Swedish text entries extracted from the Europarl bilingual English-Swedish training split. The data features a single 'text' column where strings have been lowercased and filtered via regular expressions to remove noise characters.
Use Cases
- Train a Swedish language model or tokenizer using the cleaned 'text' column.
- Perform frequency analysis of Swedish political terminology within the 'text' field.
- Fine-tune a masked language model (MLM) on formal Swedish parliamentary discourse.
Strengths
- Extracted from the 'sv' translation field of the europarl_bilingual dataset.
- Contains a single 'text' column containing lowercased Swedish sentences.
- Pre-processed using regex to remove noise characters defined in chars_to_ignore_regex.