File size: 3,083 Bytes
5a95a6c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 |
import streamlit as st, time
from utils import st_def, ut_openai
st_def.st_logo(title = "Welcome 👋 to Summary!", page_title="Summary",)
st_def.st_summary()
openai_api_key= st_def.st_sidebar()
#------------------------------------------------------------------------
def init():
if 'page_text' not in st.session_state:
st.error('Read PDF before continue ... ')
return False
elif not openai_api_key:
st.error("Please add your OpenAI API key to continue.")
return False
else:
return True
def combine_chunks(summaries):
chunks = [] # combine chunks of "summaries" array into one string of max 4000 characters
summary = ""
for sum in summaries:
if len(sum) + len(summary) > 4000:
chunks.append(summary)
summary = ""
summary += sum
if len(chunks) == 0: chunks.append(summary)
return chunks
def main():
if not init(): return
page_text_array = st.session_state['page_text'] # array, store pages. len(text) is pages.
print("Summarizing text..."+str(len(page_text_array)))
combined_summaries = combine_chunks(page_text_array)
print("Found " + str(len(combined_summaries)) + " chunks to summarize.")
iterations = 1
while True:
if len(combined_summaries) <= 1: break
summaries_of_summaries = []
# print summaries
for i, summary in enumerate(combined_summaries):
prompt =f"""
Your task is to extract relevant information from a text on the page of a book. This information will be used to create a book summary.
Extract relevant information from the following text, which is delimited with triple backticks.\
Be sure to preserve the important details.
Text: ```{combined_summaries[i]}```
"""
# st.write(f"Summarizing {i + 1} of {len(combined_summaries)}, iteration {iterations}...")
st.markdown(f'<span style="color:blue">Summarizing {i + 1} of {len(combined_summaries)}, iteration {iterations}...</span>', unsafe_allow_html=True)
sum_page = ut_openai.aichat(openai_api_key=openai_api_key, messages = [{"role": "user", "content": prompt},])
summaries_of_summaries.append(sum_page)
st.text(sum_page)
time.sleep(2) #You can query the model only 3 times in a minute for free, so we need to put some delay
st.write('summaries_of_summaries')
st.write(summaries_of_summaries)
combined_summaries = combine_chunks(summaries_of_summaries)
st.text('combined_summaries')
st.text(combined_summaries)
iterations += 1
# summarize last chunk
with st.spinner("Summarizing last chunk..."):
final_summary = ut_openai.aichat(openai_api_key=openai_api_key, messages = [{"role": "user", "content": combined_summaries[0]},])
st.header("Final Summary")
st.write(final_summary)
st.success("🚨Cheers!")
if __name__ == "__main__":
main()
|