Number of edges per node in the index graph. Larger the value - more accurate the search, more space required.
ef_construct
integer
Number of neighbours to consider during the index building. Larger the value - more accurate the search, more time required to build index.
full_scan_threshold
integer
Minimal size threshold (in KiloBytes) below which full-scan is preferred over HNSW search. This measures the total size of vectors being queried against. When the maximum estimated amount of points th
max_indexing_threads
integer
Number of parallel threads used for background index building. If 0 - automatically select from 8 to 16. Best to keep between 8 and 16 to prevent likelihood of slow building or broken/inefficient HNSW
on_disk
boolean
Store HNSW index on disk. If set to false, index will be stored in RAM. Default: false
payload_m
integer
Custom M param for hnsw graph built for payload index. If not set, default M will be used.
inline_storage
boolean
Store copies of original and quantized vectors within the HNSW index file. Default: false. Enabling this option will trade the search speed for disk usage by reducing amount of random seeks during the
{
"$schema": "https://json-schema.org/draft/2020-12/schema",
"$id": "#/components/schemas/HnswConfig",
"title": "HnswConfig",
"description": "Config of HNSW index",
"type": "object",
"required": [
"ef_construct",
"full_scan_threshold",
"m"
],
"properties": {
"m": {
"description": "Number of edges per node in the index graph. Larger the value - more accurate the search, more space required.",
"type": "integer",
"format": "uint",
"minimum": 0
},
"ef_construct": {
"description": "Number of neighbours to consider during the index building. Larger the value - more accurate the search, more time required to build index.",
"type": "integer",
"format": "uint",
"minimum": 4
},
"full_scan_threshold": {
"description": "Minimal size threshold (in KiloBytes) below which full-scan is preferred over HNSW search. This measures the total size of vectors being queried against. When the maximum estimated amount of points that a condition satisfies is smaller than `full_scan_threshold_kb`, the query planner will use full-scan search instead of HNSW index traversal for better performance. Note: 1Kb = 1 vector of size 256",
"type": "integer",
"format": "uint",
"minimum": 0
},
"max_indexing_threads": {
"description": "Number of parallel threads used for background index building. If 0 - automatically select from 8 to 16. Best to keep between 8 and 16 to prevent likelihood of slow building or broken/inefficient HNSW graphs. On small CPUs, less threads are used.",
"default": 0,
"type": "integer",
"format": "uint",
"minimum": 0
},
"on_disk": {
"description": "Store HNSW index on disk. If set to false, index will be stored in RAM. Default: false",
"type": "boolean",
"nullable": true
},
"payload_m": {
"description": "Custom M param for hnsw graph built for payload index. If not set, default M will be used.",
"type": "integer",
"format": "uint",
"minimum": 0,
"nullable": true
},
"inline_storage": {
"description": "Store copies of original and quantized vectors within the HNSW index file. Default: false. Enabling this option will trade the search speed for disk usage by reducing amount of random seeks during the search. Requires quantized vectors to be enabled. Multi-vectors are not supported.",
"type": "boolean",
"nullable": true
}
}
}