- Source: Midjourney V3
- Resolution: 1024x1024
- Image Count: 91
Curated Collection of Jovian Spaceships, Skies of Jupiter
Description
for use with Stable Diffusion Aesthetic Gradients
FAQ
Comments (4)
3418Dec 13, 2022
CivitAI
Hi, It worked fine on first run, then when I restarted SD, it stopped working and it gives an error. All gradients stopped working.
"call_queue.py", line 45, in f
res = list(func(*args, **kwargs))"
"\aesthetic_clip.py", line 205, in __call__
tokenizer = shared.sd_model.cond_stage_model.tokenizer"
"venv\lib\site-packages\torch\nn\modules\module.py", line 1207, in __getattr__"
"raise AttributeError("'{}' object has no attribute '{}'".format("
"AttributeError: 'FrozenOpenCLIPEmbedderWithCustomWords' object has no attribute 'tokenizer'"
afaik SD 2.0/2.1 are not yet updated to support the Aesthetic Gradients yet. This must be done by whoever is developing the UI that is being used. In the case of Automatics1111 i can say that if you never updated 1.4/1.5 models are still working. There are many UI's you can use but this is the most common. I'm hoping that they will fix this. I'm helping to test another project called AI Nodes which is still working with this, but i know the developers had to do some code to allow this to work. My hope is that soon (if people ask) Automatic will be able to support 2.0/2.1 models. When there are code changes this can happen, so i keep a separate install which i update and check things didn't break, before updating my main Stable Diffusion UI. Thanks so much for trying out the Gradients, i have A LOT more to share with the community
Nice! Would love to try this. *Can we get a ckpt version please?
i'll pass the dataset on to axsthxticroot later today