- Видео 34
- Просмотров 22 467
Zachary Proser
Добавлен 22 мар 2006
I BUILD AI APPS IN PUBLIC | PINECONE ENGINEER
Staff Developer Advocate @ Pinecone.io
Staff Developer Advocate @ Pinecone.io
Demo: RAG pipeline with LangChain, OpenAI, Pinecone, Next.js and Vercel
This video demos the custom Retrieval Augmented Generation pipeline I built against my own blog at:
zackproser.com/chat
It uses Vercel's latest AI SDK, OpenAI's ChatGPT 4o model, LangChain for preparing, chunking and embedding data and Pinecone for vector storage.
zackproser.com/chat
It uses Vercel's latest AI SDK, OpenAI's ChatGPT 4o model, LangChain for preparing, chunking and embedding data and Pinecone for vector storage.
Просмотров: 294
Видео
Codeium vs ChatGPT
Просмотров 3214 месяца назад
Blog post: zackproser.com/blog/codeium-vs-chatgpt Codeium is a developer-facing autocompletion tool that helps software engineers write better #code faster. ChatGPT is a multi-modal model capable of text and code generation, viewing images and reading the internet, having complex discussions about architecture and more. People often ask me which is better - should you use Codeium or ChatGPT?
How to build chat with your data using Pinecone, LangChain and OpenAI
Просмотров 6 тыс.4 месяца назад
zackproser I show step by step how to build a Chatbot using Pinecone, LangChain and OpenAI in this easy to follow tutorial for beginners. I ingest my entire blog full of MDX posts, chunk and convert them to embeddings using LangChain and OpenAI. Then I upsert them into my Pinecone vector database to build a knowledge base that our chatbot can use to answer questions. I build everyth...
How to use ChatGPT in the terminal
Просмотров 3 тыс.5 месяцев назад
zackproser I demonstrate how to install the open-source project Mods which allows you to use ChatGPT in your terminal: go install github.com/charmbracelet/mods@latest This is useful for examining files, getting code reviews and extracting errors from logfiles or stack traces.
What is Retrieval Augmented Generation (RAG)?
Просмотров 1295 месяцев назад
I demonstrate and explain Retrieval Augmented Generation using a live, custom pipeline I've built to ingest the MDX files in my personal blog. Here's the link to my full tutorial: zackproser.com/blog/langchain-pinecone-chat-with-my-blog
How to use LangChain in a Jupyter Notebook
Просмотров 6205 месяцев назад
In this video I step through using LangChain for the data ingest steps within a Jupyter Notebook that I've open-sourced here: github.com/zackproser/ingest-portfolio I explain how to use LangChain to load MDX documents, how to handle chunking, embedding and upserting and how to sanity check your data quality using Jupyter and the Pinecone console when building your knowledgebase.
How to build a Chat-with-your-data RAG pipeline - Part 1
Просмотров 1935 месяцев назад
In this complete tutorial with code, I step through how I built the chat with my blog experience at zackproser.com/chat using Vercel's latest AI SDK, LangChain and Pinecone's vector database. I provide the data ingest code in a Jupyter Notebook that you can use to accomplish the same thing, and I provide and step through the backend API route code as well as the user interface code. You can fin...
Building an AI chatbot with Vercel's ai-chatbot template
Просмотров 926Год назад
Building an AI chatbot with Vercel's ai-chatbot template
How to use Jupyter notebooks, langchain and Kaggle.com to create an AI chatbot on any topic
Просмотров 745Год назад
How to use Jupyter notebooks, langchain and Kaggle.com to create an AI chatbot on any topic
Building an AI chatbot with langchain, Pinecone.io, Jupyter notebooks and Vercel
Просмотров 952Год назад
Building an AI chatbot with langchain, Pinecone.io, Jupyter notebooks and Vercel
Reviewing GitHub Pull Requests in your Terminal
Просмотров 6 тыс.Год назад
Reviewing GitHub Pull Requests in your Terminal
Dev Diary: #neovim, tmux and Gruvbox for hacking on #aws, #terraform and #golang
Просмотров 758Год назад
Dev Diary: #neovim, tmux and Gruvbox for hacking on #aws, #terraform and #golang
Tea Tutor - a quiz service delivered over SSH
Просмотров 1162 года назад
Tea Tutor - a quiz service delivered over SSH
Can you add link to notebook in the comments
Hi, Can you share the link to the Notebook?
Please provide document llink to understand it more easy and it takes less time to users to know more about langchain and pinecone
Hi, thanks for your comment. Did you see the linked Notebook in the comments?
Hello, I could not find the document link. Please provide me here or add it in video description
That's great content. How would you make the model have a memory on the chat?
Thanks so much for your feedback 😃 Great question - the TLDR is that you keep an ever expanding array of messages and pass them back and forth between the LLM and user each time. I may add an example of this in the future. You could also use a vector db to store the history and query it at inference time....
Your tutorial was very helpful. Please keep up the good work.👍
Thanks so much 🙏 I will!
Great tutorial. I'm curious, how do i store all the messages from users and AI 1. User send a message <- store as HumanMessage 2. Bot reply <- store as AIMessage In this case, should i store all the messages in mongodb and pinecone (for memory)? because i want the chatbot to know who is he speaking to.
Hi @naufal-yahaya - thanks for your support and for your question! Yes, I've recently spoken with a Pinecone developer who is doing exactly that - he shared that vector databases make an excellent place to store conversational history, because retrieval is so fast and accurate, and because you can skip having to send all that data back and forth each time.
OpenAI v2 has vector store feature now. It automatically splits into chunks and creates embeddings. Is there a way to use that instead of Pinecone and langchain?
Hi @RajPatel-d4u and thanks for your question! Ah I wasn't aware of that yet, but it makes sense and I'm guessing it's an extension of the vector datastore they already had for processing the documents of the custom GPTs - yes, so long as their API supports query methods, you should be able to swap that in instead. I may do another video in the future examining that in more detail.
While creating the RetrievalQA, it shows this error of not being able to instantiate abstract class BaseRetriever with abstract methods _aget_relevant_documents, _get_relevant_documents
Hi, did you use the same Notebook I linked?
@@zackproser no , I'm using the same code for rag pipeline
Is it solved ?
Great Vid!!! I'm using a new api key, any idea why I'm hitting: openai.RateLimitError: Error code: 429 - {'error': {'message': 'You exceeded your current quota, please check your plan and billing details. when running PineconeVectorStore.from_documents
check the balance in openai platform... Probably your credit is over
Zachary if I have to add the api key directly not from the environment, then where will I put it in your above code?
Hi Usman, Thanks for your question! Are you saying that you're not able to export an environment variable that contains your key? In a Jupyter notebook host like Google Colab or Kaggle, you can use their secrets integration to set your Pinecone or any other API key. You then reference the secret using their library. Here's a link to a ton of example notebooks where we demonstrate this pattern: github.com/pinecone-io/examples Let me know if that's what you mean or not! Best, Zack
@@zackproser I want to ask that setting the pinecone api key in an environment variable is the only way to include the pinecone api key in the code. can i make a variable as api_key and equals it to actual api key in the code and then inset it in the pinecone???
@@usmantahir2609 you could also hard code your API key in your call to instantiate the Pinecone client, but I wouldn't recommend that from a security perspective
@@zackproser I checked about hard code of API but it was not working as expected., can you share the doc for that ? thanks
@@roopeshk.r3219 @zackproser exactly I am also facing this problem @roopeshk.r3219 can u tell me ur linkedin?
example tutorial, very clear and useful
Thanks so much 🙏 Glad you found it useful. Stay tuned for more.
Really great tutorial, thanks a lot!
Thanks so much for the feedback and support 🙏 Glad it was useful. LMK what else you'd like to see in the future.
@@zackproser I still haven't thought on a specific issue, but I will be glad to update you when I'll have...
Please how you build your UI ?
The full demo and tutorial is available at zackproser.com/chat. Check the link at the top. My site is also open source at github.com/zackproser/portfolio My site is Next.js and Tailwind CSS deployed on Vercel.
Thanks a ton for the video! Does anything changed from the time it was published? I want to use your workflow in exactly the same way, bc I like how it looks. Please share your experience of how you are working with the prs these days?
Thanks so much for your support 🙏☺️ I haven't changed much with this workflow since publishing the video. Lately however I've been shipping a ton of Next.js mostly as a solo dev, so I also use GitHub's native UI. Let me know if there are other topics you'd like to see covered in the future!
@@zackproser thank you for the answering, - I would really like the debugging of js (Next.js/React.js etc.) code to be covered in the future. Another topic which is very interesting - is writing tests and testing flow in neovim. Btw, it would be great if you leave some links on your dotfiles to have the same config as yours, it's pretty handy for the community, and allows to follow quickly, trying new things. Wishing you all the best of this world, waiting for the new useful videos from you, bro! Cheers!
No, mods does not use ChatGPT. According to the repo's readme it uses the openai API endpoint, which is a huge difference from a cost perspective. Sadly, I clicked on the video hoping to find an actual implementation of ChatGPT for command line.
Thanks for your comment 🐱
Is the GPT used here in mods GPT 3.5, GPT 4.0, or GPT 4o?
It's set to 4o on my system, but mods allows you to quickly switch between OpenAI and also local open source models
insane
Glad you liked! There's ton more where this came from. What else are you interested in learning?
@@zackproser tools related to devops/automation/productivity
@@sanusihassan7125 Awesome - thank you - stay tuned, I've got plenty on those topics
@@zackproser subscribed!
@@sanusihassan7125 Woot! Thank you so much - tell all your friends :)
Bro its too short video Why didn’t u share ur experience of using it?
Thanks for the feedback! I'm planning on adding a series soon showing how to auth as well as code review, auto-commit messages and more...stay tuned.
This is neat. Would be great to also see, how the auth flow is handled
Great idea - and I agree - I'll address that shortly in a follow-up video. I can also show how to swap between different models locally.
What similar topics are you interested in? Let me know! I'm planning to shoot my next video shortly.
thats sickkk!!! keep posting
Thanks so much, will do 😄
Thanks !
You're very welcome 🙏
I love to interact with software in terminal and I'm using both tools. This integration via 'C' keymap is amazing 🤩
Can you make a video on explaining modifying the authentication login with next.js? Like replacing GitHub with google and email?
80s synth + neovim + octo == ❤
I agree :) thanks so much!
I am digging into Octo and gh dash myself to use it more in my workflow. Thanks for the video!
Very nice - glad you liked it!
Thank you so much, Zach! Your video really helps me learn the process of setting up my own chatbot. A huge thumbs up. Hope to see more useful content from you.
Faisal - thank you so much for taking the time to leave these kind words of encouragement - they truly mean a great deal to me! I enjoy sharing this content because I find it fascinating but also I do it to help out other developers - since so many have helped me along the way as well. Happy building - if we can help you out with anything related to using Pinecone - please feel free to join us in the community forums to chat through issues / ideas or just meet other builders working on similar things: community.pinecone.io/ Best, Zack
Nice content! Have been looking for use of the vercel ai chatbot template. Do you have public repo of the project?
Thanks so much! I do - check out github.com/zackproser/office-oracle. I have also open sourced the Jupyter notebooks I used for this project
@@zackproser If you figure out a good way to integrate langchain agents and streaming into this template - I have seen a lot of posts with people having issues with it. It would be a great video.
@@roberth8737 Excellent feedback, thank you. I'm going to dig into langchain agents shortly here myself as well as meet with a friend who has been experimenting lately. Hopefully will be able to do something around this soon. Really appreciate the heads up!
@@roberth8737 Good news - my colleague just created this Vercel template which leverages the streaming response pattern: vercel.com/templates/next.js/pinecone-vercel-ai - it's available on GitHub here: github.com/pinecone-io/pinecone-vercel-starter - but it doesn't use LangChain agents. If you had the video of your dreams, what would we be demonstrating start to finish? Building a similar Vercel template that used LangChain agents to accomplish a certain kind of task? Feedback / ideas very welcome!
Wonder of gpqt engineer can do this?
Good question. I haven't tried gpqt engineer. Is it on GitHub?
Thanks mate
Any time!
Have you get the error when you tried to add comment for instance to the PR? because for all changes to PR I'm getting error. exec: "gh": executable file not found in $PATH
Ah, that error suggests you need to install the gh tool, which is GitHub's official command line interface. You can find it at github.com/cli/cli
If you already have it installed, then that error suggests it may be missing from your PATH.
@@zackproser yep, it’s installed, I can use gh command
@@zackproser I’ve just found that gh cli do not recommend install gh by snap what I did, I’ll try to reinstall with another manager
@@Krishtal3 Awesome! Glad you got it working and thanks for leaving this note. It will help others in the future.
That is amazing, thanks
🙌 Glad it was helpful 😄
Hi, Zachary, could you please share your dotfiles? thanks you!!
Hi, @cristiandelgado9987 - absolutely! Thanks for asking. I'll get them into a public repository shortly and share the link back here once it's ready