Fix tts-server for multi-lingual models (#2257)

TTS/server/server.py

@@ -115,8 +115,13 @@ synthesizer = Synthesizer(
 use_multi_speaker = hasattr(synthesizer.tts_model, "num_speakers") and (
     synthesizer.tts_model.num_speakers > 1 or synthesizer.tts_speakers_file is not None
 )
-
 speaker_manager = getattr(synthesizer.tts_model, "speaker_manager", None)
+
+use_multi_language = hasattr(synthesizer.tts_model, "num_languages") and (
+    synthesizer.tts_model.num_languages > 1 or synthesizer.tts_languages_file is not None
+)
+language_manager = getattr(synthesizer.tts_model, "language_manager", None)
+
 # TODO: set this from SpeakerManager
 use_gst = synthesizer.tts_config.get("use_gst", False)
 app = Flask(__name__)
@@ -147,7 +152,9 @@ def index():
         "index.html",
         show_details=args.show_details,
         use_multi_speaker=use_multi_speaker,
+        use_multi_language=use_multi_language,
         speaker_ids=speaker_manager.name_to_id if speaker_manager is not None else None,
+        language_ids=language_manager.name_to_id if language_manager is not None else None,
         use_gst=use_gst,
     )
 
@@ -177,11 +184,13 @@ def tts():
     with lock:
         text = request.args.get("text")
         speaker_idx = request.args.get("speaker_id", "")
+        language_idx = request.args.get("language_id", "")
         style_wav = request.args.get("style_wav", "")
         style_wav = style_wav_uri_to_dict(style_wav)
         print(" > Model input: {}".format(text))
         print(" > Speaker Idx: {}".format(speaker_idx))
-        wavs = synthesizer.tts(text, speaker_name=speaker_idx, style_wav=style_wav)
+        print(" > Language Idx: {}".format(language_idx))
+        wavs = synthesizer.tts(text, speaker_name=speaker_idx, language_name=language_idx, style_wav=style_wav)
         out = io.BytesIO()
         synthesizer.save_wav(wavs, out)
     return send_file(out, mimetype="audio/wav")

TTS/server/templates/index.html

@@ -65,7 +65,7 @@
                 </ul>
 
                 {%if use_gst%}
-                <input value='{"0": 0.1}' id="style_wav" placeholder="style wav (dict or path ot wav).." size=45
+                <input value='{"0": 0.1}' id="style_wav" placeholder="style wav (dict or path to wav).." size=45
                     type="text" name="style_wav">
                 {%endif%}
 
@@ -81,6 +81,16 @@
                 </select><br /><br />
                 {%endif%}
 
+                {%if use_multi_language%}
+                Choose a language:
+                <select id="language_id" name=language_id method="GET" action="/">
+                    {% for language_id in language_ids %}
+                    <option value="{{language_id}}" SELECTED>{{language_id}}</option>"
+                    {% endfor %}
+                </select><br /><br />
+                {%endif%}
+
+
                 {%if show_details%}
                 <button id="details-button" onclick="location.href = 'details'" name="model-details">Model
                     Details</button><br /><br />
@@ -106,11 +116,12 @@
             const text = q('#text').value
             const speaker_id = getTextValue('#speaker_id')
             const style_wav = getTextValue('#style_wav')
+            const language_id = getTextValue('#language_id')
             if (text) {
                 q('#message').textContent = 'Synthesizing...'
                 q('#speak-button').disabled = true
                 q('#audio').hidden = true
-                synthesize(text, speaker_id, style_wav)
+                synthesize(text, speaker_id, style_wav, language_id)
             }
             e.preventDefault()
             return false
@@ -121,8 +132,8 @@
                 do_tts(e)
             }
         })
-        function synthesize(text, speaker_id = "", style_wav = "") {
-            fetch(`/api/tts?text=${encodeURIComponent(text)}&speaker_id=${encodeURIComponent(speaker_id)}&style_wav=${encodeURIComponent(style_wav)}`, { cache: 'no-cache' })
+        function synthesize(text, speaker_id = "", style_wav = "", language_id = "") {
+            fetch(`/api/tts?text=${encodeURIComponent(text)}&speaker_id=${encodeURIComponent(speaker_id)}&style_wav=${encodeURIComponent(style_wav)}&language_id=${encodeURIComponent(language_id)}`, { cache: 'no-cache' })
                 .then(function (res) {
                     if (!res.ok) throw Error(res.statusText)
                     return res.blob()

TTS/utils/synthesizer.py

@@ -242,7 +242,7 @@ class Synthesizer(object):
                     "Define path for speaker.json if it is a multi-speaker model or remove defined speaker idx. "
                 )
 
-        # handle multi-lingaul
+        # handle multi-lingual
         language_id = None
         if self.tts_languages_file or (
             hasattr(self.tts_model, "language_manager") and self.tts_model.language_manager is not None