We compile with it. We don't use all its new features. Fred
On Thu, Apr 27, 2017 at 6:10 PM Michael Klachko <[email protected]> wrote: > Great, that worked, and the error about float64 is gone. Thank you. By the > way, is CuDNN v6 supported in this dev build? > > On Thu, Apr 27, 2017 at 2:46 PM, Frédéric Bastien < > [email protected]> wrote: > >> pip install --no-deps -U git+ >> https://github.com/Theano/Theano.git#egg=Theano >> >> I added "-U" >> >> On Thu, Apr 27, 2017 at 5:45 PM Michael Klachko <[email protected]> >> wrote: >> >>> It does not want to update: >>> >>> C:\Users\Michael\>pip install --no-deps git+ >>> https://github.com/Theano/Theano.git#egg=Theano >>> Requirement already satisfied: Theano from git+ >>> https://github.com/Theano/Theano.git#egg=Theano in >>> c:\programdata\miniconda2\lib\site-packages >>> >>> On Tue, Apr 11, 2017 at 2:12 PM, Frédéric Bastien < >>> [email protected]> wrote: >>> >>>> >>>> http://deeplearning.net/software/theano/install_ubuntu.html#bleeding-edge-installation-recommended >>>> >>>> >>>> pip install --user --no-deps >>>> git+https://github.com/Theano/Theano.git#egg=Theano >>>> >>>> >>>> >>>> On Tue, Apr 11, 2017 at 12:22 PM Michael Klachko < >>>> [email protected]> wrote: >>>> >>>>> Thanks! Currently I use the release version of Theano 0.9, installed >>>>> with conda. What's the best way to switch to dev version? Should I just >>>>> overwrite the >>>>> C:\ProgramData\Miniconda2\pkgs\theano-0.9.0-py27_0\Lib\site-packages\theano >>>>> folder? >>>>> >>>>> On Tue, Apr 11, 2017 at 8:02 AM, Frédéric Bastien < >>>>> [email protected]> wrote: >>>>> >>>>> Just an updated, we fixed many such useless warning. Update Theano to >>>>> the dev version. If you still see them, tell us. >>>>> >>>>> Fred >>>>> >>>>> On Sat, Mar 25, 2017 at 4:22 AM Michael Klachko < >>>>> [email protected]> wrote: >>>>> >>>>> Oh, looks like my message got truncated. Here's the complete graph: >>>>> >>>>> >>>>> C:\ProgramData\Miniconda2\lib\site-packages\theano\gof\type.py:405: >>>>> UserWarning: You are creating a TensorVariable with float64 dtype. You >>>>> requested an action via the Theano flag >>>>> warn_float64={ignore,warn,raise,pdb}. >>>>> >>>>> return utils.add_tag_trace(self.make_variable(name)) >>>>> >>>>> C:\ProgramData\Miniconda2\lib\site-packages\theano\gof\graph.py:447: >>>>> UserWarning: You are creating a TensorVariable with float64 dtype. You >>>>> requested an action via the Theano flag >>>>> warn_float64={ignore,warn,raise,pdb}. >>>>> >>>>> cp = self.__class__(self.type, None, None, self.name) >>>>> >>>>> HostFromGpu(gpuarray) [id A] <TensorType(float32, scalar)> 'mean' 84 >>>>> >>>>> |GpuElemwise{Composite{(i0 / Cast{float32}((Composite{Switch(LT(i0, >>>>> i1), i2, i0)}(Composite{Switch(GE(i0, i1), i2, >>>>> i0)}(Composite{Switch(LT(i0, >>>>> i1), i2, i0)}(Composite{Switch(LT(i0, i1), (i2 + i3), i4)}(i1, i2, i3, i4, >>>>> i5), i6, i7), i8, i9), i10, i11) - Switch(LT(Composite{Switch(LT(i0, i1), >>>>> i2, i0)}(Composite{Switch(GE(i0, i1), i2, i0)}(Composite{Switch(LT(i0, >>>>> i1), >>>>> i2, i0)}(Composite{Switch(LT(i0, i1), (i2 + i3), i4)}(i12, i13, i14, i15, >>>>> i16), i17, i18), i19, i20), i21, i22), Composite{Switch(LT(i0, i1), i2, >>>>> i0)}(Composite{Switch(GE(i0, i1), i2, i0)}(Composite{Switch(LT(i0, i1), >>>>> i2, >>>>> i0)}(Composite{Switch(LT(i0, i1), (i2 + i3), i4)}(i1, i2, i3, i4, i5), i6, >>>>> i7), i8, i9), i10, i11)), Composite{Switch(LT(i0, i1), i2, >>>>> i0)}(Composite{Switch(GE(i0, i1), i2, i0)}(Composite{Switch(LT(i0, i1), >>>>> i2, >>>>> i0)}(Composite{Switch(LT(i0, i1), (i2 + i3), i4)}(i12, i13, i14, i15, >>>>> i16), >>>>> i17, i18), i19, i20), i21, i22), Composite{Switch(LT(i0, i1), i2, >>>>> i0)}(Composite{Switch(GE(i0, i1), i2, i0)}(Composite{Switch(LT(i0, i1), >>>>> i2, >>>>> i0)}(Composite{Switch(LT(i0, i1), (i2 + i3), i4)}(i1, i2, i3, i4, i5), i6, >>>>> i7), i8, i9), i10, i11)))))}}[(0, 0)]<gpuarray> [id B] >>>>> <GpuArrayType<None>(float32, ())> '' 83 >>>>> >>>>> |GpuCAReduceCuda{add} [id C] <GpuArrayType<None>(float32, ())> '' 82 >>>>> >>>>> | |GpuElemwise{Composite{EQ(Cast{int32}(i0), i1)}}[]<gpuarray> [id D] >>>>> <GpuArrayType<None>(bool, (False,))> '' 81 >>>>> >>>>> | |GpuSubtensor{int64:int64:} [id E] <GpuArrayType<None>(float32, >>>>> (False,))> '' 27 >>>>> >>>>> | | |<GpuArrayType<None>(float32, (False,))> [id F] >>>>> <GpuArrayType<None>(float32, (False,))> >>>>> >>>>> | | |ScalarFromTensor [id G] <int64> '' 13 >>>>> >>>>> | | | |Elemwise{mul,no_inplace} [id H] <TensorType(int64, scalar)> '' 1 >>>>> >>>>> | | | |TensorConstant{128} [id I] <TensorType(int64, scalar)> >>>>> >>>>> | | | |<TensorType(int64, scalar)> [id J] <TensorType(int64, scalar)> >>>>> >>>>> | | |ScalarFromTensor [id K] <int64> '' 15 >>>>> >>>>> | | |Elemwise{Composite{(i0 * (i1 + i2))}} [id L] <TensorType(int64, >>>>> scalar)> '' 2 >>>>> >>>>> | | |TensorConstant{128} [id I] <TensorType(int64, scalar)> >>>>> >>>>> | | |TensorConstant{1} [id M] <TensorType(int64, scalar)> >>>>> >>>>> | | |<TensorType(int64, scalar)> [id J] <TensorType(int64, scalar)> >>>>> >>>>> | |GpuFromHost<None> [id N] <GpuArrayType<None>(int64, (False,))> '' 80 >>>>> >>>>> | |Argmax [id O] <TensorType(int64, vector)> 'argmax' 79 >>>>> >>>>> | |HostFromGpu(gpuarray) [id P] <TensorType(float32, matrix)> '' 78 >>>>> >>>>> | | |GpuElemwise{Add}[(0, 0)]<gpuarray> [id Q] >>>>> <GpuArrayType<None>(float32, (False, False))> '' 77 >>>>> >>>>> | | |GpuGemm{inplace=True} [id R] <GpuArrayType<None>(float32, (False, >>>>> False))> '' 76 >>>>> >>>>> | | | |GpuAllocEmpty{dtype='float32', context_name=None} [id S] >>>>> <GpuArrayType<None>(float32, (False, False))> '' 75 >>>>> >>>>> | | | | |Subtensor{int64} [id T] <TensorType(int64, scalar)> '' 74 >>>>> >>>>> | | | | | |Shape [id U] <TensorType(int64, vector)> '' 73 >>>>> >>>>> | | | | | | |GpuReshape{2} [id V] <GpuArrayType<None>(float32, (False, >>>>> False))> '' 72 >>>>> >>>>> | | | | | | |GpuElemwise{Composite{(i0 * ((i1 * (i2 + i3)) + (i4 * >>>>> Abs((i2 + i3)))))}}[(0, 2)]<gpuarray> [id W] <GpuArrayType<None>(float32, >>>>> (False, False))> '' 71 >>>>> >>>>> | | | | | | | |GpuArrayConstant{[[ 0.5]]} [id X] >>>>> <GpuArrayType<None>(float32, (True, True))> >>>>> >>>>> | | | | | | | |GpuArrayConstant{[[ 0.1]]} [id Y] >>>>> <GpuArrayType<None>(float32, (True, True))> >>>>> >>>>> | | | | | | | |GpuGemm{inplace=True} [id Z] >>>>> <GpuArrayType<None>(float32, (False, False))> '' 70 >>>>> >>>>> | | | | | | | | |GpuAllocEmpty{dtype='float32', context_name=None} [id >>>>> BA] <GpuArrayType<None>(float32, (False, False))> '' 69 >>>>> >>>>> | | | | | | | | | |Subtensor{int64} [id BB] <TensorType(int64, >>>>> scalar)> '' 68 >>>>> >>>>> | | | | | | | | | | |Shape [id BC] <TensorType(int64, vector)> '' 67 >>>>> >>>>> | | | | | | | | | | | |GpuReshape{2} [id BD] >>>>> <GpuArrayType<None>(float32, (False, False))> '' 66 >>>>> >>>>> | | | | | | | | | | | |GpuElemwise{Composite{(i0 * ((i1 * (i2 + i3)) + >>>>> (i4 * Abs((i2 + i3)))))}}[(0, 2)]<gpuarray> [id BE] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 65 >>>>> >>>>> | | | | | | | | | | | | |GpuArrayConstant{[[[[ 0.5]]]]} [id BF] >>>>> <GpuArrayType<None>(float32, (True, True, True, True))> >>>>> >>>>> | | | | | | | | | | | | |GpuArrayConstant{[[[[ 0.095]]]]} [id BG] >>>>> <GpuArrayType<None>(float32, (True, True, True, True))> >>>>> >>>>> | | | | | | | | | | | | |GpuDnnConv{algo='time_on_shape_change', >>>>> inplace=True} [id BH] <GpuArrayType<None>(float32, (False, False, False, >>>>> False))> '' 64 >>>>> >>>>> | | | | | | | | | | | | | |GpuContiguous [id BI] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 51 >>>>> >>>>> | | | | | | | | | | | | | | |GpuReshape{4} [id BJ] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 50 >>>>> >>>>> | | | | | | | | | | | | | | |GpuElemwise{Composite{(i0 * ((i1 * (i2 + >>>>> i3)) + (i4 * Abs((i2 + i3)))))}}[(0, 2)]<gpuarray> [id BK] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 49 >>>>> >>>>> | | | | | | | | | | | | | | | |GpuArrayConstant{[[[[ 0.5]]]]} [id BF] >>>>> <GpuArrayType<None>(float32, (True, True, True, True))> >>>>> >>>>> | | | | | | | | | | | | | | | |GpuArrayConstant{[[[[ 0.1]]]]} [id BL] >>>>> <GpuArrayType<None>(float32, (True, True, True, True))> >>>>> >>>>> | | | | | | | | | | | | | | | |GpuDnnConv{algo='time_on_shape_change', >>>>> inplace=True} [id BM] <GpuArrayType<None>(float32, (False, False, False, >>>>> False))> '' 48 >>>>> >>>>> | | | | | | | | | | | | | | | | |GpuContiguous [id BN] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 35 >>>>> >>>>> | | | | | | | | | | | | | | | | | |GpuReshape{4} [id BO] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 32 >>>>> >>>>> | | | | | | | | | | | | | | | | | |GpuSubtensor{int64:int64:} [id BP] >>>>> <GpuArrayType<None>(float32, (False, False))> '' 26 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |<GpuArrayType<None>(float32, >>>>> (False, False))> [id BQ] <GpuArrayType<None>(float32, (False, False))> >>>>> >>>>> | | | | | | | | | | | | | | | | | | |ScalarFromTensor [id G] <int64> >>>>> '' 13 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |ScalarFromTensor [id K] <int64> >>>>> '' 15 >>>>> >>>>> | | | | | | | | | | | | | | | | | |TensorConstant{[128 3 32 32]} [id >>>>> BR] <TensorType(int32, vector)> >>>>> >>>>> | | | | | | | | | | | | | | | | |GpuContiguous [id BS] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 10 >>>>> >>>>> | | | | | | | | | | | | | | | | | |<GpuArrayType<None>(float32, >>>>> (False, False, False, False))> [id BT] <GpuArrayType<None>(float32, >>>>> (False, >>>>> False, False, False))> >>>>> >>>>> | | | | | | | | | | | | | | | | |GpuAllocEmpty{dtype='float32', >>>>> context_name=None} [id BU] <GpuArrayType<None>(float32, (False, False, >>>>> False, False))> '' 47 >>>>> >>>>> | | | | | | | | | | | | | | | | | |Assert{msg='The convolution would >>>>> produce an invalid shape (dim[0] < 0).'} [id BV] <TensorType(int64, >>>>> scalar)> '' 44 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |Shape_i{0} [id BW] >>>>> <TensorType(int64, scalar)> '' 38 >>>>> >>>>> | | | | | | | | | | | | | | | | | | | |GpuContiguous [id BN] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 35 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |Elemwise{ge,no_inplace} [id BX] >>>>> <TensorType(bool, scalar)> '' 41 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |Shape_i{0} [id BW] >>>>> <TensorType(int64, scalar)> '' 38 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |TensorConstant{0} [id BY] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | |Assert{msg='The convolution would >>>>> produce an invalid shape (dim[1] < 0).'} [id BZ] <TensorType(int64, >>>>> scalar)> '' 34 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |Shape_i{0} [id CA] >>>>> <TensorType(int64, scalar)> '' 25 >>>>> >>>>> | | | | | | | | | | | | | | | | | | | |GpuContiguous [id BS] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 10 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |Elemwise{ge,no_inplace} [id CB] >>>>> <TensorType(bool, scalar)> '' 31 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |Shape_i{0} [id CA] >>>>> <TensorType(int64, scalar)> '' 25 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |TensorConstant{0} [id CC] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | |Assert{msg='The convolution would >>>>> produce an invalid shape (dim[2] <= 0).'} [id CD] <TensorType(int64, >>>>> scalar)> '' 46 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |Elemwise{Composite{(((i0 - (((i1 >>>>> - i2) * i3) + i4)) // i5) + i6)}}[(0, 0)] [id CE] <TensorType(int64, >>>>> scalar)> '' 40 >>>>> >>>>> | | | | | | | | | | | | | | | | | | | |Shape_i{2} [id CF] >>>>> <TensorType(int64, scalar)> '' 37 >>>>> >>>>> | | | | | | | | | | | | | | | | | | | | |GpuContiguous [id BN] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 35 >>>>> >>>>> | | | | | | | | | | | | | | | | | | | |Shape_i{2} [id CG] >>>>> <TensorType(int64, scalar)> '' 24 >>>>> >>>>> | | | | | | | | | | | | | | | | | | | | |GpuContiguous [id BS] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 10 >>>>> >>>>> | | | | | | | | | | | | | | | | | | | |TensorConstant{1} [id CH] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | | | |TensorConstant{1} [id CI] >>>>> <TensorType(int32, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | | | |TensorConstant{1} [id CH] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | | | |TensorConstant{2} [id CJ] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | | | |TensorConstant{1} [id CH] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | | |Elemwise{gt,no_inplace} [id CK] >>>>> <TensorType(bool, scalar)> '' 43 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |Elemwise{Composite{(((i0 - (((i1 >>>>> - i2) * i3) + i4)) // i5) + i6)}}[(0, 0)] [id CE] <TensorType(int64, >>>>> scalar)> '' 40 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |TensorConstant{0} [id CL] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | |Assert{msg='The convolution would >>>>> produce an invalid shape (dim[3] <= 0).'} [id CM] <TensorType(int64, >>>>> scalar)> '' 45 >>>>> >>>>> | | | | | | | | | | | | | | | | | |Elemwise{Composite{(((i0 - (((i1 - >>>>> i2) * i3) + i4)) // i5) + i6)}}[(0, 0)] [id CN] <TensorType(int64, >>>>> scalar)> >>>>> '' 39 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |Shape_i{3} [id CO] >>>>> <TensorType(int64, scalar)> '' 36 >>>>> >>>>> | | | | | | | | | | | | | | | | | | | |GpuContiguous [id BN] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 35 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |Shape_i{3} [id CP] >>>>> <TensorType(int64, scalar)> '' 23 >>>>> >>>>> | | | | | | | | | | | | | | | | | | | |GpuContiguous [id BS] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 10 >>>>> >>>>> | | | | | | | | | | | | | | | | | | |TensorConstant{1} [id CH] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | | |TensorConstant{1} [id CI] >>>>> <TensorType(int32, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | | |TensorConstant{1} [id CH] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | | |TensorConstant{2} [id CJ] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | | |TensorConstant{1} [id CH] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | | |Elemwise{gt,no_inplace} [id CQ] >>>>> <TensorType(bool, scalar)> '' 42 >>>>> >>>>> | | | | | | | | | | | | | | | | | |Elemwise{Composite{(((i0 - (((i1 - >>>>> i2) * i3) + i4)) // i5) + i6)}}[(0, 0)] [id CN] <TensorType(int64, >>>>> scalar)> >>>>> '' 39 >>>>> >>>>> | | | | | | | | | | | | | | | | | |TensorConstant{0} [id CR] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | |GpuDnnConvDesc{border_mode='valid', >>>>> subsample=(2, 2), conv_mode='conv', precision='float32'} [id CS] >>>>> <CDataType{cudnnConvolutionDescriptor_t}> '' 30 >>>>> >>>>> | | | | | | | | | | | | | | | | | |Shape [id CT] <TensorType(int64, >>>>> vector)> '' 22 >>>>> >>>>> | | | | | | | | | | | | | | | | | |GpuContiguous [id BS] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 10 >>>>> >>>>> | | | | | | | | | | | | | | | | |Constant{1.0} [id CU] <float32> >>>>> >>>>> | | | | | | | | | | | | | | | | |Constant{0.0} [id CV] <float32> >>>>> >>>>> | | | | | | | | | | | | | | | |InplaceGpuDimShuffle{x,0,x,x} [id CW] >>>>> <GpuArrayType<None>(float32, (True, False, True, True))> '' 9 >>>>> >>>>> | | | | | | | | | | | | | | | | |<GpuArrayType<None>(float32, >>>>> (False,))> [id CX] <GpuArrayType<None>(float32, (False,))> >>>>> >>>>> | | | | | | | | | | | | | | | |GpuArrayConstant{[[[[ 0.1]]]]} [id BL] >>>>> <GpuArrayType<None>(float32, (True, True, True, True))> >>>>> >>>>> | | | | | | | | | | | | | | |TensorConstant{[128 64 14 14]} [id CY] >>>>> <TensorType(int32, vector)> >>>>> >>>>> | | | | | | | | | | | | | |GpuContiguous [id CZ] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 8 >>>>> >>>>> | | | | | | | | | | | | | | |<GpuArrayType<None>(float32, (False, >>>>> False, False, False))> [id DA] <GpuArrayType<None>(float32, (False, False, >>>>> False, False))> >>>>> >>>>> | | | | | | | | | | | | | |GpuAllocEmpty{dtype='float32', >>>>> context_name=None} [id DB] <GpuArrayType<None>(float32, (False, False, >>>>> False, False))> '' 63 >>>>> >>>>> | | | | | | | | | | | | | | |Assert{msg='The convolution would produce >>>>> an invalid shape (dim[0] < 0).'} [id DC] <TensorType(int64, scalar)> '' 60 >>>>> >>>>> | | | | | | | | | | | | | | | |Shape_i{0} [id DD] <TensorType(int64, >>>>> scalar)> '' 54 >>>>> >>>>> | | | | | | | | | | | | | | | | |GpuContiguous [id BI] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 51 >>>>> >>>>> | | | | | | | | | | | | | | | |Elemwise{ge,no_inplace} [id DE] >>>>> <TensorType(bool, scalar)> '' 57 >>>>> >>>>> | | | | | | | | | | | | | | | |Shape_i{0} [id DD] <TensorType(int64, >>>>> scalar)> '' 54 >>>>> >>>>> | | | | | | | | | | | | | | | |TensorConstant{0} [id DF] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | |Assert{msg='The convolution would produce >>>>> an invalid shape (dim[1] < 0).'} [id DG] <TensorType(int64, scalar)> '' 33 >>>>> >>>>> | | | | | | | | | | | | | | | |Shape_i{0} [id DH] <TensorType(int64, >>>>> scalar)> '' 21 >>>>> >>>>> | | | | | | | | | | | | | | | | |GpuContiguous [id CZ] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 8 >>>>> >>>>> | | | | | | | | | | | | | | | |Elemwise{ge,no_inplace} [id DI] >>>>> <TensorType(bool, scalar)> '' 29 >>>>> >>>>> | | | | | | | | | | | | | | | |Shape_i{0} [id DH] <TensorType(int64, >>>>> scalar)> '' 21 >>>>> >>>>> | | | | | | | | | | | | | | | |TensorConstant{0} [id DJ] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | |Assert{msg='The convolution would produce >>>>> an invalid shape (dim[2] <= 0).'} [id DK] <TensorType(int64, scalar)> '' >>>>> 62 >>>>> >>>>> | | | | | | | | | | | | | | | |Elemwise{Composite{(((i0 - (((i1 - i2) >>>>> * i3) + i4)) // i5) + i6)}}[(0, 0)] [id DL] <TensorType(int64, scalar)> '' >>>>> 56 >>>>> >>>>> | | | | | | | | | | | | | | | | |Shape_i{2} [id DM] <TensorType(int64, >>>>> scalar)> '' 53 >>>>> >>>>> | | | | | | | | | | | | | | | | | |GpuContiguous [id BI] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 51 >>>>> >>>>> | | | | | | | | | | | | | | | | |Shape_i{2} [id DN] <TensorType(int64, >>>>> scalar)> '' 20 >>>>> >>>>> | | | | | | | | | | | | | | | | | |GpuContiguous [id CZ] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 8 >>>>> >>>>> | | | | | | | | | | | | | | | | |TensorConstant{1} [id CH] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | |TensorConstant{1} [id CI] >>>>> <TensorType(int32, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | |TensorConstant{1} [id CH] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | |TensorConstant{2} [id CJ] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | | |TensorConstant{1} [id CH] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | |Elemwise{gt,no_inplace} [id DO] >>>>> <TensorType(bool, scalar)> '' 59 >>>>> >>>>> | | | | | | | | | | | | | | | |Elemwise{Composite{(((i0 - (((i1 - i2) >>>>> * i3) + i4)) // i5) + i6)}}[(0, 0)] [id DL] <TensorType(int64, scalar)> '' >>>>> 56 >>>>> >>>>> | | | | | | | | | | | | | | | |TensorConstant{0} [id DP] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | |Assert{msg='The convolution would produce >>>>> an invalid shape (dim[3] <= 0).'} [id DQ] <TensorType(int64, scalar)> '' >>>>> 61 >>>>> >>>>> | | | | | | | | | | | | | | |Elemwise{Composite{(((i0 - (((i1 - i2) * >>>>> i3) + i4)) // i5) + i6)}}[(0, 0)] [id DR] <TensorType(int64, scalar)> '' >>>>> 55 >>>>> >>>>> | | | | | | | | | | | | | | | |Shape_i{3} [id DS] <TensorType(int64, >>>>> scalar)> '' 52 >>>>> >>>>> | | | | | | | | | | | | | | | | |GpuContiguous [id BI] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 51 >>>>> >>>>> | | | | | | | | | | | | | | | |Shape_i{3} [id DT] <TensorType(int64, >>>>> scalar)> '' 19 >>>>> >>>>> | | | | | | | | | | | | | | | | |GpuContiguous [id CZ] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 8 >>>>> >>>>> | | | | | | | | | | | | | | | |TensorConstant{1} [id CH] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | |TensorConstant{1} [id CI] >>>>> <TensorType(int32, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | |TensorConstant{1} [id CH] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | |TensorConstant{2} [id CJ] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | | |TensorConstant{1} [id CH] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | | |Elemwise{gt,no_inplace} [id DU] >>>>> <TensorType(bool, scalar)> '' 58 >>>>> >>>>> | | | | | | | | | | | | | | |Elemwise{Composite{(((i0 - (((i1 - i2) * >>>>> i3) + i4)) // i5) + i6)}}[(0, 0)] [id DR] <TensorType(int64, scalar)> '' >>>>> 55 >>>>> >>>>> | | | | | | | | | | | | | | |TensorConstant{0} [id DV] >>>>> <TensorType(int8, scalar)> >>>>> >>>>> | | | | | | | | | | | | | |GpuDnnConvDesc{border_mode='valid', >>>>> subsample=(2, 2), conv_mode='conv', precision='float32'} [id DW] >>>>> <CDataType{cudnnConvolutionDescriptor_t}> '' 28 >>>>> >>>>> | | | | | | | | | | | | | | |Shape [id DX] <TensorType(int64, vector)> >>>>> '' 18 >>>>> >>>>> | | | | | | | | | | | | | | |GpuContiguous [id CZ] >>>>> <GpuArrayType<None>(float32, (False, False, False, False))> '' 8 >>>>> >>>>> | | | | | | | | | | | | | |Constant{1.0} [id CU] <float32> >>>>> >>>>> | | | | | | | | | | | | | |Constant{0.0} [id CV] <float32> >>>>> >>>>> | | | | | | | | | | | | |InplaceGpuDimShuffle{x,0,x,x} [id DY] >>>>> <GpuArrayType<None>(float32, (True, False, True, True))> '' 7 >>>>> >>>>> | | | | | | | | | | | | | |<GpuArrayType<None>(float32, (False,))> [id >>>>> DZ] <GpuArrayType<None>(float32, (False,))> >>>>> >>>>> | | | | | | | | | | | | |GpuArrayConstant{[[[[ 0.095]]]]} [id BG] >>>>> <GpuArrayType<None>(float32, (True, True, True, True))> >>>>> >>>>> | | | | | | | | | | | |TensorConstant{[ 128 3200]} [id EA] >>>>> <TensorType(int32, vector)> >>>>> >>>>> | | | | | | | | | | |Constant{0} [id EB] <int64> >>>>> >>>>> | | | | | | | | | |Subtensor{int64} [id EC] <TensorType(int64, >>>>> scalar)> '' 17 >>>>> >>>>> | | | | | | | | | |Shape [id ED] <TensorType(int64, vector)> '' 6 >>>>> >>>>> | | | | | | | | | | |w [id EE] <GpuArrayType<None>(float32, (False, >>>>> False))> >>>>> >>>>> | | | | | | | | | |Constant{1} [id EF] <int64> >>>>> >>>>> | | | | | | | | |TensorConstant{0.25} [id EG] <TensorType(float64, >>>>> scalar)> >>>>> >>>>> | | | | | | | | |GpuReshape{2} [id BD] <GpuArrayType<None>(float32, >>>>> (False, False))> '' 66 >>>>> >>>>> | | | | | | | | |w [id EE] <GpuArrayType<None>(float32, (False, >>>>> False))> >>>>> >>>>> | | | | | | | | |TensorConstant{0.0} [id EH] <TensorType(float64, >>>>> scalar)> >>>>> >>>>> | | | | | | | |InplaceGpuDimShuffle{x,0} [id EI] >>>>> <GpuArrayType<None>(float32, (True, False))> '' 5 >>>>> >>>>> | | | | | | | | |b [id EJ] <GpuArrayType<None>(float32, (False,))> >>>>> >>>>> | | | | | | | |GpuArrayConstant{[[ 0.1]]} [id Y] >>>>> <GpuArrayType<None>(float32, (True, True))> >>>>> >>>>> | | | | | | |TensorConstant{[128 390]} [id EK] <TensorType(int32, >>>>> vector)> >>>>> >>>>> | | | | | |Constant{0} [id EB] <int64> >>>>> >>>>> | | | | |Subtensor{int64} [id EL] <TensorType(int64, scalar)> '' 16 >>>>> >>>>> | | | | |Shape [id EM] <TensorType(int64, vector)> '' 4 >>>>> >>>>> | | | | | |w [id EN] <GpuArrayType<None>(float32, (False, False))> >>>>> >>>>> | | | | |Constant{1} [id EF] <int64> >>>>> >>>>> | | | |TensorConstant{0.25} [id EG] <TensorType(float64, scalar)> >>>>> >>>>> | | | |GpuReshape{2} [id V] <GpuArrayType<None>(float32, (False, >>>>> False))> '' 72 >>>>> >>>>> | | | |w [id EN] <GpuArrayType<None>(float32, (False, False))> >>>>> >>>>> | | | |TensorConstant{0.0} [id EH] <TensorType(float64, scalar)> >>>>> >>>>> | | |InplaceGpuDimShuffle{x,0} [id EO] <GpuArrayType<None>(float32, >>>>> (True, False))> '' 3 >>>>> >>>>> | | |b [id EP] <GpuArrayType<None>(float32, (False,))> >>>>> >>>>> | |TensorConstant{(1L,) of 1} [id EQ] <TensorType(int32, (True,))> >>>>> >>>>> |GpuFromHost<None> [id ER] <GpuArrayType<None>(int64, ())> '' 14 >>>>> >>>>> | |Elemwise{Composite{(i0 * (i1 + i2))}} [id L] <TensorType(int64, >>>>> scalar)> '' 2 >>>>> >>>>> |GpuArrayConstant{0} [id ES] <GpuArrayType<None>(int8, ())> >>>>> >>>>> |GpuFromHost<None> [id ER] <GpuArrayType<None>(int64, ())> '' 14 >>>>> >>>>> |GpuFromHost<None> [id ET] <GpuArrayType<None>(int64, ())> '' 11 >>>>> >>>>> | |Shape_i{0} [id EU] <TensorType(int64, scalar)> '' 0 >>>>> >>>>> | |<GpuArrayType<None>(float32, (False,))> [id F] >>>>> <GpuArrayType<None>(float32, (False,))> >>>>> >>>>> |GpuFromHost<None> [id ER] <GpuArrayType<None>(int64, ())> '' 14 >>>>> >>>>> |GpuArrayConstant{0} [id ES] <GpuArrayType<None>(int8, ())> >>>>> >>>>> |GpuArrayConstant{-1} [id EV] <GpuArrayType<None>(int8, ())> >>>>> >>>>> |GpuFromHost<None> [id ET] <GpuArrayType<None>(int64, ())> '' 11 >>>>> >>>>> |GpuFromHost<None> [id ET] <GpuArrayType<None>(int64, ())> '' 11 >>>>> >>>>> |GpuArrayConstant{0} [id ES] <GpuArrayType<None>(int8, ())> >>>>> >>>>> |GpuArrayConstant{0} [id ES] <GpuArrayType<None>(int8, ())> >>>>> >>>>> |GpuFromHost<None> [id EW] <GpuArrayType<None>(int64, ())> '' 12 >>>>> >>>>> | |Elemwise{mul,no_inplace} [id H] <TensorType(int64, scalar)> '' 1 >>>>> >>>>> |GpuArrayConstant{0} [id ES] <GpuArrayType<None>(int8, ())> >>>>> >>>>> |GpuFromHost<None> [id EW] <GpuArrayType<None>(int64, ())> '' 12 >>>>> >>>>> |GpuFromHost<None> [id ET] <GpuArrayType<None>(int64, ())> '' 11 >>>>> >>>>> |GpuFromHost<None> [id EW] <GpuArrayType<None>(int64, ())> '' 12 >>>>> >>>>> |GpuArrayConstant{0} [id ES] <GpuArrayType<None>(int8, ())> >>>>> >>>>> |GpuArrayConstant{0} [id ES] <GpuArrayType<None>(int8, ())> >>>>> >>>>> |GpuFromHost<None> [id ET] <GpuArrayType<None>(int64, ())> '' 11 >>>>> >>>>> |GpuFromHost<None> [id ET] <GpuArrayType<None>(int64, ())> '' 11 >>>>> >>>>> |GpuArrayConstant{0} [id ES] <GpuArrayType<None>(int8, ())> >>>>> >>>>> |GpuArrayConstant{0} [id ES] <GpuArrayType<None>(int8, ())> >>>>> >>>>> -- >>>>> >>>>> --- >>>>> You received this message because you are subscribed to the Google >>>>> Groups "theano-users" group. >>>>> To unsubscribe from this group and stop receiving emails from it, send >>>>> an email to [email protected]. >>>>> For more options, visit https://groups.google.com/d/optout. >>>>> >>>>> -- >>>>> >>>>> --- >>>>> You received this message because you are subscribed to a topic in the >>>>> Google Groups "theano-users" group. >>>>> To unsubscribe from this topic, visit >>>>> https://groups.google.com/d/topic/theano-users/-bAlrdp-IIQ/unsubscribe >>>>> . >>>>> To unsubscribe from this group and all its topics, send an email to >>>>> [email protected]. >>>>> >>>>> -- --- You received this message because you are subscribed to the Google Groups "theano-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
