[OpenVINOQuantizer] Minor improvements

[daniil-lyakhov]

Upon a check that the OpenVINOQuantizer is working with the torchao, following minor improvements are introduced

• nncf.torch.disable_patching() could be removed thanks to the Remove disable patching for fx examples and docs openvinotoolkit/nncf#3435
• Comma is removed from the pip install openvino nncf line
• Copy import is removed
• Additional export is added before the nncf quantization as the exported model contains FQs after the first quantization

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/2581

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit 55ab97a with merge base 74808e2 ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.