| Property | Value |
|---|---|
| License | MIT |
| Model Base | Whisper Large-v3 Turbo |
| Training Data | 600,000 samples (1000 hours) |
| Model URL | https://huggingface.co/erax-ai/EraX-WoW-Turbo-V1.1 |
What is EraX-WoW-Turbo-V1.1?
EraX-WoW-Turbo-V1.1 is a supercharged speech recognition model built on Whisper Large-v3 Turbo, specifically optimized for Vietnamese and 10 other languages. The model achieves remarkable speed, processing 30 seconds of audio in approximately 350ms, making it ideal for real-time applications.
Implementation Details
The model leverages CTranslate2 library for enhanced performance, offering up to 2.5x speedup in processing time. It's trained on a diverse dataset of 600,000 samples, covering various real-world audio conditions and accents.
- Multilingual support for 11 languages including Vietnamese, English, Chinese, Cantonese, Indonesian, Korean, Japanese, Russian, German, French, and Dutch
- Optimized for real-time transcription with ~12% Word Error Rate
- Robust handling of regional accents and background noise
Core Capabilities
- Real-time speech transcription
- Multi-dialect Vietnamese support
- Noise-resistant recognition
- Integration with CTranslate2 for enhanced performance
- Support for diverse audio conditions
Frequently Asked Questions
Q: What makes this model unique?
The model's standout features include its exceptional speed, comprehensive support for Vietnamese regional dialects, and optimized performance using CTranslate2. It processes audio nearly in real-time while maintaining high accuracy across multiple languages.
Q: What are the recommended use cases?
The model is ideal for real-time transcription, voice assistants, media subtitling, accessibility tools, and language learning applications. However, it's not optimized for infant speech or whispered audio.