Instructions to use Photoroom/prxpixel-t2i with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use Photoroom/prxpixel-t2i with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Photoroom/prxpixel-t2i", 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
more info please ? it based on what ?
#2
by pikkaa - opened
and provide FP16 , FP8 please ?
Hi! Thanks for your interest π
PRXPixel is a ~7B pixel-space text-to-image model built on PRX, Photoroom's own diffusion architecture.
For a deep dive into the design and training, including what it's based on and how it evolved, check out our blog post series:
- https://huggingface.co/blog/Photoroom/prx-part1-architectures
- https://huggingface.co/blog/Photoroom/prx-part2
- https://huggingface.co/blog/Photoroom/prx-part3
We currently only ship bfloat16 weights, as that's what the model was trained and evaluated in. FP16 and FP8 variants are on our radar. We'll update the repo when they're available.
Bertoin changed discussion status to closed