nat.tool.nvidia_rag

Attributes

logger

Classes

BaseRagResult	Base class for RAG service response schemas.
SourceResult	RAG Blueprint /search endpoint schema.
DocumentChunk	GenerativeAIExamples chain server /search endpoint schema.
NVIDIARAGToolConfig	Tool used to search the NVIDIA Developer database for documents across a variety of NVIDIA asset types.

Functions

parse_rag_response(...)	Auto-detect RAG schema and return Documents.
nvidia_rag_tool(...)

Module Contents

logger

class BaseRagResult(/, **data: Any)

Bases: pydantic.BaseModel

Base class for RAG service response schemas.

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.

content: str

score: float

abstractmethod get_document_title() → str

Override in subclass to return the document name field.

to_document() → langchain_core.documents.Document

class SourceResult(/, **data: Any)

Bases: BaseRagResult

RAG Blueprint /search endpoint schema.

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.

document_name: str

get_document_title() → str

Override in subclass to return the document name field.

class DocumentChunk(/, **data: Any)

Bases: BaseRagResult

GenerativeAIExamples chain server /search endpoint schema.

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.

filename: str

get_document_title() → str

Override in subclass to return the document name field.

parse_rag_response(

data: dict[str, Any],

) → list[langchain_core.documents.Document]

Auto-detect RAG schema and return Documents.

class NVIDIARAGToolConfig(/, **data: Any)

Bases: nat.data_models.function.FunctionBaseConfig

Tool used to search the NVIDIA Developer database for documents across a variety of NVIDIA asset types.

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.

base_url: str = None

timeout: int = None

document_separator: str = None

document_prompt: str = None

top_k: int = None

collection_name: str = None

async nvidia_rag_tool(

config: NVIDIARAGToolConfig,

builder: nat.builder.builder.Builder,

) → collections.abc.AsyncGenerator[nat.builder.function_info.FunctionInfo, None]