steren’s Twitter Archive
—№ 7,172
⇤ Newest
Tweet
⇠ Newer
Tweet
Older
Tweet
⇢
…in reply to @adamse
adamse
A lot indeed. But my bet is that the energy used for training a popular production model is orders of magnitude smaller than the total energy needed to serve this model.
On twitter.com
↗
2022 Apr 24
Mood +
3
🙂