The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: ValueError
Message: Invalid string class label MRE-MI@51d7ddd54b39be1d9f463d5250b4bb667c5ada7f
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 2951, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2474, in __iter__
example = _apply_feature_types_on_example(
example, self.features, token_per_repo_id=self.token_per_repo_id
)
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2391, in _apply_feature_types_on_example
encoded_example = features.encode_example(example)
File "/usr/local/lib/python3.14/site-packages/datasets/features/features.py", line 2192, in encode_example
return encode_nested_example(self, example)
File "/usr/local/lib/python3.14/site-packages/datasets/features/features.py", line 1472, in encode_nested_example
{k: encode_nested_example(schema[k], obj.get(k), level=level + 1) for k in schema}
~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/features/features.py", line 1495, in encode_nested_example
return schema.encode_example(obj) if obj is not None else None
~~~~~~~~~~~~~~~~~~~~~^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/features/features.py", line 1168, in encode_example
example_data = self.str2int(example_data)
File "/usr/local/lib/python3.14/site-packages/datasets/features/features.py", line 1105, in str2int
output = [self._strval2int(value) for value in values]
~~~~~~~~~~~~~~~~^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/features/features.py", line 1126, in _strval2int
raise ValueError(f"Invalid string class label {value}")
ValueError: Invalid string class label MRE-MI@51d7ddd54b39be1d9f463d5250b4bb667c5ada7fNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
MRE-MI: Multimodal Relation Extraction with Multiple Images
This is an unofficial convenience mirror of the MRE-MI research dataset, introduced in MRE-MI: A Multi-image Dataset for Multimodal Relation Extraction in Social Media Posts (NAACL 2025). Dataset credit belongs to the original authors.
Original source
- Official project and usage instructions: https://github.com/JinFish/MRE-MI
- Original download linked by the authors: https://pan.baidu.com/s/1R8J7FDmMdp8jRNpdmmktLg?pwd=83xh
- Original archive SHA256 recorded during acquisition:
bfa93388966a56b69cd898703f49b83142758b90e3b631ce7ba3bb84d080aa24
Contents
This repository hosts the dataset as a single archive (MRE-MI.zip, 2.49 GiB) plus SHA256SUMS.txt for the extracted files. Inside the archive, MRE-MI/ contains:
images/: original images.images_vg/: visual-grounding image crops.images_vg.pth: mapping used by the original visual-grounding pipeline.train.txt,val.txt,test.txt: 13,504 / 4,500 / 4,500 instances.ours_rel2id.json: an additional relation-to-ID mapping used in the uploader's reproduction workspace. It is provided for convenience and is not presented as an original release file.
SHA256SUMS.txt records the hashes of every supplied data file. The files were verified against the uploader's existing HPC copy before upload. The redundant original dataset.tar.gz and temporary transfer files are omitted.
Use
pip install -U huggingface_hub
export HF_ENDPOINT=https://hf-mirror.com # 国内建议：直连 LFS 可能超时
hf download hihihini/MRE-MI --local-dir ./mre-mi
cd mre-mi
unzip MRE-MI.zip # -> MRE-MI/{images,images_vg,...}
sha256sum -c SHA256SUMS.txt # 逐文件校验
Only the archive: hf download hihihini/MRE-MI --include "MRE-MI.zip" --local-dir ./mre-mi
The same archive is also mirrored on Kaggle as suhaizhou/mre-mi-multi-image-relation-extraction. Note that Kaggle automatically expands the uploaded archive and adds an outer directory, so files there live under MRE-MI/MRE-MI/; the HuggingFace copy is the plain zip, which extracts to MRE-MI/.
See the original project for the annotation format and training instructions. This dataset is MRE-MI, a multi-image benchmark; it is distinct from the single-image MNRE benchmark.
Data notes
This mirror preserves the files as acquired; it does not repair annotations or image contents. The existing data audit reports empty token slots in some annotations and unreadable images, including zero-byte files. Use appropriate image-loading fallback and validate entity offsets in your training pipeline.
Attribution and license
Please cite the original paper and project when using the data. The linked upstream repository does not state an explicit dataset license, so this mirror is marked Unknown (license: other in the card metadata). No new license or ownership claim is asserted by the uploader; original authors' and third-party rights remain unchanged.
If you are an author of MRE-MI and want this mirror changed or removed, please open an issue on this repository.
- Downloads last month
- 43