Skip to content

Latest commit

 

History

History
64 lines (41 loc) · 1.77 KB

README.md

File metadata and controls

64 lines (41 loc) · 1.77 KB
Raw

LLM Fine-tuning Environment Setup Script

This script automates the installation of libraries and packages required for LLM fine-tuning on Linux systems. It's designed for Ubuntu 22.04 Server, but can be adapted for other distributions.

Features

  • Installs essential tools like git, tmux, and uv.
  • Sets up Python 3.12 virtual environments using uv.
  • Installs Swift and Axolotl for fine-tuning LLMs.
  • Optionally installs CUDA and NVIDIA drivers using Lambda Stack (Ubuntu 22.04 Server only).
  • Handles systems with or without sudo.

Prerequisites

  • A Linux system (preferably Ubuntu 22.04 Server).
  • Internet connection.

Installation

  1. Clone this repository:
git clone https://github.com/Aunali321/llm-finetuning-setup.git
  1. Navigate to the cloned directory:
cd llm-finetuning-setup
  1. Make the script executable:
chmod +x llm-finetuning-setup.sh
  1. Run the script:
./llm-finetuning-setup.sh

If you're on a system without sudo, you may need to run it with root privileges:

su -c ./llm-finetuning-setup.sh

Usage

The script will guide you through the installation process. It will ask for confirmation before installing CUDA and NVIDIA drivers (if desired).

Compatibility notes

  • This script is optimized for Ubuntu 22.04 Server. Some components, particularly the Lambda Stack for CUDA/NVIDIA driver installation, are designed specifically for this distribution.
  • While the core functionality might work on other distributions, you may need to adapt the package installation commands accordingly.

License

This project is licensed under the MIT License.

Contributing

Contributions are welcome! Please open an issue or submit a pull request if you have any suggestions or improvements.