Skip to content

Add export support for TEQ #1910

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 5 commits into from
Jul 11, 2024
Merged

Add export support for TEQ #1910

Changes from 1 commit
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
ae231b2
export teq model
yiliu30 Jul 11, 2024
e508c94
fixed export
yiliu30 Jul 11, 2024
21c5e28
fixed the quant tensor
yiliu30 Jul 11, 2024
0a185d7
disable some check
yiliu30 Jul 11, 2024
3c531fe
merge with master
yiliu30 Jul 11, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
disable some check 
Signed-off-by: yiliu30 <[email protected]>
  • Loading branch information
@yiliu30
yiliu30 committed Jul 11, 2024
commit 0a185d790cd687bf35202877cddb5802a7b7973d
4 changes: 2 additions & 2 deletions neural_compressor/torch/algorithms/weight_only/teq.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -282,9 +282,9 @@ def quantize(self, **kwargs):
# for only group_dim is 0 or only `transformers.Conv1D`, we need transpose weight.
if is_transformers_imported():
transpose = (group_dim == 0) ^ (isinstance(m, transformers.Conv1D))
else:
else: # pragma: no cover
transpose = group_dim == 0
if transpose:
if transpose: # pragma: no cover
weight = m.weight.detach().T.contiguous()
else:
weight = m.weight.detach()
Expand Down