formath / tensorflow-predictor-cpp Goto Github PK

I am going to try it on raspberry pi? Will it work?

please help me！

运行make命令时，报错：tensorflow/core/framework/variant_encode_decode.h: No such file or directory 。 项目的目录与tensorflow平级，进入tensorflow/core/framework/中也发现没有这个头文件，请问这个问题该如何处理

sh predict.sh

Session created successfully
Load graph protobuf successfully
Not found: Op type not registered 'HashTableV2' in binary running on MacBook-Pro.local. Make sure the Op and Kernel are registered in the binary running in this process. Note that if you are loading a saved graph which used ops from tf.contrib, accessing (e.g.) tf.contrib.resampler should be done before importing the graph, as contrib ops are lazily registered when the module is first accessed.

环境：
tensorflow 1.8
python 3.6

I can't find some header file with tf-1.3 ，so what is the version of tensorflow you use？

tensorflow-predictor-cpp/src/simple_model.cc这个文件中每次调用预测时都会初始化session，预测完成后close。如果作为线上服务，session只在最开始加载时初始化，预测的速度是否更快？

any idea?

tensorflow/tensorflow#13897

hello，I have forked your project and encounter the same problem

博主大大，我用c++加载图的时候遇到了这个问题：Not found: Op type not registered 'IteratorV2' in binary running on dsprecsitapp06. Make sure the Op and Kernel are registered in the binary running in this process. Note that if you are loading a saved graph which used ops from tf.contrib, accessing (e.g.) tf.contrib.resampler should be done before importing the graph, as contrib ops are lazily registered when the module is first accessed。

我用的是tensorflow1.14 estimator训练得到的模型，然后1.14编译的静态库加载的，在create graph那个地方报错出现这个问题。想请教您有遇到这个问题么，我该怎么去解决呀