project_adaptation

(TBD) Project adaptation

This pipeline is the example of domain adaptation with auto labeling. AWML do not manage the configuration for each experiment, so if you want to manage for each experiment, then you make new repository to manage it. In this example, we use CenterPoint.

• Support priority: Tier C

Get started

1. Prepare non-annotated T4dataset

• Prepare non-annotated T4dataset

2. Make pseudo T4dataset

We use Pseudo label for T4dataset to make pseudo T4dataset

# command

3. Fine tuning

• Fine tuning by pseudo label

# command