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/folder_based_builder/folder_based_builder.py", line 198, in _split_generators
for pa_metadata_table in self._read_metadata(downloaded_metadata_file, metadata_ext=metadata_ext):
~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/folder_based_builder/folder_based_builder.py", line 306, in _read_metadata
for df in csv_file_reader:
^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/pandas/io/parsers/readers.py", line 1843, in __next__
return self.get_chunk()
~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/pandas/io/parsers/readers.py", line 1985, in get_chunk
return self.read(nrows=size)
~~~~~~~~~^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/pandas/io/parsers/readers.py", line 1923, in read
) = self._engine.read( # type: ignore[attr-defined]
~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
nrows
^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/pandas/io/parsers/c_parser_wrapper.py", line 234, in read
chunks = self._reader.read_low_memory(nrows)
File "pandas/_libs/parsers.pyx", line 850, in pandas._libs.parsers.TextReader.read_low_memory
File "pandas/_libs/parsers.pyx", line 905, in pandas._libs.parsers.TextReader._read_rows
File "pandas/_libs/parsers.pyx", line 874, in pandas._libs.parsers.TextReader._tokenize_rows
File "pandas/_libs/parsers.pyx", line 891, in pandas._libs.parsers.TextReader._check_tokenize_status
File "pandas/_libs/parsers.pyx", line 2061, in pandas._libs.parsers.raise_parser_error
pandas.errors.ParserError: Error tokenizing data. C error: Expected 6 fields in line 3, saw 11
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 71, 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.
PaperParse Benchmark
NOTE: EXACT OUTPUT FILES / EXPECTED FILES ARE JUST PLACEHOLDERS RIGHT NOW ; WILL CHANGE ACCORDING TO PAPERPARSE'S OUTPUT
A curated benchmark of scientific papers designed to evaluate the robustness and accuracy of PDF parsing systems, with a focus on challenging real-world documents (e.g. complex layouts, oversized figures, dense tables, multi-column text, scanned content).
The benchmark is intended to compare different versions of PaperParse and other document processing pipelines in a reproducible manner.
Repository structure
.
├── README.md
├── metadata.csv
├── papers/
│ ├── astrophysics/
│ ├── cancer/
│ └── ...
├── expected/
│ └── <paper_id>/
│ ├── figures.json
│ ├── tables.json
│ ├── text.md
│ └── ...
├── runs/
│ └── <run_id>/
│ ├── config.yaml
│ ├── manifest.json
│ ├── outputs/
│ │ └── <paper_id>/
│ └── metrics.csv
└── scripts/
Papers
The papers/ directory contains the benchmark PDFs, grouped by scientific field.
Example:
papers/
├── astrophysics/
│ ├── astro_0001.pdf
│ ├── astro_0002.pdf
└── cancer/
├── cancer_0001.pdf
Each paper is assigned a repository-specific identifier (paper_id) that remains stable over time, regardless of its original source (arXiv, DOI, publisher website, institutional website, etc.).
Metadata
metadata.csv contains one row per benchmark paper.
Typical columns include:
id– repository identifier (e.g.astro_0001)field– scientific fieldtitle– paper titlesource– URL where the paper was obtainedexternal_id– arXiv ID, DOI, or other external identifier (if available)notes– why the paper is included in the benchmark
The metadata file acts as the index of the benchmark.
Expected outputs
expected/ contains the ground-truth annotations for each paper.
expected/
└── astro_0001/
├── figures.json
├── tables.json
├── text.md
└── ...
These files describe the expected parsing results against which implementations can be evaluated.
Benchmark runs
Every execution of a parser creates a new entry under runs/ (timestamp_paperparse_commit). Runs are immutable and are never overwritten.
runs/
├── 2026-07-27T14-52-03_paperparse_a81f39d/
├── 2026-08-01T09-11-42_paperparse_b42d8ef/
└── ...
Each run contains:
config.yaml– information about the parser, version, Git commit, runtime configuration, and environment.manifest.json– the list of papers included in the run.outputs/– raw parser outputs for each paper.metrics.csv– aggregated evaluation metrics for the run.
Example:
runs/
└── 2026-07-27T14-52-03_paperparse_a81f39d/
├── config.yaml
├── manifest.json
├── outputs/
│ ├── astro_0001/
│ └── cancer_0001/
└── metrics.csv
Recording the current Git commit in config.yaml is purely for reproducibility. It identifies the version of the parser that produced the results; the benchmark never creates Git commits automatically.
Scripts
scripts/ contains helper utilities for tasks such as:
- downloading papers
- validating the repository
- running the benchmark
- computing evaluation metrics
- reproducing experiments
- Downloads last month
- 58