Question about running the code

[woosik Yang]

helllo 

First of all, thank you for your great work

It was good to see your presentation at ICCV in Seoul

I'm trying to apply your current code. 

While running allennlp train default.json -s my_dir --include-package my_package, 

I encountered an error regarding the type of dataset which you named 'vswag'

I guess that error comes from not making @DatasetRead.register('vswag')

The dataloaders/vcr.py does not you allennlp.DatasetReader, it only uses pytorch's module Dataset

and this is why I get the error mentioned above

Sine I'm new to allennlp module, am I taking the correct place to handle that issue??

Also

when I run the python train.py -params multiatt/default.json -folder saves/flagship_answer/

I get the configuration error that "MultiHopAttentionQA" is not acceptable for model.type

I think this comes from custom configuration 

Do you have any ideas??

Thank you for your reply 

[rowan....@gmail.com]

Hi Woosik! Thanks for the interest. I think you might have a wrong issue of allennlp installed, or maybe the wrong pythonpath. Can you try following the setup here? You'll need to replace export PYTHONPATH=/home/rowan/code/r2c with wherever you saved the code.


[woosik Yang]

Thanks now it works! :)