https://github.com/subhamsarangi/ragsystemdemo
Use your own data with the power of an LLM
https://github.com/subhamsarangi/ragsystemdemo
chromadb chunking embeddings faiss fastapi huggingface langchain llm ollama python rag semantic-search similarity-search
Last synced: 2 months ago
JSON representation
Use your own data with the power of an LLM
- Host: GitHub
- URL: https://github.com/subhamsarangi/ragsystemdemo
- Owner: subhamsarangi
- Created: 2025-02-19T09:24:02.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-02-19T09:36:36.000Z (over 1 year ago)
- Last Synced: 2025-02-19T10:30:25.094Z (over 1 year ago)
- Topics: chromadb, chunking, embeddings, faiss, fastapi, huggingface, langchain, llm, ollama, python, rag, semantic-search, similarity-search
- Language: Jupyter Notebook
- Homepage:
- Size: 169 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# LLM Augmentation
Use your own data with the power of an LLM
## RAG (Retrieval Augmented Generation)
Way to add data to an llm model without retraining it. needed when llm model has a cutoff date or for using sensitive data.
## Vector Embeddings
A format of data which holds semantic information.
## Semantic Search
Way to identiify things which are similar in meaning, not in the characters. eg: dog is similar to cat; dog is not similar to dot.
## Useful commands for setup
```
ollama --version
ollama list
ollama pull llama3.2:1b
poetry init --no-interaction
poetry add faiss-cpu langchain langchain-community langchain_experimental sentence-transformers langchain-ollama langchain-huggingface fastapi uvicorn python-multipart
```
To use deepseek r1
`ollama pull deepseek-r1:1.5b`
## running the files
`python script.py`
or
`uvicorn server:app --reload`