HA! https://www.infoq.com/news/2020/10/training-exceeds-gpt3/
On Tue, Oct 6, 2020 at 10:54 AM <[email protected]> wrote: > Yes big net is embarrassing but but but if it scores better it scores > better ..... the small net may be small -- and therefore can store more > data in its pooch -- unlike the big model that only ate 1 cookie of data! > --- but the small well now big model still isn't as good as the original > big model, and both are full too of food now.................... a small > model is only good because it has room to get fat, it's a delicious honor > saying look at that free space, but it won't ever do if it doesn't find > patterns as well as the other model. Same for speed. > > Once we do get the seemingly best predictor, we can THEN start to say ok, > this one is faster and smaller, it's better. > *Artificial General Intelligence List <https://agi.topicbox.com/latest>* > / AGI / see discussions <https://agi.topicbox.com/groups/agi> + > participants <https://agi.topicbox.com/groups/agi/members> + delivery > options <https://agi.topicbox.com/groups/agi/subscription> Permalink > <https://agi.topicbox.com/groups/agi/T6761a13445e5864b-Mc0ce36c870083637d8689748> > ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T6761a13445e5864b-Mae7edce88afc1212b9a5b4d6 Delivery options: https://agi.topicbox.com/groups/agi/subscription
