Reuse Saved Model (Maxent) to predict new input in R

Question

I am trying to recreate a learning from multiple sources and trying to build a model with maxent in R with my own data. I was able to train and test and validate the result from a data set.

I have saved the trained model and am trying to predict with new input by loading the new model.

Below is my training model code

#Load file
cat = read.csv("Trainingdatafilepath")

#Create Matrix

 matrix = create_matrix(cat[, 1], language = "english", removeStopwords = TRUE, 
                   removeNumbers = TRUE, stemWords = FALSE, tm::weightTfIdf)
# Create Container

container = create_container(matrix, as.numeric(as.factor(cat[, 2])), trainSize = 1:150, 
                         testSize = 151:300, virgin = FALSE)   #removeSparseTerms

models = train_models(container, algorithms = c("MAXENT"))

#Save Model

Save(models, file="mymodel.rda")

I reload the model using following code:

#Load file
cat = read.csv("Newinputfilepath")

#Create Matrix

matrix = create_matrix(cat[, 1], language = "english", removeStopwords = TRUE, removePunctuation = TRUE, stripWhitespace = TRUE, 
                   removeNumbers = TRUE, stemWords = FALSE, tm::weightTfIdf)


# Create Container - New Input

container = create_container(matrix, as.factor(cat[, 1]), testSize = 1:70, virgin = FALSE)  #removeSparseTerms


load("mymodel.rda")

#predict

results = classify_models(container, models)    

Am I approaching this right? The results when validated are not to the expected levels, but that could be because of the input too. The recall score is less than 20%.