import streamlit as st from gradio_client import Client # Constants TITLE = "Llama2 70B Chatbot" DESCRIPTION = """ This Space demonstrates model [Llama-2-70b-chat-hf](https://huggingface.co/meta-llama/Llama-2-70b-chat-hf) by Meta, a Llama 2 model with 70B parameters fine-tuned for chat instructions. """ # Initialize client client = Client("https://ysharma-explore-llamav2-with-tgi.hf.space/") # Prediction function def predict(message, system_prompt="", temperature=0.9, max_new_tokens=4096): with st.status("Requesting LLama-2"): st.write("Requesting API") response = client.predict( message, # str in 'Message' Textbox component system_prompt, # str in 'Optional system prompt' Textbox component temperature, # int | float (numeric value between 0.0 and 1.0) max_new_tokens, # int | float (numeric value between 0 and 4096) 0.3, # int | float (numeric value between 0.0 and 1) 1, # int | float (numeric value between 1.0 and 2.0) api_name="/chat" ) st.write("Done") return response # Streamlit UI st.title(TITLE) st.write(DESCRIPTION) # React to user input if prompt := st.chat_input("Ask LLama-2-70b anything..."): # Display user message in chat message container st.chat_message("human",avatar = "🧑‍💻").markdown(prompt) # Add user message to chat history st.session_state.messages.append({"role": "human", "content": prompt}) response = predict(prompt) # Display assistant response in chat message container with st.chat_message("assistant", avatar='🦙'): st.markdown(response) # Add assistant response to chat history st.session_state.messages.append({"role": "assistant", "content": response})