https://github.com/kohya-ss/sd-scripts/blob/main/docs/config_README-ja.mdhttps://github.com/kohya-ss/sd-scripts/blob/main/docs/config_README-ja.md[Stable Diffusion]训练你的LoRA(Linux) - 知乎简介LoRA 是一种参数高效微调方法（PEFT），最早由 LoRA: Low-Rank Adaptation of Large Language Models 提出并应用于微调语言大模型之中，后来由 Low-rank Adaptation for Fast Text-to-Image Diffusion Fine-tu…https://zhuanlan.zhihu.com/p/640144661配置文件格式TOML：

[general]
shuffle_caption = true
caption_extension = '.txt'
keep_tokens = 1

# This is a DreamBooth-style dataset
[[datasets]]
resolution = 512
batch_size = 4
keep_tokens = 2

  [[datasets.subsets]]
  image_dir = 'C:\hoge'
  class_tokens = 'hoge girl'
  # This subset has keep_tokens = 2 (using the value of the parent datasets)

  [[datasets.subsets]]
  image_dir = 'C:\fuga'
  class_tokens = 'fuga boy'
  keep_tokens = 3

  [[datasets.subsets]]
  is_reg = true
  image_dir = 'C:\reg'
  class_tokens = 'human'
  keep_tokens = 1

# This is a fine-tuning-style dataset
[[datasets]]
resolution = [768, 768]
batch_size = 2

  [[datasets.subsets]]
  image_dir = 'C:\piyo'
  metadata_file = 'C:\piyo\piyo_md.json'
  # This subset has keep_tokens = 1 (using the general value)

在此示例中，将训练三个目录作为512x512（批量大小4）的dreambooth数据集，以及一个目录作为768x768（批量大小2）的微调数据集。

C:\
├─ hoge  ->  [[datasets.subsets]] No.1  ┐                        ┐
├─ fuga  ->  [[datasets.subsets]] No.2  |->  [[datasets]] No.1   |->  [general]
├─ reg   ->  [[datasets.subsets]] No.3  ┘                        |
└─ piyo  ->  [[datasets.subsets]] No.4  -->  [[datasets]] No.2   ┘

 所有方法均可使用的参数：[general]

dreambooth-style 特有的参数：

fine-tuning-style特有的参数：