Update app.py

app.py

@@ -19,6 +19,10 @@ config = Config()
 logging.getLogger("numba").setLevel(logging.WARNING)
 limitation = os.getenv("SYSTEM") == "spaces"  # limit audio length in huggingface spaces
 
+# Load the whisper_large_v2 interface
+whisper_interface = gr.Interface.load("models/openai/whisper-large-v2")
+
+
 def create_vc_fn(tgt_sr, net_g, vc, if_f0, file_index):
     def vc_fn(
         input_audio,
@@ -74,20 +78,6 @@ def create_vc_fn(tgt_sr, net_g, vc, if_f0, file_index):
             return info, (None, None)
     return vc_fn
 
-def load_hubert():
-    global hubert_model
-    models, _, _ = checkpoint_utils.load_model_ensemble_and_task(
-        ["hubert_base.pt"],
-        suffix="",
-    )
-    hubert_model = models[0]
-    hubert_model = hubert_model.to(config.device)
-    if config.is_half:
-        hubert_model = hubert_model.half()
-    else:
-        hubert_model = hubert_model.float()
-    hubert_model.eval()
-
 def change_to_tts_mode(tts_mode):
     if tts_mode:
         return gr.Audio.update(visible=False), gr.Textbox.update(visible=True), gr.Dropdown.update(visible=True)
@@ -97,8 +87,12 @@ def change_to_tts_mode(tts_mode):
 if __name__ == '__main__':
     load_hubert()
     models = []
+
     tts_voice_list = asyncio.get_event_loop().run_until_complete(edge_tts.list_voices())
     voices = [f"{v['ShortName']}-{v['Gender']}" for v in tts_voice_list]
+
+    # Add the whisper_large_v2 model to the models list
+    models.append(("Whisper v2", "Whisper v2", "OpenAI", None, whisper_interface))
     if limitation:
         with open("weights/model_info.json", "r", encoding="utf-8") as f:
             models_info = json.load(f)