fix(qwen-image dreambooth): correct prompt embed repeats when using --with_prior_preservation

[chenyangzhu1]

What does this PR do?

I found that the same problem in #13292 also appears in Qwen-Image's dreambooth lora script.

diffusers/examples/dreambooth/train_dreambooth_lora_qwen_image.py

Line 1468 in 8070f6e

num_repeat_elements = len(prompts)

The root cause and fixing are the same as #13307 and #13292.

Root cause: collate_fn appends class prompts to the instance prompts list (doubling len(prompts)), but prompt_embeds is already doubled earlier via torch.cat([instance_embeds, class_embeds]). Using the full len(prompts) as the repeat count produces 4 embeddings for 2 latents at batch_size=1.

Fix: Use len(prompts) // 2 when args.with_prior_preservation is active, so the repeat count matches the number of unique prompt groups rather than the doubled collated list.

Applied to the Qwen-Image related script:

• examples/dreambooth/train_dreambooth_lora_qwen_image.py

Fixes #13292

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• 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.
• 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?

@sayakpaul

Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

[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.

[azolotenkov]

@sayakpaul I think the same prior-preservation repeat bug also exists in the Flux2 scripts, at least in:

• train_dreambooth_lora_flux2.py
• train_dreambooth_lora_flux2_klein.py

I reproduced this in the Flux2 Klein script with --with_prior_preservation, train_batch_size=1, no custom captions, no latent cache, and got:
RuntimeError: The size of tensor a (2) must match the size of tensor b (4) at non-singleton dimension 0

The same fix pattern from this PR seems applicable there too:
num_repeat_elements = len(prompts) // 2 if args.with_prior_preservation else len(prompts)

If preferred, I can open a separate small PR for the Flux2 + Flux2 Klein scripts.

[sayakpaul]

Sure

[azolotenkov]

Done