Update app.py

app.py

@@ -127,18 +127,16 @@ def load_models_and_documents():
             n_gpu_layers=32
         )
 
-        provider = ['TensorrtExecutionProvider']
-
         dense_model = FastEmbedEncoder(
             name='mixedbread-ai/mxbai-embed-large-v1',
             cache_dir=os.getenv('HF_HOME'),
-            providers=provider
+            providers=['TensorrtExecutionProvider']
         )
         
         sparse_model = SparseTextEmbedding(
             'Qdrant/all_miniLM_L6_v2_with_attentions',
             cache_dir=os.getenv('HF_HOME'),
-            providers=provider
+            providers=['CUDAExecutionProvider']
         )
 
         nltk.download('punkt')