Datasets:
Splend1dchan
commited on
Commit
•
71c47a6
1
Parent(s):
8bb33d1
Update librispeech_asr_individual.py
Browse files
librispeech_asr_individual.py
CHANGED
@@ -76,15 +76,18 @@ _DL_URLS = {
|
|
76 |
},
|
77 |
"test.other": {
|
78 |
"test.other": _DL_URL + "test-other.tar.gz",
|
79 |
-
}
|
80 |
"train.clean.100":{
|
81 |
"train.clean.100": _DL_URL + "train-clean-100.tar.gz",
|
82 |
-
}
|
83 |
"train.clean.360":{
|
84 |
"train.clean.360": _DL_URL + "train-clean-360.tar.gz",
|
85 |
-
}
|
86 |
"train.other.500":{
|
87 |
"train.other.500": _DL_URL + "train-other-500.tar.gz",
|
|
|
|
|
|
|
88 |
}
|
89 |
}
|
90 |
|
@@ -358,6 +361,18 @@ class LibrispeechASR(datasets.GeneratorBasedBuilder):
|
|
358 |
]
|
359 |
dev_splits = []
|
360 |
test_splits = []
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
361 |
return train_splits + dev_splits + test_splits
|
362 |
|
363 |
def _generate_examples(self, files, local_extracted_archive):
|
|
|
76 |
},
|
77 |
"test.other": {
|
78 |
"test.other": _DL_URL + "test-other.tar.gz",
|
79 |
+
},
|
80 |
"train.clean.100":{
|
81 |
"train.clean.100": _DL_URL + "train-clean-100.tar.gz",
|
82 |
+
},
|
83 |
"train.clean.360":{
|
84 |
"train.clean.360": _DL_URL + "train-clean-360.tar.gz",
|
85 |
+
},
|
86 |
"train.other.500":{
|
87 |
"train.other.500": _DL_URL + "train-other-500.tar.gz",
|
88 |
+
},
|
89 |
+
"train.10": {
|
90 |
+
"https://dl.fbaipublicfiles.com/librilight/data/librispeech_finetuning.tgz",
|
91 |
}
|
92 |
}
|
93 |
|
|
|
361 |
]
|
362 |
dev_splits = []
|
363 |
test_splits = []
|
364 |
+
elif self.config.name == "train.10":
|
365 |
+
train_splits = [
|
366 |
+
datasets.SplitGenerator(
|
367 |
+
name="train.10",
|
368 |
+
gen_kwargs={
|
369 |
+
"local_extracted_archive": local_extracted_archive.get("train.10"),
|
370 |
+
"files": dl_manager.iter_archive(archive_path["train.10"]),
|
371 |
+
},
|
372 |
+
),
|
373 |
+
]
|
374 |
+
dev_splits = []
|
375 |
+
test_splits = []
|
376 |
return train_splits + dev_splits + test_splits
|
377 |
|
378 |
def _generate_examples(self, files, local_extracted_archive):
|