Why is this kind of situation, the range to gradient rather than the sum to gradient? Loss should be the sum?

CodePudding user response:

The second picture sent you a wrong, is this,
All code is as follows:

The import numpy as np
The import tensorflow.com pat. V1 as tf
Tf. Disable_v2_behavior ()


W=tf. Variable (. [3], dtype=tf float32)
B=tf. Variable ([- 3], dtype=tf. Float32)
# Model input and output
X=tf. Placeholder (tf. Float32)
Linear_model=W * x + b
Y=tf. Placeholder (tf. Float32)

# loss
Loss=tf. Reduce_sum (tf) square (linear_model - y)) # the sum of the squares
# optimizer
Optimizer=tf. Train. GradientDescentOptimizer (0.01)
"Train"=optimizer. Minimize (loss)

# training data
X_train=[1, 2, 3, 4]
Y_train=[1, 2, 3, 4]
# training loop
Init=tf. Global_variables_initializer ()

Sess=tf. The Session ()

# output figure, into the folder, enter the DOS window tensorboard - logdir=board,
Output_graph=True
If output_graph:
Writer=tf. The summary. FileWriter (' board/' sess. Graph)

Sess. Run (init) # reset values to wrong
For I in range (100) :
Sess. Run (" train ", {x: x_train, y: y_train})

CodePudding user response:

CodePudding user response:

Because be reduce_sum, it don't have to wait for all the data are ready to do the sum operation, but rather to a number, you can add it