Skip to content

📹 A more flexible CogVideoX that can generate videos at any resolution and creates videos from images.

License

Notifications You must be signed in to change notification settings

aigc-apps/CogVideoX-Fun

Repository files navigation

CogVideoX-Fun

😊 Welcome!

Hugging Face Spaces

English | 简体中文

Table of Contents

Introduction

CogVideoX-Fun is a modified pipeline based on the CogVideoX structure, designed to provide more flexibility in generation. It can be used to create AI images and videos, as well as to train baseline models and Lora models for Diffusion Transformer. We support predictions directly from the already trained CogVideoX-Fun model, allowing the generation of videos at different resolutions, approximately 6 seconds long with 8 fps (1 to 49 frames). Users can also train their own baseline models and Lora models to achieve certain style transformations.

We will support quick pull-ups from different platforms, refer to Quick Start.

What's New:

  • CogVideoX-Fun Control is now supported in diffusers. Thanks to a-r-r-o-w who contributed the support in this PR. Check out the docs to know more. [ 2024.10.16 ]
  • Retrain the i2v model and add noise to increase the motion amplitude of the video. Upload the control model training code and control model. [ 2024.09.29 ]
  • Create code! Now supporting Windows and Linux. Supports 2b and 5b models. Supports video generation at any resolution from 256x256x49 to 1024x1024x49. [ 2024.09.18 ]

Function:

Our UI interface is as follows: ui

Quick Start

1. Cloud usage: AliyunDSW/Docker

a. From AliyunDSW

DSW has free GPU time, which can be applied once by a user and is valid for 3 months after applying.

Aliyun provide free GPU time in Freetier, get it and use in Aliyun PAI-DSW to start CogVideoX-Fun within 5min!

DSW Notebook

b. From ComfyUI

Our ComfyUI is as follows, please refer to ComfyUI README for details. workflow graph

c. From docker

If you are using docker, please make sure that the graphics card driver and CUDA environment have been installed correctly in your machine.

Then execute the following commands in this way:

# 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/CogVideoX-Fun.git

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

# download weights
mkdir models/Diffusion_Transformer
mkdir models/Personalized_Model

wget https://pai-aigc-photog.oss-cn-hangzhou.aliyuncs.com/cogvideox_fun/Diffusion_Transformer/CogVideoX-Fun-V1.1-2b-InP.tar.gz -O models/Diffusion_Transformer/CogVideoX-Fun-V1.1-2b-InP.tar.gz

cd models/Diffusion_Transformer/
tar -xvf CogVideoX-Fun-V1.1-2b-InP.tar.gz
cd ../../

2. Local install: Environment Check/Downloading/Installation

a. Environment Check

We have verified CogVideoX-Fun execution on the following environment:

The detailed of Windows:

  • OS: 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

The detailed of Linux:

  • OS: 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

We need about 60GB available on disk (for saving weights), please check!

b. Weights

We'd better place the weights along the specified path:

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

Video Result

The results displayed are all based on image.

CogVideoX-Fun-V1.1-5B

Resolution-1024

00000005.mp4
00000006.mp4
00000009.mp4
00000010.mp4

Resolution-768

00000001.mp4
00000002.mp4
00000005.mp4
00000006.mp4

Resolution-512

00000036.mp4
00000035.mp4
00000034.mp4
00000033.mp4

CogVideoX-Fun-V1.1-5B-Pose

Resolution-512 Resolution-768 Resolution-1024
1.mp4
2.mp4
3.mp4

CogVideoX-Fun-V1.1-2B

Resolution-768

00000008.mp4
00000010.mp4
00000020.mp4
00000030.mp4

CogVideoX-Fun-V1.1-2B-Pose

Resolution-512 Resolution-768 Resolution-1024
1.mp4
2.mp4
3.mp4

How to use

1. Inference

a. Using Python Code

  • Step 1: Download the corresponding weights and place them in the models folder.
  • Step 2: Modify prompt, neg_prompt, guidance_scale, and seed in the predict_t2v.py file.
  • Step 3: Run the predict_t2v.py file, wait for the generated results, and save the results in the samples/cogvideox-fun-videos-t2v folder.
  • Step 4: If you want to combine other backbones you have trained with Lora, modify the predict_t2v.py and Lora_path in predict_t2v.py depending on the situation.

b. Using webui

  • Step 1: Download the corresponding weights and place them in the models folder.
  • Step 2: Run the app.py file to enter the graph page.
  • Step 3: Select the generated model based on the page, fill in prompt, neg_prompt, guidance_scale, and seed, click on generate, wait for the generated result, and save the result in the samples folder.

c. From ComfyUI

Please refer to ComfyUI README for details.

2. Model Training

A complete CogVideoX-Fun training pipeline should include data preprocessing, and Video DiT training.

a. data preprocessing

We have provided a simple demo of training the Lora model through image data, which can be found in the wiki for details.

A complete data preprocessing link for long video segmentation, cleaning, and description can refer to README in the video captions section.

If you want to train a text to image and video generation model. You need to arrange the dataset in this format.

📦 project/
├── 📂 datasets/
│   ├── 📂 internal_datasets/
│       ├── 📂 train/
│       │   ├── 📄 00000001.mp4
│       │   ├── 📄 00000002.jpg
│       │   └── 📄 .....
│       └── 📄 json_of_internal_datasets.json

The json_of_internal_datasets.json is a standard JSON file. The file_path in the json can to be set as relative path, as shown in below:

[
    {
      "file_path": "train/00000001.mp4",
      "text": "A group of young men in suits and sunglasses are walking down a city street.",
      "type": "video"
    },
    {
      "file_path": "train/00000002.jpg",
      "text": "A group of young men in suits and sunglasses are walking down a city street.",
      "type": "image"
    },
    .....
]

You can also set the path as absolute path as follow:

[
    {
      "file_path": "/mnt/data/videos/00000001.mp4",
      "text": "A group of young men in suits and sunglasses are walking down a city street.",
      "type": "video"
    },
    {
      "file_path": "/mnt/data/train/00000001.jpg",
      "text": "A group of young men in suits and sunglasses are walking down a city street.",
      "type": "image"
    },
    .....
]

b. Video DiT training

If the data format is relative path during data preprocessing, please set scripts/train.sh as follow.

export DATASET_NAME="datasets/internal_datasets/"
export DATASET_META_NAME="datasets/internal_datasets/json_of_internal_datasets.json"

If the data format is absolute path during data preprocessing, please set scripts/train.sh as follow.

export DATASET_NAME=""
export DATASET_META_NAME="/mnt/data/json_of_internal_datasets.json"

Then, we run scripts/train.sh.

sh scripts/train.sh

For details on setting some parameters, please refer to Readme Train, Readme Lora and Readme Control.

Model zoo

V1.1:

名称 存储空间 Hugging Face Model Scope 描述
CogVideoX-Fun-V1.1-2b-InP.tar.gz Before extraction:9.7 GB / After extraction: 13.0 GB 🤗Link 😄Link Our official graph-generated video model is capable of predicting videos at multiple resolutions (512, 768, 1024, 1280) and has been trained on 49 frames at a rate of 8 frames per second. Noise has been added to the reference image, and the amplitude of motion is greater compared to V1.0.
CogVideoX-Fun-V1.1-5b-InP.tar.gz Before extraction:16.0 GB / After extraction: 20.0 GB 🤗Link 😄Link Our official graph-generated video model is capable of predicting videos at multiple resolutions (512, 768, 1024, 1280) and has been trained on 49 frames at a rate of 8 frames per second. Noise has been added to the reference image, and the amplitude of motion is greater compared to V1.0.
CogVideoX-Fun-V1.1-2b-Pose.tar.gz Before extraction:9.7 GB / After extraction: 13.0 GB 🤗Link 😄Link Our official pose-control video model is capable of predicting videos at multiple resolutions (512, 768, 1024, 1280) and has been trained on 49 frames at a rate of 8 frames per second.
CogVideoX-Fun-V1.1-5b-Pose.tar.gz Before extraction:16.0 GB / After extraction: 20.0 GB 🤗Link 😄Link Our official pose-control video model is capable of predicting videos at multiple resolutions (512, 768, 1024, 1280) and has been trained on 49 frames at a rate of 8 frames per second.

V1.0:

Name Storage Space Hugging Face Model Scope Description
CogVideoX-Fun-2b-InP.tar.gz Before extraction:9.7 GB / After extraction: 13.0 GB 🤗Link 😄Link Our official graph-generated video model is capable of predicting videos at multiple resolutions (512, 768, 1024, 1280) and has been trained on 49 frames at a rate of 8 frames per second.
CogVideoX-Fun-5b-InP.tar.gz Before extraction:16.0 GB / After extraction: 20.0 GB 🤗Link 😄Link Our official graph-generated video model is capable of predicting videos at multiple resolutions (512, 768, 1024, 1280) and has been trained on 49 frames at a rate of 8 frames per second.

TODO List

  • Support Chinese.

Reference

License

This project is licensed under the Apache License (Version 2.0).

The CogVideoX-2B model (including its corresponding Transformers module and VAE module) is released under the Apache 2.0 License.

The CogVideoX-5B model (Transformers module) is released under the CogVideoX LICENSE.

About

📹 A more flexible CogVideoX that can generate videos at any resolution and creates videos from images.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published