Daniel Han

@user074 Oh interesting! We welcome new contributions but for now Unsloth supports general LLM heads - more custom heads will require manual coding (eg as in ur case)

@h-a-s-k Oh my I did not notice Kaggle had this issue whoops - thanks let me check first and get back to you! Thanks again!

@h-a-s-k Thanks for the PR again! Apologies on the delay! I had to edit Kaggle drastically - it seems like due to a 20GB disk limit, one can only save...

Ye pip normally works - I think the Xformers team hasnt uploaded Conda packages although I'm unsure. Another way is to limit the Xformers version: `xformers

@Ben-Pfirsich See https://github.com/unslothai/unsloth/wiki#finetuning-from-your-last-checkpoint

@arunpatala Oh yep that is an issue - planning to support the 128K Phi later down the road

@JiesenSun Oh my that is a super weird error msg - sadly I can't be of much help - I would try reinstalling :(

Oh great you solved it! Sorry I wasn't helpful!

@KillerShoaib WHOOPS you are entirely correct!!!! I immediately updated all `pad_token`s Unsloth has to `` Thanks for the keen eye!!

OO thanks for this - highly appreciate it - I will check it out! Might also have to edit the readme (if you're up for it :))