Add dataset loading script using transcripts.txt

my_kinyarwanda_dataset.py

@@ -0,0 +1,35 @@
+import os
+import datasets
+
+_DESCRIPTION = "A Kinyarwanda speech dataset of 102 spoken words with corresponding audio files and text labels."
+_HOMEPAGE = "https://huggingface.co/datasets/benax-rw/my_kinyarwanda_dataset"
+_CITATION = ""
+
+class MyKinyarwandaDataset(datasets.GeneratorBasedBuilder):
+    def _info(self):
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=datasets.Features({
+                "audio": datasets.Audio(sampling_rate=16_000),
+                "text": datasets.Value("string"),
+            }),
+            supervised_keys=None,
+            homepage=_HOMEPAGE,
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={"audio_dir": os.path.join(self.config.data_dir, "audio"),
+                            "transcript_file": os.path.join(self.config.data_dir, "transcripts.txt")},
+            )
+        ]
+
+    def _generate_examples(self, audio_dir, transcript_file):
+        with open(transcript_file, "r", encoding="utf-8") as f:
+            for idx, line in enumerate(f):
+                fname, text = line.strip().split("\t")
+                audio_path = os.path.join(audio_dir, fname)
+                yield idx, {"audio": audio_path, "text": text}