Skip to content

benchmarks of different neural networks using MNIST dataset

Notifications You must be signed in to change notification settings

mightbesimon-archive/mnist-benchmark

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 

Repository files navigation

MNIST benchmark

benchmarks of different machine learning libraries using the MNIST dataset

  • (tensorflow) DNN : 97.81%
  • (tensorflow) CNN : coming up
  • PyTorch : coming up

Prerequisites

tensorflow:

$ pip3 install tensorflow

note: has to be Python3.7.6 or older, tensorflow is not updated for Python3.8 yet

Use

$ python3 benchmark.py

predictions.py

played around with plotting

  • incorrect classifications
    • network guess (red) and certainty (bar graph)
    • actual classification (blue)
  • unconfident predictions (< 65% certain)

Authors

Shoutout to tensorflow.org tutorials
Shoutout to Google for creating tensorflow

Acknowledgments

  • I'm just following tensorflow.org tutorials

About

benchmarks of different neural networks using MNIST dataset

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages