Tokenless

Ml

Efficient Reinforcement Learning – Rhythm Garg & Linden Li, Applied Compute

Efficient Reinforcement Learning – Rhythm Garg & Linden Li, Applied Compute

At Applied Compute, efficient Reinforcement Learning is critical for delivering business value. This talk explores the transition from inefficient synchronous RL to a high-throughput asynchronous 'Pipeline RL' system. The core challenge is managing 'staleness'—a side effect of in-flight weight updates that can destabilize training. The speakers detail their first-principles systems model, based on the Roofline model, used to simulate and find the optimal allocation of GPU resources between sampling and training, balancing throughput with algorithmic stability and achieving significant speedups.

The GPU Uptime Battle

The GPU Uptime Battle

Andy Pernsteiner, Field CTO of VAST Data, discusses the immense challenges of transitioning AI projects from prototype to production. He highlights the critical role of data infrastructure, the high cost of GPU downtime, and the necessity of building resilient, scalable platforms that can withstand real-world failures like power outages in massive data centers. The conversation emphasizes a shift in mindset towards empathy, better requirement gathering, and closer collaboration between data scientists and platform engineers to bridge the gap between development and operations.

Big updates to mlflow 3.0

Big updates to mlflow 3.0

Databricks’ Eric Peter and Corey Zumar introduce MLflow 3.0, focusing on its new "Agentic Insights" capabilities. They demonstrate how MLflow is evolving from providing tools for manual quality assurance in Generative AI to using intelligent agents to automatically find, diagnose, and prioritize issues, significantly speeding up the development lifecycle.

Introducing serverless reinforcement learning: Train reliable AI agents without worrying about GPUs

Introducing serverless reinforcement learning: Train reliable AI agents without worrying about GPUs

Kyle Corbett and Daniel from CoreWeave (formerly Openpipe) discuss the practical advantages of Reinforcement Learning (RL) over Supervised Fine-Tuning (SFT) for building reliable and efficient AI agents. They introduce Serverless RL, a new platform designed to eliminate the infrastructure complexities of RL training, and share a playbook for teams looking to get started.

Elm & Open Source: What's Next? • Evan Czaplicki & Kris Jenkins

Elm & Open Source: What's Next? • Evan Czaplicki & Kris Jenkins

Evan Czaplicki, creator of Elm, discusses the sustainability challenges of open-source that led to Elm's development pause, his new work on a full-stack typed functional language, and his thoughts on improving developer experience to make functional programming more accessible.

Machine Learning Explained: A Guide to ML, AI, & Deep Learning

Machine Learning Explained: A Guide to ML, AI, & Deep Learning

A breakdown of Machine Learning (ML), its relationship with AI and Deep Learning, and its core paradigms: supervised, unsupervised, and reinforcement learning. The summary explores classic models and connects them to modern applications like Large Language Models (LLMs) and Reinforcement Learning with Human Feedback (RLHF).

← PreviousNext →