Yhn first seen 5 h ago, last 10 min ago, peak #23
Janas runs LLMs larger than RAM by streaming experts from SSD
Original: Janas: Run LLMs larger than RAM by streaming experts from SSD
A new open-source project called Janas lets users run large language models that exceed available system memory by streaming mixture-of-experts weights directly from an SSD as needed. The approach targets hobbyists and researchers with consumer hardware who want to run big MoE models locally without expensive RAM upgrades, and it is drawing attention on developer forums.
Why now: Running huge MoE models on consumer hardware without large RAM is a compelling technical trick that developers want to try.
Janaslarge language modelsmixture-of-expertsSSD storage
Rank over time, top of the chart is #1. 24 snapshots from 5 h ago to 10 min ago.
Evidence
- Janas: Run LLMs larger than RAM by streaming experts from SSD · khalidkhair · 91
API: https://socialmediatrends-api.osmike.com/v1/trends/2042096