Auto-GPT, AgentGPT, Neo4j in LangChain & Graph Transformers with O(N) Complexity

Statistical Plotting with Julia: AlgebraOfGraphics.jl [Open in app]( or [online]() [Auto-GPT, AgentGPT, Neo4j in LangChain & Graph Transformers with O(N) Complexity]( Statistical Plotting with Julia: AlgebraOfGraphics.jl Apr 21 [Share]( 👋 Hey, "The true benefit of AI is not just in automation, but in augmentation - making humans smarter, more productive, and more creative." - [Andrew Ng, Founder of DeepLearning.AI, Founder & CEO of Landing AI.]( Can you imagine unleashing your creativity to its fullest potential with AI? With its ability to generate unique strategies, optimize workflows, and provide new perspectives, AI can be the ultimate tool for any creative mind. Together, humans and AI can push the boundaries of what is possible and create truly groundbreaking works of art. This week in [DataPro#40]( we have an exciting lineup of topics for you to explore. We'll be delving into the exciting world of [AI Agents]( discovering how [ML graph technology]( can automate the process of identifying data relationships and [comparing different GPUs](. Plus, we have a [special serialized module]( for you where we'll be discussing what [Petar Veličković, the lead research scientist at DeepMind]( has to say about graph neural networks and graph representation learning. So, get ready to dive in and learn something new! This Week’s Key Highlights: - [Auto-GPT, BabyAGI, And AgentGPT: How To Use AI Agents]( - [Will There Be a GPT-5? When Will GPT-5 Launch?]( - [Statistical Plotting with Julia: AlgebraOfGraphics.jl]( - [Integrating Neo4j into the LangChain ecosystem]( - [How to Build Graph Transformers with O(N) Complexity]( Cheers, Merlyn Shelley Editor in Chief, Packt Don't Miss Out! Take part in our State of Tech survey & shape the future of the tech industry! Share your insights on trends, challenges, & the impact of AI & LLMs. Be the first to get the report & enter for a chance to win a $100 Amazon gift card! [Complete the Survey]( Most Forked Repos on GitHub Prompt Tuning for LLMs - [Promptify]( - Easily generate NLP prompts for popular models like GPT and PaLM with Promptify's prompt engineering. - [OpenPrompt]( - A framework for prompt-learning using pre-trained language models for NLP tasks, supporting [huggingface transformers]( and offering flexibility. - [auto-cot]( - Auto-CoT reduces manual effort in thought prompt design and performs as well as, or better than, manual design using increased diversity. - [EasyInstruct]( - EasyInstruct is a Python package for training LLMs like GPT-3, designed for easy use and extension in research experiments. Frameworks for Training - [Apache MXNet]( - Lightweight, Portable, Flexible Distributed/Mobile Deep Learning with Dynamic, Mutation-aware Dataflow Dep Scheduler. - [Kedro]( - Kedro is an open-source Python framework for creating reproducible, maintainable and modular data science code. - [VectorFlow]( - A minimalist neural network library optimized for sparse data and single-machine environments. - [ColossalAI]( - An integrated large-scale model training system with efficient parallelization techniques. [Pledge your support]( In the Industry AWS - [ML-backed decisions in near-real time with Amazon SageMaker Feature Store and Amazon MSK]( Amazon SageMaker Feature Store provides a fully managed central repository for [machine learning (ML) features]( simplifying the process of designing and building features for near-real-time ML predictions. It enables secure storage and retrieval of features, batch and streaming ingestion, and high-performance retrieval of latest feature values for accurate online predictions. This eliminates the need for businesses to build and maintain their own infrastructure and enables data scientists to focus on building great ML models. - [Automate discovery of data relationships using ML and Amazon Neptune graph technology:]( Organizations are shifting towards a data product approach, using strategies such as data mesh to create value at scale. This post outlines a blueprint for creating smart recommendations by linking similar data products using graph technology and ML with [Amazon Neptune](. The process involves augmenting a data catalog with additional metadata, enabling data consumers to easily identify new datasets and fostering agility and innovation. - [Deploy large models at high performance using FasterTransformer on Amazon SageMaker:]( As businesses explore new applications for generative AI in text, image, audio, and video generation, deploying these models can be challenging. The article explores how [Amazon SageMaker's]( large model inference deep learning containers ([LMI DLCs]( can optimize models, with [code examples]( for deployment. Google Cloud - [Scaling deep retrieval with TensorFlow Recommenders and Vertex AI Matching Engine]( The blog covers building a playlist recommendation system using Vertex AI, starting with an explanation of two-tower encoders and their popularity in retrieval modeling. They use the [Spotify Million Playlist Dataset]( to frame the use case and develop custom two-tower encoders using the [TensorFlow Recommenders library](. Finally, candidate embeddings are served in an approximate nearest neighbors (ANN) index with[Vertex AI Matching Engine](. Code can be found on [GitHub](. Spark of Infographics [Integrating Neo4j into the LangChain ecosystem:]( LLMs are being turned into agents with the ability to interact with their environment through various integrations. [LangChain]( is a popular library that enhances LLMs by giving access to external tools and data sources to improve responses and manipulate the environment. Just for Laughs! Why did the AGI get lost in the forest? Because it tried to use a graph neural network to map the terrain, but forgot to account for the trees! LLMs and GPTs: What's the buzz? - [Will There Be a GPT-5? When Will GPT-5 Launch?]( While some sources have predicted a GPT-5 release in early 2024, OpenAI CEO Sam Altman has disputed these [claims]( stating that GPT-5 is not currently being trained and won't be for some time. Keeping in mind the frequency of recent developments the last quarter of 2024 is still a possibility for a release and could bring us one step closer to artificial general intelligence. - [Auto-GPT, BabyAGI, And AgentGPT: How To Use AI Agents:]( Applications such as [Auto-GPT]( [AgentGPT]( [BabyAGI]( and [GodMode]( use OpenAI's large language models to automate tasks with ChatGPT. Unlike ChatGPT, these AI agents only need a general objective to operate, not a prompt for every action. Although these applications are experimental, they have already accomplished remarkable tasks. However, users must exercise caution when sharing personal or sensitive information, as it [may be utilized to enhance the model](. It is very essential to use AI agents responsibly. Streamline your Data Strategy with Experts [Creators of Intelligence]( – By [Dr. Alex Antic]( We appreciate all the feedback you have shared about the book so far! As promised, we're excited to provide you with a sneak peek of insights from [Petar Veličković]( one of the leading research scientists at [DeepMind](. Alex Antic: While it's in many ways natural to think about data in a graph representation manner, what are some of the limitations of graph representation learning, relative to other methodologies? Petar Veličković: Of course, the field is not without its limitations. You get all that freedom of modeling when you’re dealing with irregular structures, but it obviously comes at a cost. If there is some interesting regularity in your data, you now have to work twice as hard to account for it in your model, because it is not guaranteed by the implementation. Recently, there's been a lot of work on topologically inspired graph neural networks that try to include some information about the regularities of the input data that a basic graph neural network might discard. It’s a bit of a fun mathematical journey to try to make GNNs more specialized. In their basic form, they're very general. If you were to apply them directly and naïvely to image data, they would probably underperform compared to convolutional neural networks, as convolutional networks are highly specialized for the structure of images. That is maybe one of the main points. [Find out more about the book here]( This exclusively curated content is extracted from the upcoming book “[Creators of Intelligence]( by [Dr. Alex Antic.]( [EXPLORE NEW IDEAS & READ ON!]( Featured Posts in the Data and ML Community - [Statistical Plotting with Julia:]( The [AlgebraOfGraphics.jl]( package in Julia uses the theoretical concept of the Grammar of Graphics to create statistical plots. It separates different building blocks like mappings and geometries and allows for more modularity in creating visualizations. - [How to Build Graph Transformers with O(N) Complexity:]( The graph machine learning community is exploring the use of pure Transformer-based models, which can outperform GNNs on some benchmarks. The challenge is the quadratic complexity of the attention mechanism. This tutorial introduces two recent scalable graph Transformers with linear complexity and provides guidance on their implementation. - [Pro GPU System vs Consumer GPU System for Deep Learning:]( The article compares consumer-level GPUs to higher-end GPUs used for deep learning development and inference. The main differences are in GPU RAM and system scale, which are crucial for advanced deep learning models. Professional-level GPUs are needed for larger systems and high GPU RAM requirements. - [Streamlit Tutorial: Creating Word Reports for Data Science Projects:]( This article discusses using [python-docx]( and [Streamlit]( to automate the process of creating data science reports. It outlines the benefits of using python-docx to create a Microsoft Word report, and introduces the libraries required for this task, including Streamlit, [pandas]( [matplotlib]( and python-docx. The article also highlights the potential of integrating Large Language Models (LLMs) to enhance the capabilities of the app. Share Your Views If you’re interested in sharing ideas to foster the growth of the data community, then this survey is for you. Consider sharing your thoughts and get a FREE bestselling Packt book, The Applied Artificial Intelligence Workshop as PDF. Jump on in! [TELL US WHAT YOU THINK]( As a GDPR-compliant company, we want you to know why you’re getting this email. The _datapro team, as a part of Packt Publishing, believes that you have a legitimate interest in our newsletter and its products. Our research shows that you opted-in for email communication with Packt Publishing in the past and we think your previous interest warrants our appropriate communication. If you do not feel that you should have received this or are no longer interested in _datapro, you can opt out of our emails by clicking the link below. [Like]( [Comment]( [Restack]( Read Packt DataPro in the app Listen to posts, join subscriber chats, and never miss an update from Packt SecPro. [Get the iOS app]( the Android app]( © 2023 Copyright © 2022 Packt Publishing, All rights reserved. Our mailing address is:, Packt Publishing Livery Place, 35 Livery Street, Birmingham, West Midlands B3 2PB United Kingdom [Unsubscribe]() [Start writing]()