Instructions to use prompthero/openjourney with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use prompthero/openjourney with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("prompthero/openjourney", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
No negative prompt function?
Hello, I'm using this wonderful model to persuade myself some art, but I noticed that the negative prompts I was entering were simply appended onto the end of my positive prompts instead of being read separately. Does this model not yet have negative prompt functionality yet or am I doing something wrong here?
Info -
OS: Win10
Gui: NMKD Stable Diffusion GUI (desktop app as opposed to the webui)
model: mdjrny-v4.ckpt (obviously)
vae file: vae-ft-mse-840000-ema-pruned.ckpt (default, it ships with the GUI)
Hi! We were not aware that negative prompt is something about the .ckpt
If it is, happy to hear how to implement it.
Gui: NMKD Stable Diffusion GUI (desktop app as opposed to the webui)
Please contact the app author. The model is not responsible for inappropriate pipeline usage.
Mmmm. I have done some testing and I believe it's simply a UI bug with the desktop app. Not a problem with the model you've trained. Negative prompts do seem to function as intended, however, in this particular gui, if I go over a certain undisclosed character limit then the negative prompts are simply displayed at the end of the positive prompts on the GUI rather than in the designated negative prompt location right beneath it. The metadata and the image output itself seem to be unaffected by this UI bug though, thankfully.
I'll submit this bug to the developer of the app. Thanks for your time.
