Unable to reduce loss when doing sentiment analysis with Keras

Question

I created a model in Keras that predicts 4 sentiments/emotions based on text input. Size of my data:

label_1 : 100.000
label_2 : 100.000
label_3 : 100.000
label_4