Datasets:
Add paper link, GitHub repository, and improve metadata
#2
by
nielsr
HF Staff
- opened
README.md
CHANGED
|
@@ -1,8 +1,9 @@
|
|
| 1 |
---
|
| 2 |
-
task_categories:
|
| 3 |
-
- robotics
|
| 4 |
language:
|
| 5 |
- en
|
|
|
|
|
|
|
|
|
|
| 6 |
tags:
|
| 7 |
- RDT
|
| 8 |
- rdt
|
|
@@ -10,14 +11,18 @@ tags:
|
|
| 10 |
- manipulation
|
| 11 |
- bimanual
|
| 12 |
- ur5e
|
| 13 |
-
-
|
| 14 |
- vision-language-action
|
| 15 |
-
|
| 16 |
---
|
| 17 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
## Dataset Summary
|
| 19 |
|
| 20 |
-
This dataset provides
|
| 21 |
Each sample packs:
|
| 22 |
|
| 23 |
* a **binocular RGB image** (left + right wrist cameras concatenated horizontally)
|
|
@@ -91,7 +96,7 @@ shard-000002.tar
|
|
| 91 |
|
| 92 |
### 1) Official Guidelines to fine-tune RDT 2 series
|
| 93 |
|
| 94 |
-
Use the
|
| 95 |
|
| 96 |
### 2) Minimal Loading example
|
| 97 |
|
|
@@ -138,9 +143,9 @@ def get_train_dataset(shards_dir):
|
|
| 138 |
|
| 139 |
return dataset
|
| 140 |
|
| 141 |
-
with open(os.path.join("<Dataset
|
| 142 |
instructions = json.load(fp)
|
| 143 |
-
dataset = get_train_dataset(os.path.join("<Dataset
|
| 144 |
```
|
| 145 |
---
|
| 146 |
|
|
@@ -153,15 +158,12 @@ dataset = get_train_dataset(os.path.join("<Dataset Diretory>", "shards"))
|
|
| 153 |
|
| 154 |
## Citation
|
| 155 |
|
| 156 |
-
If you use this dataset, please cite the dataset and your project appropriately. For example:
|
| 157 |
-
|
| 158 |
```bibtex
|
| 159 |
-
@
|
| 160 |
-
|
| 161 |
-
|
| 162 |
-
|
| 163 |
-
|
| 164 |
-
year={2025}
|
| 165 |
}
|
| 166 |
```
|
| 167 |
|
|
@@ -169,17 +171,5 @@ If you use this dataset, please cite the dataset and your project appropriately.
|
|
| 169 |
|
| 170 |
## License
|
| 171 |
|
| 172 |
-
* **Dataset license:** Apache-2.0
|
| 173 |
-
* Ensure compliance when redistributing derived data or models.
|
| 174 |
-
|
| 175 |
-
---
|
| 176 |
-
|
| 177 |
-
## Maintainers & Contributions
|
| 178 |
-
|
| 179 |
-
We welcome fixes and improvements to the conversion scripts and docs (see https://github.com/thu-ml/RDT2/tree/main#troubleshooting).
|
| 180 |
-
Please open issues/PRs with:
|
| 181 |
-
|
| 182 |
-
* OS + Python versions
|
| 183 |
-
* Minimal repro code
|
| 184 |
-
* Error tracebacks
|
| 185 |
-
* Any other helpful context
|
|
|
|
| 1 |
---
|
|
|
|
|
|
|
| 2 |
language:
|
| 3 |
- en
|
| 4 |
+
license: apache-2.0
|
| 5 |
+
task_categories:
|
| 6 |
+
- robotics
|
| 7 |
tags:
|
| 8 |
- RDT
|
| 9 |
- rdt
|
|
|
|
| 11 |
- manipulation
|
| 12 |
- bimanual
|
| 13 |
- ur5e
|
| 14 |
+
- webdataset
|
| 15 |
- vision-language-action
|
| 16 |
+
arxiv: 2602.03310
|
| 17 |
---
|
| 18 |
|
| 19 |
+
# RDT2 Dataset
|
| 20 |
+
|
| 21 |
+
[Project page](https://rdt-robotics.github.io/rdt2/) | [Paper](https://huggingface.co/papers/2602.03310) | [GitHub](https://github.com/thu-ml/RDT2)
|
| 22 |
+
|
| 23 |
## Dataset Summary
|
| 24 |
|
| 25 |
+
This dataset provides shards in the **WebDataset** format for fine-tuning [RDT-2](https://rdt-robotics.github.io/rdt2/) or other policy models on **bimanual manipulation**.
|
| 26 |
Each sample packs:
|
| 27 |
|
| 28 |
* a **binocular RGB image** (left + right wrist cameras concatenated horizontally)
|
|
|
|
| 96 |
|
| 97 |
### 1) Official Guidelines to fine-tune RDT 2 series
|
| 98 |
|
| 99 |
+
Use the official [scripts](https://github.com/thu-ml/RDT2) and [guidelines](https://github.com/thu-ml/RDT2) provided in the GitHub repository.
|
| 100 |
|
| 101 |
### 2) Minimal Loading example
|
| 102 |
|
|
|
|
| 143 |
|
| 144 |
return dataset
|
| 145 |
|
| 146 |
+
with open(os.path.join("<Dataset Directory>", "instructions.json")) as fp:
|
| 147 |
instructions = json.load(fp)
|
| 148 |
+
dataset = get_train_dataset(os.path.join("<Dataset Directory>", "shards"))
|
| 149 |
```
|
| 150 |
---
|
| 151 |
|
|
|
|
| 158 |
|
| 159 |
## Citation
|
| 160 |
|
|
|
|
|
|
|
| 161 |
```bibtex
|
| 162 |
+
@article{rdt2,
|
| 163 |
+
title={RDT2: Exploring the Scaling Limit of UMI Data Towards Zero-Shot Cross-Embodiment Generalization},
|
| 164 |
+
author={RDT Team},
|
| 165 |
+
journal={arXiv preprint arXiv:2602.03310},
|
| 166 |
+
year={2026}
|
|
|
|
| 167 |
}
|
| 168 |
```
|
| 169 |
|
|
|
|
| 171 |
|
| 172 |
## License
|
| 173 |
|
| 174 |
+
* **Dataset license:** Apache-2.0.
|
| 175 |
+
* Ensure compliance when redistributing derived data or models.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|