- 26 Sep, 2023 1 commit
-
-
Artur Wojcik authored
-
- 15 Sep, 2023 1 commit
-
-
Umang Yadav authored
-
- 06 Feb, 2023 1 commit
-
-
Paul Fultz II authored
* Fuse layernorm with different patterns * Only match when using the last axis Co-authored-by:
kahmed10 <15948690+kahmed10@users.noreply.github.com> Co-authored-by:
kahmed10 <15948690+kahmed10@users.noreply.github.com>
-
- 31 Jan, 2023 1 commit
-
-
Paul Fultz II authored
* Add general optimize pass * Fuse gemm multiplies by scalar * Handle zero epsilon
-
- 19 Jan, 2023 1 commit
-
-
Paul Fultz II authored
This prevents multiple adds.
-
- 07 Dec, 2022 1 commit
-
-
Paul Fultz II authored
* Add implicit_conversion
-
- 21 Sep, 2022 1 commit
-
-
kahmed10 authored
This PR allows for other values of epsilon to be matched when finding layernorm. Similarly, the calculation now uses the variable for epsilon.
-
- 08 Sep, 2022 1 commit
-
-
Paul Fultz II authored
* Remove unused headers
-
- 17 Aug, 2022 1 commit
-
-
Paul Fultz II authored
-
- 22 Jun, 2022 1 commit
-
-
Ted Themistokleous authored
Updated each source file in the repo with the existing license.
-
- 11 May, 2022 1 commit
-
-
Paul Fultz II authored
Fuse layernorm and added triadd_layernorm fusion. This is a prep performance booster
-