As mentioned before, google colab takes quite a bit of time to train the model. Is it acceptable if I train it locally and submit that history.pkl?
The code will be identical to the .ipynb with the diference being where the source files are located.