提交 · 4ac4e2048ba5bcf8d266f6d16f4ece68a4302a32 · testgroup / pytensor

31 1月, 2013 2 次提交
- make sure numpy.nonzero returns tuples with length > 0 before numpy.vstack · 4ac4e204
  由 Jeremiah Lowin 提交于 1月 30, 2013
  
  4ac4e204
- avoid vstack of 0-length arrays; cast result as int · f54031ba
  由 Jeremiah Lowin 提交于 1月 30, 2013
  
  f54031ba
30 1月, 2013 4 次提交
- guarantee zeros in tests · 99197720
  由 Jeremiah Lowin 提交于 1月 30, 2013
  
  99197720
- missing if/else for return_matrix · 0cea723d
  由 Jeremiah Lowin 提交于 1月 30, 2013
  
  0cea723d
- fix typo in docstring · cc45e0ac
  由 Jeremiah Lowin 提交于 1月 29, 2013
  
  cc45e0ac
- change Op to return matrix; change function to return either tuple or matrix · eefac0b4
  由 Jeremiah Lowin 提交于 1月 29, 2013
  
  eefac0b4
29 1月, 2013 13 次提交
- add nonzero_values and test, also update nonzero docstring to reflect new indexing behavior · 1d9e144b
  由 Jeremiah Lowin 提交于 1月 29, 2013
  
  1d9e144b
- add flatnonzero and tests · c74b0283
  由 Jeremiah Lowin 提交于 1月 29, 2013
  
  c74b0283
- add 1D test for nonzero · 8f8b6e1f
  由 Jeremiah Lowin 提交于 1月 29, 2013
  
  8f8b6e1f
- move Nonzero docstring from the function to the class · 46bfeaf6
  由 Jeremiah Lowin 提交于 1月 29, 2013
  
  46bfeaf6
- add nonzero test · 3c0b37ab
  由 Jeremiah Lowin 提交于 1月 29, 2013
  
  3c0b37ab
- add Nonzero Op, function, tensor method · fb059696
  由 Jeremiah Lowin 提交于 1月 29, 2013
  
  fb059696
- add triangle tests · 652d9e7a
  由 Jeremiah Lowin 提交于 1月 29, 2013
  
  652d9e7a
- add Tri op, tri/tril/triu functions · c819f527
  由 Jeremiah Lowin 提交于 1月 29, 2013
  
  c819f527
- Merge pull request #1180 from lamblin/fix_infinite_opt_loop · bbc25bc6
  由 nouiz 提交于 1月 28, 2013
```
Fix infinite opt loop
```
  bbc25bc6
- Update following code review · d5c045e4
  由 Pascal Lamblin 提交于 1月 28, 2013
  
  d5c045e4
- Add DestroyHandler if there are inplace ops · 9e2f5309
  由 Pascal Lamblin 提交于 1月 28, 2013
```
Even in DebugMode, even before the inplace phase.
```
  9e2f5309
- Limit number calls to global optimizers · 7b51cfba
  由 Pascal Lamblin 提交于 1月 28, 2013
```
It helps locating infinite loops
```
  7b51cfba
- Small changes to ease debugging · ec3726ed
  由 Pascal Lamblin 提交于 1月 28, 2013
  
  ec3726ed
26 1月, 2013 4 次提交
- Merge pull request #986 from jlowin/dot_test · d13064f7
  由 lamblin 提交于 1月 25, 2013
```
Ready for merge: Streamline dot & tensordot / Allow dot products of n-dimensional variables
```
  d13064f7
- Merge pull request #1178 from pascanur/nicolasDLT_scan_bug · 9880362c
  由 lamblin 提交于 1月 25, 2013
```
Fix bugs reported by the DL tutorial wrote by Nicolas
```
  9880362c
- Merge pull request #1176 from abalkin/issue-1175 · 3420f416
  由 lamblin 提交于 1月 25, 2013
```
Fixed a race condition.
```
  3420f416
- Restored stdin redirection. · f5640c04
  由 abalkin 提交于 1月 25, 2013
  
  f5640c04
25 1月, 2013 2 次提交
- fix push dot out optimization · 780d7a43
  由 Razvan Pascanu 提交于 1月 25, 2013
  
  780d7a43
- fix connection pattern · dafaffaa
  由 Razvan Pascanu 提交于 1月 25, 2013
  
  dafaffaa
24 1月, 2013 4 次提交
- Merge pull request #1173 from nouiz/fix_crash_roll · 703d535a
  由 lamblin 提交于 1月 23, 2013
```
Fix crash tensor.roll(Var, iscalar) reported by Jeremiah Lowin.
```
  703d535a
- Removed a comment made obsolete by the previous commit. · eedd00fa
  由 abalkin 提交于 1月 23, 2013
  
  eedd00fa
- Fixed a race condition. · 5d3c9873
  由 abalkin 提交于 1月 23, 2013
  
  5d3c9873
- Fix memory leak in HostFromGpu when allow_gc=False! · 52be4733
  由 Frederic 提交于 1月 23, 2013
```
fix gh-1154
```
  52be4733
23 1月, 2013 2 次提交
- Added link to Nicolas Pinto ebuild on gentoo. · 4021fc9e
  由 Frederic 提交于 1月 22, 2013
  
  4021fc9e
- Fix crash tensor.roll(Var, iscalar) reported by Jeremiah Lowin. · c55d0ab9
  由 Frederic 提交于 1月 22, 2013
  
  c55d0ab9
21 1月, 2013 1 次提交
- Merge pull request #1171 from abalkin/headers-not-found · a696539d
  由 nouiz 提交于 1月 21, 2013
```
Issue #1164: Fixed failing test_pycuda_example.
```
  a696539d
19 1月, 2013 8 次提交
- Issue #1164: Fixed failing test_pycuda_example. · 4f2bb0a9
  由 abalkin 提交于 1月 18, 2013
```
Removed unnecessary #includes from generated cuda code.
```
  4f2bb0a9
- Merge pull request #1170 from nouiz/denormal · b7e7be45
  由 lamblin 提交于 1月 18, 2013
```
Denormal
```
  b7e7be45
- fix typo. · a18d8455
  由 Frederic 提交于 1月 18, 2013
  
  a18d8455
- Make GpuSoftmaxWithBias don't crash on GTX285 with some shapes. · 56b301a3
  由 Frederic 提交于 1月 18, 2013
  
  56b301a3
- make GpuSoftmax don't crash when input shape contains 0. · d0040637
  由 Frederic 提交于 1月 18, 2013
  
  d0040637
- better error msg and comments for gpusoftmax · defef0a5
  由 Frederic 提交于 1月 18, 2013
  
  defef0a5
- pep8 · 3e2b3f10
  由 Frederic 提交于 1月 18, 2013
  
  3e2b3f10
- Make GpuSoftmax don't crash on GTX285 GPU in somes cases. · 5e93e6b8
  由 Frederic 提交于 1月 18, 2013
  
  5e93e6b8