Upload app.py

app.py

@@ -22,28 +22,28 @@ TTS_VOICES = [
 ]  
 
 TTS_LANGUAGES = [
-  "x-lb",
-  "x-de",
-  "fr-fr",
-  "en",
-  "pt-br"
+  "English",
+  "Français",
+  "Português",
+  "Deutsch",
+  "Lëtzebuergesch"
 ]
 
 
 my_examples = [
-  [lb_text,"x-lb","Luc"],
-  [de_text,"x-de","Judith"],
-  [fr_text,"fr-fr","Kerstin"],
-  [en_text,"en","Ed"],
-  [pt_text,"pt-br","Linda"],
+  [lb_text, "Luc", "Lëtzebuergesch"],
+  [de_text, "Judith", "Deutsch"],
+  [fr_text, "Kerstin", "Français"],
+  [en_text, "Ed", "English"],
+  [pt_text, "Linda", "Português"]
 ]
 
 my_article = "<h3>User guide</h3><p>1. Press the Submit button to generate a speech file with the default values. 2. Change the default values by clicking an example row. 3. Select a language and a voice and enter your own text. Have fun!</p><p>Go to <a href='https://www.web3.lu/category/audio-technologies/'>Internet with a Brain</a> to read some technical infos.</p>"
 
 my_inputs = [
   gr.inputs.Textbox(lines=5, label="Input Text", default=lb_text),
-  gr.inputs.Radio(label="Select a language", choices = TTS_LANGUAGES, default = "x-lb"),
-  gr.inputs.Radio(label="Select a voice", choices = TTS_VOICES, default = "Judith"),
+  gr.inputs.Radio(label="Speaker", choices = TTS_VOICES, default = "Judith"),
+  gr.inputs.Radio(label="Language", choices = TTS_LANGUAGES, default = "Lëtzebuergesch"),
 ]
 
 my_outputs = gr.outputs.Audio(type="file", label="Output Audio")
@@ -56,6 +56,7 @@ def tts(text: str, speaker_idx: str, language_idx: str):
     speakers_path = hf_hub_download(repo_id=REPO_ID, filename="speakers.pth")
     print(speakers_path)
     languages_path = hf_hub_download(repo_id=REPO_ID, filename="language_ids.json")
+    # languages_path = "/home/mbarnig/GRADIO/gradio/language_ids.json"
     print(languages_path)
     speaker_encoder_model_path = hf_hub_download(repo_id=REPO_ID, filename="model_se.pth")
     print(speaker_encoder_model_path)
@@ -76,7 +77,10 @@ def tts(text: str, speaker_idx: str, language_idx: str):
     )
 
     # create audio file
+    print("speaker_idx : " + speaker_idx)
+    print("language_idx : " + language_idx)
     wavs = synthesizer.tts(text, speaker_idx, language_idx)
+    # wavs = synthesizer.tts(text, "Ed", "x-lb")
     with tempfile.NamedTemporaryFile(suffix = ".wav", delete = False) as fp:
         synthesizer.save_wav(wavs, fp)                      
     return fp.name