Create app.py

app.py

@@ -0,0 +1,38 @@
+import gradio as gr
+from transformers import pipeline
+
+# Load the 3 TTS models from your E-motionAssistant Org
+eng_tts = pipeline("text-to-speech", model="E-motionAssistant/text-to-speech-VITS-english")
+sin_tts = pipeline("text-to-speech", model="E-motionAssistant/text-to-speech-VITS-sinhala")
+tam_tts = pipeline("text-to-speech", model="E-motionAssistant/text-to-speech-VITS-tamil")
+
+def generate_audio(text, language):
+    try:
+        if language == "English":
+            out = eng_tts(text)
+        elif language == "Sinhala":
+            out = sin_tts(text)
+        elif language == "Tamil":
+            out = tam_tts(text)
+        else:
+            return None
+        
+        # Returns (sampling_rate, audio_data)
+        return (out["sampling_rate"], out["audio"][0])
+    
+    except Exception as e:
+        print(f"Error: {e}")
+        return None
+
+# The 'Interface' defines how Vercel will talk to this Space
+demo = gr.Interface(
+    fn=generate_audio,
+    inputs=[
+        gr.Textbox(label="Input Text"),
+        gr.Dropdown(["English", "Sinhala", "Tamil"], label="Select Language")
+    ],
+    outputs=gr.Audio(label="Output Audio"),
+    api_name="predict"
+)
+
+demo.launch()