Skip to content

[NVFP4][WIP] Add NVFp4 Support #287

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

Draft
wants to merge 16 commits into
base: main
Choose a base branch
from
Draft

[NVFP4][WIP] Add NVFp4 Support #287

wants to merge 16 commits into from

Conversation

dsikka
Copy link
Collaborator

@dsikka dsikka commented Mar 31, 2025

Summary

  • Introduce FloatArg classes to manage float args data (min/min/rounding table)
  • Introduce new global_scale parameter
  • Update calculate_qparams, quant, dequant to handle global scale operations
  • Add NVFP4 QuantScheme to match the expected group_size expected

FP8_E4M3_DATA = FloatArgs(
exponent=4,
mantissa=3,
bits=8,
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

i think we usually use num_bits right? I changed AWQ Modifier to it based on what i saw elsewhere in llm comressor

Suggested change
bits=8,
num_bits=8,

Copy link
Collaborator Author

@dsikka dsikka Apr 1, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This PR isn’t ready for review lol. It’s set to Draft

@dsikka dsikka requested a review from anmarques April 2, 2025 20:05
@mgoin mgoin self-requested a review April 24, 2025 19:18
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants