Distillation optimizer fix

[vlad-karp]

Description

With the introduction of ModelBundle in the distill pipeline, it allocates the optimizer state for both models in the base Tunix peft trainer.

Since the optimizer logic is completely revised in the MaxTextDistillationTrainer anyway, the fix is simply to send a dummy optimizer to the base class.

Tests

Not appllicable

Checklist

Before submitting this PR, please make sure (put X in square brackets):

• I have performed a self-review of my code. For an optional AI review, add the gemini-review label.
• I have necessary comments in my code, particularly in hard-to-understand areas.
• I have run end-to-end tests tests and provided workload links above if applicable.
• I have made or will make corresponding changes to the doc if needed, including adding new documentation pages to the relevant Table of Contents (toctree directive) as explained in our documentation.

[codecov]

Codecov Report

❌ Patch coverage is 0% with 2 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
.../trainers/post_train/distillation/train_distill.py	0.00%	2 Missing ⚠️

📢 Thoughts on this report? Let us know!