fineweb-2-duckdbs / README.md
BramVanroy's picture
Update README.md
3c8bc42 verified
|
raw
history blame
1.37 kB
metadata
license: odc-by

DuckDB datasets for (dump, id) querying on FineWeb 2

This repo contains some DuckDB databases to check whether a given WARC UID exists in a FineWeb-2 dump. Usage example is given below, but note especially that if you are using URNs (likely, if you are working with CommonCrawl data), then you first have to extract the UID (the id column is of type UUID in the databases).

Download

All files:

huggingface-cli download BramVanroy/fineweb-2-duckdbs --local-dir duckdbs/fineweb-2/ --include *.duckdb --repo-type dataset

One file:

huggingface-cli download BramVanroy/fineweb-2-duckdbs fw2-deu_Latn.duckdb --local-dir duckdbs/fineweb-2/ --repo-type dataset

Usage

import re
import duckdb

uuid_re = re.compile(r"<urn:uuid:([a-zA-Z0-9]{8}-?[a-zA-Z0-9]{4}-?[a-zA-Z0-9]{4}-?[a-zA-Z0-9]{4}-?[a-zA-Z0-9]{12})>")

duckdb_path = "duckdbs/fineweb-2/fw2-deu_Latn.duckdb"
con = duckdb.connect(duckdb_path, read_only=True)

dump = "CC-MAIN-2013-20"
uuid_urn = "<urn:uuid:4cd2db15-ae0c-482a-8688-d023d4b19f60>"
# !! Important: extract the UUID from the URN
uid = uuid_re.sub("\\1", uuid_urn).replace("-", "")

query = "SELECT EXISTS (SELECT 1 FROM dataset WHERE dump = ? AND id = ?)"
exists = bool(con.execute(query, (dump, uuid)).fetchone()[0])

print(f"Does ID {uid} exist in {dump}? {exists}")