2023/4/20 Updated yden-V3
Since the fusion formula of the V3 version is completely different from the previous two versions, the composition has changed a lot, but compared to the previous two versions, the V3 version has a good performance for the hands, especially under the condition of high steps .
All example images are generated using the txt2img function, and the VAE used is native VAE.
Due to my poor English proficiency, I used ChatGPT to translate Chinese into English. Please bear with me.
The following is the translated description:
Due to the completely different fusion formula between V3 version and the previous two versions, there are significant changes in composition. However, compared to the previous two versions, V3 version has good performance for hands, especially under high steps conditions.
All example images were generated using txt2img function and native VAE was used.
2023/3/25 Updated yden-V2
As a new version of yden, it is a model created to generate anime-style pictures like V1, and it has also been compressed to fp16, which can reduce the size of the model while maintaining the original effect and speed.
Compared with the V1 version, the pictures generated by the V2 version will be more detailed, and the composition will have a relatively large change.
There is no difference between the parameter settings and the V1 version. In addition, it is strongly recommended to use the Hires.fix function in order to better utilize the potential of this model.
The following is the introduction of yden-V1
This is a model for generating anime style, I have compressed the model to fp16, which can reduce the model size while maintaining the original effect and speed.
The example images are all generated by the original VAE of novelai.
The following are recommended parameters:
Sampler: DPM++ 2M Karras
CFG Scale: 7
Negative prompt: (worst quality, low quality: 1.4)
Hires.fix: R-ESRGAN 4x+ Anime6B
Hire steps: 10
Denoising strength: 0.35-0.6
Upscale by: 1.5-2
Visit Official Website