Commit
·
8f48036
1
Parent(s):
4485095
with metadata
Browse files- libri_light.py +37 -37
libri_light.py
CHANGED
|
@@ -92,38 +92,38 @@ class LibriLight(datasets.GeneratorBasedBuilder):
|
|
| 92 |
"file": datasets.Value("string"),
|
| 93 |
"audio": datasets.Audio(sampling_rate=16_000),
|
| 94 |
"speaker_id": datasets.Value("int64"),
|
| 95 |
-
|
| 96 |
-
|
| 97 |
-
|
| 98 |
-
|
| 99 |
-
|
| 100 |
-
|
| 101 |
-
|
| 102 |
-
|
| 103 |
-
|
| 104 |
-
|
| 105 |
-
|
| 106 |
-
|
| 107 |
-
|
| 108 |
-
|
| 109 |
-
|
| 110 |
-
|
| 111 |
-
|
| 112 |
-
|
| 113 |
-
|
| 114 |
-
|
| 115 |
-
|
| 116 |
-
|
| 117 |
-
|
| 118 |
-
|
| 119 |
-
|
| 120 |
-
|
| 121 |
-
|
| 122 |
-
|
| 123 |
-
|
| 124 |
-
|
| 125 |
-
|
| 126 |
-
|
| 127 |
}
|
| 128 |
)
|
| 129 |
|
|
@@ -145,18 +145,18 @@ class LibriLight(datasets.GeneratorBasedBuilder):
|
|
| 145 |
]
|
| 146 |
|
| 147 |
def _generate_examples(self, archive_path):
|
| 148 |
-
paths = glob.glob(os.path.join(archive_path, "
|
| 149 |
for key, path in enumerate(paths):
|
| 150 |
path_split = path.split("/")
|
| 151 |
id_ = path_split[-1].replace(".flac", "")
|
| 152 |
speaker_id = int(path_split[-3])
|
| 153 |
-
|
| 154 |
-
|
| 155 |
-
|
| 156 |
yield key, {
|
| 157 |
"id": id_,
|
| 158 |
"file": path,
|
| 159 |
"audio": path,
|
| 160 |
"speaker_id": speaker_id,
|
| 161 |
-
|
| 162 |
}
|
|
|
|
| 92 |
"file": datasets.Value("string"),
|
| 93 |
"audio": datasets.Audio(sampling_rate=16_000),
|
| 94 |
"speaker_id": datasets.Value("int64"),
|
| 95 |
+
"metadata": {
|
| 96 |
+
"speaker": datasets.Value("string"),
|
| 97 |
+
"book_meta": {
|
| 98 |
+
"id": datasets.Value("string"),
|
| 99 |
+
"title": datasets.Value("string"),
|
| 100 |
+
"description": datasets.Value("string"),
|
| 101 |
+
"url_text_source": datasets.Value("string"),
|
| 102 |
+
"language": datasets.Value("string"),
|
| 103 |
+
"copyright_year": datasets.Value("string"),
|
| 104 |
+
"num_sections": datasets.Value("string"),
|
| 105 |
+
"url_rss": datasets.Value("string"),
|
| 106 |
+
"url_zip_file": datasets.Value("string"),
|
| 107 |
+
"url_project": datasets.Value("string"),
|
| 108 |
+
"url_librivox": datasets.Value("string"),
|
| 109 |
+
"url_other": datasets.Value("string"),
|
| 110 |
+
"totaltimesecs": datasets.Value("int64"),
|
| 111 |
+
"authors": datasets.Sequence(
|
| 112 |
+
{
|
| 113 |
+
"id": datasets.Value("string"),
|
| 114 |
+
"first_name": datasets.Value("string"),
|
| 115 |
+
"last_name": datasets.Value("string"),
|
| 116 |
+
"dob": datasets.Value("string"),
|
| 117 |
+
"dod": datasets.Value("string")
|
| 118 |
+
}
|
| 119 |
+
),
|
| 120 |
+
"genre": datasets.Sequence(datasets.Value("string")),
|
| 121 |
+
"Dramatic Readings": datasets.Value("bool"),
|
| 122 |
+
"meta_genre": datasets.Value("string"),
|
| 123 |
+
},
|
| 124 |
+
"snr": datasets.Value("float32"),
|
| 125 |
+
"voice_activity": datasets.Sequence(datasets.Sequence(datasets.Value("float32"))),
|
| 126 |
+
}
|
| 127 |
}
|
| 128 |
)
|
| 129 |
|
|
|
|
| 145 |
]
|
| 146 |
|
| 147 |
def _generate_examples(self, archive_path):
|
| 148 |
+
paths = glob.glob(os.path.join(archive_path, "small", "**", "**", "*.flac"))
|
| 149 |
for key, path in enumerate(paths):
|
| 150 |
path_split = path.split("/")
|
| 151 |
id_ = path_split[-1].replace(".flac", "")
|
| 152 |
speaker_id = int(path_split[-3])
|
| 153 |
+
path_json = path.replace(".flac", ".json")
|
| 154 |
+
with open(path_json) as json_file:
|
| 155 |
+
metadata = json.load(json_file)
|
| 156 |
yield key, {
|
| 157 |
"id": id_,
|
| 158 |
"file": path,
|
| 159 |
"audio": path,
|
| 160 |
"speaker_id": speaker_id,
|
| 161 |
+
"metadata": metadata,
|
| 162 |
}
|