Mistral AI Releases Its 7B Model Under the Apache 2.0 License

IBL News | New York

The French start-up Mistral AI made available for free download this week its 7B model, including a 13.4 gigabyte torrent, under the Apache 2.0 license.

The company, which has raised $113 million to date, has also started a GitHub repository and Discord channel for collaboration and troubleshooting.

Mistral 7B is a further refinement of other small and large language models like Llama 2, offering similar capabilities at a considerably smaller compute cost.

“Our ambition is to become the leading supporter of the open generative AI community, and bring open models to state-of-the-art performance,” wrote Mistral’s team in a blog post.

“Mistral 7B’s performance demonstrates what small models can do with enough conviction. This is the result of three months of intense work, in which we assembled the Mistral AI team, rebuilt a top-performance MLops stack, and designed a most sophisticated data processing pipeline, from scratch.”

Mistral’s business model is based on making users to pay if they want to dig in. The company is actively working on hosted solutions and dedicated deployment for enterprises. Its commercial offering will be distributed as white-box solutions, making both weights and code sources available.