Skip to content

PPOExportModel ​

python
PPOExportModel(model: PPOModel, export_format: Literal['logits', 'action'], action_masker: ActionMasker | None = None) -> None

Wrapper module for exporting a PPOModel to ONNX.

Strips the value head and outputs either action logits or the greedy action index.

Initialize the PPOExportModel.

Parameters:

  • model (PPOModel) – The PPOModel to wrap for export.
  • export_format (Literal['logits', 'action']) – The output format - "logits" for policy logits, or "action" for the greedy action index.
  • action_masker (ActionMasker | None) – Optional ActionMasker to apply before action selection.

Source