提交 · 25bce5d718a650ab458eb275d8fac6caec89ff6f · testgroup / pytensor

05 12月, 2015 10 次提交
- Register the dnn optimizations for AbstractConv. · 25bce5d7
  由 Arnaud Bergeron 提交于 11月 26, 2015
  
  25bce5d7
- Remove GpuConv and everything that depends on it. · 01219a35
  由 Arnaud Bergeron 提交于 11月 26, 2015
  
  01219a35
- Some flake8 bugfixes. · 6f0af857
  由 Arnaud Bergeron 提交于 11月 24, 2015
  
  6f0af857
- Remove extra parentheses · 7a2a76c2
  由 Arnaud Bergeron 提交于 11月 24, 2015
  
  7a2a76c2
- Add a opt to deal with AbstractConvs that are partially on the GPU. · 515f7319
  由 Arnaud Bergeron 提交于 11月 23, 2015
  
  515f7319
- Add optimizations in the gpuarray backend for AbstractConv2d · 0ce3cc18
  由 Arnaud Bergeron 提交于 11月 23, 2015
  
  0ce3cc18
- Prevent crash with 1-input nodes. · 84120e95
  由 Arnaud Bergeron 提交于 11月 23, 2015
  
  84120e95
- Merge pull request #3738 from SinaHonari/issue3420 · 52185277
  由 Pascal Lamblin 提交于 12月 04, 2015
```
adding comment on negation
```
  52185277
- Merge pull request #3718 from Erotemic/enable_driver_test_config · aaa1c7e7
  由 Frédéric Bastien 提交于 12月 04, 2015
```
Added config variable to disable initial testing of nvidia drivers
```
  aaa1c7e7
- adding comment for multiple -1 values · d51e6057
  由 Sina Honari 提交于 12月 04, 2015
  
  d51e6057
04 12月, 2015 11 次提交
- Merge pull request #3510 from aalmah/ticket_3359 · 134d270d
  由 carriepl 提交于 12月 04, 2015
```
Add n>1 experiment in multinomial sampling on CPU
```
  134d270d
- Merge pull request #3736 from lamblin/fix_kron_test · 0db3f9af
  由 Pascal Lamblin 提交于 12月 03, 2015
```
Update test for recent scipy.
```
  0db3f9af
- adding comment on negation · 5bb13bbb
  由 Sina Honari 提交于 12月 03, 2015
  
  5bb13bbb
- Merge pull request #3734 from carriepl/cudandarray_data · 33bda7ca
  由 Frédéric Bastien 提交于 12月 03, 2015
```
Fix index error in scan backend
```
  33bda7ca
- Update test for recent scipy. · 05636181
  由 Pascal Lamblin 提交于 12月 03, 2015
```
We still support the old cases, even if scipy dropped them.
```
  05636181
- Merge pull request #3643 from SinaHonari/issue3420 · bfae9c44
  由 Pascal Lamblin 提交于 12月 03, 2015
```
Rewrite Reshape.infer_shape
```
  bfae9c44
- Merge pull request #3728 from nouiz/pool_grad_grad_infer_shape · a9655e21
  由 Pascal Lamblin 提交于 12月 03, 2015
```
[BUG] Fix DownsampleFactorMaxGradGrad.infer_shape. Fix a failing tests
```
  a9655e21
- Fix index error in scan backend · a6678feb
  由 carriepl 提交于 12月 03, 2015
  
  a6678feb
- Merge pull request #3729 from f0k/fix-conv2d-interface · 49ca2e00
  由 Pascal Lamblin 提交于 12月 03, 2015
```
Fix errors in conv2d() -> abstract_conv.conv2d() wrapper
```
  49ca2e00
- Merge pull request #3605 from fvisin/tag.test_value · cca20baf
  由 abergeron 提交于 12月 03, 2015
```
Add print_test_value_by_default option
```
  cca20baf
- Merge pull request #3726 from lamblin/gpuarray_syncok · 146bca86
  由 abergeron 提交于 12月 03, 2015
```
Add proper wait/record pairs for cudnn wrapper.
```
  146bca86
03 12月, 2015 19 次提交
- Fix errors in conv2d() -> abstract_conv.abstract_conv2d() wrapper · 0a2a7ca4
  由 f0k 提交于 12月 03, 2015
  
  0a2a7ca4
- fixing division issue in python 3 · d2e5db6b
  由 Sina Honari 提交于 12月 03, 2015
  
  d2e5db6b
- Remove useless code · 179d6fba
  由 Frederic 提交于 12月 03, 2015
  
  179d6fba
- [ENH] Simplify DownsampleFactorMax.out_shape that is used by its infer_shape when strides == ds · 4201bf7c
  由 Frederic 提交于 12月 03, 2015
  
  4201bf7c
- [BUG] fix infer_shape of DownsampleFactorMaxGradGrad. This fix the test… · e4089d8d
  由 Frederic 提交于 12月 03, 2015
```
[BUG] fix infer_shape of DownsampleFactorMaxGradGrad. This fix the test theano/sandbox/cuda/tests/test_blas.py:test_downsample.
```
  e4089d8d
- flake8 · 6f86147d
  由 Frederic 提交于 12月 01, 2015
  
  6f86147d
- Add proper wait/record pairs for cudnn wrapper. · 3109a772
  由 Pascal Lamblin 提交于 12月 02, 2015
```
Take 2.
```
  3109a772
- Merge pull request #3725 from Theano/revert-3724-gpuarray_syncok · ea96b166
  由 Pascal Lamblin 提交于 12月 02, 2015
```
Revert "Add proper wait/record pairs for cudnn wrapper."
```
  ea96b166
- Revert "Add proper wait/record pairs for cudnn wrapper." · b58cd010
  由 Pascal Lamblin 提交于 12月 02, 2015
  
  b58cd010
- Merge pull request #3724 from abergeron/gpuarray_syncok · 5c93d1cc
  由 Pascal Lamblin 提交于 12月 02, 2015
```
Add proper wait/record pairs for cudnn wrapper.
```
  5c93d1cc
- Merge pull request #3721 from ballasn/cond2d_interface_update · 51c6ef71
  由 Pascal Lamblin 提交于 12月 02, 2015
```
Cond2d interface update
```
  51c6ef71
- fixing error for scalar variables · 84169272
  由 Sina Honari 提交于 12月 02, 2015
  
  84169272
- update mode · 09d58a97
  由 Nicolas Ballas 提交于 12月 02, 2015
  
  09d58a97
- Merge pull request #3719 from mgermain/rc_debug_tests · 754baeb3
  由 abergeron 提交于 12月 02, 2015
```
Use np.cumprod instead of np.diag in OpDecoratorTests
```
  754baeb3
- Avoid __repr__ recursion · 968de8cd
  由 Francesco Visin 提交于 11月 05, 2015
  
  968de8cd
- Reuse theano.gof.op.get_test_value() and fix cuda variables · a763418b
  由 Francesco Visin 提交于 11月 05, 2015
  
  a763418b
- Add print_test_value_by_default option · be939078
  由 Francesco Visin 提交于 11月 03, 2015
```
Add option to print the tag.test_value with the variable name by
default.
```
  be939078
- Add proper wait/record pair for cudnn wrapper. · 1b72aa54
  由 Arnaud Bergeron 提交于 12月 02, 2015
  
  1b72aa54
- Added documentation of new config · 1c9f1cda
  由 joncrall 提交于 12月 02, 2015
  
  1c9f1cda