Add pipeline tag and links to paper, project page, and code

#1
by nielsr HF Staff - opened
Files changed (1) hide show
  1. README.md +15 -10
README.md CHANGED
@@ -1,19 +1,24 @@
1
  ---
2
- license: bsd-3-clause
3
  library_name: rl-games
 
 
4
  tags:
5
- - isaac-lab
6
- - reinforcement-learning
7
- - robotics
8
- - peg-in-hole
9
- - contact-rich-manipulation
10
- - tactile-sensing
11
- - sim-to-real
12
- - insertanything
13
  ---
14
 
15
  # InsertAnything Checkpoints
16
 
 
 
 
 
17
  Pretrained RL-Games policies for the single-hole precision-insertion environments released with **InsertAnything: Generalizable Contact-Rich Precision Insertion from Simulation to Reality**.
18
 
19
  The policies were trained entirely in Isaac Lab. They use deployable robot observations together with a compact three-axis fingertip force signal for contact-aware motion correction. The files are intended for evaluation, visualization, and reproducibility checks with the accompanying InsertAnything codebase.
@@ -96,4 +101,4 @@ Compare the returned hashes with the table above before evaluation.
96
 
97
  ## License
98
 
99
- The checkpoints are released under the BSD 3-Clause license.
 
1
  ---
 
2
  library_name: rl-games
3
+ license: bsd-3-clause
4
+ pipeline_tag: robotics
5
  tags:
6
+ - isaac-lab
7
+ - reinforcement-learning
8
+ - robotics
9
+ - peg-in-hole
10
+ - contact-rich-manipulation
11
+ - tactile-sensing
12
+ - sim-to-real
13
+ - insertanything
14
  ---
15
 
16
  # InsertAnything Checkpoints
17
 
18
+ **Paper:** [InsertAnything: Generalizable Contact-Rich Precision Insertion from Simulation to Reality](https://huggingface.co/papers/2609.24511)
19
+ **Project page:** [https://mzhsoul.github.io/InsertAnything](https://mzhsoul.github.io/InsertAnything)
20
+ **Code:** [https://github.com/mzhsoul/InsertAnything](https://github.com/mzhsoul/InsertAnything)
21
+
22
  Pretrained RL-Games policies for the single-hole precision-insertion environments released with **InsertAnything: Generalizable Contact-Rich Precision Insertion from Simulation to Reality**.
23
 
24
  The policies were trained entirely in Isaac Lab. They use deployable robot observations together with a compact three-axis fingertip force signal for contact-aware motion correction. The files are intended for evaluation, visualization, and reproducibility checks with the accompanying InsertAnything codebase.
 
101
 
102
  ## License
103
 
104
+ The checkpoints are released under the BSD 3-Clause license.