Re:DeepNet: Scaling Transformers to 1k Layers
by notacanofsoda on Wednesday March 02, 2022 @06:28PM
(#30534943)
I love the tl;dr for practitioners part. All ML papers should have one.
[ Parent
| on HN ]