Question

我是Tensorflow的新手。我正在编写代码以使用tensorflow在python中执行简单的多层Perceptron代码。在python中执行代码时，出现“ InvalidArgumentError”。请在下面找到我在python中编写的代码以及该代码生成的错误。

我还使用UBUNTU 19.04 OS，Python 3.7.4，TensorFlow版本1.14（GPU版本）和Conda版本4.7.11

预先感谢

这是我编写的代码。我正在使用numpy随机生成数据。

import tensorflow as tf
import numpy as np

np.random.seed(0)

# setting my data for experiment
x1 = np.random.uniform(-100,100, 10000)
x2 = np.random.uniform(-100,100, 10000)
x3 = np.random.uniform(-100,100, 10000)
x_data = np.append(x1.reshape((10000,1)),x2.reshape((10000,1)),axis = 1)
x_data = np.append(x_data,x3.reshape((10000,1)),axis = 1)
y_data = (8*x1 - 3*x2 + x3 + np.random.uniform(-5,5,10000)).reshape((10000,1))
del x1,x2,x3

# Tensorflow code

X = tf.placeholder(dtype = tf.float64, shape = (10, 3), name = 'X')
y = tf.placeholder(dtype = tf.float64, shape = (10,1), name = 'y')

W1 = tf.Variable(np.random.uniform(-0.5,0.5, size = (3,5)), name = 'W1')
b1 = tf.Variable(np.random.uniform(-0.5,0.5, size = (1,5)), name = 'b1')

h1 = tf.add(tf.matmul(X,W1) , b1)

W2 = tf.Variable(np.random.uniform(-0.5,0.5, size = (5,1)), name = 'W2')
b2 = tf.Variable(np.random.uniform(-0.5,0.5, size = (1,1)), name = 'b2')

predict = tf.add(tf.matmul(h1,W2) , b2)

loss = tf.reduce_sum(tf.square(y - predict))

optimizer = tf.train.GradientDescentOptimizer(0.1).minimize(loss)

with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())

    total = 10000
    k = 10
    for i in range(total//k):
        feed_dict = {X : x_data[i*k:(i+1)*k], y : y_data[i*k: (i+1)*k] }
        sess.run(optimizer, feed_dict)
        print(sess.run(loss))

Traceback (most recent call last):

  File "<ipython-input-1-859d9ba5b0e9>", line 1, in <module>
    runfile('/home/devesh/Project Work/Code/Tensorflow/MLP_1.py', wdir='/home/devesh/Project Work/Code/Tensorflow')

  File "/home/devesh/anaconda3/lib/python3.7/site-packages/spyder_kernels/customize/spydercustomize.py", line 827, in runfile
    execfile(filename, namespace)

  File "/home/devesh/anaconda3/lib/python3.7/site-packages/spyder_kernels/customize/spydercustomize.py", line 110, in execfile
    exec(compile(f.read(), filename, 'exec'), namespace)

  File "/home/devesh/Project Work/Code/Tensorflow/MLP_1.py", line 42, in <module>
    print(sess.run(loss))

  File "/home/devesh/anaconda3/lib/python3.7/site-packages/tensorflow/python/client/session.py", line 950, in run
    run_metadata_ptr)

  File "/home/devesh/anaconda3/lib/python3.7/site-packages/tensorflow/python/client/session.py", line 1173, in _run
    feed_dict_tensor, options, run_metadata)

  File "/home/devesh/anaconda3/lib/python3.7/site-packages/tensorflow/python/client/session.py", line 1350, in _do_run
    run_metadata)

  File "/home/devesh/anaconda3/lib/python3.7/site-packages/tensorflow/python/client/session.py", line 1370, in _do_call
    raise type(e)(node_def, op, message)

InvalidArgumentError: 2 root error(s) found.
  (0) Invalid argument: You must feed a value for placeholder tensor 'y' with dtype double and shape [10,1]
     [[node y (defined at /home/devesh/Project Work/Code/Tensorflow/MLP_1.py:18) ]]
     [[Sum/_5]]
  (1) Invalid argument: You must feed a value for placeholder tensor 'y' with dtype double and shape [10,1]
     [[node y (defined at /home/devesh/Project Work/Code/Tensorflow/MLP_1.py:18) ]]
0 successful operations.
0 derived errors ignored.

执行代码时获取“ InvalidArgumentError”

0 个答案: