change to phi2 model

app.py

@@ -167,7 +167,7 @@ def split_docs(documents,chunk_size=1000):
 
 @st.cache_resource
 def load_llama2_llamaCpp():
-    core_model_name = "llama-2-7b-chat.Q4_0.gguf"
+    core_model_name = "phi-2.Q4_K_M.gguf"
     #n_gpu_layers = 32
     n_batch = 512
     callback_manager = CallbackManager([StreamingStdOutCallbackHandler()])