Skip to content

P
pytensor
提交 8aefb2f5 authored 作者: Nicolas Bouchard's avatar Nicolas Bouchard
浏览文件
操作

pep8

上级 4b909254
隐藏空白字符变更
内嵌 并排
正在显示 包含 270 行增加 和 289 行删除
theano/scalar/basic.py
浏览文件 @ 8aefb2f5
...@@ -2018,7 +2018,7 @@ class Cos(UnaryScalarOp): ...@@ -2018,7 +2018,7 @@ class Cos(UnaryScalarOp):
cos = Cos(upgrade_to_float, name='cos') cos = Cos(upgrade_to_float, name='cos')
class Arccos(UnaryScalarOp): class ArcCos(UnaryScalarOp):
def impl(self, x): def impl(self, x):
return numpy.arccos(x) return numpy.arccos(x)
...@@ -2034,7 +2034,7 @@ class Arccos(UnaryScalarOp): ...@@ -2034,7 +2034,7 @@ class Arccos(UnaryScalarOp):
if node.inputs[0].type in complex_types: if node.inputs[0].type in complex_types:
raise NotImplementedError('type not supported', type) raise NotImplementedError('type not supported', type)
return "%(z)s = acos(%(x)s);" % locals() return "%(z)s = acos(%(x)s);" % locals()
arccos = Arccos(upgrade_to_float, name='arccos') arccos = ArcCos(upgrade_to_float, name='arccos')
class Sin(UnaryScalarOp): class Sin(UnaryScalarOp):
...@@ -2056,7 +2056,7 @@ class Sin(UnaryScalarOp): ...@@ -2056,7 +2056,7 @@ class Sin(UnaryScalarOp):
sin = Sin(upgrade_to_float, name='sin') sin = Sin(upgrade_to_float, name='sin')
class Arcsin(UnaryScalarOp): class ArcSin(UnaryScalarOp):
def impl(self, x): def impl(self, x):
return numpy.arcsin(x) return numpy.arcsin(x)
...@@ -2072,7 +2072,7 @@ class Arcsin(UnaryScalarOp): ...@@ -2072,7 +2072,7 @@ class Arcsin(UnaryScalarOp):
if node.inputs[0].type in complex_types: if node.inputs[0].type in complex_types:
raise NotImplementedError('type not supported', type) raise NotImplementedError('type not supported', type)
return "%(z)s = asin(%(x)s);" % locals() return "%(z)s = asin(%(x)s);" % locals()
arcsin = Arcsin(upgrade_to_float, name='arcsin') arcsin = ArcSin(upgrade_to_float, name='arcsin')
class Tan(UnaryScalarOp): class Tan(UnaryScalarOp):
...@@ -2094,7 +2094,7 @@ class Tan(UnaryScalarOp): ...@@ -2094,7 +2094,7 @@ class Tan(UnaryScalarOp):
tan = Tan(upgrade_to_float, name='tan') tan = Tan(upgrade_to_float, name='tan')
class Arctan(UnaryScalarOp): class ArcTan(UnaryScalarOp):
def impl(self, x): def impl(self, x):
return numpy.arctan(x) return numpy.arctan(x)
...@@ -2110,7 +2110,7 @@ class Arctan(UnaryScalarOp): ...@@ -2110,7 +2110,7 @@ class Arctan(UnaryScalarOp):
if node.inputs[0].type in complex_types: if node.inputs[0].type in complex_types:
raise NotImplementedError('type not supported', type) raise NotImplementedError('type not supported', type)
return "%(z)s = atan(%(x)s);" % locals() return "%(z)s = atan(%(x)s);" % locals()
arctan = Arctan(upgrade_to_float, name='arctan') arctan = ArcTan(upgrade_to_float, name='arctan')
class Cosh(UnaryScalarOp): class Cosh(UnaryScalarOp):
......
theano/tensor/tests/test_basic.py
浏览文件 @ 8aefb2f5
...@@ -908,304 +908,285 @@ FloorInplaceTester = makeBroadcastTester(op=inplace.floor_inplace, ...@@ -908,304 +908,285 @@ FloorInplaceTester = makeBroadcastTester(op=inplace.floor_inplace,
expected=lambda a: numpy.asarray(numpy.floor(a), a.dtype), expected=lambda a: numpy.asarray(numpy.floor(a), a.dtype),
good=_good_broadcast_unary_normal_no_complex, good=_good_broadcast_unary_normal_no_complex,
grad=_grad_broadcast_unary_normal, grad=_grad_broadcast_unary_normal,
inplace = True) inplace=True)
RoundHalfToEvenTester = makeBroadcastTester(op = tensor.round_half_to_even, RoundHalfToEvenTester = makeBroadcastTester(
expected = numpy.round, op=tensor.round_half_to_even,
good = _good_broadcast_unary_normal_float_no_complex) expected=numpy.round,
good=_good_broadcast_unary_normal_float_no_complex)
# TODO: Why complex are accepted in the next one? # TODO: Why complex are accepted in the next one?
RoundHalfToEvenInplaceTester = makeBroadcastTester(op = inplace.round_half_to_even_inplace, RoundHalfToEvenInplaceTester = makeBroadcastTester(
expected = numpy.round, op=inplace.round_half_to_even_inplace,
good = _good_broadcast_unary_normal_float, expected=numpy.round,
inplace = True) good=_good_broadcast_unary_normal_float,
inplace=True)
#numpy.vectorize don't handle correctly empty ndarray. #numpy.vectorize don't handle correctly empty ndarray.
#see in their file numpy/lib/function_base.py in class vectorize.__call__ #see in their file numpy/lib/function_base.py in class vectorize.__call__
#This happen in float32 mode. #This happen in float32 mode.
RoundHalfAwayFromZeroTester = makeBroadcastTester(op = tensor.round_half_away_from_zero, RoundHalfAwayFromZeroTester = makeBroadcastTester(
expected = theano.scalar.basic.round_half_away_from_zero_vec, op=tensor.round_half_away_from_zero,
good = _good_broadcast_unary_normal_float_no_empty_no_complex)#_good_broadcast_unary_normal_float) expected=theano.scalar.basic.round_half_away_from_zero_vec,
RoundHalfAwayFromZeroInplaceTester = makeBroadcastTester(op = inplace.round_half_away_from_zero_inplace, good=_good_broadcast_unary_normal_float_no_empty_no_complex)
expected = theano.scalar.basic.round_half_away_from_zero_vec, #_good_broadcast_unary_normal_float)
good = _good_broadcast_unary_normal_float_no_empty_no_complex, RoundHalfAwayFromZeroInplaceTester = makeBroadcastTester(
inplace = True) op=inplace.round_half_away_from_zero_inplace,
expected=theano.scalar.basic.round_half_away_from_zero_vec,
SqrTester = makeBroadcastTester(op = tensor.sqr, good=_good_broadcast_unary_normal_float_no_empty_no_complex,
expected = numpy.square, inplace=True)
good = _good_broadcast_unary_normal,
grad = _grad_broadcast_unary_normal)
SqrInplaceTester = makeBroadcastTester(op = inplace.sqr_inplace,
expected = numpy.square,
good = _good_broadcast_unary_normal,
grad = _grad_broadcast_unary_normal,
inplace = True)
ExpTester = makeBroadcastTester(op = tensor.exp,
expected = numpy.exp,
good = _good_broadcast_unary_normal,
grad = _grad_broadcast_unary_normal)
ExpInplaceTester = makeBroadcastTester(op = inplace.exp_inplace,
expected = numpy.exp,
good = _good_broadcast_unary_normal,
grad = _grad_broadcast_unary_normal,
inplace = True)
Exp2Tester = makeBroadcastTester(op = tensor.exp2,
expected = numpy.exp2,
good = _good_broadcast_unary_normal,
grad = _grad_broadcast_unary_normal)
Exp2InplaceTester = makeBroadcastTester(op = inplace.exp2_inplace,
expected = numpy.exp2,
good = _good_broadcast_unary_normal,
grad = _grad_broadcast_unary_normal,
inplace = True)
_good_broadcast_unary_positive = dict(normal = (rand_ranged(0.001, 5, (2, 3)),),
integers = (randint_ranged(1, 5, (2, 3)),),
complex = (randc128_ranged(1, 5, (2,3)),),
empty = (numpy.asarray([]),),
)
_grad_broadcast_unary_positive = dict(normal = (rand_ranged(0.001, 5, (2, 3)),), SqrTester = makeBroadcastTester(op=tensor.sqr,
#complex = (randc128_ranged(1, 5, (2,3)),), expected=numpy.square,
#empty = (numpy.asarray([]),), good=_good_broadcast_unary_normal,
grad=_grad_broadcast_unary_normal)
SqrInplaceTester = makeBroadcastTester(op=inplace.sqr_inplace,
expected=numpy.square,
good=_good_broadcast_unary_normal,
grad=_grad_broadcast_unary_normal,
inplace=True)
ExpTester = makeBroadcastTester(op=tensor.exp,
expected=numpy.exp,
good=_good_broadcast_unary_normal,
grad=_grad_broadcast_unary_normal)
ExpInplaceTester = makeBroadcastTester(op=inplace.exp_inplace,
expected=numpy.exp,
good=_good_broadcast_unary_normal,
grad=_grad_broadcast_unary_normal,
inplace=True)
Exp2Tester = makeBroadcastTester(op=tensor.exp2,
expected=numpy.exp2,
good=_good_broadcast_unary_normal,
grad=_grad_broadcast_unary_normal)
Exp2InplaceTester = makeBroadcastTester(op=inplace.exp2_inplace,
expected=numpy.exp2,
good=_good_broadcast_unary_normal,
grad=_grad_broadcast_unary_normal,
inplace=True)
_good_broadcast_unary_positive = dict(normal=(rand_ranged(0.001, 5, (2, 3)),),
integers=(randint_ranged(1, 5, (2, 3)),),
complex=(randc128_ranged(1, 5, (2, 3)),),
empty=(numpy.asarray([]),),
) )
LogTester = makeBroadcastTester(op = tensor.log, _grad_broadcast_unary_positive = dict(normal=(rand_ranged(0.001, 5, (2, 3)),),)
expected = numpy.log,
good = _good_broadcast_unary_positive, LogTester = makeBroadcastTester(op=tensor.log,
grad = _grad_broadcast_unary_positive) expected=numpy.log,
LogInplaceTester = makeBroadcastTester(op = inplace.log_inplace, good=_good_broadcast_unary_positive,
expected = numpy.log, grad=_grad_broadcast_unary_positive)
good = _good_broadcast_unary_positive, LogInplaceTester = makeBroadcastTester(op=inplace.log_inplace,
grad = _grad_broadcast_unary_positive, expected=numpy.log,
inplace = True) good=_good_broadcast_unary_positive,
grad=_grad_broadcast_unary_positive,
Log2Tester = makeBroadcastTester(op = tensor.log2, inplace=True)
expected = numpy.log2,
good = _good_broadcast_unary_positive, Log2Tester = makeBroadcastTester(op=tensor.log2,
grad = _grad_broadcast_unary_positive) expected=numpy.log2,
Log2InplaceTester = makeBroadcastTester(op = inplace.log2_inplace, good=_good_broadcast_unary_positive,
expected = numpy.log2, grad=_grad_broadcast_unary_positive)
good = _good_broadcast_unary_positive, Log2InplaceTester = makeBroadcastTester(op=inplace.log2_inplace,
grad = _grad_broadcast_unary_positive, expected=numpy.log2,
inplace = True) good=_good_broadcast_unary_positive,
grad=_grad_broadcast_unary_positive,
Log10Tester = makeBroadcastTester(op = tensor.log10, inplace=True)
expected = numpy.log10,
good = _good_broadcast_unary_positive, Log10Tester = makeBroadcastTester(op=tensor.log10,
grad = _grad_broadcast_unary_positive) expected=numpy.log10,
Log10InplaceTester = makeBroadcastTester(op = inplace.log10_inplace, good=_good_broadcast_unary_positive,
expected = numpy.log10, grad=_grad_broadcast_unary_positive)
good = _good_broadcast_unary_positive, Log10InplaceTester = makeBroadcastTester(op=inplace.log10_inplace,
grad = _grad_broadcast_unary_positive, expected=numpy.log10,
inplace = True) good=_good_broadcast_unary_positive,
grad=_grad_broadcast_unary_positive,
Log1pTester = makeBroadcastTester(op = tensor.log1p, inplace=True)
expected = numpy.log1p,
good = _good_broadcast_unary_positive, Log1pTester = makeBroadcastTester(op=tensor.log1p,
grad = _grad_broadcast_unary_positive) expected=numpy.log1p,
Log1pInplaceTester = makeBroadcastTester(op = inplace.log1p_inplace, good=_good_broadcast_unary_positive,
expected = numpy.log1p, grad=_grad_broadcast_unary_positive)
good = _good_broadcast_unary_positive, Log1pInplaceTester = makeBroadcastTester(op=inplace.log1p_inplace,
grad = _grad_broadcast_unary_positive, expected=numpy.log1p,
inplace = True) good=_good_broadcast_unary_positive,
grad=_grad_broadcast_unary_positive,
inplace=True)
SqrtTester = makeBroadcastTester(op = tensor.sqrt,
expected = numpy.sqrt, SqrtTester = makeBroadcastTester(op=tensor.sqrt,
good = _good_broadcast_unary_positive, expected=numpy.sqrt,
grad = _grad_broadcast_unary_positive) good=_good_broadcast_unary_positive,
SqrtInplaceTester = makeBroadcastTester(op = inplace.sqrt_inplace, grad=_grad_broadcast_unary_positive)
expected = numpy.sqrt, SqrtInplaceTester = makeBroadcastTester(op=inplace.sqrt_inplace,
good = _good_broadcast_unary_positive, expected=numpy.sqrt,
grad = _grad_broadcast_unary_positive, good=_good_broadcast_unary_positive,
inplace = True) grad=_grad_broadcast_unary_positive,
inplace=True)
_good_broadcast_unary_wide = dict(
_good_broadcast_unary_wide = dict(normal = (rand_ranged(-1000, 1000, (2, 3)),), normal=(rand_ranged(-1000, 1000, (2, 3)),),
integers = (randint_ranged(-1000, 1000, (2, 3)),), integers=(randint_ranged(-1000, 1000, (2, 3)),),
complex = (randc128_ranged(-1000, 1000, (2, 3)),), complex=(randc128_ranged(-1000, 1000, (2, 3)),),
empty = (numpy.asarray([]),),) empty=(numpy.asarray([]),),)
_grad_broadcast_unary_wide = dict(normal=(rand_ranged(-1000, 1000, (2, 3)),),)
_grad_broadcast_unary_wide = dict(normal = (rand_ranged(-1000, 1000, (2, 3)),),
#complex = (randc128_ranged(-1000, 1000, (2, 3)),), SinTester = makeBroadcastTester(op=tensor.sin,
#empty = (numpy.asarray([]),), expected=numpy.sin,
) good=_good_broadcast_unary_wide,
grad=_grad_broadcast_unary_wide)
_good_broadcast_unary_arccos = dict(normal = (rand_ranged(-1.+1e-7, 1.-1e-7, (2, 3)),), SinInplaceTester = makeBroadcastTester(op=inplace.sin_inplace,
integers = (randint_ranged(-1.+1e-7, 1-1e-7, (2, 3)),), expected=numpy.sin,
complex = (randc128_ranged(-1.+1e-7, 1-1e-7, (2, 3)),), good=_good_broadcast_unary_wide,
empty = (numpy.asarray([]),),) grad=_grad_broadcast_unary_wide,
inplace=True)
_grad_broadcast_unary_arccos = dict(normal = (rand_ranged(-1.+1e-7, 1-1e-7, (2, 3)),),
#complex = (randc128_ranged(-1000, 1000, (2, 3)),), _good_broadcast_unary_arcsin = dict(normal=(rand_ranged(-1, 1, (2, 3)),),
#empty = (numpy.asarray([]),), integers=(randint_ranged(-1, 1, (2, 3)),),
) complex=(randc128_ranged(-1, 1, (2, 3)),),
empty=(numpy.asarray([]),),)
_good_broadcast_unary_arcsin = _good_broadcast_unary_arccos _grad_broadcast_unary_arcsin = dict(normal=(rand_ranged(-1, 1, (2, 3)),),)
_grad_broadcast_unary_arcsin = _grad_broadcast_unary_arccos ArcSinTester = makeBroadcastTester(op=tensor.arcsin,
expected=numpy.arcsin,
SinTester = makeBroadcastTester(op = tensor.sin, good=_good_broadcast_unary_arcsin,
expected = numpy.sin, grad=_grad_broadcast_unary_arcsin)
good = _good_broadcast_unary_wide, ArcSinInplaceTester = makeBroadcastTester(op=inplace.arcsin_inplace,
grad = _grad_broadcast_unary_wide) expected=numpy.arcsin,
SinInplaceTester = makeBroadcastTester(op = inplace.sin_inplace, good=_good_broadcast_unary_arcsin,
expected = numpy.sin, grad=_grad_broadcast_unary_arcsin,
good = _good_broadcast_unary_wide, inplace=True)
grad = _grad_broadcast_unary_wide,
inplace = True) CosTester = makeBroadcastTester(op=tensor.cos,
expected=numpy.cos,
ArcsinTester = makeBroadcastTester(op = tensor.arcsin, good=_good_broadcast_unary_wide,
expected = numpy.arcsin, grad=_grad_broadcast_unary_wide)
good = _good_broadcast_unary_arcsin, CosInplaceTester = makeBroadcastTester(op=inplace.cos_inplace,
grad = _grad_broadcast_unary_arcsin) expected=numpy.cos,
good=_good_broadcast_unary_wide,
ArcsinInplaceTester = makeBroadcastTester(op = inplace.arcsin_inplace, grad=_grad_broadcast_unary_wide,
expected = numpy.arcsin, inplace=True)
good = _good_broadcast_unary_arcsin,
grad = _grad_broadcast_unary_arcsin, ArcCosTester = makeBroadcastTester(op=tensor.arccos,
inplace = True) expected=numpy.arccos,
good=_good_broadcast_unary_arcsin,
CosTester = makeBroadcastTester(op = tensor.cos, grad=_grad_broadcast_unary_arcsin)
expected = numpy.cos, ArcCosInplaceTester = makeBroadcastTester(op=inplace.arccos_inplace,
good = _good_broadcast_unary_wide, expected=numpy.arccos,
grad = _grad_broadcast_unary_wide) good=_good_broadcast_unary_arcsin,
CosInplaceTester = makeBroadcastTester(op = inplace.cos_inplace, grad=_grad_broadcast_unary_arcsin,
expected = numpy.cos, inplace=True)
good = _good_broadcast_unary_wide,
grad = _grad_broadcast_unary_wide,
inplace = True)
ArccosTester = makeBroadcastTester(op = tensor.arccos,
expected = numpy.arccos,
good = _good_broadcast_unary_arccos,
grad = _grad_broadcast_unary_arccos)
ArccosInplaceTester = makeBroadcastTester(op = inplace.arccos_inplace,
expected = numpy.arccos,
good = _good_broadcast_unary_arccos,
grad = _grad_broadcast_unary_arccos,
inplace = True)
tan_grad_rtol = None tan_grad_rtol = None
if config.floatX=='float32': if config.floatX == 'float32':
#We raise the relative tolerence for the grad as their is error in float32 #We raise the relative tolerence for the grad as their is error in float32
#This is probably caused by our way of computing the gradient error. #This is probably caused by our way of computing the gradient error.
tan_grad_rtol = 0.052 tan_grad_rtol = 0.052
TanTester = makeBroadcastTester(op = tensor.tan, _good_broadcast_unary_tan = dict(
expected = numpy.tan, normal=(rand_ranged(-3.14, 3.14, (2, 3)),),
good = dict(normal = (rand_ranged(-3.14, 3.14, (2, 3)),), shifted=(rand_ranged(3.15, 6.28, (2, 3)),),
shifted = (rand_ranged(3.15, 6.28, (2, 3)),)), integers=(randint_ranged(-3, 3, (2, 3)),),
grad = dict(normal = (rand_ranged(-3.14, 3.14, (2, 3)),), complex=(randc128_ranged(-3.14, 3.14, (2, 3)),),
shifted = (rand_ranged(3.15, 6.28, (2, 3)),)), empty=(numpy.asarray([]),),)
grad_rtol=tan_grad_rtol) _grad_broadcast_unary_tan = dict(normal=(rand_ranged(-3.14, 3.14, (2, 3)),),
TanInplaceTester = makeBroadcastTester(op = inplace.tan_inplace, shifted=(rand_ranged(3.15, 6.28, (2, 3)),))
expected = numpy.tan,
good = dict(normal = (rand_ranged(-3.14, 3.14, (2, 3)),), TanTester = makeBroadcastTester(op=tensor.tan,
shifted = (rand_ranged(3.15, 6.28, (2, 3)),)), expected=numpy.tan,
grad = dict(normal = (rand_ranged(-3.14, 3.14, (2, 3)),), good=_good_broadcast_unary_tan,
shifted = (rand_ranged(3.15, 6.28, (2, 3)),)), grad=_grad_broadcast_unary_tan,
grad_rtol=tan_grad_rtol, grad_rtol=tan_grad_rtol)
inplace = True) TanInplaceTester = makeBroadcastTester(op=inplace.tan_inplace,
expected=numpy.tan,
good=_good_broadcast_unary_tan,
ArctanTester = makeBroadcastTester(op = tensor.tan, grad=_grad_broadcast_unary_tan,
expected = numpy.tan, grad_rtol=tan_grad_rtol,
good = _good_broadcast_unary_wide, inplace=True)
grad = _grad_broadcast_unary_wide,
ArcTanTester = makeBroadcastTester(op=tensor.tan,
expected=numpy.tan,
good=_good_broadcast_unary_wide,
grad=_grad_broadcast_unary_wide,
grad_rtol=tan_grad_rtol) grad_rtol=tan_grad_rtol)
ArctanInplaceTester = makeBroadcastTester(op = inplace.tan_inplace, ArcTanInplaceTester = makeBroadcastTester(op=inplace.tan_inplace,
expected = numpy.tan, expected=numpy.tan,
good = _good_broadcast_unary_wide, good=_good_broadcast_unary_wide,
grad = _grad_broadcast_unary_wide, grad=_grad_broadcast_unary_wide,
grad_rtol=tan_grad_rtol, grad_rtol=tan_grad_rtol,
inplace = True) inplace=True)
CoshTester = makeBroadcastTester(op=tensor.cosh,
CoshTester = makeBroadcastTester(op = tensor.cosh, expected=numpy.cosh,
expected = numpy.cosh, good=_good_broadcast_unary_normal,
good = _good_broadcast_unary_normal, grad=_grad_broadcast_unary_normal)
grad = _grad_broadcast_unary_normal) CoshInplaceTester = makeBroadcastTester(op=inplace.cosh_inplace,
CoshInplaceTester = makeBroadcastTester(op = inplace.cosh_inplace, expected=numpy.cosh,
expected = numpy.cosh, good=_good_broadcast_unary_normal,
good = _good_broadcast_unary_normal, grad=_grad_broadcast_unary_normal,
grad = _grad_broadcast_unary_normal, inplace=True)
inplace = True)
_good_broadcast_unary_arccosh = dict(
normal=(rand_ranged(1, 1000, (2, 3)),),
_good_broadcast_unary_arccosh = dict(normal = (rand_ranged(1, 1000, (2, 3)),), integers=(randint_ranged(1, 1000, (2, 3)),),
integers = (randint_ranged(1, 1000, (2, 3)),), complex=(randc128_ranged(1, 1000, (2, 3)),),
complex = (randc128_ranged(1, 1000, (2, 3)),), empty=(numpy.asarray([]),),)
empty = (numpy.asarray([]),),) _grad_broadcast_unary_arccosh = dict(normal=(rand_ranged(1, 1000, (2, 3)),),)
_grad_broadcast_unary_arccosh = dict(normal = (rand_ranged(1, 1000, (2, 3)),),)
ArcCoshTester = makeBroadcastTester(op=tensor.arccosh,
expected=numpy.arccosh,
ArccoshTester = makeBroadcastTester(op = tensor.arccosh, good=_good_broadcast_unary_arccosh,
expected = numpy.arccosh, grad=_grad_broadcast_unary_arccosh)
good = _good_broadcast_unary_arccosh, ArcCoshInplaceTester = makeBroadcastTester(op=inplace.arccosh_inplace,
grad = _grad_broadcast_unary_arccosh) expected=numpy.arccosh,
ArccoshInplaceTester = makeBroadcastTester(op = inplace.arccosh_inplace, good=_good_broadcast_unary_arccosh,
expected = numpy.arccosh, grad=_grad_broadcast_unary_arccosh,
good = _good_broadcast_unary_arccosh, inplace=True)
grad = _grad_broadcast_unary_arccosh,
inplace = True) SinhTester = makeBroadcastTester(op=tensor.sinh,
expected=numpy.sinh,
good=_good_broadcast_unary_normal,
SinhTester = makeBroadcastTester(op = tensor.sinh, grad=_grad_broadcast_unary_normal)
expected = numpy.sinh, SinhInplaceTester = makeBroadcastTester(op=inplace.sinh_inplace,
good = _good_broadcast_unary_normal, expected=numpy.sinh,
grad = _grad_broadcast_unary_normal) good=_good_broadcast_unary_normal,
SinhInplaceTester = makeBroadcastTester(op = inplace.sinh_inplace, grad=_grad_broadcast_unary_normal,
expected = numpy.sinh, inplace=True)
good = _good_broadcast_unary_normal,
grad = _grad_broadcast_unary_normal, ArcSinhTester = makeBroadcastTester(op=tensor.arcsinh,
inplace = True) expected=numpy.arcsinh,
good=_good_broadcast_unary_normal,
grad=_grad_broadcast_unary_normal)
ArcsinhTester = makeBroadcastTester(op = tensor.arcsinh, ArcSinhInplaceTester = makeBroadcastTester(op=inplace.arcsinh_inplace,
expected = numpy.arcsinh, expected=numpy.arcsinh,
good = _good_broadcast_unary_normal, good=_good_broadcast_unary_normal,
grad = _grad_broadcast_unary_normal) grad=_grad_broadcast_unary_normal,
ArcsinhInplaceTester = makeBroadcastTester(op = inplace.arcsinh_inplace, inplace=True)
expected = numpy.arcsinh,
good = _good_broadcast_unary_normal, TanhTester = makeBroadcastTester(op=tensor.tanh,
grad = _grad_broadcast_unary_normal, expected=numpy.tanh,
inplace = True) good=_good_broadcast_unary_normal,
grad=_grad_broadcast_unary_normal)
TanhInplaceTester = makeBroadcastTester(op=inplace.tanh_inplace,
TanhTester = makeBroadcastTester(op = tensor.tanh, expected=numpy.tanh,
expected = numpy.tanh, good=_good_broadcast_unary_normal,
good = _good_broadcast_unary_normal, grad=_grad_broadcast_unary_normal,
grad = _grad_broadcast_unary_normal) inplace=True)
TanhInplaceTester = makeBroadcastTester(op = inplace.tanh_inplace,
expected = numpy.tanh, _good_broadcast_unary_arctanh = dict(normal=(rand_ranged(-1, 1, (2, 3)),),
good = _good_broadcast_unary_normal, integers=(randint_ranged(-1, 1, (2, 3)),),
grad = _grad_broadcast_unary_normal, complex=(randc128_ranged(-1, 1, (2, 3)),),
inplace = True) empty=(numpy.asarray([]),),)
_grad_broadcast_unary_arctanh = dict(normal=(rand_ranged(-1, 1, (2, 3)),),)
_good_broadcast_unary_arctanh = dict(normal = (rand_ranged(-1, 1, (2, 3)),), ArcTanhTester = makeBroadcastTester(op=tensor.arctanh,
integers = (randint_ranged(-1, 1, (2, 3)),), expected=numpy.arctanh,
complex = (randc128_ranged(-1, 1, (2, 3)),), good=_good_broadcast_unary_arctanh,
empty = (numpy.asarray([]),),) grad=_grad_broadcast_unary_arctanh)
_grad_broadcast_unary_arctanh = dict(normal = (rand_ranged(-1, 1, (2, 3)),),) ArcTanhInplaceTester = makeBroadcastTester(op=inplace.arctanh_inplace,
expected=numpy.arctanh,
good=_good_broadcast_unary_arctanh,
ArctanhTester = makeBroadcastTester(op = tensor.arctanh, grad=_grad_broadcast_unary_arctanh,
expected = numpy.arctanh, inplace=True)
good = _good_broadcast_unary_arctanh,
grad = _grad_broadcast_unary_arctanh)
ArctanhInplaceTester = makeBroadcastTester(op = inplace.arctanh_inplace,
expected = numpy.arctanh,
good = _good_broadcast_unary_arctanh,
grad = _grad_broadcast_unary_arctanh,
inplace = True)
#inplace ops when the input is integer and the output is float* #inplace ops when the input is integer and the output is float*
......
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