mirror/whisper.cpp
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

readme : add Vulkan notice (#2488)

Browse Source 
* Add Vulkan notice in README.md

* Fix formatting for Vulkan section in README.md

* Fix formatting in README.md
This commit is contained in:
toboil-features 2024-10-16 18:43:26 +03:00 committed by GitHub
parent 1d5752fa42
commit f7c99e49b3
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 11 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
README.md
View File

@ -18,6 +18,7 @@ High-performance inference of [OpenAI's Whisper](https://github.com/openai/whisp
- Mixed F16 / F32 precision - Mixed F16 / F32 precision
- [4-bit and 5-bit integer quantization support](https://github.com/ggerganov/whisper.cpp#quantization) - [4-bit and 5-bit integer quantization support](https://github.com/ggerganov/whisper.cpp#quantization)
- Zero memory allocations at runtime - Zero memory allocations at runtime
- Vulkan support
- Support for CPU-only inference - Support for CPU-only inference
- [Efficient GPU support for NVIDIA](https://github.com/ggerganov/whisper.cpp#nvidia-gpu-support-via-cublas) - [Efficient GPU support for NVIDIA](https://github.com/ggerganov/whisper.cpp#nvidia-gpu-support-via-cublas)
- [OpenVINO Support](https://github.com/ggerganov/whisper.cpp#openvino-support) - [OpenVINO Support](https://github.com/ggerganov/whisper.cpp#openvino-support)
@ -429,6 +430,16 @@ make clean
GGML_CUDA=1 make -j GGML_CUDA=1 make -j
``` ```
## Vulkan GPU support
Cross-vendor solution which allows you to accelerate workload on your GPU.
First, make sure your graphics card driver provides support for Vulkan API.
Now build `whisper.cpp` with Vulkan support:
```
make clean
make GGML_VULKAN=1
```
## BLAS CPU support via OpenBLAS ## BLAS CPU support via OpenBLAS
Encoder processing can be accelerated on the CPU via OpenBLAS. Encoder processing can be accelerated on the CPU via OpenBLAS.