Training Base Image (MPI)

This section describes preset mindspore_1.3.0 images.

Engine Version: mindspore_1.3.0-cuda_10.1-py_3.7-ubuntu_1804-x86_64

Image address: swr.{region}.myhuaweicloud.com/aip/mindspore_1_3_0:train-mindspore_1.3.0-cuda_10.1-py_3.7-ubuntu_18.04-x86_64-roma-20211104202338-f258e59
Image creation time: 20211104202338(yyyy-mm-dd-hh-mm-ss)
Image system version: Ubuntu 18.04.4 LTS
cuda: 10.1.243
cudnn: 7.6.5.32
Path and version of the Python interpreter: /home/ma-user/anaconda3/envs/MindSpore-1.3.0-gpu/bin/python, python 3.7.10
Third-party package installation path: /home/ma-user/anaconda3/envs/MindSpore-1.3.0-gpu/lib/python3.7/site-packages

The versions of some third-party packages:

requests 2.26.0
dask 2021.9.0
easydict 1.9
enum34 1.1.10
mindspore-gpu 1.3.0
Flask 1.1.1
grpcio 1.41.1
gunicorn 20.1.0
idna 3.3
PyYAML 5.1
imageio 2.10.1
imgaug 0.4.0
lxml 4.6.4
matplotlib 3.4.2
psutil 5.8.0
scikit-image 0.18.3
numba 0.47.0
numpy 1.17.0
opencv-python 4.5.2.54
tifffile 2021.11.2
pandas 1.1.5
Pillow 8.4.0
pip 21.0.1
protobuf 3.17.3
scikit-learn 0.22.1
...