unused_pages/x_LlaMA.py

# import streamlit as st
# from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
# import transformers
# import torch
#
# st.set_page_config(
#     page_title="Falcon 11B"
# )
#
# st.title("Falcon 11B Showcase")
# @st.cache_resource
# def Chat_model():
#     model_name = "tiiuae/falcon-11B"
#     model = AutoModelForCausalLM.from_pretrained(model_name)
#     tokenizer = AutoTokenizer.from_pretrained(model_name)
#     pipeline = transformers.pipeline(
#         "text-generation",
#         model=model,
#         tokenizer=tokenizer,
#         torch_dtype=torch.bfloat16,
#         device_map="auto",
#     )
#     return pipeline,tokenizer
#
# def get_text_output(user_input,pipeline,tokenizer):
#     sequences = pipeline(
#         user_input,
#         max_length=200,
#         do_sample=True,
#         top_k=10,
#         num_return_sequences=1,
#         eos_token_id=tokenizer.eos_token_id,
#     )
#     return sequences
#
# if "Falcon_messages" not in st.session_state:
#     st.session_state.Falcon_messages = []
#
# if "Falcon_model" not in st.session_state:
#     st.session_state.Falcon_model,st.session_state.tokeniser = Chat_model()
#
# for message in st.session_state.Falcon_messages:
#     with st.chat_message(message["role"]):
#         st.markdown(message["content"])
#
# if prompt := st.chat_input("What is up?"):
#     st.session_state.Falcon_messages.append({"role": "user", "content": prompt})
#     with st.chat_message("user"):
#         st.markdown(prompt)
#     with st.chat_message("assistant"):
#         response = get_text_output(prompt,st.session_state.Falcon_model,st.session_state.tokeniser)
#     st.session_state.Falcon_messages.append({"role": "assistant", "content": response})