LG

is this what you are looking for https://github.com/horovod/horovod/blob/master/horovod/common/operations.cc#L359-L378

Perhaps there exists some tools which can do the translation, need to find out.

https://github.com/lgarithm/crystalnet/commit/7b5806596b2c6db80e7abe1b48a8df50d0641e68#diff-4a7fc93aea4010278c6d52b293f3857c It can convert yolov2.weights into idx format (the format of MNIST)

yolov2 in C++ is almost done, the only thing left is remove overlaping detection boxes, which is relatively easy. how about the python one?

***draft*** ## The pooling operator The generic pooling operator with filter shape [r, s] and stride shape [stride_r, stride_s] (TODO: add padding shape) has the following signature: or pool :...

***draft*** ## Pointwise operators Pointwise operators are generic unary operators that has generic signature common pointwise operators are * relu * leaky relu

Other topics: * #35: Affine operator * #40: Support split and merge operators * #60: Operator fusion * #62: Variadic operator

``` [i] 1 34848 shape(11,11,3,96) [i] 2 96 shape(96) [i] 3 614400 shape(5,5,96,256) [i] 4 256 shape(256) [i] 5 884736 shape(3,3,256,384) [i] 6 384 shape(384) [i] 7 1327104 shape(3,3,384,384) [i]...

``` 1 conv1_W float32 34848 (11, 11, 3, 96) 2 conv1_b float32 96 (96,) 3 conv2_W float32 307200 (5, 5, 48, 256) 4 conv2_b float32 256 (256,) 5 conv3_W float32...

See the yolo-dev branch for implementation of route layer of darknet.