Update app.py

app.py

@@ -41,9 +41,9 @@ def load_model(model_name=DEFAULT_MODEL):
     return model, tokenizer
 
 def load_embedding_model(model_name=DEFAULT_SENTENCE_EMBEDDING_MODEL):
-    if torch.cuda.is_available():
-        embedding_model = SentenceTransformer(model_name, device='cuda')
-    else:
+    # if torch.cuda.is_available():
+    #     embedding_model = SentenceTransformer(model_name, device='cuda')
+    # else:
         embedding_model = SentenceTransformer(model_name)
     print('Load sentence embedding model done')
     return embedding_model