Wan2.1-T2V-1.3B-SOLACE

LoRA adapter from SOLACE (Self-cOnfidence reward for aLigning text-to-imAge models via ConfidencE optimization), CVPR 2026.

SOLACE applied to the Wan2.1-T2V-1.3B text-to-video model, using the model's own denoising confidence as an intrinsic reward (no external reward model at training time).

• Base model: Wan-AI/Wan2.1-T2V-1.3B-Diffusers
• Method: SOLACE intrinsic self-confidence reward (built on Flow-GRPO)
• Code: https://github.com/wookiekim/SOLACE
• Adapter type: PEFT LoRA (rank 32) on the Wan 3D transformer

Usage

import torch
from diffusers import WanPipeline
from diffusers.utils import export_to_video
from peft import PeftModel

model_id = "Wan-AI/Wan2.1-T2V-1.3B-Diffusers"
lora_ckpt_path = "wookiekim/Wan2.1-T2V-1.3B-SOLACE"
device = "cuda"

pipe = WanPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16)
pipe.transformer = PeftModel.from_pretrained(pipe.transformer, lora_ckpt_path)
pipe.transformer = pipe.transformer.merge_and_unload()
pipe = pipe.to(device)

frames = pipe(
    "a cat walking across a sunlit kitchen floor",
    height=480, width=832, num_frames=81,
    num_inference_steps=50, guidance_scale=5.0,
).frames[0]
export_to_video(frames, "solace_wan.mp4", fps=16)

Citation

@inproceedings{kim2026solace,
    title={Improving Text-to-Image Generation with Intrinsic Self-Confidence Rewards},
    author={Kim, Wookyoung and others},
    booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
    year={2026}
}

Acknowledgments

This work builds upon Flow-GRPO by Jie Liu et al.