---
language: en
license: apache-2.0
viewer: false
---

# Dataset card for Wikipedia Domain Labels

This dataset is a list of domain labels for the [txtai-wikipedia-slim](https://huggingface.co/NeuML/txtai-wikipedia-slim) embeddings database, which is the Top 100K most viewed Wikipedia articles.

## Dataset Code

This dataset was built using a [zero-shot-classification model](https://huggingface.co/MoritzLaurer/deberta-v3-large-zeroshot-v2.0-c) applied to each of the Top 100K most viewed articles. The labels are based on the labels from [this dataset](https://huggingface.co/datasets/m-a-p/FineFineWeb) with a few minor modifications.

[The code used to build this model is here](https://huggingface.co/datasets/NeuML/wikipedia-domain-labels/blob/main/labels.py).