Gaurav Shukla 1be604bfd3 [LINALG] Lower `aten.Matmul` to `linalg.BatchMatmul` ... This commit lowers `aten.matmul` to `linalg.BatchMatmul` under the following conditions: 1. The result of matrix multiplication must have batch dimensions, i.e., rank greater than 2. 2. The resultant matrix must have at most 1 dynamic batch dimension. It also handles broadcasting of batch dimensions when batch dimensions of the matrices are broadcastable. Signed-Off-by: Gaurav Shukla <gaurav@nod-labs.com>		2022-06-25 10:58:06 +05:30
..
CMakeLists.txt	Bump LLVM (#958)	2022-06-22 22:23:46 -05:00
DataMovement.cpp	[LINALG] Lower `aten.Matmul` to `linalg.BatchMatmul`	2022-06-25 10:58:06 +05:30
IndirectDataMovement.cpp	[LINALG] Fix shape function of index.Tensor + support N-rank inputs (#972)	2022-06-24 09:45:44 -07:00
Linear.cpp	[LINALG] Lower `aten.Matmul` to `linalg.BatchMatmul`	2022-06-25 10:58:06 +05:30
Pooling.cpp	[MLIR][TORCH] Add decomposition of aten.adaptive_avg_pool2d op	2022-05-27 07:56:37 +05:30
PopulatePatterns.h	Temporarily revert support for custom op extensions. (#944)	2022-06-14 18:24:40 -07:00
Random.cpp	Improve "pseudo" op terminology.	2022-03-15 17:57:52 -07:00
Reduction.cpp	torch,linalg: add support for translating aten.linalg.vector_norm (#839)	2022-05-19 15:48:15 -07:00
TensorConstructors.cpp	[LINALG] Add support for preserve memory format in aten_empty_like op.	2022-05-10 09:37:55 +05:30
TensorScalarInterop.cpp	[MLIR][TORCH] Add E2E support for ScalarImplicit, Int.Scalar op	2022-05-10 22:40:49 +05:30
TorchToLinalg.cpp	Temporarily revert support for custom op extensions. (#944)	2022-06-14 18:24:40 -07:00
Uncategorized.cpp	linalg: lower `aten.triu` op to `linalg.generic` (#965)	2022-06-23 22:45:48 -07:00
Utils.cpp	[LINALG] Lower `aten.Matmul` to `linalg.BatchMatmul`	2022-06-25 10:58:06 +05:30
Utils.h	[LINALG] Lower `aten.Matmul` to `linalg.BatchMatmul`	2022-06-25 10:58:06 +05:30