Kubeflow offers ML model development environments optimized for cloud, enabling Kubernetes-based linking with various open source software.
The standardized environments support a range of machine learning frameworks from Tensorflow and Pytorch to Scikit-learn and Keras. The pipeline for the entire development, learning and deployment processes of machine learning models are automated to ensure simple configuration/creation as well as reuse of the models.
Job schedulers (FIFO, Bin-packing, and Gang-based), GPU resource monitoring, Kubeflow engine logging and more add-on features are available for efficient usage of GPU resources.
- Request : Automatic deployment and service configuration for the requested Kubernetes clusters
- View : Offering list, Kubeflow version/resource status, and running/stop status information
- Delete : Delete created Kubeflow modules
- Basic features
. Jupyter Notebook (model development, learning, and inference)
. Workflow automation (based on machine learning pipelines)
- Additional features of Samsung Cloud Platform
. GPU Job Scheduling
. ML framework images (Tensorflow, PyTorch, Mxnet, etc.)
. Kubeflow engine monitoring/logging and authentication
Whether you’re looking for a specific business solution or just need some questions answered, we’re here to help