The dataset viewer is not available for this subset.
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.
PartTrellis predictions — every method's raw inference outputs
The per-object inference results behind the paper's tables, with per-face part labels, so any part-level metric (CD/F1, detection F1, PQ/SQ/RQ, VI, coverage, IoU variants) can be recomputed without rerunning any model.
Naming inside every tar: <oid>_<stack>.obj + <oid>_<stack>_faceids.npy
(one integer label per face; volume-labelled stacks carry 2 labels, part-labelled
stacks carry per-part ids). Predictions are in each model's own output frame:
evaluation aligns them to GT with the 48 signed-axis-permutation oracle (code in
the PartTrellis repo: eval_metrics_axis_oracle.py).
test986/ — the 1,000-object held-out test split (986 evaluable)
| stack prefix | method |
|---|---|
meshes_slatorig100k_fit |
PartTrellis (ours, released pipeline) |
xpart_meshes_fit |
TRELLIS.2@1024 + X-Part |
xpart_meshes_hy3d |
Hunyuan3D-2.1 + X-Part |
partpacker_meshes_fit |
PartPacker |
autopartgen_meshes_fit |
AutoPartGen |
omnipart_meshes_fit |
OmniPart |
Three tars per stack, 400 objects each, alphabetical by oid.
wild200/ — the 200-object unseen-source set (Sketchfab/GitHub)
qual200_defstack.tar ours, qual200_omni.tar OmniPart,
qual200_xpart.tar X-Part (includes trellis2_mesh/ — the TRELLIS.2@1024
whole shapes X-Part segmented — and raw xpart_out/).
train6/ — the six training-corpus figure objects
tr6_preds.tar: ours (tr6_defstack/), X-Part (tr6_xpart/), OmniPart
(tr6_omni/).
GT meshes/labels and conditioning images live in PartTrellis-corpus and PartTrellis-wild-test; checkpoints and code in PartTrellis.
⚠️ 8 tars here are truncated — use the replacement repo
This account's storage quota was hit mid-upload, truncating 8 tars
(meshes_slatorig100k_fit_000/001, xpart_meshes_fit_000/001,
xpart_meshes_hy3d_000/001, partpacker_meshes_fit_001,
wild200/qual200_xpart). Complete size-verified copies live, split into 200-object _pa…_pd parts, in
AuroraRyan2/PartTrellis-predictions.
Every other file in this repo is complete.
- Downloads last month
- 53