1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
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
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
|
.. py:module:: torchaudio.pipelines
torchaudio.pipelines
====================
.. currentmodule:: torchaudio.pipelines
The ``torchaudio.pipelines`` module packages pre-trained models with support functions and meta-data into simple APIs tailored to perform specific tasks.
When using pre-trained models to perform a task, in addition to instantiating the model with pre-trained weights, the client code also needs to build pipelines for feature extractions and post processing in the same way they were done during the training. This requires to carrying over information used during the training, such as the type of transforms and the their parameters (for example, sampling rate the number of FFT bins).
To make this information tied to a pre-trained model and easily accessible, ``torchaudio.pipelines`` module uses the concept of a `Bundle` class, which defines a set of APIs to instantiate pipelines, and the interface of the pipelines.
The following figure illustrates this.
.. image:: https://download.pytorch.org/torchaudio/doc-assets/pipelines-intro.png
A pre-trained model and associated pipelines are expressed as an instance of ``Bundle``. Different instances of same ``Bundle`` share the interface, but their implementations are not constrained to be of same types. For example, :class:`SourceSeparationBundle` defines the interface for performing source separation, but its instance :data:`CONVTASNET_BASE_LIBRI2MIX` instantiates a model of :class:`~torchaudio.models.ConvTasNet` while :data:`HDEMUCS_HIGH_MUSDB` instantiates a model of :class:`~torchaudio.models.HDemucs`. Still, because they share the same interface, the usage is the same.
.. note::
Under the hood, the implementations of ``Bundle`` use components from other ``torchaudio`` modules, such as :mod:`torchaudio.models` and :mod:`torchaudio.transforms`, or even third party libraries like `SentencPiece <https://github.com/google/sentencepiece>`__ and `DeepPhonemizer <https://github.com/as-ideas/DeepPhonemizer>`__. But this implementation detail is abstracted away from library users.
.. _RNNT:
RNN-T Streaming/Non-Streaming ASR
---------------------------------
Interface
~~~~~~~~~
``RNNTBundle`` defines ASR pipelines and consists of three steps: feature extraction, inference, and de-tokenization.
.. image:: https://download.pytorch.org/torchaudio/doc-assets/pipelines-rnntbundle.png
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_class.rst
RNNTBundle
RNNTBundle.FeatureExtractor
RNNTBundle.TokenProcessor
.. rubric:: Tutorials using ``RNNTBundle``
.. minigallery:: torchaudio.pipelines.RNNTBundle
Pretrained Models
~~~~~~~~~~~~~~~~~
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_data.rst
EMFORMER_RNNT_BASE_LIBRISPEECH
wav2vec 2.0 / HuBERT / WavLM - SSL
----------------------------------
Interface
~~~~~~~~~
``Wav2Vec2Bundle`` instantiates models that generate acoustic features that can be used for downstream inference and fine-tuning.
.. image:: https://download.pytorch.org/torchaudio/doc-assets/pipelines-wav2vec2bundle.png
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_class.rst
Wav2Vec2Bundle
Pretrained Models
~~~~~~~~~~~~~~~~~
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_data.rst
WAV2VEC2_BASE
WAV2VEC2_LARGE
WAV2VEC2_LARGE_LV60K
WAV2VEC2_XLSR53
WAV2VEC2_XLSR_300M
WAV2VEC2_XLSR_1B
WAV2VEC2_XLSR_2B
HUBERT_BASE
HUBERT_LARGE
HUBERT_XLARGE
WAVLM_BASE
WAVLM_BASE_PLUS
WAVLM_LARGE
wav2vec 2.0 / HuBERT - Fine-tuned ASR
-------------------------------------
Interface
~~~~~~~~~
``Wav2Vec2ASRBundle`` instantiates models that generate probability distribution over pre-defined labels, that can be used for ASR.
.. image:: https://download.pytorch.org/torchaudio/doc-assets/pipelines-wav2vec2asrbundle.png
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_class.rst
Wav2Vec2ASRBundle
.. rubric:: Tutorials using ``Wav2Vec2ASRBundle``
.. minigallery:: torchaudio.pipelines.Wav2Vec2ASRBundle
Pretrained Models
~~~~~~~~~~~~~~~~~
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_data.rst
WAV2VEC2_ASR_BASE_10M
WAV2VEC2_ASR_BASE_100H
WAV2VEC2_ASR_BASE_960H
WAV2VEC2_ASR_LARGE_10M
WAV2VEC2_ASR_LARGE_100H
WAV2VEC2_ASR_LARGE_960H
WAV2VEC2_ASR_LARGE_LV60K_10M
WAV2VEC2_ASR_LARGE_LV60K_100H
WAV2VEC2_ASR_LARGE_LV60K_960H
VOXPOPULI_ASR_BASE_10K_DE
VOXPOPULI_ASR_BASE_10K_EN
VOXPOPULI_ASR_BASE_10K_ES
VOXPOPULI_ASR_BASE_10K_FR
VOXPOPULI_ASR_BASE_10K_IT
HUBERT_ASR_LARGE
HUBERT_ASR_XLARGE
wav2vec 2.0 / HuBERT - Forced Alignment
---------------------------------------
Interface
~~~~~~~~~
``Wav2Vec2FABundle`` bundles pre-trained model and its associated dictionary. Additionally, it supports appending ``star`` token dimension.
.. image:: https://download.pytorch.org/torchaudio/doc-assets/pipelines-wav2vec2fabundle.png
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_class.rst
Wav2Vec2FABundle
Wav2Vec2FABundle.Tokenizer
Wav2Vec2FABundle.Aligner
.. rubric:: Tutorials using ``Wav2Vec2FABundle``
.. minigallery:: torchaudio.pipelines.Wav2Vec2FABundle
Pertrained Models
~~~~~~~~~~~~~~~~~
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_data.rst
MMS_FA
.. _Tacotron2:
Tacotron2 Text-To-Speech
------------------------
``Tacotron2TTSBundle`` defines text-to-speech pipelines and consists of three steps: tokenization, spectrogram generation and vocoder. The spectrogram generation is based on :class:`~torchaudio.models.Tacotron2` model.
.. image:: https://download.pytorch.org/torchaudio/doc-assets/pipelines-tacotron2bundle.png
``TextProcessor`` can be rule-based tokenization in the case of characters, or it can be a neural-netowrk-based G2P model that generates sequence of phonemes from input text.
Similarly ``Vocoder`` can be an algorithm without learning parameters, like `Griffin-Lim`, or a neural-network-based model like `Waveglow`.
Interface
~~~~~~~~~
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_class.rst
Tacotron2TTSBundle
Tacotron2TTSBundle.TextProcessor
Tacotron2TTSBundle.Vocoder
.. rubric:: Tutorials using ``Tacotron2TTSBundle``
.. minigallery:: torchaudio.pipelines.Tacotron2TTSBundle
Pretrained Models
~~~~~~~~~~~~~~~~~
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_data.rst
TACOTRON2_WAVERNN_PHONE_LJSPEECH
TACOTRON2_WAVERNN_CHAR_LJSPEECH
TACOTRON2_GRIFFINLIM_PHONE_LJSPEECH
TACOTRON2_GRIFFINLIM_CHAR_LJSPEECH
Source Separation
-----------------
Interface
~~~~~~~~~
``SourceSeparationBundle`` instantiates source separation models which take single channel audio and generates multi-channel audio.
.. image:: https://download.pytorch.org/torchaudio/doc-assets/pipelines-sourceseparationbundle.png
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_class.rst
SourceSeparationBundle
.. rubric:: Tutorials using ``SourceSeparationBundle``
.. minigallery:: torchaudio.pipelines.SourceSeparationBundle
Pretrained Models
~~~~~~~~~~~~~~~~~
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_data.rst
CONVTASNET_BASE_LIBRI2MIX
HDEMUCS_HIGH_MUSDB_PLUS
HDEMUCS_HIGH_MUSDB
Squim Objective
---------------
Interface
~~~~~~~~~
:py:class:`SquimObjectiveBundle` defines speech quality and intelligibility measurement (SQUIM) pipeline that can predict **objecive** metric scores given the input waveform.
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_class.rst
SquimObjectiveBundle
Pretrained Models
~~~~~~~~~~~~~~~~~
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_data.rst
SQUIM_OBJECTIVE
Squim Subjective
----------------
Interface
~~~~~~~~~
:py:class:`SquimSubjectiveBundle` defines speech quality and intelligibility measurement (SQUIM) pipeline that can predict **subjective** metric scores given the input waveform.
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_class.rst
SquimSubjectiveBundle
Pretrained Models
~~~~~~~~~~~~~~~~~
.. autosummary::
:toctree: generated
:nosignatures:
:template: autosummary/bundle_data.rst
SQUIM_SUBJECTIVE
|
