# Hugging Face — ~Don't~ Repeat Yourself

- Company: Hugging Face (huggingface.co)
- Announced: 2022-04-05
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://huggingface.co/blog/transformers-design-philosophy
- Record: https://forck.live/items/2216-don-t-repeat-yourself
- Subject: Platform
- Models affected: BERT, T5, Reformer, Longformer, BigBird, DeBERTa

Post explains the design philosophy behind Hugging Face Transformers library: the 'single model file' policy, which deliberately avoids refactoring common code into shared modules to make the code easier for contributors and readers to understand and modify, especially given fast ML research evolution.

## Evidence

Verbatim from https://huggingface.co/blog/transformers-design-philosophy:

> At first glance, the design of Hugging Face's Transformers library couldn't be more contrary to the DRY principle.

---

Record: https://forck.live/items/2216-don-t-repeat-yourself
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md
