supert

joined 1 year ago
[โ€“] supert@lemmy.fmhy.ml 1 points 1 year ago

Check out localllama community. Lots of info there.

I use oobabooga + exllama.

Things are a bit budget dependent. If you can afford a rtx 3090 off ebay you can run some decent models (30B) at very good speed. I ended up with 3090 + 4090. You can use system ram with ggml but it's slow. Mac M1 is not bad for this .

Where did you get the reddit dataset?

[โ€“] supert@lemmy.fmhy.ml 2 points 1 year ago

The same (65B versions) and wizard vicuna uncensored 30B.