add bucket for sdxl lora and dreambooth script traing

[Light-yzc]

PR Title: feat: Add aspect ratio bucketing to SDXL LoRA training scripts

What does this PR do?

This PR adds support for aspect ratio bucketing to the train_text_to_image_lora_sdxl.py and train_dreambooth_lora_sdxl.py scripts.

This allows training on images with different resolutions and aspect ratios. It avoids forcing all images to be cropped into a square, which helps preserve the original composition.

Changes:

• Added a BucketBatchSampler that creates batches of images with similar resolutions.
• Modified the data preprocessing logic to assign each image to a bucket and resize/crop it accordingly.
• Updated compute_time_ids to handle the dynamic target sizes from the buckets.
• For the DreamBooth script, the bucketing logic now also applies to class images to ensure instance/prior image pairs have matching dimensions.

New command-line arguments like --enable_bucketing and --bucket_size are added to control this feature.

Fixes # (issue)

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?