How to finish the training process?

TobyPDE / FRRN

Full Resolution Residual Networks for Semantic Image Segmentation

MIT License

278 stars 93 forks source link

How to finish the training process? #12

Closed Danielll2 closed 7 years ago

Danielll2 commented 7 years ago

Is there any number of iteration of training? or Just put Ctrl+C to stop training?

TobyPDE commented 7 years ago

I usually perform manual early stopping by watching the validation IoU score and terminating the training process via CTRL + C once it no longer improves.

Danielll2 commented 7 years ago

Thanks a lot!

TobyPDE commented 7 years ago

Quick heads up: I'm currently working on significantly speeding up training by optimizing the block-backprop algorithm and porting some code to C++. I will publish the code within the next 1-2 weeks.