Model issue, slop always higher than 0.0005

[Clare_Chung]

Dear mentor:

I tried with different model architecture(ie. only LSTM, or Conv1D, MaxPooling1D,GlobalMaxPooling1D…), I only manage to get bigger slope of your validation loss curve. But to pass the assignment i need to have a slop lower than 0.0005.

Kindly let me know what I can do.
Thanks!

[m.abidat]

Hello @Clare_Chung, thanks for reaching out.

Consider adding a dropout layer to reduce overfitting and your problem should be solved.

PS: Kindly remove the first image.

[Clare_Chung]

Thanks, it worked, also I used a very simple structure this time. ありがとう

[m.abidat]

いいえいいえ がんばってね～@Clare_Chung さん.