Speech Transcription with parakeet-tdt-0.6b-v3 🦜
This demo showcases parakeet-tdt-0.6b-v3, a 600-million-parameter multilingual model designed for high-quality speech recognition.
Key Features:
- Multilingual transcription across 25 European languages
- Automatic punctuation and capitalization
- Accurate word-level timestamps (click on a segment in the table below to play it!)
- Long audio transcription: up to 24 minutes with full attention (A100 80GB) or up to 3 hours with local attention
Supported Languages: bg, hr, cs, da, nl, en, et, fi, fr, de, el, hu, it, lv, lt, mt, pl, pt, ro, sk, sl, es, sv, ru, uk
This model is available for commercial and non-commercial use (CC BY 4.0).
🎙️ Learn more about the Model | 📄 Fast Conformer paper | 📚 TDT paper | 🧑💻 NeMo Repository
Example Audio Files (Click to Load)
Transcription Results (Click row to play segment)
Transcription Segments