276 packages tagged with “Embeddings”
Kernel Memory is a Copilot/Semantic Kernel Plugin and Memory Web Service to index and query any data and documents, using LLM and natural language, tracking sources and showing citations. The package contains the interfaces and models shared by all Kernel Memory packages.
The package contains the core logic and abstractions of Kernel Memory, not including extensions.
Catalyst is a Natural Language Processing library built from scratch for speed. Inspired by spaCy's design, it brings pre-trained models, out-of-the box support for training word and document embeddings, and flexible entity recognition models. You can install language-specific models with the model packages Catalyst.Models.[Language], like: Catalyst.Models.English.
Azure AI Search connector for Microsoft Kernel Memory, to store and search memory using Azure AI Search vector indexing and semantic features.
Provide tokenizers to allow counting content tokens for text and embeddings
Provide access to OpenAI LLM models in Kernel Memory to generate embeddings and text
Add Azure AI Document Intelligence to Kernel Memory to extract content from images and documents.
Qdrant connector for Microsoft Kernel Memory, to store and search memory using Qdrant vector indexing and Qdrant features.
Postgres(with pgvector extension) connector for Microsoft Kernel Memory, to store and search memory using Postgres vector indexing and Postgres features.
Provide access to Azure OpenAI LLM models in Kernel Memory to generate embeddings and text
Azure Queue Storage connector for Microsoft Kernel Memory, to run asynchronous pipelines via Azure Queue Storage queues.
This package contains the default English models for Catalyst. Catalyst is a Natural Language Processing library built from scratch for speed. Inspired by spaCy's design, it brings pre-trained models, out-of-the box support for training word and document embeddings, and flexible entity recognition models.
RabbitMQ connector for Microsoft Kernel Memory, to run asynchronous pipelines via RabbitMQ queues.
Kernel Memory Web Client allows to connect to Kernel Memory Web Service to index and query any data and documents, using LLM and natural language, tracking sources and showing citations.
Redis connector for Microsoft Kernel Memory, to store and search memory using Redis vector search and other Redis features.
Provide access to OpenAI LLM models in Kernel Memory to generate text
MS SQLServer connector for Microsoft Kernel Memory, to store and search memory using MSSQL.
MongoDb Atlas connector for Microsoft Kernel Memory, to store and search memory records with embeddings, and files.
Collection of text chunkers.
Elasticsearch connector for Microsoft Kernel Memory, to store and search memory using Elasticsearch vector search and other Elasticsearch features.
Kernel Memory Plugin allows to use Kernel Memory Service as a Semantic Kernel plugin, to index and query any data and documents, using LLM and natural language, tracking sources and showing citations.
Azure Blob Storage adapter allowing Kernel Memory to upload documents and maintain their state in Azure Blobs
This package contains the default Polish models for Catalyst. Catalyst is a Natural Language Processing library built from scratch for speed. Inspired by spaCy"s design, it brings pre-trained models, out-of-the box support for training word and document embeddings, and flexible entity recognition models.
This package contains the default French models for Catalyst. Catalyst is a Natural Language Processing library built from scratch for speed. Inspired by spaCy"s design, it brings pre-trained models, out-of-the box support for training word and document embeddings, and flexible entity recognition models.
This package contains the default Portuguese models for Catalyst. Catalyst is a Natural Language Processing library built from scratch for speed. Inspired by spaCy"s design, it brings pre-trained models, out-of-the box support for training word and document embeddings, and flexible entity recognition models.
This package contains the default Russian models for Catalyst. Catalyst is a Natural Language Processing library built from scratch for speed. Inspired by spaCy"s design, it brings pre-trained models, out-of-the box support for training word and document embeddings, and flexible entity recognition models.
This package contains the default Italian models for Catalyst. Catalyst is a Natural Language Processing library built from scratch for speed. Inspired by spaCy"s design, it brings pre-trained models, out-of-the box support for training word and document embeddings, and flexible entity recognition models.
This package contains the default Finnish models for Catalyst. Catalyst is a Natural Language Processing library built from scratch for speed. Inspired by spaCy"s design, it brings pre-trained models, out-of-the box support for training word and document embeddings, and flexible entity recognition models.
This package contains the default Hebrew models for Catalyst. Catalyst is a Natural Language Processing library built from scratch for speed. Inspired by spaCy"s design, it brings pre-trained models, out-of-the box support for training word and document embeddings, and flexible entity recognition models.
This package contains the default Czech models for Catalyst. Catalyst is a Natural Language Processing library built from scratch for speed. Inspired by spaCy"s design, it brings pre-trained models, out-of-the box support for training word and document embeddings, and flexible entity recognition models.