Metadata-Version: 2.4
Name: arthaLM
Version: 0.1.0
Summary: Plug-and-play pipeline for Artha LLM
Home-page: https://github.com/vyomie/artha
Author: Vyom N. Patel
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.7
Description-Content-Type: text/markdown
Requires-Dist: torch
Requires-Dist: transformers
Requires-Dist: tqdm
Dynamic: author
Dynamic: classifier
Dynamic: description
Dynamic: description-content-type
Dynamic: home-page
Dynamic: requires-dist
Dynamic: requires-python
Dynamic: summary

# Artha Language Model (ArthaLM)

## World's First & Most Affordable Custom LLM â€“ Built by a Teen

**Artha-1** is the **world's first Liquid Neural Network-powered language model pipeline** trained and deployed end-to-end by an independent teen researcher. Unlike traditional LLMs that require millions of dollars, proprietary infrastructure, and industrial compute clusters, **Artha-1 was built in under 3 days using accessible tools and mid-range hardware.**

### What Makes It a World First?

-  **LNN-Based Reasoning Core:**  
  This is the **first open-source LLM** to integrate a **Liquid Neural Network (LNN)** core for deep, dynamic reasoning inside compressed latent space.

-  **Built With ~Zero Budget:**  
  Trained using just local GPUs and open datasets, **no enterprise backing or funding** was involved. This makes Artha-1 arguably the **cheapest working LLM** architecture available to the public.

-  **Created by a Teen Researcher:**  
  From architecture design and training to deployment and packaging, every step was executed by an **independent teen developer**, proving that **you donâ€™t need a PhD or billion-dollar lab** to innovate in AI.
