Skip to content
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

Jump to bottom

Update Quantization Save Defaults #22

Merged
merged 4 commits into from
Jul 15, 2024
Merged

Update Quantization Save Defaults #22

merged 4 commits into from
Jul 15, 2024

Conversation

Satrat
Copy link
Contributor

@Satrat Satrat commented Jul 12, 2024 •
edited
Loading

SUMMARY:
We now default to save_compressed=True when running oneshot() or save_pretrained() so quantized models are compressed automatically. Also updated the default save formats as follows

  • w4a16 - packed-quantized
  • w8a16 - packed-quantized
  • w8a8 - int8 - int-quantized
  • w8a8 - fp8 - float-quantized
  • w4a16 2:4 - marlin-24
  • w8a16 2:4 - marlin-24

TEST PLAN:
Added unit test for default save formats

@bfineran bfineran merged commit 50642a6 into main Jul 15, 2024
8 of 12 checks passed
@bfineran bfineran deleted the sa/quant_defaults branch July 15, 2024 14:37
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@bfineran bfineran bfineran approved these changes

@robertgshaw2-neuralmagic robertgshaw2-neuralmagic Awaiting requested review from robertgshaw2-neuralmagic

@markurtz markurtz Awaiting requested review from markurtz

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@Satrat @bfineran