Skip to content

Exclude multimodal vision branch from quantization by default (NVBug 6293731, 6293762) #1691

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
Edwardf0t1 merged 2 commits into from
Jun 12, 2026
+14 −0
Merged

Exclude multimodal vision branch from quantization by default (NVBug 6293731, 6293762) #1691

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
9e2acad
Fix gemma-4 fp8_default / nvfp4_mlp_only recipes quantizing vision br…
Edwardf0t1 Jun 11, 2026
513862e
Exclude multimodal vision branch from quantization by default (NVBug …
Edwardf0t1 Jun 12, 2026
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
14 changes: 14 additions & 0 deletions modelopt_recipes/configs/ptq/units/default_disabled_quantizers.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,6 +36,20 @@
enable: false
- quantizer_name: 'output.*'
enable: false
# Multimodal vision branch: keep the vision encoder (SigLIP / ViT) and any
# multimodal embedding projection in BF16 by default. Recipes that enable bare
# `*weight_quantizer` / `*input_quantizer` or `*mlp*` wildcards otherwise also
# match the vision tower (`model.vision_tower.*`, `model.visual.*`) and the
# embedding projection (`model.embed_vision.*`); quantizing the vision branch
# crashes export / produces garbage image embeddings on VL models (gemma-4,
# Qwen3.5-VL — NVBugs 6293731, 6293762, 6294017). A recipe that intentionally
# quantizes vision must re-enable these after importing this unit.
- quantizer_name: '*embed_vision*'

@meenchen meenchen Jun 12, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

I recently added vision tower and visual for qwen3.6:

Model-Optimizer/modelopt_recipes/configs/ptq/units/default_disabled_quantizers.yaml

Line 41 in 60b1af5

- quantizer_name: '*visual*'

Could you rebase and resolve the overlap?

@Edwardf0t1 Edwardf0t1 Jun 12, 2026

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

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

Let me do it in #1690

@Edwardf0t1 Edwardf0t1 Jun 12, 2026

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

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

Resolved in #1690 (commit 0cf494b). I rebased #1690 onto main and removed the duplicate bare *visual* / *vision_tower* entries your qwen3.6 change added, keeping the single documented block that disables *vision_tower* / *visual* / *embed_vision* — so each glob appears exactly once. I also dropped the now-redundant explicit vision excludes from the new huggingface/gemma4/ptq/w4a8_awq-kv_fp8_cast.yaml recipe since they're inherited from the shared unit. Verified load_recipe still resolves all three globs as disabled with *weight_quantizer enabled (INT4).

enable: false
- quantizer_name: '*vision_tower*'
enable: false
- quantizer_name: '*visual*'
enable: false
- parent_class: 'nn.BatchNorm1d'
quantizer_name: '*'
enable: false
Expand Down
Loading