def test_multilayer_conv(self):
# fixed parameters
bsize = 10 # batch size
imshp = (5, 5)
kshp = ((3, 3), (2, 2))
nkerns = (3, 6) # per output pixel
ssizes = (((1, 1), (2, 2)), )
convmodes = ('full', ) #'valid',)
# symbolic stuff
kerns = [tensor.dmatrix(), tensor.dmatrix()]
input = tensor.dmatrix()
rng = numpy.random.RandomState(3423489)
# build actual input images
img2d = numpy.arange(bsize * numpy.prod(imshp)).reshape((bsize, ) +
imshp)
img1d = img2d.reshape(bsize, -1)
for mode in ('FAST_COMPILE', 'FAST_RUN'):
for conv_mode in convmodes:
for ss in ssizes:
l1hid, l1shp = sp.convolve(kerns[0], kshp[0],\
nkerns[0], input, imshp, ss[0], mode=conv_mode)
l1propup = function([kerns[0], input], l1hid, mode=mode)
#l1kernvals = numpy.random.rand(nkerns[0],numpy.prod(kshp[0]))
l1kernvals = numpy.arange(
nkerns[0] * numpy.prod(kshp[0])).reshape(
nkerns[0], numpy.prod(kshp[0]))
l1hidval = l1propup(l1kernvals, img1d)
# actual values
l2hid, l2shp = sp.convolve(kerns[1], kshp[1],\
nkerns[1], l1hid, l1shp, ss[1], mode=conv_mode)
l2propup = function([kerns[1], l1hid], l2hid, mode=mode)
#l2kernvals = numpy.random.rand(nkerns[1],numpy.prod(kshp[1])*nkerns[0])
l2kernvals = numpy.arange(
nkerns[1] * numpy.prod(kshp[1]) * nkerns[0]).reshape(
nkerns[1],
numpy.prod(kshp[1]) * nkerns[0])
# for debugging, we bring things back to integers
l1hidval = numpy.arange(numpy.size(l1hidval)).reshape(
l1hidval.shape)
l2hidval = l2propup(l2kernvals, l1hidval)
def test_multilayer_conv(self):
# fixed parameters
bsize = 10 # batch size
imshp = (5, 5)
kshp = ((3, 3), (2, 2))
nkerns = (3, 6) # per output pixel
ssizes = (((1, 1), (2, 2)), )
convmodes = ('full', ) #'valid',)
# symbolic stuff
kerns = [tensor.dmatrix(), tensor.dmatrix()]
input = tensor.dmatrix()
rng = numpy.random.RandomState(3423489)
# build actual input images
img2d = numpy.arange(
bsize * numpy.prod(imshp)).reshape((bsize, ) + imshp)
img1d = img2d.reshape(bsize, -1)
for mode in ('FAST_COMPILE', 'FAST_RUN'):
for conv_mode in convmodes:
for ss in ssizes:
l1hid, l1shp = sp.convolve(kerns[0], kshp[0],\
nkerns[0], input, imshp, ss[0], mode=conv_mode)
l1propup = function([kerns[0], input], l1hid, mode=mode)
#l1kernvals = numpy.random.rand(nkerns[0],numpy.prod(kshp[0]))
l1kernvals = numpy.arange(
nkerns[0] * numpy.prod(kshp[0])).reshape(
nkerns[0], numpy.prod(kshp[0]))
l1hidval = l1propup(l1kernvals, img1d)
# actual values
l2hid, l2shp = sp.convolve(kerns[1], kshp[1],\
nkerns[1], l1hid, l1shp, ss[1], mode=conv_mode)
l2propup = function([kerns[1], l1hid], l2hid, mode=mode)
#l2kernvals = numpy.random.rand(nkerns[1],numpy.prod(kshp[1])*nkerns[0])
l2kernvals = numpy.arange(
nkerns[1] * numpy.prod(kshp[1]) * nkerns[0]).reshape(
nkerns[1],
numpy.prod(kshp[1]) * nkerns[0])
# for debugging, we bring things back to integers
l1hidval = numpy.arange(numpy.size(l1hidval)).reshape(
l1hidval.shape)
l2hidval = l2propup(l2kernvals, l1hidval)
def test_convolution(self):
print '\n\n*************************************************'
print ' TEST CONVOLUTION'
print '*************************************************'
# fixed parameters
bsize = 10 # batch size
imshp = (28, 28)
kshp = (5, 5)
nkern = 5
ssizes = ((1, 1), (2, 2), (3, 3), (4, 4))
convmodes = ('full', 'valid')
# symbolic stuff
bias = tensor.dvector()
kerns = tensor.dmatrix()
input = tensor.dmatrix()
rng = numpy.random.RandomState(3423489)
filters = rng.randn(nkern, numpy.prod(kshp))
biasvals = rng.randn(nkern)
for mode in ('FAST_COMPILE', 'FAST_RUN'): #, profmode):
ttot, ntot = 0, 0
for conv_mode in convmodes:
for ss in ssizes:
output, outshp = sp.convolve(kerns, kshp, nkern, input,\
imshp, ss, bias=bias, mode=conv_mode)
f = function([kerns, bias, input], output, mode=mode)
# now test with real values
img2d = numpy.arange(
bsize * numpy.prod(imshp)).reshape((bsize, ) + imshp)
img1d = img2d.reshape(bsize, -1)
# create filters (need to be flipped to use convolve2d)
filtersflipped = numpy.zeros((nkern, ) + kshp)
for k in range(nkern):
it = reversed(filters[k, :])
for i in range(kshp[0]):
for j in range(kshp[1]):
filtersflipped[k, i, j] = it.next()
# compute output with convolve2d
if conv_mode == 'valid':
fulloutshp = numpy.array(imshp) - numpy.array(kshp) + 1
else:
fulloutshp = numpy.array(imshp) + numpy.array(kshp) - 1
ntime1 = time.time()
refout = numpy.zeros((bsize, ) + tuple(fulloutshp) +
(nkern, ))
for b in range(bsize):
for n in range(nkern):
refout[b,...,n] = convolve2d(\
img2d[b,:,:], filtersflipped[n,...],conv_mode)
ntot += time.time() - ntime1
# need to flatten images
bench1 = refout[:, 0::ss[0],
0::ss[1], :].reshape(bsize, -1, nkern)
bench1 += biasvals.reshape(1, 1, nkern)
# swap the last two dimensions (output needs to be nkern x outshp)
bench1 = numpy.swapaxes(bench1, 1, 2)
ttime1 = time.time()
out1 = f(filters, biasvals, img1d)
ttot += time.time() - ttime1
temp = bench1.flatten() - out1.flatten()
assert (temp < 1e-5).all()
# test downward propagation -- symbolic stuff
#vis = tensor.grad(output, input, output)
#downprop = function([kerns,input], vis, mode=mode)
#visval = downprop(filters,img1d)
## test downward propagation -- reference implementation
#pshape = (img1d.shape[0],numpy.prod(outshp[1:]),numpy.prod(kshp))
#patchstack = numpy.zeros(pshape)
#for bi in numpy.arange(pshape[0]): # batch index
#abspos = 0
#for outy in numpy.arange(outshp[1]):
#for outx in numpy.arange(outshp[2]):
#for ni in numpy.arange(nkern):
#print 'filters[n,:].shape = ', filters[n,:].shape
#print 'out1[bi,abspos].shape =',out1[bi,abspos].shape
#patchstack[bi,abspos,:] = filters[n,:]*out1[bi,abspos]
#abspos+=1
#patchstack = patchstack.reshape(1,-1)
#indices, indptr, spmat_shape, sptype, outshp = \
#sp.convolution_indices.conv_eval(imshp,kshp,ss,conv_mode)
#spmat = sparse.csc_matrix((numpy.ones_like(indices),indices,indptr),spmat_shape)
#visref = numpy.dot(patchstack, spmat.todense())
#print 'visval = ', visval
#print 'visref = ', visref
#assert numpy.all(visref==visval)
print '**** Convolution Profiling Results (', mode, ') ****'
print 'Numpy processing time: ', ntot
print 'Theano processing time: ', ttot
def test_convolution(self):
# print '\n\n*************************************************'
# print ' TEST CONVOLUTION'
# print '*************************************************'
# fixed parameters
bsize = 10 # batch size
imshp = (28, 28)
kshp = (5, 5)
nkern = 5
ssizes = ((1, 1), (2, 2), (3, 3), (4, 4))
convmodes = ("full", "valid")
# symbolic stuff
bias = tensor.dvector()
kerns = tensor.dmatrix()
input = tensor.dmatrix()
rng = numpy.random.RandomState(3423489)
filters = rng.randn(nkern, numpy.prod(kshp))
biasvals = rng.randn(nkern)
for mode in ("FAST_COMPILE", "FAST_RUN"): # , profmode):
ttot, ntot = 0, 0
for conv_mode in convmodes:
for ss in ssizes:
output, outshp = sp.convolve(kerns, kshp, nkern, input, imshp, ss, bias=bias, mode=conv_mode)
f = function([kerns, bias, input], output, mode=mode)
# now test with real values
img2d = numpy.arange(bsize * numpy.prod(imshp)).reshape((bsize,) + imshp)
img1d = img2d.reshape(bsize, -1)
# create filters (need to be flipped to use convolve2d)
filtersflipped = numpy.zeros((nkern,) + kshp)
for k in range(nkern):
it = reversed(filters[k, :])
for i in range(kshp[0]):
for j in range(kshp[1]):
filtersflipped[k, i, j] = it.next()
# compute output with convolve2d
if conv_mode == "valid":
fulloutshp = numpy.array(imshp) - numpy.array(kshp) + 1
else:
fulloutshp = numpy.array(imshp) + numpy.array(kshp) - 1
ntime1 = time.time()
refout = numpy.zeros((bsize,) + tuple(fulloutshp) + (nkern,))
for b in range(bsize):
for n in range(nkern):
refout[b, ..., n] = convolve2d(img2d[b, :, :], filtersflipped[n, ...], conv_mode)
ntot += time.time() - ntime1
# need to flatten images
bench1 = refout[:, 0 :: ss[0], 0 :: ss[1], :].reshape(bsize, -1, nkern)
bench1 += biasvals.reshape(1, 1, nkern)
# swap the last two dimensions (output needs to be nkern x outshp)
bench1 = numpy.swapaxes(bench1, 1, 2)
ttime1 = time.time()
out1 = f(filters, biasvals, img1d)
ttot += time.time() - ttime1
temp = bench1.flatten() - out1.flatten()
assert (temp < 1e-5).all()
def test_convolution(self):
# print '\n\n*************************************************'
# print ' TEST CONVOLUTION'
# print '*************************************************'
# fixed parameters
bsize = 10 # batch size
imshp = (28, 28)
kshp = (5, 5)
nkern = 5
ssizes = ((1, 1), (2, 2), (3, 3), (4, 4))
convmodes = ('full', 'valid')
# symbolic stuff
bias = tensor.dvector()
kerns = tensor.dmatrix()
input = tensor.dmatrix()
rng = numpy.random.RandomState(3423489)
filters = rng.randn(nkern, numpy.prod(kshp))
biasvals = rng.randn(nkern)
for mode in ('FAST_COMPILE', 'FAST_RUN'):
ttot, ntot = 0, 0
for conv_mode in convmodes:
for ss in ssizes:
output, outshp = sp.convolve(kerns, kshp, nkern, input,\
imshp, ss, bias=bias, mode=conv_mode)
f = function([kerns, bias, input], output, mode=mode)
# now test with real values
img2d = numpy.arange(bsize * numpy.prod(imshp)).reshape(( \
bsize,) + imshp)
img1d = img2d.reshape(bsize, -1)
# create filters (need to be flipped to use convolve2d)
filtersflipped = numpy.zeros((nkern, ) + kshp)
for k in range(nkern):
it = reversed(filters[k, :])
for i in range(kshp[0]):
for j in range(kshp[1]):
filtersflipped[k, i, j] = next(it)
# compute output with convolve2d
if conv_mode == 'valid':
fulloutshp = numpy.array(imshp) - numpy.array(kshp) + 1
else:
fulloutshp = numpy.array(imshp) + numpy.array(kshp) - 1
ntime1 = time.time()
refout = numpy.zeros((bsize, ) + tuple(fulloutshp) +
(nkern, ))
for b in range(bsize):
for n in range(nkern):
refout[b, ...,
n] = convolve2d(img2d[b, :, :],
filtersflipped[n, ...],
conv_mode)
ntot += time.time() - ntime1
# need to flatten images
bench1 = refout[:, 0::ss[0],
0::ss[1], :].reshape(bsize, -1, nkern)
bench1 += biasvals.reshape(1, 1, nkern)
# swap the last two dimensions (output needs to be nkern x outshp)
bench1 = numpy.swapaxes(bench1, 1, 2)
ttime1 = time.time()
out1 = f(filters, biasvals, img1d)
ttot += time.time() - ttime1
temp = bench1.flatten() - out1.flatten()
assert (temp < 1e-5).all()
def test_convolution(self):
print "\n\n*************************************************"
print " TEST CONVOLUTION"
print "*************************************************"
# fixed parameters
bsize = 10 # batch size
imshp = (28, 28)
kshp = (5, 5)
nkern = 5
ssizes = ((1, 1), (2, 2), (3, 3), (4, 4))
convmodes = ("full", "valid")
# symbolic stuff
bias = tensor.dvector()
kerns = tensor.dmatrix()
input = tensor.dmatrix()
rng = numpy.random.RandomState(3423489)
filters = rng.randn(nkern, numpy.prod(kshp))
biasvals = rng.randn(nkern)
for mode in ("FAST_COMPILE", "FAST_RUN"): # , profmode):
ttot, ntot = 0, 0
for conv_mode in convmodes:
for ss in ssizes:
output, outshp = sp.convolve(kerns, kshp, nkern, input, imshp, ss, bias=bias, mode=conv_mode)
f = function([kerns, bias, input], output, mode=mode)
# now test with real values
img2d = numpy.arange(bsize * numpy.prod(imshp)).reshape((bsize,) + imshp)
img1d = img2d.reshape(bsize, -1)
# create filters (need to be flipped to use convolve2d)
filtersflipped = numpy.zeros((nkern,) + kshp)
for k in range(nkern):
it = reversed(filters[k, :])
for i in range(kshp[0]):
for j in range(kshp[1]):
filtersflipped[k, i, j] = it.next()
# compute output with convolve2d
if conv_mode == "valid":
fulloutshp = numpy.array(imshp) - numpy.array(kshp) + 1
else:
fulloutshp = numpy.array(imshp) + numpy.array(kshp) - 1
ntime1 = time.time()
refout = numpy.zeros((bsize,) + tuple(fulloutshp) + (nkern,))
for b in range(bsize):
for n in range(nkern):
refout[b, ..., n] = convolve2d(img2d[b, :, :], filtersflipped[n, ...], conv_mode)
ntot += time.time() - ntime1
# need to flatten images
bench1 = refout[:, 0 :: ss[0], 0 :: ss[1], :].reshape(bsize, -1, nkern)
bench1 += biasvals.reshape(1, 1, nkern)
# swap the last two dimensions (output needs to be nkern x outshp)
bench1 = numpy.swapaxes(bench1, 1, 2)
ttime1 = time.time()
out1 = f(filters, biasvals, img1d)
ttot += time.time() - ttime1
temp = bench1.flatten() - out1.flatten()
assert (temp < 1e-5).all()
# test downward propagation -- symbolic stuff
# vis = tensor.grad(output, input, output)
# downprop = function([kerns,input], vis, mode=mode)
# visval = downprop(filters,img1d)
## test downward propagation -- reference implementation
# pshape = (img1d.shape[0],numpy.prod(outshp[1:]),numpy.prod(kshp))
# patchstack = numpy.zeros(pshape)
# for bi in numpy.arange(pshape[0]): # batch index
# abspos = 0
# for outy in numpy.arange(outshp[1]):
# for outx in numpy.arange(outshp[2]):
# for ni in numpy.arange(nkern):
# print 'filters[n,:].shape = ', filters[n,:].shape
# print 'out1[bi,abspos].shape =',out1[bi,abspos].shape
# patchstack[bi,abspos,:] = filters[n,:]*out1[bi,abspos]
# abspos+=1
# patchstack = patchstack.reshape(1,-1)
# indices, indptr, spmat_shape, sptype, outshp = \
# sp.convolution_indices.conv_eval(imshp,kshp,ss,conv_mode)
# spmat = sparse.csc_matrix((numpy.ones_like(indices),indices,indptr),spmat_shape)
# visref = numpy.dot(patchstack, spmat.todense())
# print 'visval = ', visval
# print 'visref = ', visref
# assert numpy.all(visref==visval)
print "**** Convolution Profiling Results (", mode, ") ****"
print "Numpy processing time: ", ntot
print "Theano processing time: ", ttot