Chechpoint/model consistency

[Giusemoli312]

Complete noob here. The problem I have encountered recently (in the last few days) is a lack of coherence with the images previously produced with the models using the img2img function. The notebook works correctly, the model is loaded, the VAE is the correct one that I have always used. When I try to create an image with img2img (also txt2img but the effect is less evident) the effect I get is something different than the basic checkpoint but which does not have the characteristics of the chosen model. This happens to me with all models. As mentioned, I am a complete noob and therefore I don't even know if I managed to explain the problem clearly. I thank anyone who will try to help me solve the problem, thank you.