mirror of
https://github.com/qdrant/landing_page.git
synced 2026-09-26 22:48:30 +02:00
Apply suggestions from code review
Co-authored-by: David Myriel <davidmyriel@gmail.com>
This commit is contained in:
Kacper Łukawski
co-authored by
David Myriel
David Myriel
parent
9cb2810a69
commit
4e24d80a82
@@ -10,32 +10,32 @@ author_link: https://kacperlukawski.com
|
||||
date: 2024-07-21T14:24:00.000Z
|
||||
---
|
||||
|
||||
It's been more than a year since we published the [original article](/articles/hybrid-search) about building a hybrid
|
||||
It's been over a year since we published the [original article](/articles/hybrid-search/)
|
||||
search system with Qdrant. The idea was straightforward: combine the results from different search methods to improve
|
||||
the search quality. Back in 2023, you were still expected to use an additional service to bring the lexical search
|
||||
capabilities and combine all the intermediate results. Things have changed since then. Once we introduced the support of
|
||||
the sparse vectors, [the additional search service became obsolete](/articles/sparse-vectors/), but you were still
|
||||
required to combine the results from different methods on your side.
|
||||
retrieval quality. Back in 2023, you still needed to use an additional service to bring lexical search
|
||||
capabilities and combine all the intermediate results. Things have changed since then. Once we introduced support for
|
||||
sparse vectors, [the additional search service became obsolete](/articles/sparse-vectors/), but you were still
|
||||
required to combine the results from different methods on your end.
|
||||
|
||||
**Qdrant 1.10 introduces a new Query API that allows you to build a search system that combines different search methods
|
||||
to improve search quality**. Everything is now done on the server side, and you can focus on building the best search
|
||||
**Qdrant 1.10 introduces a new Query API that lets you build a search system by combining different search methods
|
||||
to improve retrieval quality**. Everything is now done on the server side, and you can focus on building the best search
|
||||
experience for your users. In this article, we will show you how to utilize the new [Query
|
||||
API](/documentation/concepts/search/#query-api) to build a hybrid search system.
|
||||
|
||||
## Introduction to the new Query API
|
||||
|
||||
At Qdrant, we believe that vector search capabilities go well beyond the simple search for the nearest neighbors.
|
||||
At Qdrant, we believe that vector search capabilities go well beyond a simple search for nearest neighbors.
|
||||
That's why we provided separate methods for different search use cases, such as `search`, `recommend`, or `discover`.
|
||||
With the latest release, we are happy to introduce the new Query API, which combines all of these methods into a single
|
||||
endpoint and also supports creating nested multistage queries that can be used to build complex search pipelines.
|
||||
|
||||
If you are an existing Qdrant user, you probably have a running search mechanism that you want to improve, either sparse
|
||||
If you are an existing Qdrant user, you probably have a running search mechanism that you want to improve, whether sparse
|
||||
or dense. Your system will act as a baseline for further experiments, so you have a reference point to compare the new
|
||||
search methods with.
|
||||
|
||||
### Available embedding options
|
||||
|
||||
Support of the multiple vectors per point is nothing new in Qdrant, but introducing the Query API makes it even
|
||||
Support for multiple vectors per point is nothing new in Qdrant, but introducing the Query API makes it even
|
||||
more powerful. The 1.10 release brings support for the multivectors, which allows you to treat lists of embeddings
|
||||
as a single entity. There are many possible ways of utilizing this feature, and the most prominent one is the support
|
||||
for late interaction models, such as ColBERT. Instead of having a single embedding for each document or query, this
|
||||
@@ -47,8 +47,8 @@ AI](https://jina.ai/news/what-is-colbert-and-late-interaction-and-why-they-matte
|
||||
|
||||
|
||||
|
||||
Except for the multivectors, you can of course use the standard dense and sparse vectors, play with data types to reduce
|
||||
the use of memory. Named vectors help you to store different dimensionality of the embeddings, which is useful if you
|
||||
Besides multivectors, you can use regular dense and sparse vectors, and experiment with smaller data types to reduce
|
||||
the use of memory. Named vectors can help you store different dimensionality's of the embeddings, which is useful if you
|
||||
use multiple models to represent your data, or want to utilize the Matryoshka embeddings.
|
||||
|
||||
|
||||
@@ -111,9 +111,9 @@ with oversampling the candidates with the dense vectors of the lowest dimensiona
|
||||
number of candidates by reranking them with the higher-dimensional embeddings. Actually, nothing stops you from
|
||||
combining both fusion and reranking.
|
||||
|
||||
Assume we would like to go really sophisticated and build a hybrid search mechanism that combines the results from the
|
||||
Let's go a step further and build a hybrid search mechanism that combines the results from the
|
||||
Matryoshka embeddings, dense vectors, and sparse vectors and then reranks them with the late interaction model. In the
|
||||
meantime, some additional reranking and fusion steps will also be done.
|
||||
meantime, we will introduce additional reranking and fusion steps.
|
||||
|
||||
|
||||
|
||||
@@ -223,7 +223,7 @@ and serve them directly from Qdrant.
|
||||
|
||||
Our webinar on *Building the Ultimate Hybrid Search* takes you through the process of building a hybrid search system
|
||||
with Qdrant Query API. If you missed it, you can [watch the recording](https://www.youtube.com/watch?v=LAZOxqzceEU), or
|
||||
[check the notebooks](https://github.com/qdrant/workshop-ultimate-hybrid-search), if you prefer it that way.
|
||||
[check the notebooks](https://github.com/qdrant/workshop-ultimate-hybrid-search).
|
||||
|
||||
<div style="max-width: 640px; margin: 0 auto; padding-bottom: 1em"> <div style="position: relative; padding-bottom: 56.25%; height: 0; overflow: hidden;"> <iframe width="100%" height="100%" src="https://www.youtube.com/embed/LAZOxqzceEU" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" allowfullscreen style="position: absolute; top: 0; left: 0; width: 100%; height: 100%;"></iframe> </div> </div>
|
||||
|
||||
|
||||
@@ -13,8 +13,8 @@ date: 2023-02-15T10:48:00.000Z
|
||||
|
||||
<aside role="status">
|
||||
This article was published in 2023, more than a year before the Qdrant 1.10 release introducing the new
|
||||
<a href="/documentation/concepts/search/#query-api">Query API</a>. Although the presented concepts are still valid,
|
||||
Qdrant 1.10 makes it easier to build a hybrid search system, without any additional tools in the stack.
|
||||
<a href="/documentation/concepts/search/#query-api">Query API</a>. Although the concepts in this article are still valid,
|
||||
you should use Qdrant 1.10 to build a hybrid search system. It is easier and requires no additional tools from your stack.
|
||||
<a href="/articles/hybrid-search-revamped">Hybrid Search Revamped</a> article presents even more advanced techniques
|
||||
to build a hybrid search system that gives the best of multiple worlds.
|
||||
</aside>
|
||||
|
||||
Reference in New Issue
Block a user