Caius-Lu

training is so slow

8

When I train on my own data set, the speed is significantly slower than yolov5. I have 8 a40. Is it because of ota loss? #136

add install docment

After using the pspnet model, I can only test 76.64 on the test set. What is the maximum amount you can reproduce?

## ❔Question In utils/loss.py: build_targets t[:, 6+2*kpt: 6+2*(kpt+1)][t[:,6+2*kpt: 6+2*(kpt+1)] !=0] -= gij[t[:,6+2*kpt: 6+2*(kpt+1)] !=0] What are the meaning of doing this? Is there a better way? ## Additional context