from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="QuantFactory/Qwen2.5-7B-Instruct-Uncensored-GGUF",
	filename="Qwen2.5-7B-Instruct-Uncensored.Q2_K.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)