Skip to content
This repository has been archived by the owner on Mar 10, 2023. It is now read-only.

Attempting to use uninitialized value when calling grpc_mnist_client.py #8

Open
ChanningPing opened this issue Sep 5, 2018 · 0 comments

Comments

@ChanningPing
Copy link

Thank you so much for sharing this example. I followed the tutorial step by step and got this error when running the following:

TENSORFLOW_SERVER_HOST="0.0.0.0"
python python/grpc_mnist_client.py \
  --image ./data/0.png \
  --model mnist \
  --host $TENSORFLOW_SERVER_HOST

The error message is as follows:

Traceback (most recent call last):
  File "python/grpc_mnist_client.py", line 56, in <module>
    run(args.host, args.port, args.image, args.model, args.signature_name)
  File "python/grpc_mnist_client.py", line 34, in run
    result = stub.Predict(request, 10.0)
  File "/home/qping/miniconda3/envs/tensorflow-serving-example/lib/python2.7/site-packages/grpc/beta/_client_adaptations.py", line 309, in __call__
    self._request_serializer, self._response_deserializer)
  File "/home/qping/miniconda3/envs/tensorflow-serving-example/lib/python2.7/site-packages/grpc/beta/_client_adaptations.py", line 195, in _blocking_unary_unary
    raise _abortion_error(rpc_error_call)
grpc.framework.interfaces.face.face.AbortionError: AbortionError(code=StatusCode.FAILED_PRECONDITION, details="Attempting to use uninitialized value conv2d/bias
	 [[Node: conv2d/bias/read = Identity[T=DT_FLOAT, _output_shapes=[[32]], _device="/job:localhost/replica:0/task:0/device:CPU:0"](conv2d/bias)]]")

I have trained 10 steps, and copied the saved_model.pb into the models_for_serving/mnist_custom_estimator/1 path for serving. Could you help to shed some light on why this happens, or where could i have missed? Thank you so much for your help!

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant