See some comparisons below comparison . wikipedia
We will use tensorflow 2 in our course. Some of the examples will be given by new edition of Dive into Deep Learning. It is a very good open book with jupyter notebooks and new edition uses numpy for most of its examples.
- tensorflow
- keras
- pytorch
- fastai built on top of pytorch
- mxnet supported by amazon.
- chainer, company behind it will switch to pytorch, see this web page
- Caffe
- DL4J
- weka deep learning used DL4J as backend.
- Matlab Deep Learning
- torch
- Microsoft Cognitive Toolkit (CNTK)
- Theano
- https://github.com/PaddlePaddle/Paddle
