This is a work in progress repository to display data science and machine learning content in a public setting.
Currently prod has a prototype resume scaffolding with sections and buttons for the ASR chat-gpt plugin. I have almost figured out https for prod to give access to visitor's microphone. In dev I have the basics of ASR/chatgpt setup to start that speech-to-gpt service. I need to clean up some of the file communications as it is just working with a "smoke test" right now. Also I need to put the asr model (whisper) into lambda to avoid a huge EC2 bill.
Public Domain: robertreaney.com
- Configure AWS CLI as your credentials will be shared with the local services.
docker compose up --buildCtrl+Shift+Bin VSCode works as shortcut
- ssh keygen in
~/.sshssh-keygen -t rsa -b 4096 - terraform apply
ssh -i ~/.ssh/id_rsa [email protected] -o StrictHostKeyChecking=no- Set environment variables AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY
- run
ec2_startup.sh - startup the container with
sudo docker run -it -p 80:5000 container_image_name /bin/bash - inside the container start the web server
gunicorn --bind 0.0.0.0:5000 app:app - go to your devbox and goto ec2_public_ip:80
- spin up stack on EC2 instance in cert mode
docker compose -f docker-compose.yml -f docker-compose.cert.yml up --build -d --remove-orphans
- docker exec -it nginx-server /bin/bash
- apt-get update
- apt-get install certbot
- certbot certonly --webroot -w /static -d robertreaney.com -v --cert-name robertreaney.com
- /etc/letsencrypt/archive/robertreaney.com should have keys in it
- /etc/letsencrypt/live/robertreaney.comy has symbolic link to keys