Issues
- 0
The reproduced result is lower than expected.
#13 opened by hssdsb - 1
DarkNet53-pretrain
#11 opened by OliverHuang1220 - 2
DarkNet53-pretrain
#12 opened by OliverHuang1220 - 0
- 19
- 4
Does the pretrained visual backbone exclude val/test images on refcoco/+/g datasets ?
#7 opened by seanzhuh - 1
Trained models pickle files corrupted?
#8 opened by dreichCSL - 1
- 2
- 1
torch and Cuda's version
#5 opened by OliverHuang1220 - 0
How can I get the landmarkconv module?
#3 opened by shydyl - 2
Training getting stopped after sometime
#2 opened by abhinavkaul95 - 6
No module named 'landmarks'
#1 opened by abhinavkaul95