[prev in list] [next in list] [prev in thread] [next in thread]
List: kde-devel
Subject: Re: Interest in building an LLM frontend for KDE
From: Alexander Semke <alexander.semke () web ! de>
Date: 2023-12-04 21:45:15
Message-ID: 2914505.e9J7NaK4W3 () linux ! fritz ! box
[Download RAW message or body]
On Montag, 4. Dezember 2023 12:09:43 CET Joseph P. De Veaugh-Geiss wrote:
> I agree with the concerns Josh raises about the energy consumption of
> training LLMs (see, e.g., [1]). A benefit of satisfying the above
> characteristics is it is then possible for us to measure the energy
> consumption for training/using the LLMs. This would enable KDE to be
> transparent about what these tools consume in terms of energy and
> present this information to users.
To make this argument more complete, it's not only the training of such mo=
dels but also
their usage ("inference") later. For popular generic models the negative i=
mpact can quickly
become bigger than the impact of the training itself:
https://www.technologyreview.com/2023/12/01/1084189/making-an-image-with-g=
enerative-ai-uses-as-much-energy-as-charging-your-phone/[1]
https://arxiv.org/abs/2311.16863[2]
Though, these and similar arguments mostly tend to ignore the fact that th=
e hyperscalers
have commited to the net-zero imitative and are heavily investing into ren=
ewables for
their data centers and also trying to shift heavy workloads into more sust=
ainable time
windows.
=2D-
Alexander
=2D-------
[1] https://www.technologyreview.com/2023/12/01/1084189/making-an-image-wi=
th-generative-ai-uses-as-much-energy-as-charging-your-phone/
[2] https://arxiv.org/abs/2311.16863
[Attachment #3 (unknown)]
<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
</head>
<body><p style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">On \
Montag, 4. Dezember 2023 12:09:43 CET Joseph P. De Veaugh-Geiss wrote:</p> <p \
style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">> I agree with \
the concerns Josh raises about the energy consumption of</p> <p \
style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">> training LLMs \
(see, e.g., [1]). A benefit of satisfying the above</p> <p \
style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">> \
characteristics is it is then possible for us to measure the energy</p> <p \
style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">> consumption \
for training/using the LLMs. This would enable KDE to be</p> <p \
style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">> transparent \
about what these tools consume in terms of energy and</p> <p \
style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">> present this \
information to users.</p> <p \
style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">To make this \
argument more complete, it's not only the training of such models but also their \
usage ("inference") later. For popular generic models the negative impact \
can quickly become bigger than the impact of the training itself:</p> <p \
style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;"><a \
href="https://www.technologyreview.com/2023/12/01/1084189/making-an-image-with-generat \
ive-ai-uses-as-much-energy-as-charging-your-phone/">https://www.technologyreview.com/2 \
023/12/01/1084189/making-an-image-with-generative-ai-uses-as-much-energy-as-charging-your-phone/</a></p>
<br /><p style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;"><a \
href="https://arxiv.org/abs/2311.16863">https://arxiv.org/abs/2311.16863</a></p> <br \
/><p style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">Though, these \
and similar arguments mostly tend to ignore the fact that the hyperscalers have \
commited to the net-zero imitative and are heavily investing into renewables for \
their data centers and also trying to shift heavy workloads into more sustainable \
time windows.</p> <br /><br /><p \
style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">-- </p> <p \
style="margin-top:0;margin-bottom:0;margin-left:0;margin-right:0;">Alexander</p> \
</body> </html>
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic