Training and validation loss and accuracy in lstm
Question
I am doing text classification with 3 classes, after dealing with over fit model, the image below is my model accuracy and loss results after adding regulization l2, now it means my models is learning?
1 answers
solution1
1 ACCPTED 2020-11-26 19:26:57
You're dealing with an overfit even after you've added the L2 regularization. Your model is not learning much after the third epoch. Some solutions you should consider:
- Increasing the value of your L2 regularization
- Decreasing the number of weights in your model, to support your dataset size
- Increasing your dropout regularization amount
- Transfer learning from a different classification task (for example, you can pre-train your model to classify news articles in your language to different categories)
- Increase the size of your dataset with data augmentations
Good Luck!
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.
Related Question
Plot training and validation accuracy and loss
LSTM model training accuracy and loss not changing
Interpreting training loss/accuracy vs validation loss/accuracy
Validation loss is lower than training loss training LSTM
training and validation accuracy and loss for text classification
Validation loss and accuracy not changing from training
Training and validation loss decrease but accuracy fluctuates?
Low training loss with high validation loss and low validation accuracy
Good accuracy and loss on training vs bad accuracy on validation
Accuracy of 1.0 while Training Loss and Validation Loss still decreasing
Related Tags