Update taiwanese_english_translation.py

taiwanese_english_translation.py

@@ -48,7 +48,7 @@ _LICENSE = ""
 # TODO: Add link to the official dataset URLs here
 # The HuggingFace Datasets library doesn't host the datasets but only points to the original files.
 # This can be an arbitrary nested dict/list of URLs (see below in `_split_generators` method)
-_URLS = "https://github.com/atenglens/taiwanese_english_translation"
+_URLS = "https://huggingface.co/datasets/atenglens/taiwanese_english_translation"
 
 class TaiwaneseEnglishTranslationConfig(datasets.BuilderConfig):
     """BuilderConfig for FLoRes."""
@@ -131,7 +131,7 @@ class TaiwaneseEnglishTranslation(datasets.GeneratorBasedBuilder):
         urls = _URLS
         data_dir = dl_manager.download_and_extract(urls)
         
-        path_tmpl = "{data_dir}/{split}.txt"
+        path_tmpl = "{data_dir}/{split}.csv"
 
         files = {}
         for split in ("train", "valid", "test"):