nat.tool.retriever#

Attributes#

`logger`
`AIQRetrieverConfig`
`aiq_retriever_tool`

Classes#

RetrieverConfig

Retriever tool which provides a common interface for different vectorstores. Its

Functions#

`_get_description_from_config`(→ str)	Generate a description of what the tool will do based on how it is configured.
`retriever_tool`(config, builder)	Configure a NAT Retriever Tool which supports different clients such as Milvus and Nemo Retriever.

Module Contents#

logger#

class RetrieverConfig(/, **data: Any)#

Bases: nat.data_models.function.FunctionBaseConfig

Retriever tool which provides a common interface for different vectorstores. Its configuration uses clients, which are the vectorstore-specific implementaiton of the retriever interface.

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

retriever: nat.data_models.component_ref.RetrieverRef = None#

raise_errors: bool = None#

topic: str | None = None#

description: str | None = None#

_get_description_from_config(config: RetrieverConfig) → str#: Generate a description of what the tool will do based on how it is configured.

async retriever_tool( config: RetrieverConfig, builder: nat.builder.builder.Builder, )#

Configure a NAT Retriever Tool which supports different clients such as Milvus and Nemo Retriever.

Args:: config: A config object with required parameters ‘client’ and ‘client_config’ builder: A workflow builder object

AIQRetrieverConfig#

aiq_retriever_tool#