⬢github first seen 1 d ago, last 6 min ago, peak #7
AirLLM runs 2.8 trillion parameter models on a 4GB GPU
Original: lyogavin/airllm
Developer lyogavin's open-source project AirLLM is drawing attention on GitHub for enabling inference on LLMs as large as 2.8 trillion parameters with a single consumer GPU carrying just 4GB of memory. The project, written largely in Jupyter Notebooks, ranks among trending repositories globally, with developers discussing its approach to memory-efficient large model inference.
Why now: The project claims to dramatically lower hardware requirements for running extremely large language models, which excites the AI developer community.
Rank over time, top of the chart is #1. 48 snapshots from 11 h ago to 6 min ago.
Evidence
- lyogavin/airllm · lyogavin · 16
AirLLM 2.8T LLM inference with single 4GB GPU
API: https://socialmediatrends-api.osmike.com/v1/trends/1850728