Is there any error in the illustrations of the "Interleaving Stages" ？
#123
by
railgun10032
- opened
In the illustrations of the "Interleaving Stages" chapter, why isn't the execution time of the backward pass (last layers) placed before that of the backward pass (first layers)? I'm a bit confused about this. Additionally, do the "first layers" in the forward pass and the backward pass refer to the same layers in the network?