How to compile Tensorflow c++ in Jetson nano

Hi all,

I want to ask if any have finished running TensorFlow pb file in Jetson nano with TensorFlow c++ API.
Now the new Jetpack provides CUDA 10 and cudnn 7.5.0.

How to compile and build Tensorflow c++? Does anyone have experience? I have read some links below and trying to build on my own. However, until now I did not get success.

https://devtalk.nvidia.com/default/topic/1055131/jetson-agx-xavier/building-tensorflow-1-13-on-jetson-xavier/

https://devtalk.nvidia.com/default/topic/1058253/jetson-nano/request-for-prebuilt-tensorflow-c-c-api-libs-for-jetson-nano/

https://devtalk.nvidia.com/default/topic/1048711/jetson-nano/tensorflow-libs-for-c-c-integration-on-jetson-nano/

It will be very nice, if anyone can provide me any information or install guides.

Thanks in advance.

Hi,

Suppose you can compile a TensorFlow c++ package with the instruction shared in topic 155131:
https://devtalk.nvidia.com/default/topic/1055131/jetson-agx-xavier/building-tensorflow-1-13-on-jetson-xavier/

The only different is that the GPU capacity is different between Nano and Xavier.
The configure for Nano should look like this:

$./configure
...
Please specify the location where CUDA 10.0 toolkit is installed. Refer to README.md for more details. [Default is /usr/local/cuda]:/usr/local/cuda-10.0


Please specify the cuDNN version you want to use. [Leave empty to default to cuDNN 7]:<b>7.5</b>


Please specify a list of comma-separated Cuda compute capabilities you want to build with.
You can find the compute capability of your device at: https://developer.nvidia.com/cuda-gpus.
Please note that each additional compute capability significantly increases your build time and binary size. [Default is: 3.5,7.0]: <b>5.3</b>

Thanks.