Add train flux2 series lora config

[tcaimm]

What does this PR do?

Expand LoRA support for FLUX.2 series single stream blocks and update docs

1. Architectural Evolution

Compared to the original FLUX framework, the FLUX.2 architecture has undergone significant changes. Firstly, the number of single-stream layers is far greater than that of double-stream layers. Furthermore, In the Single transformer Blocks, the q,k,v projections are fused with the MLP into a single unified linear layer: attn.to_qkv_mlp_proj.

Therefore, using Flux's Lora configuration to train Flux2 is insufficient.

2. Implementation Updates

To address these changes, I have updated the LoRA configuration in the following training scripts and added additional notes to the readme:

• examples/dreambooth/README_flux2.md
• examples/dreambooth/train_dreambooth_lora_flux2.py
• examples/dreambooth/train_dreambooth_lora_flux2_klein.py

The target_modules logic has been modified to ensure that the Lora adapter can correctly train the main attention layers of both double-stream and single-stream layers.Flux2 has 48 single-stream layers, while Klein has 24.

# Updated LoRA targeting logic for FLUX.2 series
target_modules = [
    "to_k", "to_q", "to_v", "to_out.0"
] + [
    "to_qkv_mlp_proj", 
    *[f"single_transformer_blocks.{i}.attn.to_out" for i in range(nums)]
]

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?

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.

[tcaimm]

@sayakpaul Please take a look at this PR. Thank you for your help!

[tcaimm]

@linoytsaban Please take a look at this PR. Thank you for your help!

[linoytsaban]

@tcaimm thanks for this!

[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.

[linoytsaban]

LGTM!

[tcaimm]

@linoytsaban I've added the requested comments and included LoRA configuration support in the img2img script. Please take a look at this PR. Thank you for your help!