systemml git commit: [MINOR] Utility to obtain the exact output sparsity of sparse products

2018-10-20 Thread mboehm7
Repository: systemml Updated Branches: refs/heads/master ab8cccdff -> cca6356f8 [MINOR] Utility to obtain the exact output sparsity of sparse products Project: http://git-wip-us.apache.org/repos/asf/systemml/repo Commit: http://git-wip-us.apache.org/repos/asf/systemml/commit/cca6356f Tree:

systemml git commit: [MINOR] Fixes baseline sparsity estimators (layered graph, bitset)

2018-10-20 Thread mboehm7
Repository: systemml Updated Branches: refs/heads/master 07650acf2 -> ab8cccdff [MINOR] Fixes baseline sparsity estimators (layered graph, bitset) This patch fixes (1) the selection of bitset implementations according to input datasize, and (2) operation-specific API of the layered graph.

systemml git commit: [SYSTEMML-2479] Fix MNC sparsity estimator reshape operations

2018-10-20 Thread mboehm7
Repository: systemml Updated Branches: refs/heads/master bd34292d4 -> 07650acf2 [SYSTEMML-2479] Fix MNC sparsity estimator reshape operations This patch fixes various smaller correctness issues of MNC sketch propagation for reshape operations, adds a related test, and finally removes the

[1/2] systemml git commit: [SYSTEMML-445] Support non-CuDNN GPU operator for LSTM forward and backward

2018-10-20 Thread niketanpansare
Repository: systemml Updated Branches: refs/heads/master ef842da9c -> bd34292d4 http://git-wip-us.apache.org/repos/asf/systemml/blob/bd34292d/src/main/java/org/apache/sysml/runtime/instructions/gpu/DnnGPUInstruction.java --

[2/2] systemml git commit: [SYSTEMML-445] Support non-CuDNN GPU operator for LSTM forward and backward

2018-10-20 Thread niketanpansare
[SYSTEMML-445] Support non-CuDNN GPU operator for LSTM forward and backward - Added corresponding GPU tests that compare the result of CuDNN operator with the newly added operator. Also, the results are compared with DML-bodied LSTM implementation in the nn layer. - The LSTM forward operator