xplat/executorch/backends/xnnpack/recipes/xnnpack_recipe_types.py (#19083)

DevmateXplatTypoFixes Bot · facebook-github-bot · commit fd2b76850a80 · 2026-04-23T16:48:33.000-07:00
Summary: Pull Request resolved: #19083 Reviewed By: franklinho Differential Revision: D100956511
diff --git a/backends/xnnpack/recipes/xnnpack_recipe_types.py b/backends/xnnpack/recipes/xnnpack_recipe_types.py
@@ -27,7 +27,7 @@ class XNNPackRecipeType(RecipeType):
         "xnnpack_torchao_int8da_int4w_per_channel"
     )
     # INT8 Dynamic Activations INT4 Weight Quantization, default group_size = 32
-    # can be overriden by group_size kwarg
+    # can be overridden by group_size kwarg
     TORCHAO_INT8_DYNAMIC_ACT_INT4_WEIGHT_PER_TENSOR = (
         "xnnpack_torchao_int8da_int4w_per_tensor"
     )

Original file line number	Diff line number	Diff line change
`@@ -27,7 +27,7 @@ class XNNPackRecipeType(RecipeType):`
`27`	`27`	`"xnnpack_torchao_int8da_int4w_per_channel"`
`28`	`28`	`)`
`29`	`29`	`# INT8 Dynamic Activations INT4 Weight Quantization, default group_size = 32`
`30`		`- # can be overriden by group_size kwarg`
	`30`	`+ # can be overridden by group_size kwarg`
`31`	`31`	`TORCHAO_INT8_DYNAMIC_ACT_INT4_WEIGHT_PER_TENSOR = (`
`32`	`32`	`"xnnpack_torchao_int8da_int4w_per_tensor"`
`33`	`33`	`)`