Dataset Viewer
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
vae_id: string
vae_subfolder: string
latent_contract: struct<latent_channels: int64, scaling_factor: double, shift_factor: null>
child 0, latent_channels: int64
child 1, scaling_factor: double
child 2, shift_factor: null
scaling_factor_applied: double
shift_factor_applied: null
pad_to: int64
legacy: bool
stride: int64
source_stored_tic_stride: int64
align_decisions: bool
every_tic: bool
row_semantics: string
decode_check: struct<frames: int64, latent_shape: list<item: int64>, psnr_mean: double, psnr_min: double, latent_a (... 40 chars omitted)
child 0, frames: int64
child 1, latent_shape: list<item: int64>
child 0, item: int64
child 2, psnr_mean: double
child 3, psnr_min: double
child 4, latent_abs_mean: double
child 5, latent_abs_max: double
git: string
torch: string
args: struct<in_dir: string, out_dir: string, stride: int64, batch_size: int64, decode_threads: int64, dec (... 430 chars omitted)
child 0, in_dir: string
child 1, out_dir: string
child 2, stride: int64
child 3, batch_size: int64
child 4, decode_threads: int64
child 5, decode_workers: int64
child 6, device: string
child 7, dtype: string
child 8, shard: null
child 9, num_shards: int64
child 10, episode_ids: string
child 11, canonical: string
child 12, canonical_ids: string
child 13, canonical_only: bool
child 14, rebuild_sidecar_masks: bool
child 15, dry_run: bool
child 16, max_episodes: int64
child 17, vae_id: string
child 18, vae_subfolder: string
child 19, cache_dir: string
child 20, latent_channels: null
child 21, scaling_factor: null
child 22, shift_factor: null
child 23, decode_check: int64
child 24, legacy: bool
child 25, align_decisions: bool
child 26, every_tic: bool
18: string
20: string
6: string
15: string
24: string
17: string
23: string
5: string
8: string
21: string
2: string
14: string
10: string
22: string
26: string
16: string
3: string
7: string
27: string
0: string
12: string
9: string
19: string
13: string
4: string
28: string
25: string
1: string
11: string
to
{'19': Value('string'), '22': Value('string'), '15': Value('string'), '9': Value('string'), '5': Value('string'), '7': Value('string'), '4': Value('string'), '8': Value('string'), '14': Value('string'), '27': Value('string'), '6': Value('string'), '28': Value('string'), '24': Value('string'), '18': Value('string'), '2': Value('string'), '0': Value('string'), '25': Value('string'), '20': Value('string'), '10': Value('string'), '13': Value('string'), '17': Value('string'), '12': Value('string'), '16': Value('string'), '26': Value('string'), '3': Value('string'), '23': Value('string'), '21': Value('string'), '1': Value('string'), '11': Value('string')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/utils.py", line 147, 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 127, 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 483, 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 2840, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, in _iter_arrow
for key, pa_table in self.ex_iterable._iter_arrow():
~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 536, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, in _iter_arrow
for key, pa_table in self.generate_tables_fn(**gen_kwags):
~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
vae_id: string
vae_subfolder: string
latent_contract: struct<latent_channels: int64, scaling_factor: double, shift_factor: null>
child 0, latent_channels: int64
child 1, scaling_factor: double
child 2, shift_factor: null
scaling_factor_applied: double
shift_factor_applied: null
pad_to: int64
legacy: bool
stride: int64
source_stored_tic_stride: int64
align_decisions: bool
every_tic: bool
row_semantics: string
decode_check: struct<frames: int64, latent_shape: list<item: int64>, psnr_mean: double, psnr_min: double, latent_a (... 40 chars omitted)
child 0, frames: int64
child 1, latent_shape: list<item: int64>
child 0, item: int64
child 2, psnr_mean: double
child 3, psnr_min: double
child 4, latent_abs_mean: double
child 5, latent_abs_max: double
git: string
torch: string
args: struct<in_dir: string, out_dir: string, stride: int64, batch_size: int64, decode_threads: int64, dec (... 430 chars omitted)
child 0, in_dir: string
child 1, out_dir: string
child 2, stride: int64
child 3, batch_size: int64
child 4, decode_threads: int64
child 5, decode_workers: int64
child 6, device: string
child 7, dtype: string
child 8, shard: null
child 9, num_shards: int64
child 10, episode_ids: string
child 11, canonical: string
child 12, canonical_ids: string
child 13, canonical_only: bool
child 14, rebuild_sidecar_masks: bool
child 15, dry_run: bool
child 16, max_episodes: int64
child 17, vae_id: string
child 18, vae_subfolder: string
child 19, cache_dir: string
child 20, latent_channels: null
child 21, scaling_factor: null
child 22, shift_factor: null
child 23, decode_check: int64
child 24, legacy: bool
child 25, align_decisions: bool
child 26, every_tic: bool
18: string
20: string
6: string
15: string
24: string
17: string
23: string
5: string
8: string
21: string
2: string
14: string
10: string
22: string
26: string
16: string
3: string
7: string
27: string
0: string
12: string
9: string
19: string
13: string
4: string
28: string
25: string
1: string
11: string
to
{'19': Value('string'), '22': Value('string'), '15': Value('string'), '9': Value('string'), '5': Value('string'), '7': Value('string'), '4': Value('string'), '8': Value('string'), '14': Value('string'), '27': Value('string'), '6': Value('string'), '28': Value('string'), '24': Value('string'), '18': Value('string'), '2': Value('string'), '0': Value('string'), '25': Value('string'), '20': Value('string'), '10': Value('string'), '13': Value('string'), '17': Value('string'), '12': Value('string'), '16': Value('string'), '26': Value('string'), '3': Value('string'), '23': Value('string'), '21': Value('string'), '1': Value('string'), '11': Value('string')}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
DoomShift per-tic latents
Derived from the raw dataset RohanNaga/doom-dense-arnold: every tic of each episode encoded once through a
frozen Stable Diffusion autoencoder, stored as one .npy per episode plus a .npz sidecar with the per-tic
metadata. Released so that a training or adaptation run can start without re-encoding, and so that latents are
reproducible bit for bit: latent bytes depend on the autoencoder, its scaling, the precision and the encode
batch size, all recorded in encode_meta_*.json inside each folder. Made for the DoomShift paper
(DoomShift: Efficient Adaptation of World Models to Domain Shifts, Nagabhirava, Chirumamilla, Liu, 2026).
| Folder | Autoencoder | Channels | Latent shape per tic | Scaling | Episodes |
|---|---|---|---|---|---|
sd15/arenas |
stabilityai/sd-vae-ft-mse (SD 1.x, KL-f8) |
4 | (4, 32, 40) fp16 | z * 0.18215 | training ids 0 to 1999 of arenas (the paper's training set) |
sd15/arenas_2000_6000 |
same | 4 | training ids 2000 to 5999 (the rest of the train split) | ||
sd15/val, sd15/test |
same | 4 | arenas ids 6000 to 6099, 7000 to 7099 |
||
sd15/arenas13 |
same | 4 | the 13 unseen arenas, 24 episodes each (adaptation and scoring) | ||
sd35/... |
SD 3.5 Medium autoencoder | 16 | (16, 32, 40) fp16 | (z - 0.0609) * 1.5305 | the same folders and ids |
The SD 1.x latents serve the U-Net (SD 1.4) and PixArt-alpha backbones, the SD 3.5 latents the SD 3.5 Medium
backbone. Frames are 320x240 padded to 320x256 with mid-gray before encoding; decoders crop the padding
([:, :, :240] in pixels, the last row block in latents). Encoding is bf16 autocast at the batch size named in
encode_meta_*.json; a re-encode at another batch size differs at the last bits.
Sidecar columns (one row per tic): tic, action (Arnold's requested id), buttons (the EXECUTED 19-button
control, see the raw dataset card for the mechanism), buttons_raw_len, switch_requested_index, health,
ammo, kills, deaths, frags, pos_x, pos_y, angle, map_id, episode_id, is_decision,
chain_id. Row i holds the latent of the frame at tic i and the control applied from tic i to i+1.
Code: encode_parquet.py --every-tic in https://github.com/RohanNaga/Doom. Splits: dense_split.json and
arenas13/splits/ in the raw dataset (train 0:6000, val 6000:7000, test 7000:8000; the paper trains on
0:2000 and adapts on 16 of the 24 episodes of each unseen arena).
- Downloads last month
- 1,684