minsky91

I too would really appreciate it if the Forge version of MultiDiffusion would have implemented all the features of the original one, in particular Noise Inversion. And like for other...

It is indeed a very special preprocessor that can be used for a number of purposes beyond image restoration (where [SILVI2 ](https://civitai.com/articles/5834/silvi-v2-upscale-method-super-inteligence-large-vision-image) is the most brilliant example). I use it...

I am getting very good results with Hyper-SDXL-8steps-CFG-lora in Forge / SDXL, equivalent or actually better than those with SILIVI. So hang on, I intend to publish results here and...

Well, the plugin already appends a hefty cryptical substring to the layer name upon upscale, e.g. (2075337012) - do we really need that? Instead, at least a substring indicating whether...

Thanks [megagrump](https://github.com/megagrump), I didn't know about this switch. Setting it to some very large value and restarting Comfy didn't change much though: when attempting to 2x upscale a 7200x9600 image,...

UPDATE. I receive the message “_Error Failed to read JSON 'Maximum request body size 104857600 exceeded, actual body size 104861476' with message: Unexpected character encountered while parsing value: M. Path...

For backward compatibility, both in pixels and %.

A few issues in what I see in your screenshots: 1. if you want to use a genarlist/realistic checkpoint like Juggernaut (which is a right choice), don't select it within...

Doing that in chunks will not address the issue of the total object removal area remaining disproportionately large against the image itself, you will just make your work more laborious...

After posting the above I found that the same thing applies to the inpaint fooocus models for SDXL and some other model files.