Okay let me try that lang chain demo

19:37 ok so first per

https://langchain.readthedocs.io/en/latest/getting_started/getting_started.html here, installing this stuff,

creating a new environment on my laptop

pip install langchain
pip install openai
pip install faiss-cpu # adding this here after the fact after getting below error

20:11 got one error,

ValueError: Could not import faiss python package. Please it install it with `pip install faiss` or `pip install faiss-cpu` (depending on Python version).

Ok let me query

the [[New yorker]] #article I added . I recently read [[The American Beast New Yorker]] this article by [[person Jill Lepore]] about the #report that was commissioned about the #[[January 6th Insurrection]] . I used my #iphone #scan-to-text feature to pull in the first page and a half to a text file, article.txt to try this out . Let’s see how this works.

Raw data is here https://github.com/namoopsoo/namoopsoo.github.io/blob/hugo-main/data/2022-01-29-lang-chain-quick-look__files/article.txt

from langchain.embeddings.openai import OpenAIEmbeddings
from langchain.text_splitter import CharacterTextSplitter
from langchain.vectorstores.faiss import FAISS
from langchain.docstore.document import Document
from langchain.prompts import PromptTemplate

from langchain.chains.question_answering import load_qa_chain
from langchain.llms import OpenAI


with open('article.txt') as f:
  article = f.read()
text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
texts = text_splitter.split_text(article)

embeddings = OpenAIEmbeddings()
docsearch = FAISS.from_texts(texts, embeddings)

chain = load_qa_chain(OpenAI(temperature=0), chain_type="stuff")

query = "What did William Barr do?"
docs = docsearch.similarity_search(query)

In [16]: chain.run(input_documents=docs, question=query)
  ...: 
Out[16]: ' William Barr referred to Trump\'s legal team as the "clown car."'

indeed

20:24 amazing haha ! Let me try another.

In [19]: query = "Did Cheney get reelected?"
  ...: docs = docsearch.similarity_search(query)
  ...: chain.run(input_documents=docs, question=query)
  ...: 
Out[19]: ' No, Cheney lost her bid for reelection.'

wow !

okay one more

query = "Was there election fraud?"
docs = docsearch.similarity_search(query)
chain.run(input_documents=docs, question=query)

Out[20]: ' No, the January 6th Report found no evidence of election fraud.'

Wow this is really cool! I’m not sure which section was used for this answer. I wonder if it was this one?

I want to check out the section

here which says you can get a precise link to how the answer to your question was constructed, with the evidence basically.