Chatting about youtube videos transcripts

youtube.py

#%%
import os
import openai
from dotenv import load_dotenv
from langchain.chat_models import AzureChatOpenAI
from langchain.embeddings import OpenAIEmbeddings
from langchain.text_splitter import TokenTextSplitter

# Load environment variables
load_dotenv(override=True)

# Configure OpenAI API
openai.api_type = "azure"
openai.api_base = os.getenv('OPENAI_API_BASE')
openai.api_key = os.getenv("OPENAI_API_KEY")
openai.api_version = "2023-03-15-preview"

#%%

from langchain.document_loaders import YoutubeLoader

import sys
url = sys.argv[1] 

loader = YoutubeLoader.from_youtube_url(url)
transcript = loader.load()[0]

#%%
splitter = TokenTextSplitter(model_name='gpt-3.5-turbo', chunk_size=4000, chunk_overlap=100)
docs = splitter.split_text(transcript.page_content)

#%%
# Create embeddings
embeddings = OpenAIEmbeddings(
    deployment="Embeddings",model="text-embedding-ada-002", chunk_size=1)

# %%
from langchain.vectorstores import Chroma
docsearch = Chroma.from_texts(docs, embeddings)

# %%
retriever = docsearch.as_retriever(search_kwargs={"k": 2})

# %%
llm = AzureChatOpenAI(
    openai_api_base = 'https://druid.openai.azure.com/',
    deployment_name = 'Gpt35',
    openai_api_type = 'azure',
    temperature = 0, 
    verbose=False)

# %%
from langchain.chains import RetrievalQA
qa = RetrievalQA.from_chain_type(
    llm = llm, chain_type = "stuff", retriever = retriever, verbose = False)

# %%
while True:

    query = input("\nQ: ")

    if query == "exit":
        break

    res = qa(query)
    answer = res['result']

    print(answer)