一、創建安裝賬號

groupadd HwHiAiUser     
useradd -g HwHiAiUser -d /home/HwHiAiUser -m HwHiAiUser -s /bin/bash

二、下載依賴包

以下包根據需求自行下載

wget https://mirrors.tuna.tsinghua.edu.cn/anaconda/miniconda/Miniconda3-py39_24.5.0-0-Linux-aarch64.sh
wget https://ascend-repo.obs.cn-east-2.myhuaweicloud.com/Ascend%20HDK/Ascend%20HDK%2023.0.0/Ascend-hdk-910-npu-driver_23.0.0_linux-aarch64.run?response-content-type=application/octet-stream
wget https://ascend-repo.obs.cn-east-2.myhuaweicloud.com/Ascend%20HDK/Ascend%20HDK%2023.0.0/Ascend-hdk-910-npu-firmware_7.1.0.3.220.run?response-content-type=application/octet-stream
wget https://ascend-repo.obs.cn-east-2.myhuaweicloud.com/CANN/CANN%207.0.0/Ascend-cann-kernels-910_7.0.0_linux.run?response-content-type=application/octet-stream
wget https://ascend-repo.obs.cn-east-2.myhuaweicloud.com/CANN/CANN%207.0.0/Ascend-cann-toolkit_7.0.0_linux-aarch64.run?response-content-type=application/octet-stream

三、安裝環境依賴包

sudo yum install -y gcc gcc-c++ make cmake unzip zlib-devel libffi-devel openssl-devel pciutils net-tools sqlite-devel lapack-devel gcc-gfortran python3-devel

四、安裝環境

• 1、安裝python
  bash miniconda3-***.sh

安裝python依賴

pip3 install attrs  -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install numpy==1.26.0 -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install decorator -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install sympy -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install cffi -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install pyyaml -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install pathlib2 -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install psutil -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install protobuf -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install scipy -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install requests -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install absl-py -i https://pypi.tuna.tsinghua.edu.cn/simple

• 2、安裝驅動和固件
  bash xxx-driver_xxxx.run -full
  bash xxx-firmware_xxx.run -full

• 3、安裝CANN
  bash xxx_cann-toolkits_xxx.run --install --install-for-all
  bash xxx_cann-kernels_xxx.run --install --install-for-all

執行：source /usr/local/Ascend/ascend-toolkit/set_env.sh

• 4、安裝MindSpore
  • 首先安裝依賴包：

pip install sympy -i https://pypi.tuna.tsinghua.edu.cn/simple
pip install /usr/local/Ascend/ascend-toolkit/latest/lib64/te-*-py3-none-any.whl -i https://pypi.tuna.tsinghua.edu.cn/simple
pip install /usr/local/Ascend/ascend-toolkit/latest/lib64/hccl-*-py3-none-any.whl -i https://pypi.tuna.tsinghua.edu.cn/simple- 安裝mindspore
pip install https://ms-release.obs.cn-north-4.myhuaweicloud.com/2.2.14/MindSpore/unified/aarch64/mindspore-2.2.14-cp39-cp39-linux_aarch64.whl --trusted-host ms-release.obs.cn-north-4.myhuaweicloud.com -i https://pypi.tuna.tsinghua.edu.cn/simple

- 測試MindSpore

python -c "import mindspore;mindspore.set_context(device_target='Ascend');mindspore.run_check()"

五、下載Pytorch鏡像

docker pull swr.cn-south-292.ca-aicc.com/cloud-c86-images/mindspore_2.2.1-pytorch_2.1.0-cann_7.0-py_39-euler_2.8.3-aarch64-d910:ca-aicc_20231222

創建容器：

docker run -itd \
--ipc=host \
--network host \
--device=/dev/davinci0 \
--device=/dev/davinci1 \
--device=/dev/davinci2 \
--device=/dev/davinci3 \
--device=/dev/davinci4 \
--device=/dev/davinci5 \
--device=/dev/davinci6 \
--device=/dev/davinci7 \
--device=/dev/davinci_manager \
--device=/dev/devmm_svm \
--device=/dev/hisi_hdc \
-v /etc/localtime:/etc/localtime \
-v /usr/local/Ascend/driver:/usr/local/Ascend/driver \
-v /var/log/npu/:/usr/slog \
-v /usr/local/bin/npu-smi:/usr/local/bin/npu-smi \
-v /home/:/home/work \
--name PyTorch_NPU \
swr.cn-south-292.ca-aicc.com/cloud-c86-images/mindspore_2.2.1-pytorch_2.1.0-cann_7.0-py_39-euler_2.8.3-aarch64-d910:ca-aicc_20231222

打開容器：

docker exec -it -u 0 xxx /bin/bash