LLM Code Generator

This project leverages a DeepSeek R1 Distill Llama 70B to generate Rust projects based on user prompts. The generated code is automatically parsed and saved in a structured repository format.

Project Overview

🔹 Features:

• Accepts user prompts to generate Rust projects.
• Automatically structures project files (Cargo.toml, src/main.rs, etc.).
• Saves the generated files in the repository.
• Ensures proper formatting and error handling.

NOTE

• Here I have integrated a knowledge base which is based on the indexing question and keywords created from the data here.
• The rust compiler is being integrated into this project.So now it sends error back to the llm and model iterates over the code and fix the errors.
• It is not performing very good on complex project.

---

Setup & Installation

Prerequisites

• Python 3.7+ installed
• pip installed
• Rust installed (if testing generated projects)
• A mongo database.

Executing the project

git clone <https://github.com/ItshMoh/Rust_IDE.git>
pip install -r requirements.txt
cd src
cp .env.example .env
cd ..
python3 main.py

Here is the Video link for using it. [https://www.loom.com/share/a5a76d51234941eb95076a57048a2899?sid=80a1150a-2ddf-4c24-9500-675b0f254d4a]