Aller au contenu
ModelPrésentéExterne
Télécharger

Wan 2.1 T2V 1.3B SOLACE

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

TypeLoRAModèle de baseWan-AI/Wan2.1-T2V-1.3B-DiffusersVersionadapter_modelBClasse

Source du modèle

Type
LoRA
Taille du fichier
90,1 MB
Modèle de base
Wan-AI/Wan2.1-T2V-1.3B-Diffusers
Version
vadapter_model
Mis à jour
10 oct. 2026
Téléchargements
5
Source
Hugging Face

Description de la source

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).

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.

Sources

1 source
Hugging FacePrincipalVersion a68333b95a3c8f9f4ef56c4a432c8cac98f76a1b
huggingface.co/wookiekim/Wan2.1-T2V-1.3B-SOLACE/tree/a68333b95a3c8f9f4ef56c4a432c8cac98f76a1b

Vérifié 24 août

Artefacts du modèle

1 artefact

adapter_model.safetensors

safetensors · 90,1 MB · SHA-256 113bce1af109…7f53 · Hugging Face

Download file

Extraits de sources

2 extraits

--- base_model: Wan-AI/Wan2.1-T2V-1.3B-Diffusers library_name: peft license: apache-2.0 pipeline_tag: text-to-video tags: - text-to-video - wan - lora - peft - flow-grpo - solace - reinforcement-learning --- # Wan2.1-T2V-1.3B-SOLACE LoRA adapter from **SOLACE** (**S**elf-c**O**nfidence reward for a**L**igning text-to-im**A**ge models via **C**onfidenc**E** 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`](https://huggingface.co/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 ```python 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 ```bibtex @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](https://github.com/yifan123/flow_grpo) by Jie Liu et al.

Source context: 5 downloads · 0 likes · Pipeline text-to-video · Library peft · Repo wookiekim/Wan2.1-T2V-1.3B-SOLACE