flux2-klein lora train，bug

[lspze]

Describe the bug

path, 1662-1672: https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/train_dreambooth_lora_flux2_klein_img2img.py

code:
cond_model_input_list = [cond_model_input[i].unsqueeze(0) for i in range(cond_model_input.shape[0])]
cond_model_input_ids = Flux2KleinPipeline._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]
)

在这里，为什么要把同一个bacth里面的所有图像当成一组参考图来计算id，这样应该不对吧？应该只计算其中一张图片的id值，然后使用repeat来进行复制。这里一个batch的参考图之间应该是独立的。

Reproduction

thanks

Logs

System Info

• 🤗 Diffusers version: 0.38.0
• Platform: Linux-5.4.0-215-generic-x86_64-with-glibc2.31
• Running on Google Colab?: No
• Python version: 3.10.0
• PyTorch version (GPU?): 2.9.1+cu128 (True)
• Flax version (CPU?/GPU?/TPU?): not installed (NA)
• Jax version: not installed
• JaxLib version: not installed
• Huggingface_hub version: 1.17.0
• Transformers version: 5.9.0
• Accelerate version: 1.12.0
• PEFT version: 0.19.1
• Bitsandbytes version: not installed
• Safetensors version: 0.8.0-rc.1
• xFormers version: 0.0.33.post2
• Accelerator: NVIDIA A100 80GB PCIe, 81920 MiB
  NVIDIA A100 80GB PCIe, 81920 MiB
  NVIDIA A100 80GB PCIe, 81920 MiB
  NVIDIA A100 80GB PCIe, 81920 MiB
  NVIDIA A100 80GB PCIe, 81920 MiB
  NVIDIA A100 80GB PCIe, 81920 MiB
  NVIDIA A100 80GB PCIe, 81920 MiB
  NVIDIA A100 80GB PCIe, 81920 MiB
• Using GPU in script?:
• Using distributed or parallel set-up in script?:

Who can help?

No response

[Zhu1116]

确实，我也发现这个问题，但是PR还没处理