Hacker News new | ask | show | jobs
by numlocked 1183 days ago
Interesting. I wonder what the training cost was for:

https://huggingface.co/EleutherAI/gpt-neox-20b

Perhaps it’s in the paper…
1 comments
michaelhartm 1183 days ago
They used the 6b GPT4-J, not 20B. That's what's interesting, it's a smallish large language model :).
link
dragonwriter 1183 days ago
GPT-J, not GPT4-J.
link