Version, Share, Deploy, and Monitor Models • vetiver

Vetiver, the oil of tranquility, is used as a stabilizing ingredient in perfumery to preserve more volatile fragrances.

The goal of vetiver is to provide fluent tooling to version, share, deploy, and monitor a trained model. Functions handle both recording and checking the model’s input data prototype, and predicting from a remote API endpoint. The vetiver package is extensible, with generics that can support many kinds of models, and available for both R and Python. To learn more about vetiver, see:

the documentation at https://vetiver.rstudio.com/
the Python package at https://rstudio.github.io/vetiver-python/

You can use vetiver with:

a tidymodels workflow (including stacks)
caret
mlr3
XGBoost
ranger
lm() and glm()
GAMS fit with mgcv
keras
the luz API for torch

Installation

You can install the released version of vetiver from CRAN with:

install.packages("vetiver")

And the development version from GitHub with:

# install.packages("pak")
pak::pak("rstudio/vetiver-r")

Example

A vetiver_model() object collects the information needed to store, version, and deploy a trained model.

library(parsnip)
library(workflows)
data(Sacramento, package = "modeldata")

rf_spec <- rand_forest(mode = "regression")
rf_form <- price ~ type + sqft + beds + baths

rf_fit <- 
    workflow(rf_form, rf_spec) %>%
    fit(Sacramento)

library(vetiver)
v <- vetiver_model(rf_fit, "sacramento_rf")
v
#> 
#> ── sacramento_rf ─ <bundled_workflow> model for deployment 
#> A ranger regression modeling workflow using 4 features

You can version and share your vetiver_model() by choosing a pins “board” for it, including a local folder, Posit Connect, Amazon S3, and more.

library(pins)
model_board <- board_temp()
model_board %>% vetiver_pin_write(v)

You can deploy your pinned vetiver_model() via a Plumber API, which can be hosted in a variety of ways.

library(plumber)
pr() %>%
  vetiver_api(v) %>%
  pr_run(port = 8088)

If the deployed model endpoint is running via one R process (either remotely on a server or locally, perhaps via a background job in the RStudio IDE), you can make predictions with that deployed model and new data in another, separate R process. First, create a model endpoint:

library(vetiver)
endpoint <- vetiver_endpoint("http://127.0.0.1:8088/predict")
endpoint
#> 
#> ── A model API endpoint for prediction: 
#> http://127.0.0.1:8088/predict

Such a model API endpoint deployed with vetiver will return predictions for appropriate new data.

library(tidyverse)
new_sac <- Sacramento %>% 
    slice_sample(n = 20) %>% 
    select(type, sqft, beds, baths)

predict(endpoint, new_sac)
#> # A tibble: 20 x 1
#>      .pred
#>      <dbl>
#>  1 165042.
#>  2 212461.
#>  3 119008.
#>  4 201752.
#>  5 223096.
#>  6 115696.
#>  7 191262.
#>  8 211706.
#>  9 259336.
#> 10 206826.
#> 11 234952.
#> 12 221993.
#> 13 204983.
#> 14 548052.
#> 15 151186.
#> 16 299365.
#> 17 213439.
#> 18 287993.
#> 19 272017.
#> 20 226629.

Contributing

This project is released with a Contributor Code of Conduct. By contributing to this project, you agree to abide by its terms.

For questions and discussions about modeling, machine learning, and MLOps please post on RStudio Community.
If you think you have encountered a bug, please submit an issue.
Either way, learn how to create and share a reprex (a minimal, reproducible example), to clearly communicate about your code.