Update app.py

app.py

@@ -9,7 +9,7 @@ from contextlib import asynccontextmanager
 # Your model repository on Hugging Face
 MODEL_REPO = "bartowski/Phi-3.1-mini-4k-instruct-GGUF"
 # The specific GGUF file name that is around 3.1GB
-MODEL_FILE = "Phi-3.1-mini-4k-instruct-Q5_K_M.gguf"
+MODEL_FILE = "Phi-3.1-mini-4k-instruct-Q4_K_M.gguf"
 # n_gpu_layers=-1 tells llama.cpp to offload all layers to the GPU if one is available
 N_GPU_LAYERS = -1 
 # Maximum tokens the model can generate in a single response