Comments (7)
I have switched compute_20, code=sm_20 to compute_30,code=sm_30
in setup file.
from sparseconvnet.
The hello-world.py
example works now. Thanks for the quick fix!
from sparseconvnet.
I have the same issue
from sparseconvnet.
Hello. To help me debug, can you please show the output from:
cd SpareConvNet/PyTorch
python setup.py develop
ls sparseconvnet/SCN/
(Also, what OS? What Python version? Conda or not?)
from sparseconvnet.
Hi btgraham,
the following log is my output when I use "python setup.py develop" in SpareConvNet/PyTorch:
ozzie@debian:~/working/work/ML/SparseConvNet/PyTorch$ python setup.py develop
Building SCN module
nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
generating /tmp/tmpS1UlkY/_SCN.c
running build_ext
building '_SCN' extension
gcc -pthread -fno-strict-aliasing -g -O2 -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -I/home/ozzie/anaconda2/lib/python2.7/site-packages/torch/utils/ffi/../../lib/include -I/home/ozzie/anaconda2/lib/python2.7/site-packages/torch/utils/ffi/../../lib/include/TH -I/home/ozzie/anaconda2/lib/python2.7/site-packages/torch/utils/ffi/../../lib/include/THC -I/usr/local/cuda/include -I/home/ozzie/anaconda2/include/python2.7 -c _SCN.c -o ./_SCN.o
gcc -pthread -shared -L/home/ozzie/anaconda2/lib -Wl,-rpath=/home/ozzie/anaconda2/lib,--no-as-needed ./_SCN.o /media/New_bt/ML/SparseConvNet/PyTorch/sparseconvnet/SCN/init.cu.o -L/home/ozzie/anaconda2/lib -lpython2.7 -o ./_SCN.so
running develop
running egg_info
creating sparseconvnet.egg-info
writing sparseconvnet.egg-info/PKG-INFO
writing top-level names to sparseconvnet.egg-info/top_level.txt
writing dependency_links to sparseconvnet.egg-info/dependency_links.txt
writing manifest file 'sparseconvnet.egg-info/SOURCES.txt'
reading manifest file 'sparseconvnet.egg-info/SOURCES.txt'
writing manifest file 'sparseconvnet.egg-info/SOURCES.txt'
running build_ext
Creating /home/ozzie/anaconda2/lib/python2.7/site-packages/sparseconvnet.egg-link (link to .)
Adding sparseconvnet 0.1 to easy-install.pth file
Installed /media/New_bt/ML/SparseConvNet/PyTorch
Processing dependencies for sparseconvnet==0.1
Finished processing dependencies for sparseconvnet==0.1
ozzie@debian:~/working/work/ML/SparseConvNet/examples/Assamese_handwriting$ python VGGplus.py
Downloading and preprocessing data ...
--2017-07-18 18:06:00-- https://archive.ics.uci.edu/ml/machine-learning-databases/00208/Online%20Handwritten%20Assamese%20Characters%20Dataset.rar
Resolving archive.ics.uci.edu (archive.ics.uci.edu)... 128.195.10.249
Connecting to archive.ics.uci.edu (archive.ics.uci.edu)|128.195.10.249|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 8067448 (7.7M) [text/plain]
Saving to: ‘Online Handwritten Assamese Characters Dataset.rar’
Online Handwritten 100%[=====================>] 7.69M 1.22MB/s in 12s
2017-07-18 18:06:13 (671 KB/s) - ‘Online Handwritten Assamese Characters Dataset.rar’ saved [8067448/8067448]
UNRAR 5.30 beta 2 freeware Copyright (c) 1993-2015 Alexander Roshal
Extracting from Online Handwritten Assamese Characters Dataset.rar
Extracting data_table.pdf OK
Extracting 1.1.txt OK
Extracting 10.1.txt OK
Extracting 100.1.txt OK
Extracting 101.1.txt OK
Extracting 102.1.txt OK
Extracting 103.1.txt OK
Extracting 104.1.txt OK
Extracting 105.1.txt OK
Extracting 106.1.txt OK
Extracting 107.1.txt OK
Extracting 108.1.txt OK
Extracting 109.1.txt OK
................ (the middle "Extracting xxx.txt OK" is removed by Ozzie Zhang because it is too long)
Extracting 53.9.txt OK
Extracting 54.9.txt OK
Extracting 55.9.txt OK
Extracting 56.9.txt OK
Extracting 57.9.txt OK
Extracting 58.9.txt OK
Extracting 59.9.txt OK
Extracting 6.9.txt OK
Extracting 60.9.txt OK
Extracting 61.9.txt OK
Extracting 62.9.txt OK
Extracting 63.9.txt OK
Extracting 64.9.txt OK
Extracting 65.9.txt OK
Extracting 66.9.txt OK
Extracting 67.9.txt OK
Extracting 68.9.txt OK
Extracting 69.9.txt OK
Extracting 7.9.txt OK
Extracting 70.9.txt OK
Extracting 71.9.txt OK
Extracting 72.9.txt OK
Extracting 73.9.txt OK
Extracting 74.9.txt OK
Extracting 75.9.txt OK
Extracting 76.9.txt OK
Extracting 77.9.txt OK
Extracting 78.9.txt OK
Extracting 79.9.txt OK
Extracting 8.9.txt OK
Extracting 80.9.txt OK
Extracting 81.9.txt OK
Extracting 82.9.txt OK
Extracting 83.9.txt OK
Extracting 84.9.txt OK
Extracting 85.9.txt OK
Extracting 86.9.txt OK
Extracting 87.9.txt OK
Extracting 88.9.txt OK
Extracting 89.9.txt OK
Extracting 9.9.txt OK
Extracting 90.9.txt OK
Extracting 91.9.txt OK
Extracting 92.9.txt OK
Extracting 93.9.txt OK
Extracting 94.9.txt OK
Extracting 95.9.txt OK
Extracting 96.9.txt OK
Extracting 97.9.txt OK
Extracting 98.9.txt OK
Extracting 99.9.txt OK
All OK
(6588, 1647)
nn.Sequential {
[input -> (0) -> (1) -> output]
(0): nn.Sequential {
[input -> (0) -> (1) -> (2) -> (3) -> output]
(0): nn.Sequential {
[input -> (0) -> (1) -> (2) -> (3) -> (4) -> (5) -> (6) -> (7) -> (8) -> (9) -> (10) -> (11) -> (12) -> (13) -> (14) -> (15) -> (16) -> (17) -> (18) -> (19) -> (20) -> (21) -> (22) -> (23) -> output]
(0): ValidConvolution 3->8 C3
(1): BatchNormReLU(8,eps=0.0001,momentum=0.9,affine=True)
(2): ValidConvolution 8->8 C3
(3): BatchNormReLU(8,eps=0.0001,momentum=0.9,affine=True)
(4): MaxPooling3/2
(5): ValidConvolution 8->16 C3
(6): BatchNormReLU(16,eps=0.0001,momentum=0.9,affine=True)
(7): ValidConvolution 16->16 C3
(8): BatchNormReLU(16,eps=0.0001,momentum=0.9,affine=True)
(9): MaxPooling3/2
(10): sparseconvnet.legacy.concatTable.ConcatTable {
input
|-> (0): ValidConvolution 16->16 C3 |
-> (1): nn.Sequential {
[input -> (0) -> (1) -> (2) -> (3) -> (4) -> output]
(0): Convolution 16->8 C3/2
(1): BatchNormReLU(8,eps=0.0001,momentum=0.9,affine=True)
(2): ValidConvolution 8->8 C3
(3): BatchNormReLU(8,eps=0.0001,momentum=0.9,affine=True)
(4): Deconvolution 8->8 C3/2
}
+. -> output
}
(11): JoinTable: 16 + 8 -> 24
(12): BatchNormReLU(24,eps=0.0001,momentum=0.9,affine=True)
(13): sparseconvnet.legacy.concatTable.ConcatTable {
input
|-> (0): ValidConvolution 24->16 C3 |
-> (1): nn.Sequential {
[input -> (0) -> (1) -> (2) -> (3) -> (4) -> output]
(0): Convolution 24->8 C3/2
(1): BatchNormReLU(8,eps=0.0001,momentum=0.9,affine=True)
(2): ValidConvolution 8->8 C3
(3): BatchNormReLU(8,eps=0.0001,momentum=0.9,affine=True)
(4): Deconvolution 8->8 C3/2
}
+. -> output
}
(14): JoinTable: 16 + 8 -> 24
(15): BatchNormReLU(24,eps=0.0001,momentum=0.9,affine=True)
(16): MaxPooling3/2
(17): sparseconvnet.legacy.concatTable.ConcatTable {
input
|-> (0): ValidConvolution 24->24 C3 |
-> (1): nn.Sequential {
[input -> (0) -> (1) -> (2) -> (3) -> (4) -> output]
(0): Convolution 24->8 C3/2
(1): BatchNormReLU(8,eps=0.0001,momentum=0.9,affine=True)
(2): ValidConvolution 8->8 C3
(3): BatchNormReLU(8,eps=0.0001,momentum=0.9,affine=True)
(4): Deconvolution 8->8 C3/2
}
+. -> output
}
(18): JoinTable: 24 + 8 -> 32
(19): BatchNormReLU(32,eps=0.0001,momentum=0.9,affine=True)
(20): sparseconvnet.legacy.concatTable.ConcatTable {
input
|-> (0): ValidConvolution 32->24 C3 |
-> (1): nn.Sequential {
[input -> (0) -> (1) -> (2) -> (3) -> (4) -> output]
(0): Convolution 32->8 C3/2
(1): BatchNormReLU(8,eps=0.0001,momentum=0.9,affine=True)
(2): ValidConvolution 8->8 C3
(3): BatchNormReLU(8,eps=0.0001,momentum=0.9,affine=True)
(4): Deconvolution 8->8 C3/2
}
+. -> output
}
(21): JoinTable: 24 + 8 -> 32
(22): BatchNormReLU(32,eps=0.0001,momentum=0.9,affine=True)
(23): MaxPooling3/2
}
(1): Convolution 32->64 C5/1
(2): BatchNormReLU(64,eps=0.0001,momentum=0.9,affine=True)
(3): SparseToDense(2)
}
(1): nn.Sequential {
[input -> (0) -> (1) -> output]
(0): nn.View(-1, 64)
(1): nn.Linear(64 -> 183)
}
}
('input spatial size',
95
95
[torch.LongTensor of size 2]
)
Replicating training set 10 times (1 epoch = 10 iterations through the training set = 10x6588 training samples)
{'weightDecay': 0.0001, 'initial_LR': 0.1, 'checkPoint': False, 'nEpochs': 100, 'LR_decay': 0.05, 'momentum': 0.9}
('#parameters', 97295)
THCudaCheck FAIL file=/b/wheel/pytorch-src/torch/lib/THC/generic/THCTensorMath.cu line=35 error=8 : invalid device function
Traceback (most recent call last):
File "VGGplus.py", line 38, in
{'nEpochs': 100, 'initial_LR': 0.1, 'LR_decay': 0.05, 'weightDecay': 1e-4})
File "/media/New_bt/ML/SparseConvNet/PyTorch/sparseconvnet/legacy/classificationTrainValidate.py", line 73, in ClassificationTrainValidate
model.forward(batch['input'])
File "/home/ozzie/anaconda2/lib/python2.7/site-packages/torch/legacy/nn/Module.py", line 33, in forward
return self.updateOutput(input)
File "/home/ozzie/anaconda2/lib/python2.7/site-packages/torch/legacy/nn/Sequential.py", line 36, in updateOutput
currentOutput = module.updateOutput(currentOutput)
File "/home/ozzie/anaconda2/lib/python2.7/site-packages/torch/legacy/nn/Sequential.py", line 36, in updateOutput
currentOutput = module.updateOutput(currentOutput)
File "/home/ozzie/anaconda2/lib/python2.7/site-packages/torch/legacy/nn/Sequential.py", line 36, in updateOutput
currentOutput = module.updateOutput(currentOutput)
File "/media/New_bt/ML/SparseConvNet/PyTorch/sparseconvnet/legacy/validConvolution.py", line 46, in updateOutput
torch.cuda.IntTensor() if input.features.is_cuda else nullptr)
File "/home/ozzie/anaconda2/lib/python2.7/site-packages/torch/utils/ffi/init.py", line 177, in safe_call
result = torch._C._safe_call(*args, **kwargs)
torch.FatalError: cuda runtime error (8) : invalid device function at /b/wheel/pytorch-src/torch/lib/THC/generic/THCTensorMath.cu:35
I guess that this should be a CUDA device issue related my GPU device number
I should use arch=compute_30,code=sm_30 because my GPU is nvidia k4000
from sparseconvnet.
my os is
uname -a
Linux debian 3.16.0-4-amd64 #1 SMP Debian 3.16.7-ckt25-2 (2016-04-08) x86_64 GNU/Linux
python
Python 2.7.12 |Anaconda custom (64-bit)| (default, Jul 2 2016, 17:42:40)
[GCC 4.4.7 20120313 (Red Hat 4.4.7-1)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
Anaconda is brought to you by Continuum Analytics.
Please check out: http://continuum.io/thanks and https://anaconda.org
from sparseconvnet.
my bug should be related to torch not SparseConvNet
from sparseconvnet.
Related Issues (20)
- sparseconvnet.SCN ImportError HOT 1
- Some questions about operational efficiency
- About the parameters of InputLayer HOT 3
- Dense to Sparse for input is quite slow
- AttributeError: module 'sparseconvnet.SCN' has no attribute 'Metadata_2' HOT 3
- RuntimeError: CUDA error: an illegal memory access was encountered HOT 2
- voxel input HOT 1
- Cloning face an error
- undefined symbol: _ZNSt15__exception_ptr13exception_ptr10_M_releaseEv HOT 2
- How to compute FLOPs for spraseconvnet HOT 3
- RuntimeError: expected scalar type Long but found Float HOT 5
- Building failure related to gcc version
- RuntimeError: expected scalar type Long but found Float
- Dilated convolution HOT 1
- Directly applying convolution HOT 1
- Rewrite for convolution operation
- Output with empty tensor
- setup.py中的C++17要改成C++14
- question from paper HOT 2
- Segment Fault due to resolution
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from sparseconvnet.