https://github.com/asadhanif3188/mlops-project-02-end-to-end-data-science-project-with-deployment
https://github.com/asadhanif3188/mlops-project-02-end-to-end-data-science-project-with-deployment
Last synced: about 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/asadhanif3188/mlops-project-02-end-to-end-data-science-project-with-deployment
- Owner: asadhanif3188
- Created: 2024-11-24T03:32:17.000Z (7 months ago)
- Default Branch: main
- Last Pushed: 2024-11-26T11:26:27.000Z (7 months ago)
- Last Synced: 2025-04-15T13:57:34.300Z (about 2 months ago)
- Language: Jupyter Notebook
- Size: 741 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# MLOps Project 02 -- End-to-End Data Science Project with Deployment
***Note:*** *This project showcases a comprehensive skill set for implementing an end-to-end data science or machine learning solution.*
## Problem Statement
The quality of wine is influenced by various chemical properties, making it essential to establish an accurate and efficient method for predicting wine quality. The objective of this project is to build a machine learning model that predicts the quality of wine based on its physicochemical attributes.
Using the [Wine Quality Dataset](https://www.kaggle.com/datasets/yasserh/wine-quality-dataset), this project focuses on a classification problem where the goal is to assign a quality score (from a predefined range) to a wine sample. By leveraging data preprocessing, feature engineering, and model training techniques, the solution aims to assist wine producers and quality control teams in automating the evaluation process, ensuring consistency and efficiency in quality assessment.
## Few Screenshots
Running ML Application as a Flask Application.
Hiting the training endpoint of Flask app, `127.0.0.1:8080/train`
Model's experiment tracking UI at DagsHub.
Model's experiment tracking MLFlow UI.
Model's parameters and metrics
Registered model at MLFlow
Basic UI for data input for prediction.
Predicted result demo.
----------------------------
## Key Features
This project highlights the following key features:
- **Data Ingestion and Preprocessing:**
Automated pipelines for data collection, validation, cleaning, and transformation.
- **Modular Programming:**
Well-structured, reusable, and maintainable code components for scalability and ease of debugging.
- **Standardized Project Structure:**
A `template.py` script has been developed to automate the creation of a consistent and production-ready project directory and file structure.
- **Model Training and Evaluation:**
Efficient workflows for model training, hyperparameter tuning, and performance evaluation using robust metrics.
- **Version Control:**
Data and model versioning using DVC or other tools to ensure reproducibility and traceability.
- **Model Deployment:**
Seamless deployment of machine learning models using FastAPI, Flask, or Kubernetes, ensuring high availability and scalability.
## Workflows -- ML Pipeline
1. Data Ingestion
2. Data Validation
3. Data Transformation
- Feature Engineering
- Data Preprocessing
4. Model Trainer
5. Model Evaluation
- MLFLow
- DagsHub
### Workflow Steps
1. Update `config/config.yaml`
2. Update `schema.yaml`
3. Update `params.yaml`
4. Update the entity
5. Update the configuration manager in src config
6. Update the components
7. Update the pipeline
7. Update the `main.py`
## Environment Setup Instructions
1. **Clone the repo:**
```
git clone
cd MLOps-Project-02-End-to-End-Data-Science-Project-with-Deployment
```
2. **Environment Setup:**
Setup conda virtual environment using following command.
```
conda create -p venv python==3.12
```
3. **Activate Environment:**
To activate conda virtual environment use following command.
```
conda activate venv\
```
4. **Install Libraries:**
To install libraries use following command.
```
pip install -r requirements.txt
```
5. **Deactivate Environment:**
To deactivate conda virtual environment use following command.
```
conda deactivate
```