Fix a typo in the memory planning doc (#5723)

Olivia-liu · facebook-github-bot · commit c1c50803103e · 2024-09-27T13:43:12.000-07:00
Summary: Pull Request resolved: #5723 Found this typo in the doc. Reviewed By: dvorjackz Differential Revision: D63500296 fbshipit-source-id: 23ba7f287324bcaaa200a8ec55d28aac919f43b8
diff --git a/docs/source/compiler-memory-planning.md b/docs/source/compiler-memory-planning.md
@@ -33,7 +33,7 @@ program = edge_program.to_executorch(
             exir.ExecutorchBackendConfig(
                 memory_planning_pass=MemoryPlanningPass(
                     alloc_graph_input=False, # Inputs will not be memory planned, the data_ptr for input tensors after model load will be nullptr
-                    alloc_graph_output=True, # Outputs will be memory planned, the data_ptr for input tensors after model load will be in the `planned_memory`.
+                    alloc_graph_output=True, # Outputs will be memory planned, the data_ptr for output tensors after model load will be in the `planned_memory`.
                 )
             )
         )

Original file line number	Diff line number	Diff line change
`@@ -33,7 +33,7 @@ program = edge_program.to_executorch(`
`33`	`33`	`exir.ExecutorchBackendConfig(`
`34`	`34`	`memory_planning_pass=MemoryPlanningPass(`
`35`	`35`	`alloc_graph_input=False, # Inputs will not be memory planned, the data_ptr for input tensors after model load will be nullptr`
`36`		- alloc_graph_output=True, # Outputs will be memory planned, the data_ptr for input tensors after model load will be in the `planned_memory`.
	`36`	+ alloc_graph_output=True, # Outputs will be memory planned, the data_ptr for output tensors after model load will be in the `planned_memory`.
`37`	`37`	`)`
`38`	`38`	`)`
`39`	`39`	`)`