Train a Model

Training Workflow

workflow

Training Run

Move from a validated dataset to a runner-backed training job, metrics, and exportable artifact.

1. Dataset Use a labeled dataset with consistent sample shape and metadata.
2. Model template Select a template such as WavesFM or another configured training path.
3. Runner Choose a runner that can execute the generated workflow.
4. Training run Monitor job steps, duration, status, and logs.
5. Metrics Compare loss, accuracy, model candidates, and run history.
6. Artifact export Export an ONNX or compatible model artifact for downstream application packaging.

train

Training run

Monitor job steps, duration, status, and logs.

Where this appears in RIA Hub

Training Dashboard.

What you need next

Successful metrics and generated artifacts.

Open guide

Model Builder selects repository, runner, dataset, model template, and mode, then triggers a workflow run. The Training Dashboard reports run status, job steps, duration, artifacts, run history, and model comparison data.

Model Builder choices

template Model template Choose the configured training path for the dataset.compute Runner Pick the execution target that can run the workflow.run Mode Select training, export, HPO, or another supported mode.

template

Model template

Template selection controls generated code, adapters, transforms, and artifact expectations.

Queued Queued Running Running Succeeded Ready Failed Error

Reading a Run

Training curve placeholder

Example loss and accuracy trend for a short model run.

• Loss
• Accuracy

Data table

Series	0	1	2	3	4	5
Loss	1.2	0.94	0.72	0.58	0.47	0.41
Accuracy	0.42	0.55	0.68	0.74	0.8	0.84

Model comparison placeholder

Compare candidate metrics before choosing an export artifact.

• Validation score

Data table

Series	Tiny	Base	Tuned
Validation score	0.71 score	0.79 score	0.86 score

WavesFM path

Form input

The Model Builder form captures dataset, template, runner, and mode choices.

Handler

The backend handler generates the training workflow and connects it to repository state.

Python template

The generated Python path adapts datasets, applies transforms, and trains the model.

Actions runner

The runner executes steps and produces metrics plus artifacts for export.

Next steps

• Package the model — Take the exported artifact to Package an Application to connect it into a deployable pipeline.
• Run live inference — If working with the Conductor fingerprinting workflow, load the ONNX model into the Zone Fingerprinting Demo.