Mistral AI 7B Models Now on Hugging Face
Mistral AI introduces its new 7B base and instruct models, now available on Hugging Face. The models feature an extended vocabulary from 32,000 to 32,768 tokens and support function calling in the instruct variant. Licensed under Apache 2.0, these models provide flexible use without the need for magnet links. However, no shared benchmarks or evaluations are provided at this time. Discover the latest advancements in AI modeling with Mistral AI on Hugging Face.