End-to-end Automatic Speech Recognition and Speech Translation: Integration of Speech Foundational Models and LLMs