Repo consists of a small code snippet that enables training in parallel if the machine has multiple GPUs installed with CUDA and cuDNN. The script works for tensorflow 2.0 and above version.
I have added the requirments files for making the installation process easy.