Update app.py

app.py

@@ -5,14 +5,10 @@ from llama_cpp import Llama
 # Load GGUF model
 # ----------------------------------------------------
 
-MODEL_PATH = "astegaras/merged_kaggle"   # HF repo containing your .gguf
-
 # llama_cpp automatically downloads from HF Hub if you provide the repo
 llm = Llama.from_pretrained(
-    repo_id=MODEL_PATH,
-    n_threads=8,
-    n_ctx=4096,
-    verbose=False,
+	repo_id="astegaras/merged_kaggle",
+	filename="llama-3.2-3b-instruct.Q4_K_M.gguf",
 )
 
 # ----------------------------------------------------