verl: Flexible and Scalable Reinforcement Learning Library for LLM Reasoning and Tool-Calling

Transcript

Pleasure to introduce Hibin Lynn who works on LLM infrastructure at Bance seed and before that has done a lot of work on MXNet and Amazon. Today he'll be talking about volcano engine reinforcement learning for LLMs or VUL an extremely popular reinforcement learning pipeline that provides a hybrid controller programming model first class support for... Read More