Determined AI

Determined AI is an open-source deep learning platform for distributed training, hyperparameter tuning, experiment tracking, and GPU resource management, now maintained by HPE and underpinning HPE MLDE.

Determined AI is an open-source deep learning training platform designed to remove the infrastructure complexity from model development. It provides distributed training out of the box, automated hyperparameter search, built-in experiment tracking, and cluster-level GPU resource management, so data science teams can train large models and share expensive GPU resources without building their own orchestration. The platform works with popular frameworks including PyTorch and TensorFlow and handles fault-tolerant, multi-GPU and multi-node training, letting researchers focus on models rather than plumbing. Its scheduler and resource management make it particularly useful for teams that need to fairly allocate limited GPU capacity across many concurrent training jobs. Determined AI was acquired by Hewlett Packard Enterprise in 2021, and HPE committed to continuing it as an open-source project. It now serves as the foundation for the commercial HPE Machine Learning Development Environment (MLDE), which packages the open-source platform with enterprise support and integration into HPE's compute offerings. Teams can adopt the free, self-hosted open-source Determined for training and experimentation, or move to HPE MLDE for enterprise features and support.