CCCP poster style

关于NOOB V-PRED 2.0

基于NOOB V-PRED V1.0使用OneTrainer进行训练。

筛选训练集并增加训练次数，以获得更优的色彩与效果表现。

已移除style-abstract触发词，因为NOOB对这类艺术风格的理解较差。

具体训练细节请参阅右侧版本相关信息。

About NOOB V-PRED 2.0

Train using OneTrainer based on NOOB V-PRED V1.0.

Filter the training set and increase the number of training repeats to achieve better colors and effects.

The triggerword s-o (style-abstract) has been removed because NOOB has a poor understanding of this type of art style.

Please refer to the relevant version on the right for specific training details.

关于NOOB V-PRED V1.0

基于NOOB V-PRED V1.0进行训练。

减少训练数据以获得更佳效果，目前在单色风格（monochrome）表现更优。

细化触发词以生成不同艺术风格：s-a 表示 style-abstract，s-o 表示 style-oil painting，s-m 表示 style-monochrome。

具体训练细节请参阅右侧版本相关信息。

About NOOB V-PRED V1.0

Trained based on NOOB V-PRED V1.0.

I reduced the training data for better effect. Now it performs better in monochrome.

I divided the trigger words to generate different types of art styles, where s-a means style-abstract, s-o means style-oil painting, and s-m means style-monochrome.

It is recommended to use NOOB V-PRED 0.5 as the generative model for better aesthetic performance.

Please refer to the about version on the right for specific training details.

关于NOOB E-PRED V1.0

基于NOOB E-PRED V1.0进行训练。

重新收集训练数据，并细化触发词以生成不同艺术风格：s-a 表示 style-abstract，s-o 表示 style-oil painting，s-m 表示 style-monochrome。

建议使用NOOB V-PRED 0.5作为生成模型，以获得更优的美学表现。

具体训练细节请参阅右侧版本相关信息。

About NOOB E-PRED V1.0

Trained based on NOOB E-PRED V1.0.

I recollected the training data and divide the trigger words to generate different types of art styles, where s-a means style-abstract, s-o means style-oil painting, and s-m means style-monochrome.

It is recommended to use NOOB V-PRED 0.5 as the generative model for better aesthetic performance.

Please refer to the about version on the right for specific training details.

关于PonyXL-experimental

尝试以PonyXL作为底模进行训练，结果勉强接近3.0版本，但图像细节缺乏足够质感，目前仍在探索更优版本。训练参数详见版本相关。

About the PonyXL experimental

Attempting to train with PonyXL as the base model resulted in a result that was barely close to 3.0, but lacked a lot of texture in image details. We are still exploring a better version. See about the version for training parameters.

关于3.0

在3.0版本中，使用了多分辨率噪声技术（金字塔噪声），未启用正则化，所有图像均参与训练。

光影效果明显增强，对规则物体的描绘也更为敏感。其缺点是似乎更偏好绘制建筑，因此在绘制人物时建议提高1girl或1male的权重。

训练参数如下：

multires_noise_iterations="6"

multires_noise_discount=0.3

Introduction

这是一个用于绘制苏联海报风格的LyCORIS。权重较低时偏向模型自身风格，权重较高时则更贴近海报风格。

目前仅在二次元模型上测试，主要测试模型为AOM3A和viewerMix，效果较好。若用于偏真实的模型，可能更能还原原版海报风格。

Training parameters

基于NAI Final作为底模进行训练，使用36张图片作为训练集，35张图片作为回归训练集，训练前进行了镜像增强；每张图片重复训练10次，共训练10个epoch，最终步数约为7000步，主要设置参数如下：

network_dim=32

network_alpha=32

keep_tokens=4

conv_dim=4

conv_alpha=4

lr="1e-4"

unet_lr="1e-4"

text_encoder_lr="1e-5"

batch_size = 2