Change split names
Browse files- MultiDialog.py +6 -17
MultiDialog.py
CHANGED
@@ -122,7 +122,7 @@ class Multidialog(datasets.GeneratorBasedBuilder):
|
|
122 |
if self.config.name == "test_freq":
|
123 |
return [
|
124 |
datasets.SplitGenerator(
|
125 |
-
name=
|
126 |
gen_kwargs={
|
127 |
"audio_archives_iterators": [
|
128 |
dl_manager.iter_archive(archive_path) for archive_path in audio_archives_paths["test_freq"]
|
@@ -137,7 +137,7 @@ class Multidialog(datasets.GeneratorBasedBuilder):
|
|
137 |
if self.config.name == "test_rare":
|
138 |
return [
|
139 |
datasets.SplitGenerator(
|
140 |
-
name=
|
141 |
gen_kwargs={
|
142 |
"audio_archives_iterators": [
|
143 |
dl_manager.iter_archive(archive_path) for archive_path in audio_archives_paths["test_rare"]
|
@@ -152,7 +152,7 @@ class Multidialog(datasets.GeneratorBasedBuilder):
|
|
152 |
if self.config.name == "valid_freq":
|
153 |
return [
|
154 |
datasets.SplitGenerator(
|
155 |
-
name=
|
156 |
gen_kwargs={
|
157 |
"audio_archives_iterators": [
|
158 |
dl_manager.iter_archive(archive_path) for archive_path in audio_archives_paths["valid_freq"]
|
@@ -167,7 +167,7 @@ class Multidialog(datasets.GeneratorBasedBuilder):
|
|
167 |
if self.config.name == "valid_rare":
|
168 |
return [
|
169 |
datasets.SplitGenerator(
|
170 |
-
name=
|
171 |
gen_kwargs={
|
172 |
"audio_archives_iterators": [
|
173 |
dl_manager.iter_archive(archive_path) for archive_path in audio_archives_paths["valid_rare"]
|
@@ -182,7 +182,7 @@ class Multidialog(datasets.GeneratorBasedBuilder):
|
|
182 |
if self.config.name == "train":
|
183 |
return [
|
184 |
datasets.SplitGenerator(
|
185 |
-
name=
|
186 |
gen_kwargs={
|
187 |
"audio_archives_iterators": [
|
188 |
dl_manager.iter_archive(archive_path) for archive_path in audio_archives_paths["train"]
|
@@ -205,10 +205,6 @@ class Multidialog(datasets.GeneratorBasedBuilder):
|
|
205 |
for line in jsonl_file:
|
206 |
data = json.loads(line.strip())
|
207 |
meta_dict[data["file_name"]] = data
|
208 |
-
# data = json.loads(line.strip())
|
209 |
-
# meta_csv = csv.DictReader(csvfile)
|
210 |
-
# for line in meta_csv:
|
211 |
-
|
212 |
|
213 |
for audio_path_in_archive, audio_file in audio_archive_iterator:
|
214 |
# `audio_path_in_archive` is like "dev_chunks_0000/YOU1000000029_S0000095.wav"
|
@@ -228,11 +224,4 @@ class Multidialog(datasets.GeneratorBasedBuilder):
|
|
228 |
yield audio_id, {
|
229 |
"audio": {"path": path , "bytes": audio_file.read()},
|
230 |
**{feature: value for feature, value in audio_meta.items() if feature in self.info.features}
|
231 |
-
}
|
232 |
-
|
233 |
-
|
234 |
-
def _flatten_nested_dict(nested_dict):
|
235 |
-
return {
|
236 |
-
key: [inner_list_element for inner_list in value_to_lists.values() for inner_list_element in inner_list]
|
237 |
-
for key, value_to_lists in nested_dict.items()
|
238 |
-
}
|
|
|
122 |
if self.config.name == "test_freq":
|
123 |
return [
|
124 |
datasets.SplitGenerator(
|
125 |
+
name="test_freq",
|
126 |
gen_kwargs={
|
127 |
"audio_archives_iterators": [
|
128 |
dl_manager.iter_archive(archive_path) for archive_path in audio_archives_paths["test_freq"]
|
|
|
137 |
if self.config.name == "test_rare":
|
138 |
return [
|
139 |
datasets.SplitGenerator(
|
140 |
+
name="test_rare",
|
141 |
gen_kwargs={
|
142 |
"audio_archives_iterators": [
|
143 |
dl_manager.iter_archive(archive_path) for archive_path in audio_archives_paths["test_rare"]
|
|
|
152 |
if self.config.name == "valid_freq":
|
153 |
return [
|
154 |
datasets.SplitGenerator(
|
155 |
+
name="valid_freq",
|
156 |
gen_kwargs={
|
157 |
"audio_archives_iterators": [
|
158 |
dl_manager.iter_archive(archive_path) for archive_path in audio_archives_paths["valid_freq"]
|
|
|
167 |
if self.config.name == "valid_rare":
|
168 |
return [
|
169 |
datasets.SplitGenerator(
|
170 |
+
name="valid_rare",
|
171 |
gen_kwargs={
|
172 |
"audio_archives_iterators": [
|
173 |
dl_manager.iter_archive(archive_path) for archive_path in audio_archives_paths["valid_rare"]
|
|
|
182 |
if self.config.name == "train":
|
183 |
return [
|
184 |
datasets.SplitGenerator(
|
185 |
+
name="train",
|
186 |
gen_kwargs={
|
187 |
"audio_archives_iterators": [
|
188 |
dl_manager.iter_archive(archive_path) for archive_path in audio_archives_paths["train"]
|
|
|
205 |
for line in jsonl_file:
|
206 |
data = json.loads(line.strip())
|
207 |
meta_dict[data["file_name"]] = data
|
|
|
|
|
|
|
|
|
208 |
|
209 |
for audio_path_in_archive, audio_file in audio_archive_iterator:
|
210 |
# `audio_path_in_archive` is like "dev_chunks_0000/YOU1000000029_S0000095.wav"
|
|
|
224 |
yield audio_id, {
|
225 |
"audio": {"path": path , "bytes": audio_file.read()},
|
226 |
**{feature: value for feature, value in audio_meta.items() if feature in self.info.features}
|
227 |
+
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|