Feat (inference_mode): pickle compatibility

[Giuseppe5]

Reason for this PR

Injector are not compatible with pickle when trying to do 'torch.save'.

In theory, once dependency injection is resolved, we don't need injectors anymore. We rely on the injector too much so this might break at times.

In quant_inference_mode, we detach the injector and save it so that we can re-attach it at the exit of the context manager.

Example usage:

import brevitas.nn as qnn

model =  qnn.QuantLinear(3, 8)
with quant_inference_mode(model , delete_injector=True):
    b = model(torch.randn(1,3))
    torch.save(model , 'test.pickle')

Changes Made in this PR

Remove/Reattach quant injector when needed.

Testing Summary

None, Draft PR.

Risk Highlight

• This PR includes code from another work (please detail).
• This PR contains API-breaking changes.
• This PR depends on work in another PR (please provide links/details).
• This PR introduces new dependencies (please detail).
• There are coverage gaps not covered by tests.
• Documentation updates required in subsequent PR.

Potential edge cases not tested

Checklist

• Code comments added to any hard-to-understand areas, if applicable.
• Changes generate no new warnings.
• Updated any relevant tests, if applicable.
• No conflicts with destination dev branch.
• I reviewed my own code changes.
• Initial CI/CD passing.
• 1+ reviews given, and any review issues addressed and approved.
• Post-review full CI/CD passing.