Open
Description
Hello, I am very interested in your work OSEDiff and I am currently trying to reproduce it. However, I have some confusion about the parameter settings for training. I noticed that the article mentions that the entire training process only takes one day on 4 A100 GPUs, but the code does not provide the number of training iterations, only the maximum training iteration of 100K. I would like to know how many training iterations are required to achieve the results in the paper？
Thank you for your help！
Metadata
Metadata
Assignees
Labels
No labels