[Mlir-commits] [mlir] [mlir][transform] Fix transform dialect tutorial chapter 1 (PR #147983)
Renato Golin
llvmlistbot at llvm.org
Thu Jul 10 08:26:22 PDT 2025
================
@@ -11,6 +11,9 @@ The application of transform IR always starts from one top-level operation. In t
Let us illustrate this with a simple sequence of transformations on the common “fully connected + bias + ReLU” ML layer, which boils down to performing a matrix multiplication, followed by an (elementwise) matrix addition and taking an elementwise maximum with 0. This can be expressed using the following IR:
```mlir
+#map0 = affine_map<(d0, d1) -> (d0, d1)>
+#map1 = affine_map<(do, d1) -> ()>
----------------
rengolin wrote:
```suggestion
#map1 = affine_map<(d0, d1) -> ()>
```
https://github.com/llvm/llvm-project/pull/147983
More information about the Mlir-commits
mailing list