Build Tensorflow from source but it not support DT_INT64

[kongpon1995]

Describe the Issue

I implement my model on Keras Tensorflow backend and train it on my MacBookPro and then I copy model and weight file to execute on Raspberry pi and I got these error.

Steps to Reproduce

Hardware/Software Info

Please provide the following information about your Raspberry Pi setup:

• Raspberry Pi model: Pi3 model B
• Operating System used: Rasbian
• Version of Python used: Python 3.5.2
• SD card memory size: 32 GB
• Size of USB/other device used as swap (if building from source): 16 GB
• TensorFlow git commit hash (if building from source): I follow all of your step

Relevant Console Output/Logs

OpKernel was registered to support Op 'Assign' with these attrs. Registered devices: [CPU], Registered kernels: device='CPU'; T in [DT_FLOAT] device='CPU'; T in [DT_INT32]

[[Node: SGD/iterations/Assign = Assign[T=DT_INT64, _class=["loc:@SGD/iterations"], use_locking=true, validate_shape=true](SGD/iterations, SGD/iterations/initial_value)]]

I implement my model on Keras Tensorflow backend. I try to fix this problem, for 3 days Please help

Thank,