Instructions to use facebook/DiT-XL-2-256 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use facebook/DiT-XL-2-256 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("facebook/DiT-XL-2-256", 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
Confusion on Scheduler
#2
by Gaaaavin - opened
In the scheduler config file, it specifies the scheduler to be DDIMScheduler. However, this example uses DPMSolverMultistepScheduler. Which one is correct? Which one reflects the original training in the DiT paper where the loss is computed in two parts?
The default number of sampling steps also seems to be 250 instead of 1000 (https://colab.research.google.com/github/facebookresearch/DiT/blob/main/run_DiT.ipynb)