Dataset Viewer
The dataset viewer is not available for this subset.
Cannot get the split names for the config 'default' of the dataset.
Exception:    SplitsNotFoundError
Message:      The split names could not be parsed from the dataset config.
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
                  for split_generator in builder._split_generators(
                                         ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/webdataset/webdataset.py", line 80, in _split_generators
                  raise ValueError(
                  ...<2 lines>...
                  )
              ValueError: The TAR archives of the dataset should be in WebDataset format, but the files in the archive don't share the same prefix or the same types.
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 68, in compute_split_names_from_streaming_response
                  for split in get_dataset_split_names(
                               ~~~~~~~~~~~~~~~~~~~~~~~^
                      path=dataset,
                      ^^^^^^^^^^^^^
                      config_name=config,
                      ^^^^^^^^^^^^^^^^^^^
                      token=hf_token,
                      ^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
                  info = get_dataset_config_info(
                      path,
                  ...<6 lines>...
                      **config_kwargs,
                  )
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
                  raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
              datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

EgoDex Depth Cache

Per-frame metric scene depth (Depth-Anything-V2 Metric Indoor Large, 20 m ceiling) for the episodes inpainted in VR-VLA/VR-egodex-inpaint-cache-full. Depth is the second input Eq. 4 compositing needs beside the inpainted plate; together they let scripts/egodex/build_task.py retarget an episode onto all 16 embodiments.

Pairing with the inpaint cache

Archive names are IDENTICAL to the inpaint repo, one tar per <split>/<task>:

inpaint  <split>/<task>.tar -> <split>/<task>/inpaint_diffueraser/episode<N>/frame%04d.png
depth    <split>/<task>.tar -> <split>/<task>/depth/episode<N>/frame%04d.png

Episode numbers match the source dataset's episode_%06d.mp4. <split> is part of the key: most task names exist in both train/ and test/ with overlapping episode numbers.

Each archive expands to:

<split>/<task>/depth/episode<N>/frame%04d.png   one per source frame, 0-based, 30 fps
<split>/<task>/depth/episode<N>/done.json       {num_frames, format, created}

Archives are uncompressed tar (PNG payload). train/add_remove_lid_v2.1.tar (308 GB, 1,158 episodes) exceeds the Hub's 50 GB file limit and is stored as seven .tar.NN.part files; cat train/add_remove_lid_v2.1.tar.*.part > train/add_remove_lid_v2.1.tar before extracting.

Format

16-bit grayscale PNG, native 1920x1080, fixed scale: metres = value / 65535 * 20.0 (Depth-Anything-V2's metric-indoor ceiling; values beyond 20 m clamp). Resolution 0.31 mm, ~1.0 MB/frame, 5x smaller than float16 .npy. Fixed scale, not per-frame min/max, so no sidecar is needed and frames are directly comparable.

import numpy as np
from PIL import Image
depth_m = np.asarray(Image.open('frame0000.png'), dtype=np.float32) / 65535.0 * 20.0  # (1080, 1920)

In the retarget repo: png_to_depth() in src/retarget_egocentric_data/retarget/visual/stage3_scene_depth.py; load_visual_alignment_cache() accepts .png or .npy transparently.

Batches

Batch membership is per EPISODE. A task tar on the inpaint repo can hold episodes of several batches; batches.json in this repo lists, per <split>/<task>, which episode ids belong to which batch. Each batch is retargeted on ONE server; a server builds its manifest with scripts/egodex/manifest.py --batches batches.json --batch N and only that batch's episodes are staged, so no episode is ever retargeted twice.

batch content episodes frames archives depth here
1 original server's local cache (EgodexRetarget) 659 282,240 after batch-1 campaign finishes
2 every other task, remaining episodes 1,050 405,111 232 yes
3 train/add_remove_lid_v2.1 only (43% of remaining frames) 1,158 311,580 1 (308 GB, 7 parts) yes

Batch 3 is one task on purpose: it alone is as large as the rest of the corpus, so the server that takes it downloads exactly two archives (inpaint + depth) and nothing else. Episodes under 23 frames have masks but no inpainted frames in the inpaint repo and are in no batch (masks_only in batches.json).

Workflow on a new server (batch N = 2 or 3)

  1. Fetch batches.json first — it is the source of truth for what is left to do: hf download VR-VLA/VR-egodex-depth-cache-full batches.json README.md --repo-type dataset --local-dir <depth>
  2. batches.json['batchN_tars'] lists the archives to fetch, by identical name from BOTH repos (batch 2: 232 tars; batch 3: 1 tar, stored as .tar.NN.part pieces in both repos). Inpaint tars also contain other batches' episodes — that is fine, the next step filters them.
  3. python scripts/egodex/manifest.py --cache-root <inpaint> --depth-cache-root <depth> --batches <depth>/batches.json --batch N --out <OUT_ROOT>/_manifest/episode_manifest.json keeps only batchN episode ids per task.
  4. scripts/egodex/launch_batch2.sh — same launcher for batch 2 and 3, no prototype videos (see scripts/egodex/README.md, "Batches").

Per-task episode ids

task batch 1 batch 2 batch 3
test/add_remove_lid_v2.1 0, 1, 2 3, 4, 5 -
test/arrange_topple_dominoes_v2.1 0, 1, 2 3, 4, 5 -
test/assemble_disassemble_furniture_bench_chair_v2.1 0, 1, 2 3, 4, 5, 6 -
test/assemble_disassemble_furniture_bench_desk_v2.1 0, 1, 2 3, 4, 5 -
test/assemble_disassemble_furniture_bench_drawer_v2.1 0, 1, 2 3, 4, 5 -
test/assemble_disassemble_furniture_bench_lamp_v2.1 0, 1, 2 3, 4, 5 -
test/assemble_disassemble_furniture_bench_square_table_v2.1 0, 1, 2 3, 4, 5, 6 -
test/assemble_disassemble_furniture_bench_stool_v2.1 0, 1, 2 3, 4, 5 -
test/assemble_disassemble_legos_v2.1 0, 1, 2 3, 4, 5 -
test/assemble_disassemble_soft_legos_v2.1 0, 1, 2 3, 4, 5 -
test/assemble_disassemble_structures_v2.1 0, 1, 2 3, 4, 5, 6 -
test/assemble_disassemble_tiles_v2.1 0, 1, 2 3, 4, 5 -
test/assemble_jenga_v2.1 0, 1, 2 3, 4, 5 -
test/basic_fold_v2.1 0, 1, 2 3, 4, 5 -
test/basic_pick_place_v2.1 0, 1, 2 3, 4, 5, 6 -
test/boil_serve_egg_v2.1 0, 1, 2 3, 4 -
test/braid_unbraid_v2.1 0, 1, 2 3, 4, 5 -
test/build_unstack_lego_v2.1 0, 1, 2 3, 4, 5 -
test/charge_uncharge_airpods_v2.1 0, 1, 2 3, 4, 5 -
test/charge_uncharge_device_v2.1 0, 1, 2 3, 4, 5 -
test/clean_cups_v2.1 0, 1, 2 3, 4, 5 -
test/clean_surface_v2.1 0, 1, 2 3, 4, 5 -
test/clean_tableware_v2.1 0, 1, 2 3, 4, 5 -
test/clip_unclip_papers_v2.1 0, 1, 2 3, 4 -
test/color_v2.1 0, 1, 2 3, 4, 5 -
test/crumple_flatten_paper_v2.1 0, 1, 2 3, 4 -
test/deal_gather_cards_v2.1 0, 1, 2 3, 4, 5 -
test/declutter_desk_v2.1 0, 1, 2 3, 4, 5 -
test/dry_hands_v2.1 0, 1, 2 3, 4, 5 -
test/fidget_magnetic_spinner_rings_v2.1 0, 1, 2 3, 4, 5 -
test/flip_coin_v2.1 0, 1 - -
test/flip_pages_v2.1 0, 1, 2 3, 4, 5 -
test/fold_stack_unstack_unfold_cloths_v2.1 0, 1, 2 3, 4, 5 -
test/fold_unfold_paper_basic_v2.1 0, 1, 2 3, 4, 5 -
test/fold_unfold_paper_origami_v2.1 0, 1, 2 3, 4, 5 -
test/fry_bread_v2.1 0 - -
test/fry_egg_v2.1 0 - -
test/gather_roll_dice_v2.1 0, 1, 2 3, 4, 5 -
test/insert_dump_blocks_v2.1 0, 1, 2 3, 4, 5 -
test/insert_remove_airpods_v2.1 0, 1, 2 3, 4, 5 -
test/insert_remove_bagging_v2.1 0, 1, 2 3, 4, 5 -
test/insert_remove_bookshelf_v2.1 0, 1, 2 3, 4, 5 -
test/insert_remove_cups_from_rack_v2.1 0, 1, 2 3, 4, 5 -
test/insert_remove_drawer_v2.1 0, 1, 2 3, 4, 5 -
test/insert_remove_furniture_bench_cabinet_v2.1 0, 1, 2 3, 4, 5 -
test/insert_remove_furniture_bench_round_table_v2.1 0, 1, 2 3, 4, 5 -
test/insert_remove_plug_socket_v2.1 0, 1, 2 3, 4, 5 -
test/insert_remove_shirt_in_tube_v2.1 0, 1, 2 3, 4, 5 -
test/insert_remove_tennis_ball_v2.1 0, 1, 2 3, 4, 5 -
test/insert_remove_usb_v2.1 0, 1, 2 3, 4, 5 -
test/insert_remove_utensils_v2.1 0, 1, 2 3, 4, 5 -
test/knead_slime_v2.1 0, 1, 2 3, 4, 5 -
test/load_dispense_ice_v2.1 0, 1, 2 3, 4, 5 -
test/lock_unlock_key_v2.1 0, 1, 2 3, 4, 5 -
test/make_sandwich_v2.1 0, 1, 2 3, 4, 5 -
test/measure_objects_v2.1 0, 1, 2 3, 4, 5 -
test/open_close_insert_remove_box_v2.1 0, 1, 2 3, 4, 5 -
test/open_close_insert_remove_case_v2.1 0, 1, 2 3, 4, 5 -
test/open_close_insert_remove_tupperware_v2.1 0, 1, 2 3, 4, 5 -
test/paint_clean_brush_v2.1 0, 1, 2 3, 4, 5 -
test/peel_place_sticker_v2.1 0, 1, 2 3, 4, 5 -
test/pick_place_food_v2.1 0, 1, 2 3, 4, 5 -
test/pick_up_and_put_down_case_or_bag_v2.1 0, 1, 2 3, 4, 5 -
test/play_mancala_v2.1 0, 1, 2 3, 4, 5 -
test/play_piano_v2.1 0, 1, 2 3, 4, 5 -
test/play_reset_connect_four_v2.1 0, 1, 2 3, 4, 5 -
test/point_and_click_remote_v2.1 0, 1, 2 3, 4, 5 -
test/pour_v2.1 0, 1, 2 3, 4, 5 -
test/push_pop_toy_v2.1 0, 1, 2 3, 4, 5 -
test/put_away_set_up_board_game_v2.1 0, 1, 2 3, 4, 5 -
test/put_in_take_out_glasses_v2.1 0, 1, 2 3, 4, 5 -
test/put_toothpaste_on_toothbrush_v2.1 0, 1, 2 3, 4, 5 -
test/rake_smooth_zen_garden_v2.1 0, 1, 2 3, 4, 5 -
test/roll_ball_v2.1 0, 1, 2 3, 4, 5 -
test/scoop_dump_ice_v2.1 0, 1, 2 3, 4, 5 -
test/screw_unscrew_allen_fixture_v2.1 0, 1, 2 3, 4, 5 -
test/screw_unscrew_bottle_cap_v2.1 0, 1, 2 3, 4, 5 -
test/screw_unscrew_fingers_fixture_v2.1 0, 1, 2 3, 4, 5 -
test/set_up_clean_up_chessboard_v2.1 0, 1, 2 - -
test/setup_cleanup_table_v2.1 0, 1, 2 3, 4, 5 -
test/sleeve_unsleeve_cards_v2.1 0, 1, 2 3 -
test/slot_batteries_v2.1 0, 1, 2 3, 4, 5 -
test/sort_beads_v2.1 0, 1, 2 3, 4, 5 -
test/stack_remove_jenga_v2.1 0, 1, 2 3, 4, 5 -
test/stack_unstack_bowls_v2.1 0, 1, 2 3, 4, 5 -
test/stack_unstack_cups_v2.1 0, 1, 2 3, 4, 5 -
test/stack_unstack_plates_v2.1 0, 1, 2 3, 4, 5 -
test/stack_unstack_tupperware_v2.1 0, 1, 2 3, 4, 5 -
test/stack_v2.1 0, 1, 2 3, 4, 5 -
test/staple_paper_v2.1 0, 1, 2 3, 4, 5 -
test/stock_unstock_fridge_v2.1 0, 1, 2 3, 4, 5 -
test/sweep_dustpan_v2.1 0, 1, 2 3, 4, 5 -
test/thread_unthread_bead_necklace_v2.1 0, 1, 2 3, 4, 5 -
test/throw_and_catch_ball_v2.1 0, 1, 2 3, 4, 5 -
test/throw_collect_objects_v2.1 0, 1, 2 3, 4, 5 -
test/tie_and_untie_shoelace_v2.1 0, 1, 2 3, 4, 5 -
test/tie_untie_rubberband_v2.1 0, 1, 2 3, 4, 5 -
test/type_keyboard_v2.1 0, 1, 2 3, 4, 5 -
test/use_chopsticks_v2.1 0, 1, 2 3, 4, 5 -
test/use_rubiks_cube_v2.1 0, 1, 2 3, 4, 5 -
test/vertical_pick_place_v2.1 0, 1, 2 3, 4, 5 -
test/wash_fruit_v2.1 0 - -
test/wash_kitchen_dishes_v2.1 0 - -
test/wash_put_away_dishes_v2.1 0, 1 - -
test/wipe_kitchen_surfaces_v2.1 0, 1, 2 3, 4, 5 -
test/wipe_screen_v2.1 0, 1, 2 3, 4, 5 -
test/wrap_unwrap_food_v2.1 0, 1, 2 3, 4, 5 -
test/wrap_v2.1 0, 1, 2 3, 4, 5 -
test/write_v2.1 0, 1, 2 3, 4, 5 -
test/zip_unzip_bag_v2.1 0, 1, 2 3, 4, 5 -
test/zip_unzip_case_v2.1 0, 1, 2 3, 4, 5 -
train/add_remove_lid_v2.1 - - 1158 episodes, see batches.json
train/arrange_topple_dominoes_v2.1 0, 1, 2 3, 4, 5 -
train/assemble_disassemble_furniture_bench_chair_v2.1 0, 1, 2 3, 4, 5, 469, 471, 473, 475, 477, 479 -
train/assemble_disassemble_furniture_bench_desk_v2.1 0, 1, 2 3, 4, 5 -
train/assemble_disassemble_furniture_bench_drawer_v2.1 0, 1, 2 3, 4, 5, 468, 470, 472, 474, 476, 478, 480 -
train/assemble_disassemble_furniture_bench_lamp_v2.1 0, 1, 2 3, 4, 5 -
train/assemble_disassemble_furniture_bench_square_table_v2.1 0, 1, 2 3, 4, 5, 469, 471, 473, 475, 477, 479, 481 -
train/assemble_disassemble_furniture_bench_stool_v2.1 0, 1, 2 3, 4, 5 -
train/assemble_disassemble_legos_v2.1 0, 1, 2 3, 4, 5, 468, 470, 472, 474, 476, 478 -
train/assemble_disassemble_soft_legos_v2.1 0, 1, 2 3, 4, 5 -
train/assemble_disassemble_structures_v2.1 0, 1, 2 3, 4, 5, 469, 471, 473, 475, 477, 479 -
train/assemble_disassemble_tiles_v2.1 0, 1, 2 3, 4, 5 -
train/assemble_jenga_v2.1 0, 1, 2 3, 4, 5, 468, 470, 472, 474, 476, 478, 480 -
train/basic_fold-part1_v2.1 0, 1, 2 3, 4, 5 -
train/basic_fold-part2_v2.1 0, 1, 2 3, 4, 5, 467, 469, 471, 473, 475, 477, 479, 481 -
train/basic_fold-part3_v2.1 0, 1, 2 3, 4, 5 -
train/basic_fold-part4_v2.1 0, 1, 2 3, 4, 5, 468, 470, 472, 474, 476, 478 -
train/basic_fold-part5_v2.1 0, 1, 2 3, 4, 5 -
train/basic_fold-part6_v2.1 0, 1, 2 3, 4, 5, 467, 469, 471, 473, 475, 477, 479 -
train/basic_pick_place-part1_v2.1 0, 1, 2 3, 4, 5 -
train/basic_pick_place-part2_v2.1 0, 1, 2 3, 4, 5, 468, 470, 472, 474, 476, 478, 480 -
train/basic_pick_place-part3_v2.1 0, 1, 2 3, 4, 5 -
train/basic_pick_place-part4_v2.1 0, 1, 2 3, 4, 5, 467, 469, 471, 473, 475, 477, 479, 481 -
train/basic_pick_place-part5_v2.1 0, 1, 2 3, 4, 5 -
train/basic_pick_place-part6_v2.1 0, 1, 2 3, 4, 5, 468, 470, 472, 474, 476, 478 -
train/boil_serve_egg_v2.1 0, 1, 2 3, 4, 5 -
train/braid_unbraid_v2.1 0, 1, 2 3, 4, 5, 467, 469, 471, 473, 475, 477, 479 -
train/build_unstack_lego_v2.1 0, 1, 2 3, 4, 5 -
train/charge_uncharge_airpods_v2.1 0, 1, 2 3, 4, 5, 468, 470, 472, 474, 476, 478, 480 -
train/charge_uncharge_device_v2.1 0, 1, 2 3, 4, 5 -
train/clean_cups_v2.1 0, 1, 2 3, 4, 5, 467, 469, 471, 473, 475, 477, 479, 481 -
train/clean_surface_v2.1 0, 1, 2 3, 4, 5 -
train/clean_tableware_v2.1 0, 1, 2 3, 4, 5, 468, 470, 472, 474, 476, 478 -
train/clip_unclip_papers_v2.1 0, 1, 2 3, 4 -
train/color_v2.1 0, 1, 2 3, 4, 5 -
train/crumple_flatten_paper_v2.1 0, 1, 2 3, 4, 5 -
train/deal_gather_cards_v2.1 0, 1, 2 3, 4, 5, 467, 469, 471, 473, 475, 477, 479 -
train/declutter_desk_v2.1 0, 1, 2 3, 4, 5 -
train/dry_hands_v2.1 0, 1, 2 3, 4, 5, 468, 470, 472, 474, 476, 478, 480 -
train/fidget_magnetic_spinner_rings_v2.1 0, 1, 2 3, 4, 5 -
train/flip_coin_v2.1 0, 1, 2 3, 4, 5 -
train/flip_pages_v2.1 0, 1, 2 3, 4, 467, 469, 471, 473, 475, 477, 479, 481 -
train/fold_stack_unstack_unfold_cloths_v2.1 0, 1, 2 3, 4, 5 -
train/fold_unfold_paper_basic_v2.1 0, 1, 2 3, 4, 5, 468, 470, 472, 474, 476, 478 -
train/fold_unfold_paper_origami_v2.1 0, 1, 2 3, 4, 5 -
train/fry_bread_v2.1 0, 1, 2 3, 4 -
train/fry_egg_v2.1 0, 1, 2 3, 4, 5 -
train/gather_roll_dice_v2.1 0, 1, 2 3, 4, 5, 467, 469, 471, 473, 475, 477, 479 -
train/insert_dump_blocks_v2.1 0, 1, 2 3, 4, 5 -
train/insert_remove_airpods_v2.1 0, 1, 2 3, 4, 468, 470, 472, 474, 476, 478, 480 -
train/insert_remove_bagging_v2.1 0, 1, 2 3, 4, 5 -
train/insert_remove_bookshelf_v2.1 0, 1, 2 3, 4, 5, 467, 469, 471, 473, 475, 477, 479, 481 -
train/insert_remove_cups_from_rack_v2.1 0, 1, 2 3, 4, 5 -
train/insert_remove_drawer_v2.1 0, 1, 2 3, 4, 468, 470, 472, 474, 476, 478 -
train/insert_remove_furniture_bench_cabinet_v2.1 0, 1, 2 3, 4, 5 -
train/insert_remove_furniture_bench_round_table_v2.1 0, 1, 2, 1759 3, 4, 5, 467, 469, 471, 473, 475, 477, 479, 1761 -
train/insert_remove_plug_socket_v2.1 0, 1, 2 3, 4, 5 -
train/insert_remove_shirt_in_tube_v2.1 0, 1, 2 3, 4, 468, 470, 472, 474, 476, 478, 480 -
train/insert_remove_tennis_ball_v2.1 0, 1, 2 3, 4, 5 -
train/insert_remove_usb_v2.1 0, 1, 2 3, 4, 5, 467, 469, 471, 473, 475, 477, 479, 481 -
train/insert_remove_utensils_v2.1 0, 1, 2 3, 4, 5 -
train/knead_slime_v2.1 0, 1, 2 3, 4, 468, 470, 472, 474, 476, 478 -
train/load_dispense_ice_v2.1 0, 1, 2 3, 4, 5 -
train/lock_unlock_key_v2.1 0, 1, 2 3, 4, 5, 467, 469, 471, 473, 475, 477, 479 -
train/make_sandwich_v2.1 0, 1, 2 3, 4, 5 -
train/measure_objects_v2.1 0, 1, 2 3, 4, 468, 470, 472, 474, 476, 478, 480 -
train/open_close_insert_remove_box_v2.1 0, 1, 2 3, 4, 5 -
train/open_close_insert_remove_case_v2.1 0, 1, 2 3, 4, 5, 467, 469, 471, 473, 475, 477, 479, 481 -
train/open_close_insert_remove_tupperware_v2.1 0, 1, 2 3, 4, 5 -
train/paint_clean_brush_v2.1 0, 1, 2 3, 4, 468, 470, 472, 474, 476, 478 -
train/peel_place_sticker_v2.1 0, 1, 2 3, 4, 5 -
train/pick_place_food_v2.1 0, 1 2, 3, 4, 466, 468, 470, 472, 474, 476, 478 -
train/pick_up_and_put_down_case_or_bag_v2.1 0, 1, 2 3, 4, 5 -
train/play_mancala_v2.1 0, 1 2, 3, 467, 469, 471, 473, 475, 477, 479 -
train/play_piano-part1_v2.1 0, 1, 2 3, 4, 5 -
train/play_piano-part2_v2.1 0, 1 2, 3, 4, 466, 468, 470, 472, 474, 476, 478, 480 -
train/play_reset_connect_four_v2.1 0, 1, 2 3, 4, 5 -
train/point_and_click_remote_v2.1 0, 1 2, 3, 467, 469, 471, 473, 475, 477 -
train/pour_v2.1 0, 1, 2 3, 4, 5 -
train/push_pop_toy_v2.1 0, 1 2, 3, 4, 466, 468, 470, 472, 474, 476, 478 -
train/put_away_set_up_board_game_v2.1 0, 1, 2 3, 4, 5 -
train/put_in_take_out_glasses_v2.1 0, 1 2, 3, 467, 469, 471, 473, 475, 477, 479 -
train/put_toothpaste_on_toothbrush_v2.1 0, 1, 2 3, 4, 5 -
train/rake_smooth_zen_garden_v2.1 0, 1 2, 3, 4, 466, 468, 470, 472, 474, 476, 478, 480 -
train/roll_ball_v2.1 0, 1, 2 3, 4 -
train/scoop_dump_ice_v2.1 0, 1 2, 3, 467, 469, 471, 473, 475, 477 -
train/screw_unscrew_allen_fixture_v2.1 0, 1, 2 3, 4, 5 -
train/screw_unscrew_bottle_cap_v2.1 0, 1 2, 3, 4, 466, 468, 470, 472, 474, 476, 478 -
train/screw_unscrew_fingers_fixture_v2.1 0, 1, 2 3, 4 -
train/set_up_clean_up_chessboard_v2.1 0, 1 2, 3 -
train/setup_cleanup_table_v2.1 0, 1 2, 3, 4, 467, 469, 471, 473, 475, 477, 479 -
train/sleeve_unsleeve_cards_v2.1 0, 1 2, 3, 4 -
train/slot_batteries_v2.1 0, 1 2, 3 -
train/sort_beads_v2.1 0, 1 2, 3, 466, 468, 470, 472, 474, 476, 478 -
train/stack_remove_jenga_v2.1 0, 1 2, 3, 4 -
train/stack_unstack_bowls_v2.1 0, 1 2, 3, 4, 467, 469, 471, 473, 475, 477 -
train/stack_unstack_cups_v2.1 0, 1 2, 3 -
train/stack_unstack_plates_v2.1 0, 1 2, 3, 466, 468, 470, 472, 474, 476, 478 -
train/stack_unstack_tupperware_v2.1 0, 1 2, 3, 4 -
train/stack_v2.1 0, 1 2, 3, 4, 467, 469, 471, 473, 475, 477, 479 -
train/staple_paper_v2.1 0, 1 2, 3 -
train/stock_unstock_fridge_v2.1 0, 1 2, 3, 466, 468, 470, 472, 474, 476 -
train/sweep_dustpan_v2.1 0, 1 2, 3 -
train/thread_unthread_bead_necklace-part1_v2.1 0, 1 2, 3, 4, 467, 469, 471, 473, 475, 477 -
train/thread_unthread_bead_necklace-part2_v2.1 0, 1 2, 3 -
train/thread_unthread_bead_necklace-part3_v2.1 0, 1 2, 3, 466, 468, 470, 472, 474, 476, 478 -
train/throw_and_catch_ball_v2.1 0, 1 2, 3 -
train/throw_collect_objects_v2.1 0, 1 2, 3, 4, 467, 469, 471, 473, 475, 477, 479 -
train/tie_and_untie_shoelace_v2.1 0, 1 2, 3 -
train/tie_untie_rubberband_v2.1 0, 1 2, 3, 466, 468, 470, 472, 474, 476 -
train/type_keyboard_v2.1 0, 1 2, 3 -
train/use_chopsticks_v2.1 0, 1 2, 3, 4, 467, 469, 471, 473, 475, 477 -
train/use_rubiks_cube_v2.1 0, 1 2, 3 -
train/vertical_pick_place-part1_v2.1 0, 1 2, 3, 466, 468, 470, 472, 474, 476, 478 -
train/vertical_pick_place-part2_v2.1 0, 1 2, 3 -
train/vertical_pick_place-part3_v2.1 0, 1 2, 3, 4, 467, 469, 471, 473, 475, 477, 479 -
train/vertical_pick_place-part4_v2.1 0, 1 2, 3 -
train/vertical_pick_place-part5_v2.1 0, 1 2, 3, 466, 468, 470, 472, 474, 476 -
train/vertical_pick_place-part6_v2.1 0, 1 2, 3 -
train/wash_fruit_v2.1 0, 1 2, 3, 4 -
train/wash_kitchen_dishes_v2.1 0, 1 2, 3 -
train/wash_put_away_dishes_v2.1 0, 1 2, 3 -
train/wipe_kitchen_surfaces_v2.1 0, 1 2, 3 -
train/wipe_screen_v2.1 0, 1 2, 3, 4, 467, 469, 471, 473, 475, 477 -
train/wrap_unwrap_food_v2.1 0, 1 2, 3 -
train/wrap_v2.1 0, 1 2, 3, 466, 468, 470, 472, 474, 476, 478 -
train/write_v2.1 0, 1 2, 3 -
train/zip_unzip_bag_v2.1 0, 1 2, 3, 4, 467, 469, 471, 473, 475, 477, 479 -
train/zip_unzip_case_v2.1 0, 1 2, 3 -

Download and extract one task

hf download VR-VLA/VR-egodex-depth-cache-full train/clean_surface_v2.1.tar --repo-type dataset --local-dir .
tar -xf train/clean_surface_v2.1.tar

Generation

scripts/egodex/depth_batch.py (Depth-Anything-V2-Metric-Indoor-Large-hf via transformers, fp32, one frame at a time from the raw EgoDex mp4), 0.11 s/frame on one H100. Packed and uploaded per task by scripts/egodex/depth_pack_upload.py.

Downloads last month
314