mats-lab 0.1.2

Unified AI Interface - Merged AI Tools System

MATS Lab

MATS Lab (Merged AI Tools System) is a unified Python library that simplifies working with all major AI tools, models, and datasets.
It offers a clean, human-readable interface to download, run, fine-tune, and deploy models from platforms like Hugging Face and Kaggle, with built-in support for training, compression, and optimization.

---

Features

• Simple and natural syntax inspired by English
• Load any pretrained model from Hugging Face using only its model ID
• Download any dataset from Kaggle or Hugging Face by ID
• Automatically selects backend framework (PyTorch, TensorFlow, or scikit-learn)
• Unified interface for inference and prompt-based tasks
• Customize model parameters like temperature, max tokens, top-k, etc.
• Control GPU/CPU memory usage during inference
• Train your own models easily with built-in functions
• Support for model compression techniques like quantization, QLoRA, etc.
• CLI tool to run MATS from the terminal

---

Supported Libraries

MATS Lab integrates with the following major Python AI libraries:

• PyTorch
• TensorFlow
• scikit-learn
• Transformers (Hugging Face)
• Datasets (Hugging Face Datasets)
• Kaggle
• PEFT (Parameter-Efficient Fine-Tuning)
• NumPy
• Matplotlib / Seaborn (for optional visualization)

---

👨‍💻 About the Developer

Assem Sabry is a 17-year-old AI Engineer from Egypt, passionate about building accessible and developer-friendly machine learning tools. He is the creator of MATS Lab and actively works on merging advanced AI technologies into a unified and simplified interface. Learn more at assemsabry.netlify.app.

Installation

pip install mats_lab