Skip to content

FastAPI service for Amazon forests satellite images multi-label classification

Notifications You must be signed in to change notification settings

DimYun/amazon-forests-satellite-class_service

Repository files navigation

Type of Amazon forests satellite image. Service

FastAPI service for Amazon forests satellite image NN multilabel classification. I made accent on "industrial quality" code with next technologies:

  • FastAPI
  • Gitlab CI/CD (test, deploy, destroy)
  • DVC
  • Docker
  • Unit & Integration tests with coverage report
  • Linters (flake8 + wemake)

Disclaimers:

  • the project was originally crated and maintained in GitLab local instance, some repo functionality may be unavailable
  • the project was created by me and me only as part of the CVRocket professional development course
  • this project is my first "industry grade" service for NN, for more advanced code and features please see car-plate projects

Location for manual test:

Setup of environment

First, create and activate venv: bash python3 -m venv venv . venv/bin/activate

Next, install dependencies: bash make install

Commands

Preparation

  • make install - install python dependencies

Run service

  • make run_app - run servie. You can define argument APP_PORT

Build docker

  • make build - you can define arguments DOCKER_TAG, DOCKER_IMAGE

Static analyse

  • make lint - run linters

Tests

  • make run_unit_tests - run unit tests
  • make run_integration_tests - run integration tests
  • make run_all_tests - run all tests

About

FastAPI service for Amazon forests satellite images multi-label classification

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published