thangvubk / FEQE

Official code (Tensorflow) for paper "Fast and Efficient Image Quality Enhancement via Desubpixel Convolutional Neural Networks"
MIT License
128 stars 19 forks source link

Plans to add TensorFlow v2.0 support for the script? #8

Closed ghost closed 5 years ago

ghost commented 5 years ago

Hi I was trying to train the network with Nvidia RTX 2080 (only CUDA v10.0 support) and since the script requires TensorFlow v1.11.0 to run perfectly which in turn works only with CUDA v9.0, do you have any plans to add support for TensorFlow v2.0 lately? If not I would be happy to make a pull request for the same :)

ghost commented 5 years ago

Found the perfect set of dependencies' versions for running the code on Windows with Nvidia RTX 2080 Ti (CUDA v10.0 support). The requirements file can be found in the attachment. 👍

Still, it would be great to have a TensorFlow v2.0 implementation given that it is more suited for new environments.

Thanks! :) requirements.txt