Git Product home page Git Product logo

cuda-convnet's People

Contributors

dnouri avatar kashif avatar yjxiong avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

cuda-convnet's Issues

osx

Hi,

have you tested it on osx with cuda 6 installed?

Thanks,
Gerald

Import Error When run convnet.py

I tried the given CMakeLists.txt and built the _covnet.so successfully, but When I tried to run "python convnet.py ..." . An error arose: Import Error: undefined symbol: cblas_sgemm. It seems that "import("_convnet.so") " in convnet.py can not find the definition of cblas_sgemm. Even after I export the path of "libbblas.so" to "LD_LIBRARY_PATH" , the error still exists. What should I do to solve this problem?

Thanks.

various syntax errors while trying to make

Hi, I'm trying to make the project on Ubuntu 16.04. First I run
cmake -DBLAS_LIBRARIES=/OpenBLAS/libopenblas.so -DBLAS_INCLUDE_DIRS=/data/samin/OpenBLAS .

Because it asks me to initialize BLAS_INCLUDE_DIRS. Then when I try make I get several syntax errors like these:

/data/samin/cuda-convnet/include/common/matrix.h:308:1: error: expected declaration before ‘}’ token
};

/data/samin/OpenBLAS/common.h:168:15: error: expected ‘}’ before numeric constant
#define ZERO 0.e0f

what's the issue?
Here is the complete log:
[ 4%] Building CXX object src/common/CMakeFiles/common.dir/matrix.cpp.o
In file included from /data/samin/OpenBLAS/cblas.h:5:0,
from /data/samin/cuda-convnet/include/common/matrix.h:86,
from /data/samin/cuda-convnet/src/common/matrix.cpp:27:
/data/samin/OpenBLAS/common.h:168:15: error: expected identifier before numeric constant
#define ZERO 0.e0f
^
/data/samin/cuda-convnet/include/common/matrix.h:133:50: note: in expansion of macro ‘ZERO’
TANH, RECIPROCAL, SQUARE, ABS, EXP, LOG, ZERO, ONE, LOGISTIC1, LOGISTIC2, SIGN
^
/data/samin/OpenBLAS/common.h:168:15: error: expected ‘}’ before numeric constant
#define ZERO 0.e0f
^
/data/samin/cuda-convnet/include/common/matrix.h:133:50: note: in expansion of macro ‘ZERO’
TANH, RECIPROCAL, SQUARE, ABS, EXP, LOG, ZERO, ONE, LOGISTIC1, LOGISTIC2, SIGN
^
/data/samin/OpenBLAS/common.h:168:15: error: expected unqualified-id before numeric constant
#define ZERO 0.e0f
^
/data/samin/cuda-convnet/include/common/matrix.h:133:50: note: in expansion of macro ‘ZERO’
TANH, RECIPROCAL, SQUARE, ABS, EXP, LOG, ZERO, ONE, LOGISTIC1, LOGISTIC2, SIGN
^
In file included from /data/samin/cuda-convnet/src/common/matrix.cpp:27:0:
/data/samin/cuda-convnet/include/common/matrix.h:135:12: error: expected unqualified-id before ‘)’ token
Matrix();
^
/data/samin/cuda-convnet/include/common/matrix.h:136:12: error: expected unqualified-id before ‘long’
Matrix(long int numRows, long int numCols);
^
/data/samin/cuda-convnet/include/common/matrix.h:136:12: error: expected ‘)’ before ‘long’
/data/samin/cuda-convnet/include/common/matrix.h:140:12: error: expected unqualified-id before ‘const’
Matrix(const Matrix &like);
^
/data/samin/cuda-convnet/include/common/matrix.h:140:12: error: expected ‘)’ before ‘const’
In file included from /data/samin/cuda-convnet/include/common/matrix.h:30:0,
from /data/samin/cuda-convnet/src/common/matrix.cpp:27:
/data/samin/cuda-convnet/include/common/matrix_funcs.h:37:15: error: expected unqualified-id before ‘float’
#define MTYPE float
^
/data/samin/cuda-convnet/include/common/matrix.h:141:12: note: in expansion of macro ‘MTYPE’
Matrix(MTYPE* data, long int numRows, long int numCols);
^
/data/samin/cuda-convnet/include/common/matrix_funcs.h:37:15: error: expected ‘)’ before ‘float’
#define MTYPE float
^
/data/samin/cuda-convnet/include/common/matrix.h:141:12: note: in expansion of macro ‘MTYPE’
Matrix(MTYPE* data, long int numRows, long int numCols);
^
/data/samin/cuda-convnet/include/common/matrix_funcs.h:37:15: error: expected unqualified-id before ‘float’
#define MTYPE float
^
/data/samin/cuda-convnet/include/common/matrix.h:142:12: note: in expansion of macro ‘MTYPE’
Matrix(MTYPE* data, long int numRows, long int numCols, bool transpose);
^
/data/samin/cuda-convnet/include/common/matrix_funcs.h:37:15: error: expected ‘)’ before ‘float’
#define MTYPE float
^
/data/samin/cuda-convnet/include/common/matrix.h:142:12: note: in expansion of macro ‘MTYPE’
Matrix(MTYPE* data, long int numRows, long int numCols, bool transpose);
^
In file included from /data/samin/cuda-convnet/src/common/matrix.cpp:27:0:
/data/samin/cuda-convnet/include/common/matrix.h:143:14: error: expected constructor, destructor, or type conversion before ‘;’ token
~Matrix();
^
/data/samin/cuda-convnet/include/common/matrix.h:145:51: error: non-member function ‘float& getCell(long int, long int)’ cannot have cv-qualifier
inline MTYPE& getCell(long int i, long int j) const {
^
In file included from /data/samin/OpenBLAS/common_stackalloc.h:31:0,
from /data/samin/OpenBLAS/common.h:793,
from /data/samin/OpenBLAS/cblas.h:5,
from /data/samin/cuda-convnet/include/common/matrix.h:86,
from /data/samin/cuda-convnet/src/common/matrix.cpp:27:
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘float& getCell(long int, long int)’:
/data/samin/cuda-convnet/include/common/matrix.h:146:30: error: ‘_numRows’ was not declared in this scope
assert(i >= 0 && i < _numRows);
^
/data/samin/cuda-convnet/include/common/matrix.h:147:30: error: ‘_numCols’ was not declared in this scope
assert(j >= 0 && j < _numCols);
^
In file included from /data/samin/cuda-convnet/src/common/matrix.cpp:27:0:
/data/samin/cuda-convnet/include/common/matrix.h:148:13: error: ‘_trans’ was not declared in this scope
if (_trans == CblasTrans) {
^
/data/samin/cuda-convnet/include/common/matrix.h:149:20: error: ‘_data’ was not declared in this scope
return _data[j * _numRows + i];
^
/data/samin/cuda-convnet/include/common/matrix.h:151:16: error: ‘_data’ was not declared in this scope
return _data[i * _numCols + j];
^
/data/samin/cuda-convnet/include/common/matrix.h: At global scope:
/data/samin/cuda-convnet/include/common/matrix.h:154:47: error: non-member function ‘float& operator()(long int, long int)’ cannot have cv-qualifier
MTYPE& operator()(long int i, long int j) const {
^
/data/samin/cuda-convnet/include/common/matrix.h:154:47: error: ‘float& operator()(long int, long int)’ must be a nonstatic member function
/data/samin/cuda-convnet/include/common/matrix.h:158:29: error: non-member function ‘float* getData()’ cannot have cv-qualifier
inline MTYPE* getData() const {
^
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘float* getData()’:
/data/samin/cuda-convnet/include/common/matrix.h:159:16: error: ‘_data’ was not declared in this scope
return _data;
^
/data/samin/cuda-convnet/include/common/matrix.h: At global scope:
/data/samin/cuda-convnet/include/common/matrix.h:162:26: error: non-member function ‘bool isView()’ cannot have cv-qualifier
inline bool isView() const {
^
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘bool isView()’:
/data/samin/cuda-convnet/include/common/matrix.h:163:17: error: ‘_ownsData’ was not declared in this scope
return !_ownsData;
^
/data/samin/cuda-convnet/include/common/matrix.h: At global scope:
/data/samin/cuda-convnet/include/common/matrix.h:166:34: error: non-member function ‘long int getNumRows()’ cannot have cv-qualifier
inline long int getNumRows() const {
^
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘long int getNumRows()’:
/data/samin/cuda-convnet/include/common/matrix.h:167:16: error: ‘_numRows’ was not declared in this scope
return _numRows;
^
/data/samin/cuda-convnet/include/common/matrix.h: At global scope:
/data/samin/cuda-convnet/include/common/matrix.h:170:34: error: non-member function ‘long int getNumCols()’ cannot have cv-qualifier
inline long int getNumCols() const {
^
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘long int getNumCols()’:
/data/samin/cuda-convnet/include/common/matrix.h:171:16: error: ‘_numCols’ was not declared in this scope
return _numCols;
^
/data/samin/cuda-convnet/include/common/matrix.h: At global scope:
/data/samin/cuda-convnet/include/common/matrix.h:174:39: error: non-member function ‘long int getNumDataBytes()’ cannot have cv-qualifier
inline long int getNumDataBytes() const {
^
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘long int getNumDataBytes()’:
/data/samin/cuda-convnet/include/common/matrix.h:175:16: error: ‘_numElements’ was not declared in this scope
return _numElements * sizeof(MTYPE);
^
/data/samin/cuda-convnet/include/common/matrix.h: At global scope:
/data/samin/cuda-convnet/include/common/matrix.h:178:38: error: non-member function ‘long int getNumElements()’ cannot have cv-qualifier
inline long int getNumElements() const {
^
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘long int getNumElements()’:
/data/samin/cuda-convnet/include/common/matrix.h:179:16: error: ‘_numElements’ was not declared in this scope
return _numElements;
^
/data/samin/cuda-convnet/include/common/matrix.h: At global scope:
/data/samin/cuda-convnet/include/common/matrix.h:182:37: error: non-member function ‘long int getLeadingDim()’ cannot have cv-qualifier
inline long int getLeadingDim() const {
^
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘long int getLeadingDim()’:
/data/samin/cuda-convnet/include/common/matrix.h:183:16: error: ‘_trans’ was not declared in this scope
return _trans == CblasTrans ? _numRows : _numCols;
^
/data/samin/cuda-convnet/include/common/matrix.h:183:39: error: ‘_numRows’ was not declared in this scope
return _trans == CblasTrans ? _numRows : _numCols;
^
/data/samin/cuda-convnet/include/common/matrix.h:183:50: error: ‘_numCols’ was not declared in this scope
return _trans == CblasTrans ? _numRows : _numCols;
^
/data/samin/cuda-convnet/include/common/matrix.h: At global scope:
/data/samin/cuda-convnet/include/common/matrix.h:186:39: error: non-member function ‘long int getFollowingDim()’ cannot have cv-qualifier
inline long int getFollowingDim() const {
^
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘long int getFollowingDim()’:
/data/samin/cuda-convnet/include/common/matrix.h:187:16: error: ‘_trans’ was not declared in this scope
return _trans == CblasTrans ? _numCols : _numRows;
^
/data/samin/cuda-convnet/include/common/matrix.h:187:39: error: ‘_numCols’ was not declared in this scope
return _trans == CblasTrans ? _numCols : _numRows;
^
/data/samin/cuda-convnet/include/common/matrix.h:187:50: error: ‘_numRows’ was not declared in this scope
return _trans == CblasTrans ? _numCols : _numRows;
^
/data/samin/cuda-convnet/include/common/matrix.h: At global scope:
/data/samin/cuda-convnet/include/common/matrix.h:190:43: error: non-member function ‘CBLAS_TRANSPOSE getBLASTrans()’ cannot have cv-qualifier
inline CBLAS_TRANSPOSE getBLASTrans() const {
^
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘CBLAS_TRANSPOSE getBLASTrans()’:
/data/samin/cuda-convnet/include/common/matrix.h:191:16: error: ‘_trans’ was not declared in this scope
return _trans;
^
/data/samin/cuda-convnet/include/common/matrix.h: At global scope:
/data/samin/cuda-convnet/include/common/matrix.h:194:45: error: non-member function ‘bool isSameDims(const Matrix&)’ cannot have cv-qualifier
inline bool isSameDims(const Matrix& a) const {
^
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘bool isSameDims(const Matrix&)’:
/data/samin/cuda-convnet/include/common/matrix.h:195:18: error: ‘const class Matrix’ has no member named ‘getNumRows’
return a.getNumRows() == getNumRows() && a.getNumCols() == getNumCols();
^
/data/samin/cuda-convnet/include/common/matrix.h:195:52: error: ‘const class Matrix’ has no member named ‘getNumCols’
return a.getNumRows() == getNumRows() && a.getNumCols() == getNumCols();
^
/data/samin/cuda-convnet/include/common/matrix.h: At global scope:
/data/samin/cuda-convnet/include/common/matrix.h:198:27: error: non-member function ‘bool isTrans()’ cannot have cv-qualifier
inline bool isTrans() const {
^
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘bool isTrans()’:
/data/samin/cuda-convnet/include/common/matrix.h:199:16: error: ‘_trans’ was not declared in this scope
return _trans == CblasTrans;
^
/data/samin/cuda-convnet/include/common/matrix.h: In function ‘void setTrans(bool)’:
/data/samin/cuda-convnet/include/common/matrix.h:210:9: error: ‘_trans’ was not declared in this scope
_trans = trans ? CblasTrans : CblasNoTrans;
^
/data/samin/cuda-convnet/include/common/matrix.h: At global scope:
/data/samin/cuda-convnet/include/common/matrix.h:213:16: error: variable or field ‘apply’ declared void
void apply(FUNCTION f);
^
/data/samin/cuda-convnet/include/common/matrix.h:213:16: error: ‘FUNCTION’ was not declared in this scope
/data/samin/cuda-convnet/include/common/matrix.h:216:59: error: non-member function ‘void subtractFromScalar(float, Matrix&)’ cannot have cv-qualifier
void subtractFromScalar(MTYPE scalar, Matrix &target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:220:57: error: non-member function ‘void biggerThanScalar(float, Matrix&)’ cannot have cv-qualifier
void biggerThanScalar(MTYPE scalar, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:221:58: error: non-member function ‘void smallerThanScalar(float, Matrix&)’ cannot have cv-qualifier
void smallerThanScalar(MTYPE scalar, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:222:53: error: non-member function ‘void equalsScalar(float, Matrix&)’ cannot have cv-qualifier
void equalsScalar(MTYPE scalar, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:224:48: error: non-member function ‘void biggerThan(Matrix&, Matrix&)’ cannot have cv-qualifier
void biggerThan(Matrix& a, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:226:49: error: non-member function ‘void smallerThan(Matrix&, Matrix&)’ cannot have cv-qualifier
void smallerThan(Matrix& a, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:228:45: error: non-member function ‘void minWith(Matrix&, Matrix&)’ cannot have cv-qualifier
void minWith(Matrix &a, Matrix &target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:230:45: error: non-member function ‘void maxWith(Matrix&, Matrix&)’ cannot have cv-qualifier
void maxWith(Matrix &a, Matrix &target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:232:44: error: non-member function ‘void equals(Matrix&, Matrix&)’ cannot have cv-qualifier
void equals(Matrix& a, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:234:47: error: non-member function ‘void notEquals(Matrix&, Matrix&)’ cannot have cv-qualifier
void notEquals(Matrix& a, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:248:50: error: non-member function ‘void addScalar(float, Matrix&)’ cannot have cv-qualifier
void addScalar(MTYPE scalar, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:250:54: error: non-member function ‘void maxWithScalar(float, Matrix&)’ cannot have cv-qualifier
void maxWithScalar(MTYPE scalar, Matrix &target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:252:54: error: non-member function ‘void minWithScalar(float, Matrix&)’ cannot have cv-qualifier
void minWithScalar(MTYPE scalar, Matrix &target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:259:91: error: non-member function ‘Matrix& slice(long int, long int, long int, long int)’ cannot have cv-qualifier
Matrix& slice(long int startRow, long int endRow, long int startCol, long int endCol) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:260:104: error: non-member function ‘void slice(long int, long int, long int, long int, Matrix&)’ cannot have cv-qualifier
void slice(long int startRow, long int endRow, long int startCol, long int endCol, Matrix &target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:261:59: error: non-member function ‘Matrix& sliceRows(long int, long int)’ cannot have cv-qualifier
Matrix& sliceRows(long int startRow, long int endRow) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:262:72: error: non-member function ‘void sliceRows(long int, long int, Matrix&)’ cannot have cv-qualifier
void sliceRows(long int startRow, long int endRow, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:263:59: error: non-member function ‘Matrix& sliceCols(long int, long int)’ cannot have cv-qualifier
Matrix& sliceCols(long int startCol, long int endCol) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:264:72: error: non-member function ‘void sliceCols(long int, long int, Matrix&)’ cannot have cv-qualifier
void sliceCols(long int startCol, long int endCol, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:266:53: error: non-member function ‘void rightMult(const Matrix&, Matrix&)’ cannot have cv-qualifier
void rightMult(const Matrix &b, Matrix &target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:268:68: error: non-member function ‘void rightMult(const Matrix&, float, Matrix&)’ cannot have cv-qualifier
void rightMult(const Matrix &b, MTYPE scaleAB, Matrix &target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:272:55: error: non-member function ‘void eltWiseMult(const Matrix&, Matrix&)’ cannot have cv-qualifier
void eltWiseMult(const Matrix& a, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:274:57: error: non-member function ‘void eltWiseDivide(const Matrix&, Matrix&)’ cannot have cv-qualifier
void eltWiseDivide(const Matrix& a, Matrix &target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:275:25: error: non-member function ‘Matrix& transpose()’ cannot have cv-qualifier
Matrix& transpose() const;
^
/data/samin/cuda-convnet/include/common/matrix.h:276:34: error: non-member function ‘Matrix& transpose(bool)’ cannot have cv-qualifier
Matrix& transpose(bool hard) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:277:52: error: non-member function ‘Matrix& tile(long int, long int)’ cannot have cv-qualifier
Matrix& tile(long int timesY, long int timesX) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:278:65: error: non-member function ‘void tile(long int, long int, Matrix&)’ cannot have cv-qualifier
void tile(long int timesY, long int timesX, Matrix& target) const;
^
In file included from /data/samin/cuda-convnet/src/common/matrix.cpp:27:0:
/data/samin/cuda-convnet/include/common/matrix.h:279:159: error: non-member function ‘void copy(Matrix&, long int, long int, long int, long int, long int, long int)’ cannot have cv-qualifier
void copy(Matrix &dest, long int srcStartRow, long int srcEndRow, long int srcStartCol, long int srcEndCol, long int destStartRow, long int destStartCol) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:280:20: error: non-member function ‘Matrix& copy()’ cannot have cv-qualifier
Matrix& copy() const;
^
/data/samin/cuda-convnet/include/common/matrix.h:281:31: error: non-member function ‘void copy(Matrix&)’ cannot have cv-qualifier
void copy(Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:282:32: error: non-member function ‘Matrix& sum(long int)’ cannot have cv-qualifier
Matrix& sum(long int axis) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:283:45: error: non-member function ‘void sum(long int, Matrix&)’ cannot have cv-qualifier
void sum(long int axis, Matrix &target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:284:17: error: non-member function ‘float sum()’ cannot have cv-qualifier
MTYPE sum() const;
^
/data/samin/cuda-convnet/include/common/matrix.h:285:17: error: non-member function ‘float max()’ cannot have cv-qualifier
MTYPE max() const;
^
/data/samin/cuda-convnet/include/common/matrix.h:286:32: error: non-member function ‘Matrix& max(long int)’ cannot have cv-qualifier
Matrix& max(long int axis) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:287:45: error: non-member function ‘void max(long int, Matrix&)’ cannot have cv-qualifier
void max(long int axis, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:288:17: error: non-member function ‘float min()’ cannot have cv-qualifier
MTYPE min() const;
^
/data/samin/cuda-convnet/include/common/matrix.h:289:32: error: non-member function ‘Matrix& min(long int)’ cannot have cv-qualifier
Matrix& min(long int axis) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:290:45: error: non-member function ‘void min(long int, Matrix&)’ cannot have cv-qualifier
void min(long int axis, Matrix& target) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:291:18: error: non-member function ‘float norm()’ cannot have cv-qualifier
MTYPE norm() const;
^
/data/samin/cuda-convnet/include/common/matrix.h:292:19: error: non-member function ‘float norm2()’ cannot have cv-qualifier
MTYPE norm2() const;
^
/data/samin/cuda-convnet/include/common/matrix.h:297:39: error: non-member function ‘void printShape(const char*)’ cannot have cv-qualifier
void printShape(const char* name) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:298:19: error: non-member function ‘bool hasNan()’ cannot have cv-qualifier
bool hasNan() const;
^
/data/samin/cuda-convnet/include/common/matrix.h:299:19: error: non-member function ‘bool hasInf()’ cannot have cv-qualifier
bool hasInf() const;
^
/data/samin/cuda-convnet/include/common/matrix.h:309:18: error: non-member function ‘void print()’ cannot have cv-qualifier
void print() const;
^
/data/samin/cuda-convnet/include/common/matrix.h:310:82: error: non-member function ‘void print(long int, long int, long int, long int)’ cannot have cv-qualifier
void print(long int startRow,long int rows, long int startCol,long int cols) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:311:46: error: non-member function ‘void print(long int, long int)’ cannot have cv-qualifier
void print(long int rows, long int cols) const;
^
/data/samin/cuda-convnet/include/common/matrix.h:312:1: error: expected declaration before ‘}’ token
};
^
src/common/CMakeFiles/common.dir/build.make:62: recipe for target 'src/common/CMakeFiles/common.dir/matrix.cpp.o' failed
make[2]: *** [src/common/CMakeFiles/common.dir/matrix.cpp.o] Error 1
CMakeFiles/Makefile2:156: recipe for target 'src/common/CMakeFiles/common.dir/all' failed
make[1]: *** [src/common/CMakeFiles/common.dir/all] Error 2
Makefile:83: recipe for target 'all' failed
make: *** [all] Error 2

error when I tried this code in linux...

When I replaced the codes in layer.py like the below...

    dic, name = self.dic, self.dic['name']
    dic['dropout'] = 0.0
    if name in mcp.sections():
        dic['dropout'] = mcp.safe_get_float(name, 'dropout', default=0.0)

it shown the error message which's about the 'name' in dic is not defined..
well, I changed the codes in layer.cu, layer.cuh, convert.cu, layer.py after comparing files with the original cuda-convnet codes.

How to solve this issues? please give me the tips..

[noccn] Make batch File wont work on 128 and 256 images

Haloo,
did it because of the size of the images too large ?

and do you have any suggestions about layers n params that best match for the 64 image ?

I trained using your dropout, 15 classes (14million images, 7 batches), cifar-cropped , it was 36% error..... how to get lower error ?

thanks

Why should we double the number of outputs when using dropout?

Hi Daniel,

Thank you for modifying Alex's code to enable Hinton's dropout.

Is it possible for you to please explain in the README why you suggest doubling the number of outputs in the last layer when using dropout?

RHH

In practice, you'll probably also want to double the number of outputs in that layer.

Does that mean if are making a simple binary classifier, then the number of outputs should be four when using dropout? How do we interpret four outputs from a binary classifier?

error on osx

I get the following error trying to compile this on osx:

/usr/include/sys/types.h(92): error: duplicate specifier in declaration

/usr/include/sys/types.h(92): error: invalid combination of type specifiers

2 errors detected in the compilation of "/var/folders/dn/8cdhcbp10vn436v1n27631080000gn/T//tmpxft_0000c305_00000000-7_conv_util.cpp1.ii".
CMake Error at cudaconv2_generated_conv_util.cu.o.cmake:264 (message):
  Error generating file
  /Users/jjarutis/dev/cuda-convnet/src/cudaconv2/CMakeFiles/cudaconv2.dir//./cudaconv2_generated_conv_util.cu.o


make[2]: *** [src/cudaconv2/CMakeFiles/cudaconv2.dir/cudaconv2_generated_conv_util.cu.o] Error 1
make[1]: *** [src/cudaconv2/CMakeFiles/cudaconv2.dir/all] Error 2
make: *** [all] Error 2

I have cuda 7.0

cmake modifications break initRandom

@kashif

The following error appears under the following case when running convnet.py.

  • ab2c1fd Fedora 20 64-bit GeForce 610M

src/nvmatrix/nvmatrix.cu(279) : getLastCudaError() CUDA error : initRandom: Kernel execution failed : (8) invalid device function .

But the error does not appear in the following cases:

Weight array Access

Hi Daniel,
I was looking at the code and trying to figure out how could I get/set the weights/bias for the convolutional layers directly.
Do you known if there is a way to do that?

Thanks.

Error rate reached : logprob: 0.066761, 0.002564 is it ok or smth went wrong ?

Haloo

I used your dropout version of Convnet for Food classification.

I have 8 batches :

1-6 : train, 7 validation
and 8 for test

I followed the methodology https://code.google.com/p/cuda-convnet/wiki/Methodology

I started the training :
--train-range=1-6 --test-range=7 --epochs=200

an so on like the methodology...

The error rates for each batches is Good ( around 0.11)

but when I test it ( last method) with batch 8 . its 63% (final result).

I dont know what caused this. I assume that Foods share almost the same colors, shape, etc, mayble those are the cause.

What do you suggest ?

Thanks

Result

Did someone try to use Imagenet Layers ?
What was the performance for Imagenet database ?
I'm currently using it but I'm not sure that my computer allow me to use 1.2 millions image for training... With 400 000 images I have 92% error rate...

Import error convnet

Trying to run cuda convenet using my own dataset in order to train my model . but got the following error .googled alot but failed to get answer. i am getting the following error in convent.py file .although i have installed cuda tool kit vs etc but unable to understand why this is happening

Importing cudaconvnet._ConvNet C++ module

ImportErrorTraceback (most recent call last)
in ()
101 print "========================="
102 print "Importing %s C++ module" % lib_name
--> 103 libmodel = import(lib_name,fromlist=['_ConvNet'])
104
105 class ConvNet(IGPUModel):
ImportError: No module named _ConvNet

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo 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.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.