Installation¶
Stable version¶
Install deepchem via pip or conda by simply running,
pip install deepchem
or
conda install -c conda-forge deepchem
Nightly build version¶
The nightly version is built by the HEAD of DeepChem.
For using general utilites like Molnet, Featurisers, Datasets, etc, then, you install deepchem via pip.
pip install --pre deepchem
Deepchem provides support for tensorflow, pytorch, jax and each require a individual pip Installation.
For using models with tensorflow dependencies, you install using
pip install --pre deepchem[tensorflow]
For using models with Pytorch dependencies, you install using
pip install --pre deepchem[torch]
For using models with Jax dependencies, you install using
pip install --pre deepchem[jax]
If GPU support is required, then make sure CUDA is installed and then install the desired deep learning framework using the links below before installing deepchem
tensorflow - just cuda installed
pytorch - https://pytorch.org/get-started/locally/#start-locally
jax - https://github.com/google/jax#pip-installation-gpu-cuda
In zsh
square brackets are used for globbing/pattern matching. This means
you need to escape the square brackets in the above installation. You can do so by
including the dependencies in quotes like pip install --pre 'deepchem[jax]'
Note: Support for jax is not available in windows (jax is not officially supported in windows).
Google Colab¶
The fastest way to get up and running with DeepChem is to run it on Google Colab. Check out one of the DeepChem Tutorials or this forum post for Colab quick start guides.
Docker¶
If you want to install using a docker, you can pull two kinds of images from DockerHub.
deepchemio/deepchem:x.x.x
Image built by using a conda (x.x.x is a version of deepchem)
This image is built when we push x.x.x. tag
Dockerfile is put in `docker/tag`_ directory
deepchemio/deepchem:latest
Image built from source codes
This image is built every time we commit to the master branch
Dockerfile is put in `docker/nightly`_ directory
First, you pull the image you want to use.
docker pull deepchemio/deepchem:latest
Then, you create a container based on the image.
docker run --rm -it deepchemio/deepchem:latest
If you want GPU support:
# If nvidia-docker is installed
nvidia-docker run --rm -it deepchemio/deepchem:latest
docker run --runtime nvidia --rm -it deepchemio/deepchem:latest
# If nvidia-container-toolkit is installed
docker run --gpus all --rm -it deepchemio/deepchem:latest
You are now in a docker container which deepchem was installed. You can start playing with it in the command line.
(deepchem) root@xxxxxxxxxxxxx:~/mydir# python
Python 3.6.10 |Anaconda, Inc.| (default, May 8 2020, 02:54:21)
[GCC 7.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import deepchem as dc
If you want to check the tox21 benchmark:
# you can run our tox21 benchmark
(deepchem) root@xxxxxxxxxxxxx:~/mydir# wget https://raw.githubusercontent.com/deepchem/deepchem/master/examples/benchmark.py
(deepchem) root@xxxxxxxxxxxxx:~/mydir# python benchmark.py -d tox21 -m graphconv -s random
From source with conda¶
Installing via these steps will ensure you are installing from the source.
Prerequisite
Shell: Bash, Zsh, PowerShell
Conda: >4.6
First, please clone the deepchem repository from GitHub.
git clone https://github.com/deepchem/deepchem.git
cd deepchem
Then, execute the shell script. The shell scripts require two arguments, python version and gpu/cpu.
source scripts/install_deepchem_conda.sh 3.8 cpu
If you want GPU support (we supports only CUDA 10.1):
source scripts/install_deepchem_conda.sh 3.8 gpu
If you are using the Windows and the PowerShell:
. .\scripts\install_deepchem_conda.ps1 3.7 cpu
(XXXX)
in your command line.conda init <YOUR_SHELL_NAME>
to activate it, then:conda activate deepchem
pip install -e .
pytest -m "not slow" deepchem # optional
From source lightweight guide¶
Installing via these steps will ensure you are installing from the source.
Prerequisite
Shell: Bash, Zsh, PowerShell
Conda: >4.6
First, please clone the deepchem repository from GitHub.
git clone https://github.com/deepchem/deepchem.git
cd deepchem
We would advise all users to use conda environment, following below-
conda create --name deepchem python=3.8
conda activate deepchem
pip install -e .
DeepChem provides diffrent additional packages depending on usage & contribution If one also wants to build the tensorflow environment, add this
pip install -e .[tensorflow]
If one also wants to build the Pytorch environment, add this
pip install -e .[torch]
If one also wants to build the Jax environment, add this
pip install -e .[jax]
DeepChem has soft requirements, which can be installed on the fly during development inside the environment but if you want to install all the soft-dependencies at once, then take a look at deepchem/requirements