Update recipes_source/torch_compile_user_defined_triton_kernel_tutorial.py

zou3519 · svekars · web-flow · commit 72d60075b0ef · 2025-01-13T14:34:49.000-05:00
Co-authored-by: Svetlana Karslioglu &lt;svekars@meta.com&gt;
diff --git a/recipes_source/torch_compile_user_defined_triton_kernel_tutorial.py b/recipes_source/torch_compile_user_defined_triton_kernel_tutorial.py
@@ -155,7 +155,7 @@ def add_fn(x, y):
 # triton_op is a structured way of defining a custom operator that is backed by one
 # or more Triton kernels: like regular custom operators (``torch.library.custom_op``),
 # you are able to specify the interactions with PyTorch subsystems via ``torch.library``.
-# However, unlike ``torch.library.custom_op``, which creates opaque callables w.r.t.
+# However, unlike ``torch.library.custom_op``, which creates opaque callables with respect to
 # ``torch.compile``, ``torch.compile`` traces into ``triton_op`` to apply optimizations.
 #
 # Here’s a chart of which API to use when integrating triton kernels with PyTorch.

Original file line number	Diff line number	Diff line change
`@@ -155,7 +155,7 @@ def add_fn(x, y):`
`155`	`155`	`# triton_op is a structured way of defining a custom operator that is backed by one`
`156`	`156`	# or more Triton kernels: like regular custom operators (``torch.library.custom_op``),
`157`	`157`	# you are able to specify the interactions with PyTorch subsystems via ``torch.library``.
`158`		-# However, unlike ``torch.library.custom_op``, which creates opaque callables w.r.t.
	`158`	+# However, unlike ``torch.library.custom_op``, which creates opaque callables with respect to
`159`	`159`	# ``torch.compile``, ``torch.compile`` traces into ``triton_op`` to apply optimizations.
`160`	`160`	`#`
`161`	`161`	`# Here’s a chart of which API to use when integrating triton kernels with PyTorch.`