications.
Responsibilities
- Develop and optimize large model training frameworks to support efficient and stable training at scales over 10,000 GPUs for single tasks
- Participate in NLP and multimodal large model structure design, and collaborate with business teams to verify model training efficiency and effectiveness
- Contribute to training performance acceleration for text-to-image, text-to-video, text-to-3D, and similar business applications
- Engage in low-precision training performance optimization and business promotion, as well as large window training performance optimization
Tags & Categories
Answer 10 quick questions to check your fit for 混元文本/多模态预训练框架研发工程师（深圳/北京/上海/杭州） @ Tencent.
Related Books and Jobs
混元文本/多模态预训练框架研发工程师（深圳/北京/上海/杭州）
Tencent
混元文本/多模态预训练框架研发工程师（深圳/北京/上海/杭州）
Job Description
混元文本/多模态预训练框架研发工程师（深圳/北京/上海/杭州）
📋 Job Overview
The role involves developing and optimizing large-scale pre-training frameworks for text and multimodal models at Tencent. Engineers will design model structures, verify training efficiency, and accelerate performance for generative AI applications like text-to-image, text-to-video, and text-to-3D. The position is available in Shenzhen, Beijing, Shanghai, or Hangzhou, focusing on high-efficiency training at scales exceeding 10,000 GPUs.
📍 Location: Shenzhen, China
🏢 Business Unit: TEG
📄 Full Description
2.参与NLP、多模态大模型结构设计，并联合业务进行模型训练效率和效果验证；
3.参与文生图、文生视频、文生3D等业务的训练性能加速；
4.参与低精度训练性能优化和业务推广、参与大窗口训练性能优化。
🎯 Key Responsibilities
- Develop and optimize large model training frameworks to support efficient and stable training at scales over 10,000 GPUs for single tasks
- Participate in NLP and multimodal large model structure design, and collaborate with business teams to verify model training efficiency and effectiveness
- Contribute to training performance acceleration for text-to-image, text-to-video, text-to-3D, and similar business applications
- Engage in low-precision training performance optimization and business promotion, as well as large window training performance optimization
Locations
- Shenzhen, China
Salary
300,000 - 800,000 CNY / yearly
Source: ai estimated
* This is an estimated range based on market data and may vary based on experience and qualifications.
Responsibilities
- Develop and optimize large model training frameworks to support efficient and stable training at scales over 10,000 GPUs for single tasks
- Participate in NLP and multimodal large model structure design, and collaborate with business teams to verify model training efficiency and effectiveness
- Contribute to training performance acceleration for text-to-image, text-to-video, text-to-3D, and similar business applications
- Engage in low-precision training performance optimization and business promotion, as well as large window training performance optimization
Tags & Categories
Answer 10 quick questions to check your fit for 混元文本/多模态预训练框架研发工程师（深圳/北京/上海/杭州） @ Tencent.
Related Books and Jobs