HTLM: Hyper-Text Pre-Training and Prompting of Language Models

TL;DR

HTLM is a novel hyper-text language model trained on large-scale web data, leveraging HTML structure for improved zero-shot and fine-tuned performance across various NLP tasks, especially in summarization and classification.

Contribution

This work introduces HTLM, a hyper-text language model trained on HTML data, utilizing structured prompts and HTML semantics to enhance transfer learning and zero-shot capabilities.

Findings

HTLM outperforms comparable text-only models on classification tasks.

HTLM achieves state-of-the-art results in zero-shot summarization.

Hyper-text prompts improve data efficiency over plain text prompts.

Abstract

We introduce HTLM, a hyper-text language model trained on a large-scale web crawl. Modeling hyper-text has a number of advantages: (1) it is easily gathered at scale, (2) it provides rich document-level and end-task-adjacent supervision (e.g. class and id attributes often encode document category information), and (3) it allows for new structured prompting that follows the established semantics of HTML (e.g. to do zero-shot summarization by infilling title tags for a webpage that contains the input text). We show that pretraining with a BART-style denoising loss directly on simplified HTML provides highly effective transfer for a wide range of end tasks and supervision levels. HTLM matches or exceeds the performance of comparably sized text-only LMs for zero-shot prompting and fine-tuning for classification benchmarks, while also setting new state-of-the-art performance levels for zero-shot summarization. We also find that hyper-text prompts provide more value to HTLM, in terms of data efficiency, than plain text prompts do for existing LMs, and that HTLM is highly effective at auto-prompting itself, by simply generating the most likely hyper-text formatting for any available training data. We will release all code and models to support future HTLM research.