edhenry / chexnet

Implementation and fullstack pipeline for CheXNet classifier
MIT License
13 stars 12 forks source link

[BUG/Prediction] No Prediction and heatmap generation after running jupyter notebook provided. #31

Open Aliktk opened 3 years ago

Aliktk commented 3 years ago

Describe the bug When I ran the docker file and everything is set up and open as mentioned in Video as well as in the Readme file. The Jupyter is running and I upload an image of an example + new image for testing the model then the error came when I getting the AUROC's collecting results function.

To Reproduce Steps to reproduce the behavior:

  1. Go to 'Run Docker file fulfill required steps'
  2. Click on 'upload new image example as well new image for testing'
  3. Scroll down to 'run AUROC cell'
  4. See an error

Expected behavior What does it mean? Does this code not getting the prediction from the cheXNet client or TensorFlow serving or Kafka? what if I uploaded a new image rather than an example image? Screenshots Here is the screenshot. Screeasnshot from 2021-06-14 15-41-41

Desktop (please complete the following information):

Please help me here I am near to show for my suoervisor. @edhenry @kholohan