Sunflower Björnskalle 🌻 on Nostr: npub1e7vnd…5flg9 LLaMa was pretty good if I just asked it like for options for how ...

npub1e7vndtpf0s9pcg6yxkmvwvfq2g56rg5694nlgthwg7j2hkjgeuwq65flg9 (npub1e7v…flg9) LLaMa was pretty good if I just asked it like for options for how to do something. It was pretty methodical, whereas Qwen (even 14B) was more sloppy and incomplete. I haven't tried Qwen2.5-coder yet though.

So you're actually integrating it into projects? I'm just playing around so far.