[Quant] Remove warnings from using torch.tensor(value) (#84277)
Summary:
I think zafar made an earlier pull for these changes [here](https://github.com/pytorch/pytorch/commit/ce0786add26c1e117b16b58e8ae12dbe776133e1), but they didn't seem to make it through the migration.
Test Plan:
```
python test/test_quantization.py
```
Reviewers:
Subscribers:
Tasks: https://github.com/pytorch/pytorch/issues/73566
Tags: quant
Differential Revision: [D39145070](https://our.internmc.facebook.com/intern/diff/D39145070)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/84277
Approved by: https://github.com/z-a-f
