Skip to content

P
pytensor
提交 bfe1fd42 authored 作者: Olivier Delalleau's avatar Olivier Delalleau
浏览文件
操作

Some PEP8 fixes

上级 31dfb237
隐藏空白字符变更
内嵌 并排
正在显示 包含 65 行增加 和 26 行删除
theano/sandbox/cuda/tests/test_mlp.py
浏览文件 @ bfe1fd42
import copy
import logging
import time
from itertools import izip
from nose.plugins.skip import SkipTest
import numpy
......@@ -16,8 +17,9 @@ import theano.sandbox.cuda as tcn
import theano.tests.unittest_tools as utt
if theano.config.mode not in ['FAST_RUN','Mode','ProfileMode']:
raise SkipTest('Skip test_mlp when not in normal optimization mode as otherwise it is too slow!')
if theano.config.mode not in ['FAST_RUN', 'Mode', 'ProfileMode']:
raise SkipTest('Skip test_mlp when not in normal optimization mode as '
'otherwise it is too slow!')
# Skip test if cuda_ndarray is not available.
if tcn.cuda_available == False:
......@@ -28,11 +30,16 @@ logging.getLogger('theano.sandbox.cuda.tests.test_nnet').setLevel(logging.INFO)
def my_rand(*shape):
return theano._asarray(numpy.random.rand(*shape),dtype='float32')
return theano._asarray(numpy.random.rand(*shape), dtype='float32')
def my_randn(*shape):
return theano._asarray(numpy.random.randn(*shape),dtype='float32')
return theano._asarray(numpy.random.randn(*shape), dtype='float32')
def my_zeros(*shape):
return theano._asarray(numpy.zeros(*shape),dtype='float32')
return theano._asarray(numpy.zeros(*shape), dtype='float32')
def get_mode(use_gpu, check_isfinite=True):
if theano.config.mode != 'FAST_COMPILE':
......@@ -50,25 +57,33 @@ def get_mode(use_gpu, check_isfinite=True):
ret = ret.excluding('gpu')
return ret
def print_mode(mode):
if mode != None and isinstance(mode,(theano.compile.ProfileMode,)):
if mode != None and isinstance(mode, (theano.compile.ProfileMode,)):
mode.print_summary()
def print_diff_mode(a,b):
if a != None and isinstance(a,(theano.compile.ProfileMode,)) and isinstance(b,(theano.compile.ProfileMode,)):
def print_diff_mode(a, b):
if (a is not None and
isinstance(a, (theano.compile.ProfileMode,)) and
isinstance(b, (theano.compile.ProfileMode,))):
a.print_diff_summary(b)
def run_nnet(use_gpu, n_batch=60, n_in=1024, n_hid=2048, n_out=10, n_train=100):
if config.mode=='DEBUG_MODE': n_train=1
def run_nnet(use_gpu, n_batch=60, n_in=1024, n_hid=2048, n_out=10,
n_train=100):
if config.mode == 'DEBUG_MODE':
n_train = 1
if use_gpu:
w = tcn.shared_constructor(0.01*(my_rand(n_in,n_hid)-0.5), 'w')
w = tcn.shared_constructor(0.01 * (my_rand(n_in, n_hid) - 0.5), 'w')
b = tcn.shared_constructor(my_zeros(n_hid), 'b')
v = tcn.shared_constructor(my_zeros((n_hid, n_out)), 'c')
c = tcn.shared_constructor(my_zeros(n_out), 'c')
else:
w = shared(0.01*(my_rand(n_in,n_hid)-0.5), 'w')
w = shared(0.01 * (my_rand(n_in, n_hid) - 0.5), 'w')
b = shared(my_zeros(n_hid), 'b')
v = shared(my_zeros((n_hid, n_out)), 'c')
c = shared(my_zeros(n_out), 'c')
......@@ -77,10 +92,11 @@ def run_nnet(use_gpu, n_batch=60, n_in=1024, n_hid=2048, n_out=10, n_train=100):
y = tensor.fmatrix('y')
lr = tensor.fscalar('lr')
hid = tensor.tanh(tensor.dot(x, w)+b)
out = tensor.tanh(tensor.dot(hid, v)+c)
loss = tensor.sum(0.5 * (out-y)**2 * lr)
if 0: print 'loss type', loss.type
hid = tensor.tanh(tensor.dot(x, w) + b)
out = tensor.tanh(tensor.dot(hid, v) + c)
loss = tensor.sum(0.5 * (out - y) ** 2 * lr)
if 0:
print 'loss type', loss.type
params = [w, b, v, c]
gparams = tensor.grad(loss, params)
......@@ -88,7 +104,8 @@ def run_nnet(use_gpu, n_batch=60, n_in=1024, n_hid=2048, n_out=10, n_train=100):
mode = get_mode(use_gpu)
print 'building pfunc ...'
train = pfunc([x,y,lr], [loss], mode=mode, updates=[(p, p-g) for p,g in zip(params, gparams)])
train = pfunc([x, y, lr], [loss], mode=mode,
updates=[(p, p - g) for p, g in izip(params, gparams)])
if 0:
for i, n in enumerate(train.maker.env.toposort()):
......@@ -107,34 +124,41 @@ def run_nnet(use_gpu, n_batch=60, n_in=1024, n_hid=2048, n_out=10, n_train=100):
print_mode(mode)
return numpy.asarray(rval), dt
def test_run_nnet():
for n_in in 1024, 2048, 4096:
for n_hid in 1024, 2048, 4096:
utt.seed_rng() # Seeds numpy rng with utt.fetch_seed()
utt.seed_rng() # Seeds numpy rng with utt.fetch_seed()
rval_cpu, tc = run_nnet(False, n_in=n_in, n_hid=n_hid)
utt.seed_rng()
rval_gpu, tg = run_nnet(True, n_in=n_in, n_hid=n_hid)
#print "cpu:", rval_cpu
#print "gpu:", rval_gpu
abs_diff, rel_diff = theano.gradient.numeric_grad.abs_rel_err(rval_gpu,rval_cpu)
abs_diff, rel_diff = \
theano.gradient.numeric_grad.abs_rel_err(rval_gpu,
rval_cpu)
max_abs_diff = abs_diff.max()
print "max abs diff=%e max rel diff=%e n_in=%d n_hid=%d"%(
print "max abs diff=%e max rel diff=%e n_in=%d n_hid=%d" % (
max_abs_diff, rel_diff.max(), n_in, n_hid)
print "time cpu: %f, time gpu: %f, speed up %f"%(tc, tg, tc/tg)
print "time cpu: %f, time gpu: %f, speed up %f" % (tc, tg, tc / tg)
rtol = 1e-4
if n_in*n_hid>=2048*4096:
if n_in * n_hid >= 2048 * 4096:
rtol = 7e-4
assert numpy.allclose(rval_cpu, rval_gpu,rtol=rtol,atol=1e-6), ("max_abs_diff, max_rel_diff, n_in, n_hid",
max_abs_diff, rel_diff.max(), n_in, n_hid)
assert numpy.allclose(rval_cpu, rval_gpu, rtol=rtol, atol=1e-6), \
("max_abs_diff, max_rel_diff, n_in, n_hid", max_abs_diff,
rel_diff.max(), n_in, n_hid)
def test_run_nnet_med():
utt.seed_rng()
rval_cpu = run_nnet(False, 10, 128, 50, 4, n_train=10000)
def test_run_nnet_small():
utt.seed_rng()
rval_cpu = run_nnet(False, 10, 10, 4, 4, n_train=100000)
def run_conv_nnet1(use_gpu):
if use_gpu:
shared_fn = tcn.shared_constructor
......@@ -144,8 +168,9 @@ def run_conv_nnet1(use_gpu):
n_kern = 20
shape_img = (n_batch, 1, 32, 32)
shape_kern = (n_kern, 1, 5, 5)
n_train=10
if config.mode=='DEBUG_MODE': n_train=1
n_train = 10
if config.mode == 'DEBUG_MODE':
n_train = 1
logical_hid_shape = tcn.blas.GpuConv.logical_output_shape_2d(shape_img[2:],shape_kern[2:], 'valid')
n_hid = n_kern * logical_hid_shape[0] * logical_hid_shape[1]
......@@ -190,6 +215,7 @@ def run_conv_nnet1(use_gpu):
print_mode(mode)
return rval
def test_conv_nnet1():
utt.seed_rng()
rval_cpu = run_conv_nnet1(False)
......@@ -197,6 +223,7 @@ def test_conv_nnet1():
rval_gpu = run_conv_nnet1(True)
assert numpy.allclose(rval_cpu, rval_gpu,rtol=1e-4,atol=1e-6)
def run_conv_nnet2(use_gpu): # pretend we are training LeNet for MNIST
if use_gpu:
shared_fn = tcn.shared_constructor
......@@ -276,6 +303,7 @@ def run_conv_nnet2(use_gpu): # pretend we are training LeNet for MNIST
print_mode(mode)
return rval
def test_conv_nnet2():
utt.seed_rng()
rval_gpu = run_conv_nnet2(True)
......@@ -367,6 +395,7 @@ def build_conv_nnet2_classif(use_gpu, isize, ksize, n_batch,
shape_target = (n_batch,n_out)
return train, params, shape_img, shape_target, mode
def run_conv_nnet2_classif(use_gpu, seed, isize, ksize, bsize,
n_train=10,
check_isfinite=True,
......@@ -412,6 +441,7 @@ def run_conv_nnet2_classif(use_gpu, seed, isize, ksize, bsize,
print "estimated time for one pass through MNIST with %s: %f" % (
device, (t1-t0) * (60000.0 / (n_train*bsize)))
def cmp_run_conv_nnet2_classif(seed, isize, ksize, bsize,
ignore_error=False,
n_train=10,
......@@ -542,6 +572,7 @@ def cmp_run_conv_nnet2_classif(seed, isize, ksize, bsize,
print "Estimated time for one pass through MNIST with GPU: %f" % (
(time_gpu * (60000.0 / (n_train*bsize))))
# Default parameters for all subsequent tests
gpu_only=False
cpu_only=False
......@@ -550,16 +581,19 @@ verbose=0
version=-1
seed = utt.fetch_seed()
def test_lenet_28(): #MNIST
cmp_run_conv_nnet2_classif(seed, 28, 5, 60, n_train=10,
ignore_error=ignore_error, gpu_only=gpu_only,
cpu_only=cpu_only, verbose=verbose, version=version)
def test_lenet_32(): #CIFAR10 / Shapeset
cmp_run_conv_nnet2_classif(seed, 32, 5, 60, n_train=8,
ignore_error=ignore_error, gpu_only=gpu_only,
verbose=verbose, version=version)
def test_lenet_32_long(): #CIFAR10 / Shapeset
# this tests the gradient of downsample on the GPU,
# which does not recieve specific testing
......@@ -567,6 +601,7 @@ def test_lenet_32_long(): #CIFAR10 / Shapeset
ignore_error=ignore_error, gpu_only=gpu_only,
cpu_only=cpu_only, verbose=verbose, version=version)
def test_lenet_64(): # ???
#float_atol need to pass in debug mode
#needed as cpu use extended precision and gpu don't
......@@ -575,18 +610,21 @@ def test_lenet_64(): # ???
cpu_only=cpu_only, verbose=verbose,
check_isfinite=True, version=version)
def test_lenet_108(): # NORB
cmp_run_conv_nnet2_classif(seed, 108, 7, 5, n_train=4,
ignore_error=ignore_error, gpu_only=gpu_only,
cpu_only=cpu_only, verbose=verbose,
check_isfinite=True, version=version)
def test_lenet_256(): # ImageNet
cmp_run_conv_nnet2_classif(seed, 256, 9, 2, n_train=5,
ignore_error=ignore_error, gpu_only=gpu_only,
cpu_only=cpu_only, verbose=verbose,
check_isfinite=True, version=version, float_atol=5e-5)
#I did a wanted error in the name as we don't want it to execute automatically for now as it don't work
def tes_lenet_hd(): #HD 720p: 1280(wid)x720(len)
cmp_run_conv_nnet2_classif(seed, (720,1280), 9, 2, n_train=3,
......@@ -594,6 +632,7 @@ def tes_lenet_hd(): #HD 720p: 1280(wid)x720(len)
cpu_only=cpu_only, verbose=verbose,
check_isfinite=True, version=version)
#I did a wanted error in the name as we don't want it to execute automatically for now as it don't work
def tes_lenet_full_hd(): #HD 1080p: 1920(wid)x1080(len)
cmp_run_conv_nnet2_classif(seed, (1080,1920), 9, 2, n_train=3,
......
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