[x] input quanteda dfms for x. This should be implemented in S3 as a generic and a .dfm method (as textmodel_svm() does)
[x] missing data for y should be handled exactly as textmodel_svm() does
[x] the output should be a list of class textmodel_slstm. To pass through the object that keras will need to predict, this can be wrapped inside the object in the way that textmodel_svm() does. Once again, see the svm function for guidance.
x. This should be implemented in S3 as a generic and a .dfm method (as textmodel_svm() does)yshould be handled exactly as textmodel_svm() doestextmodel_slstm. To pass through the object that keras will need to predict, this can be wrapped inside the object in the way that textmodel_svm() does. Once again, see the svm function for guidance.