search

Res2Net / Res2Net-PoolNet

Res2Net for Salient Object Detection using PoolNet
https://mmcheng.net/res2net/
MIT License
46 stars 8 forks source link

loss during the training of model #3

Closed ghazni123 closed 3 years ago

ghazni123 commented 3 years ago

Hi

Thank you for your work. I came across this and training. Initial values of the training are as

The number of parameters: 70452145 epoch: [ 0/48], iter: [ 0/65231] || Sal : 7288.5986 Learning rate: 5e-05 epoch: [ 0/48], iter: [ 50/65231] || Sal : 669608.8125 Learning rate: 5e-05

How do we know that model has learned well? does Sal value have to be near zero or when flatten on any number?

gasvn commented 3 years ago

Just use the default config and you will give the a converged trained model. Usually when the loss is stuck at a range, you can change to a smaller lr.