StarVLA: A Lego-like Codebase for Vision-Language-Action Model Developing

Agent skills for vLLM

A framework for efficient model inference with omni-modality models

A high-throughput and memory-efficient inference and serving engine for LLMs