A Streamlit application designed to transcribe extended audio files through Groq's Whisper Large V3 API, addressing the platform's file size constraints through intelligent processing.
Features
- Smart Chunking: Files exceeding Groq's 25MB limit are divided into 10-minute segments with overlap
- Intelligent Merging: Uses longest common sequence algorithms to seamlessly connect transcription segments
- Broad Language Support: English, Italian, Spanish, French, German, Portuguese, Dutch, Russian, Chinese, Japanese, and Korean
- Format Flexibility: Processes MP3, WAV, M4A, FLAC, OGG, MP4, and WEBM files
- Authentication: Basic password protection for the interface
Tech Stack
- Framework: Streamlit
- API: Groq Whisper Large V3
- Audio Processing: ffmpeg
- Language: Python