mirror of
https://github.com/ggml-org/whisper.cpp.git
synced 2026-10-19 07:08:57 +02:00
ggml-ci: update input env variables to GG_BUILD_ (#2879)
This commit is contained in:
+9
-9
@@ -21,7 +21,7 @@ mkdir tmp
|
||||
bash ./ci/run.sh ./tmp/results ./tmp/mnt
|
||||
|
||||
# with CUDA support
|
||||
GGML_CUDA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
|
||||
GG_BUILD_CUDA=1 bash ./ci/run.sh ./tmp/results ./tmp/mnt
|
||||
```
|
||||
|
||||
## Environment Variables
|
||||
@@ -30,12 +30,12 @@ The CI script supports several environment variables to control the build:
|
||||
|
||||
| Variable | Description |
|
||||
|----------|-------------|
|
||||
| `GGML_CUDA` | Enable NVIDIA CUDA GPU acceleration |
|
||||
| `GGML_SYCL` | Enable Intel SYCL acceleration |
|
||||
| `GGML_VULKAN` | Enable Vulkan GPU acceleration |
|
||||
| `GGML_METAL` | Enable Metal acceleration on Apple Silicon |
|
||||
| `GGML_BLAS` | Enable BLAS CPU acceleration |
|
||||
| `WHISPER_OPENVINO` | Enable OpenVINO support |
|
||||
| `WHISPER_COREML` | Enable Core ML support for Apple Neural Engine |
|
||||
| `GG_BUILD_CUDA` | Enable NVIDIA CUDA GPU acceleration |
|
||||
| `GG_BUILD_SYCL` | Enable Intel SYCL acceleration |
|
||||
| `GG_BUILD_VULKAN` | Enable Vulkan GPU acceleration |
|
||||
| `GG_BUILD_METAL` | Enable Metal acceleration on Apple Silicon |
|
||||
| `GG_BUILD_BLAS` | Enable BLAS CPU acceleration |
|
||||
| `GG_BUILD_OPENVINO` | Enable OpenVINO support |
|
||||
| `GG_BUILD_COREML` | Enable Core ML support for Apple Neural Engine |
|
||||
| `GG_BUILD_LOW_PERF` | Limit tests for low-performance hardware |
|
||||
| `GGML_TEST_MODELS` | Comma-separated list of models to test (e.g. "tiny.en,tiny,base,medium", defaults to all models unless `GG_BUILD_LOW_PERF` is set) |
|
||||
| `GG_BUILD_TEST_MODELS` | Comma-separated list of models to test (e.g. "tiny.en,tiny,base,medium", defaults to all models unless `GG_BUILD_LOW_PERF` is set) |
|
||||
|
||||
Reference in New Issue
Block a user