Cannot run BVLC AlexNet!

[pmousoul]

Hello,

I can confirm that I can build CHaiDNN and execute the build-in examples.

BUT

When I try to execute the BVLC AlexNet I get an error:

[EP030] Layer conv1 is not found in the caffemodel file.

I'm using the files from:
https://github.com/BVLC/caffe/tree/master/models/bvlc_alexnet

I'm successfully quantizing at 6 bits using the following command:

python XportDNN.pyc \
--quant_type "Xilinx" --deploy_model /mnt/terabyte/pmousoul_data/Repos/CHaiDNN/tools/tools/deploy.prototxt \
--weights /mnt/terabyte/pmousoul_data/Repos/CHaiDNN/tools/tools/AlexNetNoLRN_6Bit_CHaiDNN.caffemodel \
--quantized_deploy_model /mnt/terabyte/pmousoul_data/Repos/CHaiDNN/tools/tools/AlexNet_6Bit_deploy_CHaiDNN.prototxt \
--calibration_directory /mnt/terabyte/datasets/ImageNet/val/ \
--calibration_size 32 \
--bitwidths 6,6,6 \
--dims 3,227,227 \
--transpose 2,0,1 \
--channel_swap 2,1,0 \
--raw_scale 255.0

Any idea of what is going wrong??

CHaiDNN is not useful at all if it cannot execute networks other than the examples..

Cheers,
Panos