Update app.py

app.py

@@ -1,11 +1,15 @@
 import streamlit as st
-from transformers import pipeline
+from transformers import pipeline, T5Config
 
 # 加载 Visual Question Answering 模型 microsoft/git-base-vqav2
 vqa_pipeline = pipeline("text2text-generation", model="microsoft/git-base-vqav2")
 
 # 加载文本到语音模型
-text_to_speech_pipeline = pipeline("text-to-speech", model="microsoft/speecht5_tts")
+text_to_speech_pipeline = pipeline(
+    "text-to-speech",
+    model="microsoft/speecht5_tts",
+    config=T5Config.from_pretrained("microsoft/speecht5_tts", speaker_embeddings=True)
+)
 
 def main():
     st.title("Visual Question Answering with Text-to-Speech")