eager quant: fix error with removing forward hooks (pytorch#49813)

Summary:
Pull Request resolved: pytorch#49813

pytorch#49739 reports a crash
where removing forward hooks results in a

```
RuntimeError: OrderedDict mutated during iteration
```

Unfortunately I cannot repro this inside the PyTorch module, but the issue
author has a good point and and we should not mutate the dict inside
of the iteration.

Test Plan:
```
// test plan from pytorch#46871 which
// originally added this
python test/test_quantization.py TestEagerModeQATOps
```

Imported from OSS

Reviewed By: jerryzh168

Differential Revision: D25698725

fbshipit-source-id: 13069d0d5017a84038c8f7be439a3ed537938ac6

torch/quantization/quantize.py

@@ -258,9 +258,12 @@ def _remove_activation_post_process(module):
         delattr(module, 'activation_post_process')
 
     # remove activation_post_proceess hook
+    handle_ids_to_remove = set()
     for handle_id, hook_fn in module._forward_hooks.items():
         if hook_fn is _observer_forward_hook:
-            module._forward_hooks.pop(handle_id)
+            handle_ids_to_remove.add(handle_id)
+    for handle_id in handle_ids_to_remove:
+        module._forward_hooks.pop(handle_id)
 
 # TODO: rename to something more general
 def _remove_qconfig(module):