Remember to read the instruction!
Please use Pixel version, if you don't have good understanding on Latent version.
This will provide you
Crazzzzzzzzzzzzzzzzzzzzzzzzzzzzzy
Amazzzzzzzzzzzzzzzzzzzzzzzzzzing
Higggggggggggggggggggggggh
Quaaaaaaaaaaaaaaaaaaaaaaaality
Images!!!!
just try it!
remember use my custom node and modify the parameter carefully!
adding jc2 version. Remember to download my nodes
https://github.com/TTPlanetPig/Comfyui_TTP_Toolset
https://github.com/TTPlanetPig/Comfyui_JC2
check your Max VRAM before you run it :)
at least 10GB for nf4 and 20GB for BF16
Following is for Pixel version instruction:
Just enjoy, and try not to touch my setting if you are not sure what you are doing.
if you can run flux official properly, you can run this one
Following is for Latent Version instruction:
This is a workflow for my simple logic amazing upscale node for DIT model. it can be common use for Flux,Hunyuan,SD3 It can simple tile the initial image into pieces and then use image-interrogator to get each tile prompts for more accurate upscale process. The condition will be properly handled and the hallucination will be significantly eliminated.
you can download my newly build node here:
https://github.com/TTPlanetPig/Comfyui_TTP_Toolset ,please update to latest version
A updated version for Flux Hi-res fix!!!
adjust the denoise to control the difference and possible latent block issue.
You will be shocked by the detailed high-resolution photos!!!
Key parameter to adjust:
1. tile block numbers: recommended 2*3 or 3*2 or 3*3
2. overlap rate: recommended 0.15-0.2! Please update my node to latest version for this function.
I also used the node:
https://github.com/yolain/ComfyUI-Easy-Use
and
https://github.com/kijai/ComfyUI-Florence2
Description
FAQ
Comments (23)
I'm getting this error on J2C node:
JoyCaption2_simple
object of type 'NoneType' has no len()
Does anyone know why? Did I forget to install something? Let me know. Thanks!
NoneType could be resulted by different reason, try to check the complete error information
@DucaAI sorry，I am using a 24GB Vram card, I just can't try the low VRAM condition very well,....
@DucaAI 24GB VRAM and 64GB RAM
I was getting similar error, but can't remember now I fixed it. Do you have a models/Joy_caption folder in comfyui? Pretty sure I had to create that manually and from that folder run "huggingface-cli download fancyfeast/joy-caption-alpha-two --repo-type space --local-dir ./" in order to grab the models. You should end up with a models/Joy_caption/cgrkzexw-599808 folder. That seemed to work for me at least.
I have the same issue, 24GB vram card, has anyone come up with a fix?
@dirtyape this is the way!
@dirtyape thank you very much, this helped
@dirtyape you are the best thanks you
@dirtyape Everything works fine with another LLM. So I'm really not sure how should I download and how to call it and where to actually put LLM (unsloth/Meta-llama) model. If you could help me with that, please hit me in DM (I don't want to bother you). Cheers
@Hogar315 my code should download the model automatically except the joy2 lora model, it has to be manually placed:)
@dirtyape what does that even mean ? I'm pretty sure I speak English but this, whatever you've explained doesn't seem like it's from this world. Is this how you programmers explain to layman ?
老白万岁
The ttp joycaption node isn't working for me. I switched to the NYJY version (which previously didn't work, and it works fine now. Definitely not getting the same level of detail with SD35 as I do with Flux.
this is awesome. Thank you! Also works very well with the de-distilled versions of Flux after a few adjustments.
What the hell...
This is amazing! This approaches real photo quality and gets rid of the "Flux oily clay skin" problem. Awesome work!
please post your output( ´▽｀)
Hey,
I'm getting this ...
TTP_Image_Assy
index 1 is out of bounds for dimension 0 with size 1
Any ideas how to fix this ?
do you change the workflow? i
" TypeError: object of type 'NoneType' has no len() " what is this ? Can someone explain this in a manner a normal, layman, a person with zero programming knowledge can understand ?
joycaption2 lora model can't be download auto, please read the github to place it in the correct location in your model folder.
thanks
The LLM-part does "nothing". Just using a prompt like "Masterpiece, realistic photo" or something for every tile works just as good and takes up way less memory and time 🤔 Depending on what the image to be upscaled looks like it might be a good idea to segment the tiles differently. I find that when prompting for cars for example. If they are parked on a road with distinctive wear going along the track the upscale tend to create different tear-patterns for each tile. Segmenting the image 1x6 can be better than leaving it at 3x2. Good workflow, but I'm not sure how it differs from just using Ultimate SD Upscale which seem to do this in one node 🤨
the reason I create it is because I can't good result from Ultimate SD upscale. one more thing I can use this one play with the latent, which SD upscale can't
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.