Hello Civitai Community,
I'm thrilled to share an update on a recent project of mine. After some dedicated work, I've developed a highly effective text-to-image model. This innovation results from integrating the DPO model from Hugging Face with several advanced counterparts, including Juggernaut7XL, ALBEDOXL, MEARGEHEAVEN, and a model of my own design. The outcome is a unique fusion that showcases exceptional prompt adherence and semantic understanding, it seems to be a step above base SDXL and a step closer to DALLE-3 in terms of prompt comprehension. Notably, this model excels in interpreting and adhering to the given prompts, focusing more on semantic accuracy than on ultra-high-fidelity image generation.
In my quest to refine this model, I encountered a small hiccup – I inadvertently deleted the model from Civitai. However, I've quickly rectified this and reuploaded the model for community access. I'm also excited to announce that version 1.1 is already underway, aiming to bring further improvements and expanded capabilities.
I'm eager to hear your thoughts and feedback on this version. Your insights are invaluable in guiding the ongoing development of this project. So, please stay tuned for the upcoming iteration and thank you for your continued support
Settings for OpenDalle v1.0
Use these settings for the best results with OpenDalle v1.0:
CFG Scale: Use a CFG scale of 8 to 7
Steps: 60 to 70 steps for more detail, 35 steps for faster results.
Sampler: DPM2
Scheduler: Normal or Karras
Description
FAQ
Comments (13)
can this model be used for commercial purposes?
@DataVoid?
What do you mean by the DPO model from Hugging Face ? Did you change the clip or did you train on the output of something? I'm very curious :)
Excellent model, thank you so much for this !
Using your model, my output is always a pure black color block with no content. Other models with the same settings have no problem. Have you encountered this problem and what is the reason?
propobly a vae problem (you got all NaN-ed), change vae or turn it auto. Switch to fp32 vae and check again
Amazing model, it's very good for prompting.
do you recommend v1.0 or v1.1 fro open dalle?
@ThricePi I'm using 1.0, I haven't tested the 1.1 version
If you guys like this model you may want to check my model https://civitai.com/models/239836/dpoxlplus
which open dalle version is best in peoples opinion v1. or 1.1?
Hey, thank you for the model!
I made a video comparing DALL-E 3 with OpenDALL-E v1.1 including my thoughts according to my own use.
https://youtu.be/8nhKLtNWFPk
Hope this helps! :)
I tried a long time to get the fur really fuzzy on this checkpoint. Hours of failure.....but the prompting is so 100% on point. ...I had to do it. Its by far the best checkpoint when it comes to prompt following - there is nothing near to it - its unique - a diamond. Done by a genius. Thx for sharing.
i fell this is a hidden gem, chat gpt brought me here lol