Skip to content

Train and do transfer learning with various yolov5 models and find out what parts of an image made the most contribution to decision making of the model, using explainable AI called EigenCam

License

Notifications You must be signed in to change notification settings

AzizjonKasimov/Yolov5_with_EigenCam

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Yolov5_with_EigenCam

Train and do transfer learning with various yolov5 models and find out what parts of an image made the most contribution to decision making of the model, using explainable AI called EigenCam.

Follow the Colab Notebook code to train the models while also using ClearML for logging and comparing how models performed with different parameters. You can download dataset from your RoboFlow using the code generated from RoboFlow dataset and pasting it into the notebook. When importing EigenCam, first unzip the folder pytorch_grad_cam and upload it to same location where your notebook is, so that it can be imported and used. Also you can specify to freeze the layers in case you have pretrained model and you want to do transfer learning. And using clearML you can do hyperparameter optimization and upload your dataset to ClearML

About

Train and do transfer learning with various yolov5 models and find out what parts of an image made the most contribution to decision making of the model, using explainable AI called EigenCam

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published