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
study_id: string
dataset: string
generated_report: string
prompt: string
bleu4: double
rougeL_p: double
bleu1: double
rougeL_r: double
rougeL_f: double
rouge1_r: double
cabs_precision: double
rouge1_p: double
rouge1_f: double
cabs_recall: double
n_studies: int64
cabs_f1: double
to
{'bleu1': Value('float64'), 'bleu4': Value('float64'), 'rouge1_f': Value('float64'), 'rouge1_p': Value('float64'), 'rouge1_r': Value('float64'), 'rougeL_f': Value('float64'), 'rougeL_p': Value('float64'), 'rougeL_r': Value('float64'), 'cabs_precision': Value('float64'), 'cabs_recall': Value('float64'), 'cabs_f1': Value('float64'), 'n_studies': Value('int64')}
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 478, 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 2818, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2355, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2380, 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 2369, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2297, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
study_id: string
dataset: string
generated_report: string
prompt: string
bleu4: double
rougeL_p: double
bleu1: double
rougeL_r: double
rougeL_f: double
rouge1_r: double
cabs_precision: double
rouge1_p: double
rouge1_f: double
cabs_recall: double
n_studies: int64
cabs_f1: double
to
{'bleu1': Value('float64'), 'bleu4': Value('float64'), 'rouge1_f': Value('float64'), 'rouge1_p': Value('float64'), 'rouge1_r': Value('float64'), 'rougeL_f': Value('float64'), 'rougeL_p': Value('float64'), 'rougeL_r': Value('float64'), 'cabs_precision': Value('float64'), 'cabs_recall': Value('float64'), 'cabs_f1': Value('float64'), 'n_studies': Value('int64')}
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.
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Reproduction bundle: TIF-GRPO (ICML 2026 #11844)
This bundle contains the scripts, data, and results for the open reproduction of "Regulating Anatomy-Aware Rewards via Trajectory-Integral Feedback for Volumetric Computed Tomography Analysis".
Contents
reward/— cloned TIF reward code from ZJU4HealthCare/TIF-GRPO.medical_report_abnormality.py— the anatomy-aware TIF reward.example.json— verl data schema example.
verify_reward_and_concordance.py— local verification script.generate_figures.py— Plotly figure generation.verification_results.json— JSON output of the verification script.ablation_results.csv— TIF reward ablation results.concordance_results.csv— Synthetic concordance results.figure_ablation.html,figure_concordance.html— Interactive figures.poster.html,poster_preview.pdf,poster_preview.png,poster_embed.html— Reproduction poster.amos_mm_20_metrics.json,amos_mm_20_per_study.csv,amos_mm_20_outputs.jsonl— AMOS-MM proxy inference outputs and metrics (20 studies, Qwen2-VL-2B).claim1.md…claim6.md,executive_summary.md— Logbook cell sources.
Published bundle dataset: sisaacson/tif-grpo-repro-bundle
How to rerun
# 1. Activate the project virtual environment (Python 3.12)
source .venv/bin/activate
# 2. Run the verification script
python verify_reward_and_concordance.py
# 3. Regenerate figures
python generate_figures.py
# 4. (Optional) Re-render the poster with posterly
python /path/to/posterly/tools/poster_check.py preflight poster.html
python /path/to/posterly/tools/poster_check.py measure poster.html
python /path/to/posterly/tools/poster_check.py polish poster.html --strict
python /path/to/posterly/tools/render_preview.py poster.html
Scope and limitations
This reproduction verifies only the released reward implementation and the methodology on synthetic data. The end-to-end report-generation results (Claims 1–3 and 6) are not reproduced because the trained checkpoint, training scripts, and gated primary datasets are not publicly available.
- Downloads last month
- 41