alt.freeweights.video / Wan2.2-Fun-5B-Control
From: alibaba-pai <huggingface.co/alibaba-pai>Newsgroups: alt.freeweights.videoSubject:

Wan2.2-Fun-5B-Control

Date: 2026-11-15Commit: b8bc1a6Signed: 2DC28CC7B26A283B ✓

Image to video model. Safetensors weights, 24.2 GB in all. Published by alibaba-pai under Apache-2.0.

video24.2GBWan2.2-Fun-5B-Controlsafetensorspermissive
FactsParamsnot statedSize24.2 GBFormatssafetensorsTaskimage to videoLicenseApache-2.0From sourcepermissive
$ fw get alibaba-pai/Wan2.2-Fun-5B-Control
# manifest sha256 61e6c9c0dbc0… · 24.2 GB · 12 files
$ fw verify ./Wan2.2-Fun-5B-Control alibaba-pai/Wan2.2-Fun-5B-Control
From sourceHosted (signed)alibaba-pai/Wan2.2-Fun-5B-ControlLast check: 2026-11-15 02:37 UTC

Get help: custom fine-tune or private deploy. We review your setup before we quote.

Fine-tune of Wan2.2-TI2V-5B.

Works with

Overview

Size
24.2 GB
Weights
10.0 GB (safetensors)
Parameters
not stated
Formats
safetensors
Task
image to video
License
Apache-2.0 (permissive)
Files
12, each with a sha256
Publisher
alibaba-pai
Added
2026-11-15
Section
alt.freeweights.video
Lineage

Original release from the publisher.

Base model: Wan-AI/Wan2.2-TI2V-5B

Where these bytes come from
Publisher at the pinned commit→You

Sources are tried in order. A file counts only when its size and sha256 match the manifest. Sources are checked when you ask for them, not on a timer.

Sources
BadgeHostLast checkResult
From sourcehuggingface.co2026-11-15 02:37 UTCpassed

Model card

From the publisher at commit b8bc1a6. Shown as plain text with links; images open at the publisher.

Wan-Fun#

😊 Welcome!

Hugging Face Spaces

Github

English | 简体中文

目录#

模型地址#

名称存储空间Hugging FaceModel Scope描述
Wan2.2-Fun-A14B-InP64.0 GB🤗Link😄LinkWan2.2-Fun-14B文图生视频权重,以多分辨率训练,支持首尾图预测。
Wan2.2-Fun-A14B-Control64.0 GB🤗Link😄LinkWan2.2-Fun-14B视频控制权重,支持不同的控制条件,如Canny、Depth、Pose、MLSD等,同时支持使用轨迹控制。支持多分辨率(512,768,1024)的视频预测,,以81帧、每秒16帧进行训练,支持多语言预测
Wan2.2-Fun-A14B-Control-Camera64.0 GB🤗Link😄LinkWan2.2-Fun-14B相机镜头控制权重。支持多分辨率(512,768,1024)的视频预测,,以81帧、每秒16帧进行训练,支持多语言预测
Wan2.2-Fun-5B-InP23.0 GB🤗Link😄LinkWan2.2-Fun-5B文图生视频权重,以121帧、每秒24帧进行训练支持首尾图预测。
Wan2.2-Fun-5B-Control23.0 GB🤗Link😄LinkWan2.2-Fun-5B视频控制权重,支持不同的控制条件,如Canny、Depth、Pose、MLSD等,同时支持使用轨迹控制。以121帧、每秒24帧进行训练,支持多语言预测
Wan2.2-Fun-5B-Control-Camera23.0 GB🤗Link😄LinkWan2.2-Fun-5B相机镜头控制权重。以121帧、每秒24帧进行训练,支持多语言预测

视频作品#

Wan2.2-Fun-A14B-InP#

|

|

|

|

|

|

|

|

Wan2.2-Fun-A14B-Control#

Generic Control Video + Reference Image:

| Reference Image

| Control Video

| Wan2.2-Fun-14B-Control

|

|

|

Generic Control Video (Canny, Pose, Depth, etc.) and Trajectory Control:

|

|

|

|

|

|

|

|

Wan2.2-Fun-A14B-Control-Camera#

| Pan Up

| Pan Left

| Pan Right

| Zoom In

|

|

|

|

| Pan Down

| Pan Up + Pan Left

| Pan Up + Pan Right

| Zoom Out

|

|

|

|

快速启动#

1. 云使用: AliyunDSW/Docker#
a. 通过阿里云 DSW#

DSW 有免费 GPU 时间,用户可申请一次,申请后3个月内有效。

阿里云在Freetier提供免费GPU时间,获取并在阿里云PAI-DSW中使用,5分钟内即可启动CogVideoX-Fun。

DSW Notebook

b. 通过ComfyUI#

我们的ComfyUI界面如下,具体查看ComfyUI README。 Image: workflow graph

c. 通过docker#

使用docker的情况下,请保证机器中已经正确安装显卡驱动与CUDA环境,然后以此执行以下命令:

# pull image
docker pull mybigpai-public-registry.cn-beijing.cr.aliyuncs.com/easycv/torch_cuda:cogvideox_fun

# enter image
docker run -it -p 7860:7860 --network host --gpus all --security-opt seccomp:unconfined --shm-size 200g mybigpai-public-registry.cn-beijing.cr.aliyuncs.com/easycv/torch_cuda:cogvideox_fun

# clone code
git clone https://github.com/aigc-apps/VideoX-Fun.git

# enter VideoX-Fun's dir
cd VideoX-Fun

# download weights
mkdir models/Diffusion_Transformer
mkdir models/Personalized_Model

# Please use the hugginface link or modelscope link to download the model.
# CogVideoX-Fun
# https://huggingface.co/alibaba-pai/CogVideoX-Fun-V1.1-5b-InP
# https://modelscope.cn/models/PAI/CogVideoX-Fun-V1.1-5b-InP

# Wan
# https://huggingface.co/alibaba-pai/Wan2.1-Fun-V1.1-14B-InP
# https://modelscope.cn/models/PAI/Wan2.1-Fun-V1.1-14B-InP
# https://huggingface.co/alibaba-pai/Wan2.2-Fun-A14B-InP
# https://modelscope.cn/models/PAI/Wan2.2-Fun-A14B-InP
2. 本地安装: 环境检查/下载/安装#
a. 环境检查#

我们已验证该库可在以下环境中执行:

Windows 的详细信息:

  • 操作系统 Windows 10
  • python: python3.10 & python3.11
  • pytorch: torch2.2.0
  • CUDA: 11.8 & 12.1
  • CUDNN: 8+
  • GPU: Nvidia-3060 12G & Nvidia-3090 24G

Linux 的详细信息:

  • 操作系统 Ubuntu 20.04, CentOS
  • python: python3.10 & python3.11
  • pytorch: torch2.2.0
  • CUDA: 11.8 & 12.1
  • CUDNN: 8+
  • GPU:Nvidia-V100 16G & Nvidia-A10 24G & Nvidia-A100 40G & Nvidia-A100 80G

我们需要大约 60GB 的可用磁盘空间,请检查!

b. 权重放置#

我们最好将权重按照指定路径进行放置:

通过comfyui: 将模型放入Comfyui的权重文件夹ComfyUI/models/Fun_Models/:

📦 ComfyUI/
├── 📂 models/
│   └── 📂 Fun_Models/
│       ├── 📂 CogVideoX-Fun-V1.1-2b-InP/
│       ├── 📂 CogVideoX-Fun-V1.1-5b-InP/
│       ├── 📂 Wan2.1-Fun-V1.1-14B-InP
│       └── 📂 Wan2.1-Fun-V1.1-1.3B-InP/

运行自身的python文件或ui界面:

📦 models/
├── 📂 Diffusion_Transformer/
│   ├── 📂 CogVideoX-Fun-V1.1-2b-InP/
│   ├── 📂 CogVideoX-Fun-V1.1-5b-InP/
│   ├── 📂 Wan2.1-Fun-V1.1-14B-InP
│   └── 📂 Wan2.1-Fun-V1.1-1.3B-InP/
├── 📂 Personalized_Model/
│   └── your trained trainformer model / your trained lora model (for UI load)

如何使用#

1. 生成#
a、显存节省方案#

由于Wan2.2的参数非常大,我们需要考虑显存节省方案,以节省显存适应消费级显卡。我们给每个预测文件都提供了GPU_memory_mode,可以在model_cpu_offload,model_cpu_offload_and_qfloat8,sequential_cpu_offload中进行选择。该方案同样适用于CogVideoX-Fun的生成。

  • model_cpu_offload代表整个模型在使用后会进入cpu,可以节省部分显存。
  • model_cpu_offload_and_qfloat8代表整个模型在使用后会进入cpu,并且对transformer模型进行了float8的量化,可以节省更多的显存。
  • sequential_cpu_offload代表模型的每一层在使用后会进入cpu,速度较慢,节省大量显存。

qfloat8会部分降低模型的性能,但可以节省更多的显存。如果显存足够,推荐使用model_cpu_offload。

b、通过comfyui#

具体查看ComfyUI README。

c、运行python文件#
  • 步骤1:下载对应权重放入models文件夹。
  • 步骤2:根据不同的权重与预测目标使用不同的文件进行预测。当前该库支持CogVideoX-Fun、Wan2.1、Wan2.1-Fun、Wan2.2,在examples文件夹下用文件夹名以区分,不同模型支持的功能不同,请视具体情况予以区分。以CogVideoX-Fun为例。
    • 文生视频:
      • 使用examples/cogvideox_fun/predict_t2v.py文件中修改prompt、neg_prompt、guidance_scale和seed。
      • 而后运行examples/cogvideox_fun/predict_t2v.py文件,等待生成结果,结果保存在samples/cogvideox-fun-videos文件夹中。
    • 图生视频:
      • 使用examples/cogvideox_fun/predict_i2v.py文件中修改validation_image_start、validation_image_end、prompt、neg_prompt、guidance_scale和seed。
      • validation_image_start是视频的开始图片,validation_image_end是视频的结尾图片。
      • 而后运行examples/cogvideox_fun/predict_i2v.py文件,等待生成结果,结果保存在samples/cogvideox-fun-videos_i2v文件夹中。
    • 视频生视频:
      • 使用examples/cogvideox_fun/predict_v2v.py文件中修改validation_video、validation_image_end、prompt、neg_prompt、guidance_scale和seed。
      • validation_video是视频生视频的参考视频。您可以使用以下视频运行演示:演示视频
      • 而后运行examples/cogvideox_fun/predict_v2v.py文件,等待生成结果,结果保存在samples/cogvideox-fun-videos_v2v文件夹中。
    • 普通控制生视频(Canny、Pose、Depth等):
      • 使用examples/cogvideox_fun/predict_v2v_control.py文件中修改control_video、validation_image_end、prompt、neg_prompt、guidance_scale和seed。
      • control_video是控制生视频的控制视频,是使用Canny、Pose、Depth等算子提取后的视频。您可以使用以下视频运行演示:演示视频
      • 而后运行examples/cogvideox_fun/predict_v2v_control.py文件,等待生成结果,结果保存在samples/cogvideox-fun-videos_v2v_control文件夹中。
  • 步骤3:如果想结合自己训练的其他backbone与Lora,则看情况修改examples/{model_name}/predict_t2v.py中的examples/{model_name}/predict_i2v.py和lora_path。
d、通过ui界面#

webui支持文生视频、图生视频、视频生视频和普通控制生视频(Canny、Pose、Depth等)。在examples文件夹下用文件夹名以区分,不同模型支持的功能不同,请视具体情况予以区分。以CogVideoX-Fun为例。

  • 步骤1:下载对应权重放入models文件夹。
  • 步骤2:运行examples/cogvideox_fun/app.py文件,进入gradio页面。
  • 步骤3:根据页面选择生成模型,填入prompt、neg_prompt、guidance_scale和seed等,点击生成,等待生成结果,结果保存在sample文件夹中。

参考文献#

许可证#

本项目采用 Apache License (Version 2.0).

Files

12 files · 24.2 GBWeights 10.0 GB (safetensors). Sizes and sha256 from the signed manifest.
Hide all 12 files with their sha256
FileSizesha256
.gitattributes2.31 KB1d1299516e6b5f5b82112358f59ce663be86424d20f32de3c4d69886bf1c0fda
README.md17.2 KB5912c1db6ce9a591b948aee5d09dc1183a21c9707335d7c2776d102fe48a8d85
README_en.md18.0 KB68e2b3d85a60f0ac2c1971d6fac975941dc7105fcf24e8588da156ebdb50329a
Wan2.2_VAE.pth2.82 GB20eb789667fa5e60e7516bf509512f6cb61f01b0aa0695eadaea930c13892b36
config.json301 Bdc20f8568e6b08121aa8e388c8cddba2ed42e9e5e57c7d75fbb6bf3b771cd018
configuration.json47 B5d1cc0f912402522f05f12a681278da402004ad37006c72ecbb461c50065ec6c
diffusion_pytorch_model.safetensors10.0 GBace4718a7c87ee3e5606a68ab79142c4395e81aece76b8120bc886f0fbbe1d16
google/umt5-xxl/special_tokens_map.json6.62 KB7b8a9f5040adb67b5805abdfd42c1f8d0f3d0e711f10726580eb3789cd0ad61d
google/umt5-xxl/spiece.model4.55 MBe3909a67b780650b35cf529ac782ad2b6b26e6d1f849d3fbb6a872905f452458
google/umt5-xxl/tokenizer.json16.8 MB6e197b4d3dbd71da14b4eb255f4fa91c9c1f2068b20a2de2472967ca3d22602b
google/umt5-xxl/tokenizer_config.json61.7 KBed9a3a8b0faa71a70a32847e0435fe036e6e112d4df4edb7bb48a921e344dc05
models_t5_umt5-xxl-enc-bf16.pth11.4 GB7cace0da2b446bbbbc57d031ab6cf163a3d59b366da94e5afe36745b746fd81d
VersionsEach time the pinned upstream commit or file list changed in this repo. Newest first.
Upstream commitDateContents
b8bc1a65ab pinned now2026-11-15 02:37 UTC12 files · 24.2 GB

Use it

$ fw get alibaba-pai/Wan2.2-Fun-5B-Control
# manifest sha256 61e6c9c0dbc0… · 24.2 GB · 12 files
$ fw verify ./Wan2.2-Fun-5B-Control alibaba-pai/Wan2.2-Fun-5B-Control
fw.pyChecks the signature and every sha256, tries each source in order, resumes.
$ python3 fw.py get https://freeweights.allternit.com/manifests/alibaba-pai/Wan2.2-Fun-5B-Control.json --dir ./Wan2.2-Fun-5B-Control $ python3 fw.py verify ./Wan2.2-Fun-5B-Control ./alibaba-pai__Wan2.2-Fun-5B-Control.json
hfThe Hugging Face CLI with one setting. Same repo id.
$ HF_ENDPOINT=https://freeweights.allternit.com hf download alibaba-pai/Wan2.2-Fun-5B-Control --revision b8bc1a65ab71d054ba4636dc0dac104aa4df2686 --local-dir ./Wan2.2-Fun-5B-Control
Packaged for ComfyUI. Most image, video and music models run in ComfyUI . Check the publisher's page for the workflow. Freeweights does not host it.

License

Name
Apache License 2.0
Id
Apache-2.0
Class
Permissive
Redistribution
yes
Listing
Hosted (signed)
Decided
License as declared by the publisher. No person read it. 2026-11-15
Text in
README.md
Text sha256
5912c1db6ce9a591b948aee5d09dc1183a21c9707335d7c2776d102fe48a8d85
Terms gate
No extra terms beyond the license.

Read the license text

Licenses are checked against the license table. A custom license passes only after a person reads it.

Provenance

From: alibaba-pai <huggingface.co/alibaba-pai>Newsgroups: alt.freeweights.videoSubject: Wan2.2-Fun-5B-ControlDate: 2026-11-15Commit: b8bc1a6Signed: 2DC28CC7B26A283B ✓
Upstream
alibaba-pai/Wan2.2-Fun-5B-Control
Pinned commit
b8bc1a65ab71d054ba4636dc0dac104aa4df2686
Manifest
Wan2.2-Fun-5B-Control.json · .minisig
Manifest sha256
61e6c9c0dbc0570caeef1de37854eb8b295b03027dfb62917c59ffe4a50986b0
Signature
minisign, Ed25519, key 2DC28CC7B26A283B
Created
2026-11-15 02:37 UTC
Generator
freeweights-manifest/0.1.0
Takedowns
None.

How to verify

$ curl -O https://freeweights.allternit.com/manifests/alibaba-pai/Wan2.2-Fun-5B-Control.json $ curl -O https://freeweights.allternit.com/manifests/alibaba-pai/Wan2.2-Fun-5B-Control.json.minisig $ minisign -Vm Wan2.2-Fun-5B-Control.json -P RWQ7KGqyx4zCLXtUFOqvfv8L3g+8fpKJ1FsuxPyOYMF+4KufbUxKq1Oq $ shasum -a 256 <file> # compare with the Files tab

A manifest whose signature fails must be rejected. A file whose size or sha256 does not match must be discarded. Verify a download

Model tree

Parent

Wan-AI/Wan2.2-TI2V-5B · this model is a finetune of it

No model in the catalog names this one as its parent.

Get it
Download
$ HF_ENDPOINT=https://freeweights.allternit.com hf download alibaba-pai/Wan2.2-Fun-5B-Control
Open in Freeweights Studio

Pull, verify and run it in the Allternit app on Desktop, web or phone.

Signed manifest
minisign, key 2DC28CC7B26A283B. Wan2.2-Fun-5B-Control.json. Provenance. Takedown history: none.
From: Comfy-Org <huggingface.co/Comfy-Org>Subject: Bernini-R
From: Comfy-Org <huggingface.co/Comfy-Org>Subject: Causal-Forcing-Framewise
From: Comfy-Org <huggingface.co/Comfy-Org>Subject: HuMo
From: bytedance-research <huggingface.co/bytedance-research>Subject: HuMo
From: genmo <huggingface.co/genmo>Subject: mochi-1-preview
From: Comfy-Org <huggingface.co/Comfy-Org>Subject: Mochi-Preview