datalake

DataLake product cli

Installation

pip install piedatalake

Getting started

datalaker auth

Visit the page:
https://console.piedata.ai/proxy?project_id=64b61f572e9765980a0640d3&action=/login
Click on your logo (the most right icon at navbar)
Paste here email and api key:

email: ########@piedata.ai
API_KEY: ##########

Credentials successfully saved!

datalaker search "Rabbits" --image ./demo/rabbit.jpg --annotation polygon[rabbit,#FF0000]

View results in a file: Rabbits.json

Python api

Simple search by text, annotations and images

from datalake.searcher import Searcher
from datalake.credentials import load_credentials
from datalake.annotations import TagSearch, PolygonsSearch

credentials = load_credentials()
searcher = Searcher(**credentials)

data_request = searcher.search("Rabbits",
                               annotations=[TagSearch("rabbit")],
                               search_limit=9)

print(data_request.wait())

Search by embedding

import numpy as np
from datalake.searcher import Searcher
from datalake.credentials import load_credentials
from datalake.annotations import TagSearch, PolygonsSearch

credentials = load_credentials()
searcher = Searcher(**credentials)

embedding = np.random.randn(512)
data_request = searcher.deepsearch(embedding,
                                   annotations=[TagSearch("rabbit")],
                                   search_limit=9)

print(data_request.wait())