The PyTorch Conference 2024 in San Francisco was a vibrant gathering of nearly 1,500 AI researchers, developers, and enthusiasts. Over two days, the event featured engaging discussions, insightful keynotes, and hands-on sessions focused on advancements in PyTorch, the leading open-source machine learning framework.

Key Themes

1. Generative AI and Large Language Models (LLMs):
   • Many sessions highlighted PyTorch’s evolution as a primary framework for LLMs and generative AI applications. Discussions covered scaling models and optimizing performance across various hardware platforms.
2. Democratizing AI Through Open Source:
   • A recurring theme was the importance of open-source tools and communities in shaping AI’s future. PyTorch’s commitment to inclusivity and accessibility aims to bring AI to a broader global audience.
3. Distributed and Edge Computing:
   • The conference emphasized how PyTorch is driving AI to the edge, with a focus on edge accelerators, scalable training, and inference. This enables powerful models to be deployed across diverse environments, from the cloud to on-device applications.

Highlights

• Keynote Sessions: Featured top AI leaders discussing the latest advancements and future directions of PyTorch.
• Startup Showcase: An exciting addition this year, where early-stage founders pitched their AI startups to top venture capitalists. The winner, CTGT, was recognized for enhancing data lineage in LLMs and reducing hallucinations.
• Technical Deep Dives: Sessions covered the evolution of PyTorch, new libraries like Torchtune for fine-tuning LLMs, and TorchChat for deploying models on various devices.

Major Takeaways

• PyTorch 2.4 and Upcoming 2.5: The conference showcased the latest features and improvements in PyTorch 2.4, with a preview of what’s coming in version 2.5.
• Llama 3.1: Discussions included advancements in this language model and its integration with PyTorch.

The PyTorch Conference 2024 was a testament to the dynamic and collaborative nature of the AI community, pushing the boundaries of what’s possible with open-source technology. View all of the conference sessions on the PyTorch YouTube channel: https://www.youtube.com/playlist?list=PL_lsbAsL_o2B_znuvm-pDtV_cRhpqZb8l

GraphQLConf 2024, held from September 10-12 in San Francisco, was a vibrant gathering of the global GraphQL community. The conference featured a mix of keynotes, workshops, and panel discussions, focusing on the latest advancements and best practices in GraphQL.

Key Highlights:

• Keynote Sessions: Michael Staib from ChilliCream Inc. delivered an insightful keynote on “The State of Distributed GraphQL,” while Anant Jhingran from IBM discussed the intersection of GraphQL, BFFs, and AI.
• Workshops and Talks: Attendees participated in hands-on workshops and talks covering a wide range of topics, from GraphQL in production to identity management with GraphQL.
• Networking Opportunities: The event provided ample opportunities for networking, allowing participants to connect with industry leaders, developers, and innovators.
• Community Engagement: The conference emphasized community building, with sessions designed to foster collaboration and knowledge sharing among GraphQL enthusiasts.

Overall, GraphQLConf 2024 was a successful event that brought together a diverse group of professionals passionate about the future of GraphQL and its applications.

View all of the conference sessions on the GraphQL Foundation YouTube channel: https://www.youtube.com/playlist?list=PLP1igyLx8foEO0qsyk3IFn1peYSVGDBFA