README.md

---
license: apache-2.0
task_categories:
- text-classification
---

# HuggingBench-Classification

This dataset contains Task Classification test collection in HuggingBench for paper "Benchmarking Recommendation, Classification, and Tracing Based on Hugging Face Knowledge Graph" ([https://huggingface.co/papers/2505.17507](https://huggingface.co/papers/2505.17507)).

## Dataset Details

- `classification.zip` contains training/validation/test set files and task-to-id file for multi-label classification.
- `huggingface_xxx.pt` is graph data file in the format required by [CogDL](https://github.com/THUDM/CogDL). `xxx` indicates the type of node feature initialization.

## Uses

<!-- Address questions around how the dataset is intended to be used. -->
We use [CogDL](https://github.com/THUDM/CogDL) to implemen baselines for task classification.
- Install [CogDL](https://github.com/THUDM/CogDL).
```
pip install cogdl
```
- Download data in [HuggingBench-Classification](https://huggingface.co/datasets/cqsss/HuggingBench-Classification) to `task_classification/data/`.
- Run `./task_classification/tune_huggingface.py` from [Github](https://github.com/nju-websoft/HuggingBench).

<table>
  <thead>
  </thead>
  <tbody>
    <tr>
      <td>💻 <strong>GitHub</strong></td>
      <td><a href="https://github.com/nju-websoft/HuggingBench">Code Reposity</a></td>
    </tr>
    <tr>
      <td>📄 <strong>Paper</strong></td>
      <td><a href="https://arxiv.org/abs/2505.17507">ArXiv-Link</a></td>
    </tr>
    <tr>
      <td>📊 <strong>Data</strong></td>
      <td><a href="https://huggingface.co/collections/cqsss/huggingbench-67b2ee02ca45b15e351009a2">
HuggingBench</a></td>
    </tr>
  </tbody>
</table>