2. Installation

2.1. Install UCTB

2.1.1. Step 1: Install TensorFlow

You can skip to step 2 if you already installed tensorflow.

You can refer to this page https://www.tensorflow.org/install to install tensorflow, if you have a Nvidia GPU installed on you computer, we highly recommend you to install GPU version of tensorflow.

2.1.2. Step 2: Install UCTB

pip install --upgrade UCTB

The following required package will be installed or upgraded with UCTB:

'hmmlearn',
'keras',
'GPUtil',
'numpy',
'pandas',
'python-dateutil',
'scikit-learn',
'scipy',
'statsmodels',
'wget',
'xgboost',
'nni',
'chinesecalendar',
'PyYAML'

2.2. UCTB Docker

You can also use UCTB by docker. First pull uctb docker from docker hub.

docker pull dichai/uctb:v0.2.0

And you then can run it.

docker run  --runtime=nvidia  -it -d dichai/uctb:v0.2.0 /bin/bash

Back To HomePage