Revise VGGish inference pipeline test (#3544)

Summary:
Pull Request resolved: https://github.com/pytorch/audio/pull/3544

Revises VGGish inference pipeline test to support internal testing.

Reviewed By: mthrok

Differential Revision: D48058409

fbshipit-source-id: 045140a0e9d50128d32ef6510bdb2f642a365c83

test/integration_tests/prototype/vggish_pipeline_test.py

+import unittest
+
 import torchaudio
 from torchaudio.prototype.pipelines import VGGISH
 
 
-def test_vggish():
-    input_sr = VGGISH.sample_rate
-    input_proc = VGGISH.get_input_processor()
-    model = VGGISH.get_model()
-    path = torchaudio.utils.download_asset("test-assets/Chopin_Ballade_-1_In_G_Minor,_Op._23_excerpt.mp3")
-    waveform, sr = torchaudio.load(path, backend="ffmpeg")
-    waveform = waveform.mean(axis=0)
-    waveform = torchaudio.functional.resample(waveform, sr, input_sr)
-    batch = input_proc(waveform)
-    assert batch.shape == (62, 1, 96, 64)
-    output = model(batch)
-    assert output.shape == (62, 128)
+class VGGishPipelineTest(unittest.TestCase):
+    def test_vggish(self):
+        input_sr = VGGISH.sample_rate
+        input_proc = VGGISH.get_input_processor()
+        model = VGGISH.get_model()
+        path = torchaudio.utils.download_asset("test-assets/Chopin_Ballade_-1_In_G_Minor,_Op._23_excerpt.mp3")
+        waveform, sr = torchaudio.load(path, backend="ffmpeg")
+        waveform = waveform.mean(axis=0)
+        waveform = torchaudio.functional.resample(waveform, sr, input_sr)
+        batch = input_proc(waveform)
+        assert batch.shape == (62, 1, 96, 64)
+        output = model(batch)
+        assert output.shape == (62, 128)