- You must have docker installed on your computer (https://www.docker.com/get-started/)
- create
generate/.env
file with the following values:
JORDAN_PETERSON_VOICE_ID=jordan-peterson
JOE_ROGAN_VOICE_ID=joe-rogan
BARACK_OBAMA_VOICE_ID=barack-obama
KAMALA_HARRIS_VOICE_ID=kamala-harris
BEN_SHAPIRO_VOICE_ID=ben-shapiro
ANDREW_TATE_VOICE_ID=andrew-tate
JOE_BIDEN_VOICE_ID=joe-biden
DONALD_TRUMP_VOICE_ID=donald-trump
GROQ_API_KEY=YOUR GROQ API KEY HERE
OPENAI_API_KEY=YOUR OPEN AI API KEY HERE
NEETS_API_KEY=YOUR NEETS API KEY HERE
1.5 Note, you should get the actual values for your GROQ, OPENAI, and NEETS api keys before proceeding (scroll down for links on where to get each)
- go into generate (
cd generate
) and rundocker build -t brainrot .
. This will take 10-15 minutes, as there are a lot of dependencies. - now, once this docker image is successfully built, we need to run it as a container. Run this command:
docker run -d --name brainrot brainrot \
-w 1 \
-b 0.0.0.0:5000 \
--access-logfile access.log \
--error-logfile error.log \
--chdir /app/brainrot \
transcribe:app \
--timeout 120
- now run
docker exec -it brainrot /bin/bash
, followed bynode localBuild.mjs
- when the video has been generated, exit out of the container (
cntl+d
in terminal window), and then rundocker cp brainrot:/app/brainrot/out/video.mp4 ./video.mp4
. This will output where the video is located on your computer (e.g.Successfully copied 97.8MB to /home/noahsolomon/brainrotjs/generate/video.mp4
). Voila you just generated brainrot. - change the variable values at the top in localBuild.mjs to change what vidoe is generated. The video generation process can take 10-20 minutes so be patient! we are so back fam
I have removed assets for download except MINECRAFT-0.mp4 (in generate/public/background/). If you want your own GTA / Minecraft / etc. bottom half video just find some on youtube. and add the videos to generate/public/background/ folder.
- Dalle 3 API rate limit exceeded: this is because each dialogue transition has an image, and it is prompted to have 7 dialogue transitions. However, typical tier 1 open ai accounts can only generate 5 images per minute. You might need to reduce the # of dialog transitions if this is the case (in generate/transcript.mjs)
- You don't have enough storage (the image will be around 12.6GB)