Any Type of Data. Any Accelerator. Any Use Case.
Support for End-to-End Text & LLM Use Cases
Built for LLM online inference, batch inference, embedding generation, and synthetic data generation.
Reduce Costs When Processing Videos
Maximize compute utilization and leverage GPUs and CPUs to process videos of any size.
Image Processing at Scale
Scale image processing workloads by independently scaling CPU and GPU resources, delivering high throughput, lower costs, and improved utilization.
Enhanced Audio Processing
Process audio data without breaking the bank. Anyscale makes it easy to run a variety of use cases, including speech to text.
Best-In-Class Data Processing for ML/AI
Text Support
Image Support
Audio Support
Video Support
3D Data Support
Task-Specific CPU & GPU Allocation
Stateful Tasks
Native NumPy Support
Native Pandas Support
Model Parallelism Support
Nested Task Parallelism
Fast Node Launching and Autoscaling
Fractional GPU Support
Load Datasets Larger Than Cluster Memory
Improved Observability
Autoscale Workers to Zero
Job Queues
Priority Scheduling
Accelerated Execution
Data Loading / Data Ingest / Last Mile Preprocessing
| | | | |
|---|---|---|---|---|
Text Support | | | | |
Image Support | | - | | |
Audio Support | Manual | - | Manual | |
Video Support | Manual | - | Manual | |
3D Data Support | - | - | Binary | Binary |
Task-Specific CPU & GPU Allocation | - | - | | |
Stateful Tasks | - | - | | |
Native NumPy Support | - | - | | |
Native Pandas Support | | - | | |
Model Parallelism Support | - | - | | |
Nested Task Parallelism | - | - | | |
Fast Node Launching and Autoscaling | - | - | - | 60 sec |
Fractional GPU Support | - | Limited | | |
Load Datasets Larger Than Cluster Memory | - | - | | |
Improved Observability | - | - | - | |
Autoscale Workers to Zero | - | Limited | | |
Job Queues | - | - | - | |
Priority Scheduling | - | - | - | |
Accelerated Execution | - | - | - | |
Data Loading / Data Ingest / Last Mile Preprocessing | | | | |
Best-Price Performance
Anyscale’s Ray Data consistently outperforms competitors:
- 17x faster compared to AWS SageMaker
- 2x faster than Apache Spark
- 90% cost savings on select instances with spot instances
Out-of-the-Box Templates & App Accelerators
Jumpstart your development process with custom-made templates, only available on Anyscale.
Batch Inference with LLMs
Run LLM offline inference on large scale input data with Ray Data
Computing Text Embeddings
Compute text embeddings with Ray Data and HuggingFace models.
Pre-Train Stable Diffusion
Pre-train a Stable Diffusion V2 model with Ray Train and Ray Data
FAQs
Ray Data is an open source machine learning library built on top of Ray, a best-in-class Pythonic distributed computing platform. Anyscale was founded by the creators of Ray to continue optimizing proprietary technology—built on top of Ray open source—to meet the challenges of the fast-paced AI world. With Anyscale’s proprietary Ray Data, you get access to additional and advanced capabilities including:
- Faster job startup through incremental metadata fetching
- Faster autoscaling
- Improved observability and checkpointing
- Fault tolerance support
- Head node recovery
- Spot instance support
- Incremental metadata fetching
- Out-of-the-box data connectors with Snowflake and Databricks
- Resumable jobs
- And much more