Enhance dataset card: add sample usage, update project page label and tags
Browse filesThis PR improves the dataset card for the BEHAVIOR Robot Suite (BRS) Data by:
* Adding a "Sample Usage" section with practical code snippets for training a WB-VIMA policy and deploying it on a real robot, extracted directly from the project's GitHub README.
* Updating the "Demo" label under "Dataset Sources" to "Project page" for improved clarity and consistency with standard conventions.
* Adding relevant `tags` to the metadata (`robot-manipulation`, `imitation-learning`, `real-world-data`) to enhance the dataset's discoverability on the Hugging Face Hub, reflecting its core aspects as described in the paper and repository.
README.md
CHANGED
|
@@ -5,6 +5,10 @@ license: mit
|
|
| 5 |
task_categories:
|
| 6 |
- robotics
|
| 7 |
pretty_name: BRS Data
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
---
|
| 9 |
|
| 10 |
# Dataset Card for BEHAVIOR Robot Suite (BRS) Data
|
|
@@ -31,12 +35,33 @@ These data are first collected and used in the paper [BEHAVIOR Robot Suite: Stre
|
|
| 31 |
|
| 32 |
- **Repository:** https://github.com/behavior-robot-suite/brs-algo
|
| 33 |
- **Paper:** https://arxiv.org/abs/2503.05652
|
| 34 |
-
- **
|
| 35 |
|
| 36 |
## Uses
|
| 37 |
|
| 38 |
For usage instructions, see our doc [here](https://behavior-robot-suite.github.io/docs/sections/wbvima/overview.html).
|
| 39 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
## Citation
|
| 41 |
|
| 42 |
**BibTeX:**
|
|
|
|
| 5 |
task_categories:
|
| 6 |
- robotics
|
| 7 |
pretty_name: BRS Data
|
| 8 |
+
tags:
|
| 9 |
+
- robot-manipulation
|
| 10 |
+
- imitation-learning
|
| 11 |
+
- real-world-data
|
| 12 |
---
|
| 13 |
|
| 14 |
# Dataset Card for BEHAVIOR Robot Suite (BRS) Data
|
|
|
|
| 35 |
|
| 36 |
- **Repository:** https://github.com/behavior-robot-suite/brs-algo
|
| 37 |
- **Paper:** https://arxiv.org/abs/2503.05652
|
| 38 |
+
- **Project page:** https://behavior-robot-suite.github.io/
|
| 39 |
|
| 40 |
## Uses
|
| 41 |
|
| 42 |
For usage instructions, see our doc [here](https://behavior-robot-suite.github.io/docs/sections/wbvima/overview.html).
|
| 43 |
|
| 44 |
+
## Sample Usage
|
| 45 |
+
|
| 46 |
+
To train a WB-VIMA policy, simply run the following command as described in the [official documentation](https://behavior-robot-suite.github.io/docs/sections/wbvima/overview.html):
|
| 47 |
+
|
| 48 |
+
```bash
|
| 49 |
+
python3 main/train/train.py data_dir=<HDF5_PATH> \
|
| 50 |
+
bs=<BS> \
|
| 51 |
+
arch=wbvima \
|
| 52 |
+
task=<TASK_NAME> \
|
| 53 |
+
exp_root_dir=<EXP_ROOT_DIR> \
|
| 54 |
+
gpus=<NUM_GPUS> \
|
| 55 |
+
use_wandb=<USE_WANDB> \
|
| 56 |
+
wandb_project=<WANDB_PROJECT>
|
| 57 |
+
```
|
| 58 |
+
|
| 59 |
+
To deploy a WB-VIMA policy on the real robot, simply run the following command:
|
| 60 |
+
|
| 61 |
+
```bash
|
| 62 |
+
python3 main/rollout/<TASK_NAME>/rollout_async.py --ckpt_path <CKPT_PATH> --action_execute_start_idx <IDX>
|
| 63 |
+
```
|
| 64 |
+
|
| 65 |
## Citation
|
| 66 |
|
| 67 |
**BibTeX:**
|