Skip to content

P
pytensor
提交 2681accc authored 作者: David Warde-Farley's avatar David Warde-Farley
浏览文件
操作

Big whitespace cleanup.

上级 b7a6e812
隐藏空白字符变更
内嵌 并排
正在显示 包含 197 行增加 和 197 行删除
theano/sandbox/cuda/GpuConv3D.py
浏览文件 @ 2681accc
......@@ -11,13 +11,13 @@ if cuda_available:
class GpuConv3D(theano.Op):
""" GPU implementation of Conv3D """
def __eq__(self, other):
return type(self) == type(other)
def __hash__(self):
return hash(type(self))
def __str__(self):
return '%s' % (self.__class__.__name__)
......@@ -33,16 +33,16 @@ class GpuConv3D(theano.Op):
b_ = as_cuda_ndarray_variable(b)
d_ = T.as_tensor_variable(d)
return theano.Apply(self, inputs=[V_, W_, b_, d_],
return theano.Apply(self, inputs=[V_, W_, b_, d_],
outputs = [ CudaNdarrayType(dtype=V_.dtype, broadcastable=(V_.broadcastable[0],W_.broadcastable[0],False,False,False))() ] )
def c_code_cache_version(self):
return ()
def c_code(self, node, nodename, (V,W,b,d), outputs, sub):
fail = sub['fail']
H = outputs[0]
codeSource = """
///////////// < code generated by GpuConv3D >
......@@ -51,31 +51,31 @@ class GpuConv3D(theano.Op):
//Check dimensionality of inputs
if (%(W)s->nd != 5)
{
PyErr_Format(PyExc_ValueError, "GpuConv3D: W must be a 5 dimensional CudaNdarray");
PyErr_Format(PyExc_ValueError, "GpuConv3D: W must be a 5 dimensional CudaNdarray");
%(fail)s
}
if (%(V)s->nd != 5)
{
PyErr_Format(PyExc_ValueError, "GpuConv3D: V must be a 5 dimensional CudaNdarray");
PyErr_Format(PyExc_ValueError, "GpuConv3D: V must be a 5 dimensional CudaNdarray");
%(fail)s
}
if (%(b)s->nd != 1)
{
PyErr_Format(PyExc_ValueError, "GpuConv3D: b must be a vector CudaNdarray");
PyErr_Format(PyExc_ValueError, "GpuConv3D: b must be a vector CudaNdarray");
%(fail)s
}
if (%(d)s->nd != 1)
{
PyErr_Format(PyExc_ValueError, "GpuConv3D: d must be a vector CudaNdarray");
PyErr_Format(PyExc_ValueError, "GpuConv3D: d must be a vector CudaNdarray");
%(fail)s
}
if (%(d)s->dimensions[0] != 3)
{
PyErr_Format(PyExc_ValueError, "GpuConv3D: 3 stride length arguments expected (row, col, time) but %%li were given", %(d)s->dimensions[0]);
PyErr_Format(PyExc_ValueError, "GpuConv3D: 3 stride length arguments expected (row, col, time) but %%li were given", %(d)s->dimensions[0]);
%(fail)s
}
......@@ -87,7 +87,7 @@ PyErr_Format(PyExc_ValueError, "GpuConv3D: d must be a vector CudaNdarray");
const int inputChannels = CudaNdarray_HOST_DIMS(%(V)s)[4];
if (CudaNdarray_HOST_DIMS(%(W)s)[4] != inputChannels)
{
PyErr_Format(PyExc_ValueError, "Conv3D: W operates on a %%i channel image but the image has %%i channels",CudaNdarray_HOST_DIMS(%(W)s)[4],inputChannels);
PyErr_Format(PyExc_ValueError, "Conv3D: W operates on a %%i channel image but the image has %%i channels",CudaNdarray_HOST_DIMS(%(W)s)[4],inputChannels);
%(fail)s
}
{ //extra scope so error handler jumps don't cause errors
......@@ -115,14 +115,14 @@ PyErr_Format(PyExc_ValueError, "GpuConv3D: d must be a vector CudaNdarray");
%(fail)s
}
{ // extra scope so fail works
//Read and check stride arguments
const int dr = *(dtype_%(d)s*)PyArray_GETPTR1(%(d)s,0);
const int dc = *(dtype_%(d)s*)PyArray_GETPTR1(%(d)s,1);
const int dt = *(dtype_%(d)s*)PyArray_GETPTR1(%(d)s,2);
if (dr <= 0 || dc <= 0 || dt <= 0)
{
PyErr_Format(PyExc_ValueError, "GpuConv3D: Strides must all be positive but are %%i, %%i, %%i", dr, dc, dt);
PyErr_Format(PyExc_ValueError, "GpuConv3D: Strides must all be positive but are %%i, %%i, %%i", dr, dc, dt);
%(fail)s
}
{ // extra scope so fail works
......@@ -139,16 +139,16 @@ PyErr_Format(PyExc_ValueError, "GpuConv3D: d must be a vector CudaNdarray");
dims[2] = outputWidth;
dims[3] = outputDur;
if(!(%(H)s) || CudaNdarray_HOST_DIMS(%(H)s)[0]!=dims[0] ||
CudaNdarray_HOST_DIMS(%(H)s)[1]!=dims[1] ||
CudaNdarray_HOST_DIMS(%(H)s)[2]!=dims[2] ||
CudaNdarray_HOST_DIMS(%(H)s)[3]!=dims[3] ||
if(!(%(H)s) || CudaNdarray_HOST_DIMS(%(H)s)[0]!=dims[0] ||
CudaNdarray_HOST_DIMS(%(H)s)[1]!=dims[1] ||
CudaNdarray_HOST_DIMS(%(H)s)[2]!=dims[2] ||
CudaNdarray_HOST_DIMS(%(H)s)[3]!=dims[3] ||
CudaNdarray_HOST_DIMS(%(H)s)[4]!=dims[4]){
Py_XDECREF(%(H)s);
%(H)s = (CudaNdarray*)CudaNdarray_NewDims(5,dims);
if (!(%(H)s)) {
PyErr_Format(PyExc_MemoryError, "GpuConv3D: could not allocate output");
PyErr_Format(PyExc_MemoryError, "GpuConv3D: could not allocate output");
%(fail)s
}
}
......@@ -180,7 +180,7 @@ if(out_contiguous && !b_strided && (version==0||version==-1) && outputDur<=512 &
//conv_rows_stack
dim3 grid(outputHeight*outputWidth,batchSize*outputChannels);
dim3 threads(outputDur);
int shared_size=0;
conv_rows_stack<<<grid, threads, shared_size>>>(
CudaNdarray_DEV_DATA(%(V)s), CudaNdarray_DEV_DATA(%(W)s), CudaNdarray_DEV_DATA(%(b)s), CudaNdarray_DEV_DATA(%(H)s),
......@@ -193,7 +193,7 @@ if(out_contiguous && !b_strided && (version==0||version==-1) && outputDur<=512 &
CNDA_THREAD_SYNC;
cudaError_t sts = cudaGetLastError();
if (cudaSuccess == sts)
if (cudaSuccess == sts)
{
work_complete = true;
if (verbose>1) printf("threads.x=%%i, threads.y=%%i, grid.x=%%i, grid.y=%%i, shared_size=%%i, nb_threads=%%i\\n", threads.x, threads.y, grid.x, grid.y, shared_size, threads.x * threads.y);
......@@ -206,7 +206,7 @@ if(out_contiguous && !b_strided && (version==0||version==-1) && outputDur<=512 &
PyErr_Format(PyExc_RuntimeError, "ERROR: all implementations failed for GpuConv3D! (%%s)",
cudaGetErrorString(sts));
%(fail)s
}
}
}
......@@ -216,10 +216,10 @@ if(!work_complete){
%(fail)s
}
}}}}}}} //extra scope so error handler jumps don't cross declarations
}}}}}}} //extra scope so error handler jumps don't cross declarations
///////////// < /code generated by GpuConv3D >
"""
return strutil.renderString(codeSource,locals())
return strutil.renderString(codeSource,locals())
def c_support_code_apply(self, node, nodename):
# This code is not sensitive to the ignore_border flag.
......@@ -260,7 +260,7 @@ conv_rows_stack( float* img, float* kern, float* bias, float* out,
for (int k =0; k < kern_height; k++) {
for (int l = 0; l < kern_wid; l++) {
for (int m = 0; m < kern_dur; m++) {
sum += img[img_stride_ochannel*z+img_stride_row*k+img_stride_col*l+img_stride_frame*m] *
sum += img[img_stride_ochannel*z+img_stride_row*k+img_stride_col*l+img_stride_frame*m] *
kern[kern_stride_stack*z+kern_stride_row*k+kern_stride_col*l+kern_stride_frame*m];
}
}
......@@ -278,7 +278,7 @@ conv_rows_stack( float* img, float* kern, float* bias, float* out,
"""
return codeSource#renderString(codeSource,locals())
return codeSource#renderString(codeSource,locals())
gpu_convd = GpuConv3D()
......
theano/sandbox/cuda/GpuConvGrad3D.py
浏览文件 @ 2681accc
......@@ -22,14 +22,14 @@ class GpuConvGrad3D(theano.Op):
WShape_ = T.as_tensor_variable(WShape)
dCdH_ = as_cuda_ndarray_variable(dCdH)
return theano.Apply(self, inputs=[V_, d_, WShape_, dCdH_],
return theano.Apply(self, inputs=[V_, d_, WShape_, dCdH_],
outputs = [ CudaNdarrayType(dtype=V_.dtype, broadcastable=(False,)*5)()])
def perform_(self, node, inputs, output_storage):
V, d, WShape, dCdH = inputs
print "GpuConvGrad3D python code (warning not updated to new format)"
#partial C / partial W[j,z,k,l,m] = sum_i sum_p sum_q sum_r (partial C /partial H[i,j,p,q,r] ) * V[i,z,dr*p+k,dc*q+l,dt*r+m]
#partial C / partial W[j,z,k,l,m] = sum_i sum_p sum_q sum_r (partial C /partial H[i,j,p,q,r] ) * V[i,z,dr*p+k,dc*q+l,dt*r+m]
batchSize = dCdH.shape[0]
outputFilters = dCdH.shape[1]
......@@ -66,7 +66,7 @@ class GpuConvGrad3D(theano.Op):
dCdW = outputs[0]
codeSource = """
codeSource = """
///////////// < code generated by GpuConvGrad3D >
//printf("\t\t\t\tGpuConvGrad3DW c code\\n");
......@@ -123,9 +123,9 @@ class GpuConvGrad3D(theano.Op):
const int inputChannels = CudaNdarray_HOST_DIMS(%(V)s)[4];
if (WShape[4] != inputChannels)
{
PyErr_Format(PyExc_ValueError, "ConvGrad3D: W operates on a %%d channel image but the image has %%d channels",WShape[4],inputChannels);
PyErr_Format(PyExc_ValueError, "ConvGrad3D: W operates on a %%d channel image but the image has %%d channels",WShape[4],inputChannels);
%(fail)s
}
{ //extra scope so fail works
const int filterHeight = WShape[1];
......@@ -149,7 +149,7 @@ class GpuConvGrad3D(theano.Op):
PyErr_Format(PyExc_ValueError, "GpuConvGrad3D: W has a duration of %%i but V is only %%i pixels long", filterWidth, vidWidth);
%(fail)s
}
{ // extra scope so fail works
//Read and check stride arguments
const int dr = *(dtype_%(d)s*)PyArray_GETPTR1(%(d)s,0);
......@@ -167,7 +167,7 @@ class GpuConvGrad3D(theano.Op):
const int outputWidth = int( (vidWidth - filterWidth) / dc )+1;
const int outputDur = int( (vidDur - filterDur) / dt ) +1;
if (CudaNdarray_HOST_DIMS(%(dCdH)s)[0] != batchSize ||
if (CudaNdarray_HOST_DIMS(%(dCdH)s)[0] != batchSize ||
CudaNdarray_HOST_DIMS(%(dCdH)s)[4] != outputChannels ||
CudaNdarray_HOST_DIMS(%(dCdH)s)[1] != outputHeight ||
CudaNdarray_HOST_DIMS(%(dCdH)s)[2] != outputWidth ||
......@@ -185,10 +185,10 @@ class GpuConvGrad3D(theano.Op):
dims[2] = filterWidth;
dims[3] = filterDur;
if(!(%(dCdW)s) || CudaNdarray_HOST_DIMS(%(dCdW)s)[0]!=dims[0] ||
CudaNdarray_HOST_DIMS(%(dCdW)s)[1]!=dims[1] ||
CudaNdarray_HOST_DIMS(%(dCdW)s)[2]!=dims[2] ||
CudaNdarray_HOST_DIMS(%(dCdW)s)[3]!=dims[3] ||
if(!(%(dCdW)s) || CudaNdarray_HOST_DIMS(%(dCdW)s)[0]!=dims[0] ||
CudaNdarray_HOST_DIMS(%(dCdW)s)[1]!=dims[1] ||
CudaNdarray_HOST_DIMS(%(dCdW)s)[2]!=dims[2] ||
CudaNdarray_HOST_DIMS(%(dCdW)s)[3]!=dims[3] ||
CudaNdarray_HOST_DIMS(%(dCdW)s)[4]!=dims[4] ){
Py_XDECREF(%(dCdW)s);
%(dCdW)s = (CudaNdarray*)CudaNdarray_NewDims(5,dims);
......@@ -219,7 +219,7 @@ if(out_contiguous && (version==0||version==-1) && WShape[4]<=512 && !work_comple
//conv_rows_stack
dim3 grid(WShape[0]*WShape[4],WShape[1]*WShape[2]);//outputHeight*outputWidth);
dim3 threads(WShape[3]);
int shared_size=0;
convgrad_rows_stack<<<grid, threads, shared_size>>>(
......@@ -235,7 +235,7 @@ if(out_contiguous && (version==0||version==-1) && WShape[4]<=512 && !work_comple
CNDA_THREAD_SYNC;
cudaError_t sts = cudaGetLastError();
if (cudaSuccess == sts)
if (cudaSuccess == sts)
{
work_complete = true;
if (verbose>1) printf("threads.x=%%i, threads.y=%%i, grid.x=%%i, grid.y=%%i, shared_size=%%i, nb_threads=%%i\\n", threads.x, threads.y, grid.x, grid.y, shared_size, threads.x * threads.y);
......@@ -248,18 +248,18 @@ if(out_contiguous && (version==0||version==-1) && WShape[4]<=512 && !work_comple
PyErr_Format(PyExc_RuntimeError, "ERROR: all implementations failed for GpuConvGrad3D! (%%s)",
cudaGetErrorString(sts));
%(fail)s
}
}
}
if(!work_complete){
PyErr_Format(PyExc_RuntimeError, "ERROR: no implementations executed for this GpuConv3D!");
%(fail)s
}
}}}}} // extra scope for fail
}}}}} // extra scope for fail
///////////// < /code generated by GpuConvGrad3D >
"""
"""
return strutls.render_string(codeSource,locals())
return strutls.render_string(codeSource,locals())
def c_support_code_apply(self, node, nodename):
# This code is not sensitive to the ignore_border flag.
......@@ -329,7 +329,7 @@ convgrad_rows_stack( float* img, float* dCdH, float* dCdW,
dCdW[j,z,k,l,m] += dCdH[i,j,p,q,r] * V[i,z,dr*p+k,dc*q+l,dt*r+m]
*/
"""
return codeSource#renderString(codeSource,locals())
return codeSource#renderString(codeSource,locals())
gpu_conv_grad3d = GpuConvGrad3D()
......
theano/sandbox/cuda/GpuConvTransp3D.py
浏览文件 @ 2681accc
......@@ -14,10 +14,10 @@ class GpuConvTransp3D(theano.Op):
""" The gpu version of ConvTransp3D """
def __eq__(self,other):
return type(self) == type(other)
def __hash__(self):
return hash(type(self))
def make_node(self, W, b, d, H, RShape = None):
W_ = as_cuda_ndarray_variable(W)
b_ = as_cuda_ndarray_variable(b)
......@@ -27,9 +27,9 @@ class GpuConvTransp3D(theano.Op):
RShape_ = T.as_tensor_variable(RShape)
else:
RShape_ = T.as_tensor_variable([-1,-1,-1])
return theano.Apply(self, inputs=[W_,b_,d_,H_, RShape_],
outputs = [CudaNdarrayType(dtype=H_.dtype,
outputs = [CudaNdarrayType(dtype=H_.dtype,
broadcastable=(False,)*5)()])
def infer_shape(self, node, input_shapes):
......@@ -51,7 +51,7 @@ class GpuConvTransp3D(theano.Op):
R = outputs[0]
codeSource = """
codeSource = """
///////////// < code generated by GpuConvTransp3D >
//printf("\t\t\t\tGpuConvTransp c code\\n");
......@@ -59,25 +59,25 @@ class GpuConvTransp3D(theano.Op):
//Check dimensionality of inputs
if (%(H)s->nd != 5)
{
PyErr_Format(PyExc_ValueError, "GpuConvTransp3D: H must be a 5-D tensor but it is %%i-D",%(H)s->nd);
PyErr_Format(PyExc_ValueError, "GpuConvTransp3D: H must be a 5-D tensor but it is %%i-D",%(H)s->nd);
%(fail)s
}
if (%(W)s->nd != 5)
{
PyErr_Format(PyExc_ValueError, "GpuConvTransp3D: W must be a 5-D tensor");
PyErr_Format(PyExc_ValueError, "GpuConvTransp3D: W must be a 5-D tensor");
%(fail)s
}
if (%(b)s->nd != 1)
{
PyErr_Format(PyExc_ValueError, "GpuConvTransp3D: b must be a vector");
PyErr_Format(PyExc_ValueError, "GpuConvTransp3D: b must be a vector");
%(fail)s
}
if (%(d)s->nd != 1)
{
PyErr_Format(PyExc_ValueError, "GpuConvTransp3D: d must be a vector");
PyErr_Format(PyExc_ValueError, "GpuConvTransp3D: d must be a vector");
%(fail)s
}
......@@ -106,7 +106,7 @@ class GpuConvTransp3D(theano.Op):
if (CudaNdarray_HOST_DIMS(%(H)s)[4] != outputChannels)
{
PyErr_Format(PyExc_ValueError, "W produces a %%i channel image but the image has %%i channels. W.shape: (%%i, %%i, %%i,%%i, %%i) H.shape: (%%i, %%i, %%i, %%i, %%i)",outputChannels,CudaNdarray_HOST_DIMS(%(H)s)[4], CudaNdarray_HOST_DIMS(%(W)s)[0], CudaNdarray_HOST_DIMS(%(W)s)[1], CudaNdarray_HOST_DIMS(%(W)s)[2], CudaNdarray_HOST_DIMS(%(W)s)[3], CudaNdarray_HOST_DIMS(%(W)s)[4], CudaNdarray_HOST_DIMS(%(H)s)[0], CudaNdarray_HOST_DIMS(%(H)s)[1], CudaNdarray_HOST_DIMS(%(H)s)[2], CudaNdarray_HOST_DIMS(%(H)s)[3], CudaNdarray_HOST_DIMS(%(H)s)[4]);
PyErr_Format(PyExc_ValueError, "W produces a %%i channel image but the image has %%i channels. W.shape: (%%i, %%i, %%i,%%i, %%i) H.shape: (%%i, %%i, %%i, %%i, %%i)",outputChannels,CudaNdarray_HOST_DIMS(%(H)s)[4], CudaNdarray_HOST_DIMS(%(W)s)[0], CudaNdarray_HOST_DIMS(%(W)s)[1], CudaNdarray_HOST_DIMS(%(W)s)[2], CudaNdarray_HOST_DIMS(%(W)s)[3], CudaNdarray_HOST_DIMS(%(W)s)[4], CudaNdarray_HOST_DIMS(%(H)s)[0], CudaNdarray_HOST_DIMS(%(H)s)[1], CudaNdarray_HOST_DIMS(%(H)s)[2], CudaNdarray_HOST_DIMS(%(H)s)[3], CudaNdarray_HOST_DIMS(%(H)s)[4]);
%(fail)s
}
{ // for fail
......@@ -173,10 +173,10 @@ class GpuConvTransp3D(theano.Op):
dims[2] = videoWidth;
dims[3] = videoDur;
if(!(%(R)s) || CudaNdarray_HOST_DIMS(%(R)s)[0]!=dims[0] ||
CudaNdarray_HOST_DIMS(%(R)s)[1]!=dims[1] ||
CudaNdarray_HOST_DIMS(%(R)s)[2]!=dims[2] ||
CudaNdarray_HOST_DIMS(%(R)s)[3]!=dims[3] ||
if(!(%(R)s) || CudaNdarray_HOST_DIMS(%(R)s)[0]!=dims[0] ||
CudaNdarray_HOST_DIMS(%(R)s)[1]!=dims[1] ||
CudaNdarray_HOST_DIMS(%(R)s)[2]!=dims[2] ||
CudaNdarray_HOST_DIMS(%(R)s)[3]!=dims[3] ||
CudaNdarray_HOST_DIMS(%(R)s)[4]!=dims[4]){
Py_XDECREF(%(R)s);
%(R)s = (CudaNdarray*)CudaNdarray_NewDims(5,dims);
......@@ -213,7 +213,7 @@ if(out_contiguous && (version==0||version==-1) && outputDur<=512 && !work_comple
//conv_transp_rows_stack
dim3 grid(batchSize * inputChannels, videoHeight * videoWidth);
dim3 threads(videoDur);
HERE
int shared_size=0;
......@@ -230,7 +230,7 @@ HERE
CNDA_THREAD_SYNC;
cudaError_t sts = cudaGetLastError();
if (cudaSuccess == sts)
if (cudaSuccess == sts)
{
work_complete = true;
if (verbose>1) printf("threads.x=%%i, threads.y=%%i, grid.x=%%i, grid.y=%%i, shared_size=%%i, nb_threads=%%i\\n", threads.x, threads.y, grid.x, grid.y, shared_size, threads.x * threads.y);
......@@ -243,7 +243,7 @@ HERE
PyErr_Format(PyExc_RuntimeError, "ERROR: all implementations failed for GpuConvTransp3D! (%%s)",
cudaGetErrorString(sts));
%(fail)s
}
}
}
......@@ -258,8 +258,8 @@ if(!work_complete){
}}}}}} // for fail
///////////// < /code generated by GpuConvTransp3D >
"""
return renderString(codeSource,locals())
"""
return renderString(codeSource,locals())
def c_support_code_apply(self, node, nodename):
# This code is not sensitive to the ignore_border flag.
......@@ -358,9 +358,9 @@ def computeR(W,b,d,H,Rshape = None):
assert len(W.shape) == 5
assert len(H.shape) == 5
assert len(b.shape) == 1
assert len(d) == 3
assert len(d) == 3
outputChannels, inputChannels, filterHeight, filterWidth, filterDur = W.shape
batchSize, outputChannelsAgain, outputHeight, outputWidth, outputDur = H.shape
assert outputChannelsAgain == outputChannels
......@@ -382,18 +382,18 @@ def computeR(W,b,d,H,Rshape = None):
assert Rshape[1] >= videoWidth
assert Rshape[2] >= videoDur
#print "setting video size to Rshape = "+str(Rshape)
#print "setting video size to Rshape = "+str(Rshape)
videoHeight, videoWidth, videoDur = Rshape
#else:
# print "No Rshape passed in"
#print "video size: "+str((videoHeight, videoWidth, videoDur))
#print "video size: "+str((videoHeight, videoWidth, videoDur))
R = N.zeros( (batchSize, inputChannels, videoHeight,
videoWidth, videoDur ) , dtype=H.dtype)
#R[i,j,r,c,t] = b_j + sum_{rc,rk | d \circ rc + rk = r} sum_{cc,ck | ...} sum_{tc,tk | ...} sum_k W[k, j, rk, ck, tk] * H[i,k,rc,cc,tc]
#R[i,j,r,c,t] = b_j + sum_{rc,rk | d \circ rc + rk = r} sum_{cc,ck | ...} sum_{tc,tk | ...} sum_k W[k, j, rk, ck, tk] * H[i,k,rc,cc,tc]
for i in xrange(0,batchSize):
#print '\texample '+str(i+1)+'/'+str(batchSize)
for j in xrange(0,inputChannels):
......@@ -403,7 +403,7 @@ def computeR(W,b,d,H,Rshape = None):
for c in xrange(0,videoWidth):
for t in xrange(0,videoDur):
R[i,j,r,c,t] = b[j]
ftc = max([0, int(N.ceil(float(t-filterDur +1 )/float(dt))) ])
fcc = max([0, int(N.ceil(float(c-filterWidth +1)/float(dc))) ])
......@@ -424,7 +424,7 @@ def computeR(W,b,d,H,Rshape = None):
tk = t - tc * dt
if tk < 0:
break
R[i,j,r,c,t] += N.dot(W[:,j,rk,ck,tk], H[i,:,rc,cc,tc] )
tc += 1
......@@ -439,7 +439,7 @@ def computeR(W,b,d,H,Rshape = None):
"" #close loop over r
"" #close loop over j
"" #close loop over i
return R
......
theano/tensor/nnet/Conv3D.py
浏览文件 @ 2681accc
......@@ -27,15 +27,15 @@ from theano.tensor.blas import ldflags
#partial C / partial b_j = sum_i sum_k sum_r sum_c sum_t (partial C / partial H[i,r,c,t,k] ) * ( partial H[i,r,c,t,k] / partial b_j )
# = sum_i sum_k sum_r sum_c sum_t (partial C / partial H[i,r,c,t,k] ) * delta(k = j)
# = sum_i sum_r sum_c sum_t (partial C / partial H[i,r,c,t,j] )
# = sum_i sum_r sum_c sum_t (partial C / partial H[i,r,c,t,j] )
#partial C / partial W[j,k,l,m,z] = sum_i sum_n sum_p sum_q sum_r (partial C /partial H[i,p,q,r,n] ) * (partial H[i,p,q,r,n] / partial W[j,k,l,m,z])
# = partial C / partial W[j,k,l,m,z] = sum_i sum_n sum_p sum_q sum_r (partial C /partial H[i,p,q,r,n] ) *
# = partial C / partial W[j,k,l,m,z] = sum_i sum_n sum_p sum_q sum_r (partial C /partial H[i,p,q,r,n] ) *
# (partial sum_s sum_u sum_v sum_a W[n,a, s,u,v] V[i, dr*p+s,dc*q+u,dt*r+v, a] ) / partial W[j,k,l,m,z])
# = partial C / partial W[j,k,l,m,z] = sum_i sum_p sum_q sum_r (partial C /partial H[i,p,q,r,j] ) *
# = partial C / partial W[j,k,l,m,z] = sum_i sum_p sum_q sum_r (partial C /partial H[i,p,q,r,j] ) *
# (partial sum_s sum_u sum_v sum_a W[j,a, s,u,v] V[i,dr*p+s,dc*q+u,dt*r+v,a] ) / partial W[j,k,l,m,z])
# = partial C / partial W[j,k,l,m,z] = sum_i sum_p sum_q sum_r (partial C /partial H[i,p,q,r,j] ) * V[i,dr*p+k,dc*q+l,dt*r+m,z]
# = partial C / partial W[j,k,l,m,z] = sum_i sum_p sum_q sum_r (partial C /partial H[i,p,q,r,j] ) * V[i,dr*p+k,dc*q+l,dt*r+m,z]
#derivatives wrt V unimplemented for now. derivatives wrt dr, dc, dt are undefined since dr, dc, dt are natural numbers.
......@@ -76,7 +76,7 @@ class Conv3D(theano.Op):
#print "dCdH.broadcastable"
#quit(-1)
#dCdH = printing.Print("dCdH = ",["shape"])
dCdV = ConvTransp3D.convTransp3D(W, T.zeros_like(V[0,0,0,0,:]), d, dCdH, V.shape[1:4] )
WShape = W.shape
dCdW = ConvGrad3D.convGrad3D(V,d,WShape,dCdH)
......@@ -92,7 +92,7 @@ class Conv3D(theano.Op):
def infer_shape(self, node, input_shapes):
V,W,b,d = node.inputs
V_shape, W_shape, b_shape, d_shape = input_shapes
dr = d[0]
dc = d[1]
dt = d[2]
......@@ -103,12 +103,12 @@ class Conv3D(theano.Op):
vidWidth = V_shape[2]
filterWidth = W_shape[2]
vidDur = V_shape[3]
filterDur = W_shape[3]
filterDur = W_shape[3]
output_height = T.floor( (vidHeight - filterHeight) / dr )+1
output_width = T.floor( (vidWidth - filterWidth) / dc )+1
output_dur = T.floor( (vidDur - filterDur) / dt ) +1
return [(batch_size, output_height, output_width, output_dur, output_channels )]
......@@ -133,9 +133,9 @@ class Conv3D(theano.Op):
fail = sub['fail']
H = outputs[0]
codeSource = """
codeSource = """
///////////// < code generated by Conv3D >
//printf("\t\t\t\tConv3D c code\\n");
......@@ -143,14 +143,14 @@ class Conv3D(theano.Op):
//Check dimensionality of inputs
if (%(W)s->nd != 5)
{
PyErr_Format(PyExc_ValueError, "Conv3D: W must be a 5 dimensional tensor");
PyErr_Format(PyExc_ValueError, "Conv3D: W must be a 5 dimensional tensor");
%(fail)s
}
if (%(V)s->nd != 5)
{
PyErr_Format(PyExc_ValueError, "Conv3D: V must be a 5 dimensional tensor");
PyErr_Format(PyExc_ValueError, "Conv3D: V must be a 5 dimensional tensor");
%(fail)s
}
......@@ -180,13 +180,13 @@ class Conv3D(theano.Op):
if (%(W)s->dimensions[4] != inputChannels)
{
PyErr_Format(PyExc_ValueError, "Conv3D: W operates on a %%li channel image but the image has %%i channels.",%(W)s->dimensions[4],inputChannels);
PyErr_Format(PyExc_ValueError, "Conv3D: W operates on a %%li channel image but the image has %%i channels.",%(W)s->dimensions[4],inputChannels);
%(fail)s
}
if (%(b)s->dimensions[0] != outputChannels)
{
PyErr_Format(PyExc_ValueError, "Conv3D: b adds to a(n) %%li channel output image but the output has %%i channels",%(b)s->dimensions[0],outputChannels);
PyErr_Format(PyExc_ValueError, "Conv3D: b adds to a(n) %%li channel output image but the output has %%i channels",%(b)s->dimensions[0],outputChannels);
%(fail)s
}
......@@ -221,7 +221,7 @@ class Conv3D(theano.Op):
}
{ // extra scope so fail works
//Read and check stride arguments
const int dr = *(dtype_%(d)s*) PyArray_GETPTR1(%(d)s,0);
const int dc = *(dtype_%(d)s*) PyArray_GETPTR1(%(d)s,1);
......@@ -249,10 +249,10 @@ class Conv3D(theano.Op):
if(!(%(H)s) || %(H)s->dimensions[0]!=dims[0] ||
%(H)s->dimensions[1]!=dims[1] ||
%(H)s->dimensions[2]!=dims[2] ||
%(H)s->dimensions[3]!=dims[3] ||
if(!(%(H)s) || %(H)s->dimensions[0]!=dims[0] ||
%(H)s->dimensions[1]!=dims[1] ||
%(H)s->dimensions[2]!=dims[2] ||
%(H)s->dimensions[3]!=dims[3] ||
%(H)s->dimensions[4]!=dims[4]){
Py_XDECREF(%(H)s);
%(H)s = (PyArrayObject *) PyArray_SimpleNew(5, dims, %(V)s->descr->type_num);
......@@ -287,11 +287,11 @@ class Conv3D(theano.Op):
// ex: filterDur == 1 && batchSize == 1 && dt = 1 (for SFA)
// ex: inputChannels == 1 """
#if the data types are not mixed, we can insert special case optimizations based on BLAS
......@@ -326,7 +326,7 @@ class Conv3D(theano.Op):
long long Hposi = Hpos;
long long Vposi = Vpos;
for (int r = 0; r < outputHeight; r++) {
long long Hposr = Hpos;
long long Vposr = Vpos;
......@@ -357,7 +357,7 @@ class Conv3D(theano.Op):
dtype_%(H)s * writePos = & ELEM_AT(%(H)s,Hpos);
for (int k =0; k < filterHeight; k++) {
int Wposk = Wpos;
long long Vposk = Vpos;
......@@ -368,7 +368,7 @@ class Conv3D(theano.Op):
//H[i,r,c,t,:] += N.dot(W[:,k,l,m,:],V[i,dr*r+k,dc*c+l,dt*t+m,:])
//note: changing the weights so that outputChannels and inputChannels were the last two rather than
//the first and last elements did not speed this up, even for extremely large input sizes
......@@ -395,11 +395,11 @@ class Conv3D(theano.Op):
Hpos = Hposr + %(H)s->strides[1];
Vpos = Vposr + %(V)s->strides[1] * dr;
} //closes r
Hpos = Hposi + %(H)s->strides[0];
Hpos = Hposi + %(H)s->strides[0];
Vpos = Vposi + %(V)s->strides[0];
} //closes i
} //closes "lots of channels" special case code
else
"""
......@@ -414,7 +414,7 @@ class Conv3D(theano.Op):
long long Hposi = Hpos;
long long Vposi = Vpos;
for (int r = 0; r < outputHeight; r++) {
long long Hposr = Hpos;
long long Vposr = Vpos;
......@@ -441,14 +441,14 @@ class Conv3D(theano.Op):
long long Hposj = Hpos;
long long Vposj = Vpos;
int Wposj = Wpos;
// H[i,r,c,t,j] = b[j]
dtype_%(H)s & writePos = ELEM_AT(%(H)s,Hpos);
writePos = ELEM_AT(%(b)s,bPos);
for (int k =0; k < filterHeight; k++) {
int Wposk = Wpos;
......@@ -462,9 +462,9 @@ class Conv3D(theano.Op):
for (int z = 0; z < inputChannels; z++) {
//H[i,r,c,t,j] += W[j,z,k,l,m] * V[i,dr*r+k, dc*c+l, dt*t+m,z]
writePos += ELEM_AT(%(W)s,Wpos) * ELEM_AT(%(V)s,Vpos);
Wpos += ws4;
Vpos += vs4;
} // close z
......@@ -478,7 +478,7 @@ class Conv3D(theano.Op):
Vpos = Vposk + %(V)s->strides[1];
} //close k
bPos += bs;
Wpos = Wposj + ws0;
Hpos = Hposj + hs4;
......@@ -495,15 +495,15 @@ class Conv3D(theano.Op):
Hpos = Hposr + %(H)s->strides[1];
Vpos = Vposr + %(V)s->strides[1] * dr;
} //closes r
Hpos = Hposi + %(H)s->strides[0];
Hpos = Hposi + %(H)s->strides[0];
Vpos = Vposi + %(V)s->strides[0];
} //closes i
} //closes general case code
}}}}}}} //extra scope so error handler jumps don't cross declarations
}}}}}}} //extra scope so error handler jumps don't cross declarations
///////////// < /code generated by Conv3D >
"""
"""
return strutil.renderString(codeSource,locals())
return strutil.renderString(codeSource,locals())
global conv3D
conv3D = Conv3D()
......@@ -514,7 +514,7 @@ def computeH(V,W,b,d):
if len(b.shape) != 1:
print b.shape
assert False
assert len(d) == 3
assert len(d) == 3
batchSize = V.shape[0]
outputChannels = W.shape[0]
......@@ -539,7 +539,7 @@ def computeH(V,W,b,d):
outputWidth = int( (vidWidth - filterWidth) / dy )+1
outputDur = int( (vidDur - filterDur) / dt ) +1
H = N.zeros( (batchSize, outputHeight,
outputWidth, outputDur, outputChannels ), dtype=V.dtype )
......@@ -563,7 +563,7 @@ def computeH(V,W,b,d):
v = V[i,d[0]*x+k, d[1]*y+l, d[2]*t+m,z]
#if i == 0 and x == 0 and y == 0 and t == 0 and j == 0:
# print 'setting H[0] += '+str(w*v)+' W['+str((j,z,k,l,m))+']='+str(w)+' V['+str((i,d[0]*x+k,d[1]*y+l,d[2]*t+m,z))+']='+str(v)
H[i,x,y,t,j] += w * v
H[i,x,y,t,j] += w * v
return H
......
theano/tensor/nnet/ConvGrad3D.py
浏览文件 @ 2681accc
......@@ -33,14 +33,14 @@ class ConvGrad3D(theano.Op):
dLdd = None #not differentiable, since d is not continuous
dLdWShape = None #not differentiable, since d is not continuous
dLdB = conv3D( C, dLdA, T.zeros_like(B[0,0,0,0,:]), d)
return [ dLdC, dLdd, dLdWShape, dLdB ]
def perform(self, node, inputs, output_storage):
V, d, WShape, dCdH = inputs
print "ConvGradW3D python code"
#partial C / partial W[j,z,k,l,m] = sum_i sum_p sum_q sum_r (partial C /partial H[i,j,p,q,r] ) * V[i,z,dr*p+k,dc*q+l,dt*r+m]
#partial C / partial W[j,z,k,l,m] = sum_i sum_p sum_q sum_r (partial C /partial H[i,j,p,q,r] ) * V[i,z,dr*p+k,dc*q+l,dt*r+m]
batchSize = dCdH.shape[0]
outputFilters = dCdH.shape[4]
......@@ -82,7 +82,7 @@ class ConvGrad3D(theano.Op):
dCdW = outputs[0]
codeSource = """
codeSource = """
///////////// < code generated by ConvGradW3D >
//printf("\t\t\t\tConvGradW3D c code\\n");
......@@ -90,13 +90,13 @@ class ConvGrad3D(theano.Op):
//Check dimensionality of inputs
if (%(dCdH)s->nd != 5)
{
PyErr_Format(PyExc_ValueError, "ConvGrad3D: dCdH must be a 5 dimensional tensor");
PyErr_Format(PyExc_ValueError, "ConvGrad3D: dCdH must be a 5 dimensional tensor");
%(fail)s
}
if (%(V)s->nd != 5)
{
PyErr_Format(PyExc_ValueError, "ConvGrad3D: V must be a 5 dimensional tensor");
PyErr_Format(PyExc_ValueError, "ConvGrad3D: V must be a 5 dimensional tensor");
%(fail)s
}
......@@ -131,16 +131,16 @@ class ConvGrad3D(theano.Op):
PyErr_Format(PyExc_ValueError,"ConvGrad3D: WShape must be contiguous");
%(fail)s
}
{ //extra scope so that fail will not jump over declarations
dtype_%(WShape)s * WShape = (dtype_%(WShape)s *) %(WShape)s->data;
const int outputChannels = WShape[0];
const int inputChannels = %(V)s->dimensions[4];
if (WShape[4] != inputChannels)
{
PyErr_Format(PyExc_ValueError, "ConvGrad3D: W operates on a %%i channel image but the image has %%i channels",(int) WShape[1],inputChannels);
PyErr_Format(PyExc_ValueError, "ConvGrad3D: W operates on a %%i channel image but the image has %%i channels",(int) WShape[1],inputChannels);
%(fail)s
}
{ //extra scope so fail works
const int filterHeight = WShape[1];
......@@ -184,7 +184,7 @@ class ConvGrad3D(theano.Op):
if (%(dCdH)s->dimensions[0] != batchSize ||
if (%(dCdH)s->dimensions[0] != batchSize ||
%(dCdH)s->dimensions[4] != outputChannels ||
%(dCdH)s->dimensions[1] != outputHeight ||
%(dCdH)s->dimensions[2] != outputWidth ||
......@@ -202,10 +202,10 @@ class ConvGrad3D(theano.Op):
dims[2] = filterWidth;
dims[3] = filterDur;
if(!(%(dCdW)s) || %(dCdW)s->dimensions[0]!=dims[0] ||
%(dCdW)s->dimensions[1]!=dims[1] ||
%(dCdW)s->dimensions[2]!=dims[2] ||
%(dCdW)s->dimensions[3]!=dims[3] ||
if(!(%(dCdW)s) || %(dCdW)s->dimensions[0]!=dims[0] ||
%(dCdW)s->dimensions[1]!=dims[1] ||
%(dCdW)s->dimensions[2]!=dims[2] ||
%(dCdW)s->dimensions[3]!=dims[3] ||
%(dCdW)s->dimensions[4]!=dims[4] ){
Py_XDECREF(%(dCdW)s);
%(dCdW)s = (PyArrayObject *) PyArray_SimpleNew(5, dims, %(V)s->descr->type_num);
......@@ -241,9 +241,9 @@ class ConvGrad3D(theano.Op):
for (int p = 0; p < outputHeight; p++) {
for (int q = 0; q < outputWidth; q++) {
int Hpos = i * %(dCdH)s->strides[0] + j * %(dCdH)s->strides[4] + p * %(dCdH)s->strides[1] + q * %(dCdH)s->strides[2] ;
int Vpos = i * %(V)s->strides[0] + z * %(V)s->strides[4] + (dr * p+k) * %(V)s->strides[1] + (dc*q+l) * %(V)s->strides[2] + m * %(V)s->strides[3];
int Vpos = i * %(V)s->strides[0] + z * %(V)s->strides[4] + (dr * p+k) * %(V)s->strides[1] + (dc*q+l) * %(V)s->strides[2] + m * %(V)s->strides[3];
for (int r = 0; r < outputDur; r++) {
for (int r = 0; r < outputDur; r++) {
writePos += ELEM5(%(dCdH)s,i,p,q,r,j) * ELEM5(%(V)s,i,dr*p+k,dc*q+l,dt*r+m,z);
//writePos += ELEM_AT(%(dCdH)s,Hpos) * ELEM_AT(%(V)s,Vpos);
Hpos += dhs3;
......@@ -258,11 +258,11 @@ class ConvGrad3D(theano.Op):
}
}
}}}}}}} // extra scope for fail
}}}}}}} // extra scope for fail
///////////// < /code generated by ConvGradW3D >
"""
"""
return strutil.renderString(codeSource,locals())
return strutil.renderString(codeSource,locals())
convGrad3D = ConvGrad3D()
......
theano/tensor/nnet/ConvTransp3D.py
浏览文件 @ 2681accc
......@@ -8,10 +8,10 @@ class ConvTransp3D(theano.Op):
""" "Transpose" of Conv3D (Conv3D implements multiplication by an implicitly defined matrix W. This implements multiplication by its transpose) """
def __eq__(self,other):
return type(self) == type(other)
def __hash__(self):
return hash(type(self))
def make_node(self, W, b, d, H, RShape = None):
"""
:param W: Weights, filter
......@@ -27,7 +27,7 @@ class ConvTransp3D(theano.Op):
RShape_ = T.as_tensor_variable(RShape)
else:
RShape_ = T.as_tensor_variable([-1,-1,-1])
return theano.Apply(self, inputs=[W_,b_,d_,H_, RShape_], outputs = [ T.TensorType(H_.dtype, (False,False,False,False,False))() ] )
def c_compile_args(self):
......@@ -62,7 +62,7 @@ class ConvTransp3D(theano.Op):
R = outputs[0]
codeSource = """
codeSource = """
///////////// < code generated by ConvTransp3D >
//printf("\t\t\t\tConvTransp3D c code\\n");
......@@ -70,25 +70,25 @@ class ConvTransp3D(theano.Op):
//Check dimensionality of inputs
if (%(H)s->nd != 5)
{
PyErr_Format(PyExc_ValueError, "H must be a 5-D tensor but it is %%i-D",%(H)s->nd);
PyErr_Format(PyExc_ValueError, "H must be a 5-D tensor but it is %%i-D",%(H)s->nd);
%(fail)s
}
if (%(W)s->nd != 5)
{
PyErr_Format(PyExc_ValueError, "ConvTransp3D: W must be a 5-D tensor");
PyErr_Format(PyExc_ValueError, "ConvTransp3D: W must be a 5-D tensor");
%(fail)s
}
if (%(b)s->nd != 1)
{
PyErr_Format(PyExc_ValueError, "ConvTransp3D: b must be a vector");
PyErr_Format(PyExc_ValueError, "ConvTransp3D: b must be a vector");
%(fail)s
}
if (%(d)s->nd != 1)
{
PyErr_Format(PyExc_ValueError, "ConvTransp3D: d must be a vector");
PyErr_Format(PyExc_ValueError, "ConvTransp3D: d must be a vector");
%(fail)s
}
......@@ -103,14 +103,14 @@ class ConvTransp3D(theano.Op):
int dr = *(dtype_%(d)s*)PyArray_GETPTR1(%(d)s,0);
int dc = *(dtype_%(d)s*)PyArray_GETPTR1(%(d)s,1);
int dt = *(dtype_%(d)s*)PyArray_GETPTR1(%(d)s,2);
if (dr <= 0 || dc <= 0 || dt <= 0)
{
PyErr_Format(PyExc_ValueError, "ConvTransp3D: Strides must all be positive but are %%i, %%i, %%i",dr,dc,dt);
PyErr_Format(PyExc_ValueError, "ConvTransp3D: Strides must all be positive but are %%i, %%i, %%i",dr,dc,dt);
%(fail)s
}
//Read and check sizes of inputs
{ // for fail 2
......@@ -119,7 +119,7 @@ class ConvTransp3D(theano.Op):
if (%(H)s->dimensions[4] != outputChannels)
{
PyErr_Format(PyExc_ValueError, "W produces a %%i channel image but the image has %%li channels. W.shape: (%%li, %%li, %%li,%%li, %%li) H.shape: (%%li, %%li, %%li, %%li, %%li)",outputChannels,%(H)s->dimensions[4], %(W)s->dimensions[0], %(W)s->dimensions[1], %(W)s->dimensions[2], %(W)s->dimensions[3], %(W)s->dimensions[4], %(H)s->dimensions[0], %(H)s->dimensions[1], %(H)s->dimensions[2], %(H)s->dimensions[3], %(H)s->dimensions[4]);
PyErr_Format(PyExc_ValueError, "W produces a %%i channel image but the image has %%li channels. W.shape: (%%li, %%li, %%li,%%li, %%li) H.shape: (%%li, %%li, %%li, %%li, %%li)",outputChannels,%(H)s->dimensions[4], %(W)s->dimensions[0], %(W)s->dimensions[1], %(W)s->dimensions[2], %(W)s->dimensions[3], %(W)s->dimensions[4], %(H)s->dimensions[0], %(H)s->dimensions[1], %(H)s->dimensions[2], %(H)s->dimensions[3], %(H)s->dimensions[4]);
%(fail)s
}
......@@ -150,16 +150,16 @@ class ConvTransp3D(theano.Op):
{
if (%(RShape)s->nd != 1)
{
PyErr_Format(PyExc_ValueError, "ConvTransp3D: RShape must be a vector");
PyErr_Format(PyExc_ValueError, "ConvTransp3D: RShape must be a vector");
%(fail)s
}
if (%(RShape)s->dimensions[0] != 3)
{
PyErr_Format(PyExc_ValueError, "RShape must specify a 3D shape ( [height,width,duration] )");
%(fail)s
}
dtype_%(RShape)s RShape0 = *(dtype_%(RShape)s*)PyArray_GETPTR1(%(RShape)s,0);
dtype_%(RShape)s RShape1 = *(dtype_%(RShape)s*)PyArray_GETPTR1(%(RShape)s,1);
dtype_%(RShape)s RShape2 = *(dtype_%(RShape)s*)PyArray_GETPTR1(%(RShape)s,2);
......@@ -188,16 +188,16 @@ class ConvTransp3D(theano.Op):
dims[2] = videoWidth;
dims[3] = videoDur;
if(!(%(R)s) || %(R)s->dimensions[0]!=dims[0] ||
%(R)s->dimensions[1]!=dims[1] ||
%(R)s->dimensions[2]!=dims[2] ||
%(R)s->dimensions[3]!=dims[3] ||
if(!(%(R)s) || %(R)s->dimensions[0]!=dims[0] ||
%(R)s->dimensions[1]!=dims[1] ||
%(R)s->dimensions[2]!=dims[2] ||
%(R)s->dimensions[3]!=dims[3] ||
%(R)s->dimensions[4]!=dims[4])
{
Py_XDECREF(%(R)s);
%(R)s = (PyArrayObject *) PyArray_SimpleNew(5, dims, %(H)s->descr->type_num);
if (!(%(R)s)) {
PyErr_Format(PyExc_MemoryError, "ConvTransp3D: could not allocate R");
PyErr_Format(PyExc_MemoryError, "ConvTransp3D: could not allocate R");
%(fail)s
}
}
......@@ -205,25 +205,25 @@ class ConvTransp3D(theano.Op):
for (int i = 0; i < 3; i++)
if (%(R)s->strides[i] < %(R)s->strides[4])
{
PyErr_Format(PyExc_ValueError, "ConvTransp3D: R must have the smallest stride in its last index, but it doesn't (if this is a problem, the only part of ConvTransp3D that depends on this conditions is the memset, so this is probably easy to fix)");
PyErr_Format(PyExc_ValueError, "ConvTransp3D: R must have the smallest stride in its last index, but it doesn't (if this is a problem, the only part of ConvTransp3D that depends on this conditions is the memset, so this is probably easy to fix)");
%(fail)s
}
{ // for fail 6
memset(%(R)s->data, 0, (batchSize-1) * %(R)s->strides[0]+ inputChannels * %(R)s->strides[4] +
(videoHeight-1) * %(R)s->strides[1] +
(videoWidth-1) * %(R)s->strides[2] +
(videoDur-1) * %(R)s->strides[3]);
#define ELEM5(x, i,j,k,l,m) * ( dtype_ ## x *) ( x->data + (i)*x->strides[0]+(j)*x->strides[1]+(k)*x->strides[2]+(l)*x->strides[3]+(m)*x->strides[4] )
#define ELEM_AT(x, i) * ( dtype_ ## x *) ( x->data + (i) )
dtype_%(b)s * b = (dtype_%(b)s *) %(b)s->data;
int rs4 = %(R)s->strides[4];
......@@ -232,60 +232,60 @@ class ConvTransp3D(theano.Op):
int hs4 = %(H)s->strides[4];
// Compute R
// R[i,r,c,t,j] = b_j + sum_{rc,rk | d \circ rc + rk = r} sum_{cc,ck | ...} sum_{tc,tk | ...} sum_k W[k, rk, ck, tk,j] * H[i,rc,cc,tc,k]
// R[i,r,c,t,j] = b_j + sum_{rc,rk | d \circ rc + rk = r} sum_{cc,ck | ...} sum_{tc,tk | ...} sum_k W[k, rk, ck, tk,j] * H[i,rc,cc,tc,k]
for (int i = 0; i < batchSize; i++) {
for (int r = 0; r < videoHeight; r++) {
const int frc = std::max(0.0, ceil(float(r-filterHeight+1)/float(dr)));
for (int c = 0; c < videoWidth; c++) {
const int fcc = std::max(0.0, ceil(float(c-filterWidth +1)/float(dc)));
for (int t = 0; t < videoDur; t++) {
const int ftc = std::max(0.0, ceil(float(t-filterDur +1) /float(dt)));
const int ftc = std::max(0.0, ceil(float(t-filterDur +1) /float(dt)));
long long Rpost = i * %(R)s->strides[0] + r * %(R)s->strides[1] + c * %(R)s->strides[2] + t * %(R)s->strides[3];
long long Rpos = Rpost;
for (int j = 0; j < inputChannels; j++)
{
{
//ELEM5(%(R)s, i,r,c,t,j) = b[j];
ELEM_AT(%(R)s,Rpos) = b[j];
Rpos += rs4;
}
for (int rc = frc; rc < outputHeight; rc++) {
const int rk = r - rc * dr;
if (rk < 0) break;
for (int cc = fcc; cc < outputWidth; cc++) {
const int ck = c - cc * dc;
if (ck < 0) break;
for (int tc = ftc; tc < outputDur; tc++)
{
const int tk = t - tc * dt;
if (tk < 0) break;
int Wpos = rk * %(W)s->strides[1] + ck * %(W)s->strides[2] + tk * %(W)s->strides[3];
int Hpostc = i * %(H)s->strides[0] + rc * %(H)s->strides[1] + cc * %(H)s->strides[2] + tc * %(H)s->strides[3];
int Hpostc = i * %(H)s->strides[0] + rc * %(H)s->strides[1] + cc * %(H)s->strides[2] + tc * %(H)s->strides[3];
Rpos = Rpost;
for (int j = 0; j < inputChannels; j++)
{
int Wposj = Wpos;
int Wposj = Wpos;
dtype_%(R)s & writePos = ELEM_AT(%(R)s,Rpos);
int Hpos = Hpostc;
for (int k = 0; k < outputChannels; k++) {
for (int k = 0; k < outputChannels; k++) {
//TODO-- it's probably bad in terms of cache that our inner loop is over the largest stride of W.... maybe OK since it's the smallest stride of H
//writePos += ELEM5(%(W)s,k,rk,ck,tk,j) * ELEM5(%(H)s,i,rc,cc,tc,k);
//writePos += ELEM_AT(%(W)s,Wpos) * ELEM_AT(%(H)s,Hpos);
writePos += ELEM_AT(%(W)s,Wpos) * ELEM_AT(%(H)s,Hpos);
writePos += ELEM_AT(%(W)s,Wpos) * ELEM_AT(%(H)s,Hpos);
Wpos += ws0;
Hpos += hs4;
} //close the k loop
Rpos += rs4;
Wpos = Wposj + ws4;
......@@ -304,9 +304,9 @@ class ConvTransp3D(theano.Op):
} //for fail 2
} // for fail 1
///////////// < /code generated by ConvTransp3D >
"""
"""
return strutil.renderString(codeSource,locals())
return strutil.renderString(codeSource,locals())
convTransp3D = ConvTransp3D()
......@@ -316,14 +316,14 @@ def computeR(W,b,d,H,Rshape = None):
assert len(W.shape) == 5
assert len(H.shape) == 5
assert len(b.shape) == 1
assert len(d) == 3
assert len(d) == 3
outputChannels, filterHeight, filterWidth, filterDur, inputChannels = W.shape
batchSize, outputHeight, outputWidth, outputDur, outputChannelsAgain = H.shape
assert outputChannelsAgain == outputChannels
assert b.shape[0] == inputChannels
dr,dc,dt = d
assert dr > 0
......@@ -341,18 +341,18 @@ def computeR(W,b,d,H,Rshape = None):
assert Rshape[1] >= videoWidth
assert Rshape[2] >= videoDur
#print "setting video size to Rshape = "+str(Rshape)
#print "setting video size to Rshape = "+str(Rshape)
videoHeight, videoWidth, videoDur = Rshape
#else:
# print "No Rshape passed in"
#print "video size: "+str((videoHeight, videoWidth, videoDur))
#print "video size: "+str((videoHeight, videoWidth, videoDur))
R = N.zeros( (batchSize, videoHeight,
videoWidth, videoDur, inputChannels ) , dtype=H.dtype)
#R[i,j,r,c,t] = b_j + sum_{rc,rk | d \circ rc + rk = r} sum_{cc,ck | ...} sum_{tc,tk | ...} sum_k W[k, j, rk, ck, tk] * H[i,k,rc,cc,tc]
#R[i,j,r,c,t] = b_j + sum_{rc,rk | d \circ rc + rk = r} sum_{cc,ck | ...} sum_{tc,tk | ...} sum_k W[k, j, rk, ck, tk] * H[i,k,rc,cc,tc]
for i in xrange(0,batchSize):
#print '\texample '+str(i+1)+'/'+str(batchSize)
for j in xrange(0,inputChannels):
......@@ -362,7 +362,7 @@ def computeR(W,b,d,H,Rshape = None):
for c in xrange(0,videoWidth):
for t in xrange(0,videoDur):
R[i,r,c,t,j] = b[j]
ftc = max([0, int(N.ceil(float(t-filterDur +1 )/float(dt))) ])
fcc = max([0, int(N.ceil(float(c-filterWidth +1)/float(dc))) ])
......@@ -383,7 +383,7 @@ def computeR(W,b,d,H,Rshape = None):
tk = t - tc * dt
if tk < 0:
break
R[i,r,c,t,j] += N.dot(W[:,rk,ck,tk,j], H[i,rc,cc,tc,:] )
tc += 1
......@@ -398,7 +398,7 @@ def computeR(W,b,d,H,Rshape = None):
"" #close loop over r
"" #close loop over j
"" #close loop over i
return R
......
Markdown 格式
0% 或
您添加了 0 人 到此讨论。请谨慎行事。
请先完成此评论的编辑！
请 注册 或者 后发表评论