Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,94 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: cc-by-4.0
|
3 |
+
---
|
4 |
+
# PhysicalAI-Robotics-GR00T-X-Embodiment-Sim
|
5 |
+
|
6 |
+

|
7 |
+
|
8 |
+
Github Repo: [Isaac GR00T N1](https://github.com/NVIDIA/Isaac-GR00T)
|
9 |
+
|
10 |
+
We provide a set of datasets used for post-training of GR00T N1. Each dataset is a collection of trajectories from different robot embodiments and tasks.
|
11 |
+
|
12 |
+
### Cross-embodied bimanual manipulation: 9k trajectories
|
13 |
+
|
14 |
+
| Dataset Name | #trajectories |
|
15 |
+
| - | -|
|
16 |
+
| robocasa_bimanual_panda_inspire_hand.TwoArmBoxCleanupPadRes256Freq20_1000 | 1000 |
|
17 |
+
| robocasa_bimanual_panda_inspire_hand.TwoArmDrawerCleanupPadRes256Freq20_1000 | 1000 |
|
18 |
+
| robocasa_bimanual_panda_inspire_hand.TwoArmLiftTrayPadRes256Freq20_1000 | 1000 |
|
19 |
+
| robocasa_bimanual_panda_parallel_gripper.TwoArmThreePieceAssemblyPadRes256Freq20_1000 | 1000 |
|
20 |
+
| robocasa_bimanual_panda_parallel_gripper.TwoArmTransportPadRes256Freq20_1000 | 1000 |
|
21 |
+
| robocasa_gr1_arms_only_fourier_hands.TwoArmCanSortPadRes256Freq20_1000 | 1000 |
|
22 |
+
| robocasa_gr1_fixed_lower_body_fourier_hands.TwoArmCoffeePadRes256Freq20_1000 | 1000 |
|
23 |
+
| robocasa_gr1_fixed_lower_body_fourier_hands.TwoArmPouringPadRes256Freq20_1000 | 1000 |
|
24 |
+
|
25 |
+
|
26 |
+
### Humanoid robot tabletop manipulation: 40k trajectories
|
27 |
+
|
28 |
+
| Dataset Name | #trajectories |
|
29 |
+
| - | -|
|
30 |
+
| gr1_unified.CanToDrawer | 10000 |
|
31 |
+
| gr1_unified.PlacematToTieredShelf | 10000 |
|
32 |
+
| gr1_unified.CupToDrawer | 10000 |
|
33 |
+
| gr1_unified.PlateToBowl | 10000 |
|
34 |
+
| gr1_unified.CuttingBoardToBasket | 10000 |
|
35 |
+
| gr1_unified.PlateToCardboardBox | 10000 |
|
36 |
+
| gr1_unified.CuttingBoardToCardboardBox | 10000 |
|
37 |
+
| gr1_unified.PlateToPan | 10000 |
|
38 |
+
| gr1_unified.CuttingBoardToPan | 10000 |
|
39 |
+
| gr1_unified.PlateToPlate | 10000 |
|
40 |
+
| gr1_unified.CuttingBoardToPot | 10000 |
|
41 |
+
| gr1_unified.PotatoToMicrowave | 10000 |
|
42 |
+
| gr1_unified.CuttingBoardToTieredBasket | 10000 |
|
43 |
+
| gr1_unified.TrayToCardboardBox | 10000 |
|
44 |
+
| gr1_unified.PlaceBottleToCabinet | 10000 |
|
45 |
+
| gr1_unified.TrayToPlate | 10000 |
|
46 |
+
| gr1_unified.PlaceMilkToMicrowave | 10000 |
|
47 |
+
| gr1_unified.TrayToPot | 10000 |
|
48 |
+
| gr1_unified.PlacematToBasket | 10000 |
|
49 |
+
| gr1_unified.TrayToTieredBasket | 10000 |
|
50 |
+
| gr1_unified.PlacematToBowl | 10000 |
|
51 |
+
| gr1_unified.TrayToTieredShelf | 10000 |
|
52 |
+
| gr1_unified.PlacematToPlate | 10000 |
|
53 |
+
| gr1_unified.WineToCabinet | 10000 |
|
54 |
+
|
55 |
+
### Robot Arm Kitchen Manipulation: 24K trajectories
|
56 |
+
| Dataset Name | #trajectories |
|
57 |
+
| - | -|
|
58 |
+
| robocasa_panda_omron.CloseDoubleDoor256_1000 | 1000 |
|
59 |
+
| robocasa_panda_omron.CloseDrawer256_1000 | 1000 |
|
60 |
+
| robocasa_panda_omron.CloseSingleDoor256_1000 | 1000 |
|
61 |
+
| robocasa_panda_omron.CoffeePressButton256_1000 | 1000 |
|
62 |
+
| robocasa_panda_omron.CoffeeServeMug256_1000 | 1000 |
|
63 |
+
| robocasa_panda_omron.CoffeeSetupMug256_1000 | 1000 |
|
64 |
+
| robocasa_panda_omron.OpenDoubleDoor256_1000 | 1000 |
|
65 |
+
| robocasa_panda_omron.OpenDrawer256_1000 | 1000 |
|
66 |
+
| robocasa_panda_omron.OpenSingleDoor256_1000 | 1000 |
|
67 |
+
| robocasa_panda_omron.PnPCabToCounter256_1000 | 1000 |
|
68 |
+
| robocasa_panda_omron.PnPCounterToCab256_1000 | 1000 |
|
69 |
+
| robocasa_panda_omron.PnPCounterToMicrowave256_1000 | 1000 |
|
70 |
+
| robocasa_panda_omron.PnPCounterToSink256_1000 | 1000 |
|
71 |
+
| robocasa_panda_omron.PnPCounterToStove256_1000 | 1000 |
|
72 |
+
| robocasa_panda_omron.PnPMicrowaveToCounter256_1000 | 1000 |
|
73 |
+
| robocasa_panda_omron.PnPSinkToCounter256_1000 | 1000 |
|
74 |
+
| robocasa_panda_omron.PnPStoveToCounter256_1000 | 1000 |
|
75 |
+
| robocasa_panda_omron.TurnOffMicrowave256_1000 | 1000 |
|
76 |
+
| robocasa_panda_omron.TurnOffSinkFaucet256_1000 | 1000 |
|
77 |
+
| robocasa_panda_omron.TurnOffStove256_1000 | 1000 |
|
78 |
+
| robocasa_panda_omron.TurnOnMicrowave256_1000 | 1000 |
|
79 |
+
| robocasa_panda_omron.TurnOnSinkFaucet256_1000 | 1000 |
|
80 |
+
| robocasa_panda_omron.TurnOnStove256_1000 | 1000 |
|
81 |
+
| robocasa_panda_omron.TurnSinkSpout256_1000 | 1000 |
|
82 |
+
|
83 |
+
## Download the dataset from Huggingface
|
84 |
+
|
85 |
+
Users can download a specific subset of data by specifying the dataset name.
|
86 |
+
|
87 |
+
```bash
|
88 |
+
huggingface-cli download nvidia/GR00T-Sim-Post-Training-Data\
|
89 |
+
--repo-type dataset \
|
90 |
+
--include "robocasa_gr1_arms_only_fourier_hands.TwoArmCanSortPadRes256Freq20_30/**" \
|
91 |
+
--local-dir $HOME/gr00t_dataset
|
92 |
+
```
|
93 |
+
|
94 |
+
**Replace `robocasa_gr1_arms_only_fourier_hands.TwoArmCanSortPadRes256Freq20_30/**` with the dataset name you want to download.**
|