Some minor type stub improvements (#118529)

Summary:
I was just playing around with improving the typing of symbolic_shapes. The PR is not "complete" but I in particular wanted to get feedback on whether or not people liked making ValueRanges Generic; it seems that distinguishing if you have an Expr ValueRange or a SympyBoolean ValueRange is a lot of trouble for downstream. Using TypeGuard, we can perform refinements on the generic parameter inside methods, although we still have to cast back to ValueRange[T] due to python/mypy#14425 (comment)

Signed-off-by: Edward Z. Yang <[email protected]>

X-link: pytorch/pytorch#118529
Approved by: https://github.com/Skylion007

Reviewed By: clee2000

Differential Revision: D53296779

Pulled By: ezyang

fbshipit-source-id: 95799914350e50aeedf1acad93d71b79cad827c8

Author: ezyang

userbenchmark/dynamo/dynamobench/_dynamo/utils.py

@@ -2066,18 +2066,18 @@ def defake(x):
     size: "torch._prims_common.ShapeType"
     stride: "torch._prims_common.StrideType"
     if x._has_symbolic_sizes_strides:
-        size = [
-            s.node.shape_env.size_hint(s.node.expr)
-            if isinstance(s, torch.SymInt)
-            else s
-            for s in x.size()
-        ]
-        stride = [
-            s.node.shape_env.size_hint(s.node.expr)
-            if isinstance(s, torch.SymInt)
-            else s
-            for s in x.stride()
-        ]
+        size = []
+        for s in x.size():
+            if isinstance(s, torch.SymInt):
+                size.append(s.node.shape_env.size_hint(s.node.expr))
+            else:
+                size.append(s)
+        stride = []
+        for s in x.stride():
+            if isinstance(s, torch.SymInt):
+                stride.append(s.node.shape_env.size_hint(s.node.expr))
+            else:
+                stride.append(s)
     else:
         size = x.size()
         stride = x.stride()