Prediction with trained and exported keras model.

Machine Learning and Modeling
, , ,
1

I have trained a keras model and saved the final model using:

model %>% save_model_tf("model")

Now, say, I want to send this model to my colleagues who will use this model with their own data to make predictions. They are not programmers, so I will send the following code to them so they can only focus on the result:

library(keras)
model = load_model_tf("model")
model %>% predict(your_data)

But the problem I am having is that to run the above code, they have to install anaconda, python etc. And as they will only use R to make prediction on their data probably just once, I do not think if it is necessary for them to download all the dependent software. I am wandering, if there is any workaround to achieve this without having to download many other software? I would specifically want to be able to successfully run model %>% predict(your_data) in R with R packages.

2

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.