LlamaIndex’s Post

This new video is an introduction to LlamaIndex and in particular its agentic capabilities

🚀 Dive into the world of Retrieval-Augmented Generation (RAG) with ease! Our latest blog post equips you with the knowledge and code to deploy RAG on LlamaIndex using Friendli Engine, known for its blazing-fast performance and cost-effectiveness. Follow the steps and code provided to seamlessly implement RAG in your applications. Blog: https://lnkd.in/gt755gJQ #rag #llamaindex #friendli #friendliai #generativeai #genai

📢 Another big announcement: Argilla 2.0 release candidate! What does it mean for AI builders? 🤺 Flexible, unified framework for human feedback collection for AI workflows > Define the shape of your data and the questions you want to ask, that's it! 🐍 New Python SDK to configure, read and write datasets, including a powerful integration with Hugging Face datasets > Set up your workspaces, users & datasets with easy to use & learn methods > Read, map and import existing datasets is a breeze with the new log method! 🖼 Flexible UI for data annotation, tailored for domain experts! > Arrange the annotation view as it fits your dataset and use case. > Access key information like progress and guidelines more easily! 📚 New docs! All this, thanks to months of listening to our community and users. These coming weeks, we will collect further feedback before releasing Argilla 2.0 as a stable version, help us! Link to the announcement blog post in the first comment.

What sorting algorithm does std::sort use behind the scenes? Most of us have heard of Merge Sort, Quick Sort, Heap Sort, and other famous sorting algorithms; but the answer is Introsort (Introspective Sort). Introsort is a hybrid algorithm that starts with quick sort and then switches to heap sort when the recursion goes too deep. It also uses insertion sort when the input size gets small (usually smaller than 16). The main issue with quick sort is that the recursion will get too deep if the pivot is chosen poorly and the time complexity will approach O(n^2) in the worst case. Intro sort keeps the worst case time complexity at O(nlogn) by limiting the recursion depth to 2*log_2(n) where n is the input size. This value has been empirically determined to be the optimal choice. In the image below there is a sample implementation of intro sort. As you can see in the benchmark, it is considerably faster than both quick sort and heap sort. Links: - The code I wrote for this benchmark: https://lnkd.in/g6niaTHt - The GCC implementation of std::sort (line 1894): https://lnkd.in/g2zaJXzz #cplusplus #datastructuresandalgorithms #sorting #code

Today was #Day134 of #180DaysOfCodeChallengeLeetCode where I solved the question - Create Binary Tree From Descriptions 👨🏻💻🚀 #CodingChallenge #LeetCode #ContinuousLearning #Algorithm #DataStructures #CareerDevelopment

𝐄𝐱𝐜𝐢𝐭𝐢𝐧𝐠 𝐍𝐞𝐰𝐬: 𝐏𝐢𝐞𝐜𝐞𝐬 𝐂𝐨𝐩𝐢𝐥𝐨𝐭+ 𝐋𝐚𝐮𝐧𝐜𝐡𝐢𝐧𝐠 𝐨𝐧 𝐏𝐫𝐨𝐝𝐮𝐜𝐭 𝐇𝐮𝐧𝐭 𝐉𝐮𝐥𝐲 𝟏𝟎! AI tools are great, but their true potential is unlocked with proper integration and prompts. Pieces for Developers Copilot+ with Live Context offers real-time, on-device contextual assistance, transforming your developer workflow. Check out the prompts that developers can only use with Pieces Copilot+, and help me support their huge Product Hunt launch on July 10! 𝐊𝐞𝐲 𝐏𝐫𝐨𝐦𝐩𝐭𝐬 𝐄𝐧𝐚𝐛𝐥𝐞𝐝: ▪️𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡𝐢𝐧𝐠 𝐚𝐧𝐝 𝐏𝐫𝐨𝐛𝐥𝐞𝐦 𝐒𝐨𝐥𝐯𝐢𝐧𝐠: - “Where did I find that Python script I forgot to save in my research yesterday?” - “Summarize this research paper.” ▪️𝐂𝐨𝐝𝐢𝐧𝐠 𝐢𝐧 𝐭𝐡𝐞 𝐈𝐃𝐄: - “Explain this new repo I’m onboarding into.” - “Summarize the code changes in this PR.” Collaborating with Colleagues: - “When’s the best time to set up a sync with my team based on our shared calendars?” “Summarize what I did today for my PM” ▪️𝐈𝐦𝐩𝐫𝐨𝐯𝐢𝐧𝐠 𝐏𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐯𝐢𝐭𝐲: - “What was I working on yesterday?” - “Analyze my workflow and suggest ways to reduce context switching.” Discover how these prompts can 10x your workflow! Read more in their blog post: https://lnkd.in/dKv7unCm

Today was #Day13 of #180DaysOfCodeChallengeLeetCode where I solved the question - Find the Pivot Integer 👨🏻💻🚀 #CodingChallenge #LeetCode #ContinuousLearning #Algorithm #DataStructures #CareerDevelopment

Day 18 of #100DaysOfCode: I successfully solved the anagram grouping problem using a mapping approach! 🔄📚 The key idea was to group words that are anagrams by sorting each word to create a unique key. By using this sorted key, I was able to map and group all anagrams together efficiently. For example, "eat", "tea", and "ate" were all grouped under the key "aet". This method was both intuitive and effective in organizing the data. Feeling great about deepening my understanding of string manipulation and hash maps with this challenge! #DataStructures #Algorithms #Anagrams #StringManipulation #CodingChallenge #Learning #100DaysOfCode #neetcode

#LeetCodeIsomorphicStringsChallenge Day 36 of LeetCode Challenge This code determines if two strings are isomorphic using ASCII values to record character mappings. Initialize "letterFrom_s" and "letterFrom_t" to [0] * 256 each. The list stores the last seen index of every one character which is represented by its ASCII value in the respective strings "s" and "t". Iterate over each pair of characters from strings "s" and "t". Convert the characters "char_s" and "char_t" to their ASCII values "ascii_s" and "ascii_t" each. Check mapping by verifying if the last seen indices recorded in the "letterFrom_s[ascii_s]" and "letterFrom_t[ascii_t]" are different. If so, return "False". This implies the strings are not isomorphic since the characters do not map rationally. If the loop ends and all character mappings are consistent return "True". This implies the string "s" and "t" are isomorphic. The time complexity is O(n) and the space complexity is O(1). #LeetCodeChallenge #IsomorphicStrings #DataStructures #Algorithms

Today was #Day34 of #180DaysOfCodeChallengeLeetCode where I solved the question - Word Search 👨🏻💻🚀 #CodingChallenge #LeetCode #ContinuousLearning #Algorithm #DataStructures #CareerDevelopment