Ray
Ray is an open-source distributed-computing framework for scaling Python, artificial-intelligence, and machine-learning applications across processes, machines, and clusters.
LLMs in Production
Ray is an open-source framework for executing distributed and parallel Python applications across clusters.
- How it is used
- Distributes concurrent machine-learning and high-performance workloads across multicore CPUs or cluster nodes.
- How deep it goes
- Defines tasks, actors, decorators, resources, and clusters, but stops short of detailed cluster operations.
- Examples
- Compares ten one-second function calls run sequentially with Ray remote calls and collected futures.
- What you can do after
- Build it from scratch by initializing Ray, adding @ray.remote, collecting results, and shutting the runtime down.
Transformers in Action
Ray is an open-source framework for scaling Python and machine-learning applications across distributed computing resources.
- How it is used
- Scales LLM training, data processing, and hyperparameter searches across distributed computing resources.
- How deep it goes
- Explains Ray Data, TorchTrainer, Tuner, scheduling, checkpoints, and model publication through executable workflow listings.
- Examples
- Uses SuperGLUE CommitmentBank classification, Ray-native dataset conversion, TorchTrainer scaling, ASHA search, checkpoint loading, and Hub publication.
- What you can do after
- Build it from scratch by converting datasets, distributing training, tuning configurations, selecting checkpoints, and loading trained models.
Related concepts
- TorchTrainerused by · TorchTrainer is a Ray training interface for distributing and managing PyTorch training workloads across workers and computational resources.
- Ray Tuneused by · Ray Tune is a Ray library for distributed hyperparameter tuning and experiment management.
- pandasalternative to · pandas is an open-source Python library that provides labeled data structures and tools for manipulating and analyzing tabular and other structured data.
- distributed systemuses · A distributed system is a computing system whose components run across multiple networked processes or machines and coordinate through communication.
- hyperparameter tuninguses · Hyperparameter tuning is the process of selecting or adjusting configurable model or procedure settings and evaluating them against a chosen objective.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.