FacebookAI

/

xlm-roberta-large-finetuned-conll03-english

like 162

Follow

Facebook AI community 130

Hi there! I have been using this model and turns out that when I pipe it after fine-tuning, the utput ignores blank spaces and returns all words together, as in:

[{'entity_group': 'LABEL_0',
'score': 0.4824247,
'word': 'Thedogandthecatwenttothehouse',
'start': 0,
'end': 325}]

I have tried add_prefix_space=True in the tokenizer, but it does not seem to be working. Could someone give me a little push on this? Many thanks in advance.