-
Notifications
You must be signed in to change notification settings - Fork 239
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.
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
[WC] Fix ratio_defining_params #2653
[WC] Fix ratio_defining_params #2653
Conversation
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## develop #2653 +/- ##
============================================
- Coverage 62.02% 29.87% -32.16%
============================================
Files 495 495
Lines 45973 45971 -2
============================================
- Hits 28516 13735 -14781
- Misses 17457 32236 +14779
... and 264 files with indirect coverage changes
Flags with carried forward coverage won't be shown. Click here to find out more.
|
ratio_defining_params = list( | ||
filter( | ||
lambda wp: wp.node_with_weight.metatype in self._backend_entity.matmul_metatypes, | ||
all_weight_params, | ||
) | ||
) | ||
|
||
# The last MatMul layer is quantized to 4-bits if all_layers=True or if the layer is shared |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This comment looks wrong.
if the layer is shared
means that it has the same precision as embedding, and it's 4bit only with all_layers
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, would be nice to correct comment, but it's minor
bbc7f71
to
15afc14
Compare
Changes
Reason for changes
Fix bug in
_get_ratio_defining_params
methodTests
test_shared_gather_all_layers