app.py

import logging
import sys
import gradio as gr
import asyncio
import nest_asyncio

logging.basicConfig(stream=sys.stdout, level=logging.INFO)
logging.getLogger().addHandler(logging.StreamHandler(stream=sys.stdout))

from llama_index.core import (
    ServiceContext,
    SimpleDirectoryReader,
    StorageContext,
    VectorStoreIndex,
    set_global_service_context,
)


#from langchain.embeddings import HuggingFaceEmbeddings, HuggingFaceInstructEmbeddings 
#from llama_index.embeddings.huggingface import HuggingFaceInstructEmbeddings
from langchain_community.embeddings import HuggingFaceInstructEmbeddings 


from g4f import Provider, models
from langchain.llms.base import LLM
from llama_index.llms.langchain import LangChainLLM
from langchain_g4f import G4FLLM

nest_asyncio.apply()
"""
documents = SimpleDirectoryReader('data').load_data()
model_kwargs = {'device': 'cpu'}
encode_kwargs = {'normalize_embeddings': True}
embed_model = HuggingFaceInstructEmbeddings(
    model_name="hkunlp/instructor-xl", model_kwargs=model_kwargs,
    encode_kwargs=encode_kwargs 
) 

"""

#from langchain.embeddings import HuggingFaceInstructEmbeddings

model_name = "hkunlp/instructor-large"
model_kwargs = {'device': 'cpu'}
encode_kwargs = {'normalize_embeddings': True}
embed_model = HuggingFaceInstructEmbeddings(
    model_name=model_name,
    model_kwargs=model_kwargs,
    encode_kwargs=encode_kwargs
)
llm= LLM = G4FLLM(
        model=models.gpt_35_turbo,
        provider=Provider.ChatgptAi,
    )

llm = LangChainLLM(llm=llm)

service_context = ServiceContext.from_defaults(chunk_size=512, llm=llm, embed_model=embed_model)

index = VectorStoreIndex.from_documents(documents, service_context=service_context)

 
async def main(query):

    query_engine = index.as_query_engine()
    response = query_engine.query(query)
    print(response)
    return response

iface = gr.Interface(fn=main, inputs="text", outputs="text")
iface.launch()