Swedish text corpus created by extracting the "text"
from dataset = load_dataset("europarl_bilingual", lang1="en", lang2="sv", split="train")
and processing it with:
import re
def extract_text(batch):
text = batch["translation"]["sv"]
batch["text"] = re.sub(chars_to_ignore_regex, "", text.lower())
return batch