Preemo is developing an open-source text generation inference server, a fork of Hugging Face’s original project with OpenSource license. The server will be modular, allowing for easy addition of state-of-the-art models and functionalities. Development is set to begin in September 2023, with plans to add a public CI/CD pipeline, unify build tools, and introduce a plugin system. Preemo’s long-term goal is to foster a community around this repository for exploring new ideas in LLM inference.
Read more at GitHub…