clarify default value of requires_grad for tensors (#61038)

Michael Dagitses · facebook-github-bot · commit 58df01c3b808 · 2021-07-12T12:57:37.000-07:00
Summary: Pull Request resolved: #61038 Test Plan: Imported from OSS Reviewed By: albanD Differential Revision: D29491984 Pulled By: dagitses fbshipit-source-id: 7e6b7f8e81d77f38c881b86a68c17d3cf5483dad
diff --git a/docs/source/notes/autograd.rst b/docs/source/notes/autograd.rst
@@ -56,9 +56,10 @@ to disable gradient computation but, because of its name, is often mixed up with
 Setting ``requires_grad``
 ^^^^^^^^^^^^^^^^^^^^^^^^^
 
-:attr:`requires_grad` is a flag that allows for fine-grained exclusion of
-subgraphs from gradient computation. It takes effect in both the forward
-and backward passes:
+:attr:`requires_grad` is a flag, defaulting to false *unless wrapped
+in a ``nn.Parameter``*, that allows for fine-grained exclusion of
+subgraphs from gradient computation. It takes effect in both the
+forward and backward passes:
 
 During the forward pass, an operation is only recorded in the backward graph if
 at least one of its input tensors require grad.
