Ever wanted to code Graph Transformer (GT) from *scratch* using a few lines of code with PyTorch and DeepGraphLibrary? :) See below my course material Slides: dropbox.com/s/i9coadk7fgzf… GitHub: github.com/xbresson/CS620… Paper: arxiv.org/pdf/2012.09699… Coding GT step-by-step 👇

If new with graph learning, here is a warm-up notebook to learn to use graphs: • Build graph w/ features and compute basic message-passing function w/ DeepGraphLibrary • Convert into graph formats w/ DGL, NetworkX, dense/sparse PyTorch • Visualize graph Code: github.com/xbresson/CS620…

A new offering from the team! Announcing GraphStorm, a low-code framework for enterprise-level graph machine learning. See how it empowers GML in business 👇aws.amazon.com/blogs/machine-…

Will give a talk tomorrow Sunday 10am at the workshop Graph Learning Benchmarks Workshop on Graph Learning Benchmarks #KDD2023 w/ Yizhou Sun, Jimeng Sun, Z. Da DeepGraphLibrary, A. Wang deep-learning-graphs.bitbucket.io/dlg-kdd23 Thanks to the organizers! Also glad to meet and chat about research during the conference 😀

Da Zheng DeepGraphLibrary "Graph machine learning for industry applications with DGL and GraphStorm" #KDD2023

Amazon has publicly released RefChecker, a combination tool and dataset that detects hallucinations in #LLMs. To characterize factual claims, RefChecker uses knowledge triplets rather than natural language, enabling finer-grained judgments. #GenerativeAI amazon.science/blog/new-tool-…

Accelerating GNN Training on Intel CPU with DGL through fused sampling & hybrid partitioning, which can give you up to a 2x speedup! 🚀🧠 Awesome work done by Hesham (linkedin.com/in/hesham-most…) and Adam (linkedin.com/in/adam-grabow…). Intel AI Intel Business community.intel.com/t5/Blogs/Tech-…

Say goodbye to data loading bottlenecks! DGL 2.0 introduces GraphBolt, a revolutionary data pipeline framework that supercharges your GNN training. 👉Check out the blog for the release summary: dgl.ai/release/2024/0…

DGL 2.1 with GPU-accelerated GraphBolt delivers blazing-fast data loading & stays flexible for customization. Plus, it works with PyTorch Geometric (PyG)! 👉Check out the blog for the release summary: dgl.ai/release/2024/0… 🧠Major Contributor: Muhammed Fatih Balın #DGL #GML

🎉 DiskGNN ACM SIGMOD/PODS : Trained 300M+ node graphs (TB-scale) with just 1 GPU & <100GB RAM! Zero-overhead GPU/CPU/SSD pooling made it possible 🚀