Fix/paraphrase transformers compat

[aspire488]

Tell us what this change does. If you're fixing a bug, please mention
the github issue number.

Please ensure you are submitting from a unique branch in your repository to main upstream.

Verification

List the steps needed to make sure this thing works

• Supporting configuration such as generator configuration file

{
    "huggingface": {
        "torch_type": "float32"
    }
}

• garak -t <target_type> -n <model_name>
• Run the tests and ensure they pass python -m pytest tests/
• ...
• Verify the thing does what it should
• Verify the thing does not do what it should not
• Document the thing and how it works (Example)

If you are opening a PR for a new plugin that targets a specific piece of hardware or requires a complex or hard-to-find testing environment, we recommend that you send us as much detail as possible.

Specific Hardware Examples:

• GPU related
  • Specific support required cuda / mps ( Please not cuda via ROCm if related )
  • Minium GPU Memory

Complex Software Examples:

• Expensive proprietary software
• Software with an extensive installation process
• Software without an English language UI

• Summary

Fixes #1499.

The buffs.paraphrase.Fast buff currently relies on the external transformers-community/group-beam-search custom generation implementation and trust_remote_code. Newer Transformers versions provide the required group beam search functionality natively, making the custom generation dependency unnecessary and causing compatibility issues.

Changes

• Removed the custom_generate dependency from Fast.
• Removed the trust_remote_code requirement and argument.
• Kept the existing native group beam search parameters:
  • num_beams
  • num_beam_groups
  • num_return_sequences
  • diversity_penalty
• Added a regression test verifying that the deprecated custom-generation path is no longer configured.

Duplicate-work check

I checked the open garak PRs around paraphrase/custom generation and did not find an open PR addressing this specific compatibility fix.

Verification

Focused test:

python -m pytest tests/buffs/test_buffs.py -k "fast or native_group_beam_search"