Skip to content

Instantly share code, notes, and snippets.

@zomux

zomux/profile.6.txt

Created May 26, 2015 01:14
Show Gist options
  • Save zomux/d51f151feae904fead81 to your computer and use it in GitHub Desktop.
Save zomux/d51f151feae904fead81 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
profile.6.txt
Function profiling
==================
Message: None
Time in 2000 calls to Function.__call__: 1.451054e+00s
Time in Function.fn.__call__: 1.404534e+00s (96.794%)
Time in thunks: 9.432883e-01s (65.007%)
Total compile time: 5.521090e-01s
Theano Optimizer time: 1.164739e-01s
Theano validate time: 1.684189e-03s
Theano Linker time (includes C, CUDA code generation/compiling): 4.920912e-02s
Class
---
<% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
71.9% 71.9% 0.678s 1.13e-04s C 6000 3 <class 'theano.tensor.blas.Dot22'>
13.8% 85.7% 0.130s 9.30e-06s C 14000 7 <class 'theano.tensor.elemwise.Elemwise'>
7.1% 92.8% 0.067s 3.33e-05s Py 2000 1 <class 'theano.tensor.basic.MaxAndArgmax'>
1.8% 94.6% 0.017s 8.70e-06s Py 2000 1 <class 'theano.tensor.subtensor.AdvancedSubtensor'>
1.7% 96.3% 0.016s 2.69e-06s C 6000 3 <class 'theano.tensor.elemwise.DimShuffle'>
1.2% 97.5% 0.011s 5.54e-06s Py 2000 1 <class 'theano.tensor.basic.ARange'>
1.1% 98.6% 0.010s 5.05e-06s C 2000 1 <class 'theano.tensor.nnet.nnet.SoftmaxWithBias'>
0.8% 99.4% 0.008s 1.89e-06s C 4000 2 <class 'theano.tensor.elemwise.Sum'>
0.6% 100.0% 0.006s 1.50e-06s C 4000 2 <class 'theano.compile.ops.Shape_i'>
... (remaining 0 Classes account for 0.00%(0.00s) of the runtime)
Ops
---
<% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
71.9% 71.9% 0.678s 1.13e-04s C 6000 3 Dot22
11.4% 83.3% 0.108s 2.69e-05s C 4000 2 Elemwise{Composite{[Composite{[mul(i0, GT(i0, i1))]}(add(i0, i1), i2)]
7.1% 90.4% 0.067s 3.33e-05s Py 2000 1 MaxAndArgmax
1.8% 92.2% 0.017s 8.70e-06s Py 2000 1 AdvancedSubtensor
1.7% 93.9% 0.016s 2.69e-06s C 6000 3 DimShuffle{x,0}
1.2% 95.1% 0.011s 5.54e-06s Py 2000 1 ARange
1.1% 96.2% 0.010s 5.05e-06s C 2000 1 SoftmaxWithBias
1.0% 97.2% 0.009s 4.68e-06s C 2000 1 Elemwise{Composite{[log(clip(i0, i1, i2))]}}
0.6% 97.8% 0.006s 1.50e-06s C 4000 2 Shape_i{0}
0.5% 98.3% 0.005s 2.46e-06s C 2000 1 Sum
0.4% 98.7% 0.004s 1.93e-06s C 2000 1 Elemwise{neq,no_inplace}
0.4% 99.1% 0.004s 1.92e-06s C 2000 1 Elemwise{add,no_inplace}
0.3% 99.4% 0.003s 1.39e-06s C 2000 1 Elemwise{Composite{[Composite{[mul(i0, true_div(i1, i2))]}(i0, i1, Cas
0.3% 99.7% 0.003s 1.37e-06s C 2000 1 Elemwise{Composite{[Composite{[neg(true_div(i0, i1))]}(i0, Cast{float3
0.3% 100.0% 0.003s 1.31e-06s C 2000 1 Sum
... (remaining 0 Ops account for 0.00%(0.00s) of the runtime)
Apply
------
<% time> <sum %> <apply time> <time per call> <#call> <id> <Mflops> <Gflops/s> <Apply name>
52.7% 52.7% 0.498s 2.49e-04s 2000 5 Dot22(x, W_dense1)
input 0: dtype=float32, shape=(20, 784), strides=c
input 1: dtype=float32, shape=(784, 256), strides=c
output 0: dtype=float32, shape=(20, 256), strides=c
17.2% 69.9% 0.162s 8.09e-05s 2000 8 Dot22(Elemwise{Composite{[Composite{[mul(i0, GT(i0, i1))]}(ad
input 0: dtype=float32, shape=(20, 256), strides=c
input 1: dtype=float32, shape=(256, 256), strides=c
output 0: dtype=float32, shape=(20, 256), strides=c
7.1% 77.0% 0.067s 3.33e-05s 2000 13 MaxAndArgmax(Elemwise{add,no_inplace}.0, TensorConstant{(1,)
input 0: dtype=float32, shape=(20, 10), strides=c
input 1: dtype=int64, shape=(1,), strides=c
output 0: dtype=float32, shape=(20,), strides=c
output 1: dtype=int64, shape=(20,), strides=c
5.8% 82.7% 0.054s 2.72e-05s 2000 7 Elemwise{Composite{[Composite{[mul(i0, GT(i0, i1))]}(add(i0,
input 0: dtype=float32, shape=(20, 256), strides=c
input 1: dtype=float32, shape=(1, 256), strides=c
input 2: dtype=int8, shape=(1, 1), strides=c
output 0: dtype=float32, shape=(20, 256), strides=c
5.6% 88.4% 0.053s 2.66e-05s 2000 9 Elemwise{Composite{[Composite{[mul(i0, GT(i0, i1))]}(add(i0,
input 0: dtype=float32, shape=(20, 256), strides=c
input 1: dtype=float32, shape=(1, 256), strides=c
input 2: dtype=int8, shape=(1, 1), strides=c
output 0: dtype=float32, shape=(20, 256), strides=c
2.0% 90.4% 0.019s 9.33e-06s 2000 10 Dot22(Elemwise{Composite{[Composite{[mul(i0, GT(i0, i1))]}(ad
input 0: dtype=float32, shape=(20, 256), strides=c
input 1: dtype=float32, shape=(256, 10), strides=c
output 0: dtype=float32, shape=(20, 10), strides=c
1.8% 92.2% 0.017s 8.70e-06s 2000 16 AdvancedSubtensor(Elemwise{Composite{[log(clip(i0, i1, i2))]}
input 0: dtype=float32, shape=(20, 10), strides=c
input 1: dtype=int64, shape=(20,), strides=c
input 2: dtype=int32, shape=(20,), strides=c
output 0: dtype=float32, shape=(20,), strides=c
1.2% 93.4% 0.011s 5.54e-06s 2000 6 ARange(TensorConstant{0}, Shape_i{0}.0, TensorConstant{1})
input 0: dtype=int8, shape=(), strides=c
input 1: dtype=int64, shape=(), strides=c
input 2: dtype=int8, shape=(), strides=c
output 0: dtype=int64, shape=(20,), strides=c
1.1% 94.5% 0.010s 5.05e-06s 2000 12 SoftmaxWithBias(Dot22.0, B_dense3)
input 0: dtype=float32, shape=(20, 10), strides=c
input 1: dtype=float32, shape=(10,), strides=c
output 0: dtype=float32, shape=(20, 10), strides=c
1.0% 95.5% 0.009s 4.68e-06s 2000 14 Elemwise{Composite{[log(clip(i0, i1, i2))]}}(SoftmaxWithBias.
input 0: dtype=float32, shape=(20, 10), strides=c
input 1: dtype=float32, shape=(1, 1), strides=c
input 2: dtype=float32, shape=(1, 1), strides=c
output 0: dtype=float32, shape=(20, 10), strides=c
0.6% 96.1% 0.006s 2.98e-06s 2000 3 DimShuffle{x,0}(B_dense2)
input 0: dtype=float32, shape=(256,), strides=c
output 0: dtype=float32, shape=(1, 256), strides=c
0.6% 96.7% 0.006s 2.83e-06s 2000 2 DimShuffle{x,0}(B_dense3)
input 0: dtype=float32, shape=(10,), strides=c
output 0: dtype=float32, shape=(1, 10), strides=c
0.5% 97.2% 0.005s 2.46e-06s 2000 18 Sum(AdvancedSubtensor.0)
input 0: dtype=float32, shape=(20,), strides=c
output 0: dtype=float32, shape=(), strides=c
0.5% 97.7% 0.005s 2.26e-06s 2000 4 DimShuffle{x,0}(B_dense1)
input 0: dtype=float32, shape=(256,), strides=c
output 0: dtype=float32, shape=(1, 256), strides=c
0.4% 98.1% 0.004s 1.93e-06s 2000 15 Elemwise{neq,no_inplace}(argmax, k)
input 0: dtype=int64, shape=(20,), strides=c
input 1: dtype=int32, shape=(20,), strides=c
output 0: dtype=int8, shape=(20,), strides=c
0.4% 98.5% 0.004s 1.92e-06s 2000 11 Elemwise{add,no_inplace}(Dot22.0, DimShuffle{x,0}.0)
input 0: dtype=float32, shape=(20, 10), strides=c
input 1: dtype=float32, shape=(1, 10), strides=c
output 0: dtype=float32, shape=(20, 10), strides=c
0.4% 98.9% 0.004s 1.81e-06s 2000 0 Shape_i{0}(x)
input 0: dtype=float32, shape=(20, 784), strides=c
output 0: dtype=int64, shape=(), strides=c
0.3% 99.2% 0.003s 1.39e-06s 2000 19 Elemwise{Composite{[Composite{[mul(i0, true_div(i1, i2))]}(i0
input 0: dtype=float64, shape=(), strides=c
input 1: dtype=int64, shape=(), strides=c
input 2: dtype=int64, shape=(), strides=c
output 0: dtype=float64, shape=(), strides=c
0.3% 99.5% 0.003s 1.37e-06s 2000 20 Elemwise{Composite{[Composite{[neg(true_div(i0, i1))]}(i0, Ca
input 0: dtype=float32, shape=(), strides=c
input 1: dtype=int64, shape=(), strides=c
output 0: dtype=float32, shape=(), strides=c
0.3% 99.7% 0.003s 1.31e-06s 2000 17 Sum(Elemwise{neq,no_inplace}.0)
input 0: dtype=int8, shape=(20,), strides=c
output 0: dtype=int64, shape=(), strides=c
... (remaining 1 Apply instances account for 0.25%(0.00s) of the runtime)
Memory Profile
(Sparse variables are ignored)
---
Max if linker=cvm (default): unknown
Max if no gc (allow_gc=False): 86KB
Max if linker=c|py: 42KB
Memory saved if gc is enabled (linker=c|py): 43KB
<Sum apply outputs (bytes)> <Apply outputs shape> <created/inplace/view> <Apply node>
20480B [(20, 256)] c Dot22(x, W_dense1)
20480B [(20, 256)] c Dot22(Elemwise{Composite{[Composite{[mul(i0, GT(i0, i1))]}(add(i0, i1), i2)]}}.0, W_dense2)
20480B [(20, 256)] c Elemwise{Composite{[Composite{[mul(i0, GT(i0, i1))]}(add(i0, i1), i2)]}}(Dot22.0, DimShuffle{x,0}.0, TensorConstant{(1, 1) of 0})
20480B [(20, 256)] c Elemwise{Composite{[Composite{[mul(i0, GT(i0, i1))]}(add(i0, i1), i2)]}}(Dot22.0, DimShuffle{x,0}.0, TensorConstant{(1, 1) of 0})
1024B [(1, 256)] c DimShuffle{x,0}(B_dense1)
1024B [(1, 256)] c DimShuffle{x,0}(B_dense2)
... (remaining 15 Apply account for 3780B/87748B ((4.31%)) of the Apply with dense outputs sizes)
<created/inplace/view> is taken from the Op's declaration.
Apply nodes marked 'inplace' or 'view' may actually allocate memory, this is not reported here. If you use DebugMode, warnings will be emitted in those cases.
Function profiling
==================
Message: None
Time in 2594 calls to Function.__call__: 1.710203e+01s
Time in Function.fn.__call__: 1.674901e+01s (97.936%)
Time in thunks: 9.490532e+00s (55.494%)
Total compile time: 8.929539e-01s
Theano Optimizer time: 7.065191e-01s
Theano validate time: 1.076841e-02s
Theano Linker time (includes C, CUDA code generation/compiling): 1.703279e-01s
Class
---
<% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
28.5% 28.5% 2.702s 1.49e-05s C 181580 76 <class 'theano.tensor.elemwise.Elemwise'>
24.8% 53.2% 2.351s 1.13e-04s C 20752 8 <class 'theano.tensor.blas.Dot22'>
14.1% 67.4% 1.342s 1.29e-04s Py 10376 4 <class 'theano.tensor.basic.MaxAndArgmax'>
10.8% 78.2% 1.026s 1.65e-05s Py 62256 12 <class 'theano.ifelse.IfElse'>
10.5% 88.6% 0.993s 2.55e-05s C 38910 15 <class 'theano.tensor.elemwise.Sum'>
5.1% 93.7% 0.479s 6.16e-05s C 7782 3 <class 'theano.tensor.blas.Dot22Scalar'>
3.0% 96.7% 0.287s 7.90e-06s C 36316 14 <class 'theano.tensor.elemwise.DimShuffle'>
0.5% 97.3% 0.051s 1.97e-05s Py 2594 1 <class 'theano.tensor.subtensor.AdvancedSubtensor'>
0.5% 97.8% 0.049s 1.88e-05s Py 2594 1 <class 'theano.tensor.subtensor.AdvancedIncSubtensor'>
0.4% 98.2% 0.042s 2.72e-06s C 15564 6 <class 'theano.tensor.subtensor.Subtensor'>
0.4% 98.6% 0.034s 1.31e-05s Py 2594 1 <class 'theano.tensor.basic.ARange'>
0.3% 98.9% 0.033s 2.56e-06s C 12970 8 <class 'theano.compile.ops.Shape_i'>
0.3% 99.2% 0.025s 9.51e-06s C 2594 1 <class 'theano.tensor.nnet.nnet.SoftmaxWithBias'>
0.2% 99.4% 0.023s 2.91e-06s C 7782 3 <class 'theano.tensor.opt.MakeVector'>
0.2% 99.7% 0.022s 2.82e-06s C 7782 3 <class 'theano.tensor.elemwise.CAReduce'>
0.2% 99.9% 0.021s 4.01e-06s C 5188 8 <class 'theano.tensor.basic.Alloc'>
0.1% 100.0% 0.012s 4.45e-06s C 2594 1 <class 'theano.tensor.nnet.nnet.SoftmaxGrad'>
... (remaining 0 Classes account for 0.00%(0.00s) of the runtime)
Ops
---
<% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
24.8% 24.8% 2.351s 1.13e-04s C 20752 8 Dot22
14.1% 38.9% 1.342s 1.29e-04s Py 10376 4 MaxAndArgmax
10.8% 49.7% 1.026s 1.65e-05s Py 62256 12 if{}
9.6% 59.3% 0.913s 3.20e-05s C 28534 11 Sum
7.3% 66.6% 0.694s 2.97e-05s C 23346 9 Elemwise{add,no_inplace}
6.7% 73.4% 0.640s 4.11e-05s C 15564 6 Elemwise{Composite{[sub(mul(i0, i1), mul(i2, i3))]}}
5.4% 78.7% 0.508s 6.53e-05s C 7782 3 Elemwise{Composite{[add(i0, mul(i1, i2))]}}
5.1% 83.8% 0.479s 6.16e-05s C 7782 3 Dot22Scalar
4.6% 88.4% 0.433s 2.78e-05s C 15564 6 Elemwise{Composite{[sqr(Abs{output_types_preference=<class 'theano.sca
1.9% 90.3% 0.181s 1.39e-05s C 12970 5 DimShuffle{1,0}
0.8% 91.0% 0.072s 9.26e-06s C 7782 3 Sum{0}
0.6% 91.6% 0.059s 5.64e-06s C 10376 4 Elemwise{mul,no_inplace}
0.5% 92.2% 0.051s 1.97e-05s Py 2594 1 AdvancedSubtensor
0.5% 92.7% 0.051s 3.90e-06s C 12970 5 DimShuffle{x}
0.5% 93.2% 0.049s 4.71e-06s C 10376 4 Elemwise{abs_,no_inplace}
0.5% 93.7% 0.049s 1.88e-05s Py 2594 1 AdvancedIncSubtensor{inplace=False, set_instead_of_inc=False}
0.5% 94.2% 0.044s 8.47e-06s C 5188 2 Elemwise{gt,no_inplace}
0.4% 94.7% 0.043s 5.48e-06s C 7782 3 DimShuffle{x,0}
0.4% 95.1% 0.042s 2.67e-06s C 15564 6 Elemwise{isnan,no_inplace}
0.4% 95.5% 0.037s 2.37e-06s C 15564 6 Elemwise{sqrt,no_inplace}
... (remaining 24 Ops account for 4.52%(0.43s) of the runtime)
Apply
------
<% time> <sum %> <apply time> <time per call> <#call> <id> <Mflops> <Gflops/s> <Apply name>
11.1% 11.1% 1.052s 4.05e-04s 2594 10 Dot22(x, W_dense1)
input 0: dtype=float32, shape=(20, 784), strides=c
input 1: dtype=float32, shape=(784, 256), strides=c
output 0: dtype=float32, shape=(20, 256), strides=c
8.0% 19.1% 0.761s 2.93e-04s 2594 156 MaxAndArgmax(Elemwise{Composite{[add(i0, mul(i1, i2))]}}.0, T
input 0: dtype=float32, shape=(784, 256), strides=c
input 1: dtype=int64, shape=(2,), strides=c
output 0: dtype=float32, shape=(), strides=c
output 1: dtype=int64, shape=(), strides=c
6.3% 25.4% 0.594s 2.29e-04s 2594 120 Sum(Elemwise{Composite{[sqr(Abs{output_types_preference=<clas
input 0: dtype=float32, shape=(784, 256), strides=c
output 0: dtype=float32, shape=(), strides=c
4.7% 30.1% 0.449s 1.73e-04s 2594 164 Elemwise{Composite{[sub(mul(i0, i1), mul(i2, i3))]}}(TensorCo
input 0: dtype=float32, shape=(1, 1), strides=c
input 1: dtype=float32, shape=(784, 256), strides=c
input 2: dtype=float32, shape=(1, 1), strides=c
input 3: dtype=float32, shape=(784, 256), strides=c
output 0: dtype=float32, shape=(784, 256), strides=c
4.4% 34.5% 0.415s 1.60e-04s 2594 22 Elemwise{add,no_inplace}(W_dense1, W_dense1_vel)
input 0: dtype=float32, shape=(784, 256), strides=c
input 1: dtype=float32, shape=(784, 256), strides=c
output 0: dtype=float32, shape=(784, 256), strides=c
4.0% 38.4% 0.378s 1.46e-04s 2594 152 Elemwise{Composite{[add(i0, mul(i1, i2))]}}(Dot22Scalar.0, Te
input 0: dtype=float32, shape=(784, 256), strides=c
input 1: dtype=float32, shape=(1, 1), strides=c
input 2: dtype=float32, shape=(784, 256), strides=c
output 0: dtype=float32, shape=(784, 256), strides=c
3.7% 42.1% 0.350s 1.35e-04s 2594 49 Dot22(Elemwise{mul,no_inplace}.0, W_dense2)
input 0: dtype=float32, shape=(20, 256), strides=c
input 1: dtype=float32, shape=(256, 256), strides=c
output 0: dtype=float32, shape=(20, 256), strides=c
3.4% 45.5% 0.322s 1.24e-04s 2594 93 Dot22(Elemwise{mul}.0, W_dense2.T)
input 0: dtype=float32, shape=(20, 256), strides=c
input 1: dtype=float32, shape=(256, 256), strides=(4, 1024)
output 0: dtype=float32, shape=(20, 256), strides=c
3.3% 48.8% 0.311s 1.20e-04s 2594 106 Dot22(x.T, Elemwise{mul}.0)
input 0: dtype=float32, shape=(784, 20), strides=(4, 3136)
input 1: dtype=float32, shape=(20, 256), strides=c
output 0: dtype=float32, shape=(784, 256), strides=c
3.3% 52.1% 0.311s 1.20e-04s 2594 146 Dot22Scalar(x.T, Elemwise{mul}.0, if{}.0)
input 0: dtype=float32, shape=(784, 20), strides=(4, 3136)
input 1: dtype=float32, shape=(20, 256), strides=c
input 2: dtype=float32, shape=(), strides=c
output 0: dtype=float32, shape=(784, 256), strides=c
3.3% 55.3% 0.310s 1.19e-04s 2594 144 MaxAndArgmax(Elemwise{Composite{[add(i0, mul(i1, i2))]}}.0, T
input 0: dtype=float32, shape=(256, 256), strides=c
input 1: dtype=int64, shape=(2,), strides=c
output 0: dtype=float32, shape=(), strides=c
output 1: dtype=int64, shape=(), strides=c
3.2% 58.5% 0.304s 5.86e-05s 5188 162 if{}(Elemwise{isnan,no_inplace}.0, Alloc.0, Elemwise{Composit
input 0: dtype=int8, shape=(), strides=c
input 1: dtype=float32, shape=no shape, strides=no strides
input 2: dtype=float32, shape=(784, 256), strides=c
output 0: dtype=float32, shape=(784, 256), strides=c
3.1% 61.7% 0.296s 1.14e-04s 2594 112 Elemwise{Composite{[sqr(Abs{output_types_preference=<class 't
input 0: dtype=float32, shape=(784, 256), strides=c
output 0: dtype=float32, shape=(784, 256), strides=c
2.1% 63.8% 0.200s 7.72e-05s 2594 104 Sum(Elemwise{Composite{[sqr(Abs{output_types_preference=<clas
input 0: dtype=float32, shape=(256, 256), strides=c
output 0: dtype=float32, shape=(), strides=c
1.9% 65.7% 0.180s 6.95e-05s 2594 92 Dot22(DimShuffle{1,0}.0, Elemwise{mul}.0)
input 0: dtype=float32, shape=(256, 20), strides=(4, 1024)
input 1: dtype=float32, shape=(20, 256), strides=c
output 0: dtype=float32, shape=(256, 256), strides=c
1.7% 67.4% 0.164s 6.32e-05s 2594 20 Elemwise{add,no_inplace}(W_dense2, W_dense2_vel)
input 0: dtype=float32, shape=(256, 256), strides=c
input 1: dtype=float32, shape=(256, 256), strides=c
output 0: dtype=float32, shape=(256, 256), strides=c
1.6% 69.0% 0.154s 5.92e-05s 2594 68 MaxAndArgmax(Elemwise{add,no_inplace}.0, TensorConstant{(1,)
input 0: dtype=float32, shape=(20, 10), strides=c
input 1: dtype=int64, shape=(1,), strides=c
output 0: dtype=float32, shape=(20,), strides=c
output 1: dtype=int64, shape=(20,), strides=c
1.5% 70.6% 0.146s 5.62e-05s 2594 158 Elemwise{Composite{[sub(mul(i0, i1), mul(i2, i3))]}}(TensorCo
input 0: dtype=float32, shape=(1, 1), strides=c
input 1: dtype=float32, shape=(256, 256), strides=c
input 2: dtype=float32, shape=(1, 1), strides=c
input 3: dtype=float32, shape=(256, 256), strides=c
output 0: dtype=float32, shape=(256, 256), strides=c
1.4% 72.0% 0.135s 2.61e-05s 5188 154 if{}(Elemwise{isnan,no_inplace}.0, Alloc.0, Elemwise{Composit
input 0: dtype=int8, shape=(), strides=c
input 1: dtype=float32, shape=no shape, strides=no strides
input 2: dtype=float32, shape=(256, 256), strides=c
output 0: dtype=float32, shape=(256, 256), strides=c
1.2% 73.2% 0.118s 4.53e-05s 2594 138 Elemwise{Composite{[add(i0, mul(i1, i2))]}}(Dot22Scalar.0, Te
input 0: dtype=float32, shape=(256, 256), strides=c
input 1: dtype=float32, shape=(1, 1), strides=c
input 2: dtype=float32, shape=(256, 256), strides=c
output 0: dtype=float32, shape=(256, 256), strides=c
... (remaining 145 Apply instances account for 26.77%(2.54s) of the runtime)
Error in atexit._run_exitfuncs:
Traceback (most recent call last):
File "/home/hadoop/lib/python2.7/atexit.py", line 24, in _run_exitfuncs
func(*targs, **kargs)
File "/home/hadoop/lib/python2.7/site-packages/theano/compile/profiling.py", line 65, in _atexit_print_fn
n_apply_to_print=config.profiling.n_apply)
File "/home/hadoop/lib/python2.7/site-packages/theano/compile/profiling.py", line 788, in summary
self.summary_memory(file, n_apply_to_print)
File "/home/hadoop/lib/python2.7/site-packages/theano/compile/profiling.py", line 601, in summary_memory
sh = self.variable_shape[out]
KeyError: Alloc.0
Error in sys.exitfunc:
Traceback (most recent call last):
File "/home/hadoop/lib/python2.7/atexit.py", line 24, in _run_exitfuncs
func(*targs, **kargs)
File "/home/hadoop/lib/python2.7/site-packages/theano/compile/profiling.py", line 65, in _atexit_print_fn
n_apply_to_print=config.profiling.n_apply)
File "/home/hadoop/lib/python2.7/site-packages/theano/compile/profiling.py", line 788, in summary
self.summary_memory(file, n_apply_to_print)
File "/home/hadoop/lib/python2.7/site-packages/theano/compile/profiling.py", line 601, in summary_memory
sh = self.variable_shape[out]
KeyError: Alloc.0
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment