app.py

import gradio as gr
from g4f import Provider, models
from langchain.llms.base import LLM
from langchain.callbacks.streaming_stdout import StreamingStdOutCallbackHandler
import asyncio
import nest_asyncio
from langchain.callbacks.manager import CallbackManager
from langchain.llms import LlamaCpp
from llama_index import ServiceContext, LLMPredictor, PromptHelper
from llama_index.text_splitter import TokenTextSplitter
from llama_index.node_parser import SimpleNodeParser
from langchain.embeddings import HuggingFaceEmbeddings, HuggingFaceInstructEmbeddings
from llama_index import SimpleDirectoryReader, VectorStoreIndex
from g4f import Provider, models
from langchain.llms.base import LLM
from llama_index.llms import LangChainLLM 
from gradio import Interface
nest_asyncio.apply()
from huggingface_hub import hf_hub_download

model_name_or_path = "hlhr202/llama-7B-ggml-int4"
model_basename = "ggml-model-q4_0.bin" # the model is in bin format

#model_path = hf_hub_download(repo_id=model_name_or_path, filename=model_basename)

n_gpu_layers = 40 # Change this value based on your model and your GPU VRAM pool.
n_batch = 256


embed_model = HuggingFaceInstructEmbeddings(model_name="hkunlp/instructor-xl",
                                                      model_kwargs={"device": "cpu"})
"""
node_parser = SimpleNodeParser.from_defaults(text_splitter=TokenTextSplitter(chunk_size=1024, chunk_overlap=20))
prompt_helper = PromptHelper(
  context_window=4096,
  num_output=256,
  chunk_overlap_ratio=0.1,
  chunk_size_limit=None
)
"""
from langchain_g4f import G4FLLM
llm = LLM = G4FLLM(
        model=models.gpt_35_turbo,
        provider=Provider.Acytoo,)
    

llm = LangChainLLM(llm=llm)
    
service_context = ServiceContext.from_defaults(llm=llm,
  embed_model=embed_model)

documents = SimpleDirectoryReader("data").load_data()
index = VectorStoreIndex.from_documents(documents, service_context=service_context)
  
async def main(question):
    query_engine = index.as_query_engine(service_context=service_context)
    response = query_engine.query(question)
    print(response)
    return response

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