Create convert_to_ljspeech.py

convert_to_ljspeech.py

@@ -0,0 +1,19 @@
+import json, csv, os
+
+root = "/tmp/dataset/voxpopuli_es_500"
+
+with open(os.path.join(root, "metadata.json"), "r", encoding="utf-8") as f:
+    data = json.load(f)
+
+with open(os.path.join(root, "metadata.csv"), "w", encoding="utf-8") as f:
+    writer = csv.writer(f, delimiter="|", quoting=csv.QUOTE_NONE, escapechar="\\")
+    for entry in data:
+        # Quitar prefijo "voxpopuli_es_500/" si está presente
+        path = entry["audio_filepath"]
+        if path.startswith("voxpopuli_es_500/"):
+            path = path.replace("voxpopuli_es_500/", "", 1)
+
+        text = entry["text"].replace("\n", " ").strip()
+        speaker = entry["speaker"]
+
+        writer.writerow([path, text, speaker])