|
@@ -33,7 +33,7 @@ W = tf.Variable(rng.randn(), name="weight")
|
|
|
b = tf.Variable(rng.randn(), name="bias")
|
|
|
|
|
|
# Construct a linear model
|
|
|
-pred = tf.add(tf.mul(X, W), b)
|
|
|
+pred = tf.add(tf.multiply(X, W), b)
|
|
|
|
|
|
# Mean squared error
|
|
|
cost = tf.reduce_sum(tf.pow(pred-Y, 2))/(2*n_samples)
|