Skip to content
This repository has been archived by the owner on Nov 17, 2023. It is now read-only.

example/ssd error:Operator Scale is not registered #4720

Closed
guo253 opened this issue Jan 18, 2017 · 15 comments
Closed

example/ssd error:Operator Scale is not registered #4720

guo253 opened this issue Jan 18, 2017 · 15 comments

Comments

@guo253
Copy link

guo253 commented Jan 18, 2017 •
edited
Loading

@zhreshold
For bugs or installation issues, please provide the following information.
The more information you provide, the more likely people will be able to help you.

Environment info

Operating System:
ubuntu 14.04
Compiler:
g++ 4.8.4
gcc 4.8.4
Package used (Python/R/Scala/Julia):
Python
MXNet version:
0.9.2 download 2017.1.17
CUDA = 0(only for cpu)
Python version and distribution:
2.7

EXTRA_OPERATORS = example/ssd/operator

Error Message:

python demo.py --epoch 0 --images ./data/demo/dog.jpg --thresh 0.5 --cpu
[22:24:20] include/dmlc/./logging.h:300: [22:24:20] src/core/op.cc:55: Check failed: op != nullptr Operator Scale is not registered

Stack trace returned 34 entries:
[bt] (0) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(_ZN4nnvm2Op3GetERKSs+0x376) [0x7fd1584c2c06]
[bt] (1) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(+0x1859608) [0x7fd15850a608]
[bt] (2) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(_ZN4dmlc20JSONObjectReadHelper13ReadAllFieldsEPNS_10JSONReaderE+0x110) [0x7fd1585111c0]
[bt] (3) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(+0x1857881) [0x7fd158508881]
[bt] (4) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(ZNSt17_Function_handlerIFN4nnvm5GraphES1_EPS2_E9_M_invokeERKSt9_Any_dataS1+0x11f) [0x7fd157884dbf]
[bt] (5) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(_ZN4nnvm11ApplyPassesENS_5GraphERKSt6vectorISsSaISsEE+0x703) [0x7fd1584c8fc3]
[bt] (6) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(_ZN5mxnet18LoadLegacyJSONPassEN4nnvm5GraphE+0x18b) [0x7fd15788118b]
[bt] (7) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(ZNSt17_Function_handlerIFN4nnvm5GraphES1_EPS2_E9_M_invokeERKSt9_Any_dataS1+0x11f) [0x7fd157884dbf]
[bt] (8) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(_ZN4nnvm11ApplyPassesENS_5GraphERKSt6vectorISsSaISsEE+0x703) [0x7fd1584c8fc3]
[bt] (9) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(_ZN4nnvm9ApplyPassENS_5GraphERKSs+0x8e) [0x7fd15778e01e]
[bt] (10) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(MXSymbolCreateFromFile+0x471) [0x7fd1577873c1]
[bt] (11) /usr/lib/x86_64-linux-gnu/libffi.so.6(ffi_call_unix64+0x4c) [0x7fd15c558adc]
[bt] (12) /usr/lib/x86_64-linux-gnu/libffi.so.6(ffi_call+0x1fc) [0x7fd15c55840c]
[bt] (13) /usr/lib/python2.7/lib-dynload/_ctypes.x86_64-linux-gnu.so(_ctypes_callproc+0x48e) [0x7fd15c76f5fe]
[bt] (14) /usr/lib/python2.7/lib-dynload/_ctypes.x86_64-linux-gnu.so(+0x15f9e) [0x7fd15c770f9e]
[bt] (15) python(PyEval_EvalFrameEx+0x965) [0x499be5]
[bt] (16) python(PyEval_EvalCodeEx+0x2ac) [0x4a090c]
[bt] (17) python(PyEval_EvalFrameEx+0x18c5) [0x49ab45]
[bt] (18) python(PyEval_EvalCodeEx+0x2ac) [0x4a090c]
[bt] (19) python(PyEval_EvalFrameEx+0x18c5) [0x49ab45]
[bt] (20) python() [0x4a1c9a]
[bt] (21) python() [0x4dfe94]
[bt] (22) python(PyObject_Call+0x36) [0x505f96]
[bt] (23) python() [0x4de41a]
[bt] (24) python() [0x5039eb]
[bt] (25) python(PyEval_EvalFrameEx+0x965) [0x499be5]
[bt] (26) python(PyEval_EvalCodeEx+0x2ac) [0x4a090c]
[bt] (27) python(PyEval_EvalFrameEx+0x7d2) [0x499a52]
[bt] (28) python() [0x4a1634]
[bt] (29) python(PyRun_FileExFlags+0x92) [0x44e4a5]
[bt] (30) python(PyRun_SimpleFileExFlags+0x2ee) [0x44ec9f]
[bt] (31) python(Py_Main+0xb5e) [0x44f904]
[bt] (32) /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf5) [0x7fd15da25ec5]
[bt] (33) python() [0x578c4e]

Traceback (most recent call last):
File "demo.py", line 95, in
ctx, args.nms_thresh, args.force_nms)
File "demo.py", line 41, in get_detector
data_shape, mean_pixels, ctx=ctx)
File "/home/guorui/soft/code/mxnet/mxnet/example/ssd/detect/detector.py", line 34, in init
_, args, auxs = mx.model.load_checkpoint(model_prefix, epoch)
File "/usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/model.py", line 372, in load_checkpoint
symbol = sym.load('%s-symbol.json' % prefix)
File "/usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/symbol.py", line 1056, in load
check_call(_LIB.MXSymbolCreateFromFile(c_str(fname), ctypes.byref(handle)))
File "/usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/base.py", line 75, in check_call
raise MXNetError(py_str(_LIB.MXGetLastError()))
mxnet.base.MXNetError: Failed loading Op relu4_3_scale_20 of type Scale: [22:24:20] src/core/op.cc:55: Check failed: op != nullptr Operator Scale is not registered

Stack trace returned 34 entries:
[bt] (0) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(_ZN4nnvm2Op3GetERKSs+0x376) [0x7fd1584c2c06]
[bt] (1) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(+0x1859608) [0x7fd15850a608]
[bt] (2) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(_ZN4dmlc20JSONObjectReadHelper13ReadAllFieldsEPNS_10JSONReaderE+0x110) [0x7fd1585111c0]
[bt] (3) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(+0x1857881) [0x7fd158508881]
[bt] (4) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(ZNSt17_Function_handlerIFN4nnvm5GraphES1_EPS2_E9_M_invokeERKSt9_Any_dataS1+0x11f) [0x7fd157884dbf]
[bt] (5) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(_ZN4nnvm11ApplyPassesENS_5GraphERKSt6vectorISsSaISsEE+0x703) [0x7fd1584c8fc3]
[bt] (6) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(_ZN5mxnet18LoadLegacyJSONPassEN4nnvm5GraphE+0x18b) [0x7fd15788118b]
[bt] (7) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(ZNSt17_Function_handlerIFN4nnvm5GraphES1_EPS2_E9_M_invokeERKSt9_Any_dataS1+0x11f) [0x7fd157884dbf]
[bt] (8) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(_ZN4nnvm11ApplyPassesENS_5GraphERKSt6vectorISsSaISsEE+0x703) [0x7fd1584c8fc3]
[bt] (9) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(_ZN4nnvm9ApplyPassENS_5GraphERKSs+0x8e) [0x7fd15778e01e]
[bt] (10) /usr/local/lib/python2.7/dist-packages/mxnet-0.9.2-py2.7.egg/mxnet/libmxnet.so(MXSymbolCreateFromFile+0x471) [0x7fd1577873c1]
[bt] (11) /usr/lib/x86_64-linux-gnu/libffi.so.6(ffi_call_unix64+0x4c) [0x7fd15c558adc]
[bt] (12) /usr/lib/x86_64-linux-gnu/libffi.so.6(ffi_call+0x1fc) [0x7fd15c55840c]
[bt] (13) /usr/lib/python2.7/lib-dynload/_ctypes.x86_64-linux-gnu.so(_ctypes_callproc+0x48e) [0x7fd15c76f5fe]
[bt] (14) /usr/lib/python2.7/lib-dynload/_ctypes.x86_64-linux-gnu.so(+0x15f9e) [0x7fd15c770f9e]
[bt] (15) python(PyEval_EvalFrameEx+0x965) [0x499be5]
[bt] (16) python(PyEval_EvalCodeEx+0x2ac) [0x4a090c]
[bt] (17) python(PyEval_EvalFrameEx+0x18c5) [0x49ab45]
[bt] (18) python(PyEval_EvalCodeEx+0x2ac) [0x4a090c]
[bt] (19) python(PyEval_EvalFrameEx+0x18c5) [0x49ab45]
[bt] (20) python() [0x4a1c9a]
[bt] (21) python() [0x4dfe94]
[bt] (22) python(PyObject_Call+0x36) [0x505f96]
[bt] (23) python() [0x4de41a]
[bt] (24) python() [0x5039eb]
[bt] (25) python(PyEval_EvalFrameEx+0x965) [0x499be5]
[bt] (26) python(PyEval_EvalCodeEx+0x2ac) [0x4a090c]
[bt] (27) python(PyEval_EvalFrameEx+0x7d2) [0x499a52]
[bt] (28) python() [0x4a1634]
[bt] (29) python(PyRun_FileExFlags+0x92) [0x44e4a5]
[bt] (30) python(PyRun_SimpleFileExFlags+0x2ee) [0x44ec9f]
[bt] (31) python(Py_Main+0xb5e) [0x44f904]
[bt] (32) /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf5) [0x7fd15da25ec5]
[bt] (33) python() [0x578c4e]

Thank you!
@piiswrong
Copy link
Contributor

Please use example/ssd
The ssd in @zhreshold 's repo is not working anymore

@guo253
Copy link
Author

guo253 commented Jan 19, 2017 •
edited
Loading

@piiswrong @zhreshold
I indeed use example/ssd,and download 20170119.
This time I use CUDA=1.
There is still the error.

config.mk:
#EXTRA_OPERATORS =
EXTRA_OPERATORS = example/ssd/operator

Error Msg:
~/soft/code/mxnet/mxnet/example/ssd$ python demo.py
[12:19:34] include/dmlc/./logging.h:300: [12:19:34] src/core/op.cc:55: Check failed: op != nullptr Operator Scale is not registered

@zhreshold
Copy link
Member

@guo253 There is no more Scale operator in mxnet, are you mixing code in this repo with my repo?

@guo253
Copy link
Author

guo253 commented Jan 19, 2017

@zhreshold
I just use the ssd_300_voc_0712.zip from your repo.
Maybe that is the problem.
Thank you.

@HengjLi
Copy link

HengjLi commented Mar 13, 2017

Why removing scale operator in mxnet? I want to convert a trained caffe model (vggm) to a mxnet model. If there is no equivalent op in mxnet, what's the right conversion rule?

@zhreshold
Copy link
Member

@masterlee15 Because scale op is simply a broadcast_mul op. Use a mx.sym.Variable to hold the parameters of the scales instead.

@demingbh
Copy link

@zhreshold
The zip files in mxnet/example/ssd are all moved? Not exist in Dropbox now.
The URLs:
https://dl.dropboxusercontent.com/u/39265872/ssd_300_vgg16_reduced_voc0712_trainval.zip
https://dl.dropboxusercontent.com/u/39265872/vgg16_reduced.zip

@zhreshold
Copy link
Member

@demingbh Should have been fixed, plz try again

@wengyc
Copy link

wengyc commented Mar 24, 2017 •
edited
Loading

@zhreshold I redownload the mxnet today and make and setup.The model ssd_300.zip was also downloaded followed the link in example/ssd page.I run the demo.py without any modification and it has the same error.I want to know if the code need to be modified or i do something wrong with it.
Traceback (most recent call last): File "demo.py", line 96, in <module> ctx, args.nms_thresh, args.force_nms) File "demo.py", line 42, in get_detector data_shape, mean_pixels, ctx=ctx) File "/home/wyc/mxnet/example/ssd/detect/detector.py", line 34, in __init__ _, args, auxs = mx.model.load_checkpoint(model_prefix, epoch) File "/home/wyc/mxnet/python/mxnet/model.py", line 373, in load_checkpoint symbol = sym.load('%s-symbol.json' % prefix) File "/home/wyc/mxnet/python/mxnet/symbol.py", line 1215, in load check_call(_LIB.MXSymbolCreateFromFile(c_str(fname), ctypes.byref(handle))) File "/home/wyc/mxnet/python/mxnet/base.py", line 78, in check_call raise MXNetError(py_str(_LIB.MXGetLastError())) mxnet.base.MXNetError: Failed loading Op relu4_3_scale_20 of type Scale: [19:44:52] src/core /op.cc:55: Check failed: op != nullptr Operator Scale is not registered
Please help me.
Thank you!

@zhreshold
Copy link
Member

@wengyc I guess you have downloaded the model from zhreshold/mxnet-ssd, the model for mxnet/example/ssd don't have Scale and relu4_3_scale_20 inside. I have double checked the new model agian.

@wengyc
Copy link

wengyc commented Mar 25, 2017 •
edited
Loading

@zhreshold I has a new error.
` Traceback (most recent call last):

File "demo.py", line 96, in ctx, args.nms_thresh, args.force_nms)

File "demo.py", line 42, in get_detector data_shape, mean_pixels, ctx=ctx)

File "/home/wyc/mxnet/example/ssd/detect/detector.py", line 37, in init self.mod.bind(data_shapes=[('data', (batch_size, 3, data_shape, data_shape))])

File "/home/wyc/mxnet/python/mxnet/module/module.py", line 388, in bind state_names=self._state_names)

File "/home/wyc/mxnet/python/mxnet/module/executor_group.py", line 205, in init
self.bind_exec(data_shapes, label_shapes, shared_group)

File "/home/wyc/mxnet/python/mxnet/module/executor_group.py", line 301, in bind_exec
shared_group))

File "/home/wyc/mxnet/python/mxnet/module/executor_group.py", line 593, in _bind_ith_exec
arg_arr = nd.zeros(arg_shapes[j], context, dtype=arg_types[j])

File "/home/wyc/mxnet/python/mxnet/ndarray.py", line 900, in zeros
return _internal._zeros(shape=shape, ctx=ctx, dtype=dtype)

File "/home/wyc/mxnet/python/mxnet/_ctypes/ndarray.py", line 133, in generic_ndarray_function
c_array(ctypes.c_char_p, [c_str(str(i)) for i in kwargs.values()])))

File "/home/wyc/mxnet/python/mxnet/base.py", line 78, in check_call
raise MXNetError(py_str(_LIB.MXGetLastError()))

mxnet.base.MXNetError: [11:04:01] src/c_api/c_api_ndarray.cc:381: Operator _zeros cannot be run; requires at least one of FCompute<xpu>, NDArrayFunction, FCreateOperator be registered`

How could i do to deal with it?
Thank you!

@zhreshold
Copy link
Member

@wengyc python demo.py --cpu

@wengyc
Copy link

wengyc commented Mar 29, 2017

@zhreshold I downloaded the newest mxnet and make it.Now it works on gpu.Thank you!

@BenLag2906
Copy link

Hi,
I have the same problem.

"
C:\SSD\mxnet-ssd.cpp\build\vc14\Release>ssd person.jpg --gpu=-1 -m ./deploy_ssd_300
[09:26:57] C:\MXNET\mxnet\mxnet\dmlc-core\include\dmlc/logging.h:304: [09:26:57] C:\MXNET\mxnet\mxnet\nnvm\src\core\op.cc:55: Check failed: op != nullptr Operator Scale is not registered
"
where can I found the correct model ? drop box is not available anymore.

Thanks for your advises.

Ben

@lxtGH
Copy link

lxtGH commented Feb 7, 2018

try to compile mxnet with CUDNN=1

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
8 participants
@piiswrong @zhreshold @guo253 @HengjLi @demingbh @lxtGH @wengyc @BenLag2906