init

model_pyannote_embedding.py

@@ -14,11 +14,14 @@ from pyannote.audio.core.inference import fix_reproducibility, map_with_specific
 class PyannoteSE:
 
     def __init__(self):
-        self.model = Model.from_pretrained("pyannote/embedding")
-        self.inference = Inference(self.model, window="whole")
+        model = Model.from_pretrained("pyannote/embedding")
+        self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+        model.to(self.device)
+        model.eval()
+        self.inference = Inference(model, window="whole")
 
     def get_speaker_embedding(self, wav: np.ndarray, sampling_rate: Optional[int] = None) -> np.ndarray:
-        wav = torch.as_tensor(wav.reshape(1, -1))
+        wav = torch.as_tensor(wav.reshape(1, -1)).to(self.device)
         fix_reproducibility(self.inference.device)
         if self.inference.window == "sliding":
             return self.inference.slide(wav, sampling_rate, hook=None)
@@ -29,5 +32,5 @@ class PyannoteSE:
             return outputs[0]
 
         return map_with_specifications(
-            self.model.specifications, __first_sample, outputs
+            self.inference.model.specifications, __first_sample, outputs
         )