Baidu has fulfilled its promise to open-source its large language models (LLMs) with the release of ten variants of its Ernie 4.5 AI series. The Chinese tech giant announced the move on Monday, making the models available to the open community. These models are built on a Mixture-of-Experts (MoE) architecture, a technique that enhances performance and efficiency.
This release signifies Baidu’s commitment to fostering collaboration and innovation in the AI field. By providing access to its advanced technology, Baidu aims to empower developers and researchers to build upon and improve these models. Alongside the Ernie 4.5 variants, Baidu has also released multi-hardware development toolkits in open source, further facilitating the integration and utilization of the models across various platforms. This marks a significant step towards democratizing access to advanced AI technology.
