Added all langs

app.py

@@ -7,6 +7,23 @@ models = [
     "https://github.com/AI4Bharat/Indic-TTS/releases/download/v1-checkpoints-release/bn.zip"
 ]
 
+languages = {
+    "Assamese": "as",
+    "Bengali": "bn",
+    "Bodo": "brx",
+    "Gujarati": "gu",
+    "Hindi" : "hi",
+    "Kannada": "kn",
+    "Manipuri": "mni",
+    "Malayalam": "ml",
+    "Marathi": "mr",
+    "Odia": "or",
+    "Punjabi": "pa",
+    "Rajasthani": "raj",
+    "Tamil": "ta",
+    "Telugu": "te"
+}
+
 for model in models:
     os.system(f"wget {model}")
     os.system(f"unzip {model.split('/')[-1]}")
@@ -23,26 +40,27 @@ for model in models:
 
 
 def convert(text, language, out = "out.wav"):
-    if language == "Hindi":
-        m = "hi"
-    else:
-        m = "bn"
-    
-    os.system(f'python3 -m TTS.bin.synthesize --text "{text}" --model_path {m}/fastpitch/best_model.pth --config_path {m}/fastpitch/config.json --vocoder_path {m}/hifigan/best_model.pth --vocoder_config_path {m}/hifigan/config.json --speaker_idx "male" --out_path {out}')
+    m = languages[language]
+
+    os.system(f'python3 -m TTS.bin.synthesize --text "{text}" --model_path {m}/fastpitch/best_model.pth --config_path {m}/fastpitch/config.json --vocoder_path {m}/hifigan/best_model.pth --vocoder_config_path {m}/hifigan/config.json --speaker_idx "male" --out_path male_{out}')
+    os.system(f'python3 -m TTS.bin.synthesize --text "{text}" --model_path {m}/fastpitch/best_model.pth --config_path {m}/fastpitch/config.json --vocoder_path {m}/hifigan/best_model.pth --vocoder_config_path {m}/hifigan/config.json --speaker_idx "male" --out_path female_{out}')
 
-    return out
+    return f"male_{out}", f"female_{out}"
 
 text = gr.Textbox(value = "यह कल का दिन अद्भुत था क्योंकि हम संगीत कार्यक्रम से वापस आ गए हैं।", 
                   placeholder = "Enter a text to synthesize",
                   label = "Text")
 
-language = gr.Dropdown(choices = ["Hindi", "Bangla"], 
+language = gr.Dropdown(choices = sorted(languages.keys()), 
                        value = "Hindi", 
                        type = "value", 
                        label = "Language")
 
 inputs = [text, language]
-outputs =  gr.outputs.Audio(label = "Output Audio", type = 'filepath')
+outputs =  [
+    gr.outputs.Audio(label = "Male Speaker", type = 'filepath'),
+    gr.outputs.Audio(label = "Female Speaker", type = 'filepath'),
+]
 
 title = "Indic Languages Speech Synthesis"