Episodes Preview Franka Panda Visualizer
1.08k episodes · 10 fps

LIBERO-X Levels 1-3, 218-task subset (LeRobot v3.0)

The 218 tasks by skill and signature, and the three seen/unseen splits: leave-task-out, leave-signature-out and leave-skill-out

A subset of the human teleoperation demonstrations of LIBERO-X, converted to the LeRobot v3.0 dataset format. It contains the demonstrations whose instruction matches one of the 218 Level 1-3 tasks (152 distinct instructions) used in our LIBERO-X experiments.

Papers

  1. Failure Prediction and Classification in Vision-Language-Action Models Via Calibrated Latent Probing
  2. LIBERO-X, the source of the demonstrations (RSS 2026)
Episodes 1,076 (2 to 44 per instruction)
Frames 371,555 at 10 fps
Instructions 152 distinct (the 218 tasks share instructions across scenes)
Cameras image (agent view) and wrist_image, 256x256 RGB, stored in the parquet files
State / action 8 / 7 float32 (see "Conventions")
Format LeRobot v3.0, 506 parquet data files, 39 GB
Robot Franka Panda, simulation (robosuite / LIBERO)

Provenance, license and changes

The data comes from LIBERO-X by Meituan (dataset, code, paper, accepted at RSS 2026), released under CC BY 4.0. This subset is released under the same license. If you use it, please cite the LIBERO-X paper (BibTeX: see the paper page) and credit the original authors. This subset is not endorsed by them.

Changes made to the original (the images, states and actions themselves are unchanged):

  1. Selection: only episodes whose instruction text (whitespace-stripped) equals one of the 152 instructions of the 218 tasks (see "Why these tasks"). The original has 2,520 episodes.
  2. Format: converted from LeRobot v2.1 to v3.0 with LeRobot's own converter (0.7.0). Episodes are packed into larger parquet files.
  3. Renumbering: episodes are renumbered 0 to 1,075 and the global index is recomputed. The map to the original numbering is in source/episode_map.json.
  4. Tasks: the original has 41 instructions that appear twice under different task_index values (and one with a trailing space). They are merged by text, so there are 152 tasks, renumbered 0 to 151. The original task_index of each episode is kept in source/episode_map.json.

Why these tasks

The 218 tasks are the finetune_set_v2 selection: LIBERO-X extension tasks (EXTENSION_KITCHEN / EXTENSION_STUDY scenes) of levels 1-3 that are backed by human demonstrations (72 Level 1, 73 Level 2, 73 Level 3). Each task has a skill (8 in total: place_on_surface, place_in_container, place_relative_to_landmark, stack_objects, articulate_open, articulate_close, toggle_appliance_on, toggle_appliance_off) and a signature (skill | target object | spatial relation | landmark | single or compound instruction; 72 in total). To keep the set balanced, the two dominant placement skills are capped at 3 task instances per signature, while the under-represented skills are uncapped (every demo-backed instance is included), and the three levels are roughly even. The set is kitchen-heavy (194 kitchen and 24 study tasks).

The full per-task table is splits/splits.csv (level, skill, sig, room, scene, target_class, relation, landmark, compound, n_demos, file, lang and the split columns). n_demos there was counted by instruction text alone, so tasks that share an instruction share the same demos, and the column over-counts if summed.

Splits

In our experiments the VLA policy was fine-tuned on all tasks, and three seen/unseen splits over the 218 tasks were used to test the generalization of a latent probe. In principle the same splits can be used to test the generalization of a policy itself. They are in increasing difficulty:

split what is unseen what has been seen
leave_task_out (175 seen / 43 unseen tasks) 20 % of the task instances (scene + instruction), stratified by skill (seed 0) the same skill and, for most unseen tasks, the same signature
leave_signature_out (162 / 56) 16 whole signatures (about 20 % per skill, at most half of a skill's signatures) the skill, but not that object / relation / landmark combination
leave_skill_out (8 folds) one of the 8 skills entirely the other skills only; the primitive itself is new

On leave_signature_out, one unseen signature (stack_objects | bowl | on | bowl | single) holds 16 of the 17 stack_objects tasks, so that skill has a single seen task.

The splits are defined on tasks (scene + instruction), but the episodes only carry the instruction text. Episodes are therefore assigned to splits through their instruction, in splits/episode_splits.csv (one row per episode) and splits/instruction_splits.csv (one row per instruction):

  • leave_signature_out and leave_skill_out are clean: every instruction belongs to exactly one signature and one skill, so each episode is cleanly seen or unseen (train / test).
  • leave_task_out is not clean at the episode level: 19 of the 152 instructions are used by both a seen and an unseen task (in different scenes). Their episodes are labelled mixed. Only the 18 test-only instructions are guaranteed unseen. Decide how to treat mixed episodes (for example drop them).

Contents

data/chunk-*/file-*.parquet     frames (images, state, actions, indices)
meta/                           LeRobot v3.0 metadata (info.json, tasks.parquet, stats.json, episodes/)
splits/splits.csv               the 218 tasks with skill, signature, level, scene and split labels
splits/instruction_splits.csv   one row per instruction (152): skill, signature, levels, split labels, episode count
splits/episode_splits.csv       one row per episode: its instruction and split labels
source/episode_map.json         new episode index -> original LIBERO-X episode and task index

Conventions

  • state (8): end-effector position x, y, z in metres, end-effector orientation as an axis-angle vector (the original metadata calls it roll/pitch/yaw), and the two gripper finger positions.
  • actions (7): [dx, dy, dz, drot_x, drot_y, drot_z, gripper], controller commands for robosuite's OSC_POSE (delta end-effector pose, normalised to about [-1, 1]); gripper is -1 (open) or +1 (close).
  • The world frame is centred on the table (not on the robot base) and is the same for all scenes: +x points from the robot across the table. The base is at x of about -0.66 m (kitchen) or -0.75 m (study).
  • Actions are teleoperation commands, not differences of the states.

Known quirks

  • A few action outliers: 0.74 % of frames have an action component above 1 in magnitude, and the extremes reach -20 and +9 (in 10 episodes above 5). Consider clipping or filtering, especially with mean/std normalisation.

Loading

from lerobot.datasets import LeRobotDataset

ds = LeRobotDataset("podolinsky/libero-x-level123")   # needs access to this repo
print(ds.num_episodes, ds.num_frames)   # 1076 371555

Related: the rollouts, embeddings and failure labels of a GR00T policy on these 218 tasks are in podolinsky/gr00t-libero-x-features-v3.


Note: demonstrations are labelled by instruction text only in LIBERO-X, not by scene, so episode-level labels for leave_task_out are approximate.

Downloads last month
156

Models trained or fine-tuned on podolinsky/libero-x-level123

Collection including podolinsky/libero-x-level123

Papers for podolinsky/libero-x-level123