six_four_eight

4 ms·

1. 1.▲
   by six_four_eight 1y ago

   my understanding is that yes it is, you cant make shoes in the US, but the power that comes from pretty much all finance flowing through the American pipes is a good trade off.
2. 2.▲
   by six_four_eight 2y ago

   I wonder how this compares to 'catastrophic forgetting' that can be a problem of full fine tuning. Or at least that's what I've just been reading as a case _for_ using LoRa, as it's not susceptible to that. I guess
3. 3.▲
   by six_four_eight 2y ago

   I wonder, is this exponential relation specific to multi-modal models? From my admittedly naïve view it seems to make sense that "...what is rare is not properly learned" would apply generally?
4. 4.▲
   by six_four_eight 2y ago

   Towards the end they state: ‘… just adding “do not hallucinate” has been shown to reduce the odds a model hallucinates.’ I find this surprising and doesn’t fit with my understanding of how a language model works. But I’m very much a novi
5. 5.▲
   by six_four_eight 3y ago

   Undoubtedly that can be the cause in some cases, but there are counterexamples. Like West Virginia which has a bad opioid addiction problem yet relatively low homeless rate. What does appear highly correlated is homeless rates vs. cost of
6. 6.▲
   by six_four_eight 3y ago

   Agree in that “Startups” imply scale and that’s just race to the bottom clickbait with maybe a little real news on the side. However, one example here in Canada is “Canadaland”, which is profitable, funded partly by subscribers, and has in