Small100Tokenizer Error

#7
by miwytt - opened

Hi! I used the Small100Tokenizer for some experiments a couple of weeks ago and it worked perfectly fine, but when I tried to rerun it more recently it resulted in an attribute error ('AttributeError: 'SMALL100Tokenizer' object has no attribute 'encoder''). I encountered the same error when using the demo link on the model card. After switching to the M2MTokenizer (M2M100Tokenizer.from_pretrained("alirezamsh/small100")) it ran smoothly – does anybody know what the issue could be?

Update: Problem solved!

The issue seems to arise due to the transformers library update for version 4.34, which heavily influences the tokenizer workflow. Hence, tokenization_small100.py only functions with transformer < 4.34 at the moment.

Thanks for the message. I will update it soon!

alirezamsh changed discussion status to closed

Sign up or log in to comment