---
datasets: airshop/Lego_task_complex
library_name: lerobot
license: apache-2.0
model_name: ACT
pipeline_tag: robotics
tags:
- ACT
- lerobot
- robotics
---

# Complex Lego Task

## Description
Pick up small lego block (pink) using left arm and place onto 2 x 1 (white), then right arm picks up combined block, places on 2 x 2 block (blue). In this setup, blue block is closer to right robot arm and pink block is closer to left robot arm, white block in between both.

## Dataset
- Repo: airshop/Lego_task_complex
- Image augmentations: Disabled.

## Training
- Model / Policy Type: ACT
- Steps: 50000
- Batch size: 8
- Vision backbone: resnet18
- VAE enabled: True (latent dim: 32)
- Training time: 18 hours
- Machine: Spark (NVIDIA GB10)

## Inputs
4 visual streams, 1 state inputs

## Results
[Insert results here]

---
## How to Get Started with the Model

For a complete walkthrough, see the [training guide](https://huggingface.co/docs/lerobot/il_robots#train-a-policy).
Below is the short version on how to train and run inference/eval:

### Train from scratch

```bash
lerobot-train \
  --dataset.repo_id=${HF_USER}/<dataset> \
  --policy.type=ACT \
  --output_dir=outputs/train/<desired_policy_repo_id> \
  --job_name=lerobot_training \
  --policy.device=cuda \
  --policy.repo_id=${HF_USER}/<desired_policy_repo_id>
  --wandb.enable=true
```

_Writes checkpoints to `outputs/train/<desired_policy_repo_id>/checkpoints/`._

### Evaluate the policy/run inference

```bash
lerobot-record \
  --robot.type=so100_follower \
  --dataset.repo_id=<hf_user>/eval_<dataset> \
  --policy.path=<hf_user>/<desired_policy_repo_id> \
  --episodes=10
```

Prefix the dataset repo with **eval\_** and supply `--policy.path` pointing to a local or hub checkpoint.
---

## Model Details

This policy has been trained and pushed to the Hub using [LeRobot](https://github.com/huggingface/lerobot).
See the full documentation at [LeRobot Docs](https://huggingface.co/docs/lerobot/index).

- **License:** apache-2.0