Shashikant Singh’s Post

The best way to learn is to experiment and tinker. Take out some time and follow this awesome quick-start that will show you how to build a Streamlit app that uses GenAI techniques like Prompt Engineering and Retrieval Augmented Generation (RAG) by leveraging Snowflake Cortex. https://lnkd.in/dPiYygRQ

Building a RAG? Tired of chunking? Maybe Vision is All You Need! My serverless ColPali V-RAG (Vision Retrieval-Augmented Generation) demo, the code and the accompanying blog post are now finally published. This is a full end-to-end Applied AI engineering demo, from GPU inference through an HTTP API to React UI, not just some throwaway Jupyter Notebook. The demo is implemented on Modal and utilizes various tools, including Qdrant. The ColPali model from ILLUIN Technology is at the core of the whole demo! Check out the blog post and repository, links are in the comments.

Imagine: two days of hands-on experimentation & learning about SWEngineering techniques working with LLMs... <ahhhhh>. That describes Thursday and Friday in Austin w/ John Willis and Patrick Debois. We are a small, focused group of 14 techies all debating, discussing, and debugging. RAG, embeddings, vector dbs, functions, callbacks and debugging, security concerns and ideas for addressing, multimodal, gptscript, langchain, using OpenAPI to expand capabilities... so much content! We could probably use 5 days minimum to really explore in depth! #ContinuousLearning #GenAI #SWE4GenAI #ArchAITecture

Today was #Day132 of #180DaysOfCodeChallengeLeetCode where I solved the question - Robot Collisions 👨🏻💻🚀 #CodingChallenge #LeetCode #ContinuousLearning #Algorithm #DataStructures #CareerDevelopment

#Grok: Exciting news for AI enthusiasts and developers! The xai-org has released an open-weights model called Grok-1, which is now available on GitHub. This model represents a significant advancement in the field, boasting a 314 billion parameters. It’s designed to be loaded and run using JAX, a high-performance machine learning library. Due to massive billions of parameters and other factor GPU necessary to run model (or Serverless from GPU supported provider) For those diving deeper , the weights of model available to download through torrent client and magnet link available in Git repo.

#45daysDSA-Revison Challenge Excited to invite you all to join me in the 45-day DSA-Revison Challenge! 🚀 Let's revise the most important DSA questions together, covering key topics like Array, Linklist, Stack, Queue, Graph, and Tree. 📚 Time to enhance our skills and tackle this challenge head-on! 💪🌟 #DataStructures #Algorithms #TechChallenge

🚀Day 60/100🚀 📌Sharing my Day 60 with you all!💡 100 day DSA Challenge on #leetcode100days Problem:-Rotate Image #day60 #datastructures #leetcodechallenge #algorithms #learningeveryday. #edslash #javaprogramminglanguage #dsawithedslas

LazyLLM is a low-code development tool for building multi-agent LLMs(large language models) applications. It assists developers in creating complex AI applications at very low costs and enables continuous iterative optimization. LazyLLM offers a convenient workflow for application building and provides numerous standard processes and tools for various stages of the application development process. Features: - Easy AI Assembly: Assemble AI applications like Lego with built-in modules, no deep knowledge needed. - One-Click Deployment: Deploy complex multi-agent apps easily during POC and package images for Kubernetes in production. - Cross-Platform Compatibility: Seamlessly switch IaaS platforms without code changes, supporting various environments. - Grid Search Optimization: Automate hyperparameter tuning for efficient application optimization. - Efficient Fine-Tuning: Simplify model fine-tuning, focusing on algorithms and data, with minimal engineering effort. GitHub: https://lnkd.in/gkUh97kx https://lnkd.in/gff8N-gA

"Day 13 of 120 in my DSA Challenge" On LeetCode On Day 13 Solved Today's brain teaser:-✅💫 Problem 632: Smallest Range Covering Elements From K Lists Problem 2: Add Two Numbers Problem 3: Longest Substring Without Repeating Characters Platform:- LeetCode 🌌Just emerged from an enriching DSA journey! The complexities of algorithms have turned into exciting puzzles to solve. #DataStructures #Algorithms #CodingJourney

Thanks to Jiří Materna for organizing yet another year of Machine Learning Prague conference! Anyone considering whether to attend or not should think no more and get ready for next year. Talks are engaging and practical, after an hour I had my mental backlog of ML stuff full for the next year. Thanks for your work Jiří and kudos to great speakers! #agrpdev #justcode