The Framework for Personalized Music and Lyrics Generation is designed to generate customized lyrics and corresponding melodies based on user input. The framework comprises two key components: lyrics generation and music melody generation. The lyrics generation module explores and compares two different approaches: topic modeling-based generation and retrieval-augmented generation. The effectiveness of these approaches is evaluated using various linguistic and structural metrics, including perplexity score, repetition score, coherence, and syllable variance. The music melody generation module takes the syllabified lyrics as input and generates suitable musical notes using different sequential models. The performance of these models is assessed based on statistical metrics to determine their effectiveness in producing musically coherent melodies. By integrating these components, the proposed framework aims to enhance personalized music composition by leveraging advanced techniques in natural language processing and deep learning.