« 6TH POPE Home | Email msg. | Reply to msg. | Post new | Board info. Previous | Home | Next

You can now run a GPT-3-level AI model on your laptop, phone, and Raspberry Pi Thanks to Meta LLaMA 

By: Fiz in 6TH POPE | Recommend this post (1)
Mon, 05 Jun 23 2:54 AM | 41 view(s)
Boardmark this board | 6th Edition Pope Board
Msg. 43154 of 60008
(This msg. is a reply to 43142 by Decomposed)

Jump:
Jump to board:
Jump to msg. #

http://arstechnica.com/information-technology/2023/03/you-can-now-run-a-gpt-3-level-ai-model-on-your-laptop-phone-and-raspberry-pi/

LLaMA made a heady claim: that its smaller-sized models could match OpenAI's GPT-3, the foundational model that powers ChatGPT, in the quality and speed of its output. There was just one problem—Meta released the LLaMA code open source, but it held back the "weights" (the trained "knowledge" stored in a neural network) for qualified researchers only.
Flying at the speed of LLaMA

Meta's restrictions on LLaMA didn't last long, because on March 2, someone leaked the LLaMA weights on BitTorrent.

De, you seemed skeptical; just wanted to link this so you can let go of that and get into the spirit of the hack!


- - - - -
View Replies (1) »



» You can also:
- - - - -
The above is a reply to the following message:
Re: google sez: We have no moat, and neither does OpenAI. Open source has taken their LLM code and tuned it to rapidly compete
By: Decomposed
in 6TH POPE
Sun, 04 Jun 23 10:19 PM
Msg. 43142 of 60008

fizzy:

Re: “I'm trying to find an open code base, so I might begin playing with it myself. Or, at least, use it as a learning exercise”
Let us know how it goes. My understanding is that the AI programs Google and Micrososft are touting require MASSIVE computing power - the equivalent of mainframes, though they may just involve a few thousand networked desktops. You don't happen to have anything like that sitting around in your basement, do you?





« 6TH POPE Home | Email msg. | Reply to msg. | Post new | Board info. Previous | Home | Next