Activity
Mon
Wed
Fri
Sun
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
What is this?
Less
More

Memberships

AI Developer Accelerator

10.8k members โ€ข Free

3 contributions to AI Developer Accelerator
Is Langchain not for production?
Hello everyone, I am a full stack MERN developer, now trying my hands on building some GEN AI RAG applications. As a beginner, as I am totally in JS environment, so I choose to go with Langchain JS/TS SDK, to go forward, and I created a project also, which is somewhat chat with youtube videos. But the thing is yesterday I read some articles which are saying Langchain is not for production. They have different points for this - 1) Langchain use tons of packages which most of us don't need so it's will only make application heavy. 2) Langchain fails at security levels. 3) Langchain changes rapidly, so not good for prod applications to change code every week. and many more... Some Links - https://medium.com/@aldendorosario/langchain-is-not-for-production-use-here-is-why-9f1eca6cce80#:~:text=When%20there%20is%20an%20issue,by%20you%20and%20your%20team! Can anyone give me some suggestions of what to do? My goal is to make a good prod ready application, as well as making my resume stronger to get hired. If I skip langchain then I may need to write come extra codes in JS for character splitting, DB storing etc. Please give some suggestions. Also which stacks should I use to get hired?
0 likes โ€ข Mar '24
@Tim Kitchens sure, will update all of my works here
0 likes โ€ข Apr '24
@Brandon Hancock Will surely Look into it bro
Get text response from Stream response
So I am creating a stream of response from `createRetrievalChain` from Langchain JS. The whole code is `ย const chain = await createStuffDocumentsChain({ llm: model, prompt: prompt, }); const conversationChain = await createRetrievalChain({ combineDocsChain: chain, retriever: retrieverChain, }); // check for question each time const outputChain = RunnableSequence.from([ conversationChain, new RunnablePick({ keys: "answer" }), new HttpResponseOutputParser({ contentType: "text/plain" }), ]); const response = await outputChain.stream({ chat_history: chatHistory, input: question, });`. In the FrontEnd I am using Vercel AI SDK in Next.JS to stream the response. While everything looks good to me and is working properly, but I also want to get the textual response from this Stream response, beacuse I have to store this textual response to DB. How I can do this with the help of Langchain or Vercel AI SDK? You can check the gist for clear code - https://gist.github.com/DeadmanAbir/8187b4436fadfa5aead77169e4638615
0
0
My New Open-Source Project: โ€œCrewAI Visualizerโ€
๐Ÿš€ Exciting News! ๐Ÿš€ I'm thrilled to introduce my latest open-source project: CrewAI Visualizer! ๐ŸŽ‰CrewAI Visualizer is all about simplifying AI crew management with a user-friendly interface based on CrewAI package. As someone deeply passionate about AI and its potential, I wanted to create a tool that makes it easy for anyone to create and manage AI teams without the need for coding skills. ๐Ÿ› ๏ธ Tech Stack: Built using a powerful tech stack, including PostgreSQL, Prisma, Next.js, GraphQL, node-calls-python to execute Python code from Node.js, and the CrewAI Python package to leverage the framework's capabilities in addition to Gemini as LLM. ๐Ÿ” Main Features: - Intuitive UI: Designed for ease of use, allowing you to focus on building your AI crews effortlessly. - Role-Based Agent Design: Customize agents with specific roles, goals, and tools, giving you full control over your team composition. - Task Management: Define tasks and assign them dynamically, ensuring smooth workflow management and efficient task execution. - Sequential and Hierarchical Processes: Choose between sequential or hierarchical processes for task execution, providing flexibility to adapt to different workflow needs. - Save Output: Track output for future reference or analysis, enabling you to monitor progress and outcomes over time. ๐Ÿš€ Future Plans: In the pipeline: - Adding more tools for agents from the LangChain community or creating new useful tools internally. - Expanding LLM options to include ChatGPT and support for local LLMs, providing users with a broader range of AI capabilities. ๐ŸŒŸ I'm continuously working to improve CrewAI Visualizer and add new features, so your feedback and contributions are invaluable. If you're interested in checking out the project, feel free to visit the GitHub repository, give it a star, and maybe even contribute to it! CrewAI GitHub Repo: https://github.com/Eng-Elias/CrewAI-Visualizer
1 like โ€ข Mar '24
Hey, how's your experience using node -call-python-package? Is it really optimized? Can it run complex py functions?
0 likes โ€ข Mar '24
@Elias Owis Does it make backend slow? I mean is it optimized for prod? Like scaling a product?
1-3 of 3
Abir Dutta
2
12points to level up
@abir-dutta-5668
A fellow full stack developer, exploring GEN AI. connect - https://www.linkedin.com/in/abir-dutta-408759223/ https://twitter.com/ItsDutta99

Active 476d ago
Joined Mar 10, 2024
Powered by