Add kinematic planner checkpoint info and download instructions

README.md

@@ -137,9 +137,30 @@ SONIC includes a kinematic planner for real-time locomotion generation — choos
 
 ## Model Checkpoints
 
-Pre-trained SONIC policy checkpoints (ONNX format) are available in the `policy/release/` directory of this repository. The models are optimized for real-time inference on NVIDIA Jetson platforms with low-latency control and robust sim-to-real transfer.
+All checkpoints (ONNX format) are available directly in this repository and optimized for real-time inference on NVIDIA Jetson platforms.
 
-See the [Installation Guide](https://nvlabs.github.io/GR00T-WholeBodyControl/getting_started/installation_deploy.html) for instructions on downloading and using the model checkpoints.
+| Checkpoint | Path in repo | Description |
+|---|---|---|
+| Policy encoder | `policy/release/model_releasev1-1_encoder.onnx` | Encodes motion reference into latent |
+| Policy decoder | `policy/release/model_releasev1-1_decoder.onnx` | Decodes latent into joint actions |
+| Observation config | `policy/release/observation_config.yaml` | Sensor / obs configuration |
+| Kinematic planner | `planner/target_vel/V2/planner_sonic.onnx` | Real-time locomotion style planner |
+
+**Quick download** (requires `pip install huggingface_hub`):
+
+```python
+from huggingface_hub import snapshot_download
+snapshot_download(repo_id="nvidia/GEAR-SONIC", local_dir="gear_sonic_deploy")
+```
+
+Or use the download script from the GitHub repo:
+
+```bash
+python download_from_hf.py           # policy only
+python download_from_hf.py --planner # policy + planner
+```
+
+See the [Download Models guide](https://nvlabs.github.io/GR00T-WholeBodyControl/getting_started/download_models.html) for full instructions.
 
 ## Documentation