Closes #366 | Implement dataloader for Kheng.info Speech

seacrowd/sea_datasets/kheng_info/kheng_info.py

@@ -0,0 +1,113 @@
+# coding=utf-8
+
+from pathlib import Path
+from typing import Dict, List, Tuple
+
+import datasets
+import pandas as pd
+
+from seacrowd.utils import schemas
+from seacrowd.utils.configs import SEACrowdConfig
+from seacrowd.utils.constants import Licenses, Tasks
+
+# no bibtex citation
+_CITATION = ""
+
+_DATASETNAME = "kheng_info"
+
+_DESCRIPTION = """\
+The Kheng.info Speech dataset was derived from recordings of Khmer words on the Khmer dictionary website kheng.info.
+The recordings were recorded by a native Khmer speaker.
+The recordings are short, generally ranging between 1 to 2 seconds only.
+"""
+
+_HOMEPAGE = "https://huggingface.co/datasets/seanghay/khmer_kheng_info_speech"
+
+_LANGUAGES = ["khm"]
+
+_LICENSE = Licenses.UNKNOWN.value
+
+_LOCAL = False
+
+_URLS = {
+    _DATASETNAME: "https://huggingface.co/datasets/seanghay/khmer_kheng_info_speech/resolve/main/data/train-00000-of-00001-4e7ad082a34164d1.parquet",
+}
+
+_SUPPORTED_TASKS = [Tasks.SPEECH_RECOGNITION]
+
+_SOURCE_VERSION = "1.0.0"
+
+_SEACROWD_VERSION = "1.0.0"
+
+
+class KhengInfoDataset(datasets.GeneratorBasedBuilder):
+    """This is the Kheng.info Speech dataset, which wasderived from recordings on the Khmer dictionary website kheng.info"""
+
+    SOURCE_VERSION = datasets.Version(_SOURCE_VERSION)
+    SEACROWD_VERSION = datasets.Version(_SEACROWD_VERSION)
+
+    BUILDER_CONFIGS = [
+        SEACrowdConfig(
+            name=f"{_DATASETNAME}_source",
+            version=SOURCE_VERSION,
+            description=f"{_DATASETNAME} source schema",
+            schema="source",
+            subset_id=f"{_DATASETNAME}",
+        ),
+        SEACrowdConfig(
+            name=f"{_DATASETNAME}_seacrowd_sptext",
+            version=SEACROWD_VERSION,
+            description=f"{_DATASETNAME} SEACrowd schema",
+            schema="seacrowd_sptext",
+            subset_id=f"{_DATASETNAME}",
+        ),
+    ]
+
+    DEFAULT_CONFIG_NAME = f"{_DATASETNAME}_source"
+
+    def _info(self) -> datasets.DatasetInfo:
+        if self.config.schema == "source":
+            features = datasets.Features({"word": datasets.Value("string"), "duration_ms": datasets.Value("int64"), "audio": datasets.Audio(sampling_rate=16_000)})
+
+        elif self.config.schema == "seacrowd_sptext":
+            features = schemas.speech_text_features
+
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=features,
+            homepage=_HOMEPAGE,
+            license=_LICENSE,
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager: datasets.DownloadManager) -> List[datasets.SplitGenerator]:
+        urls = _URLS[_DATASETNAME]
+        data_dir = dl_manager.download_and_extract(urls)
+
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={
+                    "filepath": data_dir,
+                },
+            )
+        ]
+
+    def _generate_examples(self, filepath: Path) -> Tuple[int, Dict]:
+        df = pd.read_parquet(filepath, engine="pyarrow")
+        if self.config.schema == "source":
+            for _id, row in df.iterrows():
+                yield _id, {"word": row["word"], "duration_ms": row["duration_ms"], "audio": row["audio"]}
+        elif self.config.schema == "seacrowd_sptext":
+            for _id, row in df.iterrows():
+                yield _id, {
+                    "id": _id,
+                    "path": row["audio"],
+                    "audio": row["audio"],
+                    "text": row["word"],
+                    "speaker_id": None,
+                    "metadata": {
+                        "speaker_age": None,
+                        "speaker_gender": None,
+                    },
+                }