Skip to content

Conversation

@littleKitchen
Copy link
Contributor

Summary

Fixes #609

When pulling models from HuggingFace, users can use either hf.co or huggingface.co as the domain prefix. Previously, these were treated as different sources, causing the same model to be pulled and stored twice with different tags:

$ docker model inspect hf.co/mlx-community/llama-3.2-1b-instruct-4bit
{"id": "sha256:f130735e...", "tags": ["hf.co/mlx-community/llama-3.2-1b-instruct-4bit:latest"]}

$ docker model inspect huggingface.co/mlx-community/llama-3.2-1b-instruct-4bit  
{"id": "sha256:ce26d336...", "tags": ["huggingface.co/mlx-community/llama-3.2-1b-instruct-4bit:latest"]}

Changes

  • Normalize hf.co/ URLs to huggingface.co/ in the normalizeModelName function
  • Added test cases for the normalization behavior

Result

After this change, both URL formats resolve to the same canonical form:

  • hf.co/org/modelhuggingface.co/org/model:latest
  • huggingface.co/org/modelhuggingface.co/org/model:latest

This ensures models are not duplicated regardless of which URL format the user uses.

When pulling models from HuggingFace, users can use either hf.co or
huggingface.co as the domain. Previously, these were treated as
different sources, causing the same model to be stored twice with
different tags.

This change normalizes hf.co URLs to huggingface.co in the
normalizeModelName function, ensuring that:
- hf.co/org/model and huggingface.co/org/model are treated as the same
- Models are not pulled/stored twice when using different URL formats

Fixes docker#609
@gemini-code-assist
Copy link
Contributor

Summary of Changes

Hello @littleKitchen, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request resolves an issue where models from HuggingFace could be inadvertently duplicated in the system's storage. By introducing a normalization step for HuggingFace URLs, the system now consistently recognizes models regardless of whether they are referenced using the hf.co shorthand or the full huggingface.co domain. This enhancement improves data integrity and resource management by preventing redundant model downloads and storage.

Highlights

  • HuggingFace URL Normalization: Implemented logic to normalize hf.co URLs to their canonical huggingface.co form within the normalizeModelName function. This ensures that models referenced via either domain are treated as the same entity.
  • Preventing Model Duplication: This change directly addresses and fixes an issue where models from HuggingFace could be pulled and stored twice with different tags if users specified hf.co or huggingface.co inconsistently.
  • Test Coverage: New test cases have been added to normalize_test.go to validate the correct behavior of the HuggingFace URL normalization, ensuring robustness and preventing regressions.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature Command Description
Code Review /gemini review Performs a code review for the current pull request in its current state.
Pull Request Summary /gemini summary Provides a summary of the current pull request in its current state.
Comment @gemini-code-assist Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help /gemini help Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

  1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution.

Copy link
Contributor

@sourcery-ai sourcery-ai bot left a comment

Choose a reason for hiding this comment

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

Hey - I've reviewed your changes and they look great!


Sourcery is free for open source - if you like our reviews please consider sharing them ✨
Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.

Copy link
Contributor

@gemini-code-assist gemini-code-assist bot left a comment

Choose a reason for hiding this comment

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

Code Review

This pull request correctly normalizes hf.co URLs to huggingface.co to prevent duplicate model storage. The implementation is straightforward and includes relevant test cases. I've suggested a minor refinement to improve code efficiency and style. Overall, this is a good fix.

Comment on lines 128 to 130
if strings.HasPrefix(model, "hf.co/") {
model = "huggingface.co/" + strings.TrimPrefix(model, "hf.co/")
}
Copy link
Contributor

Choose a reason for hiding this comment

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

medium

For better performance and to make the code more idiomatic, you can use strings.CutPrefix. It combines strings.HasPrefix and strings.TrimPrefix into a single call, avoiding scanning the string for the prefix twice.

Additionally, consider defining "hf.co/" and "huggingface.co/" as package-level constants to improve maintainability, as they are used in multiple functions within this file (isHuggingFaceReference, parseHFReference).

Suggested change
if strings.HasPrefix(model, "hf.co/") {
model = "huggingface.co/" + strings.TrimPrefix(model, "hf.co/")
}
if newModel, found := strings.CutPrefix(model, "hf.co/"); found {
model = "huggingface.co/" + newModel
}

Copy link
Contributor

@ilopezluna ilopezluna left a comment

Choose a reason for hiding this comment

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

Thanks for working on this, @littleKitchen
Could you use strings.CutPrefix as suggested by Gemini? 🙏

I’ve already approved it since it looks good to go.
While testing your changes, I noticed that we don’t check the local store before pulling from Hugging Face, so the model gets pulled again even when using the exact same reference. I’ll open a new issue for that, if you want to work on it, it’s all yours!

Updated: Added the issue here

@littleKitchen
Copy link
Contributor Author

Thanks for the review! Will update to use strings.CutPrefix. And yes, will work on #616

@littleKitchen
Copy link
Contributor Author

Done!

@ilopezluna ilopezluna merged commit 36c9c73 into docker:main Jan 31, 2026
6 checks passed
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.

Prevent pulling hf.co and huggingface.co as different models

2 participants