Cannot display SHAP text visualization in Streamlit

user_10604508 0
Viewed 456

I am trying to build a dashboard of my NLP project. So I am using a BERT model for the predictions, the SHAP package for the visualization and Streamlit for creating dashboard:

tokenizer = AutoTokenizer.from_pretrained(model_name_cla)
model = AutoModelForSequenceClassification.from_pretrained(model_name_cla)
labels = ['1- Tarife','2- Dateneingabe','3- Bestätigungsmail','4- Kundenbetreuung','5- Aufwand vom Vergleich bis Abschluss',
          '6- After-sales Wechselprozess','7 - Werbung/VX Kommunikation','8 - Sonstiges','9 - Nicht auswertbar']

def f(x):
    tv = torch.tensor([tokenizer.encode(v, padding='max_length', max_length=128, truncation=True) for v in x])
    attention_mask = (tv!=0).type(torch.int64)
    outputs = model(tv,attention_mask=attention_mask)[0].detach().cpu().numpy()
    scores = (np.exp(outputs).T / np.exp(outputs).sum(-1)).T
    val = sp.special.logit(scores)
    return val

text = ['This is just a test']

# build an explainer using a token masker
explainer = shap.Explainer(f, tokenizer, output_names=labels)

shap_values = explainer(text, fixed_context=1)

shap.plots.text(shap_values)

The code works fine in my jupyter notebook but when I am trying to execute this as .py file with streamlit nothing happens. It neither displays anything nor throwing an error. My console just returns this while execution:

<IPython.core.display.HTML object> >

How can I display my graph in streamlit?

shap streamlit python
2 Answers

This can be visualized with Streamlit Components and latest SHAP v0.36+ (which define a new getjs method), to plot JS SHAP plots

(some plots like summary_plot are actually Matplotlib and can be plotted with st.pyplot)

import streamlit as st
import streamlit.components.v1 as components

def st_shap(plot, height=None):
    shap_html = f"<head>{shap.getjs()}</head><body>{plot.html()}</body>"
    components.html(shap_html, height=height)

st_shap(shap.plots.text(shap_values),400)

Find more detailed discussion on Visualizing Shap in streamlit here - Display SHAP diagrams with Streamlit

user_140455370

You can capture the output of the shap plot and render it with components.html

import streamlit.components.v1 as components
from IPython.core.interactiveshell import InteractiveShell
from IPython.utils import capture

def st_plot_text_shap(shap_val, height=None)
    InteractiveShell().instance()
    with capture.capture_output() as cap: 
        shap.plots.text(shap_val)
    components.html(cap.outputs[1].data['text/html'], height=height scrolling=True)
user_35008170
Related