Skip to content

Commit 60ad6a5

Browse files
committed
Update README.md for python 3.14 and HIP (ROCm) guide
1 parent 9b985b7 commit 60ad6a5

1 file changed

Lines changed: 12 additions & 5 deletions

File tree

README.md

Lines changed: 12 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -129,8 +129,8 @@ pip install "llama-cpp-python @ git+https://github.com/JamePeng/llama-cpp-python
129129
It is also possible to install a pre-built wheel with CUDA support. As long as your system meets some requirements:
130130

131131
- CUDA Version is 12.4, 12.6, 12.8 or 13.0
132-
- Python Version is 3.10, 3.11, 3.12 or 3.13
133-
- Basic version: A version compiled without using AVX instructions (for compatibility with CPU platforms lacking AVX instructions or with AVX instruction compatibility issues).
132+
- Python Version is 3.10, 3.11, 3.12, 3.13 or 3.14
133+
- Basic version(Default): A version compiled without using AVX instructions (for compatibility with CPU platforms lacking AVX instructions or with AVX instruction compatibility issues).
134134
- AVX2 version: A version compiled using AVX2 instructions.
135135

136136
Check the releases page:
@@ -172,13 +172,20 @@ pip install llama-cpp-python \
172172
</details>
173173

174174
<details>
175-
<summary>hipBLAS (ROCm)</summary>
175+
<summary>HIP (ROCm)</summary>
176176

177-
To install with hipBLAS / ROCm support for AMD cards, set the `GGML_HIPBLAS=on` environment variable before installing:
177+
This provides GPU acceleration on HIP-supported AMD GPUs. Make sure to have ROCm installed.
178+
179+
You can download it from your Linux distro's package manager or from here: [ROCm Quick Start (Linux)](https://rocm.docs.amd.com/projects/install-on-linux/en/latest/tutorial/quick-start.html#rocm-install-quick).
180+
181+
To install with HIP / ROCm support for AMD cards, set the `GGML_HIP=ON` environment variable before installing:
178182
179183
```bash
180-
CMAKE_ARGS="-DGGML_HIPBLAS=on" pip install "llama-cpp-python @ git+https://github.com/JamePeng/llama-cpp-python.git"
184+
CMAKE_ARGS="-DGGML_HIP=ON -DGPU_TARGETS=gfx1030" pip install "llama-cpp-python @ git+https://github.com/JamePeng/llama-cpp-python.git"
181185
```
186+
Note: `GPU_TARGETS` is optional, omitting it will build the code for all GPUs in the current system.
187+
188+
More details see here: https://github.com/ggml-org/llama.cpp/blob/master/docs/build.md#hip
182189
183190
</details>
184191

0 commit comments

Comments
 (0)