## Model links
text_encoders
qwen_3_4b.safetensors
Model Storage Location
```
📂 ComfyUI/
├── 📂 models/
│ ├── 📂 text_encoders/
│ │ └── qwen_3_4b.safetensors
│ ├── 📂 loras/
│ │ └── qinglong_detailedeye_z-imageV2(comfy).safetensors
│ ├── 📂 diffusion_models/
│ │ └── z_image_turbo_bf16.safetensors and z_image_turbo-Q4_K_S.gguf
│ └── 📂 vae/
│ └── ae.safetensors
```
Read the notes in the workflow for more help, or join the Discord community:
🎮 Join Discord : https://discord.gg/DZEenb5wGc
🔗 Get the Pro Version of This Workflow with More tweaks and enhancements on My Patreon.
Description
Now With Detailer Daemon, Ultra Fast and multi Quality Options.
FAQ
Comments (30)
V3 works really well, I've been struggling to get any useful images out of ComfyUI after switching from ReForge.
One question, I don't understand how to adjust strength for loras in this workflow. Frankly, I'm having trouble with penises.
Turn on Lora nodes, start with 0.75 and adjust the Lora as required.
Depends on Lora too.
not you dude, its ZIT having trouble with penises. 😊
wait for the good days 😂
@Stable_Yogi Thanks, that helps, you make so many great tools!
Where is the lora qinglong_detailedeye_z-imageV2(comfy).safetensors used? I downloaded it and placed in the folder, but wasn't able to see it mentioned in the flow.
Ignore that, its just there. you can use any lora if you want.
@Stable_Yogi Thanks. I wasn't sure if it was supposed to be used somewhere as it's listed to be downloaded and I was missing something.
@Stable_Yogi lora qinglong_detailedeye_z-imageV2(comfy).safetensors The download link is broken
@liguichuan939163Â Ignore its its outdated now.
how do i have to train my character lora to get great results?
Pretty simple => Contact stable_yogi 😊 https://discord.com/invite/FAgfAeMqys
This is really impressive and worked as soon as I downloaded the necessary nodes (Which seems to never happen for me...). Question: how do you set the LoRA strength? I can load the LoRA but the strength box is greyed out.
A node "Math expression" is connected to Lora's, check and add your value there.
@Stable_Yogi Got it -- thanks!
Hi, I ran out of memory, I have a 2070 with 8GB dedicated memory, is this not enough? Do I maybe need to configure page file or something to make it work?
Thanks for your work!!
download a lower size diffuser like f8 or lower quant for gguf model
KSampler
'NoneType' object is not subscriptable
i get this error with the workflow. No idea what i have to do.
copy paste the error code in ChatGpt and troubleshot.
Down your transformer < 5.0.0(if)
Finally decided to work with ZiT a bit and anytime I begin working with a new model I look for an example workflow to learn from and build on. I downloaded several to check out but honestly I could not have asked for a better one than this. The Yogi hits another one out of the park, as always.
In which box is the "z_image_turbo-Q4_K_S.gguf" be suppossed to set?, thanks in advance.
GGUF box
@Stable_Yogi Thanks dude, I underlooked the final part, thanks for your insight :3
This error popped up, any idea how to fix it?: Node 'ID #454' has no class_type. The workflow may be corrupted or a custom node is missing.: Node ID '#454'
custom node is missing, reinstall nodes
I can't figure out how to use the VR2 upscale node. I'm generating 960 latents, and it takes the 1200x1600 and shrinks it down to 896x1194 and makes everything look like plastic. What am I doing wrong?
Nevermind figured it out lol I needed to input a larger resolution size. The results are glorious. Had no idea this thing existed and I've been using SD ultimate upscale all this time lol
Could you please explain what you mean with 'input a larger resolution size'?
EDIT: Nvm I figured it out, whoa what a difference.
I do wonder how TWO people found the solution, were compelled to update their comments, but somehow decided not to include what they did to improve their results...
I used the face lora I wanted, but a different face came out. How do I fix this?
VERY GOOD WORKFLOW!!
THANK YOU!
Very good workflow for very good models. I'm sorry for not supporting you on Patreon, but that doesn't stop me from appreciating you. Thanks...