Ability to fuse non-square (pruned) attention weights for BERT-like models #6850
mfuntowicz
marked this pull request as draft 4 years ago
mfuntowicz
marked this pull request as ready for review 4 years ago
tianleiwu
dismissed these changes
on 2021-03-02
Ability to fuse non-square (pruned) attention weights.
76e80db6
Fix invalid comment.
0735c801
Trigger CI
93593ae2
mfuntowicz
dismissed their stale review
via 93593ae2
4 years ago
mfuntowicz
force pushed
from
8d906d36
to
93593ae2
4 years ago
tianleiwu
approved these changes
on 2021-03-05
tianleiwu
merged
9126faa3
into master 4 years ago
mfuntowicz
deleted the hf_fuse_pruned_attention branch 4 years ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub