# Hugging Face — Fine-Tune MMS Adapter Models for low-resource ASR

- Company: Hugging Face (huggingface.co)
- Announced: 2023-06-19
- Category: research-paper
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://huggingface.co/blog/mms_adapters
- Record: https://forck.live/items/2038-fine-tune-mms-adapter-models-for-low-resource-asr
- Subject: Platform
- Models affected: mms-300m, mms-1b, MMS-1B, mms-1b-fl102, mms-1b-l1107, mms-1b-all

Meta AI released Massive Multilingual Speech (MMS) with adapter layers for low-resource ASR, fine-tuning only ~2.5M weights per language on checkpoints pre-trained on over half a million hours of audio in 1,400+ languages.

## Evidence

Verbatim from https://huggingface.co/blog/mms_adapters:

> Meta AI's most recent release, Massive Multilingual Speech (MMS) by Vineel Pratap, Andros Tjandra, Bowen Shi, et al. takes multi-lingual speech representations to a new level.

---

Record: https://forck.live/items/2038-fine-tune-mms-adapter-models-for-low-resource-asr
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md