-
Notifications
You must be signed in to change notification settings - Fork 3
Home
A project is designed to detect wildfires from acceptable content. The project is presenting a containerized solution. The projects container are api and fireML. The api container runs FastAPI and is the communication between a user and the machine learning model. Container fireML contains the machine learning model for this project. The api container have an encrypted communication with fireML to feed the model data and receive the prediction of wildfire is detected.
The branch strategy will be four topic branches:
- api
- fireML
- deployment
- research
The purpose of these branches are to separate workflow for feature-branches where a new feature is merge into its relevant topic branches. This is to avoid conflicts with main branch and the sheer complex of too many feature branches and not knowing which branch to merge into get the correct feature for a current deployment branch. Since the topic branches is not the main branch, direct commits are allowed to topics.
The api topic branch serves the purpose of maintaining and developing the codebase of wildfires api. The api topic does not handle the its docker container. That is the purpose of deployment to work on.
Branch fireML is the maintaining and developing of the machine learning model for detecting wildfires. Container and branch fireML does not handle the training of the model but instead the functionality.
Deployment handles deployment of docker containers and github workflows. The deployment maintains the running of docker containers of api and fireML
The research branch is the focus of furthering the research of various models for wildfire detection. The research branch train new or continuing the machine learning model. It will merge any new model into fireML branch to apply.
Github project directory structure:
wildfire-ai
|
\-- api/
|
\-- fireML/
|
\-- utilities/
|
\-- research/
|
\-- <subject>/