---
task_categories:
- text-classification
language:
- en
tags:
- code
---

The dataset comes from [the original paper upload](https://console.cloud.google.com/storage/browser/gresearch/webllm/datasets/descgen) which was uploaded in a RecordIO format.

See the original paper [Understanding HTML with Large Language Models](https://arxiv.org/abs/2210.03945) for more details.