https://github.com/drod75/llama3.2_vision_example

A basic example of a Llama3.2 vision application!
https://github.com/drod75/llama3.2_vision_example

ai ai-vision jupyter-notebook langchain langchain-python llama3 llama3-2-vision llm pillow streamlit

Last synced: 2 months ago
JSON representation

A basic example of a Llama3.2 vision application!

• Host: GitHub
• URL: https://github.com/drod75/llama3.2_vision_example
• Owner: drod75
• Created: 2025-05-07T18:13:13.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2025-05-12T16:05:58.000Z (about 1 year ago)
• Last Synced: 2025-06-26T07:41:36.152Z (12 months ago)
• Topics: ai, ai-vision, jupyter-notebook, langchain, langchain-python, llama3, llama3-2-vision, llm, pillow, streamlit
• Language: Jupyter Notebook
• Homepage:
• Size: 304 KB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          


    

    
Ollama 3.2 Vision Demonstration

 


Table of Contents



  


    Introduction

  

  


    Tools Used

  

  


    Demonstration

  

  


    Future

  



## Introduction

This is a repository made to demonstrate the cability and usefulness of Llama 3.2 vision, it is meant to show how the model can do in a professional use case, an example of how

it should do is on [Blind-Spot](https://www.blind-spot.app) which uses Google Gemini. The whole point is to test whether or not this model can act as a good replacement and if it exceeds the performance and usefulness of Gemini, the next phase of testing will occur after work in this repository is done which will be the use of an agent through one of the models or both, to test which is better and should be used in the for seeable future.




## Tools Used

The tools used are:

[![Python](https://img.shields.io/badge/Python-3776AB?style=for-the-badge&logo=python&logoColor=yellow)](https://www.python.org/)

[![Ollama](https://img.shields.io/badge/Ollama-%23000000?style=for-the-badge&logo=ollama&logoColor=white)](https://ollama.com/)

[![Langchain](https://img.shields.io/badge/Langchain-%2300A9FF?style=for-the-badge&logo=data:image/svg+xml;base64,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&logoColor=white)](https://www.langchain.com/)

[![Streamlit](https://img.shields.io/badge/Streamlit-%23FF4B4B?style=for-the-badge&logo=streamlit&logoColor=white)](https://streamlit.io/)




## Demonstration

The way the image description will be demonstrated is via a video on a basic Streamlit Application that will use the model to take in various images and show the description out to the user, the whole point is to mimic [Blind-Spot](https://www.blind-spot.app)s and show if any changes can be made to the model to make better descriptions and help blind people better through the use of a new model.

Video Link TBD




## Future

As stated before the point is to create an agent using either Gemini or Ollama to provide better descriptions to the user while ensuring the best possible resources and model. Any future repository made based on this data will be listed below.