mrhosseini

> Are you sure the error is in the reorg? > If you end the network right before the reorg the result is correct? And right after? > Are the...

I found the problem. Input arguments of the [`reorg_kernel()`](https://github.com/ceccocats/tkDNN/blob/1dfc69ba8916f913282c1fb68a117cdc72468f78/src/kernels/reorg.cu#L3) are output dimensions of the result. Therefore the arguments of [`reorgForward()`](https://github.com/ceccocats/tkDNN/blob/1dfc69ba8916f913282c1fb68a117cdc72468f78/src/Reorg.cpp#L28) are incorrect and should be changed like this: ```C reorgForward(srcData,...

You are right. I didn't noticed this. Actually my configuration contains a `reorg3d` layer. I added `reorg3d` just as an alias for the `reorg` layer. Therefore the problem occurred. My...

Ok. I did the fix.

After more debugging I found that, order of yolo layers in `pluginFactory` of `NetworkRT` after desrialization ([here](https://github.com/ceccocats/tkDNN/blob/6d9beb1ec5da217d8c6b1b2b51bfabf6766c3f95/src/NetworkRT.cpp#L629)), is different from what we have in `NetworkRT::buffersRT` which is based on `engineRT->getBindingDimensions()`...

Changing [this line](https://github.com/ceccocats/tkDNN/blob/6d9beb1ec5da217d8c6b1b2b51bfabf6766c3f95/src/Yolo3Detection.cpp#L95) to the following will solve the problem. But I am not sure if this works for all the cases. ```C rt_out[i] = (dnnType *) netRT->buffersRT[netRT->pluginFactory->n_yolos - i]...

> The order of buffersRt should always be: > Input > Output 0 > Output 1 > Etc. What happens is that order of yolo layers in `NetworkRT::pluginFactory` is different...

This pull request now contains the fix for #49.

I have merged this repository with Maratyszcza/NNPACK, [here](https://github.com/mrhosseini/NNPACK-darknet).

check https://github.com/digitalbrain79/NNPACK-darknet/issues/3#issuecomment-849705541