Update app.py

app.py

@@ -73,7 +73,7 @@ import librosa
 def tts_fn(slider_value, input_text):
     audio_embedding = dataset[slider_value]['audio']['array']
     sample_rate_embedding = dataset[slider_value]['audio']['sampling_rate']
-    if original_sr != target_sr:
+    if sample_rate_embedding != 16000:
         audio_embedding = librosa.resample(audio_embedding, orig_sr=sample_rate_embedding, target_sr=16_000)
         
     
@@ -113,6 +113,15 @@ def tts_fn(slider_value, input_text):
     return 16000, (audio_out.reshape(-1)).astype(np.int16)
 
 
+def master_fn(slider_value, input_text):
+    all_speech = []
+    for sentence in input_text.split(".")[:-1]:
+        sampling_rate_response, audio_chunk_response = tts_fn(slider_value, sentence)
+        all_speech.append(audio_chunk_response)
+        
+    audio_response = np.concatenate(all_speech)
+    return sampling_rate_response, audio_response
+
 import gradio as gr
 
 slider = gr.Slider(
@@ -131,7 +140,7 @@ text_input = gr.Textbox(
 
 
 demo = gr.Interface(
-    fn = tts_fn,
+    fn = master_fn,
     inputs=[slider, text_input],  # List of inputs
     outputs = "audio"
 )