Wish there was to a way to plug in my local llm into amethyst to do translations, maybe we can do it with the local llama 3.2 models