NOTE: Civit has made the download UI confusing. For the v4.0 model, it contains four versions:
BF16 UNET (11GB)
FP8 UNET (6GB)
Q6 GGUF UNET (5GB)
FP8 Checkpoint (16GB)
I uploaded them very descriptively but Civit has stripped away the descriptions so it's now confusing to see what you're downloading. The FP8 checkpoint is an all-in-one version that contains the VAE and Text Encoder. For the UNETs you need to provide those extra models.
I've used 100s of curated images to help add some better color, contrast and detail to the Z-Image Turbo model. It's not a huge difference yet, but it is noticeably better in side by side tests. I'm starting out a bit conservative because I don't want to mess up any of the great compositional features that Z-Image offers.
Use the workflow embedded in the showcase images for the model version to get the best results. Check the About this version section for links to the workflow and recommended settings. Z-Image currently mangles nipples and genitalia so my workflow uses SDXL detailers to refine just those aspects of the image.
P.S. I would appreciate any constructive feedback on how to make this model better.
Description
This is just the UNET version of the v3.0 AIO model.
FAQ
Comments (17)
Is there any way to have this in diffusers format?
They're all diffusion unet models except for the AIO versions which are single file checkpoints that include the unet, text encoder and vae.
@GBRX I mean diffusers (folders) like this https://huggingface.co/Tongyi-MAI/Z-Image-Turbo/tree/main
@Hikarias I see. I'm probably not going to be separating the model components like that. I'm not even sure how to do it.
Hi, I got an error while running your workflow.
RuntimeError: ERROR: clip input is invalid: None
And it shows Refining Negative Prompt section has some problems. Any idea what it could go wrong? thanks.
Which specific workflow and model are you using? It just sounds like the yellow connection to the negative Clip Text Encode node got disconnected somehow.
https://civitai.com/models/2172100/gonzalomo-z-image-refiner-workflow?modelVersionId=2464585
the one you recommanded. The model is using this one Zpop
it's the negative prompt clip. positive has no error.
@robinjfor123884 I had to use delete the prompt node and use the standard clip text node
Runtime error . status shows.. ERROR Clip input is invalid none
Which workflow are you using?
@GBRX I am using Comfyui workflow, the status shows Runtime Error when passing Clip Text Encode
@Az111 Which specific workflow are you using? Which Clip Text Encode is showing the error? positive or negative? Is the yellow line connected between the Clip loader and the Clip Text Encode node?
@GBRX A runtime error occurs during the positive to negative clip process. The yellow lines are all connected and aligned on their paths. I like the model, but it doesn't work perfectly.
@Az111 Which specific workflow and model version are you using??? I can't help if I don't know what you're referring to.
@GBRX https://civitai.com/models/2192562?modelVersionId=2598704
@GBRX It's running normally. I have to set it up in the comfyui node. And it's running normally.