update dataset

libris2s/libris2s.py

@@ -0,0 +1,79 @@
+"""LibriS2S dataset."""
+
+import csv
+import os
+from typing import Dict, List, Tuple
+
+import datasets
+
+
+_CITATION = """@inproceedings{jeuris-niehues-2022-libris2s,
+   title = "{L}ibri{S}2{S}: A {G}erman-{E}nglish Speech-to-Speech Translation Corpus",
+   author = "Jeuris, Pedro and Niehues, Jan",
+   booktitle = "Proceedings of the Thirteenth Language Resources and Evaluation Conference",
+   year = "2022",
+   url = "https://aclanthology.org/2022.lrec-1.98",
+   pages = "928--935"
+}"""
+
+_DESCRIPTION = """LibriS2S: A German-English Speech-to-Speech Translation Corpus"""
+
+_HOMEPAGE = "https://huggingface.co/datasets/PedroDKE/LibriS2S"
+
+_LICENSE = "cc-by-nc-sa-4.0"
+
+
+class LibriS2S(datasets.GeneratorBasedBuilder):
+    """LibriS2S dataset."""
+
+    VERSION = datasets.Version("1.0.0")
+
+    def _info(self) -> datasets.DatasetInfo:
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=datasets.Features(
+                {
+                    "book_id": datasets.Value("int64"),
+                    "DE_audio": datasets.Audio(),
+                    "EN_audio": datasets.Audio(),
+                    "score": datasets.Value("float32"),
+                    "DE_transcript": datasets.Value("string"),
+                    "EN_transcript": datasets.Value("string"),
+                }
+            ),
+            homepage=_HOMEPAGE,
+            license=_LICENSE,
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
+        """Returns SplitGenerators."""
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "split": "train",
+                    "data_dir": ".",
+                },
+            ),
+        ]
+
+    def _generate_examples(self, split: str, data_dir: str) -> Tuple[int, Dict]:
+        """Yields examples."""
+        csv_path = os.path.join(data_dir, "alignments", "all_de_en_alligned_cleaned.csv")
+        
+        with open(csv_path, encoding="utf-8") as f:
+            reader = csv.DictReader(f)
+            for idx, row in enumerate(reader):
+                # Handle paths for both DE and EN audio
+                de_audio_path = os.path.join(data_dir, row["DE_audio"])
+                en_audio_path = os.path.join(data_dir, row["EN_audio"])
+                
+                yield idx, {
+                    "book_id": int(row["book_id"]),
+                    "DE_audio": de_audio_path,
+                    "EN_audio": en_audio_path,
+                    "score": float(row["score"]),
+                    "DE_transcript": row["DE_transcript"],
+                    "EN_transcript": row["EN_transcript"],
+                }