Add paper, project, and code links
#1
by nielsr HF Staff - opened
README.md
CHANGED
|
@@ -1,9 +1,14 @@
|
|
| 1 |
---
|
| 2 |
-
|
|
|
|
| 3 |
library_name: pytorch
|
|
|
|
| 4 |
pipeline_tag: robotics
|
| 5 |
-
|
| 6 |
-
|
|
|
|
|
|
|
|
|
|
| 7 |
---
|
| 8 |
|
| 9 |
# Long-WAM YAM
|
|
@@ -12,6 +17,10 @@ A YAM bimanual robot pretraining model trained on **ABC 130K**, with a
|
|
| 12 |
LongLive 2.0 Robot-S video backbone. It predicts actions after imagining future
|
| 13 |
video latents.
|
| 14 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
## Download
|
| 16 |
|
| 17 |
```bash
|
|
@@ -45,4 +54,4 @@ and denormalize predicted actions; RGB inputs use `[-1, 1]`.
|
|
| 45 |
|
| 46 |
The matching runtime and base-model VAE/text assets are required separately.
|
| 47 |
This is a pretraining checkpoint: task-specific adaptation and controlled
|
| 48 |
-
closed-loop validation are required before real-robot use.
|
|
|
|
| 1 |
---
|
| 2 |
+
datasets:
|
| 3 |
+
- XDOF/ABC-130k
|
| 4 |
library_name: pytorch
|
| 5 |
+
license: other
|
| 6 |
pipeline_tag: robotics
|
| 7 |
+
tags:
|
| 8 |
+
- long-wam
|
| 9 |
+
- yam
|
| 10 |
+
- abc-130k
|
| 11 |
+
- robotics
|
| 12 |
---
|
| 13 |
|
| 14 |
# Long-WAM YAM
|
|
|
|
| 17 |
LongLive 2.0 Robot-S video backbone. It predicts actions after imagining future
|
| 18 |
video latents.
|
| 19 |
|
| 20 |
+
**Paper:** [Long-WAM: Scaling the Context of World-Action Models](https://huggingface.co/papers/2610.10528)
|
| 21 |
+
**Project page:** [https://nvlabs.github.io/LongLive/Long-WAM/](https://nvlabs.github.io/LongLive/Long-WAM/)
|
| 22 |
+
**Code:** [https://github.com/NVlabs/LongLive](https://github.com/NVlabs/LongLive)
|
| 23 |
+
|
| 24 |
## Download
|
| 25 |
|
| 26 |
```bash
|
|
|
|
| 54 |
|
| 55 |
The matching runtime and base-model VAE/text assets are required separately.
|
| 56 |
This is a pretraining checkpoint: task-specific adaptation and controlled
|
| 57 |
+
closed-loop validation are required before real-robot use.
|