[torchbind] fix error message when attr is a real tensor. (pytorch#151944)

Summary: Previously, when attr is defined, "if attr" will try to evaluate the data of attr, which is not intendended and we get a ugly error stack if the attr is not evaluable (like a fake tensor) before the callable(attr) check. Test Plan: Existing tests. Reviewed By: yushangdi, henryoier Differential Revision: D73460905 Pull Request resolved: pytorch#151944 Approved by: https://github.com/yushangdi

Author: ydwu4

torch/_library/fake_class_registry.py

@@ -155,7 +155,7 @@ def maybe_to_fake_obj(
 
  for name in x._method_names(): # type: ignore[attr-defined]
  attr = getattr(fake_x, name, None)
- if attr:
+ if attr is not None:
  if not callable(attr):
  raise RuntimeError(f"Expect {name} to be a callable but got {attr}.")