import os import octoai octoai_client = octoai.client.Client(token=os.getenv('OCTOML_KEY')) from pinecone import Pinecone, ServerlessSpec pc = Pinecone(api_key=os.getenv('PINECONE_API_KEY')) from llama_index.vector_stores.pinecone import PineconeVectorStore from llama_index.core import VectorStoreIndex from llama_index.core.response.pprint_utils import pprint_source_node from llama_index.llms.octoai import OctoAI octoai = OctoAI( token=os.getenv('OCTOML_KEY'), model="meta-llama-3-70b-instruct", max_tokens=512, temperature=0.1, ) from llama_index.core.memory import ChatMemoryBuffer import gradio as gr from io import StringIO def get_credit_dist(history): _out = StringIO() print("Disabled momentarily...", file=_out) return _out.getvalue() with gr.Blocks() as demo: chatbot = gr.Chatbot(height=800) msg = gr.Textbox() clear = gr.Button("Clear") credit_box = gr.Textbox(label="Credit distribution", lines=20, autoscroll=False) credit_btn = gr.Button("Credit response") def get_chat_engine(): vector_store = PineconeVectorStore(pinecone_index=pc.Index('prorata-postman-ds-256')) vindex = VectorStoreIndex.from_vector_store(vector_store) memory = ChatMemoryBuffer.from_defaults(token_limit=5000) return vindex.as_chat_engine( chat_mode="context", llm=octoai, memory=memory, system_prompt="You are a chatbot, able to have normal interactions, as well as talk about news events provided in the context of the conversation.", ) chat_engine_var = gr.State(get_chat_engine) def user(user_message, history): return "", history + [[user_message, None]] def bot(history, chat_engine): response = chat_engine.stream_chat(history[-1][0]) history[-1][1] = "" for token in response.response_gen: history[-1][1] += token yield history msg.submit(user, [msg, chatbot], [msg, chatbot], queue=False).then(bot, [chatbot, chat_engine_var], chatbot) clear.click(lambda x: x.reset(), chat_engine_var, chatbot, queue=False) credit_btn.click(get_credit_dist, chatbot, credit_box) if __name__ == "__main__": demo.queue() demo.launch()