The main limitation is the VRAM, but I doubt any model is going to be particularly fast.
I think phi3:mini on ollama might be an okish fit for python, since it's a small model, but was trained on python codebases.
ViatorOmnium
joined 3 days ago
What can I use for an offline, selfhosted LLM client, pref with images,charts, python code execution
I'm starting to suspect there's something wrong with the Taurus.
Paraphrasing Churchill, if AI companies were plagiarizing hell's works I would make at least a favourable reference to the devil.
In this case that's exactly what we need, someone with more money than god that can set precedents to protect the authors and artists that need it the most.
For English yes, but there's no equivalent in other languages.
People are mentioning drip, and that's on the Play Store. It's literally the same amount of effort as installing a surveillance app.
This sounds like the Polish Constitutional Court just said Poland can't stay in the EU.
I'm betting on Google destroying Google instead.