Attributeerror: str Object has no attribute 'Train'

Asked

Viewed 126 times

0

Context
I’m replicating an algorithm of neural network of the kind Single Layer Perceptron (simpler neural network model). I used the library Tensorflow and wrote the code based on the documentation tutorial (https://www.tutorialspoint.com/tensorflow/tensorflow_single_layer_perceptron.htm).

But, the following error occurred and I don’t know how to fix.

Attributeerror Traceback (Most recent call last) in 1 for epoch in range(training_epochs): 2 avg_cost = 0. -----> 3 total_batch = int(mnist.train.num_examples/batch_size)

Attributeerror: str Object has no attribute 'Train'

Code:

import tensorflow as tf

from tensorflow import keras

import numpy as np

import matplotlib.pyplot as plt

mnist = ("_dados/train150_mucuri.txt")

#Parâmetros

learning_rate = 0.01 

training_epochs = 25 

batch_size = 100

display_step = 1 

#Dados de entrada do gráfico


x = tf.placeholder("float", [None, 302500]) # dados da velocidade versus tempo 550x550 = 302.500‬


y = tf.placeholder("float", [None, 10]) # 0-9 dígitos de reconhecimento => 10 classes


# Criar o modelo


# Definir os pesos do modelo



W = tf.Variable(tf.zeros([302500,10])) 


b = tf.Variable(tf.zeros([10])) 


# Modelo de construção


activation = tf.nn.softmax(tf.matmul(x, W) + b) # Softmax - função de soma

# Minimiza o erro usando entropia cruzada


cross_entropy = y*tf.log(activation) 

cost = tf.reduce_mean(-tf.reduce_sum(cross_entropy, reduction_indices = 1)) 


optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)


#Configurações de plotagem


avg_set = [] 


epoch_set = [] 


# Inicializando as variáveis

init = tf.global_variables_initializer()

#Iniciar o gráfico


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

#Ciclo de Treinamento


  for epoch in range(training_epochs):
      avg_cost = 0.
      total_batch = int(mnist.train.num_examples/batch_size)
python-3.x artificial-intelligence anaconda tensorflow

1 answer

0

Hello, I was able to correct this error with the following changes in the code

 for epoch in range(training_epochs):
       avg_cost = 0.
       dataSet = tf.train.list_variables(mnist)
       total_batch = int(dataSet / batch_size)

Browser other questions tagged python-3.x artificial-intelligence anaconda tensorflow

You are not signed in. Login or sign up in order to post.