ModelSmith-AI / README.md
ACA050's picture
Update README.md
b614da3 verified
---
title: ModelSmith AI
emoji: 🤖
colorFrom: blue
colorTo: green
sdk: gradio
sdk_version: 6.5.1
app_file: app.py
pinned: false
license: apache-2.0
short_description: 'An intelligent ML platform '
---
# ModelSmith AI
An intelligent ML platform that automates tabular classification and regression tasks. It analyzes datasets, recommends optimal strategies, trains models, and provides explanations.
## Features
- **Dataset Analysis**: Automatic detection of data types, missing values, and potential issues
- **Strategy Reasoning**: Intelligent model selection based on dataset characteristics
- **Automated Training**: End-to-end model training with preprocessing pipelines
- **Explainability**: SHAP-based feature importance explanations
- **FastAPI Backend**: RESTful API for seamless integration
## Supported Scope
- **Task**: Tabular classification and regression
- **Input**: CSV files with ≥1200 rows
- **Target**: Binary or multiclass classification, regression
- **Features**: At least 2 usable features after preprocessing
## API Endpoints
- `POST /analyze`: Analyze dataset and get strategy recommendations
- `POST /train`: Train a model on the dataset
- `POST /explain`: Get model explanations and feature importance
- `POST /predict`: Make predictions with trained model
- `GET /health`: Health check
## Deployment
This project is designed for deployment on Hugging Face Spaces using Docker.
### Files for Deployment
- `Dockerfile`
- `requirements.txt`
- `backend/` (entire directory)
### Running Locally
```bash
pip install -r requirements.txt
uvicorn backend.api.main:app --host 0.0.0.0 --port 7860
```
## Limitations
- NLP functionality is disabled
- Requires datasets with ≥1200 rows
- CPU-only, no GPU support
- Stateless API (models saved temporarily)
## Architecture
- **Orchestrator**: Main workflow coordinator
- **Dataset Analyzer**: Data profiling and preprocessing
- **Strategy Reasoner**: Model selection logic
- **Model Factory**: Training and evaluation
- **Explainability Engine**: SHAP explanations
## License
MIT License