Yujian Tang’s Post

🚀 Exciting News in AI! 🚀 I'm thrilled to share a project that I've been working on a Retrieval-Augmented Generation (RAG) System tailored for the latest advancements in language understanding, demonstrated on the "Leave No Context Behind" paper published by Google on April 10, 2024. 🔍 About the Project: Large Language Models (LLMs) like Gemini 1.5 Pro are incredible, but they often miss out on specific contextual information crucial for businesses and technical fields. To address this, my project integrates LLMs with real-time data from external sources such as PDFs and text files. This integration enhances the model's ability to leverage the most up-to-date information, ensuring no context is left behind. 👨💻 Technical Insight: Utilizing the LangChain framework, this RAG system retrieves external documents—in this case, the profound insights of the "Leave No Context Behind" paper—and dynamically incorporates this data into Gemini 1.5 Pro’s response generation process. This approach significantly enriches the answers provided by the AI, making it a powerful tool for research and information retrieval. 🙌 Acknowledgements: A massive thank you to Innomatics Research Labs for providing me with the opportunity to work on such groundbreaking technology. Special thanks to Kanav Bansal Sir for his invaluable guidance and mentorship throughout this project. 🔗 Explore the Code: For those interested in diving deeper into the technicalities, you can find the complete code and documentation on my GitHub: https://lnkd.in/eBf7w4gv Feel free to connect, discuss, and collaborate! #AI #MachineLearning #DataScience #LangChain #Google #Research #InnomaticsResearchLabs #GenAI #API

The Final Countdown? Project Q* is a secretive project by OpenAI. It’s been referred to as a mysterious AI breakthrough that could potentially 'threaten humanity’. The project has been creating ripples in the AI community. Here’s what we know about Project Q* so far: Project Q* was a new model that excelled in learning and performing mathematics. It was reportedly only at the level of solving grade-school mathematics, but as a beginning point, it looked promising for demonstrating a previously unseen intelligence from the researchers involved. The existence of Q* was reportedly scary enough to prompt several staff researchers to write a letter to the board to raise the alarm about the project, claiming it could "threaten humanity". However, the Chief AI scientist at Meta, Yann LeCun, tweeted that Q* has to do with replacing “auto-regressive token prediction with planning” as a way of improving LLM (large language model) reliability. LeCun says all of OpenAI’s competitors have been working on, and that OpenAI made a specific hire to address this problem1. Sources: 1. digitaltrends.com 2. tech.co 3. techradar.com

Successfully completed the End-to-End Text-to-SQL LLM App with the help of Google Gemini Pro! 🚀 Check out the GitHub repository for the app: https://lnkd.in/gk_Gym62 🌟 Feel free to explore my prompts for better responses from the Gemini model. Gemini sets a new standard by outperforming human experts on MMLU (Massive Multitask Language Understanding). A big shoutout to Krish Naik sir for his mentorship and invaluable insights throughout this journey. Your expertise has been instrumental! 🙌🌟 Explore the app, dive into the code, and share your thoughts! 🚀💻 #GenerativeAI #GeminiPro #AI #NLP #GitHub

Groundbreaking news from the AI world! Google DeepMind has done the unthinkable - solving a math problem that was deemed unsolvable! 💥 The Big Picture This isn't your everyday AI accomplishment. We're witnessing a giant leap in problem-solving capabilities, blurring the lines between artificial intelligence and human expertise in pure mathematics. 🔥 The Takeaway DeepMind's success is a wakeup call: AI's potential is enormous and rapidly growing. It's time to rethink how we leverage AI in our fields. The actionable step? Start exploring AI solutions to tackle complex, unsolved problems in your industry. This is the moment to innovate and lead with AI-driven strategies. Dive into this MIT Technology Review article for the full story. https://lnkd.in/eC2z2yVv #AIRevolution #DeepMind #InnovationStrategy

Hey there, #LangChain fam! 👋 Ever wanted to build your own Retrieval-Augmented Generation (RAG) system with some ✨adaptive capabilities✨? Then this video by Lance from LangChain is for you! 🚀 In the video, Lance dives into the world of RAG systems, exploring two key concepts: query analysis and flow engineering (aka adaptive RAG). Query analysis helps refine your questions to enhance retrieval, while flow engineering incorporates checks throughout the RAG pipeline to ensure things like document relevance and minimize hallucinations in the generated answer. 🔎 The video also introduces a new model to the mix: Command-R. This small and speedy 35B parameter model is a great fit for RAG tasks due to its ability to handle tool use and query writing. 💪 Overall, this video offers a fantastic intro to building adaptive RAG systems using LangChain and Command-R. If you're interested in creating RAG systems that leverage query analysis and flow engineering to boost retrieval accuracy, this video is a must-watch! Check out the full video here: https://lnkd.in/d3jZZVcE LangGraph code: https://lnkd.in/d_hxinee ReACT agent code: https://lnkd.in/d-a7gjPb Paper: https://lnkd.in/d6SThFYP #Rag #LangChain #AI #MachineLearning #NLP

While Low-Rank Adaptation (LoRA) has gained significant traction in the realm of Large Language Models (LLMs), its utility extends far beyond this domain. Sharing my latest Medium article on Low-Rank Adaptation (LoRA) and the corresponding GitHub repository to help fellow enthusiasts test LoRA beyond LLMs. In the article, I delve into the technical details of LoRA, exploring its applications in fine-tuning deep learning models for specific image classification tasks. The GitHub repository, "LORA_Playground (https://lnkd.in/gGm-xbpf)" contains simpler examples with notebooks for hands-on learning.

I am happy to share my latest project on NLP which is Cyberbullying Detection. As we know With rise of social media coupled with the Covid-19 pandemic, cyberbullying has reached all time highs. We can combat this by creating models to automatically flag potentially harmful tweets as well as break down the patterns of hatred. I have experimented with multiple machine learning and deep learning models for this task such as Logistic Regression , Random Forest Classifier , and LSTM. Feel free to see my work. https://lnkd.in/dmrkxtKE #machinelearning #deeplearning #datascience #nlp

Responsible AI part 3 of 3 - In this final part of the series, we’ll delve into tools and frameworks related to artificial intelligence (AI). We’ll explore Responsible AI use cases across different industries. Let’s begin! 

I've embarked on a learning journey with the LLM Zoomcamp by DataTalks.Club! This bootcamp is an incredible opportunity to dive deep into the world of Large Language Models (LLMs), and I'm eager to learn and grow alongside an amazing community on Slack. Week 1 was an introduction to LLM and RAG, we focused on understanding the basics of LLMs and Retrieval-Augmented Generation (RAG), and implemented a simple RAG pipeline to answer questions from the FAQ documents of our Zoomcamp courses. Here’s a brief summary of what we covered: - Understanding what LLMs and RAG are. - Exploring the RAG architecture. - Utilizing the MinSearch search engine to index documents and perform searches. - Invoking the OpenAI API to build prompts and get answers. - Exploring alternatives to the OpenAI API for local LLM execution. - Running ElasticSearch with Docker, indexing documents, and replacing MinSearch with ElasticSearch. The first week has been incredibly insightful and has laid a solid foundation for the upcoming modules. Stay tuned for more updates ! 🚀 #llmzoomcamp #DataTalksClub #AI #NLP #LearningInPublic #LLMs