PyTorch Day China 2026 Call for Proposals Announced
Developers and AI researchers are invited to submit session proposals by November 10.
1Submission Deadline Approaches
The official call for proposals for PyTorch Day China 2026 is currently open for interested participants. All submissions must be finalized by November 10 at 23:59 China Standard Time. This event serves as a primary gathering for machine learning engineers and open source AI leaders to share technical expertise.
2Event Logistics and Timing
The conference is scheduled to take place in Beijing on December 12, 2026. Following the submission deadline, organizers will notify selected speakers on November 17. The final event schedule will be publicly announced on November 18, providing attendees with a clear roadmap of the sessions and lightning talks planned for the day.
3Focus on Model Training
One of the primary tracks for session presentations covers model training and post-training methodologies. This includes distributed training techniques, supervised fine-tuning, and reinforcement learning strategies. Participants are also encouraged to discuss preference optimization and various technical approaches aimed at improving overall training efficiency and performance for large-scale AI model development.
4Inference and Deployment Strategies
The event highlights efficient inference and deployment as a critical area for production-ready solutions. Topics in this track include large language model inference, quantization techniques, and optimization strategies. Engineers are invited to share their experiences in deploying AI models effectively, ensuring that these systems remain performant and scalable in real-world environments.
5Hardware Acceleration Advancements
PyTorch core hardware acceleration remains a central theme for the upcoming conference. Proposals focusing on compilation, kernel optimization, and performance profiling are highly encouraged. Additionally, the track covers advancements in GPU and AI accelerator technologies, aiming to bridge the gap between software frameworks and the underlying hardware infrastructure used for computation.
6Physical AI and Robotics
A dedicated track for Physical AI explores the intersection of software and hardware in robotics and intelligent devices. This includes Vision-Language-Action models, perception systems, and control mechanisms. Presenters are expected to showcase simulation-based training methods that enable AI to interact effectively with the physical world, pushing the boundaries of current robotics.
7Expanding the PyTorch Ecosystem
The final track invites submissions regarding broader PyTorch ecosystem applications and innovations. This category is designed to highlight new tools, unique applications, and creative projects that expand the utility of the framework. It provides a platform for developers to demonstrate how they are extending PyTorch to solve diverse and complex problems.
8Significance of the Gathering
PyTorch Day China 2026 acts as a vital hub for the regional AI community to align on technical standards. By gathering researchers and industry leaders, the event fosters collaboration on open source initiatives. The insights shared during these sessions will likely influence future development cycles and the adoption of PyTorch technologies.
Original article: PyTorch Blog — PyTorch Day China 2026 Call for Proposals Announced