generall

Machine Learning Engineer Test Task: Text2Image Search

Objective

This test task aims to evaluate your grasp of fundamental machine learning and development concepts. Your task involves working with an image dataset to develop a system capable of searching for similar images based on textual queries.

Dataset

Please choose whatever image dataset you prefer, for example:

• crawl some e-commerce website

generall

collection::collection_manager::holders::segment_holder::SegmentHolder::appendable_segments::{{closure}} () at lib/collection/src/collection_manager/holders/segment_holder.rs:238
collection::collection_manager::holders::segment_holder::SegmentHolder::appendable_segments (self=0x7fcb18c430b8) at lib/collection/src/collection_manager/holders/segment_holder.rs:236
collection::collection_manager::holders::segment_holder::SegmentHolder::segment_flush_ordering (self=0x7fcb18c430b8) at lib/collection/src/collection_manager/holders/segment_holder.rs:460
collection::collection_manager::holders::segment_holder::SegmentHolder::flush_all (self=0x7fcb18c430b8, sync=false) at lib/collection/src/collection_manager/holders/segment_holder.rs:477
collection::update_handler::UpdateHandler::flush_segments (segments=...) at lib/collection/src/update_handler.rs:552
collection::update_handler::UpdateHandler::flush_worker::{{closure}} () at lib/collection/src/update_handler.rs:518



collection::collection_manager::holders::segment_

generall

{
  "result": [
    {
      "payload": {
        "location": "html > body > div:nth-of-type(1) > section:nth-of-type(2) > div > div > div > article > h3:nth-of-type(4)",
        "sections": [
          "documentation",
          "documentation/quick_start"
        ],
        "tag": "h3",

generall

import asyncio
import time
from multiprocessing import Pool

import httpx
import numpy as np
from grpclib.client import Channel
from qdrant_client import QdrantClient
from qdrant_client.grpc import PointsStub, WithPayloadSelector
from qdrant_client.http.models import Distance, OptimizersConfigDiff, \

generall

-0.836976	-0.82773167	-0.2321338	0.01756154	-0.33699885	-0.87920535	1.121577	-0.48148146	0.12505984	-0.5362323	-0.04594946	0.21799839	-0.9739923	0.06835833	-0.0905306	-0.4682897	0.21512008	-0.19464707	-0.93475235	-0.5278779	0.07396773	-0.7966168	-0.9158848	0.3736219	0.91806734	0.16620624	0.4292583	0.24501403	0.7698505	0.80631876	-0.40066087	0.06443666	-0.94286495	0.9055134	-0.32404503	0.029998884	0.69158125	0.5815438	-0.019293755	-0.67045003	-0.02590625	0.6856051	0.43317997	-0.022387378	-0.2444922	-0.63252854	0.49054447	0.4057305	0.15405568	0.42062312	0.8911598	-0.107649505	-0.1934745	-0.21087441	-0.98097706	0.7977838	-0.3478621	0.25571245	-0.59421146	-0.18338741	0.33044222	0.11104743	0.0024537258	-0.79205513	0.7596757	-0.18430562	0.2074007	0.24703579	-0.051154852	0.3755078	-0.028878666	-0.7895112	-0.2619388	0.31705666	-0.26749983	1.5598425	0.20117877	-0.8625497	-0.08463721	0.07981754	0.3533825	-0.22829238	-0.8085298	0.83868784	-1.3574781	-0.23162451	0.75115657	-0.67257035	0.20241697	0.09624748	0.11910073	-0.

generall

# File: service.py

from fastapi import FastAPI

# That is the file where NeuralSearcher is stored
from neural_searcher import NeuralSearcher

app = FastAPI()

# Create an instance of the neural searcher

generall

from qdrant_client.http.models import Filter

    ...

    city_of_interest = "Berlin"

    # Define a filter for cities
    city_filter = Filter(**{
        "must": [{
            "key": "city", # We store city information in a field of the same name 

generall

# File: neural_searcher.py

from qdrant_client import QdrantClient
from sentence_transformers import SentenceTransformer


class NeuralSearcher:

    def __init__(self, collection_name):
        self.collection_name = collection_name

generall

import numpy as np
import json

fd = open('./startups.json')

# payload is now an iterator over startup data
payload = map(json.loads, fd)

# Here we load all vectors into memory, numpy array works as iterable for itself.
# Other option would be to use Mmap, if we don't want to load all data into RAM

generall