Source: ATYUN AI platform
Machine learning is deployed with many challenges, but the new Seldon Core aims to help its new open-source platform for deployment on Kubernetes machine learning models.
Kubernetes (commonly referred K8S) for automatic deployment, management and extension of the container (Containerized) open source application. It is designed to provide "cross-host cluster to automatically deploy, scale and platform to run application container." It supports a range of container tools, including Docker and so on.
Seldon.io announced a new open source platform -Seldon Core, the platform allows data science team management and operational models on the scale. Seldon Core focused on the last step to resolve any machine learning programs, will help the company put into production models to solve practical problems, and maximize return on investment.
Traditional infrastructure stack (stack) and devops process can not translate well into machine learning, and there is limited open source innovation in this field, which forces companies to build their own or use a proprietary service at great cost. Have the necessary skills, multidisciplinary data engineer is very rare. Inefficiency of data scientists were put into service quality and performance-related challenges, these challenges will divert their attention from where they can add value to build a better model.
Data scientists focus on creating better models and devops teams can more effectively use the tools they understand to manage the deployment.
The platform features include:
- Scientists can deploy the data model using any machine learning toolkit or programming language built. Seldon Core plan initially supported python-based tool / language, including Tensorflow, scikitlearn, Spark and H20.
- At deployment time, via REST and gRPC will automatically integrate machine learning models to predict the needs of business applications and services.
- Processing model deployment of full life-cycle management, no downtime, including run-time map updates, scaling, monitoring and security.
installation
The official version: https://storage.googleapis.com/seldon-charts
Installation seldon-core:
helm install seldon-core --name seldon-core --repo https://storage.googleapis.com/seldon-charts
To install the optional components include analysis of Prometheus and Grafana including using a built-in dashboard to monitor the operation of machine learning deployment:
helm install seldon-core --name seldon-core \
--set grafana_prom_admin_password=password \
--set persistence.enabled=false \
--repo https://storage.googleapis.com/seldon-charts
Deployment Guide
For more details see: https://github.com/SeldonIO/seldon-core
This switched ATYUN artificial intelligence media platforms, the original link: Seldon.io released a new open-source platform for machine learning on Kubernetes
more recommendations
Ten Python image processing tools
BloomReach: How do customers expect from CMS to promote the development of DXP
AI generate real-time facial animation based on an audio voice
Cisco: Smart City technology can solve the face of the challenges of urban mobility