search_engine / README.md
Vitomir Jovanović
Streamlit + Readme.md
a1d6c7a
|
raw
history blame
987 Bytes
---
title: Search Engine
emoji: 🔥
colorFrom: green
colorTo: red
sdk: streamlit
sdk_version: 1.39.0
app_file: app.py
pinned: false
short_description: Semantic Search engine with Faiss
---
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-
### For local deployment run
```
main.py
```
which will create swagger app with endpoints on [localhost:8084](http://127.0.0.1:8084/docs). First endpoint return the top k semanticaly most similar prompts with query prompt. Second endpoint returns all similarites with query (only applicable for very small datasets).
```
data_reader.py
```
creates data of various prompts for encoding into vector database. Local database encoded only 6000 prompts.
Faiss index that is used is small and not optimized, used for experimental datasets. Search is brute force, not optimised.
### Streamlit
```
streamlit run app.py
```
should be run for streamlit app, it can be assessed locally on http://localhost:8501.