CODEWITHSUNDEEP
rmlflow
Carlo
Carlo

Reputation: 441

In R how to log a model from a file into mlflow

I received a model as a file my_model.rds, and I would like to load it into mlflow. File was created as

saveRDS(my_model, "my_model.rds")

If I try the following

my_model <- readRDS("my_model.rds")
mlflow_log_model(my_model, "my_model")

I get error 404: 'RESOURCE_DOES_NOT_EXIST; No experiment for node found.

My understanding is that the model saved in rds preserve no information regarding the experiments, and this information is required by mlflow to register the model. Is there a way to add this information either when the rds file is created or ex post with some empty experiment?

Does anyone have any advice on how load the model?

Upvotes: 0

Views: 53

Answers (0)

Related Questions