SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

[Examples] Use Audio feature in speech classification (#14052)

Browse Source 
* Update SEW integration test tolerance

* Update audio classification

* Update test

* Remove torchaudio

* Add dataset revision

* Hub branch naming

* Revert dataset revisions

* Update datasets
This commit is contained in:
Anton Lozhkov
2021-10-20 12:22:43 +03:00
committed by GitHub
parent 3fefa292c1
commit dbaf49203e
6 changed files with 23 additions and 32 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
examples/pytorch/speech-pretraining/run_wav2vec2_pretraining_no_trainer.py
View File

@@ -113,7 +113,7 @@ def parse_args():
"--audio_column_name",
type=str,
default="audio",
help="Column in the dataset that contains speech file path. Defaults to 'file'",
help="Column in the dataset that contains speech file path. Defaults to 'audio'",
)
parser.add_argument(
"--model_name_or_path",
@@ -431,9 +431,9 @@ def main():
# via the `feature_extractor`
feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained(args.model_name_or_path)
# make sure that dataset decodes audio with correct samlping rate
# make sure that dataset decodes audio with correct sampling rate
raw_datasets = raw_datasets.cast_column(
"audio", datasets.features.Audio(sampling_rate=feature_extractor.sampling_rate)
args.audio_column_name, datasets.features.Audio(sampling_rate=feature_extractor.sampling_rate)
)
# only normalized-inputs-training is supported