iragent.models.KnowledgeGraphBuilder Class Reference

Public Member Functions
	__init__ (self, embedding_model, int chunk_size=128, int chunk_overlap=32, index_dir="./vector_store/")
	build_index_from_markdown (self, str dir_path, bool strip_markdown=True, bool recursive=True)
	build_index_from_texts (self, list[str] texts)
	save_index (self)
	load_index (self)
List[str]	search (self, str query, int k)

Public Attributes
	embedding_model = embedding_model
	index_path = os.path.join(index_dir, "faiss_index.index")
	index_dir = index_dir
	docs_path = os.path.join(index_dir, "faiss_index.index.docs")
	index = None
list	document = []
	faiss = faiss
	chunk_size = chunk_size
	chunk_overlap = chunk_overlap

Protected Member Functions
list[str]	_chunker (self, str text)
list[str]	_read_markdown_and_chunk (self, str dir_path, bool strip_markdown=True, bool recursive=True)
np.ndarray	_normalize (self, np.ndarray vectors)

Constructor & Destructor Documentation

__init__()

iragent.models.KnowledgeGraphBuilder.__init__	(		self,
			embedding_model,
		int	chunk_size = 128,
		int	chunk_overlap = 32,
			index_dir = "./vector_store/" )

Member Function Documentation

_chunker()

list[str] iragent.models.KnowledgeGraphBuilder._chunker ( self, str text )

protected

Splits text into overlapping chunks based on token count.

Args:
    text: The input text to split.
    max_tokens: Maximum tokens per chunk.
    overlap: Number of tokens to overlap between chunks.
    model_name: Model name for tokenizer compatibility.

Returns:
    List of text chunks.

_normalize()

np.ndarray iragent.models.KnowledgeGraphBuilder._normalize ( self, np.ndarray vectors )

protected

_read_markdown_and_chunk()

list[str] iragent.models.KnowledgeGraphBuilder._read_markdown_and_chunk ( self, str dir_path, bool strip_markdown = True, bool recursive = True )

protected

build_index_from_markdown()

iragent.models.KnowledgeGraphBuilder.build_index_from_markdown	(		self,
		str	dir_path,
		bool	strip_markdown = True,
		bool	recursive = True )

build_index_from_texts()

iragent.models.KnowledgeGraphBuilder.build_index_from_texts	(		self,
		list[str]	texts )

Create a FAISS index from a list of texts

load_index()

iragent.models.KnowledgeGraphBuilder.load_index

(

self

)

Load FAISS index from disk

save_index()

iragent.models.KnowledgeGraphBuilder.save_index

(

self

)

search()

List[str] iragent.models.KnowledgeGraphBuilder.search	(		self,
		str	query,
		int	k )

Search for nearest neighbors, this function get query: str and k:int as input. 

Member Data Documentation

chunk_overlap

iragent.models.KnowledgeGraphBuilder.chunk_overlap = chunk_overlap

chunk_size

iragent.models.KnowledgeGraphBuilder.chunk_size = chunk_size

docs_path

iragent.models.KnowledgeGraphBuilder.docs_path = os.path.join(index_dir, "faiss_index.index.docs")

document

list iragent.models.KnowledgeGraphBuilder.document = []

embedding_model

iragent.models.KnowledgeGraphBuilder.embedding_model = embedding_model

faiss

iragent.models.KnowledgeGraphBuilder.faiss = faiss

index

iragent.models.KnowledgeGraphBuilder.index = None

index_dir

iragent.models.KnowledgeGraphBuilder.index_dir = index_dir

index_path

iragent.models.KnowledgeGraphBuilder.index_path = os.path.join(index_dir, "faiss_index.index")

---

The documentation for this class was generated from the following file:

• iragent/models.py