Discover how to enhance multi-agent reinforcement learning (MARL) experiments using Ray and Weights & Biases in this 14-minute video presentation by Anyscale. Learn to leverage Ray's automation capabilities for MARL tuning, including dynamic tracing and optimization across parameter spaces, while efficiently utilizing computational resources through tools like AIR, Tune, and RLLib. Explore the benefits of integrating Weights & Biases to gain a comprehensive view of experiments, centralizing details and assets into a single ML system of record. Follow along as the presenter demonstrates experiments on Autonomous Vehicle Driving and Drone Flying scenarios, showcasing how this combination can save time on iterations and quickly identify optimal performance outcomes. Access the accompanying slide deck for additional information and visual aids. Gain insights into Anyscale's AI Application Platform and the Ray open-source framework for scaling AI workloads, including options for a managed Ray service.
Tend and Tune Your MARL Experiments with Ray and Weights & Biases
Overview
Syllabus
Tend and Tune Your MARL Experiments with Ray and Weights & Biases
Taught by
Anyscale
Related Courses
-
Evaluating and Debugging Generative AI
-
Building and Deploying Real-World ML Models with Ray and Weights & Biases
-
Ray/RLLib-Empowered Reinforcement Learning Based Recommender Systems in NetEase Game
-
Supercharge Your Training With PyTorch Lightning and Weights & Biases
-
MLtraq - Track Your ML/AI Experiments at Hyperspeed
