Enable LoRA loading on Ideogram4ModularPipeline

[SamuelTallet]

What does this PR do?

This tiny PR fixes #13979.

I applied same mechanism (LoRA mixin) used in other modular pipelines (Z-Image, ERNIE-Image, etc).

Before submitting

• Did you use an AI agent (Claude Code, Codex, Cursor, etc.) to help with this PR? If so:
  • Did you point it at the project conventions in .ai/ (e.g. via make claude / make codex)? See Coding with AI agents.
  • Did you self-review the diff against .ai/review-rules.md?
• Did you read the contributor guideline?
• Did you read our philosophy doc? (important for complex PRs)
• Was this discussed/approved via a GitHub issue or the forum? Please add a link to it if that's the case.
  Yes, see Can't use load_lora_weights() on Ideogram4ModularPipeline #13979
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

Who can review?

@yiyixuxu @sayakpaul

[yiyixuxu]

thanks!

[SamuelTallet]

@yiyixuxu You're welcome.

[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.