A search engine on top of my open browser tabs.
.
├── LICENSE
├── README.md
├── configs
│ └── dev-config.yaml
├── data
│ ├── data.txt
│ └── processed.json
├── logs
│ └── inbox_search.log
├── main.py
├── pyproject.toml
├── requirements.txt
├── src
│ ├── __init__.py
│ ├── __pycache__
│ ├── preprocessing
│ ├── semantic_search
│ └── ultils.py
├── test
│ ├── __init__.py
│ ├── __pycache__
│ ├── test_preprocessing.py
│ ├── test_semantic_search.py
│ └── test_utils.py
├── uv.lock
Example of data.txt
https://github.com/microsoft/graphrag | microsoft/graphrag: A modular graph-based Retrieval-Augmented Generation (RAG) system
https://poloclub.github.io/transformer-explainer/ | Transformer Explainer: LLM Transformer Model Visually Explained
uv venv venv
source venv/bin/activate # On Windows: .\venv\Scripts\activate
uv pip install -r requirements.txt
pre-commit install
pre-commit run --all-files
python main.py --stage search --env dev --query "python tips"
If you are anything like me, you'll probably have + 100 tabs open in every single device you own that has access to the internet. This isn't great for multiple reasons: resource consumption, visual clutter, ... but specially because you can never find that one article you read and want to recommend to someone, or that resource you've found and you are "saving for latter".
The idea is to try if having search on top of all this link repository would be useful for me. There are two main things I want to achive:
-
Monitor: What am I intersted in lately? -> Most likely because I'd want to devote proper time to read / explore that
-
Build a searcher that given a sentece can retrive the link I'm looking for
We'll start with the basic stuff. I use one tab that has an hability to export links. This is the initial database.
I'll build a small cli on top of my personal information.