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
model: string
image_size: int64
feat_size: int64
hidden_dim: int64
mem_dim: int64
num_maskmem: int64
max_object_pointers: int64
memory_temporal_positional_encoding: list<item: list<item: double>>
child 0, item: list<item: double>
child 0, item: double
image_mean: list<item: double>
child 0, item: double
image_std: list<item: double>
child 0, item: double
no_obj_score: double
multimask_max_points: int64
variant: string
mean: list<item: double>
child 0, item: double
std: list<item: double>
child 0, item: double
family: string
to
{'family': Value('string'), 'variant': Value('string'), 'image_size': Value('int64'), 'feat_size': Value('int64'), 'hidden_dim': Value('int64'), 'mem_dim': Value('int64'), 'num_maskmem': Value('int64'), 'max_object_pointers': Value('int64'), 'multimask_max_points': Value('int64'), 'mean': List(Value('float64')), 'std': List(Value('float64'))}
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
model: string
image_size: int64
feat_size: int64
hidden_dim: int64
mem_dim: int64
num_maskmem: int64
max_object_pointers: int64
memory_temporal_positional_encoding: list<item: list<item: double>>
child 0, item: list<item: double>
child 0, item: double
image_mean: list<item: double>
child 0, item: double
image_std: list<item: double>
child 0, item: double
no_obj_score: double
multimask_max_points: int64
variant: string
mean: list<item: double>
child 0, item: double
std: list<item: double>
child 0, item: double
family: string
to
{'family': Value('string'), 'variant': Value('string'), 'image_size': Value('int64'), 'feat_size': Value('int64'), 'hidden_dim': Value('int64'), 'mem_dim': Value('int64'), 'num_maskmem': Value('int64'), 'max_object_pointers': Value('int64'), 'multimask_max_points': Value('int64'), 'mean': List(Value('float64')), 'std': List(Value('float64'))}
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.
Eight Count model weights
Mirrors of the ONNX models the Eight Count editor fetches at runtime, hosted
here so that no third-party force-push or deletion can break a deployed build.
Every file but two is byte-identical to the upstream export it came from; the
two -m320 files are that export with two constants in its graph rewritten,
which the provenance notes below spell out. Nothing in this repository is
original work: every file keeps the licence its original authors published it
under, listed below. If you are
the author of one of these models and want it taken down, open a discussion.
Files and licences
| file | mirrored from | original work | licence |
|---|---|---|---|
rtmdet-n-person.onnx |
bukuroo/RTMDet-ONNX | RTMDet, MMDetection (OpenMMLab) | Apache-2.0 |
rtmw3d-x-384.onnx |
bukuroo/RTMW3D-ONNX | RTMW3D, MMPose (OpenMMLab) | Apache-2.0 |
vitpose-s-wholebody.onnx |
JunkyByte/easy_ViTPose | ViTPose (ViTAE-Transformer) | Apache-2.0 |
depth-anything-v2-small.onnx |
onnx-community/depth-anything-v2-small | Depth Anything V2 | Apache-2.0 (small checkpoint only; the larger variants are CC-BY-NC-4.0 and are not in this repo) |
rtmw-m-256.onnx |
bukuroo/RTMW-ONNX | open-mmlab/mmpose | Apache-2.0 |
rtmw-l-256.onnx |
bukuroo/RTMW-ONNX | open-mmlab/mmpose | Apache-2.0 |
rtmw-l-384.onnx |
bukuroo/RTMW-ONNX | open-mmlab/mmpose | Apache-2.0 |
rtmdet-ins-tiny-person-m320.onnx |
tori29umai/rtmdet-ins-onnx-person-masks | open-mmlab/mmdetection | Apache-2.0 |
rtmdet-ins-s-person-m320.onnx |
tori29umai/rtmdet-ins-onnx-person-masks | open-mmlab/mmdetection | Apache-2.0 |
ppocr-v3-det-en.onnx |
SWHL/RapidOCR | PaddleOCR PP-OCRv3 English text detection (Baidu) | apache-2.0 |
ppocr-rec-en-number.onnx |
SWHL/RapidOCR | PaddleOCR en_number_mobile_v2.0 text recognition (Baidu) | apache-2.0 |
sam2.1-tiny-memory-attention-fp16.onnx |
square-zero-labs/sam2.1-tiny-video-onnx | facebookresearch/sam2 | Apache-2.0 |
sam2.1-tiny-vision-encoder-fp16.onnx |
square-zero-labs/sam2.1-tiny-video-onnx | facebookresearch/sam2 | Apache-2.0 |
sam2.1-tiny-memory-encoder-fp16.onnx |
square-zero-labs/sam2.1-tiny-video-onnx | facebookresearch/sam2 | Apache-2.0 |
sam2.1-tiny-memory-attention-light-fp16.onnx |
scripts/lab/sam2/anymem.py over our own sam2.1-tiny-memory-attention-fp16.onnx (4 mems, 8 pointers) | facebook/sam2.1-hiera-tiny via square-zero-labs/sam2.1-tiny-video-onnx | Apache-2.0 |
depth-anything-v2-metric-hypersim-small.onnx |
77ukhtar/depth-anything-v2-metric-onnx (model.onnx) | depth-anything/Depth-Anything-V2-Metric-Hypersim-Small | Apache-2.0 |
rf-detr-seg-nano.onnx |
PierreMarieCurie/rf-detr-onnx | roboflow/rf-detr | Apache-2.0 |
efficienttam-ti-2-constants.json |
yunyangx/efficient-track-anything | yformer/EfficientTAM | Apache-2.0 |
efficienttam-ti-2-vision-encoder-fp16.onnx |
yunyangx/efficient-track-anything | yformer/EfficientTAM | Apache-2.0 |
efficienttam-ti-2-memory-attention-fp16.onnx |
yunyangx/efficient-track-anything | yformer/EfficientTAM | Apache-2.0 |
efficienttam-ti-2-memory-encoder-fp16.onnx |
yunyangx/efficient-track-anything | yformer/EfficientTAM | Apache-2.0 |
efficienttam-ti-2-mask-decoder.onnx |
yunyangx/efficient-track-anything | yformer/EfficientTAM | Apache-2.0 |
efficienttam-ti-512-constants.json |
yunyangx/efficient-track-anything | yformer/EfficientTAM | Apache-2.0 |
efficienttam-ti-512-vision-encoder-fp16.onnx |
yunyangx/efficient-track-anything | yformer/EfficientTAM | Apache-2.0 |
efficienttam-ti-512-memory-attention-fp16.onnx |
yunyangx/efficient-track-anything | yformer/EfficientTAM | Apache-2.0 |
efficienttam-ti-512-memory-encoder-fp16.onnx |
yunyangx/efficient-track-anything | yformer/EfficientTAM | Apache-2.0 |
efficienttam-ti-512-mask-decoder.onnx |
yunyangx/efficient-track-anything | yformer/EfficientTAM | Apache-2.0 |
rtdetr-v2-r18vd-fp16.onnx |
onnx-community/rtdetr_v2_r18vd-ONNX | lyuwenyu/RT-DETR | Apache-2.0 |
sam2.1-tiny-constants.json |
square-zero-labs/sam2.1-tiny-video-onnx | facebookresearch/sam2 | Apache-2.0 |
sam2.1-tiny-mask-decoder.onnx |
square-zero-labs/sam2.1-tiny-video-onnx | facebookresearch/sam2 | Apache-2.0 |
sam2.1-tiny-pointer-tpos.onnx |
square-zero-labs/sam2.1-tiny-video-onnx | facebookresearch/sam2 | Apache-2.0 |
instanthmr.onnx |
momolesang/InstantHMR | facebookresearch/sam-3d-body | SAM licence |
Provenance notes
- rtmdet-ins-tiny-person-m320.onnx and rtmdet-ins-s-person-m320.onnx
— changed copies of
rtmdet-ins-tiny-person.onnxandrtmdet-ins-s-person.onnx, both of which are also in this repo, unchanged. No weight was touched and no training, fine-tuning or re-export happened: two constants in the graph were rewritten, so the mask head returns its planes at 320×320 rather than bilinearly upsampled to the 640×640 input, and at most 16 candidates rather than 100. The export is otherwise the same graph and returns the same boxes and scores. Stated here because Apache-2.0 §4(b) asks a changed file to carry notice that it changed. - All other files are unmodified copies of the linked Hugging Face repos, which are themselves ONNX exports of the linked Apache-2.0 upstream works.
What this repo is not
No training, fine-tuning or conversion happened here; the only edit ever
made to a file is the pair of graph constants noted above. This repository grants
no rights beyond what each file's own licence grants; the license: other
tag in the metadata means "several licences, per file, listed above".
- Downloads last month
- 17