Skip to main content

​
Overview

The FlashrankRerankAdapter implements reranking functionality that:
  • Improves Relevance: Re-ranks retrieved documents for better relevance
  • Fast Performance: Uses optimized ONNX models for speed
  • Easy Integration: Seamlessly integrates with LangChain retrieval chains
  • Automatic Model Download: Downloads models automatically on first use

​
Configuration

Configure through LangChatConfig: 
from langchat.config import LangChatConfig

config = LangChatConfig(
    reranker_model="ms-marco-MiniLM-L-12-v2",
    reranker_cache_dir="rerank_models",
    reranker_top_n=3  # Top documents after reranking
)

​
Usage

​
Basic Usage

The adapter is automatically initialized by LangChatEngine: 
from langchat import LangChat, LangChatConfig

config = LangChatConfig(
    reranker_model="ms-marco-MiniLM-L-12-v2",
    reranker_top_n=3,
    # ... other config
)

langchat = LangChat(config=config)
# Adapter is automatically initialized