提交 · 879b8716343ce9808a78c00ef6a55ad58ce8ce34 · testgroup / pytensor

05 5月, 2016 21 次提交
- test_extra_ops.py has been modified in order to respect the flake8 style. · 879b8716
  由 Chiheb Trabelsi 提交于 3月 15, 2016
  
  879b8716
- test_driver has been modified in order to respect the flake8 style. · 0a904644
  由 Chiheb Trabelsi 提交于 3月 15, 2016
  
  0a904644
- test_cuda_ndarray.py has been modified in order to respect the flake8 style. · 19a6bd40
  由 Chiheb Trabelsi 提交于 3月 15, 2016
  
  19a6bd40
- test_conv_cuda_ndarray.py has been modified in order to respect the flake8 style. · 791d4871
  由 Chiheb Trabelsi 提交于 3月 15, 2016
  
  791d4871
- test_blas.py has been modified in order to respect the flake8 style. · 186b90a0
  由 Chiheb Trabelsi 提交于 3月 15, 2016
  
  186b90a0
- test_bench_loopfusion.py has been modified in order to respect the flake8 style. · 155c4e01
  由 Chiheb Trabelsi 提交于 3月 15, 2016
  
  155c4e01
- test_basic_ops.py has been modified in order to respect the flake8 style. · a2503274
  由 Chiheb Trabelsi 提交于 3月 15, 2016
  
  a2503274
- var.py has been modified in order to respect the flake8 style. · 6d8bd41f
  由 Chiheb Trabelsi 提交于 3月 14, 2016
  
  6d8bd41f
- type.py has been modified in order to respect the flake8 style. · fa143f13
  由 Chiheb Trabelsi 提交于 3月 14, 2016
  
  fa143f13
- rng_curand.py has been modified in order to respect the flake8 style. · 4fd12c42
  由 Chiheb Trabelsi 提交于 3月 14, 2016
  
  4fd12c42
- opt.py has been modified in order to respect the flake8 style. · 54fe4a7f
  由 Chiheb Trabelsi 提交于 3月 14, 2016
  
  54fe4a7f
- nvcc_compiler.py has been modified in order to respect the flake8 style. · 1a3948cc
  由 Chiheb Trabelsi 提交于 3月 14, 2016
```
nvcc_compiler.py do not contain long lines.
```
  1a3948cc
- nnet.py has been modified in order to respect the flake8 style · 755d7cdf
  由 Chiheb Trabelsi 提交于 3月 14, 2016
```
nnet.py has been fixed

nnet.py do not contain long lines.
```
  755d7cdf
- neighbours.py has been modified in order to respect the flake8 style · 06adacd3
  由 Chiheb Trabelsi 提交于 3月 14, 2016
  
  06adacd3
- kernet_codegen.py has been modified in order to respect the flake8 style. · 2ea6dbac
  由 Chiheb Trabelsi 提交于 3月 14, 2016
```
kernel_codegen.py do not contain long lines.
```
  2ea6dbac
- GpuConvTransp3D has been modified in order to respect the flake8 style. · e4ebf518
  由 Chiheb Trabelsi 提交于 3月 14, 2016
```
GpuConvTransp3D.py has been fixed.

GpuConvTransp3D.py do not contain long lines.
```
  e4ebf518
- GpuConvGrad3D has been modified in order to respect the flake8 style. · 11b3e525
  由 Chiheb Trabelsi 提交于 3月 14, 2016
```
GpuConvGrad3D.py has been fixed

GpuConvGrad3D.py do not contain long lines.
```
  11b3e525
- fftconv.py has been modified in order to respect the flake8 style. · 7c171df5
  由 Chiheb Trabelsi 提交于 3月 14, 2016
  
  7c171df5
- elemwise.py has been modified in order to respect the flake8 style. · 25df3acb
  由 Chiheb Trabelsi 提交于 3月 14, 2016
```
elemwise.py do not contain long lines.
```
  25df3acb
- blas.py has been modified in order to respect the flake8 style. · 27c8da22
  由 Chiheb Trabelsi 提交于 3月 14, 2016
```
blas.py do not contain long lines.
```
  27c8da22
- basic_ops.py has been modified in order to respect the flake8 style · ed7759fb
  由 Chiheb Trabelsi 提交于 3月 12, 2016
  
  ed7759fb
04 5月, 2016 1 次提交
- Merge pull request #4443 from nouiz/grad_abs · 200babca
  由 Pascal Lamblin 提交于 5月 03, 2016
```
Change the grad of abs to prevent relu from generating nan in fast_compile
```
  200babca
03 5月, 2016 2 次提交
- Merge pull request #4450 from nouiz/deepworx-pr1 · 46fbfeb6
  由 Frédéric Bastien 提交于 5月 02, 2016
```
Remove upcast to float64 in the graph.
```
  46fbfeb6
- Small fixup to the previous commit · 3974f9b4
  由 Frederic Bastien 提交于 5月 02, 2016
  
  3974f9b4
02 5月, 2016 4 次提交
- Be sure to use the good dtype and not a possibly modified one from get_scalar_constant_value. · 939076b5
  由 Frederic Bastien 提交于 5月 02, 2016
  
  939076b5
- Do the same fix for all opt path and for prod. · e3b3ecc1
  由 Frederic Bastien 提交于 5月 02, 2016
  
  e3b3ecc1
- Better test error message. · 2d13ec42
  由 Frederic Bastien 提交于 5月 02, 2016
  
  2d13ec42
- small modif · 9ba08fd0
  由 Frederic Bastien 提交于 5月 02, 2016
  
  9ba08fd0
30 4月, 2016 6 次提交
- Small doc update from Simon. · efe85d2a
  由 Frederic Bastien 提交于 4月 29, 2016
  
  efe85d2a
- Merge pull request #4378 from shabanian/bn · 1bd371ce
  由 Frédéric Bastien 提交于 4月 29, 2016
```
Disable test values in BNComposite and get_substream_rstates
```
  1bd371ce
- Add in the faq a section on small float variation · 75865306
  由 Frederic Bastien 提交于 4月 29, 2016
  
  75865306
- Fix grad of abs to have relu don't return nan in fast_compile · b1ff3235
  由 Frederic Bastien 提交于 4月 29, 2016
  
  b1ff3235
- tests the fix for 'float64' in RNNs, wastes memory and breaks when warn_float64=raise · dd58fb00
  由 p 提交于 4月 29, 2016
  
  dd58fb00
- Remove duplicate information · 9c1d79d0
  由 Samira Shabanian 提交于 4月 29, 2016
  
  9c1d79d0
29 4月, 2016 2 次提交
- Merge pull request #4422 from tsirif/develop · adffdc1c
  由 Frédéric Bastien 提交于 4月 29, 2016
```
Develop SearchsortedOp to wrap numpy's searchsorted function
```
  adffdc1c
- Raise ValueError on invalid `side` kwd of searchsorted · 6506ecd9
  由 Christos Tsirigotis 提交于 4月 29, 2016
```
- Add extra tests for searchsorted
```
  6506ecd9
28 4月, 2016 2 次提交
- fixes the 'float64' in RNNs, wastes memory and breaks when warn_float64=raise · f777fd27
  由 p 提交于 4月 27, 2016
  
  f777fd27
- Merge pull request #4413 from nouiz/harmdevries89-multinomial_newbackend · 5e5e5cc5
  由 Frédéric Bastien 提交于 4月 27, 2016
```
multinomial newbackend
```
  5e5e5cc5
27 4月, 2016 2 次提交
- Fix range of values in test_extra_ops:setUp · 8bffb72e
  由 Christos Tsirigotis 提交于 4月 26, 2016
  
  8bffb72e
- Fix flake8 in extra_ops and tests · 15db4a1f
  由 Christos Tsirigotis 提交于 4月 24, 2016
```
- Resolve :215 of test_extra_ops to int explicitly
Reason: numpy.uint64 + int -> numpy.float64 for some reason (numpy/numpy@3509704)
```
  15db4a1f