Model Library
The Model Library is where machine learning models are stored within a project.
Any model that is used inside a workflow must first exist in the Model Library. This includes both built-in models provided by the platform and custom models added by the user.
The library acts as a centralized location for managing all models associated with a project.
Built-In Models
SolutionEngine includes a set of preconfigured models that are immediately available for use.
These models can be selected directly inside workflow nodes without requiring manual upload or configuration at the file level.
The available built-in models may change over time as the platform evolves.
Custom Models
In addition to built-in models, users can import their own models into the library.
Currently, custom models can be imported from external sources such as Kaggle.
Once imported, custom models appear in the Model Library alongside built-in models and can be selected in the same way within workflows.
Detailed instructions for importing models are provided in the Model Importing section of the documentation.
Model Configuration
Each model has its own configuration options.
Configuration parameters depend on the specific model and may include settings related to input handling, inference behavior, or output formatting.
When a model is used inside a workflow node, its configuration options become available and can be adjusted as part of the workflow definition.
Using Models in Workflows
Models are used inside workflows through dedicated model execution nodes.
When a workflow runs, the selected model processes incoming data according to its configuration and produces output that can be used by subsequent nodes.
The Model Library therefore serves as the foundation for integrating AI capabilities into workflow logic.