Mistral CEO confirms ‘leak’ of new open source AI model nearing GPT-4 performance
Mistral CEO confirms ‘leak’ of new open source AI model nearing GPT-4 performance
An anonymous user on 4chan posted a link to the miqu-1-70b files on 4chan. The open source model approaches GPT-4 performance.
It's not open source it's weight available(for now). As of now there's nothing you can do with it publicly because it lacks a license and is known to be stolen.
I'm sure that's not that big of a deal to some people. For example, I'm mainly using LLMs for use in my home assistant instance
So what I'm hearing is I can use it but just make sure I don't tell anyone
Has anyone tried to see how censored it is yet?
Definitely. It has some alignment, but it won’t straight up refuse to do anything. It will sometimes add notes saying that what you’ve asked is kinda maybe against the law, but will produce a great response regardless. It’s a 70b, so running it locally is kind of a challenge, but for those who can run it - there is simply no other LLM that you can run at home that gets even close to it. It follows instructions amazingly, it’s very consistent and barely hallucinates. There is some special mistral sauce in it for sure, even if it’s “just” a llama2-70b.
GGUF q2_K works quite well IMO, I've run it with 12GB vram + 32GB ram
How does one leak open source material?
Not all open source is created in the light.
I’ve no idea what happened in this case. But it is extremely common for companies to take closed source developments and transform them into open source after they go through an internal process.