Model-Visual 1.1.1

A deep learning model visualization suite based on Mermaid

Model Visual

⭐ Star us on GitHub — it motivates us a lot!

Model Visual is a lightweight deep learning model visualization suite based on Mermaid.This kit will convert model schemas into Mermaid code and output code snippets or complete web pages. Users can easily adjust the color and border of the chart.

Only support tensorflow now. More frameworks will comming soon!

Contents

• Model Visual
  • Contents
  • Demo
  • Tutorial
    • Isnstallation
    • Create object
    • Run and get result
  • APIs

Demo

This is the suite's demo. Suite will generate a web page that include a chart.

Tutorial

Isnstallation

First,install the suite by using pip.

pip install model-visual

Create object

Then,creat an object and set some parameter.

Test is the name of an object, you can take any name you want. And model is your keras model's name.(Notice: the model has to be compiled)

from model_visual import ModelVisual

test = ModelVisual(model)

Run and get result

test.run()
test.save_web_page()

Beside save_web_page(),you can also use return_js_code() or return_web_page().

save_web_page() will create a html file, return_js_code() and save_web_page() will return you sourse code.

APIs

set_name(): # set the html file name

set_path(): # set the html path name

set_chart_fill_color(): # set the cart fill color

set_chart_stroke_color(): # set the chart strole color

set_chart_stroke_width(): # set the chart stroke width

set_model(): # set the keras model