3.6.4. Adaptive Workflows
The Model class inherits from the bf.mt.Task class, among others. In combination with another class AWorkflow, all possibilities of MLPro’s multitasking abilities are unlocked for machine learning:
1. Models can execute internal methods asynchronously
2. Models can run as separate threads or processes
3. Models can be grouped into macro models in workflows
4. Models can share/exchange data using a shared object
This enables the efficient execution and adaptation of models and model groups using all available runtime resources.
Cross Reference