Fix `SpeechT5ForSpeechToSpeechIntegrationTests` device issue (#21460)

* fix

---------
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>

src/transformers/models/speecht5/modeling_speecht5.py

@@ -2869,7 +2869,7 @@ class SpeechT5ForSpeechToSpeech(SpeechT5PreTrainedModel):
             predicted mel spectrogram, or a tensor with shape `(num_frames,)` containing the speech waveform.
         """
         if speaker_embeddings is None:
-            speaker_embeddings = torch.zeros((1, 512))
+            speaker_embeddings = torch.zeros((1, 512), device=input_values.device)
 
         return _generate_speech(
             self,

tests/models/speecht5/test_modeling_speecht5.py

@@ -1423,7 +1423,7 @@ class SpeechT5ForSpeechToSpeechIntegrationTests(unittest.TestCase):
         input_speech = self._load_datasamples(1)
         input_values = processor(audio=input_speech, return_tensors="pt").input_values.to(torch_device)
 
-        speaker_embeddings = torch.zeros((1, 512))
+        speaker_embeddings = torch.zeros((1, 512), device=torch_device)
         generated_speech = model.generate_speech(input_values, speaker_embeddings=speaker_embeddings)
 
         self.assertEqual(generated_speech.shape[1], model.config.num_mel_bins)