-
Hi,
Thanks for your excellent work. I am not sure the batchsize in your paper is same as it in the code? In code, 3072 refers to total tokens, corresponding to about real 32 examples each iteration. …
-
during training on VCR i got this error.
have you seen this error?
thank you
```
root@4a938104a7f8:/src# horovodrun -np 2 python train_vcr.py --config config/train-vcr-base-4gpu.json --output_di…
-
i got this problem.
how do i do?
thank you
```
root@da5f2b8bcf3b:/src# horovodrun -np 2 python train_vcr.py --config config/train-vcr-base-4gpu.json --output_dir /media/ailab/jaeyun/UNITER/outpu…
-
Hi, zhe;
Thanks for your excellent work. Recently I want to reproduce some results in Villa and conduct pre-training on indomain datasets. I am curious about whether it is possible to mimic the adv…
-
i got this error!
would you check https://convaisharables.blob.core.windows.net/uniter/txt_db_vcr ... server?
thank you:)
```
ailab@ailab:~/UNITER$ bash scripts/download_vcr.sh /media/ailab/ja…
-
The following does not work -
```
from model import *
config_file_path='UNITER/config/uniter-base.json'
checkpoint = torch.load('UNITER/data/pretrained/uniter-base.pt')
config=UniterConfig(con…
-
hi!
i try convert vcr dataset to itm task.
( i make data [CLS] question [SEP] answer [SEP] or [CLS] question [SEP] answer [SEP] rationale [SEP])
but i got error below.
the error start at "model/Un…
-
https://github.com/ChenRocks/UNITER/blob/feda6bb902085dbba198e201b791a7adba4c5ea2/model/model.py#L252
From the paper, I think there only one layernorm in the last step, but I notice there are more …
-
Hi,
Currently, there is only the in-domain pre-training config in this repo. Could you please provide the config files for the out-of-domain and the out-of-domain&in-domain pre-training process? Sinc…
-
Thank you for the great method UNITER (both code and the paper), but I really got confused when I read the detials of WRA, I'm so thankful if you are convenient to release the code of the parts of WRA…