noflm
commited on
Commit
•
8458c7e
1
Parent(s):
55d0e5f
fix dataset config
Browse files- EliteVoiceProject.py +3 -4
EliteVoiceProject.py
CHANGED
@@ -102,7 +102,6 @@ class EliteVoiceProject(datasets.GeneratorBasedBuilder):
|
|
102 |
version = self.config.version
|
103 |
|
104 |
audio_urls = {}
|
105 |
-
#splits = ("twitter", "youtube", "twitch", "test")
|
106 |
splits = ("train", "test")
|
107 |
for split in splits:
|
108 |
audio_urls[split] = [
|
@@ -146,9 +145,9 @@ class EliteVoiceProject(datasets.GeneratorBasedBuilder):
|
|
146 |
# row["accent"] = row["accents"]
|
147 |
# del row["accents"]
|
148 |
# if data is incomplete, fill with empty values
|
149 |
-
|
150 |
-
|
151 |
-
|
152 |
metadata[row["path"]] = row
|
153 |
|
154 |
for i, audio_archive in enumerate(archives):
|
|
|
102 |
version = self.config.version
|
103 |
|
104 |
audio_urls = {}
|
|
|
105 |
splits = ("train", "test")
|
106 |
for split in splits:
|
107 |
audio_urls[split] = [
|
|
|
145 |
# row["accent"] = row["accents"]
|
146 |
# del row["accents"]
|
147 |
# if data is incomplete, fill with empty values
|
148 |
+
for field in data_fields:
|
149 |
+
if field not in row:
|
150 |
+
row[field] = ""
|
151 |
metadata[row["path"]] = row
|
152 |
|
153 |
for i, audio_archive in enumerate(archives):
|