Install the following in a new virtual env
flask
langchain
openai
unstructured[local-inference]
tiktoken
faiss-gpu (or cpu based on your server)
Export your API Key on your instance
export OPENAI_API_KEY=your_api_key
RUN the app from root folder
python app.py
To install Docker
curl -fsSL https://get.docker.com -o get-docker.sh
sudo sh get-docker.sh
Now we have Docker installed, we are ready to build and depoy the app.
Build Docker APP:
docker build -t my_app . #Replace my_app with any app name you want
Now to deploy the app on :, use the following command:
docker run -p 5010:5010 my_app # Make sure to replace the name of the app to your name choosen earlier. Remember i have choosen 5010 (make sure you have this open)
To keep running in background run the following command
docker run -d -p 5010:5010 -e my_app # Make sure to replace the name of the app to your name choosen earlier. Remember i have choosen 5010 (make sure you have this open)
To view docker app, Run :
docker ps
It will show the docker container running
To stop the container, Run:
docker stop <container_id>
To view the logs of the Docker Container, Run:
docker logs <container_id>
To view the files inside a running Docker container, you can use the docker exec command to start a shell session within the container. Run:
docker exec -it <container_id> bash
Make sure bash is installed
To exit a running Docker container, simply press Ctrl + D or type exit and press Enter. This will terminate the current shell session within the container and return you to your host system's command prompt.
To run nano to view any file content like app.py, install nano first in the Docker container, RUN:
docker exec -it <container_id> /bin/bash -c "apt-get update && apt-get install -y nano"
Then run,
docker exec -it <container_id> nano <file_path>