I tried all the models ranging from CNN(what you have described) to using LSTM's to using Naive Bayes and SVM's but could not decrease the loss function. The loss function is not going below 0.66 and validation accuracy i not improving either.Is there something wrong with my approach?
I tried all the models ranging from CNN(what you have described) to using LSTM's to using Naive Bayes and SVM's but could not decrease the loss function. The loss function is not going below 0.66 and validation accuracy i not improving either.Is there something wrong with my approach?