Issue in running the demo

[divamgupta]

I was trying to run the demo. I am getting the following error:

WARNING: Logging before InitGoogleLogging() is written to STDERR
F0808 09:21:26.621508  2574 common.cpp:53] CPU-only Mode: cannot make GPU call.
*** Check failure stack trace: ***
Aborted (core dumped)

I have even set isGPU= False

[TheTesla]

misstake - write GPU=False not Flase

On 08.08.2016 11:27, Divam Gupta wrote:

I have even set isGPU= Flase

[divamgupta]

@TheTesla Sorry that typo was in this issue. I wrote False in the .py file only.

[divamgupta]

I replaced GPU with CPU in pycaffe_utils/my_pycaffe_utils.py line 1487

('solver_mode', 'CPU')

Still I am getting the same error.

[pulkitag]

Where have you set isGPU=False? Can you detail what command you are running?

[divamgupta]

This is the fork with the isGPU updated

https://github.com/divamgupta/ief
https://github.com/divamgupta/pycaffe-utils

[pulkitag]

Point me to the specifics.

On Fri, Aug 12, 2016 at 9:57 AM, Divam Gupta notifications@github.com
wrote:

This is the fork with the isGPU updated

https://github.com/divamgupta/ief
https://github.com/divamgupta/pycaffe-utils

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#8 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/AETXxiGZa7N2YxjgxhMFFEVOTwc2VZdXks5qe_ZGgaJpZM4Je25c
.

[divamgupta]

In test_demo.py in line 30 and 40.
In my_pycaffe.py line 271 and 294.

I am running

from src import test_demo as td
#Define pose-predictor class
ief    = td.PoseIEF()
#Name of the image
imName = 'src/test_images/elvis.jpg'
#Point (x,y) on the torso of the person whose pose is to be estimated
bodyPt = (108, 98)
#Predict the pose
pose,_ =  ief.predict(imName, bodyPt)

[pulkitag]

You should call

ief = td.PoseIEF(isGPU=False)

In your fork, you have isGPU=False in some functions but not in others.

[divamgupta]

I tried that also. Still same error.

[pulkitag]

Can you try,

ief = td.PoseIEF(isGPU=False, deviceId=None)

If this doesnot work, can you tell me if you are able to run normal caffe models in CPU mode?

[divamgupta]

The GPU error went away, but a new error came

WARNING: Logging before InitGoogleLogging() is written to STDERR
W0814 10:50:45.724055  5207 _caffe.cpp:122] DEPRECATION WARNING - deprecated use of Python interface
W0814 10:50:45.724424  5207 _caffe.cpp:123] Use this instead (with the named "weights" parameter):
W0814 10:50:45.724555  5207 _caffe.cpp:125] Net('models/vgg_s.prototxt', 1, weights='models/scalesel-vggs-epoch-14-convert.caffemodel')
[libprotobuf ERROR google/protobuf/text_format.cc:245] Error parsing text-format caffe.NetParameter: 133:28: Message type "caffe.PoolingParameter" has no field named "round_below".
F0814 10:50:45.726429  5207 upgrade_proto.cpp:79] Check failed: ReadProtoFromTextFile(param_file, param) Failed to parse NetParameter file: models/vgg_s.prototxt
*** Check failure stack trace: ***

Aborted (core dumped)

Can you tell which commit of caffe you are using?

Yes I can run normal caffe models in CPU mode.

[divamgupta]

And I tried compiling your fork and I got the following error

CXX src/caffe/util/imchannel2col.cpp
src/caffe/util/imchannel2col.cpp: In function 'void caffe::imchannel2col_cpu(const Dtype*, int, int, int, int, int, int, int, int, int, int, int, Dtype*)':
src/caffe/util/imchannel2col.cpp:36:46: error: there are no arguments to 'caffe_cpu_modulus' that depend on a template parameter, so a declaration of 'caffe_cpu_modulus' must be available [-fpermissive]
src/caffe/util/imchannel2col.cpp:36:46: note: (if you use '-fpermissive', G++ will accept your code, but allowing the use of an undeclared name is deprecated)
src/caffe/util/imchannel2col.cpp:37:45: error: there are no arguments to 'caffe_cpu_modulus' that depend on a template parameter, so a declaration of 'caffe_cpu_modulus' must be available [-fpermissive]
src/caffe/util/imchannel2col.cpp: In function 'void caffe::colchannel2im_cpu(const Dtype*, int, int, int, int, int, int, int, int, int, int, int, Dtype*)':
src/caffe/util/imchannel2col.cpp:101:46: error: there are no arguments to 'caffe_cpu_modulus' that depend on a template parameter, so a declaration of 'caffe_cpu_modulus' must be available [-fpermissive]
src/caffe/util/imchannel2col.cpp:102:45: error: there are no arguments to 'caffe_cpu_modulus' that depend on a template parameter, so a declaration of 'caffe_cpu_modulus' must be available [-fpermissive]
src/caffe/util/imchannel2col.cpp: In function 'void caffe::imchannel2col_cpu(const Dtype*, int, int, int, int, int, int, int, int, int, int, int, Dtype*) [with Dtype = float]':
src/caffe/util/imchannel2col.cpp:71:58:   instantiated from here
src/caffe/util/imchannel2col.cpp:36:5: error: 'caffe_cpu_modulus' was not declared in this scope
src/caffe/util/imchannel2col.cpp: In function 'void caffe::imchannel2col_cpu(const Dtype*, int, int, int, int, int, int, int, int, int, int, int, Dtype*) [with Dtype = double]':
src/caffe/util/imchannel2col.cpp:77:60:   instantiated from here
src/caffe/util/imchannel2col.cpp:36:5: error: 'caffe_cpu_modulus' was not declared in this scope
src/caffe/util/imchannel2col.cpp: In function 'void caffe::colchannel2im_cpu(const Dtype*, int, int, int, int, int, int, int, int, int, int, int, Dtype*) [with Dtype = float]':
src/caffe/util/imchannel2col.cpp:124:39:   instantiated from here
src/caffe/util/imchannel2col.cpp:101:5: error: 'caffe_cpu_modulus' was not declared in this scope
src/caffe/util/imchannel2col.cpp: In function 'void caffe::colchannel2im_cpu(const Dtype*, int, int, int, int, int, int, int, int, int, int, int, Dtype*) [with Dtype = double]':
src/caffe/util/imchannel2col.cpp:131:40:   instantiated from here
src/caffe/util/imchannel2col.cpp:101:5: error: 'caffe_cpu_modulus' was not declared in this scope
make: *** [.build_release/src/caffe/util/imchannel2col.o] Error 1

[sampath0339]

Hello

I am getting this error when i run demo.

F1109 22:39:26.245774 13244 net.cpp:774] Cannot copy param 0 weights from layer 'fc6'; shape mismatch. Source param shape is 1 1 4096 18432 (75497472); target param shape is 4096 25088 (102760448). To learn this layer's parameters from scratch rather than copying from a saved net, rename the layer.

Please provide some suggestions.