snapo on Nostr: Omg https://arxiv.org/pdf/2304.11062.pdf this machine learning paper is absolutely ...
Omg
https://arxiv.org/pdf/2304.11062.pdf this machine learning paper is absolutely crazy. It provides a way to have a context length of 1M+ tokens. Just to put it into context, GPT-4 with 32k tokens looks child play against the 1M+ tokensize...
Published at
2023-04-25 14:15:09Event JSON
{
"id": "2fbf74b486e4f96a5cc7bce95bb97b647230430a8fb3c9a5ef53b5cc8bb6ae41",
"pubkey": "5b380071ce7f87ea7e5c8e7fa42b8f433ff83fc7b3f9940f505eafbcfdde4897",
"created_at": 1682432109,
"kind": 1,
"tags": [],
"content": "Omg https://arxiv.org/pdf/2304.11062.pdf this machine learning paper is absolutely crazy. It provides a way to have a context length of 1M+ tokens. Just to put it into context, GPT-4 with 32k tokens looks child play against the 1M+ tokensize...",
"sig": "a725c9c4b845af469827b8168f01e25a879df9035ec9bb5de19faa92b7fce38fab46f1cacb2d75146c3f4be7f5d22ee480037f250d445a13f8760715c2c5a336"
}
