Minghao Liu

Results 6 comments of Minghao Liu

Thanks for that :) I may I ask when you generate the pickle file for under layout, how did you get that from your raw datasets?

Just let you know that image used in the docker file should be updated. This is the closest I found from the docker hub. (Basically using cuda:11.2.2 instead of cuda:11.2.0)...

Just a proof that the code runs on my machine. Thanks for the awesome paper and open sourced repo!!

Correct me if I am wrong. I manually set loss[1] in offset_calculate as zero, and still getting the same results `def offset_calculate from inversion.py line375`

Did you used 11 or 9? Thanks in advance.