this post was submitted on 23 Nov 2024
556 points (95.9% liked)

Technology

60062 readers
3387 users here now

This is a most excellent place for technology news and articles.

Our Rules

  1. Follow the lemmy.world rules.
  2. Only tech related content.
  3. Be excellent to each another!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, to ask if your bot can be added please contact us.
  9. Check for duplicates before posting, duplicates may be removed

Approved Bots

founded 2 years ago
MODERATORS
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
556
‘Thirsty’ ChatGPT uses four times more water than previously thought (www.thetimes.com)
submitted 1 month ago by [email protected] to c/[email protected]
96 comments fedilink hide all child comments
 

Writing a 100-word email using ChatGPT (GPT-4, latest model) consumes 1 x 500ml bottle of water It uses 140Wh of energy, enough for 7 full charges of an iPhone Pro Max

you are viewing a single comment's thread
view the rest of the comments
[–] [email protected] 8 points 4 weeks ago* (last edited 4 weeks ago) (1 children)

I’ve run an LLM on my desktop GPU and gotten decent results, albeit not nearly as good as what ChatGPT will get you.

Probably used less than 0.1Wh per response.

[–] [email protected] 1 points 4 weeks ago (2 children)

Is this for inferencing only? Do you include training?

[–] [email protected] 2 points 4 weeks ago

Inference only. I’m looking into doing some fine tuning. Training from scratch is another story.

[–] [email protected] 2 points 3 weeks ago (1 children)

Training is a one time thing. Tge more it get use, the less energy per query it will take

[–] [email protected] 1 points 3 weeks ago (1 children)

Good point. But considering the frequent retraining, the environmental impacts can only be spread on a finite number of queries.

[–] [email protected] 1 points 3 weeks ago

They have already reached diminishing returns on training. It will become much less frequent soon. Retraining on the same data if there isn't a better method is useless. I think the ressources consumed per query should only include those actually used for inference. The rest can be dismissed as bad faith argumentation.