Qwen-Image-2.1 Fun-Acc PDD 4-Step · ComfyUI

Alibaba PAI 原版模型的 ComfyUI 配套文件。保留全部 528 个原始张量,仅添加识别元数据;四个输出头需由配套采样节点按步切换。不能作为普通 LoRA 直接接 LoRA Loader 使用。

使用

  1. 安装配套 ComfyUI 节点。
  2. 下载本仓库的 Qwen-Image-2.1-Fun-Acc-4Step-PDD-T8.safetensors,放入节点目录的 models/ 文件夹。
  3. 准备 Comfy-Org 的 Qwen-Image-2.1 基础模型、文本编码器和 VAE,重启 ComfyUI,下载并拖入节点仓库的文生图或图片编辑工作流。图片编辑需在 Load Image 中上传参考图。

固定 4 步,支持文生图及图片编辑。原始模型修订:f7545234760e1847cd8e89e52bd951cb0b7e327f。本文件 SHA-256:6f6fc81bea6ce13f2515e21cc2b62de374fa77ecc43f18988028bbf1b3403ee4。模型授权与署名见 LICENSE 和 NOTICE。

English

This is the companion ComfyUI file for Alibaba PAI's original Fun-Acc 4-Step adapter. All 528 tensor payloads are unchanged; only identification metadata was added. Its four step-specific output heads require the matching sampler node. A standard LoRA Loader cannot run this PDD bundle correctly.

Install the node, put Qwen-Image-2.1-Fun-Acc-4Step-PDD-T8.safetensors in its models/ directory, and prepare the Qwen-Image-2.1 base assets. Download and drag in the text-to-image or image-editing workflow. For editing, upload a reference in Load Image. The schedule is fixed at four steps. Model terms and attribution are in LICENSE and NOTICE.

链接 / Links

平台 / Platform 链接 / Link
B站 / Bilibili T8star
YouTube T8star-Aix
API Seedance API
免费画廊 / Free gallery OpenZhenzhen
在线 AI 应用 / Online AI apps RunningHub
ComfyUI 整合包 / Package 夸克网盘 / Quark Drive
Hugging Face t8star
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for t8star/Qwen-Image-2.1-Fun-Acc-LoRAs-Comfy

Adapter
(72)
this model