Executor failed to create kernel error when running on raspberry

[FundamentX]

I deployed this solution over 30 times, with success, on my Raspberry Pi 3 over several months now. Just recently (this week) when i build and deployed the solution again the "ImageClassifierService" won't process any images anymore. Tested it with postman and it returns "Error processing image".

If i look at the logs of the module it say's:

/usr/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: compiletime version 3.4 of module 'tensorflow.python.framework.fast_tensor_util' does not match runtime version 3.5 return f(*args, **kwds)

/usr/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: builtins.type size changed, may indicate binary incompatibility. Expected 432, got 412 return f(*args, **kwds)

WARNING:tensorflow:From /app/predict.py:28: FastGFile.init (from tensorflow.python.platform.gfile) is deprecated and will be removed in a future version. Instructions for updating: Use tf.gfile.GFile.

Loading model...Success! Loading labels...2 found. Success!

• Serving Flask app "app" (lazy loading)
• Environment: production WARNING: Do not use the development server in a production environment. Use a production WSGI server instead.
• Debug mode: off
• Running on http://0.0.0.0:80/ (Press CTRL+C to quit)

2019-03-08 01:52:14.047760: E tensorflow/core/common_runtime/executor.cc:624] Executor failed to create kernel. Not found: No registered '_FusedConv2D' OpKernel for CPU devices compatible with node {{node conv1/conv1}} Registered: "no registered kernels"

Can anybody help me with this? I downloaded and rebuild the entire solution a couple of times with no success.

Is it maybe because tensorflow just released a new build and in arm32v7-requirements.txt there is no version specified for tensorflow? I'm guessing tensorflow 1.13.1 is giving problems and tensorflow 1.12.0 works. Not tested yet.

[FundamentX]

Update: If i specify tensorflow==1.12.0 in the ImageClassifierService module, it works.

[emmanuel-bv]

Thanks! I added your fix to the repo.