pdfToChatGPT.py

from llama_index import SimpleDirectoryReader
from llama_index import GPTListIndex
from llama_index import GPTVectorStoreIndex
from llama_index import  LLMPredictor, PromptHelper
from llama_index import ServiceContext, load_graph_from_storage
from llama_index import StorageContext, load_index_from_storage

from langchain import OpenAI
import gradio as gr
import sys
import os

os.environ["OPENAI_API_KEY"] = 'Your Open AI key'


# data/pdfs has all the pdf files you want to train your chatbot on
documents = SimpleDirectoryReader('data/pdfs').load_data()
index = GPTVectorStoreIndex.from_documents(documents)

llm_predictor = LLMPredictor(llm=OpenAI(temperature=0, model_name="text-davinci-003"))

from llama_index import  GPTVectorStoreIndex

max_input_size = 4096
num_output = 256
max_chunk_overlap = 0.7 # now takes a value between 0 and 1, before, ints
prompt_helper = PromptHelper(max_input_size, num_output, max_chunk_overlap)

service_context = ServiceContext.from_defaults(llm_predictor=llm_predictor, prompt_helper=prompt_helper)

index = GPTVectorStoreIndex.from_documents(
    documents, service_context=service_context
)
# save index
index.storage_context.persist("indices/ch1")



# rebuild storage context
storage_context = StorageContext.from_defaults(persist_dir='./indices/ch1')
# load index
index = load_index_from_storage(storage_context)

query_engine = index.as_query_engine()
response = query_engine.query("What are the prohibitions in currency dealing?") 
# replace above with your question
print(response)