Skip to content

GitLab

Unverified Commit f9663a7b authored by moto's avatar moto Committed by GitHub
Browse files

Add license to pre-trained model doc (#1836)

parent 33a655fd
Hide whitespace changes
Inline Side-by-side
Showing with 89 additions and 65 deletions
torchaudio/models/wav2vec2/pretrained.py
View file @ f9663a7b
...@@ -159,9 +159,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit ...@@ -159,9 +159,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit
(the combination of "train-clean-100", "train-clean-360", and "train-other-500"). (the combination of "train-clean-100", "train-clean-360", and "train-other-500").
Not fine-tuned. Not fine-tuned.
Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`]. Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`] under MIT License and
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] redistributed with the same license.
""" [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
`Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
WAV2VEC2_ASR_BASE_10M = Wav2Vec2PretrainedModelBundle( WAV2VEC2_ASR_BASE_10M = Wav2Vec2PretrainedModelBundle(
_path='wav2vec2_fairseq_base_ls960_asr_ll10m.pth', _path='wav2vec2_fairseq_base_ls960_asr_ll10m.pth',
...@@ -200,10 +202,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit ...@@ -200,10 +202,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit
fine-tuned for ASR on 10 minutes of transcribed audio from *Libri-Light* dataset fine-tuned for ASR on 10 minutes of transcribed audio from *Libri-Light* dataset
[:footcite:`librilight`] ("train-10min" subset). [:footcite:`librilight`] ("train-10min" subset).
Originally published by the authors of *wav2vec 2.0* Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`] under MIT License and
[:footcite:`baevski2020wav2vec`]. redistributed with the same license.
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
""" `Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
WAV2VEC2_ASR_BASE_100H = Wav2Vec2PretrainedModelBundle( WAV2VEC2_ASR_BASE_100H = Wav2Vec2PretrainedModelBundle(
'wav2vec2_fairseq_base_ls960_asr_ls100.pth', 'wav2vec2_fairseq_base_ls960_asr_ls100.pth',
...@@ -242,10 +245,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit ...@@ -242,10 +245,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit
(the combination of "train-clean-100", "train-clean-360", and "train-other-500"), and (the combination of "train-clean-100", "train-clean-360", and "train-other-500"), and
fine-tuned for ASR on 100 hours of transcribed audio from "train-clean-100" subset. fine-tuned for ASR on 100 hours of transcribed audio from "train-clean-100" subset.
Originally published by the authors of *wav2vec 2.0* Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`] under MIT License and
[:footcite:`baevski2020wav2vec`]. redistributed with the same license.
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
""" `Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
WAV2VEC2_ASR_BASE_960H = Wav2Vec2PretrainedModelBundle( WAV2VEC2_ASR_BASE_960H = Wav2Vec2PretrainedModelBundle(
'wav2vec2_fairseq_base_ls960_asr_ls960.pth', 'wav2vec2_fairseq_base_ls960_asr_ls960.pth',
...@@ -283,10 +287,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit ...@@ -283,10 +287,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit
(the combination of "train-clean-100", "train-clean-360", and "train-other-500"), and (the combination of "train-clean-100", "train-clean-360", and "train-other-500"), and
fine-tuned for ASR on the same audio with the corresponding transcripts. fine-tuned for ASR on the same audio with the corresponding transcripts.
Originally published by the authors of *wav2vec 2.0* Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`] under MIT License and
[:footcite:`baevski2020wav2vec`]. redistributed with the same license.
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
""" `Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
WAV2VEC2_LARGE = Wav2Vec2PretrainedModelBundle( WAV2VEC2_LARGE = Wav2Vec2PretrainedModelBundle(
'wav2vec2_fairseq_large_ls960.pth', 'wav2vec2_fairseq_large_ls960.pth',
...@@ -324,10 +329,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit ...@@ -324,10 +329,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit
(the combination of "train-clean-100", "train-clean-360", and "train-other-500"). (the combination of "train-clean-100", "train-clean-360", and "train-other-500").
Not fine-tuned. Not fine-tuned.
Originally published by the authors of *wav2vec 2.0* Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`] under MIT License and
[:footcite:`baevski2020wav2vec`]. redistributed with the same license.
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
""" `Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
WAV2VEC2_ASR_LARGE_10M = Wav2Vec2PretrainedModelBundle( WAV2VEC2_ASR_LARGE_10M = Wav2Vec2PretrainedModelBundle(
'wav2vec2_fairseq_large_ls960_asr_ll10m.pth', 'wav2vec2_fairseq_large_ls960_asr_ll10m.pth',
...@@ -366,10 +372,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit ...@@ -366,10 +372,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit
fine-tuned for ASR on 10 minutes of transcribed audio from *Libri-Light* dataset fine-tuned for ASR on 10 minutes of transcribed audio from *Libri-Light* dataset
[:footcite:`librilight`] ("train-10min" subset). [:footcite:`librilight`] ("train-10min" subset).
Originally published by the authors of *wav2vec 2.0* Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`] under MIT License and
[:footcite:`baevski2020wav2vec`]. redistributed with the same license.
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
""" `Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
WAV2VEC2_ASR_LARGE_100H = Wav2Vec2PretrainedModelBundle( WAV2VEC2_ASR_LARGE_100H = Wav2Vec2PretrainedModelBundle(
'wav2vec2_fairseq_large_ls960_asr_ls100.pth', 'wav2vec2_fairseq_large_ls960_asr_ls100.pth',
...@@ -408,10 +415,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit ...@@ -408,10 +415,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit
fine-tuned for ASR on 100 hours of transcribed audio from fine-tuned for ASR on 100 hours of transcribed audio from
the same dataset ("train-clean-100" subset). the same dataset ("train-clean-100" subset).
Originally published by the authors of *wav2vec 2.0* Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`] under MIT License and
[:footcite:`baevski2020wav2vec`]. redistributed with the same license.
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
""" `Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
WAV2VEC2_ASR_LARGE_960H = Wav2Vec2PretrainedModelBundle( WAV2VEC2_ASR_LARGE_960H = Wav2Vec2PretrainedModelBundle(
'wav2vec2_fairseq_large_ls960_asr_ls960.pth', 'wav2vec2_fairseq_large_ls960_asr_ls960.pth',
...@@ -449,10 +457,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit ...@@ -449,10 +457,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit
(the combination of "train-clean-100", "train-clean-360", and "train-other-500"), and (the combination of "train-clean-100", "train-clean-360", and "train-other-500"), and
fine-tuned for ASR on the same audio with the corresponding transcripts. fine-tuned for ASR on the same audio with the corresponding transcripts.
Originally published by the authors of *wav2vec 2.0* Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`] under MIT License and
[:footcite:`baevski2020wav2vec`]. redistributed with the same license.
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
""" `Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
WAV2VEC2_LARGE_LV60K = Wav2Vec2PretrainedModelBundle( WAV2VEC2_LARGE_LV60K = Wav2Vec2PretrainedModelBundle(
'wav2vec2_fairseq_large_lv60k.pth', 'wav2vec2_fairseq_large_lv60k.pth',
...@@ -490,10 +499,11 @@ Pre-trained on 60,000 hours of unlabeled audio from ...@@ -490,10 +499,11 @@ Pre-trained on 60,000 hours of unlabeled audio from
*Libri-Light* dataset [:footcite:`librilight`]. *Libri-Light* dataset [:footcite:`librilight`].
Not fine-tuned. Not fine-tuned.
Originally published by the authors of *wav2vec 2.0* Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`] under MIT License and
[:footcite:`baevski2020wav2vec`]. redistributed with the same license.
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
""" `Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
WAV2VEC2_ASR_LARGE_LV60K_10M = Wav2Vec2PretrainedModelBundle( WAV2VEC2_ASR_LARGE_LV60K_10M = Wav2Vec2PretrainedModelBundle(
'wav2vec2_fairseq_large_lv60k_asr_ll10m.pth', 'wav2vec2_fairseq_large_lv60k_asr_ll10m.pth',
...@@ -532,10 +542,11 @@ Pre-trained on 60,000 hours of unlabeled audio from ...@@ -532,10 +542,11 @@ Pre-trained on 60,000 hours of unlabeled audio from
fine-tuned for ASR on 10 minutes of transcribed audio from fine-tuned for ASR on 10 minutes of transcribed audio from
the same dataset ("train-10min" subset). the same dataset ("train-10min" subset).
Originally published by the authors of *wav2vec 2.0* Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`] under MIT License and
[:footcite:`baevski2020wav2vec`]. redistributed with the same license.
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
""" `Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
WAV2VEC2_ASR_LARGE_LV60K_100H = Wav2Vec2PretrainedModelBundle( WAV2VEC2_ASR_LARGE_LV60K_100H = Wav2Vec2PretrainedModelBundle(
'wav2vec2_fairseq_large_lv60k_asr_ls100.pth', 'wav2vec2_fairseq_large_lv60k_asr_ls100.pth',
...@@ -574,10 +585,11 @@ Pre-trained on 60,000 hours of unlabeled audio from ...@@ -574,10 +585,11 @@ Pre-trained on 60,000 hours of unlabeled audio from
fine-tuned for ASR on 100 hours of transcribed audio from fine-tuned for ASR on 100 hours of transcribed audio from
*LibriSpeech* dataset [:footcite:`7178964`] ("train-clean-100" subset). *LibriSpeech* dataset [:footcite:`7178964`] ("train-clean-100" subset).
Originally published by the authors of *wav2vec 2.0* Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`] under MIT License and
[:footcite:`baevski2020wav2vec`]. redistributed with the same license.
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
""" `Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
WAV2VEC2_ASR_LARGE_LV60K_960H = Wav2Vec2PretrainedModelBundle( WAV2VEC2_ASR_LARGE_LV60K_960H = Wav2Vec2PretrainedModelBundle(
'wav2vec2_fairseq_large_lv60k_asr_ls960.pth', 'wav2vec2_fairseq_large_lv60k_asr_ls960.pth',
...@@ -617,10 +629,11 @@ fine-tuned for ASR on 960 hours of transcribed audio from ...@@ -617,10 +629,11 @@ fine-tuned for ASR on 960 hours of transcribed audio from
*LibriSpeech* dataset [:footcite:`7178964`] *LibriSpeech* dataset [:footcite:`7178964`]
(the combination of "train-clean-100", "train-clean-360", and "train-other-500"). (the combination of "train-clean-100", "train-clean-360", and "train-other-500").
Originally published by the authors of *wav2vec 2.0* Originally published by the authors of *wav2vec 2.0* [:footcite:`baevski2020wav2vec`] under MIT License and
[:footcite:`baevski2020wav2vec`]. redistributed with the same license.
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
""" `Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
WAV2VEC2_XLSR53 = Wav2Vec2PretrainedModelBundle( WAV2VEC2_XLSR53 = Wav2Vec2PretrainedModelBundle(
'wav2vec2_fairseq_large_xlsr53.pth', 'wav2vec2_fairseq_large_xlsr53.pth',
...@@ -662,9 +675,10 @@ Not fine-tuned. ...@@ -662,9 +675,10 @@ Not fine-tuned.
Originally published by the authors of Originally published by the authors of
*Unsupervised Cross-lingual Representation Learning for Speech Recognition* *Unsupervised Cross-lingual Representation Learning for Speech Recognition*
[:footcite:`conneau2020unsupervised`]. [:footcite:`conneau2020unsupervised`] under MIT License and redistributed with the same license.
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/wav2vec#pre-trained-models>`__] [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
""" `Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/wav2vec#pre-trained-models>`__]
""" # noqa: E501
HUBERT_BASE = Wav2Vec2PretrainedModelBundle( HUBERT_BASE = Wav2Vec2PretrainedModelBundle(
'hubert_fairseq_base_ls960.pth', 'hubert_fairseq_base_ls960.pth',
...@@ -702,9 +716,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit ...@@ -702,9 +716,11 @@ Pre-trained on 960 hours of unlabeled audio from *LibriSpeech* dataset [:footcit
(the combination of "train-clean-100", "train-clean-360", and "train-other-500"). (the combination of "train-clean-100", "train-clean-360", and "train-other-500").
Not fine-tuned. Not fine-tuned.
Originally published by the authors of *HuBERT* [:footcite:`hsu2021hubert`]. Originally published by the authors of *HuBERT* [:footcite:`hsu2021hubert`] under MIT License and
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/hubert#pre-trained-and-fine-tuned-asr-models>`__] redistributed with the same license.
""" [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
`Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/hubert#pre-trained-and-fine-tuned-asr-models>`__]
""" # noqa: E501
HUBERT_LARGE = Wav2Vec2PretrainedModelBundle( HUBERT_LARGE = Wav2Vec2PretrainedModelBundle(
'hubert_fairseq_large_ll60k.pth', 'hubert_fairseq_large_ll60k.pth',
...@@ -742,9 +758,11 @@ Pre-trained on 60,000 hours of unlabeled audio from ...@@ -742,9 +758,11 @@ Pre-trained on 60,000 hours of unlabeled audio from
*Libri-Light* dataset [:footcite:`librilight`]. *Libri-Light* dataset [:footcite:`librilight`].
Not fine-tuned. Not fine-tuned.
Originally published by the authors of *HuBERT* [:footcite:`hsu2021hubert`]. Originally published by the authors of *HuBERT* [:footcite:`hsu2021hubert`] under MIT License and
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/hubert#pre-trained-and-fine-tuned-asr-models>`__] redistributed with the same license.
""" [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
`Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/hubert#pre-trained-and-fine-tuned-asr-models>`__]
""" # noqa: E501
HUBERT_XLARGE = Wav2Vec2PretrainedModelBundle( HUBERT_XLARGE = Wav2Vec2PretrainedModelBundle(
'hubert_fairseq_xlarge_ll60k.pth', 'hubert_fairseq_xlarge_ll60k.pth',
...@@ -782,9 +800,11 @@ Pre-trained on 60,000 hours of unlabeled audio from ...@@ -782,9 +800,11 @@ Pre-trained on 60,000 hours of unlabeled audio from
*Libri-Light* dataset [:footcite:`librilight`]. *Libri-Light* dataset [:footcite:`librilight`].
Not fine-tuned. Not fine-tuned.
Originally published by the authors of *HuBERT* [:footcite:`hsu2021hubert`]. Originally published by the authors of *HuBERT* [:footcite:`hsu2021hubert`] under MIT License and
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/hubert#pre-trained-and-fine-tuned-asr-models>`__] redistributed with the same license.
""" [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
`Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/hubert#pre-trained-and-fine-tuned-asr-models>`__]
""" # noqa: E501
HUBERT_ASR_LARGE = Wav2Vec2PretrainedModelBundle( HUBERT_ASR_LARGE = Wav2Vec2PretrainedModelBundle(
'hubert_fairseq_large_ll60k_asr_ls960.pth', 'hubert_fairseq_large_ll60k_asr_ls960.pth',
...@@ -824,9 +844,11 @@ fine-tuned for ASR on 960 hours of transcribed audio from ...@@ -824,9 +844,11 @@ fine-tuned for ASR on 960 hours of transcribed audio from
*LibriSpeech* dataset [:footcite:`7178964`] *LibriSpeech* dataset [:footcite:`7178964`]
(the combination of "train-clean-100", "train-clean-360", and "train-other-500"). (the combination of "train-clean-100", "train-clean-360", and "train-other-500").
Originally published by the authors of *HuBERT* [:footcite:`hsu2021hubert`]. Originally published by the authors of *HuBERT* [:footcite:`hsu2021hubert`] under MIT License and
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/hubert#pre-trained-and-fine-tuned-asr-models>`__] redistributed with the same license.
""" [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
`Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/hubert#pre-trained-and-fine-tuned-asr-models>`__]
""" # noqa: E501
HUBERT_ASR_XLARGE = Wav2Vec2PretrainedModelBundle( HUBERT_ASR_XLARGE = Wav2Vec2PretrainedModelBundle(
'hubert_fairseq_xlarge_ll60k_asr_ls960.pth', 'hubert_fairseq_xlarge_ll60k_asr_ls960.pth',
...@@ -866,6 +888,8 @@ fine-tuned for ASR on 960 hours of transcribed audio from ...@@ -866,6 +888,8 @@ fine-tuned for ASR on 960 hours of transcribed audio from
*LibriSpeech* dataset [:footcite:`7178964`] *LibriSpeech* dataset [:footcite:`7178964`]
(the combination of "train-clean-100", "train-clean-360", and "train-other-500"). (the combination of "train-clean-100", "train-clean-360", and "train-other-500").
Originally published by the authors of *HuBERT* [:footcite:`hsu2021hubert`]. Originally published by the authors of *HuBERT* [:footcite:`hsu2021hubert`] under MIT License and
[`Source <https://github.com/pytorch/fairseq/tree/main/examples/hubert#pre-trained-and-fine-tuned-asr-models>`__] redistributed with the same license.
""" [`License <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/LICENSE>`__,
`Source <https://github.com/pytorch/fairseq/blob/ce6c9eeae163ac04b79539c78e74f292f29eaa18/examples/hubert#pre-trained-and-fine-tuned-asr-models>`__]
""" # noqa: E501
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment