Categories
1 page
Transformer Architecture
Tensor-Parallel GQA Attention, One Head Group at a Time
