Comment Section for Hugging Face has a two-person team developing ChatGPT-like AI models | TechCrunch

The webpage from TechCrunch discusses how Hugging Face, an AI startup, has a two-person team developing ChatGPT-like AI models. Named H4, the team focuses on alignment, a process that involves teaching large language models (LLMs) to behave according to feedback from humans or other AIs. The team has created several open-source LLMs, including Zephyr-7B-α and a modified version of Falcon-40B. H4 uses a dedicated cluster of over 1,000 Nvidia A100 GPUs for training its models and collaborates with several internal Hugging Face teams as well as external groups. The team has recently been exploring different alignment techniques and building tools to test their effectiveness. While H4's tools don't directly generate revenue, they support Hugging Face’s Expert Acceleration Program, which helps enterprises build custom AI solutions. The team aims to contribute to the open AI community by releasing training code and datasets associated with its chat models.