Edan Meyer

Question about model architecture

1

Hello, I was looking through the implementation, and I was curious about how you made choices regarding what to use for the architecture. Looking at the original paper [here](https://arxiv.org/pdf/2006.10204v1.pdf), I...

I tried to reproduce the results in the paper for the copy task with a sequence length of T=100, but I got quite different results. For SAB, I ran the...