R/onLoad.R

In aifeducation: Artificial Intelligence for Education

transformers <- NULL
datasets <- NULL
tok <- NULL
np <- NULL
codecarbon <- NULL
torch <- NULL
torcheval <- NULL
os <- NULL
keras <- NULL
accelerate <- NULL
safetensors <- NULL
pandas <- NULL
pyarrow <- NULL
calflops <- NULL

.onLoad <- function(libname, pkgname) {
  # use superassignment to update the global reference
  os <<- reticulate::import("os", delay_load = TRUE)
  transformers <<- reticulate::import("transformers", delay_load = TRUE)
  datasets <<- reticulate::import("datasets", delay_load = TRUE)
  tok <<- reticulate::import("tokenizers", delay_load = TRUE)
  np <<- reticulate::import("numpy", delay_load = TRUE)
  torch <<- reticulate::import("torch", delay_load = TRUE)
  torcheval <<- reticulate::import("torcheval", delay_load = TRUE)
  accelerate <<- reticulate::import("accelerate", delay_load = TRUE)
  safetensors <<- reticulate::import("safetensors", delay_load = TRUE)
  pandas <<- reticulate::import("pandas", delay_load = TRUE)
  pyarrow <<- reticulate::import("pyarrow", delay_load = TRUE)
  codecarbon <<- reticulate::import("codecarbon", delay_load = TRUE)
  calflops <<- reticulate::import("calflops", delay_load = TRUE)
}


# Message on load of the package
packageStartupMessage(
  "This is version 1.1.3 of aifeducation. The package now requires the python library 'calflops'.
  Call aifeducation::update_aifeducation() to update your python libraries or use
  aifeducation::install_aifeducation() to install the necessary packages.",
  appendLF = TRUE
)