Doc update for latest features

[kohya-ss]

No description provided.

[Copilot]

Pull Request Overview

This pull request organizes and expands the documentation structure by adding comprehensive guides for various training features. The documentation has been reorganized from the main README into dedicated markdown files covering specific training topics and features.

• Creates dedicated documentation files for different training aspects (validation, fine-tuning, textual inversion, etc.)
• Adds detailed guides for specific model architectures (FLUX.1, SD3, SDXL, Lumina)
• Removes redundant content from the main README and provides organized reference structure

Reviewed Changes

Copilot reviewed 10 out of 10 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
docs/validation.md	New comprehensive guide for validation loss functionality with English and Japanese sections
docs/train_textual_inversion.md	New detailed guide for Textual Inversion training with examples and troubleshooting
docs/train_network_advanced.md	Removes status markers and adds new advanced training options
docs/sd3_train_network.md	Removes status markers and adds extensive LoRA target module documentation
docs/lumina_train_network.md	Adds new training options for Sage Attention and sample batch size
docs/gen_img_README.md	Expands image generation documentation with SDXL options and new features
docs/gen_img_README-ja.md	Japanese version updates matching the English documentation changes
docs/flux_train_network.md	Adds additional training options and loss function configuration
docs/fine_tune.md	New comprehensive guide comparing fine-tuning vs LoRA training
README.md	Restructures documentation by moving content to dedicated files and updating recent updates

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[kohya-ss]

Some documents still need to be fixed, but we will merge this before HunyuanImage-2.1.