Retrieval Augmented Generation (RAG) con LangChain
Meri Nova
Machine Learning Engineer

Codifica i chunk come un unico vettore con componenti non zero

Codifica i chunk come un unico vettore con componenti non zero

Codifica usando il match delle parole con componenti per lo più zero

TF-IDF: Codifica i documenti usando le parole che li rendono unici

BM25: Riduce l'impatto delle parole ad alta frequenza che saturano la codifica
from langchain_community.retrievers import BM25Retrieverchunks = [ "Python was created by Guido van Rossum and released in 1991.", "Python is a popular language for machine learning (ML).", "The PyTorch library is a popular Python library for AI and ML." ]bm25_retriever = BM25Retriever.from_texts(chunks, k=3)
results = bm25_retriever.invoke("When was Python created?")
print("Most Relevant Document:")
print(results[0].page_content)
Most Relevant Document:
Python was created by Guido van Rossum and released in 1991.
retriever = BM25Retriever.from_documents( documents=chunks, k=5 )chain = ({"context": retriever, "question": RunnablePassthrough()} | prompt | llm | StrOutputParser() )
print(chain.invoke("How can LLM hallucination impact a RAG application?"))
L'applicazione RAG può generare risposte fuori tema o inaccurate.
Retrieval Augmented Generation (RAG) con LangChain