Skip to content

Latest commit

 

History

History
81 lines (50 loc) · 1.89 KB

README.md

File metadata and controls

81 lines (50 loc) · 1.89 KB
Raw

sjtu-openmmlab-tutorial

This repository maintains codes of OpenMMLab tutorial delivered in SJTU. For a full list of OpenMMLab courses and tutorials, see OpenMMLabCourse (in Chinese).

Notes on SJTU Cluster

For OpenMMLab v2.0

TODO

For OpenMMLab v1.0

  1. To activate conda environment in terminal

    module load anaconda3/2019.07
source activate openmmlab

  2. Remember to select openmmlab kernel when using jupyter notebook

  3. Remember to check PATH in terminal

    PATH should include /cluster/apps/anaconda3/2019.07/envs/openmmlab/bin at the first entry. Otherwise mim will not work correctly.

Set up Environment Locally

Step 0. Set up Python and install PyTorch correctly, using either pip or conda

Step 1. Install MIM

pip install openmim
which mim  # to check mim installation
# On windows, in Powershell, use
# gcm mim

For OpenMMLab v2.0

Step 2. Install MMCV

mim install "mmcv>=2.0.0rc0"

Step 2. Install MMClassification and MMDetection

mim install "mmcls>=1.0.0rc0" "mmdet>=3.0.0rc0"

For OpenMMLab v1.0

Step 2. Install MMCV

mim install mmcv-full

Step 3. Install MMClassification and MMDetection

mim install mmcls mmdet

License

Credit

This repository uses some datasets and models from Zihao](https://github.com/TommyZihao/MMClassification_Tutorials).

Fruit dataset

Fruit model

and MMDetection and KITTI.

KITTI-tiny