Spaces:
Runtime error
Runtime error
| import streamlit as st | |
| from transformers import pipeline | |
| import spacy | |
| from spacy import displacy | |
| import plotly.express as px | |
| import numpy as np | |
| st.set_page_config(page_title="NLP Prototype") | |
| st.title("Natural Language Processing Prototype") | |
| st.write("_This web application is intended for educational use, please do not upload any sensitive information._") | |
| st.subheader("__Which natural language processing task would you like to try?__") | |
| st.write("- __Sentiment Analysis:__ Identifying whether a piece of text has a positive or negative sentiment.") | |
| st.write("- __Named Entity Recognition:__ Identifying all geopolitical entities, organizations, people, locations, or dates in a body of text.") | |
| st.write("- __Text Classification:__ Placing a piece of text into one or more categories.") | |
| st.write("- __Text Summarization:__ Condensing larger bodies of text into smaller bodies of text.") | |
| option = st.selectbox('Please select from the list',('','Sentiment Analysis','Named Entity Recognition', 'Text Classification','Text Summarization')) | |
| def Loading_Model_1(): | |
| sum2 = pipeline("summarization",framework="pt") | |
| return sum2 | |
| def Loading_Model_2(): | |
| class1 = pipeline("zero-shot-classification",framework="pt") | |
| return class1 | |
| def Loading_Model_3(): | |
| sentiment = pipeline("sentiment-analysis", framework="pt") | |
| return sentiment | |
| def Loading_Model_4(): | |
| nlp = spacy.load('en_core_web_sm') | |
| return nlp | |
| def entRecognizer(entDict, typeEnt): | |
| entList = [ent for ent in entDict if entDict[ent] == typeEnt] | |
| return entList | |
| def plot_result(top_topics, scores): | |
| top_topics = np.array(top_topics) | |
| scores = np.array(scores) | |
| scores *= 100 | |
| fig = px.bar(x=scores, y=top_topics, orientation='h', | |
| labels={'x': 'Probability', 'y': 'Category'}, | |
| text=scores, | |
| range_x=(0,115), | |
| title='Top Predictions', | |
| color=np.linspace(0,1,len(scores)), | |
| color_continuous_scale="Bluered") | |
| fig.update(layout_coloraxis_showscale=False) | |
| fig.update_traces(texttemplate='%{text:0.1f}%', textposition='outside') | |
| st.plotly_chart(fig) | |
| with st.spinner(text="Please wait for the models to load. This should take approximately 60 seconds."): | |
| sum2 = Loading_Model_1() | |
| class1 = Loading_Model_2() | |
| sentiment = Loading_Model_3() | |
| nlp = Loading_Model_4() | |
| if option == 'Text Classification': | |
| cat1 = st.text_input('Enter each possible category name (separated by a comma). Maximum 5 categories.') | |
| text = st.text_area('Enter Text Below:', height=200) | |
| submit = st.button('Generate') | |
| if submit: | |
| st.subheader("Classification Results:") | |
| labels1 = cat1.strip().split(',') | |
| result = class1(text, candidate_labels=labels1) | |
| cat1name = result['labels'][0] | |
| cat1prob = result['scores'][0] | |
| st.write('Category: {} | Probability: {:.1f}%'.format(cat1name,(cat1prob*100))) | |
| plot_result(result['labels'][::-1][-10:], result['scores'][::-1][-10:]) | |
| if option == 'Text Summarization': | |
| max_lengthy = st.slider('Maximum summary length (words)', min_value=30, max_value=150, value=60, step=10) | |
| num_beamer = st.slider('Speed vs quality of summary (1 is fastest)', min_value=1, max_value=8, value=4, step=1) | |
| text = st.text_area('Enter Text Below (maximum 800 words):', height=300) | |
| submit = st.button('Generate') | |
| if submit: | |
| st.subheader("Summary:") | |
| with st.spinner(text="This may take a moment..."): | |
| summWords = sum2(text, max_length=max_lengthy, min_length=15, num_beams=num_beamer, do_sample=True, early_stopping=True, repetition_penalty=1.5, length_penalty=1.5) | |
| text2 =summWords[0]["summary_text"] | |
| st.write(text2) | |
| if option == 'Sentiment Analysis': | |
| text = st.text_area('Enter Text Below:', height=200) | |
| submit = st.button('Generate') | |
| if submit: | |
| st.subheader("Sentiment:") | |
| result = sentiment(text) | |
| sent = result[0]['label'] | |
| cert = result[0]['score'] | |
| st.write('Text Sentiment: {} | Probability: {:.1f}%'.format(sent,(cert*100))) | |
| if option == 'Named Entity Recognition': | |
| text = st.text_area('Enter Text Below:', height=300) | |
| submit = st.button('Generate') | |
| if submit: | |
| entities = [] | |
| entityLabels = [] | |
| doc = nlp(text) | |
| for ent in doc.ents: | |
| entities.append(ent.text) | |
| entityLabels.append(ent.label_) | |
| entDict = dict(zip(entities, entityLabels)) | |
| entOrg = entRecognizer(entDict, "ORG") | |
| entPerson = entRecognizer(entDict, "PERSON") | |
| entDate = entRecognizer(entDict, "DATE") | |
| entGPE = entRecognizer(entDict, "GPE") | |
| entLoc = entRecognizer(entDict, "LOC") | |
| options = {"ents": ["ORG", "GPE", "PERSON", "LOC", "DATE"]} | |
| HTML_WRAPPER = """<div style="overflow-x: auto; border: 1px solid #e6e9ef; border-radius: 0.25rem; padding: 1rem; margin-bottom: 2.5rem">{}</div>""" | |
| st.subheader("List of Named Entities:") | |
| st.write("Geopolitical Entities (GPE): " + str(entGPE)) | |
| st.write("People (PERSON): " + str(entPerson)) | |
| st.write("Organizations (ORG): " + str(entOrg)) | |
| st.write("Dates (DATE): " + str(entDate)) | |
| st.write("Locations (LOC): " + str(entLoc)) | |
| st.subheader("Original Text with Entities Highlighted") | |
| html = displacy.render(doc, style="ent", options=options) | |
| html = html.replace("\n", " ") | |
| st.write(HTML_WRAPPER.format(html), unsafe_allow_html=True) |