The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: ValueError
Message: Invalid string class label thistle-forge-1103@15401b309cd3f33456ed68bc3d1bb7d500e7b3cb
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 149, in get_rows_or_raise
return get_rows(
dataset=dataset,
...<4 lines>...
column_names=column_names,
)
File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
return func(*args, **kwargs)
File "/src/services/worker/src/worker/utils.py", line 129, in get_rows
rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
File "/src/services/worker/src/worker/utils.py", line 544, in safe_iter
yield from ds.decode(False) if ds.features else ds
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 3052, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2572, in __iter__
example = _apply_feature_types_on_example(
example, self.features, token_per_repo_id=self.token_per_repo_id
)
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2489, in _apply_feature_types_on_example
encoded_example = features.encode_example(example)
File "/usr/local/lib/python3.14/site-packages/datasets/features/features.py", line 2244, in encode_example
return encode_nested_example(self, example)
File "/usr/local/lib/python3.14/site-packages/datasets/features/features.py", line 1518, in encode_nested_example
{k: encode_nested_example(schema[k], obj.get(k), level=level + 1) for k in schema}
~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/features/features.py", line 1541, in encode_nested_example
return schema.encode_example(obj) if obj is not None else None
~~~~~~~~~~~~~~~~~~~~~^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/features/features.py", line 1210, in encode_example
example_data = self.str2int(example_data)
File "/usr/local/lib/python3.14/site-packages/datasets/features/features.py", line 1147, in str2int
output = [self._strval2int(value) for value in values]
~~~~~~~~~~~~~~~~^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/features/features.py", line 1168, in _strval2int
raise ValueError(f"Invalid string class label {value}")
ValueError: Invalid string class label thistle-forge-1103@15401b309cd3f33456ed68bc3d1bb7d500e7b3cbNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
RoboTwin 2.0 policy rollouts, round 2
Download as one file: robotwin_rollouts_r2.zip holds this whole dataset (this README and every task folder,
rollouts/success/<task>/ and rollouts/failure/<task>/), stored without compression:
hf download youtudou/thistle-forge-1103 robotwin_rollouts_r2.zip --repo-type dataset --local-dir .
unzip robotwin_rollouts_r2.zip
Every task folder is a complete LeRobot v2.1 dataset (meta/, data/, videos/): aloha-agilex, 14-D
observation.state and action (joint targets, aloha order), 50 fps, three 240 x 320 AV1 videos (cam_high,
cam_left_wrist, cam_right_wrist), RoboTwin's seen instruction templates. Same format as the round-2 recovery data
(youtudou/russet-lantern-1102) and round 1's rollouts (youtudou/slate-harbor-1027).
Policy. pi0.5 recovery40_orig50_taskuni_plain, EMA step 40000 (the round-1 recovery-trained checkpoint;
pi05_rt3d_full_2ep config). Inference: 30 of the 50 predicted actions executed per chunk, 20 denoising steps, fixed
sampling noise (seed 0), plain task instruction.
Rollouts. 200 per task on all 50 tasks, 10,000 in total: 100 demo_clean and 100 demo_randomized (50 with seen,
50 with unseen textures). Seeds are new: clean from 3,000,000, randomized seen-texture pool from 3,500,000, unseen pool
from 4,000,000 (highest 4,032,409); none is a RoboTwin official train/eval seed or a round-1 rollout seed; each seed was
kept only if RoboTwin's scripted expert solves it. Split by RoboTwin's check_success at the end: rollouts/success/<task>
(usable as demonstrations) and rollouts/failure/<task> (failed attempts, not demonstrations; 8
tasks have none).
Rendered in RoboTwin 2.0 (stable_2.0) from the simulator state recorded at every step. Every episode was checked against the camera images the policy saw during the rollout (edge-map correlation at every chunk start, all three cameras): states and actions are exact. 100 of 10000 episodes score under 0.9 at some frame; 61 of them only at featureless wrist frames (camera covered or facing a plain surface: the score is noise) and the other 39 at single low-texture wrist frames (0.67-0.90) or under RoboTwin's "crazy random light", which draws new light colours at every render; the inspected ones match the recording. Every video decodes with the episode's frame count.
Success rate per task (success / rollouts)
| task | clean | randomized |
|---|---|---|
adjust_bottle |
98/100 | 99/100 |
beat_block_hammer |
95/100 | 92/100 |
blocks_ranking_rgb |
91/100 | 90/100 |
blocks_ranking_size |
72/100 | 66/100 |
click_alarmclock |
100/100 | 100/100 |
click_bell |
100/100 | 100/100 |
dump_bin_bigbin |
100/100 | 99/100 |
grab_roller |
100/100 | 100/100 |
handover_block |
86/100 | 74/100 |
handover_mic |
96/100 | 99/100 |
hanging_mug |
29/100 | 18/100 |
lift_pot |
94/100 | 91/100 |
move_can_pot |
100/100 | 100/100 |
move_pillbottle_pad |
99/100 | 99/100 |
move_playingcard_away |
100/100 | 100/100 |
move_stapler_pad |
97/100 | 92/100 |
open_laptop |
99/100 | 100/100 |
open_microwave |
100/100 | 100/100 |
pick_diverse_bottles |
86/100 | 79/100 |
pick_dual_bottles |
95/100 | 92/100 |
place_a2b_left |
100/100 | 99/100 |
place_a2b_right |
97/100 | 98/100 |
place_bread_basket |
67/100 | 53/100 |
place_bread_skillet |
83/100 | 84/100 |
place_burger_fries |
97/100 | 100/100 |
place_can_basket |
84/100 | 81/100 |
place_cans_plasticbox |
98/100 | 99/100 |
place_container_plate |
98/100 | 98/100 |
place_dual_shoes |
72/100 | 76/100 |
place_empty_cup |
100/100 | 99/100 |
place_fan |
98/100 | 95/100 |
place_mouse_pad |
91/100 | 96/100 |
place_object_basket |
85/100 | 86/100 |
place_object_scale |
99/100 | 97/100 |
place_object_stand |
99/100 | 100/100 |
place_phone_stand |
99/100 | 97/100 |
place_shoe |
88/100 | 89/100 |
press_stapler |
96/100 | 99/100 |
put_bottles_dustbin |
64/100 | 65/100 |
put_object_cabinet |
83/100 | 80/100 |
rotate_qrcode |
86/100 | 89/100 |
scan_object |
68/100 | 69/100 |
shake_bottle |
100/100 | 100/100 |
shake_bottle_horizontally |
100/100 | 100/100 |
stack_blocks_three |
87/100 | 86/100 |
stack_blocks_two |
92/100 | 92/100 |
stack_bowls_three |
76/100 | 75/100 |
stack_bowls_two |
91/100 | 93/100 |
stamp_seal |
99/100 | 100/100 |
turn_switch |
63/100 | 73/100 |
| total | 4497/5000 (89.9 %) | 4458/5000 (89.2 %) |
success
| task | episodes | frames |
|---|---|---|
adjust_bottle |
197 | 23648 |
beat_block_hammer |
187 | 20573 |
blocks_ranking_rgb |
181 | 82231 |
blocks_ranking_size |
138 | 67223 |
click_alarmclock |
200 | 10598 |
click_bell |
200 | 10347 |
dump_bin_bigbin |
199 | 38199 |
grab_roller |
200 | 19180 |
handover_block |
160 | 48755 |
handover_mic |
195 | 40762 |
hanging_mug |
47 | 14560 |
lift_pot |
185 | 20696 |
move_can_pot |
200 | 28168 |
move_pillbottle_pad |
198 | 27343 |
move_playingcard_away |
200 | 19590 |
move_stapler_pad |
189 | 31980 |
open_laptop |
199 | 28502 |
open_microwave |
200 | 82746 |
pick_diverse_bottles |
165 | 17694 |
pick_dual_bottles |
187 | 19293 |
place_a2b_left |
199 | 26420 |
place_a2b_right |
195 | 26273 |
place_bread_basket |
120 | 38707 |
place_bread_skillet |
167 | 26987 |
place_burger_fries |
197 | 41690 |
place_can_basket |
165 | 43069 |
place_cans_plasticbox |
197 | 41769 |
place_container_plate |
196 | 25846 |
place_dual_shoes |
148 | 44606 |
place_empty_cup |
199 | 30763 |
place_fan |
193 | 28293 |
place_mouse_pad |
187 | 29024 |
place_object_basket |
171 | 44000 |
place_object_scale |
196 | 27918 |
place_object_stand |
199 | 25305 |
place_phone_stand |
196 | 27706 |
place_shoe |
177 | 30502 |
press_stapler |
195 | 12169 |
put_bottles_dustbin |
129 | 73853 |
put_object_cabinet |
163 | 43477 |
rotate_qrcode |
175 | 26681 |
scan_object |
137 | 21533 |
shake_bottle |
200 | 18227 |
shake_bottle_horizontally |
200 | 18482 |
stack_blocks_three |
173 | 78544 |
stack_blocks_two |
184 | 52942 |
stack_bowls_three |
151 | 72807 |
stack_bowls_two |
184 | 55041 |
stamp_seal |
199 | 26139 |
turn_switch |
136 | 18184 |
| total | 8955 | 1729045 |
failure
| task | episodes | frames |
|---|---|---|
adjust_bottle |
3 | 1203 |
beat_block_hammer |
13 | 5213 |
blocks_ranking_rgb |
19 | 22819 |
blocks_ranking_size |
62 | 74462 |
dump_bin_bigbin |
1 | 601 |
handover_block |
40 | 32040 |
handover_mic |
5 | 3005 |
hanging_mug |
153 | 137853 |
lift_pot |
15 | 6015 |
move_pillbottle_pad |
2 | 802 |
move_stapler_pad |
11 | 4411 |
open_laptop |
1 | 701 |
pick_diverse_bottles |
35 | 14035 |
pick_dual_bottles |
13 | 5213 |
place_a2b_left |
1 | 401 |
place_a2b_right |
5 | 2005 |
place_bread_basket |
80 | 56080 |
place_bread_skillet |
33 | 16533 |
place_burger_fries |
3 | 1503 |
place_can_basket |
35 | 24535 |
place_cans_plasticbox |
3 | 2403 |
place_container_plate |
4 | 1604 |
place_dual_shoes |
52 | 31252 |
place_empty_cup |
1 | 501 |
place_fan |
7 | 2807 |
place_mouse_pad |
13 | 5213 |
place_object_basket |
29 | 20329 |
place_object_scale |
4 | 1604 |
place_object_stand |
1 | 401 |
place_phone_stand |
4 | 1604 |
place_shoe |
23 | 11523 |
press_stapler |
5 | 2005 |
put_bottles_dustbin |
71 | 120771 |
put_object_cabinet |
37 | 25937 |
rotate_qrcode |
25 | 10025 |
scan_object |
63 | 31563 |
stack_blocks_three |
27 | 32427 |
stack_blocks_two |
16 | 12816 |
stack_bowls_three |
49 | 58849 |
stack_bowls_two |
16 | 14416 |
stamp_seal |
1 | 401 |
turn_switch |
64 | 25664 |
| total | 1045 | 823545 |
- Downloads last month
- 45