Hot Fantasy A.I. 2.0
Requires: vae-ft-mse-840000-ema-pruned.ckpt (https://huggingface.co/stabilityai/sd-vae-ft-mse-original/blob/main/vae-ft-mse-840000-ema-pruned.ckpt)
jdolinschi/tagger
(https://github.com/jdolinschi/tagger)
Donate for respect : https://www.buymeacoffee.com/hotfantasyai
Description
FAQ
Comments (9)
Can you add, if possible, a smaller 2gb pruned fp16 no-baked-vae safetensor version ?
Works great in A1111 and is awsome! Any chance to get it working in InvokeAI too? The In/outpaint feature is much better then in A1111.
Loading this model in InvokeAI results in nothing just junk images =( some gravestones, streetsigns and so on. And yes i added the required VAE in config
I'll be honest, but I don't use InvokeAI, so unfortunately I don't know how to make it to working
Is there such adifference in technics? InvokeAI use SD1.5 and SD2.1 Models. to bad that it doesn´t create and logs with errors.
@AsaTyr the problem it is because I disassemble the model and rearrange a few things in them
yet another, huh? I get so tired of A1111 being the de facto standard.
@imn0mn InvokeAI ist pretty stable, easy and lightwight. still its extreamly focussed on in and out painting. they even have a simple installer that takes care of everything. got pretty good results. and if something is off: use in/out painting to correct things.
@AsaTyr Couldn't agree more. It's my tool of choice. Which is why things that only work in A1111 annoy me - A1111 is interesting and useful for the features it gets but too unstable to really be a serious use tool.
@imn0mn don´t get me wrong, a1111 is still a great tool. specialy for bulk rendering. and its sidecar file export makes it special usefull for imageboards like hydrus network. ^^
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.