Here I implemented a Neural Network model visualizer, using Flask and Streamlit. I trained a machine learning model using tensorflow on the mnist dataset of hand digit recognition on 20 epochs. The I saved the model with .h5 format. The python file will take care of the server and prediction.
All the source code is provided here. The notebook will train the model and save model.h5 (which is also added). The ml_server will start the server using flask and app.py will customise the page with streamlit and plot the result.
- open cmd
- cd over the
NN_visualizer - to install depencies
pip install -r requirements.txt - run the command
python3 ml_server.py - it will start the flask server in
http://127.0.0.1:5000/ - run the command
streamlit run app.py - check the
http://localhost:8501, where the streamlit is running