关于参数数量减少

wanyao1992 / code_summarization_public

source code for 'Improving automatic source code summarization via deep reinforcement learning'

77 stars 30 forks source link

关于参数数量减少 #2

Open jingranburangyongzhongwen opened 5 years ago

jingranburangyongzhongwen commented 5 years ago

您好，您论文中 We can see that the actor and critic networks share the modules (a)-(b)-(c), reducing the number of learning parameters a lot. 意思是它俩共享参数吗？如果是的话，我在代码中好像没有看到；如果不是的话，请问您这是什么意思？