The releases of versions 1.9.0 and 1.9.1 of the high-performance automatic speech recognition system took place on June 17 and 19. whisper.cpp, implementing the model Whisper from OpenAI, based on a tensor-based machine learning library. GGML and binary format. GGUF.
The libwhisper library is provided, along with demonstration examples and console utilities: whisper-bench, whisper-cli, whisper-command, whisper-lsp, whisper-quantize, whisper-server, whisper-stream, whisper-vad-speech-segments, parakeet-cli, and parakeet-quantize.
The projects are written in C and C++ and are distributed under the MIT license.
Changes:
- Added the console utility parakeet-cli, which supports the model. NVIDIA Parakeet.;
- Support for NVIDIA Parakeet has also been added in the Ruby wrapper.
Some important changes made in intermediate versions. since the news about whisper.cpp 1.8.0.:
- C++ exception handling in the function whisper_init_with_params_no_state.
- The ffmpeg support has been completely rewritten (without using GPL code taken from ffmpeg samples), and the usage of ffmpeg in whisper.cpp has been clarified.
- A —version option has been added to the whisper-cli utility for displaying the version.
Source: linux.org.ru
