language icon indicating copy to clipboard operation
language copied to clipboard

Published 20 hours ago • verified publisher icon google-research

CANINE Pretraining Code (pt.2)

Open stefan-it opened this issue 2 years ago • 4 comments

Hi @jhclark-google and @dhgarrette,

I would like to know if there's any chance to get the pretraining code for CANINE.

It's been a long time since the readme was updated and I'm highly interested in pretraining own models on other datasets.

Many thanks in advance!

stefan-it avatar Jun 08 '23 21:06 stefan-it

I am also interested in this.

mwesthelle avatar May 31 '24 00:05 mwesthelle

Any updates on this? I would love to take a look at this since existing wordpiece/sentence piece tokenization doesnt fit our data

ganeshkrishnan1 avatar Jun 07 '24 02:06 ganeshkrishnan1