GRU-Based Music Generator

ML Engineer

Sequence model that generates symbolic music with controllable temperature and musical parameters.

Highlights

Implemented a GRU-based recurrent network for symbolic music generation, trained on MIDI-like sequences.
Exposed parameters such as temperature and pitch-shift to control diversity and transposition of generated sequences.
Generated 10-second clips and analyzed how hyperparameters affect musical structure, repetition, and variability.