Run LLama 3 70B on a Single 4GB GPU
by LkStr - Sunday April 28, 2024 at 10:50 AM
#1
Run LLama 3 70B on a Single 4GB GPU - with airllm and layered inference : https://twitter.com/rohanpaul_ai/status/...7899982943
Has anyone tried this approach? Seems like a good idea, I will give it a try Smile
Reply
#2
Why not just use a Sagemaker account? It is far better than colab (which has honestly become a real big piece of shit). Rather than so many mental gymnastics ?
Reply


Possibly Related Threads…
Thread Author Replies Views Last Post
  [GUIDE] How Open and view .Sql and JSON Data Lavi 33 7,623 Feb 09, 2026, 09:36 PM
Last Post: gopoo2
  UpViral – Viral Hacking Masterclass Download spanko73 0 114 Jan 26, 2026, 12:07 AM
Last Post: spanko73
  LINUX COMMAND secretleakbuyer 3 1,708 Dec 30, 2025, 11:46 AM
Last Post: 888
  Best AI Checkpoints for Photo realism. DredgenSun 12 2,728 Apr 14, 2025, 09:48 AM
Last Post: DredgenSun
  Mix Courses Collection | 100+ GB moron 15 3,280 Apr 13, 2025, 09:34 PM
Last Post: wared

Forum Jump:


 Users browsing this forum: 1 Guest(s)