Metadata-Version: 2.1
Name: lerpnet
Version: 1.0.2
Summary: Lightweight learnable interpolation using compact MLPs (edge & ESP32 ready)
Home-page: 
Author: Somendra Seth
Author-email: akhilresearch07@gmail.com
License: MIT
Keywords: machine-learning,interpolation,mlp,edge-ai,embedded,esp32,tinyml
Classifier: Development Status :: 5 - Production/Stable
Classifier: Intended Audience :: Developers
Classifier: Intended Audience :: Education
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3 :: Only
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Topic :: Software Development :: Embedded Systems
Requires-Python: >=3.8
Description-Content-Type: text/markdown
License-File: LICENSE.txt
Requires-Dist: numpy
Requires-Dist: matplotlib

# LerpNet (Learnable Interpolation Network)

LerpNet is a **lightweight neural interpolation library** designed for **fast regression and function approximation**. It trains a compact MLP, learns smooth interpolation from data, and exports results in **Python-friendly `.pkl`** and **embedded-friendly `.json`** formats for deployment on systems like ESP32.  

No heavy deep learning frameworks are required—perfect for **edge AI applications**.

---

## Features

- Learnable interpolation using a **compact MLP**
- Extremely lightweight (**no PyTorch / TensorFlow required**)
- Outputs:
  - `.pkl` for Python inference
  - `.json` for C++ / ESP32 inference
- Built-in utilities:
  - Data normalization
  - Learning rate finder
  - Training metrics (RMSE, MAPE, R²)
- Designed for **edge & embedded deployment**
- Ideal for:
  - Sensor calibration
  - Curve fitting
  - Control systems
  - Real-time interpolation
  - Function approximation
  - Learnable lookup tables
  - Replacement for hard-coded formulas

---

## Generated files:

1) model.pkl — Python inference
2) norm_constants.json — Input/output normalization
3) model_weights.json — Embedded / C++ inference
4) Training plots & CSV (actual_vs_predicted.csv)

## Installation

```bash
pip install lerpnet
