EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

[Build] Update doc (#509)

Browse Source 
1. install torch-npu before vllm-ascend to ensure custom ops build
success.
2. set `COMPILE_CUSTOM_KERNELS=0` if users want to disable custom ops
build.

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
This commit is contained in:
wangxiyuan
2025-04-14 14:38:50 +08:00
committed by GitHub
parent 11ecbfdb31
commit 5fa70b6393
2 changed files with 30 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
docs/source/developer_guide/contributing.md
View File

@@ -28,6 +28,9 @@ cd ..
# Clone vllm-ascend and install
git clone https://github.com/vllm-project/vllm-ascend.git
cd vllm-ascend
# install system requirement
apt install -y gcc g++ cmake libnuma-dev
# install project requirement
pip install -r requirements-dev.txt
# Then you can run lint and mypy test
@@ -38,6 +41,8 @@ bash format.sh
# pip install -e .
# - build without deps for debugging in other OS
# pip install -e . --no-deps
# - build without custom ops
# COMPILE_CUSTOM_KERNELS=0 pip install -e .
# Commit changed files using `-s`
git commit -sm "your commit info"

45
docs/source/installation.md
View File

@@ -123,10 +123,30 @@ First install system dependencies:
```bash
apt update -y
apt install -y gcc g++ libnuma-dev
apt install -y gcc g++ cmake libnuma-dev
```
You can install `vllm` and `vllm-ascend` from **pre-built wheel**:
Current version depends on a unreleased `torch-npu`, you need to install manually:
```
# Once the packages are installed, you need to install `torch-npu` manually,
# because that vllm-ascend relies on an unreleased version of torch-npu.
# This step will be removed in the next vllm-ascend release.
#
# Here we take python 3.10 on aarch64 as an example. Feel free to install the correct version for your environment. See:
#
# https://pytorch-package.obs.cn-north-4.myhuaweicloud.com/pta/Daily/v2.5.1/20250320.3/pytorch_v2.5.1_py39.tar.gz
# https://pytorch-package.obs.cn-north-4.myhuaweicloud.com/pta/Daily/v2.5.1/20250320.3/pytorch_v2.5.1_py310.tar.gz
# https://pytorch-package.obs.cn-north-4.myhuaweicloud.com/pta/Daily/v2.5.1/20250320.3/pytorch_v2.5.1_py311.tar.gz
#
mkdir pta
cd pta
wget https://pytorch-package.obs.cn-north-4.myhuaweicloud.com/pta/Daily/v2.5.1/20250320.3/pytorch_v2.5.1_py310.tar.gz
tar -xvf pytorch_v2.5.1_py310.tar.gz
pip install ./torch_npu-2.5.1.dev20250320-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
```
Then you can install `vllm` and `vllm-ascend` from **pre-built wheel**:
```{code-block} bash
:substitutions:
@@ -156,25 +176,10 @@ pip install -e . --extra-index https://download.pytorch.org/whl/cpu/
```
:::
Current version depends on a unreleased `torch-npu`, you need to install manually:
```{note}
vllm-ascend will build custom ops by default. If you don't want to build it, set `COMPILE_CUSTOM_KERNELS=0` environment to disable it.
```
```
# Once the packages are installed, you need to install `torch-npu` manually,
# because that vllm-ascend relies on an unreleased version of torch-npu.
# This step will be removed in the next vllm-ascend release.
#
# Here we take python 3.10 on aarch64 as an example. Feel free to install the correct version for your environment. See:
#
# https://pytorch-package.obs.cn-north-4.myhuaweicloud.com/pta/Daily/v2.5.1/20250320.3/pytorch_v2.5.1_py39.tar.gz
# https://pytorch-package.obs.cn-north-4.myhuaweicloud.com/pta/Daily/v2.5.1/20250320.3/pytorch_v2.5.1_py310.tar.gz
# https://pytorch-package.obs.cn-north-4.myhuaweicloud.com/pta/Daily/v2.5.1/20250320.3/pytorch_v2.5.1_py311.tar.gz
#
mkdir pta
cd pta
wget https://pytorch-package.obs.cn-north-4.myhuaweicloud.com/pta/Daily/v2.5.1/20250320.3/pytorch_v2.5.1_py310.tar.gz
tar -xvf pytorch_v2.5.1_py310.tar.gz
pip install ./torch_npu-2.5.1.dev20250320-cp310-cp310-manylinux_2_17_aarch64.manylinux2014_aarch64.whl
```
::::
::::{tab-item} Using docker