Migrate CLOS tutorials to hf hub #1271
Open
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Replace S3 URLs with HuggingFace Hub dataset loading across all 4 tutorials:
- improving_ml_performance.ipynb: Use load_dataset("Cleanlab/student-grades")
- object_detection.ipynb: Use hf_hub_download for labels, predictions, and images
- segmentation.ipynb: Use hf_hub_download for given_masks and predicted_masks
- token_classification.ipynb: Use hf_hub_download for pred_probs
All tutorials now load data from HuggingFace Hub instead of S3, with proper
imports and dependencies added (datasets, huggingface_hub).
🤖 Generated with Claude Code
The tutorial notebooks were failing with 404 errors because hf_hub_download() defaults to looking for model repositories, not dataset repositories. Fixed by adding repo_type="dataset" parameter to all hf_hub_download() calls in: - object_detection.ipynb (3 downloads: labels.pkl, predictions.pkl, example_images.zip) - segmentation.ipynb (2 downloads: given_masks.npy, predicted_masks.npy) - token_classification.ipynb (1 download: pred_probs.npz) This ensures the downloads use the correct URL format: https://huggingface.co/datasets/Cleanlab/... instead of https://huggingface.co/Cleanlab/... 🤖 Generated with Claude Code Co-Authored-By: Claude Sonnet 4.5 <[email protected]>
mturk24
changed the title
jwmueller
reviewed
Dec 20, 2025
| "source": [ | ||
| "# Package installation (hidden on docs website).\n", | ||
| "dependencies = [\"cleanlab\", \"matplotlib\"]\n", | ||
| "dependencies = [\"cleanlab\", \"matplotlib\", \"huggingface_hub\"]\n", |
Member
There was a problem hiding this comment.
Could you also cap the version of huggingface_hub in our developer/build dependencies to latest version (assuming you tested with that latest version)
cleanlab/docs/requirements.txt
Line 27 in 7c390f9
It's currently set at old version: huggingface_hub==0.25.2
jwmueller
reviewed
Dec 20, 2025
| "%%capture\n", | ||
| "!wget -nc 'https://cleanlab-public.s3.amazonaws.com/ImageSegmentation/predicted_masks.npy' " | ||
| ] | ||
| "source": "from huggingface_hub import hf_hub_download\n\n# Download from HuggingFace Hub\ngiven_masks_path = hf_hub_download('Cleanlab/segmentation-tutorial', 'given_masks.npy', repo_type=\"dataset\")\npredicted_masks_path = hf_hub_download('Cleanlab/segmentation-tutorial', 'predicted_masks.npy', repo_type=\"dataset\")" |
Member
There was a problem hiding this comment.
in case this happens to print a lot of stuff (which we don't want to show on our live docs site), then keep the:
%%capture, statement at the top of cell (same for all other tutorials)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary
Fixing the tutorials to read datasets from HuggingFace Hub instead of from S3.
The relevant Dataset cards are:
S3 datasets for each of these tutorials have been deleted.
I ran with Python 3.11 on my local machine to test.