app.py

import sentencepiece
import streamlit as st
import pandas as pd
import spacy
import roner

example_list = [
    "Ana merge în București.",
    """Ana merge în București. Ana merge în București. Ana merge în București. Ana merge în București. Ana merge în București. Ana merge în București."""
]

st.set_page_config(layout="wide")

st.title("Demo for Romanian NER")

model_list = ['dumitrescustefan/bert-base-romanian-ner']

st.sidebar.header("Select NER Model")
model_checkpoint = st.sidebar.radio("", model_list)

st.sidebar.write("For details of models: 'https://huggingface.co/dumitrescustefan/")
st.sidebar.write("")

xlm_agg_strategy_info = "'aggregation_strategy' can be selected as 'simple' or 'none' for 'xlm-roberta' because of the RoBERTa model's tokenization approach."

st.sidebar.header("Select Aggregation Strategy Type")
if model_checkpoint == "akdeniz27/xlm-roberta-base-turkish-ner":
    aggregation = st.sidebar.radio("", ('simple', 'none'))
    st.sidebar.write(xlm_agg_strategy_info)
elif model_checkpoint == "xlm-roberta-large-finetuned-conll03-english":
    aggregation = st.sidebar.radio("", ('simple', 'none'))
    st.sidebar.write(xlm_agg_strategy_info)
    st.sidebar.write("")
    st.sidebar.write("This English NER model is included just to show the zero-shot transfer learning capability of XLM-Roberta.")
else:
    aggregation = st.sidebar.radio("", ('first', 'simple', 'average', 'max', 'none'))
    
st.sidebar.write("Please refer 'https://huggingface.co/transformers/_modules/transformers/pipelines/token_classification.html' for entity grouping with aggregation_strategy parameter.")

st.subheader("Select Text Input Method")
input_method = st.radio("", ('Select from Examples', 'Write or Paste New Text'))
if input_method == 'Select from Examples':
    selected_text = st.selectbox('Select Text from List', example_list, index=0, key=1)
    st.subheader("Text to Run")
    input_text = st.text_area("Selected Text", selected_text, height=128, max_chars=None, key=2)
elif input_method == "Write or Paste New Text":
    st.subheader("Text to Run")
    input_text = st.text_area('Write or Paste Text Below', value="", height=128, max_chars=None, key=2)

@st.cache(allow_output_mutation=True)
def setModel(named_persons_only):
    ner = roner.NER(named_persons_only=named_persons_only)
    return ner

@st.cache(allow_output_mutation=True)
def get_html(html: str):
    WRAPPER = """<div style="overflow-x: auto; border: 1px solid #e6e9ef; border-radius: 0.25rem; padding: 1rem; margin-bottom: 2.5rem">{}</div>"""
    html = html.replace("\n", " ")
    return WRAPPER.format(html)
        
Run_Button = st.button("Run", key=None)
if Run_Button == True:
    
    ner = setModel(named_persons_only = False)
    output = ner(input_text)[0] # only one sentence

    # tabular form
    data = []
    for word in output["words"]:
        if word["tag"]!="O":
            data.append({
                "word": word["text"],
                "tag": word["tag"],
                "start_char": word["start_char"],
                "end_char": word["end_char"],
                "span_after": word["span_after"],
            })
    df = pd.DataFrame.from_dict(data)
    st.subheader("Recognized Entities")
    st.dataframe(df)


    st.subheader("Spacy Style Display")
    spacy_display = {}
    spacy_display["ents"] = []
    spacy_display["text"] = output["text"]
    spacy_display["title"] = None

    for word in output["words"]:
        #spacy_display["ents"].append({"start": entity["start"], "end": entity["end"], "label": entity["entity_group"]})
        spacy_display["ents"].append({"start": word["start_char"], "end": word["end_char"], "label": word["tag"]})
    
    entity_list = ['O', 'PERSON', 'ORG', 'GPE', 'LOC', 'NAT_REL_POL',
     'EVENT', 'LANGUAGE', 'WORK_OF_ART', 'DATETIME',
     'PERIOD', 'MONEY', 'QUANTITY', 'NUMERIC',
     'ORDINAL', 'FACILITY']
    colors = {
        'O': '#FFF',
        'PERSON': '#F00',
        'ORG': '#F00',
        'GPE': '#F00',
        'LOC': '#F00',
        'NAT_REL_POL': '#F00',
        'EVENT': '#F00',
        'LANGUAGE': '#F00',
        'WORK_OF_ART': '#F00',
        'DATETIME': '#F00',
        'PERIOD': '#F00',
        'MONEY': '#F00',
        'QUANTITY': '#F00',
        'NUMERIC': '#F00',
        'ORDINAL': '#F00',
        'FACILITY': '#F00',
    }
    html = spacy.displacy.render(spacy_display, style="ent", minify=True, manual=True, options={"ents": entity_list, "colors": colors})
    style = "<style>mark.entity { display: inline-block }</style>"
    st.write(f"{style}{get_html(html)}", unsafe_allow_html=True)