fix: correct batch handling in _prepare_image_ids for cond_model_input

[tanishaj12]

What does this PR do?

Fixes huggingface#13811

Description

In train_dreambooth_lora_flux2_img2img.py, _prepare_image_ids was being called with a list of B single-image tensors (one per batch element), causing it to assign escalating temporal IDs (T=10, T=20, T=30...) across batch elements — as if they were multiple reference images for a single sample.

Since each batch element is an independent training sample with exactly one conditional image, all samples should receive the same temporal ID (T=10). The fix computes IDs for a single sample and expands them across the batch dimension.

Before (buggy):

cond_model_input_list = [cond_model_input[i].unsqueeze(0) for i in range(cond_model_input.shape[0])]
cond_model_input_ids = Flux2Pipeline._prepare_image_ids(cond_model_input_list).to(
    device=cond_model_input.device
)
cond_model_input_ids = cond_model_input_ids.view(
    cond_model_input.shape[0], -1, model_input_ids.shape[-1]
)

After (fixed):

cond_model_input_ids = Flux2Pipeline._prepare_image_ids(
    [cond_model_input[0:1]]
).to(device=cond_model_input.device)
cond_model_input_ids = cond_model_input_ids.expand(
    cond_model_input.shape[0], -1, -1
)

Before submitting

• This PR fixes a typo or improves the docs.
• Did you read the [contributor guideline](https://github.com/huggingface/diffusers/blob/main/CONTRIBUTING.md)?
• Did you read our [philosophy doc](https://github.com/huggingface/diffusers/blob/main/PHILOSOPHY.md) (important for complex PRs)?
• Was this discussed/approved via a GitHub issue or the [forum](https://discuss.huggingface.co/c/discussion-related-to-httpsgithubcomhuggingfacediffusers/63)? Please add a link to it if that's the case. → Issue Fix incorrect batch handling in _prepare_image_ids usage in train_dreambooth_lora_flux2_img2img.py huggingface/diffusers#13811
• Did you make sure to update the documentation with your changes?
• Did you write any new necessary tests?