Installation

GPU Machine

Deep learning models in ELIT require a GPU machine. Take a look at the pre-trained models page to check how much GPU memories are needed for models of your choice. If you do not have a GPU machine powerful enough to run those models, you may want to consider using our web APIs.

CUDA Environment

ELIT requires the following CUDA environment:

Supported versions of the CUDA Toolkit: 8.0, 9.0, 9.2, 10.0, 10.1.
CUDA Deep Neural Network library: cuDNN.

The followings show how to install the CUDA Toolkit 10.1 on Ubuntu 18.04:

$ wget http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/cuda-repo-ubuntu1804_10.1.168-1_amd64.deb
$ sudo dpkg -i cuda-repo-ubuntu1804_10.1.168-1_amd64.deb
$ sudo apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/7fa2af80.pub
$ sudo apt update
$ sudo apt install cuda
$ nvidia-smi
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 418.67       Driver Version: 418.67       CUDA Version: 10.1     |
|-------------------------------+----------------------+----------------------+

The followings show how to install the cuDNN 7.6.0.64 on Ubuntu 18.04:

$ wget https://developer.download.nvidia.com/compute/machine-learning/repos/ubuntu1804/x86_64/libcudnn7_7.6.0.64-1+cuda10.1_amd64.deb
$ sudo dpkg -i libcudnn7_7.6.0.64-1+cuda10.1_amd64.deb
$ sudo apt install libcudnn7
libcudnn7 is already the newest version (7.6.0.64-1+cuda10.1).

Python Environment

ELIT requires Python ≥ 3.6. The followings show how to install the Python 3.6.8 on Ubuntu 18.04:

$ sudo apt install python3.6
$ sudo apt install python3.6-dev
$ sudo apt install python3-setuptools
$ sudo apt install python3-pip
$ sudo apt install python-virtualenv
$ python3 --version
Python 3.6.8

Virtual Environment

We recommend to install ELIT using Virtualenv. The followings show how to setup a virtualenv using Python 3.6:

$ virtualenv --python=/usr/bin/python3.6 ~/.elit
$ source ~/.elit/bin/activate
(.elit) $

MXNet Installation

ELIT uses Apache MXNet to develop deep learning models. The followings show how to install MXNet 1.4.1 based on CUDA Toolkit 10.1:

(.elit) $ pip install mxnet-cu101
(.elit) $ pip show mxnet-cu101
Name: mxnet-cu101
Version: 1.4.1

ELIT Installation

Finally, the followings show how to install the latest version of ELIT:

(.elit) $ pip install elit
Name: elit
Version: 0.2.0