BSD Portsaudio/whisper.cpp 1.8.4_2 FreeBSD
Port of OpenAI's Whisper model in C/C++
- System
- FreeBSD
- Repository
- FreeBSD ports
- Maintainer
- [email protected]
- Last update
- Commit
fc2d9770b2- License
- MIT
- Categories
- audio
- Package size
- 605.1 KiB, 2670.0 KiB installed
Description
High-performance inference of OpenAI's Whisper automatic speech recognition (ASR) model:
Features: * Plain C/C++ implementation without dependencies * Apple Silicon first-class citizen - optimized via ARM NEON, Accelerate framework, Metal and Core ML * AVX intrinsics support for x86 architectures * VSX intrinsics support for POWER architectures * Mixed F16 / F32 precision * Integer quantization support * Zero memory allocations at runtime * Vulkan support * Support for CPU-only inference * Efficient GPU support for NVIDIA * OpenVINO Support * Ascend NPU Support * Moore Threads GPU Support * C-style API
In the other systems
| System | Port | Version |
|---|---|---|
| FreeBSD | audio/whisper.cpp | 1.8.4_2 |
| OpenBSD | audio/whisper.cpp | 1.9.5 |
Install
Binary package:
pkg install whisper.cppFrom the ports tree:
cd /usr/ports/audio/whisper.cpp/ && make install cleanTo depend on this port in another port's Makefile:
RUN_DEPENDS= whisper.cpp>0:audio/whisper.cpp
LIB_DEPENDS= libwhisper.so:audio/whisper.cppPackages
The version each package repository carries. “-” means the repository builds no package of it, “n/a” that there is no such repository.
| Repository | aarch64 | amd64 | armv6 | armv7 | i386 |
|---|---|---|---|---|---|
| FreeBSD:14:latest | 1.8.4_1 | 1.8.4_1 | - | - | - |
| FreeBSD:14:quarterly | 1.8.4_1 | 1.8.4_1 | n/a | - | - |
| FreeBSD:15:latest | 1.8.4_1 | 1.8.4_2 | n/a | - | - |
| FreeBSD:15:quarterly | 1.8.4_1 | 1.8.4_1 | n/a | n/a | n/a |
| FreeBSD:16:latest | 1.8.4_1 | 1.8.4_2 | n/a | n/a | n/a |
| FreeBSD:16:quarterly | n/a | n/a | n/a | n/a | n/a |
Dependencies
As this port's Makefile names them.
Library 1
libggml.so: misc/ggml
The full lists below include those, and their dependencies in turn.
Build dependencies 28
- devel/cmake-core
- devel/gettext-runtime
- devel/jsoncpp
- devel/libepoll-shim
- devel/libffi
- devel/libunistring
- devel/libuv
- devel/ninja
- devel/readline
- dns/libidn2
- graphics/vulkan-loader
- graphics/wayland
- lang/python312
- math/mpdecimal
- misc/ggml
- ports-mgmt/pkg
- print/indexinfo
- security/rhash
- textproc/expat2
- textproc/libxml2
- x11/libX11
- x11/libXau
- x11/libXdmcp
- x11/libXext
- x11/libXrandr
- x11/libXrender
- x11/libxcb
- x11/xorgproto
Run dependencies 20
No port depends on this one.
Configuration
Options name: audio_whisper.cpp
| Option | Default | Description |
|---|---|---|
| EXAMPLES | on | Build and/or install examples |
- USES
- cmake:testing
- Service scripts
- none
Message after installing
You installed whisper.cpp: Port of OpenAI's Whisper model in C/C++
In order to use whisper.cpp you should first download an AI model.
For example:
$ whisper.cpp-download-ggml-model.sh base.en
Then you can transcribe audio files using a command:
$ whisper-cli -m ggml-base.en.bin -f {sample}.wav
Files
pkg-plist 13
- bin/whisper-bench
- bin/whisper-cli
- bin/whisper-quantize
- bin/whisper-server
- bin/whisper-vad-speech-segments
- bin/whisper.cpp-download-ggml-model.sh
- include/whisper.h
- lib/cmake/whisper/whisper-config.cmake
- lib/cmake/whisper/whisper-version.cmake
- lib/libwhisper.so
- lib/libwhisper.so.1
- lib/libwhisper.so.1.8.4
- libdata/pkgconfig/whisper.pc
distinfo 1
| File | Size | SHA256 |
|---|---|---|
| ggerganov-whisper.cpp-v1.8.4_GH0.tar.gz | 8.206306e+06 | b26f30e52c095ccb75da40b168437736605eb280de57381887bf9e2b65f31e66 |