Add text input and generate audio output

app.py

@@ -4,22 +4,30 @@ import scipy.io.wavfile
 import torch
 import torch.nn.functional as F
 from whisperspeech.pipeline import Pipeline
+import time
 
-def process_audio(audio_elem):
+def process_audio(audio_elem,text="This is a test voice genereation"):
+    
+    scipy.io.wavfile.write('test.mp3', audio_elem[0], audio_elem[1])
     
-    scipy.io.wavfile.write('test.mp3', 48000, audio_elem[1])
-
     # print out details about ut
     pipe = Pipeline(s2a_ref='collabora/whisperspeech:s2a-q4-base-en+pl.model')
     # save audio_elem as a file
     speaker = pipe.extract_spk_emb("test.mp3")
-    speaker = speaker.cpu().numpy()  # Move tensor from GPU to CPU and convert to numpy array
-    print(speaker)
+    speaker2 = speaker.cpu().numpy()  # Move tensor from GPU to CPU and convert to numpy array
     #save it locally
-    np.savez_compressed("speaker", features=speaker)
-    return "speaker.npz"
+    np.savez_compressed("speaker", features=speaker2)
+    try:
+
+        pipe.generate_to_file('test.wav', text, lang='en', cps=10.5, speaker=speaker)
+    except Exception as e:
+        print("Error: ", e)
+    
+    return "speaker.npz", "test.wav"
+
+
 
 
 # Define Gradio interface
-with gr.Interface(fn=process_audio, inputs="audio", outputs="file") as iface:
+with gr.Interface(fn=process_audio, inputs=["audio","text"], outputs=["file",'audio']) as iface:
     iface.launch()