app.py

import streamlit as st
import gradio as gr
import shap
import numpy as np
import scipy as sp
import torch
import tensorflow as tf
import transformers
from transformers import pipeline
from transformers import RobertaTokenizer, RobertaModel
from transformers import AutoModelForSequenceClassification
from transformers import TFAutoModelForSequenceClassification
from transformers import AutoTokenizer
import matplotlib.pyplot as plt
import sys
import csv

csv.field_size_limit(sys.maxsize)

device = "cuda:0" if torch.cuda.is_available() else "cpu"

tokenizer = AutoTokenizer.from_pretrained("paragon-analytics/ADRv1")  
model = AutoModelForSequenceClassification.from_pretrained("paragon-analytics/ADRv1").to(device)

# build a pipeline object to do predictions
pred = transformers.pipeline("text-classification", model=model, 
                             tokenizer=tokenizer, return_all_scores=True)

explainer = shap.Explainer(pred)

##
classifier = transformers.pipeline("text-classification", model = "cross-encoder/qnli-electra-base")

def med_score(x):
    label = x['label']
    score_1 = x['score']
    return round(score_1,3)

def sym_score(x):
    label2sym= x['label']
    score_1sym = x['score']
    return round(score_1sym,3)

ner_tokenizer = AutoTokenizer.from_pretrained("d4data/biomedical-ner-all")
ner_model = AutoModelForTokenClassification.from_pretrained("d4data/biomedical-ner-all")

ner_pipe = pipeline("ner", model=model, tokenizer=tokenizer, aggregation_strategy="simple") # pass device=0 if using gpu
##

def adr_predict(x):
    encoded_input = tokenizer(x, return_tensors='pt')
    output = model(**encoded_input)
    scores = output[0][0].detach().numpy()
    scores = tf.nn.softmax(scores)
   
    shap_values = explainer([str(x).lower()])
    local_plot = shap.plots.text(shap_values[0], display=False)

    med = med_score(classifier(x+str(", There is a medication."))[0])
    sym = sym_score(classifier(x+str(", There is a symptom."))[0])

    htext = ner_pipe(x)
   
    return {"Severe Reaction": float(scores.numpy()[1]), "Non-severe Reaction": float(scores.numpy()[0])}, local_plot, {"Contains Medication": float(med), "No Medications": float(1-med)} , {"Contains Symptoms": float(sym), "No Symptoms": float(1-sym)},htext


def main(prob1):
    text = str(prob1).lower()
    obj = adr_predict(text)
    return obj[0],obj[1],obj[2],obj[3],obj[4]

title = "Welcome to **ADR Detector** 🪐"
description1 = """This app takes text (up to a few sentences) and predicts to what extent the text describes severe (or non-severe) adverse reaction to medicaitons. Please do NOT use for medical diagnosis."""

with gr.Blocks(title=title) as demo:
    gr.Markdown(f"## {title}")
    gr.Markdown(description1)
    gr.Markdown("""---""")
    prob1 = gr.Textbox(label="Enter Your Text Here:",lines=2, placeholder="Type it here ...")
    submit_btn = gr.Button("Analyze")

    with gr.Row():
        
        with gr.Column(visible=True) as output_col:
            label = gr.Label(label = "Predicted Label")
            local_plot = gr.HTML(label = 'Shap:')
            htext = gr.HighlightedText(label="Diff", combine_adjacent=True, show_legend=True)

        with gr.Column(visible=True) as output_col:
            med = gr.Label(label = "Contains Medication")
            sym = gr.Label(label = "Contains Symptoms")
            
    submit_btn.click(
        main,
        [prob1],
        [label
         ,local_plot, med, sym, htext
        ], api_name="adr"
    )
    
    with gr.Row():
        gr.Markdown("### Click on any of the examples below to see how it works:")
        gr.Examples([["I had severe headache after taking Aspirin."],["I had minor stomachache after taking Acetaminophen."]], 
                    [prob1], [label,local_plot, med, sym,htext], main, cache_examples=True)
    
demo.launch()