MoneyPrinterTurbo
v1.2.1
Due to the deployment and use of this project, there is still a certain threshold for some novice users. I would like to express my special thanks to Luka (AI Intelligent Multimedia Service Platform) website for providing free AI视频生成器services based on this project. No deployment required, just use it online, which is very convenient.
Thanks to Zuotang https://picwish.cn for supporting and sponsoring this project, which enables the project to be continuously updated and maintained.
Zuotang focuses on the field of image processing and provides a wealth of image processing tools to extremely simplify complex operations and truly make image processing simpler.
API and Web界面1080x1920 1920x1080 字体,位置,颜色,大小, and supports字幕描边settings. 背景音乐音量can be set 
More realistic synthesized sounds | ||
|---|---|---|
demo-portrait-1.mp4 | default.mp4 | demo-portrait-2.mp4 |
demo-landscape.mp4 | demo-landscape-2.mp4 |
Download the one-click startup package, unzip it and use it directly (the path should not contain Chinese characters, special characters , or spaces )
After downloading, it is recommended to double-click update.bat to update to the latest code , and then double-click start.bat to start.
After startup, the browser will automatically open (if it opens blank, it is recommended to open it in Chrome or Edge )
The one-click startup package has not been created yet. See the installation and deployment section below. It is recommended to use docker deployment, which is more convenient.
全局流量mode git clone https://github.com/harry0703/MoneyPrinterTurbo.gitconfig.example.toml file and name it config.tomlconfig.toml file to configure pexels_api_keys and llm_provider , and configure the relevant API Key according to the service provider corresponding to llm_provider.If Docker is not installed, please install it first https://www.docker.com/products/docker-desktop/
If it is a Windows system, please refer to Microsoft's documentation:
cd MoneyPrinterTurbo
docker-compose upOpen the browser and visit http://0.0.0.0:8501
Open the browser and visit http://0.0.0.0:8080/docs or http://0.0.0.0:8080/redoc
Video tutorial
It is recommended to use conda to create a python virtual environment
git clone https://github.com/harry0703/MoneyPrinterTurbo.git
cd MoneyPrinterTurbo
conda create -n MoneyPrinterTurbo python=3.10
conda activate MoneyPrinterTurbo
pip install -r requirements.txtWindows:
imagemagick_path in配置文件config.toml to your actual installation pathMacOS:
brew install imagemagickUbuntu
sudo apt-get install imagemagickCentOS
sudo yum install ImageMagick Note that you need to execute the following command in根目录of the MoneyPrinterTurbo project
conda activate MoneyPrinterTurbo
webui.batconda activate MoneyPrinterTurbo
sh webui.shAfter startup, the browser will automatically open (if it opens blank, it is recommended to open it in Chrome or Edge )
python main.py After startup, you can view API文档http://127.0.0.1:8080/docs or http://127.0.0.1:8080/redoc to directly debug the interface online for a quick experience.
For a list of all supported sounds, you can view: Sound List
2024-04-16 v1.1.2 adds 9 new Azure speech synthesis sounds, which need to be configured with API KEY. The synthesized sounds are more realistic.
Currently, 2 subtitle generation methods are supported:
速度快generation, better performance, no computer configuration requirements, but the quality may be unstable速度慢, the performance is poor, and it requires certain computer configuration, but质量更可靠. You can modify subtitle_provider in the config.toml configuration file to switch
It is recommended to use edge mode. If the quality of the generated subtitles is not good, switch to whisper mode.
Notice:
Since HuggingFace is not accessible in China, you can use the following method to download the model file of
whisper-large-v3
Download address:
After downloading the model, unzip it and place the entire directory in .MoneyPrinterTurbomodels . The final file path should be like this: .MoneyPrinterTurbomodelswhisper-large-v3
MoneyPrinterTurbo
├─models
│ └─whisper-large-v3
│ config.json
│ model.bin
│ preprocessor_config.json
│ tokenizer.json
│ vocabulary.json
The background music used for videos is located in the resource/songs directory of the project.
The current project contains some default music from YouTube videos. If there is any infringement, please delete it.
It is used for rendering video subtitles and is located in the resource/fonts directory of the project. You can also put your own fonts in it.
OpenAI announced that ChatGPT 3.5 is now free. Some developers have encapsulated it into an API that can be called directly.
Make sure you have installed and started the docker service , execute the following command to start the docker service
docker run -p 3040:3040 missuo/freegpt35 After successful startup, modify the configuration in config.toml
llm_provider is set to openaiopenai_api_key , such as '123456'openai_base_url changed http://localhost:3040/v1/openai_model_name changed to gpt-3.5-turboNote: This method is less stable
This issue is caused by large models not returning correct responses.
There is a high probability that it is caused by the network. Using a VPN or setting openai_base_url as your proxy should solve the problem.
It is also recommended to use Moonshot or DeepSeek as a large model provider. These two service providers have faster and more stable access in China.
Normally, ffmpeg is automatically downloaded and automatically detected. But if there is a problem with your environment and automatic downloading cannot occur, you may encounter the following error:
RuntimeError: No ffmpeg exe could be found.
Install ffmpeg on your system, or set the IMAGEIO_FFMPEG_EXE environment variable.
At this time, you can download ffmpeg from https://www.gyan.dev/ffmpeg/builds/. After decompressing, set ffmpeg_path to your actual installation path.
[ app ]
# 请根据你的实际路径设置，注意 Windows 路径分隔符为 \
ffmpeg_path = " C: \ Users \ harry \ Downloads \ ffmpeg.exe " These policies can be found in ImageMagick's configuration file policy.xml. This file is usually located in /etc/ImageMagick- X / or a similar location in the ImageMagick installation directory. Modify the entry containing pattern="@" and change rights="none" to rights="read|write" to allow reading and writing of the file.
This problem is caused by the system's limit on the number of open files and can be solved by modifying the system's limit on the number of open files.
View current limits
ulimit -nIf it is too low, you can adjust it higher, for example
ulimit -n 10240LocalEntryNotfoundEror: Cannot find an appropriate cached snapshotfolder for the specified revision on the local disk and outgoing trafic has been disabled. To enablerepo look-ups and downloads online, pass 'local files only=False' as input.
or
An error occurred while synchronizing the model Systran/faster-whisper-large-v3 from the Hugging Face Hub: An error happened while trying to locate the files on the Hub and we cannot find the appropriate snapshot folder for the specified revision on the local disk . Please check your internet connection and try again. Trying to load the model directly from the local cache, if it exists.
Solution: Click to see how to manually download the model from the network disk
This project is refactored based on https://github.com/FujiwaraChoki/MoneyPrinter. It has made a lot of optimizations and added more functions. Thanks to the original author for his open source spirit.
Click to view LICENSE file
