AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
How to Scale Your Models
This is maybe the most impressive iteration of Transformers with RNN so far. It also makes me wonder, you know, looking at this 1.5 billion parameter scale that they're investigating, whether these results would actually scale. There's a lot of findings here that you won't go into as usual, but if you're curious, you can just go to the paper. After we're lining around, first, our WKV reinventing our NNs for very transformer era.