Compare prompt variations — run the same dataset through different prompts and see which performs better. Benchmark models — evaluate GPT-4o vs GPT-4o-mini vs Claude on the same test set. Test RAG ...

Research-focused neural Text-to-Speech (TTS) synthesis system built with PyTorch. This project implements state-of-the-art TTS models including Tacotron2, with comprehensive evaluation metrics and an ...