Summarize loaded documents without vector db

[hawksung1]

I am having difficulty changing the code from LangChain to LangFlow.

I want summarize document without vector db in langflow.
langchain code below

from langchain.chains.summarize import load_summarize_chain
from langchain_community.document_loaders import WebBaseLoader
from langchain_openai import ChatOpenAI

loader = WebBaseLoader("https://lilianweng.github.io/posts/2023-06-23-agent/")
docs = loader.load()

llm = ChatOpenAI(temperature=0, model_name="gpt-3.5-turbo-1106")
chain = load_summarize_chain(llm, chain_type="stuff")

chain.run(docs)

I tried flow like below ( using vector store)

Ultimately, I want to summarize the content when a user attaches a file in the chat (the attachment feature will be implemented separately).

If you have any ideas, please tell me

[dosubot[bot]]

🤖

Hey @hawksung1, great to see you back here! Let's dive into another adventure together. 🚀

Based on the code you've provided, it seems like you're trying to replace the vector database in LangFlow with a summarization component similar to the one in LangChain. However, without the specific details of the summarization component in LangChain, it's challenging to provide a precise code modification.

In LangFlow, the vector database is created in the RedisComponent class, specifically in the build method. If you want to replace this with a summarization component, you would need to replace the Redis.from_documents call with a call to a summarization method or class. This would also require changing the arguments and return type of the build method to match the requirements of the summarization component. You can find this in the Redis.py file.

As for the document loading and processing, it's handled by the DocumentLoaderCreator class in the src/backend/langflow/interface/document_loaders/base.py file. This class creates document loaders, which are used to load and process documents. You can find more details about this in the base.py file.

Unfortunately, without more information about the summarization component in LangChain, it's hard to provide a more specific solution. If you could provide more details about how the summarization is done in LangChain, I might be able to help you better.

Sources

• src/backend/langflow/components/vectorstores/Redis.py
• src/frontend/src/types/flow/index.ts
• src/backend/langflow/interface/document_loaders/base.py

---

This response is meant to be useful and save you time. It is not meant to be a precise solution, but rather a starting point for your own research.

Help me be more useful! Please leave a 👍 if this is helpful and 👎 if it is irrelevant.

If you want to continue the conversation, start your reply with @dosu-bot.

[hawksung1]

thanks for your answer.
I'm doing my job with AnalyzeDocumentChain.
This chain is almost same as what I wanted. (need some custom)