DeepMind Control Suite
Paper • 1801.00690 • Published
Datasets:
__key__ stringlengths 13 33 | __url__ stringclasses 7
values | hdf5 unknown |
|---|---|---|
ball_in_cup_catch/2621 | hf://datasets/ml-jku/dm_control_10M@30bb74e5535f6558e0e8e99c8afe5493e9ce4450/ball_in_cup_catch.tar.gz | [
137,
72,
68,
70,
13,
10,
26,
10,
0,
0,
0,
0,
0,
8,
8,
0,
4,
0,
16,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
255,
255,
255,
255,
255,
255,
255,
255,
198,
239,
0,
0,
0,
0,
0,
0,
255,
255,
255,
255,
255,
255,
255,... |
ball_in_cup_catch/6709 | hf://datasets/ml-jku/dm_control_10M@30bb74e5535f6558e0e8e99c8afe5493e9ce4450/ball_in_cup_catch.tar.gz | [
137,
72,
68,
70,
13,
10,
26,
10,
0,
0,
0,
0,
0,
8,
8,
0,
4,
0,
16,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
0,
255,
255,
255,
255,
255,
255,
255,
255,
184,
239,
0,
0,
0,
0,
0,
0,
255,
255,
255,
255,
255,
255,
255,... |
ball_in_cup_catch/9060 | "hf://datasets/ml-jku/dm_control_10M@30bb74e5535f6558e0e8e99c8afe5493e9ce4450/ball_in_cup_catch.tar.(...TRUNCATED) | "iUhERg0KGgoAAAAAAAgIAAQAEAAAAAAAAAAAAAAAAAD//////////y7vAAAAAAAA//////////8AAAAAAAAAAGAAAAAAAAAAAQA(...TRUNCATED) |
ball_in_cup_catch/964 | "hf://datasets/ml-jku/dm_control_10M@30bb74e5535f6558e0e8e99c8afe5493e9ce4450/ball_in_cup_catch.tar.(...TRUNCATED) | "iUhERg0KGgoAAAAAAAgIAAQAEAAAAAAAAAAAAAAAAAD///////////HvAAAAAAAA//////////8AAAAAAAAAAGAAAAAAAAAAAQA(...TRUNCATED) |
ball_in_cup_catch/2383 | "hf://datasets/ml-jku/dm_control_10M@30bb74e5535f6558e0e8e99c8afe5493e9ce4450/ball_in_cup_catch.tar.(...TRUNCATED) | "iUhERg0KGgoAAAAAAAgIAAQAEAAAAAAAAAAAAAAAAAD//////////2nwAAAAAAAA//////////8AAAAAAAAAAGAAAAAAAAAAAQA(...TRUNCATED) |
ball_in_cup_catch/5859 | "hf://datasets/ml-jku/dm_control_10M@30bb74e5535f6558e0e8e99c8afe5493e9ce4450/ball_in_cup_catch.tar.(...TRUNCATED) | "iUhERg0KGgoAAAAAAAgIAAQAEAAAAAAAAAAAAAAAAAD///////////LuAAAAAAAA//////////8AAAAAAAAAAGAAAAAAAAAAAQA(...TRUNCATED) |
ball_in_cup_catch/1389 | "hf://datasets/ml-jku/dm_control_10M@30bb74e5535f6558e0e8e99c8afe5493e9ce4450/ball_in_cup_catch.tar.(...TRUNCATED) | "iUhERg0KGgoAAAAAAAgIAAQAEAAAAAAAAAAAAAAAAAD//////////2zuAAAAAAAA//////////8AAAAAAAAAAGAAAAAAAAAAAQA(...TRUNCATED) |
ball_in_cup_catch/8696 | "hf://datasets/ml-jku/dm_control_10M@30bb74e5535f6558e0e8e99c8afe5493e9ce4450/ball_in_cup_catch.tar.(...TRUNCATED) | "iUhERg0KGgoAAAAAAAgIAAQAEAAAAAAAAAAAAAAAAAD//////////2bsAAAAAAAA//////////8AAAAAAAAAAGAAAAAAAAAAAQA(...TRUNCATED) |
ball_in_cup_catch/8958 | "hf://datasets/ml-jku/dm_control_10M@30bb74e5535f6558e0e8e99c8afe5493e9ce4450/ball_in_cup_catch.tar.(...TRUNCATED) | "iUhERg0KGgoAAAAAAAgIAAQAEAAAAAAAAAAAAAAAAAD//////////8DuAAAAAAAA//////////8AAAAAAAAAAGAAAAAAAAAAAQA(...TRUNCATED) |
ball_in_cup_catch/2339 | "hf://datasets/ml-jku/dm_control_10M@30bb74e5535f6558e0e8e99c8afe5493e9ce4450/ball_in_cup_catch.tar.(...TRUNCATED) | "iUhERg0KGgoAAAAAAAgIAAQAEAAAAAAAAAAAAAAAAAD//////////1DvAAAAAAAA//////////8AAAAAAAAAAGAAAAAAAAAAAQA(...TRUNCATED) |
End of preview. Expand in Data Studio
DMControl 10M
This repository contains the DMControl datasets as used in "A Large Recurrent Action Model: xLSTM enables Fast Inference for Robotics Tasks":
- Trajectories for every task are stored as separate
.tar.gzfiles. - Every
.tar.gzfile contains 10M transitions. - Every trajectory is stored as a separate
.hdf5file.
Download the dataset using the huggingface-cli:
huggingface-cli download ml-jku/dm_control_10M --local-dir=./dm_control_10M --repo-type dataset
For dataloading we refer to our Github repository: https://github.com/ml-jku/LRAM
Citation:
- Downloads last month
- 19
Size of the auto-converted Parquet files (First 5GB):
4.12 GB
Number of rows (First 5GB):
60,200
Estimated number of rows:
133,460