Tommy
Tommy
Hi, bro., I also meet the same problem, have you already fixed it?
@KaiWU5 I guess you may don't understand the problem proposed by @randydkx. In your code, you first calculate the `score_mask` according to Eq(6), it owns the shape of (batch_size, batch_size),...
Thanks for you reply, I got it! Now, I also set the ROI_Head_NMS=0.7 to get at least 10 precomputed box. But when I run your code, it seems the loss...
Hi, I'm very sorry for bothering you again. I have already checked the generated proposal, I find that when calculating the classification loss and regression loss, for each phrase, there...
Me too. Can you provide your processed scene graph for sentence? I found it is not easy to align the phrase in ground truth and processed phrases by the program...
Hi, Thanks for your attention, and sorry for the late reply. Have you already solved the problem? If not you can provide more error information for me to assist you...
We have already posted the validated environment we used, you can try to check whether it can address your issue.
@johnnylecy Have you ever tried our provided conda environment? We have already validated this environment and the code can run smoothly with this conda environment on the NVIDIA A100 with...
@johnnylecy Already approved, feel free to have a try~
@johnnylecy You can follow the practice of @Yuanyan266969, utilize our provided conda environment, and then solve the potential problem of `AttributeError: module 'distutils' has no attribute 'version'` by referring to...