Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
loading thread…
3 ms
·
20B parameter GPT-3 model trained on a single Cerebras chip
2 points
by
keveman
4y ago