Our repo. is the Winner of ⚡#PoweredByTF 2.0 Challenge!.
Timeline:
make sure you are using python 3.x.
CPU install
pip install tensorflow -U
GPU install
Install CUDA 10.0
(or after) and cudnn
by yourself. and set LD_LIBRARY_PATH
up.
pip install tensorflow-gpu -U
Test installation:
In [2]: import tensorflow as tf
In [3]: tf.__version__
Out[3]: '2.0.0'
In [4]: tf.test.is_gpu_available()
...
totalMemory: 3.95GiB freeMemory: 3.00GiB
...
Out[4]: True
TensorFlow 2.0的视频教程链接:深度学习与TensorFlow 2实战
Feel free to submit a PR request to make this repo. more complete!
Our work is not built from scratch. Great appreciation to these open works!