Create app.py

app.py

@@ -0,0 +1,70 @@
+from transformers import pipeline,Conversation
+import gradio as gr
+
+pipe = pipeline(
+    "automatic-speech-recognition",
+    model="openai/whisper-base",
+)
+
+model= pipeline("conversational", model="facebook/blenderbot-400M-distill")
+
+
+
+# Get a response
+
+# Initialize the conversational model
+# conversational_model = pipeline("conversational", model="microsoft/DialoGPT-medium")
+
+# Mock implementation of transcribe_speech for demonstration
+def transcribe_speech(filepath):
+    output = pipe(
+        filepath,
+        max_new_tokens=256,
+        generate_kwargs={
+            "task": "transcribe",
+            "language": "english",
+        },  # update with the language you've fine-tuned on
+        chunk_length_s=30,
+        batch_size=8,
+    )
+    return output["text"]
+
+def handle_audio_input(audio_file):
+    try:
+        # Step 1: Transcribe the audio
+        transcribed_text = transcribe_speech(audio_file)
+        print(f"Transcribed text: {transcribed_text}")
+
+        # Step 2: Create a conversation and generate a response from transcribed text
+        conversation = Conversation(transcribed_text)
+        response = model(conversation)
+        chatbot_response = response.generated_responses[-1]
+        print(f"Chatbot response: {chatbot_response}")
+
+        return transcribed_text, chatbot_response
+    except Exception as e:
+        print(f"Error: {e}")
+        return "Error in processing audio", str(e)
+
+# Create the Gradio Blocks container
+with gr.Blocks() as demo:
+    gr.Markdown("## Customer query audio to text chatbot")
+
+    with gr.Tab("Microphone"):
+        mic_transcribe = gr.Interface(
+            fn=handle_audio_input,
+            inputs=gr.Audio(sources="microphone", type="filepath"),
+            outputs=[gr.Textbox(label="Transcription"), gr.Textbox(label="Chatbot Response")],
+        )
+        mic_transcribe.render()
+
+    with gr.Tab("File Upload"):
+        file_transcribe = gr.Interface(
+            fn=handle_audio_input,
+            inputs=gr.Audio(sources="upload", type="filepath"),
+            outputs=[gr.Textbox(label="Transcription"), gr.Textbox(label="Chatbot Response")],
+        )
+        file_transcribe.render()
+
+# Launch the Gradio app
+demo.launch(share=True)