DEV Community 👩‍💻👨‍💻

Cover image for FruitsClassifier with Pytorch Lightning 🍉
Peter Bao
Peter Bao

Posted on

FruitsClassifier with Pytorch Lightning 🍉

On this project, I convert my FruitsClassifier from Pytorch to Pytorch Lightning. I employ a callback module, a checkpoint mode to store and save the model state_dict. I conduct resume training, plot loss and accuracy from previous trains. I check the accuracy for each class and make predictions on images from outside the dataset.

 

✍️ Documentation:

- Convert my Pytorch model to Pytorch Lightning

- FruitsClassifier on Fruits 360

- Full Training on Kaggle | GoogleNet | 94% Acc

- GitHub Page

 

💥 Training Epochs

training

 

🍇 Visualize Loss and Accuracy

lossacc

 

🍌 Visualize model performance on test dataset

visualize_model

 

🍎 Accuracy of each class

accofeach

 

🍍 Image Predictions from outside the Dataset

prediction

 

🍑 Accuracy of each class: green(correct predictions), red(incorrect predictions)

accofeach2

Top comments (0)

🌚 Browsing with dark mode makes you a better developer.

It's a scientific fact.