laitimes

AI videos are automatically generated for local deployment, which perfectly realizes the function of "clipping" graphics and texts

author:13 Demon Studios

Project Description

This project is for study and research only and may not be used for commercial purposes. All recommended open source projects have been tested and run by myself. If there is any infringement, please notify the takedown.

Project Brief Introduction

Just provide a video theme or keyword, and you can automatically generate video copy, video materials, video subtitles, and video background music, and then synthesize a high-definition short video.

  • Complete MVC architecture, code is clearly structured, easy to maintain, and supports APIs and web interfaces
  • Support video copywriting AI auto-generated, and you can also customize the copywriting
  • Supports a wide range of HD video sizes
    • 竖屏 9:16,1080x1920
    • 横屏 16:9,1920x1080
  • Batch video generation is supported, and multiple videos can be generated at once and then choose the one that is most satisfying
  • Video clip duration can be set to adjust the frequency of material switching
  • Support Chinese and English video copywriting
  • Supports a wide range of speech synthesis
  • Support subtitle generation, you can adjust the font, position, color, size, and subtitle stroke settings
  • Support background music, random or specified music files, and set the background music volume
  • The source of the video footage is HD, and it is copyright-free
  • It supports multiple models such as OpenAI, moonshot, Azure, gpt4free, one-api, Tongyi Qianwen, Google Gemini, and Ollama

Project address

https://github.com/harry0703/MoneyPrinterTurbo

Effect display

AI videos are automatically generated for local deployment, which perfectly realizes the function of "clipping" graphics and texts

The result is an MP4 video file

AI videos are automatically generated for local deployment, which perfectly realizes the function of "clipping" graphics and texts

Software preparation

  • ImageMagick https://www.cnblogs.com/sttchengfei/p/17945462
  • Large Model Choose Moonshot, Moonshot's text generation model (moonshot-v1) is trained to understand natural language and written language, and it can generate text output based on inputs. After registration, generate a key URL: https://platform.moonshot.cn/console/api-keys Friends with other large models can choose ChatGPT, Wenxin Yiyan, etc.
  • Image API URL: https://www.pexels.com/zh-cn/password/new/ Registration: pexels_api_keys
  • Run: git clone https://github.com/harry0703/MoneyPrinterTurbo.git clone code
  • Make a copy of the config.example.toml file and name it config.toml
  • Configure the pexels_api_keys and llm_provider as described in the config.toml file, and configure the API key according to the service provider corresponding to llm_provider
  • Modify the large model llm_provider="moonshot" and fill in the corresponding moonshot_api_key

We recommend that you use Conda to create a Python virtual environment

Install anaconda configuration environment variables

AI videos are automatically generated for local deployment, which perfectly realizes the function of "clipping" graphics and texts
cd MoneyPrinterTurbo
conda create -n MoneyPrinterTurbo python=3.10
activate MoneyPrinterTurbo
pip install -r requirements.txt
webui.bat           

After running successfully, a website will be opened, enter keywords, and generate "video copy" and video keywords

AI videos are automatically generated for local deployment, which perfectly realizes the function of "clipping" graphics and texts

Generate video keywords

AI videos are automatically generated for local deployment, which perfectly realizes the function of "clipping" graphics and texts

Generate videos

AI videos are automatically generated for local deployment, which perfectly realizes the function of "clipping" graphics and texts

The video generation is complete

AI videos are automatically generated for local deployment, which perfectly realizes the function of "clipping" graphics and texts

In general, the generated video is unsatisfactory, and those with strong hands-on ability can achieve more effects through secondary development, looking forward to your transformation!

Precautions

  • The pillow version needs to be installed with 9.5.0, otherwise an error will be reported
AI videos are automatically generated for local deployment, which perfectly realizes the function of "clipping" graphics and texts

Read on