Anime Illust Diffusion

AnimeIllustDiffusionV2.0 在 1.0 的基础上额外训练了400期。我大大改善了训练集图像。现在，整个训练集一共有9600张图像，其中有4000张512x512、2000张512x768、1300张768x768及其他分辨率的图像。与 AIDV1 相比，AIDV2 在细节、笔触、动作、透视、场景关系和手部上的表现有了很大的进步。AIDV2在低分辨率和提示词较短的情形下，细节不如AIDV1。AIDV2 的风格化更强。你可以尝试写入一些风格的名字来触发特定且一致的风格。AnimeIllustDiffusionV2.0 trained an additional 400 epochs on top of 1.0.I improved the training set images. There are now a total of 9600 images in the entire training set, with 4000 images in 512x512, 2000 images in 512x768, 1300 images in 768x768 and the other in other resolutions.Compared to AIDV1, AIDV2 is a huge improvement in detail, brush strokes, dynamic scenes, perspective and gesture drawing. But AIDV2 is not as detailed as the AIDV1 at lower resolutions and short prompt length. AIDV2 is more stylised. You can try writing some style names to trigger a specific and consistent style.

animeIllustDiffusion_v20.safetensors

AnimeIllustDiffusionV2.0

aidv2.10

aidv2.10_beta1

AIDV2.91 - Weak

AIDV2.91 - Strong

AIDV2.9

AIDV2.8

AIDV2.7

AIDV2.6

AnimeIllustDiffusionV2.5

AnimeIllustDiffusionV1.0

Files with SHA256: d7d9617d035230174117e593b6605ee25c8ea5f4d3fd8be3b567fe29a9b449f9