https://github.com/hesamsheikh/animai-trainer

Train an LLM to generate cracked Manim animations for mathematical concepts.
https://github.com/hesamsheikh/animai-trainer

fine-tuning llm machine-learning manim mathematics python

Last synced: 4 months ago
JSON representation

Train an LLM to generate cracked Manim animations for mathematical concepts.

• Host: GitHub
• URL: https://github.com/hesamsheikh/animai-trainer
• Owner: hesamsheikh
• Created: 2025-03-02T18:19:06.000Z (8 months ago)
• Default Branch: main
• Last Pushed: 2025-03-14T01:39:15.000Z (7 months ago)
• Last Synced: 2025-06-08T13:06:01.754Z (4 months ago)
• Topics: fine-tuning, llm, machine-learning, manim, mathematics, python
• Language: Python
• Homepage:
• Size: 1.13 MB
• Stars: 16
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          


  

  
AnimAI-Trainer





Train an LLM to generate cracked Manim animations for mathematical concepts.

  

  [![GitHub stars](https://img.shields.io/github/stars/hesamsheikh/AnimAI-Trainer?style=social)](https://github.com/hesamsheikh/AnimAI-Trainer/stargazers)

  [![Twitter Follow](https://img.shields.io/twitter/follow/Hesamation?style=social)](https://twitter.com/Hesamation)

  [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)



## Overview

AnimAI-Trainer is the training module for AnimAI, a framework for generating manim visualizations given a text prompt.

In this repository, we will be handling the data processing, and the model training.

## Plan of Attack (Not definitive)

- ✅ Run manim code string and get the output.

- ✅ Implement a verfier to check for object overlap.

- Process open source manim datasets from Hugging Face, and prepare them for training.

- Train the model.

- Release the model.