Skip to content

Embeddings-based search of Emergent Ventures winners

Notifications You must be signed in to change notification settings

nqureshi/ev-search-python

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Emergent Ventures Embeddings-Based Search

Emergent Ventures (https://www.mercatus.org/emergent-ventures) is a grant program run by Tyler Cowen (an economist who writes Marginal Revolution) at the Mercatus Center.

This repo has a CSV of all EV winners until December 2023. In addition, there is a Python file (search_winners.py) that uses Sentence Transformers to generate embeddings for each of their project descriptions. This enables easy semantic search for EV-funded projects, e.g. you can search for all funding relating to writing a book, or starting a podcast, or working on climate change, for example.

Usage for the Python script: just download the repo, install dependencies

pip install

and then:

python scripts/search_winners.py [query] [number-of-results]

Note that the number-of-results argument is optional and defaults to 10. This returns 10 results.

The scripts folder also contains generate_embeddings_py, which I use to generate embeddings for each EV winner's project description for semantic search. Usage is just:

python scripts/generate_embeddings.py

...which updates the data in /data, using the CSV file ev-winners.csv.

About

Embeddings-based search of Emergent Ventures winners

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published