Model won't train with tensorflow 1.11.0 about deepfacelab HOT 4 CLOSED

The normalize_data_format issue was in resolved #45. Have you merged in upstream master since then?

from deepfacelab.

Please refer to existing issues regarding tensorflow and CUDA updates. #48. We are waiting for tensorflow 1.13 to ship before migrating. This version is required before we can support updates to CUDA, dlib, and tensorflow-gpu.

from deepfacelab.

Thank you, @TooMuchFun for replying

I learned a lot playing with deepfacelab on a cloud console based Linux Machine (I am using ubuntu 18.04 on a cloud console controlled via SSH), but I still consider myself a beginner.
What is "upstream master"?

from deepfacelab.

In this case "upstream master" is just another way of saying the most recent commit to the master branch of this repo.

from deepfacelab.