Added streaming and conversation history

app.py

@@ -1,3 +1,4 @@
+import time
 import gradio as gr
 from groq import Groq
 import os
@@ -8,18 +9,30 @@ You are a helpful, respectful and honest assistant. Always answer as helpfully a
 groq_api_key = os.environ.get("groq_api_key")
 
 client = Groq(
-    api_key=groq_api_key,
+    api_key=groq_api_key
 )
 
+conversation_history = [{"role": "system", "content": system_prompt}]
+
 def chat_with_llama(prompt):
+    global conversation_history
+    conversation_history.append({"role": "user", "content": prompt})
     chat_completion = client.chat.completions.create(
-        messages=[{"role": "system", "content": system_prompt},{"role": "user", "content": prompt}],
+        messages=conversation_history,
         model="llama-3.1-8b-instant",
     )
-
-    return chat_completion.choices[0].message.content
+    response = chat_completion.choices[0].message.content
+    conversation_history.append({"role": "assistant", "content": response})
+        # Simulate streaming effect
+    streamed_response = ""
+    for i in range(1, len(response) + 1):
+        streamed_response = response[:i]
+        yield streamed_response
+        time.sleep(0.01)  # Adjust the sleep time to control the streaming speed
 
 def clear_fields():
+    global conversation_history
+    conversation_history = [{"role": "system", "content": system_prompt}]
     return "", ""  # Clear prompt and output, but not the API key
 
 
@@ -29,17 +42,23 @@ with gr.Blocks(theme='freddyaboulton/test-blue') as demo:
     gr.Markdown("<center><h2>Arjun's Chatbot</h2></center>")
     gr.Markdown("Hi there! I'm an AI assistant tasked with answering one question at a time. I'm instructed to provide complete and truthful responses, without any harmful or biased content. I have some memory issues which are work in progress, so I cannot remember previous conversations, yet!. Happy Chatting!")
     prompt = gr.Textbox(label='Question', lines=2, max_lines=5, placeholder = 'Type your question here.')
+    conversation_display = gr.Textbox(label='Conversation History', lines=20, max_lines=30, placeholder='Conversation history will appear here...', interactive=False)
+
+
+    def update_conversation(prompt, history):
+        new_response = chat_with_llama(prompt)
+        updated_history = history + f"\nUser: {prompt}\n\nAssistant: "
+        for partial_response in new_response:
+            updated_history = history + f"\nUser: {prompt}\n\nAssistant: {partial_response}"
+            yield gr.update(value=updated_history)
 
     with gr.Group():
         with gr.Row():
                 submit_btn = gr.Button(value="Submit", elem_id="generate_button", variant="primary", size="sm")
                 clear_btn = gr.ClearButton(value="Clear Question and AI Response", elem_id="clear_button", variant="secondary", size="sm")
-    
-    with gr.Row() as output:
-         llm_output = gr.Markdown("<center><h3>AI Response</h3></center>")
 
-    submit_btn.click(fn=chat_with_llama, inputs = [prompt], outputs=[llm_output])
-    clear_btn.click(fn=clear_fields,outputs=[prompt,llm_output])
+    submit_btn.click(fn=update_conversation, inputs = [prompt,conversation_display], outputs=[conversation_display])
+    clear_btn.click(fn=clear_fields,outputs=[prompt,conversation_display])
 
 demo.launch() 
 #test comment