Update tokenizer files

#2
by Xenova HF staff - opened
No description provided.
H2O.ai org

@Xenova could you please elaborate on this PR?
"add_prefix_space": false is correct

Sure thing! This is what I got when I loaded the tokenizer with .from_pretrained followed by .save_pretrained directly afterwards. I got a warning saying that it needed to re-convert the tokenizer.

H2O.ai org

That is an expected warning based on how HF has that implemented currently.

psinger changed pull request status to closed

Sign up or log in to comment