Embedded Files
Talk Date and Time: April 18, 2024 at 12:00 pm - 12:45 pm EST followed by 15 minutes of Q&A in IRB-5165 and on Google Meet
Topic: Growing the MARL software ecosystem in JAX
Abstract:
We present JAX-based libraries and tools to support an ecosystem for Multi-Agent Reinforcement Learning research. Our ecosystem contributions include stable and performant software, tightly integrated with scalable yet hackable online and offline algorithms with environments, datasets, and evaluation protocols to produce reliable research artefacts. Developed by researchers for researchers, our stack should allow anyone in the MARL community to go from proposal to paper with confidence.
Bio:
The MARL research team at InstaDeep works on large-scale multi-agent learning with a focus on algorithmic innovation in cooperative systems for industrial applications. The team regularly contributes to the research community through publications at venues such as NeurIPS and ICLR, and to the open-source community, by releasing open-source software for MARL research.
Page updated
Google Sites
Report abuse