How to use nielsr/layoutreader-readingbank with Transformers:
# Load model directly from transformers import LayoutReaderForSeq2SeqDecoding model = LayoutReaderForSeq2SeqDecoding.from_pretrained("nielsr/layoutreader-readingbank", dtype="auto")
Is there any chance to use a different tokenizer/vocab along with the layoutreader-readingbank model?
E.g., i would need the "dbmdz/bert-base-german-cased" tokenizer for correct German umlauts.
Kind regards
· Sign up or log in to comment
