tests.test_embed

Module Contents

Functions

emb

test_embedder_similarity

test_embedder_search

test_embedder_search_topk_less_ncorpus

test_embed_url_query

Computes snapshots of fetched URLs and computes query similarity on them Fetching is done using celery, but semantic processing is done without it

test_embed_folder_self_uuids

Embeds a list containing 3 pairs of semantically similar items and performs 2-nn search between items

test_embed_folder_query

Fetches and embeds urls from firefox_bookmarks and returns result from a single natural language uery

test_embed_folder_subset_query

Picks subset of items (50+) from big bookmark folder (1000+ items) Fetches content and embeds urls for subset Performs search on original folder using NL queries Retrieves path on original folder

test_embed_hackernews

test_embed_github

test_embed_mixed

Tests whether content from 3 main types (GH, HN, Gen) can be fetched, embedded, and that relevant similarities can be deduced

test_content_snapshot

Not a test per se but a visual check of snapshot generation per *Content

API

tests.test_embed.emb()
tests.test_embed.test_embedder_similarity(emb)
tests.test_embed.test_embedder_search_topk_less_ncorpus(emb)
tests.test_embed.test_embed_url_query(emb)

Computes snapshots of fetched URLs and computes query similarity on them Fetching is done using celery, but semantic processing is done without it

tests.test_embed.test_embed_folder_self_uuids(emb)

Embeds a list containing 3 pairs of semantically similar items and performs 2-nn search between items

tests.test_embed.test_embed_folder_query(emb)

Fetches and embeds urls from firefox_bookmarks and returns result from a single natural language uery

tests.test_embed.test_embed_folder_subset_query(emb)

Picks subset of items (50+) from big bookmark folder (1000+ items) Fetches content and embeds urls for subset Performs search on original folder using NL queries Retrieves path on original folder

tests.test_embed.test_embed_hackernews()
tests.test_embed.test_embed_github()
tests.test_embed.test_embed_mixed()

Tests whether content from 3 main types (GH, HN, Gen) can be fetched, embedded, and that relevant similarities can be deduced

tests.test_embed.test_content_snapshot()

Not a test per se but a visual check of snapshot generation per *Content