zhaoqiange

Hello, thank you for the code you provided. After I ran the train. Py file, all the values of loss were NaN, and I only changed the shuffle=False in the...