Stable Diffusion v1.5 [bf16/fp16] [no-ema/ema-only] [no-vae] [SafeTensors] [Checkpoint]
=====================
Disclaimer:
I'm just a Script kiddie. I have no idea what i'm doing.
So, keep calm and know nothing.
=====================
All the credit goes to CompVis | Stability AI | runwayml.
https://huggingface.co/CompVis
https://huggingface.co/stabilityai
https://huggingface.co/runwayml
SD v1.5:
https://huggingface.co/runwayml/stable-diffusion-v1-5
VAE ft-mse-840000-ema-pruned:
Description
v1.5
no-ema.
no-vae.
bf16 and fp16.
FAQ
Comments (5)
Why isn't this available for the online image generator? Isn't this the very trademark model of stable diffusion?
why is this only 1.7gb but the ones on hugginface are twice that???
This shit doesn't work. tried using it as a model for a lora that said was modeled after sd 1.5 and got a grey screen. 0/10
1.5 is the SD that everyone uses, almost all models are for that category. This one works if you put it in the VAE folder (not the stable diffusion folder) however this is just a checkpoint-trained, it's best to find a base checkpoint.
I can confirm that it works, are you using low resolution like 512x768? Also recommended clip skip is 1. For reference I created this image just yesterday with an SD1.5 Lora of my own: https://civitai.com/images/71584593
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.