---
library_name: lerobot
tags:
- robotics
- imitation-learning
---

# Pi-0.5 (xarm wire rope)

LeRobot policy checkpoints uploaded by `goal_gen/upload_hf_checkpoints.sh`.
Each subfolder contains the deployment-ready `pretrained_model/` payload
(model.safetensors + config.json + pre/postprocessor + train_config.json).

| Subfolder | Train step | Final train loss |
|---|---|---|
| `checkpoint-020000` | 20,000 | — |
| `checkpoint-025000` | 25,000 | — |
| `checkpoint-030000` | 30,000 | — |

## Usage

```python
from huggingface_hub import snapshot_download
ckpt_dir = snapshot_download(
    "JayCao99/pi05-xarm-wire-rope-v2.0",
    allow_patterns="checkpoint-030000/*",
)
from lerobot.policies.pi05.modeling_pi05 import PI05Policy
policy = PI05Policy.from_pretrained(f"{ckpt_dir}/checkpoint-030000")
```