kostas2370/Video-Creator

This project is to automate the video creation.

Frontend repository : https://github.com/kostas2370/ViddieDEMO

Samples : https://www.youtube.com/watch?v=PvrX_jq4fv4

https://www.youtube.com/watch?v=bNZvK68O-Rk

How to run it :

There are 2 ways:

Manually :

Recommended python version : 3.9 - 3.11

First install these 2 programs in your computer :

1. Install FFmpeg in your computer (Check this guide : https://phoenixnap.com/kb/ffmpeg-windows) (For video Rendering)
2. Install ImageMagick (https://imagemagick.org/script/download.php#windows) (For subtitles)
3. For some voice model you will need to install the MSI from here (https://github.com/espeak-ng/espeak-ng/releases)

If you dont want to install any extra packages for voices , use only : tts_models/en/ljspeech/tacotron2-DDC Model

Go to the viddie folder and run this command : pip install -r requirements/requirements.txt

Create a checkpoints folder and Download these files and add them in the checkpoints folder https://drive.google.com/drive/u/1/folders/1Fp4sjMi6U3bQaKmQQe04qeXzk7quu0Od

Inside the viddie folder : Create the .env file and add your OPEN_API_KEY, SEARCH_ENGINE_ID, API_KEY

To find your google search_engine_id and api_key , check this guide : https://www.youtube.com/watch?v=D4tWHX2nCzQ&t=127s

After go to the viddie folder and run those commands:

py manage.py makemigrations
py manage.py migrate
py manage.py loaddata fixtures/fixtures.json
py manage.py setup_media
py manage.py runserver

If you want to add voices from elevenlabs(your voices or the premade ones) just add your XI_API_KEY in .env file and run this command :

py manage.py setup_elevenlabs

Docker :

1. Create the .env file and add your OPEN_API_KEY, SEARCH_ENGINE_ID, API_KEY like .env_example
2. Download these files and add them in the checkpoints folder https://drive.google.com/drive/u/1/folders/1Fp4sjMi6U3bQaKmQQe04qeXzk7quu0Od
3. docker-compose up --build in the src folder

Login credentials for admin page : username = admin, password = pass

You can find all the endpoints in swagger : http://localhost:8000/swagger/

• TO DO :

1. Convert all moviepy functions to ffmpeg for better Performance .
2. Add celery support for async tasks and scheduled tasks
3. Test Cases for models, functions and Views

For any Injuries or Support you can contact me at those emails :

1. kodamia@cs.ihu.gr (University Email)
2. kostas2372@gmail.com (Personal Email)

NEWS:

• Fix of the conquis TTS docker issue
• Gemini and claude support !
• Implemented ElevenLabs api voices !!
• Now we are able to make compilation videos from twitch by game or streamer
• Added open-ai voices
• Added midjourney and special diffusion as image providers !! (For now you can change the providers in the mapper.py in default providers)
• Dockerized the app