I encountered a piece of pytorch code where the trained model is saved with .pk. I often see pytorch models being saved as .pth or .pt. What is .pk format and how is it different from .pth or .pt?
Btw, the following parameters and weights are saved under the .pk file:
save_dict = {
"encoderOpts": encoderOpts,
"classifierOpts": classifierOpts,
"dataOpts": dataOpts,
"encoderState": encoder_state_dict,
"classifierState": classifier_state_dict,
}
Many thanks in advance!