- 11 9月, 2014 2 次提交
- 09 9月, 2014 13 次提交
-
-
由 abergeron 提交于
Make optimization more deterministic
-
由 abergeron 提交于
[MRG] fix gh-2091
-
由 Frédéric Bastien 提交于
Completed TODO in tensor.py:as_tensor_variable which said to use the App...
-
由 Dustin Webb 提交于
-
-
由 Frederic 提交于
-
由 Frederic 提交于
-
由 Frederic 提交于
Why do it use weakref anyway for the Link?
-
由 abergeron 提交于
Opt inc sub
-
由 Frederic 提交于
When optimizer=None, we had a crash at Theano import time. THEANO_FLAGS=optimizer=None,device=gpu python -c 'import theano'
-
由 abergeron 提交于
Fix crash in python 3.
-
由 Frederic 提交于
-
由 Frédéric Bastien 提交于
Completed TODO in as_tensor_variable which said to strip off leading bro...
-
- 08 9月, 2014 5 次提交
-
-
由 Frédéric Bastien 提交于
Doc
-
由 Frederic 提交于
-
由 Frederic 提交于
-
由 Frederic 提交于
-
由 Frederic 提交于
-
- 06 9月, 2014 8 次提交
-
-
由 Pascal Lamblin 提交于
Move Assert to the GPU in both back-end.
-
由 Frederic 提交于
-
由 Pascal Lamblin 提交于
Buildbot
-
由 Frederic 提交于
-
由 Frederic 提交于
-
由 Frederic 提交于
-
由 Frederic Bastien 提交于
-
由 Frederic Bastien 提交于
(so fix tests broken by the previous commit that use them more frequently)
-
- 05 9月, 2014 4 次提交
-
-
由 Dustin Webb 提交于
-
由 Pascal Lamblin 提交于
[CRASH] fix crash on the GPU with {inc,set}_subtensor and broadcasting t... -
由 Dustin Webb 提交于
Completed TODO in tensor.py:as_tensor_variable which said to use the Apply default output mechanism. Added tests to excited the various possible failure scenarios.
-
由 Dustin Webb 提交于
Added check to make sure the variable is of the user specified size after removing leading broadcastable dimensions.
-
- 04 9月, 2014 8 次提交
