| Commit message (Collapse) | Author | Age |
| |
|
| |
|
| |
|
|
|
|
| |
previously disabled by mistake
|
|
|
|
| |
casting on GPUs
|
| |
|
| |
|
| |
|
|
|
|
| |
class: this enables users to leverage their existing threadpool when using eigen tensors.
|
|
|
|
| |
std::array
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
orders of magnitude on GPU
|
| |
|
|
|
|
| |
TensorDimensionList
|
| |
|
|
|
|
|
|
| |
several tensor operations.
Created the TensorDimensionList class to encode the list of all the dimensions of a tensor of rank n. This could be done using TensorIndexList, however TensorIndexList require cxx11 which isn't yet supported as widely as we'd like.
|
|
|
|
| |
failure
|
| |
|
|
|
|
| |
This also fixes underflow issues when scaling complex matrices through complex/complex operator.
|
| |
|
| |
|
|
|
|
| |
I observed significant speed-up of the CG solver.
|
| |
|
|
|
|
| |
the input matrix is complete (Upper|Lower) but column major.
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|