Diagonal Examples
These examples use the diagonal primitives (Diagonal Primitives (FHEON-D)): the channels are mixed with diagonals
instead of being folded one by one, so a convolution costs 8 hoisted rotations, one rotation per diagonal and a single
multiplicative level. They are single-input models, run sequentially, and are built with
-DMODE=DIAGONAL_SINGLE_INPUT (HEDiagonalSingleModels/).
Model |
Dataset |
Highlights |
|---|---|---|
|
MNIST |
Leveled context, no bootstrapping. |
|
MNIST |
Strided layout kept through the network, multiplexed conv2; 2.17 s per image. Leveled |
|
MNIST |
The circuit of |
|
CIFAR-10 |
Stride-2 blocks on shared taps; key file per stage or all keys in memory; 84 s per image. |
|
CIFAR-100 |
Four stages, key file per stage. |
|
CIFAR-10 |
Normalized activations within [0, 1]. |