metadata
language:
- en
license: apache-2.0
datasets:
- WNC
metrics:
- accuracy
This model is a fine-tune checkpoint of T5-small, fine-tuned on the Wiki Neutrality Corpus (WNC), a labeled dataset composed of 180,000 biased and neutralized sentence pairs that are generated from Wikipedia edits tagged for “neutral point of view”. This model reaches an accuracy of 0.32 on a dev split of the WNC.
For more details about T5, check out this model card.