alpertml/TopicModelingForSummarization

1# external libraries2import streamlit as st3from transformers import pipeline4    5import pandas as pd6 7# internal libraries8from config import config9import pipeline10 11 12def main():13 14    st.set_page_config(15        layout="centered",  # Can be "centered" or "wide". In the future also "dashboard", etc.16        initial_sidebar_state="auto",  # Can be "auto", "expanded", "collapsed"17        page_title=config.main_title,  # String or None. Strings get appended with "• Streamlit". 18        page_icon=config.logo_path,  # String, anything supported by st.image, or None.19    )20 21    if "output" not in st.session_state:22        st.session_state['data'] = pd.read_csv(config.sample_texts_path)23        st.session_state['sample_text'] = None24        generate_text()25        st.session_state["output"] = False26        st.session_state["output_text"] = ""27        st.session_state['inputs'] = {}28 29    col1, col2, col3 = st.columns(3)30    col1.write(' ')31    col2.image(config.logo_path)32    col3.write(' ')33 34    st.markdown(f"<h1 style='text-align: center;'>{config.main_title}</h1>", unsafe_allow_html=True)35    st.markdown(f"<h3 style='text-align: center;'>{config.lecture_title}</h3>", unsafe_allow_html=True)36 37    # topic modelling radio bar38    input_topic_modelling = st.radio(39                        config.topic_modelling_title,40                        config.topic_modelling_answers,41                        horizontal=True)42    st.session_state['inputs']['input_topic_modelling'] = input_topic_modelling43 44    # input text area45    input_text = st.text_area(config.input_text, st.session_state['sample_text'], height=300)46    st.session_state['inputs']['input_text'] = input_text47 48    # generate sample text button49    st.button(config.button_text, on_click=generate_text)50    51    # choosing segmenter radio bar52    input_segmenter = st.radio(53                        config.segmenter_title,54                        config.segmenter_answers,55                        horizontal=True)56    st.session_state['inputs']['input_segmenter'] = input_segmenter57    58    # choosing summarizer algorithm radio bar59    input_summarizer = st.radio(60                        config.summarizer_title,61                        config.summarizer_answers,62                        horizontal=True)63    st.session_state['inputs']['input_summarizer'] = input_summarizer64    65    # generating summary button66    col1, col2, col3 = st.columns(3)67    col1.header(' ')68    col2.button(config.generate_text, on_click=generate_summary)69    col3.header(' ')70 71    if st.session_state["output"]:72        73        TOPICS = [key for key, value in st.session_state["output_text"].items() if key != '#']74 75        if config.filter_threshold_summaries:76            TOPICS = [key for key in TOPICS if st.session_state["output_text"][key]['summary'] != config.threshold_error]77 78        st.write(config.output_title)79        options = {}80        for topic in TOPICS:81            option = st.checkbox(topic)82            options[topic] = option83 84        if len(options) == 0:85            st.warning(config.warning_len_input_text, icon="⚠️")86 87        for topic, option in options.items():88            if option == True:89                st.text_area(topic, 90                            st.session_state["output_text"][topic]['summary'],91                            disabled=True)92 93def generate_text():94    df = st.session_state['data']95    df = df[~df['data'].isnull()]96    df = df[df['data'].str.len().gt(100)]97    st.session_state['sample_text'] = df.sample(1)['data'].values[0]98 99def generate_summary():100    st.session_state["output"] = True101 102    MODELS = {103        'summarizer':st.session_state['inputs']['input_summarizer'],104        'topic_modelling':st.session_state['inputs']['input_topic_modelling'],105        'segmentizer':st.session_state['inputs']['input_segmenter']106    }107 108    with st.spinner('Generating the output of Topic Modeling for Summarization...'):109            OUTPUT = pipeline.run(st.session_state['inputs']['input_text'], MODELS)110 111    st.session_state["output_text"] = OUTPUT112    st.success('Done!')113 114 115if __name__ == "__main__":116    main()