As long as you do not specify in the command line arguments --gpus # then you will be training on CPU. I recommend you try using the train_finetune.py to see results without having to wait too long. It functions just the same as train.py except you do not need to specify the model_name in the command line.
As long as you do not specify in the command line arguments
--gpus #
then you will be training on CPU. I recommend you try using thetrain_finetune.py
to see results without having to wait too long. It functions just the same astrain.py
except you do not need to specify themodel_name
in the command line.