Move ASR features out of prototype (#2187)

Summary:
Moves ASR features out of `torchaudio.prototype`. Specifically, merges contents of `torchaudio.prototype.models` into `torchaudio.models` and contents of `torchaudio.prototype.pipelines` into `torchaudio.pipelines` and updates refs, tests, and docs accordingly.

Pull Request resolved: https://github.com/pytorch/audio/pull/2187

Reviewed By: nateanl, mthrok

Differential Revision: D33918092

Pulled By: hwangjeff

fbshipit-source-id: f003f289a7e5d7d43f85b7c270b58bdf2ed6344c

test/torchaudio_unittest/prototype/rnnt_decoder_cpu_test.py → test/torchaudio_unittest/models/rnnt_decoder/rnnt_decoder_cpu_test.py

 import torch
 from torchaudio_unittest.common_utils import PytorchTestCase
-from torchaudio_unittest.prototype.rnnt_decoder_test_impl import RNNTBeamSearchTestImpl
+from torchaudio_unittest.models.rnnt_decoder.rnnt_decoder_test_impl import RNNTBeamSearchTestImpl
 
 
 class RNNTBeamSearchFloat32CPUTest(RNNTBeamSearchTestImpl, PytorchTestCase):

test/torchaudio_unittest/prototype/rnnt_decoder_gpu_test.py → test/torchaudio_unittest/models/rnnt_decoder/rnnt_decoder_gpu_test.py

 import torch
 from torchaudio_unittest.common_utils import skipIfNoCuda, PytorchTestCase
-from torchaudio_unittest.prototype.rnnt_decoder_test_impl import RNNTBeamSearchTestImpl
+from torchaudio_unittest.models.rnnt_decoder.rnnt_decoder_test_impl import RNNTBeamSearchTestImpl
 
 
 @skipIfNoCuda

test/torchaudio_unittest/prototype/rnnt_decoder_test_impl.py → test/torchaudio_unittest/models/rnnt_decoder/rnnt_decoder_test_impl.py

 import torch
-from torchaudio.prototype.models import RNNTBeamSearch, emformer_rnnt_model
+from torchaudio.models import RNNTBeamSearch, emformer_rnnt_model
 from torchaudio_unittest.common_utils import TestBaseMixin, torch_script
 
 

torchaudio/models/__init__.py

+from .conformer import Conformer
 from .conv_tasnet import ConvTasNet
 from .deepspeech import DeepSpeech
+from .emformer import Emformer
+from .rnnt import RNNT, emformer_rnnt_base, emformer_rnnt_model
+from .rnnt_decoder import Hypothesis, RNNTBeamSearch
 from .tacotron2 import Tacotron2
 from .wav2letter import Wav2Letter
 from .wav2vec2 import (
@@ -19,6 +23,7 @@ from .wav2vec2 import (
 )
 from .wavernn import WaveRNN
 
+
 __all__ = [
     "Wav2Letter",
     "WaveRNN",
@@ -38,4 +43,11 @@ __all__ = [
     "hubert_pretrain_large",
     "hubert_pretrain_xlarge",
     "Tacotron2",
+    "Conformer",
+    "Emformer",
+    "Hypothesis",
+    "RNNT",
+    "RNNTBeamSearch",
+    "emformer_rnnt_base",
+    "emformer_rnnt_model",
 ]

torchaudio/prototype/models/rnnt.py → torchaudio/models/rnnt.py

 from typing import List, Optional, Tuple
 
 import torch
-from torchaudio.prototype.models import Emformer
+from torchaudio.models import Emformer
 
 
 __all__ = ["RNNT", "emformer_rnnt_base", "emformer_rnnt_model"]
@@ -278,7 +278,7 @@ class _Transcriber(torch.nn.Module):
 
 
 class _Predictor(torch.nn.Module):
-    r"""Recurrent neural network transducer (RNN-T) transcription network.
+    r"""Recurrent neural network transducer (RNN-T) prediction network.
 
     Args:
         num_symbols (int): size of target token lexicon.
@@ -425,7 +425,7 @@ class _Joiner(torch.nn.Module):
 
 
 class RNNT(torch.nn.Module):
-    r"""torchaudio.prototype.models.RNNT()
+    r"""torchaudio.models.RNNT()
 
     Recurrent neural network transducer (RNN-T) model.
 

torchaudio/prototype/models/rnnt_decoder.py → torchaudio/models/rnnt_decoder.py

 from typing import Callable, Dict, List, Optional, NamedTuple, Tuple
 
 import torch
-from torchaudio.prototype.models import RNNT
+from torchaudio.models import RNNT
 
 
 __all__ = ["Hypothesis", "RNNTBeamSearch"]

torchaudio/pipelines/__init__.py

@@ -32,6 +32,8 @@ from ._wav2vec2.impl import (
     HUBERT_ASR_LARGE,
     HUBERT_ASR_XLARGE,
 )
+from .rnnt_pipeline import EMFORMER_RNNT_BASE_LIBRISPEECH, RNNTBundle
+
 
 __all__ = [
     "Wav2Vec2Bundle",
@@ -64,4 +66,6 @@ __all__ = [
     "TACOTRON2_GRIFFINLIM_PHONE_LJSPEECH",
     "TACOTRON2_WAVERNN_CHAR_LJSPEECH",
     "TACOTRON2_WAVERNN_PHONE_LJSPEECH",
+    "RNNTBundle",
+    "EMFORMER_RNNT_BASE_LIBRISPEECH",
 ]

torchaudio/prototype/pipelines/rnnt_pipeline.py → torchaudio/pipelines/rnnt_pipeline.py

@@ -9,7 +9,7 @@ from typing import Callable, List, Tuple
 import torch
 import torchaudio
 from torchaudio._internal import download_url_to_file, load_state_dict_from_url, module_utils
-from torchaudio.prototype.models import RNNT, RNNTBeamSearch, emformer_rnnt_base
+from torchaudio.models import RNNT, RNNTBeamSearch, emformer_rnnt_base
 
 
 __all__ = []
@@ -158,7 +158,7 @@ class _SentencePieceTokenProcessor(_TokenProcessor):
 
 @dataclass
 class RNNTBundle:
-    """torchaudio.prototype.pipelines.RNNTBundle()
+    """torchaudio.pipelines.RNNTBundle()
 
     Dataclass that bundles components for performing automatic speech recognition (ASR, speech-to-text)
     inference with an RNN-T model.
@@ -176,7 +176,7 @@ class RNNTBundle:
 
     Example
         >>> import torchaudio
-        >>> from torchaudio.prototype.pipelines import EMFORMER_RNNT_BASE_LIBRISPEECH
+        >>> from torchaudio.pipelines import EMFORMER_RNNT_BASE_LIBRISPEECH
         >>> import torch
         >>>
         >>> # Non-streaming inference.
@@ -379,7 +379,7 @@ EMFORMER_RNNT_BASE_LIBRISPEECH = RNNTBundle(
 )
 EMFORMER_RNNT_BASE_LIBRISPEECH.__doc__ = """Pre-trained Emformer-RNNT-based ASR pipeline capable of performing both streaming and non-streaming inference.
 
-    The underlying model is constructed by :py:func:`torchaudio.prototype.models.emformer_rnnt_base`
+    The underlying model is constructed by :py:func:`torchaudio.models.emformer_rnnt_base`
     and utilizes weights trained on LibriSpeech using training script ``train.py``
     `here <https://github.com/pytorch/audio/tree/main/examples/asr/librispeech_emformer_rnnt>`__ with default arguments.
 

torchaudio/prototype/models/__init__.py

-from .conformer import Conformer
-from .emformer import Emformer
-from .rnnt import RNNT, emformer_rnnt_base, emformer_rnnt_model
-from .rnnt_decoder import Hypothesis, RNNTBeamSearch
-
-
-__all__ = [
-    "Conformer",
-    "Emformer",
-    "Hypothesis",
-    "RNNT",
-    "RNNTBeamSearch",
-    "emformer_rnnt_base",
-    "emformer_rnnt_model",
-]

torchaudio/prototype/pipelines/__init__.py

-from .rnnt_pipeline import EMFORMER_RNNT_BASE_LIBRISPEECH, RNNTBundle
-
-
-__all__ = [
-    "EMFORMER_RNNT_BASE_LIBRISPEECH",
-    "RNNTBundle",
-]