initial commit

.gitattributes

@@ -53,3 +53,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.jpg filter=lfs diff=lfs merge=lfs -text
 *.jpeg filter=lfs diff=lfs merge=lfs -text
 *.webp filter=lfs diff=lfs merge=lfs -text
+*.jsonl filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1 @@
+raw/*

2wikimultihopqa.py

@@ -0,0 +1,141 @@
+# coding=utf-8
+# Copyright 2020 The TensorFlow Datasets Authors and the HuggingFace Datasets Authors.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""
+Mirror of https://github.com/Alab-NII/2wikimultihop
+"""
+
+import json
+import datasets
+import pandas as pd
+
+_DESCRIPTION = """\
+Mirror of https://github.com/Alab-NII/2wikimultihop
+"""
+
+CITATION = """
+@inproceedings{xanh2020_2wikimultihop,
+    title = "Constructing A Multi-hop {QA} Dataset for Comprehensive Evaluation of Reasoning Steps",
+    author = "Ho, Xanh  and
+      Duong Nguyen, Anh-Khoa  and
+      Sugawara, Saku  and
+      Aizawa, Akiko",
+    booktitle = "Proceedings of the 28th International Conference on Computational Linguistics",
+    month = dec,
+    year = "2020",
+    address = "Barcelona, Spain (Online)",
+    publisher = "International Committee on Computational Linguistics",
+    url = "https://www.aclweb.org/anthology/2020.coling-main.580",
+    pages = "6609--6625",
+}
+"""
+
+DEV_URL = "https://huggingface.co/datasets/somebody-had-to-do-it/2WikiMultihopQA/resolve/main/dev.parquet?download=true"
+TRAIN_URL = "https://huggingface.co/datasets/somebody-had-to-do-it/2WikiMultihopQA/resolve/main/train.parquet?download=true"
+TEST_URL = "https://huggingface.co/datasets/somebody-had-to-do-it/2WikiMultihopQA/resolve/main/test.parquet?download=true"
+
+
+class Dataset2WikiMultihopQa(datasets.GeneratorBasedBuilder):
+    """Mirror of https://github.com/Alab-NII/2wikimultihop"""
+
+    VERSION = datasets.Version("1.0.0")
+
+    def _info(self):
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=datasets.Features(
+                {
+                    "_id": datasets.Value("string"),
+                    "type": datasets.Value("string"),
+                    "question": datasets.Value("string"),
+                    "context": datasets.features.Sequence(
+                        {
+                            "title": datasets.Value("string"),
+                            "content": datasets.features.Sequence(
+                                datasets.Value("string")
+                            ),
+                        }
+                    ),
+                    "supporting_facts": datasets.features.Sequence(
+                        {
+                            "title": datasets.Value("string"),
+                            "sent_id": datasets.Value("int32"),
+                        }
+                    ),
+                    "evidences": datasets.features.Sequence(
+                        {
+                            "fact": datasets.Value("string"),
+                            "relation": datasets.Value("string"),
+                            "entity": datasets.Value("string"),
+                        }
+                    ),
+                    "answer": datasets.Value("string"),
+                }
+            ),
+            supervised_keys=None,
+            homepage="https://github.com/Alab-NII/2wikimultihop",
+            citation=CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        """Returns SplitGenerators."""
+
+        train_uri = dl_manager.download(TRAIN_URL)
+        dev_uri = dl_manager.download(DEV_URL)
+        test_uri = dl_manager.download(TEST_URL)
+
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN, gen_kwargs={"filepath": train_uri}
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split("dev"),
+                gen_kwargs={"filepath": dev_uri, "split": "dev"},
+            ),
+            datasets.SplitGenerator(
+                name=datasets.Split.TEST,
+                gen_kwargs={"filepath": test_uri},
+            ),
+        ]
+
+    def _generate_examples(self, filepath, *args, **kwargs):
+        """Yields examples from a Parquet file."""
+        # Read the Parquet file into a DataFrame
+        df = pd.read_parquet(filepath)
+
+        # Iterate over each row in the DataFrame
+        for idx, row in df.iterrows():
+            yield idx, {
+                "_id": row["_id"],
+                "type": row["type"],
+                "question": row["question"],
+                "context": [
+                    {"title": item[0], "content": item[1]}
+                    for item in json.loads(row["context"])
+                ],
+                "supporting_facts": [
+                    {"title": fact[0], "sent_id": fact[1]}
+                    for fact in json.loads(row["supporting_facts"])
+                ],
+                "evidences": [
+                    {
+                        "fact": evidence[0],
+                        "relation": evidence[1],
+                        "entity": evidence[2],
+                    }
+                    for evidence in json.loads(row["evidences"])
+                ],
+                "answer": row["answer"],
+            }

README.md

@@ -1,3 +1,13 @@
 ---
 license: apache-2.0
+task_categories:
+- question-answering
+language:
+- en
+size_categories:
+- 100K<n<1M
 ---
+
+# 2WikiMultihopQA: A Dataset for Comprehensive Evaluation of Reasoning Steps
+
+Unofficial mirror of <https://github.com/Alab-NII/2wikimultihop>

convert_to_jsonl.py

@@ -0,0 +1,11 @@
+from pathlib import Path
+import pandas as pd
+
+file_names = ["./raw/dev.json", "./raw/test.json", "./raw/train.json"]
+for file_name in file_names:
+    # Read the JSON file into a DataFrame
+    df = pd.read_json(file_name)
+
+    # Convert the DataFrame to JSON Lines and write to a file_name
+    jsonl_path = "./raw/" + Path(file_name).stem + ".jsonl"
+    df.to_json(jsonl_path, orient="records", lines=True)

convert_to_parquet.py

@@ -0,0 +1,20 @@
+from pathlib import Path
+import pandas as pd
+import json
+
+file_names = ["./raw/dev.jsonl", "./raw/test.jsonl", "./raw/train.jsonl"]
+for file_name in file_names:
+    df = pd.read_json(file_name, lines=True)
+
+    complex_columns = [
+        "context",
+        "supporting_facts",
+        "evidences",
+    ]
+    for col in complex_columns:
+        df[col] = df[col].apply(json.dumps)
+
+    parquet_file_path = Path(file_name).stem + ".parquet"
+    df.to_parquet(parquet_file_path)
+
+    print(f"Converted {file_name} to {parquet_file_path}")

dev.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0d8b60b9026b728fb07ad74c5252a0f188f6942e8ba5c02df4dfa369502ea8d
+size 30056098

test.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1dc69c656c03837f12bb231e1adbd8a2dcdaaf632143a1ab9ebb98bd6452ede5
+size 28475341

train.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0f464f1604edf40ad990b605c61c03d56911c106d09ad32becf6afcf3201d52
+size 343395646