tryanbot

I guess they just dont understand. LOL

why did you change the hyperparameter on EDet-D0? is there any particular reason?

What is the correct order to do that? pip install torch==1.4 -> python setup.py install on pytorch_encoding ? is it the same as doing pip install torch==1.4 -> pip install...

thanks, the error is disappear. however, I found a new error when I run python train.py --dataset ade20k --model deeplab --backbone resnest50 --aux --se-loss  could you please explain why...

@songguanglu : did the training finish? It will help a lot of people, it kinda hard to find the pretrained

no update regarding this?

super helpful, thank you so much.

I am be able to start the training with only 1 gpu, so I think the batch size is not a problem. but I tried the batch_size 4, and it's...

additional information, maybe it can help (or not)  this is the message when I interupt the process (after hanging for a while)

I tried by using 1 million before. the iteration number may not be a problem because the script run well in 1 gpu (regardless batchsize and iteration).