app.py

import json
from collections import defaultdict

import gradio as gr
import pandas as pd
from transformers import pipeline, AutoTokenizer, AutoModelForTokenClassification

tokenizer = AutoTokenizer.from_pretrained("d4data/biomedical-ner-all")
model = AutoModelForTokenClassification.from_pretrained("d4data/biomedical-ner-all")


EXAMPLE_MAP = {}
with open("examples.json", "r") as f:
    example_json = json.load(f)
    EXAMPLE_MAP = {x["text"]: x["label"] for x in example_json}

pipe = pipeline("ner", model=model, tokenizer=tokenizer, aggregation_strategy="simple")


def group_by_entity(raw):
    out = defaultdict(int)
    for ent in raw:
        out[ent["entity_group"]] += 1
    out["total"] = sum(out.values())
    return out


def ner(text):
    raw = pipe(text)
    ner_content = {
        "text": text,
        "entities": [
            {
                "entity": x["entity_group"],
                "word": x["word"],
                "score": x["score"],
                "start": x["start"],
                "end": x["end"],
            }
            for x in raw
        ],
    }
    grouped = group_by_entity(raw)
    df = pd.DataFrame({"Entity": grouped.keys(), "Count": grouped.values()})
    label = EXAMPLE_MAP.get(text, None)
    return (ner_content, grouped, label, df.hist())


interface = gr.Interface(
    ner,
    inputs=gr.Textbox(label="Note text", value=""),
    outputs=[
        gr.HighlightedText(label="NER", combine_adjacent=True),
        gr.JSON(label="Entity Counts"),
        gr.Label(label="Rating"),
        "plot",
    ],
    examples=list(EXAMPLE_MAP.keys()),
)

interface.launch()