This repository has been archived by the owner on Nov 3, 2023. It is now read-only.
Add new HRED agent and integration tests #2989
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
NathanShuster
changed the title
|
Test that is failing is that you need to add an |
added 4 commits
August 27, 2020 11:48
|
Tests will pass on master, depended on #3019. |
Merged
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Patch description
Implements a basic HRED agent agent that extends seq2seq. I originally implemented a heavier version that had a lot of dataset specific features for SIMMC and stripped out all the dataset specific stuff for this PR. Is a general implementation but has some room for added functionality (for example, currently only supports LSTM, not GRU) that is not in this PR.
Testing steps
As part of my intern project I reproduced the SIMMC (https://github.com/facebookresearch/simmc) HRED baseline in ParlAI, and successfully reproduced the original results. To generalize for broader use in ParlAI, I stripped out all the dataset specific features and consulted with Stephen in the process. I ran some follow up testing with the DailyDialog set, shared my findings with Stephen and my manager, and the numbers were deemed sufficiently close to send out a PR.
Logs
Can be run with any of the ParlAI datasets with -m seq2seq/hred