Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/arska/onpremllm
Demo using https://github.com/amaiya/onprem
https://github.com/arska/onpremllm
Last synced: 1 day ago
JSON representation
Demo using https://github.com/amaiya/onprem
- Host: GitHub
- URL: https://github.com/arska/onpremllm
- Owner: arska
- Created: 2023-11-03T16:55:31.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2023-12-15T17:43:38.000Z (11 months ago)
- Last Synced: 2023-12-15T18:54:23.120Z (11 months ago)
- Language: Python
- Size: 8.36 MB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Aarnos onprem.LLM demo
Using https://github.com/amaiya/onprem with a modified frontend to demo different models and RAG
## Installation
- On my Macbook pro M1 with Mac Os 14 (Sonoma), I used Python 3.11 (3.12 did not work), installed from MacPorts (https://www.macports.org/install.php)
- create a virtual environment to install in: python3 -m virtualenv venv; source venv/bin/activate
- Install llama-cpp-python with Apple Metal support: CMAKE_ARGS="-DLLAMA_METAL=on" FORCE_CMAKE=1 pip install llama-cpp-python
- Install the other dependencies: pip install -r requirements.txt## Running
- load the python environment: source venv/bin/activate
- start the WebGUI: streamlit run app.py## Data
- onprem stores all models and vectordb at "~/onprem_data/"
- we use the (supplied) PDF in ./sample_data/ for retrieval augmented generation (RAG)## Screenshot
![Screenshot](screenshot.png)