Update app.py

app.py

@@ -71,6 +71,8 @@ class ConversationBot:
             if tool == "Generate Image From User Input Text" or tool == "Generate Text From The Audio" or tool == "Transcribe speech":
                 print("======>Current memory:\n %s" % self.agent.memory)
                 response = re.sub('(image/\S*png)', lambda m: f'![](/file={m.group(0)})*{m.group(0)}*', res['output'])
+                image_filename = res['intermediate_steps'][0][1]
+                response = response + f"![](/file={image_filename})*{image_filename}*"
                 state = state + [(text, response)]
                 print("Outputs:", state)
                 return state, state, gr.Audio.update(visible=False), gr.Image.update(visible=False), gr.Button.update(visible=False)
@@ -160,9 +162,9 @@ class ConversationBot:
         self.t2s = T2S(device="cpu")
         self.i2a = I2A(device="cuda:0")
         self.a2t = A2T(device="cpu")
-        self.asr = ASR(device="cpu")
+        self.asr = ASR(device="cuda:0")
         self.inpaint = Inpaint(device="cuda:0")
-        self.tts_ood = TTS_OOD(device="cpu")
+        # self.tts_ood = TTS_OOD(device="cpu")
         self.tools = [
             Tool(name="Generate Image From User Input Text", func=self.t2i.inference,
                  description="useful for when you want to generate an image from a user input text and it saved it to a file. like: generate an image of an object or something, or generate an image that includes some objects. "
@@ -173,11 +175,11 @@ class ConversationBot:
             Tool(name="Generate Audio From User Input Text", func=self.t2a.inference,
                  description="useful for when you want to generate an audio from a user input text and it saved it to a file."
                              "The input to this tool should be a string, representing the text used to generate audio."),
-            Tool(
-                name="Generate human speech with style derived from a speech reference and user input text and save it to a file", func= self.tts_ood.inference,
-                description="useful for when you want to generate speech samples with styles (e.g., timbre, emotion, and prosody) derived from a reference custom voice."
-                            "Like: Generate a speech with style transferred from this voice. The text is xxx., or speak using the voice of this audio. The text is xxx."
-                            "The input to this tool should be a comma seperated string of two, representing reference audio path and input text."),
+            # Tool(
+            #     name="Generate human speech with style derived from a speech reference and user input text and save it to a file", func= self.tts_ood.inference,
+            #     description="useful for when you want to generate speech samples with styles (e.g., timbre, emotion, and prosody) derived from a reference custom voice."
+            #                 "Like: Generate a speech with style transferred from this voice. The text is xxx., or speak using the voice of this audio. The text is xxx."
+            #                 "The input to this tool should be a comma seperated string of two, representing reference audio path and input text."),
             Tool(name="Generate singing voice From User Input Text, Note and Duration Sequence", func= self.t2s.inference,
                  description="useful for when you want to generate a piece of singing voice (Optional: from User Input Text, Note and Duration Sequence) and save it to a file."
                              "If Like: Generate a piece of singing voice, the input to this tool should be \"\" since there is no User Input Text, Note and Duration Sequence ."