Update app.py

app.py

@@ -38,26 +38,21 @@ run_opts={"device":"cuda"},
 def generate_prompt(instruction, input=None):
     if input:
         return f"""Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.
-
 # Instruction:
 {instruction}
-
 # Input:
 {input}
-
 # Response:
 """
     else:
         return f"""Below is an instruction that describes a task. Write a response that appropriately completes the request.
-
 # Instruction:
 {instruction}
-
 # Response:
 """
 
 def evaluate(
-#    upload,
+    upload,
     audio,
 #    instruction,
 #    input=None,
@@ -90,8 +85,7 @@ def evaluate(
                      token_ban = [], # ban the generation of some tokens
                      token_stop = [0]) # stop generation whenever you see any token here
 
-#    instruction = result.text.strip()
-    instruction = result.text
+    instruction = result.text.strip()
     input=None
 #    input = input.strip()
     ctx = generate_prompt(instruction, input)
@@ -128,16 +122,30 @@ def evaluate(
 
     res.append(out_str.strip())
 
-    res1 = ' '.join(str(x) for x in res)
+    res1 = ''.join(str(x) for x in res)
 
-    return res
+    tts.tts_to_file(res1, speaker_wav = upload, language="en", file_path="output.wav")
+    
+    voicefixer.restore(input="output.wav", # input wav file path
+                    output="audio1.wav", # output wav file path
+                    cuda=True, # whether to use gpu acceleration
+                    mode = 0) # You can try out mode 0, 1, or 2 to find out the best result
+    
+    noisy = enhance_model.load_audio(
+    "audio1.wav"
+    ).unsqueeze(0)
+
+    enhanced = enhance_model.enhance_batch(noisy, lengths=torch.tensor([1.]))
+    torchaudio.save("enhanced.wav", enhanced.cpu(), 16000)
+    
+    return [result.text, res, "enhanced.wav"]
     
 #    yield out_str.strip()
 
 g = gr.Interface(
     fn=evaluate,
     inputs=[
-#        gr.Audio(source="upload", label = "请上传您喜欢的声音(wav文件)", type="filepath"),
+        gr.Audio(source="upload", label = "请上传您喜欢的声音(wav文件)", type="filepath"),
         gr.Audio(source="microphone", label = "和您的专属AI聊天吧！", type="filepath"),
 #        gr.components.Textbox(lines=2, label="Instruction", value="Tell me about ravens."),
 #        gr.components.Textbox(lines=2, label="Input", placeholder="none"),
@@ -148,13 +156,13 @@ g = gr.Interface(
 #        gr.components.Slider(0.0, 1.0, step=0.1, value=0.4),  # countPenalty        
     ],
     outputs=[
-#        gr.Textbox(label="Speech to Text"),
+        gr.Textbox(label="Speech to Text"),
         gr.Textbox(label="Raven Output"),
-#        gr.Audio(label="Audio with Custom Voice"),
+        gr.Audio(label="Audio with Custom Voice"),
     ],
     title="🥳💬💕 - TalktoAI，随时随地，谈天说地！",
     description="🤖 - 让有人文关怀的AI造福每一个人！AI向善，文明璀璨！TalktoAI - Enable the future！",
     article = "Powered by the RWKV Language Model"
 )
 g.queue(concurrency_count=1, max_size=10)
-g.launch(show_error=True)
+g.launch(show_error=True)