Skip to content

Conversation

@MekkCyber
Copy link
Contributor

@MekkCyber MekkCyber commented Nov 27, 2025

What does this PR do?

Fix Quark after weight refactor, all tests passing locally

@MekkCyber MekkCyber requested a review from SunMarc November 27, 2025 15:24
Copy link
Member

@SunMarc SunMarc left a comment

Choose a reason for hiding this comment

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

Let's go

quant_state = target_key.split("_", 1)[-1]
# print(f"full_layer_name: {full_layer_name}")

sub_module_state = full_layer_name.rsplit(".", 1)[0] + "." + param + "_quantizer" + "." + quant_state
Copy link
Member

Choose a reason for hiding this comment

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

pls explain

EXPECTED_OUTPUTS.add("Today I am in Paris and I am enjoying my day off! The sun is shining, the birds are")
EXPECTED_OUTPUTS.add("Today I am in Paris and I'm here to tell you about it. It's a beautiful day,")
EXPECTED_OUTPUTS.add("Today I am in Paris and I am not in Paris at all! I am not in Paris, but")
EXPECTED_OUTPUTS.add("Today I am in Paris and I am in Paris, but I am not in Paris\nToday I am")
Copy link
Member

Choose a reason for hiding this comment

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

any idea why this changed ? the most likely culprit is the dtype we loaded the param but not sure

@HuggingFaceDocBuilderDev

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

Comment on lines 119 to 124
converters = []
for key in CHECKPOINT_KEYS.keys():
converters.append(
WeightConverter(source_keys=[key], target_keys=key, operations=[QuarkDeserialize(self)])
)
return converters
Copy link
Member

Choose a reason for hiding this comment

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

beautiful

Copy link
Member

@SunMarc SunMarc left a comment

Choose a reason for hiding this comment

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

Thanks !

@SunMarc
Copy link
Member

SunMarc commented Dec 1, 2025

@bot /style

@github-actions
Copy link
Contributor

github-actions bot commented Dec 1, 2025

Style bot fixed some files and pushed the changes.

@SunMarc
Copy link
Member

SunMarc commented Dec 3, 2025

@bot /style

@github-actions
Copy link
Contributor

github-actions bot commented Dec 3, 2025

Style fix is beginning .... View the workflow run here.

@SunMarc SunMarc requested a review from Cyrilvallez December 3, 2025 14:17
@github-actions
Copy link
Contributor

github-actions bot commented Dec 3, 2025

[For maintainers] Suggested jobs to run (before merge)

run-slow: quark_integration

@MekkCyber MekkCyber merged commit bebfab0 into main Dec 4, 2025
24 checks passed
@MekkCyber MekkCyber deleted the fix-quark branch December 4, 2025 12:30
sarathc-cerebras pushed a commit to sarathc-cerebras/transformers that referenced this pull request Dec 7, 2025
* intial

* update

* add convert

* fix

* style

* rm comment

* explain

* loop

* fix

* fix

* update

* Apply style fixes

* fix

* style

* update

---------

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
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.

4 participants