As I understand CNN is not connected to LSTM?
https://github.com/tobegit3hub/deep_cnn/blob/master/pokemon_classifer.py#L248

Also what is point of RNN usage in image classification task?

Hi,
I hit the following error when trying to run the java client for inception example. it's definitely contacting the other end but appears to have problem parsing the incoming jpg image stream. The server end is python running the inception model as per the Tensorflow tutorial @ https://www.tensorflow.org/serving/serving_inception.

Any help is appreciated! Error below:

Start the predict client
Aug 22, 2017 3:04:30 PM io.grpc.internal.ManagedChannelImpl
INFO: [ManagedChannelImpl@d041cf] Created with target 10.1.0.234:9000
Aug 22, 2017 3:04:30 PM com.tobe.InceptionPredictClient do_predict
INFO: Start to convert the image: /Users/joelgooch/Downloads/old_car.jpg
Aug 22, 2017 3:04:33 PM com.tobe.InceptionPredictClient do_predict
WARNING: RPC failed: Status{code=INVALID_ARGUMENT, description=Could not parse example input, value: '????, cause=null}
Aug 22, 2017 3:04:33 PM io.grpc.internal.ManagedChannelImpl maybeTerminateChannel
INFO: [ManagedChannelImpl@d041cf] Terminated
End of predict client

+ set -e
+ mvn clean install -DskipTests
[INFO] Scanning for projects...
[INFO] ------------------------------------------------------------------------
[INFO] Detecting the operating system and CPU architecture
[INFO] ------------------------------------------------------------------------
[INFO] os.detected.name: linux
[INFO] os.detected.arch: x86_64
[INFO] os.detected.release: ubuntu
[INFO] os.detected.release.version: 16.04
[INFO] os.detected.release.like.ubuntu: true
[INFO] os.detected.release.like.debian: true
[INFO] os.detected.classifier: linux-x86_64
[INFO]
[INFO] ------------------------------------------------------------------------
[INFO] Building predict 1.0-SNAPSHOT
[INFO] ------------------------------------------------------------------------
[WARNING] The POM for org.tensorflow:tensorflow-hadoop:jar:1.0-SNAPSHOT is missing, no dependency information available
[WARNING] The POM for com.xiaomi.infra.galaxy:galaxy-hadoop:jar:1.8-SNAPSHOT is missing, no dependency information available
[WARNING] The POM for org.apache.spark:spark-core_2.10:jar:1.6.1-mdh1.6.1.2 is missing, no dependency information available
[WARNING] The POM for org.apache.spark:spark-sql_2.10:jar:1.6.1-mdh1.6.1.2 is missing, no dependency information available
[INFO] ------------------------------------------------------------------------
[INFO] BUILD FAILURE
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 0.377 s
[INFO] Finished at: 2017-02-22T09:49:32+08:00
[INFO] Final Memory: 21M/964M
[INFO] ------------------------------------------------------------------------
[ERROR] Failed to execute goal on project predict: Could not resolve dependencies for project com.tobe:predict:jar:1.0-SNAPSHOT: The following artifacts could not be resolved: org.tensorflow:tensorflow-hadoop:jar:1.0-SNAPSHOT, com.xiaomi.infra.galaxy:galaxy-hadoop:jar:1.8-SNAPSHOT, org.apache.spark:spark-core_2.10:jar:1.6.1-mdh1.6.1.2, org.apache.spark:spark-sql_2.10:jar:1.6.1-mdh1.6.1.2: Could not find artifact org.tensorflow:tensorflow-hadoop:jar:1.0-SNAPSHOT -> [Help 1]
[ERROR]
[ERROR] To see the full stack trace of the errors, re-run Maven with the -e switch.
[ERROR] Re-run Maven using the -X switch to enable full debug logging.
[ERROR]
[ERROR] For more information about the errors and possible solutions, please read the following articles:
[ERROR] [Help 1] http://cwiki.apache.org/confluence/display/MAVEN/DependencyResolutionException

It seems that missing some dependencies?

hello ,I run tensorflow server in docker ,and it run successfully.
I tensorflow_serving/sources/storage_path/file_system_storage_path_source.cc:252] File-system polling update: Servable:{name: deep_cnn version: 1}; Servable path: ./model/00000001; Polling frequency: 30
But when I run java client (not in docker,in local),it has an error.
警告: RPC failed: Status{code=UNAVAILABLE, description=null, cause=java.net.ConnectException: Connection refused: /127.0.0.1:9000}
And I try to run python client in docker ,it runs successfully.
Does this have relationship with docker? Otherwise why cann't they connect.
thank you very much~~

1. predict_op definition missed in pokemonclassifier.py

there is only inference_predict_op for model_base64_input type

inference_logits = inference(model_base64_input)
inference_predict_softmax = tf.nn.softmax(inference_logits)
inference_predict_op = tf.argmax(inference_predict_softmax, 1)

missed predict_op for x type

predict_softmax = tf.nn.softmax(logit)
predict_op = tf.argmax(predict_softmax, 1)

2. typo error in README

in section Inference

./pokemon_classifer.py --mode inference --image ./data/inference/Pikachu.png

should be

./pokemon_classifier.py

hello,I provide a tensorflow serving ,and use java client to use it.how many clients can request this service at the same time ,Is there any quantity limitation for the this？thank you

when I Run the program pokemon,there are some questions as follows,but I do not know what is the meaning and how to modify,

Traceback (most recent call last):
File "E:/Python/TEST2.0/pokemon_classifer.py", line 391, in
main()
File "E:/Python/TEST2.0/pokemon_classifer.py", line 354, in main
tf.constant(FLAGS.export_version), sess)
File "D:\Program Files\Anaconda3\lib\site-packages\tensorflow\contrib\session_bundle\exporter.py", line 269, in export
gfile.MakeDirs(tmp_export_dir)
File "D:\Program Files\Anaconda3\lib\site-packages\tensorflow\python\lib\io\file_io.py", line 299, in recursive_create_dir
pywrap_tensorflow.RecursivelyCreateDir(compat.as_bytes(dirname), status)
File "D:\Program Files\Anaconda3\lib\contextlib.py", line 66, in exit
next(self.gen)
File "D:\Program Files\Anaconda3\lib\site-packages\tensorflow\python\framework\errors_impl.py", line 469, in raise_exception_on_not_ok_status
pywrap_tensorflow.TF_GetCode(status))
tensorflow.python.framework.errors_impl.NotFoundError: Failed to create a directory: ./model\00000001-tmp

Thank you!

Hello,I found a performance issue in the definition of quantize_weight_eightbit ,
java_predict_client/src/main/proto/tensorflow/tools/quantization/quantize_graph.py,
sess = tf.Session() was repeatedly called in for n in quantize_weight_eightbit and was not closed.
I think it will increase the efficiency and avoid out of memory if you close this session after using it.

Here are two files to support this issue,support1 and support2

Looking forward to your reply. Btw, I am very glad to create a PR to fix it if you are too busy.

hello ,when I Run TensorFlow serving,I found this
root@3a676eb847d1:/java/deep_cnn-master# bazel-bin/tensorflow_serving/model_servers/tensorflow_model_server --port=9000 --model_name=deep_cnn --model_base_path=./model
bash: bazel-bin/tensorflow_serving/model_servers/tensorflow_model_server: No such file or directory
how to get the ./tensorflow_model_server?Thank you ~~

hello~~
I need to use the 3 models to identify 3 kinds of pictures and I want to provide these 3 service in a container,How can I solve this problem?
I created a container like this " docker run -it -p 7770:7770 -p 7771:7771 -p 7772:7772 card /bin/bash"
provide 3 ports,and used ENTRYPOINT ["card"] to run the 3 commands like this"./tensorflow_model_server —port=7770 --model_name=simple1 --model_base_path=./simple1",But I saw it just run the first commands.

hello,when i run java client ,i found this
java.lang.NoClassDefFoundError: com/google/common/base/MoreObjects
at io.grpc.internal.AbstractManagedChannelImplBuilder.build(AbstractManagedChannelImplBuilder.java:257)
at io.grpc.internal.AbstractManagedChannelImplBuilder.build(AbstractManagedChannelImplBuilder.java:69)
and i did not find MoreObjects in guava14.0,so should i download guava 19.0?i downloaded a guava19.0,but somtimes it tried to find MoreObjects in 14.0.what should i do ?thank you ~~~