E-Commerce-Recommender is a recommendation system that aims to help users find similar items to the ones they bought from a retail store.
It does this using two components. A CLIP and Dot NN component. The CLIP part uses pretrained embeddings from the CLIP Text Model and the Dot NN part uses trained-from-scratch embeddings.
- Python
python -m pip install -r requirements.txt
- FastAPI
uvicorn main:app
│ main.py
│ train.ipynb
│ utils.py
|
├── variables
│ ├──all_items.pt
│ ├──df.pt
| └──weights.pt
|
└── images
└─── model_architecture.png
Running the code mentioned above should launch a server in your localhost with some port. Head over to "bought" page and supply an index. Example:
localhost:8000/bought/1234
Dahir Ibrahim (Deedax Inc) - http://instagram.com/deedax_inc
Email - [email protected]
YouTube - https://www.youtube.com/@deedaxinc.3389
Twitter - https://twitter.com/DeedaxInc
Project Link - https://github.com/Daheer/E-Commerce-Recommender