Code and Resources for "LLM-Powered Grapheme-to-Phoneme Conversion: Benchmark and Case Study", introducing methods to leverage LLMs for G2P tasks without additional training, featuring Sentence-Bench and Kaamel-Dict.

A Persian grapheme-to-phoneme (G2P) model designed for homograph disambiguation, fine-tuned using the HomoRich dataset to improve pronunciation accuracy.

Benchmarking notebooks for various Persian G2P models, comparing their performance on the SentenceBench dataset, including Homo-GE2PE and Homo-T5.

HomoRich: The first large-scale Persian homograph dataset for G2P conversion, featuring 528K annotated sentences with balanced pronunciation variants and dual phoneme representations.