b7651
📦 llama-cppView on GitHub →
✨ 3 features🔧 1 symbols
Summary
This release introduces streaming Inverse Short-Time Fourier Transform (ISTFT) capabilities, essential for audio generation models, and refactors audio caching to be instance-specific. It also provides comprehensive pre-built binaries across multiple operating systems and hardware configurations.
✨ New Features
- Added streaming ISTFT class (`mtmd_audio_streaming_istft`) with overlap-add for audio reconstruction, required for models like [LFM2.5-Audio-1.5B].
- Replaced global audio cache with per-instance cache, requiring two independent caches for preprocessing (audio input) and ISTFT (audio output).
- Unified templated FFT/IFFT implementation supporting both forward and inverse transforms.
🔧 Affected Symbols
mtmd_audio_streaming_istft