Rename app.py to main.py

app.py

@@ -1,60 +0,0 @@
-from ctransformers import AutoModelForCausalLM
-import gradio as gr
-
-greety = """
-Follow us [Gathnex](https://medium.com/@gathnex), [linkedin](https://www.linkedin.com/company/gathnex/) and [Github](https://github.com/gathnexadmin) for  more update on Genrative AI, LLM,etc. A special thanks to the Gathnex team members who made a significant contribution to this project.
-"""
-
-llm = AutoModelForCausalLM.from_pretrained("zephyr-7b-beta.Q4_K_S.gguf",
-model_type='mistral',
-max_new_tokens = 1096,
-threads = 3,
-)
-
-def stream(prompt, UL):
-    system_prompt = 'You are a helpful AI assistant'
-    E_INST = "</s>"
-    user, assistant = "<|user|>", "<|assistant|>"
-    prompt = f"{system_prompt}{E_INST}\n{user}\n{prompt.strip()}{E_INST}\n{assistant}\n"
-    return llm(prompt)
-
-css = """
-  h1 {
-  text-align: center;
-}
-#duplicate-button {
-  margin: auto;
-  color: white;
-  background: #1565c0;
-  border-radius: 100vh;
-}
-.contain {
-  max-width: 900px;
-  margin: auto;
-  padding-top: 1.5rem;
-}
-"""
-
-chat_interface = gr.ChatInterface(
-    fn=stream,
-    #additional_inputs_accordion_name = "Credentials",
-    #additional_inputs=[
-    #     gr.Textbox(label="OpenAI Key", lines=1),
-    #     gr.Textbox(label="Linkedin Access Token", lines=1),
-    #],
-    stop_btn=None,
-    examples=[
-        ["explain Large language model"],
-        ["what is quantum computing"]
-    ],
-)
-
-with gr.Blocks(css=css) as demo:
-    gr.HTML("<h1><center>Gathnex Free LLM Deployment Space<h1><center>")
-    gr.HTML("<h3><center><a href='https://medium.com/@gathnex'>Gathnex AI</a>💬<h3><center>")
-    gr.DuplicateButton(value="Duplicate Space for private use", elem_id="duplicate-button")
-    chat_interface.render()
-    gr.Markdown(greety)
-    
-if __name__ == "__main__":
-    demo.queue(max_size=10).launch()

main.py

@@ -0,0 +1,51 @@
+from ctransformers import AutoModelForCausalLM
+from gradio import Chatbot, Interface
+import gradio as gr
+
+# Load the GGUF model
+llm = AutoModelForCausalLM.from_pretrained(
+    "zephyr-7b-beta.Q4_K_S.gguf",
+    model_type="mistral",
+    max_new_tokens=1096,
+    threads=3
+)
+
+# Format prompt with system message and chat history
+def format_prompt(message, chat_history):
+    system_prompt = "Below is an instruction that describes a task. Write a response that appropriately completes the request."
+    E_INST = "</s>"
+    user, assistant = "<|user|>", "<|assistant|>"
+    prompt = f"{system_prompt}{E_INST}\n"
+    for user_msg, bot_msg in chat_history:
+        prompt += f"{user}\n{user_msg}{E_INST}\n{assistant}\n{bot_msg}{E_INST}\n"
+    prompt += f"{user}\n{message}{E_INST}\n{assistant}\n"
+    return prompt
+
+# Define chatbot function
+def respond(message, chat_history):
+    formatted_prompt = format_prompt(message, chat_history)
+    response = llm(formatted_prompt)
+    chat_history.append((message, response))
+    return chat_history, chat_history
+
+# Create Gradio Chatbot UI
+chatbot = Chatbot(
+    bubble_full_width=False,
+    height=500
+)
+
+# Launch interface
+with gr.Blocks() as demo:
+    gr.Markdown("## Zephyr LLM Chat Interface")
+    chatbot = gr.Chatbot()
+    msg = gr.Textbox(label="Your Message")
+    clear = gr.Button("Clear Chat")
+
+    state = gr.State([])
+
+    msg.submit(respond, [msg, state], [chatbot, state])
+    clear.click(lambda: ([], []), None, [chatbot, state])
+
+# Launch Gradio app
+if __name__ == "__main__":
+    demo.launch()