commit files to HF hub

README.md

@@ -0,0 +1,56 @@
+---
+pretty_name: '`msmarco-passage/trec-dl-hard/fold3`'
+viewer: false
+source_datasets: ['irds/msmarco-passage']
+task_categories:
+- text-retrieval
+---
+
+# Dataset Card for `msmarco-passage/trec-dl-hard/fold3`
+
+The `msmarco-passage/trec-dl-hard/fold3` dataset, provided by the [ir-datasets](https://ir-datasets.com/) package.
+For more information about the dataset, see the [documentation](https://ir-datasets.com/msmarco-passage#msmarco-passage/trec-dl-hard/fold3).
+
+# Data
+
+This dataset provides:
+ - `queries` (i.e., topics); count=10
+ - `qrels`: (relevance assessments); count=444
+
+ - For `docs`, use [`irds/msmarco-passage`](https://huggingface.co/datasets/irds/msmarco-passage)
+
+## Usage
+
+```python
+from datasets import load_dataset
+
+queries = load_dataset('irds/msmarco-passage_trec-dl-hard_fold3', 'queries')
+for record in queries:
+    record # {'query_id': ..., 'text': ...}
+
+qrels = load_dataset('irds/msmarco-passage_trec-dl-hard_fold3', 'qrels')
+for record in qrels:
+    record # {'query_id': ..., 'doc_id': ..., 'relevance': ...}
+
+```
+
+Note that calling `load_dataset` will download the dataset (or provide access instructions when it's not public) and make a copy of the
+data in 🤗 Dataset format.
+
+## Citation Information
+
+```
+@article{Mackie2021DlHard,
+  title={How Deep is your Learning: the DL-HARD Annotated Deep Learning Dataset},
+  author={Iain Mackie and Jeffrey Dalton and Andrew Yates},
+  journal={ArXiv},
+  year={2021},
+  volume={abs/2105.07975}
+}
+@inproceedings{Bajaj2016Msmarco,
+  title={MS MARCO: A Human Generated MAchine Reading COmprehension Dataset},
+  author={Payal Bajaj, Daniel Campos, Nick Craswell, Li Deng, Jianfeng Gao, Xiaodong Liu, Rangan Majumder, Andrew McNamara, Bhaskar Mitra, Tri Nguyen, Mir Rosenberg, Xia Song, Alina Stoica, Saurabh Tiwary, Tong Wang},
+  booktitle={InCoCo@NIPS},
+  year={2016}
+}
+```

msmarco-passage_trec-dl-hard_fold3.py

@@ -0,0 +1,43 @@
+
+"""
+""" # TODO
+try:
+    import ir_datasets
+except ImportError as e:
+    raise ImportError('ir-datasets package missing; `pip install ir-datasets`')
+import datasets
+
+IRDS_ID = 'msmarco-passage/trec-dl-hard/fold3'
+IRDS_ENTITY_TYPES = {'queries': {'query_id': 'string', 'text': 'string'}, 'qrels': {'query_id': 'string', 'doc_id': 'string', 'relevance': 'int64'}}
+
+_CITATION = '@article{Mackie2021DlHard,\n  title={How Deep is your Learning: the DL-HARD Annotated Deep Learning Dataset},\n  author={Iain Mackie and Jeffrey Dalton and Andrew Yates},\n  journal={ArXiv},\n  year={2021},\n  volume={abs/2105.07975}\n}\n@inproceedings{Bajaj2016Msmarco,\n  title={MS MARCO: A Human Generated MAchine Reading COmprehension Dataset},\n  author={Payal Bajaj, Daniel Campos, Nick Craswell, Li Deng, Jianfeng Gao, Xiaodong Liu, Rangan Majumder, Andrew McNamara, Bhaskar Mitra, Tri Nguyen, Mir Rosenberg, Xia Song, Alina Stoica, Saurabh Tiwary, Tong Wang},\n  booktitle={InCoCo@NIPS},\n  year={2016}\n}'
+
+_DESCRIPTION = "" # TODO
+
+class msmarco_passage_trec_dl_hard_fold3(datasets.GeneratorBasedBuilder):
+    BUILDER_CONFIGS = [datasets.BuilderConfig(name=e) for e in IRDS_ENTITY_TYPES]
+
+    def _info(self):
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=datasets.Features({k: datasets.Value(v) for k, v in IRDS_ENTITY_TYPES[self.config.name].items()}),
+            homepage=f"https://ir-datasets.com/msmarco-passage#msmarco-passage/trec-dl-hard/fold3",
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        return [datasets.SplitGenerator(name=self.config.name)]
+
+    def _generate_examples(self):
+        dataset = ir_datasets.load(IRDS_ID)
+        for i, item in enumerate(getattr(dataset, self.config.name)):
+            key = i
+            if self.config.name == 'docs':
+                key = item.doc_id
+            elif self.config.name == 'queries':
+                key = item.query_id
+            yield key, item._asdict()
+
+    def as_dataset(self, split=None, *args, **kwargs):
+        split = self.config.name # always return split corresponding with this config to avid returning a redundant DatasetDict layer
+        return super().as_dataset(split, *args, **kwargs)