Web server for image classification

Download the repository

Use git to clone the repository:

git clone https://github.com/unica-isde/flask-classification-2022-G

And install the requirements with

pip install -r requirements.txt

Additional requirements:

• Redis

  • for Windows ---> https://redis.io/docs/getting-started/installation/install-redis-on-windows/

  • for Linux ---> https://redis.io/docs/getting-started/installation/install-redis-on-linux/

Configuration

Configure the service by editing the file config.py.

Prepare the resources

It is recommended to pre-download images and models before running the server. This is to avoid unnecessary waits for users.

Run prepare_images.py and prepare_models.py. Models will be stored in your PyTorch cache directory, while the path for the image directory can be found in the config.py file.

Usage

Run locally

To run the code without containers, it is sufficient to run separately the runserver.py script, the worker.py script and the worker_histo.py script. The workers will process jobs stored in the queue. In order for the queue to work, you should have redis installed and running (specify port in config.py).