Update app.py

app.py

@@ -141,7 +141,7 @@ def load_models_and_documents():
             verbose=True,
             chat_format="llama-3",
             n_ctx=16384,
-            n_gpu_layers=32,
+            n_gpu_layers=33,
             flash_attn=True
         )
 
@@ -151,7 +151,7 @@ def load_models_and_documents():
         )
         
         sparse_model = SparseTextEmbedding(
-            'prithivida/Splade_PP_en_v2',
+            'Qdrant/bm42-all-minilm-l6-v2-attentions',
             cache_dir=os.getenv('HF_HOME'),
             providers=['CPUExecutionProvider']
         )