Simpletransformers: Save and Load a trained model

[yon606]

Good evening,

Is there a way to save and load a model?
I use the following code to save the model:

import os
import tarfile

def save_model(model_path='/content/drive/My Drive/Erklagg/checkpoint-4860-epoch-12',file_name='checkpoint-4860-epoch-12'):
  files = [files for root, dirs, files in os.walk(model_path)][0]
  with tarfile.open(file_name+ '.tar.gz', 'w:gz') as f:
    for file in files:
      f.add(f'{model_path}/{file}')

save_model('outputs','checkpoint-4860-epoch-12')

that works so far but when I try to load the model with:

import os
import tarfile

"def unpack_model(model_name='checkpoint-4860-epoch-12'): 
  tar = tarfile.open(f"{model_name}.tar.gz", "r:gz")
  tar.extractall()
  tar.close()

unpack_model('checkpoint-4860-epoch-12')

and then try to evaluate the model I get the error:
NameError: name 'model' is not defined

Is there any better way to save and load a model?

[AbinayaM02]

@yon606 : The library automatically saves the check points and the best model files if you specify the path. There is a parameter called 'args' for every model which can take a dictionary of several parameters where you can specify output directory path for checkpoints and the best model to be saved.
Here is the explanation provided : https://github.com/ThilinaRajapakse/simpletransformers#args-explained

To load the model, simply specify the path in the model call. Here is the explanation : https://github.com/ThilinaRajapakse/simpletransformers#loading-saved-models

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.