Philip Rathle, the Chief Technical Officer of Neo4j, the popular graph database company which has now taken off by storm because of GraphRag, a new approach for making LLM Retrieval Augmented Generation applications more accurate by leveraging graphs, so you know today will be all about GraphRag and its impact on the market.


Chapters:
00:00 Intro
02:09 Is AI Resurgence of Graph tech?
03:46 GraphRAG popularity
05:39 Top Use Cases in GenAI
11:08 Gen AI in supply chain
16:46 Graph and its types in enterprise
24:03 GraphRag
25:25 GNNs in GraphRAG
29:30 Graphs are eating the world
35:16 Knowledge Graph
36:06 Drawbacks of vector based rag
37:43 Neo4j vector database
41:27 Filtering with Knowledge Graph
45:02 Execution Time of LLMs
49:03 Does longer prompts mean longer graph query?
54:26 Scale of Graph
57:05 Marriage of Graphs and Vectors
59:46 Fine Tuning with Graphs
01:00:46 Graphs Use less tokens
01:02:46 Multiple vs One GraphRAG
01:05:38 Updating Knowledge in Graph
01:10:50 large Vs small models
01:13:09 MultiModal GraphRAG
01:15:36 Graphs in Robotics
01:17:11 Neo4j journey
01:20:03 Phillip Linkedin Post
01:21:56 What's different with AI
01:23:31 Advice for Gen AI startups
01:26:00 CTO advice
01:29:36 Chemical Engineering
01:32:00 Career optimization function
01:35:00 Book Recommendations
01:37:06 Rapid Round

Get 1000 free inference requests for LLMs on build.nvidia.com
Kyle Kranen, an engineering leader at NVIDIA, who is at the forefront of deep learning, real-world applications, and production. Kyle shares his expertise on optimizing large language models (LLMs) for deployment, exploring the complexities of scaling and parallelism.

📲 Kyle Kranen Socials:
LinkedIn: https://www.linkedin.com/in/kyle-kranen/
Twitter: https://x.com/kranenkyle

📲 Mark Moyou, PhD Socials:
LinkedIn: https://www.linkedin.com/in/markmoyou/
Twitter: https://twitter.com/MarkMoyou

📗 Chapters
[00:00] Intro
[01:26] Optimizing LLMs for deployment
[10:23] Economy of Scale (Batch Size)
[13:18] Data Parallelism
[14:30] Kernels on GPUs
[18:48] Hardest part of optimizing
[22:26] Choosing hardware for LLM
[31:33] Storage and Networking - Analyzing Performance
[32:33] Minimum size of model where tensor parallel gives you advantage
[35:20] Director Level folks thinking about deploying LLM
[37:29] Kyle is working on AI foundation models
[40:38] Deploying Models with endpoints
[42:43] Fine Tuning, Deploying Loras
[45:02] SteerLM
[48:09] KV Cache
[51:43] Advice for people for deploying reasonable and large scale LLMs
[58:08] Graph Neural Networks
[01:00:04] GNNs
[01:04:22] Using GPUs to do GNNs
[01:08:25] Starting your GNN journey
[01:12:51] Career Optimization Function
[01:14:46] Solving Hard Problems
[01:16:20] Maintaining Technical Skills
[01:20:53] Deep learning expert
[01:26:00] Rapid Round

Kaggle Grandmaster Chris Deotte, he is currently ranked 1 on notebooks and discussions on Kaggle and is part of the KGMON team, Kaggle Grandmasters of NVIDIA. We’ll be discussing GEN AI and personalization, optimizing your kaggle game and other strategies to make progress in your career.

Solution: https://arxiv.org/pdf/2408.04658

Mark Moyou, PhD Socials:
LinkedIn: https://www.linkedin.com/in/markmoyou/
Twitter: https://twitter.com/MarkMoyou

Chapters:
00:00 Intro
01:51 Current Gen AI
04:40 Evolution of Conceptualization in ML Models
06:59 Measuring Tonality in Data Sets
08:51 Multi-Modal Data Sets in Text Based Models
11:56 Large Vs Small Language Models
13:46 KDD 2024 Competition
23:28 Prompt Formatting and Bribing the Model
28:08 Qwen2 Vs LLama
30:39 WiSE - FT
33:53 LoRA on all the layers
35:43 Logit Preprocessor
42:05 Personality of Small Vs Large Model
44:02 Models Understanding Shopping Concepts for E-Commerce
47:26 Offline Purchase Data in E-Commerce Personalization
55:56 Navigating the Problem with Required Data
58:33 Constraining LLM Output
01:00:45 LLMs in Search and Personalization
01:02:03 Kaggle Grandmaster
01:09:45 Gen AI in Kaggle Competition
01:13:07 Learning ML in Non-Traditional Way
01:16:15 Thoughts on doing PhD
01:17:58 Mathematics
01:22:22 Advice for PhD students
01:24:32 Hardest Kaggle Competition
01:27:32 Level of Grit in Competitions
01:32:59 Career Optimization Function
01:35:00 Management vs Technical IC Roles
01:37:27 Making Progress
01:39:48 Book Recommendations
01:44:43 Thoughts on Writing Book
01:46:20 Advice for High Schooler, College Students and Professionals
01:52:20 Rapid Round