polyaxon
AI Infra / AI Orchestration / AI Control Plane
What is polyaxon?
AI Infra / AI Orchestration / AI Control Plane
Overview of polyaxon
What it is
Polyaxon is a platform for building, training, and monitoring large scale deep learning applications. It aims to solve reproducibility, automation, and scalability for machine learning applications. Polyaxon deploys into any data center, cloud provider, or can be hosted and managed by Polyaxon. It supports major deep learning frameworks such as Tensorflow, MXNet, Caffe, Torch, and others. The platform turns GPU servers into shared, self-service resources for a team or organization.
Key features
- Supports distributed training for frameworks like Tensorflow, Pytorch, MPI, Horovod, Spark, and Dask
- Hyperparameter tuning with search algorithms: Grid search, Random search, Hyperband, Bayesian Optimization, Hyperopt, and Custom Iterative Optimization
- Parallel execution of processing or model training jobs via a mapping abstraction
- DAGs and workflows for container-native machine learning pipelines, supporting jobs, services, distributed jobs, parallel executions, and nested DAGs
- Dashboard for viewing projects and experiments created by team members
- CLI for project creation, experiment running, and log/resource tracking
Use cases
Polyaxon is used for building, training, and monitoring large scale deep learning applications. It is suitable for teams or organizations that need to manage GPU resources as shared infrastructure, automate machine learning workflows, and ensure reproducibility of experiments. Use cases include hyperparameter tuning, distributed training, parallel job execution, and orchestrating complex pipelines with dependencies.
Getting started
bashInstall Polyaxon CLI$ pip install -U polyaxon# Create a namespace$ kubectl create namespace polyaxon# Add Polyaxon charts repo$ helm repo add polyaxon https://charts.polyaxon.com# Deploy Polyaxon$ polyaxon admin deploy -f config.yaml# Access API$ polyaxon port-forward
License and pricing
License: Apache-2.0. Pricing type: open source.
Based on the project's public README and repository metadata.