Walshyu

Results 2 issues of Walshyu

talk is cheap, show me the code

The paper proposes to use NVIDIA titan xp for inference. I want to ask how many GPU resources are needed to train the model?