Update app.py

app.py

@@ -11,7 +11,7 @@ GGUF_FILENAME = "Dolphin3.0-Llama3.1-8B-Q4_K_S.gguf"
 GGUF_DOWNLOAD_URL = f"https://huggingface.co/cognitivecomputations/Dolphin3.0-Llama3.1-8B-GGUF/resolve/main/{GGUF_FILENAME}"
 
 MODEL_TYPE = "llama"
-GPU_LAYERS = 20 # Try -1. If OOM, reduce (20, 15, 10, or 0 for CPU-only)
+GPU_LAYERS = 0 # Try -1. If OOM, reduce (20, 15, 10, or 0 for CPU-only)
 MAX_NEW_TOKENS = 512
 CONTEXT_LENGTH = 4096
 TEMPERATURE = 0.7