Drew Arad Hudson

Sorry for that! I'm pretty sure that was a bug in https://github.com/stanfordnlp/mac-network/blob/gqa/merge.py. I resolved it since then but if you ran it before there's some chance it still affects your...

Hi there! :) Thanks a lot! Yea that's actually a great idea, will do it!

not yet, haven't got a chance yet. It should take about ~12-18hr to train from scratch and will definitely upload a pretrained model at the first opportunity!

Hi yea definitely! The flag is `--dataVer`, you can set it either to 1 or 2, the default is actually 2, haven't tested it recently on 1. I made local...

Thanks a lot for the interest in the repo! I haven't tested recently the memory consumption, but yep I'm pretty sure it needs more RAM. It should be very possible...

Hi! thanks again for the interest! :) I actually have tried several different learning schedules, including both mixtures and first training on clevr and then fine-tuning, but not waiting too...

yep! :)

Hi, Thank you for the interest in the work! I have couple deadlines over the next days so will definitely try to get back to you by the end of...

Hi, most sincere apologies for not getting back to it earlier! The model can be trained by even a single GPU. On which line of the code did you get...

Hi, thanks for reaching out! The implementation doesn't support that at the moment but basically the script/approach that is used there could be used for our model too. I hope...