FoilHao

Results 10 comments of FoilHao

How to deal with the error: RuntimeError: value cannot be converted to type float without overflow: (0.000353832,-0.000114967)

I met the same problem. Have you solved it?

Test the pretrained model on my aerial images

Same question here, have you solved it?

The README.md may contain mistakes

Is the 'ship' required adjust to customized class?

Step into an endless loop when instance segment images

Hi, cao1124! I also meet endless loop problem. I still don't understand why this happened and how to solve it. Can you please help me out?

Step into an endless loop when instance segment images

@cao1124 @leohsuofnthu Also I'm puzzled why there is no endless loop problem with author's pretrained weight.

Step into an endless loop when instance segment images

> > @cao1124 @leohsuofnthu Also I'm puzzled why there is no endless loop problem with author's pretrained weight. > > Hi，FoilHao~You need to preprocess the image first ，make sure the...

Some question

[2020-04-23 23:57:48,790 INFO train.py line 65 60930] train_dice_coef: 0.0, S Loss: 259086.375, C Loss: 1.6436882019042969 [2020-04-23 23:57:50,109 INFO train.py line 228 60930] Iter -24-976: Loss: 259088.015625 acc: 0.000000% dice: 0.000000%...

Some question

Thank you! Your job do help me a lot and I've found the cause for some of my problems

How to handle gradient overflow when training a deep model with mixed precision?

Gradient overflow. Skipping step, loss scaler 0 reducing loss scale to 32768.0 repeats until it is 0 and get an error. Have anyone solved this?

How to handle gradient overflow when training a deep model with mixed precision?

Does your model train normally without fp16? I found my problem is that there are some nans in my dataset leading to nan gradient/gradient overflow. After correcting my data, it...