LJX

Results 12 comments of LJX

Can I train a num_class is 26(>21) model use the SSD-512 VGG-based pre-trained checkpoints?

@blacklong617 hello , do you sovle the problem? I am trying training a num_class >70 model ...

It doesn't work on my data, why don't you provide a pretrained model?

> > I pretrained a model using ctcloss and it works well. Then I loaded the weights and continued to train with the aceloss. The losses seemed to be coming...