Getting Started

¶Quick Start

Use vector type in your entities :

use Doctrine\ORM\Mapping as ORM;

#[ORM\Entity(repositoryClass: VectorsRepository::class)]
#[ORM\Table(name: 'app__vectors')]
class Vectors
{

#[ORM\Column(type: 'vector', length: 1024, nullable: true)]
private array $vectors = [];

}

Warning

If you use symfony console make:entity add manually the length parameter attribute as vector. Length is your model embedding's dimension.

¶For example OpenAi use these dimensions:

text-embedding-3-small : 1536

text-embedding-3-large : 3072 (customizable)

¶Mistral AI

Mistral-embed : 1024

Additionally, you should manually add an HNSW index to your vector's column. Be aware that dimension should be 2000 max for HNSW indexes.

¶L2 distance

CREATE INDEX ON items USING hnsw (embedding vector_l2_ops);

¶Inner product

CREATE INDEX ON items USING hnsw (embedding vector_ip_ops);

¶Cosine distance

CREATE INDEX ON items USING hnsw (embedding vector_cosine_ops);