Datasets:
Update Dataloader for Clean Versions
Browse files- OpenBible_Swahili.py +7 -1
OpenBible_Swahili.py
CHANGED
@@ -103,7 +103,7 @@ class OpenBible_Swahili(datasets.GeneratorBasedBuilder):
|
|
103 |
name=book,
|
104 |
version=datasets.Version("1.1.0"),
|
105 |
)
|
106 |
-
for book in BOOKS + ["default"]
|
107 |
]
|
108 |
|
109 |
def _info(self):
|
@@ -129,6 +129,8 @@ class OpenBible_Swahili(datasets.GeneratorBasedBuilder):
|
|
129 |
def _split_generators(self, dl_manager):
|
130 |
if self.config.name == "default":
|
131 |
archive_paths = [dl_manager.download_and_extract(f"{_DATA_URL}/{book}.tar.gz") for book in BOOKS]
|
|
|
|
|
132 |
else:
|
133 |
archive_paths = [dl_manager.download_and_extract(f"{_DATA_URL}/{self.config.name}.tar.gz")]
|
134 |
|
@@ -155,6 +157,10 @@ class OpenBible_Swahili(datasets.GeneratorBasedBuilder):
|
|
155 |
book_chapter = f"{book}_{chapter.zfill(3)}"
|
156 |
book_chapter_verse = f"{book_chapter}_{verse_number.zfill(3)}"
|
157 |
audio_path = (path / book / book_chapter / book_chapter_verse).with_suffix(".wav")
|
|
|
|
|
|
|
|
|
158 |
transcript_path = audio_path.with_suffix(".txt")
|
159 |
with open(transcript_path) as f:
|
160 |
transcript = f.read().strip()
|
|
|
103 |
name=book,
|
104 |
version=datasets.Version("1.1.0"),
|
105 |
)
|
106 |
+
for book in BOOKS + [f"{_book}_clean" for _book in BOOKS] + ["default", "clean"]
|
107 |
]
|
108 |
|
109 |
def _info(self):
|
|
|
129 |
def _split_generators(self, dl_manager):
|
130 |
if self.config.name == "default":
|
131 |
archive_paths = [dl_manager.download_and_extract(f"{_DATA_URL}/{book}.tar.gz") for book in BOOKS]
|
132 |
+
elif self.config.name == "clean":
|
133 |
+
archive_paths = [dl_manager.download_and_extract(f"{_DATA_URL}/{book}_clean.tar.gz") for book in BOOKS]
|
134 |
else:
|
135 |
archive_paths = [dl_manager.download_and_extract(f"{_DATA_URL}/{self.config.name}.tar.gz")]
|
136 |
|
|
|
157 |
book_chapter = f"{book}_{chapter.zfill(3)}"
|
158 |
book_chapter_verse = f"{book_chapter}_{verse_number.zfill(3)}"
|
159 |
audio_path = (path / book / book_chapter / book_chapter_verse).with_suffix(".wav")
|
160 |
+
# skip if audio file does not exist, e.g. audio not included in clean data after filtering
|
161 |
+
if not audio_path.exists():
|
162 |
+
continue
|
163 |
+
|
164 |
transcript_path = audio_path.with_suffix(".txt")
|
165 |
with open(transcript_path) as f:
|
166 |
transcript = f.read().strip()
|