antoniorv6 / antoniorv6/SMT-plusplus

Attention weights from attention decoder

Open
#2 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
33
Forks
7
PR merge metrics
No merged PRs in 30d

Description

forward_decoder method does not provide an interface for deactivating the keep_all_weights option of the decoder forward, forcing the attention weights to use a lot of memory even if they are not needed.

Contributor guide

No contributing guide indexed for this repository

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.