Hacker Timesnew | past | comments | ask | show | jobs | submitlogin

> The context length for these models is 4096 tokens.

!!! And I was excited that llama gave us 2048!!



Rumor is RedPajama is going to have upwards of 60k token context by using Hyena: https://arxiv.org/abs/2302.10866

But it's just a rumor. We'll see.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: