RAGVizExpander

Welcome to RAGVizExpander, an extension of the RAGxplorer project, where we aim to incorporate new features tailored to personal interests and enhance the overall user experience.

You can use this program to test the effects of different embedding models and chunking strategies.

Install

• Install from source

pip install -e .

• Install from PyPI

pip install ragvizexpander

Components

Below are the components you can use:

Type	What	Where
LLM	OpenAI	Built-in
	ollama	examples
Embedding	OpenAI	Built-in
	SentenceTransformer	Built-in
	HuggingFace	Built-in
	Endpoint-based	examples
File Loader	DOCX	Built-in
	PPTX	Built-in
	TXT	Built-in
	PDF	Built-in

Document loader

Type	Tool	Where
WORD	Default
	unstructured	unstructured
	llama-index	LlamaIndex
	docling	Docling
PDF	Default
	unstructured
	llama-index
	docling
TXT	Default
	unstructured
	llama-index
PPTX	Default
	unstructured
	llama-index
	docling

DEMO

Usge: streamlit run app.py

Huggingface Space: RagVizExpander

Features

• Custom LLM & Embedding model
• Custom chunking strategy
• Support for parsing multiple types of files

Acknowledgments 💙

• This project is forked from RAGxplorer

License

This project is licensed under the MIT license - see the LICENSE for details.