- The installation is based on theano as the backend
- The selected system is Centos6 32-bit
- Install the required packages
conda install numpy scipy mkl <nose> <sphinx> <pydot-ng>
#<...>为可选的包
pip install parameterized
- The GPU accelerator is not installed, but the installation needs to download the GPU download
- Stable Installation
conda install theano pygpu
&
pip install Theano
- libgpuarray
git clone https://github.com/Theano/libgpuarray.git
cd libgpuarray
git checkout tags/v0.6.5 -b v0.6.9
- Bleeding-Edge Installation (recommended)
pip install <--user> <--no-deps> git+https://github.com/Theano/Theano.git#egg=Theano
- libgpuarray
conda install -c mila-udem pygpu
- Developer Installation
git clone git://github.com/Theano/Theano.git
cd Theano
<sudo> pip install <--user> <--no-deps> -e
sudo yum install python-devel python-nose python-setuptools gcc gcc-gfortran gcc-c++ blas-devel lapack-devel atlas-devel
sudo easy_install pip
- Keras framework construction
pip install -U --pre pip setuptools wheel
pip install -U --pre numpy scipy matplotlib scikit-learn scikit-image
pip install -U --pre tensorflow-gpu #可能执行不成功,没关系继续执行
pip install -U --pre keras
Once installed, type python, then type:
>> import tensorflow
>> import keras
Error free output
- Keras mnist dataset test download Keras development package
>>> git clone https://github.com/fchollet/keras.git
>>> cd keras/examples/
>>> python mnist_mlp.py
The program is carried out without errors, so far, the keras installation is complete