Rename factory functions `wav2vec2_asr_ARCH` to `wav2vec2_ft_ARCH` (#1804)

* Rename factory functions `wav2vec2_asr_ARCH` to `wav2vec2_ft_ARCH`

In #1783, we split the factory functions of wav2vec2 into ones for pretraining models
and ones for fine-tuning models (pretraining model + extra Linear module).

I picked the name scheme `wav2vec2_asr_ARCH` for factory functions of fine-tuning models,
but did not feel right, because the architecture code is more generic.
Even though the resulting model architecture was used for  ASR fine-tuning in the paper, 
it does not have to be ASR.
This became more evident as we add pre-trained parameters support, such as #1799.
It matters more for the weight files that for which task and on which dataset it was
trained on. For factory function, ASR task is not relevant.

Therefore renaming the functions by replacing `_asr_` to `_ft_` fine-tuning.

Note: Since the new functions are not release yet, this PR itself is not BC-breaking.

docs/source/models.rst

@@ -90,20 +90,20 @@ wav2vec2_large_lv60k
 
 .. autofunction:: wav2vec2_large_lv60k
 
-wav2vec2_asr_base
+wav2vec2_ft_base
 ^^^^^^^^^^^^^^^^^
 
-.. autofunction:: wav2vec2_asr_base
+.. autofunction:: wav2vec2_ft_base
 
-wav2vec2_asr_large
+wav2vec2_ft_large
 ^^^^^^^^^^^^^^^^^^
 
-.. autofunction:: wav2vec2_asr_large
+.. autofunction:: wav2vec2_ft_large
 
-wav2vec2_asr_large_lv60k
+wav2vec2_ft_large_lv60k
 ^^^^^^^^^^^^^^^^^^^^^^^^
 
-.. autofunction:: wav2vec2_asr_large_lv60k
+.. autofunction:: wav2vec2_ft_large_lv60k
 
 hubert_base
 ^^^^^^^^^^^
@@ -120,15 +120,15 @@ hubert_xlarge
 
 .. autofunction:: hubert_xlarge
 
-hubert_asr_large
+hubert_ft_large
 ^^^^^^^^^^^^^^^^
 
-.. autofunction:: hubert_asr_large
+.. autofunction:: hubert_ft_large
 
-hubert_asr_xlarge
+hubert_ft_xlarge
 ^^^^^^^^^^^^^^^^^
 
-.. autofunction:: hubert_asr_xlarge
+.. autofunction:: hubert_ft_xlarge
 
 .. currentmodule:: torchaudio.models.wav2vec2.utils
 

test/torchaudio_unittest/models/wav2vec2/fairseq_integration_test.py

@@ -3,17 +3,17 @@ import sys
 
 import torch
 from torchaudio.models.wav2vec2 import (
-    wav2vec2_asr_base,
-    wav2vec2_asr_large,
-    wav2vec2_asr_large_lv60k,
+    wav2vec2_ft_base,
+    wav2vec2_ft_large,
+    wav2vec2_ft_large_lv60k,
     wav2vec2_base,
     wav2vec2_large,
     wav2vec2_large_lv60k,
     hubert_base,
     hubert_large,
     hubert_xlarge,
-    hubert_asr_large,
-    hubert_asr_xlarge,
+    hubert_ft_large,
+    hubert_ft_xlarge,
 )
 from torchaudio.models.wav2vec2.utils import (
     import_fairseq_model,
@@ -75,12 +75,12 @@ ALL_PRETRAINING_CONFIGS = parameterized.expand([
     (HUBERT_XLARGE_LL60K, hubert_xlarge),
 ], name_func=_name_func)
 FINETUNING_CONFIGS = parameterized.expand([
-    (WAV2VEC2_BASE_960H, wav2vec2_asr_base),
-    (WAV2VEC2_LARGE_960H, wav2vec2_asr_large),
-    (WAV2VEC2_LARGE_LV60K_960H, wav2vec2_asr_large_lv60k),
-    (WAV2VEC2_LARGE_LV60K_SELF_960H, wav2vec2_asr_large_lv60k),
-    (HUBERT_LARGE, hubert_asr_large),
-    (HUBERT_XLARGE, hubert_asr_xlarge),
+    (WAV2VEC2_BASE_960H, wav2vec2_ft_base),
+    (WAV2VEC2_LARGE_960H, wav2vec2_ft_large),
+    (WAV2VEC2_LARGE_LV60K_960H, wav2vec2_ft_large_lv60k),
+    (WAV2VEC2_LARGE_LV60K_SELF_960H, wav2vec2_ft_large_lv60k),
+    (HUBERT_LARGE, hubert_ft_large),
+    (HUBERT_XLARGE, hubert_ft_xlarge),
 ], name_func=_name_func)
 
 

test/torchaudio_unittest/models/wav2vec2/huggingface_intergration_test.py

@@ -2,9 +2,9 @@ import json
 
 import torch
 from torchaudio.models.wav2vec2 import (
-    wav2vec2_asr_base,
-    wav2vec2_asr_large,
-    wav2vec2_asr_large_lv60k,
+    wav2vec2_ft_base,
+    wav2vec2_ft_large,
+    wav2vec2_ft_large_lv60k,
     wav2vec2_base,
     wav2vec2_large,
     wav2vec2_large_lv60k,
@@ -50,11 +50,11 @@ PRETRAIN_CONFIGS = parameterized.expand([
     (HF_BASE_10K_VOXPOPULI, wav2vec2_base),
 ], name_func=_name_func)
 FINETUNE_CONFIGS = parameterized.expand([
-    (HF_BASE_960H, wav2vec2_asr_base),
-    (HF_LARGE_960H, wav2vec2_asr_large),
-    (HF_LARGE_LV60_960H, wav2vec2_asr_large_lv60k),
-    (HF_LARGE_LV60_SELF_960H, wav2vec2_asr_large_lv60k),
-    (HF_LARGE_XLSR_DE, wav2vec2_asr_large_lv60k),
+    (HF_BASE_960H, wav2vec2_ft_base),
+    (HF_LARGE_960H, wav2vec2_ft_large),
+    (HF_LARGE_LV60_960H, wav2vec2_ft_large_lv60k),
+    (HF_LARGE_LV60_SELF_960H, wav2vec2_ft_large_lv60k),
+    (HF_LARGE_XLSR_DE, wav2vec2_ft_large_lv60k),
 ], name_func=_name_func)
 
 

test/torchaudio_unittest/models/wav2vec2/model_test.py

@@ -4,17 +4,17 @@ import torch
 import torch.nn.functional as F
 
 from torchaudio.models.wav2vec2 import (
-    wav2vec2_asr_base,
-    wav2vec2_asr_large,
-    wav2vec2_asr_large_lv60k,
+    wav2vec2_ft_base,
+    wav2vec2_ft_large,
+    wav2vec2_ft_large_lv60k,
     wav2vec2_base,
     wav2vec2_large,
     wav2vec2_large_lv60k,
     hubert_base,
     hubert_large,
     hubert_xlarge,
-    hubert_asr_large,
-    hubert_asr_xlarge,
+    hubert_ft_large,
+    hubert_ft_xlarge,
 )
 from torchaudio_unittest.common_utils import (
     TorchaudioTestCase,
@@ -40,11 +40,11 @@ pretrain_factory_funcs = parameterized.expand([
 
 
 finetune_factory_funcs = parameterized.expand([
-    (wav2vec2_asr_base, ),
-    (wav2vec2_asr_large, ),
-    (wav2vec2_asr_large_lv60k, ),
-    (hubert_asr_large, ),
-    (hubert_asr_xlarge, ),
+    (wav2vec2_ft_base, ),
+    (wav2vec2_ft_large, ),
+    (wav2vec2_ft_large_lv60k, ),
+    (hubert_ft_large, ),
+    (hubert_ft_xlarge, ),
 ], name_func=_name_func)
 
 
@@ -67,7 +67,7 @@ class TestWav2Vec2Model(TorchaudioTestCase):
     def test_cpu_smoke_test(self, dtype):
         model = wav2vec2_base()
         self._smoke_test(model, torch.device('cpu'), dtype)
-        model = wav2vec2_asr_base(num_out=32)
+        model = wav2vec2_ft_base(num_out=32)
         self._smoke_test(model, torch.device('cpu'), dtype)
 
     @parameterized.expand([(torch.float32, ), (torch.float64, )])
@@ -75,7 +75,7 @@ class TestWav2Vec2Model(TorchaudioTestCase):
     def test_cuda_smoke_test(self, dtype):
         model = wav2vec2_base()
         self._smoke_test(model, torch.device('cuda'), dtype)
-        model = wav2vec2_asr_base(num_out=32)
+        model = wav2vec2_ft_base(num_out=32)
         self._smoke_test(model, torch.device('cuda'), dtype)
 
     def _feature_extractor_test(self, model):
@@ -194,7 +194,7 @@ class TestWav2Vec2Model(TorchaudioTestCase):
     @finetune_factory_funcs
     def test_finetune_torchscript(self, factory_func):
         """Wav2Vec2Model should be scriptable"""
-        if factory_func.__name__ == 'hubert_asr_xlarge' and os.name == 'nt':
+        if factory_func is hubert_ft_xlarge and os.name == 'nt':
             self.skipTest(
                 'hubert_asr_xlarge is known to fail on Windows CI. '
                 'See https://github.com/pytorch/pytorch/issues/65776')

torchaudio/models/__init__.py

@@ -5,17 +5,17 @@ from .deepspeech import DeepSpeech
 from .tacotron2 import Tacotron2, tacotron2
 from .wav2vec2 import (
     Wav2Vec2Model,
-    wav2vec2_asr_base,
-    wav2vec2_asr_large,
-    wav2vec2_asr_large_lv60k,
+    wav2vec2_ft_base,
+    wav2vec2_ft_large,
+    wav2vec2_ft_large_lv60k,
     wav2vec2_base,
     wav2vec2_large,
     wav2vec2_large_lv60k,
     hubert_base,
     hubert_large,
     hubert_xlarge,
-    hubert_asr_large,
-    hubert_asr_xlarge,
+    hubert_ft_large,
+    hubert_ft_xlarge,
 )
 
 __all__ = [
@@ -25,17 +25,17 @@ __all__ = [
     'ConvTasNet',
     'DeepSpeech',
     'Wav2Vec2Model',
-    'wav2vec2_asr_base',
-    'wav2vec2_asr_large',
-    'wav2vec2_asr_large_lv60k',
+    'wav2vec2_ft_base',
+    'wav2vec2_ft_large',
+    'wav2vec2_ft_large_lv60k',
     'wav2vec2_base',
     'wav2vec2_large',
     'wav2vec2_large_lv60k',
     'hubert_base',
     'hubert_large',
     'hubert_xlarge',
-    'hubert_asr_large',
-    'hubert_asr_xlarge',
+    'hubert_ft_large',
+    'hubert_ft_xlarge',
     'Tacotron2',
     'tacotron2',
 ]

torchaudio/models/wav2vec2/__init__.py

 from .model import (
     Wav2Vec2Model,
-    wav2vec2_asr_base,
-    wav2vec2_asr_large,
-    wav2vec2_asr_large_lv60k,
+    wav2vec2_ft_base,
+    wav2vec2_ft_large,
+    wav2vec2_ft_large_lv60k,
     wav2vec2_base,
     wav2vec2_large,
     wav2vec2_large_lv60k,
     hubert_base,
     hubert_large,
     hubert_xlarge,
-    hubert_asr_large,
-    hubert_asr_xlarge,
+    hubert_ft_large,
+    hubert_ft_xlarge,
 )
 from . import utils
 
 __all__ = [
     'Wav2Vec2Model',
-    'wav2vec2_asr_base',
-    'wav2vec2_asr_large',
-    'wav2vec2_asr_large_lv60k',
+    'wav2vec2_ft_base',
+    'wav2vec2_ft_large',
+    'wav2vec2_ft_large_lv60k',
     'wav2vec2_base',
     'wav2vec2_large',
     'wav2vec2_large_lv60k',
     'hubert_base',
     'hubert_large',
     'hubert_xlarge',
-    'hubert_asr_large',
-    'hubert_asr_xlarge',
+    'hubert_ft_large',
+    'hubert_ft_xlarge',
     'utils',
 ]

torchaudio/models/wav2vec2/model.py

@@ -173,7 +173,7 @@ def wav2vec2_base() -> Wav2Vec2Model:
     )
 
 
-def wav2vec2_asr_base(num_out: int) -> Wav2Vec2Model:
+def wav2vec2_ft_base(num_out: int) -> Wav2Vec2Model:
     """Build "base" wav2vec2 with an extra linear module
 
     This is one of the model architectures used in *wav2vec 2.0*
@@ -235,7 +235,7 @@ def wav2vec2_large() -> Wav2Vec2Model:
     )
 
 
-def wav2vec2_asr_large(num_out: int) -> Wav2Vec2Model:
+def wav2vec2_ft_large(num_out: int) -> Wav2Vec2Model:
     """Build "large" wav2vec2.0 model with an extra linear module
 
     This is one of the model architectures used in *wav2vec 2.0*
@@ -297,7 +297,7 @@ def wav2vec2_large_lv60k() -> Wav2Vec2Model:
     )
 
 
-def wav2vec2_asr_large_lv60k(num_out: int) -> Wav2Vec2Model:
+def wav2vec2_ft_large_lv60k(num_out: int) -> Wav2Vec2Model:
     """Build "Large LV-60k" wav2vec2.0 with an extra linear module
 
     This is one of the model architectures used in *wav2vec 2.0*
@@ -388,7 +388,7 @@ def hubert_large() -> Wav2Vec2Model:
     )
 
 
-def hubert_asr_large(num_out) -> Wav2Vec2Model:
+def hubert_ft_large(num_out) -> Wav2Vec2Model:
     """Build "Large" HuBERT model with an extra linear module
 
 
@@ -451,7 +451,7 @@ def hubert_xlarge() -> Wav2Vec2Model:
     )
 
 
-def hubert_asr_xlarge(num_out) -> Wav2Vec2Model:
+def hubert_ft_xlarge(num_out) -> Wav2Vec2Model:
     """Build "extra large" HuBERT model with an extra linear module
 
     This is one of the model architecture used in *HuBERT*