How srimanikanta2006/resale-wale works
Resale Wale is a used-car price valuation product backed by a trained LightGBM pipeline. Its principal workflow takes eight vehicle inputs through the web interface and REST API, normalizes and scores them, then returns a price estimate with valuation intervals and supporting advice. Offline capabilities cover synthetic benchmark-data generation, model training and statistical evaluation; the live product also exposes comparable listings and model metrics. The API and prediction excerpts show the inference handoff; some UI-to-API wiring and unsampled subsystem details remain unverified.
Main entry points: train.py Model Training · api_server.py REST API · predict.py Valuation Engine · used_cars_marketplace.csv Marketplace Records
Components 15
Data and Analysis
- Dataset Generator
src/generate_dataset.py - Marketplace Records
data/used_cars_marketplace.csv - Exploratory Analysis
src/eda_analysis.py
Model Lifecycle
- Feature Engineering
src/feature_engineering.py - Model Training
src/train.py - Model Evaluation
src/evaluate.py - Valuation Pipeline
models/car_valuation_pipeline.joblib - Model Metrics
reports/metrics.json
Valuation Service
- REST API
src/api_server.py - Valuation Engine
src/predict.py - Valuation Results
- Comparable Listings
src/api_server.py - Metrics Endpoint
src/api_server.py
Web Product
- Web Interface
static/index.html
Other components
- Valuation User
Connections 17
- Valuation User → Web Interface: enters vehicle details
README.md - Web Interface → REST API: requests valuation
- REST API → Valuation Engine: calls prediction
src/api_server.py - Valuation Engine → Valuation Pipeline: loads pipeline
src/predict.py - Valuation Engine → Feature Engineering: imports mappings
src/predict.py - Valuation Engine → Valuation Results: produces price and advice
src/predict.py - REST API → Valuation Results: returns valuation response
src/api_server.py - REST API → Comparable Listings: serves comparisons
src/api_server.py - REST API → Metrics Endpoint: serves metrics
src/api_server.py - Dataset Generator → Marketplace Records: generates records
README.md - Marketplace Records → Exploratory Analysis: analyzes data
- Exploratory Analysis → Feature Engineering: uses transformations
src/eda_analysis.py - Marketplace Records → Model Training: trains from records
README.md - Model Training → Feature Engineering: fits pipeline
src/train.py - Model Training → Model Evaluation: uses metrics functions
src/train.py - Model Training → Valuation Pipeline: saves champion model
- Model Training → Model Metrics: records evaluation