Commit 0ed6745
metal: optimise
* optimise GGML_OP_SUM
* add non-contiguous tests by permuting the input
* change tests to require full contiguity of OP_SUM
* cuda : add check GGML_OP_SUM
---------
Co-authored-by: Georgi Gerganov <[email protected]>GGML_OP_SUM (ggml-org#16559)1 parent 5248535 commit 0ed6745
File tree
5 files changed
+71
-11
lines changed- ggml/src
- ggml-cuda
- ggml-metal
- tests
5 files changed
+71
-11
lines changed| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
3633 | 3633 | | |
3634 | 3634 | | |
3635 | 3635 | | |
3636 | | - | |
3637 | 3636 | | |
3638 | 3637 | | |
| 3638 | + | |
| 3639 | + | |
3639 | 3640 | | |
3640 | 3641 | | |
3641 | 3642 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
662 | 662 | | |
663 | 663 | | |
664 | 664 | | |
| 665 | + | |
665 | 666 | | |
666 | 667 | | |
667 | 668 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
866 | 866 | | |
867 | 867 | | |
868 | 868 | | |
| 869 | + | |
| 870 | + | |
| 871 | + | |
| 872 | + | |
| 873 | + | |
| 874 | + | |
| 875 | + | |
| 876 | + | |
| 877 | + | |
| 878 | + | |
| 879 | + | |
869 | 880 | | |
870 | 881 | | |
871 | 882 | | |
872 | 883 | | |
873 | 884 | | |
874 | | - | |
| 885 | + | |
| 886 | + | |
| 887 | + | |
875 | 888 | | |
876 | 889 | | |
877 | 890 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
1727 | 1727 | | |
1728 | 1728 | | |
1729 | 1729 | | |
1730 | | - | |
| 1730 | + | |
| 1731 | + | |
| 1732 | + | |
| 1733 | + | |
| 1734 | + | |
| 1735 | + | |
1731 | 1736 | | |
1732 | | - | |
| 1737 | + | |
1733 | 1738 | | |
1734 | 1739 | | |
1735 | 1740 | | |
1736 | | - | |
1737 | | - | |
1738 | | - | |
| 1741 | + | |
| 1742 | + | |
| 1743 | + | |
| 1744 | + | |
| 1745 | + | |
| 1746 | + | |
1739 | 1747 | | |
1740 | 1748 | | |
1741 | | - | |
| 1749 | + | |
| 1750 | + | |
| 1751 | + | |
| 1752 | + | |
| 1753 | + | |
| 1754 | + | |
| 1755 | + | |
| 1756 | + | |
| 1757 | + | |
| 1758 | + | |
| 1759 | + | |
| 1760 | + | |
| 1761 | + | |
| 1762 | + | |
| 1763 | + | |
| 1764 | + | |
| 1765 | + | |
| 1766 | + | |
| 1767 | + | |
| 1768 | + | |
| 1769 | + | |
| 1770 | + | |
| 1771 | + | |
1742 | 1772 | | |
1743 | 1773 | | |
1744 | 1774 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
4732 | 4732 | | |
4733 | 4733 | | |
4734 | 4734 | | |
| 4735 | + | |
| 4736 | + | |
4735 | 4737 | | |
4736 | 4738 | | |
4737 | | - | |
| 4739 | + | |
| 4740 | + | |
| 4741 | + | |
4738 | 4742 | | |
4739 | 4743 | | |
4740 | 4744 | | |
4741 | | - | |
4742 | | - | |
| 4745 | + | |
| 4746 | + | |
| 4747 | + | |
| 4748 | + | |
4743 | 4749 | | |
4744 | 4750 | | |
4745 | 4751 | | |
4746 | 4752 | | |
4747 | 4753 | | |
4748 | 4754 | | |
| 4755 | + | |
| 4756 | + | |
| 4757 | + | |
| 4758 | + | |
| 4759 | + | |
4749 | 4760 | | |
4750 | 4761 | | |
4751 | 4762 | | |
| |||
6876 | 6887 | | |
6877 | 6888 | | |
6878 | 6889 | | |
| 6890 | + | |
| 6891 | + | |
| 6892 | + | |
6879 | 6893 | | |
6880 | 6894 | | |
6881 | 6895 | | |
| |||
6886 | 6900 | | |
6887 | 6901 | | |
6888 | 6902 | | |
| 6903 | + | |
6889 | 6904 | | |
6890 | 6905 | | |
6891 | 6906 | | |
| |||
0 commit comments