- 19 3月, 2014 1 次提交
-
-
由 Frederic 提交于
The preallocated output size for make_vector should ALWAYS be the good shape right now as it only depend on the ndim and that is constant in the graph and we only give preallocated output of the same node. This fix will be useful in the futur if we make other type of preallocated output that could be of the wrong size.
-
- 18 3月, 2014 7 次提交
-
-
由 abergeron 提交于
[BUG] in fast compile related to inplace opt.
-
由 Frederic 提交于
-
由 Frederic 提交于
-
由 abergeron 提交于
Add subgraph_grad
-
由 Nicholas Leonard 提交于
-
-
由 Nicholas Leonard 提交于
-
- 17 3月, 2014 4 次提交
- 14 3月, 2014 1 次提交
-
-
由 abergeron 提交于
Add tile in the doc.
-
- 13 3月, 2014 1 次提交
-
-
由 Frederic 提交于
-
- 12 3月, 2014 6 次提交
-
-
由 abergeron 提交于
Update the c_code_cache_version that was forgotten.
-
由 Frederic 提交于
-
由 Nicholas Leonard 提交于
-
-
由 Nicholas Leonard 提交于
-
由 Nicholas Leonard 提交于
-
- 11 3月, 2014 8 次提交
- 07 3月, 2014 2 次提交
-
-
由 abergeron 提交于
Speed up matVecModM function.
-
由 Marc-Alexandre Cote 提交于
-
- 04 3月, 2014 4 次提交
-
-
由 abergeron 提交于
Add consider_constant op
-
由 abergeron 提交于
Use pip --user instead of --install-option
-
由 Bart 提交于
-
由 Bart 提交于
I don't know if there is a specific reason that --install-option='--prefix=~/.local' is used instead of pip's --user, but it seems to break things when installing Theano without root access (both myself and Jae Hyun had this issue).
-
- 28 2月, 2014 6 次提交
-
-
由 Sander Dieleman 提交于
-
由 Sander Dieleman 提交于
Moved consider_constant from theano.tensor.extra_ops to theano.gradient, moved the optimization that removes it to theano.tensor.opt. Modified theano.compile.builders so it doesn't import theano.gradient, because theano.gradient now imports ViewOp from theano.compile and this would lead to a circular dependency.
-
由 Frédéric Bastien 提交于
Add new subtensor helper make_constant and use it to fix a bug in local_subtensor_merge.
-
由 abergeron 提交于
Fix gpu crash and faster optimization
-
由 Sander Dieleman 提交于
Add fast_compile tag so the optimization to remove the Op always runs. Use g_out.zeros_like instead of tensor.zeros_like so the op also works for sparse variables.
-
由 Arnaud Bergeron 提交于
-