Issues with `vector_query` request in Typesense
TLDR Gauthier struggles with a vector_query
request in Typesense that gets stuck in sending state. Jason suggests an upgrade to 0.25.0.rc43
, but docker container compatibility issues arise. Kishore Nallan has been tagged for assistance.
Jul 03, 2023 (5 months ago)
Gauthier
09:13 PMvector_query
I have this query:
curl --location 'https://MY_CLUSTER.typesense.net/multi_search?query_by=headline%2Cmerchant_name_dba%2Cmerchant_description' \
--header 'Content-Type: application/json' \
--header 'X-TYPESENSE-API-KEY: MY KEY' \
--data '{
"searches": [
{
"collection": "triple_offers",
"q": "*",
"vector_query": "embeddings:([], id: triple_5296, k:25)"
}
]
}'
and this schema (omitting some fields for brevity):
{
"default_sorting_field": "amount",
"enable_nested_fields": true,
"fields": [
{
"facet": false,
"index": false,
"infix": false,
"locale": "",
"name": "triple_offer_id",
"optional": true,
"sort": true,
"type": "int64"
},
{
"facet": true,
"index": true,
"infix": false,
"locale": "",
"name": "merchant_name_dba",
"optional": false,
"sort": false,
"type": "string"
},
{
"facet": false,
"index": true,
"infix": false,
"locale": "",
"name": "headline",
"optional": false,
"sort": false,
"type": "string"
},
{
"facet": false,
"index": true,
"infix": false,
"locale": "",
"name": "amount",
"optional": false,
"sort": true,
"type": "float"
},
{
"facet": false,
"index": true,
"infix": false,
"locale": "",
"name": "merchant_description",
"optional": false,
"sort": false,
"type": "string"
},
{
"facet": false,
"index": false,
"infix": false,
"locale": "",
"name": "embeddings",
"num_dim": 1536,
"optional": true,
"sort": false,
"type": "float[]"
}
],
"name": "triple_offers",
"symbols_to_index": [],
"token_separators": []
}
and this would be, for example, the document to use for the k-nn
{
"document": {
"amount": 5,
"embeddings": [1536 elements -- did not include for Slack],
"headline": "5% back at Patron Mexican Grill Cranberry",
"id": "triple_5296",
"merchant_description": "Patron Mexican Grill is a locally-owned and business built by Martin Bolanos, who moved to the area from Michoacan, Mexico and opened the first Patron Mexican Grill in 2007. Today, with more locations than ever before, you'll find family traditions in everything from the decor and warm hospitality to the hand-crafted fare that's made from the finest ingredients. Showcasing dishes from all over Mexico, there's plenty on the menu to try. From favorites like cool and creamy guacamole to cheesy quesadillas, carnitas, burritos, seafood and steaks, there's lots on the menu. It's all made fresh and to-order, and with reasonable prices AND lunch specials, you can go anytime for a great meal. Also be sure to check out the cold beers and cocktails from the bar. With a large selection of tequilas, you can sip on all your favorites. It's always a fiesta at Patrons, check it out yourself today for a taste of Mexico!\r\n"
}
}
My question:
When I try to send the query over Postman, it stays stuck at "Sending request". Whether I use a local instance or a Typesense Cloud instance.
Is it because my vector/embeddings field is too big?
Thank you!
Jason
09:21 PM0.25.0.rc43
to see if that fixes the issue?Gauthier
11:28 PM0.25.0.rc36
with arm64 but same behavior.Otherwise I'll start the old intel machine tomorrow 😀. Thanks!
Jason
11:29 PMTypesense
Indexed 3005 threads (79% resolved)
Similar Threads
Integrating Semantic Search with Typesense
Krish wants to integrate a semantic search functionality with typesense but struggles with the limitations. Kishore Nallan provides resources, clarifications and workarounds to the raised issues.
Announcement: General Availability of Typesense v0.25.0
Jason announces release of Typesense v0.25.0, listing new features. Users express excitement and ask pertinent questions. Gorkem, Manuel, and Daniel commend the team for the new functionalities. Manish and Tugay share their positive experiences with Typesense. Jason and Kishore Nallan answer questions and thank users for their feedback.
Utilizing Vector Search and Word Embeddings for Comprehensive Search in Typesense
Bill sought clarification on using vector search with multiple word embeddings in Typesense and using them instead of OpenAI's embedding. Kishore Nallan and Jason informed him that their development version 0.25 supports open source embedding models. They also resolved Bill's concerns regarding search performance, language support, and limitations in the search parameters.