News

User Story

Engineering

Culture

Eric Liang, Jules S. Damji, Zhe Zhang

We cover 3 methods of batch inference in Ray - from low-level using Ray Actors, to high-level using Ray Data streaming.

Figure 1. Three ways of scaling model batch inference in Ray.


batch_inference_fig_1

Scaling Model Batch Inference in Ray | Actors, ActorPool & Ray Data

Scaling Model Batch Inference in Ray: Using Actors, ActorPool, and Ray Data

This blog covers four methods of [batch inference](https://docs.ray.io/en/latest/ray-overview/use-cases.html#batch-inference) (distributed model evaluation) in Ray: from low-level using Ray Actors, to high-level using Ray AIR libraries. 

We'll see how low-level Ray APIs allow you to control exactly "How" Ray is executing computations, while Ray's libraries enable you to specify just the "What" --- for a more out-of-the-box scalable experience.

Four ways of model batch inference in Ray.

Model Batch Inference in Ray: Actors, ActorPool & Datasets

Model Batch Inference in Ray: Actors, ActorPool, and Datasets

Intel and Anyscale have integrated Ray with Gaudi3. Gaudi3 is AI-optimized for efficient training and other machine learning workloads.

Anyscale - Intel

Accelerating AI: Harnessing Intel(R) Gaudi(R) 3 with Ray 2.10

Anyscale is teaming with NVIDIA to combine the developer productivity of Ray Serve and RayLLM with the cutting-edge optimizations from NVIDIA Triton Inference Server software and the NVIDIA TensorRT-LLM library.

NVIDIA-Anyscale Logo 1280x680.jpg

Low-latency Generative AI Model Serving with Ray, NVIDIA Triton Inference Server, and NVIDIA TensorRT-LLM

Based on the popular “Needle In a Haystack” benchmark and RAG, we share our process of creating a problem-specific fine-tuning dataset to extend the context of models to build better RAG systems.

fine tuning llms query featured

Fine-tuning LLMs for longer context and better RAG systems

Generating embeddings is a critical task for developing successful RAG applications, and Anyscalew with Pinecone is the most cost-efficient solution for this workflow.

blog-anyscale-pincone

RAG at Scale: 10x Cheaper Embedding Computations with Anyscale and Pinecone

This section is used to order the "Types" and "Tags" that show up for filters on the Blog Index

Types

Products / Libraries

Blog

Hosted Anyscale is in Private Preview. Click here to request access!

Anyscale is the leading AI application platform. With Anyscale, developers can build, run and scale AI applications instantly.

Posts by Eric Liang, Jules S. Damji, Zhe Zhang

No posts found.