hikapok / tf.extra_losses Goto Github PK

I have Ubuntu 16.04, Python 3.5.2 in a virtual env, and Tensorflow 1.6 GPU.
I have 2 1080ti.
I have run the "cd build && cmake .." and the output is:

`-- The C compiler identification is GNU 5.4.0
-- The CXX compiler identification is GNU 5.4.0
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Looking for pthread_create
-- Looking for pthread_create - not found
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE
-- Found CUDA: /usr/local/cuda-9.0 (found version "9.0")
-- Found CWD: /home/vision/matthew/tf.extra_losses/build
-- Found GPU_CAPABILITY: gencode arch=compute_61,code=sm_61

/home/vision/tf_16/lib/python3.5/site-packages/h5py/init.py:36: FutureWarning: Conversion of the second argument of issubdtype from float to np.floating is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type.
from ._conv import register_converters as _register_converters
/home/vision/tf_16/lib/python3.5/site-packages/h5py/init.py:36: FutureWarning: Conversion of the second argument of issubdtype from float to np.floating is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type.
from ._conv import register_converters as _register_converters
-- Found TF_INC: /home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include
-- Found TF_INC_EXTERNAL: /home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/external/nsync/public
-- Found TF_LIB: /home/vision/tf_16/lib/python3.5/site-packages/tensorflow
-- Configuring done
-- Generating done
-- Build files have been written to: /home/vision/matthew/tf.extra_losses/build
`

And Also 'make' command:

`[ 16%] Building NVCC (Device) object CMakeFiles/cuda_compile.dir/cuda_compile_generated_l_softmax_grad_op.cu.o
/home/vision/matthew/tf.extra_losses/l_softmax_grad_op.cu:181:1: warning: multi-line comment [-Wcomment]
// #define DEFINE_GPU_SPECS(T)
^
/home/vision/matthew/tf.extra_losses/l_softmax_grad_op.cu:181:1: warning: multi-line comment [-Wcomment]
// #define DEFINE_GPU_SPECS(T)
^
/home/vision/matthew/tf.extra_losses/l_softmax_grad_op.cu:181:1: warning: multi-line comment [-Wcomment]
// #define DEFINE_GPU_SPECS(T)
^
/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/arena_impl.h(57): warning: integer conversion resulted in a change of sign

/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/arena_impl.h(304): warning: integer conversion resulted in a change of sign

/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/arena_impl.h(305): warning: integer conversion resulted in a change of sign

/home/vision/matthew/tf.extra_losses/l_softmax_grad_op.cu:181:1: warning: multi-line comment [-Wcomment]
// #define DEFINE_GPU_SPECS(T)
^
/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/arena_impl.h(57): warning: integer conversion resulted in a change of sign

/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/arena_impl.h(304): warning: integer conversion resulted in a change of sign

/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/arena_impl.h(305): warning: integer conversion resulted in a change of sign

/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/generated_message_reflection.h(685): warning: variable "unused" was set but never used

[ 33%] Building NVCC (Device) object CMakeFiles/cuda_compile.dir/cuda_compile_generated_l_softmax_op.cu.o
/home/vision/matthew/tf.extra_losses/l_softmax_op.cu:131:1: warning: multi-line comment [-Wcomment]
// #define DEFINE_GPU_SPECS(T)
^
/home/vision/matthew/tf.extra_losses/l_softmax_op.cu:131:1: warning: multi-line comment [-Wcomment]
// #define DEFINE_GPU_SPECS(T)
^
/home/vision/matthew/tf.extra_losses/l_softmax_op.cu:131:1: warning: multi-line comment [-Wcomment]
// #define DEFINE_GPU_SPECS(T)
^
/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/arena_impl.h(57): warning: integer conversion resulted in a change of sign

/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/arena_impl.h(304): warning: integer conversion resulted in a change of sign

/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/arena_impl.h(305): warning: integer conversion resulted in a change of sign

/home/vision/matthew/tf.extra_losses/common.h(32): warning: the "always_inline" attribute is ignored on non-inline functions

/home/vision/matthew/tf.extra_losses/common.h(46): warning: the "always_inline" attribute is ignored on non-inline functions

/home/vision/matthew/tf.extra_losses/common.h(52): warning: the "always_inline" attribute is ignored on non-inline functions

/home/vision/matthew/tf.extra_losses/l_softmax_op.cu:131:1: warning: multi-line comment [-Wcomment]
// #define DEFINE_GPU_SPECS(T)
^
/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/arena_impl.h(57): warning: integer conversion resulted in a change of sign

/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/arena_impl.h(304): warning: integer conversion resulted in a change of sign

/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/arena_impl.h(305): warning: integer conversion resulted in a change of sign

/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/include/google/protobuf/generated_message_reflection.h(685): warning: variable "unused" was set but never used

/home/vision/matthew/tf.extra_losses/common.h(32): warning: the "always_inline" attribute is ignored on non-inline functions

/home/vision/matthew/tf.extra_losses/common.h(46): warning: the "always_inline" attribute is ignored on non-inline functions

/home/vision/matthew/tf.extra_losses/common.h(52): warning: the "always_inline" attribute is ignored on non-inline functions

Scanning dependencies of target extra_losses
[ 50%] Building CXX object CMakeFiles/extra_losses.dir/l_softmax_op.cc.o
/home/vision/matthew/tf.extra_losses/l_softmax_op.cc:274:0: warning: "REGISTER_CPU" redefined
#define REGISTER_CPU(T)
^
/home/vision/matthew/tf.extra_losses/l_softmax_op.cc:268:0: note: this is the location of the previous definition
#define REGISTER_CPU(T)
^
/home/vision/matthew/tf.extra_losses/l_softmax_op.cc:291:0: warning: "REGISTER_GPU" redefined
#define REGISTER_GPU(T)
^
/home/vision/matthew/tf.extra_losses/l_softmax_op.cc:285:0: note: this is the location of the previous definition
#define REGISTER_GPU(T)
^
/home/vision/matthew/tf.extra_losses/l_softmax_op.cc: In lambda function:
/home/vision/matthew/tf.extra_losses/l_softmax_op.cc:62:40: warning: variable ‘num_dimensions’ set but not used [-Wunused-but-set-variable]
shape_inference::DimensionHandle num_dimensions = c->Dim(features_shape, 1);
^
/home/vision/matthew/tf.extra_losses/l_softmax_op.cc: In lambda function:
/home/vision/matthew/tf.extra_losses/l_softmax_op.cc:93:40: warning: variable ‘num_dimensions’ set but not used [-Wunused-but-set-variable]
shape_inference::DimensionHandle num_dimensions = c->Dim(features_shape, 1);
^
[ 66%] Building CXX object CMakeFiles/extra_losses.dir/l_softmax_grad_op.cc.o
/home/vision/matthew/tf.extra_losses/l_softmax_grad_op.cc:293:0: warning: "REGISTER_CPU" redefined
#define REGISTER_CPU(T)
^
/home/vision/matthew/tf.extra_losses/l_softmax_grad_op.cc:287:0: note: this is the location of the previous definition
#define REGISTER_CPU(T)
^
/home/vision/matthew/tf.extra_losses/l_softmax_grad_op.cc:310:0: warning: "REGISTER_GPU" redefined
#define REGISTER_GPU(T)
^
/home/vision/matthew/tf.extra_losses/l_softmax_grad_op.cc:304:0: note: this is the location of the previous definition
#define REGISTER_GPU(T)
^
[ 83%] Building CXX object CMakeFiles/extra_losses.dir/common.cc.o
In file included from /home/vision/matthew/tf.extra_losses/common.cc:22:0:
/home/vision/matthew/tf.extra_losses/common.h:46:37: warning: always_inline function might not be inlinable [-Wattributes]
attribute((always_inline)) bool is_pow2(const T x)
^
/home/vision/matthew/tf.extra_losses/common.h:32:39: warning: always_inline function might not be inlinable [-Wattributes]
attribute((always_inline)) Target binary_cast(Source s)
^
/home/vision/matthew/tf.extra_losses/common.h:32:39: warning: always_inline function might not be inlinable [-Wattributes]
/home/vision/matthew/tf.extra_losses/common.h:52:37: warning: always_inline function might not be inlinable [-Wattributes]
attribute((always_inline)) bool is_aligned(const T ptr, const size_t alignment)
^
[100%] Linking CXX shared library libextra_losses.so
[100%] Built target extra_losses
`

but unfortunately got an error in test_op.py execution:

/home/vision/tf_16/lib/python3.5/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from floattonp.floatingis deprecated. In future, it will be treated asnp.float64 == np.dtype(float).type. from ._conv import register_converters as _register_converters Traceback (most recent call last): File "test_op.py", line 36, in <module> op_module = load_op_module(LIB_NAME) File "test_op.py", line 33, in load_op_module oplib = tf.load_op_library(lib_path) File "/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/python/framework/load_library.py", line 58, in load_op_library lib_handle = py_tf.TF_LoadLibrary(library_filename, status) File "/home/vision/tf_16/lib/python3.5/site-packages/tensorflow/python/framework/errors_impl.py", line 516, in __exit__ c_api.TF_GetCode(self.status.status)) tensorflow.python.framework.errors_impl.NotFoundError: /home/vision/matthew/tf.extra_losses/build/libextra_losses.so: undefined symbol: _ZNK10tensorflow8OpKernel11type_stringEv

What should do?!

When I run 'make' in build directory , there some errors:

[100%] Linking CXX shared library libextra_losses.so
/usr/bin/ld: 找不到 -ltensorflow_framework
collect2: error: ld returned 1 exit status
CMakeFiles/extra_losses.dir/build.make:166: recipe for target 'libextra_losses.so' failed
make[2]: *** [libextra_losses.so] Error 1
CMakeFiles/Makefile2:67: recipe for target 'CMakeFiles/extra_losses.dir/all' failed
make[1]: *** [CMakeFiles/extra_losses.dir/all] Error 2
Makefile:83: recipe for target 'all' failed
make: *** [all] Error 2

do you know the reason? thank you for your help.

ubuntu16.04+python3.6(anaconda3)+tensorflow1.6
thank you very much.

Now, I try to install in virtual-env.
but I couldn't find the way for install.
could you help me?

Hi

I am using Keras api for model training and testing. May I know how to use the loss function in this project together with Keras api?

Thank you!