Yandex Cloud
Поиск
Связаться с намиПодключиться
  • Истории успеха
  • Документация
  • Блог
  • Все сервисы
  • Статус работы сервисов
    • Доступны в регионе
    • Инфраструктура и сеть
    • Платформа данных
    • Контейнеры
    • Инструменты разработчика
    • Бессерверные вычисления
    • Безопасность
    • Мониторинг и управление ресурсами
    • ИИ для бизнеса
    • Бизнес-инструменты
  • Все решения
    • По отраслям
    • По типу задач
    • Экономика платформы
    • Безопасность
    • Техническая поддержка
    • Каталог партнёров
    • Обучение и сертификация
    • Облако для стартапов
    • Облако для крупного бизнеса
    • Центр технологий для общества
    • Партнёрская программа
    • Поддержка IT-бизнеса
    • Облако для фрилансеров
    • Обучение и сертификация
    • Блог
    • Документация
    • Мероприятия и вебинары
    • Контакты, чаты и сообщества
    • Идеи
    • Тарифы Yandex Cloud
    • Промоакции и free tier
    • Правила тарификации
  • Истории успеха
  • Документация
  • Блог
Проект Яндекса
© 2025 ТОО «Облачные Сервисы Казахстан»
Yandex AI Studio
  • Начало работы с Model Gallery
    • О сервисе Yandex AI Studio
    • Yandex Workflows
    • Квоты и лимиты
    • Термины и определения
  • Переход с AI Assistant API на Responses API
  • Совместимость с OpenAI
    • О Yandex Cloud ML SDK
      • Overview
      • Authentication
        • Overview
        • Assistant types
        • Tools
        • Operation
        • Tuning
        • Result types for models
        • Datasets
        • Search API
        • Messages
        • Run types
        • Search index types
        • Other types
        • Batch
      • Request retries
  • Управление доступом
  • Правила тарификации
  • Аудитные логи Audit Trails
  • Публичные материалы
  • История изменений

В этой статье:

  • class yandex_cloud_ml_sdk._search_indexes.file.SearchIndexFile
  • class yandex_cloud_ml_sdk._search_indexes.index_type.BaseSearchIndexType
  • class yandex_cloud_ml_sdk._search_indexes.index_type.TextSearchIndexType
  • class yandex_cloud_ml_sdk._search_indexes.index_type.VectorSearchIndexType
  • class yandex_cloud_ml_sdk._search_indexes.index_type.HybridSearchIndexType
  • class yandex_cloud_ml_sdk._search_indexes.chunking_strategy.BaseIndexChunkingStrategy
  • class yandex_cloud_ml_sdk._search_indexes.chunking_strategy.StaticIndexChunkingStrategy
  • class yandex_cloud_ml_sdk._search_indexes.combination_strategy.BaseIndexCombinationStrategy
  • class yandex_cloud_ml_sdk._search_indexes.combination_strategy.MeanIndexEvaluationTechnique
  • class yandex_cloud_ml_sdk._search_indexes.combination_strategy.MeanIndexCombinationStrategy
  • class yandex_cloud_ml_sdk._search_indexes.combination_strategy.ReciprocalRankFusionIndexCombinationStrategy
  • class yandex_cloud_ml_sdk._search_indexes.normalization_strategy.IndexNormalizationStrategy
  1. Yandex Cloud ML SDK
  2. Справочник SDK (англ.)
  3. Types
  4. Search index types

Search index types

Статья создана
Yandex Cloud
Обновлена 7 ноября 2025 г.
  • class yandex_cloud_ml_sdk._search_indexes.file.SearchIndexFile
  • class yandex_cloud_ml_sdk._search_indexes.index_type.BaseSearchIndexType
  • class yandex_cloud_ml_sdk._search_indexes.index_type.TextSearchIndexType
  • class yandex_cloud_ml_sdk._search_indexes.index_type.VectorSearchIndexType
  • class yandex_cloud_ml_sdk._search_indexes.index_type.HybridSearchIndexType
  • class yandex_cloud_ml_sdk._search_indexes.chunking_strategy.BaseIndexChunkingStrategy
  • class yandex_cloud_ml_sdk._search_indexes.chunking_strategy.StaticIndexChunkingStrategy
  • class yandex_cloud_ml_sdk._search_indexes.combination_strategy.BaseIndexCombinationStrategy
  • class yandex_cloud_ml_sdk._search_indexes.combination_strategy.MeanIndexEvaluationTechnique
  • class yandex_cloud_ml_sdk._search_indexes.combination_strategy.MeanIndexCombinationStrategy
  • class yandex_cloud_ml_sdk._search_indexes.combination_strategy.ReciprocalRankFusionIndexCombinationStrategy
  • class yandex_cloud_ml_sdk._search_indexes.normalization_strategy.IndexNormalizationStrategy

class yandexcloudmlsdk.searchindexes.file.SearchIndexFileclass yandex_cloud_ml_sdk._search_indexes.file.SearchIndexFile

This class represents a file associated with a search index.

search_index_id: str

the unique identifier for the search index

created_by: str

the identifier of the user or system that created the file

created_at: datetime

the timestamp when the file was created

id: str

class yandexcloudmlsdk.searchindexes.indextype.BaseSearchIndexTypeclass yandex_cloud_ml_sdk._search_indexes.index_type.BaseSearchIndexType

A class for search index types.

chunking_strategy: BaseIndexChunkingStrategy | None = None

the strategy used for chunking the index

class yandexcloudmlsdk.searchindexes.indextype.TextSearchIndexTypeclass yandex_cloud_ml_sdk._search_indexes.index_type.TextSearchIndexType

A class which represents a text search index type.

chunking_strategy: BaseIndexChunkingStrategy | None = None

the strategy used for chunking the index

class yandexcloudmlsdk.searchindexes.indextype.VectorSearchIndexTypeclass yandex_cloud_ml_sdk._search_indexes.index_type.VectorSearchIndexType

A class which represents a vector search index type.

doc_embedder_uri: str | None = None

URI for the document embedder

query_embedder_uri: str | None = None

URI for the query embedder

chunking_strategy: BaseIndexChunkingStrategy | None = None

the strategy used for chunking the index

class yandexcloudmlsdk.searchindexes.indextype.HybridSearchIndexTypeclass yandex_cloud_ml_sdk._search_indexes.index_type.HybridSearchIndexType

A class which represents a hybrid search index type combining text and vector search indices.

text_search_index: TextSearchIndexType | None = None

the text search index associated with the hybrid index

vector_search_index: VectorSearchIndexType | None = None

the vector search index associated with the hybrid index

normalization_strategy: IndexNormalizationStrategy | str | int | None = None

the strategy for normalizing index results

combination_strategy: BaseIndexCombinationStrategy | None = None

the strategy for combining results from different indices

chunking_strategy: BaseIndexChunkingStrategy | None = None

the strategy used for chunking the index

class yandexcloudmlsdk.searchindexes.chunkingstrategy.BaseIndexChunkingStrategyclass yandex_cloud_ml_sdk._search_indexes.chunking_strategy.BaseIndexChunkingStrategy

A class for an index chunking strategy, from which all other strategies are inherited.

class yandexcloudmlsdk.searchindexes.chunkingstrategy.StaticIndexChunkingStrategyclass yandex_cloud_ml_sdk._search_indexes.chunking_strategy.StaticIndexChunkingStrategy

This class implements a static chunking strategy (i.e. a specific strategy with specific properties).

It is characterized by maximum chunk size and overlap in tokens.

max_chunk_size_tokens: int

the maximum size of each chunk in tokens

chunk_overlap_tokens: int

the number of overlapping tokens between consecutive chunks

class yandexcloudmlsdk.searchindexes.combinationstrategy.BaseIndexCombinationStrategyclass yandex_cloud_ml_sdk._search_indexes.combination_strategy.BaseIndexCombinationStrategy

A class for index combination strategies.

class yandexcloudmlsdk.searchindexes.combinationstrategy.MeanIndexEvaluationTechniqueclass yandex_cloud_ml_sdk._search_indexes.combination_strategy.MeanIndexEvaluationTechnique

A class with enumeration for mean index evaluation techniques.

MEAN_EVALUATION_TECHNIQUE_UNSPECIFIED = 0

an unspecified mean evaluation technique

ARITHMETIC = 1

the arithmetic mean, calculated as the sum of values divided by the count of values

GEOMETRIC = 2

the geometric mean, calculated as the n-th root of the product of n values

HARMONIC = 3

the harmonic mean, calculated as the reciprocal of the arithmetic mean of the reciprocals of the values

__new__(value)

conjugate()

Returns self, the complex conjugate of any int.

bit_length()

Number of bits necessary to represent self in binary.

>>> bin(37)
'0b100101'
>>> (37).bit_length()
6

bit_count()

Number of ones in the binary representation of the absolute value of self.

Also known as the population count.

>>> bin(13)
'0b1101'
>>> (13).bit_count()
3

as_integer_ratio()

Return a pair of integers, whose ratio is equal to the original int.

The ratio is in lowest terms and has a positive denominator.

>>> (10).as_integer_ratio()
(10, 1)
>>> (-10).as_integer_ratio()
(-10, 1)
>>> (0).as_integer_ratio()
(0, 1)

is_integer()

Returns True. Exists for duck type compatibility with float.is_integer.

real

the real part of a complex number

imag

the imaginary part of a complex number

numerator

the numerator of a rational number in lowest terms

denominator

the denominator of a rational number in lowest terms

__init__(*args, **kwds)

class yandexcloudmlsdk.searchindexes.combinationstrategy.MeanIndexCombinationStrategyclass yandex_cloud_ml_sdk._search_indexes.combination_strategy.MeanIndexCombinationStrategy

A class which contains mean index combination strategy with evaluation technique and weights.

mean_evaluation_technique: MeanIndexEvaluationTechnique | None

the technique used for mean evaluation

weights: Collection[float] | None

the weights associated with the evaluation technique

__init__(mean_evaluation_technique, weights)

Parameters

  • mean_evaluation_technique (MeanIndexEvaluationTechnique | None)
  • weights (Collection[float] | None)

Return type

None

class yandexcloudmlsdk.searchindexes.combinationstrategy.ReciprocalRankFusionIndexCombinationStrategyclass yandex_cloud_ml_sdk._search_indexes.combination_strategy.ReciprocalRankFusionIndexCombinationStrategy

A class which describes reciprocal rank fusion index combination strategy. Reciprocal rank fusion is a method for combining multiple result sets with different relevance indicators into a single result set.

k: int | None = None

the parameter k for RRFscore. Default is 60.

__init__(k=None)

Parameters

k (int | None)

Return type

None

class yandexcloudmlsdk.searchindexes.normalizationstrategy.IndexNormalizationStrategyclass yandex_cloud_ml_sdk._search_indexes.normalization_strategy.IndexNormalizationStrategy

Enumeration for index normalization strategies.

This class defines the various normalization strategies that can be applied to an index.

NORMALIZATION_STRATEGY_UNSPECIFIED = 0

indicates that no normalization strategy has been specified

MIN_MAX = 1

represents the Min-Max normalization strategy

L2 = 2

represents the L2 normalization strategy

__new__(value)

conjugate()

Returns self, the complex conjugate of any int.

bit_length()

Number of bits necessary to represent self in binary.

>>> bin(37)
'0b100101'
>>> (37).bit_length()
6

bit_count()

Number of ones in the binary representation of the absolute value of self.

Also known as the population count.

>>> bin(13)
'0b1101'
>>> (13).bit_count()
3

as_integer_ratio()

Return a pair of integers, whose ratio is equal to the original int.

The ratio is in lowest terms and has a positive denominator.

>>> (10).as_integer_ratio()
(10, 1)
>>> (-10).as_integer_ratio()
(-10, 1)
>>> (0).as_integer_ratio()
(0, 1)

is_integer()

Returns True. Exists for duck type compatibility with float.is_integer.

real

the real part of a complex number

imag

the imaginary part of a complex number

numerator

the numerator of a rational number in lowest terms

denominator

the denominator of a rational number in lowest terms

__init__(*args, **kwds)

Была ли статья полезна?

Предыдущая
Run types
Следующая
Other types
Проект Яндекса
© 2025 ТОО «Облачные Сервисы Казахстан»