提交 · 22c56b8b702a460408afa2486d3d47fb025dfbce · testgroup / pytensor

19 12月, 2015 3 次提交
- Skip part of tests with not compatible pydot_ng · 22c56b8b
  由 Frederic Bastien 提交于 12月 18, 2015
  
  22c56b8b
- Don't crash when generating the error for some type of error · 2ba1d48c
  由 Frederic 提交于 12月 18, 2015
  
  2ba1d48c
- Merge pull request #3784 from mronian/master · a5c6a13a
  由 Frédéric Bastien 提交于 12月 18, 2015
```
Adds test for uint datatype for image_shape and filter_shape in thean…
```
  a5c6a13a
18 12月, 2015 20 次提交
- Merge pull request #3792 from abergeron/gpuarray_prealloc · 26400567
  由 Frédéric Bastien 提交于 12月 18, 2015
```
Gpuarray prealloc
```
  26400567
- Adds test for uint datatype for image_shape and filter_shape in… · 0a876e23
  由 mronian 提交于 12月 17, 2015
```
Adds test for uint datatype for image_shape and filter_shape in theano.tensor.nnet.conv and also checks if datatype is in tensor.discrete_dtypes
```
  0a876e23
- Merge pull request #3559 from abergeron/multi_gpu_doc · 54e96754
  由 Pascal Lamblin 提交于 12月 17, 2015
```
Multi gpu doc
```
  54e96754
- Delete the cudnn handles when we no longer need them. · e6ca1825
  由 Arnaud Bergeron 提交于 12月 17, 2015
  
  e6ca1825
- Rework the gpuarray api version check. · d6b2e034
  由 Arnaud Bergeron 提交于 12月 17, 2015
  
  d6b2e034
- Add the gpuarray.preallocate configuration variable. This will help fix out of… · 8fb5d66c
  由 Arnaud Bergeron 提交于 12月 16, 2015
```
Add the gpuarray.preallocate configuration variable.  This will help fix out of memory errors in certain sequences of tests.
```
  8fb5d66c
- Merge pull request #3788 from nouiz/carriepl-v4 · 4ad36ddc
  由 Frédéric Bastien 提交于 12月 17, 2015
```
Rebased cudnn v4
```
  4ad36ddc
- Fix typo. · c5084ac8
  由 Arnaud Bergeron 提交于 12月 01, 2015
  
  c5084ac8
- Remove the warning paragraph as the issues have been fixed. · 47ecce09
  由 Arnaud Bergeron 提交于 11月 30, 2015
  
  47ecce09
- Fix typos. · e28057a7
  由 Arnaud Bergeron 提交于 10月 28, 2015
  
  e28057a7
- Move the warning about no speedup earlier and add a section showing · 838508a9
  由 Arnaud Bergeron 提交于 10月 28, 2015
```
the printout of the context map.
```
  838508a9
- Fixes from review. · 0f4f011b
  由 Arnaud Bergeron 提交于 10月 27, 2015
  
  0f4f011b
- Link to PureType instead of duplicating the docstrings. · 99e23d19
  由 Arnaud Bergeron 提交于 10月 27, 2015
  
  99e23d19
- Fix the multi gpu script to use the new target argument. · f2c85754
  由 Arnaud Bergeron 提交于 10月 23, 2015
  
  f2c85754
- Add a tutorial section on how tu use the new multi-gpu functionality. · 7beebd0c
  由 Arnaud Bergeron 提交于 10月 23, 2015
  
  7beebd0c
- Delete some old and useless documentation. · d6156c63
  由 Arnaud Bergeron 提交于 10月 23, 2015
  
  d6156c63
- Rest of libdoc for gpuarray. · 30617ff5
  由 Arnaud Bergeron 提交于 10月 23, 2015
  
  30617ff5
- Add a listing of ops for gpuarray. · 3bf6f4cb
  由 Arnaud Bergeron 提交于 10月 23, 2015
  
  3bf6f4cb
- Add a bunch of documentation to gpuarray/type.py · f005cc21
  由 Arnaud Bergeron 提交于 10月 23, 2015
  
  f005cc21
- Link in the blocksparse docs and add a section about the gpuarray backend. · 418967de
  由 Arnaud Bergeron 提交于 10月 23, 2015
  
  418967de
17 12月, 2015 1 次提交
- Merge pull request #3786 from lamblin/abstract_conv_infershape · 22db3930
  由 Frédéric Bastien 提交于 12月 16, 2015
```
Abstract conv: fix tests and add infer_shape
```
  22db3930
16 12月, 2015 16 次提交
- Merge pull request #3787 from oeway/master · 725b7a3f
  由 Frédéric Bastien 提交于 12月 16, 2015
```
fixed cula runtime error (libcula.so not found)
```
  725b7a3f
- Flake8 · 32e113c1
  由 carriepl 提交于 11月 25, 2015
  
  32e113c1
- Make dnn conv in gpuarray work with AbstractConv · ced46b7d
  由 carriepl 提交于 11月 25, 2015
  
  ced46b7d
- Standardize documentation of the 'precision' parameter · aafc23aa
  由 carriepl 提交于 11月 12, 2015
  
  aafc23aa
- Update and standardize error messages · 7d44a89c
  由 carriepl 提交于 11月 12, 2015
  
  7d44a89c
- Clean up V2 code · 524438b6
  由 carriepl 提交于 11月 12, 2015
  
  524438b6
- Avoid using method cudnnGetConvolution2dDescriptor() method · 877f4210
  由 carriepl 提交于 11月 10, 2015
  
  877f4210
- Make deprecated dnn flags raise more meaningfull error messages. · ebd97785
  由 carriepl 提交于 11月 10, 2015
  
  ebd97785
- Add 'as_input' option to dnn.conv.precision flag · 32f11551
  由 carriepl 提交于 11月 10, 2015
  
  32f11551
- Drop support for v2 · 7ff6621e
  由 carriepl 提交于 11月 09, 2015
  
  7ff6621e
- Don't restrict 'small' workspace implementation of GradW to v4 and later · 3e4da0dc
  由 carriepl 提交于 11月 09, 2015
  
  3e4da0dc
- Add missing syntax · a28cb169
  由 carriepl 提交于 11月 09, 2015
  
  a28cb169
- Fix verification of spatial sizes in cudnn conv (gpuarray backend) · 757b04c0
  由 carriepl 提交于 10月 27, 2015
  
  757b04c0
- Fix condition based on CuDNN version · fdffa58e
  由 carriepl 提交于 10月 27, 2015
  
  fdffa58e
- Call cudnnGetConvolution2dDescriptor instead of cudnnGetConvolutionNdDescriptor · 6f1dc3d8
  由 carriepl 提交于 10月 26, 2015
  
  6f1dc3d8
- Add comments and fix condition in GpuDnnConvGradW C code (gpua backend) · bf918482
  由 carriepl 提交于 10月 22, 2015
  
  bf918482