Keynote: Accelerating AI/ML Development with Open Source Software
summary
AWS describes concrete open source work that lets teams run notebooks as production jobs with Jupyter Scheduler using immutable snapshots and pluggable backends, interact with models inside the notebook via Jupyter AI (magic command, chat pane, SL learn/ask/generate, local models and metadata tagging), and wire LLM apps to cloud services through LangChain integrations. A new S3 connector for PyTorch claims 40% faster data loading and checkpointing with iterable/map datasets and direct S3 checkpoints.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:03Customer challenges with AI/ML
- 2:29AWS strategies for open source
- 3:31AWS contributions to PyTorch
- 4:16Jupyter Scheduler for notebooks
- 6:42Jupyter AI for generative AI
- 10:14Jupyter AI commands
- 13:24AWS contributions to LangChain
- 14:29PyTorch S3 connector
- 15:53Conclusion on open source value
related talks
16mKeynote: Accelerating AI/ML Development with Open Source Software
16mKeynote: Open Source for AI and AI for Open Source
11mKeynote: Open Source Stack Empowers the AI Era
16mKeynote: Marin: An Open Lab for Frontier AI
16mKeynote: An Open-Source Stack for AI Compute: Kubernetes + Ray + PyTorch + vLLM
13mEngineer your company’s future with AI: Use open source tools on AWS
33mKeynote: Fine-tuning our way towards openness in AI
09mSponsored Keynote: Production AI Runs on Open Sou...