Hello, I would like to ask where is the difference between the training details used in your article and your open source on github. Can you provide the implementation script for the warmingup-stage in the first step? Also see that you train like(m = 3 and n = 7 in our experiments, and repeat this process until the network converges around 40 epochs)in the last step, where is it reflected in your code? (doesn't seem to see freeze with m=3 and m=7 settings in the code)
Hello, I would like to ask where is the difference between the training details used in your article and your open source on github. Can you provide the implementation script for the warmingup-stage in the first step? Also see that you train like(m = 3 and n = 7 in our experiments, and repeat this process until the network converges around 40 epochs)in the last step, where is it reflected in your code? (doesn't seem to see freeze with m=3 and m=7 settings in the code)