Video editing with GPT-3
This is en experimental home project, and is for testing or entertainment purposes only. It may contain bugs
Tested on Linux Ubuntu, that requires ImageMagick to be installed, so:
sudo apt update && sudo apt install imagemagick
Then you have to uncomment or remove the following line in the /etc/ImageMagick-6/policy.xml file:
<policy domain="path" rights="none" pattern="@*"/>If you don't want to remove it, you can just uncomment in this way:
<!-- <policy domain="path" rights="none" pattern="@*"/> -->Save.
Then:
git clone https://github.com/0ut0flin3/VideoGPT.git
cd VideoGPT
pip install -r requirements.txt
Add your OpenAI API Key in the .env file, then:
python3 main.py /path/to/yourvideo.mp4
Replace the path with the path of the video that you want to edit, then type the action you want to perform on the video:
user@linux:~VideoGPT$ python3 main.py ./test.mp4
What you would like to do with this video?
> Add the text "LOL" at second 5 with a duration of 3 seconds
Moviepy - Building video ./outputs/test.mp4.
MoviePy - Writing audio in testTEMP_MPY_wvf_snd.mp3
MoviePy - Done.
Moviepy - Writing video ./outputs/test.mp4
Moviepy - Done !
Moviepy - video ready ./outputs/test.mp4
if you're lucky, you'll find your edited video in the outputs folder. Original video will not be touched.
Often runs into bugs and syntax errors, will be fixed. Pull requests for improvements are welcome