Use torch.no_grad for VITS inference

TTS/tts/models/vits.py

@@ -982,6 +982,7 @@ class Vits(BaseTTS):
             return aux_input["x_lengths"]
         return torch.tensor(x.shape[1:2]).to(x.device)
 
+    @torch.no_grad()
     def inference(
         self, x, aux_input={"x_lengths": None, "d_vectors": None, "speaker_ids": None, "language_ids": None}
     ):  # pylint: disable=dangerous-default-value