Langchain ChatGPT your documents challenge PART 2 with Gpt 3 and OpenAI embeddings and Streamlit UI

We will build a ChatGPT like chatbot which uses langchain, openai GPT-3 and OpenAI embeddings and Wikipedia API to search over an uploaded document and answer questions using relevant context. Includes a Streamlit user interface with a file uploader.

NOTE: Forgot to mention in the video that you run the streamlit app by running "streamlit run main(dot)py" in your command prompt

Timeline:
00:00 DEMO
02:40 Code review step by step
04:04 Prompt Engineering
05:06 Wikipedia API
05:31 Avoiding Unicode encoding errors
08:00 Callback for Token Count
12:28 Full code review
13:10 Final Demo

#langchain #langchaintutorial #chatgpt #gpt3 #openai #openaiembeddings
#chatbot #gptchat #gpt3 #gpttutorial #langchainchallenge #streamlit #streamlittutorial #promptengineering

[echohive]

Forgot to mention in the video that you run the streamlit app by running "streamlit run main(dot)py" in your command prompt
I hope You will find this video helpful! :)

[MichaelHoughton_]

how can you make the chat flow int he same direction as the chatgpt UI or a normal text/messenger UI?

[prasadavijeet]

Great video! Are you rebuilding the vector store from scratch each time or is it calculated for only the new files and then appended to the vector store?

[ajmiravenkatesh]

sir i hava question for you,
how to do
create atleast 3 diffrerent doct by spliting one document

and connect to open Ai gpt3 api and do the following things
i)Ask a question
ii) it should give back the relevant document through the stored vector database (embedded document)
iii) it will answer the asked question through gpt3

[ChuckWilliamsTechnology]

what changes are needed to upload PDF, CVS, or Excel files as raw dataset to be vectorized?

[huanxu]

Love the video! OpenAI embed is truly a hidden gem.

[dexter_chann]

May I ask how to add open ai key in the python code instead of placing the api key in the environment variable?