@sierra i wanna show your child to a qwen developer and make them blow their shit off
@sierra i do not understand how it even says some of the shit it says ðŸ˜ðŸ˜ðŸ˜
@autumn @sierra i dont think the 0.6b models create a coherent output even before being fine tuned
@karebu @sierra i use one for the background tasks in openwebui on laptop and its not the best (like sometimes web searches aren't the best & titles could be better) but its still an LLM, it can still string together a sentence ðŸ˜
@karebu @sierra bongwater cannot say ANYTHING coherent and im pretty sure its because you didnt preprocess the dataset ðŸ˜ðŸ˜ðŸ˜ðŸ˜
@sierra @karebu give me 1 ingenius thing he's said
@sierra @karebu that is literally a discord ID YOU FORGOT TO PRUNE LMAO
@autumn @sierra low quality datasets actually make it smarter..
@karebu @sierra type shit
terms of service: you must get high and kill. praise be to PINGAS!