
1. Introduction
This is a LoRA for illustrations in style of Takahashi Shoutei (1871-1945). The source images are collected on the web, and I would like to serve a LoRA made with 88 public domain images.
This LoRA was trained on my computer (Ryzen 5 2600 + 16 GB RAM + 500 GB SSD + NVIDIA GeForce RTX 2060 6 GB) with SD-Trainer by Akegarasu (https://github.com/Akegarasu/lora-scripts), on sd-scripts by kohya-ss (https://github.com/kohya-ss/sd-scripts/tree/main), and well known configurations of training as follows (I don't change most default parameters):
-
DreamBooth method without regularization images
-
pretrained model: ULTRA v3 (for SD 1.5) https://civitai.com/models/228525?modelVersionId=470417
-
resolution = 512 x 512
-
epochs = 30
-
optimizer type = "Lion"
-
steps = 3,960
2. Usage
To be a simple way in your prompt, please add LoRA decralation with some weight value just as "<lora:BT_Takahashi_Shoutei_v1:1>" and put the trigger keyword "takahashi shoutei" in some place of your prompt. In most cases, a keyword of style will be better to write a later position in your prompt.
A weight value can vary a result, you know. When you set "0," no effect will be shown, and when you set "1" or more value, the result may be worse by the well-known "overfitting," and most images are likely resembled with the original images used in my training. So, you will have to set the value in your own situtation (what other LoRA you use, and where do you put a keyword in your prompt, and son on.) If you set a weight over 1.0, the overfitting may be happen. But most results of larger images are well painted by a weight of 1.0 as the following image.
A value for CLIP SKIP will be available not only 1 but also 2.
Anyway, here's a sample image.
Prompt: (no humans, very wide shot, scenary, perspective, outdoors landscape, legendary, Skyborne archipelago with cloud forests and floating fauna, X-shape, even, edge light, dappled in Ajanta Caves in Maharashtra, wallpaper, uncensored, unparalleled masterpiece, finely detail, incredibly absurdres, perfect artwork, highres, concept art, elegant, extremely detailed, intricate, professionally color graded, UHD, HDR, 32K, 16K, professional illustration, vibrant details, abstract beauty, (takahashi shoutei) <lora:BT_Takahashi_Shoutei_v1:0.5>
Negative prompt: multiple angles, multiple views, collage, multiple images, frame, border, split frame, split screen, humans, girl, boy, easynegative, badquality, verybadimagenegative_v1.3
Steps: 30, Sampler: DPM++ 3M SDE, Schedule type: Polyexponential, CFG scale: 7.5, Seed: 1433552444, Size: 1024x576, Model hash: 2289c564ac, Model: ultra_v3, VAE hash: 63aeecb90f, VAE: ClearVAE_V2.3.safetensors, Lora hashes: "BT_Takahashi_Shoutei_v1: b525f6845b09", TI hashes: "easynegative: c74b4e810b03, verybadimagenegative_v1.3: d70463f87042", Version: v1.10.1
3. Weight
[Notice] It is hard to add images on the lower part of web page on this editor application, so unfortunately I will not use explanatory images below.
Results would be varied by a weight value in LoRA declaration, depending on other LoRA you use, the position of trigger word ("takahashi shoutei") in your prompt, base model, or others. Generally speaking, a weight under 0.5 will be ineffective and a weight over 1.0 can make an overfitting. And even with a weight as 0.5, some images may not have a taste of Takahashi Shoutei, so I recommend you to set a weight over 0.6.
4. the Last word
Have fun, and thanks to users, creators of scripts, applications, models, and at last but not least, to crews at Civitai.
描述:
The initial release.
训练词语: takahashi shoutei
名称: BT_Takahashi_Shoutei_v1.safetensors
大小 (KB): 32352
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success