Parameters object

As explained in Designing Deep Learning Systems: A guide for software engineers

A metadata entity that stores the input parameters and configuration values of a training run.

A parameters object stores the inputs that configure a training run, including the dataset ID, dataset version, and training hyperparameters. It belongs to a metadata schema centered on training runs, allowing configuration to be traced alongside produced metrics and models. section 8.3.2

Use the object to record:

• Dataset identity and version used by the run. section 8.3.2
• Training hyperparameters supplied as inputs. section 8.3.2
• A shared training_run_id that connects inputs with metrics, output models, and other run metadata. section 8.3.2

Figure 8.5 An entity relationship diagram of the data schema of the metadata storage

Inspect the entity relationship diagram to see how the parameters object connects to training runs and related metadata. section 8.3.2

How Designing Deep Learning Systems: A guide for software engineers covers it

How deep it goes

Defines stored fields and relationships but stops short of specifying serialization formats or database implementation.

Examples

Records dataset IDs, dataset versions, and training hyperparameters for a run.

What you can do after

Evaluate training-run configuration by retrieving its recorded input data alongside related outputs and metrics.

How it connects

Connects training runs with metrics, models, experiments, and the metadata store.

liveBookmanning.com

Read Designing Deep Learning Systems: A guide for software engineers on liveBook Get the book at manning.com

This explainer was generated by Manning's knowledge graph from the text of Designing Deep Learning Systems: A guide for software engineers. The section links point to the parts of the book it draws on.