Is it just me, or its the Model used for T2I buns

while this could be just a skill issue on my part, but whatever model being used for t2i actually sucks, constantly giving extra limbs, messes up the face, and constant. style. inconsistency. no matter what i put for my prompt and negative prompt it feels like the creator just spun a wheel of crap models. while yeah it gives out good images once in awhile, its nowhere near as good as it could be. i wish i had the money, and knowledge to do it myself with a better model like wai-nsfw-illustrious-v90-sdxl or something similar. i cannot be the only one that thinks this.

4 comments

It's not the model, I think. I tried it on my comfyUI, chroma v48, it barely ever gives me extra limbs. It could be the resolution. The model was trained in 1024x1024 format, and the options offered on perchance don't match. But to compensate for that perchance offers great image quality, the workflow (or loras) or whatever sampler/scheduler combinations... I've tried it all with the latest chroma and couldn't replicate it. It's amazing here on perchance.
yeah, in the new engine, the variation in image drawing style is very large. In the previous engine, the quality or realism may have been worse, and so on. But in that, all the images were similar to each other, even though the sids were random. Therefore, every change in the prompt and negative prompt was very strongly felt there, and the result of the image was depending on your skill in using the layout and priorities there. And now you can get beautiful images without a single word in the prompt, but if you want to get a specific image and style, they won't let you do it.
FLUX Chroma (Perchance T2i model) is literally the best model on the market haha.
I've collected what prompt I can find on Chroma training data here (see screenshot above for output): https://huggingface.co/datasets/codeShare/chroma_prompts/blob/main/parquet_explorer.ipynb
So try mimicing that prompt format 👍
aesthetic 11