zzx528 opened this issue 3 years ago · 0 comments
How to use the policy given in the article to train our model? What should the training process be like? The test accuracy of the model trained according to my own thinking is very low，please answer