Seohyeong Jeong

To be more concrete, segmentation masks that I've made running your code gives me .npy format files and it seems like probability value is assigned to each of the location(pixel)...

Thanks for the comment! I configured running `accelerate config`. However, I'm getting another error. Though it seems like either a python `subprocess` error or a compatibility problem between `accelerate` and...

Pytorch is installed. I installed accelerate through `pip install accelerate` and ran `accelerate config` right after.

@sreesindhu-sabbineni You don't submit the trained model. Segmentation task for example, You need to generate segmentation masks one for each test image running inference on your model and submit .png...

@SafaMessaoud Were you able to submit your results? I had a similar issue and things that I've missed were 1) zip into tar.gz format (not tar) 2) size of segmentation...

I think bert-joint repository is written in python 2. Check this out if you want python 3 implementation (https://github.com/google-research/language/issues/15#issuecomment-504375411).

BERT-joint uses different vocab as original BERT. Vocab list used for bert-joint has replaced [unused=xx] tokens from bert original vocab keeping the total size as it is (original vocab.txt file...