id stringlengths 36 36 | url stringclasses 504
values | title stringclasses 475
values | section stringlengths 0 148 | content_type stringclasses 3
values | has_code bool 2
classes | text stringlengths 147 22.5k | vector listlengths 384 384 |
|---|---|---|---|---|---|---|---|
da3c6e78-315f-513a-8c15-dd5e56a92fcf | /documentation/tutorials-operations/create-snapshot/ | Snapshots | Create and download snapshots | documentation | true | Snapshots > Create and download snapshots
Qdrant exposes an HTTP endpoint to request creating a snapshot, but we can also call it with the Python SDK.
Our setup consists of 3 nodes, so we need to call the endpoint **on each of them** and create a snapshot on each node. While using Python SDK, that means creating a sep... | [
0.0079322699457407,
-0.04850583150982857,
-0.011905995197594166,
-0.019816454499959946,
0.009694121778011322,
-0.01821613684296608,
0.06347677111625671,
0.03833428770303726,
0.07666481286287308,
-0.060163021087646484,
0.047320351004600525,
0.030497096478939056,
0.049995772540569305,
-0.035... |
da4b7de4-22a1-5fab-8393-b64c709854dd | /course/multi-vector-search/ | Multi-Vector Search Course | Who it's for | course | false | Multi-Vector Search Course > Who it's for
ML, backend, and search engineers who want to go beyond single-vector embeddings. Requires intermediate Python, basic familiarity with vector search concepts (embeddings, similarity metrics), and comfort with APIs. | [
0.010091223753988743,
-0.05152435228228569,
-0.06673262268304825,
-0.019327234476804733,
0.059966590255498886,
-0.04199843853712082,
0.045113835483789444,
0.03613181784749031,
0.06943237781524658,
-0.012311789207160473,
0.038694288581609726,
0.035693373531103134,
0.06101067364215851,
-0.02... |
da50d032-4128-5dac-8297-baf8cf50a1e0 | /documentation/search/low-latency-search/ | Low-Latency Search | Query Indexed Data Only > Pin Reads with Route Affinity | documentation | false | Low-Latency Search > Query Indexed Data Only > Pin Reads with Route Affinity
*Available as of v1.19.0*
Enabling `prevent_unoptimized` makes cross-replica "blinking" more prominent: points held back until a segment is optimized become visible on each replica at slightly different times, and because reads are spread ac... | [
0.018531909212470055,
-0.07478298991918564,
-0.044277820736169815,
-0.0366998016834259,
0.049600932747125626,
-0.04018841311335564,
0.058265089988708496,
0.06305714696645737,
0.0635204017162323,
-0.013510316610336304,
0.05252628028392792,
0.009965858422219753,
0.06523079425096512,
-0.01272... |
da55fd53-f1bf-5fe1-8442-350193b49298 | /documentation/faq/qdrant-fundamentals/ | Qdrant Fundamentals | Search > Does Qdrant support a full-text search or a hybrid search? | documentation | false | Qdrant Fundamentals > Search > Does Qdrant support a full-text search or a hybrid search?
Yes. Qdrant supports full-text search with BM25, full-text filtering, and hybrid search that combines them with vector search.
What Qdrant can do:
- Rank points by relevance to a text query with [BM25 full-text search](/documen... | [
0.039362549781799316,
-0.030401209369301796,
-0.05371183156967163,
-0.04285450279712677,
0.05598609149456024,
-0.030390188097953796,
0.0937144011259079,
0.03154328465461731,
0.06844044476747513,
-0.0026020221412181854,
0.0575346015393734,
0.02403254806995392,
0.0471888929605484,
-0.0312460... |
da692c8b-d155-54c4-800f-de7b86057974 | /course/multi-vector-search/module-3/final-project/ | Final Project: Build Your Own Multi-Vector Search System | Share Your Results | course | false | Final Project: Build Your Own Multi-Vector Search System > Share Your Results
We'd love to see what you build. Share your project on the [Qdrant Discord](https://discord.gg/qdrant) in the [#course-submissions](https://discord.com/channels/907569970500743200/1429673887590776832) channel.
Tell us about:
- **Your datas... | [
0.041088663041591644,
-0.06035095453262329,
-0.027908949181437492,
-0.01103935670107603,
0.06598721444606781,
-0.02249888889491558,
0.07226964086294174,
0.02486868016421795,
0.0644586831331253,
-0.04966980218887329,
0.031085379421710968,
-0.0011342158541083336,
0.06294424086809158,
-0.0402... |
da6d5270-e588-5a45-844a-36bbdbd875ae | /course/multi-vector-search/module-2/colpali-family/ | ColPali Family Overview | What's next | course | false | ColPali Family Overview > What's next
You've explored the ColPali family ecosystem - from ultra-compact models like ColFlor (174M parameters) to multilingual alternatives from NVIDIA and Nomic AI. You've learned how bidirectional attention architectures can achieve competitive results with far fewer parameters, and se... | [
0.020144600421190262,
0.005368340760469437,
-0.0424099825322628,
-0.04787221550941467,
0.1100248321890831,
-0.02086147852241993,
0.0732928141951561,
0.05757638439536095,
0.023938514292240143,
-0.050052762031555176,
0.043193407356739044,
-0.018873611465096474,
0.037239912897348404,
-0.02108... |
da6ef47f-c6f0-53ca-89c9-997abe1b6c5f | /documentation/ops-monitoring/monitoring/ | Monitoring & Telemetry | Metrics | documentation | false | Monitoring & Telemetry > Metrics
Qdrant exposes various metrics in Prometheus/OpenMetrics format, commonly used together with Grafana for monitoring.
Two endpoints are available:
- `/metrics` for metrics of a Qdrant node/peer, see [all metrics](#node-metrics-metrics).
- `/sys_metrics` (Qdrant Cloud only) for metric... | [
0.0152586679905653,
-0.020642472431063652,
-0.02168229967355728,
-0.004502881783992052,
0.04652046039700508,
-0.03556159883737564,
0.09194314479827881,
0.08451051265001297,
0.08610091358423233,
-0.05484231561422348,
0.045797910541296005,
-0.033075232058763504,
0.06075183302164078,
-0.01282... |
da78f3d3-7806-561c-b7aa-6258f4840de0 | /documentation/ops-monitoring/hybrid-cloud-datadog/ | Monitoring with Datadog | Step 2: Configure the Datadog Agent to Scrape Qdrant Metrics | documentation | true | Monitoring with Datadog > Step 2: Configure the Datadog Agent to Scrape Qdrant Metrics
To monitor Qdrant, configure the Datadog Agent to scrape metrics from your Qdrant database. Create a `DatadogAgent` resource that adds an [OpenMetrics](https://docs.datadoghq.com/integrations/openmetrics/) check through the Agent's ... | [
0.045845456421375275,
-0.013820214197039604,
-0.047197721898555756,
0.026202967390418053,
0.058975931257009506,
-0.03700517117977142,
0.08524733036756516,
0.04625489562749863,
0.03768803924322128,
-0.022680383175611496,
0.033354051411151886,
0.014208954758942127,
0.04050584137439728,
-0.02... |
da7cb823-40cd-5661-8be4-421b3dccc748 | /articles/what-is-a-vector-database/ | What is a Vector Database? | Hybrid Search > Dense vs. Sparse Vectors | article | false | What is a Vector Database? > Hybrid Search > Dense vs. Sparse Vectors
Dense and sparse vectors represent data in fundamentally different ways, and each excels at a different type of retrieval. Understanding their strengths and limitations explains why combining them can produce better search results than relying on ei... | [
0.005623673554509878,
-0.034863196313381195,
-0.08791080862283707,
-0.0793699324131012,
0.04157600924372673,
-0.036834124475717545,
0.10413437336683273,
0.05204867944121361,
0.06745582073926926,
-0.03661264106631279,
0.01780729927122593,
0.07038939744234085,
0.046747200191020966,
0.0189676... |
da82e706-7b58-5022-aea3-1a77fc9e037a | /articles/bm42/ | BM42: Attention-Based Sparse Embeddings for Hybrid Search | Fostering curiosity and experimentation | article | false | BM42: Attention-Based Sparse Embeddings for Hybrid Search > Fostering curiosity and experimentation
Despite all of its advantages, BM42 is not always a silver bullet.
For large documents without chunks, BM25 might still be a better choice.
There might be a smarter way to extract the importance information from the tr... | [
0.05581391230225563,
-0.022746693342924118,
-0.039674852043390274,
-0.04843366891145706,
0.032259952276945114,
-0.06425561010837555,
0.0889117494225502,
0.044345494359731674,
0.030020123347640038,
-0.004955448675900698,
0.02575499564409256,
0.045023318380117416,
0.036418139934539795,
-0.02... |
da9282ba-30b9-5c12-bfd8-e86696ab3d39 | /documentation/examples/graphrag-qdrant-neo4j/ | GraphRAG with Qdrant and Neo4j | Implementation > Prerequisites > Setting Up Environment Variables | documentation | true | GraphRAG with Qdrant and Neo4j > Implementation > Prerequisites > Setting Up Environment Variables
Before initializing the clients, we load the necessary credentials from environment variables.
Show code
```python
# Load environment variables
load_dotenv()
# Get credentials from environment variables
qdrant_key = o... | [
0.005344732664525509,
-0.06750737130641937,
0.004309060052037239,
-0.056320205330848694,
0.04432832822203636,
-0.07170119881629944,
0.027664383873343468,
0.07295604050159454,
0.027008140459656715,
-0.051211319863796234,
-0.005349092185497284,
0.02831803262233734,
0.09003904461860657,
-0.02... |
daa4c909-5b32-5a09-9a77-fb8f826601ad | /course/essentials/day-1/chunking-strategies/ | Text Chunking Strategies | The Real Beginning: Data Structure | course | false | Text Chunking Strategies > The Real Beginning: Data Structure
Our pipeline starts with the data and how we represent what we want to search. In practice, this means thinking about the structure of the data we're storing. Most real-world data is messy:
- Text documents are long
- Product descriptions vary in length
... | [
0.035130299627780914,
-0.024555066600441933,
-0.05777701735496521,
-0.04183768853545189,
0.04826800525188446,
-0.05253015086054802,
0.04593174532055855,
0.024368377402424812,
0.06456337124109268,
-0.024254798889160156,
0.010609663091599941,
-0.012051530182361603,
0.09289281070232391,
-0.02... |
daabcf27-4fc7-5403-906d-a5a07b2b2e63 | /documentation/search-tuning/when-your-collection-outgrows-ram/ | When Your Collection Outgrows RAM | documentation | false | When Your Collection Outgrows RAM
Once a collection no longer fits in RAM, the kernel evicts vector pages, and the next query waits on a disk read to get them back. Quantization buys that memory back. Qdrant keeps a compressed copy of each dense vector in RAM and moves the full-precision originals to disk.
[TurboQuan... | [
0.03999682515859604,
-0.03802693635225296,
-0.06630130112171173,
-0.044362153857946396,
0.06819840520620346,
-0.01539657637476921,
0.06184622645378113,
0.05468730628490448,
0.1082022488117218,
-0.03904152661561966,
0.027788160368800163,
0.041047416627407074,
0.024050075560808182,
-0.017326... | |
dab4a04a-de4f-5fa4-939b-8a871cc89332 | /documentation/edge/edge-vs-qdrant-cluster/ | Edge vs. Qdrant Cluster | Architecture & Deployment | documentation | false | Edge vs. Qdrant Cluster > Architecture & Deployment
How each option runs, connects, and scales.
| | Qdrant Edge | Qdrant Server |
| --- | --- | --- |
| **Architecture** | Embedded, in-process library | Client-server, accessed over the network |
| **Connectivity** | Works fully offline | Requires network access to the... | [
0.0143936388194561,
0.002633987460285425,
-0.08557606488466263,
-0.03320673108100891,
0.08242913335561752,
-0.059151630848646164,
0.07687630504369736,
0.045122724026441574,
0.0801284983754158,
-0.04649524390697479,
0.03893495351076126,
0.0012891399674117565,
0.019948996603488922,
0.0089849... |
dab8b3be-c705-5d43-9ac1-e650f4b18304 | /documentation/ops-monitoring/hybrid-cloud-datadog/ | Monitoring with Datadog | Step 5: Stop Metric Collection After Testing | documentation | true | Monitoring with Datadog > Step 5: Stop Metric Collection After Testing
To stop collecting Qdrant metrics once you are done testing, remove the `openmetrics.yaml` entry from the `DatadogAgent` resource and reapply it:
```shell
kubectl apply -f datadog-agent.yaml
```
The Agent stops scraping Qdrant on its next configu... | [
0.04703069105744362,
-0.01596791110932827,
-0.007591040804982185,
0.0064569320529699326,
0.027042359113693237,
-0.049648161977529526,
0.059220485389232635,
0.03320717439055443,
0.0799623429775238,
-0.03669428080320358,
0.03513810783624649,
0.05433373898267746,
0.03526817262172699,
0.008485... |
dadb613d-12e6-55fa-b097-f2a91340e5cb | /course/multi-vector-search/module-1/maxsim-distance/ | MaxSim Distance Metric | The HNSW Challenge | course | true | MaxSim Distance Metric > The HNSW Challenge
There's a significant challenge related to MaxSim when it comes to indexing these representations efficiently. HNSW (Hierarchical Navigable Small World) graphs enable fast approximate nearest neighbor search by building static proximity graphs. They rely on the assumption th... | [
0.06977654248476028,
0.008444997482001781,
-0.060477323830127716,
-0.020251266658306122,
0.06313750892877579,
-0.061167437583208084,
0.036604080349206924,
0.02750348299741745,
0.058424606919288635,
-0.031719520688056946,
0.04258856177330017,
0.0269594918936491,
0.07033038139343262,
-0.0017... |
daf31ad7-45cb-5643-90ed-51cd1ca18a87 | /course/essentials/day-1/chunking-strategies/ | Text Chunking Strategies | Adding Meaning with Metadata | course | true | Text Chunking Strategies > Adding Meaning with Metadata
Chunks by themselves are just fragments of text. They don't tell you where they came from, what they belong to, or how to control what gets retrieved.
**That's where metadata comes in.**
In Qdrant, this metadata lives in the **payload** - a JSON object attached... | [
0.01256086677312851,
-0.01857210509479046,
-0.07816088199615479,
-0.03197398781776428,
0.0532381646335125,
-0.03240159526467323,
0.03254213184118271,
0.016998739913105965,
0.08148398250341415,
-0.050962816923856735,
0.04419462010264397,
0.0031224219128489494,
0.053189776837825775,
-0.04173... |
db05058a-e651-5d61-8d29-9843cefa97d6 | /course/essentials/day-0/qdrant-cloud/ | Qdrant Setup | Qdrant Cloud Inference | course | false | Qdrant Setup > Qdrant Cloud Inference
Qdrant Cloud also offers **[Cloud Inference](/cloud-inference/)**—managed embedding generation for text and images. Skip running your own embedding models; create vectors in Qdrant Cloud and write them straight into your collections.
Cut steps from your pipeline: send raw text or... | [
0.052389245480298996,
-0.03325314819812775,
-0.017595337703824043,
-0.07322511076927185,
0.04890136048197746,
-0.014377384446561337,
0.11838623881340027,
0.027253536507487297,
0.05502576753497124,
-0.02629593014717102,
0.019174491986632347,
0.012552829459309578,
0.06448272615671158,
-0.042... |
db0c680e-5464-5e5f-a092-0576b90c7126 | /documentation/examples/rag-contract-management-stackit-cohere/ | Region-Specific Contract Management System | Components | documentation | false | Region-Specific Contract Management System > Components
A contract management platform is not a simple CLI tool, but an application that should be available to all team
members. It needs an interface to upload, search, and manage the documents. Ideally, the system should be
integrated with org's existing stack, and t... | [
0.024900266900658607,
-0.04980906844139099,
-0.0813712552189827,
-0.03332136198878288,
0.05796308442950249,
-0.03209671005606651,
0.06954289972782135,
0.021888993680477142,
0.09192211180925369,
-0.03419448435306549,
0.012406537309288979,
-0.004081444814801216,
0.046867065131664276,
-0.0491... |
db0f2f74-4e25-5f4b-bc30-9e265b8c7f83 | /documentation/send-data/qdrant-airflow-astronomer/ | Querying with Airflow | Implementation > Configure credentials | documentation | true | Querying with Airflow > Implementation > Configure credentials
We can set up provider connections using the Airflow UI, environment variables or the `airflow_settings.yml` file.
Add the following to the `.env` file in the project. Replace the values as per your credentials.
```env
AIRFLOW_CONN_QDRANT_DEFAULT='{
... | [
0.024202115833759308,
-0.037379831075668335,
-0.02672361396253109,
-0.057101134210824966,
0.036218058317899704,
-0.020837457850575447,
-0.004692861344665289,
0.06436659395694733,
-0.01521233282983303,
-0.023804500699043274,
0.028346111997961998,
-0.03114105574786663,
0.08111980557441711,
-... |
db3ce76a-6900-54ce-9669-590709f688f8 | /documentation/tutorials-basics/multimodal-search/ | Multimodal Search | Dataset | documentation | false | Multimodal Search > Dataset
To make the demonstration simple, this tutorial uses a tiny dataset of images and their captions.
Download the [tutorial images](https://github.com/qdrant/examples/tree/master/multimodal-search/images) and place them in a folder named `images`, in the same folder as your code or notebook. | [
0.03438912704586983,
-0.04700571671128273,
-0.0399433858692646,
-0.030953284353017807,
0.06479305773973465,
-0.026314374059438705,
0.03914906084537506,
0.048501260578632355,
0.044677890837192535,
-0.027858823537826538,
0.032340507954359055,
0.03439885750412941,
0.09735056012868881,
-0.0259... |
db4d171c-5825-57dd-b1c1-ef9183823e66 | /documentation/tutorials-build-essentials/video-anomaly-edge-part-3/ | Video Anomaly Detection Part 3: Scoring, Governance, and Deployment | Deployment | documentation | true | Video Anomaly Detection Part 3: Scoring, Governance, and Deployment > Deployment
```bash
# Clone and install
git clone https://github.com/qdrant/video-anomaly-edge.git
cd video-anomaly-edge
uv sync
# Configure
cp .env.example .env
# Edit .env with your Qdrant URL, Twelve Labs API key, etc.
# Option 1: Run with Docke... | [
0.04317086562514305,
-0.02413659542798996,
-0.06400194764137268,
-0.07727383822202682,
0.04297715052962303,
0.018950751051306725,
0.006291042547672987,
0.041343241930007935,
0.06563922762870789,
-0.0003957878507208079,
0.07277969270944595,
0.03141101822257042,
0.04066765308380127,
-0.03290... |
db64ab25-ea39-5cc8-bc9f-6e3200d1645f | /documentation/tutorials-develop/search-as-you-type/ | Semantic Search As You Type | documentation | false | Semantic Search As You Type
| Time: 30 min | Level: Intermediate | Output: [GitHub](https://github.com/qdrant/page-search) |
| --- | ----------- | ----------- |
A search-as-you-type box sends a request on almost every keystroke, so it puts a vector search engine under steady load and makes every millisecond of latenc... | [
0.037463560700416565,
-0.08716490864753723,
-0.07088559865951538,
-0.01971968449652195,
0.048989374190568924,
-0.02639365941286087,
0.024503540247678757,
0.0024167972151190042,
0.06828499585390091,
-0.024306075647473335,
0.054990485310554504,
0.009790248237550259,
0.08024752885103226,
-0.0... | |
db6f0e7f-3506-576a-9f2c-f1c019fcd84f | /documentation/tutorials-operations/time-based-sharding/ | Time-Based Sharding | Ingest Historical Data | documentation | true | Time-Based Sharding > Ingest Historical Data
Time series data often arrives in streams, with new data points continuously being added. Each data point may include a timestamp indicating when it was created. You can use this timestamp to determine which shard a data point belongs to. If your data does not have timestam... | [
-0.011519870720803738,
-0.01095673069357872,
0.002978211035951972,
-0.02576984465122223,
0.04441435635089874,
-0.027842121198773384,
0.027674298733472824,
0.0491168275475502,
0.045327018946409225,
-0.019277513027191162,
0.0340723842382431,
0.02167665585875511,
0.013792709447443485,
-0.0430... |
db77cded-dc3b-5200-8852-1c9ff199a492 | /articles/gridstore-key-value-storage/ | Introducing Gridstore: Qdrant's Custom Key-Value Store | Why We Built Our Own Storage Engine | article | false | Introducing Gridstore: Qdrant's Custom Key-Value Store > Why We Built Our Own Storage Engine
Databases need a place to store and retrieve data. That’s what Qdrant's [**key-value storage**](https://en.wikipedia.org/wiki/Key–value_database) does—it links keys to values.
When we started building Qdrant, we needed to pic... | [
0.03728223964571953,
-0.03644111007452011,
-0.06264685094356537,
-0.022423410788178444,
0.06524032354354858,
-0.05491047725081444,
0.026142118498682976,
0.02301155962049961,
0.05988379940390587,
-0.026328174397349358,
0.014415758661925793,
-0.02955852448940277,
0.0458252988755703,
-0.03468... |
db8703f6-d342-54d8-acce-3b93e2d6bd78 | /documentation/tutorials-operations/prevent-unoptimized-usage/ | Prevent Unoptimized Usage | documentation | false | Prevent Unoptimized Usage
| Time: 20 min | Level: Intermediate | Output: [GitHub](https://github.com/qdrant/examples/blob/master/prevent_unoptimized_usage/prevent_unoptimized.ipynb) | [](https://githubtocolab.com/qdrant/examples/blob/master/prev... | [
0.04744463041424751,
-0.057894516736269,
-0.06584445387125015,
-0.02877802774310112,
0.10032226890325546,
-0.031634483486413956,
0.053539663553237915,
0.01641545258462429,
0.0978139191865921,
-0.04769811034202576,
0.04333991929888725,
0.01479044184088707,
0.09346019476652145,
-0.0658335089... | |
db8d7816-b630-5aab-bb59-ede9f2c5ab29 | /documentation/tutorials-build-essentials/video-anomaly-edge-part-1/ | Video Anomaly Detection Part 1: Architecture, Twelve Labs, and NVIDIA VSS | Why kNN Beats Classifiers for Anomaly Detection | documentation | false | Video Anomaly Detection Part 1: Architecture, Twelve Labs, and NVIDIA VSS > Why kNN Beats Classifiers for Anomaly Detection
Let's prove why this matters with some numbers. We tested CLIP ViT-B/32 (512-dim, single-frame image embeddings) as an alternative and it scored **0.23 AUC-ROC**, near random. This failure is ins... | [
0.054668158292770386,
-0.0379478819668293,
-0.07479869574308395,
-0.045453883707523346,
0.0783044844865799,
-0.015882354229688644,
0.02157711796462536,
0.00444134883582592,
0.08271985501050949,
-0.024399204179644585,
0.04267314821481705,
-0.011284439824521542,
0.0763106420636177,
-0.017460... |
db936efd-5a9e-5c9b-ac76-7a5692a455f0 | /documentation/cloud-premium/ | Premium Tier | Customer-Managed Encryption Keys | documentation | false | Premium Tier > Customer-Managed Encryption Keys
Qdrant Cloud encrypts every storage volume at rest by default. Premium customers can encrypt their cluster's volumes with a key they own in their cloud provider's key management service (KMS): AWS KMS, GCP Cloud KMS, or Azure Key Vault. The key must be in the same region... | [
0.04554099589586258,
-0.015698956325650215,
-0.048795752227306366,
-0.04729297384619713,
0.023097995668649673,
-0.03892574459314346,
0.060782305896282196,
0.04769453406333923,
0.05486205592751503,
0.013001603074371815,
0.015403319150209427,
-0.011182563379406929,
0.0867583379149437,
-0.041... |
db9fdd7e-8251-5380-845f-e6f8dffdbe6b | /documentation/tutorials-develop/serverless-semantic-search/ | Serverless Semantic Search | documentation | false | Serverless Semantic Search
| Time: 45 min | Level: Intermediate |
| --- | ----------- |
Do you want to add semantic search to your website or online app without running a server? In this tutorial, you build a small search service in Rust on AWS Lambda. For each query, the function embeds the text with Cohere, searche... | [
0.05474163219332695,
-0.03821903094649315,
-0.03459344059228897,
-0.017622359097003937,
0.05644112452864647,
-0.046848829835653305,
0.02703557349741459,
0.020321961492300034,
0.023779915645718575,
-0.028585823252797127,
0.07387731969356537,
0.01396086160093546,
0.04582953825592995,
-0.0360... | |
dba1bf86-63a0-5e1b-abde-859cbe52518d | /documentation/embeddings/nvidia/ | NVIDIA | Next Steps | documentation | false | NVIDIA > Next Steps
- [Use named vectors](/documentation/manage-data/vectors/#named-vectors) to keep text and image vectors on the same points.
- Read the [NVIDIA Embedding NIM API guide](https://docs.nvidia.com/nim/nemo-retriever/embedding/latest/use-the-api-openai.html) for embedding types, modality, and truncation ... | [
0.0409935899078846,
-0.07123900949954987,
-0.03481556102633476,
-0.036193542182445526,
0.05995941162109375,
-0.003477406455203891,
0.05122234672307968,
0.019754549488425255,
0.04571075737476349,
-0.0184630174189806,
0.04241545498371124,
0.00212492304854095,
0.05742086097598076,
-0.01728809... |
dbaa192a-b387-50d7-a0cd-42594328615e | /documentation/ops-configuration/administration/ | Administration | Strict Mode | documentation | true | Administration > Strict Mode
*Available as of v1.13.0*
Strict mode is a feature to restrict certain type of operations on a collection in order to protect the Qdrant cluster. The goal is to prevent inefficient usage patterns that could overload the system.
Strict mode ensures a more predictable and responsive servic... | [
0.03870568424463272,
-0.011479418724775314,
-0.056925490498542786,
-0.030970029532909393,
0.07212141156196594,
-0.03279692307114601,
0.0747290849685669,
0.026674414053559303,
0.0643644705414772,
-0.021723385900259018,
0.017993181943893433,
-0.015943992882966995,
0.06823796778917313,
-0.034... |
dbbbf3b7-e2a7-5f21-b5fb-0ca679ed28b7 | /documentation/tutorials-build-essentials/agentic-rag-camelai-discord/ | Discord RAG Bot | Step 7: Run and Check the Bot | documentation | true | Discord RAG Bot > Step 7: Run and Check the Bot
Ingest the documentation once:
```bash
python discord_rag_bot.py ingest
```
The command prints the source URL and number of chunks indexed, then the total vector count. Counts depend on the current documentation content. A nonempty collection should refuse a second ing... | [
0.03148949518799782,
-0.04826672747731209,
-0.040779974311590195,
-0.008992772549390793,
0.05783797428011894,
-0.032156672328710556,
0.03114216774702072,
0.0196592528373003,
0.051092375069856644,
-0.03423507511615753,
0.009061369113624096,
0.0030108208302408457,
0.03810178115963936,
-0.038... |
dbc5c44d-65f3-5de6-96ab-c16727ceb58e | /documentation/tutorials-build-essentials/video-anomaly-edge-part-3/ | Video Anomaly Detection Part 3: Scoring, Governance, and Deployment | API Endpoints | documentation | true | Video Anomaly Detection Part 3: Scoring, Governance, and Deployment > API Endpoints
The backend exposes all functionality through REST endpoints:
```
# Core anomaly detection
POST /api/embed Upload video, get embedding
POST /api/score Score embedding against baseline
POST /api/detect ... | [
0.04737934470176697,
-0.03797747939825058,
-0.04411769285798073,
-0.04957859218120575,
0.022097021341323853,
-0.018957531079649925,
0.06174455210566521,
0.05225355178117752,
0.06552130728960037,
-0.01298553217202425,
0.0703481063246727,
0.062010787427425385,
0.0322863832116127,
-0.02577621... |
dbd9cbaf-fa43-570e-bfd8-2b12c5a2448e | /documentation/embeddings/upstage/ | Upstage | documentation | false | Upstage
[Upstage](https://console.upstage.ai/docs/capabilities/embed) provides Solar Embedding models as a pair: a passage model for the documents you store and a query model for the searches you run. Both map text into the same vector space. The Upstage API is compatible with the OpenAI API, so your application calls... | [
0.041566502302885056,
-0.04083273559808731,
-0.036693304777145386,
-0.04261505603790283,
0.0681295245885849,
-0.026487259194254875,
0.062058623880147934,
0.03975982218980789,
0.023316221311688423,
-0.020134376361966133,
0.0238933302462101,
-0.02352180890738964,
0.04492770507931709,
-0.0491... | |
dbe36859-a04e-5084-b004-305641a4a089 | /articles/late-interaction-models/ | Late Interaction Retrieval with Dense Token Embeddings | Understanding Embedding Models | article | false | Late Interaction Retrieval with Dense Token Embeddings > Understanding Embedding Models
The inner workings of embedding models might be surprising to some. The model doesn’t operate directly on the input text; instead, it requires a tokenization step to convert the text into a sequence of token identifiers. Each token... | [
0.040938880294561386,
-0.02830221876502037,
-0.04493463411927223,
-0.03493235632777214,
0.04008426517248154,
-0.04537754878401756,
0.06294280290603638,
0.04951805993914604,
0.0664597600698471,
-0.016705650836229324,
-0.0049109249375760555,
0.0100087346509099,
0.03481726348400116,
-0.008432... |
dbf12ede-80b6-5d10-8b1d-ac09c1700c09 | /articles/neural-search-tutorial/ | Neural Search 101: A Complete Guide and Step-by-Step Tutorial | Step-by-step neural search tutorial using Qdrant > Step 3: Upload data to Qdrant | article | true | Neural Search 101: A Complete Guide and Step-by-Step Tutorial > Step-by-step neural search tutorial using Qdrant > Step 3: Upload data to Qdrant
Now once we have the vectors prepared and the search engine running, we can start uploading the data.
To interact with Qdrant from python, I recommend using an out-of-the-box... | [
0.03873848170042038,
-0.0753854364156723,
-0.03653576225042343,
-0.005720974877476692,
0.06832916289567947,
-0.03159479796886444,
0.0857890322804451,
0.06076871603727341,
0.12295543402433395,
-0.029803847894072533,
0.016235394403338432,
0.03227992728352547,
0.04326438903808594,
-0.01420925... |
dc0406e6-41bf-565d-9630-5332de1cd26f | /documentation/tutorials-build-essentials/data-ingestion-beginners/ | S3 Ingestion | Step 2: Add Sample Data to the Bucket | documentation | true | S3 Ingestion > Step 2: Add Sample Data to the Bucket
The pipeline expects one folder per product, each with a `review.txt` and a `product.png`:
```text
product-dataset/
├── p_1/
│ ├── review.txt
│ └── product.png
├── p_2/
│ ├── review.txt
│ └── product.png
└── ...
```
If you already have product data, upload... | [
-0.007343836594372988,
-0.042885150760412216,
-0.06711176782846451,
-0.0758504569530487,
0.06657203286886215,
-0.018460355699062347,
0.0472538135945797,
0.0605606846511364,
0.04891081899404526,
-0.01801939867436886,
0.030937623232603073,
0.0393042154610157,
0.07122088968753815,
-0.03521680... |
dc090e85-90fd-52de-8ebf-5789fa959a8d | /course/beginners/module-0/building-simple-vector-search/ | Implementing a Basic Vector Search | Step 3: Connect to Qdrant Cloud | course | true | Implementing a Basic Vector Search > Step 3: Connect to Qdrant Cloud
Use the cluster URL and API key from the previous lesson. If you saved them in a `.env` file, this reads them automatically:
```python
import os
client = QdrantClient(url=os.getenv("QDRANT_URL"), api_key=os.getenv("QDRANT_API_KEY"))
# For Colab:
#... | [
0.035325516015291214,
-0.07287552952766418,
-0.020540468394756317,
-0.017252879217267036,
0.06675238162279129,
-0.02895919419825077,
0.05499960854649544,
0.02722773142158985,
0.06870745867490768,
-0.03920533508062363,
0.0011328966356813908,
0.04056406021118164,
0.0606255829334259,
-0.01535... |
dc326c87-0f5b-54da-8c8f-f24fe13ac60a | /course/essentials/day-1/movie-search-system/ | Demo: Semantic Movie Search | Step 4: Processing and Uploading the Data | course | true | Demo: Semantic Movie Search > Step 4: Processing and Uploading the Data
For each movie description, we apply all three chunking strategies, embed the resulting chunks, and store them with their respective vector names:
```python
points = []
idx = 0
for movie in movies_data: # Process each movie
# Fixed-size chu... | [
0.04639220982789993,
-0.06513666361570358,
-0.03701246902346611,
-0.03907078877091408,
0.05782347545027733,
-0.026469601318240166,
0.025648269802331924,
0.011526575312018394,
0.07726123183965683,
-0.012575553730130196,
0.0070029497146606445,
0.020032674074172974,
0.10158605128526688,
-0.04... |
dc3749bc-d3e8-5a55-a307-84bfee6aaf82 | /documentation/edge/ | Qdrant Edge | Qdrant Edge Shard | documentation | false | Qdrant Edge > Qdrant Edge Shard
Qdrant Edge is built around the concept of an **Edge Shard**: a self-contained storage unit that can operate independently. Each Edge Shard manages its own data, including vector and payload storage, and can perform local search and retrieval operations.
 is an integration platform to connect applications, data, and devices across on-premises and cloud environments. It provides a unified platform to build, manage, and secure APIs and integrations, making digital transform... | [
0.04597075656056404,
-0.0743299350142479,
-0.03305410221219063,
-0.03921735659241676,
0.01427945215255022,
-0.03693421930074692,
0.07168298214673996,
0.012505542486906052,
0.04494098946452141,
-0.03214718773961067,
-0.014948010444641113,
0.021331211552023888,
-0.008045567199587822,
-0.0610... | |
dcc9d5c4-be0b-5764-81d2-664d6c9e792d | /documentation/edge/edge-bm25/ | BM25 | Embed and Upsert Documents | documentation | true | BM25 > Embed and Upsert Documents
Use `embed_document` to generate a sparse vector for each document, then upsert the points. Call `optimize` after bulk inserts to build the sparse index:
```python
from qdrant_edge import Point, UpdateOperation
shard.update(UpdateOperation.upsert_points([
Point(1, {"text": bm25.... | [
0.05471009016036987,
-0.04718036949634552,
-0.008244365453720093,
-0.02949264459311962,
0.07633132487535477,
-0.03538931533694267,
0.08039326220750809,
0.04648639261722565,
0.023412276059389114,
0.00045896938536316156,
0.014294853433966637,
0.029861656948924065,
0.032587889581918716,
-0.02... |
dcd513ab-278a-5862-a116-036ff92f0a5a | /documentation/examples/qdrant-dspy-medicalbot/ | Building a Chain-of-Thought Medical Chatbot with Qdrant and DSPy | Conclusion | documentation | false | Building a Chain-of-Thought Medical Chatbot with Qdrant and DSPy > Conclusion
By leveraging Qdrant and DSPy, you can build a medical chatbot that generates accurate and up-to-date medical responses. Qdrant provides the technology and enables fast and scalable retrieval, while DSPy synthesizes this information to provi... | [
0.016950897872447968,
-0.020870117470622063,
-0.04155606031417847,
-0.02901824563741684,
0.060666825622320175,
-0.04509366303682327,
0.10492300987243652,
0.027822941541671753,
0.04422704130411148,
-0.03037947602570057,
0.04092317447066307,
0.011368225328624249,
0.06224733963608742,
-0.0469... |
dcf4d49d-adae-55d4-975d-b50b288236b7 | /documentation/tutorials-build-essentials/semantic-cache/ | Semantic Caching for RAG | Set Up Cache Lookup > Calibrate the Threshold | documentation | true | Semantic Caching for RAG > Set Up Cache Lookup > Calibrate the Threshold
1. Split the golden set into three parts: "train", "val", and "test". Split by cached question, so that rewordings of one question never land in different parts.
2. On train, for every candidate value, count the correct and the wrong hits. Decide... | [
0.0542542040348053,
-0.0355217345058918,
-0.05111243575811386,
-0.004369477741420269,
0.046216804534196854,
0.0022116550244390965,
0.03217235952615738,
0.027581969276070595,
0.04217294976115227,
-0.021379906684160233,
0.018496379256248474,
0.005604678299278021,
0.07269969582557678,
-0.0389... |
dcf73a5a-eeb3-5074-8695-2156145320ca | /documentation/ops-monitoring/monitoring/ | Monitoring & Telemetry | Metrics > Metrics Configuration | documentation | true | Monitoring & Telemetry > Metrics > Metrics Configuration
*Available as of v1.16.0*
In self-hosted environments you have further configuration options for metrics.
By default, all Qdrant metrics have no application namespace prefix. You may set
a prefix with `service.metrics_prefix` in the
[configuration](/documentat... | [
0.029516752809286118,
-0.01131161767989397,
-0.006129342131316662,
0.010848868638277054,
0.05894869193434715,
-0.08298139274120331,
0.03819408640265465,
0.06923781335353851,
0.04508545994758606,
-0.011557838879525661,
0.009678389877080917,
-0.031632766127586365,
0.07224217802286148,
-0.032... |
dd030c3a-0276-5793-bcca-ffebc75061e4 | /documentation/search/text-search/text-filtering/ | Text Filtering | documentation | false | Text Filtering
Qdrant supports [filtering](/documentation/search/filtering/) on a wide range of datatypes: numbers, dates, booleans, geolocations, and strings. In Qdrant, a filter is typically combined with a vector query. The vector query is used to score and rank the results, while the filter is used to narrow down ... | [
0.03565758094191551,
-0.04204793646931648,
-0.04555032029747963,
-0.03836105391383171,
0.028821762651205063,
-0.015588635578751564,
0.08831787109375,
0.04832635074853897,
0.08680409938097,
-0.012536211870610714,
0.04279240593314171,
0.012339792214334011,
0.07890240848064423,
-0.01978656649... | |
dd045a62-90ee-5cc9-a870-eaba6d5a64f3 | /articles/modern-sparse-neural-retrieval/ | Modern Sparse Neural Retrieval: From DeepCT to SPLADE++ | Did we Solve the Vocabulary Mismatch Yet? | article | false | Modern Sparse Neural Retrieval: From DeepCT to SPLADE++ > Did we Solve the Vocabulary Mismatch Yet?
With the retrieval based on the exact matching,
however sophisticated the methods to predict term importance are, we can’t match relevant documents which have no query terms in them.
If you’re searching for “*pizza*” ... | [
0.05700937658548355,
-0.0499827116727829,
-0.03313765302300453,
-0.030591100454330444,
0.0644829273223877,
-0.03725675120949745,
0.041529685258865356,
0.002411567373201251,
0.092160165309906,
-0.04050479829311371,
0.0015869110357016325,
0.03957177326083183,
0.04672545567154884,
-0.02779262... |
dd0fd838-9ddf-515e-9073-09b0b0866360 | /articles/hyperbolic-embeddings-qdrant/ | Why Hyperbolic Embeddings Matter for Hierarchical Data | References | article | false | Why Hyperbolic Embeddings Matter for Hierarchical Data > References
1. Nickel, M. and Kiela, D. (2017). [Poincaré Embeddings for Learning Hierarchical Representations](https://arxiv.org/abs/1705.08039). WordNet reconstruction results: Table 1.
2. Mahmood, M. et al. (2026). [Hyper3-CLIP: Hierarchy-Conditioned Hyperboli... | [
0.05256102606654167,
0.011195083148777485,
-0.05102129653096199,
-0.043767716735601425,
0.07411964237689972,
-0.043874502182006836,
0.09054754674434662,
0.04852571710944176,
0.06436081975698471,
-0.03383169695734978,
0.033993709832429886,
0.026861991733312607,
0.07266208529472351,
-0.06454... |
dd5d6da4-f1f7-518e-81fa-f2a519130380 | /documentation/manage-data/collections/ | Collections | Collection Info | documentation | true | Collections > Collection Info
Qdrant allows determining the configuration parameters of an existing collection to better understand how the points are
distributed and indexed.
```python
client.get_collection(collection_name="{collection_name}")
```
Expected result
```json
{
"result": {
"status": "green"... | [
0.012653909623622894,
-0.06196821480989456,
0.007006736937910318,
-0.05886192247271538,
0.058513518422842026,
-0.08171063661575317,
0.043129775673151016,
0.03423536568880081,
0.0920899510383606,
-0.07089128345251083,
0.010708164423704147,
0.02548561431467533,
0.041698165237903595,
-0.03875... |
dd6be9a6-e462-5fa1-b583-9da80b2c55cb | /documentation/faq/qdrant-fundamentals/ | Qdrant Fundamentals | documentation | false | Qdrant Fundamentals
|||||||
|-|-|-|-|-|-|
|[Vectors](/documentation/faq/qdrant-fundamentals/#vectors)|[Search](/documentation/faq/qdrant-fundamentals/#search)|[Collections](/documentation/faq/qdrant-fundamentals/#collections)|[Compatibility](/documentation/faq/qdrant-fundamentals/#compatibility)|[Security](/documentat... | [
0.026651134714484215,
-0.0062124780379235744,
-0.036889027804136276,
-0.04438333213329315,
0.01493798941373825,
-0.07198013365268707,
0.06617075949907303,
0.06331625580787659,
0.10467683523893356,
-0.05789676308631897,
0.028635459020733833,
0.03745422884821892,
0.05578308925032616,
-0.0439... | |
dd77e286-917d-514a-bd80-c00576c636aa | /documentation/search-evaluation/pipeline-output-quality/ | Evaluating Pipeline Output Quality | Non-RAG Use Cases | documentation | false | Evaluating Pipeline Output Quality > Non-RAG Use Cases
Ragas's metrics assume the consumer is an LLM generator. If retrieval feeds something else (a ranker, a recommendation surface, an agent, a search UI), swap the metrics to match: CTR or dwell time for a UI, graded rubrics for a ranker, task-completion rate for an ... | [
0.007666471879929304,
-0.02012624591588974,
-0.04047118499875069,
-0.03943273797631264,
0.0567021518945694,
-0.021383464336395264,
0.09342920780181885,
0.0445530042052269,
0.05281250923871994,
-0.0209910087287426,
0.0230557844042778,
0.004529107362031937,
0.07602117955684662,
-0.0245574563... |
dd7fd2ea-7757-56e8-a809-b8fa5132c0cd | /course/essentials/day-3/hybrid-search/ | Hybrid Search and the Universal Query API | What You'll Learn | course | false | Hybrid Search and the Universal Query API > What You'll Learn
- Understand when to use dense vs. sparse vectors
- Build hybrid search pipelines with Qdrant's Universal Query API
- Apply Reciprocal Rank Fusion (RRF) to combine results
- Design multi-stage retrieval and reranking strategies | [
0.029408978298306465,
-0.03394296020269394,
-0.06398577243089676,
-0.05481033772230148,
0.08267883211374283,
-0.04758166894316673,
0.06971656531095505,
0.0362500362098217,
0.05477333441376686,
-0.028571318835020065,
0.046895790845155716,
0.01726524718105793,
0.06637724488973618,
-0.0052631... |
dd9ce0e2-fabe-5eed-9e01-4fc23baec97f | /documentation/tutorials-search-engineering/index-dynamic-payloads/ | Indexing Payloads of Random Shape | Optimize Exact Match by Concatenating key=value | documentation | true | Indexing Payloads of Random Shape > Optimize Exact Match by Concatenating key=value
For categorical attributes that only ever need exact match, you can store each as a single `key=value` keyword term in a flat string array, indexed once.
```python
# Payload: {"attrs_flat": ["vendor=acme", "screen_size=27in"]}
client.... | [
0.02144569531083107,
0.004633746109902859,
-0.05036096274852753,
-0.00969290267676115,
0.03625316172838211,
-0.06173037365078926,
0.026857376098632812,
0.011094649322330952,
0.10563183575868607,
-0.03698110580444336,
0.004917353391647339,
0.0166680496186018,
0.060884274542331696,
-0.012661... |
dd9e9b32-5914-5893-bec3-0829380d0fa5 | /documentation/scaling/ | Scaling & Resilience | Where to Go Next | documentation | false | Scaling & Resilience > Where to Go Next
- [Vertical Scaling](/documentation/scaling/vertical-scaling/): resize existing nodes.
- [Horizontal Scaling](/documentation/scaling/horizontal-scaling/): how Qdrant's distributed model achieves scale.
- [Resilience](/documentation/scaling/resilience/): fault tolerance, multiple... | [
0.05210159346461296,
-0.015439596958458424,
-0.05086871236562729,
-0.03918468952178955,
0.06968546658754349,
-0.03131329268217087,
0.0845242515206337,
0.03111932799220085,
0.06440883874893188,
-0.017698140814900398,
0.0512293204665184,
-0.0035085335839539766,
0.04719226062297821,
-0.018937... |
dda8d369-afee-573c-b5c2-e474591a5b41 | /documentation/cloud/configure-cluster/ | Configure Clusters | Restart Mode | documentation | false | Configure Clusters > Restart Mode
The cloud platform will automatically choose the optimal restart mode during version upgrades or maintenance for your cluster. If you have a multi-node cluster and one or more collections with a replication factor of at least 2, the cloud platform will use the rolling restart mode. Th... | [
0.00552601832896471,
-0.021318510174751282,
-0.04409736022353172,
-0.037616029381752014,
0.09465919435024261,
-0.04190702736377716,
0.03887759894132614,
0.04825104400515556,
0.06529008597135544,
-0.03601355478167534,
0.009959610179066658,
0.02112525701522827,
0.0092700757086277,
-0.0223580... |
ddab3e4e-62f7-565b-8565-e4b3f847d81b | /documentation/private-cloud/api-reference/ | API Reference | qdrant.io/v1 > Resource Types > QdrantConfigurationCollectionStrictMode | documentation | false | API Reference > qdrant.io/v1 > Resource Types > QdrantConfigurationCollectionStrictMode
_Appears in:_
- [QdrantConfigurationCollection](#qdrantconfigurationcollection)
| Field | Description | Default | Validation |
| --- | --- | --- | --- |
| `max_payload_index_count` _integer_ | MaxPayloadIndexCount represents the m... | [
0.008812760934233665,
-0.019868426024913788,
-0.020607277750968933,
0.012984496541321278,
0.06434889137744904,
-0.061912428587675095,
0.03582988679409027,
0.0671791136264801,
0.08424200117588043,
-0.0756295695900917,
0.03950131684541702,
0.01469529140740633,
0.04178028181195259,
-0.0568290... |
ddb9c8ce-3878-5995-bb77-a995b1fe6786 | /course/essentials/day-3/hybrid-search/ | Hybrid Search and the Universal Query API | Implementing RRF in Qdrant | course | true | Hybrid Search and the Universal Query API > Implementing RRF in Qdrant
If you would like to implement that process in Qdrant, then you don't have to calculate the RRF scores on your own. It's just built in!
This is an API call that you would have to make in order to retrieve twenty documents from both dense and spars... | [
0.042787373065948486,
-0.04438737407326698,
-0.04034464806318283,
-0.05346085876226425,
0.08861988037824631,
-0.04733273759484291,
0.04763198643922806,
0.04617198556661606,
0.06903359293937683,
-0.05263493210077286,
0.03219771757721901,
0.013299421407282352,
0.048744797706604004,
-0.009947... |
ddd54f85-3140-5799-a5bd-6eda969e90a5 | /documentation/tutorials-operations/secure-and-expose-hybrid-cloud/ | Securing and Exposing a Hybrid Cloud Cluster | Expose the Cluster Through an Ingress | documentation | true | Securing and Exposing a Hybrid Cloud Cluster > Expose the Cluster Through an Ingress
Qdrant now speaks TLS, but to reach it from outside the Kubernetes cluster, we need to put an ingress controller in front of it, terminating or passing through TLS at the edge. This tutorial uses [Traefik](https://traefik.io/traefik),... | [
0.057849470525979996,
-0.037690430879592896,
-0.06403213739395142,
-0.03314227610826492,
0.06843434274196625,
-0.053661130368709564,
0.06453292816877365,
0.023584982380270958,
0.04672813042998314,
-0.02852213941514492,
0.02126481384038925,
0.018522946164011955,
0.03394489362835884,
-0.0064... |
de1001d8-fecc-5c80-8c8d-fb0a4199d8fb | /documentation/production-operations/vector-search-in-production/ | Vector Search in Production: Scaling, HA & Tuning Guide | 2. How do I Ingest and Index Large Amounts of Data? > How to Arrange Metadata and Schema for Consistency | documentation | false | Vector Search in Production: Scaling, HA & Tuning Guide > 2. How do I Ingest and Index Large Amounts of Data? > How to Arrange Metadata and Schema for Consistency
||
|:-:|
|**"My filters aren't working the same way every time."**|
In some cases, the payload schema is inconsistent across data pipelines, so some fields... | [
0.03745877370238304,
-0.04291146248579025,
-0.08085428923368454,
-0.02535746619105339,
0.037077270448207855,
-0.0028630902525037527,
0.045977335423231125,
0.02165951579809189,
0.07122521847486496,
-0.02660992555320263,
0.02094687893986702,
0.019104812294244766,
0.03725589066743851,
-0.0452... |
de139735-f7e9-5071-af03-afb81d0d3919 | /documentation/search-tuning/when-your-collection-outgrows-ram/ | When Your Collection Outgrows RAM | Set Quantization in Four Steps | documentation | true | When Your Collection Outgrows RAM > Set Quantization in Four Steps
1. Estimate the resident footprint of your dense vectors, and how far it overshoots the memory you have.
2. Choose the quantization method and bit depth, starting from `bits4`.
3. Pin the quantized copy and leave the original vectors `cold`.
4. Set `re... | [
0.03904934599995613,
-0.07671016454696655,
-0.024815289303660393,
-0.03076118417084217,
0.08257205039262772,
-0.026364678516983986,
0.04528341442346573,
0.044776275753974915,
0.10637584328651428,
-0.0235340166836977,
0.021283652633428574,
0.0050534592010080814,
0.03616495430469513,
-0.0211... |
de23c56a-617a-5692-baf9-71393aefeca6 | /documentation/embeddings/nomic/ | Nomic | Reduce Vector Size | documentation | true | Nomic > Reduce Vector Size
`nomic-embed-text-v1.5` was trained with Matryoshka representation learning. To store shorter vectors, pass `dimensionality` with one of the Matryoshka sizes and set the collection's `size` to match. Pass the same `dimensionality` when you embed queries:
```python
output = embed.text(
t... | [
0.05498726665973663,
-0.006329371593892574,
-0.03115563839673996,
-0.06898826360702515,
0.03680716082453728,
-0.0569455511868,
0.047896429896354675,
0.03436993062496185,
0.06319364905357361,
-0.022754687815904617,
0.0694175511598587,
0.026683233678340912,
0.06664935499429703,
-0.0257006119... |
de7b0902-eb1f-5fd7-9d05-bb753d5c9d01 | /documentation/edge/edge-quickstart/ | Quickstart | Load Existing Edge Shard from Disk | documentation | true | Quickstart > Load Existing Edge Shard from Disk
After closing an Edge Shard, you can reopen it by loading its data and configuration from disk using the [`load`](/documentation/edge/edge-api/shard-lifecycle/#load-an-existing-edge-shard) method:
```python
edge_shard = EdgeShard.load(SHARD_DIRECTORY)
``` | [
0.030863158404827118,
-0.03560065105557442,
-0.013561559841036797,
-0.09022983908653259,
0.08546440303325653,
-0.05787564814090729,
0.018939308822155,
0.027226712554693222,
0.09155619144439697,
-0.023704633116722107,
0.04439180716872215,
0.04241009056568146,
0.04483168572187424,
-0.0037124... |
de8e70c9-35a3-5627-b896-935976c844ec | /course/beginners/module-6/grouping/ | Grouping: One Slot per Document | course | false | Grouping: One Slot per Document
Module 6
Chunking creates the neighbor problem: one long document becomes many points, and a strong match on it can fill the whole first page with its own chunks. [`query_points_groups`](/documentation/search/search/#grouping-api) groups results by a payload field and returns a set num... | [
0.014420497231185436,
-0.04430071637034416,
-0.05286484584212303,
-0.02695509046316147,
0.07829773426055908,
-0.01623673364520073,
0.06542393565177917,
0.023295823484659195,
0.10049694776535034,
-0.03475666791200638,
0.026544194668531418,
0.010404253378510475,
0.02258787490427494,
-0.02164... | |
deb23703-513e-5aff-ba18-cebde31f4fca | /documentation/cloud/configure-cluster/ | Configure Clusters | documentation | false | Configure Clusters
Qdrant Cloud offers several advanced configuration options to optimize clusters for your specific needs. You can access these options from the Cluster Details page in the Qdrant Cloud console.
The cloud platform does not expose all [configuration options](/documentation/ops-configuration/configurat... | [
0.025817958638072014,
-0.009996116161346436,
-0.05678548291325569,
-0.0036895335651934147,
0.06689398735761642,
-0.05202155560255051,
0.020316608250141144,
0.05456244572997093,
0.07684958726167679,
-0.058619920164346695,
0.01070417370647192,
0.02909010276198387,
0.04445185884833336,
-0.059... | |
dec93c2e-4de1-5a71-9ae6-e01690e5dbc6 | /documentation/hybrid-cloud/platform-deployment-options/ | Deployment Platforms | documentation | false | Deployment Platforms
This page provides an overview of how to deploy Qdrant Hybrid Cloud on various managed Kubernetes platforms.
For a general list of prerequisites and installation steps, see our [Hybrid Cloud setup guide](/documentation/hybrid-cloud/hybrid-cloud-setup/). This platform specific documentation also a... | [
-0.0025634337216615677,
-0.030980220064520836,
-0.06751835346221924,
-0.07388336211442947,
0.04930954426527023,
-0.04377252236008644,
0.05264957249164581,
0.031632695347070694,
0.04778224229812622,
-0.02370869554579258,
0.026317793875932693,
0.004790500737726688,
0.05323034152388573,
-0.03... | |
df10a727-a910-5c31-8259-8732c754801a | /articles/bm42/ | BM42: Attention-Based Sparse Embeddings for Hybrid Search | Why has BM25 stayed relevant for so long? | article | false | BM42: Attention-Based Sparse Embeddings for Hybrid Search > Why has BM25 stayed relevant for so long?
The last component of the formula can be intuitively interpreted as **term importance within the document**.
This might look a bit complicated, so let's break it down.
$$
\text{Term importance in document }(q_i) = \... | [
0.02154974639415741,
-0.0059080119244754314,
-0.005790247581899166,
-0.08176660537719727,
0.06102817878127098,
-0.04687073081731796,
0.05523368716239929,
0.04540155082941055,
0.12079182267189026,
-0.01562182605266571,
-0.003609110601246357,
0.004268115386366844,
0.0622878335416317,
-0.0370... |
df146662-d983-5929-bb92-a4453edfa5fd | /documentation/embeddings/cohere/ | Cohere | Embed Queries | documentation | true | Cohere > Embed Queries
Set `input_type` to `search_query` for search queries. The input type tells the model how a vector will be used, so pair `search_document` for stored content with `search_query` for queries. This example embeds the query with `embed-v5.0-fast`, which shares the embedding space of `embed-v5.0-pro... | [
0.04952507093548775,
-0.06556440889835358,
-0.052389007061719894,
-0.02008393034338951,
0.05893167108297348,
-0.04723257198929787,
0.05641087517142296,
0.06053614616394043,
0.03900231420993805,
-0.010933361016213894,
0.03203360736370087,
0.02138160541653633,
0.043673075735569,
-0.037236053... |
df5bdec2-b4f8-51ee-aa00-4d6b7bb170dd | /articles/bm42/ | BM42: Attention-Based Sparse Embeddings for Hybrid Search | Practical examples > Benchmarks | article | false | BM42: Attention-Based Sparse Embeddings for Hybrid Search > Practical examples > Benchmarks
Please note, that both BM25 and BM42 won't work well on their own in a production environment.
Best results are achieved with a combination of sparse and dense embeddings in a hybrid approach.
In this scenario, the two models a... | [
0.04770328477025032,
-0.015250785276293755,
-0.03958408534526825,
-0.06694065779447556,
0.04243379086256027,
-0.06059484928846359,
0.0685039833188057,
0.05852342024445534,
0.027174588292837143,
-0.01170451007783413,
0.057227157056331635,
0.02885379269719124,
0.0376100167632103,
-0.02324495... |
df6aa081-5dfb-59b7-aa88-2b7a070c17b0 | /documentation/frameworks/ | Frameworks | Framework Integrations | documentation | false | Frameworks > Framework Integrations
| Framework | Description |
| ------------------------------------------------------------- | --------------------------------... | [
0.02377375029027462,
-0.00836594495922327,
-0.06138771027326584,
-0.0024769455194473267,
0.09432660043239594,
-0.04959753900766373,
0.057595424354076385,
0.07095026969909668,
0.01941591687500477,
-0.05106600373983383,
0.020957686007022858,
0.0026399490889161825,
0.04298003017902374,
-0.029... |
dfab9a78-4e2d-517d-a5a8-616192b66c7a | /documentation/tutorials-operations/embedding-model-migration/ | Migrate to a New Embedding Model | Migrate Using Named Vectors > Step 5: Disable Dual Writes and Delete the Old Named Vector | documentation | true | Migrate to a New Embedding Model > Migrate Using Named Vectors > Step 5: Disable Dual Writes and Delete the Old Named Vector
Once all search traffic uses the new vector, keep dual writes running for an observation period. Until you delete the old vector, you can roll back by switching the `using` parameter and the emb... | [
0.05274930223822594,
-0.04813266545534134,
-0.03316597267985344,
0.01292650680989027,
0.02665429748594761,
-0.01562519371509552,
0.0311430711299181,
0.05327821522951126,
0.04156982898712158,
-0.025243204087018967,
0.033414412289857864,
0.03523871675133705,
0.03423687815666199,
-0.032455883... |
dfc91d71-809f-54f0-a6d6-f036ed66e3ff | /documentation/tutorials-operations/prevent-unoptimized-usage/ | Prevent Unoptimized Usage | Prerequisites | documentation | true | Prevent Unoptimized Usage > Prerequisites
Install the Qdrant client, plus `huggingface-hub` and `polars` to download and process the dataset used in this tutorial.
```bash
pip install -q qdrant-client huggingface-hub polars
```
Create a [Free Tier Qdrant Cloud cluster](https://cloud.qdrant.io/) and instantiate an as... | [
0.020168641582131386,
-0.020199719816446304,
-0.03677923604846001,
-0.010129150934517384,
0.06538476049900055,
-0.04831033945083618,
0.06508587300777435,
0.011723152361810207,
0.05965818837285042,
-0.005803160835057497,
0.022587420418858528,
0.025943364948034286,
0.054877184331417084,
-0.0... |
dfd00d07-95e4-54ac-b23c-da5013ab2e91 | /documentation/private-cloud/api-reference/ | API Reference | qdrant.io/v1 > Resource Types > QdrantClusterSnapshotList | documentation | false | API Reference > qdrant.io/v1 > Resource Types > QdrantClusterSnapshotList
QdrantClusterSnapshotList contains a list of QdrantClusterSnapshot
| Field | Description | Default | Validation |
| --- | --- | --- | --- |
| `apiVersion` _string_ | `qdrant.io/v1` | | |
| `kind` _string_ | `QdrantClusterSnapshotList` | | |
| `... | [
0.0327417254447937,
-0.016369333490729332,
-0.021827755495905876,
-0.05751582235097885,
0.05264471471309662,
-0.06366289407014847,
0.06413670629262924,
0.06355531513690948,
0.058402497321367264,
-0.06360600143671036,
0.0643947571516037,
0.01969415321946144,
0.04364650323987007,
0.009186136... |
dfff78ac-640a-5f96-a8a0-19ce0bda189b | /documentation/manage-data/vectors/ | Vectors | Vector Types | documentation | false | Vectors > Vector Types
Modern neural networks can output vectors in different shapes and sizes, and Qdrant supports most of them.
Let's take a look at the most common types of vectors supported by Qdrant. | [
0.015026590786874294,
-0.024935593828558922,
-0.05996830761432648,
-0.05990497022867203,
0.018247289583086967,
-0.04985241964459419,
0.0890374407172203,
0.025617344304919243,
0.12453015148639679,
-0.0179576575756073,
0.05257809907197952,
0.029181841760873795,
0.07558874785900116,
-0.033315... |
e00384f0-fc6e-5abe-bc6e-409f81eaf161 | /course/multi-vector-search/module-2/colpali-family/ | ColPali Family Overview | Model Size | course | false | ColPali Family Overview > Model Size
While the original ColPali model delivers good performance, its multi-billion parameter size can be challenging for resource-constrained environments, demos, or CPU-only deployments. Fortunately, smaller alternatives maintain competitive performance while dramatically reducing comp... | [
0.005938833113759756,
0.007380415685474873,
-0.019349584355950356,
-0.049807511270046234,
0.09591969847679138,
-0.02367125265300274,
0.08223367482423782,
0.028218401595950127,
0.041822340339422226,
-0.05276752635836601,
0.05623748153448105,
0.007399358786642551,
0.03340322896838188,
-0.019... |
e00eac46-97ff-5804-bb84-a4c594040921 | /documentation/scaling/distributed_deployment/ | Distributed Deployment | Listener Mode | documentation | true | Distributed Deployment > Listener Mode
This is an experimental feature, its behavior may change in the future.
In some cases it might be useful to have a Qdrant node that only accumulates data and does not participate in search operations.
There are several scenarios where this can be useful:
- Listener option can b... | [
0.04431384056806564,
-0.04176374897360802,
-0.050848521292209625,
-0.009616252966225147,
0.06402206420898438,
-0.031245268881320953,
0.037789393216371536,
0.029102524742484093,
0.05757256969809532,
0.009972420521080494,
0.022900424897670746,
0.026168696582317352,
0.04527340456843376,
0.015... |
e013df4a-e9a4-5d91-a3fb-2dbce9e83f7d | /documentation/examples/graphrag-qdrant-neo4j/ | GraphRAG with Qdrant and Neo4j | Implementation > Retrieval & Generation > Integrating with the LLM | documentation | true | GraphRAG with Qdrant and Neo4j > Implementation > Retrieval & Generation > Integrating with the LLM
Now that we have the graph context, we need to generate a prompt for the LLM. This is where the core of the Retrieval-Augmented Generation (RAG) happens: we combine the graph data and the user query into a comprehensive... | [
0.002760383067652583,
-0.050638020038604736,
-0.0530768446624279,
-0.061616700142621994,
0.0858456939458847,
-0.036133285611867905,
0.06820667535066605,
0.06545750051736832,
0.04624878987669945,
-0.052709806710481644,
0.0420532263815403,
0.007873083464801311,
0.07883742451667786,
-0.028453... |
e02590cb-c57e-59aa-983f-6b1a00b73e78 | /course/beginners/module-5/project-overview/ | Project Overview | course | false | Project Overview
Module 5
A factory fire at a supplier's plant reaches you four ways. A local news report, a satellite image, an earnings call where an executive gets asked about it, and the supplier's own filing weeks later. You are building the system an analyst uses to see all four, and none of them arrives labele... | [
0.02584575116634369,
-0.026744818314909935,
-0.05464193969964981,
-0.029028503224253654,
0.10006722062826157,
-0.03848477825522423,
0.025218063965439796,
0.05634615197777748,
0.06021709740161896,
-0.008584761060774326,
0.03391368314623833,
0.020291287451982498,
0.06408876925706863,
-0.0217... | |
e0627a8c-4718-520a-82c2-4fd1bdd1ffc4 | /documentation/migrate-to-qdrant/from-elasticsearch/ | From Elasticsearch | Hybrid Search Considerations | documentation | false | From Elasticsearch > Hybrid Search Considerations
If your Elasticsearch setup uses hybrid BM25 and kNN scoring, you can reconstruct this in Qdrant with [BM25 on a full-text index](/documentation/search/text-search/full-text-search/#bm25) alongside dense vectors. The migration tool transfers the dense vectors and docum... | [
0.06195454299449921,
-0.037255920469760895,
-0.062170784920454025,
-0.04132363945245743,
0.04355437308549881,
-0.045695431530475616,
0.0793248862028122,
0.03754080832004547,
0.04465235769748688,
-0.01168532483279705,
0.05521363392472267,
0.012166897766292095,
0.06059147045016289,
-0.011827... |
e073d86e-1d4f-5985-9749-c18d683b9bff | /articles/what-is-a-vector-database/ | What is a Vector Database? | Hybrid Search > Dense vs. Sparse Vectors > 1. Dense Vectors | article | false | What is a Vector Database? > Hybrid Search > Dense vs. Sparse Vectors > 1. Dense Vectors
Dense vectors are, quite literally, dense with information. Every element in the vector contributes to the **semantic meaning**, **relationships** and **nuances** of the data. A dense vector representation of this sentence might l... | [
0.014342409558594227,
-0.04488394409418106,
-0.07202617079019547,
-0.06465651094913483,
0.032556477934122086,
-0.03289239853620529,
0.06299261003732681,
0.06953492015600204,
0.06749150156974792,
-0.06074086204171181,
0.031471800059080124,
0.06343840807676315,
0.06198723241686821,
0.0238717... |
e074ede9-d935-5989-8429-d6586776cd77 | /documentation/security/ | Security | Authentication > Read-Only API Key | documentation | true | Security > Authentication > Read-Only API Key
*Available as of v1.7.0*
Qdrant also supports a read-only API key.
This key can be used to access read-only operations on the instance.
```yaml
service:
read_only_api_key: your_secret_read_only_api_key_here
```
Or with the environment variable:
```bash
export QDRANT_... | [
0.03540724143385887,
-0.04000585153698921,
-0.04666123539209366,
-0.04779371619224548,
0.04340033233165741,
-0.03921663016080856,
0.032052408903837204,
0.05449363589286804,
0.06509647518396378,
-0.009353588335216045,
0.017997104674577713,
0.016621051356196404,
0.07307801395654678,
-0.00792... |
e07c1812-5ec6-54ee-bfb5-c26b8338a141 | /articles/universal-io/ | Universal I/O in Qdrant | Filesystems, Files, and Pipelines | article | true | Universal I/O in Qdrant > Filesystems, Files, and Pipelines
Since reads can finish out of order when the depth is above 1, each request also carries a tag, the *user data*, which comes back with its result and tells the caller which request it answers. It works like `user_data` in `io_uring`, except that it can be any... | [
0.03067309595644474,
-0.04101143777370453,
-0.0756707713007927,
-0.015576176345348358,
0.07795891165733337,
-0.011717262677848339,
0.06163707748055458,
0.03943096101284027,
0.08182637393474579,
0.03366846963763237,
0.007820424623787403,
-0.001643797499127686,
0.04238332435488701,
-0.010810... |
e09e1dab-3a79-5b15-921f-348ba8f18ea2 | /documentation/inference/external-inference-providers/ | External Providers | documentation | false | External Providers
Qdrant Cloud can act as a proxy for the APIs of external embedding model providers:
- OpenAI
- Cohere
- Jina AI
- OpenRouter
This enables you to access any of the embedding models provided by these providers through the Qdrant API.
 and an optional [payload](/documentation/manage-data/payload/).
It looks like this:
```json
// This is a simple point
{
"id": 129,
"vector": [0.1, 0.2, 0.3, 0.... | [
0.01816270686686039,
-0.06361508369445801,
-0.05332796275615692,
-0.04348113015294075,
0.05436612665653229,
-0.00339009752497077,
0.0668715089559555,
0.07457289099693298,
0.10918162018060684,
-0.015191305428743362,
0.03406522423028946,
0.017425276339054108,
0.003109228564426303,
-0.0229635... | |
e0def89e-37dd-52b2-8ed2-089b6c4df9e7 | /course/beginners/module-3/the-two-families-of-search/ | The Two Families of Search | Sparse Search > Indexing Sparse Vectors | course | true | The Two Families of Search > Sparse Search > Indexing Sparse Vectors
Because most dimensions are zero, Qdrant stores sparse vectors in an inverted index: rather than a row per point, it keeps one list per token recording which points contain that token and its weight.
```text
Token "nike" → list: [point_1, point_2... | [
0.06489984691143036,
-0.021329576149582863,
-0.021236008033156395,
-0.03552599996328354,
0.025166058912873268,
-0.04596971720457077,
0.0811404436826706,
0.02945496328175068,
0.07059670239686966,
-0.011287663131952286,
0.024504704400897026,
0.06018207222223282,
0.044110871851444244,
0.00548... |
e0e538ec-81a1-56e6-99db-42775fb363f5 | /documentation/ops-optimization/optimizer/ | Optimizer | Prevent Reads from Large Unindexed Segments > Effect on wait=true | documentation | false | Optimizer > Prevent Reads from Large Unindexed Segments > Effect on wait=true
Write requests support a [`wait` parameter](/documentation/manage-data/points/#awaiting-result) that, when set to `true`, causes the request to return only after the update has been applied and is visible for search. If `prevent_unoptimized`... | [
0.020077969878911972,
-0.045962896198034286,
-0.04819287359714508,
-0.018487708643078804,
0.05529186502099037,
-0.026501433923840523,
0.041956376284360886,
-0.009307777509093285,
0.0682089626789093,
-0.022299623116850853,
0.04281136393547058,
0.018903367221355438,
0.0675368383526802,
-0.04... |
e0fd4d45-64a8-518f-97c7-70bd196a9ee5 | /articles/search-feedback-loop/ | Relevance Feedback in Information Retrieval | Query Refinement > Query As Text | article | false | Relevance Feedback in Information Retrieval > Query Refinement > Query As Text
In **term-based retrieval**, an intuitive way to improve a query would be to **expand it with relevant terms**. It resembled the "*aha, so that's what it's called*" stage in the discovery search.
Before the deep learning era of this centur... | [
0.04834567382931709,
-0.060403745621442795,
-0.054476942867040634,
-0.04934581741690636,
0.04691668599843979,
-0.05917276814579964,
0.04698707535862923,
0.006834015250205994,
0.08406782150268555,
-0.009694875217974186,
0.028135225176811218,
-0.002285136142745614,
0.056256841868162155,
-0.0... |
e10e2ce2-23a0-5940-87b1-9728d25f4ef0 | /documentation/data-synchronization/with-postgres/ | Keeping Postgres and Qdrant in Sync | Tier 2: Transactional Outbox Pattern > The Outbox Table | documentation | true | Keeping Postgres and Qdrant in Sync > Tier 2: Transactional Outbox Pattern > The Outbox Table
```sql
CREATE TABLE sync_outbox (
id BIGSERIAL PRIMARY KEY,
entity_id VARCHAR(20) NOT NULL, -- article_id
operation VARCHAR(10) NOT NULL, -- 'upsert' | 'delete'
payload JSONB, ... | [
0.0011732744751498103,
-0.016083527356386185,
-0.01449983473867178,
-0.06555061787366867,
0.03732810914516449,
-0.034835077822208405,
0.03244979679584503,
0.008010251447558403,
0.07606931030750275,
-0.019631199538707733,
0.033554960042238235,
0.054348744451999664,
0.08650705963373184,
-0.0... |
e1370723-0236-5633-9729-4a2395a3008b | /documentation/tutorials-build-essentials/video-anomaly-edge-part-2/ | Video Anomaly Detection Part 2: Edge-to-Cloud Pipeline | Two-Shard Edge Architecture > Query Path | documentation | true | Video Anomaly Detection Part 2: Edge-to-Cloud Pipeline > Two-Shard Edge Architecture > Query Path
Every kNN query searches **both shards simultaneously**, merges results by score, and takes the top-k:
```python
def score_local(self, embedding: np.ndarray) -> float:
req = SearchRequest(
query=Query.Nearest... | [
0.028778376057744026,
-0.039907149970531464,
-0.07079154998064041,
-0.026192380115389824,
0.04192991927266121,
-0.015746232122182846,
0.06472504884004593,
0.0026706813368946314,
0.07590661942958832,
-0.007793272379785776,
0.05359963700175285,
0.020332682877779007,
0.05574527382850647,
-0.0... |
e13c3c07-a0ec-5844-a211-9fb9c1fb6cac | /documentation/manage-data/indexing/ | Indexing | Payload Index > Parameterized Index > Tenant Index | documentation | true | Indexing > Payload Index > Parameterized Index > Tenant Index
*Available as of v1.11.0*
Many vector search use-cases require multitenancy. In a multi-tenant scenario the collection is expected to contain multiple subsets of data, where each subset belongs to a different tenant.
Qdrant supports efficient multi-tenant... | [
0.03679238632321358,
-0.037429798394441605,
-0.05423343554139137,
-0.018967559561133385,
0.06964822113513947,
-0.054833974689245224,
0.05160566419363022,
0.0012786398874595761,
0.0933004766702652,
-0.03170754387974739,
0.02810520865023136,
0.025694848969578743,
0.06604766845703125,
-0.0106... |
e13fbb0e-b48c-59c8-bb1f-9d0bf4857e4f | /documentation/manage-data/quantization/ | Quantization | Binary Quantization | documentation | false | Quantization > Binary Quantization
*Available as of v1.5.0*
Binary quantization is an extreme case of scalar quantization.
This feature lets you represent each vector component as a single bit, effectively reducing the memory footprint by a factor of 32. This is the fastest quantization method, since it lets you perf... | [
0.06063001975417137,
-0.04440984129905701,
-0.028859281912446022,
-0.07540728151798248,
0.07756203413009644,
-0.017192987725138664,
0.03504887968301773,
0.04007600620388985,
0.08710862696170807,
-0.05443461239337921,
0.03256000950932503,
-0.008335001766681671,
0.028807150200009346,
-0.0264... |
e15c44a7-1878-5d69-9601-170164503128 | /documentation/scaling/ | Scaling & Resilience | Resilience | documentation | false | Scaling & Resilience > Resilience
Resilience comes from replicating data across multiple nodes. A single node, or a single copy of your data, has no protection against failure: if you lose it, you lose both the data and the ability to serve it. A minimal fault-tolerant cluster needs three nodes and a replication facto... | [
0.02460329793393612,
-0.036083027720451355,
-0.04199155420064926,
-0.06282967329025269,
0.07279444485902786,
-0.0319112129509449,
0.05755732208490372,
0.04490257054567337,
0.03445575758814812,
-0.04029102995991707,
0.04993551969528198,
0.028004417195916176,
0.04205707088112831,
-0.01522650... |
e1633b71-2015-5d1b-801d-fcbe734a1388 | /documentation/faq/qdrant-fundamentals/ | Qdrant Fundamentals | Compatibility > Do you support downgrades? | documentation | false | Qdrant Fundamentals > Compatibility > Do you support downgrades?
We do not support downgrading a cluster on any of our products. If you deploy a newer version of Qdrant, your
data is automatically migrated to the newer storage format. This migration is not reversible. | [
0.0310808252543211,
-0.058434903621673584,
-0.01809578947722912,
-0.06314995139837265,
0.040306445211172104,
-0.04276549071073532,
0.045226529240608215,
0.030859893187880516,
0.013179504312574863,
-0.03624572604894638,
0.0213650930672884,
-0.010555685497820377,
0.07089123874902725,
-0.0040... |
e16d6933-2311-54c7-92b2-bd7235bc21c6 | /documentation/examples/rag-chatbot-red-hat-openshift-haystack/ | Private Chatbot for Interactive Learning | Next steps | documentation | false | Private Chatbot for Interactive Learning > Next steps
- In this example, [Red Hat OpenShift](https://www.redhat.com/en/technologies/cloud-computing/openshift) is the infrastructure of choice for proprietary chatbots. [Read more](https://access.redhat.com/documentation/en-us/red_hat_openshift_ai_self-managed/2.8) about... | [
0.017992038279771805,
-0.03565429523587227,
-0.055785562843084335,
-0.009665374644100666,
0.06264171749353409,
-0.04009830206632614,
0.062454380095005035,
0.0385206937789917,
0.026607893407344818,
-0.013895473442971706,
0.035290706902742386,
-0.016070760786533356,
0.026721209287643433,
-0.... |
e16da9a6-5c02-5d48-8e77-90b94a706701 | /documentation/tutorials-operations/gpu-accelerated-hnsw-indexing/ | GPU-Accelerated HNSW Indexing | documentation | false | GPU-Accelerated HNSW Indexing
| Time: 45 min | Level: Intermediate | Output: [GitHub](https://github.com/qdrant/examples/blob/master/gpu-accelerated-hnsw-indexing/Gpu_Accelerated_HNSW_Indexing.ipynb) | [](https://githubtocolab.com/qdrant/example... | [
0.012650239281356335,
-0.028652707114815712,
-0.06812696158885956,
-0.035741616040468216,
0.07175431400537491,
-0.024362947791814804,
0.05647186189889908,
0.05060073360800743,
0.08933659642934799,
-0.03662628307938576,
0.045084282755851746,
0.012572618201375008,
0.04831278696656227,
-0.021... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.