Update wizard_of_wikipedia agents.py

[Reveyer]

Patch description

When wizard_first is True, the len_episode is incorrectly determined, resulting in some data not being used.

Testing steps

I tested it using command

parlai display_data --task wizard_of_wikipedia --num-examples 100000

Before the change

loaded 18430 episodes with a total of 74092 examples

After the change

loaded 18430 episodes with a total of 83247 examples

Other information

[facebook-github-bot]

Hi @Reveyer!

Thank you for your pull request and welcome to our community.

Action Required

In order to merge any pull request (code, docs, etc.), we require contributors to sign our Contributor License Agreement, and we don't seem to have one on file for you.

Process

In order for us to review and merge your suggested changes, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA.

Once the CLA is signed, our tooling will perform checks and validations. Afterwards, the pull request will be tagged with CLA signed. The tagging process may take up to 1 hour after signing. Please give it that time before contacting us about it.

If you have received this in error or have any questions, please contact us at cla@fb.com. Thanks!

[facebook-github-bot]

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Facebook open source project. Thanks!

[klshuster]

Hi @Reveyer, thank you for flagging this bug! In the interest of reproducibility, could you please make some minor modifications to the PR:

1. Create a subclass of WizardDialogKnowledgeTeacher, perhaps named LongTeacher, that overrides len_episode to fix the bug.
2. Somewhere in the __init__ for the WizardDialogKnowledgeTeacher, please loudly warn that it is advised to use -t wizard_of_wikipedia:long to include the missing examples from the default teacher.

Thanks again!

[klshuster]

Additionally, it would be great if we could limit this change to --datatype train (so, valid and test sets remain the same)

[Reveyer]

Additionally, it would be great if we could limit this change to --datatype train (so, valid and test sets remain the same)

I think adding an argument would better achieve what you said.

[klshuster]

thank you! yes, an argument works great here

[stephenroller]

Greatly appreciated