I'd suggest you read the Outfit.ai PDF file for every single detail about this project.
This section assumes that the reader has downloaded the code/cloned this repository. This application requires Redis to serve as a message broker, and this demands the need for a Subsystem for Linux.
For windows users, WSL (Windows Subsystem for Linux) must be enabled and installed. For macOS users, Lima (Linux-on-mac) must be enabled and installed.
Due to file size limitations by Github - I have hosted the 2 ML models on Google Drive (https://drive.google.com/drive/folders/1LziscsJhf4k8WyVr89xSbSCmNjAEuKnu?usp=sharing), you should download both of the models (fashion-classifier.pth and recommender-model.pth) and place it under outfitai/ml/models/ .
- Open a terminal and navigate to the root folder of the application.
- Create a virtual envrionment and run 'pip install -r requirements.txt'
- Change your current directory to /env/Scripts and run activate
- Change your current directory to the root folder of the application and run flask run
- Open Ubuntu and run sudo apt install redis-server
- In the same Ubuntu terminal, run sudo service redis-server start
- Open a second terminal and navigate to the root folder of the application.
- Change your current directory to /env/Scripts and run activate
- Change your current directory to the root folder of the application and run celery -A outfitai.celery worker --loglevel=INFO --concurrency 1 -P solo
- Go to localhost:3000 in your browser