Support activation quantization without scaling (#2607)

Summary:

X-link: pytorch/pytorch#148380

We enable the activation quantization in the forward pass, and users can customize the dtype they want to quantize.

Differential Revision: D70522237

Author: mengluy0125

userbenchmark/dynamo/dynamobench/_dynamo/utils.py

@@ -4589,3 +4589,7 @@ def maybe_disable_inference_mode_for_fake_prop() -> Generator[None, None, None]:
             yield
     else:
         yield
+
+
+def is_node_meta_valid(node: Optional[torch.fx.Node]) -> bool:
+    return node is None or "example_value" in node.meta or "val" in node.meta