Fix: AttributeError: module 'packaging' has no attribute 'version'

[soulteary]

What does this PR do?

An error occurred while using optimum run Whisper-v3.

Someone in the community has given feedback:

• https://discuss.huggingface.co/t/problem-with-whisper-large-v3/61453

problem also in langchain:

• Solving anthropic packaging version issue langchain-ai/langchain#7306

Fixes #1659

Although the error is reported because there is no need to call Optimum, in Utils, due to the problem of how to import the function, an error will occur when running the code in the Python 3.10 environment.

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you make sure to update the documentation with your changes?
• Did you write any new necessary tests?

[fxmarty]

Thanks a lot for the fix - wonder why this was not catched in the CI

[fxmarty]

@soulteary Could you fix the error there?

    @contextmanager
    def require_numpy_strictly_lower(version: str, message: str):
>       if not version.parse(np.__version__) < version.parse(version):
E       AttributeError: 'str' object has no attribute 'parse'

../optimum/utils/import_utils.py:198: AttributeError
============================== slowest durations ===============================
0.88s call     tests/exporters/onnx/test_onnx_config_loss.py::TestOnnxConfigWithLoss::test_onnx_config_with_loss

(92 durations < 0.005s hidden.  Use -vv to show these durations.)
=========================== short test summary info ============================
FAILED exporters/onnx/test_onnx_config_loss.py::TestOnnxConfigWithLoss::test_onnx_config_with_loss - AttributeError: 'str' object has no attribute 'parse'
======================== 1 failed, 30 skipped in 6.78s =========================

[fxmarty]

LGTM, failing tests are unrelated. Thank you for the fix!