loading generation config if it is part of model

[eaidova]

What does this PR do?

there is misalignment with behaviour how we convert model and how to load in generation config part. In some cases, generation config parameters can be different from default and not provided in config. I see issue with attempt to run whisper-tiny with pipelines due to difference in generation config, when openvino model loaded and when original pytorch model (due to obtaining generation config from model config instead of using provided with model before conversion)

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?

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[eaidova]

@echarlaix could you please take a look?

[echarlaix]

Looks good thanks for the fix @eaidova