Add HF dataset loading script

tts_dataset.py

@@ -0,0 +1,52 @@
+import os
+from pathlib import Path
+import datasets
+
+_CITATION = ""
+_DESCRIPTION = (
+    "Phonemized-VCTK: audio + phonemes for TTS/alignment, paired by basename."
+)
+
+class TTSDataset(datasets.GeneratorBasedBuilder):
+    VERSION = datasets.Version("1.0.0")
+
+    def _info(self):
+        return datasets.DatasetInfo(
+            description=_DESCRIPTION,
+            features=datasets.Features({
+                "speaker": datasets.Value("string"),
+                "audio": datasets.Audio(sampling_rate=16000),
+                "phoneme": datasets.Sequence(datasets.Value("string")),
+            }),
+            supervised_keys=None,
+            homepage="https://huggingface.co/datasets/srinathnr/TTS_DATASET",
+            citation=_CITATION,
+        )
+
+    def _split_generators(self, dl_manager):
+        root = Path(dl_manager.manual_dir or ".")
+        return [
+            datasets.SplitGenerator(
+                name=datasets.Split.TRAIN,
+                gen_kwargs={"root_dir": root}
+            )
+        ]
+
+    def _generate_examples(self, root_dir):
+        # Scan for all .wav and .txt files
+        wav_dir = Path(root_dir) / "wav"
+        phon_dir = Path(root_dir) / "phonemized"
+        audio_files = {p.stem: p for p in wav_dir.rglob("*.wav") if not p.name.startswith("._")}
+        phoneme_files = {p.stem: p for p in phon_dir.rglob("*.txt") if not p.name.startswith("._")}
+        # Intersect by basename
+        for key in sorted(set(audio_files) & set(phoneme_files)):
+            audio_path = audio_files[key]
+            phoneme_path = phoneme_files[key]
+            speaker = audio_path.parts[-2]  # e.g. "p225"
+            with open(phoneme_path) as f:
+                phonemes = f.read().split()
+            yield key, {
+                "speaker": speaker,
+                "audio": str(audio_path),
+                "phoneme": phonemes,
+            }