If you like my content, support me at:
https://ko-fi.com/duchaiten
Patreon is here (but there's nothing in it):
https://linktr.ee/Duc_Haiten
I created my Discord server. I'm still quite new to Discord server management, feel free to share your opinions.
https://discord.gg/vKEW6jqa49
This is where my style models trained with V-pred will be collected.
style_27
style_18, style_26
style_45
Negative prompt: (worst quality:1.5), (low quality:1.5), (normal quality:1.5), lowres, bad anatomy, bad hands, multiple eyebrow, cropped, extra limb, missing limbs, deformed hands, long neck, long body, signature, username, artist name, conjoined fingers, deformed fingers, ugly eyes, imperfect eyes, skewed eyes, unnatural face, unnatural body, error, painting by bad-artist
Description
FAQ
Comments (7)
Sad I do not use confy ui, I am not able to use Vpred models, I would love to use some specialised style models, with more definits styles, with fix look but wide concepts understandings. what do you think about Vpred models did your position changed compared to Snapvu ?
V-pred outputs look way better than EPS, in my opinion; typically less unaesthetically fuzzy, better colors, far superior contrast, more-accurate prompt following for fine detail. If you're using the master (default) branch of A1111, you can use V-pred models (NAI-derived, at minimum) presuming you put the appropriate .yaml file in the same directory as the model. Filenames base must exactly match (e.g., the V-pred checkpointA_v1.0.safetensors will need an accompanying config file checkpointA_v1.0.yaml). Don't forget to reload models or restart your A1111 instance before generations.
Some creators here on Civitai post their configuration files alongside their v-pred models, but most don't. You should be able to reuse one from another model (like https://civitai.com/models/1195325?modelVersionId=1716418 from @nuko_masshigura ) and simply rename the .yaml.
Good luck.
I have posted a screenshot of the settings you can see I am using ForgeUI, but you need Forge2 extras, and
Set model prediction type v_prediction, then go to UI settings find and select Zero Terminal SNR
@DucHaiten thank you, I followed your instructions and it works nice , I will test soon your model.
Hello! I've been following and using DucHaiten's models for a long time. I tried using this one and, sadly, it did not work for me on reForge. Can someone please share me a link to the location where I can download forge2? Thank you : )
I have posted a screenshot of the settings you can see I am using ForgeUI, but you need Forge2 extras, and
Set model prediction type v_prediction, then go to UI settings find and select Zero Terminal SNR
@DucHaiten Thank you for helping me. : 3
I left a post showing what the model is capable of. It generated a very nice image with some very lovely colors. ^w^
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.