https://github.com/testdrivenio/fastapi-ml

deploying an ML model to Heroku with FastAPI
https://github.com/testdrivenio/fastapi-ml

fastapi fastapi-docker fbprophet heroku-deployment machine-learning

Last synced: about 2 months ago
JSON representation

deploying an ML model to Heroku with FastAPI

• Host: GitHub
• URL: https://github.com/testdrivenio/fastapi-ml
• Owner: testdrivenio
• License: mit
• Created: 2020-07-02T17:53:59.000Z (over 4 years ago)
• Default Branch: main
• Last Pushed: 2024-04-17T01:03:41.000Z (8 months ago)
• Last Synced: 2024-08-02T14:12:21.775Z (5 months ago)
• Topics: fastapi, fastapi-docker, fbprophet, heroku-deployment, machine-learning
• Language: Python
• Homepage:
• Size: 785 KB
• Stars: 46
• Watchers: 4
• Forks: 22
• Open Issues: 2
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-starred-test - testdrivenio/fastapi-ml - deploying an ML model to Heroku with FastAPI (Python)

README

        
# Deploying and Hosting a Machine Learning Model with FastAPI and Heroku

## Want to learn how to build this?

Check out the [tutorial](https://testdriven.io/blog/fastapi-machine-learning).

## Want to use this project?

### With Docker

1. Build and tag the Docker image:

    ```sh

    $ docker build -t fastapi-prophet .

    ```

1. Spin up the container:

    ```sh

    $ docker run --name fastapi-ml -e PORT=8008 -p 8008:8008 -d fastapi-prophet:latest

    ```

1. Train the model:

    ```sh

    $ docker exec -it fastapi-ml python

    >>> from model import train, predict, convert

    >>> train()

    ```

1. Test:

    ```sh

    $ curl \

      --header "Content-Type: application/json" \

      --request POST \

      --data '{"ticker":"MSFT"}' \

      http://localhost:8008/predict

    ```

### Without Docker

1. Create and activate a virtual environment:

    ```sh

    $ python3 -m venv venv && source venv/bin/activate

    ```

1. Install the requirements:

    ```sh

    (venv)$ pip install -r requirements.txt

    ```

1. Train the model:

    ```sh

    (venv)$ python

    >>> from model import train, predict, convert

    >>> train()

    ```

1. Run the app:

    ```sh

    (venv)$ uvicorn main:app --reload --workers 1 --host 0.0.0.0 --port 8008

    ```

1. Test:

    ```sh

    $ curl \

      --header "Content-Type: application/json" \

      --request POST \

      --data '{"ticker":"MSFT"}' \

      http://localhost:8008/predict

    ```