Metadata for Microduck policy

**Model card**
- `library_name`: `onnx` → `microduck`
- `pipeline_tag`: `reinforcement-learning` → `robotics`
- add tag `microduck-slot:walk`, taken from `manifest.json`

README.md

@@ -4,7 +4,9 @@ tags:
 - robotics
 - reinforcement-learning
 - onnx
-library_name: onnx
+- microduck-slot:walk
+library_name: microduck
+pipeline_tag: robotics
 ---
 
 # hill-climb-1m5-reference