Changed to q4 model to make quicker.

app.py

@@ -11,8 +11,8 @@ logger = logging.getLogger(__name__)
 logger.info("Loading the model...")
 llm = Llama.from_pretrained(
     repo_id="AndreasThinks/mistral-7b-english-welsh-translate-GGUF",
-    filename="*q8_0.gguf",
-    verbose=False
+    filename="*q4_k_m.gguf",
+    verbose=True
 )
 logger.info("Model loaded successfully")