code/qa.py

import streamlit as st
from transformers import pipeline
import pandas as pd
import datetime
import base64 
import os

@st.cache(allow_output_mutation = True)
def load_model():
    question_answerer = pipeline('question-answering')
    return question_answerer

#function to convert any dataframe to a csv file
# @st.cache
output_path = "results/df_log_file.csv"
def csv_downloader(df):
    # IMPORTANT: Cache the conversion to prevent computation on every rerun
    res = df.to_csv(index=False,sep="\t").encode('utf-8')
    st.download_button(
    label="Download logs data as CSV separated by tab",
    data=res,
    file_name='df_log_file.csv',
    mime='text/csv')

def load_file():
    """Load text from file"""
    uploaded_file = st.file_uploader("Upload Files",type=['txt'])

    if uploaded_file is not None:
        if uploaded_file.type == "text/plain":
            raw_text = str(uploaded_file.read(),"utf-8")
        return raw_text


st.markdown('![Visitor count](https://shields-io-visitor-counter.herokuapp.com/badge?page=https://share.streamlit.io/https://huggingface.co/spaces/aakashgoel12/nlp1&label=VisitorsCount&labelColor=000000&logo=GitHub&logoColor=FFFFFF&color=1D70B8&style=for-the-badge)')

# Loading Model
question_answerer =load_model()

# App title and description
st.title("Answering questions from text")
st.write("Upload text, pose questions, get answers")

# Load file
st.text("Disclaimer: This app stores user's input for model improvement purposes !!")
raw_text = load_file()
if raw_text != None and raw_text != '':

    # Display text
    with st.expander("See text"):
        st.write(raw_text)

    answer = ''
    question = st.text_input('Ask a question')

    if question != '' and raw_text != '':
        answer = question_answerer({
            'question': question,
            'context': raw_text
        })

        st.write("Answer: {}".format(answer["answer"]))
        st.write("Confidence score: {}".format(round(answer["score"],2)))
        res_df = pd.DataFrame({"TimeStamp":[str(datetime.datetime.now())],\
            "Question":[question],\
            "Input":[str(raw_text)],\
            "Answer":[str(answer["answer"])],\
            "Score":[str(round(answer["score"],2))]})
        res_df.to_csv(output_path, mode='a', index=False, sep="\t", header= not os.path.exists(output_path))
        st.dataframe(pd.read_csv(output_path,sep="\t").tail(4))
        csv_downloader(pd.read_csv(output_path,sep="\t"))


# def csv_downloader(data):
#     csvfile = data.to_csv()
#     b64 = base64.b64encode(csvfile.encode()).decode()
#     new_filename = "results/df_log_file.csv"
#     st.markdown("#### Download File ###")
#     href = f'<a href="data:file/tsv;base64,{b64}" download="{new_filename}">Click Here!!</a>'
#     st.markdown(href,unsafe_allow_html=True)

# log_file_object = open('../logs/log_file.tsv','a')
# log_file_object.write(str(datetime.datetime.now())+'\t'+str(raw_text)+'\t'+str(answer["answer"])+'\t'+str(answer["score"]))
# log_file_object.write('\n')
# log_file_object.close()        


# @st.cache(allow_output_mutation=True)
# def Pageviews():
#     return []

# pageviews=Pageviews()
# pageviews.append('dummy')
# try:
#     st.markdown('Page viewed = {} times.'.format(len(pageviews)))
# except ValueError:
#     st.markdown('Page viewed = {} times.'.format(1))