- 0 Posts
- 5 Comments
danglybits27@sh.itjust.worksto
Technology@lemmy.zip•In rare disclosure, DeepSeek claims R1 model training cost just $294KEnglish
3·3 months agoMore clarification in this article:
https://www.theregister.com/2025/09/19/deepseek_cost_train/
"But, that’s not actually what happened. Never mind the fact that $300,000 won’t buy you anywhere close to 512 H800s (those estimates are based on GPU lease rates not actual hardware costs), the researchers aren’t talking about end-to-end model training.
Instead, it focuses on the application of reinforcement learning used to imbue its existing V3 base model with “reasoning” or “thinking” capabilities.
In other words, they’d already already done about 95 percent of the work by the time they’d reached the RL phase detailed in this paper."
danglybits27@sh.itjust.worksto
Television@piefed.social•What TV series has the best intro song or sequence? What TV series have the worst intro song or sequence?
51·3 months agoSomehow Voyager does have the best “feeling” intro of the Trek series.
I’ll counter that as much as I have learned to accept the reasonings for the theme for ‘Enterprise’, I still think it might be the worst theme of just about any show ever. And then they added the country twang around season 3…
danglybits27@sh.itjust.worksto
World News@quokk.au•‘Unsettling and unprecedented’: Washington DC mayor responds to Trump’s federal takeover of city police
46·4 months agoThe fuck is this headline? She, in fact, did say that combination of words in that order but, in context, what she actaully said was ‘"I can’t say that, given some of the rhetoric of the past, that we’re totally surprised,” she adds.’ Wild.

Seconded! It happens to me often using swipe controls and is incredibly frustrating.