所以我正在尝试在tensorflow中实现DQN算法,并且我已经定义了如下所示的损失函数,但是每当我使用ADAM优化器执行权重更新时,在2-3次更新之后,我所有的变量都变成了nan。(tf.cast(actions_, tf.int32), shape=(Mini_batch,1)) #Actions which was taken by the online networkz=tf.reshape(tf.range(tf.shape(self.dense_output)[0]), sh
DT_INT32, shape=[1000], _device="/job:localhost/replica:0/task:0/cpu:0"]()]]import tensorflow as tfimport numpy as np#define x and y
x = tf.placeholder(shape=[1000],dtype=tf.int32,name
当试图在FamilyID中检索具有腐蚀因子值的因素时,我得到了行的2-3倍。例如,FamilyID =10216有27975个成分,但我的查询比55k+行多。我正努力找出加入的办法。FROM soladbserver..tblFamily tf ON tfd.FamilyID= tf.FamilyID ON