https://github.com/guslovesmath/llama3_macsilicon
Repository for running LLMs efficiently on Mac silicon (M1, M2, M3). Features Jupyter notebook for Meta-Llama-3 setup using MLX framework, with install guide & perf tips. Aims to optimize LLM performance on Mac silicon for devs & researchers.
https://github.com/guslovesmath/llama3_macsilicon
apple llama3 llms mac machine-learning
Last synced: 3 months ago
JSON representation
Repository for running LLMs efficiently on Mac silicon (M1, M2, M3). Features Jupyter notebook for Meta-Llama-3 setup using MLX framework, with install guide & perf tips. Aims to optimize LLM performance on Mac silicon for devs & researchers.
- Host: GitHub
- URL: https://github.com/guslovesmath/llama3_macsilicon
- Owner: GusLovesMath
- Created: 2024-05-03T22:00:07.000Z (about 2 years ago)
- Default Branch: main
- Last Pushed: 2024-05-04T13:12:52.000Z (about 2 years ago)
- Last Synced: 2025-05-11T00:33:00.904Z (about 1 year ago)
- Topics: apple, llama3, llms, mac, machine-learning
- Language: Jupyter Notebook
- Homepage: https://medium.com/@guslovesmath/efficiently-running-meta-llama-3-on-mac-silicon-m1-m2-m3-61585c9bc741
- Size: 93.8 KB
- Stars: 10
- Watchers: 2
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Meta-Llama-3 on Mac Silicon
## Overview
This Jupyter notebook demonstrates how to run the Meta-Llama-3 model on Apple's Mac silicon devices from [**My Medium Post**](https://medium.com/@guslovesmath/efficiently-running-meta-llama-3-on-mac-silicon-m1-m2-m3-61585c9bc741). It includes examples of generating responses from simple prompts and delves into more complex scenarios like solving mathematical problems.
## Requirements
- Apple Mac with M1, M2, or M3 chip
- macOS Monterey or later
- Python 3.x
- Required Python packages: `ipywidgets`, `torch`, `mlx-lm`
## Setup
Clone this repository and install the necessary packages:
```bash
pip install ipywidgets torch mlx-lm
```
[MLX Cummunity on Hugging Face](https://huggingface.co/mlx-community)