light_head_rcnn/lib/lib_kernel/lib_psalign_pooling/psalign_pooling.so: cannot open shared object file: No such file or directory

[hang0522]

No description provided.

[hang0522]

When run the command "python3 train.py -d 0-7"
it shows error as following:
Traceback (most recent call last):
File "train.py", line 12, in
import network_desp
File "/home/zzz/tf/light_head_rcnn/experiments/lizeming/light_head_rcnn.ori_res101.coco.ps_roialign/network_desp.py", line 30, in
from lib_kernel.lib_psalign_pooling import psalign_pooling_op, psalign_pooling_op_grad
File "/home/zzz/tf/light_head_rcnn/lib/lib_kernel/lib_psalign_pooling/psalign_pooling_op.py", line 5, in
_psalign_pooling_module = tf.load_op_library(filename)
File "/home/test/.local/lib/python3.6/site-packages/tensorflow/python/framework/load_library.py", line 56, in load_op_library
lib_handle = py_tf.TF_LoadLibrary(library_filename)
tensorflow.python.framework.errors_impl.NotFoundError: /home/zzz/tf/light_head_rcnn/lib/lib_kernel/lib_psalign_pooling/psalign_pooling.so: cannot open shared object file: No such file or directory

1. I have compiled lib folder successfully
2. I checked the folder "light_head_rcnn/lib/lib_kernel/lib_psalign_pooling/", and there is no file named "psaligh_pooling.so".
   I am confusing that is there anything wrong when I compiled the lib

[aggpankaj2]

@hang0522 "python3 train.py -d 0-7" 0-7 show your gpu id so if u are using one gpu you need to specify it by only 0
"python3 train.py -d 0"

[nithishc829]

I am having issue when loading the session from .meta file
`

Traceback (most recent call last):
File "freeze_graph.py", line 66, in
freeze_graph(args.model_dir, args.output_node_names)
File "freeze_graph.py", line 41, in freeze_graph
saver = tf.train.import_meta_graph(input_checkpoint + '.meta', clear_devices=clear_devices)
File "/home/nithish/my_install/miniconda3/envs/RFCN_tf/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 1435, in import_meta_graph
meta_graph_or_file, clear_devices, import_scope, **kwargs)[0]
File "/home/nithish/my_install/miniconda3/envs/RFCN_tf/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 1457, in _import_meta_graph_with_return_elements
**kwargs))
File "/home/nithish/my_install/miniconda3/envs/RFCN_tf/lib/python3.6/site-packages/tensorflow/python/framework/meta_graph.py", line 806, in import_scoped_meta_graph_with_return_elements
return_elements=return_elements)
File "/home/nithish/my_install/miniconda3/envs/RFCN_tf/lib/python3.6/site-packages/tensorflow/python/util/deprecation.py", line 507, in new_func
return func(*args, **kwargs)
File "/home/nithish/my_install/miniconda3/envs/RFCN_tf/lib/python3.6/site-packages/tensorflow/python/framework/importer.py", line 399, in import_graph_def
_RemoveDefaultAttrs(op_dict, producer_op_list, graph_def)
File "/home/nithish/my_install/miniconda3/envs/RFCN_tf/lib/python3.6/site-packages/tensorflow/python/framework/importer.py", line 159, in _RemoveDefaultAttrs
op_def = op_dict[node.op]
KeyError: 'PSROIPool' `

Since PSROIPool is implemented in .so as a layer is it possible to freeze_graph ?

[ccyydejqy]

vi mvpose/backend/light_head_rcnn/lib/lib_kernel/lib_psalign_pooling/make.sh

TF_INC=$(python3 -c 'import tensorflow as tf; print(tf.sysconfig.get_include())')
TF_LIB=$(python3 -c 'import tensorflow as tf; print(tf.sysconfig.get_lib())')
CUDA_PATH=/usr/local/cuda-9.0/ # jing
nvcc -std=c++11 -c -o psroi_pooling_op.cu.o psroi_pooling_op_gpu.cu.cc
-I /usr/local \ # jing
-I $TF_INC -D GOOGLE_CUDA=1 -x cu -Xcompiler -fPIC -arch=sm_52 -I$TF_INC/external/nsync/public --expt-relaxed-constexpr

g++ -std=c++11 -shared -D_GLIBCXX_USE_CXX11_ABI=0 -o psroi_pooling.so psroi_pooling_op.cc
psroi_pooling_op.cu.o -I $TF_INC -fPIC -lcudart -L $CUDA_PATH/lib64 -L$TF_LIB -ltensorflow_framework -I$TF_INC/external/nsync/public