Closed Ainaemaet closed 8 months ago
I merged today @JonathanWac's solution for accepting more file formats, and if no loading time issue because large files from Civitai, I will found some new solution to read images from existing model directories.
This node pack is pretty wonderful for the most part - a lot of useful stuff here (thank you).
Unfortunately, my main reason behind wanting to use it is for the visual checkpoint/lora loading, and the restrictions and tediousness of the setup just makes it something that I don't want to bother with - especially considering all of the work that already goes into crafting a good workflow!
The hard JPG limit on the previews is terrible; especially given that many people use the Auto1111 Civitai Browser + extension to download and organize their models; so the best thing for this ComfyUI_Primere_Nodes pack, would be if it could utilize the already existing previews (png) and model-info, as downloaded and saved from CivitaiBrowser+ extension.
Pythongosssss 'ComfyUI-Custom-Scripts' does this, which makes bouncing back and forth between setups a breeze; and while you include some of their nodes in your workflows, the fact that their checkpoint/lora loader can utilize the already existing .png previews with no need to move files around is a huge Godsend, and one that makes that nodepack considerably more useful atm.
Please implement similar (or better) functionality into ComfyUI_Primere_Nodes - it would make this undeniably one of the best packs available out there for ComfyUI.