How do you train DRN-D-22?
acgtyrant opened this issue · comments
I trained it with python3 segment.py train -d <data_folder> -c <category_number> -s 896 --arch drn_d_22 --batch-size 14 --epochs 250 --lr 0.001 --momentum 0.99 --step 100
, and multi-scale test only get 66.49.
It seems that you have many GPUs, you may use a huge batch-size or big crop size, right?
OK, I forget to use --bn-sync
.
I add --bn-sync
and it overflow in epoch 3, it comfirms that the implementation of synced bn is not correct.
Hi @acgtyrant would you be able to upload the models to e.g. dropbox as they don't seem to be accessible on the Princeton site? That'd be awesome! : 👏 😎