Model Definitions

The model definition is the interface between Determined and the user’s application framework (e.g., Keras, TensorFlow), in terms of loading training data, describing a model architecture, and specifying the underlying iterative optimization training algorithms.

Model definitions are written using the Trial API, which is an interface provided by Determined. With the Trial API, developers expose the components of a model definition to Determined’s training loop implementation.

To read more about the Trial API, see:

For examples of how to adapt existing model code to use the Trial APIs, see the tf.keras tutorial and the PyTorch tutorial, respectively.

Best Practices

To learn more about some best practices when writing Trial API model definitions, see Best Practices for Model Definitions.