提交 · c5da9237da9252e794115cd2e66d592ba385cf31 · testgroup / pytensor

13 7月, 2016 4 次提交
- Merge pull request #4726 from abergeron/fix_buildbot · c5da9237
  由 Frédéric Bastien 提交于 7月 12, 2016
```
Fix some problems in the buildbot
```
  c5da9237
- Fix problem with float64 for cuda. · 71c1bcbe
  由 Arnaud Bergeron 提交于 7月 12, 2016
  
  71c1bcbe
- Fix problem with testing unpickling. · f74a9d81
  由 Arnaud Bergeron 提交于 7月 12, 2016
  
  f74a9d81
- Fix deprecation warning about using non-integers. · a95f8af1
  由 Arnaud Bergeron 提交于 7月 12, 2016
  
  a95f8af1
12 7月, 2016 6 次提交
- Merge pull request #4683 from Faruk-Ahmed/gpuarray_tests · a96b5ded
  由 Frédéric Bastien 提交于 7月 12, 2016
```
gpuarray tests should pass when floatX=float32 and add this to the daily buildbot
```
  a96b5ded
- Merge pull request #4570 from Sentient07/new_graph2gpu · a24fd9bb
  由 Frédéric Bastien 提交于 7月 12, 2016
```
New graph2gpu
```
  a24fd9bb
- Merge pull request #4712 from Morgan243/master · 2b371c6f
  由 Frédéric Bastien 提交于 7月 12, 2016
```
Decode bytes to string before split
```
  2b371c6f
- Portable version of decode; swap 'dir' variable · daeb22d3
  由 Morgan Stuart 提交于 7月 11, 2016
  
  daeb22d3
- Merge pull request #4716 from nouiz/doc_cudnn · 9bf778f8
  由 Frédéric Bastien 提交于 7月 11, 2016
```
Add in the doc cuDNN BN and new CPU parallelized operations
```
  9bf778f8
- Fix docstring · aab3f35f
  由 Frederic Bastien 提交于 7月 11, 2016
  
  aab3f35f
11 7月, 2016 2 次提交
- Add in the doc cuDNN BN and new CPU parallelized operations · 9b1d9969
  由 Frederic Bastien 提交于 7月 11, 2016
  
  9b1d9969
- Fixing compilation of graphs with only one host_from_gpu OP · 195f9b1d
  由 sentient07 提交于 7月 11, 2016
  
  195f9b1d
10 7月, 2016 2 次提交
- Decode bytes to string before split · 7045e6a2
  由 Morgan Stuart 提交于 7月 09, 2016
  
  7045e6a2
- Corrected few import statements in scan module and fixed error with elemwise, fft and nerv · 345d9024
  由 sentient07 提交于 7月 09, 2016
  
  345d9024
09 7月, 2016 26 次提交
- Added fft opts to new opt, fixed broadcasting test, few cleanups · 3bd237f5
  由 sentient07 提交于 7月 09, 2016
  
  3bd237f5
- standardized opt name and return type of gpualloc and gpueye · ebcf5615
  由 sentient07 提交于 7月 09, 2016
  
  ebcf5615
- Missed out comments · aa853330
  由 sentient07 提交于 7月 08, 2016
  
  aa853330
- Added test for special case of get_scalar_constant_value · 430f0b6a
  由 sentient07 提交于 7月 08, 2016
  
  430f0b6a
- Final Cleanup · 0797283e
  由 sentient07 提交于 7月 08, 2016
  
  0797283e
- Fixed few errors generated from Dilation PR · 6548ed04
  由 sentient07 提交于 7月 06, 2016
```
Simplification

added fd parameter to tcase
```
  6548ed04
- Generalized filter dialation parameter and reverted a change · 9a9e873b
  由 sentient07 提交于 7月 04, 2016
  
  9a9e873b
- Changed interface of gpu_alloc_empty and few cleanups · 6701568f
  由 sentient07 提交于 7月 03, 2016
  
  6701568f
- Fix sparse tests · 8f602f6f
  由 Pascal Lamblin 提交于 6月 30, 2016
```
Sparse types do not have a "broadcastable" attribute.
```
  8f602f6f
- Make get_scalar_constant_value handle shape_i of broadcastable dimensions · 1c2f0f4a
  由 Pascal Lamblin 提交于 6月 28, 2016
  
  1c2f0f4a
- Reduced unwanted code · ea0d37ec
  由 sentient07 提交于 6月 24, 2016
  
  ea0d37ec
- Cached GpuFromHost · d7688cea
  由 sentient07 提交于 6月 23, 2016
  
  d7688cea
- Removed the unnecessary print statements and added missing documentation · 44480978
  由 sentient07 提交于 6月 23, 2016
  
  44480978
- Made the stricter optimization consistent · 8ea065e6
  由 sentient07 提交于 6月 22, 2016
  
  8ea065e6
- changed Flatten CPU Op with GPU Op · 8dfe6847
  由 sentient07 提交于 6月 22, 2016
  
  8dfe6847
- Better handling of tags and added back ShapeOptimizer to fast_compile · 6c611b5e
  由 sentient07 提交于 6月 21, 2016
  
  6c611b5e
- Revert "Handling tag 'cpu' in the middle of graph" · 6178fc1a
  由 sentient07 提交于 6月 21, 2016
```
This reverts commit 1736a6ce4ae50c6b3a6b4c286e1f23b8a43a0f33.
```
  6178fc1a
- Added docstrings and fixed two errors · dd544cba
  由 sentient07 提交于 6月 20, 2016
  
  dd544cba
- Handling tag 'cpu' in the middle of graph · bb6829ba
  由 sentient07 提交于 6月 20, 2016
  
  bb6829ba
- speed up compilation with exception_verbosity=high · 55ceca97
  由 Frederic Bastien 提交于 6月 17, 2016
  
  55ceca97
- Remove some call to host_from_gpu() in the op_lifter and make those opt only return gpu op. · 587d51fe
  由 Frederic Bastien 提交于 6月 17, 2016
  
  587d51fe
- Clean up and cached GpuCAReduceCuda · 1d8e6f26
  由 sentient07 提交于 6月 18, 2016
  
  1d8e6f26
- Made suggested changes, fixed travis · 4ebe109a
  由 sentient07 提交于 6月 18, 2016
  
  4ebe109a
- Added more opts to the new global opt · 643f5b1e
  由 sentient07 提交于 6月 17, 2016
  
  643f5b1e
- Fixed few error messages and cached GPUAlloc · d0d53136
  由 sentient07 提交于 6月 17, 2016
  
  d0d53136
- Fix for grad not implemented · 7946e06d
  由 sentient07 提交于 6月 17, 2016
  
  7946e06d