Persian ASR Transcription with NeMo Fast Conformer

Transcribe long-form microphone or audio inputs with the click of a button! Demo uses the NeMo's Fast Conformer Hybrid Large.

Trained on 800 hours of Persian speech dataset (Common Voice 17 (300 hours), YouTube (400 hours), NasleMana (90 hours), In-house dataset (~70 hours)).

For commercial applications, contact us via email: saeedzou2012@gmail.com.

Credit FAIM Group, Sharif University of Technology.