import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("kimi000/quiet-canvas-27", dtype=torch.bfloat16, device_map="cuda")
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
image = pipe(prompt).images[0]YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Z-Image AlphaGRPO DVReward - Step 500 EMA
This is a complete native Diffusers Z-Image pipeline exported from
zimage_base_diffusionnft_alphagrpo_dvreward_alphagrpo20k_512px_16step_cfg4_28prompts_7train_1dvreward_tp1_2node_nrt
at Step 500. The EMA LoRA (rank 256, alpha 256) is merged into BF16 base
weights, so FAR and PEFT are not required for inference.
python demo.py --prompt "A red cube beside a blue glass sphere."
The helper uses CFG 4, fixed scheduler shift 6, dynamic shifting disabled, and an explicit 50-update sigma schedule ending at sigma 0.
- Downloads last month
- 1