Update app.py

app.py

@@ -21,19 +21,14 @@ def synthesise(text_prompt,voice_preset="v2/zh_speaker_1"):
     inputs = processor(text_prompt, voice_preset=voice_preset)
     speech_output = model.generate(**inputs.to(device),pad_token_id=10000)  
     return speech_output
-def speech_to_speech_translation(audio):
-    translated_text = translate(audio)
-    synthesised_speech = synthesise(translated_text)
-    synthesised_speech = (synthesised_speech.numpy() * 32767).astype(np.int16)
-    return 16000, synthesised_speech
 def speech_to_speech_translation(audio,voice_preset="v2/zh_speaker_1"):
     translated_text = translate(audio)
     synthesised_speech = synthesise(translated_text,voice_preset)
     synthesised_speech = (synthesised_speech.numpy() * 32767).astype(np.int16)
-    return synthesised_rate , synthesised_speech 
+    return synthesised_rate , synthesised_speech ,translated_text
 def speech_to_speech_translation_fix(audio,voice_preset="v2/zh_speaker_1"):
-    synthesised_rate,synthesised_speech = speech_to_speech_translation(audio,voice_preset)
-    return synthesised_rate,synthesised_speech.T
+    synthesised_rate,synthesised_speech,translated_text = speech_to_speech_translation(audio,voice_preset)
+    return (synthesised_rate,synthesised_speech.T),translated_text
 
 title = "Multilanguage to Chinese(mandarin) Cascaded STST"
 description = """
@@ -41,52 +36,48 @@ Demo for cascaded speech-to-speech translation (STST), mapping from source speec
 ![Cascaded STST](https://huggingface.co/datasets/huggingface-course/audio-course-images/resolve/main/s2st_cascaded.png "Diagram of cascaded speech to speech translation")
 """
 examples = [
-    ["./mama (1).mp3", None],
-    ["./mama (2).mp3", None],
-    ["./mama (3).mp3", None],
-    ["./mama (4).mp3", None],
-    ["./mama (5).mp3", None],
-    ["./mama (6).mp3", None],
-    ["./mama (7).mp3", None],
-    ["./mama (8).mp3", None],
+    ["./cs-CZ.wav", None],
+    ["./de-DE.wav", None],
+    ["./en-AU.wav", None],
+    ["./en-GB.wav", None],
+    ["./en-US.wav", None],
+    ["./es-ES.wav", None],
+    ["./fr-FR.wav", None],
+    ["./it-IT.wav", None],
+    ["./ko-KR.wav", None],
+    ["./nl-NL.wav", None],
+    ["./pl-PL.wav", None],
+    ["./pt-PT.wav", None],
+    ["./ru-RU.wav", None],
 ]
 import gradio as gr
-demo = gr.Blocks()
-
-# Muti_translate=gr.Interface(
-#     fn=speech_to_speech_translation_fix,
-#     inputs=[
-#         gr.Audio(label="Upload Speech", source="upload", type="filepath"),
-#         gr.Audio(label="Record Speech", source="microphone", type="filepath"),
-#     ],
-#     outputs=[
-#         gr.Audio(label="Generated Speech", type="numpy"),
-#         gr.Text(label="Transcription"),
-#     ],
-#     title=title,
-#     description=description,
-#     examples=examples,
-# )
 
-mic_translate = gr.Interface(
+demo = gr.Blocks()
+file_transcribe = gr.Interface(
     fn=speech_to_speech_translation_fix,
-    inputs=gr.Audio(source="microphone", type="filepath"),
-    outputs=gr.Audio(label="Generated Speech", type="numpy"),
+    inputs=gr.Audio(source="upload", type="filepath"),
+    outputs=[
+        gr.Audio(label="Generated Speech", type="numpy"),
+        gr.Text(label="Transcription"),
+    ],
     title=title,
     description=description,
+    examples=examples,
 )
-
-file_translate = gr.Interface(
+mic_transcribe = gr.Interface(
     fn=speech_to_speech_translation_fix,
-    inputs=gr.Audio(source="upload", type="filepath"),
-    outputs=gr.Audio(label="Generated Speech", type="numpy"),
-    examples=examples,
+    inputs=gr.Audio(source="microphone", type="filepath"),
+    outputs=[
+        gr.Audio(label="Generated Speech", type="numpy"),
+        gr.Text(label="Transcription"),
+    ],
     title=title,
     description=description,
 )
-
 with demo:
-    gr.TabbedInterface([mic_translate, file_translate], ["Microphone", "Audio File"])
-    #gr.TabbedInterface([Muti_translate], ["Record or upload your speech"])
+    gr.TabbedInterface(
+        [file_transcribe, mic_transcribe],
+        ["Transcribe Audio File", "Transcribe Microphone"],
+    )
 
 demo.launch(share=True)