ExternalVendorCode/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/whisper.cpp.git synced 2024-12-18 20:27:53 +00:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

readme : update build instructions

Browse Source
This commit is contained in:
Georgi Gerganov 2024-12-08 15:48:14 +02:00
parent 472464453d
commit 627b11c78a
1 changed files with 19 additions and 37 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

56
README.md
View File

@ -89,10 +89,11 @@ Now build the [main](examples/main) example and transcribe an audio file like th
```bash ```bash
# build the main example # build the main example
make -j cmake -B build
cmake --build build --config Release
# transcribe an audio file # transcribe an audio file
./main -f samples/jfk.wav ./build/bin/main -f samples/jfk.wav
``` ```
--- ---
@ -265,11 +266,12 @@ Here are the steps for creating and using a quantized model:
```bash ```bash
# quantize a model with Q5_0 method # quantize a model with Q5_0 method
make -j quantize cmake -B build
./quantize models/ggml-base.en.bin models/ggml-base.en-q5_0.bin q5_0 cmake --build build --config Release
./build/bin/quantize models/ggml-base.en.bin models/ggml-base.en-q5_0.bin q5_0
# run the examples as usual, specifying the quantized model file # run the examples as usual, specifying the quantized model file
./main -m models/ggml-base.en-q5_0.bin ./samples/gb0.wav ./build/bin/main -m models/ggml-base.en-q5_0.bin ./samples/gb0.wav
``` ```
## Core ML support ## Core ML support
@ -303,10 +305,6 @@ speed-up - more than x3 faster compared with CPU-only execution. Here are the in
- Build `whisper.cpp` with Core ML support: - Build `whisper.cpp` with Core ML support:
```bash ```bash
# using Makefile
make clean
WHISPER_COREML=1 make -j
# using CMake # using CMake
cmake -B build -DWHISPER_COREML=1 cmake -B build -DWHISPER_COREML=1
cmake --build build -j --config Release cmake --build build -j --config Release
@ -426,8 +424,8 @@ First, make sure you have installed `cuda`: https://developer.nvidia.com/cuda-do
Now build `whisper.cpp` with CUDA support: Now build `whisper.cpp` with CUDA support:
``` ```
make clean cmake -B build -DGGML_CUDA=1
GGML_CUDA=1 make -j cmake --build build -j --config Release
``` ```
## Vulkan GPU support ## Vulkan GPU support
@ -436,8 +434,8 @@ First, make sure your graphics card driver provides support for Vulkan API.
Now build `whisper.cpp` with Vulkan support: Now build `whisper.cpp` with Vulkan support:
``` ```
make clean cmake -B build -DGGML_VULKAN=1
make GGML_VULKAN=1 -j cmake --build build -j --config Release
``` ```
## BLAS CPU support via OpenBLAS ## BLAS CPU support via OpenBLAS
@ -448,28 +446,13 @@ First, make sure you have installed `openblas`: https://www.openblas.net/
Now build `whisper.cpp` with OpenBLAS support: Now build `whisper.cpp` with OpenBLAS support:
``` ```
make clean cmake -B build -DGGML_BLAS=1
GGML_OPENBLAS=1 make -j cmake --build build -j --config Release
```
## BLAS CPU support via Intel MKL
Encoder processing can be accelerated on the CPU via the BLAS compatible interface of Intel's Math Kernel Library.
First, make sure you have installed Intel's MKL runtime and development packages: https://www.intel.com/content/www/us/en/developer/tools/oneapi/onemkl-download.html
Now build `whisper.cpp` with Intel MKL BLAS support:
```
source /opt/intel/oneapi/setvars.sh
mkdir build
cd build
cmake -DWHISPER_MKL=ON ..
WHISPER_MKL=1 make -j
``` ```
## Ascend NPU support ## Ascend NPU support
Ascend NPU provides inference acceleration via [`CANN`](https://www.hiascend.com/en/software/cann) and AI cores. Ascend NPU provides inference acceleration via [`CANN`](https://www.hiascend.com/en/software/cann) and AI cores.
First, check if your Ascend NPU device is supported: First, check if your Ascend NPU device is supported:
@ -483,10 +466,8 @@ Then, make sure you have installed [`CANN toolkit`](https://www.hiascend.com/en/
Now build `whisper.cpp` with CANN support: Now build `whisper.cpp` with CANN support:
``` ```
mkdir build cmake -B build -DGGML_CANN=1
cd build cmake --build build -j --config Release
cmake .. -D GGML_CANN=on
make -j
``` ```
Run the inference examples as usual, for example: Run the inference examples as usual, for example:
@ -636,8 +617,9 @@ The [stream](examples/stream) tool samples the audio every half a second and run
More info is available in [issue #10](https://github.com/ggerganov/whisper.cpp/issues/10). More info is available in [issue #10](https://github.com/ggerganov/whisper.cpp/issues/10).
```bash ```bash
make stream -j cmake -B build
./stream -m ./models/ggml-base.en.bin -t 8 --step 500 --length 5000 cmake --build build --config Release
./build/bin/stream -m ./models/ggml-base.en.bin -t 8 --step 500 --length 5000
``` ```
https://user-images.githubusercontent.com/1991296/194935793-76afede7-cfa8-48d8-a80f-28ba83be7d09.mp4 https://user-images.githubusercontent.com/1991296/194935793-76afede7-cfa8-48d8-a80f-28ba83be7d09.mp4