{"repo":"Atharva-Phatak/shopme","free":true,"listed":false,"github":"https://github.com/Atharva-Phatak/shopme","clone":"git clone https://github.com/Atharva-Phatak/shopme.git","description":"ShopME: An E2E fashion recommendation System","language":"Python","stars":20,"topics":["airflow","dagster","docker","gcp","machine-learning","mlops","pytorch","streamlit"],"license":"MIT","category":"machine-learning","readme_excerpt":"ShopME: An E2E fashion recommendation System Recommender systems have grown to be an essential part of all large Internet retailers, driving up to 35% of Amazon sales or over 80% of the content watched on Netflix. In general good recommender systems can help increase companies sales by majority. In this project we will be building an end to end fashion recommender system for H&M products. This project will help you to understand standard ML cycle and will teach you how to create an E2E ML system. Demo 🚀 Procedure The system consists of three major steps like any other ML system and we will be using various tools, so that at you can use this project as an MLOps template for any project 💙 Process Tools ----------- ----------- Data Collection BeautifulSoup, Airflow, MinIO/AWS-S3 Experimentation PyTorch, Mlflow, Dagster, scikit-learn, Onnx Backend Service BentoML, Docker Frontend Service Streamlit Deployment Service Google container Registry (GCR), Google Cloud Run Testing/Formatting, etc Github actions, pre-commit, pytest Below figure shows the overall steps carried out in the process. ---- Data Collection The data is collected by scraping the H&M website using BeautifulSoup and wrote a data orchestration pipeline using Airflow, so that our data collection jobs run according to a schedule. Basically I wanted my code to run on my local system when I was asleep :), plus Airflow provides a great way to automate your process. The key things to remember during this process is a dat","default_branch":null,"files":null,"tree":[],"storefront":"/r/Atharva-Phatak","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/Atharva-Phatak/shopme/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}