Open klevin-ken opened 5 days ago
In this paper, we train our model on a single 3090 GPU for training stage one, and use two 3090 GPU for training stage two. It is worthy to mention that, we will release the code for training stage two before 2024. 07. 14.
Thanks!
Hello,
I am interested in using this project and would like to know more about the required compute resources.