Skip to content

Install

Following steps are required to complete installing convmodel.

  1. Prepare Python3.8+ environment
  2. Install PyTorch
  3. Install convmodel

Prepare Python 3.8+

First, prepare Python 3.8+ environment.

Install PyTorch

Then install PyTorch >= 1.8,<=1.9. Please refer to official document to find out correct installation for your environment.

Some examples of installtion are as follows.

Install in Docker container without GPU

$ docker container run -w /work -v $(pwd):/work --rm -it python:3.8.6-slim-buster bash
(container)$ pip install torch==1.8.1

Install in Docker container enabling GPU and CUDA 11.1

Assume that CUDA 11.1 is installed in your environment.

$ docker container run --gpus all --ipc=host --rm -it -v $(pwd):/work -w /work nvidia/cuda:11.1-devel-ubuntu20.04 bash

--ipc option is required because share memory would not be enough because DataLoader multiprocess requires them. Refer to the pytorch discussion for more details.

Then install Python3.

(container)$ apt update && apt install -y python3 python3-pip git

Install PyTorch which corresponds to your environment by following the installation guide. For example, in CUDA 11.1 environment, PyTorch can be installed as follows.

(container)$ pip3 install torch==1.8.1+cu111 -f https://download.pytorch.org/whl/torch_stable.html

Install convmodel

Finally, install convmodel from PyPI.

$ pip install convmodel

If you want to run tests, specify [test] option to install dependencies.

$ pip install convmodel[test]