Results
Platform and tools
- Educational platform
School notebooks and learning materials.
- Audio resources
Tools, speech models and datasets.
Speech models
- Whisper large-v3-turbo (Hugging Face)
Speech recognition (STT)
- XTTS-v2 (Hugging Face)
Speech synthesis (TTS)
Datasets
Over 1,000 hours of audio, including 200 hours recorded in a studio within the project.