Flow-DPPO: Divergence Proximal Policy Optimization for Flow Matching Models
Paper • 2606.11025 • Published • 41
Text-to-Image
Diffusers
Safetensors
StableDiffusion3Pipeline
flow-dppo
unirl
stable-diffusion-3.5
stable-diffusion-3.5-medium
import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("Eculid/sd3.5-flowdppo", torch_dtype=torch.bfloat16, device_map="cuda")
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
image = pipe(prompt).images[0]SD3.5 Medium Flow-DPPO
This repository hosts a Flow-DPPO checkpoint trained from Stable Diffusion 3.5 Medium using the UniRL Flow-DPPO recipe.
Links
- Paper: https://huggingface.co/papers/2606.11025
- Code: https://github.com/Tencent-Hunyuan/UniRL
- Base model: https://huggingface.co/stabilityai/stable-diffusion-3.5-medium
Notes
This checkpoint is provided as a community-contributed artifact for research use. It follows the usage and limitations of the Stable Diffusion 3.5 Medium base model.
- Downloads last month
- 5
Model tree for Eculid/sd3.5-flowdppo
Base model
stabilityai/stable-diffusion-3.5-medium