- 22 2月, 2012 26 次提交
-
-
由 nouiz 提交于
Removed useless constraint
-
由 Razvan Pascanu 提交于
Lop useless restriction Thanks Pascal ..
-
由 Pascal Lamblin 提交于
-
由 Pascal Lamblin 提交于
-
由 lamblin 提交于
Added name and splited the list into sections in NEWS.txt.
-
由 Frederic 提交于
-
由 Olivier Delalleau 提交于
Small updates in NEWS.txt
-
由 Pascal Lamblin 提交于
-
由 lamblin 提交于
Compiler
-
由 nouiz 提交于
Better error message if problem in givens
-
由 Pascal Lamblin 提交于
-
由 nouiz 提交于
Fix max'd error in EquilibriumOptimizer with constant folding
-
由 Frederic 提交于
-
由 Pascal Lamblin 提交于
Now constant folding does not bust it anymore.
-
由 Pascal Lamblin 提交于
-
由 Frederic 提交于
-
由 Pascal Lamblin 提交于
It used to wait for the first attempt to apply an optimization after it has already crossed the threshold. Now, it signals that the optimizer should stop earlier, just after crossing that threshold.
-
由 Frederic 提交于
This is needed as we changed the signature of some function and thus we should force the recompilation of all gpu op.
-
由 Frederic 提交于
It seam more logical to associate it with the compiler then the Type. This also make cuda_ndarray.cu compiled with it.
-
由 Frederic 提交于
-
由 Frederic 提交于
-
由 Frederic 提交于
-
由 Frederic 提交于
make {nvcc,gcc}_module_compile_str a class with another function compile_args that get added in the keys. -
由 Frederic 提交于
-
由 Frederic 提交于
-
由 Frederic 提交于
-
- 21 2月, 2012 14 次提交
-
-
由 Pascal Lamblin 提交于
Instead of using current size of the graph, use the size the graph had at its largest. This enables optimizations like constant-folding to reduce greatly the total number of nodes in the graph without triggering an error.
-
由 lamblin 提交于
Fixed circular import issue
-
由 Olivier Delalleau 提交于
Sorry, I should have tested more thoroughly this stuff with GPU before pushing it :/
-
由 nouiz 提交于
Fix tolerance in Usmm test.
-
由 Pascal Lamblin 提交于
-
由 Pascal Lamblin 提交于
-
由 nouiz 提交于
Added missing import
-
由 Olivier Delalleau 提交于
-
由 nouiz 提交于
Fix when forcing cuda to be disabled
-
由 nouiz 提交于
Fixes for Python 2.4
-
由 Olivier Delalleau 提交于
-
由 Olivier Delalleau 提交于
The GpuOp class must be available even when Cuda is disabled (e.g. by forcing cuda_available = False). Otherwise some tests fail. This commit moves the definition of GpuOp directly into cuda/__init__.py so that it works.
-
由 nouiz 提交于
Minor fixes
-
由 Olivier Delalleau 提交于
-