Skip to content

Add support for BCE loss for reward decomposition.#573

Closed
j-jiafei wants to merge 1 commit intofacebookresearch:mainfrom
j-jiafei:export-D31783145
Closed

Add support for BCE loss for reward decomposition.#573
j-jiafei wants to merge 1 commit intofacebookresearch:mainfrom
j-jiafei:export-D31783145

Conversation

@j-jiafei
Copy link

Summary:
Add BCE loss for reward decomposition.

BCE loss decomposition loss is defined between avg(p) and avg(y).

Differential Revision: D31783145

@facebook-github-bot
Copy link

This pull request was exported from Phabricator. Differential Revision: D31783145

1 similar comment
@facebook-github-bot
Copy link

This pull request was exported from Phabricator. Differential Revision: D31783145

j-jiafei pushed a commit to j-jiafei/ReAgent that referenced this pull request Nov 3, 2021
)

Summary:
Pull Request resolved: facebookresearch#573

Add BCE loss for reward decomposition.

BCE loss decomposition loss is defined between avg(p) and avg(y).

Reviewed By: czxttkl

Differential Revision: D31783145

fbshipit-source-id: af46052812f517f5679babc9be2a7633e79f64be
j-jiafei pushed a commit to j-jiafei/ReAgent that referenced this pull request Nov 3, 2021
)

Summary:
Pull Request resolved: facebookresearch#573

Add BCE loss for reward decomposition.

BCE loss decomposition loss is defined between avg(p) and avg(y).

Reviewed By: czxttkl

Differential Revision: D31783145

fbshipit-source-id: 180134dba517ca507ab86693bf165221d08fa21d
@facebook-github-bot
Copy link

This pull request was exported from Phabricator. Differential Revision: D31783145

j-jiafei pushed a commit to j-jiafei/ReAgent that referenced this pull request Nov 3, 2021
)

Summary:
Pull Request resolved: facebookresearch#573

Add BCE loss for reward decomposition.

BCE loss decomposition loss is defined between avg(p) and avg(y).

Reviewed By: czxttkl

Differential Revision: D31783145

fbshipit-source-id: 83807361904ca198d2b5470fd0868877f8dd5a0f
@facebook-github-bot
Copy link

This pull request was exported from Phabricator. Differential Revision: D31783145

j-jiafei pushed a commit to j-jiafei/ReAgent that referenced this pull request Nov 3, 2021
)

Summary:
Pull Request resolved: facebookresearch#573

Add BCE loss for reward decomposition.

BCE loss decomposition loss is defined between avg(p) and avg(y).

Reviewed By: czxttkl

Differential Revision: D31783145

fbshipit-source-id: d58ced069be20f472bf3c64b00efab7ec64e7918
@facebook-github-bot
Copy link

This pull request was exported from Phabricator. Differential Revision: D31783145

)

Summary:
Pull Request resolved: facebookresearch#573

Add BCE loss for reward decomposition.

BCE loss decomposition loss is defined between avg(p) and avg(y).

Reviewed By: czxttkl

Differential Revision: D31783145

fbshipit-source-id: 9fe83444d47243e0507b861638644cf0819c4c8d
@facebook-github-bot
Copy link

This pull request was exported from Phabricator. Differential Revision: D31783145

@facebook-github-bot
Copy link

This pull request has been merged in e2c2674.

xuruiyang pushed a commit that referenced this pull request Sep 20, 2025
Summary:
Pull Request resolved: #573

Add BCE loss for reward decomposition.

BCE loss decomposition loss is defined between avg(p) and avg(y).

Reviewed By: czxttkl

Differential Revision: D31783145

fbshipit-source-id: fd5c46f406a708ed3f9b9323b70cfdf8252478ec
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants