Dataset Viewer
The dataset viewer is not available for this subset.
Cannot get the split names for the config 'default' of the dataset.
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/parquet/parquet.py", line 127, in _split_generators
                  self.info.features = datasets.Features.from_arrow_schema(pq.read_schema(f))
                                                                           ~~~~~~~~~~~~~~^^^
                File "/usr/local/lib/python3.14/site-packages/pyarrow/parquet/core.py", line 2459, in read_schema
                  file = ParquetFile(
                      where, memory_map=memory_map,
                      decryption_properties=decryption_properties)
                File "/usr/local/lib/python3.14/site-packages/pyarrow/parquet/core.py", line 328, in __init__
                  self.reader.open(
                  ~~~~~~~~~~~~~~~~^
                      source, use_memory_map=memory_map,
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  ...<8 lines>...
                      arrow_extensions_enabled=arrow_extensions_enabled,
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "pyarrow/_parquet.pyx", line 1656, in pyarrow._parquet.ParquetReader.open
                File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
                  raise convert_status(status)
              pyarrow.lib.ArrowInvalid: Parquet magic bytes not found in footer. Either the file is corrupted or this is not a parquet file.
              
              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.

Poly(script) Data

Data artifacts for the Poly(script) virtual polymerization pipeline. The pipeline builds a monomer dataset, decomposes and recombines it with BRICS, classifies the monomers, and runs virtual polymerization. Optional stages evaluate the extender mechanism and benchmark VPF resources.

The exact commands, inputs, and outputs for each step are in Workflow.md. The accompanying research paper describes the chemistry, benchmark design, and analysis.

Files

Step File Size Contents
1.1 1.1.0_merged_smiles_set.csv 1.97 MB Combined monomer SMILES (78,247 rows)
1.2 1.2.0_decomp.csv 572 KB Unique BRICS fragments (26,602 rows)
1.3 1.3.0_all_rxn_data_BRICS.parquet 498 MB BRICS compose reactions and products (40.4M rows)
1.5 1.5.0_cls_split_batch_5K.zip 367 MB Monomer class splits in 5,000-row batches
2.1 2.1.0_ps-run-500_10b_val.zip 14.6 GB Virtual polymerization output
3.2 3.2.0_extender_benchmarks.zip 1.17 MB Extender benchmark results and traces
4.1 4.1.0_VPF_benchmarking.zip 74.2 MB VPF resource benchmark bundle

Step 1.4 produces the classified table that Step 1.5 consumes; the workflow links the split archive from Step 1.5.

Pipeline

1.1 SMILES -> 1.2 BRICS decompose -> 1.3 BRICS compose
    -> 1.4 classify -> 1.5 class split -> 2.1 virtual polymerization
    -> 3.x extender evaluation -> 4.x resource benchmarking

The 1.5 class directories are not mutually exclusive. A monomer can occur in more than one class because the classifier tests each functional-group type independently. The unknown rows are excluded from the split archive.

Download

Large files use Git LFS and Xet. Use huggingface_hub to download them:

from huggingface_hub import hf_hub_download

path = hf_hub_download(
    repo_id="debasishmohanty/poly-script",
    filename="1.5.0_cls_split_batch_5K.zip",
    repo_type="dataset",
)

You can also clone the dataset repository:

git lfs install
git clone https://huggingface.co/datasets/debasishmohanty/poly-script

The 2.1 archive is 14.6 GB. Download only the files that you need.

Reproduction

The scripts, step order, and released outputs are documented in Workflow.md. The same workflow is reproduced in the PolyScript source repository under runners/ and tests/.

License

The Poly(script) source repository uses the MIT license. The 4.1 archive includes third-party code and data from SMiPoly and OpenMacromolecularGenome; those parts keep their original licenses. Check the source datasets before redistribution.

Citation

@misc{polyscript2026,
  author = {Debasish Mohanty},
  title = {poly(script): The Programmable Virtual Polymerization
  Framework},
  year = {2026},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/debasish-g1t/polyscript}}
}
Downloads last month
46