Class LightRAG

LightRAG/lightrag/lightrag.py:67–477

@dataclass
class LightRAG:
    working_dir: str = field(
        default_factory=lambda: f"./lightrag_cache_{datetime.now().strftime('%Y-%m-%d-%H:%M:%S')}"
    )

    kv_storage: str = field(default="JsonKVStorage")
    vector_storage: str = field(default="NanoVectorDBStorage")
    graph_storage: str = field(default="NetworkXStorage")

    current_log_level = logger.level
    log_level: str = field(default=current_log_level)

    # text chunking
    chunk_token_size: int = 1200
    chunk_overlap_token_size: int = 100
    tiktoken_model_name: str = "gpt-4o-mini"

    # entity extraction
    entity_extract_max_gleaning: int = 1
    entity_summary_to_max_tokens: int = 500

    # node embedding
    node_embedding_algorithm: str = "node2vec"
    node2vec_params: dict = field(
        default_factory=lambda: {
            "dimensions": 1536,
            "num_walks": 10,
            "walk_length": 40,
            "window_size": 2,
            "iterations": 3,
            "random_seed": 3,
        }
    )

    # embedding_func: EmbeddingFunc = field(default_factory=lambda:hf_embedding)
    embedding_func: EmbeddingFunc = field(default_factory=lambda: openai_embedding)
    embedding_batch_num: int = 32
    embedding_func_max_async: int = 16

    # LLM
    llm_model_func: callable = gpt_4o_mini_complete  # hf_model_complete#
    llm_model_name: str = "meta-llama/Llama-3.2-1B-Instruct"  #'meta-llama/Llama-3.2-1B'#'google/gemma-2-2b-it'
    llm_model_max_token_size: int = 32768
    llm_model_max_async: int = 16
    llm_model_kwargs: dict = field(default_factory=dict)

    # storage
    vector_db_storage_cls_kwargs: dict = field(default_factory=dict)

    enable_llm_cache: bool = True

    # extension
    addon_params: dict = field(default_factory=dict)
    convert_response_to_json_func: callable = convert_response_to_json

    def __post_init__(self):
        log_file = os.path.join(self.working_dir, "lightrag.log")
        set_logger(log_file)