Anyscale Courses

Scaling up Terascale Deep Learning on Commodity CPUs with ThirdAI and Ray

Explore ThirdAI's BOLT engine for efficient deep learning on CPUs, leveraging Ray Core for distributed training. Learn about overcoming communication bottlenecks and achieving near-linear scaling for terabyte-scale datasets.

• YouTube
• 26 minutes
• Self-Paced
• Free Video

Ray Observability 2.0 - Debugging Applications with New Tooling

Explore new Ray observability tools for debugging ML workloads. Learn to effectively troubleshoot offline and online applications using advanced functionality in Ray and Anyscale.

• YouTube
• 30 minutes
• Self-Paced
• Free Video

How Spotify Built a Robust Ray Platform with a Frictionless Developer Experience

Discover Spotify's journey in building a centralized Ray platform, enhancing reliability, scalability, and developer experience for diverse ML applications and thousands of users.

• YouTube
• 36 minutes
• Self-Paced
• Free Video

Scaling AI Health Assistants: Challenges and Solutions

Explore challenges and solutions in scaling AI health assistants, focusing on model management, state persistence, and system reliability. Learn key principles for building scalable AI products using Ray.

• YouTube
• 28 minutes
• Self-Paced
• Free Video

Forecasting Covid Infections for the UK's NHS Using Ray and Kubernetes

Discover how Ray and Kubernetes were used to forecast COVID-19 infections for the UK's NHS, overcoming scaling challenges and creating a stable architecture for Bayesian modeling.

• YouTube
• 30 minutes
• Self-Paced
• Free Video

Python-centric AI Application Building in Minutes with Lepton and Ray

Explore efficient AI application development using Lepton's Python SDK and Ray for scalability. Learn to bridge research flexibility with infrastructure scalability in the era of AIGC and LLMs.

• YouTube
• 30 minutes
• Self-Paced
• Free Video

How Ray Empowered Ant Group to Deliver a Large-Scale Online Serverless Platform

Explore how Ray enabled Ant Group to build a massive serverless platform, overcoming challenges and optimizing resource utilization for increased efficiency and cost-effectiveness.

• YouTube
• 27 minutes
• Self-Paced
• Free Video

Supercharging Self-Driving Algorithm Development with Ray: Scaling Simulation Workloads and Democratizing Autotuning

Accelerate autonomous driving algorithm development using Ray for large-scale simulation, metrics evaluation, and autotuning. Learn to improve self-driving software without code changes.

• YouTube
• 33 minutes
• Self-Paced
• Free Video

KubeRay: A Ray Cluster Management Solution on Kubernetes

Explore KubeRay's integration with Kubernetes, its capabilities for Ray cluster management, and performance benchmarks for simplified deployment of Ray applications.

• YouTube
• 25 minutes
• Self-Paced
• Free Video

Last Mile Data Processing for ML Training Using Ray at Pinterest

Explore Pinterest's integration of Ray for streamlined ML innovations, scaling end-to-end ML lifecycle, and enhancing developer velocity. Learn about challenges, lessons, and business benefits.

• YouTube
• 28 minutes
• Self-Paced
• Free Video

A Ray-based ML Platform for Early Cancer Detection

Explore how Ray enhances scalability and efficiency in machine learning for early cancer detection, focusing on improved model development and evaluation for potential diagnostic applications.

• YouTube
• 34 minutes
• Self-Paced
• Free Video

Best Practices for Productionizing Distributed Training with Ray Train

Explore fault-tolerant distributed training techniques using Ray Train, covering experiment restoration, node failure recovery, cloud storage snapshots, and large model checkpointing for efficient AI workloads.

• YouTube
• 30 minutes
• Self-Paced
• Free Video

The Different Shades of Using KubeRay with Kubernetes

Explore KubeRay deployment on Kubernetes, from basic setup to advanced API usage. Learn to manage ray-clusters programmatically and scale on demand using the Python-Client KubeRay library.

• YouTube
• 33 minutes
• Self-Paced
• Free Video

Faster and Cheaper Offline Batch Inference with Ray

Optimize offline batch inference for ML applications using Ray Data. Learn to process terabytes efficiently, leverage modern models, and explore LLM use cases for faster, cheaper solutions.

• YouTube
• 28 minutes
• Self-Paced
• Free Video

Enabling End-to-End LLMOps on Michelangelo with Ray

Explore Uber's extension of Michelangelo for end-to-end LLMOps, leveraging Ray for scalable development with hundreds of A100 GPUs and integrating open-source techniques for efficient custom LLM creation.

• YouTube
• 30 minutes
• Self-Paced
• Free Video